flexmix/0000755000176200001440000000000013432625350011731 5ustar liggesusersflexmix/inst/0000755000176200001440000000000013432516316012707 5ustar liggesusersflexmix/inst/CITATION0000644000176200001440000000441513425024236014045 0ustar liggesuserscitHeader("To cite package flexmix in publications use:") citEntry(entry="Article", title = "{FlexMix}: A General Framework for Finite Mixture Models and Latent Class Regression in {R}", author = person(given="Friedrich", family="Leisch"), journal = "Journal of Statistical Software", year = "2004", volume = "11", number = "8", pages = "1--18", doi = "10.18637/jss.v011.i08", url = "http://www.jstatsoft.org/v11/i08/", textVersion = paste("Friedrich Leisch. FlexMix: A general framework for finite mixture models", "and latent class regression in R. Journal of Statistical Software, 11(8), 1-18, 2004.", "doi:10.18637/jss.v011.i08") ) citEntry(entry="Article", title = "Fitting Finite Mixtures of Generalized Linear Regressions in {R}", author = personList(person(given="Bettina", family="Gr\\\"un"), person(given="Friedrich", family="Leisch")), journal = "Computational Statistics \\& Data Analysis", year = "2007", volume = "51", number = "11", pages = "5247--5252", doi = "10.1016/j.csda.2006.08.014", textVersion = paste("Bettina Gruen and Friedrich Leisch. Fitting finite mixtures", "of generalized linear regressions in R. Computational Statistics", "& Data Analysis, 51(11), 5247-5252, 2007.", "doi:10.1016/j.csda.2006.08.014") ) citEntry(entry="Article", title = "{FlexMix} Version 2: Finite Mixtures with Concomitant Variables and Varying and Constant Parameters", author = personList(person(given="Bettina", family="Gr\\\"un"), person(given="Friedrich", family="Leisch")), journal = "Journal of Statistical Software", year = "2008", volume = "28", number = "4", pages = "1--35", doi = "10.18637/jss.v028.i04", url = "http://www.jstatsoft.org/v28/i04/", textVersion = paste("Bettina Gruen and Friedrich Leisch. FlexMix Version 2: Finite mixtures with", "concomitant variables and varying and constant parameters", "Journal of Statistical Software, 28(4), 1-35, 2008.", "doi:10.18637/jss.v028.i04") ) flexmix/inst/NEWS.Rd0000644000176200001440000004445613427002713013762 0ustar liggesusers\name{NEWS} \title{News for Package 'flexmix'} \section{Changes in flexmix version 2.3-15}{ \itemize{ \item Modified the internal function \code{groupPosteriors} to be more efficient for a large number of groups. Thanks to Marnix Koops for pointing the problem out. \item Model driver \code{FLXMRlmer()} adapted for \pkg{lme4} (>= 1.1). Thanks to Mark Senior for pointing the problem out. \item Model driver \code{FLXMRmgcv()} adapted for \pkg{mgcv} (>= 1.8-27). \item Data set \code{Catsup} is now loaded from package \pkg{Ecdat} instead of \pkg{mlogit}. } } \section{Changes in flexmix version 2.3-14}{ \itemize{ \item A bug fixed for \code{FLXMRcondlogit()} to ensure that the formula for the strata is also stored and can be used for predicting, etc. new data. Thanks to Peter Fraser-Mackenziefor for pointing the problem out. \item A bug fixed for \code{FLXMRglmfix()} which occurred if several components were simultaneously omitted. Thanks to Moritz Berger for pointing the problem out. \item For mixtures of mixed effects models and with censored data only weighted ML estimation is implemented. This is now checked and an error is given if a different estimation method is requested. \item A generic function \code{sigma()} is available for R >= 3.3.0 and thus different versions of \code{sigma()} need to be called depending on the R version. Thanks also to Stephen Martin for pointing the issue out. \item Components are now generated using functions instead of expressions. \item Functions from the base packages \pkg{stats}, \pkg{graphics} and \pkg{grDevices} are now correctly imported before being used. \item Function \code{FLXMCdist1} implements model drivers for univariate mixtures of different distributions. } } \section{Changes in flexmix version 2.3-13}{ \itemize{ \item A model driver for mixtures of regression models with (adaptive) lasso and elastic net penalizations for the coefficients building on \pkg{glmnet} was provided by Frederic Mortier and Nicolas Picard. \item A bug in a coerce method to class \code{"FLXnested"} fixed. } } \section{Changes in flexmix version 2.3-12}{ \itemize{ \item Construction of model matrices changed to re-use levels of factors while fitting for prediction. Thanks to Robert Poos for pointing the problem out. \item Package mgcv of version at least 1.8-0 is required in order to avoid loading of the package for formula evaluation. } } \section{Changes in flexmix version 2.3-11}{ \itemize{ \item Examples changed to be less time consuming. \item Bug fixed in ungroupPriors() and getPriors() to work with a grouping where only one unique value is contained. Thanks to Christine Koehler for pointing the problem out. \item The \code{logLik()} method for \code{"flexmix"} objects now also has a \code{newdata} argument. \item In the M-step only the parameters of the previously fitted components are passed over to avoid nesting of environments. Thanks to Dominik Ernst for pointing the problem out. \item Function \code{boot()} was fixed to work correctly with grouped data if the groups are kept and if fixed effects are fitted. \item Authors@R used in DESCRIPTION. Deepayan Sarkar listed as contributor due to internal code copied from package lattice used for the plots in flexmix. \item Model driver \code{FLXMRlmer()} adapted for \pkg{lme4} (>= 1.0). } } \section{Changes in flexmix version 2.3-10}{ \itemize{ \item Parallel processing disabled in \code{stepFlexmix()}. } } \section{Changes in flexmix version 2.3-9}{ \itemize{ \item Package dependencies, imports and suggests modified suitably for R >= 2.15.0. \item NEWS file adapted to a NEWS.Rd file. \item Parallel processing is enabled in \code{stepFlexmix()}. See \pkg{flexclust} for details. \item New model drivers \code{FLXMRmultinom()} and \code{FLXMRcondlogit()} are included which allow to fit finite mixtures of multinomial logit and conditional logit models. Identifiability problems might arise for this model class. Details on sufficient identifiability conditions are given in Gruen and Leisch (2008). \item A bug in \code{FLXMRlmm()} was fixed which did not allow to specify restrictions on the variances of the random effects and / or the residuals. Thanks to Gregory Matthews for pointing the problem out. } } \section{Changes in flexmix version 2.3-8}{ \itemize{ \item The fit function in the M-step by default now is called with an argument containing the fitted component. This allows to use the parameter estimates from the previous step for initialization. Fit functions which do not require this now need a \code{...} argument. Thanks to Hannah Frick and Achim Zeileis for requesting this feature. \item Function \code{initFlexmix()} was added which is an alternative to \code{stepFlexmix()} where first several short runs of EM, SEM or CEM are performed followed by a long run of EM. } } \section{Changes in flexmix version 2.3-7}{ \itemize{ \item A bug fixed in \code{predict()} and \code{fitted()} if a concomitant variable model is specified and \code{aggregate = TRUE}. Thanks to Julia Schiffner for pointing this out. \item A bug fixed in \code{FLXMRmgcv()} if observations were removed in the M-step because their a-posterior probabilities were smaller than eps. Thanks to Ghislain Geniaux for pointing the problem out. } } \section{Changes in flexmix version 2.3-6}{ \itemize{ \item Vignettes moved from /inst/doc to /vignettes. \item \code{stepFlexmix()} can now be called with a concomitant variable model \code{FLXPmultinom()} for \code{k} starting with 1 without getting an error. The concomitant variable model is internally replaced by \code{FLXPconstant()}. \item The \code{boot()} method for \code{"flexmix"} objects is extended to mixture models with concomitant variables and mixtures of linear mixed models. \item A bug fixed in the \code{summary()} method for \code{"flexmix"} objects. The column post > 0 did not give the correct results if weights were used for fitting the mixture. \item A bug fixed in the \code{unique()} method for \code{"stepFlexmix"} objects. This only occurred if components were dropped as well as if the EM algorithm did not converge for all repetitions. Thanks to Sebastian Meyer for pointing this out. } } \section{Changes in flexmix version 2.3-5}{ \itemize{ \item A bug fixed in \code{posterior()}. Fixed priors were always used, also if a concomitant variable model was present. \item A method added for \code{prior()} such that if newdata is supplied and the object is of class \code{"flexmix"} the prior class probabilities for each observation are returned. } } \section{Changes in flexmix version 2.3-4}{ \itemize{ \item A generic method for \code{nobs()} is introduced in \pkg{stats4} for \R 2.13.0. \pkg{flexmix} now does not define this generic function and \code{logLik()}, \code{AIC()} and \code{BIC()} methods were modified to better exploit already available methods. Thanks to Prof. Brian D. Ripley for suggesting the modification. } } \section{Changes in flexmix version 2.3-3}{ \itemize{ \item A bug for \code{boot()} fixed for \code{"flexmix"} objects with an unbalanced grouping variable. Thanks to Laszlo Sandor for pointing this out. } } \section{Changes in flexmix version 2.3-2}{ \itemize{ \item A bug for \code{rflexmix()} fixed for \code{"flexmix"} objects with a concomitant variable model. Thanks to Greg Petroski for pointing this out. } } \section{Changes in flexmix version 2.3-1}{ \itemize{ \item Functionality for bootstrapping finite mixture models added. } } \section{Changes in flexmix version 2.2-11}{ \itemize{ \item More generics and methods exported to use the \code{refit()} method when extending \pkg{flexmix} in other packages. } } \section{Changes in flexmix version 2.2-10}{ \itemize{ \item For long formulas \code{FLXMRglmfix()} did not work properly due to the splitting of the formula into several parts by \code{deparse()}. This is fixed by pasting them together again. Thanks to Dustin Tingley for the bug report. } } \section{Changes in flexmix version 2.2-9}{ \itemize{ \item A new model driver \code{FLXMRmgcv()} is added which allows to fit regularized linear models in the components. \item More generics and methods exported to allow extending \pkg{flexmix} in other packages. } } \section{Changes in flexmix version 2.2-8}{ \itemize{ \item The a-posteriori probabilities are now also determined as changed for \code{FLXfit()} for version 2.2-6 for \code{refit()}. \item Bug fixed for FLXfillconcomitant and refit when weights and grouping are present. A check was added that weights are identical within groups. \item Function \code{group()} is now exported. } } \section{Changes in flexmix version 2.2-7}{ \itemize{ \item Bug in the \code{FLXgetModelmatrix()} method for the \code{"FLXMRlmm"} class fixed when determining identical random effects covariates for the grouping. \item A new model driver for finite mixtures of linear mixed effects models with left-censored observations is added. } } \section{Changes in flexmix version 2.2-6}{ \itemize{ \item Determination of the a-posteriori probabilities made numerically more stable for small likelihoods. Thanks to Nicolas Picard for the code patch. \item \code{summary()} for \code{"FLXMRmstep"} objects now returns similar output for \code{which = "concomitant"} as for \pkg{flexmix} version 2.0-1. \item New demo driver \code{FLXMCnorm1()} for univariate Gaussian clustering. \item Non-postive values for the maximum number of iterations for the \code{"FLXcontrol"} object are not valid. A validity check for this is now included. } } \section{Changes in flexmix version 2.2-5}{ \itemize{ \item Model class \code{"FLXMRfix"} introduced which is a subclass of \code{"FLXMR"} and a superclass for \code{"FLXMRglmfix"} which also extends \code{"FLXMRglm"}. \item Model driver \code{FLXMCfactanal()} added which allows to fit finite mixtures of Gaussian distributions where the variance-covariance matrix is estimated using factor analyzers. \item Comparison of formulas now done using \code{identical()}. } } \section{Changes in flexmix version 2.2-4}{ \itemize{ \item Model drivers \code{FLXMRlmer()} and \code{FLXMRlmm()} added for fitting finite mixtures of linear mixed effects models. \item \code{EIC()} added as additional information criterion for assessing model fit. \item Bug fixed in \code{plot()} method for \code{"flexmix"} objects introduced in version 2.2-3. } } \section{Changes in flexmix version 2.2-3}{ \itemize{ \item New model driver \code{FLXMCmvcombi()} which is a combination of Gaussian and binary. \item \code{parameters()} now also has a which argument in order to allow to access the parameters of the concomitant variable model. \item Bug fixed in \code{refit()}. \item \code{nobs()} now returns the number of rows in the data.frame and not the number of individuals (similar as for example by lme). } } \section{Changes in flexmix version 2.2-0}{ \itemize{ \item vignette describing Version 2 added \item isTRUE(\code{all.equal()}) replaced with \code{identical()}. \item Bug fixed for prior in \code{flexmix()}. \item New function \code{relabel()} to sort components (generic is in modeltools). \item New example data generator \code{ExLinear()}. \item Fixed a bug in handling groups (gave an error for empty design matrices). \item Added new model \code{FLXMRrobglm()} for robust estimation of GLMs. } } \section{Changes in flexmix version 2.1-0}{ \itemize{ \item Renamed \code{cluster()} to \code{clusters()} to avoid conflict with \code{cluster()} from package survival \item Bug fixed in internal functions using S4 generics and methods. } } \section{Changes in flexmix version 2.0-2}{ \itemize{ \item \code{refit()} now has a method argument. For method \code{"optim"} the variance-covariance matrix is determined using \code{optim()} to maximize the likelihood initialized in the solution found by the EM algorithm. Method \code{"mstep"} refits the component specific and concomitant models treating the posterior probabilities as given, i.e. performs an M-step of the EM algorithm. } } \section{Changes in flexmix version 2.0-1}{ \itemize{ \item \code{Lapply()} added which allows to apply a function to each component of a finite mixture \item \code{KLdiv()} modified to allow for determination with a discrete and a continuous version of the KL divergence } } \section{Changes in flexmix version 2.0-0}{ \itemize{ \item Model driver for zero-inflated component specific models. \item Latent class analysis for binary multivariate data is now possible to estimate for truncated data where the number of observations with pattern only zeros is missing. \item new argument newdata for \code{cluster()} \item new \code{unique()} method for \code{"stepFlexmix"} objects } } \section{Changes in flexmix version 1.9-0}{ \itemize{ \item New class definitions for component specific models and concomitant variable models. \item \code{fitted()} and \code{predict()} now have an aggregate argument in order to be able to determine the aggregated values over all components. \item The package has now a vignette presenting several applications of finite mixtures of regression models with varying and fixed effects on artificial and real data which can be a accessed using the command vignette("regression-examples"). \item The vignette "flexmix-intro" was adapted to reflect the changes made in the package. \item \code{stepFlexmix()} now returns an object of class \code{"stepFlexmix"} which has a \code{print()} and \code{plot()} method. In addition \code{getModel()} can be used to select an appropriate model. \item \code{flexmix()} now has a weights argument for multiple identical observations. \item New model drivers for latent class analysis with Bernoulli and Poisson distributed multivariate observations. \item Variants of the EM algorithm have been modified to correspond to CEM and SEM. These names can now also be used for specifying the classify slot of the \code{"FLXcontrol"} object. } } \section{Changes in flexmix version 1.8-1}{ \itemize{ \item The package can now fit concomitant variable models. \item New M-step driver for regression models with varying and fixed effects. \item ICL information criterion added. } } \section{Changes in flexmix version 1.1-2}{ \itemize{ \item Fixed a bug that made the log-likelihood infinity for observations where all posteriors are numerically zero. \item Fixed a bug for formulae with dots. \item \code{posterior()} now has a newdata argument. \item New demo driver for model-based clustering of binary data. \item Adapted to changes in \code{summary.glm()} of \R version 2.3.0. } } \section{Changes in flexmix version 1.1-1}{ \itemize{ \item The \code{cluster} argument of \code{flexmix()} may now also be a matrix of posterior probabilities. \item Fixed a bug to make size table work in case of empty clusters. \item Fixed a bug in likelihood computation for grouped observations. \item The artificial NPreg data now also have a binomial response, added example to help(\code{"flexmix"}). } } \section{Changes in flexmix version 1.1-0}{ \itemize{ \item The \code{FLXglm()} driver now has an offset argument. \item New data set seizure as example for a Poisson GLM with an offset. \item \code{fitted()} can be used to extract fitted values from \code{"flexmix"} and \code{"FLXrefit"} objects. \item New accessor methods \code{cluster()} and \code{posterior()}. \item The package now uses lazy loading and has a namespace. } } \section{Changes in flexmix version 1.0-0}{ \itemize{ \item The package has now an introductionary vignette which can be accessed using the command vignette("flexmix-intro"). The vignette has been published in the Journal of Statistical Software, Volume 11, Issue 8 (\url{http://www.jstatsoft.org/v11/i08}), and the paper should be used as citation for \pkg{flexmix}, run \code{citation("flexmix")} in \R 2.0.0 or newer for details. \item Several typos in help pages have been fixed. } } \section{Changes in flexmix version 0.9-1}{ \itemize{ \item Adjusted for \R 2.0.0. \item Fixed a bug in the \code{summary()} and \code{plot()} methods of \code{"flexmix"} objects in case of empty clusters. \item \code{stepFlexmix()} takes two new arguments: \code{compare} allows fo find minimum AIC/BIC solutions in addition to maximum likelihood, \code{verbose} gives some information about progress. \item Use a default of \code{verbose = 0} in \code{FLXcontrol()} (better in combination with \code{stepFlexmix()}). } } \section{Changes in flexmix version 0.9-0}{ \itemize{ \item New \code{summary()} and \code{plot()} methods for \code{"flexmix"} objects. \item \code{"FLXglm"} objects can now be automatically \code{refit()}ted to get table of significance tests for coefficients. \item New function \code{stepFlexmix()} for more automated model search. \item The artificial example data now have functions to create them and a pre-stored data sets, new function \code{plotEll()} to plot 2d Gaussians with confidence ellipses. \item New function \code{KLdiv()} to compute Kullback-Leibler divergence of components. \item The calculation of the degrees of freedom for \code{FLXmclust()} was wrong } } \section{Changes in flexmix version 0.7-1}{ \itemize{ \item Fixed some codoc problems (missing aliases). \item First version released on CRAN: 0.7-0 } } flexmix/inst/doc/0000755000176200001440000000000013432516316013454 5ustar liggesusersflexmix/inst/doc/mymclust.R0000644000176200001440000000205313425024236015451 0ustar liggesusersmymclust <- function (formula = .~., diagonal = TRUE) { retval <- new("FLXMC", weighted = TRUE, formula = formula, dist = "mvnorm", name = "my model-based clustering") retval@defineComponent <- function(para) { logLik <- function(x, y) { mvtnorm::dmvnorm(y, mean = para$center, sigma = para$cov, log = TRUE) } predict <- function(x) { matrix(para$center, nrow = nrow(x), ncol = length(para$center), byrow = TRUE) } new("FLXcomponent", parameters = list(center = para$center, cov = para$cov), df = para$df, logLik = logLik, predict = predict) } retval@fit <- function(x, y, w, ...) { para <- cov.wt(y, wt = w)[c("center", "cov")] df <- (3 * ncol(y) + ncol(y)^2)/2 if (diagonal) { para$cov <- diag(diag(para$cov)) df <- 2 * ncol(y) } retval@defineComponent(c(para, df = df)) } retval } flexmix/inst/doc/mixture-regressions.Rnw0000644000176200001440000022021513425024236020201 0ustar liggesusers% % Copyright (C) 2008 Bettina Gruen and Friedrich Leisch % $Id: mixture-regressions.Rnw $ % \documentclass[nojss]{jss} \usepackage{amsfonts} \title{FlexMix Version 2: Finite Mixtures with Concomitant Variables and Varying and Constant Parameters} \Plaintitle{FlexMix Version 2: Finite Mixtures with Concomitant Variables and Varying and Constant Parameters} \Shorttitle{FlexMix Version 2} \author{Bettina Gr{\"u}n\\ Johannes Kepler Universit{\"at} Linz \And Friedrich Leisch\\ Universit\"at f\"ur Bodenkultur Wien} \Plainauthor{Bettina Gr{\"u}n, Friedrich Leisch} \Address{ Bettina Gr\"un\\ Institut f\"ur Angewandte Statistik / IFAS\\ Johannes Kepler Universit{\"at} Linz\\ Freist\"adter Stra\ss{}e 315\\ 4040 Linz, Austria\\ E-mail: \email{Bettina.Gruen@jku.at}\\ Friedrich Leisch\\ Institut f\"ur Angewandte Statistik und EDV\\ Universit\"at f\"ur Bodenkultur Wien\\ Peter Jordan Stra\ss{}e 82\\ 1190 Wien, Austria\\ E-mail: \email{Friedrich.Leisch@boku.ac.at}\\ URL: \url{http://www.statistik.lmu.de/~leisch/} } \Abstract{ This article is a (slightly) modified version of \cite{mixtures:Gruen+Leisch:2008a}, published in the \emph{Journal of Statistical Software}. \pkg{flexmix} provides infrastructure for flexible fitting of finite mixture models in \proglang{R} using the expectation-maximization (EM) algorithm or one of its variants. The functionality of the package was enhanced. Now concomitant variable models as well as varying and constant parameters for the component specific generalized linear regression models can be fitted. The application of the package is demonstrated on several examples, the implementation described and examples given to illustrate how new drivers for the component specific models and the concomitant variable models can be defined. } \Keywords{\proglang{R}, finite mixture models, generalized linear models, concomitant variables} \Plainkeywords{R, finite mixture models, generalized linear models, concomitant variables} \usepackage{amsmath, listings} \def\argmax{\mathop{\rm arg\,max}} %% \usepackage{Sweave} prevent automatic inclusion \SweaveOpts{width=9, height=4.5, eps=FALSE, keep.source=TRUE} <>= options(width=60, prompt = "R> ", continue = "+ ", useFancyQuotes = FALSE) library("graphics") library("stats") library("flexmix") library("lattice") ltheme <- canonical.theme("postscript", FALSE) lattice.options(default.theme=ltheme) data("NPreg", package = "flexmix") data("dmft", package = "flexmix") source("myConcomitant.R") @ %%\VignetteIndexEntry{FlexMix Version 2: Finite Mixtures with Concomitant Variables and Varying and Constant Parameters} %%\VignetteDepends{flexmix} %%\VignetteKeywords{R, finite mixture models, model based clustering, latent class regression} %%\VignettePackage{flexmix} \begin{document} %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \section{Introduction}\label{sec:introduction} Finite mixture models are a popular technique for modelling unobserved heterogeneity or to approximate general distribution functions in a semi-parametric way. They are used in a lot of different areas such as astronomy, biology, economics, marketing or medicine. An overview on mixture models is given in \cite{mixtures:Everitt+Hand:1981}, \cite{mixtures:Titterington+Smith+Makov:1985}, \cite{mixtures:McLachlan+Basford:1988}, \cite{mixtures:Boehning:1999}, \cite{mixtures:McLachlan+Peel:2000} and \cite{mixtures:Fruehwirth-Schnatter:2006}. Version 1 of \proglang{R} package \pkg{flexmix} was introduced in \cite{mixtures:Leisch:2004}. The main design principles of the package are extensibility and fast prototyping for new types of mixture models. It uses \proglang{S}4 classes and methods \citep{mixtures:Chambers:1998} as implemented in the \proglang{R} package \pkg{methods} and exploits advanced features of \proglang{R} such as lexical scoping \citep{mixtures:Gentleman+Ihaka:2000}. The package implements a framework for maximum likelihood estimation with the expectation-maximization (EM) algorithm \citep{mixtures:Dempster+Laird+Rubin:1977}. The main focus is on finite mixtures of regression models and it allows for multiple independent responses and repeated measurements. The EM algorithm can be controlled through arguments such as the maximum number of iterations or a minimum improvement in the likelihood to continue. Newly introduced features in the current package version are concomitant variable models \citep{mixtures:Dayton+Macready:1988} and varying and constant parameters in the component specific regressions. Varying parameters follow a finite mixture, i.e., several groups exist in the population which have different parameters. Constant parameters are fixed for the whole population. This model is similar to mixed-effects models \citep{mixtures:Pinheiro+Bates:2000}. The main difference is that in this application the distribution of the varying parameters is unknown and has to be estimated. Thus the model is actually closer to the varying-coefficients modelling framework \citep{mixtures:Hastie+Tibshirani:1993}, using convex combinations of discrete points as functional form for the varying coefficients. The extension to constant and varying parameters allows for example to fit varying intercept models as given in \cite{mixtures:Follmann+Lambert:1989} and \cite{mixtures:Aitkin:1999}. These models are frequently applied to account for overdispersion in the data where the components follow either a binomial or Poisson distribution. The model was also extended to include nested varying parameters, i.e.~this allows to have groups of components with the same parameters \citep{mixtures:Gruen+Leisch:2006, mixtures:Gruen:2006}. In Section~\ref{sec:model-spec-estim} the extended model class is presented together with the parameter estimation using the EM algorithm. In Section~\ref{sec:using-new-funct} examples are given to demonstrate how the new functionality can be used. An overview on the implementational details is given in Section~\ref{sec:implementation}. The new model drivers are presented and changes made to improve the flexibility of the software and to enable the implementation of the new features are discussed. Examples for writing new drivers for the component specific models and the concomitant variable models are given in Section~\ref{sec:writing-your-own}. This paper gives a short overview on finite mixtures and the package in order to be self-contained. A more detailed introduction to finite mixtures and the package \pkg{flexmix} can be found in \cite{mixtures:Leisch:2004}. All computations and graphics in this paper have been done with \pkg{flexmix} version \Sexpr{packageDescription("flexmix",fields="Version")} and \proglang{R} version \Sexpr{getRversion()} using Sweave \citep{mixtures:Leisch:2002}. The newest release version of \pkg{flexmix} is always available from the Comprehensive \proglang{R} Archive Network at \url{http://CRAN.R-project.org/package=flexmix}. An up-to-date version of this paper is contained in the package as a vignette, giving full access to the \proglang{R} code behind all examples shown below. See \code{help("vignette")} or \cite{mixtures:Leisch:2003} for details on handling package vignettes. %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \section{Model specification and estimation}\label{sec:model-spec-estim} A general model class of finite mixtures of regression models is considered in the following. The mixture is assumed to consist of $K$ components where each component follows a parametric distribution. Each component has a weight assigned which indicates the a-priori probability for an observation to come from this component and the mixture distribution is given by the weighted sum over the $K$ components. If the weights depend on further variables, these are referred to as concomitant variables. In marketing choice behaviour is often modelled in dependence of marketing mix variables such as price, promotion and display. Under the assumption that groups of respondents with different price, promotion and display elasticities exist mixtures of regressions are fitted to model consumer heterogeneity and segment the market. Socio-demographic variables such as age and gender have often been shown to be related to the different market segments even though they generally do not perform well when used to a-priori segment the market. The relationships between the behavioural and the socio-demographic variables is then modelled through concomitant variable models where the group sizes (i.e.~the weights of the mixture) depend on the socio-demographic variables. The model class is given by \begin{align*} h(y|x, w, \psi) &= \sum_{k = 1}^K \pi_k(w, \alpha) f_k(y|x,\theta_{k})\\ &= \sum_{k = 1}^K \pi_k(w, \alpha) \prod_{d=1}^D f_{kd}(y_d|x_d,\theta_{kd}), \end{align*} where $\psi$ denotes the vector of all parameters for the mixture density $h()$ and is given by $(\alpha, (\theta_k)_{k=1,\ldots,K})$. $y$ denotes the response, $x$ the predictor and $w$ the concomitant variables. $f_k$ is the component specific density function. Multivariate variables $y$ are assumed to be dividable into $D$ subsets where the component densities are independent between the subsets, i.e.~the component density $f_k$ is given by a product over $D$ densities which are defined for the subset variables $y_d$ and $x_d$ for $d=1,\ldots,D$. The component specific parameters are given by $\theta_k = (\theta_{kd})_{d=1,\ldots,D}$. Under the assumption that $N$ observations are available the dimensions of the variables are given by $y = (y_d)_{d=1,\ldots,D} \in \mathbb{R}^{N \times \sum_{d=1}^D k_{yd}}$, $x = (x_d)_{d=1,\ldots,D} \in \mathbb{R}^{N \times \sum_{d=1}^D k_{xd}}$ for all $d = 1,\ldots,D$ and $w \in \mathbb{R}^{N \times k_w}$. In this notation $k_{yd}$ denotes the dimension of the $d^{\textrm{th}}$ response, $k_{xd}$ the dimension of the $d^{\textrm{th}}$ predictors and $k_w$ the dimension of the concomitant variables. For mixtures of GLMs each of the $d$ responses will in general be univariate, i.e.~multivariate responses will be conditionally independent given the segment memberships. For the component weights $\pi_k$ it holds $\forall w$ that \begin{equation}\label{eq:prior} \sum_{k=1}^K \pi_k(w,\alpha) = 1 \quad \textrm{and} \quad \pi_k(w, \alpha) > 0, \, \forall k, \end{equation} where $\alpha$ are the parameters of the concomitant variable model. For the moment focus is given to finite mixtures where the component specific densities are from the same parametric family, i.e.~$f_{kd} \equiv f_d$ for notational simplicity. If $f_d$ is from the exponential family of distributions and for each component a generalized linear model is fitted \citep[GLMs;][]{mixtures:McCullagh+Nelder:1989} these models are also called GLIMMIX models \citep{mixtures:Wedel+DeSarbo:1995}. In this case the component specific parameters are given by $\theta_{kd} = (\beta'_{kd}, \phi_{kd})$ where $\beta_{kd}$ are the regression coefficients and $\phi_{kd}$ is the dispersion parameter. The component specific parameters $\theta_{kd}$ are either restricted to be equal over all components, to vary between groups of components or to vary between all components. The varying between groups is referred to as varying parameters with one level of nesting. A disjoint partition $K_c$, $c = 1,\ldots,C$ of the set $\tilde{K} := \{1\ldots,K\}$ is defined for the regression coefficients. $C$ is the number of groups of the regression coefficients at the nesting level. The regression coefficients are accordingly split into three groups: \begin{align*} \beta_{kd} &= (\beta'_{1d}, \beta'_{2,c(k)d}, \beta'_{3,kd})', \end{align*} where $c(k) = \{c = 1,\ldots, C: k \in K_c\}$. Similar a disjoint partition $K_v$, $v = 1,\ldots,V$, of $\tilde{K}$ can be defined for the dispersion parameters if nested varying parameters are present. $V$ denotes the number of groups of the dispersion parameters at the nesting level. This gives: \begin{align*} \phi_{kd} &= \left\{\begin{array}{ll} \phi_{d} & \textrm{for constant parameters}\\ \phi_{kd} & \textrm{for varying parameters}\\ \phi_{v(k)d} & \textrm{for nested varying parameters} \end{array}\right. \end{align*} where $v(k) = \{v = 1,\ldots,V: k \in K_v\}$. The nesting structure of the component specific parameters is also described in \cite{mixtures:Gruen+Leisch:2006}. Different concomitant variable models are possible to determine the component weights \citep{mixtures:Dayton+Macready:1988}. The mapping function only has to fulfill condition \eqref{eq:prior}. In the following a multinomial logit model is assumed for the $\pi_k$ given by \begin{equation*} \pi_k(w,\alpha) = \frac{e^{w'\alpha_k}}{\sum_{u = 1}^K e^{w'\alpha_u}} \quad \forall k, \end{equation*} with $\alpha = (\alpha'_k)'_{k=1,\ldots,K}$ and $\alpha_1 \equiv 0$. %%------------------------------------------------------------------------- \subsection{Parameter estimation}\label{sec:estimation} The EM algorithm \citep{mixtures:Dempster+Laird+Rubin:1977} is the most common method for maximum likelihood estimation of finite mixture models where the number of components $K$ is fixed. The EM algorithm applies a missing data augmentation scheme. It is assumed that a latent variable $z_n \in \{0,1\}^K$ exists for each observation $n$ which indicates the component membership, i.e.~$z_{nk}$ equals 1 if observation $n$ comes from component $k$ and 0 otherwise. Furthermore it holds that $\sum_{k=1}^K z_{nk}=1$ for all $n$. In the EM algorithm these unobserved component memberships $z_{nk}$ of the observations are treated as missing values and the data is augmented by estimates of the component membership, i.e.~the estimated a-posteriori probabilities $\hat{p}_{nk}$. For a sample of $N$ observations $\{(y_1, x_1, w_1), \ldots, (y_N, x_N, w_N)\}$ the EM algorithm is given by: \begin{description} \item[E-step:] Given the current parameter estimates $\psi^{(i)}$ in the $i$-th iteration, replace the missing data $z_{nk}$ by the estimated a-posteriori probabilities \begin{align*} \hat{p}_{nk} & = \frac{\displaystyle \pi_k(w_n, \alpha^{(i)}) f(y_n| x_n, \theta_k^{(i)}) }{\displaystyle \sum_{u = 1}^K \pi_u(w_n, \alpha^{(i)}) f(y_n |x_n, \theta_u^{(i)}) }. \end{align*} \item[M-step:] Given the estimates for the a-posteriori probabilities $\hat{p}_{nk}$ (which are functions of $\psi^{(i)}$), obtain new estimates $\psi^{(i+1)}$ of the parameters by maximizing \begin{align*} Q(\psi^{(i+1)}|\psi^{(i)}) &= Q_1(\theta^{(i+1)} | \psi^{(i)}) + Q_2(\alpha^{(i+1)} | \psi^{(i)}), \end{align*} where \begin{align*} Q_1(\theta^{(i+1)} | \psi^{(i)}) &= \sum_{n = 1}^N \sum_{k = 1}^K \hat{p}_{nk} \log(f(y_n | x_n, \theta_k^{(i+1)})) \end{align*} and \begin{align*} Q_2(\alpha^{(i+1)}| \psi^{(i)}) &= \sum_{n = 1}^N \sum_{k = 1}^K \hat{p}_{nk} \log(\pi_k(w_n, \alpha^{(i+1)})). \end{align*} $Q_1$ and $Q_2$ can be maximized separately. The maximization of $Q_1$ gives new estimates $\theta^{(i+1)}$ and the maximization of $Q_2$ gives $\alpha^{(i+1)}$. $Q_1$ is maximized separately for each $d=1,\ldots,D$ using weighted ML estimation of GLMs and $Q_2$ using weighted ML estimation of multinomial logit models. \end{description} Different variants of the EM algorithm exist such as the stochastic EM \citep[SEM;][]{mixtures:Diebolt+Ip:1996} or the classification EM \citep[CEM;][]{mixtures:Celeux+Govaert:1992}. These two variants are also implemented in package \pkg{flexmix}. For both variants an additional step is made between the expectation and maximization steps. This step uses the estimated a-posteriori probabilities and assigns each observation to only one component, i.e.~classifies it into one component. For SEM this assignment is determined in a stochastic way while it is a deterministic assignment for CEM. For the SEM algorithm the additional step is given by: \begin{description} \item[S-step:] Given the a-posteriori probabilities draw \begin{align*} \hat{z}_n &\sim \textrm{Mult}((\hat{p}_{nk})_{k=1,\ldots,K}, 1) \end{align*} where $\textrm{Mult}(\theta, T)$ denotes the multinomial distribution with success probabilities $\theta$ and number of trials $T$. \end{description} Afterwards, the $\hat{z}_{nk}$ are used instead of the $\hat{p}_{nk}$ in the M-step. For the CEM the additional step is given by: \begin{description} \item[C-step:] Given the a-posteriori probabilities define \begin{align*} \hat{z}_{nk} &= \left\{\begin{array}{ll} 1&\textrm{if } k = \min\{ l : \hat{p}_{nl} \geq \hat{p}_{nk}\, \forall k=1,\ldots,K\}\\ 0&\textrm{otherwise}. \end{array}\right. \end{align*} \end{description} Please note that in this step the observation is assigned to the component with the smallest index if the same maximum a-posteriori probability is observed for several components. Both of these variants have been proposed to improve the performance of the EM algorithm, because the ordinary EM algorithm tends to converge rather slowly and only to a local optimum. The convergence behavior can be expected to be better for the CEM than ordinary EM algorithm, while SEM can escape convergence to a local optimum. However, the CEM algorithm does not give ML estimates because it maximizes the complete likelihood. For SEM good approximations of the ML estimator are obtained if the parameters where the maximum likelihood was encountered are used as estimates. Another possibility for determining parameter estimates from the SEM algorithm could be the mean after discarding a suitable number of burn-ins. An implementational advantage of both variants is that no weighted maximization is necessary in the M-step. It has been shown that the values of the likelihood are monotonically increased during the EM algorithm. On the one hand this ensures the convergence of the EM algorithm if the likelihood is bounded, but on the other hand only the detection of a local maximum can be guaranteed. Therefore, it is recommended to repeat the EM algorithm with different initializations and choose as final solution the one with the maximum likelihood. Different initialization strategies for the EM algorithm have been proposed, as its convergence to the optimal solution depends on the initialization \citep{mixtures:Biernacki+Celeux+Govaert:2003,mixtures:Karlis+Xekalaki:2003}. Proposed strategies are for example to first make several runs of the SEM or CEM algorithm with different random initializations and then start the EM at the best solution encountered. The component specific parameter estimates can be determined separately for each $d=1,\ldots,D$. For simplicity of presentation the following description assumes $D=1$. If all parameter estimates vary between the component distributions they can be determined separately for each component in the M-step. However, if also constant or nested varying parameters are specified, the component specific estimation problems are not independent from each other any more. Parameters have to be estimated which occur in several or all components and hence, the parameters of the different components have to be determined simultaneously for all components. The estimation problem for all component specific parameters is then obtained by replicating the vector of observations $y = (y_n)_{n=1,\ldots,N}$ $K$ times and defining the covariate matrix $X = (X_{\textrm{constant}}, X_{\textrm{nested}}, X_{\textrm{varying}})$ by \begin{align*} &X_{\textrm{constant}} = \mathbf{1}_K \otimes (x'_{1,n})_{n=1,\ldots,N}\\ &X_{\textrm{nested}} = \mathbf{J} \odot (x'_{2,n})_{n=1,\ldots,N}\\ &X_{\textrm{varying}} = \mathbf{I}_K \otimes(x'_{3,n})_{n=1,\ldots,N}, \end{align*} where $\mathbf{1}_K$ is a vector of 1s of length $K$, $\mathbf{J}$ is the incidence matrix for each component $k=1,\ldots,K$ and each nesting group $c \in C$ and hence is of dimension $K \times |C|$, and $\mathbf{I}_K$ is the identity matrix of dimension $K \times K$. $\otimes$ denotes the Kronecker product and $\odot$ the Khatri-Rao product (i.e., the column-wise Kronecker product). $x_{m,n}$ are the covariates of the corresponding coefficients $\beta_{m,.}$ for $m=1,2,3$. Please note that the weights used for the estimation are the a-posteriori probabilities which are stacked for all components, i.e.~a vector of length $N K$ is obtained. Due to the replication of data in the case of constant or nested varying parameters the amount of memory needed for fitting the mixture model to large datasets is substantially increased and it might be easier to fit only varying coefficients to these datasets. To overcome this problem it could be considered to implement special data structures in order to avoid storing the same data multiple times for large datasets. Before each M-step the average component sizes (over the given data points) are checked and components which are smaller than a given (relative) minimum size are omitted in order to avoid too small components where fitting problems might arise. This strategy has already been recommended for the SEM algorithm \citep{mixtures:Celeux+Diebolt:1988} because it allows to determine the suitable number of components in an automatic way given that the a-priori specified number of components is large enough. This recommendation is based on the assumption that the redundent components will be omitted during the estimation process if the algorithm is started with too many components. If omission of small components is not desired the minimum size required can be set to zero. All components will be then retained throughout the EM algorithm and a mixture with the number of components specified in the initialization will be returned. The algorithm is stopped if the relative change in the log-likelihood is smaller than a pre-specified $\epsilon$ or the maximum number of iterations is reached. For model selection different information criteria are available: AIC, BIC and ICL \citep[Integrated Complete Likelihood;][]{mixtures:Biernacki+Celeux+Govaert:2000}. They are of the form twice the negative loglikelihood plus number of parameters times $k$ where $k=2$ for the AIC and $k$ equals the logarithm of the number of observations for the BIC. The ICL is the same as the BIC except that the complete likelihood (where the missing class memberships are replaced by the assignments induced by the maximum a-posteriori probabilities) instead of the likelihood is used. %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \section{Using the new functionality} \label{sec:using-new-funct} In the following model fitting and model selection in \proglang{R} is illustrated on several examples including mixtures of Gaussian, binomial and Poisson regression models, see also \cite{mixtures:Gruen:2006} and \cite{mixtures:Gruen+Leisch:2007a}. More examples for mixtures of GLMs are provided as part of the software package through a collection of artificial and real world datasets, most of which have been previously used in the literature (see references in the online help pages). Each dataset can be loaded to \proglang{R} with \code{data("}\textit{name}\code{")} and the fitting of the proposed models can be replayed using \code{example("}\textit{name}\code{")}. Further details on these examples are given in a user guide which can be accessed using \code{vignette("regression-examples", package="flexmix")} from within \proglang{R}. %%----------------------------------------------------------------------- \subsection{Artificial example}\label{sec:artificial-example} In the following the artificial dataset \code{NPreg} is used which has already been used in \cite{mixtures:Leisch:2004} to illustrate the application of package \pkg{flexmix}. The data comes from two latent classes of size \Sexpr{nrow(NPreg)/2} each and for each of the classes the data is drawn with respect to the following structure: \begin{center} \begin{tabular}{ll} Class~1: & $ \mathit{yn} = 5x+\epsilon$\\ Class~2: & $ \mathit{yn} = 15+10x-x^2+\epsilon$ \end{tabular} \end{center} with $\epsilon\sim N(0,9)$, see the left panel of Figure~\ref{fig:npreg}. The dataset \code{NPreg} also includes a response $\mathit{yp}$ which is given by a generalized linear model following a Poisson distribution and using the logarithm as link function. The parameters of the mean are given for the two classes by: \begin{center} \begin{tabular}{ll} Class~1: & $ \mu_1 = 2 - 0.2x$\\ Class~2: & $ \mu_2 = 1 + 0.1x$. \end{tabular} \end{center} This signifies that given $x$ the response $\mathit{yp}$ in group $k$ follows a Poisson distribution with mean $e^{\mu_k}$, see the right panel of Figure~\ref{fig:npreg}. \setkeys{Gin}{width=\textwidth} \begin{figure} \centering <>= par(mfrow=c(1,2)) plot(yn~x, col=class, pch=class, data=NPreg) plot(yp~x, col=class, pch=class, data=NPreg) @ \caption{Standard regression example (left) and Poisson regression (right).} \label{fig:npreg} \end{figure} This model can be fitted in \proglang{R} using the commands: <<>>= set.seed(1802) library("flexmix") data("NPreg", package = "flexmix") Model_n <- FLXMRglm(yn ~ . + I(x^2)) Model_p <- FLXMRglm(yp ~ ., family = "poisson") m1 <- flexmix(. ~ x, data = NPreg, k = 2, model = list(Model_n, Model_p), control = list(verbose = 10)) @ If the dimensions are independent the component specific model for multivariate observations can be specified as a list of models for each dimension. The estimation can be controlled with the \code{control} argument which is specified with an object of class \code{"FLXcontrol"}. For convenience also a named list can be provided which is used to construct and set the respective slots of the \code{"FLXcontrol"} object. Elements of the control object are \code{classify} to select ordinary EM, CEM or SEM, \code{minprior} for the minimum relative size of components, \code{iter.max} for the maximum number of iterations and \code{verbose} for monitoring. If \code{verbose} is a positive integer the log-likelihood is reported every \code{verbose} iterations and at convergence together with the number of iterations made. The default is to not report any log-likelihood information during the fitting process. The estimated model \code{m1} is of class \code{"flexmix"} and the result of the default plot method for this class is given in Figure~\ref{fig:root1}. This plot method uses package \pkg{lattice} \citep{mixtures:Sarkar:2008} and the usual parameters can be specified to alter the plot, e.g.~the argument \code{layout} determines the arrangement of the panels. The returned object is of class \code{"trellis"} and the plotting can also be influenced by the arguments of its show method. The default plot prints rootograms (i.e., a histogram of the square root of counts) of the a-posteriori probabilities of each observation separately for each component. For each component the observations with a-posteriori probabilities less than a pre-specified $\epsilon$ (default is $10^{-4}$) for this component are omitted in order to avoid that the bar at zero dominates the plot \citep{mixtures:Leisch:2004a}. Please note that the labels of the y-axis report the number of observations in each bar, i.e.~the squared values used for the rootograms. \begin{figure} \centering <>= print(plot(m1)) @ \caption{The plot method for \code{"flexmix"} objects, here obtained by \code{plot(m1)}, shows rootograms of the posterior class probabilities.} \label{fig:root1} \end{figure} More detailed information on the estimated parameters with respect to standard deviations and significance tests can be obtained with function \code{refit()}. This function determines the variance-covariance matrix of the estimated parameters by using the inverted negative Hesse matrix as computed by the general purpose optimizer \code{optim()} on the full likelihood of the model. \code{optim()} is initialized in the solution obtained with the EM algorithm. For mixtures of GLMs we also implemented the gradient, which speeds up convergence and gives more precise estimates of the Hessian. Naturally, function \code{refit()} will also work for models which have been determined by applying some model selection strategy depending on the data (AIC, BIC, \ldots). The same caution is necessary as when using \code{summary()} on standard linear models selected using \code{step()}: The p-values shown are not correct because they have not been adjusted for the fact that the same data are used to select the model and compute the p-values. So use them only in an exploratory manner in this context, see also \cite{mixtures:Harrell:2001} for more details on the general problem. The returned object can be inspected using \code{summary()} with arguments \code{which} to specify if information for the component model or the concomitant variable model should be shown and \code{model} to indicate for which dimension of the component models this should be done. Selecting \code{model=1} gives the parameter estimates for the dimension where the response variable follows a Gaussian distribution. <<>>= m1.refit <- refit(m1) summary(m1.refit, which = "model", model = 1) @ \begin{figure} \centering <>= print(plot(m1.refit, layout = c(1,3), bycluster = FALSE, main = expression(paste(yn *tilde(" ")* x + x^2))), split= c(1,1,2,1), more = TRUE) print(plot(m1.refit, model = 2, main = expression(paste(yp *tilde(" ")* x)), layout = c(1,2), bycluster = FALSE), split = c(2,1,2,1)) @ \caption{The default plot for refitted \code{"flexmix"} objects, here obtained by \code{plot(refit(m1), model = 1)} and \code{plot(refit(m1), model = 2)}, shows the coefficient estimates and their confidence intervals.} \label{fig:refit} \end{figure} The default plot method for the refitted \code{"flexmix"} object depicts the estimated coefficients with corresponding confidence intervals and is given in Figure~\ref{fig:refit}. It can be seen that for the first model the confidence intervals of the coefficients of the intercept and the quadratic term of \code{x} overlap with zero. A model where these coefficients are set to zero can be estimated with the model driver function \code{FLXMRglmfix()} and the following commands for specifying the nesting structure. The argument \code{nested} needs input for the number of components in each group (given by \code{k}) and the formula which determines the model matrix for the nesting (given by \code{formula}). This information can be provided in a named list. For the restricted model the element \code{k} is a vector with two 1s because each of the components has different parameters. The formulas specifying the model matrices of these coefficients are \verb/~ 1 + I(x^2)/ for an intercept and a quadratic term of $x$ for component 1 and \code{~ 0} for no additional coefficients for component 2. The EM algorithm is initialized in the previously fitted model by passing the posterior probabilities in the argument \code{cluster}. <<>>= Model_n2 <- FLXMRglmfix(yn ~ . + 0, nested = list(k = c(1, 1), formula = c(~ 1 + I(x^2), ~ 0))) m2 <- flexmix(. ~ x, data = NPreg, cluster = posterior(m1), model = list(Model_n2, Model_p)) m2 @ Model selection based on the BIC would suggest the smaller model which also corresponds to the true underlying model. <<>>= c(BIC(m1), BIC(m2)) @ %%----------------------------------------------------------------------- \subsection{Beta-blockers dataset} \label{sec:beta-blockers} The dataset is analyzed in \cite{mixtures:Aitkin:1999, mixtures:Aitkin:1999a} using a finite mixture of binomial regression models. Furthermore, it is described in \citet[p.~165]{mixtures:McLachlan+Peel:2000}. The dataset is from a 22-center clinical trial of beta-blockers for reducing mortality after myocardial infarction. A two-level model is assumed to represent the data, where centers are at the upper level and patients at the lower level. The data is illustrated in Figure~\ref{fig:beta}. First, the center information is ignored and a binomial logit regression model with treatment as covariate is fitted using \code{glm}, i.e.~$K=1$ and it is assumed that the different centers are comparable: <<>>= data("betablocker", package = "flexmix") betaGlm <- glm(cbind(Deaths, Total - Deaths) ~ Treatment, family = "binomial", data = betablocker) betaGlm @ The residual deviance suggests that overdispersion is present in the data. In the next step the intercept is allowed to follow a mixture distribution given the centers. This signifies that the component membership is fixed for each center. This grouping is specified in \proglang{R} by adding \code{| Center} to the formula similar to the notation used in \pkg{nlme} \citep{mixtures:Pinheiro+Bates:2000}. Under the assumption of homogeneity within centers identifiability of the model class can be ensured as induced by the sufficient conditions for identifability given in \cite{mixtures:Follmann+Lambert:1991} for binomial logit models with varying intercepts and \cite{mixtures:Gruen+Leisch:2008} for multinomial logit models with varying and constant parameters. In order to determine the suitable number of components, the mixture is fitted with different numbers of components. <<>>= betaMixFix <- stepFlexmix(cbind(Deaths, Total - Deaths) ~ 1 | Center, model = FLXMRglmfix(family = "binomial", fixed = ~ Treatment), k = 2:4, nrep = 5, data = betablocker) @ The returned object is of class \code{"stepFlexmix"} and printing the object gives the information on the number of iterations until termination of the EM algorithm, a logical indicating if the EM algorithm has converged, the log-likelihood and some model information criteria. The plot method compares the fitted models using the different model information criteria. <<>>= betaMixFix @ A specific \code{"flexmix"} model contained in the \code{"stepFlexmix"} object can be selected using \code{getModel()} with argument \code{which} to specify the selection criterion. The best model with respect to the BIC is selected with: <<>>= betaMixFix_3 <- getModel(betaMixFix, which = "BIC") betaMixFix_3 <- relabel(betaMixFix_3, "model", "Intercept") @ The components of the selected model are ordered with respect to the estimated intercept values. In this case a model with three components is selected with respect to the BIC. The fitted values for the model with three components are given in Figure~\ref{fig:beta} separately for each component and the treatment and control groups. The fitted parameters of the component specific models can be accessed with: <<>>= parameters(betaMixFix_3) @ Please note that the coefficients of variable \code{Treatment} are the same for all three components. \begin{figure} \centering <>= library("grid") betablocker$Center <- with(betablocker, factor(Center, levels = Center[order((Deaths/Total)[1:22])])) clusters <- factor(clusters(betaMixFix_3), labels = paste("Cluster", 1:3)) print(dotplot(Deaths/Total ~ Center | clusters, groups = Treatment, as.table = TRUE, data = betablocker, xlab = "Center", layout = c(3, 1), scales = list(x = list(cex = 0.7, tck = c(1, 0))), key = simpleKey(levels(betablocker$Treatment), lines = TRUE, corner = c(1,0)))) betaMixFix.fitted <- fitted(betaMixFix_3) for (i in 1:3) { seekViewport(trellis.vpname("panel", i, 1)) grid.lines(unit(1:22, "native"), unit(betaMixFix.fitted[1:22, i], "native"), gp = gpar(lty = 1)) grid.lines(unit(1:22, "native"), unit(betaMixFix.fitted[23:44, i], "native"), gp = gpar(lty = 2)) } @ \setkeys{Gin}{width=0.8\textwidth} \caption{Relative number of deaths for the treatment and the control group for each center in the beta-blocker dataset. The centers are sorted by the relative number of deaths in the control group. The lines indicate the fitted values for each component of the 3-component mixture model with varying intercept and constant parameters for treatment.} \label{fig:beta} \end{figure} The variable \code{Treatment} can also be included in the varying part of the model. This signifies that a mixture distribution is assumed where for each component different values are allowed for the intercept and the treatment coefficient. This mixture distribution can be specified using function \code{FLXMRglm()}. Again it is assumed that the heterogeneity is only between centers and therefore the aggregated data for each center can be used. <<>>= betaMix <- stepFlexmix(cbind(Deaths, Total - Deaths) ~ Treatment | Center, model = FLXMRglm(family = "binomial"), k = 3, nrep = 5, data = betablocker) betaMix <- relabel(betaMix, "model", "Treatment") parameters(betaMix) c(BIC(betaMixFix_3), BIC(betaMix)) @ The difference between model \code{betaMix} and \code{betaMixFix\_3} is that the treatment coefficients are the same for all three components for \code{betaMixFix\_3} while they have different values for \code{betaMix} which can easily be seen when comparing the fitted component specific parameters. The larger model \code{betaMix} which also allows varying parameters for treatment has a higher BIC and therefore the smaller model \code{betaMixFix\_3} would be preferred. The default plot for \code{"flexmix"} objects gives a rootogram of the posterior probabilities for each component. Argument \code{mark} can be used to inspect with which components the specified component overlaps as all observations are coloured in the different panels which are assigned to this component based on the maximum a-posteriori probabilities. \begin{figure} \centering <>= print(plot(betaMixFix_3, nint = 10, mark = 1, col = "grey", layout = c(3, 1))) @ \caption{Default plot of \code{"flexmix"} objects where the observations assigned to the first component are marked.}\label{fig:default} \end{figure} \begin{figure} \centering <>= print(plot(betaMixFix_3, nint = 10, mark = 2, col = "grey", layout = c(3, 1))) @ \caption{Default plot of \code{"flexmix"} objects where the observations assigned to the third component are marked.}\label{fig:default-2} \end{figure} The rootogram indicates that the components are well separated. In Figure~\ref{fig:default} it can be seen that component 1 is completely separated from the other two components, while Figure~\ref{fig:default-2} shows that component 2 has a slight overlap with both other components. The cluster assignments using the maximum a-posteriori probabilities are obtained with: <<>>= table(clusters(betaMix)) @ The estimated probabilities of death for each component for the treated patients and those in the control group can be obtained with: <<>>= predict(betaMix, newdata = data.frame(Treatment = c("Control", "Treated"))) @ or by obtaining the fitted values for two observations (e.g.~rows 1 and 23) with the desired levels of the predictor \code{Treatment} <<>>= betablocker[c(1, 23), ] fitted(betaMix)[c(1, 23), ] @ A further analysis of the model is possible with function \code{refit()} which returns the estimated coefficients together with the standard deviations, z-values and corresponding p-values. Please note that the p-values are only approximate in the sense that they have not been corrected for the fact that the data has already been used to determine the specific fitted model. <<>>= summary(refit(betaMix)) @ Given the estimated treatment coefficients we now also compare this model to a model where the treatment coefficient is assumed to be the same for components 1 and 2. Such a model is specified using the model driver \code{FLXMRglmfix()}. As the first two components are assumed to have the same coeffcients for treatment and for the third component the coefficient for treatment shall be set to zero the argument \code{nested} has \code{k = c(2,1)} and \code{formula = c(\~{}Treatment, \~{})}. <<>>= ModelNested <- FLXMRglmfix(family = "binomial", nested = list(k = c(2, 1), formula = c(~ Treatment, ~ 0))) betaMixNested <- flexmix(cbind(Deaths, Total - Deaths) ~ 1 | Center, model = ModelNested, k = 3, data = betablocker, cluster = posterior(betaMix)) parameters(betaMixNested) c(BIC(betaMix), BIC(betaMixNested), BIC(betaMixFix_3)) @ The comparison of the BIC values suggests that the nested model with the same treatment effect for two components and no treatment effect for the third component is the best. %%----------------------------------------------------------------------- \subsection[Productivity of Ph.D. students in biochemistry]{Productivity of Ph.D.~students in biochemistry} \label{sec:bioChemists} <>= data("bioChemists", package = "flexmix") @ This dataset is taken from \cite{mixtures:Long:1990}. It contains \Sexpr{nrow(bioChemists)} observations from academics who obtained their Ph.D.~degree in biochemistry in the 1950s and 60s. It includes \Sexpr{sum(bioChemists$fem=="Women")} women and \Sexpr{sum(bioChemists$fem=="Men")} men. The productivity was measured by counting the number of publications in scientific journals during the three years period ending the year after the Ph.D.~was received. In addition data on the productivity and the prestige of the mentor and the Ph.D.~department was collected. Two measures of family characteristics were recorded: marriage status and number of children of age 5 and lower by the year of the Ph.D. First, mixtures with one, two and three components and only varying parameters are fitted, and the model minimizing the BIC is selected. This is based on the assumption that unobserved heterogeneity is present in the data due to latent differences between the students in order to be productive and achieve publications. Starting with the most general model to determine the number of components using information criteria and checking for possible model restrictions after having the number of components fixed is a common strategy in finite mixture modelling \citep[see][]{mixtures:Wang+Puterman+Cockburn:1996}. Function \code{refit()} is used to determine confidence intervals for the parameters in order to choose suitable alternative models. However, it has to be noted that in the course of the procedure these confidence intervals will not be correct any more because the specific fitted models have already been determined using the same data. <<>>= data("bioChemists", package = "flexmix") Model1 <- FLXMRglm(family = "poisson") ff_1 <- stepFlexmix(art ~ ., data = bioChemists, k = 1:3, model = Model1) ff_1 <- getModel(ff_1, "BIC") @ The selected model has \Sexpr{ff_1@k} components. The estimated coefficients of the components are given in Figure~\ref{fig:coefficients-1} together with the corresponding 95\% confidence intervals using the plot method for objects returned by \code{refit()}. The plot shows that the confidence intervals of the parameters for \code{kid5}, \code{mar}, \code{ment} and \code{phd} overlap for the two components. In a next step a mixture with two components is therefore fitted where only a varying intercept and a varying coefficient for \code{fem} is specified and all other coefficients are constant. The EM algorithm is initialized with the fitted mixture model using \code{posterior()}. \begin{figure} \centering <>= print(plot(refit(ff_1), bycluster = FALSE, scales = list(x = list(relation = "free")))) @ \caption{Coefficient estimates and confidence intervals for the model with only varying parameters.}\label{fig:coefficients-1} \end{figure} <<>>= Model2 <- FLXMRglmfix(family = "poisson", fixed = ~ kid5 + mar + ment) ff_2 <- flexmix(art ~ fem + phd, data = bioChemists, cluster = posterior(ff_1), model = Model2) c(BIC(ff_1), BIC(ff_2)) @ If the BIC is used for model comparison the smaller model including only varying coefficients for the intercept and \code{fem} is preferred. The coefficients of the fitted model can be obtained using \code{refit()}: <<>>= summary(refit(ff_2)) @ It can be seen that the coefficient of \code{phd} does for both components not differ significantly from zero and might be omitted. This again improves the BIC. <<>>= Model3 <- FLXMRglmfix(family = "poisson", fixed = ~ kid5 + mar + ment) ff_3 <- flexmix(art ~ fem, data = bioChemists, cluster = posterior(ff_2), model = Model3) c(BIC(ff_2), BIC(ff_3)) @ The coefficients of the restricted model without \code{phd} are given in Figure~\ref{fig:coefficients-2}. \begin{figure}[t] \centering <>= print(plot(refit(ff_3), bycluster = FALSE, scales = list(x = list(relation = "free")))) @ \caption{Coefficient estimates and confidence intervals for the model with varying and constant parameters where the variable \code{phd} is not used in the regression.}\label{fig:coefficients-2} \end{figure} An alternative model would be to assume that gender does not directly influence the number of articles but has an impact on the segment sizes. <<>>= Model4 <- FLXMRglmfix(family = "poisson", fixed = ~ kid5 + mar + ment) ff_4 <- flexmix(art ~ 1, data = bioChemists, cluster = posterior(ff_2), concomitant = FLXPmultinom(~ fem), model = Model4) parameters(ff_4) summary(refit(ff_4), which = "concomitant") BIC(ff_4) @ This suggests that the proportion of women is lower in the second component which is the more productive segment. The alternative modelling strategy where homogeneity is assumed at the beginning and a varying interept is added if overdispersion is observed leads to the following model which is the best with respect to the BIC. <<>>= Model5 <- FLXMRglmfix(family = "poisson", fixed = ~ kid5 + ment + fem) ff_5 <- flexmix(art ~ 1, data = bioChemists, cluster = posterior(ff_2), model = Model5) BIC(ff_5) @ \begin{figure} \centering \setkeys{Gin}{width=0.8\textwidth} <>= pp <- predict(ff_5, newdata = data.frame(kid5 = 0, mar = factor("Married", levels = c("Single", "Married")), fem = c("Men", "Women"), ment = mean(bioChemists$ment))) matplot(0:12, sapply(unlist(pp), function(x) dpois(0:12, x)), type = "b", lty = 1, xlab = "Number of articles", ylab = "Probability") legend("topright", paste("Comp.", rep(1:2, each = 2), ":", c("Men", "Women")), lty = 1, col = 1:4, pch = paste(1:4), bty = "n") @ \caption{The estimated productivity for each compoment for men and women.} \label{fig:estimated} \end{figure} \setkeys{Gin}{width=0.98\textwidth} In Figure~\ref{fig:estimated} the estimated distribution of productivity for model \code{ff\_5} are given separately for men and women as well as for each component where for all other variables the mean values are used for the numeric variables and the most frequent category for the categorical variables. The two components differ in that component 1 contains the students who publish no article or only a single article, while the students in component 2 write on average several articles. With a constant coefficient for gender women publish less articles than men in both components. This example shows that different optimal models are chosen for different modelling procedures. However, the distributions induced by the different variants of the model class may be similar and therefore it is not suprising that they then will have similar BIC values. %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \section{Implementation}\label{sec:implementation} The new features extend the available model class described in \cite{mixtures:Leisch:2004} by providing infrastructure for concomitant variable models and for fitting mixtures of GLMs with varying and constant parameters for the component specific parameters. The implementation of the extensions of the model class made it necessary to define a better class structure for the component specific models and to modify the fit functions \code{flexmix()} and \code{FLXfit()}. An overview on the \proglang{S}4 class structure of the package is given in Figure~\ref{fig:class structure}. There is a class for unfitted finite mixture distributions given by \code{"FLXdist"} which contains a list of \code{"FLXM"} objects which determine the component specific models, a list of \code{"FLXcomponent"} objects which specify functions to determine the component specific log-likelihoods and predictions and which contain the component specific parameters, and an object of class \code{"FLXP"} which specifies the concomitant variable model. Class \code{"flexmix"} extends \code{"FLXdist"}. It represents a fitted finite mixture distribution and it contains the information about the fitting with the EM algorithm in the object of class \code{"FLXcontrol"}. Repeated fitting with the EM algorithm with different number of components is provided by function \code{stepFlexmix()} which returns an object of class \code{"stepFlexmix"}. Objects of class \code{"stepFlexmix"} contain the list of the fitted mixture models for each number of components in the slot \code{"models"}. \setkeys{Gin}{width=.9\textwidth} \begin{figure}[t] \centering \includegraphics{flexmix} \caption{UML class diagram \citep[see][]{mixtures:Fowler:2004} of the \pkg{flexmix} package.} \label{fig:class structure} \end{figure} \setkeys{Gin}{width=\textwidth} For the component specific model a virtual class \code{"FLXM"} is introduced which (currently) has two subclasses: \code{"FLXMC"} for model-based clustering and \code{"FLXMR"} for clusterwise regression, where predictor variables are given. Additional slots have been introduced to allow for data preprocessing and the construction of the components was separated from the fit and is implemented using lexical scoping \citep{mixtures:Gentleman+Ihaka:2000} in the slot \code{defineComponent}. \code{"FLXMC"} has an additional slot \code{dist} to specify the name of the distribution of the variable. In the future functionality shall be provided for sampling from a fitted or unfitted finite mixture. Using this slot observations can be generated by using the function which results from adding an \code{r} at the beginnning of the distribution name. This allows to only implement the (missing) random number generator functions and otherwise use the same method for sampling from mixtures with component specific models of class \code{"FLXMC"}. For \code{flexmix()} and \code{FLXfit()} code blocks which are model dependent have been identified and different methods implemented. Finite mixtures of regressions with varying, nested and constant parameters were a suitable model class for this identification task as they are different from models previously implemented. The main differences are: \begin{itemize} \item The number of components is related to the component specific model and the omission of small components during the EM algorithm impacts on the model. \item The parameters of the component specific models can not be determined separately in the M-step and a joint model matrix is needed. \end{itemize} This makes it also necessary to have different model dependent methods for \code{fitted()} which extracts the fitted values from a \code{"flexmix"} object, \code{predict()} which predicts new values for a \code{"flexmix"} object and \code{refit()} which refits an estimated model to obtain additional information for a \code{"flexmix"} object. %%----------------------------------------------------------------------- \subsection{Component specific models with varying and constant parameters}\label{sec:comp-models-with} A new M-step driver is provided which fits finite mixtures of GLMs with constant and nested varying parameters for the coefficients and the dispersion parameters. The class \code{"FLXMRglmfix"} returned by the driver \code{FLXMRglmfix()} has the following additional slots with respect to \code{"FLXMRglm"}: \begin{description} \item[\code{design}:] An incidence matrix indicating which columns of the model matrix are used for which component, i.e.~$\mathbf{D}=(\mathbf{1}_K,\mathbf{J}, \mathbf{I}_K)$. \item[\code{nestedformula}:] An object of class \code{"FLXnested"} containing the formula for the nested regression coefficients and the number of components in each $K_c$, $c \in C$. \item[\code{fixed}:] The formula for the constant regression coefficients. \item[\code{variance}:] A logical indicating if different variances shall be estimated for the components following a Gaussian distribution or a vector specifying the nested structure for estimating these variances. \end{description} The difference between estimating finite mixtures including only varying parameters using models specified with \code{FLXMRglm()} and those with varying and constant parameters using function \code{FLXMRglmfix()} is hidden from the user, as only the specified model is different. The fitted model is also of class \code{"flexmix"} and can be analyzed using the same functions as for any model fitted using package \pkg{flexmix}. The methods used are the same except if the slot containing the model is accessed and method dispatching is made via the model class. New methods are provided for models of class \code{"FLXMRglmfix"} for functions \code{refit()}, \code{fitted()} and \code{predict()} which can be used for analyzing the fitted model. The implementation allows repeated measurements by specifying a grouping variable in the formula argument of \code{flexmix()}. Furthermore, it has to be noticed that the model matrix is determined by updating the formula of the varying parameters successively with the formula of the constant and then of the nested varying parameters. This ensures that if a mixture distribution is fitted for the intercept, the model matrix of a categorical variable includes only the remaining columns for the constant parameters to have full column rank. However, this updating scheme makes it impossible to estimate a constant intercept while allowing varying parameters for a categorical variable. For this model one big model matrix is constructed where the observations are repeated $K$ times and suitable columns of zero added. The coefficients of all $K$ components are determined simultaneously in the M-step, while if only varying parameters are specified the maximization of the likelihood is made separately for all components. For large datasets the estimation of a combination of constant and varying parameters might therefore be more challenging than only varying parameters. %% ----------------------------------------------------------------------- \subsection{Concomitant variable models}\label{sec:conc-vari-models} For representing concomitant variable models the class \code{"FLXP"} is defined. It specifies how the concomitant variable model is fitted using the concomitant variable model matrix as predictor variables and the current a-posteriori probability estimates as response variables. The object has the following slots: \begin{description} \item[\code{fit}:] A \code{function (x, y, ...)} returning the fitted values for the component weights during the EM algorithm. \item[\code{refit}:] A \code{function (x, y, ...)} used for refitting the model. \item[\code{df}:] A \code{function (x, k, ...)} returning the degrees of freedom used for estimating the concomitant variable model given the model matrix \code{x} and the number of components \code{k}. \item[\code{x}:] A matrix containing the model matrix of the concomitant variables. \item[\code{formula}:] The formula for determining the model matrix \code{x}. \item[\code{name}:] A character string describing the model, which is only used for print output. \end{description} Two constructor functions for concomitant variable models are provided at the moment. \code{FLXPconstant()} is for constant component weights without concomitant variables and for multinomial logit models \code{FLXPmultinom()} can be used. \code{FLXPmultinom()} has its own class \code{"FLXPmultinom"} which extends \code{"FLXP"} and has an additional slot \code{coef} for the fitted coefficients. The multinomial logit models are fitted using package \pkg{nnet} \citep{mixtures:Venables+Ripley:2002}. %%----------------------------------------------------------------------- \subsection{Further changes} The estimation of the model with the EM algorithm was improved by adapting the variants to correspond to the CEM and SEM variants as outlined in the literature. To make this more explicit it is now also possible to use \code{"CEM"} or \code{"SEM"} to specify an EM variant in the \code{classify} argument of the \code{"FLXcontrol"} object. Even though the SEM algorithm can in general not be expected to converge the fitting procedure is also terminated for the SEM algorithm if the change in the relative log-likelhood is smaller than the pre-specified threshold. This is motivated by the fact that for well separated clusters the posteriors might converge to an indicator function with all weight concentrated in one component. The fitted model with the maximum likelihood encountered during the SEM algorithm is returned. For discrete data in general multiple observations with the same values are given in a dataset. A \code{weights} argument was added to the fitting function \code{flexmix()} in order to avoid repeating these observations in the provided dataset. The specification is through a \code{formula} in order to allow selecting a column of the data frame given in the \code{data} argument. The weights argument allows to avoid replicating the same observations and hence enables more efficient memory use in these applications. This possibitliy is especially useful in the context of model-based clustering for mixtures of Poisson distributions or latent class analysis with multivariate binary observations. In order to be able to apply different initialization strategies such as for example first running several different random initializations with CEM and then switching to ordinary EM using the best solution found by CEM for initialization a \code{posterior()} function was implemented. \code{posterior()} also takes a \code{newdata} argument and hence, it is possible to apply subset strategies for large datasets as suggested in \cite{mixtures:Wehrens+Buydens+Fraley:2004}. The returned matrix of the posterior probabilities can be used to specify the \code{cluster} argument for \code{flexmix()} and the posteriors are then used as weights in the first M-step. The default plot methods now use trellis graphics as implemented in package \pkg{lattice} \citep{mixtures:Sarkar:2008}. Users familiar with the syntax of these graphics and with the plotting and printing arguments will find the application intuitive as a lot of plotting arguments are passed to functions from \pkg{lattice} as for example \code{xyplot()} and \code{histogram()}. In fact only new panel, pre-panel and group-panel functions were implemented. The returned object is of class \code{"trellis"} and the show method for this class is used to create the plot. Function \code{refit()} was modified and has now two different estimation methods: \code{"optim"} and \code{"mstep"}. The default method \code{"optim"} determines the variance-covariance matrix of the parameters from the inverse Hessian of the full log-likelihood. The general purpose optimizer \code{optim()} is used to maximize the log-likelihood and initialized in the solution obtained with the EM algorithm. For mixtures of GLMs there are also functions implemented to determine the gradient which can be used to speed up convergence. The second method \code{"mstep"} is only a raw approximation. It performs an M-step where the a-posteriori probabilities are treated as given instead of estimated and returns for the component specific models nearly complete \code{"glm"} objects which can be further analyzed. The advantage of this method is that the return value is basically a list of standard \code{"glm"} objects, such that the regular methods for this class can be used. %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \section{Writing your own drivers}\label{sec:writing-your-own} Two examples are given in the following to demonstrate how new drivers can be provided for concomitant variable models and for component specific models. Easy extensibility is one of the main implementation aims of the package and it can be seen that writing new drivers requires only a few lines of code for providing the constructor functions which include the fit functions. %%----------------------------------------------------------------------- \subsection{Component specific models: Zero-inflated models}\label{sec:component-models} \lstset{frame=trbl,basicstyle=\small\tt,stepnumber=5,numbers=left} In Poisson or binomial regression models it can be often encountered that the observed number of zeros is higher than expected. A mixture with two components where one has mean zero can be used to model such data. These models are also referred to as zero-inflated models \citep[see for example][]{mixtures:Boehning+Dietz+Schlattmann:1999}. A generalization of this model class would be to fit mixtures with more than two components where one component has a mean fixed at zero. So this model class is a special case of a mixture of generalized linear models where (a) the family is restricted to Poisson and binomial and (b) the parameters of one component are fixed. For simplicity the implementation assumes that the component with mean zero is the first component. In addition we assume that the model matrix contains an intercept and to have the first component absorbing the access zeros the coefficient of the intercept is set to $-\infty$ and all other coefficients are set to zero. Hence, to implement this model using package \pkg{flexmix} an appropriate model class is needed with a corresponding convenience function for construction. During the fitting of the EM algorithm using \code{flexmix()} different methods for this model class are needed when determining the model matrix (to check the presence of an intercept), to check the model after a component is removed and for the M-step to account for the fact that the coefficients of the first component are fixed. For all other methods those available for \code{"FLXMRglm"} can be re-used. The code is given in Figure~\ref{fig:ziglm.R}. \begin{figure} \centering \begin{minipage}{0.98\textwidth} \lstinputlisting{ziglm.R} \end{minipage} \caption{Driver for a zero-inflated component specific model.} \label{fig:ziglm.R} \end{figure} The model class \code{"FLXMRziglm"} is defined as extending \code{"FLXMRglm"} in order to be able to inherit methods from this model class. For construction of a \code{"FLXMRziglm"} class the convenicence function \code{FLXMRziglm()} is used which calls \code{FLXMRglm()}. The only differences are that the family is restricted to binomial or Poisson, that a different name is assigned and that an object of the correct class is returned. The presence of the intercept in the model matrix is checked in \code{FLXgetModelmatrix()} after using the method available for \code{"FLXMRglm"} models as indicated by the call to \code{callNextMethod()}. During the EM algorithm \code{FLXremoveComponent()} is called if one component is removed. For this model class it checks if the first component has been removed and if this is the case the model class is changed to \code{"FLXMRglm"}. In the M-step the coefficients of the first component are fixed and not estimated, while for the remaining components the M-step of \code{"FLXMRglm"} objects can be used. During the EM algorithm \code{FLXmstep()} is called to perform the M-step and returns a list of \code{"FLXcomponent"} objects with the fitted parameters. A new method for this function is needed for \code{"FLXMRziglm"} objects in order to account for the fixed coefficients in the first component, i.e.~for the first component the \code{"FLXcomponent"} object is constructed and concatenated with the list of \code{"FLXcomponent"} objects returned by using the \code{FLXmstep()} method for \code{"FLXMRglm"} models for the remaining components. Similar modifications are necessary in order to be able to use \code{refit()} for this model class. The code for implementing the \code{refit()} method using \code{optim()} for \code{"FLXMRziglm"} is not shown, but can be inspected in the source code of the package. \subsubsection{Example: Using the driver} This new M-step driver can be used to estimate a zero-inflated Poisson model to the data given in \cite{mixtures:Boehning+Dietz+Schlattmann:1999}. The dataset \code{dmft} consists of count data from a dental epidemiological study for evaluation of various programs for reducing caries collected among school children from an urban area of Belo Horizonte (Brazil). The variables included are the number of decayed, missing or filled teeth (DMFT index) at the beginning and at the end of the observation period, the gender, the ethnic background and the specific treatment for \Sexpr{nrow(dmft)} children. The model can be fitted with the new driver function using the following commands: <<>>= data("dmft", package = "flexmix") Model <- FLXMRziglm(family = "poisson") Fitted <- flexmix(End ~ log(Begin + 0.5) + Gender + Ethnic + Treatment, model = Model, k = 2 , data = dmft, control = list(minprior = 0.01)) summary(refit(Fitted)) @ Please note that \cite{mixtures:Boehning+Dietz+Schlattmann:1999} added the predictor \code{log(Begin + 0.5)} to serve as an offset in order to be able to analyse the improvement in the DMFT index from the beginning to the end of the study. The linear predictor with the offset subtracted is intended to be an estimate for $\log(\mathbb{E}(\textrm{End})) - \log(\mathbb{E}(\textrm{Begin}))$. This is justified by the fact that for a Poisson distributed variable $Y$ with mean between 1 and 10 it holds that $\mathbb{E}(\log(Y + 0.5))$ is approximately equal to $\log(\mathbb{E}(Y))$. $\log(\textrm{Begin} + 0.5)$ can therefore be seen as an estimate for $\log(\mathbb{E}(\textrm{Begin}))$. The estimated coefficients with corresponding confidence intervals are given in Figure~\ref{fig:dmft}. As the coefficients of the first component are restricted a-priori to minus infinity for the intercept and to zero for the other variables, they are of no interest and only the second component is plotted. The box ratio can be modified as for \code{barchart()} in package \pkg{lattice}. The code to produce this plot is given by: <>= print(plot(refit(Fitted), components = 2, box.ratio = 3)) @ \begin{figure} \centering \setkeys{Gin}{width=0.9\textwidth} <>= <> @ \caption{The estimated coefficients of the zero-inflated model for the \code{dmft} dataset. The first component is not plotted as this component captures the inflated zeros and its coefficients are fixed a-priori.} \label{fig:dmft} \end{figure} %%----------------------------------------------------------------------- \subsection{Concomitant variable models}\label{sec:concomitant-models} If the concomitant variable is a categorical variable, the multinomial logit model is equivalent to a model where the component weights for each level of the concomitant variable are determined by the mean values of the a-posteriori probabilities. The driver which implements this \code{"FLXP"} model is given in Figure~\ref{fig:myConcomitant.R}. A name for the driver has to be specified and a \code{fit()} function. In the \code{fit()} function the mean posterior probability for all observations with the same covariate points is determined, assigned to the corresponding observations and the full new a-posteriori probability matrix returned. By contrast \code{refit()} only returns the new a-posteriori probability matrix for the number of unique covariate points. \lstset{frame=trbl,basicstyle=\small\tt,stepnumber=5,numbers=left} \begin{figure} \centering \begin{minipage}{0.98\textwidth} \lstinputlisting{myConcomitant.R} \end{minipage} \caption{Driver for a concomitant variable model where the component weights are determined by averaging over the a-posteriori probabilities for each level of the concomitant variable.} \label{fig:myConcomitant.R} \end{figure} \subsubsection{Example: Using the driver} If the concomitant variable model returned by \code{myConcomitant()} is used for the artificial example in Section~\ref{sec:using-new-funct} the same fitted model is returned as if a multinomial logit model is specified. An advantage is that in this case no problems occur if the fitted probabilities are close to zero or one. <>= Concomitant <- FLXPmultinom(~ yb) MyConcomitant <- myConcomitant(~ yb) m2 <- flexmix(. ~ x, data = NPreg, k = 2, model = list(Model_n, Model_p), concomitant = Concomitant) m3 <- flexmix(. ~ x, data = NPreg, k = 2, model = list(Model_n, Model_p), cluster = posterior(m2), concomitant = MyConcomitant) @ <<>>= summary(m2) summary(m3) @ For comparing the estimated component weights for each value of $\mathit{yb}$ the following function can be used: <<>>= determinePrior <- function(object) { object@concomitant@fit(object@concomitant@x, posterior(object))[!duplicated(object@concomitant@x), ] } @ <<>>= determinePrior(m2) determinePrior(m3) @ Obviously the fitted values of the two models correspond to each other. %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \section{Summary and outlook}\label{sec:summary-outlook} Package \pkg{flexmix} was extended to cover finite mixtures of GLMs with (nested) varying and constant parameters. This allows for example the estimation of varying intercept models. In order to be able to characterize the components given some variables concomitant variable models can be estimated for the component weights. The implementation of these extensions have triggered some modifications in the class structure and in the fit functions \code{flexmix()} and \code{FLXfit()}. For certain steps, as e.g.~the M-step, methods which depend on the component specific models are defined in order to enable the estimation of finite mixtures of GLMs with only varying parameters and those with (nested) varying and constant parameters with the same fit function. The flexibility of this modified implementation is demonstrated by illustrating how a driver for zero-inflated models can be defined. In the future diagnostic tools based on resampling methods shall be implemented as bootstrap results can give valuable insights into the model fit \citep{mixtures:Gruen+Leisch:2004}. A function which conveniently allows to test linear hypotheses about the parameters using the variance-covariance matrix returned by \code{refit()} would be a further valuable diagnostic tool. The implementation of zero-inflated Poisson and binomial regression models are a first step towards relaxing the assumption that all component specific distributions are from the same parametric family. As mixtures with components which follow distributions from different parametric families can be useful for example to model outliers \citep{mixtures:Dasgupta+Raftery:1998,mixtures:Leisch:2008}, it is intended to also make this functionality available in \pkg{flexmix} in the future. %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \section*{Computational details} <>= SI <- sessionInfo() pkgs <- paste(sapply(c(SI$otherPkgs, SI$loadedOnly), function(x) paste("\\\\pkg{", x$Package, "} ", x$Version, sep = "")), collapse = ", ") @ All computations and graphics in this paper have been done using \proglang{R} version \Sexpr{getRversion()} with the packages \Sexpr{pkgs}. %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \section*{Acknowledgments} This research was supported by the the Austrian Science Foundation (FWF) under grants P17382 and T351. Thanks also to Achim Zeileis for helpful discussions on implementation details and an anonymous referee for asking a good question about parameter significance which initiated the new version of function \code{refit()}. %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \bibliography{mixture} %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \end{document} flexmix/inst/doc/bootstrapping.pdf0000644000176200001440000032667213432516317017063 0ustar liggesusers%PDF-1.5 %¿÷¢þ 1 0 obj << /Type /ObjStm /Length 3016 /Filter /FlateDecode /N 51 /First 400 >> stream xœÕZKsÛ8¾ï¯Àm’J& ·¦ReGÛËv$;q2•-A'’èTâä·ïa»BâKíx[2ÍÀF£ñõô #œÈ„G>$ˆ""‰œ„DFQDy‰H$ñüy’À/pö‰†ñ8ñ¹Ïá!ñ•Oî E +x€8ƒçŠÜx ÂPŸÁ$¼äÁÃøÀ€„F %"%‰és¸DràËÆ$ƒþÈšd0 =|/"!ç0FBÀÙ#¡TœpŸ„*‚3L&æÅE¸ Êg>á’(B{H"Ÿ³ŠDöƒé0Šð€dè‘æË˜!q¼}˜JD0 tü&ba\â á"À€“ E‹JÀ¤á" $Sð\ åÈþõÇ„uOã"¡ÀêŒ=]‹d¥sXs_ÃM`oοßhB_AÿEzM^¾4$öÖÅ<ÍÈ“}]É*&ÙWÏÉë,ÑÓ,™ÌɱNòÉü)¼—é¸HÒÕ .4y2ø·Ï¼ˆùžò|Á{ƼßûÍõCŠÇñ¹¾$ß’bNæ0r–é4¿Ñß¿¥Ù4'OFÏÉ,Y%@m™Üë ÎéT/òç$Óy¼¼‰\?'WiZäEßÀ»Ãtú³Ñϲtºžhþàì˜ÌÓ¼È'YrSè…/¡Ãx}õ·žä \Ÿ'Ũ½¶\ K.†È$ñõ ÞN&9¹È2ÚpE@îótš?uB|•®W‚ž¾I`jv Ä•ùÏ­ø"x Ë“}(ì:©ÐÞÙ"ûö'·bv7C­ ½*rÍvL“x?½…a‘¶ˆÄ €» ¼*B"gq/8À,Òu6‚HíÏÛâ`\ Le‰˜×@µÜÂ'K'c]ez6x üèÛH¾|Ye­.„H$Y^’Äq ×"27eÇ΋Œèå‡`8^€bHùô{µ^,íK‡c3‡’ÇrÝŽ`5ôž\’tVY¶–h#7ù˜r˱•çÛþ&›fËxú{6$tzCŠl­]ïíhî•8×f@zr8½Úöj¸ 6–þ¹š¤Sœ~P²mÖâÕ<ÎÐ&AúF!@EÁ:oÖ¨ì )̰8¸çx°Ã½O¦Å.}\™ˆ‡FvÿqÏ·Ò—þæèjw?4a„pR¦gÎB)œlà)ÓC†h`å–1FÚ²«d°ƒ  (c¦§7îÈ0ƒ÷Ž <ã”MŸ² :˜p踖ˆ8ñ’âf xãÓÆìæWí`‘our=w·°ž'tîÓ}Mè=¦C:¢czAc:¡“t‘®è”N ˜OrªéŒÎf …¿k:§ ]Ð%]Ñ”¦+Mo(˜ß$Ҍ洠Å<ÓšßRº¦_é-ýñÔè¤aèw0¬°Ø‹ø:7ª «¾ïtÄ´ï†ÏO¶õu²Ðèã­œÄKÝBæxd²·ºÍ…I“ªÁ“‘zù^E\¬-蟎O?~<€Fhx·Èwœlß|þ`à‡aò"aL•2 _ìîg>PåJ™ƒ™1Ì æ›ñ°oŸÃ/ßd>CjM¡„NÊ.á%ûÂË„ö?+¯ÜQ¿»ïq_Z[.ªwÝïV9­€÷—yîO¡:Zì°D{_L Gú>ßí=04;ÙøŒ zI?€bÄù$IŠd1Õp‰F8ÿLãôå*ÎÚŽ$],àñÆ8ÿQõ˜Uë¬í2:Ïïç7œ¨h(z‚|¾«îÖ«©ÎòIšé»\ƒ9Uæ¢ál«Àò ¸®Èk*rCµšŠßVV§®ÖøÄ©µø©Z~[­ƒýáÛ {|\Ská¸ÜúŠeŽZÊÜ7žô;âÉݿиSŒâ¶1ÜÆK”±Û¦/·×¡Øé9LU ÔX-Ls‘i3Dc­ÍªrØ¢ÞÎwD=ßUÅrÙŽÍu\¦Ó ¶ê ÎM™Õ4ü Vév,¶1‚ Ço€Òk‚²Œ¾Þ%Pò.ƒñèÍáG ŽTßÐ…µ†ð[y½××·^3‰Ü3G þêOD»Z»ûEûsk»L¢ MvĹÝq vÀL¾éz•JP êböŽÚ•´ ²w<ŸìõáQ=.×Jìµ!Ò΀½¾ád»’•e›Ðþÿ5hl³ÖAO1cýÿNš5ÈÈzÈqì°vµí– ª±pPØ¡ ¬¥Òè•ãåÕ4¦Ëõ&UKÀÅ^C+xÐÚ¼FŠÙ6¥¼.ÇÉ›@k@¡¯â~_¬Þ¾\œ üPµK‹Å·ÖHµ¡ÖL¤°O iéÅ ’Ó©†éÕ¶+šîR©Öâ`TY{»Yœ›,YêŠÐw䱇y»!nÑw]=¥ôN`ß¾;{w<0ä롈ߔv‡éo¥¬¯éïªlýÿÿЮÜÏ]`„¾LVë¼Wý£t ¨¥.¡²©b¡ÓXÛ¾šÊz{…óñéûác úFX¼má¨Ѳ¾­m}·ýü\+ üÄ{Žð.³Ø*A×eÑSÖaÔWÔƒ÷£ƒcC~¿î]Ík+â¶æí ­¯¢Ê¶ƒ À8Tlo7Ž7ó¸*dµÌ ‚âÀ¼ßn«Ž×œ~o,ïJÐÊÀàÖö Gø!³Ã€hÀ‘;g1‚ÓiÛ?$PîZß½Ë~8`¯!löÝ»JøîÝ¿Lp×?¨p²ÝDtß#m"V„‹ß<`qƒÊáåxtðÁä+¿V¡â² ÐT¨š•Ÿ‡WvºjE¬óéîÑÙN~X£íîZÖÃkPwSp£ÿZuÊÕ¦ËmÃÓJmª»2uE¯²xòY&?*¯m†ô8µªZ¥êgE*œòŽMŽjÍêêUQT/ øÌ6K/ìp0ÎŽlM`Kת6Ðx²¾E+y¢®5kEb°^™ñ£mØz!|l•¼>]°aÜ|%B7Óa ËnŒÙ æ¶6WŸ*¶IÕL¯ÅðÎÌv×*Y‡±¯ø{uˆ<¸p×’ð°Ûx»ä÷qŒwPý$xÐ ãýn|t0x÷ìP/¾ê"™ÄÕèßk.gGåGª–Íö{ÚlÜlæ›ík¡ðã9e!dë`µÿØ«z‡¾_¹Úö©o’K ºVÙº’²™†”#1wíx‚qËðl‹”Uq6vSÜõ¯o–ÿ"«`“>æwå5»«Ôpyïpƒd6Ó4DÖ_æ3Y‚q nÛB²«}n? Û¼ý!°ÒcK„.טŸÓ›3G[¤ÀÏÐJóK¬Ùß=n04GÄ&±;ð&0žÍjbAž·é&z©m}V¶Ûß$òIm£ÜØ€r¿þ‘1òl¢îG¢‡5SÉx,þþ9°(&q~$zÌ‘35 HË»éþÛØ(endstream endobj 53 0 obj << /Subtype /XML /Type /Metadata /Length 1684 >> stream GPL Ghostscript 9.26 R, finite mixture models, resampling, bootstrap 2019-02-18T12:53:50+01:00 2019-02-18T12:53:50+01:00 LaTeX with hyperref Finite Mixture Model Diagnostics Using Resampling MethodsBettina Grün, Friedrich Leisch endstream endobj 54 0 obj << /Type /ObjStm /Length 1959 /Filter /FlateDecode /N 50 /First 397 >> stream xœÕYmoÛ6þ¾_Á†…â; ’¸IÓ5kk·]Ü!T‡Ž…9¶a)]º_¿;J²)ÉÎÔÄÀ0­)ñõž{ž»£ªIˆÒ„1A”!\¢,^:¢¸$:!Z(¢aI¿œ0.9¼#L;è“„'Þ+Â9¼Ôšpiá=¬%!Ú®,ü:·LbÞN„tœA„J1’Í1 ‹ ¿ú Îb,‘ œÕ8"¹à¶’æ[F¤‚ý`¨42!Vé¤#VÅèWD ¿~9¼;U"ˆµ`=nÁZ£8“”Ã:8°´Æý˜Æ9#áÂ& ØÇi€^¢© Îí,üÂb΀XHl0hÀ2xma*ƒíµãØ€Kâ¾ûùgB>/ÈÄ‚[†„^?iG`¤QGJI²¸ŸÏÉ5¡¿ù‡F½M×~Q]=­ýRõ¼Ïй'߯‹lšM²tNnÒ"%¹/~ ¿ü6<]. ˜£_ËI—þ&KO–pŒÙáÔxÎJv†]o¶«F}¾¼_O|NpµÅù¨H ˆ‡þ3Xa¯Î¶œŒ<HßÎà|hÆ5š_W×¾õõÉ*(”naa„8JàDM,Ü^,”j€1òÙß÷k¿ uHÀ×[œ|m2p«dD†]67 pî~ÙrA–S’§_ü ô.ï‹Õý– 'iîÃaé‡WƒO<½«ÔºzéZV ëU²-Üf7Áßñã|‚Žñ ÓÕKŸÝÎêG€‰ò=ýL=ÎiFïè’ôá‡@Œ0ñ' ÎÓ[à®`‰Ñ%¢'5u¡ÛŠ$¬w]ves!ˆë-5Kï|ÛãE:Ï&Ç‹[à<^fyþ~ fÑQáï>†ˆy2"½ªl‘Jtéu2þ}<|›½¿hð #b›`¼K0Ñ!˜èK0H_Os·Llø[¹ä%c~óßËÝXÁHÈZ f¾’8ZÕgQó®mÎ1@)mðÙ@¦Â3×svv•: ë6î]?×çà@XËLXß—§•ÓÐ.¹T ©5†*Â7ŽéU5ÉÊtLÖczBOé}E¥—ô }K‡tDßÓh<¡“å|¹ 7@h¿¸IóÒiF§>é-ÑÙ×ÕÌ/€ésàúؾ¢+ ms?-ÊÖ÷¤+¿Î–7tMsÐCñ×’ÞÓ¿èWú·_/cqpn"r\°o (^û›Ù ò&Ч7ã¡íxØÄṩ,*EäŽËà‹×£Wx5u|¬ƒôii;+Û¤4ŽéàIP="A¡»…‚{}ù¸~µÎîü.ø»È ¸DàìòÝ‹RžÈ£Õ%¡¨î{óúÓ»sXþòâÄ6#oà ®7oóšõ¤µµvh ¢Œi7ئóÕ,“½Û.ô¨q~ ÝN™Ý2¿ï=?CôÃwîJ(».§øG„ÙM%ÖÈDI¥C_1šƒ‰ñ9¶øß%;E$ù!Eät,"ó=?:í'bö#dÓ©CѲ?·óiø~_žs3î)‹a…ªýë5'à]¯a‡þm¦ƒCýp¨õ@íÓ¬ü\~Y<Ô àš¶{¹…ÉGendstream endobj 105 0 obj << /Filter /FlateDecode /Length 6487 >> stream xœÕ<Ër$·‘wZÑ¡Sõ†Ø[x»±É+ÉkËK£ðÁr8zøè¡D²)r¨™Ñe}óè*²iIáØ˜P¬F%ùÎD¢~\µñ_þÿÙÍÉ¿­Üj÷p2®v'?ž(úu•ÿwv³úìÎð+6Z9»zuy¯ª•Òi“ŒY­6ʦի›“¿ _¬OxJÉ¥áj=nlL)èáö ŸÛ”àÏá-p±i»FBtÚ ;|ì w»¯=¬q‡.Œð©Se§à3œ=0B!éá[š’¢õùÅè¬âÉ£H“×Ñ£öõÇ“¨7J½:²&åãêÕ9ÐõëLÆ€ûf£·1MÉgÜa›* 7ݻ뺽nY¤…S Õ±äæÞ 5ŠÎðy•v„/‹àêœFñ8µÈO°zã­W¼ÏÏh£2}•LæhÔ‘—4á݉–hy»eZ&ÀâKœ¢ST€ô}¦šñ±ð@í Í z` ÷¼êÿæ=­q®ëˆz%§ ·ëWßãF"H>¬”mDÔ¸Õ©2gcæÄ‰Ì^=ìß ö£–lo™Ï I~ Èë˜ò^S²Á (FÃ_jøÓˆË$w°m‡;äš–ñ›DÈ9î†Ùá[ZôÂy"˜V£ñqøi­AW0ÓÚ4áçŽZi;<ˆÇW$=ÚzËôõô. ©€½ÀR‹µ€h Ý=Ph?곿 Û·¼±íðC´aD2ÔÝüLÜG—ìœÐ¨QþðHRú‹uÒ¨C†°·pv¬¬ÌÑ,¬ÑƒdŸß‹ç,æ4<[ƒØ‚5²Ã›,GiøŠw 치”ET*=Z Â*dÌ[)1¤\uBÒò† XLnâÍÄs¢ !Æ1ë²A×’Ÿ_¬ÙÀ6ôKùrYyAÜõ›„H! ³ ~l=" ”×`×… ~FV4j«ÑÞÂæ@ñÕp.;õÊ€‰jäá“`‡òõì¶ó `÷¼%Pâá¯Uv¯²IR1ò‚cc ÄMþêä/'f“âêÐêË“nŒ¥Ú€µ½9qZ×?¯O¾Yô†0 Ÿ×^:C¬V ß›èÙ~:é¾5Ãk´dlìY À<;=Ž 0¬¢ ¤C Û³l’m0rNÀŠÀ©rŒ È_~Ù%n½pß\ë,‹•Š&€á+`•ÊÕQÞ [øC{à˜Ï¶ tÖ;îq:˜ÃÑMÄjOF*XŸò`?’ý *M¶†HðvÍÌFM{Õ¬ )î4Dð¸bvÁO €AbÁ¿Y/m€È£ž&w/YC-zÑìR<ö »}œ3Ù·D~_¹7#ÞÆ:Åi‘­ŠwkS0 A²ôšeÓØÓu øl1XhØï$Rk`~È´9Oè-ŠN\ð 0ªGQÒ»²ák¡WU#Î*^Y|R§>$>ÚA Hâ1ŸW‘}ØUKÑR"µ© ´ð/, PaúŽWÁ€qAþYOC º BB½­‡K\ÒCì°¿çMa „t^ÙB쑊øZÙ®+HÝÞîØ«Òþ÷½iÉŠ*v}Ö™ä²ÊtFš§á£²Ž€ †’,çÓk^C2^@^µ¥Á9 akÝXÐ\&‹ Þ5“ßW3%¨Öô^(–@¥;c“¥;GÝ7%Ôf€ÛQô•6^s¶Tn*nmDÑ ¿^D ™Š£šVVo…1Orª6;ƒU% i€SZÈÔc•˜ÛêÎÎæí^gŸ…¹ÓÔ8FIÝ=/7¢>£€Bè $–RG ïåZAnz*)†E .2>Iÿ:šˆH@޼¨ Zn×Äüç@Ì<$-L©èa­köÒèP<Àöu%¶°,,CPÆR¸û6NzŒe&²þ°vÃ|p»N~;çⱿµsXÏŒiZeSU‘U*à¼ëTj!´Bš$EnäQ6…ô”Cñ MîmÎö£jŽØåMª°ÔÖ#„£ý ® é©jTF¸ÍÌs¯íÓ1]µN2èY êÎë¶ß 2^ÍÚ½7ÓdCÙ5Ž! nä›BãB”™ž¦±êÒ@\Õ½l¯à-U] ºW $!§8ŒÏ¦@²Ðt+¬Ìù„i·’ÇÁüÛ2ª€ 1¨FéË“Ö Â¾äÑnL£«2ú¹ŒËè^¼1'ü pI™jJ$ó¶¼›}gºJ +6ÙéìÇ·™rð_CÖ©‘™þPfTÈÁR ¬Ø6µ|Ñ»Z:@ñ¦$ÌýÚ)P ,I‰§,c~.]·€‚Se×ÿQróŒq3ò ¡æ4ùëZ´™ jg\µYŸp¦ŠÕÌúÚPòª¶I~ÙÀg×ïNn!¡@²|ï,›=Š÷8I@}YR¸`,,(‘„Vbâ{ξe½ˆ2qË’nêEySLW”¼o^ËŽ/¼•ëÝÕ]_’g`|HdÉhìÊ" Òf4g.:Y^ØÓ®!ÃGýeëm*X;‹ÅìvÔøl~SA«nt(äS`úó©*«ÀWZ¯›š¸3HWÖiH×T)Mq4•–¿‡ÿ!©…4À{ÎæÜh±ÅÛ¾]TteraÞË…lxlK!Ÿ_Õ)û\ïò@Ò[.ƒh\évXuÉOÁ =gPâBúv¹àXðîz«Û®TØhÅ×ã€Í|2-É‚íkðÎhršŽfuü4hdCo´Ž[§ì¸#˜?Ò4yÁ;©BYº÷Bد1ç(µ%YØÅÔ °wºÌž &ÅrEgZe`ë ©hà$–œ G”É‚¹…8‰Pª^f«ŒˆÜó*à”^ªÊôðb†¨JNy©çþA˜Ûì ¬n*’¬³5{`L1†e¼… ³¬ÿí)ðO­µBâYMµÐ c›c *_ß­çÊ…½ƒe4jÙðÝPÍCQ~> ‘mXÝ?qÖؤívŒ`Ø þÛ(YÛ¿äðĺšÆkËEW° ¢PÕ5$îêJ·ø©)ºÞòN(º­­ˆž¾Êñ=d—”äX#ð7è„] ’¼” N…M-¾[·Ô>SŒ“¸LžtŠÛáGÂ8öIW¶åˆ‹J…Âs|ì]Ä`£›Q¤`ÏùìôÐû~ÏÄ$]瞟¦ÔŠ€¬Ï5'*å ¦Wpb DFBØq$Z•ØÌ¸„Ü$;î“Ç"ÞóEóëºSF¼ÍaFo#à±v¢ çvÆ5và¶‚8oÔQÆ1MðhŒJ„`l¶Ç"*AhvL/5eF[EG.%2®À½B²»ò“¹WHñÎÜg>ÜIÜоDœ'½ÉW>ÔŒnÁ Hy$¾aa»õ)sÔÍìL1HŠ "†ßOûN´f2RXÎ$€Ð=•Zn¯h¬$yW­çØ2‰!6D_ #FKa! X0XB¶2£”Çò¡ŽXƒðÁ¬Šv†²ã¸”ç*:T@œ¸ˆY#B´Ï8'Œ­¼pá—~9m’ ÀMœÔthÐè‡T·iò‚ªô¡iîEaÔúyqr±bqªhÑX=ˆ4kÛbE–bd®íh`R*ßiV ;ûM\I]VxJ%—²rUïéÎà» ‘)—3ãykÐE ­˜$ £¢¬%‰ŠÒõ1 LÓú!„3D‹‹{OÚd²ä2™6yÃÇ;!NÉ*k ²BG2ØbŒ×w/p9f*Ñ).Ë#v.6ìÝ3àqt±vŒfÀ:d3—Äd1\N p锺ÚÁ¬Ýñz¾uaò?ïg ´ "›Ú·ºÏÇNw‚%JI ´~{3ãoŽ£l¼'1råfˆicë0ó©)5m|ÈôPA†bÕK4û(QAÐ{*„˜]H& ©™«õÎ"®5ÒùÝšú߬iõ‘2X<–º’5[Q­kVØ¥&xŸ½jrÖË‘ÏXãÃ\?ƒðª !oð)à¡K¬œ•öm`»‹*Å·ó¼±™SÁy5Õ€jìǡӱ.¡ *¦²O¨$wµÙÞ® Ym´–‚ÃrsL°ÊŠÌÇ*‹e@…}#¤´_¼ú·§êD“8”j½H†µ¢ÖÂIù0OXKTP;¾­Ãï1^Ïv™AòSÑlÁ )4¢Š#Ó‚|P¤@¼· Û4©`XÎå"DDu]Ò‰ H4£Ö~9]N¶j¤7aÑÔ@Š}x/Ù#jZ«eó÷KOÚÀEïèøð‰JVÆ^èç±–…ôÀàPC߉…CÎa&™ß"Go8¿Å7zí&õ)ÍVœnìÚ@é”^¬äÑ.‡(©^|ê’‘s߬g“a7¦Èf\,½ÊÌ}ûXI|=+"€j•hzº»*!ì|þ-c„f¯Khÿ¡ºÌƒÃê6‰›½rÈ·vŠ:úáB"BÃ|?k“ÎfÙOÖ§²F.dÈØ¤i äo§ä'£³œÌ<ìË„S-¦íÇ*ჸE×È&??ºÈüëÈ\ttZ7ÁöB–,óŒ{†±ÒÁ9^ÚŒ*´™à\¬96(6*´e‹^ÄPé—ÊCGõ+'Ć¢UWµ §í©î–g"°ëlzF¢ÖÖãz±-q$¿‰ZP„ïCEÝ•žÊàÜÆ±ÃQ—ÂÁoÐzôÚW¢mrÓÞitŠú±(ípš,V‰åk½”Ït¨ö+‰Ú\d=ççX¦ÿÃB#¸È׿w…‘H„Xû;~Y§Iľ¼fÓÿY÷ô+Ñ6ª•Æ&‚éNûï… ™Î=ÖÕP0³Ãêu¾¹Ò“Ÿãå–éþ¶ËÞYEëÍܾð[mNŽ:jü%äƒ ›õÛÝQü–1ßÑSçå+QmÊ),ÿœ5¡x© »! Ë÷³þäLÚ˜…h0ªèTWÔœcm™XxQøÙó¡¶£ƒ9E—{æ,S׫[AàÊöØVÝR@è«SVóÙ–ÎÇ$— Rœ‰¨m¾¡ƒ…ô‡ç-¼)麈Ü/ëðº/êð}ÞÔáÕì„…šVPÇâº.ƒ®ÐÜb0ÌʵTPŸ™ úq£lK¸P_ᮺ<œ Üh»Ì»¯Úœ…J]“9‹våíð„̈–1Ñ%sƒ}Òª¸’ º `)ß qTe<ZAÒÐ¥½º#Ñsd÷QõbgŒ¤ ËÏYê/…´¬ý§lùÇd,•°/òñ³Åk¹õ³\é{k{=8µtK²Ïé¥h(žþ?9”šmG‚Ž¡&ßO^”‚À^Zðv?¶mÅÝóÓ‘¤b¸» OxsiÄ›KRé.ðÄðK3ŸWWög†å±_3˜Ý>G:ä[Ö¹mô ƒÖ  ,d¤2€•žö¢nõ<£§%ºÍOÇ6æ›Ûjއ\Û©±k¡•\‹ú`r¼°ÇªuPYn~“XÑÒÙp”Ç {!˜òEý÷}*Äçx~ÇNYõ]òS/b.+›êhM˜î}À"ɛ˅ä ~&ì-„ Ýí’™˜^òðjy\9Êž:‚íÏÆ(»Tè1X^Ã{ñ @Ì)ÿâB*¾‡úM¾¬æY4rVÉíš«¦Æ½¹;y›—:–—Úwoø;F”1Ÿ¶·3õÖaœÕ} ?—¦ÏQm0WEHœ‘ØFG/®z‹4Hj5!bGÕœ>>0¼ðÕ5ìe°;‘2íËŒ»iFX’kR‚¾¶Ý´îÉ$¹k—mjѦM¤vز÷¦bÿt°ÿþ6QÃÙUÅûWÁÍ`9©=ç²:ZV/]š%Q®Ü]¸Î„mnØÎúSR€äè°êàšÑž!a!m!Á‘ûXصp+•E6/&kj‹ek“诶ü²çF7À ˆz6^oˆx/’nOѸ¿˜?£"%yú“Ç%écw¯‹¤¶wÿÊçCD ‘OXÒ)ä}àµé4¨û =nÀ§¯Þt …LçL~£&,TþF³ª$ízI„7y`r˜Ô[ÿéé‘.;)—i¯q&ç Î¥ÖH¾/b¼•FòŽÅ×›p~ßî^qâþà®?û¶ø»]:zi¯Ê’ŒµI¼ˆ(·’ Y×Q)óçу{¢Óeêíô™F¼çó¡Úè¦!£HòÁ: 4Yâ­%Á­§92RG·Âv ®þTÛú».£v¦‹.òÚTל»U'íx½hó_‹ùÃpS¡UBÞbÊùc ‹ŽŸ×öÁÝ=V åУ®9T½5äO"õ/ì|›ýŒ“Þ@4«»¬é #ZÚ+ŠÌ¸ÖÌãC¬§ÍWja3®͉•ꌫ¥Ï.%²!ã câ/'ÿÌSvNendstream endobj 106 0 obj << /Filter /FlateDecode /Length 7708 >> stream xœÍ=KÇyFŽk!Ç\rYø4›hÇ]¯®*>ØðC ì–Ä€#.¹¢MîÐ\Róëó=ª»¾¯ºjv–¢A5g«ëñÕ÷~õß.§½¹œð¿òÿ§¯.~ü¹ —·÷ÓåíÅß. ýõ²üïé«ËŸ?Á ~Ùç)›Ë'Ï/øUs™Ìe qŸ]¸|òêbg¯žüÆ:/ÇFø{ž3ŒrsñÇݯ®¦½&›ãî>šÉÍiwwu=íCÎ1ZúÙ¤l¦¸{ SÎ!åݳ«kçÜÞ³û-üê&3MANñ Mìæàé5øsœÒî˜ø ŽHÞÚ:ÍesNiÞ¯®mÜOÆšÝ üêÍl­ «åä¼{‰¯Ù}šì”ƒIFnâpK'™';™Ýñþ ׎Á©-½À1ÃË»§W8vö)ìîyK.¥ÝÝÓÏÞz5ø¢™çÝ-o$Ã{Ÿã€)'a«øZŠÖál뮯è“5Öï^ד½¬Û~p& ¾—3@òË:³‚d…¿ØÔÿ<ù R]/˜ríÂ>Í3ã ½ä§i ÒpíÙïŽïèÙÁ àµ,»¿Á-'\p÷m"$Wp'ÃÒf÷×+øÑN¢rŽìc(P±.ö-Œ{^!q|ÿÎy–8ÉGŽ.&¹„Xxà½wƒÈCx‘fç^¼¡cLÜú64E?‹\ÓþêÚÏþ`v¿®s¶lÝ ,›œ…éø$SÑ"¿ËˆˆbÀS‰P|8Šw\Ǿ`œÛO~w KIpt˜‡0mˆY€ˆ£ìîkl’@Ìy˜}rJ¯ HáxÙ§ÜÒ‹ºÂó+D.ç`™à®óž‹pi|H/cåÚ3ð”î€õ~Õ²ÇBZÞÍY8ÛŒ°?JxŠ-ßáéÑ¢€Ð=DÜîõ²ßkãöÁgËÛ¦éWºÃi PüŒ e˸&…Ïo®V‚_è‰FиZÚ7p£.Þƒs¿}!é Ñ®>¨õîª`¨¢Ž-öÐXÄ œ9Ál˜ýî«–. ¾ox蜽âhÄèLœa¶Ñ{ëùi¹y"¦}ü¢ÒÀ®Á›_]¿G>~´+Nu¼“ÓÕ9>Elõ°b”w¿¯/"Å®ÄA£Qïs5ÇröhrÙÔ¬y/£AË(Ø,áÿäáâ‚ÿb7Ëu#MÈÒ€]è)äaÞÑO>ز”\RÃ% ð:Àª/ô´âÔîøœwRÒþpFÞÔA¯¥ÄÉõrd&¡”øäÌršzLn6 Ò󈯪ÑGúw•ã„ØA xa·?c-Ô¢¸»e¾4%W’dÊ™ff˜`,ÂÑ*-ÊLóÞ¬JÔ_a^8Ç4e‚øöñ5=†døçúx[ßÔÇc}|WÅ ÷W=q=øûd—-É›Œ¤* ÅîñïÙDæ-)œÐ¬EÖÿQ$Ó3ž^Ãñ’6­xpA= f#ß»­¢º xH(ü½•‡©ÄÎb;ãÉŸ×Ç#‰þ9'ë-eòÉx’JpoŸá‹ÒÏ€Ä#IõË«*˜¤Þ+î㦜Þh}_ê‘.ûN §jÄ‘–6 Ÿâr’&¤°‘òáGÃL95W‹sÌ&Œ•ªe±ôP©ªkK4X™çAÞÚk‚¾‡ZäÑ]‚ v?G· ÿW•\žu©ì‹.ÁÝË×zTæ>pþe™DR%G{ÓÝK«"®ÈÝÈ)XðU29Z4}#3 *˜ò x¼ÜXGðýG©Kb92…‚Rs®õì÷Ë,Jè)Ï „œ&ÆÏyE¿´;&Gñë0¡i }½Ê¹/+#‡zÃ~í†ù$ýÚ%RZ1ró%÷À€Äá ú#¢E¤cLOÕù鎓e7QÎëeÇÒ?@ÖM²0i™±ØB"PvwÚÀwÀÖª­ü¨*K¿ª¿©èj^Çî£ðü¨ËCØÏÀØVÈøƒ.olrä>Q¦XßP‘‘©O½¨v½q* #|òïyD$WÂ$=&ÉRq:TçªÀ’{>€/L¡«dáÔ_M¿Œ-ÊÁÖ0â&iüP]•mào6Dßh>!š8”Ëh—Ç1»f Ƈý Š@Û¸åÖè”)Çs6ÇÛ°÷v5O‰ï0{E+mL:Å9¢“#”<ògáu)YYJ:ŒsD¡ ÏÂú$d–¤W<\28Û‰tȪbu6êù b`ž;!Þ%4ŠgŒÌK"JmÛÆ¦'”?yŠ"~.‰e¤‹¬8vÏ»Q«œð‡ÁôBo± u>E4µÒ ^T-q”ÚÓS3…DÛ×bmIˆi$CKžnÉŸï˜c µ±Žp>uÞè`Ñ”øTOZ¼”shë—Wk8ë·œX§¬¬’[¹_±¯.õª¤'I$>Üs‚ie„q®‚Ðþ Ñ8¹Ò»ÄmÝÖ€q£ÈØ‹#F/ŸüæâÉ¿üqÑ(‡¬ÑŒ#Ë[¥6­ýÜ!sÅX`5Ï^Q>鳦Qžolš¼„®ÜƒŽ]ÑÄÙ¸®m©œ` ®}…Ù)|V”?¥ËzÒ@êÂ’1Ó‰õ©á¨GÔd¸®9æÒ1vZ¢‹=sL< ?äŸvâùªo›¥h¦ÜœñtÆû JT¾þüT[aÛ$¾N eöaÍÄ`=,wŸ×Ç?×Ç·õñY}¼ï(`àç>`ù4O­~¶ê÷Æ ý~“Ú!óû6*>–Úh¾i¤Ï kB#Îö® ÷Ìúvqt°¾ ÒhV|…ôí„'cq“à›OûÄ|à,Êš9ö $` ÞçŽýeWžÉûG¨ÊìDç«Û]ÈjŽê+¥ÊÃ|„GÖÙt, yÑ8Š+NÜé6§LÇ:IáèÇÉ è£3,ÎÒg ’gèµ%ˆZ´S'F^÷âr¦ ìTH+"û]«ùt`]Iš‰Hjss /úX™¬T¡_ †[ ’„Ò;5ÅÊe>Õ¨ÇìMÌg9)q'U§äi p»D&_ JµoÞ©ùVÚºeMfšÎÐQ)xËÒOޏiQþÅssñ(ž+˜Sl¢”ÎJ6‘ñ=¾:ÃlZálÑV°#ù'0á-f¡jÁ§h#¿{)¿ãTp#üºñ)ÁVãdN©à–Vé둎®¬ý¢.± É^0/WI“W(t¨¨¨ kʼnEÁŠL,zíIVÀ°Ûú°˜FÑË„8|*-oø…hݹ¸Z˜PŸíÆF,NþõFÍn¤¸¡3+g²hÇê)7N=Ì®aÙ×Uœ¼¼ê»fzB9ÏÂÇ­ %ƒx å ‚"#s̶ŽM$)M7-*À¢1Z”’šö«úø›úø‡úøýo싪¾ÁÓÞ<©]´ÞLrür[Çç=ƒÑÑ >dTtàª(c5EUÇKÞWÌV¹´Õ0µ¤Ý¾Ê XÕ³räÉ(¹ºá(쀯…·¶d5¥&«DS¦5èõn}6cV*}>©¿ŠüزNtÁH³ò†ß «f#U¤!:6PÑ,WWßN-–®7mŽã*:ŽÔ¿“ ›qÓuñãm0äÌ–÷/¼5Û´ƒÙ/T(jé\´¡M÷Â¥ÑõpàG¬æduVá¾çÍc½subUxG“MùÑÑeëó^iL÷ËÆfÉ…u"7<™Ãá¾€»MíSÅe—ò”¼MY¶ƒ]І¸°÷­c¹õ"Þ–Š ¹Y¥—XB±õcqž±R9¶©Ô'žkíp'ë䣄[6Sœ%¨Ëаèë«nÀ^H&fk€rá±@Fëu¢lƒf¬+ñc«ÞIÆÆœLØÈà§ÅUÈŠæÞ>FGÜÛ&J–‰Kø9«ÙáýÂXJ毉èÒá‹ÖÏG±±gš {o¨忯æñ—úYØåÚ4QcÝÖ]"ëD¡Ï–þÛôR‹eœ­„L¾ÒOºDÔ8«åÅÅI‚è¿é¾4 ”#SÑzY¦Õ×VÀn‹4TìþM¹èiùËnEÑå8]çv‰p€M” /£ gœ†LGè\9® =w„†™ƒ¦ qtîà¯HìÈ—kq€DÓŽ…ó Ü©· f§Èt%vöØÝÐz£wz7ºZí£FQ¾ö¨çû6Mþ3é°àí{ït°©„N¤JÒŒÙ4M]§ÔD:Æ®¯–GÄëY½Òæâ³—"ìÚ&ØFV˜p(.ôP Öˆ”›Px*ÒEm!&4õúxNYT4Õ@™h+ñ2U€4„ë¢Ê ‚ã(8‰ˆ€ мÜBÔJÖ¾?x_™7½7/óõ 0±Ç:ž•£ëBlÂiæyõÁpJ¼H¦iÓ”–ËfåøÙ7¿@pëº@_è,wkW/¹Å«>)¼‚I7^Ô*šYœp:¨¸Â6$Z¶"<ð%‘£Iá,x)ÜïyÎl£®!^£®d>MMOíð4Žòk†5%"j}ž}UÖÙÚW”ÏÿŒ7…%“JZ î½S&)ÆÊPÁ»•{*ÿµV` óiÁ:m -•®¿Ä ûÒí‡Ä²wÛô|/ÛÆ¹ø/¿æFoxp;½ž¥]Šd|"s†µÉIf,©tÞ•~ï—Uì(°VHÎF•¥ª­ØR#R`<›S¥'\09PC( …¥Å"do^ëlôI>°–R¯Ei®ôŸë”"‘RKmn²òÌ ŸŸV*1 Ê„¹ðÝ’ˆ/#ùg¤˜ŠZÊ$qK" XG;+oŽYac™Я¥!§<”|'¡ÔpãLÃÒÊð²â¶ ÷ÉhWg*ê:¤7^þ”BÓƒ6FÍz]Ûð.Ž@‚z´ã¼ÌÚpMhØõ18TÐ¥¼ÖÎošDd,g¾•á‰2ÿd­òU°±àöi?¾•”QÓ“´ GäÑp³êÅÇ‚ÊÛ!¥—*ö\¡Sè€ÝY ±šñ׬üaѦDMC¼e»M6FC„!÷àN`ÀÉ‹¿ê¤MÞiZ —›ÈÑ¢|î Øý•3Ú®ÃáX椪ÖÕ%ðuSZþí¿Ø¤ÊTy`JËà ÌQe„ÿøs?‰3Q·38¾Û£à.`˜ YÏݰ± YŒzG—HÏn {8p´X —fgÀÜ2Q&`½ßC˜›†9ò‹)õ–ÀÆo3ÂA½D>`rOººx ~””kŽNnu~œÑ±®!eH#o­ÇU`M¶,&&7¸•¸͘¥Ð¼[âÆ^—›ªkuMÐÏ2o¸Fÿ²‹0ýLÛ›íM_ .¯˜Ó:ÁÚEüX_máõ¾Ü[•û°à5»ð/Íþ5¡fáD!ä–вfJûŒÿ}Dþ \5œø©v€y¸,g‚K¸1¬Ì6X™Mþ7àðóå×Óå¯/l yoí¥›œÇE¯.€ã8Pê×_^^üþ‘Î:˜sŸaN›à|R­L[@Ø¢ì5|¼ÿ$ý4¥>CÄÎgÝxíòõOíä ÿ cèÄÿ0å¦/sÚ"qîR½óˆ fÔKØ{yÌŸœœpÏŠ~°Ä­µÔßá_5VXW+º|¨àö‰Èß`N€Œÿ~ B Ÿ ˜ nÕ>Œ Î~ˆàëÝ>ˆ’ ud£k ÈÂñ,6tí¨Y@¼D­0bBÜ&úÕ&æ?;Ë$g½*émÈ™}0³º£ðNEóð¾G1Fç—>~'E{‰—3^’xGª(=ÞÞ0‘ ooU6àí¨DŒî§Õ—Ó[Ç$´fÌp¥Ø,ËœT&ýÜê5h¦j½æ‡­© œPõ{,”±A¨ʧ”Ö>ÆÆ ‡¿@ô7s¤¶¡ÿ\àw€{š&7Ì¡%uöˆ¨)å©õ(/«Ê<“1G¡@rËÃ8…bòc.œ¦0~—ì%¬½†×ÓZ3ÚÅH7}$X²Ÿ°Y¬—e¤ nŽÒ7ؽfTŠr²’1»Ÿ‚løHÁ¿¾s+Éb}»mžÄl‘ g(åÕ|zR'ǶßBœ| 0OÀPBå»áÃÄ:€¡¡mjKº0¬.͇Õ;ÏßlQ;BsÂXy‚g²»ÔPÓ¦6—-~ÐŽnÅ °Š¸"„aÖ!«Æ{~Jè½jqQd& Û>ÙO蛫¢ä¤{S ®{àáÀSÉ›û³{†2–ÏþZøïª¨üâ 5…áŽg™—~ÎÛ~;ßTG¥tÒ¾«Ò~æX O£C66záøÖ-¨(›KÊDî–ª"².‚gg¹T͇ Fémœ€~>/uëfZåb[ó°¶îð{$ª <ëÝÎbµO®ÖŒêƒèÿ®ä4-çò‰H1 HnXE}àô3ûM›‹`°dC´“ïWf„úÁâd4‰•´ º1ø†ž©Å)·6À¤£• K#@~d#šãzýÀ´À{쟄Uä&hO;£“É"Û¼[íoM¥·ÙZ ÆrU_R™ƒßá®›ÝÁƒÊ|‘g3Lí¦—/‘èô»»òã¤ËO¸¤6¹¤#‹HNcQSn‚“¡Êû`x„F&Ì*¯o»7TwiZûn +°´ÿcU ,gÚ\ш-ƒE?Ê}Eùð¹~fJY|ê>·bÈBùLšºÃMV¦S/w‹öƒ¥Å~¢d1Ö,ö‡;”kq!ÛI´1YåŸöê›(ý”{y¸¶—Ç`Ý¿­Ùxæ-eðê¢ .¤Òer”-ßšT›Ë¶‹}¹q‘/h|˜vºÊÞOÙ}ØNdÁ-_C)Gݦk•É´iãÆ«É{ýêð“’zùþÈb„7%dåŠ{.ÌðÇ5f°ñ­4>R³ÏØè¼ÓìPÂû'­úÛÆWMÂàÛå5õá*žÇß9ï<ÆÒ ¶XdtÁåw4?¬Qò#E{Q&ꋈ{ î ´ëå§Ÿ´¾< ŠÃíÞ A¡4×ÅÍr=HŘ|YÒäݱOsõq‡%Ÿ}É?zÀ8¦Ê9rš»VË‚•öÛ+1 ÑÕn–öŒ`%vå!‘=;VY}zåÖ¬o.f?K‡ÙÃ.âöbxGÒ6-7S·Š÷ñ€ñ÷¸"r>(t¬­Jt·Çy~„U)œÑçÜ݇šÁ¸jœ‚íí5Ôn±Û²C7¼ß‡¿[j·ÿÔnÇÔþ½ ñ-R‰èìªAy†«Â=†!<©¶^þ¿K°ßo–p’AŸs‡Èwx%ê°/\¥Ò:ªTQÉyåke Œ*ˇmª^Ñ YëJÞct¿§êW¶e¤ŠÞó☨&§Š*Ép“¯°j)UÔáÉÇ—#Š 2‚ˉ$™nG™°^zY2¸xÈ£d¿A*nAÙ¦m~ŸÑ­¾Çûzˆ†Ú/ê£øòÑãúºa嶈1Ÿîõa3}œ½~Hóµ¢ç”åÛe­:*¹$uñS2ޓӉˆ?Ö/Td[W]v@¦ÚÔ8fNö ÇŒpGŠÆ©¸e™Ž D© _SïvP:¾ÊÅ3[(?½mk6 @o˜¼Ç`Ð _ǽï–"|š!›|Ô°¸2‘Y|Ãï…ä{ƒ2J]̽ìôãär߆(¹ñ9ß'êU?uÑ Ï{3{DàÏëãËú(4~S_ÕÇÝ4ðÁÚÚ¼^mï«6Ú%'¨x¡ý¶ÀOèÙ«jðqóñs{'¢‡pNñ‘”C7õñE}|LGE¢œú‰! Pë÷ØÙA4Àmæ$.moj]?°U2M——ït,Î,öòÀMŸùÙ²”Éqs(ÓiÊ:®*„*exɃ#µêú{ú‰4 ¸;ñEܦ÷Gj?»¶Ú{à‹¸Ì Õ~)ú“RÆc÷^P]¯ŠŒéÇQ÷Ãq :”ÆÔ¥Å£šdTOaÿ©~ãý‰Mƒ`ŒM‡ÍÅ{ñ ùsýÆØ± ìù©´ ¨(.œ!MÅeãZâewôá H;\y=Û/k+wLøÙõSí4Ô7­©˜æ¸ÜEnŠ;‹°˜Õ„ÿ(®…^|¸×:•‡º3Žš`árêj¾dä 7•KŸƒtô%Áíi‚aßÃAr€¢¥›ÝHò§wJß¶Õ¡´4wdtö(JÁýÝÅÿ–´endstream endobj 107 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 3772 >> stream xœX TTW¶}eÁ{L4O%$ïAœpˆâ°¢Ø1QQÔ(§€µ A(™”¡D(ªS…)TD( ©ƒDÓ1‰™Œ¿“8ô·5þíБ˜óðÒý[`bV'ÿ¯¬^V‹[÷³ï>ûì[ Æa£P(8¿€ÅÁ3gÙßΔ=òsÃäç•@6?*ît|ž±xáóÏÜØëâ.JpqhndÈH¼ö4šF`ØSŒR¡ˆKÒùÅíL×DD&zzo›ì9Ó×wî4ÏY>>¾ž‹bÂã5ÛÔ±žêÄÈðu"ý#Ús]Ü6Mxbª§÷Ë‘‰‰;çϘ‘œœ<]“0=.>â•ÉÓ<“5‰‘žkÃÂãµáažþq±‰ž«Õ1ážC‰Nzñ‹‹Ù™”ïË0̤E±‹ãv.‰÷OX–˜´B«NYµ- -,|{DäZͺ Ñ1ÓgΚ=?ÓÅa^`^gÆ1Ìf³Ž™Ä¬g¼™ ÌÌ&ˆYÌld^d–0Ó™f)ãÏ,c–3+˜ÙÌkÌæ%&€YÍŒ`F2£˜g&‡‰Q g¼)ŒŒ³—¹§X£h&Kö±r¢ò}‡mµޝ8ZX¶‹íスïùÅ|‡“‡SªóBç–áø,révyàjvmsE·#˜¾#âGüøÔ99ß­ÿØÐÇ&¿V¡·ôû º£ú· ôyºTbøi¬n7èãòãó÷Að©*ö˜é}¨‡Z¸Uú·]ÆÙ¸$‡¸’‚"0™ö÷àè±ha/ÇUl“é>´Aü5ƒK§qä–¼[@žt:’=¬[¿#Ødg«•¨À-¨PÊuxQø.ô,æ¹P#pKä8ª¸ !L"Ÿù«¸*ÓPGãççû£†C曞[^Iíï-áÜú;h™o[±Ê¦À 8ÿŸS¶÷/Tî{©ä`}¯Ý¾÷ݼO‰k‘„ãkë®ÃiþîÜoÈ‘l  X…+¸–—†ù'‰L&7|C8t*#Ì–Ušù„“Ü䋹­ò³VE#ºàAtQʵ' ÇŒïÉL2wÎ82šŒ¹û"ÎĹ·¾Ç‘"1‘5 ètý4¼]Þ*Uvœ†L2ú|*©1OªÍ³æî‡ŽìŽxPó¯ø€6xÕ‘ûqbn…Á˜¼ò%Âq) +-4Á‘ ñXùÉÈwà(¸ŸÇ§¢û™]Öä£Rh³Æ¼¡xeÉë%p†ÿè6XPi^oà ð(¨’Üñ™6 ±É#mŠÐKž…c•ýY˜&|ecôž«a¬6Ñ’Uè5-Äm‹OΈÎRÃZð>p%ìdŸ]àAÙg'{>¼Ú[|®ºªÈSv‘ýAU`ª–0–«‚‚$q¡ßº3Ë«;É!Óɲ„âd2õ—¿.í锚£ÞM(þ ¡±‚vgÃW+?PÚºv)1Bž àxëÔ¬Á;ÄÛ;Ÿßq4r^׿øòÄ…/Åu·"&6úu¸\e§>ÅŸï~¢øP^-ü̆`îÕ-«,ÜWs^Ä{ñ“,nÒåØŸª¿„k6Ñ8ŸÅEfG ƒvIvàÜä—³{… ëúÇè`c#ô“@McD áå0pŸ8¨¸VS/tÑè…Ö!êØw¸“g‘›BΑ݄ÂÒý.oã"ôz… ‡µƒ;pµ•y:ò²$âDЉ€ÍŽçml¢~)] k`×à²K.Y”bÊ«÷#pÀTTA›Å–kÅÐZœcU4ÓêCí ©ÂsÂÝůuÄ9o®º)µ¶¡ÞÒV®;šV"Z‹ª¡øO:Ã}¥m™Efo%Jä“?ýËÙ®ÒvPýCìf- Ò’ {¤UF8UðÁ©8õäòÒtÈÝmÈÏÈ•v½¶BBWL0 Ç©ÞÁÔil·´–7µTÚà'(".ƼÞä?&PwâØwÛ>«bìºÅÚÈðlßÇG¬%&ƒÞ,¦ge§–ßÖ”z¬²µôDgD×¢ '¼8a‘êKòè÷g•|‘lnÂYw¯6aF³¢ñ:&<4_Wâh|M8³ë”ÒY.žÿËgݱ­™ÅRý‘êýe¦ü<êÀ2xíá”cÇ—[ªS›ÔªPmªV ­+ ¦åˆë_Y]Ú¥‘ÒRR#!TUÛ-»ra-øÑr\…óo¾sõokêãŠ!Õ+a6uy[!Ϩ*Ìl£²qÀXZrˆÇáfa|zê|zëøoÞ þs¤ $D¸òYFPdäjÆF\hûð”J?û7ûéÿŽ…›¸iIŒ}zÿýµ yïß·pßdZµ6†ÿEMV«õ x‡ï}á& ‡:žÎü>h¥ñ#ÔõçFÅál4eVîjÐPS»œFh·æ\äuC%ð?¼s¸,§ #Å Ë0Hy£µq°t—sîò£Òs{t_ÆJõ¡B NØJ7¹¾ë sÈŸ$7üˆ’FîþŽ·“F‰ ýJ¡sGcXh´&<¼QÓÙnmìɲ‚öÇÅ_y™O©®ƒó#@žJ‹uÌ´õsŠîà•;J,”W Þ•ù_Àa¾Ûzñþý[o’1±@ †òÇ.9»+Ñå²²EÕ¦xËÆö)Tiƽ:“(–·¾yxt.¨.·/±7é ¾R{(£>‚y•öeâ0›H 81G4…‚äŸIà31‹ŠÅ’’¢C55ïm¾ža7™>?|øßwˆ¢†àÇL›,i±üÂ)e˜Ü/@¥ÑxlÿQS…ñ(5öa°âh,€ÄA-®âj©îTç×ï3 ‰xš#ã&®!ÎÄo>ûž(Ï{hãb©o‰ áù‹~ÏäÈØ‡dnÆ€›ÈQ÷<ˆ¶|¿{ê˜þׄßÒìs l-|þ„CüêÆ±uèWÙÿWJÙ»açˆÃÄ7—EíwÿuYéJ˜ AÚþ_A×Î]:^™´àÿàïXÂÙK¢Ü¹‰ã•˜òÅ uÂvD‡†Z£Û©Cuƒ¦_`;Q¡À=è¨Äïû9¡^Q⢕Di¿Ÿ¥¨ÊÙè1wâ*‰¤âºL?Âi¡kȸÓlاÏÍìãÓÌt  Á>¿@.ó¿Åœô8üÞ9ÈˆyÜ@%æÐc°EÖD…îŒŒŠ¬ÛÑÑÒT×.bìè–8ë¶È¨¸0UK|SgmK“8¨˜øèÜt±Øï@F©Û+¯ö‘ñ½µ=ÎÅ%­Ü Fge$À.~{mB³µ®ªíü–÷&–LÓWqâ“Áò/üÆG¸”«1"%rõ÷þ/OêüçÙ+8âèã ìIôOúƒIüiUFôQ}b PÏBÉš‘ôÿföoçúO§l¡—»}µù›²&}j­“®¥¦R{ ¥|/?ƒsË<(áÖŠâƒ,Ù\ÈÙœQ1\tv˜[áâÔ\ä₊*W£ý‹˜ÿendstream endobj 108 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 7713 >> stream xœµzt×ÖõÙš Å€• 6‘) ÐK¡†Þ{1î ÷&Üma[¶¤#ɽY¶Üä Ølj  @(¡½„„„@î8×ï{ß•d'yëË{ÉÿÖûki!iîÌ=ûì³Ï>WPV½(@ Z´zÃäI¦ÿMæ‡ø¡½ø7…îXû‹¼3ÑúMªÿÂùo¾~WÆÙ@_!ôµj:ê¨-¿~ :Øy ¬‚À°ÄÔEAÁQ¡»½}ÂF»q˜<{öÌñS&Mší° À3t·»k Ãj×pÏ×pòÆßacûnÏð(‡Ñs|ÂÃß}爈ˆ‰®aƒB½ß3Þ!bw¸ÃÏ0ÏÐ=žK‚ÃÖ¸x:˜7:Ñüº(( Xîê°:ÈÃ34p“g¸+EQ«. Z¼8äƒÐ%aK×I—ïYáº2ÒmU”ûêh5žk½ÖyûlؽÑw“ßfÿ-['ÄLœ4yÊÔiӇϘ9kö»ï͵å8z̸„¾ã{õ±¡¨áÔZjµŽI­§Þ¢6PoS©QÔ&j4µ™Cm¡ÆR[©qÔ6j!5žÚN-¢&P;¨ÅÔDj'õõµ„šD-¥&S˨åÔTj5ZIM§VQ3¨ÕÔLj 5‹²¡‚¨þÔ@Ê–S¯S,õA ¢ì({j05„JqTÅP»©×¨ÞT_*‰  ú T‚Þ‚>Ô’;ÊŠR  ÿÓ+FØO(òVV÷¬×YÿC¤=¡ÕôwÌ|æÐkƒ^û¼whŸÑ}þÖ7¬ï'ýÜû·ñ´i↑ÿÈðþ¶[mį^od²ò7F½axãþ ívÃí¾²{eßÛ~´½Áþù`!ÖCœ‡Zðæ˜7µœ×À}-Y(Iv˜è3Ìz˜Ë°¶áÖÃßþhÄ„E#]ßræ6‰`DÛür ®sÞ¹ŠMÉWfÆ@ (ä)qؽ뻨mî±›UŒ?]®>¨i#t(ëTÇúŽ0ÒÛC •ZuŽZ#9‚Y#ÇÔ©9¤Úl„`ΟnÒ^†}pލZ•¦%QôZˆË”Õ3XÏç°Èß´ÆR‘ ßFÞºUª·E½.!ÏKƒÄ‡Q/ôŒÅe´“ –‘5ªÛ5UPIžÝ`yö2– UôOWN]8Ÿ»y‡ãþåµ[!Tµ ÚHÃ}0ú7ÅûêÖÃBpt Z͈o¼¤ñ0™µ?->|9+y‡Ä¦sAå¢í7Ú"É]tæÙÂçƒÄ< å{±HE×k2Û84@ô¬qÆ–•;fb¡äq"û¬ôÒ¸Æ=ImPwh*¡ Ú•¿'À:›NiÝü0¨ÏOß!‰Ç~­$âgsÁÕ9BÊhq!‹Âèšö£5íÀ\=7[ã~Ì\àî©k –Ä•tŽ40v ŒÄÝçÅÏ…óXÿ º'‡=ì%Bs‡F£uh‚&ãyüúßXsù}j<} —¤í‹®Ëžs #×zïˆÞ°þm;_ÑùXC„ζá.*"\ÿ’pŽÅKÿ}¤üi)í9ÉZü¬:Þíü{C03|,~|ö6bN—¡å ¾†Ñ"£Á/*HçìÌÂ5ú\¾{ívë´-„Ýs»ÏQ}ƒ }ˆvÐ8ƒÖ DCo²=q^µ¢\ëPQ(ž¾ÏÛwU÷ˆ~–ë7Cb@%¡¢é2ÿ±Üd”,²”+B jS–0Q’Zwí©~HO¿ºøêþ ñ ôúŠE½h,1eÛ\wâÿ•û~u›¦@P˜0ØC/Pæ6IÐDR&¿ÈÔ€JÃ)T •ÅêK yuuþ5.œøÆBð^çB2ê@Âl3¢#–07Ѓ¿Õ9‡í:òky–õ¤VÔ¨½I §C“ù¾ôxÈo” Õ}úÛœž€gîõÇMGÝ÷¡±`ØX„9° ÏÉôüŒRÛzCè‡(Êþá ñCä¶°úÔ;1°ˆÙæã>w¦÷é¯"9E¾*m/0ñ%Ávt4$禫ե¥œFš’ÒC®‡Ue„Îô‘O??P›'ñ­óÈrÍ"Jµ1wYþ™Á•-÷Q¯ôÉþN˜ `ò!½Tòƒ¹ìR”Ï)R÷&¦*<«Ý –dÏÆsÙFŸ¢ *©Dü°.¤ZþyÁÇ 5&êP¨‘ˬ3ú4ÇòγùÒî¹ÀdCV¡-¤÷§¥Âi/Õ&e¸Ã ðÖ„üëèGZ™—d%-nÃVXœ´ïhDíD¶×Ÿýùñd<êßÊ‚_Ýæ£Á¯Á3ðìŠ]Ñ4<mˆc¯ÂíʆH­ØÉ h”Ñ,à§Kh0aé~$‹Fñ´*T¤šà0&•‘~…khÜÿiBÓ•ê+¸Ó¡›hâj¼–ÁãbÎÒâ>6¢£™…|'ÿ#ۤɽÁUE>ŠD=a>ø˜£ê.z;P<ý'Wý& „?×H£ÛסÑÌŸ0ú}3a Q.ü³«ˆ ÊôÒVR ¿ö¿H~È-v2FæZé’2’ò@ˆÆ£~-¼„ž¶s险múªû›› ˜"Ácé¹0 çÄ‘§àPuÍÇŒ ú{·¾ }(DRSüÚÜ[\¥)²9$.˜ÓY%}KÛ™B6‡ 1íùÙñ~¸Õ½çzŽÂ¼ç Y?Ø((åã„ü4tŸÍ./>xQÍépe *Â`—&Zmº'Ý¬Ê ‡HHT¦¦ÊÞÂYvØ ÕÉó q5ö•û¡„+@µQÁà®И—œ‡4y¾?‰‘]Fœ&%ò!-;-‡°"Gª?ƒ¬ø=FA ²Bcî ÑI4š}gûÆ-ò¥pÆ›ëÈ«­l ÒIwÇ9/¸²öáW/_Ï•¨³Èã~ÝRçÉ•ž*Gp…©ài¼‰þâ}ˆ´l2}´ùé+M~rDoÛzcç4óÒUÒÞŽ£û_°iîU®mÀ”+øRr*t#½>T.óM=œÎ!GZa¿w³ÇþíE;™½Äqy€.º²ºDW™—R¿K-©j<š[ÌÑÓS$^´øàñmŠ«æïÝ ÎÌ»ÏBÏsä!áÈÒ6&ÍbGÑãívÚáÖxìdë]4;ƒ”bÉo$äGò3Y´ -ÈÊ:Þþ¥ÚÞH‡(ü¼aì6'„> Ê‡DˆÙ»7)_ÄvèÊ¡ ш!Þr9.ËHÌŽ‡ûˆ0™F³Sm„}ÝN•Þ {Ò™Z¶©uy¼Zˆ ¶vß¾£•Ʀ–Â6-B”ª’ä-šx -Ê”YñÄ?ÇìO‘¯žk7÷‡Ä’â {(,ÌÖ™W¨–+c`8ªƒ-I¾ jUÛÚï1k7÷Ù¹~rØûÓ•£ºj UYañRmr>‚¾¦þÜç_Œ²Ë’åÇò%&Æ›Üy­ú”FOlÃ~UmOQšðÆ,.„Ó#ôÈ«µÇlųÕE—»Ý–YPºžhµ:« ª /lÝ~mH¥ó=ež .Ž1QÄR±È†RPçí?u2«CSH…g¥›v¸—rm ÓOßÍàÞaÍÄei¡„ôk¢Åî:ž³hñW¦†ŒrÙP‘ì­8bkÞY|_K7@Á7\½Qä­˜DnéAœÔ¿È4_"2Ù*W}É=^Jä螟Jʶò|»º~U£(pÔH-ù1ªŠ¢É@°W¥LIx«ì°éSs”i¤ +›¡”³$(ú÷jô!‰µÀíØ¥ÄÇùÊ㥛‚–¬„ ²VbLËÒCSY!pjõ>v¾åôé Ά_G:Xý“fƒÉém}g0¯¡@ͼÓz<û¼Ê»‘sSÈ ž )‰¨®Ô×~6oÿûxà;˜Â8ñKüú÷£ ]ס¾99{AAT®ä'®HðFÌo™zM—Øü2º»®ªùŠ/…HÞ9†MÔ+Èæ<÷W¯£ ƒÁq¤y<:ª©‡z8Úã/GB[•„oRÓHþ÷^éñ鄯ö¹ ÍJËEªÎ~vê®+¢žªjÖ~FZÁ~øš_U|EOeâ¹¢Urê:kGÿº^Ðj@“læßb³sM‚Çä'A‡¯ŠdxºµA”ñci3êELH„ëºú%Ê@ )öq™$讨­%n4÷ÞãŒToQOÇ´7“ºÛGÇòN²Ëi4ê·žy”ÿé÷Çþt`“KåfÓ=†Á,7™˜}7êTf…]iB¥,:`§“ÔÅDàʰ}!-ŠÓÄ6kÎÖ.3î«?A„¢5Öè”MöŘM†EÙªùÙÕBÛ9…MÌS©#ɃÞ÷¯‚NU¨*W¨UÉt•þI&Šéôµ*MÎïzf— VeS™¥þ(]m/C}T;”¾à ;4>R¿Ç`I×<Ú¹€2´PoKæÇÄ‹¡fÒY¡t6¹ghæEâ!Á‰{XT'Òk¡ìûÍÏHõ1ïŒÄ%„sƒ_Ž%ƒÍa$,ÍIÒ¤$*åIJÎgädØ ÛÁ½6¼9àC¸M$êieçÌJqÎ/„žè ‹\ð82B®ÀKñ< »[7OBKÐ 4MENþ ͎ÿF(½ýÙÝhÄbœKð›3ߑؠ3f±EÓîeéÈêâ“B´ˆ_Ä"†þúÃi !¹KHЇP&¼(¢ª¦PWÑìU¿åƒ[†q˜žã ßø³â:iË´¯7 }QÛ !Š@—X4W„ú"êËçßþ¿)Á¯~7ÊÿIkQ b•[†À‡"4†ípRÈSM=v+GÎÜ+GPn[Ýq-¼ÞØá|føg EóÙið¼¼J]Whdå•WæŒ VùņJ’ÂƒÜ ‘Ò ¼X·?¨EQ ÌÓK—n4Ç4‡–KššÒKHU¥ƒV%KM‘AŸP”SšY^œPçá,wså\ë]µR`&.Yò¾³ÞÓ°Gå >Œ¸“ÿ|OcÄÚh?pg}¿ Ù Þ?µßlŒkßVÍm©^kˆ^º@²Ú/mH™ŒÄ—Ø·—´?²¯ùVò‘è>²‚é›o#±å˜ËR`ñ²BÔë öÿß)ÎLp^íÉ ¿¯s›Îp©žä÷¸umÉPTÈzÞ‹o+‰Aáè‹òh(‡<½6GS:`!«TüÖ’°wñhÉóxöYÅÍsp“ù‹îáQþôHФ½bVÂ+Ýc¿È@4-þ{qS»÷yƒ?,ƒàÞâÙâ}R^ ÌÍâ;Õ/ƒÄd•D¾Ø'*v@R{ÜÓøÛvQvÔo*%n½Ö<-£~¯l¯¿úàâ ñSô¿†Å¦?É–!Ó÷ÀT§>¡© rÔØÓòÔôýñiü>¶.´,È_¬1VÊêHGê;'UÛžª ¹€>»°…rýÈÛ±øuãêBpÆþò‰s7Ñ{z<Ë)ƒSËAžÛ=^¢]æù-Y¥JHæ|6Ô¸µL!ó›pæÔa³Ž/¼,ÉMmO¸͈ŸV%×&W{•…ä…bo_43p^úÑ-ÜúŠ ª&UN(z†_Góð›©Öäfr Öjëw_ð~HúÀk·¿C½$â«ð|ù™:ÉÒዾæĄ̃¸?éx¹â¢)™^&\Æ›p¹‘mò0´Sâô•)áÂ¦é  Š©ŒÒEE„l=túúÇ“¾Ú9Ѫ&¬"00,,0°"¬¦¦¢¢†ØS/‚úì2´ YÐd@YQ–AØ9±s6Û…D‘XJºTš^XÆA ¡e޲,dÎN솸᜻¿¨Eû#!ÿ^@‹™îôp(h0uG±‚’´\BèR 7(ô-ÀÓ͇s‰BZ¯. _ýZ(õD»j ¨(+”ô…“7†µn„(ûÈèÔHr]µú#Mñˆ5ªËu^ ÏãXŒ–Ú©4)Äw¥]9›“u¾½USfòžŠx".{µ&ÎÒ8JU92ˆ‡ˆ¸ⲇa;>üŸ+)‘0fŰuM†-†ÅÃǾ7Þ÷6  ?ðOMy?C &·@ô£§HTÌiu t•®‡Ã¦~æ­Ú©Ü >°Sãý»~f.„NQáE!ú sû{»ZLÿ Öq ¤6ÿNUL;/ëFaçNa§€¿Ã4WÕ|fÁ_én¶ìë5!–§u¨ÊBˆ]JRÈ•r¼²kÞÈkäæ µæ4L;T,!¡í—ž~bò„ž(¶ë»ÌÈtE>ƒ&=½€´Åkÿ|˜mžçþ7ϳÿ[ª÷`A®Ó¯çà_þz~Ô’ú_ò,–ü&üe°äôÁœý&O ôW‘±Vöø©JçK&|¹"‰8¶ËÍOããUiÊ 2 מJ3€KÍìpѸYÖœ†2ÈóÖ`‡%] ðt^¦Ò§£ý—×_‡´”´·|Jã>Z{F‹J„¨­eÏà’õÄ)1À§ ùT”Åê»R¥´‰e¯÷_ Ѧãìÿ•^!ÕwÚéû»BÞ]eávÊmŸk»¾]ä kaaˆÛD¿%)sá=˜Ÿ6¹eÞÁw¯F…ðEÙ¡oê.¥ß„[ öÆWYXgˆúVvžÀGp.f(E}®gˆ¼U::g#̃U¤x–ÊVÆŽßàd:.V*5é(øšµu¾ÎV€&œ[½r*øútäV•LW!+a2çg2'ˆç9ºÍ‰=Y‘Ûl² p½|ùƒÑöØKäüØb Å7ØW/Ð$Ä=Ïí‘hòmlƒ­¼3eã1Ð7æµTFUú$)A¥àÊ?9Ñ|˜ÇßýÞö9›VKð&ì+3ùô0{ñKžU£…Ĥ§$;ïMNZNtÈ´iõ2ÀŒ,G½øØ`ÛŠ¬füˆÞ{2ïÅ 1¦ÐAô’½Suæ2AèÎÔSãÞž·vž¯>²Æ¨Ó×s„XÎxôb~0-'¥“”ëCœ%;w+BUIª0e2$©’U°—wÉd™PÌÝi^:]èì2¿ø#IcŠQû˜¦`]`˜_œó¤§Ë‘Ù<~ü#G wVÝQf 6ò®á¶a4›fêñ ¸=&=ï }¼Žwëy·ƒ|ðÜ8ŠßI°ÏˆIT˜N3ñٮݤnÁÈb…)ø×¬Í¡£^ß¡Uz4è®í±ïwQäCb÷øÍh‹úÌ~Ž©.1>ÞÚJëäõÉu¤éõ¿ü8?C‘ž¬R‚"E‘¸#b¸Ã®LDz$­’ Íì…¤ >JGARA¦F›‘Æå5Ÿø:À°++ªÈKãÛñ/DH׆¹†yx…ì"Õ½áxôGZR[¤‘×**¢ ¾q>É»¦_‹¬$âŸóã·Hl"/)Ûƒ:†½²­G}¥úk¦—AâÿAøûìÙ°Ã.>QÁ!!ÅÁMº‚œ b 4jµ Ä¥ú%/ZµZ’@v¨`’ÓSÓsï~zshËÿÉ23rüðò»ŸŸ—™qòI>ÝE²˜9X õŒ‡øb ¼EètäÎáÚi8©=™u"­¼àþ¡-•‚>=ŠIÌÇ¡ÊCáÇL¿‰Íäw±»jS ˜9÷Ië•«¶.Úà±s­§e?o>Ú™'SNOõQ§VG.7¡9Ô@\Ø‹õÓ7Ž<ìÇe/‘õ7wn’ölå"ËbrV`F¹²¨ÿÜêȶXc¿`þÖÙÓ}|÷LÝ…¯Ú,ð7=¹WÎ4Á¿â òxd¢­]aýh?b×úèÍš†`îPVM±1¬Pè±ì„ÛWˆ%cÃPäGÐGzL6eÓŽ('ÎÍ)Üæ3xà×c'þù´´–Ÿað€÷ÙØþ磚á˜ô²­¬¹êؾüFb êw;åù¥­‚]̬Ýc§rf‰FAÝ*ÍÛü_™Æ7ÿÀÑ¡Z¼¹ Í7­´½ÿ*’°ù‡?^-~Ù%d‰1š»rArÒo„üü6bj”×6÷p8]j5—©ÎÕdešæ_dM< 8Õ9AعÝf KA§‹„8I×·¢¸Hˆˆ(…B ÿ ÞÂX¾‰•týMkù¦ÀüK¯^f„+¨ÝˆÒM¿Õ|þÍuÇëƒÄÎ:ÏÏdƒ3ÌõEŸŽ~Ê2‰ œëó· ·T}Q“¹Äy”˜ô‰?I#Ñí\]ËIÿ©+}ÖLIâUÉ`ÒÈv“N'e«PŸ½eG0˜]áë™®)Žâr5¥9PÁèc "¢#BÝ›½ÏÜxÕ”+g >÷C}1{Ü4’üª÷-ÂÀ#·ÿPLRý†:"t¶h¼^Šú^3½˜´ÆÀod+£«¼#âöÊåœJ©R©€QBAZ]æçç?–ä™ OËd¦¤¥ÈfÍý×V8l,©ª4iͲ̆·ÃÚ²_6êߟ¢{8™%†Y]Œ"[?´Ð@ ÷¡…¨?˜ß}“T­|ïä¥X¼ŽÛ†{%`!ñRssq¿“Øú,îE0iÚLÉo·F1äÞ­HËÂMê· Y¯Eý§²yjª\ÊÒ(Éw£Oã©€=Ïó‹ñloMž–š‘óè3$þˆkG½òî™x7« >€Žè¼õ!?E›Žr zȇëP½¨J#" 2ŽÃõ¸žŽë¡"ïDq©ÓAi9Ÿ ^<ú\ˆœù°•ARiP^Z]««¬àÐæ7ª¥ú €ˆ°0]Hmµ^_Íuÿ¹†Ñÿ ¢6™h‹V=1cÇÐ]~${$3ÞM‚Ëÿ½;ç…4¶MŒQ™2e/¾«R*!nî¬Eßß?ðqmÓÞp禔A ã[_Z^^TûѦÖqŸmXÀaÑ?ôüñ#Ð Ú¢&Ú–Ú¾f¡­³š½ÖâÞÃÚ¢¼Œ¬N£±V«Uq+ƒÖ8;Jär²«T3cónÝB´é§ i9:õ:Y.8õì´5œ~ö×O¼³£ñ›÷Q:ÊA’ßþ€†LÂ8 ›5A‚–ð“ÙÐÝølC(ih@gw›¦ÂÛ–©0]O¦Bäuñé!¿ÉôÃ'm$1¢‡¶þã°&Ÿ†ëœ=SS¸ÀæéÁÀŒÇÌVLIþ"^ÈëOOÏ·^±Èš~v§õ“£Õ‘ æTt´2Â4TÊÕ¡U‰d¼L„Ót÷dfÉ“z3K¼{ˆÒ.ãú`ЇbGjìLµߣþ*S¦npâÂ]uÛ™€_#´ùË0xÿj-'^Ï—lâùƒýU4¦UqÊÄ0ˆi^LI c#Õó‹ŠQ`n†^„ÝòhcïK}¸ÞV3u}_Óg÷í{I×·Eý/V¼œëendstream endobj 109 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 2189 >> stream xœmU Tgž23ZŒÊ8">&Q«èŠˆ–­Õ¶>@ô,«¢TjP(°)  (`D"InÂ#‚µ&øE’ò4ŒXŸkëºívÕmí©Û=mõ¬ö¸­Ç]{‡ýÝíñØížÝ3çÌü3sÿ¹÷~; *0€R(LL|ô¦ù‘ÃËHi’Bš MQ‚iè÷CkTS¨ÑÑ˦ŒûÒ¤VC‚»'0£k,GcÚJ©PäUÅäm/-ÈÊÔ´³Òfkç/Z´0\» 2r‘vyNFAVZj®6>Õ ÏÈI5È7okòÒ²2 ¥ÚY¯è †í‹çÍ+))‰HÍ)ŒÈ+È\2;\[’eÐk×gfg¤kWæå´kRs2´OëŒxz‰ÉËÙ^dÈ(ÐÆç¥gäRÅ-ÏÍ+(4¥ßJËJx{þ‚_.\LQӨשµTµÚH%RÑTC­ â¨(*ž ¦8j<5šD“‘ ©2ê{ůÞ€¤€w~T:#©Tª$ú9zm£ÿɸX–].YÔCC âLQŠs)z‡Ž+¥Ü¡(Þtlì€<¨ÞSUJìO¾›`.³ìY l±Žî²Ÿqœ„n¸ïYÎMk4ÛkívÍY¥B7ý™© —k¾…~èoÀãüCîKÙ<Ò䤊iõP°œ÷áÅLo0Òò‘‚tç“O®®Ô1mö]rž+Ö6Ûðö•˜Á â«Áßvº‹— äðÿ ÙÏü}cwØœÄõ›òîR Y¶G¥c8ßU'¤kÔÒ sÄy88N)uc.ç=$ ÉK‘3H(óM¾Œ¯|ýG dYÍúÚïìÔz¯Ï=}з³Uï΂ °™BNPK9r3Í8}ÃÅàÇH=Fõp/[ñ.*†,®Á_¿3Ýç€.¸^?0QÃp·þpÕß 6[éNÈeßê(ioïp÷œßØþ†Àùb!-ɰšUKIf¯4©#¸«/õ:î’O!r“™(ò¿vÆ5È^½‡>íÊÝP'ØËì{ÝÀºÃš/˜CÐPZmcºß¬¯O6‡D IsmwíÐôè»Ì·ÊYîŽ×üiÅÚ‰ËæAÑÔü›ý»„½-{¡ØB04DÁ¡ò]§¸;´¨q dŸ°'1pf‰Gߦán¥·g׿Ú,Wš`îÁE\êR<’9)*ISyrœÑ~]tý3ßå›ÂyÝ&fUvaÞ:¸vXÀvYd&&è[Ø›³fYØLFˆÙ›y^–ÈÏb”ânð¯2[¯™ø*èÖp>I¦ùµÿ@œÀ,Ú”½Ärô‚€÷ÿëM9£½]ôÀÁ ø\®ò’Ù›#âë'p®7¸G,Á±5œ¦Xy‹’ÿ&œ¡]ûáÈá ¨Ð¬±C3f1›¹³âwdöz¢®~a«Çxüd‡»ï„ñ Å!´ÕwØ÷û©'u©&‹áÄâ¬$3&ÎÁ1¥Ÿ\ñ ö¶ö±;p>ÎÌ™WQ°)Ís)\Ðr¶FÖÐ#YCDE«ô¿C#p|sK½óK½õ Û^(€$Gž]VŽiýÖýUÕf°VW/œ;a62ª‘η­¶fC6lqló‡]g0’¬&!$l‡16C¨¬?Ç9{…ßX]~Û0¥Æºz;Øû4Pd¼*]G³è8ೞòG­xJHŽ[šâÅuÃ2§±Ç…pe¸MšÁãCb‚úÉ‘¼¤â÷ìÜxgÎ~t5:÷>™¡áÊ¢!)·°„­'y¬eŽ{δ_Õ@FÅê`C~º&÷­dS6ÄBú…Ýd–öɮݣè’]c•øDšÃ—î´C%û ÿ¯[¿Ž¬ZNÂ’Ðûá/ž@ÖÙR»KmÖŠ½BþêÍÆMÀ `ØÀ%ÀñÈ{Z,Í5Þð¢;ü¹ ÖH¤B¸{hy=ÜÑ•Ÿœ{Ò>èèƒð¾ÕëÇ$žá6Jó§3;SSôú”¾¿ÏÛyZ¡2‰Ò¿úƒ¯ù¶<ÀO,ñ…p7¥™ÒžÌ8™ÖÐ †ö·]ýþ‹‡Ydòa¡¶¬ÏÜŠ¯ùíºÇfÛm¶l]7× ³FÂcæL‹;•ì.Ò ¾ÙY…ŠB–»÷,ºÖâeGó!™Ýœ¿rúRîÂÙ{Û»PWô̹q~ç68Î}ÂÀÀGoÞ3÷B(.ºŠª‡î& rñíÈC¬éÎE²H)Ñ0ÇÿœÏ³r÷òäï—Gú€µóé¸eä*o2‹§] Ü…Ï)ñ‡!%ß ù&+TU KÖ?ÅÀ–è\ô;pÑ~åû¤…htÌéšÇ0(á´­ÿK²ù.ãÔ.QqŸÇ•((¥,àu´¥¤òµò*“yì”ù £Åîsý¾SHÿùšß²ÈLÿ#™JFÍŸ¿öLÙ‘ŽžÖ³çRß)«Út5yýj`Õ²•«bWhH2I­ØV( -•–§4õˆ“ñyï]|'¾t·á´î9q2Ì+÷ȬÜDK¶NÀ LÛÞ&k§,Ó™AÕQ§µi—ÅÕ&Íú¢Ds ,…ÄCù •56»Ør0íÔS ­ÍŽÚ}N¡{àBc/°wà…—5Ü?¢!Œ®®»œ*xíG —=Ÿr °ìÊ5/~þ"ޏ0ê¯wQ=ü¿2¹¥z«¸&þI¾ä“îñMàKRh°ºòÚâZ[ƒµ‘ÕÑzâT‰tãgÍÍµŽ¦úÛpP®öUzCÜC&™{ êßë>8SsÆ)6|ì¾âéõ“‘8#·1R‰»¥¾_ïIy¦é~§_ ½ÿóLmj’Ö6àwcM’Œ8éç„‘ ]A#zœAAH·²©)êßÀrŠ„endstream endobj 110 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 5999 >> stream xœX\×öžuavTe53$±ÅޣĨ`K,(Vé¬. ,KKé°KAƒ±+–ˆe-IÖúb¢1*–$Æ’˜œI.ïå—¶øâ{ïÏÁ»s¹åœï|ßwFD™ô D"‘dîR'§‰ ÿ(  CzïˆwâPAû‡£é;”¹ýœw|%ï‹ÌÄÈ̤nÏZÓúC½9líG‰E"YPÜ\ÙöP¹·§—Âf¤û6gÌøpŒÍ¤ fØØù{ȽÝÝl–º)¼<üÝdàg³Jæîí¡µ9ÓK¡Øn;~¼R©çæ8N&÷œõÁ¥·ÂËf¥G ‡<Øc«ÍY€Âf™›¿‡MûAǵÿš+óߤðÛ,•mõP5ß.`­½lÝÜíóæË.T9+ÝB¶, uÛºÌcÛrO¯•Þ«|œ|Wûùç:~ÂÄðI“#¦Löáô÷fØùñ¬‘Ì5†¢Þ£>¡fPïSË)[j(µ‚F §VQ#('j$µšZC¢œ©µ”=µŽšK¥ÖSó¨ Ô|j<µ€ZHM¤QÔdÊ‘šB-¦¦RK¨iÔRjÕ—2§æPý¨þ”µ‰’R(WŠ¥R–”eM‰©AÔ`Ê”z‡â(žêIõ¢Þ¥zS}¨$A” Jý&R‰^÷ø°G¹Ø\/¾mâlrÂt¬éz ]'$q—œaÖ2{ô¼Ókl¯#½ûôöïýÀl•YCŸþ}ú:ö=mîi~¶Ÿ¬ßƒþ³úŸµj±MÚO:UºF*Ý'ý×€e‚|Ï*ØoÎxÞr“å~«‘VQÖ}¬2ä4h× ƒÇVþkHÙ‡BY_¡ é~ÕþލZðƒ^¸ÃÆìS§ø¡¤B1¸O«ÖJ(>)*>>E0þU¨˜{E—T ºÝJäÉ—ÿj Óèò÷M=•ÈÏ¿•ð¯èâ*T—Ÿ°+2ƒO BáˆÁs… 6…¾‚õ¦ØŠnÛ.êœtÏôष”~ ¶Â¶ºsö´JNªBü|ÑÕ8‚ÉtYw·©xW˜OK¼¸¹ïäéÙ'Vn•´Í«6ÌÛ}^8ýrí¡™œôëEÈÅ;$„iÛ[PëDû1†oX0…i]hø›(œŸ—#¿rTÎÏÃÃgÁpSM9Ú]'o´ƒ<ªB…|9 f8[ãbÓ ZHö®$ŸþÚ¾L RðO`Ís¼ÆTôæ#ÊŽëÀEgñXáú $LŽ±ÐŸþ¾l©Ï_‡QüPVËØËe‡O¡‹Ìew±˜ÃÁ´Úp†JTÄW·ªiéëíÝB”Lƒé…–Ûú WlKyœÇb)G_gï²~ëò)<Ù8^_ëDgõP© žàGv÷›`“q#q,ýq ô€?þvÃSY»¥·ž>¾råÚõËŽÆ-cÇwEBÈËÉ‘µtaeÛ­UÁ+Cü{8J}Ò¦Gk­+ÊŒ4<® ‡×Ò0Ÿ7ŋ咮Ϡ·q%9,6„)J'dëDûõðÕ±"Üf‹P¥hBt\—´%EÈ}j6iÖ„±ä ƒ°#^ ä7 æ~Ë£†@£@I|ܶ[3aêo0 >xð[Ë…3›l ø´°TUbö"­Æp/,×Áx|ÕÌL’”;` Åìùs§/݆oÑ+= !wÃl¨aaýÖ³H{ËÙû#ÒЪöéïqF0:Ó…cq]‘:?¥ô%œ„lV‘®L“g¤/ÌCŸ1Uõ»ü®ÀG—–¶»Xz‰³;ÕqIœ«Ö…´~¤˜á9þuØ5/}ÚûmÄšAÛ¶ùLظ*÷¶œK(Ù™»1†Œð£$*$¯Ë)JiÌãš{°®\Ã!e{¤ÛÃó•îÈôI$rfaº1†xú]º i3¦´¶âDzôWž7nÞ9ÔR‚´^>ÉÁa\òÒ•Ñ ÄÌB§*ÛIËÉ€©N½?ëÙ_-*AÌ º¨²íªjÞîÒmY}CD¼‰TL´ô‰ê9(—ü}F=Zxƒ“†@v5Ö0oPœ0°2Õ }ûŒpªU©nã}ÝæQÔì<·t¢ a»X°'©ëžÆÄÔÐX„oaÜ2­ycF—´ÔÑ ÁOó¼5H‹¬KQ9Êm0œs§ìô ։Є(±°É°ƒ–D#…z CÛ”¬”o¢×ᓦÉôΪˆ£‘B/%æz¡h‚Ĩ Ä8wÌíT2É´3œ4mjv éús-ªã/à–tEZø~d½¥å”è*¦¤•1)t†s#î™›V‚òP#J+J/c:Ó5Q÷BgqF¿¹–’” =êÙ2¿ãÉûSÓ¡nÑi±<øI®­¾‡‡n†ìýµªŠÂšœÆe§¤hR RòQbþqÊo/=†ç'bóu£-¸äøŠ“7¢‹YÇv3Q„[>”¼5Ô3ÌÅ"fé–=Ÿ5Ýo„aé|'ã-†ý g(é~º º3O=;E«š·>û¥iU§\á6bk·ûiè…Gâ~xª©»ñSèÙž­`äÇßóln*ëT|ò¬k÷vqÒèÀ´K]‰ï±bÁñï˜lͤ•(˜—¦’“6¹ Z3¸›ÔðÒð È%ÔÛÁëLnHþ¦OF‹ôO@¼€€¸¹Û§‘ÕØªÞf f²@FŠULd!^bPǰG':ôà y.öŽØçgL'”ÆhãJÇAœUåÛa½›†÷ðéÑ( %[£$ùŒy€“gRx2JÉÈFùLeh‰|{Xè·=îM/Ž€Un:×åanè!°E,l[€âUÑIÊ8.>"d£b¦Í½úøx H`°ö3çx•ª+S¡-®½ñ>rÁs>ÁC&c³†ð{~.6¤IÚ(ôV«IICÅÌž Âá§´Ol¬¿IÒ¥ U*ÁWpS°29…Y{—¡RN0§Ë2‘2猲êË$EñU[ú?íLÿŸ'XO¤ÞÏu'1!ó-$6‰–6Õúl)Ú<‹FÁRÜïѨß šÑÁêšZ|MæIV¹LŸ5ï“k-?]½rõrÓ*'# ž6Cp³V ×ÙȆð´íˆYhÌXe«ZnýÖÍ‚ ²–´&æ{%W!뺼´}¼Y)éšñ®1‘•­²ä•rèDgôPA9V±ÚRÔôâŸñ@Ü{ôâQú¾,X{TiH\hlbH´vuŒ ÍAsN¨î3)Ø‚çÏ\EÄùÊs†2éþÒKôψ!Ó /}»¡ªï1úö—¨ŽÛ™³u¯¦¼¨æÂGHõѬMS¸5#Œbó_Þ®+Œt=¼ÃS,ÑA¢ND 4‰”-$907=}j÷Kx¬èfvZÕÝ °ÆÅüiAmòýÓ¡kÖl^0šÊP@¡Àîì™ °é‚Eí¹ƒ”}vnÅuKéïQ †yìj·c§Ï>|êóƒ׬tuuæ'z±iIûMˆyrãzK]LmX ¯Í«HÏOÉÜ^· 1Ååyµå!»WG­OöÞÄ«r½J7!fì¼9Ó6û†ðRLE©b‚ü!¯yN¤ Zˆ¢˜©¯—Áðúâ÷¡'ÖVsYne3ŸŽE(65ţȂ¨‚d JEé©é»è/±#æ:\’S_¯á+‹rÐ Ä´€)šê:góˆŽâiƒ¶¨M¡ÅÂl8Ê:¯E†®‹ƒeoa8õiîþ­]{w°®ÃZ<ЃƒÞM¯!e¥‚ „í®ÓO/ß°ÞeÅ~. N¾ì¥:t€¹?û.îÅáùÝ¡K  kôfâCKOÜ&¾¥ÕËÊ‚QrCŽÈ¿Ô·Ø¿Fµ1—*¾ØWOº âUü“ùؤma2tôô%”SÎEÖ*´êÜ.èÞya/ó=…‹Â ¶Ýï¾× Œ¦ø"îÓ OÇxz–¾’Lj‚Êd²  ™¬,¨¦¦¬¬†ëì,N]¢+W ‹7A5[ÿMaœ7/Ÿ)[Ÿ^íιUï<˜\’\œ\âÛa-±U›·,Ë,¨ÏáöûëT—‰·d®?l¹ìy02Ÿß²{ÖÇ%Œô¥m‰[žÇ^tÌúdÓñkЫp¬_&—”¯éôŠVíþ6!Ì?–Û²wuž ñŠâ93§-:êð] ñ·éÉωL—‚@o˶P¼LºB1‰îªRØ:–FaÈc‡×v?"í(25<•Ðq*ÊbjÅÁŠ °€Ù?£0ìS0»ÌIŸ ¿¼5:íüW¥k=Xuš9«·ôƒÛùÝÞÈï§ÁºUíÜ­;µû·† ‘zFêÑl°Ù,˜ƒÄto7 8®›ÌÒxc ¯Û7t4RnS«z}·F§[ß±þýíMR}ãÛß®´…¦—qy¿Ž™ðÕ-=Dd‹\¯®SKÈþl7²Â?Ð6ÎÎkOýÎÁu²ÌÛf=§±ùƒñ/^°üþÉSNú ]Ûxޡ΀1!̯X<€_,¥ß Â?Ø¢ð"UT|B|ç/Û¡R“Æ,6'"/"_©õFÛ‘L¥S†¡PFU¨*ÊÍÌÈÌâêjªŠ Ð.¢3ù‘ùšà=¨Õ•k kËPI‡æP:¡§FT{äÍâ}ÈÆD—­d|râvs©¥ÝDo‘1—Kºm} ¯é†¸loNF¢„ä%ƒ?jeXÃ+SüJÒäæ{Àé`*Q¡–5zH&AþêkÖ ¹Æm b`%Ä=#.£}âb|y\* EŠú’ôŒ]Ù\uý’s¤ÿ¼{ÄÞñ£yŸ|([œ}FF"šE"Z«( ÷Ù:ýÞ|0þϘ“þþrÖÏDÈÛ³Üxf‚ùÔ-Îm¯RûÁc¶ÞýáîC´Ee$¯xÂÀeU¥¦¸®ÉóóQ¤bá>Ø…“¾ÆÑ`bÿD?ß3 '·àœTpFÎa.[°ÁXG´µI©KÊLÎLÊdJNGä+%¡fï'g%YºxáÄ?îœ^d7Õaåܶci…Fè°þ,)Õ>„—±^â·#܃Q#Ÿ½Åʵ\4ÂX°××”¤¦¢Æ!Rüt|Q¨¡õFA!¼[‡$•vÈ=¤ê`´ÞIåzÃëgP­¬ýÒ#ͧšsûìç•K]íùoüØo«4¡³ÌÝi·qolöáLÛ•G×}³ƒ“þ2'xñò9ƒF= ýÁâÉÃg·ÏÛi¹™¸žîЬ׻ôøÉÙ%Ë8¬šn¨#¸ÇDQ¨ïþ(þ‹ºý%\fÓŽygG •æš]­RÂÓ‚kP%Ê(I)6$}óŸÑ~™¡‘dNïÜ· Õ9ªØhÜïã‘£Q òL .JÏÈÈ$mXE¨f‡\æï[2ØW+RZýa=IÄc±~ ŸNrI?œ$uÏN¡SÅG0xh:k»æË›®5_?T£PñU>yk‰ÁMJNB (!]…b§Õ+§sí¼FnLþNîø¼ÉÛºö¼îæ=pºw‘$ÒâÜåÅ-ßëžß\|ÙRúO-…Zö‹ˆC.h-Zë²qq³ T,—¸dn) ?™™¾+•35ÅÊÏ›¦Ÿ›L·Í`ÍvסÊ"nŸú–ž¼®¢&Ùs<¬š‘þë4l•4ì<¡.Z›‰"⇪µy_;]ǽvqm†ÌÉUœ‡÷ˆX<Ù¡ÀLLÁL?æªÉÊ]k>E'Ññãû/î2$¡Y+¸jÚk¹ bs³ò³Q)S’§ŠILŽQsø.v‰ˆE!Ö(";jWÜÇV1iÉ„!‹JóÊKɧY<Þ¦¬8_½{U´Op Ká›×@pék‰±Po# Ø@^ã1lA#yâcÐÆ?éHò¤‘¨õ›çYI°“i8ÊÍÎÂ÷ÁÃJ“˜ƒUH¤"„|ÇÃ~ì€É¿¨¸¶C*J#‹rÒRr ©™¡ý#h’©A‡Õ,*ŒKÍÃýîmk&èÅ>àaÐã¢Ü¬tÒ²g'hvàÍØ çàÞG'î]Ƕ¿µª1Ušl޸اD._ÃyAïM¼N¬¹1§91òxelxn\F"Ó±¨O%ˆm˜Çƒ'„“<¡p&6SYY › vsÚ_bœÔAù!:Üü¸æ7‹…¡B(›˜ÚfK>rœŠMqï«¶7›~:vW¨ð<<÷Öx^}±5Ø<ºw¶¶F†¢ÃQt4ïåëåMÌ75 (x>xùê—š#ÑËÊùÅø»ÑíÂí{GÏ ¨!§6ù‹¤*CŒ0¯ †bÜ_kAªr¸¡¢؛Ɔ/ݨŠÜš@ô,¼0¬¸þÕçÐC·õ3Ww…¿¯G­¼!3+%5‹ë•+ÌÍÙ¹4Í’èzé{s½LBef=uffz³>õÄi|endstream endobj 111 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 3375 >> stream xœµWktSeºÞ!MØB)ŽãÆfì: HQhAG«bAK(-‚ ÐKhCoi›ônÓ4i²÷~wÒ\zoi›6½¥(±ÈõˆÈŒÈfôèâ€èTtt˜oÇsÖ|i9çpfkÍZçüÈJ²’ïÍû=ïó<ï2Ž’Édã—Äůž|9Oš*“'M“Ûpò†@‚b59vñ´‡/U$‡A¨BCú}ðý‡Ð[?C¦É(õAJ.“å,KrtÅyÚ´t}Ĭ”Ȉy .x2b~tôˆ—²4yÚ”¤ìˆ¸$}º&+IOÞdFÄç¤h5úâˆY1éz½nQTTaaáܤ¬ü¹9yi/D>Q¨Õ§G¬×äkò 4©Ër²õk’²4cÎ{Z’“¥3è5yq9©š¼lŠ¢¦fÇæè–æ-Ë_®7$­N‰KÕ¤¥¯×&dfŸ¤¨W©éÔZj5ƒŠ§§6P³¨*‘Š¥6Qs¨¥Ô\êejõ µ‚zŠZI=CÅQk(35‰z”@D…Pƒ²XÙ®qk’O“„lWü\Q¢ü™òÀx=ŽÞþÀ¬ŽNxb¢|âáP{èNªŸômX`rôä ‰ “Ž_RyeÒÚ³rT‚¾bZnâéÖLÅÉF«®2 gZ³X®¸J+ð–=lkfø xÜÞ]pŒŽ„Nwú›ôàs¸‡m¬û û ç s æíÇúº‡›€ÞÛV¸¥Œ¯¨\Çjñ–*M)ÄŠÔ¿wgó G3ÛtµºËµZToîZ67á¥Ü2uXà5ð£¼h¯_&]¼)G³/2ÖÌFè…fAååšy0€…¯ÌOÝö†Ûò9c–µˆÓBa°ciª=Žd·n^9ñt û]ëµßÃUúÓÏâ05.Áç˜Èîì=G@…B•_7¿¶pi4~€Å³ðçLN´;ÀîhbGOâ”i™Ÿ…CØ0‰«ðIÑ^Ù{ÐÈ'rÉÜ ’? Sø—8 ³X5ýãgQR!…¢j\‡-L΂Zº§óe_8WÝtÿÅ2¢7c*}›¹~CæF É`p¢=é•fûdRÞ§rT€0«Mom7Šªîd늶MAáh¹2æ£ÈEñ‰Æ ¶î ž9:Á&[þ("lÌí KJ9¬ºü1|7凂À×IÝÌ›§Oµíú½7_‡9+ŸŽ×î}«œ Tƒ? óË¿¾!G‘ÿ à”Å£ehÊFO=Ž&ãhÇ-gÆ0¦¿V~ýÕUbï<ÿ›¼aIâL g+ÿè0iê+Ñõ+r¤G_2-?à°ÊbAà« 8[!ðt X¸bv+n¯Òج/JP’ îGì½à‚Ï…~ž\Ð^+ÖÙké"å£ËÛØ^v¶±-ïì¿þ%ÐÂøRKE1ÒY>Kƒ¯³­¯_בºjsâÆ|··C~42(Ì–£~ô'æ3å™>ãÆb®„ü¨OyíRä)Ë¢ s£€^ <€<Î~§ëá­DÓ|Û+‡"ÁH¾Xƒ ŠÛ#ÁÖ¼Ê÷ë-ð|%;ÿv¯5ƒ7.•!øÑãO°Þ„Ïa˜ÓC†2Æèñ: "Íu…Wz¦Uvù8Š?.—fKÏ1ö°ƒƒvqN³‡ò2µ±’àQµ³g”§ï˜÷º¶YßVÂvëûªU¬|ÏkhmÞÒÇ6¯mA“ªÔB38`ãŠõøá|%ð`£-Õ6·SÛÚÕ.gcƒÓÙ¥6í%L›9|ëØp©¯¸…ÍîËu&Ôn­‰¯ƒ#´—ÑÏ[ÏCt9_œº²FpÔ‰®Ðgð£7üÒ ¿ Q7åµh+ãr:Ï€Xãh=¿á“…4ȘvB¿ò-Ïó&Ö¦3oãLúu‹Ó—=O:…–»-¾Œ&²¿ýó¿¿u‰4EŸþª~¼Xá7ÁÃÕŽÔî©rTÙÁX¡^›¹oóH4¨p*ÞŽ7áB¼1øAôÜgîfë ý)uä8ÍDuáDÐSüèÙ6 #Žh‘~Å éÊ;ÂÓ•}h5zéœÄ'¢8Îþ™r+î⊠nåªÙÍ}‡üþÁ]ê]~{CÃ%:OiŒåß(Ï5—éS5@'™ööÖ‰Õ¢›Ló™¦ê€ =|F.õHk̘x(AÅñ¶<àFÙm©Òrl~¨*Ó¼qû6UFF¦)èXa¸Õ)¶ŠÕì»èæ=š xŠÍÀViLñPD¯ñoÿþê[ÿVëR‹Â ÐCµvhEQå°WûÖÝvp;vÚ™·£…^ÙÉOÿ¯r©Í`l¹–$®ÒœUT^´Áâm­­9lo`]ýh xé—Y0sùÊø’Zƒ¿·£³§«¸!_Tû‡NÔõ}²oË&“%ÁƱ¼Àƒ^À «ðĘK¿ðuéÛWkô¨—½ijï}§íÃzµ£Aô¸öÐx‘“±fY35;òóÓKÒޞѳ_‰ÊjIÄ~Y½ô/räE;kÕœÈJs®.9õA0so•t½[ìÝpˆ¯Ý-}±D„N4×7{<_\ ?é6ÕƒM¦Ê’LЮO|ÇÞAÖÑ~¡;h.ÿBÉ¢.çÒÔ'|~ ;~ø¨³TwÈÃë„ÂÚ-v£HNX³ÜB‹ù*((-ÍÑq¶¤íóÚoÂ+‚*…ºº¦^»è!´¯¡s…|)d@ª=ôdæIhâ½iª°€bÌo¤«ò€Õ1yÊLLíÄñR cî2È/¤'áØW êW§H㢠RáEHs‚wPhW~Pk)ÛÄ£ÿÛlx“PÁ›hi—2LZ[áC×vûdg/"Ýu¹ô¼4…ÙÈ[“ÁDçéëiÝwfIß < ‡âGÈBS~ºø[ô‹£hRÇb·šÞX¥Þ¹|ýÎu@oŽBšúGôÈïòé½lسƮ³çǵrôCàqÆÚ!ˆ:bMÏ)ïÑâ‚Ûœ-ß¼u¨î48$¶€~Çû…Ñ‘(ò—Áq·½Fôú"qÙÿ9ŒvS¨jÁUãô ¿ÂÂEåìÛÞm­»¿'ÖºΞ µŽ™] \]îò:D)wUêsöWו7ˆ'ŽQæƒ * ñû·O…©uáÚÙÙóré·• ~è²ÛFi8ÃÔ¬=K g ÇðL<ñÏ/_¿xâ¤ÛÍšV!æ¾ú[ÙÃpE ”dîØ’½ èÔˆju"Åásìð±ãdE_숭 ® vlvYí(Ö+;}¹‰‘,F¥Lgm­/4ƒ‹Fbv/ÁS±‚<Á!Ÿ.ú+šü;¤l©æÝ•fÞf²ª_Š© ùÍ+½ˆês jU·3Cè©M@_ÈBƒ`±é‚΀“;Ðúw?ë@[ÏËu¢ãïÉ¥)–±“ƒ2ºÂSÖÚÐVÛU[5š¶½$K«6Õj»Ö“Õ>ÿ…ùš&]s k6kR`h›ôMÆåEɯÃVzå¥×Pè 4k‡&aDªã•Ùb‰zÀ)öuuõç[}@ß¼rárO¥¯¤ƒ=¸o‰šp=“™²ë”Çùaç0ëléºs¤|U¥m¹>•-JO+"f™3wíÁ·;;ØîÖ:øèÏo@ìš§ãg°ÿƒ£ &`ɉÞa»ˆ^›é`ö|&qÍÎBµ ¾‹¾ù‰‰?}»åîü1–l}p‚ï¿ÃÛ.¯ùÉ:Õ„°NÍô/ (ǽè ÄÖ¯à;¹dDטo¼·>€ô‘çð£j|&Xç~>0ªŒ?Œ*ã°ûNè¸[$Ñ(Ðx%R@”Ù ÖŠÕd£ðf(¥ýIƹ+¾¿48mn³I°9uùF­V ô6nÿÇbt²°:ó†Š|émùzVƒe$/5U·‹µmÕª{?´ýÑßËqø¢x<‰ C}A_?€&Ý’¡e“£á€œé(éÈÊÉÈOËn/éò÷öúÕx^ÏŒÅÀ;)ðDÃ}R`¿xÂÞ {`˜ï¹ƒÞÖBèKS§%J¨4øhòâey`³ô SÝåùœþp ¿EЂVÙ³‚þ\•©è8‰Và¹Rvþ&Ÿ„{ùsO€*%ª…×C#‘îY)ònI烆ò’ûrÜm‰ åP¡Âß|s¼’äx¦]=ŠD@Ö/@ r4òÓfóÔ ”ñA'áw߃2…êÀ޳ò@÷ÿ&zLBâ¶\µíÊÒÚm0^*ËÄpòWpò}ùàþÿ%ýÿ±x$Núà[JúîÚwrt* `üú†‚Œ¼¬Üâ†Ü~Ÿß;¨F‰S|E­yY:]¶®½¤£¯»»78œÐŽß’]G!r´ª…1Íßüê–dŽç8›«œVW#÷=š\§_Óè3²‹ês‡Z›k> stream xœ}[LyÆÿCaf(CÖš¨e:Ñi¶)C¼‘` ƒD#$l¬8ØbË`­•„rK©DD.}€‚Õè YïxM ûb¬q×Ý`âîB²ÙdCrÿ>X|ñÍœäË9ÉÉïûÎaHra&¥¸ìÀ¶ÕÆ®md4S’–­Ñž­´¥d“5{ve¯ý£ÅdABòõ•©,ôe¢¼«3ˆŽa¼'[Š•†€Ï}Ü¥J95)¯°°À*åÛí…Òn¯ìs×8ë¥2§ê’½N51x¤ ¥Æ-«)§È¥ª Ûssý~¿Íém´)¾ã;,VÉïV]Ò¹Qö–I%J½*•;½²´šÒ¶*ÅŠ·á”*û¤2å˜ì«'„°ª+/?ÇBÈA²d‘H*Ñ.qI&ò”É`Ú˜7IyZÈ åµNhé1Ü6Ì`RM:,×6Ñ£ÜïaÃö£V[ˆW>Ò«Mý˜¿ýü8ç9È•xje,DEƒöbZÆÊÓ ¦ÆÑ×igqÎø§ühÓÖêŠ1Œ‹,µ´¤Ôq¿övT™?p?¡T¬ãîD¦aÆa î„î ›c\if̸[þíî‹CGö‹4ðý]ƒ6–0¿›Nä¾ÓáüJ¦q´«»AÀäW’‹³†úo~E½Çÿ–ÿ1—Q]Òá > stream xœ••yp×Çw‘-b‹¤S%6 Z%”` i›áhšà†à GCŠ —°d$’,Ë–l ë²ìÝýiuX§-ÉÒú’oƒ9l $Å&@B´L™„I¡¡ažÜeÒ®¡ÿôø'¼Ù·3o¾ïûý½ßç=Ëš‡á8>¿°¸¸hÝÚ¹éºÌ³xæ¹y™%"à+ÿ‘=ûVölÑæ×–üð+Û“ÈANÖÀsD×PçSȹŸÄÄ8®«mfÜþB¾Þ 9¢6Ê——®¯[¿þ•Uò—Ö®]/½JeД*´òb…Q­ªR…ŸJù/u¥•±^¾|“ÚhÔoX³Æd2(ªj t†#¯®X%7iŒjù.UÊP§RÊ·è´FùÛŠ*•ü±×‚ÇŸB]•¾Ö¨2È‹uJ•A[©¨:¬Tè55š#U Ãr·šëßVí¨ØðôFiþ³¶+Áöbk°-Ø:l¶‹b¹8ƒîÆ$B!°,<??¯N´ZÎz9;7{eöïÅr15ÿ™ùæ %™Ate â³ÞÌ}ipŒ }Do’aP¤ )há-œÏYê´É„N¶è8™ ƒ 4A+CSMÛy*ŸO#kpÌšüï)AQV~ñÑ<»’vn¢Þ!¬Š›þÒd 8FÀÑ) AØæ2$ü¬Ïí'?E}ÙhJ¬à7WµZìf¦Ë÷Qd>µqÑ»Hq7Êáý÷Ðõ{"ôô¡ÔßÛuïø RMº“jˆÆ°¹Jœ¸°£]³Sý®Ñ*3\ÛÒ±t/5ØCÀ_e2OI½íÀB‘8´Ø]‡ŒnqV¶Ö7W7ë]úüÖ¼VË¡7Š€Øc›8v&~eä;é C’˜¬8V²ªŒ_昋r³1Ÿ¬ïÏ'„8TìÕÍf²ôùú]°xéjÕôøH<–5ýʺ·ôÝjJ»ŽÂOÃx ¢Ð CQè$õ^»±¼qoÉDÙ•û÷¿AÙi™dvgq¡Ý¨Ìo¥¡‰Á[Âqwº!ÔD–ƒš* ¨=ås½â² g‚y‰ ×A-P´…¡ø¬‡;óøêL§3Á¸Ï/›{Êwä\ÙK*6Pù.cnöµÈ!õ ÃŒQÃsº¾($C·ÜÉ:9HB›Ûç £¼LOž×´ÏŽØ‡_çµ5xè „<¾!ɨlÒsh#‡Ÿ™AñaÚŽ¥ýð{>ucÀ7 íÄ„êä^į|‘/\6½éohéí¿«ÍÕBÓN—lÓr+?sÜЏ;>"ýœt­ºßÑ Äû}ZÍRþÇp˜”Ì>-y¼]J´üO"ôkô™½*F/ ò[”u³è þ’¿)6ÛÀl B2š`˜£ýØþ£†¹UðüÚ¢ÿ Zœ™”GYÏi þ"þê¢~‹Òôå²äÀvçA øçÅ’L\¨Âå?F8Tv G§¯O OÌþ\:wxÆÍNºZædl Ô‚µŽ³l³]F™øe?o= 4¸òÉ–`´#ñË|ÁITìl„"h†VÊN[i+؈'J/ ôÅŽq²ž‰ð‡Àsؙ֣”Ì´Ù¸G0bn Ž]ÞéƒBðɶþŸ†CC]Ýg/ž9}üà–‰Ðn¸X–~r(ÁúÞÞ~n|À8|@·ß¹±P¦Q)KUB‚àþõz碴¿¦[«­©Ñj»kúû»»ûe’ÌßùÊÚÆeÖ$qÔˆ¾•úR¡ãà#âÖP¥cÔÿ£‡¿£¬BJ:ß¶ÆS‘„‡•yƒÑ ÇQÁ~²5 7`|²; ZšÀæ ÐnrX` çÍ¿ò=y|?ßgµº\ДßÐnÅ‚áP@Øçº-…þ9…£»çD(‘Y EÙâr~¹ãl èšÐAAˆd*‘ôëj'µ3ð ¤ábÛá ‹£c0Ò]f’¿#®ܧVVG$Àz~Ù¹›SèÄ%qŸ»›ÏU¬uZ„¼Í„9æŒÒÇ"²ö3ÀFÏžôÝ~ü J,Êt†þ­_Oòf1¿R`|=ÿÄËÓo|ƒ¢|”‹ âp¶r‚äuüwÒ½¥c3‘/G§ÉSŸ]nnÏÛ¢Ê}¦õš=¤rÇîj%’Z.SGQ'æáùƒ gž-Ìz%™³ ÇŸ“ƒaÿ9ŸH}endstream endobj 114 0 obj << /Filter /FlateDecode /Length 162 >> stream xœ]O»à Üù ÿyL•¢,é’¡UÕöˆ1C "dèß7Ò¡ÃY:ß|–ÃxÙFàðEŒehu[@‚‰fË¢n@[Œ…剋òB7åßO°Èü®’Ϻ½äU}„ÐiZ½B Šg]Uõ1½ ÖR L¦8ÛÝ™ÐTm“ý§’¢©Äyp 8榹I*`™~ÏxçS vˆ/,ûS(endstream endobj 115 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 316 >> stream xœcd`ab`dddsö Ž´±ôH3þaú!ËÜÝý}Ùì² üN²"÷òøºy˜»yX~ß*ô=[ð{ÿ÷TFƼâ¦vçü‚Ê¢ÌôŒdMCKKs#KÇÜÔ¢ÌäÄ<ßÄ’ŒÔÜÄ 'G!8?93µ¤RAÃ&£¤¤ÀJ_¿¼¼\/1·X/¿(ÝNSG¡<³$C!(µ8µ¨,5EÁ-?¯DÁ/17UìL=0霟[PZ’Z¤à›Ÿ’Z”WP”™›ÊÀÀÀhÀÀØÅÀÄÈÈâú£ƒïû–- ~Îg\õ½ùûÜy¢fôOêžÊ±&wa~rCFm»Üoá?ÎíõÝÝ]’ÝM;§vÌmîíîíîëžØ=i_ñâŸö Ø~ËMgßɵ“{ç\ æe`ÚpÂendstream endobj 116 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 333 >> stream xœ%ŽÏKÂPÇßsKI׬È.¦¾[ ¢³‚0¢_B‡À Û¥ˆÁ]r¬±?¡ÿÃãm‘âáûý~¾·|@#•jíª$ÍPò“Ðßù)ŽNÁYð¸ñ“£ÔÚø¡%R£?bß«¬½Âq¦,B½ÿTÑ»–¡6še•*•Ë»y´%Iet¬aCUäªÊ¤‰5™LOÕtEÅÄBÙý&!ݽbÑ4Í‚¬õ ºÑ8Èå‘©’&ºÄ=lÜá::Õ;ËFsÏÂ|*ºÖíl ª^ÇFjMë|Τøü…ÿ,²ÑdðöSs!‹yÛnuL“ZVúo'bQËu©ã¤ÅÀa¶Ç>ã‚Mf'†u]‹Ú™ß¯°mQÓtè0#ö>‚Ã÷ð$ùñ¢^,å×oï…EïU¦YàÂwIendstream endobj 117 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 478 >> stream xœÓ,þCMMI8/øøø‹ø©ùJB@Þ€‹ ‹ ©øG÷e÷b÷_ns‚Copyright (c) 1997, 2009 American Mathematical Society (), with Reserved Font Name CMMI8.CMMI8Computer ModernkLÝøˆø½§÷¶åš§÷e¸÷Ä×Ãë÷¸ù,ŒŽ˜‹Œ‡”‡lˆ‰tS‡u‰Š‹y}™‹—»‹…‚ˆ‹‰†yû"ü͇|‹…‹Šz™}ž­“«™•µ¥õ–µÂ…Îx‹R‹†‹†‰ˆ~‰}‹†P·`ŵ¥ªž˜£°›Ç‹‘•ƒŽ„‰…y‡~\qH\‹q…¡¦˜Œ—¢Œ›‹–‹Ý$X’°¡­­—˜ÂÁ¸°»‹˜‹—ˆ“U„‹\‹Šuœ|¡¦©¡´­s¯YQW_RRQRr€{„v øC•––÷p û`•« Ù  7Ÿ § (Áendstream endobj 118 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 426 >> stream xœcd`ab`ddäpöõõt24±õH3þaú!ËÜÝü3äG«,¿“ƒ¬È½)¾næn–Õ?$„¾ ~Ïãÿž-ÀÀÂÈXPÞÖíœ_PY”™žQ¢ ‘¬©`hii®£`d``©à˜›Z”™œ˜§à›X’‘š›Xää(ç'g¦–T*hØd””Xéë———ë%æëå¥Ûiê(”g–d(¥§•¥¦(¸åç•(ø%æ¦*@ª¥ós JKR‹|óSR‹òs 2¹»˜YýèàûÙÓ¼þ‡ÄzÆßÙ™Üùn!úò;ÃÂï¬Ý}3kº›*º:+Úä*> stream xœí[Yo$·~öG îqvÚ¼ElÀ6ìÀÁæÁ€ØƬ®•­Ëš=¤<8=U$»Yì©9´+;èA…6YdÝ«ž‰^Îþ•ÿG—ŸþMÚÙÙê@ÌÎ~>é×Yùwt9ûòFhOú(¢œžäWåL©Ðk'gÞú>j;;¼†h»+xªƒ—ÎuËùBkÕÇ`»?ÃSœ±¡»Mc…0áûÿXÕkÔ ž™“…ÙáñA÷ïùái}# °I~{ѽ¹ŒtÊFXlœEïU÷4¯æƒé¾™Ý+¡<,Üà…pˆ»qÂ`똼›X4JÁX ¤Ð.tGsåpë¡{•§·QvÏOð5ƒ¶. –Q(åºSpÚ˜ö½a Ü/H)P(#{Ø È÷­Ó¾ ly¡mœË;.û‘ö^N'”Èfb(o“XD´RjIwr}5eD5Ê`²™\~Š ¦Ø=¯ä?*y\ÉóJ®*ùº’ÿì=Ÿsç`]/¤PÃò¯’Œ½ö¡[®æ #Co@Ï–W™vÒtË$°¨%ÅY:'àŒºKo ʨŽÊéj®¶xºx\*hE½:{JV÷~†#8b-œ˜õ ;઒'‰ âÍU%_Wrź'{°/ÙŠ"aˆ(ôf½$qS µ¨•5²[&¡ha´ký`:>ˆ1PëJA¹xÍÕÄf4Ähˆ¥ÅhB‰Âh{9 úQWÑ€²:©} ™Í`b)ÓÁMÍXå2aƒ\àQ ï‡À Îßuý˜d‘XÝDãÐ[‡äóƒÃO^tŸÏV¡²øâÐ1M¸{YÉÛJ.Ù§÷|<üC%O+yQÉ“JÞUò’ ¿wì¼%䢸,³µ¯êÖ®Ùn*I\±,’§$ø¡’rXØuªOõ黳­‰E²Â¿²fz<5Óixþ¡’#k¡û¬’dìy%W•|]IÊïW§ìrGül²’OëàE}ªÈ{svðŠÝÈÙÔç%µ^³À*¶²ÁY­êП+yËz¿²D‹Jö•”dð|#]œQæ£uFÿ5mVÿ׿÷Ñf¢¬  RÙÿ¯©Â#™C±võ˜üÆ*\o“P¢Â_T&ñƒ(Ê׬ÇïwG‡ÊúXEßš¼ï«©ä+ãŸXE"Qº~9Qɾ’–]n£Ò !39™4ÿºj½D]=b½È£º™UÌgîP¾b øáiîô)‘xã|¸³þmyØ Ù Ä$õû;B.P¥¦¼4jÌÌ´Èjé$¥1@9û.U2yLe1Ap ”œŒA`¢N+T%¨ëC”M*{œ's²¤¯V)[Òú1³ú o¥÷`UàªEaÓÿÊ”G¤ëUYœŽ™N ›ØÎÅ­ÜiMš !zoçÖB¡¦¡ÆMÉ}ð*µ>åpU4ÆIÉ?p²±èÐÚ3Ÿ½UÈ„AŒ¬Á…È k£PÃXäC©‘ 7ÜÒÄT¤ÂÈáÆ-à !è€JŠ(ÁôZð­Ú(Ž2¬J«3RæÞ%RšPÀÞ¬©]:õ)´1pÐfFT_®SùEžM޼è%Õ¹ äu&”5 €F>#mƒ.Rq6ç m§ ÀcKbÙ£V:{ ‹PjFé,O“~C¡ÜÚ´L¾Í'á¥nÌ>áRKPÞ'kP¬žBluÀN‡ÃnÆØp€ªÈ&ôTi—™¥õPK},;(éCr6( NÝÜÓ¹à¶Á=19‚¡$OGSlN4†_‘ÊOæ?¢vIj×Ô©ôu<§DJÇž $$?;­äE% sWI’ž³h†·M‰*fKdYü›QùVcDïL)Sl~¢$ÄæËáK(êM­%eTï@€ÃÎ[”º*!ÃÚè$±7„\4µeª”Oæãe9 G8:ÊÂ:)%¬âKÓ½åo‰¨}ӹ݆ÔN‘78àü8èønçÒkäʈ"ØÃTxÅ€Üè0I#c7#coè6G8ÕØxŠìÈõS=)îêg{ÂÏö:”°pƒÆ*Çç¤ß+”Ô‰3÷oÉHkQû45¾H=ß²¤eþ–º‹åø¦$Ñëº=e+mÑó²ø1ŸR]nynªºÖËÂâvbr;"yÛH~CHÓ’>ÝÛQ ¬êYnqÙÀï dÉ­ÂÂÌÂA¨Ÿù¿«$©í—lj»á" T8Œ÷fp)6 j28]KN4ÎуpK&ë¢áÇ51½Ï¯oì<£ÐFêx”›nr˜ó*û»Jn2õQyòýºç¤ßé6Ë2‘Ö˜ìÔPôõfuôò:]7¹Qsƒªš §tSçj4"fÜ: 9Ç‹ªƒ÷m<\@Ú{§'yÝÔF\ŒQ6¢j#øXEQnÑ!ßí®iê±10¡ô•T©ôÁ…о,¨½²Mê8"ü{\¬y 3›¦¥€º­QÆKÇã¼\{šø`£€4{¨äÆ+Ó ê\ó1ü³;ÓX…=JœÊ3‘×Wùá$m<‡LÀ ?zе´«½ Y Ãcaüè(A“–ì Ë]€Õ3p “K}‚+¯K ‚8P8åîQQªg•ä‘“–M‚®¾bù‘ìj„ü´’Ûá’ Fô`Uõà„•9xS°$(-ëd%U%5¼ýŠëÃu•pNªÐË] úX5oÑ9V¾`µ™$_³Sd”¨;¿ÑDŒ{WÉ=òNU?|{,Ìj5&£¿‚Æ<Ä7}TÉû]*õE}º³-Åߢ%ø¨’÷•|Á’ÄÅ|OI<ÞßË™°þ‚»¢ä^çÿ!Šúó'.¾é)ã&ãÛL&Àö·÷ Þ€ùlÈ^ý4é[J3ΦØc[/¥ÏׇßävTƒ0~v»¹µånèA•:åx3lì’Â`*¨ÖáÃÜŸ²f¥é|/¶åA%Äœæ«kX{É€I®^ÞÔ±i0]•ù„¦Í9¢¿¦°XSÿ4~ЩÀ'®Ïr!i¹ÞŒÚåmf[ S~Ï ð(V®m(K;Ž–@°Á("÷ÈŽ[ß!2×/ãÏùž–À…ù¼L0fv’‹º õÐÖ3Ü<à|'LÓwd3Æ•¼DF<Èk³¦¸/Ý$ßXògÒJi-±ÓhHœSæn(lé.Ónæ{p$ì•6¿ÞiТüŽÌÑp&€ Ñw¶v´µ*Ù{]öÝZé&ªíXÊɖç“JÕh\%L*Õ,Ü„÷•¤`h~œP«¡‰;†MÆšeÊ`o1x„{‚]æì§Æt6)%¥ëƒôнIÕÙžŸTo:n'ÔâÞ0˜p‚©' p…©‡‰÷ƒŠÊ{o ¢Ä^ó‘±×•·-·%½ '½lÅLÂAEr…¿¡–'Ó ÁÞ—PÆÛF­iŠÂ§ß°¹KÒ™%­P6\¨xÓkY¯v¶·f.T W²Lç×ï­ýµ RHžÓvä­|‹Ù2éX!@kµêiÂCþ' ÉKYMÒ½â8ÖÒ>qr£yÂ5§Ý°}l–N»ÞVV®w6,ñG·ìØ‹ÇêGZW]NÅnßѬlˆðHcÑÅ®%;¤Õ6°K¼e_#*ø’åaE_ãØ!bÛ˜&î¿­öTu<Ï*ùIß$çºqîCùc• x¼°1ÌïsÁœ>°âlWaã¨fpé Gc‘^>–Á<àÿ°/±hm“ÓGoµ¡÷M ¿ä?£Üçu%çÉgh—,¹dÕ“ ={Änk{áãØ ñ€Ïo X‘ÍU÷¶ƒ¯°<Ÿ×·JšÕ÷ió "üí¼'1ªN(§”ÅÏØÖïIHZò¦:ºK–|З#{ï‘IL³™©ÑúIŸý²’S2MóŒ»B çòº’'•Ü3Ù‰OÐ3$hë/•$ƒŸÖ§ÛÀÇßc>·ìóm³·ìX>>¾ï'çìˆßB8NIš™­’4áóäU=m;!¾3¿!?Ñ·E%E%«7oوɾœ²óA‰Ážétì« M¡ÈMf&’™û^âI/BHûÖw~Î â¶žíö/z¢ãX¾ãžÿ8áóJ öú–74¢Â¥¯É@‚^¿Å¼Ö>BJä%ëÛˆ?–õ[Á°•t5µ·tä&ø1wè|åÕ³JÓÓ#I²_3^â¼ç~Ô®ýJŽû1ì~\eÌîÚ+©J0³®ÉG5}~Ð}<ÿx-ØAX×^Žžç‚=гª†äË¢²?ÑýslÛ{- nŒ¤o·'e²×VÐñAžùÅQìñ7§ë^¨û&QˆôO|Æ*Ñð©“S½W2NÜ/@¾­$ѱg¬ûõ•Œì&»³Ðìlб/ëÓo+ùU%×x˜jÙ.Ñ<â\I™h¨îŽ7fßüÈL»jendstream endobj 120 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 5881 >> stream xœ•X tSÕÖ¾!ärd¾ÐÞ[¨‚@2 ÈLAæy*´@iKGB:$Mš4mÆ“¹MÓ$Òy€2Tž>@J (¨ OúœÐÏu\ë?I ©ïáïÿVéZ4÷äœ}¾½¿oûòˆÎ×uÞ²5K×®éûc<7˜Ç éĽÌÏBbî÷ß–^&zÍýrÿ»Š˜ž ôè\;„]Õ—›Üê÷ô&ø<Þ^±v^ZºX˜Ÿ {mçȰñÓ¦M‰›9-lNJœ0qgljزXQB\J¬ÿ‘¶&mgbœHöÚŒ‘(}ú¸qYYYccSöMÆÏ–•(J[·/N˜·+laZª(lylJ\سPÇ>ûϼ´”ô Qœ0lYÚ®8a*Aóæ¤nL›—>ïáÂ}¢¨ŒÅ™Y±Ù;–Šw.“ìŠ[±;>auâšµIë’S¢ÇŒÝ>.rü„‰¹¯Oš2uØôW^1sÖk#GmAÈÄ4b%ñ ±Šx•XM 'Ö#ˆµÄ:b$±žEl 6s‰MÄÌ¢A\ÚyÈ}¬°Ž2$‘ ˆ ò¤A¡¤ÿPxÅËãÆ^ås½ôvß I©µx 8†t×à#ö»‹<·ÙÞ6XAyè­­ûmØž™¶›F£ X(ß f®ïì@h-äFtQ°¹ÃN:îT'Z;êªO9áÔñÂìÔ}®^Åúcád8–î$]¸_:þÊ+ÁkE áëápAM ¾[ÈeÃÉaû!å䔿£RAzðÃ_ „…½áúÑzÁ^a{˜¿ø •MŽ‚e€Ê oyyÚ`Ó>g‚E4ì<ëúÔõ@½‹¨/¢¾†½`O؆3HŽ¢é¨íÞ¿Þ»{ûãw×Î3+|¼ï /Wèåýúw‹6˜ 6`¤\yÎ,‰:_QÀ¨U ‰¦ qLéV@¡±h(ZŽ6 æ"ñÖþ„½¼•;U™ D*6ñGK’5yãýÇp>6vjù0#ÜÈÊ <ÅÖZ3ƒODB/Èã¶«|8÷])lJm| ö€]áPØyöÈÕÑ{³÷²5À¹ëÙšø´gë9fÌ ‡Ét<È.W”J+ ê@0:²9kÿþAË9@]<ºa"ÆO˜Ü!{Ém|8žƒôóJÉéOe8+ÁzØFÂQ°/\WÿºÒ¨‹ú,§¿%a?š††MD“P?6‚ Þì/ïÆÓÎ ¤Mn`fª:Û"MÒɶ0rwâ±yÉþ¨'ꊓĿõæ÷×Ou•± (^°>X<ð`MÅh&¾Ø²Í§o=ytÿë›å³w¨•Q*›þÖʵQ€ #à ·{ypl¡áR¯ÖŠšz4%õ–’rSúØhÇÁC‹´åIÉ:YLƒÍp»jImFÁ. ¡ö•«]Æ’*3㸊C7áÕ~¾<íD¼ÆPB Ž%ƒ¥è´±ÄѬ”¯ RtR¡šQ, òvÚ^Ä¡MñÒ“;ÜЬ顤¥Éý½ÉQùà‹{O€agƒÆ¨ÚrOíù™@>:j©LÆæªgüE‚Çßl?`ï¦Àz&5Y@TTj=°5Ýn|ýŸŒÅþëûÂKÒIvª4˜<÷ ñ/¯ß'XõþëÇu¨qõò1<š|NKØÇ s½|.Š›LÇU®Rœô)±37N{ó]{¼àrˆ¡tà 4] ?ìFXس–<ZÊA­ÀÝD=߆œåsnßæU/ÈÊIrzO 'ñ—J”7”×õE hò’—lB‚ü¹ÝPŠÐÃÁü­G88 >ТÎý!Çþ8j€‡5Šô« "ãR°,¥0 ŽºÜÀídìöó.:H$$ÈYö7lë y-À*£™ duŠ÷G/ïSœÖTxž®•Bš‹ø`fºGæ)l1Ÿf*-Í Pÿc^)Õè²ôvå4ŠzëÌšoÞ»à¨}›)K? >¨ºjl2E6°d°«¿ U‰Úœø¬½¹;òµ!¾úÐeØó ìcbŸ«·#t‚t?+6D¤®”³‘ÃWþ$h¬zî‡ÈGè54Ml f RpÒAûöºŠz 2Ÿ%¿«/¾g¹l<9NjwŸx¡ ½£–pç}emÄÜÿÏâœúT£ÍÁÅ© õ©ËN™?ï¿ï¡ÿU?…U¸’/ø¢o/äÉBñé}“¿w²´ö¸îl¸jU):ÙÞ@ã(ðwÔlØ—KöòÊîó;–·ß®Á­+IÔå§,ÓV£ª>tÐ+k±ñ‚&@ 41S”²Kwë$z‹ÍâE”KY‘*ITDïhH;zûâ—gë™ç‚þä*´ü“ÏeCíT:…UŽ’ÅÅdlÔô-®¶µÂ>q4­5è : $ ᨻ©øí+cÁ´øu1ìµkã1Æ¡g¾ñ0AöÇqrº¨ÌZb)¡ ä̧©Iý·ŸUB&{›“ JƒÌßè÷Ú²´)hFˆhkê*Y¥'µgc`¡Æ¦,—€@­xV•¬úoû—&0¦dûÒÂ-Ž7­‰æ½Ö=¦ÌzÜýj€«‰ Öw Gл½ÔPÕY ‡Aß“© _£JÖ+•›°ÕHmÖFcakƒ;êS } ÏËÏÙ¡û*‹áx8ׯ"Ë~sUá ¬š¾ö’— p®uù@WC¡*²c!_Å<. Ôñ‹å9MÊ)ßݰ„b»ÛuAá¨ó§sž:ái:ÌêÕz-PR¾«2\´IE¯Úvê&ä}ýÏ{7ÊejÔ)ù[XÙ´‹> 0äìïåÝ¿ ›p 1p ÝèŸÃ!ÍóQê†}[|5 à€ï×úKA«–ä±{wlÉŽÔ²Y‡`K.ÐÕmmçoê{0uŸèôA¯}ïüøŸká´ô“n4u^ª›ÈÄ…'"%ˆ¢Âˆ!üÍùø Óôð”‚kª~ºž¦lô~zF¹Î²5êK¿Ô—ò9Kö …y1Ø{ÙÉXížç û¼âkð};4ø®¤Ùc°›+¬ë[nSL30ÊÕè8tnÌØ¾S®`±7T䌣ƒ`둆%Ö·í§©øœóÁÙ 0úÏyþí«PÓʇ)^:ÏoùFJ‘yJÊ&Ö­h”Ô÷¿ú¦ZU©,cë]•¦ƒEèÉwÊã)­+•7¯ŪS¶²{=‰EI€zõõ ãâÝÉF ›-’íÁIö¹D“¢åsAµàö[pæí} {¡Qæh&v¡·¦ –¾ºs,ûl®ô èØ €¾¸ð|Lú¯ÇÅH%ø¯¦Uµljwµ}á¯WJÛö“rCáYú³cÍWÀYêÚšHÀ ÅÿoäóºýNß%oÚ·ŠUj¡>ŸU,JK½ùnÆ‘k ÷ª±H¥ UÍèõÊm™xœ/0×9, æÖ\m©%T?)¨TÚ}€B‚fz¾Çëä?š7Æ®[5òy+¼íiÅóÌc>7‚û®Î®LK¥ ÷z2kjj4æ©‘F}; q:XN#È¿ÃìÞ ë0ìäsÁ~AÍG>ýajÍVÉÕZèN–¶£emÑ.l»»D¼>jCuZE&[Ÿ^«ú,ûRNÈIye<ˆ£–¯˜71 õß¹Œ`c|cšR ¤ùŒF)D€ÖoÁ+ÌEGÊ™cñ)ÞÃ=’þ__J8˜[¦×e›Ãk²-É6•1Ù±¯¦Z/Ÿ¿†×Ss@fsÍ^ifþÍå üËac0¹ÓF¯Xâ·‘ÚÀR“ qØÖ0º(âOõšÃÔœ {Œ~>rÁ°ÆM½‹àh°ÐØ`KÜO¢hôðÛŽ°/ ¦ç„ÿ¬Ž£éØŽ£©Ñaª,<ùð“oиŽ/u`·`VÓÛó¦õµQ>wq0°‡_•tœ.W£ïdSÎ[B7í8þI1°l+|"ø¿¾àu4uº±ðKȉïÜpüotF‚|†½özP‡r¼ÝÁ:ôޝ\ô9>Tziá¾ç“§ÿ‚"ÆZš0ó k«.k>(»S¯Ôk±èÇ"‡àEë‹o”\wܰVà¦p{™½fÇN…’‘¿T,B˜—.ÒÔeø†Æ¤@`U žõ©pfî÷ÿúxûTïoá8öq¿…Ñ«Á‚iÀ ªC¦†“ˆ÷{¸Ev!ŠÄmXöû ¬nãýÊEò¹dXM;‚ƒ|óÅ5Rž‡Ÿú°î†}ùðþºQT—˜˜—R™~¨¹¹ñˆ/m¼w}›Œ‹øÓ=¦z~UðþÆUñ¹K¿õ¢-6³ Ø©òwvv<¯€yºö÷%r­ 䇊]’ŠòB‡ÝÊ<—„o®Á÷¯ð¹®? é· ïUÜ[‡ Õ(…|4ôGø2 ×Âí r¡¯èÅ»îý —þ þº~xÒx4j,Z6¼ÝÆöô>®àq>] æTº2% •†‘ £çÌSIõ™:B)¬’ÚÃ-e•̧“jÊœ&·OyZŸb=€Ž6>×n§!uïì]ðêÛ7¯cC1lê¼\X!©¬óTÕØU…©:zúȇ€úøã¤Ù#P×É\Šeh½T¡Î²P¿¯ÉRP¬ÕF»õ0(ô¿c’e€­:ÉtƒÆ ¾±û£°7µ«ð}úÑÕÑ çÌ!™h|_ÈšMN¨¢j³\[ÀƼÑOv Ã~yòã{igVìg½_úPKµ®»‚º¢ÑëæoLh*(®ª.op¨mÀÀjn3¸uáhú®Ô̘¸TvwzŠ.Q·R#Óclüâ 89ìÁƒsáS>×ÄݤKrËÅyjM’‘Êw$‰@>PÙòí9å’òD°ìÊó“2DR ¦d%â2»Õj23ÎâwVcšXÕvy¹¨LÚ€‡&wuicE-¾~»Ý!¼\Wœ³Qp&ݼsÉ‹œ@W-šFf€ £wž!y¹jÈ  )F„`q‡.³(¸Å²ž+¡ ä*ô£ ÛamòÂIxÒüVcX>Q×v°XRäzUB * è·†’ɪÜ&SQsòÔEK# ÚšÖ­=fndì&÷ k·ÙñK•ɪR·€ yïn|À>?Áþß,¹ž¾]“±Å_äî°Z8Ik´#&Ì@ƒ"Ñ b°õÎ(•»ÍÕnÓ7ýÂ1p€½¨Ôe6MæBL¡ç¿ Üb C ÐæhÅ:yÁNl’}ÍÅY N™ö¢ñÂh/ïZ<ƒ•: Þ¥oVÞ8>¥>]ðê„õÆôï÷Øť̉KÀ*òó…°|åWØù›3ñggÔ2sée«œ;ßvæÎ­ãÛ·¯œ·n:V®é9kN\ºñÙ•‡Ž®Ú<÷…‘>¶ÁÛhÅ^a\‚>¼ÌyiƒÉX,TinuJŒ&V¡g²ÐfL éM5&·ÛÌ@ŒùëUE‚‹±*q¹nÿ©HQë ejªÇ{S^ðâ°B.ÃÜË|n÷>ý¾ª1Ã'Sæç)%.5œ2¾WÖ¶õüp;ci0•áQÛç>Òµ’½F§ÐfëuøGŸ«C]²æ½f€o/†Dö.$ yÝ.àÙÜl0šŒÎ³g§6ù¬’N¸´ù®’Q§áv¦¡ÄEY­¹¢ÈÀ˜+ŒvSìTüè+ð=øbyë'µÂI¯)ŠkÇÁ—ß¹XL¹Q½_Õ*ËÀÊÕëæË|Ęz»µaøEì{îãsnüËçae-tŽÙ§¬™ùR9VÖÐßåËpmhýÊê..vØø?0Œö¦‹Y·aÖ¤÷n=u÷ fèp:_YV{æÀ ·Óœ¤X—´MÌNà8úbÊÑØU›§O]Øœx¤õÔ;ÌÐWie~yÍÍÇ\%–Téú¤h±/éç<ÜvLÖz.‡¶[vLâJq‰D•¯+ÈeÐS›­(P`‘â¥EO£¸WžïÕ-UVê¬(˳XXä€oùÞ{À mwÚx7¹T>g„?ÒÅ͸­ÄùÚödޝ­4ùÔø A»} 'Ù÷‘ŠDüä~Œ%VÑåŶbàe¹vµ †q!R»ûH*'[.;òmjÚÑbt -ÍSÉò€84£,·´°Èd)ôUÍ4ÏoûpŤbÎÌ^Gg¾U^ôê­lØ O·à&˜XÍE6à¤ù½‰Þ޵LlÈÔïl]Q¨kE!ÜêÞj ׃ÇU_­úgzË o „ª¬|Y¡Ê¡dÏ*GÀ|ðFÚ„RÑ@•F‰õÊ·dÕ†.8ó…õÀà ªàýÂMó½>Á㻵¬+~…¬P¦ÔêT u_R{ qAZBKòœâ<¡:A͈QïˆK¹EjìªB3$Yb™µ ¤€ýa%ê¾J®’(@f¨Ò’]e·á"f`¯+ˆÒʰҩC%%y®â*[½…1Xl§ EE…üë-œG`¡•¥en¥UiaG\ÉÐß^¢adsåó¿zu2¢~+D›ƒs i—ê¨Æl¸OŒ±|Ìlƒ³*x÷¹µ|ø/ìNà2h.†Õp˜Á‚G+ch©¬,3Eš§c„h´ú8„܇†éÕzÐ…JœbO£¥ÁîWCÄz`·€­Ƈ tÚò-2‰J¥ÏV,J^œåyvþìñ;i-ÛböÆ'Å×f6ÁÄôTsóŠà›Å 7‰Ä¶.ÞnW»3Ý:‹Ózt%ˆÿK,N;endstream endobj 121 0 obj << /Filter /FlateDecode /Length 151 >> stream xœ31Ó³´P0P0´T06P0²P05WH1ä2Ð30Š!Œ‘œËU¨`aS027iÆ`9'O.}O_…’¢ÒT.ýp .} *.}§gC.}—h ™±\únúÎÎn@n€¾·‚.PÆ9?§47¯XÁÂÎŽËÓEAmGTaý?ß ÿ p«ƒ¡ËÕS! B)Éendstream endobj 122 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O»à Üù ÿy ]¢,é’¡UÕöˆ1C "dèß7Ò¡ÃY:ß|–ÃxÙFàðEŒehu[@‚‰fË¢n@[Œ…剋òB7åßO°Èü®’Ϻmóª>Bè4­^!Å3‰®ªúΘ^ë?©&SœÍ¥/h0ûO%ES‰ó&àqÌMs“TÀ2ýžñΧì_=–Seendstream endobj 123 0 obj << /Filter /FlateDecode /Length 4399 >> stream xœí\[\G~_ñ#æ‰3‘æÐ÷ BH€dÀÃb¯Ñ®×xã`Güx¾êË骞³v6Š¢í©©®®ûåœÿý V}Pô_ûûäöâûŸjx~¡Ï/þ~¡Ë·‡öçÉíáÇ—„‘Y³Êúpùì¢nÕ‡¤ÑÇ5[¸¼½XÜñòoÀµŽãF|ŸCþåÓ‹?.?;ªÕ(er\^ÐR+ÒòòxR«Ï9FSÀ:e­âò9–*gŸòr}%•“‹`•¶¥h,QÛ¸¾º-”ÑÆ-¯†d7ƒížÂÔÄRícs†&ÿ:( Mý3¦þ|ùË‹O./~Gþf]^C:xãªÌÁF|öù Sp«5‡×ׇ?^^øÕy¯üáp¾_âÿ¿Á]~¡]Hkʇ`]U<Ü^h¯Òª# ɬðÄ›‹ÏV…q¬º¯b¥´z{€æÝj ’öâöN™U[›äiʲ8Í@GV…q¬ºOð{¬10¦´KS"wIéL s¹xI¯Ú ñDîœÊ&ιvV…q¬ºsåîkJœ«ð7p®Dî’òÁ4Š›/(oŠi6E €Ü#é ±ù,èh]µ9 ˆÜ5é vÉRO„8ë©À„žÊ>ɰ”ðeáxNp5 r—8™«ûA?þ£gëUÃjû$WRkáSQ¯! Ÿ¹Kúôgƒ {é"wIÎmœ½3†sï¬0ŽU÷Iù¢í¶éZG‚ó^h}@ä.í&¯Iæ„W=ç„ ãXuß$¡^´`DŸýªÂ„„eŸ r×D)tO’ÞÉ!r×$ê…ÌœnÈ=SA^´2øuJ^Òi{¤d­È+)Û5ɪ0 r×äßHù2î<²ÈìÞÞ]vMQg×èeÔÙQ=¶¨³½ mQWöMF\Ìæ ±i…ÂDî:« ~. Qºå€ˆMS&w]OÝrÈiaº‘»¤/!'Ù«NVª‘»¤.½¥zÆu™ÍêgWǪû&Zgù>ÃgW˜ ÅjGçÛ” rœsÁ€È]Sí×hŒe¾WgµªÁDQöMý$M&‚VÌgÕ±ÁDoZöM¾®zÞì'¢2¹Ù>&|]ÌÉìeb¸PsœVØdë8ÛÇùùDäÅ3¬ •?ç3‘š0­¿º/0q¢9óT ¿‘½¼AÐ…¹¬0ÞSÕ}’jmN+­ÆÏ'Ǫû–CÇ-íHÞ8k¢Á8VÝ7i‘0u05•‹†@B§e—äž‘ÚBôÏIƒq 뾩OU“?c]ÃYô˜èyÕˆ­›DE“-h3X ºAÄ&¢ã2|²F…þÝ’|.š¡¾H“4&ŽqÃ0´%r7gT8僭}XßrÖ«‰ƒtG¹a(Ö׿p&ÃI{‡â+Ó…PT¤JÊ;mØáÞ…ØÍ¶ÑN9ÔAeœoP¨=c±¡0ìTºÉ™Œ`ÑÆX'—Îb­§'\œIå×'&¼’ó˜š6q c1ÆTÏšˆºÙµDå ¤á†Â)£ûpÂ>ž|Òqá+@oRë+†:PuÚü0L]þÍέg £g>B…4ÛÙÙ,CRÌyNs nepœµPaCáÌÑ#Dᇠœì:¹":æ0³ˆ¨K‚EåÒ>l, Î"Ô¥Œ`±¡ð(DÈ #Ssn2‚ÁcФÚ,OD&¯´uúcn©ºÓu·´uhd~ééiÙÍ¡UOnȵÚÒðäv\©ž:ÇIvªXFä6„7×iÃàÂ#M+„ï8Üð˜Wƒá†¯É"5Q2¼sê%tBÈëÀIU€ô¡Øž0ñ9ì0-•¹} » ˜rN^•¤Ð)`ˆÈÃAøcDvM˜©LiÔP;)Ò¨êõ¥§QÓ;Î-úÞrJ22ûY[‚lÙ¯¤ü åE™´A÷yv®kx jipˆçEÝðÞQ%¤0qʶ·©½ƒ~·=M",ÊKàÕ»¡ˆÄa*K3™ÿVçK­o|ëçi.Ùz‡m4 }‚éăŒoˆïDóRQx|ƒÇ$³RÇáž“ˆ²Y>O~£z‘ÞêhÑÕãìƒZÅôT1xp+ד’¤2e7O“¢Èn¶EÏn¾“[g{í“dÎâFû)nô™ü¾¾KaqƒÎÞM~d2¥SÝÐR.ZQDtcT–.ÚqDJRõu«š±Ûž“T}óÆr‹Y§”¤ú˹!¿›¬ïέoά7ëÓëð Fë¦é¹6º¥’²M@Ûäì‡^‡·§í]:;¦AÄƒŽ kƒèŒ&£²l4'(|`mŽ•¨}Ö«C–õôòŸ¸A–3¡|˸ïŽ5IÄddXï»ËíÒ–TâwYè¹,Zˆ´ºFøúåíÅØøxBEÉ2ø`ø×>(úø7i|@E8n—¶«•¢¾±it›©OB;X&W•6ˆÀŠôŽ(s¬áX( ¡ÞihH p ©Os¤Xd8%H5ˆÀjò0,&áW3 åDÖÑÆ Ñ#¹cv³Õ.'}<‹©XÙBÀ <¡í±4}; nÇ›y…A9mÄJ‡ U½LÒwCx„=Þûl {pÉÑý¢rê-]§Z ÷@h8o°¡`¼#Œ"Í¿w“†ÚEÝ9æÓŸ3Ç3'Ü´CgÏ¡g>\…!EÕ=‚¡›EÚCˆ*z8ÃÈûŒ”Ršš±ÃQóÑéÓÓã,LäPX[¯ù»Ç^ZCÇpèèCtÕ›æ;i*Ã:Ëóùî˜U)ÙÄo¢éO4O'm–Óí«œÑ.W7ã>ÕU¹P•³‹tÝë´­¯ËÝ0§c¤;RDÐgº·EfðÆøvm«]ºzuä—Ð:íëzAË)½<×¶^¥\Zîn+†I~ùœÀ)g„k»«å”Ñì*ÜU¹—­†¸âºÚvMì;ãŠ×v ¬ȩ̂lŠä劜Öâ¢]ÎÓ9ë˜öä Z§äúAÛÎʉ *pþþÚµç$”ßàk~~¤kiÐ?ÈÒ}&¥#ÝUë—Üêm6ºI'woRÜU ©là”I¬§.(ôË\ÇL+OŽ›U‘ÇtcëÜrwSÀ÷º"”“…îÈ¥`†‚)Lž6®èuCê+\=ö¹Ëë±|¶«®è_ŒåËG˜íÉ.«] Ptó¨…ì1”Ç’!Ø¡5–yŠ’®é¡ 2„ÚÕcæ*Ý–aAÝXÆb‰¢Ì—¿º¸üèk[à-4ByÁ®ÙE`j`ñÉ]éi9íʾînó»tã®FÇ%åFî‡ì/ÆòùX²Dq…¾˜^MmV›Õ`w5v{È–p ³KÌîJéw—©åÙ]lé£:ÊoJ>G–—¿P­qÚúå‹ãvE§ó7%{$à’ÍRTÌjE°>Y´y'º—´«&#YÇЪ”×TŒ€@¯1B«ÑÑ¢^²Ûóò"ýÖ„<­„ƒò ¯Ê/ŸâVý8úÍvtmŒBÖ 1u< Pn¶j oràNþŒu5ïdŸrêRËb×ìñ›â‹Oã$G×Wu(‰%`ôWŽÿ~áí8“9ý›AAx ˆ‘arûÅGA`\ßOvS s›³Ãéî_˜„t\kk²…ØèÛ¢%k}ÅŸ~Ï9óö‘ùî•ðmÚf‚ô̧•‚‡®¸€ï{s‹ y'C¤ªƒæX#ÑÒTîÆ§~‡€•wu#)AFb“öŠ5½W£[.LÃIà_Ö¢×vÁõê&ÖQ´÷Cù  ‘7w¬8çM‡ai¿­q6Þñ¿l ,üŠ ¿Nä¼M‚ÓÖ{ Ü$w%ò°Ûs¿å‰ýRéyù=&Zû @F+ï+²ÍúáòRç±\ã1—€\‡ðé—ý˜èîX˜ ®oâÇ^6/~ìõ¼ŽCô[.– yšáq÷bßE¯jN 6º‰JÿIát›K¼o`|dˆm‰åiUXxè‡d_§¡¶þxì˜jó^ ÁGÌ{Ô†°áîãÑ’}Í‹mcCÀCƒ×6üp@Oú8bcŠ[Æúã±|=– ù“±|»Û >Ë»ÚÐ8ML«¬ý}3¾Ý]îÏÔïÆò›ˆ¯á>¾íÍÉ¥¿5¦«à»C%›‰˜Ü¯vÙÓ嘨é)Ù1Ùîw™|ñ!‹‰vsºÏvü”ûÆ6E1çy½»¼ãÐmÛ—cùÅ@¸Ë›±|3–×cnùíîiÜñ4–ÿË/w¡Íœ'¡h_ÕÜh{ýb,_Žåÿă·»d¸4ô•— } ʆ>³‹8t3ÊþCv„ßEHÃ~8–ûS9ƒ^%æ5?b#öÑ€Šå–·ò´_ó1„çšÞ–‰#|uý³YÝ~ÈvWÿûÚK»â·«RúiD“Xöï?Àý(ì9K{âAÓV,ïÆòén²ÌúþÔ¤fÿ Ÿ’*zMÓºêå{õwþ²@”·RÑôÆû£A],w©ÇU“›¾‡º¦7ûvûW>èfŠ›mïLúlN(j-bÆ=“‚~§eÞK²J¡¿²û¬Ó’ô^m [€Ý‚&4À:“Üúo„Ûõ?Éí …s¿½Åù^'lÁr·2èóccF3Ú"¾‡òäÖÆüûMR}Ýjµ‘µûÿ-Ò7Þ" 5oÏ¡¾…&I>Á~O“tö® ,Õ.ô1EZt0¬˜²P祰óË ¤Ý=bÿe‚ûv‹t»?ñ/‚èdLendstream endobj 124 0 obj << /Filter /FlateDecode /Length 5889 >> stream xœíuGì ¿ÏKmÓ„`~ XFÙ(é-6Ö0ÀOI!\’ª÷ žéð.ƨ† U…8j™,(£³tÊମ‹.¦~U2FzíC­€‹ðn¯q¹˜˜Žñ@ ñô¢ @ô2Y ?09þ. gKŠá‚€6¢¢?Óã4#àÙ+$ÿ¬o2ÏØ”z3ÂÚ»KÚÇFÍ|»5 -òÍÙ·Èì€;?|3AMë¼Yk™Ý›tVa¬B¶I‡’~¸Ú&ÆŸÕà¼òÕ5ã³Ý‹‚(fq—h—ŽeÁöîni§šÇßýz—XJ?:ïÉî=‡‡­t%_á¢'½bñ•Iâ•q|kÂmHüÎ׸K‘è#0”óc“èXå0‘9‘!ö˜sÒÑ›éP$&!û+ Z#ÀôZžAôm~d3× <Š}¢áy¢d”‰ù(ý§·¤µÝc–$ˆd`ËH ‚n©ØñjKGÞÇJ`Î œà-’¤®ñëœÞ˜ˆ¯ªÊåÚ2ßÒ@f¥j™rlôÎFÀÀ3õjYläPc`ðè!2‡gËË9ŠÑˆQàûµE§(ÈIAz…n46RI1¸Ï†é—-kƒ¡:'5 EwÌ`À>dAe¬ž%+Ú Œâf¦RÚ ˆºk1OV€€¿˜õ˜™2õ.ÁÅÖ“JËuT*#Ñ®"]:TmO27 Iïzi3ÓeáØr.åæ›˜ZFÏ? Ë YMô¬Ré):ÀÐ`Ÿˆ J€00‹ÂD §Ž¨È"kT*ŠOnøg€±WaM®µv+°^oðÕ Xe:5ºÕ1ÊÉÈh@A4=Ç\ªUì}nºÉ6€º·ŠÄ…ÓIp5åìÀ£¸ÍgñŒ\>ËDqê’c¸nß³É:%©ì$­c(Å­ÑøÀÍ!ù×üøó F øA5ïikçãwá¼ ·ôl°²bœßZØí.ÁÏà:ù<s/Y„_ÓY ŠÂ.8jÌžUê´–éÂI¤„ãË¥€ß?T"]¡aãÛ |xÏap£JbÑé1œ{ß‚*ÍyŽq÷ÌÈÅY=ì§Fˆó]R&•s×aÛ#šN »Û&ùsÑèŒ óöѲç-UuHùc öð /Y#¡Fáìdö-Ð"“ucdL˜µ#vOPî”Yt\Áq.ÌNl:Ò ®ëQÀÞå…%úÇÖ€bEñâR7ïBWVëK«|æÚX-H“ 6£Šb<-ÿ\M±-çžK|¨ݵõ™ßšcQU ]Í]û0<ÊŽÌ#ø¤ePÝ“¼-®é+•Nñ§á {›ƒYô ·×ë)tï Äy !Þ@>Ö±=?T&¹–^Õ’Ë]£Yç"Ç{Б"²Ü!‹z˜”ÓbVÖ'Y@Gí.K„xØ;.AuROÅåj-2[ËDŸAyƒÆ(/–ÈD,ç.³jÁÊ1$èCÐÇÝÁ- œHRòÙêø'ûÙ`šP¨ŒYÈ—m¥výlȧeì¿k34ù÷4Bê™{½¬¤á5R|¼hê莌¬†¿ù¨+³ÛÓÿ451óìˆ%%„æe[RSL×lY ÂBëy©ÖB%8¥ázអŸQ«R:µF)›cÚ…˜|O_—èׯ¤÷‚—¸¶®âZ.ˆ®|£¦Mòá‘A™…©Ø—è¤û&58''|úÁ¥bÙ€n8´ÞgÐjCðS„÷&ù«“ ª«ˆä5M?I–¾r»Û¼– \àGJ® æ«"rN$©äÌ»çôVzv׌¹‹¿“¤ bþmÎ ¤ãY4‘ñC$Sµ ÈÛ®ÎuvS>˜6ZA0™x[ÎT¥p{ ý‚ƒÝ¹§Q‹R×İÚÒ<—»of‡ªrÉ‚ÐË™‰Éyæ 'S<ÈËébŒM»÷7‰†’ä(Q\˜ÿC9ʧÿÉÓd\ ²o j_9ð"„Ê'HfdÛ,G=w†f×*¥˜að/ŠO0r?†iS2rÄ)~—|Ã"ŠÓ±d]qئªvɧÀ(ºóÜ÷´ „…ÅÆ|F-1poG? s!‹ì"÷›²–ÊUúœTŽºáú,üÔôÙÅ*©³\à8CÈÀ 0©=y——¶3Š|£üSàPW/žÚ¸+Ö§œG9¥UÔàƒ&»§Úû’Ìf‚Ü5>ÇkAÇ‚.Ô›±²wUú)S3»„´T^[HH~JÆdSºç ›é¤Id,÷ÓbʸŸ@‘-,bÓâØÃÛw"‘¹,}OÓPI$ÆÕ:¥­–Ó0 teæì7XÒU 6J¥«›³Ç…g– ñ€(«Ó¯NN?{>i¸NÑuçQ¾[>Лzí»!" U¨©(#WRüh×ô"¾ÊÍÜÒs'Šü×@0·†v‚#åŸ\°¾Rc×ÅsªÐÝæS[‘ƒ §Êï» ¯Ÿ£k| ºqEêÆoý®¿W9ˆc´f."µ í± Û4ý]»¶Q|¶Efgö/ ðJðÉzZˆ0M~{8 °BœfÿS³´l àL·ü;&øØ6¢ëD@ž1˜›Â¡Çhí¡ ÙSáyüIÞ—áE¾,ÃËðM¾î¾ÆÖ}6AIÉÏ}xÎÊðUZ¬ÏlËðª /¦ÅjÓiøI™pY†×üµyø¶ oÊðew[TSŠÛ¦¤În vå@ç]ì\”üsyº)O¿,O¿*Ãÿ,ïËð›2¼â[ÌÛ>é/[bÀ Cå»c¸+Ã]w±ûîͽíRàYÙͽìÙºŒô]Я»00žËH;°0;ÜeHBdºZ“2:â%Ǽ®ù÷,Ãߕᮠ«µæá}ò1åUò~<¯Â8ö¬Ë±·]õÁ¤œÍ=ÊÇŒ ™"¹ïnÁyóewƇ§GJY/Ž ;=t9àúØŒ6e؇íö}w…×eø¢ NÅy=ÈØn™ u_è}ÑÅïx̰áyŸp² Uj6ù  õmÏ×Ý o±bŸéØnÌ81õ{Ýû¶Õ®ŒW~²úh™ƒ½v_hÿßexzì5¦ìãßtçÞò¹óŸ½§òümˆwß2‰™øäë´^ŽBr®[rsf‚Ñ¥ý‡»6ÌŸyÕÙ-dÆoÅþö]îºRm»‹ÝtÑ~Þ]·¯Çž¶B†³î ÇôNÌŒÚ ™-Ù¼nkoËð¡k wdbDø%×r½g?ÿFU÷z7Ó-Gˆ ÔMOºé—eøY¡÷#†½€7mâŸö§`'Åu,ʪÚïKF~vìB?_°ƒ½®î·}è 7ì4–Ø-ë²t¯Ç›…šM–%oä\öïÞ§;ÄT·Qéz)âÄV‘›Sí²ŠÜXo ¿Û5…n&´wLº%³Å\` m@†ˆ§•óîsLsJ†¦³zŽL¿$>·°iŠyTõ¿ªSŒ6G òJK‹Lõám¯ó:¾Þs ˆyÛdÐiRNÄc ãÒ_M×Ê,ÛN7_ ÖDP÷“u2?gØN÷_”6´©&*ë^…Û5¬ç|{Gë<.igtB•dÉB01<®¬BUEdŸ!`ŒT5b•&L~R9À«*YÐê¾Ô\.·I½¨x.ášdT"Rtßc&®ªÚrXñx™}~‹÷­léiº‚ºpã’ƒ‚‰“žó’JŽ\Dò‘ê–1&ü2µb“Ìq·²•¤DÛŒ´Ò6õ×™@ƒW5äÇ“2ÜÅ…º·jYg½ÀŒÅèéšmÅÉ“IxG ëc·Xlp¤<æ^’©yó‚E×0{­ŽÛ+–áÛ‘ÙôV›ZpfÐÎèPÈ5UGXšêÛWÊ…®JJ;st¬›n3ÞD¬¥;?U³~1ƒmÝqjgLWñ>ÔPP"`üØzÏ÷Úí(Êó·ÕEnvé´î*§V\H ,ÞsA²yE î¼kâ¬è2Þ‘XLDkÜþ¾Ó5\¯î\A½Qjš¸¥ Hü½…žŽþ%”ê»!´Ú¯Êgb_ÿ8ÚìËúø«»~ h4Œв陯:â½J çýº±¾éÖÕ6ܼðPH÷? }Î$+…ëbx+g•Ýêdºóq¤ézÔ;ܖwné“t~§‰ ‰Yw®¾&äÂK=`ýn/ÚFyyàz4#w¶3wmgÎX‰|'=s\¹¥<‡™³Öñþ]e\šöêv¸ Z¢¾jŒ²!dòù{½p‚ÉÙ®ì“í˜F~Óe’öv ¬áÒEÓ+Ò»ë4ï} ¢9M`uE,ÐMû-t#ó¾7œŠ=ÝnÕãŸDÓ³MSÔáënɱ]ºîF ÂÊd`º„È€¨«n°™—àŽ½ØcÔEv»îúcçÄ,ô1„Mc“7ß½OF O0.^į™+]ÝÖ™û$s³ÅçÓµŒªí.SÒ™†’Ó-—Kò8ÈÀÀ®ãÓ÷ÀÈF_3{UŒM!DuyÿøuZþõ JÛ|»  ‹q˜ò!—Ó‹ö2!‹²ç/ý[ÓIñæÿGVyÝfØ1ÚÏÁ»ÿfgR£uïœ)³º£(½­ô&~dWèÚ‰Î1O›ýHÚ‰Ýr©šeÙñØ”îm}nt˜0ô{su‰ŽÁ«K\ùë­|“½ž‹t­–qê’—B·­tØë™UTÂÓþ –,«¾—6=‹Ù­rù{[š3S‡{WµªÛŒ™«ïèJx2t”ѾŽA»9¸®»T¡E¹¿(RÇò ý”«Ò!‚p<¸™µËŽ­Z9SÓý.Qºy¨8%1:º,Ë7~s»ÉoƒL©µrdv7›ÌUÛ˱™Î½™rÂMy–êQmK«I°§¬Äß/í]tW¸?V4ë7l<†¶Яv²ºæûÚ¿îN`%¸± _ôwfÝèÈîuwü¡ƒ¾ßÝÁ&°««_Cû^kFÛ'Ã(»ënÌk²Ûî‘NúiVßéÈ9åØvwÝaÕLÒ]⦻Z¿žt¼èz×]íu÷i¿ ú&ž4™ s.ñyùýY~Òb¸­t±¶Ë2Ë0Ÿ-_†ý [Ëκîº+ðÎ’OÊðÛ2|¶ž¯ªºà*‡€Õn™ô»I~èÖnYA†¤­ÂªîÓ¾\ìIì'¯Ǿ…ßµÀöcš)Pù“*ÛÇ*üyȸó]LôkÁ}ÉcUß+^nžŸö;h˜X|û¡èUÿÞýÝj´Ÿ»óç\$¥&µ×`k,GM¹~Û.l¥\ ‹U—“‰¸œ¿çuòÞ~›f¼ñAŸaËÓÿHSœ\øü\}s5ÍÅÏÙ±å0{G°1:*ž®ïXw¡¢ó¨Ô)Ö|PêtÃîb²¿¬‘¾‰ÊÀ>ðYºŒÔŸ.{J)L¤Ë|¬i5G‚]®îñìÌZáwº}à;šúßÉœw¶Gž6ßtì[ ïóä#ÕÙMö1‚ÂVo¨dÊåïõ;H‘>ÇRÝǦEÒÅkvk“Ç@Í×ÃÛþüt¡-WLþ¥h «î°ßÓÄ&ìõUµ*꼫JiÕ BÁ¢A„ÏÖlÕÞ—ådjXéãO£¥L}&uq§G$Êv§ï¾q¹Í bÊ4…›¸¡TÇ7¬Ó,=)©súUÕsÞrïNAæâ©1Eè|§€øsáK4ñ9Çû„)c•¦ ÜŸNþ"ŸTendstream endobj 125 0 obj << /Filter /FlateDecode /Length 4135 >> stream xœí\Yä¶~äG4 PÛZÞ‡ð v¯';½3³ãµçò¾^ò×ó)5‹jªÕˆ~®ºHÖñ±¾"Eùû•èåJÐÃß³ë“ç/¤]]>œˆÕåÉ÷'2ýºþœ]¯Þ?…„6xÒGåêôÕIî*WJ…^;¹òÖ÷QÛÕéõÉWÝûëèUŒ!¸îb-z­Š®{DS áEàÍ×$ìcˆ¶»ÁS¼t®Û®7Z«>Û}‚§*8cCwŸd…0á_§9±ª×"¨ž™€ÁÂêôü¤û÷úôÛ“4¿‘j²ß¾êž0–‘NÙˆÉ6¤Yô^uÏòl>˜îãuнÊc2h#  Ó¬q¨uÎúß'REK íBw¶VŽLÝ7yxe÷ÙuÓ1hë’°ŒB)×=›‚ÓÆÔýÆ)È^D)ð(#{˜‚8Ý‘ì~þÂj.²1pÁsdvÁŸ×«tŸ¤IctÆÄd&"¦Ø ;4Jó¢Ù¶­ÏºÛFÝSóuiþRšŸ7~*Íg£>¡{Qšï–¦,²¶Ùí¡9Åus°·ŠÀ]inKó¾ùôºiñc³Óá¬4ÙÄ6&€x]ñª„ˆ»ý­f”#·i <5eï–ÒdËMÞ¹ýí¦{nJóí"{Õ”}hº½/ÍWÍnl°‹æ`ƒƒÇÀðvK£Û&<ÎÛ㱘¿lqÛ4jÓxJ Ñ ÉUÆœ¿â½rè‚DŽ*ÍÇÒìKó¾4ÏKs[šo•æ€ÖÍ8ó&å´öòî]0ƒKX7‹‰…ÁïêÍÒ4wÿZÉt –zX|X`p¸h ´×ÀåRÆbéä¶9[lepH~ÚœùqI᳦À]s¾ŠÞâ^kàwÏ¢¹«Ñz[a»Š@%ÛzÊþPAy ëNO¾@qˆr¥qe„·}4+#CD…µRN†Þ¨ÕýÅê«›Ûk…]ýˆBܾú¥å''RY*ÇVÆx”vju'"ö(aÆ'W'_î¤P Å¥e‘ŸÔ¡RÕ©•ï•GÅÊ*U)l/‚ÆT¦wT ]ŸtŸ¿¾IÕ"LRJù• ¾w.בŸo¢ß’Ùo>Õ`v)Óç¹ÁÛ‹Wk-!i»þëîӛNjû³‹»GJ F ‡O„N×@Š`×Ý_l¯/nÿ¹–OPI^ h1ÞÄý^W·—_wnÓt÷3M ]/ôMð‰äÓHÆyg»;TãÒÀL¯»×·÷$j¤=”®ž…Å$_ðPR%Ô£5Án¢=´#€;ÒVǸC-¬{ÚjÛ{Ã>•"¤!”Ej|Â¥°+°½uLj|RI‘)^r©á —2Þû>TŽžThï•2ØËø´£qqõ¢X„œ¦+`azc†ž9{ êg4ï_c)p…  I^…£ÜÙ£“’Ò÷Â/êå¤49× 6b§¹¤—GNÁâáSAŽ[tvp½ô|Ê äèÄ"åcoypÓšž,)l¤È”Ž6À"sDˆ,´¨ôrQb'º8£Ü©b$åQ¨ˆ(r•®ˆ¤_šá6ôâèT ï|¯m%…}³]F…"Æôò‘0uˆ°¦åÎi|ÿ€úÊÖ\UrŸšg£ 0]%P§gsë”XG‹düt™˜T :c#pÅZ‹¦DÀµ€ÃÄ¢ú#°Žê‹#à:Aå1¨zLŒð)U ú–[4—B ¸È‘øúö†RÑ’^™8Z §#Öm¦î{°M)žw¶øë—ÕÚ1JÑK¹>êIÚÉ4À€¯ˆèôþÊ4ÀÕ7›§§DP…Ñ•òíÃ2pä 89ðU¤íq‰.sD67ªÚºŒDÀ6%Â)ªQg›Aéh;C5%&… 9î-Zê7öT£bs³¥ò”Ø”N© žò@éˆiÆèNy€ë¥ÓÑÇ¢Ã20½Õ ó¥ò”ø”(Ìì,}O‰€»‚Ê|)3eÖ3Õë˰˜€ë <Ô 8ê°ÝÐóÏ” X„УIS&à¡´¶×Ëà˜€{ŒSÇx,17HqyÊÌÕBÒcj˜0ŠþÉO™€» D±Àð•«‘ë–WîH<¸HYºN<#´OCê•©àzï`¥J<põá0{D 3paÂyÒšRbv Þ™Rwv.v–¦TPÅ^]N)#0+Ša“3¥‚reœA$gõš0;Ã’ø;Ô#S&¨NºB9jh?a‚k~jÆÒǼb™ øŒ(ÌÜLO˜€Íh‘v‚©ý50—’¶7ókmÂ\1Ðë=ç” Ø”>Øq;sЙ Ø”ë;,O92®6GL92‡ ¦ ËLÀˆ•¬£ 0Å Ü{U?2×Bù{Î)p»ãÁ Í” ø”VR‘¶¼tÎ l!ɱ|šRArÄwÕ™ ª€Ë | >£\1PÇ+ö¯ÃÆzJEÊvü<9ì±ï:zÔìAM2O¦Ž}ä8¿KŸrS?}àí^´a ©ÜJ#OÓÛ½ô¦íã5¬BEÏïc]>Mnli‚ÝÝÎB‘²3ï¬7Æ ¿KÕ} …Ó^òQ¶—üâØmº·½Õ¦¾ÓÆn £§wÀÃݳ«r£ìöq¸é&Mwûj¼(VÝc×À²²(Ñ£é^®7%<5èýêW%ÓxŽèl~£›®«U—ïÊ-ºn{—Çså¶¿ –úKï¼í¸óèM2R[ •šÙR§ôv}¸Èw{?Ü­“æ€MðŒHKùš›ª®R!‹@/ÂÙ[ê‹Ò|]š¿4¯wÜón­Ût»3n÷¾ý¼DkKz‹`…QºÛöÉ¿BÊòm‡Žuº‘·UîžnÀV 'ÿ˜®Cf<~“ä§&]&Ù¡Ÿn‹y¡¼ X„a ³o3lƒ‡»9χ…µSA«tÿ(Y@Ó¤Åé„r‡ÞX´Û¤¡€+U7*™«,)kêuÓX"Ù“’®Žr7õ”H€Éo”¦µ@3c=nÓ"Šî½ÊO\ŠÕ›î~xÊAfÄâ7‹|êa4/5Ï-Û„ ±ÕVÝ6;ØËX»2#‹B-§CŽ}3¸EÒU&›µ—«@Îc â±>û:øìš]NëHJdTSA”.2€² ÄŸ .oÚ #ƒ\t€ü.ëÉÊóɱÀ;Á„ÁòqM²‘®Cäé|¨~ß)aÚqþ6-\åêœíùTzëäpë9ésžçD #dÂh¼¹+Ãc¯½e›Â¼¿&S>>+o9$…‚‘ÂKõü©'Xº5oLVHDy›n?G-“±3HM¨–T¦»Õ•]eäˉ›bc¯"ØS8½Ú‡²ÆFVF „ó(†’1WHä7•®zLŠ˜(¤¬¥oÎRý¼•1WµUiRnvš¤¢b¨)>`§~Û\¯ g'%¦Í¿G [‰9Ìh@ª‰óÆÞ8-«*4Ÿ¿¦ã^ÿ< ‚róý<¶¿iÆC…/Ò§=t¸„Ú«E•«Ùé,Ú펡^®1¤ ôíÚBJj›küãxd§tŽioäóÇàÃ)úIôsr®ßÃzÜ'©êù=uÔÂR©¹ëHg×Þ’mä4*K ÊÒlo€yFw_²} A9ïþÆ sòåp~÷TSI”³Ñ¸|¼HGÐQv¯½£€äÍÍðQú ›ãl-á!î•öi‡‘7ªÌnÞï¡8©…0 úôc 5¸´;N†#*RVï¶ù )(Ó}·¶6ÍÒmiFÂøþ`tNÿšÅ|–¢Œ3ùtcÐ9­ì×ã‹Ú°ý¨ÑÏ2¬œ—Ã1¢ KÕd³øwõÏØ÷C´ÙÉ\±ºŸÿòJ5ÿRÂW^®b ÅüFè›& ´?ç}§4Ÿ7›”æ‹&çüµI)LÞ2®´š'M–Y®ý¶I³ŒEÛ ¶ý½åei>_b}6ÅwM˦fï¾Ñ°ÆæÝ~õ§µ£÷¨Uœë|–þâä?­Þ¸uendstream endobj 126 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 1074 >> stream xœ’mlSUÇÏ]»Þ‚m'Ò¡½ òdƒÅ, (P™ÉÂÆ› …¬¬w}™½íÚë^ܘÀ²Û§XÆÐL)ˆ2…-lp˜ÜÁ˜ˆ&N³H€@¾‹0 cp”¬]–?yÊW_bÔÙYê´L|?±1{ÉY»zŽõ×½F ˜t`Ò÷ÍÖ7¿€ ÓПƒåÏÈ‘ýŽ`¨>ìóx%~aÅ"~ÙŠ…Køåùù+ø5!ì«p‰|‰Kò —¤]Þã·+|‚TÏ/|Ý+I¡•yyµµµK]ÈÒ`ØóÆ¢%|­Oòò›…ˆ®Ü|QP”øRW@à3e.ÍìŽ` ô¾$„ù’ [‹„cPªÛ-øfB6’2²8H)&¥D$S´.‰žøÈ9f “Êb³6g=Öµª,ê–~uzšéF+n@«N­Â ‡¹÷é,Êå-§óèËVâtœqoy­¦%\%Ç÷6§à[8 uœ9|þÜç`RbGU‡vÀNpÂÖ}¾}îJÉF‹:«¥s’RðÍ$ó î):ܦærhV¨·8 Ñ 5Ìo3znГ,%c57¹vó™‡HtèSÿä=½®r¯·¼ü¬wàBºwÐF×Sçd‡'{¢£ENötìrüôÀU9ÁQ¢.`-?hVõ3÷5…ÿ,ÝDG¹Û‘ï¨î­î­![Ý×Ú«ÀÛv‹ŒÑ'ºú_XãîØHü<ôÁ°Ü1.BEÝOÊ•®dkž,v²éØP¼.ˆœüß’É´2メ´–ÁUZ¯ª_K«9úNÛhö·WŠ@“†ÊDÍå4Û¾`“óúT+êŒf7 Mø¥gááS\õß/,Mk“ýˆ…Ï`ðüét÷6WG¡SNÈu ²ÍðáG5]M'ÿk0Îg-{;Õ¸ëØáNÝ™`•©Hž³MÕ&MSú&’^“9f²ò7“J%íendstream endobj 127 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 5114 >> stream xœ­XiX“WÚ~c$¼.¥ŒšÚy_ªÖâZµNÕ¶¶*ŵ ±;²…} Ùó$! $€lÂD@PÁk±¸´jmµVëL;Úùf¾™n'xøñ ½ê´kÚùæÊÅœ—sžû~î羇;†âp8Þ›·mX¼Èóíb÷tŽû©1îßq3±ñ¾d¸ÄëwÔãkVýnògҘȅ‰cÛžz’닾˜„L£¸'(.‡“ž£Ld&%$fûÄÎñ_¼bŲùþK-Zá¿:5>3)6:Ístvb|jt6ù!Å?4=6)>»À?àåÄìlá‹Ï=———·0:5kazfÂ+sæûç%e'ú¿ŸŸ™ç¿6=-ÛKtj¼ÿƒ.|ð˜ž*ÌÉŽÏôßœŸ™FQÔÜÕikÒ…™k³²s6æEoŠ)ˆÝ¼'$!ñ¤Ðm)© D -YúbÀœ‰5ƒ ¦fR!Ô,j+J=Km£¶S;¨yÔNj 5Ÿ ¤vQ¯Q © j-µˆZG-¦ÖS¨ç©ÔRê÷Ôfj åC¥SS“(?ŠOM§ÆQã©Tj©'5–rpB9ߌÙ0f+⢱§½½úx±¼ÛÞMôÚq¯Œ»8^2þó Ç&fM¼ñØ‹]ö‰~|Ìã}OØ&Í›tß÷”ßâÉÏNŽ˜,ŸÜÈŸÀ¿0Å8åý©s¦J§ítk|†EàB»\îõuœ³ÃqÑ6äÅz5J¨•êt… £5úÓ­F'¸ ]Ó'&Î4U@­Mr=«Ô'Èçe&{YõÈwÊçÙð‹^yYÅQÒbÈ/¶ƒ‘é‚c7€NM‡g±Ù5r…i¬N-5ÌÓ9tFD V(¥t æòð /ÏóqŸ•:Ü ”0„6qQ! 䛿z“£â$è«[+û®^úˆ!"ws$ž Že%IšÈ¥#[ŽÝí@sLårPH@[¤fäBᜭ@çêª*›ÊÖF¶n …Aúæ6Ç®oçKÅñ¤¶(H€¼lQj~¬$´´¬Lz½ÃÄÔ\©=îÚ iR‘&_-bWá­R§@j-ª³W[j-ŒÏp()ç»tÐåë¾ý§”?Nõ»ëæ»GøeØ4Ï»U¾.å…WžÁ<Oã¡ä>òiþèsø˜þ äæ0øO$%…³B­« 5 U‹Y<}¤ÇËïïIr¥|ÝôÑ_—Á>}ÆC37òþæ³õýe,öÇ×ùø ^ -fc¹ÞÌ^A^hï®ù÷s_š‰Ç°>nëÃŠŽ¹‚ìW¸n3rñ¿ ¹¼?ùxž=ëüæÏЈOü,ƒÍ¸Ž·ÇõÁ 'â¶Ÿf»†ŽµúxKdHTάä`6mÛΈ]@û G渆¸8hîîðÉá•|}©Þ ¥´Ag•iU -a^_uyë¾pàœŒsp1^þý,Ä 0ÄA{•Ë@¥T€\ÃîÄ+ñ¤MË^öü4ã\Z|a/ýùÏýý]ç¼¥ªdаXlF}u)CN„w¸ÐüÑS¹ƒßã~'óå)rMœŽ.! T˜T¶™Ð·Ú¡Cãe ¡¯]â`.7[œ¥ûØN4ÛRØs d?Pšó}ðñà]Q Iilñ`„=óW°B "E:{6›oí8wý- ûÊßHŽâ i›Øą̂ @o•ݶõFƒ•à°BZ._¯"{žs‹‹l¨—¯7ö!þç`‚¨'J5É ¢%eùÎWåÁSm{×&„çˆé@\íÞÿ£m–V«pÔŸpô¨ õwp†_éâ¢Ëèþg¼‹eEÂBµˆ,Ò*tjÐÐÅ6(%5®51o!!Ô&­4¦@ż&Ê$0­ä]@vë~‹~?)Ý8Þ÷Ž”µYš|M6«yyP©/­10 (ÈèTB.FÞäÊ @bºA—è@¡.bñ“#ûåI„ÛD<ä¶‘íCÆ臣š>devY ˆXo X+JÁPZEª8_êxûZxé„Ã÷ÚmÅäûöî§úÝG€ÂùGd}2ØN' _š•³ÕtRÄœ4Á!0е’ŠÜìôˆ-ƒ ÝȧÿÎÊš˜2Æï»ts‚õð“õ—n4wÉâlL²Z.-56V;í‡ûcnÀãbæ…1~÷á•‹â%íƒn“~¸P`áÍ;\TážÅ¿Í«¶=ÜíK¼.PZ R»úB“,S–K¤(?ïõ¯ð‡>¾Á Ä=:µZbAøþØS?ùP¸ŠAW g ÞŒ]õÐBhÚg}–rìý›ø¹!´xˆë.wOçÌ #m“Wå¥*ÒäZf îS€¶D"ÈOOK:¢äÈþ¡ý_4]b­õ¦J¨¥OìuE-LÄÊ2T–2VWóWƒ@ÛlšŒ,™PÇfe¬†0zé`Ú鎶¶F¦fûae-tA}S•«¦Û>H :ºU.GɈCó_O:š+ت#Çqñ¯n±‹sïk®»ƒ(¡¥ÅbíºŠ”¨Li”³„Ìš$ˆ€ÝÆHОž­ŠW 6¨Ñ:4VÚ#„˜Ù3-ÕAÛ ›ä½,@w;:šœy¹šxeÝéØ ÷þ|õKë3<ÿa¡…C(ù8×ýz™¯H×傘–¦,ßUœv~äºÉXjMU¿²&áø)È!$Y“œYíé}šj"-*ÚNVíéž…wàPoÛûm—¢…0@(v¥egfgK, ³‚iÌ0&“6ˆMžQ“–ó ÃE½‡Êô¶#ì!ô²ÙQÞ¥Ši !Añ`O¯Ó:|u ½:4Õïcô³|Û®®Ü¿†K¨é&+’T:Ưïp~ò¾ä阋Çãùxö ýAWY¿;áH}Ïqú4²ðwã­òäÍ/nÁnùÙ¶.4®æÛ4Ô¢—ŒÜ¦°h­N¥U’*—k˜ãâœîæº5îßóÑëhACí¥·ß5 ¬úr=TÓµIÁäB¦&!Þ˜äZ)|i™ÎÄVêì*CAa‘H‚¿ÅýÓ®£¬íVëyÔ—AMaP°ñ£‰ƒXˆ1Æý°öçò§Ò ½ÆƒÐ½šƒžEïP@ÞV°–™­í¦öš…X2>q»¹´D$“ãuÒuR‹ÖÂtÂAc´ÏÓ:ª@ÖÒ*ÒÓ%l‚¾ØRì¢}ÜQ£ìB{q÷ÛHø^Ì÷è;AÃÃmL§Ó0òbTùi9T3(dtLÀ«ÓÈb ^°ìкÑX4ÍEÏzÆN|Rnr£ñágî­¿z¤ì@+â±–F0TÕµötwtÿÒhñqàèªý÷¬:©Ã÷o·Ó†ªÎLõû ;ùñxÅêÈp ßw›ôz½EϺLöfØG׋lÅqÊаޤ·¾ùòë¿51u3z³(‰àQ%9êå÷—4ElpÈôÿ¯ˆ•íWcEŃú¡îë}ÌHôm®{û1~¸’´¶–ÙD ÍŽîãQ]›~ð&g7üMþì«Lr›\­ÓÉäÌêKÔ…@oÏìhþÖyŒí¼xâ@Ñ>mU ísòj¾{ÿeîýIn7ÿ¯h¯ˆ-dñd^h@Z€ïŽœ›êÔ:´E\Šà—\Øä»84¾8B+Ku%f…mƒ£ËÓþá˜]¡/fß0hPF0—îCwË4=ohäkC!(ª@P&‹Ùæ7Ì!ã§GÎÿ”¹]„zÝ@ø«éú)sõ`|ÖÉ7ª \÷éß@¾­XRä6G““>ÒþGê÷°¿^ZªÌlH#q[[^ÚôÏD ºšôrQyäƒF§Ö*‚.t —ê¼ËHÕÍEr$â·Á{Æcµ×Ûá"Xèþ=7ÍÀ3gã ÅÇ^øê—ß(}ÿ`wyŸ¥™5;ø‡Ñò¯å@÷4“žÆ+`/%KŽsx™“ƒ¾pÇßæ·ºCøhù3ÿƒ}q4ŽÅ2\„çÝóGO¡ÝhÊAJfäÙ‘×ù‹Bo¢ ý•Èÿ òaî}ùá ?ù ?ù|ìXÄnÄÜ%!«H·¡³8ÚQ{ÃïB >©wp2o¡sÄÆµ‘é“W]þ]+y•kª®X»l¤õÚ QªJœytNl_SSÝÃ1 1s2ñ”$¢xŒ×íòVUì9Þš‘qÅ›e1Ê\’MŒ Ïû’üEÊdÿøžÃíã là–ðÒT¿oÑ—üÜeÅáÊÜô$AìîX2Ý%B[Õ^O¦EEImFAŠ8.¶«àÖ·×Þ½VË Ž{ci`¥£¬8[ž¦fü¾j [¢¦ã±Øà§_8ºŽ °oOÂá^»¾ˆ*ùx£ }u6±E<¤4nT•B%™ª„p„(¤y‡ù‡8÷Þùm;i„L#U vtî9Ö]×ÜTÁT¨¼@ò“‡b*%¨4Œ"³0(?:?Î:ŸáÁ ¡ÿ|%u öúÖ‡/Ý;‹„Š4m ™?»&êAK¬ülÑ.·¶Öí ŒLeä—Ãj# víMúÃáN £à3œÏ‘™‹ÞFÁüZ¿Ñn0›„ü£,Àa„ZPK´JU¾çÖ 5žä‚¬Î]w×Ý€¬üªr°‚92ŸW$éè#ÍßqâpQæy¾CTŸž¼7uoFM¡³Ãµ¿ÍóÇð4RñqçÑÅó\÷)·‚ߣBS0?ð¥Ý˜ÌìZ…a ˜EO¾{¯=s‘é½Û–À§4>5¢à'GÕ}ró*:và¶ëêiGн©›VáΨùlÎîðaûl͹bq »$ÿlEšè@3ßç¢ùñO|þa¯«-¼óå9@|y¯<½¬ÀóÌQ}ס‰Ý5%$J•JIè¯×2æâR Ói{žÞR,3µ\¼ŠBOïQárüÌÒe!L„0:+:,h™4t£`Ø 28¬-5$m@K^S¡³À•ÓCzÜ»­±=t_"5+މ[¿Ò!Î**M°Æ•¥€J´Z=z Vžo«os_«ãq¯áº£0þèÐ4ÒÎ,½&U™Q¨f4²xÕïï¿È ¿¡OÑ'üQ]zRJrJö¾¼–W»‹Ðe>6Öß­ã ù'¸¨+øäüEùö›òÛpÐSgÝÞH¢*+õzR‹Ú¨”bïñŒ¥ ƒ7¤àmà!e¯×„7­?¼µhsi©™}ä±·Opݾîp>œ-9_p"ó½¨A1ÐJµÚ½EÆž^a}‚?2â8J,Öé@C+J5fòþ͸à1hƒ mƒï h\䇣7H3o´ŽúÈ»ï 'ÙJœ$ÚýÈÅG©ºê›\UÌ¿û×ÐÛ£!Î °) Ár^Íݺ¿TœG>8ªiÊŒ,8B>Ïõ¢£¥¦û*™˜^‡Ÿ2 ¼6þx9uŽ·ËvIwï(2$;À$5?±¡»äøâšžia|rî@Ê2ìsðp´ÍÛ5þÒfüØeuÇ9Ê&N¤¨ÿ¦Ñ ²endstream endobj 128 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 525 >> stream xœýýCMSS10/øøø‹uøýù`B@Þ€f‹ ‹ ­ør÷n÷g÷bou„Copyright (c) 1997, 2009 American Mathematical Society (), with Reserved Font Name CMSS10.CMSS10Computer ModernRS434OÄýøˆù‹ ÷¼È÷˜Ãëá÷ÚÜø7÷Úò©ÎÒ‹Üîûâû#û…ýJá÷Ñ÷2÷PûÑäüGø÷˜÷$÷ÍPDDKPû ø¿uÐøïÍÃÚ÷±Úøeù.T¬TœB‹û4)(`™a³aµ^·Æ}àv•ˆ§rŸz¡j‹`ITI0b1•Ò7|:ãTÛ}Ê‹÷çð÷Ìl»u£\½j”/¢Q™{m¥„‘o¨‹·ÆÁÇçß»jl°øˆ‹ ÷*Èø(Ÿ÷¿Ôø÷?îÈ(ø<0û‘ü<N÷û?Úûž÷|¿á÷!÷‹Äüu¡øPœ÷W¡›¡ûb—È Ù  7Ÿ ÈŽ Ù– ¤¿ÚŠendstream endobj 129 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 2659 >> stream xœ•yTWÆ«hºªT¥,D»1ŠˆŠ€» "È"û¢a5È.4* ‰ š¸=Ô‰ר¸€‰ ÁÆ ã ‚ÈÚ0¶†D£InõyÍ™y ÉÌ™çœ>uúUwßw¿û{÷Ò”±EÓôpïÄôœÄìÔø8ÃÊN´¢ÅñFâxÉœ¡Û¯Ë•ŠL(d"A&ÆEãGl7‡GÃN3ˆEIhz®ûvÛð•S§M³_ºv]nfjrJ¶µ£Ã,gëÕ¹Ö~±öHÌJMΰžLþä$¦¯]§LÌÈHU®Þe—‘õßÿ_(Š¢&g¬Xç‘™•í½Á'nSn¼^bRrjhXºrÆL‡YŽNÎsæ-°>•¢© jL}@…RaT85•ZAÙQîTIyP3¨(jåEyS¾”35›ò§(3j5š2§xj %P2Ê’²¢†Qç©áÔâeL¶í¡*i¥÷Ñ¢Ñ,£(£ãFewÉ&ÉmãÆnƉÆjã?¤yR3…ñd¢˜o˜'L/;Ÿd«¹9\2÷lXø°õÃJ†U û.šŠwJ iJQ™7¶¸vÂþÿU£X(è ;á:Ë×þQÓØ]w=ÞOŽÿÕ!Z³=¾5“弃 ŠPƸs–ìÏní[èÒ¸Ó#³b©€m&b쩱©`Ûû¸ïìwØ^Qà*¼ªqÅcñ¨`×™3BZ€þ~‹Vaª›‰T:O UZ‰(ƒr΋R,ap¼@ŠY.ëÉ»ríà™­¢]+Ñ|× ùVŒ¯½‹3„N8ÇT„«³êã5@ÃBð˜ FX¡à&ì-:kß¶®ºasV±ð”×—§Q™¥úÆ…úïK3ÂöÉMábv“hWOwuHà Ñ7&ÍÓ<¼~Ð[3᥵§Oîû{±¼‘ݺ{ÛÞ<Ä%ç*Sîf Þ¨ ¶nªè­8‚ë&t ¸K«SÎÀõúP?c@©…®Á5Eœ0¯í€nÍ2~“˜3V,œ‡«Y¾Ò:ÔÓÙ'îë'r‘§·ekC•ó-H}^ÝÀ¹³Dáo^I`…N)LP¾bƒŠ-ôõ8Ú!/©…í{v¡ˆËøøø9ºCŒî1 ‰3a´Á:ÒA®ôÀ•?^´%³|ÿ“¸˜b+Ì;ááx‘SáòÊŠò˜êÌ»è ª,®¨ãÖ±Ø:a¥GRñÃò9Ì?ìZ€A-軳—¯­>^‹ÀŒ#á|Y>gù šê%Øó.qžÑ Œ©~ÖMÔgróæ7óºÎy¤åjÁŽ?O_ØÁðouJã¹;Ò³ÛÅ·­t£V¤ÉUÓŠGT½ûØÇ—ã;çT`3âž³í|láXúS¦|¢´Y×’Pˆ%JȉOß‘“„¢§R®¯»´õJA%iÞGW•ÄW/ëL êA?¨ªJÕåkIj}¡u¶çñrµ…Ó™´bTÃ=­Uÿ \íÊirSl; Ç´øCg%ÀŽv}Ã|øtÀJË ¦b §¾SöäªÎP´?ýŸ±²Îð“ñWþš,í,ÿý³{ê_Iô–ãà ÃòQi‚a)N`_…ßžê—)O¯‰;ã…8ÞÁE¯_åÇ=lïéž÷}/ØØ¼'".nTé"È=Ì% pþ@¾ôùº|éPÏxÙJß胲> ë’¨LŠŒYž= áqÞ݆°»vRí`½áUt<©:¨Äq.Ư«ìð‡xuÌ´©3¢ß@DUõ¿V MŽh$p„ED£KfÓ±dã'x|.·ÜÀXA·XCˆŸß'÷÷úaÌ/]í¹ó¦Æâ—{·ž¢^ô0«&®<îêÊ“Þhò&³)3q[ì2S™ýêÏKŽœ)ªøæüĵ= v ^á—¬˜Žmç|äµ;ZŠéÌÐÔ´˜‹Š¾e}2¾ÖC·p˜)c~ºëìæ`­þý3™ Á‡O¤•y6¤ô“É0ù•&¼L©žsAÁwÕÿªºiŒZÐŒ­°Íâ…xún¹†)¨8pöðiÕµŠ3÷×ZíòanzBž"+?mWÀ^n0y˜ùœ†6’,ðR ,ýíùD1Õ—ÅÍú(é n?p&™ðã` Z¡®¢H^ÃUâÈNñ’Ã7b ¹ ÏÄ8ßXMNË6=ázcùÕ Çé¤gƒUÿ[øXÎz…yϰõ+àUóË[ô݈?æÉ‘é½Pèº8Í!ÐgÁ|ÿ†—}÷(†Î+ y­æe}¾}°—ÅBã!L» 6½À6Âé,óþÏBØñ ö}ƒeÜ=Æ l"agNjpÑÀDâø ˆh¤Åk ôžbÁ_`é ‰§o@ö|ÌBàè¾S@Æ…1l£Æ‹vFØ)»8a(»H,<!GÒŽ1/†kG´î71é8d2’¢þ …`·Þendstream endobj 130 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1ƒ0 Üó ÿ @¥vA,tahUµý@p”' aèïKBèÐá,ïN>Ë~¸l#ÈGpø¢Ʋ´¸5 ÁH“eQ7 -ÆÂòÄYy!û›òï'Ø dv~W3Ég}:çU½‡ÐiZ¼B Š'mUu­1 ÖR Œ¦8›KWÐ`öJЦÇMÀ5☛æ&©€eú=ãO)Ø ¾?pShendstream endobj 131 0 obj << /Filter /FlateDecode /Length 6264 >> stream xœí=]o$7rï‚‘ß 8Üè²Ó×ül2¸ à»Øù€û6û`c#4^[Òh¥Õz×/ùë©b±›Ur´k¯ó÷°4Å!«Šõ]žקã NGü_þ÷âæäÏ•;Ý=œŒ§»“×'*ýõ4ÿsqsú—°ÂX˜âÕé‹«ú©:Õ: Æ«ÓÉMC4îôÅÍÉËÕ_ÎÖã c Á¯.ÏÆÁE§£_½¡Çi |ø O1D·º…Y&åýjs¶6F1¸Õ?ìÞº°ºOkÇцï^üÛ‰Óƒƒ>…9`³púb{²úï³?œ¤ó­²&ûÛËÕ#ìe•×.Âak„,N“^=£Ó¦`W_3èQOp@3, BÄÆÞX[öûû¢ÕÖÂHƇÕÅ™öˆzX}OÛ»¨V__âÏL Æù´XÅQk¿z@2o¬•¿›@|á–¿mÕ¨À= Þñþãs-®imkã†à=QàÝÙÚMÞ £Z­áLØ|ãJ—áP†cº2Œe: ÖÊM  \ó¾iÈfCÆ2œš F]äÀ4ë2«›Ã¡¹¯-CS6ûS¶w`³—eÈ`PeèËf(³bˆw©ÌàlÔ§/¾>yñ‡—y/om”C\Šÿ÷2ùª wex[Èÿª ¯øµf¸<0Ø2»/Ãm^–áCþ\ëh'ºVâÇi:aÔCˆ6óãï[üèoü¤iÉËL–´»6wŸ…;²»Rƒq&λ?É|#¿Á֙Ƃ O!4)%b÷|€…¬ÑG·$,ÔcÑô² G©%îÀMKä&n˜Ô¤â§„Ö}h‡ßÚA¹½óóOOò?!­—GÁúiœ20Š€‘ú{m@šâèO×:#˜+ÚùE²žã¨ã-"°}¡\Û Œ-øY²b“™ÈŽw„á÷§i¶ºÞªd§ðw^éÕ Z¯èÁ~r›ö®˜,¶Ûc9£X]bм%CíäV›[šG ·¹Æ?„Q)X„êëÉÚ=Á1Žnu~¶Ö°8¸‚”Å=àÄQ£q´Œ¦ß#tœ³«Ÿ™éäftK‡{ @›œ,* ™QæVG¿‹ôbëù¾ øÍ siîpqô€à꺑F'eœ½Œ]Ášù<ü”lî•M—½8'ËÝÔë‚sd.û3§³°'ƒ4˜"87‘97ûûÌ1ô)_îñç!&qQ°Lj¡Œà®gx×Ɇ0ƒÎÜ8ŽÆÒ6ÁoÄó¢uŠŸ]†ŒJo΀¹°cÄÝ+GT2œÉ¿/L~I P¼îZ⌈›Ê¯õÊY3ñ /™PÜÓŽ(ryÆØ²\=a¨Ãh G‹³I\£ÀÈØjÜÛŽÓê'\­€îSÈêB{AîÌÜZi[°4œ™Ÿ9¯€[C:kz4ÛÏã€1H²Ìè‡úœm’[`ºPÝS4pG@3K—ÜÙY¾tÛä!à=ØÁª$W‰ÿÚ¸™ÿÙŸñëi!bŠ[TcV¼øá4(2Q“<P€aKº4K`5 „Ç{’¢VxÄÎñ<>c1í ¦³™}J«ÕžDn™Ó<æbÂÀ¬Ýæ.“+Ö—“xuL óKº,ÊwK‹á¿„ ÒvBÉJçÅ r›¤dý¨ÁíÞ.⽿Ɉ*ÛSÈL‹BfkÁ’Ü8ð¬‚r’5×ù"›­6¤j3wïÉP)ŒœÁÎL%9Ù.Œ’÷2¼¨•`DŽ¢°g¸´Ü«õ ¸4 ç%$Ú7‡oJœñŸ+6>;k…؉Kãâ/Ö>he¸š˜x 5優fJw¡¸ú”Ü‹!ØO¡ZÎ} mÆ'å„A©};ºQ\)c]šµ!Il¹Çk:0ÀÎ唼]Í'’B¢^(ƒ6¤+†m‹XDàŽöóóeqaò{Ç<Ê[ÎeŒ¯?k”-‘ÃE4eI—¯¿§É0†®W3cš ç€wܽÉb=M3#õ,J0ª›Z,¯Â€V—»xp5JO9äOÛ&‹àù%rûŒÌP¤9YkÌ0Y_Y–#1‰ô¹•ßïHó8àX!/×¸Ø IÑ/|º{LÎ4˜ÇiU»Wàƒ^–eŽpšô ­FÅUY–îƒìHŠóÞ”!›Ý”áusÁÏeȲAÿU†e¸onöø ìg·íû„ÛÑÖV(씽60÷U˜±äa;¾é%ð,’1K÷ª„_Šþ™&…wÄy„&i•ZÁ m1UISæö°Yùe[4:ïmm0ðp­0Q4mGUYÊ„‹òB°›¨ì‹dówL§0yü›$Jk5›}5wçìIx \&ƒùN5–6eK(†³øÆÚ]È8¿î.9 ‹p¡¥x³F¤êßÓvhË;N÷Õâ€l‹[rÍ™´E€->ăD@¡uyEä%ˉWrnà8³„XDÆ SíÚLÈ+,W²ÙÈÇQ"b€‹6¤E¶hGÄ0,¶+i`v‹*2",º ûã±Q²"ÑQ-ÎÚQãJÎÚ¡q%Ááa-‹h(¸È§ôCó Rg‚%á ÜÂ#x—`0¾ëÀ?°cìyGohE€Ùó%ByO“SìÇñx˜‰K@auTdgÈa:pj„S@Ü;3x?çpŸ·R{˜Š6™/jÈ@ƒÀ}›@…@Qq£ÉáØHr¤_©c„ÿ´,±=µ¼I”ÄLqÜ2Ù,lî“DmõÓîËr»ŠûEì^ä*hQÀÃ6)kü'¦¬÷tÑ€­pô²²†ˆB0y'¯WüÆ=K‚m+ÝŒ2ݶ÷¥ZÉfùìÉYÏ*™e^ÊCíõc°¼"ªxa.®Û/еî[ŠcxJ ³TÑ%­¢¾i:"èšÍ’—•E$‘‡ytA÷L´úñ 8ÔÞ«ÍŽÅT-¤ÉÀ5*ý¤Ú[Ïy®&¤àϘ*BÈ¿|qò·*‹B£NïûµðÊ›káQ^Ч©öi ÕÂÿ)Ijt*€( òÀ$zõ–[®Å ±ÄÚ>ÓÏc*H²æ\Jþ*kë­àŒ“†XÀö4Ì{Ú£àŸéŒáÿ±8ý¦´ÀùHxaŒKë×ð»àm¶-ß$L&eYžÌI„ ktYP?á0`’‹I²7ôÇD .P©£`Ž˜guƒ“>Ïȸgnñ00ÿ))‰›9äÏl÷K"YÚ-ˆÅÙžb@thNS Gæ~qPÜBU_æéäeÕ)Jô² ÷ñLÔmÆU‰"@¹*”S@UÚp87Í8…*,àä ™È(¤ô!yƒÅ?zC‹ã¨3íQ={‘¢xƒ°HsöG&øi·gM/VFý(Z“©Š2ÛÖÉŒ(̲´[JˆÙ3]»QÕT …¥rXxÇ%ƒŒã¢HÙÊ `ÎôçÄcÞiC3énröUñánsC郅ôÜSÉ4‘rS'Ãçî‚)L²çŽW… Œ­¸Ô€nñ˜œU¸O™úÑ o—ƒÂ™8»¾ÅITu•{= Ý3 öЩԵEÆ;™ÀÅßн –v’~Ð’ å@2%}^8E*fºø¤²I=X-Xœ¶–û?GÃc¬®äB‹Jê²›ÕšAú­BŒŒ‹ 1:  žK•@ônWçL«v–üÔ%à;`qh g^¼Ì'²K€8îôƒÒ .Ú*Ç—Îy}¶(\fˆ0U€b\ +/8‹Õ1@;ÿG„NaföxÛ¡ø–~èeµ§LP½Jùj\Rû2ˆ•]ýk²$ŒWùˆ£ !\aá‡LZ ¡™H=s˜òÕÅÑÔ™N ¡w"+{j4Á¶Û¥É®›=Áaå'fA¶Çn5&}±\Ï´Ú]›{àQœ¬yNOdˆA—îJsGž›íúgsÐÎï.‹.pÈÛÊON©¯*“6EX r>.ý9ÅJãèc3çøù°ïKF‹I w¨´’ú”ØŠañÇègDY]Ö ‚X´­@4C‡ýãÜ=:åêFª¥²>QVVe Ĭ{ô²9dM§¼«Êu&¶ø¬tÑ¢N£.ZåÊû&”_–á»&ìç¥wöOev]f? yÒÄó‹2<€2U¾,ÃweÈz¯Ê°]ÙzVú‘Ÿ—áŸËõ.»æÏšGÜ47û¼,¸+CVR»oÎÞ”áe‹û& eÈμQ1k›¾*Ç/ãóæÝ?yŸëæ‚ÇæÚ»§0þc Á¤dÃQ^Èþ»&ynËðwem»¨ùÐ$û“ŒÆ6»ln– <_ · b…O†é¶½»óóæû&Rëæ‚Çìy¡?(Œ›`çTb,Ã7e8”á}nËpS†Ÿ—aæÖõ|2·D¿>k6,c²™ué_Ë›2¼®‡ùe@CÿÒNœc:ñÃ0e¯^5wxVÀó2üs2ûóJ'äÄÅ•çî`å„Ge‘ÁÛ9È£p ¼¯©:Xôï'ýU‡&“ôÚµRò85“Vðv“;=Có£¥2öúšú½Rï¤Ãfϵs ÍÕg‹‡Üé·xC? ¹Ù¤?+²6y¥V‚œ’H£,h–p´à]vß_•á×eø¿La~Þ¤µSƒµ>²§õcª)–ÃsG²1T ¯CËìÂ_ñ°ú±•ån);‰ò}Z| «‚¼÷–ú÷Y›dÂ|ë®Þ¤êCòcë’Õ¢~!¾°Ý~Þ÷N<),aœÒ‡7ºfS‘û°_ñ|Ns°àäž6ôA¶RyÔ»¥73W—4.K¶oªB4Û$ímejhO³ã¨Õ¡aÖ»Åý‘JcÖ¿×a¤N”í-j|ÐâDF©Ó>ÿ¶Ô,E{6¡êzð´û2yÀ(*F©Tëî¾fa°Së»´_Ã]!r/5»i»Åk‡mæLô0•Ä_…PG€™¼ÒȸMBÕ!š¯¢'a õjù™Ñ•U•=“¬o 4òUz g±s _‡²Ú%V¶ ål§I„ö×ùeO|ÖÎ X†ºÓ¯½aIÑKÚ[”Î@IÇʎʆØ*ËÉšH‚Mýö_%\>ïàµ%Öd _{qóóIJäØbcOì\*C„ÿâ‡_Z~6øäÔE¾_.?g’à³×ˆ]Z<Úb–Xžœgåî×…PŒÊ»•Ÿp”ª»û™™ÌjÇê±<ß}·XÁ‡|³°8ù‰!CÕ€Í2Y¤k'¥–àMóUJ·ôQ$üzj ®¼eûØs$^pa_#Ò^f–/˜%K.SÀfT[·È‡)Ù®|Xá‘JlNË*åÁ—‚ÓÅ:»Z)Rù¨£V¶Pü§3Ôh@Ó+„ʧüÁ§bwU|'”Rís=¯–5ô˹°Mg‹Â~zÞGéf®úSÓ>ürªê, ÍžjpnÀéèã~ÃÔ6«T2¸YY“Ÿ~¾8KÜô‚—xuÛåóqÕ#»¹G~²‹ÛpžÎ;p>i¬]¹Æïxï¡3CƪòçÏ€J Ô}»˜û.BUOÐùG¿¦<{”¶¸F/¤ž°4s8’ Ñ’3•[„·â‡‹A#nÑÈ·ÜÎ]Òá¸÷'v¦d[W†š.ßã›oæÈk;SÈÊ789EÕyÖnXæâÎö•œ÷•Bu ̯½öóÃE,ÐùRÖ_ˆy‘IúÒÔl{|¿¼¾Â!ø 8$sY2hüDZDÖÈévD@‰ÕÔÓ =÷½n®œñ—fdÇ|§=×Gw¹®v»¨ðñeS3ìgQÍ\Û›VÐxpÆz Cì(*'K—§í ç%­0Õ£Mî‹Q<’Þ ïë¾>j—dL¹ç/I‹Úýð"÷òDÏ`'çЉ4ž–á:‘ÅÒ³ïÓlZÍ Ütë˜ü §`:'êÆ[5 ß+c1›“XÞÈÄÀsP‚Î}dA‚x„"ÞÙ7ÍÊA‡c>Vv8ÎmŠÚ™N©­ªÓÚrI+¬(¿Pa¬bd]€ìÜéxJ´k<Ò:—d:u­Æ¿mB çÕß&9N¯µXX–Ñp*ôÆÔÏÍP>”0¿ÀYƒ×]Ë]n¿ð—_aÂ\œêöÅ„©j7šWß_˜’Üý˜ATü¨«¼¨8V47&¢ýªW”‚#óŽqЮù^P¸4%‰B-¨êYX~Ì¥!Ê{‘v#&ƒJ‹7}¼uÑÕ*µáf·ùv¼¬NÀÁ³©™ÒÉqôz.—óÎi&Rí¼ËyÙBv-·!uêl 0”TøÔUÏ!›Æ5£óõ:6Ë‘'ûHnØe^"ïåžf}p}>4n¢&ÛÝN|òaÙe·É=nªHQJêlóîêÀ/Ÿ—|ú¢i!h¿è{ú€Qªÿ¹ Á®D·Ö÷.æÇV•ʦ ýÝÔ\ás íÇå UçÅ<·&¹Ï3ÛìYœx¨X=|$?Tr.u¬w$¡“Áfìr<!Džn]I}Þ sZÿxTñ@‹Q'º¡àŽSýüK{Ÿ\¥ÛêžÚ H«ïÌ>{·ã·³£?h#LõòÚì]Ë}ö?}ƒ¨`6CÜÛàö"`èt -bÚþ ⎶[¸Y[,¦’ˆæïÚ4¾aAv!ùý²Ë÷U˜ÈJbaa•^êrL°z_4Äß»ƒ k£±Œ_žÌÆ k”xŽ^wÆo5²û]Pkûi$(5`°^ÇMîÛŒ^”G˜ý;ö™§Ä·;| š>¬0±2*=Ð5é ýˆW`›~§WR&¨Ë½JM‰›XU×¼F|MÇ©óέûŒw‹ã‘¾H™g#!­ßÑÚ¤CŽ&×Ò- ®Â“]<ÜŒb a*ˆÃRaº`sÅ3n(ÑWî]‰|.˃)|õ 1?<Á©ÁµìÇŽ¶)üJc Åö<¡ÛyoÂ,Öñ~Å|ºtŠxÎ5ßoþPS‘ªß#½*1EõÊÁÑ«ƒå¨~Ì`1â F¤˜–÷QÜüYáÇÐ.ŠÞj;ö·K¾f‡Œùv©$mBý07èŶ¿KèGcšqÈ ØÍOÀý°D"ýÚ~Ã/dÓý@7­ gJ­ÇáÌ#¸¦µØ~ÿ=Kc-ÉDþ¡.^ý€ÀZz¬x ÆØé…†Åcï}“t––x£ïÜ·žûÈïk!†Ê|2g‰p™’:5{ÂÀð†Þøz;LõƒÖ&ß-5©i½ÂeŸjÜÔ_H˨£´ ¦¼Ö´o9ƒY¬óŠ8¢´•$’kÎ9iƒ/ŽXšmépæìSFì™…çŸÑ®"é24Ñp¦¿§ˆˆ= U5ä×ò;f‹Î`¾Ð¬ìxÆ‚D˜žÚÊT-}=E{áf•Ï>½”`H'(“«†ÒÚÄë@œJ×óuÊ€ˆçg_-–¶0oNSÛ3£ÜXT9©ƒ&ï ©üî(ë6¤Ö²ºÇûu²îUÖbÇúõØÏ:=¶Ûæ¬CöMs·‡æ¶ï“­áw¹Mv6{Ó„¡ Ù=ò‰†ñ»æiCÝŸÅfYO`§ ÷‹æâi9?oŸÑìºþ¦ÐºÝÏܦ_ûÓM»§¨zÕ¤ûÜ5»¸}ó`Ϋ›&žŒo›WpѤê¾Iv~1츻æµI¿ooqÓÜíÉ–ÑwíÝ>i³jG)×_*`/Ëߟ•áç5…S÷ô¾ /Ëðª YWö»2dý×ß•á³ìcgp¤+ð÷å£öŒ%ë0†ú©ÙTÌ^²äNøºY7gÛœz C¼uÖ7•µ{³ó˜êaKý¦ÚíC5ì2düòM“ìù@›Tí7;Þͽ̶_Î0Fýn1?n´ óó1”ÖÿOéúÿ–¡Ié¶¡ßJ·= öfàm“dìÃc“NÍÍÚWñæ©Ë¼©wøUêû¶ù;f‡~j¾V`?»hþìº}Üÿеh*´g<ÃyþlºöK ½©­AêKüÛÉÿ½Ñü> stream xœ]O1ƒ0 Üó ÿ €*uA,tah…Ú~ 8Ê€…0ô÷%!tèp–Îw'Ÿe?ܶä¾(‚±¬­n H0ÑlYÔ h‹±°> stream xœÝ]olµñ}…ø +žÎ¶wOýýQ•J”B*ÐT<çnr— $’\Hûë;cû½>{6TÑ ]eâŒ=ãù¶ç˜׬çk†ÿ¥Ÿ»ÛÕ¾âz½\±õ~õ㊇¿®ÓÝíú/ˆá`¤÷ÌóõÅ›UœÊ׎¯­¶½—z}q»êÌæâÀ•ŠâZø»7ð/.WßtŸnX/Þv×r&ëî6[Ökï­a˜;Ï™íždÞk绫ÍVJÙ+λ/`T2Θ¦K<‡…¥Ñ*Lƒ?[溷dáÄpJˆ¼XXKxïœé›­°=ã‚w—0ª¸:ÐjÞImºœ&zÇD÷WäÌyͧL û°ÃãÝáqƒ´­–K׈a=LîvÄ5Êéî1²$ëþù†•PònÛYnL·Œx˜÷"0ï”Vqš³Bâj×Ãmø… .TwŸwv“Ù^¤BÄDõ#½I~ŸW.$™åO˜úîâs0-©¡lGKÙQeÁ¶‚Áüy³ÕBö »À<`(åƒ.aUÆ|ï2ø”Áo» ßgð&ƒ‡Åyeð¡‰üIŸ3Øgðu3øjÜKLhçF¼ÌàUá:ƒwÜçu?Ìàa—ÁC“Ä› ö|Î YŒì¢½îM“gp[£dÇüvSÀhVŸ\¬¾„0æX¬k¥Xü„Ò+·†;üýájýõún¥{¥5ÓëŸ!¦}ÿ~€(ø·ÆõƬ•³0i} `Њ7«L8ÆjN@ÆÄ:S;Û á’„TÎLïŽî¥ð!ª~q}‡qu+zpaqØY ëÐ;º/†gü[ØóËIqN›2(‡Ä¾év‡«7 n͸îúo»Ïîž®vW÷O Þ-çL¸ å†%«îñz;à/Bðý@ª¿—Ýõá!kÅ«œw½UY¼ bÄ*Þ'‹wD¢âEMOʲÊõL­¥AÞ@Ó\âbóІÈ#PÙÒò^Re§JHh–i H˜|ÐÆ&¤q€")ë ²™7Ÿ ³é…P£-˜u¿þ*ói4¤\é(¨A+ÓL¹¶uÂ<%éD ™•'jî}Ïì"IpX‰CtÛÜC2]"iàw(ÈDÐ+L4Ñ…&Ƅ樓 æznÎlÉúP˜Ft¹>‡Õ\©4Ø_ä~²½Jý‹òâNõÀÚfrR^B´)(ÂRê«ws²” -ªHyn+«‚E%Bú–ªÀât‘1ˆÀ½3cÝô¢,4³¦Wª4׉±“$­2í ’X¨bK â_¥T/ùâÊöV”|Á—ù;ä¥=Yˆ3Da4ˆB»Ò,tïξ×Káƒ,š~ˆàFÅ&Úô2Ia¸Ô·•£Kž$ q[ŠÒÄêqY®žD•Ç\±ènz])W0±³ä 1¸´0Ùty“*!r…ô%Ï h‚ÓŠà ½^¤x”±´P€9O±Î*9iÑÜÎÚNUÈD fÎÎ ™²J;sž 9fBRÁ_Œ(’÷˜U—5çƒXU¨,Œ¥X§•ËŒqç$ûUÁ l[ÏG”:­ àtYGcZ¡:‰çH"dBQ1’æ­"e*| ê¨U³ ÁBseâ Æ\r¼#gü3̵fL:˜x†ôSV¹­+¿Â(bR¡«Càç°’ ™ÃÜF!*IˆÉÚžaº§ª…ÂÌbÖ?Cö!©ÐM‚ù"Å1©‰ÖâÕÁ9v(+ôÍl\'êlÊcѵ,טT(Ii²¡œŠN¶”+·*³z2‡¤Bœ\Åò&cR¡{”üÔá¢N*4„Aa)÷8fê¸6s$ë¬BNBB"Yg•<<žgéÌ“$‡Êq&ƒrV®uV!½cãI¥Î*‹ Ñóùº¢Î*D†Ë^òrý”UèÙcʬ»Õi%O„CµÂ¢kQb)­-iÏòí„’BZ!¬2E$}‚bJ+Tà§gl2¥ºIgÆMÖi…lHH¨!æ]¤N+„1 ?gÏ uV¡æŠUŲ你¶c<¯ÓJÅ–š­ê´RJbY )§P®0.ÓKI…ÐSP—Íg™:§Š .½,ù”S¨?:ågĦ”SæniNxZÌ)·õ}Ï2ŘREL®gX}L)d᜺¬È1¥P…óÔ|ÕZçE™8U8Õ9¥Ð¤ÀÓÊÉ”S MÊžËé6ð×¹÷»J¯_Ð#ã‚÷P欕Pxš —¬í&ؾnVIæœt´õ%þ¸Ù*H—ާ”gÎyä%£å|jÂßM½°¢745„O© ÄUwxaXœ®ö=áû*bH¯º×›­0pr‚v„p=Ô°7ÚvcÌÐõuÆìæ>®g’@LÙ€ ó¹§,:}7¡…ül’Û›Üå:në—xÿÆΈ%H¸œ3†iôÈêÅþïBÍKôÝŽÊ¿5}·5ûËôŽ®Çäþ[V|,Eñ%Ÿ·púvP;â5(s)\½¼êx‡³ Ç7ÿҢ⦉œ§Ý†àì\Üí.+êÁ« ¾iŠ™VlŸe°ý½ÚUSv»&Â}süVg¬ÈX“!žA—AbÜ2ƒ:ƒ®mË£}WeÐd5GÉ4Ay8Ÿšn’Mjº¹‚l•åïhÏÀ¹5«ùE­øz´f· º_,§¶|“š¤ïdD¬ ª&rÊžuä à>ƒäd:l¶F8;/2j2(›¶M¡[Š» ‘¶ÑûæÞÛ>h)x>5Û$Ì›$Ú;fÇÙìèãÜZ'¤,xx"Úá©©ÚÆ"^ šv€óMEˆ&¢ÛýÊÔb¦.ïC‚¶nTÍ^ ¥ûFüL[:ˆeñ^Qù^q;^Tyeuºh _‹“[-ò)þÑìZ/µ`9ë”(Z‡ñqï´Ó¾¸¡i3ytÃÆ™°ôx‰9³ý°¼ hÕ.\aßNG´²÷Mð×8·4c´fºº#~Œ[ÇKFÖoÆ‹@æ\÷ ÖëÅ„“k¢ ©¶øì©«Ì$÷(à0šŒKùâU§å^4G»1ñB)2•,"§W—¾IOâ½í=Yy ·ÞC¼^ÕL I—¡wª‡À!‡½‡ƒÂøF!Ôo`~À¯×÷éNÚòîpWÈ ™9Ïa«€¡¸ÔÔæy%ÈÍŒÚ Ö§÷~³µ `¬ÍûΕ^~–˜×L˜À–íµÓÈÕ¤cz± \áÇå ¹‚åxïœ+DžÍx¡Œêþ]IZ{-¼É{ÐÝ` Võ3®ì˜’Õ[—éºú&â:ˆqŠ˜·°ˆÁ ö¢6E SÙq­~°äËHÒðÙ·8?…íz(Ñ‹æ52~¨ì7pÈ8϶rXœ ú6æ€æàFA…ˇ_¥Êу‡È)(½pðŒÆc«pA/õÁÁŠÝápSÞßoGjå)$ΰҦ§;ŽA¥xÚKÐÚ¨®ŠnJäE9¥Âs”´å;Ìã`ý¦È é(ÞàgúÌP_©ã `À„nØï£~aSß#xÂ~ˆìâgÓ…’/Yç !E%[±DñPI¨`œ˜FmPŽö„QÖ}1­¹¹ra· D¢'öRg,Hˆ—ŽQDByos؃”£à4ÙUN`Ý9ܵ;#A«Œ $ÒHÁˆ£«Á(±Wà=v4â`"RŒ#ax|ô%¼Ve›ê±€‹w{ w‹fœè•Á8DC¶+ÕÔ!Ô:m7dþ÷‰t¢Ô ÈU®@f·Avm /Â+áÃ~”áÁ.o1þAê³bý ±vfòt™{êšèÌHÛº£,Ã+‡Ü ¦A)1‡ø2;M(r¨j§g•˜âQ€XŒâÉ~Rì=í2mÃTœ…‘ýÈOÒvÈÑK4XzIø;8+zI„mQ.”Ù9cz6Ñß VÌöRºÑ‚ÏI›Ù_fû¥ÛqÙm:ÛàÚ_‡TÃP”ÕøNÕA4£C7×2“íTTDÓƒØ#ƒË”Ҵꨩ±'/ψšF2Yüá2Î6\ÓH3<&oÓ^M\ºªÆ£áf,wl\‡‡µûïƒ8ñÝWÆ%mþ°¬J•NŽqäMl<0 DdFýîrJœu"H[jG×!ñ㙜sÒ]5jã@eÑa½˜µÓ‚û¸H›X®„µ¡¬GkCâe¡!%@ARÄÒëlñ›nFSm)¬=VGÁ6æ=¦ Ôx´R¾žªwš¦Ð#pb¢,$P5"¤0CK†Êm±x°Ýˆµ$p“¢#ª\rá +_›ϰè$‚f"l¼‰[ð jk’ª5W#?è`E5MóGÙûHÜs-ÜÜdzEvd)Ĭd³ñ@'ËÏ|¬Àx?*÷9–÷̨Öñò¿W~æø'ByPäó·Gñ¦þ¨exùj'¦Åð²ü×Nn¾2AÒä|Bc®ôáåõKÎ'r&£Åi^dŒõ¼G$°Ë¦H´hŸè DŒUõa~®ìù mª•‚=$"BÁ¨fR×ìØ“Áå8~cÈEUô ùUýÛ\+Ìü~̱·¾=I$‘0:¼ŽrõFPW,³ùh GQ/°1=ùW¤2Ò‡S÷)îøa˜ä"ÝO†EÌ»<º— __…xœç9“r`sz6üåê?­A¾endstream endobj 134 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 405 >> stream xœcd`ab`dddsöuа±ôH3þaú!ËÜÝýãÿVY~'Y‘{ B|Ý<ÌÝ<,›~° }OüžÄÿ=^€™‘1¯¸É9¿ ²(3=£DA#YSÁÐÒÒ\GÁÈÀÀRÁ17µ(391OÁ7±$#57±ÈÉQÎOÎL-©TаÉ())°Ò×///×KÌ-ÖË/J·ÓÔQ(Ï,ÉPJ-N-*KMQpËÏ+QðKÌMU»RL:çç”–¤)øæ§¤å1000V00D2012²ÌùÑÁ÷Sª{ÿ÷ûp|ïý.ü]|öœ•Ýû»%÷³çvÄu—w7uvgôíåQgŸ:§£µ£µµ«KÞÌXãëÊýluÝQÝ•ÝP%Ù¿›ÿvý-ÿ[µ¡¡±Å·[²*œmUßÎî¹Ý»Ov¯ëªÙÏ^ÑÝ6¡­ÿ;ÃwV‰ï,¿ùXsÃÙŽöÍèÞÓ= ®Æ•¯væ€ißãçNÉö;n2û~®ï|Ür\,æóy8·NåáùηŒ‡·‡‡“­˜èendstream endobj 135 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1ƒ0 Üó ÿ VÄB†Ví‚ã  8QC_ C‡³t¾;ù,ûá6°K ŸÑã‹XÇ&Ò귈͎…ªÁ8L'+„ìï:¼?`7=øC/$GÕ¨²RG½¡5h¤¨y&ÑVU×ZÛ bó'ÉžÎfwfÔUSÿ¥äh.qÝÜb$N¥ii’ 8¦ß3Á‡œ‚â ( S endstream endobj 136 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 306 >> stream xœcd`ab`dddsö Ž4±ôH3þaú!ËÜÝýýúì² üN²"÷2øºy˜»yX~_#ô=[ð{ÿ÷TFƼâ¦vçü‚Ê¢ÌôŒdMCKKs#KÇÜÔ¢ÌäÄ<ßÄ’ŒÔÜÄ 'G!8?93µ¤RAÃ&£¤¤ÀJ_¿¼¼\/1·X/¿(ÝNSG¡<³$C!(µ8µ¨,5EÁ-?¯DÁ/17UìL=0霟[PZ’Z¤à›Ÿ’Z”WP”™›ÊÀÀÀhÀÀØÅÀÄÈÈbó£ƒïã¦?Šç3žø>‹ùû¹Y¢“fõNëžÈ±*iVr}Fm»Üo…?­ÍÝí’U››{¦ôO™Ü/ÇW¼ø§ý¶ßŠÓÙ÷síçÞ¿€‡ˆyÑ·káendstream endobj 137 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 417 >> stream xœcd`ab`dddwöõõt²1õH3þaú!ËÜÝô³ýG$«,¿“ƒ¬È½ ¾næn–Å?D…¾ç ~Ïæÿž!ÀÀÂȘ_ÚÒéœ_PY”™žQ¢ ‘¬©`hii®£`d``©à˜›Z”™œ˜§à›X’‘š›Xää(ç'g¦–T*hØd””Xéë———ë%æëå¥Ûiê(”g–d(¥§•¥¦(¸åç•(ø%æ¦*@Ü©¡œós JKR‹|óSR‹òs 2¹»˜Yýèàûy ië¾­Œ{¿ó¿ÿ›ù'ówsÑgï—}gížÌ1·ª»¡¤«£¼U®Ô*)°³›£³kF¿|Ãwõß{T»9~§ýæú­ý[ò7Ï«ÀïÂßÙí]±C~ßÞÊÝ·Vìž7·¢»©¡±»®U¾¥¢¢¡¼»°»ý;{×,ŽßÜq¢Þá¦Yvݾ‰KÏo\ômý!ù ‡Ž/ÜÝ}±{MÐtÝþüž¨I^|û¹¾krËq±TæópßËÃów9oBB›mendstream endobj 138 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 468 >> stream xœcd`ab`dddsöõõ4±ôH3þaú!ËÜÝøãÝOoVY~'Y‘{ b|Ý<ÌÝ<,+8}OüžÄÿ=^€™‘1¯¸É9¿ ²(3=£DA#YSÁÐÒÒ\GÁÈÀÀRÁ17µ(391OÁ7±$#57±ÈÉQÎOÎL-©TаÉ())°Ò×///×KÌ-ÖË/J·ÓÔQ(Ï,ÉPJ-N-*KMQpËÏ+QðKÌMU»RL:çç”–¤)øæ§¤å1000f30ø0012²\ûÑÁ÷S£qõ÷Õ/g®þ¿ñûï»§Þ2ðSG´§¿wb÷„îÙå]õ]¿Yš$ªºêºº‹9jftÏ^1©£}†\w}wg{õo…?‡%Ú«»Ûº;9Šæ7Íš:·Þ ¹I36·š´ú\Ûåêî \[GCGMwkw3‡çáÔ#oÖ}W›ÜÛ5½²¶³µªM®1=ß>¢›#µ{Û´ÙÝ=æËƒˆîéÝ7LÛMZŸ‡s]?×½qendstream endobj 139 0 obj << /Filter /FlateDecode /Length 7864 >> stream xœÝ]K“f·mÝwåG|¥òâëD}Å÷Ãe¥Ê‰ì8©xak\®”ìE{f4Õ¼<ÙÒ"¿=àí;nµ¦“Ø.-ÄÆ¸$šýÇ“ÙìÉÐíÿŸ_|òkOOÞ\˜Ó“‹?^Xþ×SûßÃç§y@= ([5Õž|y!¬öTì)ǼUOž_œË僯Ñ×Ý7ãßkªèÿàÑÅçŸ_šÍãj>?¥¦5>•ó‹Ë+³ÅZsvL¶¥Z“ÏoÑ4µÆRÏ/¯¼÷[°öüKP½±ÆD-âÏ,ا˜ ÿœM9¿S‚_Sœ›ÂX–«µ”t~yyåòf¬³çG ›\äïâkµø˜ÎψÍmŸóg4²R£-Vâú Ï$gìùå›KúvŽ~ÒSê‘+˜Ï/©o %žßÈ|)çß¼arpaéü‚¦]²MéüDRÁ÷kê`j C%¶’'icÔ×Ïù㬠çWsfÏæ°oýŠR“Vˆ¬¯šüjJ^49õ¯õûÿC‰^ÊU·”+|4dØÌ?_^Eç7èŽ!Ôó57³1•—µ5wžíç³¹tnÒÒùÓÙ|x,"ÎæÇ³s8¤ºj9¦ÆþÝå$ÿÓam6íÂGÚ²~‹¡ºÓƒÿ¼xðïÓʳ9ˆ—³ùöxvŸÏæãÙ|:›ßÍæ»Cu+¶7³ùÉlþâp<Ç”„ÿž³ÿl6¯góÛÙTÊ~tØ÷í!õÓÙ|3›góél~7›ïfóõ!›ΫI}8›_ÍŸÎæõáp¶Ãá¼8œ›ÃM%Ûñ›[°y³‡b]nææãÙ|:›ßÍæ»Ù|}Èö£Ù|p[ßëÙ|;›Ïû¾8ì ÑmšóUŸô{N)7v“ëé7˜¶§\H¹Â‹C£×ÞôÍáÆ¢v…<›Ûm[Œ„rÞoMÇŸ²¸—³ùìPÂG‡¦slßÊó[Óùncz~¸]« ½:Tûí;×›ÃÕø;×fó³Ã#ãÛÃýËCß|{ØTúS{T›Ñ÷ _vøó²£ýÉÿÒ>÷£ÙüÅ¡©KPcøX¶&S¶TRßš”Ò”¥½ÚýÓÙüh6ŸR?n{Bûï Q>‰³ØÆlWô¢?ÒÍ#Ï~0;üz63›?;ôl…4”?;ü„Õʆ÷ìÐÄŽ÷†[%üI[æ-»KsòŸ=¸ø… @â› §sÜü (2m%žlxs§×O¿=½@DñoÀ¤Î@½?!lŽ9}ýÞ £úS$ÍUÇ6B ý‹É›• Ãט¢>š»È [ŽòꌿyqG:Åh·šïc|uK òRÚlºyÉñ9“ÝŠ/÷ /l¤ä¸{‡HÁ»S,yÃIqò°¬W ¢—‡­¡†€ñE%÷0¾ayCÜR¾ñ‘ßeÈ«[–8üåUÄŠ0—àÞƒ¼b·LË÷0Ý6çîÑ\°m¥âïÏ{KÝ,–!x»yÕ"|¦Ý Bž½yp‹P ¯l>ÞÃòÖ´a¶ù¾f[¶šì)ÚŒ½êƒÅ9c·`±¸¶lå|Í9d,‡ã‡OÇÆæ]bçˆõ>ä•-‘WÃ=ÌG¼Ã‘½Ùrúpcq€] óŒÁÂéîANÈkBˆ˜ö<05ìáœÅ$°qøLÄÓsPHM ê€ûÙ…³Ål!iÊŽéÙÅç ù¸9ؼ£·² 8'¶bOrÌfÂFóT”•Iá¤/´UàHÀIE‚* ËÓæá6¸ñ¹ ÓX(+—H®HëD Ζ$y¶D–S?xá ðÑ…²r±$BÈ4p[âHŽÇ~• ‘"\¹°J&eá)Þ“B Î`Ö´Ož >T¾¬5–Ô¡) “È4‚6Bä£Ð5TJ ¬§”…²0±œÛO‹˜~ +e'4ee9®20èÐV>ÁöC<Á,as„ÕXMÙq‰$¦­0FÚeHì¢f²=Oûó9¶*EY¹DRΤ’È·o<¹ ë%|ŠêS ™­JQV.–” X»< ‚.h&PMn\–U2)+ˆA3™ÁàÅ3Zd‰˜‡Ø]†±ESvL"–_i0jÇf”mM–"|”FY(+—H‚å³F°oÞ÷ ,?ÓÑ¿ç2ÂoÊÊÅ’Š¡ü 3èÍRpTü´ÁÉéà2ü’â¡IÙq‰$8@¦MÖ*CŠ<ìI—)š”•Iäà|ctÛ.°ÿ°xDÁ9aâBY˜DÖÿJ¡Å‚œJw Ųfñ¤ MY8DH% ¹P&OªÂöÙPõí ,ÀÁÐ…¦,L,§R¨Ë‹Xèc$(ò>…Žô/ÌgЄ“‚å³2ºMW~Ú„+m¬n¡,L"'“¶=þÀ…Mœkf‚}·P‹È€ÕÔ0Æ| Z*Îx¿ ›I¢Ì"+£ wIˆ¦àeð<6pVɤ¬L"'0´8Í{8AÒàlFT2)+—H¢µ k† Ö’¯´™Ã¿ûM /”•I¡Õžílñw¿®.”…‡ÅX¾Ñg¬Xxu¢‚Í3r”—Hò…cúœcC ð€B{Y`äK|Ôß-”•K$! r‹ šAZÙ<uMY˜DìŸÌŸl”§F×%qXر—‰Ä,†5) ËÁ6 A0Š65›“ÐePv\" ö_½ÞÝ"íaË®8¾`ØAY¹Dݧ.+—ñåPõ:%i¡¬\,)Ã8ˆ°$û*±G!>å"Z”—H¢:(Úq ;Q#ø¡­;ضQgÌÁIY¹D§iK¡kuêg—9T”…IäÀ X']Nä¸r9V‰¢,L,§˜,r°º¼pÑ}þ2•IäÀÒ²•P|Gç´À‚XÈ/›ËŽK$ÑÞhwÏô’„ÏÁñnMlŽ5MY™D\€5‚ÿ[6¤Jµ]ô9Òp>#™”•‹%UC¬“á&.§(À…FÌ=—H¢³¯,’’áØCIªUàÛ”´r‰¤ÐNIpW¾ È}RV.‘”«:%ƒ1•bOu(Và# &E±Œ@*£Ù>š``ü¹,ßFüÀúè”=—HB@dôªXôU`ÿ’eÏ$‚`þiD7vås¢ŽIY¹D•^´@ûsðËTÂ:³•…ÅX˜^ÅÀükZù¬hdPv\" a‘_Áüã*ˆn–®L"‡®øÖ‰ÇA‚Š•¤7`3¢IY¹D¬Ÿ2ÌÈÁúSUF$ÃΚ²r±$GWA/›óTª‰Â{ÖÉ ì¸D¬Ÿ•2ÔDÎeÕn Û¢¥“ U”2QJ2,|¶ˆmMÊÊ%’²¥ ItwµD¡àó;ÛÚq‰$ xRÊ8ÛðõÆFØO²àè&!,”•‹%!0¥´ н3f}¿›c4;)+“ÈATÄ:™sƒD½%†k\¸çIœàV#‚ŸR¢¿/ùMY˜D| ,@·Îy1r±$,/¡p- ÆÅ•§Úb;.‘' —åTLˆ‹ürâî¸Dâ"ΟtÛÎÎJÎ`XrÚ!…(ÊÂÄr2 /~›C’ th6•¸-Êß1‰ `D%cj@9”9‘L— Ëé²ãIIîžçÔj”$ʘHö<MY˜DY¹¸Ý=§Åµ2Ý&,ÀŽ‹%Ñ¿šåì.!pežÔ¹xÁ²ƒ²ãI°ÿ¸¸-]Açåì.¦4㔕K$!(²«$˜°w+Ÿ,;)+—H‚ °V†$º…®«¤˜E+“²r±¤Š¸ˆ‘Xæ;~’'à”IøœÄ3ƒ²ãIp‚¼Þ”BY°;¬K\°çIüb”tŽyå³¢•IY¹H•'TG—ÄÐ ]>Ò™L?éä|å¹Õ=βƒ^òH•–‹*]/™|>?"úÆ>“@ß-ƒd$w7:M‚êÔÒ©³Ó ¨Ní>v„Ù©Ý>ãgÕ…c”ÑA~šÿÜÂ9óAPVmhõÌNéwÔ°,8‚ˆ‡¥¦ˆNE+¡Ë!DÖž_|q6—W®ÛK8;´9ž·õF;ž“¢jGÎÒž­¡…ÀñlÝø¡œ-ØG…ì¨mçÁóAItCѲœ½<°o,ý™½¢%ü¦:5ÂÒG%oš£–>9óºêÔ(ºWçì¥GþýàŠV‚â!Žj Y7ÁÑ}Wñgךƞä&n²çs9Ðl·“h(D­Ó:!QåÙ͘Ҟ <ïî–å)J16­Tð2©Ï®/©þ ”rþ–ÇŽ"Õƒr^ÉQ=1ÅVÄ娤•º̗ r¯b£§5F9‡hù¤ËÊÐ×Hí[+££-èB/c:ó¬ZÍ–OQNòQ|EQm”j¨^EEùVþR¯†¢Ú ¥—5Á¡$yÞË“nXnÏÒ|£Î‚bÛå¹$ô"¡I´Ìœ)lò*ÃfJséž)ÅÊî1R^&—¦ãž»29É¥|ÏAÒ̸ç’"]|.)!º¢`ïì©YÞÙS4TO×&5R-@’m9P¸ühf?"Ý';ÉHCŠªtB”°Iª£'øw~N3AŠi”vÑMh›z¿°%ÑrÌ‹w.&e• ÅO\Ó6¯Âgö¨ßiGã«l‘ýn”†Œú53(mè×Å 4+n÷¾ 8)Žè÷· XÙ°û=,Õ–™ï÷©‘!}ÚÅh¤’œÂÑV»à%ІûE%(­Z¹_8Fºh爬ßFöZO»ŒÆÊYÖ¯ò@(²ö;9¾ÉâÌa¿[‹T|Ã9/Él/›·]½„²ßZqi&WÃõë'PŒXí¼F2UîHû}(¥­M»×á2 #÷!|C9í¼i!ˆ¥÷®éc‡žWƉC÷K P6é— òÀy«€±JýÞ¸ ÊË’G˜m QZ¸^ûa5ÂîZc§´ð¹RäÈ”WB6'ÍÖêd?E°V¢R±¢‰ý¦„Œôëô+K©ÈÙì–Ø÷ðÆÙ±[LYš`·HémVPÃn@(V¿c7àÓà]Çn´ŽÜ¥C7z4*Ý"Ý+ ˜kÐ-`˲ t T¬(¸L [ º…àåèÐð7WzèFÕò|8 èFõÞ¼7 ìäØWð*Vص~óX#¿yÊôŸ|óT©{Rèî¸å,íèÍCéìĽQŬ@„†Þ<¼;´ßAôæChEß½Q¹¬Jÿeoêè õääîè~¿Ž5:Êô©4P‚Þ¬-«5ôÖϳÞÆ)3À¡÷(õß ¼Ñî+ȶƒ7³"7Ó~?`7ògAµ ¸±¿H}ú(¿.¹ijRçÒÎ(‰xfÔ6ã`Å4à*i\ꟸ…ÚÏY5œìA¸‰JEp¯ãõVTÓpWe HØe5”ưïÒ¸TrÖ¹RyžPZÑ*åÊØ&w[k!©üÊ„®5¦¡Q®î,µÕKœB8DöPúï[Œ²Iº¾‹»ÈÜH¯d,¥Ùë¨H,Å·ª’QYXzIÃm|Ç•f£Ò¯@÷APšà¶@‰|iÛ@½¯•ЕÜâ˜Û@±­6B`v×*—t¶ñ¥*{÷,2Ky…m|µç4n%¶ —Q¿•Úî< ±ð¿F[°Òm­ã6®µ·‚äZ…SIFr·ñ%±@§†Û@)›×ôÒ!`*Œ£(v Üp[ Àä4n e õ^ZC I«;ncŠ,ð¨u¡èñ¤‹Vb«í›å'ÒIÉD¯#‰-‚ï¸)A(½°#šÖghPˆPz¡Ü¢jÜÆ©6•ÚÂ> F œ7 (ëÅ¡ˆ±Í¢‚ÐünÖßhàÆ‘Üö¡·Ä;QáŒ:åG‚õDxè¸hd´C ùff:´Ãnf˜:‰b"ð!53¾kJtê–ÜM7¦déÓ“©”Ôæ>#) ŠÈiÀ-”†ûU,QDÎH7†vhμ!ŸÊ·1At1òxt¾’ëù8PDpˬAÆ7RdÁJq¦º‚•0¦7NVÉ´ôìS0r”Í,RèÀ²7¢„–Ÿiyž±À(ò­‘g¡d®äpz¾Ä·Pqæ=|mßù P¢d4zÂwX;ò D‘|éûÝñ½¯­yRi¶rÕ¸JÍÔbÊùÉ;y ¦å§7pîcZônÝg@ÂÙÿøòÊÐù(X6¿¿ì…Ó¥=¬†Ø ½¿SäwÜÛUãšè”D0ùml鈴°k~ {7½–v-]à½üâ¾›’ׯŽÉóhÉ£ÃW2Îdõ›iêq±ñ¹pþ‡IU}US¿Uö¨Ëço.±IÒ[qçkz ÂôKqü’šÅ±E/–0[0áü¥êñòµÐ³õï§~íŒ/vy•±¿*`z}××öhSðÀ‹üö’*tùu+z¤ˆöCïÎO.å…5:J¡,xí­~J'§'•]9úu¼Jߌ÷äH7ØR€Ï/«¢ærçñSŠgÔ2~~œN½ÀG¿;éßÿß°JŒ(0Œòç,˜Ý&^óÂm(ǯ޽¥{ZÄ+‚©ˆ²6þ“óI7âp~x鍨¦²Öa×ÊÛ®îŸÉc£¬õñ!9jLtü: ´ÀúѯÓ|ÉóAïÄ&GQA6ÃÜmè&ט:­¾üÃòð!Š ³ø¸ —Ÿü†&äŒ[;<"yžXÉeé¸{ fSÛnpcïx#ÃðXÑõ‰G¶síuÂ'ïq5­_ó«8øc Ò÷%¥7žnh”Í´}™ÐÝs„2ˆ wŸ7B ü(š»VžêQOR¶i¨OÈëüL£²ª±ž¹¦qºŠÇ§)T²â@4:^±cªn‘'Ÿa ¬üb°µ”¾<ˆõò\·›ÚL®Æû‡ý™¨fþ"ˆ v¡Æ¤¾Z›!ê‰2Ym§Ûìô%ÿ&½òpãÅEyŠ6SŽ©×ÌÔoß¾)¦ž;S”¿§¸JP™½OFýdR¯&õÆS˜û×£Ô ]ïyRL½UöâO=ýõ§Ã—¤Ûû§;üœzüëúðËŠúüð­²ã—ØŽ_%Óo§}|§Á«9‹ bCMÃýÍ·ÈZ‡õ}Bõ™z‚LY™ZKõ(áõa‡ÅžŽ^;~ìPÙÞ{à»ñŠÙ~_Üay´öÿJÖýàÁUµ0ïy-ö£Ùüél>9l?E¨:?˜¨ú>:ü°ziNQÿõÐdŸk]=÷‘VÉÑk‘w‘ëåjµïÚJþÕí»£ù…^‚»ZÄ_xfQu8Þ¹Ÿ–Tþý\‚ŸLêÕ¤*¹¯›Ê­Õ›Œÿ®úñ¡Q}|ôÒ­R‰zÇöÕl>o<û7lxÿºßªÕ!ñý|ÿÿ‚¾ÇN2šÇZ?ƒ{ÃöÂþî6š»¬÷ñžü×¶ÞÇ+û÷½ÞíöfùC ôp©ötEO‚• ýôð¬¸çsÑýÂÕ]Äbs²!ɪ ùá\¥—³ùx6¿<Ô±Æhÿ>›/´¾VL-ÈÃï%ð;ÿô Ë@Š·>îo©éªšu6ãmJ@Qóá'òmŸXú~ÿ¯)6§Ç{D½12>ýðÕ¸ôê‘æcÄ} ³C¨cƒù¯÷x;é©ÞnþPãå°ïñòù;ü-‰[mA5ÕB…6·c;Wrݾ€ýP8~V^!p½9Üúü;¹vöž* nŸ}ÝÛf³·+k#KüAK“§z|ÒßEYõðÃÇ›ìñù—ãýƒ¾Æ ’Ì]³…‚nK¿wûùeñ”xŠœ* ü»•ýOUúÃ`kþ‹:à§óõ“'’ˆ¡ØåÖÿɵJ63GGÔ‰ùCZÞºzÖÙ'NwISå{ÔPÞÈ÷ýHh¥¸ø{ ö¼ ¹}û»e*oòR§ÏW :5=r+üGÓª )ÌQ„óË/å{É´ÑKèæ_[˵”ÛS±*[þT+€^—­‰~ቿV¬Îé$˜üq1—*ÍOúR²ofòÚjmËÁy¶¯wšhi±5ÙÊûÉñ:Õݸ5#j™ÕÊÉÒaYQeÅÖ¬™Vº=k¦OúŸm3ã%Û pÝ…vÆîó–ÿ‚©ûBßlº.)úóK½„Íš¨èâ勽IyHËÞù¨¿ruI~P`þ‹õ‹8|hu§‡ª‹äÍé­_]šñPfB^¡m¡S…¦´h£[ÏIrÊüÒòîOú±AUú£yÊ ¿Ïô£Jö¿¥,8´Ø’Iûõûš2ú™¸³å6k])3æöõà×TùgVuA‚Ø/ÿ!ÁG"7ÁG°ªlêTÿ=L]Oí[XÁ~¸Ï9=Ûå#_MÃUƒ_ÒÌ3׫ÿæàÍr×Ó{5?¡¾> stream xœ]O»ƒ0 ÜóþƒêcA,ta ªÚþ@p”' aèß—èÐá,ïN>˶»ul#ÈGpø¢Ʋ4»% Á@£eQV -Æå‰“òB¶½òï'X d6~WÉgyºæU¹…Ðiš½B ŠGuQ4µ1 ÖÒÌî¼`“QqóJЦÇMÀ%☛æ&©€eú=ãO)X!¾o:Süendstream endobj 141 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 371 >> stream xœcd`ab`ddätö vŠ04qôH3þaú!ËÜÝýSñ§«,¿“ƒ¬È½†¾næn–5ß }ÏüžÍÿ=C€™‘±°²Ã9¿ ²(3=£DA#YSÁÐÒÒ\GÁÈÀÀRÁ17µ(391OÁ7±$#57±ÈÉQÎOÎL-©TаÉ())°Ò×///×KÌ-ÖË/J·ÓÔQ(Ï,ÉPJ-N-*KMQpËÏ+QðKÌMU€9TÆpÎÏ-(-I-RðÍOI-Êc``` b`0f`bddÉÿÑÁ÷3¨{Á÷m—¿—>d|ö]åûüïòÌ?Ê¿_=5ëÕÆîïß#çvÇÉýnaK¯oH“ÿóŠ=}Zý¹Ýl§ÎX'ÿ}öwföï¿7Ï,žÖ½°[òÛü9Ý+åÿ˜}ÿ*ú½ô»ûwޱžµ¡¿ùåùJþˆ›ó=!Ûï¤éì/¹¾ËpËq±˜Ïçáü.3‡·‡‡mŠxendstream endobj 142 0 obj << /Filter /FlateDecode /Length 3298 >> stream xœíZKs·¾³\¹çÆòi6%®ð~ødùYvd›¶V+’"eS\™¤¬8ªJþz¾FcÙYËR’ªR>b€F?¾ïCÏþ|¨ÖúPÑõÿ'Ïî~­ýáùÍ:Y©µQÊä8<¥¡V6¤áju¤Ö>çMy¬SÖ*·ªœ}ÊÃÙêÈZ,¥õð½—sJA.ñ·²° Þ•×´RQ¥á…Xøšf$gL[L®µ]™¸VÚèáOÆ—}ÕÚæd}.é5³NÊ ‘e){´4bs^N”QzØÞ¬hïèmgÒSš3^NV47¸ä‡6ɦ4|{S;ãºÉWxjSÔ! çlHÆ{_Ó•“‹lªSÊg7ðÖVyØôŒž×C>§ErÐ:á4“Ù;»d§œÜE¸I:„ãcs†'/šÏ:O6ÿß´7ÿz|'«ŽÆL9²~Bà|)ŽÎ*ب‡ÍÊ2' ¿´` ïßlàG$̽U1ÇûáÃâ N¿;( ™ì(O=¦iëÉÏcìoiˆ-tL%ö*yE™.Ÿ_¯¼^#%ȑӋÏamÖøÙêKµÑB¶©[ᢽ·¥ÅR2žCÌoh1¼)7¸b—käM¥Í4’ë¸$"¥û…Ì`X‚„b°ÚZí`vY4ª€@ÊL–g‡—-¿9©•B±9Åì¬h±[aX™2gp2roôAKÛ—2Vg·І€ZƒÔÕSž•µQH€´ï Qp>É–õr~Æ™„T¾YF$ZÄ"™‚j ãä©„/¨Ý‘uzí‘jGÚ®½Ë†M/Õâ ª¥îmL~7eŒ ü\0Îh`€œ½qÞøþ’7Iä®2?å§ÊáaÙÛ)Çu‹Þöq™›=’Ãc“lgW'ÊÌìà„ªZybÊ oŽM%Æu§½!)ŽÕå0HT-óµH¹ÙâÓþ›Ö´©.Èó¡¨.J”€DQyxÜÂÍ!A÷TPÁŸÀf3|[À›÷³¡ü3Š«zʾɽ™ënÉ•1ú•Ï=dj+¼t½¢l"ÆÂÇ8ëÎKžŠ!ú. tãš0f^#KGoUɶ¹äœdçó³ùÚ ‡†'>oÓ×­0¸ªÑ\)±í.P0'q ào»-Ê_V>DÕ‘xî>ZÞôÑf]ÿšpüÁ0±#òLµDdVm8Æ)Ù$'ÓgËÔ§ÒM”ö.mÜPH¦’Î69¢¶vM>š^ñÑE>’èóK {.m&â@Èt2û ‘Üœý‘”œ*-íÙðˆQöðÈt•Ÿ6Ĭ™‘³¡Ù¤†Ø8¤ ‰ÇÒ៤FË‘QÝÈQñ·  '”#èKÆ….Ú”M‰²É$ãyš•1}MŸ×ÁQdÇõ>á5Ryȇ“SRÂÁN‚¢axôˆˆ>$—HBFm% 탵 )6Ï›-‚/™÷ÉíóbnÕå• C—#DIªÙ3>­Sv¯.Ÿ€¨Óå»×-Õпäª1vä(ørNGEö®› ï’ƒ‹'§2'jàá§á> aî]QªÚì\e4V;î\bÄîJÈ=þÞBÇHOz¼Æ‰Ë(ï¤a¬Ëä;=JÍå +5!퇓zú {ÊD(ž/šhz(2ùh€g à›JD( h³( ‚+zÔ™'½q逖[%™•ð§Ï AP»$+_CÓYUÔ¨!áP- :ÍP˜ÛvG§b}!îˆôÀ'X¹<ÉÖÞÏ¥–¼â§ºÐÏèB!£Ÿµ¢ò…ð1寨 E®ɼ÷ÞÈ•¥ /^²Ak…uêÊr¹ù.²Hs nÔûŒsCà`j9éCi¥§{G¯i·`¸<&ˆ"a$©%+‰€Ð :˜XWƒ2_‰M}–tîDà ²/@TL‘hÆØ 0tùmx9Šðw«D¼d페©:á²ÿæ¼ZŸ·ÏΚ ‘RY°b)Žñü½:8mt!¹ l°ÁÙ°öÉŽ2+/ ù.qΔŠA5YjÙ#5,ñrȹ>DÈ4©-eNÒºk%ýlì(ëøº¬³õÞËgâô$ —“`ƒb‡wíœÏT½Ø‚ÿtÉoI1‚áÉòlE¦ú΋W”ǰç ñNS»o$ÿoË•œ|ø+YîŠó·OxÀ¶ûš?{XFÞ+ȸgOšmÛëº2X@Êz~ÑІ‚õö‘ø-õr^×¼ \匷ǿ†\H˜Uúss܌֢²»Òts<—eu6]n˜ }ZkçG˜˜w¬jqm¯iïÝÕÅA$khàqÉûY!‹F_4¥|-Pùj•®Î Œs×*yÒÞc:LXGéùðÃV—RöŠt¾)—>§m¨ h³VÒÍ; Ø¥ÝÄÜÝØ­q+fð‹!L.3^Óƒ¸Ìè$àºï)©n'¡§nK\-¶7ˆý8½«ùZð,HÅ-ÃDûÊäPqÈ:Ãr>ÛÒxêèûU²÷Ô ($¡’‘GCŠeºpRK®awø!IdÑ#ÍÝÛ ‰-¢„º û/ˆ «&¥é‚ö î1E–pE~à‹ œ /2⦳Üp®Û…N ‘;í›ÊeƒÆÞüò£ê Ü‹g½Ä0ï%â‡÷;…Á@€¢?Þ S¶ §(LßQI[ÖÔ)Xç:¸ÙŽ. ΛÐuÊùuØßÑì?.J_¦§ƒiÊ,g6,u 9Ï›` ¹u6qFò/gµ¢‹ÆKRºàD÷¹k$×»9F¿¬¼/N.ð:—] aáûÆ’ø¤!»ÜB><_òRV(•ûòì•RYÏJBE%¶#c;ȼ¯uÀ…×å…µ@Àem½§½<÷v²î ´+.dóV`è'04¿ äÙ¤_AMUÈó!Øú>¼¥×÷€òuø¼?y:£Ï¼Î=BŽéÈ(a…*3¬b½Ò¶4+Ûã «,È@ÏŽs—]þ…“{$ÇVL¿z¯ýcñLÔ±À¥òøÁÁñŸ~è(§›”·vjŒ‰áÎ'£"ë6cãÈítkJ\É jVR†ÌǽM…Y·ÆF(ÿ4oÖŒ¹_*¹Ü ž‘·læq£¹4óÆ~ªäÀØ6ØÔkÆ©Ø{vá/Ù²|`®*fÓùm{O× ˆ?õpJ {Iµ¡÷+OÊʇR]ç]`Šød /¿ÝÈ/“î[å䙟ºj:¹i3½Ï¨‚Û"Wù ¨}n’v_5»å)oØÃ¨Èás®¤ò1%/—«éqwÛš¼Ù·Â§Cì4Oç"A¶ô˜ tëµ:u}, GãQØn)èjì͘:(‰^–·Óü½mÊ·¶qù.©ï•CÈSW“ýmn¼Ií»ŽN{ÞýtI•‡_n,þq̯!äUm½røBo5B@øóN·*£÷dÁ1ï»Kᜠê=²Äv|ÜD¢¼w°I#qµ{¯Ëò¾: ”…bùÑÃõþßH·ô m ’:׸Ó%˜+qNU¿ÉÔáíâði^µá¦ ×møi^·á‹ÅÝÄbï·áeèSýø]‡/Úp݆›6¼]5w½ÍÏH¼dMfï|B²7ž±0£öyx:Vj$ÉË:Ëmom×뮂¤©·7nèKΙ¥üÿ¹æ¿Â5´rê[&¢“qÊ(N¢Ôëo0iß­vHʳCv¾‘Öœš}$Ã%_‚P#º`Ñ£*Æ0Æ'Úô¦´0‹²ÆeÂír•G¯qùyôtkÚÛÛ’±þ©P«±H?±ðåâ¡ëŒñƒ}ArxèûVQ"ê\µåÇNs&FÏUÜCf3;Ž›ß\êãi{7ªš±å¨3¹?;ewÁ‚”øËÇnaNz¦tÖߨð5ŽØ¸ŽÏ Åîe‚ômNò•ÉcÃùMœ·ø Eö•¿3›¬øßc:K?æHvÄò%Fœ&éôusOÚðb‘ÿ,®+V¸™/VØë¢ ßoÃÇm¸};V> /W [ 1 3 1 ] /Info 3 0 R /Root 2 0 R /Size 144 /ID [] >> stream xœcb&F~0ù‰ $À8JEþgà±=d³?…Þ[¡ÑÐ#6ܤÿž²%‚BL> Dò‹ƒH¹ÉùDJœ‘|] ’¹Dr€HÆ× ’i˜½̶“, R@DЇ€e€É— Rd£„7XM!ˆdí‘"¡`½é ’{B½ä!0ìþR°š÷YưÈo°ÛôÀ¾`Çþ+°øYɫô`Æ endstream endobj startxref 109560 %%EOF flexmix/inst/doc/flexmix-intro.Rnw0000644000176200001440000010260313430477461016760 0ustar liggesusers% % Copyright (C) 2004-2005 Friedrich Leisch % $Id: flexmix-intro.Rnw 5156 2019-02-12 08:11:16Z gruen $ % \documentclass[nojss]{jss} \title{FlexMix: A General Framework for Finite Mixture Models and Latent Class Regression in \proglang{R}} \Plaintitle{FlexMix: A General Framework for Finite Mixture Models and Latent Class Regression in R} \Shorttitle{FlexMix: Finite Mixture Models in \proglang{R}} \author{Friedrich Leisch\\Universit\"at f\"ur Bodenkultur Wien} \Plainauthor{Friedrich Leisch} \Address{ Friedrich Leisch\\ Institut f\"ur Angewandte Statistik und EDV\\ Universit\"at f\"ur Bodenkultur Wien\\ Peter Jordan Stra\ss{}e 82\\ 1190 Wien, Austria\\ E-mail: \email{Friedrich.Leisch@boku.ac.at}\\ URL: \url{http://www.statistik.lmu.de/~leisch/} } \usepackage[utf8]{inputenc} \usepackage{listings} \newcommand{\R}{\proglang{R}} <>= set.seed(1504) options(width=70, prompt = "R> ", continue = "+ ", useFancyQuotes = FALSE) grDevices::ps.options(family="Times") library("graphics") library("flexmix") data("NPreg") @ \Abstract{ This article was originally published as \cite{flexmix:Leisch:2004a} in the \emph{Journal of Statistical Software}. FlexMix implements a general framework for fitting discrete mixtures of regression models in the \R{} statistical computing environment: three variants of the EM algorithm can be used for parameter estimation, regressors and responses may be multivariate with arbitrary dimension, data may be grouped, e.g., to account for multiple observations per individual, the usual formula interface of the \proglang{S} language is used for convenient model specification, and a modular concept of driver functions allows to interface many different types of regression models. Existing drivers implement mixtures of standard linear models, generalized linear models and model-based clustering. FlexMix provides the E-step and all data handling, while the M-step can be supplied by the user to easily define new models. } \Keywords{\proglang{R}, finite mixture models, model based clustering, latent class regression} \Plainkeywords{R, finite mixture models, model based clustering, latent class regression} \Volume{11} \Issue{8} \Month{October} \Year{2004} \Submitdate{2004-04-19} \Acceptdate{2004-10-18} %%\usepackage{Sweave} %% already provided by jss.cls %%\VignetteIndexEntry{FlexMix: A General Framework for Finite Mixture Models and Latent Class Regression in R} %%\VignetteDepends{flexmix} %%\VignetteKeywords{R, finite mixture models, model based clustering, latent class regression} %%\VignettePackage{flexmix} \begin{document} \section{Introduction} \label{sec:introduction} Finite mixture models have been used for more than 100 years, but have seen a real boost in popularity over the last decade due to the tremendous increase in available computing power. The areas of application of mixture models range from biology and medicine to physics, economics and marketing. On the one hand these models can be applied to data where observations originate from various groups and the group affiliations are not known, and on the other hand to provide approximations for multi-modal distributions \citep{flexmix:Everitt+Hand:1981,flexmix:Titterington+Smith+Makov:1985,flexmix:McLachlan+Peel:2000}. In the 1990s finite mixture models have been extended by mixing standard linear regression models as well as generalized linear models \citep{flexmix:Wedel+DeSarbo:1995}. An important area of application of mixture models is market segmentation \citep{flexmix:Wedel+Kamakura:2001}, where finite mixture models replace more traditional cluster analysis and cluster-wise regression techniques as state of the art. Finite mixture models with a fixed number of components are usually estimated with the expectation-maximization (EM) algorithm within a maximum likelihood framework \citep{flexmix:Dempster+Laird+Rubin:1977} and with MCMC sampling \citep{flexmix:Diebolt+Robert:1994} within a Bayesian framework. \newpage The \R{} environment for statistical computing \citep{flexmix:R-Core:2004} features several packages for finite mixture models, including \pkg{mclust} for mixtures of multivariate Gaussian distributions \citep{flexmix:Fraley+Raftery:2002,flexmix:Fraley+Raftery:2002a}, \pkg{fpc} for mixtures of linear regression models \citep{flexmix:Hennig:2000} and \pkg{mmlcr} for mixed-mode latent class regression \citep{flexmix:Buyske:2003}. There are three main reasons why we have chosen to write yet another software package for EM estimation of mixture models: \begin{itemize} \item The existing implementations did not cover all cases we needed for our own research (mainly marketing applications). \item While all \R{} packages mentioned above are open source and hence can be extended by the user by modifying the source code, we wanted an implementation where extensibility is a main design principle to enable rapid prototyping of new mixture models. \item We include a sampling-based variant of the EM-algorithm for models where weighted maximum likelihood estimation is not available. FlexMix has a clean interface between E- and M-step such that variations of both are easy to combine. \end{itemize} This paper is organized as follows: First we introduce the mathematical models for latent class regression in Section~\ref{sec:latent-class-regr} and shortly discuss parameter estimation and identifiability. Section~\ref{sec:using-flexmix} demonstrates how to use FlexMix to fit models with the standard driver for generalized linear models. Finally, Section~\ref{sec:extending-flexmix} shows how to extend FlexMix by writing new drivers using the well-known model-based clustering procedure as an example. \section{Latent class regression} \label{sec:latent-class-regr} Consider finite mixture models with $K$ components of form \begin{equation}\label{eq:1} h(y|x,\psi) = \sum_{k = 1}^K \pi_k f(y|x,\theta_k) \end{equation} \begin{displaymath} \pi_k \geq 0, \quad \sum_{k = 1}^K \pi_k = 1 \end{displaymath} where $y$ is a (possibly multivariate) dependent variable with conditional density $h$, $x$ is a vector of independent variables, $\pi_k$ is the prior probability of component $k$, $\theta_k$ is the component specific parameter vector for the density function $f$, and $\psi=(\pi_1,,\ldots,\pi_K,\theta_1',\ldots,\theta_K')'$ is the vector of all parameters. If $f$ is a univariate normal density with component-specific mean $\beta_k'x$ and variance $\sigma^2_k$, we have $\theta_k = (\beta_k', \sigma_k^2)'$ and Equation~(\ref{eq:1}) describes a mixture of standard linear regression models, also called \emph{latent class regression} or \emph{cluster-wise regression} \citep{flexmix:DeSarbo+Cron:1988}. If $f$ is a member of the exponential family, we get a mixture of generalized linear models \citep{flexmix:Wedel+DeSarbo:1995}, known as \emph{GLIMMIX} models in the marketing literature \citep{flexmix:Wedel+Kamakura:2001}. For multivariate normal $f$ and $x\equiv1$ we get a mixture of Gaussians without a regression part, also known as \emph{model-based clustering}. The posterior probability that observation $(x,y)$ belongs to class $j$ is given by \begin{equation}\label{eq:3} \Prob(j|x, y, \psi) = \frac{\pi_j f(y | x, \theta_j)}{\sum_k \pi_k f(y | x, \theta_k)} \end{equation} The posterior probabilities can be used to segment data by assigning each observation to the class with maximum posterior probability. In the following we will refer to $f(\cdot|\cdot, \theta_k)$ as \emph{mixture components} or \emph{classes}, and the groups in the data induced by these components as \emph{clusters}. \subsection{Parameter estimation} \label{sec:parameter-estimation} The log-likelihood of a sample of $N$ observations $\{(x_1,y_1),\ldots,(x_N,y_N)\}$ is given by \begin{equation}\label{eq:4} \log L = \sum_{n=1}^N \log h(y_n|x_n,\psi) = \sum_{n=1}^N \log\left(\sum_{k = 1}^K \pi_kf(y_n|x_n,\theta_k) \right) \end{equation} and can usually not be maximized directly. The most popular method for maximum likelihood estimation of the parameter vector $\psi$ is the iterative EM algorithm \citep{flexmix:Dempster+Laird+Rubin:1977}: \begin{description} \item[Estimate] the posterior class probabilities for each observation \begin{displaymath} \hat p_{nk} = \Prob(k|x_n, y_n, \hat \psi) \end{displaymath} using Equation~(\ref{eq:3}) and derive the prior class probabilities as \begin{displaymath} \hat\pi_k = \frac1N \sum_{n=1}^N \hat p_{nk} \end{displaymath} \item[Maximize] the log-likelihood for each component separately using the posterior probabilities as weights \begin{equation}\label{eq:2} \max_{\theta_k} \sum_{n=1}^N \hat p_{nk} \log f(y_n | x_n, \theta_k) \end{equation} \end{description} The E- and M-steps are repeated until the likelihood improvement falls under a pre-specified threshold or a maximum number of iterations is reached. The EM algorithm cannot be used for mixture models only, but rather provides a general framework for fitting models on incomplete data. Suppose we augment each observation $(x_n,y_n)$ with an unobserved multinomial variable $z_n = (z_{n1},\ldots,z_{nK})$, where $z_{nk}=1$ if $(x_n,y_n)$ belongs to class $k$ and $z_{nk}=0$ otherwise. The EM algorithm can be shown to maximize the likelihood on the ``complete data'' $(x_n,y_n,z_n)$; the $z_n$ encode the missing class information. If the $z_n$ were known, maximum likelihood estimation of all parameters would be easy, as we could separate the data set into the $K$ classes and estimate the parameters $\theta_k$ for each class independently from the other classes. If the weighted likelihood estimation in Equation~(\ref{eq:2}) is infeasible for analytical, computational, or other reasons, then we have to resort to approximations of the true EM procedure by assigning the observations to disjoint classes and do unweighted estimation within the groups: \begin{displaymath} \max_{\theta_k} \sum_{n: z_{nk=1}} \log f(y_n | x_n, \theta_k) \end{displaymath} This corresponds to allow only 0 and 1 as weights. Possible ways of assigning the data into the $K$ classes are \begin{itemize} \item \textbf{hard} \label{hard} assignment to the class with maximum posterior probability $p_{nk}$, the resulting procedure is called maximizing the \emph{classification likelihood} by \cite{flexmix:Fraley+Raftery:2002}. Another idea is to do \item \textbf{random} assignment to classes with probabilities $p_{nk}$, which is similar to the sampling techniques used in Bayesian estimation (although for the $z_n$ only). \end{itemize} Well known limitations of the EM algorithm include that convergence can be slow and is to a local maximum of the likelihood surface only. There can also be numerical instabilities at the margin of parameter space, and if a component gets to contain only a few observations during the iterations, parameter estimation in the respective component may be problematic. E.g., the likelihood of Gaussians increases without bounds for $\sigma^2\to 0$. As a result, numerous variations of the basic EM algorithm described above exist, most of them exploiting features of special cases for $f$. \subsection{Identifiability} \label{sec:identifiability} An open question is still identifiability of many mixture models. A comprehensive overview of this topic is beyond the scope of this paper, however, users of mixture models should be aware of the problem: \begin{description} \item[Relabelling of components:] Mixture models are only identifiable up to a permutation of the component labels. For EM-based approaches this only affects interpretation of results, but is no problem for parameter estimation itself. \item[Overfitting:] If a component is empty or two or more components have the same parameters, the data generating process can be represented by a smaller model with fewer components. This kind of unidentifiability can be avoided by requiring that the prior weights $\pi_k$ are not equal to zero and that the component specific parameters are different. \item[Generic unidentifiability:] It has been shown that mixtures of univariate normal, gamma, exponential, Cauchy and Poisson distributions are identifiable, while mixtures of discrete or continuous uniform distributions are not identifiable. A special case is the class of mixtures of binomial and multinomial distributions which are only identifiable if the number of components is limited with respect to, e.g., the number of observations per person. See \cite{flexmix:Everitt+Hand:1981}, \cite{flexmix:Titterington+Smith+Makov:1985}, \cite{flexmix:Grun:2002} and references therein for details. \end{description} FlexMix tries to avoid overfitting because of vanishing prior probabilities by automatically removing components where the prior $\pi_k$ falls below a user-specified threshold. Automated diagnostics for generic identifiability are currently under investigation. Relabelling of components is in some cases more of a nuisance than a real problem (``component 2 of the first run may be component 3 in the second run''), more serious are interactions of component relabelling and categorical predictor variables, see \cite{flexmix:Grun+Leisch:2004} for a discussion and how bootstrapping can be used to assess identifiability of mixture models. \pagebreak[4] \section{Using FlexMix} \label{sec:using-flexmix} \SweaveOpts{width=12,height=8,eps=FALSE,keep.source=TRUE} The standard M-step \texttt{FLXMRglm()} of FlexMix is an interface to R's generalized linear modelling facilities (the \texttt{glm()} function). As a simple example we use artificial data with two latent classes of size \Sexpr{nrow(NPreg)/2} each: \begin{center} \begin{tabular}{ll} Class~1: & $ y = 5x+\epsilon$\\ Class~2: & $ y = 15+10x-x^2+\epsilon$\\ \end{tabular} \end{center} with $\epsilon\sim N(0,9)$ and prior class probabilities $\pi_1=\pi_2=0.5$, see the left panel of Figure~\ref{fig:npreg}. We can fit this model in \R{} using the commands <<>>= library("flexmix") data("NPreg") m1 <- flexmix(yn ~ x + I(x^2), data = NPreg, k = 2) m1 @ and get a first look at the estimated parameters of mixture component~1 by <<>>= parameters(m1, component = 1) @ and <<>>= parameters(m1, component = 2) @ for component~2. The paramter estimates of both components are close to the true values. A cross-tabulation of true classes and cluster memberships can be obtained by <<>>= table(NPreg$class, clusters(m1)) @ The summary method <<>>= summary(m1) @ gives the estimated prior probabilities $\hat\pi_k$, the number of observations assigned to the corresponding clusters, the number of observations where $p_{nk}>\delta$ (with a default of $\delta=10^{-4}$), and the ratio of the latter two numbers. For well-seperated components, a large proportion of observations with non-vanishing posteriors $p_{nk}$ should also be assigned to the corresponding cluster, giving a ratio close to 1. For our example data the ratios of both components are approximately 0.7, indicating the overlap of the classes at the cross-section of line and parabola. \begin{figure}[htbp] \centering <>= par(mfrow=c(1,2)) plot(yn~x, col=class, pch=class, data=NPreg) plot(yp~x, col=class, pch=class, data=NPreg) @ \caption{Standard regression example (left) and Poisson regression (right).} \label{fig:npreg} \end{figure} Histograms or rootograms of the posterior class probabilities can be used to visually assess the cluster structure \citep{flexmix:Tantrum+Murua+Stuetzle:2003}, this is now the default plot method for \texttt{"flexmix"} objects \citep{flexmix:Leisch:2004}. Rootograms are very similar to histograms, the only difference is that the height of the bars correspond to square roots of counts rather than the counts themselves, hence low counts are more visible and peaks less emphasized. \begin{figure}[htbp] \centering <>= print(plot(m1)) @ \caption{The plot method for \texttt{"flexmix"} objects, here obtained by \texttt{plot(m1)}, shows rootograms of the posterior class probabilities.} \label{fig:root1} \end{figure} Usually in each component a lot of observations have posteriors close to zero, resulting in a high count for the corresponing bin in the rootogram which obscures the information in the other bins. To avoid this problem, all probabilities with a posterior below a threshold are ignored (we again use $10^{-4}$). A peak at probability 1 indicates that a mixture component is well seperated from the other components, while no peak at 1 and/or significant mass in the middle of the unit interval indicates overlap with other components. In our simple example the components are medium well separated, see Figure~\ref{fig:root1}. Tests for significance of regression coefficients can be obtained by <<>>= rm1 <- refit(m1) summary(rm1) @ Function \texttt{refit()} fits weighted generalized linear models to each component using the standard \R{} function \texttt{glm()} and the posterior probabilities as weights, see \texttt{help("refit")} for details. The data set \texttt{NPreg} also includes a response from a generalized linear model with a Poisson distribution and exponential link function. The two classes of size \Sexpr{nrow(NPreg)/2} each have parameters \begin{center} \begin{tabular}{ll} Class~1: & $ \mu_1 = 2 - 0.2x$\\ Class~2: & $ \mu_2 = 1 + 0.1x$\\ \end{tabular} \end{center} and given $x$ the response $y$ in group $k$ has a Poisson distribution with mean $e^{\mu_k}$, see the right panel of Figure~\ref{fig:npreg}. The model can be estimated using <>= options(width=55) @ <<>>= m2 <- flexmix(yp ~ x, data = NPreg, k = 2, model = FLXMRglm(family = "poisson")) summary(m2) @ <>= options(width=65) @ \begin{figure}[htbp] \centering <>= print(plot(m2)) @ \caption{\texttt{plot(m2)}} \label{fig:root2} \end{figure} Both the summary table and the rootograms in Figure~\ref{fig:root2} clearly show that the clusters of the Poisson response have much more overlap. For our simple low-dimensional example data the overlap of the classes is obvious by looking at scatterplots of the data. For data in higher dimensions this is not an option. The rootograms and summary tables for \texttt{"flexmix"} objects work off the densities or posterior probabilities of the observations and thus do not depend on the dimensionality of the input space. While we use simple 2-dimensional examples to demonstrate the techniques, they can easily be used on high-dimensional data sets or models with complicated covariate structures. \subsection{Multiple independent responses} \label{sec:mult-indep-resp} If the response $y=(y_1,\ldots,y_D)'$ is $D$-dimensional and the $y_d$ are mutually independent the mixture density in Equation~(\ref{eq:1}) can be written as \begin{eqnarray*} h(y|x,\psi) &=& \sum_{k = 1}^K \pi_k f(y|x,\theta_k)\\ &=& \sum_{k = 1}^K \pi_k \prod_{d=1}^D f_d(y|x,\theta_{kd}) \end{eqnarray*} To specify such models in FlexMix we pass it a list of models, where each list element corresponds to one $f_d$, and each can have a different set of dependent and independent variables. To use the Gaussian and Poisson responses of data \texttt{NPreg} simultaneously, we use the model specification \begin{Sinput} > m3 = flexmix(~x, data=NPreg, k=2, + model=list(FLXMRglm(yn~.+I(x^2)), + FLXMRglm(yp~., family="poisson"))) \end{Sinput} <>= m3 <- flexmix(~ x, data = NPreg, k = 2, model=list(FLXMRglm(yn ~ . + I(x^2)), FLXMRglm(yp ~ ., family = "poisson"))) @ Note that now three model formulas are involved: An overall formula as first argument to function \texttt{flexmix()} and one formula per response. The latter ones are interpreted relative to the overall formula such that common predictors have to be specified only once, see \texttt{help("update.formula")} for details on the syntax. The basic principle is that the dots get replaced by the respective terms from the overall formula. The rootograms show that the posteriors of the two-response model are shifted towards 0 and 1 (compared with either of the two univariate models), the clusters are now well-separated. \begin{figure}[htbp] \centering <>= print(plot(m3)) @ \caption{\texttt{plot(m3)}} \label{fig:root3} \end{figure} \subsection{Repeated measurements} \label{sec:repe-meas} If the data are repeated measurements on $M$ individuals, and we have $N_m$ observations from individual $m$, then the log-likelihood in Equation~(\ref{eq:4}) can be written as \begin{displaymath} \log L = \sum_{m=1}^M \sum_{n=1}^{N_m} \log h(y_{mn}|x_{mn},\psi), \qquad \sum_{m=1}^M N_m = N \end{displaymath} and the posterior probability that individual $m$ belongs to class $j$ is given by \begin{displaymath} \Prob(j|m) = \frac{\pi_j \prod_{n=1}^{N_m} f(y_{mn} | x_{mn}, \theta_j)}{\sum_k \pi_k \prod_{n=1}^{N_m} f(y_{mn} | x_{mn}, \theta_k)} \end{displaymath} where $(x_{mn}, y_{mn})$ is the $n$-th observation from individual $m$. As an example, assume that the data in \texttt{NPreg} are not 200 independent observations, but 4 measurements each from 50 persons such that $\forall m: N_m=4$. Column \texttt{id2} of the data frame encodes such a grouping and can easily be used in FlexMix: <<>>= m4 <- flexmix(yn ~ x + I(x^2) | id2, data = NPreg, k = 2) summary(m4) @ Note that convergence of the EM algorithm is much faster with grouping and the two clusters are now perfectly separated. \subsection{Control of the EM algorithm} \label{sec:control-em-algorithm} Details of the EM algorithm can be tuned using the \texttt{control} argument of function \texttt{flexmix()}. E.g., to use a maximum number of 15 iterations, report the log-likelihood at every 3rd step and use hard assignment of observations to clusters (cf. page~\pageref{hard}) the call is <<>>= m5 <- flexmix(yn ~ x + I(x^2), data = NPreg, k = 2, control = list(iter.max = 15, verbose = 3, classify = "hard")) @ Another control parameter (\texttt{minprior}, see below for an example) is the minimum prior probability components are enforced to have, components falling below this threshold (the current default is 0.05) are removed during EM iteration to avoid numerical instabilities for components containing only a few observations. Using a minimum prior of 0 disables component removal. \subsection{Automated model search} In real applications the number of components is unknown and has to be estimated. Tuning the minimum prior parameter allows for simplistic model selection, which works surprisingly well in some situations: <<>>= m6 <- flexmix(yp ~ x + I(x^2), data = NPreg, k = 4, control = list(minprior = 0.2)) m6 @ Although we started with four components, the algorithm converged at the correct two component solution. A better approach is to fit models with an increasing number of components and compare them using AIC or BIC. As the EM algorithm converges only to the next local maximum of the likelihood, it should be run repeatedly using different starting values. The function \texttt{stepFlexmix()} can be used to repeatedly fit models, e.g., <<>>= m7 <- stepFlexmix(yp ~ x + I(x^2), data = NPreg, control = list(verbose = 0), k = 1:5, nrep = 5) @ runs \texttt{flexmix()} 5 times for $k=1,2,\ldots,5$ components, totalling in 25 runs. It returns a list with the best solution found for each number of components, each list element is simply an object of class \texttt{"flexmix"}. To find the best model we can use <<>>= getModel(m7, "BIC") @ and choose the number of components minimizing the BIC. \section{Extending FlexMix} \label{sec:extending-flexmix} One of the main design principles of FlexMix was extensibility, users can provide their own M-step for rapid prototyping of new mixture models. FlexMix was written using S4 classes and methods \citep{flexmix:Chambers:1998} as implemented in \R{} package \pkg{methods}. The central classes for writing M-steps are \texttt{"FLXM"} and \texttt{"FLXcomponent"}. Class \texttt{"FLXM"} specifies how the model is fitted using the following slots: \begin{description} \item[fit:] A \texttt{function(x,y,w)} returning an object of class \texttt{"FLXcomponent"}. \item[defineComponent:] Expression or function constructing the object of class \texttt{"FLXcomponent"}. \item[weighted:] Logical, specifies if the model may be fitted using weighted likelihoods. If \texttt{FALSE}, only hard and random classification are allowed (and hard classification becomes the default). \item[formula:] Formula relative to the overall model formula, default is \verb|.~.| \item[name:] A character string describing the model, this is only used for print output. \end{description} The remaining slots of class \texttt{"FLXM"} are used internally by FlexMix to hold data, etc. and omitted here, because they are not needed to write an M-step driver. The most important slot doing all the work is \texttt{fit} holding a function performing the maximum likelihood estimation described in Equation~(\ref{eq:2}). The \texttt{fit()} function returns an object of class \texttt{"FLXcomponent"} which holds a fitted component using the slots: \begin{description} \item[logLik:] A \texttt{function(x,y)} returning the log-likelihood for observations in matrices \texttt{x} and \texttt{y}. \item[predict:] A \texttt{function(x)} predicting \texttt{y} given \texttt{x}. \item[df:] The degrees of freedom used by the component, i.e., the number of estimated parameters. \item[parameters:] An optional list containing model parameters. \end{description} In a nutshell class \texttt{"FLXM"} describes an \emph{unfitted} model, whereas class \texttt{"FLXcomponent"} holds a \emph{fitted} model. \lstset{frame=trbl,basicstyle=\small\tt,stepnumber=5,numbers=left} \begin{figure}[tb] \centering \begin{minipage}{0.94\textwidth} \lstinputlisting{mymclust.R} \end{minipage} \caption{M-step for model-based clustering: \texttt{mymclust} is a simplified version of the standard FlexMix driver \texttt{FLXmclust}.} \label{fig:mymclust.R} \end{figure} \subsection{Writing an M-step driver} \label{sec:writing-an-m} Figure~\ref{fig:mymclust.R} shows an example driver for model-based clustering. We use function \texttt{dmvnorm()} from package \pkg{mvtnorm} for calculation of multivariate Gaussian densities. In line~5 we create a new \texttt{"FLXMC"} object named \texttt{retval}, which is also the return value of the driver. Class \texttt{"FLXMC"} extends \texttt{"FLXM"} and is used for model-based clustering. It contains an additional slot with the name of the distribution used. All drivers should take a formula as their first argument, this formula is directly passed on to \texttt{retval}. In most cases authors of new FlexMix drivers need not worry about formula parsing etc., this is done by \texttt{flexmix} itself. In addition we have to declare whether the driver can do weighted ML estimation (\texttt{weighted=TRUE}) and give a name to our model. The remainder of the driver creates a \texttt{fit()} function, which takes regressors \texttt{x}, response \texttt{y} and weights \texttt{w}. For multivariate Gaussians the maximum likelihood estimates correspond to mean and covariance matrix, the standard R function \texttt{cov.wt()} returns a list containing estimates of the weighted covariance matrix and the mean for given data. Our simple example performs clustering without a regression part, hence $x$ is ignored. If \texttt{y} has $D$ columns, we estimate $D$ parameters for the mean and $D(D-1)/2$ parameters for the covariance matrix, giving a total of $(3D+D^2)/2$ parameters (line~11). As an additional feature we allow the user to specify whether the covariance matrix is assumed to be diagonal or a full matrix. For \texttt{diagonal=TRUE} we use only the main diagonal of the covariance matrix (line~14) and the number of parameters is reduced to $2D$. In addition to parameter estimates, \texttt{flexmix()} needs a function calculating the log-likelihood of given data $x$ and $y$, which in our example is the log-density of a multivariate Gaussian. In addition we have to provide a function predicting $y$ given $x$, in our example simply the mean of the Gaussian. Finally we create a new \texttt{"FLXcomponent"} as return value of function \texttt{fit()}. Note that our internal functions \texttt{fit()}, \texttt{logLik()} and \texttt{predict()} take only \texttt{x}, \texttt{y} and \texttt{w} as arguments, but none of the model-specific parameters like means and covariances, although they use them of course. \R{} uses \emph{lexical scoping} rules for finding free variables \citep{flexmix:Gentleman+Ihaka:2000}, hence it searches for them first in the environment where a function is defined. E.g., the \texttt{fit()} function uses the variable \texttt{diagonal} in line~24, and finds it in the environment where the function itself was defined, which is the body of function \texttt{mymclust()}. Function \texttt{logLik()} uses the list \texttt{para} in lines~8 and 9, and uses the one found in the body of \texttt{defineComponent()}. Function \texttt{flexmix()} on the other hand never sees the model parameters, all it uses are function calls of form \texttt{fit(x,y,w)} or \texttt{logLik(x,y)}, which are exactly the same for all kinds of mixture models. In fact, it would not be necessary to even store the component parameters in the \texttt{"FLXcomponent"} object, they are there only for convenience such that users can easily extract and use them after \texttt{flexmix()} has finished. Lexical scope allows to write clean interfaces in a very elegant way, the driver abstracts all model details from the FlexMix main engine. \subsection{Example: Using the driver} \label{sec:example:-model-based} \SweaveOpts{width=12,height=6,eps=FALSE} <>= library("flexmix") set.seed(1504) options(width=60) grDevices::ps.options(family="Times") suppressMessages(require("ellipse")) suppressMessages(require("mvtnorm")) source("mymclust.R") @ As a simple example we use the four 2-dimensional Gaussian clusters from data set \texttt{Nclus}. Fitting a wrong model with diagonal covariance matrix is done by <<>>= data("Nclus") m1 <- flexmix(Nclus ~ 1, k = 4, model = mymclust()) summary(m1) @ The result can be seen in the left panel of Figure~\ref{fig:ell}, the result is ``wrong'' because we forced the ellipses to be parallel to the axes. The overlap between three of the four clusters can also be inferred from the low ratio statistics in the summary table (around 0.5 for components 1, 3 and 4), while the much better separated upper left cluster has a much higher ratio of 0.85. Using the correct model with a full covariance matrix can be done by setting \texttt{diagonal=FALSE} in the call to our driver \texttt{mymclust()}: <<>>= m2 <- flexmix(Nclus ~ 1, k = 4, model = mymclust(diagonal = FALSE)) summary(m2) @ \begin{figure}[htbp] \centering <>= par(mfrow=1:2) plotEll(m1, Nclus) plotEll(m2, Nclus) @ \caption{Fitting a mixture model with diagonal covariance matrix (left) and full covariance matrix (right).} \label{fig:ell} \end{figure} \pagebreak[4] \section{Summary and outlook} \label{sec:summary} The primary goal of FlexMix is extensibility, this makes the package ideal for rapid development of new mixture models. There is no intent to replace packages implementing more specialized mixture models like \pkg{mclust} for mixtures of Gaussians, FlexMix should rather be seen as a complement to those. By interfacing R's facilities for generalized linear models, FlexMix allows the user to estimate complex latent class regression models. Using lexical scope to resolve model-specific parameters hides all model details from the programming interface, FlexMix can in principle fit almost arbitrary finite mixture models for which the EM algorithm is applicable. The downside of this is that FlexMix can in principle fit almost arbitrary finite mixture models, even models where no proper theoretical results for model identification etc.\ are available. We are currently working on a toolset for diagnostic checks on mixture models to test necessary identifiability conditions for those cases where results are available. We also want to implement newer variations of the classic EM algorithm, especially for faster convergence. Another plan is to have an interactive version of the rootograms using \texttt{iPlots} \citep{flexmix:Urbanek+Theus:2003} such that the user can explore the relations between mixture components, possibly linked to background variables. Other planned extensions include covariates for the prior probabilities and to allow to mix different distributions for components, e.g., to include a Poisson point process for background noise. \section*{Computational details} <>= SI <- sessionInfo() pkgs <- paste(sapply(c(SI$otherPkgs, SI$loadedOnly), function(x) paste("\\\\pkg{", x$Package, "} ", x$Version, sep = "")), collapse = ", ") @ All computations and graphics in this paper have been done using \proglang{R} version \Sexpr{getRversion()} with the packages \Sexpr{pkgs}. \section*{Acknowledgments} This research was supported by the Austrian Science Foundation (FWF) under grant SFB\#010 (`Adaptive Information Systems and Modeling in Economics and Management Science'). Bettina Gr\"un has modified the original version to include and reflect the changes of the package. \bibliography{flexmix} \end{document} %%% Local Variables: %%% mode: latex %%% TeX-master: t %%% End: flexmix/inst/doc/myConcomitant.R0000644000176200001440000000130313425024236016415 0ustar liggesusersmyConcomitant <- function(formula = ~ 1) { z <- new("FLXP", name = "myConcomitant", formula = formula) z@fit <- function(x, y, w, ...) { if (missing(w) || is.null(w)) w <- rep(1, length(x)) f <- as.integer(factor(apply(x, 1, paste, collapse = ""))) AVG <- apply(w*y, 2, tapply, f, mean) (AVG/rowSums(AVG))[f,,drop=FALSE] } z@refit <- function(x, y, w, ...) { if (missing(w) || is.null(w)) w <- rep(1, length(x)) f <- as.integer(factor(apply(x, 1, paste, collapse = ""))) AVG <- apply(w*y, 2, tapply, f, mean) (AVG/rowSums(AVG)) } z } flexmix/inst/doc/bootstrapping.R0000644000176200001440000001562013432516252016475 0ustar liggesusers### R code from vignette source 'bootstrapping.Rnw' ################################################### ### code chunk number 1: bootstrapping.Rnw:11-34 ################################################### options(useFancyQuotes = FALSE) digits <- 3 Colors <- c("#E495A5", "#39BEB1") critical_values <- function(n, p = "0.95") { data("qDiptab", package = "diptest") if (n %in% rownames(qDiptab)) { return(qDiptab[as.character(n), p]) }else { n.approx <- as.numeric(rownames(qDiptab)[which.min(abs(n - as.numeric(rownames(qDiptab))))]) return(sqrt(n.approx)/sqrt(n) * qDiptab[as.character(n.approx), p]) } } library("graphics") library("flexmix") combine <- function(x, sep, width) { cs <- cumsum(nchar(x)) remaining <- if (any(cs[-1] > width)) combine(x[c(FALSE, cs[-1] > width)], sep, width) c(paste(x[c(TRUE, cs[-1] <= width)], collapse= sep), remaining) } prettyPrint <- function(x, sep = " ", linebreak = "\n\t", width = getOption("width")) { x <- strsplit(x, sep)[[1]] paste(combine(x, sep, width), collapse = paste(sep, linebreak, collapse = "")) } ################################################### ### code chunk number 2: bootstrapping.Rnw:95-100 ################################################### cat(prettyPrint(gsub("boot_flexmix", "boot", prompt(flexmix:::boot_flexmix, filename = NA)$usage[[2]]), sep = ", ", linebreak = paste("\n", paste(rep(" ", 2), collapse = ""), sep= ""), width = 70)) ################################################### ### code chunk number 3: bootstrapping.Rnw:195-200 ################################################### library("flexmix") Component_1 <- list(Model_1 = list(coef = c(1, -2), sigma = sqrt(0.1))) Component_2 <- list(Model_1 = list(coef = c(2, 2), sigma = sqrt(0.1))) ArtEx.mix <- FLXdist(y ~ x, k = rep(0.5, 2), components = list(Component_1, Component_2)) ################################################### ### code chunk number 4: bootstrapping.Rnw:211-216 ################################################### ArtEx.data <- data.frame(x = rep(0:1, each = 100/2)) set.seed(123) ArtEx.sim <- rflexmix(ArtEx.mix, newdata = ArtEx.data) ArtEx.data$y <- ArtEx.sim$y[[1]] ArtEx.data$class <- ArtEx.sim$class ################################################### ### code chunk number 5: bootstrapping.Rnw:225-230 ################################################### par(mar = c(5, 4, 2, 0) + 0.1) plot(y ~ x, data = ArtEx.data, pch = with(ArtEx.data, 2*class + x)) pars <- list(matrix(c(1, -2, 2, 2), ncol = 2), matrix(c(1, 3, 2, -3), ncol = 2)) for (i in 1:2) apply(pars[[i]], 2, abline, col = Colors[i]) ################################################### ### code chunk number 6: bootstrapping.Rnw:238-241 ################################################### set.seed(123) ArtEx.fit <- stepFlexmix(y ~ x, data = ArtEx.data, k = 2, nrep = 5, control = list(iter = 1000, tol = 1e-8, verbose = 0)) ################################################### ### code chunk number 7: bootstrapping.Rnw:246-248 ################################################### summary(ArtEx.fit) parameters(ArtEx.fit) ################################################### ### code chunk number 8: bootstrapping.Rnw:256-258 ################################################### ArtEx.refit <- refit(ArtEx.fit) summary(ArtEx.refit) ################################################### ### code chunk number 9: bootstrapping.Rnw:280-284 ################################################### set.seed(123) ArtEx.bs <- boot(ArtEx.fit, R = 15, sim = "parametric") if ("boot-output.rda" %in% list.files()) load("boot-output.rda") ArtEx.bs ################################################### ### code chunk number 10: bootstrapping.Rnw:300-301 ################################################### print(plot(ArtEx.bs, ordering = "coef.x", col = Colors)) ################################################### ### code chunk number 11: bootstrapping.Rnw:318-331 ################################################### require("diptest") parameters <- parameters(ArtEx.bs) Ordering <- factor(as.vector(apply(matrix(parameters[,"coef.x"], nrow = 2), 2, order))) Comp1 <- parameters[Ordering == 1,] Comp2 <- parameters[Ordering == 2,] dip.values.art <- matrix(nrow = ncol(parameters), ncol = 3, dimnames=list(colnames(parameters), c("Aggregated", "Comp 1", "Comp 2"))) dip.values.art[,"Aggregated"] <- apply(parameters, 2, dip) dip.values.art[,"Comp 1"] <- apply(Comp1, 2, dip) dip.values.art[,"Comp 2"] <- apply(Comp2, 2, dip) dip.values.art ################################################### ### code chunk number 12: bootstrapping.Rnw:373-379 ################################################### data("seizure", package = "flexmix") model <- FLXMRglm(family = "poisson", offset = log(seizure$Hours)) control <- list(iter = 1000, tol = 1e-10, verbose = 0) set.seed(123) seizMix <- stepFlexmix(Seizures ~ Treatment * log(Day), data = seizure, k = 2, nrep = 5, model = model, control = control) ################################################### ### code chunk number 13: bootstrapping.Rnw:387-392 ################################################### par(mar = c(5, 4, 2, 0) + 0.1) plot(Seizures/Hours~Day, data=seizure, pch = as.integer(seizure$Treatment)) abline(v = 27.5, lty = 2, col = "grey") matplot(seizure$Day, fitted(seizMix)/seizure$Hours, type="l", add = TRUE, col = 1, lty = 1, lwd = 2) ################################################### ### code chunk number 14: bootstrapping.Rnw:414-418 ################################################### set.seed(123) seizMix.bs <- boot(seizMix, R = 15, sim = "parametric") if ("boot-output.rda" %in% list.files()) load("boot-output.rda") print(plot(seizMix.bs, ordering = "coef.(Intercept)", col = Colors)) ################################################### ### code chunk number 15: bootstrapping.Rnw:425-430 ################################################### parameters <- parameters(seizMix.bs) Ordering <- factor(as.vector(apply(matrix(parameters[,"coef.(Intercept)"], nrow = 2), 2, order))) Comp1 <- parameters[Ordering == 1,] Comp2 <- parameters[Ordering == 2,] ################################################### ### code chunk number 16: bootstrapping.Rnw:439-446 ################################################### dip.values.art <- matrix(nrow = ncol(parameters), ncol = 3, dimnames = list(colnames(parameters), c("Aggregated", "Comp 1", "Comp 2"))) dip.values.art[,"Aggregated"] <- apply(parameters, 2, dip) dip.values.art[,"Comp 1"] <- apply(Comp1, 2, dip) dip.values.art[,"Comp 2"] <- apply(Comp2, 2, dip) dip.values.art ################################################### ### code chunk number 17: bootstrapping.Rnw:461-466 (eval = FALSE) ################################################### ## set.seed(123) ## ArtEx.bs <- boot(ArtEx.fit, R = 200, sim = "parametric") ## set.seed(123) ## seizMix.bs <- boot(seizMix, R = 200, sim = "parametric") ## save(ArtEx.bs, seizMix.bs, file = "boot-output.rda") flexmix/inst/doc/bootstrapping.Rnw0000644000176200001440000004740313425024236017044 0ustar liggesusers\documentclass[nojss]{jss} \usepackage{amsfonts,bm,amsmath,amssymb} %%\usepackage{Sweave} %% already provided by jss.cls %%%\VignetteIndexEntry{Finite Mixture Model Diagnostics Using Resampling Methods} %%\VignetteDepends{flexmix} %%\VignetteKeywords{R, finite mixture model, resampling, bootstrap} %%\VignettePackage{flexmix} \title{Finite Mixture Model Diagnostics Using Resampling Methods} <>= options(useFancyQuotes = FALSE) digits <- 3 Colors <- c("#E495A5", "#39BEB1") critical_values <- function(n, p = "0.95") { data("qDiptab", package = "diptest") if (n %in% rownames(qDiptab)) { return(qDiptab[as.character(n), p]) }else { n.approx <- as.numeric(rownames(qDiptab)[which.min(abs(n - as.numeric(rownames(qDiptab))))]) return(sqrt(n.approx)/sqrt(n) * qDiptab[as.character(n.approx), p]) } } library("graphics") library("flexmix") combine <- function(x, sep, width) { cs <- cumsum(nchar(x)) remaining <- if (any(cs[-1] > width)) combine(x[c(FALSE, cs[-1] > width)], sep, width) c(paste(x[c(TRUE, cs[-1] <= width)], collapse= sep), remaining) } prettyPrint <- function(x, sep = " ", linebreak = "\n\t", width = getOption("width")) { x <- strsplit(x, sep)[[1]] paste(combine(x, sep, width), collapse = paste(sep, linebreak, collapse = "")) } @ \author{Bettina Gr{\"u}n\\ Johannes Kepler Universit{\"a}t Linz \And Friedrich Leisch\\ Universit\"at f\"ur Bodenkultur Wien} \Plainauthor{Bettina Gr{\"u}n, Friedrich Leisch} \Address{ Bettina Gr\"un\\ Institut f\"ur Angewandte Statistik\\ Johannes Kepler Universit{\"a}t Linz\\ Altenbergerstra\ss{}e 69\\ 4040 Linz, Austria\\ E-mail: \email{Bettina.Gruen@jku.at} Friedrich Leisch\\ Institut f\"ur Angewandte Statistik und EDV\\ Universit\"at f\"ur Bodenkultur Wien\\ Peter Jordan Stra\ss{}e 82\\ 1190 Wien, Austria\\ E-mail: \email{Friedrich.Leisch@boku.ac.at}\\ URL: \url{http://www.statistik.lmu.de/~leisch/} } \Abstract{ This paper illustrates the implementation of resampling methods in \pkg{flexmix} as well as the application of resampling methods for model diagnostics of fitted finite mixture models. Convenience functions to perform these methods are available in package \pkg{flexmix}. The use of the methods is illustrated with an artificial example and the \code{seizure} data set. } \Keywords{\proglang{R}, finite mixture models, resampling, bootstrap} \Plainkeywords{R, finite mixture models, resampling, bootstrap} %%------------------------------------------------------------------------- %%------------------------------------------------------------------------- \begin{document} \SweaveOpts{engine=R, echo=true, height=5, width=8, eps=FALSE, keep.source=TRUE} \setkeys{Gin}{width=0.95\textwidth} \section{Implementation of resampling methods}\label{sec:implementation} The proposed framework for model diagnostics using resampling \citep{mixtures:gruen+leisch:2004} equally allows to investigate model fit for all kinds of mixture models. The procedure is applicable to mixture models with different component specific models and does not impose any limitation such as for example on the dimension of the parameter space of the component specific model. In addition to the fitting step different component specific models only require different random number generators for the parametric bootstrap. The \code{boot()} function in \pkg{flexmix} is a generic \proglang{S4} function with a method for fitted finite mixtures of class \code{"flexmix"} and is applicable to general finite mixture models. The function with arguments and their defaults is given by: <>= cat(prettyPrint(gsub("boot_flexmix", "boot", prompt(flexmix:::boot_flexmix, filename = NA)$usage[[2]]), sep = ", ", linebreak = paste("\n", paste(rep(" ", 2), collapse = ""), sep= ""), width = 70)) @ The interface is similar to the \code{boot()} function in package \pkg{boot} \citep{mixtures:Davison+Hinkley:1997, mixtures:Canty+Ripley:2010}. The \code{object} is a fitted finite mixture of class \code{"flexmix"} and \code{R} denotes the number of resamples. The possible bootstrapping method are \code{"empirical"} (also available as \code{"ordinary"}) and \code{"parametric"}. For the parametric bootstrap sampling from the fitted mixture is performed using \code{rflexmix()}. For mixture models with different component specific models \code{rflexmix()} requires a sampling method for the component specific model. Argument \code{initialize\_solution} allows to select if the EM algorithm is started in the original finite mixture solution or if random initialization is performed. The fitted mixture model might contain weights and group indicators. The weights are case weights and allow to reduce the amount of data if observations are identical. This is useful for example for latent class analysis of multivariate binary data. The argument \code{keep\_weights} allows to indicate if they should be kept for the bootstrapping. Group indicators allow to specify that the component membership is identical over several observations, e.g., for repeated measurements of the same individual. Argument \code{keep\_groups} allows to indicate if the grouping information should also be used in the bootstrapping. \code{verbose} indicates if information on the progress should be printed. The \code{control} argument allows to control the EM algorithm for fitting the model to each of the bootstrap samples. By default the \code{control} argument is extracted from the fitted model provided by \code{object}. \code{k} allows to specify the number of components and by default this is also taken from the fitted model provided. The \code{model} argument determines if also the model and the weights slot for each sample are stored and returned. The returned object is of class \code{"FLXboot"} and otherwise only contains the fitted parameters, the fitted priors, the log likelihoods, the number of components of the fitted mixtures and the information if the EM algorithm has converged. The likelihood ratio test is implemented based on \code{boot()} in function \code{LR\_test()} and returns an object of class \code{"htest"} containing the number of valid bootstrap replicates, the p-value, the double negative log likelihood ratio test statistics for the original data and the bootstrap replicates. The \code{plot} method for \code{"FLXboot"} objects returns a parallel coordinate plot with the fitted parameters separately for each of the components. \section{Artificial data set} In the following a finite mixture model is used as the underlying data generating process which is theoretically not identifiable. We are assuming a finite mixture of linear regression models with two components of equal size where the coverage condition is not fulfilled \citep{mixtures:Hennig:2000}. Hence, intra-component label switching is possible, i.e., there exist two parameterizations implying the same mixture distribution which differ how the components between the covariate points are combined. We assume that one measurement per object and a single categorical regressor with two levels are given. The usual design matrix for a model with intercept uses the two covariate points $\mathbf{x}_1 = (1, 0)'$ and $\mathbf{x}_2 = (1, 1)'$. The mixture distribution is given by \begin{eqnarray*} H(y|\mathbf{x}, \Theta) &=& \frac{1}{2} N(\mu_1, 0.1) + \frac{1}{2} N(\mu_2, 0.1), \end{eqnarray*} where $\mu_k(\mathbf{x}) = \mathbf{x}'\bm{\alpha}_k$ and $N(\mu, \sigma^2)$ is the normal distribution. Now let $\mu_1(\mathbf{x}_1) = 1$, $\mu_2(\mathbf{x}_1) = 2$, $\mu_1(\mathbf{x}_2) = -1$ and $\mu_2(\mathbf{x}_2) = 4$. As Gaussian mixture distributions are generically identifiable the means, variances and component weights are uniquely determined in each covariate point given the mixture distribution. However, as the coverage condition is not fulfilled, the two possible solutions for $\bm{\alpha}$ are: \begin{description} \item[Solution 1:] $\bm{\alpha}_1^{(1)} = (2,\phantom{-}2)'$, $\bm{\alpha}_2^{(1)} = (1,-2)'$, \item[Solution 2:] $\bm{\alpha}_1^{(2)} = (2,-3)'$, $\bm{\alpha}_2^{(2)} = (1,\phantom{-}3)'$. \end{description} We specify this artificial mixture distribution using \code{FLXdist()}. \code{FLXdist()} returns an unfitted finite mixture of class \code{"FLXdist"}. The class of fitted finite mixture models \code{"flexmix"} extends class \code{"FLXdist"}. Each component follows a normal distribution. The parameters specified in a named list therefore consist of the regression coefficients and the standard deviation. Function \code{FLXdist()} has an argument \code{formula} for specifying the regression in each of the components, an argument \code{k} for the component weights and \code{components} for the parameters of each of the components. <<>>= library("flexmix") Component_1 <- list(Model_1 = list(coef = c(1, -2), sigma = sqrt(0.1))) Component_2 <- list(Model_1 = list(coef = c(2, 2), sigma = sqrt(0.1))) ArtEx.mix <- FLXdist(y ~ x, k = rep(0.5, 2), components = list(Component_1, Component_2)) @ We draw a balanced sample with 50 observations in each covariate point from the mixture model using \code{rflexmix()} after defining the data points for the covariates. \code{rflexmix()} can either have an unfitted or a fitted finite mixture as input. For unfitted mixtures data has to be provided using the \code{newdata} argument. For already fitted mixtures data can be optionally provided, otherwise the data used for fitting the mixture is used. <<>>= ArtEx.data <- data.frame(x = rep(0:1, each = 100/2)) set.seed(123) ArtEx.sim <- rflexmix(ArtEx.mix, newdata = ArtEx.data) ArtEx.data$y <- ArtEx.sim$y[[1]] ArtEx.data$class <- ArtEx.sim$class @ In Figure~\ref{fig:art} the sample is plotted together with the two solutions for combining $x_1$ and $x_2$, i.e., this illustrates intra-component label switching. \begin{figure} \centering <>= par(mar = c(5, 4, 2, 0) + 0.1) plot(y ~ x, data = ArtEx.data, pch = with(ArtEx.data, 2*class + x)) pars <- list(matrix(c(1, -2, 2, 2), ncol = 2), matrix(c(1, 3, 2, -3), ncol = 2)) for (i in 1:2) apply(pars[[i]], 2, abline, col = Colors[i]) @ \caption{Balanced sample from the artificial example with the two theoretical solutions.} \label{fig:art} \end{figure} We fit a finite mixture to the sample using \code{stepFlexmix()}. <<>>= set.seed(123) ArtEx.fit <- stepFlexmix(y ~ x, data = ArtEx.data, k = 2, nrep = 5, control = list(iter = 1000, tol = 1e-8, verbose = 0)) @ The fitted mixture can be inspected using \code{summary()} and \code{parameters()}. <<>>= summary(ArtEx.fit) parameters(ArtEx.fit) @ Obviously the fitted mixture parameters correspond to the parameterization we used to specify the mixture distribution. Using standard asymptotic theory to analyze the fitted mixture model gives the following estimates for the standard deviations. <<>>= ArtEx.refit <- refit(ArtEx.fit) summary(ArtEx.refit) @ The fitted mixture can also be analyzed using resampling techniques. For analyzing the stability of the parameter estimates where the possibility of identifiability problems is also taken into account the parametric bootstrap is used with random initialization. Function \code{boot()} can be used for empirical or parametric bootstrap (specified by the argument \code{sim}). The logical argument \code{initialize_solution} specifies if the initialization is in the original solution or random. By default random initialization is made. The number of bootstrap samples is set by the argument \code{R}. Please note that the arguments are chosen to correspond to those for function \code{boot} in package \pkg{boot} \citep{mixtures:Davison+Hinkley:1997}. Only a few number of bootstrap samples are drawn to keep the amount of time needed to run the vignette within reasonable limits. However, for a sensible application of the bootstrap methods at least \code{R} equal to 100 should be used. If the output for this setting has been saved, it is loaded and used in the further analysis. Please see the appendix for the code for generating the saved \proglang{R} output. <<>>= set.seed(123) ArtEx.bs <- boot(ArtEx.fit, R = 15, sim = "parametric") if ("boot-output.rda" %in% list.files()) load("boot-output.rda") ArtEx.bs @ Function \code{boot()} returns an object of class \code{"\Sexpr{class(ArtEx.bs)}"}. The default plot compares the bootstrap parameter estimates to the confidence intervals derived using standard asymptotic theory in a parallel coordinate plot (see Figure~\ref{fig:plot.FLXboot-art}). Clearly two groups of parameter estimates can be distinguished which are about of equal size. One subset of the parameter estimates stays within the confidence intervals induced by standard asymptotic theory, while the second group corresponds to the second solution and clusters around these parameter values. \begin{figure}[h!] \centering <>= print(plot(ArtEx.bs, ordering = "coef.x", col = Colors)) @ \caption{Diagnostic plot of the bootstrap results for the artificial example.} \label{fig:plot.FLXboot-art} \end{figure} In the following the DIP-test is applied to check if the parameter estimates follow a unimodal distribution. This is done for the aggregated parameter esimates where unimodality implies that this parameter is not suitable for imposing an ordering constraint which induces a unique labelling. For the separate component analysis which is made after imposing an ordering constraint on the coefficient of $x$ rejection the null hypothesis of unimodality implies that identifiability problems are present, e.g.~due to intra-component label switching. <<>>= require("diptest") parameters <- parameters(ArtEx.bs) Ordering <- factor(as.vector(apply(matrix(parameters[,"coef.x"], nrow = 2), 2, order))) Comp1 <- parameters[Ordering == 1,] Comp2 <- parameters[Ordering == 2,] dip.values.art <- matrix(nrow = ncol(parameters), ncol = 3, dimnames=list(colnames(parameters), c("Aggregated", "Comp 1", "Comp 2"))) dip.values.art[,"Aggregated"] <- apply(parameters, 2, dip) dip.values.art[,"Comp 1"] <- apply(Comp1, 2, dip) dip.values.art[,"Comp 2"] <- apply(Comp2, 2, dip) dip.values.art @ The critical value for column \code{Aggregated} is \Sexpr{round(critical_values(nrow(parameters)), digits = digits)} and for the columns of the separate components \Sexpr{round(critical_values(nrow(Comp1)), digits = digits)}. The component sizes as well as the standard deviations follow a unimodal distribution for the aggregated data as well as for each of the components. The regression coefficients are multimodal for the aggregate data as well as for each of the components. While from the aggregated case it might be concluded that imposing an ordering constraint on the intercept or the coefficient of $x$ is suitable, the component-specific analyses reveal that a unique labelling was not achieved. \section{Seizure} In \cite{mixtures:Wang+Puterman+Cockburn:1996} a Poisson mixture regression is fitted to data from a clinical trial where the effect of intravenous gammaglobulin on suppression of epileptic seizures is investigated. The data used were 140 observations from one treated patient, where treatment started on the $28^\textrm{th}$ day. In the regression model three independent variables were included: treatment, trend and interaction treatment-trend. Treatment is a dummy variable indicating if the treatment period has already started. Furthermore, the number of parental observation hours per day were available and it is assumed that the number of epileptic seizures per observation hour follows a Poisson mixture distribution. The number of epileptic seizures per parental observation hour for each day are plotted in Figure~\ref{fig:seizure}. The fitted mixture distribution consists of two components which can be interpreted as representing 'good' and 'bad' days of the patients. The mixture model can be formulated by \begin{equation*} H(y|\mathbf{x}, \Theta) = \pi_1 P(\lambda_1) + \pi_2 P(\lambda_2), \end{equation*} where $\lambda_k = e^{\mathbf{x}'\bm{\alpha}_k}$ for $k = 1,2$ and $P(\lambda)$ is the Poisson distribution. The data is loaded and the mixture fitted with two components. <<>>= data("seizure", package = "flexmix") model <- FLXMRglm(family = "poisson", offset = log(seizure$Hours)) control <- list(iter = 1000, tol = 1e-10, verbose = 0) set.seed(123) seizMix <- stepFlexmix(Seizures ~ Treatment * log(Day), data = seizure, k = 2, nrep = 5, model = model, control = control) @ The fitted regression lines for each of the two components are shown in Figure~\ref{fig:seizure}. \begin{figure}[h!] \begin{center} <>= par(mar = c(5, 4, 2, 0) + 0.1) plot(Seizures/Hours~Day, data=seizure, pch = as.integer(seizure$Treatment)) abline(v = 27.5, lty = 2, col = "grey") matplot(seizure$Day, fitted(seizMix)/seizure$Hours, type="l", add = TRUE, col = 1, lty = 1, lwd = 2) @ \caption{Seizure data with the fitted values for the \citeauthor{mixtures:Wang+Puterman+Cockburn:1996} model. The plotting character for the observed values in the base period is a circle and for those in the treatment period a triangle.} \label{fig:seizure} \end{center} \end{figure} The parameteric bootstrap with random initialization is used to investigate identifiability problems and parameter stability. The diagnostic plot is given in Figure~\ref{fig:plot.FLXboot-seiz}. The coloring is according to an ordering constraint on the intercept. Clearly the parameter estimates corresponding to the solution where the bad days from the base period are combined with the good days from the treatement period and vice versa for the good days of the base period can be distinguished and indicate the slight identifiability problems of the fitted mixture. \begin{figure}[h!] \centering <>= set.seed(123) seizMix.bs <- boot(seizMix, R = 15, sim = "parametric") if ("boot-output.rda" %in% list.files()) load("boot-output.rda") print(plot(seizMix.bs, ordering = "coef.(Intercept)", col = Colors)) @ \label{fig:plot.FLXboot-seiz} \caption{Diagnostic plot of the bootstrap results for the \code{seizure} data.} \end{figure} <<>>= parameters <- parameters(seizMix.bs) Ordering <- factor(as.vector(apply(matrix(parameters[,"coef.(Intercept)"], nrow = 2), 2, order))) Comp1 <- parameters[Ordering == 1,] Comp2 <- parameters[Ordering == 2,] @ For applying the DIP test also an ordering constraint on the intercept is used. The critical value for column \code{Aggregated} is \Sexpr{round(critical_values(nrow(parameters)), digits = digits)} and for the columns of the separate components \Sexpr{round(critical_values(nrow(Comp1)), digits = digits)}. <<>>= dip.values.art <- matrix(nrow = ncol(parameters), ncol = 3, dimnames = list(colnames(parameters), c("Aggregated", "Comp 1", "Comp 2"))) dip.values.art[,"Aggregated"] <- apply(parameters, 2, dip) dip.values.art[,"Comp 1"] <- apply(Comp1, 2, dip) dip.values.art[,"Comp 2"] <- apply(Comp2, 2, dip) dip.values.art @ For the aggregate results the hypothesis of unimodality cannot be rejected for the trend. For the component-specific analyses unimodality cannot be rejected only for the intercept (where the ordering condition was imposed on) and again the trend. For all other parameter estimates unimodality is rejected which indicates that the ordering constraint was able to impose a unique labelling only for the own parameter and not for the other parameters. This suggests identifiability problems. %%------------------------------------------------------------------------- %%------------------------------------------------------------------------- \appendix \section[Generation of saved R output]{Generation of saved \proglang{R} output} <>= set.seed(123) ArtEx.bs <- boot(ArtEx.fit, R = 200, sim = "parametric") set.seed(123) seizMix.bs <- boot(seizMix, R = 200, sim = "parametric") save(ArtEx.bs, seizMix.bs, file = "boot-output.rda") @ \bibliography{mixture} \end{document} flexmix/inst/doc/ziglm.R0000644000176200001440000000220513425024236014715 0ustar liggesuserssetClass("FLXMRziglm", contains = "FLXMRglm") FLXMRziglm <- function(formula = . ~ ., family = c("binomial", "poisson"), ...) { family <- match.arg(family) new("FLXMRziglm", FLXMRglm(formula, family, ...), name = paste("FLXMRziglm", family, sep=":")) } setMethod("FLXgetModelmatrix", signature(model="FLXMRziglm"), function(model, data, formula, lhs=TRUE, ...) { model <- callNextMethod(model, data, formula, lhs) if (attr(terms(model@fullformula), "intercept") == 0) stop("please include an intercept") model }) setMethod("FLXremoveComponent", signature(model = "FLXMRziglm"), function(model, nok, ...) { if (1 %in% nok) as(model, "FLXMRglm") else model }) setMethod("FLXmstep", signature(model = "FLXMRziglm"), function(model, weights, components, ...) { coef <- c(-Inf, rep(0, ncol(model@x)-1)) names(coef) <- colnames(model@x) comp.1 <- with(list(coef = coef, df = 0, offset = NULL, family = model@family), eval(model@defineComponent)) c(list(comp.1), FLXmstep(as(model, "FLXMRglm"), weights[, -1, drop=FALSE], components[-1])) }) flexmix/inst/doc/regression-examples.R0000644000176200001440000004061513432516316017601 0ustar liggesusers### R code from vignette source 'regression-examples.Rnw' ################################################### ### code chunk number 1: regression-examples.Rnw:11-14 ################################################### library("stats") library("graphics") library("flexmix") ################################################### ### code chunk number 2: start ################################################### options(width=70, prompt = "R> ", continue = "+ ", useFancyQuotes = FALSE) set.seed(1802) library("lattice") ltheme <- canonical.theme("postscript", FALSE) lattice.options(default.theme=ltheme) ################################################### ### code chunk number 3: NregFix ################################################### set.seed(2807) library("flexmix") data("NregFix", package = "flexmix") Model <- FLXMRglm(~ x2 + x1) fittedModel <- stepFlexmix(y ~ 1, model = Model, nrep = 3, k = 3, data = NregFix, concomitant = FLXPmultinom(~ w)) fittedModel <- relabel(fittedModel, "model", "x1") summary(refit(fittedModel)) ################################################### ### code chunk number 4: diffModel ################################################### Model2 <- FLXMRglmfix(fixed = ~ x2, nested = list(k = c(1, 2), formula = c(~ 0, ~ x1)), varFix = TRUE) fittedModel2 <- flexmix(y ~ 1, model = Model2, cluster = posterior(fittedModel), data = NregFix, concomitant = FLXPmultinom(~ w)) BIC(fittedModel) BIC(fittedModel2) ################################################### ### code chunk number 5: artificial-example ################################################### plotNregFix <- NregFix plotNregFix$w <- factor(NregFix$w, levels = 0:1, labels = paste("w =", 0:1)) plotNregFix$x2 <- factor(NregFix$x2, levels = 0:1, labels = paste("x2 =", 0:1)) plotNregFix$class <- factor(NregFix$class, levels = 1:3, labels = paste("Class", 1:3)) print(xyplot(y ~ x1 | x2*w, groups = class, data = plotNregFix, cex = 0.6, auto.key = list(space="right"), layout = c(2,2))) ################################################### ### code chunk number 6: refit ################################################### summary(refit(fittedModel2)) ################################################### ### code chunk number 7: beta-glm ################################################### data("betablocker", package = "flexmix") betaGlm <- glm(cbind(Deaths, Total - Deaths) ~ Treatment, family = "binomial", data = betablocker) betaGlm ################################################### ### code chunk number 8: beta-fix ################################################### betaMixFix <- stepFlexmix(cbind(Deaths, Total - Deaths) ~ 1 | Center, model = FLXMRglmfix(family = "binomial", fixed = ~ Treatment), k = 2:4, nrep = 3, data = betablocker) betaMixFix ################################################### ### code chunk number 9: beta-fig ################################################### library("grid") betaMixFix_3 <- getModel(betaMixFix, "3") betaMixFix_3 <- relabel(betaMixFix_3, "model", "Intercept") betablocker$Center <- with(betablocker, factor(Center, levels = Center[order((Deaths/Total)[1:22])])) clusters <- factor(clusters(betaMixFix_3), labels = paste("Cluster", 1:3)) print(dotplot(Deaths/Total ~ Center | clusters, groups = Treatment, as.table = TRUE, data = betablocker, xlab = "Center", layout = c(3, 1), scales = list(x = list(draw = FALSE)), key = simpleKey(levels(betablocker$Treatment), lines = TRUE, corner = c(1,0)))) betaMixFix.fitted <- fitted(betaMixFix_3) for (i in 1:3) { seekViewport(trellis.vpname("panel", i, 1)) grid.lines(unit(1:22, "native"), unit(betaMixFix.fitted[1:22, i], "native"), gp = gpar(lty = 1)) grid.lines(unit(1:22, "native"), unit(betaMixFix.fitted[23:44, i], "native"), gp = gpar(lty = 2)) } ################################################### ### code chunk number 10: beta-full ################################################### betaMix <- stepFlexmix(cbind(Deaths, Total - Deaths) ~ Treatment | Center, model = FLXMRglm(family = "binomial"), k = 3, nrep = 3, data = betablocker) summary(betaMix) ################################################### ### code chunk number 11: default-plot ################################################### print(plot(betaMixFix_3, mark = 1, col = "grey", markcol = 1)) ################################################### ### code chunk number 12: parameters ################################################### parameters(betaMix) ################################################### ### code chunk number 13: cluster ################################################### table(clusters(betaMix)) ################################################### ### code chunk number 14: predict ################################################### predict(betaMix, newdata = data.frame(Treatment = c("Control", "Treated"))) ################################################### ### code chunk number 15: fitted ################################################### fitted(betaMix)[c(1, 23), ] ################################################### ### code chunk number 16: refit ################################################### summary(refit(getModel(betaMixFix, "3"))) ################################################### ### code chunk number 17: mehta-fix ################################################### data("Mehta", package = "flexmix") mehtaMix <- stepFlexmix(cbind(Response, Total - Response)~ 1 | Site, model = FLXMRglmfix(family = "binomial", fixed = ~ Drug), control = list(minprior = 0.04), nrep = 3, k = 3, data = Mehta) summary(mehtaMix) ################################################### ### code chunk number 18: mehta-fig ################################################### Mehta$Site <- with(Mehta, factor(Site, levels = Site[order((Response/Total)[1:22])])) clusters <- factor(clusters(mehtaMix), labels = paste("Cluster", 1:3)) print(dotplot(Response/Total ~ Site | clusters, groups = Drug, layout = c(3,1), data = Mehta, xlab = "Site", scales = list(x = list(draw = FALSE)), key = simpleKey(levels(Mehta$Drug), lines = TRUE, corner = c(1,0)))) mehtaMix.fitted <- fitted(mehtaMix) for (i in 1:3) { seekViewport(trellis.vpname("panel", i, 1)) sapply(1:nlevels(Mehta$Drug), function(j) grid.lines(unit(1:22, "native"), unit(mehtaMix.fitted[Mehta$Drug == levels(Mehta$Drug)[j], i], "native"), gp = gpar(lty = j))) } ################################################### ### code chunk number 19: mehta-full ################################################### mehtaMix <- stepFlexmix(cbind(Response, Total - Response) ~ Drug | Site, model = FLXMRglm(family = "binomial"), k = 3, data = Mehta, nrep = 3, control = list(minprior = 0.04)) summary(mehtaMix) ################################################### ### code chunk number 20: mehta-sub ################################################### Mehta.sub <- subset(Mehta, Site != 15) mehtaMix <- stepFlexmix(cbind(Response, Total - Response) ~ 1 | Site, model = FLXMRglmfix(family = "binomial", fixed = ~ Drug), data = Mehta.sub, k = 2, nrep = 3) summary(mehtaMix) ################################################### ### code chunk number 21: tribolium ################################################### data("tribolium", package = "flexmix") TribMix <- stepFlexmix(cbind(Remaining, Total - Remaining) ~ 1, k = 2:3, model = FLXMRglmfix(fixed = ~ Species, family = "binomial"), concomitant = FLXPmultinom(~ Replicate), data = tribolium) ################################################### ### code chunk number 22: wang-model ################################################### modelWang <- FLXMRglmfix(fixed = ~ I(Species == "Confusum"), family = "binomial") concomitantWang <- FLXPmultinom(~ I(Replicate == 3)) TribMixWang <- stepFlexmix(cbind(Remaining, Total - Remaining) ~ 1, data = tribolium, model = modelWang, concomitant = concomitantWang, k = 2) summary(refit(TribMixWang)) ################################################### ### code chunk number 23: tribolium ################################################### clusters <- factor(clusters(TribMixWang), labels = paste("Cluster", 1:TribMixWang@k)) print(dotplot(Remaining/Total ~ factor(Replicate) | clusters, groups = Species, data = tribolium[rep(1:9, each = 3) + c(0:2)*9,], xlab = "Replicate", auto.key = list(corner = c(1,0)))) ################################################### ### code chunk number 24: subset ################################################### TribMixWangSub <- stepFlexmix(cbind(Remaining, Total - Remaining) ~ 1, k = 2, data = tribolium[-7,], model = modelWang, concomitant = concomitantWang) ################################################### ### code chunk number 25: trypanosome ################################################### data("trypanosome", package = "flexmix") TrypMix <- stepFlexmix(cbind(Dead, 1-Dead) ~ 1, k = 2, nrep = 3, data = trypanosome, model = FLXMRglmfix(family = "binomial", fixed = ~ log(Dose))) summary(refit(TrypMix)) ################################################### ### code chunk number 26: trypanosome ################################################### tab <- with(trypanosome, table(Dead, Dose)) Tryp.dat <- data.frame(Dead = tab["1",], Alive = tab["0",], Dose = as.numeric(colnames(tab))) plot(Dead/(Dead+Alive) ~ Dose, data = Tryp.dat) Tryp.pred <- predict(glm(cbind(Dead, 1-Dead) ~ log(Dose), family = "binomial", data = trypanosome), newdata=Tryp.dat, type = "response") TrypMix.pred <- predict(TrypMix, newdata = Tryp.dat, aggregate = TRUE)[[1]] lines(Tryp.dat$Dose, Tryp.pred, lty = 2) lines(Tryp.dat$Dose, TrypMix.pred, lty = 3) legend(4.7, 1, c("GLM", "Mixture model"), lty=c(2, 3), xjust=0, yjust=1) ################################################### ### code chunk number 27: fabric-fix ################################################### data("fabricfault", package = "flexmix") fabricMix <- stepFlexmix(Faults ~ 1, model = FLXMRglmfix(family="poisson", fixed = ~ log(Length)), data = fabricfault, k = 2, nrep = 3) summary(fabricMix) summary(refit(fabricMix)) Lnew <- seq(0, 1000, by = 50) fabricMix.pred <- predict(fabricMix, newdata = data.frame(Length = Lnew)) ################################################### ### code chunk number 28: fabric-fix-nested ################################################### fabricMix2 <- flexmix(Faults ~ 0, data = fabricfault, cluster = posterior(fabricMix), model = FLXMRglmfix(family = "poisson", fixed = ~ log(Length), nested = list(k=c(1,1), formula=list(~0,~1)))) summary(refit(fabricMix2)) fabricMix2.pred <- predict(fabricMix2, newdata = data.frame(Length = Lnew)) ################################################### ### code chunk number 29: fabric-fig ################################################### plot(Faults ~ Length, data = fabricfault) sapply(fabricMix.pred, function(y) lines(Lnew, y, lty = 1)) sapply(fabricMix2.pred, function(y) lines(Lnew, y, lty = 2)) legend(190, 25, paste("Model", 1:2), lty=c(1, 2), xjust=0, yjust=1) ################################################### ### code chunk number 30: patent ################################################### data("patent", package = "flexmix") ModelPat <- FLXMRglm(family = "poisson") FittedPat <- stepFlexmix(Patents ~ lgRD, k = 3, nrep = 3, model = ModelPat, data = patent, concomitant = FLXPmultinom(~ RDS)) summary(FittedPat) ################################################### ### code chunk number 31: patent-fixed ################################################### ModelFixed <- FLXMRglmfix(family = "poisson", fixed = ~ lgRD) FittedPatFixed <- flexmix(Patents ~ 1, model = ModelFixed, cluster = posterior(FittedPat), concomitant = FLXPmultinom(~ RDS), data = patent) summary(FittedPatFixed) ################################################### ### code chunk number 32: Poisson ################################################### lgRDv <- seq(-3, 5, by = 0.05) newdata <- data.frame(lgRD = lgRDv) plotData <- function(fitted) { with(patent, data.frame(Patents = c(Patents, unlist(predict(fitted, newdata = newdata))), lgRD = c(lgRD, rep(lgRDv, 3)), class = c(clusters(fitted), rep(1:3, each = nrow(newdata))), type = rep(c("data", "fit"), c(nrow(patent), nrow(newdata)*3)))) } plotPatents <- cbind(plotData(FittedPat), which = "Wang et al.") plotPatentsFixed <- cbind(plotData(FittedPatFixed), which = "Fixed effects") plotP <- rbind(plotPatents, plotPatentsFixed) rds <- seq(0, 3, by = 0.02) x <- model.matrix(FittedPat@concomitant@formula, data = data.frame(RDS = rds)) plotConc <- function(fitted) { E <- exp(x%*%fitted@concomitant@coef) data.frame(Probability = as.vector(E/rowSums(E)), class = rep(1:3, each = nrow(x)), RDS = rep(rds, 3)) } plotConc1 <- cbind(plotConc(FittedPat), which = "Wang et al.") plotConc2 <- cbind(plotConc(FittedPatFixed), which = "Fixed effects") plotC <- rbind(plotConc1, plotConc2) print(xyplot(Patents ~ lgRD | which, data = plotP, groups=class, xlab = "log(R&D)", panel = "panel.superpose", type = plotP$type, panel.groups = function(x, y, type = "p", subscripts, ...) { ind <- plotP$type[subscripts] == "data" panel.xyplot(x[ind], y[ind], ...) panel.xyplot(x[!ind], y[!ind], type = "l", ...) }, scales = list(alternating=FALSE), layout=c(1,2), as.table=TRUE), more=TRUE, position=c(0,0,0.6, 1)) print(xyplot(Probability ~ RDS | which, groups = class, data = plotC, type = "l", scales = list(alternating=FALSE), layout=c(1,2), as.table=TRUE), position=c(0.6, 0.01, 1, 0.99)) ################################################### ### code chunk number 33: seizure ################################################### data("seizure", package = "flexmix") seizMix <- stepFlexmix(Seizures ~ Treatment * log(Day), data = seizure, k = 2, nrep = 3, model = FLXMRglm(family = "poisson", offset = log(seizure$Hours))) summary(seizMix) summary(refit(seizMix)) ################################################### ### code chunk number 34: seizure ################################################### seizMix2 <- flexmix(Seizures ~ Treatment * log(Day/27), data = seizure, cluster = posterior(seizMix), model = FLXMRglm(family = "poisson", offset = log(seizure$Hours))) summary(seizMix2) summary(refit(seizMix2)) ################################################### ### code chunk number 35: seizure ################################################### seizMix3 <- flexmix(Seizures ~ log(Day/27)/Treatment, data = seizure, cluster = posterior(seizMix), model = FLXMRglm(family = "poisson", offset = log(seizure$Hours))) summary(seizMix3) summary(refit(seizMix3)) ################################################### ### code chunk number 36: seizure ################################################### plot(Seizures/Hours~Day, pch = c(1,3)[as.integer(Treatment)], data=seizure) abline(v=27.5, lty=2, col="grey") legend(140, 9, c("Baseline", "Treatment"), pch=c(1, 3), xjust=1, yjust=1) matplot(seizure$Day, fitted(seizMix)/seizure$Hours, type="l", add=TRUE, lty = 1, col = 1) matplot(seizure$Day, fitted(seizMix3)/seizure$Hours, type="l", add=TRUE, lty = 3, col = 1) legend(140, 7, paste("Model", c(1,3)), lty=c(1, 3), xjust=1, yjust=1) ################################################### ### code chunk number 37: salmonella ################################################### data("salmonellaTA98", package = "flexmix") salmonMix <- stepFlexmix(y ~ 1, data = salmonellaTA98, k = 2, nrep = 3, model = FLXMRglmfix(family = "poisson", fixed = ~ x + log(x + 10))) ################################################### ### code chunk number 38: salmonella ################################################### salmonMix.pr <- predict(salmonMix, newdata=salmonellaTA98) plot(y~x, data=salmonellaTA98, pch=as.character(clusters(salmonMix)), xlab="Dose of quinoline", ylab="Number of revertant colonies of salmonella", ylim=range(c(salmonellaTA98$y, unlist(salmonMix.pr)))) for (i in 1:2) lines(salmonellaTA98$x, salmonMix.pr[[i]], lty=i) ################################################### ### code chunk number 39: regression-examples.Rnw:923-927 ################################################### SI <- sessionInfo() pkgs <- paste(sapply(c(SI$otherPkgs, SI$loadedOnly), function(x) paste("\\\\pkg{", x$Package, "} ", x$Version, sep = "")), collapse = ", ") flexmix/inst/doc/mixture-regressions.pdf0000644000176200001440000124241713432516321020214 0ustar liggesusers%PDF-1.5 %¿÷¢þ 1 0 obj << /Type /ObjStm /Length 4348 /Filter /FlateDecode /N 73 /First 605 >> stream xœµ[[sÛ¶¶~?¿o;v$q!¸gOgb;vÒØ‰+;Ž“3} %Úæ.-¹"•:ýíçá| $Þd3®öÈ4/……õ­ÀˆL03ɤJ˜bq0Í’H³˜%Ú0ÃÂ@ –°0 c\³PÄ! ñ§"…‡8ÇY(ÁBÉ"à^±(‘xI3!Éðª0• “¡Â}¤ ^b*ˆB†?%饈) ¦Œ Æ˜¤fè ÕQ_3­ :‹Y¢ßÈ€`Ð%ÌD¨$f”Ä9Äd‹ˆ º„`Æ BbhÊK”À=†šàFÐÐÂi¼2ÁÄèRbÈ!Ñ,iØREŒÆj4-1àÐ๔Ğ q!@=h#©“h9Ò•Ñr”ÐëhY„Fcä¸CFËBÅŠ¡ ƒX)b´O¼I€ ´,ÃÀ€?tF(´,#ƒ'hYJP… Ä…Æ¢e©b0-Ë\B/¸Uš&É€–eBEhY…:giRQâ*" ó‘Ó[hY)Ð é•&¶“°<&°ò‚–5øÁÐo¨£@ÿÏ¿ÿÍøIV¥³´J†° ãVU‘ϳRfïOÓÜHwsþí>c|õ‹Å ûùgÛÄ«Uu»X²{YUåó”-ÿoþ;\æÙl™OoÙq–—ÓÛðÞ2K«|1?H«Œ½8øW„)…t™0RBE?á?‚ྵxœžg—ìϼºe·èy¹Ì®Qü.ûöçb9+Ù‹ÉOì:Ÿçhí.¨VKœ³¬(b7Ù<[¦EþW6c4œt¹.š.æÓÅ]^¥óŠ}M—yzUd%š=YÌž"ìt¹˜­¦(;:=fG·‹²*§Ëü¾bÉËH£ÂÙêê?Ù´b/p}žWZ;,²‡“ü]dËcgѿء£ùÄÑ\ºî7Ⱥðd±t>£»oùüÆ^£Vi«œ¦Ëô.«Ðè~ö+2ü]îü/s3jìÀN‰ýѤÄN¡;7lj»KÜôÇq]EÔg]Ÿ]CaÕgUŸÝóØu þÕçú¾¦"uëu+ª.öéº7]÷fj*_^ׯ‰ !ìtŽj*£šÊ¨¦2 Ýýo^€L¯Ù6¯²yUB¹¹×‰Yžî-ÀBêP%ê%EòÒ$Ôx<&&Y¹X-§hZ{ýPU$Ga\7wˆpWW‡ Mϲ móÓƒCP”=Thô矛ĵ§3z˜/ËŠ96ðã´¤/ëšYY­§œ_~þÂLø¶Ãè— TÒ|UDú{ê êÄcÛ¤Fz-°oç•sBêL’»d’<á³X´¦Lí’2¥”)ý÷(Ó»¤ÌD ÊŒxe{'„Åò.- ïOOXˆJïXµ\e¾ö¦7ÿJZf¶CþîøòòÝÅû'{—@? µ5#õTƒÕ‰êþ-4.Ó;M«%¡ÒC±áºNlÕfe»¥ÎCOƒëîS>«nI›ir2p$"¶Ü{îOÄäDnȺuÍ0j½iàË»·…Êbø,Æ*N[G£uhIGbl•¯¢­pujî‰kY›öXÙrTUðUìüäf±<ÝäÐqÂÖªKë#JÔ—¾5‰Zë³$¢ÙÞùmmÁË)É]Œ^aÜïßdùÍ­¿Å$‘\½à¯øßçüñ·ü˜ŸðS>ágü#¿àŸøžò+>åÓE±˜óŸåL[^òŒ_óëëœÛ¿¯¸+ø5Äšßð[£~›ÍyÎç¿ãs¾à‹yÆï9l}¾˜ñ%/y™?ðŠW·Ë,ãÕŸ ¾â_ùŸüûÁ*=Kø?a°!/EzCžŠœ=3[ÂI³ãùÍ•æE/5Š7HzsÚî·ptòé«ù #˜q’—%DÝŠ¤#~VewÐÅAShòÞCϗã—ÑÁfwžØëÁ xÂ>xÔ³ÁoMÙÃÁú°€‡j‡ÔýùgôVlÝŸ–ŽÀÞG8„öLuÇQýnP·jñ¢RðPgÄp´ àtÇ€• øË zêvûÝÛ< £mA¹À‘úðzP~h–$¼dYPŸE ÖÏíé>r$+,3ërz.’‡´Œñψ:ÓÑd¯gAª?5ˆ¿×o€Á_ø;‹Ã÷üCÅó—ü³Edz$•p6a~Š"åÓ|9]Ý]Ã1­á Çó.mƒÖ’À³ù,-oyöÇ*-,Ž·@ù&ƒ1™Ï®Š²‰êÿ¬q=‡ ¾7Y¹"»®ÜÕÒö:‰ù÷Ūäð?V‹*C³¶¶¿q/XÝÝåneö]’®( ¢x­1òbÖÑü/þW¶\´ÔqzH}üS®Ácb'¤£@LW´ñ=V¨p¬þøõøàèí{§ š PŸ¿š—ùæÁF• êí©’†+Y«=R•ãÕ‡ ‰Ÿ8„6T'V™à0W­—eÕ7‘öÑEØxE óˆ9W0yTC9`GÃÎkCÂ%—ˆÝ%¢þ.øtbšàs· ð9èèN´¦±ó`q˜ @ÔpRoeÞI«—ц=kŠa˜H/†"êš1*  Q֑¤+…)-†ñ1ä—5[ˆ{=™¼8:ÞûpŠÞÎß¶dR{úQ£–ôŒšë†¢á­}Ï/¶ÎêÚ#D¼(TlÏ$½ÿt¸zÖàqň=ch’hÊY³e#@aÈ(ë jˆ§AÔjß ÚÊ1&Æaß§9d ]ÿ”>¢Ã_-þÞŽ£¾ÈÐåÖ£]FFx V@¾”p55œR c+A'•QÍže"íаLöv‹eÚn—zö¨e|:6ÇY›|ÀYÜfVz>ãSZ?P[´¾P®Z?²ƒmã­´:¼u0o£ÝÆý}Ø è6Â<ò7½Æ¢°‡°`¬Û¸ÖÏ»ù 9˜ʺ£Ä5¶}i™k?®F¹÷áèzÈ#\ÄÎ{³&gDèÿ’6-Bæá]YYö¾Ñ-ÖÇ6Î Tɇ-j¿¼PYBj¡»ødÉFÛ22VÅótþ—ãÃ_.ÑÙÙÙh?ÄôóQ7`F+颙ïSUPF-yÎk¥ Ÿ»¡#âê0ÒQ—ù~Œä~UN?ÿòë+jÿ¸­|»ñödŸËÏ£¸‰Å§ñKá[â:pÛ Úl½7¶ ÛõՌɆ°¬‘L¿n¶As»Ý>®U;ó¡šB³gíY´ ‹ïpÉÁœ@î îlOæÂ [( Ö}¬ Zí²½wÄ­‘IvâÖ‡±`—æI°£JOø^_¼~Cæí|4ØÅ€ö‚ŽhlС¬Ówµ¹þoO· ÒŒx;¬õ÷Çó¼úQ¦íÕW/:Q š<óínŸsùº‰8}å4Ï]ÎÁ]» B­ü§¼Á«tɯ–é4³¡¾½rq>]þžUëǸv]ÿq¶( 4²v#}Îb•ÝðZXÌ–}g²ÈÊòï$*êÌÄØ¤„U«ù ¾ðt±ÌóQøPuR®TÑèu{©(öÉ‹µ@Fº/¦#°iZY›ˆaM;:7ô3v­¹ùÌ/ zº•vaR꣯ eKʶô:ð*¿¹iÜU›ë»|¶¹©àÁ¸ T³¼¼/R¨·$Œ¶SõT°™-Ñ*Ù.[T áêå£:“?Rº¢`[‚ ']G—ï>íÑB§» Ù7èÏP{2ö §¸eREØ:zÒõSPÏû¹ÖeLjÈëÞ“ ¥.éÞS"CÚ¨œÔ”+]×Ðß»À§™ë(ׯ‘ä›=4ƒ±+«u‚j‰~íh·¹í/BR!­[õí­ì¥¨Z2V»=Of£Ìó¯ŽÞœÿxr¶w2:A¥“¾8F]¥7ia½]ªhHÏ@™vfÅ´ޝù¤1%ëùÊQvå‘ÞïÌG/ÐáÑØ ¶…üýõÙÅ/_.¬ƒÕT ±qìe×±§DåcÖæ±ß!|w8úѾþîœÒN„ÆœºÛõœv\Þûe~—5qoÕûpuÎQs)î…°í¹Ñ«Gß~º<µF¥5Ãû›•ž)îz¸fì"dHF¡Ëì–rl©Æ­ÆÕÆ_[Ú¨á$i€›{›Ûl½Ø;Úq;=›L>°Ž¡«ÌLŸá]n¼mM’±^Ö¶Ÿ–í DëAÍhÓ&!9FÀ"ùŽžmG~§}5´ô‘4THA#µÝ™êÞv®öTüí˜ññ4éþ§') ˜ŒvÙ·aÖÕnN»H†ÛRœô Voft{ftË‚9¿ŽÈx;fhçŒîMI/ŽoqiìŒDOøµYºff®•[gÝøäúµ¹‰Ž™Ï«¿VÌ;M„$Þ¤Mâ„y•31”mRÕØøùý›É«ë›Ò¢{ºXöUƒúNsëejx?à@¶øqSص}[µrÀè[½¾åS½½mnŒ”Š­«¡ÏHÝF,ôiY 2Í®§),M‰é[ÄÁ°ÏÓî(ÜPô!ã÷§€Ÿ€X‹ƒDøn8ØçÎN¿õÒ͹~N‚ü;¸ó½–£†ByÄÔä××Ù’2¹6æ`,x7ŸÒ­«>§= ѺŠ6Ú{Æô9b'±l?$Zgiïr3ÅLûÛZifâŽO)[ éÓÎÌg¨é{®M²™>`›ü käºÝÃÚÙÍ€)ó{}#wÛ\Ý^Ç]5Ù†ùuAän“ߣ²³>þ m‚)6K+^Ò5u–¹ìì®:–9;ŸÊ5ÇIðê,Xã›»¦õõ‚ÝøÕü¸Q†zó]FK:ì.»gbç¬à®QNä·’¨~|©ðFb•>Ln%Wm¬ØŠëâï^^•u¯;¢¿ßù®åÞ®“Óç]ëõqò ܲ·m½¸ÏéƒYZ·a êŽÅÊÿãE´›endstream endobj 75 0 obj << /Subtype /XML /Type /Metadata /Length 1751 >> stream GPL Ghostscript 9.26 R, finite mixture models, generalized linear models, concomitant variables 2019-02-18T12:53:52+01:00 2019-02-18T12:53:52+01:00 LaTeX with hyperref FlexMix Version 2: Finite Mixtures with Concomitant Variables and Varying and Constant ParametersBettina Grün, Friedrich Leisch endstream endobj 76 0 obj << /Type /ObjStm /Length 1949 /Filter /FlateDecode /N 73 /First 627 >> stream xœíZmoÛFþ~¿b?^Q$³oä.¢@lÇIîb_ »qÚC>PÒÚf«·Š´ëô×ß3|±HK¶™†¿ÅCîììÌÎÌ3³«8'¤p^¨(.Fiᥰ‰^‰$ޅד^xƒ¿ÏV¨Ø'ÂGBy ¶X¨Ä`Ü mœÞ íÀí¥‰:Q^$Jé𬅱F ¬`œt"±Â$X5‰ð—ßÇÂJmEâ„UÐ(ñÂêóam %$Ô³PNI%lä V±ä*i„Ya½v "ªJˆM43;s$Ò•LDÄ+)%Ed ˜•a-èY0*eDŘ®,Þ‰(Ž GÅ"ò¯”±ÄTkc©Y`‚—ÐRÄ:Ât­D ]AhØ.¥ˆ#‹éP..—Бˆc9ØíØyædS”†doY`"œ„• ;æ4,P˜à4›l4ì¢2F8AUc…‹#Š@8È1p¶+ß8xÆ)xÐ%¼¨ä›­¬D$À&e¸GYĆâéìv1?l¿Â{cy"Äðê¡`Á¨ ®·ˆe9hŒûÇO? ÚŸÏŠ0+rÁ»"Å@ÐQgéÞüFüÏpD½Ö-磓P@6}:8tn ýùgߥ_<ÞÕ-êU7¸º¥›¶OÓ-îU7¸½¥›•›uû0….û·:Ò—ÿ #LpÕNmVz/ÍC)—Žþõù×÷ÿþqÿèäãé)ÒÞÎFóq6»t–ÍÞÌòlõâ0[æÅþeºäR¯ƒ–Ù¢˜/œJ?¦5' \ ‹ruÖA5ªTKŸeãâ2‡%ò‹¿òö³zÓP÷¿iÏ^qÈ ³šÛ–°i–li¸¾ÚãÚmûÝMÂׯ±oòÇV §7×å—´Ÿ.Þ‡ìâƒ@:bqý“öhŸè-Ò;ú@鈎é 蔾PJi>ʲQ¶]M+ºÈ&ã@iAC¦K.ÓQ˜„ó¢¢–¼BIþŠÛ× «æ“ù ÿN§))P©…?¯Ò …›Ñ$Ò9g×ΑtAË€tXÒ%]~[\†eô;ýAš„<§)ÍhNóY -8¯Ê5KªZq•æcZL®rúój^„ñpBKÊ© ârÍ銮fã°ÌGóe kú‹nèýM‡åüA伩¯40ž'éEε¯Œâ½*Ï_¡œ”Ã1*oþ×jø0›dp«T>N§aCÚ}(ÒI6z3»˜ñŠ_¥7-ïeyŽl½9)Âôs•d«¤jå#}©ÝmQ$êè ¸(UãBõ†+´¯±e`¸ŽÕ´åúÐÐ(k¾¡c¡~Ýð£j4ø9긡ÁÓÌ#FŽšvÂÔ`>@ƒÐ Ýeߺ¶Š-€õ}¬÷-|õÉпBÑßÒÅùü3L¡ç B¹À–b×`pþË,S€Q5K ×Äÿr|öî·߇Éu(²QÚY¢v@ œõ:8»Á­c ˜ùcàCSú1ò¶þ—¹ø‹–l«ïºÔv©ü2.ªäYqznŸÊx«ÆËYe8£Gµ¶#ÍÆIÍšRvÖ¾¥Ky<ʯÐÖ²²µ­çÈRƒRšl•+o× Ø¡Iknõ¸º ¹àžÒÙýp;¼EÎfvaò¸È¨8¥i6ÖÍh–·ÆF†Ä<\_óì†òIš_R¾¢ßÊï"#à¡FFÛF_ &Üc º¸Ë»¸¸ém`”w€½Ò!d3>€}ÀYß`1|ZðkÞ6 ×3vÌf0‹UŸ`–ؘ%Ñ“ÀìøÃÉÙ烕^íÍ'ãä4ýC“軣v‡œnªµ»¡ïç|üSÊý–ß Ó Õs£)ÿõe á÷Õ˜ÁsýÜI´À¡|¼‡€`ˆ> stream xœÅYksÛ¶ýÞ_í܉HÄ«ÓéŒqn¦uëÚI“¦ãŒD[œJ¤‡¤ûß÷,DJ|Ȳ¬èÞŒ $ìî9»‹]Š+ÎBÆ•`\[|K&”ÃwÄ"!ñ­˜ߚ鈞f,Í·ÌIŽoÌ -&hÜRc€»ÒÑv‚¶Åm-é‚îDŒƒ]´bÜI,Õš Îé‘aBømhMh• ™Ð+ gÂlfN`9.d(éN„,á˜'¹¤1ˆh`˜”P›C°”~Ǥ ± †H%°Êr&5§À€–[ì¬&[†Xn! ­Æ8ÌŽ$™c-‹”„qÖa@Pº‡;޳È(€D"§¼Ýx¬BG€1@Ä@lN« SÒÐ#Ë”"ÃKE2¥­9S–` ÐNÀ$å°¡#¦yhC…P!è6,Ñ‘ åƒˆ&;¦UH<„Lû„& ˆ1‚Ó#ÉL*I9£á‚+ ˆ<6†„rx†#¡Üb@B¹c6„ã¬´Ú `m9ÐÐÀÂÊï~ú‰'ù"«Ø+΂Ӥ¬ØßÌÂ!/Yðñ¯O˜4G2‰ˆ³l1›±kœ¥æ‘Ëúy¿Æí«ß’‡Ê»±¿ºˆ‹›ëúªHî™YŽß¥Õ,aߟç“dÆÊ»dœÞ¤ã¸JóŒÅÙ„A“tî/`?ÿìõ¬µs=íw# Xíj‘ÜÇÕZnÇó¤JŠM[×È•pÑ•¢m4ârˆAÔÁ@w00ObÐWî}™f·¬š&,K¾°›E6&õâYZ=öÍhFö;¤F³Z¸ê_R <**Â;g,yˆçw³¤/ǘ¾X12-/¨åèÍrj3£ŽÔ㤊_}žåã’¢d“¸ŠË¤vb¤‘`wE©.ÝE>YÅ{@Èòv1ŽXY-&X[²4cŸÓ|¿Ë3zÒä6§Q²/i5e÷qñH¡B)bœgecæ]Úå&ÀÄasmÀlø€éA$»€õ#¹ñ¥>`µÿ¸|Ù8Ÿ§`•ÆŸqw à ‘ äãP©H­#QwS×Ù¢@Þ*Øxg·ÉFäAh" f@ìÁ@墅…ìc!ôÈš~”kÕ‰r׎rÓå©å]'ÿP¤yñ#,dù—ŒMŠô¾ë¾½#y¨ b¤¤hh:šNr]OÒÑn¡÷#û”ù«4»™IߟZLG‡eZu˜Ö_kºŸQ#mùÀ×u÷y½<dëzÉÕ&ëÕA­wQÛz§ö÷ó¶ j;n Ö ®1aÛš«¾ãÔ AõÜè)¨¯‰‘jH‹ÛŸ}HZ¨ì_ÓBµÿþNix?é1²èázÖwRI/h¯ó9ÎD æ‹j–çÿ¬0¸”h›–2VwÐåÔ*_¢ù൵€Í˜©fL-K3Fg(:»‘W kÚ÷N¦qA`,q‹9¬ÖÖ3$|cB”‰z«¥ÃÑ*Fs=¬ÇK–Ð;¡õ£×ËûçqU¤ž¿Qèí [£ë¥³®wºZ|®<ôD ºwùû,…tphoòÏ=“Á‡tRMKlÁ/ÚŸ·Û>Æ[µé£Âõw¤hD骑&pméE‚ZÏADs„bíu+O<*Ç>à\Nâ»ÿ&éí—¨®|ÐQöù>x¼â`œÏò,˜IpLƒ4˜ó î‚"(ƒ*¸~ðIÈï÷ŠÞVg³ø¶„ìN†Ë‡Vr/µŽÂ³t†p.\çÁßP~è|[¡e·HS ã<-)W{ä½ÉÁU•ÌÿWQ››¯ÁÇÚÆ]ç–päÊ¿.Úà’O$²§ÜvƒCĶíª3bSÙ4uÙK‘ H´mß‚-»Û5ÛÕ5‹ÜîÞ"toHéœÍ9¼J ˜Úö^îo”ÏP|¿ô¡­¶œa›Ï³õô ÷pMÛx)èÕnã[OúîK #‘MžIã586eR/ž¾:Ý·|í°ÝëÄÛ'†võ5¯©Cù=wf—ÛùÛ5Ãoˆ‘>F‘ðoøŸñý+Ã80‡Œ:¡Û É>}âÿ¬QØ{XLL“}^k>s¾>‰ÃFÛ‡öºƒÚË;>À÷ñuYþ×ï¿üñæuÜÕÕ[îX—GjP¿‰¾V]ÎÍ®u¹|®þÞí£%£¡ò?¼ì#üêõGúcQø~Unp¯U•//WUùEp\ÓÇ»i’¡/ƒE§MýÍ%ç¾ïÖà‚ä$ãŶJpÙ$ªu Þ§n×Yã¹\ó¡§ü~~öé UüÇßíî*ÈCW‘Wé·ðO¹Šòõ“’bõ×SUÓ½<º$ª M•ê4UþrE*´RÔFUmÞìvÚ,Óhåh«k¼ÏÚÆisòYÖ”ØÖ9QÙüžx)tèqþå|´÷ ß¿Û9ð@9¿‹ÑKOÅ­Hœ¦77 L#[þö%UW×b«ÐùÓý 4ÿònã‚¿äÈÊendstream endobj 223 0 obj << /Filter /FlateDecode /Length 7324 >> stream xœ½]IsÇ•¾cø#úØ¡îÉ}ñÍòÈš±)‡-q<i €jp e‚E]ü×ç-¹¼¬®A‚tø f!+—·~oÉò?Vj§W ÿWþ{qsöïßÚ¸:Ü©Õáìgšþº*ÿ¹¸Y}ùG¤•Ž;£½[½øñŒ_Õ+mâ.&½ŠFï´Ë«7g߯ÿ¸ÙÂä9gŸ××µs)çhÖ/ñ±Ç6¯…ÇFeÓú|¬àqÌqý ÿár†7`ÌÖF·‹Ù­ÿ¾É§Ä“X¥‚Kë×8w†iòúGxÏ&˜Cí¼²Î¹õ{­Òú§ó;¥òÚУLXÿn³…¿ã’°mµKÊ{­Û6t^ßòt.³~#¦{ÉÓYeáðbÊ:ù"Qç¨Þ»ç]«ôp˜»rÚ”×oû$ÃâíçÕæ_üélkÓN[³Úí³iõâˆÿ‡M}÷(hy{qdb%ØÂøÃôtåñ~ƒ< ¨„³z§4þOŒ¤êüà*“}…åˆóë>ä%<Ùºùì~‡{“£o/™&°¥÷¯øŽ$*¹P àùgözfb´Úvbµ ßÞ±p¦MôIT¶b²D»‹Ö–Žü»¬=Lƒ³RÞD^¯•·~ØÚÍ,&‚7Ñ€A»„àNXaaÞTØm »`á÷Öfx;ÞÛ—(¤É%mÛ¼šN…ó& `=ŸUšéªµ÷v}»/¬ }Íz4Pù5oÔØàp:»‚‡Aå¯úÏrk 2¬r6f%‡­ïo7/þâÐÁJYÓA´rLcmwÞó„bÞkÖÇ+Ü»ò*®÷·´X Ï-<¶@•ËIsv (zÖÀ¿ôúÏø8) ,ÓÞá}Zÿ,&¹&re “󈀯q`¼rëÿ¦e€ >¹ŒV$ò— ØI®Je€Ü³Êhã˜{â= Ž©ú|$ñ2êœ.æHÀ€š@&´çýÙ÷ëý>XJnýœgtQ!úi~#Þ‚8f7'2ZiØX :ÿq“ ™±,y vìÙæ17‰*Q¥â{ùZ¾æmFíNH\æ¥0x•<ôy·®¤áYk/y6†öY@;ƒsÄÁ5ŠÉnÛž„—ÜþQ—“Ý΂1¯8ç ƒm£½û«ð{oº ½h<À%Ôe%€QÎVÙ¹ö/X,R6d>¸«ØeÂ=`Ò@B“P“wÁzYåÛå‘âœ@DLnšÖ$í-+‡‡sˆ•+åAH%1¯†Cµ9Øô:`œŸÜ­)u€E‡bõŸ ÛxLÈD(pêÛW,mk¨3 ëmw+[¡íRó÷œ8k¦²ð|Á8üÖAÒ^ˆÂ 2R`€re¾j´ý†¢Ô„A;œ@y÷×]qG’©¬‚¼:‘?RÚî¨nxêP=ñgTz8=âX… Ú“šXˆø˜ùŽ'q-˜÷ ¬!V ML¾¸iÇÞ ƒöÛA & P;:¬¬X‰wç…¥!ß.ñ8Ê…Ì7¤WåU!š"I°—Šõª› 7¸}ˆ«o3+ ¦gyØÞD¸'†»‹vÙÂ胺wØFô˜Lý©Sô€Ö"gØ!:JÂDžºšš*¢/Ö€²ßu,¤päÍÀ_ü,™½N \’ù#FÚD䢪‰¤èä` z‹ŠQ6ò/‚c1ã>I±c^°2‘„BTçí"ÔOHÙ~jD^©ŠŒß ’ß…ä½¹'tÙ8ÉT²E„±LE;Äè?„è÷Õ¿àͦ©‹ð+ÓuH(ÜÌ‘ÁwX:”eþB–M+ÈGðÍ`λ—àóxšËàMŒ¨c1¼ld`=4Ý™«Êç!èÃëZbŠÎý›\bà eIu²ò*Þ“[QÈÝ—*,¦9?ñq/Äã£(¢Ï-èòB¡¼X ›c3*sh ‡LYv AY”™ü+B¸s"´A3 (#"÷_@5–Ÿ[ ‹ËüÓiÅåúχŠËØáÀz3T—¨tð+ÑÜ.!iÜ#0D|W»Êâú¿HI¦„ÐZ©´/Ý\ÊR+Ŷý>búßa¸¿x#þòj#Q--4N³-À ÚàÖXÌ`ÅÚ¾AQ„©¹Ë£P‘êD77\~ nö™KMÛÚÉtˆ fjŽd±]áº-ÂÕ;/“›è>1Õ©ð'vû1Øžwü™ÑÂΛ0t9 &á{±yQÕ(h øæ }²P€t¢ÜGâ]R»€ÙG±ò U–»8Yø qÒ°rÛuÿ€kX"²èòp$ïœl¨LJ‰`£çwNíx+̼uŽ^¶Ò\ò‘°¯íjÞ±° KÕK3ó$’²ß„3¤ Bõw¸¼¦ÒcEŸÆqËY=5 ¹ldalN"Tæ(cÔFa T¯‡n0!ÒH‡AD£çMÔ%ÞÆÓ`êeAgÊt—´MàþÂÙc ‡Á&î/iWŒ;¹@Áö»)™¬èl!mú Î» ¹ Æ"˜Æ•ƒŠR~ïgµáâ„×Ô¤'¶ÉÁ<›ö»B–«­NŸðåûΔ»N-)8¬:Ú›aÏ[rPè’ûF‹½´kLqïl”$Äç͹ä\ÍÜ_ÓÊÕR^rÏrÊôn“,¶FtQ¡Xêš×!wuÂsäž„˜øì{*Ú¸"9Xº(·‘f¦Qå²²Ì/ ‰ ã)X'Z׆F–Á×Ýi)bB?Î58Ä7 1•Ũ[@4N%ù˜Ã ë² ”‰¡oƒ÷w¼z ‰Á·û(d±zR4„CCÉ­ãÔXÁÞß LW‚fø[W•šËE¿h§º"O ·ubrÎQ5æüÄ"è_;<X4~zÑSÖ~t2 NÖ†ŸØ¯;X— 2Á%?…q87&ö…GÙ Aþ‰leÖZZí’Fsq0«eL ªDø² ÙK¸HF»Dê)AŒãeò‚E«Ù&âUÇ£ÂkI ò€x¢òüþD~Ër²ñ³fÒƒ ÜVÖ[ ‘–{Š1IÔ°ò”Iõ9c˜UK:¿µà¬ØW¯]—®—<•«íßô޽´]«–á;ÎaEãã$zŠ |¯šüÍw#×)÷€øÐf.I{g¦G_´ÂLwŽš±óWt@púý½Òâ|BW1®Hy‡¥ö½Ð˜K>¦:ÿ“Ó€ oAH0õØæ·™³`/epãQ¸ý­7’#vä:k ‚ñظ¹)‘Õy²óM÷v_ ‚Opl´a\Pº?áS»ê¿ÀáBl|è/J|Qñ=“ïºy‘vK >…À®4Úù0Þ.(5 g\*®{Â&è’;)4ÚÍ=[^ÂS :þ2&Z>’³à6w +&ûèÅØ’YÓCŸ€ŸÚÀ3?¬ƒ2éŸ “ÚAø';ßôpò“ˤöØ7¡•àÒ…°ýÏK9+Sãl3Ç2v¼­íáÄa?"`ÿ/…y§N8y`M½4V[KÈþSX)!NaN$>©™21ÀÆÓT$ÒDû1´nýsÚ)«2ÞFW|Ãe"(˜Š¹òQn#‘õ9ͳ0Âó§ø" ìÿ\FY¶:N¸œóSœQÌÿj.#*šrù›MSN JIóQ­Ó˜ó¹î<쪟¸ØXXNªŸ &‰ë¿rL3Áù"ù{=“´üpRº¬±Ëâ³KËx/ÇR`Àu=A ¼½ Ó­oÆ­oÁMÓ¨IÊDw^/ØÌ'¹ €.`ÇV&i€a¶^r¦x^{‘  ð;`±ïÇôN¯rÙèWòOßË Æº§­çêÌÕjï1ƒÈ²J™ÜR|•7Üû>ÏÕd†¤Ô˜ùÁº<çÇt e¶y¤åçðE ààL±ÊgxŠ/¢"ýtëÂJí+µÕ.Òø­¦nfnÿs9³†–$FLnšuã>ݸÖü3—öX)í'¥<š6¯WʃŸ6NºDH.k+D_?qÑ «!íªƒ,:R·NGý÷2ˆ Ø­Ë=yeꇹHz;V+‡MÛ‚k[iSäÞjc°¤#ס00¤öjžg>Û~?›ÏœîT V¢Ï;ƒßÝY"qg*e§çÒŸ¼êÄ”™ ¢ÿap`OÂpÈ8îã³@ ·óFË€»ó1ÚúØC$ÌþM1ª35‚kÃ*€M6ºÖJ›:6žËr/.p7­µKå›…›ß‡ÛÚÓfjà`ë²,ÒFœÞbýs'ÔÒ¶¢‰>cKÉ^¼aÝî4P¹dÌ^ÓGµ·Gجí™,±›Mú 6ćÁKµ»ø’">_ëôêí"¿$kI?Ïa§ˆ¡ÂÃñ-'þ~À›"F)Ô¾»rg»õÈ'ÑÑc×0- 8¾|Ђ  xo0eí ˜‘Ð0-'âÎJOîL*õ-½´k53…e•aÕV7ˆ s× YÐKY”_9 *æºê©•Ó 1‰vå~ŠÃ¼5#ªÒ64`ÖØ%à^*õø’]¢¤ˆže/»¡¥€gµL€[zï».¡sNïpFÝN/€Ùœ? ;%®]¹ 顪=M,å2ë k,vº]g¹P”v>ÉZa/Pc]_v¸0.Ì¥@¤=æ‹õeTÚÏéŒÞyˆ;-† %<À¯Ý$åC°ƒÚßÐ%çljË׉ßÞ`ê`›ž‚›ìCã1‚¡‹ÙHBJºP~„%Î1_Ì…Æ…K^Û§i‚Ÿ=¼7_®Ÿn­1Z›"ÌÅÎ!Lݯû Ãvê°/èpºú Ä´±­>– š9y#„[œl@{íéÝÉŠc'Ý%ßœø‚¡…52gãÄJ»Q¼h”t”²Ê:©uÔV‰R»¤þ®ªá¥ê_º[=`@&uSj„ÖäÇk.!Ñ*ŒAár/43¼«Ü>²\œ‡²±«É­Ïy€¯5~Î¥ÅÒ®30¯—xâ"¯ÆŽ‘¡È»Ð7&ÅBQ’Z- `ÕÝLn¨p©$Á“À!?WRN+´‘Ø|k‡¯P‡ú°¡]DÜnCÕdéö·LK r¦+Àà „o½êœÙóí0dÜK€]Y†“EOIt¥´3ãyK­—§21¸]°u/æ]ÄİlŠ;3”Žg°«ÎSìjsÜMÚ¾$Ñx-ýŒü‘ÒœZó“Ùä‹ëVæŸjlÂYêgé‚Êá1,·.wv!¤Ä¯ÿÀlªê I•%¯KX¯Íxîês¥­¸á9@zZâα\?©Õ\ß)¿{ë¶{ Úq1tf#dmí4²)•a*~ûL¶GLG”gqÁ?Ð}¨’œ;­[ò}H0®õC_õ-ÉŠâ ˆ8ûÞOí–ëpøIkUAiÅj{­¤³,„nÇzƶîyD¾3<§^ÿY›ß„$ïeóÙ¸ù‘¾ëC7Ï‹%hCµ8`,ðaG9(MÕtèPêe¿œ8ß \þ N’°¾÷Y*„S9š•ÎxGÛ±Uúƽã$èAëŽÐËÑ£Òç{¦-u6; Ñ9 ¶t›ú$„MÔ¤ùm·ßÜKí>ç Ñ'±»á¸ñI¦œäÓH¶I&&úDx¹c޶ã´}(C;Ön¬·– 2Ça•”ßÖà‹â³ZÔÕ÷»n¢•NÑ‘Fì÷¬{”š®KFO(C%Í“àÀ¯]—q}:¨9œ¯FPü.Œ=,ôÜÆœí R»aŸ>¾ùŽw‡wë&õtÚ´ÌÉ~Z>Mѹ$2ÇÖŒùVäóÚ*qZ†§çQž­éÂõHþÓFÙz6„‰ïg¶NxmÞ‘.Ü­ÿâ+>HC¾î(j92¬¯éê%ØWHÕ¿NÆ¥­AÂÜâ~”²Ó^]‘ ÓCHÎô¾™r¨Ñå°áŠÚç!PªŸWñcÕdp-“ØÆäXòRMø&]£í´mÒ¶­Å+8&§Ø¿`À¡†0–Mpß”¤$oŽEM)U rýó‘û ÏÂd/-rçF:¢D§5Ž£¸·3S.YP´ñ D÷®úQûl÷]ÇW•ƒõ»ÙCÜ{¸ï8Ón5}¹èI¹œîº+ÄVæQadÝù¾¾÷ˆsçe+¨FSC†ü†ï…aGP‘¸0Ùø‘+•M㨲ââ ¡t—ß*$.ÞÄ›kвK_ÙÖî¦Úò1 X“ºàùÎ×è æSÙТ u¥ÅC1—y C–}S½¼l0•aÊ´C]XMeÅŵéå„zeo1}Æ÷—ÃÜÁáu1ü‰ð¤¼ÇoIÙ+ÍÍ$"6k!²ÇΈe-¯½ßâŒÔ•`¬ää)[&ý¶uß\ù¶å3.‘¦¿´“‰ôÑ&Xe1ˆgÃdý"óBJÖÒ©]‚nJ#ôGöñÉ`ø’—ói¢7!=Ì£ •ñ»ú¥ qÅ`‘db§GðjøÛ¯³Ô>³[ŒäÉÂäÁ:‡ÞðT)‡‡“«­öFë~üªåtHÎiýÎ82¿OºBJj€ÙÂÐðÏp—”›Æ»¤Õ]H´ÎWH™š8f¼¸‹hãrÀ¦ö@`4þe¿£ñÿ ñf‚#[üB7~ÏKyofÀµëGIÖõ„÷”¡úå­ás´²kY*Rù^OÒ~ÔÎG“~½ãYňéi¥wÀ¨ñ¨”À†"}ÑÒæL‰‹t×»nq–ïIâ¼é¤¯N¼$ÔK·îê§4ÃoÞ´ÔPy·ëAÙÙÑ)ˆ tËJJ}öiš†®Û^D/ôZN‹JÙƒ,Çt@#oé͆-ÓD”Æï)™ñ Oͳ£Û•eLãǯ æpZÜn”2yXUÎ _ùVĶ~79µˆ«+rÞï6‰nVõ~.yAM¨Ûr=÷Œ ÜÏ!–b°<¤s{ ! àã[mÿÄbų¾9¡ ó‰6Ä:À×c/E ”Ãy6ü AÀÿ§™Aa{¿¬YìÚw¿Tj…Fš Q¢láâè/|nÍŠ«&ôy'8ödÜ|ìía‚“‚óiÝÅ«º¼Jž^nÁÙ|¥"í¶Ì?ÕL2~Mæç¯Âx¿‘š|†3tÍQÜLLË -ú¿É(ñ~MöìÅå@œã/<æ6©Ýßœ¦«} øà“•‘?žñ‘fšZ’÷ûÇÁÆgØ+NWÎO.;C·…Î×þdh D¦W‹‹4°­¡œ1eÏð:6øÜÕ¤¶"§«ê=a¯&âMâogÿAÛœ£endstream endobj 224 0 obj << /Filter /FlateDecode /Length 7967 >> stream xœµ]K“Çq¾C ý† Ÿfœa׳«ìðRH²lJ!Q°/”ƒÅrر úâ¿î|TweVW X x@³·ºžùøòQ9?\Ls1áåß«W¾úÖÍ7w¦‹›G?<2ô׋òÏÕ«‹ß<†ßò”ÍÅãïñ§æ"™‹9̇ìÂÅãWvöòñ? mðªm²ë-´üôÑw»ß_N;oâî%>šÉÅ´»¾ÜO—“ q÷^›œÍœvÂ9§wϱŜSÐ`ÿ™wÿ} ½Ãgz˜!çy¶»7Ôtš|ÚÝÑÛ)Yùýé–;9Á”ÿçñlÖ·_&½wábä©¿¾ÜÛxˆf6»Ók쮌÷qÉÆ,Wt| ÏSšmšÚÁ§ƒ7°:¿;\î£u‡4¹Ýcx›¦£Û=]?ǽHÙL3,eï²…MÊ»WØ¢ìˉ&•“É»§Ôóìæu?³ŸLŠ>L&Pwe~Ô;øÉð&¥è¼— ä ôÞ¯–ø†{ŽÙîÞbë”s 3NÉes€OG¼¯òäŸÖ­Ù‹e“_]âÑú‚z}%žiã|B4uÓRwÓ\2ÚU´Ê)$X™èâﻺJ$›ß=~ô×GÌL>§ÉÏo>”ƒÜ y1[{˜bF6ún÷— 8ÇÛ:UI"HøÑ8àyz‚N¼ †Ý‘ú˜lrV·7ùÝoÄÖºERÔ‡SöÁ±Úiš.Å|è¢}Ê3[½è¿_Ò¹%?߃/ÄÄp{?%êqoÜ!øl™aå‘i#ódmÜÝ2F }Á3b÷~µ¥>b¥7Èé“5Ö«×|2:BÄ9¦P"1á,¤/øÎÀ—s!Ô Tk[§@-PXˆïÄœñ;3Mó”ÊwDBÆ,c@ÇÇ×õÄ_ׯ/kccFyUeèñUæ‘ó¸^·’¹ÔìpÜ%{Ë3ò äÑ“ÚÃ;qüôè§¶vHºÓ÷0Ï/C y„" é µÃ•õ3QŽõÑëY»ºY^”{|¤%Åœ@‘쎯HpÍ&;ÅULHGF½gJó.ÚJ1}:ŠjgpV ×S’”X¤¬‰9$%^Ú‹*ˆoE/§K;ï~ª¬·*ÌE‚PH(@œÍ‘ùÞq[ÔS‚ÀOüv23œéªF¯¹ ?i6»Ãg?MDVÿ¨_µ˘èf/QÀS±(ÞbÅò_xñì&â¤Â˱Ýñœ(E8JøP P¨Tè»YKj1ÞñªYø”Âä­“Ó;ë¦ÉØw¡¦~ænà’ŸŸîHQïæ†ÒxV“Zïz^©œ"Êåg̰“Öæ…;®¹%Hà…ŒGÚÇ} @Õ?WÀ§ÅÙ:É›}å ©ÈqCg˜BÔà׸XrVŸ×­o’aë¦ –q×(eAcå±­ cµ/ÒhÁ@Žå…Û}O2bšãíxÏ ‡âBj õà~BŠMÈE'!0_`€ð@T3DÉpˆsüDÅ€˜ñÁ]ÄÿZËÚúßZ ²bþš'Bª…ÐH /u»øî ¡éO]‰yÔ ÌA1 M·¼‘@iØŽ<§°FwA*Îz˜â¬×NH¥,øK+¡-â¢f[´G—‚¦y"XD?âQÃüg)ÔÞ—¥Ápê«WØXh–s@}ÇÊ Zy®gtÇg’‘ô(ŽöVçR~ua8QsŠ>¤û(œ¡MT2U8Ä_W„-OäîR˜L®Ny…À[lòä úm1:Ÿ n¬m^ì%¥­aeÞâðWÐ%¦z[Éèf™Þ¼“~†c•võ e &Él¹ãÉy¤`Ü]$C¡°N—¤ç‚MSqRù¥RCì´Ðìµ,‰@o­9Gµ:AqôQ(èS]ÿ 9Pæ­xü-“ÒÇ#îa2”ÄÍSiÝJÎ#ÄËÙz½ÑŒþÉ!" 2x Ÿc„Îð‡@‚ÎGü`wÃL :5£ú˜¡µu;m©Òwqrƒ­ùT˜ažm_D~ö¬7™ 0ÚË:QA} Õ …¾ÎzÃh¸ph³ûH Ö;#%û ’ómÊÍ æ||Àz‹ÔnŽPxÐ@N æEç’8Ô:W¸öPƒÌqfÔ'²FIxg7Ç ¬Qfµp=J«tNÀB*·`ϹLKЉ¢_fy›shrl¾*E¼ëO¥ˆq§¡2?W ]¼cèÍØê'òº ª“rô)£÷hZX¸$ÖQŠ«º½§wRª‚%wõ–ò*âL#¨“ª˜¥`§õŒ(½@êirgPBÏ~9Ýc`㚤išyŒhÙLŠ|Ž¢õ‘›LSVȃ C+ƒ™^åê+@¨ å¶¡Y¤í[gÑdÚVÒº‡?;ůÉÈcÏ@,æÌ¼ iþ\:9/]eçŒW¾7íªÞ{ë(6säG Žb¶înåúÛ†lhD´xÀ¶qL˜Ðã¼û o‡‹qV=¼‘JJ"ò±–±¤ƒçÊæu¯ÈWÌú‰è‚BpäöTΈÆxc¾Ù8”sã ¤†«sÀŽêߣò©~‚gØJʹվðw9iÇ\cþtp½%j±Ìnºù`æÕM÷´Ðƒqh3e„Ýs!ÀÙÉ~Òœ"ð§ðuèÍX6îVÍg•ÍÛóÝšı_¢¡óÆ´ò6¶‹‡€Xd––>BˆB;¶UJ.µ®V#Trš–›Åhüò’f>'åE8ôõÉ{ io¢B‡ ®€ÆhÄÁg^KéÓ²QŽvìjþ˜÷F³'áP‘zÇ+Âp!Åbà9´þ]Úç²u‹RòšYYóuÙu=Ð(fVkåݲ(h¶ÞkP´! [r[?«ìt]g/IöXfLqÑæ<%­íW íËV­-1ŸÆEͳÖ`é3€ÕLåL¼Éì<3°üÃåêÊ« ŒÜ;î`ŒsHõÀu3Ík÷l‰ƒíâOßÁAœxË~.¤Š6ªÄo¿!€“, ÆÊ&»BòŒ$ȳúöˆÓ×d8ôüëªÞ>™[=’ùqv§ ÒÝg§= ˆîV?¯[í¼9dl½ž¶ôˆÔFíWQûýê Ĺý‘˜• EmÁ˜ñý·¾Öø@ïì§: ç0ÊEˆ€ílªNKVÞÁ»(û+ò‡Ã5EþÃâ¶é}¢Ò¿³f" 5Jûˆ#g/".¼—‡Sš)M ‹ Ž’1ï–lø­™* $M)w¯EÏ%@⇨è)Ò˜#§Ù ÿaÆ!µ¢Â¢ØK>§XjpxÙÓ3¹‘Üu:A`1Ž[Å¢@’P–•¾d)Ø„¶²ÄÊd^m½ (#ègxŠ@¨7HLˆô…kÅdÐîÛ¸Wv÷RHGÛiÙ“KøwâõŸøuepîÛ©M3£¯CL~Ð9{S È•m>$°Îi„§vËÃ@sF]ÊåšõÍOåŒÓ .@ùÀT"aä Œìœ€eß=Ä‹/ÙÖg„™:æq]^ƒ`*^j­'u1kø´5wâ&3´¿xž˜pÚD†8½±â.•›Ã–(•´RŸ—ì«5N¬8½çd(Ä{ÛŸekõÓIH«_xÁ¶É=xyw%8g­³€.bªc›î$8#ŽF™-&´\“L`C<¾ÆØo“´c ³ë‡ì%Òm³p\¤´¾¯—®Ñ›Ð76zÞž/{îàfw±_Ã4À³?—fØ\U‚qՊ㘕¦¹l‚m-sé&Z|\gS‚ÔøÈeÃÞÉ™‹ã>VVzYaÀc>-*áŽ;óÓ&õ×8ÊØÙòâdZ^tÆÛ@{è5KÉs˜móÿ¼˜- Ù‹Ûà,KIÿ) áÁ†ý}׆¸¦gj¤m³q¨ÛÏÝÓÒ)!!ŠL žú)Ë”ç=BhøáÜxY\ç”z쇵+'êòFŒÓôU6éêÛ@%ȉÒ5ãuÚŽ€.æ6Ö7c#„ŸÐü]ñ² ްݘÀt\—‘tÖüñi¥×Ü50`›Åޝ³†éPìFš1C-»&µ†ànâ]° œ[áÜßßôÙÿýʱ_¨p§öSÑŠÜŸÉ]†kÄ,JôIÆÑ‡D™¬†‚ÿH‰Ÿ,+š$ý¢Âk‡‚÷gÚx’—`îÄo1ÿb`øŸ4b»vSÆ6MÂn`r!š×UL®®ÜCT;Ê Ï6™h=²ëŒØÁžQbeÚŸëOÜhF…'˜I`VQ+¦w]²ÖF4¡¤8sÜiÙÞ)û+“Úá]ñË僋^h®®13Nð a&ÿéïXPnÝ3áLrea½¾Ôìö$Ë"eo­Þð“€Å/dzֲÿ'!ûÙ:™Â’ŒlPSÒW)“ìwO ­ÉMÑøtNO` 0¾e¾H?þ>ÒÞ•­L–£ŸU§×Í1ï—èîDÕ\æÐêDÕxÝ“>öAÈ’Áj°:5þ AWeaÙ=èíÍ6Í ^‡|.M¼ ÝdxvMIØ1æèÌ®Z@‚.á´8òŽMFìØ JlLG2.„¿œ¡ä’e“¦t¿]j<Ö¼­ Z©ï©æ&qIå½.™~ØÅxs'¦¼ý[ŸI˜@à*®ò™ý†s:ÌÎ^øB×9¡á!~ÃäQ‚Êþ„bŽE‚¥f{Å Ñ]?©Xrk“æFÊPajZû`›,z>òWxëo·}@ò-˜Òüéí?èפX®y&<_Túf¹½‚·®y!( ·±pJ£yß0%£%áo! V©Ðnm°-“ñ<Ä^Ì(I?€AüîEeü›©ÁùHÿ‚ÐD°I o€Ìr>ÄœƒÄ7ØÚ} 1ö$¸&ß9š6¶Å ڨ݉D‘X*ï~ŠD˓̼]Ç> Ñ1@­&˜à¿^f­PýIC^ÁÐæ’Ö½ÕB0jÄÂ{mµ#åçUoúþ&ÂçiVø¦]u'ÔH9Y+«õØœæ.x{ËÚ *Æ—3öÀÚuÍ“ñ^š]N©~¸ëzÖwÜ1j¤†S– qÊ2égg‹Æ8Ç( ~£º¾¬µ‰¹¾i8ƒ3ÔØ*3yê rçy¿¤&ô®œ#ƒÏiÝYò:–ëw( ÚËs®¤­ÌsÚ8 A M÷á+| p~rå—–çpáƒ?€ÈÊê)Å=\eT£¾7§öo;òëy|Žô\çnÔ0 ‹ý“j…çT‘‹DÊ[íÄì_¹ã‘èŽ2èñ-ºvn„|;¾®mkð7©ëH,®‘ wf²FCKRH0"lÀö ü6ÊYQ5D'Ï“>®º®ÖŠ˜¯Ô,l%+û»WR2M+Ë^Ïk.ØÆµÄÁ2Œ2=¥ ’ðéʰù½[mø¾G¹ŸåÒ¦™âÒ#ˆ-{X¨Û‹I™Àm&ÆR{í¨ŒSLrŠ 6šu'¾=¿ ½›7ö}ƒS‹Ã‹æ¬w½8(@"^fÊÂÂܡ■ž-61¤úwŠ£z×›´Ã`ùNzäÙáf¢“ÖT!ߨ8,%±†Ï"&ñ¾Îì.¼5‡èâªVÔðB¡ÈﱡÔl ù ¤kôd¹ 蟩?A»Eyiƒ0ÞÈÌ³Û rÞò¥ìv r®hà™:6`šOí—óÂe) )wÛ¯ñœ„zÈ5¶´"ޝbû£¥Yu€(ÁMiäV&> —ò*•Å 0ù‘Ú R NQEÐÖ‘6µÐßSëÌ ljÖf¹þ@y§x&óLYË$ Å-¬öl·ô5SšËî·Â¼?mÝÑ€Ý&Czõ¹Äžòüíšá=ajü„÷…”0]áȆû`¡qÍ»üš]Ë)ÒÌN@GH¹2±³i÷g…ÆØ¤ïǯΆüòó±{ûk‘k(Îȶ{âÄáàÀ*Xƒ€‹\ kb‹kûøº>þK}üªûøÛúøm}üº>þ¹>ºm÷è.Â,__N˜³ÿ¦>žêã?êãu}¼ªoëã¡Ûƒè÷¦>~UÅŽÝ!^tÜtgöo—åòx¤chó(਽ç @åñº>¾§Ç¯¼ªÏëãû%’Z Tù ‡Ù,n®ž;R †² lâ—H¥x]¥+„á^0xH¾g§ðVÐàüš;G¿õ/:ÖÎ!±DÎRØß¸¢ŽNôÃïðF÷==ŽØ|6®•÷‰KÜôK$hïË"ÇÇ^ ž¾m ‡òv_ã]´ç] xá¦4ÆPòñN<óò°ÞÇU3‰YÜ´suŽØÉQÜ/ÑÀèeü} c`Õ¨6Áª7œÎÇ+´0Bxæ6ž£ì¶K@Êv÷†7Þ¤êiDáCoG¢³sïǽuÑ=%Θ îG.x ˜ÂÜq¹,ù°¡½ŒÓÃä‰'® öû)¯;š¨ooPÏ̶" oõµæß/¹}GãwçòÐß½"rËsÃLìî%jYMˆ:ÈVV‰¢ü®L—NIA<¡5ªÅKÿ5ö%tèu}|ÙÕ¡`³¬ áŸêãõñyWÜvUÊÛîãuw0zÄ"?·šmäøØT>Óà ‡¢%%éj+îR!ɇFÕ'ðµ%˜CXæògs³-×RÄ8h>$ÖLx‘¡™ùe):`w\ér‘÷dLBX¢:ÍiuéZ…ûÊc»ª›~¹¾~É—ó!¢EM‘çx2ƒˆÚM÷–P2…Äœ¦²&Mî«oˆjp>Z<Ðd ÌÅÞ'¬Ã‘™M,.)ƒÁHWØÃ„ÿ7cê7+ŒgSqC˜€õž^¡@"£üÁ“©kÁô„×Ô:ƒã…dXJÚ]Tî@¢/®Ž´0kJ®¼§6@°Ñ·>;Ìâ ìO²ïˆ-2˜f0­Û§<>ÆÄHb*oŘÏë0¯Äë#‘ƒÝL›™”g5“®Ï*8Џ³6¢uE•c´ v1mÉ ¨¥`IºgÑml3%ß§ðìêezÉ#ÎIUÅ”õc<€­‘& iÚL÷)¾¨]Sò~çò8³>J fìMé,£TFìíNšÆÌ íñF¶Ï\¶pç؆a€æäªð.¶5ž¿f¢ €§å¢ _v ons‡O°ÅA¼wgÂwŽ’UÒ•‚Ikd± ¯ Èàó BsŽ÷¨‡Á.Ÿ²æqÁžBWSnèjÅ­ÛL:ʽæ‹Á^Gïø­#Ë%.XqN‘ _´¥Çö‚2 ¶ì8ÆçgJªo@µ²X¿ØÕöôw%ö‚ý}¿*‡â´S(€@k½ïÿÔ®´6v@*©ÁàL!Ó׃gÒ¸XU¢‘‘<,Œ¾gËN›jÜsœ²fãæBs§˜‘iè~ßK(èà¥F²S<¹œÒýn›‰kÉU& ÆTzÕp‰™©Pqc6ª¢ÊÕ1º-ü‹ùç ÿR÷*ü»Ä‘ÑÓLÍòòØ„¢­GçÚ(èöì¼ê{œ±ˆ»m4¤w\0 ÿd/‡{óL%öqÇ¢\wQ8\nÄ’ö2$­6K19 ³Š üBßýƒƒ_]AºÁˆéÓ»›ïŒz†E"@ûUÖD¢•ý'E½ÔKÚ^X ÏN2'²xôW㩬[2"¿Î€’t½º²,dp§‹+ça†]¹cÅtz²’Úà6•à•’ë™ÌÙˆüÒu£ ú6ü0 àš›`žšÌgøÔ·`›Æwé ¤îŒùa¹­þ½7M›º²´Øö¶—„·ô¡Kg~–`¯ó©M^_T;ôËÉîý[‡aשrεQEŠ„àÍÇÉ´HÂ̤ îw½ yæ°lðÏüN~VÛ¨R™_›_ÝŠý¥2(‚®)à*-jJ°æŠ5¼™ÁÚ 6è·€Ù.J±½Z•ÒŠãëڳؠ¼ífÜýQMžìç5ÒÈS€³ú# w0U©¢4mÛt¶~9•÷u›cj u¤«<ß•¦¶îÐj?¾î¹[7(…ó9Èà¨$tRâFàÝ•ïÆåZTör)š&¼£|øuÛ¥Œ¿nxu±ãeˆ´& ŽïH5%õ¸ä¡Ñ? °IÔµ˜íÓ–奃`Î*Ù„µ ù‰9(ݘðšœ÷QTu“GæÂHolò©£ã+]DCTA©¸×f]¿I#@´úpF—tÖ9é\¸¦îpSnMÁ­Lû•“xAáä샔]êúÀ¾Ÿ€ƒýb6mÁQ+uŒ•Ò®*†»æÏU>—¹½8¾LÜÙk kQŒé w7Åsð­ç+dã>Ìü¡ÅB4vZ*ÂSUdŸ¸~"uÜfñ-o®Ž’¤—ÓMBë~(ZHq¼î>{vˆ0Ú×|™4[\e¾bÞB#Ú@”ÒI`zí§e"9—bÇž ûë<=¿N©1ÈkßÛbWÒ1_Ÿ¸ >#ŠHµ¨Ð§Ère²4GS…hEÖ eF*+×æÞ‡²‡nUª‘¡NÑžÙG·•ƒT¾©5î°¿ÉrETtV€:ÉY7,¿íÓ(*¦¨õÿÉ÷ƒƒ§š…rÚÛëÅ*¸ÕŽŽ I7â×ÜÀO: p¿¼ò¢wiSôFjùXT»oðKÛ´­ëQC8)2? ÍRIsÇÇ÷·'O­O‡p˰<“<—wUÜih9ÃãF• 7 ÁYÏ%éKW²(¿Ýw£uÇ*!ωŒ@¢ñ._Rn‹,Íšù'ÄÔs6à°F`™M¼ßEn¬eÒ½R‰²Y}¢ÚlæD8™`´þ_7¢*•J“ߨÜÜeMýáÎÀ0:¯°ëÿºQç— ;Á¢AXÇ6³.µ'œ¨/ íƒý1+øTÍ ú!Èä¥7Q›·Y_ý’ÉÝåúƒbŒri9j µóµÌï× _*ÜSÞÌ òPò‡¥|ÉÙ„>’¤ÉÓOØ•Zð½RDlp¡)ôýÍ9JE„D·^2ÌLaãL•Iªë‘ŸùÉ'¼/òg\xréìW]*%^ž4¬ûšAüžK!ä1·Äšãk`OÃ0=³MDo ¼¡¬ýÍBïü–•¹ØÀG‘U]«†ùH¹_) ôÑY(<ô¾ð\WXÞ¨>ÕŸÃxÍ-BÞöúß>ð“¦ÜêmQ‰íO,!)JjoTtg-åÖ’à“U†@'1Bbš2Iþ‘\©®`×vè{]„,VQ»ü¦%¦ c,$µ[ãÿ~óÛF­5\tÈâÓ®°N)略½L>#È)€£¢PlVÙ˜Rdó¯ð’èHÅ7òwÄ„JÖwî7-ïŇ½àšûP*ø_ý?ä§Ðcendstream endobj 225 0 obj << /Filter /FlateDecode /Length 6598 >> stream xœÍ]I“äÆu¾7u±o¾8ꈲXpî i]¦2e‡iw(A[5;Åžnr†Ã!uñ_÷÷2ÀËPUèé1tP Èõmß÷^&¿ß‰^îý¯üÿÓ×WÿøŸÚï^¾½»—Wß_Éô×]ù¿§¯wÿ|7 =飈rwýâ**wAî¼õ}Ôvwýúª3ûë?ã]kªwƒê•QxÿúÙÕ×Ýo÷¢WBéºú)…v¡{¾?ˆ^Ç ­ë~Âc£ô¡û7z#Æ\÷ ½ácˆ/´Vø‡ïþ¸GëøL£ÑÛ½WÝ›ôª&toÓSÿþî67`cèÔþ¯ÿu6¿Ã0胶}p.ý;4f¤×Ø1õábŒª;¾Æs/…µÃ4¢ñ¶ûOE´RjY=ƒçZ(© F‡qè^‡ØÈc Á”/£ÅãWS—ÏóËFÊî¶z:¶L“uÒY¬!k‚Mœ¾Ó!¨`»—¹»ˆîV¶áǽòØùÒÞLõûƒ1ÿÐÝ5=-[ö*õ¡u<¶;N4t/¿IÛë¢É=(¡tÕož…w˜Ðgcoy“l¨KØ^iƒ=2½°RåMú{ú&!\Ì9¿ƒ„z‘¾q½Ýdk6¿ÿíþæiU÷lØS£—„BBb„ßU_ÿ&÷&e¥·X}"{+MØ)Õkgdþ&¤obtÖÓ7±2J³»þÃÕõ?|ÝýSýW{§mþúYý×Ù’˜Þy(¥Â²éjEØ*žY² ‚¶´ ¢Gª¾y‘6ÝK纻7y¯½ôÝÓIpHçЬ€êg¥ ÁB+IN¥^@¥n÷ʑĮèB¾èœS6’Α5ÐÒ';Œ¦!IÂÆ¬;ÖE»U‹0ŠÚÒ ´#s±(º·&Þo çBõËùãÞBâ¡PÕRüœ ªvÖÌU=51ªºmVs\“lÁä^Vª˜­ˆSZUk ¶a5C¸t5a}aOÊâÿ˜÷Wª¢c^²× &$¤ÜËÿtƒ-a¡qUSÑoÛ¾Ü4í{lz¿¥é3ûM †ÏseÛo§M[±“L‹øÏ7GÆÜr»ë’3ÚꤨEo^æQ4üØŠ—ë{Ä!%1(£ó’¼O­ …-ž*ÆŽ„6Eµ2š;Ã%iÄZ8'â°C?ŽcYÚ K–4ÊZ®"Ù/Jº¡p¡¬<ç¥Æ]¯',—,c5þ,Âü—å_“8ìYþ“®PÜXã.œ&~:åÆèì7y AèNî=L@ˆbÇÆ?ߤó)ûŒ~:ü´ù'{š~>½ðÇIh"iÕ¸–§£z• çÚÂ×ËG?Ý´«ªY›f®I;Ç—¿l]©ì•3Á#pÚùÓUÆôšîØи´£­,V;ŠÈËÛ€%q+‘ôÀèMØhM"£!l|ãro|Ó=É«.½\óBË^†Öé‹ë«¯®2,5¤ñ»7[±¨Ò²—ÎìœFD-Ò¯»ßQÁØ S\­á:„µõ)2éþ´/i9ÜŒÚñ¿}Í%÷6/7-Ë1i˜PA«<ÕüøIlàý…kvd\˜§û´\š¤{i1¶.€FL!Ä«È~kB °Ð>@³~:¸hGkwß¡gh‡¾Ÿ,S_ù€ö ½‡’6oп$=NC³zb†¡û¤Pæê“u<8«àIí ÏÉÝC“þ^”¤0ŒQ4’¤lRË‘xF8ð’·=‹êGH.3Ç'tÊX]9‘¦BM "êx×wI !aE§g¦ÿf2\oK1Tp`hȾXR ¨vÇÁÛÆôdê²fà•’bÎæ.÷‚Q¯´‚K!d†-“šÅ¯e&FT=¿ªÞ Ò˜a;Šƒåvþ.Ûd8ÂÈœÂÓz>ä[¢4 ôLz‚X¬aåÙ4¾™˜ƒîå+öÙàòØæ<‚› Òa‡Õíaé’Â%ýŠÑn÷IŠì€aw¨# ƒ% FjtWVN!)‹± í“ôOËþ,?G4ž8>-ðíŽOõT!Y;Ãc–,…?çV§wpmaÿÆSFl¬†÷¿d¨(/]‘Ù&æUóy ñs\äëutØh’¶ÿdò`ÒÑ(Ö²±–ãhðû.f‘˜´•‹©Â¸ü±â_˜a+ÈB¬ãÛ< \£±gw3MêWÜŒaåMî/H='‡FÑ ¦…±ÕT&½˜ÍOâKõ(:â1(|Èd¨èþ^Îxà>ÄK2F3›Dì÷4—üpÜBS…Õ«"Nmê¥&°Ê†y‹v3AÌ⠌†Ü•êÎË­½²œvfôñqúù:E4QÉ+¥Woø&Žº8hÖìšEôšÙqQ9–»—ÜŠŽ zLŽÁExæà¶ÉLÉ0(P¢Ù›)ÀP‰4I˜ç£6ÞC%ù~ŒÑ„Á+Í#»ƒ.¥Y÷Kc—KÐT9˜íP r÷w464š‚&.“5…UK*@¥%J½úæ,c;EøOFXõóIÔ~Äõ:¹;¨äâÂÉy=¾iÁ=}T}ÓZˆ3Îû$8!“2–ã“BëáÕE:%ÒÀÇ6ˆØQ>П穂†œNT^cæ0~¢¸Òç36ƒ/ ôg‰à6ƒ[Jdqí–†Uþ"ÓFªî¬¢•ƒ¶ˆiƒlèäⲈ~ %H{]Êð,‰ní°àíñzæ`s<ÅýÅÍäòØ:4ùOmCN æÖD#ÆL'Z–¨Ìr½(+5ç¹ü³qô&jê¿›ØÛ—µa-ÏhOEu¼ÁÆJÐÌ\*Ç›äç$µ’¹^þ˜"1Ģʚ¸ËÒý¹—-oÒÒ´1áRjm±8íYn›¸ßebsO¹GnqôêlU™ (ëd±e-·Å:Ô_cÊŽ¨Á7“ YHÂ|2é÷r}íJ¹\+çn»¿,°bDýPQ å퀭oK<7°Ö¡<¶Û¹ÉŸ¨Œ8=TçÅbà™pæ@4еz#âJ8PZÅÜÀÆ& •O™Mo‡Yµ7Ä4³Ïo0ȘYš ",•Và’k8+6oÓô¬m¨sI99I¦¥sé1ØUªê) “Î\ÒÉô³t.«„])°dÙáZ•G£Â2;ËEL®ŸjÞn¬ìÕHlÎTLõ~z{¹®;™ à×À'¸’‹My”¸ª˜f}ä+Í|MZtªbQ'‚XÚ¡ZÏ{×q( ÊIÉKr·’4±Š4󘬊ŽE—CÕËâ¦Ï¸}M«8¤Í'Ïšr„!ÂIvŸ’BZú_§~цt3`ݰ*P>²hÒšQÎ5ìúíÀ"†,ì Vn™C3ÀìžNJHjBÆyTû}V1ødžj ’T©¤Kî+ëPûE~J¶>«G|лªý\ÕŸ)Ì) ãE|\>f@H¨4(#ïe r‰%¯+ Ò¨x$X×â¥ÉQ¢ŸÃ\ ""V$9-мâ­Äƒx+Wezs&Š~Uõn'1£ÇpòL!+)'k] EcF÷ÄæåMÕÝ·b+1Û>n²<¤´5W)Ñ‘E¦²ûí>hÊTù*”È5©l­¶pIƒ¹¸cûò¼lKhRÕy¶²jzȺ&}Ï!&e&ÙwËC:|e¤œz•í8(M.l¯Ý¨€”ç¸UèØr« ‰çy¢ã\œã ô,ÎÉ#dLqý‘uª>'–=Ôp4Lm4¢ÁYz–Ž6AÒ¸™šê–¦dºiu¡™pTX0±÷$¸Þ¤ôöïÎ8ä2CCDȉò¥4R(5áGß¹˜Ìì$v~‚Is]S\±ÐøëÜŒ a>s/ÍT²ŒJ7E““ñz²«3,J˜SVbN2\h%N†Re›qY|ê¯#Þqÿ¼._«À ×˜êŒÊìç³$Ü\G¢ý8?cá躗½¯uŽßMÎ3#ñÆyñ|Z@(åV¶šké_öã50«×ë.ª)6NBN~À©‚û¿:2†Þ!T[BƉ%8/Zä·FJÛDKm–¬ææˆtæÑ7qù¦õµ á°C.¼pšèÃTœ5¯jhƒt=@f}=ÃCQ”Ûcø¿‚d*¹]2'Rë2Ûÿ‹—ÌrBý×g$1¿Ÿî¤>/™z€³üXÛãH愲O“tœHõDZü•mfF Ûlf5ªs’É/Ò?+™ëÑI¦p÷ánôt$æóVz›Rjª´¡xo}ÁEöK÷Ñ6 ¸P¤+ML—§aa.U“Åê¯GP’)®8o½³hXqyÀÊvb/þÆ{çYQqº¯ýò°bË:|ã= çTXQ_F7AIët]PþzV}5gí«–Xû1#qšµO|L—Xûư8™.¦»€´¯®wŽþ3>„ñ3e¿ž ôì–™`"6žÑD¼ta%!sê?Mrj&³D²''†ÿ×6“'2ܼes¯ ·ê•kUß³›cï^ ÿŠÍ9DÇeô$ñÀµß+ZÓ?͵©[}óÀ\»ëëƒZOá,Õå^œý(1:Œ-ì9wpKv~=-ŠÐí¢´(\<ý¢¬è˜7äÑÄxõÝpzó É2—*LÂp©Þå÷½EE0ϯÂý±Û%r[úýà2]Jø¨ åýÿ/ ¼mŠƒ6B®ç77DÒåÀÆ{„£*Š÷FyÞýûélþã¸ÿ±üàÂYû÷BËâþ‡ÿpÐcº½¿–êýie1Þ¿¸ós•›n”V´K›o”žÀІŠ-^yÈ€/yå +ðõGµÀ¹æYznÁŸô³óC%&Øé”^sj!.}uõÿ+·&xendstream endobj 226 0 obj << /Filter /FlateDecode /Length 5552 >> stream xœí&E¹Ü퉜¼R1ß w!6¾à]â~n ,_òy6-?ðǨŒd+A;Æ6Â& ¯¬¹}Óá&Tso·HŸ´’Ür˘è–ÛµØãnµÆÛŠpµIH†$ K°d>ù7h#\øØ'FfÀ­ËébÝá%~ÌZn_âw?èAåý40çò¦â"ÐjD®|•ì(áÿqµB–3@äÁ(f±ÚK,ÐÅtùqpÞ[‰øö’éç?¯Æ[t™W¢ ™9i¿Âœv•7x÷˜ºŸæpÍmJ>8ÐeÁÅj nù²•²õˆóZ™³I ÔT rµ1_’<;Ro0Œ¹WÁ‡VMlŠ"aq6q¡ìÆX¯I¥*TÄUiH|®¡õ°«6¸Þ¦ÿ2Ë'¹±ÈôGÞˆÀÅ©Á û€ÌfnÌ{¼¾âÏ>ï Îd ( Ÿø³Ú¼"ŧ“ Ëwû‹{ÀuÝç@è¦è‚û-ñߘ‚2¬6óŒû•QHù‡ mO§­ ´Ùt™ ¿ÇUYú°Î“h”[#ï8ÀÂ~UMÊá¦êqæ"‚<öI€_ó9Èí2§Óïxc´JR‘d¾Š*‘>«:¶(±^½ü¬Î˜3ˆ]MIgg¢ÜHƒŽ(íì’î‚d2D£` i)Iùœ+dO.¼˜J;ð‰óJ‚}(Wäf°ÉWsæ=ܸuÍ÷aiù¸dP±£—ŽL&Nñõ¦;|'È{¬ÊJÜ…P¼[‰¬PmâïxGiÐ%õþŒªK%¢týÊ®†54åx1ªÕnd«ërK8ÆD4?²f@±ŽVç1©7"ú7óêM2ݱƒQ)e‘¡’n€Ï3®Ôn+4|¢ñÎny(œ¡[Å5°f”&»£ƒ…9Žl â@}ýNÜÙ;ÿdÛpù‘6±Ú6@o÷Õœ‚G¨#¸Q`¯^VšÁð@xÊTÚf¼È á]"T'‹`;0oß±ÎæáºÜïqzyÙ$#wŽIWßTÞôŽÝÐ蓟p@‘i¹7Ÿ¨­·˜@ó±ç~3F2ëˆÌä%»QaÆcõø3CœU2Wù‘V=è€ßžŸýáŒc0( ·o畉¢ÈŠõµÉƒåƒ•ihÁ~-ž¡@ýSl’s±EùÛ\l!À2ÈIm‚š€üã÷Ž)yž‹4Ém’“Ç,Á zÄõgBîTØ?…|ÕB¾}G³XdÌ¢b݉gö ¨ªA“`/1ö¹ ˜Äƒ›0“ÖÝ“”ÖžÝD—̈øEÂä¯/ÈŠDþhkÁŠÌS,²¦±àϼð¨PRô™£àüs‘9&Ôvòd©ð ,¥•Úh=|U-uë[K 8}­€|Ze/òÒš°ÀaÌEòž‡‡Aõ™l×£wŒr+øó>&ÜmÛ1¹ =¹ダñoÜÏ·|LPO5ƒÓ<qêÄÄãÎãšW<,ùúÅ7%CQ ƒÐ_ž#†ÚÙÈ~²ÓàÃ8ÛÎXˆË x%Ûñèep+¡¸çÙ Àä¸ñ†à»±Dõd"ºÕhÀÐü'?ÓäóœUsHo/pܦ…P“ \4h°F ·Ý•“}?„ÖdyxW#öÝDŒ×ãέ7¶³_¤jXy`i¸õ§Þ9Ć ‚ä¡©onðPöšð ¥¶ÂÇšqëEä"4‡ÇR"ÍdoJ8É”K}×¥.Lâ¦dbä¦á4Â`°2¸aÿ<!M“Ô’1'ênRn_²ßÛªÍç𨠨9d6W!sfMÆdת²Òë ¶à”­+‘Õ°”`°à|®ÉDÝ !Çû Ž&fE¡ñ%Ž’¬xUý^bðbr~®@`˜Y•ÂT¨”©›ŠÑ>]cáòg »dªà™ÌðºÖè·á‡Óè§ùÜ%N‰šÍt{–ղǡ0Þö¾:"Œ0(É™ã `e6Sð”taÇÙSÞ&ù` }ê„|1´—YÈÑP´øÔG>™£ Gˆjé]é4ê^Ò#£2)k8ÛÔuín"(Y•;ž‹I¾F¨I>!û=DXKÉ-5 ß˜–¿%"€¡Hc~U“üàîâ]æ°$¼ŒQ+WM¼å”™H?m¾ECÝí>ŽÁKŠy\÷ÚdTÛËSž0ºÔì:btw’ŒçÀfšQÅ I° 2“q®¥™ì•°'åçÛü ¨ÿ²b S‡UTÞw¿’äñç#^M¹DôÚòÞ{ê6gö_o+ØW«­5ã6Ÿ V«.…à œ+éëIîOf°‘™,à ˜DüL03Mµiülήl±Ã-8e‡Fà Î{PBaªª,°4t˜3p—ã9OM#~dÄ- Iâ=ÆìCk”F*±F‹Ñ´Îé]Õh7ÂаÝK`”>»'†ñzfŒŒ°±m¶©Ç8 „øù4ñ³®ˆƒ¼µ‰œï<ÜÖáme±»:¼©Ã«§HSÀ¾>=Xõ9€ƒMQ”XF0˜a cµ³P9gÛ²“†ôuÕÉ­£ZMPa޳Xy 2·­I.a5J‘,“Qét7žHî1ÕÉœ4%ïl+ÅánU’–QÜ=2 ö$DRÅu»ŠIvÛ&–¦`2±«ôqàú6 ×6ïDã-¯Ce÷®ª°Ùà‹aЕFysÞ›FÓðœ)n‹ÔX¡ÌÔ6’­µ”ìY¨±’MzÙê„E…iRG†1ïT>OÓ”­r#Wê’ épq½Êξ®ÒãZ!`ý<9ús•£ë:ÜÕá¾+÷÷Ýá¬bÀ!DÃåáUî›pì¯ëðP‡û:\w7{_‡Û:¼­Ã»:¼y Û—ìP¡å»ëqQ‡oºvÝÓ¾êžöê1 ͈‰öÛ:¼®C1áó:œÑ¦³[näšWÕl¢þ[æUìÈ®Y¶JÝ_fºÂùHÁÌ´úo­kj,jƒ„#ÐÎÆb´Bw‰ŠÑ ÇàŠ·¥°èHÕhL ‚2¼EJ©¥BÃ=(Ãn¼ð_~ù¯89Ì(e¨rÌ”SÚË‘gþüÙÅ5ÍOä[r€æ)hƒ¥ 5ªð’aÕ©¾e—Ìp›eÀ›¸£R4õ\Ù½QTnX…qvþ§ÕfP‹èö¥DƒU§ +¸F‹S”î~UôÞ›V‡uÓå¢@nõœ»Ä9SNÀÈüXͤ^ñ!˜™×B–*vËí]Í,Ý!j’tBžŠ³Ýœ7MUc#|¦¤«¨n&k5=ôfå•%8=?¢˜µG? à›–LÇo'ñKëûʰ³õšõ*Æn„ÂÞÏA×y?iœ i¨·t&­=ÞxÛ†3öP‰HÍóÕešd¨þ qCMÉÖÓRÉ7Ë.oq›ãB«L®˜±jÝã:Ðfj V@(\“ “È>H 3^xTÌ|á(u7üE¬¼â•X‹ÆÿT[BÙ6EWÜÜ¥¶9ä’b5cRàÃÏ>0QLªLšPøÙSa6ÔePus×-©*ÂlÌpZ$R=~ȬÉg€Ï"Å÷8.L’Y« )U éûîGº}œK<ÒªI,q9Nmb(¸ïD\*ÜÆQÁmcŒâ¬CE±ï“OƒÜ€6›”y!è±×0Ws³¡V¥ÙÐÅk6äxû¸Ê–®€ðî,#;cFF¥²õ¯u¦@îA¬keÎÅßÕ%jN$¸-8ìêW'1I“³ uKIè~0î+ƒt3ZN&”¿ZQ[nŒôã4Œ:yþû&€ýém7HVf1Eb²7 ÇÉ?õ„§ý­AØ2a'ÒUÿméZS”‚°ÊÁX+*JCM=ŠjèE×ûy‚{÷Áÿî)?Ù:¨Eؤ0Ð\¾šgÍl@Kt‰»×6 OlÄ–‹>‚ T-F=Æ`ÁÏHÔæ<`.w®†£¼› <{Ph‡¹V5 EKv “6±°É?TÕÅPa‹sÁðÛ»Cl×Ã0íD®[6-û,aJ º¸eÿô weQˆ†b’l.£ç<ÞR3Á|Í|lÉš)3Lšëƒ®eÝ{žŠÞ¹tM…ÃÀ‘\inSC .0Fо^M¬!¿»ÉÞ&Žl¬Ñ3„4 ˜äY€ï¹ñ–¢Ž¥ú„f-ƒ(íå~ä; 8€q>_uóÒ•æ?£¯ãÁ„Oø"Ê.ž/Š8É™m^|%ùgªëMuÑ/KѶ`Fç³U)VNÒ³Ú»Ü_£>V­XÇÅ2Î ÷ºõØIi©“"ø4*oRÍg¦Q©2ð·qö¬âGrpÏðºS½“3;Ôàø6µìOGOŒüÔŸö _€d®†ühRþ3n4ø½7°{FóÓ;a7ï*|wSÕR·\áÊ7ÓÙÓ|0TÍ œ¼¹Î¾Ÿ0:}ÿrdj$ðyÎG<’,7sȨ{¥OüHáŠ$aÛúlîc>»)î·é9Gñšqõ˶vÓz]ÑZ;`M?}:rUã°lŸ‚Оn3ðäÚtØ/òÄ0*Ü´E2›—$G?‡€ zÁ Ø.t[“ñœ]€ƒµÙxŽ¿²’âCir¹Ô|±qœÐûy–ÝkæJëpö|ßÑúíIXÅží ï³[¨©F1›”)RÔ·Ÿ))ƒ?ïâ@ªt$žúä¤ ¸‚1¹ß‰QÀ’‰xJ¾›û( 2þ$Æ“# ›Iço÷£sîJKDPòö'¿¢9E ;mƒ.=åç½]-*ð‘Ÿ#Žúò©-Ë„ÖðP)0G"¢!û Ýü¬xCrÙäÄ!Š´~ŸÅŽ+ÊN7`»¢ÿeìî ¹ýï–ÈMö犡xs {&UÆ:êP‹uܲWRSÜÐÕ‡]´q‹~Ï—uø¶·Ý¯ú‹†ìWuØï_‡·]xÞw÷ÍxÎ#÷óßo”á·uø]}Æ®»Ciõ²}^œ|Q‡o¦NÛ@KOéWuøô6Ði‹ëuw‚ØWð”{˜£¾©8ºl-(&Øáu¸¯ûçúu]¿~]¿¾¨Ãÿ¨CÃ÷Ýû *d ñ‡:ÜWJÿµ7uøEþ¾¿› êÕ¡nèÛ£õ£òûéÔ¾û?Fí»j— ¢¯Zp{¿u[ðõ<í)ýÚ‡îfýô~×ù©Í_§Pê±{yVµ*Õߦžñ×:|ß½Ëî Ü˯=Z?­øÔ;øMw_Ý{Û½¢Ëîi7Ý}oãq&%uÏS‡‡:¼¬Ã«:¼©Ã?Öa~zÀö%f/¸¹g8᮳¢ñ—«üV™On+]˜!‚h“]ÖS $ž‰{û¶;÷¦2ÙWuØ÷LŽÝ}%÷Š,âZk½ìrƒ¸Ë¾‘UuÂð Þ>y©æ6nf5þ›*÷7u(žÛ»È ¼ê~×Ð÷u®»§_UÚÿÔÕ'ýA¯§§M%Š*Óó;Ó>Vë¡þ¢ ¨8qÝU|º7]ðûs_wOkli (¤\ÀÞSù”¨Ï¶;f,Ô¡¯Ã(‡Å HåY† ™ú’M¿2ÉÖ ¯uQIw¨Ã}žhŽ©ÎÚu'\ÖŸçúÃÙÿlX¯æendstream endobj 227 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 4972 >> stream xœXXTgÖ¾ãÀÜ "*z¢ÞÁÞ»5‰IÅŠÁŠŠeA@šR‘ÞgæÌ€C•^¥"c bM¢Æ£nš1¿Ñ¸›Mtõ\üÈ¿ÿ7€IþlvŸ<û\x(sæ+ç¼ç}ß3Æ¢#‘HX'—E›fÎ2ÿ:K&‡÷GHlnÒéj9‚é¿èí³äo–Ø‚l,š‡Û7Ù‰± ?z `¤IPX’SОˆ`_oŸPÇ ;':Μ?îÇY3fÌw\àì»Sèè¢ õñ P†Ò?ü×íôõ pœð†Ohèž×¦OŸ¦ ™ìýæÄ)Žá¾¡>Žk½B¼‚U^žŽÎA¡Ž«•^Ž='ÖóÃ)(`OX¨W°£K§Wp Ã0s. rÚ³eq°sÈÒаåªpå>U;]<½v¹zû¬õ]ç·Ác€[¿©ÓfΚ=ç/¯¾kcË0£˜w˜ÑŒ+3†YüΌeÖ1ã™õÌf³‘™Ä¸1›˜EÌfƉ™Êla3Ó˜%Œ3³”™É,c–3³™Ìfó*ãÂÌeV3ó˜@Æ–éÏØ1ƒ˜Á Ï a†2ØD&@Ò—ù M7cÁÄ1ßKÖHêûLîSÜ¥«¤UÒZ”Z¾akyK6TÁ l7›+ä>·ÚauÁz‹õ•¾³ûú~gÓÔOÞשּׂms×þ Ø9°ÏÀB» v ì¶ÙU 2È8¨kp6?˜wáS†pC>Õ¶À„3L⊉¸½s>ŸT¤9Á IIŠ º®çöI‘  R«S! ¸…ìþ=¨*¸ åšwmF›Ø0ð‚ œ´,ÐëÓÛqˆ=–Ê.“–S²z}4A=ü•Ý¡SXrOŒä‘#-–$ZfÛi &ÑÚ(A)Jp;J¤b5^â¿ó8Mú8¹ø¼í+h±Œ,H´T°d€§œ|ì¬`ËõŸ@5}^î,2_´_9Rò–@ªþ(„µíÀ‘¢®UÚß9‡/‡´0ám§u§–U,’H¦‘±dñÀ‰d"jÐY´EëÂd]T|<Äiå d áÝf ˜]´î²O“ß9@~È¿yòü•Û³OÁmÀ~ÊZ²XŸ¹À•ƒ¾BNKE6šp¬QÚ(yJ«e ·ÃÃd2ïŒëe•aÇ‚ßG<ÿ Ç¡0í>æº)Üo§ü KþR½®Ì-³ØÂ9Õ^Ð ÜÕK®ÄžŒuÚ´R¡(;éO·t‚âßéeÅ«R±«s0Ÿ™  ƒ+K•Е&‹#ñ–­²Lõ×38„î>\Ö5¦ëab,¤B‚Ch” b±¬ë-•²DbµæÑ82ŽÂî …]Â$§@húóY\(@v£ŠLÁþÄUN\Ê‹8‰½ûq«ñ(Þ„båæ³ÕgÛ²Ý3³wô`ùéÓ'ÈIÑÃxcÎF+!:.i?DpÆý••ÆÂ–6·j÷7|•ÑÝR™÷{tߦ٩/¡¢ºs(tÝ(tkè×)ó‰íq6-¼uç¾ë>=óÑ YepÚá>œî~qÀšsì÷™;É[±V)[”¸c²°kp´ gCå¢<Òk•ÙŠï$EÇZIÓGù‘Tì'¾ÅCFT²¢„Tu\RŠ:¨`—Á¸2 ÈQ8 ¬.E^•bLN‡£ GƒAɽ9T›Vt É%ZÝ~àT •–ÝIùz((÷9 Eàpæ¢Ã©½Æð"¹Gƒ¯aCöÊœwràwõ>”¢Ô°$X/èb2Í +…´r¹í Ž‚ÞÝ$Úõ ~ÚSÈã~þ3“,Dã~°–Ãj=½²GŽeAA;ƒÃcüã•°&ä¹Üô<ò>  <)üøxû¿G7²ÿ¹‹®}žßÞ"oð»’\.dÐV_N[5á[%’Ÿh?ôk•¢·8–Ç1&2×)eÚ î&¨¹Ý_’:väÐOn;KhWº±Ëýßkå´§-iþ¹¶_韉«ù—hØÄ¾µ}õ‚·S+Ï ø=KœÅI<;þZàóŠ[ð¥Iн&Ã…Kë’ÍrÑ‚µ_íE÷hš.¬îÊ£…Iæ­JúŒïž|Ytu Û¨ ­ôy =Ð1¯ð ¥TäM1Q´å3òÓ3/gb½5ð€= µÝ+°Ue)IZЦÄˉÉ&<6Xž3ÉB5KhP ¬½Ýa—Y\Üe—µOŸR©Ï*á(›õìÐé)öã3 Ó2.šwÔÄ€7Ý#¦÷ŒlÖiˆ¡;¤ÆËG‘ƒ„Å&Ë æÊ»Ò 6ÑslĦ%ˆIOHKÎШ“´ê·È{Â`±9>T³”Æû;ôÆãÔ®éIeàC1Yò ‹ìŸ‘‚¬0óŸefÈ(¢mJ6¢GÎ1Jh‰<Ìú¤À3üÃE‘ëˆuÊ\e}DUmMiSqRÑþÁ˜UÅÀ}Ôâ5_¾“%³ÈìD:¹ð럞n=V"ߊ…6Yi&”•&@´|•  œ#’óüÔ¹~[<Œ'¾E¦î‡­mù´†¨\¢CÎÁ¬Ì¯ÀÁÄ®£Uð¥éU@POŽŠKS’¨Vk4ò9'`?ËÖ—7€ ½17XœAÖRñ”4ì{z=ÔQši…êß§:¡]F¦ü1Z¢” ³T*dúk”Œj)³ôD-a{0ûŠ7´Q¦Ã³ xv®àKó?¶²…Å…Å%mÏÅ¥Ä,<»K5wÄÔd„ëÖ}~òîüáSgŠZék€ Z¾Ö†ì’û»o÷ƒ¥°ëTt…ÈÜ_Hšv=‡å¼R–äIÓº 8GJÇ&ÙÍ,Ø»€È °û¦=~Ñ a²Õ(^5IÊÌUKûó†¼ôÌ+f”ùjRÿ?Žu'!F­NÖªÍ8>@c­Ç!4‰A4Ñ®½8¾Â¢Kר”ØèÀ”XÕv?ç%Tgb!Π×ÕæBg ) ðQùn=¦|ïÎÙ[ÇJhÕ”"¨^6Q7°­¥¢Zñ¹q©ÕƦIûö,›Ni$à”ã—t8‡f\Ò¤hµZV®VÇDA0§lØ_YÚtðäÂ6ç%TÄX2ü»é8gÖ!—!ؾ8Ó›§ñ'ŧæMjqO’Äà d*Je¾4+×d¦ÕîMЄk#€ UÈÊõŸuëðç½)”5g$$-0]«ƒ2N<§c±²ëù8C|!8‚!û@Övö±×™S,«Ò?…ú<…ªî·;ž™ˆ/«‚HPq¿…‰ùö3¹=6«l/…mgÊwÔ8ÓÛ7›8’Ñ^E‡s'‹Ír7%4‘%’®¼rõ¦ªCdà ©Y*k’׿d(‡8aÜA›5;¹×Õà›=æ+á¡TÜ„{ùrTÖ&ÝÞvR®l]wp ]aÆ«c‰q|2'ãäãÏ ó£ 9R«ŽI–ï]±\evcµ8ì¼¼ÆBߢk.m,®?\f‚çElt›9[ü¾»31„ºV˧÷ÍR`¦}™‰ô}"{úa1G¯Õ„¨ø„ý âvÖG*kÌ?Öâݺx‘zN»Pq‹¼ø7RŸEWÏoÂös‚'q¤”ÊöEß”áëûú 2S¾&óå]Sz N¥ÿ94Òç9ÔôÔÅWħÃü¥Ëa.™.'rl§R‰l«ÇYo×cLƒ¤î†<3Ü‘â\ÁŸÚ{òÍ]zîÓÛc³å5é…zu µü1œ*oß¡CyÅ¥õJ…‡*B%xTxfm¢yÖ¿¹Ä?Û£ÕW¾_„enEù®ÒÈd—PX˹^]†«ðµ»go³¦&8Wp¯X ³éX±RtŠŒØ&Ê”™ºüœƒö5ðsàú‰pýÞ=pÞ¶ œçÈÏwþæÇ1n>>k¨'´÷>ßtå2äË_ fXýÁÌ0nËⳫúAö[:ºøßÏ _ÄU&m¾g|îhË€û)÷A^abZÌ>mRŒVž²Ù_+!éZâCnPTr{Ò-Â)I”ðÄ éB87m¾ëzíòºÜ¯R¼ˆm?â3¥Ò)å[v×yzøûzyÕù¶4ëZ²˜,§wé7ó:ߪáJ/ð\ÄÉfb5u²G%<À›¤˜!®â'”©?<®Íx©£ãÞV2´THS¶¸×¡+kv‹Ijm|‚ Ø\º¹y¥›ÑoÍ$’e[ó¢ågܪ“Ÿ†>ËÕ”©ÆÔÁ&N¡zƒXÌ&òZ—(h‹ -ü¥ctíqŒ:]V¶““u°²òâ¶;1æ©bƇϞýÝå‘TÊ_ #º›åGŠ_£‘G»Ÿr³s²®›Õq¯&œÊÁ^X¡=Üž[¡NLMN¥êHVq–¢‚ýWh˜“Qׯ+lAnþÁ²"ìGfØg$RZÔrñá-(Ø ý‡4´ ò µG6#!âtÑi©iqy”õþkÅÍGR{ÓÉC™Nw(½H_¢+¢vӬΠ¨°Á‚Þ3–³U ‡ uMªN a\×@–Œ·†X§T|å¢ Î{fvFŽ´¿¼Áñ›ÉjYúà6t¹K'<;¡ bG[~§É=7½Ñ}Ó¿Aøƒ„Òξԇùt2¼á`zö-à M2$Ò#.~™Æ2:^TÇ¥DÛe—ñã-ÿf’ùi&ÓÉß&@pwÔk~¥-0;½lÈ*ål_,øÍˆ¿£ç[Úù5mÙ‡žgˆÅ¸­Ký"…ȯ—毄Éàæ¢rçþôÌ/ùòÌå#ea þMÿþ‰Öv¸…¶'/êÄòEzÃ]³GØL]ÖZ*þ‹ ¤Gß@´4© ròF×Z²Tô·üΜi4þ0©7ì.ÛB2¢‹!Î4àï&Y f"­šLéMÑ]~æ ûÓÌÞ“úÌ C.•öèÔ¥4Dvÿ·¤gsÝT×+4Ôhy>•Š…´¦ Z·ÔTPSLƧíÉÜ qfM$#½ˆ•|ܯºò»:ã ôüãWXò1¯Çì£o*nNhS¨ÙV«…¸¸Ôh ²¡%QLÝT_\üP’‹íR¼Š‹ù‡¤}:kÆ¥»8FŠû>éfÏÝþFÿænÖ¡’C+Ÿf:V"Áh´”â£N–¯Ñ€Ÿ°p%‘š?KÚ§(–†vÃу\W‘+ØVý?à$}žõvÁšWAk¼‚m&Éyt@W,õèGÝ Z•䟘´€ž•#ÓeMõßþ5?¥_}Ôßshíxƒj‡å¤•S6š¢ÊëKO´ú—ú§ GO´§—÷ðØÛN 6:y(ä$€„Ç'ДF8ìç™·Œ=ڌھ‹ƒ[%Ÿat1 qÞ¥³Œvß# £ZžÖà.7eå×A-×Xã¡ TN¼ûãüÏ¿}|:ø;2½Jø¦êýàsî¶Óe"'ó]_ÛÜy¨æpñ‰òø†­BkËG`î>8¤n£Öx¯ï^—&V®MÕ$«! R¹ý(éI'uæÏéðS³ì-Á™|‡i’¸PV©MŽÐ@T¬@:ºÞ±Tà"#[£D•¨õš´IÔK‰A<¹A ç–7p N5þˆ–¹tQæ™'ÆibßXE=A\€k-uÅgŽÓ1åäÆj5/l‹Ü·Ž¶‡{ÎÖÂ$½Z¯.â÷ËÉ1Jáq%Ùz]v¦›×xìkše˜ú†Œ&ýw­Ê»à+o?x´¬qoÕ®€Ý[ç|>ûâ´Žû8mçÝ##vº'ú)äÝCb<‹ÖÕ’“È£7’â"ü_ÞOû úlYk[óUjÕÁ^©‘A°‡ó® ­?d,<öÞÎ÷G‘dÁ’· H>Æ£ýŸ¡#:NºEú†¯„Ww (!ïñ+!øljf´¾€,û·Ÿ¹ô^?^-ôÀgüm‘ûWl“v‹?»x‡Ižâ°¯p˜)ÜM>•~{|ü|ªw=\_Ý,`àÃAÆ>~AžŠÃÁõ-U‡ë…ÞN?smJ_ö:>×ó©dÌc–cÖÙGs!¼Vð ½Ü®ªcuyÓ¹íÇÑBOñ%ý„ÿØüKØJƒÖGNnÿÑëâ4¶›¤ñçÓ7±Qï ̇èÿ'ñúªÿ ‚òØ ³±$“è‰þ+:úgýÙ*òÕýÛ _Ök"ª„€„(uU™ûŠã8ÊJ±¹¢kî(ÉΕ‘m¬É%}k‹¹%6V Y66()·é§3 ÿ›¼O2endstream endobj 228 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 8620 >> stream xœµšt×¶÷Gk†f‚• 6-t½B(¡w0Õ¸àŠ{î¶\dKÚ’Ü{‘eËÙØlŠ6BI¡×@ prÆ9¾÷½3²MrßË}¹ß[ï[¬¥…Ê™9»ý÷oŸ±ˆêÞ‰D’…«ÖOš(üo?PÄêÆ¿+vÅúß”m VïR}|ôîdÙ‡~ÖÐ[ ½»ï4ö… ïÚî‹\Þ¢Ä"‘H€Àˆà]ž^¡ƒG¹Ž^²$t©|Ùîåa.+Âw®Œp]é¶Ú}ÇZO¯õ»6xoô±÷Ýä·¹ÏxÛ¨ ïOœ4yÊÔiC§Ï˜9köïpœkë0jôظÞãzYSÔPj 5‹F­¥fSéuÔÔ{ÔzjµIm¤FQöÔhj5†ÚL¥¶P ¨qÔVj!5žÚF-¢&PÛ©©÷©ÅÔDj 5‰ZJ-£¦PË©©Ô jµ’šN­¢fP«©™”?eM)©ª/õÕ²¡¤ÔÛK½C…Qý)[ÊŽ@ ¤QЦb)ŽÊ¦â¨]Tª'Õ›J¤ü¨>¢ž¢^Ô&Mª;¥ =ïæÕí’x›øf÷]ÝOYµª“ “dÓº‰ÙÌ÷XУ¼Ç¯=ýz^î¥î=ª÷ƒ>ëÑÖ }íú*ÞøVh¿nýtýx›6MÒ¡o÷y;“]ÇÖ¼ãöιþô¿o»Í6×®¯Ý­9§\60q`ÞÀƒ_òtðݱï^âFp7eβÛ~8$tˆyh¡W‡• ;=\2¼a8z/þ½{#¦ˆqh¤ËÈ=£Fûñ*ë¶0£­f~™ATÓ6_Ìm[É&ç©3¢ TÊäìÚþ­mÄ×h{ ãK—iêêÁ ­êͱÞÃÌôÖ`S…^›­ÕÉŽ þV$ÇñxmŠN )v~ ó¥ëõ_À>8G4MjaI½b2µ 6òÙ,²Â·­°\bÍ·‚™·j’mP·+ÈýJéaÔ ½`q)í¨€¥äBuÚ]%T{ïí¸÷R–¡•ôÏWO]¼c¿žÃ1ÿò·›!4{´†G`ö­)ò6¬ƒàà°Š‘Þú…ÆCV¾´ôð™IÛd¯´‰Ì¢ò¶õb~ÿ‹ÞE;Œ5{kžhíÌ´›ÆA ®°\õäê¾ôQ¡¯V')°ϵÅ,²Wë“3!ÕrEµŒ™öR-wð…UZ°ª„> %I%h%6Ûâ™89&Ìoǰ#.Ó6éj¡NB½Ê²ûÚÈJ(†,}jFŠ¯Ï”X·¥v†{ÄK_ŠëÚæ±¾åtW®vej7 šƒ84 ­Eëñ@4 Ï“á·ÿ>˜µ$+¿¯£^J>…+ò–E×fMƒ¹€Åák<·…®_7‚¤_ÞéÓ1¦0ƒÍÞ¨(Ô=þ­ó,^ò?û?-§Ý'ZI_TÅî¼ðÁ@Ì ƒßÁý^Œ@ ‰ÀÐxÐXÍè‘/‹‡Ðà ‰ñ tfÁê'ˆF½¾xpýnÓÔMD“æ*>gFµ{Ehд̈́ƚô&1t›í²óº¤ åã ¡ ðGÄEvixJ¿Èñ™.3¡â`É4…ïnJ’t”>+Aã¡rc!–1&I’[µŸëªwk~`—,¿¾üúQé-ôúŠEÝh,òÈ¢–Òÿ‹ô~m³® @3°ÔÄnz¾:§^†&9xr3C§Ò„ÅAPm,6åÖÔøV;sÒ[ ÀsiŒ3)äÁÄÌf3:ÒaæFFRæNÛ¶ýÈ*}S„uúÛ¤á›Î"ô¦ÇA^ iÑßew<#Þ{,7 ¥wÜ[‚Æ€iC!æˆÁÖ<§0òÓKljMÁŸ Sè'ý¥OÚÄSîGÁBf‹—ëÜž§¿ çTyšÔx`b!)B†méHHÊIÓjKJ8tÅ%‡\kJIîÒG>ûò¸ßž¸\™w[¦K&é/r–æPWÑøuK›ä«ã´ YDr˜© ªR~Cüc’› (ØÌ¿gÑ—éO¼êúK£ù§Ø<ù×`² ³@†ÐûSÓ.9óÐlTû‘n5<-ݪ†~ªWxÈVÐÒfÜK–ãmu¨…(ÎÖÚs¿~3 üǧÆþÄ(°Ã«ñt<»`4OAëeÒèkp·bïYR+¶ i@f4Ó úù @²t>?œEÃÌxZ,ÑÌò<:…‘…«iÜ÷y\ýÕª«¹ÓÁi‚«Ká›"®L>5£½fK´>ómüOl½.çWn–x©æ“à—Ū΢DZ¸ÅÒñ«ß¥äÏuÒ5÷Á ¨³äOý¡%a P+.ø«_©UåM¤@ÞPK8?ð;‰FÃs¬ÌôljGIy Dã‘oÊ/¦§n_²`FJs+‡¾êüæ6¨&Ëðz.LDã8iø)8TUý)cþÞ©oƒžˆ‘\°_Ÿs‡«,›Cìrƒ9–UÐwô–©s¸ aÏwÈŽ÷ÃÎ=ÑsT–=—+JùfQ #æ§¢GlVYÑÁË¡„ªý5¡;t‘²A“ á NIQ¼‡3mqwT£Ì#‰«³«ØÅ$¹ü4Ô¡®Z?K¯¤/@ª2Ï ÇÈ6=F—œyš•šM²"[n<ƒºó»Í¢FÔ~$F'Ñ(öý­6)—ÀO®5wOE}€!@¾+ÆiþÕ5O~ºöÅ™6“Ün/ï¥óá¹£“Ú]ã@ºóp·^O±^D=:Ì*(ö?жø˜9¯N0KNÌ’ƒ?8êÂ;̪×#È{%¨“c'â,ÛQ¨&%W#@å~0{øk¶¨CÀü»¬j€Buvlš2Ç/#|N¶‡ ɹdc©]+:ü›´ž+ŽA®*×Ùà_mÓbô‰‚téiù?¢zÛŸpCªBøÈ.ô¾Q)Œ‘fäc¬x:éÙ£MÓ­íÏÐŒ+×8øGn²©®•.ÍÀ”žÏ¿';¼^¬Tx§NãMo¿gƒÛþ­…Û™µØa™Ÿ!²¢ªØP‘›\»C+«¬;šC8èèi·É2ZzðøÕǪ•~í Þ NÌìÁ8r“Ëpä@I3“„f²#éq w9nÛYwìdÓ4+”bñïIÈçg°h)šŸ™y¼åž@ÃA*’ž°vuÑpz($@T||b¾Œ[mÑÕÿƒ‚\H#†LËpizBV,ÄÙù¹t°W˜/ê`_'öÒö°;Õïв…ÖeåòZ1:€âØ=ûö-¯0×74 i¤öÓ‘ oÒÅv¤E©:3–L=Qñ±ÉÊUsmçþ˜ `aºd,+4ËÔQ°´Aþ´šæ5?`Öv<îµ}ÝZõnß+tfmTC“º¼ƒäú¤<(cuíù/oŽ´ÍTäE‘äKHˆ¸}ö”ÎH°a¿fO—ºÐjÆ".$§‡‘GSÆEŒëâ/:9Î"(íÏM´V›™_ÙÌH†4m½>° ÖFÿ€%D™ç‚³CTµBí¤¡´9…ûOÌ,‡ÃPTî^±S¿v‚‡z¿ãn_wo{pÏòÂoz(&=ÇŠh±«ç:´ø+¡!£6X¢x/†àû"’wœÁï¡÷Bþ·\­Y⩚H.éFHê_Dš/–Xåb,~Èˉ=óSHÙV\hÑÂ5м#>fMa$ãâ5êä¸Xc‹ÅȘ’­N%UXÑ%\G€"ÿ¨FŸ[ó½Ð˜v‘mrlŒ·2V¾1`ñ H&ÃZx±95Ó…LExah˜<ÖϱÉóØ…ÆÓ§Ë9k~-é`µÏLém~cƥȟE3î7Ϻ ñ¬ãvª”~ˇUUŠö|>oÿ‡¸ßû˜ÂoqÒ_ðÛ?Œ"é: õÎÎŽQP¥šóŸ°<Î)¿iÊA4Mfýۨκ\Å—ß#eÛh6Á¨"›cðÜ7¬£¢Áa¸e¨=JÆÂZ8Úŗá¹RÆ×ki¤ü{·´Ø4’¯v9 ÏLÍAš¶>¶Úö«’®ªjÐn™U>ƒ†?V_ÞU™x®$@“ÚÏÙ€‰Û(j2¡OI´ðï±Y9‚à1y‰Áák„اbOÿ©¤u#Ð#‘àšö> PC²]L z@€} ¡ÑÜs·fˆQ=%]ÓÎ’ÔÍ¿u’]F£‘¿÷Ì£üÏLqìKû×;WØ ×:³Xúý($döE8¼×TàÒîØ…&É¢VDúmw”; \²/¨QušÀaƒî\õáRó¾ÚD(š¢ÍŽ9‘dŸ˜Uñ³ªÄ(ºm2›«Ñ†“}è[-)ƒ¦@S¦Òj œi/ù‹HÑiÉZMª’×þÂ6+N«É&2JdüQºš_‘ õÒlS{ƒ7lÓyu$õ+p”µÏ£;rί-0Ú99ár°%麣46~hÚ Ì«l<ÐM†v³¨FbÔCéö/Hõ1ïÇýd$çü2† Ö‡‘¸$;Q—œ V&ª9¯á“ ¶‚ëžÐ¿Oà2Ô«‹åm3*D„^‰ÛÜÑ39ã±d8]Ž—àñx*ÞI°nžˆ£åh,š‚9üþš‹|òQ.ñùƒÇhØ"œ…‹ñ»3Þ—Y£3±ESfE¨ûågÅh!¿E ýõ'ŸÒ@R— ÁLhaXeu¡¼Á£vÓÇÓ7 á0=Ç÷¾õW3Ä $1Ócо tAúž æWb†®°h®õFÔ½—ßú¿+ïÿpdñ­E…ŠXTÖ1>‘ )0d›Bž"œ\àehý™‡eÊlªZÃn 7êZÎô—þª@Áè#v*¼,«ÔÖ˜d™¹eU€y #C5>ÑÁ²Ä¸ÐW’HŒ ©PVdØШ*æù•+·¢‚Ëdõ{ëÓŠ…£Ðk)É ˆc"óâ ³K2ÊŠâjÜœ”;]8—Z½˜ ‹èdt7í–ÅFGxƒ#m£À7Ïݶ&ÒÇ\™…?lDÖ¨çÏ-·ëbZ¶Tq›ªÖÁj¢—ΤõIÝmR&ÃñvÄâ–ãGö5ÒËÎJ¡î0Í~ÑÖѲŽ8KYŠ:XVŒºÝdÿÿ½Í§5‘î zõÇ:·n •I|Q«Ù† õHƒ¬æ½ú®¢¿ô7Š^³(—†2È5ê³uÕ`æ)êž‚ß[2’½Œe_”ß>·™ï°ä!ÉáÏþ$ þë©M_¤#š–þ½ÐTÝ®}žà Ka9¸6º7zžTîævÑýªtȉU@B’F¦\ä Û ±%æyì]Ûˆ‹Ûj7–Zßc™–QŸ×67^|¹¿ô9ú˜_ÍâÁ‚—²”DÈŒ]nªÑžÐ•9ªërÓ²”´ý2é©ü>¶&¸4ÀWh 2W™JkHGj•›Û&VÙœªºˆ>¿¸‰8äú‰·eñÛæUn'àŒÝ'ÎßFñLÇtN«eNçx‰vXæ·$&.‰óZïW½³q2™ßÄ3¦ ™y|Áí@YNJKÜ‘ŒôyeÒž¤*Ò l¢PÌ¢­ gøÏK;º‰[wBuQS¯ÉNU×ðë`~3´ºœ ´z}íqó®‹žOHèq÷{ÔM&½/—ÝŸa .9C:|á×|ˆÅ+®ÏZY~Y¦‡à—q‚_ne 3˜vL˜¶"9ò‚¸‚TC>T2†€€ˆ° Í§Nßøôâ7¤¯¶Mè^Rîïâï_R]]^^MðÔƒx}V)šß ª7¡Ì'(Ó$n›Ð6‹mG’p˹R>šVP‹ÆB1IËlui ( ”!9± b†r®¾’FýOp„ü{–œp¥‡Bþ^¡;’Ë/NÍ! ]äÞùxšåР.QP“èõeñë7…RK´«öÀÍõï…òоxòÖ¦ a™N~W¥=«+%ŒX­©îøÈÁý8–¢%¶]r:á®Ô«ç²3/´4éJöôWÅB¡ìUº˜ŽÆQ¢ÉV@,„ÅÄÊ‚­mùЮ¤’1ý;€­Í,[‹‡Žù`œîiBë9¾ß_ByA &—@ôÓçHRÄé t®…ÃB?óÔlWï/Ø®óüC?³B›ˆ(‰ø²}ܶœý#®Ñoœu\gi£ Paç¥mC¶m·‰øûl~CeõçÚΣ\Ùßå¶jJƒ.%ª”j%^ѾÛoàuÊ|Ë„Z}L–ÓÜÅÄ´íàܵÃKº£èößl3ÂÓTyPº´´|Ò¯ÿó#Ë<ôÿò)Äÿ•jÆ<žŸãøæéŽ7O/Žv„þ·Üÿ"þmq`ñéƒÙû¦öSûjÈX +ºø™ÆàM&|¥*‘8pLûN[<•Õ¤ªÓÉ0¼ç TX¸Ä’κkNC)äz?¶Å²öùx¯Ðè éèÿåïo@jr®rho³Õ'èãó€Ä'#5ÍâÓlÑìöôŽí:>e:Ÿ™QŠA!œ:"O¢/x)?œýþù;I3C±hÃ0–übüùzt‹Fžf{#+Éó›G®gUBB—A‘]¤$zö­9#ÊAÅbTÖ°gpñ:ݰ“‰?!Ÿ+ù¤q6ÈÙÅÛ$'4Õ¡à˸;GNà|å‡$©„¾›s³j‹Iû9èQº#C˜ü&àSÄ| Êdí)rZ¨ž×¢G¯Åhãqö¿Ëa}éo ¢S|1¿¼ÍMË#r¡cr²1)ññ*ÿí ’c‰\«íY 9ù©ÙÙza¡ÜØfkí€2ˆygt…»Éw½®ïønV¡ ¬A;'ø,Nž ÀG©“çœ}-ì(œ€›¥‡¾­¹’vî0Ø_cÝ`­)â;Å9xgá"\ÎæÑœùèå¼z`OÊ'Å©×9Éü¶ïRk5!ê$HÔ$i ž‘¶+PÄ’ÜoX2ZìïäüQÑY/Y]²Ùû˜ú@ƒˆOŒÓÄçËYóÍO¹5Ü_y{Xi‡ƒÑ3ïB<Ü|‹xÍb§ Ð3ŸÛ-4¸V6Mp¼³Áu"õã—æ‘üvâûô¨•p¼‹Ïµï"B‚L¤ T¤â{XYLGݾG+¨ÿ›c?l7£ð'„y{´ƒE½f½ÄÔvç(/Om¦ ÊÚ¤B}¿ø&/]•–¤Qƒ*Y–°-l)¸ÂŽ ‡ÒD½šxš‰‡Ä(>JG@b~†NŸžÊå6œ¸­`Ú‘Qè¡sƒ­Œô7ÒYÖ„¸„¸yí r·þxäY=B6µE¦òòH“wŒWÒŽiWÆ î2鯈ùé;$„—•îF­C^ÛÔ¢Þrãu᥿ôh<ÿˆ=rØÙ+"0(¨(°ÞŸNXI§ÕjÑALŠOÒ•«dqqd‡*&)-%-çÁMÔ“C›Þùß,³xŽZöà›$až"·gËŸö—¶#[t-§ p¾UßRÞ¤Üeä¢]ÁÑ·<ÚX¸'kÿ9§Ofâ·ñ@2äɤ¿ab=GâoÿFÖI'‡Åœ´} xl ÛΠ^ø»üÎ) "šá.3·k^>-<«Ì\±l!€º¶Fzú­ÙAÀÊ’Ä¥ü¥[¢S¦oŸ‰ù@þ5›%b–28E‚­A—RÂK¼ñN‚Lú¬ y¹é·!ÄÓU¢ˆšƒErïIx 7–Á{$€ŽGŽÀál®…†“ú“™'RËòú´£RÐg§Q±ùXè!Tq(ô˜ðøq¿ƒÝ±'e/1ç/5]½vqóÂõnÛ×xqÆhöˆ£­p™y6ùôÄ‘³çOˆÐ969p9q Á&‚¥¯ÖùOÛ0jÀŸ–þ‚¬¾}ðk½â°{^•½â3Ò…E}çV…7ûÁj»ùmž5má§ÎÔ\üª¹ÃýõÏ–ñÃ÷/†Üž i+FWYÇÛ±.Ò^·7;”Y]d)Gù%¸-=±ó+Ä’9jò!ÞGFL>yã¶Ggn§c¨ |Äà~_FÝ9é¯ç¡±©ì ƒßú‚à{!¢Ž —Í¥ •ÇöåÕ¢¬ÝUä˜ë“ºv03w™ÂY´tÊ;oýßõßþÄE{°})úHXióèu8Éæÿ|µôyx»˜%¤8wÅü¤ÄßòË»ˆ©V_·ïÊá4hµ\†6G—™!Ï#ÂF0"½ñ±QtêS´Þ„f ¯â¶ù¤Á¤géôni0ñÉIIJnë’¥ àcXeÚØ^´»€Hø¼E3pÜçËÙ7®~‚ºçCóîVÙ%cÉõpZÌé©Zm»ýß'%ÄB²Ð—²²³u©9™œ‰ŸÀ†-ËC³Ñ—òJ·ÌŠ©Í 7Ëø5í#ÿô Ÿ%,¢×>^ô_î)œTËÍü@3*4Šš컀šÐ¹*qÛJÔ‡UÍYáíö°¹Ågß‘ã>?rgÿåÒº'Vz¨í5J 3ºÎu˜Âƒ‰¡•lãÇ%Λb ¿ë„èç>¿IõHÔƒE«æoâ’¾°n27qŒãŽ @°·›zå_=XÛÀ]Ys6êb÷…ÓºGL»;jdïªÎ€™yx ¦î S/›=5$ˆsþ€5:ø\UP¿¹ÁcÅúUIÀÌ }qâÞÞú­)k(½×*"Ç Fz¤àAÅDêý^q±±Ù@ˆNµ·-DwÙ‚0Â!FÖþ$&ÂÂJ @Æ‹7±ùßDËÚÿ&‰îø&ßò×F…®¢3Jˆ«úËoo8Üè/u¢Ð~ûÄtè<œan,ülÔ‡“·‘©ºß\ßÛ¿£T‰ö².rE ­…?I#ÉÝCãIß)+¼VONä4I ´·¡Å&fiP¯øDm†1˜]î힪+Šàrt%ÙPΣòÃ"Ãã‚]<ÏõŸOµ¦³endstream endobj 229 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 3098 >> stream xœeVyTWÖ¯¶¡ªâÒ*±ªÝ ÔÑDQHˆEÅ…VA ‚Ⱦ‰"4`cwßn@Ä…]EA»Ã"K©Q‘,Æ1'føŒùb’c‚5:N21·ð1ß|’Lþ˜óÎé×ݧê¾ûûÝßᆱ`F0 …‚[¾Ê“÷ÐWy¢Bv!OR }aXí8‰ë¿tÒ\qI• F+a´C«Û˜ 'ìyc1j£T(Ó,OLÊJ‰Ý“¦õˆôÔú,\¸ÀK;×Û{¡vY|tJldD‚vUDZLt|Dý±[».126:-KëñzLZZÒ¢W^ÉÌÌœŸ:'1eçO/mflZŒ6$:5:%#:J˜˜¦]­ý5Ñ9¿nËã“ÒÓ¢S´«£¢S†™¸lyÒŠ€”Ô´7ƒ#vD®Š^7Ûgóæ/øóÂE 3…™Æ¬e¦3îÌ:f=³™ÉÌbü™åÌ fÈø2*ƉQ3/3æOŒ33qa\™‰Œ¥ˆq`²¢ÈQÜá3âˆr™Ò üÐa¢ÃYG­£žue ¸y\0WÆsü9þß/Õœ7²zÔ”Q¶Q“ªÐ]’ƒkí§•r€¯F*öB",8E,ƒŸ`È6øŸ¡c[,­MÐ ×á]ãåÑS%Îä •–¢r‹E¼Œc±–ý¸;zÑ‹C´Áw`~p&GÉqdI“#Ùê>ÓKò IqgËÆKÊÒM5”eå›L „7׬ºüFU¸+™M@]¼‘iKCwg¥‰ê'—8Pà¨ãn”B´8h Ôq–­-”·Læ!:9Õ Gä~I3ÐMvÆJœ1 Ö öF'/çp3Τ@wa™‰²Eô×±hì×ÈýèÎÝ‚¯÷\ °m:s€(rCã¶…½ºo5𪃴¢6 ˯…ž‚ó$åÀN<­Ñ±éDHœKVO¼ÐÖÃþP¾L”ð¼Ž]–>SXåèÅâ<8圌á%v7ÿ¥Ð!3ø5ÍiÀ‰UÉñ4t%N]ßC)cž£ÊYÝ…Ûñ9²è7ÄQ"¹AÛ,]Önh›`€/.çÔwoŸDÕ3ÁlÎÚ üŽæÌ3gškÛ®n8³QPw@äæ´•¼JÞl°Ë›Z.D|†¹ôÃY}w¢¤y§4¸$þÆC¨¾Ó’°¾D°d[ ko€’zñKnH>M°'JH®Œ9 |<¡Ü8GÖ$ÕìÛbZ w÷óê~»áNÎ×¥¯@úä侎\¡ðh!ì> i"Qp{ ¿ªÔ Çj Ô+²vÇ]à›p {G¦-¦QTß:whq%Ít¡ ÿ-¡_â'*¡‘’ò'y²†œæ´ß¦öyWoŸpU·‰ ŠKM\ 7둸K×Õ— _ÚÐ?=6çìGdº¨Îö‡Í ©™ü!Ò£Ábî´íâ™j`1:XŸ%&ìØªƒˆº–wŒ–÷°¡MÛ¦hÁ—ÑÇ+qPž¥ÉÚgÊ€|þwx÷ h- ZF8—G^8çŸE¾ôè~ÈË2›r …ä•a{6Qω€3º¯þ 5¶Fc¥¨zñùo¥÷¸*ÿ‹Æ—;p‚†OéOê!6¿”!ó}ÈBºÙœhL5šhOæS†zk/m%àÊo"Há²`¯%ËšPd¦¢åeÉÂáÉÁ§e™G +À¥+.Ãf™L° wÖ^ôÚézô›5 C{ #‘d.ô O¡È¨°WžÃEvE3_ðX)ïÀM]9œn2ôm¾,¾c­ ¤ |}§‘ñÄõÙLœ‡s;žŸ¨Ú…{M…û bFXH\0}b²]ÞÏ9X[-jÏjk;rð%°gK mVNôø¿§›]Ûžà§O–t9«ûdwy®†LoŠ,[w>réh¼ñìËÄ·z¡8L¿ûß6|Ùœ§¶m_Û\F)óZ>kJð¹­µéâ¥-ç "•W?üŽJ°.ãXöÉdØÊ‡%Nõ#^5èY ˜« $ýwï{¿Ìj-=,twßÚòÐÐ.¸ð:þCT÷*Ýó®¦Ô¨Ñÿ‰¢ßWâgè¯yBÞŸI-øWZIƒÔY£À\¥Ä_”š:(OÖ›à@¾°$ddŸ©«aCåƒj~ð(u\gÑs¸D×sèüµÚáHÔj½8¹ERôâ4 DA)ÅÚ¾™ùoí? 7,…}”׬Ôz¥£ë²ßÜ”à1ÜÔÿ!“ɘ٫|Ö\Ìnhn«»|%âxv±p¦¡¥Âüýî ¥ÑA+D²•Dää‚ 2]²äeCGê;šiãplý+NjiuºH¯Ú1Ô5N¢ûlg5aðgüYóim_/ÜçïúÝ çŒZüÖkáMI-]'›ß}CÅA‹`kì<Üü×EoíÊ1ç©q‡Éd.4ç™My`àÕƒú¬r¨:YT€ç&LŸš¼±øl˜x®ìäIhç»âš¶GÄ$mòyò6j0¨ïÙ÷=zîS)ðà £B’(Å}C^Y‹Þš'ÒLy!{ÚT’˜a6gç äÇÁ7uØNÄAǵ}?<‚|ûÇÂÒ—°ø±ã0îÐ §Ùà8§^týóUzßNqVÿB ߪAîõ‡Ä#!Ô§p×XXa:O»…ûßÐñd©©"×h†ƒz1$=Ô~Z\–Oýg~?è÷‰¤‹š2§®ÒZ|¸Thí¾VNG…~xõ5QýÂfÑ©+Kz#»õLÅ~5üXjöÆüÕó¿˜ãõ/èûôªÄá ñ'ûEwj¨,®ÆñZtsVâJ|¬©æÙÚŽÒc Ò5ûGPçõi; ³’ ž­Olmm8Õr#âãÉÄ,œGJõ x½ìµö@ÉïóÔÛŠøÃWô¶Ðº÷‘1¢z0ÖA„ÿ3ùX³’Z Žïk†>¨âqü5©ç1,áËÜ…¡œjåC8^qSúj¨]òCMyXÁÆ“p6 ÈŒ’ýÅÅæ2S9¯ccH©£Ä–^YYl­8tNÐc³ëƒg‡åd"—€ëpè¸TrKèæàbÑÅR©ì“Úlíü0xùÅ}í1**59:è½TÒžïó –Ñ|x„ä‘)‹èØ*ü})Ž¥å¶TÚr›'Ê/¸óì)|ùËõ¡yÅâ‘]°bù•ÉÓ*¤Á' ·âz+1O¡éˆ±…‡ÇÄ„‡Ûb::l¶´;ü×*}…¼¦ ·Õ–W°dk)'Dv”0ÒaAÍè—ÚJGF¶qôËhÃü?qZ…7endstream endobj 230 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 7447 >> stream xœ­yx”Õ¶ö #Ãbâ@ðûhR¢""H1ÒÄÐC3½×I$“i™>kf2=½Lz&…„’Pƒ˜š€¨(   êi–³'îü÷¹{8zÎñœ_ï½Ï0žãjœ6G0Ñ3uªzE<Æâ°ÙÉÊõÉ‚ìÔØè˜ôéóÂçO_²bÅòÓ—.^¼búÚÄÈÔØðФé[CÓc"CÓÉ/ Ów&‡ÇF¦gOŸ÷RLzºàÅgž …‹BÓ%§F¯ž¿`º06=fúŽÈ´ÈÔÌȈé¯&'¥O=41rúÈF|[Ÿœ(ÈHL¾59"25‰Åb½´6i]òzÁ+þ©¯¦mHÏØ”¹Yº%+ìµìð­9¯GDm‹ŽÙ»3nWüî„=‰ E‹/Yúìó/¬xqÎø³X3Y¬Y¬m¬Ù¬í¬'YsX;YsY»XóX»Y{XO±YO³ö²Ö±°ö±Ö³²ö³^a-b`ù³ža½ÊZÌÚÀZÂÚÈÚÄZÆÚÌz–µ…õë5Öó¬­¬å¬×Y>¬dÖ£¬ ¬‰,_5™ÅgMaMea%²ac­"Y`=Är±w²¿µiTGÄùãC—Goâúrcöy›Ê¦þ:6nì—ãšÞ:Þÿ‘Q˜}¦ù¼ýè’Ç~ìÍ 9ŸX:ûZŸùøY^ö¤ÀIßN>îçÏÇÿlŠaÊM5LË›Ö3íãiÿõDÔŸÒUL$óÝtñ îŒÍ3×Ìò›u`v¼Gë3(7Úïöl¬`÷>ÂA»Ðh^ÈË!šL¯Ï=¥5ä9éh0Õ‚š´prü,s!”; 00*C´‘¼o7ØK¯£‰~(‹ëÄ/ަICdRÈ’€‰nƒS;´Â!m‹w±¥Ê  ÔÚjuLÈ WïÒ›DÀF©’Q ˜Ã«Â·FãH®§_æò,t±QôÚ:ÀA9h=Ïì²Tš]…§ÀPÚPTÝýþ  ºŒA™[ƒñÃâpF«ÍƒL*¸>úøW-h¾Ù¡¥t¹Z!ÌßT¦¾¸¨ÆQk«f*z;ÑCÐG}´ËµÅ€, ­<ÛÑ L%f…K¶Ž’;À\h0¸ÌtÙµòn j!I&ÒfiDÌÜ«SéÕ åËl¹¥Ör+í3¸“„ó² vOôÜýCÂg“}¿òðó‹ ?,¹Þº…]9éšxã®ò,šøÖÁw'ûþ„íãuɻ尛Ьœ±Ý|JDŸ2Ã0Rå’ÂÌô䜠×û¢ßAOµ#ŸžOW•…Ùiß“-Ѷ£Sª+¯Üªk“G8éxBrJfÉ©.­-8Úvxöt íû¬¾$®W.)Êp#±Ûóä}:Ùˆ"y–*££ ¨B'”Ù4æ<&bI2C Èâ=ŠBB™Š·Jõ…úJ­MbЀN‘œ±T'KL‰Ø%jÁŠËˆßW–ÝFœÿÄ*H3BaüÿDaÏ}?ÍüòÝž¶ÆžæŠv‚Œ““òAw ËòÝh}a“sTè™Í»Ë-uÞøJnš—_F’ÿʳ œNmJš\ Ž`RüSÖB õl_Ò™Þ–ÆÆjºl÷QU9´AeM±»¬½ €bx«z}†ŠïÌz-¨¨+z°UÞû½åÙÆñLö ž­Éb¿ T¥c˜˜5L8„jà BLaÞ“3¼¥ŸŽ” ƒ§ u*âòŸ™Ù2«Î: ðNè„£Úï"«—å Ü V§,Ãè¹¢ÅEgÑ’Ü3ž×*u'¥§¦§K¬J‹’®N1Å‘|çã1»e'›ûš>­»N›K,Å¿1: Ï à µøbR?z©ý¿ZÒ¤\®KôˆÝì{ßq<-DXë­¶n Š &ì*“‚! ¤… 8h önÞÛº†aPÀ-{Ë é\/¤5bL Eùe­Ùá¿‹„(¦FÑiÔ ö¨vWKM­°0S» Vo8~ñÞ—ïá"Ì¿àþ©—  E(žû1ôO™¬Ï1%Kxa?¼Wx¦ö=÷G´µÜü[¹Oƒ@¬-¶6­)¹[[J:lkaã©â¨A?œƒ#ï4^9ŒA/õk1†ðø™9añaoµOÔyÄn¨vv1GÐK—£ƒÔž¶IJíˆn™Z¡sMD¬+èåɾ ¿öóœûÛ2[3î® š_T]¼2V­§}»fÅ—ÄOÅ</ÀsžïñŸñýàtUvœ Î +ï Þ®ˆßúâ.àTô7žt£±eG˜šž“DyÖ†êôjŠDÙ>‚éóƒíÖó½†V•_yë²…o38 PJY5f% ©ÚtˆHS¬7Ñ* dÉìz3S¤/PÎÊÎÉIð¸Çï&ºnk²Ù.¿ÒeN¥QÉDB˜6Â!Ìñ`í¿ Åæývè4†èÔ. ¯üV‚p;Øì[“¹ŒiE³CT$n²ä‚Žt¶øH½Ä[rÞª¢ÁaS4éß0Ü6lùÅ„Äò˜hƒÔ*u“¾.sÝB)·Š\ìÞ;èäŽ'Éó/\QÄ=Œ.0”˜™ºON<=ƒcîCÈÓÏñLò|ϳ·Z·rŒ´!FBÚ7HÕ‡šÂ†UÒ®‹¨$Ké ï†|'Ñ|D¨i¥zm–’Ž_‚)¼Ø+Õ~!=Ù8´¸÷VË3»Bæšø§»IÅg'ûþˆky‘xÅÚà}@·› D}·¹ J¨J‘3[¡ÚØûæ÷_|÷§úÚ`1’Äÿe:6Ú÷›$exÀ¶©ÿÛöC4F¸« Gr‰Úov˜›Ð»ÏÏ#¼}*-%rŠªêª]í'BÚ¶<ð:óú7}†ÿøÏÅf…S¡Ñëå zíÂ¥š v§v÷õÖýP{œ9tédk7!àn]qåóÓã# ºüÓKœŸ&x<¼‘¥¼J,mƒçf¸Ê²ñWCçývzút./óÍÕ=>T Þ) Òñ2}vžEke¡Åäöòè}§Bе@i2;ŒºZ(&¸µä— ¾§ÞÏÀúΘÊbà‚ÙjqzƲÉÛxÆÐ…¦€6RÃí@ˆ@ÛöÏ`ë}ÝwÑéB.çE¿94ˆŽðúûŽß»—QU¤¦¤*:n‡D-‚l;²j¥%9Uñ@­ Y(̳êlc™±¹jÚ Ê.ys.‘d›“'‰ð x¹ßÎ]Öñ'·Û`0Zóé¢ÊºîvpMV–Z",Î"BûÛ3hRµc› “½ï@lAdÚ ¥ùBPƒÔz«­²¦¾žz`g‡ëw¸/pLï½/Þý¨¯à)Ë¢1Ïbf3æ,ݶ†P-êÇ¡®ò[ž,7Zøa¥‹z'®±‘õ¸¥Žû弊[ôž©¬øš­ÍÙ{ˆ$²RW(JT‹Ó@He”ÉKjj*Z†U…ÍOÅ“bIgÇ£Foò¦@HPVÌœç®+Ý*Seò±’ÀhC–IVÜh.ª† Ê•í¥¦ c‚Žf¶}߉¢y§G^Þ^‹.¸Ùƒ“>ã ·ÑUz†‹&¢©Ecïl¸ˆÙ îý7¢ßÑ÷½f¯ˆ!<Òý`0õ ÷Ú©”-ÂÜù‘ÁLLľ¬ ðD.É6 SÁ÷\Ñ^½wW&ûþ€¾èåe.—îSe&ÇòÆ]*8+l•Dçæ•§d'ˆ#ÂÛ²ïüpãòr±=›ó‹‰ôÊ–^ÒtE’†öýºA,¨™ŠÂcð<<ãùcˆôúáí,( .¡"^Þ¬ˆ}‘^»u7œÇÑØÊcLÍ™®îÃ@³/W0ÿ@wÏ– ü¾þç }ª×éó@Ç—ÙsË+ K&:¿¸ M²ÖÌí‡ Æ²¦Š¦£®ý_Ì7ÆÄ$&<'f¸H¦{]è´›íÉFì–Ͻ Ÿ¸‚F›ÑJ^MPÔÅ?ÂB•fÑîÝLbÒ¾ô @Íá¢1ÍcѺTÅšb¿Ó¹6mi¶SìÌ€hjñª¹K׬9ô}•¹Îä¤Áh0È9óH¯ÔyOˆ#uJ½–Èi•Yc±Œí,¨mr¶í¾˜{(RÕ4MlPv¦41©n©iA}–9Á®6Æç4A'õѸ{¯~ÃÚ½R“AëÅ£1õæ/û-s8Ì1wP̹ƒOöBWÑuÿü(¢Ž%Ý9ož«| ³-tÍ…Žóä£kŒÙ#ã¬C${"Vm~6ˆm:C«ÓÔæ¿WpT[f×[—îVҾߡ)ƒ+xõijZZšÄ!¬®®qÕ,ƒÛƒÛ؃²6Žçñò^ Û 6c”õ}•x+±²òmŒNF¼…Ž’x÷b4•™èF´ÁfB Þ¡=äà ‰tReoÃüH¿óþ³¬Lùù`2ÑF£Áà¨=†fÔW"P-Ü*Ð(•º<â¼ð”¡–¼˜<ù2ö‘ñI˜™SÐk:§áŒöÔýiW…S¹L.W F-$›ÔùPDzþý®–Twívá•»Ø÷E%Öz»‘>…‚ÛÔ3a;” ÑŠIãS´þXj[—«¹ÜI»Ú‹Àô a!Ù›µ!(®ª¢ÊZo03æJ0@!\ñ¯{Í^UTYYç>ݸç$á«Ì Uªc@1ì²\õeíµÙ-A{BBÃcé›D¢˜¯Ð °Ÿìf£ ÷]h$¡‘@„FÜK±Ç/óf,^Tà®–/\æ&³‹ù³gÎïö§OMRÄCÈùxü ÷/)IhT4!òAÞö½s¿3¦R !—¿çPÔñöŠºšBº´µèâý°ªU ÖÒÊÔÿ¬Ðdàÿ¬Ø¼¡©œAB98—3˜á¹Âëx»ÕÙf²œ@ŒˆQg—Ñi Ц‰‘)iø@Ê2ØôL‰¾ZI„½D&Ô(ñ‘!¡ß^OåèÆ^ƒ£hÄ©ié(×FB,ĘâVJv~Ò˜—jdÎ ³ êଜ„4û~¾êz}íùÇÛ®`,¨l8y´ï*P=°7W LÒå5Sø/7PèÙ‘5üYô?o]w¶WôßœH+®–Aì‹}ãÿຌ،€³ìO…ƒÞB¼rB¦£Åâ"´<Ì 8°‚4JåßÛ Ì; B6û+OÇS…l¼b8‡‹{h7W²aùE(~d£ÑˆÍA©x.Qer|\b\JYNm‹»¹Ñû × ‰øØ èÒŽç´GÉëP£I˜·~åAÌ  ÷/ Áëa)•˜AS.ßëDO^¢;¿jGKá6…O)yñ!~ô>:Þzi{ÿŒ«¨ÎC‰[ÖúãC! ˜Œƒûö‚×£‚Lv#vÛ4È{€l›*_¥Òës¥ôêÙþÊ$­Ú‹îŠW~ñ Kë1JRÄ+ä«€¯á5¢Á™£Ðm: ]УíúûS"FÁ•@ŽAbÈ0© „×,F»!ßûÙ׬^[‡“ƒ®hDã]hÖ;´à=ÞÉOÞít—¡EŸ~qBcVY…ç­ÀA›…–î›GÐøö²¼V‰J­V*6èh‹4_R*)jÆëR¹¹þÒûhòé<&x?ùìòyÛè AhZh ÿr™?è‡P`49e«/;ÔHÌF½°&§6ÛÑA:ߘ´Îù`T£aÄa7B2DØDùѶ{ÈñèôÔðå^%ZÖXÙè¹QÁîò¬ãx– @Þ°”4Qµim¢*%GCk呸)½j]m§$ÉåÌ!#%°‘èQ2õ¡` ¼º3I•.PÒqxB0ÿ ¦^Â/õ×h•ÚøQ«Äf).°9i»Íní‡2âc[e…êj¹K…çûi¼Ã1 •Yªv”:«,tšcÑÛÀüúK…%fUÙ3häL¶2u˜è5 6 § Þ3Óõõ­Ž6 ™®×7sÝÂÛ&=ßp³e!~‘ ÒJ»Ô.$*úãeË]PM]~ÿò½‡ã¶ÏÄ6h,Jº ûñj2ª²ÈV.[ùÌ–à£ßíÿìOÇF¶Òvë—gv ûûÖÚë(ÿÇóêç¥ÌH Þ™™r p+ÁD¬UÒp<ßÔADI‘¢,%6;)zOàü7wn¿}Œþ3šQ„ü¡‹ªô4ž„'c_<Ï ªÁ¡1çº:޹iì·™÷ ¼[ÔÞˆ¦ÖŸd \G¯¼Ô©’!ëCˆ‹|b„'Pò½TÙh&ß'ûþåWIÃwà6úW%ªHŽMˆOH/Ö·¸›Ü´×­ºúCô°kàö¤Ò&Þ@ûϹѦ&Îà O4n4Š8£Ê‡(3%'1«@æÒÓ1Ͼ| qâÆP’´àzAkCyUuUŽ-73Gœ­¤µ^lÙ ¯­êË·ŽìvC(µrÎÂÅ/=ÙŠ9[ÿ‡ªNÚ¤u'êtWwO»À#͈cLßõ£¾‚FèÊ6¬£vÀ½W œ4 ” Dyuú <O$íÈžÌÕ '/)6ÝúU™Ï@5ë‹…hr¿‡Ä>—q2¦‡zrÏQØ~ì(P%¥ŠÔhAdT¢°¢NNÌ׈‰~wø®kÖ­†aýÕ¹^b¤ˆ‘Fq…åmÎm•5îbúÿ?§¿ÞùN%äÐXÁ-ûªâ›Â ȇ”ÓÖ¤çAÄýé²W÷¼ÇA*¦ÖFà:{áèÍ?_£ç¾Šåûeû¥ûùÊÉžhðä$eÿäÂveµ~Þ…æ 'ë‡ÅõîÊLtzF¥œ”Z¯Ù_ÌË]½oËÁp)Ù5á­Ii-C¢GÐÃÍâÚøðÔ˜¤ø*yeE‰3ßLÐ=øó|á×p`ž¾ƒžºƒæÜáxÞ!Á¹=2ÓõR÷jn鵬쎣Ëyªƒ(ÒZ(ÉNWgeAê𔡺¶üÐÑàªÐÐĨ,:3LŠ'QXÌmF¯ØKl%Ö¢¦Æ“P Úò\™6)†—ÕU¹šŽVìÄÓâç¯ÓÂÙ£7y/ŠDeÝÐ8ÉFI FÌW¥‰cþ}Tîö Éè™Fz]ÄÜ ø;ñÚœ-²4¾Zýú^‚0?·¨Ê\^5Ãëéñ{ûDíWûú;]tÍñâ^G/å“áò¬w¢4c‰‹‹CcÜã®> stream xœ­UkPgíf »UÝÐ%íc Æø@Ôb}¢q3Š1ŠÆå1×5"ÙXV¬Z³»I,ªÄ”XÉ®±¶êkò‘ÚýfÐÝÚM~l¥öOWwõíïÞsúœ{hÊÓƒ¢iš UnZèº[&½IK³=¤92À»‹ðšCùnÚ0'H¾¾Þ¼eàíÙ3{Ò­_!ût”á‹¢§Q2šÖè‹BµiYº¤„DƒbA컊e«V/R®RlLUë’b£5 e´!Qm ‡;µ±IjC–bÁÚDƒ!mõÒ¥K¢SõK´º„õï.Rd$ªõjQ§Ój ŠmÑ©j…{Î%îk¨65-Ý Ö)”Ú8µNCQ”ŸF›¦ ӌљ±Ê8ubÒ¡Ô©+(j;Aí vRÔGÔ.*’ÚDí¥Þ§–P›©0ê·T8¥¤¶QЧxBåI¥Ri%ÝLÿèQçñX¶^vÍSéyÉ+Ìë3ƒ)a°2ÖÎs¥R¹tDIæ Ÿ!#F&e¢!þ‘þ>–…î‰Ù®Ì/¼¡ØKÅUAœ…©ØË µzனÅ|Ý{ž†Ô,’ýI¼^ßXº\ÀÍá*Öa¹jí~45N”°>RùQ§ô†ƒnA~h;ò“I)HË£¹ËŸáY˜_„çã·GW£7ÿÓa¤ða¬äãA}¶ ¯¸îAˆpµº½¦ûÒ¹Kp:4Õ)Õɰöƒ " “ ãâ À4x·ˆæ:¤iNz˜ Š$ðah3ÓªïÊïMþþÏè=°äK쫉—_añfÄh„ZSC9C7~ì§Pä~¾©­³©¸έØ+"w|àÐÅÖL¹2Aä0¢†‘ íBFþÑ)ä5"˜ÍY9 ábÚ3.^l·;qaßþhýoÒó7 ^ýŸÜvYú­Ð G¹‹¸¬Ï ":'¢³·hBÝ[H&ʤckùñÓªÑñÇ‹UL—mœ0ßA¯û<;ÍŽTÆ„ÈEÔ©bBJc‚„ltö)ƒA]ðIìÁ‰Œ$uHo¶Ó‘ò¡ =‘Öñ PUœQP’'d×h« ]®›ûqóáº|y{Jwñ@‘³ø^>(¹ˆ¨#Q¡ñ~¶PvÒlËNG rL³™Pt¦Ê ug„ÊŠ“Õ•6§ö2œ"Œûvþèî¡¶Üò䎄ª˜ãªš]5p»÷¹¥áIåV£U°±–5w*Ï&gu"ßF”&¢FúkÄ<eh4—GSE<íT1æ`-fÞ)ã>ÅÍ,¦¾2~öÅŃÂmÕ.vWN¢& >k|¤óevišHK[¤)ü±zÛ±{À‰¬Î¤1ë Xs-„,[Õg.0ƒ¹¬P¾ßC6¯;"£1o5¥Âaˆ´&¹‹†X4o|RÍáZS=Ìl‚ŠÚcõDkÙGè@ZìÒô$ä2ÏJt‹¾ñ워g˜–èÊlnk·_nO?m¶öŠãÖVà/ìÛ"O`ñÛåxÆ>x‹ûõhÊ£{½]ývÁ»¿.3up¾±ÐZ`‘«lpìf~ǯ O×Äz»‘ò¯»`u¡#ªx)Òè²ò£È¯æDeÕ—0SdSL1æ,ÐÁ^«v Ñ´¹¾póBÿ ˆòpÛkŠ5$[¬®‚,òÃÛæâù¹™ÉÄM3Ul«õš… wLv·{Y¥Ìšwü+äé?Œ}½ö«˜V褅 púLîŠ-ÄßÄß¾Nºƒˆt2š.C7¥E|QqA¸×DÜßxa-ö_ˆxÁ·kHÙì6äY{2 ”™óË„’Œô­!ÍêÉê1÷ÃUnдggjMåõ‚ÏáÄb#³ß-©Ž´´(ˆÇ;%sѹR[ÌÄjâˆAÒÁ¬/O/7B p*æ’E´:à\59ÜÃf°F0ZK,:Xà,'uZXd®Ì«È·ÃÌF¨8QY‡²¥þ–ñ¿¸íõ5Ù=ðºÜöZ,ŒŠ8ŽRE:×Î!Fòs ·ÝØ#‰ÐÀØ&þµ­#ب®›8<%` ^€}‚d½='Z:å1,ÚzÌKd•%Ð#—þÎbfü[Þ¯‚xÐ4åutÂ]²ûœp«®ë”³£©ÚÀq¤YMºÆ“|ÐV¿Ú¨ÏeÒB¤á›ª,-# h^ŠùeË Ïóÿ¶Í@³ú¿·ŸÉ„ÒlSin©\¿m‡1–Ìã hú]y‡§å† ±O€{xžAþ:\€~š ïï‹Ý£Ìè 3±3ÝǯÿâX@­DÇ/n @ú9¢d(Iú‘ïKèŒ>˜˜xð`Wâ•ËŽÎ>€C Ÿwÿµ&Û,׬—¡n¼ú¥J)`â(iäöÄ9ÆÂùÿž°´¿BÒcà•jIû±?’öÿp¾ê;fý7ظÈ4!óÉ[ „žÍvîÿ”‡ÿ[dÞt{¥¹S­#¸¤d‚«ØüQY9” Šl‡ë!üKo<õ ö’ìP}‚_þVô­ûù7,våúWNC_w›£å*1 Â*SfŽÉ@Â0¿ÂXŸWéê—ýãJr2-J›iÔé‡ò_ˆ!Ò*¦É\¥Í)…#…÷R¡RL/$©æ¶P/ AÛB!DéÒ‡<¾‚²>µRD5в×Ô2x+NFÔa²gp£÷$g•·7¢:½§Z¼}(êŸZ0Á¡endstream endobj 232 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 464 >> stream xœÅ:þCMSS10/øøø‹uøýù`B@2>w‹ ‹ ­ø5÷k÷f÷bou„Copyright (c) 1997, 2009 American Mathematical Society (), with Reserved Font Name CMSS10.CMSS10Computer ModernRS34OÄøˆù‹ ÷¼È÷˜Ãëá÷ÚÜø7÷Úò©ÎÒ‹Üîûâû#û…ýJá÷Ñ÷2÷PûÑäüGø÷˜÷$÷ÍPDDKPû ø¿uÐøïÍÃÚ÷±Úøeù.T¬TœB‹û4)(`™a³aµ^·Æ}àv•ˆ§rŸz¡j‹`ITI0b1•Ò7|:ãTÛ}Ê‹÷çð÷Ìl»u£\½j”/¢Q™{m¥„‘o¨‹·ÆÁÇçß»jl°u¡øPœ÷W¡›¡ûb—È Ù  7Ÿ ÈŽ Ù– é¹,endstream endobj 233 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 2875 >> stream xœ•VytSe!MxÐR$¥É«ƒ( @Aì „a©A„–6tK›´MºoI³¼÷îKš¤ii›Ò6mJ›tg©lQåèqÄ \ü^ü:çÌ ÅÏ zÎü‘ó’“wïw¿ßïw÷Šˆ1„H$»tmüšùÑÁ¯óù©"þÞ1ü41à„_uÉ4bbÜ’i ¨Åyá&†°ß½vLB=AeQÊ]„X$RëŒKÕš¢ÜôÔ4mÔÌäYQó.Œy8jAtô¨§²”¹éÉIÙQk“´iʬ$­ðC¯NNWj‹¢fƦiµšEóæÌMÊÊ›«ÎM]<ëᨂtmZÔež27_™µ\­Z—”¥Œ-tîèc©:K£Ó*s£ÖªS”¹ÙALÉŽSç.Ï[¡Õ%®I^›¢LKWm\Pñ0A#-º"kú ‡ŠX–aÍr -ÀAŽX0ÒEÔ6ÜjRZÌOœ• àîãY»À_°ÝŒ®µ–«³Ö’…ÒuOC7zÚÞB5Øí ìØceY^c½·½Å×­iKY½eã¦<ò(ò!?:Ô+ê Ì£nô‘ìé_Ŧ"ºX8Ô+=ˆvKr¥¥Ñ9󀌑@5ön»cH€‘”¢iÞí÷•C![!¼èB:ÉÈ¡`ié›»Œù40ŒZ0ÒeÎd*\ü«Ïöì…—à ¤G%‘)­¨ñØôÝØŠJÿX³è³ã(þ¸˜ŸÍ?!³ºÀ 6ÒAÛ«, ”—** ¦ŒÎíP$ž¾sþ énmK1µGë3 ^«€udzî²û¶¬oBL Ö ö ÐKÓEZ|O$c,¤±Úâ´s\K«Âao¨·Û;Òõ{D Þ<6Xâ-j¢²}9ö„Úm®ø:8Brðº»ù ÉuþÄ/¿ü~­CÁ±e1hR­9Ž“Û¬Õ^A?ë+ì{¬à´í¾ÕÑTÔŠzD'?DÌÛb¾ÍYrŒI´¡*«°¼HÑÓ\ë:l­§Ým(<ä+ŽÄ<°bU|q­ÎßÕÖÞÙQTŸÇ)ü}Ãu> Oú¶nÖ,4Ű hY¼Ð„Cca.¹øÛ’W.k¨Q,IßÚu¢å½] [=Wã ñ"»ÌœeV)wæå¥§™˜Ù¹Ÿ4 •Ö 5¶ Ö ñ‹vñÿ#Ú)3›æÌ2Tåhv¤,fÙ*ºŒ5Ö=\'ôó떽úŒ\1írpïr×Ô|y5òÔ,§¾¶,rЊU¬ ;wÂÚ&8ò~vOÐ4þ­‚$ ;ä´CsP;˜p%ò&süðQ{'Èo‹AÍhXµ Ú­Ö Nˆ0g9Ù&òä_R¢ÖЖ¤DcÕúï" õÁ.…ººÆ.+W#ÈÞEæ°+™È„kÞ­HÕIhd<©òð€ü‘_ÔÃ_$¨N–+Ua"‡âe@ÆJÿcÆ_ò=vAc߀¼[*8E*=v@ < ©6!ámZ¥oÕóoJ˜ ÿ6FÏV2z’ß- ç×WzQÏÕ~¯èìE¤¹&æÿÆGÈ61æ 'ƒ|z;›÷Yê[‰§á0D ©ï¢)?5¼Ê¤uQá¿Î½ÎÀ¯ëÅè§Àƒ2sËikzBz‡cFhK^Õ¶ ¿]`wØÚxƒñ³·(‘ä-‡ãN«‹«è–p‚ËþSd«°êë@^ —½ý䤳GÞû½µöÛÞ¬u/œ… µŽš™/]SîðØ8‡î2ï³wW×¹'ˆ'Ž•æ* ð›#§"G»uŽ µ¢³çÅüëU2<ém7:Rq¦Þ>'NŽÄãñ8ôÓ§¯]>étRúÕHö‡ý·ªSF[œ¯Ú¹5{3)+ÛÑlG’Ãç¨ÁcÇÛöy±-ΕÔ(wY­(Î#:}9#Y‚JdíAµ5Û<à‰ÆÇö/ÅS±DøLÁ!/úM|I›ª§¡Š±èÍŠæÅV Œ¼ø÷.Dø}‚´ª[e}葯ÛüfèX£ECý¶)!.K¼Ùv ºv“Á5å±ë2 ,÷*úîOyt¤é÷szt òÂ0Ó}›ßÿo•Z÷§yªbíT8òíâšpS„–ÿ"Fƒ±¬­¸-K™—šÝZÜáïêò+ð|¼A6º]Ü^.†ëÿ`¹è憭í0ƒLçíö@ûè>Í’[DÝ¢T/F‡þ\*ÿHp/´Ã¦ÿ/0èZ÷ “tr QHæ/É<ûz;Þ»?RÅ$³È€ ¿ù‰äXCN±žaªJ)Ü8’ÅSËÚ ZÞý:x‚ÀÓèe „ͰѪxÁµCXÕF~©.°Ñ wƒµÚœˆ4|ˆ~þù†prÄM”ø¥˜ÏEq‚11Æ‚ÕeÅ‹…5³4$ÒÉB_/“¢$ýáÝÃC¯+$/86ß»OZ™Üg®óùÚö5©M,°Eϙӧ€|¿'~E\Æ"<‡Q8oÕëÍåP&çÇ öDóo}/BI7®Þ£S‰Ì¯­ÏÏÌÍÊ)ªÏéöú=½ ´1Â[Øœ›¥ÑdkZ‹Û|{öt) pB+ZrSt …ˆÑê&™~Á–g¶î šÖ“Ýìh@c~DwÁéç•ÚÌìÂ]9}ÍnWS®óðK›PvÃ#Å;võ?ªÓ6Îã #ˆòœ€Ûendstream endobj 234 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 6870 >> stream xœy \W»þÄÀdTÜH§‚Úê^k­{•Z· еŠâ ¢";‘Öö-9aßׄD뮸T\‚Ú6®_mµ­Š­mmÕÖ~ïØCï÷?a ½õÞûç–Éœœå}Ÿ÷yžwQý(‘H$Yììâ2mªéÏiÂH‘0ªŸð†8Gú—N–oPC-|c:?é`d%FV £Æ$X .ÃàðØ9”‹D²ÄŲÝr/o…ÝÄoÙM›;÷½ÉvÓ§Nkçà)÷Ùáhçì¡ðö ðP »µ²>žŠ»‰ó¼ŠÝöï¾6Å# xŠLî5ÿ­Éva> o»5žÁžòPÏvËd »Užv]Òõk±,`wˆÂSnç,Ûé)¤(ê#‡À ‹d‹w/Y*_ü¡"Ä14Ì#|ûʈΑ;WyîZí彯g­¯‹ß:ÿõ®[ßq›âþîTå´¨éªÑ3gÍ~oÎè¹öcß7ïƒùßZ8ióÛ“)j4õ5—C­¦ÆRSã¨ñÔZjåBM¤ÖQë©I”+µZDm¤SïP›¨%Ôj3µ”z—ZF}HM£–SŽÔ ʉšI­ fQ+©Ù”3õµŠšCͧS ¨!ÔBj(5ŒÚBYS[))åF½F¹S,õ:5œ²¡l)15‚IYR£(šzƒâ(žêO  Þ¤Ró(+jµš¤² "¨?DJÑ‹~ïõ«ŠŸX¬±8g¹Åò>=ŸÞO WÉ/Œ²¿¸PÿsœT\2°ÚÊÖJcõpÐòA5ƒãÃò!xhí0Ûa!ðõfë2ëÿH^s~-õµ}¯=g­ÙYl{âõůç¼~bø¸áõ6ólªlÙ¦7B9âÙÈå#o2ê³7ƾq‚ëÏ9q\;??d·Þ.ÛîÞ›¼P1X¨@†ß `¼+ª|Å`î²ñTZˆ”(ÞêÐÛ¨%©c“’âP4PƒJ¹çtYj؆¼øÊß-a6]9ÆÒ+ ùT¡2þ9]Zƒ “ób2ù¤ô…¼XaµôUl´Ä6tç¢Ðfp1X?1‚‹q¸ôK°^cQñn­”“‰jP ·¹›¯`]DæÝ#GJÞ–ÒÒcOo8}¶Hö‡ÃvJ:Ç՚ƽÓw\ýlÑyœôËåÈÍ'<œé\[PD71…¯X°…e]lúLŠâÇÐQrà_‰*ù%xü|o©«D{ä]·‚È­TÌWÒ`…C°-.µ ¡•Ádíjòîï]Ó#ÿ¬ÿ¯·T„üý–iEwB àf°þÞQÆÍ$ÌN°0Œþ®ÂÙw»Ÿã$~, ëd않£gPssÕ=,æp(­2í¡•ðµ*ZúbwŸih°¼Ô~Ǹùª}9ïˆ X,…Pàèï¹mÚ¹z&ON2À—Ñy#TÅ‚ø“§üŠ-°Å”‰x–þ4úA¿Ÿ~k{àY¬ƒóíÇß_½zýƧ©Sœ:˜¦ˆ59ÑA#|qI,„ wØT‘—˜Ì©S:Zî[·U·¤%óŽÀNxßð:0÷Ûí F±ñ ¨æß²ÞŽ˜©³þ€IðÖƒ?Ú/Ûj_ħG¦)Ó„ô:ÓbXn€w ðE7\²H¤î‚=”²/œ½|뢙o/uZú÷ÓoqÈãH|#b@ôä'Â鿼¹Ù-Â߇—>©Wyþ \uÿ¸è^Ü3B™Q|I¶{d1_Ñ¡ ¡{¯`; Â̂™³`¶å±¤ã+¨~–€e+ƒûŒ‡`1ß•ò¸wc¸ñ˜q 9ÊŸO l€y>§÷fç6òí’rTáãBøeX'‘v@ —lGʦ}¨¢œ{Ü}[¹ÃÃÁÔëép²Ï2î¦dORž×Ñß Ö†¾%ïaÙƒáÒc ‡Uì†Aœ6¸4¬1º’Z]Ìá9VD¬tá¥_îëš ÷ë*‰*‚¹»¸ÍÞ|EfTcq÷*¾¦—è¹F‘3ŠaÔ±0Š~åÞG/¤¿bïÃÌÙÛ¡òíï)tqug…)yWºÍÖ7 cð7 —>ƒÓÃ*2ÂÒå™ O˜šÆ=?}S”Îi#Ó"÷tË(1a68E•¨æÜõî(œ õ}Å\¯w_Dþ â¥[¾Ž^?b×.ß©[Öæß‘sÉe)ùAˆ1e„Ÿ$Q"yCn‰¶¹€kñ:«®#€›ÞøÍI^ú ýŽY½_&ÓY$0ÃG Ó¡†”Ú:ðd«LLÒ¬æ ¦c°Å¦-ëc¿(˜ï|ýÛ¦Ë L•òi‘Y15Ý[¾sϡɱ>QœwÃŽb²g«±x$vÄ+Ƕ0ðüá†J‚º"Ýž/ ¿éD¦¤O'É 4 sÌ1ÄsîѽH›;³‹EMtõ§ÒoáuóÖÝ#íeHïí« ä4Îk∙ÎTw1‰‹ S=$6²Z?=*CÌMº¤ºó¨*Þ îÑYý³ûþžfã7ímä RòϹôÛÆà›œ4ü:”×\gfn'9e¢Jªú˜áT‡ S}®ôG™P“rnDUÂn±°ˆD¤¡O@ú›SGc¾EpÛ²îo#zù¾ ~\à£Czd[Ž*Qþ^Ó>Sôà`±AT"ÄŠ…­¦ô$Á(‚ÐÜ)/å| ½Ÿ¶ÔÐ)5ÑÇcE\NÍ÷Fq‰±!©‰;0—ÆhhW8mÙÒ¥¶ Þûq=jà/áö EzÔAd{¥ç–ªj‹Ò+-éÚŒûg$D–¡ÔŒÒK2*˜žtM3<5XŸ3nkg’²P¡_#[áRs1u]*ç—™žÀƒ¿äúºoñØmãТ½²ª¸.·YË¡­V—V¤-D¹ˆù×ÿ%¼ô^šŠ‡l|{IJËNÏ9ih3jË>±‡‰%Üòž„‡Ð^‘n(1ÎÛ÷}Òr¿Æeð=tDL‡‚œ&Z¤‹j{òÔ‡ TË;Œ_ð̲¦{{w‚ºµþ ðD<ϲÜa~úwe+ùó÷aH;b)ë‘ar¯w]â¢0€e¯ºD3bÂÓ?1Ù‘E‡¡PR\ºjNÚâqˆhÍÈ>RÃK£6#·o´¸)ù‡>™,2Bñ2âÖ>gžmFVs‡j—ùJ°ê‘‚4R¬a¡D’$&u 7û ¢#þ$/Æ>Ñ»’"ým5try¼>±| $ÚT¿Ö{h_fÄ!5ÒØ"µ|n´<ÐÅK¥AÚÌTÈTG”ÉwGFyìÛÑòôØägp½Æâ¦‚ÛÅÂ.±u(I§K䒢÷8 föâkߟ¬ ŒÔ‚ây­w¹ª 1UúÒú›c^ø5[ý8¬ß÷Ki7LÆ¡Á ºÖ Õ­b8"X²1^ñI¤8âNÌ€uV\•©³¶ë¼zÔÈw¨$~¡Ý‡¹×?^ÒJò·^Á.Ñ9ŠBÏÌĬ„‚ÄTˆŠrrŠÒ³¾ƒ¬‚³Œ¶#ˆ~å‡U’F}7×9KQÅ1½@n1¡N’ÉêuÚtTÊì Aá~Lû&$øò[%½Wi|·„†iRQ¤­O*ç„!tEÒ¡bæ‚Yø=q…¤§l¿èèþ€þyŠõBªƒ\_š²^A³ÓiiK½ïö’m#±èíÉXЇ>šôÁi+:\[WÏ€ŸÅÉZ·9ó—|t½ýçkW¯]iYëbÆéãV%1_#Ü`cöF¥ïḞfLUw¨äæ«?ÌH’ 2- +;R ½K45ȶ¡ ý/dUKzG¼i†Zu‡LÓ­ã£.ÁkÑ9#T‘@΃µ¬¾5½âü:øöâ¢?,ØœxTm‚VDBjx"²a]¼ -D O)ï3ÚKlÑÍ‹ç®!b˜å¹c{ÓÃHÑ­4>y †,“îƒûÁ'è;Ÿ£.%wg“®²¤îÒûHùþü­3¹õÌrø ·ï#(UfAßízV Õ "¢&…mj“‘ææ£Ça }ìX‡ªåÖ²lýúmËÞ&2•xÄ%p8îl½d]!Ü$¶Ÿ\øøÆpé¿cAKØu'Ξ?zô̧‡·¬_ãîîÊOófÓÕ-ˆùáæö†øúÈ:^_P•Q¨ÍÚ]™˜‡˜ÒÊ‚úÊð=ëb7i|¶òÊ|ïò­ˆygÉÂÙÛJ}‹Ãy)¦b•ñ!#w‘<7ÆÁ }ˆb™Y/VÁkðÚ‹¶ïš#Nm¨å¢³=*¦!f)€¢QBZJB1E±EJCiEy Ä—Ù -ËmlÔñÕ%¹è&bÚÁÍr_¸mBwywB[Ôé!ÄÂ8Κ¶å¦fƒU¯à`ÕÿÒþt„t§5ÛØmS­ÁÑèaÔ‘²RÂTÂÇ7莮޼Éíã©üb\üØÏšË èsÁ=<€ÃKûB—@ï ìSàKKOÝ!Ϊ5Ê*BQò@N( Ü¯4 N¹1—«>;ÐHú â¦4|‚zW¤ ?{åÄWr1õ ½*¿ºwŸZƒçÓed¡MÁv9òÑ}Àh‰Ûð >x:iÆÓhZúLP[Ô…TÈd!!2YEH]]EE×ÓÃYŸ¹ %W=¯š&oZ¶.é›04“qݶzžlSFíΣ6å°¦LSª)óë6¿Ø¦ÓýVd5ær Ê+Äý27¶_ñ:SÈo?¸;ûƒ2Fụ́̾À³ °=Ýrò: (~Ç?‹KÉOÒõ¸Y›.žÀmoZWàFܬxá¼ÙË;~Lx†æ‰¢(•éÕ8ˆ“í0¼3/‹ÞPL§{«^ïx‡F‘È3È{·§?1(&-*ÐqÊfꥡŠÈÀ¿ C0n?X]ᤅß^.þ«1‚­lzì¦Í+:•‘]ü¾ÈÌïgÁ¶Cå*1÷Hÿ­eéb¤„‘úµšÌO C@bÙÔǤNéch<O¶„] :™)·¥Cµ©O+Ö§3Út—þú©¾w»Êt†f€yzÿ®_Ý6B¬I¶Èñz´„¬Ïö!+ü#mçèêºÞéÌ¿9ø±AB¦yÕ¨_i<äÁ»OŸ>ø†pÿÝ«þ÷\u–îëĤ³}®ïõáä×M;LЛœM™Õµ¿ýúîuÇu-«GMü||1z}ø‚Äm&WÓ”RßÛ M/Ó“‘0ŽU7„¡Äø™?ŽÛ$^¡Ýtò°+’¡È‹‡%Ò–öC'UE:qX% ïáœÛôI&ÆèéÛ£wʸ¨}²2_ÄHïú¡UpÓ mä ƒ(_Ј G~kœ×‡‘hw|?aU¢«FF’te lÒdÅW*P<é+ÂPé+¶à-ÉÎÄ:¡LMyâ§®‚]_8¯ËCס\^KßèÐezW‘Â6—8“<]zö§qônŸâ­:[^ƒ²Éý*”ÙdÚãÃË1$ã*’ñs/ßìθc„Nþk žÜçÚ±ËÔ’BÕ5Q«0U,Ì…*¶  55y£_§c¼‘·w*0 üô-×D°_xÆÆd'¡T"8jM|2ìh•‚"™Ä<¿ªŒ,mv÷GÛ«ŠâÓP²­ÊkÌ.ψÉîÔ&‚-0L wÚØæ€?O¹¿ÿ®† ÆÕ{ö˜`;¹g#“ÿÇ}ÌÑ¿TèD„ÉbÁíå6»D‹PSU’ÄáoþZ–¨Ô ”`Q©«È..Ìêe\Ñ)#|B*..°°j Çý°Å„·d«žãáÐ,žÂ HñÆ9ìN0&Âx㯷n^{ OÀð€:7a€S¦WO…•²¨LÏ);ð0­ 1Yº¬*’Òˆ¥X=#’×$¡HÆL“ /å¥ôŠÚ¨!}|RXRp@•&¯Ñ€(5mƒí%(¹ <¬8—x1¤íÜS/ׯóxT:Úö00K‚ªS¹4®.Þ«­îîÉžëDLöq¥ûY`xRLL2ç9Û56\­Ô„ªÉÚò²äâòCYŽröS%fÙÌ0Ão*Ý]»FNõr$[€š¼Iìc¸¿HbwÉIA7½ú:sêâÏ/°ñ IÉHŘ‹rðC®jº­t¸ŒÃñôÅÓ?ß>rðrëñ+èk¬Æßƃñ€ysg懲-ÕW–Ö$ÆgsÅÇOì9‡˜{_n±tÓÇŽ.¯ õEÕ¡c®õ›kVfx#&’ÔsPy a0NMÂëË`¤Å~GkýÔ[P ƒ—IˆyáqêΈ“=–>øIw·˜×f¡S¨Žy$!±-Þ_²¿ø¤¶ª+u ÁÚ8>X‹û¥©[»Ÿ%¥]2=,>ÚþQ;ì6î øDöuÝyô)skÑçX‚­–n^æ[­Ò™ÎS””—šÎíÙsÝÔªqÊË- Ê{wï$W{©×¤D!ÈH3"wœþõôòW­]9uÇüüA|ff6iF˜ªp\å7ýÉÇD †÷ÃcNú]ßrѱÁ´'!¬œ®Z?€ß†K¿…±%Q%ÊØ¤ä¤D.@¤TÎIÈ.ˆ. Óû ÝH¦TD†E† FY¬,ÉÏÊÌÊæêjJŠP1>…1…ѺÐ}¨Õ•T–êŠë+PY7GS¡¿NT ä­âÊÆG“Æ4ŒñÍMÜÃA&¦œû¸°åf89uÇ|õ¼ ÷&æøp2:%k¢Ãü~ž[âç’Î ·~ œf[Ô¾ÞäCƒ`Ëz#÷Äí! ¬Dsƒ+hßÄx?—K"¢±,#3/‡«m©íÒ©Ý=»Üa–ãšÅÛÒ ÍÑQãyB[ƒˆò`¼ÒïF¸£B¾M¥¥ùõ\&4Ã;°ÈpQW––†ŠS„ÒI%¦G?>($œ÷èöHåÝþÒ °•@Úèb„J£é â(†v‘ó±Ö3­'¿ºs~³ëg÷EüWþì×µÇZÐyæÞì;x ¶zožýšã¿ ⤿- ]±záˆIgÀ0°þáá“ÛNôÜ<ÜÈÎql5.\þþ‡ó+W-s\;§Sgî Ñ¨èKÔO÷~ÿ‡ºóJ¸Â¦ŸðɉFá(06ݽ»Ûh£ÒCëf–iKMIßö¿è:ÌÜ6h&™3º¶CôSuŽ…6ý`âÛD~½´¡%™™Y¨”©ŠÐÉÃ"üÊ–~ʆ”Ö0ØDRIXlœŒÅãç\Ò§IÝ“3èLé±C ›ÁÚ¯ÿüÖƒë­7ŽÔ)ÂJøß‚ ¨FÉ(9C‰â—ukæp]Žf,þ)tø¢Å«ÄoŽþ¢áÖ·àòmI¤õ…++Ú¿3üzkÅ•áÒ¿ôpêÙÏ¢¸¡ hƒÛ–±Ì*P²8Jâ–µ½"êtVF^ªdê‚KÃâ½R¶Î¹0˜Xn—©WØÓ€ªK¸\~ú>º†Zd¿âqµŒô¿ÎÂNÉÞ”Sª’ q É1(š!½V_UÐü¥Ë < kîìü‘\ÉyJp¿è  ÎÂÌóg®Y¬É[¿F'OlËëlx=QËγ[›Î¾ú¢éÿ¯<¿ôñê 6zgxTBRª*)™¨âÈÒŠŸ?‹Ï=½‚d>²Rï–Π´Áè6Hmƒ1m¢¦Ö¶Ö?È·XXür›‘—–IêŤüÁ‰áDðpÍ_KSâÔÉ(Ñ6¼DY^•]–ŸÁÁw¿³·v]ܸq×j‡æÕ§N5_¼ÅÍ\ÂÆ„×=Z^[PP¾ÛÝ=|7±ëˆzÞÙÙºy“çÇ ?ødíÉ–nw¬;z¨¬¡°°,`»{¸,Æ„äV½à®ë²*«%›Ÿ]˜ƒÊ™Êðe|ª&^Åá{xG|Rt ·EÑ9±y‰p{Úŧkˆ””T–“w³y|-M…ázíÞ5ÑÁE,8ÃS¶`/±>»Š'¶^µ‹XŸ½ÄúÀ <™-j&w|M¦èO:Æ—Üi&NöïûYC€Ÿ›eÚ*LÈËÆ÷ÁÓF•›ªGŒ2> stream xœ]’Ánƒ0Dï|ƒñ:‘¢½¤—ZUmŒ‰8IýûÎ,M=Œ¥‡™õ¬½»Óùå<÷r÷¾Îé3ßËaœú5ßæÇšrÙåË8U]öcºÿ’­éÚ.ÅîôÚ._ßK.ñC6~k¯y÷á}eŸªÍ”æ>ß–6åµ.¹8:§ÇaÐ"Oý¿-/›£ž¿vºÉw•“nò]MtS•;`U« »{¢¨É¹.jr>9bVP€52Q@z}«&çêLD ˜ˆðQÎyî6^MÎa6j6Døó6 ÙÀט·¡7Tjr.°Á_0o 7 ~°C†½š€ Ð <Äbcʈ•–’ & oCPU¬²°² ªXeaeAU±Êb•_¬a ‚ €¼ éÕì‰hN¬Aaƒ—J9yígF;7òÜÕŒDDˆ#2FD„h1"cD> stream xœXyxSuºNÖbë2q—sÊU¼¬z•P@A@(bY ”6mº'mh–¶iöœœïdßÛ¤mº7¥ ÔRÊ";Y”EÇ«^EEæêÌ83Î/Óy¼¿´8×Y|ž¹÷Ÿç²µkWÍ››x;/þ?þð„ø#‚B®èÏwdÜñ/åÅ¥ï×÷-¸¯æþ{ïÿâgrá‹ÂXñ@KœNŽ­ÆgEùèÏõç7Î é*“Ì„*¤¯ D< n’µ¹Z¢_¤º¢®¨³Éöt°ŽÚØñ/ ÄAonÁÚ­ÜÏ*wQz9SZb{äЭ>ô¸Ç§Zt•™T¼\ðÚV ÔF_S‹·Ù¥ê  pš¸´«Ù¼­«åjÒ|jw,2A™£SUæW¯ ¡óÍDzQ¾Òp¸ˆ¿¦ ‚Öj6QÉ#ïBl„ãÇUñ9Âùhí±:€+\?àð±>ÂD hr7좳  ÖÚËàè”GM5 Ðú 8ìÒíñRh2×Ååq vkÆöÊÍ´HÃ蔺 Ù {í1èƒýtoâJW|:ÖJéÙ"´C]m] ø-â¥~ñÍï8ƒCË2`A©Î±h=j¢Æ².ÛÛ¡Þ¢[§¸PÐZÉZíêþœ¯V¢¹òáw.ÿÝÕÕÞѼ¯yØd½õ > Y t䂨ž7½F'íf(/³B)Tk5=‘ÿåíjÜŠ_†B{»ßqŠü‰ƒ"„Çì02(¡¥P{á탔ZãtwX+Ôr™ÅLqU\75…À† Ù||!4ø5,…téj¨¹]C‡¾ ×0…ûˆ¥íZ E@gsϘUš ýC( øg°Qã%h‡«tûí4øõ ´*Á¦¦*l&Àííq6ö ©©ÎzÖîñÉ#;öÄFfÅøo~)aGV íV·žf­–\±XʼΠ@Ä•s{8Ž|n$Íúìýþ¡ãÔW…¿;áÓÕd¶¬¤¾D¾ü¥ô—^[£Þļ©ÇÐÜÓèÙߢIÔ[WnýˆoÏ.æˆlwÒážÁq’Éqà2bhæ7âÏ^ä’…†|«q'5ZPœÕ8µ½ÐMw%Rsqj:¨Þ‚ó  ùî*XÒ.yD {Ñãè‰[..Ù²-/§€ÒØÕ˜;x½J•¯Yô?>2ˆå´Ö¢¦Îk…î¶¶öHmO¬Ã¢¿Y²~}w§JL坨V Ä*ÍÇ.ØìLŽwÆá0²ë3zú¡.ÇjX„釣®ñB=…>IB9h ÊC¥hê³h"÷$Å}ž¤ÿwC0Â0ªÎê¢A †¦kˆ¡·Çf?ÄŸ’¾ù8ƒKݪà~ñ,7*É]´„K‚»#©¬`bu,Í‚ì`sF0.•·¥…Þ7Œ~~K—Ä á Yª¬\¶Gã4ùµd£"(ƒ<"ëUNÅMÚ,ï9vªíZìcÒv†0 öm™%æ¦éÇkc«s’¿;ÐÔ„' …e:©QEíäM&ì"8ÞÊ!´è4úä™<²ó‡[_ð{­Ç7ßn Q_¸ˆHµ·ªFePIÚ¸{Í(3[U Â[Y³<£¸(ˆçà½p—sÿ!”Bù;]!h ôm™û7Ñ´}µ?Ö„·Ž),Ó–Õ8šÉ5™ &æ^’\‚ËýdëêwMpßnîþ  =‡‰“ÒZuY®Ëä5~m‹ ¶A~É ÙfY†D‚µr»´ót»;æáv~q›Ý®øt!c·$ÊËFB~Z†J½õî~G§Ý“¼á Ýr”Ñå9ã¢a¬ùA4¢V¯ªAU,7™G䎧~ˆÞóöÚ|×@Ô‚ï5ØLT),Ã\/åö’±‹5? îÁÑîÔ#+öö+vh§!ÊàD­@¸ƒãlÏ-…ˆÛµ×R¡cÁãÁch"RÁEÒpDÅ»½V;®d? Øûñ 'èè_i¥=µ›Õx«z0®ÞÑFƒŸ£¬ÏƒQ~ÇMté¦}†N ]­M7/ƒš¡QSjV”A9QéW¶E»"ýÇÖ‡òÓ%ò²ìâë¾ ÝÀ ËmôØm^ÕýÝÅ–n ê{ôéEꕹ\U¹V’µL ‡Š`‹3ÚíDÛOUq™¬0ó¨bðbïe— äöæ ñûÓHØ’wz«Åh¥ Ò^./"§²«Lÿ •×pYÑãä÷ âêø½BGBmÝchÔUšªô¤Õl(²¨0¥&éneª¥j׊U@lÖö …Ïuß êÞpøHļø{$Þêïìþ2£’ÊžªÚÛ‰§Þ.>»¿;ÜÞNj6ÕlÍÎå‹K21 vØÕûë ¡€Ïë b®Å¤¼ rë–ÁÜs¿ÿý—èŽv,‚˵QÔ}£?Êã=Tò• ž¿GX‰¯®®®¬²ªÁH”G´õmÍýG¶÷-çâDÜ£ÜÔygý=úÉw—É5Øj0‘‹¦s|¨bôÈð _wŸ£b§O8 ݰO’â–Šµ—±™Óp‹fQæݵÈØºe“Â¥†:§VÂþ}ýõå¯V1ŒÎBÉ´ÅF Ëäøµ¿}šØ/PÿmŽÈ_4Xe˜#ZT?æ7môšeÌLˆä fo]] ä$ëß~1¾|ˆl.Ï^Öe÷ÆNu·EëbõØ”ðçç½.æøå¯R¹øŸí U…Ë7þx©;pÝgˆ ÙK×l+ÊÌ%“Ñø*„øœÿáðaê¹CØP]'—JeReHÝÜÔÙÔI." t`Ùmó4ü”Iù‘W+,:Ê ÛÁ•»t2MTÛ»%‡q|Ó¼ÿŸjMÁ«›oÇç 5½…aùŸ‹s}üø’ø„îv‡+`÷ÚöbU#~wÐò|•%Ç"£‹;tÂd7zý6O­lC/²ÖvQ´ä豤r`£r·<•±šqa “Óìö±Î:yÍi é@ìOj‹ÁD×0t‚Ì1}.­MQž ãÖ®¶v´aßÁ&||;Ó–`ttZÛð%ì¼de°É¨»ÁuÀ‚ÓQ;.hø‹Àxÿ\:óÕÿ¹ÿÁZoÀE:½ÑZWK7ýQ%¬#XhµÆšØ6×÷gël«ë‹’-ƒþ“#šŠå›±YWjü½aG“£Žr„1e}paMdzþX¤««©ùðñ¡GÀ…ãµ1+v®&ŒÍºÖç@͘{ÀœÞß)ß»£4ÓðÜ22_œ“-®"’Ñ›Úèˆp€?ô¶ùÿ1#%5´ö0«7$2ºo”¦µV#TŠ6 Jímiè “‘}3'ñÊõÙ@CW¨™u:ëqÿ?ÜÓ82Ë6·º# Øûl– û·+]þV:/ã¦Ú¬õ:¨€ªª=ÀpãU©/Ç[ï7‡Bþq—½‹†Ý¤ ”lÂñ*mj °GLTÝ)V_f§~5ú»ÂAAT§'H$ÿy¸›Iùÿ>#(3–Š©RŽP,±(ôù`%®¹ZºoÈâ’”¾Ü¢ýù\l àôŸ—nÞ™‘K2Aº»´¹¬±ÐYó`ëjµÜ|:«sÇK—D³þŸ-]uµñí§ ¨¾ãúm<.ì(o.))//)i.ïèhnîÀÞè7\Q#ZÏià£Jô¡³Ñ·œD¸Æ§¦K‰žä}Ÿ®ÁX¶Šªü5áÆ@Än#Þà±}hv—ô ¥.ÃþƒÍíÞ}¥ƒÿÂìyz¨vij×úö9jEÎ&|ÃñeÆ,cò ävn]Ím­1æmsb_P¾÷D‡»ÝÝŽ]@$§1ÞÝÀÿ8Ž#ó‘RèwzœxõœZçè›(”ªq[B«×kõNSÈHÅ£;Š_ÕF³ ô"­Wï·c¿ÄJÙɽވ–~ÇG«¢´“^,ÛŸS (+•6Èö‚¯´ÛX–‚e™ê$ë (lïñùvNÿ‡¿D“ð’vÿ8¬°·xþH+âu½Iºv¢ïn¿îçÏçBºM9wŽËhDÏDÑ÷ÑP9£üï'¤|Ì{[ðý„Åßó0 ¯o<·0½´R&!7fo|R¹„{ú/)©ŒÅiò$¶^k¿Œ:Rë× æ^ƒkp½çààá¡®óð\Í=–×rfïésp•àVr …³ÖZŽ}´‚Ñ9m¾^«ëº‚f݃҅°B¼lS‘Ô§°Qñ[@1¥ÇìèƒZ¶3°/ÏYÄfÚÒË—–æä¼¾2oKÄg®`ÅÁ²q¹¢Ÿ^EO£) :ÓS´çtIÛˆ¾?ÃGŸ H|’Ý‘TÀM×ï‚•PRO¾o§ßT DCc¤Á6Sˆ*¹ °ÐwŸð#PRR°º›+@Iq7’ö€vn—äÈ21ÿÐ<6‡ÇEùåìêˆÓI…ÜݯswgÍ5TáÞ e!Øéiï ¡!°b?CqYè…±A[…™3ñ¢s½‰×󊑹BÖasa ùŒ¬YÇXU’[ø—K¼â>¸(Âç·ž@iŽ€;üЯû;n]™®L#¡¶q÷ Ô0óƒÝÿm2`°TgÐZÅB ÂDà6'8l¦®{hš‘Ú7àÖŸÞjôwz[¢ŸÁ qJúÇ5}õ¦’¼§ŸZWl<ân&²y±!>§1‘Í‘ qò‚ ¾jä)!Øì~ E¼5:†V[Èt®’{ÚŒ]¿¸‡E0óò"”l4 E5Ûuℹ§ƒÈNÚl®æñ¨Ð3N4 =èAøUÆMŽòÜC£+„ëìËzà·€¦|ŠÒÐŒ ÑúÀû–Ó™ð4p“ŸÁÔy²‡–§ïµ.t-ʸˆº/ âÛâ´ð¢ QÓ8Š›ÍõpªYœàK4ÍDýHNrÃÜ#Âðù†û¯´é;…ùvĵzÃVÃæò锾P¨Í¬xV¼Lò²xë. *:¾piýãk¾Oa =ãº.@o¢ a]¡lš¹g“šÿëÔ™/ˆ%…ëð—ü¥$‰{Jn7áú¢ï.\è÷Û¿ÞŠG¾ž*RäªsHß2p£Ç³/ñ4kO4¾Ìï¢âh—å¿36ùê]ää‰ó¦LjqM™Âãý_@Öendstream endobj 237 0 obj << /Filter /FlateDecode /Length 7519 >> stream xœå]I%Çq|l¬›/B_ìRîË2`I¶`“´Œm”Íž™Ê3ÓÔ4Wì¿îˆÈ¬Êˆ¬Êz¯Òƒ‡y|]•k¬_,ï/çjÒç ÿ«ÿ^¿=ûÉïm<¿¹;Sç7g9Óô×óúÏõÛóŸ?‡'¼ƒo¦¬²>þꬼªÏI“ ú<ú8eëÏŸ¿=ûäðó‹K5™œS ‡—jòÙ›_ÀG­TT‰ü Ž9exßÚu‡«‹KkÍ”“?ü ¾5)8ŸïéY¥\úçÿ|æÍdU2çðK0X:þâìð?Ïÿ|Fó;í`™ìoŸ¾„±œÆg˜ìW–c4‡Ël1¹Ã?^$;e"L«Q°¶BÜMPÁÁ²^°÷ßÓ1ð,|Òʆt¸¾0·ž¯Ëð>ëÃÇ/ñ5›“õÖYwxL)Xçä{ó¸ßŸüÞ)~Æé ¶÷€û¶´o[¾´~J!”?§»PÊäãáùºll–dá,ôá-î)8ÕÃíÅ¥‰“‚/ùîhÓÚêœoÊk¸9üs ÞÂwË–¯îð5§”϶DÏÚ$N宜\óá†ìW0±Î!V‚¡‘ß•|¶‡Oñïp¡ùðíòR9Ÿ]j8gà à •×å^·CØ:È—+.ùÉáO|<Õİ5‡›àäÓüô·ØËÃÁó‡ý2v~úÏ»«·“²J-CC§™3Üá§—:Ä)éxøšn)e›Ë·ô‡óFX[{´zòÖ,CÿéIøçÃÏðÍ·žRÝA<nŽªP™I½9×v 6Ò>BZƒç2ŸáÒMÀ6ÖÁÙOð¯òe¦?6’ØÀ'£ìù¥±SÔÑ”—þ_‚Óª. &n/9 {­ò¼—Ÿ•4µ.—Z7©x<ãª,½-{ù›º¾* s<¿ÔÐ@öƒEÉö“†?ÂAñw -¥œƒ#Z2zÙJµÖ?À1BNðèÖÄ:ãÂ#¿ÚcSZæÇ_õBFœ„Ÿ¬ÓÊÜÿ$Rô8‰¿#¶2xíûlõ£]BÕ@(OîOÚuÑÝÏ¥FÖB=f(ÈzȔڃ”DLéó¦Ä©‰)a°ï”)Þ“'£ýÿÆ“ò:ÕWÁoó“Ã/ë ¢ëòbá¸bʺ»Øß% Rw7û¢Ÿ«»Y룲ºÚ`á#»ÚË2¼Y/Už{ž½ö´¢§_aå&W¸Z–WnJÞƒbïœ` ¤°!´6Ìa½ÇÕo{cQŒÄ÷"x@ÖzÚ›:Aw1Öüiw‚r†Kküä@I 1ËXØy`Óf•uÕò· &Õ\Ï.n_/Ô u=çý‰uBì^GóÓ/ȉÀÆ|ò‚œ6ànÑOR)š¤Œx¬hç`4ð(è‘ îB¦í´ñæ™ÖàшÍTžŽÜ»}_Þ‹œWøY£+r¸zC®J³þMù6ÂÇÏ™“pµ¸KWo;÷/h﬘†;Aå=£[vÞÕ«æ ò5±1^7WnÙlâÎ ó.¾iþ9ƒÍ±,ƒip#'è]{ïnåÍp™èF\,²„ËãŽ~-w¦ˆê§€€Þ´áÑOq1J?¥0 qBÖöpE+T&YâBü}Oé…Í'}óÙrµ•R`çü‚ÞÍd°òÂÚ¾Àå4“R(€À_s³6:Eð6O TªƒK‘QüvøŸMFÒàভs8Ahü¨géN¼kò-žÀ̸èÝH1ѬtŽLtÔñDWU1“Þ¶^á|˜uFÇ‹BÃ;sxvÚÿˆw>º=äìðÓ¾óg’½êán6bäô8iJŽ˜DÎÄ _0–*øˆøŸpÊJQh3. AbÑø<|K(õÀSâI`Y‹É?Ü?áõ-D`p$ ­—kKK…òWº-—mq Hr"ÁJgô‡¾í•ŒµéE´øäS*ƒbú™3óJ‡‚„M'o‘-õ–C~üû·…wœ FJq$x¸¸¸Q¸°Œ‚”ŠKïIZå¤ë„½ú”ÍÇP2Iz¨VbJöUÅ¥öÐÆT›ÍãÝ÷7©(OMH{‚“PEqd•´cxÀ,^36¼Eŵg”ø#+€Cu[ÈÄ«Ø1-é˜1 ‡0U+®üü‰ìTÔÙŠ¾(öõü —ßVZK B¢ž¿U ° ï ¦|>k¦ &O6‘ó{xCD˜ Ðl¾-…OñK%)ä´õœ÷n71-TÅŠ »B8b9lÎiDtåã§mÝü ΀DŒ@&–”*n9íëÙ¨‰©º\&ä‘°l=Œ…ûüç©YŠÓ ³ vI¨à¢ÙðS=JpоÀ«®M½˜yfã×£CC{+œÅ‚EËaXÁ´݇H˜1QŸ3òd¢¦=ûb;âÕž%¥3¥Oªì<\b_ÊIö‘§ƒe¸ÚÎ5µÜ›Qt|ìêøI~y±|ütÛ·zÙ O Z2'5û°wH’Ÿ´’Ú’qáÔ>rßBÓ¯ò; NaäÌó Ø-wzW ¿A0Y³¬Lf º3a9ßú³ Gg>‰ŽJ ÍÔº‡í£=ÉÊ“‹9œ‹w:ÇS¡ã™kø¯\Ùÿ‹¹£hâ¸|Ìé04TG`èØá{ —€Å•i<„ÛQ Y™­cŠKE–èM3Jg` ÇñæÖ"ïϺoËGš?"0Å¡¡z €’ÊS=ò|d7*k'¶S@àl`þ ¹å²hÓMãº>Fj4•щÿ¯ÛÒ˜@‚ãj`‘wîHg’k~W–ŠFq/‹„‰nî7®3Ñ×8猈WÍO8§ÇšµMXðÛDvÐ"˜I oùE­(I2’-¨Î£¹*cÛê±ØgÒ‹šaHgÕýœ«‰Öɤ ƒÚGà µ°Ù*D»‚C³ò4Я‹µâ¿¼kfï@‡¾/$µÝ‹|x2F»Ü+äZ0ÊÓ†7 ‡ªEì†ymŽ( ¡Í`ö«/zÁÖã¹pyÁê~s¼¥«íÂL· gX¹u›%ïŽR™®.0¯W¼ÛCvhoÌ!›¡…ù8™¸À|Üb}Ó$à™Gk ·¸V‚+Ôõ»ûCå]l¡Ü¥5±‡ÎŽ9 sCà;:Bó0Úg#Zc¡xo±/× F=aàѲ÷HÃh¾2wì]â2ËŽÀn©²b»ê³#PIf‰:3˜ÇÔ䌟´>rÒ˜}x/³.ÙêhurŠÕă÷Ë;GÒLDÈi<ÑC“?Ö¢Vk!j«Ø|vÂç]1+s+Œ Î$6e:›ŒDüP57ø÷ÍÌÚØ(Pʳ±"‰‰2®f€Jø+žR#êøµœY}VšÌžrO=|³h‰ß# ™J/1Ó1PÀLØŒ±V …`Ngs5MNÀ«ø&Ê!Ÿ=ÿÛzÉ ‹ç5€ölZÌŽh&ôÏŠƒ Îå’%1f #€ƒÄÐ ,$kÅÓÄ€YõìPˆÅ#k«’©f¢ñ ƒXDœÂ쀎p›QšRq1À/ñfF:ô§gG„Jp'f×GPÚ釲½ÏÃyÆ9;’£rÞnnÿ‰ÙžOt”íAìÂ¥žbbu|_ègap‚qžþ Þøe¿È ÖWqƒõyfžäýKgu‰UË&x†øJdûXPLÔ Ñ.™Iú·ÕZ²Êd0s‘!5@k¢wÆ÷ààÌü{/Î×hkl^sc|LÃä.tŒ¯rÁãSM óz©ë …Kãw8UÓç|¥¤D¬l3*¥„|‡(6œ´˜øia =­#ô PŽÜwŒÀ$˜È9ÊN¼ühƒ{x—¼†7Þ“#ú•òçôÿ•‚ +沋$;‘Ha/9€§$s/¶ó²>Œ¤™3b$Á¢OcÿTy ý§w…JƒÖ¯ª+–g-gA¡[]¹ÚŽÿ2Øg3çmΠšìBÉ4b¶¨lzDgapÎ:Žußr?®O×ÁSÉNøU|<ü»“=î»ß Ó¦ï ´s²„™¬G'¡ 3•ÔÉŒ™c¨z7¯UTÓ¢+±$t5Ý6’¾¨$\q99/­‚6J$gï/Q™vŠà}ð~”Ä!Aõ%µe?ÇÉ¥)øÓI+9$Zô hŽ`TÚ<[ Qøçx‹+BZ¸Lfºv„Ôd›ŽphÞo#–ˆ·,íêc##üDZ)2ñ4bɱØ*5d÷Ø+#iä| ýïQ…¸×D­qn.Æf* K¶vùù6p*ò•QòߥÌá3ÞÎËl™Ð8ŸÍñé)ÆE¼˜'€}ÖøšaåoÚ¦å½ÍK©^"r²©Y¶3"x0´ºˆ´&¶L/±èƒ‰{e „áJ©Oˆµ¸EMMMahØOyLÖbÉaè@ð€)2}ÒžùXûÔ\ÓÍXÿÀ]Ûl¡ ó\Î;–ª£/ûñYcÍä[\vÈ Uç‹2iW-ÒýuS>ìyØŠeçʺ‰ fì³¶tÀÌž÷µÑÔšº’AŽÚᩲ»¨Ðá*•ŽvÏé7¯é5%=Ã;)»¶òæ*ÁaÝ,Ö KïíHØ8ÑëÉ]Þ‘º¾lK´n#1>YmA" Í»†C¤}Ó.L ÆQ‹•™mCdOcʪÙfâs—Ø !$:ŽSÊ­Ö!œ\ ”’îîÑA߯Ü:éÓ²ÿxÔòÎ=«o^nÝüå®»…‰‘à{84Ñ:]ÔÀG"X8h›Žð×÷xPOQ—.joÊX©¸>þwÍt0ÈÐù°T9,9 ŸÅ/¼R0Ä:}‚`P­‹ Ü€Õ³é#eÂZƒÐåE•³Õsê HrÂ.æ“&ª'6Vi >MA1á¨úA‡²®Ìx–ÌË-R¦Å†IÌóŸ·o¹µXò´àlš³dø«ì5ˆ—-^Züø’U5B;—•¬+µ ƒ^_ÌŸn{®öNbw…ZUE/†) Ù»Ð%pÔµmôH2{M’æ¥]ô3é€@( Eèâ:ó$o½soA•µ³¸]U\£mo;3Äg: 1YÚ=ñR­âÍ$8æ‡t¨týö¤n|«‡y©`ù¬+ž>È0îW¨>ú˲,k:÷ËáJ½·}{ÚRÖ÷.V¡ûPy‡¼ë'«,nÕÊwšŠÈçpÀS ó{’ÌÍy¨‹«°Šèr~ˆrŽdpÎÛã)ß•n£ ‡÷3Ä[Ò^qp”y¯ Ñá轸’1€ù‘–)Ù´ÛԠιÕ/îz™³ÝÕ@Ž /”d‚Ë„`öŠ|\*^ÑŒäÇ{¤º:a¸æ!}Ÿ¸×0ìîÑE Ìe»˜Ï_÷æCïÊ€TÔö1 <Þs$Á-ËUõ­& bC¶ ¸5¸D`'ìJ*ë’! ÇÉYb·›‘ƒÏ¹òiÜTž]•¹j”üÖ Qç®…˜`ª‹)HQÇbʬβ?ù„ÍÝ}òãàæÉwuA†ÒþX) £×lÔãøv‚³ñ¹Ãø¸UöM€@JÌ"¡‰S|ú-Ð Âf×Öùr‚ëÃ*bB‚S©·éó> ™Ý5d…iœæ+H++¦c ëÓ¢ 0ƒ¡ôpAhÀõ­Dpm@yŒB ÓÀ÷!ïbÝÉR’'*YHhÙYªƒñ›—ÑG¢¡äï2pšîâ¿:¬V“¿(Z}¬æ€ ˜õÝ&M’ÂÃe#öÄ!·[Ú8+ ^½Y î¿ãIïuq”û  DK“Šv™<Åñ-œ&"«DDþ«F\á‡Ð%-´©ô°þáùÙïÎJ‹Yçêý¸¯lçÓÔ¾²Úkì#qn2ºÊ¥­l™¶ */ÁܯJ^¢­¼CéÍëºÉPä8@Ù»;ü†¡£Ì–јºËѧá Ï]°¸ï¦ hèJ“o}P?å¢Ã«GË3D½¯ï"Ì|@¯j–˸¬'Óq·åÏÌ©(Ô¿}]M>nÔË–¿×ÜðlC¥b›_¢$û)º?â ²»7“-fqÏYð5]‹Ô$=ZóDŸÀKµàýcM¸•þ U;œœP]/šÆp”%…©´½”ƒÓ÷HÜ¿ìY¿\Ûv/Ãì±=‚GY tžý£è\åÉx-÷É¡¨Rç´ër§pX1:ˆ“²Ð]Y¯ÓŽ++™¢BØ>î*ƒ÷–š=•ûFæ§qÙæµí{-Klg‘ Í€Ú+µ+gוÇÔ·/ê;Þ•‡=uÓ´ [õ¸Ñ&eH%ö£FùǪÖ8[•fAaV§xåv-±óSl5€•‚&ãFâb˜¢ƒg1y Øè UØfž-Þ+ŒpAbŽÊÀ k{“kÊ<4·Æ1q•‡úãr­¼±Oòmå¼ë“=©g—(7*4©ùÄtºÆA§ 3Ïóð":Ô^·î•;[Öö¡?§‘-¯÷jRÕï CõMHꦷ%PÝ„ÂQ«.FJaõÛ9Å´ªöbÀ2#¼:ôá!Ê9Mõ¡8 Êùº~Üa}ïãnÊË‹êSl)Î}üž¶ó·­ßÆ’u”8ßÙ“Kó®³L”QT:ÕCÙù¥¨¹æ5ôÂáJC/TÞÏG`Y3|¾•$Ÿæº-î!ÈÝÑv¤»‘÷ˆ]P—PÖ~·\TRj£ÙãéDÜ×U÷ÂF—þ]/;úªgvÜux_ƈÊM: Ue™…Ó˲R›»®• û¶é¢íVd7EwÔ%™ö©]ÕM­g †??Ë… ¥†XºoœJqT¿ïzÙ…>•q™ïU3 «ï ç¸ £·´VQÖ=PeW£Øí2nË×›Ô[(–vQõ©®ËönÊ ÈiÕ=…åK‡uóê'9Öes¿ƒA2¦t¨ñaXV‡ÅÖ~VB8Ëj†bŽüË7å-l¨VrÔñ’íÜ6½ïfõÍe‡X“NãÄ9÷ÕÈéð÷ó­øQr(þX²ìo[®¥›eîc]±ƒVI%:YîEGDžûv¡¯-Z¥ÿxÌ-_ìšY³aoäÉol©iÀ„Ÿ–Ý?ÎJràa¾xg¿v'”ß!QlüqÛîˆY…šÿFÈ÷Q?ù‹z6lšV6`LiɈ¢Œ£ˆFÇ!øƒ8’Íx…yLçaEÆöá¿·r£°|Páù\²G7.zU‘£ZŠg ¯^ÿ‚L¡õK ˆFZPŸ¥æ¼ëàŽ6ÖÜ%ˆõ—±º"v‡k–µÝyrmò›­õáBK‚ì¼:ƒ÷ñ¨ît8öC鋌 z>{•<˜²öìKyÓ@âíy‰Dœ"&Ì`}>Þ™ÁQžV©ËÌ8Ý\=ÖÚî,26~±ÂQUh+¼–-¦AúLsØí$(|€´ÈܨK¶92X”ÍÀ~_ÀMi{£(¨y¢ÝQÇ^í(S=€CÛeq7Ò‡Ûvˆ*Hÿvº×*úDBX 5w¡?„å{¡ÖH_”Ϲo µ¹ó{Ý y‰†rãǶìWÚd@ ø^ì‡Í ö±÷Eã<¿@rö!ØÝ+[”øÉï [õ­¨ëÓ}¹ž,v(Áú^Ò¢Ô®^¨÷œ~W÷YòV­‡©œh3oý ã• Múš:ìqðåºI&Ö±ÈÖ’¸2º6VÅ÷mn¹ƒÌd ® ²]k *r°:è¸Çꯆ³ýR§ôéê¾`É" 󀩀|ñÏö¥¢ö䟣sÃük[H2ØC¦yÛÝÓ¡"¿w*^_…3vzجÃ|­«p6>SB¡®nñŒaáÂe hP%7½²ßRÆMÆÚ,‹ÿª¨#ØGZ™Žû;Âvløyïa86™yÕÜgS¼ßÕuñæY%-Î\iÖû¢\'_œ˜Æ–ßÁw¿ÿ„_ ÈîŽÌÊ6N¸T÷ðµTû~n%`cIjö¡·2sàœ;±ÃzJ A$®+-Äç"˜ývE¿ˆQ¢¤/ùM4ìûŽ1÷"’Òw^ÂpÄ%tØç¢õ'½ßOï®Êúû¶,Ê0ßd) ÒzöXÜSúFT‰š¥o´.dYÜ““<_#úŸàù¦ôžï°|%zÃo<­´¦S`âôò–‡^$Tß*¨o^!ZG5\r]§†•¼ì{±¡õmw²»Ò®vCGl…kÐ,Vqò‡æ– šs2φ%“¡Õe°ÎXc£±@–Yf|Êøbu Z4þîGñ ©Å•(ôÀa³'…l£Oä4J‘ö³÷À¤–ÛûªloiÞÔ˰û’–}6ö!) kÃÒÈ¿Úe`tjGøýBRÿºéXúíÏ…a»ÆcÂÒ僵u2ôK±@K%ŸdZºTöèqh \tÖ%Jäó½+„†±ª3TòÚ¯l>ñ¢Œ€w¡¯{ÓžÑæÀ¾û)Wܱk¿&Ôq-b‡»“r%³A>[­Ï‡?`òªVV¥¹éok,Uuö]?=±wØÀz*¬ÎÄ—ªÇVQI ‚/c\mtIb?D?ä²ùóä¥bµ—ö+/µ~Ío»v›é2¬‡ÊÜÜÀ ÷q¿~ÎfbÝ€ð0ÖY7¹rr·³í'¬¥iGKÌÞ-‰vpXÒÛϹè!ñÕLqv„1ïo S΋r®ö¦l?$‘x ôAð'GŽ'Ò°ºš1ÈB)ª¿;û_p)*]endstream endobj 238 0 obj << /Filter /FlateDecode /Length 240 >> stream xœ]‘Anƒ0E÷>…o€=($4›d“E«ªíŒ=D,b,‡,zûÎ MUuñü‘=Ÿæt9_ò²Ùæ­®ñƒ6;/9Uº¯ÉNt]²ñ`Ó·Óg¼…bšÓK(Ÿ_…,hÞý5ܨy‡¡ÓW~Šk¢{ ‘jÈW2£s8Î3Êéß'ßîÓü'*x@VÀ—zÑ€ À QaX½G ó¢TX¢GTÚÈÚr°Õ°'Q>DpGQ¶æË˜±ëPq®ŸD{TX“(ßH`ºäsÙWš{eã£Vʛ֫õIkK¦ß?PÖ"S–1ßwz`endstream endobj 239 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 1574 >> stream xœ“L[×Çß³Á~$ÔY#½Úàç¨ÓšŽm‘[ú# j˲PQ<ó€ŒM±còÛ`샂 iCmÚèB÷G((¬¢­P«6­¶®ÙÖVÚªµºI“¦î<çÒi—¸Û²ÿ¦÷î;÷ÝstÏç|Ëä©–eµ{Ô=Qn\Ù–+÷²Ê:•R¢¾ùã¬7¿„Y³{gI…°ã„ ÕP˜waækñ³;qi Î~‡YͲ¶î¾ø3^¼²°ôÁ[§Û.µ¶9 Í¥†òªª>`¨0« »:D»d6Y LŽ6±Ãä ?CÍ,‰·acu›Ãѹmóf—ËUfêè*³Ù[w”>`pIŽ6Ããb—hwŠÍ†½6«ÃPkê 9ಜÙcëèìvˆvÃ[³h·š%»Ù"vt[R§ÅÝjMÔ'>Ùm²Ð䤯jE!Z’µE²Jw—Ô!YLön«äí]&KîŽf›ƒa˜;Z޵ª˜UüåE•… s˜ibŽ0“Ìólv ±a6ÂFÙ££j2y̳À>ξÏ~­ «²êCê×òæÍä¯É÷+:e€„ÓJËyEJ±hyO3Ê7p_¿~_yéCÛ¶Ô£oe:讄îþÏ !$K<ÙIJPÀí¸ ¸«©-!ÙNÏ 4Eµp —øK–©³õX«8ey13ùÂŒ^×u>[3®!÷&µ™U™ÕúUywµx 2OÒuÃü M@oõendstream endobj 240 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 2803 >> stream xœ¥V pSu~!¥< Ôƒí®¼tQP{0p ¬ÑDâÐØÈÀ» _,eÒ夺Œ ÆóÆDÃÿØø~¯ÞnÐÐ-EHW”– Wê½Ñ¨'âí!ƒûFÐ,8Ž?¸ve~@H&Šû·BHdÊ]ÒBí6 qÛìñ[17á;:9xÌ%©W˜äúj’Ö7¿`VídîÎ27%`Æea½?ôùÜ„ÝåI:»ðÌI©œIê"­:ÏÓ hýÀå«üÐA¢oÒQÚ‰ŠÝ»øŸÌR’ù>]¡fÿõAGb7(Í ¥&?§x­I«}?àh^:š}~“Y­b}˜I#‹7?ò3pfiºÎF¸ÁcóØ\xfŠ™fÍ9ÇMÍEžI^¸vàMƒ£'.FŽ“Þ{+„ð‰â¾íKv0‹4Ó¥º¼v"öåHbð`«V&ÓÉÔ•dé"] TàKߟüàÀ)§ƒh/U ÂD_u;'À»Ìn†6‰„z›"oà2]8ÞÑštõ®žà)G¨óÃ,W¯o \xO£K£ª×I Df*ð̹çPé·Ü”1…ñ>*³T$+=†V=m ˆ ßþ SÃÌ*©Œ¾9Ñw¾ûáë·{XôÇŠ÷üúχ¯¢×IYôZ YÀÜl®†]8Ãù²ñ5´òm4þ!‘9i¿vñ{è©sh3{·]๺ۆBXi”l×ê|³¼9O'ЂLV=˜pKÝÊgÊŠÚß?=Ýÿ1áý&×€he z™AA–ߥx ŠñÅŸ ϶·÷ñ-oj†á" }è?=€n…“øµêU.½ÛD´iû$PBÑ2Q¾h[uå6À‹«¢oÅÜñÀq2ó‡[¦õ•J¦îäYl&/XÁÖáóºÏŸ@ØG‰®ñþ½]£5`…ÅmtÀ"ÊxYzH@g5@¹}4¨@n±j‚Òci§AÊÆ&Ú¢7dQ”Ùl¢®,ceGg[m­ïAv§ÂV‹_)—Ò9ù|ú= ¶³‡˜š¯×-ó‡+ÉiÉg_•¼Ù®õ‡»mÖ0‰xéí–Ü„PPM,[Ù`£là—»Ç޾ݓ…A8Ê`À¡o«wè²]Z YŸÕí´´h§«Û½ûlÖVâDú% (?àŽ „*0’åVµ[™dmð•6¼ˆr.ãœþ¯Ñ…¯¸):u¯¸….gOªïhŽD’ï=ß¾³PP].#êæ‡Yí2œÌM¢«­ôÙœ>;}ïÔÁ#€¿U4j¶ëkHƒ°þ!“\µ¥¢”mˆ¡!p  †ÀŠGú:¡L°ýxüèÞWûCD¤à€ì k]Î)tSü?óˆlÜ·eàÕÒÈP§+ÊvÓÝùÂ'Ä KÅÂ0%ñÄ¥ýSR}ãëÄ;ÜT05ç€ ÜS*5É,b5A›t¦†\Æe¡JV>9móÉ/£'þ—ͯ×çŽEM[¡pÊæcÉ®h7¡Ú¬-¼\½£L\M °7öu8FYt+ü õ•M…Ãeï~‡f}‚æõ±^.ÑÆQÏ¥=qÎÈǨ‰eúÞÔ|Þ“z4â˜Þïw¼þú3=¹Ìí Á~øËN?ޏè¡O¿õ·jÝF#EMÄc÷1i |KIâÑÎËco#LJ^°Ä„,!ÃS¶E‚QÎðGÜÔÀä2žÕf÷³Ä;Í­ó.Š–›Íõ´ÖŠlU|Á¨3ê«MD“¯ëo„lf.s+C2Kî8“û×·&Æ;Cdhý_hö ˆîótübž%FBõ¢<¿|zž…ímv?ùÄ‚mëójß.:2Öû¯È¹ûõCC?~Rc–,5ŠIu!ÐÚv|öÓ H¤ÞåùEºÏÙq¯Çåw™ÿıѽÁ>Ê£Ýb€B°L͘޾Èþ¤tpÛ¦ü⼫OfŒrFÙdüeAÄú”þWj¹ñŠÁ¬„f¶–œñšÃÃÑd¤%‚'Á‰»(·ZG™,M™TKó&yiàRC´{Zµ™?üyzåM®¼.LÙ½®n›×æLœI$c Oð½õÛ¥FYKi7m0Ö5çf™š µ,A×§œè—Y¿ÉfÉ_&µbvãÈ5’-/üf$[ŸW‘SDè/<ʇí°QùâZé—ã…P ’†¦*6žüßñ Äå\B\.*žà%E‘º©XX¯ïíïŽíf3Â|¦¦ =OÝÕÉA&t™ç {,Ó]JoS•¾Ze&òʶ& ³%FVÖ˜Ÿ°{v£¹:ÔIå8|ÇÎØ¿w¾t´ »w“±ï\1'ëîlgÄûÊÏi¤¨…hX¾‹¹I[Fë2yv“Z(ß xaMÏ1OŸ«›d+b^êBkتPnœ‹rƒ=f¿%)®lՉ¢½ÝÁ€•­uõ´8ØÑýæÜ”ñyÿ.€o5:tj‹¹ÁHˆç>NNCO;I£uLúàÓì®\ÄÎÒ…Ì¢%יּ|öýž“çÉÈc'L!À_KŒ‡} ZhÊ@‘ÊÒ*ÚÄ5%˜Q&þD5oÝsJؘ_Ôwa ýû±3ä«Çuï‡ ØSîß|Z¿Êµ ÏlèJ­õ¶µ¡º®t¦Ä;«ÎGs‰9ivfÌŽ;220ìßshŽ> stream xœ]=à …wNÁ HÂ’!bI— ­¢¶ ÆD DÈÐÛ—Ÿ¤C‡gé³ý¬g6N·ÉšHÙ¼0Rm¬ ¸»#ÒWcIÛQe žT*lÒ6Þ¥<Ò´€ºòCnÈž¼«­¶šÀ)ܽ Ò®H†¦ƒÖ‚ U#^ ‹>7y/Št¯[‘PŠ¢„<#ˆ¢„]¹vùòáñJDám,”œ9ž±ø{Õ;Ÿ]4‰|³q[Sendstream endobj 242 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 1024 >> stream xœU’mL[UÇï…öÞ –n,4+óÒ^u:HoÑM’ "Cyqcn£‹}¹@¥¥åö–®à\lƒû\(clq,` º/òecjݘe J2ã4nf|˜™K$jâ¹í)ÄÛ L?œœçåä÷üÏÿ’Ð$$IÒeÕ¯ÖåçÅÃ|åiRa”ŒD`ÿ¾Éj›g´Ħҽæ›t‰ ÓLEۂ…MèÆf"™$ݾîÐðÔÌW ÷ËÜž€àhl¹L[—_X¸;›+ÈË+äö¹xÁa³´pÕ±‰wYD5qr‡Ü6/¸Ìâ&Qôåæúýþ‹Ë›ãK²²9¿Clâjy//´ñv®ÜÝ"r5Ï­éÍYÛÊÜ.Oä®Úmç…«`±ñN¾An„Vq#t9ì^Ÿ+.@¥ñÇŸ4¬ŽÆFà¶ûl¢Ýáõ8-sÿæAÐ/ØV_G ¤Lö‘ýdˆ ÕGBC¼Fœ# ò½„´oBDéÑGu«ÆÑÈïaRICºDEŽÕàÞ:„ÊÑ–/È2Œ0ã’,˜:(ÜŒ™|ð&VJ·ÊÒ”é5%Ë“æ8dëH¤bu«aB–ÇG$ùl¿Ù9^öåÛ¿Úðß…%F)¥'%Ùcª¤<’ÔjVA€3ï⃨ÓŒÊae”¥~Y—“ÆF’ ÿ¥Ñ€RêлXD,jCŸ#+ dF­ÿ»‡†ÝlÇU˜ vJ K;'ýß¹þÂg{ÂP Ž›Y}ÿëhÆ ©Ʀ Å*»NI½p6=8#¦È"õõiJØ[Q³«¥ãL¹ï&ö …–•}ƒC}ýJ‡$ŸÂ?*µF…£¨½{¿ÁLP/¡´+‹q9N³>Ò©~ÕQ6iW˜Dî'*/²Ñ\ÌÂ÷U?Û¿¨½V>Ž>úm‡yF¹CÃM@ièÈãó!èƒafd߉^8ÙjÊ|+«¸ ë{üÐ LÌJƒŽJÁÎgkò0UÕ¿g´ËÒ¥®…†Çeaœ¡‚Lì¡êòóËÙhó± ŠfT†±$¹kÀl cÃí6´½÷\„o=«Ya£=s¬V7G"Ú©\F;YÍ µh˜£qºóÚ-stž¾ýÉUd4ͱØ{˜FéÓïTšWçé×›a£é°â[?Èj>V.³ØºÞ¡õ¬vB%-±š?®“¬æÏ8¥ ;ªËöçžÙiWÛ™¬Vü¦jó2¥œS žWmHWèï¾kˆ¤•«ÚO õÎu-µÿT?Ññí±+Eªá8]ѪÃ($©ç´ÔÇ!¨L®4cñ¿N„3 ­½¬Þ¦XӳɳO™’5·.ifP§SW AüX×ô‚endstream endobj 243 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 2426 >> stream xœ­•}PwÇ7² ¢íVlu7¨=ÀRE¼ŽèY±Z[±µh©Ð‚@yÍË "o4!!0F|AäÍŠZZ¯­¶WZZ­=uìû´^[ëoéÏ›¹E§íÍÍÍýu³3;û›g÷Ùßçy¾Ï÷' ¼¦À{Cô+S˹'ÜüiÜ!àšÉºÉ2ïÄ#ë×-g×fÏ?!øyïÛ19üQÞ#(q!2²÷oPf¨R÷ÊÕâ`iˆxùªU+CÅáaa«ÄÏeÈT©RI¦8Z¢–Ë2$j~¡Ç(¥©2u8x\­ÎZ½lY~~þRIFÎR¥jïÚPq~ªZ.~E–#SåÉ’Å›”™jñ6I†L<µÍ¥S· ÊŒ¬\µL%ŽV&ËT™A<–©Tå¨ó$ÒdÙ^yêòð~68ä)‚x‰x™ˆ!v;‰Xb=G¹ÞÐxŒYYêâ‚:žQT1*DA·BCI•4Æ`,+Óƒ PY«iÌlL ‹‘¯ïêͬѳ'3ìºËÚÁ²OŠa3•œ°ìO™Ÿž+atÍF³( ”ïcñ,R¥ÍL¦Ã‡³¹±Ùl>“4hp…æô¿7”ùžÞΖWÇ4è@Qoƒ“Tw{ß·È˼JY͘Êëôu@µBÝa~£ËùŠÏp£«M›@ó{„h·F!n‚¶*Dư¤Ð¥zJùv’Øç‹‚Ë'ßú˜¹ ØInR¤È¢àf+3“»XjãæºÜKè3Úqáhs¯‰r“r½’  âAQ×&ôês¡J Uº2¾—æì‹•Ípªçuƒ6ÆMîÑÇÃ>ȃ×!ýá7oƒ¹²%-Á?Ô¨°€êš›¨)¥¸ªÓý•Ó%迎2'Æ{„\:K:ñ\ú– [ÓuØÚé(jË53½£À3_¸¼ŠUÛõQú™+d…Û!zæÙïŸ9u²1Ëœ²~ nÔ~ÉžWì “Š4y)0ÐÈDž¥—¬Ê–Æîö ½wé 3O1ã=.îoní¦çþœî¸ØÛàþ[:x22t,PC¥±¢J+ÆU˜B--¼djç9zLí> stream xœEÐßKSqð{ÝûM—VpƒPïnP¦"þ*0+$ -ƒü‘?jétÖæl»ºNÓýHÝÙÔ6)32n…FÅ‚ zéÁý QáCOX|7®Í z8œsÎáœI$&$IÒU5åU…{e~4Œ¦'D3d±šXÙ "µüRF‘¢ôJ Èe O|>Š•Gps*n8LÈHR×m¹¬ë2ë;ÚÕ<—Õ’Í–”çrE%\™V¥ïhQvrUJ^­Ò*ùx£ájt-*ÞÌe]Tó|×ùü|£Ñ˜§ÔòtúöÒì\ÎØÁ«¹j•A¥ïQµr•ºNž»®Ôª¸ý3óöSE·F¥'"¡¢š 2‰3„,þ ‘H¦'dD‡Sby â"^Émñ»ˆÏŠ;¢ Oã-f ö„5cuàt÷Aÿ 9˜ (ð-€`±¸¬vÖ)1§õ.„ÛéH–Ø0…Ó Rù¸Òmw[…½aNÑÏF<vw‚2÷‚Ùß§1{“bw„n¼iµ€Å,@€†¨ LÂè¬Û½‚/À8š²yí}Cý.Öù`PY~§^ÓÐW ¨–2˜ ¢À¢o™[´%hzfÄJïÁ»è, bÚÒ̸I=ØÎÀ4,Ûè75ÀÃ=°¡]3 í^㘠 |X†iEµŒ›^B ó×ã¶ËŽRbm°‡§!‰?D|CÜŽ»,àó‚Â9’u²w©uÊzÜûpžÂ,ðCÀçôÚ<Š6_Ó¦aÕ1Ù¸y[·W€“VH§ÃR+Eþ#Œ;÷9ªí¦8ôwÓœª‡|ï¸/1+/G7à „÷××m?5 ×´ 'MHªÃ}Ì<ÿüõµZJ’I*b¥t)ÂÄ_[ õÐqÒPST/ÁŠvÓ0]>3ÚÁ«Œ×=æö=~߀ ©òË9”C¥çceAJº;M¯%­%³I‰Åóòƒâ¸\Cñ~Y’endstream endobj 245 0 obj << /Filter /FlateDecode /Length 175 >> stream xœ]O»!ìù þ€—‰Í…æl,4Fý– Åá¸Â¿—‡gŒÅl2»3›2žOgï2&·ô2¶Î›kØ’<Áì_p€íüª wÁmźIkT’ò3 R9X+xówÝ0ÙeçŒËB¹ì æX¨(— Noßv_}\#ÞRŸ[–³Æs¾UcˆÕ… ÐûÝY[endstream endobj 246 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 473 >> stream xœNÍkAÉÆ4¦VÉ!Ô$^Z(IZHýx‹‚õ¢'—t’¸»éfâRªù2d·ó©ÉjhRAD<ö’sï‚ÿ‚ø?Ì,Ûƒ›òx÷x?ø=¼!ô¥3›/ïÎ]‚/A~ÃÃ#mÚe~z!?Œ¬Ç6ªAhÀ;寮±íEöfá«À ¡Rjiµ¸£òAËÙ´–JÝYEëÉd =’±VÈŠ ʈD²HÜðmªÙ&;hù¾DHñ^"¡ëz\”KqUËo¬¬"½@$ô—°öo¡'ªBÐSQÆè|fü\Óª\,¬¡Œº…5¥¨d < ÀðîÏ·yßOÿr3hÿnN¸|Y’û&²¬vüV}TiF­u°‹ËU­V»IÃÕZ½ÒèÃzŒåœü¼¬TöZ´®ŒêÖ—Ng4Š2÷š)ŽÒ3ûf††#ës«K?ÅŒ¾s›Ýzqêwÿ2}ÿÙK‚ýŒé¡Ã)=>.ÓÝØÙßn™2¥‡± ;9™ðí#ø“s%Ô;èèÐÿKžªÙšôÁˆ:×ÏÒF•št?Lý½¡9þئmÚ¡}:øê–¾Ù&>'j]œ]š]ž—WøUsÆ©endstream endobj 247 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 369 >> stream xœcd`ab`dddsöu°±ôH3þaú!ËÜý¯æ×®nVY~'Y#y»T¾næn–yßß }/ü^Àÿ=W€…‘1¯¸©ß9¿ ²(3=£DA#YSÁÐÒÒ\GÁÈÀÀRÁ17µ(391OÁ7±$#57±ÈÉQÎOÎL-©TаÉ())°Ò×///×KÌ-ÖË/J·ÓÔQ(Ï,ÉPJ-N-*KMQpËÏ+QðKÌMU;SL:çç”–¤)øæ§¤å—æ‚ ªL­(a``` ``ìb`bdd)ýÑÁ÷ë뿚¿w2ûUÏücæŸY¢Ógt÷v÷u÷tOhžúûÕ/‰&lß…»oÔ쬹Sp=`…-Ç*6ƒï-¬ÿXÙr»º2äˆýiùÃóÇ«©¡«»»M²vF÷\¹_¬lϾOfÍa+vswúÍ›ãRéÜáÔÍñ'ˆo'× n9.–Ê|ΫxxnLçáe`Í-†Ùendstream endobj 248 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 1371 >> stream xœ}’kLSWÀo)àU‡›Þ[?812DeSŒ7t›‘:§f<œJ)•–WKéƒÒ´ôÉJ->¤U ± (¾p°,1›MTŒ[Ø'Ÿ˜ùa;­×»À²,ÛÜ—“srNÎùü~ ,2c0Ñ™lö®ffëCï1BË#B+˜@å¿ÚÞµ[üÉöÈm‚XˆaBLdïò¨ð¤ˆCÅ‹ç-ŒÉ`”Wê2…"…XPÄ—°’¸kXëÓÓ7%³6¤¦¦³>.ã‰\N9‹Í‘ðye ½(eír<‰‚•´•/‘ˆ¶¬['—ËS8e•)BqѶ5É,¹@ÂgíãUòÄ2^!ëSa¹„µ‡SÆcÍr¦ÌŽ™Â2‘T³ØÂBž¸ðy;¤ò*Ea †­Â¾Âr°\,Û‰ea8ýG,302"LUdJ2Ǿª†`èeþåñ~Ž×YôÛA 0™&“ Vu xÅD#’ \$B)üwj3ùúíh¹Ä*7ttø@ßI‚ &%þÔ߀ÆaÀ+͹Ÿ"Š5TÚÚ•dö—©ÙÔ|À©¤è*ú‰+–j¡w ”>É )'¾”ŠÓ MJ./±V[R‘øÇðCLjãüŠ Û|Ç ÏéË[•G%ÊŒM:—Ëz¬ÅJœzzÕ׸׮‘È4‚ZÉßXsŠðäÛüÉ›CC~?qáå\NС,ø-:=ãCmÑëÌdyÚŽàœ3Á|qò3ô8´4¾ÑÞÐÖY“mNø¿¾Z»­ö™¾rûriu Ò¿©{>vâÔúrÂ…1Bs@Wx0[!?\ðè ¿Aâi>„ü¤´­J\,?tp”;ùZ:‰âúˆØð²ÙxБ+Œ»SÌÐËpr|CÝ Úåõ5õT¤.¡º^S\éw¯Ýbtõ*°˜5²•ŸË3Õ'lZÅRŠ Þ¥VSØ=öÔ‰ >éɺQ×gÀïõôùÿÙùYÎ\çn'XmdƉx£,jÕvN­pnáÉÛCÓ ²÷Ú•‹€ßõlÕ“fuTl8I@'®YfDýt}š‰¦Þ„]‰«œÐÞc7›œԀŤ¤X¯ÇLJ0‚{uîÖŽ&“°;Ñ{à†ñŽöÒ{Fs­YE7S‡ïçM<ïGï7ÏEa¨6Ú"aF ·µCƒÍKÎ à€ûiÝÛìÝîÎöÁþË#W‡¿'8Ì-Z¾^%ÙlÆ]gÚ/—ò²öä"²—VUã±²ÎPfk[uFSù­óúL-$DnòÆÌïoЉÁ°?ÜÐßîendstream endobj 249 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 401 >> stream xœcd`ab`ddduö 21ôH3þaú!ËÜÝýóõO]VY~'Y#y»|¾næn–µßÿ}OüžÀÿ=F€™‘1·°Á9¿ ²(3=£DA#YSÁÐÒÒ\GÁÈÀÀRÁ17µ(391OÁ7±$#57±ÈÉQÎOÎL-©TаÉ())°Ò×///×KÌ-ÖË/J·ÓÔQ(Ï,ÉPJ-N-*KMQpËÏ+QðKÌMU9RD8çç”–¤)øæ§¤å1000Ú201ÈÄâ?ñGßïÝ›~Ìß´}>ã÷¥™¿ÿþÉ/º²½;WÎÓ[µ;½›#…}i÷ÙîãË9þô²guשÈ%¤°ê¿Ù}ovjßË£¼)]¥{Æy¾_Öß÷wÛÏxëçæŸÇ¾?¹¨{éÒ‚îù¿l5Ý……‹ºgÊÿôý]%:kP¦¤»Vþo[m PfY÷,y¾²?œg}ÏŸ:yÛïÄi웸îqËq±˜Ïçá\=‰‡çÞ^—3endstream endobj 250 0 obj << /Filter /FlateDecode /Length 162 >> stream xœ]O1ƒ0 Üó ÿ v©„XèÂЪjûà8(NÂÐß—èÐá,ïN>Ë®¿öìÈGôø¢Ö±‰4û%"Á@£cQ+0ÓÎÊÄI!»›ïO X d7~×ɧºœËªÞBè ÍA#EÍ#‰¦ªÚÆÚV›?i vwžVg†ªNªø%Gs‰ã&à#q*MK“\À1ýž >ä¬_.?S*endstream endobj 251 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 306 >> stream xœcd`ab`dddsö Ž4±ôH3þaú!ËÜÝýýúì² üN²Fòvµ|Ý<ÌÝ<, ¿¯úž-ø=ƒÿ{ª #c^qS»s~AeQfzF‰‚F²¦‚¡¥¥¹Ž‚‘¥‚cnjQfrbž‚obIFjnb “£œŸœ™ZR© a“QRR`¥¯_^^®—˜[¬—_”n§©£PžY’¡”ZœZT–š¢à–ŸW¢à—˜›ªv¦˜tÎÏ-(-I-RðÍOI-Ê+(ÊÌMe```4``ìb`bdd±ùÑÁ÷ƒqÓ‚ÅóO|ŸÅüýÜ,ÑI³z§uOäX•¿4+¹>£¶]î·Â‡ÎÖŽæîvɪÍÍ=Sú§Lî—ã+^üÓ~ ÛoÅéìû¹ösï_ÀÃļ ÿ kendstream endobj 252 0 obj << /Filter /FlateDecode /Length 7274 >> stream xœ½]K“Çq¾/uR„#>íqÖæ´º]2t(J¶I†E GˆVÄ»XÈv ,ßÿu磺+³ºzv–:hØè®ÊÊÊÇ—ªýûå8˜ËÿWþÿÙë‹_ýÉÅËÛû‹ñòöâï†þõ²üß³×—¿}oLž yÌæòÉó þÔ\Z›ÌeœâÝtùäõÅ_v¿½ÚƒÍ9¥°»¹‡)O6‡Ý×ðÓŒc“üù7|9æ”§Ýþ[¡%ú$÷g÷ÕÕ,™cêm–‰ÃM3&ÉÔöÅFØJCÜ3¾%í'n¾àA»_ãfån¤Èx6ÎOÒÌã|€QîéÇjü•6ŽÓàìH¸5ÅœßÙämKx„Už!áÖ­%Õã4Ú(fƺŸðÞ'°Ýªã¤ß“-¹Õ ü“~C¸l2JºühlšÝU, 5}?j<È5p ËÆ÷ؼõf|„}šgpïLÒ['äì¾Hóø¾L¶öJ¬Jö^ ”<ÍZ1¾‘)å—`t`lJ»çb˜ãÛòtY~yx†BLðk4ÔèÓ¦Ì/úD€Ãç®oýa¦*9¿Æ·ÓF™drFe@<–41²4=„“N./}|µ7;î>Ÿ~NüS<åŸ×~Góç¦Ø£Ö…!þÏ“0Š.&©­Nº)9å›n³9”và›@}a™ ùü­îÎí d·]”]nJGŽ$_\-Rõ9>MšÔ¾¿Úm@%½Âëµ6¦œÀî¢Må齌 >ÿb H5à X¾ø èaj†la‹>GcδÅÄÍ]ªoD4"YÅûh=È|w§õ>.ß5ûÞŽýaòñ1ÛNoã8…àÏ/Í6Jž’º§«àþr¨„p1x %äýUuÇ, ÁJ îê»ÇÖîGxÅ“¤1ËñŽ·‚ñ_—øsÔ"z¤ÀUéºî݆¥!ö%{t»œ¹õ(™—þ;Ú§Å£ìdN0û*–®JËf×;k\ ê=1r‡68ýö L#Ú|6ÖÛl$Ïü˜?‰žŠ"1jç4¾Zx}3k]Þ}*dö ~ n¹>Û#Q{R~J˜ Ý^—±W¬fèF±ŒöŒBXYc`ò1ŠSù©bmZ!Ì}˜Wn¦-’næ7ŒÜfñ2 Â^9Ç"3i¸ù»¸Ê–¦>K B)”'çÁvÅi¶]®³|J? æ5Š…$Lû1‹cˆ†4ñÓ'_^pFȃÌAñv; Ô@§’ò“Š.ÃbägH AäMRˆå)1‰ìzŽYpsk¾çíù¤ž-ÚFØÞrBhºƒdh¯ ^wYåcW oH€Ä‚‡ÿG$†KCDé"ÀË„\ðw\lmG7ðoT¾AX % $OàŠ\–±Ç/(Zÿ9é¤Å©HæÞ!>"ÈÙY„õ‘ø´Jpª+öÆô· [?¹B¾e=Þ4j\„û{^|Û >ž²°°¬åˆ®xí¥a¼;ÀFrÌŽÝísB9²Wbõ]±KfpÓJŽ–MÀ4ó kv;M#‚´ŒÏÒ{ú5­ ‚Žeà>ÊrÁŒ?0yÃ2ÆÈ—üµQ¢†{À¥ˆŸ-áÞäMÙØÑ4(Çp ù š^W×ïø’îe6‰{Ê$UPDÌw•†ª7/®z‘4ÂL˜iȬïål€FçgdâˆË™èE žÑä ±QÊ;xƒ%Œ,€×ƒËgo€Y)ôƒ )®Ëód{iœ9?ð°?Š–XÔL’µoªe—€â`}}œ“l2nºç…áænËœ°ðœÊ^Ä=ï©0, ¢B‡r»ñÿÞ‚á&p+¤öB²ß0t`…Ò’•:®²Î#€ÛI‹Î”øsTWÁ¸¿©kYŸãS™zZ5Y¬ìUý)Nßlg¼÷ó%Â[òÀ~2ŒážmTµÌt{§¶›>ÅÔ¾p¼¨cy™˜yfqž¬¶â3°±Íéé¬1ŽlRCdpÝ`a¤¼ü¿CS)߸)+Ia‹ì)­m\ٮϠ?\ÞÖX „…ÒÑ0(¹äÞ)}eêf{Eläwv EZã^* 7:ˆ¬ñ£ÑÎ…fŒvó ¬ÆùX'@ÈqÉ5ŸÇB©sd› SžÅDž›€žðÒkÃ?ÃÒI•d™/TjõçŠXšèÐaÉ×Ä`¤q’ð."Ç3"Û»j•¥¡º©o«ìƒ3tíjBOQóµ ´aÌÁ:«T)|nx JÜ5¯“ ê+OLϸhñÏÅãt^¨8âpEY¤‰Òá´_T7ì€þ¾ ½©ƒjñ'¦'¯™N¢ët5G V¾‘Èâ@c¦ ™¼$ÿRDîñN¿¼L7‰(Xu¸WÊŽ»;R¡úŽ•<J.«š¯òëû™Ú©âg ù‰foº!ß²ê6ºýyb¾® øž¢>Q XÆcGÇ´Õýó¨¨O¤~!ä˜eìžiÊohÇ-6Þ?N$žië£ó›¶~³÷´”íR}ßt£Âœ¹ª¾mÿ´­Æ6\ì¦Ï¬[³±˜ ‘l½¸ÝöH:ò0‹íËöe=oÆÉ+¢- ^›&»ÈCÍõ"fØ'å’¡M¸Ó¸Ð÷ê4$ Ãây>B{è€×[ñp€µäZ"xD<ÜõT¯jDØ[*ì¬EȨ¾ÿ§’v¨xdK3†Ùægó·Í*,hEÔ‹H§å ØF!`Ø@¹0Ó‡çDÿSz´?«ÂÛ[µó¬É¼·=…Ü›ìÑz5îq¬Ù¶iôÊÂÖ~Oˆ“®Ô,Ò˜¾ÛÑÚG¡ 'ò{‹4NˆÕã0bs<}ôÇ«¥»¬€Àµ¹”œ2†µ²H‚^Ê3¾–U›#“l\^Ù<ʃmø¨Z ã| üïÆ‡ m@l¹«ÉBm g«A€ÆnÂ`Î<7„ÊPuã…ØÂÒ€“NW‡qì0®ªwž+”gÔQ¹Z2æÉ‡Ux°^ر08±\QÚ=êV.{¢:J1øRS1Ѽ¹fÎYÊy …l™x{î"²nQ-A"k +l½f»U©Ð®Z4·én2€!oº}J7zj2…îiڒ曎箻²ÿ]œåòɘÏkðétÜ‹êÑaßì/6F ÕÚ(ƒ¶}&«U–Îr `Ùæ€p´£(!jÿÅ <³–¹]‚9íWA†£>ë³T·$Oº5¨ë¥û¹˜ˆ µè·†ý¶ôï€i ŽÙèÓ©çÂZûšÞlÞoËã @^÷E´gDh7ÓþpˆÁ4)úçÎ9ÒM<µ –2&-F¦»,ùZÖ6@Þ,½ðO÷)ÚÉMKŸ¢S©®{¤˜®”‡‚.$½ Sýv@m·xmÙm‰jQl86îO™^âÐÊôÅ£>z*¤n̸T¼Ô©Nfß©FµòÂÜg˜ôJE<ØœÖÃò³C£ ìÊ]5q:í[˜K§bÙèƒUšl5ú Bž>È€Oñ `çHI井ý}Q‘ªò!³y> ±ŸIÓ6ýiÓˆYŒ÷3ñûÀK ¼Ö â†[öðDÚ SNo¤Ù$®’˜»ˆ#NR~àAÐ*4-ÿx*¦f×^œq^j}Ay]ZÇÍì–¨²ˆÐ|i•´­ˆÃÓQ÷…É9ìœèL²x{Sû;Ae§`6úo)=ŽüP†˜C£Ç^é¸x*ƒ6ƒ^®ÎDËRÕ¬GÒŽy=¿„7óN©†´…'‡úS84ÕŠ$2³¯æ]Ïhu–y…§f¡.h¡ÙuÕ߇‹šcþK£#UXzcZž<])!«ÉHø~’Kìå-?“•ò.£9°hgål¬—Pøƒ°KJP§¸ˆâ¤)þÞñs0åÍ,87˜º­#i•¾\‡SM“h»¸æ¡›ôú‘ŸbÃVŸŒ ˆ†ºÍ”_Y¸YA– SÉNç¥@(ZQÊB„ÎK˃D3o7ìqeóî(¡ç¨)—¨#½îùó: Õe5mmDÂr·½ªëLV%#©bM¸–2ÇækÍv\‡Hö³nw‚Ùè†Ð4¤îø9Þ!³³¨×.œçÔ ®ECãyFxñqlòègãøy]ë¶J?àyÓCýyž±ÃÅâ9/iìú·I•B©ñMù-Löù`#®Åã3a÷ïåP˜³ê´ÇòUç!Òý*=x¾!‰@ ëv¶P ³}[Yh£sT啎Eâ²—Ù¢/¢k4é›’‹eüréCõÄ|È9u´˜jôŒáäë¯yŒ [RôH…4 µ|Gö’|ÎÇæ¾k÷wu-›× ÝóØÞô¡¼)qX‡fóZÿ¹Ùļÿ[7fšWº…,èê﵃èëå­YpЇM«·»ˆÅ¼ìÞ†$NfŠw_td)E©Ô"¢zCÇKšZ(ö°Žö)9¨ŒæDœg.]òLºãâÜfOu›@a1Rõâ|¤è[pŒœ6MþãT¹ÓÐ?ˆŒL™{ždîE$:Ô–\ÑQй´Æ&µEÒv\:O,jŒ2Ö9ðZ¢Éî”t÷ºÖï OSV FÕGº`· Ä‚¼- ?ÿ»Ê4öÍÄÊ®—y5hÙLÍ…<¸Ä×ËÖƒ·çÎÐYІŠÚpI¢Sž¢ãæ¾PT1_ ¼h;J?\š›â \*Ûu?K¢£RR nø1ÖÎmŠ)ƒ¨%¼G´À" š&C§¡~sWÎq…ˆ×™¢·…ÜleX¡N÷•™ê¹;–ê§Å¬{Œ^;f}ó\6XŒÔ@AScKâój|¹W%SVát~Dß„SL>/K# y”»„ïX`79Nú˜“9¼'L7ÃÂçë¨$ê–ŠÎ:éT0Ø¿%®%¯||ÍSÛ”;Ñ…-7üB±•wÌ&øÒD@÷¼{WºÓBœÑ!ñQ@&!?×ü6÷ßTÔµPº‚ ¦ØÊ¥iEõžÝñ Œî ¤"VÿZ£eÛTjã’ÂoYS"nØbôûc4—2á„xcÒ椑¤àVÓìÂä«nÖ~ªÜm¤½§Q<‚H±Øsþ§Z%%L6erÑ” §£G}‰oð¥kJÛ]¢ÚØ>žÏ„-µyS±ó#ëã€MKä ;w—ªª.Õ¤WåH&fßD[œ(:È[ª²Ðµø¥WœT©¢¶¤¯¬n^U°¥%wã6À­Vz*nÄœX1§Pct“‡Ñ/m¹Gþw4”ßᬫxs·XIÔJ%”50ÝÜ\¦ÒfºWŽã#¨‡ïÀƒsãê*•úT +íþ3!µ5ÙÖ×ùÊ¡Uökr¦éú_ΤˆêJWÜ·³¯ðö¼û do©hâ Z5ó©6º:©$|1˜Í|…‘,!‹ÊÁç¶ä±ä]‘xº{âŽ_ÆËcðKB8'ŸÜñ´s³íºHÝtkWçÅgd®ê %ªn."¹ç™©Eã91lp(sœ ®ß1 f®ÛF`3ê:QfSæE5qrCTYTJI_ˉÄÜçZÞªµT/¤ÛÔQQõ¦m•[”Ç3·Ž ×*ökã:  oyˆq ÊÔí{8¿”ó˜z™Ð)`oÑ•M…7¥_¦@%:Þ»òÛ«@½‚Ktu#иû¯«ó$ã© ¾AÒ»õs¬J{.öùzÛmˆ²>MK£5Q³(Þm —4 Å4ûÆm^ÕÏ›dæU±b7Œ=ƒ¦O™‹dË‹dØ,+“6dÀxÁô*×R˜$£„‡ˆ3”«?ëž y*í9ÅÅX"Ãû²WOS¯.R𢂴ß«Hßû&eƒ[í©r'½BÐ㨠‘'ñø±Í#ÕÜá xKSŽ©ÑËú˜Q‘1Ú¼+µ'NTÆÉ[_tµUÆñ¢ §”Š:˲²©º Œo¤S‡êgšN†ÈíeÏúo4à‡Æô <ò†bYn ÇlÃýàjf·ñÄ4HWU§±UV¡Aœ1«ò”¾¡±Ð¡£4|ôĶ[r ”U¹¿b=¦,éB‘u 0Í)„zÃÞcŽÃ.9dÇ«9Ž*ý4CEìiÌ”0ѶnÈf¥U¸ÖûvæçÂÕv4ókCâ®™Œ0¶µe$#k2¶½éÕP(¢€Ò©ŽXÞÈöJ¤ï§QPäV¶uãdm^9XŸðøø,Óç´‹"Eè)ÛËé 7!´ö¢‹rd|¶×m¡…²³ÚB¹"œ&¿¤Ã›~ Yš¬97’µG—Ö68ÁÑ;t.6µE·™ ˆnø;ô‡20†Uý8Ô¥‚—ºêëߎRÄužÛWŒà S>á˜&Þ/Ï=ÚLü8Õ4W&9«Ø9,þº 1úäÈ?«r›o0)®žoéGË"²ÐA†´…Á¡Ñ[÷ôv.žþ‰ÊÀ´oÞÿùÔÌe—¾Ðn"÷äŸØ¡ÑK‘›nÙosº‰Peù³#uFSB’ok>u¶4 µ `iÛ£”¿·‹³qÛ­¢!Qâ¬î× ½S÷ëìGéLv ®ìCa#zÿüôý’úz i3¸@HÕJD§Æmë;;nwaŒ„ìG~8ŽÓ#bP$WÆ ¥¥9ä^¼¡‘&4 ‰Ëä’£Gq”òÌK’¡¦ì Œ>wõO½K4ºë“õ4f#ëØ'ËiÎ]îuʘj =¬Ì%[Á0‘ùÞeš™>ÍòNœ÷tŸµÌgÀ¸c7uþžºz¢w,â®ÂPuq¯ëU&;:ñ&ïî–±òÆYuiö2´ou?,ñÌȣʲ×|¹Qºuâ턉g·wæÞåmŸÂ`cÃâ«@¼¯«ì¯_Ó|¶˜`u6;b8Ûÿ1<¢9_÷»qi)ØM¬pý7 ™¼‹xût¹6Tø¼ÃKú=ÿÁ*L'`ú®¼ 2øà/sÂæº,n'/ᦵ°åæúãÊèmÂêƒÍ9uæIFÌ¥®²©öBÀ‹Í²ºüCú²Xd#áhÛX¬¥ôPÓàJ¤äI-Zxº(µ¤\µjÍc7<…ð’î/]Ÿ‘¯7~]ö05C… 2E¾`™$¹\‡0mâ4 ñ­¡ TŠï7Õäè Af6KƪþÝ<‰ß—ÿõÆ»endstream endobj 253 0 obj << /Filter /FlateDecode /Length 8348 >> stream xœÝ=ÛŽ]·u@U£ý…y$izIœ I…¢@Ò‡#4v-Í89±òõ]nÉ33¶ …|´‡›\$×ý¶ÿx±ÍÅ‚ÿÕÿ¿|ûìçÒÅÍý³åâæÙŸúëEýßË·?}#<>9–¥˜‹ç¯Ÿñ«æ"›‹Ò±¸pñüí³C¼|þ?06x56Û£õÆ?¿~öûÿ\.G»oâá þ4‹‹ùðêòj9º’]ˆ‡ïà±)Ť|øŽ(%çxø G¤’K€WÎYøG:üç%̯9˜a9†RR²‡w4tY|>ÜÓÓ%[ùþÝ-OJ>ØËÿ~þ‹Ýþ®V ¯\8æôÿ€É¼Iû9ÎV—ûÌ–Ì1ær¸{‡¿qñpø±ËbÌÝð†ûÃéÍÔ ¼ G`¬|µÌ{ø¹”`ö/Å,oyQ[ÌáÏø¸¿hãéE³,iÉø"9®iц"ù[‚•Z^´À„^=¾½´ fM¾ç™ËbêØ§[\c±ÙYXnù®Â ƒäÚ·¢v }òÖʱ§7„±xùô/´W¶yñNŽÁ;‡C_õ)o\ëzJúìÚÇ¢qpÖ‡Ûmð½XžW_BŒp»ï{ƒ7ð±ŒÍn¥¾Ùì¬ÑŽàU…m±Œ4qñÖç Ç2¦`6€pÙ«uÝ+ãŽÁËË¿¸¼²¶“ ¿à—eqžwS¿g:ÉÅ2!å`-_ì²¾ËÒ¡a»ØsÉ4üÄT¼Fåf· ³)Èlœ6ùFå/ KR‚£²i‰MdÓNªï£ˆ,¥‚`RrW/y6XP¾¸§­ka&/{#¸ Cñîņúo65L¢bL,&X2"éK“¹ÎT!HÇ‹d t@:&ô„Ží´!(äˆbM?rDZ"J½ÜŦ<Ýn¶ÑDvTÖì£Á*ÿ-~ñ‰¿ä*èÑRHDt† íi“‰(&g‚¨Ðk=êÄè„-µ%/“WJ~9`r`‘(<6JÜrø £¤MJê¬p©ú Ž˜\r§¯m^K‚Ä6HIp‚e4KÆQp&’©ÝáÓ<Ý÷†5(vCvrµ½ØÍõd'¶_ÈÍÙÝçU{s/—¸Ñºƒu‹K‘|êÔ p‰–’¨&g|l×äÌW~€ŸŒY:ž^3¾ÊÍíÆ§{ž.*)Jcw…õÑÃÌp~+²}³ÝÓ†ð§ÉQ’xó *U¿®òáH€~½_­cû¼ÃŽí+ëØ?YçIËoD$S(-è« £žñ:ˆßÚ^ÇÃŽ!(áéêÎÉc,Ô† wd¥. l­YXî-Ïaa™jB¡;•Y¥[ÃYêPíñ£™ã8YѼ]\Š|m¯éQ¬ñ“mÛ/׃îÝ1ÑÏ9Ž „žèé¸ Jgqâé,ü#ݸí€&Î uW›Á]Ã;Ž´q+1(×¾j—²CݦÃGý‚Ï<’ÎÈK1 FxglÙ¹TÉ›%Í\ª³°Üí&Pn 'ŽS`€)Öà ìÅÂ^©@C _0?ÂOô~¡ÁM…W„_-BcØ$#+~у²"¼\2Ò10J[¬wBZç[@üƒö÷í|t{Fø IKðþp n-¼—z“àRÜ|˜.Ò±¤¥úõ0GŒ£ƒÍ¯*>]½ñÉ` u¬zç—=Xj!qõ’h¡ ÊÖd¡dpÆáP½å˜¦áMS1ƒµÔlÁ $ ³Àýöˆßrü¯û…º+õG|æBmê—ܬ?·Ep6~Ð3÷æÜ– Âd4¥OÚ’£ç±S’µK“÷¶Ù}SOX5¨™?È­ª¬kÜõ&ÖÐD}´FL™øe¬¿ )¢ý†!Ï™¼kZ¥˜ã;Ög@{*£[Ëæ`së üZʇØ'Q¬E˜™a )1ƒNÓe®ƒéF¸N„gˆv«„¸ÓÅÖïc,ªŽ}¸¢ÅÚÈX¤!³±¡. Õ"˜ÿÎó Ã$† GÁðùÊ;¿ƒ¡É@[H QAâlÄÇ Òɸ9د˝¤¾ÎW-÷4h“éг $‘¯ÂýNp„„fÞS|k{|+Nûf{u„P ÅRêOÙ'¢: Zű/̼)ö0ÚÿUvK¸¸² [`ÞÒS Á ¬i-:mäåío¬Ž?)Ä$D£j‡j]¬×c@Î^((ºø4¼c»h6(•×8µâüñˆf`•2p4ð—u`èäk´É]ˆp ·;z¡\6Îý ^E^þÝîZ‚†Üù”.@8Ûç7–þÕE¢.dLZ“ZLt½øu˜3‰{˜#ƒwaAI!÷taßN÷ÿº°¿2h³´)ãq±)=‘Ü~,¦¸ã. = sÅìOX¡ Ä_lÊp=pE:Sš6\úä£ÓŒ)?Í ¡Ø‘PןL(¦<™PrLßP÷W!ÔQã÷¥aTü5ıCìµÚ˜y½l™ Ã+0eøœIÑjcâq")ÿhôL$úÂä¯G˜ûkØRM`&ÿ?³:Mské¬ÎÏ;U±ËlÇp=°íNUü3˘O²™nl_Á+Nú'GTé{Ö= Y[Õda‚:¬åܳrVÇý¾s§Cá;ä €¢”SÐû‡È¡+½Ql̃LDøb,…Àù"fxÝós0€Ås¯½£Â’Þ¬ê›qcx.Ã%4þñàö›+èÓþ¢Õ%”#ØèöîDùζ5¶ñ™,$0÷"LbïÓ˜ñ•ðéµÇbŽ„GŒp*tàŠü«ïø³•‘g¾õt>ý–Ðb àydž:˯,¤-®sž|M›+y €à8•wó™2ò¨ Ç0‚í #y”¾›&é»à©CQVŸ¶<tOQ:šÊ&Ô§îÃBЉ·ú1ë΃8|v#Ãß6Ï×7P(œ˜Ûj6^öBXKî¿ÜæP²mÅáÔs©ÛÝý¬¼c„^¸»ð0j˜GZ¦†ÄåW|ØØíB#–ä0§ë(“DHQ0Yp$”q´1g´ˆÆSÎØiL ËPÖÑsyÿBmîð?ý…uËøcV°E÷ÖNH`I­£ww4jì²åÉcnT^(¹àLd'Há ïR/ Èh?òà“~¹¸¡Ë£ÓÆ3¸„cÞDîÃ8 üÌ œ=‹¨ä›Ï}VS8všBˆEwËrÐÐÀ ŒfuNÒ¸;W„>ýsš#™‡ý®AÌ-gn"qk† Â׬.¦KÀIS oyÆ"IâÏ&1t¤‘/ŸÍØÙxJ(âë8ÁÏ Q‰ñP1u9B|ˆ ̇P–2$•N'ÒJte ²šâVF®(æ’–yùLPÆ55 oŠ# å]tHœ0 wÅá]6óð>t›—~IQRè·c¥\„ËÁŠnwWИ€Ûœ:zCÚ™ó–ÙQô`UVk—‚·Uã:4©êTûgD¸\YGõ寉f×Ï1É%ëÜïuÞ}Hr_‚p¿^òÝkMù”¦}è´ú©PuÁSV—Êf‚*Ù„C_cÖ§­i5H{Jíîä åjèä \“iÔÒ•~Cúúß †°¯V]º ^™,ÓâÎeHï1(þ¨¸©Æ1“ÌÒ; +ŽÏ–ËÈÓ2äµ :‚+ɺt§s*}DßJ%€aäÑÀZ:i°9À ´¯&ؼ›ƒñ¼otYP2ë±ññ¨òÍæK±åº:P#©¬ó7„: í0_½)rœ=h…ªr›y—™.k1hˆNfîÄ Ée*™zܸ–EÐñÍ—¢ŽD$ßâôèB@h‚»:&GFWŸ~ˆÏÓ2‡R j‹aÒ”ÎT÷íkW̪ü²²þÈÀZ‘#’-^q ¦™Ÿ«=òˆü‡“TÖÒ˜yWO¸D·(á¤ôN”.|Åë y´×$°„,8½hŒML'ürª{Æ& øÈ0¼¦Sà×rš§å®#¼!û¾cL¤È¦ÍÓ1ì×ëĹç:\§ÖU•Ö3{YKû2Xsµ =U£ Ö!Ÿn4Ñ@£R/Ð*µFÖ4׃¨½=n òÕ%»˜NËiþò¢V&‹n…ZÆÇ;Óê,åŒS—ÄLˆ}* ¡Ø8rË’šû€[ÖÃQ û×—W `ŽÁ؇ì. €»Q©%—5õñW]ÍÓ¤®óDÍYc‹«XŽ@Z‡ .oÆéL¬¨ºßªgXÐ’×Ló ÊúÀEyg¸Aˆ”„5ÔCX¥–Š›|xÉ7ãbêÓ(Û®þ¦òËZgÉ](ìõê‰÷²èdÌu’åŒ%€\Öšp/µxL„T±Säœkó`ÇpOtsöÌ¢Õ.›ŽøÅDáað“f7ÛééTèµðqsâÎ2w%sW{úž×ðªÒe|jöÑrJ’I/ú9` ¹høzLqeZàõßãdv1á¯7„×ëò©gâ4÷T¦(WmV&îlð{ˆ2Š76dûpÙú«ÌCG¸F±ŠåoØ/¹ 6‰›$Øoõl¥®çŠö_¬PÓ¢_<_y€˜öÈ$@& PZ·¨!èÉeÂãçØSƒSt—©Åm)LDÃ4°ºbÑ2)“Ñd­sÐ+¬Ú …‹•°ð@«&FT=VbUò’PN,üAͧ@îÊ!ø”¾w}>åâÙœ‘u•Ö¤®]ÛfEUÖÃP% ù‘šU%¨äžF‚Õu“|â\G¼z'q’w<8I<_ï&êDɶ‚.q¹“­W–âü”Í1Ùð.µ"5wSÖZ ÉoN:¼9£¸£ñ;Æ‚QÌ¿±fP0JŽÊqᣠi…¡4ïU:_/ÙñN|£á;Æ5MK5ûæÕjÜåK·‹8‰–A“:]q…:`×u-üéN¥% ºzÁ‚$OQ¯`0iqÓœ ‹èn0ã¸ò°o0#|<ȽKËœ¡ù¾áU¢–ÞkV0*VGÿp™ÓÍ+†Î›±¸ø>ÅÙ4aÖ6–HMíDÅùÞó°8ŸâSM­$:ÕÐéæ³½Ûh‡îp³Ñö°_ÁmL«Þ'ip2”ËÒNëÅg°÷Gk[ äAS*q9¬ðWq[¸í„tˆ ÙÛ±ØëivX;]–ÁÖÌI••qeqÀáqäÁ|$ÿè«#™yˆÑ©{‘“º/î»ê¢£"ÝεŠ1Ô$‰G¹j7™”\¿ãY0Ì7Öý¨¿@ %Fcc¤Ë)-Ar”‘è½Å)d¶0ýµfůcÔ§R&êS&Ö¨pE(½Éƒ¶Õ· 1*b‚Qйþ²©&¯ê]f¯w5ï0$9ò’ &°QĸÚ5o Óµ-·ëãd?Ý®U¿™o8U] *¯‘®Õ‹­–_cðø 'XØqg‚NPØ| ÿ2…Qt%÷èÖ$MtîtŸ·ï@ªžèAn`^{1ùi˜DV€ŒÖ²oK!ø8ÒÄç™­RƦL‹!×À†mKªö–<®©@ÜûzF$±8,€7¤»N×Õé®wó'LEw0Y·n¬ôvˆå$'î0µv*¤s0›5¤L®‚–u/š,xQà’EÎD#³•ÊI³Ù5Gæ Å¾!V4FƒH`Žbí»èŒ#gÍÝÇqÕC¾ÅPµÎàäa®-áCÊGÎHN½Ñżå90úÞüüù³ß>ã¦ÿ¾€µž.Þ=µÓ?PÁ8ò–<—°Ý͘¬Ø(v"Θ›S*Æw ì ’Ør삚™åÅÖ ìn;µ÷<Š+w|)à©›v)}ŽzÓ¤ÕtKžQ\ë´‹%·&Ó8 Æ`´_§öb×¹·67¹ç×6›§š‘ˆt˜ú|¶GÐÙnƒOjîÉíâéÌ&íHºÒhÈÛ+Ä‘úóžß*ÚízWŸæÜE´áôQH¡Ó&ƒRT?å%oxñLŠßG&I:Xú\ÓÁ„b9‡i VÉ e&_«]8Ú­åH=Þ0OW)3}oÙ­(C9Uë|c¤–ræÛfoäù‡]¶X…ÿQÆ©Ž¾(òs19:j°ûri^/JØ\=5)éX㸑å[žÛ‹ÖµÑ–" Ã_h³½ŠßõºÇ>‰k£-ÞFzÜgsFɨ³Ý¶ÖdÊG¿2ˆ¡gïÝzxõJ´VúhüY8—£rú{|`È ¯ë'¿°\2l)¾«yæœÞí±ë +C­eœX&WæØ¹1•#WŽÙ u'Zãè%&VöÁbÕdïšwKü§nÚ0Vè0ßîHšª…Ô™XôÕ%NÏœXš ú¶ãœ@áJõØ¡—ñ4õâ§zõÝ7Ã^ï*xpÍsD#@Zðö¸UP CÄuW%Dg½'k·–f¯:–[|à܆u×]‚5D+×n¥BϨ«I?æ WQ.Îå5y_|R¥€…¨BÆÃ­ xÄ‚ÂÑ©ªíàºQi~?.J«;Ž×³x”~Ë©#ȉEFLŸ®¶#&±úDè{Òùfò퓹¬/(°·àš^û÷ZÀöŽô|\"蓇Ç_î5®3t$¦sN,§ÃquíÓH1fã/T¥©”¸ïW(kËèõSÁc¶±ËlŒñ܇üšk!¹ÃÈõ© ô2eÖÎDH5ͪ/9‰‡ÒÕˆtŸT¤€´ÓùñtJÔ´wÒàá ¿˜[™zDÃ;ù¸†è€-+Ui«Ô"‡ÅÙ­\\œØàÓÅûoÃî|¦5(bŸÕÙ׈ UÎÆc¢ W[›ŸãJ8]ùÀબ±ìJÏ"f·«œÁ–ø¼yù¥÷÷á¯}¶~%~"¾sø³Æô½mªÐ§}àŽzÙŒõÁhòîѫک·øÖü"–©ÔÖ8B ¦[Eß×8YãF8f¦îë¿îN«Gò«ÞÕŠs³vøõ6î:FöV Mú Á vE=ñ(‘'ГI´D?n¹ ë|?B8îðˆ¥Á ©}𕉉L áãˆu ªîQU‘ËV±Ô¶úéf5?*ä,<¬Û±¡O¡Šùú€Ã¥ Hú× Ðú‰¼à%=É¥•×f(À^G–{CaÁö­ò„u{†+Ÿ }Eg…Œ’h>0`y‰]*ca¬à M /8ü®‘Z£õ·uˆEµ> T çv׈Šàxø£b†mãÑh«·jÈÍIT%Ïôao1;C+¨LÅ!)àë.+“°z\ìȆAèR°Vê„UÉO2ãÂvÏ b€ÿ†÷X¥¡ÕòêHå$‹§~>ü!9%àà&­—ú`F·µÞúºWQwÍ¢Ùºf=5Ãì ˜Xþ$8BýC©+œE3¶Ë Õo&-~J„©À ´4¦×¬‰»A¬Ž¸ak ¬kÙ\ÔìæÊ;—ЇÂ%©Uç-`¬Ä]i†ßW‘Ï~kÌ“E¦ˆqý:OqÍ‚kÞ«ý,+Z¿+æ…Ÿ\Œm“´¢ûÖZ–Õ«Å«9•H(jƒ'x§¾/ó(Ûa6ƒRìœRæþ»u¯RV˸T»Ž¼3S ³Ê…Ž˜Ø…zü¼^éc%¤|쓆©¨.“n7ÆjªŒH¢x˜Þs:D+Ï‘ÙKNýçýFÉxf"Bá¹–%Qt´´%u­+Þ¨3 ,Šœ,i*ö±üÚ‘æšçÃ ì‹Æ=»¯p“üÛgÿ uºxƒendstream endobj 254 0 obj << /Filter /FlateDecode /Length 3810 >> stream xœíËnÇñÎø#:Í&šq¿FÀv"C‰ØdÇX>ňäÊ\R‘rp~=UÝ=SÕ³½¢HIˆ†.5kª««ëݵ?/Ä ÿ+ÿ?¼Øûü™ö‹ÓÍžXœîý¼'Ó_凋¯öÃX¢ˆr±²—?• ¥Â \x뇨íbÿbïy÷Õ²ƒŠ1×/Å`£UÑu×J!¼ ?Æ bì8#à&)aÔdršô\¨ Ü#7FÚïr ²2B{® Yͬãj¦•¤ YÍžƒzŸàBLGÂŒå†À ×xµli³Û•VÛ ‰)й26„£÷Ú[Z/zeÁ ¤ÌŸ=KÛ8cb÷§eo•@dÝ·iG]žÀ# <'ð'/ T#]×ý‘V{Z}L«O üŒgžÎyH2¾ ð„À3ßøCGð[/GÖB÷  PÐgiõ’V ÜxÝD8" µì;ú *ú!HÝñöÁ1k ZÊ¡Æ.ýŒà¬!FîX $«Šÿ‰ì¾t© oô)rÛÇjç‘CÄvhâ—™-ôîê‘u”UOì ÂûëLC Ýý['PòúnÍê÷sj"e\Z¾a•<"ðôô¸j¯  ´°©ß“$|ÍjÕGA #kâ¼}°J½ ×ÀºuS%sêñ†ܰSï¼á7ì@ Xø[çtÅ_ꆻ$œc /Fá„̧ŒŽ:L‰¹uÞCÈl\…î:5] Ê”Fo\&NäU:Šp_kÞBD¬ƒ06¡ã>FŒ=«`àJK¼*L×~‰5æ^ N½É{_S+ê¦ÝãË©¡”a—Ü'jLÍÞ.Y‹µ¢ZM¹í[†ðwßòDn˜ök¶w¦¦Î»²×óW>!ðë62KX(1ÁÉwÞ—²jS.!·sOžò9]ø#‰A5` <Àµ>3tÚ6]öYlÒM\9ª áÑZ•š~jÊ©ñ¨u,:¥}փɞLŒX!µÑ)ø¯|2ôFÙÅì‰ XsHÑKJ£Á÷I¿áS­*ªNIP!Dj»—ÉBRX/´•&?8”¯°Aë è¬éŽVÉ!ÞØÒJ/8iK”ÖÉù*`IÅ{v3H‚ïs~È~2/!À³%Ï øŒä”Gsw•¥­²‘òãë€ó8y? ±ê3oòªÁ2ˆ{µâñÓþÉ-¶5«Áì'Á3äØTÀØg=lþŠñ—ý½ï÷òû\“0~qµûÑ©Îôè$ ]z¡"j>:}™”q¦zÝx™Ø–‘ogêÎŒƒR5#àª~úÇ8°1PݾOþ4®°ay_Þ!]E©yXyåÁ9XL0®Þåàã?àÚìàRût2H¹x$Ü,«Db[ 0¨á‡1¢˜Ë­Ï–ÍC–ŒÔÏKÙ,mÏÈpÞÈu4?KBÂ') 3bë“ ãËÖA›‹Kâs}‘ÝÑ® d£Xg*€_v A×éÛiÎDÔ¦«ß¦0„{æÎs!F<åÂÈb,/Neÿ1'(Ý‘ÂbOÒIYèrºJ^ʼn¬^²‚ì§\d°øT ±À`Î8¾ÙBýÀ¯MoE.¿–ÅšÖ¯³›µXdÃ$Àè1 ˜´lr›S‘gñ+nNi`SÏ­7ÜòA¸|{ mŠ_$S^`d™«tàÉ«›˜z³¬F6/Ë*°ñ¸Ù 8½¡s¶ë¢•ºl„”ƒwiV¶3”Ú4˜Ó{>ë«BîdN/z‘¬Ö©©“cqªÏJ<ŽEVªÊ•ÐIL-gõTÕ¯–˜ÙUùë5íƒE«LÀË]çÈV÷©8†¬­g³ÿ”N{3‘HÜn×g ZÃ5æ4í( úº±Ý\TÐéo/¸ ^j•¶{ïrµ ÊDpJ€˜µ/8e8×%ÛŽQ¹žÂq»Þ§žÅuQž[f'£ÿg™Ð;†‚%Æ.ßtÎ(=áC±/FÉ97øËLöfè¯G`á_ºj64Æ[Ý|ºµÏ1«&@F䲄%ñÙóÈë¥5CšúeMžÀϼ3bžÎ¤- TCÙòÑ=ÔGÜ,Žòw6ŽA3}Çó|q"«gšÖë\Îà‚ï‘ ”TS§­æ4ÄH‰+Ÿ‚C0)¶V5ðð\g¥ì¹Ç†ÙkOðrzÊû[ÝÚbÃ`ÓUØ@u"VñÇ@"pazn–¸lƒŸ5lP¬±ßž.ÉUuíŽ÷Ö]ÃÕeG×íxl]]N…­ò$nH N¶«¢ø]µw]ÂmMñ½i¾líß'9ìç?À_r°0ÉåUo]…Ï äÊ,Á:dM¾uÊߥŰÃ3ÂÛÜ·3/fzvŸgE6¯ÃÆ-®›«üÁïßF‚á²¹#6?Äž9_6é^5yxØœãb#ljìeù†ãæ|ØB`Ãéçü³ d#sl¦½=ÈÎèÞešñ.Rÿ¦)õ‹Ûæ’·†ôçO½\!ØÅ1ÖØ›™a\|ŠûÏMY¶çr^ÈFô˜ìÓêºIÑ=§Ïz?œ| iŒš3ÖÚƒí=.š¸Y¬å()wÑPþ¸Pù—µ÷™ógzÓJeºð¶9QØö [ª07ýönÕ5µŒæÃgRš‚m«ÊA“¦amGÓ9cön?½µï¿¸ßÄs/¡2Ä)Ä;ù‰Oû‹’ù1—û›ÿ¹¿ÿ™O,—*)ÍE¹iå‰&ý"é„VWM²g²ŸÛ½% |@M Œõm»­š?à§(ýeÈHçåÔ{¤P-…ý¤Mï~ƒólÚ—ýh0Ïš¼œ5y¹lwkú}S˜ ›»’'·‘l§‡M„WM hA£›ÛoRc$î¶ç¸÷Ì ªŸ!N ¼ï÷þ)»M endstream endobj 255 0 obj << /Filter /FlateDecode /Length 6253 >> stream xœÕ<ÛŽÇuï´àoX0ck'u¿Q$[° " `Æpo¤¹»Cï’´6þöœKuשêêÙ¥D=|`MouÕ©s¿UÿíDíô‰Âåÿ³›gÿú½'W÷ÏÔÉÕ³¿=ÓôדòßÙÍÉÏa†Ã'»¬²>y~ùŒ_Õ'IŸDwÙú“ç7Ï6:mŸÿ&{×LNqr„žŸ?ûóæ«­Ú¥›kjeCÚ\lOÕÎæd}ØüuÎ:¦Í·8#ç”ÂæΈ9eN­5ð#nþk›Ì^³°‚Úùœc4›;šª”K›{zª’÷çÍ·¸€Ý7fû¿ÏÿcqÀÓ èSëw)ý9®k”2°ñKXÎ&— ÁnCè7g´[ Þn74™a³·ÀHÊxëÎäñ ¬ç“Û.qìwAëÍ[˜¢rö)Ãæ§óa/x†Syó…Øòk«œ\Ü|Éëe7ï·Þãà¸&,‡ì6ï`ä0>7d¸ç× ‰¡ Ö5“¯®hºSÊÃ*÷âU(¥ï÷>z¨•Š*5ðïßò„”|;EEø…ƒß®ªE1D\§œƒ…%¢·nsÎkXù'çØ¶§&î”6'”ârxX”üÿ©OœËÌä’s…,ˆç—uùùË‚”xk''¨ÝËXål0ÓÄ~]ìùH¥d[Ù3¾•Á•ãÙB3çv 'uZ5BÐN'ØNµÝy`èD"­äýŠ–LÆ™æ1s»‹ N•M—$QÃþ‡»"¨Ú Ä4Àx4a—”Žˆ~ü3 Ö"Pa5¢¢½2(9H!Ÿ@¸D|åUœF³ËÁtrÏð ‚÷üšIÖÀcdWÁI3 2®;,L›DÚH gV³Ê'è¦e¬ÀÂ*Jíô«æµº6£Ó›ä‹—‚/'|¢Ð®HÓϰë3uA»Áöõˆçü–Ïz"At( i´qD dnÇú³–2¼-¼nv&Íú~>K–*ø¾ÀYp> Ðè / x¹×RÝ0û•Ö)±b·éDë]öÞbŸ ;µÀÓÞjÒ"7ä`läx{¥-{…¿Âæ;œœ‚2‘- <·ÆÑ»9wöVüå½€xÏÐ)¯=?ÍÀ3¨ÚéM`™X$:˜د0¡ˆ`@!‰Í_îÄ¿'uÖ p¶#¥LK›¼‘½;'—@l,P 6æ7Í”û È+ñø–£È¾xUijkøB}FËeï6//HŽMTEV:ð°¡‚·´œÓ`à%Zꔑ¥6{Ì'Ï¿yöü·«&š-mNŠ8l%)V©ó÷,.`±âfOöD¥hl£jÞò«)§Ž•yÁ$5Ðþ5ê%;]SÚ–k0B d£:·û‡çÏþôŒý5à»xr÷¡NØ©]pþ$xP»ˆ©`øoX†€ý®¶b«^ÞÀH·ü›ÖÈ|Œc8@§‰rŸ Xh….çO=¤]î1óöêM¤-8šìN•§o™ŠY…V[’%²9JÍ»5»YRiž±Ùn²öe=øqxQ©ÂžAòÉ4f‚•6©dáͱ Œ`˜Œô”Ùfjp¥d|¶ÑÙ D¢VŒ;2…€Ÿ ǤwŽÊPú+,aÀù¦AÊ´iç? Ñ<ó‘ô`/$%ÖË€4ÞÎë ©àt!RY'¨0¹„tB)j²É 2UhgœmLÍdü¿j‹U­×AË÷v»¿§…£èvø«ËOòJ8þsTrÁÛYð‹Ùny5p(6/Ʀ[x®gèo€ãÐûÌ3 Ø#ÔÞôj«3 ìÕÌNÁƒ–Š©*¯ÝNx5G\œžȧW÷å§v½+6­Ô½ˆ#“Î"uS/tM«{…'¹îšgIcGä€ñœ2›Rñ0äzãŒfÀ~Âam=ô9èHÙß6Û4;ã»Ú¢ÖÕ£ï'Ø+›S©³Íò@ ñ‘0ë)®gÁ¶¼/’Iàëç™4¡¢°¶ÄÐYå\!?‚GßW*‚T7ÿaˆpŽŠÀ=¢8!¿;‡5.6_ÄYïL« Ïy— ÑM¤@Àe¹u£ ßÕ0\:é3ëŠAoäÇÃ~Žœ©i](ßàÃAV_°­:O‚}9(4´{7sì‹Â܈E=1÷uu{Ë12ЮÄÕÁQÉf‹®)Ó;gu*ô2O 5Jã¬ò¿˜»sN>¾Üù“j†ФðñZsðWŽ-£!b¶¼QââÛÊÎûëÊ­h‡’†—CËðó°OËpp/V»âebNÑ™P1†¢s½&&Æ ®ðÉÒât (‹~ùüºº€/£ª×9eŒóÈÊt ØpŠÇ–¬4»2ãÕ$:p §$­f&L˜œV$ÇŒã;­Ÿ f`|°Ù_’Y€!¬¥oîx¹äˆ&,~WOR’/\ =Û]=ð’§Úñ 1£§‚êá~â›lOé¢ê(©a.j/ùü=c KßI"¶N)ÁÖY¦ín{ãQŽÑýùì!-é–@tZõÝæ@æB|m®À¢®áò=OF‚ËW=+œ€´x‚õ©8‰SÞ¢Ì&…KÖFÞ ¬x ¦ˆûžŽ&PfŽ"em { “AÒk¡Õt€ØXÂs5¡B—T.áM¯ó0ãMw^@¯Q•mb J áaÕÑÃ6ùHP¯°ºÌG–C7Ò3ƒ&²*Åëŵ0ÑÑ_LqÍé•ôß -]áé=ª">M­MWÐ[Ëк`â(µ$óñœïN@­AÝ¢€³Žž-FFÒImÐQ•nXäñÀ€ ƒ³âÕí% ½ÛΉӅ¥ MS†2¸Bh› ážR§ÊD¤¯ÆŽp<ð‹‘¤l8°µçW[`¤¸œµèBLéÃD€ ‹•’+»\úf7K‘G=ˆÃ¬pèy1v—}žäá3ˆûa‹˜rß7*[']©ê÷Wõ(k9ÊÆ“~LAúíÇãˆ.É€g˜‹³ü*W¼Á§QA`1eQm– }Çñ¨Š‚¢,pœXÃE©×í„gE¡Tòú9¦œ¯ê˨3#Š3¼iãk~/òRSZ:5"{@Ø(_¹Æ*\t·³ ð!ðà2ç¸.0G“¶ßO¨ÒQ *Ñå›EÚcà…h‘]@Bk8¨±Sn“+§¤dD>³¸%;–b§iRŸ¢’w¦ÑA¯ªùüQäRÄY×x@VÌ;Gõß…öc±ÑøÎK¦Ù~7eœ L &+Þ–Š@ñ+ˆŠ&·Œé0Ôü–Ó2>ê ¶\ì0a#¥¬Œq4Œ NœŽ 6¦Ã»é𻡞sS¡ÊëAº;"M4[Œ«ŠŸ7sò#š9>ÜKïµX‰6¨Á s D”/©Àûm ÍìŸçÉtp}><ÎN‘NYü6zŸO«S%kæ í‹·ëH¡ÅÖ mcÊȶþÔh›“Î2”¼xÍ/’e’¢ ˆ;O;k×$ìÿÆ™Æ.¶4 ¨wD8òüܲÈ_fÉSt›I¼//j-+Í’×¢Õjä¼É ñÐB„-È*Ý/« pÜ8®ùYÚž ñT¦<÷+POÔÏà8(½®ÛY$4¥˜'‘Ui©E–Hÿ7õ¸2Hí™îê[½å àüÉôä¦Óºc-ÒâSÐ'ÒⓉyYœçh+8?‡Ÿ]1‹:""eÎV²â|¯ÁQê\†É9=\±'å¬n…B8XmxpÅH /`ß1+‚d)pxzìÅ›ygcW{‡RJ­ÆäC÷n@ÛkÂ:cPä’NaU¾ˆ½ø®[#õ5ø¯ fRjþ}«sE)}p•ÄLÛPÐ%ÅoÄýìd˜³ "ßÓf0#ÄN»µ¼e±Ô&6®˜TÀ"r-ìv¬à_v2_ç>ž¾¥…`³<,´fE¸S(Ã:‘‹Z¢g²^³˜—Ò‘ªEÂxÌåB\$#‰OQý8WÇr"ÌbÓ"#E³ÉŠIŸ6ëü%|0´+=+½!±ï¹”Ùi&ÒZrº>½®«­„}û•ÐPºIR5c=6'ä³Í.ü §;§i-³__¼bt(íØ‰-v‹g›©vX ž}~±…dN¢¼¬5c_Ëv¬CÒjØ1uÕça²(\0Lkã«I·Ù“ɯQIê¥ÖÔw>V9@k\D¡µxÝXÎY€Çmûš5¯¾ BñÆm6­#׿=å=ŠZ´^Æ'ËòJÔ½”â؇ÂY8.­>½ÕééÞ?»ÆÂHš '¨M¦OH º0XN[«Û¤Ò~œ€äd-V&­ŒuÆÅŽ?oœàÙ7‹Dðqv¯Dhƺ&SJqè#ÔLÔëm×c9’YÚXÙeoœÃšŸ›=ŠÜÐÇ”xQ-‹hEVøt2å2Š*Œ?U˜èw«ñ4b#xšÀYÈýR‡3"ÌÆ¢^ TsåR‹†Í›ä}',fkœ÷@Á•œ{GÜIõÍRáì ¸s3ƒ|UÒ㮿;~ŒJ#ÝŸ(ɳMY÷ª,3ÊÚ€–:Ó“6Ñu¥œÜXÄ{–/‡9ǸH=27Lu;qÔ_ÍÉD‘ˆlòŠòŒW…÷Ô¢—£þÉ8Á0‚gÑ(k}ªA0¬²#& ¸,ôSè‚{0œŽÁÏQ™ÈáÚìÉ øaSa-ùriÑóþ‘:Û@íœq'.Bàw ° í¿·É¢Š M*é ÓdJS©ð»…¦ƒ3OÐâݸ­&­m¨ú¥ð†Šcª²Tä¯+_,à°²·Xbº@ŸtŸãn§´ù¦P Âæ¾ðsmÊ5uhíºû p¥œðWˆmƒMµ+qyÓ³ýOÓ¹QÕ –¸¯›ÄñƒMƆ#¦2¼¬ÃWuø¶‰A§ñvxwÅÁIËÓFJÂf³Ö<ÑÓtÞ‡ÚøÀQ8î,f÷qÅ.ÅIÕt3÷ô7·X¯S&@êêóÊØm9Á3¹ZIªiñÜ9a?o|Ѩäm‘µëÍš{^+º _%c)áH‚ŽšFTª}_P,ùÇ…Ÿ@r”IlÕ!nˆ¢žP3›ÁoËÐLÿaUñP,`Þeo&Î<´œW޼šŒkI“-Ú÷g­ëNôg¡…%E~/›–ú[éäÓ­›Á§øˆ³>æ‡*Í‘šþx ”‹²f‘r!yzß•®Š¶ýë°È¢[dR=5å&ÈetGl‰ŒOáGã ÍY‰ÀÃXhÎy‰ Ã ÷ï Dxý¥ÍÝáÓ b•· ,º­÷ý ¯^”£æ©°þX‡#Ƴ”Mgn«YK>}ÜórW.ÏàÞY4f0fW“f³Pv„ b›˜t3%q¸÷N9oÖ²gËõ1†¸µ MFmŠÙ…±›€…¥v RY‘Ú+WJBð¯y2–":^„ç /%'7WÂi¶ÍíÝMkÃ&q½Ä1çu!yéïAËZ¦sÆpqüvËà›#Ç>9‚¯%wšÌ=?wÈ®. Î>~ÿâ à•ðå@uè±ùì" FptÛ+Ø|ŠEÔßç0^t¨7®aßÝE˜ëJ;u¡ésùµ87ãƒ>Ã!¶äc—üjå–ŽPPü_¶.U~—w××*ÿkÕœ‰Xz^ýj†in[™ Ú·÷Žºà»'C«zå¼ëü*Ñ%R2õy·Ñ?5‹Fãr™eåÆ`ùUñ[jNŒ…€Ý9ºêÆ=ò.ù(×ûsÓ¯Ãmlˆú®a-ý‰˜Âv‚§uZÝŘ•—\´ÈAÄàš<šL,Ølª¬ ¦3! ûpo´-ÙÿK¤½ŽÉ ö2^&Ø—MXkY#\‡VóëSf¶M¯Î/©!Ÿºü:Aœâ“Z.†ç^íF›ˆ"’|Ó½rL‘Æ.õz§­åJ‰Õ€ÄcõPqHäÄäþ%U Åôš?&?BEx0‘ǫ̀G‰öÓYÝõ”5騴Kc.EGÊeŽo§>^£]f"æ”uKÙö†¸ÝÐ;wÚ›®xN¼p•…G!\eáË`­B¦0é4WȨ˜ÝË“ð\ðY°k¡a†ÙÀPv/k¨öæÓË­°]õÖJÝ£à=}G_G£>±x$ÁW€{R®w¨ÒVJbí+µÌtln'6¶L÷kåz2xs¶YôSîÅ}¾ç •o+økþZ_%ã+¬Ý÷ÌŠŒŒø7º«Ÿ›x]½`» GÚýVCß,‹|zT`€21Ô‹Í"Z °»á6€sªééûdßîÓ æ­”ÆáÉ`Kæoàum©ÇJçFƒù®˜9NŸ7Õ;›ÌººÏ¯_f5…ýfVÁ ;PÕ´ÿØÚ¨Ž»v© ølÂÏüìÚÈ”®|0ŽzÚÛå|$ÊjL`¬ôðsÓŽf5EïM™ˆE›¶ÄfA@ûò•ŸœÉû…Z…¹«÷Ѥc‡“ùÀð»„øg¦K­ õÞ#0Ýw Wˆ"ÚÊȼè+U6q¨||9+Y¿¿à•mùÖÆ‡Ôk?¹±VE\ˆ‰Ð†×Õ3~àIS)v£€ƒfÂ!ƒ­ßyƇ; h·s™ôŠWU:?n§†ua9[–w~ðWsU~q msâ–ÜC•¢Uw¼tðéZ)™.cc7EÂ~*à‡.#ar1‘°·!6eYSO¥Ñ§;ütš®ŠUä|šÉšNiÜŸ„5Z×”˜ÞlG÷–eþe\«\Ô=Ý–ï/fÒcé\óØJÔŸXæšíˆ& ǯ·ó—¸VŠ­OéL%©öªþ?ÆÚÛ \|NN^f"6€ µ€&ƒ>°|$ùcé£^~óÞ‚¾PðmÁ<½*Më1N_*•\Þ1â7„¶»)GQºãÆÌ k¬ºÉ|´×xº&Öù[(%Ú“e³ØRÞÛ]ŸµÈšžfjxòfõ«AS]õ“êÅŠVÚ´E¥wõbwc>ç{ݽù,ŸÔ*×Õ?øëÈåCÜ«åæÓámå¡›‡]An俉Zʸ("üP6Ž«1bÂbEÌÄ&%cÅÆQî²ø§&®Â/¼Íùg|ë£ÖCÍc9þ_ª:½WªqÿlSQˆxôœ‚Û·ðþCæF‹‰R˜¯e³ß•0êxo£-šº¯Ã;ùÖ¨†'*l·u(;‰·ÅÙ(ÛµÎÆO¬?=ʛⵟUý]šˆW¶F4¹’ò¦Nø]®›#Æùç«/J¡pr0Š~§Lé>éIõ®­úßUº^‡¢àðA‹uÜðD½~9Ž«‘t;`I×›á~çCR\Iüí­0ÍÂ<=©¡CPNŠÈuøu~9žü„Rs}o[ÇŸVBÿ¼ýz2™1™ ^¨2ô§gÿNljìendstream endobj 256 0 obj << /Filter /FlateDecode /Length 5829 >> stream xœµ\Ks\·•Þ3þ]^5«Ìk¼Ù%Nœš)©f¢Qª\•Ì¢ER´‘-‹’m­æ¯Ïyè{[MŠ.-„ºDãyßùp€Ÿ7jÒ…ÿÊÿ—·gß¾°qss¦67g?Ÿiúë¦üwy»ùóK¨á|™²ÊzóòõÿToŒI“ z}œ²õ›—·gÿÜþùüBM&ç”Âöú\M>{“ÃöµRQ%Y|ƒ•cNÙoïà«MQ‡°Ý_Xk¦œüöoðÕ¤à|Ú¾§ºJ¹ô¿/ÿớɪd6ðÍ%h,m^^mÿïüåOgÔ¿Ó†)þöÏíGhËé`|†Î.pd9F³ý†{‹Ém¿?Ov2ÊDè F£`,b„8› ‚ƒa]‰ß¿§!:c .”´²!m/ÏMÀ©§íܼÏzûìfs²>Pe•1a{Ë”‚u®ÿ]íçûí §ä§'˜ ìÎÛXœø__žýýìçÍ!M)Ó&åõ”"üÒBkzãML“3¼ï:ûÍ_öø›•mﻬۮaÛ=,½Wn²Vó¶sWÊä(—áæã°P°gÉ&XJX »¡·ZýñüÂ9 ·ÿxŽutÄò³*V¥­Û[êȧl¶»{Ú‘…5¤mJ°9ó>‹ÝÛݼo²´»åêZù×¶É£Ø ¹W¢Œ{AëÌZärR.nÞ¯«ÎòZ…»Æ5¬â·G1ù7Û_±k³Ž‰§® È`Ѭ²¬4•¨Ü6À‹>xT ÈjØØ—}P0¬×U˜;-¢ÊKå”òÙ± ²'›'ZÎb ¿o{u=ë÷ö7Ò ã‚ÛÞÒÒ+gCÑ/ÞúßðsRZ§´¤ Sœ{y×:Ù±‚%ã¶ÿ>÷šÓDCnö46mÓF¸`°é Aü5Xë¡Ï]8švNÁhlÀåÒ:£±Ã~à!Ð$ä||¢ù:ã§`ûnÐz‰r†QÎcgá-X€<Ìâœ÷2ë»m" hÞ~*Bú*Ì$a­ÁLšê‹@ý~!I`¤0Hi¦MJ Q©Øjí†,ÏX>Õ…†Ñ9aKÃF–±+0èFÀUÜ5šïûæ^Š& vv†kÒž\6ÕbÝ☌MGêº Ñý{þ¨\6Ó²¿»a+·—úΫ­óÒæìîp9AÄpP&eؼÎÒИ2` Ó¡—W´Ù» -Ð<ê°jjMƒ°)÷Ü„…Ê»;^ôdÉbâg\±Åõ0Û¯ZÓkà~É`€4Pà ¶ãrá xG¦½Es¥x¢@YðXNçê垣æÈ9!d¡‰€Àˆ`âÙs*ƒ/ïVàWv{BÄ]r®ˆOrè çeù¥áÞ3ï4ªSýü”¡ ž•!•¥ì€óš,¯‹0n;HV𠦣*@4¶È s0”0?°ÊÝv`'øNî;#ïlì}jóºÂyÜóí€RñG"Ö@ÇuÙPl§:0X°OAÞ ™”|ÇLyžd |*7á“—Pùrð'E„$_ò¤00‘¨±Åž‰é0H¯oFö:™§×6hœ6Œ°bòB¡—é3ä2!×Õr½>ITüŽ•M¾ÑUØF€Éròg‡‹©‚ô‚Ò (ˆ²ôŒ¦›ÃáÉhîTdÉä°Ð—Aö † ʺŠêb‘@à6­¡ºûsáxfSÉ£5ô £.(ÂŽ8Xç3 \÷ ™Xà {[­)¬åÕ2è}Ë?LúDàÆcêmöÛ†`:h -’XúÜK¦x`AÊÞ÷r?‚B¶’Rª?qÛ1uj»/=‚…YÁ¬_‰ÏÄàdŒÀÅøw­‰ 1“9T—Œí°y¯8Øúa0Þ²Ø;Š©R¿ëgZÖÙH»ÐäTbœ-Ĺ<n²îÝ$2êLÙ f°ÞdñsHñDÅàiôÎàRèÅC0ªÊ„Ù´åî@Êc1Ž:q¤à=uþâm¨”î~ÇË Þ‘8r/¹=›—cR‘•àfnOP5âÀÏ ÙU¬\ .Aȵõ'Œ×à§ðIŸ¸á¤Â‘àØ¨ÓÌ´Çkщ/¢pê‡BCò<ØÍ+Þ÷»ÿí ßq$K&g\ƒ3A[•xKñºkÅÛV|³XÈ®Z>_dÕ(@0³Nõ[Ôs’ÃÀ5ÆÈó/¿§®’6qû¬hÅ×­ø¦?´b-——G ð&Z5/Ö@Þ\€-‡ÙЬ8ÍãJº+Q­5º ðˆ”Q`¡WÏ/ –Š]!?öµ•¸›úË2³$g–#ëÄþç8&å½Jµ¶ÃÎõ„ŠpÙù)鑚9¨Kò¯¨RDXŸ“úÎRÎffE6ÐÒ¸# Y”·^âÕ"òÎ!øfîŽ*ƒ_Û2³Fp4òæ` üY!9ˆ3Ê©Ûcà{„Ü~b¶ØÁ¨“ÙØF*e ç·ZÍŒÿ#ø›¦h²lTÇ[dbeÿhÂ=û7°HIÖÝZÞ€]!òàÀ[¨´\Õýè}¤Áö ±Ø÷hJÌûrñ±4APJF3³a-ƹbåñoDs£ŸÉZq¯CçªÄ’ÒZ©[êdÖ2±¦ËÇŒ}-(³lJè›__µ„n~h¤V †’¶]¼õ`ÒÓ@-•Ñ.: 6«qô_7¿ô}+>kÅZñjÑËÝ·â‡VüzÍÇYá©ä!K ½VóH8zRuO ã B½Þ%f(ô?:¼iÑ[]gþnŒA‚“¥eÖÚOÆä~™É¡~Ö=?oÅåUï ‘¾*¸µœ„ìæƒm†rÛŸ,˜ ,é’x¹¼”HˆÝzä?*3ž?n s2ø3¯Öâ3Ù‰ˆ%…¶¼zsÀZ]—±é#'X"€^0¸[Ä‘“8‰ÕpÖM®¥Kfà<“¿;$6Èv÷ÄFÞã™ÂY¿áøI[ ðP•aDã÷Ž@ šUFÿzÄKrk ¸ÞWÍù½lÅ}+Þ¶â»Å w­x½øUÓe}HnŠÍ¢ÍŠõ‚"”Ó‹åXstNE<âÃc:„z§#`Ô0+Ûèñ2’j!ÕS°dCD­Å¬åW ‡˜ 2cMÛûàH°šØ´SÅAeP´HrŠŸ•Jk‚*è!N!ü¾Â6ßuía/HM!/!%X¥FNãv}YôhÚƒÁþûs>3]0Eu>aŒÜÖ ¬DŽMιÁËTžYÈ‘6,Ÿ míª]D íÉ‚v²-›Å|›œF­™ùŠAä{³®SH¡#]vå|ƒ% çÓlqæÚ‡M¼~å̕Ԧ?sÅ @÷'0’}zGµxêc#ºwŽ0ʇ`e…“R|Ž-4i‚ß/¥'%Wâ(D¨Ú¶#ئ!C~Î!­ßççÔ£ƒÏÌÖfJ!x*¼jEA6Þ?Þyðû.Hª7#:Nó¢ ¯iË”P±‰ ®ÔdȪ¹žídÑO”Ò’eaáOvôeÈfc*š•œ†8b;Â…XspêæB ¥éõCþ‘!–\± ðÆcá(\dñÎu_ d>‹ÝrÚ‘¸Ðs8À+´&ãé~MÃ1\O Ô$Ʀô發hÇÀ'ü|×Z”úºÙȽð·Ý9#`ÁFw0UœúÓW aæ Ms<ó±-A™,â….7bi¨CžHU“ZI’r„vÓö”¤Œº!§¶•‰ <¿¿Ò`‚zû;×”R»{Û`Ù ­iIO€O$" ?vËX•‡Ý¯>¥#î*àiü Z\æò0™r-VQ¶'êVšJP®-wqàÿ]€ùGQu—­¸oÅ»VTÝûźÂu-›å`z½Yv`·/ð‡Ì«Ì/a˜úðq'íÆ˜‰¹óY£‘—RžDÊ%*†mî^ >7ñÁ¥¿ލ–%=‘½˜£çÅT!Þ‘\bZ×Û«M "¤)tË@‘ ¸E«ƒ—Þ„XÒNÌ,lä+þ0D3%×lþÊq6]np5ˆ¶^ºŒ[ü2¨pM6ˆý ÷û20 _sÍÆÓ0vâµìú.óW}ç2ö3-Ç£Ææãª.6:·Ê,|â1÷ Íâ ˜!QòðnŽ 1!häçd™ˆ¸óÝ¢yÛð—àß~kEÁà½Y¬pÂqq„pÉúdžK¼Ìv ·õé"m}ÅÅaHˆˆÔw<Ì„ÀÌŸ@þÀST¼ÄŽø‡0sÎ}ÒžÎ2×σe†R‡£ßî”A^½îÝHp=‚¡â‡V¼nÅw­x°óODˆ"h ÏzvÑnÿkÆ_ÅÅ&´Nâp‡uôO+0LQ¬ï1–àd›ÁqCè·ào¿Hµ~‡Sìã}˜€½¼`¶ fØQltÍ5îΚgã#Ðr»e]‹­ˆëþs@ƒ2·ú¸µ(i$¨À7ø–ºdŸæä·\{h¢Ìé¦,­»K8tPßÑûu؇K˜²,IXwóA®p½ƒ= KQâ š Joåù"Ïm Lyç]Kþجùžå‘ð©ä®ÝJZ.›íÅþ.²ÚHG÷õ¢‰#%Á9—& –5ŒÓ›ôf%½ s#ÀŠô¦rPû’M4ú1"Ù)JZX ’•oèp Î4žg”Ñ=ðÔ’~Av\ij”UÍàFÀãåÛ»ÚZOaÝõ“òî1ÑÔóÏd›&´S §:÷e%Uw¥™t“HŠ)Zá4k²’oLW4ò—2Þ‘R×ÜH€…ªË\6ÝÂv¬`ª4Ö#snLFŸÚâÐÕ(['£¡ÝÑ Âa¼´ñ;R8—_q.Ð<€Ù)íYІÈÊi)zƒæÃË‘s•?6'³$w0œ8R§‰QñY+þЊÏ[ñ»V\1XnÒxN$¯Ê«!c/ýhâü‚¬RÐÎw<ÞnåÞöï¦ü¯(®)q c¬Û»«“k—›pŽ(HœÕÞà“_h^|Î:œ°7…²YHÑãyâæsDQÃÝÁe—rzK6@+èz ¾]sïvçËû0ÛïðrÞ`¹{Ç>]lÃ7£ûÔ ŒÈÈ•¯œðÊ/¥5ÌYO•º\ÕîjáìòVȵ öß¶%àqUÉxå¥p&w9;©/Pï÷e¨¹¿ûQgpZšß Ðéš{ÅtÛ?ñùMÕ¹%3¿N°W¬Çn6Sê:>kWÇçلεðõmÌBèfÀ{d×.¡ïuÇ•9ødOɉñ®úõ`ŸYzmîjÏVz¸Þ½Uç{÷n¿çƛg“·»võF"^&ó½“ itÎ5Àp^ùjÔ{߯Ó%rbtm…ÄkÅdžC3‹+ŽxÑ$§Á‘dFt†¯ÏÎí3Ë鳤î˜òXŠK+7µ(ñ„›‹“ù)µC–þ6'NÁ#ó û,dœjÌÇ&°bsqð2dRÞ…žŠˆxÞ~DÓ";ì”ÛßÖ˜©bqYªËD×r¤¿Ã“eØðhÁrâÊ@í -]ix÷LE ÊNÊs›7±¾Á'Ú’Ìu*†´>VsXa‚¸d§ºcaÉ\ÓStÚPάûçŸl‰‘N$Ë-ÒTİӲÐEÁ`|Юr=–¥/ÒkMÌJ×!!;UÆdÕ>á ¤¹{ñÕòð"arÅízJÏïcü*Ä‚º!PA€ã©ËÁ37Ü^Í[Øb¢[ÞbÃ÷œ°*f@ôW¬gir|]òÓ5H±TR¾°r ýÕ¹Ìë3ŽËív—eóÛævn¢îÜêu³~µŸ­˜G^Ÿ»¨ëл«ȬÜî¥Yàx\ „_¹çÊ®¿2`„!øÀ^2 \=ºpäÚôœg;¦æk¨«ÌpýƒÒ1|‹”Î4èOA>7âVÍÊãò yŽ+|ÎŒÓcg×ÙÂuõ(pñˆ{J|ÏS=1PWi•wçÄ>€LíöcÇÀW¬¹ššÚÊÓ‚Ò„]l¥xC´K÷£SiʈMÃê×›Q*‘”Ånxœîà8d ÃyùÓ!¢œ²™ãVZˆƒ§tì?Vpó‘hÏùÝ âQè3¡ª+é´ƒ2C9—÷¬ÜüŒÍ2JÉx¤½ÊÅ Ï¡œ~wx)”þ6J ŒüG7©¼%H/J¡V ¯Íš rï¹mÍ’˜lþò³©õ‡)P”–TwÔŠ`N?•×mÍ7EMÔ½Ó~Û–·80“àìd'<½ÄãÑÇ0‰™ŒÀù$æÞ÷+»’žˆ‘§ /,>lF^\¾Ãw¼Oìj4`NÐèùø-èçç% Еù]Pv+n¦{Vt_ÚAïò;¼Þ‚M#ì<¸ÅF™D=M½œ‡,G.eõM¿þL Ûvï  éø£k‰ùÃ]»÷<÷!ÎæÖ’ÛV®ý]•ÍÓqx ‰ô‚3GÃ(aÉŠJ9mK†NY{þQ•çøöCB+Ñ5?û@lÉendstream endobj 257 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 5701 >> stream xœ•Y TSWú1äõ¡VEûÔ¾‡Úº‹Û¸V©»hQë®U„$öYn¶„%ì‹€Vm­Ë´Õº¤N[Ej[«VGíL«¶ÓÞçÿ:çüo4éŒ3=s‡Còîûîwßïû}ßwá}¸tÍÆÈM›¦Os˜Îàq#ûp/ó³Q.÷ðñjÁËÄÀ%‹^žÁ†gýù @óH&e07;÷ "ø<^Z®v©85W’´71#t|Ü„ÐéóæÍ™:cÚ´y¡‹… ’¤¸XQèšØŒÄalþºQ—”‘:~AbFFêü©S³³³Ãb…éabÉÞð “C³“2C7$¤'H²âCWˆE¡kc… ¡O] {úÇR±053#AºFŸ ±T´m‰xiê²å’é+32WeeÇæìŽÌ[#_›°çͽ‰’ömJNFMÙ=-oºl†|fþfš;zþ‚…áã',Ú1i2A¬#æcˆ7‰WˆõÄ«ÄXb#1ŽØDŒ'6[ˆ‰ÄVb±„ØN,%¦oˈ0b±œ˜J¬ VÓ‰b1“XM¼AÌ""‰5ÄZ"œ@¼N $AÄNb0E !v/ÑM %ú!ŸA¼L0K}‰QÄ¢?ñ"± ˆ ̼ ž’w¶Ïœ>äñ ¨Ì\!WW^yAB¤ª‡ÖþÜ7§oO¿œ~ßõïãÅy/îÐoÀ'c½4(gÐàø _‡þ¿!áCâ†t ùlÈ£—^{é(ÝŸ^N9´gجaÁ/Ç_ ‰9òpø¼áEÃ/ˆÑ8²ÏÈp®~W\ð¢‹Ç…]âs.:š”¥dQ3°±p éhûÛÒ€œ';á®Òà ¨ ¹¾sÿÊ­ÑYâ= 0K™r¯`–²T°OÜ ªØCä6ô±à-?K3ü-5Á™–ƒÖ–Æ§ŽºõnYލHŸ¯W³_89ö¥ˆ;N·¶üJ*~å·édQ ¶4ÝÇ šš@g«ä³cÈ5P+w\ƒ\I8éïvïZb`3AvW©8ÍŒÁÔ gã {ÒT­*,fôzyºV“Ò¾Ç8 D‘(»>þçcGjÛ»Yí @oÐêõ:-Híå Xê æz8´ìÕÚè‰^¯¤\týTǃ‹pfäH.ðñÍ$m_ûñW¨Ë“i™¼©‚i¾È|MÆ#¥VŠã®™A|á§Ë¾ävVZ¢aä‹–iò :í`4U?Û’I …‚ ¢qÈ  ä9U! ѲÉê<ÙÞñ«µZç¬.-;VÒÌš[KZÍ­Ö+Áé÷žûændðx¤ ™ð´v½ki{‡1‰êPWýÕç !¸„@¯† ·}©÷“‡{HàG¾C>Œ^q[Ñ8á/ØŠã×EÚ[0/E@Á¢@÷+^§H4}¾ ~.ØïóÀ»Àk³™<‹îZkA©uÀÑA=3 ƒOó9‡ÛxC/åQ?RžÚûæqrz_ 'ñKUª«ª+úòD4)«·#A0þ¾ÂP*ÐÃü±ÿpœÑ >ò:ÚkÍãGp²Æ ýzC†1¬‘·#vpØ˜ŠŠ³.ï¦ R€ìŠ×J· ;¨Ö.ÊÕ9®.Þç8¬"x–n–Aš‹ø`aªSî,;Tr’©7w2@ý£äM™F—­×°Ëµ¨¼ ‰Ô§6Þÿ¼µù0S“ÚY| P-ßäÊ`Îd7|E«“´y{³Óòw&jëÞÆŸÀÝ0ÈÄ>:.#t”t<•DyU¯Ô²ÓƾùPÐÞðÜ÷Ðx4 ÍìòE R^pRA6û x $f= ~ ;€îgùì^r,œ%ðžþÏ.(è­pUÜY7­XþœsŸh´y˜œš·ðÙà„‘ùÏõð}tçªo°3ù¼Ûû^"ÏöÅ-ý"ï&åâÞ/¸~¿­!˜µj¡Nžæ­!Enf£8˜KqñjnóýéíÁ7ÐgºžDýP¡p¶Õ­!›!´=e3î» I‰ЇMÌ•|wäTo.5Û@9eWÕ‰¤IʨÝmâ#7>þËéVæ™¶ÿr š¿ãs9GÛT6I¦F§b2b2wjþŽó—zÎÁ ÈXÛ¶‚5è : ¤ ñˆ££òðÅ0°­úšSV+.„MÀîQ^~Wqm­¨(MTc6ÈcФjÔ)z•j;î:DmÀÒn,;WÎàâú„Bg…ù '$½Øއcm,7ï/i(;ŠUÓ­ç9ÇZWtE1j ý‰| çq—ÇÏ—çT4+¯vOÛ:‚ÛÏþè4|¾äΎƒ¬¾X¯*Ê}Tæ7’‹¶«éõ»N\ƒ¼¿~wëjíÊ,M±°p+Ÿ·|e |!©Ã4öcÐ(ŸÜÕ‘ p„ôä8NSÔ)3¥‚¥þÝ8‹È¼§™ü‹¯MÈ!¿äDéOŠÍ³âmRJbd€y¯~<_rñn_‚øì1p#Ýn_‘“»–¡`Ô À¿ß.€8ôÂífµÅÒ6m÷ŽœX@­ ?ø#[užnìé9{P?€¹é  Ó=(Ü›ðàEw‰ìWbɧ¡¢¤ÎÒÝzè0Åt@ÙÛ­>Z ¤ ¢ãJ÷ʼ©Ôúß{F&?P`“Ïä¤ÞÖ,ÒµÉÜðÆA ‘pØ]ðàÚÖ/Påü—æw²Ïb&!1ïé÷[Þˆ‹\>–íÑ…Üópñ™ÓçaÔyžíìáKPsŽ….ºäƒB#¥ÌœUµFëP¶ËÞÔ·¿½ß¨®WÕ°­özS•Á,qZåtV·T+º6dÄ w²iΤòd@½ú‡S÷:RŒ6'C¾wgjnϘ¥X2©å7Þ€Ãq¢Ó®Gp šXňà ôNa÷ÑÏžw½·_ÔšõëÖ²"ô=qÝ¡5öCï4±­+¸ ¨»wÁòÈWãÂØ§s¬G±Ã|Šý|¦»ÁùŸÇÓ8¤üOÓ±§MÜÞÛQ†¿^¬î6DÆ‚§é/ÞéºNS—7^E­ú/t‡ýüùîœ¼I^«Ø™«.–è Yå ùÔëÇ2ß¾Üv«ëO‘LDî~Rµ+ Pª¢’«¹­¤‰-i4·€*jˆ Ô«ì餠‹^æ±x…ü¦k[ìæ žQû†ón!O=âs㸠tcN½8%C,Isf5·55µ1hÊ#û¡qÒG×qä‡ðª{oßFûM{„·¹¸.ï&õ·¡é!ŸÛýÜmŸüD‹ó÷e¦íÛ‡µ™*H“€ ¶-½*Wš’·â.8 §…ì‘Ræ2·Ø[¥{÷á :N–i‡ýî¸2‚LÇäWïU® ötŸZïR“x0è4Ú)ì ØF¿ql±¯éŒz6´Á~X*nÂ2÷/ŽøbŠÂ|•t?‰Ö¡Igà÷þà­ö|Գ׿½Â|{ÅF«©¾ìøÝ?ßGSY~«úúb“ê‡ÂˆI‘ËÓ#÷ ñ ¿ú?È7ŽrQñ¡ÊEKÒŸM’ží2€˜5wàl*knøä®n@UØô{Uz-VÐXdõq(m°ÌÀØ‹A„ncIÄûçšæB„_…=¼_¹i|.6ÒÖnÐÝè.“ŠD˜Ø ¬î¸dçÉ\¯ó¹ îGZ^Žj q3j4|U°B­,¹”Þ”Ðn)1šq©ïwáÞ:›Ê¬7ƒ§½ÅêP™•¥žÄ„A<¸æÃÛéöŒ–Ĥ$a‚°>õ@WWûÛnµ™ÜÃ;ævgêäÿè÷®èþeøÁE>ׯ½DCú È{O›#Ñ‹‚?õ¾ GÀM0šAvô-½*þÖO0ò'ú핃³¦£‰ahÍX/€ðî”ãâ:‚;O—4”¾Y#UÙlèpO£ÇJÂdl*š¯ÍRmG£µ!*‰>Eܨ·Û~ÖíTËçkóYu¢jSa°¿â‚“«ElP3BC˜¼ J­=F[ÅYÜÛT€NiKjõ…޲cF[oK1Àõ¨ŽÇMv_xyÃgSÙ³¤EJµ†‘I¢/UËôY: ¤”ióÁC5õíÌŠ'³“ýnò ³Ýlžë|œ5O ßäs§ å#¶[›çÈ t "ÿ¹D§Ä¡Ó‡äÚ¤Îvs[™¡·çÄe¹­=|®Œ¦!uëôMð õýëWp“1z X+©“Ö·8š*ÔV™i8ròí €úì³äEãPàl)Ψ´E¦,.òOï’Õ X–ƒ ŒjÌ1Ëä™`'vCAþþ/<ðãT®ÆÊWNÝEUy˜ÃâV½…i×ß׃ :m^šîÕ15ØWÖ’»NT£ÉÔ©ä³tR­\'ÓÊŠv«÷  :J™§ÌÂ^8j¯ëO°ê°t—X™`|â÷àxUÒ`jkM7ÈòÇeÀè¥4žw_~zÂøü€¾×¹!jÅâðqÒ™Æ$l‰Éf Ts¶=sØV0é—=ßÃПyð¾øÔºý̽ª¿ÍÔ¹ÍQ š´yÙ¶ÄŽ¢Ê†ÆÚ6kq)00ºz v@?’/ÊŠI±{R…º$Ý›¹‡Ùƒ™;Û°?.Oø\w®Ê¯Í-(Ö©™bwr(êÒŠ¼ZimØöH…É™2œ~òªÜš ‹ÅTÂØ*ßënÄúc)®PÔfÔÈÚðˆÖáh¬n¯kÆémÞ ˆé7.¤ýTq ù1'Ð5æ)“™ ÈsÑ{Oˆà‚üb9Èó: ™‹Á*¿’¸Ògb#ÙÊUÑr=z @Ò k‡ ÎÂsø_a#†Õè®]Z¿~ÓIJ€B¯NŒAUÞðj(Y¹¼Áa2•—3ÇO|lnTOÇæM“¦,™»ÝqTÎV”V8°¾ÖÈD;ÀÖ‚™7·ÝƒCaÐCøÒýÕ×ǤFk2wxxýà–“…9û€6Ù0»ÌTCNƒ8[˜W̼… ĸ¡h´‚N “‹æÒÍYÎ4azjZj]vK{cKsï=™Ñ£\¼Ë=ð&j(¼I_«¿z |N}¾üSÔ Ÿ0ñµù?ì+Õ1 ²f®S¨i_®€}à+¿Â€û§öž^ÐÌ,n§×lèüèlÏ©¯¯¿ýæÒÍó±ä_ o<ú§«_\¼{çÈú·–¼¶ÂsŸo8¡'ñ.(øðá'œ‹6˜ŒåØóêüFaŒ&V©g²Ñ[:¹@R;šLG 3`Ìï¯ò 0\…E˜ÆõýwÞ†Î=W•?rrÑÄV.®°Ø*0‡ês«¤êB]Q>ƒž Øe‘g=È­R™5ð$J¶¸¯X©šj[]MAi‘™EVø†ûRníùº‡wñ9#|@Wva‘Op—ãÉ<·Èw¸uí+4™v¸å?Åý‘Ê$üämüÄçKl k+K+ ÔäW›ÐI˜,«P= òr¹ÖÂÒbV Uè:Š,PË @nHfM~uY¹É\æ>Ô<çãtŒ³‡tÑfØ -ŠòW¯çÀ>¸+¿Ñ·C!j))/6ÊZèÈŒCUèut8Ö<óÜL¦5îܺV@5YêÊXŸ©{ØÔ®? ÞU»þ»ÔCë\b@IÔÙ…ò2µUÅ> ¯E}À2ðšmG"4L­Qáô¢ ÍÙÍ¡.„ïxïÞwÁvü‹w磻gø\—N+*%v<ƒL™‹gÚ§Ÿ\yù³3®ó-ìþ+%ðE`¥î%ÝE<4ME‹ÑV4þ^(R~4bì¥gZ@ …ÓR%Õe>“+‹Mݨ‰·`œ ‡Á1pÐ'U[%8>Aè:9yÿÅ>ýòÃî¶=kÆG,Øê×eŒzü" §y{5OÎF¸;K}q ¢þùF°6ã®éU‡¿&k¬;ÿ1£öÀð:Þmnþ wpé+ï#ÉF8Ú`Æ£‡1¤Z^“%,áR%A“hœ¯òÑd:gx5Ð…Hm¹îâUaôü+‡uÂ¾ÞÆb4vhñÚr©Z­ÏQƒlJQ™í<¾xôžøÐ®˜´½É{›³:ÊŒÀ`b(+¹¥åðõJ¥ƒD¹¥/¸ú^êÇô È÷$ˆÿ!ÿàÐendstream endobj 258 0 obj << /Filter /FlateDecode /Length 12615 >> stream xœ½}[%Ç‘žŸüûtÚ`ŸÍûÅ6¤µ°+ ì$?4çFjg¦©i’"õë7¾ˆ¬ÊˆêsšC1½žŽ©ÈÊŒŒ{FdýåÆýÃÿÆ_}xñ¿õæÝã wóîÅ_^xþ×›ñŸWnþÇKz"rî®û›—o_ª¿iþ¦æzî1ß¼üðâÔn_þ™žÍÉ<ÛÂ9¤@Ï¿|ýâ§ßܺsp>ùrzŸÞÅÒNonïÜ9ös9ý@`ß»¯íô;<Ñ{kåô5ž¨½õLÜÅèzú?·4:¡EÁs﵆Ó'~Ô¹ÔN u-hü‡2@îínÿßË~ñë—/þKÞŸiM¹„|v´ŠÜë9Å]Ä?½¹ù¿7‰Nÿô‚†õ.…vóW"Æ?Óÿÿ|•tžþ¦áº+EÓηrîÕßOÌ$z}zAd!*n3úY#&¢´§é7G³+FÌíL¡û´`Ààã¹µvSb>GšêŠ)ösññ¦8‡¿tޱŸ]è7¥„s++æÍø©¤tvK¶:”|Î`ÌqÅN‡Ô퇛RË9§%ÜØú9´BüíIœ—pc牌1—0c%íÒ*­Ù[^ÃŒíœ|¾ÉµŸ}ˆKX§c­Ì:!,aÏ!Ó€þìÊ’m‰PãíÏ4ÍEÜíÀ‹-Ç‚)’˜tWI‘¥sMKö%‘ïS\²/–ŒÔY4O†eÅÆã!…æ–Œ˜ê9’’Íd»¢[²Õ)“û@Z¢´sv+vÆ'x‘VMF«¯˜£¯0-àÒdm ÷ëÂìÏ™V¿`DrÃLLéç¸Æh¹NºŒ<ÐV¯`ž±!D†1¿ÂRƒC©Ìà±.±ýäø¥a©S[±1ÂÐ!2‘ ë#ȰV08ìÖÖ|.=Ñ–Hu¦±Õ…´ã7¯ænDÆN:mÍÙhÌ6».ÑŽäwSìCaÙÁ¾d@RФxŠ+gŸ×p#É^³ÄúGÌ“ 9(k̹ äæerE[_Â×YªÃæÉ¤ÂhŠ™FNmMãÉÒ E¬Kô7²Á³LuÉ¢iŠŽ]žt.Kâ"äÚM.´-KæWÏ΋kÊW"E’X~RŒq‰»C! #·»Drw–Øi$Ob&­ÈN—%n#iÆT‘û!±ñKèûŒÌDƒÛ¸FëT ¥¡uH×FäÝå›åÎ=¯I$8Y5™U¿&D G”Ü8¢9/1«Èu¤Æz¬,‰9‚‡ CM¦k = ¡‡¡F¢g‰ÞáLB®œ=Y”oìG*ÿ‘¡¹ùLqQ!:.ñò3jˆ´hõ¡¬qO<ˆ,kYãð@d !) çVp8«ð§Œþ»$kä3™_xŽ~M`”G™ÏÖ¸y$+ìñÐÈKȈc ÚâBz7¯ÊÈ µœ‘ÙZ2 ùȈ‹#·.Ys$­¦.…%ÁeŠÎ“Pq‰—çI=ø„s‰N |‰º%&l ‘­¾/ao(džU²ÕKì`pçV™¼&¹L#Õ&±[MküoR8)õ›½bŠ8‘£EgÚò5nYFnG²5K˜1BÛ€ý­ãK;Nu”3Å1‹¢"ΩHP,ÙgOsK™MXÂÝ\B5v¨È5î ™UÒ‰…üDz*} ɾ”5dLg,r{Ö¤XIJ"ŽÈ*¬ñæ)Ž®pl+E«kã8¹ÌäH,9¯E,è+Z¿FÑB×x¡ôñâ¢óyžþé÷ÿô‚Ï0s‘\N €>„ Ú†3qb¹yÿ"$ÒH/OÈëý‹?¼ðÈÎÑ”Jòˆ¥h$_1Jê¯ç˜è)ò’Ó™Ò'Äba¤À®9hC£“gIs‚®€äñT€žé^CX<§JÌÏÛFD!ÿæT Iµpp…T»'±XûœšG2 ŸÉÁsâ`¾"ª—9U¦Ó„°xN Çzœì gÚ/šSk•g^ø3h±9!4!‹ç<=…<é5¡S p“×Rã  B‰T4ä€Å#¥Ây¢i_™ RëBpŠRd&ˆä€Áá¥Qôƒ! èxi±žù°š"ÏÄmÄLö 9` ÁË‘ò<£ž˜H‘¦d)3‘Äbsˆ³“$—Ïœ×A’l0S¦˜PCXÂL„š$b!r¬ÁLdšÛ˜ƒ2xÛ±ayõÂÛ$]8Ÿ#ßµDýv±X2â2pM@6‰–È‚éõàL^G ™§¨ ‹‰MÀ3'¡nˆ,n?#ØÆ¸1 ‚kÿË<-J$ 0 'Y]v‹TFà©$Wq¼•A{â‹X¬Ažv†­Ï(Úa)#4’AÚ‰ ì+v)ˆAâ%Eâû ž‚ÂO“ÿ9ÔZâšZH$‹Èb>!KŸ8>ã°30¿bJ4z´( ƒ'Y«X,Ù0rÀÀØ‹ÀDò¤ô9K‚RÞ¤`‚$^›ƒk“!è´6G’X]„³%ƒ#ªÍ@,Ï4^ô8Æ”"iaàÚ¹ò.1M²X,a!žU˜§DxÈAšIA7ð‡ŒEIb±duåvØ^Z8«l×*scJUž Ž-Š6Äb z“ÆyQsÃ’‰‘ ``–èêüË>>›Ôؘ”U c#ÔŠi±™±&Äb‰BÃ" 4qÂDiY©:yب¯±±XÂÚˆH£è¦ÎŠ´h­ÈN;&G¤b6‹%êš"y>N#‹_Ä€RoETjeŸ!VÑErÀ›FV؃Ý›¶iô°$tÙ°i•ù'ä€5ÔucÆ-qS³•è Ì0nIÔu:Ëiê±Xûö;øóÀòÖeïH¯†P}7G„„Ô×N6‹qm€î,Ž<.mš?-Šð5Žñ*ëMfêÙ§Êd"|`v(5Ó…!Tv™|X˜ÊAø#Ñ.eVa¾ÑØ!,¡²—§2ñ [üJ&ª4¿˜œB.deB,’ˆ= z¥î¡“!ö¤ð¤†¬‰4â©Lä 9`ÉHÄ‚"ø×æÙ-H(uQYãkˆÅI#EÅɨiVEXJ×éqîÂrEúª KFBjà5‘Ù mì]œ ùÞÿ´Ë²Zâ f$)dY®±K˜#-×@n'Áv€Å‹ÊìþñI+¯©‰‡oÁ:& Ö>£‚$ èì,º»'vˆß Í~ñ:<¬_mrÀüìyñ9¢Ø„Y®u\¯ÃµÎ rÄ®5—&¸œìWgñ㽬‹_ÍFTCÊ®¨Ù«Œ\^25‡}ƒ5­ KFêmx£Ó@@„° J‚ûÆ’€A®VA,–1%;mÄæÌÒä'Tqõ67Žœ™†R‹$"¾GiFjÄÿ¼¸F„e Q‘Ì!SÈÙ„X¬a¦ÅåÂr3 ÞÌ}ÃJ‹Ÿ6‡ÇÉd—à!Âv"c¤0oŠ €÷#{±X| Vbßm3Ò½&æãÔö8ƒÕÀøÓ>/³qž£C¶U Ã붆-ê!ÕŒ({B,ŽL™rD=(σ8† ÝÈNˆiïH±X# ü|§Þ$2«üTÀƱñKDV6öb±DÊ(Pf^'CžÄ×Äü8mÃ2•Yu)ˆÅãQ¸¾ž+ „Ü O³ÂÀY ±XL'¨ÊÎ9ÁB|[Q˜¯½˜Ï ¶ˆÅW6¦‰iâÅ%˜˜,ÖKtX,Mœ¿ 1H" ¤–ñ¶ÒØBA@àBÁÛ%góG^.T 9` ”*^k¯ô.¹ä@$ÚèIü61Hâ4àÄ)KàÅ\É "Ÿ‡è©ÇÍIàhmB Òî|àh¶ÕrßœZ~îÉ‹¦!çƒô™X¤!ùE2RTÂܨ6ór‘å“è{öþÄbÉHÄðàD¿â„”Ù0,Zb_SA S‰ÿ3”©§Ê*¤vÞ š"X¯³‹Ì§9b±DÏâ æ Œ¸ J5µÊæ; kÄ ÉŒ6¯­¹‘½¨pô¡·2φê$U£ kXý΄Dí2O) ûšÙŽ2o"!ý±Xb÷É)d¶ÉVŠ–ldÔ“Á"`Q<ËrÀùc•Ä<'äAÞÄU,bD&Äb‰´Ñ{°+Ù¡ú™¥-Š&I¹Jº„ÍX/rÀ¡Ì(·”#‡2D'‰6Ä{)óÎ)ˆE’Ååa\“¨! bÊìÐfñç[l kÈId}S6G" Ñ‚õ ܆&þk;[€EL€³BvË„a%5Á®1ñNK›GÍ®ñ„X,1o^´yЬ‹a߸¥ ‰Ü,a!Ù±rÀÍ ÝÁÊ”, ;[H¶3Qh¬;JÁö°8²o¨{gg—“³+)¢DK=ç®4Ä" Å%ZA*¯+A÷q*ŽÓœÄIíÝ„„ÚF9¸ä8Ô}¤ðú mÁ ¼†°ÄvûÞÆáµ‘QA1Jdy%mÄÄà€#»$¢Š¾ ËÉEC"…,ÙHÐK<µ(2” sœãbWb4Õ´>¢ %ý¤ klZ?ŽØVÒÜDƒÕwÉ™„‚ø~bFQ˜( il#S’<ÙÊ.!NÖº†°F<Ù$£žPֆ¡À. 8,œï 9`$]aöÇz½DhwHB“(ú'pÒ`,ŽHG±TgWï瓇Þ$%Hn“È}B,–P ™VÁUFò©²Öð$^’ ©¬!I, G H¬´L‰#É Ãb„®!,‰l\>( FŠâ߆ÄiÌ ‰;!¬á•&xƒ§“ax¥wÎ]”;'/ª…X¤¡$Ù°ÓôáÞT‰28O%ÞLEk7‹54IFW”°†ÔH?si€;KB´QôóâiUa0hfUé…Ñ ¶aP^A,ÖîiC§béÒDœÌÄÌgìf‹[7[=?B>Î@STÁ a8¢d79—D˜Å±3 b±„R”¯PÄZsª KÕK¢À7IdLˆÅÂH8Ûl8ÅDU1Ñ2 ­PtCoËú¸Rš•ÑÃìQ~M»¿µ;³»—áÈl^ö“§v*ô·§Y¢ ~>µAôS­°žTOmõT@³ùªj^D?ŸÁyýÔÑOV¤Ö¨žz®óð›vF¨”l-.r…4%œ¼HyÁOîö.Bù| óg?éL`»½#=€TÅɻل¾· W—Gv{@ 4†ûqá)ä Ä<ÕCÉŽí©ÑOQÌ“Õ#ü§ùw1dóÈ€˜§JzýÌ3c%ê!µ¶ÏÜìÁçÇÈÓà’ ©âà#¡ÞAÝ;½Žõößáå7ÂÍSRð|i7&÷4zªw Ù ¥‡á!ìë|ÑruÝMd¢û÷°&Š p–âøÓ(]Øi† &ð‘RܺO?~Ü«0µ8œƒÿ”»pþˆvâà)¯( ‰8¶A'6|ï%–=ÝE*ÑÜ’Rø°Ü©ºnŽïÖ…tLhGŠ‘[ˆVŒˆ”™(®¾¯+ ½N†k^9"Ú‡F§Ó¢ÑÄÐWnŒO\o¸pŠè’ëaaËAD†)¬”>ÅY9E./Éyå¢3‘nm@³á ©F!s¥üª­ŽäS¢/w!;¢Þ§û•z‚³Ûq¥M@¶–šËÇWÔþ¢ø®­0T‹û•[]P‡ 7îô%E«'ÄK瘸×`©.ÃåkL@¢j©‘AC› E&r23¬T=¸žÝªëæˆÃØ.=nI‹WL£Fyݪ=ŠF2÷‹·%ÍØ8ëô~©m%> KŠ"i-´­m¿yåÆàȈUVmLäÓ˜¥–VƯô˜#Ú7Pè»L…£v$pÕ²È#¡Éb)Q˺ҟp(Xi+Mk%Ü-]3ÊBÂÊ®ßÄ©ò²’›dW׉ Ù˜Œ œuŽ*4CXÈÞ¨Ð'Z9"îB@úº]áLäB¡¦P¦ã0vâAÒÝ/õ'ù<|ò¸ÌŸ(¤ÀK_Ê=¸ñª®ôqçß‘³ÎwìçÄ•‚«èˆCn„ý+c™x.e¥ùç„BY9Çp’¾6E‘øÖuŽ®Ã…uX< ¦­õnq0½4‡‚Jà•6&V”À®uóH–—º<°Z).]5ñßöµncp&͵ÒËÂ’@²3+Q”‡Ö•6â%û•ú=4y¹nlK³2\ÊÚWîLFgÄJÿ€|³ÔÂ|é[¿ÔC©ÜP²2=ªå¥ço­Ÿ=Y®ÛÔÓ,1Q —VÚ-X·Ô¶¢B¤¶¥a—c¯<½åê>\º·ngø2É¥é[T/µµbs̵§Ž…C땉õ‚¦ågÅ…«:Ô+—™U)£Huôldö F+¡îÓu±X(šˆM—¸î(¢¼êËØäšâÜá2¢Ö"ÖÜÏ|ª>! ã$VîKR1ŸP¹…N$WFwDjNúÄbñŒ:Q›ÖÑ—‚¢–äÐÆ‚mD·-ŠlbÏÜ$¥!kŸ×´º,Mr<§Â¿AÊsWLv±X<’+•+½ͳʜxcæ´®NN:¬5ÄbñH)ÄÑUµÍ ý½ÝÌ ö&õÿ;ä€Å#áJNîftYŠ/SA±(JkË AB‡ŠÜ°C,Ä—¯&='{v3ìPJb±dNu²ï#•.…ì¯àZŒl kì]ærßIñ–¹™_Ñ·!W¨þ´(ÂL@ Š-‰QÐ;y0¢t«d±H<ßKŸWFR¢Ljò ®­¨É@,ÖàïÌ¥ws¤Ž«­,{ó {$™yÑ`œ)»(þæDË.©…¿%` ‹‰d3†Sxa&‹&)™·[A,–¨“È7•¨9á{6ð ”öN ±XB&”iê)¹Q£«÷Iô™‚$^ê KT3"«€›+æë·L±H¢â¨: Ó)ñnÔ˜NˆÅ’ëÌ纇; "îæ“ËQìùj©j ‹ç$UõJD'7)ìòŽKhš…ž(b8eçø¼Ãd@¦ˆœÁFÜõ¢¤Ÿ;&ŠÑmh™hb±d]£`î>ª‹ùXig?»j K’’@©u0\ýS8` ÖŽÜIgY»õ#kŒ§ÅÝ6Ö2·­£åºé]jX‹Ù·–0w·³LæF¿ˆåîTìn‘Dq;Ö/¥s×ôvà;uZ/KYn×Q‹3¬[•fì}ßR6Z$ ójˆEZ›ÛL'#E4 j>"­Ý¹ûKA,Ò¾û\–3÷,ð5OZ!V‘+½‹KD$Juû)£ë˸Eè"pÙ@,–ðvL|Ù‹âmÒ#ÝxJü­ã)°„ÚNø}j$ÒÒÒŽ¼ëÄÎ訖<ÈS;GÆž¤kç? ¥kBX¢Üh%O™ëŸSŒR¼›Ê AB_#7eîÖX]`?aò6XÅiŠ'4œ&½›G,‘·G7áî›véøUž(¾^Ð Äb É•vÖ¹ºŽkAK:ÈfB,–Ð _sÊz¤äG7ᤠr¬Í@,ÖXcßpÆ­K7ùôü޽X,±K¸„P«J¸/Q+݈Žýò4HÂáð V hTªzŸpa‡7‹4ÜniõPNj*ÍÆ9´™;`íÊÛ0+ï^ž(ïvTÞf¢ø( îÆ× 5m+>5ä€4œÆêsZKçù•Vøª% ±HB%Q(¸{ÄwC¹²AC,Ö0àâ‡)?°ÈýoÊëC;§u „>ôh#/ô¡›8¢ÅÄ,‰M2_Ä¥ÖFÄ5šÍ;3#‹#‚À–¬ÑsßÁä@Ü+oxô€$óir/ÖÔ¹¸ë„Ûw Ë_ù3Zø€5â·Š§Tœ‹»U¢Ž(îhoÉ@,–Ⱦc"f\…µ»–pZëb±Dç’Îã>ïi¿»4Y*ƒ-WÜhˆÅ1Aô®’ØxT¥"ôm¢0ĹéÒ;Iö CW\XŒ¿}@!‚‘4l«’ª^ÙÝ›¬x@‘­Ç5cÈ¡9/·x$ßGK{艿æÕæŸöyq#HëD³áºn_ŽNƒâX»CšqHø(I9$|ËqìpsˆþÆí”;qV4,‹ »u,,Ö0j‘Ó Ó±M$Å$ øÄj ‹GêI® i[¯{Âý¸Â§níì4#üS‹%j66¡Á ·aNzîžï‹Õ‹%sÎÜŽŽ„kÔ]•HvB,–Ð ÚÉó…6š(¸¿ÑøÿIL¿ë^Uz+ßm–pIAçê±+ ¹:ÌqBX"ø¸ÄѦ¥p…IK¡È­Ú´”Åa#7†âfúŽh™l~ðEwÀÁM’+šv¾ÄÐÙ¤Øp‰ ‚@­ –¬®ÈmŽÓ×’&råYáº-ùêÒ1H²o¸f ¨À=a{¢ÜIR"sŽ‚X,áJܽ–ŒÃ]·@í^ó×9'Äb‰eó¬ÆW¢ý7®,ïR–¨ìMLϦqrT;2¢?4Äbœ‹Ü ÂR¹ŽTG¡¸Ž´Ú@Õ` njäï#¹a§Në=r_ìé€%wr›ïÔ—=òÝ6j¤Ö8]ªÇ¶Xb¸½d”]Šý ³4’+ÈKö.Œ k—:\Æ’Mšµ:Is(ˆÅÎ75+:áÂ+•ȹ) ±H²uÙ\@´qZwšãL°$š|Á£NO¦rHf¹âKGãkD‚ríóŒppD0g|>j"œÖÈáyŽétO®JÖ9<‰EuÏ`‰*ÈìÐÌ)q P5S‚.bF·[+±4A_c5‹%ÆEöÆ3mIÄ~®¤v¹I%Í-–Kž½U¥jlEö*&û}Àž)|MiÞ 7cél­®]|À)¡$"¦ùݦ¨I MQ±†7èX‰ª 2¥Cþ¯ã®Y“[<`‰ V™wô™€óÖf\‹ÖîwÜn¾!ÆzÝ1½îãÒðÁ›®í“w|O¸²F`xk±XÂL©«7éd\”†Ž}ãɰ~ñ… í[9©vq¢ ßI§¿Ÿ>µCÔ•ñ&UæSD=•\aµ9_8ú”v¢æv>´AôSÈ໢ŸÚ ú©ÃzÔ Ùu ¸\9‚ñÖ^§0»Ø+kà;„Â~Mþ…§¸³oBÌS­Ž{ÿö§D?Up' ßm¾=µAÌS¸àžµßþÔ€˜§¸-ê§Ä<…ã4á”í©1Ou«§%~îå i÷ í³N®Ë:*<¶Í#ƈÛ¶A˄ҾùÆ- ¨7§öáÒn —E?AßQ²Aæ} ûbŸ@Ž70DçY®×߀)#R¾z Cô¸;Û[¾1·0\~cýÞÞHBM95¹þ÷â7·´·Î…^O_ã§wd Oï¾»½#R÷^k8}"xt­ÅvzC¹5øäÿûí]ÂÕ“>œþ@O$_cm§oo¡{nýtÿQ òƒ4Tcœî1` Sæ ÊãåžÆ[JÁ;p€Öb.§wühk¥[øã-îI#}—éçc¸{´¤ÓÃG MνËó~„BívºÿxèÆ?}ƒ×÷â};½ŸtØ×{úÓ ˆäK¹ãûÐoy‰¡t&ÂxöO·2bçAB¦w¬œ|c¢Ó¿ÜâPž€²€BäÀ ÉÈ’Ú3+Ü–E…(Fðê\Â40n ôóݤ®¥˜¥Þصç2tÄw™eµ®·TõîkùŠEGæ´iÄ^/_3}»cѲ·Ÿg¼¹õ”;äüÑ[~aðý" þo'—O¯4×|¸EƒIb¸Þ1oK fpé; η2\ëYóãáÍd¤W1$õˆiþjÎíÕœ›Þž& !Ž­ÜÖ±ö^]‰ãß*=| h6o~­JëíôÝMÉëäЯ'[ž¾¿Í™ŸÐºIíî=£µL5¾TRÒÓù!*Œ×3ÛõL;(‘ÁK%;HÎ Eb¼âWƒ]L¬`DéÕÓ—LWR’^kmÍõƒ!{ó†O_ñ¦’LÿÕeÞÜ8‘îQ^NÜî‡Ò£U(âóBkßÙ©“ÌAYI«»© Ë’Ú'õÆ,‚MDÞë‘y¸DT>°‹¦'Ͷ÷¯nY(C"1ÍÚí[X74û¨UðD$.7²]µnŠÿ¼O›ê+” éÀß¾xùßþxzÉosü›ÚgV# iOowJ¬`ÉÄeÊì^QƒÌrÞ—XÓqj›¬Ç&Ã<ö=ñ…Y »Ì„X‰6K?ýÀ::Qé%eŸ˜ .”¼f=_ËÈ™¶ð¯l ÄHËT¦aTOlwUjÙùÇßç¢]"|ž>—Ûdˆþ‰ˆGÔ{˜??ΟßΟŸ.>ûþV^c=/|5‰ Áöš¡\¢'îKXBXxð=ŒåÒ°DèŠ\J.q©p4MK¨™£ÕŽ?Ê`±Uãˆ6@ÆLB´A¦e=õ+fΞ\zvÇ–ËÛé­I)`µ‡ÚSdUøâà.ª. õ l3žÏéôçƒ ÞÌ©–Ím…g­‚«2±§~âý4S÷Z/>Îg.ñÒ¸0elê?0›wö¥ÇÏßΟÿ6¾š?æÏóç·óç5>ÛþÃE–ÃÍëŠåHÛà3í•¶ï7·ø®ƒ›ùIhS½µÛSŠ¿áKZW~œâ¥4¹¸žô’lá«©œßù¡3îß+W½XžÀÜîågoE¿ïþšC÷ZVRh%Ê¿‘J0VùÛ1´«ƒ«Å|0…/•g³Ïù›9 å ?ܲגaÕŸO y}]BX#!OdY¢˜WØ\ ÏÒÂ<²l\»Ó(Š¡5a‚寮ñ9¼…7ê7ì8n$u&‚{4²s&k%ÇaÍŒ×SS‰Q©þï˜d^êQ<ñB¨´CœÈpì–É73àûvLÊÕ뺟%DúyjO½D±ý÷Ø™àÌÛ¯)f|º`Ýìáé÷ûíXjÏÓ¦TÛN‹V ßÜÕj"Ö ¿™?;þÛü¹Ââý„&;•]MîÚ›üѧڛi¦U‘¦é09Tbý_‹Ñä]|?MžRR¬(†¿ýÔ˜º¢ Г½ˆû^lêš?A¾¸´ ]÷Å7e:]±{–ãa >eµÑ&u¡ ±cãP´!à‰eÃW~ØnÙÀô9³…JeŠÞÏŸ÷óçãÅŸ_ÏŸo'üxÙûiç‚&m•<ν3r¨•º B9]‰0wˆ1Ûdœ#H ï¯Äæ÷·03¡hÍðãŒ"†_«G‡­òÞÕrúb{"œþ§òM=]¸ßÉ•´&„EÎ꓃~Í?)Üæ‡·wa¢ãžáó’à nwŠ}Zš[eS5~~š?¿ž?ôW÷¬ïŽñÛiÄæ²ÒUÁ¿S ePåô>˜¨ ßT!{wÒIG5†NøÜѲ‰8U¿Iè=íÙI¤]CsÆrjç.v–Ô¡ÀsY?ú7%{ãa&È[ùTœ‰…$Ù’ÁeßÌÈI%–t 7œ|o쌌/lúì ‹ðW(v5«øBið7•ùyþTœu?þp‘EP¡Rj°,R=tÚ'¡HuÏY4&v³I {NËÒz}°ñ+ls'‘QÙœcTfÌüû†´¥@š5c–OÙCÎÊçMf_SZ±ÎÚ૞1º§÷¤Úã` Û>ÎfÅ.§¨Ø5õ÷Sfß\”ï//Ê÷£F»´íÎâfÃÆ~Ö0ö3ÚÏdÒ1GAVñùæ8*»tE¿ƒYǧ‚z“äï@}; àE3ÆÁL·Äa{ôæ;ñÏ//†\í¢ŒgãY$!¦ã$:rio&§=Ë#1n¤VCp#%­ùëìÝRŒÕCâ4*^{É[PV3©ºwÚpò~埒¿Œ²¤¨ó0ïî”0*ó«fòï—BB“ÜÛêŽônoõÆÃ4ŠÏWhkÏœÞ:çд4jaTÓ{-o\ÄAsçZa·Y?>¯rñébìÿ\ùÌŠÉô:s4¢mŸ¬#†¯Ú¾Cw‡6fèc æe‹îryÉ?]Êú‘‹{Lp ¨|¬w† ¨°^»toÆ<\2ž¢áo¥–-S﹆OcÖ4¥ÏÊ"Êûsââ!.åØžO[ƒo|W/~Ê4åô»(î羉Þ8_÷r5èËÇ%øÊ›Jñ Ië¾4¹xæõK¦ÿÁuŽýþ»é.š3¤»cEØÉëˆëáVœ=ž¥uõ„m8‰ñ0Ž ZKO#üèC7~—2Eê¨á^ö“„âΣÙY_Ò‰>¦|ЉS{_Ê+þ„R,dzíÂX€ÛçO"Éåç¦ù·Ã]uΦXYR¥kÕÛßmg:Ï>Jæ¸Hf”|,WŒàå—|£†2Ƚ¢—Õ{¾Æžt3Ó£Ò¢k¶W1×Ýö¯w|ƒB ¯>úwBºC asÙä˜'ȵ3VÓ_®ùá™s¥3+âãQ-뼴ʹދ@f2oûÃ=¹½´œw’poÄ«ÖWA:çr72ýQÎýzÕñî0ÂÀ•\\½uN4…To>]¯$»\×Ň™ñ†v®!W‰…½Hì¹Â´+Ã4Be=ëL´Z5_I3þÔ äNñ OlJœ J¾ÙÓR.½Ê2˜ŒéoŠ=®%ǤàçªI©óÿ—ê ²R÷ÚéJÃ*Uþ˜EÚ”†Ö{É—ùV„‡}8,¦´„£b >#“£³7V2Ž{ƒoEÆy^ý§Ó”ÀEL„«÷*!š]WéË{n1©pDÅ`µP@IÐnÕt~öð­­Rì¸~ŽY%>OÆî^ÆžFÅè6ÛqåhL$—ùæžø6âcò\$¨œ]¢Þ+C†êQ¾€V'ʹ”Áá`«/ϤV…<ë uªòï¬Ç:œJâuÅ™€äA ´,}pk"c*ùøœ± rPvU+ÌCŸ/d©H¹ë‰žçR@’yî 0H¥Ã¡ôb·ñ×´Ï8.ŠU»KÎ {«¥hU0¢úûùóÇÛK€ïæÏo/ú&Híâ¹1ö›¯©v%¡q5!qÀÇm£‚­Ö°Ç’Z´ø+™D…>™ª£÷°ùWäÊבº¯’æ×Òãw¸£ûQ&F¿‹3"f DñÒ\Ff?ÞîUÈW"¶GÃlwhZh®H80’ _͸{¼çJõ²Ì­â3hG?’š’:^Ü8ÌüÌéÏ·{žýêAõ P=:: EÏŠnÍè[ƒµQtÊmçÝYû«ÜöËgÇÊmñ§rÛUrìªÛžÕÉØÑmô»Ê¾¼i.=QD[¦–)]2z-dÙ{>–•˜’&\+¬¾õu1Ûá±DëÈQ^®ã|2fû?}RË”xãI²<º Æâˆ*çªOUÓù¥ˆ_1I#r£ÿ{ù“yÄçê'g}³>ãѶÃùQz9sö(L»Èõ6ã²e¡ÿ*“ÿË\Ï}èC}ûçWt¦Â JU°ë}šÝ"¹‘ɼãÏ/ùr!–ëGÑã‘Æžø1’c<ž5`bÖ;ü8&Ú«ª]˜„xÐÉŒ‡w’”Ê9ëQ]Ôêåçùóн'7%sÉXy&Qe_*›<S1dV¡Éa.†|š!ÖüÍkp¦þ/|¹GÔ÷_6~Åå©ËI9IÔäXT-P1uY×+ŹF­€- Ú·üä`÷ö¿Ÿ!¦MêOøã|¹d÷ú· AH)Ǭ¢ÒûØ?NúíöÖ}”õ’ORõz•…{Ð^¥*£{8ú5{ù¨@àfü´cˆ«¼pi U#ôsssýÉÜpmP{™H37ŲãtA¦`-ñ7ÓÎ>YÁ¨ÇÒ ’^’¢çPúëR‹K^‰ ‡L9• UªP²¦A‰_âÓgvá»÷dPIˆ÷ô²jR®ßOŸJé³#Ú#¸P7 fSâr?…Y v¹]Dü©gA Ó¢^k0°=ƒ Åvöô4räø7–’[}/;›8ÝGÑó÷”7«Ì¯Ö“Ò‚³ÏŸª_ ´_I^y3Ž1¾’÷LÑÅÓÝÓÆKn4š_ÃŒ^UÛß/K0£åûÉqß!ƒ\\V1#Z|_þ™¦Xó ¹ÓÕó þN'nTBAOó¿lÆÆì¹ÕJš”Ù¶¦$ÑJ¹Ht®_:¿3o€í w-èz©#ð¬;ʽiqâg“o‹ÊöÝÛl2 ð#´Ã% ƒk8‡_ë×2Ri6½ÍÐV7+ÍÙ“+§PL”ž˜>ŒI#£ æ·»êw}-ø¹—k"9€Õ]÷И[–´‘dxuÑŒò·ƒ}a´Nt>ȬœoZ Žù”Q­?,\q¸Øk‹L¿¨‹eäØMjŠ)K>f«1ªILÝE|°Ûÿ?¤ŠñaÑèËç{Κäi«kP÷©<åhL’hœ¨ÇG]¥û_’5ö dŠv6£ý˜¥“«$“ø/·ûYÚS#!‰×Ç©ÂßÈ,Ñ¢‚á>rZË|ä9®Åäô?“¤Cè¬/‡oÝ­Wf3U@ŒÃ½g‚öçR¡ã-?ΈE))ÁdY³ÄÍÖÿâS .Æç*kp[Pïjùi ¶úK{¤Ü˜ÐJÊͶäJG; e¾Tù÷¤jjÐår}#‹$>G[Ÿä|´¤<Ù=í²GýBRµW­§%ræCne´öŒº·ïú|0««ñö‚…(±Šò[µ’úNÇ—†x#]ówã-Åû'a ŽÅüs,/w7ü²ÄÅ>‘ »¨‘<w 8šEÞ¹‘ï~#§WHc]ÉóÈþdOfôt¯N¢¯(¨×ú(úRðà¯ìËúÜ"¡â9 2÷gè[<›¨+nö1‚ ®qQÁ|\*…ÙÏ£ðrsT«нÔÓô­zT÷+qËã¶!ý3c&&v¾œàWkýIÏ^yëÝbBnrÏñ<ħíé˜ä¾(ž¯çŒ¯xðß_Róß}cÚ‡Œ¶Ì2åQÒ¸¸ÞtVìªëýÊ]ªÔÿÊÙùh“â×$g—?]žõ(x‘»• ðQŠWÊÁòÈ;’‰a•+p?}Ù+çj{QÌç²Ó®j­§µ2Ÿeq^ýiãKÀåZÑË„ñ<[Ã.ŸÙ¨Z¬ËM&¼ ±$k€~.ç5ǵ=|}Mã…<—¯¹t¼zÅê¼–q7=Go|â™ï×X¿ÙÞ1«£þ¢kÕJÛÖÝ]®ñÞÒݱÚÛýõñôðÍô7.n3N#vøÛå=S¯¿ÜV“ AûI×´&ß\´7Ê ©ÚßϱB¸`t¶L²)hœ˜yF–Æ–¼=pÀ rÖÇ&4Ÿ¯œEFªj<ÅcÚ{˜©¯¦¥Ó±‹­¹3¿Åϳ0‘?@v<á»t7W¸½¨|„ žÔIáŠàéc¨^¼oæOuý‰j4W5ëcwå÷•ÝEEØ,x?ä2ø$öB%Çñ¹‘µsµeBFµÇgO„‰Ä×ò¾Ã•à6|½¦=Ût„,Ä¡IîP’cS%Ó—ÝÃx |‘Ê©öÏ=_«½.E|~Þ½ðxòƪæéîS^àEÆÖ޲}MºJçˆô±ÛñûƒÞ{ Ê «öƒ”£ÝïkºTë¿§3ÿõÅ Òzendstream endobj 259 0 obj << /Filter /FlateDecode /Length 162 >> stream xœ]O1ƒ0 Üó ÿ @Õªb¡ UÕöÁqPœ(„¡¿/ СÃY:ß|–mwëØFàðEŒehvK@‚FË¢¬@[Œ;Ë'å…l{åßO°Èlü®&’ÏSuÍ«r ¡Ó4{…$ê¢hjcA¬ÿ¤=0˜ÝyÁ&£:ãæ?”M%Ž›€KÄ17ÍMRËô{Æ;ŸR°B|oÚSýendstream endobj 260 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 3593 >> stream xœWy\SW~1ä½'*ñˆ&”"»€+‚ DÙ”µ" P *¢ˆG¤.õ¢¶¶¨U*ˆ j«â‚0 bQ, ŒŠ,„@ÊT«vÎËÜ037ÐÎòïü~ùåwo’wî¹ßù¾ó(1”@ ÐõIJÛœ”™š¯ÝYó¦~ê~ª0§ÿ}¶:[ÄOÓ£žéé”NßlÈS¡p<ÄM „Á\Vá¡+fÚØØz®[Ÿ‘šœ’iæä8ËÅlu¶Ùïߘy%mLMN7³$‹ÍIiëÖË“Ò3Så«7m4[Ÿ¾ñ?þ¡(Ššî±Îs½—,c£wæ&߬ø-«³“Ö'§¤†­M“¯ˆ³wpœåä¼Íeö³¹óæ/°²›iCQA”LM§B¨”%µœ £¬¨p*‚²¦VRT$åIEQ^”=%£¨¥”75‹ò¡|)?j6åOPÔB•-ŻР[ÉŒÅ÷ZçÉ)¦ k€ÙÃ>gÄŽûró òLÃÑêµÑn¬8øcÀï­Ó~ØÑ/äýxnï;ï±>ÁÌ`¡¿õüŠD˜ —vnù1·t3Zcõ©G\Ò‘›%ÛŽî<º«œu¦`ý–˜FPšÜÛü¶ó“매®Ç½¥Ÿ@å&Õ×ϵ<¼”¶Ob >ž©€= 8£T’‚Fò‡ï̇Œ,ÙØa“%>gn¬’Dצ÷£A¼]°aw¨]±ä½¬°Îs슜gfhÌY !lÌâL˜†!ò—Îò®G’RkQ#Aã|æcÞºEð¼G ³aú<<ÝËû/3:üRzó‰ïö}Y&icþ´gûÞ­ˆMÎ+,—îc´Lx©Ò°¹ô,n7oá7ñ%óp #®2[.sñ¿z_Â3ó4VŒKóò_%b¯T}¦º•õ`ȹ¨X-+¼y)„µœ³–¿¤µ Іð·ŒÀž/ÙÑÛé¼­hý±ÆÖŒ,h°Õ”ˆ^Òoy‡ß4¢‘ç–z®\œ(…³ XâN5ª©€ÿÕ¿§+Ô“¤“Iuñ -žB`t¢A¦)];/wMvn 2ÉEÚ—}€]Ìí(Ú}Ag¿9qìôñ£'ŽU@®Ú`ò¿©žð~-¼åvßÚþCÖÙäjÿ³„üæŽX{`·A30‡‰]í0©P:—λ*Þ ±aí0Œë;n«\íQH„úd´ôíJþ RȧA<" (ÕrÌ18ex?··–Ý×Âj¼ß‹Øç¾vÒ5ð$¬éc’ñ)•‹‚¾xü»+÷£½Ç%­Ì–Û ˆþÜc?Y,u’y·ibzø˜þã*† |£ü¬‚!!é˜MdÓwPUÑ•‹•Wß@Y˜¶ O—àúa¹ŠæMuàˆ`ê¢]]#¢´„ÓÛ`¢¶Â¤³_쇋Æâ©¼Ic#º[ækŠÅÎX/t.YV!½[“Q²Êìz›%®°ðZSv7_2‡þÚº#hÔ~:u¡âjõÑfãYΧᜲf 6Áâ•K\fG¶“jžöIÿppR^A[1paÛ÷{ˆ}kÍ»ìt~N±ädvaZÃŽš¸ÒÿöŒ%q™Ë¢$pŽ4”ŒÕ½ù`ø w!‡ ¬Iúó4%=´ø­Z®3×¼—”H¥…¾%ñ7µ)ù Më|ø|ØTEÿ!\(üHªð‘–\Ú VÑϺA†ËEïhü)¼&Öè+2£qΫåNK¯lÉs‡åä-ÅÿkY¯}‹/þ1‘(ñç·ïµ4]Lò‘àaíÚí½K‰¾Ú-?y^;S¿98J’ÖÒ±bGŠÙð‰?û„1ø»×è™ê#PËE4ž¢À øÑÇ4žC–ÓÈÒ‚† a¹èWæò¯a¡æ5Q*Dª#9l§™ j¥AêÀב%và'ˆ´,|/è/„åï9 ‹÷ZsÈ*VG’N–Mn†ó†óD·iÈSçýÞu_t ®Bù ÊÔÉœ/ ’¯‰Š]–iƒðï鯨‚µ‚Tx˜mzó@²¦&ø¬bÝt^ݰÆ+ñêX›™ö1o ¢o ½¥€+…p˜?Ìi+ÕÉ!Lfåà©Ùì2­çôñ Dzó…üþ~N3–þæî÷—ûëÞ4Lþåö­'hÝÝØ%þòŠï|Ð,äC掌¤íq_Þ_ýÕÙÃ'K+ÿ|¦±ÝwCÜB>ôO–:„c«9«¼ó±“ Ÿ6Ú†ye‡!/\:h,n… ÐÇ¢Ëé×·â\ícªÛ)%NR˜T´¶\Öš2DÜÜò¥¦½H©™sN*~þàÌ5§À„íØ[,rÅv{$Jº òÀ©C'НUž¼ØÎ1n+³Ó·J7æ­Ý¸—¸Æ3(i¹øPÀGó\ADΦ5(å|™um¼õЖ/³‘JŠÜeËÖÑûJ÷(,Ë­,QZu5·…Iü&NÑ´ÔÞ)(ÀeNÄÝ×?7>PŒXÍ{h}/€1D蟎 /ÑÕ{ŸVíŸ a;üÇØØû!< a§ûØï ™ oÓÎ` ¸ p®7%˜P7²MÀ_ƒbNS ã þžÏ„*œÂ½ãûà‡`'ðûŒÙ0Z‰-ªñBDþ¡`ç(ìæŒGåCP› Û„üy'@‚ 'à·6 [Ú ³”/9 ¡‡×~K?ÓUëܯ§×S¨§OQÿüç–endstream endobj 261 0 obj << /Filter /FlateDecode /Length 4813 >> stream xœÅ\m·‘þ¾¿bàÃÓæû‹‘3p1'/ˆ%÷ÁwŒvWkùVeWk[Á÷×ï)’M{ºgv#F`¸Ì­.뽊œü}#F¹ô¿òïË7Ÿ?Ó~ss!67¿é¯›ò¯Ë7›?¼†5X£ˆróâÕEþTn” £vrã­£¶›o.¾þ°Ý‰QÅ‚®·b´Ñªè†÷¥^¾&dC´Ã[¬êà¥sÃ~»ÓZ1Øá¬ªàŒ Ã]„ÿzñ— «F-‚Ú`Í ›WÃÿm_üx‘ö7Ò€Mö·ï‡Ð2Ò)±ÙŽ8‹Þ«á³¼›føã6èQ 屸à…qH§q°už¿K,¥€ H íÂp¹UŽŽ†2yåðí5}¦cÐÖ%d…Rn¸'1§é¿›¶ ó~þÌ®e䈣@tîHçþúÅÅwÐ`€î¬TvTzc >švvsw½ùÍÛ“ª–2lâhÎÌu-c£Ñ¯”Wtýìpx¸¹Û¿Ù^mÞîß_ß½>ÜmÞÝm•/÷/_ß¾~ÿúú~óåF^ï„ÙÒI›DŒ ›Ÿ±Á_ðÏàé› ©…!v]ôfôvó¦®x!õ¨ÍæöâùT(Ga8VYé°¢†UJŽUV8–RÖŒÁ2¬i¥Ã²^Ž^q¬²Òa…Fç8VYáXéœRA“vc­P£òÄl¾TWÒ –“ÒŒÎ0¬i¥Ã‚ÙÁî9VYé°¬ƒciŽUV:,¯ü(Ç*+¼uÆWYÉç …¥`0} ‚C”!"ËÈ‘¾:¼y·ÕfŒö0ndf>Z²ëìs¡)fG^Y"Ça~zªh{(y˜íB(U0T„dJç‘>2ýŸu¢ÞÙú\æÖ“ÿ)üÇäþ“ßÿS©_#åØÉ©Ô¯~¥Ô†ûïVyï'4ëAn ¸²pýGš Ž%ŒD " 74ïqB„(Ëô)†ëíÎZmºA}±ÝxX@ê|‘¦l™ ›MèF˜á]wyíÃpÛ†X‡÷ƒ¶yC­ˆ=ÍP&SÑx›æs"Z)µí:3;lwèÑ{ÑøËZt…Â¯ÚØîÐOè>f]'JúRå‰Õ÷ÃïðaŒ6„˜hð¶× ü¥oøzáwÛ¥qNluÝûð€F/‡;éÑ ¶â†K6ªËqJ«áž-¶Ý9”JRebâ3¾¬G…Ž2+&¨hw` @ÍB÷y"è¢éÆ“iêUÄÈÅþr ] xÆááXWX–:üC#–“¿K‚ è@ЍxhàûþçÐà7 ” c»(lƒI 9íYä¤îÄG2+ঞ1X5üœíZB˜™vÙ¡!µÙ˜5¡Lu³Œ½d*“™CgÁ+r–ÃM²L£F Øg³Ê$ujjvsx5 j}Q8ì)΋0t$碭, ¸ápŸæÂØKó¹G1;`ódò—4y4/®»0gÝ'ÒÒí«¡xv²¶:§&“&6 »aŸ Ì ÏGÒŒn^žªÆW\nÄ™ïpޱ’ …hDIèöâÛ‹Ÿ~Ÿht I6ÛûïLö÷…$ø3évëi³¼¸M>€¦"fõÐf %FÎävÓEAØd×ý1’6¬‹Ÿ‹ó¿%6?5—ÄuÓÆUÞà3«Hì3Ü´8쯒}“¾Vû6û°°œ2œÅ  òÁBBäØÖ9Ýí—ÂD€f· ›ì6a²RËOÄY¸Jb“t÷‘Tƒ$ /e%ï]‰w Wù‰% ¿ãÀgPé7ý‘+x™B‡¦À,k¨(²Dpö<2-Ü´0Yö†Å÷öšiȪóÄÇ!ÝÓ DiÃÑSd™¸è# K Süˆ% -UÌ… mJ*‡ã«ÎpkÈèl1»”³)·ì(p ï]¡Aq’eæ=d=‘È=ž»þtͼo†Áƒs²B!IЍéªõÑäýÝg#6.ø0“\ñˆä™BYÝ9é‡|é†TÛWh,NýŠUCmn—-1²33_;0ì¥Ô©©Ò“éÝ¥´N§Ê#¯øºïX’d*=ÖÒ!cuœ6J)\Ãë ·Ê¡à`A¤VéFÒOÑ%M1ÿ˪,Ž ×ÂùŸB"šy…––…í¢'‘£¬K@8*dn;ëV0;ÔO P´3y0+–ö5ñwq$m·~™ør2ôQ‹]+¿ÍÈ(ÚÖX϶A>#yd%6Ì׋°iÇ®ËâvÃþ]3×wŒÄm“ʇm­q˜¹òýn’!í¨x3Æ—¸U6 {v)$¨¢â âҴᘳ®]MñèŲ³òÈëçjp†×¯ûFå}ÝyÑ,ºÚh^ïg.“íÛÌêÛ–W®™ð<œ]œ*[np¹D7вËHdX1âôžJm4’ z-®1ŸyKÈ©†øiY2=éØGMš’éæ[L^SÊÃv©óeà¾w üÀ’Òùe-|ËÕñPN‡*§S=÷Z^pûg½*2Šç¶N.‹¾L4èC/ýÇ$)"¡C‚.¦-ÖäGm?‚^N?]-štÝÃûGé}¯ûy+~Ý@ÖÁó}EµH *VÍ~±ÝYô4@J‘gæö}Ó-Q·^~Ú”h Åë–‡eªezŠe;VH¯Ó@ ºÈ.µ©‡Ï¥ŽQQæžš×I›]HéÆ]¦ÊE˜Ž×¦×-@ô&hœh¥¹lµÈþ¡µàùH©Ã¬üÙõ\R¥ò!#SXç¹NŠaÖoMt—Ïë§“ý_•héW‹Gó™X®¾K]}GY)æÂ3Rz€Ôs·Gã:†kÉ¡ŸŸåi€çö Z‰Ò,“ØO"®î1ëÿ2_ëuÖ~VgzM亪¢4@0œú;m§/¥á-ÒJÕu•‘)Çð&ñ0‘öëåÚ#BˆXHPžªo°F0}bN”S%ü¶ùC=Œã¥Âa>ÂKÙçÔÄ”vMð3O⌠ӡw^|Fv¡ œ>B-…‹ÕÒaªžw!*+C¨Y.ÈÖn(Â²ÙÆÄÜ9£$‘'éË4j©~ÙuÍ #à$}ºÉbŽÝü¤yf§ä¼Žú%I$Â(û˃zIqàÝ|VŒUv ª9S6ë|ŠÁCT~꫃˜OIñ¥AÞ7ý»àD¯¶ Óh£ò×W„K©`yZÕÍøB98uä³QœJO׊ˆë²_Ðë:{¿@Ï® ‰Žînýlù ö‹oõFKêÍWIâÁ%5RÒ¢;X«ŠZǘz*#š.‹}Ï ‘ÚöüJçÕFB«Ýyõ±‡›O&éô(´œ±¾ÍVa¯³I ©ÖÅ>üåê! ¶™òW:ëææûdÐù–æ¨NÍ{(†î¤[OŒe¿µ™êÊvÛvO¡@•$Ñç»åyg¹­ ¦\­O…Ó°¿µ$acóœ£KR¨Ó€µ Én´Xp×'5DÖQC?]ø™˜‹›2Z<¾ð+âKİèK¶ Φž'„.úb03oWzòœ›ëa¥ˆ\‰PW™#Ÿ­6»…ŸR¬ôÒÒ›©ß®ÝDô¬UàÊ<´¯KòáUWMÝCcÙÅS¥2aá»Ú–U¿—•µî$ì'@ü@bŠ™}°FKk­ÝÔû¦1?“ùž]ß67*ÌÑíäSëø,Ã/»öü!W¹ÉSÕùKWErÂthåçIŠEaÁ©s_òçÝÄ^_õæˆæŒ‰åíC¯È®€n—[ŽôR°wìdÍ$Z+Ò¯×VAW­Óé~2ÖÐ~Ö^A#}ÉR–ÖÝ„åÕãߣ¥Š"}f£^›‰Sj¥;Áò°+¹½•à0?§ÈûÄõfi´Þ’Ù8aU×­]¢g¹‹ðQw[6åÎ5_gbºÏüÓÛ‘ÎCÚƒ)ž©g~Q4ð2»Š£·5ILú±ÓÐ!§uaf>*.#m꟯F³Z|æ(™«?qÊ·dO_ÂÁä(¼ïæôOs¯ þkåòMª¶õ&µ<H¯EŽÞ!Ô;r©Ó£†÷â £ëyy™‘G?Hr..T¥éUX_wƒ ]X&,'͵¼ÊFwé8&G‚®U¡#x!OŒEð5ÀÌV¶~ŒÇ×J3ÝäÈ|zþÖd6 £Ÿœˬ%X|ª°V0§dÔ~ªÚGI©¶7þœQý·ònº'÷2S÷¶~€nmV˜-ÿzÔoœîL,{†¶¡´-tõ.ÔBÄRø#ã<º›eRáByh‰ìüŒYkøù3 tM ÍD}Pò¬E‹/Ѽ)ôîbz‡XBDÇ>íG¡ë†ß·Õ][ý¸7! f½“<ꌪ²Où¾{˜ÓƒûEA|8ÅÚ¯&Ë ~Öøs;ò dw/9Bù, T=f¢<4ðªìÅ0{GÌ(|Öè>ØgÇê½ùº¡Ý/ª…)kÙbîuSáóvÌ÷‹‚›H¾n«w‹à¯ÖÏþÑÀŸÂ~Q›\C;) ½mþ¶¸ŸE|ÙÀÿmà?W‹:w½ §Þ¡PMûsß6°ä’7]."¼[T (']075‹bè8îM¿¥â1S¨hè¨h›.±eº¾‚Ó­Fðûªsœ17Ý-nÌè~ÚV;p!ÿ²Ý9Òy⢤٪\”¿o j’‹ç±‹ ;eäªX’´_¤u¥[I“x™ßpwú¥ÿÍÅ×9ƒ¦ŸŒÅ‰‰´,ΩȟSQ8§"qNE‚ˤ bYRáœ_8¾Åo®­]SKN~c'¯yÞþúº7ó 9OG¬¢[°cAòÐÀ«ÅÐÉ2Þsw†f>¡jSÉîRHŽô“ÁòÿŸó ]+ÍÓ·fsêOÛXŠÙy?JÚÿyp Ù³àe‹ \Þ”?öY?’oÏíq“:I3ŸC>æÅwRìiÈ#ô†ÊŒŠE§ŽFç,¿F÷üžt÷u~¥{´´AýcžÊœá5©*ô¹¶Úð'Ù½ð ÀÅ’•‡—aîÐ.[¿»øðÅéendstream endobj 262 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 1662 >> stream xœTkTSW>7ä>Ä’ôV47íTªTP¦¢.ª¥.Ôò¨¶b‡$DIbIx ]kUÖ)Z‚€ÄGAy´Aky‰:â”* k,Îjµ3Ó±ÕÎàªKt_æÐµæÛÕù=?îZçî³Ïwöþ¾o)ˆa˜™úÌì¼L‡9ݰ(Ú–á …ÈAŒ§”!ÊõJ³ò›È:ØSl;{½Íþ28€…ûÀ2m …JÐúÀyù3‘, & ‰!º ‚å°ì ˆ†5ó“(©‚°âÍ^ݳÁbV­Ü8òðaßȨDì0 ¾‡iV·ê…*í,õYxSî‰ö§riãZ×uÛ†±Ïü ü!r4wÔÔ+©7ö™ô­ñA±x“Õ¼NPß„é6ñzWìoçoŠY:üヮáë>»å߀Dѽȃx²ƒÂzxõú[·j>,ßS¯çø{ð.¼'²£uØ#Þ_îw€ÆAÁtÓÉ´IyŸ¬÷·d„ÌÙû’Žø˜ª²ëó¤ú‚:ç¹"X±av[©«ç [3ò㶪 5ÅU»ªv5Ñln?ñýÇZˆÄ_â?ÕzZ<-Ç{q?¾–uvõÓ7;ÃUvN¶¸{†NÙß¬ÐøÃßàÑ]æö~Ä(˜G´dž.t„ü{é¬áû–cïhÐŒñ%{ß-/Å‚ÑYÕ-A×_ŸÔΆ»å`U'hõë`.åU>R'Úí¬Åôó Xu‘“à• $À© ®òâ±Ã#X8ßXh*t;¥‚2Œu¥Ie³µ…’qˆr%þþåóëÕ¨7bÛûeÕ;XÈ©«IXÄåé›W§Z?òœ>ê¾äÒtì¨ü`_íÀ)¯L[ä–èÐnÕÀ/*§LXÄŸ, ½À«{^oíÊq fÂÒ¿äÞ0“Œýq­q8§æ˜Spñ0ôŠ#=1/Ìߤ£=¼ß=|Mòvúž» 33òÈ+\ø¸ŸßÙûÎǦ±ˆÂP¥~·(IyåÞ³°fŒ ?n–ˆ†+JËÈÙˆÓñ¶ºü“ï¸ß;¾·OØwWwitñ§Îâ§2`. üÙû ÕCƒøŠ¬'Ë8âšÔ±d:tpZ{ø žlA$ $ÀEÖßûT]æf> stream xœ31Ó³´P0P0´T06P0²P05WH1ä2Ð30Š!Œ‘œËU¨`aS027iÆ`9'O.}O_…’¢ÒT.ýp .} *.}§gC.}—h ™±\únúÎÎn@n€¾·‚.PÆ9?§47¯XÁÂÎŽËÓEAmGTaý?ß ÿ p«ƒ¡ËÕS! B)Éendstream endobj 264 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1ƒ0 Üó ÿ @‡v@,tahUµý@p”' aèïKBèÐá,ïN>Ë~¸l#ÈGpø¢Ʋ´¸5 ÁH“eQ7 -ÆÂòÄYy!û›òï'Ø dv~W3É穾äU½‡ÐiZ¼B Š'mUu­1 ÖR Œ¦8›sWÐ`öJЦÇMÀ5☛æ&©€eú=ãO)Ø ¾@®Sjendstream endobj 265 0 obj << /Filter /FlateDecode /Length 5319 >> stream xœÅ\kÇuý¾üFõÈšv½F 6$ƒÄ"‘ b`´»\SárWä’Z {έª®ºÕÓ=³KQ ˆÅš[¯û8÷QÕü~#F¹ô_ùóüúì·_k¿¹z{&6WgߟÉôë¦üq~½ùÃ3P꣈róìÅY*7An¼õcÔvóìúlbûì;[Ó?ºè1àÙÅÙóáË­•Fºá5¥Ð. —Ûu ÚºáÝ2FéÃðQÄ‚^…!Zì´Vø‹þ}Ôˆa3ˆÑÆè½Þ$R!LÞ¦^Æ×Ån^ÓztÂjû_Ïþõì‹ggÁÙNmB¡Þƒ3ˆ¸QÆâO»ys¹ùÍë£L’Rnâhvɹ$½ð/ãh”!V=~ĤU£uzؤ=Ðr6¸ÍNŽBy¯‰]Ã߉£T¼á?ÜgVKåù’Mg]æõð›BÓoË_ÑØBsßhÜ„õDbFïŒpü©W‰Paâ“¥0*l~À ±íÍw`ʟΤqz„"@ÔnÄô×µÇ iGa6¯ÎžžIïÔè§*=œJ '±KF5õtT}*rªÒéÒÞÖç•r”iU¥ [{ÒÞ&*o >œ(wäÙŽ‹¿ñp’~ÔšÅYL!Å,ý?Þ\ßn5ö(¡ïã&+¡Ø(è²µ1ÊŒF6E¥#)‹ÍÂþÚ‘”r4݉*Ñt¢‰¦ˆÔ¾\{_¥0¿±‹[9iþ‚Øa㊋>o’wÁ’LâÜè÷êo.àO½ £½ÝHi°UÕÎ^ô µN™kFàŒ3V>l°œ¶ô³=€è)™Ÿ0¿ΔŒb4žqÍ$ál“Ú¨6Í4 6:Õñÿ˜²bP§¤2—^aöDˆ —ù¡Š#Lžþ¿ysu|Øu¸>Á×:1Ái# £„PÍ#.„‘¨äN¾ž¼¾»|s~y{÷Ͷ¢ÏÇáéêÎ ¹ š¸áÀÔÓ…Ôž~chQz&𙌰ÍW¬OW-µN7™j7ÛcÁGÁ„à>76„QdOƒ=>\å¬÷°ùáˆ1MÀ Cy;šb˜/~$q: "S€ücˆq’ði[:@_àÛà‚ƒ…nÏã$G ‰Èe š+ƒ Ù‰“â°Š+–ñ’ÔkÂGÑïuÁÌA#³»ÙÁ(„ާ€cmУÀcu’ÔEläB3È“o†û¿ªŽG7Îd Át)9~”ŽÖjƒ¸€S•žŸnX…}¹ð±Âj^N;f°Vb%ï»èŽQÕžF5Årªõ0ªË1ªÚÓ¨Ï i Ü¿†I˜Qj•$òèj°hÚ›”Ïì´ Žº¥m ©ˆ^%‚ß©ñƒ¬Ý²t«1Oý Íš&8エTU•ŠIâq§0Á }Â14~ 郚cøaö ËUcfÑT6qý`Ð=Ô¤-aÆè"à°£·0”ÑJ¹ºGè TÓÃ_¬Šq”qƒßŒW‘+¨{šˆÃ®vAQÜÂÌÇKON.5¨{”Š[¢ y¦Ö¥2Þ‰Û'cÜ£·Ë|ÒGOPKÑ£u3Æ=µÛ¿<þV͸¤ŽÙüò™P@Žt‰%ba¸zG·\NˆåðýZ„ ÓMšòt—&ý»íÎHh›ÐÃ3¢.³üÚ宬ƒ=ÃÅ–jF^ûz·Ã š;(‡ý;60_Þ!§uÃæ Õ ~¸Q µÂ7wy˜s™ÏCBÃ͛܀£ü*ð“´%éƒÁ"˜M£ßEˆÑY_®ýø„:ìßWãXôÏ­ù?­ùÓboA]Çè‚üeÎ$¯'­ùº5;6.Áéù"Áí¢ °•*‰]ë‹êÁ|ŠY¤µ‹‘*¥ž.' mkmÆÖt‹ B“Ï®õ./Á&3+«ØdXljÞ›ÊÀd3QM6y¿Ý9“9GO}kjnˆµéù/Nm—Ò/ö²¦êxŠ„#í7ž’›×òyëÎþiq ægw‹rM®ŸµÞ®Iœ&ö2 á†sßš]d*©½v¾wM×Y¯YœÂ/öªSr‹è´z‰×¢Éå¤ É•æ‰éÅí(ޜؾküíšÓïO[çËÖ¼šÃÙÜm¼à’¨°ÄàŒE•‹ Ç<Óïª9Z2FúÓôz1t›V%–>-I;'kGÄ.RN1Ýgm…®™ýó|z?Jƒ#ÓcÚ"N)ÓŸTF֔ˋjDñä‘è‘åÒ‘–ÏAlÒ8‡|È9Š:.®D• pt¡nóØpy”z$ûâñ|?b¬öçì·ê&ÿ%÷»ÆT5"ì¬òþ4ÛÆ ¹’áyQ2CÌ™vÓ':;eà³Ù©ˆŒG–šl_ò¡¢ ´?Õv´Ñâ?¨â3塯Úûë;šAŽtÆ#õz!eN?J¾“ƒnuä¡8O¥(gr —(ô§)œtó’u{a¤cD¡ƒ}@Q©e5yGÁxž_ª)E;ÿŸjJìöòƒËJYBU êG·O,›Z/MÑt”z¦ŠÓÔI,1ÌSÏn>)¡C¡O`)—Ôbܧ¹£@6W^ü§ Ó•#xëþMˆ45$à%}åÄÈ=Ž ,S¶çœÞZJA]ÉÑ¢ÊùÙ_>Ž–®j­4IkµfÅ*šd7¤¶ \‡œȱb=›„DG´J{‹E£\XÈfùºÙûU^:æ¥dÂ7<¯þ$‹•ᚪ œ‘—yjÝtÁ+Õl—?Ú@n麤&âT–|[f¤ì_·f’<•(ÿBd"\µÊæ{¥JÏN‘h1]¿Ý"Ï¿XaSý6Ñ—ß¹j”g ô½‰1#Ý<î:@‡ôFT%ÛN÷š=Xyül€WÁ§{Ná‰0Y¨Û“TøªevbmèÍd?ñ Ôß&´v2¨IèFšÎ¢/»6¯ßLÜmVÅ1|_§r× îÞä~àÁÊÈ™vô$ŒÜ×OX‰©ÔªbTÓ×B*£#¸GN‹;°âž‚\½-yUJ×»8N¬)|x á=¶ü•§îíŽð«”*ªÝ-^ÃÓnnxrÂ>n^äÓ9Håˆó•ù>é—¬…ֵߖõf{›{ ì^N{Óý×d‰A@Š°Î R$-€ÕÙžo—Ž»ÌªA¶àU‹(Ò6…ë¸s“$DhÍ÷ xÞ5‚=W †<û\ÿµÂ€C½Ãž&ä!Î\R}˜ÕŠ‹ñ¤C÷~íæE¹yÔZ.ßúŒéóšî³âóº†.¹ðî :SQöÍàþvúŒÏ.ùZ¯ì,ʤ-•ÛÌY½`šž; ›¾Ò»SrôNQùÛTÂç]ÿ `ÓÁ…ÿxŒ ˜¿Ûw£/ž.\¤ ×Áe)H`ãº8£¥µß(i=fاX&jYîIÒ6\©ª§Ø«â]ai9`™î&ˆî~šù­J—ÇMv±âFJ&HÝNè™×švtôREÁuëƒK•¤µK•‹| MŠ˜K ‘à)˜uõ¯óe´%WáaÚ1¿Ø™”N„µhž)£`5åx&˜j†B΀¢¿Õ™ÎÐÃCIÖ¼"ÔÚ1E¡y¢ûîÞ‡¥³/tå]XØSÑøËm-jü¹5ÿ³5¿jͯ[óª5Ùe)».a%0V»oM~…ºr éIõüf·^¯iI1@ïM;aƒÀHÛ¥ÀÍ7³¤’9en,RãÊ»Æßå<ñy27I• f‘ì‘*‰G$®èxõ.™¡WÚ¯UføuiJ¶…÷nɽ¡ØÕYy¹¼|Uv±h5¡¿h6–¸r×£-Öë&ÕtkB¼M\Ð!ðÀê.[x£DÝ^Äc1/éXˆ,Úæ’¿¦^O í·KvPV WP)ŠÕô‹^µÎlKˆzÔ”ò'4Y/Z°Py‘-iÁÐ'kûsšñƒÉ‘S¦¸âþâæ]åâm#I˜íð0õž9”iíHŒ™wâM‘†ˆ)¸üï¥à²«ß>O·K“±%;‚ÃD87w[˜M5é4›@f+ósª˜Ôm˜baz…—ê6á”M"GŒ‘u¼kºËƒ×<8¹v!<&C˜š mÚÚJpÉêJ–Çu)îvÖpPÕçhEŽ8))?œ*uà†©±¥’kIì«<±‡sZ‹¦jñ‹ñâ> âÐïP?X~lŸ"ENöq_ï R §°__%õ/‹_½lªœ-@FÛEš¯'q>âQ›—í¾ˆ=ƒc/験9öôí]k²xûE68¿l!VzǕ͚ÊE*Bpqr>)®íT·{^+…ÑPxÒ=³`ÏZ9ƒ_dpSY¼¤!XW®jËâ¿ðby™eϺ§,ÀKþlè¶UU>žVka¡ÕEs³ð?'œ‡p›>¤$ ‚,·½ýÜ-;iöyº<Á›yM{û\–2Xèúô"âËmÐ$@[Þâ;×®f.óïI­¥wõW– ‚·hÃçxe%L?ÊJÏM]ºT9 *tˆ¸zÃTŽÑ§j¯ZôÉ»óÊÒªyÝ»y×ε>96£®‡–22e'#ë¥Iç×A'WF½'5Ä»û©•Û˜&Lþ’1O¯¦ÒS‹dˆ”úˆ· )ú4DúÌpG…»A¾8·Ät5âÜï,ò¾4²Ðc „¦m-§ÁDiª}³0©¤Ä·ÙÀ‘eöåÕÕâÅòÝŽ$±fu³šË>yH*}‡µ{—8ˆ´û‡‡ÜŒŠrÉü¬žþ]‰ØÁÒž!á +z£*iÝz»îßÅ ˆÀe‚sM ø çEY2œÓ%ðeÞݼÌââi¶ߺ’™Æ€W-wã<ú?Ê ék‚Œ#Á@>ÓRíN©¢ L}P ZÎs¨ á5Eu©QV’{Öü˜/XÖrÚÅ¥œÉˆ>±oXÚ,%SÏþ‘´ÜæÕ¦³d³û—9$î&&pÿÿ÷é}½Î¯Íß´WGO¶‹%V‰Y}Y¶X“!¯¡˜V!äháG*Õ^¨8Ù?'^ µûkÑZ®ä¯&îò"°ÅƒÚPSœ@MÒöï[Â-mŸC$/ª©_¹ð*ˆînô> stream xœ]O1ƒ0 Üó ÿ @UuA,tahUµý@p”' aèïKBèÐá,ïN>Ë~¸l#ÈGpø¢Ʋ´¸5 ÁH“eQ7 -ÆÂòÄYy!û›òï'Ø dv~W3Éç©:çU½‡ÐiZ¼B Š'mUu­1 ÖR Œ¦8›KWÐ`öJЦÇMÀ5☛æ&©€eú=ãO)Ø ¾>5Sfendstream endobj 267 0 obj << /Filter /FlateDecode /Length 4700 >> stream xœå<Ûn\ÉqïôÂß0ðÓÐÙ9éûE‰xo†­a t``‡)Žd‘Z”V+`±ùuWU÷9]}¦G”dÇA°Ûlõ©®®{UWÏ_WjÐ+…ÿÕÿŸ_ŸüëW»»µÚüõDÓ¿®êÿίW_œÁ ‡3CVY¯Î.Oʧz•ô*ú8dëWg×'kmNÏþ‹½ë§8„ჳ‹“ïÖßœªÁ(ítX_áP+ÒúÙéF 6'ëÃú˜Ö9ë˜Öß⊜S ë¸"æ”=,ØXkà¸þ¯ÓdøÌ5øœc4ëW´T)—Öw4«’ï§Íö7ÀAŵ9ýï³ÿDœÇy£mŒ_m¬Ró @€ ”ÊðÕ4ÚP·anÃÈl| vPºó)åµjá Mú6Œmèó׃s!¯6&*§‚êW´,8—‰.u¸kÃW ©gâðP5€© ë}›½l³ß´Ù£À.Ú»nÃG ®k³vœMë³vî}¾nÃm^µÏþ¼nÓ/D2?ã³Ów¿k³o8`iøçÓ6þ·ÓV."ƒgåø‰¸óˆå…ˆDwP”mï²Y=>9ûåwubò›6¼âÉÖ_‰Üú¾ _´á¶ oÚð\„ì &lC·Ž˜ûï'¢ÈÞµ!Ã…ÉÓÃÌÇ-l›UmÈ´–É[:ݸ¨ýú×mê·møe>jBà 1«Eø£=ë ¶ÉC&´>Ci”2`AŸÃ·6‘¶ÙÂ_E¥­ a4ÌÙEOÌ:xßYã LòÄ §£˽½*Àbä:ÍÁ}ßì<3ÎÛ\R-9’> ²æÈÝá §”Ï®lY†»Ý3öÅò[å™ ­TTèŠÉd³ÏÙ–Û×¸Ä )905&€wñpój”±§*T2Ú¸zÆ`|æ´aHÜž¢…ö@¥ÎSU2ß1¨/ˆ.ÆW<àëF#vŠÛ¶ïœgŒÓêñÏ;A¼ádNÃyØqÙu_ô¯I¢¬†õ1^üyÛvbLâD@2¢ü9 ¾y; Aa$~‹óà¨\\O‡+–¼LʹP”cX j0ñº,Í©OÁ:Çàr¢Ü6häCÇÃm&[Šg¬¢]ªH Ùž^À\@C׳' ß„‰Ê9›.² #G >}HÙÌØî")©×(dF׉vXœ·8JJǺE Þ€ÏôïË,,ÿŽà£!Ú_‘`‹ÓšF¼-t0`ÿ¶[ 3øô ‹Ø/¼«|¢!Ûø ;1ÒÇ$gL!0Êl“aâ$Qlr¶É?SÀ§í@o0†[‹Ê†çð`w4ôÓ@lR~cæª{3ÒÒ0 Q)4 á[—Y´6;£Êò [Îæl0ÝñA¯’‡ä¿ÚtÒ¾7³&“üD®"ü¤»"˜äÎ|&kÃdR#¤@Q£”îS#2³!¢…à*3£®pÚUª‘ ﹈¡Q è!,q¯G²øjé¬' Yƒ;kdCH.€àˆæZH,ˆ@œÈÞ#öAë¹åÆ3£9Q)åIÚDÇ·»h .Ùâý«2u·z{Ž'õ ,Ìóöá{É[Õ…׆q1Ä´($dÕg—ÀÁ„™Ã£ÈzG*­F”‹›6ÉšÎÎ1zÝ4õÜ}r8ê“LD® LøI ¼É¡õÏýì”ëVO4÷3wÈ„W‰‡W! Ñj;¢ú¤m#cZÊ=®~ZXî»þ»ƒ°ÍÁ7½>mÉ**ç5";#e‚u•”¤f;1½…䢒[$û¥ê²Xøµ¸à•tz ŸUÓ9™–£ÇÑÒôC&¨C5žÄš€Ät.‰é_ÕT;/rÕäi;îa–‚«_żÝöU‚Ê1s˜ôBYñ·½\ï‹·Šà®(ÎlÌŠø-˜»îhw²"_ÔÕ)ß/¯ÿICögŠ,¬Àô<*u5 !ƒi1bvÆÁù©ÂI7Ê1¤m>âò¯ÏNþpRjN.ƒ§Ž«W-49@8z³×<˜`°ÚôÝú÷ÜW5žiŠÝÇ’ ˆÉsµèåd[!æÝÞ4{5Qد¿`Sù§|>pŠh:u[®!÷RX+«¤y(9´‚€]§•OSƒÝ!rP¢*˜lç‘fýG¢Ç(,b"†þÄ€”kß»b¾Ø¡<%(·v)ë4 ÌM/RQ9tqÛÖh0¸À€Ë²OP‰³®óøû.ÉclZ`ðëSª‹ÁÚwe£Fç-scEkLPaI™;'E0P7ÏÙ~‡žµ‡¶àXï*Ysçk.d¡=6ó¼Ÿ5;¹}ÚÈÂ5£X6ðÈ­¨Ñ«qàá§èõ]A ËÂà²]šÖj°­5±ªª{òÇ ñ‹gºª,¡" ŸžØ·å¶»Êž«ªº{_TËÄÏ«t(먔à,Ê–£xVëH•g‹4³Üˆ‡ÌÖæ(ÞÚ>µ»¨»¨‚.ÂvjfŽ'Uâ.û i&Dii)'¼(à$ú]˜€û%½XÃxVV`}€ÓïM³?g*U$*R0y*!tp`öºbÎSH^R¢œ_éZ *Q!±…|‹Ò<Øfsêâ|TºW1áx² §Ï.;á N„³6j º“9)zWXƒ²{1¥ÊµN¤ôLw«„Ø{œô§p“b"›W²<ª_ø¯H%9¥ç*»eÎb†Þ$•€Ï¢ÑÚ`ñ“õãí5ò"d™”Í\ݦd€Œ öˆàe •ÜU¸a …5^·$Ç÷Y£#ùg-:„ê§…˜à¢æ0NHá{ –ò¨ý<ÂÕYJm6XÊèxm§È10ŸÛºèüùhÓí²IŸl÷]îí½. 4È;¬c`ð=åïZx.+$:ô½Ò³‚EXðÆ ™òm‹2›÷££9Xr*~jv`mA×:áø ù¡àÀ`³øµÞ¨hÁ&ì˜à´¤¤ëÿAé$BüiVüß¾ãe±›‘] þbœ}Lê qgH]Xq×ìR‚€ð‘»£Pë¤WÞWƒçýIS‹ä!KÕ¡ßÌezZœ‘>Æ0!Kf‡8Å:c¢Ò×eÛ¤8;˜‰Ò[4ŸFó;¦«,N_Pˆå#ª«² d.÷)AÐ~Œò#›Þ3Ú.…WÔ~OšþpõGJ¹ƒÜ °ð`.y€¹gnn©š¼­<%E9xñj[ŒGJ6­·Ü‹–$†ø!Õ\T+;»%Á£aR´/……!Û)V_`Æ(M™ß®ìKӀÔWjçéa £Ë×ÁŒIV.‡€¼Á.URn:xðïä +>ã‚…´ò…(ý] wÕ¤‹Ák`¦KMÁÆÌ ™9ù@@ż,c¼…ãa¯T‰6¡ ‡K#ѱ2üç`f]ÙôšÁy[N´Hïî–¥%ÛåÔ º»"õÕ[œu”LÞ•t>´\y˜Aûܱ b–·ì6¢0¢&×çe­=¸:­någ05D=s±fz&§Ç9Îל‰§qT:Œ S÷ùkIh$ˆ*ʼn٥a9|iÈóÏ=•дG¡¸mQpÍ{¼Š]Y‘Õ‹7×…abT)Â.tU¸{-ôSüÇXàtvX¿eU]Öñm²‰Úð›…S9ùßÛì¦ÍÞ‰3tnç[Ì[nXëÌmx݆/Ä¼è¼ ŸŠßÝ´á… â+!ÖŸÃÚ“ž·!köùü£œ6møÁè”ò@¦îéׯjC݆?¶a¹4ÀÛ«º1 oÚðµ¸àU~^¼ÒˆDï•þ4Ge?Pui ÖGƺ˜± _‰]kÛðO¢Ð?iÇ; :—‘½K‘=²ø2©×xý«6üÅyl·­¸1Û‚IìÑ#3)¼QÿI&LZ¶¢8]óµ$Ç\”. e!¬{èÕïEï¥(N¬ôQ²ÆÏÛZv‹&7LÞŠ[x‘ø"Ï:s!Qü©ÈÙÊ<e‰3¡/E¸¯˜øŸŠ‘Ô—fíè³LsTÚð—ï1œZ"‘‰¬ý„`§~E÷‰€Š­z$÷´ê™H%b«^ !—0­G÷Œ þñÔ¥»çkv–oGRZ,a²·«’á½7s² cUvè<˜pë (¢¯´: J§ÅŠׂíU 5ü`ãÔõ ña‡é;p¼‘@†²kv×Y…"Ú÷yµTÆ®ANÁ¸×‚¯O§òÐØ©H· •Æ ÓäW?YvgDF(4Ÿ7üÝ•ièòv¾RÁV ÞcÍ5hlùÙ•iì¹ãd>:Læ×¡\ Ÿ4cï† t9-ÝŽ°¾»¾™Çbk9j¡4§X‡Þ•{b¤®J5Ê:ÍÊAûY‹"‘N%¡«à”K7¢WåÃ4oVþDF¶#ãÓtmz~`</ì.ÅÒë;k…§b*_L­ã ·Ø#з¸òX¨óŠQ_˜¹m¢xUàC F=Yô˜,kK>ä ZKàA,ÅzO÷¢<'ó`š–U›¾3‘n}W`Ø©^êÁN÷ŹVæã¥æîò†Psüt#êF/ NÃÙÒ…ÏîŠzêæˆ-yDz¥/G’½Ö«€ä”º†lþ¤YªâC¯/è»â½JU„,¶Ð'÷‘Á‰¥¦¿¬-ÌÆ6—ÃÒEÛ=Ú‚^7DmÄ"áf<Òf*–ŽUŒbÿWŠ…Rp‰Ï— ¥m^͇5÷²>ötìu²·eeÆy}ç™xëc'­¥žùÛ±§âw¬*p!ƒ°mç ù§/3òZ€TÒmø#ñûÏ´vüƒ[”¥šÌß¡ HGüS~Û†ìñ玓YJ1ÿÿ‘Õ¨¯—Çôª{P,ñâ§¶àLäü}ú3×íû„kcIÃþ¯Uæê/EÌØ“ÓGmè$~T‰³¾R'Üh=8ìʨ¯ÚS˱:“~¨åPfy÷"+¿_€0 wâ&z/ùpZ ZÑXæÓ}f2^pžAbç €É‹Ï„§_[ ¿.~ÌÈl硲9Ýh5>ßf?i`L¦]ì’ämø5/‘OOâM›e? Û0‰³ìu³üš\q“ܲQ\Ëàq¨9„#påרl6‹Çt"Üt _#"ÉÖúI'mÇüa´ƒö½x`™ƒas‰"2·eJ/G‚ëĵƒ¸Àˆ¨›cpí1:,ýÒÇÈA'ª£ý0f:þ ïÏÌ$ˆÂfšcDÏâÚ|ŒèA„ÅCàzq‹£ø*>ûW=¿ÆBý8L©y¬ìÞŠue~óÂzûø“R¶DJ±¼BP¹¿?aWÙŸ(’ïD|TPSÍHJÒu¾'KwøZ)ùÙø–tKÛvKq#C»(Ðß{"+‚¶Ë¥‹{o§ttk§ø'¾JCÆuчß!ñðÒ¯«hß”y,ˆõÝuHT;þ8G)t.TV^ðÊý”Œ‰ O&ïÊÄ_² ›ãM÷°ˆýx«!°Ê«€¼G.9 yª?Ý’÷L zH~âÑ[v#\Že£™½Ü£§ÃøÓ¼Ø¸+}]¬Ømÿ‚¦k´ž“(ú!êÐPšÄì¹(‡ç|tDkòdŽfmÑPÒ4–¿Ïco~S2±ü]S¸çºÀâ_æãÄôðºžöZ¼IÏÜt1ÍZ%§ ¹#k÷ì:a8ÝD먽{¹^>p¡÷µüÐOm`.ä×úáoûðä‰ÝoÙ¥{€ÙkްÊÎïkKÑ3êq§Úfɰ=î-§r¿™=H§ÙtïQà£A0^üpíÓ¦å_–r¶Nó_+¡Gù~‹§áT^­g­¢€ukLg-ØìÇ ,ÚÙðÀ'­¦· 8ù.ælendstream endobj 268 0 obj << /Filter /FlateDecode /Length 5695 >> stream xœÝ]YoeÇq~g ÿ† =æQï‹‘ˆœH¡É„@ȆqÅápÆ™!Ç$GËSþz¾êµúðœËe®$0buJuz©­¿ª®Kÿe'f¹ôŸòÏ‹'_¾Ò~wuw"vW'9‘éßîÊ?.>ì¾:‡5 ÌQD¹;s’?•;¥Â¬Üyëç¨íîüÃÉ÷ÓW§gbV1†à¦ËS1ÛhUtÓ=†R/¾#fC´Ó5¨:xéÜ´?=ÓZÍ1ØéPUpƆé6ñ aÂÏbÕ¬EP;ÐLÀdawþúdúïÓó?Ÿ¤õ4Ø&ûwßOŸ0—‘NÙˆÅÎhgÑ{5ý&¯æƒ™¾> zVBy,†Ýì…íNã„3ØÖköýmÚ¢Q ¼I¡]˜.N•££‡émžÞF9}wIŸé´u‰YF¡”›îHLÁicÆïêtÞ/_Áu ŒœqèÎ-5üŸÏOþ * Pž•ÊÎJïŒÁ7"BUÂÎÂîn/wÿ±»>¨k)å.ÌÐC0ƒ²c˜ƒÐ;« …Hºž~wy}yÛV~Ƥ4áÍq"»sBÍB«l@ÿt¹¿{÷åù©ÄJBêéæ~ÿþ”dV±°£Âî'ìì÷ø¿?c²oN¤ÄöŒßYo#¤¶û F!4Êû“o\.IÇ1®J¸ v®B¸‚…Æ4ç*âzL&2µžË™Ä•¾õl¥K‚†,‰Yì4fv^%¥“1ªhÔU^%˜¢ 0E'ÉP>0ŠR³3ù †gs®Ja\ÚxøçjÆeD2IÆÕ(œËa<0ã!†35 çòþ?pU ã‚ÛÍQq®Fá\Q!¨ \•¸¼‰³NØ(Œ+?ËAò¹œ™ípÆFa\QIÄIÎÕ(œËÇÙ gl”Î¥¸f>U#pžhFv ã’Ái˜ªQ>%ßU§p.çf®Ja\Zi\Aœ«Q8W³öW¥0.£Ãì¹ä;¥s=æØv6ÆXÇ=›â™Bè´°/¥h'ã)¢EôVM2œÚç'i:U·¡R§gÑlp/ÙÓ3ácZO’†qÊL&3à.’²Ï úÐ÷Bÿ̵ϔÌC«Ã¤;ƒ¢a@¨ÖSì §É$¢”œ¤¢½Ëhâ$Ý)ð€0~RlaÛ#8ÝQ-^!KåÍì!#ë“BÊ=µìq*Š_›ÑÄã&™Ý#‰3mƸi+\r¦ØÅXyY@î^}ƒ³Øz¼mS;x½²Ø37%2H\WмÓdJ@·'Ó{†ž9#®H×%ð‚=ÆŒÊÏÊ@®pW­ÔQf„± ø–EÈÁ?0£1É nM³zh™ø ¬—jÈÁÛÌQN_5þäQt^5äüQäY5ä4"ÀQN]4—2qm¾'"ÞG`ãf‚ÏòxU\¹ò!Š—ù)ÙÑÐïn¯òŸ¼YÙ ÿþÕ7‡¿üöYÁïÒ‘²)òALÊgÿþÓüNWš››þæDY@4 »‚­³ hÐ(ŒËaÌãv#p4 ãòÚ#ùä\¹uü°`£0®€ÔT€¥Qò¸Ù ûjÎ¥‰°4JçÒ¸¡cW§p.@®Ja\RGDkÎÕ(œ+:J9W¥0.…8-ù;…qi‰˜ÁÏØ)œËü¸*…qdσ;…s!ËÃ…qYdØjØW£p.˜¼dß(ŒË zìÆå%R¾a_òYÀSæ1qgFiχÊ ‚²ÏödÜežÄqx"_®ÀÓ蘀g]ĵ¥ OCϪ[Ò³‘綃3œ·m¯Ó–é3¦íHÁ™6C_nÓk¦­À™6ýŒ3m†¥át[n˜i+tq¦Íhs|ø­. %ÿ1!?æh nç FŽ֜njfÖÇØ¢ HΩ–Aþ¨Â1fôtóÃj£ÄŒÇÈ9´D€‰€€‚81#\¹ Ã?µuǘÑã™°sZÏÞEÕBÌÎÂxôQr7ÃŽÆÛ£Ø·pTøN[ôÇ£Ò³ƒŸ“}+¹š¼¼4wëîmgkW³Žº· j>ŽÊ‹w[*HËcȳ¹7¥—á¸î 1Šcº· ~6â(VTÝIˆ:Ž¥W÷†ïÇvŠw“ªM<Šª³wSØ5ú(z©Þ-#ôâèÝ6Ü^«{ü¬TõqXÿÊøÔÄÿÿSÿƒß¿(ñ'm{µš÷«ã ó@Þ¯="Ÿ°g£° (j†Ã5 çB0ff¸"Ðߘ™5Jç2@phÙ)œËÅ13ëÆ%Þ?c§p®@f®Ja\ô¾ëù;…s!MRràªÆ¥é=rØW£0.#f¹ì;…sÁ?̰¯Fa\VAÎÕ(œ Wdß(ŒËMö˸…qy!É^W£p.$ˆv8c£0® íXëïΣ7ƒì…qE-ƈNá\sj8c£|V毽Màè@ê_ÓdÓóö§¥þ4lNÒ._œêd%W‡œsæï¡º3ªû‡ON™Cµ5b©˜0•2‚ >%þ宽=QÞ_ˆ,íGÆ!:Ö` ÚÍ•«7T.ÞиZs@ãbÍ«54.Ö@AW{ãy¡àÙåˆm›fyï¶Is¦M+fÚ2ûa¦-«?~.n /á@®ê((Óx\9@1Ž úŽlšú[¨g@˜c¤¥&S˜“ƒ‚â1p«TÔ½â±G±Šÿ_šO5I€º£¼Vv©ù( oÓPòÎcd?MCf]ž/Õõ”ÄU+ú,÷x(ZĬ§bâCüOÁÄ¿&¦[U®¿…éã ó˜{vÉ‹èõ;*7}s}{Óõÿ9Ób¿Rè jó]œn/÷÷—¯û]z¬¨ ÇÕtÂ{«¦üâ`Sf6¨euæ®°á0¸6Eê(±×§K—–ÏiE¡¤º­ÎŽü\ì™R¯ø—÷ºð>½€IFX7R¡Ð2ßc²YK| SµJšÝ±Ï×ýk/SÿûºPê&_þ,Â^u@ˆ’޽\UÊ¡M¶àHoýûW¡|>Xb)¨£R.ðJ‡7•rh“íkûQþu¡4u¯ZëØ½[5ÐUÜîJïë×ÔA,„Š>7‹D˜®>Q/°Ã­eéLhù’Z‹éÙÅLæ·Éé^PÓ+pD jYN Ç‹ÀhBÀ±§l³…âÎ?žRGˆˆyfìKúéš:’Ô¶ôO{Hbú@D 9ýDŽ¡ôrXì6Oà©}÷M;ᆾéÔN-µŒqÜk|žîê>Üô†‘oêô`gÝäo{ƒwÍÀкºy#vYkƒQw3Dmžk ¿Åli刄gÕ¢„¤ƒÙBqÆøqóQ°å 6óMšZßµ}"´§sxèò*5šGkät“T!TÐjú˜b³öÐ9p‹È­îoR3}48F•Âþxì ZØIgr‹º£“`k|ëç§;0 Ãâÿ>õæ»”IÍêgu;gäÇ&ª¼+fÒœK5ýHóP^àÐÑmÙXÞŠ•°¡iÆ~vðC²ȯ}nÉ¿Á,yRÖ&{œá¿²Á+]~ªœÕÉv)®=Øë>oTå§ý]ú-,VrÈ[rˆºÓ|zf]j™ÇmzF½þQúl£°Jëìá¡ÔÅðã‰Qêd&J"#¿+ÒBb¼Ïv¢e"Ö¸ã6—>Äýé7§~]õ ͤ{d׿dbw´¡´U_ÛŽAMM?2•$ËVfvHÑ‹e—p„Í~ΑÃQH›Í¦ÔÒ>³JmGåh¯‡€±Œºä‡ 9Íc Ý–á`Ý;Þ®N£œ(;G(¿NIF`„Ú´ 5÷Ž!*!ç kP0ÔBqó‰iâc²3Ìa‰9eˆÙË“-Ι4Oó03‡aJ¹^Ûwô]‹#o;9wŸô{—€{&ð 3XAÛýFˆä£³U”XHrŽ_õe83Om•×ù3Š6?žZ€yltÚ¿g:ÌÆ$} ’üúIç ±DÔôó$Ѝ4—–3ïsXA $—¨ë±³Þ|`1êcŠFNXEÚnÑ|óÒe=ød¾ynrÂä ô³[ýcáĪq³ë5®8W÷”~–•÷Tè³óÀiO£ÆË°»z¶ê¼dÔ6ï…Ÿ7Ç~-¢™~ÎC$ÆE[aâÆ›Reµ.¹J¦É§»I[AÆ-Lñ¾ü¬ –þ³úl¨€l–/žn@z¯ÓÜ@˜oýÒ÷¼þ›µ«þ›µÁº¸Xœe­¿iKJq€­f Ä»aÞ—%` ‡šé `ÑÆ<øs´³/7Œ´|òüÒkAgÏÅŸålöO¼› \ 7[xQR4#ôÉA?Ï7£¾ì²)ÃâÿÓ7ÞÍ‹3Gƒ¹;{þ9åº ‘ŸÕ%±)VóÊ KGh=ý.t¥gl3 ÈG$¤Ü†÷}¸ïÃéÃw}øs~ýßúP×=¸éï:õ¬S¯ì̆P·“†7}øº/ûð}þaêãV™ÙûÕåÞõáÏ}øõc ¿©‡ ðãF}»úÙghŸý}~ѾêÃoûðw}Èxÿpš¬²ÙïNÎÿöÿžmÜ®îìýªÞÖuü¿Â þÔ‡zÕL˜â>¼ÌÚ¿xl^f1׫'fóÖˆ†'Œ½X²"äøìËð¾¿èÃbˆËrÂYý,çïRB{©S¬ ”õ<@'[H Ô!z]€Ý£wyjÂg¨Ê07íˆOM’o(`deårñ.Õ,Ö~}}±¼ÎrƵ|‘ À*jÐ[Èÿ}Þ¥Û‹4*KLçÜ©7&aWß"¥Êò°”„0£å$Uõ«‰Ô]—QF’6 `tQZ8ÄᢽÉ+P¥ƒg-ÉVu¤Üvù6-mì ÏÓ"ò=/”—‹K{C&cFEéO@~³Un`X–çÎ ×r “j8€ÎH5o¬M y)ÇA\Ÿ–¦ßÒžf¬s³ÐtCøYP·R¾ ¾MjÎÉÛ5™^¤ŸùVRc¨ñ.ó²RQªwìG'O þ\‚òb·%Æj`ú>}= Ó†\©ÿ<,·1˰ÚeþŒò器–<¤F°?%j‹Wq8^h:þ'6êœÌ¢ízH’€” …¤Ú€ækçï<ýˆ·Y´8Ó†\8g‘¶Ô;Õƒu¿(GN:‹jáÈD…âW²æ’Ñ6=ð¿ÎQÌŽK¸Çï41㺩÷‡¤òëxÔé&ádir•³¹ên°>æn¬†—µˆb,”|Zt©™ò#•ždDvaq ñ9ˆJÝÏ÷ ˆ\!Ó§xP¤¢SÜØÆ¶¥Í¹¥ ª­—l­„ÿgûO—YÞtuÁ«wM1¼&ÇçÍ¢fQ*JÔò–‘Œ‡ýë+VºÎ<ÔÈêS#ý—ˆTÕÓ3òóžV4EkwZºè¹Ô<é¡tc>fzbó}¿fXÒ¾g‡ÌdÜpq(°’Û/Yx÷ᢄŸø@Á) ýA¹é¥˜u:úxcî³¹[J¹|²édÄ™p†%²É·Üø3Öl”+ö,e‹P÷C-eµ$‡ûs¨ªö`ÅÀÄM)Ù*«É+n>õrK.¬RŸ‡ñEy!áj#É@`ˆW©üªËŠ ˆ ¹ˆÚñìZ­&= ªÖ¬Ì¸V­‰ååN=úª¥R9ÖßêP›×$¿¸ûø]7ìZº¨2æO{õ)HÉÕC†*RRŽG¤шéÕ¡BðR® äž®¥ê!þ%ˆî,w§\ª y$T`¨µ ¡‚ŸØÍœƒ« fòýX¿}´ôU&OÙ£êPé+§¨ËBÆí*õC®—B.Wg¸ëÃReHã¿BaåÙU†–[7Ü&¤?ÆÄB Yf²Z+>°’ÄÇ>œûÚa$@÷ÒÔgÏ Ó1ÏÚ¹XQÿ¢›ÅM^öá›.å¹ÏWU}¹ª=¦Óuº^å}áŒáu/–uaˆ !·¡íÃÀõ¸F5}{iéÿáj­`Ú5øBÓáAàÛÇŒ`=¤\¬2|\>~¦¥Qä>g«2™WO¯W‡î1ù¹U½¬¯¶®±:ïújfu5¹ºš[]"ö¡_]Íóý®U+[ÌVñ¯§íy·¼ÈÂÿbÉ‚lTîaái—õ¥9¦Žè ‹ŽvŽîïKî¶ÈÆ\“V*¤­„C§_g¤lTÊ©!’ó”gè´ ßÕ ›„¹hLYÕƒ§ÈFYÑaŸ:]D´Æñ,b‘L-o+j_µ®=®w}Ýö!+`³ê=«pXåªákʧŸ`Y©êâ¹çfSWRŠ46Üs°Éx” Uh†Z ¯7 S~ˆN~”k±XPki9ŒÏNËóÚR \é¤`‡Ë¸&‡küZYL%7´™}ŠÎÉh*G3+_3#§))Œ£ý5.ø 3ŠôW2Ç*YN ˜+êÇ \×CT–ŽË†¡,³n®jVHd}HÉÄë<±‹™s'Ðán*ú‹ønPb+(ü’NgüØAŶ|•VÞxÑ¿ÏÓ§j×›²©Ÿ°^X¨u2¶S©¶³™šT%ýxa«¶8VÁIjFh^fá¥$^È`t–ý=ȵX ,Uõâ<]T$Yíó:=XJÉÖÿ™ÝbL˜¬nÉ ôÛiz[aUÆQíæã…ö÷‰/wcZLG‡3òŽ1SˆŽ±¨ÞÝejê#“Ó.È\Á£:ËÃò >N†@“P9ˆ¿t ­¯e©ä+âÃn_Új¬«¬”¿ê‚‡ú­R‹«+°\ÍÉ(ƒFCù ÆPã¶Øþf8w~2Ó¼aelÒÞ(sÉõJøöÃSú©=]ʽ˜‰m¸ÛXçkrQ« úA²ãs;ey ˇái_¹L½*ýoÊ´îëÕr1P-~l†¢Ó;Y_ä½7d:#m7jÎ/™~ìp‡5•Nš{^gº“oí™á«¦>XMÞ1An°Ý_³Ù.úlël|µ‰ ¡Ȧ‡Ò>6:59ÕÀúº¥#ZÚy-`-½n]æsVÝðŽ»ìØ]²zÄóƒõ‹ñ—å!tÑyj]­ÆFoõ¬pϯ‘GÍßõ þW›·úè5åÀ›¯U5žÒkU~[¡?ˆ5ôÖ^ÑO)óT½rE\÷Àν„÷Â/D5þùÃy\ù±éˆu}ׇÿÙ‡¬¬öªYÏÙûU`Ç{…6:X¬Ÿ¥3m'°îHbþÇ«ýÐq™äÆG%ò½”ι¥ÙÇ\^ß?,´&‡ã’Íý Ñh¹ Ž4 ™ÊzV²/«G¹¶˜”Ýù j¦ôxåÍÕxp ¿f³÷kEôÔ¹šV÷›²áùÉøU¿Ûhª/k(?\+¿}ÇžCínØ‹·À牌×nÑiTÀÙø{ ˆ[­><¸rlؾÖC8ùM ­7ãÃe^ÑOÛ¥„úD²¿º*Ÿ¦¶˜>®ÊKââ j/=#Ù3ƒaÿcݵË_ý*µçÿÃOøÕ¯®³·ÚLu!Ÿ÷4ÂÝnóÙ£=¡.áòÛÜÿûŒÍåendstream endobj 269 0 obj << /Filter /FlateDecode /Length 3993 >> stream xœÍ[ÛnÇ}gŒ|ÃBÈÃÒñNú~1â‘ìD@6Œ%)Q²I.Í•, òí©êî鮞­Ù%%Ú1ô âlOߪêԩꞟb ÿÊÿgWGùFûÅÅöH,.Ž~:’é×EùïìjñøZ|2DåâäÅQ~U.‚\x뇨íâäêh)ÍñÉÐØš®qðƒ‹^89?z¶|z,%¤‘ny‰¢Ú…åóã•t Úºå/ðXÆ(}X~…-b Á-_a C´Ð`¥µ‚?üò_ÇA ðš†Ä`cô^-oSS!LXnÓS¼_Û\czpÂ/Õñw'ÿ„9Ã/tÒ«qÖ+m‡à\žû7ÐGŒÎ˜¸üÛñÊ*=èâ4=…ñbšD_7qÝÄ¯šøª‰¿Œ¹å_ÛÓU{ºeû%£Ý4ñi/Ù¶¿$ц—WM|ÕDÒàÛe“ÏšxʾwÝÄs¾‹Ï›ø¼‰ë&¾nâË&n›øÉ¸)ayÒžnØH¿—íµU?|:ß·ÞþK§†V¥µ” £åߦׂT¾Œ–Äu_7ñŠm{MÛ–Áìò?M|rø5®™Ù'iæ«qê+©k¢Ê+øóñJ‰h¡Ê“=mšxÎÜe3äÏšHìôË&þ›õ•ošxÁš7™µ·¬Vy«¿l⻦ÉÏšøèö¿94Úš˜ –ÅýìÔ4Û–Ìç–µô¶3;µ†d!,N¾<:ù¸³¢orôqª¿bž²ú&6wÖħýN×›šáÞ:Šö b˜Wz\ nØþi?nó¿ƒÈF•4Žðã8ÿ·pr;ã¦öŸoKýøQÉ_Š=Ÿ°®G܉8Ù9;0ïY|¿$šð·…+¶í5Û¶÷îêYˆ®û â†ÕË-û”à o<ÏÙ¿(êÜ£ûû™bë˜÷=iÂà”]¬’sH™#Í“CPzÓÄ¡‰ Éɾo ¨¡4êzÐ9‚Öu¡Žm^ÕYÓ黎ϛø¢m×À«ä‹&^ßCÁglƒ¶ä4ã¬èÞr«µM M‡DMû%ÔŒè×NM}]ãÌÌî>ZdïØUhv:ŠŠhèåïi'¬b‰6ùÌ#o;ï9ipÞ‚ÌŠn:ï(ôا‘î4§L~42¯AÞPåû&ØUvAžm›íE @‘àr KOùõ4WõÊV³óä7 O5ø¾‰º©ºU$ ?nO¿h"Aò_­Ð í£™¨žíŸ5É&~×tE–kš(Yt T}Z=“ï×±ý†&Fv[–ÚWˆòRã¸Ô¬È(!Tô)‘ÖXOªúhëS2 #½öVjþÞË \à ÎjZàÁ§Nh©ÓVµ„f1ÒAN‘Às/Ç‚S4Þ‚Æaº"ê°|{ ¿!=6ÅÊ‘uƒ…ù\‘¹mRÇ1è²ü±«i]ÖêÏ`VA…‚‡°KN)Fƒhã8̃lKиýuïXË•ƒ±V>䫸ôPƒï)\sëòj0!ÖÙûÙÝK›f'Á«$^6­_g…F)gZTóÉ D &Luƒî+—ëÔ"ëœI×EíëÚ\£±)pFyh0Öö3Z£t£Á'sþcj!Œvªk–š‹©scŒSÞ¶ß?UÁ VëRáÁì>Ù®LE£br &ƒŽ“»IJ5ÉQIƒTg V)»\_uN Mø»V³BW ©åæ6?wð|YšŒ°k©R„}ð3CòžO&rÖ6mC'r“]_(—Ê¿ø*l\é¥kt}#•¦v°Íó1`/ºlÊ4”—u!+ìï×]´n0ð×Rê훕‹Ô‰vUuÆ=ÞR­ð]QfŒÇ'¤´jù3J2zÒm ø™*þÆ2` €)HðŠ~ˆì‡É-æ} z0‡Ÿ­M΂×z~“„Å›®‹ªy—è¸ÑSïbP†ÅIð]=#'uÓÓÞ ÍÝÄ;¦@·ÚtŠÜ ²D“g¸ãü¾Ò üoD“ÜÖ×åiÿÖ :Œ5¥7'*äøæ]~1ÀæŽ16$•h 0‚N›~—ó(VE·ÇWì-²÷Ué] Ã¥qžWÉLtLU“æ¯ël'ÎÑÍ tá"÷'¦q¢µxž[ |ÔLôu2$P¢õj pŸÏN4º’ÊTh2édª†äyƒ†£ŒÑkBwïµý%+ÿ¨#J)Í0rpÆ–tCÜ•XdæÈî÷ËÛøGCƒ1WݼiÖCöë<ßrÀÝex ÞK|¯ÆGºW/*Äìbz \44dÛ0ÊšKî`½²r«ânð^û¦å ÓmŠë7dÔ¼ Ê —¼3zˆÑ¡¬7õAviSZD¹“¡àâ¢v’5²®àQƒrR½Ý¹1ÕùÁûXKÔ…©¾õZÈæ¢—Ë&^l בì˜Ni•ÐwxÛE‚£èÀZ~F¶¡„êƒü6ï2âu~ ÔÚ,gL\•Ä-Eüô*€æ6$YJÀ޽€¿.7/r7v_6‘ @Ž-x³,F‚á˜Ý€P,3Ç L¦ï8:†&ÊSä Î@àÜœ6_ZŸNÐz“ÈÓVhâ’ýöBÓÍÞ“R¬¤×P~‚ñcò—3T“YU@~© ×KŒÍ|¡ÉæzŠñÙØK0Ó–‚6¸®Ç~ÿ{Þ o½_^PGË5Ž´JFèêHY/ÒÞW£!w ÐJŽcä‹væ9vBMÞ ÀmÕqÓ€w–ì/µ*J9Îs68j¥›ü4F7CrîÄΩ§þ`oÑx€{hzç-íRYç“¢{2Âñ-.Iü¦ÏD&@’žLRKKØµŽ”[Mb>60`÷MZš'}Ô|†¶žT.!°€IîԵΈû”,Ñb-è&¡– :޹—ž³eYTÁêôHŒƒ•%ï¨A’¦)7uܲ(’ ®·¹_dD“Z>ö@ )(õfZÙ<¡i…•9Ù ÄýHÚ_ž‰îøó<ºIÀBÌ%OPX:¥;,qd؉ ~’xOç“í¼#}9hæX€å&³ÙsV¡ê2™¬û²äÆð£ÂKщ±à’ʯ;Ä>ïQBj@å‚ÔyÕëž8•þZ ]•·ôw1ã–S/'Ý‘²Ý˜l2†äŽ>ÝMO¢ÌÖš8{ð]½ˆ 3­Š±×“Òà vb×¹Hàë½iÌWéÕÝyÎùm—R¤KŰ›²s„+ôc@’‚«ÜƒÆJá*8€(S¡Â"©z¢ ¬oµY’‚S…ðMv¯ñ6óx…¹£áÕ‹ŠË{HO ÉE°WÚÎÓ²+äé„ÈÕà6tùP%åOítLE`:áÌÿ„³»üÓCï1ª_•wÅl)=" @~Çš9TƒÞMo'ÛÜù˜|Μöè á{¬'!ÏC•ëú³lî¦@[Fœ>–T¦ÚŽè-Ü\àkž]èCg-}#Ùsz²U|ºAš&gKþZ°Žèåb*ª¡=ýãäèë£üµ füâö¾Ÿ{ÍBÛ“2é;‹× Jß÷îص0Žö÷¬wŸ´bdÅS"™v"N™dQÞŒFz«˜`´·i*=Séи&¬h¹FÉTeš@ðûrì®Ü¿2 vçpÛÜ€Än¼.ã.µ¬ÝÂpJ5;›¦8ŒöÁl:­ÆõvÙßDÌì/ ©…ÛC?ÒÚå$•9çº-C,O‚eÀ‚ÓM?$„ÚàÏ ñðá$—¢íçÝuù¦v*v{?9*…œ­új¯(Šã²õ>gï·t·vú@^o!YcÚ˜ÁÚüu‘û¯r¡±t*\vø1b:K¯)ŽõîÐ¥Uö£—35Íu%4säGjäif§ºŒ,à`ݘ5BÔ© _¦AòAWÄãº!…†Ü±Kg¸òz6—TLl&uà ÌÑ<”0”Þ¿˜bäĽ ÍKn×hQÐL xŸ_yœ š„ê;»Õ¤~™ Ò]ø‘NðAj2$¿ImÀØŽ‹õ´†^vÂîU-Y ­4›]·êíN/ŸãÓÈVÐÚMÝûO{`.Ÿ–•¬B¤ˆûîžÞçrŽJ¯žž±ß4ñà—’öÕÃ] µ"À¦&×w–?ÅŠ:Æ€Q,ÔWÈÇä©h{Nnjšv”Ü!ó÷7Û—JûNã}ŠŒ°æ0­ss‚ùL×io–Ӭߧ+[ôäf7Eÿp½7’Œ©^VÙ)Ô'Ø›Ïqh «M‡ƒH˜;SìòÑ—ùUØ æü¢G¹Ç2‡\Ÿ…”‰¾øp7E<Þl4³3ÙÙÐÂô-ø_Y¼…¿ô”ˆgõi˜v‚<ȸL»|JKˆE欽{ ]öŽ/@oøûG9 Ëóê+eùb%°×’#pê™ÜŒ÷›¼!%5¸v\v™ƒM¼ ×W6”cß´ƒú{Z‰yùD¦rUª½‡.N‘².H°¦\+mîn”®ÎôP˜ƒÃ¿sã¿>!_›’«3$$И~­ï}{Íö–]Å}¾™½ÃkU$÷÷_L7uzºÆÀH?“øõ¾Œä¾>>ã'A¾ä¿Ëã?¹¼eÛþß…’¯ëÏ›ø¨œ(¤„n„ÂBRƒ‰ÌÞ1_ïùŸšO=i"ùÀŒ|F<˜ŠLÃÀ*¹sƒõuòF¿ËĤ{»“êâ¢Mg`»ñ”˜pOÉG(vJbÐÕ=S‡ºuìlT«m~}ô?9K®­endstream endobj 270 0 obj << /Filter /FlateDecode /Length 3669 >> stream xœí]ËŽÜÆÝ7ò„Wl C×ûáENbF6dad1–F93{$'Ùå×s‹¬Ç!‹Ýœi¶Çh²zΜ:¬Ç­ÛuY÷B?7¬ã ÿÅ¿_ßï¾|)móîãŽ5ïv?ïøðÛ&þõú¾ùú1´"¤óÌóæÕÛÝØ”7B¸NÞXm;/uóê~÷}ûõþŠuÂ{çL{³göZxÓ~¢œ1Ë~|ÈÖ;¯Û„Jg¹1íõþJJÑy§Û„ g”víãÀeL¹¿½úv§E'™ aÊ‘˜k^½ÙµÿÞ¿úq7<_qEÝ„ß}ßþBZŠ¡==ì*ôÌ[+ÚßO³Nµßìì–F½aÔèaaFQ·Þ@ûÇ¡‹JâÒ'Τqíë½0a讽åµçí_nB3éÔf sÏ„0íÇ0MÎH¥¦íÒ#Âx¿|©®P¼£¡Ð:„qsþçW»ïh -uÒuŒ7JQæi©˜î˜noš¿6Ž®5ç®ñ4h\lî}ç•lŒ×´Nb\ì—ŸÞ=^ß7o›Ÿ>~ºy|ÿðØüô¸¬}øáú‡÷wï?½¿ùØü¡á7WLíÃP†nj2%\óOzÀ·ôçGêÓ‹†“I5Æ2ÛiÙÜ¢LÇEs·ëw\Y¼FVF€eïœCVF€å¼è8GVF ‹V•ìM« À’\tF"+#…5Œ“Kß9Ûh!‡e¡§rÆÉl32ô-±¬µáo`EY†Q¬VB&,nu'ð‰ ™°¤ [÷ÈŠÈ„¥”ëœBVDÆq³1šßÎ1mÑļì˜ÐŽæ€wŒü ™XKÖË‚m³F+ÚfBöNQ!hk? ´Ë"—["9 ²³‘ËU§9»€ê¬K+ä¸j¢î`¢hš+²X©9£ª' ¬Œ+š°2¬h~ÀÊHa%ó+¬‚+š°2RXÏ_(N ¤ÈH5­ºç£+ …Ú_qòÏävèA?ïȉÓjøÁwLÒãJZóâ‚Ëæ{ÏÉ«wd\E.(ù²Cž„Ñj{ØaBOÎw²ÃȵÓ×Ä;bs?,jü=š<`ã ã{¼Ù½]§ÐÒG–sJ!ûì47Qf•̉¼…:.³F!­Éc ô1ÆÊN)˜dÖênww6­¾â‘–u¦ó[ÖH¥>ôëlZ}àY²PÔòd™´I eÌZÅ~M«¯xdÉŒ<³À«‘Z}è×Ù´úŠ\ ï=‡Z5R©‡~O«¯x¤KV@¥=ôêLJýœ·+g´_wô*%x+:$Ò¡VÓ 5µÛº¯ÜËÝ2kMãlZµ#¼Ÿû«» Ò¾]p—§*UNõ~Á;- síÛÚ÷ž¬tqÑ}qÑ¿-ýÄøûÀ™uõÀŸ0ÃüÐ…H¤S5ctÄO /@œWáÿ÷Åj+cÝ3Tyùâ)*ë!…ë¸1'bLahÃS°aDðGCHñLJûŸöRuÞZÓv Ÿ†›§ye/vBkZY|Qˆú­Ñ4AÈʰ¼>+#…%99B…o# ,éM°8`eX†B¼0ÌÂʾxnð'4YÂðFÆuÊbð—£:‘â=ÁŸ×9ø‹¿q¥Í$øÃ¹‡Ð5#àÆ™VF€gX)¬4Ó…U`Å™VF€gX)¬`ÅB‘C¸ÿ¬ W(?¼n)ÞOÚFènKŒ+4ù«*~õ–¦^êèàÖ(4—†œ'Çc†°üª^@f­¢ó>›V_ñHËišYxÌ)ß}­RHEÓ¯T}8Nuôý1ʬrb܋ۺ¸­‹Ûú »­ûè^pk, síÛÚ»¬T;ÁûÊYÝ-!³V±WgÓªÝj8á):1Of¹F*õÛ÷{ºÖÆ8wýÌ;;?/Î]oõ”8÷ *'Ź :² t §@Wœ)Ð}â^ì¤SŸa¤˜€(*f(°N€#,Å{x^`)ǦauA€e<›^ßXžñŽOX™^¹Ç!AD–€Å!N€G¤Œ+ŽXV°2¬8"`edË-®¤ ³~‰;‰ãc„ÿ¤K\Í„†œŽf&\݃3é +g9dd9$VÉrH,ÌrȬœåYåY9Ë!³ Ë!³r–CfA–CØ Ò ×)qZ /­—¡Q™$åC>•šLR âiÍ,FüK‡õ5 }«øp¿9¬rÂ~ NKÏ9“›êuép5¼K‚9Y¿ Y«ñ[î|Z}Å#-iC¿±eTêC¿Î¦ÕW¼Á»øé½ÄR©ý:›Vy“1Úœf2¢™µŠý:›V_ñHËrúv˜Œ¨F*õ¡_gÓê+i9ᇼ¡Ò²F*õ¡_gÓê+ÞÒeÈ:g!˜Ÿû¦ûÑÍL.Y«xœ=›Víï+·u·€Ì[Å~M«v°÷ Žj©ÔoñéZ‡}q؇ýkvØá©èX9ÙÉ5R©ß.8öÓµ6¾&X?VÏÎßÏ{M°Þê)¯ ž rÒkEÁ‡í‡_È3½&xÒ¾;8€i]Bê¿£NP×…ócß¿ µ©y;–I0ï˜kßýª(‚sžÇš ç¤koöWʸcÜê¯öW–+Š^yû§}3Òò6ÖRxeuûv¨Ʒ׿@eÆÝP! å'$HQ‰ü4cpz6³íçñ™ž›öáí¼øB›É % [1_|ß~A*ÞkçüГøñ®|¼)ÿU>Þ—ï _ì—*?tù]~øÃû+RiN~„¡ µ7ŽìÀ´¯¡ìäÓ0…FHÑ~,oÿÜ{nCáIšL,W‡fS•ÅQXÊs;yö¨ìCï1…Ÿ—å†Zn -^©ëiÿ±×¤3êÐõÐOG–É&u5&z*¤t;Õ^jsˆ/'‹l9Œö´Ä½ûª½ˆsãÃ2®\áïJUÑ0ãPb¬éñÌÕ«à1¼Ñ6,çUH*p²¹ çe埊Ã)FûMµÁ0,×L˜`¬C²R˜ÊqšlmÿÞŠóÁƇ²)ZŒë¡(É+á‡"eÈ­ádq¥6k¤8«­mÿ4£ÉCå7ÃÈ„§èrTv9£ÃáFÙà`Î_g$m¨÷°Ÿ·ÎH*oÂÕc)ë^wsÓ`Y°‚¬XĬ„ +q+!ÀJE…•dÅ"`%X˜êDk«Âò”J„`=ÐÆ„wDÀŠÈ„e=ï,V#%dÂòô%-&OŒ²$§¯œIÍRB&,!m˜O`Ed[Ñ`jFÿšêŒ¸¯ì$™‰…,hÀJ²¢ù+!Ȋ欄+™_aeYÑü€•`^gÄyP:~æB£Ôý“ ÊK…F¸Åæ…FéwG RJ~ýŒ2¼SÓ:£ƒ”œLd…Q2ØÁÁÄÌí2údÖjšÁ¾]«¯x%ëZÖH¥ŽìÛµJ;hÅÌmhY#³VÓ öíZ}Å+Yç…W#µ:f°o×ê+^É:­©Ô1ƒ}»V_ñrÞ94¬€J3Ø·*õsÚ‘ öÔ#EF—uqY—uqYŸÑe…³¿ nd®}[{¶“•ú§†¶ÇkK%g%ú´Ê)¹Ù3^F¬RJ&5Ì@Ì .3°€ÌZM3©·kõ¯d?CË©Ô1“z»V_ñJö3´¬‘J3©·k•LjЊÄвFf­¦™ÔÛµúŠW²Ÿ¡eTê˜I½]«¯x%ûZÖH¥Ž™ÔÛµúŠw¬ôå0çHéËÅy]œ×Åy]œ×Ãy•Ò hY#•úí‚“;]kcx¼~ÀœDŸmðœâ3†Ç©îãXx¼Pø±ifuÿ„š.Ô½<µæƒB éMk¾Ú_MÀÍÆšI¡ŸGk>ˆâ™^¯ù ¹éÜøÔ|ŠçÏRòAs¬hž]ò1¬ž–|äÒ |†’2Äゅ kÿ8ÖO m,Ñ-ú ÃͬèC†š'7)úXãâï‡*Æ1ÎçŽ[K+úfjWWܩ޶i,qîÿµä»Ýò2ªÒendstream endobj 271 0 obj << /Filter /FlateDecode /Length 2586 >> stream xœíÛnÇõ]ú |(àeàÝÎýRÚÔZ4@š}qü@S«•HE´]+èÇ÷œ™Ý=g–³¤¹( yðÉhæÜïËŸ¢“ ÿõÿnnÏ~ó½ö‹«ý™X\ýt&Ó_ý?›ÛÅVpÃàIE”‹ÕåY~*A.¼õ]Ôv±º=k¤[®þ—­).ß¹èáÁêüìuóÍRtJH#]sƒ Ú…æbÙŠNÇ ­k>Á±ŒQúÐ|‹7b Á5×xÃÇ-\hµVð?¾ùÛ2¨žiÀ :£÷ª¹OW…0¡Ù§S¼‰í¶ˆ@wNøF-߬þŒ<;Îs;0ÝšØiUæý×€#FgLl¾&pGàmtlîìÌäŒí¼qÑJÝYBÆþš®½$Pø&=nË×±çM.[«t'@&A 1Ò‚ÀH ­‚†ÀÀ™ŸÈ €éÿr¶úò5=ž_¥*O†‹€ (߉žoýÙÍ3¼~Šyê*ѧ.®¨g™Çœ²”«úRáV9¬Ê¼¡­RK;T4„ªnÞ.Á–dóÐFÝìÞb$÷QýÓ‚Aùf}RD66[ô² ñ-{x…!KDÄÑ"7Â{×¼K­4˜|"¤Ÿæ‹)AHTÚ%PÄhCìóLpV7çùƒTöqi ¦µf}“XsÑ4§l,ÒÛ>¿ÒÍ%ÜÐÁKç˜. PK~ (þ¹T® Bzðbø³é„”¨@,TÐ ‘ˆ#@)R ¨"8c²f5bÄ^qdCäàløc“4.¼(ós—28èTÛåë!ÔA–c° ˜oÇDÎÐJ’d½%IFeBNÖ?.{¼1cé Æuò*íáZ%Û&QðCn¤l€ÖŒe>vÞrí¤jâbPF%ã;aT–sò”öWeاŒ *ómÀ&`ap`SœÞb÷¹Äs4»Ë\­€÷Ãwä÷ù‚ÞÑéh‘^Äh@®,¹ŠÉÆ‚ºIbYí<§°£:Z«‘R€̰¢à¿'ð"6º5ï ¼%ÝÝò»™8£¬Ð=ùvL*ÈÅ÷”Ú¾¢„÷vÊP߸&ݽ!pGà†ÀTñ2ñ_WŸ¥À:Ìß/†¯´šTQ1ÂãåÐ×󀃡fýÈÞÿXåx]UÊ;÷ÄÚªª”º.oèÙ×UÂÛ*†B—£˜+þ¾ê4Ïr°vÔ×´«ÒΧ¡!ZˆdÛAè°Rá‚4X]ð®´^Bfë¥Mèw3Dk°»7cO¹d¨á£wk¢[g,Tû…JdÎëñ8)íõ8¼$ö®«–¯;Áy=tžÕßV)bˆ—OÚi *:ý¥Q;6EÔ²àa1w²›»+'Tã£gwvªCœqé“t¥)¶¼¡N>" |IàJbFÀZŸZŸ3ê38EâéSS;pruæUm±SK äêrÌ㿇°òØþÓ°‘†¡T¹¾È3ÎNú½±¡»GTûSì¦ ß:7“BE•H€5y^èg„… ¶Ù¥>ç]Ì7™_eœiø\˜®l æB1|d‚Ê ×+44ïIÙöáÅj­£´ÆDV‹Szn‚ªeuhzh•ë•÷’Àëjî ÐPTj~fÀÏAèW»$¥r…þ78¬Y@IÚÈcCºú²Üèt’¶|8J k »ãÙ¡¾¸ÛO „ïœáߦ+Âh§ò<+¥ÓÞ¼Œf"Âs=ií6lòü×Z>ßÿ*žyý¸IÇ`Äȧ<¦Ái‹ª4R[9¨í¯b®ºÊÏ@)G6E„k Á ³p ê¬í•„*Äð.]½Ð|b©_ާÉl(ÖSyÊlR^³¼tžyrb6Þ?’’˜õŽl ª+ è—”Ù£k~&7i™]òv°`ÆC„”²ÙEÁïKÏí¡§»‡¬5ÛÛ9‘aóoÞ§Dï úkZBÈØyg‡Hc«–È £"¯Ÿ9f¢z•u ®‡(Ja{‡OÂb®{Ôšiä¨[¶0ócj¾#½Ìlßו•J#©—º\r°ÜŽ÷..;³¸w€üß©7xÃ@P”Zé‚VñýÖ 9Ý¡¤`ÚèÐѤFVFFMa3¦¡‡ü΃ÉÖw´-»“*ßY0¢Êß1ú²;M`½óNRÐÀN¹ÇDÎ ¨Êx.vZý³¹]`.dümz%©¢Ÿ™*©"›Çu!„"Adò $É“÷CfÕGÎÞzܦ•ä.9•Ý>[z±}Âa™‡ç0€¨ÂE3Q« +0ýDn_–²“¾·‡5=×¶£u$±\V‘¾yðª_ájÝA'1hTš0³¨cMÇzÃô}^ü ¥&fÒ£1ƒËÔ¿ sÒ=_Á $·Ö¾g°\dO7‰9Oœ/Gø!? 6†bZ´[û%çŸÛ ßv… îþýÜ2—5jŧÁ^†×¨ÛºçképƒGÖ]ÖtGÜåúM¶†ÑÚ›! ¾ I6ƒ§ðCj¥šÑo“v‰´ƒžÑ[)’óª*R]T“~„ò1r°x> í­ Ò=Ú Àqº2AcqOWüÑ/Có %TòÛýGùÓ*tTÎŽ‘c ÞXÝœ«Î‹þ·£€¸I AcpI?þ Ÿ÷§™eÏ7Lùÿo®?wsýjuöWøïßyšSendstream endobj 272 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1ƒ0 Üó ÿ @ÄB†VUÛÇAp¢†þ¾$„ÎÒùî䳯ëÈ6‚|‡/Š`,ë@«ÛL4[uÚb,,O\”r¸)ÿþx‚Ý@æàwµ|^ª6¯ê#„NÓêRP<“說ïŒé±þ“J`2ÅÙ´}AƒÙ*)šJœ7·ˆcn𛤖é÷Œw>¥`‡ø?qShendstream endobj 273 0 obj << /Filter /FlateDecode /Length 2856 >> stream xœíkoT×ñ»å7¬ªJÜ¥ìÍy?*ˆTTT´ja?D"QµìãÆö?REô¯wÎãÞ™{wÖk&DBþàÑñÜ9óž9sŽœˆVNDú©¿—Ç{_=Õ~rp¶'&{?îÉü×Iýµ<žÜŸ†5°ÒFådþr¯|*'J…V;9ñÖ·QÛÉüxïys:­Š1×ìOEk£UÑ5çJ!¼†h›XÕÁKçšÅt¦µjc°Í_`UglhN3®&|?ÿëžU­AM`Í &óÕ^ó¿éüß{y# °Iþö¼¹ZF:e#l6KœEïUs§ìæƒiMƒn•P6nðB8LÒ8á °µ"ߟfR€ Ú…f9U.‰šW…¼²y²Ÿ>Ó1hë2²ŒB)ל%5§~×m‘äýê©ÔÊÈD;$¹¥O‚ƒÅ™ÐÁLÛ68WTð][Å茉ÍcO<Ï ¨9fy+xŠà’Ex ¡ðÝ„—L›.+[vÕdІ cWÃt&­—­ÀÕA²JèZ%%FVgVéV€—ÎX ‘Ýͳ¸<Ý»*–YÝGpƲî¦ÉM¤n­‰j2²7¿ý¼™£aOÜGpÁÚ혵ñ ‹;g½„|ÆoAVu±ËI«§È‚âNBèZ$t“ó ã]ÞIø-KŒx†ÛE!²œñN"(nr ÇÐùȯ˜nãòì<ÙþVýx—`&ô꟡0¡¦¿?ÃÅCƆÊ"ø’†OÏ1Ôš2Ήs†àŸ27àH· ÆH7ldk´ µÝâêmöR”ç—*g“¼oeJh—´Õ±#Ï'Á‚’ßTè vŠä¼bEâåHjÒWC^EŽš/Ø )³áÒÌÃeDŶЊ‰ëëý’œe߇ßÞ€øMò»M© š(ÛÛûV )øÀòÁp/À÷ÂwÝYîJ•u­1f2Ëé7ÕèDéÈ\#HrØkÖ#tf Bœøí^Ÿ^õ‡l8Ÿ³ã•°@ðœV—¾>=cVT£=îC\=eÁ5]íëéÏHá'–³#/X&ÿÁîVË]†¿FðfW¡n%=φŠîÌø)7Ö‘æ»^Òbhں̔ˆ¶jw<ó­éÅøšoÖZ–¾Í!-ÿ]–îšåž³Û¸:»úû[k=v¶Ðd•KøSRØå’ýLn¡Ë5Ë<]bmÏâº]½¥¸D âf‰/kÁmè§Ó‹‘sÞç¶L|nË>·eïØ–çjئé4=¬åŒ!óŽÏ¹/Q®•àuÐ ¬è4Z'm]ÍÕ" ›WÓ4{ôÚ‡Ô¡$ýì,ûŒ6šò]PAMgŠÇd²ºÈCT 0M*Fãó`RA‡â¤¤»÷ÓLŠº ÛoLmµaO’„F* Ô2á|N\uö¹žænV*ÙÐÙæ—„ZÆ|‰ÊB±a›K'ÃgEKF˜æ?‰‰ ¤/ªó­a0Á]'ØÍjF°Ùâ(aÀg2„œ?ôÄ^ ǒ0±ÎÐBIerÓ¥cÚf‘§»QK¯Ðp’2Z,›§ÊD¾³ì~: æúÔQ·0ÚÍÓÉVHªå¨ÈâA–bÊ`<å~àU÷Ú:*ßé@]ICÖ9MÇÍeÐ-5xvº»”Q”¹Ü wÒ U5Ò„‘ÓC®sê÷õKuìüR'—™iÛ}gÎL¤sLk´¿L’EiU`—˜ŽÔ»:¦tyºŠ##çUà£YÐø=#VÌ— ¿-ÄV…ˆ“)_ÌòÎfC§e­æE–Å[´˜ÙÛ⽚½TX ·.4‚¾„ÄÐiå£ ÙºL?}qÂÂúI–K­®Ëžda0‰µ $V%\\ªÆ¾ •š%ÈÆ$†2^®—i;Hv1_³á¤¥E§ºë\—0ŠWv†'=’ÿA.ãMú¼OßB›¬¦t‰äÀ›–ÄC eÁ¿ M䉶Õ… ÆÙˆƒ :·ã |ë•[~2Ó (jÛN=2^¢žôa+ ¼•«‘axÜOÉ.  ¦‰gx·ÑxÛjg;ã?ÂÚÿÁoü‚O<@ÌÈhã%‚‡¾AÎê™Ü#$õI­;VÁMƒ ­•ÿ¹º„‘v{„š$)¨ü ªÅšiê8¯¦©Þ”¡Q.†P…Sj\w–öãBZ†¤åb ØW‹×Rcz·ö©ü òYÏé.Ý…ž¡å°Q&¶Î‡QY Rrbvâ-ù}•ÇÙ–zqʉ£bàЋ\ý¡9ôcùû™[£ C 1µ‘)+WZœ£Îý»qËS–ùëcÚV ²¤L®“}´^…ç´ «^ pIÔ²FX+§‡Ò©Fš[\hŸ‹à·Ã\\uï®ÀȰm(ú zkÔ+#;<Ä4t¿‘çš #ˆdºX»žGWùqîQ±”ÅÏSûÒž6Ǩây³sØd¢«é¥¨Ty†E¥*Ú«¢hìe®Û*¥HuàÛ›˜DÚŸÚX¥ìÀ‚‹£œÂUª:)ÝÛ&Wz‘kôϹ % ÁAÜìc}^¾‹rÐ%¯ëjš®¢š¨ÊJ£í!>òöš™°ƒþü wpÚÌ•Ÿk[·Ûeã hÀ<²?Xž`M ³Å3ÏY„[>ÀÀJ¡£M£]$Ë•mô'åpˆyÁ2V Rù®”eX!xAIùZçôøxÂ%Šá •³vzw"ÐÃëNkOEü<Œ£3ù·¾ã´õ:Ý;8å{‹«[4¡ËtJ’n¢&É:ñz0ŸèÐS’’é‘Sþê)Ê÷u·¹¯=ÒÕFŒGþEà/ø!ô]\áê#\}‚à·~ŒžŽ /Xj‡ìvÿeƒí÷ˆð‚¥@Çz×nüÝÙânü^ Ê.ÄKÓ÷0°H¸"x†à9Ÿ]~ ±ÙƒWIE=ò0ŒZ6 Ëæòx=AfìÄÕI€“ø½`½~ŽzÒW2ü1.Ë'—ܦ³:aq‰áß"H&¥UcÜ]rðâÉçT+£è_n¼¹ù¼ñrl¯±ÖÞ°jÝ –×ßOâ–!Þ䝨H%ƳÙ✥{D/W{ðýÙ©.3Êoi öà/>ÀÝGð„Í4l¤w 2‰tÈW(j\LßlYÜ̹œ©ùæM³¸«]–\°Ä^°ìðÞDpIˆ'$¡òK÷”²Þ%b¶%«Ó‹K¹ùV‚3ækÖ˜W 6Jfc·<¹ŽRI;³%m¦êÁ}3Ÿ³‰Ä V ²Êw­;UBÔG«Þ T‹whaÆ*<òÅÔßÿ…•Ü´Ì’¥D^…¼d)}„³ÇyLÒ½-"Q¿f­Aºir[Kæ1‚|sÃç%‹ðš¥PýŽª!ˆñ}Ôuž”5‚~§3§Âøa‹dU³í9‡ X„Xêd/m÷šûü&TÜåÏÞþ‰{»ý•}}翯Ɔ¼çâ#€àÚõu|öŽNkãã]®úA^}&†MºÛ4—ä¹Ï@-‚AÉ"Ä)?‡ê¹ ïdø†eËú>«¼<òm=jß•a2«»Œ¡›m?H/ÔÃäèpŸMMC¤G»þUΙ"¤7Ý9SóméôfÝ×'|ŽN(üý‘ä;ÃF‰’&I‘³è?„®f?ã£Ñ°tã˜nVÿȶeHúü¯\ÒäÃùÞ?áçÿð½’6endstream endobj 274 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1ƒ0 Üó ÿ @‡.ˆ…. ­PÛÇAp¢†þ¾$„ÎÒùîä³ì‡ÛÀ6‚ƒÃE0–u Õm &š-‹ºm1–'.Ê Ùß•<Án sð‡ZH>/u•WõB§iõ )(žI´UÕµÆt‚XÿI%0™âl®]AƒÙ*)šJœ7·ˆcn𛤖é÷Œw>¥`‡ø;¾Sbendstream endobj 275 0 obj << /Filter /FlateDecode /Length 2159 >> stream xœ­ZKo䯾‚ü{€L»»ºú•£Xˆ‘¼àÃ:a¥Õ®£ÇZ»Ž‘Sþz¾"Ùdq$-g¤±YüØ]]¯«Øó[ç¬ïœü7þ}w»ùæMÈÝõçë®7¿m|ÿ´ÿ¼»í¾="2$¶ºê»ó÷›áUß’ïr̶†ØßnÞšo·;g©ÖR’¹Ú:k¤šÌ\zç²+úò£€s-5š;HCÉ>%s±Ý…@¶–hÎ ¥’8óÐcãò¯ó6‘lp…:ȸ`°Ò_nÌÿ¶ç¿núùÙ3ÔTÏÞšß1ûD±b²hVs&ó—a¶\Ø|¿-Á’£ŒÉ ƒ.JCYMr‰¡Ö¥zÿ¡W‘‰€Å•w!ónKI–^̇aøX½ùÇ•¼j 1õ`_Q2ŸÅL%æå{m Yï7oØi{‹¥À²n_eá;ßü8JÉ6úŽïuÿÀåWÝÏÝÝ&Ú”C¤î¸òüûÎ?ÛøàɦÒq¡¿v·O©Â0“äfóÓ„‚W+V«PM"¨¯“÷gL•1ƒÌž«…_ûXúîþöÓŠ;›Û޶b "ê÷^ÝÃøÓ[bXSŠ‚Ž#ìåŠhÛ$\ 0¬)ÆjkÒ¨&ѨÙÆ QM¢Q•«u¤QM¢Pä9ب@M 1!ë²5‰BoiOP§Hl8ëjL½ÃÛ]¬É"u„ÞvW lPÉ8yä}ñ¦÷W~”‘u7q6»„àJ£êÞÃCœ “åú•8D€“øíY[LjyX»ÍÈòkx ÃD /ÅŽ|µ>¦€åPÂt›÷C8¬•Ÿ³K¶Tè­ 0 ¶ŽÁ8ÞX‹ 0ª”I¥X̧S†h6šÓBÁÖEú$ÁŸõÊ8Á”㣭g`V×O"üZ*Ëÿ»‡ëõ—`¶ÇzéAÞœ0ÈzÄ#ºsN¼ùʈµ‹ ÌABÞübþ~÷åêáÝÕ§/¿l'ú<…}¿®üÙ†RöÂq*±›D§EˆÕEú7‰BJ`…š$É[˜F¡šäUTB‰,öì=*¡íŽ’ÃžA+¸† Ø"€¡Á'Æ/¹‚˜¡¿œ+ž7ÁS\1=)ìŒíà ®8ø“8® ¨˜x@¶„zJsÅ:FqE~ Š+ ¶Ä%WDŸmÑ[5h Ô©¸bÝ+ã xW¬¾tW¬ò"® x+Å¢¹â­y¿ ˆpÔ^æêöç-ɾÌýíÕÝ2–_kí5æ±°D‡ÎãQ¢³½D×ÜŒj…bï“ͺè™$ Ñ"¹&‰F%”8²Sͨ&y¿„ Yæ=~A Ú#2Ò§/Uh¸“KЗLÍHÕ29‡3vz§7WÎÉÛºL˜†š“¯¡tò‰£ƒ8*½œ³ž7ØSœÅg*JPS àGœu«¤ÁÚ…Š }ZpÖ:FsVŠaYßÀ_Ñæ°0š;tY¤]ÄS§ã¬U§,wc­½sa­Žñ"¾bñ§Emóï—ñ$%ÍA*Ï•OàêÄ£sSØ$ºu 5±-4nÚh•º@XZêÑ&ñ)Ø<$j踇j’¨+¬QM¢Pä°'ãzFM"îùQ¡šD£8ÀdU£šD¡^жI£ã2,|Ѹ”—acG~lÔrÄ8!=éÐ’”›šŒðç“íZSùøvíY_0Ö|Z¶k ]`=ð†”<é©nm§¢ À«èä|ÉpZvH·k©JFÍaÚú0¦­¥› îÃJü)š‰ç}²×¬5K/s”Wºµçß:¦]ûÊ(/ì׉ƒQ¶Þ_ƒn/þyñð€8FT˜ÍÇ«ËÓ”_.ãlª»Un7‰f€$5ªI4ª2*·¨QM¢PÁÇlI£&‰F…Œú.hT“hÔ@æ 4 ^ÅK¨]ÁKeÉKR,owh ñ·ÿ€äåN>E¹#¹“4‘›°xÄOw‡/¦¦ç]ö5Mè‘r¤KF@=榀s›Öx§ ¾vŽ—£RäDÕÉÇ♜‚ÏUÊqENˆÐÓÝDN­‡<9­{e¯;<ŽœÖß:¦?<59QõXcÔädÐ ~9IÁu ÒgRtC14‰NùŠ5(H«ž³šÐ…Ï$Ñ(–îW4I4*•*¬PMò*‚ à»”ë£ÂG*Æ3”7cç—9œ²Ý‘ã±ò‘OP.Ô¡'¤Ç=áœ&­ÛÓi2¡ZÂM 9áŽ[9—’JÌuíu¤º~ÔÑcö¾£í±ûñ-鳡ðQNè‘ÿÝÇ“ŸÜJ˜R°€y¾ëO‡QîÁAˆW—åP¼?®œ£ù³)WINŒgñ0g,•è»-ÆtæË0#¬7ž±IJù,Xvè\¹?dw…Šã Ï¥oÕéüEjÎlôÉôçapvÑ\È©<;*äÌ[ÄÉ¥¦k¾ïî‡ë?Í'ìú„¼?8G]—ïïfñdÁ¼üIõô·øÉ€ÖtðybóŸm„ÿ°FsqÓŸ¬'¡­Ä¼WÓÞ? Òìª÷ì»Ò¦·W€õ>”OŽ_.À“>7=#ÐJöÃÉü[ó ÅŔ°´â¼/mEY~6ða6ì½²÷ÍìµÿŽ—ØÖéÙ\ô?<¨‘½<ŸVξ›µ¼î5Ûªí¼Ð\·•O=,£É•1%«œú^ÜBž½‹q„}`Õˆ´ó ¹ò…ŽC;¹SmVÿýXî:endstream endobj 276 0 obj << /Filter /FlateDecode /Length 3244 >> stream xœíkoÇí»Zä7Eï9¹í¼A 6ÀE4ŽÐp㬗•X:E’]Û(úÛËyì’³âiO¶€æƒà¢G\’Ã× )Îo»¢—»"ý«?÷OwþòLûÝã˱{¼óێ̿ݭ?öOwí†I+}Qîîí”Oån»Þú>j»»wºÓ)±Øû­iƒï]ôðÁÞÁÎóîÛ…è•Fºîu¥Ð.t‡‹¥èu Úºî,Ë¥Ý÷ #Æ\w’0| ÑÂRkÿñÝ?Aõð™ ¢·1z¯º‹Œ*„ Ýe^AÁ÷#³õY" {'|§?ïý-Éì¨ÌËA襶}p®ˆþHÄ茉DðçÅÒ*Ý ¦ó*pŽ@w%»Ú#YPtݺ‚{vÕSºeÇ­Iµè3~·³÷ðy÷4iP 0SèŽ@YBMwµW_%ŒªïÄ¡Á|Ý#@p1j' £ ±{\”´Ô —õ³ »7°j¤×>¯,DâW½"‚dÝAà ÁjðÒ¹n}QˆxØØi®޲^,•Ç>L´²±ñ²×…žÚG×YS"ç‰OtR†nE|ê=ø’Ì^•$„õª(/E)=4_Rž«"H½)<.*›ÈïXo±ab„3DÞo“`;¶Á d$ùˆpœK+ß=†ü:3*hE7õ¾˜.Õ½]X›uš¶d„ßzÞÈ-Sl„Oªð½HþA²ÂgCíTµ¸U±æ• ÀÈ!c;¼ÒÀ´i$@É–"ã©ÍÒ.—Ã6—R÷ÖDUv[½Ó«êI¼Èäp˜º`hðöf“‰´„0"ŸÑhHZkJe†çåC<çYé ÉäŽÀœbguÖ˜s„¹,(…é½ ƒ0‡"xʦ%{í>D}Úp° á’älÀF©Æg‹.•‹“7_Àlhü} 8? dÃ%dù|TìåœVΕW”nÇ@¤néB5Þ(ö žX»u¡šÈö×G¶1´ ƒ†ï ܉®22Ä…nM«lpuÉÕí‡Ø‚ÛK ;ÈI"}‘¬˜œ&-§äû2kÆK¥[¿ÄÇxY½8pÆ3žÝėijֈ ¡xs:mªÖÁ “‚¦¾¬Bo¥üñ‚ueâà'^!øSGàëø6¦}9ÞZ 2Ä@sÏp –îéüu5…<ÃÛÅ7x»¸Dæohüqà A²×÷üVî@G,ø‚½·€ælC Zl­•î F%ÛÞ:³:à7‘E`YhôU÷‚¼©É*–ÄPnû°+õš´³qžn0i$÷Yr¤Ò3]‰Ù„M¬Ý ç$ŽE˜MØŠyÏ š&Æ lž[Ïr ·ˆ[Â"² I –‚“¸dé#6„I`ÿÕ»FÜqˆœ¥‚A(6˜˜˜ Ã[á6ù˜Ïd³†opGüA<ë$þvåÁÈ®ò,n þdâs4æ+2Ûõ hY°9d·OÍfÎ$MÖ#«# ɲàµ( [ʲv0¬†îbdÑ&Y¢pøý¸HnÁÇÓ„=½‘ZªÇÄK6‰à6‚Iqóe{z€*¤ŠÔǦç#{£umÜtjC¿)Îlï´ócoç!rhÀZðNÈû^¦Òàò ¶:ägÝ™€’gªM¯b˜Û’—’Ý¿¤&u#Ͳ¹Í>ª·±œRã"lo8£L¨ØÞK'æõnt+ïás'òöŸ$ïÔO®¼ªÎŽ.ü Ä| Û=R½µŸ 4A𥲮7Æìæ>a¬]…OïY(¦g1”Ò÷=‹kÜî¶g14‡î{áSz±Ûô,ˆ8wÕ¾¸ïYÐK ç$wݳhõpƒ³=‹ÈV9)^ëEÝ7*Âï´QñêO´%ÄÔô³³¤ý°)F·oJÌšˆFð–eáæ(Óó¥ù,RÆÛ§á°½ÜySBD 7=ùÑÞĤÀ%&&†'™`öü%Ùï#ð>д¸ó—7<z6tGbü Η޷1ü]´*rrž½>ña)ÙU~âj6Xùë «/ŪƒO—¼³ðI@Í¢ü9,X{½_|S£brÆþ~z÷­ qߪ¸oU|d«b:¢‚­‹aè3¦#Ÿéòœ>ªŒ9Õ›4”ÖmÄ1 ‹ô´ª£Ÿ o¦1´:¢–iÈ<@ÔŒÄ#>«»7 ÍeŽA7ÓoÍÐC}–‰Òªpmh«Ž8Rì:z'ì …4qœFµÒh ’”  tc£Q_!x°iPÎi3|IF7‡Í(ÙˆwYÔºšŽÁ&Y½Ðd@kÕ¼J#–¯ †“¶t«óayšî<χ¹4äJg67ª')':ë‹TIôvk=ªÐÓ­‘iÀ?‰ºÉk–6ÊfR¶Œ|)ãLwL$éœÜþb4[åa< É|Z;‹:L澦×~Ú€#£yØÅËÌpšGW ò†1ËTk€hH†MØ A’*aXáóÄžÖub/QvB:ô~üŠj>S€yY,&}ö’D¼£b.:J5Ó‚éè¤1º'‰h—EždýÕñŠÄÕÙ°©f¸u:b]‡TÇQÎõ")Ú4œæÁ˜t¬ò²MCÈ›§&³"ê8zPFé-á>&êoÐkÇâù¹<ÒBš½Ú¼n ‹árô®.qõ[\ýÁ!H?Cð˜åFJ©#ÌI'¾C6U 2éÖž²$H÷=Þ‰¿FðOˆ@räš%vÉ‚÷ AB÷ ä6»ORè°òþÁ_YbÅÈA*MIÛ}^†à­IizHí§ˆºBð‚~5‚÷Á3¯:Ù»e½îož#ÕœõLòq1âæïXÇ;aè)ß‚½b-rÄò”uò· âÅWt•³ÿKÖèij'=/ïÛWì*}Ÿ¯7sĈ ÌÞl ÀìBkÎ_/Yo#îxà ¥Å!Tgøˆ_ ¨çá/?ÿbó9ö„ˆÃm‘ŽGŸþÁNèèÓ n4‡GRÓPx„àSóÈ1~á/pÓŸÆ/;Ú Iëi‹qïö”ýÈgvýŽ{fG:H‘åFèò)Ç^÷Óþò}J ¡à^SîSpÁ ôþVž^@Ù§Ç2@6ET}Ë$S gälûƱ-¹òÍ1¿K[&Y„÷n¸µé”ÑŽ l£oPÊ¥ÊJC…#OnF)iÝWЍ9ŒF\r[Ì·ráL-7¼^ÉÒ¸Û¿pƒïRµóorÅ+ZQÀsÓ»¼õ›la¬¢e [^ƒ¯*¥>ªÔ 9v-êK7-ë+°re=Æ73o“Ù”PzR€Êi_Å 7mòÀ– Óg|Å,!èFïIæ'{;?ì”÷Á&‚BüîÅme{:S±×éåìéDÈÞ/é[“Ó2=¥ôž—[;)×´i)s¬è¢á¿æ™©<Òå]¥„-ÙØ¸xqb^ŽÎJKØUq¨ð=Ugc-RxYÆYÏMO­àCP2x«Ê€NžˆÈ¯ÑqÊNhç¥<Àù‰uìuYM¥îŠÆey ­¡2kžt¦ãÒ‹jÆ­˜ •UjÍ$Új¯cZÞ–:6ƒv{Q¾ôÂ65¦¡2ž–x"=ŠTô}êšp§4rÄH.H{ÒÚŸ}¾¯t…Zù*d ú‹~º |sÒÈaÛø+hÚgmE…¸ö68·(ÈÃÁ7Èú4-º±æmg¦v$Rµ[b¢l¢½'ÑLOß4™¤·NR7§Ý5ò$¸Íü¥ù¶\{xÔÁj·"DÝtùꇔ¾ª9Í ucŸ@‚em·Ú'Ž7ÿYTÚÞTaj„"-’ ÁŽçh: ¿¹ë•HÄúgŠJƒHý¡1úèy´‹5É^91þ°ó?!p‘…endstream endobj 277 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1à Üy…@B¥v‰XÒ%C«ªíˆ1C2ô÷ „tèp–Îw'Ÿy?\gðGôø¢Æ:iñkD‚‘&ëX+@[L••‰³ Œ÷7ÞŸ@°Èìü®fâÏ“8—U»‡ÐkZ‚BŠÊM妑1’‘ÓR Œ¦:ÅEV,þCÉÑ\⸠¸ÆH.•¦¥I.`ýž >äl`_@Siendstream endobj 278 0 obj << /Filter /FlateDecode /Length 3201 >> stream xœ½ZmoGþ.î7,t•FÛy)®ýÐ R$@›ÈÉ¡PmÙukY®ä4MQôþú‘³»C®<òJ¶q(RÑ—Ã!9ÏKýZ‰FVÿë>—“/^k_m&¢:›ü:‘éÛªû8^Vß„5Ài¢ˆ²::´ÊJ©Ðh'+o}µ­Ž–“wõ7Ó™hTŒ!¸z1VEW_)…ð"pò…} ÑÖ—ÀÕÁKçêùt¦µjb°õsàªàŒ õ:É aÂ޾XÕhT<@Y¨ŽN&õ§G?OÒúF0“}÷®þºŒtÊFXl†–EïUý¸]ÍS?›Ý(¡<,Ö°…Yˆ»qÂ0ë„=¿N&¥@()´ õñT9Üz¨jÕÛ(ë— |LÇ ­KÂ2 ¥\½A7§>×/ûýâµ<ÊȶqÀ}+‰z4ùB xFGÛÈPÏ(U)ˆO#eµ^To«Ë‰mœ×VU!–ß¿Ÿ!úÏ'RKÕ¸P9­u#tµœHå"x&s.&oH*ׄÀ¥:JÝ–M ñuÆólŠ’)­ŽÞðm6=Y-¯¦`ºàõ¦RSôlÏC®XÛÆuKF&™ä‹vSS§+‰Óx‹æfްMpí¦œ q\*s˜T7(É¥2‡¤”ERÄ!©Ã$Á1!úÊFøtÝqÃ$™Îd´±˜Î”2ªNl¤àÕØÙ)˜"9†‰R‚ÿ`³RÀ&Þ’&€ ½ºÓ_´Q¦ÝcãážM¤…¬)/È(؉„“¬+¥]•ƒå&§m°ÆY¬”4\G µí!šd£ÁÇ1%O ÏiçÁkjÓJÄ@Ç·Ò 0ÎÇÃÁéVè{w“`M¿ b`ˆÿ_­Ïöx \wÓ2®åõó}´Œçwl„÷Î ÜÁ±¶±ò‰ù]¿¯_\^/ÖÇ‹«ë÷ÓŒqââëŸO”•¾ñƒsœ9ì´{H±%ã”ÊQˆ8LJùˆÉʤ2‡IYH6ؓʜûà‹T5pñE#ŒH‡02S‰öéõLmÑe€:@Â=›fˆ3Ê€‰ÒÜgv{¤„3Yº‡¨ÀÆ›8³‡ FÅ„8@‚¸¥m!†3Z[¬rÎh+uìg0Í¿‡U>¡ÙCáÌx`ºzÁÃpfü©}pf-w  ×ÙwõénW Õ‹åÛ©œº^-—ľ·ÃGQG;àxB†PCVr©ÌaRP†ó(/‡¤ &†ãÆ¥ˆÃ¤´ÒC4$“2Á ±‰8LÊ™˜®M’Êœû ˜†FÄùx£BÒ€PÒ{‹´ji—;þ)¼\F¿ÓäáÈ¡·8ÀÕ(J÷ŽþŽ=„PÑ·ƒõÙ`@? §nÇ)!èP/¦3m ÿ@:|9á]냭Ÿ¤±Q«e%Gý8AJ“h yû;Î_¢´ Ç+Än×´!ªôåt)qô”V ±HAk¨ë Êh7£I)TFó!Î’²æij%ö¦Ó’†?Ô­pzŽ,#TÐ çC¸¤“LÍ«ËÖ% zÍ¿¥q²ƒYR1Á•‡#ªcbg÷Yné%ŽŒTšט™4;«›Z þ `ðEA9pÀ¦ÕjD¬OÙº«5òáIØSüÓÀâöI9t†ßjØ^žºññ×Ef Ù¡°-xl“@‚|·;ƒÕ@ Ô> Ó:¿vû¦ž§Ý¹pîñ)íIAk<ôO6⬘€­îŒ™¹JcÉ–žÑþÐöYoü °°ŸEõY%tŸUZ;Èúy›ŽVºî¦›àÈ+æÈ9;#óåÖÌ2ÚwŽˆëƒToŸTRµ„ÞXƒÄGàaÓ[˹IJbÑe„á‹l$4d ¸Þ ë[›‡¤ ?ç?’_/øü4ûµhZ7€¸d”̹p5ʼn– ÉŒL¶eáa(bˆêŸd–dw R`Õ… S`[ÜLùnsÜOú ®¦TRÝÃG><ÛQÀH&s­gíÜ ÈߤøxqÝܧ¼^±´mP$)ChÆqÏ@Y‡H-+Èn(Suë¡×ɡΘX=Y¥lëW‰ëñd¬ˆJöþEä/EemƒT¾¶½¬­?o1»‘¾?]K¹æOe’É.ˆ¼$òšÈ÷Ói‹ªÝr U¡Š½-cO·“!‘?›ì1–d,Ñ/¦ÞyQ Ë·Dω\y]Œ‰,†ý¤˜±×œ[Šõų|{RV–CåC±)Ú°)š~\A!á”ìÞÆŒ…œ]Ÿ ‘éí¶L¿-¸£†öÝÐ,[ÅÞ19Eä‚H悆È_ЧUGpq ½îŒX¢¨Iøž3‹\Çw—StF\±Ã—™¼ójèKÌÛ;ºŽ%k9±_ ¬‹Îg9u´ø‹-yÓ©‹‰Ü½3)Œ§%öóôaK<œ§™s†ào°›/;‰qÕØfl1ñ䘓÷ÎKÜÓIù^ìðÒŽ®Š0ó-;Š ŠãÉH…®‰@9õʾ;нf‹üñöKŽÝ@Šy; ŸÊ‘\CÂ2 \#ÝãÌþ`5ÇGâ²âšAÇ\`¤·d²å~±,P®ßCI`6ì¸î1Àÿ¤?)žDJv4ÙŽº]Àw+²ëøb7ÈyJb›bäX<Ë938Åù¾)î˜Õt Åâ)q×ErŹù±?ˆü­èõrS´ “û]q5^MäŸDþQäö­õÐÑ}kÝiM{Aä%‘ÿ8d÷ómÒ6é877Teà+ß7e@µÅ Kõ²R©|»yR6z)–I¶0«2XŸi‰GÄ…>ä”bÌnÍ%‘oɽåÆsPedÿ^s,l,ð,® ìW?mQŒ«ÓË<í8wÿ¸Ž*óE{kAÛÇoF‘9ÔxÆ[Ñ7ôý9‘gÛç}YT>·å÷å7Å ¼¿L†W>Ãâ!ØAÿ(£u÷¿þ §ŸÛwþ\Þùüšÿ)]m²¥FoUCqK³¨Ñ ™‘²¼¨6ŠalK^Êâ–Êû@7©[u¶ûûìƒÕÓ7VRpY„ýã'‘--¶ñÒ‰q¿C>°÷–“do±Wº½½ Úïcðv¢ÜlËV*üarÎáÏÚ3ÏYª³@Þ‹îMf-Ùoþ"Óendstream endobj 279 0 obj << /Filter /FlateDecode /Length 159 >> stream xœ]O1ƒ0 Üó ÿ @‡.ˆ…. ­PÛÇAp¢†þ¾$„ÎÒùîä³ì‡ÛÀ6‚ƒÃE0–u Õm &š-‹ºm1–'.Ê Ùß•<Án sð‡ZH>/Õ±ª:M«WHAñL¢­ª®5¦ÄúO*Égsí ÌþSIÑTâ¼ ¸…@sÓÜ$°L¿g¼ó);Ä<[Scendstream endobj 280 0 obj << /Filter /FlateDecode /Length 5179 >> stream xœÅ\[o\9r~×ù •YP)ýìÑ£5„f3•2fÖौœñB#ÎÊìlT1“ì<ø VÏR‰SY¯ õ$*/ÝìvÖ…ˆ@ v¤Ða1N4ѨY‚µÃ.½•ÎÎrp$f—¨4„3+‰YB¢R`jRŸUæ¥u…Ê͆fÔVêBåf2•“¦R‰ÄËk?]{UÐø¢ÏÎP PE' Uê!*#œê©Œ öŒJ·¢2«7Õcú>! Yé ¤3¶fö !H‚RD $¸9âdHùÛ‡ÝÁ`›Ê†¾›è8á¤|‘‡}°sÐRÙÑl 6øBÀÒÞ®×…ÉMT2ŒV†o‹e¦—øzPH¤¶š\ ^€ÍãΦ#³;m¡JÂî”ró8¼¼—sƒ& ¤ˆKàQí C.„<:šÚÁhœJ%'ª=JKà:(FÕzP´¢ª=ÊX ùþj£Zí™K¡Qý 5¶ ‘Ôøå$ö㽟5¬diÙÉ´¦«ßZ&b!~’4 "E¯dƵšÖ À¥Ç€Zdލ`>Bè`Z©" ]2X•ªôp*ØP=[Ũ–žŽ "8Uéé¨ÊZ[ýeNò>à%l†Ö¤Sv{ú *aÂ.(ü¡øšÿa’n:b¡»Ú1¤,=^Öíõrd}N)R „ðÒªlïn^]ßíN¯wG¸i²u"šéáíå»ëû´‡&òb^0iøô/w§WÇWoß½}øH–©Y–/uÜy­Ï“÷öä]0vÆ~’÷†‡†P¼ä‹‹÷NîÞ;"þHÆC¨D“ °xo8?•Ç0;™½·O.6@‹•[x@œwmE Éëì¸p¬N.ÜÓLpØXWqáÙ.8¸Ç› W’ú$9ÖÅ…w­äÚ…;²e ‡ÜÔ¹ O’è\øâœsá©ÿ÷.<èÝ—ûñ gÛœâÙÞ•ç2»EÕü%vcƒ}w¢×€f@ÄvþE*èfA,+£*èÓø"m!­[¯mí¯UóÌOhVƒBÉBÀD NnÔ^ÀÒªŒ“Î*sr‰r¡‡ˆrm€‘ '’‹(WP,\‚SåÀ9¢‹9Êšó2 -ᩉ†,¬S˜ñ5á˜@ììð² ÀãkŠ˜…+s !(j ‡#Â~E«ˆ:Ϩ…O‘6Rèñê‘X£d=šbtðHÁvˆ'Á¥P:`Æ3N<Ñ[o( (0Q'ºð)ÇÞ`ë2Ƚ]`’ƒ 2T… ÐYYX%”H¬§<¡„h´[6Ø@‚È]­@£ÒNôË‚eà¤BŠî^Ø.RÆúa‡€ ¢aM ¬ú~æéüÃA”}!ÕrÃ/P!8ËÕ¢0âÆÈÁªd€!ñÁJ<¡£I`í î9(ì¦}Zpy&L9ä(Á˜¼2aŠü™&­T‹õ¦(iwÁ²Rtüp)æ,ˆŠÉ¯cºlá•zHã5áuq<.¹,…̼øK6Ø‘³,D^¤œRÙ @|†|ª¤PKê5£D1·øo³§3I‹ß‘•a ³ß!üQ¦ŠôvASEŒÙgM&AΖp_½NN_ åÊÅë8SQÑ{ îuhB³ö:1Û(æu<Ù¶•×qyF¨”â@á(˜<Þæ*Šáºƒ…³R™&2hQ÷½DŽ~!×äY7!o5q ÜÇ¿²ú‚ x9ú"ófD„e €áÂ6Âà y¼Ñ,ˆB<´Ö? ï¢ÕlXÔo #8c GŽõ§³‡;Ÿ­\:Zð¼ÑRFGN\zzkÉF"ÚnDéiêÀ¨)€òuéùÂÐZ“ ÀdÎ#tD|[«9øÂ¨rÅ¿6ح€ŸÁ™kN/Ï áÏàjðQ1ô®ŸÏÒ))³X…Ê˳ÓÍû=L¬NM3¨~¹ûöú6…GÀHÁß7/þnHùÝF[H¯¦ÓM4 TŒçÆç£–ý|FÉ;‡ÆèU=¹¤›Z“ðj¼LúæÕQkM§°¬T«¥ê©‘rŠ¿¤Ä!‘TÓ ¢.\¾'ê`¢ê¨s‰68«§{"†ü-%{£‚ni>ýMâ—‹¿G¢‘Òiox¡÷*ó&&ïÑk .ËR›Ng‘ÔI%‰²Žú@ß#üÍtɪÎiδ+ÿ˜$£ánxïÞ’žˆÖÇ\ŠÆÚ§×iÏ^ÒÄw¥B-MWÙ>^Ò@'5I(QØ(»UœnÀç…Ç–hùbîuȽëøÝîA`¤°XQbGÃ^³­. qP’'쩆NeZ‰G•áñ–„ˆ8Iw¢ý Í’NçÔó %ÔpΈé—êþm•íL¤òb¢æ„æ ‰Üñ’£D )”Ò!3x¾§¡3&Nÿ´? Ÿr˜éëÔëyL¿kÍg­ùŸSk¿6ÿÔš†Ûïóê=±ÅQ6©šý²-J¶æÛútã([3´¦j͹5õ ”%õ¦ul@,i¦ˆAÐ}Ç:ˆ._}‹tâ@~ó&~6¨M&6äþKwŒ°Ú¤³•Í‘4æ2n\Ql"P üjæDúí†ô4գɾ%¨SâøOº³¡¸•Ú¯[{­œÅ`\³•‘ˆa¸nQó{¨4BjÄØ‘þúqwy²jÎkY[ˆ|õß7“tňr#œ6½i¸lˆ¤­¨'&Î E†5Ça6 4²˜™ÉÛ‚TŠÂԇÌ9 ‡­R‘|êîiƒ2@üd(ÓIÙìVÙêX˜Y‚ñkmX¶Jk9,‹¡¸Ô–ò?¿¬Á[ï¹é4ieåGFg^­|4Þ–ÕY)µìÍ<]a ¥ùd @׌ś^ÏpR®@’1F> µåˆ˜«§xZH•ÌoC3æÂ¹’ŽïÒU(Φƒ_Mƒ’*é9‡'é¡ Êoîs™¥“{ÂdDLiVYó X–¡8nz04cC³ÀHŸM ÷µÅÔÔs<æ^îN_¦73dÉPÕC½Î2CÆ»BhHÝ£âXììÄéMo)|Ë·mË$@ƒ™Ê¢|5}¿L§#é£ énóY3Ss­I6jLCzLLÈ`‘™Md‰É£ÂùXÒ—i‹ªÒl¿¯Ò9:œ/ʄ騌+›Í"¹£žQ¤ƒÄÎj¬¢„ºÒ}vYR¶ Å«ÿ¸·6ÅQÓ1[z£ø|Ì!‚n1&Ò@Sn›âaZí|¯])äÞ> Þ-‡ŠB%Ëç`ÛÏMz ”ìÈ ÉUÒ2D?ÍU&w² _—^XîæI:ÐôôªÉ¹_P%Î®Ò ±Éã~¥¯å“ÛtÙŒ6 _òªy›ûlL3ÞK†ÞzÓq¼jZÏ&gÞò”ŽF½I¹E€90‹r¯¼©#oû`7YKL!Õ–K?Ô·'Y&³6XöAîºxy1‹"ðÃâ|—×äƒ~š;õ” ™äN“”`'Ìü43ªŠ,2ª¢Z\x’vQ¤³ëŃ-VÀËNwû¨/'0Q/âai öʮ⻜´4«½ÅwÑXÕ“ä +õ-e |ŠSéÝŽ<¯3…Æ$_3ûù»ä².=cÝ{à³4¼÷¡úäqúñm ÎO­yÕš×­ù®5íÂÁMÿØz­÷×­÷›ÖüÖd?oÍ7ÃÙnÖi],\iþÔš%]JmF|lÍ›!‹w­ùqÙF˜~ÕšÛÞ·æiÈì~Ød´·­Éø~Õfûä>¯[ój¸ÞÿnÍÿ2ˇ$â¨zœvúûâ~ n±P7ôº5o[ó3¨Í×ÃaŒYÉQËtp Ë4­xü¡ËFyîê).ó\TúÃþ Û²£t©Âlf>µ¬˜ÍÐe©g‚mR9+öQ7]Hnßr¤ýÈÞÜæB(-é!/“ŒWݼPuo»~ØWP|kkJ?EâÐÜ®(§ë-µ–TÈ]bcÉúê5¤œ6o7’ôƒí ¶‚Šç)SV°¨A™‡›¹x›Ê£Á+c]⑼‹Þü³ÛšŸ^õ“càèbí5öÖw\$F™¯’!‚åzËlèÇB‹4N¤{Äæ7òajè~wÍÁîù؄܀s•È÷Q¡CƦü+âF?Uˆ1“@oò"Ô:®óVóCƒq‚ÁúªYÝ/GGsšû­ô²XÆõ{™C Œ4šð´è‹¦‰i~ë®éfÃÖ€œ_Œ­uƒש,ÔRÀ¯€Îƒwö¤êø¦°  ŸË3 Ë;Œü$3‡Q6Ô;Œ5JšÎ=Æå¢?£ŸèžUúÉczÆ©§ï’Km,Mò,N‹d´2 '€(Ûûˆ|h‰ëÖË”cÖÂÕCfòqûÜrM~ðìv¢Î×\o‡¨é­všÃ½ïRÐFryN“¬ëí' é]@—´œÎ_X£}ÒyšEÄÞ~¶I^ÕIº»øC»Ž®Qã÷Ùf¬^uí>#”ÑA™¾m¯GÝç AVœtã½dƾ ¥êæŽeÏ6l¦L*+ûPzÙÝ÷*ãYù²%žÏ^µEŸ²”Ä4YÙÕrV8ùܧ ë‡]T-ísö[ùc˜ôip3ëôÄ],1AzÜÒ»lÕ¹s6í£Á(µ™g׿X©÷ùÏô3öŒJõB޽ĨÑ[±jÎÇž0½ŒH½L2„´hz/ò™eŸêöë¾{/|]º»QíeÉ C,^'íÄÉóŒØ¹àžÑ2Éå(Ž~G¦x9[†ñ³ï³?ï™Í‚?(}Zœ±[em«%ºË»ë£;ö(iý:YÁ ÇX N„`fÑR^AÀ>üö”Ò¡ûG~õmrq´±ë¬I÷W9x˃œˆ¨Å»¢6ÃS 9¾Ö3“Œ°{ÒŰÀ^&¼[ý\ºƒnºU=¶b侬Avî¶y°«1çË›­ßwú¡£Ç’’™3é˜:}|¬ T*|¦vÚ.‘ N—ý2 ’©ˆHÜTuÌš’uN©R°™^•ŠäËWO­jú\ŽVšk߯ cGô-§'B7Í1©-ÞÇ…IÎ.×dÖO,—ý÷ºgð(ɈP†›O fè{úï§ë›vÝUâòKqú t˜6Ë\Ÿzõºæ#§þˆ8ƒMv×w#ýûLë:ËtbãYÌuþ}¡F…5A=büt¾‡ã฼±)oZ†ÙTÅÝÇÌ.· «Î±¤9®KïòJ¢Yn¼ßíë6§sà)K¡'šƒW°SþQ¾¦Ÿ½··X¿ÒÈÞ8ÝѰi³âèT~ùz_/ ÒaŽÿže™–PkÎO)ÞVA¬~mP~óW‰ó=Äendstream endobj 281 0 obj << /Filter /FlateDecode /Length 6835 >> stream xœÅ]Ûœ·uWŒü ‹<ÍÞÉÇ;Ù¢nšMì5„"€Ó‡ñÞ¤XÚQ<’-å/èŸÝsáGrÈÙYÉv‘S~ä!y.¿s!÷ïËV],ø¿üßë×Ï~ûµ ÷‡gËÅý³¿?Sôÿ^äÿ\¿¾ø·çÐÃâ/Û´$uñüîª.¢º.l“qÏ_?Ûh{ùüoÐÙÙ¦s [Ÿ|ðüæÙ7›?\.[½(«üæ6Õb|ÜÜ^^-[“¢q~ó~V)©7_a”bô›—Ø#¤˜t¸2FÃ?Âæ¿/£ÞÂgFX¶.¥ôæ{êº,6nôë5|_&Û?àfë—°Ñ—ÿóüOG ¼Z‰¾2n½gÒ_㸅\†Y’%z¡éݼ…æ’’‹ió;‚ GtƤLfôÎmD…Î?ÒÈË¢aIL¨ ÚÉá^ðš]Ú\óV%Üf/Izsy¥=,Êò Ëì8¥_–˜ÔæázXeph¶†÷(zc- àNnçµ8^6¤7ŸANQ3I¸«š-Ú=ðÝæ¦v•ýªv`r ̼¿[OGÁÀ…ŠWõüwH±WΚ€®ãñA:/R›¸ÕN­Ì÷ø÷6ôµùemþ¥6¿ªÍßÕæo†übᜀ¥×i¶…Vì [ Dxwq¥ü6è°ÊA4(°`âWŸRÒÃ%À&y]ƾ#Z€»í_nÞÖæûÚ|]›/‡þºíËáҜۆ¨×Ùwx”‹ɹ©,6$[mcŠRêó<_Öæ_jó®'•vûmmfR¹=!Nd5O)ÙvO|½(ï;^4)`ŸÍ·õWÁhù³$ûú#E»ù®Šû°J±g­õz˜¨}’J‹ð‹Ó$Êð•Wa³#ÍŒ"1-Ô)ï%[Þ Èp[ á ·sq`ð_’öÓh`Vu³g&ÒÑèz8ôìàâ¼7¼þ¸(Ì%6à“éb9m‹{ˆŠlBÔVo> ÒÖÛÉÎÞŽcÄmT^þ,(—B|F4Ùf¿n2!€v­v@R«m$!ZNB䄲…íY€=òNÐVÄï<¯Ž‹582Ù˜LJU™uÇ磉›]/1Ùø ‰™ÈâG}R»$kŽÈÛŠ)Š¾Ù·w—Xy÷meVä´OUìHF\Z&P¡%mÊJ#ÛRZ;ÄøsPQ Ðâ‚£ÄBªÖ %af1n£‘%<,Ôï$dARz l.‰È€u ]/—úwˆÀ1»C>¶FWñòÀ· þ!o(D6¼Î*y<7Ãm¨|ŽÙÌß›µúdÁÉ eëctpZAûM‚„ìƒÛ0䕤5Üߣjö0ÞÔÎBÖ¯~¨N—ô”¤–e?!XPe#}¨ÎØ'¢Î Æè¢³˜Pj’3ÑY†Èož¡Ø­ðQp“Á#5û’–íASeŸ•Öǰ´ü™ JþgòN¾žNìq`éýÏ#ñÛ¯õ¢×ÚCG›.®4¨mPAÈP›ÿe'¼óØC,NÄsáoòbM޹r±W¤ö2ç¼ÆñI«îZ•M´[: -.ŽåAñ\×C‡L?;¬~Ñ„!×Ö< àµ^wÁtvñGü}¤5Yšµ3€D¾j€ÆžGF*Ç´ìš‘þêp¥ 㯔ˆ@7ç£ó+Ù…6Ket•õÕªE–)Þ~Å"û‚­&5 ˆÆ›‚wÊfмØ8d/§~YÁL‹¯VŒÑà$È`š©hu(Ç 0<¤&Ï`'”Ï«ª´ºp ˆ›rxNo*‡ï«Ó;Ãö‡¡;;«0—#3ѵü&­0K`¿?Å?ƒjãµ²^»ßg£ðî/Zû 3¿Îô+{†:ß]‹½f©rp†FîÛ4º¾sù‘Ñ£'°haùme™¡¢¼Zõß•ÖÛÅÇŸ@Q°6€õ~ÌJƒ«ÕoA«4ó3% Ú€s²IÀ³UwHeîéÃpŸ {š  íÑö¥Ó²­ EïÕ)g¿Êî[V¸*6æ„‘xkñ ÔQ9»`Œ¨ÙHí$ê@úÒÓWN’ÊNNþær½Lº“Â4áù¼ÚÒ„$Ö—™Š6âÑ0“D‘,ˆŠ‚ŸfYu›+qx‡#yý˜´ÄoVeº`a_ïVõf7cKH‘é=i&X£Ò+;ÚÄQjC¾¡@mk4O«9ýÞ˜R@Ê›`+J'Zh±*³‡1 P7!¨ƒªRYáZt0Ä™xRñ°¨‘VD²4’ÐÊ×î;Ü=Å)‘ª_Jï&:ó–E‰É&#ÂÙ6Ǽç.Ƈ4]ð¤DŠqß —9X%ðÇÚHÓžNqmôˆïÉ#m€›>³‹¨aDñÝ“t„Àr¢k ‚fíÛ7áo‰»ÍÎ48û± ´ ‚XcŠ1H: €O*ú =8 ô[ ËØ6=•9ŽOÍ·ÃæmmÞÔfôs{è·i«uÉ(æäE{Qa6¸Æ½m6d€ñÖZÛ©Ç[µÊÙÈM7l¦²« P1e”eAF7è’*$g®o±ú,³³’Êü ,Üà‡hŸ† -.¾‘ßwcA-E"*Šaºµr@i5hÎ-P8ȦÝÕæ«Ú¼­Í÷µùº6_;ŒslÎoµð?ö`âF+ƒbC"DÆE3¾Á ¼“^òù¼sªÛ5b( Ò:Í›JÖ÷Ã…Ý Ws]›OÌaá*ujx{V³6ž®+Á’i bn«nÔeu·®/K­†Þ Û˜àù13΢e»ú ïÆº°ËÝ‚VÃN˜ÇÇP.`«'Š@nË¡}‚ñÓó¦uÛTó…AÇH¤sÌ Ys2„üqô}žfÅ9](yÖ1Þ yôiŒ‰™ªæŸÈ— ÷a˜ù¬Z¾6ȃ?ÐÎxN/32×M¼V ýÚb&î©¡y$ ”ˆ“=…)<ûFŸç vnÌïhvʤíg@ SÈ9Ú*–òÝîU‡)ÿ¡ íìš@’Åîx ÍÆwÁêUŽw¸VÒ¸>ë@@qåPN1Â¥£ú©åôLKâ0ÿ¥>Ý’l;Kà])G§9è`¥Á!ذë>,n6~øs‹PGmN¶(ðèðg·à¿<žÀ­¥0$¶£ÙGÉrÎ m‚kz<ÜæÃS„kŸ“Ð|ád“ƒžì-†`‡®Ë|ÉäÁ9m>»æÑk{]@Hþƒ»z9,9; 9ð‡XþǺ˜—¢ÇÛº˜y„5^¡¤`2ŽÐÉúÂÃ)ë€ü¾8ÅùïuÙ< n‹.§øðᆆýÝ\ïë(‡ÌÊ”}¯$ñ§‹¦ îM »óÐov—101塳œñ5¦guã£Wέ“äŸé„ÀûQ4HþùPW6â]c¹tñüËgÏÿé›ÍÀ+ kÂÌ;û‘{¨\ÑôÑN=â—&ë4r¡s)HŸ ¯c±Ç_1áÚãv÷ìñ¥@É©\ðx¥ÈMžM9™|1Ö¬™r*PMàŒ¸ýdÜwà4޵HX=½ŸþÓ80Až…¼ê°ù#uöõåW"ö{Èa)„m#߉ɷÑZap^¬ã¶áE^tâšÒ‚5ÅÏÞ=ŠùteÓ& z®Î^Y_WzL;'™Ù`I¶Ì˜Žp'ÌÛªB–\`íÝó¨5~–(/Sõ–j,§Ï3+pèæ×‚¸>`›ÁÛ$ü!ØJæN޲»áq:{_ƒW暸ôM¬iGÄS áÎâ¥ãát•E] en¡þ*ÀM©Ù é6Õb°Ì)h*³y>:LŽg^³qªxõj]Cë1X;º±RóËÚe•_Õæ×µy?ÄXMUcžç…°¦¤3rThÀ’áxŽ2N!ô>é°¨¶M·aÂ1àY[ÂÃø',™§Ê°G“ÓS •— Wo¯}‡ñÚ"×óJŠ¿®ÍûÚ±%»oæžá¾ÑјBª ¬Rê%û„,S`u©ªª/R6û¦•.â$grÏ£‚m½ apBTZpƒˆ!Tk–^ •Ñ¥)±;šæs a9Ô^#w–ˆ ¶æ -CðÁɼ²°V]$€2$JK3²úQh0Ç~ԣש&.UØ*¶êëV:½†r’êR7UƒÝÖæ¡6_É|X j2¾Æ"¶nèáíQñ ‹*‘›/Ð9÷ˆÍŒs)PÆÁ¯sJå†bø×3–%I²:e¬ªÞ3y]šJä¦&aQÒÌ"¬òC•ï{&CyOÄÑHâ.Y #ƒu„l”ÉÆóÔ ëz<Æš§ñ’:¡ã±"ð'q¼‡…ë¦ðà8IHHOj¢÷ùËÛBÝœ‚q€ö}á=ülsÀE%ÜÔ-;•x!ߨ€µ*GÞ:™þ`%ö´[Dg×½|ΠfÞmmÊ}ï£&­à¢nE½îÞ¿Ó¶—Ê£r ½õn){ý¯—tÙ&‚A›ØÙÎã¶Á­[6ŠkAÀ–‡-¨êá·è6a8\×´Ð%°ÍŸó·Øêð.”¹¿ÙüËÉeÚm´ R+áª>x֘̀ËÁëýøÐI—ø`g²aܹY£…M‹®±×‘ŽB·Fe“εSúz‰Á²ñZ‚ÃîÝIÿ¶*z¼BÕ}m~?´]ï†Æm74 (kª^»:ˆ" ² BPxÑÑ݉Íߪzm3è¥~à-ˆyR}‹Yô?3‘ šÒ w÷(l}¨M‘T8ÔæÛa‘ƒÐy:Ö\u`ÐX¶ß@Ò6|>2Cmmäi»=l¯te°ŽË]Ñ3:\`Ý]¢×ÛtøÉñcy’O›ßÖhÈÒDI%+| WA߃ݹ’÷"ÒfŠ‘ïŽ.¾àþƒ}3Mä_Ö÷æ8D;ܯ/ËE#iwDÝØùu•©C¬UÛÖª¢À¯ðßWûž?XuTjLãæ§)µ“A–µÂ¨Å›<_›÷ÚñMBåmï“ôV«’ãRäëÏ-Kt– áUÍ–ÀÀ~ ™ůÖº¤  ~¥–ò›Ï+RÛ2à«j²¯Ùjø ò Þ5j ÔR󛺅DýØ eÁ¥ý»¾w¯´]ð©Øà£"Ò>ãÚ°»Þ=¡æûÞ¬e«524Ȩlϰ3¦ú5'ÊSs!_[¹½µ’_ ô˜eOJ™Þ$±G¹Œ |8`wµ'Û%>!b\/…ᆘêõô0γâNNMÀs~c¥J£%•õÈ¿Tµ…¾§X ”ø¡Ìa!ü*3íjóû¡_¼“~qi^K5„Œ†Í’>Ž ¿€³XE‚îÙ.»à¥O1‹··â±÷J¡Àcï•8²¤½E6)£×xÛW5dïøWL“ü䃴 ÇɺÔ#f؉ \o£+p,,¸‘Ìó¢JÌ÷oò†pßÛ²g\)ºÓ'ïíáyU86Fÿïª×Nw ‘åîá…¡nàjE Ö׫Ò?vD1¦¥Z i3Q7´ÕøCŒ(Èîl‡ŒŽCšÞΞéèCš ñ iT "Òq»wÅ¡nËc‡š ÆÚúÍø*úì,EZ¶ô•õ(|ÕÄ&KÙuм cYlsÅU^SœË7¦ÞŸCH’Åí¯É ­Üú?T¼"® 3ÂAÌ  ­in´BXâ3m2W&šq GÅ|#Üo•+zQ"wÒ…/q\V—D—6¼DâŠ^ìhOÑ¿–UµA°•3ÃÍuƒéžæËß¼F™£V·ÊÊSQ"^n/5º²t.Ú'oHõϪæiåèèÙQWoõcˆÉ¢*/éÊ,l%0ú+Ú­dMëÙLŸ} ªôT¯ngºaGßñÊ·(WK¢wêuõá˜bíæØsëÙö õ>%˜FÞ뙸ޭðfê§V<Œrt¢jõnmú$¼.4¸Ä¸õ­“Ó+Þç[ œ—-©1è&kl>Þáùp)JUìD˜›g‹,Ô”´õ@¤'7ÒÅVMÐ"ÞÀ¦,æ¼ìÖ(ŽÂëÚwÝ:z†AÊk¾ ‘b+SוɅ~FûutMçò³Ñ:UK«°¢¡&±NeÐ>%{F¶ÖkP²%šØ<IÐE×qÀcßéZÜd¶ãt,ŒÖ>òVÎ3¿CbÃrâê*Ž€ê@¢ÜrFºI I¡× E¹lãÒÈ$*²Ô£Ü˜û”úfÎhÏ•#“¨ÆÄ}—–eëRý°않͛)¢¨´Cb§ÖjÄ´ê†ýY«AÎxF,`V§¨Ñþâ>_%¹*ïî!)#Ý|8‘™0•%ÇgÑý¥ê¶Æ wOØ ™ø¾Kâ-’ TÊ[þ¯ª´}àa*½‚ŒÆÊó·°phtÚ>â)€HþåøÜÎñ¤‡’'L{:aùÇ7=4NŒåVì.ÂDmKÅo2ùqe )ýÎ~}V •¸ËuüÊ•‹Ž, Fü\ß 8õ˦€t·Ha8|êÔUKøö‰ƒÍ8{ò2¼þq *¼Bkü õóK\ÛrX"Õ? X‚÷´ÆãÛÝ.R€NÆä©¯e¿XVü¬~ý‡Š)ÿшmã±2ž­N>\6€vŒ qv¬Ï8ñ&æý[Ñ[¥u Â}ÚfdÇîÛ^\-¼eI Q;ÅfšZ0ˆ*ï¤<ËÅ×ËñYÁ4¾¾šý>9-¤žßjÄñÅQÈÈÀiíŽëT Þó3‚ ¥_Êg;.z€õn¾«`æþ¶Š[fÖVR¦[‹‚ùë÷Fdÿ¾Â©sÕ&è‘0v¬M"ïàñ§tê预@¨} ŠuUÐ>ùšoB÷,ͤ —:•tüZ%ÓtB"pÕ¼0{$Ô¥5kï‹L_izSYÿ-ÛZ¨»Lô©Hä€8ôËé  ÍTqúÔ«w”±NFÚ–œ²&Uì8S\»-8G%²(3ñZX‡Ìk1Ja`ÄIkEƒSöaþ`Õ0»RŸ1ºáYPawþŒ’=eà«î !J&µ/=¬COÊÛÚ'KÚR·ÌŸõ×I[÷ãú·¦lÈÀôVI¹©¸>›T»ã‹+\pbè »6Z˜ù´É^硞ŒÙÖãnÞP®4 “,ê–/|à‹žÿ) ¯¨åA¦¾uzN¼lš⬗M±²5Ë®-O š ¾°ÖÝŒ¡§µkœ~v_h†‹q`w̲ü I©âxrYaö€!°¾˜‹y°^¢¯¶˜§ÊÔcïZÃPËøÿo5è“ÀIªÚ¥’iCb[€Ã)Úƒx¨Ä5‹`vxèÔæy‡*¹½2Ý?Hñè;ç9âkxæ¯.D÷ófs†÷èafNÚaârúûÞ¿«b%AÂà囇·×Ñ$b“ùó–ç˜WU'£¸EíÆŠo¢U늦dñ'"¸bÑGÙS¥[/­7ê¾´0íJ\Í׉×w×—ÓIµÕÚ·fëÑç΋(ó^ŽÚZŒð¥­ÅªG‡`Có%hòй6bê´m&¤—SåT‰2|(~tùA<]''؉ZÏö‘Y>³Ø?]¸¯ C2NìNœ¶Y…lã åÓÖs/äãèäm鼗雲”÷ë‡ýóäÔ(|Ú0´ïÐ÷ÞïqzEþ̳m§{`ò¨Â ?Á†Y¸o‹MYŸ³[¼Œ_‰û~‚¸™ÑéÍ;½¼Úh@s.ùù»XWÆë R‡»õ­½rVxÁöD@GadÑ‹WÇDHH`Y‘$ñhˆÝL3‰èáì=1q]Íœx³_ÊG++lTÂV¹ákÍÓÄe]ùÂИñ¶<€ž¤¿ sÎ߀ ~2'âUyï·B múØm…lkx¡ƒ”†Æ[g]r?ð>E"d¹ÿq—Þ‡—ðŠA^ (Fæø‰+œ‹2/%zÈ„®KDŸx-;Ø¥Õ '¯ZpÝQ”à'0>Œ|=ªôðþä;ƒÒ³ÉÓI¶ˆ» €ëÇ®lòÞ¨é»õAjnœ¬OŠèzù)Œ¢8fæø{BR îò>'~%6% ¾õ‡çh —QyMåÿRÝåÞÊ·x«/üÊoƒˆ‡*„hŒoæ~[‡¤íuTgɉ®.±"Ýtéò•Y0æéüÎ¥£E4EaüÜŠ`¼W5¿‚Áìèúº®ž¤/~æçk6ùtmf¤ ÖSý¥ËŸç<ù+¡}´°:ÁW¬%ÚñÏÇ·¦¥ÃÌõaV¾ŠŸ5²s®¹WÍ…ÞÞÆ\Ÿ/©~‚ßQÎì“ëàúÒsÚ{Ó]·vMF±~ÿüÙÁÿþêØbTendstream endobj 282 0 obj << /Filter /FlateDecode /Length 6304 >> stream xœÍ\[s·•~g¹ö7°ü4ÜÊŒ×ò–¸ä½YU‰ÃJ¥ÊÙ‡‘H‘E”3/û×s.èÆ9hô´½Ù-?{p98—ï\€¿ž;s>àåÿ¯oϾúÎç‡û³áüpö×3C=/ÿ{}{þÛKø"xèÙå!›óË7güSsnmÚ¹hÎÇ0î² ç—·gßo~{±v6ç”âæúbØ…lŽ›OÐ4Ã0I6ßáÇcN9lî ×¥ÑĸÙ_l³»œÂæß ×¦èCÚ|¤o‡Á§ÿ¾üϳ`wnHöú|‚ÁÒùåÕÙæ..ÿrFó{ãa™âoßo>ÃXÞD2L¶Å•åq´›_ñlcò›o.’ÛÙÁŽ0¬f€µˆânâ=,ëJüþ#-Ñ[ ßBË .¦Íë qëió–‡Ùl¾½ÆŸ¹œ\ˆô±ÉƒµqsdJÑy¯7Mûýê;?È3°Þì`+p¸ohãö¼ua—bä-RãáîG7&Ø|äa!¡î“*Åà6·â÷¼Lã|PD£î”Í0ÒÁ•<ð08ØÈ<Þñ¦ÒÏÀ ±å4<¬#1` ·óCؼÁ/â0À<›ãGì‡cÒ $¨ÄÛ‚_šivê=Þñ(v0…Ð)$ c@oÊ9†q³¿»°##;ìÆ1RÞ|¨Ü²§#ÎÎÀpûÛ†¿£ Þj<¦åœ¯ô½gÒøa”Ë?òtpBj'{ä˜bÜü€+ŽtÓîÐâM•ÏâW7t&@çM"V"RTž;Ò'ÙúèÕ/o/P²F¸»@ns>î†82w}_v‘’KH«ú»÷Hà.7»‹mH‘Ýüû L`7äÁL4¡_•=Áž¥<ðÞE–È€cÑjÈÉŠ>ïêNïë¸ðí´ò­q»à³•¢O<úáb Ó‡ä2HñÜ»§‰ÒhÓ äXî¹ÛÚÝ`,Ïê‡!dÙ¸Hª™YÙœ7šch §{÷ïq “œ“rx?}åŠ>qo6IöÊÑp‡ DHjäB‰û³‚úå«>M %QQüj °¦Ug'g¡¯Mt£×ërde¤–¢B¢*Ú—Aó)â;¯ ¥7âJ2Ó–g¼>Cø»7ÖÉ_ ~¬†…&Žƒ·? A€áŒs“ |¨êk>‡ \¾åà-É&0ž"òMm ÂîY#PT$ŠIHd’~âƒ!FÇv¨|À¦2{ æ¡È!Ð*P_!Ír²0ÆCyE“Y».XpÊNYpR’f)£â\{ÒfaâƒìÇP¤5à{ÕϲmõÄÔÎáò¥#íŽFÖlýý¤»hŒ"¡  ^IÉá8­  ˆ´KRñíoxÐdÂD'•^´Ù¿ª,ÇgÁì+†ßÍ ig@ÛÐ4ßž]þë÷ˆ€™hW#(‡4Ò€aÏMFë½çQÜ¥4Išs2J_IêâÞFT³¹H‹°‡0 rå«úKhüÃÜãiFW&$.‘»¿á“1«rÍ”²S\÷ãôCÕ;o˜Í«´f¯«êU_‡¸*çƒd>™.²kcî&bÞBÑF)¶iô8Ñ›œUëgË0F’‹*§?°¾q~guSx>€Fr’bê˜ù CoäRå,×UÁM†0H]¨flH¦áhHŽfÐÀý~ó_•›zØuëü#„ÆH °J{»?¤Öz’)‹Š$£ÀÞ2tÅ`Ý•M"óÏMmIqMÃ8F”ùù°´²ÆñàJåpžŽ"­ØöøMY Øæ¿7!Pg ’†éƒ‰°&€ßáQ´æoÖ§-sÁé¦{*wt¶Â,Cj‹|@°Å,ØÿxWÏ~}/ŒBÅVöZ®JÿšÎä!Va/qzi òÞõlrë 8„‘E*ðÖï…¸°z u=V`äŽ÷@–Hت½6(Õû¬qÙè¦Ùy@ƒ²à½ q¨6‰Ò²á€×6/‰6ÀDa³Z€qì •°$@iàJØ„®ÄŠ¢Ö—>â¸×¼t?(÷¦,¬¹d™þ–ʘñ%ƒ1?¶N†c4†>>H‡pÖ ëþg—yáò]J øˆ½Þ³ 2y©Ñ@DŸ^üºj, M¿¨Ð´aVœ$ä¨<:ÉØ|èP,­þXÝ9{«v4ºûïÄà¹'ÙÍ"4¼W zQI9Ÿ`4w"‚âq½A¿˜é½ôÛR›¥Z|+ñCÁ^Ðú¤ Á?’Ì~½ç^—šˆÍ•’XúVÝ@j?ÊŠ@Bç±óxô*ãÜÃGîy$à=ŸÐÍô}èie—aœçjåja$6÷)‘d?T2kipøê;Ù eò:³€ñ§„Ç¥ähéwxDÅ¢,?ƒßLÑZüWÜ|M½Ñ{Šîa{pPçë#þÁ˜0e`cĦå›O¢{OÌ™}нK>eŸ Fƒ‡Ë:×ïB¦Íúäí¼sïYwÀšãæÕMÝ ù´ ¶á•Ï1(‰ßÀé^1)Á^gÆ¢>ETÿ÷åÔƒ ÝdŠA½Ùà€:A)öÕ½36+ÕÀ‰ àœ>é{R 1„¨º‹3Q*¶õÈ3.é´¦(_ l.“Èß,è¤`pì”:BP"cÕ\²—œƒ­—œ<µgu6Ák6‘ÌÁa‡ïynŠw+ø…×>«#{êž¿)eTç ¼šÔÔ—|S›ßÖæŸjówµùåE—…ðÌ4¶Žs!™¯8‰_Te¸ð»˜f’”t0½¾è`7kÇržyÐïñå˜üÐ:¥«þÆX“TBìmÝ×µ>™­Íà“~ëká·Õ]2-¢Ÿõjl›©11d$Jô;àq­ô¼þNÁ$ÁŸ×ü{¤ÇjlvÇm »Ýs/Ì•úòÖ‹‚ŸE™É–g̲N¡þƥĽ§±s&`5 ï.˽’yaí‚­Žr¡öjB†æ¨Ð‡aç Ìɉ\Y\ô0ßU‰œJò3änâÕ¸Š4^/K|Ãå“÷°æ_&Ò/´ÁÉÜðJpU²ƒÈVJÝ%#B_1Ž)»Ñ8¦RÞ HÃDdQ["®bÀíµø‘Žî\qV“×Ó/uf†¹ÑÔàeßZ–Lœ#Ä¡Hqo«Ñî™MvçGÔÒÃPÜE³Œyqbü÷¢Ýûõ£BJÍ mƒí†Ä)Rë¹ÍAþ~¹‹b¬ºÚtRŽX_N ½}Q³Ñ¸Š³KϺÞR…Ó˜@Uo„?*ct¨,¢ä¹ˆ0‘°ÔEB«µ¸®GÕ:âFPƒˆž3¤KŽÂÔ8!ãÙü¥Q¦sqŒ”RôÀL®¤/y‡Ú‘!›}ÉŠ_WÓÚ5 ë»LL›ß`±Q»pGlŒó€¼³Lú¡4ïjóum~ªÍwµy”?Û‹e2qóçMíþ±6U¿xèöî*në7ÿ|Ñ…sÀàppIWÂ4²Çì.Š•–YèöÛ¢€Âç’·U7ŸrgÏ µº²À5óò¹µ‚[f¤gÚ4=–í),‘PO-b|ÕðNúɧ'Eùh@´ù²šH£žâ`ܤ"¾yy÷<Zn¼åÎÙu£2ÃÕ"5Vl7Ä6¾ð¨>yQ ô’{G8r i@”F¡p¸Fí®˜b …ÚØ€–K4¨vtç[ øÔ•‹~¸®Í…)âÙS 3!VéÆkúŽëý‘Œÿ£?H'̽ég輫ɔà.ãñWl,¢YH±û:H¸Žk¥¶«EmŽ·¸ªx˜‹²sÿš˜0,UtdB]I¦ìÚ/³óÎ>Ç~=Æ$Ðÿ÷ì÷þ—5_DÔøO0_>sÚ*Áï©He%öq`œ€ÃT¿(¸çA0U‰¹œ1š¤rO¢îheɯh•pì_$´g„*m)Ïù.ìÍ)€£3+y̵„å ÷r¨*²b0*lÒ2²2±´‘•©ÿ¹‘ ñrdeÕÁ·ÁZQõ㺺z‘8ø¡Í»4AsÄ*n^Âótþ1û³ëàN¡úàvÑg ¾%e¥Å:cH"ÏÚªÍ|Á¢B~Ân Êò¹©bÂ4ßàÚ¼p¼‰ô# 3Êáè¶¥§â½Æµã¦®ÑkmNqaL0K^y| =;Ô ÀÎà OÔzÄ‘§–•=šÁžö &<Çéëߟ0½W1©aºƒ¢*(±Žuð­,‹ UNY0§é>ÂÐþL(€?k‹À«ÃÛª‹9ö6m•¸Œ.ÑôâÊ“ß_ mX+"íñç=*º&Œ)”œsË0泪‡e= Œ‘Lö Žü.†!5è]àôcm LÛ…/7µ¹ï‚*ùî&z¨})â ¢ ~%ª"â—¤oÀ¥4kõh¢"¼[3n×8ïñÀaÍ.ÙÞåU'⎶< ã°ëØ–³”Óz¶a#¾·ÏÕ §l ÌiÃI›ò<- ‚ F?NZ²Àö»Ê‚û–‹³Ùõ%a»Ñ§gðÝŒRš´Üž›qÑ”ŸYm/•¦¾PPû'¾B)? Ö¥WEÑæ¢x¤ø­°¦TS^ ‘X?ñª²JÿÊä;vu÷³˜q®L÷$ó¸ZK9¯·{ý±©]sœIi&¡/ªöÿœæ \æèw•Ûõõ´¥)qqRݹÏbUˆ"ÓÿŸÉÒ¸”Äßõ¾uà¢=ßì v©8Kˆ+<Ÿ‡UÄÞ^¹zΈbÙÙoA“Ö®²õ¤âÀ…ä”â•×pðN‘çŒOµÓsm`ëŒtooâà ÎàgÜòÁ!ð~ÒO”yç®-´EÿÍe«™q˜Êýi‚€7Aæ\‰dK.ü/9y.2´X;7±ðU¡¬¡úWªÄ”ÜjÁ!­Ñè²óã²Ziµþ1è2-O‡J¿©¶àÛÚüSmþ®‚cQÑ”¹y_›"³ï~+>A˜•¸I—f ý"Ðñ(è ¶6‘Âs ‰}䓯™8J6F%âÛËtsÉ>-v>×½‰»Vr(ëjq”©óßUyY?¤.ì«Àz»:Q^ lk©g¨l§« x`§/?öS—ZÎÑ¥°èdmëOã$!‡EÐGeì Çèü£„¶38`UA¥ ªô«±ëT`¸VËEAÂU Ï¿<Íè$Ïs/J]6™™žcxѺÙaxNõÔmm~®Í›ÚüT›ïjó®6ÝÁŠpŸ ŠÂQYÛÞ !¬ˆ@¢Eà¤Wâ‚ÜÖe.Ø]#s°¯)Ñ:¥##^uùŸN¾d°$¾6ok³ÛO"Ÿ,ǽ/„t¤“•Ĉod”/PV‹'î9ßñß~' ýzå+6D$´6èËjmN™#²Â=êT÷»ƒõë‘ʦ&že•Aÿ޹t~€‡Å©¥XÓuVà+ޱ(Æ–\üˆéM„™~ÙjÊ´óŒ[ t¢ÒLÖ² £š"ÌŽ/v¯P®@JŽàëÒÆ²ÎÅ@Ö}ÙEáÅ]mé`ÑV½Ó“ÅëÚ|Ó% FÛBÈ2SU`!ù; ³~³–"À¿û¶„`®mž0™&¥ù®a¿åUbMqe¯\5‡Ïµ%m_T$Ë;Û륥»[4ÝÔÉG¤@­9êÄ£¼á÷eÖ­ó"÷¾2=rH³” Îf9Åö•ÅxÅ:¡LAESì[çÆÑØE:Ñ´c~mpqÌVŠàšCÊk¹°ÕзØK]”©wCÞW_ìpÝê¯D„o±Î±pk#~Ò\¶–€—$fýÖMIè‹Ë³ßŸñ{X™Ï?®?‚ÕdVË#XƒWÁ[t퇑Áúãó¸6 5«×§•Ò!]]ZÖiâ9E`Ûï°{J,Μ%OA½è2÷>,†öPS…Œ =—*fÀÇ6Ò¹uvgC¡ à‰Y N_hÙZ¬Iþ|‹IL—KÜ{åb‹»è]lùæ"£Â|¦{& ÍLTwVÞ"IÆaÄ´ÂÇ’£G(!°yK7Q²Iɺ;\Ì7eä5“{ÞIö¹›÷¡»%ã|7çD”ÜN+wGú®ÂZ©ø„ïÿy ÁñÍÖ@Þ̉"_È êdyN&yàgcOóiéÄ¥6&Gܼó½ä`ù™(a;L%¿åJp¯&]Œ}{AD9CÏSÑmm¾ÎÉÛVq¦Æóž—j…@aº¶é¥ž¯*ñ_Í9ÎîăR×u÷ªV5í‰$ÙVÍXvàµÃÚ{YâLu”ïT ›‰ÝNkYVÁ•«ÒÈ.“IQYuÎú¤Œ^k¿B|YÆ Âi&6=ñ‘é(Ëœ&Ôz>ö`¨ÄÏäk!‚/ªá‰U‚ßEßN`±Ì7nþP™Hôá#ðêãAÇ»•(‹¥˜#ó!µñ~lr._« ç=ƒ¼(ÿÌ%ü.! ïé„ü1µ¢<¸xy¡÷’VÑF9`ñ€*+¦§äJÓV‹O—š2Åý/ш’t vR8ʳiÖÊw‘osù¦P[…>û¥CsØÞÎUŸSpGÓâ[#É‹ "Ò#ŽŠ£]ÁÅ6¤mÑ[ÍËÞ½í‹Iø-"?‘tR%éøwƒ>Œó%å#¥eâA¼cëZ‹Gsгþ\އl«Ÿ‹ãYÆ'$qºÓhvÁŽ×òëÚ|Q›/õ' ^6˜½jqY¤Þ‚ã]gú?üœé§ñú”ú ·öøxÉ÷¢ÔsÍȤq`<ÅFgAo¬7Ÿ[ƒP!é^×b7†V›à`fýqSAƒ/#ÔEáßÔæ¾6ï»ÍwÂmž›ÝSغŒò܆…d¢øÐ>ë¤2ÍkÒ“8Òm&BXXÑšNÕá­ªAÖo̵ ,d'—³žõîÉSÞPPÑ»!ÒÓ£]>^¹O.×ó¶…c%b“vŒ‘Ÿ´¿¼˜Skë/hw¸wáò—=kȤ"×ýç§„OÜQ1ÕBp†ácªˆoïPçxê¡@oáheƞ΂o âR^½ÏNe2_{7Ë íåÚC‘޼)ëLêÿkÅ?5ìÛzÈÅ5 ãE× +®T§œ±óðŽíÍI“éI‰LöÜwõЏÈåö=QÂ-Ÿ›ªsßÎOÌ«ßI©·8E…Ðâi`*ùKþp­öÌòvdÖÌâ) wÿþìÖe*pendstream endobj 283 0 obj << /Filter /FlateDecode /Length 7584 >> stream xœÅ=Ù’¹qïô†¿aBOÝM»pU~³ë+¤K¢­K~hrÈ!%ršË&—¤¾Þy €LÐÓ\®Ã±[¬AáHä}õ7ÓÁÜLø_þÿó·Oþþwn¹¹??™nîŸüðÄÐ_oòÿž¿½ùÕSáñÍ!MÉÜ<}ù„?57ÑÜ,a9$nž¾}²³óþéŸapðjp\sZàƒ§wOþ¸ûçýt°“ñfÞ½ÁG3¹9î^ìo§ƒKÑ…y÷^›”Ìw¿Á)Å8ï^ãˆ%Å`À­sþ±ìþkí>s0Ãt)-‹Ý½§¡ÓäãîLo§háû²Øé'p‡yZvvÿ?Oÿ}sÀÛuÓ·.â<óÖiŠ)Å)îp‰¼ÜÝ·>åˆç|¢ä—°;~€÷°‰8y·;ÁöÜ”sL»—øi\Ì<ï>ŠêŒrš%cœÙ'4~÷ ‡Ûi²â»~™,ù*žìîø† §à.èuœÌagx0‹º>Ì`w÷¯öv9øÉàÄüYtëNið‰O@Ïò=_R°i†3Â*ÞXG›3Ó´ÐÞãg@#ClYµ»uûQ‚`…Àäxi?ÙèèCg¸‹ÝÓ[£Õ»ý-ì A›†÷ð8Ú(?u÷yØßzÄ €ÀSü(£ë+ºNŸl”KG(æž»Š“w³Ln]æøž8fw//o­ÂW0cT,ŒAÄX%˜× }äxæ·ÈïŽwÏ¢ðŽG#(uâ· /!Z‚¹ˆîºe ²ÔÍR“Ô|Os\Öïq>ïóËF»Ëº\ß+ÈÁC à/é‚BŒ‰¨=?¾¨ŸëãÛúøº;àO;ñ¼ïÞwðÄ"×å[´àÇ™1bAðše*ŽGƒ‘°ª".ê¸'­'X¤;Ô@1;½®´ÀWÈ 5í*zGz@ˆŠ`ý ¯’)Í„Ã4{&µP•8ÀÜHȼȒer¾H|Fa=#ƒ +Ó `€H9ô<7O‡‹JI‹ õãös“È»ÌyÓÁ”½9Ï;)«Ë&N|À$íËÍAI¸1 â¼himÙŸã¨! éÁ€%ãgd{Á-pßÑ·íœêþ½’«v?95óój† }W‰U’¥2ÞZ³ ÏV´ë¢Ñ,ö²*@ö0WºÍêæâPCxU'>î{DïP ·V}4i¦<¾¯oëãÇúø¦>GÌûÄÍ×ê®#î• ‰’€±%…Ɔ´„ïav¼5ԀȧiNYï÷ËâVtœI±ÀÁ ×ŒHiëf Yó\Œf¦c–ЊîBÔ÷¸ p ÞcÝž²XÕplamãÛè¿ €¥§—øìp”{ÖÛvÖ]-RÔTýQ.^¦s£-,Âù¼QÜœ¾u~ëÆ’SZ‹Y•æÃbŠrW¥Ë±>~o{XiÑá4¥FÅ@üëhniŒ’¬aLAà.ÑEàÌ\²Å¬yO4í”s¤¬%%QvŽèÎ0 ¶ƒRÛU–Ó˜ÁJYrÀF®öéTf%é¥xˆÎü:)„…sÚ“™fç±|Vñä$Xæ iÁóC Õ5¬¶±‰issãÇÀ žÃÞ?6µZR s§pÁü©Ü´,Ʋ¸x‘|Â,ºã× Ä“L†¯vŸå£5Vû—@e]ŠL8>#ô|°#?ëº;2. ±œÛ‡<>/ðo…GŠe){X+0 tœòž” ËÎ$Þòؙħ¦šÔk–9q ˲û‚z#Õ@p]¾ù`m u~Aû-µêéâHÆ*ç¨Öâ2¥I˜Ì¯`íø®h][4OfZ¾Í5êàb¸¢³Ñ A®ô•Ah -M(¼§s½ ‰ã,,°¶Ý³Š‡b6æò èÿÜ^nIf° oE$w•ñçÍÖ²‹O˜ðoêëʯð‘<Ïšï¹2^6ž¸Láoêl‹øõ¬xyc).ìh¼ßv)àÖ·û¹è/¢B€“!P$9f÷¥ªWâfn+0e˜h9ó¬"ù±±OÊ3ÚSa!‡•Ôq„çU[£IeÔ–$Ä:‰5Ö,)¨2³TeÈ(£8Àëj„|®Ò¶ùXl›®sq4DäK y\ÃÄݰLEMZ"‚@ݳ¾è ¦YÝÎë*éäp ›²)g]7qßÙЈÜ…gmœÎµÀkÃpŠT‡h`\ —ÅÙåãŒ=(¤Û,,dp¥Ù,Èö˾(Uö³Þœ=S ½Ì³S¤(XÑ«õºLŽ*€"=ÌâÜRE,%Kqã¥öøZ‹ te×*µ·ëùµÔDõZ˜,ìØŽèØ—êÝöÊ& RÆ\Ã`D<°u[¿ÒPèòPX'‚æÀ~ýäéßýq÷o$VØø"¿ ªt~w:{ÛPs³3='×h – §‰Ó»ìW[XtVÿ‡VNÖýÉŽRξ#Ð:ŒXŠ9¾ðhT„Ñ$àó7õâB Gé.xP=?‘ ú>É Šoÿz¥b€§ )HVb7£id ±WGV« eÈ™'g…¹n†9Pö—XĈôpß ®ìT“9鋼ïBoã³Øyïç89w9RÁüå»z¡|Ü×ßpXÚ^2AŽè»^w§ll¬ñ µÕ’³ˆ‡ÆžÎx3`õ±GÃ!£nÝLi ír±G•ç.[DºbAŒ‚ã>ÙøÂSXîOliˆÐõ*{ýk½ì†K•eNbSgÞZ5WE‰p0jIÿ$_oí7 …ÙŒ’wü! ùGäß‚9BT¦û& ¾*qe:I(=æ“M5Œä…âW»çÅàµó·;¡~"-2æÃä'ð¡£Àü/<5fÊ|Ï  ¶¶‚€(iã¼WqÁž0ôài;Ê9{žI ïXÌȺB vE}„wz?Ëy²Œ…ŒØ$÷àŽÖög rOÌϦ@&iÙÞ]  xÆÌÞ‘qµ˜£F²‚M¿ÙçЛW&åö¸‰r{úþ ET­rûF)A‹¼ÆŒý©ºœQ~fGÛ©>ž÷=Oœ ½ß÷¢F§î€+H3€<†Ô×Ì‹¶ùP9ÄwNƒU¦Ö¼«˜5p˜ÈÑ,ÐÐí¡³CVkOd‡Õc4ÑÛö‚@µzr©‘4|Q«=ÕÇs}QYÌ}_E\÷Ô/ŠŸ‡å¢/î%áf”j&²-kÈM.ÈóHy<&5dOCÉpèâ+KL2Gd‹ŽŸêãOô'{€µ© LVÓA,Úì…eŠßF¬·ž'¡UÑ8ûàuH„ÍÀÙW4´_îI ‰¦aëzMø‰ êÖà—½Ê4R¿p>oB#œ,‚uõÑI ÜÇêåÆ*ñ©ØÑЦgQ£µè¥Â§ÂºršÖAbæˆ>´F‰ðHËu‰$D«åâk¶¸‹œÊRØ %Ü¿È_Niœ3$nHR‰Ìù,`v–гœ¤A Yæ½L4y!4ŸÂ·!Ÿuìû§O~û„Ó‡=¿_nÞmΰ[çn‚õèÒÄÄá?îþ€‰[&‘_»îôU%Œ®GŽ< P °cJeýU%Z ›/„Æ.~·~|êœkf®Õ3°¾ðPàÔù{ô—*gÉß,šU¿Ã×LfGF3 Aõ=‰Âêû¢eK¾©¯½?[ä–úv@:ÜJÇýÏ€~+É.rsS—~êÖ1WÄ·[ßïE”h”ºq˜OyzüiF-Žæ‹Y1&s‡M¶#©—ÂcË 4ðE8eH$Š «4ZN7šÝ⻾ÂÏüf¢ÿÒ-.Ç2p€/"4ðSëµÍÞµ¾OD$¨ÃÃ¥'§X>Çdp S;>«"TXtýp‚23 +eÌ j"q@<ˆq2¡I‰ªízjt' ïà 0¯ãAU/?ñh† >û5á‹—%Ý`³*_êím£˜Äu ¥ÇC ¡(-Ï÷½¬‘ëq®êã‹úø¾¯y°Œ½ùöXû­Žü¥àˆ*™*ËÔÅÖdJ¶wxƒ¦ë¤\NYw}ÓÕ„?×Ç·]¥X ¸BuŽª0V‡ëgDí maG®¶Ÿ‰‹®*²±¥ [V¼ÒëAëóÈ-ЦvåÑǼ¸ŸÌ(A¢Ð8æGÌ8G’;:çDöÚ´5\¢M€—»Ï§S¾ÁÛmùχœK?‘ƒ¨¤¾ßîñq1É õJyfa#¶™¸‹9,ÀætBz4”lчxCÂAPOz.ŽÂs¡c8 ÆßU “9yD¢¤'˜ÐµËAý„pžy>o¤CãTR3?áŸ-Úc~›'èýè2–ÂAû²bÝ æ* |çË!ïâÏu?Çüìâ¼q”R¡â²¼ä ±à?V?rÇK£ö÷ºH ã…å[þ.îñ9òW0,ýî/5 u¿‘^[Õpˆ9ç£ ’«‹.?ˆ“à¦gàÖFÇŸ7KÍðk&²Ä©úxþ/ÔÉ7‚ÄG8º›Y÷û}EB•ÐOÀB«W½fK¶‚NMD½4~ƒ^ ã7‚Ê Fb\”^JòSTvÿÙÉ„m=ÿdøYRjà2Ÿ‚JiåÛ]šúÇã{ždNWÖ âØ©I——l‹Føi‘:ÒÉÆIðŠã?óg©# 1f¼Há-G¡µÝ·q›dœ€Ú•º~bZ°dW*Ö5ü=£à—h¡`™û¦^i[úµM‹ gƒò>ê¸Ù+ã÷Fo„”¶ùñ-ªl÷Bü$§©J·-0õ_±¦ vEÝó ×¾Ühksƒ¡B×dW Q4ÔëŒæ¯5ðfLh*¼».¢¼%p"ª Ÿz@‰n».-ïÔ%1%œu †!U ”¡$-WLå·ÁãæÌîÏ¥’}+5±ÕgâÜFªa†‰ò’tZ)½FéÍI¨ÀQÆØ^U¤¶ïIñÂÐ$¨†n¥@ÛXA°´~eM7Þúâ¬_«¥ ›C¨ñBÏZú| –¥ÜÒ 3Af› @L~P'óПïœï ¥ ¬Ÿ.ÉÈêÇ6 ÅÁt©¸r™¤¬Š¦D‹‡Gª3ZWø!óF%ØFÒ)GHêa芜.Çxɳ\ªùRIvf¹)b”æD ff7-àSª)ü¾ÆÍ´Ï \¥sÃ{1 `âô%bÃRµ¸l Á;;PnWÚßý«¢wTTÒy„Ï”bõödHÖsqÅ pȸrOúqpÔ¥O§{‰qœ¶H­®.µk•ïGÚŒÅEs“²bX ½`±¬¼_»L¯Ìúb(Îv[Þćd§±2¸_*à|’St½Á’¸úYÿš1ÝN*6Ôæ¼æãjÑ&]Eúë‡ú(”6!îQÚ>vzZ”2„yTÛØ¤‹P‚JC/IûFŠ¢c®x¡‚0N-$,CAþu_²]^äM—´ò<àÿ㹞ÀlC¹Vc” Û)5þª’)!YFË•½]ˆ_…™l/UÜÉ ¥D‹Š&Ö*ÈšÏÓfŸjû*-k~é@(ßÕú“aÇŽBËÞ-<ËÚ¢ñ±FEÜ£rÉõ3dÙbiµ'Õ BïKRŽËÑŒ\äÞ—6R !.ˆÁÏØtè[Û ^SÙÛ ”ômŽ3k•å˜Ôµû—} {ýššÊR¢ZùxnŒ6¢qVì5tÔ^2zT ‡Â(^O¿qæÑ‚ßâÌãƒ(6]ØË¸´Nûç€ðñÔ„ßÅA¬RhÛÔ¤EØMÑ¿Li¹¬ÌrÇ ¯u mú„S·~uu{š¤÷öØèY{×ÝQnQTKa訰!•ˆ1áæ~¤ªWpNz–Yºxßú)~µƒHv¾[û‰^Ñ$«3ŸBÖw«ëÌ«æº%½Kj”=7Á½¬¡é"¾Ü‚v*¡+ilTbc%—ìNw) æí,ʕęšô›§)wâeCÒ| r^i‡·%/ÂÒ’Ø©Î-}aƒ=.Ç!„f¸y0öړ夥Ü'jMTùÕÀ|‹Ò,¶`éBøsò %ä]S0JýšiÕ2dõ÷“&Uú2`âÌ€AB†˜R¬TëøÀs£n!z’K=K¸äs¸ º¨’8ß­-¿”Pï' KØ28ƒ-—BlTæ!ƒóX~‰Æ¹Ž=!„0e¼“}:»éšî'Á1_ó:±ääô÷¤&”«ÿÛ"V•ç÷ˆbØx0|.íkä1½ÖÉ£}•áR»qÏ%•ç HÛ6°I…ÎxhHq ÷^QŽs¼ãIæÜ „ž'•p>šå ç©J>䜢å`Ü<q¯§väHŠ+í¬Œ–\vdÑ‚ "¤³ôuÛ­˜Ï[Öb^fbœØÛ°¹û»*JE*ÆPžûÀíÔ$md’IÑŒ¤É¦™$ÜÅb8G'H¬+4¤…/È¡dÀ˜‘ÁÀž2¸æ47]’ð$*0uæ`”ÏDÅ…ÊXU°ñ1 ßó’Š‘Žñ¨?•‹%–8޽j\kÔMW€Ot—‘¸%b0ª×ê~”ùl_ Aè%*’[OÑ‘ ‡‚Ë–™žºøèZÒ~OmKÞU¸l›¬n)…° œô1ÒÉ”ZœÝ´l~ú=§ŒAÊD¦/ÙÃö¼º³ËöÈ <ó`—qä':ôy½Mm,M¬óêDTÿ€Yvˆ±ý™„2¤€ãëšÞ©óÀ+r¹®½|d—råÝá•úgP"UdÞnr¼ˆU5©îát=³¼1´6}=@ù5ªÒ4;A0u'Õ&¶åš£[›ŸMá…*=û:¤¨!s'äv_EêÂÁ5tå]ŒT˜ (ª[fãP ª¥Èø%ÛM¬M[Dp”Ãü—­/fE³µÓqð{ý2¸ÑŽ%ã¡gû8¤ò=ôß"'b“WáŠßº\®H«(ÈÍ%Ó&x· Åý4…œÌpU] Ì?ÒÁ`oŒ?8° Ⱥú{sëá`ÌŒHWH$lb=á¿–Ýö +dvÈDà­³9*ãSœ-û{ó£|ý€ÉÞ‹‡b¾£c¡EÎò¼ÆG„ðhÀijÛsÏ×¼ÊÏ€úÝûþ*?â‡@ò+gÃËeÀ'8}Ö<À7w)j…„õ@«&mÅKΕ†}DdP*¥êrÎhuVoÛ:çx¦î·ö‚|¨‚eå|š­öçÀšPì7}‘òÛK¾ø}”ßm^™S1ì¦ÙݦÓ=/‹cGN~;p¬‰t»AWѾMĆPî„] æÛÜ«šH‚lóƒX:¿¡bâýjÆÚÚ?» Ý‚’y<Ù >óºèâi53|E¼ ¹ç›䊶Íí…³ý±æö¸Üd7M€TUÑz)`”Â*ÞT4ÍŽýÒÖd'ÔªzåŸâòÈ¡ŽÓK½°NnWØiv³Ö%þtKe­ÊÛt<žHeÞ´KòtF Àw‰[ë®ËAV°nÁB øÉÕÚò?žóll·U$³¸ºh!¢hR¨`džIÕú~«¹QÊè— Ž¤½á²_,«] tj«Û+úV¨>/=a;Ðør‚ØbäÖúE˜íï®y >“NØ„§xóWzç17S”'ƒcœ'¾Œ9ÁAï«ù…·Fã‰ë/¼eŽß´/´˜iKU']jû{%eªú‡äÔgÞê }!ónUb<[§ÀYÖ’O¤ñû<³ß¤dÞ/Ó·6-¼¦¯MZœu*GMñ8/7ü=ƒãº¸ö_7®ùq±AZ Ñ\„ ƹ†©ºù‰i!$rÊÝßu’²bÓÊ?ïàª&!?°ZÈ-s cÀh•ë …÷ª±ã¸ï ÙqVWæÉ'ú+þ@ß ®z ÈÏü_S¤ð ŒQ~¥Q§ê¼õ´±pŽÔ+O¤&""¬3¸«“ 1„ºf¥x±r5ïmôë•öo&ÄÀý$Ï:Èbz(9“lêÑn]pd\ØÆ ÙÇE[›†Pá8)$(¹ˆ]·Z "–¸´“.®ÎŸE¦ó¦7H†‚&¨Ÿ¿¯_GÙª3Ø^…¬:L¨·Ñ¬ùëlÏÁÌù·Ø<úÌÎÉ­f^˜°y$ýßóoÚdŒcLÁذî; ÷5âáE–†Îè°£^4“õÇí1áŽ9åk"›k¶³x¾Ztpõß=_>®”È |øXãw/Hã²ØÔçØ/÷0K‹ÿw:/i­cæ•Ï%sûšÀ`hz$Z´fÊÅ yÔ V8A5 mw›Cò2zÐÎ5¸ÛzP½]ÂP@ŸP: üöÉÿBÅ×Õendstream endobj 284 0 obj << /Filter /FlateDecode /Length 7942 >> stream xœÍ=É’$·u>·:ûØ¡S•ƒ]Lì€o’MZ¶9i4Ž`éCqzáÒÓÕš’C^üë~ 2ñ€ª«§Ç!Kæd£°<¼}Ë¿žO;u>áÿó_¿9ûô¥ ç7gÓùÍÙ_Ïýõ<ÿçõ›ó?¼‚Λ]š’:u}Æ?UçZÇñê<¸°KÆ¿zsöÕæÛ‹i§SŠÑo®¶ÓÎ%§“ß¼ƒG5MaŠòñ;RLnsoM ÊûÍ~{aŒÞ¥è6ÿ ouôÖÅÍ[;M6þ÷«?szg¦¨Ïá0Y<uy¶ùŸí«ïÏh}«,lSüí«Í0—U^»‹]àÎRzó ¯¢Ý|¾f§'`1ØÍ{;ÄÓøÉ[ØÖ¥øý[Ú¢ÕÆÂ“šŒ›×[íñèqó-Oï’Ú|q…?3)çi°J“Ö~ó€`ŠÞX[ÿn^ÏûéK;É;ÐVíà(pxnèàÎ|aÜ.zÏGþ7œ`­lÄ3kãáaó§­ 8çAœ·á•³&À#Âuš\²›Cþ¡ÞòsPvó 3˜Ð¹Æ¨H‡7ø¸ÊX·Ùßò$Q0gÈ$Üæ† 5v#!Æû >8Þ_Hm}ÙªŸÜF.~€×ðÉÀÅ-8 §¾Å'“Æ©q “÷œïó¿M^mùÝ~^P8p ¯¢Æ™µA«Íᜒ³ ȵgð&ß¼Æ*DK¹â±xB×hÓhj¯VºäeYQüP Üfÿn‹Hg¬ÏHö žR´¡{•—øF¼X½Þ“5^o~ÂýkÀÎÑîÊ¿A “15ÌAÂXŸ¢¶D9ó).”Ù9›4nHÑ:n~¥GïL…ˆ|‡mÒ¦¢zi'Ô° ÍM (äSLˆ÷ôCøÆ@ŠOõ Üá‡ÓÕžÞ¶'À‹Í=œ>Àþ<¼ÄžÚkSGò®ÝöÂ) §Ñ›ßó–€UlZºÕpgiéü(ÐFÞLa…W3¸ÒægâA¼e.Ú`AC!Á‘~W'zigSª)¶™ïçž1„2$9nŠ XÃ;|Sò.,W¨pEx;9ïÍfumÊÀ à|3U¸fªXŽjÛ¥gT÷º,ˆˆŒxj‚¯)Ÿ¯NxYÞ YÊx£ÃnÇ-1ŠÇ,ð^g.>#%Nå#qñ™lžö 9vu‘[IÞ ò]ò(BÚ"w S> ¸å7€JªB6È[ž$€”4$ ù1ËAf5¸Cdé‚yïiýtœ¦€Æ€t€ÆvóJ`Û·$‹à×#1²œ>œšP°¼>Ò¡@0ÑëEÝ j º}œÌæ_Ü“7À4V:±3>H¤“Ø•ç0ø/åz¥ž»¼`d$JÈw”œY0™=ÀOô‡»…Ò—åÔæE!¾Îà€Êc_.H+9÷?l-°Sdr•ª·#áÒËèK|­ «ÿc!u!¦Vc”íú®óoó´€hÀÍÒV ÓSÈÛjPèëíB?¨L˜±a%À€¨õ››«Â®îú\\(ZÌçà*à(R’˜JX+ÇÄôL©n¢ ¥º¢®Y8ÚQШù¶Ä<7Ócì*eVCh.%Ü-ÿ0 x]X€DÌÚJ²ìP?£rômPA/[º-úÎeÞ?¨kµ©€SØÉ6LßNÊm~SÐãÏߟ¨¡-S¼­Iô% Ë9ôiÊÚ|Ž ´o‹¥~•¡?é¡}qWf¨|3^ÑxbcÖø€7ZüÌw,ØòÇŽ³ ‰ãcgìe­DÏ–¢¥Å¥ ”MFã¼ðX); ; ÊMÚ/ÚØ7#jß ýÔÒŒÍ{±ô‡8v´äŸ±Ïš(9@מ3r ü’§æU“ ‚Ïp®]k­í#6$¤Ïe#¹"—]ÜK—æ5ø-˜×µŸ‘ð qFßCïpÇ38´YZM‹4¾ô3±ØrR—:´´‘]„H‚8 X£Éá‹€½œä†Ãëœß3}ðijþ^)'ãŒÅ‹ d7Ì›mF«A©©î\"(Mì#Íj,©$âÎðÒ-„IŸÑ•RK‘yu¡Õjk†™öYœ²bÍZ«Øš¼¥®A ´6zñµYä ÝÃûÚVèû ¬'opq¿“)†çDøz[S,;Yû\ãu3G&ø¢züÂë…˜ŽQ*B¹&¸Çw! Fâ$CcŠ‹šbz­¦Töt£6ãŽÐòÐÐ8O\;M]‚J¦aùTjàÝÝ“³,ML˜ìjâ}ÖBaŒõS ˜½ÝE<Ûbl_ïÁù9ëHc›t (t­Æ›ŠƒÁì;¦áÐNÑdòâI·Þš¤ŠÌ dÑs• ·Ðr>Óz:QMq„Ó÷?ç²!$2 ÁMÌ1!|M@SøÏe¡úËEâU!^Øî0O1i¡¿‹Cq“ŒMåLÑg$’Î3ŒF` 4ÙK‹§õ$½qÖØ²9Ê›”æ¨Äà=sho‚•qÃpï·L4SCtQð#±vßÄÒ²ìyà9È£Ht0Y+Ftí^ÉÙÑ ¹ ‰:ÌÁ‚!"-ˆ(׊I‹DåZ‘ƒ£—˜¿ß’êå–@ûñ÷3¿é“K ¤ÈV‰Ly‹1­mDÁ[P°…“¢ˉ¿)ƒ+ÍD¸^WAn Tr8 U3fl_I¢í%é¿Þöü™æ„3†£<Œ}­yôlšÏDÓsAeÜç-±a3=9'Փߊ[àÛ¢NÏaV- ßFþ§®çÀyh¸z4a’ma™MBOs"èNFéšRXCÓ‰€”7ÔZ…ø™ÓÀù,~'£M®ùô¥w•£ÎƒaiíŒ',FµíÿÍ´¤y|KŸcAVFnsŠ€ò 'Ã0M¦ÊŒSb×o‰µ*§‡ZÖ£Âé×£àA£PÂk«F¡äÈ•Œ8”ä‘aø¬¸-Ñ=Q1{Ð.w áÃ$®bÿ±µyç¬I #]¬ØO²q–C1%NÆüàˆÙº㬄gä] Ùøf¸N¾p±4¢šÚMJ¼ãéŽüYåYhé}v(Ï ì¸»Í[gÑq±Ù߈³eð‰lf  è“ä`LÞïûÖBå0iºGf&îœÑ‚̲Ü4½/$w/ÅÔ²öAÈÒžR õê…ûGÿäÈXV-ôW£í#s •[jymFžþÚê_´’ÆP9´·˜&ÔHŽŽÚ®¼«Í‡ «1NÈç¨PJ³µ …êåJ"‘1e6/àªv³Šk0šSÉŽ2â&¯N°-«³xÁØv¥ÕôòvÄÄ'dö\ñz˜ \Aƒ¤µ×ånÆz¹Õ¬©‰éx¸JP5o&?3Û袮D—Ð"Äžé&µ™º 0Û ¥›P<6ªàbú)$Õµú„vb‰PÏ÷CY#eŸßVèINW…QfQQžC˾X‚§>a¹ÏÊÛ…ÔAek"Ÿ¢¶Ÿ…°¾ÉZ¦÷u¨•ïÙ›ÉHø…µÄ^F ó1Ëscðwã1†cêÍmy¼*ïËã›òø]wùFQi½í*6ÎïJ¤Fð<†¿'1Y£:Jî*Õ­RïP¨Ô­Ý-þ x x:T¤[ò²´ÁÓâä{é)dŸ€-Ÿ(Ûå(ÓÏpÙišµ+g vóþh¡1 ô€Kž4³Shw¼mÌ·ÌWèq`¾\0±²†©¨3-u,ó,é§de܀ߒXm‰}Â8ÄÞ°“#gý0Æ0R˜pê¨T•À Uñ¾þ=¯ˆnY¨M ‚ >E“(ö%á‡2ñš›R0‚L¬“ô5œ:€Ü3‚KÌÌ.OYŽä4Àµ_ã¤$××}ô•ê&ûj@Õñœ‘'ü$Kµ‹¦òÁŠˆÔ#^Z€Ô‹ír`ÁoÏÓ¬¼T¾/‹·¨ÃÝÏ u©—ÕMW‰ï„k îÔ³HìØ³·ÃÛÈVbJ»ìáè)3ûyˆŠöJ¾ÂÙ€Á…dNñTçÑèªÊÒ]PÒ6&޶®#Öy7§'|$—ÛÚ1ïŒI•àðAÌ¡ŒÍàÀÅ1ƒC«==ž“±Ü¦paP¸kDƒ=ûÄ™‹½wËï1 áTwæm6Y²Œø4ŽQÈ,s2°Ï$'{àéìø{pŸ!‚ŽoáOvŽ ‚îráäZ’´ØÛ Å1tm&¨D)< ƒlUbc#(¥‹Vú;Rb#µ?/_”Ç/Ëã‹òø²<Þ”ÇÛòø¦<þ®«{%]m ^(¦ò|#Š–D”ÇbÚqfÀÁ©Jàœy™ÊõÍ,‚¾ çZCÈõJûµRÈ~=UÖ"g¼¢Ñ<´‰¹‘¿0mnøþ)»degšËnä&Z²øI?ßöo~,±È…n€úÒ{µÃÆ9\ÈNYƒÉ™¥0u÷³=5Rvª&üj³«ü\*zS;9Ç×eµ!ž´Öz'åŽæ_i“ó¯D.eñ5fÊ|g]ØZR5’Uùyyü¢<~Y_”Ç—åñ×ÖZ%³ô¦<Þ–Ç7å±OšÁà‘«¹Qxè(s¿)\|à|»ä90JM)~šS5å$¢*hXe¶,Ó·«Eýeß5 ”ÛÛðAÊçåñ‹òøey|Q_>ØÀü¯ì»¨C±tpËD©VpL›¬)Á“tdÔ¯–,L#u$b©;˜æ4P'6½6A‰ê’!=ʦkJ#X© ,Ѷ;‹eo%«}àéS¯™õåJ5œYO9_‡˜Sc ó|OõnàïB R™g`/?ܱƒ$h£át>5ÅcÄ™½âf¯)“{±P¤cí$Ç+­¨é§óó¾KT81¨~¨~-ß=[å ÍŒÔÓ9ÆY´x$;¥¡ÍK(”L{” fc/TébëRе·î ã@KuÇ®sŒ.Ä9ù˜Ï#É€KÓÈÄ<â†Ïq¤¶¡½ÿ á0qI€ùXÂŒoÊ£pÞ þz‚Cögí1ÛjÐÇ+Kª(²f7H­ÖÎÓ¯ju—}å×»­6$U³ÒI㢺†ôÇ—Y§8¹Aéò…ª¨üž/g9E[¸:Úd`CÙ‹—}Ra¯9!F¾Õ“"T‹Kí÷„ܽv*3ñœœMv/JÛ“ vªÌ·äJS—<”ü¶ÊÜ3ŽdöÊ«I)!µùß‘ ëÓ%alO¥Jˆb®žÅýT}S¾‚‰˜NÎj^'†SV³q¤k“ÙëD¿jô°|G‰×‹Öÿ ;­tr!TQä3A¿VA)®É•7¹Î4ÆyUÍu÷u‡‘×ã:V#£öË-›Q&¢4ú…Dùæà_¹êpý¾+í Œ\õìòFki…-€°€Ììæûv7°së\Ë)0ºI˜"2›B˜3òœ²s¾”êÄ•\º•0Ч„ÁŠòõX˜MõžWnà6W¯å¡c;Iz!dZÏbç~qöêYºZ3|D’þãÌá¤Ç<3鄺“×&ùÜ<øc¥µ•ìNæêh;h,%´µ Ho½zô®xjGOVêµ"ÿä8sö¢ÃÂM6ø£É vOŠÆôÒ5.çÛhÝCÇ´6ÔÎ{².!´†«òø®U6H?”ÇËòxõ˜¿/ïÊãÛò(¬÷å1«+üÜWW”R»©v-Ѩ°%†ÉŒÄDCìšt÷ñù†'Ç‚»¡–0SÏ í_ÊgñêiãÄìr¦ Tãr&—£tf­]T:+öÓDqD7«•¯8ìô­Âtµ8 ×»eÏé‘\#² &ÒMü«R”?(àÙKe¬–Ò|mØŒi¢ýRÈ8ÐͥáΞeæUÇÊ}/uª‰2¶¦ÐÛîãvÉ]« ìݱbÀ·q”(¹È5a¨T©²˜Í1…Vw—"~l¤¨O¥z†Ø»l6PN°v¶Dºâ(|>¯_ðk¬ü”µ¦sUqUÆÒ™\àv÷ZµECødŽÿ¶ËñÁÊãOݱÿÜ+f¸ï¸ëN&Þ ©s‚­ |~çúÖ? A³¤÷wˆ¦ß7ïŠ$Ë€ë|ƒjU¦A‚¶&̧7ÖaØV'áp²L öHœA28ø¨y™ŒÉZJvlc²U6öñ]ÿ¤"ÒlÂM“OKáÞSÛ.q•ã1ݽ Épa©ª ”÷Äe7¥I“”¿ËÄyÜÎÛ%„½ÌZ]Ã\¸ZW 0Oª ‚›VkÄù°rWú´is•»… èõœJ¡ø³ ëáCö¨¦4¦«Ùå*Z¬]l—ÚíQ@µvFÑ$È"× Îv™"b¿7s¦è#EiÄâ¨;gí¨¥žƒâ˜ÀõÊâBèFŸ˜}R˜­R?#«µÔÚtü©”Ö$éí® F³”üûçÍ}ÆwHºÌ ~&W×µ´nMU ©‘í‚X„¨#G°-Ò#f›©l8=¼*jXÊã»ò(J_îËc6¤ø¹oH!öØnÖ ¡C¤J !¦¤)»²¥¨jš‹@­’šR|ƒ]*ý…Œ!>—@Ðë9]†dÁÞNÇJô ºa®t/5tEþ¸ùTõq\ΤŸxà©ÐG¼/0é·‡im<éUzñÊ ŒmŽl:ÿ€ÌPáîáϽï¸+WÝ·ÂÔçb`ˆUT»ƒètЦSXív 5¦é÷Ä6L•²LfYð-i£rÉ館¶ôÄ–-šzEúŠárD~´+c½Gkä²L|rû%ùÍ#}2*ª5Y1Ô!ô÷¼ &ýÊ«~æú99ÕZ ×È<ŠÿâûèÜt2À롃2‡èÄ,¶R±':5ôÅÕ5ð“ŸSvÆ—éž’ÆŒÐÁ|±öÓ¸-ÏÏ´º0J¶MêŸ .‘l펎ö&‡/] 3Ë'äú-SEŽ~‹¸—‹‹V%UÔ¬eR}ÿR­ð_û4êy…ÌŽ[(`žÐÛ„¿löÕͶ×mJˆöúýV”´ú̱ÕÕqÇîø‘ôŽÛÜ\¶2-]¡*‘„ƒÑ?wš#”¨ÍcBÐâ±xDg¥î, ÁäIyŠÝ¶µ~çDKÚδ¢?–ÆT:âÙÉ;¬å‹àë"f™¬æÂ2L©XÚþŒ{Ùõ¼Â9Ÿ¯j9c…ù޵™9–™®¦ií´9-5µ€ÿSÿ@ °»¢ú`[†AÖJ|iÚʶ›hÛP:båÅ7%€”fU;LÚÌŒNFÚfÐòz/ãÜ2Ü8¯# "ºU‘Á–íª¬Ížží*xÎC—F#˜Ò D0|½/µ1ÏÒv5)ôOÒúünòN<­ïC‹÷ %”þx½÷ÊS ¯BhôCi5Œ±1ZôkËcuËÁAKµAvñ“ïW9ªº,¹´æ˜Ü&ã?=­¹eS"“0Ûï/roK•ýÊ`ég’Q!WàD¦Ç{J‹cKNˆ±½$eN"²aÒÒ;‹ ]\[ E¼g¦i›ðšÜ^ð½_øŒµâwÇëa&ô ͽj ï°9OãÌ \ ÇéRìõ©³¥æýÙeÂbïûa¤[Ä]?;ãº<Š’@áQ9!eÆZÅUÿÁ©LM«Ýaü5ÌIlVѽ‹Û¼Hjª§ššK©•ÞLˆø>¨ù«i¶S.|¤F4„\ÖÓ-é­XŒ žLYaC} ‚§Ö9Ç,ƒ`÷® ÜtÙ+ÆëæÎ®šÊÏ]êð$ÓøÛ šrq—ì3PíTÁjl?t“óž*Xî~:­ŸŸäèÝŠ&µ3¢V˜-÷Ý“Šó?­Œ†Î_’VÉ•c¥@EÑeêÿ™#?•1¨#ütMtµ2,à¦m.6çô –Àßâ{æ~ ±ò̶5 Ü6T&·ò(©JUôk•Êöþ®$=Ìu´&ÍÊÅ*±½«ßËß-,e Ò·jøÑ‚ÿ#Q¶Ûè¡ß iÛǽ`õÈßtÙ[¤TkK¶Ÿu±qÃÌ9)YvÌl6>}©•;W˜gš0¢hs5Äg¦äÑûý›ûÛ«Âi‡Ž—ÿ½zÁ©¾»C%ÎMÔïÝ·WùpòKjÀ™` X4?]ý9À7Ýÿv­ÂM§Gª1„RöÀŸMaìïʽŸù3&*Ǥ>0ŸAw‚PëzÎYYmùý:Ú%à{lØ|öiŠT(7Thdw ÞX£Íߣ>‰‘‡e Peý…Ø’øÖÍ mñ%EkUÁ¦¾~Û:SoNåMáCþ‰…5PÜ¡4w49§Ž`ÛŒ…U2œQ‡|j³wü]¬¾•#€;îÜŰºl–?Ú0×þó.›/è”s5 ~K‡á 0iÆE)ãӂسÀDø¯Ã6Sà &óñ¿-8K¿êËâ˜L ÎIcŠp—$i°Yšž¨Úp!à[±Ðn±jžñ±<¸°ða8¬ø±Ú¾`£°$—áïü=£ñ 14;=îðK\Xhœ°Ðxð©[Â;̰_áõüQ#QÜWGæÛ¶«ºH(¼ìê|B)~×o{¤­¨­xÜÍ,T×ê;UïñÈi$’‰ìê¯Î  ŠeáX¬ ²!`é„)ÕÒèûÀµ5³Üyð€Qõ¢6Tö¹J·îC×P_·]‚ O™6þÒæÐvrÕ¨T¥~i*ækÑ…úCôÕ3§(ÿvTS-*EäùÉ¿™Ü¶Àsì˜"X”h•¬£xÜ­q:¼wjHéõ4óµÅ‰ìê?ˆ®<ZµU»Ì¬=nòœÇÈ"Š Ä»G\wJl3½í˜Eþ¶Í)0ÕH¾—µ∵n1o´¢¼ª·L»`}aâÒMµ÷ûµƒ—¼;j˜Ry>„Yr“ek^F†=¤¥Cj²WÞv ߀¿BÌãëT2Ý µÇžnØôƒã‘9}¬Š™‰›žÒpåG_èû¦\o];ñŠÇoÏ@’prâÞªB¦Ë áE9”þQ~æDá×í ç.è&Å68Gç‚Ed·Z¹¸P÷¿ ^ Ü¡bSôÂÙ··É ‚"ñåÎìV6߯k5ùÛ,ÀJÝ^÷8/æõ]Fƒ¶8‚,lj,Œ*®¦¿ŽéÄH:) >û_8݃^endstream endobj 285 0 obj << /Filter /FlateDecode /Length 5020 >> stream xœµ[[o%·‘~Œý yê³kux¿ØbÃ1d={<ûÐ#i”è2–4·üú|E²ÉbŸnI“8œápª‹d±ê«ÏÏ;1Ê ÿ•?ÏnN~ýƒö»Ë‡±»<ùùD¦Ý•?Învß¼…¡™1Š(w¯.Nò§räÎ[?Fmw¯nN-ö¯þ bk:âàG=>xu~òÓ𻽕Fºáš†Rh†ÃþTŒ:mÝð Ó2FéÃðŸDcn¸" C´ 8ÕZá/~øŸ}P#>Óà F£÷j¸O¤B˜0<¤Y¾¯‹ÝÝ=:áµÿÿW8ùîÕÉŸÒÙ5ÈåÎxFïw:º¸SÖÊÑîî»ÿÝÝžXœË)»û!üÿýbûþŒÃèÂÎ kFëw7l&ÆQ©ÝõÉ'ZX£j3ÊáF/U›iT´ª AŽag•ÄRD(ƒö£4u†g"wŽ•Nd½­æDe†yëp在Ìp¢¨î“•F䤴¸EF4Ïp"í¨øÆçNd¼åDe†9T*NTf8ti´Q™ÉRß6)qCcð–[ˆ”Œs;+ã( YÉOë½Äß"Ôûþ0=ÞnßîMØ0|¾¼:Ü’N„¤;IöE–t¾ú)t]š‘Ld¸º}È_žšQû€-Ó·Æë­o·ôµ…q¸:{›—…]èÎ3ëN××y­[ Â>»Öùû³Ì_‚ì¿{|{{uöf¯Äˆ­Øáz:ûKæN2ðFuG)äß^=pN5´AXÛóüþp{~ÀùΧ„n¦ë"©FLèX^ß]¾¾9\^Ýîþ Øh_ï‰3AéÀ¯‡ßã4÷g‡w™(a MÀš ¬pÐ}ß@£ª3Œj6ùFUgÕ«£zÑîLtÀ9ŸõB¶Ð@cp4tN4[±ˆˆ}X†5&©Êk*U›JÅÀ¦RU´©T m*U…›JÅà¦RU¼©T ofª83œJU§R1Ä©Tr*ƒœJU1§R1Ì©Tt*ä™ ¬uP!DÚ0Š,ä ç„ÝÜÖ:eѧÑ Räÿ°ºóÐ5,£<¬z,5ÿíx[œÉß¿€Éóî;Ò0ï­)Ò#…Št»:{ïoïnÞíaÑH©Ô0îX>óKHøÉýÿéËÓEƒÑÎæ­çTP(äq);$ÃpùžÒ7'Dˆ©ÒB‚N)¢¢¯ƒ¤úÍþÔK0ˆ5O+›·DLTyI.ìÙ”:é,òÌG ‘QÙyrx“Øåœs"„ðÚ›ŽÉyæí„ÎZÚy‡Y{P²Kgÿ…VŒÒbOgl:Ÿ‹«ŽúvF ›¶'…ð‚²Ø|œð-95€´Û e|ù÷)«öùñÝ);nÞ®Ç`•áëû6†@NIÍ£³>gÜ{uM º“IÄQz¾Áº“kúNŽåpAÓ³$ïó¼¿î”$Çò¿þÁ:®iT³¹Ö@«b'Ryl«/Úð±péõ)¢Ź´óCg‚FéazØoV¬`ü0BM‘<Aj Á ëœ^œ#IÎ@ˆ_µuRÃÅÅJLø)‘Ñ–0[îNa”ÖqÓv“’![Ú wtl/­Pnx—ÔÓ©B$VøN­‹îI:Ì - 7‡lt:Ú¢iæpWˆaFïØÁ¨¸#aÂXã.KÍ« ·º…I—¤J« |ÛÐðµ™Î¹¾K:èHT÷!dQB&åÍ#œÚµÒíÔšs›Þµ¯ò†àfÝð~q[¹Þ”Ùeº½mZj:¨tüH½å5yNŒ «ug&ÆÎuª³u-Œv “ì”5@M,H›( ž•ží¨©óCRáÊfàsîø€oñËq¯.7ÿ{Ö÷²Å¤ï6ïsÊ`†-¡.•õ ÁÅ–’~~j¨¶P8B ežN{}a…DKc´îʇlI6=Öe'±v€“?ž¼úןfyÃZÛÂÜîò¤~Fí´É|Ñ8ˆ/é'½`ž’0SkIçŒXC¢ïL²§!ÿ¨eLjʬÐû{k“gyké»’êô¦I.ë*òŽLQ9X~7æ˜VlZ)ù‡Ÿó6Cùâ‡,žú‘]ÎcòI&¤À&é¤D¦“áHt¾’z­Æ†qôAêyÉ¢ìfõL­%êâÀ„ºëÉÂù±g~ðn_ÜPЇE¸Ç¬¬D1VöÂ˾&‚÷mÈÕ»úM×{ŽJÜ>Ñêí1ûn&ô˜)BðGß)|ç‡K†ÆèÖetž›iÏÙ¨¡Â),0ØYÝøo:O‹\YèEø®R]?»Þ’²:ÿ±?µJ“XÒæ!r!f%Oë6¼mÃÇ6,ªŸÆŒÅuÞ=û[ïІ«»Ø`ñ»%ñÒDÙðІçÝÖñ׳XG„»6¼iÃw«·«ËÝ®îç¡­öïm¨V·ófuµOm8®bÕ´ºðg¶¶].ùp[³¿"¨”.V°£ž|"‘AáNçO`»äÜ}þÒ’}D§´Ï±¡!.2Ô ©¦i+èonø6ÍBK“¢±Ð ~øŒÎm¥„·Ì^\´Èg¡ydÓS +‚ÒdE6h¸öôad†½ÙL)¶Yrya° ß)ûÈ@ á›ëv‚‹T,~¯“ï$ q3nS_Q'Û0!R èXiW±•þQטù÷è¼ ‚“GH=GN4»L¤A¡cm%6eˆlž%Æ Ž2o§M9#š8±ŽB÷pÍúrN£( ™ÞìçÔºÞ ¢HY~‰4«*LyˆªK=y˜Âc…Ë»œš »î8ò pºÎÛ&ÛŠßëÁ˶£5›Ï2¾¦-B¿ÃÁë|È’ƒ¤‡÷G%1}lÀ#VrfˆR»D:U$E‰×¹Ì–¶)§Hvl±g;æð¢aõ' SK'¦R¨é;âŸiéˆãv§e_?‚ŽéºÒne¿ËO>EvúQåÃ0H!+•“ÍC&¼9ÒX5Ï»¦Ï .6»{Iª¦]±#}Ø‚•eÜpÈBÞ[´:¹÷£ò TE¦xÒ´™Š`=HÅdÒsJvS—o‰Ò‹:*10Ð2¬¦yºŒ•BRG (C² ô–u”@Æ©Q-®;%©Ä#ö泈ýL ‘¶’»n>C˜t.öÕš–Pg Ë÷“L‚º´¡‹I0µÌ¨`ìÂ⩹¾7ëv\oXså$fKãX½%+yÙWÙ2%?°šfNâ´2}­ù¨x=k`WñqT%ˆsXédJÉ1IPðDÕÄQÕ¤ï dEƒ‹9= U<>Ôäåš¡Â*Æ<ÌŒ‹r$ê)í(Û˜N›uÏU]K彚Ä•飷Ut®sy¥š5¿±ËÂ%¸/å͹òÄŠ<ÌÏ1åcµŽ<ësÝz ì7Š+¥6°|¢d>Kâœï2ozôªO;M0&Û'­ïÍÜgV§ø¸¾Cñ Ð.×›–£Zb¡Itv0B§ê賎 ÇM³½äSnšK—ºµÊËC³¢õêÁ(‘%ÍÂûÕjºñÇ6ü¿6ü¯6üÕj=qÐ^vÎxQ/‘‘§ªÜp¯©‹b)6\Öþ%aü}?½B‹ó=ω|kâ`{w~›9Ð5²W©Ü‹¼o©÷{¼*Q€Ñ›ZƒMOMä/ëŒ*GªvʨQ ŧ״R³×e_ÌÐ; 9Ã\¬ˆ:UÂ~›£äÓ©¸D¢ã±³¤y÷T“Ä —Ýá™`Tç§×¯«,AfÉøNåÆMé ÎÅ‘|ìøMŲw´]Ý™>ùr96±*ºÎûQN‡HÑË¢N7ëRßE©Ê4­ö&OeLISW#ÍY*N\µácò¬£Ö0PôÄI4–ˆ!¨ÃÃÑE8œ¬K„² ‘`ê®™Á:„Яmò)#+.’ŒGEó©vü̚üþL„ ~Îg¡ÖíÖ 1וÁ§öYýç [\P’¼Umt®Ù%LÒ¤Ô7ûªKl¿‹bB­TjmÑÞ'u•¨£§VP+¨Ál/¬úž;˧VWmøØ†Å©÷ÕõÞ©M•×j˜G­õ­}.Jz›brZÌøþHKœ±ÝÓ›‡,>#ûl%¸i p§É.ÞhY‰(%Œ¹Á¥Äz.Oà¿L拗ʰåਸ਼“y-p¯©”?ÿJnÙ#ì"…Tóx}ç Í¢<ÎxYœ®ü‚áL~Â9Ê­§_µSÉ\"³›rÀ.*¹èÒwù~zÒbË£ú ä,nr(¹ÿÏÍ ¾_ÛN×ÙMQ€U¬´ñh†£ü¡¼÷¥^(Å‹]8P+¢í­E*N“›MÙ*²=)iûvUY$U^UÉD¾+ëÁìÓtóîú@ol |GTçëÄY_\Ý^îSYšTöñí¡ü…²½ÔHŠ Ä1?òëµÔÚìö¨f±‘ûÿc7ET‡a³§gdpÁuyKßí!æQºã¶IyóuÚ·MÖê׬ִ¯ÕTdÿ܆߶ázøÙNòz–7­2û²40âpÖ‡•Ç'YzѪyدáz•ô|™Ñt°Û¾!]_o´lC_±ÏÎÚ-^ÜœOµ¢ß´Ä®ŒÀKùù4 …r*2*ëÝÖÛÜ”¿Úô¤sQÃ,ÐðãÂ…›Ê€—^&®¿£çQó/T%Ãbcð~'ýfÁäßœÿE2HŠÞñûé‰b†Ñ樢4•¢ïü\h(ã÷š}¡‘Úq¯¶T4Moú·_’mâ¾ rÙ¥¤Ë#õû…cκËi^£¯—^ÌjChÈ<2[Š!åªQ,ÚÀ­ªÎÞOåìþ©6pÚ =ŽÍ¥íÛ~xÅû¡é;iRí>ORž>­b%öâ¼-‚ÛÎ>‹åxí¶€ oþÒÒ¿M'“jSRŽTªcþ*ÇÔõå^îò¿3bùŒžfiWúá…¸qùX=}èDßrby×WéÞ)<´=mhÜò!¯£¹ûgÛRëo¹Q=Ì0\ø†X9¸OòY{x¾(Ý!«Ë¿Ç9€p˜{³†®Ö´—¾áe)¯œ?ý4êŽGïKv»Ö‹ä[ÚÄ*]oŽ—ÝèJÌ¢¿»xý›÷Þ'?ü‡(×ÍryjÃ^ì?†í'LÇá[“YiKþ€ð^ê­Ÿ¸”ؘÚaìwräÉ¥Cendstream endobj 286 0 obj << /Filter /FlateDecode /Length 3715 >> stream xœå[YoÇ~'ü#˜'ÏÆÚIßGŽØP !ŽCdÃX‘-›äÊZI–ä¿§ŽžéêÙ^ŠR¤q ‡=Ýu×WÕߎըþ+ÿŸ^ýîk/vGêøâè§#M¿=.ÿ^v+>³ÊúøäÑ¿ª“>Ž>ŽÙúã“«£ÁšÕɰػfqŠcÈ^89;z0|±R£QÚé0\"©• i8_­Õhs²> /á±ÎYÇ4ÜÇ9§†Ç¸"æ”=,X[kà‡8ü}•̯YØA>çÍð”–*åÒ°£§*x>l{Ø1¨8˜Õ·'Fžƒäy­ƒ“5Çkëdz~ [À Jeb½Ï+¹«ä³JžW8óÆŽJ…áãJ>© ¶o´ÙL>^îàSšùÍPé«J±bUé;oI.;ŸUò¼’—õµ+yY<®ä®’ÏúlÞ£“gò»J^K‰ÐÌÆ1×ÚŽÞeÃV½ßÕþYWåÂìßURدh‘è;Õ§Ý3®+yÚöª«:¡¯MWØg];}¾ªäç݃Åfï˜3éoh’µ£éxmâ¨rb“…ÞääËÚu×þ"-ù¦1G¬~RIUõù´+ì³®b¶,«S#f>v@ϲ²‚s#D^URp:VRWQTw«d®¤]­µØOÃjmTö£2°ïÌ´­dªb«út¬d¨d–OQlж¦½wtòÛ·ؼN`_I± ‚ÀÊEXw¸×É_'{¬¤—jàÂbœ‘•eró{GÇ«÷ ƒZOÅò²««‹ªŠ{]—ÿQ*¨Ë†ƒ°³*ó9À ×톭GëmžXYKßë+²GŽÒ˜Rÿ“02û‹<ÿ¨’WRËíg•ðÀè¼Hõ¬:¼[I`̘®¼º+N?ì™jØ}VŸÞ­äç•Üãa)®ˆ5á¦ý`;øâ -”3yL ¬\‚ó‹U²Ïü°…ÄjA7ª>§àí°½BpU0Fjmºj)ã­§Üït´€å.p¿r9J’ ÙÊÃ÷uÁ9h³à`vÑSN:x8D¼'àœähƒK´† r ¦<ýìcA˜ú˜¥±Êhã@šµ%("D8U&¨ü3u½‚ß;m,VXÜ.¥4üŒ“ÒQ"PV/¾_AÌ;„ Ä›RQ!.%YòäÞ6A BçIî¶9Å3|¿/ ´^¬¼5[Ó„Î!;£ëùµsfÒÁÚí£Yžâ D·Fò)Ò_‰=vFCºÔjN l–äb±f0>Wkêá‘ØpK¼‚òµ‘ø‹æä ès=7ld—œ+ŽD;ÏŽ4é.;ÕJ~M~iÁ:ÒØÌ&z•^vÓ›”;¬¥™R¥©p1ø‡UN)eݰ¹æˆ‡ì5:jÚ TøóêÖ»ò¢Ï®ñë3b,™0ä‹»°¶±&FÑøQO1úuÍWŸL¹"¬¯í®ºäºûÚWoËþPŸ®ëS‘ÕŸW²<ªÞk²ˆˆ+ùC%Ï_wÔ“¹úÿ³‡f>ªðAHÿ°’?tUyÚ5ѧÝ7u Ë}sŠ#6ËÍ–Ú-¼}ÿnuýiwÁ·+<¼¬äÙò^ØÝjg`Åûhß— Š¢—+néf§îÒ*ù›J $&õI%û þiWë‚‹óî¿F—Bç±ÆP…¦’ÉΕ´‰“ˆžG~ø–ìô^ d?ª]Í¿x²A0î­Dtý['0xöË&eî¨I&²zÔ÷îTR`÷okôö×¹v¿wŸ ¡«vUÕy¿H•´•T]ÒKVnßø C%s—l8ÛËoîu‚õÛíþSs@Ü·aôŸ¦GôÕ¯2Z„SýŸE‹XßS´yÚ^´˜w-}¶Ò{Š¡;sxB€¾Â¢þe…½ft‹2õÃÐæ¿X)jE½k:+ê‚”­È&ˆ»Uë“ÅVû¨´œ ´M,¬°ðÞâ1·ùpv3hÛ+zÑg¿ß)'àcÑ*ãŒÁꜩ3‡&›Gìšy${÷ì=[Ñl¨:Øò¯¡±m¦[ê¡+¥$Sæ¹U¥ÎXHFòÃcn£Å¸3ƒÉ1¸fh°ÛM“ e›ÉƤ¢‚WŠx‚gÉ!pðx,¡-·¤ý ¸È¢ûŸß|ÊȦQG3 €ŽtJìt6k7Z ]Àpßã5¼æ¢±üV 7Ê);€Ž„úÛsLƒä3–]è ù†õ9&¾,ÏÑï¢#(•—€&µ®Ïø9x¥^—½öüù3ñêåjæ5¥ñ#WÿÈÍ<ÈzáÑ€d€è š#æH’ì+š!œ¦=縳ã^ß.Û¶©ÝÀÛyžá|ót%çj`ñœmÐånU[×:äã¿/Ù0gr}›Gx;MçTW߀Ã9\(SÜæ…b;édb…¿”™úúbx甆5ì¸äWOÁ…óbÌSFIpÈ Ô©Qfy9ŒlF… eöWÉ*C§ˆû6³xZÖÜLŠ÷dš¡HMΘF¼¢8œ{bÆaQóðeð{xwg•³ÁÌ«!hiTVnÊ9íš B3Ž£µ¨„øéu?LeÂàôªS;‡<¢à`9M`”’_!€x  "œ¡[F MQ¸.Ù‚ÀÚÉ.˜QTù†ØS&Ùæ@FXt1)ˆû39—Ûò»Tevâ9')ÈQÎâþ4µ‘¦¸]ÛR;[ÂeÃùÔ™¬‡4µ¬.{ "W^~x ¯p¢rvLÃÉ CÁ‡`›Âþ–5—IägŸb¨{Y§ÉìÜǤó^ô’°{³d‹ì´f_Wl0k†d|¤13tYà¼0¤ì¦° oŠä+~ñSû½ï4`ŸR‰ñL‹ÞæÐÛ&CY¨*äkûDFõí¿%÷sQ ¼å\ ’åç$ŧ±åŒk3»èÔU§rC¹\ °J>(áV5~LÞIûá.>žbx=1Âc¨Yo…'Ÿ5&›Ñ:INc ü²F¶UŒPð±Ò†æÛ\ÙòHžb†ß<\øKÉ=b‰Àe´]†íðö‹[žï ¦›gC*ÖÞQ€Êô+Å|g#Œý‹X1³áor[XÁrÛ ¡9+IV8™èöœ‹  ç(–ÕHn@| –v-“BÎËnÊJ +|_\6m›Ù»ƒ_Wí&Á7a|gBô)ÇH0#t}S0G¤!ó‰ ފ܋ڔZµ§(€S2øÁêÀ…“®y פÜ>×ìã…¡ÖÉ{¸Úødÿ”¼½ÖFz­XÞíß·ˆnÚ.Cö5¢'¹Õíä\6Î&<ÝIà1â-ºwåÄÄ €=Dvq¤Š)—>®ú›.6ÙÚ3*ãÂ»Æøóµ–ËÀáSyÂ÷| :J6¹lØÛu&˜u@‰œkU¾ã,d*kY{Ó>.bê¦eÜHÞ9X]ÄËt°Sâê4Ðmªl ›<¤Ó^q5ø@Š»®e·PTÒ˜;Vº²èÁ6óê ¤àÀ81äT“\9¡†æ#Ë‹‹óšˆçË;uÒŠ_¦®ê}â¼×6ÀÂÞ²Õ?­ªc1xt˜}¦Ì$³Ñ¶U+ß7KŠa}2 Rn.®Ée¿})6Øl#4ÏŽ¬sb_™ñžW/—Ѻ®z)×Ï,]›æŠµ´Kòn¹l€pº|Míiˆ9§–MOŠQn²?¨O§ì µÌS‚ë¾ÙšÏ6*”íûxon )qn:ޤ¸{—Aâê®#È §ylÚö¼P¿ {˜N¿¡AYòm4›’üô¹œt¯’ÿ¨ä]¸`¶ù²ëÃGt(3}Ûcb)-8~ˆÒÉ`,‘(6ÎÌ>äö:GEÉTâhkÙ^ìêã;¼ ÅTq_-_S¢ábœ†…7.9ípŸá¤‰:ÑwkQEí |v‘üvDµ6ÊYŒ´Û‡“² WY&JÞËp¨(ô1~nªâ¢GäÄOs8Ÿ£å°½‡àý `L\xTJ^@ä“[ÀÑ~Óƒ3w‰7`ìí¡y¥|]Õ›× fPéÿhÞJÆ0 È&Ñq «íñAõœ–³I¿fjCqšæBíC‹=ÓÔºywcç†_9–Öƒ£a[ô™— ŸYk±v®É´Ívýn©RÌz[¥ÎÅÐe ço©¾íÀ Æ((Œ¼ÙÉ¢$¾0Å…O –=_³à5·Ü•Rµ?¸ Îï÷RäÊVpç‡/ëê{÷åH¨0rÓ஢|Kãöè‰IÐ+^6í·…·‰À8í~"`£ØcþµßëúÃVrË‚9É%ÝŒ‚ÁçÑÄ9¶O³XHt>lfžÓ ïcæI“m-AaYÎ#÷ŒAã—Wj4g¶ï(áñ‹½Äož>/lšÔÞ s2ËA'ì’ÊÇJ½©‹èîÏJQ9{ìxo§Í-ݚĹñB V h—ÆÙ´óO§àÏ"¦}lº7áo°iÛþ•´mb`‡×ˆ <÷¬E@îY ëkb¨\»ìùÙË6‰P|X×NSI¼Þ­pw Ø´Fåš1’O¡Èøgp”}<}&ÚOß²š,¦ì-&” ­(DË!ý$E‹“äýe3Ï+ÚfžXÔøMÅeJV®ç»ç¶™¯·‘Ë2âå4mÐŽŸ¢"ÙDˆ Ù>’䘛û䆓%€3MV©v¸Ì–4”Ë{—]R–%¤›n¾†5+GGt[š¯©=~W ¥¿4ßcãï5äI õ"KY~øÏ ¹?í|Ê;ºª°øiQ€¦¿Öší@Aí5T¼)(DæÙ®…÷ŠÁ”î›§â”Mwº?drÌ·œ×³#™o&·ËóWâÉì#©"`‹¤–ó˪8ãá-y“Ðúá¸Àýéäè¯ðïßôö²&endstream endobj 287 0 obj << /BitsPerComponent 8 /ColorSpace /DeviceRGB /DecodeParms << /Colors 3 /Columns 826 /Predictor 15 >> /Filter /FlateDecode /Height 427 /Subtype /Image /Width 826 /Length 57847 >> stream xœìw\GÿÇ7Aņ ööØÐXˆ±·DQT¬ˆ½÷ú³E[4ÁÇ{ïŠ1DÅÞˆ »±KP *JÀ‚"òû<÷}œg³{wÇ5¸ïû{íîÍÎÎÎîμ·ÌÌgýõ—Ä0 Ã0 ÃØ*Ÿ±®1 Ã0 ÃØ2¬k ÃX”9‰‰‰ÖNäIpùdÊ”é³Ï>³vBKúÆ0ŒE‰ŽŽÎ“'µSÁ0i•Ë—/W®\ÙÚ©`, ëÃ0å?þÈ;·µSÁ0i•ß~ûíË/¿´Ì¶ºvízýúu~˜§àÝ»w(Q¢„%7ʺÆ0ŒEºæêêêææfíä0LÂtçΛ7oJ–ÕµjÕª]¸pÁ2ÛJ[ܺu«lÙ²–Ü"ëÃ0EèÚÿýßÿýë_ÿ²vr&mðÃ?|óÍ7ëšmÀºÆ0L:‡uaŒÀººUøøñ£e6j³|þùç“&Múþûï%Ö5†aÒ=¬k cV×5ËlÑÆ®M:Ub]c&ÝúÆ0FÀºf °®1 c/°®1Œ˜P×’’’É”)SvîÜ©?$ëšÖ5†aìÖ5†1SéZXXXãÆß¼yóõ×_=zT`Ö5¬k ÃØ ¬k c©×µÁƒ?|øfY׌€ua{uaŒ 5º¶}ûvxÆ7ä Y׌€ua{uaŒÀ8]Û¿¿ŸŸ_xx¸XÒªU«àà`‰uÍ(X׆±XׯRªk/^lܸqLLŒXRµjÕC‡9;;Ó R¬kFÀºÆ0Œ½Àºfrt•䨘3eÊDÓ?~üóÏ? å @RREåèè(ÿ ±šÎ˜1ãçŸ.P¬¡:%Œ¨k8RW¯^­S§Î»wïh r¾bÅŠaaaâ±® ëÃ0öëšieÉ’E®Mr .üøñcšnÕªÕîÝ»1Qºté»w絛^½ºoß¾4}óæÍråʉ¿*W® ÀD"##©¾÷öö abccsæÌ)O˜ÜÏÊ—/¯øvŠI†èÚ‘#Gºvíúüùs±ÄÓÓ3000Ož<ò`&Ô5,7dxƒ¾z*#4Ö5†aìÖ5Óòúõk]º†êöܹs4íêê‰É“'£ÖÑ>GŽoÞ¼‘þ®t}úôY³f &2gÎ @Z^²dÉû÷ï‹uÏœ9S³fM1[§NÓ§O‹Ùo¿ý–ê9Æ8ôëZPPиqãîܹ#–´k×núôé_|ñ…:*Ó>]Ó¯MbElçÎmÚ´Q‡Ä9†3¦{õêE§;ggg1“+W.ùvgÏž=räHE„Ù³g‹‹£éáÇϟ?_ 44YŠó<::šãnw)3gÎôððк;¬k ÃØ ¬k¦åÉ“'EŠ¡é)S¦ÈvD äîîîããC³¢z ‡ºimóæÍ]»v¥éóçÏ£ÒZ»vmïÞ½iÉÅ‹¡¬¨§úé'???š>r䈧§§ü_ŠÍ¨]dþƒ.]Ûµk–ß¼yS,iÛ¶í÷ß_¾|y]QYW×àV­Zi ,âJšF^^^4ýÕW_8q‚¦Å]|»ÅŠ{ðà<*Ü¥Ô¨QCÌBæ tòTuèÐaÛ¶mÉ&[ëØ’}ûöᾪ^½zÖN£Ö5Ó²jÕª~ýúa"W®\òïÊ\¾|™L+C† ‰‰‰º‚}øð!sæÌ4ííí=yòdTØ4»{÷î-Zˆ—.]ªZµ*&èÙ®¸ÐÐPL`ÖÑÑ¿Y³f_PA+ *”º}µkÔºvêÔ)8Gdd¤S·nÝŸþ9Ù|6“®5nÜX 7ÇŽSÓ£k’æ .NBL899½zõ kÍ;W<9óõõýå—_Ø쑘íÔ©²BÌ*t "xðàA𯆠^øMHHˆŠŠ¢‡Ê¬k cF^¼xѦM›Ó§Oã2¦ûÀ¸]Ë—/Ÿµ“ÆüÖ5ÓÒ¤I“C‡IšOÓPê †:†^€:;;ãè‰pÙ²eƒ R,œ>}:¹‚@Ô[¸¾J—. u>Ur~~~ ,ÀÄСC-ZDáß¾}›-[¶”ïó_„®]¹rE\­Zµ„ þùç*T8wîœPmý˜I×ô넺¶sçNñ<VŠ­gÉ’…ÞT¢Üxùò¥:Bºa¿÷óæMŽ9DE ý]×6mÚÔ­[7šnРÁñãÇièÞ½û† ´&uaR.K\Òïß¿×ú/nñË”)sùòe §ŠÑ ëši•V† 2eÊ$÷Ô¯_¿Ÿ~ú‰fÅ—dÿøÇ?þýïë‰0..UhÑ)i^E‰g$‚5jPíX²dÉY³fQ‹%...ô¦µÚ³gÏ/^,i“І¥Œq]ƒpË£âÈîß¿?{öì†GEçL£F>¬?dõêÕÏŸ?/YP×@ëÖ­©g8€SZœ98òçϯŽÙ2cÆ yP¶@Î ,øôéSZ"×5¨-½;–7Ä1Ö5†I¢MPåÊ•QñÄÆÆ¸¢››ÛªU«PY«U#±®™”—/_æÍ›W׿8ÛÅ·Û¥J•ºwyr°~ýzýÑúûû‹ï²‹-*†-’S¼xqÒ¾^½z­\¹Õ* oîܹsæÌyô葤¹NkÖ¬&iÞL³‡Ì'„®AˆåÏG¿þúë-[¶¤è•~ð{ÜÊê w§7݆èZµjÕä–UΜ9#Z®k’foöGŒSKëvCCCqgr÷îÝ.]ºlÚ´I¬¯úî»ï(Œ\ײeËF&ƒ‚‚ †úS¢€ua’‡ëàÁƒ!!!â™qŒ5 Q±+XÖ5råÊ•*UªHšªK]˜ß¸qC|o.ZÉ%[YâAE½‹ê–¾%R " nÑ¢Ž)*Hñ›`6>>žÂ´lÙ2•;kç]ƒ(@MF-ï¥M›6¸štµ Q`ÄͪqM pô½I†éš¢¡@ÕªUé¬Öí¢FxðàÁ!C2gÎüöí[Èk·nÝè¼aäºfàã@­°®Ùü^Àp4¬_¿·D¾¾¾¦Å_éÒ¥{õê•þŽФŒ3Z;Z`]3!Û¶mkß¾=Më)Ìå ô—ù00Åå€ÊOñ)œL<˜¡¿üòËëׯÓww÷‹/¾zõ*W®\´ÚæÙ˜†P75€—L™2åöíÛ"Lçαešþ¨ÈWœ¡ÔúÏhÓëׯ%‹ëš¼SJjtt´bCbvïÞ½Mš4¡ìرcýúõ‹ˆˆ F ¬kLª ¯ ­Š4Ãüùóýüü´><0 æ‹ÙºàC}iíThuÍ„ 0`ÅŠÒ'CÒ w;={ö”4/‰Pí鉰\¹rTý—*UªjÕª[·n•4×J-yCuëÖá&G’}ýýÓO?ýóŸÿ”dOãÄÝ;¬N!0F£«#˜¼k\ooo莞·ää+ 6ÓQ#†ððð ÚW+¨óöíÛ'i*BT™ð°üãô׬Y³F%B6kÖŒêNTçôuZBB=G}ðvéÒ…¾WË—/ß³gÏøkÑT¢¿›ÜK—.ÕªUKþX´Zµj¡¡¡ZÛŠÚrËP >y„âã¶AŒOüøñ ¨#$]»wïÓ’5jœ={VF®k➤H‘"ô¥á°®ÙB׊j°vrl”wïÞ]¹rsæÌAaò×  êׯ/ï_4í‚7ru͵‘žNhQÈçȑŦ³dÉ"o['i†Š¢Æq]»vݼy³¤ù=::šÞtèÐAts%zâ@„Ù³g§µ±ÖÆ)À¿ÿýok%J” ÙJ•*A,$YÝɤ†d¡Bæ£Ìlذ¡x¬ŽÒgÅÄ[i–u 7£G¦é~ýú­X±ÂÙٙړᬣ!7’®IšÎ`Q’æ#víÚI:tM>ØZÓ¦MéE΀–/_®5y¬kö…Ð5õo¿ýÖÚɱQnÞ¼Y¡BLlذ^©ìÙ³wŠ7N}äâhpp0 ¾âÅ‹§>NÛzg]K÷ˆŽjÁÇuÝû}øðî¥ëëLJ¤iÓ¦‰‡sòzH4e4• o©¯ÝeË–¡nÓ3¼0** 8Ê8ÖÆí##0pˆwpðàÁAƒÉǃ”,]º´X±b4k]]«U«VÁ‚åÅÅÅÑÃZT‹ŽŽŽtçœ7o^:`c¢fù \µ®¡Ä ÙéããCŸ²iÕ5éïÃ$ÀeQfâWJdd$ «ÀÝä2ÿAèÊÇ)S¦X;96Ѝ'„®IŸ‡¢fZ°`n¤~yFÝÜÜpͧ²m©-CO5X×Ò=¸ Ä Oz:¡•ò¯&_¾|¨ÈEǤ[·nmù¸@t­ÅÄĸ¸¸Ð’»wïêú°]Ô”<¢ÀáºFàPNž<™ž6;vDóÅ_XW×´Bë"m¢¹+ªe4sR ͪ_¿¾¤M×ôlT¡kþù'Pñ1œ:=jX×ì Ö5CЪkjÄ—°ÉR§N7oÞˆŠ'úÆ0é”ê±}ûöQ£FÉ;FîÔ©Ó–-[$ÛÓ5ù`ŠÙEÿm’æ„- YÅðhZ7Šõ-úÁƒÇŒƒûÿØØXJ<îCJ”(M×®][k„¬köëš!¨kô¼í÷߯Zµ*lL}>#«Q<-]ºÔ¼Éµ%X×&ýaœ®§Nòòò¢~ò&Ñ5Su9žVº.g]³/X× Á@]“óúõkù-×Ó§O3gάç5Pz…uaÒ©Ñ5IóYØõë×kÖ¬)ë3ÕÓ5»‚u;`]3#tMJ;·hf…uaÒ©Ô5Á¯¿þÚ©S§ÈÈHÖ5#`]³/X× Á8]c$Ö5†I˜J׈;wîÞ½{õêÕúƒ±®)`]³/X× uÍhX×&ýaZ]3Ö5¬köëš!°® ëä?X×lÖ5û‚uÍX׌†uaÒ¬k¶ëš}a*]Kö³zý ÿ*ÿþýûÍš5ÃDllì³gÏ$Í(õ¸Œ%McÌÈÈÈ”¥Û0R£kùóçwuu=uê”9fû°®1LúƒuÍ`]³/’Õµ¸¸¸ìÙ³ëZ½dÉ’4-| þ¤u1 X»vmÏž=Å_?–ZóáÃy?„rnß¾]®\9š¦FÞOzŠN¡U«V?~Hž†ëšðNììþýûab¨//¯Ì™3ÙU‹Q×µððð2eÊ`…赜a= øºyó&ÁzèÐ!OOOËl—uMëš}‘¬®½{÷N×x2ÀÍÍíÚµk4-D×48¾¹s玉‰¡Yùh0 xþü9EòôéSÅÀÏrL¨kýû÷_¹r¥¤FšÆÔ…ºvòäÉßÿ†2Ô ©Ûºu뫃 2<©i×µ«W¯V®\ÙÚ©`˜´ÊŽ;Ú¶mk™m ]»qãÆëׯ-³Q›åóÏ?_¶l5§e]³ R¤kÕ«W§ñn¥J•Úºu+M'«k ""ÂÕÕ•¦¿üòKº?óóó[°`-Lö”Z×°¹úõëc븀ÃÃÓßçO˜J×Ο?Ÿ)S&#jýéÓ§7nܘÞä¦Wl\×6z¨†óÅŸµ“Ã0i”½¨5$+飀uÍ.H‘®8 ›]Ó¦MöhzÈ!ð³’%KÒ¬‡‡‡zÌÍ$Í+ËŒ3ªu ¿ñññ´$K–,bE,‡„a×èå#*c ôïû÷fÍI£kqqqþù§"V]KHH7nÜܹsõä‰dΜ9&&ΗþŒÁÆuM >f̘€€k'‡aÒ3gÎ;v,&®]»æææf™²®é‚uÍ.°¼®!’zôˆ¦õè–`¹zÓÍš5 ‘ŒÕµÊ•+cZÏ®¥{ä_ëä?¬¢kŒíÀºfRÙÔ@~È ×5°hÑ¢aƉٖ-[ËÈ[¤:töìÙŽŽŽ»wïnÕª•bëZuM$æôéÓµjÕ’4ÅÊ–-[îÞ½»mÛ6íâÅ‹%̓·„„ƒÚä„®1ÿþ÷¿ÅÓMC`]c˜ôëšÃºf ×5Ñ!…£umøðá .³eË–½uë–<ÀŒ3&L˜€ HÛ›7oÄòÑ£GÏš5K¾u­ºæàà@_Â"p£FòäÉS¤HEŒhj°sçÎ5jôéÓgÿþý’Ž+d]c†uÍÎa]³©üvMþ,Êp]{ñâE¾|ù çÍ›çïï/f}}}·oß.i¾‹ŠŠË/^¼èáá!OV]ÓúQêàÁƒçÏŸ9£ÙÔ· Íš5«££cê?5+X°`\\œÖ··6Â?ÿùÏŸ~úIb]c†uÍîa]³–ojð矺¸¸P¯9Y²dqvvCÈ¿‹jß¾=½µÌ“'ôN¬~þüùêÕ«ËÓ£U×0qâĉQ£F]ºtI¾uÑ{ˆd¢Ž>>ðžzõêéÉFÛ„uaëšÃºfÌ¡kZ¡uóæÍûòåKZòèÑ#ñù‰%|X¬X±ëׯW¨PAk­Í\R.ä 899………åÉ“çùóç8]:”š8mä tšr3V‡uÍÎa]³¬k†`´®QœújØê°®ÉÙ½{÷íÛ·K•*åãã£Ðµ“'Ož?ÞÑÑÑËË háÑ£G±0&&& ‚H%$$|ûí·’æëÌwïÞIšqši®Ç¬Y³öë×gMÈÿÅØ»wU«V·n]Zxâĉ¯¾úêÚµknnnZS¾uëVXcþüùÛµk‡8±döìÙ(Ú´isøðá: Â]»vÕªU«víÚ’¦Õ6Î^}¤5iÒ¤|ùò"¶K—.:u êÔ©#ï•uyëÖ­ããã9R®\¹-ZÈÓpìØ±Ë—/gÉ’ËÅw¨ÁÁÁ¸ÊP mܸñÅ‹]ºtA"%M/<‹/F`ì)’…"÷ŒæôéÓ0ÚL™2aJ—.M çÏŸ4Só¦íÛ·ãã_lTÒ4-‡ûNžIó¤î(Nc¸à®]»P "˜­W¯tW(¢={ö¬xèEº¶pá¡C‡b1 žgϞѱ&]S´vGœ¾1cÆ`ë>‘C©SÉ7víÚU‘¼;wúøø@áX´Dñl©mÓ¦\רñ>Mxx¸ôiÜ—;v´mÛ–ò G“2§+TO>³Ñ°®Ù9¬kV€uÍ Ô5Üm£ŽI¶4DM‰24õ#ZÖ5‚êéwïÞ‘pܾ}»\¹r¤k¥K—¾wïu#izõƒŽ¼zõŠº_–ŒÒ51Þu%M¼téRy`„${£z]kAг®`Á‚Ø:Í"H'tŠt «Ð£ ¤Ÿ“R‡Õ!$:¾Þ·o_óæÍ±#Ð&$8**Šúß¡l„ŽÐˆ#X‹lOúÔ¿% g®R¿¡¿H×äý&bëHƒ|µêVA!†½?mÒCÙ²e‘ bC$7BDBŠòtMŒ×g…mÞ¼¹sçÎÒ']S<ÔÄîÏ;Wq˜è4Ã1…b-\V”aé÷ööÆÁš0aþBá¬H¿Zרïñ#GŽà|“>u0ŽÝ¼uëmg&Â`+bu“Ô³¬kvëš`]3]ºF7¾(4hpæÌ™”F‹b¥¶è¶×–I뺆¬¦< RªkEŠÁŽ(ê`Ò5­Ÿ'BÊEk„®A’èyÕŽ nݺ⭨ ÿþË—/‡={öLkAŠØ*V¬H}ˆHšã w)tmÓ¦M8·i]5jÔ¬Y³þ-Uªvú½Sä½Î¦…ÎÎÎ8.ô}—†œ§¿páàŠ“7UëÖ­\¹2NKñÁœ"Ï…ºµmÛ6((ˆuI=¬kV@èšx·Â¨¡Qú»®-[¶ 0 ™e4b”*øî§Å[#êÚ±cÇjÔ¨#ù¼{÷î–-[Š)ÕµªU«^ºtI”TÑÑÑ8d¤kNNN¨¶QÍëZ×Tºæãã³sçN­¥%½”/鱕/_þÆ4‹ý={vJuíܹsÈOˆï| ¢;  °‡îÝ»Ãá$ݺ†›“|ùò=|øPrCt­_¿~°]Ùk ¹råúðávYë¿Ø5ú´¿E‹8UÄr…®ÑH*pG¤¤C×™ÈRý•G`` Ü®¦È–°°0 _~ù9ƒ0b¹Z׉››n)q­Ñú^>¼c]cÌ뚺ֺuk\Ì|´‚’:+i¾,AŽ!»zöìi¦mmÛ¶Í$—è¡W¯^)]%êÚæÍ›a0¨5½¼¼nÞ¼Ù¨Q£Þ½{£vRªkô’ø@ž¤‡t n@_ëS`ˆ»»»X×TºvàÀ¦M›B&D‹KH½P‹‰‰qqqŸXì5µåľ£F}°áöÓPCtM<¹mܸ+ž={ÖÛÛ[ô¼nÝ:œTò若ê¯Y³WüÉI†!º&tPÜ+88ØÓÓS m¢Ü”N™2%F®‡ ×uÞ¼yccc¿úê«£GŠ]>éÚ¡C‡°!Ì¢4˜:u*V¤‡îZu>’ÃåL­YÁýû÷q¨“ô™l]ë_H î:ĵ®Qc‹=zà(Hš+´xñâÈ:d ÄºÆ˜Ö5+ tÍÚ I3(FmOs jßWNÕ5¢k×®ð¶"EŠP?rŒhŠª111ÎI6lb ]ƒ¨‘Ó BÅ5ƒ*ÑH¸„Z×p"¡ ÇÄܹsáXÏŸ?§›r/]º&i>ÒGõ©Q£ÆÚµkgΜ)þêÒ¥K``àèÑ£ûöí»}ûö &Ð_Ô¶´V­Z«W¯†°Âº¨m£!º†êÕ<ücäÈ‘ÈL¬‹…+VÄîÛ·‰„ #ÍâÑ.]{ñâŽt±eΜy̘1¯^½¢6ŒÉêZ¹råîÝ»‚<Y$¥¼©v¶`Á‚8‚8LP[¨Û¹sçž>}*iÞTnݺ•d‡ ÅÙNºV @(é³gÏ|||êÕ«'F´ÓªkVódåÊ•uêÔÙ°aöˆÂ`ëÈÆ"Ú¥K—Ι3¢Oˆ@ïN:…ÜÈ‘#ÇàÁƒOœ8!®FDD [>ŒïÞ½ ]ù@¹‡+n×¼ysäêëׯ)çY×s¶kÁ4 ëš½aoºk ÿõ×_ýýý?~öìYjH¡k(¦Pí7jÔèСC¨/^ŒšUú40îªU«° ÈP‡ä+ªuíÝ»wâQœbÒß›(tMÒ<8DYBÑsê{Ô訡‘ùH\–GEEùúúÂÛÊ—/¿nÝ:z¿oˆ®A†Z¶lÏ“·×™7o&†>uêT±k-[¶Œ¾ÍR¼NʇÀ>'''„Õº†m‘>|èѣ܅º©!Sªk²B™)S&ˆ25Ïœ?>ÎOOOêU˜NŒÂ… ?~üXú¤kX?Æ™ŒËAþŒVO øržêçç'†d@Hl§ ¬T©RÐ52IÓÌsÚ´ip;¬ªU«Â­EGwz¾ü“4çøqã°!¨$¥D–~~~Ë—/ÿðáCоl9rdãÆ›6mú—EF©²[]ûK÷P`¬kúa]#\\\bbb8¬kvëó¿†ôÄܹs! VLOJe̘Q|s­•ÚµkC˜Îž=kK“§Í>uM¬k†`áÕfá|°®Ù9¬kÌ=”>½½²Ùj>Y\]]©¡ŸX’-[¶ ôîÝÛZ…>ëšÖ5†1Ö5;‡uÍÞ¹téRÕªU%M_ ð6êþ[41K‹¼|ùrß¾}:t€·‘XÖ55¬k c¬kvëš½Ó¼ysȤбdÉ’46¶““Sll,$ÃÚ©Kó°®©a]c#`]³sX×ìšk×®U¬XQÒ´0xöìÙgŸ}6dÈ%K–Hš±Ú´icí¦yX×Ô°®1Œ°®Ù9¬kvøjM\ÿ4¸ ý+F'dŒÆŠº&ºµ5âââ²gÏ.±®1LJ`]³sX×ì—§OŸÒ>®®®ááábù¨Q£æÌ™ƒ‰½{÷6oÞÜjéKXQ× ÚeÊ”±Á <))éÎ;’Qº†Ýyùò%n$I‘"EäÁM3gΜ'OS¦U/ 111?~Dª‰±0‘‘‘ê aÒ¬kvëšýR¡B…›7obB žM“¸ó§TcE]Kéæ,º–?þ¨¨(šVÎŽÓ'³ÉÒ—°Ã>hMŒ…¡”ðÕš¾a]³sX×ì”ØØXgggL¸»»_¼xQñïàÁƒ—.]Љµk×öìÙÓ éK/XE×J–,I£˜Û,ÔeLJu-,,¬fÍšŠ» ÁÕ«Wœœ°ïZ7·aÃÓŽ`ŠÕç“'OĽYquuˆˆÐU\߸qããÇÐt ¤„±VÔ5îýÎ`]³SÊ–-K/¤öïßïå奸7..5Î\¹rÅÄÄX#é«èš#nRªkgÏž­U«–(² ¯Bl×®]­Zµ2"µºèÑ£ÐðÄWትèY¸¢¸VÄÉujúƺO×øÔÒJÖ¬Y_¿~íàà`m±®Ù#?.Z´¨¤{êêÕ«ZômÛ6((óæÍó÷÷·húÒ¬kjŒhêîî~ùòeÅB*»±¢AŒâ|†NAªÔ±É ½wïÞâlwqq10ý¢ÞÂaMJJ’ǹ}ûv__ß„„„Œ3bvá…ǛÊժU£Æ×ŠÝG¡ÊxéҥأbÅŠmÞ¼¹víÚº6÷ôéÓ H¯DÕ…9*øï¿ÿû˜7o^L÷íÛW¾+V¬Ø±cǾ}û‹-J»]-Ú ¬k6HöìÙ_½ze™N¯X×ìTgΜÁÄñãÇ4h +X¦L™m¶aš€uMÑyà¤Å©«ë–žo×ðî=Z·n­þ‹¾Ô„äQ6†Ó½{÷7*³{÷îV­Z‰&Õð!8\×ha¿~ý|||vîÜ)o̓d Üß¶m[³fÍnܸQ½zõ¸¸8Ÿ®iýv wYó5@ÇŽ;sæÌõë×#Ù"1%@ýúõéYûÇé.ޱMlA×p¦U®\™µÁÁÁ7ïß¿g]cÌ.uúÆ%YkÙ²åž={0±`Á‚aÆY(}é Ö55FëÚ¡C‡š4iòîÝ»,Y²¨ÿÕ¯k›6mêÒ¥‹ú¯gÏž+VìÂ… )­ÿ:vì¸uëVEù™¬®¡ª£Ç„ƒÛ¡!C†,Z´³sæÌ5j”ž¶ØdTºŠk­º–?~ìÔáÇi¶xñâ¨Z®_¿.‡ûñÇ1^¦L™“'OÖ©S'E™ÀX[еR¥JýþûïÞ´mR®\¹Û·o³®1f¤iÓ¦4}亹¹éL—hžúùù­\¹RÄìëë»}ûö+W®hí\M«®ÑgÔ3ÜÍ›7!‚¨ì§OŸ.¶Îº–¶0­®;wWIJeË Ïº¦€u1#Ô#nXX*׊ŒŒ,\¸0MóÙ’RX×Ô­kôòòRèšÖ»ŽÐÐÐzõê‰Y¬uäÈqä§±Ñ-C;tèðË/¿¨¯77·7nÐôd z¾]ëҥ˦M›höúõë5jÔÀÞÑìèÑ£gΜ)¢…åÉ“çíÛ·4ûüùó|ùò7. @‘€bÅŠ=xð@L?|ø¦½œ`ë:u’X×Ò&Ñ5œááá8?~ŒÃƒnàŠ¬k X×s»|êYJ´M3Ô³fÍÂÄÖ­[Û·oo–ô¥SX×Ô­kÈCXKÖ¬Yå …ßÈÉœ9³¢ÅŠâÈc@øþý{uødA„?~Ì–-›b9"ŒÇ¯££#~Äæ`EØ–˜Í!‚‰u“’’¿&_N ýø—Jl+ªÏD Š`X.t‘NDˆ½C„rµE¾eÊ”ItY‡Y°LoŸŒq¤^×PÕ­[÷åË—t±®¥Ö5Æ\з̒æÅ¢àNœ$¨Z$M-ˆºÍ\IL°®©1ZׯžI®={¶[·nò…¬k©u1 /^¼ÈŸ?¿IŽõÂ… ‡šúxìÖ55¬k cÆéÚÁƒ±–|`hooï½{÷J¬k©ƒu1 'NTã.'))‰ZŒ–-[¶xñâzB)RdÅŠ¦M^:†uM ëÃAJu-44ÔßßÿÊ•+bIýúõq¿uɽX×Rëc>~üHt𡦅uM ëÃáºvýúõÖ­[ß»wO,qssÛ³gO±bÅhÖºöáÃÑ®EMŽ9òçÏOÓwîÜA‘÷(S¦ŒÖÀbxÜÒ¥KëúIruuÒ¤Pý©Ïýû÷QÇÑ4»+ è+Ï\¹rÁÃtí ÁºÆX\iôôܹsa ÖNNúuM ëÃ!º•©]»vTT”X%úõ×_óåË'f]}°k¥gÏžk×®•G%éîj`À€âŽbÄÀÀ@Ñ¢èGîgÔs<6j»#fcbb`cŠ”¨éÖ­Û† tý+±®1Ö‚uÍL°®©a]c#ЯkW¯^õöö†¬ˆ%îîîAAAE‹UGeãº&}*9iZ²dII3ZZΜ9éÑZÁ‚###Jšçˆ×®]“GÕ¦M›]»v‰Yu 4oÞœ>òÓ ëcX×ÌëšÖ5†1]ºväÈ‘‘#Gb¡Xâéé9{öl=] š[×BBB¨3AñâÅ=<<äQIzuMtê.ɆLD$( iá… ÄæÖõôéÓ ˆY•<1Zumøðáõëׇ†‡‡ûí·”0ý=!°®1ÖuÍL°®©a]c#PëÚáÇÇqa jîîîú£2·®é÷ CtMú4ä.Mûûû+V VJ³0ª©S§ª#tvv†ýøãcÇŽ¥%Û·o÷õõÍ‘#Çëׯi‰V][¿~}÷îÝiúþýûô0O Y×ëÀºf&X×Ô°®Yø)ês‘±„®]»v %6ôâÖ­[â_¨R`` úÃ|­Ð9P­Z5(‘!àèèX¾|yɲº&iÚ¨7—?~1ô­"Â}ûöÑH‰3~3eÊôñãGÔk#FŒ 0Éêô‹d̘1!!AWÚX×ëÀºf&X×Ô°®Y˜›7o¶nÝšª=ëø:uúù矹Ò1¡ky󿥗ƒ ™ãÇË_ÿ%‹ÑÊnˆ®!rõ!–/1\×hà…3ÁÕD#Su„ð³ÄÄÄÍ›7wîÜ9""‚šŽBÑœ)ŒV][¾|yÇŽ“’’°-[¶¤±RêÔ©sòäI]ic]c¬ëš™`]SúfI>|X¬X±fÍšmܸ•“¨´ÌÄâÅ‹G­ønI€qëÖ­\é‡Ð5GGG”Øb¹§§çRd fÕ5­§k 887bvÍš5½zõR„‘GØ®]»;vй7>|øpõêÕÃÂÂD›T©R{¤'a¬kŒu`]3¬kjX×,ÉÉ“'ëÕ«‡<7·¨sæÌ5j”þj…_ȇе˗/?{ö ^"ÌE 5 #Òê‡òF2oÞ<ÑÙ ˆµk×–,þt¶(ï@§qTT”ëVaBB·»k×.ò¼›7oB­RªkóçÏ>|¸ž„±®1ÖuÍL°®©1B×V­ZuåÊ• ,\¸ðÉ“'ðæÍ›ËŒ1ÂÓÓ³Y³fÈ1Äß¡CyÓ¹Õ«W‡‡‡çÉ“§wïÞ...´ðÀû÷ïÇ ¿råJÔ@¾¾¾ò8QX»¹¹ 8pñâÅØ¨··7ÕX’æcämÛ¶ÅÆÆ¢ÂS¬E¬]»öÎ;… :th²E9 áñãÇüøqÖ¬Y˜}ýúõôéÓóåËI©â ¨¨=z´eËœK3f|ÿþ=òþEæwïÞ½R¥J˜F 7qâDì)µÅó÷÷Gà€€Ôå¨ÛZ¶lY£F yÌHêš5k¨{ñ1ø?þˆœ¿pá±cÇ`4ö<2­Aƒ˜7niŠ:«¬[·9–?þnݺÑéAÐQ€š,]ºQ¿~ý .œ¢|H7¨›lܸq̘1OŸ>a{3gÎ,X° þ¨è\jذá‘#G Üz*›ÈO`ÃumÙ²eƒ R,œ2eÊwß}§N›ˆ°B… 8ó3eÊ„ÛÁÁ!>>z§_×:uê„k922¥x6¼oß¾¦M›êJëcX×Ìëš#t …&ì ?ŠWdæÑ£GQ‹ãöWÀrd׆ àpøða*Ù`eË–Eïîî~ëÖ-Ô4¢§o¿ýöûï¿oÑ¢Òƒ2÷àÁƒp‹={öÈã„ÂÌJ”(‘-[6¸ʼn^^^¨/qPBBBð/¢å>¶R¾|y¨aõêÕQîŸ?^×›A90³9sæ`p"s$‘Nyzzôè­1$ÇÀŒ3°õ—/_žÕaØ/ìÈ7hâŸ:uªnݺÈRToVÔXôl/** YqýúuØð={ölÓ¦ &°ºvñâE裢ÒAª ߨQ#l.降Á¦±!‹ãÇ#ð»wï²dÉb`&¤'tuä±|ùrø½|D8ô„ ôi¢e(4Ttx‹S1gΜgΜ¡Yœcp2]îÝ»'<-Á½ÖàÁƒåa’mjàää„ÓLÿÎJ¬kŒµ`]3¬kjŒÖ57ܸíFÆŠvp¤q¨é)£ VXyGÉš5«&H”g;“®µnÝ:((HÒ ¿6bĈsçÎU«VMç•+W¨j¼pႇ‡Çû÷ï¡««kxx8" 8Äâ{šÄÄDÜÓÃí¨¸—4ÊräÈaÈ>"rØL—.]6oÞ¬õE Ò%Zºt©!± Nœ8ñÕW_½yóF1¨N²ºíââË—/T)44 ããã¡JE‹r€|·FÔÿ2º ¥Vü;fÌ; ´]lÙ.šþa!’Ä`ùƒ\Rtâ`?èï&Gsذa”QD×®]qVãÎAUšÐ5zHFÓ—/_F¡!ŠALˆØ´FσíÉ;N3\×è~Œ¦å…ŒÖ5Æ:°®™ Ö55FëZll,î°%I ’WÛ(vÕ]“©`¡xsºuëÖÎ;#ˆ‡tíñãÇôríÝ»wÐ,__ß_~ùEĉYTÄ£Gp,Dœ4ØnñâÅaŠÒ§Å:äéé™ÒlCå”(Q"""Býê§dŠ¢5Z×DÕ€Y! k×®íÝ»7´IWÏ^FèêNX5¼¶‹ƒ2yòd1Â#Ö¬YS-”?_Wë®&š›0\×nHP2Hš‹QþåœÖ5Æ:°®™ Ö55Fëšâkå^½z­Y³FÌúûûÏ›7O¾ÖŽ;Úµk'¾2Æê¸Õ~ðàÌ€lf ˆ³R¥JW®\³â©’€ºP<Ø“>Õ»wïnÕª•Ñoë¦M›öÝwß +5 &ѵL™2‘LŸ>}âĉzj #t ñË{Ò"#”o½~ýúØ 1+Wj»ÂÀ!Þ¡;È.ÜNˆþ/`à 4À¤p«ëšV°ndd¤øBQþt\úôØŒ¦ÅehÈãºéNrœê’¦ Î|EË‚u±¬kf‚uMѺ†<¤b‘Þ9Ž1r@PìŽ=zæÌ™òµ° VTûAûöíq¶“®É=‘4nÜøàÁƒb¢sìØ1ùº8Å‹?yòd:uD0gggì¦Qù5jÔHña„……Õ¬YâÝŸIУkô¬"** Âdˆ®-^¼xèСêQ´FèêBjîGÛE„DsEÖ5º&ؾ}û€èÌ$p÷²|ùrœ`æÐµsçΉ&)©Ô5¸šT<ÿ&èÖ‚¦Åi™"]“·’ W®\Ù·o_­áQ†ìÛ·O!ëcX×Ì뚣u 5Mÿþý%M{ù;G”­êØÈBV¯^Ý»wo±ðøñãÐéSSÜFO›6MÒ4ûoÓ¦ü&[«®Qçm۶ݱc‡ôékè1cÆ`–¾ô¢~ž(¼ºEçÂ… q> 2D¾š7oÞøøø©S§Ž;Vnò+UªT¯^=CrL K×&OžŒLÀ46:iÒ$CtÆç‘·É@õ™-[6í¢E‹† öþý{­=JhÕ5Ø3bßÁYïܹ#¶ÎºF¤T׈ÎpyC„AƒÑ¦Õ59z2'ÛÆ9¥ åáu­«uyj)±®1Ö‚uÍL°®©1Z× (€j&)) våÊ•/^¼(hÕ5IÓH¿~ý†Þ¥K¨¼âBõ é|nÖ¬YP%â—§¢U×$M³»+VÀo UPׯ_c 4ô/½ËÃ_ˆÎ-£`äÑJªÇ8¾=‚"°F¸£ºµVyò¹!ÔJÒ|îîî^¦LäŠlþšpãÆ Êó½{÷"ÿ‘i8Žôî ³¥K—._¾<Œ|Íš5ò1Yׯé㦻É|ºf°®1ÖuÍL°®©IÍ·k5jÔ@~‡ÄkPu j£3f¨×}ñâEÏž=ýõWÔ43gÎlÒ¤ -ß®yzzž?¾]»v«W¯VĉÀØ®:Îàà`øD§eË–6lPü‹â»W¯^ðc]K ¬kŒu`]3¬kjLÒÔ õ¨›0Œ-“z]“4 BCCqA‘¯³® ëcX×ÌëšÖ5†1“èÓ~ïÞ½«W¯¦O0 uMtmäÈ‘Ø cn`4Vš°®™ Ö55FèZ½zõNž>>òÔN:uþüùÑÑјž>}:U±zhÖ¬ÙþýûoܸQ¾|yÌ&$$8::–-[öÖ­[ò`ˆªzõê76$Ç@­ZµÎž=«X(D {·~ýzì¦ß¾}‹¢ãàÁƒ9ÒŒ3(åXˆ¿D ÈŠñãÇ/^¼k¡â˜7oÞàÁƒ%M%’˜˜H«‹­L˜0áûï¿—þþ¡ºZæ yß}÷]TT®”aÆÑi/ÒY±bÅJ•*áè`¶ÿþË—/70Òv®k8ñ2f̈_Ió?&p½ß¿ßÚ鲬kiÖ5[†uMÑC¼)RäÎ;(¦Ë•+z÷î€JÝÃÃ#[¶lðLW©RåæÍ›XÞ»wïµk×®^½{öìiÙ²e—.]È´hˆ÷¬Y³¢ÄGz°úõë×I8DœØ"þÂM|®\¹¾úê+L`9¤ "‚ÌuîÜyË–-?.\¸0­…ø!ˆÐǸ¸¸^½zýòË/ú÷;‚”Ã`H˜Ý½{w«V­Nž<)wGJÄždHŽ ._¾ìîî®~º†_¤¼cÇŽ˜F:³gÏÁ%¦”zZ—š2e Å{;räȪU«¥8”mÚ´ ›Óÿt ²µråJÅ¿‡jÒ¤ ² ™¶sçN¸oëÖ­ƒ‚‚Ä^ã—Ì’>}Û±cGÛ¶mS” é;×µíÛ·ûúúbÂßß?<<œ:‡Ç•ˆk'ÍB¤g]‹‰‰¡KE­µÓ¢÷úTfáv%uj¢b]³eX×Ô­kâ,hÄ‚$ŒàJWßjß¾}b×·o_(-Am‡-ÆÆÆâ¢#] †ÃIš*°R¥Jr/Aœ...ôL•?~ñÌ Q9;;îvíÚ…YØ$*áÇϟ??EyR°`ÁgÏžQ™€´Aà PPIy”Ò~~~³fÍJQÌ'Nœ€h¾yóB&_ž¬®ÅÇÇ;::Ò,’‰i˜Yƒ EÈ(­››3gΨQ£téšÖ¦ÕªUÃ5‚âš:¯êÙ³'„ 鄚àèHšç+}úô,¦(Òv®k¸Þ¿/i®Aœ•+W–4Ýö¢2µ“7æ¦Ñ5K~a€òQOÕUµjÕcÇŽÑ´HRBB¬È‰3˜9rPyÔ®]»mÛ¶¥&*Ö5[†uMѺ&¯æš0+×2‚S¡ÄÈ™3'-9wî\5``ð0Ò5EœuêÔ9yò¤˜E`ÅûDaÉ’%¯^½Z±bE Å à·ÿþfÍš=}ú´@)É ù0f̘ÀÀ@¸‹««ë€–-[–¢ta´®É_›Š ‚Z“Sc„®!þ|ùò=þœfÏçðo½zõÄ<Ìúúú&ûÀ2]bϺFg&:uê„˸´qc"<<—Œ•ÓgLötM¸‘¹iã& E­®«W¯¦H’-ëZê‹Ö5[†uM©t­_¿~+V¬³êØ‚‚‚´¾5ƒÀ´êš‡‡ÇùóçÅ,DGÜþ÷îÝ+UªÔÍ›7Ë•+'‚IŸÊ=\‰¸å‚h8(„===k×®=iÒ¤/^è)åR„!ºFß®¢k?þøãøñãõò©×µU«VáÈÊ·^¿~}ì…˜e]³C]ËŸ?TTêqœÉt«€[&zÀ–'O\/ÖN %Hú&ÿ U ÿ>×–uM¤ õо}ûRëš-ú¦Æh]»sçN™2e0{ãÆ 77·Y³fÁ (€V]»páBµjÕt•H¤kô f=z„LÆü’%KDœj]{úôi¡B… ‰P éÓs)ñŽ>ÃQ£jô¼ˆ€«…††¢*BŸbX£ß`èѵ5kÖôêÕKúôÐÑ] nݺõÆ»víª5m(@FŽ™"]ƒþâ”ˆŽŽ¦]ö÷÷_¾|9½ö’X×dØ­®-X°€.(ooï={öˆå8±éÄ8uêîs¬–>Ka+ºfxa$º¦[¶¬k&„uÍ–a]Sc´®¡Ú†<᪯W¯Þ™3gÞ¾}+¾ûÔªk xñâ¸ö±V‘"E’’’ Y°=é“®•(Q7x$j"€Ûtܬ‹8Õº†ãˆò'>>Þ† zG#œ”,YòþýûçÏŸ÷ðð@R!:tÇ !v»‰‰ùóç>\ Xº³nÝ:CrL KײdÉR½zuü‹â±|ùò†èÈ—/lòêÕ«åÊ•ûøñãÂ… å%Æ–-[:wî 'ëÖ­›:1}úô#*ö}åÊ•ýû÷?|øp£F?~\´hÑÊ•+C|ÅÖY×»Õ5ú²S]Š>|øúðà LuúÆ ]C6yòä­[·¾|ù³(øêÖ­Û®];j¦!Ú).Â!C†,Z´ˆ¦q„»á+W®Hšg`“&MòññÁvÿ—”Ov…” ~Â글Q0[Ï‘#ý›z]Cm7p(¡0[³fÍ€€T7y¢¼–47Óô ß‹/Ò¡C‡¢jß¾½øììÝ»w¸§\ºt)Ö={6JUIóA4êl¥V­Z›7oFÍ¡N[óæÍI¶D3xIóð€šôSéF 孥䰮Ù2¬kjŒÖµï¾ûnêÔ©˜Å¹Š Jþ¢Ô¨Q£ÔŸá#Ûaááá4‹Ë388˜Ú+®ÁèÆEÄôéÓG-S. ‚÷ïß—*U %­Õ±cGêc‚@¡T¡B…Û·oSéäîî.Ê ­¤£ì‚Q¡ÐUcŢїS*¨Y†ºp@Ñ„²…ºE ÅžêÑ5˜Í¢¬C–¢¦¤Y//¯ýû÷ËcF *þÅñB>?xðZ¬N›ØDÆ ‘Ï”äíï¿ÿ.ßkÅ…¨8¶oßž¢LHا®mܸ‘º›Ñú7NE\k’æ;T¦VHŸI^×èÓZ­ÑÅ&Úk(¥'.]\âj`E¸—06a0(C'Nœ¨üèÑ#h¢”j]Û·oŸÖƒ*š>yáBïÅGŒ1oÞâNò‘?~ À´ÜÙÙ®ƒ›Eˆ 5›*P Lñôïߟî8a`ð°ÿ&EvvëÖm̘1Ø„x Bêoj .&$™®………Õ¬Y“Ö«W6‰½†ZÑW®âA—ôé‹ZšFM)n *ôäÉšº¦=gUqŽ ¤i­M »téBÓ… ŽŠŠ¢çmH<ìMѪ_Àºf˰®©ºV§N///CVY¿~=ʸÖ;Cã8räȉ'l¤£ýû÷Ÿ>}E}ŸÇ0rP¡àt=|ø°dOºF}%bBþQ©‚²eËÞ¹sGÒ\A–'i”äu ê ÀDÛ¶mwìØ!–_¿~ÝÍÍíéøvna%<~üX,ÇÙ†$Ù@­"OOÏC‡ÑtDD„h£KýF¦F×°!ê[OÞ€4)) ¦Bb$Ö¤I‘ xÈ'ý]×¶nÝ íC a´54ö·`Á‚C† ¡ó,K–,¢cO]-CE -çÊ•+•*Uҵˬk¶ ëš¡k Ãè*eTšTrRkn­Á¨+DIób”:ÇN¯$¯kÔÑM#ï h•+Wnذ!<ÃÁÁA(‘.]ƒZA°$Ͱ0X…¾N@àµk×Ò£ËZµjážRƒÜ™¨_DšÆ}ð˜1cäº6jÔ(y²ãããEÃ~I›®‰%ÇG=*ž«9’žÿA.éã¢xñâ¨)ÅìÎ;Û´i#f…®!÷âââ2dÈ ß„ðìN… ™£§#wwwñ™­¤y>nÜ8I7¬k¶ ëšÅç¡ Ã¤ý7ðéê7~~~ê§rêÕ«G_7^ºt©J•*JŸÅ1¨©AŸ>}‚ƒƒ½{C8bcc“Õµd?2€aÐÓ;RÞàK¾ü›o¾™>}zj¾]K61ݺu“¿ßœ1cÆ„ h6î(Ét ¢#ovNaaaÕ«W—>u­®H°] uçæ X×lÖ55È Êk'„aÒIII%K–ÔÕMqzB4¿‹ŒŒ,X° ž·nÝ¢Æy¹sç~ñâEzý´1ey\¸pa×®]zH6bĈ9sæü7"ºVºtijæ“QƒróŸ}K §"jÑMÓ4Ú1MSûöÔèš$lT][ |ûöíEû€Ó§O+ÆìSô‘¦_ׄæ‹3I2@×WÈ1ùF1‹ºMÏn²®Ù2¬k Ã0)eݺuÔ)`Ïž=×®]›lxQÞ»wOk3ät@òºÖ©S'ñÐgveÊ”K©w“uÍ–a]c†1XÄ€$M{ÏÀÀ@ÃWDEŒªùþýû˜>xð`ãÆÍ•D+a²1CM”¿ëšu“†`]³eX׆a 'wîÜüñG*#Qþ–`]Kó°®Ù2¬k Ã0²sçN“D%zàO7°®¥yX×lÖ5†aCøë¯¿ÆŽK}­éz i:©èÔ©“®`%K–>|¸‰“hUlH×–,Y‚dà€õîÝ[Þ-£Ö5[†uaÆ´ÐÚ5kž9sÆÚi±6¤kŒq°®Ù2¬k Ã0¦…t­FgÏžµvZ,ëZš‡uÍ–a]c†1-¬k¬kiÖ5[†uaÆ´°®±®¥IX×lÖ5†aÓb׺æèè(YbÒ>>>ÔI/ëš ÂºÆ0 cZìZטtëš ÂºÆ0 cZXט´ ëš ÂºÆ0 cZìT×ÒåÀõroܸQ®\¹ôÝõîÇòfÍjÜê¬kf‚uaÆ´Ø©®Ñpìé˜èèè>ÞÚɱ_Ž=º~ýz‰uMR52“^w95°®™ Ö5†al*‘räÈñêÕ+k§Å®Y¼xñСC%Ö5I¯®åÏŸÿÙ³g–LŒíúf&X׆±X×lÖµÿâääôöí[]ÿfÊ” vbÉôØ>¬kf‚uaÛuÍF`]û 9sæÔÿVþñãÇ… ¶X’lÖ53ÁºÆ0ŒíÀºf#°®ýìjGýaV¬Xѯ_?ˤ'MÀºf&X׆±X×lÖµÿ``/¸éoÇSëš™`]cÆv`]³X×þƒºÒ¬Y3s'&­Àºf&X׆±X×lÖµÿ`ˆ®!ÌŽ;Ú´icô¤ X×ÌëÃ0¶ëšÀºöøe¨°®™ Ö5†alÖ5uí¿$kl¾¾¾¿üò‹e“&`]3¬k îÝ»÷ôéSk§‚I|ùå—° k§"½Áºf P#ýëÚxÿþ}öìÙ“’’ô„yòäI¡B…,–$Û‡uÍL°®)2dÈ’%K¬ & Z¯^=k§"½aˆ®]ºtIWï ¨XÝÝÝÁŠ+æìì¬'iÔ§D‰¹råÒõ/|Iºxñ¢ƒƒƒ|‰"B”„1114;wî¢E‹Ò4n_¿~-‰„9::b‹ùóç×µ›É"tíüùóFÇ£Öµÿ’'Ožèèh]ÿr7¹jX×Ìëš‚áÇ/\¸ÐÚ©`Ò'Ož¬S§ŽµS‘Þ0D×ô?UÎ ‚mܸ±k×®ê÷ï߇ˆPøjÕª;wNüW+S¦ MgË–ºµ—o÷u‹-Ò“°~ýú­X±‚¦½½½CBB´¦ºFû«gt!6wùòåÊ•+ƒX×þ ޽“““žér¯Sëš™`]S tmÙ²e¸­²vr›ãÚµk“'O–X×̃%u ôïßåÊ•4=f̘€€š.Q¢Äƒhé)]º´b»Y²dy÷î<ªS§NÕ­[W̨kDbbb† ôÐ ëš…à!ÞSëš™`]S tGa´rìØ±† J¬kæ!EºvàÀEu‰r¬yóæŠ`ztMÒ<ßŠŠŠ¢ðÑÑÑÎÎÎ\¾|9ý‹Ò€¬ERÕÚò¿@ýúõýõW1«U×ræÌùóÏ?#ÍÏŸ?Ÿ3gÎo¿ýFpR}õÕWºR¨ Ö5K€jÚ´éÁƒÕá†þÅ‹–O’ƒ‹n>à~~~ÖNNúuMе‡ŠOOFpôèÑFI¬kæ!Eº¦_ Ô5_Å‹§é\¹r¡^®^½:Í–(QâÞ½{êa]HÖºÿ>-A08 RñUºV]“×ïòñvìØÑ¶m[yª`AAA/_¾D)Ô±cÇL™2?>K–,økÔ¨QÙ²e“´éÂC³0ñîÝ»üñÑ£G;wîD’7n\¡B±­Û·o,X°S§N¡VX×þÇ7ß|óÃ?¨—#ƒ¬þéLBBÒfS9ÿñãÇéÓ§cÂËË«fÍšÖNÎÿ@Â:tèP±bEk'ÄHX×°®1úa]3+–×50mÚ´ï¾ûN±0C† >|·iΞ=΄‰çÏŸçË—OÒTÜð8l •‚d€®É#Í›7¯ÖY›Ö5c°¼®µiÓ‡ÆF/Caô±±±«W¯X× uuM׿çΣ.È Ñ5éï_:ß¹s§L™2˜À…9wî\Z8oÞ<ÿï cAAAbüÖ5«“¬®òÔ`¨àä#$»ºÛM b Cš9sæØ±c%f½~ý:{öìɦGOù_º‚±®iÁ*O×Ò5õrÖ5C`]c]ÃDæÌ™ß¿¯+˜ºŸ5kV €cñáÇß~û­jÕªô/’¡è‹œ1 TQíÝ»·E‹bI²k±®™Cž®¥{pZâÔêÓ§OÉ’%_¾|¹nݺ={öÐ_òÎ8Ì ëš¬¥kTTi5kÖ5C`]c]Ä££#dKW0uMÒ¯?8swggg†ÝÙ×–™‘/iÚ´©˜šC$â,³´´\¶l™»»{``àäÉ“,X@Ïç [8&)>Ná–-[ž*pppˆŽŽFúèíí}ÿþ}ê°&©ú¬Õ¨QƒÂ!š'2GOåºF†ÇÇÇ#>½®[·î½{÷$ÙI'©&íÏŒ3 n’j¤ø#GŽHÿÔµ;v °ïÞ½CcKÙkРÁíÛ·i-w5ÈW°®åX×4Ppu ÷»¿ÿþ»¤jjÑŠ)2ò³®õêÕKÜÙg³ÂY×X×$Õ;8\ìåÇRFF t.é©kr!`?ä+a!Ë—/WŒ¿CàáèèHaù8R‰‰‰ô­NÏ´´4sss¡k¸u¡ÑÝÄÉ»téÒ)S¦H™Œä$t 6 §{G‹±`ÁJ–, ›\µj•J ˜øbìéÓ§¤nØÍf-Jf&ÉféÁîÐÂP©!‚´öàÁƒ¢ÍùðáÍ2$ôQþ2T˜™xÝ\ªT)ñ~–u-_Áº–O`]Ó@ÁÕ5´Ô—/_–Ô>.ÉϺfDX×X×$ãu5T=@#""$•ÒìpÒ?_AJ²“ëèѣݺuSO$$$¤J•*÷(¶MOO—ëš¼"ÎÍ›75j„$Š…®!ȉتN<ÏCä¹sçŠDžìN<¤åB×zôè!n¨~øá‡H²{¿üòËÂ… µÖâºÖ§OŸýû÷Ó*$NÏÞä6Àº–¯`]Ë'°®i  êÚAy',Iu‘Tsœ }ú¢E‹ÔÓyõê•ø_àlSP‹yuð¿NKKÃù¢]×îÝ»G#‰‰áyÕuMñuÝĉW®\)}|D´øÒ?õùúúÒÇab×B×ä!Ôumݺu¢;Ef(t­oß¾hhUÏž=i. Öµ| ëZ>uMB×ÐÀå“Â"ô´Lw]#裺F?&&F± Ž –_ip÷çççGóæ)))ÖÖÖе6mÚÐ’N:É;ÓuèÐaêÔ©P.yMŠÉw%Mß®É猓T'l+>þÝ»w¸8©–uuM2’®É•ˆDF…”i ¸!×Y±b…{=88X$3]ËÈÈ@3ºxñâ'Ožà't‡®pø¿ªëâ£A§™ÔŠ/>gΜ¢E‹.Z´(<<\RÕ¥#©zlá*BawwwDÀ¥…~ 'ëš:¸ªÑˆPWW×ÇSXcW4»Mš4¡p¥J•ÐF`_ôóÆ™õc]c]“tÖ5CRãì › 1 Ž^ðô¢póæÍbœ‹qãÆQGܱ7Bš-[¶Ä¶¸±¹|ùò²eËð‘T¹‡Nq<==ñ¯ ¹{÷.-9vìX×®]%YÏÐlêî¬Pü½té’XˆI7½¼¼¶lÙB [·n]²dIzž-ÉfÖÒQר*h„dµE‹(>]Ü>}ú466væÌ™¿þú«¤®Ýºu‹^ûJª>+VD{%ÞŸªÃº–£P‹”×¹`>úö™V­Z‰ÇKùø K¤ýÛ5´Yh¹$¾]+q£e¤áÚŤ=’jÚD13®`÷îÝô)±ôO]£Æ=88XŒ%}º ‰®p¸¦¦¦ÒªÌz†.Y²Dtæ—/¤y?4ºƺ&éótMúNkÔ¨Q7n¤%+W®ƒåJ²©ÁíÛ·4h ©ž*ASº†B9::*æàq×ò Ûêƒê,p𤥥á¼À΀v5_[___ùEZ0}]ˇ Î-,,pœK×ÄÃýë_¸Æ‹GÜS¦L¡pÛ}á‘®ýbKлwojµ ¹®EGGÓg=b§íÛ·?sæŒ"'â(Ò2.NòK ¼‡šFX× ³®Í;—ŒÁNnÞ¼™Y´ððp1}¸:piÓ¦áà§µJ•*Ї:8¼Û¶mK}6Ë–-+ïÿìÙ³M›6á Ä­l'¼Dô”$V¬Xqé񴯯Xœ#8/ÄgdNmz=*/BÕªUiÕ$šøçB+€ßྈ¢ ]£3à@×wïÞ¡D¨õ’B¶–.]Šv*¨¨˜×ÙÉUX×-¸¹¹=zôHR}6× 4°òwLÖµ<ÀèºFiÖ¯__ãx³Ò'òóó#!“T_³EGG¿{÷ÎÞÞ^D£a?«kòQà-,,° 333ìEt&ÅÉVµjUõ<³®±®1…Ö5&3¢¢¢Ê—/Ë ÚÕ“'OÒxò±EMÖµ<@G]Ã]5ÍÓüÕW_É__„h\YÅÕOž<ñöö¾uëTÉÎÎÇ1ü¬S§N5kÖÄZÜÔ]öFß>}zðàÁÔåÍÕÕõÂ… ׯ_§neñññbç²eËÒÌ#GŽœ5k-„óYYY!޼môèÑôäÌÚÚ:44H$¹&Ž´7nºaïÈÙdTcQÖ5Ö5¦ºÆdFóæÍ©ÇñÅ‹qíöˆQxLÖµ<@G]+ä°®±®1…Ö5F#OŸ>­Q£†¤z~-©˜¥73ð6šSδa]ËX×tuu)„°®1ó¼=xð@ ]¥JˇM‹-ò29ëZÀº¦ ¬k&©k¢PqqqŠeF’MZ%:?1L`` }Òóå—_¢éŸM‹¯Ù6lèçç—§yÌqX×òÖ5]`]+¸ºÖ¬Y³ºuë®\¹Rã*k¦L™2ô5$ÃÈIKK£©¯Åì« CÓ P«V-±œºи¡Ð5H[že1ça]ËX×tu­Àéîtg̘±aÆ޽{»»»O:uذa¿ÿþ»¥¥¥ˆƒ¦V>&ÃdÆp åu.˜¼'22²B… ’ª_ZDD„b”µðððJ•*!P£FÀÀÀ¼Éb®Àº–°®éëZÓ5Iõ«qãÆUªT±··¿}û6þ‰®®®òbØ$Åü› Cà’DS°0áYDY×òÖ5]`]+ˆº&ɆY¾wï^ݺukG½aƒƒéŽ™aä\½zµ}ûöRæ×f¦P!¦1¨V­ÚÓ§O5Æ pww—2yüf2°®å¬kºÀºVàtíÝ»w-Z´hݺµ¿¿õêÕqœ?þüúõëæææ"÷ e´Ã=C95kÖ¤Wœòaáգʇ……9::æZörÖµ<@èÚŒ3`lbòuF€Û#Ü0Ñw£¬kE×p`?zôˆFfvss[¸páµk×7n,ïɺÆh‡uÄÅÅ•)S—HWWWíß¼Þ¾}›®hUжäVsÖµ<@èZ^g¤`ÀºVPtM0}úôÊ•+:T}ë£Ö5Fƒ‡IY=Z#p£Hcy(zš ¬kyëš^øøøˆâ …S×´ÀºÆh‡u!Ĥ:>0c³eù(®€Âº–¤¤¤tïÞ=¿ÕüÕ«WÓÒÒ:tè‘‘‘×yùLzzúÌ™3©ù.ˆ°®)`]c´ÃºÆíÛ·?wπjÕªU{þü¹¤ºœ5kÖ,gó—ë°®1ÿKKKˆZ~”•a]SÀºÆh‡u=rssC D‰qqq:n¦ÒT'9`]cþ u~^²dÉ”)SLµ#tîú¦€uÑëºuëvüøqIåmôŠSG*W®ŠÀµkך6mšSùË XטÿCîg|HÖ5¬kŒvX×ñÕš¥¥åü¡×lu¯^½1buëÖ½wï^Ne1/`]c¤?ʇZHNN¶±±ÉÃü˜¬k ÆŽ»nÝ:‰uÉÖ5¦eË–W®\É~:5jÔÈ~:ùÖ5æó€´DëÖ­/\¸‡ù1%Œ¢kgΜqppÈÜå6¨‡E‹8p@b]c2u­óîÝ»’%K%©Š+†‡‡%©üë#ÙÙÙ%&&ŠŸ†¹££èšIb˜®A[-,,½ú·mÛ¶qãÆøøøúõëïß¿ßxyÌp2º»»‡uÉÖ5¦0’ÙÛ””kkë\ÎŒéÁº¦@_]{þü9T‡"ê°lÙ²4¢µZ8t‹-úþý{„ÀßF‰^2ãÆ[»v­îíÛ°aömÛ¶xñâ ÀÉ>~ü(¶…6a/ô3$$¤J•*/^¼¨\¹22`aaannŽ“¥cÇŽ§OŸ†S"2¤ «Ž;Ö½{w„“““·lÙ‚]ˆÝaáÌ™3.\WËÈÈ@1‘a,ÄŽà‹7nÜhܸ±<‡¿ÿþûO?ý„ªÀ^Á”ú Ëa]c²„u)tŒ3fýúõWá*BB&;°®)ÐK×RSSa'°´¨¨(I5ÍsçÎ¥¾ U=JLœ8qåÊ•úêÚܹsçÌ™ñêÔ©Óo¿ý=’´êZ5é…,¤mÓ¦M“&MB¶a{–––{{{c« &¬Y³&88Ûʳ­8Ž ü³íì캆 fÏž=8€˜0<äÁÅÅEÇÒ Xט,a]c ø×17ýZ"äf~LÖ5zéüÆÙÙùäÉ“'IõÄ—Þê¢kðª—/_êÞó€t ¦Í¢dgÍš5þ|I«®5hÐàöíÛ%K–|÷î"=zôûï¿OOO;vìÆù\½zõ¸qãÄÏQ£Fi¼Yúûï¿¿üòK…® :tûöí¦úDMë“%¬kL¡CûgìwîÜ¡ƒƒa]S —®ÑWbâÃ/u9ËL×ô…tM¤ƒd§OŸ¾hÑ"I«®Ñ£T©Rqqqˆ³ìÒ¥ t ÷ðáCE®&OžLÛ(ýµk×¢6Ôs¢Q×jÕªõøñãÂÐ\³®1YºÆ.Ê•+­åÀ¥˜ÜÌ’éÁº¦@/]{ò䉫«ëÓ§O«U«FKò\×è[º,u­M›6W®\Ñ’+l»~ýúQ£F©¯Ò¨kÍš5»~ýzah®Yט,a]c ÎÎÎÁÁÁÚãðá‘MX×è¥kQQQ¸©ÚA›è¢kž„„ÝÇ Ñ¢k4i4­ÚºuëðáóԵ]»v <øþýûuêÔ¡å==õÕµíÛ·:ôÈ‘#Ý»w§%8¢äsÇ™ ¬kL–°®1… ]ôÚ¼y3.N¹S…uM¾=C===ýüü A-Z´@ëüäÉ)'»hԵŋϘ1cÍš5P¥ªU«¦¦¦f©kæææˆ}îÜ9ÄÙ¸qãØ±cÙÎL×’’’Š-zóæÍFÉ—c¯^½:{ölÓ¦M‘·V­Zµk×NÇÒ Xט,a]c nnn=Ê2šOÿþýMi¤Ö\†uM¾º†zƒxA•P‡Ç×½g(LHñ±¿vúõë·wï^¹®?~ÕªUôóûï¿?zô( iõêÕ®®®øŸÂƧzõê0H333øãÈ‘#=zô ]ûðá<² ëšÖ5F;¬kL–°®1…‹Úµk?|øP{><² ëšÖ5F;¬kúòÿTäu.r4§¤k 4ðóóËëìä¬k…íïCýýýÝÝÝs-3& ëšÖ5F;¬kú²|ùòyóæ™d¿à²õ÷ß#`ffV¤H‘¼ÎNîÁºV¨iÑ¢ÅÕ«W3[›””Dsþ0ú¦€uÑëš¾,\¸ð—_~Éë\09ëZ¡¦W¯^üñGfÇÙ‡uMë£Ö5}ºæéé™×yar„‡&''³®v2{ʆQ`]SÀºÆh‡uM_„®q£mª¸»»°®v2Óµ””kkë\ÎŒéÁº¦€uÑëš¾]ûøñc¡¢¯ð@Ã8°®v¶lÙâåå¥Xhoo“'ù11X×dG×rÀØÂ6Dm~€uM_X×LÖ5æÿHNNVï\CsìäI~L Ö5zéÚ½{÷êׯ/~±±ªY³æøñ㇠’Y„öíÛŸ;wN±Ç}ûöõíÛ7‡ÚÌ_U¤¦¦"üîÝ»/¿ü2'ö¢ ‘‘‘*Txôèj)÷÷κ¦/¬k&ëó_ÔÏp>*ŒëšÞ®uêÔéôéÓF<,qÌ/X°@Kº3gÎ॓/ñý÷ßÿñÇò8t*pš`ÃI“&-[¶L±°bÅŠáááâg:uîß¿p›6m.^¼èâârãÆ 8_óæÍ‘çcÇŽQLœ¶(;tçðáꢰ5’§Œâ ŽzN¢¢¢Ê•+‡Šuuu•/·³³+^¼8nÛ ‹Õ«Wô葼ÔmÛ¶…oݺÿe]^j„uM_X×LÖ5æ3AAAh÷˜;wîœ9s4Æqtt|ùò%0zÁº¦ÀˆºöðáÃÚµk{{{Ož}ÔsB] º†êš={öÝ»wQvü\½zõ„ n߾ݠAQê7oÞ”.]{733“?Ô Ö5}a]3yXטϤ¦¦Ò(kK—.2eŠÆ8¬kÀº¦ÀˆºöÇôìÙSÞ‰×*gggÜ{P¸S§NP.Z…˜ò <²£kp__ßAƒíرCÇüë‚aº&rHiÒO˜™……ÅÖ­[‡ªewzéš››Û£GÄîÞ¿_¬X±ß~ûí§Ÿ~’Ô$Õ`g•X×ô‡uÍäa]c>#tmÕªUŠÞpÖ5`]S`D]Û»wo¿~ýä—(Ê”)ó×_Q¸{÷îGŽ¡U?þø#ìÄ(ºFaøLш3L]×>Ü£G-»ÓK×ÔO¤°qãÆ#FH¬kyŠ.º–e?•ÂÓ‡ í —S¥J•®]»+Ù+W®<}úôÇ’êz:cÆ ü_èª:jÔ(›ì$κÆ|FèÚêիǧ1ëš°®)0L×:vìxæÌűƒÆÝ¹s‡ÆfKHH(^¼x«V­ 1ø‰°››Ûõë×)rëÖ­/]ºd]ƒzyya×° ¤©kɳ"3]ûæ›opÞIŸ$L]ÃÁ†ËªzÅŠZv§—®uîÜùÔ©SâY¦¿¿?rFmH¬kyŠŽOײÔ5 T­ZõÅ‹|õÕWÑÑÑê1å#«ãÔà (_Û¡C‡³gÏR˜N±J•*‰û ¬UªfòäÉòÃ^ääæÍ›êNOOϱcÇâ€DXK‰2£D‰ñññtëÖíèÑ£¤ NåÊ•CCCåKäú[ªT©ì¤ÏºÆ|Fè®çÎÓxb³®ëšÃt >®I>„‰…ÿý7ÔÁÙÙùùóç8b‘,¿wï^ݺu%Õã4Ønv---qТeÇ¿Àˆ] :TmÜoלœœpiù믿ÌÍͽ½½§Nª‹®—   °°°Š+"åôôôˆˆ\8EÊfffÓ¦MC­ªçäõë×åË— ¤.«})¸dÉäZ€Ýáú!ïÁ*±®åÆÕ5ú.SRµ]8³4FîׯßÞ½{)Œû¨öíÛS˜ ã‡ÿ ôéȤ…h qÚÊû/ƒråÊEEE©ç"èáá‘Y†Ei}1º®Í;wÞ¼y’ª†‹+†¿(#ë“S]“T·æ{೮뚃'¡‚$%%)R$---99™^¿~½Y³f¶¶¶ÄÄD¸Ô¾}ûhWT;6IHH Îä>ÞÊÊJ¤O-ì³gϨû$v°S,ü BRÈÃÞÞ>..ù‘Ÿ/ÆÕµcÇŽuïÞ^¹Ž3féÒ¥:êšô©ó®£r ab-õ½ÀZdéGFFJª«©AѪ÷ïßgddÐ&ôññÁrhþ.\hݺµÆR³®å&¹¯k§½R”>Ëp 1Œî.púPX®k’jjñEðóóóôôÔ˜¡kò1nähùtG F×5Ü¡$Õ)yþüy±\T8uÁÉÎ.Xטϰ®å¬k Ö5Ó“'ORRRp6iÒD,ÇkXX”¢lÙ²ò§G"õèÑ#\6 .#FŒ€l‰£ž§8È‘BóæÍai4¾ \›6mJû‚ω½GGG—)S†zU×®]CÊò겄 Q¼xñ"&&טêÕ«_½zI]Aß¾} [¥hþþþ0NA¾mHH.¸vºººŠ\€‹1t“:r’¤"óËã฻áááQQQH­ .–C­ä1é4U𾰮鋾º¶{÷î~ýú)ÖŠ§A:êš|‚8™87i @ZròäÉN:QX¡kŠÆPñhMÊäéšXˆ“wÔéX<ÀÓ u]Ëæw{Å‹Çí¤z舺¥…ø_6î FXטÏà0ÀòåË—%5]‡²““®‹tÀžïR³ ëš‚ìLñ®ŠCM<–<}ú4çöÈÖ5}ÑW×D^è¨k`öìÙ , 0ÔíîÝ»6lØ–-[D4¡k¥K—Æ ¤ú(nqûlCËnß¾Mñµëš¤£v(©Æ>§RœqãÆíÙ³2giiÙªU+d£B… ´vÔ¨Qëׯ—2yº&*‰ >|ïÞ½IIIUªTYµj¼WÆñãÇã -yûöí?Nw5W®\iÙ²¥zÍ ÞP{Чk¸zâ*âà6 ?Q°=±5_§NõYטУGêF§þt'5Ö5¹¦k’jÎyóæU¬XqÇŽçÎûã?¾ÿþûÝ#“}X×ôE_]Ã9Ø¢E ù*1±îº†Ý•-[."_hccóîÝ;ùgB×àÈÂmÛ¶Åùˆ%C† Á¹Y®\¹Áƒÿúë¯_»®…††BhþþûoIÍ;6l(œ@ËLRøçŸ¦ž7Úu Ç=¶\¼xQ¼ô'jÖ¬ID_½zUQ“DPP³³³ú·kò7¿_}õÕ«W¯PÉÐ8Zâãã£qƉuQ™®i|ÆO×t„uMAnêG\Éž={V£F 6ˆûl&?ú¦/ÙÿvM .-é£k’ê+Ü‰Æ »ˆˆˆ€~Éã]Ãi‚á!ŒhÅŠ£Kð¡Ó§O¯^½šâëøíVݺuKü„ŠMš4ITÈŒ3‚ƒƒ›4iC uÔ5I%Ræææô5'aff†B%&&Ò†’j®9ñ „xúã?Šò4å] –-[¦qdÓ±cÇŠùßÔa]cþ–§kŒÁ°®)ÈM]c "¬kú’‡º`]ïß¿§0´B£èø)t­zõêöööÐÈ“§§gÏž=±<)) v%dE—ž¡$dò%Ôu¡C‡nݺ•bGâ Ku ¹õ÷÷—> ¬M éå»wïJ–,IKΟ?ߦM ÛÙÙÑÜwØäСC"K™õ /såË—‡¿j,)ÁºÆüÖµœ€uMë£Ö5}ÑW×/^L£å ¢ë¢^ºV·n]E×''§ù¹®AS6lØ0zôhìÎÊÊ*22rÊ”)K—.?~¼]ÓøtM1Á®(]XX˜£££úru>;C:ˆŠ¢…iii4Æ›ˆvìØ11Ä®¾º†²”-[–Fó–2ÜDëóX×rÖ5¬kŒvX×ô%¯ºÈ?´‡mˆgl»víêß¿¿ˆ¦Ð5Iæ7æææ 666ÚuM,|ùòe›6mÄpkHDXŽ(|V:I]{£©Ij#ÔHÙÓ5I6±¤z—§ýšËºÆüÖµœ€uMë£Ö5}É]£ID(LC–.]ªDK¢¢¢è5I“®‰QvÅ.tÔ5zGã Ê'þꫯèy•üƒý””[[[ ç]£òÊl×®˜ B#¬kÌ?ºÆ‡„a]SÀºÆh‡uM_rH×*V¬(¦Ü¸¹¹ÑòO»®^½Ú¬Y³ÐÐÐÊ•+Ó’Æ߸qƒÂêºÙªU+KKË={öÐDgzéÍ8"©4Q<Òóòò£‡ ýråÊeddx{{Ïœ9“æ]ÃQ-e´°°ƒ oÛ¶ 'e‚éë*Ç%މ¼ÎHÁàܹs/^¼@¥áÌá^Ÿ:R¥J4LZ"°®)`]c´Ãº¦/9¤kè/^|÷îÝâ >>>–[×âÅ‹§M›&iÒ5uôÒ5š<—Â4+šôiJºÌ²-å]£éÝDlE“Ž=ªY³¦ÆÌ›¾®eùÿc˜lò¯ýë÷ß×uMëZ®«ˆ¤ºŠ‡‡‡ÓLV p7Û²eK+ŸÀº¦/¹¯kæææ¢AKIIf 8ê’’’$ÙìF×5I5:ŽIuCzè¡ š”Š+Ê£íÙ³GÌß0yòdooo)Ou­B… b|šÃ-$$÷ü´º"Þ&+`]c˜ìº¦/¦§k¸\uéÒ7îhRó:/ÿàüùó8ðpË>iÒ$ùGÙrpEÑq¼AÆ qÕ9|ø°ñrúX×ôEG];{ö,4KRy¡MÎíÛ·i1´hÑ"44ö/¦º/@ œ×8œ¨/§ƒƒCÍš5oݺE6ƒ#&3P'((ˆ¦ÝbLZ\Áal’êEY‡4î4mÚº&†}ðà¤ÇÖÖ¶ZµjØ£x‡fgôèÑ’êÕ-½‚´··wss£µ¸c133C"bxˆJ*©\M,uC„\¹r…æÕýꫯ䀈ܼysd#::ÚJß…#ÃbÆ6???z¥‹_ÕªUå3 ‘®¡²/^,F7f˜l‚æ€fÄc]ÓÓÓ54µÅŠ£Ïwò:/¸ví@ºfØ`ר¤téÒŠ±ìëš¾è¨k&ôËÛÛ-3 JòôéÓ)S¦œ>}šÖÒsÁ<Í á"]ûþûïÅG‘ “}þúë/ÜHI¬kú£¯®áU‡»[܃Òdíâ㤤$ÔíÛ·o%Õ8“bâÔ©S*TÀÝ*nî­¬¬<<<è>ûÕ«WØ5ËÆÆ›Ð-²¯¯/ÜÓÓóÞ½{h­­­ïÞ½[®\¹5jPj¸b“)f¶~ùòå³gÏ>|øÐ½{w\'è»l+æY Äw¥J•!P.ÄGd;;;¤I·ÝØv¶ËÖÖÖÅÅEŒ†%jW™é2@©á¦þÌ…‹‹CöœE¿3gÎ@L›6mŠ |øe„ØAªð–†¿0QEûŒ­àppDÝ+<3]Cµ (Hƒ ä3-¢˜ÈsBB"àòœ‹< 9EÍ`Ïȶ¤zb‘’’¢{ftuM_X×333œƒh%èMŽR48©q§G‡kÁ…ua „uÍ` Ó5ú,lA_7ÓPãáÌ™3;v”+-§7 ÿOþ ø‹¦ZÛ§OŸ“'OBMêÖ­ ]{ÿþýùóç{ôè¿ø——ׯ!.ØãEGGCÈ~ýõW1@€Æ—¡pAÄ¡‰«§N 3¦‹ìAÔè³Z²eËìêIÕ ‰â˱U:u´| ®Ž–—¡” B×._¾ÜªU«ú†ÖHßßÈ7á—¡ù Ö5çšYT‚8¡p¦+ŽÞ ëÃëšÁ¦k¢÷Ù¹sçÚ·oäì쬫àmâ§|ð'аìp¦úõëC×Ð>|ø°víÚ4 RhÓ¦ÍáÇь@à 1"eø™ø[£®Ñ(PòÖ[­Zµjüøñâç˜1cÖ­['ÏÏÖ­[‡îêê ½“ /€¨a_êeׂaºVµjÕiÓ¦!ê!±®å7X×LÖ5†1Ö5ƒ1L×D3EÑûÐÅ‹¯_¿^tŒG%{zz ?ÆݻwWŒó íëÙ³'ÜKþïëÚ“'O`KØ R€è éPL Cˆ,iÔ58_@@€¢u•ÏrˆÄáj06y„ôôô=zˆÚ6lèçç—eiG_]ËÈÈ8p :/©zÀÁåÓœ°®å7X×LÖ5†1Ö5ƒÉ¦®={ö¬zõêP«çÏŸ÷îÝ{øðáHM̾,œ†ŸPÅxðèËÙ³géÕ*™v];zôh‰%ðO¬Zµª<)ñá¿F]£q¡´´®Ø®9jÔ(õU8T°Ó%K–À“Ä`Q£¯® Nœ8±|ùrÔgŸ>}öïß/ß„u-_Áºfò°®1Œ°®Œaº&"Oš4iÅŠ¯_¿^´hÒͽÂËR×°ÐÆÆ&99™~0`÷îÝÚu nóòò¢lò/ºH×Μ9Ó¾}{†}¿uë–ç366VÞÍS‹®Éã@I8 – É.R¤ˆÆ‘™2ÃÏÏÏÓÓS|‹¦¾‹ÌtMD¨[·î½{÷äKJ•*…âèž½`]Ó¡k&|5/äP_(Ö5†ÑÖ5ƒ1L×P¾¾¾×®]ƒ5kÖìêÕ«GŽA:uúôiÈÄ?ü•¥®uèÐáܹsh  V0?ŸÄÄDíºvåʕŋϘ1cìØ±¸.>|ø°sçÎØ£˜›/^ƒ…Ô&U¬^½zXX²cÛ°aÔ)SŸ²©ëÚ–-[ }ûöµµµEæ/]ºDWå[éÛÕ^U¦L¸Šf®BRé&þ"ŒQö›7oR¶ñâ Ë8p`‰%†ºÿþ;v 4H$HO(/^¼HSýSnÖ5}ºvêÔ)~ºf’Œ=úÿ†&f]c}a]3Ãt âÒ©S§>Ìœ9sþüùô)LhÀ€ïÞ½kÑ¢®ñˆY»vmDCRëy@¤§§Ã¶mÛ†µÞÞÞøï@†,Ô*44!¤ 7²”7RWB3dÈçϟ׬Y XµjUùuËáXþþþ‹^P¢+VÌ;é{yyawvvvòrÉ{‘‘‘Ó¦M;xð`ZZ²„|*FD3`I5>\›6mpJŸÀh¼¨S÷Û¸¸¸©S§îÝ»799¹J•*M1Ê<‘””4lذ}ûöÑO£_AX×ôEècÚ°®1ŒÞ°®L6¿]ÓÃF”•oeX 9ŠQ²dôr=AÖ5}™7oî ò:LŽÃº–ã¼}ûwÉ™­Åå¼[·nþõ×_-,,p]Ÿ1c†ÆÈ«W¯NMM•T×°I“&É»kÉ×bùäÉ“V}È鋜˜~X>’Bîúf0¹£kLÁ…uM_BBB‚‚‚òÛÝctX×rêË–ÙÚž={:tˆÂò;{‘׬Y#Þž 6lË–-bÕ‚ fÏžMáÁƒoß¾=û97 QŠ˜˜£%Ý¢E‹«W¯"P¡B…W¯^7qÝa]3Ö5F;¬k £ÖµG»®É¿†ÎR×@ëÖ­/]ºDá+W®Ðì„ÑÑÑåÊ•£­ ø´Å¸ˆRÀihBF#ºVÐ1½)ÞãºÆ0a]Ëq了ÚNLL”¯µ°° áÚ%ÝtF ?SSSÍÌ̬¬¬„ @ÝÈ$ò Ö59¬k X×í°®1ŒFX×r…®i‰©‹®IªVŒºÐƒÎ;C–/_N?ÕÇI—T#¹¯X±ÂÏÏ/>>’ã:t¨<ÎŽ;†+¨šÈÉ“'CÅÚ—/_þðÃ’ªJŸ>}3mÚ4Xˆ››Û²eËlmmÅ Œ¯_¿Föh«ºuëÒø¥5Z¹r%õë×777ONNÀÏŸþùöíÛ 1b„ØÊ!{÷CÇŽå#H¬kÖ5F;¬k £ÖµG»®É¿Í×Q×@¯^½JOOO___ÅB(ÔÁƒ K–,IóUKª!zö쉺RÄBmܸ‘ÂÏŸ?¯V­…ׯ_?zôhE]\\¤Oj¢ž[1ž‚( ²hÑ¢d03fÌX¸p!PÀ¥K—*¶µ¶¶†õîÝ›~²®tX×í°®1ŒFX×r¡kº&ºë@ÑŽ_¼xqù’Fݺu‹ÂuëÖuttLNN¾}û¶••Õëׯ±0 ÀÝÝ"8;;׬Y366öÚµk´Dˆ”Ð5EcS?›ïÙ³‡–wéÒ…ÑÕ®]{úôéòV©RåÅ‹†¥-^¼XÞ‹Ù.W®,çñãÇ´äúõëô@‘u­ ÃºÆh‡ua4º–ãhïj`°®]ºtI>|åo¿ýöÓO?É#@ÔÄ›»wïîׯŸX5gΜyóæIªr4‡4.œ¸|ÒÚC‡‰§Y‘‘‘0'ùÓµR¥J¡uqq¡i1°ÄÖÖ6))I½êß®‰UÐøÙ˜1c°Ó›7o¶k×®D‰´ÊÛÛ[ŒB"†í€h†……I¬kÖ5F;¬k £Öµ''t J—.'_¨˜Å¹S§N§OŸ–TÍ‚‚‚4¦cee•––†¬¨N:´0%%FáƒöêÕK®k¾¾¾<éÓä6ZJ¡E×fÍš5þ|±üÞ½{õë×§pBB‚èKŒÌKª‰nÒÓÓ%Öµ‚OþÔ5™[·n:t¨XÒ¥K—“'OæŸæqÙ²eK—.}ûö­bÏÜu…ó§@¥ÏºÆ0a]ËqŒÞÕ@R½v\³fb¡b†ÄªU«Ò GñJTÇ=Šå›6mòòò’ëZ–~©‹®%''Ëû:\»vF$Ñ’ZκVÐ)(ºÖ²eË+W®èÛ|¸Gþé§ŸH†Ê–-%±®|ôÕ5ÔÛúõëûô郻”»wïÚÛÛ÷êÕKaíڵУ‰'^¿~·%•+WnÛ¶­8Ònß¾€ã 768Úå^¼x188w>:tëÚš5khz·ÔÔÔI“&iÌ´ ù—gæÈ‘#‘‘‘[¶lYµjƒØcË–-Ûêe<4P* [mÞ¼YÂéO£ê;vìÅ‹"‡ûöíÃGß}âÞ•ܽ{÷””ØOÑ¢E©)„3=99ÙÃÃC|±|ùrD9r¤¤êŽÊOKKCÝÒZ”‹4ÔÚÚzذaŠq_¾|ùýû÷8æå¥ =uêÔØ±c}}}>|X®\9dLK‘Y×F#¬k9Žº¦ÚvöìÙ , %¢{.c 4 …¸öÐ'k’ªZp ÿòË/¸Œ½}ûvÉ’%aaaÑÑÑX¸sçN1°’2d®7ÔX\áèùœ^º&º¬Ö¯_õêÕhß±„dQ‹Š/í$Õ%ÊÓÓóàÁƒâ…ï¢E‹¨o)ëZAG_]ƒ @:vìxíÚ5Tæ£G¬¬¬p[ZZR:¨ )>>>H·ðê"›sçι¹¹eddàÆCEÒ§wµkׯ± 'صk—еzõêA¶ð_‹W?PCBBªT©‚¤ªU«B†°/DÃrœ5<@âȧ\uíÚU>ÿ/‹/Ö«ÉuqqAY¨çuÆ áaXB‡\›6m`œ"5wwwS„Q^ //¯ýû÷;99ùûû#‘wïÞ‰’·bÅ ¨-nöÐ 8?%•_ÚÙÙÑ·Ð8¤ŸÏæ± ¸£ŸŸ_éҥ߼y#Ï'´Ùûæ›o° õ€t°íZ ?CÁ!²pA,ÇY 'ˬȬk £ÖµLjº†k®=ôSÞ—SR5åh1)|áÂê4ŠË‰ƒƒƒú˜jr×üÍ™3G=c¸‰ŒŒ¤qnõÒµ)S¦À·äI©»¦q@\ÿp-T/¸|vTÖµ‚ŽaºVªT)z¢ ç@xåÊ•bØ:¨.\Hb¡¬@æàO›6m¡‹ôiT¿¸¸8Êõë×›5k&îv¨².] è;ø"Î,Ú/t­\¹r"‚ö—¡4±¯M.ôåRl˜¥®¡­xøð¡¤zN6f̈,Å—/_nÕªÕ¾}ûè)¤º’k¥öyZ±V¡k>|°±±«½|ùRú4<Шƒ9隘¹_¿~h±If…e]c°®å8h%Å[˜lêšü5èÛ·oK–,)€“Añ¥—¤ºÀ ÉÆE ÖâZyÂ¥«fÍšbC$…özÏž=111håÑ”ã'Úz‘ŸÌŒ33ýÚ¹sç/¿üÛ£ŸâfZûÛ^, š9sæÙ³gq9Dézõê5wîܲeËŠ8ÂJå#ýæ>¬kc˜®‰é:p`ÛÚÚÂØÄÑE•z¿œû¨|q“˜˜hgg‡D=ZKJJ¢ÐOž?….ãî‹Í&]£!¤O%d]c°®å*ÙùP7>ò͵l䓲 ëšÁ¦kò>4t["®Ì܂ƛ¨U«­Â¿àÒ¥K4󇃃Ctt´âÆC];xð`Ÿ>}ÒÒÒ詳:9ÔÕÀ0];~ü82#©IâöÆÏÏw;RV6–eu]£w^¼xQ¹reõH×%úžoÓ¦M#GŽd]c}a]c½a]3ÃtmÏž=}ûö¥%Ð5úž~jѵzõêÉ¿jÇ’²eË4Hn6Ò§uѵS§NuîÜYî% òP× ¦?ÖE׊-š””„:¤êè«kaaaNNN·nÝòððqÐöb¿ëà Ö5†ÑÖ5ƒ1L×ÚµkwöìYü„¥)R¤wïÞb®³ÌÜ¢Y³fÉÉÉwïÞKbbbè¿¶zõê &}¡¯-uѵÈÈÈ *´nÝúÂ… ´%—d²C‡ð?o<ƒƒƒŸ={F ¥umàÀ>>>P"jåÊ•+¥‹®Í›7oîܹÈL•*U(©ððpˆÐW×hÀþýûïÚµ ?‘ dfܸq¨g‰uaŒëÃè ëšÁ¦kŸþyøðá8—^½zm¢™¹½¡sssƒ‡ÙÙÙáïŠ+D4úÐ fðüùó!C†$&&’®¥§§Óxýƒ ºxñbRR\[‰Ý­\¹râĉ½zõZ°`´¬_¿~pÅN+W®|úôébÅŠÙÚÚRD·«¯¯oãÆ4hpðàÁéÓ§>|™×E×>|øààà»B X²d = #P!;vìØ°a6eŽŽÆ1l§©Ñ“<”öŒµèC‡ÁÏÚ·oß¶m[DFúômëÃÖ5†ÑÖ5ƒ1L×._¾ À•ÞÝÝay'-‚à%P+úA¼½½Å“øwÀEΜ9Cý`°|óæÍÐÁÀÀ@y/<ýˆˆˆ^»vÍÅÅåĉNNNò˜ð¿¾}ûúûû#Ü´iS‡œÏ“÷ÄÄÄÐØŠ¬k #‡uaô†uÍ`„®1ŒvX×FN¡Ðµ.]ºìÞ½›¯—Œ±ˆ­Zµ* ðññÑ“uMÁ„ Ö¯_ŸÙ #øóÏ?===ó: “_0e]“OÄ09“““Ɖ³¬k Ã0Lö1e] rqqÉë\0¦LåÊ•_¼x¡%ëÃ0 “}LY×BCCiðñ’%KÒ«+†1 ééé<@@Ì¢¬k Ã0Lö1e]KHH(^¼¸¤ÂñÀyÆtxÿþ}±bÅ$Õ`ª;vìГua†É>¦¬kÜ3”É!¸g(Ã0 “›°®1ŒÞ°®1 Ã0¹ ëÃè ëÃ0 “›°®1ŒÞ°®1 Ã0¹ ëÃè ëÃ0 “›°®1ŒÞ°®1 Ã0¹ ëÃè ëÃ0 “›°®1ŒÞ°®1 Ã0¹ ëÃè ëÃ0 “›°®e jæ‹/¾È|1z“ßþ¬k Ã0Lnº¦2ƒÌj&¿©ƒiÓ©S§Ó§Oç·£”ua†ÉMX×4àéé™”” X>f̘õë×›påCfÍšuòäIíÓ¨ç>¬k Ã0Lnº¦S§Nõöö6áct„ua†ÉMX×þÁ_|A›$$$ȯ¯ŽŽŽH-%%%55µD‰Ti›6mêÓ§E˜9sæÎ;‹/¾téÒÎ;‹ ÷ïß?lذÄÄÄñãÇïÞ½áÅ‹kÏ@Ó¦M­­­_¾|ùàÁƒ)S¦ ç>>>­[·ÆÚ .¬[·îîÝ»666…yóæÑV]ºt¹~ýúŸþY§N'Nôïßyسg^Õ%ÇÂÂÂÒÒòáÇ;vľþóŸÿ+VLä°]»vgÏžEóçÏ£6þ÷ÿ×ÍÍ áü±aÆ»víª_¿>ê R}‚‰'îÝ»·téÒ«W¯þî»ïľV­Z…ÚKNN>|ø‘#GPQ¿üò –£ž)j/##C=“cÇŽ=tèþ5ÈŒØKîÀºÆ0 Ãä&¬kÿàÊ•+VVVß~û­ôÏo×üüü` 7nÜ·oßýû÷“’’$Õ•ØÞÞžQ©R¥?Bªüýýa£FZ¿~=mqÁ½[·n*TèѣǿÿýïÇkÉdb oÁ™ d`fff...ãÆÃ~_¿~=`À€R¥JÅÆÆb+x¤­­-œ &&¦L™2EŠyÿþ½"Y)ó¯ñ4f!£ƒ êСÔ{«7n|ãÆ „QQ7oÞD²ááá¨ülÛ¶íìÙ³¿þúëmÛ¶¡ð¹ãÇãQ±bEäjÇŽ¾¾¾ ,˜>}ú¢E‹(AøëŒ3`œ(]ûöí—,YrîÜ9,§]Ì;ªÈ9J KCàßZB…£à:Í(°®1 Ã0¹ ëš2“›)S¦,[¶L±|áÂ…sæÌ0Á$Õ#ŸÍ›7Ã$ÌÍÍ¥Oºq9qâÅ×ÞS«ÜÝÝ¡}P±¸¸8DÞ»woÿþýSSS‘ |ÃÀÀÀš5kІ ‚Ÿ§NÂ^Zyýúõ&MšèR"í5€RÀQ@˜° -ºóôôD„ZµjA1K—.––6qâÄM›6‘ãˆ&«è'éZß¾}Åã@yA<¨È9ülûöíñññ666’êQ¨ŽE3 ¬k Ã0Lnº¦½t »HNN>|ø0-¿wïÞ¯¿þúáÃ+++é“®½|ùò›o¾Ñq×õêÕ»{÷n™2ebcc‘æéÓ§;uê]³´´D„û÷ïGGG§§§C¤Ú¶m;kÖ¬ùóçÓ¶-[¶¼råŠÜ{èÞ§U^TQj-ºöüùsØ "Ô­[õ ŠP¤Hd~çΔ… 6lØ $]ƒ!¾zN4ê,­xñâQQQyÕE—ua†ÉMX×4 —®!²™™YùòåÅ’?¾xñB®kz=Öª_¿þ;w„ëœ9s¦cÇŽÐ5 ‹Ò¥K¿{÷ÎÁÁž´Á!: .¤máCƒ :vìX×®]u/lfÙ>ÕÀîÝ»  ‹®9;; ]³··óæ 9"2l‹ô322"##)Lº–YiÔ5$ˆ ˆˆÈf1 †ua†ÉMX×4 —®•+W;oúQ×úöí‹R@Œ*V¬ˆ˜>|°±±º–––†Ÿ0E˜\zzºî…Í,RæºKûÏþƒ°‡‡²š¥®¡,–––ÂÏ kÐVúPŸ®1 Ã0…Ö5 d¦k£_~ùåíÛ·%K– ·mÛæååµ=çÏŸ?{öl k×µ¯¿þºG«W¯–ï:3]kÑ¢=Ê¢˜½zõ:|ø0é, ñãââ~ûí·Aƒ999…„„Èw„%ˆ©2 ºV´hQGGÇÀÀ@I¥ªQQQYêÚ’%KP!0Zú¼x{{Ã})l€®¡È³fÍŠŽŽÆ.ðóåË—¤°¹ëÃ0 “›°®}ÆÇÇçÒ¥K° ðÞ“‘‘QµjU˜Eˆˆˆ€`!Í:˜™™ÁÒ¾ýö[ÄiܸñíÛ·»víjnnކº|ùrÚD»®a_={ö³ÿþ«W¯*V®\yòäÉâ'jóæÍ¯_¿ÆÕnäááAËOœ8m­¬¬† R¥JÿÖ­[°¨-[¶hÜãO?ýäéé9|øp±—ÿêÕ«Ãf +111°=Hɺuë6lØA ݺuëû÷ï[´hBa_]ºtÞA}¬­­ÿýïS"ðHü[!sb°4ÄÄä$ËJ P4¨'ö…0Ltûöíò"x{{GFFŽ3†æ€Ý¾}ûŽ8oÞy ¹7oÞ,Q¢ö¥è6›Ó°®1 Ã0¹ ëÃè ëÃ0 “›°®1ŒÞ°®1 Ã0¹ ëÃè ëÃ0 “›°®1ŒÞ°®1 Ã0¹ ëÃè ëÃ0 “›°®1ŒÞ°®1 Ã0¹ ëÃè ëÃ0 “›°®1ŒÞ°®1 Ã0¹ ëÃè ëÃ0 “›°®éÇ€¾þúëE‹%5-ˆY rzGùœ­[·úúúîܹ3¯3òX׆a˜Ü„uM?¾øâ‹R¥JÅÆÆ%5-ØÚÚ¦¤¤˜ðGGºvízâĉüV¬ksÿþýààà¼Î£7ýû÷Ïë\0Lá…uMPW¹91ežì1’ßêuÍ`þçþGLnË,LøbÁ0ùÖµÏà‚ '  133ûøñ#¸В(È/·X+ˆƒ˜">"cZEK4®R_›åŽä[! âËóOé#ŽÆmu‡ò€ÔÔ“¢Z;¢"ˆ½c ­¥äf–óÌjI{=ÈÌNI ƒuÍ`X× .&|±`˜üëÚg Å‹ÇVLž<ÙÛÛ 7n|ãÆ ZKÑÊ—/!¶ò÷÷¯S§Î¡C‡zõê…ŸÅŠKHH UÇߺuëêÕ«ÇŸ5ŠÖ6lØðöíÛèÓ§-¡ÿ…\eÿ7N˜0!55á¢E‹^½zµnݺòü/\¸pîܹéééøæèèH«à›7oÞèXu””ëÁƒôèÑãðáÃb•§§§¯¯/Â5ºuë2^©R%¨IPPÔ—ä5kÖ ÂæÍ›Q ,X°y£œ#3/_¾´²²¢-Z4sæÌ€€777ëüüóϨ> ܶmÛsçÎéX4£Àºf0B×îܹS®\¹¼Î“¶¶¶8µÑ*J¬k “§°®}~@rF¢·3fÌ®]»RRRäq¾ùæ؆XBºV¿~}\~nÞ¼ùí·ß2dÛ¶mX5bÄøŠ««+¤®FŸaÅÇÇC ±Òãçç×´iÓ3gÎ)RäáÇîîîâ­_™2ebccåÿüÄB¸Ýþýû‘7D~þü9–––"oø{÷îÝzõê%''›™™ %¢Uºÿ¯)>œ ™ìС¼ *j aEº¶aÆîÝ»;88T¬XKPù()êêâÅ‹mÚ´9þ<þ&%%¹¸¸ØØØ@b(ÁÅ‹Ϙ1öóèÑ#{{{,wvv9éׯßÞ½{9?~üx·nÝ k^^^ÈϺuë :Í(°®ŒÐµÿýßÿýúë¯ó:;LÖ”,YòÝ»wëÃä)¬kŸÁ…¿wïÞrãíí=uêTye¦k·nÝòðð —ÐÐPé“®ÁÃè¹QXX˜““\²d‰ôI×kÔ¨¡žu]5jE\ä e 4ðññ0`€ÈÂX¢žÚÏ?ÿŒ¤V­Z¥Cµý7)éÓûÍíÛ·:TäD‹®Á·jÖ¬‰¨û÷ï‹"Ô®];22ŠC)ÀêÆG¥OºF2§žèéÁƒG) øõë×qqqbI.ÜÆºf0¬kÖ5†É°®}×ûüqïÞ½B×à7$:ò8u-::š®ßˆ`oo#}Ò5ÅæôœIú¤k™Õ¿º®AΠhŠÔæÎ;gÎñsÅŠȰ.•“eUHŸZçÝ»wÃuѵ   gggD¨[·î½{÷PoÞ¼ÉL¤D‚¤k™ÕƒF]C‚ŠWÒ¹ ëšÁ°®8X×&?Àºöcéø)e¢k;v|ûö­ÆÇlèZ‹-îß¿OB²_X`]3ÖµëÃäX×>££®9::†……‰%YêZñâÅCCC‹-Z«V­ñÍ–v]®#_hnn^­Z5ì1>>‘‘A=¶DÞ2Ó5úhÔ5777* !kܸñÿS¡]× VõêÕëÙ³'êE˜7oÞ¡C‡žE±-¤ ê‘B©iÃK—.M˜0å•T.Uª%8gΜùóç+ò†Í]\\2«páÂ…‰'R‚Æz¦¨;¬kúVà`]c˜üëÚQw2±D÷^‡Š˜êß®é²_ÃÐ%÷¥o–äñµTcžä-‡`]3ÖµëÃäX×r]t)ˆ°® ëZƒuaò¬k9뚩ºf0¬kÖ5†É°®å ¸¨£™Ã;÷wÍä(¬k“ßtíÍ›7ß~ûmfk=<|ø`aa‘Y‚±±±C† 9qâýüóÏ?[´h¡o‰ŒëÃäX×FoX× Æäu-..®téÒ§zõê5kÖ}ºråÊû÷ï¯^½ºXÕ¬Y³ëׯc“6mÚܼysõêÕC‡ÅrÜU/^¼øêÕ«ÈFçÎwìØaiiI› K#GŽlРr‚Œ9sÆÜÜ\$:pàÀ{÷îõë×oÛ¶mb9ÒAÊGŽ)_¾œ~.\¸pÆŒú&˜ç°®1L~€uí3¯_¿~ûöíË—/!IgÏž…Ü`¡­­-½ÅÏ÷ïßãòÿâÅ‹aÆõêÕëàÁƒb[º/¦¬[·uü…íÁ¥h-´ÉÏϯ^½zp©ºuëB€h@Nè t ËSRR¼¼¼°;´•ÅŠ“Tí;ò3kÖ,d©wïÞ¸ lÞ¼™RCÊ ÀM9. wîÜ™>}zBB–GFF~óÍ7ÈÆÎ;‘y¬]ºté”)SD&‘¦««ëãÇu¬UR¥É“'#ÿ;vÄÕèèÑ£b•ª@èrû믿"«©©©¨„ZµjAàð™:uê’%KÜÝÝ[µj…\Á•a™pÐöíÛS‚UªT Aö&NœÓ…ªÂe±üáÇ0l˜‘‘¡¨ó7n4mÚõƒ}EDD ¶GÇ¢éëšÁ˜ª®Q—œ³EŠp§„{œÑâÅ(ŽIjX²L0ÿÀºÆ0ùÖ5%ÁÁÁÎÎΡ¡¡•*U ׬Y3~üx___z45vìX¸l 1)‚F]kܸ1®ë¸ZÓÏÚµk'%%‰Ÿ¤kbFBÑÏ­¼]×®];&©Úwl…lHª/‘ãããá@’ê!“µµuÍš5a?´!öByƒVÂ`Ä7Ë0!¸.QbH%Eyu¬ćEQ¯7”ž'|H‹®!Ÿ¸’áßAC½c­££cXXX5 þþþ”‚““Ê‹ÒOÒ5˜¨zN`l꺆RêaÞô3<<\þ¸ÎX°®Œ èšFhÛAƒᾈ–9r¤{÷î’jZR´ôJ´jÕª¢`]cFwX×”hÔ5WW×'Ožˆº‚ý888ˆæXÊD×ÌÍÍëÕ«7bÄZ¾víZˆˆˆCº†ûoù MâîÝ»/^¼ ;õ~ø¡bÅŠÐIÕ¾ ?À=:¤„Rƒ·¡I=|ø0JªH 2T¡B…Ù³gSÌœ?^äÁ€Q0äÊøí·ßÞ¼yS¤¦E×(nÙ²eذaXYY}øðÁÌ̬uëÖ(#EX°`N$HºÑ#QÉѨkH¾~ýz½ ¥/¬kcº†3«]»vô³M›6ø)ÖN:ÕÛÛ›ÂsæÌ™;w®ÄºÆ0Œ>°®)ѨkeÊ”‰•[.À»víêß¿?-ѨkhŽK•*±KÞ¿¡0éšzý×®]VW¹reÈG@@.'‘‘‘”ààÁƒ·oß.ýS×Þ¾}[ºti¤Ö°aCEjؤX±bòn$ňPz7n|ãÆ É ]ÛºuëСC€¤BU€÷ zEú /_¾¤0éZf‡¨F]C‚3gÎÄ . .°®L~Óµ7oÞ´lÙRR5ÚçètrrÓI)À7oŽ3QRö8鎞jã¤ÆýŽ^JÆÓ§OVž€uaò¬kJ4êy‰èÍk5N:Õ±cGŠ q ´Å]»vÝ»w¯ÆiÔµ}ûöAwnݺåááAKàYêJŠ&uãÆâ£fjÀÅÅ êXü,Ñ®kâÅ¥Žº%=zôòåË5îË]C‚½{÷Ö®PÙ‡uÍ`ò›®1YºÆ0ùÖ5%aaa¸ë½pႼ/=½æØ¿Ÿ>}¤Oš‚k¶x,Ô¶m[l_¼xq±Uçί\¹òþý{±äСC½zõ¢°F]Û¶mÛ°aà gô‚rËR× VVVX ¢t MÔwWG$òáñ \,ÇŒ#~Â)Qd^ÇúÑ¢kð§ñãÇ/[¶,66–j&K]kذá³gÏÄ”ØàâÅ‹¢æ еúõëß»wO,„G8°D‰òU?ÿü³“““üy›vßÅÅ…ÞºnÙ²åùóç¢?~DçR[°`~Ât/]ºtáÂúÎ¥èÙ³g½zõH‘"E‡ªõ†ª@1‡ &ïx±bÅŠˆˆyRc¶ Ê–- u¾wï^$ˆr¡¼:–K/X× †u­À‘‡º†“½mÛ¶¹¼S&ÿcÂÞ¢Ö5†ÑÖ5ƒa]+pä¡®]¾|¹U«V¹¼S&ÿcÂÞ¢Ö5†ÑÖ5ƒa]+pä]0`€è}ÅNÌÍÍçÎKcˆš°·huaô†uÍ`X× ùA×vîÜ™C60ˆêÕ«ÓÄ»&ì-Z`]c½a]3Öµë“O`]3Ùb³®19ëš.hœ3c„ «W¯–T§§]^ä‹ÑR¥JÅÅÅI¬kL^úf²Åf]crÖ5à¢>{öìsçÎ…„„H™4©uêÔñ÷÷‡Æ™pËcª°®1y ëšÉ›uÉ!X×4K[½zu¯^½¼½½iT<õ8µjÕ¢1ù˜‡±.éééºÄd]cä°®™l±Yט‚uM;:uÊR״ϘÎääs%g3)ˆZ‹-æÍ›÷ÝwßéŸu‘úf²Åf]crÖ5íhѵ±cÇÒüåË—Ïõ¬1zc”®8þ›6mzûöm a-[¶Ôe+Ö5FëšÉ;ÏumëÖ­'Ož,Adl¢XÒT €ü`•ü£ìŒŒ l…¶²²²2¤^>‘””„<ØÙÙ!€L"ÿ° Z•€+VŒJ=RÅDÁ‘„‹)’’’‚äþÿöÎ<(‹#ýã“UW…Ë"¨%ˆ‘%Þ$JˆˆPJ (Zë¼KD¼Q#Šw0`Šh<’*p誈7–¨¨¨ ŠÑ` A]ŽrY%‰×þ¾5Oٿμ‡ï;ÀËË›þþñÖLÏLOw¿=Oúf!W$üGbâ<‹4dIÈ7èÝwßÕ~Î<äßR縦_×,IFá>óÑ£G§¦¦2XPÚ!C~øá‡ &H5kd tùûÉf¡Â|Q,t-øG7Hò<EK?Y0ÅUv?.ÁhÃ2›ÞXåçç»»»Ó±!ÿTNNާ§§á÷›ƒ®Yl´ÅµÎ;———Óé† """víÚ‚S aÆ®pŒ¯ñôéÓ½zõ¢;a,¸s̘1ä ™ \‹ŽŽFEŽø€×¯_?{ölI¦.'''zÜýüüØØŽk×®yyyBÇß»w/Ÿ;w®_¿~Ø ¤¤¤ÀÀ@~Ø5X7ÃÿPÊúóæÍ[·nN[¶lùàÁºDØD^¡Š (¡c¸ÛÙÙÁvã <<IÇ;v  cÇŽ%.A‚deeuèÐ<|øða‹-Ž92xð`r!}}}ÓÒÒÈE³34>>>,,Œ„k/_¾Äëžó   ML€k1110ÎüUÔ;wî,ŒkÇŽ£j³$רQà•õ<às&tãqí›o¾¡!%»ï¿ÿžùú¶Ö¸Ú±¸#†D’Ë/ÃKº•À5‹¶Š©0"—.]‡(aíá-[¶p(Ò y'½Á5°Å!Cø4§àˇÉÉÉ`£³gÏöíÛ—ÜQœÿOŽÔê µ²²’i÷dee)†[»v­±)C"\#Éô=zÔÏÏOÒ‹ksæÌ5²p ‰ƒfÍš‘Áâ_Áp áïÖ­›fH´ââþ믿šgþ¸¦Z×ê´âZAAÁŠ+@QÌÅÍÍméÒ¥z&0\c=oÕéÓ§ÉúˆkzÌ…¸Mž}踤¤ÄÙÙ{÷î¥!tü1Òóòå˰!ŽŽŽ}ôÝùèÑ#T×a%`–ßž}ûP£Æ+† Æ™‡(DPšHÚpmÿþý°¢dÒ%Ù%y#8¼¢²²ÉK)ƒÊ6݃RŒ,U«V­@«ð3??.îîî¬}‘×ëׯsrr?~Œj?¬7òßÄÀ{Ž÷ôôÌÍÍ-**²±±éÕ«WÆ %¹:&FÀðìû￯ç‘®Yp´UàÚÁƒ‡Š6~üøŠŠ 6 Ÿ2®®´"\;~ü¸¯¯/ï®×üýýwî܉€ñc#¨É¥‘Î;kˆÊ%, qíß¿?¯ÿ/M\CÍràÀÒq 8*Õ…kĵ8¾Àé áÚõëש¦«V\spp(++3Ïü)pMµØ2¹×ꋸ†¯ȲgÏv>ÕM›6 >\¿? ×´–ýšºsçÁ½)qM’‡ÞjN_€+..f§ºpMâ´hìï’‘‘Ú82?‚?òòò@‡`Dö8JI“&iú¦µÍìÒ¥KÄ|š¸¦kãjhÐÚJ•d €*ü­ì©éÓ§ÓÁ„"ŒÊ ^Ý»wG±Ec`$yª5=à(PJ’;¯¹sç~õÕWô¡†Æšâ«/=¸†zP•.-Z´ˆFìIzqð6oÞ\ó]*p-<<é{ͦ;™ŒÅ5$ך5kôÌ_û“épøða0’Àµú#×ðâke—@°{ýúõ3dK1†k*d ®Qƒ 3b’‘¸†o_.ïÒ¸qc”¼ÿºpæfÑ1ìž“““Ä‘o‚À+€B­QPã¸öý÷ß?^?®i†“’c šBâPj3\Óôf­)Ô_À \³Øh«[w-00ðÀ8¸qãFÇŽ™;H9õ¾ÁIz®Åð!Y[[Ã7dwú¼¼¼233Qt <8888))‰q *©+W®ôðð@M®[·n°8` òˆ{0àêÞ½{GŽÉÿ}¨¦\»vÍØ©Zq æF'++ olӦͫW¯ÞŠk'QHKK™!âˆ> ŒV\ÃUòçêÕ«ˆ)»™Üa^a;œÓÓÓ]]].\x¥¹u.cqÍTáªO2×PÆ †CÈËŠ(R«V­ð±hæ|–Áj\cÇŽMHH Óº5¡4Q ÉR{“¦®Áø°{{{TDñ/ЬOCd\Ó”:\ƒPOæÚîÙ³‡õ’x\ëÔ©¥LUU26óÿùóçÔeÉçCØÌ   МInÌ;xðàƒ>ôltttdd$]Ukz—tL5`¸&ÉÝ—ÉÉÉÙÙÙlJ,ký¢±Îä‡ÙàÑ£GÔó+ÉÊ›7o–8\“äµ Ö®] ›™——Ç‚‚ÆÑÑqРAÔ¹„ÌÌo€¡À5‹¶:\CN"JS¤LQQQXXØþýû%yé Øöm #—kŽ]ƒMßµk<ìÞ½;>c|ÒÀ2Ö=Oƒô%yÌÇÆÙ³3gÎÄ'Ak||òÉ'¨Ä°æ%p[xxøîÝ»%yÈB||<”`80ŸáhË–-KJJè~"N˜ I®Ûþà[¯^½BBBf̘Áp 5*¼—á¾d|–tµ  §- ‘bãð"¼XöÁðaÐZš²(ÀÁÃ#GŽà¸f&K¯Žkü e!^ÆâŠ”(c4/á뮨¨ Õ¼Äwë× Ph!›ÚMQBÍJëÈpêÉ“'ÇŒc86+†k°TøÆÉ_"LLá"õmÛ¶ñ´­G/^¤nACpMkƒ:\:€<ð™3ØðË—/ó÷¼uf(*Ïla3fèh()p€ê=UÆ / 9@CÖÝ\'¸ûü"GÖ¤ŠzÊAð½·´jïØ±ƒþ&D¤¬¬Lâpͨe«H²U_µj&Nœˆúš5kVYY©+I®Yl´«¹«AíUÍkUÆÛðûÙš¨K+]O™ÊŽkùùù¬r)„2~åÊ•TA7×%P'v')ò ÊŽ¸VSY ž°"ö­©¦Pì988 ˜DaY}ßTŒï MHHˆŠŠBEŽ]=zôÒ¥K]‡Ze‚™¡zZXµuëÖÍŸ?_áȆµ®)0¤Ò³gϘ˜>MXš3OøÅAØãàÐŒ$OheY½Np-00pß¾}äÈzœlll¨E‰C«u*¼E>a äÎp-((ˆÍ fAb+=Q/Öpò¸f±Ñ®óM¨„,U†ãšBÆÎ e‹ò…"+¢„Ä?ÈpwWÜsõêUZCÇÀÀÓ(…#óïòööfËá(Cmá4tQîÕ«Wqq1 ãýû÷óÝ—™™™Ó§OÏÎÎ~÷ÝwwíÚÀnT¼ÎÕÕõÞ½{’ÜÞCËLP*):Cñ.ðÍ™3gÀ(‹/ŽŒŒd-Ó”2G…aDÙ}Q Š}ûí·t' *WAAA[·n­ªª9rdjj*[GšÐÜš!9yò䀸Fã>Q¶ÓÓÓQö |šŒ;vÇŽÔL²iÓ&„GWdu­»†¸ :”߈ †è[[[kõ§¾àZÛ¶mYŸ/ri~~¾òÉhz\ú£€À7õ!Ju„k´„]å—Æ¤±1Ò›¯‡køûز/×ÔIàšÑ¸¦ZÕÁ5pɉ'4/¹¸¸ðã»OIzqMÿX=IJJúüóÏŽúq–uøŸ¼?ï«W¯h3\j)QDﲤ‹Û·okíp¤e]¸FÍr,…—,YÍNÉgrôúõë ´k×NÏ„=Ëä"W¯^íׯ›õ ß<==/\¸Àvêd2\ÃßÁ¯šœ€¶—?%y<í+H«K’#ûßM†klº<Á /_¾¤ñm¤·â-ꉃ¨¨¨¥K—2÷êàš$Ͻ{öì™$rËÈÈ@â{Ùu×jPׄ„Œ–À5Õ2\«Îð2u¸¦(é”JYð ÛƒX!¸F€Å^GäØDEšÐ†åš»•+|“ô–‘§N "Ô Èù-IÌ×4…¿ ??ŸïËã›]AHH^Fx@€‹Épí›o¾™9s&ïþM„–ŽßŠkø:ŠŠŠxÚGµš¸Æ½SˆnéXàZ JàšÑ¸¦ZæƒkÕQ â-ñ¥g}A¸Ö°aC¶øŽô&»*üñ¸fmmMkqk•[¼CÉÉÉ‹/f$!‰¢££©\¯%\‹‰‰™3gë !íû¬ë*M„ϱ=¬Ë@òÚ:wî Öá·x7רf6l3hö8BHm®¸w²Ç׬Y³~ýúŠŠ ÐOpp0RR“ÿøµâøðÜ»woܸqÌ‘–á½ví[¢–Ï tÀO5ðóócËÛò>?{ölþüù©©©4бQ£F=zô˜¸V­kl™ZpÛ\+..nÕªUlllXX˜Öw©Àµnݺ]½z•-µzäÈ,//¥T›¸FBÊDFFÒª $  `AÈU³¸ÆËd“ÊšüÖPÕøtþÚ[ úÒ:ƒÛ@ \³Øh \ª% \S­ÚÀ5=S KKKù5â~²µ Œ’V\C–°³³W!$µíÛ·‚kÒ¢MSàˆÜ…P±–šj°uëÖ/¾øB3$´|[Ž•D[mÒdxèää?ÙúXµk¤£Gâëà»GÝÝÝi VíášeKàšÅF[àšP-IàšjÕ ®%$$Lœ8Q’—±€#¾±µµ¥•«HgÏžíß¿?P¦aƬ–···!\â%pÍb£]­kd>L‰kׯ__¿~=¬•f…[’÷ýlÒ¤ ?ª×ôÚ·oßðáÃÍ×$™­‘\l’WKàšj \«wª)\#Á”uèÐAѲ¨Kׄx \³Øh×ÆØµš×ô \òÉ'ŸÀ¨ÕaÌ×øškV×TKàZ½SÍâšQ¸&ÄKàšÅF[5®¡NOOïܹsjj*¿³².\+((X¾|ù¡C‡***<==7oÞÌWŸ={6tèÐ3gÎ8::¢¸Z¸p!ÿlddd||üï¿ÿ>xð`”dÎÎÎäîííGèXѪuÉGZW QFØÊËËÉ1??ßÝݽwïÞçÎão]© Î;w.]ºé3a„M›6‘ צL™²}ûvDüìÙ³l tJJʆ .^¼Ø°aÀ€¤ Û]úì³Ï>ŒL¸jÕª¨¨¨÷ß§-[¶D¿þúkØèÛ·o¼h‹/V„?::éöôéÓ+VDDD»®=ãù( Á‹‹‹ýýý“““ùddC°'NœˆWûúú"ðZû ®©–Àµz'kBf"kmu¸¶qãÆ°°0<üzüøqpp0ßH£ ×À“&MòññiÒ¤É?ü°fÍš¯¾úŠäJkÓ¦ °©Ëvº…€h'Nœ@ð:vìxúôé{÷î­\¹’.áí ´S§N?^kð¥ЧW¯^‰‰‰ptrrÂ=´6ü¡å¹—-[ *--mÞ¼9{¼Õºuk¶“j×®ÝÝ»wãââæ#GŽÌš5‹–ù&\sqqÁ«mllú÷ïïææÆúg¼~²AAA………ü$5€¼UVV6}úôï¾ûá\°`¿zõêÑ£GÃÛÜÜÜaÆÚÀX´Z¼jÑ¢ÒdÏž=ðïfeeIòxm¤ÉßeeggK2رÕÃ=<<îܹ³{÷n<\C:<|ø-¼Žÿ1úøãñï#ŽóæÍCx˸¦Z×êÌ×vìØ1vìX¿]ÈÜÔ¡C‡[·nI×,O*p–¬ä»­Pv†² ö8¾}û6`N×<€wÞyo7èòŠ–¸Ô5Õ{{{é˜ ¡X×®] V€tÍš5£U§™háM`ásâ¶mÛš––FÖS’§ªÑ4=Â5¶IÜüù󪈾¤mÔ?NgÏžÍ6à#\›2eʦM›È¥²²’ŸÅFyyyÒèhFáêÕ«l=÷§OŸ‚›áH§×L jâZii)j5éééŠUø“’’–,YòèÑ#|òŠ‘šªuæÌ™3f ‡ >V·µy;;»'OžÔ•­¸&$d¸¦­|akkëááñ믿޿†’¿¡E‹`¸=zü÷¿ÿ½qã†$[±Þ½{gff‚|÷î݃}jPÚæååõéÓÇÕÕµiÓ¦?ÿü3˜£¤¤„­Ö3Á "´·nÝZ¸páªU«èR÷îÝh0”?ýô¸wVVVÒ™táp¼ò?YšqT73tÍš5ž››HQ@  Œ”ôâÐÚÚºI“&]»v}ðàÁÌ™3çÎk®!1½¼¼hNNN¬åË—ãÕ ×p?<Ù±cÂoccƒtDFFF²ÇcbbæÌ™C9‡¿)77—Ü£££Q¨ã)ü¿øÁ‘´¯6]¸fU×(åyæ†víÚ5vìXd­ÀÀ@dBäÌê‡3--Í××ÿï”)S·k{¿#FܹsGñu3éŠ` \2 \Ó®”””ÂÂB‡àà`ÍÕ_>|ááСC@=À\PPP\\€iÑ¢Etéúõë×®]CxL¡"AP‰44nÜ€Ò¹sgviݺuŠ!t @æ' l×®]»‘#G*BXUU\CÆÒŒfëÖ­ÿüç?íííH¹ˆ=}ú4øµK—.ýû÷'GÄÉA…JNNŠ:¼š®1÷îÝ‹7zzzöíÛáÁÁÀé꣊…èH€0¼ åççgeeµsçNP ?Š/ÊÊÊBšX;vì¨xüÂ… ¸á7Y|¢áoÝ"TH·!C†ð|ýõ×øO)¬U×T«6pæÖÔ¬)5jTbb¢ÉÌ#j&¨ê]]í\,pMHÈô§ÀµÞ½{/Y²„ û3|Ÿšš jd+œ Õ µPœK׌—:\|?~üLC«3\Q5jÔhÊ”)eeeÇŽ{úô)NªBÔ.@E«W¯Æ[Ž?ŽšLÕˆP­ºxñâˆ#èλwïfgg³Êƒ¶oß>##§¨œ ÊÄûŒzΕ+Wà¡‹‹ ÃMŠÂÔ©SKKK) ’ÜêLK ¢fH£ ^¼xª£"¨¨ŠdffÂ[Ôaå˜;¬:*¨Àï¿ÿŽŠbÓ¦M}}}U'ˆÀ5!!sПׄ„jI׌• \£µ1[·n]^^²eˆkööö666 ¨—/_º¹¹«pzýúuölAAAÛ¶múé'Cˆ5hÐàßÿþwUUù è)))‘är`&f3Qš5k;ä}øá‡|(..~-‹y ØÁ"ÙÚÚVVV~úé§ÉÉÉZ£oܸA¨6Å#………’-ÒR@1ZAž<|øØŽÆàFDDÄÆÆ¶hÑ¢¨¨¨:{² \2 \R/kÆÊX\KMM `kß>%ŠÎPø¦Õ”tÚ´isçÎwww£Â©uTå[q tE\Ž=:xð`Ü0sæLé-5jM’ä¡lÖŽô¶ÎP­£*===oß¾M«†¾zõªqãÆÓ§OÇ¥7¸Ö¾}û¼¼<éÍPN£˜•—À5!!s%ã \~~¾ùì;.dy²±±ÑºK•.‹k;vsüöÛo´ -aá¸öâÅ ÚMëÄj=R‡kŸ~úiZZ;6mZ||¼$Ï$ø×¿þÅÏkQH®5iÒµ…-[¶Ð)x´¬¬Œ`‘pmûöí&LÀé£GœiOa£$pMHÈdɸ&$$dn2ד"KJJ@<†ãšj©Ã5ÿƒ²ÓÐÐPÂ)ÚDOUàü_¶lÙ—_~I§]ºtÉÍÍ¥{×RRRh` È•™™Ù³gOãSBàšYHàšéd,®5oÞ¼¼¼œ™©›7ovêÔ©®p-999((ˆ9R'£!¸†Ðfgg³Ån4¥×æÎ»nÝ::¥fHkBB–*kBBB¦“±¸TÔªU+œN:uóæÍ†ãÚóçÏsss2Ænd¤×: cƒÀ¨‹Ó\£•ávïÞ=zôhºZQQA‹’Ú·ooµ-Z´nÝúâÅ‹têàà€PѤTkBB–'kBBB¦“±¸öêÕ+kkkWW×ôôôãÇ7àRWS h}i;;»Ÿþyç΋-â·ÖƒkÒ›ÎJÄ}̘1çÏŸ2dïùÊ•+/^ý—¿ü¥iÓ¦xü…,°š••áZUU®zFEE-[¶lÆŒ+V¬ :qâÛmEàšåIàšé¤b!²²2__ßììl??¿mÛ¶µjÕ*++«[·nìÜ£Õ”©Xȃ4`À€“'OjÝöwàÀÿùÏbcc’<®á)¶E/Nù‰Ãà­¤¤¤  â¸-11‘o]“ä‘pË—/§¹À£¿ýío,­xýãÿ¸}û6ƒÏ¾øâ )X099™!)áÚ¾}û%¹%ÏÞÞŒèååeT"® ™ƒ® ™N Aø­Ø„ÌYׄ„ÌAׄ„„L'†kVVVu¡·¨aÆÉÉÉÏŸ?—® Õ©® ™NZ;ø„ê…Da!$T‡ú?j4endstream endobj 288 0 obj << /Filter /FlateDecode /Length 159 >> stream xœ31Õ3R0P0U0S01¡C.=C Âɹ\… Æ&`AÃˆÍ ÀRNž\úž¾ %E¥©\úá@i.}0éà¬`È¥ï 43–KßMßÙÙ È °±Ñ÷VÐÊ8çç”ææÛÙqyº(¨-> stream xœ]O»ƒ0 Üóþƒð`@,téЪjûÁqPœ(„¡_ C‡³t¾;ù,‡ëåÊ6‚|‡/Š`,ë@‹[Œ4YeÚb> stream xœ­V{P[e¿!4Ü"¢­Æ1¶Þ›.´¬Ú–ºVÛÑê«Öú¢ÚâB!„”@BÈy'7'òæ@(¥/¨ÚÒ‡µ¶]wëLÕV·î®îìØ]uÖýë»ìíÎì¨ã>ÆqÖýëæÎ|ß9¿×9,3ãp8YEÅ¥ W¥Ò‹8ôâ úQ.0³7f_÷(vßú]E®+Ê….ädŽ,¾gd!Š-@ÚûPõý—Ñ©¬E2¹N!×*…˪ „…kÖ<ý¸pÕÊ•k„¿¨)$U• ÂâJe­¨¾RɾH…;dU‘R'\öl­R)_»b…F£Y^Yß´\¦¯+x\¨‘(k…ÛEM"…ZT-|IÖ –TÖ‹„s….Ÿ{Éêå*¥H!,–U‹ †-loP4)Õ•Új‘¸V²SZ¿ö`Û°|lVŠíÄvcë±2löö ¶{[c%Ø"ì,á”g¼˜qž»Ž;žùDæÌ¼Þž5롬I|þéü²ìç²Õ´3—ž6Åéqs®ÿ¹Ñßß=øàSP¦/þ%“£ßKš3¨ñ}‡ê'_È Ž6—³ÅA˜%ò€·X‰È¡`œL¾;vÞÄÏ+Ϭ+Þ/~í ᜔-%4šÔÚšÖ-@á–v*q»~¢ërïå1À“ÐЦu¨r=3é0€ ‚¶`[,Žõøˆ\ÚÉìJ¡s…¾rûGæA¾Ejµ‰7š@k PídRž!Q*gs–ø£‹Ø@§j„®;06“aÅ`í4àˆƒ²Ðr”÷IùÛ/mÞ½Éj'ÍgÄý’T¡ÚÑB¾¡ç]ùè"àG{kví0 4•dí†2)àŶ?»½n¯'HæÎ>)t-…ŽâÌ¢V¾/pãa8íNp: вhìÆRæ‘Û™À™\^ƒËñÇ×I¾ƒÍû„÷›`³Lg•Q(ÊNÜ_Ôíëðo …ÛçöƒO>¥Âà$4ÏëšÖþ ï*r‡ŽÜ£ @ x('r Ä { vÊvÜ‚ö(xc^bm¶Xð "ÐwwóôfК‚Ð},&'P,!Ì¢»ãæZ›¥º¯?’t5/n·¿ Ï¥WšúÐÍw;Yb nrQ;:Á÷F'Ñ}¿‡v†~C‹CÛ \ÝméŠ uy9 +©v׿Ë]òÿY½ÝiYüÔ¿òúù¯ÎÅ9ï}€6áÒ éL¾ØAÕƒ7 ñDOllªêø&&§6¯¢¦W2=ͽö×Mƒö™6(ÅåMO2â²Îi#qÚàÁûtÁVKµ£´äRí;Hx q/|üb÷¾!÷‹ƒ0‰'û¯ßžH©ëD.ú€uÎâZç ì‹\¢óù¿ãõ„YÝYAO>ËG…n›¥]²vifk—j¦pÞjÞ·@Þæ1v ´±•U^üâ,ZÖž6—ÑEíD›´ºàUÖ\†Àá._Â×EæÒ»æìŠ–_ŠC÷~Ä¥µô½üM‡N«Ð*Í;ÛaBî­f¼±œ‰2ó÷N¿9v'q…´w@o:¾w¥„Yj3§eàövúˆàHò/oØ›¦Fõ6²Šá:Ä Ãón6ŸDk§Ð?’i~}3v•KËè…|o¼,`K‡®ÎZ „Ë¥–Då `8Ló“¿jzóõ‰‰ääq2R2iJÀ ˆ§:R=G¢—À‹Ï%ˆKi#4Ïhke€«ì}½sMîˆñm[Vo|õAôì(Êî'“W'g^ü­À*«ËijŠ®Èô~4Ι¹….ÜäÒÕôc|•­Ušo7]i(“·ÆÃÒ#e€/-cštï·§»LÜ<}z ðD¢yŸÒTe‘’v•„YDYŒÅÎ×À*Ðu¦‚ˆâ½†°¾Q£’WÒOŒì'‚MÉ×Þü‹$ˆÕPz;©~yé^À+mS±¾àhô éïO†{oÅo½Åº¢bü8‘ddÑS³OòÝ~Å4du7»–š6¸Ì.Ðá¬P:Ï‚Ùd§,.±•1F©ŠsGû°·ƒ,íå[å[žú¯0_ñ/w’Nü¯@FŠ8zŽ%žw ÙǸHZøGÐV@ 'kŽo]ÂyÌ‹ùçŠþ€îÿèË΀%hsº\f ¹«èÙÀ7š?=Ÿ ýqþ8zò‹p ðñ„L"dÖB]šE¦"¹ñeœÎc¯ÈøOm)[ÕzŠpñL ÅX9à?ap&É_qvÛ¯ÏONdóQ*z~-Œ ÆH(‡N¼ÃÜ­Ô(ôu¢Í·þöþÛÆ Úη,Y·ðÔ­èé4¯û89È6|ŠÕUð)å²Sfò_˜ùð«¡8Ãrãñ»¿‡eb¹aòïÞv¶º(p ZCm=ÃþdÀC°aìO\@|‚=P7UÖHdlòœŸ«xƒ˜9{Ž‹nÌrù½†ˆ^ר,S±ÊÄ’³ùn€oQA+˜ñm'Ô‡?ý÷‹ ·³´­} Ùy4v†¼ŒV‡Ùi̪^qYfÿNFŒ&ï°ÑZc£MoÕ±î :5ûHŠƒö_çΊèßò»¯ôóྈ»'ÂÞBÊAJÉ`+È<®ôD¶Akºä1WÜõ`¦Ú(Û*æï70YK—ä¾½lüã„oÔßOþ•Î ùƒWAÐbãÔ v²j(1Ô€È#Nfa3 wpšmà¤ldÁ]ÂRZh0Ù·7þ Ý‹2Ð4?ÅÂå5Ågáܹ8Gå÷Ûìk*ÞuSf6hZ»Æj§Æâ‰áN"v¸ómÖSa ØV°±ªt:ZœæÖ½ú-’ï:+wvb.+gyiýÐöôR5§ÿÃRU&.¶ëûû`않Ë~Ì‚íÄ„'|ýÌ?ØÞïïkﻈæßIWº§YÃþEGndzX$Øu@` ¶~³“¡óð92}ÎA™h¦óûZbŠ:Y½LÛ¡:> stream xœå[[o·~_ôG, ˜ª)ïä1à6ƒv€8J! ɃºZIÛìEÑ®b;EúÛ{o‡äŽei6)Š<øËÑÙ3ò㹑óÔõ|Êð¿øï|=ùÃ+i§W» ›^M~˜pÿ×iüg¾žþé4Jú |zz9 ?åSǧVÛ~zzºžtÂÍNÿÊZUÊÎöf°ðƒÓ‹É7Ýóë㊛n…3i\·˜°^NjÓ½1n]÷5†Á9Ó-QÃnРp"¥€ÿ±ÝßfNôð3 X¯‡ÁZÑÝzUÆ”ëv^Êœ€ßç‡m7h@ö†ÙN̾;ýëä³ÓÉ— †¦¯a>Ÿ8!z!§FiC™®IàLÏÌt5ù D`P:”©žYTÒªçlȯÕÈÖÍ­î]­ÕÊ@K°¡·¦ÖÌöÖÃ"I9®0Tœ"£Ê±‡_U–²Rkµî!ÃÐF˜’ –õÊX&Mïx¤ƒ)Õá=m&Ãnv2ÈØ ‘ï‡Aón¥?%¸"xžtm¶R€Üf²û¶›˜Äïž X{ñsPÖø8×½ xFð%ÁWÞ0ŒýÃX5&œEWÐ3=¿SVjGythÕ‹õn©3#±U¸^:åRpM†éN6™¿Úeþj—ùk¾Ž… Ú¸"í`B½À9ŒõV-*øàœà5Ð@Â# ÏÂý.¹g„wž·ïÃß]tJ<9¥‡lB/]‘ô-Yã”,D–œŠkU­_VjWôÑœ\o!·“ºç[y÷²É³°ÈèuHü°a,»'y‰¡ãÝÉË{ãP´p˜¼àóŸ$ÿ¬Ò†dªM'¢W8#i“ND…«Ê3•éDtR~É%îXë—‹öA+Á[‚k‚wW)‚«ÊŽ8SnKÙV10+µœ|4¹v«©Ö²çFe¦ý}=³õ“… Ê?0’aÌ-Òî.¶ÿnÔ,0.XFr‘9-Mέž|I0%äŠgNK“9 ðŠàŠà»rÃ~!N€¡ýqAðßÔz·ù”üÔòSJ‘;> ´¦¢ 'J]#’RM'û!<*T¥>üü|ì c‹AÅTƒÊJí0‹”_si{^¾^üUe)+µ¶]ŠÊáëq‹½†P¹Ž•®Pï; Uí kg}íá¾® b"¾ i¡pMpKð"䄃Ù5CÞ5žÜïž|FÒ+’.Hº'éK’n ^D©óÏDðØQº&é9I÷$½%é’¤oÐð”V5qH SŸs_¿Ë¯nžÜ¼#xKpÑT:rð±-¥0Ú'ÕÞ[©ß>ÁÁUòa·p•·?Ø…á–äyEÿ‰Æû g† 7'›L ‹‘0ƒíu¹)ã¶©¶RVj7×ãà <¦Á"D)%s•vÁ©Méáž–$Ýܤ”!$H\XŸQ®)o)_\„ÙLø\×Çsægó‚ÖöÜÃÃÞ¾’òA ._¤œâ—ÈUŒ/ýc()âÌ5ÁåÇO©–<¥¾"ø5½ÁgͰ1œU‡)Y,+½r¤MI%RjÉutÐQ°ô’îÑBLYëzUð;  TVj‡yL"ïþžCù½¦|gKð¢Ê‚RÁ„ëÎüÂôy|r5'ü«ïŒ®Zè¾ é‚àRØ“ôeíØQ*EVxM [’^P K»ÍýŸí¶±½`°æí<ÅLM»¬TÑE;£X€Ç.ÐnI™&±Zä¾7—RNï¶o!fú·¤š‚#êuþžà-Á‚ð;ú™7áƒ<¥ 1»ˆ|¼ „`<»xFÒK’Þê®Fu·³*‰pý>c生øxúŽvYH$Š¥ÈJŠTd1KEж¡pSYi_j®û=ã±a¡Ëüã)IÙ½x¥a}¤)¨iYQ5+µä=:™V¥húg|¬W³é€ÐE2}ÈÐ- ±’Ò2»¢2)Öƒ_¹›ÜÂJ} ík±(<'á®þ;çÞ¤†§V¹é iÈ|´Ñ]´¿/¨Ën/˜ˆ^Í/ öyBúòaÔ)<‹B6(>ô®t\qå+6d¥–GÇKXlmÜ1ñòà $c½(ùG[½AVjßé˜ó ŽÇ· ‹!%ö¯ðs\]§G'™;ÖÛr’ãpª!f¥vÐE%¬˜á=/_6þª²”•ZÛG~iܼxôÌâËþfÊÜ‹2÷–`QæŽGŸu!|J°ˆE™{3ª°-”7÷!¢àËåFêHÅ›‹¾wU¼Å)ð#*^[ôT_,¥M§†{*ÞX¯ªˆSE@ëpSñÆ'Ÿ…2Wº^J‘ÜzÓ¾+*Þ¢}W¼üW¼,õåŽñ»CW;&ëÔ[Hí$°ÁB‰á/ÁŒµŽ Œi£xq2ó[®‰•òÉäÆÏ7üØø3Ý¡û¬c†ìjÉ¢€lËÊ)ñ_¦i•âŠT‹”uše;º’”ì Æ+I þ®/G_ )é´c<6¡—0ÿNÚÑ„>„¥g䔿;/¶!éGáØ½(3m3úžã|‚Íž“­]}^]–~‘3 ©"ÕI…‰Ïa&ýwSŠÃ¤¨{p7…ÒçCz9¹"¸#¸HÕ¹|¤¿<à·ÆKi%›s*2e†^G§(œ˜ƒç)ŠT0UÅÕ°4˜j€I§q‘ H©$^öZ7†+;Y§±|¼·•Àb˃·=ÿõõ2ÙFÃ/ž‘G_“t7êò$½I5¨ý5rŠñ.ú/šSÄÇ=è ð¿I¡zQÜa DnÈ•jºßÏîûzçR`{ÊüOôÎc[î5õ×—¯^ÜÜ5ö)óQ‚´»©"×6£ç8…tO ž6’1ïhˆ©Í˜;åLAþWP%Ò¢¢JVjÉs|Fs©ëxF#`[Uw—ÓÊA‘R;Ì£.¹ ’ g‘Àsêàls[gABÌt|·ù !.”O8æt Å“6bÐàYË»¿„üT0ÖË”¥òtû“‚ÛY{Ý&v•Ò5¡Îÿ˜^\ÛÐòωÛY™P„^ª­®T7&‹ër£í¥gô³7á [;å’Ú“¼¾s’öòM…Kåð´ú#²RË‘c»?b`X\å+'‡·GŠ–PÑüÙ¥P㪶ö(•ÛÚ3ºÁòIÖ×jñϼv”Ñ}W”šˆI¹fÝËõp×\gª8P\™ô&Ë%ƒãYU«Lu×:­`µªY©]磳0a°"\ô—c5¯”½d&·F×­¸mTtxûºèqáJq{QùPþš¤Ël¶8ï¸nO9^ý¬¨öÍ†Žø`÷{ŸPT—!Š<¥³®9µ‚·T£\Ög]ÕYRHÒ"ybMˆd q0Þºl¡¿f´#é‚àž,Cÿ‚Æð5 ýEièÄTW÷ ë[–¤TSõˆ[–8’ƒEÓ U_á½ûÐÜÖ¿¬ßÇ虤ço‚‹“ÕIŸ¼$…s’v—£Þ·ïýʬ^tSUµ?<'r¶¡)¶%CÀËÆæI‹Ã¶g$=8ÏóÒK‚Å  Mk·è¾cðð ûÚ*lFÜt_‘¸±pψÐÌT$>VÍJ-k ‰Ê[„Ä/}gЦ8•‡B·´$_U“9¶ùØéÙƒÜ, ôŠç¹#ûYB.oâ§y©æ*+µ³÷øýÌy¯@$”èÓe‹¦ ŽûŒ¤GÔT…ÞÙ¹Ÿ‹ Új §×Ù\-9›«%çí§ì?Ùÿ8KëtäpPÞú0Š´x=šyŸ{æ¯Iº'¸£äæ›Ú…,óï(rO¬Êaùžñy§ÛÑtú†|ÖSj8<§ºí_På÷U8‚P /¯Å´ã3úýw³{>9‚2ÐVßt$&WìÎJ-ßÏnoIøO¤%W®âo ® ¦CߨˆŠÒ Áõ;6‹.žÿ4†ûÅ‹WO¨&BW€^ôI}*Mø`öàÙÕi¦ªÙËJí|Ÿ„rp¨á‹ŠN¾£h„Nd¸¿+*8ÔÓª8YºäTñ‹l&|¨ÉI_ÝáWÔù‹j`)s.¤9øí´ÂÍÍ?ž(p\ÌŸA…¹AsÇ£º1¥½g3y(ÖÊ/µÑ˜e ø³~µ½xs…Þ!Æ«ŸŠoă - Ÿ+¾öÞžÌ,>ßåoqd ¸wÁö~ÄV8&ªïÏ/Â#5ØžâíÚÿ\øÛ&ð:Æ?í+&œ¬Ílf3gþ‚4¾%,FøÝðQÞ„…ÿÑå„”ô“Ç´õ;ö<ÍÇ£d8¥Ã>–cè]³n9œYëQìÀ=;—÷À—“s endstream endobj 292 0 obj << /Filter /FlateDecode /Length 4517 >> stream xœÍko·±ŸÕ ¿A d•øÖ|?‚$@ÝÚA‹H}háôÃY'ËJ$#ÉIí_ï É]÷fu'ÙU‚¢È˜âç=C‡?î‹^î ü_ùïÑùÞçÚïŸ\í‰ý“½÷dúë~ùÏÑùþ£C˜a ŒôQD¹øb/*÷• ½vrß[ßGm÷Ï÷žu¢W1†àºãÑÛhUtÝ5€R/Oq²!ÚîFuðÒ¹ny°ÐZõ1Øî UÁºË4Wþuø÷=«z-‚Ú‡1YØ?\íuÿ=8ü~/­o¤2Éßžu¯—‘NÙ‹-²è½êäÕ|0Ý“ƒ {%”‡Å€´ ‘'œ²VäûËD¢Q æ$…v¡;:PYÝËŒÞFÙ}yŒŸé´ui²ŒB)×]¡˜‚ÓÆ´ß K ¿ŸAu ŒìÐò­bbÜÏ mûà\fùâ@yäYó•âsõ"ªîùÁ@t,º Îj”´20º~Q`éi”Ÿ×>t…‘h¼’QBØh@s€NH£»7(5E¼šˆ¦alt Z*Â+‚—ÂyŠM$›)=Éè`J·Nô[DÙ½GV”Œ†CŒÎú·Ú4³ÑŽEtJ«¤ä$%Ù}×Uóþ+ÎÓ‘Av_ŒnñA’tÝaþP7ã «Ê"]û—d=Q‚.¾;È8XØò:svEÝŒØbÁÙFÚ¤}$:¨÷ÉéZœï6Æd {FeM½ç´ •ׯ2 @L«*¦J©)‚ ÆÏšFéÚE]à.ÉN©»,êR÷ÖD•™(2‘eB­Y+Ýúy¢Ö ߨµ÷ìïVƒÏüt`-hR;dUíUMX_dÜdüªˆÛË |Z€E7Ö¿Nž)@RÄíd†ˆ$Ë EÜš²}wrLÈJHઙ.ÐÚÁ’Hx­Ë´6D 1ËËSÍÌØ[²åˆ²mÂãˆáyÅ»ÌÑ/€$~H.¢5ÝIxÔ­Ss¼§8‰r1Û†aƒÖ6N‡wàãŠØÑ“»’'@HUÞžpT³a‰Fš£ª…jÛIÀ.wEîÂ(v± ÿÒ ‘ a´k]2Çq1'aÆ(ò"ñ p#žì²óÑd(Tga >%˜äk£]%6€JQ"c7“\2zj»Ù军~ _¸º*½øåÞáGÏ0Ê!V¡b^VðÈìwÄFNRµ|ÖRÏòg¡("eÔ;:QÃZö³a…÷*»ÄwHùZ IÛv?ãp µ$õ´ÖŽ‹ oÍÛY¢‚†HªÞŸórJ¸í•ÅO$±V —¥ˆöI$Ì%Ϩ}ŽÂlб6l%qzZ6Q/Éõ‹2ÎL)ÝnÈ?0CI”ÒuÎÎRy Îùª¯[£á^•5Ròg¦ÙÓ2x’Éb0’¼u¯sR§æ%%Ö€+žÝOR=VB.Ç@P´[ ä.À=µµ2Sý°AaLì>?XX…4åJ Hb^¯ÙÑTOøƒ ç|Áb#Ÿ=ˆÀ0Ç-wTÁØ ' ´!{LFºÏ*øAð¢‚gô³ü¥‚ç(¼Ô©6"ãÔ^±†Eæ®é„ÑF?« ¨úl¹ÑôF«ÁtÉßeÁ)œM¯|ØnVù¸yU ã5ÉHÒöeßðñÌ vHÂ)¦)ŠÛÇôi”)Ò¡)äáSëš“/pÓèâõÿT¥ô— ®+HäAJž¾‚*"@MfýŸ¯‘ˆÿóI¯Ù,ŒŒsJSpjÒÉ\{ ¥Zäµ=ÜÅ1›%¶`ˆ,9<éEš|<Ù |9UêÔ[Iøþ¹‚/Ù¹M.€ªÝ9 ä[õÊKšwª@Y»“^#‹—ŒzúÙ–%9|DñÛ0h–Þ¨ã{wz}κåÜÉÓî šŒò šwO¹Ü]Å$62åôÖm;OúV“äí0Î ›ÄãR— ÷£‚/ÙÍ_’hLÂõ;·9ÛA²µ9ª[ð­Æ *ظÇîœÉß©Zã£@`ið¬1ðùÔm#]ß‚^ÞŽd„H¦Z»sà³÷þ¤#CÃvÐSsØ”?6Ác÷D+ØÑÆ«f"6cš]mkµ·5­ó6ÀÓÛ„‰!üßo H‰Ç§•—À»õr¾^'£’ÊŸ¯c%}› ïL|h¶[–PìjM¨Ù½ˆÐ,ÇDñM¨¹GàµMæ^Qd£âo“ Ë¥a%Ò„;ÙŸyøÍ¢ºEZàC˜¢«Ý)$ð;“ÆHîÑHýOÎ,ùý>ì¸nýosÒÃï øH.©ÈîT;ðÉ‚?Oq·8éÙß·n(yæùR ÐHðG¬µÃýò1?ÿ›†ÉÕá¼×C’ßj¿±ó¸ï 'c,¿³Ík¿‡Šü·"ìÎãï¿¢Ùé}ïÊßEêe\jQilÀA#ßÖÁS6>ó^O~¤ìk@%:ßús;QÃ'‰ˆÝî=|ª”o:$þNÊ­ûñÒýôçIÛ;í¥~¾)ªo¢÷½Äíö èmulcžf˜ãìRà°‹jÓ«¸•%çËÏŠIïÀ‡Ü…RÁ°+©€÷…w&NS&Xl亮—û õ}zGûäÿOz焪úàì¨ï³KÁ7²¿ÈãÅÓr=p&jÚn•…²._ÇO¿Þ{1ã/ÌF¡$æ¼_ ~Ëíh•ë–“[øŠŽ¡¹ì¹Î7MeH—}Ò#foÈ/Ig2óøpÜeb€rxÿr¾ªåj胒Î@ û þ¦·I.ã{¯´°©_ÉGH¿RZz{Ö­ÓÕáÔƒ°Ùi¥ðåÂmºË|R4bš-÷ŸÅ¨Áè‚‹ï>/Œ–½Œ©Íj¸›~VoÍ÷M›U‘Îm%¢tèƒo¥õ»¼2ºG6É:Œ15CÝ•rÔ¥šP~­* Ý-W´Á`è@ m+œkà_ó—Áñï¦ÜÀ-£¤û‚ô¬6ÔŸmÈ•f²ÆºmÛŠ55Þ«{g7§×¦.ê©gîJ ­àxÙ§U ”¾Ñ‡1MÔnŠu£mïFnð.õÕÔ~‹riÞ¸ L‰-ùÃ| +€Õ*=¹”Z "¶ðØnýGrk¼ M£Ò®óôÛŽ®4vƒzJ Š*Îv8àm~àÜG£Ø: ¡¦-0Š÷é—µy©mpËtc-2Ýth-Ïê=÷Tù•ÎÒ¢Èã›^ñÇmRÓžzá>Y>ìkŒ£–_®øG?Óä›D¤Um×^irL!?‹~FözÞ%Žf;öªT3:L^¤ çv¤Ô޳ï±ê`j ¦Ý{ …‡iD6 kdêTH$IÕtÇq$Ïtñ-Úü³Èiû$·q¸IƒHnÚrÀÕà´ÃÂëdŒ…Q  M¿M«G÷k{HIÑ*ÏHØ^8ºíèhÈ-U ’Цí+ÔÛÞ`«/È̦ƒCe¨Ü§T,þeóeÐ.õ«F0l[Ïò2èAŒM¬/½ÅmëéØBL75êä–ãI¤ÏxÍŽíIIˆ7µÁùÜ­¶%¾BrOœéáÿƒu“ø@;Ÿ Í|Ü2§2Й0íÈMÄÛIkpVˆmµÚE•npÍtŽtMk{H ô®"é Å$b ¡M5Œ@©RæÀ6´Ö~•"½³¶1ë™P™ú¥„MA.€x~éþK-Âh8ìroÜ`;kpQ,Z°,žÖ^Ÿú¦{Ibò ËÇ“:¢ÍÈë"1Ös:´+FÌK+©q6ýG^ÉÙ¦ˆµ½Ò& +ýaÚ5)à`ÏI6-œ ìv9€âÆýýÛHbó˜æÂí¬X08Áæµàéxê„uN7É–Ô…WÙБ[gF¿¯F3ÓÕG#çLs+©SVÅð$z@JÂ&µ¡ø›&ú—Õ6F3Ö´‰uyÔБ”˜µ‘þ–×DŽi,~æ@ëË|ÝX ¨£|œ\­ÿ:S‹·üOÉÆ3;&òÙ¿Ô iuªÞH76‰Ó)üEuÓQJJÒ§ø¼bx]Á$ª±Ùã°ó}5¬œšÍ ¶=bZ H¢£¦ú[ÔG+šv¡1xÜ)†ÑkþÙFÀ S·F÷Ø-÷ ù² iMŽ,²…Eç*™ëô¨D4ø›¾s€5?‰ëm)XÖ³ø ¨‰ÜÞÕY²Ùè\ååLº$°K‘ÊÙÑú0{𿟦’tQK9Z ìðŒ‘̘Íâùx$ÁïñÇ ­“qoŒ 5“=l)£g ™­å`y.Êa¥CÝ'¹á!Ëæp9ÓH9lr£”J‚Ÿì±‹©húŠ™pòº ¸}ñ¦IÔïàðßDôû¨ÔÖæÃõGÿV@vPŽ"|–RÄ%ܶüy£ØP"}гoLkâ¿SJ28™mÜú}â³ôyµüNO »x.ˆö…çhÊÇ´™7*‘3ÈC9£±µq&!ƒ Þ:žÎz©£ðÉá|>0'þ6{f‹è"Ÿ‰¡&òCs^6Ò;·gåÞíãÑ&?Y>àMäÜ4‹ù 0!5ÙfÏEòi¦8´©~]Ø‹M >­EæEö^‰Õè½ô í*cȧé Ġ™§ÙÊÊK¬L“ö ž6Ïï\DÛìdX ™‰WÜù]ÖGª¨±‡!iA>Ù-%?þfïœNendstream endobj 293 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1à Üy…@B+u‰XÒ%C«ªíˆ1C2ô÷ „tèp–Îw'Ÿy?\gðGôø¢Æ:iñkD‚‘&ëX+@[L••‰³ Œ÷7ÞŸ@°Èìü®fâÏ“8—U»‡ÐkZ‚BŠÊM妑1’‘ÓR Œ¦:ÅEV,þCÉÑ\⸠¸ÆH.•¦¥I.`ýž >äl`_>ÕSgendstream endobj 294 0 obj << /Filter /FlateDecode /Length 3427 >> stream xœí[mo$5þ>¿bÄz¸¦ýn#‚€=qZN¼D'¤¡Ù$“Í1/K2! Çqýªl·«ºÇIH`ï´âÃ>xªírÕS/îv¾Ÿv­˜vø_þ÷h=yû 妧“nz:ù~"â¯ÓüÏÑzúá!H #m肘.'éQ1•Ò·ÊŠ©3® ÊLד'͇³y×ʼ·ÍɬkM02ØfPtë<‡g(삦ÙÀ¨òNXÛ,fs¥d¼i>†Qé­6¾9²]§ý7‡›ÙªÎË)Œi“ùéáñ¤ùÏì📸¾Ôd¿=i.a.-¬4›£fÁ9ÙŸ(°Ûô ¼ùñÄKÙJ5uJ€±ÌtMÚ¶RNW“/atVÇ4þBFƒ{B‰R£±õÄgZ?””ìBëìPJv®uŽ©E#\¯¤ª²õ\õôÐ`¢^f<3 ]Ëh; m°ÒrB{ 4°XÌ[›½žÍƒk;¡|ó#Á‡jÓl ¯‚iå¨úûšàÁÁÅp.ð`Ètó.@£À±ù ‰Þ[§sÀ_ÍÌ…¦qah®"40`–º“½‚o½7&b ‘ìµ½QC# âæA 9\ÜM="¸#xF²[…ÇÀm×Ùæë1ÒÛÅ5Jè(œŸ;§Ñ5ÁK‚+’C;‡‹„æ Ò°†i~h`/^5b6·~w¡ùz†Ã$dó¯=£Û`Tk3z607: Ýp“Ñ!ä÷î ‹Y'aJßIÚ( ÿ‘“4¤Óù”!ž4?Á¬†­‰Ì!H-@¡¹Ó0 ä¾ =ä¹¹‹;×ÍšF}24$²7`ب˜ üª¬ðˆ* ïãSŠ,~°oE.Œº½ÅV,Bc»Þ™ºÂÂóÖÔc]¦î†v°(;XÓà Ù°ÆÆ}é­Qr…¤\!c®À¬icK£‚G™›ìŒàŽà‚æÝ zuŠÅeÍâNh(0ÌâÙº‹¡±îeq¯qJˆgm(Y 9ÂsŠÓ›9ÂÈH܃>¦}™WË2¯–eÞœ‰óè%ÁÁEŸm,Æÿžñ FÃ.˜ñ²¡Æ+BcsÞÝxB@«tµ°¿läV4ï\Öxä‡u©‚D‹ +Háó’L~9H¿±¨VF¯†£Q˦ÍAiz [c¦†®­’¥5ÐNs¶gç V„Æ.¼»Ã¤j!AÀ¢}o|F)tÙ—5(lŒîg/Æ0º lžÒdqâœqÑžÁ ‘ËYªr?S•ûMµËzäÅù¡íS¹+‰L~IpEîY‘,i!¹WQ‹Z½*Þ~—fž'ñ÷ó\§r:Îð9îZ”>ÁÆ,°ˆ ‰íCb‘1££ž±›ŠSxTBÛ¸§ OI`G£ÏÒ}1ì÷ôboOÑpµ¤¡ ˜D1F¾ù!‹Ð•þ7t *´ÂÇFÁWƒ)E”Faym£Ðň^‰{rl‰vu^OÇ£}ž.¹·!«/É- rËÁ liô|ä7"^äìh ²§ÏQ ®þ˜ëéuMBËM¤àUxkA@âıL¹hÁ^ &ã€Ê÷£` 8…Æl]©ÛRá1…±(3fÀ¢"4æÕ3™‚´á®ëàp–3YßëÀóþŠF\Às½ ãØÄº}@Iè òÏdr ±kðž¡ÔÏÃ-e`¨"46ݽR>¬SšVºd§JWý‚>Æþ:e¢Q-6>ReAu»§#TsW 3ûD›rüqð­!÷ù*‡|µ]Œý%×Å ÏVàŠ ,!á4Ëñj¾¬Uó5•‹Jõ ‚VDöÝ‹ïFZÉã »’»—„Æ¿—{ ônl¥ ŽþM EÄ s¯ûŸß¦ ɪ¬ë½"øå°Û?Ô^ŒªüHK:ؤ„q©×˜Ñ0ÇO@D ÙÅÈW‰P‘½/ŽÉCç·äÍç${@‰óñð‚ ~Ið#zì›}Ç{ÐÚ±¸îì,¦qç÷?¹4œ +J†åýÏïw’°ûg‡ñ‰"#b’Ú?I«| ]].BCß»ûŸ$ œŸµ¯O¯OScLå3ß,BcVÞ»öƒçc]Þݼ>(¼Â£Bh OT™’¡1mî}P0à˜ösP0ŠhàÑ”20T›îÞ]›VôõIá%Ÿhªy d_ü[„Æ¿odezñ"ªo^<šê[³?ÌY"~2㺻gõê€çQ•-<°zûáþm:Ìcª·é:¹”`JEì@)ªiﯔ¦uùÜO{Já«ygI©^REh¬æ}>Ûê8#4–ö:C9ü\ÃÞ—÷ëóKS°÷ìÒ@?Â/ TA€KÇèÒÀ@ªßsQkl—[¢´r}Å‹©†VQz™2ð#¼¾apé&G<$¤ÓK¼ˆa!7Aó’®…x¯b"ÒJá}8w½‘ cõkþ "]è¬r‚ß")7C~˜á{´Nª&ß  ùyšÍÅOWå"Ì6ÛXA BI-wO¶›¤œìDNh»fwpØ=vñf±Á‹& Óìg–mt—á ‹x©%(á·ZOi¹U–¶³dŠ˜VÊÊŽ¤gƒcºyïÀ¬ÒÊ^ÌW¨&(‘o¾¤Eâ•>7d˸šFñ =,Ñ«sÄVKVQV66Y ›í°Ä›Gº“^Évt”†2™,%öt…fÆá9ž‚¯Øs ·…àgá8ÏØÈ ¯æÚ¨¤óE¼Ðd¡ŒkYt×kN–Å…çýʯ¡5OËg{ˆtÁ]Á©PgÓ)©U3f†¿Æö×tÊù+Q+¾hž¦Í¤>Ñk‹Yt°Õײ–èfq‰èá<¸}ušfCÖ^?Y4Ày’tBÞäwÿ3³VöµÁøÝ^d³˜ ó,on ×)J3êo‘,Ðiœ»rlû”(ÅcæŒbfES³UêWãØÅ³‹´´‚lÉ–N‰Ù4$Í"Þ'³B9Œ*|ÒÏ×Î,†@¿žÅ?àC"/Ê+ÚôvÙóB1j×=1ˆÀxÇ/ák#3ÑÄá…š{‘‚Ï9»bKquq ½ï·'U$wŸÊ´ ¢O8ô ê+ðÞßÀ Dß–ÄÓ<ÜÁƒR$>ž¾õ¤ùf½4´ÏïÍæš|ìµÆQ‡Õ–à†àQU`MðŒàŽà¢:Ù®_ØÂ‰¡ŒÎiô>&øÁÏ"4Þ÷:DxIpEpGðŒà†à¶:t™ÿÒ«†‚2ú” §Î2z7È›mý)mãG‚¯˜Öc%ã6VÆLyt›UϪ~YT'Ûývg˜›]Áv)o³È’FWO¾¸Í+L ï'â–Öø…à ÚÏÚðqÕfKÙ‚'9çOžVþ®:¯¬Ê®«$8®®¶ªÎË¢—æâ&jx!ûЊpKð˜à ÁÁo nrÃ5ÈÚ¾Ô”½™ßa…稧7có~åa¡ø ¾AÃ;>òš4ð?ÈoÊU/7 °4‹¶ù•±­^ÇöëØ~ÕbÛ„ëc›Ñð’àE5tOÏ«ñú¼¯¿b²ÓÇn?ßóÔ8Ή²AÆ„ª§Š——Kj”ü”^‰f'e9|ÿ®º[wæDÆ“u²Ï\ËzPž¹­I$Õâ+;vÆ@ý’~ ‚«1ŒÓ¼Së–—Õ§NÞ7 ³ì¾­š`]5âŠ[®V–äɆK¤ý¨ïXW-w\Í«ªVUGÕS÷Ù§wZ¹læÛjl=¨†ú¯¨_µyŸÿ!RÚKH^±”-nîRe•2Vÿ×~`ü3GÕùÖz(zÒµ]è‹«H{Uf?P ›.ª²?qVܵèÅ ¾G°#žWM_°M%^ÇO…ã÷<,5n ® 2][‚‚6ÓU4Á@PÅ‹tñm%µÔfR[ô´ñŽF[‚–`ࣸñQ¸Ë.åm»4™€Ã¯¼?çýŒômv·mØ4|ï©JåF¯ÙºÎ$Í›³7÷J¤iñcKy{¹ªë”lÁÞ{1ªÇ-TÕC›Ö©.ä |³|ÏåŸ–ð†  ½*sθº%k°å¾`Þäè7ûC–k—2oéA®fe뤣&ú²á'#ß!edu¿¢ºz°1è)Ø>¤ÑO²³òžãí²c<­‡ ›ÁÌê/~ñþFpÿÿþQ1‰×ý㟳Œ_å?ŸüàA™Üendstream endobj 295 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1ƒ0 Üó ÿ @¥vA,tahUµý@p”' aèïKBèÐá,ïN>Ë~¸l#ÈGpø¢Ʋ´¸5 ÁH“eQ7 -ÆÂòÄYy!û›òï'Ø dv~W3Éç©>çU½‡ÐiZ¼B Š'mUu­1 ÖR Œ¦8›KWÐ`öJЦÇMÀ5☛æ&©€eú=ãO)Ø ¾?rShendstream endobj 296 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1à Üy…@B‡.Q–tÉЪjûbLÄPƒúûB:t8K结Ïr/#»ò=>)ul"-~H0ÑìX´ ŒÃTY™øÖAÈáªÃë6Ùßô›ä㤚²j÷zCKÐHQóL¢kš¾³¶ÄæOªÉV§:÷ ‹ÿPr4—8n®1§Ò´4ÉÓï™àCNÁñ<]Scendstream endobj 297 0 obj << /Filter /FlateDecode /Length 6721 >> stream xœ½\I“7r¾Óþ>vøà¨Á.vÀ>QI£*¬¡h;bÔ><õFZÍ~»9"¥ƒÿºsA TU?R9xà‹jHäòå‚úñdÕÉ„ÿòÿç¯}òÌ„“ë»GÓÉõ£)úëIþïüÕɧÏa„³ðdLSR'ϯñ«êDë8¯N‚ c2îäù«Gß ŸîN§Q§£.wÓè’ÓÉ÷ðSMS˜¢üù‡“nᩉAy?ìw§Æè1E7| OuôÖÅá &ÿûùŸ9=š)êxf#LOž_<þw÷üÑúVY Süí»á-Ìe•×.Áb§HY AyµíðÅ.šQO:Àb@Í´ q7~òȺï¿!­Ö0~©Éø8œï´Ç­ÇáOï’ž^âk&Eã< ViÒÚwȦèµí{ó¸ßOžÙIž¶j„­À9ྡ[Øó©qcôž·üN¼V6âžµ‹0¥™OŸ¾ ¦"ìØêU=Œ·b«ô0ÖøÌLz¯ðJL D;IMaØ_Ó1'¼w8ÈN“K6“äÍdZ6–qÀ”€3i8ÀÓ0N ¦¸!.z˜àއZªïÅâ{Á\Éþ F¯Üp(œQÍ^ÊXIçþJAŠÚêáµX‡‰ (QÒ[EîI#ˆãÀSxåfÞ¡×FçS¡÷æÍjÕ0õn挕âóBîü¦Òqû* s`>]=.+I~µCq3¤>dñúnk§’zÚTL:5Oow°˜UÓ|~цå1D8†ˆ'ÅôiH¤Ç§Zîÿ@sD7YmøP$çâä¼·ùc40ãëV‡Nç]*3:›4oŽßÊeÒá¤'Ãlõ*øà¸©Ò,D¨±^¨À82zg†ý-?ôÊ׬î@cþ†<+cp9šƒ‡ Yû›:T.¼ÿž”ÒÄ) âhÅë‚w·0:ÚŸ®_ ©çIdÚ÷íl0Ô*mäÐM­q’[(Cá¨Ná@g ôÀá˱…Èž8*3ücqϳE޳GGXÙ<×ÏŸ?úË#öjøfÃÉ›mWÖšÑÙ•òŒ6¤ØÂè4{²MßÒ£@”½ï!©ØÓ1M:‚Ê_ÔÇOÉoFP›æäA’ŸðÊTyä–Ÿô À`ÈÇ2Á:3Z×2álWôvÜFÜÂð„ˆ7QÅómÝŠ¾ó…}3 ø#·u¸íµ4h"ƒµ4fþD"4M:2A³±ú¦âs-ñÄ(4rAêãÅý†TÁò÷ÛuŸ•Eæu³9S^ê¾PÆ÷¼pœT£ÆbÄaG (7tÇïYå; ç“ròi뙊ºÇÁ05,}S1 ó^kNEN±Ã/P·EÏß[wm±ÅlhƒŽ“n| A#eTjÍÍ¥pœw¼ŒE;Fdj8©¸¯ú ŒQ8q TäuÝÌžX‚¿hôÎ}ΞLÛ4{2†0p´1}Æp§Å«Ï„‚˜Hó{·ÂäŘak²à˯Ûãb¥i¹DLR³¹Wv&ßµJ±—Âx¾“SîžÖ½ï†;m½Û¹piYøT’î†×·:=LÀÐÚƒ1ÄÉë=óÕ›`%v.Ó½c ÆYØÒŒ¨ÒIê“ä:ƒkgLÊò;#Õ¢’3¾ö€ªøÎÀØÙùÆÞY?i¦êÿq嘈 üóªþ|YÞןà…êïÝn ¤[3zë‹ìý„ôÜ„Fî…q¸`梺TX.£ „^ƒ‘MŸ\äçM|@ôQ…›½–ÂЈÐG¡XÄD4ÕÈO1C~V‘CÚ˜?ë;ÈŸw[•HbÀŠøÇJ)É¿J U ä¸ 2ÅL¾î7pø§Å¾U¢Ù€»mÉ£Ñf&sl’àÈÎ’š&dÙKËÙ9?¤Ã¡1¼Âß] ?‹uØNè©;‡ÓÝü³AjõÎH“ÁCËwø›—¾a²á /ePÕú€ÞVRí{ ƒ=Å`Eæ6È;ˆ0Hžž$$½]³ŒGxuËn@^&w›æÃ³Æë‚H§Øqö´Ç™uXÈ {—F|, Ú7 n'ByãÂrðÌ6Z+µô-K;Ð ´® ¥–·e×wÌf22­dAºê¥Ÿ;é=l–¤£Ên@ìg/µVð5!“Nšàæ¼E¶}Æx)iNâ8׈7âÙÕ ðÞj ht¿ æ|Yµ¢ (¯hú,Œ®0†r7O†~Š ½ßÕµÅQŠXr+]…óZ9Š1•ƒ ϨýÔÅë4ØSü$m³Ô5–Uís¬?'•ÁþZM*tC»`íSëóEÖs&G°ø Õ³¥-ÀËœÂFèÌ®<¡52Yô2G ¾È¿Ì$¾-Þ´ ìF+7UÑX L‡ßù¡Çÿ°H¤/B•Ëê’å4É‘ÓyKmXðw× £,E6§‹ôº3>‡Ãþ>Ÿ†%<9*ĵÃV¼&•ãªÈ_›Ô£9P-¥Ø½áÇ¡³ÒìúHœ÷½S2 à[ì!2Ñ·í@ÚNAÀç6t.?Ä[*]”ã+L6¿KB-ÆqòæÄk3K8™ôR=£¨Z-" ™kþgÏ[Çüy›^ðsL²>ÃǬ˜û« õe6AT¤„6¾gV¯8ׂUÀØÅß’Kƒ=Ž)tûÿ·6bù;°9ª¯u9]÷Tf°%&‡ümÓd:cæð7íV?NF·´äÔ!¡¾ÇìÎQ QìL5Š.p @å•„’mfŽV†Ýv¾$¦DpÝ)b©÷M1©"¿ÆY¸Ùÿ€)NȈµúÕupÆð"ø>yJªð`ò”ŠC §Ã^8“/|}æ2†ÎÀ$’òœƒÙ× ‘¿éõõ r<ì_ÿ9Ò¤]`’ÑÕHŠÂSiã²S{W¹%áÞË{¼kÕ¨JGŠ oé“ðæN[ -iH„Á†DÚê±§ÒÄôÖk¤òT‹ÑÉÉ©Áêy4LîgÄC6ð䇯Iž›LÐ}¼%:òöd!µQÉ˯¾Ü‰Yêè[މ<t‚„€—Ù9ÅÎF0(#&Œ#Ù™äYnhrguV¼—ñkGqª,Öéa“`÷9<}¢:…È.-`\êfy'}>,«))‚-´ÝÚüœßàÊà2Rq}TPek_Ÿ^ðêÀõ‘ö•oh²@Ëeê>æ©[$òÙÕËSÈÔ6CoBÆ}æ.Ì)ûJku¢j¢(1âP«\Sù]ÅF™b?™Æ†87YǹH¤3j©ù–¦úÅ#6šªÌŠúìa­Vj„˜uœ7HaˆäK1¥€pJf,Ï e4ÍC`OApM“ôg;0ÍúBžòºMi£Tð”ö±ÆÙ?c’þ‡š•½–¾N §5;‹¥LÑ™U¯¨T˜4Jdy/´‹ °R@Ä;QYB55ÛÕÜ6ðØ1¯¢˜™TîšDebˆ»\|çÒhŠK@¹žLãƒÖËY¯×Åï²(ñªk÷GÅY^oª`ëOõx!‰D:ó$MÄY(}]ÃÉk±Ã6&%Gá¹]%ÛÚMq…ìà$¯ÃÊ †8&mz`6~諸ÍK¬R þ¬¶ŽèJ¬‘´*÷!ÄZ³’xÝS>+’¹÷µy}ùŒqÆ*åóÔ”Vø¬Y¦³q1ñíèpež•<"ÈEj"áYšrCöNɵJú)ØÑÛ2'‚[WvsT''? ²ŒPUÊrŸ¹Bo›”§ [d&–f¬ÙØ>[3‘“°/ZÊ‘>²Ì~h‘Ðl÷²¬* µ\í»£"‰žç{ˆ¿X–ÁÂOÊÈö²z3Aö*QÛŸŠE [*wŒ®–W}ÒÌG±`]£la¶dyÈ®CühÂÎCÍ ž0ÍŒšC.ò=!DŠxè|g`I€îà= 4n;XLhn»ÈИ€,H¨™ñPÜpq]Ÿ¿Ãå˜[Z'ÎêäGѰ§9ƒá ,PŠÆÌ&R`½Q;—™x© ¶£Ïwý@+_ðd^YYÊK`.:§ ´nûÞ"žë&eÀÀ•¤ll‚CóÔ{( ôø0¤­˜ùr&È ëå›yàù'}Þ¯´Ä)P‡žš¨'öÛõ¸[„5/š#8Å húb5ÔÃzg”DØûíÖLu°Ö}‘ë’z ÕÇüWõw_T w¶ãb [íÏ%¶ê" :´˜r\”ƒó=%bLÒ^ÀÝÎdžÎ¤ÉJÌ75ìŸc§€ŠÚGc)P.Oj…q€.O½áƪ }i’‚?ÔS»ã9±aåÒ’úöF|:a¥FvÛ­•°˜•"ëᯌM²Éƒ›4JIËv± ̆Ùï> a»ÆFC‡(Ï t¼‘gºÉÓŰ`9n`¾F!Dâc£;@FJ€—áóIP–° Ìu|,.ïu}ñ†3Ê*ÉXÁcKƒÚª`Ið¶e©œÞN4¹«ÓyµN5˜§&ZErn=VêAf‘¼ær—c[¤Évµcœûœ¼lÊk0W¦ÚÚŸ^ú :œ«ÆÆ–6ÓõÁ—s…rQ!ªáðš,þuÃÛ-¸Çé’L2\ψ2¸®Õ}®þXwý¶#~¥åPV÷zS RÇÊ÷5«rx[¨÷L\—Ó¶ óGè€Fj+B²µµ7£x¦í qv:#““s?vß~ZŠÿ«~i[]>_äò‚m¾ác›i¹J°¸­|nôWäæeGÏ¢uŠ¡vî5©<¶mNYÙj¸Ñû÷Ï A€ÿ†ô*ÌÙ4µÒZëÚÌå&ê¾"râVsÝmq••™_kAÔ˜ûÿÑ‚Ä©ŠºÇ.VêеwæR‡®çºZ˜J› ƒà+:L¾yLi… }{N: š®Òq¨ ð)xC×™tÁÊOrI+ÄÒÄ€=N?TÑÈè©5Î _çŸà2"!yI{*SAXá)eï –WaØÙ~a¸ ÐæK‹‹Zf!)(Âf6kf]€†Ú-Ë)ªÌ5îoÎÔ ¬öŠg@lñ”woÄGdx×€˜WâêlnQÿ:s‹üßóD–¦ôý®”—DKNhêó¼bªþÿD Ó™íNç;C†3àlŸíJöÐBÜèXú F,”3l‰F†¸´ˆ2‹l(œñæ±ÃK¡è ‘òq+%•!Ø÷ó®ô?¬ö7»æüûDAv…ÔÍ l \nO­ ˜ÍÆ ˜E «ØÊžÍK{¹WSŸîš–^£ŒQ–+PŒ¿Ð–ËcÀž²ýϲÕGjœV:¯ƒïØY¨ì#Ä Dz¦h‘Òr0¨›©\G¥ÿ2äV©‚sÐ&йçVA°¿(söE•$G䘴‡ˆ_ÖLé×Y1ª¾úÌÃ,¬—iMrª)»;¾óÇzvF¨é›]ôö챩8W¸ŸHË2K’ `è†_½cBáò£ÚÄågì̶M£†¨`s(ª¼Õ¹?0 Ê ŸïJè¸ÑÕºc^&¢QšÀ>ÐZS4 é¡VAÞåDâRê.®Œ7±¾˜‘ŠJ+¯,J‹+r?ïÄ‹$ƒÂ¾¤Z„`Ñ_÷t½·¸óƒ¤&%ÝÒÆUS‘ùfHéQ¢l)e<ÔX“’MªØcØh;Šìû·£Þ‹ßX•ŽÉ›ùžPiy(‡k*h’ÁR§­@Ö$E¥¹,ìÀ!*+-%æê:àSï©ÓÌidO.J¤Ë VJ¬vKîËXíòÙ-ÌÊdÆ©VÁŽ›,q§D“0Y¥â3ìJv-VÌàÇ)`0gÚXó‹æ¼KcÐ(ƒiTgÐäZp‹JÞ­-†:@„J­SÛÎøi¢xUá3qîù16iÉǵ uãšÆÛêßÍ3'ÖP>Ÿy⨥O?È™¹ +/´½©ÞéžgHà_¿äŸ>5¼Z'¶Ë94¾b€¼¤¹L¤P^ããn7tq6R1Ñ•!"-IkÈg„lßU¶ou‹ÄM…MÝdž0o7õPÊÔn4È‹ßÛ™ˆÌ¡ŸÄŠ"7Ü\áÓxÙ%¥áKüq¹éìcû)J¢¦&b–2›ï¸€‰r«ww 2Ä¡B®­ÛÍÅõ&“MйA€˜û×Gkß0Ø—±â  ‰þ.•€ù«ÝŠ.¶û9{*BÙŸc†ÑØõ|G0ôY0i«C¢Šu+ànW3yuÄy1²­Ç™æ{§‡­Ù€;âzí¤ 䌲âa²É¹EY-/r/¦»$2B¾Sº7x¶áÉNc«r  ˆO/'½Ü~/µIæC+j)¤û–Ux/ßâW3.x]@ûlYæ ™äzì+àîä›%n›­±{ɧÜFZ|ØÖš\¹À4iAgMÀŽQ›¢£úV»g:-¾Iz]Ý´BnÅtçäé!>ßðØÎ¬Wd­‹0éõØYACv!l „»*ý´lb"nFÄS!fg­ôïå¬_xåÇóc×9ëjbØj€ïJ¹kí»ÎqÄop×ßÙ[#޾wצw×Ê+‚ggÒE‘‘›‚™bcädV|ý þu¦,˜¨< ,¯µýbÒ5ïóê"ßRË]Û‰s$b%¿Cw‘}öÖªzëK‰)ZŽÒ(öõü¾š;=9ÃMÅbØPhóÃËîHšåU& ¸hÇ=ˆt|ñ¶ºxÛ#Éë®'«û Œz¢+ÛõDé£?Šû-×,®ûxC8ïrªJàSq ÜRžŒ•±íz÷Û‡8W$Íõr¡u± , Üòó6äˆ×‹ÒÍ—:DI*ÑãæPƒ­÷¢$þP‘·½ÏT{I³øû;)JÇnÉÍÔ4¶|'B'Po`Ùžs”Jc%a#´…Ã6µ}õ³_šÜX P Là™ùt|ÅûË8¯Ç ÁfÛ%똠DöÊ÷î÷Û Ÿû»âÿÂ?Lür–®¢RÁ– Õ9 ãg·í€ævT.[½|×JzNI4*p8 «ŽD§-@}žÞyõ‹ n–û¿x°ééqXd÷ Ð5|L¾~¶ˆ® •gß ¶`ˆ 6¾Š#*rè|Â% ÇE?Ó´ÌPf PéEªU–hðnw«+¦˜%šæäL2xîÜG)?"´%aó0÷5aõ:QÛ¸£0xšSç™ò Üðô ÂÆ­[^­{>5{`t›8Š.º"NÝÇ?]à=+º:øÄ²§TLJÈP>AÊVÌ÷snhKœ*¬_Ƨؤñm]Xv@•‡òŠŒ[e[¡¼%,ª›·sK›ñ†’óÒÊʈd£ÉDT.åÅßÚYkéF[Ó ÆLÂàõi^%€áù³2D}±„HòS¢_xµŒ8÷ÅÿQWkjLøq’$ËEþÐÂÙÙóõDÊ%Ï…“ø+…APï\öÄ©$Ù65ÃIÞÏùJÚª9Ò:{#Ž|/¿®$Û1»&P*Žm52ºnY¯œ5¡iÇç3n· cÑÙv3î·U ÕSú¢éBñÿÄ×ÉZ™?{õ~5¯QEïqm$XÂE·ð‡Ü…ÙZ,?FÛó²‡],ÂÚG e±¯|Ï bmØì𕌰´°þIòuå¹Nƒû>ùúÊ2ß;^0À=­©9zê1€åäÈs{Ï1÷O­6)ßÌ“¶S9'R¯A S·‘¨Ø¸:/àÑáú=ßQHÇÓ§óZWú' ß½µmhö¶¢É7Uæ·cr àno€1PxƒOc!‘(§óUd{Q‰ž!Ør¸ž$ã}5)³ÎëŽ|[ƒ³MÕã#Bµ[± ( ÷h=‡C¼¶“kVLøÅÀõ.nP–D¤QÞÏ|k×*nNa4µÙôªfÙ^à•-"Vï£4éô­º…Jî¿.›Aß ùË£ÿ5Ny.endstream endobj 298 0 obj << /Filter /FlateDecode /Length 6485 >> stream xœÅ\Ks7’¾kçG0ö0Q!ÖÞ€o”Ÿ£ 4^KcÇÆhm‘¢8ÙQ²­Ñaÿúæ$ÐUTs¬Ý Ü*¢ðÌüòËêïGÓ¨Ž&ü/ÿÿùÕ½ÿøÁ„£‹›{ÓÑŽ¿ßSô×£ü¿çWGžB ‹OÆ4%uôôÅ=~UEu\“qGO¯î Ænžþ ;Û4Žaô)À OÏîýuøf3zRVùá5þT“ñq8ßO£IÑ8?üUJ*Äá¶H)F?\b‹brÐàØ ÿÛ¨GxÍ@ÓèR Ao©é4Ù8ÜÐÓ)jx¿ ¶»ÆÌè§0èÍ?}¸·ÀãyÒÇÆÑ{žú—دž& ãp~‚…À*àiœ¦É§üÔ%í‡÷ø8O&lµ£R÷8è­‹Ã}|jÇ‹û žNiò&(9Q^SôÎ ?oŽu]4iؽ¦ Š0Îð.÷œìð»KózK.¦šÅgüwt´ÆÍ1<§AŸ=;á·aS‡pºSŠ6 ÛküÍ‹;Ûà±(oõ°»âŒ²ÃŸ°…š¦0ÅáŠv…é 571Âò§ÎXÓKÛí;Ú¬”¼ ´P•l˜4ÎÄ%5|/6î-í,¢9¿kÑâ9î¸2 /²ÅÑ‚v+‚ä´RöÅæfgb†§´–¸—¸‚h“¦Ö´^8í'¢?ZŽR1 Ñ¡LJ+˜ìôPÛnIþ¢7ÖÂ{(i&‚hj=kƒ˜óó:Ü×|ðOTÚ™úr¯­·ÃÅî˜3Yã5v|l|ÂÕ+ØGX€;h¦ \Þ8b8:†g<­Ø*§*Or¤ÆÆ;;üû¬"Z›ˆú3¯ã‡ÃEénáÅZŸBÄ7Ø ¢¤íÄS>_ÀÀSʇpúé<Ê `m}Ž]×y¼óÓx†D”ÄÏ€¥ù|£ÆiJ…É) Te­eé>Õ§Sý©ªÔ/† ˆ€}yKîW”’êòXhš|~22Àù øÝm´Çí±Ã/Ø”‚a›UÔÕåSKQ9<5ê6õÛÜ€j‹@Ig :‚®Ã:UÀ©dÒÈóÚ²°eÐ~ò¨BðÜN°&›áÜL0Kaþ@: âQ‰Š†oßѨ0¨5=üãLA%¾[ô(Ï_ÅNÑH…”s®!jR㯸 @ðB<†Nrï¸P<-ç½á“ãE¼¯‹¸‹à5`k»gâ<à¢'e#`ÌuEÝ Ðè ãêJpÆ(„²•â,ªOëŒ~ÅÓ n¹ƒ'-˜³ð¶àçQcj¹•ËØ+?Ú˜2XÅÈóÈ{ k‘`õœ­µÙÍæ^ki{sS˜””o –¤4tÊŒê Æù@–cÞ¦/ɺٔZ»"ãOÄ{_§…ë ŒÂcm!!ï² Ù®Ä5ë;±¥m}üšñûx"ZÈþDë~˜¬æ7B;öš$£²qÍ8™!¨ÀðGü`ix ‚+Yv¡?÷¦m'8…“ftX]TQ€ÕËY^ʃ.7Ðò›xyWI…,Iÿt£€¬üd|ïs‹8 ÛçËŽ»4kÛ³Ma#r/ÊÊÀ~ GÐY¸¸gáÌ„zâÀÂI\$Œ#ƒÒ–í®˜=I½¬TP† ÙLl(O—˜Òömn+ÛŸiuÀxúˆ3.É 8ÝîÛ{š^¶DÓêÐ}ÌBÁkJ·®ïM@T„³…÷|Vͧœæ:Ááë& ÀÈH¡‰•À\µŽª…FЇ° eR·¸›ƒÏ¾ÃFBÆ XLT#lýžK/Ý™ãâߟ³?9aäk£!»êY4ÎÛ*Ì]öЋÔäšÝMðÆg褙é'@ñ=÷;·àž½Ò³ÌïÑtnmê-ûäÙGNò¿¤+ÛC'Ùˆ[ÀóÐ(Á"¾Ðg«ÿ)g8gø`ìÌj›Á“/¼a`ça1{Þ0€fÒ·`äì4!å¿"gD9ÝÆúâ!ãÆ¶¾;œ|äØ$ÆN;EŠIžn™«FŒ¬ ÚpÆM<ˆáãGõñü¦L p(/Yà?‹É |»Î)X 9) ‘ê?5S²!ÀVÖv˜ÃÁX"Àõ³gªÛß6%n*7…|¨Y™+ÞþS‘‰y…GaÖ°Hô+ãž; J€Ì‡ÚæuJïªz »< zév²±Øó|vWÄYAMÓ!N¸àøÚ(Á÷iÀdX¶`?á¨g—ˆyÍvÀ–ÂÈÝÙ÷Ž…£PcÖGˆ›Ü–snaà¨ÚX3ê±VÄÃ+Xä¸3ÒÞ9ìLD© ;/ŒÃ;39'¢7kÔY#a w Î‡Ø@]ôw›|‘ŸÆ>*ƒS™šlø+5Ÿ#&ñ w0SÏÔ©ßÉ‘ã’Öæöا`ž îÞû\¦Ç c¯€yÁˆój‚…ñaÞµì€ÝÍͪ¹¶O°_Dp skSØoºÁãBÖJ}43HÊ‹< *•AZù}”{ Ø´tÈp»üµ¶jDá‘JÊKá¸qF…œC¤ÆÐ (p5|÷„[ãã.®é6ñÂŽPâÉãô‘œRÌ‚S"‚Û–ÓÈTYÆ6‹“4°"1Ia1¾È‹S l«ì„»…UŰÍí”Ç'o*„®@e;çC²fe c`S#L˶NkËúS¥«?ðôìó·"†s™3R€’¦0 AÔ_UZò`”—@Fì©—°Â—•/Ÿ¥ $U¨é#˜Æ«Ñî¹y2þ¸‘ š–B•aÓž™äCxl€5¡s´Œ"”"X)–œÂ5@F/$uaµ¡tÕËúâVÄtÞ gµM,ç‘ "ZÈp‘àP✉Ëeißc­}1El>ÉGÕŒ }[}ËW3öÿ‚xasÊ ÝFtÙ¶‡œû¹¦7@j‚s%™&Aÿ_D2ǶæðeöVD øuõ&w<â›ÔÔ<0µ “My§£™6Öéya×Â.,zçËØµ,­²'u¾åŒ,…螃|Ä`4‹FTMSæÞTµÏÎ:E¶¢‡ÕDy€qÛD,38àX(b˜ÅÅxñw[&š~@ÌD@ê/¨Ôp„¡õý¸"ã´GZý.¶)6AÏ'µWaÿ¾Û²^û ŽJL_Z»‡¹C¥ú²+Õ—]ݦ2V#z-”È —PŠï¥€.!:‚P=j¶d˜~[!>Ì÷( SÓt¢»š@; ÈØLäùÿ\õü}ÕÚ–}•®‘Gæ‡\÷§Ò¾=Î)‘¨³–ÏRjXº58[R.¢a’ök+ê;ûÊ2`WEXˆÏé.×ÑQöì¬v³gI¿Ò!ŒÊ™VèsxË®:²è‡Øveqåh€À›_¡¶Ll ç¿ O·Ù7L逴2¥cP{ȶa¨H]@0í¥¼saQгo[J ^ª®©æ© ázó^ Ï{ýÁ.«ƒÚ' ‘“Ó]ÁÚ ™Ñù¡&]UÐã9/0Lf %]»`c¢ÝÒ¸&õÔN53"ŠêBç^°ýsÖë?ŒÃãL,ˆ¥Ÿ˜Ófè›T_zK'c–Ý`3a˜Uãmõïän”Åê¶ô¶ìùoõ$öxRÒ#¾.þ'ÜyДàÉ‘Še7W(’î즂<´µ7Œ[ G'¼‘=iÆ‹£jØQå2Å[U§–@˜Ó<.­æyŠØ2ö)NN.8“M°ò_çÈžàØÊce‰ìíFœvh"Vá.;¸{é(W çK,l1Ó£›Z‹‚•f?Vy[áN\Vº•`%èÖGLi°’Ô`%0/£ÝçÃip%(ÀþÂÀfœÜ秘Š!œÆ’+íd%ørÜ;~o†é”´úÊÕôÓsÞFV¥žUÏæ“‰‘åŠÌœ~!k'Z°Í ]eþ;Z}â<-‹Ã_äíI¾§F áËY¤sîÂÜR‰m¸«Ìø‚ÍÇäGãý,6§¨È!M}m‘x«œpÈEã‹2‡eM·v´à° w»-AÓáûÂbujbB+ÁH¶æIMAf¾ºË2^Éj¬–o®‘ ä‹>ÌÊqNEåʲ.ŒÃTÚOXX¢Rgð›¤h¶¬ÒàI ÇÃU‹W*©Ý‹s`/0Ë ]ÅH±¤—ÕO’qƒm%˜DÍ*®å˜wJD mð>áp@QÑð?x‹“ °óHþé¯9L[ù’P㈼¬Y–¼’³<&|)Ýܯ£›t^©¶1§„áqlËW×Óüš®pR!ˆv­Î­˜\ÒC¢ÿDˆÎxÜÏ¢SŽJ§ˆ>‹Ðe,¢Þ°»Õqr0êƒ ¦<–žþŽ8?޵V'CzYïö^Åâ*±zÅ.ðµ©KÏþN}ŠÄÐ%–8<{v*oȰsŽã®å¬_ç©åºŒ¬ÒMíÐoÃÆjZŒrÆî®;=FÜì.<ÖÛëR/:&A·R—¯<¬ÍUf%v¾µ¨#‹Þç­ÞXÓ‘±>_© õ ƒ—Ÿ¹6Ôã5ÐÕ†â O“ÈàŠ¿‹[_Ç|G· ›<⛬^X8*b{\>ïü¾·ãn# «‘Uœ.Ææñ¢c¥“ï¾M®yï˜ÕKñü².âuFÿ¡ay¿s÷.:ÞpÆSºƒS_*†Äiû2åµRwòämRÝ´W7 ‚3&,Ó[AÄŽòdC(/ŒÉ/SÞò§†òbj?¸‘ ¶ð‡Ye”sÁÑú]¼] ÖÀ¹Rj[£4í7戈 {õ¥Ë˜üƒxš£4úS%Ï Ex­ üšÛKسñ9ªu€´] û‹5FëLã€ZõWsð)ÆÙ!ãÅ v•”œ>_¾œÒ¡énÇŸ¾˜ƒ¨ýg)îÀcñRXt’ÇÚHW¿g+F™æ—EŽ„9µ®# /–‰T“Ò­–øîJ$¹/ÄŠž¦±\Â{@ü»Î£ùnL³ÇF›ÑöÁjQWL1`%²G V$ö­E¤kô› ¥"Ýâœ[§ßÁŒ_aßœ —iáמñè˜æø v…ñØ2½rS¨YŽF¥mpXA–¶w§¶Ÿ¾G“ëÖdtt#\®\sQu¯¾†*ÚÏÝ,~CÈ֮˕ŠdjnLMžÜ†§ÈÆiOÛ{±ôÕ»*0m‰ãMûÖâžáî[ë/BxE¿¥Ò,aM:®ø+zÓ‰´†½ð˜2ß—Ö©“RürÉO’%•û~_NÏHAe’¸ Ü$ÿ;8«¬|i¤ùÞ‹mNéì"~ ãN{ß_¬j4›œ¯©Q5Ù/Õ´æ‘`Z!ËŸÓÚ5Òu@ˆì Âp熘(±+}Áûó¡˜áTúë,)`´÷tœêÄy¯sßpq)ÑÐPØ)üØÎƒêaˆbâÄ´}¬f +Ø#K¦È}§Bª5‡„¦û{’¶¸Z8òÕj”„rÖ¶V™vÓ• èNø]–O_˜^^ãgÁÓN6A€|ðX’ü S'×(ý&Y+,ÞŽƒ÷`É-»U/:ò ¯¦®(«˜X®gÀÆûÚ\¬FÔ+\¼:læ¢z…G•XÉO~ìyo P:ìů˜8JŽÇ†“ RéäÛÁ¶c]¿´ «“ØÏI<·|¡¿)óqþ©oÿ }lö³Î~IÊÝe\ó]qñ!“|£3¦ÔñM|ªðyßHg»îñ»êêý³÷ÿz'¯ ç˜Ô¿Þ‰ÔÎŒ°—òªÉ„®Ê…ð{qXD¯„Ýôì>*¥u~UêX¯ÆOZiŸ/PïU^+Ü„c¸;<áõoIfó'J>9€ˆUzpn߉¢¬¦ÛûJ^Ë`Evª¯öϼm­â'cÛ"ñó’º Ì}:iD 2¤)æ GXÉRfeÍ¡¹c òóZQ£•ãòGh’ñ\©­¨LâîFtîî†¹ïÉ7ùzåôž=û¦*rãÂj·©kêRÁüv`bºµJóÐj¦cé©ávàÅULSâo\ýi÷‘Ž|QF¾‰—i¨9°À¦n©$8/c)ÆÚ¯6íY´¹Kðâj]©~¸õþ<ÙR‡ã‰Í]¢r‘û²`‹û‹¢2e ¢àöƒxmíœsH´W‡&¤ùGþ²FY rL¦lôo’™Ëܸ¥+pA^…ø7¡§¢±üô‹¤S@k¨j®¨õi!éIxG ÞÉ.¸¦6­ m5H'c@éj¤Bò%o)êÜó%›¯pðOä>Ë·¹dÝ÷~ 8©–Nú@¦åQý„˽« Ûh×ñ2ŸJ÷%'a>ëM)g±"q•xMv•w±A·6 {EßPÊŒ L0*KJ gFÕ}ÊTüôØ–jN‘NY¬*‹Ÿ StÓ•+[λÿŸµG¶˜ì/¤Þðp1†ŽGÍm³ÕD=ï9eþ£5Ÿó[Œ{ƒ80e`“!|j ëïÛ>úØ|ŒùÑzG[¿Ñw˼øëëDË¢}ƒí·ñc(ú)ÿ-$ß|š9.p¨Äk#\ØIÈØŠ ù¼ 0Ûßõ®l#?ý%•™J¹–ã•xÈÑ–: ÎF=`=5WÕ€£j†î#D_?½÷ŸðßÿÁ¼Åendstream endobj 299 0 obj << /Filter /FlateDecode /Length 6437 >> stream xœÝ\[sG–~×nÌoPÌÃFu„UÎûÅoøÂØ`< ñì„Ù‡6È‚AÐŒ¬‡ýë{.Y•'³«„x#vÃNZÙy=—ï;çdÿëPúPáåÿž|~߯ÃÓ‹uxzð¯M=,ÿ{ôüðËcèá|2f•õáñ/üU}hLmЇÑÇ1[xüüà§áËÍ‘MÎ)…ád£FŸ½ÉaxM­TTI6Ÿbç˜Söà øÔ¦¨C¶›#k͘“þŸšœOÃ9õUÊ¥ÿ:þþÀ›Ñªdá3—`°txüø`øïÍñ?h~§,Süí§á5Œåt0>ÃdG¸²£>ãÙbrÃíM²£Q&Âd°k+ÄÝ,ë±øþ9-Ñ}¡¥• ix´1·ž†'<¼Ïz¸s‚_³9Y¨³Îʘ0\à1¥`k¿7Mûýü¾SòŒÓ#lî÷m=mÜÁž¬S¼åþøBΙ†ÓvÔÚè›_´1áñÁ&Ç MsÝŸñÇ1EÜ*îÉdh‹#¤fmžvê"Pí66#tÖÃÃúª G‘«x¸Á§ ±ß¸9rÆÁLaxøð6žd‚ ’¼›Ü)N›Gè2à`FÁîƒìE1ñêOx6»áîf>1Â[×ïär^/Ê‘“ ØÁ¦w¿pÛƒ¤Š‹•½yeY~ÌÒl´qÃöŒä.ÀBÅê~åó}3Ü¡}'¥u*Ç¡"(`Ó{{^®o¸ŸªœàvO6(oÖ»1›ÌâúÓpz¾FBîdTÇQTz]1.¸a'Uw:•ÚËê5'IÍñNß…i)÷ë¾–ô,ŽÊ‚”N½Ç:Í7¥cA.ÁrΊy— ï+¼ìþŽÏ†û9‰nH!/ÅǯgSGÂêwŠfqîñŠÔN)“#mÝ‚Çfè܃$Y~ó Î̃æÁ5>ƒÈ÷&Ò9ƒLñ³Ð]2Úl¬Ýï’­†ËyÍ ;³‰î æÌ°”ÿÀ&ˆˆ²Ã×=²l¡kÊÐ Í`[î«@ÕnmPÝA?¶…RvV•ù5áä§Uš¤œ-zKgJ¢•r>.]9ÌÔNWþ٬𥳑üWiêìiOÙšAoÇO`æ¡ÉÆS 5Y8­r³O‚ËðÆÅKoÀß± ú,Ž|dL³?„ÿJ§ØûGVìýü·={³=ß³÷bž=˜sßz6œP)Ð 4ø`I4èð“:ÜíiˆØü´ç{·ZI;²÷³fcæ¢2kYx[;¬cZ@\òã&!Ò1ÌÞÿ #D;€e›/6èwáfRë<–Ͳ:áÑ:×1ÿíâ—š©È¨`ð¬t8ª¶Ã›^Cµ ªÁdó•X2U,f­±}TÏg÷œÎÔ*y.¯ØÖ?:'!ß¾@O‡HÆoTOœû–}wz>fûsõegõ:ß8µI#Ð÷@»˜Þ”æùQmöšü™v …~Z¸Û­8¿w´L½âQ+Àø W8rý|O°ÚÒÏ~=œçñlæ2œ} t@•‡L)’@0ÀÁíÇé€È¨nŸ7ðºjÝ%Hž ­ £‡Ÿ‚5Æ*ZË> 4 -÷tߣ!‡3I³ïB‹ûšt$%+‘9É:ØX_k¬ÃéƒáŒÍ`£&GDŠÓ9€©3î/iccl-9þíŽ^wC ¤±ý‚,ê¯kæn‡#ÀØ®ø ¹I`áÅxç¯ÁðfWàΞÙwÀ+rh\Jƒ9:Ÿâ2"‡ÒÛ OÑ>­ûg§ÂËr4{°–̵f[«/%€Ý‚¯&•‡¿ ¼v‡æøæøà¯°Ó ¾3Ý;¿‚†%v˜5~to9å˜>!Q ÎeºPRp–²ù²6¿¨ÍÏ›oÞ×kóŸµy±8ñvñSÑwW›¿ô}I7F©&só¼6OkóóÚüµ6Mm¦Å¾OkSÕ¦“}ëå]Aç;ù™è|Rh6Û«Æ‚L,b#ìè#Üv|ÍU-µ=ìp ®¾wÌ–¶á™Óp·§!Òìø¹z‹@ÈAÂ9x„ çwdø  ,^æñäóZ†|Îåß«÷G7 œ³ðy j0ˆÁмãÝ$ÀÄH q;`ô?‚àhào?9Áq-*…XåÝÚY qÖÎÁö¨ç0gA˜c‚ÅÉzÆgå’aOwv§O7ÒÅà¢2pj±Y0þx¶ ®p§óš$Þ õ)³=<@ž§Ó14£m4¾lŒ¤ä1ÁIåŒb˜¬ Œ”˜ÁÞ]«ñˆEûö8„Gbb}$BjÏô=€FßÐöáˆôðo{Ѽ‚ë©ðŽ‚±Í©9,°peäÖú“¢ Z”ïP~ŠI8œv÷²Ê|QQTË‘a[€f ¸.0´PÖ¶@¥…!V{ï\‘ °‹î\Gä”5âp`fÀ•·‚—ðˆ¼)£/7sh¦êPÒ¿S´íq »s¾ÿXœp¡?‹ÞÄ»öBnrM[šxDiÚ½ày0þ°ë"5G&€N¡‹ U>r¨ìÞ¿VÞÀ&:EÙù§oI«€Aªö •Õ¦±=혤þ|3Eq]ï0t˜lƒÈgTò˜kèÉS|/h©8j 2+Ž<Ï¡”xYÑVÔ0‚¢áÔîR‡I¿`Þ¬m‰cO22E.¤rpȃ’2ð‚È{ Ð?À«ÈÆØ¹ŸãVÁ[‰íœlÈZƒ‰;~Ú^ \·&ÚD\I×Cí°µà }qŽC'ÄF^κ/BûËo™?±ÌYØI3¨/‚¬gÕZoOK,'é8Ü;© ü†¥Ì€/ûÇ<Þ®rágÂï-_826¨%öã&)ŸÂ-Ê2!T)‚t4‰‰¦¯€o†ãÍLô?ãO‘k/û­Ÿ+,“A'›¹)&ÇVØ7l‰à„Õ|÷yB k‚>U. Òa‹B±\ÃÖÁÓÁÊ à®1?³µœ¸q“ËWJ—§²³šÉyÂâòA™››üÞRÖÞ´&íiuÿm̈ºòÌÞaYK€3á ØÌupÅ !ßKÓaûcôùÓÒt‹¦9ïÑti“ x"Môˆ¨®hÒá%+Gx×{Ûn:ßé+Pµ=·*3Å”± d~Ûv!’€On` °EÂÈ×ç»Ðу"+ öÅh*˜G¶…Õ6ZÚ¹w%Ügóe•ó%;ææ ôõÃζbÝvöW„S" 6ÅØRˆÀsôñèNç_£1ZLO,.úÏö¢E’öŒí–R!ÕŒ€ÂȤ3™üš¢‚šæväÒÿ)C[Ï6àk &@8Ü 6 ¦yFmm'»‚&?±Ÿ†&¦È:?m›}b@üFåzøð«F¬gü$ÁNÚ<±‘<å™1( é¡4Cl@É-“ÂíÌVÄGkÆîõ²ý˜Ï3Ph¢=Dtt,ÊM¤WŽ` ùïò§ <í³ âOwçUí$)YP1öó26¸·fÒ%kÛRU"p mQϬþè àl`í´þž´–µ Ö*$36ð*IÙNãËA陘’Ç[†ŒR&˜íXŒŸ/këÇ ¶äÁwùÀºv …iÅF|žŸ¬£rÄ: ÛYe:=öÿÏt û »ì®•Lbì9íã¥Ïj/éìæ6éwg2–ûöƒ`êÚb‹œ(у¹·—ro¦ç4a XJ‰¨ß îI¤ea‹3nÇÀ&üݠÔÿ‹Žl5Ù~8GšP‡°©M¶CììoÂüðBˆ€L¹=Û ›]¸|¸˜Žâߢ V=‹ ø¾«_„t— ŠšÒ RL9~ך@«ˆqǦ¦èé·´°~)L­ºð´„7² Œ#"`¯c‰“%0¤Ð ¯Õ!úí¸_œ×Á%`ýÀ5¦.©DÊ=ýVFÿ!‹Ú@Ϭ+q9ßÞ3k÷òˆ\ç½Lÿ¥šÑ» ÿ Ô–ÎE¤–€)i°ßÐ ˜Ù?Æ”0ãcS_uͲHÃùc ë›+yšÙJÌDGä¨À“¾% O+„hcSöí‚§„3‘¸ÿj(¾(€Nø9 ¦KÎME>1“Äeß×Î¥Z¤ðËqDN_)@ÔvÙ€TŽÚ]EmÚÚô…„«säÂ]bŒX Â☜ì@ž^„nÖ]«³£ö•Á ÷÷r/t–!ÕAúKŸÔ8EW ËAùÎN©¬že¥š£1Ä¥ô4¯†Ì»<«·:Çx"hÄ·{Ež [ßÏý.>™5ŸuZ‚[˜Ò$Cºå/)x`ýy3VÆOc.ñÔ2ªö:rt ðRrmòNR¾Nå©Â“ÂÌyŸUÌü,iO¨gò`Zßà°´x Ì! |Õ­U¼âåá|Gœìw¨)ÀP_ðŸAÐå÷ywÉLüD á„ûà~"v*L·¼[‘;i“qK‘ µ‡Æ7¥¬ªéJWmºÖ '~.Ó¹'$äd£)^Ë,¯ À?z=|aN03SÙ2º¹~¸0š`´>Ü/g0}9Ãt­F_Â"0Q’I7~—z—-\}:t.¦o ÈÖ½Zl¾¬Í/QTåq–Ÿ*#D‰ÃWµy¿6oÕæ½Úû‚¢ƒ) `DÔVœ×¦¨kE'µù¨6_-ζ[w¹ÚáqmŠ"‰G‹}ï׿‹Ú<©Í7µy!Gø˜ˆ€fÊ:°›Îû刔¨O›cº±A4v6ˆÆ,•¼»®vGFèáCŠÖ€ÔVfs¥Ïpt',ÝzM•-g½)ÂõÚ©^ª|OÍYYÒÎùJ âZEïÓT)mŒâ}’júiàµlýÚI]ñ/&Ûæ Î…ä°õ% ‚ìÝŠ—ÉÁÔçŠIp/qü‘ïxßD_ Ü£Cöò=ßWÊ™!‚âkÍ» Ën^»þEÉ‘õiL¡­Z˜È„viÍèÅd Î×Àö˜öj†Ú¥œÂ_áH}¶A AYÚ} á±p§PЛB!´ßsšCë`;Ñ~+Ÿ›Ÿ/6ß¼¯9næ×GfÊl8ÍŒšXýïðìà&D¼<ñkmêŦèû´6ųƒæ‰ÂÇPn¬wb­±Ë”Ûêó‘0"iâ”·¹‰e}õ„ŒASöæáûì(R «‹×ÀeË…üm.Î%Ü—oض>˜|Víõ©ðÔ»×Õgõ%à°¿ªyoÇsRœZ—ðœXIßúp>#ÿQN†€UÉ8‰“–™û5Þ$|GÀÏài•¹MþÎ » 1•’0€Lœóá0s©õêƒøùW,×ºîƒøiÎ’ýôx£6¸_Hžž÷ ¦ÔâU·€0—â) ºª$z:B)ì¨Éë¥VZqQ¯XªA[{­5V”µR㟀<ÖßÍi9$¤;¿RÞUÚ_^¾F ˆxEÂWZYüL©fwï5~AU[ÂyÕÃv+,%\çaû^ÕúòÃv…ÅÜiÐ`þ/áV|_8É¿0e;!ÐÑ#Á%d¾gù}j¾}‚¢ééähJœæÒ/ õd—îɘ®x:·ÃÎË7[E¾ÅŸbM[«È)÷f®Aî Κ?I < ‘ôÉ'sÞ¢þÆAï§ëÛå¨ùM¬§I’àï€õößGY¦$ˆ˜xáqå9’¿Ñ{¶ùÒûž’’˜úË?èH£¥j#ó¢nô1ß„OýÃÊYÝ—ßXwE¦M¢¢HJÿ¡.o_V¢×¹0ú*>öíóóðíVxe!¶{oÊ1ÚØVù¬èÇÂMÑ÷I!²Î´4z‰a¥_ýwÀúL!¥?T`ëʱIcЮ@Ä™îÎ+ø¥u|¿Ê'YÔñ¹¬«1LÙO>:\¤ùÒ(žÆå‡/;ã ¢Å:L>æƒy¯2å¼ïxišÊfZ™,ªóºÊáÚ»1¾àQ>(™‰ÒÌùO}–AJQ˜FË0­B_ü#ÔLþÑ£™ô·,+£Ê×S² ¿"ëúl± _@bQtÙø©7ü9îôs2s'VÇ“Ðoo]…U?‰%Á•€ñ•)²ï‰ÁÁÖ›-Ô=¹ÝÖÿ–ãóp|œãÔ{ Tƒ/ÞçRAR*Jég žMÊ“ß~£ty=Zz0¥§üÀ[a $¯}]oSêúâpdzà`õGûJ@\Áh!èÓ³kaãwOªÛ,׈'ü÷Í«NQÔý‹ŸZ}7Ý—£ÇBø0_[IdéM[ZÓ98ïÇäû2×Ð÷У׊‚W=YFiâ5ðœcß±ÛÅH®8™ú£ª¬%Í­Óã—ü¬_dï/x@ ö®ß-£(³&¸×ÂXˆ¯ïëë]Žj0áâÀ.ml™¹$Q‘ò‹A)F M¬‰ê쌄 “îeB›­[¬—ûL<6ìªÉ…¶¿Þ0zDQóïË,;§z4Ë%ºpD muÊYŽ%ûÿ¡˜³*H`\Ô΋^¨¼*‘@–ɽ˜PÔÍœ÷‡á¢LRô•b“äªKx(6Ñx9tš•üj¬¥}o*¶‹ú×CJ.è Çbð˜PôcùkŠ€(Õ5\ÕƒªðSíʼÑáѧp&¤_c3‘ì¾/—Ô¨¢ Í…aÊDéý¢²ÆÙî•çœr{5•§MLKú0$pRmÎâkTƒ–}–BD–@}Bû3!.IUc—E©wa'~­x³ákóË‘‰ö'AŽ0ÜK¾]ì»e ¨ñ¡ô‘jjIX.%pzCùV~T»ü3'’ºÞôW}hE©üH ª.”çÕ_OX6L «_× ù÷ª.¢–žØþa°Oô(E{Å‘«~…˜éYu¾ &À®|ÿ( ˜[Zû ”Ùªî9ó $$ê.0½‰ÁD×µi ŬԵQÔ¶þEU³A6¤þ%ú¦®@Éà»åPæû¯ÿ@ÿendstream endobj 300 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 785 >> stream xœ-{HQÆïuÜiªu5c‰¥œ¥ƒÔ5#±Ô4Ëè±a¾*zŽ:®fûhwÖUË,É\¼¥¥•öué…AeïÀžfo³’(++BH(âÌr‰š(ÎßwàðãûF¡!cÌeXrrÅ›ÿú8e"V&…(Q ñ³‚4Q(|^ZÔ cj¥Žh¢ í î„êq`)1;½»2ÎJW©µD¢ § ñII‰Ó…fs’n“\¥…¢]°ˆr‰deuÙ$ä8 K%¹RˆN.‘eçì¸8¯×+Úܱ—5uÚtÁ[*—Ù’[r•KEB¦Ã. ËD›$üÏû_36§G–\‚ÅQ$¹ì¡QnOVvΦ„rQ>ŠG h&² âÔž(U¡›Xƒóñ#üÿV|:eüö‡ð½ €é£d*éúuëlRLòë×XŲª÷Ìêà GÍ”£á4ľ¿ ³ÌxV:µ³gËËŠöz¿ìwïs‘rn^z┥)žl¾lN·8èxO a0ÂõפKÆçœÍæc—%/¸¯Ÿ ?¾V¹ö2¯ƒZÒ4ùñ ¸¸\¬Oû—Yý|„¼{¸:xMzÚÎÖ,×@;«ó”¬£°"p$ÀÒ‚ÖQ½cÞŽåÇ„&úµ£{›´Z„þÕ¾\?endstream endobj 301 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 945 >> stream xœ…’]L[uÆÿ‡–rµˆÉQ6䜳¸l,!À–¸ ²,æü˜,X²‘é¬p¤eý¢èN¿PZFw^ÊWY)m)P>¥¦ Ì48à Ã%êÍ.¼ðBã.ÔD#ÿƒg–¹Ý™x÷>O~Ïû¼/”Yˆ ˆÜÚºš‹o©ØåR!½š%+À³“½s.»åל.>ÊœÒj@­µréoóKØU€-ùX÷"RD»à«5[«¾Udz%͇Ù#••ÇKÙ£•lµ‘³ê›µ&¶NËë8£–ÏÛ`nÖs¼À–œÔñ¼¥ª¼Üf³•iefkë©Ã¥¬MÏëØz®ƒ³vq-ìY³‰gÏkû˜]¥ì`¢f©Þûà…Þ½ïnê·pÎ7˜‰Ž»ÁÙ)Š­-iÚìéî_ˆùƒF#ý”&‰¿¶Ò†DRƒaðÃàS‚Ýîq»é«K-C‹|ò\%ï“ó~øè»¯Ò‹±i&¦ûÜ‘‚4Ì$B“ñ•ÐFƵ êE«—ö´ §õ@v{Sÿ‚D¨^ð‰žëµN¼~|s}çç˜Äo÷ÃqX†¹Î`3©‘æ=I<‚•·“Ä6&p#Ö(¤+C]Ä÷ÁG:†••…åÕµ+«'äúýò1ùåÊôëøPü¡r^¿•JÁ°Mê믗®>qœ@6îl¦æ°föK&õèax‚0+ÆlÏhõI|,Ià,Œ°ÿkžÅ&j 7¦·’Õrµœ'¿"—ÉèÛ·pÁ±{Ìb\@6˜–ÌŒ¤FÌh’ÿâæÂüâ\ân$äÊ”åêù8œg4?mà\þ:Õ ‰ÜÙGůÝ6~¬×ë,“]óŸ%Ë´ìx2Eõ:|p’ ÷ÚÖÿ\ÃE·v+¾vÜ}´§ÝRZäõ˜Ìôû™ŸqÏpd á½—.¼Ö$u¹BŽ AhÄgÙrh·×*ÞÈäXËäÀÄãÌ»=Þ)¡ú‡üÃ01At‰r~oá'¢ ÀD 0:6!Šc´L=‰<û­s[mé™ÄdˆþÿËkìQ齿ú§£*ùƒ@NzÏv½G)˜Õ¹É ZÝ¯Ö ôz[ÏÃendstream endobj 302 0 obj << /Filter /FlateDecode /Length 3287 >> stream xœÕZKoǾAî¹9³8ê÷Ã'Ó–mØ¡[¢mV+’"I¤LRVdÉ_ÏW]=ÓÕÃ!å8’€ööötW×㫯jö‡]5ê]EõÿÑ«ûlÜ=½ÚQ»§;?ìèòínýwôj÷£C¬p43f•õîá³~Tï&½}³õ»‡¯v6‡ÅbïºÅ)Ž!Gøjr•Ǭïâ+×4®^±Å†)GgºÙç½ÕµÊ!”«ÇdMròû£×¸CŒØjt“\„7î¹ìp;ìoŒÃÒÈ&ªtKaÏ”rÃ;aD>ÖdÇÚ·ÚZí0[öÍÚCwóÏhÖã##B<&.ÓP_¨7®Ó'´‡qøäɸêéP©u%”TòÊÛ=ù¦mxIº0¸ ¶#›ì‡ˆNawOÃO\6l‹=*¼é¯FÛÛl¬íLQü]‡¥ÒŸbÞ¤áát¼I€Ž¸° n5TY5U|VjâŠÆ)çàãZÜ=zóä@ð6ç(¾(1 ¼ýâLØìœ±Ä§4|WpG[çe”¾l¸CTpÉ“Ùé1øôð“¤m Å`|¦ ÆXø ]Á­BF§T´Íè\È»{Bè 4ÕŒaqJ)KA»[£mŸ–ݽŠ7åózÇ‘6GZŽk•a¼<<¼GóÂúá‘P1ÝW'¬^ å¼,ÓYã gïxŸâ<à3³î €ðò_€Ò´™,øeñ¤!ãÅÆc™¶~8natRÜ’0첡á§ŽWˆdœ¥ƒ†M_·§^wQOJƒ{U÷óNcWoF¥4Ç´U ºæÍn8s Χ&|û[üýˆ‚#…lSÁ ™²Þ¶œ&©`Zv.ºál>›u í”é ’c‡§¼ßjðŒð¬Eðd?Xjp×GÎØÌAÇ.kïĵºÇ©‡tó¼9~ ^<òv1E>Ò /›&»ì9/>•r\6ôÛkÎYb 4Úà*²¥´l7S¶ )ÛÓ{•{À·NÁéžo€ßç /Xg0û¢åcKŸ‹ æA\œGrÁ‡|"È€lŠåc~ë1AÞàÅÚ%а¡+ÎøŠêßmÉ @9ç8Pˆöá”<0áD=|E œö¶L&0Iª¸i y“¹P„‡ ·iRò“í9oèó«~8¨ç‚?±‚„`"…zF¬€x¨Lêôis‡77|Ç3Ô㫼‚Ó—-ªæsÓppÂcøåððAU†Ê f“ÃÖ|tD6†'Oê¾&jöYCæl\Åí#‡%Ãí¿¢_0ƒ£¼RX,È—s].¹˜T ~$1±-9°(çiGDp€}äŽ,«qÁÍ›;yj ãÂà ¼Â† 6ä)%r|»ËÅ  ³‰‹ž}{“_Gãôµdë̾Q×VJŠÑ‡­(X‘Æ#ulÉ"lVâb@Ôñl®$R€Ó¿.—õ 8rVw‚=$C„Ywóó¡ì«Hz_äPa#–ƒùÈ-vìü~³ž*Ân&;­²OA9_•lÝ™nüœÉ×K2 ³ƒÝн§Ð&ý3P¸žÁV›GãRE<—ˆ'™ÕqcV¹¨"ˆgÜ(½hj콦=ȉ ê3ä‡s xÚ¼é‚ê%E–´ó1ÏRj]`Í*Šên*Òž<Ùgyr6ÎxÔ¹n{¾ix)¬~ÂXÉÆGÊ òfD½-ºE^>¸AE‡õ’iؽ™w»™ñ µ/¶í wÆ à_Ï$N^VƒÀÛ?eP ØØWÚ…$£pϹf¦”àØÛ—‚ž¥,òMËŽ}oe½#º‡ð&nÃØº«Jš¦ÙPbìG¾[@KW­|” “»Á™ñð¯Là‚O)ÔøœqºYG7†:ôo5ôMøW nnZmÿ5ð™àVü7±Ã–{jqƒåáU’y0VÿŒºeAmï?BÑOuü3DèOò ºžüÕg>€z_.Erßß#P 7õ¼BN¨JªÓÛ"Aˆ™k< `®]1[yËz¬ËÏ?hV/KDÀax°søÇï;(cÄõ&‡[^ˆW=äÕpLÝÖB*Í›ÁµÅ(IŸ°*BtsE#hbø½–*çÛèwåWßWt,# ­`Réâ =(Mí”qCS{db_”t¬ 6Ýì…Ub%Û&ÜJÎE¹ S*k χ:iç˜CZÞÆ#.våwÝu.¹º…ý‡ýóƳO%”¾%Ȥ„ÕA༅XÔÊ¢DA±-NíùAÔ¢w!V ¿hlçþdbßui>Ýhx;1óýE s*RÄ©ú–^möÇòÎ¥P å¼®M®¹ñOže“Ê]r¹bP˜â¨+söY‹¢àöùÜšÿFh^¸ÂË‚jÙ)“dþy#óBÇt}m@ã½[swµðóíäfYÚI•üÔež9mß@¥gz’7ê©Ê©î/Ê'y…ìp<”fk¾­>½=ž}óš ã#ÁAØ…Jþ”‘5g”Æ$7yò`µ¯o¼ª(¢‡Q“3†;6S8ÁÙ÷et3m ~¦%š–fjÛ"ÎìÉŒœŸlæf«ä†]+µµ“ü +]>9Üùz‡@I5–×÷—·¿íGI±ò¶_›’8v©á§syåÏ]§Z9ž´áõêðyž·á¶ Ç6ü¬ /ÛðÍêib³Ûð¯Õ¼hÃ7m8¶á¶ ¯7M]¿æô2U kg5Vž/p% Þu?[dy÷­ø Ã/Ì~ÄoÒÌœoüwmyoý·tvîÕßAPþï3$hUÝ,‘ØW¤?òðÿÕôg-®âðkiJ$:‘¦Žß·ö¨ ÏV“âÁê¾b‡«åf%¥µá‡mø´ /~]ª> /W [ 1 3 1 ] /Info 3 0 R /Root 2 0 R /Size 304 /ID [<8a9ad36fcfbe38aadeea2f83d1597d4d><9f0ede477c4b510476bbaa01c6ddac18>] >> stream xœcb&F~0ù‰ $À8Jò?ƒ`á ›]ƒ/6ŽÆàP#ÿ3(„ìÆÝƒo£q7tÈÿ š»wÙ²Ï@±&Ÿ "¥@¤Ø/)rD*±ƒHžR)W"9R@$“dä™bK‚E΂HÑ™ Rv%ˆd´‘ì ’Û,òD²FƒHΫ ’Å!Îd &OƒÕl³_ƒeÁ$SØä- Rìf)±äq~[|ˆ4šN‡`]ü`× ‹ƒ]rlÚ0ùìZA[hØ„N°ÉA$ß|°ËSj¸‹lÎ {y!i, v9ã °É­àù "%_CLÊ€Ãì~~Pˆ1>:ƒÐñ)WX–ìª÷y×"ØRºd‚$ØRù ’9ªÜ`õŠ ’ Ee endstream endobj startxref 344764 %%EOF flexmix/inst/doc/flexmix-intro.R0000644000176200001440000001331213432516263016405 0ustar liggesusers### R code from vignette source 'flexmix-intro.Rnw' ### Encoding: UTF-8 ################################################### ### code chunk number 1: flexmix-intro.Rnw:31-37 ################################################### set.seed(1504) options(width=70, prompt = "R> ", continue = "+ ", useFancyQuotes = FALSE) grDevices::ps.options(family="Times") library("graphics") library("flexmix") data("NPreg") ################################################### ### code chunk number 2: flexmix-intro.Rnw:323-327 ################################################### library("flexmix") data("NPreg") m1 <- flexmix(yn ~ x + I(x^2), data = NPreg, k = 2) m1 ################################################### ### code chunk number 3: flexmix-intro.Rnw:330-331 ################################################### parameters(m1, component = 1) ################################################### ### code chunk number 4: flexmix-intro.Rnw:334-335 ################################################### parameters(m1, component = 2) ################################################### ### code chunk number 5: flexmix-intro.Rnw:340-341 ################################################### table(NPreg$class, clusters(m1)) ################################################### ### code chunk number 6: flexmix-intro.Rnw:344-345 ################################################### summary(m1) ################################################### ### code chunk number 7: flexmix-intro.Rnw:361-364 ################################################### par(mfrow=c(1,2)) plot(yn~x, col=class, pch=class, data=NPreg) plot(yp~x, col=class, pch=class, data=NPreg) ################################################### ### code chunk number 8: flexmix-intro.Rnw:382-383 ################################################### print(plot(m1)) ################################################### ### code chunk number 9: flexmix-intro.Rnw:403-405 ################################################### rm1 <- refit(m1) summary(rm1) ################################################### ### code chunk number 10: flexmix-intro.Rnw:426-427 ################################################### options(width=55) ################################################### ### code chunk number 11: flexmix-intro.Rnw:429-432 ################################################### m2 <- flexmix(yp ~ x, data = NPreg, k = 2, model = FLXMRglm(family = "poisson")) summary(m2) ################################################### ### code chunk number 12: flexmix-intro.Rnw:434-435 ################################################### options(width=65) ################################################### ### code chunk number 13: flexmix-intro.Rnw:439-440 ################################################### print(plot(m2)) ################################################### ### code chunk number 14: flexmix-intro.Rnw:483-486 ################################################### m3 <- flexmix(~ x, data = NPreg, k = 2, model=list(FLXMRglm(yn ~ . + I(x^2)), FLXMRglm(yp ~ ., family = "poisson"))) ################################################### ### code chunk number 15: flexmix-intro.Rnw:501-502 ################################################### print(plot(m3)) ################################################### ### code chunk number 16: flexmix-intro.Rnw:531-533 ################################################### m4 <- flexmix(yn ~ x + I(x^2) | id2, data = NPreg, k = 2) summary(m4) ################################################### ### code chunk number 17: flexmix-intro.Rnw:549-551 ################################################### m5 <- flexmix(yn ~ x + I(x^2), data = NPreg, k = 2, control = list(iter.max = 15, verbose = 3, classify = "hard")) ################################################### ### code chunk number 18: flexmix-intro.Rnw:568-572 ################################################### m6 <- flexmix(yp ~ x + I(x^2), data = NPreg, k = 4, control = list(minprior = 0.2)) m6 ################################################### ### code chunk number 19: flexmix-intro.Rnw:582-585 ################################################### m7 <- stepFlexmix(yp ~ x + I(x^2), data = NPreg, control = list(verbose = 0), k = 1:5, nrep = 5) ################################################### ### code chunk number 20: flexmix-intro.Rnw:591-592 ################################################### getModel(m7, "BIC") ################################################### ### code chunk number 21: flexmix-intro.Rnw:727-734 ################################################### library("flexmix") set.seed(1504) options(width=60) grDevices::ps.options(family="Times") suppressMessages(require("ellipse")) suppressMessages(require("mvtnorm")) source("mymclust.R") ################################################### ### code chunk number 22: flexmix-intro.Rnw:740-743 ################################################### data("Nclus") m1 <- flexmix(Nclus ~ 1, k = 4, model = mymclust()) summary(m1) ################################################### ### code chunk number 23: flexmix-intro.Rnw:754-756 ################################################### m2 <- flexmix(Nclus ~ 1, k = 4, model = mymclust(diagonal = FALSE)) summary(m2) ################################################### ### code chunk number 24: flexmix-intro.Rnw:761-764 ################################################### par(mfrow=1:2) plotEll(m1, Nclus) plotEll(m2, Nclus) ################################################### ### code chunk number 25: flexmix-intro.Rnw:803-807 ################################################### SI <- sessionInfo() pkgs <- paste(sapply(c(SI$otherPkgs, SI$loadedOnly), function(x) paste("\\\\pkg{", x$Package, "} ", x$Version, sep = "")), collapse = ", ") flexmix/inst/doc/regression-examples.Rnw0000644000176200001440000012412713425024236020144 0ustar liggesusers\documentclass[nojss]{jss} \usepackage{amsfonts,bm,amsmath,amssymb} %%\usepackage{Sweave} %% already provided by jss.cls %%%\VignetteIndexEntry{Applications of finite mixtures of regression models} %%\VignetteDepends{flexmix} %%\VignetteKeywords{R, finite mixture model, generalized linear model, latent class regression} %%\VignettePackage{flexmix} \title{Applications of finite mixtures of regression models} <>= library("stats") library("graphics") library("flexmix") @ \author{Bettina Gr{\"u}n\\ Johannes Kepler Universit{\"at} Linz \And Friedrich Leisch\\ Universit\"at f\"ur Bodenkultur Wien} \Plainauthor{Bettina Gr{\"u}n, Friedrich Leisch} \Address{ Bettina Gr\"un\\ Institut f\"ur Angewandte Statistik / IFAS\\ Johannes Kepler Universit{\"at} Linz\\ Freist\"adter Stra\ss{}e 315\\ 4040 Linz, Austria\\ E-mail: \email{Bettina.Gruen@jku.at}\\ Friedrich Leisch\\ Institut f\"ur Angewandte Statistik und EDV\\ Universit\"at f\"ur Bodenkultur Wien\\ Peter Jordan Stra\ss{}e 82\\ 1190 Wien, Austria\\ E-mail: \email{Friedrich.Leisch@boku.ac.at}\\ URL: \url{http://www.statistik.lmu.de/~leisch/} } \Abstract{ Package \pkg{flexmix} provides functionality for fitting finite mixtures of regression models. The available model class includes generalized linear models with varying and fixed effects for the component specific models and multinomial logit models for the concomitant variable models. This model class includes random intercept models where the random part is modelled by a finite mixture instead of a-priori selecting a suitable distribution. The application of the package is illustrated on various datasets which have been previously used in the literature to fit finite mixtures of Gaussian, binomial or Poisson regression models. The \proglang{R} commands are given to fit the proposed models and additional insights are gained by visualizing the data and the fitted models as well as by fitting slightly modified models. } \Keywords{\proglang{R}, finite mixture models, generalized linear models, concomitant variables} \Plainkeywords{R, finite mixture models, generalized linear models, concomitant variables} %%------------------------------------------------------------------------- %%------------------------------------------------------------------------- \begin{document} \SweaveOpts{engine=R, echo=true, height=5, width=8, eps=FALSE, keep.source=TRUE} \setkeys{Gin}{width=0.8\textwidth} <>= options(width=70, prompt = "R> ", continue = "+ ", useFancyQuotes = FALSE) set.seed(1802) library("lattice") ltheme <- canonical.theme("postscript", FALSE) lattice.options(default.theme=ltheme) @ %%------------------------------------------------------------------------- %%------------------------------------------------------------------------- \section{Introduction} Package \pkg{flexmix} provides infrastructure for flexible fitting of finite mixtures models. The design principles of the package allow easy extensibility and rapid prototyping. In addition, the main focus of the available functionality is on fitting finite mixtures of regression models, as other packages in \proglang{R} exist which have specialized functionality for model-based clustering, such as e.g.~\pkg{mclust} \citep{flexmix:Fraley+Raftery:2002a} for finite mixtures of Gaussian distributions. \cite{flexmix:Leisch:2004a} gives a general introduction into the package outlining the main implementational principles and illustrating the use of the package. The paper is also contained as a vignette in the package. An example for fitting mixtures of Gaussian regression models is given in \cite{flexmix:Gruen+Leisch:2006}. This paper focuses on examples of finite mixtures of binomial logit and Poisson regression models. Several datasets which have been previously used in the literature to demonstrate the use of finite mixtures of regression models have been selected to illustrate the application of the package. The model class covered are finite mixtures of generalized linear model with focus on binomial logit and Poisson regressions. The regression coefficients as well as the dispersion parameters of the component specific models are assumed to vary for all components, vary between groups of components, i.e.~to have a nesting, or to be fixed over all components. In addition it is possible to specify concomitant variable models in order to be able to characterize the components. Random intercept models are a special case of finite mixtures with varying and fixed effects as fixed effects are assumed for the coefficients of all covariates and varying effects for the intercept. These models are often used to capture overdispersion in the data which can occur for example if important covariates are omitted in the regression. It is then assumed that the influence of these covariates can be captured by allowing a random distribution for the intercept. This illustration does not only show how the package \pkg{flexmix} can be used for fitting finite mixtures of regression models but also indicates the advantages of using an extension package of an environment for statistical computing and graphics instead of a stand-alone package as available visualization techniques can be used for inspecting the data and the fitted models. In addition users already familiar with \proglang{R} and its formula interface should find the model specification and a lot of commands for exploring the fitted model intuitive. %%------------------------------------------------------------------------- %%------------------------------------------------------------------------- \section{Model specification} Finite mixtures of Gaussian regressions with concomitant variable models are given by: \begin{align*} H(y\,|\,\bm{x}, \bm{w}, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s(\bm{w}, \bm{\alpha}) \textrm{N}(y\,|\, \mu_s(\bm{x}), \sigma^2_s), \end{align*} where $\textrm{N}(\cdot\,|\, \mu_s(\bm{x}), \sigma^2_s)$ is the Gaussian distribution with mean $\mu_s(\bm{x}) = \bm{x}' \bm{\beta}^s$ and variance $\sigma^2_s$. $\Theta$ denotes the vector of all parameters of the mixture distribution and the dependent variables are $y$, the independent $\bm{x}$ and the concomitant $\bm{w}$. Finite mixtures of binomial regressions with concomitant variable models are given by: \begin{align*} H(y\,|\,T, \bm{x}, \bm{w}, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s(\bm{w}, \bm{\alpha}) \textrm{Bi}(y\,|\,T, \theta_s(\bm{x})), \end{align*} where $\textrm{Bi}(\cdot\,|\,T, \theta_s(\bm{x}))$ is the binomial distribution with number of trials equal to $T$ and success probability $\theta_s(\bm{x}) \in (0,1)$ given by $\textrm{logit}(\theta_s(\bm{x})) = \bm{x}' \bm{\beta}^s$. Finite mixtures of Poisson regressions are given by: \begin{align*} H(y \,|\, \bm{x}, \bm{w}, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s(\bm{w}, \bm{\alpha}) \textrm{Poi} (y \,|\, \lambda_s(\bm{x})), \end{align*} where $\textrm{Poi}(\cdot\,|\,\lambda_s(\bm{x}))$ denotes the Poisson distribution and $\log(\lambda_s(\bm{x})) = \bm{x}'\bm{\beta}^s$. For all these mixture distributions the coefficients are split into three different groups depending on if fixed, nested or varying effects are specified: \begin{align*} \bm{\beta}^s &= (\bm{\beta}_1, \bm{\beta}^{c(s)}_{2}, \bm{\beta}^{s}_3) \end{align*} where the first group represents the fixed, the second the nested and the third the varying effects. For the nested effects a partition $\mathcal{C} = \{c_s \,|\, s = 1,\ldots S\}$ of the $S$ components is determined where $c_s = \{s^* = 1,\ldots,S \,|\, c(s^*) = c(s)\}$. A similar splitting is possible for the variance of mixtures of Gaussian regression models. The function for maximum likelihood (ML) estimation with the Expectation-Maximization (EM) algorithm is \code{flexmix()} which is described in detail in \cite{flexmix:Leisch:2004a}. It takes as arguments a specification of the component specific model and of the concomitant variable model. The component specific model with varying, nested and fixed effects can be specified with the M-step driver \code{FLXMRglmfix()} which has arguments \code{formula} for the varying, \code{nested} for the nested and \code{fixed} for the fixed effects. \code{formula} and \code{fixed} take an argument of class \code{"formula"}, whereas \code{nested} expects an object of class \code{"FLXnested"} or a named list specifying the nested structure with a component \code{k} which is a vector of the number of components in each group of the partition and a component \code{formula} which is a vector of formulas for each group of the partition. In addition there is an argument \code{family} which has to be one of \code{gaussian}, \code{binomial}, \code{poisson} or \code{Gamma} and determines the component specific distribution function as well as an \code{offset} argument. The argument \code{varFix} can be used to determine the structure of the dispersion parameters. If only varying effects are specified the M-step driver \code{FLXMRglm()} can be used which only has an argument \code{formula} for the varying effects and also a \code{family} and an \code{offset} argument. This driver has the advantage that in the M-step the weighted ML estimation is made separately for each component which signifies that smaller model matrices are used. If a mixture model with a lot of components $S$ is fitted to a large data set with $N$ observations and the model matrix used in the M-step of \code{FLXMRglm()} has $N$ rows and $K$ columns, the model matrix used in the M-step of \code{FLXMRglmfix()} has $S N$ rows and up to $S K$ columns. In general the concomitant variable model is assumed to be a multinomial logit model, i.e.~: \begin{align*} \pi_s(\bm{w},\bm{\alpha}) &= \frac{e^{\bm{w}'\bm{\alpha}_s}}{\sum_{u = 1}^S e^{\bm{w}'\bm{\alpha}_u}} \quad \forall s, \end{align*} with $\bm{\alpha} = (\bm{\alpha}'_s)_{s=1,\ldots,S}$ and $\bm{\alpha}_1 \equiv \bm{0}$. This model can be fitted in \pkg{flexmix} with \code{FLXPmultinom()} which takes as argument \code{formula} the formula specification of the multinomial logit part. For fitting the function \code{nnet()} is used from package \pkg{MASS} \citep{flexmix:Venables+Ripley:2002} with the independent variables specified by the formula argument and the dependent variables are given by the a-posteriori probability estimates. %%------------------------------------------------------------------------- %%------------------------------------------------------------------------- \section[Using package flexmix]{Using package \pkg{flexmix}} In the following datasets from different areas such as medicine, biology and economics are used. There are three subsections: for finite mixtures of Gaussian regressions, for finite mixtures of binomial regression models and for finite mixtures of Poisson regression models. %%------------------------------------------------------------------------- \subsection{Finite mixtures of Gaussian regressions} This artificial dataset with 200 observations is given in \cite{flexmix:Gruen+Leisch:2006}. The data is generated from a mixture of Gaussian regression models with three components. There is an intercept with varying effects, an independent variable $x1$, which is a numeric variable, with fixed effects and another independent variable $x2$, which is a categorical variable with two levels, with nested effects. The prior probabilities depend on a concomitant variable $w$, which is also a categorical variable with two levels. Fixed effects are also assumed for the variance. The data is illustrated in Figure~\ref{fig:artificialData} and the true underlying model is given by: \begin{align*} H(y\,|\,(x1, x2), w, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s(w, \bm{\alpha}) \textrm{N}(y\,|\, \mu_s, \sigma^2), \end{align*} with $\bm{\beta}^s = (\beta^s_{\textrm{Intercept}}, \beta^{c(s)}_{\textrm{x1}}, \beta_{\textrm{x2}})$. The nesting signifies that $c(1) = c(2)$ and $\beta^{c(3)}_{\textrm{x1}} = 0$. The mixture model is fitted by first loading the package and the dataset and then specifying the component specific model. In a first step a component specific model with only varying effects is specified. Then the fitting function \code{flexmix()} is called repeatedly using \code{stepFlexmix()}. Finally, we order the components such that they are in ascending order with respect to the coefficients of the variable \code{x1}. <>= set.seed(2807) library("flexmix") data("NregFix", package = "flexmix") Model <- FLXMRglm(~ x2 + x1) fittedModel <- stepFlexmix(y ~ 1, model = Model, nrep = 3, k = 3, data = NregFix, concomitant = FLXPmultinom(~ w)) fittedModel <- relabel(fittedModel, "model", "x1") summary(refit(fittedModel)) @ The estimated coefficients indicate that the components differ for the intercept, but that they are not significantly different for the coefficients of $x2$. For $x1$ the coefficient of the first component is not significantly different from zero and the confidence intervals for the other two components overlap. Therefore we fit a modified model, which is equivalent to the true underlying model. The previously fitted model is used for initializing the EM algorithm: <>= Model2 <- FLXMRglmfix(fixed = ~ x2, nested = list(k = c(1, 2), formula = c(~ 0, ~ x1)), varFix = TRUE) fittedModel2 <- flexmix(y ~ 1, model = Model2, cluster = posterior(fittedModel), data = NregFix, concomitant = FLXPmultinom(~ w)) BIC(fittedModel) BIC(fittedModel2) @ The BIC suggests that the restricted model should be preferred. \begin{figure}[tb] \centering \setkeys{Gin}{width=0.95\textwidth} <>= plotNregFix <- NregFix plotNregFix$w <- factor(NregFix$w, levels = 0:1, labels = paste("w =", 0:1)) plotNregFix$x2 <- factor(NregFix$x2, levels = 0:1, labels = paste("x2 =", 0:1)) plotNregFix$class <- factor(NregFix$class, levels = 1:3, labels = paste("Class", 1:3)) print(xyplot(y ~ x1 | x2*w, groups = class, data = plotNregFix, cex = 0.6, auto.key = list(space="right"), layout = c(2,2))) @ \setkeys{Gin}{width=0.8\textwidth} \caption{Sample with 200 observations from the artificial example.} \label{fig:artificialData} \end{figure} <>= summary(refit(fittedModel2)) @ The coefficients are ordered such that the fixed coefficients are first, the nested varying coefficients second and the varying coefficients last. %%------------------------------------------------------------------------- \subsection{Finite mixtures of binomial logit regressions} %%------------------------------------------------------------------------- \subsubsection{Beta blockers} The dataset is analyzed in \cite{flexmix:Aitkin:1999, flexmix:Aitkin:1999a} using a finite mixture of binomial regression models. Furthermore, it is described in \cite{flexmix:McLachlan+Peel:2000} on page 165. The dataset is from a 22-center clinical trial of beta-blockers for reducing mortality after myocardial infarction. A two-level model is assumed to represent the data, where centers are at the upper level and patients at the lower level. The data is illustrated in Figure~\ref{fig:beta} and the model is given by: \begin{align*} H(\textrm{Deaths} \,|\, \textrm{Total}, \textrm{Treatment}, \textrm{Center}, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s \textrm{Bi}( \textrm{Deaths} \,|\, \textrm{Total}, \theta_s). \end{align*} First, the center classification is ignored and a binomial logit regression model with treatment as covariate is fitted using \code{glm}, i.e.~$S=1$: <>= data("betablocker", package = "flexmix") betaGlm <- glm(cbind(Deaths, Total - Deaths) ~ Treatment, family = "binomial", data = betablocker) betaGlm @ In the next step the center classification is included by allowing a random effect for the intercept given the centers, i.e.~the coefficients $\bm{\beta}^s$ are given by $(\beta^s_{\textrm{Intercept|Center}}, \beta_{\textrm{Treatment}})$. This signifies that the component membership is fixed for each center. In order to determine the suitable number of components, the mixture is fitted with different numbers of components and the BIC information criterion is used to select an appropriate model. In this case a model with three components is selected. The fitted values for the model with three components are given in Figure~\ref{fig:beta}. <>= betaMixFix <- stepFlexmix(cbind(Deaths, Total - Deaths) ~ 1 | Center, model = FLXMRglmfix(family = "binomial", fixed = ~ Treatment), k = 2:4, nrep = 3, data = betablocker) betaMixFix @ \begin{figure} \centering <>= library("grid") betaMixFix_3 <- getModel(betaMixFix, "3") betaMixFix_3 <- relabel(betaMixFix_3, "model", "Intercept") betablocker$Center <- with(betablocker, factor(Center, levels = Center[order((Deaths/Total)[1:22])])) clusters <- factor(clusters(betaMixFix_3), labels = paste("Cluster", 1:3)) print(dotplot(Deaths/Total ~ Center | clusters, groups = Treatment, as.table = TRUE, data = betablocker, xlab = "Center", layout = c(3, 1), scales = list(x = list(draw = FALSE)), key = simpleKey(levels(betablocker$Treatment), lines = TRUE, corner = c(1,0)))) betaMixFix.fitted <- fitted(betaMixFix_3) for (i in 1:3) { seekViewport(trellis.vpname("panel", i, 1)) grid.lines(unit(1:22, "native"), unit(betaMixFix.fitted[1:22, i], "native"), gp = gpar(lty = 1)) grid.lines(unit(1:22, "native"), unit(betaMixFix.fitted[23:44, i], "native"), gp = gpar(lty = 2)) } @ \caption{Relative number of deaths for the treatment and the control group for each center in the beta blocker dataset. The centers are sorted by the relative number of deaths in the control group. The lines indicate the fitted values for each component of the 3-component mixture model with random intercept and fixed effect for treatment.} \label{fig:beta} \end{figure} In addition the treatment effect can also be included in the random part of the model. As then all coefficients for the covariates and the intercept follow a mixture distribution the component specific model can be specified using \code{FLXMRglm()}. The coefficients are $\bm{\beta}^s=(\beta^s_{\textrm{Intercept|Center}}, \beta^s_{\textrm{Treatment|Center}})$, i.e.~it is assumed that the heterogeneity is only between centers and therefore the aggregated data for each center can be used. <>= betaMix <- stepFlexmix(cbind(Deaths, Total - Deaths) ~ Treatment | Center, model = FLXMRglm(family = "binomial"), k = 3, nrep = 3, data = betablocker) summary(betaMix) @ The full model with a random effect for treatment has a higher BIC and therefore the smaller would be preferred. The default plot of the returned \code{flexmix} object is a rootogramm of the a-posteriori probabilities where observations with a-posteriori probabilities smaller than \code{eps} are omitted. With argument \code{mark} the component is specified to have those observations marked which are assigned to this component based on the maximum a-posteriori probabilities. This indicates which components overlap. <>= print(plot(betaMixFix_3, mark = 1, col = "grey", markcol = 1)) @ The default plot of the fitted model indicates that the components are well separated. In addition component 1 has a slight overlap with component 2 but none with component 3. The fitted parameters of the component specific models can be accessed with: <>= parameters(betaMix) @ The cluster assignments using the maximum a-posteriori probabilities are obtained with: <>= table(clusters(betaMix)) @ The estimated probabilities for each component for the treated patients and those in the control group can be obtained with: <>= predict(betaMix, newdata = data.frame(Treatment = c("Control", "Treated"))) @ or <>= fitted(betaMix)[c(1, 23), ] @ A further analysis of the model is possible with function \code{refit()} which returns the estimated coefficients together with the standard deviations, z-values and corresponding p-values: <>= summary(refit(getModel(betaMixFix, "3"))) @ The printed coefficients are ordered to have the fixed effects before the varying effects. %%----------------------------------------------------------------------- \subsubsection{Mehta et al. trial} This dataset is similar to the beta blocker dataset and is also analyzed in \cite{flexmix:Aitkin:1999a}. The dataset is visualized in Figure~\ref{fig:mehta}. The observation for the control group in center 15 is slightly conspicuous and might classify as an outlier. The model is given by: \begin{align*} H(\textrm{Response} \,|\, \textrm{Total}, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s \textrm{Bi}( \textrm{Response} \,|\, \textrm{Total}, \theta_s), \end{align*} with $\bm{\beta}^s = (\beta^s_{\textrm{Intercept|Site}}, \beta_{\textrm{Drug}})$. This model is fitted with: <>= data("Mehta", package = "flexmix") mehtaMix <- stepFlexmix(cbind(Response, Total - Response)~ 1 | Site, model = FLXMRglmfix(family = "binomial", fixed = ~ Drug), control = list(minprior = 0.04), nrep = 3, k = 3, data = Mehta) summary(mehtaMix) @ One component only contains the observations for center 15 and in order to be able to fit a mixture with such a small component it is necessary to modify the default argument for \code{minprior} which is 0.05. The fitted values for this model are given separately for each component in Figure~\ref{fig:mehta}. \begin{figure} \centering <>= Mehta$Site <- with(Mehta, factor(Site, levels = Site[order((Response/Total)[1:22])])) clusters <- factor(clusters(mehtaMix), labels = paste("Cluster", 1:3)) print(dotplot(Response/Total ~ Site | clusters, groups = Drug, layout = c(3,1), data = Mehta, xlab = "Site", scales = list(x = list(draw = FALSE)), key = simpleKey(levels(Mehta$Drug), lines = TRUE, corner = c(1,0)))) mehtaMix.fitted <- fitted(mehtaMix) for (i in 1:3) { seekViewport(trellis.vpname("panel", i, 1)) sapply(1:nlevels(Mehta$Drug), function(j) grid.lines(unit(1:22, "native"), unit(mehtaMix.fitted[Mehta$Drug == levels(Mehta$Drug)[j], i], "native"), gp = gpar(lty = j))) } @ \caption{Relative number of responses for the treatment and the control group for each site in the Mehta et al.~trial dataset together with the fitted values. The sites are sorted by the relative number of responses in the control group.} \label{fig:mehta} \end{figure} If also a random effect for the coefficient of $\textrm{Drug}$ is fitted, i.e.~$\bm{\beta}^s = (\beta^s_{\textrm{Intercept|Site}}, \beta^s_{\textrm{Drug|Site}})$, this is estimated by: <>= mehtaMix <- stepFlexmix(cbind(Response, Total - Response) ~ Drug | Site, model = FLXMRglm(family = "binomial"), k = 3, data = Mehta, nrep = 3, control = list(minprior = 0.04)) summary(mehtaMix) @ The BIC is smaller for the larger model and this indicates that the assumption of an equal drug effect for all centers is not confirmed by the data. Given Figure~\ref{fig:mehta} a two-component model with fixed treatment is also fitted to the data where site 15 is omitted: <>= Mehta.sub <- subset(Mehta, Site != 15) mehtaMix <- stepFlexmix(cbind(Response, Total - Response) ~ 1 | Site, model = FLXMRglmfix(family = "binomial", fixed = ~ Drug), data = Mehta.sub, k = 2, nrep = 3) summary(mehtaMix) @ %%----------------------------------------------------------------------- \subsubsection{Tribolium} A finite mixture of binomial regressions is fitted to the Tribolium dataset given in \cite{flexmix:Wang+Puterman:1998}. The data was collected to investigate whether the adult Tribolium species Castaneum has developed an evolutionary advantage to recognize and avoid eggs of its own species while foraging, as beetles of the genus Tribolium are cannibalistic in the sense that adults eat the eggs of their own species as well as those of closely related species. The experiment isolated a number of adult beetles of the same species and presented them with a vial of 150 eggs (50 of each type), the eggs being thoroughly mixed to ensure uniformity throughout the vial. The data gives the consumption data for adult Castaneum species. It reports the number of Castaneum, Confusum and Madens eggs, respectively, that remain uneaten after two day exposure to the adult beetles. Replicates 1, 2, and 3 correspond to different occasions on which the experiment was conducted. The data is visualized in Figure~\ref{fig:tribolium} and the model is given by: \begin{align*} H(\textrm{Remaining} \,|\, \textrm{Total}, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s(\textrm{Replicate}, \bm{\alpha}) \textrm{Bi}( \textrm{Remaining} \,|\, \textrm{Total}, \theta_s), \end{align*} with $\bm{\beta}^s = (\beta^s_{\textrm{Intercept}}, \bm{\beta}_{\textrm{Species}})$. This model is fitted with: <>= data("tribolium", package = "flexmix") TribMix <- stepFlexmix(cbind(Remaining, Total - Remaining) ~ 1, k = 2:3, model = FLXMRglmfix(fixed = ~ Species, family = "binomial"), concomitant = FLXPmultinom(~ Replicate), data = tribolium) @ The model which is selected as the best in \cite{flexmix:Wang+Puterman:1998} can be estimated with: <>= modelWang <- FLXMRglmfix(fixed = ~ I(Species == "Confusum"), family = "binomial") concomitantWang <- FLXPmultinom(~ I(Replicate == 3)) TribMixWang <- stepFlexmix(cbind(Remaining, Total - Remaining) ~ 1, data = tribolium, model = modelWang, concomitant = concomitantWang, k = 2) summary(refit(TribMixWang)) @ \begin{figure} \centering <>= clusters <- factor(clusters(TribMixWang), labels = paste("Cluster", 1:TribMixWang@k)) print(dotplot(Remaining/Total ~ factor(Replicate) | clusters, groups = Species, data = tribolium[rep(1:9, each = 3) + c(0:2)*9,], xlab = "Replicate", auto.key = list(corner = c(1,0)))) @ \caption{Relative number of remaining beetles for the number of replicate. The different panels are according to the cluster assignemnts based on the a-posteriori probabilities of the model suggested in \cite{flexmix:Wang+Puterman:1998}.} \label{fig:tribolium} \end{figure} \cite{flexmix:Wang+Puterman:1998} also considered a model where they omit one conspicuous observation. This model can be estimated with: <>= TribMixWangSub <- stepFlexmix(cbind(Remaining, Total - Remaining) ~ 1, k = 2, data = tribolium[-7,], model = modelWang, concomitant = concomitantWang) @ %%----------------------------------------------------------------------- \subsubsection{Trypanosome} The data is used in \cite{flexmix:Follmann+Lambert:1989}. It is from a dosage-response analysis where the proportion of organisms belonging to different populations shall be assessed. It is assumed that organisms belonging to different populations are indistinguishable other than in terms of their reaction to the stimulus. The experimental technique involved inspection under the microscope of a representative aliquot of a suspension, all organisms appearing within two fields of view being classified either alive or dead. Hence the total numbers of organisms present at each dose and the number showing the quantal response were both random variables. The data is illustrated in Figure~\ref{fig:trypanosome}. The model which is proposed in \cite{flexmix:Follmann+Lambert:1989} is given by: \begin{align*} H(\textrm{Dead} \,|\,\bm{\Theta}) &= \sum_{s = 1}^S \pi_s \textrm{Bi}( \textrm{Dead} \,|\, \theta_s), \end{align*} where $\textrm{Dead} \in \{0,1\}$ and with $\bm{\beta}^s = (\beta^s_{\textrm{Intercept}}, \bm{\beta}_{\textrm{log(Dose)}})$. This model is fitted with: <>= data("trypanosome", package = "flexmix") TrypMix <- stepFlexmix(cbind(Dead, 1-Dead) ~ 1, k = 2, nrep = 3, data = trypanosome, model = FLXMRglmfix(family = "binomial", fixed = ~ log(Dose))) summary(refit(TrypMix)) @ The fitted values are given in Figure~\ref{fig:trypanosome} together with the fitted values of a generalized linear model in order to facilitate comparison of the two models. \begin{figure} \centering <>= tab <- with(trypanosome, table(Dead, Dose)) Tryp.dat <- data.frame(Dead = tab["1",], Alive = tab["0",], Dose = as.numeric(colnames(tab))) plot(Dead/(Dead+Alive) ~ Dose, data = Tryp.dat) Tryp.pred <- predict(glm(cbind(Dead, 1-Dead) ~ log(Dose), family = "binomial", data = trypanosome), newdata=Tryp.dat, type = "response") TrypMix.pred <- predict(TrypMix, newdata = Tryp.dat, aggregate = TRUE)[[1]] lines(Tryp.dat$Dose, Tryp.pred, lty = 2) lines(Tryp.dat$Dose, TrypMix.pred, lty = 3) legend(4.7, 1, c("GLM", "Mixture model"), lty=c(2, 3), xjust=0, yjust=1) @ \caption{Relative number of deaths for each dose level together with the fitted values for the generalized linear model (``GLM'') and the random intercept model (``Mixture model'').} \label{fig:trypanosome} \end{figure} %%------------------------------------------------------------------------- \subsection{Finite mixtures of Poisson regressions} % %%----------------------------------------------------------------------- \subsubsection{Fabric faults} The dataset is analyzed using a finite mixture of Poisson regression models in \cite{flexmix:Aitkin:1996}. Furthermore, it is described in \cite{flexmix:McLachlan+Peel:2000} on page 155. It contains 32 observations on the number of faults in rolls of a textile fabric. A random intercept model is used where a fixed effect is assumed for the logarithm of length: <>= data("fabricfault", package = "flexmix") fabricMix <- stepFlexmix(Faults ~ 1, model = FLXMRglmfix(family="poisson", fixed = ~ log(Length)), data = fabricfault, k = 2, nrep = 3) summary(fabricMix) summary(refit(fabricMix)) Lnew <- seq(0, 1000, by = 50) fabricMix.pred <- predict(fabricMix, newdata = data.frame(Length = Lnew)) @ The intercept of the first component is not significantly different from zero for a signficance level of 0.05. We therefore also fit a modified model where the intercept is a-priori set to zero for the first component. This nested structure is given as part of the model specification with argument \code{nested}. <>= fabricMix2 <- flexmix(Faults ~ 0, data = fabricfault, cluster = posterior(fabricMix), model = FLXMRglmfix(family = "poisson", fixed = ~ log(Length), nested = list(k=c(1,1), formula=list(~0,~1)))) summary(refit(fabricMix2)) fabricMix2.pred <- predict(fabricMix2, newdata = data.frame(Length = Lnew)) @ The data and the fitted values for each of the components for both models are given in Figure~\ref{fig:fabric}. \begin{figure} \centering <>= plot(Faults ~ Length, data = fabricfault) sapply(fabricMix.pred, function(y) lines(Lnew, y, lty = 1)) sapply(fabricMix2.pred, function(y) lines(Lnew, y, lty = 2)) legend(190, 25, paste("Model", 1:2), lty=c(1, 2), xjust=0, yjust=1) @ \caption{Observed values of the fabric faults dataset together with the fitted values for the components of each of the two fitted models.} \label{fig:fabric} \end{figure} %%----------------------------------------------------------------------- \subsubsection{Patent} The patent data given in \cite{flexmix:Wang+Cockburn+Puterman:1998} consist of 70 observations on patent applications, R\&D spending and sales in millions of dollar from pharmaceutical and biomedical companies in 1976 taken from the National Bureau of Economic Research R\&D Masterfile. The observations are displayed in Figure~\ref{fig:patent}. The model which is chosen as the best in \cite{flexmix:Wang+Cockburn+Puterman:1998} is given by: \begin{align*} H(\textrm{Patents} \,|\, \textrm{lgRD}, \textrm{RDS}, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s(\textrm{RDS}, \bm{\alpha}) \textrm{Poi} ( \textrm{Patents} \,|\, \lambda_s), \end{align*} and $\bm{\beta}^s = (\beta^s_{\textrm{Intercept}}, \beta^s_{\textrm{lgRD}})$. The model is fitted with: <>= data("patent", package = "flexmix") ModelPat <- FLXMRglm(family = "poisson") FittedPat <- stepFlexmix(Patents ~ lgRD, k = 3, nrep = 3, model = ModelPat, data = patent, concomitant = FLXPmultinom(~ RDS)) summary(FittedPat) @ The fitted values for the component specific models and the concomitant variable model are given in Figure~\ref{fig:patent}. The plotting symbol of the observations corresponds to the induced clustering given by \code{clusters(FittedPat)}. This model is modified to have fixed effects for the logarithmized R\&D spendings, i.e.~$\bm(\beta)^s = (\beta^s_{\textrm{Intercept}}, \beta_{\textrm{lgRD}})$. The already fitted model is used for initialization, i.e.~the EM algorithm is started with an M-step given the a-posteriori probabilities. <>= ModelFixed <- FLXMRglmfix(family = "poisson", fixed = ~ lgRD) FittedPatFixed <- flexmix(Patents ~ 1, model = ModelFixed, cluster = posterior(FittedPat), concomitant = FLXPmultinom(~ RDS), data = patent) summary(FittedPatFixed) @ The fitted values for the component specific models and the concomitant variable model of this model are also given in Figure~\ref{fig:patent}. \begin{figure} \centering \setkeys{Gin}{width=0.95\textwidth} <>= lgRDv <- seq(-3, 5, by = 0.05) newdata <- data.frame(lgRD = lgRDv) plotData <- function(fitted) { with(patent, data.frame(Patents = c(Patents, unlist(predict(fitted, newdata = newdata))), lgRD = c(lgRD, rep(lgRDv, 3)), class = c(clusters(fitted), rep(1:3, each = nrow(newdata))), type = rep(c("data", "fit"), c(nrow(patent), nrow(newdata)*3)))) } plotPatents <- cbind(plotData(FittedPat), which = "Wang et al.") plotPatentsFixed <- cbind(plotData(FittedPatFixed), which = "Fixed effects") plotP <- rbind(plotPatents, plotPatentsFixed) rds <- seq(0, 3, by = 0.02) x <- model.matrix(FittedPat@concomitant@formula, data = data.frame(RDS = rds)) plotConc <- function(fitted) { E <- exp(x%*%fitted@concomitant@coef) data.frame(Probability = as.vector(E/rowSums(E)), class = rep(1:3, each = nrow(x)), RDS = rep(rds, 3)) } plotConc1 <- cbind(plotConc(FittedPat), which = "Wang et al.") plotConc2 <- cbind(plotConc(FittedPatFixed), which = "Fixed effects") plotC <- rbind(plotConc1, plotConc2) print(xyplot(Patents ~ lgRD | which, data = plotP, groups=class, xlab = "log(R&D)", panel = "panel.superpose", type = plotP$type, panel.groups = function(x, y, type = "p", subscripts, ...) { ind <- plotP$type[subscripts] == "data" panel.xyplot(x[ind], y[ind], ...) panel.xyplot(x[!ind], y[!ind], type = "l", ...) }, scales = list(alternating=FALSE), layout=c(1,2), as.table=TRUE), more=TRUE, position=c(0,0,0.6, 1)) print(xyplot(Probability ~ RDS | which, groups = class, data = plotC, type = "l", scales = list(alternating=FALSE), layout=c(1,2), as.table=TRUE), position=c(0.6, 0.01, 1, 0.99)) @ \caption{Patent data with the fitted values of the component specific models (left) and the concomitant variable model (right) for the model in \citeauthor{flexmix:Wang+Cockburn+Puterman:1998} and with fixed effects for $\log(\textrm{R\&D})$. The plotting symbol for each observation is determined by the component with the maximum a-posteriori probability.} \label{fig:patent} \end{figure} \setkeys{Gin}{width=0.8\textwidth} With respect to the BIC the full model is better than the model with the fixed effects. However, fixed effects have the advantage that the different components differ only in their baseline and the relation between the components in return of investment for each additional unit of R\&D spending is constant. Due to a-priori domain knowledge this model might seem more plausible. The fitted values for the constrained model are also given in Figure~\ref{fig:patent}. %%----------------------------------------------------------------------- \subsubsection{Seizure} The data is used in \cite{flexmix:Wang+Puterman+Cockburn:1996} and is from a clinical trial where the effect of intravenous gamma-globulin on suppression of epileptic seizures is studied. There are daily observations for a period of 140 days on one patient, where the first 27 days are a baseline period without treatment, the remaining 113 days are the treatment period. The model proposed in \cite{flexmix:Wang+Puterman+Cockburn:1996} is given by: \begin{align*} H(\textrm{Seizures} \,|\, (\textrm{Treatment}, \textrm{log(Day)}, \textrm{log(Hours)}), \bm{\Theta}) &= \sum_{s = 1}^S \pi_s \textrm{Poi} ( \textrm{Seizures} \,|\, \lambda_s), \end{align*} where $\bm(\beta)^s = (\beta^s_{\textrm{Intercept}}, \beta^s_{\textrm{Treatment}}, \beta^s_{\textrm{log(Day)}}, \beta^s_{\textrm{Treatment:log(Day)}})$ and $\textrm{log(Hours)}$ is used as offset. This model is fitted with: <>= data("seizure", package = "flexmix") seizMix <- stepFlexmix(Seizures ~ Treatment * log(Day), data = seizure, k = 2, nrep = 3, model = FLXMRglm(family = "poisson", offset = log(seizure$Hours))) summary(seizMix) summary(refit(seizMix)) @ A different model with different contrasts to directly estimate the coefficients for the jump when changing between base and treatment period is given by: <>= seizMix2 <- flexmix(Seizures ~ Treatment * log(Day/27), data = seizure, cluster = posterior(seizMix), model = FLXMRglm(family = "poisson", offset = log(seizure$Hours))) summary(seizMix2) summary(refit(seizMix2)) @ A different model which allows no jump at the change between base and treatment period is fitted with: <>= seizMix3 <- flexmix(Seizures ~ log(Day/27)/Treatment, data = seizure, cluster = posterior(seizMix), model = FLXMRglm(family = "poisson", offset = log(seizure$Hours))) summary(seizMix3) summary(refit(seizMix3)) @ With respect to the BIC criterion the smaller model with no jump is preferred. This is also the more intuitive model from a practitioner's point of view, as it does not seem to be plausible that starting the treatment already gives a significant improvement, but improvement develops over time. The data points together with the fitted values for each component of the two models are given in Figure~\ref{fig:seizure}. It can clearly be seen that the fitted values are nearly equal which also supports the smaller model. \begin{figure} \centering <>= plot(Seizures/Hours~Day, pch = c(1,3)[as.integer(Treatment)], data=seizure) abline(v=27.5, lty=2, col="grey") legend(140, 9, c("Baseline", "Treatment"), pch=c(1, 3), xjust=1, yjust=1) matplot(seizure$Day, fitted(seizMix)/seizure$Hours, type="l", add=TRUE, lty = 1, col = 1) matplot(seizure$Day, fitted(seizMix3)/seizure$Hours, type="l", add=TRUE, lty = 3, col = 1) legend(140, 7, paste("Model", c(1,3)), lty=c(1, 3), xjust=1, yjust=1) @ \caption{Observed values for the seizure dataset together with the fitted values for the components of the two different models.} \label{fig:seizure} \end{figure} %%----------------------------------------------------------------------- \subsubsection{Ames salmonella assay data} The ames salomnella assay dataset was used in \cite{flexmix:Wang+Puterman+Cockburn:1996}. They propose a model given by: \begin{align*} H(\textrm{y} \,|\, \textrm{x}, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s \textrm{Poi} ( \textrm{y} \,|\, \lambda_s), \end{align*} where $\bm{\beta}^s = (\beta^s_{\textrm{Intercept}}, \beta_{\textrm{x}}, \beta_{\textrm{log(x+10)}})$. The model is fitted with: <>= data("salmonellaTA98", package = "flexmix") salmonMix <- stepFlexmix(y ~ 1, data = salmonellaTA98, k = 2, nrep = 3, model = FLXMRglmfix(family = "poisson", fixed = ~ x + log(x + 10))) @ \begin{figure} \centering <>= salmonMix.pr <- predict(salmonMix, newdata=salmonellaTA98) plot(y~x, data=salmonellaTA98, pch=as.character(clusters(salmonMix)), xlab="Dose of quinoline", ylab="Number of revertant colonies of salmonella", ylim=range(c(salmonellaTA98$y, unlist(salmonMix.pr)))) for (i in 1:2) lines(salmonellaTA98$x, salmonMix.pr[[i]], lty=i) @ \caption{Means and classification for assay data according to the estimated posterior probabilities based on the fitted model.} \label{fig:almes} \end{figure} %%----------------------------------------------------------------------- \section{Conclusions and future work} Package \pkg{flexmix} can be used to fit finite mixtures of regressions to datasets used in the literature to illustrate these models. The results can be reproduced and additional insights can be gained using visualization methods available in \proglang{R}. The fitted model is an object in \proglang{R} which can be explored using \code{show()}, \code{summary()} or \code{plot()}, as suitable methods have been implemented for objects of class \code{"flexmix"} which are returned by \code{flexmix()}. In the future it would be desirable to have more diagnostic tools available to analyze the model fit and compare different models. The use of resampling methods would be convenient as they can be applied to all kinds of mixtures models and would therefore suit well the purpose of the package which is flexible modelling of various finite mixture models. Furthermore, an additional visualization method for the fitted coefficients of the mixture would facilitate the comparison of the components. %%----------------------------------------------------------------------- \section*{Computational details} <>= SI <- sessionInfo() pkgs <- paste(sapply(c(SI$otherPkgs, SI$loadedOnly), function(x) paste("\\\\pkg{", x$Package, "} ", x$Version, sep = "")), collapse = ", ") @ All computations and graphics in this paper have been done using \proglang{R} version \Sexpr{getRversion()} with the packages \Sexpr{pkgs}. %%----------------------------------------------------------------------- \section*{Acknowledgments} This research was supported by the the Austrian Science Foundation (FWF) under grant P17382 and the Austrian Academy of Sciences ({\"O}AW) through a DOC-FFORTE scholarship for Bettina Gr{\"u}n. %%----------------------------------------------------------------------- \bibliography{flexmix} \end{document} flexmix/inst/doc/flexmix-intro.pdf0000644000176200001440000062456113432516317016773 0ustar liggesusers%PDF-1.5 %¿÷¢þ 1 0 obj << /Type /ObjStm /Length 4143 /Filter /FlateDecode /N 72 /First 593 >> stream xœ½[YsÛ8~ß_·ÍÔΉƒÜÚš*qâÄN<²sÍVh‰¶¹£ÃCR‰gý~ /ÙT¢Ú’e$Øh4úën4 Ÿy,`Ú0Éd ™bÚL3£3ÌDš…,Œ<1áI‰Lˆ(Ä?&å3?åiTpU† É|)æ4óÃP1aX <Y†¸1é{3©CÃ|Á”gó}¦ü ÀËLI‰ûàÈ(±|¡®™Æc¾a:0!ˆƒuƒN#f¼ bÇŒ/5 3:ôYàc QÀ@2 4îKjƒºb‘Gí4‹üP2°)…û!‹BôÐx=P•‹çcd4Oa|Cö ‘ +—C“¨ðÐ-( @H %„Ä%ÉÌxnD4V$Ê0ô˜eßC+OtÂÀˆð}I2@Á@9 FÉÚ€]+lrV!0@ʘBH”¥¢iT€¬5(K ùjP–R×’ ƒ² Á3¤ƒ©DС|Ì€e€gQA’ŒQH ¥Á¹!0’! }ó·ý‹ñó¤ˆgq3LÇ&Œ¿[ót™äÐ,[¿ˆoQ)^ýuŸ0~„öóÕ-ûõWKâ`]Ü­2öì$K“Y–NïØY’æÓ»ŸÐ2Kâ"]-ã"aÏŽÿ ‹<_„ÂWÿðÄß=ïïU;¢q_%ŸØ·´¸cwè+Ë’<~“üõm•Írölò3»I—)¨-Ò‡bẚ%óügwe×qžÌØt¾Î‹$K—·?³9º^¸ç9Ë’Û,Éspªç«ÙS|]d«Ùzš€±—gìåÝ*/òi–Þ,zîk4¸\_ÿ'™ìÊWi1µ“yòpž>ü“°—É2Éâ9;ÉâE‚üÁn0È7€órçv,^ÎØ™ãõÈò:©yeé’M~ªÄ}´Z£ Ô’¿I!’ÃÐÜ„ödÿÖö"ìEU7ƒòêê¡kº»aIŸ‹k9Âwº•Y^]ã/•^8U©Ù\Ò`rÒY§<çÉ,W`™ºV‘z£JñÝ}¡×3}©j“$_­³)µÅËË‚f+ròЇ(•4[M/“”ùÅñ øI üõ×&km*H³¼`%½³åÈu^6¡.²æû?Qˆ/|'xÖî·n'üÖ{¡UÎÐÞ54ŒgXÒh¤m ÊÂÓŽ=Ä/Ô<ˆJ¶C+!´?À“‚O¦fм5]@ 5Õ/Š"!”H¶UùÀœK´—Õ´$NéJß/µ[Χ¤_ÉÀcß¿JÒÛ»ªŠÉ ýyÆøÁOøK~ÊÏø9¿à~Éßó<æ×|ʧ«ùjÉg<á7üæ&åöï+jPV~Ëï8¼ô]²ä)ÿƒÏù‚/ùН– ¿çpÞéjÆ3žó‚wY’ðâÛŠ¯ùWþ?ð¿~²öÕ²ø Ü/4`ßR aUá°Ž}†-²œqOOÒyBAL°ÁÆ[xØžºž"NI§Ë[a û<…;]ÞZ%³òç—E²øÀ(Lm¨aCƒ{xxøæìõ :˜ÀÇnà`*N6p}8èÌH8(xi…È’¾a *; hèÞ¦½o Åôö ­EÍÞ¡ Wj;öë—ï:*Žº³0À‚•Ñ“BÁa €CÏĸ¤ìô)n$´{‡ê61ˆñi„–‚rÒ›º¢mDãêþ׌ã5)e} ÐyÀ&v*——t|ÜÍÒ!‡â© X"6= fÈ>¤w4¾¶t[mobë,”¶£ ë)taÉß_¹½:yÿö5ú»¼l¡ 8X,ótscƒ³ÐëãÌïà ‘è.®FKµî¦!VZ3áW¡•2ºVΧdóÐ<˜îõÚÑ•FzÊéJg.ÇúC>ذ¯9¿}xyñ›ÕÌÓñpÝuf8::”} vgÇVëÙ™4$ï=QÏ BºBïŠa¬ÔÉVƒèç««ÏW§æ´ãŽÊVl⎠/Ü~ì=—öõZðïë#CÝÄ;íÇuVÔ滫iŠŒÃ}´Y_#ÿñÜT Ô—Aó|ãßÓÚ­ß×øÿÚ¶OÅÃã÷Ãaíu5±­'cõâÐÑÑ0–²¹.í ZB»d4es7ºYÙ¦†nêÖØGÒ¨¬‡uYjGýÜ«¢Ñºô¥©ÁD©¥ÁPû«Õû%–㳄¹4æd‹RKõ¼þm0öö÷Ëi)qÞ îuo } ˆȆLXƒÅà•V¥.KÏ!Pêjeª½2;lÜÿMÔ®Õ¦<þ³IŒ‘J˜–'ý¶_¦,3%¾"ôzÃI½­øÄ˜¤Ý¥SvçIRÜn×D!­&Ê Y—&OFîíÝ’WÇuªê:)â*Oæ(#&¿ÏS8g¤ÞÒFn’§”GZðÅP­cïû”£)–µ·x»Ìíܽþ’vk*Ë‘2•#å@*G÷R9M½¨Õ£s9Þ¼úhÖçvR¸2(µª÷Ãë èjº×ã6wbjU{ìÓÈÏ–%3ðNÖûÙ·/ëP÷6Þ†¸›®ÆY¶úæòŠ×qƯ³xê޶ä$®Ó¯ñ¡il—8žO–vëJ;t›š¤Œ‹ÝiŠhƒOÊ-oIÚvãÙŽodצ’E5m¥³‡m[(Ô[(¤{*¼0´Ô;:ÖKB´U`¬Šé`´)|ûêèìô_tÒÚ•¶×: xýÆI ·n‰ÆšB¬4Mô=þzüë ,—;Ó/(ªC&J¶,”l¨zå0½½ml¢¸ª;)2KóûyüWU-’‡‚çkx]:CR=¬oÐã–ÊDXWm5JôPtÔ­g•zé’ÎŒŽTßÛ¶î憎¿œP“pä’WF}Õ‰:ª£Ç4Pˆ5•7Ýÿêã—V+ôuIlTkIl«õìºDE¹ûF«J?ßV;¬”\"Ú™›^z¢%º±`Û–'½™¹<8»º¸°©å/ª³2„5²ÀSJä±ø½œ¿Á) ý]rTßóña>»“L›KIvÕz’Ë0!KÉмv1G{ÛÆÒhÏi"ÖsÚúèI}âÃ$°€gÛ&ˆ ¤!ƒ|3à5xózµƒøð“!÷ú>仃øvÐŽƒk ¬?ŠÓ.B->·£3Š:u<÷X{ƒðì¥a{iŸÔÙ¢Û6† ºËm×Э+hVšŸØS­mUý„¥ú1Ó~~ÂÒÜXeeÌ·ÛOX†¤Ó”ȈHj[ô´Yjn“ŸDš´ª,Ú,©*!*¢É×qzs“d´­@¦Îœ6Ö­#WÛ׆m~Äwèù ÐÛ/1ÊÝì(¥u[éMÖÉoÚ3f­¼%kæ8äÞËs²NRt ÝL+»—Éê­L»bsñtŽ/Ö¤Uü>¥\‡Û¿tCÍS{xË™1æ ¦]ÄÚèb_ŒÂªßÌ·û¿( ßâŽ]²òæž:€BØtël»Ñ²©¹QGf¯µYG‡äª½9ÚImìüQ€²Ù®£] ¶ \Y#ru®pzû e> stream GPL Ghostscript 9.26 R, finite mixture models, model based clustering, latent class regression 2019-02-18T12:53:51+01:00 2019-02-18T12:53:51+01:00 LaTeX with hyperref FlexMix: A General Framework for Finite Mixture Models and Latent Class Regression in RFriedrich Leisch endstream endobj 75 0 obj << /Type /ObjStm /Length 2199 /Filter /FlateDecode /N 72 /First 614 >> stream xœÕYmsÛ6þ~¿ÛÉD  n:ñKܦS§=9‰“vü–i›SIÔTêܯ¿gIB")Ê–lÝen4¶p‰Ý}ö»5,`Ö2¡Cf#&µeÖ1¥ ‹¦É"ÁŒŽX$A…™b‘ X2„ ÑL‹¹aBI̱Q¨@aCü‹VE ¯ˆÈ…Ì &lê$“ `N1)ax$C‰u $è “â8Ëd„Í$sÖ¾!Цq S²Z‘LCÅ”#Æ E ÍBȆ… ‰À²Ð‘ÐA%«GŽi!ÔÚ!˜IQ!™&$„PLz]„8¢ÑLG†V 3ToEÌÈ€6tHÐÈz £@( ± =R„ô(Ú’i Hx j `XÅ7ÑDÌ!XH‡©¬`B¡É‚YZ2³†ôR I4°¯Õ°ƒÒ°p@ÄI4ds2,‰9 …°­Ô4°=À!9ÂBÈ" JZˆ q#­iO0„D‰€Ø?~øñ“l9/ÙkÉøiR”ìOÁóÆŒúü˜ËŒÈDóåtÊ®?KsÐÁ7+²_cš4ï¼KJòÓjò{œ'ØÙ4³<ùšwÞ§å4aßý—D0™ÆEÁòä.OŠ"Íæß³¬$käq=yŒu#aÅJžšiÔajM‡VãYR&9Öé,.ØÀÁº|´p#è=ŸÞÖµB]ÞÞ€"½Mãëtš–_W,ŒÃ53Ñcf•2ê‰È‘iEÛAîêþ¡Hçwìlš<œ§-‰æ~<ШpžÜ¤ñqöé íôòD¡!h×ê‡5õ8)²e>I F»½y(º(aÎ*+‚30¨"²,›\$¤9ÿýô ²‘Wï‹„v¿Kz‰úAŽ$RDÇðÈ1ûŽòçËi™.0Jç7É"™“}àjÅ"›IÑw Ž*FAõ8†ƒ+ì½Áöcð*7l–ÄÅ2Ofù ¢ÃÚÀtl`_`ÕCDZˆ«v1 ";ˆÎy6eÙ-+ïöæœÅÓ»,OËûÙ.î°¸D\Ü pÑ¾ŽŽ‹~²è@v 8Z–Ù¬öŽì&™²"‰óÉý Þ-„‚ Ú(Hñ 6Ï ÛOžÒެèç3´ò™“­|æºAÕÉg݆&ˆãÇsšÃC!&;ˆ©øMÿ0“ÚBcºñäº9Íçú€KD ©ÏÙùë¢Lì&O¿$ùò°8„ôóqpýóO…räl?|V§ùwž•Ä3dö²ú€£l²uPTÇÔ üÁõók¿ˆ‡ž;\,g³8ÿ w¸aÙ²œfÙ_+Žã"©„äNßþöþòÕÉùÅÅñ'”:üÍ|’UQÄ/ÓùѼH× U¬žÜÇ9%ïJIH;ÉÓE™åÔ¬BxEs±¼.+IQáõ­9_¦7å}]-*Ï+/ÙQ1©ôBÜó“xñs’ÞÝû)v%¿ããï+ *Jˆ|6ï dZûõæÇkS­ª·¯ê'gé4Amµö*ÀÞ–ñ4Íï€%¦çiAîT M-t+“ÙGêxZZµðhôÃ'Dí«ИwŒâÞi?¦®Î©kÆ(åý‘4Fß$fïË…¬öŘÞkÎ ò¯•í¨²¯5^Ù¤m°Àm < eÖy\æiä(¨ Ó 5ºj™š¶ê˜Zá+û0OÁaáË€Aë‡vmüµ[þ/nO!øuŒ0m«ÑlµáX-nvƒÛÆöŸßœ|üé󫟓闤L'q‡…Ø@jÓÛ©}íº;˜»ù»¬ªº¡=ÑðÍšþ¼õ/hþo~lu(íó©¹u¥"_ó o¹Ú7‚weQ´´% ýõãÚÝV\×ÓU\Ÿðë<žü• Ý–~œ%ŸðI†|Æ^½Éo‘Ñù-Ò0¿ç)ŸñY:_|Î3žÍ¾à‹$O³žó‚é/ñ$áå}žàÿßÿ›?ð¯üßIžµRÉk)"ŸMÂ^&©ž¡3¯Dî%“VõX'“M÷z,›òM6¡¾èÑtÒ:º[–µŽ®`u¾îqt­ÃëÝۋ˧ký_gÓ›NŒùüûxŒé~Œ¡yiŒU„y¯n{øvʧ?OEíošÌIcŠ˜zî%¥ï¨2*­×Ïæžz#’¤ëDR5]Ÿ<æ×–Û:Fîø]N gŽX™®¢ÅÇJ$u€ìj?âÝ _’­Cb‹KìR=»ƒzÃñþgÇ ¯mŸ<jŠº ­ƒ˜¢•nÞ6ë†Ð1߆Œ†ù±eac“1Žcíi»´×EÕõl»¸Xew؆Á_s5 ·oköj5wµ¸ú†W³¸¨¯º,¾ª»ÖׂùFl Í”W;æÛ”±è¯¼Å*æ3ôÞòT©»‹Ð»ª˜2B·]üiöõ"ßÙ} /ò×C/÷"úaaÝcl3Î/úïxÂë?jµ=öï4R»„ hìeè4pm¾ƒ¼^¸ÿKÒA+ššõ¹ z.ÈÎE¬|ÎEìÎíK”oÑ»6ÄC]ý’h‡Ü eî­±j«ßb·ÿå—]¾ûðêäüíû÷;_ 9·YÄ«ÖOUMîXÄëÊc{…4ÖÚë4Þ¸<‚EøÑYŠöU’è´œ“ªÎÛ%¯ñ/d§üÙ«z¬uÕ6½š·ÕAÕ5oÃ~±?´$ì^0ÑJ]ú’Õ/ÿÔh óWýàüxÎú¾iß”¹™*û[–Ò¬•*Ôê"tT1”bwÁd0Î6õ=lƒ®:©Q='5>÷ØÜv| Öþ;½½M iòg•)_y±VÙ£•xI1ÝÝÐÄÕìöýödÿÿ%í^ìé.‡ÇV„û9´TAendstream endobj 148 0 obj << /Filter /FlateDecode /Length 6487 >> stream xœµ\K“Gr¾þ 8¶Þù$iåõîJa/IÇv}hr†à˜3Å!EŽ.þëÎG=² ŠŒ!C …¬¬|~™•_Vj§W ÿ+ÿ~{ñ‡ÇÖ¬öwjµ¿øåBÓ§«ò¿ç·«ïŸâ »Òqg´w«§/.ø«z•ÜÎF»ŠFï´Ë«§·ÿXÿûf ´sÎ>¯o6jçRÎѬ¯ð±Ç6¯?Àc£²Žiý3>Vð8渾Ƹœá¼F+ŸÃúÛÍî¼Òëï6[=¬Yÿ >*Ç”­RÁ¥õAþã ,ñÊ:çÖÓwQ1èLxÍf³ñ´Öe—­ÖÞŽt†mÔ.dëb\¿ßX|ÖGÚÁ(€@[ù 7p@7¬_l›`†µ[\¬‚BÎh…Wa×u§q¸&ÑÅÌ¿•LЭrMr&ª"¹Â KÎDcá›øÔ¥`ÖïÞˆ%BJÿóô/ÎìBôiµ=fÒêé%(rP po‰ý¼¾¬ª„ n„Æîê±…A#—üÖ¬¢ç GàGJù­ÍéUùׇÕ;ØÏA*óë6U7](L*&z¼Û 4}T¾ò“€åÇRVÒDö$ `’Ôk>GÀ °nS¡lR@=5±…ʘI¥µ\pØ „Á¼t kÃ.éND¢~Ü÷)«ÁÝ@Þ\½5*îtt«-jÊÍߪÖkÙz“ÍÉWKPÖ˜"Ë€¹Ë7â9³G>GSãsë—UO &XšA1ÁÇ8P¾#ÆûBa>BôYóÖϬKÛw©X×Ð5àÄ>0åFä׿nLÄØ@{°=Ä:Ø„c%ø¬Ÿ⃠‰Í,¸$k¶”`?”‰R‘6\ÿßæéÿ^`ØëÏþ±žÐ°¬Ù¥äÀU·uG¤¢Ã.;c–¨õ» ŽÔÞ¡À¶Á“ÑÁ­¦Ýú{²¦d œNiˆ—}7¶7m!  µÇ¯Ðà8²“”oº¬Ð»µŠÀåÉÖ`-ÿN&«ŒDLÛ¡õ{¹žùǧ»°»œVïATºHÀ?õ.šÕí…7¦ÿóæâÉù\‘Wy—ƒ 2U˜<Ç•Ïà’SÅwä‰9;g×ÏЉkʤe§²qžuk ¸HVn3&ŽõôœÂ&ØD´rM3ÀÎÀVû°CsÐ ´¥ ëê© û’è:ïØªt²bø¾ZÜ w6Ƀ΋Ф;–)è_kô#$Äøó¦Ó¸*4À¼ÞoLØÅÎ:e'„B¶ æá´-›·Þ‹¿ÉQRÄ@W•ð6YkÒ=Lë×D \4“E”C=kkׂÔ`ôŽ¢@¥À¨|Jh­x $Ó]WZ³.'ÐÇêÍGPD\° «Ìlkå#è)ñÓ•° ¤æì×bw9x“‡”,¸,}.p@•#ÿ\s,?üB‡µÜÇ»€ºg_âÁBš³#lŠu€J¥ø1ød±iǾ@¶r5êù[ÚH²SI`ÙP¦†”ydÇ$bhƒ)UÈ-†§`É “²1EHÝ}5·sŒ, D^àã>²~"|VÆIä-Ѧ(r-¶¤ä˜( ß¼&x†çcòp3¸áŒ'tpRËÖd0Êe”3ŸþIÁpÑ@€Pb Ã? ¸$©²æý¦úØ$W¿Ùx 3Q™ÅÔ¡rù®GfÍìT€È\Xk=>ºgq #¬ =?ÐcB0?“8ñ ±íZJ†s{o Œ´qÓüÏí"W—|JS0œM–6xÇû±xœË2ë=E£”´œµNbÇhpÔ@qÒ9´ôÈ*H’¨\pË6åX¹PŽˆÃì˜_U8AWà&Òö7ÍkÄ1ÄŸ£ßáZp è)@t­ß>”Ÿ‹ Ç§p&é¡e3”“ˆ/ñ1 ^ŠK«¾KÁ  ó;2:Àõ»h›•Ñv ÌàAÝñÙÀéã ÞÁþ™Õ{fJ‹8£ÒŽx'3«ˆ3ÿgÚ…dgF. 1@·d«´Ä›îT!Ä“fM(ÂJ2Õ¿\"U¢> Ø'ߤ÷xôÑyØ&ÇÓuñÝL ‚Tä¬|Of«½,—0&±Æ²Gà±(L<> aΈ²ßu£½ÁŠc…Tűgº6¥5ÐäGÀóø%(ÓGÄ!¬øxè°npK´ P©Yßn¶3¶ ËéKúïU?ñóë ¬ýºñ"2ØdÁw!ê(O9•cšE'64üÑTN3(®´@êG–(Yc]rºéâÙ…—”ðK9R¸h§}KκPcD?s‡çÂÚ'´g0 ` z„JrB|è”e]C  Ï>¬`¥Ç\V¾I¸ >Åx FXù\˜åvª)ä°f’¶$‚ÄÛÆ8F@˜QTåBÄ znX¡! H{{…8\¦¹ÝI¬:eCºÞQAË€"ƒ*±g_>·§ ~¯»–$ãrÏY &M¾3¡ûA]±2ÀÓåßRQ¢´+ÁßU¯¯éfÙ ÙE´u >D.-­—£C/±ó{†CA¹XÈ ?š…þÂÛ:@o®à…¹=ÆyÝQã¼Xy2ÓŽïKµž{]>£Á6†©^:ëH„A?lyP<ë–ÀˆŸÅ†T©i‚xhö¼6‘„©™>á)hc/âØQq¶© ýhÒ—‰TÖFܳ‡§]¯HÅj¡°#?ÅFßô¼£5 ^Ž¢*•QÄÑR‹à*ÅJkQ;¯dÏà¬×³µd¯ ¬­ö;¯R5óã³nj%Ö\ÑüLX·°ùÁ)ºÍÇ ™ó4ž£i¡ªLPn†X 3Çv,ó8"”fR¤Âl호æI·àøËn®]ÓXŠuÃã üô \: ñÆŸËò1q,™fªm©ù«5a®¢³AÀ Ÿ´0VfniÅ膰þ¢“dƒÌXëñŠŸ±«ÃƆ? =µÀ˜WÁîvØP¢l5ü íÁ_^÷–¯0àI4{êä)gñÒb_X…0ïLa“s”ÚÜp›“· ypñ!OûvÄS±zL7"#]K¯’+H?;e´h¨t*´+@¢ ¤ š\\Ø‚C@P™p‹gåEONHïaŸ¼¾7CÙÝ#Š+ûîÞ¼#Ìó»ÉH¯™¶àpá÷kËzÔ¼„$¢2cä<ôÈ)Ÿ3~ò!EÙü{ËD “‰ãßA•äÆ+¬)Ç ôVr–¢4¡w]Â~ÙϨ ârÛâx&K„œkm7ÕXÞ‹¾Mûbš% “±åÏ .p=¸­s5´ù°Ï‚*Û¦4õD»Ø áí¾È^ù &(¯ó/³*G®ÓÚyð'ÑŠÜÖsŒxè-.×X5Ý÷ _Ó½Ô)Zâm"â`”Œ„¼Z`$éK %<=gXœ"WÓç ÑVé õBCÔF’ô TÎ Xý±}ïÃr™9ÔKÖ(#/³™“ŒK"›HºÅOlšÕ/-­Ý11›N[f²Ú^옕sYiW/ˆ¼bkî¨ ·@b¢’\~èˆMœþ8ùCð;Ó1ÿÜ@øl ŒY¡z5B¯I´j Ò ü#ë‚ËJÄJL&Éa³”(&">ןIF *u¨ÆžT$0&˜ý)Ò§, ‹³•Rž¤-Tþ›X}ÕóJÂçp'³ò®D–zÃ_ dMµ&©÷ŠÊ õ‹hŽÞ N$<Ö x\òSOu3ÓÏqÒmÿÚ³M»À›ÆÊµ—€ ’Ïw½<9×ì¼oÑSe·•"Ö_º´Å‰m¹¥È¦"¢ü,JÅÖUøÀòJ L‘x‹K%½K¾A½žÃ`Þt-ÒòUÏ4w, ÏçºKø9XåÐ]Ú¢Uh°·0Øü(µ±ÝÇêÓíÈÕÍÜsÊAGÏ9)nTamk°¶•ÍÔé´–(–Nî) õöþÓC}Kd k¾6Œ&êYŒƒADpRã4È P°|!’ð6^X7!P š‘^ñšIàð’lÂêcKõ{KÉ!Z‡ò½A?YÔv³Øbô0Þ »?5k` Á˜yé¾|úž)æ>\rê^ ¥ù¼P4-¸™æð ¨ã¦€²€÷‡Õcî áœÏDöof೨£1Ϥ’÷¼@ç/“Iv£?Ðé0ú{™ÍxÒŸî¯HLCjgüÕB¿§$õ¾$ Ÿ]»O„€„/臌òôŽq! È°4ƒ1D4;¾-v“}«gÅbÈ,žÌ®/ É,Q†”cźúQ<á× §ôB4¸õ0³Óø¿âöÓ­‡™«QÄ~ʼ ý)¾Ç3/„/'„®˜ învj-‘‘‚˨LqYšAº¡=€®©ÖNË¡ðîó3Éñ÷_”ÓЬ˜Ã¹dàS®5HrbÔ§ºbï4Ò.$h½z~˜BËÃÔÏ¡SØ·Cq`,*™ÄUñ\Ú!§ÓÜqÁïåz‰TÙì$$sw|(~a†Rª€™¿¸çl œîDc¤ –4êAþ-̈ÊbˆœPŠ3‚¡¥6BLºM0Õ~t‚ œÂí¬ farúƒ_ÁÆ»=Whd {ÌZÀ€WŠKÿLþ 9+Wæué1Ñ]÷y<[†+á{àØïž¿Ÿ\oä"š: F÷¡ÂqÒ”K Zƪ²O@r4õXQ˜“ƒfeÖÈ»·ÌäaŸë–ÛÁMòÖž%WIçâIäÏ-—F?sL¤¹°åyÜäµ$3¶ïJ–D'ÍNºé×5ÌØIc+é.DöÒLdä° ·ª³¢– bò!“Ö6gJÔ5¨YWA¹F8¨†Ê øÿ¡éà^«1UŠ*Ha­œrÔ¿QGj];T€g" ØdC9’O:Xã„O6‘ž‘Í¡Nü'‚C G¹aŽ"„‰®.1)y$P ¬¶|û¤ç¶²I£¥¾wÊ»‘;‡(«¶@‹@µDµ·òŤd8Åg­ê]Vv&ë1g!Oôï™0^Äpnþ+þ…>ÅØRð俏2êÁG¤è¯xŽRœDÒ’)ÞòN:“n8{€Që©”†ïäñ‹6Ùaà”[wØÃj ¼4ÌcÙ/§Ñ\™+ZÞ‘¬Io/C<>™!]6~ê­2WxËCÄ—ß{¾iƒ³‹[ÏÒm“Å@ŽÍÝíBl“ì ÉÓý¦gW¡€Áª¦g=ÿÜôüÓä£[·–:RÍ®!Ng®ORê4Ïy¿)–ÏCÕõ¢ -Ÿ!ñäib´ÙÃ* ÌhC(6@ź¸ŽÅb¶à2ÍeÃÇɈ«S¿8uI¤jšiŨÃü·Çv¶<%/¦Œ¢•³‰cJIMÚŒ Ÿ‰g'6úÔUìô4E9cú!—3Ô'"G$á¨ÚÄ41H‰ý¨«T—¼èL¬Æ™âCË~-Œë(v<îïy9ݵ 4tÉÛK,LQ ®Éª¾fZ© 0žÜ¾´TÏiT‘uÃ=²…Vbl¡ÙT mIþ%ãJë©ô)Z’[Ç~Seâˆ*W`¾s†ó¤m?KÊÅ dØÆš•@hŸQ€ÈWcˆ‚ç™㨓CüìÕbªf¹cÍYÿ«¤Uþ³™CÝ0|,#«Z0¸<¶)ÏF ¬í,ö<éXÇ|_„ ó»o5ˆBæŽ÷¶4ž&ÜýÀ¤Û˜RÐi8  ÔÒ’i}]Ê\Æj!h©Šzn=+n10>õO§lË®c°­m§ÓXIƧl­{,7f a6Ë8oPÈ,A ѹg7@ï¨11¬*g1±qÇû8]F“NP ¿Zé†ç0?6¶p\è̦)ZÆ'ä­Ž#cR'öÀÖ÷bõñ]7à×§§¡Ä37|ÌïŽ-áiq€½d»'Ú×Äô¯hªIY›†$ÎGs½WÎÌÓQcªI^—$YÞú¹ªGæ¦Táðø¶…ÜÞ÷zEP#ΖK)&ד{4V¾ÛBtŒ¼E~ÎëÞnàëXK·s~¨²«í™£4À3ñB¼ù$ˆÌc)¶|êÍc™Jõr¬¼ÍTpÝQ`²M„&ÆäÌcÔX4«žô;YÙ!Øó([°Ñº_7oêyé8Ç*JÆØ÷ÅVõiñAÁK´Ød+ø˜£“|{´¿á ÓX1z§]$å±W!ª´ëí®Üg‚çß­'QÉ2<äWjz÷¹ïÔˆ†Gm̸®nýÊàX­1ܘù‘έ±®,½>¤LÇæ£*:iL&~Å„ R›yì‘ïÌAñhsYƒþߥ`K%Ioƒ°«7Ÿú~ÏÂA5¸ 8ÙÊ·S±ô߆Žî§¶F¼w‘•A¹‹fˆcïh©Ð8€Øä!ãÇP17Š~ò±š¯=³³Z¶7Å»" 3Ĉ_y©îMbhƒ-cæ <ÿåÑ ®_Yyxõâv»ˆSJÔõÓD¯ÐŠ‹C8шS +ÌgDø_4¡¢M¼k±ðU7õ!:µ¬±ãñþ¹iN;ko!žÓú­†ïÙ’ ©; 1V;2Dë•„øt†³q^Ă嗢òOȇŸ}Í×7iÏbífo[ÆbU\sô&Ó âx ;V Âîe;µì5Bó¡ XjÊ/Fk×å*|l¶`}†õË¢'ºP©÷üp~;³Ü¢O;òžIä|þfe*M"‹Ê·A™¡diS0ºÜæ8,1#}1Åõ©•j\…²âùÒ=I[=´ ó¬U3–4K}ã?±355ˆ§ò7–xË­'ñ½^‹“÷ò{{¹œ3Âb‰Tò (pI#ýÛ™ptÉù±•÷‰BnUzH‡ˆå¿‚ä Z+š·Ã0:þ}“ §5á’çDÓǬÞB‰%¿þ#~§sL¢vî“SoCÀ@x D'Š¢J¤ßä×n<嘼ç÷U¿;0çØëþÀ ÂDÚÍ€¯³w²A0•Lº½#ZYÙy×Ú ¥_`#1(!!MÐá¯Ø€™”²Äf¨ÅDCLDg6Fœé“ͩᆖËx‡·N4b™u\å>_èÞÉá<ÈlI8gH}*r˜l/3Ã÷ûdžµ¼ü´tµŸ_9í½7Ü/²[úXS…䘿†£Zl™@Ô5qÝQi4ÅÏ¢UK3LÅÃè'¸ ã°í_…½OââazÕáݩҩ•9ñÙQ!X6=ÈYáL^ÏN*œ!º¢$1t7æ µÓšM<.MCí›î^¹ò.¨û€mžA´æ–‘ÝÙAèÅŸ‡Üw#?NŸ\oÑÊ×Ý2ä•”ì~&°Ý6p!tݯòäu"»‡™õ¤Ï´BèÝŸ€¿冢CœïÌ•ûÙ©b_’¹©$ì™ßKxK³¶Ö¹YÌ!ª|rW‚”ée˜ õ¢m“M²)X¦Rñöã\eþÓE%Ml» JKgö\róxÆš«WI_ âyÇŽáè­©=ûÙ•Ö?¿ôT#CÍx¿à _ËÒ%"*æñ–gÀâ2\_ñz¼¬¡»@ëåE ˆ+Åç "^l†$Ã~¬£C¤™R_¦fŠ‚¶ CT³"¯×4Ž_}ÿ´2Ôxõç›d!q<ìB’’ž¯oÜÒ¯î-;òË"ÂÄïÝQ-é‡"—‡¢é'©F,éBµëtvÆ4÷§£W9CÊé\-ù†yÀwÊ»JçwèÛ–Ê…7õöÇ;0QZR{¢õ·“‹ÒGŠý3Ãcîc ƒèÕ¸Áïºäãé¦ X„f|5({²/Çç¶®Áìì`açºÐbÐä’Ä¡y1µÜª{YûQÿÒœø¸a=ÔñÍŽ_/«OÌ¡œÇXHQ î0kp7t>}q‹!/Yð²wý¶i#•Ù“”»ÆãB}4ö­µ‚ø¿žD:KC¦Ù×é.bòºªnc/‹+äq­~Ër5ٜ˒À¯Ð‡®»‰yÃÞ…tЉÝÔzmF#ª¾78dû^ñæwÌQh¦JŒW½Zx–6©3œƒÞG»´ø 7‰3÷„_CJ8Ù"]‡U-G;Üv¾brønFý¯:ý%Úã ÒU„2à¢ý#©MWŽ ²joÒì^SLä-vÄD{ã¿sAŽ?M2 'RÁ#~ 2žÞ™²ŒÛÏa&º„jpæÙy{+ßÓðÙƒ:à.ì”·£äÈÃ`¬{ÎðûI©«.½î„‹ýÜ+‡6ôâ P‘0qóŠ,´–n„Ç"`Ë׸ÜJ`á>,˜8– SPý…€›Ç¸æ”Ny¢.[Ë7Åõ7Té®¶·c&FÛfÅö{܃n½°làÿÞ`¢›MÝà“Þ̶iг\nP盩A€ú^&ÛöêÖò nyÁ»B S»Tc<«5ÕØ èÙŸû¥pÎÎh8âUÑ»R›<ÿvñÿ’k«endstream endobj 149 0 obj << /Filter /FlateDecode /Length 7449 >> stream xœå]IÇ•¾7t™æ`øRÇê«û"Ã{,G2`KÆ€<‡b“jÒ\šb“¢è_0?{Þñ"2³ªšMž:°Tñâ-ß[âÕOµ×…ÿ•¯^^üæ;k6×·js}ñÓ…¦¿nÊ?W/7¿ˆ#"|³Ï*ëÍÃ/øQ½Iz}Ügë7_^lÍåÃÀX—»±6죷0þáã‹¶_]ª½QÚé°}µ²!mŸ\îÔÞæd}Øþ_ëœuLÛ?㈜S Ûg8"æ”= (Ï}y¹sÎöYûó+|È磡¯u‚Òö-~T**z±µvïtÿ®:½ >y'Ÿz'æ}ƒ#’3fy®›Ë‰{¥Þ>†oaÆçn¿DåS6Û[žÁ—wñ?ÿ ìuGà¨÷6ĉÀßUBñàþävV¥½6›õû?ò1J™·Oá½6¹lÊ¡8¥”u<•³r*cö&(Ó½7Áè—^ ½Ñy]]ô°? Ópt?7j‹ý¿¡)åÒöæÒPeïÂö%­š©üä)¤³Í8[€Ù¬‡ÛY`qû#Ž J¥¬·7oðû¼ÙÁá¹äñt¼*g8‡í>§h’2’ñht¶pV’ÄJ¯Ä±^ðk’Žð}e—¹ì×Hì´îYŠ&^¶«Ì|Ísößás>"áÿøðâ¯,á.'åâæÍ]ÅÚ*xÖùM„#†ãDÙ¦óµ†H´ý"Oö:éByZÔÏxŽF»ÆÞ7¯ù@PáôàSNÆ™ntáUN/ãÚ¶ÀE%ܼh’7Ÿ^&¼çú7^ò“Ó¸Û] òÜ•$λ}öIþ~9qXb+„ š¤`›XÉ+d™«à|¯ünyÉ–”AEŽø™Ô#2ï®qîlg£¥5î´Ý{`!i,ÅôR«¬B!{瀳‘ ÛºWxð?aûü6gu¸Æ•AÍá ó(¯3©¢ù´Û~!æ{ÕX\ð²¶'<ŸÝ#EãYS¿4 ìtpÕ¦¼­Ú\½v@ö@¾óTpýú>Œ¥M/‡uŽ«å9ä9?¦ý¯ktå¢Ô{åjõ>»I[ÊÍ\5%$ ¨|åí¥Ð$‹ŒÜmá|mÔÓk~$Ý5P’ÓƒÚµëÇr'Ó(YœÈmo~d¦ñÉÀ[àŒBÊÝɾ`£Ü3©ÜˆÓ Rþ|é=Yþí¡i†gâ1–ÀŒ¶FKùº àÞ¨L²ýùËÏ╇Á “„½Ž±^JLMæµhcšÖƒ¶qÁˆh×Ù8~“K ÷Ú§íC±l’8ã6S1'†Od`m{¯ÃtÐ>÷.üê2Y</·r`¡0¬1мÝ~àå%|h‹~ÌßP8„%PÝÄí™ÒåzØVO†¥ßhãäÄg±~rÆ¥{¨~xõ^ùÜo÷A‡²>QMû·lˆ§?vÕ´¸uýªÁ`Uö;õÂ5ßÐ ¯ Ðh’ÚoñÅ}t«ºO:ßUµ_yŒMžqcûÀš…¸)K5¹¢ƒi †7£™ÇŒ•’íùíšõ ¨7è5œ$†ègŠCgã‚C å¹u(øÇÆ0¸‹‹;{ü ÈÏe€KÉö\óŸ=@›ìÙ? !|¶‡¯–µà5{:J[”Tuh¦•‚^Ù„¤÷`4³† Y¯cWª·º(éà-˜]:/0îÅ&1¹"Þóà}Ž\s»ÀÐZ«Ö„)+O r+î®Û’Sv´¶¨¼Y…=’­ˆ0¤`«ÎåYäDѸ†{·kzr­h8šùei½¾öpV£ý­ƒ=PŠpf_­Û§u~¢C_£g§ß ²²_íMÕSrÝÏ‘f:1p/!àr{/y0ÆÈå“Ü7ÈÃ΀74~§8Þö\G_º¸x¯‘Q'låLÖòkãŸ69Ÿ·Õ&ËÇ$“X?jë0C!€å9:“±€Û29èÊ÷ÂY|Ša“&d€¸!ïñËÉYä%)\ D¢°ºØ{´¸4 Ât…ö°é²g2H7re½üðž¤!»áïÐÇ’«aÀˆ1«Êz?ˆ•I´O›LW¤½‘²-Ö܃(z¸Kú`E‘™¬‡wùéyKì­‚¤<Ðâ›COÄ•°äZ‚àk©1‡¾%vºó¡oÄ„…¡yd^<¯ä´Ø&†ÜXª,‘¦<¦‚¡å Ð1##ë:`@Ö¤qо–y&z)ònE@XKnÖÇz¬_öZì[è”I@½Ÿ7;cúHÄíÿrÐv­K1U{µ®7À¬âÿuñ<ÀD<ã(†ÎŽÝÊaIÖbK–0X!N¿Ž6MÙ(…kúu‡?­nˆa â\„mzÌ|2LH_OabqÒ3ÉSä, É¢”æ7üPÔ†ý(ú¼(S©!Ú(låàlãp§HXòFlçf×P°œâq³wý×4I è*²Dv€[nʪŽloÞ±öp|å%à€ý£ À{f0Šß¿šhX¥*‰óë2ž4ÔÝé Ö5Hq¡ÚY‘”õôÞ: ‰*ªëö©Š°X„#°*€;!g<]ç¿m~yQI[ùŠë‚·€î8}]ýáe¿¦ãŠAëyÕ}ÂŒt÷£{Ø)’ݤ vFïÍÙzä¿-¥ÁåÔLÖÀ}NÈ€IrÚNüŸ”—ð{))áüü¨úÂI‰ ÉOc_7ÒØÆ#¡¥ò¤ÍÙRU%‡~Y­x75³r ÌÝøé1oÆöðˆÌaeeÁkPyhV3ÂgÂqÄ}²¹Bµ¼jïÖþ][ð}¯:i+»v% &ÖŒ³¯Á‹WM‡Ì¦Ã•©eyè»3¨v¹­yº0LV…Àp¿Õ…Xì ÈxñÞF¹#Ã!*¬Å´ÝæH—ŠI9ŒºD˜úXHìC9(Ô_ŠÀ©åË´ìÇé<‹)²ú±îK8$Ç2GŠ´Ê ½OÇf5‚Ì|wÞÀ™-¸òë)ºÂ–y€b&b°Q4[ âË%÷à±!(½b\|`¯Ÿf:¯#ˆê@¬ä¢g9FV”¯*Ÿœ ¦…H;'5 ³aÌ ,¥Œ?M‡ò‘¼/’2:þ¨™n±äβìD~þ•Sý âµ€ànÚO:4»vÌ$- &ÅŠŸ³Ð̵êr"Â[›×ÄâªÍ-FŒ³Ô4¯~pêh[~íLlå¼ø­ÌVóXÒÞórLP¤©˜,zaw˜é<½:.EÎbÕ ò9Q!ÀêuÛÇäë C.FH|Ï£ ǽ!âøÆ¤ë™ÐBŠ.<Ê(kª±^ ”6neŽ÷°ø™¡t}•²wF3Ébh—U_ †„}UÖWBõMR¢TLwà˜W‘<@úDe¢^bRBRËËkµåôë¿ÃZüy.iý¹A¡1TÁç{(ˆGs°˓ǜÆHyLí׃Gß6ƒU4 õd¹ê’T"溦r«½÷ÎíÞÓf|É/2“Í!Øä)v@Ll.Sí.Lr^éL>U¯äû([ ·Ocr9&Q3‘AÚÁN^Ëþ®Ÿv†p¦ù‡8÷nzSS:õK¶$¹`uo.~e£%¨yÉep6+>ÈóÁŸ%‡g4*Œ]˜@à¢øŠ´øô˜_åsWi"À{†u`òп۲»äGÐõ># X|*‘xvÙ<ôUg £(œðìoÍälC·_µÌWèŠM‹2ˆµZ5ZxRV ×aÚøÓþ¨tZ á‹÷:ò0>͚ʯ= šñ ôÆB=Ì c¤˜ð£šVìšð…Žs´íœÿ¢âÇœü´Ö{ˆ€á®ÔÚ)3$ˆ8ºgYý(Ð"¾S?lV82Û…(çû¡+ã’§Dá k‘Ì{S…H–à“<â)ZµKFYìOp>;äWÍ5C"A{ׇ8‡¼"íp\ˆZ™åØè™ð ¹1ްpM;»â..!ñÔE=¨»)îê»1P’Ð;UEHO“Iü~y5WîmE‹|i VkS>ÿ{,°¹ã{þOÙ…Œn³þãAÔ‹véB­¼h-_Ø8q¡d¥†æý_/Jp(hñoFKµrÃ_ælŽ„nyÊôI`:Î…UÅ}øˆ×`%Û½-߂͸GzØ$Mcï/<ø¬´Ý¢®ö)¿sÛ«ž:M–ºòti§‹(ÕÄâ$d}Ìo ½Õ¬çZØÝ¸}²Õ—JŒ<8¸tÛGƒŠÓƒf»î-§ÄCù(™s‘fçõÏœtyƒ²;Wö• e«½-^:oãgB4*‰;- sB¿I—ûQ„‚WVÈa4v†êz]Æ0½3LM>¡ÀhA,^^ë>žðÔ¸D{¯ÌÆy€4´–»Oá,œ‹˜î‡ÎB>ml´»kÇ¢~Zü@Žux +ùAþ²£¥?Í»œñà¾Eïã¯ÙI=…³!B()*`‚í˾t©¦*–£pǪ<ˆ.kÒòž_htoÄEи» @ZBé5 /gPaíÆÄé…Ê| \r}! µ¦Ë0DÜÆ¸^ Œíj],~aU'|°`'ž‹¤Æ(ÿ_X*¸)âHâ=åN†:»å$⊠ƒÄÅý‚ÄÕ‰AáÖ‹€×»6ÇZèn ÍîB,§(^KR+'FŽÅ‚VŠZN]æ<ÜN³™cb”‰½Í¯éØ`,1ýÕð__5l7Úí­ †¯Î‚C¢ãfgñŽm*Eœ†ÎÝ0¨D Îò·DLŒnøíX4âÕ6ÎV©Ä"fT0·Ò%R=N¸½Bðà[*“SÜò*r\i‘XubrŠÁÃ÷x:u}<ÄF»!·lJ±¡ÑÝúgâ¹<p!©hñk¥¡xRfÚüAB {EpÁÖÜÓIxS*zÀɾëU6æýIKÝ}°ßí“ÚZ¤—%%I6\çÛokè7mI‹—?@FDª¬+ Ü•¶XHÞ.“È åZ¼£.u%Š9ät&B¾—ê‡+Ñøà÷xÍK~Ú¶´Á€ e]ݱ±}|‰ÛÚËõ’ö‡Ù’º;óHi°0Óè Z¢ŸÚî•YšÿÂb®¢Ûþör§ ÿÅí¦·O³óªp¯O˜˜èöíïú0ëo¾ zú(ªéxÀõÀ?HCÌ~ùÓ`7yŸƒ LdÐjH€Š+ú±ôž¿•Å•æF~4žËÌÌ8¦ÍóéhÔ´"/žr{Ð m+¿cÑ×&m5SxO/r8ìØ÷â,k”í¯Ê^ä²àèuÖLû>ÆýÊ¢úÉ=h+|¨{æÇQäFk7ìÏá1Îg16{6‹p‹N°Ø¿6%»p˜: VÒ§¨fçT¨{™SÃz€“¦²MDc¿ÇG¶vQÐá%X²²(·÷Ñ=àˆ¼Š¯Ñ®pGP@ô~Ó=ðo£à "§÷Ul/cäç‘é Ù;œQOÞ lÿ9å0èh&9LŸ_³f9y<œ‰øÔJ(ÿt1%ªè¨\)¤ê/D‘¾_̈ûª2Œ=¨²ž›‰mÅYOYRbg@§ÇçÆÍ$ù »3ÚªE“‰å)§c¢š¨Ø.b е0æüNR¶XÇ88ùIIt Dä™E9ØV.²Ú÷ÇÁ ÖRÏcpÏ;:ÅÌt#¿0³Ÿ•ÝËHzUZ bYi#¢(Úþ>°iV¡ŸÂÝ3ý©"©m:R[†9K±b Ë+@½¨W ,…ü5¹’¾šÂJV¦ŽjYta”H•qƒ‡D¾2õ4(L!ºõ‘Ž"Dåe©JÞóSðVâ™rñRM®: €Á)á<ðÝÒPy‰€9õòzîÈ 5û•üQðóýÑŽÖtA¤tãO"„ðàu¹Î‘t‡J®;öí*Kú²};~äâ±_].¨0è–5JHXu!< ?@àqÿb`·¨¹æŠÒÏ^‰à7c«­49BŠ(i¿à…,¯ï`+äHßvx_ÍWùå)Ò Úß‘Y áÔ†ôò†@ )R%¨=2¦¶å£Rh|ÏÝŠ(X¦3- weIsP‰Q°]÷ÐXÖ›EÛ‚S2¡Âî%9Wᢩ“ìíÑâ”q¼ÎˆCÒ Øk áz½Ín©ø´/ä+j»Ï2L !¾½xøï?lÿDdåõÎô™!™3•~´ÍûÌ´òS=²Sp`yAaqöVj-¨#»r å ÓÍt»emfW`ô¢ÜIÎk1õ³}%¶H8Öwœå ò-js¤I$@VnŒ.ÞFãìavVw+ž&S}UáÕ°©âîŠJ”«vg\]võ¿'l'æ©\µ_ŒìÓiÐ[.¨|G¥ä#©‡¥.X,AåB\ÀšÝë'ÐvÂÉÄË|^WIªÏ¦ó\éL×v¼¶¬‡Æ¾áÿz=m8¼žŠf´:3ÕžU`Õ~ŒŠ³8­ÁeGE*1îê‡{-‹Ý›ÜxƒVt“™.[¤þâó±àšq8í\9ú\WäŒ(b}`†µZ˽e¬Õ 7Ñ\ Ê\ xâ+'è#'Mæ'x"?1ÊÀ¿žQò˜í£ð2‚]c4­t8µÌFS¾_æøoŸÃ @®Gi8¾&`ÝuÏ Í CƈîÊÚO-˜ ºÕ­õ¦ Vœe ŠÂ(  0=,¨©2Jß§2 û>""mó•LaÆŸµë/ëVCS ×Z^Ê ÿ]]2P ïÆaô.ªÓÍÒÎèHs¢I£vå$ºÂÆv»»sÙ¦ή,ºk¨¼7’×îÐRËÁÀóY=ïÁ­4+xÌÓuë;õÜ-]¥¢@¤)†;Þ­’R@nø†c¦ZÁ‚Xîî1õ`ÁuнœB}AT L¬R!o‹À›r+1ger);+_¿å"/@i´È_ƒ…ؤ°Ý^·ÉàÏþ\ê§Ê—·—­J~~Öx b²Ò©­x-º¶ È]ñ(„X"œZv-ÅÜm»^~×ÖA›b¹À-àö †Ÿ¡2ŸnÝ]ÖR¶÷o—{-ˆ=çžFŸ”„»Éí;EB0g¢‰Ìgi2^áØnAR¤}SõF¯ï›ë!µ¸±‡BéRç€DéÖuÊSUAz¾/m­æv¤ø$1¼0w¯ö„xÁ™~K}{€ͬ²ˆ™‚±§üM 'üͳ€ú’€dÏÏÚDµŸÔ‚"Ï»{p>´å8-'dÞðL¡¶HHd8–»!>á¢uÇØ$M¦hgU°ëžt%^ðKb–AÐÃX„¦Aå™õênl±IZeƒD‹qçíµèiÿ–_–8él±81§3+ÞNÿ€Fe]H×©Š¿Ä2Äõ‚y/%Lî\ÅÏïAºp£ NÓRàâˆù¦)ò»úJ Ë[Ïæó´õ5ØM^£š0å6Né^aã@ŽªmúšÂÞê;Míée/Æ¢{œ§ý‚ÿµÇ KÒjàPÚA„X«QNS¦1JÃ+úAœÄQ‡` –¿W‡àÖɸׂ›’;åD⊞*P#ÒT~•‘£¾Ôºî+Þg@Í+âmKv6G?®pÕ×MJ¿þI0ÊŸ›äFo^Èk·èà\ƪ?ïQCŠ›tzl)ÅGì¶Ë%¤dн-ý^µÅë!­¼K4¶Å›Š¿™/Üjªµ5#_ó9ï79h× |?‡w–{?yz!TÀÔX þ5ùëÜ~äÀP¨^oµÁør`v¼8¿«àuv¯§y‘<“ÕÁË\@„7ýoy”®êA6F“¿hSm&RÆŸ·¸û5ü¥¾hAF^´Ž¦‡1 P‚7¢ÛYg1ÛÜ7µ‹ö`2ùvR>i2Âa½Ù¯Ê7ècg,wzÜ|:vAs)Ì+žë®Iþ#t3ºèxµ%e…n9p¨9Þ€²s„ ñöõØ]Œ¼kºCz›§\bá¢@'eñõ€F?ÜŸªúÞ6ö„9·»ût+e­”ù4X5ÝÚJCVФˆ;ÀJ¬ôØ»i…e ÷æWIa-Yn c©¬c¹ˆDvÆ‘UY}oà¶EaÛEaܱ–åš×©Ê ð.±Gó‰Úø¾pŒ©ãóWPPk-*›¿ˆÈäu½«ÖudÒzæO7Ë@²)ãÍ1QÒìÉørçºuÎZ¾¾Y›A†¾?ko¨ú6À;̽³ Û½…’\øËx”c A¦¤ÂGÝ·˜Ý ®[€Ýj?ø)nô ˆP/;-k=N™¼×¡Õô½ç3æo÷Î¥ ¶ùÁµQå)j~x¿ ½ãJ©‹M }Ðݪ«>ù¬;ÝŽ9ã¸Î¸có)²Ù9Ù’îÉ6ÖÖ>Ç2åÍ{€é_Ã9"šö0öÌ›—6ôë7/.¾_óÝe’šÊ N§$•àß_X+Á¦–¶«èî ¶§<ÿ9Ê€w`M‚(*4=ëöήgþ¿rà'¸Ÿ% ¡lÔà»XèJ¬ ¸™ñÞqì_/þÀWßendstream endobj 150 0 obj << /Filter /FlateDecode /Length 6385 >> stream xœÕ]KGr†¯ƒ=íÍ·¶OÕ†º\ùÎÔb/†µ†-ÉÐzç`@²&‡JâLSlQ"õëùŠÌªê) t`«˜ÏÈx~™üa3b3áéϧ7ÿü_Jnî7Óæþæ‡A»I<}ØüË-´Ð¾Œa bsûü&viÕèÌÆ7e6·7_Ùz5ÊIºáõvÕdíðív7.ø`†gðÍNV{3ÜáW‚s’šJ¯¥„¶ðKLÊúáéVZh`üðb»SJÂo1|ñ »©à•±ÔX„IJ;á§ñViÝöËSüÏíàF߆4~”vr{w3¨ííw7;­Åf§Ìè­Å¯_·8„œ& 3{¤‡}ì”p£žìðj»ƒù¬pb8±µž&ôð#.n²jR´k¼5 vŠ´˜&íÓVµ‡%^ãxþÂáx]ŒRa8<ÙâÐZ쟰&q D^VÚ12â:¦äóƒ0Ê: íEy ¤+ãîãg;‰áIÚ"´(7Ã\°Ò„áÈ…FÎ::bll¼OÔ˜Üä‡Cü: ³Öïþ¨‚÷ù»ÑÊ x²ZHý,î®.cO{õNú 6P§yRú½Kt~ØóÃ"ö j2 é$LÃèüXg¼ÇñÇ8qn‚µÆ9ác¸ÞÇ¢ˆàÛž nÕ'°£TÚ9ô ,HˆŽ}Daw½<ƒS°ÿMÓ‡¹!ïê¹.L) m87¥:?%º!g)Ü‚£ô)þÄHÍÄŸìküù§],Ê$,¯:‘ºŽù‰ˆîD€]ˆ<Ú0†^?mzÓô9y"båDNL©fSÂÐí” '2ó % 0Xàö€%D&“*æNÜÇ Î‚`þõ©î*ºÅɰm|έè/ ˜_¥ó¤_ÖÀòp¿"ƒ?ó ¿8!ÀOmt9ü?×Hté$ÀzÂÈb0É•s°jÊÎJK&´ãP‚QÁ–ƒøïÄcàµ.Lƒ„ Ökè~!¨êô=¡Ý’ª@ÂùÐlDƒ§í£ì±‚+´HŒ¨à'ü9!g¿•_T/|9“ÞûžyZ<ÀEÑ~2Qa!k.«åtm­“›¦ÏwVé"O ÏTõüy5>·(×/ µgÓ‡©‰Mo«f©´-¢IQ óø×™­«£×,V"oÂb‘|æ€køõï—¯›%¶ ´Xiê´I¡¢·jI)ÀV p×Nƒ×‚FuççQwö4SÄÖWê#ózI EÞÈ÷Yà&¿@4=Z g} ÑÌÑ4þ‰{ùû´—Ž¿š+¼P+‹êù»‚¥Ï#83ÍQœóŠ#Hy½â˜ÉèŠÃÛëý‹nQ$G,žšIm×ÇYà‰¼“…³¼ x*}æþE#Y­“p¥.ÿÐËV;s0œ‚¥±.À9âVqÆ0i¡T„0Ôò‘'â’ï y˜´¡HjÎÈhTˆ²àwˆoÖÑy#¥‰áxF:^.â ïâλ„tЇãgJ.Å‘7„_…àqÖþ-Aa`;šˆ…7¡ÏðÕèá—U¼„öØäG ÿº"{ ÑààKÄ4 Ä—í)½!2ÁØÑñ&üíGwãöD³öÓKÊTP JÍCÔúª†zœâ/#@c û×iwÁó³åÕ3h,‡ÁR01íSZ7Z™€l£ak3\=\>Nž{ˆS À1úØ-(rdßyÆk‘ýÒ6Úq-ºË+o³99bÆlα¦Ü~&r^dkjy.b{´‹žežJ“i–Yyªe33Þ+ß‚,ûÇ ¾.q@8)ÌÿS”Pƒ9S gàÜgðogmUÍÞs¼æéLLt¯í²¶‹›- ¬5d£!Bø3š „J _Æ ~ö/K„vˆúÀ#ÛUTûEã!m ”GoŒ?»½ùëM,0ÐÀBU½^¯*èH˜ª À©Á±Ø^´ŽeÿJ› &æ  ß´ˆmNý‘SF²Ï¯F{ ⾄Îê0’ËUX žÅ̳q#®#ÿ´MãŽý¸0Ö ”GНì]«D¶kIe&?‚µkIö0/ëÓ¸Û82d)±?*_²3þmÆ¡f?F”lCQ²,`ÒˆGREC§#*öIÄŽ<¼è=€v²²8õ¾(–“¦ÕYTk¾žê%m!TcûÛ¼v<Áù˜Ø^Í q_K4Ø”k‰á*YQá‚Å9›© žì}Å×uXó<ÇkeøÚ¢ãM,áXÒ'«˜i–*fŠ*ð`ÝRcùùYÃÿÆ@¡U¥èžGPÓÁ¯zÖèœrM¡Q›eNù<ȘÇ¿¦ÿ,þÕªSàeä~Õ{±}Hz3\/VÀ÷û>ÞéçKšJ°ð± ‰Ò^IäÈâ2,y ¹¼jÊañ)O¾èòcä-G8yu‚eŽ®j˳ÖÝ‘13Váǽ“FÛ$ :ÅçMQuJF•ñjà‰Åû(ÙІ›ëª§ß„QërsgL¢Œ&ž dÎþJ–DtôF{àPŒ«nùÖ ˆŽ©áãEÚR* SÈ,IÃÒÞ ¶ÿŸÓ Ÿl"ƒOOxQ£¥_ßÀ·Ž+–6COµº ÌÆó%1X¸§ƒEE— “§—"z6z± †B„QxÅà ÇÕ£üùEǵQ'àƒP!Èõé~K~6¯^ö"ÏdûÌË€³™ãøÝ›u?2wät®‘!—1×Èh²<µi³|T¬3Q èƒ[ÍàS]R°XA¶g¥K?º”¼‹£»ÐÚÉcÝSœÕ¡ÆÊG´jB¬GçØ@O¢©úS@È£íZo_lŠ+0b%„I†QoU1Ì«N›šï“ÑEŽ¿eÀS¬ùλm寥ê¶+Ú'¢ca¨Išx”õm=û”ÅÍ”,EqÙZBè(‘aKr¥C?w'òTfôžÔŸCl+/ô½þ Š”Kt‘ƒ‚쓞àP™}‚±ŒZR´ÂRlJVñ›©óÜëw¨ÏqAï°Á‚œíJX¯Í?_E²Óû†l䟽úmÒH×{¥pèšü³Uê\)ËóR)È¥úåõÃ_­_™[x²GÏ’= v©=y²%ò:uku”â!Ö.~c““Dè}ßåÛ:ºÑcàÖ»¡½†}p«Í ºfcPqRþ"ßz?c`Ò|EU­%üÌ7ZC½fïâpƻƒV ÓÜÊ5f$Ý‹9a#éäŠÉé½–> bæ÷óÖ2¸ +š”fzR1ì Wn<,'ú¼BÂÃÓ…,JWÕ«+ŒüÆ s¶µc¢0ÍsMó]CÇ&sÛ &^V²¸€…½‘3À;î*%¸ÂJ–×òÑ^Såð–5Ë"%º´5ü,]ÄÐÚ©g!°tn\€Ø8þ<0ý^ŠRmƒÐ²ªÔ…©¶ªI™¬]Ê©ŠÇR7ÎO‰z¹É˜î–ÛR~©æ·fÞÖU¹ލ"=‚Ñv¥ŸÕ“/«‡Æe_9Ñ’ë‚ü}¾hY‹@Ó·9$¼ÁäD‘?á-¢”Er·Ö¨´F#èDq CwYøÂ"Ã:¶Bw'…_ZYñû;æÆÖ˜«ì@uúìW³ìoëOŽÜ¿©4)hN!™æ¹éå<-‡¯Z…BCÀYð¸ªMçÛdŸ@[…7¯Âð¤ž?§l4g Ä1Ÿ¶Â7¯ó)˵ÚåôVºžëMuÃ~ª*†ñÀŠ <¦ãC£˜ÎÓ‘mÙÆŠóèjϲë)š‰+j£™tâNZ®¨ötHÄ[6‘ ]„G ‡ŠjЕÑNŽ>U ÀÑú3ÉøëþPu ƒ¸Ôt+¹#„ÐÖJĘދ°F@#-"q`Zh4$—§§•ó}•ŠÁ™‚f÷¶[m‡3ÂÖoá‚ÊÄáoœ?ËOnï8 ÀÍÆ³¸=¹ö‚bž}ÿ¦JÉ=Ï Ï ˆäEb Ó4@UŠˆ‡%ôÇ&l–x<€¦’%°ro™ËFJ§òÎnkjÓ$­©½®ÜB–Ù/Õ)ž(IEiÐêúÊ2Mþg}Ê>$XÐõa‘fæ*’T9i:,Nõ=˜EÀ~ø•y¼>k~Ûw±–O^ö.ïò„v扤;3»ÓÀ=»wèÍPzå5Nâ(µQòµý£Qûî|Ô^Øén»´õ è„ä¿,Àýbuu¾ØzVƒQZÓåÏÛѨÎò8gpq>|Q§p”K®ÔøõËK.¬ä¿l$E¬À‘ÝVö9|¾xý ï©j±~Ë]£-Ÿå†ø£uó– ëVÑvDÿ²Sé¡´¥„ €Ê_Ìê3 µÈ]&?ñˆ’±üÄá´Ã‚"-úB¸¦WñNÁµzµôy?½º>åº^-}H¯:VmWºTs¨Éý1vÔ¢ypá€_5f"Rn´O`ô¹Ñ<Í‘ðtuEíf•3-#kã=óïúŽÖƒO_¨ôæ,¯‡8ÀÎmšîÈÈxŸ?e¢9Œtªÿ´ê›k´ò8Ê‹BîX wlµjÏTÁšºˆ–g½Z*L_q¹cxˆCË0{h†Æ‘ý2ÖEUWÌ÷¢ÝûI©¥Á0BØÁFÉÓó&f1õ@¸Âä®(¼&ù@tÌ`’(œª{¥×4ÖRy¿ú%_XÁâ1Μ_†YÎp)M¨è7ß´ü#N‹iüóæ_*°4µüæ ØûC Ž>rSFÚ—¦\ôÂèRÏ™™ÖµþâÍ\œügªå_å‘Sž€‚¿ÄüÍý™3$—.<€˜âµê¦O $h«IR'A°¶ ›3S®½†”ÿž‡,ÅÉ5ѱBp+ö}Ѿ¸õ'”ÚŒ¬-VÞbÊÁ «ÍçS=†NÆÌ~A+.,ß/Ãà› ˆ%÷ÑËDI!êi )èÕ*Ý“6|51]è'öܱîÌ€N““k™ƒ£Ko®á[MPý÷5‹õà‚®½šôXÌ#¾Êâ‘háõªr¿£¾^å väLôßùR[pòZÈÜx×gS`϶K Ì/ÄœÐn!ßrª"˜ÆŽ¯?D.i³$ô÷~ g®{”ˆ£YT”Ñ•èä'ÔìÐ âWnú¤†Àˆ`2]DÐ:¿(€*›¸tafϤ¾`•÷NÄ7_”Œë R•GãPwZ-ÙZÆw±5‚Ôìò÷ "+Ã$SìŸ0§$ãàÛ:§;‘ÑK N=¡›†Z-Ùù15™t«sª…£åšñý, ^Ê õ¶Àç½=^x=„=Ýtöm¶—ŽßFì‹Q¥"ŸªC³Ûk 3uuØÊ|W"f²ãa vëÛÔúR“‡ÉÀøT%e´èŨuËHAhJC9ÛÌß?i‰é0.‘ð鋤f¨”¯zÀŒÞ"§{å¢ ××r[ËõúÒ­‡ënuØ5Û}WQðVðv4£i??Vy\I¢T£kÛ{£ù%+–˜b†ëð[È_)`>Ì.n})¯Ù”0åå#Þ½[‰&ùçî!)L*æç 1M æA’V³+Éãg±A|CrÁŠ3³ƒ¥f¥ª°è‘NË¥¢‹ˆlGñ°nòÊp«"Å%Äj‚-\¸ æ"ûºWžðΈj}'NèÚ›eŸoª?sêý£4òG¸Â§ðb0hЕàæXº) ?ঀ’„kðñÒãxᢽŠ{Œ_ñÚñ2¸þ¼Qãñ1¿VC0NaYÕ&!¹u²}&™¨  GŸ˜Gôn[°Šå*ͧÕ*±nŸà]b;ºÔ`ùźÂ{mú·èÿ™(¾Ôº–Oâ†ñÉËXlÑžC”PúÓ ä–4p¢O¥Ps°ƒMÖ¼–0/«‰OXjÓÞ^R½ïÙ˜„@!œpž>4’L.sCöË5D¹eûŒ±6Š4Û/÷ªñh„ÉÄäbîÑH*ƒîC VÀ«sbe ×+öüRYx\‡é/+7V†Z î©po"*ŸR=’Eê_ÆŒªC ¯8t³{1°(fý]È…G$ùzÒ‚Q¬¿3•W3¯n.˧Ë, ¿çü9{\7§ö%¸PÓÔ–ÔJ˜g±VY³W¾Ûüw›cŽ>AÊzqJ˳áha:¾£µ‰¶,…B؈ðq>þ®Œv¨¯{5¾>Âᥨö@c5^Y§êñ{üŒå˜,ZaË_®vlĽ«‚+Uëøœ¹è x^W:•Aéó¾~,&ô£«ÑÊ¥–ÞÅN9žÍ‰_ ØQ=V­\DÚü.,† îÂ+S½¿3¬¿Aa¸ö(|lD†˜ëF &åZ" ߳§ `ðßw=þ @~þ\ÈŸ­Ç—üÜË×§¿e=~{åÙá#çŠ^Y2.óÈJ)3˜¥Ìµ Õ#l ¢,̲SÃ)kRs~óŸìóÛÖyŽ6/«‰yÕ*gkr¬‚Bì.FÛ“l\%öè‹Ø§úÒug²ò“Õ‚µÈ½`%¯³NÕ¯>Ö¨e½¢÷«h—ñãÖ©¿Rý¤Roéq_Žf&ÙæX†ÜOFòæ]áù?²Q=(þlœó òÉýúï‘Wï¥ùJRÓ•pÎQ Ìê_ŽVZöÔÌ•tËÝ6(»ÕRþ×›ÿ=û)endstream endobj 151 0 obj << /Filter /FlateDecode /Length 7044 >> stream xœ½]Is\Éq¾Só¾À>u;íÚ—¹òXMØ3ÂMih ÁEôÅWÿlçRïUf½÷@pH*t˜V£ºÖ\¾ü2«øË™9Ø3ƒÿkÿ½xõäòîìúþ‰9»~òËK=kÿ¹xuöÛ§Ø"Ã7‡jª={úü ÿÔž{–c>TÏž¾z² û§ÿ mCUm}:äè¡ýÓË'ÚýóÞœ±Á¦Ý ~´Æ§²»ÚŸ›ƒ¯ÅÇ´û |mkµ¹ì~ĵ–’v/±E®¥FjÐ~÷Ýþ<(²Û—ýÏ·ø£XkÎŽ¾¶z(»7øÑ˜lh`ïý!X=ÖÜŸb¿z+ú}-Jpn½¯ÓþÜ僱Îî.á[˜ ‹U­—¶ÁÄRÝîž{ð0ÇõUü×Ó… ŽVmp¶Ÿò´Á?ÍÅ‹j|îª=dÏÎ}<””ð7»ÿãsóúÜÌ¡äb§n_À|LõµÂòޏf_]N¸(®Ôšbæñ´¤ÀÚ HWëäx­ƒ1±X,®ÊBŸr±×òÌ^Á÷©Ü´c5äÝ gývÏ¿Z:S+ì#íyðòÀ^ˆN¯ø7ÁúÝÅ~þö¦Ë×g–l >ã‰Pcvï±ELž°K& £`cØ•WضIÀ‘ŽÚ:¹ÐW{—`q o»\¼âÌâ>B[Üî$·ÇËMàf)¢#10§ 9½nSª:œ‡yM‹5&@‹gøµÉæäÏNLêë),PÍÿ ‚©Ùî>kKiò©E x\þ$w´ÆÕ¸u#r¨ÙPãx¥štvRšSü’ XœùîÏ| É„¸&pðÑÅâÎÔï¿Å½„å€l¼éòJ2 ]æ&¡ð H¼“¥xm–X(rù“ê£ÍŠIæ-tƒ¸¶£Iùé¼ ®î܃ ää`?ü!†êx=B®H†ƒÍ™u±õ–T4Á¡´¥D  ¾–3D££€ ¨~´20íůXjjrÉ¿Uišš@¤PÞIŒI¬ÙMg¤N¼ìfõöÊD¯oSÁ@*e[³¦,AÈegK†J_Kªp¾7$ï5ÐRìƒO¶ :Ç~,ªÍÚÚÝ7´ö6ûèÆx/{ÀiSòMXÕzXØØ™(ç€#xë½ åó$¯ä|µ?c¿ó¢÷{ÚGhÜþc}d‘ ¿­*‹uàlâ´3ÏØ®†Ê*=;‘ß=}òÇ' B-&ä³×ŸŠÀ¥L¨gpr‡è- ËàŒËÒoº'—ê÷… œÜî(…c–AËž÷áø·²€ÂHµ\ÚËà“íW[í§®0wHà Þiïø62”~Ùªaœ1îY\øœED¬é5ìçý;è:ÌFÎÍ8•ÂRôýí~Þåm~É®78a´¦L>‚žl”çb°”}Ö&çˆÑ„ÖÑ„nª à¶Aövگ⢠îœ{+yÿh`ôºÛ¡ã-YŽä—¸†Tˆ`MC<¤bê×J>€›µÒ„žÚ—àšÇX™†¤:øEêÄWÿ8„„#á›/ GæAÄ·ÃD§µ¬Á[8ò8ÀìðÓ` ´9L0¥Úé÷+0%—Lñ5¬¨­|A¶Ì‘7wá6™èäž*Çzp/6ý± )°D6ø²Ç òø_s'|·¢$5 ]Ò` †hMŽ}Rz·¡˜åãšTöÜEÄÖ~07Bá¥èÓþÄêª2;·‰¡~éhäíºæm7%R;F\†}’†4بÐî·}#ŽÔÑ ÁŒÁú-©m‡gœKäi ·PªÒˆYnð 2ø\Ø£4΢ɩõ!h6÷Òaªa+´:êÇ˹~ÁëM€+Ùg ôôš;Ïv³—«‡Õå€Xà#™èA;Ñ€DÔÎä¡—…v®id=€†Ùp¦~sºí&è¦{?8«ùXðÞ&\>ø<¨ÒÀû"ìœ×i|$À cÃ2Å: IßÎVàÄ"“ZHŠÞ¿6H‰ã(¸®ÎUü-ôgˆEJWsÏÓyÛž·1íCÁ6¶@Sð;RØLäG¨„ѱÜÖëÁ2 ñmÚ0î²óWÜ9âåuÃq±N³ˆ€ûr}®ëqÆñ ·€€@:LØ8˜äÐaÁÍîíš¿Fß%÷ô¢[š6 O>YØÎ[² ™‡h|ìH.ÅåÏÈO—äC‚pšíÌ{lŠ!ydÀÒÄûr’Ÿ2ºÜã”`ñn“ÑöüÁLü„¡Ñ`6¶+)z:|0áøÓe’·»(2¯x"k`òY~<Ÿ’塬“‡ÂD/´_Á> –Y#6×(ŒBö':¤9âîé*~Ø ŒÚHr 2Gߣ¶_ºtÄHƒýd‘»»#Ê·ZogäqÅr NFbêFÁ¼æÔí“”œM¶å¾k`Ýã³1„残ƒdtú4®ú4î§Ó±dЛàY=Úš–p,‹ eÌœÐÎ(&EZáüñ\r$ÄG2Ÿ˜¤»ÞŸü%Á® ò‘4BâÅE vG}ÿš{ÌV ­ÄkGaû¤D!|-¢§¢'¿gA«9…B¤î÷¬%µœ+“–\u¬'9ÄwûÉ\Mþœý€?Ÿ†a±mƒØ¶ÀD„ÏKëGØS6õ#ãY-òcÛ»T»¤ëè:G!9# ëÊpŽJ’…¡9Jã­¥ €N ó =¢ŠmÃuú¥Mpâ[0ì¸ÞÏ;G,'7}k7QÛM_£$}õQ·‰N*5Ì_Ê>ƒÍ1©Ó‘‡Ï”4—×ô “àJG&oÞÂDÊëT´¦Y9‹#åÐG~àjhǃ_“Ô)³˜¦‰^XÕQ±Ij¼õó‘›´Äü¼TMØ ì{˜þú@jh´žÖ™ÈZU4>5èW·þ½Úl¡Áˆ5l&>]òX02»ƒ0 í~è[y$ f¢ÉJ®Æxµq@iÞs^xX826÷ÈpjbŠ¿O$@tý(2j5rEb+óB¶Noû4ÿŒKgÜ==Ä»m«Ál%Ï0iÆßŽ¡oŠ"®}uñÌP2œp,­ÆÎ±²â*3&,²šýG»¿k¬¦&¤<òó³o1 ¦Á3ßþ}[øWò`¸7ÕmçÉúç·óæí ŠDÞÐXK (³¡'jè:¥×ß*¹’i ˆºÃÁú2°æ}bëïj6á;ÒJv.-׭Ͳø)gFêáÙðâB Ü?iŽŒ¯rHÊÌ_pkDü¿­l F€ñ)bŽ1*G(²Ä/Ԇⴠ•bÁ] ;¾Ì.KË»–A° §ÊøSùO‘³>^Q¢#Qí™ 8ÿÀv:A»)Ï3§½+¥ŠãJB7Ró~•)N°oâìÈþ"¢‚»n@dØÝ¡âz0Á´NÅ`©å9Ñ%?ïßÊi°3‹kq$Ùåußvž½öM´#D÷åÙz#%uׯèð@§Œú\ô‘2襘ªÓÈ÷ÝjËnî¹9&ÜŸ3¢7XÅqjH¶úd×ì%ÀêÊœ'?]ãô`ú6õ°ò0ëGkìÏ,ô1ÕõüÉ9ÄBcé6z;YXLÖ%ç¿@¾@¶ñëh¬ İÛàÿK»ß_^ÍÙÃÙzJÝr v#L>´÷ šßI&À; ixI#EGÌIoÄ_ƒ´ -åñkP–³4Ñ®VA8{(Îä³§xòô(%wî:“ܱÖÖÔ¦b¦ õ¿[ÈnÑ ¡#[‰gH&o@°>2Š›”< Œ@yRçõ˜ü†ˆ<‘0r—ÝÍnÁȵ¤Ð7Ò?ëj»E£ [“¨ä…æÔ'1Ï­°}‹C‘­7HØO-3ÃM,ºÌ¨QPÅ?®ÎL؈µ`°µßsÏ•b`2wÞ‚¡˜éÆ“`6¶Pg„Äñèm™™j‚"|#b¿FÒI²Îðâ•shÄ9V°el ]±Œ£¤°ÁÛá¸ç¶ *}7c†ÁÅ#ÝcYlì¡ö˜ÆW©£=CtÒ”~kܲ™]ŠjÇ¥%?MåfI¶î·€³(«œçïer÷8Ù›6HqâlÍ@©NŒê{üî2Ëß¿[ejN¡^3…üRÊõKzHˆ°X†\ÿƒšKå*ošÎ@ à¤3›^zÞd‰r'%6å‘Õ¢m¢÷ƒÚMìÕX©6W4‰a ×Ê ‡2ªói%’ šÒ•EœŠÄÑs ¬U¸â`L ¬Ø“èýL‚¡þl¡µíÑŒX¬¼ †( —Ëäj>ÖÎÖÜJ®1ë !€EñÙÖp#$ç†@¼CA})¦uK+v%GKÀ®XÜ6,€W½#´`f¥Aª¹hƃO¥î&ßÑÅ­ˆShzM?wXäý'è–¾ïrñÝf±`Dg•&ÞªŒ‚æ²DU‡(; ‰h¢ÏpXØ…/sÔæm“ŸÖ³NŽ(ò5 ws_8hÉœf$¶âŽ¿UgÑhj¸ÇþñnUe¹gÁ ˜—#ˆ ‡&}{¤A˜,0ô3éºE¨¿p€›î$”ÚR”³sZ%ìÍઘª!¤ß({Od_(È·U…݂ǒŠ™7S-ýË Ã & îX0dŸçyŠrÛݳ.½Ô¿lBçÇ».pµLG.Ò±v´z²®ûÃdú£üëÔXfAXúˆÜúÀ-@wwÇßà( `äwa«QÖDߊ“›XYéÃ^8"‹ 8 mÀÉ&ÐX[{ºbzºMð@–‚6Ûæ¥¼N}knGšœµ`eY¸ B£¼Ó4 ˆÖ’•ț֢úǵb¡Ûþ;Q(u¥oWc^5 ç+óGSMMüèqÞOOnPÎW..Z¯ðÐmQ˜bGˆV\DôU_vˆ˜}«ét•…äßö3§ÿ¥µÁ+ïη[2Füß±Ç)æš²mnµÕÝ,~r3_“—þ®7Xå(Ê¡"ëдä÷b"ÏÑF"ÞükL~Í]dI]â,©ap_TùŠc.VƒÁbÉÇbgAÒ}“„,e.ˆÔ  â©-£ª– hŒM7U9YC=f‡î:Aÿˆ¤’”1Z©6ÌÇ9äPè4‡-º°u¥ƒ‹£.Éu‰Á¾ ŒÈà…‚£fº…–}Ìê¶Äpªãç8ýÛîõ¹N=£ uîS‰Œ¬âL`e—e…ÏõÕ ý"%ÓdTV"ld…•¼æ¾ -¡¤(®¹ÌDÛ›!Åkq2?Ée-•²«:úáÆiáFZÂt¦-ÖsR[¡eç6¬íT&T1=0iè‡VßÒw[çÙ„¨:t´äO6iˆªÞ·à<Ø$f|‚¡„ UƒêÂê‡ý·#×:V#ñõQät¥};ïß·Pç¥ÁÀ­j!¥½&~Õícæ+°Û*ìn,ÖðÎM;%鸗û.cÃ^¶¾Iìf4ñM&ˆmj”Mnö3Ÿ"¿¦Ï†4iG©´EJÒhùž¿÷ ˜«•…SÛXã&W- Ôg;@¿C8ñ`M>´ÀÙ‰ÒÙõç">NT*ìÝbmùjš|þ¸¯~·&¯ÕBü]±W‚9øV.á9:ÛÇûOõEZ·xKãæqQ¤>•^¾à©'«ÁÿÝ’§z»zö”­ïa±~6f‡¯o©´sVÔÅ“8ÓSZ¨¾‡Ú¤( \ìuÕn`ñ´F^ˆf—ìp' i)CÜZíÍk¦ÂRÍêžáç'@5}#.”·0ùÛO¾Û»E8f {7«ä»(¾˜ZçGŸ›6söÊox-Õ¨uŸ¾åJ¯¬ÃC_âµøþREÝÈæ-«›wé¼AjVc—ù¸‰ªcŠgW7ID°¾.‹°`;Ý©|h;ïÕà_DÍžóÜd¥ŽÀï—¨$éyâ:d@»cÁ«MÅíþ£kÙ‰ ·¿Ð³60ÆÁ¥ræ°Ü:~zFäìß¡²`ྑZXpÑ•Ió€…7˜®<|›ú/ãcB-lÿŒ7t0TlÔKù /¹ 넊qv aɽ_;õ Z§¾ïöíìï¾Ô»Pt>W=âÓ-£$(‰¼µY¹¦ ˜^isnÕqφ¥™ /p¡DÕpÀmñÞ× ÁqzVä0Ëägˆ¨Ù!‡úÕE( ³£ÅÏ!Æ ãÔûCnòK,Á¬õzć2XK°‘ÖF¿0d¦W|ÿ_\39QÀÃîLþíO¤ qþÚ]¬¿q_ÿ€Á¬ZÔ#Óø9 âž|qÀ`Rñé¹5{:=JFe¾)?ªjçùŒi7¼×"ól]5n·ñ4âú a/ÖÝÛÆ(‚fŸ‹X—W¸ÇmeTypæt;]%j[D¡‹[z1‹|L†ß½%êX¼¶1ùg¬am”2Þ?Òº¨RVñ£çÁDzuJlͺoÂ;|ÿmÈÉàw€C§ `¯æç àÌ^rS,­?­4ݰê"¥¹î á2³×¼D¤WÑžŠ×ùÒââ.ãU[¯aŠ&o¼N(Æ£(èîí·.±gc×nX§ŸåyݦQÝz±„HúBœÅ ±§NàOlfpÍ_Woëzü6s¸³2Á"aK…µá#yÔj_—·¬éEXÀö(q£Âax¹ñ¼-oͼd;ibÓ ]ŠÜ®=¿ë±Óv}JF¦(|©ÚWYƒ€]ãå¨÷û™!ß~4QÄñ|•`·|ްóU븬‹‰d©[·˜JX}#íwj$Ö—K#׃óÑ&F–âlÿ¼cqãƒl¬@‚Ÿ¹v9óÓ²ŸvID “ýi©ÚUqɳKŠ=y1žÛƒ×®ÙÌó×o Ë"¼ó•Ùä¿TP¦Ã7UTßõU¼"˜nÀVuàÅë!êšõµ|(tùÄuKvö`ò=øL ò` ÿ[aï@ ¶+ÓVý¼~Ý:×ÜÓ_¿§O~àéà=¦‘O§YªÂƒõƒ« ÎU6]=+E{™KÞ€¿â郱’ Ï ¦Ý2ûˆW¥Å\Œlë®jîêzûÕ”Ùÿ"·ï+¥GÅ«À ËÀ F^Sß0ë‡4>R]3Aâk!~m«˜mñÎw[®v]¿6;?pd7ùaáYJº 6ÕœDç¢öhWÓE3U|¼vIb¾ÚMãPøÅÊ^X}@ã$|Ý%ÿ2ÎÿÖ€. Šþ÷$$%.DKïü= `}vÓ‹0âm—á{ÞJnŽy£Lùx}Uhé7îg·äXz€Á5`ø¸ëlö‡÷¼pI¾UÏ~ßwífÀtAýStG‹s›…Æ1„Ì¿Nž$h¹Ûò$C¬Jìû«ÎYN”b.K,/Ùȧgaüð6êð<ÐH¹nÜ_ÆO¢QW+¼›äUE\ºxm‡À‹¬ú¨Ñ”Ú…’ï<Ù† ¥r€_‚¢ÆGr<£,!Ɖ_œUM?ðDì"þû:È ÖÄž\²‹6Ø3ù·‘]äc^¦¸ º)·ûÃXÜÞ„ÔûÉr¥ £ÇÏa+aåѽ_ƒ­´ "@9 ’•ásÈJ0lœù¾ñ`TÄ%[|‡Ÿ®¸L„ѧÕ½Uöj­mãÆfw2MwxQ—%L>,óâê7ñ 4þݵgñ&ìØx“cë –~kPâ~ßÈýfÅapcô¿íÝs|e¼,@•›÷¬ÓóªUF[Öt뺗Õà›4¿¶ò¯lm¶EG³åþfßiŠ_ñÏ'µN íÉO+TŠ.¢{øÖýÌó%UmÄ›ýMȻ֋Οü?žqzÕendstream endobj 152 0 obj << /Filter /FlateDecode /Length 2961 >> stream xœí]$7‘ç¯ð6â…¸i\þöÁ!…¤ Ë)9-é ÒÜîìÜ’Ý™Í~${yà·Se»íêÏŽöt!E÷pµžêr}Øþf.z˜ ú—ÿ?½šýöµ’óÍíLÌ7³ofçÿN¯æ…¹´ªwfîŒëƒ2ó“«Ù›îÏ ¯z)¤ën¢WÂÚîb±½ >˜nkVXíMwF«&çdD•^K‰¸PÖw§ iÁøîÝb©”Dº—kúL¯ŒÈ„”¶»EÐx«´7lñÏ“¿ Ž‹!ï¥AINÎfYœü‹0Ôt¯´•„±”·µ0_*Ó{k óM§p/Œßõ‹¥ô—ëþJâíA!3Q`d2ñh@¥»íåp¦7B£šDï…’º»$d¼•Ð%áH0ÝCÎia»Ïi„ÐÂDz:Ø1Ê6-ñ–€òiæK©z >1BI!$²üŽXö:HO;K©MgU)Mw‡ ¨ÊЭ¶Ì^gñCå®GÛ-Ñ8g‰êYŽ‚äxwD6}6S^rÒÙ¤A;Ó]OMæíÈóôhïd³7Q‹ˆ+—ü{? ñ¸Ëë n*xYÁ« þ£cðbÑÒ® èÈNvç$»Â5ˆ\Il2îžë?DeN¦EVµO®/È-/‡Û´ª¼Æ M[àÛ…tdÁ•‹ºwÂw<^RDJ@¯;gÖ\¬ÙêŠÙѪìòþš>Ák×ýŠ@/|a޼u·Ôhú€>¿(à>'ZYÜý2:ºE &Ë÷ yÆÈ`«‡¹h¹Œ<ÒîVš0Ú½4™Çk´ó²lN‚Y” e‘© M“» fZìùGÕò2 ñ´Á0B2ÚïX‹›ÂXÊ(&³Æ`0f©yð™+[.ãKGúó$ž64Å*®…ô«ÈòäUK¾€Š3¢`j{Û_°) Ö—/~7J:Sò(¥ö Èê(å3ïu˜%[г!?*,l%Ÿµ¦»ÝMŠ'‡%÷´ÝCM’3¦¶åº–¾›ÚïÞF&œÂ4½z[ >Û2E¶£‰‚eìvmämËM;a¸g›v¿H“á4 €Æ”Ó[…4’áX@8‹•{>úæÅ‘x ‡1ÇXñ8…Y9›ÎRÊœ²2d©¶wMRþóG½ ë"î_\Ñ4§Í‹&£›GÐÒ ¾œüúœgM1îš«ÄxUÁ/šš`mcø0—LzVm÷ûºº¬«U«\dfÐmÝîß|@O-Þ~¶hž©=Tð« J†¼¨ð³J®*xÇW î‹ ¾ª_Tð¦‚ë nšݤ;fø Vl»ÐÙŽžƒÇ~Z©¬*x9#ÁçÅåÍðùyó›u?Ô˜‡íšϨÝ7½qUæE8˜ÿÑÁ¢‘ã‹o8 ³ò}o›imÍ Wt}Ûtƒï›Ÿ1Üì|ÐkµwBƒù Œ:!‘ùht„7BåXp°DÔý«JŠNN+]·üö€"Zyš!l+xÊŠþX±8‚Ú™h¦®^4½¹s®¶§2 » n+xÛ<«¤¡<j|ÜÄS»tŠÊZ¼U±1áçévÏ(FçqùÃáf8Ø¥!ºÝ×ÃÄ`éÈ0#ëCd>™S“›†8o¦ù‚*²Ë–xn,œq—‚‘DÔ l•\óA•Æ®xƒ¶ )âÙ·²ü¤³ Ïl®sÇø8:q/‹Øq3k$Y3Zèï(¯£ž±]—£³t®º|Ÿ$ Ÿ¦`°±Á†p¸oz?bǽgì8{¼»æ¢–l•ÕT‡B­•µxI•æ>«ÁÈâ}×üìúX¬­›«wͤ{Iï'4gç˫鼳ŻÔdLö„œ—3ÉQZn§RÖD°<×·uûY3µÔiáºIŽ­K-›u‚ñ&§Q ¾‚¶ ê :þYcÔø@¥al[é)óåͦ~ê¥SÙ*똔¡‚¦t«È.³+77k¿j*=ž1Z‹óü! ±UÕ”S5å4G,T:ãÛ*ÚE7dqÍzˆý‹j2ËLÅvÕM·}HrðÉyòVb?뙀Lº;íý¤‹m‡V'ëÝð@sr¯ùÿ–v™ÛÉË5“ ¾,Ôÿ+¹6˵ºÉ#Yí`™ þ§ ´¤{´2émS"Ù.ðÕ,Á2¶Ú¥»`Ýᬪ=?.Éosâ%–P½Ëñ¸÷šg‹Í³PÔIå#o…ó×™Ép^º¨M÷­tÌi¥çLñºfÅí<~šGD[.ÓÒã6·—9õ¾=Àó÷_ãËcþvƒ¾TøËžF*ùÅjr¡³<[ûpÈ/\é$QD­öB,ßtGÿÂTŒR$ðØ8ð“ïsA½OÊJå—˜ìñfUôâ–Žž2•K"^§…¤HÂŒ‚ß]'Ü&V1‡æ<z –¦ÿ•“é?ÞM~9û‰xÆBendstream endobj 153 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 4475 >> stream xœX XS×¶>1$爈{TÚzÖy‡Wµ·Ö*Š#¥NÅ!A(“L$Y F™G0ˆ(Äñ"Õjoµµ“õõöÖÚ×j½·rk[‹ëàæ~ïílû:ܯßý|!dåìµ×ú×ÿÿûȇ!ŒL&c=½—m3×öç\éi™ôÌiœÈŽgbú|ã˜Ë^çvîŽ38ÉÁÉáø3£çºà÷O`Ñ ÉÈe²È8­gäÞÄè àX·©»§¹ÍY´hÁL·¹‹Ü–†F‡ìVG¸y«cƒÃÕ±ôM˜›oäîÀØD·©/ÇÆî}ÞÝ=>>~¶:ÌDf3™ñe¦0¯0S™MÌff:³…ÙÊ,c¶1žÌ,æUf93›YÁx1+™9Ì*f53YÃÌgÖ1Ï1ÞÌzf!ãÌŒ`\˜QÌ“ ÏŒfžfÂeÃwZ\ÆIcîÉVÈ. Y8Ä<äž|¹üŽCºÃЧÛ]ÊÝÊÏØö8‘k:thµãHÇÇÃ܇=t ®¾a8:pá5âÎÈÔ‘…#;F>z"ÜEîâëÒ3jÒ¨ÃOFK:ç¾;`E«´¦Z&íì[Äk+õb ôÙÚDbì8V›úH]´.â€KT)™Þ„&h„ëP§ÿ³Ó+Yœ[&S^7Ž‹5Ê+ÄC1K¥_¹qÂKž¾gWÕ/W’If“IdñÇidêÑ YtFÇŠ,crz:¤Ä 2“ð[æÀ˜Wé{%¸=ô ÿ,»~æâÕ÷‹Î Àáêf²Ü”Q%ÀÕ©^¤E$›­8É"i“= u4ÓÝáQ2ƒ÷ÂW” q'£ßÇ=ü '£0û6yÚgk|ènñëÜ·…¦ÜDSžGSvì[Ì÷ß¶¯®l2¡ŽA7܆s_"®Øt½W°k™}Ë2wÍ^Á&œ Äyଔ §;p–^βHnͲöw1é]¹4\ZÂWA~r–2’…]š6[Y¾Ç\8™dô³82®%[l̶dåÁ‰ŒÑ æ^ôÍÖuå=‘BVµÁ¸8 èbE²  -Ë7Ayµp¨êTðëP ®ð‰t=e‰¯ý[CÌ›ŠÖ¿\ g¹·oC ÊÍ+¢M‚1¥À†È­qŸ~VÉÅι8–¢÷ñµ*côn ëa5¬7Ñ-«pü$4¹;:>%,] aj©÷õ€S1t‚ï+Þ?Õýû@ÄößþÚ'eÝbkè1¹À•@Q¹è,­¦SÉZqIµì; Ýár ’&ñ8ÑJ&¢¯Zi˜ê7uªŽ{íSÒÂŽ¿ûáG'/~$t«·°«Ã#Â^†kutü´þ\×O‚½Òzþ1¶²Kv®_üRNÃï±Äë'œ¤³S®E<¬ÿ>µ Æç•¸Ô¬°²ÞZ8.J”.ž£@ì±Êp-îãƒU¤ŸjzM {½ú{ˆƒŠm3݇Nz݇6;tlw¸“]#ñV™”)9óùeyW€³²Az=øÃ^ÐÃÆ;°µÙZ²ÓE2”[¬ÊXý  j ì ‹Ëû] LÙuàZ¦ÂjŽ}…¾i8Ÿ_‘›ɶB„>‚è)ƒ9²…ç …®“.>KÛoØ:ïCØJóØ6u™ÍÍ<’—‘›UÙ ×i º%dÃXÂ`•->V¿’Ƈ‚„Æã¬þ'òó´µàZL1PPÝ‹•c{Iyaœí­k-äšó+9gÉšeAÿFœo‘µÒùÛH^…çù»ËÞ"ã}‰cöõ‘ÄÆæ¦šö*må¾bÁRXUÀ½Û¸HÜÍ’¹dÞ."÷@.þ½Ïuž¬÷€ê[¡KYSµ5°_\g„r¨ãˆì"?kAbè«þ–Ó_!ÓòÏíùG´‡w¨æ kñÁ‚ÏÀÕÊúÒ.„Ðòª Ò^£ªšl-<½^œ?m*Wt>Þq8lŒù€E²‘êÄ”ÔýZíBpU±ošLÐBi¦Û•0As ߯üñ>*PNžV¨UÊzÓ5JFÍ”YìQ+X;fŸ²à¦.ÊtXl#쀾5üÏùÛN¶¢ª¢ª¼ºkó…´”Á…Þ[8 ÇͺCÆùlOõøñGÏž¯ì¤Ÿ2jõNسG óÛž +aÏÙýµ" a<Ñ6õÜEk½’hYWç&õ «Uùª~.l†=@”1°ÓÞþžñ*é’ Âd»EzÛ*«µu-SÁ›Kó ®ÚP¢Ïùÿ86ž.Ë ³áøy›m8Ž¡EŒ¤…öÄñU½û'e§îÈNuÕì õZ9 if“±¹Ú9KLMxXŒ&dûIõ›7_ÿèd5í¢ŽR•¶v*Ü+ÑQ.ãP¾$ Ò“ †ÔlA›°w•;•›ñ€3O]6â“YŸm0ôQ§KI†hNݺ¯¡¦ýà™›„5o"^+¨n³ä™¯Ýq ÎiA._p~t~°NSÎHl‹4ã(žh¥£dJe!´*×”g0Då$êã ‰ÀŪ”u¦¿Hæ'ƒ>#–µU$&7"Ï`„ZNº`d±¡ÿá4sz¸V€¹è@16÷ k´•XÙhz­ôz_/õôZI« ÈÐp?‡‰m÷f¹Ýï[ö#¹ídãªv5yÑÝ?;yq#¾y]/œ©²·°èõ¡I,‘õßáÕë·¦P"#G½C}Mm»Ørñ²¹:àd¢eÖ¢¬6ômÂí>)ã®\ÚŠQ|šµ7vœÕ¾7ØÄý¹IÄ…¸}ï3pƩފ²dÈJ2èR²Ä¨5«5~4b’Ÿ¾(69˜:ŒÇkÚªŽ­µÂC($NÆmœ3Þ˜LŒ¡ÖOñà¶MêÃm´¯´’aß+¼Sn)6ôf!9=ch¸ÝGÕ¶•ìê\îAä„&-U}DýŽÔÒ»—µc÷îGá/§²}‰Ç•ø§¿þ=23?'‹Äþ™v¨SémôzMö>°ø”Ôã;,Z¹w‘ˆØM¥þ2ÙqçÞ½qSZe-71¦×|SŽ£q 6ê"”Ù¦ôÂÇïwE´¥‰Måõy&]6õÍ)œ¦4áСҪšúÄ#j•¿&Q#ø×n¥u^yqEX‘gˆ¸/!1˜2·ªnOMRL–w,lä|Þ^…ëðù[¯ßørCSt‰àW¿æQo¾ ²ªüÔvÊ”Ʋâƒ3óóá½Ó§á½/¾¯;Àk¾x‘øñ×ßOÙ¼Ú·±AÛ¯^2ñ±ë¶Áê7Œ÷äW—‡Û\Õ?•?§£Kÿ±ñîû[ªEcÅcl³º ô£Ç•,¿SÞfù^çî?{‹x ýŠò` !?žÊAøB¬ÛKêu™9Y9TÉ2Y!©Ø_CÃVŒFxïGh<Å–—”¬­ÄáÄcl~&¥E—û[oz‡†vA)tÚe3 â!͸?7'7­”²Þ©VI²k–ôìi9µ7}<Ô‡ò*MÕÆJj7m꼘 K$,̱ŽmÔëšrŒˆãúŸ`ɄɈ#ñÌÁ§. ÒÂ^›3r£ón?êÜ–PË2w ÷-zs° õtÙþ M¶ïôƒ~ðÂ-þÙ)v—ýWÞ÷9¨»ç‰Ãäí+C“„¤ÏW–­…°Å[ãÇýác­ìÓóWŽÕÆ-þéú!¬ó3†Œš¾aVÙ£éùüJ“ù–MÁ·Q´‘Jó2ˆìò—°ßúœ ‘¼Ð¿‘¬”Â_[•¡úÙôxÓÃn±„!ãúâEþaUFè§Ñš†ÂLˆ ÀŒqæ}¹6gH]`¾¹—Iê±èÙ¯6' ü· äåWQ¹†ËïÊJ°[Žoãrþ.évgm ƒx 'Ê1áÃ9 x-Ìßßv|`) ÓjçZOVËp?*äøMË7é!TXº–Èm(TUÊ£Ðm>që¯%¢Ší4ýgèÕ;ˆ72Îö8Я¢k—Uv]ÑŸ”K& ¥þH§Ñú$§gjC2ågweû‘¯þ»¬埽Û÷8ttû€²©búÚ™›­Éu-m5§;ÃjÂò„§»ój€»{ò%ÏÅ›=ýU" 'ñéø‰®û¤…¶%SOG]ŽÞÂ';ÛdÅ!èm“¦R¼EO0Áo'Œ'Œfun«Ÿh-,kf®#¢É_¡v¿¿ Gà¢O¾º.úkâÞ(|Ùø—·àî†ç"‡E>Ïo;žt¨éhÕéºôÖíùBgÇ»`î6x…‡çì f1*$J¨O5ÄrôY:ÐB·Ï ÕörRݵ=þÁmB°çð=ÖéÒReƒ!/:QÉ©ééY¡ÂcDId*¶Éô åæVøfжL§îBŠäÉì»üŸÀY–oQQBoÊœÆHÛj¶Tηf6R•¬‡7àZGKÕùSÔ¸€’Tƒ²Ó…I¯¦ùRHúo¯Ðšt&p©¾O$')©¥U™ŒEBIiÛÉÏi•aÖ ž0ŒØ³®ô±ûà‰Ú¶¨Æ=á¯%nŸÿɆ³{nãHt^ø·Û/3T%Úq…?bKâ~ .ÒíðkÀÙ俎N5¶@™á¾ô ŸC&Þg©Íw,:QÜñÍBXzJ Dq{cZ-‡ëÚ/ì¼4™lf.LþÉ®ü‚>ð®`̆`‘Üø­Ï¥Ùv‚Á»Ž#*3°%Ñ7å&ñ§u)a›õÉ @]+Qé4#ñßfðçú¯¡ü}ööÖ¿UÑ'6 áÉzˆÒ$T¥qtºSK$ŸBÜU]T¢$;òY«#ʆ Ž ª†¶:9¡¬Îi¸Ñö\õÿú~üendstream endobj 154 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 8306 >> stream xœµ™t×ÖïGk&¬L°ŒèB/ „zïÅØ¸÷&Üm¹È*[’{/²°Üdc›bZS¡›Ès|sßÉ&¹ïËýrß]÷-ÖòBÒÌ™³÷þï½ûŒˆê܉‰D’¹ËV+üoßGÄ÷íÄ¿#öÀ†WÊÖd»w¨sf½³2àÈßì¡›ºuÞÕwäJÞ£':ع½I‰E¢ ðä¹Á!ÑaÛ}|#ú óÞoÜ´iSFõ?vì´~³½Â¶{¸õ[æáëèA>ô[ì±Ý+"ºß°é¾!ï½ûnddä·Àð1Áa> Õ/r{„o¿Õ^á^a;¼<ûÍŠè·Ü-ЫŸuŸc¬ç†È#¼Âú- öô ¢(jÝì s‚7Í Ù5˜ZM ¡ÖPC©µÔ0j5œZO 6P#©Ôjµ‰šK¦6Só¨1ÔêCê]j>5–Z@£R‹¨ Ôbj"µ„šD-¥&S˨)Ôrj*DÙSÁTêMj6Õ“r 4””z‹b©·©HªåH9Q½©ÎTª/•@qT.ÅPÛ©7¨.T7*… ¤º‹ºˆºRëI4ÉE:QoÑ“N¾>oÙy{çSv#í$%¹´„nf60¥oÌy£â_»v¹ØUÓmX·»ÝöÃí“{8õP¼Ùç͈žzê{òKš¥ÞêþV6»Š­{Ûóí³½ÞïuÇq³c¾S§½óúL쳨KŸ¸>Í};õMèûIß»ï¬yç·œû‡LÝϱ_|ºÉÿ¦ÔÀ=ÙÚ2èÑàYƒ‹?ÒgÈ™¡Ý‡~1ì2¯¶oM ÚdáEu­³Åü€Ö¥lZ&+bA­L‹Çmß9Foôˆ[§eèºýúF°@‹¦N{´Û@ ½)Ì\iÐåêô²C¨—ÉqKQÏ`ŸË";|ÓË%ö| Xx»f¹Éuº„¼.õ’DÐS—Ó. XHjÐÓWA%yö.Û³*à  -¥¹rêÂù¼u«9ÿ/¯Ýa ­eÐîƒ% 1¾Äϸ æ€ó¶àeŒôÆ ÷WØÐÒƒ—³S7ˬ^iYD­«Åüþm½ƒ¶šêvÕ=Ô9YhO­³& <` xÈêô^­I¤Ñ¤*°ÏpÄ,Z§1¤eCºä›Jöë í«ž^ËtÞzá®2ú$”¥–£¥Øâˆ§â´øÈÀ­CÀ‰¸L׬¯‡&8 jëî·êbª r éYyhòvDãqŠ^ePÚ å1Û„Ø4®G7¿vt1zGÆ cߺ‘Äø¢ í±8 Ù]tæéœg½¤< ã;±HK×ë³sèMÉÓ†Éë—lž‚Ųo“Ù§e—.À5æÁ»wðÛ\[§€ É ½ÂGÖVAK_øhpá|_ rzôðo¦]Çâ|Ù2ü‹{Jœ5Yõ2T‰VÒ?6Y8oôA2{^­0ócM¢ê»¨ö®˜÷G~,zkèKÜ÷E¸'f‰Þ@ÝúI9ŠÙmàZr(윃FØ'*T:\¾À~yå¶Êm°Š¸ÒÖÉ]å.®A›€y1tリÿ•¿Ç6û8½ˆß4Úâ÷UúP5ZûŠ£"Ê•R†§ár¼}”š zÈp*7§ä,´ ”'¸ÁV›5P¥ôç­(EÞø;G¼n·_Òh¸Bö¶®´‡)‚þ@»G†ÖÑö­ÞrKëh‹Ã©ç¼C˜¹—ôÉ)ô‹=Ðd<-GS~@2ÄäBV‚”IZ.ì7 ‹€YŽ7îBÇdÒŸP Ú°ë=‹‡9}<¨ )C¹ %Ñ%¥P© 1…Û4Û»qãÑ‘à$}‚WàIxÞFÜÝ‚×[Ð(ïÔfvxy ßí%}ŠfñÝØ‹4$èüJ–ÒU&½ÝæÛ‘0»t-úJ¨‚cš†?&×t:‡Îˆm^÷%0¨ë/? IGüˆ;ˤOg€›k¤œ1à"…Ó5ÇŽÔæê¹ ØwÿpÊl/ã¾Y|9èÀÈØóvñ¿YD—Ìü2³˜?ÐÚ—ÍÊ#?e1ÅJˆäÚÖJXng–d!»ÚÛh*yØ8I›SNU€Rœb2 œãÝ$%Èh&IÁ²¨Ñx0xªÄ¾5½=ìÜ}^úLÜÐ:“ ¨ ;´Ú¡ÔN4qhZ‰Vã>hž)Ãoý½k+¿Û–Ï$ŸÀ%ù±yûWæL‚€ÅQ+|6G†¬^5„È‹¯h÷és¤Ña×]TL*ÔmþÍs,^ð¿û?-§½ÆÚIŸV'¸Ÿ¿fŒÀoãžO‡ †Dà2ìÛoªa (€Åýið–‡'Äû‡83gùCD£®—ï^»Õ± ]«ZŠùVþg¶QŸwƒ«°H|Õ³I ð‚Yàkµª=éqnA ô_\õ{i ú¹Fºæn¸ Vý„ÓX[„ZpÑ_]EJ­Â$o& òšZ¢ø>_±ãh4(ÏÎB˜’u„¤B4ú:]ð|zâ–s¦¨·pèëö_n‚z¼  gÀX4Š“F‚Õ5Ÿ0öèïíõ­ïC1’ öò¾â*˦»^ÌOD÷Ùœ%û/ „¡ ÒF@8lÕÇØ:e“6/¢ Y£R)ãlGÜÕ) ˆpõN•{ ”ˆ+P»F!à¡ ´öJú<¤+ Ð Œ3ãõiPé9é¹D¹rÓÔ™ßaíCÑðûbt cßÝ´f½rœñáZòk+ƒÁòíñ®³¯¬xøóÕË×ódºlò¸] ¿’ÞWçÉ]5^ZgÒ'€—ÕðFúHð%åÑÛfV©@±ÿ@ÙÒ£–‚Á,91KAࢲ™Õ¨5E“ÏJФ%ŒÅ9ŽÃP*_+@Õ0’gi7jÂ!‚:¬j‚bMnB†2/0+jNs…ŒiùdcéwØü ‚õ:ÛG!_ïð¯Žñ†Áú̌ŸP£ãϸ)]!|åT›oÔ SŒù›*{|ÈäÐ|cËc4åÒUÇÑý/Ùt*·ÃÀ”Ÿ+¼-;¶†^¦Tø©fpÈ™&ȷǧÉsϦâ-ÀL›ï¼(ÐSY]j¬ÌO«ßª“U5É#tä´çx™7-Ý|£úCõÒÀYÛÃ6€+óÞÓ°óyÈE8´·ì0“Ц²CéQs·»lvo8z²ù.š–IR±ôwòƒø),Zˆfgg?v[ áPµ?€¬†í4œÉ›””’Š/âGtå¿siĉ`.ÏLÎI€D§@O‚ö óEìnÇ^zìH< ´l¡uYDù¼NŒö¢D¶v÷î#•–Æ}E‡Y„jµ¡$Èëõ 6Y”k²ÈÔ›”¦\6ÃqÆOÉf:AQQŽÑz‡v‘&v€³.ÄäË Ó^ñ#fGã®[V­Ôìð½RoÑUB 4k*l 7¤@˜jêÏ}ñåPÇlEA,_rr\ˆÀíµºSzÁ†=ÚÚ¢õ5Œµ¸M4!ïæŒ‹#×E—Û9ÎZPÚž˜i.»°ê0#=QÔ¼éZŸXþ#–Ê<¶9ÇFX+f‘; e Ë+Þsêdv„ÆÐ ¯JwÃjpoÍŠ —^~ëÀ|*›¿ ”ô;R‹=Œmq ã’´š´Ä!XëˆÅȤÊÕ¤“,¬l‚2Π˜?V£ˆ­…¾hD›È1-!ÞO™ _< ¤‘a-ªÔ’žm‚b¦2ª8"RžèÒìsôü¾Ó§+8{~%é`õ›ÌémxoÆ¥7P‹¦Üi>žs^ëÓÀ¹«•À„–FVWKj?Ÿ¹çÜó]Lá79é üÖÈ\{סn¹¹I &T©á‚Æ,NôFʯŸ°M’Ù¿ÖžWýªùŠÛb¤lÎ&›Ôds žñšuÔt8²µGÈXXG:ør®’ñ:)ÿÞ)#!ƒèÕ) ÙéyHÛÚÝQ×vEÒ‘UM†Ï­³ÊgÐôǬâ+:2Ï„hS£qpÛYGâ0óo™DÍfô ‰vo~0›“'<¦ ¢9|•û$Bì™?—5¡Nz$\×Ö=YHsŠÏ"á@w °¯ 4šŒ»ìpÅ 1ª‹¤£c:YEÝÎÑqü›'ÙE4ú{Ï<ÂÿòG‰ã:¨q[å:aÃ1‹¥? C‚²/ÀÁ]æ:•wÆn4‹F¸ÅE¾MpeøîÐ}êÓ›ôgk–[vן …¢9Îâ’Cöý‡©±o5?­ZŒâZdzÉùZ]yÐ5’"0j‹´;Õ:-D1me‰:#M§MWò£Úž:æ$ê´YÀBV™Œ?BWÁáçD¡¾ÚÍ?ðƒÍz_›¨Ÿƒsˆ¬m&mÓ\`9šcr sròÅ0«è:£ 6 ¿gÞÌó\ÜÇS†“w°¨Nb2@ùëž’ìcÞ„{ʈæz¿AûƒH\–›¢OKÖ(S4œï q›À£6¢)ð#¸ÄêRyeë”J¡G×çâV/ô˜EÛðH2œ.Æ ðh<»¬ƒÇ¢ùh1‰& ¿aGâÞß B”†|~÷8çàRüΔweö茵آ‰÷²M"Ôùâãb4—ŸË"†þ棊 µˆKTE'@QYUSd¬hò®_ÿáäõý9LOø ßø«â:’XhÛ´»†¾‡èðs1ŠD—X4C‚º!êö³†}ƒß‘á—8²ø‹Ö¢F%,ÚiJÐè¿ÙH"ON.°ûN´ú̽v:T·D^Gs®7´¸žé%ýUÂÐ,v"<ÛY¥«+2˲ówVïæ ÑúÇ…ÉR#‚=ˆOÒag‰qOð>u0O.]ºÑÛ¶SÖ¸«1£T8zƒV¡JS@"SXœ[–µ³$±Î3ÒUéîÆ¹Õ»äÀŒ™?ÿW“—y‡,!.Ú|i+^–È1þÛÀƒ™ûãZdºürìfCü±ÕÜúêU°œÔËmªóOßa’&ƒð%vÈücÇín:`},¹:äuó6 —Ùà¬i)²±¬uú’ýÿwö6\WÄx1èùóܾ5Bn"ñ=nB-2Ô#-²›ùüûÊ^ÒW(½dQ> ;!ßdÈÕ×€˜G¨³ žþ&{–À>­¸yn2ßcÉ=<”ßý‰þïS›¾ÈD4-ý{1¡©†í»} ÂbðØçµÏ礲˜›%wª3!/AÉ©Z™ržotl†”cñOn9F_Ø\¿¶ŒÐz­uZFÝ_:\ùáÅ^Ò'èC~9‹û ~š£$…ÌÔá¦:Ý })G nZ¤ÊØ#“þ”ÎïfëÂʃä!¡!¦PKµ¹¼Žt¤¹¥ulµÃ©ÊÐ èó ë‰C®¢ŸyG¿eYVäyÎ8]>qî&zß„§ºdr:%(óÚÇK´Õ:¿¥jµ‰©œïêÀ÷}ãÉü&ž2¡ÿÔãsn†ÈòTÇ/Ç0Ò'U©µ©ÕÞ塹¤B1ó6Í43ãÈznÕ õm£67Ôï³uøÍÒéó²8Ð õÇ-Û/ø<$}à[? N2éUx¶èÎ#qÉÒá‹¿áí^ñxÜòbñE!˜Þ‚_F ~¹‘#0L?Ú%yÒ’´(åŠÒ…PÅTFƒƒ£#C7œ >}ý“ ß’¾Ú:¦sMxEPPxxPPExMMEE ÁSoâõiåhv“¨ÑŒ²¢l³¸uLë4¶ I¢¬çJ…hRQ= ¥D–¹šr( ‚!šØñ8É>ÃÏpˆü{û¬šð @á.¡;–¦çA—Y È¯O²Øà…6‹^^¿|(õ¤vÕ@-ìÕîÕüž(Ïé 'oôo^ÑNQ1ª(r]µîc}9aÄmí:oƒ×q,E µú´LÂ]éWÎæfŸ?Ö¬/Ø3HÁ„²—éãm£L›«€ˆŒO$”ÝÛ;òÿœIÉD1½lÀÖj€-–ÅF¼?Êw1£Õßó/¡¼;û#“%ýè ’”p#äÀ^]¥ëá ÐÏ|´[4ÛÁ¶è}þÐϬ‰Ð*"•D|QŒ>l]ÌþWKè×Î:®7ƒ™´Ñ¦?Taçå­ý ¶n·Šø;laSUÍçºö£\Ù_å¶hËC .¥¨•%^ҶïáõÊBë„ZsÌÖÓÜùÄ´-°­c‡Ÿ Lè…âÚ^9fEe¨  ô…¤-^ûçWÖyþÈó-Ä«jÆ?˜çòúíÅí×o/ŽØBÿ*ßæÀÒWáâWý‰KOïÏÝ#0u &@KÆZXÒáÀÏ´F?2á+Õ)Ä#ÚÜñD>A›®É$Ãpí¨´:pUÛôî¶{NC9äû˜>tIJ¶Ùx¯Ðéþåõ×!=-ß9·µ:’ I@â“•^€¦ñŽè½¶LÛ—N¶o™ö7A¤2*„SGäCêÅS^ÊbG„?Wÿí£´“ÞVV§Òûº“ø*ë +ÍÑß+ÎÂcø.ÀÅìe¨ëõ3ä‹ÑeÃr×ÀLXJ ÌÅ’¸Q«]„#u5‰s“å€Q„B®‰ÑáÖ·Ø ÐGpË–L ƒ±‘,ÕRÉ´I%Žãü€{@¸p/ÎêÿW±âW±‚ÿ‹• AèÅ‹Ÿ,G_ ×om<-½Á¾|ŽÆ"îì xæ„$ãna{l7ûÝñkŽ‚¡¤!_et¥oŠ´jnç§'šÎóíþ÷¦½¿iúÚe2¼û)„'ÜIú‚§%Õh™nÒS]“âRS‘.%¯¾B&¿A;Q—Ÿ>1;4£Î“Fï?žù¼—Sh?zÁÞ©:s™¸í΄S#‡Ì\1ÓÏUc1šjŽ:Cg9r± ˜}'åã5«B]e[¶«Ã´)ÚpM*¤hSµÄHÛŠ,(áHî4-ŽûÎrÝ6«äc_YCšÅ»™ÆcP¸¼ëØ'‹ÙûíÏy4ÜYzs`¹Í먷…w#n?|ƒ¸Mc' p4›Û!4Â@Íp¼½¶£÷ƒg–¡üÌØdµp ŒÏ¶m'™I¶¨%HÍ¿ag5uú-5¡^wŽþ¸Å‚¢Næ×¡­,ê:í¦¶l‹õõáÐÚ¨¬O­#´Ðãò·™êŒT­Ôi²ÈäÍ‘ Á¶f9—§4ÄÓL¤ÄÊð:R ³ô†Ìt.¯¸éÄmhóÖìèbo½'lb¤¯HZîî麕”ÅÕÇc>6¢D¨¾Ä\Qcö‹÷MÝ:éÒÔY&ý1?¤Üó²ò¨¥ÿK‡zÔMnº&üé%ý æï³gÃnó - i4æf¦Òët:`ô¯òO»t™,1‘ìPͤf¨2òî~‰ºphýÛÿÉmVÏñvÞýv/Ì#äùxñ£^Ò6äˆn³4Ó7U4+·›¸8p ¨ˆ3׿ì9ëúÑTüîC†áP™ô6"fÞ$þîoä>éøï±˜“¶-ï‘[Ô_g—BàYeQôa¸¥ÌͺgŠÁ2 Wê  ÈnM1>+¶³Š¸œÿô†è”ù»Çb>„Éæ¤œÁ* ¶MªŒ “øaw‚V†âœóùY™7¡€ÄÓC¢ˆŽEr¿q¸–Á`@—C‡à`.wŒ†“†“Ù'ÒwÞ?ð‰-SÐg§Q ±ùhÄTy â¨ðšr ¿•ÝZ«Ú%̹O›¯\½°aîjÏ-+|9SûEÓ‘¸È<zìÐ÷f‰Ö»4;sy‰Maf‚¯ÏWMZ3¬wÿŸ¾@vßÝýµQqЫ™‹*Í]²—êÆ¢3ª£Âr§Ù³6L›4÷“»gê.|}ØæþÆÇ÷vòƒ÷/~Œ< ²£+¬?n]³N¿+„;]Sb /’Ç&{.<áþ5bɼÕùï#¦?¿vs´Ë6ÎÝ% f1¸ç7ÃQgNúë9Ø×¼ó ƒßü€…­p>º Ž‚PD—7UÝ]Ð@ȳ~{‰K¾úRØÊLÝ>bÚkx}9úx/³%]¢ÕSÌ#'EH ËÀhŒ‚8ö‘ßíÐLI‘í«xž(¹‰šuzC>¤; ­$V PsïáSvxå¿þ)B‘‘eP$C><ÆÏ”ÄÙ¾*”¡‰’¸Y«Q'‚ÊÚŠŠ yî6:e‡VþéO¸/úžµîm¥u5k3DÁíý·ÿŸ ßü²Gµx]9š%ÜépÿeIΟþün铨61KyÆ’Ù©)¿ç×·S£¹¶®#%3ô ÓqYº<}v– A¾©üv‹ˆÔk£Líû–F“èÁ¥¥vö­¥ÈIñÿhqë\t‹}íé¶ï÷ÿ^ϾKÛß~wõh_a+è˜epÎ0×ç~6ìƒñ›ÉÌÚsFÀÍßA¥LwQ_y„QK…‚ÌŸ¤‘äVžqßÉ€ K|—Oá’µ© 4…cBcJÉÑ¢®I{IŽd0»ØÏ+9B_ÍåéËr¡‚1ÅFÆD%†y4ùœ¹ÿðª NW ¾ðGÝ0{\^_—Ãû¶JÈ#÷ÿ° ®í 4JÀ”k6Vù ™ÿCV!Åõ?CÇvÐúñŒÝé¬@Y1%¸'rðG«4g7šƒz ž틨 ʤq °t%·wJÄbBÝ3òp÷“Øî,îqq.0™é†,ÙïK£X²v32°pSº¯Ev+PIŸ&£T©”„ßÊ¢e? ;'ö<ÓÏÃob‡›Ó”éªÌÜGŸ#éÇÜ1Ô)‰ážpB7µÜ‹E¼Ý^1?Ň~…¯Ó¼ÕK ¡,2¢â9\ëéø×ÙJ>Iâ!Êh„²Bk$.>=ö䙹òÿ`+"+ƒåòà`“¼ºÖXYÁ¡uoWËMÁ‘¡ááÆÐÚj“©šë€ð€ýˆ~ØhÅð¥…Áý(ºK@üPV‚» ïü ÓØ!9V+DÊIzW«Ñ@|Ü…°õ÷÷~RÛ˜aäÜ5Ê`ˆeü*Êvî,®ýxmóì1¸ëF,âþ]`·ÕA¶åD¶oØdëªc/„ïóèPmq~fv.§×ÛétÚø%ÁË]eJ%ٕʪØü¯¾B4±ø–mìÈ0‘‰y_|ò@̯^ZÓ²iFòð£Ãÿô³Ñwך•®^ª4.¨iMF0£0³Sÿîȼÿòä“Dû%‹ìè§wš?=R:NKÇh"…¥.¬*YÉÈ`šnŸa­q{\ß>>µ‡î˜‚ÿåÿ!vžêͪØ¢r’Þ£þÝØMXíÍE4¸73¿Aùo»Áçßö¢öáìÙƒ=W4¤Wqžšäpˆy~lic/7ñsKPP^¦I‚ÝóiK—K]¹.§»½aÊéÖí’±[wŠú?:UÀendstream endobj 155 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 4494 >> stream xœ¥X TSç¶>1s´ˆJ‡sp'œ­×¶VEVq¨¨Q ˆÌ3"2$Ù óèD˜AˆÌx´Ž­·Z_kW{Û«µ·ÕN>ní>øs×{ ãkïzo­·² !Ùüûÿ¿ÿÛßþvdŒÅ0F&“±«Ö¹xÍŸg~9_/“& “&Êìznêßh9‘å²bâÆ}Ýã­ÁJV-TÎ6øp æŽBÿÑŒ\& =´*<"1*80(Æaºß ‡ùK—.™í°`Þ¼¥+C¢‚ý|ÂÖùÄ„úÄÐ?ö9l÷ ˆIt˜þjPLLÄËsçÆÇÇÏñ žøÚŒÙñÁ1A›¢¢âüÜÂÃbÖû„8 mtÎÐÓªðЈؘ€(‡uáþQa ÃÌX¾*Â5*Ú=&6.Þ'Á×#Ño]’À†·ƒ6‡ì ¿`áâ—–,}ÙŠa&3˜ÌTæMÆ‘ÙÌla¦3[™mŒ'ãŸ0Û™UŒ3³ƒYÍÌav2®Œ3qg^gÞ`2kf³ž±a”Ì‹ÌXÆ–±cÆ1ã™Pf&“±`’™ïeoÈ®[;ì„ÜJÞaágqÒr¸e¸bŒ¢‚]Áv²Ï¹ýÜÇÃc‡_á6Âô„’­6[<²Çz“uö¨9£F}=:oÌ‚1WmæÚœ³ùA™-i¬ûûADGQZS!kë?!—Âú©RkB„CvÆ¡D¢øÎ.3Y“±¸8µ¢YÎP-pÎhz¬¦ˆ¬v”ésŠõz¡GZ¢Qq8ZΦ9O Zásh œÉ’ÇRˆ ¤Þ’$(¬ûmhÞ§½&TЇ7*l•]ô…¤"7ÜÔìIý5C3ÍsU{Rgþw7 `Qö ûf£1n9Oªþ4¤”ýak‹Ó¬m›¼"yå}dX²"ÃRÍ*»n€¿`-½ŸÙ*)M²| ðE¹Ô‚a*7÷)YBþ2o±'£?wÂWðÕ‡OqO ‰‡Š(™‡]pþh£püL»±Ú¡}e1¶ÀNn†(yëþ»©¢ô\”ÝBgIÓ-ï/蟯:E‰éZí¡ þëz^?æöÄ@œÉ âC|>£7 %ŽB¶â>)%’tÂ!2ŽŒ\ïË`Iµçuõ•ƒÀp9™›ï¾óîÃ’ t7ðdy%Qæ$èµUÀÕê µÅ’lq’I’õúµÚ<£(æá‹¶ÊO0ÿ¦‚ãz}yykÛ…üJàn_^OlÈ$WÏ5Ôpi·Zz8Î-³èêLkgþç( ý¼ÇdêŠmûcå7Ý,q5ÃHA ô†Ì´ü—LÌÀ§„Q÷Û(¬¥PºtNÙÒKƉýŽW&ÔÂûpö#ºË¾l£4Z”ÕJçåRŽd¥Ê/Ï-¸œÈFiÃtQ» qz3|ìñ:M†N§ÍNGrˆk-/‰Š½Úù:UÌÝßÁ°[,®°**ƒ2°? Å9ÅFŽªÄP–š~£¼?–fÉ+Ë-¼jΣ ÕEC¨ ±CYŠºu©M¦N#L&ED§-/‹Š§6|Àǰ0ê*kÜZÐA¶9òU²Ån2–˜}µ¾:_Ø{ Á?¢óÀˆ‚¤Ü´°¯Ü₣ϰÔî)ýÝ{‡Ë™¦P7œBg“M Çà"3ÜqÒˆûª}°ëŸü9EE)TW¥@а^OOWÅQN{Jl­§.‹ÞÌvCøfåPª-=” Úìì%Îv3µl‘:m„ЛÙ;v—ÅyăØ§¤„@Ø öT4 ôyPok+åŒMLÈË׃¾ Pøe8ŒŒµôW+jõ× §¡º´§£V3Ô(M4᛽¿Ên2©ð쯲é/ |#²Ý—-•Ý…Ý[Ç}ƒ2ëü˜L£ÅâÛâã¹|Ò«Â\öDÓ¹º^*ÅÖ@FºªaK¤¿æ»+5\ÁÿRÚaÊ#w¥GâÐ^£Bk;(´nIdqnªôÇ‹Š="ƒ°!`¨ô™GTS¥;,…–¨MÒÛ[d‡_Pž3ÄøhÊÅß1ÞÐuÙi:@XZéJ¬ú7Œß00)fgìÎ÷Õ`´úŒœ¢ÜªrhâNGW‡‡%ø{‰¾×î¿}ãÊ z·…´j•5Ó^ì‚cä8 ÍR%îׯA:÷3‡.ÿMâ¾’¦åˆýãÙ8_:…\AùHKÔiS²øH ^´³€Ng/ŽEUÓQ¦L°~þáOM¿(ý‹®/u Š’jSkR!–JV”Žk ˆÕéÂ5Ñ-= e¶-Whck‡ Ð6Ø@¢ØDHÒ'ÂrtTš9IÔ³X3ðmQ|iV ØW@ááÜ"lˆ~[…)ç+h£Ç?5 ÂÐË`DÉÆB*¤FÑ+„VÉ‚ºÉE.m¥L),1 ˆ3€D~@®8HVYŠŠü¾zSpW†&=•–yº}lç¥ÑŠc¸ÛR­HwÜ¿žØ·X1ÄCQ’òÐŒhž¹f“¥—:T;X\û‹DJ?ü–“Ä“Uv·l÷Ìß;ž8N^H&»/gãtÊÇëp®º¾†Ã›$‡MIôÝB½±yç\ªmêÚÅâz¸­!•aô@ÕTâ=NãËCÎ*ã‰\ò¥Æª²NÔgÞÛÞ#¼eÚvÌ®°hÑT2†Œû~&.ÆÏŽK€¬$mÖþL!nç¦54b’íÏ §- -úvãéÚÖÖÒ:Àá` ¶ú Êû¦¶õ+Ú]ÖJt–÷·â—*ÜG¦ádE‰@¦“·ˆ:)IÈã4ôá,ÈuÕFp9éûáÞwáÜÊù{ç_ۮ߮¹J}Â{á­o4¯Í} ÜÁ5{Ûnר­ÀYã?Å#L2*;Ì#%ÇáÒFÕ õ¤ÏM­ø£ÓèCV$#Ÿ²8ìâš¼<ÐðÉ© Ày7%Ôžj0¶µv-›AbI à}²Ë„³¿¾k˜6›†Ç®(ÇÕ(‡Ï]Ø*LÅBܤZ·»ºàöçŸÃë»wÃë b¦:}^[CåajïGwĨÓÉG„SGŒ†½6K›™\ÔÑÄÚڣǫ«’šv%«5A{øÄRãJ îÔ-Ë\}+|¨§ëgƒ2÷ƃäÜ­!)—8ðà¼zW ]?ºzçTÆÍ×ùÍ'7À,ˆ„]¥ßSÚ Ç 8÷p~!wƒl¥R¬¨GKXè¿•Œ%V;Ξ½vÊ…A[?DDÙÏÍZŽ QEûÿXz¯7ýbx|ò[úZ÷÷¥š¢D¬7a½hCµp¦ãLtþG­r?šð_*¬Ç,.6,[¿þB– 8+N…Ü©óïÀ»ÜS2ì3²‘ý³,ÒÒýÖìîáÛ_JëYe=i9tB}pCuM.ðì¸Ü{#›Úå˾8RIÕ(N—–¢²Ôa±¡°ÒïfµkK2ÞÓ帳¥Þû%ÓûƒÃÎ#œgóm?#±U~…1Ò6q1èê/¶^ßmh‡Sp^kD`«|Z,ÝVu6úxy{7u´›;éŽèÙ+£¶XŽ{1M%íú]Óé¡ žÑ‹œs?-è†J¶®ª0¿¢ü)±±+:x$4\z"¤ðjö”á!m„&hÔ5jͱI¬7hrSŽeh‡cŸ–—Ü6÷ÞM%F¬1Ä õˆ*ÈÑf TzDïž3o)¦êî§t2<«müyvx¾éw£ ÏЭ²ûvßh¬ŒíßqçÿRÊöyµMç¹}W$ÿÕÜœX›v‡xptxúe|ìþy|¤gË;+dx_ãýrU%G¦jáP:ÿÚ¦©\¼ºBqzõWsåDP³9Ï ›>žAç«XóJÔX]ÁIÍ¢ì NE7äåR9FÑ&®‰O_{àPjæ ØOKÖI!¶\èè:ŠÏn‰ð„CvÊPé¼nþÆsÉÕ ­•=|Ž$çòuÕÍ%&àœu_áàîºZ »ˆOÊAÚ6âí¥•攩 茖-ïãÄæ›stI-é³³­’²ýüAõžñÞxÀÝ_~ƒæyaÙÚW¼ë#š»jμ½µ$[Ï7ì,¬îÓœµ{StÄvJ¦à«Õê²ti:må@jb1ã;(ƒ^j˜6%Ò3÷ÔNátQM ´q]!õ{|‚"¼æ³Uè~ïûð41ô†= Óê¹ÁM^§u<Õô޶¹‚ãþr‘Òz2ÕCŠÎ.²¯~E¦‡mÓ„¨y´cOf•h©:~‰–5Ú’ƒê„S…M±Û2½a9l;Y”N»¶¸º_ ]´•§T–r ø–³—Šiý>‚…¯Êç.àD¬¢=ò®øð&C]5ÝèEïÃÑÉžéë_úè%Í+ÄEß~ÖÂà±ÏtGך1äq=ŽqÀ ¶ÊôÀ'ªã, ÂØQp¸Z¼dºN³QS˜•¡\pUxKKumó Ÿ›“Ȳt1Éã•ÏáÕ¢WÚÜÄåFß¡G¾þŽCÇ{d¤ ˆ…Í೜ûÜTyBDKÆ‘ý pŽq8æ’ØûV89ÂJGs'I5Jù8FvKü›ÙétI_©ŠKÀ@ñVÄqÊ;—«+ÒsjE) ö¢øÃ²²\CIþ'p”¦]¦Ø²f±XEÆ™nïzòˆyïògY8—s®@,ºm¼ÚÔ6t=FÓmÂgæÃOp $Þ€Í×3¯¨bYpIߘœäï½Û !ÐZQšs¦º¸Æ˜“þþÑ‘îç7Ñ67ç†Eôð·KI™<ýe2Žðß­ÀQê& šïpD |ÿ-|üãÕªmi¹Bé^:Æs‘Ó–ðf¡Áyh‡ódŸà<9¦IÃTAMÞ?KdGSSOÚ,þðÞÐ÷N}¢úÊ-T7™¿QÀµh;ŬÈ=OgÓê|È_ÿLª¤Y,YIm¯´öÊO‰Ó·ŠÏþÚüe1+¥ŠOÊŒH?ο.Òt¦®¦íž[dM¦ú“á¼ÓO®áKbkZ"m,ÂÝÆâÙUÀŠ#Pñ?ÂbI…ÕðÖ++Tœ´©·²f˜ÿÔyËendstream endobj 156 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 6900 >> stream xœ­Y tTUš~eIñ@ŒB(H‰¾ˆ ´¸!(Ê"¢(&¬!ˆ!ûB’J‘=©¤–Ôöêý¯ö%Ie©TöT°£¢€¢Ø ‚Úm«Ó=ÓíÒ·ÂKŸ3·*AéngÆž™S''äÝóîÿÛ¯€¸óB Œ_½~ó+‹ þsq`† pÿ„9|üÍÀpܸˆ{V­x *óÈ`L¤;Ûïð³){&£º{PÒ½„P ÈÎ׮Ζ礧¦åÍœ—8æâ¥K—,˜ùøc-¹2+9'=1^:s}|^ZrV|þOæÌMÙ‰éÉyÅ3ç=—–—'[ö›ß.ŠÏÊ]”“úüü3 ÓóÒfnLÎMÎ)HNšùR¶4oæëñYÉ3G_tÑè·ÕÙY²ü¼äœ™ë³“’s¤A<³RºmUöjÙ‹kr^Ê]›—ÿJÁºÂøW‹^+N\_’”œ•š¶1}ÓæŒÌ­YÑ å‹[\úøO>µlÒ‚˜MDK‰‰(b±xˆ˜Kl"&6óˆ-ÄVâ"šx”ØF¬"1Äjb!±x‘XD¼A¬!^"Ö‹‰—‰Wˆ'ˆuÄ“ÄSÄkÄzb ñ:ñ Fd÷“‰)„‘'¦bb1ƒÈ"î&–âÊw>Á¦;„w· /ßÉÞ92îѶñ³Ç7‘i䧘‰ÄÄ≽ëФwo›æ½çÑ{>½wÅäû&_šbœò—pÃÔˆ©âìië§}1½7¢@)9y_äŒ5÷/¾?í~Óý]ˆ°=p*¤ï§¯Ï4Ï|–ÖŸLذüh»?ð²Wpjøn!ڌƉã^ˆ3¤U,[,Épånª ÚÌÍà‡¦ŽNzÐZ un h8ZÇ¥šðÏÖ gÍhJ*¹ùeã s•q*%)+ÀLõÀ>s/tæ+ø°­j+5 fÍF:nXÄúX³$r0hu*2“ŠøkãødQXà”ÊXè Ô!´~HˆJÐj±Õg«·ú*WÓVÕØÿÑ šb Öïäï*K¤éL9;[SÝ…æ[]Ð*ÀXj 42Ùü @°žª&W³£‘öžèCw ùéfßö¥oD)(í‘ôö8H…Âä’'ÞF³Ï4¡Å×Ðô…?üaàc Üóè“»õÀÐA”ºÍ\…Â;â·úтЮ‘ï ÏOk25LK–cˆi­z;Ý‚ÙÜ ÐÅøC=Ãl®PCÙ6°¹löfK5½͵—4% ~Ÿ»Ð<4ÿ¯‘G"·Ç¥¦Kiå`lEί ‰äÚlúTžØÑuæê[@ö»6fÄËø»¤¯ÒÊœ¸5ۀܠ¾á0sf“ãÄ0J›á—?¢—ÅštÜZ îíiÚ€6¡t”\ö^LóçW§9»íÖEŸ\å§GIùȧ¦·o|1þY çÜ…°áe™§ý¨³K0¼®G8üRŠ9“Ù éVƒù’WNmà©´uü@òa·={µŽûBtÁ.—iÙR”£Q ‰©gs›Mµê-$ n L’ %À訢‹ò—¹L4„XGgð%%h²MY/ÊŒ &—fõlyp7Xª8K­‰j@kÌÍ:(ÉHª¨D Å ;xAS΂ÖPJó÷tŽ–IòS™¹"/pœÝC†–ªêÑÅ«Œ‚ù×…ÈúÄœ¹‰?+´‚·,MÅd€žT8‹škýU{¯sïz)5&_N©N$Õíú_Š`»½ÍaÂ"8×ö àÚ>ß#DÑoÅŸ‰Î;Ke%9~ ˆal«µuÖPÁ,œ°BâÚ+X¦XO¿(ÏÁÀ_.:‡*nl‚课̗r™"&fåã~±VoüŠZ¹ÁLïƒýæ}0‡˜þ œ˜å §Õ¢U਴€ÉâÁú°@å{çZtá¨oʕڷєwv\ž~ŠT÷«a ™!{vNþë19uÌ ûÁDÖ)* ò²Kb_L}=ҋ¾X^›à¤Â̶¥:Ü×XáZK:ÉMe42P“*[IcMsÅ„½¯ð¦ÂoÂóçËZuXKªòý¨ÌxhLN^FÉb[ƒÉuÈJ7Ô: Ör:Òq3ã ÖÜŠF…*7xh‡¨jØJ¶žq°P0j²ó7ª²²ã’6+“\°ô]$lDO|‚„ÿª !è„Iþ; {êûYhö.ô´ÑÎ\_ªì`²™ÝdºUVâG«½XMN Qe`Žø†¨Æ=VñgE=hž¥ L¸ÿõ%VuŽºûuÿĸ§E%· }M´šßÇ r(“Ät&ÿó±1wW[¬¡ŠßH\±H4UÛ¡ÚÞŠ[×9&í¿B‹‡„W`†Ødà 3“n§0K+Õ©U|¿¾Œå IQ¶4k'±å;‡:ßtvÔcc¬#îòÇ-JãçéFI`ª²PË·ƒ@ºÝÌî\µ¬,‰Þ½f÷Jˆ&Ÿ”ž<ÑÕÞÞHÕn9 «ƒ¨oòøk{+1(B¯Ê²ù:ªlSÑké@ª[¯*U¿wQÂÀô;:l΋@Ö»BÂl !žI€ˆ3'Œ÷xW>#¦ Íß?Ò§ÙU®y$åP¬²í!€÷A`ö²U^…ëÁîæÀBwÜX‰ªz->" ¨“ÄÝJ¿4/'/Oa×Ú´TãnsTÙ;x ?~‹êhç`Ç-PÖj›çWV'óéÂHØA>vNz =7€þÖŒMÊ—ï;ÿ§@™_ðÍwÂ@öV»£HÆ„SgÖÐX˜tˆ…æÁ—ZW¢À°7ú‚. BÚPÆ“#)E+6®ÙŒK”Ö¤é3±&Œ=²××ÕÔ\XYÀDÁókO&žûæ}åÃÊ¿`l׋‡Ð¢!”·}/zN¬Íf  ŒTe>³u |Xy²ùCÿ§”½Îúk·Ãß»qpKw¤7çvd÷35Øa»+ÛyRöÉNÁiØß×þ~û…½hœ ©Æ˜1»$!?:#áM cä}û\£û ½=gó¹öaîa+ÕÊ0lGsË ¯Ñ7Ð CÓÃ?F9%voï)èþÕ¸»€šnc]†6]ÏRáýŠ2ª3fðB~"¿€ŸûôÀšèð÷ÀÁú}GÈ“È.ÞÁoÐd¬_¶$;4§ÚúÑ„ÚýtÓÐÀÑ>Ä›¢ã¬Þ¨ÃUvŽbúÌp¯0Àž£×Іº ï¼k“885¤Ý`ÕRÃäA$›ÓoeXƒ­t[¡ÇšU\R*Wð?ðWÑŽ‡ã,HêP;–pþM„sÒOù÷Ÿ‚â?ê~/ô™÷Â>ècö†h ¬Z(ܧÍÑa­¥»ÑDãÉwØJÁˆ-#™U)dµöš{ OBm!ÛpXÁ‰ëèèua@xDœ¨†œ¿oÙí± ÒÍ»¶©©üÓÛ9@>¼…ŸžBpÕVºåóã{÷ÙV——˜«Ü^žB«XƒêMùº‚g#\(©jw8›ÀC6Ø99»’öç îìm¯¡¼¨šüîºÏ;Ö}¹ž–=“d¡¬âx£û€£‹vï÷œqtVoy똡 ªËÈ[‚€Rö Ðøwì=!z{x±x4ÌAb2j–e¨y%«b¡H¢ÂãÆa“VÍ0 `©y#4˨ Y²dÿÚËèN4=‚FÇäô‚ŒtŠ çìªÿè ³» ‰h{#˜<Þ¶}½]½¿ÆÇA(pJ˜ø^ìì¶¹>Ò5j*,ŒX ˜71ëã͉¿¡”´BäÃ)ÉVseDgqã̇ƒ£d™"-•ñxOò£Úmq1PÌÇ{N\ðæxU¾)ÿ~C:äy{zøh¸YœÌ/]¹3È7Êz­N_í·V´@5Y/w+“t›¢ûÒßúþ«ïþ½‰úS€âl&ìA’Û=èиð“j#£fü_ígŒð7VVŽöõ^ínâokw‹ct¸ZFRî–7´4úzÄõ¼zkÖ™wê•/ÑÔÏþÃcÕ¸5–Uk¨• 7”¹%§ðDË͇é=çv÷cî7zȰ›SGôîÍç„7'âQ‚’A‚â!°„æ§Šò1\UÅü×#g"6>#¯„—üÒT7u¤¢lS²2Ö(ѨØârc§Û¡ËìêèØ¤‚ÉZ¡å”ôF“±<·6K5’Z#8ÑÐÈw¦Ðz@R V»Í˜0,À?ægœýG èÁî,LÏ?Jö±ÜwÎíC>÷9 ¿¹B4Œö‹O ¾Π¢ê0§”:j×F…^Å v5+«K2€\·°°ÜntTp¦ +íðµ5õ€j5>ug)ŽdPZR®’Ëùgø%›6¿±‡?µ œÉn¡ªê[ú{ÁUmNu¡§í?žDÓ]£ØTèâ˜7Òã äöBgV¥ÒRzЀžµ;ê›Z[É[ãlˆ¿!_Nþ üÝÀ+Jý²–xÜ ÛäÏÁú#ûè_o-[#Õˆ‡Î×ŸÝ ä?™„ØÚl—Ü%…"`XƒQ7†U™=‹Á*ºˆô½B¤Arq;¼g>\wµ΃HÙûêlþÁ¹üšÅ‡Ÿþö—ºNõþÞ^W¿½…¶ùÄÐ3ß™]@îk¥Ïâ—Â.: âã}u×E~´ð·õ>AÎutGí8~*ªq¡v¹¨êCs­ç’£Ç}b~ßzc¥Üé>tÜ/#A×†Ão»ÃdÇŸ¨òÜŸ`¡Î°hË:K“·ȹ"Y&F³÷\ø.“W_¸ÈO¦øš_ ‡àñÜÞÿçQõ8Dˆö—¯.cÊË6Ѫ†2("Weœ½Þþ£Ãn´)¬^ÉP%›²22ÌbZ:šÍæfÚÜ`=õðvÚþmµxÈ=«E’O¦ ©oz+þÓéát=(¾­JRÙ£¸Jš2®¥¹²³²‡>‡^pµá®ÕIÞJnJY˜ÎÏ×ÿltÎÎæ¯CF§ÏËUIµEtÖ3Š]Jü‘”%0ßȤƄãßt¢G,Tøïô•ï~¬U€Þ¸€v^OüûÃ8=ËâI6vSÒ{/r…Xõîáøk|´oë Œ¶1>Çp¤,âx©ƒ©)v—¹ó!•|lùïX±çûk‹ÙM‰3aȹ˱%ƒ;Œæ“Z–Á©Qg5ØìœÉå Üͮʞ-çJ‰Y-D÷¡)mÚ¾ÝtŽ_i^ÐZdÍtêMéž’è#?ýøw7¾i]»r7«5äSlÙ¸Û÷†w‚?À÷†U=ÂÀ<(‘ãনòû†²õxNRGÑF®FR,¥É\k¦ÚÑZÎVLÄàÄM¸D…Ø—uÅQüÚ½"ôǪ 0[,`6S&ǹš¡Y­õH d—¨ Z­±Çzþ¾‘®ò´rõb£³yXécpÂ| ŽÃIæØØQŠ×­†RºT$‡LÙf½ª°[cÀ‰½´åÒ'•£>tîÔÿÛí­Nu íìÑŸ+€ Ђ)Ã~ ÅöºúPNÏA_g›òõV ááÿg×Rl+Z‹G~™±ÁÛ`o嬴µ8¨„ kZ^s6TÕ×·øŸÚ;p³Ò®âØx­> 4¡ïk­ím.îŠÝŸ˜Nm|E.O+Á Àª4,Þ/øæô¿¸É#Øõ…P*Ùº'åp¯·¥©’ªé®:7ö¶zèJ›S²¦(>$?»lذ<¿~x–_€’‡ç.ˆ÷½×íî1;87àðh2:UT.ȘÈ‚]féO÷ p°t5Û¨ÅaL¡*4hùý#…ÛõãÚOp®ªÑtÍP)È$ÿ|¿ »_8j*Ï1ѧGFLEœ¾$œ+ÀðàÏ·5¯¡­asLõmG ^r¶•Ê´Rc9½Ž¯ü§K”}èÉÑg$ÿôÐÿ^‡¯oðîX³sg¥¹] ‘°}Wú›ÿ7>7KÆztsžð¦*ðž¸ëLcõa‹‹³8ÁEràRRy°›É… \Ȍۆ\;]Ãvé°…*õå:oI‰È¨X.d˜Pc«;Ï‘Ì%§Æ¤£S ‰I -‘\"¨ÑXà]û-ºRÍÌëÅQШSóe#+ð:el=Ûô'¹ÕÌ_\7³4¹Ý:ò7®ÀdÀá‡j«ÃƒZáÑsE « Æ ¸‚?Œ"ß|ŽlBôŠ×aE1W˜l6ÖÓ°ðÑXXŒ`Puú¢àñ²jƒ§Èá|ˆCìq;¤# D¥jP…röøQ€Æ!åœûäõÙ»²ví®-iîòw¶É÷ G`tM8‹ÎŸŽ´â}z4¯~v/Ф¶/ŠãWÃãd&ð4ºïÝoúÐC穾¯{Ñãð ÉÑŠ3⼿ýô#t¸ûºç£“¾. ûöd½ºr ¿'n¿#f[434ÖØl?xØañ-;tŽeK•ÔósÖh¥Œ>Èä’à˜ÅsÄÖ}ˆâ'ŽH5õrhF¥Ñ úô›ÀA`þt§9­) „Spùf‡¥ÑfrrôàûB´àCñÑÏ/÷ùkÑ¢/¾:HL¢ñËO.çç-åc×Úâú¯îG“zkË»:½^‡5Ÿ3R6¥EJRš2ëu¥ÚÚzþ#…ÂúÉžázrɼ(*VŸ½f‰j °!jT˜L8-9Zk÷´ã¹¬µ°©¤¹ØŸ¿Ûø´Ê}ëÀÁ@—%$½ü2dC’CnIu$93AåF#K†:ùý54ÀOð­ÿrû¾KøNQõ¸S[퀺UT7h\V\>/:’ÎN3'b ¦ÝøáCexÈjUuÞ6cdñÚÑ7°«NùMª3\~¡›ú¾{ð+¸BÞXwd9×ÜÙO§ÕI{zºÖ©üÊàÎÎhÖQkW@)™ž‘--ÔÕÕ^ýê›?ö.^-WóQüÒÑ{´zôD{}{àŠWp0°JXŒ¢Å¡8k&›s9&K·»Ä@1êdþV·ª¹O!u¥¸K ”˜ 5 ó º̾R]žLKíâ'ïä'½È“ÏñS_c`´LF ÈIƒ]á°y*nÊépÚOA-»U•úFµO“ÂÏ0Ï¡ dAÞUWén°Q=h®u€ $­ç+«­º \X0¨ébmNÈö  çBunEäçh£QWõ£¾„hà-±ÙÍquÆÅÔiŒ^ܽ;¾üÌc×™ÆÀõtla¬& vB’KZ­3L¼¨QǨ^à߉`VÔIœ í&³ÝN;ý–³ sÉ›¶e³tvA"½sYT~ VÕ×½ìͬ «ÙU×ÖàUTåEë¢2Vxås4‡ hÁÕWž|3¡Dš2*(ûÇ)A àh+¿Oÿó/ªAøÐ'è·â¹7;=3#3¯º°µËßáæ½Ð€w9tÝð൶Ј÷õéxÈkÃCÚqÛ-BÐk{ê›üê>*½‚Þ Û˜$n-”P¼FTûµ÷ß*Ï¢0Ü8Ä4åì,‡¤±¾`:øˆŸ9<¹2‰Ÿ%£ŠŽ[÷óMÆÑKü°z»j»r»D»[±5‚g"ŠÚ˜w,êþýA4=ÔJ×-õèø¬ú`y„»›ƒÙz»G\ú|Ì«;•ø­1>³Ö^‹îAw£»:Ëš3sÒ¤ êzoµÛb¥Âò}Õn”kªö‰øx÷xÿÄ wQï\â4Áçœ4‰ þ¿ÿ´endstream endobj 157 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 535 >> stream xœcd`ab`dddwö 641õH3þaú!ËÜ]úãïÏVY~'Ù€œ=Ÿùºy˜»yXÖþ¨úž'ø=ÿ{’3#c~i‹s~AeQfzF‰‚F²¦‚¡¥¥¹Ž‚‘¥‚cnjQfrbž‚obIFjnb “£œŸœ™ZR© a“QRR`¥¯_^^®—˜[¬—_”n§©£PžY’¡”ZœZT–š¢à–ŸW¢à—˜›ªq¦„rÎÏ-(-I-RðÍOI-Êc``` f`0d0f0{Œ…Á…¡Ÿ‘åGßO¡îßý¾ý0ãë‡ß/ßaþþåûVÑï_Ÿ}gù¦ñê·œüïÞ¿^O¾oeû.÷ý£èw¦ó~f6>ž¿Ùå+~ßÎÆ÷S¨ωï3ÀúoÝaþaþý–è§•ç.ußáx÷›ñÑoe¹ß­½~¿øÝˆí{ÀOØþ¸ÿàý>ã» ûw޳..Þ¿¹äù~ì/½ðãýYÆÃ·¾o¼Åü#õ§žhÈš9NÝ¿åM45f&nN”ß·½þX탲©Ë‹æW-ÌêNàð ñ4H2œzÉ\®ÆêqÈíÚSÝ’ß%žøÎt&gwébù˜½YSôάîÏ]Ú21E÷vŽc?¿¿;+gƒ_éÂs¾‡/œ½íwÒtö\7¹å¸XÌçópžèãá¹9‡—¨7àüendstream endobj 158 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 2136 >> stream xœ•UkpçÝEÖ²#Ú/I©›hÀ&iã@H6¸…`×áåPÀ`a K–,K–_±¬·vïJ–dÙØXØ–-cK2(Å5¯IšP¦™>RIš„0!™¤3%ô[³îLWØíd2„™þЬ4Úï~÷žsî98–2Ãq|jNÞÆõK3“_3ùy8ÿè~¾ï™ÆÖIçc3³WÎ/ÐŒ"¤J 5%öè´³Pû#¨j&Úó}L‚ã:“#G§¯1¨KJÊ{*—.[–õ”òéÌÌeÊUZ•A½§¨\™Wd,Ui‹Œâr£nZe¬Q.XQj4ê—gd˜Íæ%EÚÊ%:CÉ‹ ŸRšÕÆRåU¥ÊP¥*VæêÊÊü"­J9Ñç’‰GŽN«7Uež®Xe(Ç0ì‘r!·Òhª^¿'¯XUªÖ,ǰ_bØFìIl¶Û‚mÅc«±%Ø,[‹­Ãò°yØ,,;Œ“8L™;åŠäg’Ë)RZj&^"þ25{ª@jÈ‹¤À{düYˆótç ®HP-ú‚ %¤7º4Ò‹í.½­Dи´ OSÍ2 Á¥I;ÌóöAœe³§SÓýqéV#D}Áa_—"x2x²ù¤0ôÛёءáŽA u™·×3¶Wja»SUy@‹õ\Åßñ¹¯CqàfS_`Âô{Ûúr—l^UQ/—ñžÆ(ŸÁß¼ŽNýMÂQB’>0áG‚LPtú{Ï¡D# ¥¢ÇåB«à tY-(õhš?£½þnSÈÄz!Åh°Òµ ͆MšB eü¼ Ðí%ȈnQá;‚ÌVò ë¢Áã6C&á`Áá©Qìº*·kÐ,!Îñ:wÊ;ø„M0âÞ®ÕÛBVù–@¤=Öø»á Çoß’Zçh¬3©:Ú¢½]±„¾§øåm[ +å²1%ÄÑH:ŒŽ-’ ºJ}H\ŽY k<µâ¥Qâ$:(5õ™æŠ ³ˆ¨ÙŸðF€F$æGw>ÖÕ¬E|1„LÒñSÉÖ"Ä;ûU`›âéñWcY´)ù×ßU8¿†O`Ø3~±ÂÒñƒWÄ_ÆË#ü³øGçÐÆs~ÿ<å |dÀã·»h¨—[l"Î}ý;¡H!}ïÒWÕÆ®ZÅ!cÌ9b;i{Óù¤Ú°ú±ma4Ã)g;Ào+°OQ˜›ÆØ€7éhrý×Õ-øÛÛüþ>õ°õè‰á»g‡ë¢5aEy¬Â¿¹eGhc+œ&9øý ód60 à‘ÛB¬¯• „}¢@ÒDÌŽ£çºp$ÅìàL¡tb’!ˆ¡õ"‚u‹·ü$Ãã1¥}Hìú<5b ô¢öØH<~ø ü`ÜÛÖöi ,ÙÌk özc± È"ë±V®‰ ŠÐœ¡¡Oà臗%|?ŸO ”•z–eiã6€ç¾TNµ§JQ%Ìrjì…»vÒee«Èlv¸ÓÏurMŠKè«´š„Ùn“©²n„j2?¾ëŸ7ó×–€œc_ËB³Z¼p€ã8ÚçmŠŠdXü‡¼ôÛê5,ãûùßIPí¥\ÎÅ mö ýîâYÖîyµ‘ýÞC\? Á³ÿþÊÅ\-½4tìïhnþô³´7­-5à¦Áj«Õ°â®Ç¸ ÞqK³‡’êÄ·CT÷Ñž€þ¤qxó­´»hʹÑ3þ~ 'Ö1zV%°ÝkáÄ.m ³PICU]NïqírØ ¾L³µ%å­­¼\³¯ ‘ìZ¦Ê Ø[yÿ¤æ"`"%´lL ñ1<Žò7%cRÔJ€í¦ «\A誡Û'êö€âS~Ð/ò÷Ð B”t‰'vC1¼%>±à$ ÝÄZUÆÆÚÏŒGþ·Œ•md¬$ñQ4øÙP¿réoKøøÙT!ãÚ V²¶Åöw¾~9'¶V˜/¤ sDó!Þ_ù4÷ šjvx]v–±8åû~¾aß+@nË;‚04ïOhÎöKLé€BvoÁÄ8GïHб')WËéÅzžx@‹Yãw¥}Çã@O6x„õBÞfâì}J¤•¹p.è qÍ" )'ÚÁ¿qŸÅkmº!3úzL–Æ‹ÆÿüMòýQô€cpŽ&=€ ‰,4“UÄú†@ÄÇÄr7ù˜?ÑÔz èHOaQ nh4 ‘6± ‹E­À¯\“ð¿[G ³ é Û,ʬê‘l …ÙBš0MxB˜þ÷5·oœ¿ *¬/#ê;µ½®ŸòT[[¥Ù»½|+Åk{ÖéGÒÑwÃgÏõòFO¶3ésŠ î´Ý(;‚¿uÅ%]‰ê¨Þ¤Ú:}耉¦­Êæ Rñ3GHyù×hæÛˆ71A›q[]òW3V4ŠŒüꋎˆÒjꦎ g>?Ð äǰÐlbn½â¿é™„  ¼] |E]wÉèzvKþ>³œå.¡/‚ÌOÇÃß ”‰`ŒÂy&1Éïÿ¯ù­Ó$ëWÈP,i'ÐŒ»8Êý— I¨žÚ­®¬²¤¼»¶/>0— K… ÔD N¦àù¶ïHÁwÞÛ Ga˜éŸ¼l‡z'Òä-“Þ¿p Oàƒ¨M‚N=\*ßdP$ ½pšzà w‘î.~["ÄST¤®W³O§Õš:*ãƒý±˜|“p•JÚ«Ôz­Ö6Æâ}‡å2S„Ï £òÖ@„vvmº|ZJVWê÷"¡ÔT ûCŸáIendstream endobj 159 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 340 >> stream xœcd`ab`ddäpö ö44±õH3þaú!ËÜÝý3à§«,¿“ƒl@Ξÿ|Ý<ÌÝ<,+¿úž-ø=ƒÿ{ª3#cAy›s~AeQfzF‰‚F²¦‚¡¥¥¹Ž‚‘¥‚cnjQfrbž‚obIFjnb “£œŸœ™ZR© a“QRR`¥¯_^^®—˜[¬—_”n§©£PžY’¡”ZœZT–š¢à–ŸW¢à—˜›ªu§”vÎÏ-(-I-RðÍOI-Êc``` b`0f`bdd ýÑÁ÷S²{Á÷='¾Ï8Ìøãÿæ5ßo‰~WÜtée÷ Ž'Æ~ Ëýneû­ñ׋õ!Û¹ïY¿³}çù}‘õ Ûïý?…D¿+³}gÝâîbþ[Qþ·*_éÂs¾‡/œ½íwÒtö\·¹å¸XÌçópžèçáa`íÄ{Áendstream endobj 160 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 2148 >> stream xœ­UyP“ùþ>Bò}(‡îúÕcõK\\×*xŒõZÜŠ h¸ áL 7ˆÉ›rª( ÜW"„3ÞV9¶ëê2S×V·ãŒîŽºn»:þ>æg§ý‚×´¶3NÿÉ$“w~ïû<ïó¼IØÚ$IŠÜ¼v¸ºX¿¹r‘Ü|npàäĤpá´cÛŸ¸ sÁ^ö¶½ó§mÿ5ÎD':ƒ¤"鈛2!M-RK–…"qݰaý Éj— ’íñ2Utx¨B⪎’Ňªùq’ýÊðh™:M²ls”Z°qÕªÔÔÔ•¡ñI+•*ùgŸ¬¤F«£$ûdI2UŠ,Bâ¡T¨%Þ¡ñ2ÉÔœ+§>Ý”ñ Éj™J⥌©A0 e‚*Iœª EÊ£¢ãâÖćØKì'¾ÄAÂØAŸî„±“ø‚ø5±‹ð"¼ Á¿àÙ l‰xb‚ô"[l¶ÙÙÜlXl=mÂÅÂá=QEPžT=Ÿ6ÛÙpÅŽÜe°pù‰Ðz$p4ÎÜIº‰nþa{¬îgÞ–/”Rãå!ÆãRªU?jè^Ѷê.Ú/²x …w-kë ×°¸e—”2éÏzaFµõ¯J(G®¸ÀÌ}h"[Ñ,´Íp±HÉ …kãy˜Yµ/Á‹ŸlD¢9ß?@'b/&dgró»` Œ`ó•U=çΞƒèRTÆVÆÀa)øåEçEDªå@óhðA Zhâf˜É<"?¾jÀÎŒrµ'sz€FÓžÿ}ŠœWÞöA¡YŠHñ…Ý­ë ªµuÅ©ôËcïCõà†˜¦Žî¦! /™=±–ø<×Ö®;r[_ù£ù¢æÎI$|Äêti  Ã:SÛÚ:Ì¿ñi>šôE2«ûA„7þ3·Fýazà·`*¶·–'Ž-0qu’]ÈkB€r[˜:¨LËOÍ=šÍ¦W)+d@»­Q-<Ô’X“#îŒíÉ>bÎË/Ú'(3ÈMa¹™ÎЕh€N‚µ“”Žœ.7@Íi¶¬ôDeY‰YÙ'yjœz¾¹3בu\Ó%/«VùVÁzì}ÝÃ2Ï«Ï4ÕÝegyÈó ÌÈ©%XÐÖzò;$úÞ"@þÜB9X°Ú/éÖ+±hi-ÿ·P˜¸ŸòÕ·m·FÙkR_Ê7#J_Õ±üÒx€äµ·DW9oæ )¾”[”·çnÝ™+,ú‰Âžïèʦœ¿V>k¾ ·zXýNr?&´P^G¡WÌVË+-½À„‚;Ð/­Š¶C«­ÖY‡®0O·ßÀN1x¶Ö5بiéèlèïL>¥Ó³ ¥†v G›ïË)¼¸Ï> Ó¿z{g¬Ï8ÔÀ¦ÀÁØ~Q} 4Öçrõbi ‡‹®2ëv%+ÂÔ}=ÈÍ©i6ð”5‚½°è:20OЬªãeå÷`®…ŠÕ†éÒ@å+€¼¢uµyîËçlC„pØ .@2ˆÑË Ö‚[š…½â%YšÞKs¥T»á‚ž®k¦¼K)ôE†ìŠûÈvÎì$ ”ŠÚõ_š¡ µÝS;yº}xw;™É.ÞÝÓÐLºÌ­`Žäçf€š~CÄÍíÍ›ñœu.X‚—=ÛÄ—Íï@¶Õ'² /³H—SÄMMöÜrˆ;“Ö›Ö«‚óô8 OWk‹k_Ëd– í½6ÕÅϪ”áÉo•âCÃJ’€ÆÓ7áexÅO[‘ ¯÷xk·8ŒBžïÔñ ‹^>cÔòHeD‚¢)»'·Føc†+5Ɠ殦>èSf‹Œßs$¯ñÝ]hãëËõTÀ-G ¦©\ßúHŽìð*̸®á-ùó:4ÍzÞpZ…éÚ¬Bq’÷Þ”p~[@3GÄ]¶úK%ˆzôSÀ‹ÔbGtcj‰(ÁDþˆˆ[è áÍóزì¹ÙýA_]R ºR6ïhV:$Ò!FMs{gC_ÔÛªÅaXÄ.Ý+½_ü‡£QŒMhÅ&¤î#›ù»Ê"¹þ‘I4$žsp¶¢Z¯?~Ô”6ÂÛÞñÆÄÝK ƼjqMEué‰R]±¶òhÕ©´ææSug3a™òâhÞPªä9ø¸o © oS‹ã‹#!r ½±Ô°>4ürG®_ÝF+ñÇ€i¶ü¯ŒŒ_¼c·oƒW€?x{ˆ‡p0sw =("Æ‘³lxxtN‰ßÄ‘uÉï'Òç‡Ãý•¬=½º×S{yñŽ$ÔÎ/àçk ùÍý”w‡†DE…„£úM݃,ÞÝx¼=Ñú †~h‡KZÓÔC^œó«§¸G×^½<¹‹ù× ;ùö|r ¿ö ß~òw|û¿›ß;ix6Â/Õ<ÜT nà¦ØCÿŸ²ø¿‹ëËoÅÉ'ÚÂãâbx\ùºEÅPZ8R’X‘Ù¼ì±CŠßèñ=¬èÚòïÿ¡°‰—üŸ(8ƒ=¦Öóüé©ãC¸\«ÉЪù Î)M©Í.³öË —ZuÍEZ8÷zu[Å!F.Ì·–­ÜQ“®\™Q™y,}¹K(E…˜\Λ£ä;žŸ>‡Þ)slåÝÏícð¡s«9ŸJÔPU-Âå”e"¦³Ól××ÛÛ™ËííÑmï ·w$ˆ˜ù[endstream endobj 161 0 obj << /Filter /FlateDecode /Length 159 >> stream xœ31Õ3R0P0U0S01¡C.=C Âɹ\… Æ&`AÃˆÍ ÀRNž\úž¾ %E¥©\úá@i.}0éà¬`È¥ï 43–KßMßÙÙ È °±Ñ÷VÐÊ8çç”ææÛÙqyº(¨-> stream xœ]O1ƒ0 Üó ÿ À• ]ZUm?eÀ‰Búû’:œ¥óÝÉgÙ×mù_ÁXÖ·$i²,Ê ´Åx°¥`ƒøJ‹Sƒendstream endobj 163 0 obj << /Filter /FlateDecode /Length 339 >> stream xœ]’Anƒ@ E÷œ‚ÄÀÄ$R4›t“E«ªí`0‹"dÑÛ÷ÓtÑÅ·ô˜±¿Íxw¾¼\ÆaÍwïË”>mÍûaì»O%YÞÚu³¢Ì»!­¿ä1Ýš9Û_›ùë{¶¬ßø­¹Ùî£8îýS±%¥©³ûÜ$[šñjÙI$žú>f6vÿŽÊÖÑöÏ«)nªÚ2-nªÚ@ìã¦ÂZ`QFND.‘vO> stream xœW{tSå–OÄ”‚Þ§U9§..÷‚" ¢ õò–GÒ¦MúJÛ¤MÓgšWO²ó~§Íë´¡OÚ¥´ ‹òï—8‚8£®q÷;åT/-Þå¬afÍÜ?²’sNÎþö÷Û¿ýÛ¿Ï?ŽÇçóI]³få‚ùñŸ Ø'ùìSãØé‚<Îz¿mX3a:oê«Ë§§å»™ HßþÔc¡/EíSLãMâóe¥:Úbsú‚±TY‘ªDš#Q¤ÌÊœ²`ñâEsRžŸ?qÊ+âifFaÊš …D\¡Àù)d™R±B•2k©D¡(Z2ožR©œ›Q Ÿ++Éyyöœ¥T!IY/–‹KÊÄY)oÈ )oeˆSÆò;ö•*+(*UˆKRÖȲÄ%…{ÄŠŒ,q¾"£HZ$—Ê¥9ñb|•/+\Àãñž*,zmUùjÕšŠ·Äْܼ‚%/=þò‹ ‰É)O>6•Ç[Ë[ǛɛÃÛÊ{Ž·7—·7÷ïMÞJÞ*ÞjÞ[¼©¼iÇðÑý³ 7ψè*½ „* ið…\'i¶øÞqƝ’Œƒ±7Ù‚®6³­¡ãÄío€pgç®ÙÆý¶r7¥Q˜jAMìè–¼sw?ú½Ë£ºè*©|3wÃ6 *tž¦˜»ÙÍPá}hœ&.îîM]°m•¢‚4œÚÓ‘éPžU§ª”V§A=Qç‹ÇlfdðrähoMn­®ÙB%²Àmî@sFg^ p‰"­Ô¨ÛD­ʵözÕmÖVØ]t'¼›0Ã·ª·‚ÓíëC‹œ1Ê_Ø*yD GÑïÑî®Z¶u{NV.Uwrw4{4J%­Y ôÏ&Њ´º¾‚º 9[ZZC ÝmÎ ½Í’´´\’¾¾=7ˆ•5×°X}TâðGÐ1ÌïàïþR€^ø‹¨.˨]бÁY׺!L¡/„( mE9H†ž^ˆÆsÏPÜ-aùØãˆ?d2ù¨F£ƒ5Ô›hº–øÜ$rw[¬‡ø«ðÛ뛹¤mJî ¹qTaöÒeÜT ¸ Âj0‚Þ\g¦Íà+Xì!"qø«±DX;Kd²ÖǘC¯ ¥¢'“î~|ÐÖnu™=f6Ú©%K¡„–CdYsâpêjA©³ÓNÅݨUBoy‚;‘ôúÄÝcñ\…äNÙ­µè)¤Ò2(„׬…£/×üOÛážéJú5&>°¶^¶B+Æ„?À£@8ãÅÔ@y¶ BNcm Çý'üÇÑlD$!‚c|Eñ-'ì6iÔjPª}`!{¡ÏÚ‹#œ¤™¿B j¹Æ]ÕM$²©ÿ-”qËÏðÛî ‹wèKô¾È±·éÎ%°C3Dkde ȉJoy Óê=ž®“lVÔ’%C›<…@ÌXωJâTqY-n Õuo(ÖD¸[³.¿bE6—BU®‘dì=( Ì³3Qh%ZJ]U%Åyéï*û‡z.9¬¤/»'ë,?œF¢X¼‹4Æz‘ÊMySžDVegîCo?æÒÒ%œÄþ»ÈÝcõÞ¢ÝaŸ T.,¦s¡^µüR/5&˜% Æ 2ºMPóÈëFÿ¢ØLa3˜Áüg”Ÿdó‚Ìqh4&C¥Ž”ÎÛÅ ¹W€xî¡Ôbk¸ ¦ïfߘ®ð.xÏ Ø öQ‘-€ã8‰Pµ»ª®R_¥!m~½JW¬+ÒíäV$ÕWí~}%o«û÷ žëºI5ÂkGˆ¼ý[çds35c]hi´“-w{Û1´® ä•Ôëʩ̧Uëañüù‚3»‚­­dÍ–Úm™›‹¥âÂtLÏÖŠƒà‡€Ïãöø!LtÙ깕۶ögŸûá‡ÛhB+™xç ÷ŸgÅ9| ˆ&÷¨nTP\’°LP[>¹˜ô{žn2B1$sÓÿ—öä~3âHªË¬È[B†“Jã ]T't[±.A½ïWDÔQëÍê4Ól·zÑSl_’YxãGž¥ÖŒé›ì‡×îdg OÁ·¹#—ÆJ–üK—P­ðÖ»Cp˜>êøU‹ØâuÁª)V_t1ÈǸ.òQòuÑÉÁ¶XO˜Ïèéùækæ(ØL7£Ž6j dÁö²J9È¡¶¡*XÞTÉbyúÜ2ƒUïv[ì >2hvAº&mo•öBÎîE©å• %½¹B²t£šé²X¼²©»½¥oÍ£‰ËB•m@|7ˆ~Ób¯·k5F½Ž¦ZUž JP9ÊœR•­ óª44ªþŠI³×DžÞPÇgVÂî3‡Íà©Á,—`–KAbÍoZg¹\Ôi“· J6V™hnüȺ$®˜ kC&³ì ¡R3Ö öó¸ÿã 0V¢d½ÚT®‹‡”È£1«©u-‘%JbcI6=íuÈ2r#ÉYa5ºÁ `m±šAE z‘Ṃ‚ûh-ªµÁ'–w£—Úíg @ô‹½Á ¸gŸáRgžYú-šñÅ=ŸCïÔŒF­ž\:‹ãC-/i¾èô™;> Œè šóCÃ^ Þi)”ÎàþöàªVãn ßd‹:ÐÌ[!†âú莀°¿ÝF<ØûGa mÑÀe‡/Ù6DÁ†+Víª¯“@¡Õ†÷¶1ޤ7í‘®VÍ* õetœ{Œxãv‡©!áòJõLÅü-\%P:bu÷@Ѥò•–Šwö˽ß{æ@ˆì<ß„¦˜†«£öe÷ñѸÃ@½ÃËDñÚ)^Õ‹I­Imã7µèêHZÉ͹Z?*+ÉŠˆÁÝØè ØÉð§ÌWcö#Ùâpí3;¬îŽS]-LcGO¾Ã^iÎ&1Ç—¿Ei̇¹ªÍy¯müµ i<<€&à ñafßòÕÛóÓ³ÉDvÆÒó%keÐã·ø5wPáyûôðQ¸:(W( ”ʦ––½-$—þã"Qù²íë7`5¯•»Ý‰ë!Õᢒ²œÌÞÒ£—¾>ñ]Œ¼Ã޳G ÖäÓÝ[g® cn³9ì"cùà} ¡úŒ13‘È1oÎ~åƒìðÅÁ¯ÿßùÜ>iw 5ŽXýy%¬ÔÓuÆZcÜ¢¥õfooiÜϱ~ïû`!FA2ÕWÓ¤òUÅÛØD•×x{‚¶&[#e b¢xàÃÕm ½¡ÎΦæ£'Ž><,&Ÿ˾Þb“ŸÖdAíèlÄL:خطS–®}1•”г2ÅUDâýqcNsxêßë4Kt2m%ãå²z¥F †äø»ŽX§‹ œÂåv´úO ÄãÄú N¨ÜLiäõêÿ#'ÂÇû;Ó8'–¿½ks6iòÓ]²æ’hž½À¶U ÃéŒö¿¸CIMÚÃÜacƒVüÀ«¢v¸Ç¿qO€6žµÉ› åòÂÂfy[[ssfÍY£÷7Dø§Ø°«†wˆ°ÖØð´õjÜjŽÖÓ$÷ýOk xXƒ)YíÖxýVgÐ_üŽË¢Õ ;/ÂG•è?Dö¨ç؉`­§‚–™$’ûÝȧt-f…1¹Ê[ŒúBV isû@s;‹ê#p 4·ºwî—õC7455v{üüBTüI•¶ˆ"zo„ñ½ãê¦ZÏÚà9`kH¶7á£ãÜPl’jÉÜÚÚ­ê º.ÙdØ> [åûN¶9[­Ø«„²¢lW„E‚áE¨\äµ»ìàŸÖ®¶¼‡I5N#Ð@¨5µÆ®è(–áÌ#øS­3èA3ºw+ö~CÊmŠ¢å÷øh%#@»Ì¢¡’ƒY¹ÊYQ¤xŸÏos{H«Ål6a6›ª_‘¬ÍÍ¥° Äñ ¬îvïgF1‚ç¸ÍQôO ú™ 0ÈÎð7õ:ï¼àçq/ÿÌc‡D×6ž[²NVY,!7fn|¦|÷ÂS“Lõv½+nu,fë%Ô–^ÛŸ}®Âµîþ£G:/À'p%ûxNlpßésp…àVpKD5µQŽ=ˆÒTg·xzŒ^<5Ýà7hZ'‚×Å©[ò‹—ÅærÇþ<ˆ'0qZ˜ÇMÙÄMɘ¯­Â•ÕåZ»«u¿ ‹ÿ]ì(.#Š^?U¸KÆÙÏ¢÷î Ø²áù"³ÍâÀíâÑ™ u&£ªžä–ü8{™[¼†’qü½'QŠÍ‡swŸßô¶=pŸÅu%5j;÷ö‚0ç/{þU¯ÍcõfZ­\‚ OÜb›SÒ³ ÙIûá&Üýë¡#Ñ~o»;Æ|‰Kpª(PQRR\Zã뵤W+†í°jËbIΠϯÂÇ®±ã úiàj½ `0¼ÐRQ\¹{[‰ÅÜäμ~MBÉh š„£ùý'ã~mËì¹Òîû¼û Õ÷ñÙÎý ×ÊüÊÅÒ·©¬´MÅYñ¤x«è*ÃïB]Cv;K‹†TˆšÉQÜ\®›S=ׯ n£§ÑÔ‹$w–›. ^ˆ|Ö{¹åØþS@s¬Z¿Mû¶|¥É©ÓËŠS%oŠ·í"·¬íë>ÇàÞo1Ggãuò:ÐŒþ;×è=T&j A0P7Ù …Íÿrjð6Â`#¾YƒoJ„ÜãÜàžãÆ][zïÃ{½V“WcÄãNCå+³+ð {ÝÖ¾›Ý®ñv)æz½~TÀ¹ ï#“®L&'_I˜s$$ðxÿ ‚k6cendstream endobj 165 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 1112 >> stream xœM’mL[UÇïi¡ÜëœI dÐ^ã’mq²2‚!êFDa™B”ÅÕri»¾±ÛÒRÞZZ Ð§´PhÚ2Üìd¡n”˜™é·é]ôƒŽD?.ç– Æ W’œsžóòä9¿ÿ?Y"!”SYS{±T¹»,e ¶HÄ‹ nUf<ÙÅÄ¡³o_0®e¤/†ü¬[EYß<‹W㛇ðÕg!K[_ðêÛK•–V'£×êlÔqÍ ª´¼üµ“Ôi¥²œ:c¢½Fm¦jÔ6mRÛø‘ªµhô´ÍI¯ÐÙl­¯Ÿ:åp8JÔ&k‰…Ѿqâ$åÐÛtÔ´•fìt3Ue1Û¨ójM ´%B¨´˜ZÛl4CÕXšiƬf‹cAËÐjþœ¾Ò¦6ò“ÞㆶêMz£ši3ëí4cU ‚ÈE-—µÄó²ÃeÁ*¢‰¸‰üP £ ‘Ç»DdG‰;Ä#Ä  ´#Ò‰6Ä”ØÍJñ-|ñ úý‰×á ,i5›­V³yÖšLÎÎ&åRv &Ø–YVGØø£§Ø¤oü€Qx$ Á(õ‚û˜²þŒ_εç@34—ufz¨Í?à÷ù zG!âà[ÕF‚n-É5ä€ «ô¢é{ÈèD#|b@Qj˜ë¹;8Ã"qK¬x§ sö„9bûõzj¾#ydi«–çÉK‹ñË™&Y b8ò_Éq‚3‡Xl]—`õ ÝÈæÊ{êÐHwÆÿ~¼ös@ŽyôE˜KßžO. Ïi°u.Ô˜r¯ F&} æ4_VEô@zÜàÞ¨¸ï[3LÖ=‚B]5Ö1gHØ“½_xËmZ¼OJ31Ü‘F¿±Xœ9†;dÓ1ˆÇСØ^t8ÁáˆÁ´BºÕµ›´¾•+Î<ä“êÔ­Ÿz{šª¼Õà']06Ž…ä×ÿÙx„³W&BÁaC¸?8ødš;ø+'üÁ ½.p W/½«ˆ›’ºe¹Ó)q ¯Ä< WœÆ% v>.½Š‘Mކa"¾ÑÎðö*ž,èûx=¤Ëëq»Æ¢ v†ƒí8ß ý…îqOt`RÎÎs*AºÛ)I—}_P&–â_Úíh=ó’8s–­M¶/}S‡k öñvP£}÷æ/I§ãÿì&þð)»6Ÿ²‹÷´pK×…Ú-øÙLÄ;àík÷Ë}/ž«{ª¡q¹a¹~壕w€äÐN¤4LB Œ!,zxç{¸«—Ö.ÝUÝSý$FXô§BÊÞkK°Ê4¶Æ°_‹ÙÕÌQÙÈî7!#Þˆ«§ßåöÉ˸Ölî[In˜ïÂ^¾ ÁÐh·ï¼0Ò=♄Â)EC’;Í*eøs WıÙRëlæÍ w$š“ÎMçÉs³žkqåH_ËÏçÇA‚ø:vFZendstream endobj 166 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 2110 >> stream xœ¥U lS×¾N‚¹@HK™i²R_w JµŽ×Ä Ù+$@ JÂÃ"&ïıÄoÇßëÿúývçé8 v“0’ T„2e´ ÔšÖuZ»NÝÆŽÃM¥Ý$›4¡Uª6]éêÝ«óçûþïÿxXFÆãñøyùù?ß4»ÚzŽ—Z™–z>]ÇV?¾3}pÁóØS;^{~_Íä_² 323W.úòäX†ŸBåOc<^ÜdÍ«—ªd•å Ñêã¹¢ ›7o\+zåå—7‹¶×Jd•ÇÅu¢|±¢BR+Vp›Ñúã•…J´úÇ …tËúõÍÍÍëĵòuõ²òm¹kEÍ•Š Q¡D.‘5IÊD;ëë¢q­D4tÝÜ;¯¾VÚ¨ÈDùõeYw¸BŒaXfÝOwç«Ë ªªkW<ƒa{±UØZlvû¶ÛíÁ °φ-å.Že`—xÛÒŽ¥oK”Ñ¿ dÁþøÂ+¸ ¿›¢²¦³ñ祖ຎJÑ&#âuDÂéI$&ÏMÜ|¸«©ô ˜å+K} Ç ÏÕÿù4…üf‡Å@[Œ”P¾Y\&\cö÷ôx»ý}Ddt-„Kø‡EÃy[ŠJ$uBËÔÑ“oB9Èšy©±hÜä±zCŒ3æ';~9xÌ-kTµ*ÍUmÖî³êN°ß϶jIXqEÔꈃ¡ÃíMººð¬i©¯¼iêa:ÚzO`’Ðæí€+  Ô… @ŸóQ9:Ž 9úÁª¿³¹ûW¾JÏ} B{b4V›ÒS_’qœ-å£E÷v±YU:vãÙ âèWw³‹gsù&;à¯ÝkwãY¿5_:•L½$°Ù[ýÀ€½=è÷Ü»Œ°»‰® 'GºÎ8½L˜›‡t[Àõ”ðã|5ÈÀÄX ÌÎm-:PÚCÄFxmm4@Ó¬¦mfK6EY­­ÔÌZv2ûºå¶ûî@Ng¢Œ-¤QÊi›’ØÃ_ îVí“Wb¿=“œg#g– «ÃŠöÚ™(ü6[‚ŽîŠ@GÄj]M4Ù);øÁíés´¢5ÙèU„£Lœæ@£Ó”ã6ÍI =ak1Î!Ö·ùGíŒOx™ÿ)PTpg¢aD£÷h’C®y†M}!ðŸ²ûnÞÃáëAE°¾(Ð7Ï,y?û•ÔUkŸÍ^ 9ìsóú„ í?.³|Æm(nhÜE益t™ý„z·áD(œÏ|q¶z°Ån!Ê™Z7¼.G´2õ‹l†ÿù rh˜–0ä„ÁpûRë§ùÙŒ}aæú…íL˜xÿGè¥<€»ÃshÕ”jž!Ú>pÙÝŒ ÏJ™Œ]çãhðCZõ±`x$øäNÀwërìúµSSÀ}œ3Œ6ZkÖn4«@ osgC_sG5à[‹7HmÐÒæ·;N"40î„è%{-c->nÕÔX_¿m¸B\)mؾ£ºŒœS(¯ßî𸄃CC£ÜO~STy²¦Ã0øß®£§‡|§¹…&VBY|PZÌUltÕzÅ#hÁÈÙÖLs°Çç $gxãwÓSoM¯0vGÜeõ¬5­°6ÒF¨rta¾ ¤‰4Wµ6 KØ"S{Ýx'ÒöY–`W¿xsïoÞ›ºÐÙAtì|ŸÂYèõ¶w‡/ƒw[=MÊH¡î ²¨ p…):ŽñÚŠâEM ‡cïœëÿG÷1purì,à7¢; VY.Ù@èK6ã,þ»Y<‘úµ 8ÙÝ{Ûû½î»Õk$, ¤¶ÃVÐ;L³M ¡ZÌúÍG\¥A :h…V«†M›y=›%SQs rüŠO€.ï =sVùŸºëÙT”ë.44ó™GÃÐÙöbœÎvŽäíì±xê÷q$ø”§ýRÜNOm˜ÎDµQymC}ƒ2¢ìŒÅcBVòÕÚ‚’× @Z'Í EœýàÃ{U¡¦:©¬J2"{÷*JE™ƒÂ‡è±#h÷B$gêè©ý/ä³"ýìì 0ŽˆG8øÁ… çï êªt÷Y©óÆ8™S½ûÑ¥ÏÒÑ™o$6t«JÈ~gæWV-×4¶uXMzO†BÿXìQ`lм©„ÂYri#¥§9šñ}Ç.ņ¢!aïHÛpãó±@ª¬B}iîRÀÈîþtõEm€ßíxÓ3ÒLCSÏŒ¼ !R^ã(ÞØ®íîì>”ï/:ZX(Ü_(kÑX¸©ò“ùÄšÞòDb9üî^»ßîJܘH$c oð‘ÎÆÃrRÑr„°4¨w‘RíÞìVµ¥–üd”ôü!ûëso¡òalàf·ò¿æÞ&±äksogayþ¡ùþîŽ"8 šƒy–+‡â¥P²&u%—%ÿ.g«»Ðžxê{<ÔŠ \Qo\x—Ư®4Wé¬B–˜yǪá£sd$§ Þ´¤ó~»>©¹àâM‡ó›š¸=0î b_ºc.ÎB9®nÿ)Nò¾f·¡‚”i6mªa—Ó¤T¢Pæ¨õuÊ€—V÷]‹yݽ\oš¹Þ¼‘D7⼉;hœs‡ÙÄæ²ëÙ!V¾&Á.ú½„6 ^¤²±ËçБ(íàW@¤< Úb(!L•‚Röðf6=ð7Èûg¯G#²šºRyþ@I»øì1ÿ“‹ï..ÎØØ™¹(îÌÌİb óŸendstream endobj 167 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 973 >> stream xœMmlSUÆïé˽¥tE Þ¶—@âËØs eBhÄÁ@‰°è’Ú]¶†v-í£ÛÐͽÐÞÓÛÊ^>¸¨3qJ€ìS¿Ì±@&lj4ÑùƒÂÑhœ‹ÄAò?ë¹]¼{KøprÎïž<ç9b :!ÄUVá•€Ç+J!ÿ¹I îÙ±£¹¹¹ÐíBõN¡Ù+5ÇŰ:'Ö ‡’pÄí…帅Ë[eÀl’ÄP¨CAwHlô‰§%—·¾~ –-Q(P×ä‘ê¼á ÏY!Iâ×&†´Ïêçcü£ÖœÀJwïÓ`z6Vƒƒn‚©× ‰ “Šqb†|¸; ~óÀ( EÁIÓ@nk‹áXÌ.ËUG÷RuQy'«¥7´´“îÆ´ «æ15vÑ™ï©s†ÚþÝÚß—ßǦÁ^<àx< ©Õ“›|úåL̯“lº f€Ó ¶ü‡ïôÉñlºÐ‰;.&{”¨ãjú•nœ¢ØD§nÒ4V]X­¨QÍ4ŸÞ=F7gOe~ðps¸ÆïÈ$V4š]j¡íó´-…H¹z5‡‚z¸³CŽá‹yíýøûü4û(ÆÚ÷â‘g[{…éSؤîeáÙßÛ¯$p2'åxw?½CŽçCK×4=jfKá-cfš ʲÏaåÃZ÷yÃ×o˜[¬²àGͼnë–(ç?×Sp–¾ª¹?`É%òwoŸæþnVp¢k@==¿>—Œ²°ÿy¡?6Ñy¿å—SíßÖŽìÑÞ¡Y0bÔZb%,ËÑn9ŠÏãvòI¦!WTv¼{°¼ü!kóËÓÈ7f[k7"ËšT¯Å¢­,†ù!ÙâZendstream endobj 168 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 652 >> stream xœeoHqÇçíî¦Þ¦3. óv዆ÿ¦bjì•%…Õš—·ávëvº$k†Ó_J(ÎóOlR§X%e˜Ðß(A¡½¨¢7ÊïŽ3èæÛøÂÃó<xø<0¥ ÃWcsEª)Qó0õhššC¨ iN"dÕ×å7µ/ïY!CÚ4«eç € y²Ð¥l€c˜ÿFK vIÞ6Aæ =E\Ye¥ÓÁ•—–Vr§ü¼äõ¸\£[x¿[6†vîœèñòrWX-Èr°ª¤$»ý¡bQj«)rpa¯,pÍ|ˆ—:ùV®A ÈÜ·ŸçR’Å©âýÁ™—¸F±•—¢¬üDÕIrÀ!À€<ÀÒx ˜ÀY°‚åbMصªI¨¨O•—q E>ãh]³1‰¾Aöšç|ÈÍ>*Wàû„y„º8XÁG. îÀWF¾Á…èkº@(GôѼ=uê ú³û[ÁÞî"ùŽæÐUf<öBû°ºö~7£ÌãŸt›NW—–_|~olâIìÙ›ag6×/CóÏ¥jgA}­ž¡[ìzŽ~8Òû tDµ“3è4ÑNFý-‘›÷{ká-hÖÈy”«¨— û¸ÊjyLwÞéï¿}—Õ7ö=„ÍUæ=5¬ì#·ઑ¸}`.S5ÑÑyû×=åU§Ì>¡QV´Žx„c?Ž£+j&3œ½AÑ›èHÎN&’¬¾iúogÕÖ‘eU-bkÚ\ыé©ì¶ÿM'»CP–ã0f×,ºÀĦ NÙ6È´A¬ª+†Ä‘á RwRJÆV&›arÆéôäCšÞ§-üëOÃendstream endobj 169 0 obj << /Filter /FlateDecode /Length 169 >> stream xœ]A E÷œ‚@§+“†MÝt¡1êè04, „Ò…··PkŒ‹OæÃÿäè‡óà]æâ–>(së¼I´„5!ñ‘&çYÜ8ÌWOœud¢¿èø|Eâ[€ìî¯z&qyªWÍ^Â`h‰)i?ë¤TµŠ‘7O°Fû“,h@±®ÝÆ"-Ôú,?¦ãšù\Á+Xáqž¾»ÅK‹obotFV&endstream endobj 170 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 357 >> stream xœŽMKA†g\3)³/<„©sTµ.eD_B7+òÔqÑÁÝpwmwl‰Âšƒ•‚˜¤¢£¡[÷ ?Щÿ0»è!óð><Ïí…ÀjB[,ž8Yÿ·°á†Æ²Åðp”š9ãsÂf÷v> stream xœ­’KO…ghÕQkQ"‰2•„Á_D¬Õ¢A£PÊHIi§”2 LôB€RJi)-T)DÔ|€n4qcŒ¢!Á… wÓRÖºá¸9¹7wsî9ŠHÓE7—ªTe…§üØ^4¶/-–%n},®Ü”…¤Ÿ8–¥n\ø!™dÒéõôÝ¢~—X—.VïD$(j´t•R&ÝP¯cˆlmq ¸¸(—8XPPL7tƒVc$TFG4Lri$.PÚ’±Ù%:†1ÎÏ·Z­yƒ%¢ëääÖFGœ'-$ÝLÖ§(#CœÕH"å2/¥¥”ÁÔÄ4¡¢êHÚˆ ˆÄ¨7 È9ä r6õ"E^ ¾4Ì%sBl€Æn­HÄÇâÏ Çû“ã$`flzº¶£œX§×é öz‚xøÓ¤ 6 µËiyJÁ›¸ZGOÛÍ<ÍSU`ÇšÆ:C¡±‘ÀÞ7è„'¾dz¦úCƒ7½3Ñgssáñùé瀽†Êu a»¦°·ò,´b•ë—~-Š™}î[»ÛÑbÇÙŠÚòJÀZº‘°gÌQî?7Á2¶ªš9Z¥dK”¸<žÍ ¢ðuT«Ÿ bbõÕ‰¸ÏÍèíï„ZÝî„´+³Õ}à ¬ÍÁ;—ÓCô8Ùñ{9ÓÉ‚z0:ÜõûqÿxØ#¼q¼eA¼9\®6°C7V¶L¾ø÷¥\öØ[8WOMº$aÞ„Þ°â¯À|84yÄ3ˆD>_?Œ¸¼œŽo£ ÓOñ£³ÁÅy½pùŒúŠö*~éZ£ùz;&__ù×HüÛŠ$føœáhîÒÿcµ•×T\¬½}p:ì %cõÝ^wÀËT¬ˬ…§qû;õÝPJVY²³¦³\ø8y/ ÄLŸŽý@X8m·^!oŽÄJ‡}>ÑÙœ¨Þݶ¶ß&- ˶Fûe2ù^þ¤¥endstream endobj 172 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 401 >> stream xœcd`ab`ddduö 21ôH3þaú!ËÜÝýóõO]VY~'Ù€œ=¯ùºy˜»yXÖ~ÿ#ô=Uð{ÿ÷fFÆÜÂçü‚Ê¢ÌôŒdMCKKs#KÇÜÔ¢ÌäÄ<ßÄ’ŒÔÜÄ 'G!8?93µ¤RAÃ&£¤¤ÀJ_¿¼¼\/1·X/¿(ÝNSG¡<³$C!(µ8µ¨,5EÁ-?¯DÁ/17UäH=ᜟ[PZ’Z¤à›Ÿ’Z”ÇÀÀÀdhËÀ Ä r‹ÿÄ|?¾woú1ÓöùŒß—>dþþû'¿èÊöî\9OoÕîônŽö¥Ýg»/çøÓËžÕ]§"—Âv¨ÿf÷a ¼Ù}¨}/ò¦v•îkäù~YgÜÿÝm?ã­Ÿs˜ûþ\tæ¢î¥K ºkäÿz°Õt.êž)ÿÓ÷w•è¬e@™’îZù¿Alµ%@™eݳäùÊüpžõ=êäl¿§±oâºÇ-ÇÅb>Ÿ‡sõ$ž{sxxÿᘘendstream endobj 173 0 obj << /Filter /FlateDecode /Length 161 >> stream xœ]O1ƒ0 Üó ÿ ÀÒ ˆ…. ­ª¶Žƒ2àD! ý}I€ÎÒùîä³ì‡ëÀ.|D/J`›H‹_#Œ49uÆa:X™8ë dÓáý ›ìÎïz&ù¬•*«z¡7´5O$ÚªêZk;Alþ¤#0ÚéTWÐ\.Xü§’£¹Äyp‘8•¦¥I.à˜~Ïr 6ˆ/IìS‚endstream endobj 174 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 6354 >> stream xœX\×öžuavTll&€š¢±D±G‰ , XbÅ ¢"dYX`YzÛ½KG:ìÒ«Š±+–'–µ$Yë‹yÅhcÉÓœ1—üßÿ. KÞó_ü¡xwîÞrÎw¾ï;#¢,úP"‘H²`™«ëäI¦ÿN†‰„á}„wÄ©8Rпr±|‡ä4ÿ­ç"+1²²hnÑZp{Á¶Á”X$ KZ¼#Ræëí#·»õ}ûɳf}4Á~ʤI³ì½d¾[=ƒì—yÊ}¼=åd`¿*x«¯—<Ò~ìl¹|‡Ã‡*Љž¡ƒeÞsߟ`¯ð•ûدô õ’…{m³_$·_îèeßuЉ]¿î“{Éì—oó’Qåì´Ö)xÁŽ…‹d‹CåaÎá ψ-K#·.‹Ú¶Ükû oŸ•¾«ü\ýWnúÀ}⇓”“£§¨¦ÆL›>㣙#f9¼7{ÎܱïÏ·aüŠA}FÍ¢FR+¨÷¨Ï©QÔhj5†r¥VSk¨q”µ–r¢ÖQ ¨¨õÔBj"µZD}H-¦>¡&SŸRÎÔTÊ…šF-¡¦SK©Ô2j95—HÍ£Qƒ©!ÔFÊšÚDI)wê-ʃb©·)Ê–²£ÄÔPjeI½CqOõ¥úQïRý©Ù”5€r!É¢,¨Hê¥H)zÑç£>Õâ¡â ñc‹•§-7Z¡ÇÑ¡ôU‰äw&†yÜwN_C¿wúìÏ÷ϳ²´r·j`? c lÝ ðÁïÎülHø›Ö¶Ö§­ÿ”Žn—¦JO¾eõÖη°V¬‚ýã팷_ÚxØ|o«°ýÒÎÉ®x蘡žÃøaAÃö ëî<¼à»w޽œ—$T *áŸ0ÞÕ ~b0 ·Ù„½*m BJ”à…tèmUHŽ’ÕqÉÉñ(† ¬A¥Üsº¬ 54*7_ùOK˜AWŽ´ôV €À*TÆ?§KkPCaÊÎØ,>9# E#/ÂX-}-±-ݹ)\0¸¬ÁÕh#ý„·Ø&T¼•s¢•2²P *âã æL¥‹Èº2¤ä=`-=üôÆÞ§Š‚?ã°b›¤s^­iÞ½çEÓÏÖœÍI¿ù¹ûFD0{ *ƒh¯à.†Cð- – ·¬ ‹Mß AÑüH:Z†*Q%¿ž £-u•¨±AÖõ(„<ªAÅ|% V8 ÛáRË0ZJö®&Ÿþ³k™P$çš'x¥<ì¯L;È»#nwƒõFˆ6n A`šp”…!ô÷Ëü¶ø;ãߣau0{©âÐIt¹¾üs8œV™ÎPJøÚ-}±£Wˆ44Xžo¿eÜpÙ¡œwÆ,–B8pô÷ Nîë·­˜Æ““ ðAtÆÕF±à dljO°¶˜8ÁÒŸ'@èóó/`ÍaO/"„[l ªŒL‰OJáÔ©H#ó«Û¤[KÒÎ’u‡b¼Èox˜Ú4‡¢¸9 RóIï;„nA̤é/a¼ïeûùӛиŒ¨tebv#½Î´–àC|Ý —l©Ûà¥ì¹³§.Þ8ë4mü"—Es|îãðh‹ýžš¢Ç?ƒMùõÝ î‘¾¼ôq½ºÈë/àê¨ûŸÀE÷€äŽÊŒâó°Ý3‹ùŠUÝ3‚-4ØÜ{˜Ó¦Ã8lÇcIÇQVPý"Ë6÷ýÂb¾+åað輆ñ°q#¹Ê l y=§›sòšøvI9ªðõAá n1ÖI¤6’-H¹{ª(çu?Vná° x€z Âv™BÆ]—4&ïôå:úšÁÚлäà#,¾g#= 2XÎþtîÀiCKÕˆ©Ð•ÔêbÌ\»$r©+/ýfW×R¸OWITÌÝÆÌ#(2£‹»wñ3ýˆža8¹£æA Ãé7ž}8øAÆÎ>Ä‘æ•_¯xO¤‹«;+LɻѯÑl}ͱ†ƒôœ€\Vž©Èee~R€¾`jšþ®(0$ƒÓF¥G5vË(1a64U•¤æ<ô(‚ õcù,ï_D=TñÒG­‰ÿˆY3tûv¿IWåß’q)e©ù!ˆ1e„'Q"YC^‰¶¥€kõ>¥®#€›ÒôÝ1^ú ý³zÿ,¦ëd_~Ó‰LñžBâ hfšgÞ¡{’N£Çí}ýÆíƒíeHïã§ â4ËVÆË3¬î*bWS:_ógüÑÄjýõ¨ 1×é’jÂW!HŻú3 !U_B“býg£æ&*%ÿ9#o ½ÎI#ö£ý;[ê̤ ä"ÇM,EõJü^3Ò€êPaª×xo¯y”)a©'á¦AT%ì N$" ½Ò× ¬:‹ðM,‚›–u™ÑCµ 4Hð£_Ò#»rT‰ò›MçLÕƒ£ÄQ‰'6™vГh„¢HR™tdh'³—ó­ô:|ÂRC§ÖĉÝy1-ßÅÄ…¥%mÅ\ª‚ÑÐnp²µKèä(¢çëzÔÀŸÇí™òŒè}ÈnÊÈ+3Tß×eT0Z:Ë­÷ÍL,ˆ*C¨e”dV0¯Ó5ÙðÔ`}Ú¸¹–‘”… }šØŠ€cš½ˆ©«èÏø¬ŒD$WWßÅïm…œõʪ⺼-‡rµZ]z‘¶å!æï'òÒ£xQ´nüÐÅ]žsÒðt!çh#GÊú# Ñìðm‘ÞQî(1˶ìú¢õ‡•É¿f¢÷¾bØ2hk¹.ª}§îÀË{9¡U/·¬y±@ZPY¼¤9¸x͚͋Ǔ@™J<ò<8ž9}6·®?aÛ/Î~~ÍFú{¨`!»Úóè©3‡üÛkVzx¸ñ“}Ø õ^y+b^¿ÖÞPUÇë ª2 µÙ;*“v"¦´² ¾2¢quÜzï&^™ïS¾ 1,œ?cs©_q/ÅTœ2!,p(ò)’åÅ:º£OP3ýÅrx Þzqáû–Èãkk¹˜ϊɈYD'¢”˜’QlQ\‘F‡ÒQfzQæNúã‹ì˜ù•å55éøê’‰ƒåoà`ÕÿÒŽý?š1º³uX×í­ïÁÙèiÔ‘²RÂ$ÂÇ×臇VlXïþù$~ ®þì—-å´ŸùaÞÜËzC—@Ïú÷ªð£¥ÇogÕ€š‚+ÂQòD.(°Ü¿4°N¹ 1«¾ÜÛDZâ¦5|¢z{T0:rê"ÊM¨äbëåzU~to?µ¯§‹ÉùÁa(Ûe†Gô£%¾€ôÂÓ13žFÐÒg‚Ú¢.¬"88,,8¸"¬®®¢¢Ž{Ý>YŸ¼ %—½.›o…Z¶.ù;šÆ¸m^1;x}fíVγ6õ€¦LSª)óïöضÓxVd5åqû ÊKÄx2×î·_ò>[ÈoÙ·#gN#}æPæY൵;Ñzì*ô+þ ›ËËOÖu{_°í2¿)Q‰Ü–Ý« ܉ùÏŸ=ãÓ#Îß…ó›©y,/Jcz4ŽâX;Øt†â…`ÑŠ)tO•ÂÛÐ( y…øìð æ#ŦG§:NG9L¼4\4ïW´Fí«Kœô‘ðÛ£ÓÅ5F°3‚ík»iû†&aX¿;™ùýØu¨Ü$æöÄñߺ….FêO©O›Éüä²0$–»{™Ô‰½Œgà –ð¢kC3å¶v¨Ö÷ê‚z5%ëoÓÿ¸Aªïî÷!¡ég^> «ÂW7g’-r½†×ZBög{‘þ‰¶wvs[ãròw~jeÞ4ë ÝûðéÓ{a÷ï^ý«¿æª³tß&&í5¾Ó‹“ßîÝ+žjƒþ¦SË-F±êŠDŒ×îz´‡Ç$ÞáÝÅ~¿ëžáÈ›‡O$ÒÖöýG÷WE¹pX%‰xÍ7ÿ2éÿœ’BlÐS·Äl æ¢w—ù!FzÛªBÃLix5’ÜUEîzúÕ»Ýwuî•› ŽÄz»ìù*T]µ “ÄÂ,¨b v£Ý»}P,¯Ò±>ÈÇg7*0Müõ­WD°GxÆÆæ$£4BµjMB èl˜˜Š¢˜¤þU™ÙÚœîå…—Ë‹ÒQ6²«ÚÙ”Sš›ÓÉ´0Daˆn]`[kü½dÛBšöV76š6SÿJ®&ˆ÷WƒØœ-B¹LYt©",):&™Ãßý¹8I©A(Ñ.²8JW‘S\˜ÝC$¢ãFø‚)β°|Øà>ØbLçk—åϱ ô‹§0Hˆ}p.;ÇúÁXm|rãú•÷ñ<Ö÷›ÃwFŽ›~^§”EÕh_FnÙÞû黓­Ë®"‰\„E1S£xM2Z‡‚™É䢽•ÞÑë4¤=MV$‡Vivò5¥¥¯f°ƒ"÷“Aä§“Ž!†tSõ2m‡g =¨‘éT¶ßµá@uq³¶º»ÕhxNçrŽKÝo—"’ccS8¯nqj¥&\Mö–•¥—ïÏÞ{ˆëp˜$1«A¦Í“èžè‚â ¹Õ«alÚíCËý9O‹|v“ÄtW ©Ê“7¡Þ(þê,›˜œ‚TŒéE»@SÓí#‚9œ@Ÿ;ñË̓û.¶¹„þÁ€Õè›x î7{ÖŒíuq¥úÊÒº‚”„®øÈÑÆÓˆ¹óͦ©‹Öîìú¦PŸcQ:ìV¿¡fi¦b¢H!„”§’ÂäÔ$¼~ &AÚZìP¾Ê_½¥2x±„ð«7¥îŒ89c齟u·‹ym6:Žê˜Ûâ=%{Ši«ºRתçCµ¸Oººé„¤Ÿ7½~<ÔþY;ì0n%ÔþXÃ=¶mOÝô7æ†ÓWX‚­mXìW­Ò™îS”¼3-ƒkl¼jê@¿>îíí³#Œ÷‘©½Õ+S£ b¤¿™ªŒ;B?9ñéçËW-´unþÑ>++‡xl¦*B'“‡EûOyü9á]›ï>â¤ÑÕçœLgbÁÊå²õ=øÍFú½à,ü-‰.QÆ%§$'qÁ!JiÑób b z_´+åQŠè 0É(‹•%ùÙYÙ9\C]MIÚIô¼0¶0F¾ Õ¢º’ÊR]q}*ëÖvÊ ôՉꯀ¬M¼ÒØ„Òo)¿¼¤F²0µ¬—¹øÔ §e@Ý2æÀ º9)×— ¦CQŠ&FÁà;dðÜ?—t¹í.p˜NÔ¾}4$È¿Cƒ`Çú ¤-a ¬„$³7Ç´_R‚?Ë%‘HÞT–™µ3—«mÚ_v–pÞw;¹|¼ð³‚—äž&ÍÍ!­——„G&l›ywXÀ'@ÌI6÷Wb˜º²Ürf"ÿTín/™Ãl¼ûÓûh*‰ËJ!’ICË5ÕºÒ†Vï¿#ì1ÀîœôŽ §' úõ.X=µ÷ᤂr‹r߀=ÞϺ m­ ƒ:[“­ÎfÒd"òG›‡Ðq«‘šu)‘ÿ çÿýö©O§;¯\Ðy,½Ðb2ž!´5€èvÄK}ñ„û0*ä·»´4¿žË‚øœ çteé騈1Eˆ!\iz£á‹Â"xÏné/ï¶Un€MÒFW#TM/ÆC1t°NË·l;öí­3ÜV.ópâ¿ `ÿQ{¸aî̸…ûc«f;¬<²îÛNúÛüð%+æ÷h* ë‡÷ßt9ç¨çfã&v¦s›ÑpöâÏ,]¾ØyÕÌN¹ GEÿ¢bÎS?ßùYü/êÖ¿(á›qÔ77E  ¸ Bìa«Î¯#tšU¦-5%}óÿ>£ë2³.@ ɜѭbn™ªó=¨aãñà9cÇ£ä­ /ÉÌÊÊ&ínU¤.D¦ˆ ô/[HÚ[RZC`=I$c±qžIrIߟ$uO¢“¥‡÷3ø½LÖaÍW7î]m»v°N®(ákü Övª›¥ ”L%J`\W¯œÉu 8›EL°øOÃç,Þ¬lç 7î‚ëÝ $‘Ög/-iÿÞð䯒K6Ò?õpêÙ/cº£µh­ûÆ%qÌrP²8Zâž½¥"úDvæÎ\TÉÔ…–*¼S7Í<;˜8n»É76 ên¯–Ÿø]A­ÁOð¨ZFú_§`›¤9õ¸ªdm|bJ,Šaˆï¬ÕW´|ãz ÷Ûɵtß$Sr^Ü'fÉ<äˆB³1³˜++w®ÙƒN cÇö]ØiJB›^ðÐuy„‚’ÍÏ)ÌEåLeD2!M“ âð¼5!9&EØ¡˜Ü¸Iðö²-JÈІ,)/¨,'Ÿæðx8[š°âvåÎÑ^ÁU,,ƒ§lA3±Û‘Š'NµXŽfb9àžÀµ'~&3òëGž´Wô×ó¬$XÈË6&æÇçàÀËV•—¦M@Œ2"VA>MâavÆäo\Rç!åå±%yÚ¼"Sjfé_…Yt$5輚EÅI™‘xðÝM`‡À=Ý < ý±$?'3¹)º ¼Ûâ<ÜÿÈäÝë¸SN7Wi©ÉÒåòæÅö9{DΠÿ¦>Ç×\ŸßŠY²"1:?)+‡™XTŒ§“†lÌvÌãa“¢IžP4“˜­¨®‡Í»y]¯³N ü#:Ôöc,j ï ‘lZz§ýûØe:¶Äý¯8Ühýå8^F Ä ñp<ÛàOñ'0Ûýƒ»gêë‚Q|TŠç}ü½â|I“C FÀûÏžÿVw8~y%¿_e7zž¿u÷Èé‡}îàà¾r³é>ôáÂüè" ºxUgýµ`‹_ÚHÿ C6kðÝ®»âY«LG¹ˆÑ–—Ææ©µ¼¬ô“æ" ÎÙäæœd¢ú­¼ ½Fû%ª·ËCÍ~~È/!HëœÂaê¢*O’"&2:"?¹dK´S+b“b’‘Â.íã`ð5uˆÆÙ% ¿æfÔœW¯ùR]Ù¯Œ¤ï³WX³ˆ8w‰ÈŸK$Ùñ»8bxÍÏ&¼irD·õ0ç Ì}mŠˆý…´Ù€‘tKÂ#ñI#†x$ÝÛ5a^¥xˆÞšÃhSa °;ƒ^¶Q›Ÿ™Bd5º8ª´éùß aÛ[åþ^õ²æìmz70._Xóòi™#1ô3öçúYD[õ5XY­PÔXI‰endstream endobj 175 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 5146 >> stream xœ•X XS×¶>1æxª¢= UÏ¡Újg­Ö©NU«âóç Y•xæIoЋzuõ4¦7(Öö{û|/Y¦_ N’Iâcã¤áoE Ÿ8cÆô1á“&L˜þ®0F) _)‹FJñ‰ákÅQñ1RYø[³â¤Ò¤™ãǧ§§‹¦ŒKb猞/ _“#I‹‰_,IÃWF cŸ‡:îùˆ…I©ÒIø qtŒDDÄÌwE›æ‹$-’,N‘¦.M‹Ìؽ\µ"zeÌžU±qkâ×&$ Çî7a¢bÒä¬)S§M:óY³ç¼5rô‚JD3ˆaÄ*â b51œXKŒ Öë‰ Ä(b#1ŸØL, Æ[ˆ…Ä8b1žXL¼GL$–K‰ÉÄ2b ±œXA¬$zs‰>DÑØNô'^%v41€H„|b! † Á݉D/ââ\¢+!ã xZÞÓ.3»ìçOâ·vØõ¿[WÈ-¤›ü½[j7HÉ©Öîë»è±¾Ç=WöÜ×kH/å+s^¹Ö[ÛûQŸ²¾ãú¦…L Ùâ 9ׯW¿È~¥ýûô/ìϽºçÕoèÝô…o Ø2ðâ‡þº0Ôú8l4çéÍy€ïW÷J%›È}M¼zG  rH9HY{ÑžgÞÐR‘ öŠë@ û”tÖ‚}õ)@ÌýU§’ùÃi¤" $ˆj?.õ‚ýu) ‘E¯qÝhçA×±‚ZÊ”@‚, 3©J:…‘KáU×ÊçºúèþD^|KºjðÉ@ÉF“ûáŽü#&;¨»µmß{w¦‰÷0ÀlÎ6Êÿ3ßw{hÍä&tQ°¥ÓI“:ŸT'Û9j«Ot¨ã¢Þ÷m°ñ*Ÿ³ÀBvs}, A½P_Ä"ê«%?À>°7ì ‡1H…¶ÓKvúnüv÷ÎíÏ>X·pìœaý‡¨}\÷ ®ß—Ü-Úd5å3UšíL—ksÕyŒV£–ëòâ÷í*Û(4½ŽV¢ˆ¹…H|ôÐoaŸCãåd©†MBüÑòD@MÛtï œ…¯»4_JffU¦Ôýîb»×Êà‘ÄÇ·Óf†iþiÚ#i]|ö‚Ýáë°ë¼ÖlOÎHfk€3úä˜ü¬“ãE‰þH8 &Ò± £B]¦ðäÕ‚Z`vþœïôþ哿uñèÆˆ˜8iZ;Y`ªîôñà^ØLäQ«—-%qëJêl%¦ì‰¹˜€9¬P_‘hPîÊcв î@=©OÍ‹:*¥B[Ú`.©²2ŽV‹ XðÛ>?ë ±&Pú¸`é9–<L¶Â3æG°S~……D˨ PE‡¬.âЦûèiAúBgs¯“¶Fׇçû¯î>¦¨z] ¼nïùÙ@5zÉr¥’5;­UÏõ…º¯¿ÙqÁ^ÿÂ:"uéÀ4”¨ä7˜].œþÏæâ@úþð ò(-ƒ‘à^4ÿiú!Á’ÕÒéÄüñíÝůÀíä ÃÌòñ¹%Ü4:Ƴ(0ߦGÎÞ4ãûÌd²Çó.gƒ]”¼ƒ†oß\|MÍè܆R!CnÎC–j1È£äUy¹Å¾¿’ù úpcFÍ=ôS‡Ók…•l‚Gl›íØ\¸¦죪êN@ò®)QabŒÚœèã ˜öƒÖš‚SÀBå•K¥Ežžiîªï—Jœ‡¹zˆ#i8+X4™t~݉%BC¦BÏdŽLæÿ5Ôz9FW6ó’øÒ/×>oõCš–­‘èå¼…º,@-ûKíÀl)cŸ_Ƀ$֤Р¡1°&`}IŸLC´bŒ&S»-z –ÜeùX½¬­ÎZg«s\Åùœ|éOîF¦@Dê°‘?ƒº¦ãöúcŒEä.@Ý謧`S ÃM ‚]ü—@…‘ S‰›ƒ½á?Eç†Oñ)jŒ_YZ‹«/*uN¤*MG_¬ƒ_ö#h¡ýL/yÝ·ÇU€*V *«‘jÇjºï±÷KÏÓ^ä¡·#ÌNr+ÝÍÖ3ŒÇÖ õOë*…ÎnÔ±‹ô¨0 QÔû­}øa‹Ã{„)OÚ¯ýPµÕb*ÕÀ–Ê®ùŠÖÄë3cÓ“³vçÆjclõÁ˰÷ba_ˆ”‹ÂI ]Ï)¨vÅzA;aøªŸ U/Éæ ù½…F£©‚A ÕŽMHgŸÂ>­¨ .í9¤Ýý°øŸe±±äp8UОýç>(èè¤%Üy?Y̘ÂÿZò·Ÿéô™¸äº0¿hKÝà´™ù¿ûî‡èûÿwõóVa~´ø£ï Ç4Rù¼m5è±›TŠ;¾ÁõücÿÃ\Ð Êäöþ—çç Ê€ý¸D¯ü¿3iøví!QO”+\¡¯fM]Øzõ/Ó‹ç;´¨1Æ0 Û5=G¹{ùƒÜhË·9A!UšS)’Ç«·ï®½}ñÛsuÌ‹¾ô´ÚþÆç2 væ8%©:Mf#Ù•ºP3·¶´¶]€!q4}k2š &@ÉAÜQWcñ‘«ãÀV´t 3ô­k1Æa=t3AMÅp*º°Ü^b+¡Läìg"<¨©ÛÏ™ÆÞ攂² ž6,R<Í •n­VÆQFRn,ÐåçTÈA& "ž3ÐÃú¡ù±_rb‰´$-/Øê˜k·&Û÷ZÒêp¯¥ÏµÕήV,®òvr½¼%¡©™{ê#@ö½P74 uýbþýÓ'܇X£Ö¨9”ÿ~æÝmÖЫwœ¾ y?üíîŠ÷ÒtZaîVV9cÑ{KÄ©s«SY_o¯ª?Jr†ÍHŽàtyû–$¶ óCò"23hz_Ì ò¯œ§ÿø™Ö&±©ƒ°|ÒJm_Gâƒ[à«>Þ½V؈sß×Ò ¥à¯p𘦅(õ@½ñgÁw³ ¸tÏà…^+Ïf“wo͈ÔŠ9‡~bKZèê¶¶ó7õ#x;E Ƽ];¼êŸ&ý:M“î¤Õm*²VÚÔ5ß–]MÀ ¨ÒÇÁßòY;£TjuæxjXýG¿Ü X_bmt\°¾œé~pþãÝ$ iÿÑjðå›;,Z?øÛÕ²¶ýÜëðý屦«àumí $`ÐÒCwس3ßýÆ­ÿ™;äÍ¢m2VbÌeÕ+@È¢æ~zøZýÝjÜò Ò2FcÎŽ4¼;æYk¶zk k­¶Õ‚ª¿x4÷ì§ ‰^8ñ:ùMÓ¦Èõk–Œ|AíÛî Ø-}ô„Ïà.ÑÕq¢T,Iv§yëkjê4ö™™Fý:¡q&H×ä_àvìÛ$ÚÉêä‹}ö¿uWø\ów– J«g€”âÝÍë £±‡ì6fʨÕâÊ4¶.É«ù2ãÓÌÐS*O,ˆ¡VF,˜œŽz”šm Ÿ©Î°)r@‘Ëèr$hÐüd¥µðps,öŠúC<)éo!ñçq²JØ¤Ú ë°š [b¾ÆœèH©‡¨ —Ïß‚ÓQ&HkªÉ/òX™‚¾ hE ÑŸÀG!X„êD Ä:mゎlivX<§îúï¼ÍÃA„“:yºAÿ‹èhRé¸CLÝk u»ÓhíÜR¸íóaŽ–¤¼X0×I˜µ5bMx«.k:¨"§16ǨÇ}09/{¿øFÉuÇücÀ\î¸óÖîŽRç0ª# Å ñÙIR]mª7èX¥«@ëïiA!Üûóëqì°º÷7Ï%ÂjÚq8çw¤×HUˆ‹;þ—¸÷™6\oñ¹î'ZYˆw$ ó;g zmi¨J£Öe´Ä4Ø­f[!{^zá̱m Ì]ZëpåØÔùiÀÜûñὫtƒ´6.>^#ô$ljj8̼àõÃkð,^æë¹WiH¿yo¢Áøc ŠD¡W^ ‡ÀApÜÉ Rô½4úî/pù/0ü»ë‡¦ND£Æ¡Ã;Yoß“J7Æ/»É†ÅáÌ)M“ç©5:F!Ùþî˜fBJm—{5—{˜ÅϦ Æt*¿%hh§ùKý¶ûw)–µ®âs#6šÌøÜŠLW†P+Ì60ˆüïù5ÆÆ&sÊÝ ¶úS‡×Ñûõú=t´ñ¹žp' ©»çî€o¨Gs¯ã9:têºRR)÷Ôº«jŠ4©:zæð%@}öY¼¨û49¦[9Ú Pks2,`#à}² ”{µ¹È~P~é*SÁ6†ª½I@s‹ÿ1W¾‡Í¸¥·Â³ôƒýk¶/~wÎùdóY kµ8 Šò¦—¦n›²G?Ýó†ÿúôñ‡â"ö1Ζ}¼Ô…õWQw4zýÂMqyÅUÕõm>01›ÚL¥€j9š-JÛ#b÷$ ñ†U:¥ƒˆÐOìŃóá3>×ÈݤK²*dÙZ]^£PíN‚\ ÉÏ-ʬWă½`D•›*U`6)KdåEv»ÅÊ8‹O¨ÆR°k‹TÒrE=6â®ê²†J/N¿Ý!AÂÇuÇÅgÓ:Ÿ¼È Õ2 ab2ä¥ |F„fgi• ³=HH"K;õØ÷‚G¬%ë¸ÚD®FHÒk£NÅÛÖ°Ãjö/…úNÆMJ€Ê¨‰Û…JB ¦ŽR*«\Ka!sêôE[ Úׯ=vþ„ÈÍ®J¶(¿È…¥^®¬m³'ßÙô€!?ÃW.»5,i§.uk»:à¡‹pÈçЃo^ÑaÿÞÀm,Ui´9y@JIKeµoÉÉ–Íû¢7ÑRü·¼µäï°ìûw8ô‹µ±>EÃ|‹NÒb *L«WxÔÍÀ‹½–ÙêÀÿ6ž8b³o¾ ¨o«î6qÁ¬ÀïÉÜÜ¡€g?Eëô#&ÍB¯M@“äˆZ*µLåj°V»,L3Üü+\ ÇÂE…e¥V‹Ùb-EÝõ+2Ï%Ã…}¦^fPåEaWì_ÆœpÊìÏÞvC;VÙ(ø øð2ç£Ms!°QeYÕÂ]ºHµ‘IG[ J$5ÖX\.+¥pן¿ÕÞ‚àR܆¸\mA›Ð…—ö¥·Ý¾.§Å‡Ã ½ ³.ó¹ÜYú¬¦!Õ£ÌÉÍ JV"­?mþ°¼mÛùáEŒ­ÞRŽ7Pÿ8NÒË“uŒA­Ï0ð‡1Ë€º¥/xÌG–B"#)C§¼²ZMf‹ÙYbub¸¼©®´-`µbÒ§[îä0Z±^†y$s¨Ë½ÖÊBc­4Y*a—âßÁW+/LwRNzmaL=8>¹|òb1åBõô>Í)XV­Y¿Pé×ÊÇnn'J—IÙEX@Y‰\“kÈËbÐ3™¡ÎSãd%96<ƒbBK³ý¿£ÊËœ•åÙùy69àûþ½nlûºw“ñ93|L7áñã‹?‘™þñÑèo¡_¡1´Ë?XýO©ŽÇOã'ÁX²a]Qœ_ œ <«HkAg`L¨¢H •™¡’ʹùZa"ßB˳5Êl K-Ï*+(´Ø üå™áþ=—F„·”yëià̵« ß¼•»`{»n†B8Àn-ÌNÊ‘ëJB%h.:i›|a2Su!¢P5öÊ6xÔ|ÔX® Žk¾[ý·¤æŸPMz®²@ãÈaŸÌ©@]ÀBðŽmF"4P£ÃÂSR¹¶tï!X gÃcþ°0‘g·Á9•¼{Ü:>ü;±pœ²ƒÉj8Ôdî×V¦,OjEx„HÐhÀp0™‚†âõQ ar§Ì?TŠBA¬öhŸ¨Cùp¿‰¯Üª”k4Æ H§TÅéî`×/Ÿœ7ïØ•›ëMk,0“…é­.æ¹Åj‰dùÝ|=Z{2=ºÊĽºÄÿB…Nendstream endobj 176 0 obj << /Filter /FlateDecode /Length 10574 >> stream xœ½}[IržŸûÁï~ktÚž.åý²À °­¼ÂJXøa×zH‡»$›"çªÿvÇ‘UQ<½3ƒI‹ÅT+²2#ãòEäåüÛ­Ûü­ÃÿÆŸ¿½ù›Ïc¸}õñÆÝ¾ºù·Ïÿz;þóüííß=Õ([wÝß>ûòFXýmó·5×­Ç|ûìíÍ¥Ü=û½›ºy7–­æHï?{qó‡ËoîÜœO¾\ÞàÑ»XÚååݽÛbo1—Ë÷Dö½ûÚ.ÿŒ7zo­\^ãÚ[ÏüÂàûÕÝ}JqkºÙ×óŸß)÷^k`²oÔB»|GçªãÇ·äí·Žø[ô”“æúFµûo´Âõ¶ïîCÝœþò‚¨ÔÁ»/‹ÁåÖÃ壴©×Gñ¿žý 8{#àê·Xê.àÏAÉË­é—ï£k›·÷1o­”…^è%¥~ùÛ»ûâFÃäñÒGë—‡ùøÅ||3_ÎÇ?^æó¿ÌÇßÏÇWù^ÍÇÿ<Ÿ_ýœêÏÇ«ŸíÃ(£…ÜÚÞ?~3?ÎǯçãËùøáê»cœüüv>zõÆyÓQôt„¶åF³AÿM¦…§ÃßÝ{— 4—À\÷¾n>uzÏÇ-çã­}¦úxž¼™yÁ‘HžýîæÙù5¨þý‘i@úi þÞû-ÅÐN=}] Î…^/_Q±Q/•QyXn9„lŒ«ÞåLO÷‡­<ˆ-‘K©—¤F½ÃË¥çÒn7=ÕÌ“åzö>zþxò5’ÏØ öv²k !o™Lp—Ìu#P:õÍ||{õQi¥Rñ®›†âóê'Õ$‘šŒžþýìéÃ||s~äuM ¾¼Êõr>~µŸ¯çã÷×õå||¼*Œ·WÅi,ûPÊ_ÏG%Ãw“úöG¸éãñ¿ÎÇß^·Öïçãÿžáºá~6›{1©W}ÆÃ|÷×óñ_æ ¿¿êT”«yuõþڮí0¦9‘AP¼e€fÃFß_ˆ×OÌÔ!ÜWßýwíºwß_mLµ dô·óÑͱ|ø1ѾžâÁ²r÷á‘”m<ÎÇ·óQõuÓ&x Æ]}!ÍÇ®©‡7UFìôã} µ¹ CšyÊÀMê6Ë|ìZt»Yç+^áçŒ<üØÈóÕ•_%¯ç³5¨@~ i(™(²ïýTpIŠzÙ‰XÇôQYin(0!?Æï"å :/UÒ‰zy7-‚BAì°ü0:ÉFøfŠà•²ô@s©2D'©»ñ–äó´/mÅŸAÍÓV[µžàèÃKéýÅ*ôå»ãßß‚Xz½¢Gÿ‰~p•fˆ5«uÒ§øK4ËZ®'—BJ6$Z”úEG3ñòÝnmµÔ0”þÔWùšÅ‡Ü6wT؉%ÊËÆöÉ7´î†íט„‡•˜Þ ¿ñ©8¹—´E²ù[ÃO!ùÒ ®ê?žûz óÞY޾Ãmh‘ˆ……âÎŽò#ÝD.ëݨ+´ŠLŽ¡>É¢4J»Ý¯‡{ñ-§"’é6”£¶!ýÈ ;‰BoÿzJÇ»á³KU“Bñ8UÂqƒ–KýãÁ×Á’'£êiÿDâ—#¥»Ö!2¤QÕÒà4N<Ðg¬»ÎûÖ$yˆq«©œ’‡Qå$>"‚.ý¥@Mo¤‘W±Ÿ¼ï1ò.ù.1E´|†Æðé…HžAyiº#F>cõðDŒˆÉÏ|7A#×\ÑîPœˆCášäOp(‡Áê°±àJd-ä®s×"@Oü@_#›Ÿ=x¢ê©4Vùxó—ÓQ½¿Eç L/%\É“‘ó‰°3Û}”B)sU k—¦û<¼ƒ|¡F1…-‡£òª; àP#™DbKrW#•ZW'É]^ñŒ†<ïç|ȘZ‹¤4£oÈ`2ΫuíT.¾ß{iã$ëalÀ¿˜ÁQÇ(ñ6µTÄ0v .µ=$PP}RÝßxð£|CûN ååXC>™wÉS¹çæ"ÉLM+ˆêˆš0Õ=!ƒäQÁ"i¹zÒÄüXÙF•& B 1žV ®Y…˜éÕèæûRþéá­¶þsÂÛié†\d0÷àBPßL±*WôB$EÚxyx3+£‰¡.S˜õòôbuãAÉU\pEH%ýEàÉJR•þ«ï= ;¡‡þ§¡N8Y@”îÜ V" áŒÐÓNqFòžÞœ]¡z%Ðù°û(NÚ”§Tð社P§{J57ƒgôJéÿ·,ïF­Ø'„Ñ' ááÀ4¥øEÆgõÔˆa)x¬@-ðtÅàýô݇ ñ=Â, u{¸çÊ®€žü›^ñ¹ÐCo&”þ Ÿn„{Ü6º²ÎGNÞŸX~1ç^½¡5öá‰jË»§_Éçû¹F {E4q9yŠoñä ®Á üÞ {AuýPŒ¿TÀÛô­º=qÓ9Åú ÄÞûãèp<ÅÂíùç}Q£[däD])º :ܰíÒ|~>3Ÿ'a‹‰,VÄ”9F\¯B½»Þ#N«‡÷{—-ŒRe´‡MØWíõæÍÔɇͬÎÿó›ÿ-6WoÉ5¦-ä[£}£ìÄ2Ë퇗·ÿóöÝ»ýÇÖ@øÝ¿ý'úÿŸžÜ²Ñ+å[ÔblzËF@mí–3j‚!ÏÞÒ?<ûpCbzö§£;?£9_vUÜÊ{Sm l„VéŸB*ZÐ`ˆ~(r`1ù=¤l6Þ’²n„ët0óRó-e·¤&qň+J0·¢LyAƒµl¡éa ,? |F²%WôБÁå„L û“LÉÉF*HC§Én+Ô°Á•e’a%±ÆNb£ª#Ÿ· ƒ)n¥ŠÖT·B¯r²àP;MöŠ!gµçWÌqÍH‚H ©Á°`ľ•-zø.²—ºbN(Ä*ÊÃùÛ/×Bš\BqÑ7êáŠ)éܬ…y…o­”Ì.Pž˜—8×ND°ãÂðŠ€Ç+ ÔCBe·&—OL!9†­Ç%†L¾SÝ©ƒ+o” ®ˆ ys@šPvR©½m-®˜ãFበ8Ä„À5 «Î¡ÊW8kG`¦d†Â=®°d×°¸MÚX/PBÿä¦ÙNV`k`†KnX_ 6¾DšoQdÍaÒô(}QŒ'{ña ª)X¼â€‚ËXC=Ìäk:¹ÙºÂy9À荓»À½’ÚÀôxMcň‘&gqÎ+&™B^ΕCÞ\˜¸b̯ȑ)cÄ|*¦°$ç&ã uáPÀ XM!tMf\¶²$aÄ>«.ºÝÖ¸Bš ¨LAYs ¦pâ2g¢kRÚØ|)¡@UuE) SÒ<¬o…÷/˜”ÊÈ«ûKÒ—`Çj Ð@&"‹o~ói… +õ°5n¸ô=$ô|à\tE™Á“çr¾p¨oK:ˆ²ÊgέhÐR2®0”D îŠX;[’Nøˆe†œ¼Dº¬»†¬–þq fÀnš]4—T§Ú¡6ätÒ ËCˆ¯œŸ`¹h…³©„fPM „ W˜2ÊHÝaèBÊš? ­qmQŠ¢¶%å)J ~HFû’E–— JÈ^1'Дq!ÃJCÀ†Âu°a Æz"µ—=,q5 »'hÀ”3. ¡±ŽR6I/)Az²;šã–¹‚±[îaÄQˆ¸&l[c%lKÖc*¯R<á³kÊÂØl-\bÆÔ¯6ª^K&:Xxeb…øp€¹wŽï~IÙÙ"B öl®Hw7q‘!Ó—L0J¸pý‰Wˆ'º3絋JS G"nc›_R§©”1RCXÀ«KÊ*©Ë*-<öDC %9b—$ŒÁk a ×sd!ö¥y0}þ7”^z^,®ØÑ›nß%—-ðÂb'7sû†(¼ÙÖP,×››½ñ ÓIÇ*å2‘ò-¨ <}ÆHßÜøŠmUS,Ú •RJõ(ÌRj \©-|~ ߯kÛ†b¹¸GëˈþȲ*ºÔ;¹j‰p Ô í År}B7 %~ýè“gN6zô)5C±\ÜÍ!ä¨ÜPC”B“Ç´‘WÇçŽ-–b(†‰ÛÁæ]$)Òž q“·m(V…N  1·ÍümY¸™Œ}õÀ343gŸ’TÙ Ò)þx¦'@ E1LÜŽ÷eƒµy˜ÚñØÕ€Atäàò$áÐ Å0Iè¡WðÑUúCJ –¥“ô§ó*ŠaVXA*`F” ó˜úoߪÌåwúoË"„U T‰O–‡èðN‹e×–$¥3EÑ<ÜJ¡™ÁJx&ƒ oIíTÀ4`’¶‡føR±SÏP,×ÐhÏZÞ©ï””³FZÁçöC‰FÓ8‚¡X.n©:·A}IiÙ2ªÏ<ï¹1²åïgäŠbxD ±­õB9°RúaOžµ†äJÿ¸¡–:)'.é×Pa-FZäï×NùK¤fÔ År ¡êGMž˜I›g eXI¯ØK`(–‹G—xUî ›dÑR¢±ôÀª†$r!Î?íûbï1`Bo óôVÌ;@BôV lÞKG†b¹DB¤aúYP‡g >@­§Æ$ëk %7ËÄ]BrÀË@•wJR—°°tE]kÜšµ ë&Šb¹D±C† ¨je5*˜R@è(ýöËÍP,ÓÐ#Ï¡¢Ó¸ Ï%û˜òËeëì-P«ISN\28lü†=RKMXÛZIJƒ .°#TË%òÆP0›)‰ ñ[X¸®¨y‘7ÇM±\‡n#VÔVEà8åÈ­ò’Ü®Û1Šå·ÖDÿ Í"qˆ|0Ë»ËÙÖŠ«“4Á2‰Ÿ…IDñR•ýcFÕÀ.”t¹±“¿N¯Kän!ŒZÄ…Œý4ï8n3TC1L#’U6"Ä©(>» áB«‘¾«¡X.±| “d5]ì•켊¼ÍÁUS,×1÷0ˆJßë}Ÿ{ØvéX9Øçž1ȤX.n)zËâZ#Ö¢âÄÓ¤&Âçʆ –GôÏÐþØÙa³CÚ&x’€òCÓË$Ò.È)؇BÚïçÏU¾bÂf`C±\"í˜Ùk5Ș…Y†YQFž% ‚ÑÃ$##ÄÖ‹|­ÉÐÈÒáùÙµgè‚ò¢(–KÆFsjÔ'˜XŒ—*Ò6†æ¤M‚å+#(¢LÁÊ;{#Ä"X>Ï{ŒŠ³çÍP,×°×Îv“#5ÔÞæÉ-ŠiU‚ȽÅ0ÉÈkk€ø$¦ÜUžl\3Â#!— UË5Æ&Z›QDcsŒ2y,F©l®Ù,“ø~”À0\$D¬×„2º7y²K‚É'Å0 Åö\i¼ÌÓO®?cÒpoO6 œ{”×€ÖX±‡¬ ¶Î=r•àCÁ7‹H&Årꈘ‘&*õð×ÝP “¤D‘!¯$…s"BðžÁ«IRD!'ŠåcëAâ1YR †÷IwDS/˜fd2“b™DHU†ÛG4Ô$7‚‘d’¢X®³Å&ؤ«Älq÷ð]“¸Š(’4åÄÅ-ÁƒãEg@r¤“(X MO(†b¹$ ‰ô=x…Žíœ‡ä(¾¬´Ñ#†`™¤KØ0‹Jå8Á-Q|;û²üjþ4Ò›ÂP &6l¨¡(áÄ áÛ¼6d(†i$i•¥Œ.)º Bud™¦(†IL ëûÈÀÈC à‡s±Ypæ€yîÅr 2&Ä*fÙ÷‡DŽ™­;&@ÐDcÝPQ “4”£xb1û>´egöm×êI1<"  ªÐÈM‹ÃÎ0)ĆˆÎJŲˆmàŒOFñÃEl÷ÁØ‹„ X]Ñ[™v •Àdçf°]XéýžúFަX.ñ±^ò Ì®šBÕ±o.â› NpÜÐËuÀD™Ø¼˜ÃD™ˆrOÙq´JQ,×0{1ÄVÒžòù!Ê8 ä§·¤þ>±H3I0‰Aât欗Ý9c)Î#QaR”ÉÁ-´Ð¤ ½ú[S”Ò¹–În^RHE1LâZɧÁCtŠƒE”úØo&™ ‚-‰c»¢X®Ñ#ÇQ"ÎÐ%âëØcˆÊãÖæ2G E±\#Þcu:^¤b@lo‚—úɃë1šb¹$Þ{qK[%ÞYŽ€M”Ý“ê?O,Ò!‚Ì©LYršç óFÀ'Z“b¹ÄÈ<GØ©:”60íÎRéB«ÕPö÷GÞ"Ø ÊvXz`LÀ …3\JÒ=Ïñ¤œ¸† s’ïÔÝ…QÜäZBª(Ý?(–Iœ*%`\.ìa/lׇꢴÄ…’3­ÉP,×(6ñNTþ\E$ ŒV½¤]¤vœiŠåÑ«ó 6²qÇŠX’ô âT÷^žDQ,—8h`L€ö>p½ŒÃÂ($0‹bû“29Ž" ï´:*q©3@a0!WODM°Ü¨4fFšb˜ŽÄ¾.8¤=qŬu\R;žbþ´,£'5³†-\â„gG,×Di¤ö¤(–K:äG"{¢X-Û>.YMŠa•ØÄë-H·ŠÔ™|ç:'Ê'\éò€u’$î”×ðõõÈÀdî+v¾U)ÉT£òÌuŒI±\Ò'Ê—Œš@ÊnŠä»d¥o(‚åÕ8'Õ`쥈3’D–JBÀ>ü” ÅrÚNä4©;çØX@ˆNL”¡·&X¦òÔ#ú8`ƒ£ï’^GïvàgDšb˜D¯£© £¨Ò®ÅNêšĪ$û“b˜(Í (™J$Í¢×”É! „õ¥QæâÊ‘ˆ@¹ÅQx©‘'ES,ZâÅK,_“•T¹ÞÆp@™þ&xEv2 ëEI9/ŒcÄ„†3ær8”"|œ'EPí饃ÒQ¦B“0_:(ó-^aƒwŸß;(ê­(ŽN½uPÔ[XìùÖAQoÆ£G8ßúK7ÞSƒØ æÜiS)Ï-™Â~Þá¸é4!•—©ò£O¸°j,ü˜pj£G¸”–ö›²xåø8Že=Ç1sP¢GiPD}奸8(æ­Y™Õ[ƒbÞÊXgÕ/ Á¼ƒ”[VDö—żÕ%ÇPo Š~k g¾¤Æ÷Ó&“pÙðè(P§G”ÌÇ\DàäR!ßûè:–‘/aîõ=©ç|m>”öPðD©å°C®þä'mU×lw·Ù“íþ|õ \È|F}œAº| SlpþËÉÞ’ËïŽ= è]ä³-ÿ_n#ˆX„DPÇ®[g"< +öÇ ·L|)ÈÙQâÅŠö°?E)ܺ`À)#‹^Ø^ÄP]Ö2rÜþ¸NŸ±jŒµL€׬; %m¡‘ ^Sž`HX‹Ê~áQâºrÎEZíùÔû"-ä;»z,ä×´Ð-*—…=Œ¼–‡Å°brL8°Ö5˜8§²ÎŒ±ôèœ\jÓóŠ#÷ë!ª}ÍÇuí%Øš_×ÖÉP\YÕŠ’ð««L„U0,ô1ÈhëÂþ¡æïä¤ö’CÁøµŸ„½jÄ„r·„Bq.KöÉã‡Æplf])ØòUVúé&—-ë!Ôn ç„ðGºEsyᨮÀQ.´Yg'ïgZç©[ݼ÷ “ì‹+Õ0QËÂ,"6þ ’uÑ?²°ÔðpåNYˆªîúR/˜u•áa+yY…àx[uX§Ö¸â½ 8¿¬A—hrWfïMÐåBKN¸ss¯I¹ñv¾uÎçZÚBЀ=t+uÆdhKQfEö¾L€ÙÉ=«ˆw¾}lYb‚]ëPöû(ŠaâaCšvcÇkÄ.*ôšoG¯#vQÕd(†éèÖ~qö…wáq‡¸æ^åÔ÷'{C±LÜ¡†_™ä=[MŽê>É®éŒ]NüýæåD•¢X.î’'¿À—]†£bzàÿ›¦X.n)ŽtuL±ÈY¡ºLøA$þa„I1L26àŽ¨z„Ÿä½»Ç÷É©å-‚ejr[Ã14\˜€Êè–µ°H¡(–kÌwâ¦yËrzi.Mœ„TË%š„û¦âÔHêQ³úGþ•kŠaâv°€ËøûPÉ„X™µ¢8Ï¢MŠåºYݵrËù­ÜŽÏTjå6\Ò§’xxšmªnZ)ê–ØÁ«(–‹åíø7”ÝbïcLÚJŽäC1LâH(ÑC0UŽÇßÍ÷ù˜t3Ë%RÊr¦xÎ[©çyƒM‚ááá̽×"Jø]”®E”pR?C±\Ò¡.§ • {IÛ;î`a{ŸË%Brrɶµ¹ÂBÂEu\ Îr‚5&z ASN\2mNtù𛳲wAä½âÓX&îoA©F¹q‰‰ñ®òóŸ†b¹¤Gä'¼ñn®Éy«iðnœrVË5ÆæÙ+N•Ä1O3ß·n™Pb™†´ßt(RörûÉl'áÈ_ÖËt¨v°šÝd¦Ñìø‰fŸL l|{Û>g¬|Ì𘡌óÏÑP “hv©6f'Üì(åõC³_UË%-É1Ј š|c·Ã£wIã«Üv4Ë$¡-ñÉÒ9e¸øÚ™ òø2C0<ÃeKÔœñ7Ï$!c±VQ,×1÷!›IÃÁoÜ!Ί[¤c¹ÄB²\z0[¢²³™|(šUÓÒ‰‹[â“ Ué#´¸džIÑ È:$>)<#¤8®Â¡w²ÓOQ,÷Wå2õ‘D›Æq°Ý‡¾2Ã#æŠÍ]˜îÌ\ަe9 EÞÉ£iÅrÉȼœ œŠí£œ&œjìáž¼¡X.1µ(÷PM4šª5ý”}µ±ÿÄ$íäýö›}l·­=õˆ`™dh”yõ¦ÒkF"]ýNʉIzú¸•c‡þØrÎÇb  ’M2pâ =cCy]ÒºGr¼Áà#Ë$j]älÕ4œë4Î'âz‹fÐɉk@mlžW=jX_×È?5È"QÃ4v?{lküðØ)|â±»íQÆ)FƒØp'7ˆ-79º©(–k šÂ†£\[•{à”#K’Ø(Šå)ų”’Dc%ÿ‰”âYJ gWRÔa–=½~à &Åð`/r‘ÐŒû€ç½™¼*ɵxŠb¹D‘œœÏœþqÔQÃhul~<¤a’d„“žý„ã¶ÕÀ'9Ñ—Šã+¼¢—+zÌ*'–šb¹D°farí 7¬L³(½±©ª\Û0IKPA âêÆ+âLp0`×2 päú»CBA ;JBÔ²I,¸[8W*ïfæä~øV Å0 Ä޹ΠŠQEp„.=%¹•sR,—ºÊfV%i0GŸLyäÄ$ö_6LmĽÔdž,ç–µG0\2¸*CQh‹°>¾WY†¢(–kXHäðrdìÉUQã#?§„Yî SË5rÏö7Ñv,Œ&¸®¸Ö6Ša’XBÂUA’‚Û¸íP>ø × Eóˆ×Æ9V3i|ÜÀ˜‚k^LåõÄuÔYªí¸±ØÜ“,õTy±\ÃkË Ó¯Õ.? 9½jAÝxº—Œ÷÷˜bTÅ)fÍVäöJM±L"lœbö&$ÇwÙäLH:q‰Ý"o±0 ÕSFYÅà¿3—ô)5‹n/ñbÛgІb¹d겘ΡKaÜU65¿d°å‰idr“£ªFf7NêÉvdR,בÛ6“Ý ·­&»ñ¸£Íd7'®Q´Kvæ8u¨Õ휹3—¸€&—ßÎ>¹.ƒÎ $ò‹“b¹FÕ–ï[Ôr’ƒàZ*ù´Ørbe 3ž2¼™Š8°[0Ë$ùÏ’—n ã8)ö¡üØmÆc¹0•˨UÀÉÖ‡ƒ%`ÚåÁI±\ÃQÊÞ€YÿMr•*÷â¾S'±LÃ$£Ýg×vG\;g3'Ë$p‹òÏ”H½è².ˆz® (Šå: wJ'Ȫ…Ü…!‰…ÜÉöWÏ2rÃjC×Ec>ž^ År ¸Á¯Œ%]’ ¼Ú3!.œÉÍT$ Ï’ "®ôÁ„G²îÈÝ®þô­I™Çþy¹2èë&E½…º–ô[e¾•éá£g¿Šz ; œþ⤨·N#Òcü%—DÜÊEæêµ>]»¶ „ymêÚ‚ã xç+sÞî„âq]Û•wßá3ú\gÈWÔîï ‚~'Õ*—¯ìï ‚~§ pÔÕ;ƒ ßiyܯ¶¿3êèRÜ•ˆßÙ ú1ÔùŽûϼÑ÷;f¾c·m§y£~­LÓþ˜ÆcåÛ%"~r4LSÄe×±ðíòƒô®LÙÔ°Ì7EN+Ú)óƒc¬'ÂùZúxFÜ]~­Ò`œÄúZ”*B—¾ÖཹÖàúÑMü 1~][]ób4i•'—Öø\â7wn Î…^/¯ñH¹ii—WßÜÝÉôZÃåîDhäG//q.ãôÿ+²Ÿ„…œpù×;$X•¢Èåë;,ÚõÜúåájäi#—4ˆ‹4pø©¿|y”kh.üÄ¿”çÜ’žÙ¯”v¿”7ÈH‘î¹J_b-ŧ¢æÓ³" äÍLðQ¹xy>çJ©ÝÃ.¢Ìª"'z2[‚œô¿€}:Ü„}yøbê |2WtÛª#ª×¯Y˜ä…òP¶’Z–Oc:u?¡æ,¯î/_Òx¹ 4]´sQº¨óBÞ.¾)¥z”¯Q‡.ß²Ûò„©g¬‰XPm×D­îßÜo¦2ʘY`¸ü0t¦Gˆö˜5ÝÊÙýýß%0ûù»CÂ{]s‰ònÞ{òz ¿ „Þ>W ¯zõÍt«Ÿš®<¹·’#Ô·°–.Z;ôÛ‡Õ3ðœ}i$§¨÷9঩ÈãÁG<Ä=BŒ«µ˜£v$ãÇÕ><ŸˆO”ºD\¼›G|zv‡::Oj…dÇŪû©Œœç¹‘¬mÉŸQ—ñ ¤Äÿl_N ´×¥òp°ùOü6jq>RÔ›¯³¥:Æ2Z>Ïþès!$uùwEVÓ=dKÉá·RÞ)±þ\QâÊúLé—%ûäaÎjtO¸4öÇ,ž•'Ý›x¡¨:Ð?"š‘äëå;yÁ÷ø¤mpqž¥zUfQSäñ¾™>L9«¯åã÷hAÛøã×2_œ=+àn Ãn2aL4A¯÷^òÞ;\ݾ‡ M˜ó}xlÄ0xHWŒ¸1bé,½SØ“ÂôÞ™áo>§,GÏît}÷Uu‡ ’Ô²sãñÍ||9¿Ÿoçãë«/üÕ|ܪHæK;üþqŠ ÷Hzê—?‰š»Ö½ˆò*Š ÏeFz"´%™=ù V¤ëô©ÿÞ€2êmÇùɪÿáò;Õ-«Š‡Î<¿ã_‹˜Ì{ìî'L{J¿Ä ÿÀY±½ÑµAŒ¸5¨^>GofÄî ¥ýw°+¯Àx@[(/ S¨mÇÉ„NÄ7‰žß Rî¼;ÿá¢Ïöƒ »¶¨A¤ŠßøÚø²¢ŠQÑP>,’ë§è‹.PTÑFüZÛGmTÛ«a <¼1ÔX‹…¤ðG•²Vžöcü„S¯y‚,ÌÒÿ#£Bendstream endobj 177 0 obj << /Filter /FlateDecode /Length 6547 >> stream xœí]ÝÇq?û)oy[èi6Ön¦¿»%(€cK†I”â@N€%ïx"Å»¥x¤E¾øoOUõWuOÏÞDÊè­¹žþ¨®_UWÍþ°™÷b3ãéßÇ×gÿü’›«Û³ysuöÙ ¿nÒ?¯7ÿz=t€'û0±9r_iÕÞ™3n”Ùœ_Ÿ};}¶õj/g馗Ûy¯fk§§ÛݼwÁ3]Â3;[íÍtOMÎIê*½–úBKÌÊúéñVZè`üôÝv§”„¶˜>¿Ä×TðÊXê,Â,¥n¡i¼UZ·ïå)þûüßp£ŽoC¿—vr~q6…íù³³Öb³Sfï­Å§ßÂ*ç½N9·áçy¶aº†ÇnVÒ¹)-(hg¦ê-aI° x¬çÙÖ©ŒWÓ‘ºÌfvØC ¯šé<œ,9ð5Ç÷|mgF"öô†½w|…ÏÕ>9nb+Ät|A“K¯$M)æÙÍžx¤aì,g1í·;­ðàÂtŽOEÂ!Qa»^.ãÈZX8@<âyÖ0vø`;âf¤2^æÉ¶;xd‚Ó3¶‡K.(U”""‹Èºñ9N´Ó³Û+m7;¡öF'øÅÌÏÂ!YúZ‚|~¢êÆmœEÏŠÛÀ¼/àˆ-H…‡cåCG>q¨‹ˆTYAþ޳hŒÃ9I»üµHg+M¨tÖÄyQ®<0ÍĵjÔÛ°:=ýek (sRÅE™­(é³`´åKêä˜V7#q‘¡”ûÙèÌ©ß!#Á!+î²d6ÖE}zŒOʉ3âÓWñ|°U!j z‹NÛÁf ?T®2/ZS±ËkŒ<2… ^ÑGu°4!Xã­a¿)©¥€…·¦£,WU—Œï"Áœšý=LÇá9=öÐ¥e $œ…CŇëtÈÝOòš;öà„¢5k!øhŒÂœ¹ñTP§£}äÆÂ?/¶¥yx̶ŒˆBÓkí¦ÿ¬6œq,ã°(Ð=­­eÚ'@Ë?ey{dF‹ö[9“•ŸÛ@3ka&¹c½/jï„ð>™&éj"bymàe•Y>’¬©n[j«óaGÅñôr¾4eÀéQ5WD 'náËú/â Iyåu,Q&ñÈÕw•ëz6WÞ kOª²4=‡¸Äü`‘5™ùOpyܬk´ø¡²è!öŠõòÝàˆ ‰kà¼ÈP#BZµv´’°ªÝƒm¬ cv“ X”·E’b2(B’̳ .QJ k2ëj¨¾L†_Ä¿Ç^EAj°AQF eÝBÊ€yÓγ¥ýŠÀM;ƒoìtÜÀ]ÆáŠpsÔØ 7cÊ%´RÉÅW¬Ô*¨òÞ¾ª¹{ÕFˆ}7‚07ŒªmØœ~vþOßN 'ÇY\|Q#f“À±g¡4Ù3ãÿÙé rkP€gñœÞa×ѴÛ*!5#ÀkQÓ 2ð²’Q !p‹]’Á0ÓÍE<8ë PÃò‚·qsà0ðqJ)"\ì2 ˆró’F˜‰wK’ Ôç¡iÙòÍmÒmF˜4嬓°Ä§C?N@úBÂ?&ª ·"²´L¦ky¸ðÖ“òf9âTÀ½±s8‘äh–ãF«ì´Uœ‡ãRM³TΜ•!«’·UjF °Ê Ðü™t90ñŸ'|3€6wv4°V§©wûàæèxsH ôÀÞ*'D|GÄw¼µj4Þ;Ò·iÞùx»Ö‘ü›–LÉGÝÓØü¸vX.O°åÉ€ŸVçÜ¿OŽ,€áÐ¥™Ñ·åïüy[BšÇ±yô^já7 $Óù´¡9Ç f?šÆ¢Ÿ8ËÍ®y©op•»8%€}`=ò1ý¾__Ï JiWõÝØ·{7”<³Þ+²ž ó:*Ï’3Üü Z¶‹ü4’Rñ†c{Å´ß fÔbŽš¦ö"ØX3ý ×@{Ù˜~Î-´YͰÃ7±iMYZ@£TÈf<Š5Gv¡63†+èàn wõ6ƹ#hŸ²¡~Ø– ÆëªÜù[; å¥A„$MZ¥îÓó³¯Ïb„Y°Ònór=¬Üò[ +K@macÁÁ´cX´éù³<ôÇóû  /*3ÜB)æ(åP¶1w¤sñá¥Ph„<Þ«.tS‰VÙ©;(…8\7>ÀHÈw–è•ÝìP#rȤ,b¬WŸBQ ý-m‘wpݨš&øoÀrÁŽrïg…óFCXìR"¢o°BÀ•Œ§nÓ«ùî €Ìe}xtA`tÛ IoÁa£õè`¦À°l ÚMÿŽœýÔ&ì4Y¢2xŸP¨í áDJ«2Gá‘àeGƒ?ñ­ï{›×€½·À&A<~!} ä|¨îa šð¯Vi/ÿ˜ö—gÎ\¯é‹j7 ¥(5ï°È÷˜É ï0™ð÷ç1 §z—<ö½ 5T¼'QWÝA5u7Õz”´ó˜ô€q€`F²c_e[`'4¯‰mýi¶U¤öHIwo¶%°õmq\4z¯­_A²-‰$ §ø4$ú¯ÄPc°k {Á^ ¥ŽBÊ€ƒ÷N(4½šˆ›‚æÀs2fîÐܽLCXw„¦Ü…û‹­·ï@lÕ’k#ï¢ 8ˆÎs‡«c¯(f£º±×ùÖ¶Rt)à`5ð‰±Qæì¶‘J|G<{ãEËM¹’+T³ LNW¾sFg(˜¼kuÅ.o¿½XX‘˜äA«`Dã !GàÅl¹7uŒOƒ—ü’ìŽP’ {ou LW4.R4¨1¸Æ…)Ԫƥ( {™EãÕ—"_ƒ_@⌷é¼5ªçê– <·³ãׇèøô%,žØL[b³C¦XÃêLÈ~ÕÄ>F÷à,]¢ew6œ#Û}ê¢â–‚]¸û8÷+~r³$½Ç‚Ën)ô6“p‘Àå(vW¡´ÿ:ìñaù¢>=Ôæ«Ê½‡Úü¤6Æÿ¥Y&ö€ýGãJÞ—<É|Ò‘¨¿ÙîœRf?KN‡cm^ ÉLjúIÿ”&{Z›·µÉȈJíÏjóóÚüSm~Q›ß éÀf¾Ïñ¶6oj󯵹¯ÍßÔæÇ£½©Íÿ\©a;B¢¼P`J~éÝ¿¸k÷Œ§ž Ù–Íñt¸6cņk8ÞÅ3·Ã¾7Ãq¥)W´k¯QÁýòå1ê~€Ø1íiV”ö´’šƒWØSo*Hy>ü]ƒn\ %m€{§i^ðô¤[;pÐèE?Ç”Ût§wdwVã‹*fúÐ¥MRë.»p#ˆÉÀH¨N†³ÜG£Ç"¸Æã†é#`4¯é"÷·7qj+J¦Fð}J×ñœJÅ'{Øö ù‡8Ä Kˆ”Д8÷ëÚ›ù{)¿Ç©½¨ÑÚÄ/<ãÌF‹é*Â@ºûlÎ"“o—˜Š_Èåµ4[k­Óv´Í¹² °ÇÌp³\”Eò{c‰°^„ª1Ð÷f\WÄP“§oÁ±IǬÑc„ã];¦áƒW¯×¸¡æáŠâ+/š4–,L“à­`/É}*]ÍÁå÷ã+ykà¾øãžÚâƒ~{¶â+æ(‹¡Ø•5ó•tà>Îxø2’åa¨ømŸ®0û2Iœ\$&×mÒ>62¦5Ç#ökáš:ìª|‘È€³Ц÷ýXBOô7)® 8K ¯‹µ8Œ„^`z­Z>j n0ì—/Â#Íà‰o¹+½£ûâ­èpª–9¤E6q1Þ—òEA-1£¬«`×ò0)1+€kÌÒ:›”½.–ú[×{ŸÜ°4Ìǜʚ8nã`.Wƒ>ÅfPOÝ–À ºÄR:¹&xó"Å`]Œcy0êieBf08›^dÚ4 `]ˆ£PE#œ;5‘ä=K³ VI—ºÑëô]ÞOåøn[«ËÚ|^›/ÆšûƒÚ|]›¬ó)·¨7ûÚd¦†ù$/‡ö…Mü|8[䊥ïECÇ„PiAªs1Óz¨ñvÀ·qÌq!IÑx˜' W´ ìB~¼oIT ÕqÝ›#<3«Øð. V|¾Å§ÆZµÈ.×1·ÿû¦V`W“o©·òkâÍêƒXÅŠ”¬æêÇI‚ê²á2å†ñRTÑüz¼$Q“×ÔD‹i@#)>Å«cV)”Ä!-‹–ɉÌ%x‚s)m“œG;sàiɼMpê{C}Šr™ð¡k~šf‹¼)¦f% µ½˜«]â$ÂÛ¤D7¶§S^¬ƒŒr½ÓÕ:Æu .WeÉhTðï2ø5sÄvzùGÁúd_)¢ü<Ñr^7Þ‚TÙµ¡ ÑV|Ï/iAWQ|º2ÍJ²¨\”Tq6”µ‹L-Ûbœ,5èÌ×ÐÝ+æõ̱¼íï‹â˜Ï‘;ñOÓ5×£œuk9ÌëArj1ØÊNR)‰¨²)M‚*6—ÖñLÐdûX-ëJÅÒË8ŸK—â] d\f«ä×Ë”¢ †s…pº’öMz)G€¬Ô}­2¥ðÃ+³”ÙŸ“Q$f¹¦…Ýì¯Úwsülè•« ^8³b5JZ†IwƒÙ{H9®-ùçÚ;€1h/[F«c*+ŒJ§Û/Vkµ|Ã×>ãÌe鎿£r£ÉËÊoÈèÄʾâ|Ý9ËRÅ ¢_!´"œ¹+¥=¤¤‚C'ŸÜþÐuÓ¡Ž|ˆ»„%Eï$Z¯édJIÖ`‚ ¤¨eóØ “‹}P%½b~^yŽÀ°€nå²·"¦Ð09hάTÕéFàXYæ0F›/ÂM+V‡–µ »ˆ ±bµ¬"îàŒÒ«bµw´AÂÈÖð[ö"°¼°ªôQRˆDå[“Z©@Œ´íÐO2 ²J"ôY ,EL¯±*”¶Z¢jô¾2]_±Èk1½Âr–údŽÈ܍ޓgŠ$•ûê]”Îy”¦è÷΂͊…Q˜^ë$øÑJ‰*(mrÓ¼Ã?£°VùÉâi,«‚;7+¡ÞÈ’þ“=ë‰çcp{žl#(V×#BF[²@-Cζuñ1M!F-E¦o‹=À„ô!›ë˜]Û':Jpœk³kCŠ—œÄrÕvfPŽWɤ Dç6–ù³cøž;T#Åú¶X—Í] @óg‚úDÆXmñ)­™>ÓðC¥>—µ¿MÁ‹’j/ŒÚH«W*^ÔϨxQ?·áøxï£âeWKŽ ÎlúzÅ s®„ 8A.Ö:™3œÊèÇœébzíêmÌç-x]P«gxvöjþ=hU¤üûó}¿ˆQÈüFŸf>cr5¥™—ŠÉ;2ñ ÙY`^± z‡F?òK<ÍŒ[Õ@J3o÷¤07[l$x²eª/ ¾Iª4P<͘’ï(çGaϹÑߘR=)²H­@ ¿œÏ~gM‚k{RÜDØ+´ÓR€õU˜ÑRôñM°-žÊ‚ê Êpï¾ëDuÌ"5H96«È6‘n<ë½MÓxêZ ûSKl냭SíCÖçîZ+¾«ZLâÇªŠ¨/«–O¿}ÜÑ¡­'{…7ÀT Èc¿ˆb`%S?©þ¤J×'Oø$°«ë[;P4¢=Ö…"^xƒÿæWáf;Åé4j,°jV¢¹)×&!„{B™Qx¯~ ,…÷Fw™ýÂh 5¸û¥’ßmé«Nüúá_£8 ¬èm\´ŸíJ”âJ¼ïKß›2ô¬°VG»S ä^Š R3:äÙW9^ÒÆjJäɬEhúŠ“SîC(§×™}õ¼3ßð6RúªÞ °Z«móU å·=µj+†=‹BàÐëHµ¬_õZ°7WÐ3<¨˜“)Í…ñèªÂ„°~µïº¡9áÊ„F¿Kü&%ažfÁ°× L€,ç8UŠèBo¿ž-°×óðüŸÒ%LúõI X‡©(ïžpL‚Y¼Ž5¦¤€8hP¹Ý=A?Ü”(|@& …E—÷-#ÖþÕª/¿záÕß5R2hçßR:]<‰Ü›JnOñy_Óȹ‹Îô‹¼V:wó#8¯8#2ÃÀF€w7×g :à åÉó³ÿXuq9«_†EÒP…ŠîWÑŠ9@ø'$5P¤ç^eñ IÀ€ ¸îÒ¬ÒkYQÝjŒ8vÀOGeJÞ©5D„…Óʼn£å‡k<Ó/ 5„wÿ¯5þîµ~ôà^|)!ȶrPg­ÁÝÁù½|P/Ô‚‰m?BcRJê”–á(•ç®/|)€–Îü’—Õ‚ä\6VI^^g$?ÍIëÔCf_k!o§Gò#±B[]}ŸªS.Ò©²ZL[ÒºÀ³›»Â$†¾K5ø&Wsy k¢oº¿ïÓØ¶Í6åß3¿Ž‘}X¹¸>+Yœ÷TÞãë§tïòž„±ôí¢;p±sÅ{¢ üÀšž~{÷ˆyˇ›œ^%x€6~¥*œ¶¹ uµÌgbo~ˆ_‰÷”¨}ˆ•—@dÕÜÒÄmÊÍ#§•xÒtp{Z(ëô‚Ø{çšTÍõt„å‡Å·žèÎ…5—x”-j@\Q€šrMJ¥üª6Y^&Kç\-×ÄøPIêïî_ÓxaMùÖµž@-¦Mô~ý5ö9ìQís[ÂŒcnÇ©»¹äV/ïæ–"ÙÝÍ-DœR}Á04IšÍǨóŠt>ZÓÝ5—oAìÍÊJnýÝIÈì{§V¡I<¦2~”9ËøÓ×Ð›ËÆ21ê:xxlæXó€’nQ¼Í"T+ö§¥Åï3Ç«E\¦Ô”…)l¤4'䡯͊A8tÙ¡=ò¦*zÉ÷ÖCî™|ÛSa"¼«ª?tðÑIdÕŒW"ˆ³ðê}Ç5%ý ‚š4f kÊ<ž~W­!V‰•6‹£qŽísZed ôxˆ)Õÿ]Ô¦WüaBJý¥Ì"‘jñó\sP=¾Ì\ÊÛ© ¬{˜×zè’>âpºÓ.å M.]ºàµ«‰¸ÍãÉ5³Óþ” î5}<&-ùªÖø4…픺çgmV¿Iy•ñ‹ß.iré$@Záú\:^’V¯aÛŸ”9°Ï°5°ëª·õ2ÿQ£r${õ{ÍWÂS\|ø©È϶„/~0…¾C;rùE §¯±¿·òJ±‚¯Ïþt3:Þendstream endobj 178 0 obj << /Filter /FlateDecode /Length 3403 >> stream xœíZ[o$Å~Ÿ_ÑÊ =iê~Y¤@Ky ‰äõŽÇf=žeÆ»,Q”ßžïTuwî){lˆ”HDhñé3_ª:unuù¡lý×ÿ½Ø.>|®U³9,D³Yü°éצÿs±m>9#„§‹"Êæìr‘›Ê&ÈÆ[ßEm›³í¢•byö=À&NÀFuJx48{¹ø®ýãRàSéÚ"¥Ð.´ëåJt:m]ûl£ô¡ý!b Áµ×„ð1D›}»gË•1º \ìuùù–Ù½W‰-$„öŽH!¼Hk­;#§}R_ ¬á­Þ0¹{B£T]Ön¹R¾RÉö%¸ ²q2ߤaCTí!KÐc};û ¶r¢`/;À?EàÏÎßЪ*‹±±Ö¹N¹F[éiöëæ/Í탋/…m\_z¥B§½k¼²t’Ö]ïvw»Íþ|Ûì.›×»ÃÝz½Û7¯÷KÑ»ŽíîÅù‹ë›ë»ëõ¡ù¸‘ë•0Ëq˜ºsÞ6?¢/ñï{ŒéóÔ¥¡ÊÆ oèï¶p¤PìêfñíB†@Í@=ƒa”Ô®‹ŽFGé¨;ï9jàp”ƒ:mà¨ÃQQøÎhŽ8 •fi€¾±Á„ÎZš€1°š8rÒ,{”Óà¨3AYï`bÕs&(o¬“£zμïÂDVÏ™ bŒ0pŽê9yž÷[X é"·/©#œÂaþ‚\Š/)¼¬Tç•ÅYÙ-}«ÒO¢QðiU♞'1g‡±Ï <¬hÀˆ‰/ 78 o#Ù2ޱpInp 40 f0¯*†êÍ‹¡FCõæÅP#‡¡zób¨‘SPO\º¨Lc]èþ&G‡ R…Sð¡Ò‡ì›áúáñá8,~.~ŸÔàýÐÑ-Z!F)h¦‡0U Öê¡°8Š0ô &Î#ˆÿá_„õì7h‡À͈’ÖtêlDJ‹ ]..‚5xå:bN@ªQAÂGl/é4¨ØÕ¬7¤ÃN#bgA'1°¡ ­÷¤®µé=DVé+Ý98"¬4K: "KG%`Ô‘¤´pñaL§Ad¢ëìdg’Nƒ˜_Í@Z˜ÎÓK: ‚¤€/w ’£$$}û¸|]u„“>4Hï^D5=¦‰òdèC’¿ÀöÉkÈ_N¶Â çãâ2žþ'”íB˜”"XP›J‰¨s„út·}½„y£jrm×ÈY´ù%ê=1öÏZa•¯FËûI¬“†£Gy‰@c9jàpTį~‚8 e”–ãeÒÈá(_†2ÔÀ™T-OI"CŠ$À0ƒ&ÈÔcéS…áާ—PÒË$‰pÅ—üW8%ÿ jf¨‘ÃP½šjä0T¯f†95¨¹  ‡¡z53ÔÈa(Ôz]0,ÛçTô$Û ¨R*(^*ލ±TQ¬TQc©8¢X©8¢ÆRqD±RqD¥âˆb¥"¹¨ÖÒžãgZÂñCä*²HiáD=@•#ö²Ç© óQ÷ñí4V¡Qu‰cÐ$«<Älõ¡œù$9|ÉøpÎ|ˆ<áÒx¸ 8!ßA]}\YMJ”ÓØ”€}Ùyñ¡¦EÓ#@ÌOçÝñ:î HBŒ‚eÖ}¤ÆYÒiÐ/¨N;ÐÌÓžV œnuºx„Œ'W•”SþÁj@ýªGý›'¨)X€§‹Ö|.¾~´µIgPNà—t¥EšŸg™gB¢6ÏÇF-ë;t/¬Ž^/I6„˜¤zrWÈ»Bþµ-ô¶š!–˾ÏÀû\)ìÐ j( ²sÑÄáÈŠ:cbûñreáòé=K†~ck®kW¸«Â½,Ü›B® ù®LîuÐO/Ñ?ò¶t÷¯B¾Èо_È/êÊzWÈ¿RMô6Êøg!¯ âeµÝËçÕÕ;/Ø ùuü¹ûB® ¹©vüª*72H壚‚ÝÙW‹³ßÞ·ô‡¢õ7Õ…cäy!÷ÕuãË9±©‚­uâ! QØúa¤Ÿ–‘žòfN&ÏŽ'zYm³.äϵPfú»ª*¶UeÞp ŽÖüQ!OX~ø¿åÓ+…$€JdE‹m(Q@{UÕ¨:²8l(¤‰ªIW1ðH<_Åë{–y”~¨bÿÁ£áˆ}]Æ$0¥}\HQ”¶?¥k¶p»4×UÚ+ ÂŸªŠ¹Õ®ÛB²±v…”ËU:šCöU€@Ð)´vä N®”ˆ¨kUè§Èg«œ ´Ídz(ð”áªÿÂp?|.±åæñ¬7²¯í{Ë÷ŽÂjQlqÇ‚à¦:ÓM1¢¯ª÷ŠO¯: ìe½1÷ƒ‘¤;ÃéP¤ì´ÕqÊŠëi$·†ºNkX;1¢a2Ý»c{Ý”3 gO¸úq—vMpO;Ø’UÑñ-Ö¡(-¹þ0¬©ëŸg½»7e2[ŸôšÝÑXa©†ôM¿+a 6DqLÞË–mîë`ò²:äà$˜.&ŒSÇÈŠ³RjÉgüuƒzݤ9ô^ÀûI«aHa2ó$1x„šM~Ò;1 üÝ"€n$ÙÒuì%¶Ë¼–¦¢‰NiŸ›Ú!ÈĶBj£‰m}¹µð×R­ÖNs»ÔuBf»´ð$—ê¡°+Òn0soÑMܾëËÌu1rWëÌV`“oCH«äÛÄ•Pg 4aŽbójÏwªG£PK©!P]ðRx$o«e󾊽©fUôâÍX¯³@̇¨6u1`Lå¤Ü‰ÈAÑ"©)ŠÞ*è²3úTúè ¢E]Õ|áò2S‘\©»£§K³:OÈNÚñéR®¶ÒYÁM!×…|WÈm!¯«€¾¢ã'sÒãõÐ}7=0[í»èç{;æ }Ø €J#Dp…’àŠ‚Ž¹ý#(ÐÝ •+oJ˜EbB@´UÀ):ô¯ÉÎÓã¸`œžèyK©ÄE ‰IÞ’ºýòÃb›ØÿÞ—C£}Ò j^Ð<Ê 2 ¢]»Ù,cž{zµ¬3×$‹YI¶Öäœ!X %iQ"¥ÈRÉqÉ9UÚåï˜íßõš“ó‚…6B¡ypÜmVLÑ9´ ¥&‡®¯H 1þ¾| Ë{Ì;`d"z)8˜Ûäe`Ön*u5ö!N»xKËh]E· Rn‡„€µÒYÖ0¹‡¹¯"h;gæú侺†»êϱRÉÕ³±Y6Šû°9£Þгjž•R/3ÆxÎXE’3Y*E7¹s›Î†ËRå*MÒóÂ[2iå)¨‘ä!¨å^@¿¨šÚ=ûvimzý97ݱ0ßå!AÙŽ?Þd†·ë‚<6©CGoRœõÒÕV‰Œè~†ãkÓä ¬·éúîxP¾Kðnx&ûš/Äf]tÎîÑéq°AV7žîyžvA%’w®úP „ôlϰ÷uO–d§”äò¾KG…yqýC.žLÏLë‹ìÖÊÀôo2ÄGŸ4þyutƒ¡Q•ÒˆVÃéð£.0ì¯èƒþo_J°cÜÉQýûå¨í¢hoÇ9;(ÖîøjÖŽ‚oªKxt{´š×sÄ\ L7ì¬f[Uÿ»ªn&‡95•¾=Õñ‹j {àÍjëjo¬x¾©ÎâP%Ùñ3;Ãû©Ò±mS Å«Bžr_È—…üM(!;¤Ò!©÷ö?­:§Eïjh2 !ìP¼²›&fõ£þË*—Y0»)»«bë&ÎŽ ¯ªÂ˜á¿œ^k°ÊAWŽYkÁ°~ȼ>…½ªö¶ãcGö¬xþªX ; UU²~*ÊŽMM!åoÒŽ"GûëPŽ­˜—Ó'ö3ž¨]_`·…|[ø¾ªÅ5Ÿ"{ òoøÛý}endstream endobj 179 0 obj << /Filter /FlateDecode /Length 3978 >> stream xœåÉnÇõÎø#ŸzM§öE‰Ȇ 8p‚Ä`NrŒÈ%‹äÐRKùö¼¥ºëUO·HÇt$ðÁ¥bõ«WoßæÇcÕëc…ÿ•ÿŸ^ýæ;kŽÏ÷GêøüèÇ#M=.ÿ;½<þâN¸ ;}VYŸ¼<âOõ± ¶þ8úØgëO.žw_¯’í2±»Y©Þªº×«µêcNÙw[Ø *¸ä»3Üõ9Çhè¨IÎ8 +­lHÝéÊ8àS÷jµ¶ÖÀZwßnñ3›“õ묌 Ý–>ë\ûÝpÅßNþ€òƧÞxxÉÉÙQ§õê䇣µsúxm}ŸBÀíçݳ«ÕxßîÖ*E“”‘ÀùYÚêœá)k“cuDÕì¢ïvW+ñ¶# ðË•D»‹òYî®aÏéhcê6D—ì <|s‰ÐLÎ)…n+ #8­uʉÐð9o%ßwx:å|d2¤ÐpÓºÞÇÌtxÞ]ÄÐ{]—Wuy]—7³gwòÀáï£Uz¸ñ °W9x{ÃFÉè-¿Æõ^À2±“)Ï%\TÛ!ûs6Ý;þO¾D:§¨‘Ò…2Žn®pMКëÞ“\e,d¿&Q@l¸p”Sùå÷+†8P¤ÙóĆإTT(žÈé`|˜RG=l(Ú @R 0 ŠÉ}‚çâiƒo1®». 8àþu½„¥-%›X%“‚—¤B BÒíi8r2Ît»¸¯€i©[W‚ o×:!Ô|¼Ö¶÷.æé†¿p^w)r!<~x[NòKÛ}Tˆ»Ñʨ½2È„5œöÉfT*¼Ä«Øh©šÓ:HUC$hÔjËZ^ H¶d§ÒuG$ÔÖj7àWäx "š54\>ª%í*e»±aYM±{‹8ƒÙ´8=AÎØi$¯+ä°YÅ‘~ †IÖ4@ 9”—²7’#HåØ\ä‡ì¤F½®L»ªòqÎðINÖ‹/ªô²R&ÏZ Ÿø¶ª€²dási° *è2O“‹R¤°²qÆÑg#6¯„¯Ï€aãNpFZ-Ž,cˆ^ ïÛ£Qà°72$mÅx^¿Z›ÀÁ¦î¯r^ ÆçêúÝɈ½¸ ÞÉï„S€/Ó©dëûò’'¸:+v/ùCˆ;U1:“W cGê0ðÕTÀ&¬Ä>®¨{ ¤­ò㤡ßÉÀìš=:¿F„­¹B;¢iM}Áˆ$Ü3Z¬þâäF¾%©‹yÕ÷íè¯Ixû‘æ%ä¶ÇZ÷Ù{ƒ!·Å¸%€Wùöèä³çE‚å`läï\¢¯ !°¸íþ+tÏðpŠ(xwôøƒ5†âürú²O’‚Ê‚ª ?½ì…s= #D>͇hd -`(« Àm½àÓø±/±¤×%—„ÍÑZ;;¿ÓØØW³ˆRUj|ƒãæß¢€¹ž;BvØÐ!°×cz=±ÅxsÂz j Ô;Ô[¸gLœ´VmÒÉ' Ef×*©X‰£¹m!~¬¤V°†ÔæPäÇä‡EBpyX\¥•xg"`ЧiRvïªêJEzCÔ&6v¥ #Ç­° °ÀŒn4{gö“QÜf «6ÂÁíx^: f M¢›t{lWLï¼mÜìÔYJ“±áD=€(7ÐØvÁ;êœó•¸;Y%Ó“pHŽ^±Rc¯ëóU±0Œ[ëC…‹|Õ¸Sk)W£Ð$ä€Ö|á/‘yêÂB†vÝd|òŒ]÷ÖæÃ§#ÄŽmÕ£Fi"fDcjƒý2“ ŠbäѤ2õÊYÚóg5Ž+Œ^즮RCÆZóÌ– {䅭зøjÿiAdb÷ׄ´•n®Í³Ïì–ê'"“۬ꔶ •ÎD]ƒP#ë5xØ#¢"(?`ÕØ&A£S´B4ƒÊxð¤îð< Ô¬MåùmjVŸ ÊnM­ƒª–1O”‰K™,àæ=ܸ›÷ðÈRs†Ij‘zÂ!ÝÛÈ‚ŸvlbëRFxO'ñ¼«©öÊE;xÈïVXÒ r™ßC``,¥B\*MÍ]vC÷»º»®»/ëîE]nëò}] ¸¯gPª^Öêòº^÷Ϻ|?,S÷ëºüf5) 4Z¿¯Ë¿×¥‡Wuý¤‚;«»›º¼•»ãÙÏëòOõÀŸWseÕm]žÏ^üf®“g‘¿£è2?lU¯ B¿]]^Õåm]ÞÌž½¨´þ¼./f™¸Ÿ…+ùyÀüŸWž£Žªú{™ k$àµ`®Àö/Éhÿe…²©Ë‹é’>­ÌC/Î^Î~µ­ËW•^β÷fÚݬ>ofEáMÿ›*Šœ3Ä…ƒªÞ›Y*¹J¥'u÷—UÏ©êìg #©ÿÈí¡ÇRÏõHij% !¯Ð5¡*wu9o¶’¦#ÉÄY¡Kÿ˜ýLœ-:¬{gÍTôj­}ÔX#2üŽöÜð]16r¹Öà¥ñëXïK,yÔQLÆiýhW—Wuùváñãò|ö€ÃSy`¤ÙfÖºÜKj]wU]Úz@È“T¡ŠBöæµ}^"… ïç+hÌGº>ã¨QaêI¯Dψ Ü\¬?ÇXgÊIßa\L À–71ýšÌdI^D²,é®N=~Y¢reòÚhü¤Ž€WúÜöFïj^ÃàЏ?³õ$\Öè^ï°1Ê´‘ Ýù®¤ˆf¨ú—)€y— cÚ´êÁKÍ7Q‘?ß ¾Bc±Cn\'¿Fß|9åôõ²°Â}qqÐÀ yó¸äãY‰r¹.k;ôL£–¹TùsòSžÕ2Ä´};àÐÚ¥RìÑQÖ7K¯©”{,µ²0aÉÄ[̳¼iÛ‰â]¥‡ÂI×e¼¤­dbòUKuÏVk3PQ7jbÛºå ¼ÖvQUnHTTÀ«_WM­“š' Xb}Åßùle’¼ç]Û&f»‚hŠÝ'âŠ[Þ†xð§æøcO÷YjY¿[a¶éChb-ˆXF1v4ì€ßao¾Ú:/üÍ~56W»tüî´Øÿ‚—FåýÂ(Ç £9TûXPzY ‡º¤K1C\° @S vU_c7RÏ*ù>Òɪíüù®o)Éñí,ÉÉ€ÓžF9$÷+Qk™ð|†'>Àwh ž õm´‹_ò¾Qd…hµï¾ÂHÇ‹c“¦ê$žð ЉµçE»‡±ˆ1iÖâö¤@w^æ*<ܳÐÄ}ÕH?!çG±çûUÓ¶ÛžžH­Û þˆïhp`CyXìýj¬žÝhº-W®‡–ú¤LqüÎÕÁ-Ql¦&8¬‚rÍ6µewñKj+ÛS¤‰Úˆ®ø4ÌÙò›R[þeó 3%”ˆL„& Ðb%f" |€`[ûG¨°mýÑRL&°=+L¾é˜Ð¦£˜¨8…¶&OJLþÈ…ÁR6ë?ÿ…Áâ[ÿë«ñu¡é#_Ô¥(£ì%î©ñ-pe¾±¡ëÙ§uég!\ÝÇ•ëÙ+|ƒÚbA?ä}ºjàÓºü¬ÿË™2¿yD°¾þ<"®µŸä ,Âòë ¥ÙTî ºXè«P1 t8‘A? Üdm³BãpdL?±~è‚S‡p pg«ÈÖãT¦­È¬ëÜÞX`“¥“=‡Ó3“Q‰æñÝþàncÆQ™7Óˆ©Å4CâaÂpúsœÇp+ƒ¢ÍA×B;†+¿øÓÉOOúä +?ú1ø›Ÿl{ç½Jž‘œùS\â|Žç¥ØååÇï 0N²ÀÎ7(ú~Pqã5V§€Í`ÖNÖ?6âwóÃ+“”ÂÚ ò6lpŠéÍð2=¤DT™dš §ºà ÓRƒàa|¾N5¿`9œB ÕK#"Í”4‹0u'FÈ,ƒŸEãÐM‚³nÆ85xl?Ttü0rPy×ÁOrf,»¦ßÌcÁ›ÇÔ69sÛ" ³Y…Iåuœ ½k‚\“Ô6LK¯'z’]É@è}6 ŸF¥p;·“rþ :¸úî$òȦ 1Œ“-üš„®ð}He\éùB÷£ªN“¡ãž5-{Œ“LbÈŠ§,³êRËñ€@M¤¼MÖuø;þ%ŽHÆ‘0)Mï-rÔLP- '^Ëd©ÂøP@ãü÷$Ï]·yí¸G¿žð®ûa’3$ãtœÌ'‡ÖX žAÅSûËIQùÞHÙß×3sî6²zÛO«|TÇû鼯սs•™hÖœíåP§Hv?YÍÆ$Ù-•AËœ˜-­ƒùÞvÞóÛ”úh!¥ËfÒ¯¤l¿-JŽ¥,õºý1ÃÈ>$ÛW'GÿþBÏÜFendstream endobj 180 0 obj << /Filter /FlateDecode /Length 4787 >> stream xœÍ\[oÇ•~'ü#?Íd5ºvU H'°Òî&áCgh“͘äÐ$eIyÈoßsNU÷9USÍ‘´Î~P±§º.çúKûÇS5èS…ÿ•ÏoO~ùGkN¯OÔéÕÉ'š~=-ÿœßžþæ gx2$•ôéÙ›“üª>ú4ø0$ëOÏnO6ÚlÏþ “]ª&;3à…³‹“o7ßlü©778ÔÊŽqs¹Ý©Á¦hý¸yuJ:ÄÍkœ‘RŒãæg„“§ å½—ÛsvˆrÙkþù_ò)…`豎°BÜ<áP© hckíàt½×²í#ïä[oź8#:cúkí·;¥Þ\ÀS8 ñ©º/‘Aù˜Ìæ1¯`áŒý[üÏÙïÀ^Wz€i3ÿ¸*OŽQNÞYmüéÎú!ŽãüÌH£sióëíÎ;À=7Wô4(…ç]†O<|ÍÃ=/º¯Ýðð/ßò0ððÅ|ˆqó%?ý ÇÃßòPÌýËv[®?Êë›4x ÀéÙ«“³_|[Þ¥{O<¼i‡´âKZ©ç€¤eGžúÄÃKÞóßt‰!Hô¾K–ëîIÃ<¼gÂý‡ïyø<Ì4ô1.«Ñø=ÿ—‡FLÞòxaT,l§§ŸäÓeî¯xøŸ<á¿yøÀÃK^u7>ç§{ÞuÏðÐ{Ó=Ù O¸æácwÝBÀ¨MØüÄÃK>øX‡ ô¬«ßñï{>ÊÊq<ŸÌoO(Ì ñ Úbg h¶Ow$«Éä­~`ÆÿЇ¦«swüô¡+©÷ÝÅüúíŒ#ãâ)ÙÀ„zƒgj'tí-»ææRkÙU̪ò·îkbnQg=8kbC+½ÝiVUi ÝÃAó´ùÙjjÞÃÈ!¯£ÌL¼Î9¯³çáZ¡Á2ì[jÁÆs9a!ÝÄOß|Å-?ÕüTèˆP ¡½B÷úB¬ÐWäÇbÓk4B6]¥Ù$Oø‚SΣÃ%wê“Å} vó½ðÝèœG¢òf²1ªŒeG®éx ·Oð"ºï`C”îpað꺷[X7¨qÇ-FpÒD Ÿâè-R 6~ƒcƒ‡ƒS2,Ø£Ú&©{>úÍ>ßÎDk*QŽÆAÀ'nq®„7ÓÜ1$€ceòßÄ~× ZÄWù*)VÇ¿Ì3lràÎy—ßá¸Lÿ-=ÏXnhÀŒµ§Ú Öf†–üÎÆÁšd³:9|ÇrÁû.|(»ù:Ÿ€åŒêTÌüõJÙÌu˜€HÍoî.®ó*cÒ ‚x¯à`¯^|—áiñ^e|õò `Ø|/Œ¹çzŸï´{"½[nÊ5FŸòÍþ‹p£ Ù"MÇl¶Y˜`ì¦ë¢‹“SÍæ‰àqÒÖy\§Œp•‹jPmáf(X(…Êó‹ÖÃîWåMŸ4Ÿ²å]ˆ0£Z]%zç}Áº›#© zÏjDOó~€¸ùâ…E9TÈØ:j+ßzÏ¿¯~Z+€Ýx‡Š•›éq&ój¸RÅËv’U[,§@`#x†ê"<sŸPѵ{óa-\Αë:Q“-DÍÆÒê88¿D_™‡a ^re&#Xñs>à´pÓU¬ÈÏè|û‰æ3zC.ê0þºèKi–ÌèBcùK<aŒr·wÂD4ÑÑn¾qí«‘Ñ žˆžvBì‹Qp![,#üe¤©¾GâÕÀIîÐG¨ò½(Ï£#ŠX4ö“¤S¾„ª¼F_/ö™$:x`ìžx¬àxØø¶+ehŒ/“µ!àüÆP,~—'ë4®YûµÈwÙOÿ2ogA>åjŽÇ?*ô]ƒÕöügLuº`Qºˆ`éå1_7v†ôþ}>Z€·]¦W¢D|ÔÖj´j®›Ĭ=O87\ð¡È0ŠŽn”=s:Ux Õ{8….ܳ1Z‘ÜûËñÓó>Ý&[$ %™‹¹FÐU¤\«F½}à„(dL ôn^’•%y Ò´†¨.XÒ B‘Ò \ß×g'8É)ðÊ…Ó‡OMCiTô§nôàŽ5æ¢rŒ•ÇñM·¨[p¡DÈ A•M 2“f°ŠÀ*·â¸ŸzD~<h©ŽAR†KªH$BJˆ9VY ¹[ÍœÉÙt¥}D™–8m슃Z¸ížÍ:9+o–ÀÊÂmÖ²N @†Æ¸x¡{ðé§t¡zOª¼™®.`””Wµ¹Ã“3¯×Þ;§ãžÄÓïyÇ~BN¨Nï¸a„£l8°g…ò)Ô'}­ Z;Ë–E™hAÕ²­T€Wã¢Ò¤P™ /aæ—õ±¹Áxû±Z ¶ ŒÏ²é„–*~ÌÇpùðA> –FŸöNœ>ï ç5Aa HêJl*‡«¡?¦y‡a p Ê^eïkÀØJ¦³¬ºró‘ÑËLB’Õ¡ ƒº)?å†1¸™u_r8ù _ñðÏ<|ÍÃ/»¡§ZݘמDpvѪO{ª!ÿ9‡ê§»nyx,Œ¾lŸRö艇ýËF`‰Ræ!JÛ #ü×Ô#‡ùÞÔô ³|ÃÃW<ü3_;72 s.eíÇJ¸(8‡ø¯’¹ ‚½p$À"i:(è  ­Ñe†à¾âæ‹Õˆiö€"+ˆv$uÃßAX§zK'éo/+»@N&ªJMÅy¾à÷ÄÖb(Is1'+Æu|Ó-¿\å 1±Z¥$.,Õi‘(îoí$üËdK,N^¦;~_$$$ßóªÆaÞ'ž`@t8ZÞ—­C«ŠZ°¿E^rN_d0謋ÛɆó%ó¡›¹‚‚.ÙàÍW¹zV+†Ö anI‹4Ý[ö}"'£ûîk²œ ª /xø¡ûôXbÛ·ò…°¸þÅ]¶¶?Ìf?¾¸c÷y\¼(n¦ëíäy‡¹5½ùk£™>:­Æ W?qÑZ¤åâ3Ð[ºÙg DC æßÄà‹Ìë'| zZ•jùp?å¬tœQÐ…P ây1*¨2Z+£ó™8”â)áçdè‚ ˜cD €`€»81amµ»­…)ÀóOÑP à!ºq&Á×$àáK/*¥ÆÌÀscê‹7¹Áœt‘FèZ!YeÌy¯Í(ß°2ilIæ¡ÎûJ$tD¬,R;û»œÉŸ]l Â0/‘Ÿ³?“¡0Û5øÃk4åö&Œ3,pÔÔÓ ¦Ë¥W|òSYº$:èzÊÈû݈{ê“¿ ÊBПƒBÎy(*y·<¼ïNØ•’‡U§ È…}îWu„…ïk°‰‡a˪OCDÏŠëüÒ;ÔªBHmº.¦OYWªO Nø “Aòj%“/^ÞCt3Ó=ëlýj%RÃ2ð™²7§ 4fBa  m•£Î8L¹® Q–й|ø¢/€}xø« Ðßä§s¾½k…ß2•R¡]Àx“mžžRƒ‚¤8¬`×ï–òÐe~Ë©ÖÑÄÉ¤Þæ3xëÖá5 («RÈ]kÕú]x €œ^»lÞ+L»úžŠTJû:Q£K:F¨³ßËð\¤JoVÒˆbî÷%No£9?ÃñD@qðÀåUÎY¿y>²sàe€¤²ß©(ûW]#ò'~½®ÄÚÎ ¾hrb%ù&")uB¯‚Î 34=ú:)–­}Œ–RöcPªz’Œ^ž{‘ŠàM¤ïoó\û©9K!c2Œng’h]èl…{`sCÚÐd èL¸È± Äâ Qîõ%ËEz›ÞYyˆ…ˆG0¥>©X‰õ:e(¯ –ÕBâ÷wå܉KÑѬyVàéLt€ä‰k¬àLFYÅée÷³g+K‘tz†„@õÍ‚¢& :²! 2ŠˆI©ï¥Œ¸5´Á6^íð_"$z £yôÜuÊ%µOÞlgÔ4½l3”óµ>± Z(Ï>‹ƒ3Iç.1™ºw —:“Ƭð¾ºdyó«5c—Ÿ²Å”] ÒúîW*ö°\ûÄš‹hä«\£a¸AHaµ4´êÿÊ{o¤m5I¾å¼ÂoL/òÙ™V¨e5¹òô–ý¾Xä)¯‘TjÒ'ÏIиw-Öéïlí‡m?Œ2Xrp¼Þ‰0jªóÝŒ­Ž*Ѻ.1èW…õ C‚G‹*“nNBR' ¸:©ZzþPX¦| *¼1~óÄõž""ÎŽ”1í)ò5ϵ,–‹!ž¨J‹¥¿c¶MïXùÜÚæ –å•MDE·TlôòÈK-ÊUÁ›0Žæ<Ÿ[ƒp¢¨2çñF­ Ø—´ªK&f¥.ª°}m ”d©T´Òfÿ–=#» YÆ&…£8–?¨¬8lEë7–V’'VuÍ%ñ¯tLEtÀgÒKÉ …? EäÑäé L8å‘`œ}þRfÀ£¨ÞÙp<‹„íéSˈpŽì”éj×’¾-ˆ§ ‡ššÔ æÒöîCR~¢ÄŽ©ÊÿY꣙1!"¼™¥P<üYzà6!JÀ³Ï±n +‹2íp‘Ï0jWY²‰A÷DhIFhšbŒ6C÷—£1pÆe…Ñ!ÂÑ›L¾H¾rÎp¨BÝ"v˜ï'•åÄÌ•^ª bµ¡–$‘­ê€ìßÓTÅàFr.@WêU«Š"Ýs|È“±i«‘ݼFÙ)ÐVzœJB¥c¿/™ö+®è"¯1ê±AT`±w˜J¨¥5¬2¡\HU˜Ï‰üÅ1Z!”ußbåÕˆþž'T¿o ÏÍ dF9Yõƒ'N1ÂTõk&î 0ºdïë™kgá9/ëÀúHCËM•Å,L–ÌEöWìe¯ ü è‘ô¨js;Ä3/6tðÍ¿®óeż…æÍ/€êŸÐãr,W+ÌÍþJ7¹´eµ&X+« •\©¦uZÍ-í)/—sCu†J|žRˆøÓ 1¥ÜW2Üò“¸Õèç•@˜RèÑ—1¦hœ‘‘‘°à õϤÇëÎÒý>Kw-2é9‘‹ý”ÎQüjcþ¢„³³8~eôG7î8ðø¡É7ßo9ß¼Eëø€“Jfu‰êwu­ùãúlöä»ãpÿ2{°¢ù#x,0ƒÆöƒ´fhªL‚H@ 1ûˆÆšžTx=ŒËÿe£/È]q‘°ÿ³—ëŠaÜ8§ û2>ìüqÝ'¿±^m“5»c}[›…-ëébFúìm%ÃsÕp1c›6z\ "?cÚFؽn&Ztû·•sZn:4I|üФ>qÐãu‘/í7?óׯ˜ò‹`Ib]1×åW[O¯­£" }k> stream xœ½\Ýo$·‘òG‹}èÉYíæ7iÄ@îŒ8¸Ãú!{2΀ÆÒ¬Vñh¤HÚ/wûUÙ¬bG»;ñ~p-UÍ&ëóWEöüýtÕé„ÿ•ÿ_Üœ|ùÒèÓ«‡“éôêäï'ŠþzZþwqsúoçÀaŒŒiJêôüÕI~TjoÆàNƒ c2îôüæäÇáÛU4£žtîWÓh&ï‡ëÕÙ4†“60æ'o£.qÔ¥‚&V­ÖÀ ”šŒÃÅJ{`pqx½:3F­†|̤hœ'f•&­ýð¤‹ÞXÛ>7¿â¯çÿ rÚÅQ;ØÉùåÉ Ìêüo':?ùˉS<}òøóIv«NƒQz„Þð€õ£Ö§Û“ÿ®Îâä€Ã´™Âp¶: ÞŽÓ¤†WÌú†çÚu_öÈ ×eÔ¤á–Ix,t~VgÊÂøäéølè–HØšy%o˜|ü[]ãĆ&þH§a^=Œ«3Ÿ`0¤á™„QX X­¾Þ¼bt°y•×üŠ5^ÚQ‡wËûü:‹2ÅáœÄ ‚òÃË"iðºïyôO¸0PE ÃO«ú 9ØlŠlP[a›h‡!µ¶Y™k-\‡Œ³5¶bœÉB´røˆJ‘qþ£z}YƒOÎŒ>‰E•ÈE1Ór™Ÿ½(¥1ài˜2ŽÉùDÁP"ꤹ!2%Ov]ß2¹®ߢݣù™b÷“‚(Іï&µÙV•&-Ô;´Ù@¦ƒö;[Ô3ƨ ô³ˆÞäÜðȀ晆˜üŽÉo²!¢à Œ&›,¶óŽkÓš§°ÉÂðšG™]fÁm9ç]¼dò{"U‘Q‚Á0¬li“>;*a“³ªõW¦¥A|¶ú5$ L08eõ¿â0°=H4{уÈýè1‹ÆëwóFdpäð3yÏä “o˜Ü2 óje)¶¢V!¹_²·_3ùÀáె \‰ÁìÈ:¬"³»)¶l}1k˜<¢­Î£·LÞ3à Ï@¦†oÓ=…X¯þ<+¯QheZªø…"Â)ÁâmE;–Ìš%sã›CBÒ¸ERÌlÉXµ7hõèæ¤O ÷grà  D@eø6 ³—þ̾»fòÉ “¨i¥H̘!q^CQ'áÛlΦeôÉG&7LÞ3yÍ3ìxòŽ<úûn'‹x‡H-Êl\™–*?âx¿S¢$â;ls2ð÷‚pp¿d¹†¤PÈG"Aº m]3¹eòL^.'£^ñè5“»î‹¿)¤QapMV6Þ-ˆÜ1¹áÇÄ(æ(•]ñlÝ|UàeREnoò‚#œˆø àUvèÿ-\^fÜ-=È{vºuIª¯ýcß®,¸Ž•£ØPcW•iiiŸ`7:Y˜@j®¦¶œèok¢¿âÁL^׿ÿr,ð¾"Šp/V½ÔûIŠ˜±Â{^_°˜?0*ÈÒ×Á;Ò7&2Bú$éØJ¿2µúˆGH?`Ué`Ê4B:&áÏYÖYòÎBŠ’`Ç£·L¶(_yE0ê+Žà_ñèe·"xˤPÊÁB‹ŽÀÍÁúC+þœ[nØ 6]?ÙåœNm)‘?!R„ZhÍä};Z²ós½`rÃäŽÉG~lÓÌ« €U¥z™\?QÊÎêoL¢2-äó3yð#Z„PR™ñšÉ+6û›&¡RyÊõŸœæÑçL^0Ã-‚‘h=U€í'‚xÛ‚ì=ñÎä1ìUn4åKžò“+7¯”£ôÎ"öF•i©œ#b#¬Zá”#åÜÿYnØçÌ,ÖTÞ/×ÄLËU~>°£™èA¨5€©ÕÑëuðÔ)ê Ü-#…’ ^2yÍäE"ÑýÇ2´sŒ–9V„öÂp°¹¡7娹ß3Þ~:…º„™ÌâoTR™–J:.ˆ¨Ñ¸T£ø~­,jiQks }ÏÕEnë(‚ËŸ ùœ#Ü'ECÁÀÕGÃÙswÝÀ!âÂ;Ã?ù‰êëëö}›«—X.¸ˆIóf©÷FÍ•i©øÏÌ€Üõ„S¦ÑNv¯ÄêÇËmSbEém»jÒÐ~ñyÅä#3¼^šG üsGå a2µ¶Ô£D>g† ÒŽG» ÷øâ”½³ÑOM;¢ºûÀä=g«[}·—­öº{"GpŸÎΛ8z-¬‚, ´VQ™Z; ÇççÃhm:$¼e bQeÍ¢*Ór™Gzñ½X&°©~¬&ªþ½›[p‰ÔD-82‡¸B]Tê¶0š '‹Í³ »¥Q¹c†G¶¯g­yäe^²Å…¹—r^ŒÞt»‡{e†~Û×ûf ìtÒŠâeV¦¥zÎøð6L›3~¯öO£¦š3þ–‹°ëš¯xðqö£X“ úÔÅÇbÓc—á>G·¹S_ÈÏÍ\ÇïÃôÙ«Åñ)•Ì5áD\ú–qiißuOϻ«À/o~y:ƒiH%I†…b5%U¦¥mTƒ‡{&‡…ËR!&*Ðs;ÛpûV“¹ÛiߊÑçL^2ùªmßî'AÑk½b“zÁ£×<úKkh ÇÃÅG!_0yÍä/¸ª‰ðìÜÜœqr[în˜µóuk eô±=Bû¼y+,.£í¼µäîä=¥˜•Q©ØMcK•ii]ŸoKð|„x…QÍëx ïMv£4ð²€fQ•i¹ÌãNµ<øFL=Õ¢oÆØXL6 õ=ÊèšÉ-“dò“×<•Fhª½ÖW9™•û«÷ g(µh}A} õß²#SFß-ú4sg¬µG좚+#Ùý’Ë&˜Í(¡hRªoÍL­ê½¼8`ã¤FˆQ7Ë©›™*Ór«…}O&3èn= ‰ÅL5»Þ±žç³•P y0¤qŽE/*k3“2T¸ÓézQÑ;nâ=òè}77€{Ç'<*Žz„ zç—üÈÃíûšƒÚDÓíá"w°â…îAZs UDq0‹ãa1f«g·±â»7æ\ÑöTìTûÊͬ=;ü~Q`“o^°Ç¶¥§2( ¶ÀþÐï£üKŽ{­˜•Þµí—íDmºbÿ7k\ó¨äøRrìéËñ2Ý4úªLK ʸb!w*©ùòT3SeZÎ}¼æ¡¼Š)dÍ_w±VFߎZ=óí@³3ºp¡¢O½bò–D&ioÏH z¸cf†ÚBй¢NeZŠë( j'œ = çŽw¶æÝ·£…|Î 1x¾eªc¹~-¡ÚþqL~µ°óBösEw­†J*ÿ'* <žt·;°ÐjyFUØÞÂ+­–¹ƒ7'R5wËu·LÅKxÚ̉ôS¢¡ÖT¦j¶áe|º×¬Ü\lÃàžÜ¦hG+®^’˜¹UŽF’G7¬‚h`tH›„›«)/— ª<Ë»LŠ0d¡FÿFW ŒªW z§ÿDþ?^%PÎB“éíBt¥ÆDwì|k&ï›Q¼õ…†,€Ù%3Ô< óm²Ã|Ëò#¾n"êMØl±ˆÆH*ÏÂlŽ.MŒ#vúv´“×€ç×7Kšy–k<úR£ ~TþŸ»Ô¸'\?y¼¶}³ØZ³“ʳØÛÑÁÔ€nt¹3jzU ìØ:UNù:òwAFày‰¶¹ALð:/nÏ#ò9s•ä’i¾A.˜Ê ò†k–J]U+¹§£g{‹~–T„—=ÎÅþ·x‘œ)…|§J»‡«7x%ßC-‘Tù@ ÆÜÖàz÷Õê SWíG óÌ>¨¡„Šå²?Z1¸ˆž¾-pàCŠrÂ0l0N˜Å‡ø ÄÄX¾”°ÑâÅ^\µ×. W_!TI*ÆüÍX©üæA¹Ñqñ~¢4±r­¤½`rËä&˜|\õ¾²°qLzªoxÞOõÄ·d6À†LþÀä “Ln™|Ãä“}tj´ÖעcÝ5‰T÷+°Ð3{™°9€LÿŽL $•] ú$²T$ñP~‡RmÓÍת B̉CÒ…g³r(Sv<òênTƒ'lä0ºŸMÌ!¹…~t êÖ½¦œ }û\À*Û ¯g)% ŬXú"õ¬FMÑBk5QéP¬†ÈLþÀä“·LÞ0y×eØ1¹éŽ>2Ù·Fì7/YDàÛ-ç¢ ®]Êà ØÁcCRÁÏ ãµfð4úPj6( ªÅˆ(ã’CŒ1 ±¦PÚÌö©2u?Fðî¼yÊ¢ºød#ÚØo1§xißA(€·œˆÛÖP¨3Ƈü2‹³…mÐãå­…Ö4)0.†Èÿòí5<²AÆÁnù2PFç àf3¼»;…,`)ÝSŒ4Ž*>Mô\NÃÙCÞF²ÉÒ„JÂ0>e@vwø$ ,ZF²*¹æ}oWFPª¤òC÷™Cžïjí¤òâäü÷G`X ýp†ÕŽ §÷‡¿¯íCj¼š¨u:Õ¢“ãjeñÔ÷ºæÃD缜ïÇ‚·œÖN~/ F÷n…ò¬bÐi,U.Šb1à/xªµÂ10‚­—xMƾ ¦6yB ["P~oÈ€HËðÔ°jP>÷1¾i€ueTJreŸñ\IëPsÞ^†ù7¼¿_ããW à3‘¼1êQ!úÞäU0òïÛÌ*´E^ÉNV2ï„h.X4 o¯)ˆU–³í­x¬›Sܘ J´°ŸÞvc¿È÷ÝÇ~½ê¦‡3ØçèœY æb1âá~ì.݂ͨ—¹ç2#ÇÚuþ¦*ƒá—•s W¼ótµðø/_Â[$(AX©9¢#ÄIZÑ^q†|•·šZÖp´mnéƒøŒH„wH€Í¹zºpÜ‹VƒµÌŠyÜipB‰aÑv¤/àXg+ `¦{µÍ%;ÖBFcL4b›Â°!x<…Å·þoYª´uŒöZ"ļŒdèVØ&¿»fgY(¥ ýLVW¢7®Åúµ]2”_ÐÖÛÆ3d`k1p¡Ák­I€Su®-òƒ»":pJ©˜<ˆ!IÏßä5c‹ÃåçJƒÄ#^‘ —eí†eQëà_~r%.âÝŒ„b©„ ä±VÓiÀn?Jq)Àk/àQâoᎬE¾ùXYb!$[Ã?ñ³iø›ÈØi#Bß”Y¸IJJî±Icþ/ŸyxªS y9v¹¥E»¤’o™\3¹í_T§† ÂúNbT‡.0‹úÉiLô¸YHNòïyÝjM ùÔšþ-YBñá)LqÑð@3±SäOúBv@™MB¬¬»ü8þȇˆ¼Ž7Ý7aè¡­$s° 3ó.aFñ‰}œAm¡eáe'tÊK“9l$·[Á÷±bëLAéâÍ23e÷ðÖÎîAä &wi”["V)²·#‘Á%Á9(Ö£,ñ{„"ÙÕžø¯ßƒ¡–K‹NÂ(xÍF?¥sÔ‡GÒ¿ áB.~%AñŒ-Û¼üY5ÿç7çÝü–èÒá-™ÚCOÌ IJÉä0vnR”0q}]½v×,nÖUj"¯]_JÕUîkáS}Ìp+t.AMééÈ´6»–ËFò 2VŽ+fDwûˆ…$¤På;QÅçWáÁQsÌ€?F0Žjé+¯›Äˆ ص’n³FØZÓpe!ìDó…v>Y¥‚õ¥€æÀM¢œ\”¡¹¯éŸy†]Ë%ÈÃÅ){Økj¸“•6 àÁ¨Ìð¯[LÛ<ÔØm4Þo –PœØîC‘`RòÀ )`3œ¬Ó²¸»œÅ­²V#U¼Øó ¾OEX9ø3Tï@~ÑÖ1Än¼D…@§&ˆ|!¶ȗgž7åŽ\Qi{xЭL…>îóc¨Ýß æüóYÔïl#Y-(ÕOÌûX » ¸­Q^I“ÈöWÝ”+J×[yi®Ctø¥Ü§ha†3² *+‚„RFï•4A+H4Ý èM+Íðš ™.&õðQ®|ȳFJ=HZí5â«7vsù^®]‡ÎBÆ[â5·«î ×Úß;Žå ¢ ôóbA+MŽ“{~,bIM28Ø!™uõázuö[.~“ËŸÿ  g ‡´£°ÚM²ï2¶ùrçŽêç¦åTßü^¤—ïVu“Â×G>xL£^xÏò¨,·J¼—Øo]‘‰^>†‚ÉB¸qMº@>x™¹1qJnÊaxCfÒ¢¼3 5B F›Æ9"j#Ÿ¼ÀtæY9bêtªîŒ£>Pè7án²tUöN¦Áê‡jj Ex ¡:¶³ñ}äWã± X=VÚýR¥ ø’ ´é±™ÌSql2FðŸsDÌ·¿›æû¢ VˆÀƒlÕì­‹C÷LŠ–×µdèâ|3È‹wv·\2ÔI Ûë\€äð0¼ÅpÝðÛm•,[‡gÆ8Ð~Q-:žšœ2[®WÔBž6s‰lt‹£ÒÆc”lMð‡ÐŸe‹¥ór1³„й܆õ¶óå—;n¼ˆ²‡mÈZKmÁ0`Oçø6yy… Õ»<Š‚‘#!¹ò±ƒzã©¥ð¸zM†?)'¿èŒÓ«•—AöŬˆe¯¬Ú¡˜N¸*:A˜ ‘ìu%+ !6¼mÏǘxQpÝ:uÔÏý§ÙÜzG’Âﮘ|Íäc÷1ÑîþšÉs&_2ù=“êº6ÞàRº6Òð"Ê7&sè.I9àt£·‹ÌuÅEZë9ÆäSž|ƒï1ä6É-ý¬> stream xœí\[¹qγà1اsM§y'¬Ä± 6dÀÀ&gG£‘¼s‘g¤õNòÛSv³Š‡}fF¶óèAT‹‡—b]¿*òOgódÎfüSÿ¾¸yõ¿söìêáÕ|võêO¯ ýïYýëâæì_Þ`_¦2söæÝ+þ©9Ëæ,…4ÎÞܼÚ¿óGèì‹êìídç?xóöÕ·»_îgø§ñ&î®±ifóîr>O®dâîGølJ1)ïþ{”’sÜ}À©ä¨CýÝO÷çÞ»)Ëa?´ÿ¾Å…RR²ôÙd!ï>asžÓL;ç&oô\ë4´‚—¿ú,ƽÇÙ[;ënnÓ4kvoá+,І¢öKd˜C.v÷À#8Xãxÿõæß€ÀÁ('3A·…À¿[ ÅõÑ;ã¦ÂÙ¹ SŽ‘s¸Å¥Í>à"Ï}È0]Üýyoã”g“–åŸ/Œ—{õ~{ó³KDœ¹øŒ›ÀÁæP<¯ G¹‚h§­áåÂÄ2z‘aJ¦mhÚŸg;OÙx8åìð˜Ãîîž×³ÛÝà:c & '¬#˜ë.¨ÀÉ¥z´s6n÷Ã>:ÚÝŽ°8S¹¤þì@{ÊÉæ™ÎÖ‡‚Ç·ûx„Ùwšp¶ÙYÞv=RA‚:¢s\ȼtÇõ»b$c½ŒuÉ=<ìï|e§ILbÔZ{“Ýðš9%5į¾Çã³³u[ü(ú¾oã –VÜ]'©Ñ$Yh£Æä’õÚ±Çì]´Lsc¢KFYIñÀ;q@΋öõŽÏ-€ÀÝã±dë‘nÈvÎ{4.L$—ñ Ôšg·»“â- 2{ë3Žpî °jQA‘ñÅòHõ´ŒÏ¸{µà\nö8?Œé•°À‰S—g[O?Ì g¯Ÿ÷“c€Åà!BÓ#ozäó…5½…ž’5å²q˜8C—K¢xFDzt:nU­¯q€H²¶Í•8…+x´ð5G ´ìÌk‹3.5žË6%cëî*+lÎ0Ú;1 ‹‰ÏÑNã '–á®Í=R=6ÅÉøÕáp¨²2²ôÚü¡5§Öüsk~jÍÿ܉ö~?R¸¾LqUwßVúçì´Ícõ‰L£$X²õ-ÑØ0“™À3Èiwh,8–q- õË‹½_h î,J<{–¬e‰ÝzÛVyE?w,Ê\ª¸ä̾¬ªpvž&šA0„±J.yΛ\+Eê¥,^†ã~ÇíP•h5@ï›¸Äæ'k–`}Þ²fº5Øe“Ï·u¶ì5ýXlaùO›!,¶ØZ¤Qæ‡!Áçål}©û‘ç¡L˺Π…Xzƒ %?âBqŒ:ò1XäHV—fœáÔ‹ü0´#·ËáHÇçP±7Îí`Ú½†ÜýÇ:ÖgÖÞ8·Rë!ù™W`}ôʈ|¤éÀô‹¹lZX–›ßxèÇXí!|7Èd%³ÌóuÚDTŽ™b–}g½Sʎ̘3àÇ:%hôf‡Œº°q¬0ÏÂÝ‘3Ž­¼¾²ï˜‡nÛW6á=¦? VO0,ÊØ±ü.”úÎò §Dds=íÐë}ç1Ú²§üBßG+½IÁ!ÇÞZuÑ­Þ1h6ñ™ô•Ch´öÿ¬ëÞϪ¤ÍÁƒF2@y@èºôþ»f†GÊȆ¥³÷ðäR šá¼Èà×ã°Ø‡0¯‡o™ž¨`? q<⃹øvÞÂÛŒÖi52bŠ X¹ ¹à¤ƒøÐb·¾¶U0Ž 2º. Ϻmêà¯õÛÆÛtþªrkVì§®qžA¢ØûØb¢÷mÙÜqS˜É;éÄ®±Z ò\±{9c&­ø„ۙâû\a'0IKçh+Š—9Ü€ÎÒTÒ¬„ Œ:ì-¯\ˆtŸMš;?%g"¨Uõ£§„AÉå¶0x'd©ï´G°z­–P—ÃÏSnˆšƒßQ¸W{£­[- ôC8bƒâ?-°4Ñ>5ò»®ƒ{ô÷o¹ÁA=¼êÇ SÌ'kc7S…¸øÞÍgöœÁúû­¸ó’I #i@×qšE`MªÌ‹¾$v…ð`[g(Cl<ô2¥ó©¾.šeí_Ÿ;àF"D3P’”ccmŒi’DˆxAÙÇ:C¶ÊE~ßBó-s«cRayiÀ˜ýv‹ûžsj¿†5‚%z•Š#—ƒE ´ïêH•Xô ïJVkŸ[DÕËb(Â…sdÈ–õ¡™wßÔˆA­ ‰ˆ,e˜}r@a¤³Sn€Jþ2˶ gƒý&IŠ'8[Œñå¦_àM@L$ ^wBìFnúù²m±@)yhÍ+jc,85o[óК׭ùuk¾iÍßµæï[óCì+•in:ZkJs•IyØ+ÝY{(ˆ–ÏÀôGî’MÜF9¼ó„’)g?[KŽA-žà¤f•Š“")ßÕ›§ç— õª{wm .Ó¦<×ýØVO¦Œuñ–%mÄx4Ý8HþïÜø…”ÏSó•9´–'(¦5Š1nð+ø±vUNs÷œ6K07ž Š ¦!ž‘ä”/°¢~›ÑQ%}§ü`J¤PÁ40³ÍÛN©°1ÑgJ¾ -‘´ ̱9•¯îx`ðu†Ž\ÊSŽîËâ¤i5vls Ép6 QéªgO¨q/b{½ç„¾Êùš€`Àß!‡  Âv`= 2ÇÎbãWøÇ©D›éb ÏŒoiHˆa¶¦q:Ã'›Ìq2p˜ê{­ÈØ«o“àGa ßÞí×ÔÁuk^¶æ­yÓš†ž‘ˆe¬[§×ò=ºX%Î ˆN ;˜)‡Hór>Q††Ây|qƆ™@ ÄA€=B{J­pÝœáz:ÄÍm¹ãW¼…Ùd¥ (Ž[ÌAù±WçBÇŠE§Qgdo¥QnJN(¨¡ÔåQ%rn­Ÿ Úì%é9Od»SH<ñ?"/ÜÝ\”*b‹±˜G)•`\÷LdS!ßvéÎSE¤¤á*M;ù"J8»¥7f,±:`¶”££XÆ)2‹ ébO^ s`’É•P >³ZC}_?à…*ʤpî ‚Éíòx`˜3>ԯ愱üì˜-ë„oÛ„pö’þ Ùô®çÈ‘QSY éD¨†3™„fXb ª,¡ìN>Ž!•€O_cÁ‚+¶ªk,:‘SΩqr³3+ˆ†säÓphpn0Ö±s¦kä]+a:ÊÁwÆLÞ°ø@ò¨0füèeDwY~!¬Ü¯‹\áz׿{’…b@T­KzÛ1Q TÒ¡k^ª½[@Öjhx×~âZÃ'½ %“ë×ñ$G™æ#ýáÒTZ”÷„þ(“I+,ûœ$¦*(èfΓiÊ'dœ¬tº^³l¥’:“…à«© øœ³Î^ 蔢ك$'a&/yLõqõ‘ÇH%œ€€q’ìÔh}*@Û8ô¬QÉRC•ǹȌK?ôÀìþ¯êÍéH»¤wöÐXq«”f Ëm-(¬Pa Vì·z(`pÊ}ˆHéÅ•ƒ#1é=Í凤ØðÝ*$…¨ÏÐÿ,Àßqeý¯šÇøËÖü¦5ÿКãJá•~v¸úµâ«(¥ùjè´f,§léÀÃÓÃ'“ÊB ÙùˆÖàDì˜&S[¢RŒBÜ®†b,Ö÷ý‹Ö€NA€§§r¶]„Y¨Ë,uê~¨ÞW{ñ¥„ X¹¥ÔEUEj&G† ÑãB OlL+rÞÕ{ô`w‚«`ˆŸ,c~}z<<áìâ5Ö·’Ö—H:Ys̓+M«à:ï tH]*ëS¯4d‡Hv2G‘(á{F!>„n.B}hFR5Œk“rUïXMòtÉ]¬ Í0fè´–Ró‰Ö'káÆñx É!¾nNÁù«-¢ñºbI”K+ÍA}±øISgmE:¯¿‚㣠QùeÀþ;eWKùÀ#£õÜ,Ò¦Fô•­‘µŒÜÛ•ÿÒçhl—¼âm,$X¥t£4`»äùV–€ñQ!ON˜÷!ðs'ùöê}ƒû¹,¨.Sº¦ÊÕy½šŠÄäÄ òNð‰ Ç 4ÞÌü­bsÃá”qå@¤â7Y;Þû;*å€9X«‘Òjþ[U€wÚÿˆÈ—Ï»R‚y¿h×Ò0áBIilzdk½ÑñÜT¬uÒØžžÙö[_ÌYŠ?4oáb tITÝ2çÞ"¤ÎôX~}´Xy¿öýØšÄ} BÜ6×z B»\ŒoÓj-W™Û@,»Ë È ï]­ žy, zÿÓá´Š–i„Ò׳x“§®X¬üŸQ¥óÝ~£ÂVÂâ87V ôEX¿xóê·¯øš‡ð lÊýK/žyÐbÆÇ³¼—¼ÅÛgßr¸- Ûƒ¥_]M8å&ýjÑÂfµpõ(D…€’¦ìà"}q*RÕïRyyì ”5™SÎØ{ Q漄y)1L@+“Ï€˜„Ö1¸Â¦aýc)ˆn)æà «j/y•I«t}Í%K•.²` %#@ؘ—ჺ]íK¥dºÚa‰¯nÀ¹L/·Üšh@²‚FÅøY]¸ŽA¿j~î/[ó›ÖüCSé{?w£ ¹§”wÀÍŠOËîB¾8ùg7××\ìŸO¸« Mëët4¾¼Wì6ÛuǺ‚\”ÞW9}±m™¯z'‘˶ñzÓVëŠ" ¥²Ñ£»fÆëÔ¥Öâ§špmsÙ –†a2ûgD¤R€–eh…D%¥âe+ò•eÙ‡Ã:ØÒ:¶²º\@^>Ò<©o|ͨ&>à,RCSg éŒOx­ô„‚9‘³y/z1Gé¿ò)$~•rPU7Ò Ìs寡}Éý$PKÍ5í˜!ŠProªoñ]3Ì‚9è×yäC0\)4ëâµYm/yÆnûÛJb»"ÈG"?Èùäe(g‰C7 VînxÍö”խ𪗆Yç§µ'yÆ:ëູ¨«DDPwÛîíjO¶Õ–m^¯^×Ñ% @Gº3c¦‚ÜOÎ[ݹ3“‰¡:oÃÄpâ<&)ÌÎâ2fã€ÿásÀçfæš5 ±!Õm[|Aoc¼ ¬Íëpétà§Î."‚O½8zx¹”ɃˆüžÔ¬—JÎBTß0 sð‚‘+¬Æ S¨,ÃcíÜ%–º§)!ú©—ãêŒ?ì±êt^C0žûž{X7¾fc4!’øæÕ›¿ÿ–ÞXÁÌ)½MÁW”a¼õoÑŠk|hb‘–Â×5Ûc‚¿Ù¿$iùÝûÐIqùiE"EÙëqB=Í„$¬˜ÉÕb}3Ô^Õð)€L:±´þrˆQŸ‚¨»µù|å¹€ñë3/xùcþ¯ýøkµW/}…§ú3Ýeâj?%‘îD®M÷ÝêÉ=®#T†ÌúÑä)áËUòû/ ü ŒŠuÓL7¼×@<&óiøU‚Ê_µæoZS "ôøÜšÃj`±rò©‹ˆÅ,_ãîŸÚ×óöõo–«zÙ–ëzòîZÓìz½~ý¾5¿nM?ì+.ñ 4]Te‰{:×ÃqÅ6_/äkósk>´æöMŸ>~¬î©C~RòfØ zßšãƒSÜsœêÐ]äe©?oKo4]÷Mñ§#†~7ü•ÈY); >¿RãfHÏkIÄUh¾nÍg²ye…ÿgsÅæô¾0M/5,±ïÇáù|Ø8Àõ$†}ÿ[*µµïÄ#ˆ=ü¬5çF1U<3Úúy(?«»U»Ö¼iM±Öi¨Øçk3´fÙŸCü`ð‰„Ò¾Zð?æ&ˆ"7ÆÖ mãsû:É£f¦wŠá%»´OíÒ¶fÔ»œ=zT’ ^ul8µ¦jÃy¸áø—nØ=µa'7±6SÛ° ˆQã.öÃMä§6,~æú +sõ…;÷/`è$y{ÄÐñ©óµ CÓ]–ß¾ú_üð¨”endstream endobj 183 0 obj << /Filter /FlateDecode /Length 6791 >> stream xœ­]KsÇ‘¾sõ#»Ïl£z?|£,Éaå•%z}7b‡rIbh‚I;b÷¯o>ª«2{ºPvèÀFOuUVV>¾ÌÊ*ýõÄìì‰ÁÿÚ¿go|ù£w'—×ÌÉ僿>°ôëIûçìÍÉWO°E†7»jª=yòüjOŠ=É1ïª'OÞ<ØØ´}ò?Ð8TÕ8¸3>xrþàçÍ·[Ú`Óæ5>ZãSÙ\lOÍÎ×âcÚ|„×¶V›Ëæ{lQk)ió[äZj¤í»ßnOCð»"»}9~¾Âb­9;zm ôP6ïñјlh`ïý.X=VïÆ‚§äW7¢ßwØ¢ç–û:lO]ÞëìæÞ.V5_bƒ‰¥ºÍ5÷àÆåYüדï€ÁÑ*g»ƒfƒìŒâÆÞŸØ°ó!9l|j‹ÝÁú¸+)ñ'? µ$ r­ußütóˆˆÆxç7oˆ~ÞÃk«^a‰Þ‰÷Ÿ€üvÈç&ÙÀ’l®Îù}…µ; jt˜ø^|úzÛ)Žy¤±8ÙúO X–Ú¼”HžZïv¦Ä“Sçw&âdqfOð+gŒ«yó:ó%TG«L†¿êæ--JöÓT€ä0q?”xâm1÷¼Ø ô'LÚ íò°Í= S|ÎÏÉú&–4ü±´×㪴 ¤Þr òí5¿ Å®ËâÇ¡$JÆ{©×[¥­ãgË*#¨ßoQ¶a"À‚j+ù(sè­’†Cö—g¡Øü ûµ…uª+JoVzîËZÚ²ÒÛýÙÖ¥],.€ÅHê¼Ù_¶ÆVLPt¾Eéò¡ìÒ°YÈ@XMX¦Í´Ð¹ÚÍs¨LØÞµ5¯ž ‚7¥xÐ"øâ„zììÃü‚ Á±k–ât° >(=ØÅP ~àñC´Š¿Ôg0&ÂÂ]áHÌÒS˜Ì®‚Ô¢ãs2(1]?¤Ø~àÎê•{9Ì%Éb1ƤªM§Ð7dT"z|¥×´8©&ƒ<•–s‘K¯•$u6 ƒ°D‘õ¿©ë¿KÚú200ÈH3Å™zXp©=~ +¤,5HZ° ÚÅôјÛÖÄÁn'ÛR¯e‡'¬MÙP® ¯‡z‚Œ%½˜(ö÷S¡`]ømš„_LVÚ›à“k=[4K+~\¶fdŠ›±ÈÔ²æ`'ø²q¨ªžd‚Áìû GJ”“D&ÿº1Æ@{`² ÛU_Ÿq!{µî¨é“1e±9ûÍߨqIÑ+}@—¬’]×…$Àj‚"/ ‰è=I(v‡ÆIvw  ýµ‰#½œ èÃW”=›q)^±‘q~ÎH@^!Œ>5‡I"$EÌ;#ÊAnHÕ²MI)æ{^GT¶%$Úì¦ažNм{K¼Õܹ¾›!NÒÚH—Âs8Àfö7ä/|´d°È5ô<8¸?›&¢®×Ãf(ð×i~¹,ÉÓ(° 3eÆ×ðÇæR¯ˆ=òDÁœ ‹,y6ÎM^ŽuÿÛVà„AÈù$_k@Vzv‰ö:©A;¡f‡ ÀÝ#¨ë¯²ÙcµjØ6 $¦E `þÉHx uŒ }ƒ¾B{ ÐETˆé°¥)¤€X®×#Ð$µá…r6äZ 厔uü7Î&Ìt i«%Ja׎W€ÛGøcŠHül‚À’io L:½Þ Cx‘e´Ù¥NӢŞIÉ@‹»ÉátLV£ ØKö\7bA¤–!Þåì5ã¦UßÓ—üpÅJãÂÎŬœõ¯[t_ÉC¼›KmâÔš‚i ?K;ÿrˆ­ Ð,U»Ã°æîhX¬»0þUCñ—äêÙºÃ[ À™ÐŠaklÜô1$«L(e~‘x ;@ÃökÀ©'Œê5PaDW•ùB sÆSw¢P¿¦=ÒšM ^‡•Ì™äŽ\~WžëÖ·ñ:ï²w>ÌûÜÐwÝ !¬iKXÜNvÞôÖ}ØØ;Xi1b‹ýK9n÷ד`Æ P„ÆƒÖØZÑ›N6èàƒ'¡±Ï<»”j™9\ ¨¿—VVȺ\”f_…`#©qäÙ5@c”””K½kùœª) Íßçd¢jŽ9 pežÁê À#âšïg 4ƒ|dŠ ™Ç4®øu2³™ÒÛ¨38]IxB'­®À¢mm@ŒÉü~1¼olµRNUäÅ6q½h+”ë—ìMYÁHR9†ÍÀ?N#"{*ÐÃá78Ì8)Ã-l–È þý±4õ‚GÄ1>;Ž;Sˆe­·_‹wZTNÒ§áÁì̲ÀÃJ(òb>õêBBKOÆ=£)¢ï’›Á˜œSÓÑh ` eóèì{|]@ϲ ò/)µàÁîÔ¸²†’¤7ܹӚxÍoƒÅÜ_ |PêIÁÏ>t‡Ýq³6>ãÆ~‘@ ¼q)a¸Í“-´˜’W¤Mó.^.W|±q¹¹ÙYÀzÒž/wLYnxŽ0ÿ yl™ñª y~ÙÂb\Š_UÆoY72@Z×±ÍçÐ*£ŽòFŒŸ‘IÊ%¬ï¾œN4È$Ó²™o›® ‚΄=²VÔBÙY[˜ýb4xÏoKÑÌPÄ3‹N[sˆ'Ë1e‹Å QäµLÞøavÛWÞfŠd±³ " ÑÇU‡×ËšC‰Ðäžé±¥ü@ÅDhÔbvÌ~P†´a¬$[ܳ5ë>®xìDò³ _gqK³Ї‘T“:y[¦º¥S'nŠ ÿ@/ý&Ù —0Ë0`G©Ü‰qÖ¯íCh{ŒÙÙ× àÅ*Ô¦®]S›í4n %®²Ÿgµ,`“%æéák=Jb4lýg§S[P8ß¿YZâ–ÎñQR´ŠÖG’A†˜¢ùáŠ'øL.™Ði½vƒÒýLU(ÚSÑíÿ/#ÿ?CÇ, {35†Õ$å®·ÆâÀÇžüûÏ›¿l‹Go8—)™ x-"»•Lø›(Í—°ÌûK6²`zÊ„GAj<ƒŸ4Å,p÷÷‡F’w~`ºkºÝØ@Œiføºu€¹qÁaõ–k2îN¿óêËè/%'ÿÏÁš’ C¡ä•</§Îž®ÄݳÀVec&K×¼ÿ”ó=ÊõááÆ jÈ”NÆ¡ñ9í^Þ;鸗¾²»£ÒëÓØ’ É6_ÏD­¥ˆàÂ*IÏv-OцóU”$IH'§n›:©Ômc8še–ªì¤x\n¹¢$Y á²ßUU„ƒ ÿ0_ÇÃx|?¯·‹u¦SzåéF›Øož<øÓ® J!Ÿ¼ûÜÂ`_a%ËI„¸ `\ð÷‡Dü(aÊž3¹&D­M¯˜ŸÎ^ì Ÿóë8wµÌ¦.îìÜ ÿ#”ßoÅT?wz¬bNIÏïévæ©À„P«õjˆ›¯§ðT;RjÙ…ìßs 4ªª…–hhu•÷ܤR­UmWÜG²J§Åʯ÷@ ¥ ¦r "æÛT¯º¸¤_¼ÕOLêÎW.S9rM£^•x6ò¹hÿjZ†ûe¬†3½ÎT8ú)~aVÈ‚e5Æ%.&•Ð)]ã":±_û‡C7ðvTëtÐrª+SYžÉ5UNΆ£üuÛNc¦=´W‹¥«çMšŒ†Í“ñä4¤éÉ©tÑVŸ^Š”‚*xbb)6_PæÝæ?:®9Ò×E7ô`"jy{Jñ<æi@š4YHy“Z®VžJK¸â5»H\3»Ü½ç&”Tü³}ìÕ\ÊÿRBêhmH.WË0:¡XÈÄå€Gá¡q<)xzuˆ4;ËœŽ8^œ½­äY•ka›°X+:Ä —ñ$Éz=•ÞYksÕ¦yK,úFY©¨þGõ7ð+†tmÐ0â´2¯R7´µ¥®ô°^Ž Ñ7aá•ÌÁ¿ˆUt¶àt± Œ\(#±r¤`±ŒHì!÷¶r­¨Ì“60T]Û` s•r¢×LB½y–¹&ÚǼ·•Ùˆ¦ó %ڴïTþ¢¶ûœ.E±·¶d¥fEæ´zÈ ö.VÅ.í9‹<µ[imÏ;ã.¸_̺ìÇ?ÐŒL©¹ 1#•„¸šV9®ñÝ-y6íDVê/Ü›¬ÏŠ¥¿mKÔOå¯ñTKsõ)‘Ä[É'ëLüü¼ž¾óx.Äã9½\™-¿£!øäŸ±¢“W|ÜËg—7oùtVûOsYàºÍyhëåVô2Z_Qs,0¡]<ú…¶è¢éAlû½¡¤ —'Å^Š^^Sç1¸¦Ó·ýš&©Î~=bÛfrILF¢ª\•àÔº9 ²À̲° 9¡h‚ó3:–œ)½oÏytD!²¸oÏ#R>eß\Ø‘Þuœ’ÜõjFgv”¥u!KĦ’³xœHɉ°Š¤Ž“X€¤öœÈ‰.âîùjFè•g8è0è, P¬ÀUöÌâÔéÜà êtäHÊEx`DB¯D[>…%D·åpŽÖmü®¯(M³ã&˜Ùø°íU+xà…\å{Õ¢%4‚k5 Oj°ÇŠK'BBõkÒ‹LÑVW¶_fAQÓ-^$ˆÏs Ö õ[Z€YÓý’Ý °bdýdy8³šX,A2SÞ²¡-y»,|]…³+¥N´ëÆ£}8—ÃÉe—!Š:YÞ|;ܑ䵘áç¾I;G¦ ;GÉë|åtöpJ.½›I)›¾CYî<ϳH-x³Q$ç±^Òjã}ˆõ]Œe†ZN[3ÕЙœUÓU.à¢ÝØ‹:,R.ÊšºÀgÇ2ÝL‹/wÐÑ'0uêgV÷ŽU¤ :›¼|œWˆ û¦š²]$ Y…xc^ʧ-\ŸÍ:i’d°x­*Cèí©¬VÊÄž°í[ðd&ŽÌ˜žQÄÔSž­}褹;&ããÎŽª´÷³¹ÜZ°DA\ÓCØÛø{‹Ñ ^r~Ö¥ œ {‘¨ˆ[Ý"å5•»‹.Í+BÆÇÿ,Ò&ÖYûO`í”ÌãÆ¦ädØß *K%ÃrŽo0w‘¸›“8¿¬!Y°ð ¼ßp-ãL%FÄXXÜŽ°¹:lat°åíd¸a!‚ªôò-”ʸ9¿ïå²Íƒ•pïâ¿”GðÞ&õÖûô¾•ƒõÅâ ÅÒ™É×íf†ŠgT.Ýaú¯‘ðÝOCÔÚâÐèœÞÐÀ|ŠC!V”D ÃlŽIâçÜu¤4å,Båê >¯¹žwŽÏ“<šo²¡Žî=Å^ðÔþOË!¼¢b`%Þ¿à¹à©ªo±~Æðµ§´äw# p[žûóO7ƒâ@ì_8`s.É·O·}›âÂI£ŽÂh|¥„æ]ã „e—2µ·§Ä°½•Y:Å%™Gtpòï+aæþ,DuD­™²ØLŸvŽÓLqóõ¿‡”?’„¶["Z­îR=Ú–uÉÒIœ?°i¢¶¢PGžï{ÿ¢nk?¯k'Hi‰§Ë®÷Ó@ekÙ5]ÑÙÈ»ÃãÕÅÍ÷ðh½µ›r—q¥ës¯8ž±óyî›#Q7 æ*Gè©CP)·!9þ¨e-)RV à‡ˆwÊ.„ù³¸¥ã±¨Lš °°ìè–",]96U)UD³õÄáiW0aT¤ô‚d1…0•‰[UoÇãoÇã—‹îzÜÇëÅÑö‹oÅgïÆãÍbÛË¡\RÏD \Ü-¶=7ãñËñø¿ãñÙbÛO‹£½ZíËŶ¢zïùx…|Æã~qš‹Ó|±Ø™2Sÿ@ž„4V ˜Þ ·Øäƒ[:±ûxœØÓ.B¶!ö;†².>Óȼ!¶ï¶ýú‚ïùmB,+\ ˜¯ò9.ÅvëˆûáëÞ¡±°&­©6VÓÊ„¾Ùö¼ëÁx CiczÉ'ò0‘ùaf\Z‰‰´È¸1„(ûšÜâ¸낃!+„²íÚT,Ä=x‚Ü"ÝLèLXTí5J4Ú”‚dWŸ@GoÂŒHdmò¼šD¯‰èþñÑ-»DháÙnfïtfâÔlÕÙ€³{–. u­—ö½”]‘‘”L“‹ãð+ÓÍvIQä0|¾Ó[Èý…Ë£«€BÂh–?8KĈ"%.(Y>Êö+Š ©ôqˆ©+¾„¶¦9E)üĹŸPçºÍžéöפu}±{©«H H^‹êèÕ"r<º7¶ê9P¤Ç{žlqt‰KŸB/øó›?>lŸ"Ä U” òF϶=Í7ÛRw|•ß×ÜZJY% %g–Û¾í®IèÃÓ§ßK9åƒqnV;?6LÅ÷–hóXde¯é“Ew¢_å´}ܩשfPQ+Dqöa¢¢¼?hìÑØ ã0?îsë%K2ipÁÃ`rˆ¤ä «}ö-ëì—Pî™j My‡y~Óóž;«5®iz‡,€‹Víë³4T,ä“%0ÝâˆG¯‡ã˜ŽÌjÊV*²yK7ªD­Ö¿.†$<¹|7ÀÎAD)l áyJÓ `Wµ¶-Lr ˜4ˆ™ ä ƉÏÛÛ2w*Ê^ÇDÞèG4Ó±òéʉß¶ýÖa‘§ÑÈÊS‘¬óiæ‹´‹’ Ê÷¢H„/¿ä&”,9ëáÂcg}6œõËa6>Û6öI,àCæ„Ýl“GéÙ‘»1@aèèa¾uèU, VB¤¾Ø¾ÇžuÑ•Y·‹Î‹®§HÅþÝ—£kº 5D`ÍBG»…foˆÙ‡Ù-\ÏCˆçÖƒ”3õÐ:ªlùCnWb¼wCT)0Ot:±÷ñxê9Îp[)Í̲  ™Å¢;0ÇOqO(š2ÛÊ•´‡ v;Þ  ͘‚¹QJI]mÿ€Á•÷/ô½^eoiÂÆI2W®ßZîM•IÉü.?1ß°¬íè,y¡ÌÁ´âÂQ‚ +G”Ô½BKæC3Σ¾"˜úÕÉàÔÃÄÜy¹–BžœnTúú±Ù4ۯ̣TÚ-Wæ[÷ú³¨©¥{wÚe¼8ºý»<œÈY• «¯rMÝZÕÉs66ð­Ê¨ÈD9Z Ú$«±|±ˆº‘³Áö/„°É4Þk ¬ëϘŸ‘?±ig}¹¿Áë•M›ÈÿÇÝàà$¶÷…•$ÿwW¦ms¤¸‡¯zðÿÕj?endstream endobj 184 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 371 >> stream xœcd`ab`ddätö vŠ04qôH3þaú!ËÜÝýSñ§«,¿“ƒl@Ξ¿|Ý<ÌÝ<,k¾?úž'ø=›ÿ{†3#cae‡s~AeQfzF‰‚F²¦‚¡¥¥¹Ž‚‘¥‚cnjQfrbž‚obIFjnb “£œŸœ™ZR© a“QRR`¥¯_^^®—˜[¬—_”n§©£PžY’¡”ZœZT–š¢à–ŸW¢à—˜›ªs¨ŒáœŸ[PZ’Z¤à›Ÿ’Z”ÇÀÀÀÄÀ`ÌÀÄÈÈ’ÿ£ƒïgP÷‚ïÛ./}Èøì»Ê÷ùßå™”¿"zjÖ«Ýß8¾GþÎ펓ûÝ–^ß&ÿç{ú´ú r»Ù6N±NþûìïÌìß-~ožY<­{a·ä1¶ùsºWÊÿ1ûþUô{éw1öï b=jCóËó•.ü7ç{þB¶ßIÓÙ_r}—á–ãb1ŸÏÃù]foÐw‹endstream endobj 185 0 obj << /Filter /FlateDecode /Length 151 >> stream xœ31Ó³´P0P0´T06P0²P05WH1ä2Ð30Š!Œ‘œËU¨`aS027iÆ`9'O.}O_…’¢ÒT.ýp .} *.}§gC.}—h ™±\únúÎÎn@n€¾·‚.PÆ9?§47¯XÁÂÎŽËÓEAmGTaý?ß ÿ p«ƒ¡ËÕS! B)Éendstream endobj 186 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1à Üy…@B‡.Q–tÉЪjûbLÄPƒúûB:t8K结Ïr/#»ò=>)ul"-~H0ÑìX´ ŒÃTY™øÖAÈáªÃë6Ùßô›äCµ§²j÷zCKÐHQóL¢kš¾³¶ÄæOªÉV§:÷ ‹ÿPr4—8n®1§Ò´4ÉÓï™àCNÁñ<øSdendstream endobj 187 0 obj << /Filter /FlateDecode /Length 161 >> stream xœ]O1ƒ0 Üó ÿ €ÔvA,tahUµý@p”' aèïKèÐá,ïN>Ë~¸l#ÈGpø¢Ʋ´¸5 ÁH“eQ7 -Æå‰³òBö7åßO°È~W3ÉgS]òª.!tš¯‚â‰D[U]kL'ˆõŸ´F³;ÏØe4',þCIÑT⸠¸†@sÓÜ$°L¿g¼ó)Äm^Sùendstream endobj 188 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 2018 >> stream xœ”yTWÆ«èîêƒ(]–€bw‹#ÐB«¸"DöEdŒ¬ `³ºE (qâ¾<Ä“9A&Ê„Í ãDpTŒàŠ‘ ehPFFA4s«çufæ!YNþsêÔyKÝû¾û{õ]š›Q4MO Òdhò3S“'gÎÂ\Z°5lEëŒ%ÆB‰0ÏœBæ"d.®¶e/Éà²%ì˜I3)M/÷Ù½ &ê='•ïæœÂÜÌôŒ|åbõ"weJ¡òÇ¥Ÿ&/3]§œOšìÍ9Z.?,S›òQžr}².ï—Àÿ/EQ¶ºÍ¾9¹y[·¦j22µ‰®êE‹—¸/]¦\±0^EQáTå@ERë©hj#KÅQ~TL…Pa”5ƒ²¤dGÍ¢xj6eEÑ” ušZMM§æ>”˜||ˆ£Céýô?ÍÌâ̪̾ù‰öŠjÅR±£x™‰ÏKfHJ¾bf1{™R‰ÔJºTç,„[¨Bˆz”Q!ëèöꇒ>+î|‡PÉ›*ûá²”{ð¯æŽ¶Ë©¡rüß>A)5„4Ï—sjO«Môa¥$>¿[PuÓ pË ‚“Â/Xb‡ý°ÿ =8Á‚ço` „,À*E±?Üì…g㙑^n®QÝÀw·û¥Â˜–ÿT}J7¾ÄëÀÝMÝ‘Zì›nð4„ÍëÚýîFvæŒ @c¯ª †uaöÄîÊܡӆ¯ËôD,vtB ÜR°k¹³]W«8{\{,š”ªœõ¤ÔÛ“¥Þ0Xqncß§˜†˜¦¼vÄ‚í Ðà~KÁ +œÚEemZÇöH9•Ã=€í¡ü‹ÖUØ¿¶J½8ªfÀŒ»½Ï0@8n«Ø_!{H’[q÷þñ¤´d7›Az}&åÔGŠöR²â=Y.b„<8EðšvDBˆ`ÁžüO'ÕX¬u^yyÌÔ*ôÛ¾.ª.@i6±qú$iþXU ßy|ßñýuìæ(žÞ óˆ~ëþãú÷¯ØŸTx|xBW…êlš®Ô¶sA}DN8¿I 7fƒ«P)YÈØ›T‚JâÈüΤR’*S¥d˜ܾ7¹Iˆº©¸ 5<øÁ'$Ëp–A”DÅ€/.Åëp…d„H+œ(eHÖ=5Î"'Y ØÑ¤ÂsHöÅ ø›ª³V¥e ›"ôû#…GÙ5ÒòÝåªÐitöÓª§>?^uâ2-¬FsŽpÉ‚qþÀÏo=›ÞzÖ‡À±Sc1öÁžCJ°Ëž.˜UªXÎ-ÿ Ù±nÑ]0¬nëÿÑÑâSªxû£BP;¸wÓB)ìäQï¾Æ3bQ‡QËÿ í#晴ΙS{vTÈk KsQ;e¡ÁÐ;ŽÞIùkãäP+µ€Ü·|ÇÞÈÚúWó½gBl…©²áÆZñr»~ÆÂøøW&Í™|¬¸/òéS)÷ÍßïÜk¿ÿ¥&HŽ˜\˜œÞ»°)xr*Ì“ÇÜtòO.ˆˆ“g7'×"–Sû£„-‰„s8£ÓXßIƒ¬OT3xû·7óø¡Þ!»è5mx-‚õ¯y{û×Ì$ÈG =}eê†DpƘΣpmZ\âÚ|„ç°ø`/æAÎO‡Ý üh8¡Mžv=âlb=Å#לñFœ’èâäš0ñmtd # eƒ"(ÊxSÙ 1=RšE[w`ÛBvídŸ)šÉÝ­ %ÞôæÓÖ/ê ·Æš­¿»sã1zŽZóš“/&׿÷§ ´Ûæj>N:DºSÒôÉÙ²šê†¯NßBlok¤g䇱¡é ·¼`Ù{ðb!›™êƒÝ2A10dÅ=„-0Àcê˜W7’ÜÕÑÑj焦ï÷)ˆÝ"K5åYuþ3F‰åæ¼×—Õ*¸gm§Ï_4f®êÂs±ý»xáAù SÜpô䱪ŠK 5w«¿–๱0{ÓvEÞ®¬a‡Ù·Åƒ[ ½¤þ^ðå±ï ø¾é±2C¤¸Ë/q•ÂÍç2 ’·JÁ 7-x9ÆXÞdÆl0Í–†ízYÝPÈ&/+nT˜+Xò¯V_@í,0î=¤ëX.Y%ÕåÜĪ”Ô9ØzÔ«üöX÷¦¶¬¼$çF1 å|Wc’o`\’÷šø«W»Ü¾+l\î¥^ÞÒ?Ðr—4ná ÄvÐÂ%¨àMÅþB1ó‹þÖ'‚bgöoët0ÖOüìÞ¾µPŠvnM8Äj¤m‡.6¢g,˜K-ò«…ÊãU–õódÚËwô%ææ}¥æÓ)ꇲ=/endstream endobj 189 0 obj << /Filter /FlateDecode /Length 4479 >> stream xœí[mo$ÇqÎgÆÈoX4+hGýþbÃlèƒ}brìy’KñE’…üx?ÕÝÓ]=œåð|:A A¸foMuu½>U³ûÝFŒr#è¿òï›ë“O_iµ¹¸?›‹“ïNdútSþys½ù÷SP˜ˆ1Š(7§oOò£r£œ½ÝxëǨíæôúäëá÷Û G%”î¶bÔ¹ár»£!Úá{N8ìpF»6FïU"UÁ(Z¬¤Ð. o¶ÊÀ†á›íNk…µ¾<§Çt ÚºD,£PÊ ÷XÚà´1ýsÓÿuúGº¨ç×P6ŒÊâ&§g'ƒßž~{ò»Ó“?“”UÜXëܨÜF[6Ò;³¹;ßüçææYUIa7n ®S” £önã•¥“YS¯‡‡ÃÅÝþzsx»¹=Ü?œß]î6·w[qºŽÃáõþõåÕåÃåùýæó<ß ³¥‹$1õè ûpÆñÿ·é‹i£aM/¼UÜ\·)Ô(üæê䫬ƒ 9մè””v4ŒhÚà4lj¦Nå´MàTÓ§ Q†KUwUº¥1~tzcƒ £µ$¿Ê‚©;é–… GùÑ*F5ítTÖˆQNUv:*-ÝŽQ•Ž*ÀMäTe'ßà¸ïİ1cäž#5 7Ó£ˆŽ¹¼>Ž“ÅVI¾zYé]týÕ¨™¸B¸!‹ª„0Á ”3$EH4)N(BVŸÂýærq¯¾x Õ”dÇ:T!É0À_BŽ>Äœ“~{¸¾Ýjd¯èÝ0nä,¿¼zWdÿâDË(ÇÈ }Ýa%Ñžd`Tӧ£œjÚáTÑEZ3ªi‡QÂ_‚ËUw8• aŽxêN@Þ¥lh: ~Á1ƒ—TÖÊFªSÙHŸteƒ+¾U¼¶Ó*Þ¤fFUwUQ3£ª;Œª¨™QÕF5©¹QµFUǪ̂ê£r€B ãÊ9Ãq §j¨o¢â¨¯RUÔW©ê«TõU*†ú*UE}•Š¡> >- rïßUܵT’NbÑuIµRܵÄu¤y’¼-Ú© J¥x ®peû¤tóJ±JÂ\ð™ÂµN>h"!ãó¥ôDäîàI™ì­ÒP@Höi™ähãDp(÷pîyô"~Ï@²uðÑè>¬|rXWà_@ô%~=vfAön%~ý©õÿï\â5Ý65ØÇK¼úJü‹dÿóQÉû¹Ê$<Áj`G4{ÓhÅN}óˆD.iJã„Q–™Q: çÛ¥8Ù ~³Ý…H3$3œuáÂæ;™á.äp˶¯hD·‡LBs¨kšCEj†2PŠÆÛáÛ"Z)µsTc‰ËÇá°Ý¡·ÈÁ4µ"©|K²/wiÄv!š‚ëâPÃÞiÒôõðë-Ȇ²¼jËó¶ü±-¯Ûòr‘à×Û¥)—¥ªÓ–ïqõx9|ÛéhF 7¼a#¶¬‡®d¸gÛŸlwІë‰Ïæ²%•Tf2*NÆFÛ•XXÁ½‡}žä9Ø÷f&“ƒ-yÎ2<9¼¦ùžAà%Ú€.,«1b¤n~›4¤òE×iyhˇ¶üËÐÖ×m)ÅvQÛ”6µˆÓ™EOAvÚ#••å.d¨†²cKè’Ó!;ƒ¨D1Ûì€ø‘Zë¬`/™JKNƒ¯4u¸H~ébÔNûìô¨èÉ–¡ @9ÞNÓU_¬ oŠóØ" óܦ£l éä}æâ,ÍäñÄœ€ )ZˆŸ4äe¡Z)ö‰µ´Fû&jH,®V‡ËÜ£ö¯‰@Xáù™ñÍ»8Wï$£åœeðH–ÃXY9 ›Ó/ON?þ®IRyíC¾dB¸8üó–2B4H%‰«Á¶6Ep­Œî4u“˜¨«½aÛIó—×\Êû²Õ1‹í²Ö"®÷œQ %€“ü\¢œ¢ñ”.¾ÙÂèpBòôT”ª8Ê„Ð\›}>ÐÚÒ_Ãë¶»OÓþ¨¥Wí&²Ë:‡üB E-µ–·F™Ð…ù$¥Èq¸iÞp–U`“–È.1:KŸ5;ߥ:aqïÇNRvv±„c¾×êBwìVMI³>¦ô¡MQA¦tqSS[6Z0ž3cÕqŸÏ³Ð†>ž‚Ó“¾ú±c´û›"›ì>ï\¤KC“´) ´µIè7Ì»I{BØ”ÉÉidåž<%ž£~Y“þuJ‰H¸aŠFDSï¥WÍK¿§ó”P}`ÝéH¢OJÞ ú˜Ó?† ŸhžMrîŽ)uÓǪ=5Ù¸®{Ë–P¹/y5ا®F§ùN%(œ±…Dƒ·&hô­³ 1{K¦Œ3< Ù=¦ó(ƒÎãuö)&CçÃþjÔ°,}ÅÓi5+—è>û>X¶L¾ÏhÓ­IµS%çKLŸàØ ¿Ë†ŸZMâ’ž5.cv“;IÐ(-Ió÷Ì«›³ì pµ‘_HDS(/t7y×_nœÕÃ>¿ DN/6‰ šÎÝT©QÙ…r]Á½©å­÷Ùg¦ä vQÂe:1…"@S)3aMŠѬØ`g¶f¥üû-Z;ò¹) å¬Ç+|–4Z“° £ç*TÓÅ-‘"Lêµí˜Qbçˆl=Ì`Ù£ Á#”eO“ ³Ìxh*ý‰ñÈ–µ*•ªä×ÎŒAÉ ¸)ë8$ÎgíÍÒx|Ò»à2<³ ¿iÑ|Ñdë°KÈ~»ÍÎ^dã@’¥BŽ_.È9ah4¥%ï=p ‹ë\ÄZ˜õE¸ òšéÓ=O¾ôœøì" ç:@mûh$è.¯çôe²&€ó\UIˆÈÓá,j!’OtI»Òö`§Tthc!dî…ò(h@æéQC¥6Ź*÷ A¹N¼7¤s'¬¾i" `îë-…T¤оE¸pž?‡… NrtjÈ”i॒¾ejI¦Øó!v‰=å)ö]å:EÛ¨Ÿ¨?• £ÂõP¥ÈÖC•ÎsPújB£d‡×쩊æAÂM{{! ‹ ‚Æ4Ã)} &Pˆ›¢Þ¬d<Ôo–ñ̰gù`ª˱Âhï3­®oŠª]¹B/:+ÈÍ!8¶@Ž2¡€³¿båô*ŸC}þßÛ|Õ%?GŸ}á§G ¤6tï]@°ˆîƒ0YŒ"iR÷ûTæ‹%¡£‡‰ÎO~þº2fs¥\h0$É$|™I¶ ¶;t·›øØ€ŸÇ§ôVÖȧQn‘#"œÃq7ó¨ÕäOÐ(hÚå IñÑèÂ&ámdJçû‹}a®LLñ[ž|\Ÿò(al)Êlâú1z‘Ç92ŒÊë z7Zäÿ§ÌAª2r–=A(ËÆš LÎ>BF³8º½ªÙìºþ²mƒ7>ùððoù>ÒÇéBu$µãXS;~ûÄæ%^UЊ"i©†u¡T—”K`“0¨É±*‹tbVØja>[¬|o¶õÛy²ë±J5Ó‘:8]6@¦}$OŸ´Ÿ¾Xœ÷G~•).X^&XijU½kÖKg¸Œ{Eõ7M£ü=\Nrú#­SŠ€ úæá¬’p—¯áqÅàÖQFÍÙ*e;Ü^PK°} ÊRÑ’ÍYf̘K-›ž«@«I' (¼úäE1,•ÅÅaÕ\÷»*ï4ˆ1áEÚç}0Ÿ,“Xh°Ö¡PŽˆ4Î;ÏâéŒßZËÛ¸ü³aJ†šÂ”È…O!DoÆ ^ßOr|z(*ñÉÄ‹ýöÓwú_•F¾éS`®Ö}Ò“ŽyàFâE×£§û¸ATÔ€9bÂ.ùÈËS‘¦¯$ät®âšÍ<ù÷xÙÛ–óÜ{Po¦Ì“¸³{W„±ÒpbV¸!üª]®ê¯›,Úrwg±É°Qé_Ñ-ÖIý°4$Q:šó!ÿ“Õ}°$¿¿Íìú"4õz/ g\Ü‹¸”J5EÀ»¥RÌy¸cGékŒÊi5Õ&7üdo¼™.è¦Áa:â.o{a‡nÊS5sd>Ä1ÑäQ¢Ci©ØäW,û9“:è˜ô!R.Õ—‚rwiÂ<½¦ |ÿÑŽtGF¨î%?ðÈÛÍìËi¸É|ù±9ûuáŒhøa›r¢é•ÍK®Ê,5ú.ùMTnËiÿшaCºO²$;ÀyL‘ªòV9Ù”Ýê"ûSš@ÕvefÏòõ)ú1„6ž¾ ðn¯¸µqÔÛmè»ÊÖ¥÷܃b_¤~g~ÐÊ9ÎïëÙÛ'øŒH4y`iböB´œj2XMz|ˆ™µ®µÓ»Ð|ZC6•ºCæ¹ i8jß#°Dô«æøÜÛ÷Ù­Å|ÀÉæéù ½¦‰‹†.Ø» £óä^Ø"T%à'lŸäÊPÆu ÿ…U:NoŸ2/£^›çÏŠy‘#vÍc™º‡÷z+>ý0föV¼ºË§¯PÊù^܍Ч——¯¶ôFÛ´hŸ#G+ Ä“ºƒÃ&Ø^–r"pÃgmw×vÙcçmù¶-/Ûò¡-ËËö§Ç5Šü²½"“çd¿o=.òeË}[2Ùÿº,Ú1¥Ì¥œ ýƒ¦ßgd¡ÿµ ýÛ¶<´%ã}Û–#?‘NA‚ /ÎàÚïÙý¢®/×ôðÀmX-ûÕ¶~Ëä¡-ÏÚrœhC‘!íÞ-.|·>öS[~ßömɾ óØ–çí±?-žVL˜ÖŸ·åÿ¶åO‹»Åš»NÑed^x&{ý¡-oÚ²Sã’ ½Y$¸]4D©–صm±èj‘ ´¥=B»“ÖK¤ÅâÒŬu)—±-Õ"oöÙ-Œk_V!—9„¶Ô‹tK$„ßrLþ¸Ý9¨$ fQ§L{,ÞÍâÒñ劸vM§LrÉiwp…øÆE¾fqéùn•ì³Eý³¯ íepÍ®·ÝnIš&õ²áóc[þ÷¢K’Ûkçò—áíÂvõ¢좉Ě]Ä¢þâ¢*-_V»,óeéÅ#Ìšd†[`Ép¾ª}×ôÛ-§Ï¿j›—my1Ogó²ñ–[¢¦%–Îmy¶˜äXeú ×X.žhñ;d,Tþ’mú•âG„­çÖŽN;¯¦^óãÆ¾[.³§ßDõ{)GmuÓ¬z [ÊåC5ý¾$¬]ÉKÙ_‰…ÕÓ{ oÐêYžùòÅ÷8"¼ ôÝ…Û'’‹.ÃËøsÊkyÇ*ï1¥ÂÉ­ßû(-?ZÆ”4)o¤À“~¬ .Ѧ÷ |¸™¿^î EúMI?úø 7§„õaàf¯èÉŒ¿ à\<ÌAü"8bõXq‚ÑŽBA™¬,ªÅ]VBC³c¶Ìw™À6'ø¬-Õ"‡ŽöŒ‹ê ‹ú‹šfê]œñˆFVÔëù²Ýpþâê}/ì) ýÖh½åb†s‹&êZ®‰EcØEM¹ÅæJr ¬xÁ2³_ÂZ¥½þr»Îš·ÀØÀØc«ðÏÀX÷³Šûžˆ-?bü²EƒÖendstream endobj 190 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 1662 >> stream xœTkTSW>7ä>Ä’ôV47íTªTP¦¢.ª¥.Ôò¨¶b‡$DIbIx ]kUÖ)Z‚€ÄGAy´Aky‰:â”* k,Îjµ3Ó±ÕÎàªKt_æÐµæÛÕù=?îZçî³Ïwöþ¾o)ˆa˜™úÌì¼L‡9ݰ(Ú–á …ÈAŒ§”!ÊõJ³ò›È:ØSl;{½Íþ28€…ûÀ2m …JÐúÀyù3‘, & ‰!º ‚å°ì ˆ†5ó“(©‚°âÍ^ݳÁbV­Ü8òðaßȨDì0 ¾‡iV·ê…*í,õYxSî‰ö§riãZ×uÛ†±Ïü ü!r4wÔÔ+©7ö™ô­ñA±x“Õ¼NPß„é6ñzWìoçoŠY:üヮáë>»å߀Dѽȃx²ƒÂzxõú[·j>,ßS¯çø{ð.¼'²£uØ#Þ_îw€ÆAÁtÓÉ´IyŸ¬÷·d„ÌÙû’Žø˜ª²ëó¤ú‚:ç¹"X±av[©«ç [3ò㶪 5ÅU»ªv5Ñln?ñýÇZˆÄ_â?ÕzZ<-Ç{q?¾–uvõÓ7;ÃUvN¶¸{†NÙß¬ÐøÃßàÑ]æö~Ä(˜G´dž.t„ü{é¬áû–cïhÐŒñ%{ß-/Å‚ÑYÕ-A×_ŸÔΆ»å`U'hõë`.åU>R'Úí¬Åôó Xu‘“à• $À© ®òâ±Ã#X8ßXh*t;¥‚2Œu¥Ie³µ…’qˆr%þþåóëÕ¨7bÛûeÕ;XÈ©«IXÄåé›W§Z?òœ>ê¾äÒtì¨ü`_íÀ)¯L[ä–èÐnÕÀ/*§LXÄŸ, ½À«{^oíÊq fÂÒ¿äÞ0“Œýq­q8§æ˜Spñ0ôŠ#=1/Ìߤ£=¼ß=|Mòvúž» 33òÈ+\ø¸ŸßÙûÎǦ±ˆÂP¥~·(IyåÞ³°fŒ ?n–ˆ†+JËÈÙˆÓñ¶ºü“ï¸ß;¾·OØwWwitñ§Îâ§2`. üÙû ÕCƒøŠ¬'Ë8âšÔ±d:tpZ{ø žlA$ $ÀEÖßûT]æf> stream xœ]O1à Üy…@‚Tu‰²¤K†VUÛc"†DÈÐß7Ò¡ÃY:ß|–ÃxÙ%÷èñI ¬ciñkD‚‰fÇ¢U`¦ÊÊÄ·BW^Ÿ@°Èîü¦ß$ª9•U»‡ÐZ‚FŠšg]Óôµ½ 6R L¶:Õ¹¯PXü‡’£¹Äqp‘8•¦¥I.à˜~Ïr 6ˆ/=•Seendstream endobj 192 0 obj << /Filter /FlateDecode /Length 4318 >> stream xœÕ[Ë$GÑç<â̹ąj ‹|?Vض>[|d–‘@2zwgz‡Ù¦ÇØë;¿È¬ÊŒ¬®žî±Bfs²###ãY‘ÿèÄ ;Aÿÿ}}wñ«—ZuÛý…è¶ÿ¸é×nüçõ]÷É%AxÌ QDÙ]^_䥲 ²óÖQÛîò˿ÖÄV»Á[ øË7_öŸ­Ä „4Òõ·4”B»Ð_­ÖbÐ1hëúo0-c”>ô¿'ˆCpý Aø¢M㺫µ1zíMýù=-²1z¯Ò´ ÀúG áEÚXk=ÙîU0¤½0²†¯úŠá} ˆ`”ZƵ[­•„T²ƒY¨llÎ›Ø lˆªßg 4.Ÿâo—ÿ[Ù0ØËA;?1øeaTn%·Ö" Rukm‡àÜ$“ ‰{¶9Øû‘è¤ ýëLn4Þ&6ˆh¥Ô’¹; 28¾±±ƒ j"’˜P!ˆex]‡7uøX‡íÙxµZ: ´Áy¦~œxîµ'©¯ Lxï2£-Žäû¯WOFOBI§¼©BÙ¾])GŠ­âF½ÉÈðW¿M8Œ6šÄÀqÈåMg×BIeúÍmÒuÆÊo0_˜ö1ƒ:S¬!/IYb ¶yB­îhž©¦C˜Ëiè`ÂÅi˜†i΄]žQ·¾&ž:éLÿ¶òæ5ƒØÝ‰—Ò¹Øß' ±Á’ÇTЪÁ÷žÐ)Hùò†ÒCUËa÷ 6™yTÆÎŽm%4ëo0´qvÑïä>öI­´VƒörR£´(ë€vè\WvÒªµQrpÊuk©k¢Ê«ßT%$Y€¿1F•¹î„QfM6šk2ÉFÙÆ¬ƒÂÅ%µ—jZÄ š,Iþþ'MÚ |˜öÜVÛ¼­Ã»gY¯&ëõÅ—¬¶Â³ŒËj°!!¹FíI›EðJϼþ’9ò³?dt}_7ÉÀàêîU¼f¿yU¨d–M”Á`9={ÚØ%UÜì3Êò59¡  øW3mgn*¿a7_®ÖV€¦à{b‹ N›Ö5±ñ’À•†vá¼MÒ Rùч§ámÞ×á(ú4þi>,¢¸®Ã›:|\ÄpD|'fÖ0I®dA5Z,ò¤'ÙbC8°ØÉ£ mFCq –¿¹)®»u“ÅçM –°æîòw—?ÿ²¿L^C³Ò®"ç"–ÇȾǫ °kÉO­Yx!Ÿí]²åå–Dè,P’±?Tƒü¢—ƒóvÙ`Õ`À„b°9 ÁSƆ@ 6ª6±)†ÅƒKˆ¸óz“œåF3s!Ìñ¬0ÇW;‚îSìƒýº&êì[c˜0_Ïó=ê€(–vT ‰m¾½ªÖÅ3PY“‚­&d–¦®m@8¬Š9À6¶ ÎË,=Hp/uN4§J PTn!;Í c€ ZÏ(5Á˜& &X ÃØñy.#|ìwÕfö\rÄ•Cˆ%q›*Lö™ Õ˜éMÙ»E š\!(å1û 7U&͹á§MúÝu.›b1 l¬bü61Ùìx••M¢ r%½¯3 Ñfö”È:ª[ÆD†œøQ7€XäûÒIáÔåt|“RšR¢äF8÷Þ!Fg=·z¶’N2 iÍ`=â—BvŽ2çS/kŠÂñçlÍTJÓËÙx†±Âlk*2dUœüO–ÁÐ íl‡ êEŽ0Dް°ÓÁ!ƒ–#‡óy…Œf1^Þ©»fÍG”ƨ»îÕ|l{µIÇŠ4?Å,ENà ܋|UÑž€Õ±à{ñd% eÑrÊÏAài|ßÌù6+Í¥ð¤_³âᙲTÿyYªï'K¨‚R T„þyHM¬–¨EMæ­ÿ.‘?€@”—Øžm+˜¦&1V&¯/$Ùn¿½©‰Ïè T{£px‹ÓÊÙ‘óêLÒ(|SžÆß9VDSæJäùr¸ŸâB:d‘S·Ñ)»So\s–%ú PYSîma|H?b‰,§mCXæiè·9ÕÒ„쾪õÖý N"´hU‰wsè9%Z"æE͘‘…)ˆ0”ÍRYEÙ ÝE ô8šÍF™1]›·¥«H¯XÐ*UI´™ªŒ{]Ã4 Šy?O7—Ü!Û·‰e'ôW¢ä¡˜ÖÕÁ¶úK—)"tz ÌÁlä´QÍb‰wã¸zC®¥vpL¸À`µ_.yI9#·ÕrŒŽ¿¤Ó¦;¿'êàd9â‰ë¢¤c&ŸëI.«¦/1=ElíyŠYÅyÏ*¾#ùámÎQ˜ðP2âÇ‹ïTJ2éoç7Ôcö03âÿ»¼øãEþ`໌ïžû @Zduà"ØàÒw€ä{GÜÏÆG¢G÷eÎ8éàOÎéâ]¦ë&¬´7ïGî7ÇÙ7‰è¦dù67]%±¬îXQ™-Q¯¹JâW®2¿A†6Ç3¯5Ç?4q+@5eDئ’ ~vºA¤ZÞûWk«(Ûõü~NM³®ÿu]×Yv?ÏnøØÁ7uÈðÞ,ðÛÀux_·ûWæu(¯bºÅR9SSg7uøÈg ìGuø‡ ðE>ÔáUn7~·ˆWqØtç3x¾ö¡lGDø$ʳ«Ã7‹L½­lø¨?«¿«Ã¿Ôáïëð%¿Å9ï’6âz‘¿wuxS‡·uøa‘;?­÷u¸[D¶_2Ø÷uÈðþuÕŒIÄ÷eÅß×3µÈ 6ÜÔ!»ûpÀµù:Å ¦‹ÊÙ•êd¡YE>­¤nêðv>L_,éØõ⪫:ü®&Êl·È»E~Þr&.éñ Óÿ£¦OU3aVËý×›=õÿ´ÙK?HâzÊ"ϱ÷fQ(•çûEØo¹Ô ìý"2†©èÇu(*ùÿ¸È°Ý*_"EE0Ë(™/ÙÕá]2Z‡:”õ0bÀÖ¡æ®n-‘@‚``È8Àýd=A¬C_y êìP‡¶Ç0ùùïx`uêÀ¦]ØzIÍìtJÅÅ»$éåSšÅát="eû8©79ÿþg«ŸD;^ÁŒEÙí"ƒ¶õüÌ¥0Mǹ²H†±ƒ/_ŸAIºziI‘ÈI­.÷ëEö²á²(Ü"¬â³å0<“aá‚y·ukt!k"^ÉØ|pøMeÛoë)Þ‹º?ÓŠ¸hÃâ Ââ™¶}Rg[‡ŸÖ!£!r b–¶¤x~qÈ,ͬJ͈šNYú¡“¨ÏQÝtÚ*!Þ™îáªûs÷þÉÒO ÛQÙÈ ?¥Â€Ò¸Óå1ФÊïån÷¸Û>lîºÝuw¿Û?^=Üìºû‡•ŒØ.ÿ_m^ÝÜÞ<Þ\í»;yµŒLê&°Ýר!¡û;húüB¢j¦ S«`P&uwuFSEw{ñ§ ,µüp i¦))í`Ì4Á@Pè Æq˜i†9*^šfPˆô©•M3(ÏÀ ;Ýѧ®Dº³‚)3途³žÈaPãL…,ÊF“Û ’M3™²ãÚfèã8»ÐÇW ØÐ½fºTH7ök²I£Tg±Cð1_õ¦ËwÑ)ú(¥m¾7y”-ÓmQ¯»—¨Š QaÃ_3E@íleM`Pe¦B’¯@e‚ÁŒ’g@e†A¢gPe†A²gPe¦B=“ét,å;â ðo2@d«µ‘>rá5þ!4þ0üÇ é;DÛÃeU{LŽcÔ—‰t0Ý Û+-E&÷±lºˆwpUmᥭlTÛ«ÿ‡ÿ"ð>l±Žº«IHÓ§fç"¦-¶¼¸>ÚàÄ:²õ„©Ói Ò+9ùC èƒÂˆé40Egá1OÐtˆ´V-¤ðqp#¦3€€Ih—<ŠÞôÅ.c: DV`BûDÒIà±p œäL: Cöé ’ìÃÍ0­m"†w3‡@žÌ]›Ói ? ê¤-NØ''¯ ÍèR#ÄñqIvM²>²»“«.®èâ8^~~Ž“ŽQ(4©†$‡j“.eötŸîîîWÚPåú¡“3¯õ}Ø{‚öÏ/´Œrˆ<(35¸k“I`@Ó ò’ŠA4Í0 Ÿ&<šf*¡„LpšÊ 2*m4Íð¬ä9aH#púHÙ˜RåÞ†¡_N…¡y¦0”~iÂgy u¦FЉà ªÌ0¨‘Å ªÌ0¨‘Ç ªÌT¨‰ÉªÎ0¨‘Ë ªÌ0(‡<Š%~€"ÇĒê¤‚Š¥‚jJ PM'˜š N@<$sÒ¨A}§$z19~( 2¥:‘hi1èIw{ § Õ{:œDêEẻtâ `²Èˆýaìjâä@Àäb¤.ß'C÷@À0kO¤g‘éPCJjz:ë: CfJ}ŸO î4 ðh0RœÐÊ3€¾GpÚgFû¼$àôªÓIÀ8žh¯ ªîÉ$@ýIÀY´ÿñ™_É•E¢HVfr¬Õ–ßÀmYçLÛ 0>VÓc³gô –KDv‹Ï.ÿÙ%=»ªæWpìû:¸l;l½‚“GDëÖˆ}`ÜØ¹ñ ½”ûâuTZçŸ"žè…¢ÆÛ|±O_Ûó¦O](¨ãø^'·â| pÒñM6¯XskÐc;òCo&JÃñw,y]Ó·563_žìR“r&kÕØU:ža?áQíUyû£êRhmúÆš¶Ó qDÓK= ûA‰Eºïëõ÷h1t3Äð}Ù¿^1єΧÜS“Ú¹|Y#þ$êÙ šÔs2¾`K¿ÁÒËœÄR%BÛ _3¶(õˆIæ0цâ"Þγ_±Ö¯3Ú… 3½ J]CSóúQ2šÞbê$–¶ß•š¸¶éÞIÂÔ¾ü SÔ«¤^e¤¾È¼ÎYëÊ»-ºÖ­/¿Ú‡. ™á¦[°‘WQã%µVE?Á!¹€-Sß}Trl­Ikr_7ì7y¥Rï8ïjU?ኒ ý Çsþ–Œ os_=0µ2!zÓ¬ÏVAå†qÚ-5:rûKÑIÊc½@ÜàSòl3Y9ÑX@>HzVVZÙ[ÆÃnEˆjl[9Ô¾Ï$ÆJÌŒÎü®ƒ®‹Ê‡Ÿ¦½4½h€§ré 8uV6ò÷ÕünkÌ*J1~JŸÚ7³›´¿:5p$F{´¡¯Ñü‰èöóý$þ0ÿø<,‹_+£çâ7‚Þ~¸dƒÐjz’ؼGΤX™øJÀ:øÑg[t³ÆïbHùÕQêÕÝg$MZ™Û½*­‰ÿ¬Á[G™ Ž«¨Ó1/“&}é7)Në·Û½KJ}Ž¿Í8„L’JL!2¸þ¼^•wÍ#¸C¯çmá¾sÍ£‹ô-·Ú6q“«Ê$6¨³LÀY…âŽk,V”:ÜÇ÷ôä!² ×”¾¼å}BK‰l?=gJúý>O“};·Ë|´íÛ*±Ö­ùcÇ1‘ý7ê4Á±endstream endobj 193 0 obj << /Filter /FlateDecode /Length 159 >> stream xœ]O1à Üy…a鱤K†VUÛc"†"dèïéÐá,ïN>óa¼ŽÞeàðE¬ó&Ѷ„ÍγN‚q˜«n:¾?‘`7=ø]/ÄŸRˆºêŽCkÔHIû™X/„ê­UŒ¼ù“Z`²Í)/ªAbõŸJ‰–çMÀ-%ò¹6­MJçé÷L ±¤`û:S`endstream endobj 194 0 obj << /Filter /FlateDecode /Length 159 >> stream xœ]O1ƒ0 Üó ÿ ÀRUB,tah…Ú~ 8Ê€…0ô÷%!tèp–Îw'Ÿe?ܶä¾(‚±¬­n H0ÑlYÔ h‹±°> stream xœì½[¯nÇuø~~ÅF¿øœ†¸S÷KбÃh -6ò ë9ºØ7)‹’eç¡{טcÌZµ67IúÓC #Ñ>ƒ³Ö·V]æ½æü§§ðŸþOÿûñåÝøÛœž~ýÍ»ðôëwÿô.Ú}Òÿ||yú‹ÏE™ yžaƧÏõŽCãSjù¹×§^ûóÌõéó—w?{ÿ—F~N!õ÷¿ýžshíý?|ø,<÷9f}ÿË…µÐʨï´ÎÙ{2Ò4JJ‹výCnãýÇ©-‚:Þÿý‡ÏrNëïøþo~‰ayŽ\›ÇRjï¿YÖÑr)÷qþ?ÿü¯ñ¡ýüŒTÇsªëK>ÿÅ»÷±~øüI,å¤ù¬”øôY®Ï£5£û3’vRÕç–{Œ|ÔÏÞ¹^fÎVÊ|ÿõõç¯?|VS~¡½ÿC{Ó¾AþÏëÏç|áWoSêsÏaê•ß|•ŸsÍÓ_å³ë™éú³¼ùg?ÿ™¯×þ»÷ü‹ëÏ_ÙŸuŒùþϯ?ãõç¼þü»öeŸµôÜSœOŸÅü\ËL|Ûÿ|ÍÖ»þü/ןÿñz™ãÅçù¶o}ïñ9íÃg±öøâû¿¸Ðÿvýù_®?_«®?o,ןÏo¢ãüó­íXÒs‰k·}þ7ï>ÿßöþslÞBš}mäµçÇš¦±NÐÚ㹄ºN ÎUeè\å˜æ: -!ÔYÞÿþÃþóËëþèÏøó#OÓ,½¾ÿâ+>z=çýÿøðÙ:D-öÈœë1C‡¬¦„ƒ| įŸˆ½¿×#ÖIÓÆ2‚Ç[}ÁW¹ÞPûǬ‡ÿ2`„ØÏ­un¢¯í{×–]?øñö Ô4Û) Œe¥ÒlŠeU2,Mk,¾{î<£àÃóvLlÖB3xÌÚ¹›¿ùp®Ñ~Æ7|2#~:€¡@È,ÜeìÉãoÍ£ßÏ WzŒ<Þñ¥Iбø|§û[Û/ö‘Ï=d¿¿ç4ãó×›}ñ/—ˆþ®õüá³²lŸ“/y<â ýú…‹œGéÕ)Ö˳¹6`]pž?ð¿¯-x1Ï/ìQ£§nºì×Ô`C§¦z¬ùˆï_jNíÒkKé;¨Ùúñ’ßè‹Æå3ñ½BãæíçÎmóòÓ6KŽëÏÏ6 ÿ‚[½•¥kÿ+²öÂí³þÇ%Oo’]¿8¨S:5ÅÃZmqüЇ'-¢?Gþ7ܘ¶;¾º~ý<_Q”L‰Úëµ'¾¿?¹NÏ€Þú®è.¥:§ %ô þR©Ñ¾hu¤÷eéÉ×Fú ÿdÕÞÖ õ~5ÍiØÂÞ¶ø}éA1ÊZX|üL‘Z#Ï€l©ÒÍ笉ýº—ñœRôšÉ¨ÖW´™ÏmvÎîoì‡ó/·­p;T'ý/øNx“s·þÆÔƒÅ@lE¿õ¦Ç{ÜdÞ&x­l®Ÿ˜1ݹëžãß_ÜåÔEøÊëKþŽß>¸üßÎËÀýÂwŽÚzù†Éÿ=@†Ñ4ü|´ÿãäÚ`ãÜžzªµ÷5ÿöëÄ÷§VÀ—¯¡¤üš‰éÕLrÙOŽóôúüzûëï"àÿvùÖ¾°_X’( cüÚy¿Æa¬¦~÷ÞÅ’73æk2¥9bý–RkÝãw|ÆX;<®ÍLÉÀ¶ÁwþâÕ¼µ˜Áï¿Ô°µÏɹJ‹¿bµ©áM,®}Ô<…Ò¹uMkgó7n‹ Ðzî?¨uýrn·Ãr¬×_]*Å„”)¼N‰·O¾É>.L_>î_ü…æ·ôàm)û.ÚG:æמüúx%­_‰kXåÚmÿzýÆ]³çþ/Ïë#}ÿ_ Ìõ×]î]æ… î»«dýhX?p S¡cêrÊØŸ_\þúúóëëϯޤýòúóϯ?ÿòúó?_þÍõçO¯?ÿë›N‚eЭôøJžÜÎÐ:të¿Íü=Ö‰gQ,m´š_™Fñ-à ´â¢¹†n '<Ä7üy5çúg0¬R~¥Ê¹NöÖÅXŸÏÚˇí{ù×ÛŸòrýùñúóËëÏß_~sýù»ëOy¶NÕ뙯0°Óv¬ýÇ-ôÚã¶Lë(>¾Ä~nϹ÷ÎAûa{²þ“û–úùQéò8ýúÙ…þêB¿¼þüåõç¿\Ïý‡7 NÞÿyýùñÍßøýõç7þ>ãýÿsýy8Ã~r¡ÿóúóϯ?Ë›´Ç:~}ýù‹ëÏ_ž«ûÖs'üÉöÇñ>ÿpýùÅõç¯ßüŒïâz÷z}F]|Âø8aõ ÷G0Œý§i×ß”Ä|Úo¼½!¿ysÕ_Þüó‹ëÏß¾>œßÚd·~Q|x“=Ó:Üzx¿¸þüòõŸrÜâûKù€_½9ê—ן?öègòë7gãåÍùüòœÄ}Àÿüúó<’Ú¶ÿÿ‘ü”#ù½¢Û_-˜~ún‡óS„ø÷ŸÉ¸~$ä'S7§K‰ß¼¹‰Ž}xÛe{»|ó&íÿ:¥Ä¦ýÍ›;žpLôºþ ײþöÍyþÝ› !wPYÊwtœë#Xörýy¼ëlj×Ç„ï ØöóÏþ™7tiOõyéÚÇ‘Éןéúðp¡G §¿9¬n>Tß`cŸò¹é‡>÷|µëϱ>7Ø2'à ç°ýåézïr~Ø|ù8géøssàñ ÿ”oÏ?ôíé\ÉsÊüÛ½p ä½µêÇJ¶OøöúúÛÿ \>aoúàðCÏ7ÿã?øêæ·£ÕÆÔlñßÿÙ‡?ûÿ|Ðüí?ó› tÁ~®ÕëPúkéõv(=S} À7Céîbù‘¡ôŒª¿Írêùé?*À~°œã(Í7'áìåÃÓþÓ»¸DÍsX¶Üˆs™pOqžÏeýoÂO·§ßþòé¿?}õ§Å§ÿð.>-Îõôˆð®“×ÿßr}úÛ¿z—š♟RO|YÈÚ]ñB¾|÷Ó…Á1¬õçÊ‘9ž©âìkDZXŠÏ#U(ËTީƪО¨.$÷çâT½<7Œœ0¶ù¬¥í_©øwJ æ³„dP'½=¢¥cQÕö\ª=kÌçv!¤ZÏÀ–P¤rdY¾ATËîŽx~ZTT1—ç|!¤ŠcÙø|‹e`¿ȲžËðYM£)#éˆóU-ýÈRåñ•µ~9>Ÿóž¯¤ùjxW›¯ü<6°§Þ¦kl—— ±˜Ýžˆ5uåi}öòañ½ !ÕŒÏص؎TÛœ0ïõª%ÛàYÝR”!Õ:ð ØšÁér¤ÚðGnxþ(Ïuø¤Ž Ù ”€µñxíµ^=à Ùìa6,WÅA~¹×KMA¬|‡ÅÜ3þ:MñBÄÖtt²ëÎÍ ¦ØtªµÝØuzN§¢â¬9¢°L‹á5ìî— Éë¼lvŠSª±¸Æ¶–t!z¯eË•¬ùBZÃÖÑ&í|×¾f§p“|‡#â"Ù¾;×5_¤rdìùŠK 0Ö¼¦œl~&;kŽˆQNcÄë²ù$D‚|…Ö±³í½Øè|¯b«!D[b­Ãx2qÂãêH\ ˆ‹ŒÅYôÙ¶oâh‡\iÖÑœœÀ`rÌ‘\Ø­µÈÆ_±ÒÚ5i‰½¼íÀi¼ÚD5©™&‚´†™ç¤oqQ$Æú–‹ùHhµà‚Ó‘ýÝ×LVpÌõò{Q­Ny#F…ƒÜ@µv…ÝF?r1¶„¯ñø¥nE.N´ÍF€4•Ok7pX›ÁYäÚ&éNÛ¼\ÈV ìØ—©Óº‘%!ç¦Ê`8ëÅT%S ö¦#bk{3YØÜöèF–,«›jð´®ï,b~Óö»IºÂs¾Öw&ÿECÖ§nyØ;gy1$ÉÖ%JÚˆvMÂÉ]“Sýí…€·…-5;‹+tÉÃJ…‚ˆäa0r-J†­òå~íSð×CÒ­ï°3GdïSìÀXíô½È%5יަaG™ºer^ˆÎtÇœ¯—ˆÜ—«ÏÄÚáÝxlÒ/B<Í Ñ ãà&ÕI%R=ùL¬U´8¤tÊÚ–€Í¿Yz%€ÅÅ»Ó,ÕúEìãDö´Ý<«‹_!³àTkõ+¥¨µ˜Q߀NôÚGXT„Èy  PéÝ§ÂøÄ[su9Îh÷eòˆNR9R>ºøŒI™ÔÅgâ2/B݈¸í±õ+âlBJºø BM—aÈL‘¬€ñˆ<$ÃÉËX:ª/ üÆgš!$D3ZÈÍsßJµfZ¬æ+Q¡Y'Œ<´«Þ î³ÞËL ‡Í‘Ì¡uN‡ÏúšÇr!Ú/™§y©ûZc‹úbÛk Tì˜/Ú;ÞB¶6*EU#•#K ùŹÉ@-2IÅBª¥šÃÖX)`mØ­Câáé±J“ž²EN–º¶È1d1Ö´™G¥Õ»Œ,Q¼m˜Hu©ï‰TŽ`#o½.˜Ú·ÒÓÉЮª8Jø/¡ù6í©2û§È”¶]Èõ$ûœMòr ÍŒRí©Å4Ú©¼Ãúé¢3˜`ÊûFNå=Û³HœÖ³Ê¤D#²wÖ43%™é×{á\Úþ[_Ô¦c¸ý¶s’éáH΋Ìâ2ùý¡×nÜñB$íµÁ×[R ±ü×ìzCÄZÙÈAKáê ѳº lÝD*!–º©Ä—t×{¹™Ad;(Œ÷¬U×{9²fdsÅÇ î‡ëÝ®m]y=ßKÛ£eëH^ü³ï™€ ŠšÞ~pµ…HbšVi:E´KeÛ%Ü‘¢E£h}ÚØ€Ö'>'*Ï®(åÒzr£dX˜œÍ%¯!ÚÍ.ÅvõËPúïy‡î(!|ÞÛF®yF ñr ë‹/F·Ã~Mwí‚È]SYŸ(–äÈzû}»ëtñ-8MæFöQl“N†ü(rñ­õ‰É ¥RM%´OÌv„h » Ò’¥p^È;qO}´ ž¹6õÙDk>h)TXØâF¤!kºö/ÂÌ3¥CN· ·Ç¦h†b¹½q†,…g¯2WÿÕ•ÖYØZr–n1£kɆ,½Ü¹;f" jOEM¸-˜:Mh™9 nm;»Èf[¿&>´NJ¼yŸ‚}-œ’^Ž,^¸Oò:ç¶£Öé–|^z}¼MwÓþ©²7²väö+i§ƒ*“w̰ÿ¹·¹IlË’õmnÒ¥œJj茛gS ‡ ­^½q¨¡)ôË”X¿lê ĶI‡¤sD‹gjþPWV‰ä¥”­bFÛ÷¬gݼz¶Ú›äO’o@øž?§rO®çIlUN¥½y0Okå)pdžv8#Üó%ge«t¾åñQ™YJ®79²¶vñ³2äå^ Ks7Äbª›­Œû´ÍÝ Ysæ~bS');“Ö¯šÞ+D 89K^¶áºª!Ð&}+ÄÀZ†r"»^2ÑŽYpÍt´lç& NÈ2žG¿¶C§:.¦¾ŽÍÜÀÞ xuøîšï…L»¬û“–Ó,0âÍ3è #BÏÀšS—„ç´;¦Ýþ+|½yƒIEt½M¸í‹€÷ûè6ÜÈD.Ÿk1Õûç²7àB(÷æ#†„²]„¬“¶`¾;Nîž§'EÈÖÁèuŠ>]B°ÙúžSYK/ÓƒÏgldË6Ó)Àöa5¤ÏKÄcéaJîÑõɱmD¯>M €wV&œ¸¤Ã6•M±Åk± >ÑÏ5šu‹ç'9߈äX¯K°¸†¹³ødÍ ÙÓ’·qfýNé2NX¼Y ÁTÝËų‘q9ç1ÍX²¨¸Â2Š«þ©§ ã0%KqW‹!=û¶ZúQ¢‡܆á•Vè#"UuÐ:¤Ë÷å@ÀmöÚTÿ¹É—ó7E¶‰^¤æË‰#A‹1üÐK-²…’…6zöˆ ózSÓûÖ MÉ>Õúõ7¸Ž[õÜœ\;ÄßOlÔ²ˆh¬6¤r¤ûæ„mSè–ÎeãB´0Ù8!|ÊÒ…”pĽ–j¯ßòpÕ="šòN§Jw·¶Z FÓàåƒ_xñgžmšw„Tðé nsàQ¿×»&hÓ(F†M„T‹¯ôB¾Níd#ã²b[l4›«y!ñ‹eÐl&Bªœh$¯¯6íéæ…Õ4Nž‚¬·Ôà±!<öQrÝæÔH×­zì”]qêYñ—ºR­5Ëö‹òï_È:^þ¬'­iú†ìÃóÜ€^¾šð¢gèåÖÞ—+zM<#Ü%IG°7§švÆÊKLòñmdMžûP*^óÉ6=7{]l¡n€4øàõ!…€íxwõ¥g4d\½9}!š¥`êÞìP}_6Ën¯]¢¥—Y¡ýY²WØ4Þe«i:2ÒVšºï ø ì÷•@„³Íí’¨n`¯Z³ÏLH` (j3*u ïfëî“òn$Ï­ƒ×µ× 3»3—…±n„Ï &–‹v沑½Öq÷†ã—nø^èr¢ Æ€IN½‘e~TQÕ%¢câ¶g«­Èq#¤šIG¨ÈªÙî;U§ <¯=F#£-úx!š†Á캭_Ųo3TLm Ú 2+m`-SD8‹ ™‡Ý ³Ö 1 Óíð.ª²x㸣ÂYIöÉ’Û™—t¯ RHëÌÔAšL#ŽK!ózù'²­´ŠÆÙ«¹×©09@—Ìž«¢Ð€¦Š"°g*Xh7Š#8¢p¿¶{3Gt#º'ÚT ”ˆD¢¥1§<Ž˜Jê?ñF®ÏD4±-dóaÓ\—ì âµ‘<÷öÖD„6/¾˜ŠŸøv!›{˜Þ‹Q¤r58Ž 9\Ö ¥VC/D'šÒ­¬-©g92ÃÅ@K üf4à…G æ æ¡½W—ÜÚH½8ö:mnC¦ùÕZrŠD´áÁ]É—K á|gêTÁŽú…쉠g:*ßa#‡ZÑ´§“¶5ógœ>ÚV:“¶v#‡¶®Íe²sÛS¾Æ¸±‡Ä ´+–Y­Rý—œáÞªÃô%Göï1¢3õY&áÞ[Å”rcQdXé Øs5ųÈK7‚ ?Í쎵Cfe7¿(ÿ©£SÍY’{öÝ'¤„²™_%ó2ý?‡}5.D³ 8+,6R9²tMõÚu¤× ÑLjA±="›ÁÛ^3”6ƒ7d17“03f“ãÖ¿¯Sï5€43R…^ºÓôã5eéøD-.hŒu™<š¨DçŸñÇhLbΕ#‹C^‚¢)EÑp±¬¹ÿ©I÷|˜.çÞF w-ŠÈ ·‹YBÖ²©…:äJÙÇö±ÌÆ5su¶ ÷ß…ðYpö|9æ[ˆù1_ʲ8£™ùRB¤^2;Êòx… Ê”Çõ¬.'µm¢g :®‘ƒI*GŽ™hkûXbÈòÍuÈË ÙK v.WéË…@…vß\…Û½ÑÊ ÅO½1‹ºCëÏÎ_QhãBúÚ€‘cNj7¬aãÔ ì1Le.JYÙH½RÀ[§9K=äÏ !ÑzõÁRηmâvv·ÒøŒCêS[jWÙ€Þ;2Ê’3"GrÙi4È÷ lHös;"&jò©C9Ž´°E¼F‘nKžù¾ÞûFô}m'®6R9²¦ÇùGcZkÂÕ·(Û¬‹£Ô}Ä ï/*ÓϪ##xŠL“*÷:CT[«’i ÐÊ( · Aþ3~xl)»uß¶ ¹K&Mæ¤1€º‘µÕ=vUÉÜL¾§º‚ÆD¤ËtSváQצ»À³çZ—s«yÜz¤oœ€Ni[kå«›ºŽ ê”ÏB2Ï&ô)ÆGzM ªämSa[8&tgXŽäÃ3Ȥ‘õ"3 g²ÃzèaÑÔËr“NçòøüQÙüë÷\W+ܲD¶5œév”zE£SSc®TÐ΃()²m/æJ):w!½í]Ü–|;i0[ÉÒð]¢u¶ûCKö•mNŠÝÜ€éM¦IRäúi¸-Q‹ÏÃp‰jÈ‘aŒÝf {ië’·žÙ¸÷òà¹LÛKëH>¼´fl!NYƒ+)ô¸Ü ÊEJÃFZÝ~Ç–ha¦Ú\µ ²…©)m6…©åGR©›ðy\*Oc²-c¤ÂØúV¸… =Ðå⌦ÔgšŽsRý5@“•9YÁ3C6’¯Ô¥†ÿj²¤Ë1M7]ˆf«Ò¸cLóå@štúÂÆCn.ý¢›¾-D_˜x Àߣ3Cr¹X¤å€oHqÏr›öK%Gzõ“ç¿\x’K%Æú-+T²k&椑+ÉN¤Ž%¼„Ä8wÊ3&Jò.âžõ~!šöböQˆÝm!GÊÁgF0ëµOí+¨ty›ùYNKö]%`JßT§Öö¿T,šäˆ¤ZK±¦ÜÝÈ’N}ë¾òª/ÅD “ª¢"í:Qé8#*{Б'›S!¿EîºfL±¸­¢$y$å t$·íš[ Û~XùN 󅈿3DŽHå›GÈ:–{‹fæ5 Ú$×L“³†È¶ãÌ‚%ÜÝŽR¡q/ÉZB ÙÍ:ƒ Ùz9gÇìš—)u»fZc” yk›y­F¾!ßirûÒ‘5»®øÕ¬”Ü4_‡ôBd[uó%„Ñ4_Žà°ú| (†–•aK=j¥cÒ] Ê$€ Üº ä¶¥xÏÉt.þZšÆ[B¤†ñŽD¦·éå@ʵÌÐ-ñ&”­Ö&¹N®cµÒtò1%NÁ- ¼ù6]£¸>âê(}£BöÒXzâh l:‚«(ûˆUfð¥™Å·ûð(ܵ4†w¤¿BÚ‘¦ë£µ1´”™Úg|F‚¢_Ùæ|–[TÛæ3B†Y‚[ë1™½Íuwæï×- ¡/™&[•ý¸y¥YW̼IQ]É1AÞ.DJ¯"…â[t#H?ueÌJOÑBêÛÆ¬c#ÛÆœ²™d‰ Í”7Õdlp§Y÷a^PG4óÅâÈG2bˆÄ+Íz}c  1ýzH‘!¦yIòÚ+3‡ºkŽðËÖñ3AàŒÙÌà‰xy!âGÝ&!Û˜Ü^0dM`ÞˆŽÀósù±&o±¬Ùæ}ÈÚ:7¢C]èÈ]²|#0›=äLéIމ–(ë„l}3Z¨¾È‡´à½ºŠ ’ˆÌ¡"Æ•ÍÔ­½xLGH;.L¶¨+Z—•‚ü;í‡É¤¡ËJq¤—+XÁdcŒSÞ³L/3íšn2ÖoæáwäbôØwWŠXüí_½‹ƒIúKÚ)¤ð²®Õ©ýwÄ,[éDî£ðèÇŒ§ ëÞBĽPØVÄé1ɱGй¹²'!ïj\£Óõ…ùpg@5ŸTœƒ¢¹â;M^R„¿ÇxþÒ‰©°Em\G(®Þû({î ãÍÇbææ\òœ~˜‘y¯&ŽÀ&pcÏÁ–‚6Éi×µ"Ž2 zÒ¥ÏØÖâ§pÄIJ££Jo놨¿xˆÖJòÀÈm&;Ú°‘®«æU0´«‚™nÈ}mÀH£ÙÝÄ/¼?‰†uò*®Ý›)7äÕ Nuæb÷ äû8 ¯'ŸÐÔj±/ä>ŠO äÔ0#Ÿ”ì®Þ:„™¿am q!¯Fq# ÆgZUpaœÊÄ mìÍÂÃpÃç4r§¾x˜i·±÷¢ã¦hº3ÛFî£øm‰YвlCrqa󘔈ˆ·ã•ä>jï$D$TÑVw„?„s‚­^r ÷QÜØVö-IRN^p*8•‹-ø2~úùïûNR·¡Åæª. ÷¥0Ù²%yF#úK@ŽÈ}”½|ÅÙÖ$KHâ òá¼0ךåÖpÄ9꼉W²rr&ÝÑp¶T%8•ä>ŠO #XûT^_»†C1 •úÅ]©©^È«Aä$k{d;•ò D0üôZS&¢íÅvCî£øJÕü`à7•oÔ˜N…êà dÞ?Û }Zà‘è“1?Ë’‡w¡˜ß6š…#¯FmYbü~)>dØɶdÕïÎÉ/ä>j? ±IŒÏûIPû°[t©dÇëBî£øuÅî‚YzŠ®XJZÂÁ/d‹ð–r¤Ù¦“ÇJSÉˈ!23 ¨¥ÃÕéæÂ|ê@î£(§WŒdQ{EV)¨Rc&(*çšì8û(m”fêyŸ—Ê첉áuÌ;·|¢¿à@î£8O!r‡Ó0íµ‘Á&º"ãS ^ÈëA|%Ü?·Oñ›ÜëÌ—QB¸òbæÜIS0­Á¬Þ…ÖüÔ0lê ¹Ú‚þ ó|E̶å›3üÜr%#›³-ÝJΖWcÖßKiöØ…ÜGI1ÏÆó“ØòHL#hY‰èkç4ûÜ ¹{äÝLY³ `ÎÜ6å²DûØ ¹ºXÆ_¶š¤OVêÆqp¼ä6ˆÖ-ÑuíÒª2 Ø·ð}ÀmªŒë¢7à>ˆ«ÿÓ’ø%¸êÚy狞y¶ ƒ4r¤5+ÔxúuÃw‘8!‘Ò>ªÜ¤AÙ£ȶ6í(¼½]ÿ¼ÓËJ0uÁ’%ÍGå­°±R¸‡YÝá@“SjÃ7<ù6ªgaQQíj ä¹â“2Y¸ ;§º2¡3…ÍT šÀr%&RÌo7l™f‚˜lŒ6",SõF•VÀ…ÜÑREÆ'~-NÝ×õgçyíÛãÜéÓ˜2¶Fè‡}•µ#õ¿ ¹ ²ç ÇÙ8n(R‘Ýe„R’±ú ¹Ò!³š3Ì•!;ÌÒÄ'Š Vâä>j[”«C˜¸‰(fêþq©Ýû(éLƒ¤IªQÇ+ÒJ‘–õ^oÀ}ìùBK¹*­qñO:}ªÔ¤tÁüB^Ò4Ej.¥³Ú‹MÇE穪At ÷Q’–3 °Kê[©4ÓÅ.«<Èm¹udÊŽÙû~=º‰ÅâWõ@^’b *‚´¢¤OS´fM‰•ž;‘Û ò´msQ’4–+€þZ¥*Xyž¹rnD.63ÓÉÆ8`Ð¥FnÁ"³­úú5ÌsW…Ã}‘ó@î£ôF<Éð:ë’a³L&8¦›li+ßv"÷AÛ’5IŒH­x›+xZYSîG¾!÷Q2d˜@6˜íkþ%sÑ+•?1³xÙÜGÉp´{Œ–b0xØd¤™-Σ˜r ÷QÒ²â>AZV³c‹h m)x¦hÝ:òj”xRÞš•u;J¦AñŠÂb-¼€Wc¶fŒ;rpoFàÎAwŽxÙ©Þû(iGIò¥Ð1³Øm—Tš\»j:Ã…ÜG‰q:Cð{­Ëð?ÆS"ã(•¶ì…ÜGi[6ºÂZôêZ@à-5ïx %Òv!÷Qš§lÆ+Üã»P¦¹â»… íSX‘í@nƒÄ· µsØ Ü…åBqµB²„&ÎÜi’èdƒç!ï @kn.ÀYj®¢ ¹Ò‘“èb ÆC”¤9TúÓ/ä";w€à®RE4ÝÏÇ‘œ#b¬Ž¼E…¶ÑKbŒSu^å[éʯ´ºgùÜÉÿÀb…©äjµ;m¼$ºrãÒrµíFšÔ$—ȰÌäp(Q®¬ÜGÉçͰ6¤Uݵ8L"¡æïôê ¹RüÔýà!óhƒ#]gæòˆ¸Fò ¹’ã¨0ªÙ³kÿÖö D›É|`pá©Òt·)Œ [ôÊŽÍnCdÚXU:“¥^œ³ÑWSŠ‹2k*xCî£ö“²Y¼×c²£ž;üAŒymä6Hê_fÔuîÚªiìØ¶ßdÞõF^Ú‚›F¾a ¸ÇsÁÚ›E»ûž”é}2»‚çUNj]б"¿Xø ¹Ù*;mä·ù#ë`!Ó('WÙoÀ«Aú,y*’Š@[Ág›(³dÏéˆr¥œ]L´;½k…Y¬õP=y¨ä6HŒŸµk,`//DÔ R~…ùìMºû(ÅÍ"}W]e,ÍCÌŒmRÐñ„rïÙM2S•ÒZdWáÉÆ¡7r´=«æßa®¼<«æE¯Íý¨3r ÷QÜרA †ª¢ª¤ð³Uß¶èÖ…ÜG)Þd—'kÏžƒ3mqÌŠÊ< Fr$Õ{J+‹BÌ)Š j;‘W£¶RcAÙîw[Bg Ä@Õiñ†ÜGm×zŸ2Š»Öí´°Ý<ÌË8û(¶´ÆÚMÐvÅL.A{!׈Y`šI¾öSúL–ÚËÌ‚p¤±ÂYó{ VèÕ4j”1âIo–`w"÷Q:û,HF¡%òÒÁ%¥]ܹ¯ä>J6-mêQ¦¢jxcÛi*î­¸Bº!÷ARE訄—š‚È•H%ﱂsÇ¿ïC¸…2ý¦P,ë..Öh+ ž…h™ô'r$Óˆ‡j¸´YyŠ®[¦Y¿·1;;¡^¯ìîeìZžC¿÷A;bn „ˆÃ,%ÒÇtψR7r¤y¦æl7+lV‹Ö³¿¤3ÞxÓ…ÜGéÜ[ýRXsʼS˜ tîÔ+;ä¹Ò²2Þð”Òy5Úbž#z W¢îz ·A2@‚ü‰ÅÝá…•ÂúîüŽ£:‘û¨í_£ƒZUn,0k2º{/*ëÞÈ}”–_i±úÇe¡¨À¡OÁ}†yCî£$AX2nÄá)Ùz"˜;Gi|rÅw 4Ìa SðC!§KUw-#n·åx5hT-ô5£'ÌTzBz« ÇÍÄ*´r¥¥£k†DYdNfÓ˼’úÈ’ r¥ôˆÅõªOOÑÒ#p­Öt¡)£ éC.VÏ|p"B'‡ÙK>Y{z<ÿí»÷_¢K‘õ¯¹ÚÒ|û‹2ït.>èÙ$È-O´ü:Õ&~éÊ‰ŽŒ“Ìò –bmŸTö®‹ˆ³ßˆDÆ T ¬.^BªQfY——¼©¦4鋊ã=Ï(Jß‘¾¬¦þÉóÜèœÂÅ^*Ç%ÁœèõðBl›jzÕMÕ¯b^û&dúZ(<~ñ­oúÔ¢[ ®U-»åS')±znóóFÖ%séhŽœÈ “&\äJb|±v/ƒU*ë} ÛÊ i8•°ƒJãøƒ“É.ª1ª(ÏNz|EEì¤ û*ë·³¹P¿ ¬„©«T¸8/„Ö^a¸?¢JìB Kä|»ÊÒß‚ßÓB{3ìyq€³àTÄNª°os=ò ÁC+ˆÝ}2Ã*½Á]c;GeòÆœêßð“DÔ‚¯«qØw¬Eçâ—æWŠp[Jˆ®Å`¹¦Rƒ·šrdìëIo16Sª]×fò ûvý› ;¨8ìÑ+ \k±„‡ûéGlçó–‚ûɧ4®Qð«ìJ56æÉQ’pßDa_wce'‹'ãàTÇ”¾¯ ºÜAÁ Xº¸o§¾*Bp1Ðm‹‡N v«èÙšÿÛçU!Lác{“ÿüGø0POÌ‚uÿÖÞ» w:Q÷Ö[›¨Œ?¯À½Ûn׳ž­ú§‘ÍG{G™¦:K%ÜDU÷6Qº“=ô‹"®° {¸çø¢omdôƒ{®,%[¥{[&‹"Â\X¶¥—øÜH®WÍté‹ô3z=1‰ª3×v{Ð/¤ïk oL}±3()Á«}¹ùu¹«À¤M¯B¹‰8Œ¼| ær(^£õ íêR»zwm*b'UÙ•|¹ŒH|µMòœ×„~â2FÖµ>vó×b1~Ô\+Û 5N‹ûÚ-¹„]ùvï8»©¢—“0»k²çrúp%³Ûñ‹íÓõB”Ϻ® ¿0vcé/¬™'ö)*b'U¸Z›©`Ñm#R%CDð\µv"a‡i—‹{Yidy-è†DA7wVmªé-k7Ç=^oŒ®ü˜½Èò?(MÄBª¾e5 ½ºë¦ÞnSqÜ£Ïײ«ƒÕ€Xêñø¦Ê`Mëj%Ç¡ùQú‹‰ŠØIÅq f1wßÉÒs¥Xõ›‹ðšŠØIÅq« 9ˆæ”a£˜+Æv¨‚†C7[¯úºN$ì â0íuÀ Å…àî±»û»èï¦1è WM7Ò3ó·zk+Ô™œà€6±“ªÆ?GÐÒÕiB´ø›UÜ(6rzHTQ§hž¡’⨆_2Ù®õ~ÌÒžGøt£ î·,Ã?p+¡êŒëÐÆï $$k m÷½nlªcwûJØA¥qßÁh,²VšwÉŽìgêˆbS,צ0æËŒ‡øÆIEªBd–XCó°ÑO˜Síæ›ªìBläθ\ÞMÓ\FÜc6I0µý>]Y`ÇZË£‘­•³JÊ\]86±“*^®wð8›¸éµ«Kâå-T‘ºÚ»JØAUý"Þ[[N(p%y`ø(,¾*'ɶf¸g#eW~¬w¹#Å úC \j¥pŠù!ºd¡nŠR™ó!~`bÓlÇða»²N/h~û“HÛyJ*;n™JÛEÎ7‘A' G=Zö!gßz,CïÓ5°ZX.ygP5f¶ Æ^ôqS;©ÒîjAzFy/“ E¤:½]ãN´sŠ7Õu3ó¡)G±ÆÓû²ÿ¨Ý¦ã‹&è~ØFÔËùEÓ‡=Þ¢µê<ÄA¼lTo( ñé1“.×U®uO¯v¥¶oªìIò›Šã¯6+IºVæÓã@ð_ÒYœ¼ÛW`Ë6Ýî:h4êñ¼?À쵌™G8ñ“U‘òøc‚¯Vsí™êÉ °«.DÁWÖG)(:N*GZôd·âޒм½o1m9ÌzÛ7±“ÊoýãJ\”Üdˆ  cIêulì¦"vRåö§ðÈ¢öÏpùô<È¢€#äËæWÏp7‰yqƒ‹Ê†Ér rzyqDÅéÕ.[UÛ嵜JãôƒA–¦»%Ù®>¸[rS;©8N•—áQËû½Xõ*íI›ŠØIÅqúE^J¸DWÁí‚È€­?kSuo>¼©.1øàc¹s¨‡ðð(sµÆ‘ãS•¾ÓÔáeÌõµ$"bQÝÍZ9A™®9””ùô48x¥L"{áù\ý~ô¦ú…lš¾‹Ó?ÖmÛ˜I±LÍ…³4Ù0¢£s²/ß•Võ©\gÝÜK臇ž–Uh ´ñ¦îó½»bßI)øhâiºè¸úëmªá½)6U»š¥e5£®ì`¼‡+¬“kzE5¼‰Û¦â8IŸN¹Åö¤/GÎUXì¢ÚõÉ6ÇI*&=ß{m©‚½5ùòžA›jxg¡Muuäzp¦—Ü­}òºApÙÛNïÕˆ¼á©¬Ö°ÑI%ì Úµ˜­5ã(÷úéj3+Ùݹ,iI§:‘å--E5öŽu*Ž{¼³zŠ‘èÄÕõÈû¶OñÓ9³YììXëÅãºzzˆ ËÒè•}ÈÃ:ƒ—´4³ÒÏæóD«“OHÀ{FîXýî@QUn¶ûÔY$•°ƒJãHÕX$E´Ù¬ª ›§jÌ»oS;©8Nº 5Ø4­dÙÁ¢ÓøÅ,ëTÃû¿o*Ž{|pÌÊ£ãÅCüuðºZçaËxùÔµËZ•˜Tª–ùÿ‹Ê†ý f­..\'º=º®c=(T"}®‰,Ùù³óüv.Î÷ëµ™¹ÝÖô4U«}UÇVjE1úùŸI/„Y§È¶ ¯¤Säó½¢!tÙ(ýRe;€êÑh(–£vUz½¨¼fìEU¯˜ ¼&çæTvžõk2÷bÜîv§vPiÜÖ“¦M~K°Û]FžñúŠªyCÕMUÂnP ÿkëNï­ØYÈzJ=©ˆT§_ä=3º=«Ä|é•¢vPiœøkÐý½Í* ˶òŠêÊævª+›Û"$¦›UÏÇ8s5-íÕ»am*b'Çm? «E{sª¬™¨^Óí¤"vR]•çë Ï•)µùmo˧Ëä€;(Ü6”RÛí&ê7~ºù>Ù*Ô(»†%ÿlMÛ®IªÎª/›JãD•íRwAÍØ Jd^ÁÜçUT-ìd¢Ò¸ÇËQ0_‹=&)êcfèÓƒ@êY‹ÐGãéC»Â#ÞCjSyÛ¡‹ŠãÊ.º^ÚX—{ †¦ë¥WgÌMEì¤â¸}Þ-7B5ùy·r$õºˆ¾©ª_DßT÷èÓÇæÏe>ÊoZÆcžKù!èÛŽëjٺƃƒû…­. `@cÑæœ½‹àFZ~|ñjTôùIa³Û4Xˆì%…'^È.{g8•°ƒJã¶ÝèÓ%¾Øp(^ fE$è Ò°Ç»³õS‚ùtkeq%«á_Ö©ìTo×Jî¼§"vR]­_,„e'n°4³E°Æ½¦è&ºŠ“:‡q5P8#J¢†R²*<îÂtUW‰»‹ª^ÅyêŸìF®ë<üP@èÍ}ÏzŽê;ŒsŸžüüó¿æ’FH‹e²ÅÑd¼:’C¼úL¼AÕLŒ;rR¥Dåñ¢räFÕ’=DJÈIåoqQïõÇM¦ê3ÌY„1ºN0ÚtK5ü¾ùê¡ãôóu-å´:4×s$£P!Ë î¯yœÛò­Ýï»þÕîÿ{…Í«Á*³’Þÿö—ßüæë¯¾ùåÏ~òô‡¿ÿ‡ÿóŸý$þ‡ýš08—'ãj•›Þ•~¾Y„}ëo‹8øKõ‹&ΚÒÖb\ø[¼Ì|9vO Úè¥A”¶1±Í~ ¦jj¥×M\j {tÖ\éc\H£4ÉMÔ¬‘¢™åÃM÷ás'¯, pÒ5“øT±¡¯À!‘ý-€°æ{Ž. :=€°!æ.Bë^þÓ2A×?“0»š©¥Éæç•½¥ a±Ûj©ìd®ßµžˆÖ›™À•–ÉŠí á?ÙË_Ç—O#ñŠFVS˜…óÀeõG´îèæ"ÉUÅñÑi›¹ˆY_œ¸†i”Õ5v!‘srÐiH×gÀÇ9—KC + 5…0·¯¢”&½±ØB] 38 ùü8•ïÓX‰¶.9ž4ÊŠÖÕÈ[ëßkÕÆÛZ3Û7 ŽòpÄ¢òè;T±Àg(¼ë±Á;.!1Ë–4feª£¥!ÉR[ÕcØ«ò1 K^Øû0Ÿ6úJ5k•ʘ4…í«¬U ï­éË£!å¹úä”Êüp- kÙõwnqL2¦ ý§¦múâ]ºl±ØG;k¾:ïÑ/ónkžì1Ágg=°Ú]G•^7"©½ÇV‚I¯3x«½45½^ˆ?'«Q0ö±å@·DÕi!ܦ@¢ŽË7"ŸK$ƒ!/h¿ÎkJ㎫EaSˆ€ËØ8F‡9Leê«ëÊ:̨¶`@×y¯VÝ»4ufXÈg=ó¦Øÿ9ÉCsJ0 !;+²dž'C@’TãäUŽÊ&R©=ó±™ýO°M4Õ¼îÅ[c/¤sòÔ“Ç:»C¹BY-®\&XÈÐ~DXÞ”ÜÁH‹Y|];« ©Üèˆ`ñmTº­ î%ç·tÑøîbÏ™‰y Æ×p4;¸;c‡¡‹÷1ï:®}G,t»ª/¢]•x4ÿucˆ‹!—ý>)‘øšÁ¼ê¬¹0t;Ó)À¢u5ÞF²ùêjDSô‚, ÎÕŠÏ{-än=ûÒXô46rlþ?{B•Œ¾ô86cO?_ø‡¢kiqA÷꺒(Üšªþ>j‰U½ô2#ÔüÍUq»’•><ã^Rs\KtçWÕ‡¹åÌN‹þi2Ô2ä¶· ?T‡RË%óÐVý`ÉÜãS‚—Cmj”™ËBž3 ƒNÍùÝp%)º[å´³æbK¬àYGvfg2¨Óƒ1Í~:A(/”â)Þ‹Gs%ƒ„X &ìçÀ.(œsÎpòTQ´æ“—\µùlA»+©ˆ ²Ù„$–áEx§ ©êµ:ü×SRð,k¶k’2çškÎmŽå¨X%_‡Êܧ˜ÙБI'dh  Žš½Šª$°VÊ­ÁËI ~ΓJÊàTS!ÃsL ¡\PudJÄI0Âb1¾ Ÿb"Jñš iBö R¿¼ ý%D^Xzd?'é9µé˜[êbåu+2¢¤âÏ¥tíZä1Û8ÓY¹Þâ,YI£fÓï\+ëû€ë‡ËCA{A’ËVtÇ']‚Ü­°P=AŽjÛJî£o™±Þ?òâôOM>‡!éÁ9,ͤûùP1Å»‹DÓM¢nð›öj‚4ªý³é¼æðXH“š¬B ^ZÏÄ—¿L/.¬ì§St'µ”]è•BI”ÔÁmM¸åéÛ܈'UvfA.7§"·ÀÉí|l³zÍæ;•´P˜BæÄšõƒb?½3§lf ágù)ºBe™öözh@ÏQª`hÂ_õô}“å°"e²&p©êVømý&gkVévÞÞµ},9f Ó¦A¦2fýÍýÁ뵎˜°ËÒ_ÄD6ff‘,nŽ>$ºs߯f¢ÑôúäË[qÏ”æ‡'œµ%¡‡ÐBàþUpJãÎl]mIòZG¾V]Á{-Ù()ÿ¶™^÷½fS ”´kÞ,¢wöʪÅ!ºð¨e¥Úéb¾ÕH¨³Ðu ¢gýÜ­%ÍÌdÞ IlR5«÷vË…m1pŸ“IMȤ谦UÉêÎGo6‡Î¹HÚm9¢P#ˆeÄ1ÙGHf?ÛQØF È“êò—Áǰ úUZ¥~™Ze|–¸Wç…ËuZãVM?«X¸ü VO;Ü{-"¬±·ÆPÃ]hƒëjª¤ºzïTØ4-P³Zˆ©½gŸÔõ¬•¬µëÊÏ2_˜Í2+¢þuM­0»·>^ûd¨ñ49 ªNÖ)G•Kïüý¤ž7õh^W ÛÛHû_ˆ:Jx›,kNl-gÒÔg£$½5¯+¼cdM°yOxh1Xíó¨º"†tElÚðÇX¯’àÍ"¬¿ƒį³Ö°Zgzwë–ùY‰ë$B•×Wš”± ðê!ÐyD”ihjXP÷¸GÑÆœãUª‹uû¶â(¸7®•ÒEú2ýÉØBªî¨-°´q+¤VFÜ&‘& îÔªßGT«à\¼Kú¡H‡¶î„+ìÌS¢‚Ó»¯wÑ|%ïØ6jWPGʽõ>¥1R¢[+‰ì"þé)øëÌÈFƒI,ÐzÕJ„œ tåüFoò¯•%¹@سÅht^Ñ,²¨øµš§•À®“©k ÎÂîqîã/YIߪ9iÇ¡³GS·ÝñÌë#ó§»÷¸ÙcFb2E©jL:Çà͸´ßÐŒ„mgŠøÀœCÝ‚¦,ú ö–­ ·ºuÏiîö 1ª›ªkà!6­ÚF„uÞPëFo·9U›Jº«Ù¾U•P]D´,$±ñ׉ 2Y¤enÄ:»E¹Øí9Öž!UyÎðãV;5GöY›´£IÃÐgóp¾ÒØKqá…nyØéè@NM`·é®ÞRŒ®÷'kÜ-yþ÷d½9¹­Ùwen››xû#mÄ0ì±Ù»ìyÕY+ý;¨Ü†J»{´Ýuª°Ï ú€Ð™…þAl¬>8¶–øJ4îÅ7 ô[k¾[|³%k0ý»—jn½5fÙ-àXáMˆ%'3oߣ.{¹¢MHG‰ªiW~ØèAc­™¶!ê å¡r”žÄá=K¬?KÒ¯Gµj¢É‡7tî–Ù2eï/½Í[rm­4GðÖ–ÅìɺޘB5üâEìj;¼} º [×»¦ðd]7’Ù\xsõál$Š.̦PEœ8ˆkDe_eYÖä…©Ÿª?~vƒ"0Ò(¡Ù®©2¶_S»f• vc*‚ ²¬ðTobËʰ¤öP±E¦›âQÁ-¾¦úIÙIL˾M"æOßÙ}¼IuŽÆÜ-‚u"x§‰ÀJçåú&Ýõ³­>–Š@ñ—‹5 Æw¨Ení*·üvùF°ú–¦âZ%^“Šè:0ñº^ob±Î­8x\ õªš)éH1Û†¹›ŠúИ÷¨DT‡²)ǯ/ÁÛè•A¨†Qó?u‰ôQìî_M2L`š›Ã­L^OQ;Üۼɒ}“. Ù<§©«Voö~eê2ff…zâ°0Oµ 8 )*«5ï'Ï©Ù!´6ª44œ!± y-’m1­À«NÕ%¡ÊCUÝömÐÔÜŸP8\®/>x²ßŸ•¯2$†ªj¸™¢z!ûï‡Î}&Vc’Û/ÂÙ2(õi-ÅÙx‡¦Ðÿ!IS¬";l¨=»±²Gò1òîrßuAO^›*ª"pM†âOnÌVèQål”‰¶4$FÑZ´¨e&§ñA5$¤™‹ŠÐÖ±Óó.ÖGp nÅ}!™p(5pMW««9ìš/ŠUë׫9•uWßó­çÕpÝ`²tZSߤÛÛfâ·29'V,—®(/7J§ÝŠR`xf•Âæ:ÀÁ†·=ËÇ.Öi@Å’ÍŸ¬ƒÍx_¡'ÔZ~ЋaüÔ–Ô0àœ07÷ò5oŠešd·¼g³xEW^ ö©îáÆ5Õ'»sãd.C<ÒÈHIpÏkR[…tø¹Øb Í™ØìÜ“„P2dº5Oár÷¦T.îÚ64Qq‡×χYa–\Îõ§ÿ$3Àdëojóâðõ ƒ63ö%[˦lÎjãsÊUÌê4g=Ϻ»Y·i ™RËÈ7[¡ÿ NõÔ4#:"jŸq‰›òSðxŽéX¶Q‹$¬µsOV‹Û¢mxºë ¡ÀO`©…ˆyÀÍ_5ÊšQ"‚ÏèTì,@î$æÚA¼ˆ –g´Ì滛,*}B !Uˆ –»ÈÝkÍÈ…ˆËq¿^dMI†„µ:톸nÊEÌ;Šk¾ûòèMKÜq`ú”!ÖŠ 쫉ŒWBî´á7d*xÖ‘iü 4ÃÅ‚ ·»'sˆÝÃce0® lˆ8—éJÆjš²gþô}bs–Âö›é—^†ÎT; [µ{²8°iÕÕìhr•3£0BIT=!+Ç¢²lë—ÌÔ¡¿Y—£,߄يÃ#1‘ /k­®Ãt/î’\ƒaºGÐׂEöæ’3žzäN‰´«B}XÖnµÖÙ5Ò¹±ýGÕEùZ2Ƴ ‚jÀΠÂô1A²"Ÿ»“5Ü–„a_ážÀë¢Ö(%Z<–Ę ›ÖãøXÒHpO.œ‡–Î:]~Á'ÔÔôDif C»8ö$;Ú»…Œeà™ç"{}刌UÛ¯!û/S`pÂ(òp£.™Xlj]jý´:… -Ü&ÀGAäEýR¤ dðÊÞ‰!k9Ó~_dͬå,nlVë¨ ÖGW4TmòTuµMÍŠc’Âo攓ˆe¢)VÌ\R.P)3,BèbªæöÆÔ„æÎA;_ëÝ _²ùêzà ÝzxÂn .£ŠQ`sÁP=h¶YágêÓ…øâȶß(pÔ7ÙTŠã­…¼yidk4³Z-n¢@{°0ÔB²#Õt¦…—…Hž?JD [ûSÙ¦U¦2ÊçÂ7 ÈЬA²HÀ¨®k vl––a"9Â0q¼Þ¸(´…½–ñ Sj¤]¡x2_‹Ö8ƒpO’gtj†æ{dè½_q´CRhÙŒFß TŸ ,  dÒýš+3Ñ)ºúÿ&4'‡'-°×45·X+ë㚌´RÈ:~S2“¦à ¨Oš,™ÙWoióì2594~èmöY9tî‹ö0å¬ÇL34jlQÜV.Ãp(KàƒÙeêȹÜyÂÃà)Ë0¬©?ËÉ°Þ ê‘ePfèáø†ì>ÏlF|{òˆ«ÍE£iAÙ —§Å;£›³JByïÃ43Ïê~¦øXnîÙWbí»Â‰ß*(Ë4;xT(^I0ð(Q¤©Vˆ|PÍ„gÉË¥)ŽÖŒ#Ør !lk*YÝ>B˜;ê¦,UF¤X ‹š”ÞèÙñppUŠ 9íÛu6QV˪ꆕ­öä*åî-ؘïªîîVˆÑui–;‹QUÿPœéš‰(‡`&UfÎn……xêΑa ¯ÂlÁu»ÅàM³Y]³¤u+fŒIÕs:#ؘx2¸Üå,;$ßÈÇM¹5s»†–w ¿5!ÕSÿ˜£Ãˈ•vë_e$Á=P…lº$“v:rDé[ „º¹Ö6’T×LùÚØO2Fÿ¥kí2…†$+¤í»+*Ï×nýñùdjaØ•Ý(q€÷mjâë_ ç"XBöïæèÜé€-©½àTëtÌ/\;ÃEVÚTße9´N_]÷¼×쵘aìûršÚÖöcz¡îäµl[¤ó^ýqÓ>zŠÞPo:}Žéƒ€²Ç)•~`3“L÷ÓŽœjp\Ë‘eáUW©öâÈØ¥òê™X…i¦Ö¦‰ÎŸ`™†$™AI_~d`N=†Š'NyTµ½$ņn…ôÄ2½ªUÅ­NO‘{Yºr’^^‡s‹Ìö:˜™œ¹W©~˦±¯–•R.=Éðæ,†/š¢ô6ÉŒ]2qZLȸa;ÕG±‹X–„Ö"Y£UüšTN¯ùgõëjtkéœçªÄto{o¦´Êv‘tN!R8°•˜ž6c5IáâÕÂðŸŠø°ÓXá¹”‰'pØ¿jŒ#V_¹ª€¬úrØ«FF‹«{‘gÎ5‘{ÝgËT ªošåCdªDËž²„Ü5^Å®òä«>Àè°ÝäÛi""%ÏX1Í[HEePÁB§Ýƒtq»yެLþxã…¶ó‘|Í4‡vØr6[Óu®¢ÜÞîM¡÷Nƒ_[Û“™ª–QQ|›[x)xò‡…©ÔEÊ’Ü,%î TõE M5× [0}Ù¸Å[ÓƒÓæ6Óüî;m áÚÂŒ[,ÅÎoGW·èîñÍ]®%Ö©fÅN ±$]»5?£ô «Hjóî¾Y$é2¾žïçÈÊ ž)b>O ÷÷+mWÍRèÈÉ$‘)"¶+ßg§«fõr7ŸÃy1!­,ÃÆ»ÈY%--ïЄbvµ6¡IdÚ¥˜Ì3#v¢ŠáFùdƒØùTJa K"°\wk(ñ”š29öØBq½ ;‡å¿µ¿èÉÓâ‚ÄEsOv¦¬¦„ÑÅŸB¹ä93–öÜé,‘¢:î*mR)ùyÚ³D-*&m!]œi.jóde–ä·ŸL6 Ô6¿Ó0쇑q‰pŒ c«@ƒ|‘¯-{-ƒ>ì}¤¼¾°•A 7`\Cd?QóQºˆ´6VWð#Ð3dÎ$é´t0XŠ#Ood¶è”^Ü9/ûNd–•• Q6$o ¨-±CzUªKLJ\ûjÉY† mf¸uOÃîHbvv¼„·g,$ö<•…ííÂ] ÕlçÂÁÖe>u2¿1&iÿÓ½K~C ß;žví\&|ª~ßÒv+jø¶]OTÎa¶j•%mˆ6"3Z¿dÓ=›™™ÝI²î¥¢Á¯Ð©¦Ëf˜EË aBUó_¯JÞR ÒsPpQ!ÇWØom¥×®L”¿¸‚•¦Mã¡^Û-fÓxˆ¢Hek;P‰ö‚[mÓ5^(° % É»tã•ê_óËF5ÚˆÚHYöäFy¶Šÿó£¸ì¸þ».—h´ÿ“ÿøÎ8ü±Û}tÑ…¼þýñÝþ}§ð÷ó'¼ú‚ï~µm`Ä^–è°«—¯¯\þ;üð¼ZdÕ°ìáWE_re‘”u.ìækÜ·\±E′bêAdk‘ Éæ¦ŒÈü9â%åýïpþ²¬µSX*T½ž°ÿÍßøøn#HA p|é *#½Cÿþøn¿Å¦Ð[ú^}6Ê·S­ÿÝ~ìÙ×-ÌaÕÀªC½O·ÍaÚ$.x°~…ºª‘UˆQz‹îZýû#­h+äHÕÊÇüßöß].?íñ¸14Æõ ú7è6…¿£pûÛßÊ­ø÷ø?Š_À²† ]ZÁdߣ÷ùÕ–À-%zqÔüY@²€’]¥°­[˜J U)æýïï®°NAã~Âþ7âã» Q j¡?ƽ]‚Ñoèßßí·p ½¤?àþâo´Øûw÷¡?nS (Œ™K´ÊaïË­TBFÊGü¡J ßWäÅ‹'|Z±—lE.:+= Ì…P-ôðÓêåe‹ÖU{ÁùvïO| ‚^–vƒ÷r†ù 2œ3(òWf–Û´Ì’Í¡%¯[âéÊQÙ‘Á™AÒãQ ÿ®Cäéª]­,Ä´ìEcÀ;¶“gÔýÍ ¹zlÍ#x†+ƒjë”? V üÜÕ"OHê}ÄÁB– HÏØ¶ˆM‹b>§ö€äp§OÇh5ö€Ú¤æÀ·¤'¤$? B³5‡dNõíb§ŸºÊ™òèÇGq·Dföv)¼Áœe€Œœ‡ì™CH¦ëÙ׋¿§Ì>2áïë ›E¶ËóÛ?¾—5¡öË“WwgJ,R¥NÛ‹!úç_È}Ѓ9î-[,?X‹”L8®Y ÛßQÉýS¦±²nÜc]° ;h$Ú”ÏðÇ?â *þCÎ ˜ôAªÝCTÄØ¹$¥?dÓ¬½¹|zoÏÖ–Âb>„ÏûRdï=¤%xœ–?öüa¶ô{ä¸DÊñlš5sˆ“' nà?ÄøAëb»?`† Oxñ ±‘G4"1΀ KxÈ6ìÖ¾Àºª=ĨG¬Ù‘u²&xà´møˆ®†8ÉÅÿeË`Â8Êêa_œívŒ½éC” ŠâòE®³[0œQpaNëÈþ‘Œ|µÁºþõ!_ŸGÔ4û#ÞW?#%ÔèQ®ƒ †êš£\Û%ñ„ÔÇ(ä`b[c*¢"묰ríc–xVvªŒqu6/ÎHyÈóês·+­Áš>FõO hŒùóI+ŸÄ#úÊ!k3ÙÝ3”LxvØ&DzÐCvM²@ŠÛ•‡œã­\/Q?Âp#šÚÕãÄ‹ûgñßÑÛýÓÅzÄÂÍþ5ià^b1ßÓC$hUìÃ**>Ä>‰T­s{Oz˜ÅØžÃc<¾È½#o ‘&¸-4èäþcº ¾vÞ!;„Õà½Cô‘”¦¢y²¦Æ‰ÜG=:°€t1:NmBÎm³[ÚK3}PüO"b'UêyV¬b¨eÝvÆ-(ÒƒìfTr󱓪ï²[,Ž›ySغÿ»yƒeh»ïäEÕÔÃò¢*»“pÁm9V+°êW bET+@ÚJØA¥q¤¼w˜i·¸’èÌ|EDì òül{w–š¸æa?-ÉÖÅ݉ˆT6L3Ê’ºe„ŸcWL+mœÒ+*b'UTå4ü Ý´Bu–éSPÿ<;„êÍÎDäÐEäÃ4 ™•‚š«ïnI[¼ùqtÒ„ÝN¸@ž[ý ÜšãðŠ.Φ×DÄN*Æ9ˆvÙ/áèåa¿‡Ò v{V™ÿ'±“Šãøêhm£B,ƒ;´òÒcÊÅDZ“*ªü*^¾ZIV«­Ç÷¢üÉÊÆø{m*b'•ß_¶)eö}„Øæ”òsp~ÖWDÄN*ÆßS®ì»hŸYøÎêuìãTÄN*Žoõg&ÐU~¶zåú=ÑtxÒ½-s1™D¶ÆÍÒѱ0á5 ±“*Š•´’ †Lû²Ïh:nHŸTÄN*Ž3*ôrA8|R;©8NTÑJ+#W sƒ¢q¯‰ :i8J+=©Tì]ƒ[ĦIAKw*aU(w*a•ÆiÙZžyƒ ëÍ%o·zR;©Â¼ž_¬ü’ÜKXT4´rí§"vR•´=KVÚŒŽ)ïL±z*"v£!t‘d/eo쬦¨KvP­wÞkb'•ï–’Ùõ!ÖíÀ±†KÀÌKv"bQ=\AÝÂåua˵©óu’E%ì  ‡Ç ™eÊäCÍ/ËTΗ½©²Ûì›ÊKršÅ*/™ ðLGd%ŸœÇ^áME줊mûá VÇ6oõÖZ«Z}—K½%•°ƒJãÄ<èD"†k¥ô|6¼Q ;¨4ŽoÙ¹M]ñö%nŸóZ§Ê›o;•Æm9g¾v„•œ³úä#_§S •¾½¨¼“ìOåa6¬H¢Ê‡i¯£mi¸©ˆT­\Ò0³d\(¾S­Æøš*ižD4åÚ©°!i«­ýßã(‹¾"vPq˜f”RQÄ´·³ÇÚß—ÿÌ©âöÅ9•ÆÉ¾’ÚsËKëma7¬{¿S ;¨4N:ûJ†µ›t(k¹ºô‹îŽ/*b'ÇIWañ»SÃÃx͉—^**b'UÏ›‘$V 3¿Ô¤Ö–ØaÎ.Ö½¦"vRUõþ)½‹SÎ1œIÒOY¶ghS;©¢úS˜tRùóë,NVQIçYUºÎ¢¨âq­Êä$ÓíbºÚ^-oljtÒpß­N¬'•DÕȆ ¯‰ˆT-Ul­ZùsláмÑA$ì â0þÜ`è'•­8†~R¿‡MÕ·æàTåÒàc¶¿§Ëñh-ùÌ3;ç+*b'ÇêÛÑò¦®²è!ÇKi­“• Á°ŒŠâÖ媘ûîå@úZn¡¬&,˜˜\©) *C&–C2EâB¬’‹õÂYhæ"Ê(:mís†ÿ÷ÆŠ~ìÒórs;…3 }[©ÓY)B3 æ ¥ÒY˜‡_/•;¦aiò‘'­¶\Õ¸L© y$ßÂ… Ü©¨´¶^(YÖr¶çe#¤ª™­÷µ;Ì)Ý¥FF­µ4YÓÈWŸš½>¨Þ¨r¦SëÃ~âBô‹ê#4=¢A}ÕêTÞj®/>«±‡’mÀ±ÛfR9R“ÛYax@ùk2¤àöÅd6²ã–哺ÖYmÐÞ>³bŒSeT–h¢:Yr~&J— <¤$•í.«Þ[’¾ÑêŽìÝÕTuÃ:Á]Z!ˆjZoŸ¤¸DžèB°R$úk𜫔pÍU¢·±ï̹ª¼A+D«Coê×uΕ#,ƒÉÍlÁ +5fŠYF±{+jv1[SÊ'ëÉÅZš7§²T¢.fTh(g´Øêy#:`™¥˜Ë ¿/š†é¬ÖMõ,ÉÚðê•isTc¿ÃNÿÐF¬Q¡¨ ÷Z¥iU]ÁCRb+Q"ä4ÞCµ[ß¾— A[·éü6P/GjUáËgv¿B*\‚2þhAl¤+À›QÙª…6Sìé¢,,ôÉb×/Š͒6•…×Êf¸(77²å1‡4Ä6R/æm’«±\.ºê…\²«©âç~–!e{í3ŠkÚdu¥ãd´wÍ¢ÉÊœš–tp6‚äT½”ÐL+iâY–PˆžEý£[Å >KHÉn”­C` ¶ ã Nib‘æ§Qß½— À‰ÆÖËŠdZxû-/›©>€6¦’z{»—[Þe@dáb´0l¶ùP8«Í*3ŒÍHåHÙ>&;MV_;v–˶¥7 Æx9ÊïF9’w–PF×P[ÔõZøéx!Zš¨þ¾œÿ§: ÿO–(Õ¤WõàÏsŸ-kOmE„‹¤¸#(UJñüÈ”aÞÍA,*¾}½ü{s+ÕÞ:#s-ÔIˆ¸&Û˜£Yeâ:Ò;K+b]Ù?jýT¤Á”Ñ 2^ˆVŸy2•÷—AŠžÏ4£,à–äÓ(µŠ€—éqLì²¾@Å¥vŠl“ì,¿© wnÏ|÷\ü÷âXOš@>íH ¬ùSòîI 9ë¤0á¦÷í¬ÌøéJe»3UÀnÂeH“ Ê¥Ë8” ;Kn©ÀÌR[J2­®€-t§Šj½²Ôoh”œƒ|iƒYGz iAŽà‹·4î]Ulååɦel@ÊYzž¬.½EÛo±^ÅÝ#u²Å:ŒÛÆ"â/2“‡Ž3º!Y­]&è¼Pxõ heØOXùV)Í=)5Pƒ*†‘‘£È¯µ -u3rÔÇ%5”ñ/Ú{BZØŒúªñ‚9aÎÖ(§oD›É-ô^.aê~V£:¯ŒV DÀæMRúòæNv›GÖØÔXíÐ>GØ¡: ³ÍÝÞáã3jÙÛn)LÓ€zQ|]à,²×ØZô†.É¿‘<ö–²bÑ“ÊcÛPÇF4ƒ%¡Ù<ñå@p%ãP™ƒ•u¬:U;…hW±ØvkJ¨ÛæÌWU»rHkÊ ¦†ÒBø^ë™l.щ é} Žºä¼©Ñ­êÈ#ý¸ôèíU¶Ê5p!#nîaY½‰"»qVc×;L £bßž-B_,µˆÅ}­Ç=d›Qq"b3Vc{=+8Ãr¤EOS\ì‚ÕÒ!Ô¤‘0IÇ‘mOZùä>= _̼´µ’NÃF“™+ZíÓŒêYà›ïBG„)÷u´ÎKå\G!·uÄn’î못³éÞ=D†5¿q—¶­/ÊV2ûå@ 1·¯‚½¡z"Ÿôrm½Í)J9ðxã{ÈJørPuÊÑ¿Œj6BªF'!¼*dXŽàlº°l:Õøbd˜Ùt!b"<Ã`oTBÀÞ|7XSeŽîßlÝ]âFH•/ÊlóÇnU\&öÝ…¾¾Œ2¯sÚ|¬Ú §yˆÕb28!ÔX¼&u<Ôrm¢%®*ñ_ýØ;’wìYªÌÁ§S‘™ÕeBØŠÌ,’&A9g’+ÛGà#;˜1ÀY(]=òÖcÑDÆìõ©ûFÌÔØ^$3y{ í²Ðdú¶˜éU&†#§mWgc«Ý|³'6h¢·OÜ€H'•#ð¾8‚‰]Í;·CctQˆdõÄH£ÉS~!³mhýÜÌ£ÒŸ#1K®Ñ{I®÷ºm¤'÷#gT¢€/ké 4î"Û¬¹œ¸£K³ud¨õ†}_eú­c¡Û·uW?t,ëg0©èQïwe:]J£é«u R×êøÞAL­'Ji Òme4ʻќ‹šd‰Ñ«³S8¹èFúÅE «´ yè!¸6°?/ªICÚŸgH¯ÏÛpå-ëb]¢óœ¨ 3:U!=ÅëäTéòy¸¯e›.D®Úà#•šŸ#5oÁ¡@Ã4¸‡¥2 /d¯3 Ó"¦¼ìT—8ô¦Cý£‰ìh DÜ–VR1GŠß1·RS5ú!-ÿ(ùµŠºc(¿üBz¹˜-û¦Î ù&¼º¾0_:âHÛôÖNvd©q»ô‹ºn×í›d÷!R{ÔŸdLWÇ„Œtiæv…kR8ñ|Mú‡lŽe'NgXÌxy Z‘ S©“±íͰ“ÜŽ#ЯxOÆù€¾†¥§ÿ.Îr!’óì=ÕcuG¨#kÝ Vb…÷Aä˘<«•Šì@‡˜°W9Ѷrê@mW‰†™chªNê] ²ÕáÂ]bÙ×â DôJβ{j„tµ “µÀ½˜k–›nH Ù¶st¬¥æ†‡!§20ÙÐm¤îžWïíž.%®'ĉà„ÅQ†”ƒ7~¥vÓ÷ÙêØŠa¤žDF$` OÝu4³Â;{SHK›žVaŠŒã;ÐÓNàÎèQgb,6iÐM-^…h×±µÈºÂt!9]Ú8B#³60ôìy![Ï6ÍT\Α¶©VJ¡ÕåÛìlÞèˆ~vúô Š#˜(÷“Â Í MÖF°ð҅苚ÿxœh#szÆ—íkóë·]Ff+I!Û—dn¼4˜¢~!õ²ƒz‰ ïŽ-HzVç¹°™\1bC¸—¹3zÕ±8xwÖ%"õ„=»á8¨rNÜ|}¢‚‚¼ühó¥  Í=åa{Ë6vdÔËû\킦´©lMGæF´»2ÅlÝÒÙ‘Q=ñ?w×Ò[w—{^ÈÿÛÝ·ìl’×íç)zÙ hy¯L?€ h'X€4ƒEÊè& J²üø®s‰¬ÊH™à’‹ú©¨¯*+/q9qÂáRV°g{Ð[2ëŽÞ¬ìºþ}-óõHlJÍ“+ü³`G®¯ð L¼«Œ“úZ·Ä³µÝÜ6#dIË;ä²dð`â Ÿ[°½%\Zx–À1 ³ž«l‚°:ûxҫѬyŸá!yÅ”Z°õ­š°r!ÑUmÚÌ»T7óHê|²ƒØ·ó3ìQÀo}K¼GN¾5ܨ켧%Wy"ãH'àþ½Ä¤¿T•cÉvŒ[‘Ôµd\{Òc0ií\“½üÞ¹®å6ÛO®Ë’ñÊuÁlÂ÷醥0÷?·ÀžÌ£¾vàB0®miu¦·â€Dîú’ø™–-¥åC|KàCîÑÒVÙ]yR‰Ô~$ÛÍ–)qÙZÛ’1÷iØÉ@‡mw8 ‡ ¨\Þã A³:]ãL¿%3ç'68‘„ ‡g¾i l]LE¶*’"'ðÇtUHvÚ ý ùÊ=¢é÷ïξ%¶yº‡oT $÷hìÈ|_ Z åŽÞo9˜"‰·˜®…ºä~IÆc³Â‰e¦wD*‡ ‰* Y¤“^´æ{ìÿËVUrÉd%4ç‘Ø¾˜Úšæe“-$°º{\åö†l„*LËcn‰®jöõZò2Ý’+?cÚl-²Ó¾J{àhqJpJ™á.…¤¯m%ÎXò8É8JCöí2ϦuFüx›8ߦ#’3‚NÞ°-96,2B2Û+yªv¥×mu´~œÍ’›™ýî{Ep 1"XÂâ± l‰­º*ϧ\‘oI凋roZâ6êøÒÏš‡] Ol[2Æs¼Õ¥ãó "(†‰w¾Ùˆ·X27´”K–ŒmúïÊZŒë´ò±[¤ëí1lI{ðl匣rìE¡6ó!ñ·–•ÊYècÉ’Ùzҗᯈ£¡ß#ñ c÷š)ë÷GPë+=6Ÿ°™™.÷ß4̪1œÔŠ![rm'á‹ʈT‡ï5£Ië Ãûœ·Ú µShïMwozÃâîym‰Í_u÷¹f wÁ8W{ÃJŹg×*ª–{®ùš¼û-8uxŸ¢¬ƒ;Ü8«[¹ñn |ÐA’Zz7î® ªìØ]¦ï`¯å‘œZ©( ‚ÞÅJ,5ZSGd\UçrÙÖ|ÐòÛ9Å~UU“›£”mYhá—䃖׮Ϋõ–0™ñL8s5¾M­ç_’SKÏM/8þÅôz¬¤%ûß’SKãÔI§‚Ê…@MÅ­¢_z!’¯Cr(é˵âßpÝ"êúX©-› Åh¾Grji”º$Û ö%PÖ-u—•ZµÏuHN%wç’'ß’‡;ýÄÒ—ûbc"þ-9•<+³š'gzïAÃÀñ¢Êý"ìGr(i¬/:%•2B×bTÖJs︇ßÜIäÔò©À„®rÖ }ý:¶»šÉ©¢Û  }Oíté0Aú©zdU ü’œZ>LÜ╘A)­çÅïÄ ˜mFñX*ó%ù¨Å;¡Ä4¹Â¨¸Œ»Õlª’ü „°%§–F©ä½ªÌ ÇW6ò¨i™&¢aÞ’Sˇ·¾¢*d DdÌ/6V^’SËÏtýä©#†|'䜑;:OÜý%9”´àPúiz¤ë’ÅH~ÿÊñnK$väÔÚG%gêè.¬ÌYnRo)JNnÿy^¯§µZÀ$+/•!†÷Xؽ%´<³U8 ô—è„pÒt«E+©`äœJÞG”é†1¥º{,-~êf–æÛšdÑ×Kp*y¯UA$Ž-qÀ,Å¢½=¹«„U Ãí%9•|²-ãrŒm` ¨rw©UÌ#ýNoÙ•›&쉂¯yV H›Ž?mÉ©å=Ò#›]×@£ Q8Ä'¦ p²¼§’ÞL¬ˆÉ$™’à$µ¼ù‹eŒÇ¦i ¢Š£uä۾ɩe;2Û'ºÌ`ÆŽdSPchk~õ·äƒ–\…B –lònš& Ò~!^’SÉNÉÔÏ¡kNíTõs½ûŒN—fûKrjé倕­Š_‹â§¦Ù’`¢â^§IXØGrjyoeÒ(. ¹W®2“X•YNh«ª”z$§–77`¼>´´Š´,Š*|_’SË'[Ù9ï&ž<€$?÷ùÁžGrjiœ²ëÚV<Û+»@SO—8„É­m•¦Ÿ¶CI׊îk8”v/É©µÏ6DhùÆ=ζ6ëq*bò’œZvM={jsy΂Yâª|íÙ³Þ’Zz&Y¶½XRìäs36Úß‚SÇ3¼³F1SHNISດ( SÀ)ðH%¿™xÓÖhÓëçŽ:clïñçJÙ’Z>™ŠKm«&·* `ӟƱtüùº^7¸Š&Lª]ddÚÓê{90%óHN­í×0¾)†c;6„”~{#Ôç%8•<ÊYUGª¨ã\ +f‡Œi\‡àTÚ‘¤bóÐgR—…Žœî´éY¸@_’Sk/þËœõ²#°ø»ÝˆÒbñÃd~IN-/¦ÖJ/Έ60ÃиKÏ~/抗äÔò—SÅ¿O÷ïŠp#ÛÆ>>7ýøGrjÙ–G *ަ¶[•4+—åíç%ðÚ#9µ¼ÝvÅ2Ò>®¤pV0$jùœJÞ—«Î ߨq¹€«2”Qû!9µl˜¤O,;ìV“†YÝï $éKòRñŒ\2vÉœq²1øÝMÞyÿ¶x/É©O£\æê&‰ÃÛg.Ù$û—Äãþ‘œZºÓj¶–éM@EƒkT…?œ§Ú’SËà´—4r°I‡~fŠæä-9µüù£¸eö`} ŽâååšÂGrjùN„3îˆç6wŠ,ZÒ{PÖÙKpêȳU(ˆ¬öl ÉZÕÑNnz[òAÉ6 ™³vôdhΩÏ{ÛŽ-<’SË›±Odké:§NîLö‘ß’Zº“|Ÿû󖈔WÙŽ=†@õœJ6Ÿ›‹Bî=ަº„ûÝ≶äTòÉk®Ý«DX}ÈBÐÅ‹ÌvH%íHEieL ±€ƒÎÄ/ËÜW ÝJ=$§–ï”…âÈîo@&ËU÷LFp5×Cr(ÉݾOƒ‘eð B¹ìÛÂ(îòàÛ¶Crj9¡ –·MRÃóøðG*®‘Ü’SË£í=M‹¤6™Qe)º{wSéï#9µ|'AKÁ¿R¶É26š üpp>’SËÆ­¹2[ð)ÁçPû SöÊÊV=’SËç[“ \ºY‹òtY^JJÇiɯCrjÙ,27åëczgUú z@Sª¢«‚Bg–|ÔÚ&)ÏjÓîp51.©Ä¼-R.Á-9•lœÉy \š;«z¡ÂëÛ¼$§–g“ÀSàð‡ëÚrJÓnzßZ0Ë—äPÚR…ƒ•ŠìwŸž/ƒ9c’Sɶ¶q8›,éBšJ Š±Ÿ$tðKrjm‡ð²ròœemÎÕTªBpêhé.µÜhcDÌ%©q (®hô1,΄Þ#9µ<ØKgGó׊ˆayûáë¼$om–r!]..á È+¼ÍҼɩå‰ÍnK=¹¼{ò5”¸ürw·äƒÒ~¢,ûþy š 0N¶\ê!9”¶D;MÍF8Hò¢ªŽ-9H‡àT²I¢ÔÊe â­å´ÅÞS'ÉKr(i÷o ° R‰-Å`4È"B—ž:É¡´È›Œ¬¸²wWìs/É©µ´È‹ÙŸhØÐI[*ï`<‚‘-8u¼ƒ(½‹ïï©§Èj²ÁbéýçyùNû§ñ»šA÷¨é@ú¦ŒB]‡äÔrd\uÉ-‚‘­€‹,C]0¯ò!9µöAD’¡©»{–%eú{Œs"˜ø%8tvÈØÊW`1ÿFChF`1ÿCp*íØøbb ˜T±qNè!6Eç锜J;A³7R£LgЩ•óMÆòuN%{Ý$cɤ¦Y󾋯ù›În~ûGr*y{ e8Áº\_̦÷zݼ/ý,ÉØEe×p OWùô«ªûz§’G§)v}E UºÖQ·%§ÖÈBüEqñ´ŒÃ€›áà:c¼$§Ö“_gªÂd r/i§_q6#SÉ©åÈHúɨ?[ÅÈÞWØÀM+ï%9”¼8ÄrÚ„n¬†Z‹nÿâ~[¸<’Z;E—+ð„JÁÑ-wá ÇY²-ù åMD%{(«÷i¯c!Š”v¬ü%9t¶‡ÍóÓòÃpeø\`{؈€¾$§–?›“õwÌ™gûΦ6‘Š¿$§–vµI´>Ì6ÁÒ׊8|ÑGâF[ú‘Jb“m%ÉüÄF›ó66eä½$´<‘„¦èQžK:xáEÑ™$§–WîÔu[`Ë»H¥éÃ@µzKN-ÛvUÆŠ/gôBqËîŒÀ2/É©µÇEˆaM»Š'ÆSšÉä-8uöæÏ1éMH_ChOÅpQpLÉ©å$‹L%¨E‘Rc-pJ(RoÉ­íôw;t5|þj[¾Ûºô[r*yÓVÁx9DÜ‹ˆmâ{S»f|oï¶äÔòFÙ]}Ù½õ/ÝP”xÖ°øI_øHN­¯UÝuXÝöfDæ†é^É©åQRò ¶iò¶¤ŒúÝkt­Ü|P±õ¨î>0ZÔ?&DWZþqJÉ©åAR2˜<‘Î×5—×À­N³á‘œZÛ¢e¾z•k\úLMX+›´Lª=’Skçëxz‰œÆùºL•;¾0a§ØÄ–œZ!ãÌêFÚ‰ Šä-TYù[p*9€<ŠSÂ7êÕŒÙ3Á™VÉ©åݤfœ½ârâ؇f„¤`V/É©µ]lúKÙuÆ<¾xÜ¿×Ê>öû!9µ¼/uy‚ÓŒD´iõ®Ça–d@è–œZÞOû[pêì (Ø 7i@æÂãˆå_[pêøyÜÅnFïO¬î‡#¢uÉëKrj9Á˜§MlÿYs¹,™Ö÷ö/¼ü–|Ôò ¨›#O}#Yû2ôä¶>.ñá½$§–3¿Eè×™ãNHýº_Õ0¸7~äÔzŒ-ž@Ñå, + FvÙÐ?ýœZ{Å °[±ä†#³) ï??¨h[*Š0 U˜¨ šD©×ÍP푺¯¿%§Ò6þš9¹´„ñGêÄ+ü5cxKN%oÛEÑj1šŒ¨ù´gžÿ%8•ž¬/wáP›é–/ Iò©½œ‹Ú’S wúÐu$zû@1•û0"±ëíÊdpgÜÃûÇß|úŸ~ÿ¢áΟþþþïý>ý7øw,æ ’@7ö½Ô ¿Å/®ÚÔƒFéòL g¿® Éë*Nã{h^¿’÷Užâõ\¯«Ò§ßþy¾òùé>ÒSJ'_9¢~ ô?|å¿úücÿòãíx"Õö9}ùñ*8HËgJáM¶Ï9}ùõ?þ½ùÌÑXÏ?\pgA%Ú•Ä¿¼í·àu ›F« _ׄà} ¢lbó5¼®‰Ÿß×¼Ÿç/$ ЩÊÓßÇñq”Ê¥Ü0JWbOÁ?=Jûû‘äzÍ«T†Êi=Ç«|¼ç"¦|E£¬{ƒ)SýÔÿ+¦›fÝ7‚_zqz|þãoþõÿá÷ÿú›_ýݧÿøÝ¿|ýݯõwù×`·† ˜UÄçÈ;>ÿœVùõæÄçKœï–Ñ2ë6^2>#ÞuE·kó¬Ú_p÷×9Ü=4E=út‚ Ä´°ÙD‡UsÑõ–ˆž}ö²®ir¥Ñqš–Y%{%Uç"ù@`óT9lßDeû¢VCê,ꌃÞÉ—´LȾ;„U”¤JLQWш]òêŒgN úbÍÒ)¬(wǦ ñA,Bìa×\ÿKK¾E]êeUQ¯'ŒSúRCEþƒkâ´äó˜ñ€¤c8·«Òß«}¡Ûk²KÔô«ݦ«ùj-bS öRgsÚZ$ß/]£<¥{ìÒ—Xz¾&„CÍβ“ú‹‹“£J,¹JFÎÜ’în‡3ÆKZÀúêñªÊw•©P‰| 9~ÊAÔ Ô+zƒ`(Êp«;JØ×x˜Üó‘ ¯|=žNûư–°¿õ?®>ÁF8óùØ«¶f=øìê] £á{vöÒ”ñ¡ ’ ܾ‰†xšü›8\ØD[•%@"·0¯ø4l8—ð(dƒ†]€Ðo§-9çOzàAúˆB€ƒgÅP§ÖKãp‰p4GÂÜ ì„;ž9›ÐÔsćœldT²Ô9Í…è@-ÑðR~y¨©Òì¤ÂXàušPWÓùVBÂ.¯-í57Ø7kß •5J>͆iHмŒGH4JÓ0HôagSæ‹u{sA0 ?®˜ Jõ8ˆ+O½AÒo ¶&ç[ê6ƒ=00Z¦s($E’ .e«sdHZŽuM½ê¥®¥þ#>Ú—GAŒEX¹9U5Ÿ¢û¡Ê¿zJØAðF—,& Ì9oK;Ø8ô%Œ/“ÿ–°×e†FØ.{ôàιÀ°CoiñŒŸØÇŸA?ÊnÍëŠÇ˜­´c±˜ª`Çp¡©ðVXMÝ=Øøâ¼d˜C¨‚ßë-˜cp2ç J5C祴1ªšº‡ÝmÁÑe@ŸóR߀¢:OöÎ]ûè:7Ø¿út/Ðû³öOHlå{Îüú>†ÿéeãªÝ6Ì+™te—ëðö –2Úq–ƒÃMv-Ï¥)äzXµ³‹¼áᓸ+Qž´]²ÐÝK&jæÃ™ ).N;t•-¶6ÕWq¯iÂýUãðqZì±s˜µÄÖ ÷¡±Vœ£™aú  gÜFOƒÍQ(Ю9w]öÄ ðÉT‰l2,“ÔqTª{;O]~=Ñ€ßÄÍÆR…ðý ¨f/¨ðAzÕz&ñîšâx$Ÿ+>Š[@wE·KM,ıSV·tmdhå ¢`©4%<дU¥ŒàF[ÕFÁD@6œÝ¾‰Yfk6+'[Ȱ*>E†öÅ{N»m5%ø­r ›ZàÍÂgÂi©]uhäáØâÇá!—m°‘&–·£ÿìu†—DÌQg.™Ó°E-æøçyF85°Ü×®8Þј„ø—˜9õVpêjïîñÑ»‘$–ôl&yÎZ—Ûr"•ï •UÅž´ûqútlÜ)ŠáxŠã¤àe…t¡…pT›Â˜bSaÄ›âhå\N½†•v‘`ƒ}†}ºbë½Ôû»HkÉÚKÕíÒhk²W;âÀûàÃ&“¢Ï59i¹uH4>@Rû5®t­æåzlVìE^SÜçR…pªîõÅ͓݋«‹å÷À‚EW“Hd˜pb!‹#¥äzžø,´º‹Cëªî,ܧ"= ™`z¦±3$šWÀ é—ºpJI8%“  ¾ïPn2oãi)'vÏ'bªÚ‹ƒ¦nÉåYŸjM±>åpoo‰#X>Æ#O™óYIJ@,Ä6mA˜mkKØM;Å÷> ‹»Ó®ó§hïÍ ´ƒ'áeóo]ê•QÔ¾„5c`=½®š]´Ÿ ðTWVtX÷€Š Æ»Ìyp¿ð’+jÌP«!udŒ¡çЦm¾åþ<õžþ¾Ïy¦é$ÔÉ·ÝÄ_i5§c§<Á;`g§òø‹‘5»œ´p­Bm;g(^aJ\f•;Uƒéj^ÊÐÔj¤úlîVSB ¹Hþ”1º u“/©¬]Þ…?*cWPFê«EÌ^¶ˆ:gJ"®\ÌM¼ÅÌ=ÀF(U!2ÚŽY–’• Äw×Áí¬K\+8ŠãÅщ›Äiø™!Sø '´›ác“ƒ;z ม Ÿœ? IJPk³Ð"ÈëT.P.6.™t1à„—ÖÊ&†uAØ‹:ªOb¦Óô*ggƒ ábPB¢ƒ°Q î!¨Oáç©5Ì›7@H,àcSz0œ¥Tã0R_ǧbŠƒL›»x‹`¨¶'®«ÎGqØ«ÅÝ“}²½V¾bG™N|¦kï(Ã¥=hgF8OR޽ĉ^{Û’ÒÕ#êÌg À‚ÞóÜí¾­½ŒDa'.->uRú¦"Êf®‹nÃqXlòcjó•È‚ìv™4•B t#ŽXF;A²!Þ’ëÁÜ$ê¶› ¶žhòôXÞð´®ZEt¸IþšdÑ‚qR׌:%ÃßjÃßÿ€9:sT™ñº@FìÕÃö™X/לÐdÅH€œË¯Y´C×ÑÃÄwˆ±÷Lf1æMåe3ï0r°5ÑçãNޤÊ"ÿ¦¾y•¿ÕbpÂZ›˜QÏ*£Z÷1YuþVB•Q²ãA^yÅ~«‰³”ôX)*õTÈvKFXUzZª1Í,þb®ëjcÑòÚ'¬˜Nû<åÜ©{ÍÈUcHÇä%ɨ³ç¤ÉÝFMœæ@݉Ìc†¯4 á8Õ\”öf1Ì‘wu@¬ÏhKG Ö·‡¡Rk—QªÅºÚ°ŽŠlJ–btoo–L{å÷¸T- ”1·ÉéÆ9Íè6.¹›eõ»¾wÛ&tz¿xG.ÅÔF!›9‰Ó»˜žn#cŒ‡hÏ›ëj›wH’/Ä£ÎÀ£Ö”Õo`›6Uæ ªEYVÙõª°µ°•襣LÐ5tµÈÈ!R‡ÔüyúË *E`OɆú|8åþ´ÇøË®Ü¨€¹Ô ËBâÌ97ïÄ‰áø ›ÒOS²·&¶WDPxpáÆk§Ð€²^¢é¿ÃZŠ, عvø‰½"—+ Øà 1•¹YhÙ‹®«ƒ£† Í0dëSÝWI„·@~ñK§SvrIvû?­[VÔ“GAº,¸¼ë°r?ë-^†Ÿs™É¼ŦÁR* 0ucŠ´ (Î<æŠF°üié³è+ôhºÝ;Ø‚}ò./`v7'ñÃÎôè´6Üߦâ@äj•TSø =;ß¶™ Б'l ÕìUœ£$ÔDÜEOXš:˜E‚ ¤|œRê1q²ýÿŠŠy¾V6@”×à㹕j܇œ¡¦AOC'¦GPìZ㉉ r>¾™¨Uä7WfÚc1ÝjÌAe%$Õ+°É«Uyh¸A¥0¾ÉŸŒ6ƒ-;^‡ÎJRµ9ù¦>éâ@4E‰Ø/UUÇüJÁû™A>6BÑÀ‡m‰'.1•›qZ– ™C—!¯H)Ìì%{­î(¹â.#bÙÓÌæ:áÒ£I§r.Ï®í6wGß°‚éydw‰å*§ÆåyŠ`˜SÃÑl+zàûIø:³ŸUË4¢f­ét\‘í`(uTþ¦Ðôzž¸ýŸT$ýŽ-‡¤·ý-eé\ËÛœ,înÒ ”ÀÁÞ/†!m9ˆq³gõS×3ɽº†cøØ¢ÓÃtN™8×5³¹‘L-‡æûLƒ+"Ú`„Ùˆ¬’µè†\Ž“^Nª¯»¼y•;ù?«ZdÔ‰}æá¸Wxt®Ë ›G ›õŸÒÅÑDYÄx²mÚÈ)cÌé|© Á³ûÓ´øv¯ãí/L(/Ã~m¸ ÒÞll¸ÊЦ…0{æLpxžy¥xžÅýo^±]Γ"Ùh jq㋎ÓZG"þ&î³6`% üŒ® Z»K.å ‚m¶Ål[Û¬˜w8m!QØ™¨vJŠò+ÙQß;î¬Oƒ¸'ÌMü¸îÌvéj®ìm´³­ ßÁûsSºïÙöá€dâU›f»ï ¾k¶‡½4¢ÔA]ÂHÄÊ`k›}î­ÕSßN Ö„¥Å7”ÐÜ„ý@ùÈèâTD'í¼p©0¢ha©C§¥fÛÄôlµ.7KžŠ…TxT<8ƒAºÖ9Ü,ÿdÚF€¢™TÝ7¶^NãÖkºa_|–{â¸X|^ƒÉŸŒ«åmd„œ)CënšÌg!—ô¨NDÓä1N^‰’º<‰[t9 ªö{hØ{µL'8¤Dk'¹÷Gü)Æ ÙeŨbÇŠFf:ƒŠ†iÖà{‹ÀŒ !A‰c>Mñ/rÍV5ªèì ËV?±0×—t1Â"°µÚöjØ6ÎG‰c/ÇØ÷–ii¦~ïaû51–Dxœ½°ŠKi5õÆåºa58áÒU3:ÜÆÀ-£ðSF+•)zâ >s?s,"/6ŽL™>/÷tªÿp·> ‡s‘ÉU_­{Q¢Ó‘*!K   ýDþ¡ìÔ¨}¹`¸§•©µ‚•³ÎÝuxa•«ÃÔ¬$OSV™ÌË…'‰ð”± È•£z"ðTËÍg‡(‡¾)(ri–ûoÙþ´ò$ÝÇʉ@m¬ªSËi‚l?*µw7VÛp§^ÑÚŽg±Ý<Ü׃±®ÿp#ù’ þ;3ÂêDŒfƒ{Zs?ÝÎUÕh.µrLØEúD«ï{ÆfžçÚŸK*ÞáŒò)3©ˆÂ-˜ Ù3›ÙJO’†X ™Ú[kª`|F¡ÒfbÔM$ß® #ø¤oI‘øO(h:|Œ™–ÈPÄæf¥êÛL'‰o #QK׊¨¸ê [V 24§#èÀ¨dK”ýjY¥¥ÐRŠ‘!5ßù²D(DÎ$@uB­z<…G$bÌiFª¨ªM.ßÒ×̤xÛ‰*ŽŽ–C —öXÄ·ƒ% þРû´×Ì$•çýí"Î]»ø+¢Uø§*¬« ùëHB—xÕ¨RB`^UÕBJ!Ø‘¬$OAUu4gÂØn!wÑ•‘¨ò{5—1ºÌÓGÏŸê¾-èÖRãBë à߂ƹAH°Ò Â“GŒKSß"„л_J{XVKJÄlêò,1þ’&ñΗ­~¯ð‡‘ÃÃ'n4ÒâÕ¬H+-[eÍzþ¶Ó’ìÏ„²>ܦº 'Í ¢s˜©Í•ÓÝSMÛèÏ”XÃÅ–Ÿ7Q &1¡E¹;2ˆ@Ĥf[·DA!¤¾-Éìó„*8/Uµ,ì­$I)Æ–ï>: a'æ5±x…éL¦«å',5»‚‰›A³zKÀÝ’Ÿ×ûQ3ùµ¯Ølàན·(ÙйÄÀ¦Å÷®&`mò#ÒTB%#¿oˬ’¹ ¶*2Bd¨çÔ×,ºÏT‹  €æÜgC–f: ÐÀŸ4¦Ç ‡ÑO­Kï9WüÔN,oVê’‚¼ ÑÜE µKÉZWSáͬ%ìs©&¥dqYeo‘[ÒÝÚJZ\ŠkÿºLÎ)€fÖH"ä®xÀF¨hT¨5Ról[4âÔm1 ¯çäU…n«kSJWŒ(‘ÿÅq·†ŒóRi‚7ð)w§ɇ£í/L'‚/T©~ P­TÍ.<š ŠÄ¼¸²žMD7c{¿Ë wDHðÚDW¤á@z›æ¥ÈF25ÜP(ìr ±S &ÚšÔWcÐ}öwxÈTZâ>úŠZjÁ,,e¯ä&Isޤ·‘:«s‰"¦öØ /!˜zœ&z=™/馲 9nŸ—n—W§ÚS”çÖB|›zuH+3ÆTŒ¤UQ†0 W¢˜¨Ž'|Ï$Ip>pæ:àUköVzd%u(-€˜­˜)£& ^fž®e`AðHnÏ”/ÑÖLÏTzqmˆzr”37–¢hª¸QVŠ´¹ÛM•m3Ùr÷"lÆàÂTE"ãûÀ*AÆX{ 1aüûpÃó°`n:›®î”2”ïÕYvУ¾x¨ÿ\.*(îíBÆß=†D<.³«´¢Ö%Íä ßnQòK-uŠ&ÜÑY…­S5’Ö¶oÀh”šB• vëdª¨³4¨$Ú>õQI¶ö§«$l»wÏHÁ^º÷“Œ4¯Òv¨y¿çV^Q@kí~³¼†×œOÐŒ„cžq^NJ’íáÌò޼­Ù€“¾g]„PXmÒéA"Ü*¾%‘Ïà}K˜²ƒ°ë[2Br]ºzŽsa¾—/•÷–KÉì>±à. År ¹Áû$p6ŠŽ"êØÄPÖ¯£™\n€Ò–}¶¨ïom]!ð‚b7"†0ÔÑâ΃{A¶SÇ!`q0w “ã `ÇošÆð7@bvÇ .»¼’D/77Ùmày"5€ž ž=ÍW˜%݇ ™ ô[ËD›®X°*5G)aTñ"gÑà‘ÔÄñ„e†yÓÕS•åz±é”xq÷j¾˽ÔþÏoÊ4uàCB<Ë®èFꚦNX©Å¨òÐI3ÆL'Âè gxDsi‚þ~Ÿm8ó§ß>…MHɲ Ú#ÑÇR¼œ]ò‰-Ÿçþó+ ™Çÿ—‡²Õ÷ßú¯?<’Ìõ½oÔÒ}çý þûëûâŠýˆ¾Ã‡—øúÃ?oïHÓÛ`éåÇ’Ë¿ÍwÿÿT‚˜îy}TÌÞÊ•w½×vaíkÞu®˜(d€!ébcÎçûK2èkò£Þ§×§‡xbâ™Hz¤˜ø*†ßŠûoÝúëDÍsõÄhÛswýýõ‡çÇ}Å~8ßáÃãcŠüdý·öŽÍThl"O¨å½ak*”c*Tðé©8w¼ï/‰Z83ãŽÀÕ‰ha—{œþû+¢¥óõ¿—z¸nõø[?€Ë-ÿɽ†¥Îš¢ñÜÝýaÿþ¾ÂÏ·ÕÏ7àlø€âoð5ÿª½hWÆTïSþℨçÞ@Dv51}I.¶\»¾ª¡¥+‹ˆõ÷WXDbáÞW(¸o°ÿÖO|ýᑨPý¹ ®¯ŸàŸ_Ø ÿ½ÐÚ^ÁûÇxéßâ{þU{«òyןn…S¢´ún÷¹ èGº ¦×Ý`=#µq÷¤¨ø¯_n30¥²®Ïÿ®™æçßþû—ïµ™Ò\ùóoy½·Ã:?ÿæËp-ëŸÇùò#œÀ™Ê/n’Óíµ~þ·/pŽn;|½ÿù/¸óµæêŸÿ…FB™ýóouãußágüdûü×–ûåïßÓÓÝë²}þ¿Ï?ÿ W¬”®k|þ÷[ šãæ'CÏ‹}|Þ÷BÎó~¥öù_~dçí’?ÿôûý³W¹/¾¯È·[¼>ÓãÌüù?pÁ-ºÍ\¿2-|¡|ÿì=V¿Óµã¾ñûn¸8ÏûEççŸû‡ßóÙî›´Ï?ûæ×ý¾BµW}ÜO…JŒþÿù‚Íç~ÁÏ?ãMʪÙ7ôÈþü{}žÛ…¸oÿúÓ/û3Ç~^d݃Á¿Åñ2÷^ƒñôÛ^Ÿÿçççž=?ø›×sþ3KXzÛÇéÓ?þÓ=ôg»ïðÅw»¿ÙÏ|íTf-›¦~ Ó¿?W¼~ïÿÿt„îOíº î_#´?ñ¾^âÇx‹3c9|•÷¤ãàåÛs¿^“êõ;¹Û%½_ê¹g*êÚÛé=>Ûo÷Ø¢Û§{/ ØVüi¯ŽM ó?ü?õi2gendstream endobj 196 0 obj << /Filter /FlateDecode /Length 161 >> stream xœ]O1ƒ0 Üó ÿ €Tµb¡ C«ªí‚ã  8QC_B‡gé|wòYöÃu`A>‚ÃE0–u Å­ Fš,‹ºm1–'ÎÊ Ùß”<Áf ³ó»šI>ëË)¯ê=„NÓâRP<‘h«ªké±þ“J`4ÅÙœ»‚³ÿPR4•8n®!ÇÜ47I,Óïï|JÁñAíSlendstream endobj 197 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 726 >> stream xœuKLa…ÿ¡GlP‰](v:+ƒQy¸QŒŠR0añ%ÆZFhm;u˜2”ÒŠ”é´ÜJÛ R^åUT4>ˆ5HLÃF4×ÄD ã?:šXÔ­«sOrr¿{Ôˆ ÒPYQUUT¸:ȹ„¼-CÖ©@”ßÉxm(=¬;n]Î 4êäwkæ7áK0³©Â!ø ¬ÃÅ™këx:Ï´ƒ.*.Þ»‹ÞSXXL±1œÙd´Ó•F¾Ž±ù´±Ò§X“™á]tÞ:žwì/(!ßh«Ïg¹ÚC;vÑ‚™¯£O2õ ×ÀÔÐGY;O3ÚúßùÿÔÀÚNžáèJ¶†áì!5g²Ú:…ÊP%:†Ö¤ßCjäBÄG9‘-'ÄAü+EÈî.y¬•š!Rˆ{{“‘¾Ù5†ÑÈ›ä“љɻsÐ ÷‚NÑÅYßïNôMÅæ^–M•ºN‰Ӕ÷¦eì;9e»ä ã݉¨þút×LÈë+BPjÖ¯BSømŠÀº%lYQÉø‡¶ÍÕõ¤cÀ;818rçÙÙE«ä(¹ŠþÄ iÒJÝ7=… ¸ OBzâ‘Wc0Eö6ö»[ÚÄ€D . «d)ìƒo˜øüu¹¯WòÄþ’d_šd[RÉyò¢6¹…(™dÁK)›3½v°°IèÖȤ‘ÜàqKùkÇÓ/Óxk<Ü>læüW<%6åçÏtÜŠ@(ÜógsŸŒÒÅ^è\PanÕB;´·:œv%ÚHv¢ÉPt¾‡þô;— »œ¾Dòèù£«ßÿ2Þs–Ýú¦ê .²ä„G#C7zZcM×(Hè‡áÎD|¦>±säÒ…¤±úâe³ò÷6>¬Ø„&O`¤±C'ø…@ ™ÝÒ%¢¸´+SiŒ®Me-­§²Ô.V³.Õ©Ñ ôûDQvendstream endobj 198 0 obj << /Filter /FlateDecode /Length 6790 >> stream xœÍ\Ys\·•~WåG°æenW‰-ì‹ßè-Ž#»[Ž'cÍC‹¤HY2›!µD‘«æ¯ÏYpƒ»ˆ¤meRz „ÆÅz–ï,ÀßÔV(üWþÿ|ïÁ·Öœ]ßSg÷þ~OÓ¯åÏñÏ?‚.CÍ6«¬=½ÇŸêì6úƒèã6[ðèç{?Ÿo’ÝeâpµQ[«BžmÕ6æ”ýp uA—üp‚µ>ç 55Ém¡¤• i8Þ˜ |Î7‡Ö(ëáá)~fs²>Pc•1a¸†¢OÁ:×7ñ?¾Ä…F¹ ãÓÖxXÉ£“{ƒŽ›G?Ý;tNZ¿M!`õçбÊ*بyr*'•š†SJY7<Ù∠FÜ¿ )%åóðf­ò6%7|IcÍp¿Ôª<|‹}¨ 3ÍÃ;‰¸M›äs Þ¶=}‚Ÿ¥LŽÃ7\”›Ú›ÎÙ=Þàÿ“WÎØa ½'£úáñãÏ64uŸl[ÛÍ—Ôš9+ág.ç”°Ã6ZÃf8ÙdQ¦¡Ã°:–pžÚéŽ.pI)êdÔ±R‘·Öjl†¯6uhÑöÔ¯ ÞÉÏ^-ÒÓ)ÏÆÂ¾ÑI tÒr>D9Ù*ßMâŠ(åRÙ%‡' ;§ƒÓ•Ãñ®è”ÓdWê¬ Û¨ ™ýˆ;Ž«sã GVÒi™zpHL¯è L²9;çï‚ÎÃÇÈD9g†ÒzNÞ o±¤sô…|¨³ë³ܰ»è¸ (^…ƒCm·ÞeÃSýŽVmLÃNRÁå¦vú¢í#¯öÎЗ϶!þcÃ\,&¸Ðô]·9楀=ö‚=^@9òÉ„œ-PS†iÁN´Ö0+ 2)(×u‚” £Á·ÈN8­¬#J œ 1ø)·PZszÜ"°t‘\„MƉVÕñ°¶çiËb6ƆnÑø%soá ÜÓâ¨pd>ù%7±óxã L ¸ymìÖó~ã5, –tŒ«K!ÛTˆžhú”¶9i‰x#pÈ[A‡÷y+¢²]ǧ s°13¡Ž¼¹Ôñ5÷᫱œSðqQƒp!Ž4p¿UiÜ)8@ïU%RX$l–úÖ 7Œ]£&‰ :m°ïlô[šñ!³‚Nð­984 ÆLäÏHbj`E=¼f>ƒ–Ä´+,L‚Ut 8j²ùXðÑw­W1¯/v,!z#*Ik-ʧN}Y::íÅ~ÒS±ßídÏyÚÔãÞtâf ­Õn¸èèõ°JA6‚k¾-˜táhIì%êf¶6Ú8ØB X“r×ݧt¸(‰«d *NwÑ!hqóÒèÿU­=‹Õ®÷ØXGO¢¥­ûúýä«­Ûz¯Ç}„MwŽÏò“‰˜±É„<ì.i—ôÚµÉò >ì â@4PD¥´ñ¢)£MÕòy¸¿àÞ´nö¢ËBô ²½‰q«½í‰¾ 9í yƒƒ¹¿àO•‚…œ6Õ ¢Ãà¢'øä>Vë-"&‰vvO'ÚS%0RÇSõXÞ–^´Ž>ã#$éãx;¢} @+à AÀóU;xy"¿ÿŽÎ^uÜ~øˆÇH@ß5È70{‡Â¤ ”†7€mDØ»+)© [ð´áýWÚì úd×|^<œºtÔ°ÍÙ ‡B€Ha²C¶¸xÝ}zR†Ôf<"iAG¯Ú˜s8ˆ‚ÑIHq5åpîä¢ur†‡pï 00A „BÙ#Éi1D¿6 {J5o‰ø:,*¾1í¡šdŒãy&G Êì–yÛ@ Þ߃ÆC“5ŽÚ)Úm¨GŠ”òºHvÑ—Ù‚`,ð¬ q†gÖ'”H0+"ü¬‰Š?£0dÞvøî˜}B[ª†3 + ˆw›M•p61àPŽ˜DàŒóOzíQõÁ*”Ñh8X¾N[·—X™µõ$¦ ˜¯à¹—ü=lâàZ¯º=±8mÖ’Ïðm2)À¢Šë ÉZ—t’eÞ®›ÈZD3ãcCûGÊ [´›ˆtrBGP1ŸäÈ’UÛ—ÇÆÅ>k"ž9K¥J†+K„FÃ÷’ˆk½FRÙÚ/œõ±¡G¬I€R`VÅ’H©=ÕZ.„®Œ›º…©€”¢°™ í›d qn¼‚óÉLÁìÒ*v6ïS4(öre¸b@ËI´\ªÉ+¡x2îqÛð—J m+¤!u˜Híý°¨ïˆ´ŠŠ5 ïÚáY‘ûþÛF–‰å>{tï/÷` u\$ÿÊÕº;&…%wŒUz}<JØf€Óä!ëT$IÄ¥âe+~ÔŠ‹on*n[ñzq´Ýbív±³Ý´32+Î[ñY+^´âYoÛ”â~±í¶…årÒŠ¯ZñA+^µâ®Ÿ.|ºøÙ[qèïñ»M€f9h6JLýAO1³‘Lg¨t<ÁtX ˆmŠé*˜zÙ€™T}W î·®Ña8j†èÈßÂú°]cLrgyb;Dw$zMÔ‰ :Dr¼”Âï²³zФšZf‡%*Éž&ª' "_4Ý)]B×›Þ;T][!J›±ì;£Fp'¾«›áÓÜýyS7It …#b§ˆ)».ñˆá» D"84z]FMÉ`)7±*"©~®¹7›óÌ®)*}ÁŸÒv™ ([»OVÑ ±ÞNÈßXõÓë!aÇËCþmžÏ:óaTpÀ‰ZßÅ×…E ?")âÀ]2Õ»H°Î,EvuAmêMÝeW×ÑÆ8 H4Rý<wM\¡Vƒº ËÕŹñÿå ;±uÐÎw½Ñ‘g¥‘¬ÉA–ÀÂXquóDMæiç*X¹‹Þ-’*˜Û{·l®$’«w+¾Ç»•Ü[AëwÁê™{ËÀ¬coèÿqSÍcÖ(âÀh)Î…‘Ä¢ÑÛ{NA" ½pÜçjD?ÕçMÄïžo¼‡ãÖt®ÖY8cì?6Õjêê"O{N ž îæpŽÛ—;˦V¸EäÈ&ûKR,6†8‡lûTl¿¦ÈÐö=~¯œº“°ôÆÄ-æ*ÕОIÉ%í†K!“H˜m§Î4X”ÿgÛV¿,¿Hð+ß˯‚®*7°øJÉ&é"–†à„iƒGcVvÒ ìˆxŒó¿ÿL°v/±ý1û1‡ÎÊ¿,âR»$…ëL,•Ÿ àM/¢VÓ ìã²ù ‹)Xñ]ÑõìýÌp¨ò»K1«sQÿ¬-âæ À‡ÃUr Ѻ(¦àŒAKß^ºâ±øè§YwakéT‰Jr¡l“-B¹LÈC—õ;¯ÒTÝla{VâmÊ(¢Ò P¸\Ž+†ÿ¥È1n ` Ýâ·‹wìRŒ™ãÇ¢TÀA3Ì[4ºìÞïÀG[ÛÆ ÄþTyåDËiU!B¼e:RŠ–þAÔþs3úR…7}ˆø×­A¯`©A–åUÿ…r‹”*p"GŠkQL§g¤Q¼sà+™:@zäÄ…âÈ“Üde·˜`ûêÕ:i¿·3ƒ€d:±¿¿¸ŸÅ>Þ>€à¢Ë26‘ÙM/л©ß·àï·ï¶m»f’¶YºžÎ—-¯„ç FÅÙ²|„žÚAÙÅ1BÔéÒ D*€ÝQ%‚› ¨Áláz&¬]¡|KÞH]µ=åà‹Óä#ëæ_Tåû‚«.^’—*F:µè·@/,W`êIÈLÕiu?’ƒ eZ"—“"¸ä¼œÐÜÙL.Êû¼’4üUØOâ„%¸¨ µ+"c!GS·˜›7¦MìJH ;ùsµ—6j¬4_쯚½1uPêy^¦§Ü ŒaÆ0æ¿$!©˜´bgW’*gÔãØ‰ÝëM¯ŠçV~Ù’"gZýÔû“ ƒ™÷¯Òk`ƒçšSS(7 6’wù¡$É.i«1±6(=âmü+?/}LâÜ ¡ÜTM®ôåP6Ã^¢ïµ´§J!»Ëv(—Ë”'#^øjê±.lguÒÄ™˜Vç¼ß3Â29 Úµœ‡›Ì‹ËV{MÛ‹ŒZ2f øÙÆz(¡wVO\PF#JÀ4Yvbs!"Pù7‚°‹Á`ó (7Ú§ %‰ …C¤PuxÓæ&-©ÎÀ¢?´“e@E9`bç„Ex7dÙÇÂ.wå@rÌPQt[f)±ËFi›Χ‚Ã’,²néÔ¥׬§ c€@]ꂈšœRŠ\Vª#G6çGh­ý;Âf¬FýfSý•ç5fø¶‰˜¯èqÓê;î xcø+B*hkÜÐË!®ÝÙ8u –/f9·Ì«b¤v3&µú5(y6a×Ãq¥ý>õOÍ£ ÿS—°3د;tü.¶M|ÿMq Â÷vŸRþ¡¯|—]ÄD¡S@TÂ8`¨ôà?f˜ŸµhÁe8¥ ,(f(.ƒÿº¼ò;Fä— u]|µ@VzÔøÀŒ“öyÃ…Œ‘ÏÐ?l75—{-j€2$?C| ëKµ©›íD“_„Z€øÊßâ×Ï<ôr&,‚‹ñæFËö¦U¤tÓè•P/ÕÁ ô^øQ–®rôÅŽ@ÜFd¼MÂø¯ e€ür R2ƒ4Ž@À—M„\vT jÎÒ…/@òñÄuà“’>¯Ù"•¡O©XŽnÈÒÄ”‰ðá¶ÈúÝÍ¡wY™„Þ_È@jý¥‡´ÂI‘\×₞d¿ \u6ºagþØN‹¦žLNQJpN_¡Èú ¦“t ßçãÇŸ·ÃéRÛÖÿcS)M…UTÿÑæ0Ft¼ G8¯D.´Ù!é³BW‰³Ã˜5¯M/ycwÓ$Ù ½³3ëqqšðBíìùØWšP½_B]Ñm®z…Tíß«á›ø)¢üUc᫆—{[Ëôç$ùŒn,‚•®¹/ÆÓü)&ž0<ÜÍ‚ Ìì,R–«Ýb¦/Ì9E:\¦ü1EÙh|—Š{Í­At™’èÎD€¬çðAð~5D‚©UØ»ÏÓŒªMAÚN ¬îYÉ`0i[[8”ZúRÅú§Žó~‹©éëú i':³æmº®[\²ê¼Qx¹/Å8›Ž‡“É“‹€bjjÌ¿—¿étIMºÚP|šdÈ¢…âÐÀ¿ñö*œ^ž?ú6IM¦iÊO ìZø ùÔÎg˜øêu2zšã+Š—­øQ+>X,¾¹©È@üІ¸Å¼ÁÂþDí‚scz4e¾lÅÝb­h»oŧ‹m·‹m¯Zñ¬´âëVÔ‹EÑöY+ªÆ”©ü–Li£Á6 ùÀZP%É‹LéJ¬°©Útjœ"ûœyy,$åÃÇ©@etƨ0nùݬ”³ÆQ-€>N¬‹D”HëňŒüðéxÕ3Þ˜ÿöÞ«žNÆ\þ ¯zÊîÈÞIx»u%'¢&BÜ”w¤@"𥛙âE†=$Ç/ZÜø‡†Ý„ú™å?Rèøí QèØ]7Ù3QŽ‚F3 Ów@ßÀYJHuÜ ÕìÖ&Jè0ÞÚLšù-zÞP6›É´Š!Ë¥c×°û^Ú/â†êšç xiºÃ ÅÉCÎòSn–Ü<±ƒzC ˆ- Úö"KDz™¤n s@=¾gß –úˆÏ=ÄG 0PûpCù‚ÊM²˜¥Á,ï ‹Vg•qu5²_ëå„GD¸ü™»²Ék²ÌÑ"8²qÊŠ¥,2˜8'hòû˲TÆÊÜ©èêŠäC›¡LÒ(.JbëÚa†«3óº äy§G¯Å®Ðõó<&Î’³¦Ižôa­ÞƒÛ[¤yÎSˆTÏaˆl ß á‘( æˆÅÀŒQ”¡0Þ§ÙK_s¸EÓV˜ÊÝå2' ¨›8eZÞŒ&«& HÂwÈû=I*^ÌY¹Þ0M+Æ‘ÒxMøNiÅZÞÑ_ºÆâ þ'©Gmx«—yÅ—ë¹GmGE(S,s7\N&šˆCÿÛÁÄoL@/n "R”“W›œb*@…rœØö÷Ú5o†¤Ó÷¥@ì•’Žk¬,B§SîåYüST u}ÊÃ`¾Ïã(35j§jÔ㥟†Ç¤°æý,y³>Õ^¯žÉϼߑÝâôÍa5,õ±!Ñx+GlÆJòTŸ…@é´OÇžÃðÅŠ´rb'½£œ]Šâëœ=2™|QòÙ‚2¶L°xÁƒ£„¼ã›½R³1£Û`¢ÔæO<ð žðÑÄA”§HÖ[»÷¶üòÆì…‡…G x1Fz^'/LÍ€¦äjâ¿7³Éúß7ìzu¯ 0|yª;"aû´|–e¸{’E•À@4~rߎúUx)|¼d×â*_ñ¯q e²”ÆÏÁJqCäÏBÑá³8€6é^h™—ÙÿXgÃrÜonáÞâ8åI°·xp5òO{©ôrIŽ'\R¾JÇSE´ÄŽ Ý~•fFPyZö˜N´¾ 1 ÝSÒˆK}pƒR‘µE³Dƒ2¡í º^·cª™ðs“ÎõYu½E`)ryÂÅÑì†ç,4Ú–È×¢¯þÊEQ=oyˆÐ.,,]5‚„_¼xHi%(·ã¶šRs ûéSE‚öl¶h„§¼•™ 8N,,ÇËœ d¤ëߎ5G”W„ü QêWØo»™1'ãø½zcè8y §…W­¿>mMÞpÇÔéßjlm/uÆó¦bÄÒ¾þÛ}ž3†<¾—úþ¨ d¬r w§‡lcr-Z2ÂÒˆÙVBXZ†°<ŒHikLó×òg5õ}-D° ^`ùou¨YþÚ¥aF˜ÇöÖ²:øKºF³§´aN’Ž ÊGhbå„ipVtMµ¨I¹»Tà°ì–×\ž#õhvúï&¾æÏ¸¾çb/eŠÎßÁ¼@îb5>¢¸,Œ?XÀOqV…3Kå§i¤<{K‚Z×gÃv³¡Ž³N½‰µ¨$fQ·©^µÒKpJ]¬Ü¤bg";8åG¯šü‘êzb](ß›ÝygWrèÆÙOš¶œ>7sp¬¨·ÞÅ L¡ÀE]È@XŒ’ˆ9ö{¾<íùLJ—o­…«ç2Ͳb×e’\Ø=wpÍMðiÕäWp2=Tu Áó=ãÿ@mx"Í(áM‘sÎüŠijbŽ÷G†îÙyU² Æãc\(3œ&}‚]„ù‹$ýõRL€óž:{üX&_’bµØ÷Õ+(¿°¥’<LG9–§—,èa³û˜uŸæQâí¼í˜1º|¡œE/€‡þFP»ûRí´öháŠãèšg §3ó\Ѧ¸¦¥½ƒ-žd“Ñ ˆñ.®£{`/™2få‰9=/¨D+“s Þ'PóùHù=7¶Ú²ºÛHòµ§ÞdH L쟔ÅOirvøoq|³P¼ +\)^àºæþÜä ”0×N:ž”8jôÆ•œMÊ_›Fj{å•ðȼ¹•]ö»§£áCh*O£éP²Y·Ë\F|lÛjiªމs$ ¬Ýýþö‘ÔÚ…un—Ï–éB/ÉéúJ™œ¼Z¸9-Õ²X¿øî’—ªúŒë /ÜQ).OF“¬S‰HI@[gª³„-¤Ì¹¬ãL#xß*¡]׿ØZÌ^À÷L§ŽÐÉS À¢¹š=¡Ð’Š9Ç·È ÷ÏÙVÖZÇì#nVÝç QaºÞÓ¦Œ/ÓÍ/w“ã‹Z¡¸4øJ! ƒ8œµËMËå¶“X4ä@–vñâ¶×R"áNZ=h­D¶ÑuP™¯ÕhÊöþ¿04aæ° æ]< ù| 5 ï?Ø»xïõÐ×âq+>[l ¢¯;Ÿ‰Çå–_µÛ-¼Ü@ ÌI‡-.]Ï(áí¿hE‘K"BW‹µâ³ãÅ"]Ed|ÚŠßµâ'­(B¦ÕªÅ¢]<ÓoZñª÷­xÜŠ§‹Å“V|ÖŠ­xÖŠ×r¿%á%wá 'ÌÉKÑêù/÷þHdendstream endobj 199 0 obj << /Filter /FlateDecode /Length 1907 >> stream xœ­XK·¾‚ÜskäÄ<-¾:eœ8Fœ5`K£ø`å0šYi7ûyQ`Î_ÏW$»Y=îÅF ƒj9dU±ê«¯ŠýC'ÕIúWÿß߬^¼2ºûp¿’݇Õ+•íêû›îó-íX’LªÛ¾_•£ª‹ª . ɸn{³*öÛb³M³ÍVZØVß‹¿ô*«¼¸&QIã£8ï×r0)çÅ¿±¬RR!НiGJ1zqI;BŠÉå õÜË~m­"W{Ù~¾¥C.¥t^V¢x QÊ ³acÌ`ÕÜÖ¤!Û‚ä,?õÈôÞÑŽhµ^Öuì×: Ri%X…ƒÚ¥Ù}s¤‹I‹û¢ÁÀÇå[ücûìÔ,ÀA Ø6øÕ¨²yžºµQfpÁukã†è}9ó] b®‹_VJilu×X¯Gw“ î®mH°€[bÕ§d¼ŸÕUxþgœ“IzÔìàë¬O'©Å.-Y”x—ä¢I+qX&xD €u:—W½ â­h9DÜÛžlE¥5bèׄƬxûvSüI¸Sv3HikPeTæÉ¼Nºk§Î‹Ð;Ë(Âup‰«~k©ÍSÙd)¼hŠ fب—Œb󱟴}\Úq·ïµ<9wQ.êü÷lóñ®&DYñeÉSÔ–2‰-QÅ“â e¯ bšÍ£‰bwM[¢Tjd¹Ø=;H(C†²¯¬—q1ŒÓÖ[\Äÿî®ùGÊÖ€ÿ`l·&¤Ú¤‹Î¯³[È“›Â–£¶tF,÷Ш²+P¿ïKqØY%YÑ©ñ_a?t¤Fq1e’CUv*GSK©S;’+¤BTÂ3(A0Q—ËûºŠ¸þ)gÛ¦ìw»ÎŽ6§`u¹B’Ѝï¾a•‹b6HðoŠíÌ~ûÞiT1Å ™¢ âQ:ïMåÑŒëcÛpÛ4,±†R~PfâðÏN)fFø¦3iŠ¢êU´ÑB÷‹ºõà´‰L·1ÚƒÐê'ç†Acu6¶–H·Ö8§~%ùHîsLyb5/þ–aâ˜ÝMK!òÝÕD…&‚A‡*ðž¢M ix£”"–›b+á8§.ÀDR×, ö„à œ”V!‰,Ã…¼uQŸ°˜LÌ5QŪI ˆ#"Ikã±Ð¢ÀaÚÖ›Öj‹‚ éÌCFwiÐ3Œ­ç-8ÿ Ùˆì%ñSu[ ü©ã-Ñ­àCJü6—í>f»›åž]»^úZUH‚ŸÙ8dL<¼Ó幉‹¦÷HžE’|‘Ûm|Ü;²ì\ 4~f\ŠV8Í›þ¼‰¹@UVŒqœ¦*3óCqZ:0ZÚ±ílÏekÌ\#Væ#äû°TÏ.†Ñ:ЀVË™7œC)sã¬8mF`Ýó¡Ÿa«Ç¬ÏK¸Ef k8V¾¬ ¨:ŠÓë€k§K0 øìì¨ÉøÚ»7d-ê9¿Í?ë'Áе+Õå]&FЊ 4ÑÁÒ9HN)ç*ØëöÛ—íÅþ’›oè¶g«í¨bP而«a Ê¥b>¢møRõÓ~6^‚uÆÌ0˜ç07/d@'qÆ]Ð!µö³QdO³#f6[ø¤Õù4óoÿZKVÙXÆ'L=¸ÿ¨ŠÚ\.0€7ñ‰šá¯•‡šØdʸè%HO‘e<£0<`(W ‰Ì0â?ô¤Ë`·(þŽÿöý¼f!ÇbsÛóÇâ]:Ä€ºšSv`Ü”‰;==¦T}8ä"d³ú®\< ƒjþ:Ì%À¸Ù;Eæª^z>ï³áôP6Pž¾È«¨…g 1jñ÷<Ǭ0©Ý !¿a—ç¨úWy,¤Ê:#b”:N¡Â|Êažüd~ÌPÀ¤´!y’™Ý˜çhOóŒQ¹|2ÍoÃÿ%Ë…eòÛïÙyœgû*³žFË犯/]wäð3Šþ®ÕË{©[ŒîôÁ¼öŠVêäEñ ¥=!̦ÆA¯Í|˜œÌÛü,¨;ŽìËïNøÛÛëYõÁsÔ‘ñlpüÝyÙ ÔˆXfd*}…+–h|dïa^]·­uN”¾ylh=•ÏJõ­v×ÄË&ž7ñðÜÞ}/š84ñlQ/Ópª H•eñM|×Äc¯šøØÄ¡‰»&îŸÛðзØÿš¯|­hK¤ß¼b=éìÿŸ]¬ùhf&YÅM|ÙÄ‹â§çÄa1ÌÚnqõÙ¯‘ŸÄËÅ䲄]7ñæ¹ìšxÞÄMüyQïù¢;Ìße´¾høùvõ_ï–®Òendstream endobj 200 0 obj << /Type /XRef /Length 223 /Filter /FlateDecode /DecodeParms << /Columns 5 /Predictor 12 >> /W [ 1 3 1 ] /Info 3 0 R /Root 2 0 R /Size 201 /ID [<554e844d2232aa37f25cfe39a6334681>] >> stream xœíÓÏ Qð{ï(¦Ä ÉøY`¡ìå<€¥­…Wð’Ø(k+ Y[’’”• ;{ù¾…y†)³øuºsÏé«ÓUŸ’B„é[(È¿ÐÆtáÔ~îÎüwçm‘kcwæ[Ë´`²Ó[¼Àè æ7¬-˜j@Õq”úœõ ê+nÿå‹çch 7õ!ë"Ôî¼É÷®Ø«5Y³W>™§ã]Xf¯y…¡>'aì=ä1³L¸çf–7Êä‚3õ:Œ¬¡ÿìþ”X?~ÊQÁM¢U9yKñmá/= endstream endobj startxref 206719 %%EOF flexmix/inst/doc/regression-examples.pdf0000644000176200001440000062254113432516322020152 0ustar liggesusers%PDF-1.5 %¿÷¢þ 1 0 obj << /Type /ObjStm /Length 4618 /Filter /FlateDecode /N 96 /First 809 >> stream xœÝ\[sÛ¸’~ß_·“©™ À­SSe;‰“ØN<¾Ä±·ò@K´Í3²ä©Ä3¿ý<ì×Iñ&™¶•“­-Yâ l¯_ãbŸy,`J3ɤ,dJFL1ã…L3#ñe‘ôYÄ„)ü0á‡øLHOâÇ0d"`"òH2ß÷q ™¯ Ò)ø‘fB³@E¸oXi¤˜ ´`¾Ç¤ ó“Ú¼ÌdAHÀB‰—|”H{>óQ.ïûŠ)ûžF‘<7L{ÈYêÀHxL+™ ¦ÆÑgF@h0ãÓsÉLˆÌƒá¾b‘J@TQ[ƒJT_?B©PÏ@¼t}@Ïðu^’Ѓ£€ITLè'Д€@T *Â=& N¼íâD ZU† !Ù×Ð`Ɉ ƒh¼Br!Uq¤XYÅã$KßÃ#H– Å IíôHAr(p¹ˆ0@©”O' í É¡„|É¡‚ÉŠT¨ YžGXÀ ZJA²’¨ òŠšIS‹*” M ”F É*‚†5$köÑ’N jfíCÓ’Ñ:h"‚†*´¡ä¨!Y”°¢©E $ëúG¾ÂPÁ $(hĉïSâ$@3H6J0šNð.0$Œ‚XC`ÓPNɆàŒÖ&BÁ À 9‘E-êĉÊpâE„õ(@á ¼0€à¿þùOÆ’<Çy ÀlŽÿ¸È'é4É`1öú0¾Æ…t'Ý%Œï ýdvÍ~ûÍŠØZä7³9{±äy:ÙîüßÓ_Ø›yšŒçéè†í'i6ºù ïÍ“8OgÓWqž°¯þmBC~„ÁÏžø‡çý£LG÷ã“ä3û–æ7ì9ÏçÉï%}›ÍÇ{qô »J§)¤Ý¦÷ùbŽãlœL²_Øu2Mæñ$ý;3ªN<¯fÓÑì6ÍãiξÆó4¾œ$ÄÌÆìp>/F J¶{¸ÏvofYžæé]΢—¾B‚ãÅå¿’QÎ^àü$Í'¶uw7IG¶Þ›]µ loÍ“kœeHQò§Rµ;³J ò½þ83j×T‘ýEÛƒk.á®”KØCX<+޾{SÀ…¸ûÚ¥W\Ïeq_éBY\rT‘^ÏU™¾<rtQ6í®¿”8rЪª:Í“iž›ò’qoÏîQmÊ(ŒÂ—pŽFŠ—°„/ôú/”Ђ@’öú>ß=Ω9áš\‚7ÈWEr4åè8É!›¾zƒ%÷9„þö[½pÍf‘γœ¹êñý8#N EÊWI–WÍÄ?Ÿ_Àú_zÔ½DùÙt1™PÑ?PnÂóJ³)ª]àæÝ4·h#àüÔ£$½I%-ït$Å“TT•Ìl´d¦^²èy%‹6Y2¥j%+ ÿ¸’m ³ùm<Û=<`‰öX>_$eêenå+q–Ø ùû½ó‹ÝŸw¶?£·ç¯áÔÆéôš:W,‚êÎ =¦w€Më¬àY ‹¶.Òhë½r›-e.Ê2¸ìÎÒq~CHùî(ÐVwOÿÂwÚGO^ÿvS”ƒæ×hCSÜ×ð[ÆP+IaìE´§\Ñ´u‚&ˆÜËJ®(HjC• ‰(ÐÊ!ZC…¡kW "CtF·©Š§È,E *žx¶€”¾‘ Ò©zÌlDÓÈéÝÛ$½¾)/Ñ |‹oóþ†ïòw|ŸðC~ÄOyÌ/ùˆf“Ù”ù8MÐs¤Oø¿ºJ9ý­üšßð”ÿÁ'ü–OùŒÏ¦ ¿ãèBÓÙ˜ÏyÆsžßÌ“„çßf|Á¿ñûŸ¬ó²åú•H3‰¯3ëô€íÒ\ì3ÅÐ÷‹{ú&$˜oÍ‹|ˆo“Hß7¤£­é5êz¢Ó›^[hYõóã<¹ýDÄ®¾n;Vp¾wpôûdp„ž¯fºm=6>Ù´î¢?© Ajék‚ jÞº{$Õ'†ÆG_ºg¯!Ž”rè×/J¸,©ÉbˆÞ‚Ùo?Uàçàªö™%ð€>½A0ŽÀSí;tm¬Ðeª•:‰@‚.„·+h‹‚äü´<^XÝ·U£kßU›”fX<§ûn*Kê(ïQöt¤¯•íQ1¼e¦3 ›fö˜Ù+þÚšÚ[Û{¾g îÿhð“ðd˜ß'~ÆÏa„ñ-l)‹§ãš9‚UÞÆM£´9òävg7<™ºÃŸ‹xb͵²Øô+ìwRZîu‚.a:¾œd0bðÞ›d [þWeÍSðÙ¥IS_5I®rw6·vþ'ÿs1ËH² Ê —Æ^¹SòYòÙdé=Ï&TʦøJ‚ÿÅÿæ'óYÃU6û\ůÒsOÑ6‘UzËYˆ¶³hÚò`_!†úŠO§»gÇÎÕETÄu΢Æùœ³€y sОÕQ6?R¶]OˆNoÕ×+~›£¢g–‚‚|À&…X_ZMC-õcÍÒ™œëïœ9;!#±Ø¯oñ äÖ1ÛÁ)ÈÃÊ.MRlÞ Ò’LU mh(Jv Déﯷ>}¢ š=Zi.KZtQµQê{QJÞus(íïêÊ^™W6Q¶®LÒ¯zKÇéSv}}} ÒÛ^=J´ìe<ûû;Sˆkø<@Ç›c¾tÞiE¸î –E¾³Á­D$W‘žE¢—ZÕ‚XÄ&N†âµ‡üsQO2×(?¾?þt¶Ìލ~¦[Ó,]ÞXâÓt½¨/Zø¤4âvH‘&ë $ë tTSzÕE©0DTÐÔ:h¼;½ÝR¹l«¼¥…:ü¡¦ÿfk÷ôÕÉßoÚ~رýÝ>=¤{¾íû=„xµõ;"ê®J Z>uÔc|Ì:~¹Ú´™¥Wx¶üŽW ~¿ÖkÙË™$¹o=C!¼"¾ç‚³®§hÅ`j±r ‚^ʾ…Ú+j[¨Üc­‚ízO±·ýáô‚úǃwM’å Áp7$C1¬ê¸‘á LyÅ@ƒv¿ëˤ¯C» œ|#‹qo *š1ëÙ  |[þµ=Ÿ*­G‹ÐÖHF=HõšHõH}‹ çƒ m.‰f•`Mø$¾½ÇüvQFw©¥W×xš[JfCTÅ ¨z‰¯ƒÔÚø™Cj ;ƒû´Á£ûÇï¶­)œ7ÁÙ.]lv"ohð°wìPxìGU Èâ|…O}5@Ößr >T¿Ç±û~Ïùå<%–ÁÛ3ŒâÆ-P˜œ~'Ét”H$¹p¤c¾˜"@ô=ˆah9jÑaÙò[ BthA¸w¸õi÷Ü¢¼EñK±Da ŸNñµ¢îö¹¡ßÿ͹]ª[× úM'è×aOînbç¿=Ž¥k¸ ÖBSm²¢pi­Æ 'DrCIãùÎùö›ƒ©»´Ò»V` MGXÐæã4u7ЧE¡ý>®™”|ÈûHõ Q©•èÚ“ÍÆnÇU'Wtl_4›<(›\ÂÓªÛî´ªDZ¹­vÚíÞl•ÁNdÕXÖ’sIÑãQ^ï¾?@f¯›[¦3°e¼.Úqñ0¯ÓÇ"áífitn [ 9h‚sœfw“ø¯å<¹Ï#½ÕîÝ>ôo6L1^7Ȧ®¶Œ?ܽ<;Ù½ ¸üÈ ”eO£<KLiç÷L­«aêõo½ÑJ36 j ¦Úð^`š³q6¡5:.í*su¹IíW¿îN™+ "\’cKg#Z¸%iIQ樤Ü÷!ªÑ4{ÛWŒâÇÕX}R Ã_×±uê“EÖ3mf‡ÄWˆ~1¡î!"¢3ÞÒà¨mNkQ›ßí«½÷§ïO-éit"‡î˜c0ŒÏN§)¤&ÌŘj³½ªª“©Î ®ìFä²Müe_4V#Så¢Ñ>0ªëµi澦rwY©¼Í4VÒ ²®À¾ÝRwg<­©Ú–ƒÃ­ÃóýíÓ#Ûšj Å õ˽w×w ž=³ãv´u£oëuxÛËeðÛ~¨t-!¼në–7’–Ê;ƒNM|—Á‡ xúÿœwçGˆ;e™^KÚÞRœ+&U½Õba_lt±pXëYXùã /G^žœýn'ôC¬»¡[è·Aÿ(b½)JÚG}׿çëöÝßÍÑýO­Ï@®û©õ&xur?šÄ·Óë§‘k;ðòg›c[fÝêÓq2ÏF³y2Œ`GQ“`ûž{Ü¿ ¤ä¥µ!š¶%Ö}­õèCY¶z˦¶'×á|LŒðØ÷KW\§ªå ôÒ4«K›iT„6Þ5=ˆÑôÖÕ½ôʱ¿êìKÍSmB4h‰¿nŒ§¾°¯©õ«úˆºo§u×¢>^;TQ®ºM§ð…÷}«»³»x” ’fÅR$ãF´x¶CLýΈDéëÈ)5GµÆcð ãE|wõ 2.cX}ÍF²&–òÃŽÓz¤Ç-;ˆáqׂOô¸«¼läSŠsɨ<§-í ï;LoúGöTz Þj!çñɇÏ4…xò®±óníŸê Wuçø†ºí`ÍâêúGÓ?¨f=TМÿ0ZÎ9(ܨgW¯ú”¡û–ç®$tåÎèžäòœ¼å2! ]Ñ$ìîu] ¶Ùý·µELE|{"ù\9(1` bvJ9›ÑŽ®ïz×OÚô>ÌyGýÆÿ9ç]®_zJ§GÿÕ@÷ÈÓ{Àn“úmR]oRó„&]öaÞŸ}zµdb¿nÏ&ãZ«ŠîbÄžÈCø>lð¶¸fèÑêiŠÀ¡N,ʾ”CúÂõ3+$¿ 3Ê À]—%¥£±Q5ÝwÏèÀ”©×£-bºtOG[ù_&Ò««dNKc-ˆœyVÿW£jŧ%'”¬!5Mi´ íj²)a€ØUJÊ¡­±nÎ¥Ü<±¡,h\c•™]ïÖ^ˆ¶©úüv‹ÙE¦ÖÒwÐβÀ›‘‡’ÚmœVðÕ•-7ÁÅ6/-÷:Û•-Ê©ËmÊv#U}C³ý×.Å®jVl²þ>Z¡Õ±?ÈÂü=…Û´®Ø§ÁøíÂ*ò.¥ nŸ†]êë˜/sƒ·* MËÙ)£‰ Mª ÔfU6$`¬/ݪiË.‰»D÷¸™Œ¾Wó ÖÚiAÛÞë1èŸM•4è?N-wbÐ?‚`Ë ¬¶£ãÚÿ/ÄÚ.endstream endobj 98 0 obj << /Subtype /XML /Type /Metadata /Length 1706 >> stream GPL Ghostscript 9.26 R, finite mixture models, generalized linear models, concomitant variables 2019-02-18T12:53:53+01:00 2019-02-18T12:53:53+01:00 LaTeX with hyperref Applications of finite mixtures of regression modelsBettina Grün, Friedrich Leisch endstream endobj 99 0 obj << /Type /ObjStm /Length 2300 /Filter /FlateDecode /N 96 /First 871 >> stream xœíZmoÛ8þ~¿‚w±¨)¾“Àb$mv{w)§ûr·ÈÅ‘¡¶Xr›ì¯¿g(˱9Mm·¾FbJ¢83Ï<3äIÂ&Á„Ðø–L:´ÅT0øÖ̸€oìóø¶Ì+êçX0 ߸'ê€?)]1¡^F;ô’ «¸p±| G†Éa™ƢᘔÏ…gRKz˜4 q!­r˜ô¤±”LI¨&$Ô5ºH†Ã8Ò0e =²°DAé˜$]z4â£À´T謦U@%˜6}ð‚¶ÐN@°v„¬ÖAÑÃŒÐd¥Eƒ„bP# ÝñÌhY4Èvô3·…$ìÒ’™0 VÌ&0WhÍlÄOdEØh‹é¬³&>òÌZ‹‘u`Ö;È‚<ë=F6‚ÙàéŽdNxôcœ„…Ñhx œ&±Ó¤³qÌYBΓ†& A.° ó‰$§ æyÝJ4=RÌKüð×änèä-P̰šÞrÌ;a=ó̱r£€Þ!º ÑqN² ‰2+hxF8Í‚Í8âÈpgp’î€tÞCt h©™©>’6Oô¥¾‚º&’LòDUûy9ËGz+jYjÅ7Q„²‰¡P Æ%tdD/’áˆØÄšÄ‘ÜÃ~äæ$ú&†Xó·düuVVìOæoCÆÿø×¿êün=Øl1™°KÆße÷U ÈØí/«˜b¿¦t%–ªÔŠÉd£b"IZªýZæ³v—Ž>¤7O²ûi~¿RnƒJÿ=ðìÀ¡ZRWð,%泼ʆ¯ó¬dŘýœ.Ê2OglžÝàV \Ê.0z¥…H:jhgîï"£[È$ëÈѯãù9¯òY1ÍÓ ›7yÕ«÷Fmµ$JwA³…TK…ã¬JÙÕ¤}ÈæëÐÌ*ô.™TËîgÙuž÷LrM0*˜>@h\®†_"3ÌÊb1Á$íÍ}õóE•ÂRÊ¿±ÃiAºH߀SŒ.22ŠŸ¿>…n¤ù%”Aóá.£ÑÁ®ù¦c?rÖ€²}Ç~ßoÿÒ!ºGÙ-à€&édÀª9üЉÞ+$ʯC¢Â.¸nZ±n@ɽ ‰Hž…Ķ y?ϯŠI¾˜ö!aöŠ„6ëHh» ÝŒâ”$ò‰ ¶ûŽíwé¬(‹iö\ÖèÊÂÔ4ÀtÐM¢P»¼l'Šv”ö$Šó"/Ë¢?³mÖLÛÃüÖáƒn«":y5½šç#6N“ª7Eؽ²ÀÈuµ „ê±?¨î¼"l¿ýµ+D;gŸ§dv n¿0´Ò‚Ù%-ˆn¦”J hyÕÁ? C;S^dù_ b~¯8¸ÜNtè¦G¬b´dë&…Žåí„x4…’e:™³l2IYZ–é»N«´Œ°_0t ³ O–άÕ]0Úk»vf€£É"æ–ήÙxAé‰}*æVX§eÕåg'çÇGg?œœ½‹ˆ¿™ŠkZÒò¹¶ŠòäÉm:gàP´ÊŽæù]UÌc·JŸ±ÕyübqUEÉRÑ\ ü=¿®nK묰׿£>ŒßpœÝüu¯ž~ÖèÉÙŘ¢jØÆ|ŠrÓYb¿µäü=Oy:½Ãbç#>*¦Ó”_óŒgÓë´¼åc>Î?f| ó~Ëoîn³Ïù„OùŒaÃï8FÈ‹k>ç%/ó{^ñêSÁü#ÿÄïùÿ+›ßG²FM_QÈO'éM‰‰î=®£é@O1eÚhÂeýü4ŸdŠINSz¿UžZó|ÜFûòiêž|:w|;_ê| ŒhËñù„°îJÌ3råbÍV®7–5uô‹ÆÕ˜4‡¬h YSµÁ fÅþbzì·ÎS­:OmSç½”Avz¸¸Û¼Mtó!TÆ5mÍ–Å×F–= tnèZY’6£›ÊìÛ†Ù Ck¿Õ’nUKz›jéÅ¡%Ȱ?t}sÀÐ Éá 3ÉеB‹Žþô*œ^°¼=`˜5þÛW˜5+Èzs¦ÙOý* u@®ì²X5õ‰å!g0s@èÔеŒÎÁÍvéêiì÷è´rf›…\_¥¼©:~¬‚¿e.Ù›“¯CÑ싌NÇ_ <Ÿç¿ÜQ{Lôû=Œ°­JÖnSÉ~†áÁzžuï÷ØÁ¶BÚ~ÍÚÌ7'‡ˆˆæ o›ˆ ؾÚìE‹u®oÁïý(ØÖBÆ}Í…L8àBÆî²è¥ßùØí2_ Ñ,M¦Û}Ž ßˆ¹¾ËÁdþåûó›öä79ÿ…i»™Jöu0Øú]Ûæw5ÛžYôŸS¼ § ìwÚ'ë xñ Aù<Íql>g°†Ôÿ3Æ(¥‡ËF—UnØ®;¾®î!œgG³2¼ÑMVŸ íMüÿ»ïÅ1ñ »>^ý0ï?ÕR¤endstream endobj 196 0 obj << /Filter /FlateDecode /Length 7176 >> stream xœ½]IsÇ•¾ÃŠù }lL=¹/s³#ÆŠ±uð‰9Øsh Ĉ†Q"uñ_÷[ry™U"íðAÍBVæË—oùÞ’å¿íÔAïþ¯ü÷òÝÙüÉÚÝÍã™ÚÝœýíLÓ_wå?—ïv¿y…#ÜNǃÑÞí^½9ãWõ.Ùƒ‰Zï¢ÑíòîÕ»³¿ì}~¡.ç”óþþþö\¼r)˜ý[úé}ÔûË#‚I§ýðÜÁïˆCú»'ž“ û»Çó Ý!f/ð¹ *íßð㿺{+þ@Sf—­ß_ã°ðô=Q,¤óþ<Õ9jcé5¯¬snÿþ¡¸ÆŸfvi_(IÊßð"!Çýðt{ïæ¼Î6L¬à£<Í«`9Yü™rá±ïÔß»ãå”ÖeOÁz‹¬†h¿¿Â™£Ê*†ý-ó…è«ü߫ߕsÕ‡ì½Ás…ótÀžÝ…3‡àð@¯à@ƒä$—´m;ÒÙ'ƒÜfʬÒL°Öè¹;2¯2°þb฽fA68$"À¹Fcwð0¨CàUÿ^ŽÓƒ£pݳ1;9lÿþîüÕÿã>H,¬”5îãB+˜ÔúÝ…¶ïyÂße*c•ß!íÊ«¸?ÞÑ: } wðØoT=óœ]ôtæÖÀ¿ôþ÷$· ΔOÏæáôîÅ$·Ä®¬ar‘0ãç°p&nÿ?´ œÄ.£• iÿã¹ ¶ÔµiŠ“UFLJ'ÞqÁ1÷Aa‚È(m<œ§Cvº˜CJħ¿#›Pëû³¿ì?ðÆRrûoyF²¡ïæg:[¥|vk£•úùÏ¿=ÏdÌXV…ûýÛ~š{–Í@¯Äs–uúyyná4aûïŠ åý·¼Ïäm,SÀŽcÕd”CÔ£œA•ºG ±`6œd<R””}=—zÞÄ";*åxQ"näYNgâì6a^z&oú9ÐɆCF’—“n9 '{ ƒñYÊßoÈê$ãȦÀæ”6z%ì’¶:çA¾G‰v(g¾]Ýv¡#ìã·ê»ÿß.·oY•´N‰T Ì€\7ý_¯Îþxf9í~^}s–À1›á$wïμ1í_·gÞö^i— úA:/3¬wøcÎ ÕÞÙýk4 5e:e§€pÏ›4J®°v‚Ëp¼Äçd"Z9¦I_§àBûv„/~>«? ˜9nf2áàÑ|î=ð+€º¹.šB±ÝN‰9m›zñq¤Ð–§“9¾.ÖHÏÏ3(,O–ˆ™RКâÌ€0CµE(†ïÁû=>ŽgFËh¿!d—ì{àp$ÑB®,'6ûùün}8‹”3! 25>½zõ3Ɉi”º+^(0®oæºoæøÀ§Ày¼c´¡GóVʰ·Z¼'Öx,ç Çò“ˆAÿ›ú0F {0U6A8b‡ä4|œMš¶®!&X˜4öÜ€%=+|ð¨ˆÇ:5Bù®˜tŸœ|ú†w dèýE3 ¤•ðQº5“"x)~«Ãé»Lô¯ÝdñiT•[ð÷ÌZUe›Þ=óž¤RÜ“RxdóéNhðÔ÷ úCƒ&4€Òâ^*BØ ¥ë² ©¯¤~0}Ò–n+éRWŽÂ£\ñã@F&‹¸Öû6™ô7ëîé®=-¬¼Þ ËÂÒ¨7§_ë³gVÁ_¨"¸¶=‡Ž’d!lòÎU èçFmÙ>2ÔÅût~ÃóÕÍŠ™É‚k‚Êðžf[_(E\ØoVQïróFaЀ#)>ÂÁIt³)RDÇÌ ”py'5ìf^™™ÜŸNApLAyÊNlxqöuéüÒs³)’~®çÀÑ<ü3$K¹€!h³¼|"d¼IâàWðRœ¶ô¤þ]¼5™”6ñIÀ¦bß™b~gð}”h­Æ£Ÿ}äðÏ•,Øó¾Í eŸGðEœØçñ)/Äp “æœ^Œƒíd¹Ÿ‚בk¦`\J¨Âc±ntJDJèöØö¨;c,‹.¡»a: ÀKLhÌ„tÀü-OÓàå8êŽ`æ&S%^7¦ì¼•‰aݳÜtÆYîS’ðF®ÌŒ3@ÝxdFAéAB m{Ú)u èq¢Væ6fP1úµÐ€÷]1†@kˆ2š‹™ãÁ7Ûûɸ 9‰Ç{¡¤]ls‘kW‚Ë®G¦ìé„ÒÜOÿMýF!l=’¢ÔìÍTõ“Ž»äñ åÓÛÕŸï7ä¿Q6Xw~“YàTܪ…Øêà”û$"À:‰ó}äÝèøxEíuŒ£c»^%¶Mâ6 7#'z|‰vʃoÀuß ñ+¢°‚EÎTŠUs„ȼf`šjØS¤½è,Iç]'‚‘Œ“k' 3%R†ÀÛ¹ ÇLJ—ÁÂI‚7’Mœ“KïGŒ&°`á…}YÚ‘ûjŽ8à H¶Âü»Mi#Ê ûj\)Dl3™QÓ˜à r'ó4¼åô,Ÿ#L>½fÓ?!÷‚SnìMå¥ÙC£CD/9j„|³ýfBÐn`cÖ4ó5Ll«'O‰Ë…5žcŽ’–ƒ$ g8=0Éýš[?I ,̆4ô'öðÂdŽÏœR~ÌÜÌ ,ÏÖ^r„¤œ²NÅGÿB€#g—ÿV„'üF]²±Fæ˜Ff-Oú§ŽV’ªù(#7ó"ñý‰ö¢½QÊ$!G§3Ö„Õ¬u*tØkâŒa†D3ìËÒªÞu<7ªYGqBÍèá¶CãU-—Ç*r*a=z·Ó†‘ðrÝ«ô¬ b-ÞÍ)rÛ¯ÖøÛU}:%…½ä™Ñqˆ™`û B‹L²:,éŠ7ßq¬“´¼hO¼¡°[ƽ×åñþæ(}É ëÑlê½µBèU²á¨³Ú'¾®âç÷­®ŠÇRl¥ï~nÂ`K—«¡+ûþY fñò& ïåd•\¼'«ë¢è~,ó¨Ý—Þî¥x| ÃsF:”¤•ê>gj-EE*sä.Œ*&d¸t:Øløè5æ9ŸA‘5’3r+’þác«@ë ÀFçÁ„'_;ê?ŸlMŒï‚zhtð;€¹@#p â@€WÔ<¶ÿoÒQ°a*Vã?|iS–Ú^.ÚoNôÃûËÄ_ÞžËAÔÂŒÞß­·«ÀYñsH²e@›Øò`-¿çCçqìµ½è[Ë€´!ˆèœzŠê]3|ëªk•Ê:¿­ÉOZ›±-É®¥†s×@.]eb!}Wq½7 Ž“Á£^5õM4 >*ëJV¡èzñŒÎ˜¢êaèp‘ºÁ–ŠºÈÆîa© @ ü¾\š"*½ŠÈçU­ÿ ÷]Íy æB˜4·íǽ… Þ‡Š*ÉŸb²»>Ù Ï tINÁoKb6ÁD1…àUÙj3ÊV9[ú<‹úз¦]‡¯% &åçSÅt5W¦K§&BŸî[¼š,  ;ü¯ÁÖäO™£f ”Ü)檳Ÿ¥IYW#aúͰ…úo0ªò‘|?ËcòÖæ-§pÙw$Fl¶z+ôàœµÇ곕ŽvÚ4ç>Rª ÐlD³âÐz³U¡Áp¾5æ¡Χ´yƒ ×Öy<à#w8ã¸ÿ†$Ù`7aƒ"Ér–Nê/ÛhŒ|-v‡nÃS7pÆur8Êf'uUX˜b·Çûþw¡ W̬ŒgMr ÞŒG“c:äÿ4%ÔX\åýú.„rÞ wÒ`cAºØ;‘–Ý1ɘˣ·ïdm¾®‚)'Á¯G|QwÂzQ;Ëai8Öﱚ,ÏàAŽo»g¹ã!Hí›n¤ªž‹ªgôÅ~€!7ÜÖDa?‹96¨ü”˜«LV³äxƒ+}(ÙLe¥„ŒX¥s "içߟ`òI2²šG"„1fáÞQX[–˜2«<µ¼”,´ù3ÜN– ö —°)<]Eù¨Ã³`?£D]ÛÂÛÙ­–VÕën0ê¡—N6ÜçèH7ó Ü©+ †0D²#•XHÒ”ç6¥ž^„ÁGý)»–ÈJ¬+]zl¢Á8Ô¥¦<<’©i#*„heƒÔ9Ÿ¸2‡ïÙìÇÓo‡»DÂc¬üÌ(¬ YÔLÕV¥ûqP LpmÞ*£9{cƨ„7 ªN­â°«P¸EslŽhh¯'&ë1Èàl$¦n†ã”ª+#˜—ÆD™0mïÁ4zß‚¡ph(Ì`U‹Ü„Ò;²ÂØ> T]å~Í‘û¶%+ ”+½' ¼” %?K#Tl×=Ô24ï¸s ¡nfÍs­•„nˆñ¢ä?º(T­›¢6„ÐC_ÿæÐÔj-ðx]³ÒN2o4¦ë„¯áLk¨XòÄîÖFtzo5êÂ*¬n«üu?&”(ˆæ›`z*wO\ÿu»DÍ vÞë5x_AsÔüÛs¼¦´—AÑ‘¢ñÖKŽ}d>"ŸÁqZ³•áS?¾é‚϶Lc­,®e|¬i«¼7J™s±å—nÛ+1|¶ù×óš÷ˆ ÕÂ{:¾4”âüVzf¶¬¸[©5µœjæPŠë¥wÅR7n©=:0›Ç÷óÏpŒÇ»z:Az@9D˜€®´Ò³¦ºäÜþuŸB¦¦œ¶ðè§òCŸ'ÚpÌVy 9鸿•QãDeL¨êˆQÈ`)WX^.vZC|t ãŸ¡ÅDä*¨0¯ýRÂ>@<2~Ž%Õ T‹ ´·ÞZVÇEjè8òO¢7¥ïÌ:·•9íÁÙ-¿Ó”p5ؽh¬T¢öÒ†|ÿB÷Œëýéu±yÁaÒË˧SÀÂ|¨f j¶æ¾ÁVb­Ç…g$¡7"Ìz2N(«¬ŒýH+E0^٢Ƌ_\D3QÏI ¼9(Óò“ÂÐËUIÀ j(ÅëƒÍrÛáêQâm¶»„òO’=$UØe‘ݘÆ,´ˆ¥ÈgËØU$ÇÄËxSc¥Ã¶“qÅσ–Çx»ús„U¾^´\¾€PK&,ò©( ¢ŠO -Q¶V/#læçÃÌÊ0ß \ÆQ­á9:Ð.‚Sˆxµ¯ve/‚AÅô*€NfBšÌtÊ4s[X>ôwõ¬°(/³bÜ»C¤zdeÝT†iµ”vÞ¶ðbÃÒ5è­ß=gÉ” ©I ¹E†%ýf±’R‰ óÍå £õÜ.eA–1ˆÄ§>M¹F¿:“©.¸L5Â?ÂVƒÔjXGï}dz­‡3¡å‘"›d| +ùR²p”Áá0¨–OF$ ŠaLž¬Ç<³R5ÀVYý2Zhn½¢ 38/&ÒXq0«ùv}AÚ¥…]ç&šë¹­ÊG·Rñ7%qáL²Ýñ|Ÿš%|#{ËüÒ€s!ã… ™?ÂŒœ²žj7]—ÖÑR‡÷ÁõÆq~) e±fow&øƒr ‹ï'N—½-IßöNY…é¶··hzÄßþ"+çw¼û Z¿upUö©|Á³Â'5<+3˜üÇ p_ЈŸ"•Ü?ãÙ~Sþ °ÙPÆ'ŠeІϴsQj"ü\xP´L4&m¸š1ÏUSº³{ҟ륿Ðâ‰a‹‰œ„ÌFæR¶}íPV³Žº…NEèù–á–í‘(÷ÃS¿ÇvÁ¾Mþ¦ÈŸ¶OEÁ¢0(*ŠÖõBÁZE±¥àdAqŽ‚I£ž NŠó…‘ÎwÝ»:Õv:üêGÁu»ÖFZ°ÞU‰¨–þP~Cfqêõ½úÕ‚½s²³ÙS £W÷O"^’ekaÓ)]ýX šÌ8ò%…)/·œÏ(=44sþ"­(xWÎR›äþÏë/’Ö9æ }É¢(Žˆ»å}v Á¥!ÎzäY°4&™/$ÊqÕ;üA³¸F\mý2‡ü^Ü6EÏÁ‘¸dПU“cªõ“u"5œß¦ÎÊÔÃèT*¶« ×}¿·¸JÍ’‰ƒQ²š6´C;I8%§“ûŒ hPÝpNZ?Yp§ä³+·ëXî02RÍ\vƒ½•‘¨>Bý  ¥¤Ó9‰JÈZÔ™™´æ<ÆJŽF×dw¯¾={õïýN¡ÉånH‚ƒf• |IzÍeê¼îЋIG/¤·ôÍP‚%9`â –H N¢±a-ØŽT>¾â¥úNn“óŽU—¹a5ƒðDÇàºÍùtüË[.&‡(öÏK _LÅ(ÂϪÝâ2>o «hš:œ }Y¦NÅ8â¥øÖbýii(Ÿ0$SñD&c•+Þ™ÙŒÚQ‘uEŠéJÈ*Tç½Rrá4ôb£eò«BÞ¬ Ô…½+¨8Õ¯>X•YÇ)nµôõÅH,¶û  ¸E%îãn|›A1 }½o†ÑÂ¦Ž˜¸ÿfμ2ôD% ì–ßL:"IxGë…8ý9ΧF6¤K–%T±¦sÿ7¡—,ô!C!×5¯V/ËN2ßc±Šst%Õ[Ë:ÿ-¥š¯xm!QKSB”É9W{èísí#"Ïý Êà¼ÄãVåA,ç‚t)z.“´´•nó=ñÔšØZ/ÝàÊpt¥›ý1Xo 8m$uË‘Zj#ÀùRrRJ>"Ù{oT2¾ê(úgË%‚fïe™>œ¢~w)ƒi»FI8Î1k‘]iÖÊ17×\æUåîàvNü4Ó­·†Äâ ‹æ4Ì?Çïãˆ~ Û2Zeit{ùÔ5\/uŽ­ücws_Ÿ_xÉ9”€3éCì—|ŽRôˆ~¬…fB²x®^®õ®¥²£[h4é¨9cl:ÂܲìsÙŠegðÛò\:ò£¾„Dÿ¿%JI «·_ñ˵BOÝ3ŒY\Ê—´ÄP¿EUD…nÁp›¬0¢kg*6XéÊÌ()جýŒ3›Ê·æè˜6Œú§2ù¶t™$úðñL—iå17b„økîÒ¢ƒÐ¢…[nLZ6I—×Ñǧ©äâBÑ}£ìÄw]lPX¥_‹Ë‘*?èÀèm}¦ÿÔøl±•?§ñyóBäh­E½>C¤)Åcú¢GHµªˆŸB\Œ!OˆAdÇ|+Îgs˜£=xªÔWêWÓ’$›m’ÓqØqi®"ßð±ìEÙÁÈDhøë4üRHbyä>aº¥qIoXZ!û¥ïÜî|ó…Yà_ê`ÚÌL…Íóµ œ}ÿIP´‘Qxà9BžÂO H«Ñ×Èø@>èÞÿ7oŽÇv¹³ÊfW Oº©?Š#/阣ŸÂâ‚û nÅOÆþ‹Nƒ÷¾uü o²VN='.Ʊ·S+žIðí~—Êö©Gut…òÑA`,îHÊÍPzÁSš\ªÀ¯ë®»Á^&½IªQ<åÙyF,0þË—·³ì%Ú̘½ ßûfÞ-Ö$Tß(\bo|rÞoçüpš±Q<ÌË ©©áLdʸ@à¬z©¤Ð À¶þ¡BÙ¶!¯Í)y¢Á-ÊŠ•‰¿\ p†ê$: 7>—!‘øÐëSi‰Â<‰¦Â«µ¬ÈÍñëU’›éñK.=ÖïzÓÊx½èÄăˆ²+ËÇn‘Â-;uНšÇ»¾ÿ©¬5 ƒV’—‰WYžY´¦ýñìeøËÄendstream endobj 197 0 obj << /Filter /FlateDecode /Length 6250 >> stream xœÕ]K7vF’ld ³éå½w¥ø&=ðb8/ØAf¦ì,®ÔR[3­¾²Ú–¥ùõ9çE²Èûhµ¼p©šÅÇyŸ‡¼ß]Ì“¸˜ñ¿ôÿg¯žü󔺸¹2_Ü<ù¿^¤ÿ={uñù¶°ðf sW/žÄOÅ…θ)(sqõêÉFn¯þmÍ\µÕd¥„öW×O¾ÞüîõvžTðÊØÍëíå<™œ“›[x-¼ Þo^£„Ϭß<éõî{x?{ég­x›ý<ëYz%7÷ÛK¥Ô¤ —ý |–пß|̺£ñU=v-æÙÍ~óû ñ[x-%t–_¦ö¿só¬=5¶³–ðxÓ¼6Ž]djÜ—Æ/©mðÁ -rÏ|úûí¥´Sð0Ñk¤–°Ò\jîï–Ve|¨†ùß«ÿ\1ÿráè¥2“·6ò5‘BÀ|¾¥œry­$|+¨Å Á[< ¼eü¸cÔüæ qÌ Tæâë !žj!Ã,‰¡×KßAG’1A/õDÊ8D~¡D‘zž |™:ѳ…¾sãý–èiäæíÖh œØìˆíA ø3[J[r++ŠßÇéY×b~þ<%®Yáe!“ç³Ø‘ÙY&òŠYm•ˆÅ}ñºŽ½Y‡§[é ±›÷q>ȼ»ÅÏü $÷\2ʪÄ@*á=’(tWF¾‰Ý çÕÒAª¡fï•Ç…¢x)­@<Ôb®Gö¯â¾BW^Íõ}láEmM¿-N©’q¤*q‘¤¯ûwæXÓ版^%ÑÛÝ3ìÁ€mþ¼5†|Éfwó¼ØœhuuåF¥ÚX±¨ EnÆ#uÑ7‰ 5¨ÀÐ{q±hm» “°sXFiíRdo ¾…©Œä€Û¼ëÔ¨Ò θ7é=ùê¼°¢™]嬿‰¤ža¬fŽ^2‹¹å¢ƒ®|Ý¥ßæÑ·¡>›H…%ç°Š’¨¯ã£T¥Ð wðXíÈ+á3Ò4š,k+A¿‰]/L£è«¬‰n$ûª—Ä~e@àPŸ°kã1@›eg4`YŸ¸¡e®éJž¨ZNhÊËÞQü,L‚ëÕ>¶ åÖŒ»î…cð¯ñŸØ·ži"ÀܬqIÿ@óÉdà ûÍƒ·EëÈøBk0ê"ZðÙŠ¡p+pË×û²è_òn`ðAä…ºÁ,x¨O8ã]µ¬w™EèùߣLƒßÕ£‘÷¬gb£áq(…dx44÷~1<\Ò”£Kƒš9PcC|ª Ã(–|[tƒ±•¹džƒp…ëWÔáP9FB a¢ÕIöAÕÀi¸9Ù0ʲr£UìÅÂäh5œï#¼MÃ@‡<äÓ~ÍCÔËÊ kŠÿMì{†¥í˜Çº^ˆ-ªT`÷:k \‰÷1wb¼}(X/[:¥?Þ k;ÁÓ¢dEd–‰ÀüŠxŠèà1,´ôVi]©áînˬ>©‚ñ@C»¨Â%#ß.Ƙ½Èþ®2L4Žºô àðI©ñ -î[¡Nƒ×qÚ.ùrÏmÂˬ€·…l»§E혉dÙK_ä¹u&›“¹„hµ gÿ”é@º›"S5ˆ#u ‹Üõåå»2kfÒNNS™§äT0T~_*j4@90ìEKRkÌNHp™Öt-îëŠÌsí¢l­á==`™Ê0Ž[’¯$ÊN+‘ƒ±ÊÌìXÞS;º´Xà {#!Ѐ„dH‹•Î÷cÖä¨7i¥µ3øÊþb‚ƒa!¥«nq²`qêð"Ê[€‹KlüS'ЧøMžUƒÅU÷ñCH*Z¥½”4¥ù˜Çûø u ºÃP ü‰•°DwˆÁðÌþMêX:ȇ+yÊäø^V0YÀ C/"õ‡ƒaºù‚à$Ë_í9pNV8¾ñfЉFé2xñ í‡ @qfÝ.v1Ï¡“ÈÊ*â"#\ Œæ#ïžÕ\Ž“ÓÃôì‡n^Ç–ÝûX®~•«Ñ¤:ÅËÊY›ãey `®5FÎn‚OO³ˆò¬t2Y`¬SR›^G“¥],ðÕ<+£Ž"›ÚE,Pܯ#ô€úµ‹ BíÈ0uÆ×ÞkOºÊäï“…!FE Z‰dæ" Eá¼—6p¼³5îзºõ’ú Pá}^kƒ)…8“?`È#¶Û˜ØÔt ¸×…¦=PUml¤é,Çz‘ùÞ •Ž™™€x‹²¿,†û-ÆÛa5²5L‡¾zRÚJBoÁÛOÚ›‹Kå'¹ ·Çw f¾73þËm¾Jfd«` dñóDá?.ÑA–9kròRL¼^Sãà‹š¹§„Ø”ä:W°öŸíˆØrñ é=µÑÈ@7{šø7›²©5­QcAÙ¤š¬!®æ_·ÝbÖ:òàÑê*t,efÚõŽó†¥üÙcu“øEB­ZÿF^Ìjˆ>w«`É;mÕ(·ß%%ÇXâÑ÷rZÆøÇ-Š”±V œÊ·Ë²Dm1XÐXèD¨0¼Ô.æ’˜%Ë„mY”Îø²$1L_y‡þžN$Åd=L@÷EÓÆáÎGA¸»cpp}—ÛÜ”â-j»´ÁÜR2ǺÂßó2>muØT±Â¥DŸ¤è0‚‚2‰ý¿—µõ”%LÀÖ3|³)ÉZo =Áš‚[š¿oû6š·†èß8ŸmÔŸšèÅØºk¯AšbcЩ¸i(ëþ ^œ3'~{¼?ŸWöc´2Uoa¼³.û‘m×ʧúõæo[Œ¶¥°ãTøf‹ÁƒÇ?o>[qÔ_8âæîªÉ[Œ˜KO_ÿq‰(çô…­ðàKƒ›húÔ$šVù?í «q€.b†pˆì¥žÅ&Àkˆ,ÒH’}£¦ À|/‹ú, Àž‹ö»šÐ°ƒÝ5ŠdtóO‰Ù|R´4 ¬æ=˜”jx޼âßtdÚ¶2-üãÈÉ‘þzóQëדÇXk”“Œ›üײ§»Ó‚|ÌZëÏÐK/‡j)Û5@È»4þUtÄ …=I‚ø•¨óó2m1u·`,Â\²!NΑe+R|ÖUÜæ×­îTz9ÄæDÆÌ~ ›( žEWÛÈ{ˆY€Žb˜&ÕiT2Ä×@ZD@œt\ÔìñRIC¡n౸҇õ¦ OQbx£dDâqï|í"²gוB­§™‰áß’–È€!XÏ•ÂGáÕ@"Ñ ’DZ¥­9‡Æ™¿é,g%‘Ù5E’û¯È½ªÇCζ•ÉGŽdDl´¡íÝ,”`` Z;,”`I ¤' …©»JÌ ³j0O»-ën6f¢ðˆ‚ ºýCɇ¡jA_îâàöäÒ€¥æý¤Â€õÖÀ>{RT Ë4^wö8h÷¥Î]cüd¤?Ý$à4“Ü(Àa“|¾È`'H¹:¡PmøT6ßj¶Ð¥ïJêç.œt4A}k–øG—ð½CPpžt°爞øIX#U3N»µ£ãæÜÛ²yÁ £E+âø³ÆÐ{Rcõ$L¶¶¿nWlX€µý8ìÔ$ížk?VN­›µÂq|AÐ-#¨–† Ê¿öj«% —ƦÝâAýWlÙVd½½™:á*-üΦôžž…N5T ~ ¯ÍäE³!›mÑ®µV@®æN@Rï#u¢É£›1þ…5À‘%î]kÇRm#«'+ÇÒÕÏoqdÚß­õŠ?gÜ 2ì\É2`ô¸K­·õk¶õ>ƒ9¶Yî}¬Œ›í}5¯@‹¬›º aµGšM™Ô9Xæ Wîú¦R “ˆVw ·&>Inp4‰’ GHm¬ú ýòº’-ÒÝ õû0ÝUqôü€ça¥8äºàÑCF:Ø\Yp›c3¯`øx,áç}[V„IŒ¬¶Œªò”ƒYÌr𕦶-N¡œ*€I])x)ô“G¤šÁ¿HõiùlµÕ æÇôˆ.”6±Æ¡z!V’ê§B,°õûXˆålœÓ¨gáy†Õi†‹°=¾0uÕnùòÓ2Þm™'GTWÅe´YM0øm•mXf¶àÌÌ>–j05oxp³œ·gžfû{šÐg&4u ‚O@S¨àÔ#^1*§®p›P€ª«í¥GºËcVBNÀ û¨0,›Ç ð³A ôá§GaË_4 ¡ô)(¬wægô˜*|˜œüT0ìç=M5RÇvM y•jV?•š/pÑ?<ÆeàÚ;ý¤=ØþJ&ªïî “7"Ú÷bXO« °þœ¼]Ptx!GGP3ÿÊ"ŠÙrÃsV<çá©PÄyb vÉ4Yv^*·ê 8ˆø„x<ÏQ…‰½ô¼.îü¨Ëw¾Ë Øl"Y5,ƒ—ºì·=I—˜˜Ãy“®ný<†â [€VÅyÆÙÀЍxIÂ:¨/'yâí!)2mqükž®¤ð´\Ìë½–nÉñM]ðЯ+ü£ÜëŠÞGË6Ï6ôˆ¡;«¥˜k®œÔ0ôA8‚ ÀrYî 8‚0vÁFæûñ€a#Âhôâ&´“g{”üÍYå4¡ˆÕY…#${‡›süò‘/©AôÚêK2|D`sýÇU×5/•%gô³ðòç… GÁ_ÁZ¡=R‰‡G%§Zy¼|>é`€gè¡bê þþrS‚Íb¡X2RH"nÞ­˜Öžkg%Iûäþ-³yàNx…žns ˤûGùe?ƒ]JVƒ°¿ÙbÒ¼9.ÈÙCìúÁBôWBdw\‚&ÃtDòÏ.AÊ^‚”?9QCëqV;M€EÊ”D8·d^á¶™‰e-æ\«¼M/½Ãq½K*žÇïTs‚jt«VÑý¨Fèºò$õ58dZ]Púz´Žß5&]4òãjàW‡_Ò.©µ<‰ø{F–¸ÝK‚LÝc9ˆÕÞôZÞÇá°Â‰ïå²YÔç¦ë³-q} Eßù¶ˆ£ŒŒr&—*È’ÆíãGóÜT ºvïFcs°1ò°ëó7Ù&×e8Nª|S6—÷œÓD 0ÈFr M^³£î<ìÜgrYÛÍË…TuuÛ YÏGm6ó¶>Fø"YyïÒaFÒ?¦F\¯‹då;wý5òô¹sùJê0¹ÃŽtªÁ­÷d@´kÏAVw…Ñ™TSyì5Óé@?W¦h…¬¬ïþkäŸùÙ7ò?¸7 <³seì¸h}04„­’ü”tà9qcJaÏðÀ â$¿ A‹&S}ôYt|øsÌÇ ƒYï¢ÍíüªD#6Ü›ªËEë›ÚüÃ)éE]n| pÍp±ù³·CR*»É°¥öŒR€¾j-Ë8Ïʳc(Rº¾Âq |!£@Ûcêwˆ«*ÓžŸÊ¥ž:,qÕ*qºNÌÂÑP…ø•:¢€Ê@Ìqðˆ è ˆ îÕSXêŒÏOÄ‚¢ûYÔ*v$¼†4kSH-ü\ËK-Ó…¸³ó:¶Àú¨.(Ê.êëÛå*dª,¹{šø,F®¶]™¯µƒh-Ù'x•š { O ´åʶØ—¯p¤õz!ÏòäÈ-Ð5¾²(âò “ÒƒÙÉI1k0¼ˆ%Í¢S KØ=“ƒn{ìèÖý³‚µáõ1rOËáWÔqñÿhË$÷®—àAqd¡ F@Šà!+ÀPò‰1,o×A9áÆ!?6€Pù ìsYv•Qt%¬V Ma$ÝÂñ]<ׯ¸ï—íú”NÖÀ·ào²ý—‰\ßÝRŽa¯3Ѫk¡§YÛÜCÈ×A×ͤ Zk~eã›Ã[6üÆg„êI@È¢ÀdÔNû¬b¥û6É]mÙ‚™óŒ&‹°çáÄ ÊÊÌô_À±OñÑ£‰ìm|Ì0¼ë’̬¹:üfPžñÜÙB¡µc?tNGá½Á*ïÊü1z6´]"A kuÍçú:X:OPx®ZoB/˜t™õ}\H¼…¹àÔ«êÐ.a\“´˜\28,4xçpׇÐÀúxšîPÀîˆJGdØë.< ÒÔ¶ª¹*Щ˜©Ç ‚©á®ÏVãè–ÝÚS|¶¸Ò± °2“Ò‡z3úïŽMwƒC³ ó+@Ð@;aÔ‡içoKƒ#ŠŠyžóYuÎÚôXYÈUlJ>gkµI • ˜2©¶ð€ qK9FÇ…4`îdM©!|üÕ‹±®Î09>§¯W”]6­ûƸQ¼º™Îzó;4xñ¬äA*óôÛð0D0°—Eêãøœwm1Ú´P_³}ƈêu©ãÛ·ñRn¯ãîÝTÞÜ~ƒÝµ™ÇÅéa\™~i@çuS÷¶:=BÑËbÙÒ¾DL,Ú«ù©¼B¥úܯÿ/×.-e Èaošk’ØAžŸôÖ¥ƒ¿Úpà‚±¹ÜJ”%´.:pß;e$–t~¯.»a¹bνÈ# `\xµÝúžÞ` äËÛÑ¥o^[US¿ÂøÄavÄQÚW±kàܯþs·âŽQˆµåAZ"2&@õe8¥˜“‚¯¶yö_’ †6¡º~m”¡LÏ.î3XåtŸ’ßZhNŠ…#ç~#£Íæ ¶ô¨WÚkÍíLgQ—¨ìÒã²€ýÐ8ÝCh@¦‡?ÊÂoúýj›÷`vï˜å*Û¬ìòEnÚþ²Í‡a›1K~upŒWÓÄëxua(Þ^úÅ6_JñUô^TÛ¼ÕÁ!où9P€ÔwÓöãWØG€æ­qon™©ËŽ D£%|B“ ëÁ«6oËãóòø®<¾¢GãÓݪéñ]y¤…/ÏÛîzïhpõAÒ${ÑpàÕŽÕEÙt•¥!!Ôx]Ehï?&„‘ c(yJÃÂ÷úœøz÷åº;ð]y;ØÊ©n«ŽòG[·ñK?ðÙHµ/®žüþIü%h:kwñæÜŸNR)\iRÄßOúzó%-ËCrV°ëÊ+zw¦tî44 >£žFõ=Âr!0˜Œ3|œ 8$QÝC§ŽE2¡ú63à!ÔÁžx_ñ’EîV÷ëïÈyo]#›Ô–ð¤u*<Óè…ãAÐMí¼Ð²8!G÷å3’qÝx•tr­½?v}=ÅÞ$](MQ>·;Ü ¦)é·,†7Å.;mÚŸIP‡¾ùuƒïÖ=+`îè r×8ÜÆïpϬ-­!ꤘ_qè2€4öê÷úR ùZ °Îé‹?™¸ÖG¹‰^öøm2ü€T4’gP‡ú½¸TšQ= »@ µŽûÍÂP%Ø/lôÛü¾Bæÿ~2Ç(³þ¹¼~vV¼¤»s¿ šs#@49s¾)ÔªS8%èJ“þ|VM>Ê$?)~ØæIšÈª¢Œò‹ÊžÄø]H§W}ÄÆÏË-º.§=eWe)Õç;Ö¿? (›j'-±«ÎÙÏà}—xŒÍÙ%þþÉÿÀz Ñendstream endobj 198 0 obj << /Filter /FlateDecode /Length 7869 >> stream xœÍ]ËÇyr\é’K€ —N³¦Ýõ®²¡C DÊC2l‰ È9 ¹äˆÖr‡æ’yÊ¿žïQÝõUuõÌìŠ6bÔìíîªúê{ü¾WÍ_6ã 6#þ?ÿ÷Ù««_}kÌæp5nW¹Rô×MþϳW›ß>'ÞÒ˜ÔæÉ‹+~Um´Žƒñj\’q›'¯®¾ßþöz7:¥ýöùõ8¸ätòÛ·p©Æ1ŒQ^¾Ä‡CŠÉmïஉAy¿Ý_ïŒÑCŠnûÜÕÑ[·oèÙq´ñžüǕӃ£ÞÀ=ácqóäæjû¿×Oþ|Eã[eašâoßoßÁ·¬òÚ%l‡3K!èíçnŸ]kKÛøó.©í×Ïñ5“¢qžViÔÚoï‘LÑkë÷¦!p½°K^î¶j€¥À>ຠ­ÛÂ’wÆ Ñ{^ñÏ´ã¨SÈ“3Ê»i”¥Qt0ƒW¼Cc‚¡ÜEJ",°ƒÕö¢ ¼è¶;±É4û{!¿WšlpÛ×<†ß”÷(üÓµ0ß”'zgšmÿÕ·¾"ƒ¬J1¾‡ „?%x:m¿.—¤Ka^ß”ËoËå¡\Þ–ËWåòE¹|Y.ß—Ë?mÅõõuoÇ,-3M•öF¥¤B$z›¨}Ze!Þ¤w†xxÏ·MŒÛ}¡Òö€|nÞ~µE3Â÷îðËVi#Eñ¾ðfŸÐjÞM³9–Ë7]â½ëRwß%’±‡éÊaLLv´Ûã^n'F‹6H)aVEzX`$瀑”EÊ œUÛøl•ŠQÒù®ôðyy¤G„à 0ÍîŽVè‹çåò¾\¾í>pÓ]ùÎŽ·Ù)›´¤,ÏŒ±ƒ¦a)¬øè¾¼ªéî,‹89vçüš´Þà7` 2öRMÞ´ú¨&M‚ïZU³ ßËrù¾\>/—}r€ÊÔÚÕ|@Æ¡PÁJö‚1SÁl?-wqp îìVªáy¹±ºýIߊ=0ý¼6Z*ðaÖf]*¥8à¯.D +ØÆ<ÌžxaÔÑœÝE­€Fpk9Á¾Ú{^./ÙF±_ûQ׫èƒÜ/ä8Ü?)¹ Ò #’R3£5^Ë­‘J¾‘”Þ_”í•{wK̬¯Q¢YÞ<£wF¥aŒ&‹h̸â3"‚·v²ty,—o ³¿*—ïÊåm¹Ü—ËϺuj°Ö§‰¤Ÿ£Òcô"­}­g™Ÿ®”ø=-Û‚]O¶·l tu^×Z¯•âûrù¶}à€›¾-EðýõlAxXCÔÕϺðÍ¢‹°žHì„×¾ŽO¯w΢^SÛ?7Ôh¢ÆVºáÍ0 *A&¢¯Dû˜Hi´0óþ}V(ôe¹üº\ ðr×¼ûrùvE2çË>Ky7ø4ïZX®OÈt‰O˜É}+‚€L<›zù¶Ø×—ÂìÞËË®(6mõ8šf«k䌶0«ÀÊ]9â]yâ@‹ßPrtRmg× ªO8ðEu‡4C˜lXD6¨©¢·BÐØ|DŠE,¥ž{F_•“¤©–O°çáŒIr¢?_ãâ÷FBk^Ÿ7@×Ê„öüÌ9 }÷(qã´!@À%8ì áU¥xahPz.t9?êš8û#º,-'¦!Àìã02/ÊÈ¿ßóúÀ„ˆ•f/C›¼K™Ú]ËNh¾`¡dLà.†È6É ª·óÞô6éi_£åÑ<"‚| †¼šÜQ ß´9ÀµGaåå'ïpµ Þœ@Í…,® ÁövP6±œ8¡ŸGÈú~xß²›Pnlµ­F•KKÐüuy‚m2oUåsVHŸ€ù½w÷,&1š(_Kèß=Þñà÷ ñœGÒß@´‚ öo€ƒ2ô5isÆ µÇ=ßEÆÎ¤’BÔWÓ¬à’CgÐŽ6E‹ýW°þ²½GÚ^B”‚"¤GØ”ý=ï(äÖG¤-ÍÑš‰qxyQÛ6ªÛ 5†Ñ5L\æŒPþE«°`¿ü Æ¿›„Ì?›Ž½ørÅ­=E]óð$JàaD†.ÿU}E…‹£¡aÙß½émv«¤„TGè–@’,’ˆ72÷{Ajד×!QAÅZNô½HN»õ@çczMâŸû/òʵzS„ «@GÖ ËJÈgaº²ÎȦ«Š—v!·”ð»„a–ÿa©••Œs€ÿ­di§â}ø€eYéYb>~´˜³^±5þš4!Ο´´;Ÿñ€i$—1ÔäüïPDÝ6AøÌZG~$!úºîÂü®é%?‚c@|‹Ö¸Dy¹žÞô &tJ¹3%Ù_t/En^$äW!a5½Õt$ | ÕaWD„¹±ŒO òB =ÌG·åUA aÜKþŒÕT Ñ(ÖlÎw`J(ª-€q¸m(=Q’üˆ#¥wì¥NÏÜï(ØSéŒFöz“¢¾,íßò·“ò¬àÝE̽Fk4iRðYhê¥~PRÈÛkžË9´~dŸn*ï¦kt?0 ”[ƒ67yQIÉ9 k#(ÄU`ý<`T&s 8©‰ˆës£Õ¦rMweÈ—Äî螺âMå­Ç€ªL0ì)‚8™$"ªnGj‹S˜ŒÞKÈMÓR0ª²˜©.•’(ÎavÏÓ[õZ)Ö«8û"7³Žõ ·§ wuû¶Š°JIÉrLõqô9Œ,¬eÚk멽£Àé6˜§#Ô4“MÀÉ ª¥âüìvÏŸ3«FiŸ'c œåº÷·3÷<Ðf¢æ½\õ⑆,ºâ€ÝæÁÉY2x¬¬ü°¥†>ÙȉÅJù9ß¶—9’ÄíÀÛ ²·“oÉ7'õó þ,"¦}Þì褂~/x·hx1È)Ó•¢5‘vo–µLäEÔ&à¢"OL÷aðnNƒz¹ñÇ·|;2ïh>–ÜÌf`Qhât«œìf•ô]‰šôì³R¨£yK»Y_Qÿ%ugÙ0¡n&*UêXòÆH»/V.d¸y>|SÉ#êä’ßXQ¡o™ú‘ù%jÄ«Ó~ô+f°KO‡êjŽçþލŸÒèm—žåz.ó9>-p¬­ÚEÏÕ٢ⲶAÝÀ¤LÌ7Qf±½À W9„_*I#êǶë! ë=úN -dˆFÆFš¢J|½¥Ûà]Wç›âk,šPì…¸3^e7Ü¡ŽÅtÜð¹ü:QJ-çŠVÑ´•.×ÿ‡ô4“…wzB‡ ÒŽàyeŠã>oôñšÊéJïy;0ÓøY§¤q%€ iŒÿ,L¶2#­]7”p¼-ð¢**)\+„”Y,L‰Œòš\é¤)B(¬Ý‘£ÒÕ%¹e‚Œ©AãBvÚ|Æ,=aÝ‚Wpim5Çv:&tAª€ÃaÒwS@ôd™4DAOwu¸³ Ø,eÿ§:èò›rùm¹<”KÑøø ÉÚô©ƒ .,7˜=útN5^Úp¹âƙ藅<†Òì‚›æ-Rµ=ïÍ û“êLÿŒ´jJ—«_"­ì.ˆps¤ë1Ô%ÝsäݬkŒD`ãBøj›äAí^M1 НÂDò‡¢¾úz"àN/ªÊsD…¨„ÊØ?-÷¶ ¡9î§ÐÂu‹Mn©‘¶•õÕ¤¬ì墧@·áõcéj«sbyú¢ëËŸ’V+åOý8€¨Ð­Š…Wº¿Í¾ÆZ•žÀAF òöÄôp»0cn¯« ÀÐ÷:‡²+¿®¥ •ØÃà=@8±ù0QQé?rQ)ØOPeaÌ¥Z…„^Óàá9+²sà2Á„»Qal¢á–|Tè®Îµºæ%°óiVf¤¹Ü¨@Öz+LA*¿™ýOG Xszø“ÖäÔ³å•ÄÓ îaÇ"¹w_œ&«¯<’ÏÄqŒ¢êSŽ´ ¯ÃÚxÜî(+Á°Nz3؃W‚'u;˜ ì²CÎ9XrÐAcË|yW½óIæØÝüŽCø´v‡ÂnÕ¦[x…Fù×'WÂé~¾7_]éÃýÆå6¯®À¿„!Í|çöê»ÕYßo‘X¾‰>$°µÈþ~%¿‚_ëìʰ©ÃNÚf ÁS¹TN{¾«ÿ ³‘ô³0-] Ȭ؀’[ÕJAcÖƒž LÈ Êi»Œ4Av=hðpô¥á”GŽï<‚#Þ1"L6ÀV-ýÃØ¥‚:„Þ‰§ñ¾Ç´ò"°:ÇAÓ _Ó¨#˜pGÀ?´ÑR–Îxš5½ÂƒõQWÑ%!öë40¬ÿdÆÎ5É¢!”€ÌboÅN*ªÒc^é„=îQÊ,ψ ÖŸºÀrRàCã1EØ, VrÂú-E “žY TÉö™ Ú~‘¡P°Y¸ä‹µ˜'€ÄVê’³ÙöÿGõÎw­ hªQ…Q-m_bwK'¶ºÑ„´¾·º³·µÅ]’£µ0¯¶¸¿˜Ø°*©Jî({%`ýƒ)¦¿šϺ(«ÔIïD 3zÍöÁÀß 1t ¢(MÑÔ÷â]L‹U±ÊÞa ÄÄQ°%¼*L©Ã;¶¶IX0oæ=üÓ›fœ²~_Ðb ås%ÌûÓZ̶Q$Y²ó’ÑÚ*qÒ}n”ûÏ”¶ýº\þ±\þ¾\¾*—ïÊåm¹ýÙ/Ë¥hë>v?–]i¾î»Ò;ãáAˆÕ·i_^äú ¦ e¦0Æä- úrƒÁ]ÉG=Q܇)šj×­á”ÛËô×D=°om†9·¦vè’ŠàXÉ ‡€©¶±/T.k dœ_T=TJˆ@8F¢Aˆy­Ømã¥P­•©‹O¬%ýW¾Ò$g@‚ ûjÁ& ­ÆGâ…ìg\€FE]AnÄ…í‘2²Y¹g ƒ!Xø îêËEÀÒÁ¾Þ3©"Èd‰v;˜Glm§¿T¦ˆƒÈòy=µq`¸Þk,f¬Ý$n¼BxGfþôüâ^œÖ®ò•žÒ¤Ýt=‹•SU«{%â‰'¿¾Þ›(/Ýtdà”@ Uå±wó<úÇÌ”q²Xº­æ”ç/Ðk‹jÎe—2ÎÃ.tDän›¯ŠÀïß•ÉßWoN-Œ¸›>‚rÙªD5ÓóòÚ¡S¹Êê ;Y Þ-; t š#Ë1Ê#5˜äAbQ)j/‹7" +ÁÈ+‘5;ø"¨>yRß%U`›’÷¹ˆ•ðP¨z§yª ²GŽ­êûïÊœ×O=ûCë,ÓÓ)ú¤Wùp–î9æXb`âžtÏ1OZŸ<Ç|ZÓŠ Xž9Î^j¤ê ·5<ÔiÐ>Öø+„hñx 'GAfôHºÇNlQíįgPƒ) ë¨è¥³XKiö‰*¿SÍž ×Ì"+r¨#>uqg¾Íòe4X„JªšnîH ÛÜ0b‰[HP%†‘ÑäFÒŒ5Ej0KéûsÏx­îL-Wh’i¶ûš ”½8ýä<¢´:öl-š€ÉhôÝ5xvÉÇÕ)»¨Ÿumsl/V&åy¢¦êß9ŸÉ~DD¢èØ:᳊”ÃOgEÎqkb¤öŸC*§Nˆ+ANÞõ…îf µ'ÿì¦ù®Vø—Dõ(ê‡$Ÿ°^×ï#ÙRZðè‚K Ÿ5Ô×,/ é%†N¹¯a~èã•߆m¹q:>ÛÚ¡é žÍÅ™¼‰kÌe“61Mº©é# —9“\Å]äõi±èýÎÓ^Iöví£|ŸAèJO˜_cK›"yrlx¦å_­hx6·IïùÎ@w{-R(vTÊ…ˆ—?ã¯aZð#dò’ÃÔEøD#á—:„%Àg®ÿG_×0â6&êq¢Ø9É‚øõ•Ô ¢ø|v·¨M›¾,î¥y[WBôf —ˆÎñ¼LYì2L+GƒˆS¬ë ½X(t¤8ú£„Bc¼Uó)¢kB1 -ãÃÑ™y'ÚS–y×H(ðr¤#ðüS#öÊ󡝯([‹ay¤$2sõ»õh)…ƒ>k»H"¦UæSPâØn!ÚiüÍÓöÞ+%aÙO\µk°VEsæëçùXž1^f8Á\OüäGñ÷8¿–³ÊRß”oÔ&©ÍRž–ÜyÄ šK†Wqˆ¡eQNÒûI :„c¢€D?zò°÷;Tu©ÍrL¿(ÍÌ´eL‰×ŽNÛ¡Ë‘~u€Ú¨b „&ܳòå|¾¥ò“„. ¿¦ö*:~„hÉ%~¬Q9©„pµèßû¹öñ–})J6",ýÒƒ­êÊ’E«=a%Ìr+ŒÇþè ºcŽ~Dá¡Õmûµþê3:Œ_ž"ÉÊöqãr'ÐÄUDìXŽöB"‡J†\*°Ù]9ý#««p*ø,š¢P›'\úÒ.ùå•“¯º™»ÏýzíñˆüKPû’6”sÎûÕûÚä𲂛Oio{Êgà­,fq¹¡3jº€kži ý߬j²C‚òk±åÝ´ÚF­§;çD.Ž¯È²Ñ ÂUPº>ðub NˆQÖýì¯aœ `äj—樨Ü÷ýåucáéy°jÍ-RÅ}„À”¦/0åÝ༥Дúq)mÁ=“_û~-Üw⌬ÜÄ+žèþìˆxtY‹‚¶‚|Ao ¨Zó‡gkz]¯Š´Ì5Eíïhª_–Lª1Ò™‰üT2¨]üŠÌ¹Tˆ×ج…-+!÷eÿ[YDÏ&ÌóŒ3hïtTÕgG`GR©pøÐ~¼m¡q.Ĺ4îÏ3]ú R‘cË«½­‹™¨‡8njcÎo¿ç€x‚¿¹Þ)h;Ï|[)R³#Ã=]ëOÔÖèg {ÄdÒ¢ù¶i3‚§¦÷ÿ®(­þþñËk§oe7&Þ±M ô´ÏÍÌ‹žÏ¶ÇtG½–Ñc§(79Ñ[lgX DµeÊy*ÃÂs¿úlÛÅý£ ·42½º¬A°íƒ´çw.aâºA{šÆF\Á¥ý“Øý&NòçöwÓ ƒíŸØž‚úãÈ{»öJÞÿ¡•²fK\ÐÞŸÛ’v„Ζyû§–1+&¹¶àØÁ&å—5‡ÐÖ{<hðÍ®zçœ>¨Ù¹Ž\ƒu–š~UWéiM£!7–9|Ê‘ <Û©³°É¥£j”=Óü ½@Å ^:ÝÉÛœ†¼Í )KÔe–`ñ ÔÓ3¬6©Ãíž;ð3 ²|.s7FÒâ4•*0¹}ör)Ž©ûñp„±¦°K~æu™…øL÷àüé ‹më!¶¼{š؇šœÙ ÕÄI0« ù¥¼è!Þµ¾gƒI¹òðƒÚ‘Yæ/ïPÈœQ›ž}ˆÝÙšS± Ð}RèÑ{³Âe<'7¿Þ’iµ/M7ZÓ,F?#ïXl]Ê!õ+¹¢†zÎúí½%?+0hS˜ÔD˨ü,x]):ˆŸ~¿à¢* ±÷Žßçrcµ¢äA”¥Wç­k7П°¦Ôƒ€[5al?;shªm?ܨ„樂…Ï ’íQ§Î\+øZ]­ž¯1©«…°úѺ˅UcfËýÂŒX>³³$+¶'lXq”–ñ Òųk­…M(ñ±ñöɹüÃÕÿÎ0Qendstream endobj 199 0 obj << /Filter /FlateDecode /Length 4119 >> stream xœíËŽ·Èqaä[n!€{M›ï‡c_‰ðcìf_ãµfgÖ»+[2‚|{ªH6Yìaïh¥5äÄ‚*qØÅb½Y,ê‡ëùŒáŸô÷ñÅÁ»_I9[]°Ùêà‡~¥¿Ž/fâ #½gžÏÏâ§|æøÌjÛ{©g‡™~s5«æ2Ù!`þáÉÁ7ÝãË9ë¥wR›îr¾`½öÞZÑ­a˜;ëëÎð™qÝ1ÎH³—70ΜpLI:g»X1á¤è®ç )e/Ëö aø]÷YhÖ—õ:ˆš3f™ëN# Å» üJx ÉĹ0U«îYÉgOÉW8Ã)!Yœ í ¿³Œ)&¦€«Ñ°ö°qD‘¹q]&Ÿ‡¹Þy¼È˜)ùÛùB˜Þ; ô¹Åзšñ­Ã®´óÕ2ÿ:ü;ÔTÂ_ ]HÝ;ø-ÈõOð½÷F)ß}TÀm/[òÝeûŠ9®foølÁe¯•û'eÚuo x^@²Ð²9¶­…ì3Ý×aT;7LàIûa®K4„Ñ«&¸¥£À%e{Æ»Ÿ †Ë„e×|ZÀÓ‚á‹æjßvþ €ÿ.àÏÍÑoçÏ‹šÑÎEF?+äó…²\wº)8¾”TÈ´°Cm9îQ†gPÐ5çòÂ9ÙÄ@–ДMPé½_@ÑD&¨ô2¸hi ²wÊh¢”²|~pøN%øÎ7EBFY“ùD$ŠÊlHȨŸ`~-›wË›È\s5Óœ+è¯]:zM²Ç œÏ ¸)`å‚2xUÀãæâ- 0ãìň© ‚·M Ÿ}‘/g–dÔQQY xm¯n‚d®(Ri[s[‰¢î¼ºƒY¢ËŒ¢_W NÅß¿.ƒç\•dÈΨX³°‰’l xÒT+ß+^‘ÍcX‡Ì¥NÔD*‰Z÷vÌåêЯ{#͹Ü;}¶ÑÛž£+»=ç½ÔÒè÷º Êÿ7• 2 ·oK–ózKDÔ»ûདྷâVœqü…÷1A¼€íneXE<¦Œ°èÞrs¯äêW!7« Ÿ"×rÿÊôNñT„”5}úv4+øÀV® Ó-³‰ŽÄ,$‘gf¶`LÞÝWö+ÃúVõ’‹””éÿÙÜwøìךûVlÞŸû†Dë7“û¾Ö”ª‘åZ#q?¿½D÷W Š×˜Ô.¨µ´Dïš’%£DôÄxýLˆÞ6AMçf®¶Ë&²¶è«¼ùµ‹þ¶¤¶:ø¼Énßd·o²Ûÿìv\诲]鵌˜Ci™1ám÷P%è~*ƒ;5d¯¬Åg4M­S-Û-»ß“òt˜­Ó^¥J{@}Jlæ€CqnB¼`Èkê—ï–¡èî•ð<Ò!!82+ßVZWaÆ’ .TU?‰¸µ—ÁÑhg¤RU™ÿI2œ{àAœ›Rét%ð™»¼Á)ÈX15%qOzѽ5Ï5zrÛ0AÝñ<£8-x5¹f5îÅ.Ó{ƒxõ`” Œ6°¸”ÎÌh90ZòB²bººb)÷„s7A¥€–}ìÈÀ 7ˆÀAÈÐQ§@ÖÖ)ºÄfÞ¼¹ ·"dήM='E/1çg‚0:8ðRÁc Ãå„¿ÄrxðåAì5PÞAz>»ºkƒ`àœœ)æz/,v|Ó=>/ÞšF'lŽe?Qw&D„¨ Óq^ý¾6¬Àl´¢ët`Õ~‰ÙÙËÒ§]P”šö‡•c»ÊÑ(S¯rt?Ìw86ÚÖ ÂÎ ³£YÙu ÃÓ¡?ô€ñI,ã@p I¤dxXt?M­jïN‚s”*‘v‹§˜ çrÅTRðVå¸ ªêNÁ m³.$q•â ‘wKul\Á.,¸‚ǰgpCða¼¸–ÝO¸] ÐvÛ`KÞ0¦¨ÇŽ6­¥±p¼‡¯10Ïp ÃÎ6»M¹Ÿª¹…QV•X¯#:ê\ÇcJ'Á +5Ã(¨¦’| d1T‘¶‰Ì´Uù²hLówJO çZáÉ8–ÂEXX}Æt/l]ã,Õ’ á£ÿ4¬K~Y˜ºòUW.· õ@ê•i5d0£r®ÎiG˜C^—…2ªì »a±¸‰°gC¥ÃÉuT€±+1—-¯w•é­ÔP2ƒVr?ÚèºX^6 ÜriR ¢t0f '~Îü²èkê7·pú¨ÂÀTUzᚥwÊYp§¡î˃º­ˆf{ºÍí<˜$ÜÅLÇ‘‹³`ŽìóFÿ 5`>œÒ5ä‹tž¦ô_–aÊ8Æ[¤šië&8qn#JIúàSæÁ­6U~˜¥çFÁ:ðg¨´E]¥ƒ8»zZ2ЖÁßWè\Ï ¤‚`Ó2f€¯’þyNOÝ7;©tbÅ”Ýt#,^ÇïW;’†éÝŠÆ:j“äÑE&ì({Ìç¥^ó^m²ï~¥Eu %Œè…â³–Šñ5˜ðßêH9æšÇsƒ«‹~2у;éKÉ8˜UõÈ„œ‰ûOªê&Ö÷˜¼Ô÷ãªqEÖI|ys8w˜ˆÃzÛʲַó’.Õ€ä/{W„TwÅf¨«x@ãúi‰Ãô–¨:7ÉT ö/EæGÄ)ÅTÝæýT5’Çh‘#±Æý$=îz~7Ö±ºYm]Þj¬O9ü¹{´#J7³½·,JRöt¢Þwǯ¿öÎÒ¦ºè]hsÊ^XHüêŸ;ZJ×qáv¶€%O‰äu°…|=­$È7²÷Ü”ûºGq.°¿iô]ÍhØ Çd•ƒíþ˜Þ‡Q¢ žZ¸‚à »º'ˆ+ ÖÐ`ˆ~óáN!.D9z°Of lRå‰ÃÜy U›=Öû±š=sC½œ´{ Çgyw»çs|Ûý!ÊACøà á,-Z|×&M•nòûV‹OMLlǵ; à ™ ¯É”Iÿ?çÛr¼!%˜.`µ‰¤U˜=;N<“e@´Ipôؾ·Ï`”x g“%z!C ±ˆ*5äöô¶œœ–§RÎXsu!Cßå¸~ô–qE‚m}Y_‘,–LºоuMºôw(DÛã M,DÛoAin[% 7iޫР¥ZÿO!)c—dú,[ZdŸÅ³ì å=ÕQ"Ðâø‰”É"·•SªÊ³O9¬OL¹D¶¤OIÚ½¾ffbÃöêŠÀ2Ýlãñ™(w9HŒ¯è#O£y -u­~ùìNv7þpâ§:û.—#³ÉTœDl:œ[YµŽM™±j6.q`Œ79*Å^ÃqCùEóÚ54깜©=Œû±u­²ÜúÓΘq/ÀÈ/+ltÈn/pÛt{ÊÜ/â*Z2LøŽÃ„Q=¿µRi Q\õà/“Ãüjžû-?Ú)-mœl¿ £©×7À xTÀÓ}(ÈÜuI#'iï|ÒÄ{Õ¤áai[½,&ïŽ ø¤9¨ÈiAö¨€Ê„³¦:‘ÞÚgTÉ2xÞœ@ð¦×£ÇcmÁÝ…ëmrý¢4Õ¾_Fet5ñYS!ˆài¤wSöIžÐ‡7yIdD^¡|WÀë58,£Û&†ê‘ éÌÎà«“SºÕ]÷Ÿ&iWwXã¢97²ÕÊÞÒÂQsèx}Ÿ$¥1z©úgð(ÌÃ1AÐNk¢5M1Mx^TåQÛžaGƆß^­ýˆ˜ ‘øÉ>.›v}Ôdk[QŽšäýj»‚—¾QšçVú_ÎÄ›]½Ø© (ùòmYÀõ·ÑßÉ; #2¿j~÷´‰xÙT¶ã¦lˆ£}ãu^Þë§rQÀÓè_6cæóeå2˜Y¢“èÃ貉ö¼€ë>/ðA™pÔÄ@ÈÙî[mÙ\øÝEnÈåã~ÜHœZ û:’%êmá€rõëÔö“ÔÓû%ày“”ó&)›æn«:H_zEAõ#ŸíCx?OÌwØÄFPÜ=Rç¾d2@&œDÆ r­î°È9òä±ý<•¼”´tB~`I^Œµ‡fíÿ AÍó­É—ÿx"Žüendstream endobj 200 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 3917 >> stream xœX TS×Ö¾1{AÀ©W¡ÖPœpXUìÓV©(*g,U2É ‚ †d' È Ì“È@¦8©Vûª­µ­õu²öom}«Riµºo<¼µþ‚mW‡·ºÞ:°BVvÎ={ßþ¾}0#‰Dº{­ô™¿Àôç|q¢D|n„8I dç“"£·å$fÔÊ—&Í2:GÚl,:Ÿ§‹_Á¼Qè?š‘J$±*÷ˆ}‰QÁA1Ž3öÌtœïæ¶xŽãWW7ÇaQÁ{”áŽ^ʘ €0e }ê¸9bOp@L¢ãŒÅÄì[êâ?O=/"*pùÌ9ŽñÁ1AŽ›¢¢âü="Âc7(ÃÍg~qÛåèáÎ0ÌÌá+#Ü÷EyD¯Ž‰õŒW&ø­ßãå°×;0hSpÈÖаyó,\´4ÕÆŽa&3¯0SoÆ™ÙÈlf¦3[˜ÌVf3‹ñaV2;wf.ó23YÅx0«™ùÌÆ“YȬeÖ3Ï3^Ìf3–Ç<ÃðÌD&ƒ “ŒdfÑb2ÌAæžd£¤eÄì•#DéFé‹\‹,§[Ëìdjv «e¯p.Ü%«ç­*­­·Ž|n¤ÆfŽM¼-o[`{Ù.À.Óîê¨å£nŒv½stñû1߈9vÆoÁ€®qm•DÜetãUêCÑê,U"Ñ>¶W%:"'*'bKTÈŽéÞ†Fh€P«~ÃfŠ…ˆ(Î-.¯ÇÛcµì qµœ«µèú¡Zà'¨ Ã’¯Å$9ÒmIÈ쌖`­õ”¢w¡D*ÇËü÷~çÈw¯ —‚ Ž‘e– öð—“<l­î#8N×Óç{`0‹Ìç}WOTG¿(†? aíŒ]4Í7ôXkàT…ÿ‡ÏI;+yÅ£Á~'…èâÛwî}¿ä:±-”£sÃñ[p†»»øs²H [É{<Ö¢'û zÙŽUþîÄJNf’/y|}Y´*'Ì®õÁK +·/g¶‹Ïê%Íhƒ%h#0‚lj.÷É|²xÑ2žL¸;çãâ¯ïãXèÈFžŒ´ºuÞ¨l—×tª0À[о¯<¢*¶ƒ/·GxÁÎØ—jÀíU¸¶ý ’wÐIÔöHiÆE|-äÆ /¹o>»¦n8 2L%{‰Î$3QÈ¢Z—gj“ÓÒà FžNæ~û|X +6_ ê¹(À¥7Î\¼zs è,Ü´U6‘—ué…P\-èêä*²Í€Sõâ„vÉCŠV>ÍÛÈlÞ·ÈêcOF½ Nzü%NCaÞ2ÑÛ'>düëËú#„Ì*v?pq#',›ªÒ”U Ç*O½ àpÇ” ÃÙH}|…ܯ58kѺâWŠá,÷î¨Fiþª( M9l‚±rkåvO8J#_ƒ8ÖÌ¡hOI„ûùO ²hµ#„Àð„ :𲦲{¢âSBÓ”° fõºá*úŸ€6ð üƒS}Í gÿ;/¯}VÚ×-o y+:¸(*£x{Òæa øb•ä'Ê0Û)ŠSyt6gܬ”iføÎ˜‘ýþifnÅ|ôñÉ‹ }Êí¬gXxè+p­–v‰%­?×û« à#qÿ” >ì‹»6,{)»þ‚€÷Xâñ+OÒØé×Â×} _íR®È·4°^*蔋¬ø<%b?Ž)´\xÜ8G ƒ,P=”tM‡@s½,û‰…‚m× @]Ðn¦ÝÁØ—U-ò‰Ñ_´å Ês .g`ÃÕ)~2¼[xR4 ÉN“O&G‹–o™pñ†}àjØ1u™ÍÍ8”’—ž›Y Y ÎQir^$í ƒ•¦øõj¾6sÇ$æ©jÀ¡˜"t¸êVØ?"e…±¦·5›_PÁÙ‰†L=ú5à"½¤•ÐÏ¤Ç <Ïß]ùqÚL¬³+[š«;*Uû‹}aT÷~w€›|K…»‰Ô¹øëŸœë9Y%ß Š…^Yõa¨©N‡òõZ(ƒZŽH.òs'†¼ê§?ý2Í?ìD“ =¡þÚ:)<ü%8ØÍj5C8( Â\£Êê,õ&µZ¾hæ ´µìyšqlŽùEW²‰ŠíôÔ*ÕpP°oëtÐLE Ž›M+!îP´‡åh‰R2ÑR©Õé®Q©h¢}oŽZÅšõ¬·öRÂb“êù×ò¿•µ{ØòÊòʲªÞmvQݦ3iî·d’÷k !~òC¾|ÛÙó=ô³Ñ@Æyî‚MÑ{塾¯¥…ÀjØ{ö@ gg\É׿C}“êæÎ3reÏæ#é®ÏO%c‰ãWœ³O=*/M†Ì$MNJ¦´N–å«B‹üz‚åûƒ¨h*j÷V'EgzÅÀ&ÎûÝ5¸—Þ~óæ7£Jߺu°N°»!K«(Hí "uX[Z|„Ñùü"¸~ú4\ÿúkðع<É/_þÆ)Ûƒ‚6ÒñÃ>ðbÇÕ+P*:›šhõ'ãé´W_3?È~«—þ÷ñôóT}œOè±Ý0vtš0.Y<†ÿ¦d»þ#x“˜|›x ƒ–¿òpŸŸ‚'Øqçp¤¬ ZRk"›‚éÀ¾†.¿ÞàóA·45ÀýTòíÑòŒÜ”*E#ÏÚë@u-ã.7.9³Oõ1aôT$ªxb…ít#\œëæ½E³ˆ¼ ·Ãw)_ÄÞÑÙÄF)F¥|÷ëÍþ~¡ÁÍÁÝúæn¼L<w¢×m:¦†ª,ö÷šKö;£1gúáP¦þÊ'Ë~s‹Úmþ•¿¢T½ëžXL{muH’ôÕêÒu0¶{ÅùrûZ%ùâü•5±Ëþ‚·#„5¥D9s¥˜ðÑeü_õóÓ‡vQ†ê=~®ád• ¥ïY¾Q !ŠuDjºs&(*emЗßu„¬!rÛ£ûÎÐõh@2‰5í‚Öxz ’‹è€ÞøŒTÔauÑœ8•wrZ†j$S)q‘u´|÷¯Òb”~ù~/ÜãÐÚñCÚø–³ÖÍÙfH®mn¯>ÝZš'tîË«îîÉ—Ü—ms÷SÈI‰OK§LJtØ/.1=2µ«ÛvŸéi—|Š#Ðˤ¢Gñ6aŠï.p"Lœgn«¯ÜPXÚ M\wx£Ÿ2<\é2°•^BÝ>ûnà\Ô÷Ä¥Aø¦áŸïÀgÜM÷+DN,ܼ—îèL:ÖØVyº6­õµ¡§û}Èîx„…eï¤#Edp¤:@ª‰×d«3s@ÙÜþ|¨2—“Z„é>Ÿ˜4kÎçû ³Ä²zM^T¢’SÒ?øŠ¥O‘(ØFÝ}*#­pØagQ##xò!aøÐõEî’>‹?ÃU| 'ÒËâÄ/q¢3(¬† ú¿}A!AÇ_ïjk9Þ)`øø¶ýž E[TKwC[‹0¤¼øäü ´©6]þП¶Â€¸…Ï&Î,Ö¬‹ºŠÛ!¾IMK‰†HnoCt«þxmÇ…]—¦Ñ„æ[aÚ¯Î÷»~Á'¸Š­Ï×ÉÉÍ?û\œgî(üϹ8ªbø¦C§ÿÍC¼°>%t« <¹èD,É,z"ù=‘ÿÿ|ÖÿX¥óì›­Ÿ—·¨„°ôä8: ÇN¨<ȹ°v©%¢w!î®**‘‘¬Á%#k‹ÅU6V­…66(©µ±Õšþiõÿ¤dºÂendstream endobj 201 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 8047 >> stream xœµ™t×ÖïGk†fÀÊȈZèBè½w06îÆ½ w[.²Ê–äÞ‹,[.ÈÆ6`S°é z „v„œ1Ç÷Þw$Û$÷}¹_òîºo±––43gïóßÿýÛGªs'J ˆæ._3~œùãù~¾'þ¡ Ö½•·Ä[}@õœóÙ#[†Ìµ†îBèÞywÿÞ6ü–ÞèPOäÑ‹ê,øÇ+æú„íðð 0ÜeÄ€ñÓ§O5`¸qÓÌöu Úáâä7`¹Sˆ§›¯SùÃgÀZ—n!á†Ïð øxìØÐÐÐ1N¾Ácüƒ<>1j@èŽÏk܂݂vº¹Xàï2`…“¯ÛËBÇX^çúûHCÜ‚,÷wu ò[çéâDQÔªÙ~sü7Ï ˜8?hAðÂEÒÅ;—„:- s^î²<Âu…ÛJ÷Užkv¬õZç½ÞgƒïÆ£#ÇŒ7>jÂÄI“M™:múÇCgÚ ³>"¶û¨Nݬ)jµ’L­¢>¦†P«©©5ÔPj-5ŒZG §ÖS#¨ ÔHj#õµ‰šC¢6Ss©ÑÔj5†ÚJͧÆR ¨qÔBj<µˆZLM¤–P“¨¥Ôdj5…ZNM¥VPÓ(?Êšò§zR½¨Þ” ¥¢ÄÔ{K½O…R}([ÊŽêKõ£hŠ£2)†ÚAu¡ºR ”/ÕC tt£V“M¤:SA_Ažà4Â.ÂüÎÂÎI±•¯Õ=‘–KK™¾L$óU—ù]ʺ:u£ºíïÚCÐcKWÖ›¬ç­^n½.öÜ[Ûû­Íc±÷{#ÞK`g°ú÷Åï7ô™ÐÇ¿ÏSÛ:;¿¾Ýûì;¥ïº¾Ú~]ú­ë·£ß½þ±ýo~ æÄ\©„‘¤ 6`ÛÀÑýÙ rôÓृËß"ráÃO† †þ“WZ·Äƒ m6ñ‹õ‚ª–ÙB~PË26)G• ‘ ”'Ec—ÖïmÃ7¹D­W3>t‰f¿¶LЬªRé>ØDo2–é4™­äêc…@tÖ(´rPØù®…·®Õ]‚=°©TæKÂé•&«f°Ïd‘¾i…¥"k¾L¼UƒÔ`ƒ:]Dnûˆ¢Nè9‹‹i,"7ªÑ4iË¡Œ<{wÛ³Éà -£¾|âü¹¬õk8ýo¿»‚@½‹Akix&ŸÚè/ýj˜öÛý—3â¿Ðx Ìʇ¼”ž¸Ebݲ‰då‚í5Ù É=têùœ—}Ä< â;±HMWkÓ9ÔKô¼fʆ¥[¦b¡äI<û¼èây¸Ê<{¿Ïµvò)ÝÐÊ<$­¥´ø•ìS.؇ï/BvýíáôkX˜-YŽo±¸·È^•V-Aehý¢aÌ¢y§‘XóJ™‘gTÜC»î yoäÅ¢÷†½Á=pï‘X€{cöçPÔã§‘˜Ã؞ݎŇ‚ÀY¨…½p¬ô@ù¡Æâ}pöK˶—m‡Õà°^ê(upôÛ Œeç[&ºÿ@ÈÿÊßgëΞN®ÆDû¨\Ô૵’=º>?,T&QÈ%x:.Æ‹Ñ牙 …»bcòAÎD{(ç€+8Á6“Ö|A!ý5¤Ëò‘;þÞo¤ÛãÕê.“µí…ËP«4oKý©2s¯­§­[Ü¥¦–Ñ&›¯yïš cñ³è{» )x釙A#ñû¸÷ó¡ˆ!yºõû •Œù°x ÞáþÒà˜hï{`æ¬x„hÔíÒ½«·&m µ6ƒÔÚªÞ-@ý¡-Fô‘Qg¢þ7ÙŽ8¯ŠP–U( ™ƒ?Ûµ—þ1ý<Ë{ŠÄˆ ƒD“e>#¹ñ(QÔV ¬†òuùXÂEþHjÕz¦£*­ù~vóæÂ›}Ä7Ð'è[u¢±Ä¼ÛüÍ|öjµ°aŸE¹;éÙª¬Z CŠöÑõ4-¨µœR Œq”¡Ð˜]UåS¹ߘ‹¢·“@Âl4¡Cma®3¢ÁF!«eÛzèY¿+•ÝMR(õð¤½T¼èQS#Aêô™Oóúˆ›ŒRÛî.B#Á¸6s$`kž“ø)E6ÕÆ ÏQ¸1äó>âGÈm` Š»‘0—Ùäé2sªÇÉoÃ8eŽ:9˜H —`[:³R4š¢"N«maѧƒêb"gúÐWßõÝ›-ñªrMwJ'¾¹6kQΩ¾5eõP§”ñ>ZNŸAŒÉ”"ÉO–²KRAl §TÄÅ+”nÎEvÏÚmÑZÏ|ÿr©Dü¨*°BþM4ÉQjŠ×£ ÿ¡Å¦<ò¬é#ŽâŸcs¤û\²€É€ô< šCïMNù’˜Ž»zÊ\` xè̦SE?ÖÉÜ%Kiq#îŒÅñþKð–ÔD|asõ™_ŸŒÇÃþW[ð®Zl8Øáx ž€°š„'¢5qÔ¸]¶û4©[™ Ó#š¦ü|õ%*ÍaÑ`Œ–‰ÔÓ}ŒP0Òoq%{>‹­½\qùZÈOBØŒ’‚ý4¤Ÿ…¨üÔ! Û´mý¬Na¯R(dât[ÜUÉsˆpµve{¡ˆËW½Và¢ñµt4ú$Ës|ÐŒlS£µI9ÉÉ™D™RÃ)Ô™ßiÔ£ÎhÄ!:ކ³c7¯Ý _§<¸æì]eµþzéŽhÇÙ—W>zuåÒµ,‰&~tŽ#¹‡ö52‰h;Œ5w‡Ãçš#Çî¡é©¤ !?„ŸÊ¢EhvzúѦ;;¨ô&ð€5°Ã²!ô>ujÄCd\\B"¾€›mÑåÿBAÎ¥CHw1.NψX;_W5#¬…›k`O;7Óëag²ïsË6·.“ ›×Ñ>ËîÚ³çpi™©¶>¯Ñ,‹@•¯:lòmL›,ŠUé1„æ#ãb’äËgÚÎü)Þ o©v——¡·\¡^¬Š„`¯ hÛäK Q7®|YÛѸÛÖÕ«T;Á·.Óš4eP ªÒ6ês  •Õg¿¹>Ì6]–IÄ`žviNh öªwu® ©d,æB4=Ø€Ü:`+ŠÀV]j§-‹¡´>3ÒMzny##>’×°ùj¿¤ñ‹ˆ3Ï„íö‘á©òYäLCh²ò÷ž8ž^ ¡6°Ô­ÌY·œÁ]µÒÏa§›×zpÒà:BY:($=ÇŠx±‹žçÚ¼ø[sCFYlHöa4ìyDwmœÁï¢wCî÷\µIä¡GnéJHêßì4_(2c•“¡ð>/%vt_ÈO$e[v®I“ïÜ(ìµÒ¶ý1©ó#Èx§V%ÅÅj[,DE¦*™TaYqmñ{7úœÄšë‰F¶ l“b¢½ä1Òuþ –B[+4%§ Ÿ) Ë •Æø:4x9Wòd)gͯ"¬úiÑLzŸFÍCMäÇ¢©wŽfœS{ÔpÎJ¹/Ä0…¡eú‚]_ÏÚû)î=S¸'þ¿÷b8‘kß*Ô=33”ÄAå*ÎoÌ’Xw`Äü†‰ûÑd‰õÛáíu5 ‚/½#Dò–l¼AIÇà™ïXGI€ý˰vX[ Õp¸ƒ/‡@c¹„¯ÕÐHþ÷N)1)D¯vY KOÎBê–¶šÖË¢ŽªªÓ}m™(¾‚ºßW_ÚQ™x¦(@Žý[ÏØ’ €‘Ï h0¢/Èn÷å?d3²Ì†Çä$@8‡¯ˆdx²•Q”úª¨u"Ð#áªÖñ2PA’]tÙtOT€VÇ]w:b†ÕUÔÑ1í,¢nçè(¾×qv1†ýÖ3ó?ÿ^âØ‡ö«Ý^¶Þ|A#0‹Å?GfeŸ‡ƒ»U *îŒh"•,Âw«ƒt»YÀeÁ{ë•' ÖiÏT,6í©>FŒ¢!ÊäAÖþ»Ù®?½Bˆ¢Z&°ñÙjMyЧ>•¢<ЫóÔ%J˜֢?Ù‰:%I£N–ó£ZŸÛfÄjÔiÀäBZ‘„?L—Cãk¢POõ•xÁ­g›¨_ƒ}€¤uݦ9ßb4Ç`C¦Ùø AÑuF)l*q߸˜×™¸Ÿ«ÇïdQ•È ƒâ럓êcÆÁ½%Ds}IëƒHX”™ MŠWÉTœçñ›ÁeWHïçpjIÔ…Ò²–©eBޝ…-nè)‹¶ãȹ/Ä£ñ$ìL°n ‡ %è#49pø[üûîûÊEÙhè×÷¢Áóp.ÄL+±F§,f‹&ÝO7Pç O Ñ\~.‹ú»Ï?? ËƒÄ<.VALH~hyež¾´Î½zÃü)r˜žás ßø³â™è¶1hO:gú¡Æ×BŠ.²h¦uGÔ—?ÿ Áo~w°ð'­E‰ XTÒ6>¡‰0p‹=Bžh>_ÀÎ%hÍ©û%Jl*šC¯¡9×jšOõÿ*CAè3v¼,)×Tå%éÙ%û€yÃBÔÞQA’„Ø"$Æ•…d()Ðïõ¯WóìâÅu‘uA%’Úݵ)…æЩeŠ$Ä29±ù™Ei%±U®¡Žrg'ΩÚI'fÌ‚Ÿ:ÜŒ;%1Qá^àɈ[(ðÉq3…®ŒðÞ.ÌÜë5êúsÓÍšè¦M܆ŠÕ°‚øåvHÔx'ï4)“!ø";tAÓÑC{êè$§EPg˜¼~Þæ’¶ƒ%KY ÚXVˆ:]gÿÿ)MÇ•n zýû:·n ‘Èþ5 f“ ê‘YÍzýCYñ[‚Þ°(›†È6è2µ• æ1ê¬À.þ—¼ŒaŸ—Þ< 7™°è>Æá¯þ@ÿ÷ÙJ_¤"šÿ=ŸÐTÍŽ=à‹` ¸Ô»Õ{—ïæfÁÝŠTÈŠ‘A|¢Z"Ÿç[ ¡)úYÌmÛðó[ª×Zße™–Q76×ÞÌ¿ÐGü ÍçW°x€9À/3äÄÈ iªÒÓ–;ªéHÓbEÊ^‰ø§d~[Tìï#  0š*ŒÅU¤#5KM-ã*lN”žG_Ÿß@r½âmYüžiyžë18ewéØÙ›èžæÊiä Ïj/Ñ6Ëü–¨VÇ&ržk|+ë'ùM8uâÀiGçÜ d)šb/E0âg剻+Ü‹3‰C1ó6Ïê7+åðnõ1åyu­:3”ï½eøMÓh³Ò8ÐètÕGM;Î{<"} ËíQ'‰ø ¼\|wªž¤äéðùßñÁ–¬¸Ê™²p½¿xhàÆþ'¯}qþ é«-c:W—úùûù•WV––Võ>Õo…òš>üÆÀ†µn¡#߫МÖF¬TW¶}Ϥàv‹ÑB[µ6)•pWòå3™éçš´ÅföôSÆ€?¡ìåÚè¶ÆQ¤Î”A „FÇʈ­mù­¤x¢˜>mÀÖb2[$‹üd”=îjDk8¾÷ŸByD &·@ôãgHTÀéôût…®†ƒæ~æ¡ÞªÚž°Uëñ»~f)„qá!šß²„ý=®Ðï’uTk#i£u¿sóÊ‹["lÙ*lðwÙܺòʯ5í®fdwàÚ¬.$¸” ”«äxiëN[¼–×Ês-jåI0ZÎ\жÂöŽ~ifB7ÕúÖ6-,E™ MIÉ%mñê¿­[æùÃÿÍÓõÿ–kF?œåðîTþλSùÃm[ÿ6»-…oƒ…o’žÜŸ¹×ÌÔ¾*5kaiG¿Rë½È„/W&Žlu¶Å“øu²*• ûNA™% -êØ®un»æ$C¶‡a¾-–´ÎÆ“y™ZGHG÷o¿ ’“²Ý‘}k‹­.^—dÒ’sÐt>Å}ÜšÚö¦]Û»Lû/&¤ÐȨ̈ŽÈƒøÅs^ÌaC„?VÿÚCn%¾£µF¡õÎu&uΠ‚µƒ%Xô/ˆñÇ×£4ò0±Ø Y‰ž]?t5£â+¸xEdBxzTœøÙ´ò”  Q%Zɞ…« »)ÁàB^ÒYC«BJ›kààÁ!Zw”ýŸ¦EÈBWüv­^p‚ï#䗴س)9¤èµLV|†,Z§äðßþ1')†˜®ÊN–Ÿ•›œ™©3_(5´Ø{ï¡ô{B~;ºÂÂí¤ÛžW·ý0=ß Vœ@ç1Þ ’fÂ'ðYòøúYû?¾zŽÁõâßW]L¹ ·쯰®°Êþƒì <…Óp.¤+BÝ®e ×^/ž¹fÁ2â eK£F­ñu0Ÿ‚+ÉÖÔ™è(àª5¶¼Ç–‚6„[¾t"x™eõäVÍeLkQAìxÎÛÌ\ Ê탇íÌåMWfÕFYRö6Rø6Òœ²DJÍy„è—_~2Ùù9>iC`ñ öÍk4qgŽÁK;$«Ùc'¬=º‚šìú²ð2Ϩ•\É—ÇêNódÿÇÓ?ÙL‡CBnšV—šÌeå×»Í`Ü–žï®u…ÍŒø-i+ƒ‚]Ý·'[s4â´Žø–êcii„Ñ+Ú3qÛä‹#Qg‰øWļú‰Í<ÎKŠw¢æolªQw©áªù¥øh4ÿ€=|p»gx@``A@­>73•`V£Ñ£…h…wâÜeË%±±d…J&1E‘’uï:êÊ¡ ïÿ'—Y2Ç*¹÷dÌcäútÉã>âVd‹î°¥4aÉ^µM¥ ò.Ê|J£ ù»2öžqü|~÷#ók Düë3ï~ÿ7rxÂXȉ[‚û¦Ð­ ꆯ±ËÀ÷Œ$èpèÌäšh8®;ž~,¹$÷Á/Ú*}u˜„@eBŽ˜ÿ›Êoc·íRì†æì— —¯œß8wëÖ•žœ!Šý¦îp3\`žN89nØÇ³Ç„k칬غ #!Î׫ý&¯Þwà«E¿ «ïïýZ+;èÖÀ…Gf.ÝÇ sbQÏ™a¾°Ânög§OžûŽSUç¿mlKíÓû%üsú—cpïïF Îœø×³PßPrŠÁ½>e#aøœ ¯ƒ#`6ÑÆâºò#{rj,Vï(pÈöN^Û˜i;FNä,†üÛ=Ÿ·þŸ¦oþ½JM|?Ê7îí9‡úÑ™ aË2ÔƒUÎXêåëac“÷žCG÷}}èÖÞ Å5Ì$ä®Z¯–“ö6µã(ò1,(.)Q­”¬é°pû†˜<Âm8"úØ™¯¯“ 8yoÞòÙë:¥Î­™Àé°-0ÖÛMº»ìÅ·§÷W×qWžŽü®Ã¹“ÚL«ªgn+ÉTÏÜßWUs«<)8Ûþ k°ÏóÝ¿¼È¿vcûÒ5Ëý™òüØÝµÇš ½"»}hÌ>Á‰–Ñ–¹è6›Wz}DKZE‡AhhäIøïñ6·í“(IëßDQmŸäZ~Ô7ÈLp5™PŠ*¿ùþšýµ>bG ã§²ŒÎÂ)æÚܯ†:a ózÏô¹ù[o/Ò\Ðæ@ÁºB³!òÇi$º¥¯?î3q©çŠ \¼:̦Üdn jÔ-n©‘Á f—x¹Å‡h ¹,mQ&”2†ÈÜЈ°Ø —:S]1‹Ã‘‚o¼QwÌ5Ï{ïìèA›ñÈù?t"³·qo¢æP½ e¢îWÍ/fs3òkÙ²ˆrÐè8¹œS«Ôj50*ÈM®JûæÜ’ls¥ë˜´¤ä$Ù´Y¸ëÊR‡ý5…åefsûO.³æmÛÙäÅ)!ºY3˜DàÞÈÆ­4М=hê‰zç´ßD¡“Ç_ˆÅ«¸M¸S,P™…{ÇVgpÏK2IMÖ¥I~»5Š$÷n@:nÊPuÈj%ê9ùË8`ä …œ OQ¸äÇá'ñDÀn€g¹ãy¸¶‰iKšÍ5¡NÙH÷ÍæÄ_x)xýòÙK!räÿÉ–†–ùúK¥þþiÅ.}Y)‡Ö¿_!5øû†ëwU \lúìGô£Z n.{jP {8¥Å8KpÉŸP§Æ6ñ‘jszíÄ÷Ô*D÷Ã] C¾x°ï‹]µq!zÎY%÷‡HÆ«,¦¨¤$×éu ³Çàn›°€û«`J´†×[´VL´Ö¥MkŽö|p½K‡Ôò³SÓ39­ÖJ£QG/õ_áh/‘Ëɪ™eߺ…h‹ß” ×ÑñÁ‰ç'‰ÙO>ÿëg€OØáøƒ(e"ÉÃ~BýÆáTœŽN-A øñlÐ|fwbÜÎì0ÏÉ·Û`>Å@ædä~áÙC!¿ÎüS0m")bD>o<ø‡‡ƒµž»×®rtS$q~ukS€…™˜ú« Üÿô<W™Ï“¬èçw¾<\NMG¨BÍc¶\TOîxˆe0M·O†•<­nJÚ…Ò$ãþPŠ«r‹"¶*ìÄ÷©¿ª”‰k\¢Â¹'ýf`Fã.D69 AZ‹ÛGž—÷ž.­Iò/ç\UñÁ ÒìÈÂ$ÆZjàç ¿¬Tƒ;gÓ¦®»q];OÕwïbÈèÞý¢¾{Šú?çäl±endstream endobj 202 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 2214 >> stream xœEU Pgîa˜éÖÀ(ŒâÕ3 D¹oð@ƒx`%# rÇÈ)883o@!ˆÊ!ŠÊp…£Ñ¨hvã‘­‰Á$»)4êºëjÔ×øSµÛbínuÕÿ÷_õ÷{ß÷¾ï½–P¶6”D"¡W®]âíõæÕ[˜,¦ØS¥@ ‡û‡ódS©q+ü¦¾?<3YvR°³m›2f¦#Ö8`Ö8Ü3ž²•HâS˜VÆ'è“¢#£tj×07µ÷âÅ ç¨çzy-V/HŠÓÆ©×juQ±ZxØ«Ž‹ŽÐéծˢtº„%žžiiiÚØdø¤ÈÝæ¨Ó¢uQꠈ䈤Ԉpõêø8z66Bý©ÇÛme|lBŠ."I½6><")nST„NKQ”Ãò¤dvW˜—÷Üy>ó,rµ¡¨éT0µ‰ÚL­ VR«(Ê‘RR(–z—r¢&Rc$fÊQ,eKeR¿J,’omvÚ|//­”Ùnµ‘­—=“è1t$ý/Á¨]xaMµ¤c¸^*Ä û°¹§s`ÄÃÁüzbùç{†Lc¾/0©y«¥·°Úà*´/ÚÍài“TXŠÊ,ÕE´—aüÏÄE6G¼Xô:¡~ƒæÑ‹ïÓä‘âœ4ÊHº\!ÄŠy+pƦ>Ç—H½D…“²Cñ>‹2š,É—ièë%®"×WkèfKwa´ÂM°ŽÆòÁ•´r°ÿ*žrf³>â˜]Mi M5í—77lå”Ýþ¶MÈ(„m«0¹É±õ íw˜#.NÊAŒDžÝ]²¦xæúC¨ºÓ·©˜³dZ j€©ƒâ“ªŸè*8¢?h‚ôp.±"êp0±ÄˆSXuBõ>U{T«a0‹QY w²7Lòó„”i‰9\ÁÑÈ& :‘Ðéw¢¤*k8 ÔVöÄ\‚j`ÑÐ¥3­9êŒJ9Þsøƒ i°¡ÿÍ£oµä9Êq,/}.LcI=­þ5建¯ p—5!t@LrüF¸y’ÃQ4â‚ÁÚ<ëSWW#I+„«k,ëÏ¢»Õ± Ç|ˆèƒr'eª0vÝ ;þàzåÕŸCÝÉlÈV­³@œd0†Zu‹¸ÅÁy¡ÍéõM5_œM?n,äÎn²|Ìf­¯*šVò«Œ¤$8OšãõßrÊÔŸ€ï¨ý‚Ù‡Þìz¶gvRHXsR8÷èÅ"N1ì(¢©¦ZqãåXŒœ”™¸Gpf±‡&þÿyä O_¸*S^È/½°yh2º=¿*œèþˆ8«”™+`[\rs˜ô±xˆ®oîmè@ìý5°)1\·kGn øCø•ý•b!K í¸vI+NÀè Åa6«Ï0¥Bó_z÷sx# XNh‰æ .8‹LÉÑ,د7›² ¸ÄÀíé!Àଞ«€ï"Û|Üh¬P)„Û¹VáUÒ-ÆOÇ Rá4ÞfŸ˜1ŸˆÁLòÜ\-$ x ÷=åHáÙH®ÍÈ»@܆[¥—>ê/«ƒèH¯ =¾¦Áb¯[³<³bA$qxÁÀwUK0ß‘â«a)[ e‰¹&8Ç}4RIÓTËAŸåOUÌÈQ¢ÒÐ]E/á‚ø¼„®Ñ>!júM$Ñ×pZ+/¹†3q5rRá(&±¹1-¹?șΒóm—:»Ï¡ü—›<_ØR•x$¯Èl1“¹*ÒMë!»¶¢ðPi ×Ös¥¬˜!˜·T¥|½f»äÀâkZÎZØP'r¿üYeræÖ¼u î.Àñœòúüã>*F½Û‰Ï­½8þôY8\‡jœâ¤Á@|ÌVÑ€òšÎ’Ê:þŠõk8ç'u‘úˆe¢OÆ·µÕn½®½1L!‹ç“bNù–YÚ±š÷ý!¹_¤¢úû_qª]ˆ½J9’"ZZë˼ 7حЖ,£ àƒWø¾Çà7Ë–»p£˜„ŸE8÷6ßÁãw6ß{3¢· ÉlÆ‘øãpžé¿_ÒþËÅ]Ÿ&F‡Æq]Il¿µ± ¾bžyu9Qù/›z"ðn(§Z¸;4hù$ÑbTˆ¨¸ÏÏ¿¡ R ùOÏpÄ7€mÖó }{»’ZRa;³Ð=#È-èË»w/w6W‹‚IaÍë Ñ Ô©0ޤ°qõk¾Öâ„Äs~a¸íKô£I++-(fÊ÷åïÓîÞ§çˆñÔ‘@±ƒ‰Ã9bÃ/ú†Ð>ª¦´¬¨¼Øx,*ÅkÇÜ,G{,m‘ÜÄé·x,ãoátép?>dq'áГÉ!òñ!ÛÈ.œAœ1bǸáNnDI¾aƒanÕ–û»oü×a ôçö[½wëÿßÃ1]+[Ö•ùÁXS0oûjÍG~É‹Z¬¼XM4);¯!ùóXÍ˹å†#¸³¦¬\Nv”ÐüX”¿Ãµ]Xm7¦½ÄÎågìì-v Šúm%€Þendstream endobj 203 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 2276 >> stream xœ­ViPgîv˜îVtcG¼zÆx`Œ÷£Æõ@t 1ЍŒ ÈÈ%38 —"r L¿Ý"‡Š:£ ÜŒ0*0Þ'š±LÕšì&Ù²VM©1YcmÕ×äc7û1¨©­Íý±º¦kÞê÷}žïyž÷£)MÓŒ_ÐòY3û~ÍRFÓʘÊXà ?ûö„¨ÇRÞË—ŽÒ3!É fPæïPÍ0”ᢆR*š6¤äù“3Mñ±qfíämïkg-X0ªvöÌ™ ´Ë’ô¦ømQmP”9NŸe&/;´ëŒÛâõæLíäãÌæä…3f¤§§OJJ™n4Åþþý©Úôxsœö}ŠÞ”¦Ñ®4ÌÚà¨$½Ö=çt÷ÓϘ”œjÖ›´AƽÉ@Q¿Ì`L6¥˜SÓ¢2¶Åèã×íH2—¢Þ£ÖP!ÔZjµž ¥6PaÔrjµ‚šNùS+©j.DSЧÞ%lPTõ„ otøEµ^uÆc¶Çuõ$õf6#3_±ùl·žû~àB¥Ø«§\h´K µÑÊœž¹|^m®” ™y;q\ï>…Yb±¡8Ýb†,àtl½ä”OÃY¸l©/zŽw±ùþP%I²µ\ÓŽTjdd\x´ú#ã°þ œÐÀQÜWø!‹¯)ŸðOñ5f¼”ËàRTú)bÐ|Ĩ” t‡r«ü6F¯1âK/ÍWëØ;e£ÁwVêØ:©[n#_¼e©s·^‰ô,R}åºXm+œ#àS:Ö!—Û¡º-¶þÖK)ÞçTÞqÐuh8Zƒ†«”DdäѸ9Oñ(ÌϘ'â Ï¢wÏ“‡H+à8ˆßúã9ùÍpZÁçË›*ÚNŸ8 ç ÙPžXžá°t–Ÿ³Ý Aƒ7¸Ð8‡2ÔI?$ˆÂH#dǾüJäÏ4¤´îm úéÏèä;ýkìµÇ°]sŽÅþ} *-ÇŠ!Ò¹Þýÿ u¥ÒÉ×6¶ÔvwɈ½±6lÃæÈõ /eI?‘õy©P(JãFêÇ‚(fîÝ”^_ßdw^9¾%*å©‚øƒþ'·­R§ÜmðÇ×§5—'ìs(£›èæ/PÐ*ôHỸòÌüôœ‚laW…ñ€8¿9¦q›Oí¬Ú«iJlËïÊsæßÞ A\HDV„ŸÁuw—PtH´f—ûÌL³w´L†ª£BiÉ¡òR«Óxj¼Ûî?¸µ£qÏAMBslYô]Eh\ànß—Ž=* L“)K.²W¥'äQûœÈÛ†’]h‰þ1O\*´QÇ£!.<­Ó1â|#f&q±ŸâS,¦¾IûìËú{ÝÂ5](º;Οȡ€ôµ·DW•`þ )¡¬_\pàjñøýÀâÀ_éÊf}?7¾8yîµ Ò*ùïW»Ø h×(a­¦È® uÑÊ*e0¿¿Úºÿ6p.Öd1ˆ&H„HyDèÕ±ebŽbQ®f.NûYÕ7\ŒA ´$ÁN“ãÝEwX4¾w`ÅÎJK5Œ¬…’ÊýÕDi»ö9Pd#šÖ§èhvŸuæ¡+üóe7±wa™Ùšqª±É~¶)õˆ( ö’rpÝ'ÃWibY<¡‡÷¸ž%>¸}¦µÓ.¤Á†ï„³Œ­ jl¹rޤÑYá Ø9Ì\å礢ÍgÚÐäSuR&”Õ´xå¢Ñ $óÏÐðŠƒ¥e_ÃH›h‰3Á›dc?@¢h±:׊ÏRD©»úÀm²Ä€¤X¹¯à‹†ãàqx➌⥑:¶A¾ ‘!à†ÅÞŸ,©HÎ>ð òðyˆ½Õ[tLƒô©|ê¡ÃÒâ®XEè!îövÒÍÄ݃Ð0º¬Låóòsvƒ™{CÄÝe'?Æ>ófb-žüb)Óˆ<*íܬ"qo‘Pž¸ baÇñÌöÌv±Îsw }z´ÒR\-xý@ ö˜`~E©"-#šÍãuŠ˜w¢Ðš #±^÷¬·›TSŠS‹Ó ¸tsZrÉ8 ç-÷°él¤É’É ç”‰EæÞ‡¥Ù%{í0Ò%K«Ð.å¥Ôû—i:¦Õú-ñ;üZ݆œ¦<~æÂ1ä#$™M\¿x‡;ÐÚknìa}úíêYþV¿!lDk´58<ØwžŒ§þ°©Î´¬kÑD³(ðWÍþƒÅLï Þ»ÝÛÁP›Ý–Ó·Hò9áJUëagsíhGÖ)=éº8ou3Zø:OŸ«”)ÈÀ×–IucÑ@<ó³æž'þ}Fuþd?š…»,…{ 5)ÁkÓ¶‘y< »¥iö.Yû¸ç€Ç›5^¨HëåµÇh&ýQ*¯ü“ïˆm‰Ú·ukkܹ³Ž–¯Æ~â­·¹Õ(]ÏB\zÍrâKDq9Ç¥üë}ý’_¨”=³ÜÉ•'Š©f!99¦yË‘µä܆ùÏÁ!öÈS g¸#ÿQê‰ÂöÜý– ©G6A4·9}ÕÂسi ñ(”¦¾I±wŠ•KÒñB]ÝÍÈûí0½ûù÷hPwF—©Eu#¦üC»±,þ@NIüAC ²ó57 _§/¬Y¥ÄЯ£¬µòøZ?äÈž¾#¸Ï°÷"ØD6Ú9²yº^kžl´ž?¦~q¾É-wvõÈoÖçŠð˜°d!ãÑ"k"øÁÆhÃîÿ´Kÿ·u{Ù(Ù½‘Ðb‚KI ¸òÅõEÅPȳî<ÙD žxÈV¬Öø®ÕÝį~ +z…ÿö?,v¼ÿ‘…#ÐÑÖè¨;O¢ãY¢e–ŒÝ䪒{KÒª³Kûúíæ&õÝÈX´Kñ·Ñ¨¥/35h&ÿ¥k‰²€©ËŒ» !+WÀݽj*Äô²Ý<w ÝmÀ%Ä'äþ‚Ïá­¬WN¥RŽ"ì• ÞRƺ!j°0Èc¾Ís ³ÌÓQ-žC$O/Šú7ãýœendstream endobj 204 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 337 >> stream xœcd`ab`dddwö 641õH3þaú!ËÜÝýãïO/VY~'Y­Ÿ*Þ|Ý<ÌÝ<,k¿ïúž!ø=•ÿ{’3#c~i‹s~AeQfzF‰‚F²¦‚¡¥¥¹Ž‚‘¥‚cnjQfrbž‚obIFjnb “£œŸœ™ZR© a“QRR`¥¯_^^®—˜[¬—_”n§©£PžY’¡”ZœZT–š¢à–ŸW¢à—˜›ªq¦„rÎÏ-(-I-RðÍOI-Êc``` b`0f`bddñÿÑÁ÷S²{Á÷='¾Ï8Ìøúá÷[w˜˜¿%úiå¹KÝw8Þýf|ô[Yîwë_¯‡ß/~7bûðû"ë¶?î?øD¿Ïø®Âþãl€‹‹wÀo.y¾Ò…?æ|_8{!Ûï¤éì'¸nrËq±˜Ïçá<ÑÇÃss/MØ|[endstream endobj 205 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 3240 >> stream xœµVytS×~B–x€1 G¬Pôœ¤˜a/'ê0‡5€c„6¶ñ"Ù–ÉŽmíï½ß“,YÞelË–É ˜%Äì4@È 0Ã4m–¤ $…´¡™v˜ût.Ì•íÎ’&93ç´èH:Ò½ïÞïû~ß÷ɨ¨q”L&¿zSòÆE #I3dÒÇI3圾ާ˜IMI\9sNø¹¤ˆ–CtTð‡S'NC'þUMAéS)¹L¦-²®Öê Y™ú¸Ùûãã-[¶tnÜâ… —Å­ÊÕdíO͋۔ªÏÔä¦êÉ—œ¸díþ,Þ7;!S¯×-_° ¤¤d~jná|mAÆŠø¹q%Yú̸mšBMA±&=n­6O·95W7zÐù£o«µ¹º"½¦ n“6]SGQÔÓ«òµº‚µ…Iú¢ÔÒûÓ522³rr+£çRÔ³Ô«Ô,j µ•J¦~D½FͦR¨íT"µ“šG­¡^¦ÖRIÔ+Ô:j=µ‰ÚLY¨\j25ƒàCEQ­²$ÙqKÇ–¯_2)–(|ÊÅÊ_?JÏ¢Mžš`8uâƒIþèiÑ'ë'÷ÅècK\LØ !´3$­k“õ…×Ê¥úð:†kªË K¾òäalQjfþ+À›„*ÞD·‹ÇÇ  ®ó=ÂÙèYž Â¼µú„œ]lÛÑÎ3¾²Ç‹œhNup¨ ÷–`¦]ÿ Cð6 'x²V¬u5‰^ºL¹ø:Û7Hæ2þH‹”1ÒyI*¿LÚrCŽŒèwLË#<«Êž£¸Ôd×™3pŽ=—å ¶,·jõ°ì9±âgtÁy~ r4wH±SWÍ«­9YsÒ{ÒÝ_{úÌ©`÷¯è£m%»Ëù*óV6 ï¶iÊ`¨Èþß¾‹Û÷¥ËÇ6߯îò†Õ‡»ºÖÎOY•_®Ž ¿N ¼îGGC2éÎ#9š~‰±ç4A/ø•ŸóñPVÞ,X~üÄë(ä*sí¥\”DN,ÍP¢!ÙãG÷..©e¿n}ð/pŸ¾ûÒ £ÆF|“9œî^Ñ{T(Zù{ߊŒekâ ,žï1ÚRhwÓÕÌ£€mRþ[Ëâý à(6FâªÒB¿ì½Ûhø×r©Õ0Hþ³1…ŸÅ1˜ÅªYþE!bP4z^뱕Ñ.­CÑGѸÐ9öÌí›Õ-@÷Ý)ÇQz ¦2×±9Û^ËÙ4!o¡¹~iN@&Ü•£b¼”ÙhrùëºÑ‚êN¶Þп÷&Ðh:ŠEÉí” ¿Œ_ž¼½²Š­¿†_a°ÙQ8‚›ð$Ûº¿6Œžß çøÁðõR7óöÕ+mÇ~ïíWágç`å’䬣ïTëÍzø™éÑLËŸpŒÙ ¼`Wç(žŽèD+g`÷àv›Æa_*AIžsLvö‚î }:ëÄzg]ªÜ\éñ7õ¡—ÝmlË»Ç~´0¾ÌZe€:7`m t¶ûtévmßQHh#´Ÿ ¡áYxŽõ¡˜O”ׂ•; œ‘<4 <‰) ”å Kò½TyyÝ}nÏ)B%­D3{Ÿ©€R¡’ü±)ž GŽæWþSƒµ˜ž7³‹ŸôÚ³ùÊå *Šü4èúŽ’Ù¹Cܨ,³• •^¿›(ÀGQWù¥[eŸ^@ÉäÒé§Œ³œà¢=œÛâà¡¢\]i&xØöì…  ñ¬‹ÞÈòéÛŒl·>h;e>i~¯6ÓYkžÙµ¥M¶©¸+ÁÇôøéXÞ <8hkµ£Æ-Šmíj»©ÑíîÊ2%„¿0ôøüPYÀÐÂæóÝ)u{j“ëá,-§è­·`a_œÚ\+¸êEO‹‹@( ¡·BÒó!¢ÉÃ[ÐÆ3èv_ƒQgx“O2@ :Q ÅÅN22!å/Z¬E¼—à$Äà©è§ŸÜ8w˜­/êÛ_O–Ëå#â%s5=„~Ò&C1ʬÒs š¥Óž¥ ¢„ô²yÛÿaÇÅ~¢Üƒ»8Á­B5çŽ%x*8¤>r66~L(+ù·*ò-åút Щ¦£½õbµXCØ—úÝDc¿UŸ’8E·Ò ^‚ Ùp …våõ:kñ¨“lú/³ zé á-UÔÿàp@vãÒ=”K?“¦3;x{˜èŸžÖc×V×á™8?ErEywåÐÓçÐäZ¯Õi·|¥M}0iÛÁ­@ïÚ4ˆ(4ãè©?5]æ3{ÇT>ptBvã–\z?¼žÁÓ¾E®£.ž³M¾¬S‰dP§ãX<¿€'ýæå‡w.^ª©aMóº]ßÃp¥ ‹sìÎÛ túºNDµº‘âÌMvèü…’0w:mÛ­bGïœÛŽý²«·Q À•¨ŒéŒ°Ôêòƒ<4š˜px5žäõŽº»üßÑ”DÊ–j¾Ælá&»ú U: ß|¥QAÏ ¡¤ºD?þ²¹èÏ ¾¤H°:tl º<¢Vä&EgÒ=9â#~¢Œp†'(›?÷žv»«Ú»zü!¨…fÎËi-Uz0ÐeµÅÁÎ@Ëà‰´¶ô/n™oT[öÚ³­o°øH™‚oY÷Uº‘dýFâ¹C:’§Êè$²ï€ˆfšˆ·q8­m»üIÚsKÖt¾{õ_èO.ÕJ‰Œ”ÓUÞòÖÆ¶º®:[zÆ>cn–ÚT—Õµ è9‹W,Ö4ë|FÖbÑì‡× «Yß\™Tšöì¡×ü:ŠþÍ>†Æ¡ÉXñêdežh @¸Å`WW_þ=ô£ÏnÚc;ؓdž¡‘dT3n`röºâuÐ9ÄúZzî®Ø`v$éÓÙÒÌŒRbxÙ1ó·œ<ÝÑØ?ÐÁv·ÖÃ@ßû7/I~žýs“ŒÈ,R&%7z—q"·öÑ‘÷âöÍKÔ‚x}õ=ê[ò¤åvˆÑ’€‹|ßH þÿVÍÍß»OµèÝDÁˆ8N ÉehíÈÑPXÎt;rµÙ…yíÆ®PooHámÌhó+>¿£øô‰p†Æª{H±§:G ÄU‰RŒ<0,ë“õ£F9þþqü ýä"脳üῼHø—£u-|à†<Üý¿áOHÙ¾7_íølMÝ^˜)«Êó1–tý)ßùäoÜêoKÅ_™R‰“®ÿA†R¿~ðµ] +˜¾±8» 7ßИßùÔhûô@ikA®N—§k7v»»{Õ£&6ñƒßö£Øý@.µI¯vÌF’š£Hx–‘y¡•­Háûêáñ÷Aå‡#v¿–~Ã| <ýˆàj8^sþÂîÁ¤¹xêŽg l)¦ ¶ú»°iìöÄÿ›?Gwh¡r¨ÎZ(€å/ùø^%  NiG+Ë¢(9ÚИïzuwÇsi–6·ÝÓ„ÆýMi€«¯kôÙy¥ ùƒ­¾Zo:¦È/­nAyõ¿§5ŒM¼5I=1ji[ômt4Eý'/ÖÓ?endstream endobj 206 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 2892 >> stream xœ•–yp÷Ç%fCŒ!IÕ±B²+J˜„pä¹j „p΀±eK>tû–-Y·ö­V’uúöbcÉ6>°!6WíP›3NšiÒ”œ¤“’¡¿u4t·Ó™´ý#£Ù‘4»z¿÷Þ÷ó¾ObÑÔ)"±X<=mýúµË–N|\ÆÏóMá–ÎûûÌñíÓÍúõK?:þHA $K yjÛCÉïÞNÞ‡B³y¶ˆ‹5v7ëפi´%zU¶Ò(_±P¾lùògÉŸXºt¹üå|…^•‘®–¯O7*ùéFáKž|‹&C¥0–ȬTÚK–-NÏ7,Öè³_X¸H^¤2*å›…¾P‘)U£6ÊßHÏWÈ'³]<ù–¦É×zùzM¦B¯> 0¦ç¥çÈLת ªìü‰óŒé"‘èÚ¢uÅ%o(¶l]ñËç’¥²9÷‹D[E E;D‹D;E»D«DKD¯Šž=-š%ЉfŠi1ˆ±WÌŠ~!tI4UdÔíåTÙÔ+Óž›v+i÷ôÙÓ¯LÿŽh$>¿çøŒ¹3zî½ïÞÞÂ+-ÜÐM´àê NŒdÃhá°„ŸÃÏjè¢( ÊC&Žk®ëØôEœš5ßó×L 9X9h­„Îø,N.†ŒägKĉƒeÑB]~á[›‡3>EÏô¢™ÝÅ-æ(©«ýûBÙÁŒ* Íço¶œ¶gÖ)ãcÐŽ.´£Îvñ¸¿% u{ÃW8„ƈÜ”ÒÜPÃ*V §“çÙ,Pd Aƒ·|æí3Ð G*ªµDÊxÿ$äãnþº4ÜWß~ƒ%üQ¦p•Jw¨@ÉæLðàˆ¶„!H Ó‘B(·ÇD»ñÔ;›R±Žo°ÕÓŒdøÁ¤"ë¤3ù/ ìLðZ™Qš»Â-sXèb»ß¤: ‹=,¤Ðí>üoÎÂVÆBmòÚ8h„*ÆÏFP*ß’ês€Ç_‰¼w>M­*e=!ˆ°þzD…Ð-‡žãÄÇÇPÝa ڀʤ ¸æ=Ýt¹Íª‰~ÅÑW±?ö(N›vå·ÿ§/Ï[?éˆ2í5ç©'íE‹nÕâD«Z5?‚Æ ÚÇ—ÆÅhõedº,—¯“²ž*«“¦ËÍäîͺo¶¼&TýNÆÓñÌC{c¢þ¦é±ž1ÁrÙ²å/\‡%šÝìof$ãcða;ãtz„i&]®-¸Ü„Ãï ø¦:Jž~{Óûå# C¡H†f ê4USƃ&ß Ùa«WR%`ˆøèÆÿüñùýxj®«$ Öh˜aë|=A;{Äü‹ü÷Òª¸/eCÞà ØÂw®çM®L—Ž¢]óÄÁ¬=ñk¼d+úµ^XY¤JÉ;ƒI i{ñ¼*•ö8ËÁ#dé¬ 3þÚùZr¨-¢7).›Ãm¦ÝR·[³Ü9Ȳ‚3E Ð÷B+‡áE·Nè€6OˆðÄö°N_^¥fm,Ô~_ ‘‚Úà¶øÓÛ´uHš04«ÕƒZÝlH$š›B}ÁyMèuŽ_Ò(Feè{©¿)|üD9\êÖÐJ+‰¹ó;·Yp8Ì1×5EëY/é Å ÅÚW#\†Þæxè­M?tÁÁƒµuݱóâ~RI»ÊÝdék%¡ujä¢'‚]T‰Øšð_ÌP8pr9u´ÊFîÁÌ;-éîJíܽt; ‡ßITÅ«â”P ÞÖ„^º-Fk9 ÚÇHGõ½™9Ez¶Qw8ó…Â$ëe2†¡Ë_VnÈÉ¡OöòãâÜøÝ#}ooBOsè.WÍ!?'¾;eÖÇ¢‹’»S^¸+âG¥×·^X±IS¦S’[3¶>Zü"~ò³Ri—ßasyö2J¤6lèϺ Wáz×@ÿÉã—àŒe f·Œ¾c^ƒWH+œA°š"ºÒï w{"8!ˆ9 4m’Â+Š´7ó´á"o>…[˜ŠÖ ²õÐ5L[ôH¶?ÙëÝdxI“™¹mMöf ^TŒ\ C¿º„ܧcèI” ²‘®¼ Âtݨh_Ü(â­4ޤGu?^k¬-C+\R˜aÂÚe=^·¯2⨱7âýühjáȶØ œ`±= uX˜ë–Õßà–TœÀ­f³Ã²Òjsmm( ½{Ï҄Ñg§$¨ž¿GЦ%åàÖý°L  ªÂû„à@46Õ7À먦Ôcð®@ìPÕ;‘A%%ꡳ¹Š)|3©°o2S·œÄÄØ½¾`€<õáˆ`^ÄpR.ž¹ ÏL_j3 Ú‰âZ[¬-ÕÇÁ;=¹GÑ#â#Pü’„¯úWü 'áÇßZŽï}òì+_üè3Ñâ:8™×Oa þAº+£{¬-ú‡®³Ô±ÎU «ç“àÚ¼=EËU;¨ÌÛtÂ?¨^%pµ]åÄÇFQ稄ßÍ»¥£%ˆš)¼wá’ÇXòš‹¡>d$ñ9ü°´îRã¾+­§z~;Ä©ÀºÍ»l; (k®Ô²·ðEšrµb×~ r _ Œú–J)àø´H$†ò¹$œ™Þ>cì^rÆÔg“ïi $'‹DÿÌãlµendstream endobj 207 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 1002 >> stream xœM’oL[UÀߥÐ>vΤS¶Ñ¾EÔmA(ˆˆfÉTL”E760Myý3ÚWÖ¾òÖ*´ü-ë)´-[ ¶¬3vb¬PÍ\ ÛD÷AçüfLL4óƒqËô‹qÞÛÝM}]£áæÜœ{ó;7ù“‹˜â"!¤jkï8\§Ïõd;";ŠH¥èTΔ.©d6ï{¾rO®ê€ÊP^|îÎWá…-6ã¡%B÷h8–hsôyœV³EävwsuMMÏTsõz}·×Î;­FƒÀµD o7ˆòÅÆu8ŒV^ôp»Z,¢Ø×\[+IRÁîªq8ÍÏí®æ$«há^ã]¼³Ÿïáö;‘{Å`繂jM!µ9ì}n‘wríŽÞ)ðÇÝÖ~ƒŒ¼U“Kî$Ñg±+€\VÕ\•&’ŸL„ŽD}ƒc>@ÛHûJè—ÊFÜ7~z|v*†ä%Ñéž¼÷hd 2| *NC8޲´žè5ø¬RV %jW*׺¨¤Ûcªì¦l™vSñV“¯¼4;_^.ïæ_>þ-endstream endobj 208 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 1007 >> stream xœ=mL[eÇïm‹{]q$Í4Bï5² 0¶Äl5 ÑMÌcB(pé }ãövm¡´`=@e…Òò²VhA@Á¡–°F53Ë|Y¢1..û´ÄøïK¶˜óáœÿ—“ßÿG"A’$*(*z7ÿXÎÞ{‘Œ¥b©BàËwwKR‰¤ü7Sì¦9$ ‚Xú·-_<ˆÍI¸é9"$ f'¸ ô+«R(9:½.ƒ>–›{"“>ž““K¿¥eXU\GÉ9%£•sñ ¡Kõu*†³Òé¯)9Îwô¨ÙlΖkÙzVñFF&mVqJº„12즞~G¯ãèb¹–¡Ÿ¢f?Ýz­ÁÄ1,]¤¯gX\cPÊkNN„ÈlIÄDé"HˆW%DĤ[%ø;Ö)Ù½Ó‰=!q&ð‰p&NÆ¢KݾþièC-š–†ö*»¬¹Ðšßrª] öøz»ácä4ŸÍƒàŠ÷ó¥Ç2O`h ÆÐ­³¡²—*øs«½0â–ÍaáÆÐ( þa¸À]bªþˆ­äˆaìW|Rú8zÑÞ#‹«Ëâ”YNkËL× ×Ö¹óÑ÷†>ŒÃ×ðþ8/å÷ ñÁG÷—£”_¾Ü°èöüZðJ+´Ù`wRº³ Û…Q€âæS¤¥µ×nÏMbáì-*ôÍ×3_Â_0Ì Üݬ»Œà(V²Ï4UwaO,_įFð½ÈT³26ç ð'¥®®N;¸Ðž¾ôûe½£Û¸£»gåí{º_â.>Â"œ…ý YTOS‘Ùåoá{ô[Ñ,þe>ò‡2¼Q%56³Ð èr5xF×aÄï…Žqj>,U{ô`Ô˜çäÛ[…ñUÊcë5C3(šøšóšÊJu! ì×7ÿ ô{†(ßVò~Â$>ð'Z˜°ÖkŒi¦zJÒ<{߃[ÏðUž}KûqB¢l¿Èª?q‹ÅÝb Aü)ÚÙendstream endobj 209 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 976 >> stream xœoLwÆï8ăvÅmi6\Oã20¬2uCœ.Ó.n’ul€Î9(PiÕöúçJ)`ÿÑ¿ü í…BÊŸâ" Î0s‰0…¸½˜fs.‹¾aYö½X~×Ý›Uó$O¾Ožï‹Ïƒ"ÙYŠ¢9 ¥òøgW ÿÊoÏâ 0 œH¿“®ÞT€ä}¿`wz—MÄgOýwø%hyjòaÝCQ½Ù­  6“¶IC“… Eä[ee¥ÅäÞ’’2òˆNmÒ6¨ô¤REkÔ: çÉ*ªA«¦mdá! MîÙcµZå*YN™šÞ+*&­ZZCVªÍjS³º‘Ç”?w¥3Xhµ‰TRj“A³¥¡êà»RœD>@ö#ùÈ ÈæÌ>$ ¢%h4«&«=ë> á..µwýz•CáöU¸o ãßàÅRÚßl&ܳƧûÖÊ“Ja›ªø” 4ÍSĘm0ðkÖ³âŸcËYªåŒùtg0ƒæ뷺دA Ÿ2°Ž¶zGõ§kê?áÑy˜c½hø K|mdÁ |hðÇ¿¯^sã„„\/Ëä=ÀxD¥¾–ÅG?Ý:·pgv}tY֛쎂!üvíÔW…õ‡›õ÷ô‡˜ÞnbâÔÌ,ÀãQM»i»F¦Úá®MxÑÏÆ;¿,Þ„‰‹5 ¶9°Æ®³ñÄÜÀ÷ Œ3~ÖáúŒ^Âþ…­²à´{˜Œ^a’2&¿J<|•¹[ ž´2Ž ·Ù“Á¬sq0¹1Í¡©ßaë_ÿ&ÿ²´Þg?¬¸~¢£?r+¼ºz> stream xœcd`ab`dddwöu041 H3þaú!ËÜ-ó«C†5»›U–ßÉAVë§J9_7s7˲mBß{¿·óo``edÌ/m™²Ð9¿ ²(3=£DA#YSÁÐÒÒ\GÁÈÀÀRÁ17µ(391OÁ7±$#57±ÈÉQÎOÎL-©TаÉ())°Ò×///×KÌ-ÖË/J·ÓÔQ(Ï,ÉPJ-N-*KMQpËÏ+QðKÌMU€8UB9çç”–¤)øæ§¤å—悌ÏÏKÉ,.ÈI¬„óKR+J˜"»»˜Yº¾üèà“a]ôæ#–»eX>Ù¢ßy¿_a-bKQVtÿÍž[Ö¦Yø;¤›ãÏ{¶¼œ4¹·¯{‚dwow_ËÔ?i?%~ì`û.Õý¬vJçÁæGU·"—TŸ[kű„í7ï÷µ¬2¿“ØJº»º:Z»:º+º~,ý›!ñgÖŸÏõÝmÝ]’ SºçÊñýÚü¯æÐ/ÙÍŒ?,ÿ¬ýýýwscWgw;DúçE¶‡ß{N*rðð3Ë«ns+þ­t‘-Û÷÷?'Mº¨_²»¿«§uÊï›?‚$~(°}gëþ^Õrñ·?ЦßëXÿ^d+èêÊ‘ç+ŸÏ&ó»r>û.®]Ür\,•ù<œ›'ñð1/¸PÕºendstream endobj 211 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 3394 >> stream xœ•V TSWÖ¾!š{Dѹÿhkï¯*øB;¶¾ZDª¢¢SEmQƒá D$‰d‡‘7HH b@|‹ŠZߣvªƒÚ©þXkûk§­ã¹éÑͦ«³þÇÌšu²îÍ͹ggßÞûÛ[D ð D"ÑÀàå«fõ~™.¼)Fyo‰›ÜÏÜ»¾E Y0ÿ-÷¸ ðƒ÷€æQC‡¡N_¤‚6 ¥Ä"Qâ¶ì`Er†2v«\%í'>{ö{“¥3fKƒeÊØè¨$éò(•\–¥" Ò0Et¬L•!8O®R%Ï™6-==}jTbÊT…rë~“¥é±*¹t•,E¦L“m‘.R$©¤+¢eÒ^7§ö^‚‰É©*™Rº\±E¦L¢(ŠOR'¨LQ¥.‰Úž-Ûº*6,%Pœ™­ËË1ð¹x*<˜wñªÓȉ¬(äâ·O‚çÛyc®Q[L5ÕXðšäF9EHÜœ=bT…g²ýͦO®íÌÝ Ž<žt¾ooY™±–¿Lc¿^ªÿhÞÅ¿’ý?ô u¬ípGë`î\{‹0,3¼à1ÿGB2…:¸p­/IÒéúÒ„Î÷r„zQKÒv‰‘ Ìê `ž239½a×.@» 3K“J7ƒ¥ ¯kM2éøÃ‰¶üëš“»îdÁbf˶Èio'Ý=£æòË f 0™»ÇCéLÈ)ßc4ÖìãÌæÒr³ùÄæ“z{o|ŽßúC{ÒgØËÊç„•äCBÑ'µp˜i²ý 0ÏVpÆÜ"]0UýqŸNÒy° ͲˆG7å£Âù¹° M6Ožªc±Æž3Ú®¾ø'î\ÂjzQBŒ,Vq$wØ!Ç©EôHXÀþÂÔzfDèü™ùÇ:8ô%'÷rØ º<ö§ƒa*úÝÓŸ›9ã ò+è¢çâÑ ÚGö÷lÿDŒr„çìASÉ=®Î%™«ÓÃB²ô0×D(«£ï™rd½)"ÓåÌãRâ$÷MðYp¿/()ô<]ÉÁ¾Òu eƒë‰Ý!:þ%ußrŠ…Y¨“}6¹#`VPü’ÄÚ »£ÆÒP·³:Õ̵Öwáéܹ-³ùz¥.D·0éÙŽ•ÇÌü˶›·N9låÌ2»ì80%_~ÍŸM§7ïÌL‹¶R.°“õŸ½-:bcKû.ß@fÂÓyID ]BjÖÑÒØÜärµ´Ø¯½… Úyð$ô¡¢kô…Pƒ:O£É 1辶‚xS0ŠöTÖ÷žˆÖ…A6dÜNì?qŒ†Ó+~ľ#¤JEÌæ@ýÈ8ºÕTDJ°.Áþ~SdÁ>ØWVa2×754Ùø°¨ZÈÕf¥˜ÞÚ=5ä¥Nhê?´rÌÛœ¤tÎãMá.QíC±ð{ô%[¾µÄÕëœøù°’ú½iÔWd‚ v´y)ÎT©­$žެs­œ‹Þ Û©C|ÿ‘ÓP¡¯–¡øçF'.\Œ2³ ø¢[)42Ž´ÊÔt•:a㱘®žö—%…„Є':ø¸Æ)jéAò§ba€0šUìÊW†I®IwØk÷5]^Ü4óS°{áá?MDÃÐè4¨¬L :mž!GÇiÓó&³|F;šˆÆÝîºU|2Zø~ãÑv4«Oë‹nˆ…éD(¬{ÀÖ³þ¯x4ê?Å,¢&›Ã;^Ú u%êl½v·žßê?S—AÂY›Ø–ØW¡‰1ÚÙb$ýª®'âÑFaÓ£Qȵ»(»QÔtåÞk¾ FA7Ù“²@Ýžâ¢â£ÊCyuÀ<¿uóîÁûSì|E™«ö”1_——™LF¹z_Eu±­ZݼqÇÇùQ‘œrß–êO€™¸`ÑûŸØåµé|ª27R`³%Áº#"kóˆd‚ C#/òè¼×4®šûȹü€)zÉN…שּׂf{¥…¯¶9Ê€ù&¥ÆämÛI )âµÀœÄóÙô[˜»úƒÈ‰Ûºj«.<à}ÜÖ4‡Ö…Úè¤kÑpté§ï›~;\-$¡oØ»–ïÎÃCæ9þöãð½8«ä»ä+RµŸöIéW}I—J /måÐÃ;Ñ0ÉmpÅå6ÆÃvX ¡°éẖ˜syÀ|iyдÊ3³ [kàóBå© X »Û2Ðpf¸Z{>ç&öhf0]Åâ!hìÆÒ:S±q?X€yŒFž€Ås·Žá}P{__@ÞH,Bƒn‰Q„ðŠmT5Ä%mSÅËÓì-µvûãDÝ®ï&êVýk8•P¡¥¿b–æ›ñEtìiåþM¢ Wг+b”%žZùI'\b®zå ²áwef® Û¨-ÿ»&£µto¯ÈÕë²µœßrFΨ·¹ö«lñÉŠÑ“XŽ$hä7O¾ÿrI÷x+ÏqáÜgîO??V:ge`L}¦Ýi±6¶ÆW¤¹£gï˜öcïη0)*+“OLPéWäo7ä@ÍdC5×›¯ÄÑ…'H,F/¶9ÙªˆMMVÄ6¤:š­ ßð¿~óq_GƒÛМ6Ñw„Ø­@ÙJ Øl)°“í)Ù™*•*y÷`,g+ëÈNzdg:Ù©#;d>æڟÃ,¢¯‰ÑClb÷–AY¡Á¨.çñ[ˆE+Mty¡7;Û*ÊF03æü¢œÌukwårø ,UàÀÞ€UOq5‰ùÕ&Ò£—‘‰…o3‘d9´þIÀDásHkÌ}™†GR> stream xœ]=à …wNÁ ø“šKºdhUµ½1!Co_ M‡ÏÒ³ýYÏd/£w“{ ú [çM‚5lIž`v1ŽÓùëZÕ‹Šˆ W_ï¸,€ÝýM-@üÌZ‹íÖ¨4$åg@=¥²·V"ðæo$v`²Ç¦Mœ³N¢^°ŠSÁ‹=)ÙDi§Ûµƒ«‡kÄ#Ö[Jàsû£å¬ñœ‡ß«1ÄJá"ô"PYÒendstream endobj 213 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 551 >> stream xœOÏoÒpývÝF§È¦3ÔußÃ`–™ŘùƒÄxÁƒKL4.‘°fTWŠmgG€ý("Ý>¡ Ø–˜Lgb² у¯Þ<ùø'|‹5Fäð^ÞK^òÞ£Po¢(ªß˜ytí¿òš(ób9Bƒe˜OÛZß¼skäR{,à; öÞ¦ÙŠŠÈš‘÷‹Ñ˜Äχìyð”Ïwu_öz}ø¶ÀI|(Á æ„ Ò1 xF ñœÃîaE‰^ŸœTUu"(È¢4?íÇ*¯„ñN椗ܾ+F|?(p¸;t¢Ë~Qˆ.*œ„â'E‚rGóòóNC8*ñ‡¢ŸÑ^„f© ºOQ/G?¨s¦î Ÿ-£ñ{¦N‘¡M¦ÛOœh* ,ÆÙ¿^[÷ö Ya滣†y¾[ýF›SdØYH°L¥\ªí®m&¶GïÊo×ÉÀ×á¼±[ÈÅœQ„³“ÞXkÌS&îluý ®$F’Y±•-åú¥åtx ®xr9¶T̼}5ú}aõ‹j <NéKZJ_Këk 1±òr%Ÿ…ÚKl?Ö•>µª•JÕB½˜y¿±ÃfŽôOúGÆAŽæ‹:uH24©™çVisŠÌ¡)†VÃÉ kýãϬ€.ÐrëE½–ÊB ÈÁv•qÈûí›~‹Ý±µN´N¶jv{§úʃâendstream endobj 214 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 6282 >> stream xœX \Göîq §UeìjºI4jŒ÷$FŠ1Šb¦fR#¨e”5’ú„z‡E­ FSnÔj%õ.µŠK¹S«)'êSj5žZCͧ&Pk©ÔDj!õ5™ZD9SS)jµ˜šN-¡fP®ÔRj65²¤Qƒ)+j=%¥†P(–zƒ²¦l([JL ¥†QæÔpêMŠ£xª/Õz‹êOÍ¢P.$X”Aý!R‰^ôy¿OØRœ$¾iænvÌ|¼ù9z Ý,™(ÑH~f1ß÷ èû¢ß¦~{ûÏîŸo1Ýbû€~T® œ9°ÐRnùpPö Çƒ‹¬úYmµz,]"Íb?$dHñÃCž³³Ù,vÿvoì³fdCÛTØünëc{r¨ëДa †•{4|Üðˆágßòf!gÏ)¸Bõ@¡iׂAðƒN¸ÅÆïQg  ¤Bñ^x@§ÆF()-6))E3õ¨œ{NWÔ¢æJäÍ×ün3èšæÞJX‹*øçty=j.NÞ“Ã'e…¢(ÄàyB(›A_Â:slCwm çµnZ«Ç:pÓYK¿aÛ‚J7sN´JNªG%ü!|~ƒqSé²î9Rñ`-=ôôúžã'KdsX¹EÒ5¯A?o¼é¼(úÙê³8é7‹‡ox8Óµ· ÖŠöb8ß²` ójºTÿ™`Å £ä(0 Õðóñ¨Ù0ʼªíh–^“Wõ¨”¯¡Á‡b[\nJ«BÈÞuäéï†eB‚«žàUæŠÐW_éwPt{ L Z«Ÿt¥[Kœ À4á ƒéª]ý6ù;åGÒ°RÆ^¬>xg®-½Å£Õú3Ô¡2¾¡SMK_l3qQ: æç:nêÖ^r¨äq‹¥ýC³“Çš-˦ñdã$-|£ÖAN,xCÙqÂl†Í&ŒÁƒ±ô—qÐúüò¬8쉧³Ž®7þtéÒ•«]&MpëÈ÷zÂÉËÈ‘5ti]—ÕªçU£Zþm«>îöqÛÚj£õ¯kÉá54ŒÁgÍñb¹¤÷ô7®$‡Åz7Åj…|­h¯¾>'Â…›lª ŒHŽKLæÒRPZ´Ü¯q}Õj‚0–˜0»àÅ@þÂÀüØqgŠW  4>ñ]‡Mˆ™4ý ïÞý£ãÜ©õ%|Vd¦ª1»‘¦Jo–ka¢¾îFf. Ê-p€rö왓®ŸqšöÞ—úÜ;ÊáQfû<Ä· D)XNùõ­µ¾¼ôqSZ‰×+8îlüÿpL÷zó¶*tâs°Ý3KùêNu(Ý;‚M4Xß;˜Ó¦ÃXlËcIçVP?’€y;A"ƒû~‚-±Ø!Õ†.3¬u‡tëˆ)>Ö²ÆõžÞ™WÐÂwH*Qµ¯ Sr q•DÚ ¡`-Ù„T»w¡êJîa÷kÕ&[ÃH[E‡£Ð]z—q×$;’¶ûr}yÑlšÝ0^ ïZK–²?Ÿ}¸Œrebª«ÊªböÛ¯^±Ä—~³Ë°îcDÈ-|ÞÁ8‚caq÷.~úoÑs '6Ša4²0œ~íÙ‡ƒd½æìƒÙÙ©ö3ñ÷#ÝéžÄ±ºª…m€ÖZú ŽC>«ÈVfÉs‚²?*BŸ3õ-;~ù¾$08‹ËˆÌŒÜÑ ,DÙub·A³…´~ ˜é=ñEä5/}Ø–ð]ôª¡[·úMZ·¢ð¦œK®H) FŒ>"üX‰ É› Ê2Z‹¸6ï“ipSZ¾?ÊKŸ¡ß1«ñÏaº’¦já )j:¨'I¹¼ØZ}ÑêrV£÷Ó1ØlͺUˆq˜Œ†ñßývs÷¦ÊùÌÈܘúî#ßuæ°äXß(Χys©'9³ÅH< ;ã%¿c[èzsMŠ”OÜóµö·*‘>èSH0äÍ‚½Ñ‡Øþ6Ý‹´™Ó [_ïáTú½¯½¯]¿u £i|üÒÃ"¹t×åq ÄÌF'êx“j)º«b¡3);u&k›ØX‡=°x¼ú¶·B×uªåÆÑ«' |p˜œ~}92 ÀMìÖ ‡?[Ø  ª@Ì5º¬®ËájÞnÓ]Øz…Ê| aMvˆñŸ…v¶pP#ù÷ô{ºkœ4|Ú·½µ‘y¥ÐÂ1=7P&9°Ç˜t@uª1e2Þc2ÒŸ<åÜЊj…mbÁ‰Ä¥ÙÄ9}7ÒX„o`Ü0o|eF¯Ãšià‡E¾UHƒl+Q *Ü©?gŠu ÖŠÊ„X±°^¿ƒ†x#E"EG„tñi%ßFŠ›§Ó)õчcöE\H-ôAq$bCS7c.EɤÓîpÜ¼Í /(¼÷ãÔÌŸÃÙŠ¬¨½Èv/Ê*¨ÐÖÝË(ɪf2è÷VÜ7;¡(²¡V”U–]Íô„k²ö©Öê”nc¸’… }ZØê€£é{ÓØÍJžq9Y <H®¬¼ƒGn|9jTµ¥­ÊÏȨÊ,É(FˆùÛ‰€ù¼ô^Š-?}oè .Ï9iX+:ŸwdK*Üû¢”¶DxGz ĸnÚõyÛ­ðN6ßS‰ÊòÃ^³@øÏ|/]ÒЧ¾=ÔÙÀ;Žšó̼¾‡4qWy5ˆ›½4ôÃcð <Ý|³ñ)ô5D+ ð?‚e¶4—õèò®wÍÝ ÖhÁ¼—㢈ú²aÁåß1Ù™K+QIñª:NÚæ¹0Þ0Âã¥Qk‘G„oƒ?5»&ù7–4"X¤ƒ$â…Äí&6Ï0"«µS½Õ8,zȉ ” e!§$‰ž£ÃµÂ.­èÀÝÿ‚äyØ7zkRDb€m:\¯I¬œ‰6u¯‡õÞÆ/³ãPJ·Eiò™Ñò 7ï´¨t”‘“Š™ºˆ ù¶Èˆ`Ï]›Ûž›Âl®WI]ÓAH‡XØ "¶1%©âÒ”‰\Rtø:GÄ̘wù§£ ašÏQ<ŸáS©®EL­¦¼éÚäç~Œ‡OÅ?¿VÀïúµ¼&#tЬ]n‡ºv1ÌÙïø ¤ ’qG¦BQZn\­©³Â`¯µðj‰X·1·{ÀàÏ Á’Âõ±[vt¾¢Ø+'17¡(±£’üü’¬Ü ·è$“ÑL¿öÃjI‹¦ƒ+©¦¨ãƒ™^ ·é`_Õ¼d5UY¨œÙŠÂ9üöKHðã×Kz™¶2%øn S¦§¢H[ßjTÉ –tu.ªB¥Ì£üðÂÕ’ž´ýº  ŸõôÏc¬7RïåLˬûš2;…–¶5ùm*Û8 ‹Þ‡¥xÐý±œ¶£ý M ø›Í—¬ð°Ÿ=ÿã+._º|±m…›§Û!Œø|¹p•Ù•µ 1™‹ž:zG˜pƒ Ë aIgj±OYz=²m.ÊÚà ¹u’Þo™,½[M ?C´¢S:¨%Žœ+XM%jzñ¯ø Üÿ½ÑDË |6 X°9r¿N­ˆ„ÔðD>tõÊxš‹æSýÈdœcK®=u‘A^0²7<0˜$ÝÝã»bÈÕóð@ܘ½~·j´‘”Ÿáó£L¥ÖH(£ºµ×-¤jE¤„¤‘Äv€4}ç̵û:ÞÇ+LDa§ÚDxk\,ÔfÐ,\µjãÂ÷ˆ£ô)qOŸ:ëÏY5 דíçg>¹j-ý{,¨a>»ÒóÈÉÓžøbÿºUË7lpç'û°Yi{mˆypíjGs|Sd#¯)ªÍ.ÎÈÝV“¸1å5EM5á;VÆ®I÷]Ï« }*×#füü¹36–û•†óRLŪâC‡"ŸyAŒ£úÅ2Ó_,…!0äÅùZ#Ž­nà¢ó<«'#f€¢QBfJB1%±%éU(eg–dog ?¾ÀŽž{à`EAKK_WV€®!¦ÌÑô s7ŽîNï.h‹º4„X˜‡Y}‡ºHßr°ô55Xýšàÿ¡6tQŸv‹e+" œužº*’V*˜DêñUúÁÁek×x|2‰ŸGƒ›?ûek¥íc~œs÷ãð‚ÿM;ùÑÒc7‰¾kF-²ê0Š<‘ ¬ô/lTíBÌ…Ú/÷´nŒhºÀt>!mk¤ >yåÇ×p1M º°º·žZ×Ó…ä|á¼0”5ôo›€ÑŸÇLðtÔˆ§·ié3!ͬ1´Z& •ɪC««¹žNÒêÄ%(»äuI¿x4°Iß+Ñ4Æ}ã²Y²5Ù ›9φ”ýééåéþÝÛtiðêÜ’–no Vu‘hpæê½Ž‹ÞûcŠùM{·å}XÁHŸ9TxyíFGl·½ýJÇärY¡…IU=šÚÆÐ$G&p›v¯,ò šZ~{­w õ¯^¶:°é‘›6¯é—†껓±¾ŸÛNµ»Iïø/“¡"õ'©O»^üä³` óÝ&"u‚‰ ñ <Î^6t1–ܶNõunÒŸ­¹EwdßDÃ-T—kú—è¾I õê†bõ´EÌkîá²?kR¬ðÏ´³»û*—çàçf Yæu³žÐØòîħOï>Kî_µúW¯Æª+uß "5ß6©Éo˜¶Í'Û¡¿þ[û ÞaÓš•(1þ†Ý5è3Ÿ—x‡u'û=ƒaÈ›‡$Ò¶Ž}GöÕ”DºpX- ï©7^™ô_§$Ùô”ÀMÑ[d\Ô.Y…b¤·üQ :$TP{íËÄV5±õÔË·ºmu6‰Í¸¿Fàq&cgƒœ#…ÚË¢va’X˜ µlÑn´{·Šáñ:ÆùøìFE]U‹` ÃÍólk`½¿—< `kspËžº;ôζ׼TT‰ÎãÄ‚ÇKK6¯,¡|¦"ª\šÄáïÿZ˜¨JG(Á6¢4²ª:¯´8··ˆŽéàs‚ 8ÃÂÒwÀ÷Áf£»n–>ÇÖÐÌžÂP îñÁùì‡.ÐÆÀ(Ý“ë×.¿‹Gã1.¸ß‡Ý×`µÏIÇ›Lüp±ûn*<)&&™óšáž¦JKC2F^‘\Z¹/wÏA®Óa’ÄX@³˜D÷ÊËA/‡±Eh·ñI ÷×I òÙM|UÔ 4ä7 I'þê Ÿ”ŒÔŒþFƒ?×w‹ªp‡ãé³ÇÝ8°÷Bûá‹è;,FÝÀq¿Y3glmŒ-×Ô”7%Ççq¥‡ì8…˜Û߬Ÿº`Í'În¼áb!óœþ’ò`ÇǰM·™”¢Ç‚î²íŸ5žF_0×¾Âl±`íB¿:u•~±’¤í©YÜŽWôÓ×Ǽ=£|¶…òþÁò4ï´å)Q1Òßô¨àÓOŽ/údéŠ%“6Ï.<ÌçääMÈÔ†WÉ¡QþSBê„õrÒÇèʺ³ÎÍzì 1`árÉê.üf-ýApþÆ–E•©b“’“¹@Y°JMZÊ„‚è¢èb¥ÆmC2•"RŠ"U©ª¬07'7kn¬/+AÛ ÿÇGW…íB ¨±¬¦¼ª´©Uts¥úV‰š.ƒ¼]¼RÙøhÒ(¿‚Ää`ÊÕ„ cé ÔMãèCxAïLÌ÷ådtJNV2øƒN†9<7ÇÏ%]Nn¿œ¦vêX¥ƒtâä¿C³`Ëú  ‰›BX‰F-‰«i¿ÄxWJ"¢¥";g{>×в¯â ÉÑ?orrù`þÇïË矒æç6)Ê¢ã·ØßYf0øÉ}sÒ¿?›ý+!xC”[ÏÃLRx¾u‡{×Uô ø‰-·~¾}íBe±9Éä+‰¸0¤BY_WUÞÜæýÅX’1CñìÁI_à80sz¢_ï€ÐS;pN*¸#÷HM Øá}¬ ÚÒ¦Ô¦å¦ç¦å2¥g#ò•Q€Ð1‹éyi儮Ο;ö·['9Nw^>¯ëX¡U+:¨;MRu©×Ø/ñÅÛîè‘ßîòòÂ&.Za<8iÏVUdf¢Fï!’ütRY„¾÷E¡á¼g7UUvËÈÔÂzi›jtú‹œÇP ¬“ë¡öíG¿½yz­ûr× Nü·ìw ‡ÚÐiæöŒ›¸?¶x–Ãòß~ÌI›¶xÙÜ¡cN…Á`õàÞã.g5Ü,ÜÂÚ;·ë´g.üôàô’¥ WØw½[pDôO*úõËí_Äÿ¤nþ“.²YG|ó£Q8 ŠÍÚPŒ7ØdDe…5¢:”S‘Q®úÆÿ<Ã`ÌÌóÐJ"§sï€è›úì õlôá˜÷P<òÎ+ËÎÉÉ%íYmDU°\è_1ŸÈZ’Zƒa $a±n²'±¤ïMºÇ'ЉòCû<2›uXõÕõ»WÚ¯hT(Ëøz¿¢ÕDø¦¥§¡d”œ­BñŒÛÊåöœ¡hƒ³±p fÿ^¹ñY³×Wó³ÚëwÀíÎyH«3wü }r}ñEké_8 Mì—Ñ<Ðj´ÚcÝâXf)¨X%ñÈÝTu<7{{>ªaCÊ•ñÞ)ëíÏÌ&–Ûª—l;šQ]·G+ÿˆ.£6ÙüN#ýÇIØ"Ù™rL]¶:.!9E3D'5hj‹Z¿q»ŠûmçZ»„Z’«8/ î½xrD!¹˜‚YÌe³åÛW}†Ž££G÷žßÞ¥Öx aˆ.Û­ô¶/;«¿~­ýÒ‡Ë:ÍØè•^áQ I©êD¤b¢J#Ë«}f_¶xyË|eå>mz§´k„ Uª\&¨ØÂ¼â|TÉÔ„©âSÓãÕ¾7Ç'E' p[»=~Ä^6%ñYé¤ð–UÕT’§y<Þ Îæzº_¾}Y´Gp ®ð”-ÚI˜w+i£‰ŽQo%Ì»“0/¼ÀãØ’VòÆOÏÉÒ1~äM+¯žg9XA®þ<¨8¡0.ÿ^6ê‚ÔŒxĨÂcáäi"{±3&?±‰]‡TTÆ”de”蛩yB¨ÎQ•»’E¥‰ÙExÐõ`‹À =Ý< ý©¬0/•2ùÉUÁAx#¶Á¸ÿáÉ»?åN:ÝXQ‹˜úœª|Þ¸Øg„…_ÀYAÿõ÷}Ž­º6·=1ò$eBTabN*öXTЧ“¾dôVÌãa“¢HøQ“«¬k‚$% ·:ǵÐL~‰¶ÿÔ ÚÅÂH!‚MÍìRA¸LÇæ¸ÿe‡ëmöƒã%*<ǃ°5^„?‚Øìîß9ÝÔ(Cq‘©(.Ž÷ñ÷Šõ%ZŸ ¼ ï>{þ[㡸¥5üb|…]çyîæÃ§ö:ø‰ƒƒÇòz{îk„yøá°¼p¥ÊêkÁ+¿´–þA.«õÝZuÙ³A•‰òSU\QZS–ÁËË?Ú\ÔY›¢¼â<‚“lÔ´™ÃƒæeÖg|‰šl ÐN??ä”áœÌaꂺ %"[etDTxaR1‰–h¦–Å$F'!¥m ÚËÁ «iÁéÎ(È6ùí܉v4¥™VÏõŸœ„ïã—X#79¸é¯Å’€Ü¸]Ñ}Æwã^Ž  éåx°FŸ,£ôI.Àî,6Êu*&%.5"™P¬>)Zž}´[>ß°YèïÕ$ß™›—‘™Ç Œ-æÀœBGäI´ýtý¹~f2‹¾Z ÅŠú?æ¼[Îendstream endobj 215 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 632 >> stream xœ%oHqÆïvÓ]îœVHE¹»EZþÂ&!é H›à|QEçv¹ÅΛçm·‰æœ·¯8§a&nù›‘–¦i#ÂzôÊ7E/ ‚è]¿[÷¦…/žçyóð|pL«Áp×™-–«5ÿÇRå8®œÐ(ùS?”¦Œ|,§ærþ™ÔÉ.PPÚyå×!Ôv¹r#Óâ¸àé ™·_t¶8$¦ÀVÈ”›LE̹²2SÍs¢Óƶ2Vrp<+¥‹‹ilNNò3—’ä®,-•e¹„åÛK±¥ª°ˆ‘’ƒ±ríœèåìÌ¡UbêYžcö–ì›Yà݉‹`çÄVÖåv°†id=†ÝÀÑfôØ\T† ©àªbXÅ‘éPuZ娘P>¢‘¼ÈôýåÑÙèüÄ+ˆ’ ï#Þãèlé1†CÛ’[ÅØÚ¡w¤ úÈ÷5Ëyx?³6¹¾öÇ8>3:1òCí«J4ªGÚ:¦`2XÔ˜ØKÎ-Éôöú蛪.`©RHãùöletkÃø°c¹e‰7Ðî˽ºhÝÖlçç»Æé§þéN¡Ûkâ­œÙÖí²*ün&>KІÔÖ>Oå ¯HO¤T‘·÷ûqzmœŒu@@ ÉýFOesCÈPx2BÐi5÷Í) Õ;j–zV=¦R?Ða”ù=¹¸N/Õ'ý¯Ü]ÜŒÇ| ôÀ½~ºÏç ÈЃHž"Uý­¼ºëîVia>½˜C™Ï·é•íÙMø KÖÅa¸i¬–4¼ÍB…zc–Ö/Pv’…²Tö0eÀ°¼‰endstream endobj 216 0 obj << /Filter /FlateDecode /Length 162 >> stream xœ]O1ƒ0 Üó ÿ JLˆ…. ­ª¶Žƒ2àD! ý}I€ÎÒùîä³ì‡ëÀ.|D/J`›H‹_#Œ49µã0¬Lœu²¿éðþ‚Í@vçw=“|ª¦)«z¡7´5O$ÚªêZk;Alþ¤#0ÚÃyÙœªº¨â?•Í%Λ€kŒÄ©4-MrÇô{&øS°A|-žS)endstream endobj 217 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 306 >> stream xœcd`ab`dddsö Ž4±ôH3þaú!ËÜÝýýúì² üN²Z?Uüùºy˜»yX~_#ô=[ð{ÿ÷TFƼâ¦vçü‚Ê¢ÌôŒdMCKKs#KÇÜÔ¢ÌäÄ<ßÄ’ŒÔÜÄ 'G!8?93µ¤RAÃ&£¤¤ÀJ_¿¼¼\/1·X/¿(ÝNSG¡<³$C!(µ8µ¨,5EÁ-?¯DÁ/17UìL=0霟[PZ’Z¤à›Ÿ’Z”WP”™›ÊÀÀÀhÀÀØÅÀÄÈÈbó£ƒïã¦?Šç3žø>‹ùû¹Y¢“fõNëžÈ±*iVr}Fm»Üo…?­ÍÝí’U››{¦ôO™Ü/ÇW¼ø§ý¶ßŠÓÙ÷síçÞ¿€‡ˆyŒÃkendstream endobj 218 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 590 >> stream xœC¼ýCMMI6/øøø‹ù7øMB@*ù${‹ ‹ ©ø·÷h÷c÷_ns‚Copyright (c) 1997, 2009 American Mathematical Society (), with Reserved Font Name CMMI6.CMMI6Computer ModernsuTV‹EøˆøÜ«²é÷!Ü««÷>É÷RÈø]ø n„€r‹|{—z¢Ÿ¨™¶Î? Jû-f&`:ç{¥‡Ö~Å€‹Z‹vzggwdv`‰n‹h‹Mn¬®‘œ¦‹£ }œpqkv\TÂa÷ ÷O³÷¹¬}¢tŸk§f‘[“Z”[“‹²Œ‹Ü÷¤½‡o¢ùu«ø«÷9Ùøy÷žlû„o‡{ce€kq`‹A‹Ì·—¸·÷–¤‘‹ÃXªZ%Zûv†‚™™Ž‘”Ž§ê¿¤¨‹Ÿ’~tv~i€ph3]‹g,ÒhÙÏ´´¥¥œYÀz°‹®‹¥žž©¥µ™Å‹Š—y|‰„{‡{LvZd‹iˆ¯™™Ž•“­•±‹”°²÷-Ž—Ÿ‹—‚ pzz€}„‡„„l†xv øC•––÷p û`•® à  7Ÿ ® sòKendstream endobj 219 0 obj << /Filter /FlateDecode /Length 3727 >> stream xœí\[odÅ~7ˆß`­"qfÜôý‚)‰v%"ˆø)Éìoë`Ïx}]å·§ª»Owõ¸fÆÃeÅES§ººê«KOWûb”ûÿ)ÿ>¼ØûÃ×ZïŸ^ï‰ýÓ½W{2ýßýò¯Ã‹ý? ‡ƒ•1Š(÷Nöò£r?È}oýµÝ?¸ØÌìà_ÀkEÇ+ôè”þƒ£½¿¼œ‰QÇ ­.gs1Ú½WÃ9,Ëàcà sa8DŽÂ½¸uTFSžÕh#TÐj¸žÍµÖ£)«¤ÈÃûd£eÚ_÷û h)„a8Î"Œ.ð)A%—yÕšáu#Éc·d‹+äF)V7¨v…Ïy!LHÌNäéÚ²ppQ­qݘÏo Ñ¢-ªdªþj6WnŒ=BkI§lÄ£VyçéT6Än›üåŽóç“GçÚŽÁ¹ì׃´›*úá%š8˜¨’5‰ í¾æ1‚Œ9­Ö±ˆL”NçG!%®=–Ì)µŒ— ÒR§_çU#,âµT‰ÕTGù9'ÍðbiâðÃèƒéÐÖ @w“™#è~N€¼Hê;¡„Lz‚ê4Z6¦­©b/»çI5\¶g‡3ð½ Ê ßͬMƧ•Ù AÓáÄÚ.4ÙI›sDˆ‘Z‹ë„X«¢ëìVN¢ÆZÚ(A%7¬í-b0~í€Õ¥ËÆ›¶ Nƒbƒ¥‘G“ÈYƒù¼áu™OÊÌ¥-€7ët’° bºXŒAá⣨|qOº'!±ÃòËÆpœŒp *œÕʆÆe UÀµNÑÜŸ]ˆÐ_¢CÁÔ6™TAƘñe„°Ñ ]NˆÆ[ÆüWÀŸüxˆòbMúÙcœÍ½2)Ö>GŽ‚ e–ìÚ‹ÉJ–nM2ÞuÍx7ù©OõñÓ*qôeæF÷“=è©W˜5¼ÔQx‚ZAè©–9º…íõØ…¶„l£À+Ùv4;scMÈé 0èj:S[ó<ç ?ä¼›™Ai‚1}üLFéðDO¶”!äH(ÊçH!Ÿáû–5­we3Þ½ð_·!2ÞOh²sFºÄ„vÌ nJe®äå—-[gßxÏ-óqí¶ü ÿSA?á%$)VÔu“E¡|”ô$6´d¶&çp=l¢Só¾šÎ‘˽ëʽr v¨HÆ 1­…˜TÈãF¾näE#ÏX† „žÍ¸n0‘] µò V€F£86|qÞÀLӉУê!ı!è¾Àø×'¹Ò€ ëœY“ǵ±o26äRêšëöàF÷r~˜kFh´ûòt 覦 Ç¼L$äÕ8«½" Äe ÄE²#ô«6ö‰*(xPéá#pQ”ø d@ ÷è,­êrŠÒ2ù‡­YWÈ "¤ÛÒO!6ˆkEÿC(©L-×&ܯŒ,ÖÎz„/ ž®´Ý¶ëQîÛ :P)ä…ݤ*6`ȉióijýx“™ô)‹î¦QT’Ù-%Š–yÙÆÔÚ͉~õ Ò·&Ng-2ú†6© ­â*iâ"4ªÅgÂëÔz§‚ªÃ(™ÀTgîU'-Ryƒj1_Ê›ä¡G½œ0…Dh“oY9¯OÅ^Çû©uëËUº{E©h÷·*«a­'ì`†:šNÜÔXK7ü€”ÍCRòÎZHÑ*V°'ݼ^EcwÝX¯Ê}¸k*hH#¹xÑ€D-—Ødtô)’M+"T&i¥$Ìt—5•£/ϱ'5N懾n ò³ÙÜ* ÅÔ—dºžo96’ç%äŸ+U#C#E#ý¼š>ܳ¿Ø;xºIw’ÑIÂ~ÑÈ«F.ØÕ7¼ÂOyò˜äÉsn>Ü«û »ºA÷¿²g&ºŸ²…r§îMZ:Zj‰>‡ Ïß5rÑÈSÚFaaø´‘Oãv_D.EöûÛöe;ÛŠ…;1êy³Î'muÞV‰©¿hä·$»}Í:‹ ‘`Ž¢à?Í”äôª­þžew̳Ù&$2bnX’ÏjÕë]_¶³=LHцþM³*ñ±êGm•^5òˆÕçœ ’/$ÓaÙV¯ØÇ.™í CÊ7aB=5-ÎŽÁN•jÓCùÎYû+7SÁxê>)°ºÿÓF>jÚ#¹î571!ÏÀè†Í‹KÊÀ¹œ ÷‹F~ÛȯØoY ܰš-w‚ž ü‡.w¬ç‘;Wl?:¥’vuË+7òi3~Gr]O(âö<þËY<,ÕUòÉ.¹|åÚPãñõÁ6lrËÖ׋u¿Ü³›ä½¼ŠTÄÃqûHþ¼o$·Àû]3j$1#éÉ5#_\€ Ü]èõYcâË=1ï­ÎN5¿aƒglð{Æ–Ì+Ö¼Wí±ù=›ú‰Ño©Wêc_±»Ñ¨ÿ¬‘ÿnäìjqæ¼3s-ÂEjò×ç\6’O”ê‡,Ã%ë Pf.…Á›?½,L{#¹,’+¤i¤k„ .är…n¤els€bˆ\ÅÊ4pLë%óö1„sv7×ÔyÚV;’Ix¯›ßT#e’5ünrfVìjÜeqrD¨eˆ½8‹V.Yõ›ìõëX|ÍÌh[ï4î+v™W³$tx Õœƒ£n¡¶Å–%u³éC¼ÝÅÁd²y3YGN ¶„¿iÿ÷¬‘§ë¹i½ œPã¯uò÷kâH¡ù¸™NL=hÿšNàË ÛáÃüÓ ¾ZÚÑiç뛼§MzG²Òý(1¨·H—rÔVÇI:ïÁ’’ÝS›QŰë@^Êþ@$0îœ_eª­"ó)ä½OÁ«®þäÞ¾uKXcñÒËø˜ÚÚGÑvü%µÝ`P·³õ¯¢æx‚G€dC5zá‹2•™kˆž(Ü~zËK³ñó;Fu·c¬­ÌÿeÏ8¥©_£gœþ,û{F¾ÄF–$%–4>)±]%%V4Í,+̰ÂV=Ï7>|gDvãûš®ùú)}ͽÛGþo>õÃΆoIøv[±í#/7²r#}¬ û„•Ë+ù¨m|ûÛß_FÖþÄÒ‘õŠ¡®¸?ø éY’à*°âíØUõví¿«C}ט¾kL©Æ´ª¾¹1uªìÿv_ú˜mé£õ¥šéK§vé]_ú ÷¥oñ]æä‹y[&o- @[I;k)ù“º¤®ê‘z\]8§«\bYGw{ëUz{‹jvß²Ætu‡ñ‰eøû 6à^Áí´‘àSßÙŒò×/Ç’þÍ(Á Ù@VK;Öü¥Œ˜,¾ý@}ºëe)yãÿ®+}ו¾ëJßzWjÂÃÚÒz&™µY›B¥}jþ›ûö1Tlú‰5ýutþy·³y8… ï§Ù¸ 52-Ó^á@G?(”dƒœÿ|—eÔ‘NAä9? ëæüè·q;#|??R°½áG÷tF†Î!Oò6 ,åÑ Ʊ›"ž­«mN¢nSë6ìî Y™¼ÍnU2$µš–ãÚqÓÏõ”ÄAã ¦“eú©áeùmýX žÙhמöóEu¸¡ÉH-f pÐä y'¹’€!ªnŽ$û![4OQÁ½£ªw†þÊáû õÃÉ5›‘GÇ•V8iMÒ$ nôpº\s@ÂÕ¸4%HCg‘l ÿ¡)ðûEÔ$ˆi¸'‰ AôÞ Ç¢Ñý`öMÂãæ9O›¡¾\öRmŽcqn¶‹32ÈI·û %iU`&r˰Úî¨@…p'8ºñ=«ºT ž Á·é tP„Äi¸Ù 7*È«•[Aã ïÈÏgAãì•/ƒ:§?¸u•²ØÐÊúÂBæs éŒCÊ t}®)¦v]ø ' ;ÁHfVwI·JÑ&™7DDoJϾH³½dœ¹Ž®Mþ‹2'Ÿ£NØT½û÷!Ó`Ž •‘e+”£3H%‰Z±yÞg®=œA®ÍLjoÓ„™Ð]^Ì y¼Ù‚i<’¡."®Ÿe®ãe‹dI8q—cûøMû¥¤^MF÷9}=¹nàtêÒ¿—&ãJX”CïnH~®ß, Ü™ñ¸hc²X‰pŠ÷–¡)oÓ\2ÍDDž:V®¯×ý8ÛÉ]QÊCçä½;ëSèÂ)@:5{O‡€Î6Œá*/ûôq¶þg#QØ,Ya š6 !’q¿2užµ§?r¼™•·dÊ}bCúÈB×oqQhÝÆ¶@éã@tJ«³ôÅgÀJ@žöqÙŠ¿|PƒdqÙL7FG’vŒWaÔO' «~D5„üÑϽ¦Eaú‘rX…'´#r2MYûÊi¬ÞŸ±ÃQÞÍFuqýG &KA¯£Iõ)ŸFª{œt+ƒÅá§ Î:¼æQðW#¤¿Jˆƒ ~ËoÑÂÄ»9ÁØP›á_’kDmI…¸e-£¢«oËn[ò¦yâ(5‹¬²©íé’pýð?„}Ú&W¹ÈH›/º_ÿ5cô¶\Ìòt©,Ÿ¬IÂø"}OnŸ/¸m1•‹W€‚JÑö&ŸÃÃÎ;¿¸°ô4ëú1”ô¢{­ô–ï?wÓÞ`<ÆV£ì ®\oDQÜÕ7\þºoZµ$}–BDÅ÷¡ZÖÉc´(’–´ŸÙθ°[Ð^šŸeç$ÿ}Yλ0;M£úZzbÖÙ+¿D*ëié>Î>vЗãñžìý þù/]ûendstream endobj 220 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 6365 >> stream xœ•Y XçÖž3ŽK]; .3¨­û¾kÝ÷}ßaKBBB ë—BKØ•º´ÖÛÖ¶jlµŠÚj[«]´µ½Ýýí7>Ÿ}þÿK@ïmoÿKäy„L¾9ç¼ïyÏ{ѱÁãñ:/Z³iõæÍ'ø~˜È àq;pƒø™HÂýñx¥`ÑcáüA£¿´¤;èÆÝ:Ö 2²7·¼<ÑèIðy¼‰n‘0Y"ŠK 52lâÌ™ÓÇ„Mš0afØ‚ÄQ|TdRؚȴ¸˜ÄÈ4üCBØ&aT|Lš$lÄ츴´äYãÇgffŽ‹LL'ÅÎ9&,3>-.lcLjŒ(#&:l©0)-lmdbLØÓPÇ=ýÏ"abrzZŒ(l0:F”DÄŠIÛ %/NY"Zšº,myúŠŒÌHñ¾Õ’¨5Òèµ1û×ÇÆmŒß´ùà–„ÄmácÇEŒŸ0Q6irö”©ÓÏ2sÖK³çÌ1rÔ®Ñcb±Ž˜I %Ö/ˆ—‰Ä0b1œØLŒ ¶#‰­Ä(b±XHì c‰Äbb±‹XBŒ'–ˈ‰Ärb1™XIL!¦«‰iÄb-1ƒ˜Kt'zó‰žD/¢7Nô!^$"šèK„ˆP‚Oô'bÁ,Ñ™èBt%f݈ˆ•2¢#!á xÞ£³:æOâk:ò:– ‘I¾ÛiZ§sÔFªˆúµ³»K¿.®®»ŠºÂnë»Áö½ðu÷]Ý_ë1ª‡¡çªžw{ô‚½ú êsäÅÁ/ÆÓ/÷å÷]Ö7¡¯©ïѾúMê§í÷vÈÔÓ¡L¨±?Ùÿ@ÿæƒd €Ç„ƒÊt‰éÇD1ÿË.d+ØÏ¹Šî\ð>ôr/”ó¸‰Üç´¾FçŒ”Š”‚T}íR’JÊRÀa(b‘®Zp¨>Ùãp*™?TAÊ’ÁÁ¤ZàÂo×€Ãu© Eý¹N´ëˆûDA-e|¼I}¸óò±kiØ‹¼c_—¹c÷¢aìËä.ô*}fà òu†lõæ•|ùª½ð–—w¿6^ásfh§aǹ×Ç¢^¨ê‰XD}¶üØv‡]áP)P8½<Â{ã÷»wnøúæÅcçè;Déå ¼¼/0~Ÿp·h£Å˜LTqŽ+SªÉUæ1µRªÍ‹?´·d7 Ð84­EÛs ‘øè!_Â^§Ú¤ƒ45›Œø£¥ €š¶ýÞ¯p6¾íØ¡åýô¡&VaL?ì)´ÕX|G$òÂñm´ÙƒË´ð ºBÔ˜tPðØ v†ƒaÇù7&l O§°ÕÀý<9&?©ÃäxÑóä€Ca ÄeÊYE^-¨&×/ù®š·ßkyPo[„ˆ‰“¦‘%¡•'Uâ*9¨=${Ã5pãï;!z±¨×Zú{öÉG“„hÈd4õaÇÌ>òònàz:¸~´Ù ŒÀBU‰­²ƒzyD£pÇŸX„+ù"êŽ:cø·æýxýÌñâR6Å ¶¸ _ Px/‰æàÄÖì=rÎö™÷³~ÈbOæ]Ê{)=x ßQxMÉh=úâDô¹ª­úlÅRGIŠÒR³íp9ózôyàÁ 2ïè§eÖ&–³+„Ö9Îövpˆª¬;ɻƙ‘1hTÑ'A#09mG,Õ§™r䕦¥©ey:Æßîp²žòòî`†v‡•´ šÂž]¬cß&¥ËwŒŸ ¨e»nB.†$ýÃÛÍ{ÙtÙ”ï+c4×Áid¿B¨U¹Æ`§ê´ ûÝ8ÅÁh Z"Ðð[#¾~¼¬¡™Õu£Î`Ðë@r3°ƒ2`+7Z+`ß‚³T]•ÿ¬v5»âý­œçcQ8Ê‘4œàšLº>âo¢>K¦c²Æ &ùœŒFJã®õ¾ðýß®}tÕvFŽZ¤eäók³µ.¶“¹„}N¯kð̨'‹kÛqD]Iyr;¾åä´G€b}3Lð—×øH‘Df=¯°k€bүݱä~”ñ××<­’×ë©vüû|Ì@´lŒ:K»;z% Vê=%ù¯[jXk¥ÎZ缂íO?¹ýR†ŽüÔ5´ÕŸ`ÌIÀ ¨Á:ÐÇP\(H:P‚ßüY#AP/´ {ÉwŠÖáS”Φ¶Ò%‹:œH%‰¦£7Ç´]Ðvf y}k‹+• ´”w#õì`òŸsû¯üTN“ÃÑ›‰?T¤º¡ºn°Çx ÍIX¹ BðïÆ"P€à݆‚£à=ýêÝç0öÇQ <¬)ͰÁ˜fZ –ÕnÏãÅnàv1Ç9HÛ›)£”9 ÓñJþVcN (ÀÙDµ¡:Ýû³—÷1†5 ž£kd‡fD">˜“ì‘{ Z,g˜ k(ÔÿXÖË´úLƒ–]¢Cö€FQ«ÎnúîÍ ÎšW™ÒäÚ³€ª­òÇ&WŠ5Ýø­Ž×eÅf¦dïËԶت#—`÷fØËÌ>Ó]. Wèé~J6Dµ‰p (c' [ÿ‹ ¡òOê~„|€F Ñhª`O-Hµ'd²`«¨‡ .ã)ø}úÞËfcÉapª -û¼PÐ>닸s>Z›°*ý;9g<Ñê²09µ¡>.ö€7LÌ_;ƒ7ÑýÿjÒÃJÌä ¾èÛ‰<­(¾IÔè'ò>R.lÿ×õù‘†Y«NÔËSÚFZžÖ‹ao.ÁË+½Ç¦·¿¾GW¨+ÊM\£«&u]èžÓ +k°…f%.€!h÷NWÉ÷­Þ¯—¬ùV°SŪò$i¼2|_½ðøí‹_¾SÇ<5®Bë×|N y´Kå¥kÕY*&-foú@ÍÚuájëyØ 2Πs°FƒQo”Äw7¾zeØ…VLAc†Œ¸6×8ôìw&Ðý1œ‚¶—ÚŠ¬E”‘œó$I tÿí§LÈ`osrAI ó·û—¹p:š’¶;iƒ<Ž2ºwöÂm¾ªL ²µî)+X_éÿüØO8!^ªÌ ŽÕ»œólñ–ÛsFžËÕ ¸‘ ð»ˆ#h§ÃQª©ªLÅ É$«U'TªØ%Õ[ƒ©à¼Á³þ …ÞSääfqhj(f8ÆÚd·²TœÂªé/9b€±Öç}Þ^ U’ÁD¾Šû¸´Ç.ÏÉhjVÙþúu ñn¨Š:~¼ðÛ7Ny²AT”/Uæ9ÉE;Ôô†=oÜ„¼o¾¾{£lY†V“˜»‹•Ï\²ì©ÃóCRþü¤<„fc„ äpN›wXfN¡‹þÄz<|>å’p¥~¢±Š¬Šc 4ŸÄ•ÖCíú5ð|ÑË»w6âÜ÷ÂMtC1øÓ´… .ØÉ† ÁW³¡ö}ÿ^Ÿ‚:4‡MÙ·K ¨5sþÄ] «Z[ÏÝÔ`Fªè yÏ [%¼â›ø½ƒ&~gÒâ1:,å¶æº–›À¼· XUÜà<òî ¥P²ØÆ*³ÆSCÕÿÞ¿<VŽÝîW{¡WRŒ­ÔÓ°' û} ;þ|sÛ'¨‹ÄÿÁëbcý 3‰yOÿù]íª¨Õëæc‡øuAr.xï 0üÏuîÕ«P{ž½t.ÞœsM”2ñ•™Ì¬[Ù ;¨Ÿî}õ]•ºBUÊÖW˜‹ŒV‘'× (§¤¶DÑ´1-R“¸›MñÄÛêå)“ÆÇºLZVœ&?€Í¢¤§M W,éÔ’Û«`Üè´÷Wز„3I°½;±ùÔÅs¼¯:¾fúµlú‘µ®åHiqˉj¶Îí×õí·`ÉꗣƱO7z¿b (öŸ3ÝWœÿzQBjÁõœÀïZw´üÞð÷+%­ýúȸÁðú“MWÀ;ÔµM7€A+þ߆Ìgûûœ¹CÞtì–¨5"C.«\T ›š÷zú±kõw«°þäÉd ÉgoU{2¥Ê³Ô:­õ–jÖRe­ET¨PÙ˜B‚&z±ÿÄëäMÛ#·l\>òµo{ÎcG{öW>7œ{Ÿ®WÒ„¢OFM}uu=ƒÆ>1ѨwP5Îè:œ|Þh«}Û‚†í½ùly‡}.Cíe>×òü/V+4:€^.ÜײÙíy§1SFm«–g°uÉ5êOÄd…œVTÄ‚jíºE“3Q7ðC±É ò߯ª’Y.£U‰Ð$ óo#åû±2æDìeå›x(Ó_Bâ›â𳋨äZ±ehµØš¯6%8S«ÁQêü¥s· À8%) d4Uç;*,Ì¿ØÊ~»w SqÛ¨c•ËBü¾U×v©¹ï³xÃ9l*¦`ˆêMA{å‚€] ¶}®¸Pw`]D'ÁñиÀ >D¢uhô{ðûಯ Àsʯà-}\ð–nrš+ NûQÈwh|ðã4Ø%€jr;n:ßÜæsÇp!šÛÎð«^ß`ñÙˆ~O_ºh%ݱïäG… ßädÏÃG‚ÿôßÖd¡¨Ã¥_Bþ}Hüàªû€ig4)ÐÏXý~¾oy•žMž —U^Z”úl ÷'˜„¬µw~AMå¥MÍ€r¸ ±*ƒ«}$r þìúÂE×7òOpSx®Ìß´/J©b¯,ˆÏINÓÖ¦ûöçö§g• Žõ©|F Ýïýýíqì¹® Çþ.yÉSôgñ#Ð(4T€7€¿,Æzàð­Çt^Í-ÝCa,Ö1È–ä*g¡èìíØÕ lÍ÷MŸZ>3~ŠW‰£“j”l¶?àôu±•t1úCŒ!ìÓGë÷ü/]8ˆ Ä@Òl/ž"ÆjÿÊ RF ²Ì›±¦ýfß÷vcöЄg«›B’“´±Àú¾ ÿꂳ›á}<g7þqm´­XÚ^ÄÃa$âý1T€&A¼—«hg3hnŽó­k×HEˆ‹kN¿°Á^<¸öæÃ{W膴ڸøøÄ˜ÄŠä#MM Ç|\ÓÊ{ÝwÈø1yÆ Ïã´rÞ?¸J>÷Áã´5ß’TY–[,ÎSää1O6ÿ±2O¡SƒÜPI±´¼¬Àé°1Ïï»kð­+|®ž{‘†ô*È{ į(…\ü3ÀÍ0‚AÅè+zEôÝßàêß`ØW×NˆFCk†µoݽ¿–ó¸1>Õl+˜KUœ!ÍSªµŒL¾`‘ZfÈЃDJi“Öm)­h`–>™&Ô1æÀR5|–Âõ|näãþ´Ñn4ás}y%jsô "ÿX¨W-öø—ÔÓ`­/0¶·§Î'ä÷¡³•Ïu…4¤î¾s|A}?ï:6XCf£ŽkEåÒŠZOeµCíÔZ™Êãg޽¨?<88êúùMáÙu‡˜o•|j¨ó[® Îhô–ÅÛãó +«Êêš|`dŽ4µ‹uáxrtRÆÞ˜$vr¢>^¿^+7à¢ú#ÄAp ØÂ'|®‘»Ie—Ir4Ú<#Sì;˜r:?בU&-‹À~‘"÷`zš H(y‘¤Ôa³™-Œ«ðµæ*¬6CQ–V*«ÇË`£»ª¤¡¼§ß¾&^®3{œCiÚBò"'ÐWI€„ ÉtäôÚ"$'[#YmAB@J!X4|—ŽØDÖqE´‘Ü€~ Q{Y½p*Þø¿U¸¬&߬Ó9[) ƒ:n/*jkZJf—WºÍf»9ýÆEk Z·l=vá„ÈîSrÖ‘ïpcu,•W&íÛr&ßÙþö…½~/~·òÖÐämú.wx¸£þ­í4­Õ Ÿ4õŸ€&IƒW ô…»ÁRå63-pÇC¸Ž…}ö’b‹Ùd¶àÞk‹ÏÈ<·—!è²t½"/ /+¾™ëª§M~¦þ|×à ƥɽE›]x—·R•âJafbv”†Ù‰ „ØU9Áa#A3èš OJbjrJryfmCUmMûCH“†{y×ZáY<Ãàúfųàcêã%—QÔä¨Wfýx _ÏÌΘ¼Œ¥&|ºv€/ý;~w6öÙ5Ì‚zÍÆÃïžk=ûù­“ëm™…‡Âûô‚M§>¸ñÉ•oïß°sá+Ký'·=ІÛr||x‰óÒF³ÉŽ#/É®JÜ«T˜L´S/èArcµÙí¶00 îýû«Ú’™yâºÃnWa îÞõ°‚ŽFGΘ‚ yˆ]v»Õ‰£G^.LOQî^|%ã2gÁ‰°]rm$^éñhØÜwÇCêo—=àaô”ôëͰÃ??üä'€z´z…AËÔ@nÐd¯Z²}^ÛNˆýþ ,Ç\?®Ë¿KñvtþOõy†Ç{S]ðâ’„\‚Ù—ø\_Œå[ê†t[Œ\•›d”¤(­þ Ó›¥­»Ï s0Özs)0ûMe²Nš¢eôJØ Ç/C¶uÊ\ô ˜ Ö½ºâh$™â0/,F“Ùä*²¸0»jÒÝ;ÁÙ¤vÞQ1!v)ZJâT–ÖXÊíFÆRnr˜Ëa‡Â_ÁgkÏOwQë\ô&{L=8 Þ»ôÚÅBÊêéCêói` X¿qËb¹/ÿw=\¦b—E;l.îö I‘T«ÏËfÐ)Væ)±I‘ʪ…gPLHqŽï¯Ti‰«¼4'?ÏÊ"'\å{P·µ~ÞÊ»É%ñ9ü™.lƒ+Æg{~"³|ƒ«Ñ§÷Ÿ¡1´Û7Ò|ï Rß9†ß Ä’+é²ÂüBà¥Ù1!2‡{ *K¬H“8só5,t èZ£–çIhzivIÝl-ðÁ3Óó8C“„cþ¸rm û˷İÞÔn×Ã0öµYìùÀE9sÝéQ¨ÍC¯FZ'ŸŸÌÔE_W¨j[y8ê>j,×'Õ_mø:¹eW(‘:3W^ vªØ_ç–¡`1xEˆv $ÔO­Ua!¤r­™5G1EçÀ¾°î{¸þ弇ÜLßó.‚vÚJíx4”Ë ä*^­dPו•°»©À§¡E9.IŽH§ÁÐsÌÙv v¥¡éÒL‰Ü–W”Çþ´uÝ PK• #TeW:ò1[ØãJ:¢tr,‰šPiQNqae~•1ZóÏíöË!ÿ\…qVZQR^êVÙTVvøAÀû ~ü '™ãå¾ýÁ I@Ô«BtYm»¦/2¶Ã|ª5bN+œ[λÇmæÃbÿ‡“û2¬‚CŒV¼ššBK䥉š$<ÎEh´ 88LEC ܧúP©Kâð¿l"Ö»´§!|xØH ×î’KÕjƒX 2)Ea¦çuØñ“__¶ìÙ›{0¶&£±ÀŒf¦»²[d‡ó •nIò;y»\íÊté(vëLÿ`+«endstream endobj 221 0 obj << /Filter /FlateDecode /Length 151 >> stream xœ31Ó³´P0P0´T06P0²P05WH1ä2Ð30Š!Œ‘œËU¨`aS027iÆ`9'O.}O_…’¢ÒT.ýp .} *.}§gC.}—h ™±\únúÎÎn@n€¾·‚.PÆ9?§47¯XÁÂÎŽËÓEAmGTaý?ß ÿ p«ƒ¡ËÕS! B)Éendstream endobj 222 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1à Üy…@ÂÐ.Q–tÉЪjûbLÄPƒúûB:t8K结Ïr/#»ò=>)ul"-~H0ÑìX´ ŒÃTY™øÖAÈáªÃë6Ùßô›äCÚ²j÷zCKÐHQóL¢kš¾³¶ÄæOªÉV§:÷ ‹ÿPr4—8n®1§Ò´4ÉÓï™àCNÁñ>×Sgendstream endobj 223 0 obj << /Filter /FlateDecode /Length 4830 >> stream xœÕ][d7~oñ#ú‰Ó!m|¿ ÂC¢dì)ð0ìe˜ÙÙ]²Aþ:_Ù§mŸ^÷ôôL !Àã­S§\庺Žûok)ÔZÒæÿ~µúÉScÖoVr}±úÛJå]Ïÿ÷üjýé ͈$“ZŸ½Z•GÕZë(ŒWëà‚HÆ­Ï®V_OŸn¶Rè”bôÓË.9üôC%e±~KÀ!Åä¦×˜51(ï§óÍÖ-RtÓÌêè­‹ÓM†•ÒÆ?žýbå´02ê5æl²¸>{±šþ½9ûó*¿ß* 2»ûzz\Vyí^¶%ÊRzú¸¼-D;}±‰Fh©^j$hé(¤Õxé-ÈzÑ=“I´Z#%Óóö´ô8}S뤦¯^Òc&Eã|VIjí§7ĦèµËçv¯ õBJ¾—¶J`)­ÛѺ??[ýLTx½¶A)ýÚª˜À‡µö* kÖ7/׿[¿¾UÜI¯£:ʸw0"BœDR–Ä=½WõÍwGHȶ*CiáýÚ%MëÏ¿¯øœ°ÎI·þüÿý3ž{²RF¡íÚ)i…7ë«•ÒÁ×?/Wψ6X¨ óŸ HBxWAæ? 6bsTˆò×ÀËØ¿fþsŒÅò+Èü'áp/©*ˆÓÉ ¢S™i[#‰møgðSÚ÷ÄV·I‡mT&UžŒÐày~k…´PÅ3j€¼àžå©8íDJ **a2?ëŒÅ>W…!Y¡MUg”–^‹¼ÀT›é Œ!½ì¡êLŠz¨:Ó N—…&9³¶ ÞCÙðmõf«…)L`3ÆÑ$KeŒjRy¨hVoH«3IiëN™ UÖ¡ðn¯„Ó¡)-l‚Ô á铆ûrpvm`.Ô¿WhPe1$X"­Ûéf5]vÖä$|JäaÍÃÑ©„1¤¸V¥èsJhç°Þ$\1`¤0ሧf Ðz‘4v:ìPt‰B˜V8PwQLìC— ÞÈ$‰‡¡Bd`.tîšì)hßJiR’ÉÆÒ ËÒ#âza=„C¦\e ]d`¦†‡‡";ßi9Œ4 „– Dt'¤1yû–ýM#ùú2Ø0•¼ kÓÊ$äàD@LAf18Žmèá|ȃÁzkDZäcCR…ˆíx¬˜²ðÄûÐpìkiDr9ú…®%¸Uàƒg‘j8u²Ýö–CQk{ y)j6‘`ÅÐc86 Œ¡3Å48Åb]Á:£²u –C&@mÆ’Pö_KW6M^qˆÝ'c“L‚æˆ$)Cœ£ƒšä“"cÉÈjyJt¦ÄˆŽ!Ó@äIñÈ<Ô¨b­‘rìE•ìd~H!…‰à'GÆÚÀûÅ”ó>¥8?¸?lFÈw&_Úe‰[Åáþ ÍjVž-„T—R§Eäp§Ú¸\Î#ÓMÁ%§ ‘ÄÈ"0ÚãˆÁŒjõ ÕÇ#ŽƒUÕ³e¹$Z »_© Lܽ…*ÀËÁÿ®o.îðÔêÕ€²ËÓ'wÁr´¤4¨®ªd @`ˆƒL*ÅÐ÷zýÉZ>o#}ÁßÇ€¿Äõw>3änÃÑóö0ŽûñÖ Š:°…KŽ8}÷HŒ½î'+HÌ‹‚diIƒ—Ò¿6¨6ÓAå¥/¶™*$òï=TiPV"UýÛLE4ÊJjñÝꤩ”Ý×ß+T­¿W¨®_¡j ¾BuUø µ«ÂW Vˆ¯0µ_ºZ|…ªµø Õ•ãi'è‘úÿn-Ö¤˜W(uÉ€¡b'•íÙÙ“"Dk%hðyžÁEòy¡DMI2$͆Òyª.h,G}9DøPd-eÚØ¹àn©j5 îYì4¹(áxMɱôà„61'§’#¿7ðkV’¶ÑYGrš”ÎæÌ*qÔ‰‡TtwÐs/YT0vT1`QÁ€äÔe¶‘CÊ9³R9w’ƒ‡†jä!‡Þ3HÙHY•z¬µ ‘¶ABn]ÌBáHŸ ²]MÉÌ¢Çŧ®I¤Ê.û„‘»ÎV'9™@¸†\©CXÅ‘à·} ‡ÌâZ¼…pI•!lŽb"ùªb·=ÞìêüäÃ|ú`¦óQRVŒ«)Ÿ½a5ÇŠKE"– †Å0 ®!KC¶Åã),™*uI,›”¥¥á0]6‰\#0‚ã¨ËDÄ_ 2C>!2CÇ4!òœî™”[ rÌ4‡¡‹§ó “”ç ²Hªü(=44òþ…:C‡+¹AZÈ"!É¿BêÜ{Œ"Îñôi?Ï:©ˆsü©»qî€å>…C‡WñÃŽzö«áTÀj8ÇŸ9^ùŽ{±Ö"Ÿ¥ÌG+âÜ‘ðÖí¤†…×]·[<¼_t»uPu¦AízÛT›é æÞ¶ªÎtPso[UgT_¯ò2PÝyט¸û³ï]ôTIní»? ¿!lÛÌ.@, eCRþZ¸ú×Ì.@BíÖE¹ûór¯fé¥Ó´ú¥Ý1ÔB§)µÃ‚,b–0°þÄß8Ãq’¨#ÁAnJ(Žö­`âàé<™:ªäƒh!×êÆ>éôcpêVòH¤"GåKS3:v’GüåÁb?B]âé´†¥ؑݳ4µš—çׯéh“Öïé „^$› Lc~žª=ôõŠÏ™.S‹¬†»'y{„´^o -GQMÓ1x,&Àié"Ͳc(‘ôyÁЧ1‰¤É"V,ÇÿRæ>¿¬ÒŽÅM#JeJŽÆ(-!Zo B«àòߊö Óaà(·¨@=Å$DÎ, õ“7™BÅagr;NY&‘Ùæ®ÓT îžÓ{&+†Astã“e0Ú„,¦K–öÐl­K«©ó¶‰½ÅÄÃ( …‘£Ò©‘.RŒ¤ì8*:wŽ—€Ä³uù\ªhG#«B6'u,!‹LÂXid±i¨á§» ©çìñGF ‚3Û– DÙ|è“5Ùpt¤Ãx¥˶fù$ ï Åx[£‘êTŒWÔ,NÙçR ÅsKQäªC‰CXZ(ªPv Oç¥Ï§„0ðëôå;¹$Űb+Üì”9%êH§†‰l\ÇæŸ­;ìpej¿„•9½­ñðS§´5Þ‚å!m>ÒaØã¶5ÞJú‚¿ X?¨ðÚ[Û?s÷¶Æ[p< ­ÑƒÅ©UÄÏîHwkkì*Ïu¦UžwMŒ ªÍtPscUg:¨¹‰±ƒª3 j×ÄØ ÚL_?ñûo:W.¬ƒ¢óIâ|ýü;ÑwÞ~þ­Á‹ùóo:µé?ÿ&þíš[=~לؗä+T-ÉW¨®*_¡jU¾Bu…ù µ+ÌW V›¯0µ6_ºò|…ªåù ÕUèOæ±5¥oÑc¯¥ûÞs ?¼â@žr·AŽö_=W0p—:”vâ(©h¤92ÀX`ËòtoÆ LÌ7LmJtó ”Šªânü%øéý T¥³”Äñ™ÅV$µ÷ô±Ù8V¸oû¦©®dµÝJz}>£áéd¤ªu.OIáÇ­,òmžÚ¿èÐ f•£DmšJAÎfšºaéæ²{Ô·ÅÑ7B7ßP6„,7Gú– Œ"+àéɳ2fŽÅG>ì¬(-uÞ¨LK‘ØhЇçs|O d.­œÃ QݰìÃq6LÑ”3 ô!¶ñ7qб4¼9j>/{† § etˆÎJ–]M%ªÍBŽ®vòÑÌB渄Za ™BËÒ¬«,(,- ’#4¡†gº<Ì¿ciÖõt•€ÉZt®€'?s÷>læ¢ó®ûñªÎD¸u¥AŽ*`¶4ÊMŸ]ž¿y³¾Ï¶`EBO>6¬š«“ƒ5V³(¸Œ Ž·X" Ü=º7ѱ.¹¥}bs—šÖm˜¥ÉŸ¨Ý‚ÝÁ’êñ·Ù‡òò*åzœB·:·¶tX>ûzú‚ît–R§Ð_ï|ñnïh##²¶é%]öl¨©wR?Ýl­¥>_==Ë7Oât~ÕÝýWz2õdw÷Lç£ûLßѬJI…8¿^{¹¸Lû›òF—ì2ËxðúOôJø*£§7Ý­Óùk|píŠë¿o(þÃCç„WFڋˬ¯_/å/ïbœ^µ»ÛÝÜÓõU!_G·Ggeؼ@šÏ»'óÛ±TçUÿö´k»ŸïÝŸ ÁǤ§óË‚0‚íýÛïóÛž^“¹n¤5^Ï\Ï—í5ýƒ‚ÆQB.±Ü½Ì¾Û0[J="ñ-‰‰:3óýâO,%omš~¾ÙÂÅRqeúežÅ Ót݆/Úðe^¶¡ÞaðÓÏÚì¶Í~Ñf¿jÃß·a÷â§mx1|ÛU¾ÊCDï)³x¾oÃ?Lm|øe¾ØÑ§OÚð_mØ=¦Ûðãðzˆ÷M¾=áÅ—CÒÇÈú%ÿeˆíy€®‡p«6¬ëØþÑ‡è” Ë²{þ°éa `»ƒØ"Âpu›ýx³ÕV:új¸·n†r}7üyÛPŸ´áó0¯&ÿÕ d›ý¸ÍŽ…©:FnÆãNÈo³çmxÓ†_Øs#±œ5€§møÛ6ü|Aqx wòÕêì£C*ݱýÛ6|;vÚÝéüc…WÃÇ^ölªÃ«Sýû¡Æª¡;Ä×½Žtór(¶_ž„al5v²´””ivšó|ȨNGÞ“æÍPsþ:ìíí¨ýµßŒÅ2¶¿o‡Ã—CNÞ“Õ”·{î|HÄùPÞ¿.´{óÅqõßýaywry݆c€«}¹ìËð|ˆìípoÜæ±3å¿*绡ÞÙÐyÈë!²~ótŠüÝA½ãåØ~ÚÖñe~6vÿ3³9/ã'Oýò'T}¨vÖáë¶BÕ†l‹ c74m(ÚеaêíwªÍ(²,$¥ÛÝÏÿÓõcò¿c¯ò?õœѤ7£_Õ!:Îr‚VR¾o6”âÀ…,R;€OÉ Où’€%© I C($>9岈õ{äŸ*Ë‹9ËH6¸ò³?I%“æÔ'c© Öb–Ùeo DŒáD%µåx}vCïsÀ.AýâÅù:È׋ݯù©Ol¯ay½HÞ'söÖ?·ËØä2=Íü.Ãëœb{I‰iIÓrêû¢ýœÒŸðºÈ¶Ë˜cvÀóûºÔ²üt“Õ–²Ö>¥^²J.2ØÌ ¥dð{K%ÌŠ~'JTTb:O¿M‰» ‹:®†ÃÎÉtÂ÷c…ï :Åè®7:›Æ¿Î7w’ûü0œu'HV÷œ‰HîS¶ÿØX²iá-ÙZÇØ6.†u öÐgíß¿mËýͰoó^õâ’ŽÆŽYýic¾œM?°5ÓO?˜OsÃö¨ä¿ï5 ½+¯}Ô°/†Cì~¾F߆]Ñg &í°÷›Õðùw^â±¥– êþÁ*”°Fߊ²¬BÝycÒóO/ÝÊ­é 7ƒ,‘8”g%Ö=€Øºq€û±ÔüPjð“îWö»'T´ O`8 ƒ4âÖ0ÓÒÃþ‰ öÈendstream endobj 224 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 4034 >> stream xœWiT×¶®²éª "(mˆvóÀGeFd§0*2G‘)’ˆ˜Äx0&&Æ!FÔ ‚BpˆDDm•M4wWç´+ï4$ïÞü}kõª>Õ}j×ÞßÞûÛß¡)qMÓº~ñ)™ñɱ1Ú;+Á”¦¦‰ pêóÕ9baº…ôDHO§dÚ¤[†‚Á$8bÑ)M/ôØi¹&tÝkkÏÍi9éɉIföóœÌ6æ˜ýõ™WüÖäÄT³Yd‘Ÿ²9MŸš”,߸m«Ùª˜Ô­ÿyðÿgŠ¢¨%î©›=Ó¼¶¤ûlõÍðÛ–™“½1 '607.(>!$1)4yÕêù:[;ûyŽNó˜-\´x–«åÜ9ÖeNST5ƒZIͤB©UÔlj5eI­¡æPk©uÔ»”FyRs©pÊ‹²¥"(oÊŽò¡|©y”µœr¤ü)'j>HQ‹(j"5‰2¤$ÔdЧŒ(cЦ¦P&ÔTŠ¥¤Ô;ÔiJ—Ò£œ ò”µ„Ê¥€v¡Kè?ǹŽÛ?®eÜ+Q„¨IôVg¶N‚Χ:µâ ñ.q¯øwÆ…ÙÍÔ1ƒ¬›Ì6°pK¹ÆwDï¼÷Î>]cÝùº+tÏêê¾?aüÆñãŸè­×û\ïÅ„ˆ ÑöèsúY“ æÄd”܇súj;TÝ‚}-¤Âf¾o¿fðWj¹ë1øÐ[¹ø¦ 6‹áËnog@ÄúB*B;’Š Û»Üúá³>cÉ÷íB1¯)î‡K¬¤åßí[/ÅHñŸ}‚;èß8K*±wEaò(NÉjߪ€} Ø®0ìQA*He,yÓc æºv¬êRyeQ=jBÊØÛrN"Ü=WÖÐ9]Ïú!¾,î|ø7ވõŒ ßåa;ø±õèÌ®“Ù§²ËQ,ŠÍ“oÉJÏNݵŽlòÃyØY¬ä¢6ÒÅ5æûMêqTaR{ùlÛÝòÔÕ{¥úêc Ø­€Ó ºšTH˜Àó¸i1¤g1ØÈz.6q÷;}eƒ4¢>u sàû+è‚-ˆzCçIŸ3[Àë `gä€<3B#xGb1ÂÆ΀éØÂ~é®è¹/-±73úp.£C°j£úD°Ÿ€5f,Â3¼|Ö˜1kÊS[Ž»÷óRi;ûáî{r—˜ Bø1«Íþ(`W‹è>•0^%®Ìð€J-·ÅmšñÐfK@ƒV›éŽ0mZ [ú`_ß²NcI¶i$/Âu¬¤Æl•·Óò˜ w¤»HcÉ:µ¬úU*ñêBµ§kïqäñÝ£9¥¡Ò¢„bñ\[j\±™àJ– )ú¢Ý"\3ÛÑ-›‚x&3[ccFv ØÏ˜W‚Ýï;ñ¨S‚}}iõ‰„ïjùŸ~‚>F\êûGNÉà«ò­Ã¼[ж¸DéÖ´|ù§ë¸~fÿO•gº÷ðâæ0Ù6vObξ;±îöœO6}´%% ùr6­Á¿·6”\o–~±æTÆuô-úzoé~Ï_m.ÈMÏHNÙø~8âüãÏ5Ü(/>$:xt_é!î?qÖÀ ¼à m¤†8B¨Ø†O|â"ñsL Œq”x„m£ÕEB`7¸‘&ïƒzÒâ‘×Õrþ­¼ó.+© ºÐÚa ÒaÒ˜à¶à –y¾›'ƒ3,ÌÂżj¬·ÿÙÛ„ j¶ˆî%½àN¬‘\ a ¬Ñ6NU¨'°§rÁ356x*ØoMɦEy 9yIÈ$}¸7g·Œ=ºóè'ÇÑitæ«ãßœ:väø7— O­?åÿzéi¤MðŠÿäÚŽï³Î$Öœñ ÝdnuHß»›9Lêé„Éd ™¼…b¼g·º&‚ñõî'íÕ="z€ŠÔÞEt§Rد IF#€XJµó,Nzk€SÔbL0Q À@¬ÄZ¸;ýZÖR„©Ô0¸U£'^&¸uàn~&ˆÿ‰î_¢B´¤ôTÁö"鉜é(Ó"Ê€3Ý£3V„Káì_,xE«Eh•JÐ!@Q»˜¾ŠØy²=þ(úËõ‡ä& .a’Nì:õéÈÐË.w)S\ü¾õs`èzO‹ÐÐ(Bf:Dkl#rälAIÖÉÌCï¡ œÓú8K©>¤rÃË7†­ý‹H ªÀŠ ´HSÜÇH^©å: Íû’")^uÓí*&t¬ò¨n÷á÷ËbûTcR)N–‹ñ‡+«^¤Ë`y„jëÅj‚â2cS¶¥fFå‡ g´öXÒ¥´ï><_XC¦U¡ß¡õgbë|ú“@„ÑÏEWÊk«Îµ ¯jµ¼q«íöùx?)~«ýA{{«[S?Ç;&3$\šÒsÂq{o¹e}÷€ÕÿÃkìêíFH”©ƒ§j†ðaH<›Á Èr:YZ0Läê¯ ,^€‹æ…XÛ!¯éÁ×"Xõš·°x­—YEê0BÒ9Äuœÿ6_|ƒ|uþ_åi7}y*†EPªNä—£`yBxÔŠ k„§rxw/æÁ¬¤úv‚Ù¶g‘­Ò„º3~ˆsÕy~Å ¿‹7FYϱ| qeäùXyÒpP)‚ƒ$›šƒJuâJ6‹²¶ãi9Ü íl-|,4ZX<,>ä5ï0_Ýü®r°áeã”_n\{€†ÐÍ­1U1•ë¾õCó‘véñ;¢?%2šù¬ö‹3O”Tÿpºq½7Wº®|/, Qf·[.Øà[€L„”± #(» ٰϰ±älÇü×LóâZ´“ýêÕöV‘µ¿$#Úfåø£›*¼ï%}3뙦?Mª[pV&h=ý}]ÇT˜¸¤›b‹¥Îxîn©’)¬ÞwòëãE«OÜ@\÷•H×wsRâre[ó7}´‡ ìzhè%ñ÷‚'=•àù¦Ç\Högq§&BlËB}ÓH‡ûcl ƒÝ`7H Î4u¯ǬÕ‰1kËŽ6"îy³¿ùÌ@k·ÈªÞLâôÁ•6o¹°æ~J/qÚâåk°3ÛxZXìò²RÃÕ1 µ›œtèËÏ¡’ÿXøŽÏÝ¥í˜B®È;>Æ?Ú+Û‘È,„õº]^Qr;…h0yÿ&óža‰÷ê-kãe…àÛøË+t]Ž=ìÍ‘ÙëÌ4[Û/_²8ðÞÓá›·µÔ¶ÁN ÖÞ|ñ¤¹U!Óòœv(ÿÙnXßO$²{9< ù0È&~‘•„M`á”×Íe—Ñ}î©o‹õ2ÇÈoî¸H%x’gB¸ÝT¬§pƒº–”K‹=\UÉIRn»Šf°>XW4·¢›&øâÔ¾óTÎÕ—•Õ×U$†¯OÛ´4R*ùLð~UȆ rHô iú¹ñÜ…'? ¨ý×vZ8¿óšlo!›™©3öMŽ&ZÞèx$‚„:fcãìð<„î`ÿ—Dß`Á"\8!p¬W%˜sZ“aÄàE(â5…ÞBá߆Àó‘H…“ø—`|üÌCàþ³Á˜[Í(±E-vAä8‰ñ«#5”_Áí4hÿŠx ±Þ8Çz¹´¹eô3J„â#zpÓaæ‘®j|÷gzz}ô&PÔÿ»SöLendstream endobj 225 0 obj << /Filter /FlateDecode /Length 162 >> stream xœ]O1ƒ0 Üó ÿ €D¥Jˆ…. EUÛÇAp¢†þ¾$@‡gé|wòYvý­gA>‚ÃE0–u Å­ Fš,‹²m1,Oœ•²»+ÿþx‚Í@f烚I>«úšWåB§iñ )(žH4EÑ6Æ´‚XÿIG`4‡ó‚mFUãî?•M%Λ€kÄ17ÍMRËô{Æ;ŸR°A|qµTendstream endobj 226 0 obj << /Filter /FlateDecode /Length 161 >> stream xœ]O1à Üy…@B‡ªRÄ’.ZUm?@Œ‰ˆ¡¿o ¤C‡³t¾;ùÌûá:8›€?¢Ç%0ÖéH‹_#Œ4YÇZÚbª¬LœU`¼¿©ðþ‚Í@fçw5ŠÓ¥¬Ú=„^ÓRTn"Ö5쌑Œœþ“j`4Õ)βB`ñJŽæÇMÀ5Fr©4-Mrëè÷Lð!§`ûAêSlendstream endobj 227 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1à Üy…@B‡,KºdhUµý1&b B†þ¾ÎÒùîä3Æëèlþˆ_”ÀX§#­~‹H0Ñlkh‹©²2qQñá¦Âûv™ƒßÕBü).]YµG½¦5(¤¨ÜL¬oÙ#9ý'ÕÀdªSt²B`ñŸJŽæçMÀ-Fr©4-Mrëè÷Lð!§`û@®Sjendstream endobj 228 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 3445 >> stream xœ­— pWšÇ[È qLÂDYTnOHÌ$€I`&S91G˜Ë1·ol#!˶dYÖ}´¾nÙ:mù’-È6¾m0Û„pC ¹˜2[K2;»³5;ûä´kkŸlÈ’©TjjÙ*UéªîþÞ÷ýÿ¿÷"j!f&$&oˆ_ù~Fž?#¼@(ãS¿ËœØ½€˜³ú­¿˜xn[,Ä!&ª}þ“ÿ>|æ Œ'¡@ -2%He*ynvŽ"nqú’¸øW_ýÕ‹q+–/5îí‚LynzÚ¡¸Ä4ENfAšÉK’¦çf*Tq‹_ËQ(d¿Y¶¬¤¤diZAáR©<û%/Æ•ä*râ¶gfÊ‹33âÖI)â6§dÆMºtú-AZ +RdÊ㥙òCAÌûÐj©L¾®PQ”v@•žXš‘™µ5'7)9/¿`)A/z_ôÕLóÌoÈñY‡f‹g?öøcµ1¯Ç|ø¸5võñëyOØÃ¶Ø‰ƒBi¡pB@€6¡(q~²T±Ó¢³íPJê½ÀQÍÐÆµ@„˜VYèòAƒÇÈšéBNS àwW{j?Cây¨LäãW²jN–h—h  ÄWWÐG¡‡ë‚£ÐÅtF®vVA½Ïå»,®V†Þ?AÚí H4`3›´$hr­mGw‚üèØðˆ>^„cƛΠ‹šÅΦÐSòì,MÜÏÇ”î¢ r«ŠÉ݃÷:Ñ’ ¬Z»­ÌJreñI@–]Áf_‡;@·~ÖuΑ§§ÞHÜ“½ï eÎïÙ 8¬/NUf•o†4V2ËTÍù†ó]@¶Â!­ÒZdÕЫùa«ì`•hÝÚú ·¾®‚ŠXƒøQ …áë·…¨jâM±*²z4´µlŒ™æLž4䘛A2õ§jiô±-@3‘èo7¶žzÝKÿWã×·ákòã䫼€âÍü'âK¨ÂÓçÂÝ“ Y¢›u[Ó^ym/¢ù üŇ쬯¸ ?ý1êFoŠn4%¦<»rÉÏ騰mºohÆTuE6£Nñ_7ž{‹åçò?çóÄoßûÅ¢¹h!ZDñÀ7‰Ór{®hF3Žž¥{¯Œµ÷9Öºg{j!—LX“"Ïß—ßB˦gòÎeá×üSbc¾Éœ ¤NJ£‹©¤[!ĵA;t2¡ïçmUÑa¨ñ°®´‰öÊ[rF€D¼ô¥èù»{¯®ÛôþF“…6œÊnÊý‡†Ql-£?,»;.Þ>dOCÖŽ¤\þÉ’4:gíi>‰æoXëàܸËõèwŸùqÅK~'D•¨_ì¨Fs¾„J8Mš2«² ädq­±¦¾­ºgl½Kº7³èý,Êp>µFö–FCmDs±4Fñ«C€. ›â/E—½eRµYfÓÑÀ0|K*}œÃï Î!]´W6,]•R¾¥–ÿÈ×DW‘ßÓíb;±fŠþ³¹`½Ü¦²)hF ئ.¯G½ƒjFk/ŸJ©ÉwE¥÷uvŒz0¸Hþ™É‡Eñg/ p}0ǘȼ<ØÝ^#¨i³(#Ó‹ ?§\ü½re< øüÚxJžŽg[™°’:—&¬«ï:‘Þ·‘Éy>5«¡Ø£§êÔ –cú˘’IYáËüŒìþu’ÃãÈF•»Ü˜aMÞ|6ç#w ?¸³¦v·’9³Ý0L¶6]»5*ÎsQ±è–óüZ=ûŒy‹ÄÕy§HRJ¿&@ñ,Çz€“xL5J¹¡{?ƒ^%úÞk·D ü ¥´xB{ÛÒÎüy-®ŒBggtJ›Ÿ±ä]L ëhME°¢/¹é¾‡^:Û0. 3áWøƒ ²Fë)—å•Àá‘다$-+g/û ÃG/t¼B»ƒ•Õ  ûv-Ïå_0"ƒbþ ÊÝÞúí9 «|–B¹þpñ6:oEy dñç O u:)_FGq7ôC Tªë®: ršiv…™*ùµ2G d‘¥±áA¥ÜN\†¥á¹b‡¸Á>cµ*Ï””ݮɭ’„ð3ù—øE+F6]lî£}›‡õÁøQIõbóá-«¶¹ƒ¹Q5Ú‰f×Э—†ÇŽyÁµÂd·é€Áí˜*å¶ hÉ. ÑØÄËbÎÅ:pa«¶¿`˜§±ì "ñtü£`Ð[“ÝB-š\bȆà“¼x2ùŠBÑh9ZAÄÎéj5ÅL>&Þ§ù`¨ß7ÐôÏ´ÃËy^oÿÅë§Œ±a)Ÿûlô§i’c·­æOóÓ“’@ò¨<ÖÞT[Ké÷kSrRòöeç€ ¤Õ¦Þ“}Î_bJg¶¥ôûëÝ¿}ÛLMu }12ëÌÂðêð,±Ü^¤„BRÞdñx{ã#i½›Ðûù‹[~ž¸ý¿Ëè6Ûìvƒ‘Ú‘ð–4ÈÙðùûïºpªë8á4øw“±ßšˆ &þ;{ºœîK@Öyîû‡Ÿ%*èÕü½Éç½¾Ì4ÙÙB¬–ÇE?²==(}g7SÀH,z»Êà¶WÐG ƒ aäwÜGþƒ€PÀJ]ÐÜn=î›ÇŠÎM~Ç)9[H0µ*ÝÕá™QøgþÙÉßþLº ‡nè^¦ûAìÀ`*K·ÇÍV²Ü•.Ô?PsA€V£Åâõë÷½ FÒˆ'å¶ÊEÕ4·õuª¡)¯NîWuùÍègA·™3êì6ƒ.\¿½D ¥Pæ”UçÕj1l ÓbhxxàÛ¯î êI·ðjíå&J¶§LSjлT-ê &dÂ¥J,Ž Ëy+h_{WCPgð›ûK¼øÓáü²2½ž|àÓ)[D¬z{‚ur?å £ÝJmå šNy(õQü[]yÄQM'7ˆM²-+Ô¿Km´M=­Gy½Ž)ºŒ,]BTŠÊÄÝh+ ¹ÃY}[òôóüšEã _ýP…ô}n0Üì<î:âí§ñzùÏÞz ‚ÒÜ8þ7GÇ¢ÓØ•õ_„õ!ôÊ'Mì:úôºµ`ªßø_ª¿)òsöÔýÞsÜ7Ú‹[Ô5Ê\“Y:Ré×׃õGû3‚~)çŸÎ5Qż zÓ÷;}^´fR¨ÝÂÔ›%º4ížl°‚œ-¯ïuU´ƒ›lÐxK ÕÁŒ^Åà­ƒVvàÍú—x³jFW±MæÜ¢óè1zI„žDóþ‚fÿaÃå8šïÿ1Gü]²‹lªÝÐËõ`íöOk×ývi®&-‹§ƒyWt»ÿà•ᥔ$ÚjÛ¿{9ü\ÑÄrs*ýN!ÔÉþ„\¬ŒË…_4yËVnÇé@RîÑÖq¶º8êbA¬ü?ä‘“óÆ´¬\)Þ•OOWH}.@?¢O'„â¯TuX--Ây*Ð\ßJñ›&]bc”ƒÜÖ_|ô_þþA‡ã·ãø¨­lnõ÷ÔŸ¢Ï£Užœcê%cÛ2–ä["ogë=LÞ&Pê›KM*Ì€‘âÐD|›íº„Ò. 'ˆ‰×ÅìT"=F`lØÍFjß–¬OÞ6Ð,~Oò1EÜž ºQQoù¨¤Ê:¨u3Cšú ØG.ý¹åo¬éÿ/׿l¦Xw4¾‰EÇØÕVê}>ÃncŒ¸­7NE,[奆/´뉳Ùãx¯z =}3}xg-mäÖ6*yž=~ø‚¼sóî—(ø¨ür(µ¢3Nnv‹ÈåãØ©˜¸Ó™è„{1]Gœ•nÎ…Š]swм®À˜bSGâ¢)±ºp+jT;ÚÀaÈ+‰Ø…š´cˆ3`Qoæ7ÍÃušp8Œ`Ô˲75Š–45£ù@v‰‚`5™#úˆ¤Á>C–Q¿ $Úi]{ÀMŸ€SÜœ„Qf$"]¯ï¾µ¢RPãÃ’Ó³P‹Ã3ÇyÉiX\ÿ¼zZ¨C®]D¡Ž£ôfÛu5¼# X=ŽlÐkŽöö6·vÖQ ]ÕBåC”ÓlWnNR¥õu4çé¦+ |ðŪƽîPm 80Òs|`è&±GÇ2¹£ʦò~0ØÐÙªlß¿uÿ޽TÒ;%F•‘ŒE}`b^»àÞ™é¥ü4¢ï/eî$Ëpú)“ìèÊ9ÑñSõGýW± ±ñ-V˜Ên³–Ù å»J·ä>L娉Áé7!Š=l‰s§ÍúÿpÌÝ™œM™¯í î†m°+?{ç£y«j]ÅyûÐ/œ-•MÎÆÊ&oï4ë^¤Òu ¾DEoàëp'ð‰C¢s—?8%£Óð¯õoB”}NÜ®ð”JóeùEµ¥Ím-G©ØâÆp‚ru">Õ;34ûòcÔì¨_bf<11ñ?°-áüendstream endobj 229 0 obj << /Filter /FlateDecode /Length 4226 >> stream xœí]o·ñ] ò}ÚK}~¸MƉ‹NW €Ó‡“Î’K:E'%6¤½3$w9äqï$ÅiÓ"ȃ'w8œor†÷ÝŒõ|Æð¿ôïÉÅÑÇÏ¥œmØììè»#þ:Kÿœ\Ì>] #½gžÏ–§GñS>ÂõÒð™Õ¶÷RÏ–G/ºOç Ö ï3ÝË9ëµ×›î@ΘeŽ‚¯q²õÎëîF¥³Ü˜n5_H)zït÷WÎ(íºë0—1åþ¹üÛ‘½dNÌ`L9@æfËõQ÷¯ùòÛ£°¾â È${ÑÝ.ÅÐ[ eÞZÑ=Š«Y§º§s'{Á„…Å€´ q7†d­É÷×D%̈3i\w2·îºW½ö¼{ö?“ÞImÂdÛ"›œ‘J•ß Kà~AJ†Ê@(ÞÃV@¸o‹ûþø¹)¦,°`!uï`;Z~ô"‘„|›Ás ޼þ¬)­ï3ø:ƒ« ^fð¤‰Äi„ l™ÜDÊ‘ÜçMMÝf@Ôè¶IØfkÃ2² ꦚ¹ùBY®»¿ä¡/2ø$ƒ³®L"ƒ¶‰_Ì[¾C²ÞYÇÕø= 8d®:Fp™àwF7£¯‚ç´Ò:Øs˜  <#—Üûî-ƒÓeívç{çÑÅ¡£fL{ƒëvLÉ!œÑ²»Š˜ ·qnÁS78¬-!²œäï’SõÊb(åoPmÓ"VK…A$,â ÈÎí@0XOt«mpóÎ ©.üµPFuä@B‘¬èÆH,ØŒL5tx‹£i:zÙÆM$alÝ–Ázà¤ëŽ1ž({y…a½ìVça/dqžÃÛ§zXà‡9JP#'âóYÐ*)Tï9èm\l;ÆNç¤ëVTKÖSZt›‹$Jî;×N('’RAäİ3"ß ’u<©S6)nàY©_ ä/‚£t)G’€oJ"°Njj D_s>´Ñ0É…/´ñ*¬.T¤Z[a_gQ“€Ãèx@_!ᨵx¡¸Ç\cŸ%4(GKðÙ§unÌp¨’C.IC︠¢ïƒzB¾&‚@CŸ¶Ä9ÍèJŽã„h®yÍ|!0SˆþCä¡çZ¸böë¬øtvÚÓtím)bH[Ë”IòÜÒ °ù$Lv3Hc- sm¯õv&{ä}RŽmÔN VÖr«’-:}2C/y˜ÌòGëûU ,ÖR¾^Æ©(ýã¸Kwïâ @ò2º¶D¹ d‡ÆQì’¦‘Å6Mo•aþÀ6 Ÿ :«g Ûe…Ÿ`Ðƃ’ d\oKŠÊö?ÂÈç•ÒI/S®‚$ÁäB¸Ç9Š#%0UBý'£ýB^Ý  Ïiö“¤0H7JÎüÇ]¶P/SÊ@h«•ÀWã ÄÕ0 4lÊîJ¥ ’ƒžrðH Hñ¥ÓYø|9‡ÔÉ€G:!˜*ˆZµ™jˆÖ¸bzØ¥ñ†ïr¸J|Ï`/1ÝæøÚÇ˜ÏæàZŎŘƒFúÃ* 4+x éàOŽz\P<)8DùB´©ìÎBxd C3 7iѩҎ¹ l˜¼ô{-Χ:x¥à]1~!¯ •ÌØ` ØDú„“Óî*幊rñ† bLs/Ð)˜@ÌK¿KâN88zôË”æÀ/®Û!iH~Ìyƒ JŒƒã™F Æ1Ähd^àm/¸ˆÂ_}uËw Ü€²³ëéš]yˆjv\ð~¦@â°f× ¬%Ü÷ÆgÄEñ½ˆ°õxžt²¨K9<ßbm $8¼%Wþn»mwiÖSŸeðë¦Ò?ÏàÅ;‚„œÓûèÞiSUýºÒ…‹ðnµŽUuæl¦OªI÷Aé‰åz+c 4ÈÇÙ‚]‡:ÑÃG‚»Ð^nÐãòªÑ*žÇœéÇóXhóÁ ÕìºÂ– ¸„t®0aÒCG¼åé€!<;ÎL¥ Æþ Ž 5õÃuA=Q!Ým‹-ï.Nˆu<¼ÚžÞjØ´a-„¦HçׄÉ|íõdx%BBÀ`Š|ÊWž§ U¿8j2Žën$•¾h  ÃLNŎãëˆ;‡&údö½‹¨o3…R½RvPNR $§º¯3øeɳ0r$§ErÕBoÆËë2ÐØ "%é¤tNÂ6üt_ÃÀâÉþ‘%«ÜmàÆx¸Ã —S|ð{:vñìëéç`‚Û¢Û=F½{5õÇ®éÏäà‹«ÖþOª>…Õ}^uê«èäþ­ú`ñû‰ßmÕ7‹ÜŽndŽ“EÏi²{—¶LèNk ¹ü0G¿’àÂ.àHÐ+ç÷ä>‘²ÒÜçñÞ66g)¾ó\15m¹2“/%) Úí·çùœV¶–Ç÷½tÅxâ^M4sŸ­âÞ4wªÖ¿Ø„fŠTkETüÄBý¬ —«g2öëÞáYÇı!.Âøîá]Špx^è‘WwíYFépü-‡êQR]­gþ—¯€¿÷Þ²ƒµé:3ü­·ì!äLÔ¦—c§¸0޾¶xPÐ}?¿lÏZ"ý°gíÿ¤ÐMä`YµÝMõ³ ¨Dªn$¢¿öút½÷}ýZöáÝV;©ÜÿÕħ“$¹h‚„¤¤ünGékŽýüެıÐ÷ýÕÑ¿wT™3endstream endobj 230 0 obj << /Filter /FlateDecode /Length 4490 >> stream xœµ\ëoÇ‘ÿ¾ñ±È—ÌÂq¿ ‘ÏFçCtr€V|IÉ¥IÊ’ƒàþöûU÷Ìtõl¯–”æ8l4kªë]ÕUMý¼½\ úßðóìfõõK­×W+±¾Zý¼’é·ëáÇÙÍúO§á°ÓGåúôr•?•ë ×Þú>j»>½Yuasú`­¨`…îR€?=_½êþx·½ŽA[×ÝmNDocô^uר–ÁǺ·X*|æBwFôöû"¨ Œæ0»[¬PA«îas¢µî5°ì.i­€?t_±ƒnÓùº>‡PK!¼ÝEFadwC_©’\†¨5ÝDzdŸ½gGÜD0JÙ´y€´{úÎ aBvÂ(,¯fÛ6‚qB1Iã¡¿M°1DK²˜0sòw›åú@è9IK:e#±:á»N\Ù«cþyú—Õž®þFv¡,”׋^ùµ¶*¬>Yß_¬ÿ¾¾ý¤ùDµö=YI±PÖ €/öÁûd@/.n/îÉŠÒ¡OÄHØN­Ô¦Ö¬€l¤$”¯ºï.¶o¾>ÝH=¤ÿ¸½ÞLŒÈJ{»þÂþ‚ÿ~ºV2£X[§CÛ¿YIël´s½ú¯ Ê çk¨q§‚2ÑAjØ© "‰Wq¨a‡ ŽHÄõP²4ZU ZöNg!É,‰X¬êCŒŽ³#Cf»ÚŒ»¶µ«Ä¤¤d£¸T€p(äZkŒî'iˆQB‘¸½z7}3íx©e2óÒKßûPAå ÊyÕ{”7jå{Wav*¨ |Ryv8…R .äúå`ÈŒ<T“YÃJ½–ªR”7½sð.çz Ž (Øïý ~9 ú¹¡;çÉ^{‰Â…VõѪ%JXŒåá+ °{é‰eú)[S” E}¦vŒCÌjbþLíx×[µÂQ;Î Ì/©aÊ%ŽÚ1ɱ©î'¥™#ûP‚§òH5zð‡¯dDÑw†þ}õ„¯V— Â8–—?<ËÑ?Ï¢2ZÄ ÂiF¯ì^\¿@]Ë%ûI¢ ¿"@-Ÿ›Tˆ¡Gö*ÑvÜáÑVk”“^1¨q§‚R6ôÎp¨a‡C©ÊOãN¬†áVtåêDéBo«‡JAõÃNÍ£ }¨XLŒ Z’gÌq§æ~[ÉtØ©qù€¼UáÊ;µL% y]É4ï,Ÿ é% ‚`!ûåÁG+䪮 mÑN ÏEh`š(5•ì…X¡Š(R$©b‹ p „ù:,€PKÄ £×N†~‰ü¯qI¤¼Z¹n „p{t¸‰¨@($" Qˆûà"T¹O!J îKe²SÚ$ ãYZã’ä¡]§ÌÛuègÖPŧ!»ïƒO[”÷‹xÌèÓ6à²(H'ŸF•çäY|Ú#ûúå|ÚzÜ~âqqrjA—í圚\Æù%´<85iY«%”2:5­Y$YNMÊÑ‹(%;µõä1M~A…x´ð›ˆÏ+½õ”Òû Xž]z+*½£n–ÞjIÁ!ú‡•q°iÍšUÆi$]5« Ò«x¯jšZUkU@S§jb*â[ "óç^ ŒŽ°f^T;¼ä4ØÄ•C ;”Å^E5ìÔ¸ö*Ti£Æ¤"qÁ1åå a£}ïac·óåÀEŠó¸ÚËò›±¶ µš%Êc\ 4êØÙvà¹ÂRa 8š&ÂÏ,g&ÍTËi†º Tq£bÈl|\R1–¢ÀYxRŒ¥×àDΣg ž—’Žõ””ô,ÏNI)ÉHÓLIzIÁ.K´ì­)H"ÞÊLôîöñ~WlásQ".Cy4L€‚ìî/¶çe´P¬ÆõAâãô]4ÖL˜¡3Ýô¾¢®Š£-Í„ðe¤ bÊUN¤fϰ“#Y\Ñ·Ï;׳>ê³S{j)Ó¦4u¥¬×}Ù C^­%î:·&¥úèPýTž—©Ü–üãaãúKû‘©òÂ!¸øIghªD7>Eàø­ Â2²Ç/.’#âö‹ðI^¬šw>Eâø1Ë"œ>v&-7ôà[b/rañªû> ”nGÃð ÿe5àÐŒ™Â¢ïn7Êa¡m÷> Á½ö4&Ç&*WÙ½NCr'½¬»Ïò†Á½ý®;gÿ4¹GHŒ±¦ç ƒyépÝ%ÛÞèΞ¼©¦ôY4@–sÕ¿z"qCDEˆÚT0·$#…¶tr”¶ÛÞfµ(!‰1Ú¶P\Eg7³’ù]B­ëï&:ͺ¼J/$¢5²Û½/¯7îRHÖž¢\XF. ò&ØeåBÍÒöŒvÒ"“q‰4Np“+ã«ß§—.eQs2’s_²6ÓÄ š\¤juâZšQ}ɾ.2€Éy€ê8<êÎêŒkî¶(vö6&JáÓC ö`¬ڽ˦®4ÇzŸQ‘áÅG"¦²Îmzèa„–J™…‚¹ë7'ôv€”vŠÝ ,Ìž1à ¤Ü§I¼<³a"'QR™ä%„ ƾÍF¢e9„Œ€?‚Ù¥/%ãá&kÆ}~ô:{?ÄôkÞC ø”…L5ípšžý´’ÌIÜ;3šõ «—C)y¡õŒBõˆhҪͱˆ,4xÓ3F:3<¢0!3‚'KÌ3Þ41gÇJŽa\_gƒM m¦€5÷*"¤@TdC@0*¦ˆP’^‹!A Ã]J ¦x2%—7 ÂD•ôNø €¯ _5‹­˜ý@ß‘¶k؉sêß–wjÌÜ+#˜È AžˆôÆÆŽ¢ÄA’C|UŽáÀ<6M§œçÏHã¿lPKÐ+´n{>CHaÈkÒ§ÉSOâ&GS/]rž„ÎKË·9° þ‘GŒç1^wüaÙ]ŠGNXU½Ã;˜p õ…)Ž]ޏÉs“8Cñup™3(.7>Iˆ‘ÞÜHj|7Ð4ìße× ¦÷p•ƧG€c,`BÏVÉÈF«E¢¦ñPĽ—€Cß hµÒ¸JÆÙz™'㡊â:èaë˜ÝgS…ÇZ~ø›üÊ%cñ³mVŒ…Øù$³ô"QÎÂÀ˜ŽGkžý/Š6ø»ÎÇL}A[f ç#†{è<†§GuþÍfLÔÕ6'j8=¦¢«²ü$RiêR¥õ³rm®¹}º*}úd×_¿tU½ŒòÜjÂÿ¸:ýWÝ ªp£CÁ‹ì?-¯çK»oÂÉ23‚‡úX–eyW|_– íEY~,Ë›²|ÛøGWÖgeùºùÝmYž·Q|×$h[–™;¤§˜Bé°|(ËߣV¡ù>DZvweÉ0lËòº|vR–߀‹ægGÉùǦ`ûß&i÷Ï8㦠{Ëa§#þìDÑJfùùESÌ·s1ÏîËr³ã6²k*˜a¸.Ÿ}[–Ì",¼üwYþµ,_–åW_KFƒe¥õeS¶ ømÛ¯E ß–åo Àë&¦’ݱӶ̓Ù°¡–yŸ³–m“öLpÊŠIËDzܖå뜔5jX7ÝÅ®ËïweyV–ïš'Ü—åï79Ëxë4û®i"ºizÌb™m2{»;† ²MáÔ÷’(H÷ª8ºÉ]= h;0¡hÂþ‹8ÁÞ5‘1 LŸ(KQôyÌ ˜©í²à Û¥œ ž%¡]YÞ4sH_–²0#šª,MY‚iqÿø®¢§E¸”©&·¢ìöei9,[·³üúÖÔ籟Áš:Æširi¿”5}Œ5Ë5Ì•=²Vñ>±¦ŸÁš? À\0I©yÅ49æ«îw›ßíUT–þ´AÊ:HÍÅrU¸þ±é¤ï¸,štÛ{ÔýùP’þØ©&EâšbuI9iJ’É7Ð3f[ ÃÄ Ot,-°œ÷mY†*³¤€àP+gU‹àþ\–Ìà¾i†T¶TMÎ\S ÌkþTvÿ\–¬hiÌ\Å5 Žq¹ÎÜjÞÜF¸ŒŽÞççZûpßAEºÚx~Ë`=æëÒ—ÆR…Øûàxƒ¶´õ_é 6SŽõ»ó½Ãi‘ºst„ÅÛü™¶ºž°–Ñ«\L×A‚U¡ê9ü¦ºñLÛ³Kköé’•H–ÿ}ë|^Ý`§v&íS§âÛU ¬ys•Ú1ûÝŠûA?É ëdŒ!H›é‹ANsAB÷U÷,Q6¿Ú‚Âþ]:u-øöe±­]i•Ofæ6èÆßó&)ÉáÑAÇÛÔIŽ·nÞ5Ê·]\qõ~÷{ w¿?PóŸú±yo¬ä ·óÒƒ¨G*ùï*eÊKƒgÌÍ!M\.'£¬ç0Sßa瑩¥ªû'ý䩯 ½”VwcŸúL¬eÞ&l›9§žÊ5wB`p}”¡ñ·­¹Ÿ@<Œ×ä®Kƒ‹Y §êC»ÔŽiK*æŒø}1&žÛŠ=ÞQÆ=­…´±—ªšÀìßRØ-ôcY¶ï9ÛÁÕôBùñ”lÅÒ°¶òÓ¼kàøûGƒVË›Wƒ¼%¶Y›BX.!ÐÒÀ`Ç;»«¤á Ÿ‚=$CðÆéCa5ÅL…s:7È}hÌ'Sª§¿ã~“I¢™h&k“^ilÁ§iìÓzz3ò˜ef‚1£ZÓœëm>' ‹ïë•°e‰äAíöõƒ˜ ªÉG‹—J]|F0Jc?yhÊíÕ  À†ñ–½*Þ|ŸÑyQµ‘Y²âɳÙ&U½ ªäÌ©ƒpW–Ͱ¢@¼ÙÎò§Iøj^ª‚¨Œ3Ì«ÛÃáób>È0F“çf¸eS°«z,HIÙKU›þP¥ôqBÞ»–0MLÆ6Ê„µíX•õ9Þµ‹Ì¦ÂtϪRV=îÖi´•ç3i½73ªæ3Ùœs&§U5¡Imrå‡ù%·U†·žLƒ æ_•Jµ=.ff»Ë< Y[ð& n, ã°B•îùœºmö»ÙÌ=O™ëpÍyÊÒ;Ï x„9Ó§p`ô<ÿ1Yq›LMDmèý‘ï«1ày™Ñ[šTMOõDõÏt¤·6 ÑÇÏ*ÏËu°ìŸìgËùœg(“t˜YŸj —³×"ò›îêv*C¹{¤qiœ—WP&\&õ&Ú\ T³ž1Ä»­Pã4½éØ•9ÛA/%‚B¬ŸwlsfsÚì½èt?ÉOèS; §³!iLÁûÚöcª=!ú*ÞMžÞ™Ôa&¨è»“r Y´d½ÓMáøt°.]XêK”Òó®»fÍ:ø‚×@½e~ÁckÎrHšÕ̾í*Éeã ½·Sg†Û=,ðÅ”oû¦^N~Èä“Óòí\h§gJì5‡8+²× £êX5û§WNFªëÙò§¾±}ý2T½7ÂÎÊé—GA©õ‡±!ä7íÛc.>…d(Ø„twô;6õ¼h³Dû×&ml²úýKÔ´üŸ²d]W6 b×06`ó‚wÍ‘l";+»l¢u}l> stream xœ]O1à Üy…@B‡,KºdhUµý1&b B†þ¾ÎÒùîä3Æëèlþˆ_”ÀX§#­~‹H0Ñlkh‹©²2qQñá¦Âûv™ƒßÕBü)ºKYµG½¦5(¤¨ÜL¬oÙ#9ý'ÕÀdªSt²B`ñŸJŽæçMÀ-Fr©4-Mrëè÷Lð!§`û@²Sjendstream endobj 232 0 obj << /Filter /FlateDecode /Length 3532 >> stream xœí[K#·¾ ù #‡Vbµù~°6’… ûàrØø y1gÆ^ßò×SE²Éb7µÒÌj/Á"p–SúX,Ö‹Uõ[ÇÞ1ü_ú÷|³øê­”ÝõÓ‚u׋ß<|Ú¥Î7Ý·§€ÐH<ó¼;½ZÄ©¼ ÒðÎj;x©»ÓÍâ]ÿírÅá½s¦¿\²A{-¼éŸaȳÌÑá-‚­w^÷÷@•Îrcúõr%¥¼Óý  g”výcÀ2¦Ü/§?,´$s¢šrÀÌu§‹þ¿ËÓ_a}ňI>{×ÿ¼7B{Xl…’ykEÿe\Í:Õÿkéä ˜°°HÃ@"!îÆ0£@¬ 2ÿ1ˆ¨„,Œ8“ÆõçKap뮿‰ìµçý—8Mz'µ `B59#•ªçKà~ÁJ†Ú@(>ÀVÀ¸oûþçéâg° Ûi.ô l'µp`&¦»ÇËîßÝý̹ïÜ ­¦6æFޛΠâšhã·Ûíóöúq½é¶WÝÃöéùòñvûØ=<.¹³qÞoÏÖg·w·Ï·—OÝ7¿\1µÄmìKÂ2ïa•à¿_A¬7 Îût–Á6e·!¥&»»ÅÉ‚Ki# (H3 ö$ L cA›ž€2€œtW” $ÓTvȬÀgµ²À$°cÀŽùL k&”aÂÌÔH©P‚OxŠJ” ¥ÀA••(J;60AQ‰R¡,l<Š ¥B9aS¡%jc·z˜<€ÿR/ôxJP“ hœ°‡ÑéÁ×AÇ^é|Б(Pðg$ÊÉí 9h‰jDr5X“ˆ&#a“<mFr@‚2BÀEdpgê¼ZB1Žx¯–Ê ^WÞKP™BPÉ[ *S*¹+Ae A%%¨L)¨Ña ªP ê¥Fã\3Ýi°;éx@³­`ט'à‘þ`þPôC?qå r&ù9”¢* Ï1J8ü 4@Ê|»’ŽP˜(KHB²„¼[E$ÀçÀ`,¦òg`8`\ç¼ZXiquÌgÚ Ì ¤”7‰Ó~pR°œ74§hgEwРLfÝ-npÁ£ñ:‰8é)/+íâ“™sÊdV’ëh¼Nf8 eŒ¥3ç”÷ ×Ñx$\mk*ŽqèTrˆ½˜¤SÜDWÒ‚J1§Lf¥=×Üë7ÉwæÞ¤§ÜÕÌWÆës }Ž¡:†öVÑ;Î’½Qâk$e Ï8Œ{%-‚í‹ó ÿ¿{¼Þ? 67—‹2yûæ&{yð‡ú”W Žed)áßxÈ·Ý<,%ҵ臎×öG¨öÃr¿Y@ÉçZ‰“j]BGãH›Q¤±ÿ¤Œ2€ E) (ÈÐiX ¡€$eiÒfí ^V~ Å2Ã… ;·æS”_Då¤t ¤¾L*& L!¨¤c‚Ê‚JJ&¨L!¨¤e‚Ê”‚Õ\P…RPè³ÐÖ‹ñ×™` YŸU’YlZU¦À–››YŠªªÌ@¨vtMÖʘPüçu”ɬ˜§Çë$âÍùÂp¨ h†oP&³’\Gãu2ákÓ0z.7(3îA®£ñ:™áFc[p¢±YØ>:`å營„ŽB§3äPÌEŸýô³Ÿ~ØO7Ñ™4Õkƒ2ã~Óðç×óšû=æ_Àyª×9e:+Éu4^¯®3÷ŸE“CëEuæÞI‡Ô™û™¼¸Î tº£Ê”ªLñÑUæaR¿YH!lu¿\¤¦“P>Dü‹|¬¨n– €,H@o– €œ‡X$m!bFVP!ÔwÔãVHI”)¤pÂíHü“|ž6C ™BPi7•)•¶CP™RPã~ ªP^…)¹ƒ3Ê+LÅ X„“„¢¸×õåÉIF•/FýR £ò—E¾Ȩü¥@F‘/2*)QäKŒÊ_ dùR £ò—E¾À¨•\rÜó«êl°–ŒªTÉÜàL¥Ê±Î–Ü«AN.j§×€0 vóK…ê6÷pR Qu âÓÒ3®A™ÌŠ'ÕñxÌp©ƒ‚:iƒ2ãä:¯“ˆã´ª’^™º³jP&³’\Gãu’pvoíÅŒ5uÂÍÌ™îZ”ɬ´Ë£ñšûý&zSª)3î7ðx=¯ÏQôÿEx¦{KÛÉ»e:+Éu4^¯®Û÷Ÿm“CðEuûÞI‡Ôíû™¼¸n—Õ\`¹³r—]¹"÷Ï;¥®_ÖŒ‚;Þa[|€BŸ†‡M XÛþ&H_= o…@(ïû+$Æœçýß IÏ´ý]y-ôy8çû‡ðHÉJëŸmM%úíUD®ÂS*潆‰7Õê¡8ïÿRøµŸ`Ñ7I#gÙoÈ3®ír%  ù wçY@7P÷LU>9ؾ’ÊÔuØŠ³Â1Qñ~J{a–JMw»•ãô.QÈ9á½ÝÜr0ˆů ÕÀcgÂÉZ’û% ¢¬VÉâá…šñN(QÜÂõïƒ{Kí¡]|°õ.= ÂGHŠt`Ú'²rô&k¬F1ËÆã+5ç¤õ§™תmŠVàø¼mX®,ÃjSôß#‚y§l| g`^|X·…veÐøÞ È»| Yä c:ÀQ_$NÊ\I ùÔÍ팖¨ùâ_Á¿ Dd€×AçßÅ).Q÷ñq èžïUÈäæë§¸¢bœ¾$$ìH`EIÁ¢¿¾ öeœ¬aÁ÷ƒëkÑÿ 6Qo4àBQ;®Æ][0ÿ _úy0AZNf&ž:>œxç*3u¤ÜAþ™åEgDµ1Iç(–ê~ßc‚ÁÃ"A¡öaÆ´12>ŒtŒs7¦‹o «½¿7Ð}óÑíSÄqcü #.Ié{9”•£ÃÆVÜ€Q íÎÖq+³Ì˜\— “$^Ùðà3nëH]:F¬VÒîâH'9¬&÷¸ÐÎtwPšß[hÒ!QIçÑ5ƒ;3I>i3$ûóì.hlã… gˆ|×9I¢ò)J¤x½Åõ}¤ð‚³ì:TØl cåÔxc¥6úGí¢_½……”«‘÷ ªie\áíŸï¥|ÿÍr%ÂÀ¹„TË`Ïë2|lR7exY†ÏM*áðT†ÿéËøl7²òOex[†ÆËeÔ„©‹-}%¯pøˆ:*â»0Mc9´-ÃM>”áP†–C.ð¯å >šC,eWy[Õz^ ¼-ÃË2¼*úÚ6ù¾ ï_`áó&à¡É,µ’\ ÔŠꦵ_OÕŸ‡†*¤5$V©?z½£ óæÂº ]“kr TÕ\­½MA§å¡mnÈ7‡&8FphHN\œþíÕqÚ´11, LbØvvh»Ñ+— €‹Q½†ª—5ÕK¨mc¶•N°ÄÚ^Ô^XÒi­Õ\“ê釯F–hGHÛ}ã[Pe Æ™ôƒð$}Ú¹®FL•M÷WB¤.'‡p(¡v.gfì3 MS251aâcdgá¨_ãyb°—“ÕÌø‹©¡K¼ŽE‰†ó­®š :Ö'©PNõ)”Èyž& ÷BEÛÏëRï/ª¨@k<Á8pæÕ¯{6ص c“„ƒ¢e½šu7»˜íT‘©h&Í`ù•­2¶áGF^òÐrÅß+y[•¢ågHÛ³ØÑ l€Î²¢‰ü¤S!ÔçP´@æÑ>žm­oô¬Ù.ÖÀÂk¨÷J£°ë†ª³”HëhÌÛ,/"ÔÏWÖýÞz+mä€z«]ßZè®™îè!}Þÿ^†¤àúÄÅYH ?•ámþY†päÓq«PÓеá_ªÔg Ëz̺ã0Õ@yFUñ¢tV”΋ܬœ¢PÕ²ù#¸$KZ9~Ìì1ìí¥ÛUíE’6–äsÚ'}ªVùËÒ*`ÖŒú;X€á› A+v²Øûæ.ÚeúºäدËð€i­näjªÔpŒã‡4Í—äœéËø´ ›`Âíyã{ŠÍÅù×exÞâ‹2l7ÿ÷M!›Ø»&ß/‹_|ôî#ÂÚ&Ë…áéî*\‹Ù,ï0·ÎúÕ8½:ìÿZBê»2$}+é/I O‰¿Áo %Kù,4ÃÈýõ2#‰UE=»T9ÆYˆ,ÍÊP7ÙË}O™¡(•âëª2´ÍÅLs±jˆëâ5ª/‰áãÍûK œç+«š%}wû?溌lendstream endobj 233 0 obj << /Filter /FlateDecode /Length 2882 >> stream xœí[[o\Ç ~W‹þ†EQ g ŸÓ¹_‚"€8@‹h’}Sý°ºXV"i]\;è/çr9G”V²$y03âpHÉùæ,ç—…äB¤ÿ꿇ç{ýAëÅÉÕžXœìý²'ó_õŸÃóÅßW‰ÃÁÈE”‹Õ›½2U.‚\x뇨íbu¾×I±\ýÌV4Ì"Â;˜°:ÚÛï^¼[ŠAÇ ­ëÞ-{1ؽWÝ ËàcÝ)J íBw˜8*÷úÆEPAMy6@¡‚VÝÕ²×Z¤lÞ$ZüÐý‘,t‘××í:I´‹ÐFvçi–Š ’+¼ÀjM÷I2í†,q™8‚Q* ›ô¿*úèØ¨v™8¼&­,'Œòd6l#ž$ióiæ!Úä‹IrV_ƒ³œê6Ë^¹!Pô(yK:ec£ÝY¶Ê†Ø,ózõOØP×ì~?îh¯íàoy_å²·JàŒN€¨1±2 vD`˜H¤`G#’†ò&u¤¬‰j±ú×ÞêÙ~§±®bW°,Ù¨ÖU~1ŒËþ×úÉ ’ç8ÿ’D/¥;x-â¢OfÙâÊ}dzÎÚñ:O훹&„n„x„ˆ÷+!|Òžð»nYcyRÓ]-QÚÖ(e† Í¨Ñ&ç“‹1ªÂtìcm¹@¥¿ASÞà’§H^³ä1’GHþ§Cú€e&"ÖH¾bWþ@/‘&ÑsÈ/]ÜhCHñUÍ 59ò¨F²Jž3¿^r•ÁÊÁ9:?Åã·¸ðÉ­éa¬D´ƒP÷JÈŠ½Cr@R¥š'üo•G7Hn•PR´ŸÌjòlƒD"ùÉ×£»w> øÐ×Û"¶eµÖßlxDÒ )pÿë/ÁJ°,IäV[YˆÓÕg;ݰ!žv¬§òáN¿ïð™;§ñÿÃîYarIÒý!µôä•·PðÜ,3SûLÀ¥ ##œíÞ&éµ#èˆd\&QêrèÖDÞº ¡¢ï>.'¸•‰ÂF“ ¡ät…“qU¢\^Z$›",rËßaõœü.:P4ƒ#8Ü¥’P¬)3pH-ù¬Ìó1RÐx5J É‚`¼Ò ^+ú*ã ”~ö£¸ÿ&XXú¥*'…›oGKÓ©¤ƒ—Î5Q.…ÐÅ-NRnt:…’«êʹäéÀ_b´#ùÉS$¯ÉAÒ5‡ fÂŒ¢Nx›­T.¶—À´D¢7’rZ„ Cn‹Ý)èd‰Õ‚+(ÊT‡(5;–‰Áx>ŒFpžG¯æ a3Tc ÉóuFJ§½it%!˜O@x0]ÏNŸÃš BcüZ[Õ ÎêîOäör˜‡a#½,”û†Õ΃õàJ#µ¥™3^X‚§£›šÁ%$úžOR¼°eGkZ¯h¯ì,YÓåËFÝ;½²ÔÛYÊ3œ6érœÖ§:2ÕrÕY“ºtT4rw]…Þ/gÉ—÷®^?=Ü?ÕìÊ•ŒÖ¸&ž—U‚´Ý¯$=Ù•÷Kk³ôTï’BÊj€Âêf]ãÍvM.¼G£ÃLÝä̼!ÖæËhô.ß$SDi–€ùœë•ô¶ŠT±f‰ÉK§DÍ’0Ñ@=)¦Æ`3ªÌ·M…ëe)í2:Pø†/¶Ô‰_ã‚nïG[úŒà.&ñøý OÝ$ÏY’€ïK$?"Iñ4á P~ëuŠ8aEfr öˆ­Åg|Õ=`™I‰^#ùŠ-çün¹,ü™…d”^2f4w =Íý[|) E£™U×—ÈvÅn±ùœuà5uöä‰Y²‹ò¾ÄÑK–ÜÐÑ€V: ÖLÞ³š‘è¸i”%ü›]FÒ7HþÉ_ÙѺ‡}ëè„è“£W¸a—H#¹fÝO6—¤Ë˻ږ¨üÍ ~Bù=Žò€Ÿ |ÁÉ ‡ànGó¢—ÖËäþ­(ŸÜœ"K ºÚ´­=ŽZv ¢™§š‘iÜÇ‚æÆÍ©C„‘ºÓ³úz\âY Ry ó¹¥ý \(猭~‚ú_ÇÚô­±šèH^ Éo4‰C–TZÈ—pþI`¤BBak€veø¤t÷ø}šÎ+¼°F‡Iîß6|€4.ád+³Ï=î|CެvdüÿzŠäÉ<†æ§:9ë,$†Èw+ò=D9f¾Æœõü“ʶ?©”ZZ2á«ò«ËìSÜà׫ä=Cñ É‹÷ƒLλG¼”ƒ¶zºªm É&¾ÙEµ¬¶™ä¥lMÊ"ï²CF«{e;äƒí¸Cy5ßë°f?Jñº%Å^zŸR_û4úŸ¥¯ò÷ë{—SUþu©NýªäL’ýL‘î˜u%™OÈ¢(Ü¢»I‚>9ªrüíãèèrÜ!Ç'CŽö÷Éϯšõ±cýüÏ<5Ë;°£$Éî‡/7îàâ.n³cwp±v4(ÆîÀâî3ã,~™`ñwö™‘´˜!­+|WŽgýé üG­{Ú³ vµÀ’’ô›ÂÅFÜaÄFüÂ0¢U‰¢ÄyÓA¤éuE:]JïRî,šºWdm”ÈÝ2¥iÉ9ÚUr±:뼡]5GcŸ¿¼òðCZÚºª$áÞÖ ´¦mSÇe,išNrψðZ´}XÅR%UÓÚBzXHÓÏ›JHÿÏz™ß1X°S¥¾´¬{ÑA„¶i¨iØ*O/Ló$µ+ÈTÎïR‡ŒþøšTºn¬ŠM{éÖ9ÈO/œ íkÒ·—:ÚÀow¶ØýF»Ý¨Ú^#M»9©_¥ö’F¦ ô&iÜ¡ºµ¶NöÝî$,u´ag˜7ìH!²ÑBÚåt©fö9¾kº¼Ê:ewüöz Z™˜`Îñ5Vtë3k,,)Tw¢ÈD-aÛN×g|Ç*(¤äÔÐG:–øÏ¯bÀp5X1ö;åÌlz¾Ö´ý”ôÎ]§©”RÍšC‹@];Õ¬R–fxº\x¡¼lºÌÎ!meìeÓ~WW ³`§l‰Á€:S? d2k=ÊŠwõ¤’ÖØÃT‘ D˜îg’yäÍSQVئ‰vM—#­±mwfu ·¶¥áTâˆÛ®Ê¨š¶’ѬJ½kìl¹s[_é)–À–.²ÓLëG%'­†1cN6IÔti¾\í}¿WÞÕˆ'ã—}Lgà Z/àÈ‚¿åuûÝ Ò6Œ½¨?£úmG§Ó£õjpf¦\u¦˧²7¸!xÛ®ûÊž¥§mŸa†•fð·ÌXN>œªÞ@¹NüôûÌ=§¸÷9X+h¿,ÌM±®gÅ"KŒ¥ù´¶»· “¸ÒlŸÉ5)¥˜ïuí7͇ù3À*àÚY$'Ò‹ÁïæÒ¨yrƒ)Ù¼Jl >A@( xÆ8-àP³åI©Në#@¥§©pÒyûù.ÖzH—UÓ_>Õ‰ãâ {¾9ÀáRăãÑÔCX»ûZ…³l88šç 9¥q/ä}‡=ì]ê¶¥}¾«éX?ôÕyžÜœ¹AŽ/J+0ؤ=-ÏÞA!oE%ËQÃ3@5RXR—ZÇ“d'm"C¾LÏC¨—)-bœ}¹?õ¹—Þ6ôä>y›5ªyõAãcm@‡ =ÄÄØÐ÷&es½qºy“œ‹oˆRød—mEu¤6·WÈÇS,N`RÓ'$&FõƒÊµ!¡1\nþ2w½ýq «Ñr8³×Óëã¬^VÊÁB’Ý$„ÎpŸn=p&XZ †‰a*ßïý…1÷»endstream endobj 234 0 obj << /Filter /FlateDecode /Length 5014 >> stream xœí\Ío$Ç­ÞQÉ)·wøâžd§SßûÞˆ8q`ÉZ€ 8ï0úØñÆÒŒ"­ì]#pþõ¬ê.VOµF’eçÌí©É"dUéï ÑË…ÀÿòÿO/~ýRëÅææH,6G?’ôë"ÿïôrñÁ1´°ø¥"ÊÅñ«£ÔU.” ½vrá­ï£¶‹ãË£/»–+Ñ«CpÝùRô6Z]÷H)„“¯±±!Ún _uðÒ¹n½\i­úl÷øª‚36t×ÔVþïøOGVõZµ€o&À`aq|vÔýsyü·#šßHl²ß¾ìna,#²&[!gÑ{Õ=K³ù`º–A÷J(“7xaâjœpØ:cý¯‰E£´J íBwºT—º¯Òð6Êî“sì¦cÐÖQc…R®»A1§©û SàzAKŽë@ÙÃR@¸n©qá¿~éª6+2XiÛøN"øfŠÑAÈ#y1%Aʱ{¾Äy¥î­‰jqüÉÑñ/¿$VsÓ7…¦á]S=L£8\÷¢Ìž>)ä…ü´©ßMÓ ;¯¦6DkyËäÕúUS—M!]ò]è‹B¾Wœ4G`ªÜšmÝœø=®qâWÍØ’™RÏš¬_ü®×…¼-dÒ@àWÁôF:Y zfëB¾á_[Áô͌站F²onøÛ¦SyVfûº)Õ”ï¶)&Ô«æ`ºÚ£ä~}/Ñý®”ïEös²×M¸ã ÆÜ4Û~Ç¥8¶½jv3•-ñú›Bв¶pf𬳼-veD/0n¯Æ¨3 Z»B^6cÓ­,‹Í¦¾j¹’ÖK`$떾º”“0IîFl«²ZQ¾ö…dm}!Ýhø\š:´4ËÛŽdh.M=ni–·e N`EU¡´Þ„"º÷—ïï!Û;í¥LhæKH˜X6eÕÌÿ3KþšË¢É‡±½×"fÔô~ 4IÙk«ãÀʪ)TwHê¡if¦¬€ƒ‘³fxyÑ4£ìV«’qd[¤õq!™•=/È¦ÈæÊdsé“£i}P¾~\È ¹ÇÃÔˆBÓàd“ŒÜ"³Î…\€£µ  ø7B^1 Øã¥·À¦°ä ¬PÁ›îõÉÒD¬}wñúörÙ„Ü€Ý Ã@¿`Û&v?_bbáµUbA?D)‰äu¨°ï«½µKïCeÉ<dE{ÊŒjÿŸY«‘KeÓu¥Õňá0I9å³”fÒ^‹äõ<¨MÎcí-¡öÙ=‚¡°•h¬Ã”ÎÞž¼¡¢D£D“¼ÌQðݦäSs%j«£º ³S W§€Û%d¤™Ú¹Ù‘ ¶º;r\œÇ Ÿ…M ü ÿÈu“´nGÄ>‰Q¤-9ÔÈþsü$çk¶Êój•Éâ}•,þ2KJºû…˜)¢É;±Y`φáaƒà ˆTiÎi6óį¬&—áåG×É+]ž@&–Þ:‰Óƒç êPN-ä4Ü´sxôW's‰ ÑàÓ ¬™b ,5Kq4©¼ÊôF<†é/È&@Ç ¬çe•dOUø 7E ó\"«X[2ÓˆŸ]¾•’-Ÿã¦ùœ ;âäLýGF˜—€Ð<¤iYœ s ¦8v¶¿AuÉèí€s`~<;¦ŽS¹f«ôó)$*5 ÅK“œ)îË×éô "ãÞQÆ*1Scv>¶WL3‹—·Kªð8Üû¯h¨œÙUv*ÐNq‡€À‘yÌØ|ªJ@p„þM,¯)æú–CÜA“ÕÅ·5I 27‡ ëÊË}Ãó4b‹ºv‡_…0w¸q*iI7äGTò&­”5ä ÍÆl‘ˆçfÎãÕKᫎ ÔfS¿*¾ Ç@ÓiÉZ öCZ/€M4SÒ•ïÔ³±#äˆU¦»˜,CwºHµ }©p©­™³ê™Ý¸kÏÃaÖXiAöj¼Âl+'%O/F9N4º®C&δ2zbøä©%%žÊ¡’-¡;€N «Ób §¬WòÂ)WaŠØÕN ‡Ã²:V¤1=\ᬮ¿R·©Wˆçuyk’›72ÚýªaÊ žÆì%\o ŒâÑ“ƒ.üjFQ%<È"ÉÙ;ØÞš=èâ°{%Ôá…J˜P`ëàiC}ÜÌKNU4[ê CRWȰ”ÐOvoJ¸à®h]íÉ‘ü®æŸ•Ö%ÍáëG|òr7u~û{ì‰bñ´ÔµÀB¨—ŠbÍ8… r´—.Àfׂåѳ¡Œ¨ïH´Î×D vÞ‘‰Aƹ#«‹\?j¢b*¦<`´ü|¾åm™>L9-ä×Í›BžzÍÅ®ñ_ðn#ÉžZµ_s±lÜáUÐPùJ7ÑÛ*:~€|?m6x[dö¿åëªù ˆi“½ÅzøkÐû‰„¿ÈcŠ;øˆî¬=Dû¡%ãb}hàöWf2?ÁÃÐ{ð›® Û ¯äyÛBÖOöòͼµù {È·®+U¿ï("$ ª¼ «^ñ±Ç'Ï ©›[÷©‘’8>)ä…ü´i®’–ľ¨œyDúèç—­ºWÍN›³µ_·ßˆÞó¥+ûb"/ ù^ipRÈÊö†S4UòÜ]ùý²Ùk]HÆ›lÿ½éj˜£>?`fÊž—3#coÊÛ ÚOØÛ!qË<ô­3IüÏM•ÜNUrǛ˃o‹¹Å~Èù\5gfÓ1+l?­¨·,òìÐëæ^a ®›¬L%15(fqù´rlÄ·ãi/7¸ ‚'ùTî±ÂDÑd£S%Z?/ä/‹TäðhT?¸×,F¼/Oâšuy¬,{E§ý³%ÈæIÛEêˆåë{—üÓ]%솧 “â4|5ôú¸TOê’çXçceL~f‘j¬Xs’S,¦L0i¹ÁÝqâà5]•Ÿ9 «•Ðh!N‹ÝOü(Ká}Aí°þ^ùò(KcAÔVgã›Ä¿Ø¿œ–d`<½’ºº’pÇÓ+šï¼¼ÂËr ¸zQ¼ôDÂ]÷ *óüЗWxV*¬°NX¼©ëªÞ6És¹Ú¤C›¹’*Sxuaÿ‚ép€ž´æ´¯n0mó'ì­#?9JÖOQ‡²G§†øHâÙãÃ`ÜH~~(–²âí„åý ‘ÿ—èïãöO ÙÌã×vºûaJl›K¾åSLÃ*^Cu¡V@ºÜo6ó`â…YN±1ƒ{×Äf¬ªpðÏćÿjËCRþŸ¹>ùμ#/û÷Þ™-öãáà¯þ+1SCùq BO¨ñ{hm5¥úÈzÚŠG…ßFAˆy]–^·ËD?RñÈî9ðGþ…¨Ÿ ÆÌtsÙÔÍ—wK7/;ÂH~ÞTÿA[:mÎÆºµ´7XÛ·¼AkÅ{SLé´så tAUðØ7M#fòŸjËìy°Ö<Ûl é`©3{kÊ2þrô/í“®ÿendstream endobj 235 0 obj << /Filter /FlateDecode /Length 4898 >> stream xœí\Ío\Ç‘rd|ÚÛÞF€¼Ñz^úûÃHbkí@›‡ÊF¤H3&94):–xÿõTuõë®~Ó¢Diƒ†lõtWWWýꣿ޷+1Ê•ÀùïÑÅÁ¯¾Ôzuzs V§ßÈôë*ÿ9ºXýîZX¬£ˆruxr@]åJ©0j'WÞú1j»:¼8x6ün½£Š17¼X‹ÑF«¢^BQ áEàÅ3lìcˆv¸„Z¼tnØ®7Z«1;üÔªàŒ Ãuj+„ ÿ{ø‡«F-‚ZA @,¬†ÿ[þõ o¤6Ùoφ[ e¤S6Â`ä,z¯†h4Ìðdô¨„ò0p#€Æ!ÎÆ g€­cÖÿ:±h”‚¶P’B»0­•é‡ák"o£ž¾Àn:m]j,£PÊ 7(¦à´1m¿iœ/hÉq(#G˜ èç--NüW_J!WRŽÑZ…6„°Ñv Α ×ÞŽB ›Ø¶Bo†WWkGÊ ÛËÝÍîâź;$ÌÎX½:|zpøI¡ …Pѳ¨<U)Á¨-)mIë2€p·Øõk¸xo¦Ž!kÊk°ÊBØhHÒ;Ÿ€mÔ-†™}zxð§B»‰A¿º^†x;× âÀïdXy!Ç  A<ÃDÚawž´€½á¼jÿ‚ÙÀ6;: 󺜸FAc'€LŒžn/ÖÊÌ¥ž¯7P´Áf;Êh›¬ìŠ!Fg=›í›ÎP 1 ¡ÛþeèÒ~ˆ$µ0£Ò†3€ä"÷mY[vÎúºº—q½ñ€wPÏð9N¨8 $ï"7ÙŒ?Âp27oY›awAM(jKEP1¨¯ø•Ý ³ðí)Vˇ ƒDr.Rk‰®íyß+нx€šw –ªUÓ$šFf¸?Û&€ìòP’-qr=Ê8ëQÏ0ð$Ò<š=çßM=Ù¦ÙÃÔÊh¬ÖàOÀ€“«ÜYLý²M³Wµ¶¸ÜÝ?:5ìHüFûdĘÚ@<@-€¬ |B”(<÷´F§dƒ!1ó8(4-lk±7D¤Ò7ÁóJ!ëFqÃéY‘;WÇ)QRrƒvX a3¶n2Ñaõó½@:wŒ2èY.F*ƒàxQê DàÃP¦D·l¼ó:^òÐ"x„š xòTE4²TÂD}EßyŠ"ÚBhQžÕ³F¢‰êa{Ó˜r±}Ñ8nª4x²C‰ ßDp €°døbŸDÅ­Ý7&01ÜšÀö†QÇF˜Pí!¶Ú‰­h(Té`G„mcMlLMb÷ŒZ`ÅÉ=~sDœÃˆÀþl»*Ͱ ·”ƒeO†° à^e“ºíj-óz̤~Ž"ÑhÙ¨†l[IÕz¼KŒu³‘I½«Â| Ú2³?ÂlDIËf¹ÏDžuÐÖī%!‡Ä¥²Ðþ™Ï ±XläÆôo³h°1=Y@z =ŸYhnÎXÞ‘±Z°~Ý`‰…œkš¸ËÙ7±úuÕv–.m&F ùQ1~¡³ üb­­/#9#muâ3d‚.0 JšHÜŠéÐÐÐ ¹=ìçEÇj©`6G¬#I<™ÑsW†T Ë܃A¥{+ÖÔQSžZ4µ0Îzc•\€ímE§^ËAôB¯•’ël•_Wc-l4fô}µ`æòì¯a“G˜¹Y í¹4À±§±ƒ”¼š÷;Z§°m€ã„{Gåå„{òf˜^1¾MZ€àB–Š06‹[¥EDc[É]˜8|‡%HHä”c¼ù¥.£o¼Þ1QB®JE-½¬KèÙ¼šeìdó•óÝž%å©'K²|ù™0ÎÄq]pè#Þ!óR/Û‰1äÖ¸Éd´³&³œÕÜdv7ÕCp“ÙÑ䡃䴓Á#˜Rm©£®vþ¢_ ÀÝßoêúm1 ›ÜrÇT—U¯tå²ÉypY0;ܽ¤i W¡Ù¹”<²Ù±9ï,ïX°³Ë*…„›‡Þ`DëÈJ›´‰Ø£–8L'C:ÖŽ)ó;ëçø”‹€‡‰´°qÉýnYD½ê% Ûš5s€³üùtcñ´¼P EhéŠS•&p· w£Z¡-p&¦ˆ ÿ,q‡Â ôc‹’ÇÆ]ƒ¤j…$`¨coÇjçqFI †)—êhš…¦6k‚~Ùòæ‹…-áßp¬ èø ’j3ÛÌú MÔê’¸Î$‚Y0–Ö®`ò o u£8îÞ¯yóÂ%@67 rýGÍ6Û€¬lYðL¥ ¿*­´Xƒ&Já šøYœh³ó´±ì/o¬¸”©°ÁÛ àg~Û$iž¸ÏVtœÉ­X¥ì^zäbZÈ÷–Ã<*ìï†%ôçÕŸNÔ8Xý¡K™ÊrZ?¦P°¥-F˜'JØà£Ù1÷÷‚x3"odeßq\{rŸM؃6’wäê(ª)Ò‚]Ó^,.5wHr6En' ™ ­\¸Y†#?—eb¸7¶IÊ06Ç`‚°Ä´g‚KQ-m. ü>LÀŸoíQt[Õ îçsF;5µE×\q½´< à­Û&r^Δ›1·¾´Æ­/ôžÖ&ùLÑ"¡ž¹’Îú%é£E2JÅYòi;n¶£ÜX?n¡ëØÝR†ñ›#ƒv}•:¦Ú⇉#; ,l0M/eøÅ$›­’™£¦Í~7‹n©ØxR£5ãóô¶&}½«MXÀG°+Pˆ*mÁÚ‡ì¿ …ŒÜ3Tfž6í'¦sxŸ#üt±´ˆ1éÇìÖ­h­»j1;NhBêÀT0_±Îµ+bÚR‹,4Ÿ®Pp!EíÁE§À÷Ú}É„_„€¾²zß‘•‘Xµ²hxÝK(ÏE,ÿH)!ü + ~úPÐç.…‘&J ·¡õÖïñTÂ{Ê·|§Ö€A>΃O%<ÈÇÏY_g%æ{„¸À·ëê¶M¢ ´ž—Ý^¶ÃÿqÍ×éôðÁØÛH ¶õj£SžC¦øYÍCzG~E‰Q\Ã' >p{Ü ø0œ$Œîµ-‹™¿6þƼ1(‚TÙ:ûYMúÜzëEi2W“¡8<ÞNXù1Bv‘N6Á}8 AÑiX¶§î_MI…È]\s¾±0{È–úZäUþŸç,6ÁH‰ðjCjXègþ,¤XœPÓ¬Ó¦œ÷k5³‘â%ÀÎꘊÿ“Ή¦FÈ-Ìø^`ª•µI‰Tñ>ü쟭NÞlŒMPK¨i'>CÍÜ©;J0ÆøfxãîiŸ-ed±¥ÿž¿`ƒÇ²B5Z«€5ÉÛÄ倚¤¢»Y °˜Ú“º£R²«"ƒÂMØ“îiÁ'5 "ƒ­‘‚ €QÞO &ïDÏÌ险Å)µ˜»À[(“ÈA²‘ãôÒ¢¿œEÕ²ŠÇ"ªŒòA‰ˆÌAîÆ¬À´QZ÷A¬ ŒöˆÖç1é#H½÷œjÆ7Ü5si‚7‚ïì5> ©yXä¨ÕÆ@–}èç´Zÿ‚«@@ŽÒyÅHŽ• 9‹œ’) š¶\ƒÂ“j’›_%‡‘1N¦7aL Š4elvlQ©% 6öªúqð±eàœÍ`ÇX=M›q°81YOy:Ÿ,´Çä2‚à º™=9’V#ÚâFͪa‰;¥²Òûm¦e^{Ó <žå›»”ÿáöÜÂ~Ð9õ›"³D‡íqÕut--dÓÙ󱕓 Æ,lí¥BàB¸sUÎ`š ÚSc½ö%J=:Ýð›õÆB„ÉG`-àÒãZ|Ù­å–ò‡ÝÆ×µøª¯ºÔ.kqW‹7ÝÚ‹Z|Ñåá£iF!fñú˶jñ›nƒÓZ|Q‰=®Åkƒ“Z<çÝJñûZ¼¨Å³nFÐŒ ͧ*t­¯¶Ã7ùµxV‹ßOÄÜðëZ»©µL/»âgC<©ÅónÛ7H†Ù\mÏ»ý.kñ¸Oâ“®†¶Ý~ F²ÖnÞŽú¦‰Ú]Âl¸oº˜Sݶ4ç óEÎ\œö›Ó¦âä實^¶’µƒ®½>Êç¹[»qö_à E„•€ºŸ(Èy\‹ÿo¾a®—„%3$-¼Á®«(Öí¼«†'µÁÓZüs-~Q‹_v] sŒæG^o ¬ñ¶Kí¬;Ü«×9³=»J¸`ÛÕâE-²¶[Bžƒˆ«Í„¼óúû‡ûpËmÛS·“ŸÕó-Ç]äýX‹Ì%1ÜœöÛ'w@oŽ–t³_ÎIÕìR3.]âäÙUuç×â£ÊwSìEúåøÀLå¶FY‘Ù]ßF¹x®»*è«ëeŸÄ›Ç®»,äžb‡|?‹ýU`¿¯Å]-^t9kQ¦að€(ºöħ]gõ²+¾šP[ÔÿUäË® «iZk¯»Å¯-Ý~¨Åﺞ†ù”[n ÅÏþ±;wa¿©Å¿×âÝÚ¬ÌM#f¶éiî´ïkÒs$§]5'¬r¤NÕªM¡#¤õD¿-WV)²®JÒtÛÆZ4Ý¢®*ô] n¡[Ï·±¤GÔ¢¯C<ªµMqÊ`SA&—•‘óy-^Öb?Ãdîá¨Û€Ù'w6ë*Ê 7[¦¢Rt]uj®ú¢NÝU°å Ф6]±÷5Ë(„.±¾fÅB·w®YTç¦ê­)N¿U+ϺÆÉÒ;æO¸&ŠëcŠgÆËÒQfÒWˆ¤m‚§°Œ9_~I[kmä°i¿¼$Ó*ù¦Ø'ïG‰îêòRŽw®3yñ:Ͳ¢ìªÍ¨bxã)1ýîÏCŽF«;iÒ<ä½ç±À¼Âw ™w~jò(? k©à¥B?íé¾~íüJwoiõ‡0\ô^’ªJÏãr×_’aA(v“%ë@Ïw¤é!‡p«tP§üð4Jí§QeåùS"õž©IÐï7‘Ê«¹·N¤JÑÖ¬ìÍr*fÏÿ>éÕ?)§šŒˆåTL‚kë5šeI—©šeB“]‡®ÐäÛeZªÛ6Tb¡K¡ V|—9tÙ¾+åš%ÛÿVÙW‰ö?e_?e_â÷_3ùÚ¿8Q’1“®È¾öÊ¡Á›¥±yXNý¥‘zzðmö[ùÞÄ1õsÒñÛ°³^ù!Èìþ6ö×Ïüî,c#_K_h®ä—ÔØ†æ éåÄ‘ÿgß1U6âÍå•À|ˆ¾eª™Ñ{Ö¼I¥götsñù^ó*b¿×Iª‹·Æ5~´„:;×ÞÿØÉüd;+±ÐtðvÝ=\/øqù­;‘ØR1mÚï¿»˜®”÷n7oAÊ¡yJ”n·æ!ÙÍéþ»ºæn{\‘Rm|Òi›j÷î(›/ì=í_4(·$ÓÛ•éRj"AÏôgìÒ§_„åv¾›î’›ó§ÒðÇrd¤/h4‚(öOjwZè½ËâéV…hŸÂì.èöGz¡yµž^Ëð÷Êý7ÈéëZƒ²+òåe\ƒ=ß^'ÚQL³Å»9:™ 7c ×ð…È UãÕ¸?®5$%Qû¦ÁYš‹ÒÓG7‚V’.Á‚w)Ìì]5rW…DfC)mSº !LN;K„¼³Þ/Bséš8ëxÉšçHÝ~Kæ$‚)Iô“5ðç@LÛçô%!ƒ/gÎÔxuÏÐO¶·ç/oö%!­è;Ì4æ´Ý{Eï÷ðsøÔ!Úav¥‰^6óK€éÉQì¿ø=¥ñ~ƒ`áMÔïd騆ìSÀš¼‘@PÐ}¼Z®Õ.:¦4ýåÍ5Èwü˜ôžÀ‚á@:gó'Ñ~ËæU/¬}Ãâíkyzc>B€xkZ>ÞÃû 0Ô1ªf¾øR÷€|ÏJp£fÎúºÈml/÷m D4lßæõ]ûðD”mö&¹ÍJHnâ!ñècyótÿ´ßZï|-L±Qñ Õj° …7MÜÈŽf,ÿZ7ï–^PÓ0Nè÷‹y¥ñý£\Eücs‚þźŒ':O·é#¯`ð‡è2ó-ÕÞWÒ‚K—A«“âóÝBå yÛ§<Ö>¶3{¯ðûU‡âA€5ç|á(É¿NR¾ª‘¿Yõ>e¨&ëM6½ªn?¤„}Cœe±é¥¯Žü:ï¶ÿU¬løÉMÇi¢:øQ]Þendstream endobj 236 0 obj << /Filter /FlateDecode /Length 3335 >> stream xœí\[o\·~ßæG,‚¢9›jy¿q’ØE p) ôa­ËÊ´+kå:6 ÷¯wx9‡ÃÝÙ[­Ä­#äÁŠçÆ!ùíy9d-²ð_þ÷ôzðèHÊát1`Ãéàå€Ç¿ó?§×Ã¯Ž¡‡-­gž/i( áZiøÐjÛz©‡Ç׃“æ«Ñ˜µÂ{çLs>b­öZxÓÜɳÌaòEèl½óº™A«t–ÓLFc)Eënþ ­Â¥]sû2¦Üßÿ:Т•̉!´)ÌÜðølÐü{tüAœ_qb¢¿4¯€—âFh“ƒdÞZѤ٬SÍÓ‘“­`ÂÂd Y„a5†b¡ñ·QD%ôŠ3i\s:&,Ý5—‰½ö¼9<äwR›Ø™{&„iAMÎH¥êqÝa½`%ƒm oa)`‡°nnÙªÏXÆR·Ú£ ®`&ïR¾™rZÈ›Hƒš}sXÈóBÎ 9-ä]!/ ùãh4æLÙ–ñ†•æ¶¶ž$y!åh,˜×-™™v®cIQHYHUH_H^$“$34ŒáaˆÙX Ù2pcZ¶4$éñ‚ú)>/­|ƒËV+/†Ç‡ƒãÏO:F{þ¥³B"k!ËÞò”ìpCÚ;9Iiš1¶eodN±Í l±7'¬×Ø…ö‚^•cÒchá)&w7½$%£EG^`±éƒ‘ƒeÇÅ„ÙýýûÒø‚ o¼/ yMÕÛùÀ¼g¤g, ù'ò)sq£ªíƒ‰Öy•S×gTæÒ­‘ÆŠÔå${|d_‘4{Ûò ØsÞJ-}Ç~Ÿðåô¤RÁžã¶-Ér^/ ñê:x«¤ØÈ3­ƒï¼Ž5ÂÃ6®ÝF…Uöc—%.ºµÜrŸòêû‘·ýEå]§Tpr£M7ö³X0È4 ž ú[f»ý<Ö1ÂÈ33 '(¨§ß—pþšÜίɌ’°ˆТ‡êmÜoû2¨ïȼ&š}ÏK6ù¾(ì®gX£}Â{RZoIrŽ[ûao ùÏÒaRÈ«B¾*äyIXÏÈÙò~é/ ù¯B¾%[ac z׊îÌøPŠ•úxJ±ÿú •WÈȨ*3d_†[w¯¿.¤&}%W=! Ò:AIUÏÉ}åF”¿%;üLnx7$_¤Ê³m¸)kCEáy!Qû¢§…D/®%/ ‰ŠÔçät4ãoÉ?“朑²¿&—1!eŸ^°Ã0*Z/ð2ƒÃH¡['u¨i”ãÂæcF$Ï ™5éC²Ç¬ÓBÞò²[…. Ò˜ÙŒäû 1ªhê¢vÜ-¬JÇñ–œ1ám¬Ò¥ƒÒ…³ˆÔ¾UËWãp”.É΂Öâ©%\%{eu¾¬\¨èÚÃÁSV=‚+.<”rwi ‘8¿H´Î÷ó™õeìl¥Mò˜ðÍ'¥±¿_„èÌ«Ž­s.ËïÆç×è5àêIÓ&L3ð@Bq’®Õ5 —ß ¦LÜ)ñ¥ú"I41õwùó~j‰oÒѸ)š-5 ýIáqŠ.æ'IÝÞ⇊«b7I6ëeõ ïótqÍïVž@ú‡‚¨7ü 'Sü®7Œ‰Á‘ßBºààmW¨D­†®ø¿E¼ÐsKš Tâ fL|Ê|½bª™ß¦f(&â{KîQ)8;ƹ‹ ºâá‘d½Úzú´ª§ãi8/eõ0 g Ç\êNŤ‚c<(‡¹p+ؼ  “3àÆ`!»@gpdži¨Ÿ•”ÁdÍ#'àÑ™¡¯Šº^ãíTLáæ¤Aa|P`|gžΪ™\E)ŒW)<²æIfpÐ_qš»ÔìÀ£ƒdžžñ†3âå|³!s`¿:ë#v`CKî}R®ŒÞý:ô€6°zÎQaÝx’ZI PØ\V“ÄÒãÐÌîŸÏÄyÔš)j¯JaáÃs]7”fýr 28ÑLÆ‘SÒˆœ£{ÇgFãP݃žrJuæÖ*¤^E(\°Ðw©xFr4§lòØa¸Ú)fs¬¤õó94,2ç²(È AZÞ©„]òèÐ!dLìvéÉRKé—s8¬°íã:zü<¸å:¤í›è>ÚÁ¸yÌL3[ÅG¶'¯6”‹r‰JûÞŠO¡½n‘%¿Ñž½ˆÎjÕë|æ,«%Ÿ‡³Pçò”/óæ{ZY0Ho„øÙgð¤W³ä]Iv8êãì™Ó˜õx7˜uBêf’*.³'ÆÙ&HÌl–9/Žä¢5¬­Nºë'Rƒ¹~ÿˆq̶g´Ž¯òt\VY¬ÛÁ™¬zcÅEu€òõûkÔ-¤icª]zÞi„»”z2„%š–^ÇÇÝ ëËì ÊÑÓW(RJ´¡ÔXç½ãPÂ1k uJT°ÿ;Ý'úlGߟWGª`Ä7'!vrh¤¾KgQÅIÈ"Ä—ßtýÎ]bÛY ÉpEª1».'úpƒÏRO ‰Îè–Ýý£³É¢œcÞ‘wŠ{t¨óÑûŸòh[×vПÃÿX.rOI¼ÚÃio‹5/’—ï(w`¶äQ«Ï;êèœvëUyM.™¾–½"•ö´t@E#î¨SÌw)\Ð5òŽ‘C«’>ähô•Ч¥º6™“Ì$‰ú¢‹O±-¨Ð¢×Y]ÒPò¢ ¿"e@{^8Ä}N«\öç¶êê¼C8Èu_W’ïý×å†È‰^lc†o*äãåýbÅÍ8¶ ÕZÇec:- O¥S"òêÇdÚ'×\Ú½+$#—ônëê6ЛoN‘IPƾ&I´Å£$û†¶$}3Š6käkœaÿêbïŒ,6ªmå"?Þ¶©îEîýQ|5¡0;f¢³™@Ï騃#[+r̵勱|;NÃh$EõÀÝ› rôdô²þE!i¨m´ôë>zþߊd-ÏèxLJgö‡gö߯3ûÚñaçþMïÜÄf»ÏÎfÓvÿÀHÚÞÇŠÐŽôÏQnÝboÚÈH«–le% É·‚"n12šu¤ ¨u«½÷À¼nÃFê¹RßìS²¹j‡‡’íc.ÙŠ¸ëK¶{wSŦÜ~%[¿¨÷†Fþ*ÏQÔNðÛF¢=è€úñèÖKæ×d¢¥+â yß¼Ã0ÊnÈanI=]“¶Àú=${ w†)i¢Ë²¢·ƒC’­$ï+—Ð"&™a=>PÙ€0¨à;“àND‘²6¡±<ã,1¬ÂÖ L%Fù¡¾k`‡cÎá¢Ãû»4ÿ„ÜÂh‘¦ 0,„s X–È-ÿо[ZúA¼*X!­ÃEDšZ׈ mz(nцèÐ,ÂI±×O³È¬!D.ÐŒÝ ¢îyJ€E<1À4SBâ5\#øÑ.€Œà`§Ÿ`°R‚D)Æ¡ré³úôEoŽˆB—âK+Ìš¬–³%xEǵ>Ô=õ>š`q ha D<ǘ&ä:ç“ãÁwƒô) ã”Þ®ÿþD*Ý÷'8×Î8äýéã÷' š†})óÞ›Ÿ`­Só[`@œÁÄ­×Z¤-χ1*^[@-”!Ïòb­‚X¼;ŸÝÑ1Îåö ï@Á7k¢~-ˆ¥ƒúž•ÀÎø+\ð ú);W‰øÉì++ð˜ Þä+÷`Wpô–Yв€ê&›á‡ðiî!à'Í; ï0P'9Û| eçI }Qã§è£Ð]Å-5GÎRåS{§0­ãªËª1Mœ%0rPƳØÚ9‹m4âd„V…´¸¯æ‚.´Òµê` X4(Á´½' Iã[k ž§Õ»÷>éy0Ӓ裊Ʋ9¯,ÇæîÛ3™‰CwÕX–Û!ÜæÏ‹e1â-í[lˆsmÞïtM)½/¡Ùz¿Ê ËÀ=½ïó(ÛR:8rñ†8ïáý,/#p< ‹±qÓoû7%âx=IêxüSÉ]ñüae™À OŒ=ÑIã(4çÚà1¾fð›$Q¨vv€"l9ÚýèOMóZa\ ÆÐ¬}=a,F2v°†Æî²Û¸úD&L(8†L#9®Šò®Ê°£§ÔÔÏ‘›&A=qÙ6GL6õØÝ™ª¾HXF楊eLU²à*i¢%د@B`ЪFˆër*æ²SÚÍq>ŒžÑä«UžÖŽÂZg– ²²)|7ø*Žõendstream endobj 237 0 obj << /Filter /FlateDecode /Length 5442 >> stream xœÝ\Ío\Ç‘öÈøº‡½ ö¿Ùõ¼ô÷G² [ñÚ P$1 ìaDJ4’ÃpH[Ò!ù×SUÝïuõ{=3¤)Ø…j û£ººêW]ýþº½\ü—ÿ?¹<úÕ ­gÛ#±8;ú둤¿.ò'—‹/¡‡Å_ú(¢\¿=JCåB©Ðk'Þú>j»8¾Ûý7üª‚36t7ÔWþ÷øެêµj¿™“…ÅñéQ÷÷åñŸh}# Éþöª»ƒ¹ŒtÊFXl…”EïU÷EZÍÓ}½ ºWByX ¨@ £wã„3@Ö)C$¥ /´¤Ð.t'Kåpë¡û!Mo£ìž½Áa:mu–Q(åº-²)8mL=nX÷ §äø(#{Ø œî[FÜø¯^¸ªÏÊVÚö~'|+Å茉ݦ4/KóššÀåØõ¥©—+«t/„ëD³ƒ*MÖA.WÒzÙ MÖw¥œ p¼¡_m± Ã!Ï@¿ö¥©JÓ–¦X&æHo*îø^D8ögGÝçËÏgŒ±½Ó^ÊļWÝE“+geÓÏ õç¥ùΊ&Æö^‹˜éóÖIÙk«ã@ÊŠsolÆÒôMþ»fÛ<Á?u…}§¥ù¶4Ÿ”¦ä¬›ZÒ~WÔFɸXIÝ[UÚÃï ¿-M&z¿.Ęæ~ms“ºÙAð_Gyû²üúmi~Uš3hcº4}S C³¯åͦ¶jPVòxü¯ºc‚H!Tô æ{À»ø²2€?Tà3‚*¯=¢å çÞ'Ñk+3´g5‚޳°úKkaf zAø ÝC+1©eŒ;4NP¿Å”Pjsƒ¿«ÞÁ|iÁˆSÿPM’FF¬ñ×Í%ƒþkÔr˜àms•fWBf ÅÖ"‘Jãq½(4’4ä0üÌq󤨆·Ÿ1êN uˆl.Â4{…¨ •¬Ð›ñã¢L7²ÆvkFýi¢°š±{p†Íq2r ñK £Êæ@g2Óƒñ¸6°6Ž'r¥]h´ÈÇ-M·&ëèbP&Û!øU‚¼FI*hÅ÷7Ššï8b9M dÈÆ’sÿÛAHÈ ¯‚„li×d£QO|¥£%Žœ™üq §¦v<Üh<:ÔÙIË—~5ÀµšÖµy›AêîŒL»°‚n6-Rúûã£?%ÏÇÆ/nv»;µòîŽô¾7N/Ãz8(tw%ÓóÜž…^z>ß+Àš•Óàq€ä톉ÄPMÖ:;Ü Pð*uØÙ:KÓ‰‰ö½ÇqÁ™\¢ ºzó:i$ÐÝm.Ò™x bóv8µX«— äÍë¢ Éï 6¨JÆÒ*© Ç3¾)¶“Í}´ý(ZÆW)_q“¯Æ}Sê±u6ÜK<%Ž c_¦¨ê¦°±ü8Ad˜q=/J”O"@(¦rwœŹÄ@Œ4Q IÒTÛâìe@_´I""ËUv…ɾlYÄc`±ÑÜÀ0&­4\Õ«2ÇÙà°“ŽgC:.£ó|2Tf $‚,¼Æ¿ÃŽ%HÛ¨ÖÉ|ÖNRDUN °wRÌøEiÞ•æ¶4oKóM±Ò7¥¹en óŒ¾.ÍóÒ¼m6ÙÄÌ¡z^šëæ°ìEÍàA8ïa¯ýÈN:e…§ ýWŒ§õ©Ï HøHhŽ"+ Ù3n¤¼—1 &dTs¡‘Èšn‡•f†÷4Mgk‡c“~`Å”¬—´¨áØ‚ážTK 53Õ ï¤òfª8ØÛÆu~±ígín”v"Nzëdbz*9O lðw÷xÄ;ŽÙgk{Œe;fAÇÆ‹h´ä`øa9F¨;vüNàôËäQ  Æ1þxš¨ƒMò¨ôš|',GÍ:FBm¶ÇS;ã‹·ˆ îÕF{ŠÉK°×üç¬,+KÀ£òAƒWƒï“º[Ud}ðàãÝ?«1tª‰®÷Ɖ‘7 ¯H 1ÖçÉÃÂ÷.ynèºñÍð›¶-„Œ¦µ°ÞhÞódâbÌÈìbY¾ÏŠ”É>•‘aç6+ºq—^eºcØA8 o‚Óy'½Z¬à)”¦Qß’Òžk€àP©À€ž†A©À˽I’­5ÉÆl`èó&wD¹¨HÓÐÐÖ†Á¢„``ÒƒC˜¨ù „ƒÎƒjøÄß4¶#1d„yV L¯Û!+ÀjùžÙq’ãuu¨=šC{´üÌØ°ÿ·å{г¬¥Œÿ¡9Ì”pÏë`Í«Qû$…z¬›Ûq;ž{l¡Ó“v¡^--)Œ*ÔƒVo€üªPoÌiïHNŸ¦qXﲫP¯y ¿Mãtг[øÙ‹I>Ï1Ðɯ±7ty,užÝìò{Ž:INE´L§ûÝýêôêZç\–FÙõ“LÞC¯™Æ*–Ì,©Y3êG6ë=i}Zø“”èYßãd«Dwi¬ €R…^ºû©+ôsRŸ‡gäñj‚Õ%$9SV'º2‹¤ \tÎYÝ6]OÙ>©ýÅ#0+¶¦t«"+áç7)êrÏ‚Á|Õ4)¤[ C>l=Ö! ’¸°ûÆÉT3¨,¬©ýBKi]3¨@šüÄæ›Õ£>UЉ“‡AIŠªÃ_¸aIN³†²:VÜË»$uAðTÓeZÊÁp“cé\²°ƒŒáI¾îâÚê¢ÄžjÃ4Gu‰ÌoŸIêp, Œ¡Ò?ø¶Ç%~òÛnüøAÅ’³š™êL¹§N=¼ÕTAÇÂŽÝá-)ö°QÞÿÏÝÒO9 QêËfº½¬.›“ØÔ?¶ˆ™Õfùˆ°.=&»Wi{]T˜™ê–d¨kŠr®4m¨vƒ¾Ù ´€ýR°o` ¨*y¡«‰Ì:‚5W±ÜxG:¦5ÝôfÜÐ_ýü vLoа.‰¹VJjÏ| )ãwýkØF “Ô+˜ZNê¨ |}ZL–AÌ$ÛæÑ}°|†õÙ0NÊ]3¯oS—°×ûˆÊÉÅ‘Aôœ'ÓÛôtm” õc‚aá{?M`‰cm[™oêm6ï=A—ë¤2þ¼  J¬Ê˺$ä&xT¹[¬¨žªv~9µg$6»•lâk‘ÄÞd Ñ^;¯‹Ô®yÑÇÖNŽ Ì pFWS ŸhW ‹ñOÓ¯m{œ –ÙTk=fW±L]è’JGÁ)•Êé­öÅohÄ@C”ºQ2‹O jÔ_ŸàH'ÁÝþ¡P¿æÇŠò÷V€‘6ðƒXÏÞ±ä½Ô–ÅcW3E–OõUJ_©xÍ¥ô¤Ô¬ÿ/™'ý4•„!8ÍÝ)0,râÛâê*Ò-Þtõ(…Ÿ–àSVfZêEå_¡ŽÇˆü\ys.kƒ?Õ+FªW|JÁ$ÉŸÖŽõŠžÛ äÙi)Õøz£õävÃlœç¸XYŒ Ù׬ ꪔ1þ¥cV3¸+ÉJN}àHW&ŽA™?™LW‡+ašàíOÂþŒˆ‚¬r ãäÙ•Š‘8^¾]C]i3B²{”H®¯P˜™ðÄ[ïyw]¤°ëY¹gðÆUÒùzæ’“Ts@y3š*.—äeaîԉъª(÷”¹î«Rtól bÑîW“¸q¬uŸ&càg/ý‡'ÕîÈ:°×Pl†÷k–F¹j ÛÓƒkwëH Yýcá Œ§Ó«§«c³ °£ðòQQû<(øHa;½§Õv¡Œé-˜…džíàlõ1òé^ÍË-¥ )ûh±ÈòmªË…AFª8Äñ/ßœ¸»YZü´e›YL8 ãõ}Ò˜ñU§zh·N]ða×ĪÀ¯hUÉ,ÙqÔﲚô'GÍSSåq&X%ÿGz‹ÿˆßs}攞AÄdbÉs÷œéøÙDp7*ýÛñ0±²:Éš:X¶Ës{°]_1«’ ±/<0³ôº¨å.ÿmìÀ–˜¼M†¡·ÉÏÚ ñP–j‡‡j–N_ž|„£ÓÎô ‘|øæ¡oÈ ç¦¤/‹"“&çLeV³”£ŸŸ€ži&.ë=0§™EÈP9Z,YÝ–~4òë‹DTŠ9lñ*…œ£$çe6äÂÒöë@ï§00Æyi­èýçJ9¦Ú_ÀFØÊ®Î³599Èã äíè5+üÚ€l#®zbÀM!ùnèæð`…¿1-&ö® Û¼'°;gk¯Uê¼*nÖÍ© ?ŸsxYˆšvàñ}ÂUZ $ŒŒ<µe]oW¦c¹Û/êHlW>»žà+&Kz=Ïg¹.-c<Û /¯s%-ø¯rôh“‡>[õâàd›úY53\rÓƒEzè7Ç=úêOýL%EÏÞ¢g:eÐ]4ü°väkv¸ú³üaæ@%î6¯;ôJC¥4¦—"égN#h˜sý>õÆTm~b©Œ²i™ÞdG"¶É{úæs§mZao°ãê29:ùŽiÞ¥ý¤µþÎä°4æmnƒÏ'X³c}bYí!ðz_l'£’RAІPcšwIÌ­´k]½Ža7o;Ò€¹+&| ~¡A ˆJ~êôä+P¬W¨>æOušÅ4‘Òbì¾”¡-é‰U Ê'Ú0¿Å<ýòfú}ñ«§¯À25÷úžztT²ÜÊŽò',ÌZ?MZÕŸ *½UfÉשý‚”䈾@!ï÷â¬inÓ$!N°ñ2¬Jo)¦t›î¤©]ó¼—€]·;ìKÓ[¸²ólGúFGû)$7ƒô™ƒ %¥Úê~Ú•?v1 3<’â>0È~üFMîi¤m;2UÊ£6O“T Ú!O9ž—Ù?ãå}ÆíÏ/‹Ò>jØKåtHã›C{>9DÃÏ¡Œç¾ÉÅóDÛb)w<µü´áÊ ó_(U…W%çO`{–CRSÕRiÔåŠTÀŸxY)…jÁMdM ¹(1U?Ú¯G3ÊêÞYWoà3 ‹,‚çë<6˜QÚŠ)éËtòû“Ð> stream xœíÛn\ERÚG‹×}Ø·BÊ™$sèû%KÂ% « A°«°Cì8†±'dlH¢Õî¯oUwŸÓÕÇ5›$+´ ¨´ûR÷KŸêùi&z9ø_ùÿ㓽÷j=;Úì‰ÙÑÞO{2ýuVþ÷ødöÑ>̰8ÒGålÿÉ^^*gJ…^;9óÖ÷QÛÙþÉÞ£î£ùBô*Æ\w8½VE×(…ð"Pð'û¢íNaT/ë–ó…ÖªÁvƒQœ±¡{žæ aÂ?÷ÿ¾gU¯EP336 ³ýƒ½î?óýöÒùF@“üíQw{é”pØ1‹Þ«îv>ÍÓÝŸÝ+¡<ØÀ…`ˆÔ8á  u@Ö?O(¥`.@RhºÇsåôÐ=ÍÛÛ(»‡¸LÇ ­K“eJ¹nƒl NÓ®Ž@zAJŽÊ@Ù) ¤[i$üý‡®™³02ô³Ð¶ð·Ä†Çpl+DìV<¯à¦‚g<¬ HÂ*Ý áº»|V'¬¯µÙOw°! ø]WáM+x\ÁWü‚ð‚l<¯ðí¤ÐÔQ‚Ð{òª.»[Áûuƒ ~Ë¢ö°‚Gtß<á9ñ¤‚Kvò1»ÛKaÛ½›& 6ðlŽŠ§µì­UYÑ>¤¿WpÂdîiÉ ·Ó 0äÞ7[HÝ[=)iÃ$ªEôéŒUN¢édߣ †’¯¦Z笢’eïUð3VøT9'pbŸíÁk†ÙBù^Äâëˆ1^f`SsÞ°s_QF_×ÞªVPTµ#äÅ=cY¾Î´Ñ ôìIUl¦õã4Ë3 ’À“ Lû ÊJŠ`'ø * .¤0Àj ;Œ£¢™ D´½PeBŸ€¶r@ÔѾ‚}â@2‘R~°·óWÓ®vÑ®X6 Ýz‰´ëf”!XPÚw¬)•#èæ9ðÉØF¾¢æöºó‚žíöRþjÅ2è¨Òÿ€Uù)WX4 ˜±à\ü•²×VÇ•EÝS³ìu¬(Ȩ©`¬ ­Äа@‰wYV®ÄA¤dœxâ{•qŸW¨ÞŠOŽg)S,9”Uß>ª£ŸWðã ^Àá¢ýì°µÈ*ï …A7Jèz©Ù`†+/>ððÔ’ôê„—¬Ã~É(2ƒD>‰&Ÿñ[ðÁ“wú;Ó'b¿L€šš2ì=V.¸°tü6÷Ü¡ ) ^Œ~öS6Žò…äI|:¬Þëkv1´¾Î%8Ø¥! 5Äœ›0Ë™3êêhø%Ø}‰óP,~š,LMOçrG4¡qT*¨ª¬ú ’Ñà .*(+èjö³Ž6àyØÁám «#H}Í'¬›#¾í}6& ‘¯uÊ1¬TzVe$Â5t€læX¹6££,؃{vÄBÛÝV p£b—~Jö³Ë]õƉí8Ͱ Ü⮘K…·ë¸ÃimìZTÝl@\‚ýuåÚpIB6{[áü·^~ðÙgYÙäiW=Ñ#¹k_ͦwvê|åd«yv"N¿Kô¤> ¿eÑðo·ÐØ©‰¼âý†¼F¡ÁaÙ#øBCï:b{54fƼ¹0…Q¹ÆñpþÊoÑI®¤p”””¿¯’bÚ’CJ ü:äbÞù›yjôñ!÷ü(Cm‚‚t;}”wMm;*BÑx›Œ×Ig-~sÇòÅKo ÁÔÉg¸à$M:B ÐAäÑ`éèÓ96Dy힇Œlú·>G¸Lþ8c)s¦\NæÚ¤2e¢2Ž¥KjèÞ*mR&Ó­Oó¹N4b±Ô©u˜5RÑ6©‚är5;©Vl·Öó|ˆ‡ì.D¦8à‰NN³ž7¶h­òº t÷ Ë¥ðWpN“mÔMwYá<:ìÈpêEƒÐ[ZJË (•ñfPMÒV¶ÉÜÐ!W&ç,…thÐǰ¡¸Úv8Ï6Ú¡¸”Q…_F€hw#Áû Ql¿p´;­t²)ãLÆ/õñmÁºËÅ4wÄiä’*úŒoÎ äø[XÔ–ts_ž·óÏd›¨D­Sìµ3R76q^m‚L=Ã#ÚF¤†ÂŒQ ¥éœ†ÉXTl°µÑ\Wya€$#‹+5@Žfµ>Á¿ØÏ¦®Hƒ»%7À´.“(Z•H ª[BHïå8º>MꨂVŒª€®ôƒem2Î:¨äÀ×I×Ç£d cá䳂³ðÝúI^æ ñø¹nF° Þ‚šög‚öLªËÑ€êD:ÉõJjÈëä*…T²!i@>™iÎáoNýme³œjË’uHï¼mv;É“UpDo׺ï+F‘ø’=“—çU ôlb_ßó+/XG¡©½jBΨKŒ€d–,zX«TcKâ»ó°ôÖÉÖ”ÆýŽp; Tµ¡¦±eœaD3ƒ’]‘–*"v¨Ç^…‰ó>$º~Švh’;ÎgKKÍ…““[&/ç84l…ÄuAvGU‹ë¦úƒSá_h™R½M\G§õ 5Þ©Zú˜,\&:tô”Âä6PŒ'#•ÖZ š—ÜU[ÇòÓÁ®“ˆù'…¥îáv Eh c¢ÿ}ÅüœHtÀW˜ ø–H“£•Ä&k­|pÍ[êèñ¥òöÛBx⃌ÖM;ºpÐ:p[ÏX£+A&O› ”ÔŠ‹B'eKñR8ªÜ$=B@&âe κ¨Qñjâszj5@_R¼ƒÆz«},óÙ˜å‘D±8fØ2’.÷¡EÒ›²shF×GÔÄHR÷”çôH¸¿Z²„GÚh'£-¹hgh.TÓ`6d ºÙŽâqZ0£ŸçÖ‚„tÊÈU)Á[žW]$ÊF¸ð„æo….[ÏÁX i¹¡ÄPs^'IKˆçqÈëœÀÒw 5[¢.5Ç,P™¢'âšüˆ ÿŠùÇÙ<}àðÄ€n]H?4Ïç_’4•„EÙí Ëh:J#þÁ`ˆH.TLчaíàM“ê žV.§y.ü«&ãèýtœP±I(¨óá³_$øÓý½¯öòkH?{¾ý‰K[ËO\¤é­Ÿ©1OÆ.]ÀÚ¯l}ííOºß£dú “Y€2¢šF‰ÄÈFJ%6fcEfçô0@9J§¼Ì{ЗIŠeJìH­E·#zF!Åm&¼,D©.±ò¤+à¼HPäß MŒ<ñ@˜ÁÈ¡Þ%!ÿ|Îú¨Qcà±Ö…uMå–S> *[2 R(e HUÐÈÆÆ“üT'P=¥>(ÊÒ1)C›¥ý’«¬¤òOkG®Ê¡\!å·Eeó!?GÄ­Ñ9M/©J¹U¼9©>i.Z²y¡†ê)4‰cÈ@­ÇPÁÞ#àÌréÙj¬ˆ”¶Vëé=¤6ÛªuVHD½ûqãrõ#äLÊ>âmî ·°bh]º7‡BB±Ô"XˆBu·Y®NÖ§‡ø®'U¦`@ËòWpÍËÍfùrø§ê–gË9ûN+] GíóÂE®õòB’;Þ¥9†&éTêkÂã-jL8D.ÓÆ©¶å•Έ2jü1äh/3>6&ËT‡õ¸I¢DcxÉà­r QiYHÒ)&/H$æÊþdó0HŸÄ úLî ¸| æ¦×pDöÑßàË@™ ’Ó‘ÓÜ5ËÞ7ÞJ¦kxpÀ:¹t#pƒ\âøªnØ7oeÑÞ ª-ßuULoŠQ>ôFizNjæ^#>‚úõÊLQŸ× €ß?"Tj`ríu{ÕU46l¹,[—,Ýã:yM̤u5GŸÃ€?M528#.¹;lœ=`¢¬Í½¶—›¥œ8<¼…@=\W½L±@;kðkNãéÀÁQ¶|  FãkYŸÏü¬:ÎaEt“" ÚKõùÙžf&·ñAŒÙ퓹í)¦ñÆ[——’Upl,Bÿ:Ý9Lwön¼×ÿSõj<µÞz1âú¨5ü_R Raæûq)ùuP+(¥g N§m¹êÿz(|DYâšwÓ ¥šÀ‡ŽºÇ°Düo§(6w“œÇlGj´ë´h“¥_–£Y¤!é2u7Ÿ ¶«LÖM4¼¦"R¡¤³º¿ä Vüo”˜»$òͤ¦B‡ÒSÀ]óe®vð¹è1)Ò½ƒ’3«]#‰FïÒeÀV½ké›hÇŸ§¶0!Œ¿ç¼6‘¨WÔû^D 2Ëè·jxû)Uˆ˜µ‚4ïÒı½œ@¦a M- BnìÙôNN£v’axPçyP:Èá’Bòv"3õ§Yt7[Þ­gùç+wVQ-Znh1"DC(®Œ¹ÃÆ‚0«ðPc©ÙÂ÷2}N«°» \¼’˜eA”òJ©üyXïJµ”]@¾xÄÙ•jµg •L™ó¬bA¶á(†´:@DH5E:ºâ/² ¡Gb?G”Y—Öß&nQËl3³fÍôHú è¸À§ °ËƒÌTâí)§‰?Jé+$š¦È¾0Ÿ™IPçº[¬KŸ}]€Š';£¸U;®ÊyеÍô¹Ëèš_”[¬lOÊÅZà\ûS"ÆñÐ~¸Ë©pýò7eîjwmc>}gêIÛW X4"ƒœGJÆ¿n$&|³¥=YïVpÃN&oõI¿éq!]2‡ì²»/éÙºWAÒãH:£Þ­ ù=ˆgu”tÂ>®àìòk¤‰—üJ9<Õ]Ñe#Hž]ò?ñð‚¥¢ôÏ6¸v¾Zý•Bù¢‚¤‹éEm]ú Ž.Øà»uI3îý)f¯Ã&újšü*ƿٷVDHÃ5ß²½d¥½a—ñ¿óAš³OY:W,HöݯའòZOhû±v¿ÝÍ—×3›1E!r·ëTò3äí)iÿ|F·AM7ËåR9­½»Ui€WB¾ûmUuŒü$Q!òß²jü°‚Gt_Î&ˆïV·7øÛ-!žjÍn¶aA^Ýxg¸“N¢¥,¾ÄÀ^TðVÕ‹UÕ ò[2G¤¯ ^Ôu·²Æ(bâX†OÒZú{9x¼&ùj￾º´bendstream endobj 239 0 obj << /Filter /FlateDecode /Length 10651 >> stream xœµ}ß“]ÉmÞ;­Êß0å—Üq‰W§wûÍvâ$.§’H¬Êƒä‡IQ+94¹›ÝÕ_o|¾s‡»´&¥Rí%·O4øð5ιÿvsœÃÍÿé_¿ñ«_§tóîó‹ãæÝ‹{ø¯7úŸ×ïoþþ4*IÎãáæÕ^ÈWÃM7­´óHåæÕû§˜o_ý‰”˱)ã|´J_xõæÅoO÷ñö8§ÑS©§·/s£µxº'qèmô~ú†>Æp¤ÚO¯¡¡Úwß’üè±9y‡ô9±§xú|û2¥tN4ÊÃð9Òøýô w¡|ý´_C‡ãhG?½•!r8½Ç·â )UÑ%Õ’O?¬îkß¹K|‚FÏ1b0›ÿg™OÛÔ>A£GÆ•s=r¤ï.ÄeÐc7š)ú£[ØÈ<ýDƪñôpû2Öóè4Ñ7°V¨±Œmv÷|W¥í2ÿòêŸ^ü×W/þ£†x®á¦”VÎǸ‰£‡s ¿ cžK½ùôöæÿÞ| /úo/èe¤Vn¾'Wù'úÿŸžt¬‘oò¹ôP¼_…F#ôrSéJ­58¹Ò§4Ir°9Ÿ¯/ž3¹`míÜc|†ñò9“)Ȉdéò ãÕsÎt¿©ŸŸc´~ÎqÐhô·ò ãuZxSùzxŽáh1ºYŠýY¦—i¿ÓíÜv†ñê9µDó«ç£Ög¯ŸSé4¿qîä4ñx´)Óx‘¶]Ï0Å$Êõˆç#çgÖ#à~ó¹çgXßAëqÄçó—Ñ)¨§gÛ¼ñ ÀIAª†z®¹ù"-C¬t»©Ñò>Çôò9fŠ}1õ9æWÏ1QìËé<ÂsŒGË)­Û”ô ÷Ç*WÎϰ;"%½0x³õã/‘6E ÄXC¡UIÏ0^¥\y¼RÿòÌ)ˆ’¤b‘ŸcoDÊ)óÞˆñn7ÒjðrÐmç«ËA_RèA,çŽÀÖ)òŽ›÷$¡ÜCPeJî_ü†deCã–+%ß÷KB"#L¥r®œŠ2oТD˜Â$¢Õuáz{ï$:¦EkE<Ðß*O+ÑmaK©D´èZKÁc©„°'ð—j•s‚³g2 …&` ÆçþË:SPi"S§`~¥ÓÕ‚ DÞ4ÿ)*N*T^6Q™šö‡LN_(­wÖ ˆ8ôŸò÷B‚šÏ•5TÐC È9•òyÐüЀV>™ÖtJD‹`n½¸ ¬5%¸Ñ¹º ˃ÜF [1Çp®&,Öˆ"®¬ ]ç5' O ó$ç¨<óœi‘‹IX+Ó•Èå¬ÈZSÒþ«Zf4~ËjI“Xä)-JZäL‰×Í$4½`ZEqËT±$V"“Gˆ2'ŒGåSf še7JEÖáà™²9©O‰h€¦GëIÿ•Ë©¤c¦•á ÎV…¬ÝÇ…­ÈÜyl¶šo«Ár<óBYŒ°Ù”°J &E›[´TÒ#íä>µ(‚cêŠO½ÔCnF$¢E&Ž$«Іh©¤¥gW­„}J22š ¹›D´ÈÊ0C£ŒÔEkJè¿5L-‚qÉU P¤t1U#mcM‰«Ñ¢cöuœ›Œ…ÜL"Z¨dyÞ$TÉÓJê’A·p˜JX‹AöWn)ïddÛª•lb)`H`¨äcID‹¡@F±(‹Ö”ÔhѳRø˜+,Ϋ])ü´%­O2²ŽÄ³)iûjEX³P팠5׸*-²4J£¸d^*霎¦VdÏl¡©ç4ü”£¦D´Š XhG®(’Fá$Å©EžŒ¬A¾'+ïãÐ&Öjd'X[}^Q%äÙÍ´(O^Å9û†ü¶$¢…T mú!ZS’ɦl*Z™#eC~+&-äJÌÞþ~IF̘ٞ¹ü6»ÛÃäÅ{ý8§odø#À3d¤Ô‘ߊID pZþcË©¤v|_µÕhXžŽÈ±ãŠIX«#YBFqN’Ê”4½™}ÆÜ;ü JHu&¤K ^ß› Qöަ°‘œÃð)ïß‘3‹wø)ñß‹ä¬JJ2kG.&-d!ÈÈéB•+ª¤Œ©Zä°!í"ºyDÔ‘6= š‡‡:=àvÔIÔ‚—%a­¤ ÝŽì “”âÖœÖÝ2ùï¶€#qbw 8%~G’¤µk4N{›c $ÎäË$αͱe¹"Û`ðQ¢CÙd,½œ Hwn›ÑBÅk t ×õ} Ý£û¦ÀDh 'gõt ÓÅÍê4zDXV_’eu’‚›ºh:é–D´5!£\qˆÖ” sQ’Gεé‚ID Ys¥u;DkJhStÓâŒE1¬†NEÛ¾$¢…¬ ­`­)¡¬]µ(ƒvÌ›"dºjÑBÖÄ®§-Zª\Q$ÑV•¨úh¼Ë«ì¯è–„µP~0”¦$Óô‚*¡©”©uðgšªº_BÁ–D´45}p^’n®œB”„e`2˜ÜÁ$ 9Dš4‰ƒ“)äŠfAÀ'„ &ð² ª'\’S¨†h×5xiÅP$؉D´F”0I•zrE•È–­îƒCâ q‹ v¡a \dHà,à‘–ŸׇyËÏ48ñ•Ÿ—dåg*ý4Y‘¯r”I¨ƒU"Z…‹º»¦Ël’–g”IþÏu åáåP-nãð?ý äô‘˜ÍÍÅá’ˆ2%dðEY• Àé©ÕÁµmc‘žEŽÍÅõkJüXéà*ÍW‰É>< ‚³À·+ð%š”Ô5´ ‹Œ„„•L"ZÈX|Uà¯IùØZ™÷“¿?Êoõòþ(SÆýþT²Ý_m ×êq÷ÕKH–É/ Iܦ.UÙ î¨»àž‘,›î&qÁ=K›N•ArâZ1˜D´,¡E €aæ’Ô 'Z‘·ÇZD”Š!n‹˜ o޵ˆ&p‹˜±ó›2)#ÇÕ ÈÐèãh2K²€LÊdxd &6ø¹V &-äKÈ(àJp4 H’©Õ+8š…׊uà 4þ!f˜˜aIfHå˜ušUø¤m ù²ncM‰‹–½†m,ä¸p1òeØÆš?%©¼…®Ëî¨ù²zG5‰sTXHœÐ4÷r­hQB¾DM–¥ê˜øŸV $’’¬ '…á²IX‹ªR1ƒÐ]ï„Üæ0­ÊEÍÚ=\(–}÷TdËâwIÜîÁ)ÃØ†¢ WÇÅPÈ–ÛH*ðbiÛ–æ*1_Œ„\™·¡¦ÄE[²®üIZ¨EºŒZûP3ÌŒj—Q±“òv‡\%^Üa;º`i›—IܼÚQ¤°±Eä*1í‹Ø1³_D“¸EldxÆ‘k¬û±(c2œ^cM‰+ó‰·¿G$» ¬Ü8·e4‰¿ÇªiËÆB¡X/Bsáá>Ö”ø±¼qm³ÆB»«“¥Ù6–IÜXýÈRÛTNÐB‹&-$OoM0˜-T¥,µ¹jÅØv÷BЛw/“8÷ê¸Ð¶Œ\.†};r綯Mâ–±×ì ÒÂmîæB¹¦Ä›‹L´m,仺5;«Ë$n¬q$)pÖXHyáb,äΰ5%~,TEŸ¹b\5¿haŠOŒ&q‰qä-uA gÏ«$-¤Ï-™Ä%QW8vEcº¸Å®•ÞºÅ)ñ·HµA*nZ™kÆðɲgqÓZ’5-’¥sÜfE‘wìÁ+ô¹øàµ$kZ$“êÌ…„—/ÆBòÌÛXSâÇÊQ¸Z»£°R^6‰h!y©›Üâ”PBΦ™»1f?s͸$¢E†d̾I³O²¸ãÊÌ5㾌4~e„½îÑ$îáåÙŒEcÜ“_EZ”X’%H…¾á^ÖBÚK&-¤PhÕ ¥«IúÑ&ÉI²(ôÍ„79 í-‰h!…z¾$ Þ ÍŠ«e/”޽_Ø«3áíí5%Þ^8KÃø³zÍ\:Æ_¾ÒøÅö-•¸–dAHº×Êþ…Ò‘‡HD+ÙaÞ½w’ZÍ c”ZÍ"tæ ²mšÆ//g„^’¡I¸âY~ 2ÅÝï#Rhô~oç÷±_ñ*ȰGh? È6Û›ÄÙ>vÍ_v\EîYˆÆÏ;º\w‰l¶{ä*²ì÷Hµ¼XÂîÑ$îN1ëv•b¶{LÈ¡> -‰»G*ô¸è1¹ŠÜò? /¥ßÌÿK°ò?ÉŽ^f®!w†FÏ;¼\·³ —hÉccaÇ}¬ Ž{ø±LâÆ¢hÏÔΊ>Å‹‘?·TàÇ!«s¹“ ”j“ø&ÑBö„UñI´¦¤2á­ZÇ-3JÈ~‘2eϼe “¸Ì8TÛ¢3—Ñ9#}nÑÙ$k,îˈEÈtD\òg*";::Êlõ\­³ËÍŸ¨8´OÔ]@%¸Àc­%Iëh>à,Ôk-‰ÓšÇKË$NK–Ò,#±LiIœÖ¬ƒ––I––%Ðe“8­ Ûxk-­/µfSÜkp0ßâVqºBCP8­‰Ûn~{:n_f.ÁOq_©Ÿ²}N§*Ÿ©âê§ŽÏÌõŸÈâ·/ÎI'DùG…R o[‡°õ›ûMIm-;–+Z„9=4Ü´ŒŠ6-ÇEÛøÜ‚œÖ”lZÊÍ8-•lZ:W§åfÿ3Öö‰…@ ’Z¨zh¶)eô"óxÊKZ—´_±íòp.­¹=5%•O-„ìžIüNüJßJXÉ·(.Ž¡¾õ_înŽZÂéGžô2BJš4—VnÎ=ýæí7þîÓÛÏ¿úïß}úl[_Ñcþ[^#|qŠAxê_èZo®$Í‚RŒÂ¨¬Í/‚ešßШ2 ‚U:µhóC#dúg&'Р{¡‰Àr«Íý«ºØ›:–*®wÙ­;!åŒê“üHÂHÀœ¦D‚Má:¼¢Ù ŠÖ”Tã_¹Ô™[È”ƒ òâýýÝç·÷ß|x‹™’ ƒ¾÷êŸ_¼ú›ßž^ÝŠ£çÓ§·wß¾ûá[çÀÜ…2$Í&IÅpÆ #`›ñqqh•[(*HZ|[z:zW›`×ÐÑvG’D%`úVÅ)ô¾¡Azô©m´Î*Ž•%Oúk,Ixú€lŸˆÃ¨Ì [| ÝK¼}z¾ð-šÛ4à” ½‚èçà›€ÇFôò!Š“€"Q@5\…þNäÞ3ôD&ñ@W7…fïT<ójNà‘€*vÖбšÍy fP²H¬ËkPw½ZAòu\5 õr)í›*Ÿ¹·æ§:•O{W !àCtfkRä>Ë{×NDó ¼æ³)¯Lp¶ö€‚eFÈtš€÷«ÑæPŠ`µËI8`®6ðò<°µ³„)˜])½Ï/Y'IÕÝ (¬ö¤·i}ˆý¼R«_£ãVî]E«rL°š%Z–n«Õõ€Éê]@ëJ`‰v €%¯,˜m“$²fÔ¼%³DôA£sUg'ó•¶4C;_GåÇpÅŽÉë<—·ÃîZ©Ö‘uѶ8;ކ¯²ÊŸh)ÂkM¤ÖðpC%j L<ƒjHäcj8ÚePÃùwPCz`Èb@ÝâÆá4~`hÃiH¼‚'N+ÀB‚ܧå¡5¦á4›P+$‰R-+ߥá4òL¹'Ãi‰¶¿`;i´×%‹O˜–ЗÉ_™8 OjtÁ`ŠÓRœ€aâ´fz›8”x€á´ˆ4Ç)zâ´ˆ5ÜpZl²¼ëÙR\©0àG“bÂ=yî)L#¬Ù'˜F`S}Â4‚–Q‡Â´ È­Ïç*€Y 0­M³ízÆ¡¶©Ã0­r9ÃÆQ˜V žä¥ÕƒîVá• 4’L$¢0­â”FàŸÂ4n7Wà&0­è4à&0­²kð·&Lä}Ü@h8mô¦&8 .ÆqÞpÚhEq÷Äi£Eë«y[›!W vQ~5R#s‹Dpº´Ù~¦áìŽ/=QÚ “ )J3rJð?ù’¢4´ˆËbξÝxL¦( -LŠ·¥¡‘Åœ(mC ¡´An,ð~¢4lk¥¡591dX}¢œq&FÃI°äȉÑ:y–¤lk»ìEÍÄh½G© £áù%Õ™ÍMû²VCbSxj ­­¢q5ÂùoB4t>å £!¼ (êõP,cÝt¥)Þ±ž¸¢O<Fët÷WÚ,{V§ÐÖƒh’«B+…h8XçnÚÕ¨•§a&Dëù²Ì ZŸ`u?!óò·&Dáà1Ehh£8¦­§C9ÌÐ:^W àKhå¾'@ëQÓØh¨íÂ)>ëñGU Ÿ¡?­lø¬ÃE¢ø [FVâ³tO>ëh˜aO|ÖaØÏ´SÌÁ33Ê¥&<ëÇ\—‰Ïðd“”Ÿu$vþÖÄg}–æ†ÏЃ,¨n´FéAðìhx¾BÀêhgÚJß PKº®†@6h8€âCwÜ@C :Chü%ŠDZëŠD×Á,OØàªî!€MáYëÂÙ©f7¤¥ø¬õ‰&>8i>k]#‡á³ÖAΆÏðh‹|ËNÞZן­á…¢£ ¼€Ìp"4DhA‡¡AÂr"´&Ì9„Î9Ê—¡á1A‡¢¡Ÿ)K'õoÁ×!Zå0ßÖõ¬{ñmève$µø¶ wêø6¥M߆'n´ߦöŽoÓ3ÀÅ·I{ŸçÛfãýâÛÈ……J›t[ÐbÈ`zÇ›€?¥Ûº±v“nkQ1Ú¤ÛªÄcÛŠÁ.¶-ëÓn‹mKZ&ÛÛD[“nÃ[£M¾-h«úâÛŽ*¶Z|Û¡†ãðxÜØø6txd“o#Û”ºòl›\سmÊ­ÛvHÌÜØ6E|ŽmSnm²m‹[sl›pkŽl #Û&¬õd›RkŽl‹d[zÁ±mÚW²±m"pd›à+O¶ ÌXd[UhäÉ6öÚlKȶ°a9P`|ÜàØ6ÍÛV„sl›bBǶ… ¶-Ë­;ºMŸ&Ýè6oŽnTèØ6^Ïm*ͱmJ¥Û–n.ºmBIO·)—æè6åÒÝöˆmã ³±mbQO·ÚÕÀçFé*ÝÖÊ„n“o£€,çŽÆ·%ÎñžoÃj1çe|[Ñ¢ØñmC Ç·•³e“oë}‚°É·å>¸É·…¤O¾æDeòmòðçÛ(HþŸ|ÛQ•ݘP-õ™'p«U0Í$Üî„)›Œ[Š rŒq[lÚdÜÈhBzM¤FéQîÀZlÊG/Æmžû,Æ­7·yx`Hþ¤ìÊDj`O› ¹I¹õ¦¬’Qn`äŒsRn´xY€Ù¤Ü²žÈ,Ê ˜ÚA5©Y a£ÜžZ½ñŒ ´G¸Ñ,<áVûÄ{“qCâ¼ñ„[žki„Û"å&á–ô±E¸Å¨ c„^æqÁÚ¦~c„ÛSgn#êÁ²n]³×"Ü¡åDsn¨µ„<›„[×2­†IÁ)áFhrš:7J†IN4'å–‹r“q›9e1nøÃθ¥C´1nEVcÜâÙ‹q›ÂƸ…i‰Å¸éöuŒ[¾`ÜÆ±Šö¡§5ŽqKšqƒ×ˆŽ1nUsÞbÜ’ædcÜú$–㦷°1n’=ã&’/RnByÊMI8¦´h76ù§ãÜÂç¦ÍŽsCûÒçÜò˜¼œqnM)˜Eºeeé6‰›EºÅ6I·Ôõœv‘nE¶‡#Ý’žµ-ÒMkøÅºÅ‰WŒv‹E,Ú­\ÒnJ ;Ú-êÂ,ÚM{ëŒv ý¬ðnònòhüÆ»eåóï6yÌÅ»µ»ñnÇP`i¼Û¡\øâÝ=_¼›ž&9Þm"ñŻũc¼›Ò¹©@“+M¤ºLf3‘è2™ñ¢ÝŠî‡E»ÍƒåE»]f´[/X´Û,í†ü»‹âY Q±‡7ª" cÝ(qB¡ë6 ë6é±E¼iꈷC¡E¼µ¡hΈ7J>‚@xkGñÖôYÙE¼5=bpÄÛdÃñ¦‹#ÞâDoF¼…ù-#Þ=¨5âM qÄeEeoU3ð"Þªž,â­–Ù?{^mCèoèÊܾè»çÃS›¢¼¢…[i&¸JäÙ·²®o©àg¶¥®ž>t–àí«Økü6ÒÓÿ|xóöþ&\´ô©8Ýþgןyõ2O¾:có‡*=¢ÿÈ/>(!éË”ùmÄïø½É8rß›œJüfÀŽ4Ò©ÿííËFá­‡ÓÿÂû–Gùô{~‹q£( ¯+¦­Dùu¾Ñ˜vcÑ·0 ǧÿw‹‡Ž˜6…7¸ž‹¤@ù/u>Ýñ‹¨inelïtæ—3ÊEÄÜòQNÀŒ{ µž>‰¸ŽíuÒÜ^·,÷øíÍ”šÈ·—0³^Âüg'þŽÇ ПÕDt97˜{£³¾#›"êOw»e`fä…fÇ_íÃÏóáhÐwÓ>ÿ«6˜7KÃ}OR2U þMÔ2DÃK°ÿ¨v>~†a~±¤Ò>špÜE°G_þ-®ÐǨ¥é‹±KU–²ù¥ y.%Üå»ë/½þ,3ËG£•\÷)KI¡™älÐAyãÂÛ+²_ÎY¾¤¤†wÔbª¯ÕúGʧ‡÷·8_(ÞñÖsòFÊØc¾º¼msÌpWZðªWäëÕæþýáÍk\LP^6Ž—”ƒÅ¦¹ÒJQvéGhx8þŒfÝ7k'¹åû+Ì÷9…Í`â~¿v¿Õ ì[<0Ý‘}ú|ÿxúº×ŸÎ¼ÜG ²vþÕ¯ëk¨‹5 ÇÛ—èJ'÷:ýíúø7ócÝ?ÊxÏñ|NÆ»Så]õ”În^6<ÂŽ8ˆ…̘òÀ£U˜OÁSDdø€÷Ñ-cuy¥)½ÇpúðúžÕK ýôÝg±(ÝŒlî‚è÷/`aý—Ó/f;à½ÞˆœÖV …NZ¹ïxÐ$óž¾ûÄ×"ûÈŠ£½€6þ÷·è{ =²]ˆu îãô¯kZj†Í¬/Í8sEÛ>Ìð¿e‘c;ݽ¾eZ?`$ œïœ ˨yû!€ŒìIDwòï>-ÇãoV*Ljà×ë‡?ý~v ZÆqmöpŠj¿7ðÚí«»´¯ïÈŽ¿g¤Oìà&òNñ9F?Úîj0ÆQ¶ð)R,ÖÂDJ;ÙI?¸‹|³~ö`ÿ ™Qüÿð#9s ‘ûè[R1Íw2@§Âùb$ÂIeÛ©|#„‚jæ!@*¡ <¯‡<¾"©3ÖãÜUgš»tl÷ÃkÅ@÷ÑýÑ×d¦¶hA‚œ„”“ú•Ô[œ—¼TlÞzð3HQñðáÕ¬u¿€Ã8º@!«£kN«ÆM{Ú~¼ˆŒa°¿;"ú–È»:÷s÷+ó”BÃÏG|w3ùŸÕ0OØf´]ú%vÚA@^3x ù)ÿ°ï•çJ »W!LÄÓ ËT5zßšwi·¡wHeW»Z>ËÜ©¶Ú„äƒÈ Ùb†ý^ÿAWwæ»™…–Ó"–3Šw…× |]„#\™f¹Ù¡˜˜`Ð7Ê6Ö‘K¼üùq~¿;î$,Å’ü}(sñÒi¼û#rMvì{wÛMÓ²›fÞõŠíŽ­®ð VÇAÓ‘Oïî$’†O{bYpÚlÖ«ABX¢¡·ûz'W&ß%liQ×Ý¢ƒ‡–w÷Ζ²·¡·]vù3.8"û¯Ðø…—¯Ž÷sXúÍs„ˆºÀ/ŽšŠ‚Í-ï>òWõ|cPù~ÝØÝïW q¤b™==á šý£ÏÊ‘Êiªw¦“ÿÚnöZ §Hd©|– 6J"¯V±q wùmÝ{öµÄø¸lhnÓ%zûê_ËáõiãÒð'€Ä­ í#2#‹iª§?ñþpY?ÞëUÊ|+s„/~¾uñ,Y»u=l¾ÀwG]u/W¾g?“]B*mÄ- êR áÁüº‚£á+~`Ÿ6ûޝð"x`±}¯%ywúh!ÜyßÇɧ“íc\<ħàÚêÝíµ"à|˜¬©¦Ã0H­\OÙLJõñûõñw'÷ùööš©qìHŸ£ÿòÂíë…Û'òÖte*ß­ï¯~¼[?­?~Õ\qÖAc^þae²«vKçîÒÇ5üýU»}ûõvëÎl‰3?ãB¬>íɧòùuâî÷+9:'3ߌlÍ:R«Wâì€ï…Ÿg;oñ->IÌg”Š:t[ˆêÌiÅÕ½ÀÞ¶ÿËØ+Ž_vô鶪OyÒ„…ŠÇ§~Giì£X©è½tüR|©xÀévða$ íë ,ñoD¤ùè"Z*þÚù(· [ÎK3¶#1x½Ìãó”ÏÀØâ‘ÛÜ“9ù_/OüÃU~»>þpuç}sUᯟÚa…åyñﯢØíçï8â¦á„ñ¡-°!L“ùžÀ?Y~¸žàÞÌëp±PQöã£WÀ c³èõÈœÔÁÛõñ‡õñýúøÍUòù‰è…ƒ¹uù æè%ŽéZÀ–A£@¥ÿÁ ¥m1œ[ ,½ö÷u`Òm“§‹8 ‘wðÍîdjGÐ 3èê“7¢Š¼»çWôÍ6YÏW_+ÓW­÷$ÌÃŬêSøô W£Èçᕎkg{±—bšK÷*i9ãܽsdÂÃçÝœcï¿9ÇÛÉ™«'üYÍâOCbføýàV¦IB‘êhO(~Fx°Íޱi[…ð£]ÎWG{ä“9¿ÁîûCïn/Á~ᛆÄ3mÞU»œü öz¥†‡=¹D(‹O;™é|?Ź‘ÐfË¿z\úØÉޱÐvMŸ®^âd”Åž‹wÈ8s£ExºüÀäc<ÍE©Fá|†éÕðDÌöß»ãƒÖøé_K½¨Zq0ü{ D7ùÔ!Z XåÍv/<ÿÇ!êQ½‘ÍÂïµ²h¼S%²êm#Vø Ô͆²Ýæ9 whõÁò”­ehÀyÛAW×ÿÇË=¡7±Ã¬G´^ Ð/#2^ĆŠçî£òËÉ]L—õc‚î[á‰ÌpIöÕ¶QÞËõ;ÝÞ¿Þ^#'¼O0ù3*Ý㫊“¯(øª +.¸’¿yýƒOVO˜ŸôcŒ Ÿwã>œ]¾rÎ䆸d·ªœ(nþx-9ÝŸs¬y… ÿƒA³‡ èU™Ùzº‘€þãâ0yÁ†›ù1È­íS®+^Ôµq†.Ó~\;ðá"0ñmt9تŒ9õ='lèÈ3m§N¡z?Ÿ'E1_ž±ò±û56oPñ¶Óè›*£\°. ¿ût}Y=×à¾ùûeµ=ƒËüdz‡mÛŠö½w2ç!Õ .‡êÀ‡ph†Û~êÚ¹ø¼íþÕ?o½îî?¸Ù]  uX.@çê(WŽRöŒk-g4|àméô·=M¶§üú)Xt±¥¹©dCHba ï;÷K¼&Ð8‡ö}7èmxÒìŽ}áÈefâ½½ä­&سÜa"lô"üæ×2Ì–œöŸ=%÷ô¦ó5‘¸þ"†ƒ¥õOð iÃÕŽ( Azçeå@®î.ë2}  P<àfäçÿŸØ¿C‰ûŽ£âŸêa°K–ëØqšd‹å/Ý”LÛÖ2£ÿà™)))™´1wÆ£M6[¡æ°¾ãÀ;äB«ÛÆp¸ð‰µ¹{½Ó剡?«||ѱκåû™ý1z£O#©‹™½`PK<±#ìGï)6O²¨XŤ­1>QA°ä— äJš×¾QÛ øz¬éœÅG°ýxco3ÉLïÍËŒ·œ€¾€5þ¯$X^mÛ€]Ê‘…ÀÒü¡¸ãÍ)Gœ|Oˆ´™è(Kû«£Í.3°À®Ð~«·]‚ÜŠ6ˆøF”oÜ(÷.ŒüðÆrdž¾<«‘¡÷s'ñÁŒ·d^œéÄ݉ 2õ·÷_ñÐàV¿H/ëEo…ÊŸC/'îÐK[c#êiñ[§õ<*qµ´!kçU /æù™§UoK¶£€ëÕ¡ Nì_hJá{ ñ”η“k)Ž'äW²â‘ò‰Ë¦FÕØ»×*¹°Tø€÷Çõ‹x+?ÝUå£ÞÕ+­Á%õŸ˜;ç)6ÎFp À!¿Àq™e+BÍ]$;ÚÂÕ©¿¤Øpní²JßcœÝÍOîÉ£.€<®êÂðdí¶o_úc×FHº27âÃÄ£0öè”,íb-ø‹ÑûQü‰›ágëšã]¶{¹z°î³bÀwØÁá—ìû…²ØåÉKœ©ç¦}uRûb{jËýÔ¼v[íŠöMÏ ¡x÷IžsaKþ„ñË×ÏòÑŒ¼bï–»8°ÓgNþ‚g“?endstream endobj 240 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 5658 >> stream xœ­X T“WÚNŒÄÏji£F;ßGÕZµjµZ[km«”ZuÜŠ¨e_BØ—ìy³¯– ,Üp)ŠK«ÖimuZµ«íü³ü]æ?æœÿôLgÚ9§3ÿœr?î}ßç}–ËdŒÇ`2™Â7ïØ°liàÇeþYLÿãü¿aåÑ ÷ë‡ãƒ~ÃxlÝk¿Y8hW·Á‰ÉsL•PçLG)tizü¾Íä´Õ\G!3P ÛA¿Tœ/Ž—ˆ¡DìÙ  =°ºÔÅf'ÔVÊ@­¶z5Tü0[ëÑÀ€J®Y4‹Ó@ß ¢SØÁþs±‡‰Ò†Ðæ!*Cá“Ç\oòTž]MkUc߇7€8¢-ÚGO&Q¢ uq-iÇîu¢&» ä"Д«HŸ¿`+EZWU“Ýkm¤Ü§zÑx$>ÙáÙ½jOd‰ˆ”Ïh‹‡4(.d—$‰¢@CHí`ªÔé<&²öZÝq^àIê•€z>¥Qh• æJ¬ång¥ÎBoÇå|׃øBüw¾Èú|zè=?Ç?±uY ôû¶}}ÖÊWž¢Ù=ƒ2+8(¸ùƒ»ðq=êÍ$éKlÎ u¾f(SƒF%¤èY#ƒBÿ’!SÈÖÏý³ ª)ô)ÍFЄï?}³?ÜFÑaôMý8; ³Á®3S×Pgz}Ïü|îÂÕsèqT°ßú ¢ã®!ç5–ߌ|œo£®® §9ôlzÞÜ‹›?E#~ðÓ$m¦ÝœäTßõ^Äj?Muk;Äñ–¸¨øÂ¹™‘oÇ®ØÝ@Çú†û˜hág¬á“Ãk8:£Î FB¯µJ5JT¿]uku péD:“.¤Åô‹‹H˜hß1»” 9ÈÔÔ.z =eÓ‹@¼°ü,š}¡ -»…ÆQW¾þºÿ# ~ìzfE®2ÔT -ƒ®ÆHâÑoùТÑSù#ßc}IOåȲdêd-Q(7)-T3†o ´C§Ú7Š@ _§„»Àl7[¼Æjª ͳ”5%ïgšü5òxäîø´ %Œuæý T(@ Ï¡Îp¬nž¢Ï¾-3OOâm¢Äyñ»€Ø*½c5è z+<üÆÉyêèdoìf ¿ŠÄÞ`áƒÏ^A¹•&Ó7ÒË ƒÒÔ«¨%è3ö‹€/×–Ëxh4 Pb;˜}­‘<ƒøz®S *¨‹dÉë%…/ñ{i­]>é6b4e¿Î׈Ôù”V©­<ÅÆ*±VO6 ƒWEÀIc—I¡Td÷ È*´ W•SôÌ‘YÆ#܇€ôç³Ý ÓY\D°•¤]½éÂ=Yp›…¨—£3ô!Î]0A ¸…éu&( ‘­Ä[ë«:0°Ñ±ï´˜B)9•\·ï?œÁ6K«Ug0 ×ö¨õãÚ¾ÒÍBWÑÇœOÙ—måü2•/ÒȵªGŪ3̨3‰¨½H«.U’¥¯ ò0 ×°/!çÂMdÿÕ“õF¾ºD]@iA¿X«=¿¢V0Páá ôÃQu_’6'†d(){X+ 7ºð´.’xÞ¹…–\9á ¹qG~…¼³÷wÓCï#@1œ#Ò>)ì$2ù«çn5'MpôD¨²¨ §,vË`Úûha îÿlMm¢ ý1Çœf=<³±þÊ­æni²ƒÌTÉø %$æ²Æ¯ópâ ôÄÄg¢ÉÐûðÊea‹‚Fw0_p}(Üçâ< Uúçrî°kv»šÝæ«@kW_f’æI‹0Õ&Ó˃V²ËÂâ;œ>¨U© äÆt$ üåäbƒ¶TF–îIzm|hª¶@µ¥»ãI=;„– ±üvÿ,ŽÞŒd 2Wq¶œ'Óëè>e)h*DÜ’^v±G:†:¾lºBYëMUPGœØç‹_’NÏWŒH_e$­¾æop8Ô¹ùR¾0™ÊÈ] ÑÄŠAÞéSmmdíÎÊ:è†ú&—¯¶Ç9ˆ :ºU­¶PA ·—ü6ˆ\h®z¸Õø[]:Tu-;ÎòKý“9ûÅ>^A^AÈ"7ËÉÆ\C&Ö¤œ½´‘ž°Sr¢c°ý³æë¤©Úìú•»ÌZY {‰¥—xçÐËýèo^L{žBÏå?ú…>æ7ß±üXb,-k.ܛ £ð©3 öâ áh;œìp@­Æ£¶j¡T…¡‰‘Ô%¯m‹Ø‰Þ$ëÕkõ`¢ÇÓÙä-®,RGÁ+ëO']úæë¿ò`öZôàÔˆВ!”‰ý8z™#ÏÑd½øV|PyÚûïÒRgúµÇŒ¡Ÿ€\àC†5Ûߞӧ®Áœ½¿²í¤+õ ÿœ‡C½mï·]9€–À)â—j I™³Ë £3ß"FÐ{Ȧkt¡¡—ÍûA0rðáÉù>cJ8Ë­ñ„ ÆôêÐôÐÐÿžã8vwíÿÕý¿‚š~‚þLy†RK†ö.ɬΜE³èGèEô¼•ýR¡uÁ‘úƒÇ‰ÓÈÂÙKo•en~ip÷Êε𡉵‡¨¦¡þ½ØË4E'h´JWÙ¾a¦ya¸‡åWûŸç ß¢Å uWÞy×̵êì:¨!,*“œ,‚…µÒê¦&÷þÉ ‰ òèi˜ïèqA¸(ÆÅtQØëF&Š7KE\Y†(: ¤Pb¸ÚLUà&<¥vA^^qzìá¢îï{QjW ¥Üw]ô1‡§}ÎBï¡«ô,… Yÿ‹&Þ^ƒEŸúAìß`À«¢0îúÀ»ìk's7—/H‰£Ò“cJ"±¹ÁA7—ÉùÑ7°+Ò©Ûü+ÓC@_â½ ŽQådp“ö&aµAðn«³³eE]ni–09©»ôö7Þ½QG"¦£Ñ…É8*HâOE†~Û*ä·ÄÏ¢ÇÓèùô“+®Ç‚ôÃI8Üët—Q'–Þ(Ëܲ ÒNÍ Ç±64±þ(ÕtúHN{çm/ÈþçwÎÖâ,~ö߀9=ä3­ Wb+¯«¯¬qH£«M³´Œåy[—N_Ûîn?|ñÚ#Ï¿™žž…ûœ]èÁ>åA>¦¿1;¿dù§¼R›UoÁ¯v¨âÒa±BµdçN*›S°ˆyl)ĸëÊgð;âæ«Wé)$]ó«á¸í8ð¯ápU!ûЀ \¨®n§${UB(!Öe^¼Ýö£Õ¢1‹DZ¥XM–mÏÎÌ"[ÝÜî54¼”¡Átêálú¡]µØ›ÝŽ îïCÐÔ³GÏ$|2=Ôn£9œŸT‰ÇWI&Ô5{+;*»©KèU{+îZ÷LJSêâ zòïŠaëðÞU eA¾„'/¡²_íãW˜¢ÂóF$7&|ÓÉÐ/”þNS±UXÄ/ÎÉs˪Ü-u­ØV+ô /ma¢=WPÜÖð#ÿx Ôj%dìöä÷^×cÖ{Œf⯠ў·ù”YíQ¸TÇ…3Ê­êšR‡ÐQiÄÒ5O?÷Úk]ß7˜š ô:=†œ£bMà„Ñt tjl2&•Ù¢ÓÛ­¤Ãém·Wvï¼T>žjš‰BZå½¹íTžOlXÔRbʲ)õ®²vè%>ùè‹;ß´¬_›«•« I­0è'· ÃqàóÓÝÌaI7ËÜF‘cÕYõÍPÄ÷ ÂÍØVK£(û ! ”Ro¨5mh½ÎˆÃ®‘Jäˆ]ŒNQE¯Ÿ¡~XâƒÑ©×ëtvïQôdK=âÑÉn•\®©À.gÇΊô ér ±±HåuNNœVŸ|Ý)”SåldéÕcP¡ Ë ÞaÎ!æ7çÿͱ2‚åBY åÜ·ºRõ¸››*ÉšýU—p @L©¥š”ç•E”$ä÷ïb<<ø÷‹»-Cè­¼»;zg}ë‰ÃƒWè‡]å|9OSAm¤+vŸv­[ÃýÙ¢ÿœ nou‹Ë&eW£ëb!vïËxû¿pù‡†QäYæ]df¡wP$§·ßàÔ›Í þQÐÑP‰4 eIàúà µxÙ9ø‘‰‚“…ò.r<‚úœÌ}Ùûrk˼¾Ž¶@R)ô ÏÀÅœx]¾ÈòøåœƒJ4愯ÞK³#ÉÝKâépxŽÈšB3ßý¦=u™ì½×ƒžƒßôÀˆœ“ïþø“ѱýשîO{:èíÊÞ´6‚îŠ_DîÙ ØX1Ös|ˆéfúc‡9abU ­¶\L¾27BÎS+P)#0®ãæýGIú‘ž,S&]\ÙD`¤CŸá0~õ‘Q»sP2¶Êt"]¡A¡Ã5ëm:#vˆdÅ'œtë¾ Y9.;8Fý7Ä.—‚dT镸fñ §À^Õ†&{М÷YhÑœw×ë«EK>ûê &¬9½†ž¿ŠŽÝXlŽï»yMî©­Ø/R(• <{: iE &x©OnKM-—?DQ(¸÷(ÿEú©/Ì"cù ù Ñ/H"pè Á©×cÕ²¶Ôváè-ÅMeÞR_!ŽhB?Zçx˜T”01ùÍ7!’­cš5Ù–¨Ðh´Äè]e=ZÞVßæ¿áfñ¯cù—¡hΨbo¾N­È-S‘ji ½P«Xçíñ쩎2(1ÆH‹aÝwž¢€/'÷ÑSâèɯÓÄËôÔç"Tj¹:³ „Ê"²š]N«ƒ´Ym–sP nØ/©T6J=²TzÁ U ™©ˆ¢¥½®ÒÑ`&»Ñ<³Ö fà¶\®¬6)œxÏ ’R¥ò<ìôä ¥.p„ÂDÜEÛ<ˆ¼Áü«õ^g¡Cþ3ƒGÐá“ØÕuòfÜ矺, ½Z­Òj”Tlq¬,â ÙΫVTz~¨F¡–¼J¿3C«ÖʇÅÏ¢7X,äÉógl­¦îô;x³‹’¨¸—¢ Sñ|nqkÿÇ Å‘ód]kƒ[TU­ˆÊ\wxÃ]4óü,´èfÜo'–ñRGÓâŸíùöÖÁ&&š|®·t°Í̉_h½ÙJ·Êù5”Ml+ÆÞêítú‰âäܽ ñÛÈGw9 _6¾ÛÎwàÔ±!ð@‡ÊžFìQ&Ê…ÖRogmKωÔÖÈU˶=N­¤'ÒË!‚à½8 7(ÍF O𤃶ÄÅc¶ ¶zòc4É3ôû¯0€Cn Ýmç}hC;kx•?ó‡«>4ûºú2» (·,»Ä)ñhÉôoÄíÉÞýf~x\ k]CcC™µ¼¨LX*'Õ–Ya°»áëwí´ÇA±z]B»^~j?zôlË ½¤AíãbàH_ß“–..ÜG¾;‡^¼‡sø‘RÝ:bìÑ¢õ ÃPíPêé7ž¤gÒ!"(±åPz%hñKƒ1¬ _Sô,¤s¾ZŒV!ßHøçó…'ÒûÉ‘§–sì•=GQ]#ËK㧤f»›¥£`g}x²>“+×p)»]œòWb6íMã™Á0Tä–ZôzMêz3“òÒy™ ÒzwµÃh"ƒ =þpÊ×W{Øt‚c‚ï‘+“ÈGÆ¿àž<Ñc›<™Áø?aÙ3^endstream endobj 241 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 1662 >> stream xœTkTSW>7ä>Ä’ôV47íTªTP¦¢.ª¥.Ôò¨¶b‡$DIbIx ]kUÖ)Z‚€ÄGAy´Aky‰:â”* k,Îjµ3Ó±ÕÎàªKt_æÐµæÛÕù=?îZçî³Ïwöþ¾o)ˆa˜™úÌì¼L‡9ݰ(Ú–á …ÈAŒ§”!ÊõJ³ò›È:ØSl;{½Íþ28€…ûÀ2m …JÐúÀyù3‘, & ‰!º ‚å°ì ˆ†5ó“(©‚°âÍ^ݳÁbV­Ü8òðaßȨDì0 ¾‡iV·ê…*í,õYxSî‰ö§riãZ×uÛ†±Ïü ü!r4wÔÔ+©7ö™ô­ñA±x“Õ¼NPß„é6ñzWìoçoŠY:üヮáë>»å߀Dѽȃx²ƒÂzxõú[·j>,ßS¯çø{ð.¼'²£uØ#Þ_îw€ÆAÁtÓÉ´IyŸ¬÷·d„ÌÙû’Žø˜ª²ëó¤ú‚:ç¹"X±av[©«ç [3ò㶪 5ÅU»ªv5Ñln?ñýÇZˆÄ_â?ÕzZ<-Ç{q?¾–uvõÓ7;ÃUvN¶¸{†NÙß¬ÐøÃßàÑ]æö~Ä(˜G´dž.t„ü{é¬áû–cïhÐŒñ%{ß-/Å‚ÑYÕ-A×_ŸÔΆ»å`U'hõë`.åU>R'Úí¬Åôó Xu‘“à• $À© ®òâ±Ã#X8ßXh*t;¥‚2Œu¥Ie³µ…’qˆr%þþåóëÕ¨7bÛûeÕ;XÈ©«IXÄåé›W§Z?òœ>ê¾äÒtì¨ü`_íÀ)¯L[ä–èÐnÕÀ/*§LXÄŸ, ½À«{^oíÊq fÂÒ¿äÞ0“Œýq­q8§æ˜Spñ0ôŠ#=1/Ìߤ£=¼ß=|Mòvúž» 33òÈ+\ø¸ŸßÙûÎǦ±ˆÂP¥~·(IyåÞ³°fŒ ?n–ˆ†+JËÈÙˆÓñ¶ºü“ï¸ß;¾·OØwWwitñ§Îâ§2`. üÙû ÕCƒøŠ¬'Ë8âšÔ±d:tpZ{ø žlA$ $ÀEÖßûT]æf> stream xœ]O1à Üy…@‚ªv‰XÒ%C«ªíˆ1C2ô÷ „tèp–Îw'Ÿy?\gðGôø¢Æ:iñkD‚‘&ëX+@[L••‰³ Œ÷7ÞŸ@°Èìü®fâOq>•U»‡ÐkZ‚BŠÊM妑1’‘ÓR Œ¦:ÅEV,þCÉÑ\⸠¸ÆH.•¦¥I.`ýž >äl`_@±Sjendstream endobj 243 0 obj << /Filter /FlateDecode /Length 3290 >> stream xœí[KoÇrÜ9ä–ÛÂÍ:ÜI¿JhÀ²¬D„ÀÔ È9,ß´–\šK:b`8=UÝ=ÓÕ³½\RPbÙtPaØ]]]ýÕc¦¿ý~ÌZ>fø/ýp>úÓž”ã“ÕˆOFßxøë8ýwp>~2ƒŸ´žy>žâT>µÒð±Õ¶õRgç£×͓ɔµÂ{çLs4a­öZxÓ\ƒÈ³ÌQñ [ï¼n.à©t–ÓÌ'S)Eënþ O…3J»æ*ŒeL¹Îþ>Ò¢•̉1ž–pRTàÑo¼5ÏKÆxåUÏ#Ö;× T­P‚û"êèvÊ%¦RùÖ 1•1ªÂ¬‘\¬äFwÕ°¶^ô $´Ž÷Éî“p¶ÞC®øÆ¶Ú *þ"w/×@†£0$ɤݘ—E~­†ŒÆ2°1çkšÖ*¹ón ‹XFïXÃ-xÖ¶Fgã¬ç“€gˆC(AгXÅú:¹Û¨ø˜‹b—eA+ÕÎÖb®s™­ j’AŠ&HCN—ù?鮘UT‹ŒóWd•3j"< c©UU‡+ð¯ÂS_8¿?k'PØÃ94 ¨ŒŸ§aI+šè¬$cwe™ÒÒ5'wãˆ{0YùÍ!Ö´he j@n˜u;´.u0ÂxÚ•¥€ ½LNËæ ‰A3Ó·}´ÿZÄy¸Æ°ÁSsàœ~^½y¥Ú».ÏÄ8NGÍÒ0Hâ8ö<4W@=¦¹‚+€×PÒRìð’R¾ù|2´ 4l2(Æó,^WŸB˜öò§Y|™Å£,žnÓF4ìtö`•©=Èâ›á(>Ó‹G2‡)¦?Í޳¸ ÓzñmϳxV@ôñ$ ¾kè:^ŒfŸmrúù»yŒ8ú,‹o³óþ’ŸNóÓUU/Y˜øüYñöMÂK²¸_w‘ÅÃºŠ½êa­²x™ÅeUñªªa'£dVÕpE¹Ež6Íâ;‰Ð‘а@Vèyœå¸°IWf½üS·°Æ®³Ìâ«<ö,‹×UÅ;Á†igÄŒj°å™‡æZPô.³H Ö"Ãp7‹e/²øMé{Y<©‚“˜C"{;"«§Z5É·ÛÒËVx/·­6¯.L– ˆÝºe‚Âêé?eñi{•qq“Å“: wúħ mN© †¤o‚˜‹j^ºªŽ­ãˆ €$Äz¾£Åë|8oxD—CG ¼¤j^ey@›EòT…£v´U#ލ•µ•eUÙ›Œ%ù—@‘$yN0»Y|YÍ*§Õ´3/€”zÈ¢Ù“ø™ ^Éšd®ž³øYÆB!Ö:¢¨ðîÂL sS… Ia&˜½ÝЏ÷Qñów8Oõ¡øeÞì<‹‹¡4>αìzï¯òÐë,eñž-ÄÐüµ@\Ûñ!É"ûÕy$¡ÖUìU "§þ‹n!¶‰oJ•¤Ï³øc_UÓÞõ†Ý#d¼åRßàÓlj½YÍ–f¡°Ö,ñ&‹w6 ˜U>ø¶€Øóá´wUçô´¶òËê)ŸnKQk!º“½¨\ñÊY;8¹VØBR1ãp¿ä»œrYÕNŽlIÔÞƒÉØSûjŸ!ˆ²UÕ[ŸÓƒìýrµí˜ Ø„7/¥[‹·>å›)ÞË,žg‘ØÚf‘çͰêòTeQgÑä^L¦?(y»̺ŠkÖcã÷#¶íGW·Æ \[ü6 "}* w­w4ö~®]Êm»¬•ܶKVݥضK2ÖfѤž“K^\#t1úbÔ<šHŠŒÓ±ˆët)pà ö:‰Íca …| òmyÆ«P=ÓÂ*jP\—÷ôf" @ö·ÝÊ&vŠàB5?L´ü‰fN¯TÉIÚ“×ÊP;½‰o³¼Jd.m°ós¿bòÓ›.@®â¡@ÑQöÎ6srÅ|˜†pCo\.âÚNs¨4`èFJ2Ôøx /º—ñ!ãúððýÕÉpg’Òñí|Ÿè%'ÜAÊûx¡—î¡zͦ¼ *.ƒ§ÝrexCû¦ s)0ãÅ} »¼Íâ5!¦Dš’æ*“Ž¢.°¬Ü[¡“7¼½` °)È,AmÉHÙ@NòÉG½ Q§襦Ï“ž²´ˆ£­×1Õ„Ñ”hi (ÅøáBâöÑ _ÞÅ]'‹tå >§Ó£P< rd±°©ütlP²Ã<ÓÊ0ÎÑ W^ó-;ÇÈYJ”X Ñï ð °Û¨sSøÜ*Hʶç¢lˆcòx~ñ ,ZTUäÞ¤àÆÎ…–®ÝA㾚¾Eò¯4+;¾ÚÌøP6ã—ÕB&4à[$üB'yRýpum­¢ú^#ùƒ0ʦÎáô¨m¹£gÚ³¢ J¥õ‚„ -}Ë8ÂûH-†  Pï¦9=’Ãé0çœÇÁ¢g2‡¾ï·kç𘶢×iÁð­®§ØDôÁsãìæ6WÄLËjE ÚßCr£H@³©ƒKÖæã¸ :ïiˆ)¯9„Lï›âš“0o˜92æœé öµ²­u–ÞÖ÷PBcÿµÌ)èe¶ã‹ß›ÿ¶™”i‘´ç|…ÿæX¥ &æþ›·º°¯`ó­Ñß <”þ¦­ã[éo±¾cê¼kô“œøo.ty‘ÿ¦0¿Ð1þ>¿7.(ß@€c™¸F€ã ŠŠOü7(ÇšZµ‘ÿ¦ùÝü·GÀCOðpbåžR¯.¾ ×!6W¸’;÷ÿóä c×> stream xœ]O1à Üy…@BUu‰XÒ%C«ªíˆ1C2ô÷ „tèp–Îw'Ÿy?\gðGôø¢Æ:iñkD‚‘&ëX+@[L••‰³ Œ÷7ÞŸ@°Èìü®fâOq:—U»‡ÐkZ‚BŠÊM妑1’‘ÓR Œ¦:ÅEV,þCÉÑ\⸠¸ÆH.•¦¥I.`ýž >äl`_?rShendstream endobj 245 0 obj << /Filter /FlateDecode /Length 3817 >> stream xœåko·ñû5?bÛ/ÙKs ߣ 8‘ )Y@ 8ùp–dY‰¤strìA{gHîr¸âé$y]´(Gƒ¹áp8/‡Ü_:ÎDÇñ¿ü÷èbñÉRÝévÁ»ÓÅ/ íòŸ£‹î‹C¤°€aѾX¤¡¢ó¢sƱ Lwx±è…\þĆWÄ<0î, 8<^<ë?µäL¯Œí_-Wœ™œ“ý9 …wÁûþ @)¸²¾?BŠL½¾<÷Òs­(Íæ`Í¥W²ß.WJ)¦€ËæÂøûþ2Ñeœ_Õó kÁ¹ã¾?I,´è/p” ’M´@jtÿ¶€dØk2ÅRx-%2åß&yT¨D»B ǹƙ9³\KO'h`áÈp‰Ï"mðÁ .FÎQ|ʲ²ß,WÒ²àAÐcÔ–°Ò„Jºó¸*ãC5͇ß,þz¸øC0jè LºNé; Cº«“îïÝå­þdçºIqÅ%ãV¿À¼sуžž]Ÿ Å)ïÈy­©Pšq£; sB ÃgýÁÉöÕær{òÉáRHp®ûÍõú|9.Lƒ®”3Ýìø÷0|²ÆZfLg4̈î0J0/GÌùâi¡òB1QQe ¥²Rj -T¦¢²Ê2UQe ¥ràŒ+B5`jæ,RS-Àö µ‚`62št…–àü&À0z) ÒZ‘ºEi)eÆùѸѣõIï™’ ©Ì‚8^+6ºUÃ@[$IðH¯Çñ#Ɖ¤”¤  ÝÁªn;¤;ð#§„¤*’R‚u "ÏdCPxØÕâfXRôAœäaœ!ؙ־ÅùNá·Ç‘wé4Dm1hn ƒQ"€ÏäI@hæaðÿîêô£/‚Q.OîÂe¯OO³‹òÔ£K?>½½>¹êÄœŠ½Uhä/”}°—+åq³.NžÔÇ•ô°ÉJJ”15̨ME•0•à–9U¦¢Ò -Œ‡B•15•ƒÌRñʘzÆà˜÷ÕŒ C©¤=ú@£=cj^¶ØJ]SQa™¯¨2¦¢R ÍR©>aæÏ/ s©`ƒéÕ£åJkH.ÜõHÍCUå›wï¤J÷àéQ@~pà¤ç’^²ÿºÄÆ”ó%â´°¢zpXǃìŸ/W¯B¾Ï÷Ö(¼üGvx®ILÞÚéól¸p4>œ$–vúúa$?!äÛANÕ¿@ÖÞ `¾É³;.èC†—Õû€4ÎP‚ö‹…JYôå]CÒWøEÖ°aõk²‚㬡k‘ŠJ³LZ¸þh9bA »·`Ë(GðÚQؤ؈ 45]~‹rjpì~óºñªñ‚Ì<êK¨lÈ^áÔÆKrñy†ñVi]?3Yá&¬qƒpFm5ȶ‚È‚?Q@âÝø°Ã⤣±„Œå Ö‰ŠÁœ*úïâºÂ ÊÚ z6«(zþFØ ½ªh¯3Zˆ~_æ€ à,Ía@yŒîJV £"©äÅIå%kêÒÉ•¬öir‹Y z—³9%QºÓ9𢪾C E…yÞ à„Þü@&µ¸\nó:Ëc¤PVCˆëD£ÃŒ5d 8¾à»¡¨Ð äd¦'ÑŽ+qlJ7PG¡ÒÉJ^ÆH–6Œ–‡Œ´Mn®ðFM n¾ÓqyP¿¤§I6&†~]b¢â[DÛD{ ¥r¦`#q<¸ª‚´}N‡þ·ìÜ‘Au®¡ŽVÃb8hïÓR’«91 çáJÓÇK“¤=®üWŒRPK¼d\VÌZ(“Íé‘6H“¶E3ˆ*8®ñU^r¹›yª­ÐHÛûE|¨5*,% 9dǃ1²ŽËô< ÙíNÚéý}<·‰©WWgò¸›ó46ô!)*)¯ÎŠ£À¬80ë“[íÛPNÂö¥æ·‹Ãžõ—x°ZcºÁó)ª ý£èÀ£Íƒm!½.àI__°=)àÛ^ð¬IðC_à£>oŽ»,àq›ÅAS mÓ:ÀB¿)àe·<)àǘ!ƒÊ2D‹Ãu×N¢:mmR…¡£c¤4U{+·G„Ò†ö§çE‡ß¤Üè¾kë í¥L:QäŠñšÏNi“/Ï‚ü¨|×?j÷‡?qœM;q€6ÁVMƒ$Ј¦šT}T'<&‡G¤È_z‘^unŸž‚#o·ãÀ»¾Î+ðá–#qÖäzú Uú²êuê›èæû&˜±wVE¿q8ú?ù -éŸçï»Ð¶\¡NG<é‘ü2i)ϵ‘Uk(µA²Huì/‰]pç 3˵ÃTŠ­ƒPºJÑÿ]ßK-+ôq£ë<eê­Øóªj}ƒlÉØÔ¦œ/[=Ç:6°!…ÔåÛ½˜¿è#þ´ÉbzXà¤Í1¤QGZ3ɨ©IS÷ébÇbË ]EªØ²Â9øþVKÝâ,ᲞtJDì¹ê .βÆÓN/íTJm]¢úöc¼”8%ÄéýjšÜsâMž†©´ÃÛæû]Å Éö")Óu=^{*ráyo~Æ€—ÊïYlJƒâ¹À¥¤¯7¨¸Í*º±ÅùŽˆ/mòÑ›†+ís¯ÍrÉEîÌ!t9Ú9úŠ&ïŸ9Á&jÆ»µ˜£uLò-é w±ƒÚj×QÒê«ØúÞ%¶`¹ñí¥ßÙ¦é€bn›nJ\åq“oiI*Ü٠΋¦Q±Ih.«VpÝm‡ßU»Ã5,dtŸïð8& Ìzç—‘4¿ˆö›µŒuÆ$ÑÝ[l2Ç} *Çt ´§Ýdarî¬#3·q±ôØL?)Î…IŽ;òdÒƒDÃ=Y§rvZá;/†sPêÁφZÌå6Ê´¿õ²€¤[±nV‰ä@ýº€ÏK¹÷—‚]5ÏáÕ°ß“©87Úb7ºA|YÀvß“4+žìYsØI¡ý#­×[GDRÖæÒ}ÒÇnãâaÆ 6¬š×w·@»9zÇ^z\æÛ¶ÛƒoÛ†;*àÞäq›ÅûjSÿwôÒcFƒ“ ¤§!¥¥Q±IZŸ¥«nJWÝ”é/àÓ2×MÆ©eº„¨›í.“ÿX}zƒrÚtN"9_ï÷Èÿá;ñؽK&^xÜý¶›™è¯ xZ¹áÄsò+†”øˆ¿Ï ɬJq-'y;Öîíw˦ÎgëwOU±±!ÇÍC–ãQ?**ªÀVi°{ jëæ¢ í’üoÜq½1ÇÞVæX–‡›ß/þ —FŸÇendstream endobj 246 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1à Üy? aI‡ˆ%]2´ªÚ~€1Ä B†þ¾ÎÒùî䳯ëH.qñˆ^˜¸ud"®~‹€|ÂÙk%7Ree¢ÃM‡÷' ß h~× Š§ì.eÕ!ð× £¦Yß4ª·V1$ó'ÕÀd«SvªBBñŸJŽæçM[ŒH©4-MrGø{&øS|ûCÈSoendstream endobj 247 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1à Üy…@‚Tu‰²¤K†VUÛc"†DÈÐß7Ò¡ÃY:ß|–ÃxÙ%÷èñI ¬ciñkD‚‰fÇ¢U`¦ÊÊÄ·BW^Ÿ@°Èîü¦ß$êÔ–U»‡ÐZ‚FŠšg]Óôµ½ 6R L¶:Õ¹¯PXü‡’£¹Äqp‘8•¦¥I.à˜~Ïr 6ˆ/>8Sfendstream endobj 248 0 obj << /Filter /FlateDecode /Length 3307 >> stream xœíkoÇñ;Ûq äèú.û~u€ÖH‚ Ê*üÁÍZ¢i¥"©ˆR^(úÛ;³»w;K-õ°.ɇAìñrnvvÞ;s÷]ÃzÞ0ü/ý}²ž}r$e³ÚÍX³š}7ãá×&ýu²nþvŒVzÏ9žsFÄy»½ZœÏÇã)˜´ºùØûþÿ(~1ãxT­³Æ£¯aEòÞ‰qå|ö2c Ã( þ³ø]Y8©¦(i¥À2Îîa¥•Ë3Þ ÅĹCd¦MREpÅAóx&ѵ` ¨ÖŽÅz.,ÊqQz¥Ó¢ªajŠéÓ¢¹¯RorèÀ™l´5ð·ŽJåóÎrgz&[‘A™uŒ6ÉKƒ7ö¶ášƒÅ¨† î{cË­h—@ pšƒ= ϯÇË ¸² Rè^¤a¡@²Þ÷Š"¥Š„qYÃÆ¼9úxP[…¥0w+¹(>hà= Z0z`õñ­ë¥D‚¼7FLHP„W#¢Êõa”ƒíyôKWàñÓ ó0«ïIƒˆ)Õ ÁNÂaÒd1ð>ó«¨] ’< ¿¿Ê= 9¥@…ë™ôüºœž …¬¼W p{=…!ôÑh=„Äѹž{H i`ºwpø³¹\Ýã©ÙÛ c”ÊÑ÷¡rgj½Qâ Ç=Ðt=T‰!·¾8¿Þ]-/>¥`Ãôl+¥î9”N•l áߦl«¡Š¢ÙÙQŠIÊ%e$få¢\bµ4b ÕÒðóX,¤XÆZiD"µÒ€4”J#N.•Pô’»{S§¾‰ áþÐ+“û°@»Ä EJ IIˆà.ã¤O_  †\錣NßFzS¥óé :6T¢Æ"ONA@gÈ¡YWøžÕFÖÌDÕÆHpªjc<:“c&μ™Û‰òÚ/gHSÕ#Á©Š­QA‡ËÁGdœ;£å^T-3“¾#•ßýÔ}Rù=¨<8•Kã{mª™\L)×)yVÌ‚p(»=8L/vW‹ÍòzýÐËM¢83‘èvóöz7M¨*¸^Ú~µ8]nvEPy|âSŒÖcðfÊkâ@øêeÚ¸8RwH.Ï«$önµ„ÎM,ð>ý?& À‹½ÎÐ,\…¶&ˆËyÚš`§N†&©â»®­z6ï0íbCÞÒΧn¦³BÆÞbl´.®Åyot܈{eÊ÷sa± %)éÍ\@ Ê¥n¯CÓÒJ‹mMX4^ðöMhjnyÑɼŒŒV„õȽ1î™ô`sÀYäÆxûæ^îYîKorçt…t!Es5ráÊn/…CãÚs« §²Øk#£ Üö-2ê,V·é,Ø`&½æw„¿åð ò2T\ù¢½ÆUx*ÊËb^M-åÔ«NÒR{ÒB½{'” }é²cÅ¿„Håöþí½{AQÑÝ{°H€ŽËö8tÛA¢ÜÔ(iëšìóÒ¦/í`lÓå$%¦[bŠx’n8JšÕʹx¢‹lz‹MÁ×è繚ü›ü‹0ðJxg 躪]$±€ÙkCû5nIÇ?œ“y­ ='Ý| 2[4®àrÛ¸ ´é껢á/%Öüqk2zI{PÃÙeŒr¿w¦·e@œ@vo7’nÓB9ëîU(Ö)BDï•Úç|žý»ƒ1žá“#Sé„áùG± ˆhÌbÞiø(´Ï2ødM FÂP‡Â`ÿV2(²ÿ1;~ò*!¼ÉøÓLpVA‘>T4 \WÁE/3øSA]5Œeßfð,ƒWuÇUjä¹7ü*€à«BÌà« .2¸Éà*ƒAy®i %ú§,ÑÜfÈî"ƒ}y Íî­UE,û,s°ËàUõ”ëêÑîrзk_VN3Øg\ÂÃeÜÒÕñ±Ÿ3ø}•³ó ^Lrðhlâ~V6íÀdþ4ƒÿÍàÏÕÕ¤Á®³òÉÏþžÕ•6ðË ÕK>©Z$AØeÿ|žWŸçÕòê‹ n3¸©úÍ5Ýâ°ó™~T·éã eð_ü,«°Ë«ŒZÉê š*®¢\ã4\­UÐgÐÑkLŠ*]ƒ¬S`}Úr´>–M§Ï ‹Å“VXK ÅÏ¿« Ú êhzÃceOí˜á&ƒ$0Ë"Áð¤Š@,–†Öù¼óØ‹ƒkèZ5UÙéꪺK«¼*}QÝUµªªt]•‚-8ˆU­’UY]]f°£Æ0‚&hŠs†m÷˜j?Ä”ÛcJ¬WR¶‹"û }áIæ±Cªˆ\v™DTühø,ƒ«ýóìk˜h¢Ïº$ç! <=`#ø,çQ–JÕeñþ£ˆÍéðÔÇñɲ¼ÑP#;^ŸdòX'o{Ž2»…<ç½ÔÒäëQƒUA^ßTÂuл»Žd9/DT}ó¼WRÜJ3žƒßû˜®b· ¬Ð0Pö¨@uÃ-÷Sò«§á·ÿEù=$T0r£Ç6àÇѱàk½À¦fÊìmx¢“àFž™¦Â;z|Õ/>Tý5&ª(½>dèÛ3ôï¨êsåh:7«~å 6ðTý‰GrK7P(ïž¿EÑãµ¾»J÷UÐ> ägÕU^´À«Ä;uSwë‘(§¦=²ê³öú±²ûPù}¨ü>T~¿—Êo‚9ƒã ÷¬J—ã™q\æ´8ŸÁ©=¸,™ÚœÆUÃŽÇ3p`™_( “°E}þôà™¾‚åSq¨8ŸñŒYkíGͶœíÑ6È>Ð@¦êµÜg}ŽBÆp¢Ÿú…-è¤x1¼³Å›Ûø½Þ0À¬ –|Ñõ#¸PÅòiܧ U˜ÌÄAÂôw¦¿Œ‹0¹+çý\‚tÂ0,Æ:Õþ0ÇQ¬6Fî –GäâK¾°µ„#Òaò»ludú÷SÚ²Òþ‡ˆ Z!pt›d*‹o —ñA ò½ É÷˜^Þˆ/ßp<î¶|A l¤X[Óo0É[0ã dÈÝåóÄ[:¹ß°ý~®uøÐrx{Ä “·“¶ô•ÈèJ¡¼ìð*Ix™ …+äÉâ{Ïñ;Q_cs9öI‹M¤g Š)ßJJŸ’½Rñø¢µ(1ø@N½iÆaúmÀ;´ JTÄ0Ùßy=äÙÞxtoŠ Þ`˜¸}ŠùðIá>Â|•A2ñ$õÛ*ƒ/3xMwK¹¿äÕ.¯Ö/÷Ë ^dðó žWqÉ“\ëSÐtãðIßTŸ#ÃÑÓ:‰£*CõÎDp}•ÌbŸæû¹Àn«,º áâ&ÐRÇj³Ë´Žr½OZüë .2x–ÁÍ]«« b†.ÿË ¹=>M—ÃÄelþœ¯Xÿ¡M…Ù´žæUrK Ö{EWkÄ Ú7‘-sM“¤¿CláuÉÍÝRUà7UXWy ¶¹¤œ=_ #ø—žÞnUÖ¯v$¸ÔÖû:¹©Õ±«ª®oãa?2)ÕƒKÝý6U„¾ƒQ­ÕË—Zþ–Lz5ùÿ@m¿endstream endobj 249 0 obj << /Filter /FlateDecode /Length 159 >> stream xœ]O1à Üy…@ÂÒ%Ê’.ZUm?@Œ‰b!Cß@H‡gé|wòYãud—@>¢Ç%°ŽM¤Õo &š‹Vq˜*+„n:¼?`7=ø]/$ŸêÒ”U{„ÐZƒFŠšg]Óôµ½ 6R L¶:Õ¥¯PXü§’£¹Äyp‹‘8•¦¥I.à˜~Ïr vˆ/>ØSgendstream endobj 250 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1à Üy…@ÂÐvˆXÒ%C«ªíˆ1C2ô÷ „tèp–Îw'Ÿy?\gðGôø¢Æ:iñkD‚‘&ëX+@[L••‰³ Œ÷7ÞŸ@°Èìü®fâOqº”U»‡ÐkZ‚BŠÊM妑1’‘ÓR Œ¦:ÅYV,þCÉÑ\⸠¸ÆH.•¦¥I.`ýž >äl`_C)Snendstream endobj 251 0 obj << /Filter /FlateDecode /Length 2893 >> stream xœíZkTÇý>áG\¡H¾c{.ý~ ` Û€l”à• üaØ'fvg½³‹Eþí9Ý÷QÕ—^0l)RdÙ[··ºº§ÝÞ_ÑÉF¤†Ÿû§‹;OµnŽw Ñ/~]ÈüÛfø±Ú|»—8Vº(¢löŽýVÙÙx뻨m³wºh¥[îýf+ f;á6ì,žµ÷Ï—¢Ó1hëÚóåJt6FïU»Á² >†Ð¾©¤Ð.´û‰cà^_b]„Ñœg{Ú´jw˕ֺӲ=J´‚üÐÞbåóuyN-…ð"´‡½#ÛÓ´KE¨äz^°ZÓ¾!’m»bG\$Ž`”JÂ&ýw½>:ª]$/„I'‹Î £@Ï–m„áI“61¿Ì¼1D›|1IÎêk8Ë©v»\)×ÅE’·¤S6Úm²U6Ä☟÷~\<Ø[ü-ÃIÕ9ÙXëm'b£b.àÌκæâ°ù{s=Z൷Ío€Êø÷—kMc:¤å¸’‚m¬“xÂt±€ŽÀרΧ€À¨Lcƒê¤½¹8;m}ƒ`Ãë7BçClœn¥º±<­B§”jœ43úæòœïœµ³9Ü7g„GLCã<‚ýÍå×Y%‘Šõèf‰Ê¢$ÅÆ‚è”o´U¡Qüˆ[HžA6!Yz¥;a&П.Œ ˆ­l?U¸hEçc—’ÒwÚ3.Za\VÊ®`O®š3M+Ä¥µ@;ÆE+ŒËEà$p®i…¸Œ¨â’ûaZa\3ßpoׇúLT ÿˆôÿä,Z|HþÅ?4ÿ0Ë©ìMEÏ¡|z•";¬¸ˆØ ÕÇÿ}.ëBgi¥à ê{Î5¬p.¤¯ê—5®\×°Rp9ºÈOW ®Á"ÆÅlüKÑZ¥°IäšóÄè ¦™H±PR‰.|(þaòâ›>ÿM ô‚J`jÈßÎ{–˜ãŠ‹0ÇY¬©óžÎŸP-ŸGÊǘÚ~¿Ý¦úC.ÁH!Rª#·Dð½K¾?\Üyž4_5÷7/_/Q„²ž/ÉØOh¿Ï°€v"5Š[+Có3N?øXyKÕiÊJ’7]þëOÀy6aa¬5NÄЙ„¡±®8ŒˆaZjˆ³q ÆP/œGûÄo6Smp4:(œ\ÿ,¹*tIÓ^oÒu«Là Tªä“†N(Ó)]hèÁ⸆ҤQ©Ð᳎k`z©!fÙ~WÒ°T ™¤0OaŽ(S¶Qb‘3‡™¹ ‰†ß¡Ec˜r<ý&N+×ø¢ßåaƒ+v +ŸS]CïR’&ÏZí£ÇOv¡ˆŒØ²÷xÑ>yùæò ;Ýn–_0X~ÚuÁd»þóÈ)TôÃxçÓãþˆî´¨0I ˆ°ì6X0 Å-/oŒÂ·kv©Ÿ,0%N¸‡+{ï%$Íú¬¯"ˆn¾ñê .Þ§=§Šªô@Îygï ÀI,°jЉÞËñ/ž2.{3B¬;N_ï÷¡*L~3ð[öUpÆa{3äAìà£1‚WtFYDùãÆç¼ðp‡u‰ÕX†0€¼×js„A‡ÊÊm_ä˜+û'†Ùs”¦ɩݵõ{{VÖN¢ìäGôQÕàÜN¨žö 0®¨÷$-|Ÿ€ô"ûÆI/y]¼è`@îsrO°EeX_Q g‰Ç¶ë…×2³hé Í[*OLÛß+“QÞ×=‰ZÂyÒ¬xÀJ‹ÑÙ±cX)Äà)Ï[Òúe'{©ê­0Á¤4™Ž8ŒéËÁýA5©æ™:ðö© ˆ¨”«à@sqx™µÆ+]z"€”QY›á±Œ£ú<#­(;"1 •?53òFŒf\·Ý¼êVäªoÒP•ŒaÆu qÐׇçÖ9ÌK'ÙOŠ¡‡Õ©)ÓeD*«0óÈPÀEŒˆ¶XþÓrr&_Þ/æ€þéfž_d×ý›%YO™Wô"ÌëeKªÃ;Àl0ê;T ä>ÑOÉêÃôöxÍU>)ŠA>7®áåVà 9Ë’É„3Õq9èMø¹;=Ø&pÜy4—q¯€iìÞûòæÛôpâPDÛo–«tÇ„‚ŒPN`ˆYyY]E/™èÛDU™_yAäK"÷?&áŠÈMU5¦Ã×£E!çbªÓa–É}"_Uމ<$a÷ˆ¼M GDnø¶‰|Cä)‘/« LîpUÏ©+Þws¯?©2¼pí_huE«»j,‰<'òá<„7qÔ¿L?¬Æðª—K"wÎ߉”ÄÀpÄôÙyP5cSE Sò1‘ÿ ò ‘O«HÜTÝs4÷TÊcœÆMÔM×™Ág™áˆÈ5‘§D2a"ßyÈÛDž¹­ ÛUIÆ{V•ûõ²ïxƒEeÇû ýT¤'!Å3¡Àq ›„ã{DþN$«4["ëEðqõŒ³j( OŠŒ¯Ò ‡ q몸u|GÕm/ˆ¼¨æ«“u Í1¦ú«ªfªÊ{6×,càÃlàµ$̶š†êyç©+0v)«ãXM•лD~I(ÈZ[½¾&³ÒÈš×i•dE›Uê·u ýçŠ=W÷^2ö;2vMäfNf1w©ƒõ/~IÀŽX/‰<$òC-cž^oªî¼Æ¨¶*øS†Œ]µLüÏ·Œ5×zâ×û3;îíÇÆ¦óª{˜°ÝÐÂdgíôñ9-CàÚüÐ2Ò$•ý;úD ,7ëó}ÿïÕ>‘âšJ sÚEÕ«çÕÈé÷ ôL%*½ÂãÊ?`ä¼*žÁeËj31ã}ÇœxÏ«ÂÞ›ª³Ýß)È1‹4‹Ó¶Ïƒë£4r6D±Â¾%ò´Z—;"%#ª žÈH¤[®¤õ2ý¯qÅyWÊÉÐáN¬ImEÖ Zíˆ DNŽ8ùLÓÔÇLcš3CÓ¨ý<{XC±Dº¡'K‹‘f„ìãEûÅò‹÷š¶íœöRŽtSõÅ1™Êêƒï+Æv^‹áI šä?Ú+U‘²ÓiúTY‘ÌÀ‘Qsµ®’²JF2†w2VY•ºWÅÎPVNu^Éù Û}rÜD2”Ýs#ëxrU’¡ÌPÖ|K«?ù‘ïé0OWÅœª®êê<¬ÿË#q½ú×{üe]ħOÕóÖǦ±kƪ²o_7aÿ¹ ©÷ ×üy¥¨ÉIvúk›èæeþÁÚÌÜUõp=Æ~ª6–sU½´zQ%·|uÚöŽÈ×Õ®Çf 6q@I)ŒOIó×êi¼H°&ûO"ßUWË?¦ùîwd endstream endobj 252 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1à Üy…@B+u‰XÒ%C«ªíˆ1C2ô÷ „tèp–Îw'Ÿy?\gðGôø¢Æ:iñkD‚‘&ëX+@[L••‰³ Œ÷7ÞŸ@°Èìü®fâOq>•U»‡ÐkZ‚BŠÊM妑1’‘ÓR Œ¦:ÅEV,þCÉÑ\⸠¸ÆH.•¦¥I.`ýž >äl`_>ÕSgendstream endobj 253 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1à Üy…@‚ªv‰XÒ%C«ªíˆ1C2ô÷ „tèp–Îw'Ÿy?\gðGôø¢Æ:iñkD‚‘&ëX+@[L••‰³ Œ÷7ÞŸ@°Èìü®fâOq:—U»‡ÐkZ‚BŠÊM妑1’‘ÓR Œ¦:ÅEV,þCÉÑ\⸠¸ÆH.•¦¥I.`ýž >äl`_@¯Sjendstream endobj 254 0 obj << /Filter /FlateDecode /Length 4043 >> stream xœÅZIoÇrd|Í!·†I“i×¾8‘8²’vâHl@ña,Ñe’#“”—üÛó½ªî®WÍâP' ƒO5µ¼å{[U׉Av‚þÿ>;;xÿ±ÖÝñå莾;é×nüóì¬ûèf8Œ QDÙ~s—Ê.ÈÎ[?Dm»Ã³ƒ^†ÕáKL¶¢š,⠼ÂÃçOû?½Z‰AÇ ­ë_­Öb°1z¯úS Ëàcý H%…v¡F3ÆÙ›+Œ‹ ‚0šÏÙžƒ6B­úËÕZk=hì²ý†h…ýCÿ;è<¯ëshk)„¡?Ê[ÙŸÑ*Á’Ës1ÕšþÇB²e¯Ù4#¥h³™ÿËÌŽk4à aèd18aÈãŰœv`»Í“OÒÜ¢%]Ì;'ö5”åT¿]­•b£ÏI[Ò)+îN“T6Ä꘯?9øøðàŸ 'Õàdg­·ƒˆŠA¸€2뺋£î‹î(ú˰Q{Ûý¨|‚ÿ_Þ¬h:3Ø -Ç•–;ã¨`†è5 Pº8“ØÄÏö3n0ª#Æ ÷ÞÎGŠpÚÖØ{o§¬œî€AKuapL甆AÝý¥•cüá7½q¼Ø@Þ߉÷çOë8M`‰¢Ñýù‹Yo{ÛÏj˜Wím?m À¬÷æÊ»Áz¿7û·pÀ_„}åðÔ@âîÉyµAlW¶CìD2‰{‹EzÜHVèý™é7{ /F‹jsÎ Jɽˆ+ÝþØÓæfoÞ!=’Y°upÙúP"÷&® ÈlvÞ}´5{ゟŠvoÉH{þÜÞr¯ö(NaýEƒìµ{ ö. ‚ö ¹¿»TBVí üY$qëìÚò¦•EC; bP¾ÓV*ìT5bçEÁHu¤ŒˆýÆÌ%çÙF}=„2rzð¤1«Œèt,ÍRÎ2–Í*#e–†R„ vâ<Âf-¸à|•Y»ú©¨:?èÚ•£B¾ÄAF¢^H‘ëi¯„X­½N•voÆ(@¶wü—€Ì¥:+Ô…ê³yÄyZdÔÚõY–ŒYFªYAúŬq„Ïrº!(6k©f! hÙ¬q¤šeµ<Ÿ”ª9£½º,âÝ¡3¼ÁÇ5ÎQŽãÐüÒ¡eÕ.ÉŽÓSF¶„¤(Í;¡i`ry¯ÔIa û$ÇõÜ•–àú@• ¥³ƒ ´B#SL 'ÒÆ* %ÏÔ¦‰Ô¹Ð¶IGiI &Õl±BÒÈ5"diÄPªQT×%+`ÄCœ½D² ‘ e:=uiÑVœ€¸*å?¢„ƒ‰2ƒ†J(C#1=”ô^aŒ&ð‡ª‡;)\FGÛ¡x¬ ’+ÞíäŠ êP\ó!„A+®ùˆ¿‘)1ËÓߢy'ì'™æD¿)$Ó<€(3&Í;[Ã4ï ƒä÷“æÑr¡äšwT<(®yg,’iL@tÏUoÁR\õèn²ÝgÕƒiú…©ÞMqoV½6F¿} ×<Äëè|B÷ïù.“ Ògå“à$]Ydç˜-BU—ç †â÷…€ˆçMå .Š›–•ƒ©|!&‡g¾@È®òØÁ†Údö â È–I·ÅHÎʰ]‚Ìœb#jJí™AP€P‘Î ‚Ÿe˜ 'K!¶Äë h;}!‚UÃôîPä¹+89)uR¼£\î¸âQÿÂZ†û!VW¾ EÈ(š}AC)2̾ Ñw®y§qV¨|•ÐÜŒŽ³/DRÍ5ïˆ Ç£³bá 1C¬ø®ÂEœLÞ:û¯Ic³/c9œ/ŒØ§¸ˆ§FòkÓ¨bpXrqD5œ€î4ù[@»“j8 M¹2²Ë»æÕ°)š¾zy›œ LkÈáéÆ;%Õ϶ÏN;IYhRñíðÓiX­~Ë2gó˜/ÛM nÅM)™.l…Š~¼œN·»Çé6D™î¡ª ó­6jLï>X­ pî£ëÿ±J—pÑô_§kaYòý/ŠMø^¾“–ÞyËoª¿_)OýÜt9€ƒþy>¹ ¿#ªÐy¿I÷ú`ÍÆêŠ<ï«eŒé^œX;éb<íùÝú‹êÊ:Ë¡ûoHºà¥Ã)_g©ab.týPÖ66¯Ë&§ÍçÆ»9ß\Ñ&!F€»ß\&%haCu¥Ž9š‚F¨ ¶ÇGL¿ùOo/Úóy/Mÿ CeÒO²)$ÐZS£þT&Œ¯7¦Ê QEÉü$Ã_GÒ^HÁ1° ²‘ÓïÏË£‘Á×hd‚Ñí6æêߎ¹ê¸ÕÕ[Äzb}-SxNü?õ)´é·g+ 6V¨ô¼Dw‘^\ò‘~Ú1@úœpl°ßxzDf+oHü¡¦^·yF rBÒ8fÃüÕl¶ßBœüD–ǘBfE‘„ôôLC?Gôï£q¬Ýh”ăéI|4 ðÒô惤’4¡e†xþ´4¤óPZËzZİÃ÷žò@q’' ábV¼‹ÂˆX©*;B}Ðx×ËÊ<ûH0>ÐøN¶üù¿n²-“û°-ù¡—yGƒpÀfœçQ×Ëè]¦ãhªÑ}6ß’Q wí‚ÓBayÐÉq uàrGÔÂhÝû¿oò”•fñrGREkÓˆí?JjEµ­Þà‰á,s*Gì&®CÿqbÅ@¿¨nÏË[ì–?ž$VÈ×)pެ?¦øê—Èü>Y±’8DÓ´ŽºtBˆ2T8¡.ãÛ÷7åòaA}òüiQò|ƒTœÖ~–Œ[ Ã%d%£¤Ò0²Í7Íè` ǧmÁœGfs¶}‘ Òlº_×d¿¯çó&ù3…Î §Gm¯Bí ô;e´K:%Ð7i•½ çáÊ»OÊ3:wiöè΄ݬң°UýO9¡+]Ùðyü/܈Îur,A$,Ñ|üº yö‘EOPªzPµ¡ûöÝÅ]?>ШºéÊÑ SW&]=ö~¾u¼ûÇ ôHLW¢l¿§3‚Ï“o†€r± @WŸ Lïíúæçö¤Dòð2¶’¶ªhYüg´HõÄ ÚÕæƒ· ‹À6šrs9 2+·gb…Š—ô•ýUvòǯ“àÔ²÷GÌk×LIŽ|¼FÕtÔž"© r0(Ê~DÐlÐ/VèëÚ¥‘"ÜqŠKï?FóÆvJO].NqŠ|¹ -j’ÁR¦€Kæè•Rýir3¯´ÏI49*çÎ"x‹ükþÕÓÖIÆïË—$µÂFòs$c¸‡;ÐpÏ%뫱À~ßé«ã“¹”ü¾“æ\°¨¡¶Ð_ó'9Ì([Ô*Ο! 7ìPH¢K¥†,û×âºyU-S¢µ"L8àè?Ï%$åÊoª2ËÂË2cä´ÂžQ2ˆœÎ|¹«fР;GÇ3Mf!þ8%×±"Y¤Ô¥r¶QqÞå·)l#‡?œÃǓڧ¯§üÇ…þ¢¤Ø¶Á¼õbžM°StY‹ÚôA]Ü‚§€¦>z‘YÒCpÁÓ ‚ÓÎçåO¦È/Æ%®úÀmmauГÐm›r.¯úrÉbuô"…´X¤è4˜¼è2Ñ£>Rl‘¢¤ËQ¬ù‰d!—ëjMÓ÷T¶“t3n4-î=~ªÇyrƒíXZ†K ñ5ÙCÆ–å4œÒêVl8ºŒºö½›»Ù_.7®ù ÝœY™ÓâËÛ““Ytéãßr÷1BM»”Dºø¤néìN¼+_¯ ªvxÞ¯–±kaSíÈ< ›j!dÛ}Ú6]íŽU`RÙŒ Q±Öt‰"œ®††*¯{;c«žµeu%)MÌñ1Ýa KÕN6´@_$²á@HÓá”j3ÖÕ¢9’`Ë<|Íélú b0ÝË5Íä e£Ý VÁ€´ãm·F¢½õ÷·\çà †¨/•R9ë yq&ã4‡²ÜPábkj¼¡šóª°À¶iI 3D1štõ:a@Ò׈hšv‹OÎÑÜ]|Ô™Š4  è0å>&ÕqöàœFJõ0sN„€Ít/‘>H Ì.4âÄPÅ£µD"£* Âbt”ýæŽ1],[W5;njÒ²p­jRtѸÁ3RW—Zõ•ÌXBͽG¹E–”­o¯æ6yYZ!s‡D;€6Þ‘¶¡,xPÿájMOÎð“Ô K<¼)äUsî5ÓïòÕm[ò¼9möû‰5wÃ¾Ï ùm"-Ý m y\È£i³@Qb"ß-¾)ä)_6“?ò¬'Í l_`“ 2÷a»,ðYh[ÈçMõòó´>«ïet]F•ÑO ùe!; y¼äa©’K­nnÓÛí§Û¬õªÛæf—M’Í=ße­›MÄ”vÒpì̆oi­ËÛŽ`>ò¨i¢·Ä27ççMs^5Ï8oN¸,æü¹ŒKæ³ ù°¿/˾mÂD7ç2v.šü¾Ú±YÊc³ ¿CF‘^ŠGõ½©ËÎÖ|PÈ=;ýõðÆÍv˜mÓ”ýÕ½­Íxxv›¶'´áyÕä¬â¡%У2áÓB~ÙÄ7;øuÁÀi!¯ yRÈóBnósžCÆ×fª4ÏÊï£c%zö;¢>>,ä¶nUÑcùµ,àó©k”™jl+ŸÌsû ï,Vd³Ž í”߯X¨z]ȳ&ɰ{QÈŸ É+ÿ^ð-ǵšäÖó3ÛŸ‹Ü›Bž.É´ã%ÿÓ»HÞà²L½*äQ!wEó¥|?65{ÒœÀÕy«2n­Ö. ^~.ä[Fsæq,°ØÅÂÍiÓ¿?»ÓožÕþ¡tŒ1)„húþi¼3‡/áàÒÐMÇ₟UÄÛ¦íÚÚ¨iWþZùìÖ²rW]š†ièd©¡©‚×,?ßÁ'lݪiôÿYQu ~éŠ#‰0>©²¸pÑ´Ú–OhUl«·š-¶Ùµú3ñúa!E† ÞFû 7`~@¦ÏýŒ\à™ÅÙm!Ïšar(¤,ˆæÇçΤ]­¥õ’>à7Õ¨r2 1ŒËÛºH+ÊèPH_HYH3×o)šºM4&O(¤)¢± ºˆf8ç·ˆ¦šRzþÝùÞ½ÿjendstream endobj 255 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1à Üy…@BUu‰XÒ%C«ªíˆ1C2ô÷ „tèp–Îw'Ÿy?\gðGôø¢Æ:iñkD‚‘&ëX+@[L••‰³ Œ÷7ÞŸ@°Èìü®fâOq>•U»‡ÐkZ‚BŠÊM妑1’‘ÓR Œ¦:ÅEV,þCÉÑ\⸠¸ÆH.•¦¥I.`ýž >äl`_?tShendstream endobj 256 0 obj << /Filter /FlateDecode /Length 14647 >> stream xœÅ}ɲ%ÉqݾÄx+Ú«E]fLÚIf\Q@›qAhÑ艠Uw‘Ý€þ½üøq÷Œ¸¨žÐi”É(t7rˆéøñ!ÿýåx¤—ÿÏþ÷³¯ßüÝoJyùê»7ÇËWoþýMÒ¿¾Øÿ|öõËÿü-NAó˜éå“/ßð§ée¤—Þúc–öòÉ×o^óñö“•ÆíØóqôS~ðÉçoþùõüÛÛãQæ(í|ý··ïŽG›³÷üú^à4úãõòŸ9察¡…µþô‚#£–µÍ‡oä¿ë‘Gɯ߽}WJyéå×øï,ý׿Y.ô^¿ì×A×é8ú1^¿`5½~_å)·t²­4mõõÏ×.?ûÓr‰oÑbÔœÑYÜÿw¼Ÿ2·[û-úqT\ùxœGÍòŸ_=Ámʃ£‡¥·hüm;ÇlxѳÞ~‘—uæ×oßåó1‡Üèçx[éÌmnw÷^Ÿª¹]æÿ|òoþ×'oþ·LŒ™¥Á™_JéQêKMc>êxÉgZ^¾ýâåŸ^¾ùÁY4óËxtÀ±Î¢|ŒÇ<ÇK‘;rL¥÷¾úÝëoþöW¿{‹)¥wðÓ;F§ï’¶Êó|©¥ÊèøŸ_ÿñm™ò"ŠL§/¾ùãwoã Û£¶v´—ÿûù¿•~~ý&5=ÚùÒZm”_¾^q>d]¼óÛ7òâ¤Ó²´º¥Uíã±5r`i3æ‰{^rµÒ;+©Éô}©§ÁŽfÈoê:KRš,7ReÃ9Ï®³D÷›wõqÔšêK’H^¶™×öý’õŠ¿ –³l è-3[Úìûõ›’s~̶¼q™²õ‘æöÆ£U¼ñhµ¼ñhoäÎŒuݳ¹Ná²x‡3 ¾ãˆÃv–;ya‚A~ãy’ª,”vãÊ“~L=ä…~Õ;桘kB•¤Ã,vD¿£ÃF‹KNÒî/M±ûë'”Xé³ÉV3„ZÏ;Þ`ŸÒ‘t(C3náê„ê+¼gs•õ–~ð]´• l ±  jUŽqò^&öN1d*"'Â96D.5"j-‹ ßgÙt>ŽLDÌ' r:· (2\}Cj—»%Ò’"Ð8󆜠®† E„˜ô¹!bfëùlŠLœB+r¦Ã$œòD ©3oHIÃzÖŽMZ€–‡õ;º"²Úª!“WêYiC†¼&í·â%™Å¢âŠ‚tù/„ì3úJ{ª¸¯É.CUaŠ”fƒUar©ç#YÏEï¹·Ó«âm‘£ï0¤±ŸÞ9X2öúïÑ1a èsö9l¬ä®´—1Ñzú #M«  ˆX2ÚMÃkP*™Sz{£á\'’Ù±P‡œ‰ðG=ÙH  w3 ïL"§¾Qè*Í:îzí)ïŸcÕ0@À@­Ÿ©ŸbÇp¾É[Ó71Kµ±:!½‘Ÿcu>ª>ç”±T"'ÛÀ„Õ“H#BT»|Ò˜—|Œƒ51 ÈÑž Ñ7,'j²š|Ÿ9%°QP=Tr~tþ»é#%¨¦v+›‡)%î.YvÇшœüÑY9LÒ=2éuåŸ:…,4Ž‘ ƒL˜g@ŠÝæø)2ô¥ßÇÐn+G6CÊ“ÈÐKƒlê0A R‡I]ÍB÷F]ÓÞo>'‡)uNꌓ¤M"vuÙ£ô´‡„¯÷\Žƒã¹U³ð5)Âk•,,=c¯êŠÈ50.ø ØŠÈvÏ_É.¥zD.Þ3Îpôœm/‘-:T²¥$ÌK@»±¥Ë¬˜‡‚@´R±Ž12 ³>V•×Ò´ãù qáRÿ)›~ÅuxÛ@dƒÒÀÔÐ+CgVýI–_ ¸¹Uñ—Nù^€n—2•µ_þ@öû„Êö¢×0m1HÕ_J˪ b[nn¸K Iµ "7Ùïí¦Ûݶ–8HÕ'}“Ý §³úr›lø:H2“35tb âS¤M=)+h¸üSÏüV’þsPGr²¾:Œ•J,B¹xð:tˆd¾u¶‘[‡ÛgÎЛÁ%EüÕá÷:DÝçÙ©ondn^ù”;¢[Ù£xÿ§ìMª¾sÒ ÒeëÐ1šºa‘‰<Э ‡¾ó.säI¬h®^ô>9qb‘Í:T—ÅÎµÜ…ÊœŠœŽe—ÝÕ³¿›.^”ÝLU÷kW'q€ôØv$ó Èx½ãÑ ·ƒ¤ ™ƒ'¸V¹% ´FÏ2 ™mdTz–ÛjìGYi3iÈËîêùé6'†¼KõCÌÆƒÙ}Ku…‡cž< „¾¬* ²]à)†,e¥?ÂçÕ{P‡<©Þà”Ä` ™ØÊZò¬CÝGC¨{Í@žÓdÊ„ÁX 9”Û@o@–÷§9œ\ub—Ç#Èö”t¬¦,q‘åÀ DZ_r9äFe³¨S¶õ©€Ü'\Yr úúàÕ¡š½9‚ûD/0‡"§2¥vÈ¢mì·W U;dëé¼¶lOSì¼ÞlË@’ÍIž¥Ã~ó[åÍÏž•ŸË¸eDpú€ÔÕ[Ì>¹;,t™<ÂÌîdgÅÈÝ%¸$°Õr§ž!«už· $‡Y’wéBâ3NùŽÙuÃ;„’8e6¸r‹ &{— bÐä[g]fYö£ß!H$HŽƒ~‹ëq€ÈcGÂÞuËʃÃlb/•Íþ£wøQãYÅÝvvÃà²OlV@Îb&ÛIœäøy˜}ÑÕ¹Ìk·.óÚÎùżîtb/æu7Ã=Ìëîx˜×ÝÍâ0¯»¸—yÝÍ|½Ìëa†èe^›A±˜×à È˼´/ëz‡,Öµ™‹umgÃb^O³ë.ûz‚‡oöµ°Ë¾–kÒf ûZºÛׂÐB ûÚM‘˾NÉÞiØ×‚P}ûZÓn^§ì†»›×nž\æµ 4•ݼNFy/û:{ò°¯Ý^¹ìk0×ռ mæu¢Ýì¶µ/—m-m7­e>åݶÄìz·­Ýš¹lëÔmT¶–ýMï-Lk8Ñô†yùi-È ‘ì¦uVSh5­åT;vÓ:'3õôò‘wëÑ4ãݺ–»8wë:›0rY×a…y›Ñ¢0¯³™ú—y-Ô¤qóZ~M®çæµ íA7¯Åâ Ó óš?Ûº$cía[—ltÑëRH›Â¸–9ª»Äe\—Ó¬¼0®Ån¢]Æuq;”Æ5¨:¸q-³É.ìÆ5 ÍÕ¸®þvø®¦ô\Æu5Ùä2®kwSÚë:ê‡q-v×AÄëf«é2®[år¿ŒëfÓú2®›['a\·¿0®aÑÑvãZöqÚa\ŸÍ,¡0®e礡Æ5ölƵìT'»qãº7{¬0®±5ìÆ5BY^VÛZd'¢Æµ _V×c˜‰Æ5¦—g×ÓÓ0®§=f\÷Cc‹uÝ] »ã¼Ñ§v ¦ÜAËØ,ll¬\+naÃV¤nv—¹\i›Ýõ<[ìë^³öܾÆfÌiíö5 LÊ"n__Öž›Ø]æ³°»L›DÃ×,ìù²ذþNêf`÷~˜]év‡­wã6ÌÔA“[_öx60»c حؑ(z/n`lS/‹}=R55Ðìë¿Ù×88ñܾpš²Ù× “ˆÙ×°íW´¯q8PhpûzÈSsWpûzÈÚ£Ékæõ~m7¯§Êb^Oœøún^Ï4m1»y=9UózÊ\é›y=›¹¼Æá`ˆ™×³w·‹Í¼†eKÆÌk±3]4óºø nÚ×í­€SÀìk5=¹[›-¦§Kf`·£»TF[I±õbvK‰ËÛíë–rçzvûº ³äåöµÐ?“]ÌÀn0Ò”ü¸-‹/ss»eÈb`ËþÀý: l5šå4°±•šnvIÙ~ävÕôðØ!àB°‘•-mº™à<¹Ã\ª`„åF NK¬£û"!åÅ¥;ýÉ0Šo äÆ,K¢ÜxƒØÐÎ;ïPÇŽsÛêöcø¬,<ÙuqâD¼%p1;ÇäLfÔ~_,éGMb‚¿Ü“½¸3©ÅÅH¸Ø‚­‡‹-p±.¶@ÂÅHxÙþÂpdñ²•Ͱ†ƒ­lôÿxr­=ùÕŠ1ðð«à~µÂ¯VŒ…\~µ@œû_ˆq÷Û÷¿çþî¶¿¸ÿ…8÷w·½qwÈ_ÜÿBœû»Cþâþî~¿¸¿;Û/îï®õ‹û›'ý¢þîI¿¨Πþî7¿¨¿ûÄ/êï>ñ‹ú»ü¢þîᄄ¿û·/êovÞ¨¿{³/êïÎë‹ú BûÊ©¿­nÔßÓõ—7yìÔßµ¥‹û»‹Ù¸¿kD÷?ºnpLð¶qÿð÷w)'kš´reVê¯ ÌX©ÿåç çšk)—s-¹EÎ5ð‰¾rÿPA.çZøhù†tй’(ßp®¥ný„ÍåˆË¿†4¼²õ¤®ì?T„˽–qyײI~—w ÆÿòKŸVox׊ۯá]“ëñe…wMV -Ñð®‰Îéhì&7ß§³˜Ó¤˜ÎþO®T§þ'íKçý§[;FûáqäqÚÿ"x§ýByh9퇑ZÏ•öÃ{H³Äi?D8úŸÂ«Ê{®´¿c–¥•öãÉ̓ç^µ¬™3 퇗۩Ó~µòúJûû™l ¹W­ƒó–…öËÜÜnµ£XHA¸Õd´ݯV† îWÓÔÌ‹ö«™ÄwºúÕÆù²úÕ`ñæ—Õ¯&Äýɯ&s= ï?Rq™ñþ£Ò%´DÞL£ýI†h´•ö'ÍùXi?'ÆJû‘ 3vÚÐ…+Ãé'f)~O²ß¦G=åQµZÁQ¢aB’\Õ®"„FÆ­âÿùö«Ÿð«7_~äÎÖ^~óëŸÒ˦×}4 Svû„ne6×ÄdÉÒdÉYóë§ß|õòÅ_>}ÿ¸2&o{ß?ò(Wâ¡Uk£#kæaéCÇ¥•![«IoúÒjº}ɈT S׌HC¶VE–†LÝ¥•!¿(×±`Ó{C®c€?%×ÑÝ E†Íž,cR¹ò^¯V²´âq¾4r`ic{æÒ(«ÕÏ“ˆI•Y]ªêô:©ßå·ïdçÃù"oõ]›œçkÖÿ„Gëµ^sÜs+¯™ç¹•ëÌ‹V1ó¢Õ2ó¢U̼hµÌ¼È ™™¡ËÌ‹V1ó¢Õ2óÖ½ÊßìiÎgAÈE½#c‘ÎKô7oÉ€´Í2´vÀ}9šå ½Ã¬¶ÍÊpK ¹åh0°qK 9s4— v‡@f9•EFkwd‘jN%& "^oKÄw„ŒXd#sŒ[n·Ý %@Þö=cñ¶}Á3ïÛh,!«¤Ü˜¾'KX6Õ;"¡,}ï¾Whé{÷ÍB沘õw¬bO=»ï-õì¾ÁRÏÊÌJÜnË"*ñs·Ä1‹HVó#5)ªzæh¤3,ÐhEh¤£Eh´ 4R„3hD„B 4Z =b¬:f`´)ÂP$Æ-ÃŒ¬aFŠ0xÊÂŒ©Ìôa˜†"1ÌH‘¶Æ)r®Y<Š0ðÄ⌈¬i<Š F1ÎH ÅgTT+c,ãŒaʆÅ)’y-Æ)‹3R¤®qFŠ00ËâŒav’ 銬::1Ψh@Õg¤ˆ%iœ‘e3RÄ‚“g¤c„f¤@gH㌱Ä3*xû["–´ÃP#EØ„¡F XŠCá4¶P#EæjTTH|¹"¨k¨‘"œæbP„ãk.†¢aZ ¢‹A‘²†)Âq1ƒ"|éæb(¹Å_ÑÅ ߨ¹¡Le.†ÍÑÒkèeP„ý ÐÍl®†¡Ób„èmP„b£[|@:‹ÔáP4ž‹MÔá €åÑáP Žk""u 6RÄjÔáPTZx ƒþÛrèo(æâÍ™¿A‘¹Æ ï"¢×ElBƒ"”„ÍßP4¶‹NjâŠðÒæo(ˆí¢ülþ†6¯±FŠl±F±]}Mä)*æÒA‡CAl—¥ÑáP ï¦5‘§ ¸‹OaEøöÌáP 3ÐÀEó`è Ã¡@k"O9‹kæp(ÈŒ™«Ã¡hˆ{V‡Cѯ5‘§ ‹¥9Šy1ˆ‡‚‘) æp(§3y1‡CQ¹yu8”ÓƒÜáPá5×LA\!5‡C9Ûa.$àeÓá H±h)ó8Ò,È<‚t“ÍãP f×5§ œ‹Mèp(ˆæ¢ÀE£¹V‡CA4W^£ ¢¹¨^›ÃAbÑeæp(çâhšÃAW”Íáv`9*æp¤š`o‡¢>þŠ1m]u6—CÑ£fu9øx?p9”®Øåq(8ŠèÅ3CQ?ÃËåp(»áËåo(š†Cg‚úJÇùºFœ^ô/ÐÝPв¯É<2%ëcs:üx‰5*ˆÓâfiN‡‚Þsgèt¤yžN2ÍçL»×¡à \SyJïÅÜlæt(]vDzÍé }ô8„eqÇ5C²C–5§à”äšt­lxÀ¬{Ä”®î; Ç¡à¦Ö8£‚¬ÕßP†ÌH‹¢¿¡Œóð^èo(…µúÊGâü¥¿¡ŒéÉ“æo(ó° A÷7”‰ I„þ†D¢ï€þk|ÐQ î†2q{t ÐÝ òY§G‚þ†2»±÷7”9#8‰þ†zƒ!¶æo¨âr™ØCC=d†èüvƒìQ'×Ò‰<Ø.óÆ«åñ@Õ:êŠåñ@Pèw”[°4žÛt=á(ÌõiÅ'Â\GÄín®§ƒçûe®§ÃqÜ\‡ÝÊ 7×Áê×´ 58©’„½»u·×a¸²·×á$p{=ñ^ÜX‡5Æ¿»±ž,\þ²Ö“¥ÿ^Ö:²ˆig»µ^L¸¬õòXó‚0 ?¬õêI?a­WzÂk½Ú{¸¬õæ™8a®7/æú¨jYÃY³‚ÔÍk4 "ÿËTv¿—©>ÜT;]^+{q3}šVúlžäVú´%V:ªD˜ymf:ªDԵ膣‡‡¹™ŽÍÙ3Óa¸žìÙÌtäñÜLGY[㋎xY¬ô^ú^vC,TË. +½×èÆ¬ô.=CÝJï2ÚlbFº¦Qã¹¶JÃJï^4&¬ôa1²a¥O7ÃLŸ(H3} •A|f¦kU ZØf§Ï© ;}xÕ7ÓÚv¬vú¨–‚vúhvévú8OK-r;u&,ÈìtØÑœn¨ ¶ávC}Fy"7ÔgŽ(C³Õgñ”:·Õgm^»ÃlõÙLz[}Êdáöé¶:Ά¼ÚêB'MƒŽèÀG£ú,:V1ÇÊ£ùßsµÖë¬"x€XòêŒD±|.;L91C*¢SµI˜¸–è@˜x•F4Ën ²lµÖ+Üè[XÕ Y€ÌPrk½¢Ðkf¿f­—#”ZëØóf­ÃfÓ·Öº 1ßúÝIAØ6o©sm9A7Æ(°˜ù}‘2–dt_‡¬Ä}_–´tŸ{˜IK÷õg¥¸ï )ð$¨ûnÐ’ nt°£Îõ}šŒåTÝvV–úƈ–¥¾-äÁ«HßÖbYd÷EXæW k4ß%5z‰æ»b¡¼¢ò}û–åÍݺŒQ¯øÆšõŠï ÿ²zÅ÷=2Sï i´úÇXȩܱ­þ±˜¹bÊÜ!]²þ1 ¹VoèÏ ß8$,/\Ò|ÞñÄV_øûÏï‰\JšºF%Q”Y¢’ò“¦éȃDMî’5qY3€5ÏÇZJxBÔt 4Móö/šf ¡iž¦y]š¦#—¦iÞ©EÓ $4ÍÓË…¦HhšÔuiš„¦éQ]W’Gu]!HŽ\!HÕu… yTׂH„ y ×ƒdq]W’Çu]!H×!HD’z]!HèuÅ ¹„~Å ¹„îªf(è¡j†`²fæ!k»¦‚yhšäšf(æ¡i†bšf(æ!k†b²f(æ¡kzWÈš!˜‡¬‚yÈš!˜‡¬Q^!kz”×%kZ”×¥jz”×%kº^¾D ^)",Ìk‰@2½üŠ@:\ùô$}¹"L-_"Ì®\"’—%ò$˯$¹B²§IG’—— ]I§¼v·s­%¬R8£B×Ôðä—U×L‚º&ärJ”®k¦PC]×LÕk»° ¹üeÕ5S¦®k¢Ü0/îºfrÅ"tÍä¢jèš©{y_×5“•“ YS&E 5Qœ‡ˆËšiÚ@…¬™¦Ý`Țȣ¥hi²föªª!kf/Bºfvý6tÍáF.lf 0aSþÙwa3W.ra›ÙZíHKgva3Û¦Âf¶½æ6eEÔ5åYs*¦!mc i³¤'i¢9çcD •â•‹<©x}Ùˆ@’ý×*{’<04"Jßëi…Àˆ@ªîɈ$œ=T)=©z,]D Aƒc$“G Éÿ2%"jß“žU5?¢¤°ykÒ³Jæ{’If<‘©›èž5s\Ý„džVu’yZ“žU2·nLÝ„f^7uwIéÒÔMÍkfaª›'jú¾,â&Ds^ÙÅͳû"squ”ׂGª™S'tm¢¹Uý5mSKtQ5m%ºÚ‡„]sÍ|.¨­l=›¶‰ ê¹f> ÒLum9ÕmÍ|.=G,©Ë_,zÉ"‘ ›scðH$-ɵf>ËJO–›í‘H]æ¨I™ŒDê2ÙØÂ‘À"êˆÔ›vŠ@¤îa}ˆÔO+Qá‘Hýœ®uZ$ pQì´@$|󑚤"õaø\Ú,]ÞÚ’÷\†×¬pe³Œdß„te³ ¯`xcÞó÷§>?²Òõ'ìŽÉ{þþ_ýœ¼çèå¯Ë{–ÍeA0# 3Dÿþ~‹¢ˆG>_¿øüå‹/¿¼þùÙúÍØû2Îï}¨ ù»~Ö[Aþðo~õÛ%Á÷gJ·€Äà5¡z$‹¿þã·~ÿéïÿðþüÏ-w¸`§Ð/ÖBÐ Ö@N;©~û¦(™—V²´‚ ƒniÈÒª4 s_Z²´j˜_9°´uS×{¿¥Õ(Ê–V\­ø.N¡…J RÕo‘Ñl^oÕq‚̵•![+ðúz®­ Y[iÉ…Ö—VŽl­rÕzieÈÖªZGcieÈ_•• 1¤%~Ë·ö-߇g¥#¨AcŸÌ UPÁº¶œžûùØ&f¨]—Á¨S¿½ F´ŠÁˆVË`D«Œhµ †·ºÃ[­ƒ­b0¢Õ2Ñ*#Z-ƒ-)&Š×±8¦òbøD!ÿÀ–´$Ç¢-jã`Ë7Q—THÏJ”ê/øq±X›";¶VjFÉCºëPdDK;"8¤éZò aÜ,âÛU?ëJÏ9ÝVõ«öÕ >­1‹Âï:ÇR¨à…ø×¥U5+EkmñgËVÀÉŠ)ªo¶`/gµ%%¿ROVRéB¼b—Mœ2Gµ2ÑÈkç·%/ä´bM†T”}dê ÑäIµk.™(yG†•¶„»éŠ *mGº•ù …HwDv´:vÄ‹:‚æ'dòƒ†"lñØ‘™­j×… «UîøNÚÚ¤dš¤[¥¶@sÓ7ÄK?-ÈÞ ú÷°.d>úöSÓÏЬ¶²½Í98•/„_™XYHmìHß÷!?惞v¤ÇT6„¥x6dª+|A²$^‘’­ÊÙ… +ÜH¥“}A@™ŸñØ/¥cúaµ¾.¤ÇÒ2Dc^wdڇי羴*¢÷©\ WmolkŸÊP Æ2¬€_ bïSåDŽ'¤?Me# +r¬dµ ^/-ž­´Ý…ÌÇöŠØqû“Ïýï´Ï¤õiUÉ’yšÈ5õ§UUsyZUòFù 9ŸV÷‰\«ê+ÒÚÓDÆç!Ó2¬Ä~ ½>Md|cq_Xut+8¾ ´!í8žVƒ¶½©ìó¸!¡p{_ ÕØw¤ä§‰Ü¸²½¯VµÌŠ´ô4‘á$8·wÚÎútF´~<­ŸO «òtFÀ“°Ÿm¶§……s²ùY¤ é(uŸÇúY¬´#§•D $¾fq!ói.Ÿ§•¿»2ö OsjHÚÞ jOˆ™Âõbìò†ŸnuÉžyq!óé€f‘¶†X¿ö„Œ§µ…l´íe m_ZýOKËJέÈ8ž–$i{a}–§ÉÜç|šÌã8­Äf ¸Ÿù4™Gn,Öw!4¦.dI¶rZª*•…—²úýËEKå5W“pœ–"FQ•Û‹–ž Ž V:½¶^°R$œ±0³R(MÆn•Î>-™ÌYéÍÛ+XìÙX)’Ò²qPî€`ˆÆeÉJ'œ}¤Ä¤¥²ÙwÕœ– bª“±Ò†Ô6ãÑd¥ ªÖ`¿d¥‚TT Žò0²R±ö㬴!GNûqV*ÈÉ~œ• 2솤´A.³nÌÄÇÿZ7$¥ Ž놤´á/+' ÒíâœTë vJIÅœ9h/9%mâN^›”´!…¯ñW¤¤ $r§¤r"dΧ¤ !”*å:%dPwJ*¶¬•’tJÚ ñ©‰ã”T«±+%m¸/ VJ*†a~ð_d¤¨ÿí}0RA'‰3ÒwŸÀ© •)¾ÎH±Ôg¤-ÉQxð~ÉH¢@ù#Ò†‡í¼”é.é`:¦óц§W‰Óù¨ U¢ÓQÌ@u:ÚPãN}¨NG|k¼a££‚؇)œŽjYúÁ‹“Ž6}‰¼8éhƒKŽI:ªÆ©Ö©7:*À`ì«ÓцÀVþ„l´áÝÛoÈF9ýBd£ £‘øPd£-áV6ÚàüãCm1˜u6*ˆe‹;mé´Ð`g£M]†+Ä\âÎF¡_ÕÉhKÝ|>¡†%¿w2*ˆU`p2*HÜÉhÃ䆀Œ6eJF°%£ èaÿT2*üÉ*e8Äü–NFµúo×Ȩ ¬ôà\´é”Ó‡6.Ê/ ¬\´!Þ8óWä¢-»Oß¹hC õ48mÙ=¸ÎE©;mÙ½ÎEB®Ôà\´eOMw.ÚP©¯³ ¹¨ ÕöY㢠k`.\´e/äàeFôÆI`R§+Å^²qѦA~ú ÆE[ö 碂LoC.Ú²—[p.Ú”l×"UÊn‡\T‹ìp.ÚàH^©hC-¯M&*@f’‰3QA,u™¨ ô…8¥Ø”ˆ(dÚfbD´! ì\‰¨Š­ìG‰hCX߉¨~d‚`DTaCgD´i&{!$óÃ2ND ~ž+m"ã¹àÒqî&}9mˆå~mD´åaîD'¢‚$›¤FDö"ölDTŸJFDJGñ•mÈvåÔ6"ªßÓàéfD´åyØ~gD´euþ(B"*H! r"*Hc†€цëÓˆ¨ ƒœˆ BW¥óІ@±òÐV\¬s*H±ÏxhC>ß©ñІZ?®Žöi5ÊBEt‰q]SGq„–Míüu¡Ž"»“{¤«£FÒÞÕÑÞ‡Ù-¦Žö ÐÔQ背ÏÔÑÞ­u¨£È5+#U^0VF*ÈémLígõk™:ŠXœ‘*ã°;6uŸÐ³§2u1=yå¤J\*eê(òg9Æ®Žjl>º«£8ËI—\%"/4q!FÜù]Ey)Z&Ž‚Xq;6qTã”Ø­‰£½X xˆ£H &5mT£Œ¤RE¥+»´i£Ý3TCEÔÔ±i£ ‚Ü]EŠ2wh×F}ÅYêÚ($¶sÓFA2ùÖ]íÙŠ"†6ªž_LÕ1’ñ™4Цie¤‚4;ò\E9—F»ÞÖ¥Q| r%¤úq¹ÒŠ˜?cæŒâƒ’ǦŒ" ­¯Œ”Ë»1e¦<6eq}e¤ú9K^+”Q£FÄ”QyµÔÆB…(¨ï+”Q†ÝþõãáOq­ê×½ÝõâGôS‚¿~¼“¿*LJ?ñƒˆ DÎçÿ’o^ü´'¹Úip/?`¡7ެ:AYtÙZ×oCŽláE‰Tm /JNÞ–V´f½/G¶Vµ´=pÈ‘_8®€o]üÄÀ¡ëÇ¢‰ÎEF´•µ,-mÂclc¸"Ú®V²´2⺴ dieiiÈÒŠGñÒÈ¥íøK£@–V¶±,­¹ZýüC%ýþ¢¾WFìªб%ÿhøÊŒ×Ä¿È?f’´ëyýKæ_ÐA{•=÷Zž#u­‘ZWM´ŠU­–U­bÕD«eÕDX¬šˆ[VM´ŠU­–U­bÕD«eÕ¬;±×¶e Qúy^UeõÍ•ãy0 ‹òE™Ér}\¾ÐBÝ\g‰5ÏÐ ¨_h-϶êZÞ…Z@öùN†ÈWKypõâªëê…¦!˜VAõB“sœe¦«”ìý¨z¡Ymu¥é¤˜æ9KVÃöð¨°néJŸº«Q ÈÅ ŠÉ§Ç=\ x±!JÑW„âÅŠP¼ØÁ4± ±ÍeA¨^lHgRÓ…P½Xªb9,BõbC&«ü_Õ‹¡z±!ƒI˜|±"ÉÓQ.¤3IíB¨_¬H$º\ÈÞ Z–₨~±"Ô/V„úņ Nä ¡~±"Ô/6¤ïÛPìúbƆt›ÊPÁØÉÜ ¡‚±"T06d0YõB¨`¬Œ ýâª`¬Œ é±² ¡‚±!“i,B cALÂØ•0V„ÆŠäô4•MÂXJ+B cCúÓTöóxA(alˆ¥‡]%Œ ™í…QÂX¹ÿÆ‚˜„±"”06¤?­*“0V„ƆœO«Ê$Œ Q cEZ{šÈ&alÈ`bñ…ôú4‘MÂØÎôÊ ¡„± &alÈùØ›&­ÀܹµK+RòÓD6 cE(a¬%Œ ñÚrPÂXJr>-,“06d>&a,ˆI2˜d}!Ô0V„ƆôÇ”ºÏcÓ06äŒô†xUÒ™OsÙDŒ1»"ãi.›ˆ±"16Äòñ1cE(blH:…†PÅØùtH˜Š±"T16d<­-ª+@cCÆÓÒ2cE(clˆ¥g_eŒ ™O“ÙdŒ¡Œ±!ói2›Œ±"f[ò‘/­¢W×Ä-ÇÎÐ,g¥.Ëáu•˜ V:ð²3c¥ÐÅ:ÞŒ•B‚Ÿ-…´ï~6òR\Ó¦Œ—ŽÙœÝ/Å5-xÌxéô¯#/þõaç¥ð–X|™ñÒéŠYðRÜDY½jrISð‚—ÎbBAL'—àBL£2]Ó‰÷†˜ÂE5ׯ‚\˾zÕôN)I91…ˬ®^5¹S×!˜"JΜhFLçð( '¦ˆ’3ŸžÓ‰Zø«[­À£Hõ×Ýjxš¹†yiywêðîVƒƒ“n w«áp¿âÛØûÕð|ԞݯvïÇkZ¥`%¦ ¡Ã‚ºèXCU>ëGKË·èÄkGsß´;Ö2j˜c 1gÃÜht¬áô5ÎK+îŽ5-öǞͱ†²¶Sy1áž5¼ú Ì±–ðé÷ͳ†X0se™g 5†!ô¬á™Íû1Ï¢ }dæYCÐÓ¹yÖP ¢¯´´¢xS_iiÍðž5„/õ•–V+«g cÖèú£g-§êYË¥î¡^1GcË;@„Q£Ì²H*ó¬!ˆ—îYC¹ýsË;@ ÷ ÷¬¡žÅêUÕCïƒ;ÖÃÓ7Ç"vÆæXË(ÓŽͱ†2ðÇ–veY¸•9Öòt9Öò¬[¨—᱆zU„Ìl~µÂBï/¿Z9¬ÜDøÕüBG‡ûÕêrš~µ’¬ðføÕ0UùàîWö7R*SU)Ñû˯Vàó±È.úÕŠÜ)'­ûÕJŽû1¿Z)ÇSÖA)Ùö̯¦‹`Ë:(ŽƒîX+Åwzs¬é:Y³Šl±»c­`®Á^º”¬yÖJõ0F÷¬•vø•̳†b$v7æYCªô¹yÖJ;ý©Ì³†j[)ÕÚ¥æQ4Ï ×q)ºg­œÅâ̳†"[<©Ý³fKûýåZÃçY͵†oÔš#Í\kµÌwê®5l+'­¥ûÁhžµÒI¸k ì›k­tÞs×¶ O3×>Vb¿2×¾zh—2×>ƒÀ&æYC9‹Ñ2Ï·§÷—cM8çñV†G»c­ çîX+Óê }O¨WA=§-í  †5pC+öq§õÂioq\ê…Óõ°YÆzᜌý)Ê q}$8®’¢Ä÷#WI'„ÝšJ:#úÄUÒ9|w•³§YURÄãñ¤r•tžÝ,WIçiË.TRLÝ⺨’Îæ;¦«¤ZGh*é¬Å{6•Õg¬gSIQ7\ÄUÒ™çNG©œ´&’bÍ–%í@Vu‹ .Ф°Ó&’NOe‘oœ7.’¢H9ŸÉ4RÄ·‘¹FŠhÍ: :3‰ö-Ãä]"Å77‰¥éxÇ.‘|ÚŒ=›D:iÍ~L"(ÌCiéÀ|ù+“Huæ%Ru·_™DŠ3ë™éÀ÷Ú–(¯‚x2»¿©O{”WAè˜å˜DŠï¡ñE¸D:ò±GyÉñc‚AH¤x–¸D:P}nå¢ÑØ$RÔñ©÷E‰ÌcòÒÚ?–a©žy÷E‰¥öyÒºDª'%“Hûpî©Fpñê&‘¢ ¢gP"í}ÕG­e™ÔGQßhK€-Ͳ€}ÓGˆÕ, ‹ú(¼€–Þ`ú(>@°2QýŠÞ¦ŽvÿŒE¨£ÐzøÐ®ŽâSfêh¯vŒ…: úb$ÓÔÑŽv¬DTˆÌp²jê(ÊLÎ5ÀKˆê7uSFâLÚÕQL¥%ÂK?Kh±YTGã;¡Ž"jnê(t/kcꨒ»5Ä« Ê®dê(Êiqª»:Ú#ŽÝÕQÄ>kˆWA¤“åR˜:Š/^¤5Äëúdc¨£øNÆ0ÚIu‘Nɨ©©£ó_‰¨ã¹ê(>оÆx±ÜØ㥟˜lkÖ~Ú#mêè4þâè°y—8:ÖÔWù§…‡0:|&…0:ª_Å…Ñaê.atxÞI£Ã>s £²w•]í>ÿBíÎBí—³ÿ–x£ïx.㥨¶È¤ï«íõã¿ú9á]?ÐË_ß…«äÿOµ½~ÒC}1dÙ‘}qÌ­òÀÇæ‘ÃgÑjÿVîPî~ö×?èN¹êëWzûNv—ã3½~ûU¼Ç(ãõ‹·ï`ÓÊ9÷ÚÿûÛw"ǯÿø6£ nþñ-è:ÞdFký4{ý†MÒùúGô%³¾~ŽMX`Ïü¥pð^¬É|ý´Xï/aK|eۣ˭þ Û6i°£ÑólQ_ÿfµm‚(¸þì‹·Ñàs>êùúß‚|ÒðÓ÷h9e‹yýÓÖ?6â’äv¿ã„¾~ø’—=ú£7ö_”°‘óõÃ×oAOaÔ½þÛÛw¨ç9Úë‡o…eRâµjk{­G»^kÁ]Dtqb$Ëö»ÏôA‘n%¯Uà>Çlò‚pªO¡:¸%ŽôùŠ’MTNî×:½!PÈF(óé“Ïe }þñwñ^‡KÞÛX_Ëï^1AĸëkƒõÖ¾Ô›¨3}olû»·þxíõÓo– ~¾?[?½dï÷þÎoþV”]]ŸAG_Ë)ÕßõÑfÝqŽ òÕm.ê½ÿM¦Œn?qk±1u<@Íæ6ƒ¾Õ—}uØrË­Èý^¯Ø¼”/ØC™2XsŽ!x§¹žiʳ#ŠñÌmn/ð=~‡Oe¾m›y\Æçvï_ýË[¸…ärëÕ—­÷^m4zÂàkO p½ 4þ>zËåïXT¹e6~Ã=()öTÝᄹ 1tqzü¼½¯¨ï|A^¤ØÙ®‹Ô…k^é_¿Ò‰þw¿©uíá_b¡úÄÇdŸC–†N‚ŠPy矢—k†Ëƒ¥4^1ðËÃüÜÀ×A{}z€O—­ásÀ)û³íœ +˜#^G­û"ñ¶s ËŸßÆ¯Ö²ŽÛ8×=ò¿-ƒü=»Ëu½ïØ…œ6©tŸù +B˜]®™ë3M}F®OyÙúªuYˆñsœóõÃWoñG¹TÁÇæ<[ý þÛfåßês_+¯¿Ò[8Ná³:µcµÊ˜h âã|ýä-ºoçYþb‡†ÞxLÙHñŽºXEëÆõa=ïø¬³=OcU³ìâ_¡;œkç¶Gÿ'ÓhrJ~å(ô4½þžû¶‡¯ÞógB,åÍ]7§ë¸lëÛÿô3lBg:«Ž1ä#y¿¿¦ÉzqîØý”«/›“mY#í'ú²uÈ.©}·QWø;{cß·ÔùŽÎ\ò:‘–+ë&{mìõ›mLâˆùüz¼ßó ¬SÞ¦‚}`—Òã çÃyúª}Ú¯À®åÅ~ä–=oþì#†’¬†Ÿ²úü‰8íí&uÚ7Þ©m²òS»i}¸i$½É:Z7ÙOÿ¬³HìÄ´¾¹kFýéêâkvQ„‚}únéƒd¡ £ûð|8ygQ»bW¾>e|Â8f„aæ‰ÛxLŸÊd–Á–£æ¼ÍI9c÷e㺬µíÌÍ0xäÞþó­¬`9F‹í°Xr ÿÀs€endstream endobj 257 0 obj << /Filter /FlateDecode /Length 3012 >> stream xœí[YoÇ~§ÿ†…!@³ŒvÜ÷ÁX,ËŠÈ@"íƒ%”DQ¹Ks—Ž%Î_Ou÷ÌTõ²ÈYÆŒd‚ô¹Y]]WWu Y?ND+'"ý×ýûìxçËGZOW;br¸óãŽÌ?tÿ<;žÜ™…M+mQNæ/vÊV9Q*´Úɉ·¾ÚNæÇ;Oš;Ó™hUŒ!¸æ`*Z­Š®Y”Bx(|•ˆ} Ñ6 XÕÁKçšýéLkÕÆ`›?ê ÎØÐœfZ!Løçü/;VµZ55€Y˜ÌŸï4ÿžÎ_ïäó4 &ùÙ“æ xé”pØ,I½WÍ­rš¦¹7 ºUBy8 ¤ ‘0iã„3 Ös²ÿ4‹h”Z@RhšgSå’ê¡yYØÛ(›‡i›ŽA[—‰eJ¹f•Ìœ6¦Þבô/9êed ª€’ÞJfÅ è<Ó¶ Εÿ•!„Š&›¨B“å0Bm:Mµ2ºéÄ‹ÆÛîÈ@Ô{UŽ|Òü£Aæ…Ü*JnÚàC”=ùµ–8}\ëq`>Mäœç]ÇÂC„=³Êx¿@‚è¶þŒðá+–€ð…»“œ3ÄðeθŠ%¿g ~F›}…«3\%G¬ÙÓˆ©ï!^°ÁÀß&ÂííX:;aµ'ÌV,$´$è ßâö”ˆ—è€,äÛšõ&I5„ï!_†¶Ì˜Ûe’ï³&á9óÑ›»»wlÝ—ÇÖ@ÏÔÕ……`á.Z¦‚lÅOü´ÚºÈj|ÌBRlÉ…|û›|qiõBƼŒZߢZû6af³‡Å7vY!éá«×ÀMÃ] ±Öã+XkôÕ´ÂÈø¿¬Ã!ÏÇ„ßgãŠUi‹G)¾Ð¢ÚáSSEjšýF£B¢ä±ç¬ÿŽØLHbï!ÂØ[ôˆM•$d‰84 ?¾ú3œFh_°÷úšýJÑ»e8m÷ºÁÙ}üùrŒyÖu·æk娶œô~ßg}7ä[™òLy膻 sÂr'qI ¸Ž†ÏžH{Â2;×e~P`œDÊ' ËbuúCH["'Š”­¶zø…ÅŒFo_>:Gh‰7iB%eäïÛlDu)dæT때ç7h¸Iìí¡ž•–2bËÆžÅ€»ƒ«~‹ðœ çoÍÈ S=]½°+PÛzºÞ®€/¤7$‘¼æY\½±ØÌÄô;Ôx[–¬tƒ žs‰««zü](­‡lwÉÿ;”¼%ø*BÞb|Å!Ÿ6³ävµHKd8eá’®ÛÞ!ü‰•Œ¼¤Î*!ûôWö4šHéý!yTý‚°O•¡íð[g.,òß@IÌ>c ˆÇiÌN§•BVRQ›s•Ƴ«Š…¤˜ñU‰¬Æ1f–®NáiɆ…¤^~…ç@VÉ«y†®”~µÜÅÕ 2ý{Š4öü¯8ÈU&>^°´giSÿíN†ôcÖÓcŽ8Þ²G:ºmã9¸)ÙèØñÉÛ=—HkÈ]„$hÍ‚ËëµÎoÃë ^#þ[ŽºS°¶V¬‹TÍ·s'D‘®>”ì6ÇLVG=+¶÷ì¼ÇÜCx¥/Ò[Zÿú›±V 4bvâNhq,Ј“ {D`ùz “Ë̲§ñ1ç®!…0ŸN®žLêï"Ã#m†ñVÁ¾ðؾì<ÆŸ¾b Aòj%oàC‚<8F?Vá:¤Á7R¨6DsY˜úQòG{»È¾‚<ûî“Îö=æh ¬úröPmZØJ¤ÅÞE–éÏH­.åYô[ëqð*´6\j°Ê p&Ùà­½ôr‹Þ~{yíõÈÛþOå½È¨äκ~ïÍr±`@öWpªõÂwÒ”?ŒÕp¢põÇÉßÞ–æ\*0€ŸúÁëêÉjÝV†î?ýî:BRgù«ê†¯—£ÏRËé*÷âãùF–YuÚõtä9qÕnà÷ÐûñOgÍšWo*Ü?Ü.w1ßñ0°íý½õ+¾*F«Ü #º­ê˜?û€­a ì–ñ³x£ömmÆÌn©8ôè8þIÎwšü…è8ûQ\Ⱦï&n4¤ø[_ÝÃíCªú’ÇÝ8ˆÔh!â™É+ä>ºü¦9bÊ»H›4ÃeêwzÕ£uÛ÷hŸZ³O­ÙxÑúÔš}<­Ùæ4 iÕº¿¤,¿žéèຩjô2O}„(…o>›¦áPåe=ÓQ¦•T¦Z^L¡4R¦áÌ9Äò×Å.:!LI·[VÇ•Ù-cþë®´Í{ét‡ÓBç PHÿnƒ[¨CCGfÊÉ0º“‡PéÜh'º°(ºÉi¬Û´\¤R ŽtÐv`µßÛØ<05Œ ¬ 3#ªmËÎî)ËÐLžQ!Ó18”ÛÇeÒô+Ê[¦cœÒŠN,½íì%Uxµ¹/¨ L5 DGzö³¤RNönb¨³ésõ! Ýö¡IFl)´÷¹«|ðy™3Ú©ÎÊyœ9{:@€„^ñà•n²íTÔÆËZ°¬ƒ„î_ž–uH—„øenÅ%Ò¯“ûÀÑ¢Ì:&²Ôy³^ýº¶n1¼èç•m?Œl|g½<ù´Og§³g ¿Òöܦyš­jƒŽt8j=-bÒlœ†ôÕý,¤ÀÍyÊÕS" ãƒrœj¼9+—4¹à&ÐKµO£íøÜmÔ†ʷE=H;'yªÌÉ *~$íØí#LIœ«†Ñ5›,‹öS2—Œ>ÒÈ\ ‹§%'˜òžUÀËšI²ïæ;ƒÿþÿ@N½endstream endobj 258 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1à Üy…@‚Te‰XÒ%C«ªíˆ1C2ô÷ „tèp–Îw'Ÿù0^GgðGôø¢Æ:iõ[D‚‰fëX+@[L••‰‹ Œ7ÞŸ@°Èü®âOqéʪ=Bè5­A!Eåfb}ÓÈÞÉÈé?©&S¢“‹ÿTr4—8on1’K¥ii’ XG¿g‚9;ØAìSlendstream endobj 259 0 obj << /Filter /FlateDecode /Length 2509 >> stream xœí[[o[7~Wû#„"@ŽÜš÷CM€^Ѱëè!‹v¼¶ìuW²\ÉÞ6Á¢¿}IžË ¥‘ŽÔÄ›8kä!_˜áp83œËùë31äñOó÷élpx¬Ôðb9àËÁ¯‘þwØüu:~5Ž6Œ0ϽŽÏõT1tbX™Šye†ãÙ r4þ%žsÏxeÄñÙà§âËëgÊ;elq=*93ÞW•,¦aX¸Ê;W\(WÖ§‘¢¡>¹ ãÜIǵÂ4ó«€5—NÉb9*•RL.óóˆeàïŠOÑBWi}•¯Y Î+îŠIÍB‹bgID²5m 5ºø šv‹–XD §¥ŒÌ:ù—µ<Êg¢-"EŹŽ+sf¹–^¬ 69 nñe¢õΛ¨‹Žs_eYYÌG¥´Ì» èYÔ–°ÒøLºiÚ•q>[æïãƒAÂ-[‹–Ê0gmm×ã0ß{«µ/žJ# zIRpîÓ6x ð À$ÁïeË×_Àh £ç0:%FÌf[V3ÎùâçðK€€—ß¼¸ §-[]ñÀqß´€7g$í¦í–88‚9À z÷ßB¼xP¬·à'£èTB1£½ŽミŠÏƒro—É?kœ¼Á£ÁŸÜÇÓn.ÈiO€ï)HŽ”v‹ÍIÙcjðš´ÀÌ(÷›“؆K’íÁ/H|4Ö̹bCt¸æÏH]OIs~G_‘Ñâà8X'ç¤_ÏÈÝ#Û¿&íùЦ½$±$iÑùý «º[mNîâœä‹ì}CоW0ØæHNȪq¤o.N’Ç(£CªÚ²àüÿ÷ç}¼–§¤p\¯l,»Ð³SÚB±bFB¤P0y½ær ¿ã|kÔ›<<¶Y¡&+f¬vm”ý6{pº ǧ]˜6íôsrÎàŽéuM+(oÏImÎH{L±¨ òr7m?¤íÓ¶”¡â •¢2e#yHÙ+¦ì¶4#:#Wž’Jù"íÀW_<®#¿ª˜´®üðÿS€3À ÀçOHâK’[—² ˆnšÔš®Éœƒ˜-Iˆh¯"¾OšÜ#tìP›cã\½sëÎIHò0¡`ˆø¢zh= ½£~æÀïIw¢9 °s®à¤>Ã*¦ éŒûFy×d@¸$µ° ch&@{M2[bKt¢>Èáp íÒy i¯š3.´¨]ÅÔ{Ey{pIÊ Ø ' *€¨–УÐáë j8t£ˆƒ‚B_€üЇQÐô-t†ŽªYöÙ»ìÛ»$Õ`ÃÞM%âÞ& 6ÌñÞ{6¬ð.W6|x,Bµ–WmÉÍÅãÑãµ¢Î0«*!Úø9%tûGýrùa­bèpì÷õ:A’ôm/E¦Œò­(%öZ½´ŸöÐ"câ‰Òê|ž‘¾Õ†+Y%…_)`¾Åý¹ÞS "¥¥ iÁlÒ| p¯Ž}§\õ¨è”­1eq¯SZ’y—ß#°hr M ¦¥D$m¯æù‹L9õg* .Ùö6w›|Êv5\æ•à™lýù% ^’Q ù(ª{QÍ ¥@uJïOv™?wá‚Óm¤à’9¯·µ‘±£µU§ê`ŸAš}ÅDLÛ»w©½á2ËŸä¢J3é]ß–P“~,7íC0­äVžõ>ÄÎûØ ¼t̸­ ËìN$+\â7¹Jìðu`wyÍ»‘—Ý©¼›”1Ý-§ÇõÁ s$™“¬âͯŸg¦TáynóÏœoßFÊõ6²ëoîa#ÙN»ƒF²N“ﲑ4h™u]Tz§‹G½G͉.ÝB QI–‰t͉ġӻÞÒF"fw•ÉïMïH÷xô§„ Óþ©öŒì›Uâõþ|@/Qmrµaä$ÌšËûÐFÒŠ Ï[VÒïnÎM%M@™S“Kd%%dÌhkÕ}sb{Nwfïûpº;øÁµ‘½N‰‚5yž¶»Sf_9)¿—XÈÝ;`úg“{ÝÆ»ˆÛÈåP ªª ŽØt‰5³¼ŠÛÖ'f¸Ƈ†ñ¡aüXÆÕW9Yi¬é~Û.•ôá¼ÙæéG¥ªæí OB>AÏ\š!^W&%7Å¥:¾öZv7‘³ gÕeOeÚ·&Âã1‹i=±ò¢ø-P8n¬UÅ?Ñ š(žp^ð*Ä™ÀÍ„y‘ N³aé“ô\È[¯ë‡+ÍK QZØÈÀ¶ŒÿïEXyYOÓ)ÛÂóz8ü£ø%îÔUÂÚÔ*6¯cðž®kâ`Ðø ©ž(ðó,ü¤&PN·Â›HÐ1>IbX.C ¼˜€tÿHгÂåï†nFéƒ@Ï¿EnŽ‹ªyÔ®Ö‘^µbšÈ¬#9‰ñ×8«´®…óLG)ºgSg­f«°Z};[1í»˜^#Y烙ø•Ró"Ë„©ì¹Q.SôŽÜ%ö­äNkKðû®Qú""¤¨åôa+Ù°•÷Je+|ž%?_lìZõ("YÓ*5(Ϥ…T˜íöPôЪfaWeã§y¥B³jŸ+oC\Aºþø^L½ÇËЈøî.€w°¾Dª¤aÁÅÛ#yÅ!º {ƒG©;¯èò*ê‰ö{¾ÑÁövk#Y~§ =Z» º1´v§rýKuçuÇ š[>«}´²ÁÀW¤hÇä¹AkFkâm_N­\Ã&ÞÝõ…ìì1P›"˾ï»Ù½ `·Èyçw³³þò>?¦BW5·=¦2òíS¡›ÍÏ©þ¯sz·ç³>¾'Tþ¿Ûœþíxð×ðç¿UËÌüendstream endobj 260 0 obj << /Filter /FlateDecode /Length 160 >> stream xœ]O1à Üy…@ÂÐ,KºdhUµý1&b B†þ¾ÎÒùîä3Æëèlþˆ_”ÀX§#­~‹H0Ñlkh‹©²2qQñá¦Âûv™ƒßÕBü)ºKYµG½¦5(¤¨ÜL¬oÙ#9ý'ÕÀdªSt²B`ñŸJŽæçMÀ-Fr©4-Mrëè÷Lð!§`ûBŒSmendstream endobj 261 0 obj << /Filter /FlateDecode /Length 161 >> stream xœ]O1à Üy…@‚TµCÄ’.ZUm?@Œ‰ˆ¡¿o ¤C‡³t¾;ùÌûá:8›€?¢Ç%0ÖéH‹_#Œ4YÇZÚbª¬LœU`¼¿©ðþ‚Í@fçw5ŠË©¬Ú=„^ÓRTn"Ö5쌑Œœþ“j`4Õ)βB`ñJŽæÇMÀ5Fr©4-Mrëè÷Lð!§`ûBSmendstream endobj 262 0 obj << /Filter /FlateDecode /Length 4930 >> stream xœµ[ÛrGr}‡ýûàè‰ Zu¿èºp½ R+‹Ôê܇´ ÒRt„Ý'³ª»²z¦I,ÃŽ  ÕuÉËÉ“Y5¿nÔ¨7ŠþWÿûòúä‹­Ý\ܨÍÅɯ'šÿº©ÿyy½ùêFxê³ÊzóìÕIùToŒI£ z}³õ›g×'χ¯¶§j49§†ó­}ö&‡á-šZ©¨’l¾¦Á1§ì‡ôÚuÃn{j­sòßÑkRp> ·ûÏ^ßi‡mŠ¿=Þa.§ƒñ‹ÒÎrŒfxPV‹É ¶ÉŽF™ˆÅ°…½ˆÒi‚ Û:ßßò1‹–V6¤áåÖ:z.Ëô>ëáñ9}fs²>ð`•1a¸#1¥`ë¿›– óBKAêÀ8=â(ÐÛx:ø·ÏNþƒT´¡ï£UÞ˜œô¨ñµÍvôas{¾ùysóQ}g³‰#i´)[GÎçjT)’²C«â~´ ñNJ—¨=Ö¤.M]0m²a6ZŽõ£ÖÎyÙÇŒ£Âh·zsŠ£ÅB·–½ÏNvé8¦¬E—N£ƒ(ñ=í5vë[å›î{—hʶûàF«iû!°@Û­u¸#Ú~Š^.ïpÈ´9õv ÑÙk·!¸• ¶-ÏŠeÕ€ÐMÄÖM‚*?)ÕÁd,Vþzûÿ§éþ|¢£±£r³E\Ÿ8K·­çêäé‘Q­Çò²4ʘò×6ªõˆQÉÄQ˹ZOMäÑÈQ­GŒÊVã娹§ZžHž±ú\£'oL2©íi€cÊÁ(üÃ'­æÁÕ˜Pä?ÿÃØž}jK.ÍZ5z%¡×ž«p¬šÃQ>ÀF[€BÉQSO7Ê`?ÁÈQµ§å Dݪ=ݨºW1Jìþ'aŸ’Ô5 ßâ\±Å‰ˆÏXlΛÁ–¶Ã(H¼ö»8x1&p³ ¸ÈHK̶?õ„ŒÍžù`=ÒcþIkîÙØ‡WåÌÖ4|³¿;ßì_m~}÷úfõú漂B•àRYÞ*âFµ¿ïß]ÿr~KÝžoMÆT¿msLq8G tjx»»y»y¹¿Úß¼>¿£Áw»«ëýÍùÕÕ® 矟@–ìÑFbFí¹:ÚCZ†tµ…‘èûö$ 58¼ò~9gZ:Ä>÷ë1n جÌ={&¸ VÛÑ…{ö<=yI#i aniówhþ¬‰VÒR²¥ãêXG‘™§ÿªtßž*5Ÿ ðäîÛ3‰M!¶Á4ï×3‹Mg6ºûõÌHžLý Fþ_ĕ›ýlÖ"‡ßYPäA–×ôZyëêV”6rºhý×b¸sÃëäXäPܵ%å“§ÓÖR`ö¦ìýK© à%ƒË&ICÀ´ôצÔræ…Ì®Í ´¶ˆ îeó”óF˜4Y&†AP <-2Ä9G–”»ÙÞ5ü|# ¬z—2>Ly‡ È¢s>VçŠ$Á³iA#Å.œv™À‹:“ ëˆAxÚÿý!mr”ÞÕ'ì=Vj (@gö9 OÅá^6 ºÏÌQEkªfH$·äE‘c_Dh»;ô©´CìE ùbÝ‹a†ÇGîò2‡º;ád à’P6D Y†#Pó)RÚŸžåºWÀL·š;ø¹ÕÂÈ,h©¬×¼k–à¾ò0£ôpqY”cfA3@Oß°ä³} ]PMcBDFHDƒádg îHÂ>ùÙÛdh_q‚EžàÛIê=âB$åÖ÷Wód»†Mò³ËfA’k€cƒ¦Ê+‚ó +]‹¦Ò5„Hp·á+Ò)òÍ ;d.æì5´½Î¥aìÊ|ÄbWk² ¼ν^“Í`õÇj²ìHrµODY´EÙˆõJì9´sÊ0‰ÊhøŠmâÌïJ7Éßeˆ·}ó’=#E¯ó§cf¢²¦ÅN`yÞæðki7¦9€€³!æ“ÍØà]_ö>5H9‚¦DáÔd§Ÿ0•_ ð 2ä@¹: Ǹ„ïíðâÅÃò=-èe©â2ù؉U`wWeÂø'Û™íþ! ° ¹.ñ^²ì Ô°}]öSz\†:Pc™½hAŠ›Æ&êÝ!£íS  ÝÓ$“‡7ÍOvÍ©Þ7K§*;2.p7ï\ð’çlNþÛÑ-Ì1Ú¸®ôÿº…3QÃÃ[FºFãKž63ê9¤rðŒÀï§àÙ{ãt|á+±ÿpóÄýYÕN«þØ éør;˜é|XÀΤqíÌšµínë% ¢ä“" Æ£I»vʧz,ו!uä%-eÆÚßtNúfäºØLBž "ûV´w5æ¤!iccÍóŠwÜÍ0É"ÈRrr…L˜ª¡¢0é)v…ibbE7"Zœ•jŠBÍàçPwÝ®7m’w-³ãýWgdf ÂÞ-s±=†WÚ`6k&™“!qþ»À˼cÀÔ*ƒ&ŒT®wÄ@ \Ò7°¾Ïr"Fr"g /í/g#Üv .ixåÿ/^²|@XxÙ0à}Ãu`BÞwe‡`Ò®±ÿ¿1JÖJÓ3Ú9³X¦!˜fûz;™Æ°¿n削ž`u™ qž¥W§j\ »r0§dçúÊ]]DÛ2ö#GO] äÐ$ƒŠIâÛ²»7 ’#ÎÓ‘ì?˜d{âZ­Cjÿ£ØY¨œ¨%NŠÖ÷(ZÈ`8¶`x95ÉÎÌÒüj;W¥h¯V[«]¬‚L×-Ò®ÁÈò¤×(x¥²u7Ë`Þ¸Xƒ¢g<2 ôgàÖóÁo5Hk†ùL½Á ‡ ø Á³0 šP1,|&ç³¶ä”O¨ sOîÃJ9Î"ô …þ•?ôÍdg ?ŸV©ýî”逸 £¿j–p/âDÅúJ‰åÛ7ÓoD)a—¼¬¤ð:€—ó0õo·sêý/¢'ø’D·u™ y(¦49€»OÙîªA‡ ówesØÇ4heœ¥°ëJà ½Öi‘#uôz×÷â\…úiªjL¦xVM‘î±PÄÍñ³” H£ePêÃÁGêÁdQnAìÎ+áRðÖMèôÞ/ò´01Ž¢ð§YUmAn:¦µŠµ¦y?æ$b‘*©ÄÉÄzCˆ”{Câ±E<2.(%V»áàeñUw°¢èâÖdé²Ò} ‹q£rG‚¬T¹×ädÖÙÆLˆÆ•_z[Ó±²Z&D^*ªÕ”ùb@ÞH/àb–æŒöâ>H[A@”Þ¤âðxW9üä—zõZh%çWT&Ô˜Žç@æÔr­ r©±Q‚"¯õâÅ"+>RóZÍŠg:-áߨ‰/˜»8/jBÍJøŠUkìÃ-oX§$ê“ \vÌ-A~¶²£ ·M/—&»Ô|ÇuæˆöãÖóÃþ(ÃÑêeñ%à!VAƨ·å{˜9Ϫ[Ó·ðy R3¾M­!@JºK>é å] WòvÝ>hA}Ð" ¸QÙõf¦§ªGÄ á'iĹ~›¯)WP`É› ,»ª~–Í\1›½j芩·4tªx‚Ó² Цxþ3•uî„].vЍ^RÝŠËûòQÁ~ñS ªÔä ¶,†IËWW"ÁUž0ño;~Þj­Ùú(1Ž b(1¦HíáX# ü'QÎxÜžËÒ!Käç‚·ëoÃÑ×…”»D7š~] LybÈÔQ²Þ?6ß´æ—­ùÅÑæïŸjŽ­ywtµÝÑÞñèd»ådœ}_¶æëÖ¼iÍ‹Ö|Ûšû£cÇÖߊ-ÙŸÉá¿Ñôåñòö—·/øŽäñÑ24§9òÙµ¬ªß•Iœv+¦Wišé '}s5ôÇ–«ã;‘~ùÑnWǶ̟J ¢Üï/"mcE;f(Ëö€eÒšŠ7X¯MWÆ*…hþ•¦8˃B×bæ‹Vš˜lóo]µŠŸ9ÄeÌçÁIå.ë퓞™˜­ÜLÕ&¹ÌËÔ×VÐU¥ôõvz2xü†JܨsI¹TI× !ÓÃ:LëT‹fÝe wHÉArÿ]f£zéªn©²aù^sfÊ•hO<}þïY¯´:Cï©HTAÇ)®Š.ÊÝ´we håÆ}åû²ƒ‡§‹el‰k=y™µ5§u‚ôÊ3 1ºË†ë#^b`Ó3eqãùö²`l—ÈŸœ_$­ w+n¹†ÿÿ¦6âí§Yгúë¸ÿW8@;endstream endobj 263 0 obj << /Filter /FlateDecode /Length 3688 >> stream xœíZIs·¾³R¹ç6•Cª'¥ia_t2eY.ÛŒÊ mWJÌaLR-næbÙq*ùëù€‡n<ôôˆôvHUÊÁ xxË÷¾÷Ðß-D/"ýWþ=<ßyü¹Ö‹“›±8ÙùnGæ¿.Ê?‡ç‹§ûi…ÃLE”‹ýW;ôS¹rá­ï£¶‹ýóN¹åþ·XlE³XÄ^x‡ìí¼ìv¯–¢×1h뺫åJô6FïUw†i| ¡;ÅPI¡]èÓŠ²z}‹yTFó5—¡‚VÝÍr¥µî5v¹|•Æ û‡îì ‹|¾nÏI[K!¼Ý1madwž~¥"Dr´K­é~¨Cö³;vÄuZŒRi³Qþ’GÇF´ë´Â aÒÉ¢wÂ( O&Ó6ââi¶Û¸ø4¯!Ú¤‹qç,¾†²œê.—+åú èQÒ–tÊÆFº³|+bsÌ?ö?Þ°þj°èJÛ>8GvÝË›!„6Y[2 ¥HN¬Æ„ 6ß½†86ôNªîyzˆ/»ƒ.m¡Oï:…»,³ÅuTZwýr%£ÅÂØY®Œp}²ÛMòà~íÃlFgì`ƒâ?ÕÐÙbRKÈÄ ¶>î„Û…C„Ì>¡…ƒs·¢ ¬Ñ¾{›´„ôÝeöŸ¨%NxCšÂV¯ÒÁKl4‹}Ÿç[HX”ß yZ#«uó´…ŽêE'7RoõòÑy-{eM—«%—óIr»Ë%÷ÒÖà^#>ÜÐ^‡­ù-Ž—ÑÝÞš$RѸÖ2Р‘JCâl€u÷~Ö6)(û¸4Á˜nÝÆÒj@$ÑJÄ¿÷‹•Ô½5Q‘dŸ§UЃC5xœõŽPE4ÀÝ›1ÙH9ár¬iSOL—f-"Ÿ("|;Âý“Ó²ûó’ÂÒ–QöÚÛAc9´"DL†w¹Ë^ˆ±'"S]0­iu@” ÐF?ì‘f_¥YÕ{8êìw·Y;1JøÉR&ñ¾ÀTÖSuŒ@YþÃÓêø‡K«’÷è ‰T“æìFº´ö6ÍÆoy)vLþ|½´1Ã/ç°N×k­>‘¶,|1"> d†År){èYIÎq5Ào”*àóÈu*Ÿ ǺîËÏëM÷ò¦ìï|¶ï´ŸÓòõö,îf³¸²‡C,p2J™ü%:¹ p™Û<„ =^Õá“:|<;|{ß°¯Ãoëðföàõì,[{Y‡¯f×öK r¨¾X‚<2ýÚs€”:©ÃÇuø}ÊÙ![{Z‡¢_[Íù³‰˜tÀa.š>zI6ìg]ù˜–ÈÅJÁïdÉÚ”áv–¨VÉß{käjd9WOHZ”˜^¦áDJ”cd?¢Ù”A>M[˜¤ñúÈü¶;f`xVãôg4D°R$¼Bê%,ÈŽž\jŠÒ˜Þ5ÄÔ$7j©µ4 ÚeLr1(£²‹x¤€`2´ð#)9–§\~`\”Í]¥[f·»\Z€c”)AŽa}<¢ÆY¦µÑš’ ¢xZ7È"J×†ŽŒB·?Ϊt Û]fCÂ\ K„†¿®T%¡³J))Õ9Ìýq°—)¹™v;v+¡8Kæå‹mÃË?J!@¤†¤V&Õs“2„ì¸ ZRNǵ%•˜ô¨:NÂ2G6ñ.FJä¢4^ãw)»ÜÕbóz…‘¾&+'ÇÙ.‘`™ ¡Óêáå‡>‹[jÚ¾8$Bã‹J¤yí|] •†HŽ‹O¸ LlJ„Â#ï¶kg£«¸õu7æøë“ú~õ‚ãû[ŠzDV÷÷qÃRÝ…Ù7µÈ` õ ;(OSì"½:^”tlºb†úhEUoØi1‘ÜeÖN¡…l_Å ðu2¶D<`œFƒÀ‹<®`‘ØBê È|D“a¨2à~(|P!³Š—Ýô› ›M«QE]жå ׫ÓÊÁHW&m_‘ÍvÙ¯àèº)‘X­r¾ëÕ|^ª­$/àöÊ©2É)¢jà§”þ’:Ê »Ô׬àú±*àf³Ðµq†¢:U‘’$&ÉhDi+æ}AK@#ni@ À6-ŽVÈË/‰B¬ÿµ;ÜÌóûn´2Œm aîuŸ’›ä^Ìåiæq°/r–©Emj‚8ÀPsù’!Qa¾avÍÔÐÇß½’$¿ÑBW‹øBQ6Û…Â¥°ÝåD9dIÕÎÇå:ǃKR¾®àT“Ý–Ê™ ÔÕ´¥m®MXáwWÁÿA•­¶p/°ÁDøC„g»TÙæÙ2Ô8¿«àzüž%¯¤Ãö¡ˆ^ŒIª.¾!!Å_è/t÷AŽ@­eÕ.»žMŸsV; q˜ÈÂG¥ôL.*ónð ¤,ÒR€Ü6yXAù×xr%b.sIAvb›’¹Bî´f†;XD|S3K‰‹¸wÔÌ¡)šI+²)U·Tþ'# ð'rE9ÕÁÖ…Sºû_$ä ùaQ:@$†ç À¹-¹€õCKµ“1SË“Es¬¼Ï·=8(YBy º&K³v^‹ht0 sïò„5¹9Lmé»ÒouUÄ6+М°Hl;á ?¼¥Ø’ .G4íÉŽ ÈCÚØ#èÞ 1ZÅ^Õ‘ ¦´~9Äh¤y–f}¦Ý6]KKê.¤¿²]yaš#:ŠÑYMF<`IÈÜô¿N'`uhåuõhJÁZÁ×&)‡ P[¸+¢ySñŒ}¹(¯N*²ßm´õÊI(d2 ,@¥Ïâ+¸–Bq@èjjˆÌÁ¾Ï8â„ægSAq‘Ë-;ŸW¾Âœ¯ØRÕœ‘Ô6ºI ”ê!Á隨-p4ê«x‘O?¤¡ÁxÖðzwG ²Ú>¬oÙ`p§é¯ÝÏÍ()çš•ªiV2à•ÞþsL¡úíkÛP[×È hC¹|8ðÎ/üf}AÚ x7l¯œ“jš÷™½rH.䡸Aî3,Í¿–ã1û¢Ã–ÙÓ¹!?©¥1"£{ñ¬(CÄ)wÓ\`½É•ÁÏÏY2a‰Ñ]ß„¶ÓXR!¼š Ê’_— ²X5äë4ÏK™r+x…»"{Û\%¢‡âwì¿ePBO{EQYPãõðÎ:)ýò+k k³’»5Kù¤JeµÝ(fèÍYRÙÕ¾yμÊ×µAG>Û>¥‹¡† ï¨RÍ ®Ï¬DrP‹p‹eúšó·tÖlÆÆ #Ï›c`°åE’Á,‡Å©Î¨éÚ¶€Þ…z+kGôgàßËÎŽý2µƒV®ÀTè'³xø=´•ÞXRïòÐ$²«L\ɳ© g‚OWþc6«ZQØ"ƒ %G™^gïq>’aÓ PtemYh[*‰²üâIýŸÙ;åbýý¿¾lš–änVE7<§â“ ÙçTωròÎÞzÂÞÑ«]“ `ÖPÄ÷ÑQôuÿI (ŸŸß3ÙŸ^–÷çÜ;a¾Ÿ#9ó \¡äee©ÓLª.ôX m ½ÕÚéOBÆWSž]¨u"sC®Ê¤áêPgÚ9ˆt¯ö6&IÇÿÖ\çš öïv/*G;á)ìmŠe¼›>'­ë "ša k²·ü˜3ÊÙló½ÓÌÚ9¥àǃ•m“Ïž/%>±—Ý/&f‚–€ŽYž)|”6½S$Ó+Û[^Ô[rŒ_ÕÈüÉÇq½ÄæWFù«-ÊÝÉtÏ®ü}Ò<[o£ÓþŠŒ`M¡Yke©,`ÀÌ9¼˜xúzp4ÐíhúÈÿg=”Ú™ÃöM”æµäõr¨™Ø×:<=–kLdO_§¸0çÝbâÕë\›å¦Ò-Átþ¢‡îL„0SȦÁ˘[ý\æOƒ#wàÉôt;ˆÒ\.#otïPÙØºbÀ½Ë#œ2Ÿ³CåKá4ÍHå3’,Piœv~PY)o©5œ³r1ÖGÒ$©ßàa_ªœ::ª^jz~º_œëðvvÈÞ¬/êp]‡}~X‡ìÝünö4¶Ù{uH/ÿ¹­÷¦ïê°¯ÃuÞþšwóå8ÙûÔú?ß*§XAŽÊ5 QÖ»o$ì{·Â÷8~4ŸÉÕÀ£·w òûºÏ3ÉÓþŸÛ~×ÜÆ¾ucOiG´ Ù‰Eú³$½óé¾Zn¦¹¢ÚäR½ªd—¸¥QHeWô…Þ6¡u°×Ãg–Z&v–(_Ìv3Bv¯•é¥.÷0ï}AàÖ~“Ó³Òp@¶ñÙì¥ËЬ~–OšïF»³¦¼ÑÁQRk£e$“Wx8±Z©LÇ›æƒ öÐÉï>|9“ZƒÌ~‰ß„.P=’B_âŠ5ٱϘƒðè"ÎÏ1Ї§¤7ý²äýÚMaûþOä<èEœ`}.9±ôÆ’ÓÑ}këðõl*ܛݗíp3Ý,'2öÉÝ{uøM^þ²yxß‚_›Aá)½±…_|É_(÷~{óbÎÝžÉÌ0Od®êðI>ž¾½oØÏ’¶žÝ°?û¸2OgÍËLvV‡ç÷Ùÿ¨ëðqþ{vßãYq˜¼óþʾ]ülç¿îµ˜Vendstream endobj 264 0 obj << /Type /XRef /Length 250 /Filter /FlateDecode /DecodeParms << /Columns 5 /Predictor 12 >> /W [ 1 3 1 ] /Info 3 0 R /Root 2 0 R /Size 265 /ID [<3a4faab3992dd16470dfe4db3eec2081><7e0ade8ef44e949a650d6739cb5d2367>] >> stream xœcb&F~0ù‰ $À8JŽ’$“ÿ„êÙì@iéùóÑ´4J’Ÿ–”jAiI6”Š$¶€Hyv)˜"ùW€H…S ’“DrÙƒHÆÕ ’—Dò,‘,®`Ò D²jH¦8 ÉÈ7¬þX¤DJ>‘Ì—Àâ¯ÁdXlÿM° °ø0ùD k€HY$]¯²|À²g@¤Ð$q1° Áæ‹€Mλ<ä61°¸à.©ù DŠ—‚Hö›xUl~c*k f¿@ˆpOGW# Š`[5ÝÇṪì#a°«ø71"“6Ž endstream endobj startxref 205652 %%EOF flexmix/inst/doc/mixture-regressions.R0000644000176200001440000003122613432516303017635 0ustar liggesusers### R code from vignette source 'mixture-regressions.Rnw' ################################################### ### code chunk number 1: mixture-regressions.Rnw:63-73 ################################################### options(width=60, prompt = "R> ", continue = "+ ", useFancyQuotes = FALSE) library("graphics") library("stats") library("flexmix") library("lattice") ltheme <- canonical.theme("postscript", FALSE) lattice.options(default.theme=ltheme) data("NPreg", package = "flexmix") data("dmft", package = "flexmix") source("myConcomitant.R") ################################################### ### code chunk number 2: mixture-regressions.Rnw:500-503 ################################################### par(mfrow=c(1,2)) plot(yn~x, col=class, pch=class, data=NPreg) plot(yp~x, col=class, pch=class, data=NPreg) ################################################### ### code chunk number 3: mixture-regressions.Rnw:510-517 ################################################### set.seed(1802) library("flexmix") data("NPreg", package = "flexmix") Model_n <- FLXMRglm(yn ~ . + I(x^2)) Model_p <- FLXMRglm(yp ~ ., family = "poisson") m1 <- flexmix(. ~ x, data = NPreg, k = 2, model = list(Model_n, Model_p), control = list(verbose = 10)) ################################################### ### code chunk number 4: mixture-regressions.Rnw:558-559 ################################################### print(plot(m1)) ################################################### ### code chunk number 5: mixture-regressions.Rnw:598-600 ################################################### m1.refit <- refit(m1) summary(m1.refit, which = "model", model = 1) ################################################### ### code chunk number 6: mixture-regressions.Rnw:605-612 ################################################### print(plot(m1.refit, layout = c(1,3), bycluster = FALSE, main = expression(paste(yn *tilde(" ")* x + x^2))), split= c(1,1,2,1), more = TRUE) print(plot(m1.refit, model = 2, main = expression(paste(yp *tilde(" ")* x)), layout = c(1,2), bycluster = FALSE), split = c(2,1,2,1)) ################################################### ### code chunk number 7: mixture-regressions.Rnw:643-648 ################################################### Model_n2 <- FLXMRglmfix(yn ~ . + 0, nested = list(k = c(1, 1), formula = c(~ 1 + I(x^2), ~ 0))) m2 <- flexmix(. ~ x, data = NPreg, cluster = posterior(m1), model = list(Model_n2, Model_p)) m2 ################################################### ### code chunk number 8: mixture-regressions.Rnw:653-654 ################################################### c(BIC(m1), BIC(m2)) ################################################### ### code chunk number 9: mixture-regressions.Rnw:672-676 ################################################### data("betablocker", package = "flexmix") betaGlm <- glm(cbind(Deaths, Total - Deaths) ~ Treatment, family = "binomial", data = betablocker) betaGlm ################################################### ### code chunk number 10: mixture-regressions.Rnw:693-696 ################################################### betaMixFix <- stepFlexmix(cbind(Deaths, Total - Deaths) ~ 1 | Center, model = FLXMRglmfix(family = "binomial", fixed = ~ Treatment), k = 2:4, nrep = 5, data = betablocker) ################################################### ### code chunk number 11: mixture-regressions.Rnw:705-706 ################################################### betaMixFix ################################################### ### code chunk number 12: mixture-regressions.Rnw:713-715 ################################################### betaMixFix_3 <- getModel(betaMixFix, which = "BIC") betaMixFix_3 <- relabel(betaMixFix_3, "model", "Intercept") ################################################### ### code chunk number 13: mixture-regressions.Rnw:728-729 ################################################### parameters(betaMixFix_3) ################################################### ### code chunk number 14: mixture-regressions.Rnw:737-750 ################################################### library("grid") betablocker$Center <- with(betablocker, factor(Center, levels = Center[order((Deaths/Total)[1:22])])) clusters <- factor(clusters(betaMixFix_3), labels = paste("Cluster", 1:3)) print(dotplot(Deaths/Total ~ Center | clusters, groups = Treatment, as.table = TRUE, data = betablocker, xlab = "Center", layout = c(3, 1), scales = list(x = list(cex = 0.7, tck = c(1, 0))), key = simpleKey(levels(betablocker$Treatment), lines = TRUE, corner = c(1,0)))) betaMixFix.fitted <- fitted(betaMixFix_3) for (i in 1:3) { seekViewport(trellis.vpname("panel", i, 1)) grid.lines(unit(1:22, "native"), unit(betaMixFix.fitted[1:22, i], "native"), gp = gpar(lty = 1)) grid.lines(unit(1:22, "native"), unit(betaMixFix.fitted[23:44, i], "native"), gp = gpar(lty = 2)) } ################################################### ### code chunk number 15: mixture-regressions.Rnw:769-775 ################################################### betaMix <- stepFlexmix(cbind(Deaths, Total - Deaths) ~ Treatment | Center, model = FLXMRglm(family = "binomial"), k = 3, nrep = 5, data = betablocker) betaMix <- relabel(betaMix, "model", "Treatment") parameters(betaMix) c(BIC(betaMixFix_3), BIC(betaMix)) ################################################### ### code chunk number 16: mixture-regressions.Rnw:795-796 ################################################### print(plot(betaMixFix_3, nint = 10, mark = 1, col = "grey", layout = c(3, 1))) ################################################### ### code chunk number 17: mixture-regressions.Rnw:805-806 ################################################### print(plot(betaMixFix_3, nint = 10, mark = 2, col = "grey", layout = c(3, 1))) ################################################### ### code chunk number 18: mixture-regressions.Rnw:820-821 ################################################### table(clusters(betaMix)) ################################################### ### code chunk number 19: mixture-regressions.Rnw:826-828 ################################################### predict(betaMix, newdata = data.frame(Treatment = c("Control", "Treated"))) ################################################### ### code chunk number 20: mixture-regressions.Rnw:834-836 ################################################### betablocker[c(1, 23), ] fitted(betaMix)[c(1, 23), ] ################################################### ### code chunk number 21: mixture-regressions.Rnw:846-847 ################################################### summary(refit(betaMix)) ################################################### ### code chunk number 22: mixture-regressions.Rnw:858-865 ################################################### ModelNested <- FLXMRglmfix(family = "binomial", nested = list(k = c(2, 1), formula = c(~ Treatment, ~ 0))) betaMixNested <- flexmix(cbind(Deaths, Total - Deaths) ~ 1 | Center, model = ModelNested, k = 3, data = betablocker, cluster = posterior(betaMix)) parameters(betaMixNested) c(BIC(betaMix), BIC(betaMixNested), BIC(betaMixFix_3)) ################################################### ### code chunk number 23: mixture-regressions.Rnw:876-877 ################################################### data("bioChemists", package = "flexmix") ################################################### ### code chunk number 24: mixture-regressions.Rnw:908-912 ################################################### data("bioChemists", package = "flexmix") Model1 <- FLXMRglm(family = "poisson") ff_1 <- stepFlexmix(art ~ ., data = bioChemists, k = 1:3, model = Model1) ff_1 <- getModel(ff_1, "BIC") ################################################### ### code chunk number 25: mixture-regressions.Rnw:929-931 ################################################### print(plot(refit(ff_1), bycluster = FALSE, scales = list(x = list(relation = "free")))) ################################################### ### code chunk number 26: mixture-regressions.Rnw:938-942 ################################################### Model2 <- FLXMRglmfix(family = "poisson", fixed = ~ kid5 + mar + ment) ff_2 <- flexmix(art ~ fem + phd, data = bioChemists, cluster = posterior(ff_1), model = Model2) c(BIC(ff_1), BIC(ff_2)) ################################################### ### code chunk number 27: mixture-regressions.Rnw:950-951 ################################################### summary(refit(ff_2)) ################################################### ### code chunk number 28: mixture-regressions.Rnw:958-962 ################################################### Model3 <- FLXMRglmfix(family = "poisson", fixed = ~ kid5 + mar + ment) ff_3 <- flexmix(art ~ fem, data = bioChemists, cluster = posterior(ff_2), model = Model3) c(BIC(ff_2), BIC(ff_3)) ################################################### ### code chunk number 29: mixture-regressions.Rnw:970-971 ################################################### print(plot(refit(ff_3), bycluster = FALSE, scales = list(x = list(relation = "free")))) ################################################### ### code chunk number 30: mixture-regressions.Rnw:981-987 ################################################### Model4 <- FLXMRglmfix(family = "poisson", fixed = ~ kid5 + mar + ment) ff_4 <- flexmix(art ~ 1, data = bioChemists, cluster = posterior(ff_2), concomitant = FLXPmultinom(~ fem), model = Model4) parameters(ff_4) summary(refit(ff_4), which = "concomitant") BIC(ff_4) ################################################### ### code chunk number 31: mixture-regressions.Rnw:996-1000 ################################################### Model5 <- FLXMRglmfix(family = "poisson", fixed = ~ kid5 + ment + fem) ff_5 <- flexmix(art ~ 1, data = bioChemists, cluster = posterior(ff_2), model = Model5) BIC(ff_5) ################################################### ### code chunk number 32: mixture-regressions.Rnw:1006-1013 ################################################### pp <- predict(ff_5, newdata = data.frame(kid5 = 0, mar = factor("Married", levels = c("Single", "Married")), fem = c("Men", "Women"), ment = mean(bioChemists$ment))) matplot(0:12, sapply(unlist(pp), function(x) dpois(0:12, x)), type = "b", lty = 1, xlab = "Number of articles", ylab = "Probability") legend("topright", paste("Comp.", rep(1:2, each = 2), ":", c("Men", "Women")), lty = 1, col = 1:4, pch = paste(1:4), bty = "n") ################################################### ### code chunk number 33: mixture-regressions.Rnw:1362-1367 ################################################### data("dmft", package = "flexmix") Model <- FLXMRziglm(family = "poisson") Fitted <- flexmix(End ~ log(Begin + 0.5) + Gender + Ethnic + Treatment, model = Model, k = 2 , data = dmft, control = list(minprior = 0.01)) summary(refit(Fitted)) ################################################### ### code chunk number 34: refit (eval = FALSE) ################################################### ## print(plot(refit(Fitted), components = 2, box.ratio = 3)) ################################################### ### code chunk number 35: mixture-regressions.Rnw:1396-1397 ################################################### print(plot(refit(Fitted), components = 2, box.ratio = 3)) ################################################### ### code chunk number 36: mixture-regressions.Rnw:1442-1449 ################################################### Concomitant <- FLXPmultinom(~ yb) MyConcomitant <- myConcomitant(~ yb) m2 <- flexmix(. ~ x, data = NPreg, k = 2, model = list(Model_n, Model_p), concomitant = Concomitant) m3 <- flexmix(. ~ x, data = NPreg, k = 2, model = list(Model_n, Model_p), cluster = posterior(m2), concomitant = MyConcomitant) ################################################### ### code chunk number 37: mixture-regressions.Rnw:1451-1453 ################################################### summary(m2) summary(m3) ################################################### ### code chunk number 38: mixture-regressions.Rnw:1458-1462 ################################################### determinePrior <- function(object) { object@concomitant@fit(object@concomitant@x, posterior(object))[!duplicated(object@concomitant@x), ] } ################################################### ### code chunk number 39: mixture-regressions.Rnw:1465-1467 ################################################### determinePrior(m2) determinePrior(m3) ################################################### ### code chunk number 40: mixture-regressions.Rnw:1509-1513 ################################################### SI <- sessionInfo() pkgs <- paste(sapply(c(SI$otherPkgs, SI$loadedOnly), function(x) paste("\\\\pkg{", x$Package, "} ", x$Version, sep = "")), collapse = ", ") flexmix/NAMESPACE0000644000176200001440000000763113430477461013165 0ustar liggesusersimportFrom("graphics", "hist", "lines", "matplot", "plot", "points", "text") importFrom("grid", "gpar", "grid.lines", "grid.polygon", "unit") importFrom("grDevices", "axisTicks", "hcl", "rgb") importFrom("lattice", "barchart", "do.breaks", "histogram", "lattice.getOption", "llines", "panel.abline", "panel.barchart", "panel.rect", "panel.segments", "panel.xyplot", "parallelplot", "trellis.par.get", "xyplot") import("methods") importFrom("modeltools", "clusters", "getModel", "ICL", "KLdiv", "Lapply", "parameters", "posterior", "prior", "refit", "relabel") importFrom("nnet", "multinom", "nnet.default") importFrom("stats", "AIC", "as.formula", "BIC", "binomial", ".checkMFClasses", "coef", "cov.wt", "dbinom", "delete.response", "dexp", "dgamma", "dlnorm", "dnorm", "dpois", "dweibull", "factanal", "Gamma", ".getXlevels", "glm.fit", "lm.wfit", "model.frame", "model.matrix", "model.response", "na.omit", "optim", "pnorm", "poisson", "predict", "printCoefmat", "qnorm", "quantile", "rbinom", "residuals", "rgamma", "rmultinom", "rnorm", "rpois", "runif", "terms", "update", "update.formula", "var", "weighted.mean") importFrom("stats4", "logLik", "nobs", "plot", "summary") importFrom("utils", "getS3method") export( "ExLinear", "ExNPreg", "ExNclus", "FLXdist", "FLXgradlogLikfun", "FLXlogLikfun", "FLXMCdist1", "FLXMCfactanal", "FLXMCmvbinary", "FLXMCmvcombi", "FLXMCmvnorm", "FLXMCmvpois", "FLXMCnorm1", "FLXPmultinom", "FLXPconstant", "FLXfit", "FLXglm", "FLXglmFix", "FLXbclust", "FLXmclust", "FLXconstant", "FLXmultinom", "FLXMRcondlogit", "FLXMRglm", "FLXMRglmfix", "FLXMRglmnet", "FLXMRlmer", "FLXMRlmm", "FLXMRlmmc", "FLXMRmgcv", "FLXMRmultinom", "FLXMRrobglm", "FLXMRziglm", "FLXgetDesign", "FLXgetParameters", "FLXreplaceParameters", "boot", "existGradient", "flexmix", "group", "initFlexmix", "LR_test", "plotEll", "refit_optim", "relabel", "stepFlexmix" ) exportClasses( "FLXcomponent", "FLXP", "FLXPmultinom", "FLXcontrol", "FLXdist", "FLXMRfix", "FLXMRglmfix", "FLXMRglm", "FLXMRlmer", "FLXMRlmm", "FLXMRlmmfix", "FLXMRlmmc", "FLXMRlmc", "FLXMRlmmcfix", "FLXMRlmcfix", "FLXMRmgcv", "FLXMRrobglm", "FLXMRziglm", "FLXM", "FLXMR", "FLXMC", "FLXnested", "FLXR", "FLXRoptim", "FLXRmstep", "flexmix", "listOrdata.frame", "stepFlexmix", "summary.flexmix" ) exportMethods( "EIC", "FLXcheckComponent", "FLXdeterminePostunscaled", "FLXfit", "FLXgetK", "FLXgetModelmatrix", "FLXgetObs", "FLXmstep", "FLXremoveComponent", "KLdiv", "ICL", "Lapply", "coerce", "clusters", "fitted", "flexmix", "getModel", "initialize", "logLik", "parameters", "plot", "posterior", "predict", "prior", "rFLXM", "refit", "relabel", "rflexmix", "show", "summary", "unique" ) flexmix/data/0000755000176200001440000000000013432516322012640 5ustar liggesusersflexmix/data/Nclus.RData0000644000176200001440000002062413425024236014645 0ustar liggesusers‹]zy4•oÔö1œsg0ã¡ÈTÆŠzv$•¢TRh ÈŠ"…Ê<$TJƒ” i0Üh@$SÆÌTdž ïù¾÷¿÷gXûÙ÷³®µ÷µï½¯µn«m6Ú$@à&p \¼œW^n΋@á|Žîç½ "•óa7ª¾ŸéÍ çËÈ—m´CÂ|*yü#Ø÷ïÖ³¦/±±0•{%åjè§ãDr.þ±>„ð}ëW"š§äb ®G—Èa?ývÉßmÐâ…±½'#‚°QÊs{IS'Ô­Þ•~º aÓfjGÎîÁZßßw{TŠuœov¢eÕj¢þ“Ø”•“ª{dŸ¸CN}Ò °‘;¹ëË]èÞÓ‹Þ{p?hØŒÍØ¹4_妠þÁol:î–Ìm発!áäˆý¾ü‘ùñ>ª™°X)..=nãe}~4•¢ì6‹}4ò6 «÷ÆKõ£ •W±ÙpêLµ Ö_’u‹úô)ZR\UôÛûF{ß‹Fu¤K+x°ÏMéÒ1vƒ¨ |ºž Q±·ÅBéjq‹è[µKîþ{;Kœx.ˆM^YÊYà2ÃfLN˜â;…z×íÍÄ JÇnxP…gr™ßYA¬^ámNý[ÎâÙ&qé4ú£ñ3þ@6žP×=á-6Ó÷"K»û;63ÈÔݺ6kÁŽšZ›… Šõuf¡ 5ØLcàã ßhšÕ÷ÌÄ>M9´<4xgþd(<’¿Šõð_wÄÃMx5¥/Ç>FóõCU«…Xáì:ÓcïuÐŸÔ šf 2”Öýøƒµä{•j:ú¢‰MT»Ø=Û“V*_*P¯ùbMÖØÛUVMÁ–‚Ë"\Ãæÿ–)|ز‹¾}s»MÈú“âöº’ϵn)Ø]4†U¿p»ôWq ôqΣ F±S܋u¶åDQøp⦵¨ÙFöñxéV´ÐPàCSǪOí•/ínÂÃF»m;бUìEÅC®elºEºqKÓ!Ô²çì¬IwÖÁsëÅÁë¨k¤¨ú{¤&x ½\j…šëuÅ R±¡¹M1yÿb°ß^fK†éÂhæ¿^òß}/Q“ñ£ ÂKhJ½®5C÷7÷9º”Ôº¤%2ö%ä¢ù-\Œ+?ß ÚäZ†»n<°»J8 ÓaWDÊÙ¸"½EËÈ.l+ 0PxÕr{§æ`»Šk]VLõgÆ#BÒ ß=–t¨¡¤¯µ†eLê‚´9?#}9!Ö0éïF`J­N‘~Ѐ5úÜÍmò¢úBÛWAŽ?ƒ[•øØ·jB¯ž‚\uEÂÂöë »v÷õ=Á‰À>S¨ÞÒc’–ÍŸf‚üª®Ik'°¹G¾ÍËõ‚Ü¡9e;¥{ µTÚ}³ÿ7HÉ)OÑH¶Àöqñ°Ÿ%÷sNɧnŽQ~ú }1‚’DZö÷ÍŸ2¬AÞ¶û¬êá>`'o9ÿ« ô­Ä³,@^äÞê€_9È4~áŸËy2­I¢'„{îyÌì7Ò<ÜI“˜vªM}ÚÁ] à´vîx7 ØTÓ?{Ê|AÚèÛuå(ߦîºýd¹¢êù@~úÆqi)µ{fw*Èý°1Z©< ìÑ*ÇÓÉ›@»dOžjÈÔ~R±©÷ÒAæwf‚l È ÖàKe{÷†éfaN>ÈÎî<ÁÁ_8Ó¾þŠ'°÷«¼.]a\{…œ¿8¬¡ñ”‘‚4M+}ËØJSùù@þ ïûç[G@Æ5>YãØ]õcÎþ ÿݶlRVäl½ÔœºC€­x°+µØA:Á!I¥O¤Å‹S°ö¥ß¶w·ìAaÏ¥ÄÚ÷í=ð ›Ã;¶d·À7ë@áQÜt;'Ž?÷Ê. ß…Ó?b¥šö€Tž¼¸%(¼]µ!›Bé-Þô5œü>Jkȳز-qå3#@vGׂ¯HŠîk~›¤rÖ¡*÷½ÄAþzË¢û£E`¯¹¸÷ÔȰ­s›nû¨»ÇýÉ =—#çuºX¯t¯,¢•ÉÆ•«§qZ --X[ 8Õ¸ðxqÁoÏ/¾Ú•‡óß®]ŽlÄñ“Ÿf;£ÐëÒØO.4tFûüéEmìÊŒš5À‰Çœo_ÅêZXÛ²¼qò-“æ0~\¨¸?a߃œ¸ä]ð(-(ÄuÄ`óJ¡õ‹¥¿qŠ·ÿ]Gÿû¨ºïh€gÏq4’¥0Ľ-o4˜#Úâ"ÓÙvOÒÐ4»Ì±Öë%ZQ8sïη»8-áù¹êÄ\œ­õs´àšžM,V/BS½IÖ‰æqÒ™— *€æ3_nUWÿ5æ¬N¤òÅ¡A‚̱T4ä[qð¿í8É@ld-{û!{/¹í}0-ki DZ–ÐnçSš6¸`¶hÚÃÞ´ÕÚ4ç¾ͽylu±Ùõy†*ŸáNì¯Î€æ;Ûm8wÀÓ{¯q7œî/™œxÜç yl9[»„s‰0KM: pÞÂ?F=ƒpQmVêÙ§V8!|Ï/Ä%å W|“Žâ܆—w)hà„ûäbƒ×Üh‚á»'³l-ZÞôY—E¹Õ*6ËTã|‰éÍÎU‚¨Ëiý.M[´My‡Ã\Xó®¥A¹5N÷øæsÍ -MۿΜ´C=7¼«•*ÄJd¢œƒ{зðmà‘޽L•“^khõÌǤÕhƪõqŸί^îo£¢ƒ žbìg_AË•¤æ=hèÖ¾¢œŽ)4f21)(Wˆ&óÞ.ÝÚ»Š•¿¶+vûÍáGAŽÀ%Ý8‘ÿ󆎇h¼óÕ1Ã7hÅJôô%žm¨çv GSelaï\¤ñ4šÓP9ñaJçyÝÛB[o&[d$Omfጪà5ظ€@hËíWùhu{nuD›Zø¾F§¡þ+NY]֘߈æz¶&>ôÂùŒcjüzŸà|¦ïBN×[£vøZy-–VíþŽtŸ{‚p‹^§•…–‡¦þjÙ…VG'O®Ü€Ó|ŒŽ,g+¡ ï•­ÏÉ89"φ`³ððp(=z -ê¿bÝÔA1c%57sX7ÏwÛOY¸ §óÈJcRêfàç2q·ôÈáEsšÊeMß¾ ‘üS)E kÑèN©Ü[Û†p‚Ä‘¿ ß3pá=Ó¾1¶u8]þ/¡‹1èc{»ã˜ N|å–w W§Å…XíP|^p×J]ÀyƲ{>‡&ãe-–I‡pa÷¹ ÝYM8Ñ1ößO^nÙ0™e®+8­Ÿ´7Öa Z¬ÕJ:uŠÓŸª-‘ÐlÍâJÑÎWï‘9úRD¬£¸4Ÿ€ ÉÏVw1 ·‡äo”9 tIÅâê`>;B >/OÕ^ §=ŸÖj%‰xSû=ëî–Ÿ‘Êêz¿]C¦¥ rëÁq…J`”è'ÞîÕÅf½ï>ñ`0%žir{ÆûƒuSB®TÈ¿ÄVˆƒ€OÚ»ò÷Š4¼£¢´-Œ¸¥9`:VÑ¿=> ´›mü'¥€°Æ"/Èú Põ÷o˜vĦ‹ßï Ž˜úLíle$];¬Þ‚¦Ï´Õ|ñ³ÛϤ®ãœeÎç÷ù@&UMû bÔ»½™qg€t»1¥³è%‘IðØø¦¿øHèÞ>/[†ÏnàS³ºö(Dáǹéá r&m÷ry<šÏqÝÖ!‰¦ÄÅ‹ ^F±iy Õ—ªWj@€êñèêÀôÎYË ¢ü÷ó”í @î)”ußTCæÝwåR@}ûl¿ñeàþyT*Ç(eƒg³¼ÏuÁ-œjï ŒÓ1¤–§ÊÀŸÝž¸‹“/;÷Á \@±1®;z 茶Žqåà¹£â  B¯4ŸÆÖ8S­¿Àˆ­ë¨ýƒC= ¸[äcj±§Ž]¿¯ò”Ç­ñ½è§t#Íh=wÉVt¶á2oÊéê¸^ ­ß1 $`è=PË”“Ö€šðè›hà»É‡Eœ€Æ'à\4icµFö= jò$EG[êïûÆD£Kõþ³ý¢ªÎÞUþØ¿øŒ½v£ Eþ ŒÿÖ4Õÿ>ÆíO¨å™½¥ÀLÿ¶¤{õ 0–lþdd¼kÅ–,5ÅnL¬ƒ Ba—c9þzÑ‚+B@ndýJ²~ üÏe(}ÀÔ"‹¯³¤wÏ—ÁKµÏ€ötÿÂ- Ó¶'7fQ€7Ç0=¾Yèt5eöî ”zi4åtIÝméãšP+¬í]M¢jAÍtá6`h*ï³>¬ B_’v«ÅøöPUe \?çHˆáÉà•‰,?•Fþè«aE:¡À<1¡¸-nÐÃäÜ[œƒ)²t‹ËüÐý§MjÖœf ×ü÷ˆz ÷Å+/ÆÁæéöc"À°˜5î¿Â×ýŸ'µul>Aþ {æ+í ¤ ñKWÊ|¥Ü€{­O_YÙQ–½àqz{00Î ítSú‹N«¾§€V—E»•4½Êv¼h·XN¥€Èu§èˆ! Pû]{€iþöjM&ß7:]ò(öN|=± lq¦š'ÕHÖ%åÇE~aÃÝ ÉÖ$û{q¨I-&8D4Û<@ÌúThí). ­Ž;Õv!Ï'í³•Õ <”Wóo(D^ñhåĤj÷}ÜÔãÝ[º'IxÀňƒòZεÌà ÂÇ^´8æt“¯‰î£@k[è; ò$çmz?´Á]VQïÛ@D|ËûÕÎ`,–µDŠqpÄÄéjlÈîÅ‹6 F X¦ºÃýkÀg¨bìtƒS¯™Š¬¿G€‘w—[àìnòg¨Ú€`hª¤_žˆtÆ{ühˆ? ïW"RºgM¢§ßÝsøQà1 ù$Ø?òÆ¿ßÂG øåz·êE^`ü|¦‡Ÿ‘ˆæ‹ZÆŽ l”÷¶£[˜Ý‚wÜT¼@¬Ä´Ó39„zÓiÓ¯‚€SšûèÊ80ÆŽvGmßzÔ1Ï À9b÷Ai0ÿ2—ªƒX ¨'ÑAκqÅÑwoŽ€H´îZ"· 6ßn·4¿üh ßœ¿ÿæõ$»$#`¬\$´áô¿‰šªÃ&¨‘,¾æq>Ö7Y"Ó͇Uxg¶/7haãoaówÞoØBõKü­i6xâØ0š;z«iã6àÕˆr6 %¢_b¹Çè\KÀMM¯w±}€†ÎÞ>Áù~¼ù×tH½½ó£sqÏA49FN ‘}޾Z›¿º2kƒM0^ÚüÌtA-Kÿj`ãn´ÂÀr¹—‰ •ëö|ŠõW±Õý9:ý•G¥Û sOl „Ì”÷>Dí’-Ò©Z·Ñ²²ã~ køoÝA;õX÷ߺW†§am±ÁžX~DàÚlš+h®jh¦cßT3—ê+lv×Mtlà3Ê :°iß½Ô'Pé\›Æz° çÞ®e4:½§z|ùª­î <è€þ†mOMë±Eó¶ô›ÿ‘ÃFŠ7¤XÈv zlT-?2½ÿ‘ìhâƒjì[<˜µØÏjYí]MhZ{~W{1/Ö3vU[qÜ _P4¥œoFs”‚ÔÝå°®sù¹Y-ýªàB-:^çôœG£7ï] GSƒ %ãèí­7¬YSØB×þTÀ»ÑÔ=â{l`úG±X /6h9Y3ŒÍ²~o)<åŽ ÁW2aFûë²E ÕNUüP¹²_ ]©‰Èɽ™†ó~M¯ùô-ˆ¶D¹é¿B­–±zÖ‹_Ɖ¬(m}l¦çü~§Wç±1I Þ‹h:¥ý[eÈ~lâ t"/šl[×>V†ºw~HõZÔǪ¯èýçŽS€¦Ìo©‹Œ²C´sݰßÜv&5Îà¼ÆïêXÏw–¾.?ö¯Ýÿù»¶fÔjX_[ꈥ®<̶j¯FK–±Éb¿V°Ê‚½#^†mhykx-ëM N$OþjMùŠFµæ¾žÄ&ØçÂþUAýÓI”ù“GPã§i;+qžñžÄ½ÆVÎâÅ÷wE£Ï‰‡îþæ*Cã ¶DÁy´œ9Õ1äu-ÖL„kßÃ&÷}:ã0„ý\½ÃÜ„³f ™6`ç]9›Ž-ÅŽ;x2û«ÿÖÚìÀy|Ä.i Q/ëôµ;±ƒh¹ó^ùÞ[¬_ù?õx}'¾y¡móV[&ùo‘Å©ëzÞÃ&z¹—ÚQïÇÎÏ`ïž(,$ßÞ üú½uŽÏ8óK”ìyá7اI®Krê±ñòÀ!E YÜûGç$]Ø›£j« bo—ÖöÞ±ñKîm5~ý »RþÓ–½¶tWÕ}‰­äØ 2 +ª$£]­ × âJôÈ—,`<`J—ꃈG‡øhÎHºDÞӲ݌¡§È ‚ì+]|o‹7zøcß"PµÍÍŽÙ£©îÐ]Ÿ ú£ÈT­c?ˆ”ò²÷Tiw®š© !„izxD˵Vïèé€Hí¿‚´?ÅboþDH}– v¯= ,Í›ãO›”Dq掔À´ª¨’¸ë,׿˜”X`z*˜ãàúcþŽq ˜Ä­”òV >kJ¼ªœ ’ã•ì*EU‰ØÔT b¿«ºN+ìˆÌ©ŠÚ·¡i ˆÇÏç®WñÑk¡1ú-@_s´ácû; k·&ϨƒPMGÍÞÆÿ@¤ý¼åÝ[âOW²JÞ³Ü8ó‚ð°ŠóMËëAVN“ôDoäͳ/±yÙ5›HN•˜.Nѱ…J‹‚Ï ¹ç]’kH*puû|üŒM–\Ò º%kê5ÍÄu½=$Ù! ò»8>n9$º0ÓéØt–lXÒñU_³Ýª×@2«9zó C“ì¼a°¹$ªçý¬{p!FC$<]ÝL @|³_¹§è°ê#b*W€¯õcªu+ˆ^>,ê>ódc²ó5û@FAãBJ¤&ˆ3m^‘ò@´âûÞO*Ë úè‹4µç#H«H_uVAsÜÅE~`í`Œ0K¼@èó ݰBŽÿM^kÆ&q`ùr>í<2Úô3!÷5Aê…]ê)tÄ"ä¢g·ë õË‹ƒËƒY†›…‚TƒÜCwÒ_÷æì>аZïêôð)ìQÖëÂ3³ s@Qïn$˜•EB%µ S«$õï3ÈÜï©¿j2V);ðç8ˆõH|É&‡ON'uv=p;uG70zž¯]Y½ RǺMÌ ƒ”­§‰²¨ He\!8sÞ¢ëÉ£‰Àk;œeÓâÜbn®7Ýš; ã-³Á®± ××b‘X´ $=Ø´]^†Üa¢ªõ8ˆjÎG]¾’‰z¹mãU éWz¥q‰Ò%¥/ÃêTÝžàÕàZ¿]‰â¬øÒôDöGèÖ3êAâ µïð¤g^üðì¸BH=Ñ#{AÔiyûðT1H^vß”Wl¢=¼]þpdDì¨ÄI޾*V75âÔa]sbܘ>H>Ù›rÄÄ·~S©©Ž çÃâËÀo)* Ñ5öZ(ª؇]yê@öÁàræeŽNŒé™ù2ͯk"M‹A4ªÕ³<·d¶+}¶>¡ÒMÙ aOŒA¬Ö;Û啈ჯkA\ëk¹±£H××" +øµó#dP†33ÁÀìÈð<³¯×éÒÀ›-‡A*a@åÄž9`?×5ö°´‘Ç›é<Üï$ïpeÐe™w寕6±¢L% ‡$J ºrÿ6#+Æ!N>².I[5Õ¢LƒBRÖØÃ÷ê8zÒòa[-0x6= 'ùsãµà×kz€K§ð~Ã}/`<ú'^Ø |yr×°msµ¡³€ŸdNÚ‚û}Ú% H: ¸P{Póz§“V À½säÖ2üßÍè‚ಎRîq Ä•Ôeÿ—4W< ?DZ;({XV@Í~ïsµ@Z{ûLË:à6›§Œ ÉÇ3‡4ú¥Þ§§kl¢Ž^ÙÖ¢€º¾¸ökûVìo¤‹ìícwA8Ë=>Jö9ö÷èÆÊ{m 6:všÊ RŠ_4ð𺖾Mî±J¿À|É@ïvXRšïRÛÚ;ÝÏ¢@d(SeÇí Îøph™Ê7bj€ÿø«Âç±+À?òœ¾¢ä|³sŽÞ58~Ú•|˜pè—)t½ªçÀ¯Ñ-Õò3h)ôÁë­Â Q%O|{¼˜ê–BäRÜGN…}iú²`øĪŽïUótÁÉ«ÏøÌ>pâ³ü•攳|*ÐØ×7Úy_Aµœê ¦Pݾü =uú“ààϵÀ+×søsjx~Þ!è”~Á éÛ@*˜Ùî>“Œ¯g7¦ ÜæÕ;¾³tNŸ*ÓU"¸Ôi«Ñï˜J}bš|3aÓ‘ âN ,GˆR“â€eWªŒÕ£z7+£%Ä?4’6ú¿jq¤ÁŸJ ±¼³pjèÅÀàMþ¦‚v‹s8ÔaýÛcœ¾ùùü©¯ŸGðÂ@ì’sÆ Â¹©fÇGsI Šëd÷qt,å\9ýÆ…@æRùÞ²u„_n^z|ûõ祥ìÂã2ª@ ¾óÛ}ì)úÿR¸VÅác2é%#'HãþÄTß§,Eoo`^!uh_jvKý´·Àx‘cù$Û(Üú—ºl€Ï½ G¹ÙÀ;ì&¡‘´™ç3³_-Ž×½k d¹Vûý@+Ö)/9ÏÙO_Ú•Êæ@D&Ü4Â(D5Ô—o±b@ôžðhÂ… ºE%nj¼ÄQ‡Ã§èñœÿ5M±(cɽ%êÎ@õòÑ?1vÉ*Sþ²ë0G¨Ûlf_qZ(}—.Ëå65 Ë ÝØó~ÐõNê¾<43®ù?knÒÅ_t( ¨Ÿd¿,ö¦ˆ—¡SÐ^ w§ß÷ç:Ï‚ˆq;PÊÈZjÒ€rpËß¾ è™,¿ýª-pgïж’7Ê̇¬•¢jàÖKûN výTÖ×àÆ~Moæþ— ¦ ¢v=ý6Ð×SfJæ€Ä°U®ûıi½“zÜ V.·;bÐö{ä jÖ ‡«'è@6·9à/ §‡9}ëÙ{Æ{9uÀ<°F`­u> iô>0…ôÿ¹gp±rK© ˆiÈêS‘*˜ßpôˆiÕì’¶Y{¶­@(—-LqëÓ@Z^|ÆwƒÓ/üÒ‡<\¬€é‘e¿Êt‘JÐG)œ}d^,×òåq ©F u csŽÉx{?'.ƒ-Õž²g€±> «»[ŒÆÞªTç<¢Í·.è9kE Ä/Ú”0ìî­m2ô†Ü ð_ž|·öº6ˆ¨éh²j:¯aÁ­—#À«ÀäK©Ѱ¨¢×+—A´åñôzèº>çMwqtÁj Ù#Ø ÉY‰ÕcÀ´›<ŸÚÏ^L|Ù 4Éë!*-ÿçð ÏI—3“Ðÿ?ŸBàVúß'aåÿ]ÿMÖˉ»"flexmix/data/NPreg.RData0000644000176200001440000001026013425024236014567 0ustar liggesusers‹íšwTTGÛÀ—. M©Ò–¢(²}—¶Ï²,¢Ø»‚½Pì`°$–ˆFÄnìŠb‰€Æ{‹¢ XP£¨øZA ‘D…ïÎfvæsÌ5yÿxω{ÎïÌ̽3ó{fæ¹w×Ò)ª‡Ð¢‡Ç3äòxÆ\Õ˜«ð x<[s®4i×!epgÔk˜qXsœÛ]·FñxÊ—Šòª~<Æ´*áJ§õÇòPûF¯ ¶\é1»gA6Wºœët;ƒ+½‹×Í@÷ùËCìPÛ®SÚÁÍ\Ù`–ޏÒgD“Ÿ¯¢ñ!—•s¥ëžh“g\éngÌù~-¿¦Dã½ JKQiœU'…<Õ£ælGãO»–=äúݾ¼T\Y“;&OÐ9à.×~«´Añ˜×·©›‚æýÊû ïµ'ׇŽ?Î IäÊF&NÈÑø.Ú8=~s} š7.hçU––54DãšÎm‡æóš0¶ª†+M²Î¦¡}àgIbâ¹ÒªvM.ŠÓÎóìªSܸ² Çvhœ[ÉÞwhþÊ>ó¿=Œîï.ÍÝŽæÕ¤mBý½^—¢¸lû]ìÅç®×uÍDã¼;ö~öÇ«PÑ…+ÖänàöM¹Ý}ÇC4δxu5Z¯ç• –¡óp-C×mgÅ¢óºVݶ7šÇgÛ¤1×>ÝÀu ZÏË#]Ñ:=Êž­@çâxÔò‹& ñÎW5¶Ü9(«BÓá£sÙ]—ˆæ±ozè ŠÓ²üÐþ7˜QœÖQWqíÿóMº6žëa‹Ð¹Ô½+ÜÖóîYßàÍ®qÈçÔpf_´å…EEÈgY`éˆÆ¹7Ÿ=µ–W¤ ¶i·3¥(¯l†i({ÿK÷Ð:]òëç£þU!ÖMÑþDnÉGm·µ³hœ×‹6sæ£|)ouµÍ¶_z¥½žóÔù—Õ¾FùUéš®=¯%vícPž8.@ùíôÆw¼3ê¿Ø«éc´ñØ#ßÚ˜pt~¶ãîAý½æ¤×NCçò?!«¸’WY™ŒòÚÛÁê›´.¯>]FiÛó ølÞ,݇|[-Gûô.Ùå“Ýê c”÷|ÛccÔß®$7ßg3‘Ë/å‹K¡õ[Z–å½ë­´ãÃìÞ¢çÑ¡Í3[Ôï?ª¶hþ·qNyh¿¼_µÏAyíÞ§Ð~xD[•¢Ò%³ú5*Í+Š £qWÌU¨, ¸p_›'I>¡(o]L«—£õÖ‹Ý…âsìxíŸsyM—ShþݹùUhM%¡~Î]WŽC×·¨îpû®|Z±ÐÅa/3é‡òÕ¦üfk”goãKï¢ë/FCã,÷§ü€ÎÍ6¢qšö9h;Ô å§W¶@„ö£A›iRíþ~ý뤻è9¶¾ë‡ÖïùÔi:Gû?ÞLGyÖèäpmšlû]û|8ÌFq[;Ú8Ís4;•Û/åÎçœÐ>•5R¥ÝÇ¨É è=âõÌÔ^Ð2¬÷ ¿Xhpü]Íó‰1  ð{•œY õ乞Y|eÕ²]ÉÛÖ×AÌ£Ø ‹Ïí¿¬×'þ±`æŽÕ.S UñOíÎ 2WV[Ïk†£{Ì› Ñ;·HOœ¶{Móšý² |ûÖÖŒx} "/ÉiÒÌ¢ïMÖx JØßïÈù¢N}'._ç Ã[œM|ÖT¡î·šqûRòèfËGK”5›ünlî™R§¾ãf矣sa÷¤eÐxéúIcy©5ÔÛàÆ]4oþÄuÝ£› 0X2>îX&4_W-Ž]ªìªÝa%'ÁjXÚӌν@2ºì¾ÃЙ`s‡—7=ߢ¶Ø™Ÿœ4¤6ƒÚÆÍ }VE¼é\Q¿^\[z¢¸¸WÛ­ËOw?©;ô`Íޣ鶃_Ðôz>f3 Ò+õä½ lñO®ñïv€¢ûO£c|ýU¯TüxÔ>Z-)ö3µ-z^Y÷¯É_ÿð˜§ëATW¾ƒ‹%·g¡ÇùkÁãûUÇWÖˆ@YÏÂàQD³C¯wµzX¿v;Dx&Ô˹7ä6’°ºÑÝe$T$ChÃË•“#MÁ!½}ÏÖC¶£dÀë¢ý@ÕèàÁ/6€ãŒóµÒ³=ÀyLÀ:+¡5ˆ®ùØÚcñ•k’n‚]hüék9›!Ðmö÷ÅÛ½A½×8Í:t„<úyGåb°ô¿Òcµ´¨úá’ûQ!„DlÝUód1D”Þºp" T+‡Ü,΀ìj³¢Ö ¾=¹ª±:4;öÏ]]¾ š¯±üµn14ëc÷5®ÐÒ?©êĪ6 +ï=áø„¹ÐØ7'íuÖ.=P’ßç7šWÐfsß¾wÃÓ.¶B냱W3ú¦AÌΔ[߉VAäÁw-,RB Äõî•~A ©09v¨&"x«Çž™” 'c{w EŠñMÏöCäS—ágs¾†u[KÓö'ƒæBê€u>' ÊÞ|†Ÿùa¹à›Õ:e´\9-w±ü „¥¶ŸòC/DL||={™#¨âRU[‡~Â…µ×_”®€ÈLƒÌ)·÷€tY‹%›#ãA G ;Žl kYw"Ãý=B{§Cx^“’Ø®ë!ª}ã#&Ss!ü°å‹éoB!ÊêîżvÓ ¤Uô±¦ ¶Ùzul“ ’®ì˜1ù„½Ý‘’yq=(íóîúy5(ìl&k¢AÕ»¿± ˆïú>_»k+¨GËTz~ Q)…Ž=!ôEöð[‡A0mÿ¼fž Ý’päîaP]Kú%·ÒTùfg,Ì^Jë¾ô·a ²ž”Ôåì)­[¾³L`ÂraÙkPû˜ßK2î â9óÖ÷€¶ÝJÏ j‘ÂkÒ‰‹­ ªe¤&êÆðí¾ vÝ ¬éÅ;_Yf€:ýfõJßY 2oý¸K!¨½^}¢¦H‹ªc¹sðézà+ÇÃ\sŒß ¢Ûnj¾ º1„.êÙÍëÁnˆzyà©•I¨ç».yòl,¨{y̘¸´¨ÊøÏÿhS š=cß¾²°õÆe{z®¿âßöØWÜšžºÊ T‰3ûµ(Ø‘¯>Ý-ä¿›{Áö©¥¨Þ{\"ˆÛÄ;[®yêÄØ /Çwá[a™ýŽó'4‡m»ƒjô{‡l_éÛ› ZrTkLƒ¿[»¢¾Ýi§ÅD€Ú%R'‡póYŠG>KþX.Ÿ–ªUò«yíAÕ=ËùYíqˆLŒybs5(ÎqÊ»$‡å›*’o„È€ä3qVæ ÎØÒ·Ãú\žŽ;íº¨¨×uUŸ2‚Hÿ}­ zEBÔ†ƒmº»¢Ö5é·OÂ|øïÉï@Ñéb—yuQ >6~ öì摲÷‰?´¼›áî&²éIAýsï,!²¨÷ºŽêlîý0Ói_w;&NxúìíBúçíw ÷»—ÞLßW>ÃmžžÒöý¶î£›ý ß@5žxtóù2¥®¿ÛGâá1ó93m>‡=Ÿ­žxíõÄ«‹çcûÆÌKî3ó’yô•nzÖ£»ÏÎÏÿ:ññþúCâ`KÝ}&Ž®ëÚì¾ëî³ñ°ñ2ñéÆñ˜¶®ßyÃîã§æñû½ÏéÏæ…®íÆ´õäÏ~Ö£¯Í®›=gf;žÍצL?gæ>‡nºyøzúñ™Ò›¹¯o}[{Ù2žÇôófúáytŸÖË®³‰žyu÷uï)öý¦›‡ÿgùÁ¹é{¾ôÅÞ¾÷¸=ÓÖ·O|f6=ï;þáéùèú³m}¥¾ñŸZê÷±¸>7>}ãõÝÿ§ý?wúúý·öãïz>æûÔx>öù§ùò¹ëü»yó¹yò¹qèë¯oü§öûoŸÃç®[_œŸëÕ7Ïß}^ÿéy|ªïSãÑßçúþéûè¼è¯QÑ÷ˆÁþ§0üÂÿFìs¢»g„1Ƙ`L1f˜zsŒÆSc…±ÆØ`l1v˜˜†{ŒÆã„qƸ`a\1nwŒÆÃÇxa¼1>_LcLŒ¦)¦ÆÓ€ ÄaZ`‚1Œ#ˆ1Œ#ÃÈ1 L&† ÇD`”À¨0‘5& £ÁDcZbb0­0­1±˜6˜¶˜v˜ö˜˜Ž˜N˜Î˜.˜®˜n˜î˜˜ž˜^˜8L<¦7¦¦/¦¦?ff fП¼÷ñ¦Ã§ –Êû3sQtÕ@ «u‘®"ÖU$ºŠTW‘é*r]E+†‚`RšÔD¤&&5 ©IIMFjrR#!q‰CHBâ‡8„Ä!$!q‰CD"⇈8DÄ!"qˆˆCD"⇘8ÄÄ!&1qˆ‰CLb⇘8$Ä!! qHˆCB⇄8$Ä!!)qH‰CJR⇔8¤Ä!%)qH‰CF2⇌8dÄ!#qȈCF2⇜8äÄ!'9qȉCNr⇜8Ä¡ q(ˆCA âP‡‚8Ä¡Ð9ŒÁÁìÅ8¬ªîá7ÀýL‡ô8fd W«ý‹/U}?~ŒŒLLÌê1˜3X0X2Ôg°b°f°a°e°chÀÐÁžÁÁ‘Á‰Á™Á…¡ƒ+ƒƒ;ƒƒ'ŸÁ‹Á›Á‡Á—¡1C?†¦ Íüš3021´`f0ß=V†u¼?s…|i~ùý|¢£7Òý_¾DßçŽè?|°îk¦;áqºÇ µQ¤6×ÈûZû>O$ U!£1O96èÿ«N}IÊQRâÚ—ßvÿ¦ßvºª€V…´*¢U1­JhUJ«2Z•Ó*µ ¨M@mjP›€ÚÔ& 6µ ¨M@mBjR›Ú„Ô&¤6!µ ©MHmBjR›ˆÚDÔ&¢6µ‰¨MDm"jQ›ˆÚDÔ&¦61µ‰©MLmbjS›˜ÚÄÔ&¦61µI¨MBmj“P›„Ú$Ô&¡6 µI¨MBmRj“R›”Ú¤Ô&¥6)µI©MJmRj“R›ŒÚdÔ&£6µÉ¨MFm2j“Q›ŒÚdÔ&§69µÉ©MNmrj“S›œÚäÔ&§69µ)¨MAm jSP›‚ÚÔ¦ 6µ)¨¾K„Áú~ÕX ê?¦ÐîgOûˆWûŽßgðð1flexmix/data/Mehta.RData0000644000176200001440000000067613432516322014624 0ustar liggesusers‹Ý’ËOÂ@Æ·-"ÁH<èU/„ØÝÅÇY®z@\,jR©)Uþ{uJg> Ao\lòµ¿îÌ×íÌp0ÒµQM)å*oËU®GXr鿍’ªÒsë6|N¥¼Vö†Ô uU~•IÞ"ùG5~ªBÌÉ 1×¹Ž\NAê¸Wˆ»+{Wײ¯xÉwï’NI‡¤}~î,©ÅïOHmŽqn“÷´9ÇçX“×YÞY×Õ=f–º;œß/Ô5ì×Èמ|·»!9›ÑÒ•£ð#ŒfD{ ×<êÝ…sÆí›xš&q´:|ã(˜IŠM‚q'DŸ…ß#c±`•[6žÛ¤ )ËÌF³Î¿]ZºËÕ³Q8àöH+:›«›Óýâ=>™ã h#`ú—W× ®òAd@Ô]€.AW xhxhxhÑ“#®mÞr¯§Ák(½.q°2 goñtʦ‡8 "^”Éû“ðýK®T¬&ñ¼W¬ÚýÿYÖþ2CµÇ z“„:›Ï‘úüLj.Èflexmix/data/trypanosome.RData0000644000176200001440000000212613425024236016136 0ustar liggesusers‹íÚÑje…áIRQŠÑÛ(™µ¾þÎ<èxÔÓMl#IA¼½„^‹Þ’˜ì½^-HAÅ$¸&ßG2›—@æaòÍ‹—:}y:MÓñt|2ÝžÓ“ã›/GÓôìéÍüôÝÕßoÞ^^_¾¹˜¦“³ÛK§é³›+Þß]v8çáü‹ópŽÃqÿÇç7·ãû¯Ï|{üýùÓöøùÁÌ/ï~¹‡3_ßgþ²=~=Ì̯vÿ‡ùÏÝ}~˜c~±»óqͳÝýt˜ÿpîîƒÃüwæ½?¯=–yßωÿÓçÓ<Ÿ|ðªð“·›7×7˳éî%áö›O^\l^e¿¼¾øàCO¯.xþ‡Þ¾UÜþàhÞ/Ú/Þ/µ_Ú~YöKß/c¿¬»åx>Ï6gS6g«l-Û’­gÙÒPJCi( ¥¡4”†ÒPJÃi8 §á4œ†ÓpNÃi8J£Ò¨4*J£Ò¨4*J£Òhi´4Z-–FK£¥ÑÒhi´4–4–4–4–4–4–4–4–4–4–4z=žFO£§ÑÓèiô4z=‘ÆHc¤1ÒiŒ4F#‘ÆHcMcMcMcMcMcMcMcMcMcÝ7NæósÖ™U¬f-ÖÆº°vÖÁJm¦6S›©ÍÔfj3µ™ÚLm¦6S5Q5Q5Q5Q5Q35S35S35S35S35S+jE­¨µ¢VÔŠZQ+jE­QkÔµF­QkÔµF­QkÔj µ…ÚBm¡¶P[¨-Ôj µN­SëÔ:µN­SëÔ:µN­SÔµAmPÔµAmPÔµ•ÚJm¥¶R[©­ÔVj+µ•–K„%Âa‰°DX",–K„%Âa‰°DX",–K„%Âa‰°DX",–K„%Âa‰°DX",–K„%Âa‰°DX",–K„%Âa‰°DX",–K„%Âa‰°DX",–K„%Âa‰°DX",–K„%Âa‰°DX",–K„%Âa‰°DX",–K„%Âa‰°DX",–K„%Âa‰°DX",–K„%Âa‰±ÄXb,1–KŒ%Æc‰±ÄXb,1–KŒ%Æc‰±ÄXb,1–KŒ%Æc‰±ÄXb,1–KŒ%Æc‰±ÄXb,1–KŒ%Æc‰±ÄXb,1–KŒ%Æc‰±ÄXb,1–KŒ%Æc‰±ÄXb,1–KŒ%Æc‰±ÄXb,1–KŒ%Æc‰±ÄXb,1–KŒ%Æc‰±ÄXb,1–KŒ%Æc‰±ÄXb,1–KŒ%Æc‰±ÄXb,1–––––––––––––––––––––––––––ÔÖ’?¿7úö»Íõþ½ÑÑîºÓW›w›ç¯¯6Û;›~Ûž¿Á뺱­&flexmix/data/candy.RData0000644000176200001440000000037313425024236014656 0ustar liggesusers‹eËnÂ0EÇN @Ë£´,à*ÔtÓUűbkñZÐR5 !~qfFtâkæŽù×ÂÕ5"²d# ±Å˵ªXKK¿_‰¢—¼ˆ¨:t{ ïå]¡”Á¨€ÂÁ÷6@´À3hƒýZÈÞ€$à¼.èqo“s‹DL‰‰ ç3÷—ÚûŸõaù •™_îü6ßßñ4[ÿýûßþ Fn¸òGßßdȼ+¯f¿§~qV‡[L"âD"C‘‘H*2™ˆ|°Øä]-Qsjµ¡ÚH-U«MÔt†Ó.̸®(¥©Bflexmix/data/tribolium.RData0000644000176200001440000000055413425024236015567 0ustar liggesusers‹Ý“MOƒ@†‡¥¥B¢6ÑÄ4ñ4²ý ÞLêŃ—ê¡×•.†„¨þûê.쎔XõàI’—}˜Ù—…ùÝ‚: *A‡ˆ›зÅh—yôœÅÑ:0ä¬Ð¾+t&t®t)t¡âòùXè@è´5ÊZ¹Ž4í©Ø¡ª¹R’ó}•+çNTž{{ÕõßFÑ[ÝS£!ÒÙÐål´óWPw^fȨái †­Ân³B×*Ë YPf¹ Í޽íù…ÌŸ%_‘¼·vcÏXQ²”˹ ìͲ4\øl=°%O‹ºÜÔå_nj»)K¸îAG»Íy¢4J_T û”•,þœ]ÅQÀJ®½Ç"^´À<{4—w¿=šFÆ&| S 7 ˆwä!Q¤!ÒiŒ4Aò‘¦HèAу¢EŠ=(zPô þŽ#r–¬dƒ0ýª 6¢Z7Èäflexmix/data/betablocker.RData0000644000176200001440000000115113425024236016030 0ustar liggesusers‹Õ”ÏjQÆo&i4%P.\ ¡¹w’¦(t0Q‹¨¨ŒàŸN¦SµfÚÉÝA)ˆÒµn¥ÐGðºp]½ÉùÎÁ̾ 7¿ïÜ?ç;sïyÜíér¯¬”r”“WÓ¦ ŽýË)U)Yžé‡©ßâ`+L”ÊÏOÆm;g[Í;­¦?¯^_¯#¾à xsL|ÞEÿÄ ˆo#^Fܯfòs=]ðÖuƳãÁóàSŒß@\{è€x ô=âÃL}—ßÏÎ÷2ù«™º=÷«¥øGûÌõ?Çøø= މ¼û=âûâîOâñC›¸ƒþ„ÄûòØ_#nbüüø®Áw ññmœœÏ3¬{ÄõþÆ|:Xï>Æ÷??"Ï.ü¿¡®Ô‹çI°þóál]o~c¬ï#‘w³J|‡<{´NaŸ—P7:å¾ð}àûÂçÆ÷¨ òù_ù}à{z ä÷¥ò=o€4  6AÜÿ­^ûý8kÿ&Ùr'Ôœ“i3ŸÄbŽÂhhU3&½§:ñ MâˆÃÕ$ôÓp=³v.ˆü!/ÍanqÃÒØ~^ÕqvúÀßyz§wmæ×CDs«qê³k±Òɇz•¦5lÛ®LÚR¿­ÿ›º†¹ ͰpY4Y´X,²h³X‚p ¢¢´(#ÊÕÕµ(ª-J<´xhñÐâ¡ÅC‹‡-Z<´xhñ0âaÄȇ#F<Œxñ0âaÄÃW<\ñpÅÃué<ó2©¼î§~}#±'K—IÿDàê‰âflexmix/data/fabricfault.RData0000644000176200001440000000054113425024236016037 0ustar liggesusers‹]=NÃ@…'vb#  ‚ttçŽQ *ª´KRRlˆDƒ‰QqZNWî€D ˆ ÞšÝÁrñü>¯gæ÷`§Ç~Ï'"‡—RQÑÁ£@T­Àç#u89éGê|”¹Ëi)ÑTn@·8ßFÇü€—Àð}Î5ü ß®àÏxŸÁ?áë8ÛƒnÀk^å~€¿£.†¿A_àoø+êÁ÷ÿ{§÷†j^{*C%h.#Ïœ¯šzßxÙô­÷þ® åbffÍôèY®9wrWV«Ó!–¦ª¹¬tàþp|”Û·]}Ÿq¾±?R±m´éþ@%ªM03W^™œMÙ¬ºi)ØBh¡i¡e¡m¡c¡kaËþa°)±P(Ôj µ…:B]!É`É`É`É`É`É`É`É`É`É`É%#”ŒPgühý†‚Pßflexmix/data/patent.RData0000644000176200001440000000442713425024236015057 0ustar liggesusers‹•X tåÞ„˜ð$‚ Ñá›’}GìÙÙÙÝ!³3ËìlHâ¡v‘‡ ïDêÁ¢PN…µ 1y…gh©ÊÆ¥ (-¥4ÂûÑ‚ìÉÌŸÍcsîÎÝ™oîwïýï½óïÊ®|KÇüŽ&“)Ù”ÜΤ‰)%?’L¦ÌxN*ž8£Ødj×M½…’‰âFQïf1N§¤(”À8%™Q$™çT[Á»Ý¼À3 碼RÀÏ+Œ@ùeW«‡^pr²"™Ù ÇÉ+µi=ÏJb|w"ð'{‘ò{Ùǰ\PáYFP¼È¶/dÐ ÙofЉâL$ÞÇÉhPÄh}F*¹‚¬Ð,ÄÃëÙ ý~¡ ®ý€‚ŠÀFIÁçazˆñýw2²+‹Í–³µÌH4|>:B)2“lj’u%žcIJ¹x6—èÜÀ˜þ;9Q¸‚¸+Lð›K¹ä 'ÁõrÊ|@‘³¯€“ÑËàYFÆxÍc±Ž°x¢ç¨ ŸÇ+2†Ùª¢à½Œ¢p¾ØÉo†çœS–‚¯Ù/‰®§ZUr ¼À‹jÁS–çD–‹oÂK.®yòãáѨ_IÜIòcý$œh;Ý9$šð2ŸÈ žà9_±•%'Ž*‹Éöâ–)~ýË8?eʃ/ìÄòÀ9¨>‘%gû‘ 1<Ž}³_P÷ññ˜M£²<Ù®ìXLð>^ñæ jĬ!‚äÎÄëMÝÖ¢u¶‡‹±l_àL`w‰Ç÷z~âxõ]­n˜äøñv65þNlÒ~M™ÔŸVÉú9 eÊã(iúµ‡PžFé­W§~n¯?›bj:TLj65¾IçLÒõ^Ï×­(S"0êQ®s©¶¶éÏu‰°iÒ팋ˆÇ¤ûþ(J'=nõ(D”Y(E(ÃP2P8ÝÕ®¶ö9=¦×±4˜ê¸på1Ö ¯Ý®ykÙì³°Gýœ² v®<”²h¦‰î¸ñ“Ÿ]Í;ï]è5qÈɰLÇ¿,í¹™_Qð-–}üyÞ¬ýàë¡£>„M**Ô ÊFægŽ÷„Ý÷Žý£ÇÒ!ðÛ×7­îtî;XÐò·±—¶~?e{/m„OÛç ¬›_ «‹?ºðþ?a‘-åòèÝaß6ç#׿€íßéûÈ¿?ƒ]}JC¯¢ç½¡âšµx­Ÿy^ÎÊCð»[›~¼8m9üj¼ï«[›êa‡~6ø'¼|ÞSñÈ+0K¿^ò­ëD·MƒaÍÝ!S/VO†N¼º¸òÌ_`ù3iÍ÷”¼ûÉ{‡/¼{ßv-úæ Tt¾–µâb ”Wßü÷u‹aF»îj¬Òí½ÉÁÓ¡òííRçÅ>(ѯWlq¼œ±j TªÙ|á l™PøØÖö9°¦qà­;>(S?§ö€%=¿?l¨½´±/¦–Ôßñ›’§`c]Æý9Jµå™ %æsûÌž ¥æ.y)=WêŸTK+aCÁnŒ`/”«Ë3t”-­ß÷déZ(ÖãÞ^˜Éî°SK“ ö¨Ñ BÙM-Prû•]M‡$••T6,;½ï¿…Pá,¼þüËŒº¡3óëýÞ7«L¥kç}ðâØ7;$Ÿ…]¿)êYï‡[šûCiý>zWþl”ƒNÝŒáu¾wÙw¥öÿrj}¨æÞ$¨›ûÜ·Û¶|H§žðç†WÒíM¶é‡N„óXuÕÇ·À¥Æ|Ó™7¦¸&èK'í×®<­%2|°ËþïÊ °Ÿ¸ý}øþVµ @mÉ*5SájùVŸ€;[g§ïzçrø\ýÜú£á;»Ôö‚kùßî§ùÞPty÷ä’ûáYùåæ  #nØN­>òœÔÁµáz­=X:ýÅ%ÿqÌÕª”×Ïš¾üãpFÏßM5šàïáÞÑ»W]¡SG%}4#{ݵæãÏ•#¿׬»ÞùìØª´òºù_7ü'|}¹št×:5¼…tÒÜì îÒ€´¶J–ÝaŽ–`¸ûÒwÇrn>¼öÜKwª’uÿ7æ¾ÄbONM?\׸εZ[= '3úź>U©—/VO:î¤S{jëHwZ„îíÈ÷úÝ*lÇM˸ckùy)çk™þÝæxÎåÀg%çûqÿ¹>~¯çù÷~gú¹+){-çoBûcýØ¥œ«ç¯†ñ6iÝõ\íï!\§uÔ²´n—ÃiÝ.åÜØWR†Ÿ»’óµ½ÜØ¿úñò¯äw¯—–§y±^ágl«~tËh}ôx\Ë›p\ׇ~â3—ßâsÚ&÷:©½º¶µ½ºæÜ°©g“±û Î'½/Åu­Õzùýbòt-Íg?ïŸÑ½Ãøz{&ôºè5zì*|çsFŸ½tÜý§¿{\ɹ×r®¯éùÎצ·5®›áx¾žãëõJÊÓ÷Œ8æÞžëátÎÝù÷7Ãésœ¾—èsŠúäuóºßH–÷áõpZçxßòsã|Q—t½®×\Ë5k)ã:ô…>#ê½sãýêZúÕûñæøÓ¯ñïý>0Ⱦ·CÇkÚº‘óõr#û:×b_輌–i}ãÚ÷~‹ïoÃÀøéG×H¼x½ß‡p®ö«¿‘ëtìÖáœè‡öµ—çíÐ6jYSÎë|Õ±ÑçúXoußçB|‡÷l¯ûØ?=K×ýõ€nW\“Zn|&Õ5w®ó²½>þ½>[Ç÷Š+9?Ž•Î}¯ðã×ò>x{|{¿ë³ÞãxÄ¿çh_N½‹\{½ôÞ¦sZûéz8m‡Þ[õãí†ñû³¶OÒg\ÏzßNúÜÈy:¦Þ>þ}|ÖÓg6;Þ—¾æ´|>y_êûˆ_¯c<ȹñÙGŸ«â;z\OºÕÍ )ÛËóKÿ6q=|Ygÿè½AÇÆÛÇÅç‰×U½P[Ýf­‹ß߆áK‡ôÙÔÛäã: §÷ÍÔ¹éû¾6Çþ~¢uÒ2|SuÞ ò½·ÓÇEÇOç®!Ïö²âó´>3é<öru­ÆgÌu8׳Ô-ßúÇðòü§:æßëš×ç@5@çY\gêY¼ûó ÷Ž£›¤9qn^ÉïÞ~}›ú;€…Ÿ§ï{Ñ,}¶ö2tl|ÝøÜÒû„>wê½^û ®Íø>çS¬—Z6öé×ëßáÔ­Î3­÷ñçÚß=ôýMßWVáX|Ÿ‰sfìØ*lå3–7W§˜µTæ9õ™ªÛR?MmSõyN]—ÊYê‹súúÜq›ûnéú©þœ+ciþ͵oìÜ¥ñY*kîºsçù\æ\›Î™KS×/ͧsæÿ9e/͉¹þ™:® Ï™Kmzμ9'w®¬sûlª çÎýsçÊÒž¹¶ž»NÎm﫹öNÍ«¥öMµkª¬¹¹rî\Ÿ[SS×ÿ7ó`iNW÷ç¬ÿÇŸ'ÿ?ÖÕûÃ?ïzÜûæxÖ§o8|¸}8þvùyÿäºÍë÷·?ùeý²Û×ï÷~‘Wœ±b\Y­“?ã-5W›¬eŸ³ç–ÂXýâïë™l='–1EÁØy±¬©¬‹‘ë§r§êËžë'm{,3öûØ÷KÓ<Ö)fÅkcÆ\û¦–fÌ™[Ús<­‡/ë>Õ¿ñŸCÇÆ-öAÌ›OK ÍÍñu9ÇîX¿MÕwl.O­ç©rôºx|©]:cý¬mŸšcçOÍÑ©ù1Uç±¾œêŸ8Fq LÝ”¯Kc8ׯSf•3WÖœgSóQûNÇblŒ–ÖÌT­¥¥9¶4¦ç¬×)cÆæÓT_ÎÍÅ©:DÃææÊÜýzn޵y¬m«‘Ÿc}56wæÆplýOõÝÜšûKc97—çê9fÀÔœÓý©ypÎxùX›ÆîwñØÜóÅ”Sc85Žså3?ÇæÏTKLáT=æŽÏÍ3û±gŒ©53U祜±9wNŸÜ~ûÃý퇧ÿQå—§í?ÎÈI]bflexmix/data/salmonellaTA98.txt.gz0000644000176200001440000000012213432516322016550 0ustar liggesusers‹-‹ËÀ@ïVa ¢Ù_gI÷A’ËŒäñÛ0<,Ñ8¦ƒÜM¦PRe)Vé^ÌÃ.H3$UüØ»Vª‹ÁI~~½^yflexmix/data/bioChemists.RData0000644000176200001440000001451613425024236016035 0ustar liggesusers‹íKŒÇyÇkgÜÉ%©‡eAò"¶â$@(n½Çˆ°´c@€q|ˆ>xI.¥E–¤°Ü8‘O¹äìCN¾ðÝg Ÿ}÷Õ¾~C¦¬l/ë«_Ï·Ý==”ƒÁZýª®Õ÷øUÕìhø_þ†ÝüƦ1fb&ËKgûÙéÊäì?KÆìlœ·î=úÛw=>}lÌò³[kgûå³óÿ4Ï·çÛóíùöÉlKÏ÷çûóýOnŸ<ߟïûòóýÝWžïu_ýÜ×>ýÒ‚ûúÙ¾Qöf¹¹u¶o—}çl?[W.Ëš²=fèñ£=¦ -}ïé1I×µÔ¡Ó§«ËÕÝW___ºêœW÷Ø:æÝÒé³M—ý‡ê›×Ÿ!?Žy¶ˆýçõe^lͳÎÑëŠ×y±ü,1ÛgÇEó}L½C>\$ûžåòP߯䨨vÌóߘÜ^´mC÷‡Ê,jû¡rcãuÈ7}¾Ãcꊃ>; ùd(Næi½?ßC Y$ž%6‡ú¶Hÿ†rw^\-»‹ÔÿqòcȆC>ìӞ狱¼è‹ã!ŸuÕ;/öÆÄÌÍçÅÒØçcü0Ö'}~ž—{C15¯ÿcîñû<­±öêß¼XcÛ1óÚ?ÆVcòrŒ†ònlûèËÌŸ2׎¿}xüØ<]q6Oš»Ë_=|XNWÿéуæbæ­Õ»Ç奥RríþÁÝÓG'ggl-\û Ô×Ѿ JÂ1I6^c’öYaÌ{cÊ-’@}v×®±CÁ=Æ~C>˜wo¿ÍK¦!›X_ÙyíûõϘXí³S×õPŒw]Ï‹…y×cò|¨ýóú9ä÷yq1‹°l¬iã˜zÆÄòP\Ï{6¯-‹pkè8¯_cl5Ön}ñ=6nÇäË<,šcÏšëcý1Ï}þcÿyñ5}>›órs6 ÅÏ½ÆØ¼ÏcÚ7ÆNÏâç.?|œ>Ž»¡v?‹¿†|>äËgááǵͼ¼ãßElôqãѶͳû<-ÃÏSÚ÷crëY¸ýqãjŒMÆØB÷¯ëºï¼ïÙˆö5KÐÉGfv©ºöõ£‡ï–«K_=899:¼÷´ð²^2KÔÎïë.•ãD»Ê´7]n¹ÕÅåÖ{]ïöÕ»Ôq®]í˜t”ÑÚc´¤¾®ú»êëzoÒó¼kkk-™Ù:Úï.w”Õåºì4Ô^]¦«ÎeõL—koË=×Ke»Ú7fëòM;ÍLÇqHs¢Î»ÙUvYÝ{–>öåÙP\Åp_}›Ž§¶¶Žå.»·õÛy?”sí²ó6OºŽ®{]92okó£+Ÿútºú1”—úyWŒ‰í‹vCùÝõn{ëË×!?è­‹}}þîâ‡ØDÇ‘®Ûtœ·Ë Ù~‘ñ¨¯­cÞÑÏ»òml}uèc ûÊËÖ7†ÎÛÚSš¾ºuù¶f_lvÙgÛtn=K^uÅlwu;»ÊtåM»þ±¬îòK›W]ßÿý×þçüùv9J9Ñ™|÷›»Íqùõ~«õ|ÿ÷¥~ó“Ÿü¨]Oi×þ“ÙöíÿªôçÊìýÚž«ßüKÓî‡Ô+ö\)ïÉóëå}y¯\ï¤ì¬ía~óãóûåùÆSûÖzä½µÙvàÏÒÏ•r,v¬ú”÷–¾÷½æ¸ÿ ¥»òÔÕŸbŸ_?y^Ë/•öHû¥ÝÒ.iO‰“Z¾÷ÿ÷½ïÿ¬mé¯\¯”z«?J\ü¶ø_êÛyÕÏ%~oo—vj¿Êµøñ·%OJì??”öÔò¥œèŠ¿·~üÝoµí"÷ÅK¥퇕RŸ¼'õȱÆ[y.vÕñYí7kÿÚ•‡ŸòžÔ_íVì+Ç¢SóJâ]Êëø7åzëíì¶ÛmT9©©ä¯\K>K;/ÍÆðgÿIѯÜ»–8îI~׸”þIž”þ}Xì~iV§k\?>‘ü-ºÜÃé‡è׸Þªç’ObßµÙ<ïøÔ¾:_«?Ä®’kÅÏÒ¿Ùþíÿºèì(®jþÔx”øRq(~÷7ÅÞ¥¼Ø_ú-וߊ÷bÍÑÑþ’ü¹¦â\úaŠ}Äoåºæ¥¼_8)ý¬œÐã™ÎËj‡â‰GÍ7i·Šû:îŠß·Ô{ÒOÿÚoÚ>E§rA·Cʵ~±+ý«ö/y¯ó¢Ä}Í­/ãúfÏø¶4;ÎW?êcÇr=éñ×Ú,w+OÄŽÊÕnÒ¯¢Sí öJŠý&ÊŸ:¯´Ý7‹…K¥½û¼õÞ¹ÞDú=u~#õJ9±ÇJé§´·´£ö_æuøéæí.{¯©ø‘qS8«çk’¿ò¾žwŠ^™gT]±WåEy^ò°Î$Þ'Š 2_ø@ñó·jÜ?>Qy,q+í—q\gÂÍj÷+RO¹žˆ¿Wd#í(åä¨çÍuœ-G=N_ëæè…x”x|½¤Æ9–¸­q-ñ.õÉüM8%ù"í’zïö?û•÷ªžŒ×Ån•â·R^úaÔ8&ó›Éì|·Æg$Ž>–ú×§¥¼žÿé<Õó•‰â‡ž?Hß席òJòL¯o$Oë|¡´¯ô¿òBt®©õ‹ŒŸÒ¯:/wÔq¸ÌSëzg}ÖŽŒ£Ê>zÜØÖ|/å«K6%Kû$VfóúB>_ï™®ÍÎKëø#|®ó É«Òï:Î weRì©8WLJº.–uD©Gì¤çߪz¾ZóPì¦Æ-â§Œ¯©ø}e· ë~á嵎YWy ãEž‹½$?'ÊOÂuÍ¡Ê5Ïë™VY+Îh¾m)~_X–8Rë›:.¬«y굞XSóg5Ͻ°Þ•ùÞ/§$.©u¦ž¯ûïÿAæWeýþ{µ®ùÕìüëB ïj{æá:O¤_ªù¶pGù¿¶[>‘þ\ÈYÏ(nɸ ç½kŠgr_ÖW’§zý"\ÝPíªãRá^¯ÊøÝc‡º>ù¡¬—K¤º¿R¯^ÿKšr¬Ÿ']ý9˜ž7èqçÂü@}~#z“{Š¿¤=¢_ë•qM}^!þ¬ŸKÈ|FtU~ª<»°NÒ\Ø’qNµSÍ?kœêyŠ|N ~–8ÿ—yÎþï„“j½-ñS?GRŸ£þ®ôkSµ_óìuœ(^é~»]àÿŠÊ÷kªŸuü–Ï”Ý×f×=•{uÞ3’Û}ëáR冚OŠ>’qUâJ­džX×Ý%/©ÆO½.YSë×úyìì|F>wÜÿcY]ž?k|iî _Wçe-ù ó?ç+ªýêsŽ:×ëÝßúù°'¶T^H¹úù¨çëø²¦âWû¹ê+^^›]]ˆõ8.ñ{i–óu.ﯩy ¼'þ”þÔu¢p¼”ÓŸ×qT­g/ÌÇD¯~#ófµ¾–8‘|ÚüTv‘<½>»#>Ôx$q­ù·5ûw v©ñ>wuž²©æ™Ïuž5;¿«õÖùjﶚϫö×~n÷ð <—ù¾\¾óÚüþünQó÷ìO¾#±Rî7Û•§ïÔûͶTÞ¿\Ž­zVK9ù;ùºáïår\/ïêm£´aµœK}Ë­ö4õ¾RÚµ^ž7Ûv¹nÞß2ÿ6+ﯘٿão¶žsñ;Œò;TrO~/Jê¸^Þ¿Q®?Ûº^oé/—{«†ß~ºTêo÷CøI¾¶Õj§Øºí›vßåw¥Ä—[66­~¼løÍ­æùëžm?¶7i‹´yÒjksÜ6üž•”[-umbB®7Š›2;H}ÒþÍ–ÖÃïvÉ&6ß.Ï®–>l·ÚiÌl,].ådÛ)G©Cb 3Ò6©cÝàk±»äЪ!Úù0ié.•²bÿW[íÝ,}mÞùT¹¿ªÞ;ɳ«­öJŽ®´Ú!6Ü2Äâj«Ó*ÓlŸ5øÐ”:¤ÿë†Ø‘œÞ6³>lÎ_jÙR|¼]úÖŽ‰ ±äèÕRן¸`Zv›Rv«ØëjK_ú*ù$m¡<—v|ÕÿÂHÉs±ë5C<¬µŽb7ɉi¹'ý“¶H¼J\­•ç/ÍÒc.æ›Ô'1µ]ú y"¹¿SêÖµmÐ<¿QÊ—ß0?/'\’˜lçqsïr«Ô#ešýµRÇF«Þ C̬´®…×[íY5páSÅÖ²]2ðîÃX!í“x¸RêÝ0øRìÔî‡ôUlþ×¥œØpËÓÂ0iç¤UNò|·ôE¶öo:ʘv½ô©Ù?oˆSîí”6‹-Äfâ[±Ãª›âÃvÞšbƒ&g®âîÅR× .›Ù1h«eG±‡Ä·ðp¹u- Ù,š— ^1øLÆÙþI?$§%W_.çÒa“ø@òKì¿c˜C6Ï^5Ìm$^^)š2Gz½<&‹e~J_o”:Úó•ÏŸÉØ(¶ŸÊoÇnæ>¢!çÍñ…¢%åþªœß(û«-ý—‹®ÛÈ|ðŧïÍü2ÐÆ½ƒÓƒãƒ;‡Çfö½|í‹wî>8º»û¥£GwË?ž²ûæî×wÿîÑÃwT-›§Go>>=xðžªæš»µû•ƒ‡»öÖ­½Ý½[_ðSýÓD>6Ì2Ïéqpr*§÷È郃qñ?Ý“¡hù½wïÉí‡O•À¥ûNœj‰Õ7¦7o½ó \Ìöæôý÷Î{s¹HÝQûý§ÇYë}ûàøæ¹t#ÅãkÿíøÎ±\Ù+sñ¨k?é®ýµ/žœÝ=>|¼{ôp÷øàñé®Û}ÿäñî£û»ÿðî—K¡—ß>|xïðä »{oyâàøp÷Ö[ÍAâ¯ü¼Íó÷ÏjºõÖÃGåÑKÿ/îž4ÕÝ}÷èøÞÉáÃÝ¿Ùåéö™Äî{'‡OÞ‘Ê>S´{çýÝÆ“Nf[¦c÷äÑ¿Þl…Ïù%̓¥=9±râäÄˉ„Ï’´h)ÉI–“i9™ìݪg{õÌÖ3WÏ|= õ,Ö³TÏr=«¶jتa«†­¶jتa«†­¶jت᪆«®j¸ªáª†«®j¸ªáª†«¾jøªá«†¯¾jøªá«†¯¾jøªªF¨¡j„ªªF¨¡j„ªªF¨±jĪ«F¬±jĪ«F¬±jĪ‘ªFª©j¤ª‘ªFª©j¤ª‘ªFª¹j䪑«F®¹j䪑«F®¹jäª1­Óª1­Óª1­Óª1­Óª1­SÑXÞ»u‹Ó=N-§ŽSÏià4rš8Íœ¢¶‡Új{¨í¡¶‡Új{¨í¡¶‡Új5‹šEÍ¢fQ³¨YÔ,j5‹šCÍ¡æPs¨9Ôj5‡šCÍ¡æQó¨yÔzÔüû—MýÂèL‘­ó¯ÿÝ<=¸Ó|oorãi‘‰|§ø|Ûÿ…œ5¯®ýg;š˜c¡Š:þ©ÊóÜCí3ïÊg¿ø8û¥H3pücÏ?/²Ù|ùõæý“ƒ‡¢üµm$Mflexmix/data/whiskey.RData0000644000176200001440000000732513425024236015247 0ustar liggesusers‹í›ÝnãÆ†eÉöÚ›ÝdÑmР(Z%íAƒÕ÷}ú; ° ’4mÑ i‘œŠ­¬…UäTV³Ø³ž÷´—Ò‹è…äzi)›œ—ˈæÔ¼=#j8ïpæã3CŠþìƒ/éþ—÷;N·Ó=Œþö¢ìa7úsÐé<:Ò{/.ç×Ïg/;Þã›bÎÃèÛïo‹Ül›Ã_ÓŸEÛæ°“ø»ÑöFœïÆåR[äÙ9޶߯Ÿ{q¹Iœ?H¥™ÏÝøøMžãÏéréÏoeöõ2ßoòÿÊì¿—©ïx‹7>ït]§©ï6ŸßË|Nê8Jù¤ëL§§[Î{Ó›í'9短7ÉgÛž-s·3»/홣ÃL™ãÌù§òÙ²éíÍœñ?JåïmÉniï?·Œãa*=É“Ôq”9¦—Ú^Ë´mÛØeÛž.“‹“Ìw½T>ÙŽRþ'[<7ÛÜqÌúæõ}ºÍÿ¹c|ÒÛIª?’>KÒ_äóӌ׽-}’íŸì%}t˜)ûFƫ۹{,¶õÓ/sÚ‹Óœ:·ìËÆ_ú»þŽvt~|¥Ó͹=HßQg{»²Ÿ{qßm‹…<^¥û[ÿü*Ç/=iÎ&ék[Æ<;Çw”Ùö9Ýö¤ w±gWl¤½óÚxW}Ûâ±—Ú—oNýÝhn}çßïý·t‡ö½ªžÿ¾÷›n•íÏlù¦‡¯í­;¶|ƒª)¯Ÿ‹ö¿îr»ŽqÑn5e|›ÒΠ jZü7­½uUv^nÊúÙÔ:@×úÄV;v•35~E׋¾ÇOPÐ>J7çÂõ¤GuûÕöý±/ãîK;òä{ûLÉöú#¨ÝjÚ¸7­½»äËýQ]•'w•÷å¼|W˜Ê©­çÕ¹úݳmªúœªíýÙöóÓ¥ìü[4žòî§›öüÂ÷öù¦¦ô—ëvêz^äú<‚¶Ë÷õ¶í¸1½Îp]OÙymW˜~¯K÷úZ÷óï}áZÓÏÓ—öÛŠ;Wï§7mÝÜùöKùºþѽN ñXO®ùàËø•}þ£Û·©Ç»Ò¾>×Õ%Ó×i‡W¥û}éªÏ­M«)ãÞ”vú¦¶ýŸœ­ûNÓ놶ŒG¢¦ðÊ÷~÷½}¦ek¨+Ž|/_Û¥[¾ü”ëþ6ýÿ®Ïo_åj|ªrØUœ4->uÿ>äëùW}OË·ùÞvý¶ãÃô}”îß ŠÆKS®ªßÕ“©ß±Ú6Ö}¨nýEÕ´çÞ¾ú×}^eŠû¦ŽÓ%ßyâº\É÷óö½}ûª¶ŒKÓÎc_Þ¨zæê¹§íû\Ýís}¾Eëõõþ¢¬|éﺪû>ˆ-ÿ}×¾®£‹>+ZOÓÔ4^Vå„íóhëõ’•¯íóµ]An¤‹¾Æ•¯ëmÝ~¦ë÷už¨ºŽ¶½þÖý»©©ø2u}»æƒk×rÅAÝíðå~lßã)+_Ö¾¿ŸaJ¾¯oš÷MÉ×v••/ça{¼mñÒ—þÕ%W×eÝ÷W\É—÷Æt÷Ÿëñ¯[N·tõ¯íyÒ÷ëÈ÷qß%_Û¥[M?]ïGÚ’îþöåù„îãóêñ%îš"_ãC—tÇiÙë¬ióœ/í0-ßž£dëóM¾¶«¨|ç\[øa[º×s¶ï§mÇ¥®ûϼïëþn\¶>׿ƒÔýý[W;tÉ×û£¦pÐßç;ßÔôóòµý®øè‹Úþ|·ûßâÌt;\Å—ëu‚ïÏeëŽ{Û8‘ÈÕsc×ñZV®ý}—/®¯ìú©êz«¬lù”•/ë-[늦̫¾ÄGUÙŠs[ï—èö·UŸ+…~q#_×ù¶Ö“¾Þ_×-Wµž¶Þ×»Û÷E¾Ì£M¹´µ({_áZ¾´#QÝßCuùå•+ZŸ-é꯺¿/øG‰|]÷˜ª¯©ÏqÚ.ßÖ¹¾ÜßVå—ëëÚ·û‰¦ú›VÛÏO—|yßÅ—õ„®ë]׺L·l¯S|›ÿv¿+‹~¿Ë§ªòÚéÛýgÓî÷³ÇWß²õVÏ¢õç•«»¿ìñ¾pÝÔó ]ãä·v]‡¾¥®ä{ûvÉUÿûÒ?¶ãÍÔsE×ýXV®¯wÛãkzÜ«®×«¶§iñf[Mç|iGžLóÀ–\õ¯ëù¬jýºës5¿T­GW¹²¾¦ÇÕÖ8¸–oœ7¥¶ÝçùÒ_ä[?芟ºã¼k½]ôx_æQ[ëWÇÛRÕû,ÝõêöÝÕÝëº×WYߪ׋o|Läzž5}¿¡k¼ªÆ™«urž¿îøu½¯Êźó¿«yÓÔuš7®¶æ§ºí.zšŽ³ªõšâ_Þ÷»Æ¡j=eëoú}`Q5u^Í«O÷®zý˜žÏŠú•õ±U.™Ó×iÝ8(Ú]ózUÎ韺Çíª¯n»óê+{\Ñz‹ÆqÝøªËç]õõÕÕ?A·*;/›JmË×öêŠ{[2Ýo¾ÄKžtŸ¿nnW•îùÕ”lÅ]Ußê¶K×:J÷z¹l{ŠÖß45eþ´UoYÚRÑuuÝúu«îºÁ–o]×­§hýU·µ^3§U}‹§›3Eë÷uP´þºóïëé  2½kzº¯ ýè§Úïm9 ¿äz¾ñ¬W®û}_Æ%ôS9¹Ž»ïAAAAík®‡t¿Ò    Ór͹†Ôf”È5Ú’5C®ã$Ä[PÐåúz éö4¨˜\SHõ¤AA:å:ž÷=Þ÷­\ÅEÛú±ªšr½Ú÷º>®û)¨˜\Ïc¾Ž{ÛΫ©ãé{ûŠÊ·ópÝOMçŒk^úïAAAAAAAúäzÒ†4¤! i›ÒNç°ý9ètFiïbþM”<Œ¶h÷Á÷Qúf¦ÈITd9ýfvåß–ë<ŠËn¾?è'J2œd$É ’Ì0ÉŒ’Ì8ÉLâL·ÿDåú*G*Ç*'*7P¹¡ÊTn¬rʃ”)R¤ZÌ–_Ï/.æç—Éùov-æßÍ’¦õ>yûiœ}üÁìÅtõÎõm+Ïþ0ýj¶ˆ¿úù'W—Ëå|vöÅtñ<2¾=Ÿt‰·2%>›]¼òýƒ÷/çßM¯£ýϦ›}?dž%‹7Ã×:üp5û[Òòß-Ïç³åù,säéêêÅ»ÉÑ7¥ÿÑÙhó°ùU‹óE4,±ÅAÒkÓõôݯWÑñ[Zõú‹ËùõóÙË¿&=ÝÛ<Íî…(kF”õÆ£Ô·ƒÔ–ý¬¶ÛðQerʽ'Ç‹Ùw³E6|žFÝp‘tÓmDC±Xß}ü¿L,==_ߌÑ[ÚÞÍ´)ýYíÛÞ½{Û¸Ÿ$îIkO>¸Š®¡õü<‰¶£øœ?[iì¦ÄaR¢§N8ñ‡Ã?¿ü6‰µ{O¯ÖëÅìâö¸£ä¸ärjÃC[»ìG¬éüð:#cÍ`flexmix/data/BregFix.RData0000644000176200001440000000573513425024236015115 0ustar liggesusers‹•˜ XG€F.QDÄÖA *êÈôôt÷hbG<6&*n@ШD‘x¡‹’M<Ðõ`¿Õ5jâqƒHVÔÅ! ¢ÈaTä¹@¢‹ÉT•S3IßXïU½zÿ«W¯ª½§Îg{ÏïÍ0Œ9c.c~ý1½Ì»ÿ1c{›îÖjrH`дF6 [µèþÙuÿÒµcðhJgkA£^WQã´>R¢Ÿ5Ô±?O}ë¡o•Ô¸^׺è[…ihœ~Ðú°Ö0>ìßÝÐó‡IÌã¨ùœÄ|{j7Ž“^7òï,±>¥³=ëx>µ¿¸ßÃÐçŸ^Ê›T~è88ʵ C=Fù`LÛÑþŒú©8_£:£oí%ü ;–jTëHÍ“¨:¿˜Gå‹~´^”ºNÇÓP—òÿÚü!»‘”N׊C®oQ>é|¡y¨~P½éýí;=î§ïcòAù ë#ÿÖ”´.†êG-Z§TþGR:²s¤t䟚ã“Z/½RçLêþ’Sq*%ìC;Fÿ1ŠƒÞ/j]xéûÙË%úiŠÕº¯­©q¤£ü£¼Óþèý¡ïK䟡âCþÝ){ÄGõMÍÃvÖŒÉÏ÷¢xtÜtÝÒy¦üÓõŠû)އ֥êõ£ìézFöè` ã–â¢Çèw¤tªNp?ý]#õ^ ë“â+”ý{ßÓ´dÇP­#Õ*¨yu€ã¡ê—~Œê½—Ð8U'˜ãBéþ_—?‡\">?ʺP>ÃÏCõƒÎ§Þ¿Ñ¾KÄKû•¼Ÿé÷]ïôy¢ßô{…æRßwtþñºèïê}gTßÔ÷›Wb?é8èóèý’öÙ³†vŒþ=ªÂRoøVÞ†Ê ؆-‚ynÒ¼%ÿœ8y;ÈèÿžN™êSfyÜ; <*ˆ:ÄMž*&¬äFÜŸŠ|‹?Šš¨M‰©ÛNúŸcÜq´ þÒ:«†ÆXîâf•[xTÌæ¯ƒ9íàvú1øbøÎùǰëMQ0Ðô5ý¢wÃ+Þ›7:Æ€2ûàÉá-¹° òÁÓ?ï,wã’S>(_ +}òfMÊi„GN†¿HÛA¶­Í¢/†€KŸœ÷s­;,›`g35æíð›ù¼ÄV\Yå嵡è\ò2Æ_þfôÉhxâí :Uë yÂÂôÕ«Î|õ5x¬ÉRJ XX–Òž¨µ(÷ÞxÜ¡ &Û›|èû°DùοNWŸf¸_Ⱥ@«õŒó¹i÷Áómwl ³a]Ó#aä'N Óíêyð±´Ëµ>ÿŒ\ cÌÜhþ8¶F{6}ù?¨ÛÐÈnÕÑ;üúºNL¶ÌZÉ[¦Ý‚nßî¼E~v³V667~ý¸‡à™M­Ì~hº¾8pþÞx`0ÒGZ“,ý¯DÝǯ|½¤a”g2“¨pÞ•Rö?Öý’|œ½à¸/*V¯tÎ=ÖNþ!>z§Z¶ù O:uä¯:t)ö¯Ëá1ßy§ŸßŒOf‚£ö¾}uèZ¼rjQÿ/@Óû+›ud,ÊèÒnÛ®µôës¢ðƒ}°¶âÒØ³ÿ íÙ‘~—·Æ''³,ÔlPÕ瞃[Ÿ®ÊÞ]¯{;´s-n÷’ß IfÊ',Þýä"ÔmÏžž{â#¸×²·o»ã{Zæyõˆ‡ÓAEjÙ<[ÇÃp놾yñòáÍ-ü§n°~ð¿3Ué ÆÏûX¿üØ6ÓÆûÀJè–ß±;• Šü|¬ZçÃ}þwoä­€Í\ë[5øáБ} Ã.Áî9µü\Òµqê²|˜°ó—C‡ü`‘ë‹Î›Ì\½y}jKÌ-r½?£®lI‰=3J;Îþ=4a¼Ty³'‚F €×ÂïzLˆl…Ã,¾ŠÏ O*÷ÄNŽ ËJí¶ Ï·Ì*}[ JÝ9:ôÅmë:c_|WÌI²»òh*J4 v×<ß×öÚ7Èñh,ÍèZØWp:‡¹£Ä3¼sïç|æ +çÉu"È\ëþxéÛÏ`Ýø5£¿{"Ò?»9Ôs ¬†ÖýwÙ»ÁÊg3gäè´ÌÌMÂÁˆ‡°bÁ˜o#vµÂ»nvŒ¿ fލ‘11 »9u…yô9P7ÿÙü„­Ó@êˆkj­føÎÑ^ëþër禅ŒJvgÙÏÔÀ㧃w?ÚånñÚí{”¦ùjü< NhñêïŽ;©l_r$ù à3øƒš¹v¥ñu0[QçÒ²n ¬H»Ìí¿4¸‡[F;‚öΕq}f؀ʋv7ò¦'ÂÖ±•Û4^_ÂNù­P϶HÐìSå=shîxq˜k58ÿ÷϶3Ë\A£ÃÜÛ5 ¬ê@q: ¦îxc¬ƒ.>x¿øüø´“0÷Ô½°¤ã»à=g·u§ï­ƒÕm®[†/<Ÿýåö yÑ-P5<²ÖeÉAØ”¢ÚýN‰«=bZΆeÅM%Y@Ó„…î©Fƒ†y~ œ%3Ú7ßzWë?”9í·_¨•µt|gùÓ—4¦®àø2ðtÊÐÍî å åÉßÇŽaŸ´Ô;”ç~f‰ž?‡É‹Òœ& ªí³ÝúÇeÝ‚…—‡\¶: Ê3KÛìŽÁŠM‡Ä>ëAé1ß+2ÇÀf>ÔwÒ.3ø(>?¶Wó†‘õe~û{ÄLÿ37ÑÒ}¦ìÌLŒ™ÒÍMø”²§mLõ÷Ä–7Å—òÝÓšï<šýºœHÅKÛüÙ”Ÿž|þ‘|ÿ‘úx©9¦æIÍíiO}™ðeðßÃKW¬_ß-Øë _vZ.Xº6„2µ\¸Ùšëmͼðòƒ·³û‡Ž¡Œ ÖMm²yÝòMm‰L¯Ë^‘i[ÚF&1.Å|Õ^Ê—Œ7Å6cŒcíI¦·›¶3cŒ×c*‡=åˆ^#½V:ÚÎÔÚ{ê“bKí_O¼×9©ý—ºBzÚKzo¤ZSûü Óð¬ DGËBß)‹øµë¥h¼V/™mDB8u„ Ú„¬ W¼ê4yå˜þ*°HP!C‚ <$ˆHРД^XRb‰Å’ K–ÔXâ±$`IÄf°˜Áb‹,f°˜Áb‹,f°˜Áb† 3T˜¡Â f¨0C…*ÌPa† 3T˜Áa‡fp˜Áa‡fp˜Áa‡jÌPc†3Ô˜¡Æ 5f¨1CjÌPctž7Û0Z?éÿðËëA߸÷óÌüN=3•˜JM]1™úÊÔþ’©µ¦fM}J–ÿ™©Ç¦>0Õ2õ…©ÏM]eV™ºnêëùü‡¦fLå—÷}~M³þ}¾­ S5ª wŸóê¦>avŽsVFûúêð­±=Ö+¬WY{Ÿïq f›k™.o}?͹§ ÞãúÇ™™+ÌØšD‹¯JAËÙØëÕÒœ}ÕÙ»]Ÿ*\›±{ðJ¾È*1復Sù>š¨­âz»Ð«¯AMÖ›ž{omµ©æ9¼bµ/X/Ï5yOZ¥:ï<íRÖ(”Ý£íí}Z|/j¥ª—Ö뜿Vº–™Òj…cêçTëdv?V§/˜k®­Æ[ß*øófíº¹ßêö¾)ßWÿfý_ÿõzëc¬vÔÿ±tbVgHóÕ±åc\u«R™x~”؉ ÇÕ^dÏOƽ1+ùïÙÍ?½ákÁC'Ñtý½„^FùlY¸‚^E¿D¯¡_£óhõQ…j4@C4Bc4ASôô:ú-zý½‰.Xýu¤·èoÑߦ¿M‡þý"ý"ý]ú»ô÷èïÑß§¿O¿D¿Dÿ€þýCú‡ôèÑ?¦Lÿ„þ ý2ý2ýSú§ôÏèŸ- vý÷‘®¯¯¯¯’¯’¯’¯’¯‘¯‘¯‘¯‘wÉ»ä]ò.ù:ù:ù:ù:ùùùùù&ù&ù&ù&y¼GÞ#ï‘o‘o‘o‘o‘o“o“o“o“ïïïïï’ï’ï’ï’ï‘ï‘ï‘ï‘ï“ï“ï“ï“’^F9îðåøÃßP;÷ÇH3æ3æ3æ3æ³Ñ|ù±ô*{Ù·¥qK“k£oé'ôKǯ¶itýìôeÿÕ©}Ø-f?—ÎÚÿ4_<óÇV|k”5ÚšÀšÐšÈšØšÄšÔ>sýŽ8_œ§ÅâBq‘¸X\"NJJJJJJJJJJZZZZZZZZZZ0a„#F Œ@0a„Â… #F(ŒP¡0Ba„Â… #F$ŒH‘0"aDˆ„ #F,ŒX±0baĈ… #F,ŒX‰0a$ÂH„‘#F"ŒD‰0a¤ÂH…‘ #F*ŒT©0Ra¤ÂH-£êw:ÎúÎ*gµ³³¡³‘³±³‰³Žæ;šïh¾£ùŽæ;šïh¾£ùŽæ;šïhÊÑ”£)GSަM9šr4åhÊÑ”£iGÓŽ¦M;šv4íhÚÑ´£iGÓŽ6| ˜wõ/ïí/¦Íï³ÓlþÅÀ<½á—SïÍßÒi.õÝflexmix/data/NregFix.RData0000644000176200001440000001017613425024236015124 0ustar liggesusers‹Øy\Œ[ð©i›"º5Z´i™}&qNq-7KâZnH‘HDBÉ’rI–**‘êf 9oáUZnEšVJÊ´£¼y¯ÎqûTzß?žÏó{ž9s¾çœçœyžgì§.f+.V¤ÑhÒ4iiMJ¦;Êtš¦ÊèÞËÏñtq¶Ö›F£«uÊvoCº·,PÜ$ø‹ú!ô"9¯Ù{ªT ‰ÉìðFõí9ºu·@ù ¶ûÜI Rg㥦<ðax^â‘sàÝLù×÷Ë ¢=œ!šŸ¦U}YÞ(ÛÇ7.Ÿ‡Êf7n}© ššBíL=% #õ>ÉF” 3'ñÐMôvšëÇ”õw@¹ðYÃè‚3¨™¼Î:òwÔ6êñx)W%P³?5Tôe2êç(÷(Ø•’Ê­¬¸¥j]¼AÁÅ $>41(Úæ h:ô{~°D½˜>¦VX‹^Õ_ÞÍÝ6Ô;&m:ÜMGÖ& ó ÍtÍF¡wÿb,«ª£¦>¯n¤xƒÊ«šŸ¼µÜ_²ï`¨UöÒd.ƒÊ1s²PÇhð"7 ¹AÍ5&݇?ÿ‰êu,Ÿ6k)Eßkm“å.F#f¤µÇMB¹ÆÚk˜­|T&Å×¼>Ó´–åF¶çׂςx»µk'ö‹œxEH2uQk›O(Jg¼¹VqóŠZõb 5>eX™|ì@9k´ÌäâæçêÎ3KÑ»%S›7ÕÕ€úÓ£Îî:Œù>.ûkM5ªˆ4Ý(¯T®£  (ðæMÖÉÑíIN”~ÑóA{ì½'þЦ¨9tƒÞ‚ |T²ÊoKi&ÕD~Xg¯ òíè³:ý5Qº¿ÈÏ<2Ò¼ôØ'ê¨Ñ¹fg1¤»åïz­´U¬-;8wx4héºblx¤U%uªš*€¶w«g3TÁ}ÍßJ—-IðÎÄYiút,sò„EÍ(ßÃÞºU¨j#ˆ:÷ÒlÞ´0¶¡¸öyë’g_-+N£'[#ÿCùŒ”ŒÐûkNY1Pùk» ‚Ai×Ã-›¯£ñܳ©3ï Êµ^³o•ù¡ú¦”§§&¡ŠÇg5&6F£ê¶=·gì¶âG3ó¬T¿Ody‚æ=ÚNK?A±Ë–J ª~BÍç²}-ï}åò£¼•èUàSr¶‰oÑX$¾ihr$Ù $¼ðH«üå ÊÙßÞùüõáʸas§£ Ã’œ'Œ6ôåá¶âÍ)[@i‰rQˆa]ÔJŸg müZ‡{ú¨¼šÎà®._¼âqÌv)ÎôZä9ÕëdÉì_³$Zå^¾3Dè=Ðj, ¯‘¹ÛïëPÃÕûU‹#Q÷_žµ·AÍä¸,C»¨â·ê¨é‘Ψ*c~gë§K ~Ò©˜òaµ¨Ò=6zõjkôÙ¿/¹îá¦Ñ’Ç(GlcÊ 8ZŸÂL”æf°sâ4ô–6¡EÚ‚*/É^UWUTµýAvvÐtt®z.ø¼H:àî=kФ²M¼wc*÷žŸ~eŠ%E ¹uúfê2 QõÑLÙÊŸ7–éBU[˜›i—PñdKy¨„ynï…GAEYÉ™¡Áó!mÍÔ‰|öÐ1¶ë¡0Àu•]kE?ÉŠwŽcQKÝéä}TuÇu©•Ÿ¼îˆÚ%¥Š>úÿn|còc¹Ý´FËÿjណ\.‹ê3¹çd½@&Ov²n…’Äg,ót6µö>š‡‹¡ûÏ#vU;Î妄«c€œëÏíÊÏ@•“âY—cm@³ô>ÇCåÑ«MQZÜ5@à¼M;O5ƒ‚õzž¿”NCÏçø§¹ÏQCÅæ ›$®UW×U®f ¦ú¾â´  óÜÎÐf†;xy´0JÃùê ­+i•š²?R…ÇìÐËwìo†É‚fg]ëÏ= ]’Z«j½Š¢X]ÜTÏDåæ3b)ePcýÎ’Yz<ÿµ2°pV8È©ÿã÷SNÕ Ãr]yÃE¤Ô°†Úæì‹¾Ï§!ñâiWî¨ïë^ß;íväÞCu.ê3n$%Lþm¤¸Y%gÆs-By¦º1ÛÍ"‰”A¼Eºåù{ÎͲ=†¨ëùŽrC t!˜¸Ž÷5veŸ¼/ûå§¹/ žŠÔÏjŠo'€Ö:7ºòâ7è {v×<£sèíºëzüÛPG(—;ßGD˜¢¸*¢Eïa›Ll_V-a(UdϸÉ.K@÷ýDåë}äï[ íÛ-gé^ùûÏÚ~T¾/§wÙ¾ìÞmê]×@ßÌçýÕÙ»½ýõu ­/·¿ö÷Õ¦úÔ×µ¨ÿ˵ûÑ8þ¿se0ýÌXþh|û¯µ§WÝÿx“swÙêâ¾¹;©~+ñõ¬”EO`õ*/»ÒÝisOq©o¥äV;­ôòðìN]ß-Ã5õG—e°Ó¢¯Ë<Ø!hjþ/Kw0S¹¿ú~4e;]§–Í@?+ƒ×–Ì`–ï@eû»fýåÁ,ãþÆ`°ûÁÌ™þúùÝöuYÑ¿ÐúYI_?•îù´Ïeùu­IÿýêTeUk1Ï™;AmøœQ­Ô0Åêã!Ñ©”zN í'¿³pÜ 5_ )††'ùÇe¡Ìí‰ËÞ5†f&wKß}™vÓ¤ž4B®À>êmS¨ÔXq-wßj˜PK¢Ù5ÝÆ¹Õê¦è”Ö3¡Æ#ÿT§ϧFwÕ©ÏY >½“®¼JMŽEÇŽ²(Ó‰¹¾±/¡ÖîƒöÓç·cš>å˜l4¢ìr¦&b-îïãÓ° êªt-J߆̠,?™e”•ίAÊv2éW©Êi’‡´£ÞEAÉ“ ù/驳v:Pºü§áyåש±ÖŽ…{/ƒz¬´åìR)Ŧœ»±…Û ¢ÚÏô-Y ó£™AaAÔ·ÏÕnY>’Ó¬¯ ô£ÀGùèK1»^BígïÚlôŒÑs;Ϙ[4c(3z÷/΋§Bz’[ò _eeèÐæNé¼0pï\u¦ˆ^¢r’Òœ½U  Î^ÞÓŒ,[¨ôjÔ†Í-^Ô¸á jźvÐRér íãN¨òÖuÙ½˜Û ëÌŠCËl¡qÍ˰™»‚ œ'Kª› ZÞõøÒd¸g¬´ðÌ{²5´³»¹û“ê-þÕÞòë(û1%m=Ô}f¢ÓwŽÌ{hä Ù!«^ …b5¹~ºP>¬Úä߯) …'_¨m3aŸÚ˜lÈ\ï]Xô,jÞùu7/9JÓÛÔO‡Aå «eÛ?uBîѸ´-k¶CùBÏû7zCÃ!I>§FÊÃ!·ØzûYÀ¡;'ÕÂp-8F"ìŒ7ƒÌ5‡fEÎØL±¬ôïýqÐ4nûëãËtÈ«°î|p#—½a7îQº%nòVPåVÃcq¤ŸªÊÝøjdrœA­Ý¥îuÌ(›qA¹Y“=®¹@Ê|×-ÿËà˜Ù4ãÚ7PsÝ~ë?ý~ƒæ›òžr,m Ü˜Dóü3ƽñè«Ö®ïÞ©|»ñ|{£úG…î"œÖ»|}TûV®‹ÜÊþ[FÊçÛÉï_˾¿Ò{Ýé}ìû*ß“é½rïã¾ÊöuÜWzŸëý @ïèßKõÓ×þúÔó¾lé^õõ~ÎèmöUoï>÷>×W}½ë–îÃé¯_½Ç>À¾¯±ýѵø¾Ü@eú›/}õ¥¯ëÕ×õî¯Ý¹ßÚøÏ׬žÕõuE}ýüëIioNìoIjÛwËíohÿ¨áé±Íìû:³hß=Iþ7à9=Ûx=ß=AØD=McYà„›ËbãÄÁ‰‹'>Nœ„8aƒ 66ØØ`cƒ 66ØØ`cƒ 668Øà`ƒƒ 68Øà`ƒƒ 68Øà`ƒ‹ .6¸Øàbƒ‹ .6¸Øàbƒ‹ .6xØàaƒ‡ 6xØàaƒ‡ 6xØàaƒ >6øØàcƒ >6øØàcƒ >6Ø`C€ 6Ø`C€ 6Ø`Cˆ !6„ØbCˆ !6„ØbCˆ !6DØaC„ 6DØaC„ 6DØõt–…‰,Ù$rHä’È#‘O¢€D!‰DcE4ÑXDcE4ÑXDcE46ÑØDcM46ÑØDcM46ÑØDãC4Ñ8DãC4Ñ8DãC4.ѸDãK4.ѸDãK4.ѸDãG4ÑxDãG4ÑxDãG4>ÑøDãO4>ÑøDãO4>ÑøDM@4ÑDM@4ÑDM@4!Ñ„DMH4!Ñ„DMH4!Ñ„DMD4ÑDDMD4ÑDDü–°¿þ–ôùg•â*'/'³ÕžÝ´¿õ»þ«í5m flexmix/R/0000755000176200001440000000000013432516322012130 5ustar liggesusersflexmix/R/rFLXmodel.R0000644000176200001440000000566713431371006014122 0ustar liggesuserssetMethod("rFLXM", signature(model="FLXM", components="list"), function(model, components, class, ...) { y <- NULL for (l in seq_along(components)) { yl <- as.matrix(rFLXM(model, components[[l]], ...)) if (is.null(y)) y <- matrix(NA, nrow = length(class), ncol = ncol(yl)) y[class == l,] <- yl[class==l,] } y }) setMethod("rFLXM", signature(model = "FLXMRglm", components="FLXcomponent"), function(model, components, ...) { family <- model@family n <- nrow(model@x) if(family == "gaussian") { sigma <- components@parameters$sigma y <- rnorm(n, mean = components@predict(model@x, ...), sd = sigma) } else if (family == "binomial") { dotarg = list(...) if ("size" %in% names(dotarg)) size <- dotarg$size else { if (nrow(model@y)!=n) stop("no y values - specify a size argument") size <- rowSums(model@y) } parms <- components@parameters y <- rbinom(n, prob = components@predict(model@x, ...), size=size) y <- cbind(y, size - y) } else if (family == "poisson") { y <- rpois(n, lambda = components@predict(model@x, ...)) } else if (family == "Gamma") { shape <- components@parameters$shape y <- rgamma(n, shape = shape, scale = components@predict(model@x, ...)/shape) } else stop("family not supported") y }) setMethod("rFLXM", signature(model = "FLXMRglmfix", components="list"), function(model, components, class, ...) { k <- sum(model@nestedformula@k) n <- nrow(model@x)/k y <- matrix(NA, nrow = length(class), ncol = ncol(model@y)) model.sub <- as(model, "FLXMRglm") for (l in seq_len(k)) { rok <- (l-1)*n + seq_len(n) model.sub@x <- model@x[rok, as.logical(model@design[l,]), drop=FALSE] model.sub@y <- model@y[rok,,drop=FALSE] yl <- as.matrix(rFLXM(model.sub, components[[l]], ...)) y[class==l,] <- yl[class==l,] } y }) rmvbinom <- function(n, size, prob) sapply(prob, function(p) rbinom(n, size, p)) rmvbinary <- function(n, center) sapply(center, function(p) rbinom(n, 1, p)) setMethod("rFLXM", signature(model = "FLXMC", components = "FLXcomponent"), function(model, components, class, ...) { rmvnorm <- function(n, center, cov) mvtnorm::rmvnorm(n = n, mean = center, sigma = cov) dots <- list(...) FUN <- paste("r", model@dist, sep = "") args <- c(n = nrow(model@x), dots, components@parameters) return(do.call(FUN, args)) }) flexmix/R/rflexmix.R0000644000176200001440000000742113425024235014114 0ustar liggesuserssetMethod("rflexmix", signature(object = "FLXdist", newdata="numeric"), function(object, newdata, ...) { newdata <- data.frame(matrix(nrow = as.integer(newdata), ncol = 0)) rflexmix(object, newdata = newdata, ...) }) setMethod("rflexmix", signature(object = "FLXdist", newdata="listOrdata.frame"), function(object, newdata, ...) { groups <- .FLXgetGrouping(object@formula, newdata) object@model <- lapply(object@model, FLXgetModelmatrix, newdata, object@formula, lhs=FALSE) group <- if (length(groups$group)) groups$group else factor(seq_len(FLXgetObs(object@model[[1]]))) object@concomitant <- FLXgetModelmatrix(object@concomitant, data = newdata, groups = list(group=group, groupfirst = groupFirst(group))) rflexmix(new("flexmix", object, group=group, weights = NULL), ...) }) setMethod("rflexmix", signature(object = "flexmix", newdata="missing"), function(object, newdata, ...) { N <- length(object@model) object <- undo_weights(object) group <- group(object) prior <- determinePrior(object@prior, object@concomitant, group) class <- apply(prior, 1, function(x) rmultinom(1, size = 1, prob = x)) class <- if (is.matrix(class)) t(class) else as.matrix(class) class <- max.col(class)[group] y <- vector("list", N) for (i in seq_len(N)) { comp <- lapply(object@components, function(x) x[[i]]) yi <- rFLXM(object@model[[i]], comp, class, group, ...) form <- object@model[[i]]@fullformula names <- if(length(form) == 3) form[[2]] else paste("y", i, seq_len(ncol(yi)), sep = ".") if (ncol(yi) > 1) { if (inherits(names, "call")) names <- as.character(names[-1]) if (length(names) != ncol(yi)) { if (length(names) == 1) names <- paste(as.character(names)[1], i, seq_len(ncol(yi)), sep = ".") else stop("left hand side not specified correctly") } } else if (inherits(names, "call")) names <- deparse(names) colnames(yi) <- as.character(names) y[[i]] <- yi } list(y = y, group=group, class = class) }) ###********************************************************** determinePrior <- function(prior, concomitant, group) { matrix(prior, nrow = length(unique(group)), ncol = length(prior), byrow=TRUE) } setGeneric("determinePrior", function(prior, concomitant, group) standardGeneric("determinePrior")) setMethod("determinePrior", signature(concomitant="FLXPmultinom"), function(prior, concomitant, group) { exps <- exp(concomitant@x %*% concomitant@coef) exps/rowSums(exps) }) undo_weights <- function(object) { if (!is.null(object@weights)) { for (i in seq_along(object@model)) { object@model[[i]]@x <- apply(object@model[[i]]@x, 2, rep, object@weights) object@model[[i]]@y <- apply(object@model[[i]]@y, 2, rep, object@weights) object@concomitant@x <- apply(object@concomitant@x, 2, rep, object@weights) } if (length(object@group) > 0) object@group <- rep(object@group, object@weights) object@weights <- NULL } object } ###********************************************************** setMethod("simulate", signature("FLXdist"), function(object, nsim, seed = NULL, ...) { if (!exists(".Random.seed", envir = .GlobalEnv, inherits = FALSE)) runif(1) if (is.null(seed)) RNGstate <- get(".Random.seed", envir = .GlobalEnv, inherits = FALSE) else { R.seed <- get(".Random.seed", envir = .GlobalEnv, inherits = FALSE) set.seed(seed) RNGstate <- structure(seed, kind = as.list(RNGkind())) on.exit(assign(".Random.seed", R.seed, envir = .GlobalEnv, inherits = FALSE)) } ans <- lapply(seq_len(nsim), function(i) rflexmix(object, ...)$y) if (all(sapply(ans, ncol) == 1)) ans <- as.data.frame(ans) attr(ans, "seed") <- RNGstate ans }) flexmix/R/utils.R0000644000176200001440000000321713425024235013415 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: utils.R 5079 2016-01-31 12:21:12Z gruen $ # list2object = function(from, to){ n = names(from) s = slotNames(to) p = pmatch(n, s) if(any(is.na(p))) stop(paste("\nInvalid slot name(s) for class", to, ":", paste(n[is.na(p)], collapse=" "))) names(from) = s[p] do.call("new", c(from, Class=to)) } printIter = function(iter, logLik, label="Log-likelihood") cat(formatC(iter, width=4), label, ":", formatC(logLik, width=12, format="f"),"\n") ## library(colorspace) ## dput(x[c(1,3,5,7,2,4,6,8)]) ## x = hcl(seq(0, 360*7/8, length.out = 8), c=30) LightColors <- c("#F9C3CD", "#D0D4A8", "#9DDDD5", "#D1CCF5", "#EDCAB2", "#AFDCB8", "#ACD7ED", "#EFC4E8") ## x = hcl(seq(0, 360*7/8, length.out = 8), c=100, l=65) FullColors <- c("#FF648A", "#96A100", "#00BCA3", "#9885FF", "#DC8400", "#00B430", "#00AEEF", "#F45BE1") ###********************************************************** ## similar defaults to silhouette plots in flexclust unipolarCols <- function(n, hue=0, chr=50, lum = c(55, 90)) { lum <- seq(lum[1], lum[2], length=n) hcl(hue, chr, lum) } bipolarCols <- function(n, hue=c(10, 130), ...) { if(n%%2){ # n odd n2 <- (n-1)/2 c1 <- unipolarCols(n2, hue[1]) c2 <- rev(unipolarCols(n2, hue[2])) return(c(c1, "white", c2)) } else{ # n even n2 <- n/2 c1 <- unipolarCols(n2, hue[1]) c2 <- rev(unipolarCols(n2, hue[2])) return(c(c1, c2)) } } ###********************************************************** flexmix/R/models.R0000644000176200001440000002641213425024235013542 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: models.R 5079 2016-01-31 12:21:12Z gruen $ # FLXMRglm <- function(formula=.~., family=c("gaussian", "binomial", "poisson", "Gamma"), offset=NULL) { family <- match.arg(family) glmrefit <- function(x, y, w) { fit <- c(glm.fit(x, y, weights=w, offset=offset, family=get(family, mode="function")()), list(call = sys.call(), offset = offset, control = eval(formals(glm.fit)$control), method = "weighted.glm.fit")) fit$df.null <- sum(w) + fit$df.null - fit$df.residual - fit$rank fit$df.residual <- sum(w) - fit$rank fit$x <- x fit } z <- new("FLXMRglm", weighted=TRUE, formula=formula, name=paste("FLXMRglm", family, sep=":"), offset = offset, family=family, refit=glmrefit) z@preproc.y <- function(x){ if (ncol(x) > 1) stop(paste("for the", family, "family y must be univariate")) x } if(family=="gaussian"){ z@defineComponent <- function(para) { predict <- function(x, ...) { dotarg = list(...) if("offset" %in% names(dotarg)) offset <- dotarg$offset p <- x %*% para$coef if (!is.null(offset)) p <- p + offset p } logLik <- function(x, y, ...) dnorm(y, mean=predict(x, ...), sd=para$sigma, log=TRUE) new("FLXcomponent", parameters=list(coef=para$coef, sigma=para$sigma), logLik=logLik, predict=predict, df=para$df) } z@fit <- function(x, y, w, component){ fit <- lm.wfit(x, y, w=w, offset=offset) z@defineComponent(para = list(coef = coef(fit), df = ncol(x)+1, sigma = sqrt(sum(fit$weights * fit$residuals^2 / mean(fit$weights))/ (nrow(x)-fit$rank)))) } } else if(family=="binomial"){ z@preproc.y <- function(x){ if (ncol(x) != 2) stop("for the binomial family, y must be a 2 column matrix\n", "where col 1 is no. successes and col 2 is no. failures") if (any(x < 0)) stop("negative values are not allowed for the binomial family") x } z@defineComponent <- function(para) { predict <- function(x, ...) { dotarg = list(...) if("offset" %in% names(dotarg)) offset <- dotarg$offset p <- x %*% para$coef if (!is.null(offset)) p <- p + offset get(family, mode = "function")()$linkinv(p) } logLik <- function(x, y, ...) dbinom(y[,1], size=rowSums(y), prob=predict(x, ...), log=TRUE) new("FLXcomponent", parameters=list(coef=para$coef), logLik=logLik, predict=predict, df=para$df) } z@fit <- function(x, y, w, component){ fit <- glm.fit(x, y, weights=w, family=binomial(), offset=offset, start=component$coef) z@defineComponent(para = list(coef = coef(fit), df = ncol(x))) } } else if(family=="poisson"){ z@defineComponent <- function(para) { predict <- function(x, ...) { dotarg = list(...) if("offset" %in% names(dotarg)) offset <- dotarg$offset p <- x %*% para$coef if (!is.null(offset)) p <- p + offset get(family, mode = "function")()$linkinv(p) } logLik <- function(x, y, ...) dpois(y, lambda=predict(x, ...), log=TRUE) new("FLXcomponent", parameters=list(coef=para$coef), logLik=logLik, predict=predict, df=para$df) } z@fit <- function(x, y, w, component){ fit <- glm.fit(x, y, weights=w, family=poisson(), offset=offset, start=component$coef) z@defineComponent(para = list(coef = coef(fit), df = ncol(x))) } } else if(family=="Gamma"){ z@defineComponent <- function(para) { predict <- function(x, ...) { dotarg = list(...) if("offset" %in% names(dotarg)) offset <- dotarg$offset p <- x %*% para$coef if (!is.null(offset)) p <- p + offset get(family, mode = "function")()$linkinv(p) } logLik <- function(x, y, ...) dgamma(y, shape = para$shape, scale=predict(x, ...)/para$shape, log=TRUE) new("FLXcomponent", parameters = list(coef = para$coef, shape = para$shape), predict = predict, logLik = logLik, df = para$df) } z@fit <- function(x, y, w, component){ fit <- glm.fit(x, y, weights=w, family=Gamma(), offset=offset, start=component$coef) z@defineComponent(para = list(coef = coef(fit), df = ncol(x)+1, shape = sum(fit$prior.weights)/fit$deviance)) } } else stop(paste("Unknown family", family)) z } ###********************************************************** FLXMCmvnorm <- function(formula=.~., diagonal=TRUE) { z <- new("FLXMC", weighted=TRUE, formula=formula, dist = "mvnorm", name="model-based Gaussian clustering") z@defineComponent <- function(para) { logLik <- function(x, y) mvtnorm::dmvnorm(y, mean=para$center, sigma=para$cov, log=TRUE) predict <- function(x, ...) matrix(para$center, nrow=nrow(x), ncol=length(para$center), byrow=TRUE) new("FLXcomponent", parameters=list(center = para$center, cov = para$cov), df=para$df, logLik=logLik, predict=predict) } z@fit <- function(x, y, w, ...){ para <- cov.wt(y, wt=w)[c("center","cov")] para$df <- (3*ncol(y) + ncol(y)^2)/2 if(diagonal){ para$cov <- diag(diag(para$cov)) para$df <- 2*ncol(y) } z@defineComponent(para) } z } FLXMCnorm1 <- function(formula=.~.) { z <- new("FLXMC", weighted=TRUE, formula=formula, dist = "mvnorm", name="model-based univariate Gaussian clustering") z@defineComponent <- function(para) { logLik <- function(x, y) dnorm(y, mean=para$center, sd=sqrt(para$cov), log=TRUE) predict <- function(x, ...) matrix(para$center, nrow=nrow(x), ncol=1, byrow=TRUE) new("FLXcomponent", parameters=list(mean = as.vector(para$center), sd = as.vector(sqrt(para$cov))), df=para$df, logLik=logLik, predict=predict) } z@fit <- function(x, y, w, ...){ para <- cov.wt(as.matrix(y), wt=w)[c("center","cov")] z@defineComponent(c(para, list(df = 2))) } z } ###********************************************************** FLXMCmvbinary <- function(formula=.~., truncated = FALSE) { if (truncated) return(MCmvbinary_truncated(formula)) else return(MCmvbinary(formula)) } MCmvbinary <- function(formula=.~.) { z <- new("FLXMC", weighted=TRUE, formula=formula, dist = "mvbinary", name="model-based binary clustering") ## make sure that y is binary z@preproc.y <- function(x){ storage.mode(x) <- "logical" storage.mode(x) <- "integer" x } z@defineComponent <- function(para) { predict <- function(x, ...){ matrix(para$center, nrow=nrow(x), ncol=length(para$center), byrow=TRUE) } logLik <- function(x, y){ p <- matrix(para$center, nrow=nrow(x), ncol=length(para$center), byrow=TRUE) rowSums(log(y*p+(1-y)*(1-p))) } new("FLXcomponent", parameters=list(center=para$center), df=para$df, logLik=logLik, predict=predict) } z@fit <- function(x, y, w, ...) z@defineComponent(list(center = colSums(w*y)/sum(w), df = ncol(y))) z } ###********************************************************** binary_truncated <- function(y, w, maxit = 200, epsilon = .Machine$double.eps) { r_k <- colSums(y*w)/sum(w) r_0 <- 0 llh.old <- -Inf for (i in seq_len(maxit)) { p <- r_k/(1+r_0) llh <- sum((r_k*log(p))[r_k > 0])+ sum(((1 - r_k + r_0) * log(1-p))[(1-r_k+r_0) > 0]) if (abs(llh - llh.old)/(abs(llh) + 0.1) < epsilon) break llh.old <- llh prod_p <- prod(1-p) r_0 <- prod_p/(1-prod_p) } p } MCmvbinary_truncated <- function(formula=.~.) { z <- MCmvbinary(formula=formula) z@defineComponent <- function(para) { predict <- function(x, ...) { matrix(para$center, nrow = nrow(x), ncol = length(para$center), byrow = TRUE) } logLik <- function(x, y) { p <- matrix(para$center, nrow = nrow(x), ncol = length(para$center), byrow = TRUE) rowSums(log(y * p + (1 - y) * (1 - p))) - log(1 - prod(1-para$center)) } new("FLXcomponent", parameters = list(center = para$center), df = para$df, logLik = logLik, predict = predict) } z@fit <- function(x, y, w, ...){ z@defineComponent(list(center = binary_truncated(y, w), df = ncol(y))) } z } ###********************************************************** setClass("FLXMCmvcombi", representation(binary = "vector"), contains = "FLXMC") FLXMCmvcombi <- function(formula=.~.) { z <- new("FLXMCmvcombi", weighted=TRUE, formula=formula, dist = "mvcombi", name="model-based binary-Gaussian clustering") z@defineComponent <- function(para) { predict <- function(x, ...){ matrix(para$center, nrow=nrow(x), ncol=length(para$center), byrow=TRUE) } logLik <- function(x, y){ if(any(para$binary)){ p <- matrix(para$center[para$binary], nrow=nrow(x), ncol=sum(para$binary), byrow=TRUE) z <- rowSums(log(y[,para$binary,drop=FALSE]*p + (1-y[,para$binary,drop=FALSE])*(1-p))) } else z <- rep(0, nrow(x)) if(!all(para$binary)){ if(sum(!para$binary)==1) z <- z + dnorm(y[,!para$binary], mean=para$center[!para$binary], sd=sqrt(para$var), log=TRUE) else z <- z + mvtnorm::dmvnorm(y[,!para$binary,drop=FALSE], mean=para$center[!para$binary], sigma=diag(para$var), log=TRUE) } z } new("FLXcomponent", parameters=list(center=para$center, var=para$var), df=para$df, logLik=logLik, predict=predict) } z@fit <- function(x, y, w, binary, ...){ para <- cov.wt(y, wt=w)[c("center","cov")] para <- list(center = para$center, var = diag(para$cov)[!binary], df = ncol(y) + sum(!binary), binary = binary) z@defineComponent(para) } z } setMethod("FLXgetModelmatrix", signature(model="FLXMCmvcombi"), function(model, data, formula, lhs=TRUE, ...) { model <- callNextMethod(model, data, formula, lhs) model@binary <- apply(model@y, 2, function(z) all(unique(z) %in% c(0,1))) model }) setMethod("FLXmstep", signature(model = "FLXMCmvcombi"), function(model, weights, components) { return(sapply(seq_len(ncol(weights)), function(k) model@fit(model@x, model@y, weights[,k], model@binary))) }) flexmix/R/plot-refit.R0000644000176200001440000001125013425024235014336 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: plot-refit.R 5079 2016-01-31 12:21:12Z gruen $ # prepanel.default.coef <- function (x, y, subscripts, groups=NULL, horizontal = TRUE, nlevels, origin = NULL, ...) { if (any(!is.na(x) & !is.na(y))) { if (horizontal) { if (!is.factor(y)) { if (missing(nlevels)) nlevels <- length(unique(y)) y <- factor(y, levels = seq_len(nlevels)) } if (!is.null(groups)) { if (!is.numeric(x)) stop("x must be numeric") x <- rep(x, each = 2) + rep(groups[subscripts], each = 2) *c(-1,1) } list(xlim = if (is.numeric(x)) range(x, origin, finite = TRUE) else levels(x), ylim = levels(y), yat = sort(unique(as.numeric(y))), dx = 1, dy = 1) } else { if (!is.factor(x)) { if (missing(nlevels)) nlevels <- length(unique(x)) x <- factor(x, levels = seq_len(nlevels)) } if (!is.null(groups)) { if (!is.numeric(y)) stop("y must be numeric") y <- rep(as.numeric(y), each = 2) + rep(groups[subscripts], each = 2) *c(-1,1) } list(xlim = levels(x), xat = sort(unique(as.numeric(x))), ylim = if (is.numeric(y)) range(y, origin, finite = TRUE) else levels(y), dx = 1, dy = 1) } } else list(xlim = c(NA, NA), ylim = c(NA, NA), dx = 1, dy = 1) } panel.coef <- function(x, y, subscripts, groups, significant = NULL, horizontal = TRUE, lwd = 2, col, col.line = c("black", "grey"), ...) { col.sig <- rep(col.line[1], length(x)) if (!is.null(significant)) { if (missing(col)) col <- c("grey", "white") col.fill <- rep(col[1], length(x)) col.sig[!significant[subscripts]] <- col.line[2] col.fill[!significant[subscripts]] <- col[2] } else if (missing(col)) col.fill <- "grey" else col.fill <- col panel.barchart(x, y, border = col.sig, col = col.fill, horizontal = horizontal, ...) if (!missing(groups)) { if (horizontal) { z <- x + rep(c(-1,1), each = length(x)) * matrix(rep(groups[subscripts], 2), ncol = 2) for (i in seq_along(x)) { panel.xyplot(z[i,], rep(y[i], 2), type = "l", col = col.sig[i], lwd = lwd) } } else { z <- y + rep(c(-1,1), each = length(y)) * matrix(rep(groups[subscripts], 2), ncol = 2) for (i in seq_along(y)) { panel.xyplot(rep(x[i], 2), z[i,], type = "l", col = col.sig[i], lwd = lwd) } } } } getCoefs <- function(x, alpha = 0.05, components, ...) { names(x) <- sapply(names(x), function(z) strsplit(z, "Comp.")[[1]][2]) x <- x[names(x) %in% components] Comp <- lapply(names(x), function(n) data.frame(Value = x[[n]][,1], SD = x[[n]][,2] * qnorm(1-alpha/2), Variable = rownames(x[[n]]), Component = n, Significance = x[[n]][,4] <= alpha)) do.call("rbind", Comp) } setMethod("plot", signature(x="FLXRoptim", y="missing"), function(x, y, model = 1, which = c("model", "concomitant"), bycluster=TRUE, alpha=0.05, components, labels=NULL, significance = FALSE, xlab = NULL, ylab = NULL, ci = TRUE, scales = list(), as.table = TRUE, horizontal = TRUE, ...) { which <- match.arg(which) if (missing(components)) components <- seq_len(x@k) plot.data <- if (which == "model") getCoefs(x@components[[model]], alpha, components) else getCoefs(x@concomitant, alpha, components) if (!is.null(labels)) plot.data$Variable <- factor(plot.data$Variable, labels = labels) plot.data$Component <- with(plot.data, factor(Component, sort(unique(Component)), labels = paste("Comp.", sort(unique(Component))))) if (bycluster) { formula <- if (horizontal) Variable ~ Value | Component else Value ~ Variable | Component plot.data$Variable <- with(plot.data, factor(Variable, levels = rev(unique(Variable)))) } else { formula <- if (horizontal) Component ~ Value | Variable else Value ~ Component | Variable plot.data$Component <- with(plot.data, factor(Component, levels = rev(levels(Component)))) } groups <- if (ci) plot.data$SD else NULL significant <- if (significance) plot.data$Significance else NULL xyplot(formula, data = plot.data, xlab = xlab, ylab = ylab, origin = 0, horizontal = horizontal, scales = scales, as.table = as.table, significant = significant, groups = groups, prepanel = function(...) prepanel.default.coef(...), panel = function(x, y, subscripts, groups, ...) panel.coef(x, y, subscripts, groups, ...), ...) }) flexmix/R/plot-FLXboot.R0000644000176200001440000001421013425024235014541 0ustar liggesusersprepanel.parallel.horizontal <- function (x, y, z, horizontal = TRUE, ...) { if (horizontal) list(xlim = extend.limits(c(1, ncol(as.data.frame(z))), prop = 0.03), ylim = c(0, 1), dx = 1, dy = 1) else list(xlim = c(0, 1), ylim = extend.limits(c(1, ncol(as.data.frame(z))), prop = 0.03), dx = 1, dy = 1) } panel.parallel.horizontal <- function (x, y, z, subscripts, groups = NULL, col = superpose.line$col, lwd = superpose.line$lwd, lty = superpose.line$lty, alpha = superpose.line$alpha, common.scale = FALSE, lower = sapply(z, function(x) min(as.numeric(x), na.rm = TRUE)), upper = sapply(z, function(x) max(as.numeric(x), na.rm = TRUE)), horizontal = TRUE, ...) { superpose.line <- trellis.par.get("superpose.line") reference.line <- trellis.par.get("reference.line") n.r <- ncol(z) n.c <- length(subscripts) if (is.null(groups)) { col <- rep(col, length = n.c) lty <- rep(lty, length = n.c) lwd <- rep(lwd, length = n.c) alpha <- rep(alpha, length = n.c) } else { groups <- as.factor(groups)[subscripts] n.g <- nlevels(groups) gnum <- as.numeric(groups) col <- rep(col, length = n.g)[gnum] lty <- rep(lty, length = n.g)[gnum] lwd <- rep(lwd, length = n.g)[gnum] alpha <- rep(alpha, length = n.g)[gnum] } if (is.function(lower)) lower <- sapply(z, lower) if (is.function(upper)) upper <- sapply(z, upper) if (common.scale) { lower <- min(lower) upper <- max(upper) } lower <- rep(lower, length = n.r) upper <- rep(upper, length = n.r) dif <- upper - lower if (n.r > 1) { if (horizontal) panel.segments(y0 = 0, y1 = 1, x0 = seq_len(n.r), x1 = seq_len(n.r), col = reference.line$col, lwd = reference.line$lwd, lty = reference.line$lty) else panel.segments(x0 = 0, x1 = 1, y0 = seq_len(n.r), y1 = seq_len(n.r), col = reference.line$col, lwd = reference.line$lwd, lty = reference.line$lty) }else return(invisible()) for (i in seq_len(n.r - 1)) { x0 <- (as.numeric(z[subscripts, i]) - lower[i])/dif[i] x1 <- (as.numeric(z[subscripts, i + 1]) - lower[i + 1])/dif[i + 1] if (horizontal) panel.segments(y0 = x0, x0 = i, y1 = x1, x1 = i + 1, col = col, lty = lty, lwd = lwd, alpha = alpha, ...) else panel.segments(x0 = x0, y0 = i, x1 = x1, y1 = i + 1, col = col, lty = lty, lwd = lwd, alpha = alpha, ...) } invisible() } confidence.panel.boot <- function(x, y, z, subscripts, lwd = 1, SD = NULL, ..., lower, upper, range = c(0, 1)) { nc <- ncol(z) if (missing(lower)) lower <- sapply(z, function(x) quantile(x, range[1])) if (missing(upper)) upper <- sapply(z, function(x) quantile(x, range[2])) dif <- upper - lower if (!is.null(SD)) { SD <- lapply(SD, function(x) (x - lower)/dif) for (l in seq_along(SD)) { grid.polygon(y = unit(c(SD[[l]][,1], rev(SD[[l]][,3])), "native"), x = unit(c(seq_len(nc),rev(seq_len(nc))), "native"), gp = gpar(fill = rgb(190/225, 190/225, 190/225, 0.5), col = "darkgrey")) } } panel.parallel.horizontal(x, y, z, subscripts, ..., lower = lower, upper = upper) if (!is.null(SD)) { for (l in seq_along(SD)) { llines(y = SD[[l]][,2], x = seq_len(nc), col="white", lwd=lwd, lty = 1) } } } setMethod("plot", signature(x = "FLXboot", y = "missing"), function(x, y, ordering = NULL, range = c(0, 1), ci = FALSE, varnames = colnames(pars), strip_name = NULL, ...) { k <- x@object@k pars <- parameters(x) if (ci) { x_refit <- refit(x@object) sd <- sqrt(diag(x_refit@vcov)) CI <- x_refit@coef + qnorm(0.975) * cbind(-sd, 0, sd) indices_prior <- grep("alpha$", names(x_refit@coef)) if (length(indices_prior)) { z <- mvtnorm::rmvnorm(10000, x_refit@coef[indices_prior,drop=FALSE], x_refit@vcov[indices_prior,indices_prior,drop=FALSE]) Priors <- t(apply(cbind(1, exp(z))/rowSums(cbind(1, exp(z))), 2, quantile, c(0.025, 0.5, 0.975))) indices <- lapply(seq_len(k), function(i) grep(paste("_Comp.", i, sep = ""), names(x_refit@coef[-indices_prior]))) SD <- lapply(seq_len(k), function(i) rbind(CI[indices[[i]], ], prior = Priors[i,])) } else { indices <- lapply(seq_len(k), function(i) grep(paste("_Comp.", i, sep = ""), names(x_refit@coef))) SD <- lapply(seq_len(k), function(i) CI[indices[[i]], ]) mnrow <- max(sapply(SD, nrow)) SD <- lapply(SD, function(x) if (nrow(x) < mnrow) do.call("rbind", c(list(x), as.list(rep(0, mnrow - nrow(x))))) else x) } if (any("gaussian" %in% sapply(x@object@model, function(x) if (is(x, "FLXMRglm")) x@family else ""))) { i <- grep("sigma$", colnames(pars)) pars[,i] <- log(pars[,i]) colnames(pars)[i] <- "log(sigma)" } } else SD <- NULL range_name <- vector(mode = "character", length=2) range_name[1] <- if (range[1] == 0) "Min" else paste(round(range[1]*100), "%", sep = "") range_name[2] <- if (range[2] == 1) "Max" else paste(round(range[2]*100), "%", sep = "") Ordering <- if (is.null(ordering)) NULL else factor(as.vector(apply(matrix(pars[,ordering], nrow = k), 2, function(x) order(order(x))))) if(is.null(strip_name)) formula = ~ pars else { opt.old <- options(useFancyQuotes = FALSE) on.exit(options(opt.old)) formula <- as.formula(paste("~ pars | ", sQuote(strip_name))) } pars <- na.omit(pars) if (!is.null(attr(pars, "na.action"))) Ordering <- Ordering[-attr(na.omit(pars), "na.action")] parallel.plot <- parallelplot(formula, groups = Ordering, default.scales = list(y = list(at = c(0, 1), labels = range_name), x = list(alternating = FALSE, axs = "i", tck = 0, at = seq_len(ncol(pars)))), range = range, panel = confidence.panel.boot, prepanel = prepanel.parallel.horizontal, SD = SD, ...) parallel.plot$x.scales$labels <- varnames parallel.plot }) flexmix/R/lattice.R0000644000176200001440000000746613425024235013714 0ustar liggesusers# # Copyright (C) Deepayan Sarkar # Internal code copied from package lattice for use in flexmix # hist.constructor <- function (x, breaks, include.lowest = TRUE, right = TRUE, ...) { if (is.numeric(breaks) && length(breaks) > 1) hist(as.numeric(x), breaks = breaks, plot = FALSE, include.lowest = include.lowest, right = right) else hist(as.numeric(x), breaks = breaks, right = right, plot = FALSE) } checkArgsAndCall <- function (FUN, args) { if (!("..." %in% names(formals(FUN)))) args <- args[intersect(names(args), names(formals(FUN)))] do.call(FUN, args) } formattedTicksAndLabels <- function (x, at = FALSE, used.at = NULL, labels = FALSE, logsc = FALSE, ..., num.limit = NULL, abbreviate = NULL, minlength = 4, format.posixt = NULL, equispaced.log = TRUE) { rng <- if (length(x) == 2) as.numeric(x) else range(as.numeric(x)) if (is.logical(logsc) && logsc) logsc <- 10 have.log <- !is.logical(logsc) if (have.log) logbase <- if (is.numeric(logsc)) logsc else if (logsc == "e") exp(1) else stop("Invalid value of 'log'") logpaste <- if (have.log) paste(as.character(logsc), "^", sep = "") else "" check.overlap <- if (is.logical(at) && is.logical(labels)) TRUE else FALSE if (is.logical(at)) { at <- if (have.log && !equispaced.log) checkArgsAndCall(axisTicks, list(usr = log10(logbase^rng), log = TRUE, axp = NULL, ...)) else checkArgsAndCall(pretty, list(x = x[is.finite(x)], ...)) } else if (have.log && (length(at) > 0)) { if (is.logical(labels)) labels <- as.character(at) at <- log(at, base = logbase) } if (is.logical(labels)) { if (have.log && !equispaced.log) { labels <- as.character(at) at <- log(at, logbase) } else labels <- paste(logpaste, format(at, trim = TRUE), sep = "") } list(at = at, labels = labels, check.overlap = check.overlap, num.limit = rng) } calculateAxisComponents <- function (x, ..., packet.number, packet.list, abbreviate = NULL, minlength = 4) { if (all(is.na(x))) return(list(at = numeric(0), labels = numeric(0), check.overlap = TRUE, num.limit = c(0, 1))) ans <- formattedTicksAndLabels(x, ...) rng <- range(ans$num.limit) ok <- ans$at >= min(rng) & ans$at <= max(rng) ans$at <- ans$at[ok] ans$labels <- ans$labels[ok] if (is.logical(abbreviate) && abbreviate) ans$labels <- abbreviate(ans$labels, minlength) ans } extend.limits <- function (lim, length = 1, axs = "r", prop = if (axs == "i") 0 else lattice.getOption("axis.padding")$numeric) { if (all(is.na(lim))) NA_real_ else if (is.character(lim)) { c(1, length(lim)) + c(-1, 1) * if (axs == "i") 0.5 else lattice.getOption("axis.padding")$factor } else if (length(lim) == 2) { if (lim[1] > lim[2]) { ccall <- match.call() ccall$lim <- rev(lim) ans <- eval.parent(ccall) return(rev(ans)) } if (!missing(length) && !missing(prop)) stop("'length' and 'prop' cannot both be specified") if (length <= 0) stop("'length' must be positive") if (!missing(length)) { prop <- (as.numeric(length) - as.numeric(diff(lim)))/(2 * as.numeric(diff(lim))) } if (lim[1] == lim[2]) lim + 0.5 * c(-length, length) else { d <- diff(as.numeric(lim)) lim + prop * d * c(-1, 1) } } else { print(lim) stop("improper length of 'lim'") } } flexmix/R/multinom.R0000644000176200001440000000340113425024235014114 0ustar liggesuserssetClass("FLXMRmultinom", contains = "FLXMRglm") FLXMRmultinom <- function(formula=.~., ...) { z <- new("FLXMRmultinom", weighted=TRUE, formula=formula, family = "multinom", name=paste("FLXMRglm", "multinom", sep=":")) z@preproc.y <- function(x){ x <- as.integer(factor(x)) if (min(x) < 1 | length(unique(x)) != max(x)) stop("x needs to be coercible to an integer vector containing all numbers from 1 to max(x)") y <- matrix(0, nrow = length(x), ncol = max(x)) y[cbind(seq_along(x), x)] <- 1 y } z@defineComponent <- function(para) { predict <- function(x) { p <- tcrossprod(x, para$coef) eta <- cbind(1, exp(p)) eta/rowSums(eta) } logLik <- function(x, y) { log(predict(x))[cbind(seq_len(nrow(y)), max.col(y, "first"))] } new("FLXcomponent", parameters=list(coef=para$coef), logLik=logLik, predict=predict, df=para$df) } z@fit <- function(x, y, w, component){ r <- ncol(x) p <- ncol(y) if (p < 2) stop("Multinom requires at least two components.") mask <- c(rep(0, r + 1), rep(c(0, rep(1, r)), p - 1)) fit <- nnet.default(x, y, w, mask = mask, size = 0, skip = TRUE, softmax = TRUE, censored = FALSE, rang = 0, trace=FALSE, ...) fit$coefnames <- colnames(x) fit$weights <- w fit$vcoefnames <- fit$coefnames[seq_len(ncol(x))] fit$lab <- seq_len(ncol(y)) class(fit) <- c("multinom", "nnet") coef <- coef(fit) z@defineComponent(list(coef = coef, df = length(coef))) } z } setMethod("existGradient", signature(object = "FLXMRmultinom"), function(object) FALSE) flexmix/R/FLXMCdist1.R0000644000176200001440000002123113425024235014067 0ustar liggesusers## Note that the implementation of the weighted ML estimation is ## influenced and inspired by the function fitdistr from package MASS ## and function fitdist from package fitdistrplus. FLXMCdist1 <- function(formula=.~., dist, ...) { foo <- paste("FLXMC", dist, sep = "") if (!exists(foo)) stop("This distribution has not been implemented yet.") get(foo)(formula, ...) } prepoc.y.pos.1 <- function(x) { if (ncol(x) > 1) stop("for the inverse gaussian family y must be univariate") if (any(x < 0)) stop("values must be >= 0") x } FLXMClnorm <- function(formula=.~., ...) { z <- new("FLXMC", weighted=TRUE, formula=formula, dist = "lnorm", name="model-based log-normal clustering") z@preproc.y <- prepoc.y.pos.1 z@defineComponent <- function(para) { predict <- function(x, ...) matrix(para$meanlog, nrow = nrow(x), ncol = 1, byrow = TRUE) logLik <- function(x, y) dlnorm(y, meanlog = predict(x, ...), sdlog = para$sdlog, log = TRUE) new("FLXcomponent", parameters = list(meanlog = para$meanlog, sdlog = para$sdlog), predict = predict, logLik = logLik, df = para$df) } z@fit <- function(x, y, w, ...) { logy <- log(y); meanw <- mean(w) meanlog <- mean(w * logy) / meanw sdlog <- sqrt(mean(w * (logy - meanlog)^2) / meanw) z@defineComponent(list(meanlog = meanlog, sdlog = sdlog, df = 2)) } z } FLXMCinvGauss <- function(formula=.~., ...) { z <- new("FLXMC", weighted=TRUE, formula=formula, name = "model-based inverse Gaussian clustering", dist = "invGauss") z@preproc.y <- prepoc.y.pos.1 z@defineComponent <- function(para) { predict <- function(x, ...) matrix(para$nu, nrow = nrow(x), ncol = length(para$nu), byrow = TRUE) logLik <- function(x, y, ...) SuppDists::dinvGauss(y, nu = predict(x, ...), lambda = para$lambda, log = TRUE) new("FLXcomponent", parameters = list(nu = para$nu, lambda = para$lambda), predict = predict, logLik = logLik, df = para$df) } z@fit <- function(x, y, w, ...){ nu <- mean(w * y) / mean(w) lambda <- mean(w) / mean(w * (1 / y - 1 / nu)) z@defineComponent(list(nu = nu, lambda = lambda, df = 2)) } z } FLXMCgamma <- function(formula=.~., method = "Nelder-Mead", warn = -1, ...) { z <- new("FLXMC", weighted=TRUE, formula=formula, name = "model-based gamma clustering", dist = "gamma") z@preproc.y <- prepoc.y.pos.1 z@defineComponent <- function(para) { predict <- function(x, ...) matrix(para$shape, nrow = nrow(x), ncol = length(para$shape), byrow = TRUE) logLik <- function(x, y, ...) dgamma(y, shape = predict(x, ...), rate = para$rate, log = TRUE) new("FLXcomponent", parameters = list(shape = para$shape, rate = para$rate), predict = predict, logLik = logLik, df = para$df) } z@fit <- function(x, y, w, component){ if (!length(component)) { sw <- sum(w) mean <- sum(y * w) / sw var <- (sum(y^2 * w) / sw - mean^2) * sw / (sw - 1) start <- c(mean^2/var, mean/var) } else start <- unname(unlist(component)) control <- list(parscale = c(1, start[2])) f <- function(parms) -sum(dgamma(y, shape = parms[1], rate = parms[2], log = TRUE) * w) oop <- options(warn = warn) on.exit(oop) parms <- optim(start, f, method = method, control = control)$par z@defineComponent(list(shape = parms[1], rate = parms[2], df = 2)) } z } FLXMCexp <- function(formula=.~., ...) { z <- new("FLXMC", weighted=TRUE, formula=formula, name = "model-based exponential clustering", dist = "exp") z@preproc.y <- prepoc.y.pos.1 z@defineComponent <- function(para) { predict <- function(x, ...) matrix(para$rate, nrow = nrow(x), ncol = length(para$rate), byrow = TRUE) logLik <- function(x, y, ...) dexp(y, rate = predict(x, ...), log = TRUE) new("FLXcomponent", parameters = list(rate = para$rate), predict = predict, logLik = logLik, df = para$df) } z@fit <- function(x, y, w, component) z@defineComponent(list(rate = mean(w) / mean(w * y), df = 1)) z } FLXMCweibull <- function(formula=.~., method = "Nelder-Mead", warn = -1, ...) { z <- new("FLXMC", weighted=TRUE, formula=formula, name = "model-based Weibull clustering", dist = "weibull") z@preproc.y <- prepoc.y.pos.1 z@defineComponent <- function(para) { predict <- function(x, ...) matrix(para$shape, nrow = nrow(x), ncol = length(para$shape), byrow = TRUE) logLik <- function(x, y, ...) dweibull(y, shape = predict(x, ...), scale = para$scale, log = TRUE) new("FLXcomponent", parameters = list(shape = para$shape, scale = para$scale), predict = predict, logLik = logLik, df = para$df) } z@fit <- function(x, y, w, component){ if (!length(component)) { ly <- log(y) sw <- sum(w) mean <- sum(ly * w) / sw var <- (sum(ly^2 * w) / sw - mean^2) * sw / (sw - 1) shape <- 1.2/sqrt(var) scale <- exp(mean + 0.572/shape) start <- c(shape, scale) } else start <- unname(unlist(component)) f <- function(parms) -sum(dweibull(y, shape = parms[1], scale = parms[2], log = TRUE) * w) oop <- options(warn = warn) on.exit(oop) parms <- optim(start, f, method = method)$par z@defineComponent(list(shape = parms[1], scale = parms[2], df = 2)) } z } FLXMCburr <- function(formula=.~., start = NULL, method = "Nelder-Mead", warn = -1, ...) { z <- new("FLXMC", weighted=TRUE, formula=formula, name = "model-based Burr clustering", dist = "burr") z@preproc.y <- prepoc.y.pos.1 z@defineComponent <- function(para) { predict <- function(x, ...) matrix(para$shape1, nrow = nrow(x), ncol = length(para$shape1), byrow = TRUE) logLik <- function(x, y, ...) actuar::dburr(y, shape1 = predict(x, ...), shape2 = para$shape2, scale = para$scale, log = TRUE) new("FLXcomponent", parameters = list(shape1 = para$shape1, shape2 = para$shape2, scale = para$scale), predict = predict, logLik = logLik, df = para$df) } z@fit <- function(x, y, w, component){ if (!length(component)) { if (is.null(start)) start <- c(1, 1) } else start <- unname(unlist(component[2:3])) f <- function(parms) { shape1 <- sum(w) / sum(w * log(1 + (y/parms[2])^parms[1])) -sum(actuar::dburr(y, shape1 = shape1, shape2 = parms[1], scale = parms[2], log = TRUE) * w) } oop <- options(warn = warn) on.exit(oop) parms <- optim(start, f, method = method)$par z@defineComponent(list(shape1 = sum(w) / sum(w * log(1 + (y/parms[2])^parms[1])), shape2 = parms[1], scale = parms[2], df = 3)) } z } FLXMCinvburr <- function(formula=.~., start = NULL, warn = -1, ...) { z <- new("FLXMC", weighted=TRUE, formula=formula, name = "model-based Inverse Burr clustering", dist = "invburr") z@preproc.y <- prepoc.y.pos.1 z@defineComponent <- function(para) { predict <- function(x, ...) matrix(para$shape1, nrow = nrow(x), ncol = length(para$shape1), byrow = TRUE) logLik <- function(x, y, ...) actuar::dinvburr(y, shape1 = predict(x, ...), shape2 = para$shape2, scale = para$scale, log = TRUE) new("FLXcomponent", parameters = list(shape1 = para$shape1, shape2 = para$shape2, scale = para$scale), predict = predict, logLik = logLik, df = para$df) } z@fit <- function(x, y, w, component){ if (!length(component)) { if (is.null(start)) start <- c(1, 1) } else start <- unname(unlist(component[2:3])) f <- function(parms) { shape1 <- sum(w) / sum(w * log(1 + (parms[2]/y)^parms[1])) -sum(actuar::dinvburr(y, shape1 = shape1, shape2 = parms[1], scale = parms[2], log = TRUE) * w) } oop <- options(warn = warn) on.exit(oop) parms <- optim(start, f, method = "Nelder-Mead")$par z@defineComponent(list(shape1 = sum(w) / sum(w * log(1 + (parms[2]/y)^parms[1])), shape2 = parms[1], scale = parms[2], df = 3)) } z } flexmix/R/plot.R0000644000176200001440000000656513425024235013244 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: plot.R 5079 2016-01-31 12:21:12Z gruen $ # ###********************************************************** plotEll <- function(object, data, which=1:2, model = 1, project=NULL, points=TRUE, eqscale=TRUE, col=NULL, number = TRUE, cex=1.5, numcol="black", pch=NULL, ...) { if(is.null(col)) col <- rep(FullColors, length.out = object@k) if (!is.list(data)) { response <- data data <- list() data[[deparse(object@model[[model]]@fullformula[[2]])]] <- response } else { mf <- model.frame(object@model[[model]]@fullformula, data=data, na.action = NULL) response <- as.matrix(model.response(mf)) response <- object@model[[model]]@preproc.y(response) } clustering <- clusters(object, newdata = data) if(!is.null(project)) response <- predict(project, response) type=ifelse(points, "p", "n") if(is.null(pch)){ pch <- (clustering %% 10) pch[pch==0] <- 10 } else if(length(pch)!=nrow(response)){ pch <- rep(pch, length.out = object@k) pch <- pch[clustering] } if(eqscale) plot(response[,which], asp = 1, col=col[clustering], pch=pch, type=type, ...) else plot(response[,which], col=col[clustering], pch=pch, type=type, ...) for(k in seq_along(object@components)){ p = parameters(object, k, model, simplify=FALSE) if(!is.null(project)){ p <- projCentCov(project, p) } lines(ellipse::ellipse(p$cov[which,which], centre=p$center[which], level=0.5), col=col[k], lwd=2) lines(ellipse::ellipse(p$cov[which,which], centre=p$center[which], level=0.95), col=col[k], lty=2) } ## und nochmal fuer die zentren und nummern (damit die immer oben sind) for(k in seq_along(object@components)){ p = parameters(object, k, model, simplify=FALSE) if(!is.null(project)){ p <- projCentCov(project, p) } if(number){ rad <- ceiling(log10(object@k)) + 1.5 points(p$center[which[1]], p$center[which[2]], col=col[k], pch=21, cex=rad*cex, lwd=cex, bg="white") text(p$center[which[1]], p$center[which[2]], k, cex=cex, col=numcol) } else{ points(p$center[which[1]], p$center[which[2]], pch=16, cex=cex, col=col[k]) } } } projCentCov <- function(object, p) UseMethod("projCentCov") projCentCov.default <- function(object, p) stop(paste("Cannot handle projection objects of class", sQuote(class(object)))) projCentCov.prcomp <- function(object, p) { cent <- matrix(p$center, ncol=length(p$center)) cent <- scale(cent, object$center, object$scale) %*% object$rotation cov <- p$cov if(length(object$scale)>1) cov <- cov/outer(object$scale, object$scale, "*") cov <- t(object$rotation) %*% cov %*% object$rotation list(center=cent, cov=cov) } flexmix/R/flxmcmvpois.R0000644000176200001440000000153713425024235014627 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: flxmcmvpois.R 5079 2016-01-31 12:21:12Z gruen $ # FLXMCmvpois <- function(formula=.~.) { z <- new("FLXMC", weighted=TRUE, formula=formula, dist="mvpois", name="model-based Poisson clustering") z@preproc.y <- function(x){ storage.mode(x) <- "integer" x } z@defineComponent <- function(para) { logLik <- function(x, y){ colSums(dpois(t(y), para$lambda, log=TRUE)) } predict <- function(x, ...){ matrix(para$lambda, nrow = nrow(x), ncol=length(para$lambda), byrow=TRUE) } new("FLXcomponent", parameters=list(lambda=para$lambda), df=para$df, logLik=logLik, predict=predict) } z@fit <- function(x, y, w, ...){ z@defineComponent(list(lambda = colSums(w*y)/sum(w), df = ncol(y))) } z } flexmix/R/kldiv.R0000644000176200001440000001006513425024235013365 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: kldiv.R 5079 2016-01-31 12:21:12Z gruen $ # setMethod("KLdiv", "matrix", function(object, eps=10^-4, overlap=TRUE,...) { if(!is.numeric(object)) stop("object must be a numeric matrix\n") z <- matrix(NA, nrow=ncol(object), ncol=ncol(object)) colnames(z) <- rownames(z) <- colnames(object) w <- object < eps if (any(w)) object[w] <- eps object <- sweep(object, 2, colSums(object) , "/") for(k in seq_len(ncol(object)-1)){ for(l in 2:ncol(object)){ ok <- (object[, k] > eps) & (object[, l] > eps) if (!overlap | any(ok)) { z[k,l] <- sum(object[,k] * (log(object[,k]) - log(object[,l]))) z[l,k] <- sum(object[,l] * (log(object[,l]) - log(object[,k]))) } } } diag(z)<-0 z }) setMethod("KLdiv", "flexmix", function(object, method = c("continuous", "discrete"), ...) { method <- match.arg(method) if (method == "discrete") z <- KLdiv(object@posterior$scaled, ...) else { z <- matrix(0, object@k, object@k) for (i in seq_along(object@model)) { comp <- lapply(object@components, "[[", i) z <- z + KLdiv(object@model[[i]], comp) } } z }) setMethod("KLdiv", "FLXMRglm", function(object, components, ...) { z <- matrix(NA, length(components), length(components)) mu <- lapply(components, function(x) x@predict(object@x)) if (object@family == "gaussian") { sigma <- lapply(components, function(x) x@parameters$sigma) for (k in seq_len(ncol(z)-1)) { for (l in seq_len(ncol(z))[-1]) { z[k,l] <- sum(log(sigma[[l]]) - log(sigma[[k]]) + 1/2 * (-1 + ((sigma[[k]]^2 + (mu[[k]] - mu[[l]])^2))/sigma[[l]]^2)) z[l,k] <- sum(log(sigma[[k]]) - log(sigma[[l]]) + 1/2 * (-1 + ((sigma[[l]]^2 + (mu[[l]] - mu[[k]])^2))/sigma[[k]]^2)) } } } else if (object@family == "binomial") { for (k in seq_len(ncol(z)-1)) { for (l in seq_len(ncol(z))[-1]) { z[k,l] <- sum(mu[[k]] * log(mu[[k]]/mu[[l]]) + (1-mu[[k]]) * log((1-mu[[k]])/(1-mu[[l]]))) z[l,k] <- sum(mu[[l]] * log(mu[[l]]/mu[[k]]) + (1-mu[[l]]) * log((1-mu[[l]])/(1-mu[[k]]))) } } } else if (object@family == "poisson") { for (k in seq_len(ncol(z)-1)) { for (l in seq_len(ncol(z))[-1]) { z[k,l] <- sum(mu[[k]] * log(mu[[k]]/mu[[l]]) + mu[[l]] - mu[[k]]) z[l,k] <- sum(mu[[l]] * log(mu[[l]]/mu[[k]]) + mu[[k]] - mu[[l]]) } } } else if (object@family == "gamma") { shape <- lapply(components, function(x) x@parameters$shape) for (k in seq_len(ncol(z)-1)) { for (l in seq_len(ncol(z))[-1]) { X <- mu[[k]]*shape[[l]]/mu[[l]]/shape[[k]] z[k,l] <- sum(log(gamma(shape[[l]])/gamma(shape[[k]])) + shape[[l]] * log(X) - shape[[k]] * (1 - 1/X) + (shape[[k]] - shape[[l]])*digamma(shape[[k]])) z[l,k] <- sum(log(gamma(shape[[k]])/gamma(shape[[l]])) - shape[[k]] * log(X) - shape[[l]] * (1 - X) + (shape[[l]] - shape[[k]])*digamma(shape[[l]])) } } } else stop(paste("Unknown family", object@family)) diag(z) <- 0 z }) setMethod("KLdiv", "FLXMC", function(object, components, ...) { z <- matrix(NA, length(components), length(components)) if (object@dist == "mvnorm") { center <- lapply(components, function(x) x@parameters$center) cov <- lapply(components, function(x) x@parameters$cov) for (k in seq_len(ncol(z)-1)) { for (l in seq_len(ncol(z))[-1]) { z[k,l] <- 1/2 * (log(det(cov[[l]])) - log(det(cov[[k]])) - length(center[[k]]) + sum(diag(solve(cov[[l]]) %*% (cov[[k]] + tcrossprod(center[[k]] - center[[l]]))))) z[l,k] <- 1/2 * (log(det(cov[[k]])) - log(det(cov[[l]])) - length(center[[l]]) + sum(diag(solve(cov[[k]]) %*% (cov[[l]] + tcrossprod(center[[l]] - center[[k]]))))) } } } else stop(paste("Unknown distribution", object@dist)) diag(z) <- 0 z }) ###********************************************************** flexmix/R/allClasses.R0000644000176200001440000002256213425024235014347 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: allClasses.R 5079 2016-01-31 12:21:12Z gruen $ # setClass("FLXcontrol", representation(iter.max="numeric", minprior="numeric", tolerance="numeric", verbose="numeric", classify="character", nrep="numeric"), prototype(iter.max=200, minprior=0.05, tolerance=10e-7, verbose=0, classify="auto", nrep=1), validity=function(object) { (object@iter.max > 0) }) setAs("list", "FLXcontrol", function(from, to){ z = list2object(from, to) z@classify = match.arg(z@classify, c("auto", "weighted", "hard", "random", "SEM", "CEM")) z }) setAs("NULL", "FLXcontrol", function(from, to){ new(to) }) ###********************************************************** setClassUnion("expressionOrfunction", c("expression", "function")) setClass("FLXM", representation(fit="function", defineComponent="expressionOrfunction", weighted="logical", name="character", formula="formula", fullformula="formula", x="matrix", y="matrix", terms="ANY", xlevels="ANY", contrasts="ANY", preproc.x="function", preproc.y="function", "VIRTUAL"), prototype(formula=.~., fullformula=.~., preproc.x = function(x) x, preproc.y = function(x) x)) ## model-based clustering setClass("FLXMC", representation(dist="character"), contains = "FLXM") ## regression setClass("FLXMR", representation(offset="ANY"), contains = "FLXM") setMethod("show", "FLXM", function(object){ cat("FlexMix model of type", object@name,"\n\nformula: ") print(object@formula) cat("Weighted likelihood possible:", object@weighted,"\n\n") if(nrow(object@x)>0){ cat("Regressors:\n") print(summary(object@x)) } if(nrow(object@y)>0){ cat("Response:\n") print(summary(object@y)) } cat("\n") }) setClass("FLXcomponent", representation(df="numeric", logLik="function", parameters="list", predict="function")) setMethod("show", "FLXcomponent", function(object){ if(length(object@parameters)>0) print(object@parameters) }) ###********************************************************** setClass("FLXP", representation(name="character", formula="formula", x="matrix", fit="function", refit="function", coef="matrix", df="function"), prototype(formula=~1, df = function(x, k, ...) (k-1)*ncol(x))) setMethod("initialize", signature(.Object="FLXP"), function(.Object, ...) { .Object <- callNextMethod(.Object=.Object, ...) if (is.null(formals(.Object@refit))) .Object@refit <- .Object@fit .Object }) setClass("FLXPmultinom", contains="FLXP") setMethod("show", "FLXP", function(object){ cat("FlexMix concomitant model of type", object@name,"\n\nformula: ") print(object@formula) if(nrow(object@x)>0){ cat("\nRegressors:\n") print(summary(object@x)) } cat("\n") }) ###********************************************************** setClass("FLXdist", representation(model="list", prior="numeric", components="list", concomitant="FLXP", formula="formula", call="call", k="integer"), validity=function(object) { (object@k == length(object@prior)) }, prototype(formula=.~.)) setClass("flexmix", representation(posterior="ANY", weights="ANY", iter="numeric", cluster="integer", logLik="numeric", df="numeric", control="FLXcontrol", group="factor", size="integer", converged="logical", k0="integer"), prototype(group=(factor(integer(0))), formula=.~.), contains="FLXdist") setMethod("show", "flexmix", function(object){ cat("\nCall:", deparse(object@call,0.75*getOption("width")), sep="\n") cat("\nCluster sizes:\n") print(object@size) cat("\n") if(!object@converged) cat("no ") cat("convergence after", object@iter, "iterations\n") }) ###********************************************************** setClass("summary.flexmix", representation(call="call", AIC="numeric", BIC="numeric", logLik="logLik", comptab="ANY")) setMethod("show", "summary.flexmix", function(object){ cat("\nCall:", deparse(object@call,0.75*getOption("width")), sep="\n") cat("\n") print(object@comptab, digits=3) cat("\n") print(object@logLik) cat("AIC:", object@AIC, " BIC:", object@BIC, "\n") cat("\n") }) ###********************************************************** setClass("FLXMRglm", representation(family="character", refit="function"), contains="FLXMR") setClass("FLXR", representation(k="integer", components = "list", concomitant = "ANY", call="call", "VIRTUAL")) setClass("FLXRoptim", representation(coef="vector", vcov="matrix"), contains="FLXR") setClass("FLXRmstep", contains="FLXR") setMethod("show", signature(object = "FLXR"), function(object) { cat("\nCall:", deparse(object@call,0.75*getOption("width")), sep="\n") cat("\nNumber of components:", object@k, "\n\n") }) setMethod("summary", signature(object = "FLXRoptim"), function(object, model = 1, which = c("model", "concomitant"), ...) { which <- match.arg(which) z <- if (which == "model") object@components[[model]] else object@concomitant show(z) invisible(object) }) setMethod("summary", signature(object = "FLXRmstep"), function(object, model = 1, which = c("model", "concomitant"), ...) { which <- match.arg(which) if (which == "model") { z <- object@components[[model]] if (!is.null(z)) lapply(seq_along(z), function(k) { cat(paste("$", names(z)[k], "\n", sep = "")) printCoefmat(coef(summary(z[[k]]))) cat("\n") }) } else { z <- object@concomitant fitted.summary <- summary(z) k <- nrow(coef(fitted.summary)) + 1 coefs <- lapply(2:k, function(n) { coef.p <- fitted.summary$coefficients[n - 1, , drop = FALSE] s.err <- fitted.summary$standard.errors[n - 1, , drop = FALSE] tvalue <- coef.p/s.err pvalue <- 2 * pnorm(-abs(tvalue)) coef.table <- t(rbind(coef.p, s.err, tvalue, pvalue)) dimnames(coef.table) <- list(colnames(coef.p), c("Estimate", "Std. Error", "z value", "Pr(>|z|)")) new("Coefmat", coef.table) }) names(coefs) <- paste("Comp", 2:k, sep = ".") print(coefs) } invisible(object) }) setClass("Coefmat", contains = "matrix") setMethod("show", signature(object="Coefmat"), function(object) { printCoefmat(object, signif.stars = getOption("show.signif.stars")) }) ###********************************************************** setClass("FLXnested", representation(formula = "list", k = "numeric"), validity = function(object) { length(object@formula) == length(object@k) }) setAs("numeric", "FLXnested", function(from, to) { new("FLXnested", formula = rep(list(~0), length(from)), k = from) }) setAs("list", "FLXnested", function(from, to) { z <- list2object(from, to) }) setAs("NULL", "FLXnested", function(from, to) { new(to) }) setMethod("initialize", "FLXnested", function(.Object, formula = list(), k = numeric(0), ...) { if (is(formula, "formula")) formula <- rep(list(formula), length(k)) .Object <- callNextMethod(.Object, formula = formula, k = k, ...) .Object }) ###********************************************************** setClass("FLXMRfix", representation(design = "matrix", nestedformula = "FLXnested", fixed = "formula", segment = "matrix", variance = "vector"), contains="FLXMR") setClass("FLXMRglmfix", contains=c("FLXMRfix", "FLXMRglm")) ###********************************************************** setClassUnion("listOrdata.frame", c("list", "data.frame")) ###********************************************************** flexmix/R/relabel.R0000644000176200001440000000555413425024235013671 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id$ # setGeneric("dorelabel", function(object, perm, ...) standardGeneric("dorelabel")) setMethod("dorelabel", signature(object="flexmix", perm="vector"), function(object, perm, ...) { object <- callNextMethod(object, perm) object@posterior$scaled <- object@posterior$scaled[,perm,drop=FALSE] object@posterior$unscaled <- object@posterior$unscaled[,perm,drop=FALSE] object@cluster <- order(perm)[object@cluster] object@size <- object@size[perm] names(object@size) <- seq_along(perm) object }) setMethod("dorelabel", signature(object="FLXdist", perm="vector"), function(object, perm, ...) { if (length(perm) != object@k) stop("length of order argument does not match number of components") if (any(sort(perm) != seq_len(object@k))) stop("order argument not specified correctly") object@prior <- object@prior[perm] object@components <- object@components[perm] names(object@components) <- sapply(seq_along(object@components), function(k) gsub("[0-9]+", k, names(object@components)[k])) object@concomitant <- dorelabel(object@concomitant, perm, ...) object }) setMethod("dorelabel", signature(object="FLXP", perm="vector"), function(object, perm, ...) { object@coef <- object@coef[,perm,drop=FALSE] colnames(object@coef) <- sapply(seq_len(ncol(object@coef)), function(k) gsub("[0-9]+", k, colnames(object@coef)[k])) object }) setMethod("dorelabel", signature(object="FLXPmultinom", perm="vector"), function(object, perm, ...) { object@coef <- object@coef[,perm,drop=FALSE] object@coef <- sweep(object@coef, 1, object@coef[,1], "-") colnames(object@coef) <- sapply(seq_len(ncol(object@coef)), function(k) gsub("[0-9]+", k, colnames(object@coef)[k])) object }) setMethod("relabel", signature(object="FLXdist", by="character"), function(object, by, which=NULL, ...) { by <- match.arg(by, c("prior", "model", "concomitant")) if(by=="prior"){ perm <- order(prior(object), ...) } else if(by %in% c("model", "concomitant")) { pars <- parameters(object, which = by) index <- grep(which, rownames(pars)) if (length(index) != 1) stop("no suitable ordering variable given in 'which'") perm <- order(pars[index,], ...) } object <- dorelabel(object, perm=perm) object }) setMethod("relabel", signature(object="FLXdist", by="missing"), function(object, by, ...) { object <- relabel(object, by="prior", ...) object }) setMethod("relabel", signature(object="FLXdist", by="integer"), function(object, by, ...) { if(!all(sort(by) == seq_len(object@k))) stop("if integer, ", sQuote("by"), " must be a permutation of the numbers 1 to ", object@k) object <- dorelabel(object, by) object }) flexmix/R/boot.R0000644000176200001440000003401313425024235013216 0ustar liggesuserssetGeneric("boot", function(object, ...) standardGeneric("boot")) setGeneric("LR_test", function(object, ...) standardGeneric("LR_test")) setClass("FLXboot", representation(call="call", object="flexmix", parameters="list", concomitant="list", priors="list", logLik="matrix", k="matrix", converged="matrix", models="list", weights="list")) setMethod("show", "FLXboot", function(object) { cat("\nCall:", deparse(object@call,0.75*getOption("width")), sep="\n") }) generate_weights <- function(object) { if(is.null(object@weights) & is(object@model, "FLXMC")) { X <- do.call("cbind", lapply(object@model, function(z) z@y)) x <- apply(X, 1, paste, collapse = "") x <- as.integer(factor(x, unique(x))) object@weights <- as.vector(table(x)) indices_unique <- !duplicated(x) for (i in seq_along(object@model)) { object@model[[i]]@x <- object@model[[i]]@x[indices_unique,,drop=FALSE] object@model[[i]]@y <- object@model[[i]]@y[indices_unique,,drop=FALSE] } object@concomitant@x <- object@concomitant@x[indices_unique,,drop=FALSE] } object } setGeneric("FLXgetNewModelmatrix", function(object, ...) standardGeneric("FLXgetNewModelmatrix")) setMethod("FLXgetNewModelmatrix", "FLXM", function(object, model, indices, groups) { if (length(groups$group) > 0) { obs_groups <- lapply(groups$group[groups$groupfirst][indices], function(x) which(x == groups$group)) indices_grouped <- unlist(obs_groups) } else { indices_grouped <- indices } object@y <- model@y[indices_grouped,,drop=FALSE] object@x <- model@x[indices_grouped,,drop=FALSE] object }) setMethod("FLXgetNewModelmatrix", "FLXMRglmfix", function(object, model, indices, groups) { if (length(groups$group) > 0) { obs_groups <- lapply(groups$group[groups$groupfirst][indices], function(x) which(x == groups$group)) indices_grouped <- unlist(obs_groups) } else { indices_grouped <- indices } object@y <- do.call("rbind", rep(list(model@y[indices_grouped,,drop=FALSE]), sum(model@nestedformula@k))) object@x <- do.call("rbind", lapply(seq_len(sum(model@nestedformula@k)), function(K) model@x[model@segment[,K],,drop=FALSE][indices_grouped,,drop=FALSE])) N <- nrow(object@x)/sum(model@nestedformula@k) object@segment <- matrix(FALSE, ncol = sum(model@nestedformula@k), nrow = nrow(object@x)) for (m in seq_len(sum(model@nestedformula@k))) object@segment[(m - 1) * N + seq_len(N), m] <- TRUE object }) boot_flexmix <- function(object, R, sim = c("ordinary", "empirical", "parametric"), initialize_solution = FALSE, keep_weights = FALSE, keep_groups = TRUE, verbose = 0, control, k, model = FALSE, ...) { sim <- match.arg(sim) if (missing(R)) stop("R needs to be specified") if (!missing(control)) object@control <- do.call("new", c(list(Class = "FLXcontrol", object@control), control)) if (missing(k)) k <- object@k m <- length(object@model) has_weights <- !keep_weights & !is.null(object@weights) if (has_weights) object <- undo_weights(object) if (!keep_groups & length(object@group)) { object@concomitant@x <- object@concomitant@x[as.integer(object@group),,drop = FALSE] object@group <- factor() } groups <- list() groups$group <- object@group groups$groupfirst <- if (length(groups$group) > 0) groupFirst(groups$group) else rep(TRUE, FLXgetObs(object@model[[1]])) concomitant <- parameters <- priors <- models <- weights <- vector("list", R) logLik <- ks <- converged <- matrix(nrow=R, ncol = length(k), dimnames = list(BS = seq_len(R), k = k)) for (iter in seq_len(R)) { new <- object newgroups <- groups if(verbose && !(iter%%verbose)) cat("* ") if (iter > 1) { if (sim == "parametric") { y <- rflexmix(object, ...)$y for (i in seq_len(m)) new@model[[i]]@y <- matrix(as.vector(t(y[[i]])), nrow = nrow(new@model[[i]]@x), ncol = ncol(y[[i]]), byrow = TRUE) } else { n <- sum(groups$groupfirst) indices <- sample(seq_len(n), n, replace = TRUE) if (length(groups$group) > 0) { obs_groups <- lapply(groups$group[groups$groupfirst][indices], function(x) which(x == groups$group)) newgroups$group <- factor(rep(seq_along(obs_groups), sapply(obs_groups, length))) newgroups$groupfirst <- !duplicated(newgroups$group) } for (i in seq_len(m)) { new@model[[i]] <- FLXgetNewModelmatrix(new@model[[i]], object@model[[i]], indices, groups) } new@concomitant@x <- new@concomitant@x[indices,,drop=FALSE] } } if (has_weights & !length(groups$group) > 0) { new <- generate_weights(new) newgroups$groupfirst <- rep(TRUE, FLXgetObs(new@model[[1]])) } parameters[[iter]] <- concomitant[[iter]] <- priors[[iter]] <- list() NREP <- rep(object@control@nrep, length(k)) if (initialize_solution & object@k %in% k) NREP[k == object@k] <- 1L for (K in seq_along(k)) { fit <- new("flexmix", logLik = -Inf) for (nrep in seq_len(NREP[K])) { if (k[K] != object@k | !initialize_solution) { postunscaled <- initPosteriors(k[K], NULL, FLXgetObs(new@model[[1]]), newgroups) } else { postunscaled <- matrix(0, nrow = FLXgetObs(new@model[[1]]), ncol = k[K]) for (i in seq_len(m)) postunscaled <- postunscaled + FLXdeterminePostunscaled(new@model[[i]], lapply(new@components, function(x) x[[i]])) if(length(newgroups$group)>0) postunscaled <- groupPosteriors(postunscaled, newgroups$group) prior <- evalPrior(new@prior, new@concomitant) postunscaled <- if (is(prior, "matrix")) postunscaled + log(prior) else sweep(postunscaled, 2, log(prior), "+") postunscaled <- exp(postunscaled - log_row_sums(postunscaled)) } x <- try(FLXfit(new@model, new@concomitant, new@control, postunscaled, newgroups, weights = new@weights)) if (!is(x, "try-error")) { if(logLik(x) > logLik(fit)) fit <- x } } if (is.finite(logLik(fit))) { parameters[[iter]][paste(k[K])] <- list(parameters(fit, simplify = FALSE, drop = FALSE)) concomitant[[iter]][paste(k[K])] <- list(parameters(fit, which = "concomitant")) priors[[iter]][[paste(k[K])]] <- prior(fit) logLik[iter, paste(k[K])] <- logLik(fit) ks[iter, paste(k[K])] <- fit@k converged[iter, paste(k[K])] <- fit@converged if (model) { models[[iter]] <- fit@model weights[[iter]] <- fit@weights } } else { parameters[[iter]][[paste(k[K])]] <- concomitant[[iter]][[paste(k[K])]] <- priors[[iter]][[paste(k[K])]] <- NULL } } } if(verbose) cat("\n") new("FLXboot", call = sys.call(-1), object = object, parameters = parameters, concomitant = concomitant, priors = priors, logLik = logLik, k = ks, converged = converged, models = models, weights = weights) } setMethod("boot", signature(object="flexmix"), boot_flexmix) setMethod("LR_test", signature(object="flexmix"), function(object, R, alternative = c("greater", "less"), control, ...) { alternative <- match.arg(alternative) if (missing(control)) control <- object@control if (object@k == 1 & alternative == "less") stop(paste("alternative", alternative, "only possible for a mixture\n", "with at least two components")) k <- object@k + switch(alternative, greater = 0:1, less = 0:-1) names(k) <- k boot <- boot(object, R, sim = "parametric", k = k, initialize_solution = TRUE, control = control, ...) ok <- apply(boot@k, 1, identical, k) lrts <- 2*apply(boot@logLik[ok,order(k)], 1, diff) STATISTIC <- lrts[1] names(STATISTIC) <- "LRTS" PARAMETER <- length(lrts) names(PARAMETER) <- "BS" RETURN <- list(parameter = PARAMETER, p.value = sum(lrts[1] <= lrts)/length(lrts), alternative = alternative, null.value = object@k, method = "Bootstrap likelihood ratio test", data.name = deparse(substitute(object)), bootstrap.results = boot) class(RETURN) <- "htest" RETURN }) setMethod("parameters", "FLXboot", function(object, k, ...) { if (missing(k)) k <- object@object@k Coefs <- lapply(seq_along(object@parameters), function(i) if (is.na(object@k[i])) NULL else do.call("cbind", c(lapply(seq_len(object@k[i]), function(j) unlist(sapply(seq_along(object@object@model), function(m) FLXgetParameters(as(object@object@model[[m]], "FLXMR"), if (is(object@object@model[[m]]@defineComponent, "expression")) list(eval(object@object@model[[m]]@defineComponent, c(object@parameters[[i]][[paste(k)]][[m]][[j]], list(df = object@object@components[[j]][[m]]@df)))) else { list(object@object@model[[m]]@defineComponent( c(object@parameters[[i]][[paste(k)]][[m]][[j]], list(df = object@object@components[[j]][[m]]@df)))) })))), as.list(rep(NA, k - object@k[i]))))) Coefs <- t(do.call("cbind", Coefs)) colnames(Coefs) <- gsub("Comp.1_", "", colnames(Coefs)) Prior <- t(do.call("cbind", lapply(object@concomitant, function(x) do.call("cbind", c(list(x[[paste(k)]]), as.list(rep(NA, k - ifelse(length(x), ncol(x[[paste(k)]]), k)))))))) cbind(Coefs, Prior) }) setMethod("clusters", signature(object = "FLXboot", newdata = "listOrdata.frame"), function(object, newdata, k, ...) { if (missing(k)) k <- object@object@k lapply(seq_len(length(object@priors)), function(i) { new <- object@object new@prior <- object@priors[[i]][[paste(k)]] new@k <- length(new@prior) new@components <- rep(list(vector("list", length(object@object@model))), length(new@prior)) for (m in seq_along(new@model)) { variables <- c("x", "y", "offset", "family") variables <- variables[variables %in% slotNames(new@model[[m]])] for (var in variables) assign(var, slot(new@model[[m]], var)) for (K in seq_len(object@k[i])) { new@components[[K]][[m]] <- if (is(object@object@model[[m]]@defineComponent, "expression")) eval(object@object@model[[m]]@defineComponent, c(object@parameters[[i]][[paste(k)]][[m]][[K]], list(df = object@object@components[[K]][[m]]@df))) else object@object@model[[m]]@defineComponent( c(object@parameters[[i]][[paste(k)]][[m]][[K]], list(df = object@object@components[[K]][[m]]@df))) } } clusters(new, newdata = newdata)}) }) setMethod("posterior", signature(object = "FLXboot", newdata = "listOrdata.frame"), function(object, newdata, k, ...) { if (missing(k)) k <- object@object@k lapply(seq_len(length(object@priors)), function(i) { new <- object@object new@prior <- object@priors[[i]][[paste(k)]] new@k <- length(new@prior) new@components <- rep(list(vector("list", length(object@object@model))), length(new@prior)) for (m in seq_along(new@model)) { variables <- c("x", "y", "offset", "family") variables <- variables[variables %in% slotNames(new@model[[m]])] for (var in variables) assign(var, slot(new@model[[m]], var)) for (K in seq_len(object@k[i])) { new@components[[K]][[m]] <- if (is(object@object@model[[m]]@defineComponent, "expression")) eval(object@object@model[[m]]@defineComponent, c(object@parameters[[i]][[paste(k)]][[m]][[K]], list(df = object@object@components[[K]][[m]]@df))) else object@object@model[[m]]@defineComponent( c(object@parameters[[i]][[paste(k)]][[m]][[K]], list(df = object@object@components[[K]][[m]]@df))) } } posterior(new, newdata = newdata)}) }) setMethod("predict", signature(object = "FLXboot"), function(object, newdata, k, ...) { if (missing(k)) k <- object@object@k lapply(seq_len(length(object@priors)), function(i) { new <- object@object new@components <- vector("list", object@k[i, paste(k)]) new@components <- lapply(new@components, function(x) vector("list", length(new@model))) for (m in seq_along(new@model)) { variables <- c("x", "y", "offset", "family") variables <- variables[variables %in% slotNames(new@model[[m]])] for (var in variables) assign(var, slot(new@model[[m]], var)) for (K in seq_len(object@k[i, paste(k)])) { new@components[[K]][[m]] <- if (is(object@object@model[[m]]@defineComponent, "expression")) eval(object@object@model[[m]]@defineComponent, c(object@parameters[[i]][[paste(k)]][[m]][[K]], list(df = object@object@components[[1]][[m]]@df))) else object@object@model[[m]]@defineComponent( c(object@parameters[[i]][[paste(k)]][[m]][[K]], list(df = object@object@components[[1]][[m]]@df))) } } predict(new, newdata = newdata, ...)}) }) flexmix/R/glmnet.R0000644000176200001440000000724113425024235013544 0ustar liggesusers#' @title flexmix model driver for adaptive lasso (elastic-net) with GLMs #' @author F. Mortier (fmortier@cirad.fr) and N. Picard (nicolas.picard@cirad.fr) #' @param formula A symbolic description of the model to be fit. #' The general form is y~x|g where y is the response, x the set of predictors and g an #' optional grouping factor for repeated measurements. #' @param family a description of the error distribution and link function to be used in the model. #' "gausian", "poisson" and "binomial" are allowed. #' @param adaptive boolean indicating if algorithm should perform adaptive lasso or not #' @param select boolean vector indicating which covariates will be included in the selection process. #' Others will be included in the model. #' @details Some care is needed to ensure convergence of the #' algorithm, which is computationally more challenging than a standard EM. #' In the proposed method, not only are cluster allocations identified #' and component parameters estimated as commonly done in mixture models, #' but there is also variable selection via penalized regression using #' $k$-fold cross-validation to choose the penalty parameter. #' For the algorithm to converge, it is necessary that the same cross-validation #' partitioning be used across the EM iterations, i.e., #' the subsamples for cross-validation must be defined at the beginning #' This is accomplished using the {\tt foldid} option #' as an additional parameter to be passed to \code{\link{cv.glmnet}} (see \link{glmnet} package documentation). FLXMRglmnet <- function(formula = .~., family = c("gaussian", "binomial", "poisson"), adaptive = TRUE, select = TRUE, offset = NULL, ...) { family <- match.arg(family) z <- FLXMRglm(formula = formula, family = family) z@preproc.x <- function(x) { if (!isTRUE(all.equal(x[, 1], rep(1, nrow(x)), check.attributes = FALSE))) stop("The model needs to include an intercept in the first column.") x } z@fit <- function(x, y, w) { if (all(!select)) { coef <- if (family == "gaussian") lm.wfit(x, y, w = w)$coef else if (family == "binomial") glm.fit(x, y, family = binomial(), weights = w)$coef else if (family == "poisson") glm.fit(x, y, family=poisson(), weights = w)$coef } else { if (adaptive) { coef <- if (family == "gaussian") lm.wfit(x, y, w = w)$coef[-1] else if(family == "binomial") glm.fit(x, y, family = binomial(), weights = w)$coef[-1] else if (family == "poisson") glm.fit(x, y, family = poisson(), weights = w)$coef[-1] penalty <- mean(w) / abs(coef) } else penalty <- rep(1, ncol(x) - 1) if (any(!select)){ select <- which(!select) penalty[select] <- 0 } m <- glmnet::cv.glmnet(x[, -1, drop = FALSE], y, family = family, weights = w, penalty.factor = penalty, ...) coef <- as.vector(coef(m, s = "lambda.min")) } df <- sum(coef != 0) sigma <- if (family == "gaussian") sqrt(sum(w * (y - x %*% coef)^2/mean(w))/(nrow(x) - df)) else NULL z@defineComponent( list(coef = coef, sigma = sigma, df = df + ifelse(family == "gaussian", 1, 0))) } z } flexmix/R/flexmix.R0000644000176200001440000005221313426776521013746 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: flexmix.R 5150 2018-10-11 05:04:23Z gruen $ # log_row_sums <- function(m) { M <- m[cbind(seq_len(nrow(m)), max.col(m, "first"))] M + log(rowSums(exp(m - M))) } ## The following two methods only fill in and rearrange the model argument setMethod("flexmix", signature(formula = "formula", model="missing"), function(formula, data=list(), k=NULL, cluster=NULL, model=NULL, concomitant=NULL, control=NULL, weights=NULL) { mycall = match.call() z <- flexmix(formula=formula, data=data, k=k, cluster=cluster, model=list(FLXMRglm()), concomitant=concomitant, control=control, weights = weights) z@call <- mycall z }) setMethod("flexmix", signature(formula = "formula", model="FLXM"), function(formula, data=list(), k=NULL, cluster=NULL, model=NULL, concomitant=NULL, control=NULL, weights=NULL) { mycall = match.call() z <- flexmix(formula=formula, data=data, k=k, cluster=cluster, model=list(model), concomitant=concomitant, control=control, weights=weights) z@call <- mycall z }) ## This is the real thing setMethod("flexmix", signature(formula = "formula", model="list"), function(formula, data=list(), k=NULL, cluster=NULL, model=NULL, concomitant=NULL, control=NULL, weights=NULL) { mycall = match.call() control = as(control, "FLXcontrol") if (!is(concomitant, "FLXP")) concomitant <- FLXPconstant() groups <- .FLXgetGrouping(formula, data) model <- lapply(model, FLXcheckComponent, k, cluster) k <- unique(unlist(sapply(model, FLXgetK, k))) if (length(k) > 1) stop("number of clusters not specified correctly") model <- lapply(model, FLXgetModelmatrix, data, formula) groups$groupfirst <- if (length(groups$group)) groupFirst(groups$group) else rep(TRUE, FLXgetObs(model[[1]])) if (is(weights, "formula")) { weights <- model.frame(weights, data = data, na.action = NULL)[,1] } ## check if the weights are integer ## if non-integer weights are wanted modifications e.g. ## for classify != weighted and ## plot,flexmix,missing-method are needed if (!is.null(weights) & !identical(weights, as.integer(weights))) stop("only integer weights allowed") ## if weights and grouping is specified the weights within each ## group need to be the same if (!is.null(weights) & length(groups$group)>0) { unequal <- tapply(weights, groups$group, function(x) length(unique(x)) > 1) if (any(unequal)) stop("identical weights within groups needed") } postunscaled <- initPosteriors(k, cluster, FLXgetObs(model[[1]]), groups) if (ncol(postunscaled) == 1L) concomitant <- FLXPconstant() concomitant <- FLXgetModelmatrix(concomitant, data = data, groups = groups) z <- FLXfit(model=model, concomitant=concomitant, control=control, postunscaled=postunscaled, groups=groups, weights = weights) z@formula = formula z@call = mycall z@k0 = as.integer(k) z }) ###********************************************************** setMethod("FLXgetK", signature(model = "FLXM"), function(model, k, ...) k) setMethod("FLXgetObs", signature(model = "FLXM"), function(model) nrow(model@x)) setMethod("FLXcheckComponent", signature(model = "FLXM"), function(model, ...) model) setMethod("FLXremoveComponent", signature(model = "FLXM"), function(model, ...) model) setMethod("FLXmstep", signature(model = "FLXM"), function(model, weights, components, ...) { if ("component" %in% names(formals(model@fit))) sapply(seq_len(ncol(weights)), function(k) model@fit(model@x, model@y, weights[,k], component = components[[k]]@parameters)) else sapply(seq_len(ncol(weights)), function(k) model@fit(model@x, model@y, weights[,k])) }) setMethod("FLXdeterminePostunscaled", signature(model = "FLXM"), function(model, components, ...) { matrix(sapply(components, function(x) x@logLik(model@x, model@y)), nrow = nrow(model@y)) }) ###********************************************************** setMethod("FLXfit", signature(model="list"), function(model, concomitant, control, postunscaled=NULL, groups, weights) { ### initialize k <- ncol(postunscaled) N <- nrow(postunscaled) control <- allweighted(model, control, weights) if(control@verbose>0) cat("Classification:", control@classify, "\n") if (control@classify %in% c("SEM", "random")) iter.rm <- 0 group <- groups$group groupfirst <- groups$groupfirst if(length(group)>0) postunscaled <- groupPosteriors(postunscaled, group) logpostunscaled <- log(postunscaled) postscaled <- exp(logpostunscaled - log_row_sums(logpostunscaled)) llh <- -Inf if (control@classify %in% c("SEM", "random")) llh.max <- -Inf converged <- FALSE components <- rep(list(rep(list(new("FLXcomponent")), k)), length(model)) ### EM for(iter in seq_len(control@iter.max)) { ### M-Step postscaled = .FLXgetOK(postscaled, control, weights) prior <- if (is.null(weights)) ungroupPriors(concomitant@fit(concomitant@x, postscaled[groupfirst,,drop=FALSE]), group, groupfirst) else ungroupPriors(concomitant@fit(concomitant@x, (postscaled/weights)[groupfirst & weights > 0,,drop=FALSE], weights[groupfirst & weights > 0]), group, groupfirst) # Check min.prior nok <- if (nrow(prior) == 1) which(prior < control@minprior) else { if (is.null(weights)) which(colMeans(prior[groupfirst,,drop=FALSE]) < control@minprior) else which(colSums(prior[groupfirst,] * weights[groupfirst])/sum(weights[groupfirst]) < control@minprior) } if(length(nok)) { if(control@verbose>0) cat("*** Removing", length(nok), "component(s) ***\n") prior <- prior[,-nok,drop=FALSE] prior <- prior/rowSums(prior) postscaled <- postscaled[,-nok,drop=FALSE] postscaled[rowSums(postscaled) == 0,] <- if (nrow(prior) > 1) prior[rowSums(postscaled) == 0,] else prior[rep(1, sum(rowSums(postscaled) == 0)),] postscaled <- postscaled/rowSums(postscaled) if (!is.null(weights)) postscaled <- postscaled * weights k <- ncol(prior) if (k == 0) stop("all components removed") if (control@classify=="random") { llh.max <- -Inf iter.rm <- iter } model <- lapply(model, FLXremoveComponent, nok) components <- lapply(components, "[", -nok) } components <- lapply(seq_along(model), function(i) FLXmstep(model[[i]], postscaled, components[[i]])) postunscaled <- matrix(0, nrow = N, ncol = k) for (n in seq_along(model)) postunscaled <- postunscaled + FLXdeterminePostunscaled(model[[n]], components[[n]]) if(length(group)>0) postunscaled <- groupPosteriors(postunscaled, group) ### E-Step ## Code changed thanks to Nicolas Picard ## to avoid problems with small likelihoods postunscaled <- if (nrow(prior) > 1) postunscaled + log(prior) else sweep(postunscaled, 2, log(prior), "+") logpostunscaled <- postunscaled postunscaled <- exp(postunscaled) postscaled <- exp(logpostunscaled - log_row_sums(logpostunscaled)) ##: wenn eine beobachtung in allen Komonenten extrem ## kleine postunscaled-werte hat, ist exp(-postunscaled) ## numerisch Null, und damit postscaled NaN ## log(rowSums(postunscaled)) ist -Inf ## if (any(is.nan(postscaled))) { index <- which(as.logical(rowSums(is.nan(postscaled)))) postscaled[index,] <- if(nrow(prior)==1) rep(prior, each = length(index)) else prior[index,] postunscaled[index,] <- .Machine$double.xmin } ### check convergence llh.old <- llh llh <- if (is.null(weights)) sum(log_row_sums(logpostunscaled[groupfirst,,drop=FALSE])) else sum(log_row_sums(logpostunscaled[groupfirst,,drop=FALSE])*weights[groupfirst]) if(is.na(llh) | is.infinite(llh)) stop(paste(formatC(iter, width=4), "Log-likelihood:", llh)) if (abs(llh-llh.old)/(abs(llh)+0.1) < control@tolerance){ if(control@verbose>0){ printIter(iter, llh) cat("converged\n") } converged <- TRUE break } if (control@classify=="random") { if (llh.max < llh) { components.max <- components prior.max <- prior postscaled.max <- postscaled postunscaled.max <- postunscaled llh.max <- llh } } if(control@verbose && (iter%%control@verbose==0)) printIter(iter, llh) } ### Construct return object if (control@classify=="random") { components <- components.max prior <- prior.max postscaled <- postscaled.max postunscaled <- postunscaled.max llh <- llh.max iter <- control@iter.max - iter.rm } components <- lapply(seq_len(k), function(i) lapply(components, function(x) x[[i]])) names(components) <- paste("Comp", seq_len(k), sep=".") cluster <- max.col(postscaled) size <- if (is.null(weights)) tabulate(cluster, nbins=k) else tabulate(rep(cluster, weights), nbins=k) names(size) <- seq_len(k) concomitant <- FLXfillConcomitant(concomitant, postscaled[groupfirst,,drop=FALSE], weights[groupfirst]) df <- concomitant@df(concomitant@x, k) + sum(sapply(components, sapply, slot, "df")) control@nrep <- 1 prior <- if (is.null(weights)) colMeans(postscaled[groupfirst,,drop=FALSE]) else colSums(postscaled[groupfirst,,drop=FALSE] * weights[groupfirst])/sum(weights[groupfirst]) retval <- new("flexmix", model=model, prior=prior, posterior=list(scaled=postscaled, unscaled=postunscaled), weights = weights, iter=iter, cluster=cluster, size = size, logLik=llh, components=components, concomitant=concomitant, control=control, df=df, group=group, k=as(k, "integer"), converged=converged) retval }) ###********************************************************** .FLXgetOK = function(p, control, weights){ n = ncol(p) N = seq_len(n) if (is.null(weights)) { if (control@classify == "weighted") return(p) else { z = matrix(FALSE, nrow = nrow(p), ncol = n) if(control@classify %in% c("CEM", "hard")) m = max.col(p) else if(control@classify %in% c("SEM", "random")) m = apply(p, 1, function(x) sample(N, size = 1, prob = x)) else stop("Unknown classification method") z[cbind(seq_len(nrow(p)), m)] = TRUE } }else { if(control@classify=="weighted") z <- p * weights else{ z = matrix(FALSE, nrow=nrow(p), ncol=n) if(control@classify %in% c("CEM", "hard")) { m = max.col(p) z[cbind(seq_len(nrow(p)), m)] = TRUE z <- z * weights } else if(control@classify %in% c("SEM", "random")) z = t(sapply(seq_len(nrow(p)), function(i) table(factor(sample(N, size=weights[i], prob=p[i,], replace=TRUE), N)))) else stop("Unknown classification method") } } z } ###********************************************************** RemoveGrouping <- function(formula) { lf <- length(formula) formula1 <- formula if(length(formula[[lf]])>1) { if (deparse(formula[[lf]][[1]]) == "|"){ formula1[[lf]] <- formula[[lf]][[2]] } else if (deparse(formula[[lf]][[1]]) == "("){ form <- formula[[lf]][[2]] if (length(form) == 3 && form[[1]] == "|") formula1[[lf]] <- form[[2]] } } formula1 } .FLXgetGroupingVar <- function(x) { lf <- length(x) while (lf > 1) { x <- x[[lf]] lf <- length(x) } x } .FLXgetGrouping <- function(formula, data) { group <- factor(integer(0)) formula1 <- RemoveGrouping(formula) if (!identical(formula1, formula)) group <- factor(eval(.FLXgetGroupingVar(formula), data)) return(list(group=group, formula=formula1)) } setMethod("FLXgetModelmatrix", signature(model="FLXM"), function(model, data, formula, lhs=TRUE, ...) { formula <- RemoveGrouping(formula) if (length(grep("\\|", deparse(model@formula)))) stop("no grouping variable allowed in the model") if(is.null(model@formula)) model@formula = formula ## model@fullformula = update.formula(formula, model@formula) ## : ist das der richtige weg, wenn ein punkt in beiden ## formeln ist? model@fullformula = update(terms(formula, data=data), model@formula) ## if (lhs) { mf <- if (is.null(model@terms)) model.frame(model@fullformula, data=data, na.action = NULL) else model.frame(model@terms, data=data, na.action = NULL, xlev = model@xlevels) model@terms <- attr(mf, "terms") response <- as.matrix(model.response(mf)) model@y <- model@preproc.y(response) } else { mt1 <- if (is.null(model@terms)) terms(model@fullformula, data=data) else model@terms mf <- model.frame(delete.response(mt1), data=data, na.action = NULL, xlev = model@xlevels) model@terms<- attr(mf, "terms") ## : warum war das da??? ## attr(mt, "intercept") <- attr(mt1, "intercept") ## } X <- model.matrix(model@terms, data=mf) model@contrasts <- attr(X, "contrasts") model@x <- model@preproc.x(X) model@xlevels <- .getXlevels(model@terms, mf) model }) ## groupfirst: for grouped observation we need to be able to use ## the posterior of each group, but for computational simplicity ## post(un)scaled has N rows (with mutiple identical rows for each ## group). postscaled[groupfirst,] extracts posteriors of each ## group ordered as the appear in the data set. groupFirst <- function(x) !duplicated(x) ## if we have a group variable, set the posterior to the product ## of all density values for that group (=sum in logarithm) groupPosteriors <- function(x, group) { if (length(group) > 0) { group <- as.integer(group) x.by.group <- unname(apply(x, 2, tapply, group, sum)) x <- x.by.group[group,, drop = FALSE] } x } ungroupPriors <- function(x, group, groupfirst) { if (!length(group)) group <- seq_along(groupfirst) if (nrow(x) >= length(group[groupfirst])) { x <- x[order(as.integer(group[groupfirst])),,drop=FALSE] x <- x[as.integer(group),,drop=FALSE] } x } setMethod("allweighted", signature(model = "list", control = "ANY", weights = "ANY"), function(model, control, weights) { allweighted <- all(sapply(model, function(x) allweighted(x, control, weights))) if(allweighted){ if(control@classify=="auto") control@classify <- "weighted" } else{ if(control@classify=="auto") control@classify <- "hard" else if (control@classify=="weighted") { warning("only hard classification supported for the modeldrivers") control@classify <- "hard" } if(!is.null(weights)) stop("it is not possible to specify weights for models without weighted ML estimation") } control }) setMethod("allweighted", signature(model = "FLXM", control = "ANY", weights = "ANY"), function(model, control, weights) { model@weighted }) initPosteriors <- function(k, cluster, N, groups) { if(is(cluster, "matrix")){ postunscaled <- cluster if (!is.null(k)) if (k != ncol(postunscaled)) stop("specified k does not match the number of columns of cluster") } else{ if(is.null(cluster)){ if(is.null(k)) stop("either k or cluster must be specified") else cluster <- ungroupPriors(as.matrix(sample(seq_len(k), size = sum(groups$groupfirst), replace=TRUE)), groups$group, groups$groupfirst) } else{ cluster <- as(cluster, "integer") if (!is.null(k)) if (k != max(cluster)) stop("specified k does not match the values in cluster") k <- max(cluster) } postunscaled <- matrix(0.1, nrow=N, ncol=k) for(K in seq_len(k)){ postunscaled[cluster==K, K] <- 0.9 } } postunscaled } ###********************************************************** setMethod("predict", signature(object="FLXdist"), function(object, newdata=list(), aggregate=FALSE, ...){ if (missing(newdata)) return(fitted(object, aggregate=aggregate, drop=FALSE)) x = list() for(m in seq_along(object@model)) { comp <- lapply(object@components, "[[", m) x[[m]] <- predict(object@model[[m]], newdata, comp, ...) } if (aggregate) { prior_weights <- prior(object, newdata) z <- lapply(x, function(z) matrix(rowSums(do.call("cbind", z) * prior_weights), nrow = nrow(z[[1]]))) } else { z <- list() for (k in seq_len(object@k)) { z[[k]] <- do.call("cbind", lapply(x, "[[", k)) } names(z) <- paste("Comp", seq_len(object@k), sep=".") } z }) ###********************************************************** setMethod("parameters", signature(object="FLXdist"), function(object, component=NULL, model=NULL, which = c("model", "concomitant"), simplify=TRUE, drop=TRUE) { which <- match.arg(which) if (is.null(component)) component <- seq_len(object@k) if (is.null(model)) model <- seq_along(object@model) if (which == "model") { if (simplify) { parameters <- sapply(model, function(m) sapply(object@components[component], function(x) unlist(x[[m]]@parameters), simplify=TRUE), simplify = FALSE) } else { parameters <- sapply(model, function(m) sapply(object@components[component], function(x) x[[m]]@parameters, simplify=FALSE), simplify = FALSE) } if (drop) { if (length(component) == 1 && !simplify) parameters <- lapply(parameters, "[[", 1) if (length(model) == 1) parameters <- parameters[[1]] } } else { parameters <- object@concomitant@coef[,component,drop=FALSE] } parameters }) setMethod("prior", signature(object="FLXdist"), function(object, newdata, ...) { if (missing(newdata)) prior <- object@prior else { groups <- .FLXgetGrouping(object@formula, newdata) nobs <- if (is(newdata, "data.frame")) nrow(newdata) else min(sapply(newdata, function(x) { if (is(x, "matrix")) nrow(x) else length(x) })) group <- if (length(groups$group)) groups$group else factor(seq_len(nobs)) object@concomitant <- FLXgetModelmatrix(object@concomitant, data = newdata, groups = list(group=group, groupfirst = groupFirst(group))) prior <- determinePrior(object@prior, object@concomitant, group)[as.integer(group),] } prior }) setMethod("posterior", signature(object="flexmix", newdata="missing"), function(object, newdata, unscaled = FALSE, ...) { if (unscaled) return(object@posterior$unscaled) else return(object@posterior$scaled) }) setMethod("posterior", signature(object="FLXdist", newdata="listOrdata.frame"), function(object, newdata, unscaled=FALSE,...) { comp <- lapply(object@components, "[[", 1) postunscaled <- posterior(object@model[[1]], newdata, comp, ...) for (m in seq_along(object@model)[-1]) { comp <- lapply(object@components, "[[", m) postunscaled <- postunscaled + posterior(object@model[[m]], newdata, comp, ...) } groups <- .FLXgetGrouping(object@formula, newdata) prior <- prior(object, newdata = newdata) if(length(groups$group)>0) postunscaled <- groupPosteriors(postunscaled, groups$group) postunscaled <- postunscaled + log(prior) if (unscaled) return(exp(postunscaled)) else return(exp(postunscaled - log_row_sums(postunscaled))) }) setMethod("posterior", signature(object="FLXM", newdata="listOrdata.frame"), function(object, newdata, components, ...) { object <- FLXgetModelmatrix(object, newdata, object@fullformula, lhs = TRUE) FLXdeterminePostunscaled(object, components, ...) }) setMethod("clusters", signature(object="flexmix", newdata="missing"), function(object, newdata, ...) { object@cluster }) setMethod("clusters", signature(object="FLXdist", newdata="ANY"), function(object, newdata, ...) { max.col(posterior(object, newdata, ...)) }) ###********************************************************** setMethod("summary", "flexmix", function(object, eps=1e-4, ...){ z <- new("summary.flexmix", call = object@call, AIC = AIC(object), BIC = BIC(object), logLik = logLik(object)) TAB <- data.frame(prior=object@prior, size=object@size) rownames(TAB) <- paste("Comp.", seq_len(nrow(TAB)), sep="") TAB[["post>0"]] <- if (is.null(object@weights)) colSums(object@posterior$scaled > eps) else colSums((object@posterior$scaled > eps) * object@weights) TAB[["ratio"]] <- TAB[["size"]]/TAB[["post>0"]] z@comptab = TAB z }) ###********************************************************** flexmix/R/flexmixFix.R0000644000176200001440000001415213425024235014400 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: flexmixFix.R 5079 2016-01-31 12:21:12Z gruen $ # setMethod("FLXcheckComponent", signature(model = "FLXMRfix"), function(model, k, cluster, ...) { if (sum(model@nestedformula@k)) { if (!is.null(k)) { if (k != sum(model@nestedformula@k)) stop("specified k does not match the nestedformula in the model") } else k <- sum(model@nestedformula@k) } else { if (is(cluster, "matrix")) { if (is.null(k)) k <- ncol(cluster) } else if (!is.null(cluster)) { if (is.null(k)) { cluster <- as(cluster, "integer") k <- max(cluster) } } if (is.null(k)) stop("either k, cluster or the nestedformula of the model must be specified") else model@nestedformula <- as(k, "FLXnested") } if (length(model@variance) > 1) { if (sum(model@variance) != k) stop("specified k does not match the specified varFix argument in the model") } else if (model@variance) model@variance <- k else model@variance <- rep(1, k) model }) setMethod("FLXgetObs", signature(model = "FLXMRfix"), function(model) nrow(model@y)/sum(model@nestedformula@k)) setMethod("FLXgetK", signature(model = "FLXMRfix"), function(model, ...) sum(model@nestedformula@k)) setMethod("FLXremoveComponent", signature(model = "FLXMRfix"), function(model, nok, ...) { if (!length(nok)) return(model) K <- model@nestedformula wnok <- sapply(nok, function(i) which(apply(rbind(i > c(0, cumsum(K@k[-length(K@k)])), i <= c(cumsum(K@k))), 2, all))) wnok <- table(wnok) if (length(wnok) > 0) { K@k[as.integer(names(wnok))] <- K@k[as.integer(names(wnok))] - wnok if (any(K@k == 0)) { keep <- K@k != 0 K@k <- K@k[keep] K@formula <- K@formula[keep] } k <- sum(K@k) model@nestedformula <- K } varnok <- sapply(nok, function(i) which(apply(rbind(i > c(0, cumsum(model@variance[-length(model@variance)])), i <= c(cumsum(model@variance))), 2, all))) varnok <- table(varnok) if (length(varnok) > 0) { model@variance[as.integer(names(varnok))] <- model@variance[as.integer(names(varnok))] - varnok if (any(model@variance == 0)) model@variance <- model@variance[model@variance != 0] } rok <- which(!apply(model@segment[,nok,drop=FALSE], 1, function(x) any(x))) model@x <- model@x[rok, which(colSums(model@design[-nok,,drop=FALSE]) > 0), drop=FALSE] model@y <- model@y[rok,, drop=FALSE] model@design <- model@design[-nok,,drop=FALSE] cok <- colSums(model@design) > 0 model@design <- model@design[,cok,drop=FALSE] model@segment <- model@segment[rok,-nok, drop=FALSE] model }) ###********************************************************** setMethod("FLXmstep", signature(model = "FLXMRfix"), function(model, weights, ...) { model@fit(model@x, model@y, as.vector(weights), model@design, model@variance) }) ###********************************************************** setMethod("FLXdeterminePostunscaled", signature(model = "FLXMRfix"), function(model, components, ...) { sapply(seq_along(components), function(m) components[[m]]@logLik(model@x[model@segment[,m], as.logical(model@design[m,]), drop=FALSE], model@y[model@segment[,m],,drop=FALSE])) }) ###********************************************************** modelMatrix <- function(random, fixed, nested, data=list(), lhs, xlevels = NULL) { if (!lhs) random <- random[-2] mf.random <- model.frame(random, data=data, na.action = NULL) response <- if (lhs) as.matrix(model.response(mf.random)) else NULL xlev <- xlevels[names(.getXlevels(terms(mf.random), mf.random))] mm.random <- if (is.null(xlev)) model.matrix(terms(mf.random), data=mf.random) else model.matrix(terms(mf.random), data=data, xlev = xlev) xlevels.random <- .getXlevels(terms(mf.random), mf.random) randomfixed <- if(identical(paste(deparse(fixed), collapse = ""), "~0")) random else update(random, paste("~.+", paste(deparse(fixed[[length(fixed)]]), collapse = ""))) mf.randomfixed <- model.frame(randomfixed, data=data) mm.randomfixed <- model.matrix(terms(mf.randomfixed), data=mf.randomfixed, xlev = xlevels[names(.getXlevels(terms(mf.randomfixed), mf))]) mm.fixed <- mm.randomfixed[,!colnames(mm.randomfixed) %in% colnames(mm.random), drop=FALSE] xlevels.fixed <- .getXlevels(terms(mf.randomfixed), mf.randomfixed) all <- mm.all <- mm.nested <- xlevels.nested <- list() for (l in seq_along(nested)) { all[[l]] <- if (identical(paste(deparse(nested[[l]]), collapse = ""), "~0")) randomfixed else update(randomfixed, paste("~.+", paste(deparse(nested[[l]][[length(nested[[l]])]]), collapse = ""))) mf <- model.frame(all[[l]], data=data) mm.all[[l]] <- model.matrix(terms(mf), data=mf, xlev = xlevels[names(.getXlevels(terms(mf), mf))]) mm.nested[[l]] <- mm.all[[l]][,!colnames(mm.all[[l]]) %in% colnames(mm.randomfixed),drop=FALSE] xlevels.nested[[l]] <- .getXlevels(terms(mf), mf) } return(list(random=mm.random, fixed=mm.fixed, nested=mm.nested, response=response, xlevels=c(xlevels.random, xlevels.fixed, unlist(xlevels.nested)))) } ###********************************************************** modelDesign <- function(mm.all, k) { design <- matrix(1, nrow=sum(k@k), ncol=ncol(mm.all$fixed)) col.names <- colnames(mm.all$fixed) nested <- matrix(0, nrow=sum(k@k), ncol=sum(sapply(mm.all$nested, ncol))) cumK <- c(0, cumsum(k@k)) i <- 0 for (l in seq_along(mm.all$nested)) { if (ncol(mm.all$nested[[l]])) { nested[(cumK[l] + 1):cumK[l+1], i+seq_len(ncol(mm.all$nested[[l]]))] <- 1 i <- i+ncol(mm.all$nested[[l]]) col.names <- c(col.names, colnames(mm.all$nested[[l]])) } } design <- cbind(design, nested) if (ncol(mm.all$random)) design <- cbind(design, kronecker(diag(sum(k@k)), matrix(1, ncol=ncol(mm.all$random)))) colnames(design) <- c(col.names, rep(colnames(mm.all$random), sum(k@k))) design } ###********************************************************** flexmix/R/z.R0000644000176200001440000000063113425024235012523 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: z.R 5079 2016-01-31 12:21:12Z gruen $ # ###********************************************************** ## Backward compatibility ## component model driver FLXglm <- FLXMRglm FLXglmFix <- FLXMRglmfix FLXmclust <- FLXMCmvnorm FLXbclust <- FLXMCmvbinary ## concomitant model driver FLXmultinom <- FLXPmultinom FLXconstant <- FLXPconstant flexmix/R/plot-flexmix.R0000644000176200001440000001353013425024235014704 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: plot-flexmix.R 5079 2016-01-31 12:21:12Z gruen $ # determine_y <- function(h, root) { y <- h$counts if (root) y <- sqrt(y) return(y) } panel.rootogram <- function (x, breaks, equal.widths = TRUE, nint = max(round(log2(length(x)) + 1), 3), alpha = plot.polygon$alpha, col = plot.polygon$col, border = plot.polygon$border, lty = plot.polygon$lty, lwd = plot.polygon$lwd, subscripts, groups, mark, root = TRUE, markcol, ...) { x <- as.numeric(x) plot.polygon <- trellis.par.get("plot.polygon") grid.lines(x = c(0.05, 0.95), y = unit(c(0, 0), "native"), gp = gpar(col = border, lty = lty, lwd = lwd, alpha = alpha), default.units = "npc") if (length(x) > 0) { if (is.null(breaks)) { breaks <- if (equal.widths) do.breaks(range(x, finite = TRUE), nint) else quantile(x, 0:nint/nint, na.rm = TRUE) } h <- hist.constructor(x, breaks = breaks, plot = FALSE, ...) y <- determine_y(h, root) if (!is.null(mark)) { h1 <- hist.constructor(x[groups[subscripts] == mark], breaks = h$breaks, plot = FALSE, ...) y1 <- determine_y(h1, root) } nb <- length(breaks) if (length(y) != nb - 1) warning("problem with hist computations") if (nb > 1) { panel.rect(x = breaks[-nb], y = 0, height = y, width = diff(breaks), col = col, alpha = alpha, border = border, lty = lty, lwd = lwd, just = c("left", "bottom")) if (!is.null(mark)) panel.rect(x = breaks[-nb], y = 0, height = y1, width = diff(breaks), col = markcol, alpha = alpha, border = border, lty = lty, lwd = lwd, just = c("left", "bottom")) } } } prepanel.rootogram <- function (x, breaks, equal.widths = TRUE, nint = max(round(log2(length(x)) + 1), 3), root = TRUE, ...) { if (length(x) < 1) list(xlim = NA, ylim = NA, dx = NA, dy = NA) else { if (is.factor(x)) { isFactor <- TRUE xlimits <- levels(x) } else isFactor <- FALSE if (!is.numeric(x)) x <- as.numeric(x) if (is.null(breaks)) { breaks <- if (equal.widths) do.breaks(range(x, finite = TRUE), nint) else quantile(x, 0:nint/nint, na.rm = TRUE) } h <- hist.constructor(x, breaks = breaks, plot = FALSE, ...) y <- determine_y(h, root) list(xlim = if (isFactor) xlimits else range(x, breaks, finite = TRUE), ylim = range(0, y, finite = TRUE), dx = 1, dy = 1) } } setMethod("plot", signature(x="flexmix", y="missing"), function(x, y, mark=NULL, markcol=NULL, col=NULL, eps=1e-4, root=TRUE, ylim=TRUE, main=NULL, xlab = "", ylab = "", as.table = TRUE, endpoints = c(-0.04, 1.04), ...){ k <- length(x@prior) if(is.null(markcol)) markcol <- FullColors[5] if(is.null(col)) col <- LightColors[4] if(is.null(main)){ main <- ifelse(root, "Rootogram of posterior probabilities", "Histogram of posterior probabilities") main <- paste(main, ">", eps) } groupfirst <- if (length(x@group)) !duplicated(x@group) else TRUE if (is.null(x@weights)) z <- data.frame(posterior = as.vector(x@posterior$scaled[groupfirst,,drop=FALSE]), component = factor(rep(seq_len(x@k), each = nrow(x@posterior$scaled[groupfirst,,drop=FALSE])), levels = seq_len(x@k), labels = paste("Comp.", seq_len(x@k))), cluster = rep(as.vector(x@cluster[groupfirst]), k)) else z <- data.frame(posterior = rep(as.vector(x@posterior$scaled[groupfirst,,drop=FALSE]), rep(x@weights[groupfirst], k)), component = factor(rep(seq_len(x@k), each = sum(x@weights[groupfirst])), seq_len(x@k), paste("Comp.", seq_len(x@k))), cluster = rep(rep(as.vector(x@cluster[groupfirst]), x@weights[groupfirst]), k)) panel <- function(x, subscripts, groups, ...) panel.rootogram(x, root = root, mark = mark, col = col, markcol = markcol, subscripts = subscripts, groups = groups, ...) prepanel <- function(x, ...) prepanel.rootogram(x, root = root, ...) z <- subset(z, posterior > eps) cluster <- NULL # make codetools happy if (is.logical(ylim)) { scales <- if (ylim) list() else list(y = list(relation = "free")) hh <- histogram(~ posterior | component, data = z, main = main, ylab = ylab, xlab = xlab, groups = cluster, panel = panel, prepanel = prepanel, scales = scales, as.table = as.table, endpoints = endpoints, ...) } else hh <- histogram(~ posterior | component, data = z, main = main, ylab = ylab, xlab = xlab, groups = cluster, ylim = ylim, panel = panel, prepanel = prepanel, as.table = as.table, endpoints = endpoints, ...) if (root) { hh$yscale.components <- function (lim, packet.number = 0, packet.list = NULL, right = TRUE, ...) { comps <- calculateAxisComponents(lim, packet.list = packet.list, packet.number = packet.number, ...) comps$at <- sqrt(seq(min(comps$at)^2, max(comps$at)^2, length.out = length(comps$at))) comps$labels <- format(comps$at^2, trim = TRUE) list(num.limit = comps$num.limit, left = list(ticks = list(at = comps$at, tck = 1), labels = list(at = comps$at, labels = comps$labels, cex = 1, check.overlap = comps$check.overlap)), right = right) } } hh }) flexmix/R/glmFix.R0000644000176200001440000001622613430477461013520 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: glmFix.R 5156 2019-02-12 08:11:16Z gruen $ # FLXMRglmfix <- function(formula=.~., fixed=~0, varFix = FALSE, nested = NULL, family=c("gaussian", "binomial", "poisson", "Gamma"), offset=NULL) { family <- match.arg(family) nested <- as(nested, "FLXnested") if (length(fixed) == 3) stop("no left hand side allowed for fixed") z <- new("FLXMRglmfix", FLXMRglm(formula, family, offset), fixed=fixed, name=paste("FLXMRglmfix", family, sep=":"), nestedformula=nested, variance = varFix) if(family=="gaussian"){ z@fit <- function(x, y, w, incidence, variance, ...){ fit <- lm.wfit(x, y, w=w, offset=offset) k <- nrow(incidence) n <- nrow(x)/k sigma <- vector(length=k) cumVar <- cumsum(c(0, variance)) for (i in seq_along(variance)) { ind <- cumVar[i]*n + seq_len(n*variance[i]) sigma[cumVar[i] + seq_len(variance[i])] <- sqrt(sum(fit$weights[ind] * fit$residuals[ind]^2 / mean(fit$weights[ind]))/ (length(ind) - sum(incidence[i,]))) } fit <- fit[c("coefficients")] coefs <- coef(fit) names(coefs) <- colnames(incidence) df <- rowSums(incidence/rep(colSums(incidence), each = nrow(incidence))) + rep(1/variance, variance) lapply(seq_len(k), function(K) z@defineComponent( list(coef = coefs[as.logical(incidence[K, ])], sigma = sigma[K], df = df[K]))) } } else if(family=="binomial"){ z@fit <- function(x, y, w, incidence, ...){ fit <- glm.fit(x, y, weights=w, family=binomial(), offset=offset) fit <- fit[c("coefficients","family")] k <- nrow(incidence) coefs <- coef(fit) names(coefs) <- colnames(incidence) df <- rowSums(incidence/rep(colSums(incidence), each = nrow(incidence))) lapply(seq_len(k), function(K) z@defineComponent( list(coef = coefs[as.logical(incidence[K, ])], df = df[K]))) } } else if(family=="poisson"){ z@fit <- function(x, y, w, incidence, ...){ fit <- glm.fit(x, y, weights=w, family=poisson(), offset=offset) fit <- fit[c("coefficients","family")] k <- nrow(incidence) coefs <- coef(fit) names(coefs) <- colnames(incidence) df <- rowSums(incidence/rep(colSums(incidence), each = nrow(incidence))) lapply(seq_len(k), function(K) z@defineComponent( list(coef = coefs[as.logical(incidence[K, ])], df = df[K]))) } } else if(family=="Gamma"){ z@fit <- function(x, y, w, incidence, ...){ fit <- glm.fit(x, y, weights=w, family=Gamma(), offset=offset) shape <- sum(fit$prior.weights)/fit$deviance fit <- fit[c("coefficients","family")] k <- nrow(incidence) coefs <- coef(fit) names(coefs) <- colnames(incidence) df <- rowSums(incidence/rep(colSums(incidence), each = nrow(incidence))) lapply(seq_len(k), function(K) z@defineComponent( list(coef = coefs[as.logical(incidence[K, ])], df = df[K], shape = shape))) } } else stop(paste("Unknown family", family)) z } ###********************************************************** setMethod("refit_mstep", signature(object="FLXMRglmfix", newdata="missing"), function(object, newdata, weights, ...) { warning("Separate regression models are fitted using posterior weights.") lapply(seq_len(ncol(weights)), function(k) { x <- object@x[object@segment[, k], as.logical(object@design[k,]), drop = FALSE] colnames(x) <- colnames(object@design)[as.logical(object@design[k,])] y <- object@y[object@segment[, k],, drop = FALSE] fit <- object@refit(x, y, weights[,k], ...) fit <- c(fit, list(formula = object@fullformula, terms = object@terms, contrasts = object@contrasts, xlevels = object@xlevels)) class(fit) <- c("glm", "lm") fit }) }) ###********************************************************** setMethod("fitted", signature(object="FLXMRglmfix"), function(object, components, ...) { N <- nrow(object@x)/length(components) z <- list() for(n in seq_along(components)){ x <- object@x[(n-1)*N + seq_len(N), as.logical(object@design[n,]), drop=FALSE] z[[n]] <- list(components[[n]]@predict(x)) } z }) ###********************************************************** setMethod("predict", signature(object="FLXMRglmfix"), function(object, newdata, components, ...) { model <- FLXgetModelmatrix(object, newdata, object@fullformula, lhs=FALSE) k <- sum(object@nestedformula@k) N <- nrow(model@x)/k z <- list() for (m in seq_len(k)) { z[[m]] <- components[[m]]@predict(model@x[model@segment[,m], as.logical(model@design[m,]), drop=FALSE], ...) } z }) ###********************************************************** setMethod("FLXgetModelmatrix", signature(model="FLXMRfix"), function(model, data, formula, lhs=TRUE, ...) { formula <- RemoveGrouping(formula) if (length(grep("\\|", deparse(model@formula)))) stop("no grouping variable allowed in the model") if(is.null(model@formula)) model@formula <- formula model@fullformula <- update.formula(formula, model@formula) k <- model@nestedformula mm.all <- modelMatrix(model@fullformula, model@fixed, k@formula, data, lhs, model@xlevels) model@design <- modelDesign(mm.all, k) desNested <- if (sum(sapply(mm.all$nested, ncol))) { rbind(ncol(mm.all$fixed) + seq_len(sum(sapply(mm.all$nested, ncol))), unlist(lapply(seq_along(mm.all$nested), function(i) rep(i, ncol(mm.all$nested[[i]]))))) }else matrix(ncol=0, nrow=2) model@x <- cbind(kronecker(rep(1, sum(k@k)), mm.all$fixed), do.call("cbind", lapply(unique(desNested[2,]), function(i) { kronecker(model@design[,desNested[1, desNested[2, ] == i][1]], mm.all$nested[[i]])})), kronecker(diag(sum(k@k)), mm.all$random)) N <- nrow(model@x)/sum(k@k) model@segment <- matrix(FALSE, ncol = sum(k@k), nrow = nrow(model@x)) for (m in seq_len(sum(k@k))) model@segment[(m - 1) * N + seq_len(N), m] <- TRUE if (lhs) { y <- mm.all$response rownames(y) <- NULL response <- as.matrix(apply(y, 2, rep, sum(k@k))) model@y <- model@preproc.y(response) } model@x <- model@preproc.x(model@x) model@xlevels <- mm.all$xlevels model }) flexmix/R/initFlexmix.R0000644000176200001440000000355613425024235014563 0ustar liggesuserssetClass("initMethod", representation(step1 = "FLXcontrol", step2 = "FLXcontrol")) initMethod <- function(name = c("tol.em", "cem.em", "sem.em"), step1 = list(tolerance = 10^-2), step2 = list(), control = list(), nrep = 3L) { name <- match.arg(name) z <- new("initMethod", step1 = as(c(step1, control), "FLXcontrol"), step2 = as(c(step2, control), "FLXcontrol")) z@step1@nrep <- as.integer(nrep) z@step2@nrep <- 1L z@step1@classify <- switch(name, cem.em = "CEM", sem.em = "SEM", tol.em = "weighted") z } initFlexmix <- function(..., k, init = list(), control = list(), nrep = 3L, verbose = TRUE, drop = TRUE, unique = FALSE) { MYCALL <- match.call() if (missing(k)) stop("'k' is missing.") if (!missing(control) & is(init, "initMethod")) warning("'control' argument ignored.") init <- do.call("initMethod", c(init, list(control = control, nrep = nrep))) MYCALL1 <- lapply(k, function(K) { MYCALL[["k"]] <- as.numeric(K) MYCALL }) names(MYCALL1) <- paste(k) STEP1 <- stepFlexmix(..., k = k, verbose = verbose, drop = FALSE, unique = FALSE, nrep = init@step1@nrep, control = init@step1) models <- lapply(k, function(K) { if (length(k) > 1 && verbose) cat("* ") new("flexmix", flexmix(..., control = init@step2, cluster = posterior(getModel(STEP1, paste(K)))), k0 = as.integer(K), call = MYCALL1[[paste(K)]]) }) if (length(k) > 1 && verbose) cat("\n") names(models) <- paste(k) if (drop & length(models) == 1) { return(models[[1]]) } else { z <- new("stepFlexmix", models = models, k = as.integer(k), logLiks = STEP1@logLiks, nrep = STEP1@nrep, call = MYCALL) if (unique) z <- unique(z) return(z) } } flexmix/R/concomitant.R0000644000176200001440000000647413425024235014603 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: concomitant.R 5079 2016-01-31 12:21:12Z gruen $ # FLXPmultinom <- function(formula=~1) { z <- new("FLXPmultinom", name="FLXPmultinom", formula=formula) multinom.fit <- function(x, y, w, ...) { r <- ncol(x) p <- ncol(y) if (p < 2) stop("Multinom requires at least two components.") mask <- c(rep(0, r + 1), rep(c(0, rep(1, r)), p - 1)) nnet.default(x, y, w, mask = mask, size = 0, skip = TRUE, softmax = TRUE, censored = FALSE, rang = 0, trace=FALSE,...) } z@fit <- function(x, y, w, ...) multinom.fit(x, y, w, ...)$fitted.values z@refit <- function(x, y, w, ...) { if (missing(w) || is.null(w)) w <- rep(1, nrow(y)) rownames(y) <- rownames(x) <- NULL fit <- multinom(y ~ 0 + x, weights = w, data = list(y = y, x = x), Hess = TRUE, trace = FALSE) fit$coefnames <- colnames(x) fit$vcoefnames <- fit$coefnames[seq_along(fit$coefnames)] dimnames(fit$Hessian) <- lapply(dim(fit$Hessian) / ncol(x), function(i) paste(rep(seq_len(i) + 1, each = ncol(x)), colnames(x), sep = ":")) fit } z } FLXPconstant <- function() { new("FLXP", name="FLXPconstant", formula = ~1, fit = function(x, y, w, ...){ if (missing(w) || is.null(w)) return(matrix(colMeans(y), ncol=ncol(y), dimnames = list("prior", seq_len(ncol(y))))) else return(matrix(colMeans(w*y)/mean(w), ncol=ncol(y), dimnames = list("prior", seq_len(ncol(y))))) }) } ###********************************************************** setMethod("FLXgetModelmatrix", signature(model="FLXP"), function(model, data, groups, lhs, ...) { mt <- terms(model@formula, data=data) mf <- model.frame(delete.response(mt), data=data, na.action = NULL) X <- model.matrix(mt, data=mf) if (nrow(X)){ if (!checkGroup(X, groups$group)) stop("model variables have to be constant for grouping variable") model@x <- X[groups$groupfirst,,drop=FALSE] } else{ model@x <- matrix(1, nrow=sum(groups$groupfirst)) } model }) checkGroup <- function(x, group) { check <- TRUE for(g in levels(group)){ gok <- group==g if(any(gok)){ check <- all(c(check, apply(x[gok,,drop=FALSE], 2, function(z) length(unique(z)) == 1))) } } check } ###********************************************************** setMethod("refit_mstep", signature(object="FLXP", newdata="missing"), function(object, newdata, posterior, group, ...) NULL) setMethod("refit_mstep", signature(object="FLXPmultinom", newdata="missing"), function(object, newdata, posterior, group, ...) { groupfirst <- if (length(group)) groupFirst(group) else rep(TRUE, nrow(posterior)) object@refit(object@x, posterior[groupfirst,,drop=FALSE], ...) }) ###********************************************************** setMethod("FLXfillConcomitant", signature(concomitant="FLXP"), function(concomitant, posterior, weights) { concomitant@coef <- concomitant@refit(concomitant@x, posterior, weights) concomitant }) setMethod("FLXfillConcomitant", signature(concomitant="FLXPmultinom"), function(concomitant, posterior, weights) { concomitant@coef <- cbind("1" = 0, t(coef(concomitant@refit(concomitant@x, posterior, weights)))) concomitant }) ###********************************************************** flexmix/R/condlogit.R0000644000176200001440000000553213425024235014241 0ustar liggesuserssetClass("FLXMRcondlogit", representation(strata="ANY", strata_formula="ANY"), contains = "FLXMRglm") FLXMRcondlogit <- function(formula=.~., strata) { z <- new("FLXMRcondlogit", weighted=TRUE, formula=formula, strata_formula=strata, family="multinomial", name=paste("FLXMRcondlogit")) z@defineComponent <- function(para) { predict <- function(x, ...) tcrossprod(x, t(para$coef)) logLik <- function(x, y, strata) { llh_all <- vector("numeric", length = length(y)) eta <- predict(x) llh_all[as.logical(y)] <- eta[as.logical(y)] ((tapply(llh_all, strata, sum) - tapply(exp(eta), strata, function(z) log(sum(z))))/tabulate(strata))[strata] } new("FLXcomponent", parameters=list(coef=para$coef), logLik=logLik, predict=predict, df=para$df) } z@fit <- function(x, y, w, strata){ index <- w > 0 fit <- survival::coxph.fit(x[index,,drop=FALSE], survival::Surv(1-y, y)[index], strata[index], weights=w[index], control = survival::coxph.control(), method = "exact", rownames = seq_len(nrow(y))[index]) coef <- coef(fit) df <- length(coef) z@defineComponent(list(coef = coef, df = df)) } z } setMethod("FLXgetModelmatrix", signature(model="FLXMRcondlogit"), function(model, data, formula, lhs=TRUE, ...) { formula <- RemoveGrouping(formula) if(is.null(model@formula)) model@formula = formula model@fullformula = update(terms(formula, data=data), model@formula) ## Ensure that an intercept is included model@fullformula <- update(model@fullformula, ~ . + 1) if (lhs) { mf <- model.frame(model@fullformula, data=data, na.action = NULL) model@x <- model.matrix(attr(mf, "terms"), data=mf) response <- as.matrix(model.response(mf)) model@y <- model@preproc.y(response) } else { mt1 <- terms(model@fullformula, data=data) mf <- model.frame(delete.response(mt1), data=data, na.action = NULL) mt <- attr(mf, "terms") model@x <- model.matrix(mt, data=mf) } strata <- update(model@strata_formula, ~ . + 0) mf <- model.frame(strata, data=data, na.action=NULL) model@strata <- as.integer(model.matrix(attr(mf, "terms"), data=mf)) ## Omit the intercept for identifiability model@x <- model@x[,attr(model@x, "assign") != 0, drop=FALSE] model@x <- model@preproc.x(model@x) model }) setMethod("FLXmstep", signature(model = "FLXMRcondlogit"), function(model, weights, ...) { apply(weights, 2, function(w) model@fit(model@x, model@y, w, model@strata)) }) setMethod("FLXdeterminePostunscaled", signature(model = "FLXMRcondlogit"), function(model, components, ...) { sapply(components, function(x) x@logLik(model@x, model@y, model@strata)) }) setMethod("existGradient", signature(object = "FLXMRcondlogit"), function(object) FALSE) flexmix/R/allGenerics.R0000644000176200001440000000621013425024235014501 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: allGenerics.R 5079 2016-01-31 12:21:12Z gruen $ # setGeneric("flexmix", function(formula, data=list(), k=NULL, cluster=NULL, model=NULL, concomitant=NULL, control=NULL, weights = NULL) standardGeneric("flexmix")) setGeneric("FLXfit", function(model, concomitant, control, postunscaled=NULL, groups, weights) standardGeneric("FLXfit")) ###********************************************************** setGeneric("FLXgetModelmatrix", function(model, data, ...) standardGeneric("FLXgetModelmatrix")) setGeneric("FLXfillConcomitant", function(concomitant, ...) standardGeneric("FLXfillConcomitant")) ###********************************************************** setGeneric("logLik") setGeneric("clogLik", function(object, ...) standardGeneric("clogLik")) setGeneric("EIC", function(object, ...) standardGeneric("EIC")) ###********************************************************** setGeneric("FLXcheckComponent", function(model, ...) standardGeneric("FLXcheckComponent")) setGeneric("FLXgetK", function(model, ...) standardGeneric("FLXgetK")) setGeneric("FLXgetObs", function(model) standardGeneric("FLXgetObs")) setGeneric("FLXmstep", function(model, ...) standardGeneric("FLXmstep")) setGeneric("FLXremoveComponent", function(model, ...) standardGeneric("FLXremoveComponent")) setGeneric("FLXdeterminePostunscaled", function(model, ...) standardGeneric("FLXdeterminePostunscaled")) setGeneric("FLXgetDesign", function(object, ...) standardGeneric("FLXgetDesign")) setGeneric("FLXreplaceParameters", function(object, ...) standardGeneric("FLXreplaceParameters")) setGeneric("FLXlogLikfun", function(object, ...) standardGeneric("FLXlogLikfun")) setGeneric("FLXgradlogLikfun", function(object, ...) standardGeneric("FLXgradlogLikfun")) setGeneric("VarianceCovariance", function(object, ...) standardGeneric("VarianceCovariance")) setGeneric("FLXgetParameters", function(object, ...) standardGeneric("FLXgetParameters")) setGeneric("logLikfun_comp", function(object, ...) standardGeneric("logLikfun_comp")) setGeneric("getPriors", function(object, ...) standardGeneric("getPriors")) setGeneric("existGradient", function(object, ...) standardGeneric("existGradient")) setGeneric("refit_mstep", function(object, newdata, ...) standardGeneric("refit_mstep")) setGeneric("refit_optim", function(object, ...) standardGeneric("refit_optim")) ###********************************************************** setGeneric("group", function(object, ...) standardGeneric("group")) setGeneric("rflexmix", function(object, newdata, ...) standardGeneric("rflexmix")) setGeneric("rFLXM", function(model, components, ...) standardGeneric("rFLXM")) ## just to make sure that some S3 generics are available in S4 setGeneric("fitted", package = "stats") setGeneric("predict", package = "stats") setGeneric("simulate", package = "stats") setGeneric("summary", package = "base") setGeneric("unique", package = "base") setGeneric("allweighted", function(model, control, weights) standardGeneric("allweighted")) flexmix/R/ziglm.R0000644000176200001440000000405313425024235013376 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: ziglm.R 5079 2016-01-31 12:21:12Z gruen $ # setClass("FLXMRziglm", contains = "FLXMRglm") FLXMRziglm <- function(formula = . ~ ., family = c("binomial", "poisson"), ...) { family <- match.arg(family) new("FLXMRziglm", FLXMRglm(formula, family, ...), name = paste("FLXMRziglm", family, sep=":")) } setMethod("FLXgetModelmatrix", signature(model="FLXMRziglm"), function(model, data, formula, lhs=TRUE, ...) { model <- callNextMethod(model, data, formula, lhs) if (attr(terms(model@fullformula), "intercept") == 0) stop("please include an intercept") model }) setMethod("FLXremoveComponent", signature(model = "FLXMRziglm"), function(model, nok, ...) if (1 %in% nok) as(model, "FLXMRglm") else model) setMethod("FLXmstep", signature(model = "FLXMRziglm"), function(model, weights, components, ...) { coef <- c(-Inf, rep(0, ncol(model@x)-1)) names(coef) <- colnames(model@x) comp.1 <- model@defineComponent( list(coef = coef, df = 0, offset = NULL, family = model@family)) c(list(comp.1), FLXmstep(as(model, "FLXMRglm"), weights[, -1, drop=FALSE], components[-1])) }) setMethod("FLXgetDesign", signature(object = "FLXMRziglm"), function(object, components) rbind(0, FLXgetDesign(as(object, "FLXMRglm"), components[-1]))) setMethod("FLXreplaceParameters", signature(object="FLXMRziglm"), function(object, components, parms) c(components[[1]], FLXreplaceParameters(as(object, "FLXMRglm"), components[-1], parms))) setMethod("FLXgradlogLikfun", signature(object="FLXMRziglm"), function(object, components, weights, ...) FLXgradlogLikfun(as(object, "FLXMRglm"), components[-1], weights[,-1,drop=FALSE])) setMethod("refit_optim", signature(object = "FLXMRziglm"), function(object, components, ...) { x <- refit_optim(as(object, "FLXMRglm"), components[-1], ...) names(x) <- paste("Comp", 1 + seq_along(x), sep = ".") x }) flexmix/R/lmmc.R0000644000176200001440000006511413431461313013210 0ustar liggesuserssetClass("FLXMRlmc", representation(family = "character", group = "factor", censored = "formula", C = "matrix"), contains = "FLXMR") setClass("FLXMRlmcfix", contains = "FLXMRlmc") setClass("FLXMRlmmc", representation(random = "formula", z = "matrix", which = "ANY"), contains = "FLXMRlmc") setClass("FLXMRlmmcfix", contains = "FLXMRlmmc") setMethod("allweighted", signature(model = "FLXMRlmc", control = "ANY", weights = "ANY"), function(model, control, weights) { if (!control@classify %in% c("auto", "weighted")) stop("Model class only supports weighted ML estimation.") model@weighted }) update.Residual <- function(fit, w, z, C, which, random, censored) { index <- lapply(C, function(x) x == 1) W <- rep(w, sapply(which, function(x) nrow(z[[x]]))) ZGammaZ <- sapply(seq_along(which), function(i) sum(diag(crossprod(z[[which[i]]]) %*% random$Gamma[[i]]))) WHICH <- which(sapply(C, sum) > 0) Residual <- if (length(WHICH) > 0) sum(sapply(WHICH, function(i) w[i] * sum(diag(censored$Sigma[[i]]) - 2 * z[[which[i]]][index[[i]],,drop=FALSE] * censored$psi[[i]]))) else 0 (sum(W * residuals(fit)^2) + Residual + sum(w * ZGammaZ))/sum(W) } update.latent <- function(x, y, C, fit) { AnyMissing <- which(sapply(C, sum) > 0) index <- lapply(C, function(x) x == 1) Sig <- lapply(seq_along(x), function(i) fit$sigma2 * diag(nrow = nrow(x[[i]]))) SIGMA <- rep(list(matrix(nrow = 0, ncol = 0)), length(x)) if (length(AnyMissing) > 0) { SIGMA[AnyMissing] <- lapply(AnyMissing, function(i) { S <- Sig[[i]] SIG <- S[index[[i]], index[[i]]] if (sum(!index[[i]]) > 0) SIG <- SIG - S[index[[i]],!index[[i]]] %*% solve(S[!index[[i]],!index[[i]]]) %*% S[!index[[i]],index[[i]]] SIG }) } Sigma <- MU <- rep(list(vector("numeric", length = 0)), length(x)) if (length(AnyMissing) > 0) { MU[AnyMissing] <- lapply(AnyMissing, function(i) { S <- Sig[[i]] Mu <- x[[i]][index[[i]],,drop=FALSE] %*% fit$coef if (sum(!index[[i]]) > 0) Mu <- Mu + S[index[[i]],!index[[i]]] %*% solve(S[!index[[i]],!index[[i]]]) %*% (y[[i]][!index[[i]]] - x[[i]][!index[[i]],,drop=FALSE] %*% fit$coef) Mu }) } moments <- lapply(seq_along(x), function(i) { if (sum(index[[i]]) > 0) moments_truncated(MU[[i]], SIGMA[[i]], y[[i]][C[[i]] == 1]) }) Sigma <- lapply(moments, "[[", "variance") censored <- list(mu = lapply(moments, "[[", "mean"), Sigma = Sigma) list(censored = censored) } update.latent.random <- function(x, y, z, C, which, fit) { index <- lapply(C, function(x) x == 1) AnyMissing <- which(sapply(C, sum) > 0) Residual <- fit$sigma2$Residual Psi <- fit$sigma2$Random EVbeta <- lapply(seq_along(z), function(i) solve(1/Residual * crossprod(z[[i]]) + solve(Psi))) Sig <- lapply(seq_along(z), function(i) z[[i]] %*% Psi %*% t(z[[i]]) + Residual * diag(nrow = nrow(z[[i]]))) SIGMA <- rep(list(matrix(nrow = 0, ncol = 0)), length(x)) if (length(AnyMissing) > 0) { SIGMA[AnyMissing] <- lapply(AnyMissing, function(i) { S <- Sig[[which[i]]] SIG <- S[index[[i]], index[[i]]] if (sum(!index[[i]]) > 0) SIG <- SIG - S[index[[i]],!index[[i]]] %*% solve(S[!index[[i]],!index[[i]]]) %*% S[!index[[i]],index[[i]]] SIG }) } Sigma <- MU <- rep(list(vector("numeric", length = 0)), length(x)) if (length(AnyMissing) > 0) { MU[AnyMissing] <- lapply(AnyMissing, function(i) { S <- Sig[[which[i]]] Mu <- x[[i]][index[[i]],,drop=FALSE] %*% fit$coef if (sum(!index[[i]]) > 0) { Mu <- Mu + S[index[[i]],!index[[i]]] %*% solve(S[!index[[i]],!index[[i]]]) %*% (y[[i]][!index[[i]]] - x[[i]][!index[[i]],,drop=FALSE] %*% fit$coef) } Mu }) } moments <- lapply(seq_along(x), function(i) { if (sum(index[[i]]) > 0) moments_truncated(MU[[i]], SIGMA[[i]], y[[i]][C[[i]] == 1]) }) Sigma <- lapply(moments, "[[", "variance") censored <- list(mu = lapply(moments, "[[", "mean"), Sigma = Sigma, psi = lapply(seq_along(x), function(i) { if (sum(index[[i]]) > 0) return(diag(Sigma[[i]] %*% z[[which[i]]][index[[i]],,drop=FALSE] %*% EVbeta[[which[i]]])/Residual) else return(vector("numeric", length = 0)) })) ybar <- lapply(seq_along(y), function(i) { Y <- y[[i]] Y[index[[i]]] <- censored$mu[[i]] Y }) random <- list(beta = lapply(seq_along(x), function(i) EVbeta[[which[i]]] %*% t(z[[which[i]]]) %*% (ybar[[i]] - x[[i]] %*% fit$coef)/Residual), Gamma = lapply(seq_along(x), function(i) { if (sum(index[[i]]) > 0) { return(EVbeta[[which[i]]] + (EVbeta[[which[i]]] %*% (t(z[[which[i]]][index[[i]],,drop=FALSE]) %*% censored$Sigma[[i]] %*% z[[which[i]]][index[[i]],,drop=FALSE]) %*% t(EVbeta[[which[i]]]))/Residual^2) } else return(EVbeta[[which[i]]]) })) list(random = random, censored = censored) } moments_truncated <- function(mu, Sigma, T, ...) { Sigma <- as.matrix(Sigma) mu <- as.vector(mu) T <- as.vector(T) S <- 1/sqrt(diag(Sigma)) T1 <- S * (T - mu) if (length(mu) == 1) { alpha <- pnorm(T1) dT1 <- dnorm(T1) Ex <- - dT1 / alpha Ex2 <- 1 - T1 * dT1 / alpha } else { R <- S * Sigma * rep(S, each = ncol(Sigma)) diag(R) <- 1L alpha <- mvtnorm::pmvnorm(upper = T1, sigma = R, ...) rq <- lapply(seq_along(T1), function(q) (R - tcrossprod(R[,q]))) R2 <- R^2 Vq <- 1 - R2 Sq <- sqrt(Vq) Rq <- lapply(seq_along(T1), function(q) rq[[q]]/(tcrossprod(Sq[,q]))) Tq <- lapply(seq_along(T1), function(q) (T1 - R[,q] * T1[q])/Sq[,q]) Phiq <- if (length(mu) == 1) 1 else sapply(seq_along(Rq), function(q) mvtnorm::pmvnorm(upper = Tq[[q]][-q], sigma = Rq[[q]][-q,-q], ...)) phi_Phiq <- dnorm(T1) * Phiq Ex <- - (R %*% phi_Phiq)/alpha T2_entries <- lapply(seq_along(T1), function(j) sapply(lapply(seq_along(T1)[seq_len(j)], function(i) R[,i] * T1 * phi_Phiq), function(z) sum(z * R[j,]))) T2 <- diag(length(T1)) T2[upper.tri(T2, diag = TRUE)] <- unlist(T2_entries) T2[lower.tri(T2)] <- t(T2)[lower.tri(T2)] phiqr <- lapply(seq_along(T1), function(q) sapply(seq_along(T1), function(r) { if (r == q) return(0) else return(mvtnorm::dmvnorm(T1[c(q, r)], mean = rep(0, length.out = length(c(q,r))), sigma = R[c(q,r), c(q,r)]))})) if (length(mu) == 2) { Ex2 <- R - T2 / alpha + Reduce("+", lapply(seq_along(Tq), function(q) tcrossprod(R[q,], rowSums(sapply(seq_along(Tq)[-q], function(r) phiqr[[q]][r] * (R[,r] - R[q,r] * R[q,])))))) / alpha } else { betaq <- lapply(seq_along(T1), function(q) sweep(rq[[q]], 2, Vq[,q], "/")) Rqr <- lapply(seq_along(T1), function(q) lapply(seq_along(T1), function(r) if (r == q) return(0) else return((Rq[[q]][-c(q,r),-c(q,r)] - tcrossprod(Rq[[q]][-c(q,r),r]))/tcrossprod(sqrt(1 - Rq[[q]][-c(q,r),r]^2))))) Tqr <- lapply(seq_along(T1), function(q) { lapply(seq_along(T1), function(r) if (r == q) return(0) else return((T1[-c(q,r)] - betaq[[r]][-c(r,q),q] * T1[q] - betaq[[q]][-c(r,q),r] * T1[r])/ (Sq[-c(r,q),q] * sqrt(1 - Rq[[q]][-c(r,q),r]^2))))}) T3 <- Reduce("+", lapply(seq_along(Tq), function(q) tcrossprod(R[q,], rowSums(sapply(seq_along(Tq)[-q], function(r) phiqr[[q]][r] * (R[,r] - R[q,r] * R[q,]) * mvtnorm::pmvnorm(upper = Tqr[[q]][[r]], sigma = Rqr[[q]][[r]], ...)))))) / alpha Ex2 <- R - T2 / alpha + 1/2 * (T3 + t(T3)) } } moments <- list(mean = 1/S * Ex + mu, variance = diag(1/S, nrow = length(T)) %*% (Ex2 - tcrossprod(Ex)) %*% diag(1/S, nrow = length(T))) if (!all(is.finite(unlist(moments))) || any(moments$mean > T) || any(eigen(moments$variance)$values < 0)) { moments <- list(mean = T - abs(diag(Sigma)), variance = Sigma) } moments } FLXMRlmmc <- function(formula = . ~ ., random, censored, varFix, eps = 10^-6, ...) { family <- "gaussian" censored <- if (length(censored) == 3) censored else formula(paste(".", paste(deparse(censored), collapse = ""))) if (missing(random)) { if (missing(varFix)) varFix <- FALSE else if ((length(varFix) > 1) || (is.na(as.logical(varFix)))) stop("varFix has to be a logical vector of length one") object <- new("FLXMRlmc", formula = formula, censored = censored, weighted = TRUE, family = family, name = "FLXMRlmc:gaussian") if (varFix) object <- new("FLXMRlmcfix", object) lmc.wfit <- function(x, y, w, C, censored) { W <- rep(w, sapply(x, nrow)) X <- do.call("rbind", x) AnyMissing <- which(sapply(C, sum) > 0) ybar <- lapply(seq_along(y), function(i) { Y <- y[[i]] Y[C[[i]] == 1] <- censored$mu[[i]] Y }) Y <- do.call("rbind", ybar) fit <- lm.wfit(X, Y, W, ...) fit$sigma2 <- if (length(AnyMissing) > 0) (sum(W * residuals(fit)^2) + sum(sapply(AnyMissing, function(i) w[i] * sum(diag(censored$Sigma[[i]])))))/sum(W) else sum(W * residuals(fit)^2)/sum(W) fit$df <- ncol(X) fit } object@defineComponent <- function(para) { predict <- function(x, ...) lapply(x, function(X) X %*% para$coef) logLik <- function(x, y, C, group, censored, ...) { AnyMissing <- which(sapply(C, sum) > 0) index <- lapply(C, function(x) x == 1) V <- lapply(x, function(X) diag(nrow(X)) * para$sigma2) mu <- predict(x, ...) SIGMA <- rep(list(matrix(nrow = 0, ncol = 0)), length(x)) if (length(AnyMissing) > 0) { SIGMA[AnyMissing] <- lapply(AnyMissing, function(i) { S <- V[[i]] SIG <- S[index[[i]], index[[i]]] if (sum(!index[[i]]) > 0) SIG <- SIG - S[index[[i]],!index[[i]]] %*% solve(S[!index[[i]],!index[[i]]]) %*% S[!index[[i]],index[[i]]] SIG }) } MU <- rep(list(vector("numeric", length = 0)), length(x)) if (length(AnyMissing) > 0) { MU[AnyMissing] <- lapply(AnyMissing, function(i) { S <- V[[i]] Mu <- mu[[i]][index[[i]]] if (sum(!index[[i]]) > 0) Mu <- Mu + S[index[[i]],!index[[i]]] %*% solve(S[!index[[i]],!index[[i]]]) %*% (y[[i]][!index[[i]]] - mu[[i]][!index[[i]]]) Mu }) } llh <- sapply(seq_along(x), function(i) { LLH <- 0 if (sum(index[[i]]) > 0) LLH <- log(mvtnorm::pmvnorm(upper = y[[i]][index[[i]]], mean = as.vector(MU[[i]]), sigma = SIGMA[[i]])) if (sum(!index[[i]]) > 0) LLH <- LLH + mvtnorm::dmvnorm(t(y[[i]][!index[[i]]]), mean = mu[[i]][!index[[i]]], sigma = V[[i]][!index[[i]], !index[[i]], drop = FALSE], log=TRUE) LLH/nrow(V[[i]]) }) as.vector(ungroupPriors(matrix(llh), group, !duplicated(group))) } new("FLXcomponent", parameters = list(coef = para$coef, sigma2 = para$sigma2, censored = para$censored), logLik = logLik, predict = predict, df = para$df) } object@fit <- if (varFix) { function(x, y, w, C, fit) { any_removed <- any(w <= eps) if (any_removed) { ok <- apply(w, 2, function(x) x > eps) W <- lapply(seq_len(ncol(ok)), function(i) w[ok[,i],i]) X <- lapply(seq_len(ncol(ok)), function(i) x[ok[,i],,drop = FALSE]) y <- lapply(seq_len(ncol(ok)), function(i) y[ok[,i]]) C <- lapply(seq_len(ncol(ok)), function(i) C[ok[,i]]) } else { X <- rep(list(x), ncol(w)) y <- rep(list(y), ncol(w)) C <- rep(list(C), ncol(w)) W <- lapply(seq_len(ncol(w)), function(i) w[,i]) } if ("coef" %in% names(fit[[1]])) fit <- lapply(seq_len(ncol(w)), function(k) update.latent(X[[k]], y[[k]], C[[k]], fit[[k]])) else { fit <- lapply(seq_len(ncol(w)), function(k) list(censored = list(mu = lapply(seq_along(y[[k]]), function(i) y[[k]][[i]][C[[k]][[i]] == 1]), Sigma = lapply(C[[k]], function(x) diag(1, nrow = sum(x)) * var(unlist(y[[k]])))))) } fit <- lapply(seq_len(ncol(w)), function(k) c(lmc.wfit(X[[k]], y[[k]], W[[k]], C[[k]], fit[[k]]$censored), censored = list(fit[[k]]$censored))) sigma2 <- sum(sapply(fit, function(x) x$sigma2) * colMeans(w)) for (k in seq_len(ncol(w))) fit[[k]]$sigma2 <- sigma2 lapply(fit, function(Z) object@defineComponent(list(coef = coef(Z), df = Z$df + 1/ncol(w), sigma2 = Z$sigma2, censored = Z$censored))) } } else { function(x, y, w, C, fit){ any_removed <- any(w <= eps) if (any_removed) { ok <- w > eps w <- w[ok] x <- x[ok,,drop = FALSE] y <- y[ok] C <- C[ok] } if ("coef" %in% names(fit)) { fit <- update.latent(x, y, C, fit) } else { fit$censored <- list(mu = lapply(seq_along(y), function(i) y[[i]][C[[i]] == 1]), Sigma = lapply(C, function(x) diag(1, nrow = sum(x)) * var(unlist(y)))) } fit <- c(lmc.wfit(x, y, w, C, fit$censored), censored = list(fit$censored)) object@defineComponent( list(coef = coef(fit), df = fit$df + 1, sigma2 = fit$sigma2, censored = fit$censored)) } } } else { if (missing(varFix)) varFix <- c(Random = FALSE, Residual = FALSE) else if (length(varFix) != 2 || is.null(names(varFix)) || any(is.na(pmatch(names(varFix), c("Random", "Residual"))))) stop("varFix has to be a named vector of length two") else names(varFix) <- c("Random", "Residual")[pmatch(names(varFix), c("Random", "Residual"))] random <- if (length(random) == 3) random else formula(paste(".", paste(deparse(random), collapse = ""))) object <- new("FLXMRlmmc", formula = formula, random = random, censored = censored, weighted = TRUE, family = family, name = "FLXMRlmmc:gaussian") if (any(varFix)) object <- new("FLXMRlmmcfix", object) add <- function(x) Reduce("+", x) lmmc.wfit <- function(x, y, w, z, C, which, random, censored) { effect <- lapply(seq_along(which), function(i) z[[which[i]]] %*% random$beta[[i]]) Effect <- do.call("rbind", effect) W <- rep(w, sapply(x, nrow)) X <- do.call("rbind", x) ybar <- lapply(seq_along(y), function(i) { Y <- y[[i]] Y[C[[i]] == 1] <- censored$mu[[i]] Y }) Y <- do.call("rbind", ybar) fit <- lm.wfit(X, Y - Effect, W, ...) wGamma <- add(lapply(seq_along(which), function(i) w[i] * random$Gamma[[i]])) bb <- add(lapply(seq_along(which), function(i) tcrossprod(random$beta[[i]]) * w[i])) fit$sigma2 <- list(Random = (wGamma + bb)/sum(w)) fit$df <- ncol(X) fit } object@defineComponent <- function(para) { predict <- function(x, ...) lapply(x, function(X) X %*% para$coef) logLik <- function(x, y, z, C, which, group, censored, ...) { AnyMissing <- which(sapply(C, sum) > 0) index <- lapply(C, function(x) x == 1) V <- lapply(z, function(Z) tcrossprod(tcrossprod(Z, para$sigma2$Random), Z) + diag(nrow(Z)) * para$sigma2$Residual) mu <- predict(x, ...) SIGMA <- rep(list(matrix(nrow = 0, ncol = 0)), length(x)) if (length(AnyMissing) > 0) { SIGMA[AnyMissing] <- lapply(AnyMissing, function(i) { S <- V[[which[i]]] SIG <- S[index[[i]], index[[i]]] if (sum(!index[[i]]) > 0) SIG <- SIG - S[index[[i]],!index[[i]]] %*% solve(S[!index[[i]],!index[[i]]]) %*% S[!index[[i]],index[[i]]] SIG }) } MU <- rep(list(vector("numeric", length = 0)), length(x)) if (length(AnyMissing) > 0) { MU[AnyMissing] <- lapply(AnyMissing, function(i) { S <- V[[which[i]]] Mu <- mu[[i]][index[[i]]] if (sum(!index[[i]]) > 0) Mu <- Mu + S[index[[i]],!index[[i]]] %*% solve(S[!index[[i]],!index[[i]]]) %*% (y[[i]][!index[[i]]] - mu[[i]][!index[[i]]]) Mu }) } llh <- sapply(seq_along(x), function(i) { LLH <- 0 if (sum(index[[i]]) > 0) LLH <- log(mvtnorm::pmvnorm(upper = y[[i]][index[[i]]], mean = as.vector(MU[[i]]), sigma = SIGMA[[i]])) if (sum(!index[[i]]) > 0) LLH <- LLH + mvtnorm::dmvnorm(t(y[[i]][!index[[i]]]), mean = mu[[i]][!index[[i]]], sigma = V[[which[i]]][!index[[i]], !index[[i]], drop = FALSE], log=TRUE) LLH/nrow(V[[which[i]]]) }) as.vector(ungroupPriors(matrix(llh), group, !duplicated(group))) } new("FLXcomponent", parameters = list(coef = para$coef, sigma2 = para$sigma2, censored = para$censored, random = para$random), logLik = logLik, predict = predict, df = para$df) } object@fit <- if (any(varFix)) { function(x, y, w, z, C, which, fit) { any_removed <- any(w <= eps) if (any_removed) { ok <- apply(w, 2, function(x) x > eps) W <- lapply(seq_len(ncol(ok)), function(i) w[ok[,i],i]) X <- lapply(seq_len(ncol(ok)), function(i) x[ok[,i],,drop = FALSE]) y <- lapply(seq_len(ncol(ok)), function(i) y[ok[,i]]) C <- lapply(seq_len(ncol(ok)), function(i) C[ok[,i]]) which <- lapply(seq_len(ncol(ok)), function(i) which[ok[,i]]) } else { X <- rep(list(x), ncol(w)) y <- rep(list(y), ncol(w)) C <- rep(list(C), ncol(w)) which <- rep(list(which), ncol(w)) W <- lapply(seq_len(ncol(w)), function(i) w[,i]) } if ("coef" %in% names(fit[[1]])) fit <- lapply(seq_len(ncol(w)), function(k) update.latent.random(X[[k]], y[[k]], z, C[[k]], which[[k]], fit[[k]])) else { fit <- lapply(seq_len(ncol(w)), function(k) list(random = list(beta = lapply(W[[k]], function(i) rep(0, ncol(z[[i]]))), Gamma = lapply(W[[k]], function(i) diag(ncol(z[[i]])))), censored = list(mu = lapply(seq_along(y[[k]]), function(i) y[[k]][[i]][C[[k]][[i]] == 1]), Sigma = lapply(C[[k]], function(x) diag(1, nrow = sum(x)) * var(unlist(y[[k]]))), psi = lapply(C[[k]], function(x) rep(0, sum(x)))))) } fit <- lapply(seq_len(ncol(w)), function(k) c(lmmc.wfit(X[[k]], y[[k]], W[[k]], z, C[[k]], which[[k]], fit[[k]]$random, fit[[k]]$censored), random = list(fit[[k]]$random), censored = list(fit[[k]]$censored))) if (varFix["Random"]) { prior_w <- apply(w, 2, weighted.mean, w = sapply(x, length)) Psi <- add(lapply(seq_len(ncol(w)), function(k) fit[[k]]$sigma2$Random * prior_w[k])) for (k in seq_len(ncol(w))) fit[[k]]$sigma2$Random <- Psi } for (k in seq_len(ncol(w))) fit[[k]]$sigma2$Residual <- update.Residual(fit[[k]], W[[k]], z, C[[k]], which[[k]], fit[[k]]$random, fit[[k]]$censored) if (varFix["Residual"]) { prior <- colMeans(w) Residual <- sum(sapply(fit[[k]]$sigma2$Residual, function(x) x) * prior) for (k in seq_len(ncol(w))) fit[[k]]$sigma2$Residual <- Residual } n <- nrow(fit[[1]]$sigma2$Random) lapply(fit, function(Z) object@defineComponent( list(coef = coef(Z), df = Z$df + n*(n+1)/(2*ifelse(varFix["Random"], ncol(w), 1)) + ifelse(varFix["Residual"], 1/ncol(w), 1), sigma2 = Z$sigma2, random = Z$random, censored = Z$censored))) } } else { function(x, y, w, z, C, which, fit){ any_removed <- any(w <= eps) if (any_removed) { ok <- w > eps w <- w[ok] x <- x[ok,,drop = FALSE] y <- y[ok] C <- C[ok] which <- which[ok] } if ("coef" %in% names(fit)) fit <- update.latent.random(x, y, z, C, which, fit) else { fit <- list(random = list(beta = lapply(which, function(i) rep(0, ncol(z[[i]]))), Gamma = lapply(which, function(i) diag(ncol(z[[i]])))), censored = list(mu = lapply(seq_along(y), function(i) y[[i]][C[[i]] == 1]), Sigma = lapply(C, function(x) diag(1, nrow = sum(x)) * var(unlist(y))), psi = lapply(C, function(x) rep(0, sum(x))))) } fit <- c(lmmc.wfit(x, y, w, z, C, which, fit$random, fit$censored), random = list(fit$random), censored = list(fit$censored)) fit$sigma2$Residual <- update.Residual(fit, w, z, C, which, fit$random, fit$censored) n <- nrow(fit$sigma2$Random) object@defineComponent( list(coef = coef(fit), df = fit$df + n*(n+1)/2 + 1, sigma2 = fit$sigma2, random = fit$random, censored = fit$censored)) } } } object } setMethod("FLXmstep", signature(model = "FLXMRlmc"), function(model, weights, components) { weights <- weights[!duplicated(model@group),,drop=FALSE] return(sapply(1:ncol(weights), function(k) model@fit(model@x, model@y, weights[,k], model@C, components[[k]]@parameters))) }) setMethod("FLXmstep", signature(model = "FLXMRlmcfix"), function(model, weights, components) { weights <- weights[!duplicated(model@group),,drop=FALSE] return(model@fit(model@x, model@y, weights, model@C, lapply(components, function(x) x@parameters))) }) setMethod("FLXmstep", signature(model = "FLXMRlmmc"), function(model, weights, components) { weights <- weights[!duplicated(model@group),,drop=FALSE] return(sapply(1:ncol(weights), function(k) model@fit(model@x, model@y, weights[,k], model@z, model@C, model@which, components[[k]]@parameters))) }) setMethod("FLXmstep", signature(model = "FLXMRlmmcfix"), function(model, weights, components) { weights <- weights[!duplicated(model@group),,drop=FALSE] return(model@fit(model@x, model@y, weights, model@z, model@C, model@which, lapply(components, function(x) x@parameters))) }) setMethod("FLXgetModelmatrix", signature(model="FLXMRlmc"), function(model, data, formula, lhs=TRUE, ...) { formula_nogrouping <- RemoveGrouping(formula) if (formula_nogrouping == formula) stop("please specify a grouping variable") model <- callNextMethod(model, data, formula, lhs) model@fullformula <- update(model@fullformula, paste(".~. |", .FLXgetGroupingVar(formula))) mt2 <- terms(model@censored, data=data) mf2 <- model.frame(delete.response(mt2), data=data, na.action = NULL) model@C <- model.matrix(attr(mf2, "terms"), data) model@group <- grouping <- .FLXgetGrouping(formula, data)$group model@x <- matrix(lapply(unique(grouping), function(g) model@x[grouping == g, , drop = FALSE]), ncol = 1) if (lhs) model@y <- matrix(lapply(unique(grouping), function(g) model@y[grouping == g, , drop = FALSE]), ncol = 1) model@C <- matrix(lapply(unique(grouping), function(g) model@C[grouping == g, , drop = FALSE]), ncol = 1) model }) setMethod("FLXgetModelmatrix", signature(model="FLXMRlmmc"), function(model, data, formula, lhs=TRUE, ...) { model <- callNextMethod(model, data, formula, lhs) mt1 <- terms(model@random, data=data) mf1 <- model.frame(delete.response(mt1), data=data, na.action = NULL) model@z <- model.matrix(attr(mf1, "terms"), data) grouping <- .FLXgetGrouping(formula, data)$group z <- matrix(lapply(unique(grouping), function(g) model@z[grouping == g, , drop = FALSE]), ncol = 1) model@z <- unique(z) model@which <- match(z, model@z) model }) setMethod("FLXgetObs", "FLXMRlmc", function(model) sum(sapply(model@x, nrow))) setMethod("FLXdeterminePostunscaled", signature(model = "FLXMRlmc"), function(model, components, ...) { sapply(components, function(x) x@logLik(model@x, model@y, model@C, model@group, x@parameters$censored)) }) setMethod("FLXdeterminePostunscaled", signature(model = "FLXMRlmmc"), function(model, components, ...) { sapply(components, function(x) x@logLik(model@x, model@y, model@z, model@C, model@which, model@group, x@parameters$censored)) }) setMethod("predict", signature(object="FLXMRlmc"), function(object, newdata, components, ...) { object <- FLXgetModelmatrix(object, newdata, formula = object@fullformula, lhs = FALSE) lapply(components, function(comp) unlist(comp@predict(object@x, ...))) }) setMethod("rFLXM", signature(model = "FLXMRlmc", components = "FLXcomponent"), function(model, components, ...) { stop("This model driver is not implemented yet.") }) flexmix/R/infocrit.R0000644000176200001440000000277613425024235014103 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: infocrit.R 5079 2016-01-31 12:21:12Z gruen $ # setMethod("nobs", signature(object="flexmix"), function(object, ...) { if (is.null(object@weights)) nrow(object@posterior$scaled) else sum(object@weights) }) setMethod("logLik", signature(object="flexmix"), function(object, newdata, ...){ if (missing(newdata)) { z <- object@logLik attr(z, "df") <- object@df attr(z, "nobs") <- nobs(object) class(z) <- "logLik" } else { z <- sum(log(rowSums(posterior(object, newdata = newdata, unscaled = TRUE)))) attr(z, "df") <- object@df attr(z, "nobs") <- nrow(newdata) class(z) <- "logLik" } z }) setMethod("ICL", signature(object="flexmix"), function(object, ...){ -2 * clogLik(object) + object@df * log(nobs(object)) }) setMethod("clogLik", signature(object="flexmix"), function(object, ...){ first <- if (length(object@group)) groupFirst(object@group) else TRUE post <- object@posterior$unscaled[first,,drop=FALSE] n <- nrow(post) sum(log(post[seq_len(n) + (clusters(object)[first] - 1)*n])) }) setMethod("EIC", signature(object="flexmix"), function(object, ...) { first <- if (length(object@group)) groupFirst(object@group) else TRUE post <- object@posterior$scaled[first,,drop=FALSE] n <- nrow(post) lpost <- log(post) if (any(is.infinite(lpost))) lpost[is.infinite(lpost)] <- -10^3 1 + sum(post * lpost)/(n * log(object@k)) }) flexmix/R/lmm.R0000644000176200001440000002370413430477461013056 0ustar liggesuserssetClass("FLXcomponentlmm", representation(random="list"), contains = "FLXcomponent") setClass("FLXMRlmm", representation(family = "character", random = "formula", group = "factor", z = "matrix", which = "ANY"), contains = "FLXMR") setClass("FLXMRlmmfix", contains = "FLXMRlmm") setMethod("allweighted", signature(model = "FLXMRlmm", control = "ANY", weights = "ANY"), function(model, control, weights) { if (!control@classify %in% c("auto", "weighted")) stop("Model class only supports weighted ML estimation.") model@weighted }) FLXMRlmm <- function(formula = . ~ ., random, lm.fit = c("lm.wfit", "smooth.spline"), varFix = c(Random = FALSE, Residual = FALSE), ...) { family <- "gaussian" lm.fit <- match.arg(lm.fit) if (length(varFix) != 2 || is.null(names(varFix)) || any(is.na(pmatch(names(varFix), c("Random", "Residual"))))) stop("varFix has to be a named vector of length two") else names(varFix) <- c("Random", "Residual")[pmatch(names(varFix), c("Random", "Residual"))] random <- if (length(random) == 3) random else formula(paste(".", paste(deparse(random), collapse = ""))) object <- new("FLXMRlmm", formula = formula, random = random, weighted = TRUE, family = family, name = "FLXMRlmm:gaussian") if (any(varFix)) object <- new("FLXMRlmmfix", object) object@preproc.y <- function(x){ if (ncol(x) > 1) stop(paste("y must be univariate")) x } if (lm.fit == "smooth.spline") { object@preproc.x <- function(x){ if (ncol(x) > 1) stop(paste("x must be univariate")) x } } add <- function(x) Reduce("+", x) lmm.wfit <- function(x, y, w, z, which, random) { effect <- lapply(seq_along(which), function(i) z[[which[i]]] %*% random$beta[[i]]) W <- rep(w, sapply(x, nrow)) X <- do.call("rbind", x) Y <- do.call("rbind", y) Effect <- do.call("rbind", effect) fit <- get(lm.fit)(X, Y - Effect, W, ...) XSigmaX <- sapply(seq_along(z), function(i) sum(diag(crossprod(z[[i]]) %*% random$Sigma[[i]]))) wSum <- tapply(w, which, sum) sigma2 <- (sum(W*residuals(fit)^2) + sum(wSum*XSigmaX))/sum(W) wSigma <- add(lapply(seq_along(z), function(i) wSum[i]*random$Sigma[[i]])) bb <- add(lapply(seq_along(which), function(i) tcrossprod(random$beta[[i]])*w[i])) psi <- (wSigma + bb)/sum(w) list(coefficients = if (is(fit, "smooth.spline")) fit$fit else coef(fit), sigma2 = list(Random = psi, Residual = sigma2), df = if (is(fit, "smooth.spline")) fit$df else ncol(x[[1]])) } object@defineComponent <- function(para) { predict <- function(x, ...) { if (is(para$coef, "smooth.spline.fit")) lapply(x, function(X) getS3method("predict", "smooth.spline.fit")(para$coef, X)$y) else lapply(x, function(X) X %*% para$coef) } logLik <- function(x, y, z, which, group, ...) { V <- lapply(z, function(Z) tcrossprod(tcrossprod(Z, para$sigma2$Random), Z) + diag(nrow(Z)) * para$sigma2$Residual) mu <- predict(x, ...) llh <- sapply(seq_along(x), function(i) mvtnorm::dmvnorm(t(y[[i]]), mean = mu[[i]], sigma = V[[which[i]]], log=TRUE)/nrow(V[[which[i]]])) as.vector(ungroupPriors(matrix(llh), group, !duplicated(group))) } new("FLXcomponentlmm", parameters = list(coef = para$coef, sigma2 = para$sigma2), random = list(), logLik = logLik, predict = predict, df = para$df) } determineRandom <- function(mu, y, z, which, sigma2) { Sigma <- lapply(z, function(Z) solve(crossprod(Z) / sigma2$Residual + solve(sigma2$Random))) Sigma_tilde <- lapply(seq_along(z), function(i) (tcrossprod(Sigma[[i]], z[[i]])/sigma2$Residual)) beta <- lapply(seq_along(which), function(i) Sigma_tilde[[which[i]]] %*% (y[[i]] - mu[[i]])) list(beta = beta, Sigma = Sigma) } object@fit <- if (any(varFix)) { function(x, y, w, z, which, random) { fit <- lapply(seq_len(ncol(w)), function(k) lmm.wfit(x, y, w[,k], z, which, random[[k]])) if (varFix["Random"]) { prior_w <- apply(w, 2, weighted.mean, w = sapply(x, length)) Random <- add(lapply(seq_along(fit), function(i) fit[[i]]$sigma2$Random * prior_w[i])) for (i in seq_along(fit)) fit[[i]]$sigma2$Random <- Random } if (varFix["Residual"]) { prior <- colMeans(w) Residual <- sum(sapply(fit, function(x) x$sigma2$Residual) * prior) for (i in seq_along(fit)) fit[[i]]$sigma2$Residual <- Residual } n <- nrow(fit[[1]]$sigma2$Random) lapply(fit, function(Z) { comp <- object@defineComponent(list(coef = coef(Z), sigma2 = Z$sigma2, df = Z$df + n*(n+1)/(2*ifelse(varFix["Random"], ncol(w), 1)) + ifelse(varFix["Residual"], 1/ncol(w), 1))) comp@random <- determineRandom(comp@predict(x), y, z, which, comp@parameters$sigma2) comp }) } } else { function(x, y, w, z, which, random){ fit <- lmm.wfit(x, y, w, z, which, random) n <- nrow(fit$sigma2$Random) comp <- object@defineComponent( list(coef = coef(fit), df = fit$df + n*(n+1)/2 + 1, sigma2 = fit$sigma2)) comp@random <- determineRandom(comp@predict(x), y, z, which, comp@parameters$sigma2) comp } } object } setMethod("FLXmstep", signature(model = "FLXMRlmm"), function(model, weights, components) { weights <- weights[!duplicated(model@group),,drop=FALSE] if (!is(components[[1]], "FLXcomponentlmm")) { random <- list(beta = lapply(model@which, function(i) rep(0, ncol(model@z[[i]]))), Sigma = lapply(model@z, function(x) diag(ncol(x)))) return(sapply(seq_len(ncol(weights)), function(k) model@fit(model@x, model@y, weights[,k], model@z, model@which, random))) }else { return(sapply(seq_len(ncol(weights)), function(k) model@fit(model@x, model@y, weights[,k], model@z, model@which, components[[k]]@random))) } }) setMethod("FLXmstep", signature(model = "FLXMRlmmfix"), function(model, weights, components) { weights <- weights[!duplicated(model@group),,drop=FALSE] if (!is(components[[1]], "FLXcomponentlmm")) { random <- rep(list(list(beta = lapply(model@which, function(i) rep(0, ncol(model@z[[i]]))), Sigma = lapply(model@z, function(x) diag(ncol(x))))), ncol(weights)) return(model@fit(model@x, model@y, weights, model@z, model@which, random)) }else return(model@fit(model@x, model@y, weights, model@z, model@which, lapply(components, function(x) x@random))) }) setMethod("FLXgetModelmatrix", signature(model="FLXMRlmm"), function(model, data, formula, lhs=TRUE, ...) { formula_nogrouping <- RemoveGrouping(formula) if (identical(paste(deparse(formula_nogrouping), collapse = ""), paste(deparse(formula), collapse = ""))) stop("please specify a grouping variable") model <- callNextMethod(model, data, formula, lhs) model@fullformula <- update(model@fullformula, paste(".~. |", .FLXgetGroupingVar(formula))) mt1 <- terms(model@random, data=data) mf <- model.frame(delete.response(mt1), data=data, na.action = NULL) model@z <- model.matrix(attr(mf, "terms"), data) model@group <- grouping <- .FLXgetGrouping(formula, data)$group model@x <- matrix(lapply(unique(grouping), function(g) model@x[grouping == g, , drop = FALSE]), ncol = 1) if (lhs) model@y <- matrix(lapply(unique(grouping), function(g) model@y[grouping == g, , drop = FALSE]), ncol = 1) z <- lapply(unique(grouping), function(g) model@z[grouping == g, , drop = FALSE]) z1 <- unique(z) model@which <- sapply(z, function(y) which(sapply(z1, function(x) isTRUE(all.equal(x, y))))) model@z <- matrix(z1, ncol = 1) model }) setMethod("FLXgetObs", "FLXMRlmm", function(model) sum(sapply(model@x, nrow))) setMethod("FLXdeterminePostunscaled", signature(model = "FLXMRlmm"), function(model, components, ...) { sapply(components, function(x) x@logLik(model@x, model@y, model@z, model@which, model@group)) }) setMethod("predict", signature(object="FLXMRlmm"), function(object, newdata, components, ...) { object <- FLXgetModelmatrix(object, newdata, formula = object@fullformula, lhs = FALSE) lapply(components, function(comp) unlist(comp@predict(object@x, ...))) }) setMethod("rFLXM", signature(model="FLXMRlmm", components="list"), function(model, components, class, group, ...) { class <- class[!duplicated(group)] y <- NULL for (l in seq_along(components)) { yl <- as.matrix(rFLXM(model, components[[l]], ...)) if (is.null(y)) y <- matrix(NA, nrow = length(class), ncol = ncol(yl)) y[class == l,] <- yl[class==l,,drop=FALSE] y <- matrix(y, ncol = ncol(yl)) } y }) setMethod("rFLXM", signature(model = "FLXMRlmm", components = "FLXcomponent"), function(model, components, ...) { sigma2 <- components@parameters$sigma2 V <- lapply(model@z, function(Z) tcrossprod(tcrossprod(Z, sigma2$Random), Z) + diag(nrow(Z)) * sigma2$Residual) mu <- components@predict(model@x) matrix(lapply(seq_along(model@x), function(i) t(mvtnorm::rmvnorm(1, mean = mu[[i]], sigma = V[[model@which[i]]]))), ncol = 1) }) setMethod("FLXgetNewModelmatrix", "FLXMRlmm", function(object, model, indices, groups) { object@y <- model@y[indices,,drop=FALSE] object@x <- model@x[indices,,drop=FALSE] object@which <- model@which[indices] if (length(unique(object@which)) < length(object@z)) { object@z <- model@z[sort(unique(object@which)),,drop=FALSE] object@which <- match(object@which, sort(unique(object@which))) } object }) flexmix/R/examples.R0000644000176200001440000000411313425024235014067 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: examples.R 5079 2016-01-31 12:21:12Z gruen $ # ExNPreg = function(n) { if(n %% 2 != 0) stop("n must be even") x <- runif(2*n, 0, 10) mp <- exp(c(2-0.2*x[1:n], 1+0.1*x[(n+1):(2*n)])) mb <- binomial()$linkinv(c(x[1:n]-5, 5-x[(n+1):(2*n)])) data.frame(x=x, yn=c(5*x[1:n], 40-(x[(n+1):(2*n)]-5)^2)+3*rnorm(n), yp=rpois(2*n, mp), yb=rbinom(2*n, size=1, prob=mb), class = rep(1:2, c(n,n)), id1 = factor(rep(1:n, rep(2, n))), id2 = factor(rep(1:(n/2), rep(4, n/2)))) } ExNclus = function(n=100) { if(n %% 2 != 0) stop("n must be even") rbind(mvtnorm::rmvnorm(n, mean=rep(0,2)), mvtnorm::rmvnorm(n, mean=c(8,0), sigma=diag(1:2)), mvtnorm::rmvnorm(1.5*n, mean=c(-2,6), sigma=diag(2:1)), mvtnorm::rmvnorm(2*n, mean=c(4,4), sigma=matrix(c(1,.9,.9,1), 2))) } ExLinear <- function(beta, n, xdist="runif", xdist.args=NULL, family=c("gaussian", "poisson"), sd=1, ...) { family <- match.arg(family) X <- NULL y <- NULL k <- ncol(beta) d <- nrow(beta)-1 n <- rep(n, length.out=k) if(family=="gaussian") sd <- rep(sd, length.out=k) xdist <- rep(xdist, length.out=d) if(is.null(xdist.args)){ xdist.args <- list(list(...)) } if(!is.list(xdist.args[[1]])) xdist.args <- list(xdist.args) xdist.args <- rep(xdist.args, length.out=d) for(i in 1:k) { X1 <- 1 for(j in 1:d){ xdist.args[[j]]$n <- n[i] X1 <- cbind(X1, do.call(xdist[j], xdist.args[[j]])) } X <- rbind(X, X1) xb <- X1 %*% beta[,i,drop=FALSE] if(family=="gaussian") y1 <- xb + rnorm(n[i], sd=sd[i]) else y1 <- rpois(n[i], exp(xb)) y <- c(y, y1) } X <- X[,-1,drop=FALSE] colnames(X) <- paste("x", 1:d, sep="") z <- data.frame(y=y, X) attr(z, "clusters") <- rep(1:k, n) z } flexmix/R/flxdist.R0000644000176200001440000000554213425024235013735 0ustar liggesusersFLXdist <- function(formula, k = NULL, model=FLXMRglm(), components, concomitant=FLXPconstant()) { mycall <- match.call() if(is(model, "FLXM")) model <- list(model) if (length(k)==1) prior <- rep(1/k, k) else { prior <- k/sum(k) } concomitant@x <- matrix(c(1, rep(0, ncol(concomitant@coef))[-1]), nrow = 1) prior <- as.vector(evalPrior(prior, concomitant)) lf <- length(formula) formula1 <- formula if(length(formula[[lf]])>1 && deparse(formula[[lf]][[1]]) == "|") formula1[[lf]] <- formula[[lf]][[2]] for(n in seq(along.with=model)) { if(is.null(model[[n]]@formula)) model[[n]]@formula <- formula1 else if(length(model[[n]]@formula) == 3 && model[[n]]@formula[[2]] == ".") model[[n]]@formula <- model[[n]]@formula[-2] model[[n]]@fullformula <- update.formula(formula1, model[[n]]@formula) } if (missing(components)) stop("no parameter values specified") if (length(components) != length(prior)) stop("components not specified correctly") comp <- list() for (k in seq(along.with=prior)) { comp[[k]] <- list() if (length(components[[k]]) != length(model)) stop("components not specified correctly") for (n in seq(along.with=model)) { comp[[k]][[n]] <- FLXcomponent(model[[n]], components[[k]][[n]]) } } new("FLXdist", formula=formula, call=mycall, concomitant=concomitant, prior=prior, k=length(prior), model=model, components=comp) } ###********************************************************** setGeneric("FLXcomponent", function(object, ...) standardGeneric("FLXcomponent")) setMethod("FLXcomponent", signature(object="FLXM"), function(object, components, ...) { components$df <- numeric() if (is(object@defineComponent, "expression")) eval(object@defineComponent, components) else object@defineComponent(components) }) #### setMethod("FLXcomponent", signature(object="FLXMRglm"), function(object, components, ...) { components$df <- numeric() offset <- NULL family <- object@family if (is(object@defineComponent, "expression")) eval(object@defineComponent, components) else object@defineComponent(components) }) ###********************************************************** setMethod("show", "FLXdist", function(object){ cat("\nCall:", deparse(object@call,0.75*getOption("width")), sep="\n") cat("\nPriors:\n") names(object@prior) <- paste("Comp.", seq_along(object@prior), sep="") print(object@prior) cat("\n") }) ###********************************************************** evalPrior <- function(prior, concomitant) prior setGeneric("evalPrior", function(prior, concomitant) standardGeneric("evalPrior")) setMethod("evalPrior", signature(concomitant="FLXPmultinom"), function(prior, concomitant) { exps <- exp(concomitant@x %*% concomitant@coef) exps/rowSums(exps) }) flexmix/R/group.R0000644000176200001440000000061513425024235013410 0ustar liggesuserssetMethod("group", signature(object="flexmix"), function(object) { group <- object@group if (!length(group)) group <- group(object@model[[1]]) group }) setMethod("group", signature(object="FLXM"), function(object) { factor(seq_len(nrow(object@x))) }) setMethod("group", signature(object="FLXMRglmfix"), function(object) { factor(seq_len(nrow(object@x)/sum(object@nestedformula@k))) }) flexmix/R/robust.R0000644000176200001440000000421713425024235013574 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: robust.R 5079 2016-01-31 12:21:12Z gruen $ # ###********************************************************* setClass("FLXMRrobglm", representation(bgw="logical"), prototype(bgw=FALSE), contains = "FLXMRglm") FLXMRrobglm <- function(formula = . ~ ., family=c("gaussian", "poisson"), bgw=FALSE, ...) { family <- match.arg(family) new("FLXMRrobglm", FLXMRglm(formula, family, ...), name = paste("FLXMRrobglm", family, sep=":"), bgw = bgw) } setMethod("FLXgetModelmatrix", signature(model="FLXMRrobglm"), function(model, data, formula, lhs=TRUE, ...) { model <- callNextMethod(model, data, formula, lhs) if (attr(terms(model@fullformula), "intercept")==0) stop("please include an intercept") new("FLXMRrobglm", model) }) setMethod("FLXremoveComponent", signature(model = "FLXMRrobglm"), function(model, nok, ...) { if (1 %in% nok) model <- as(model, "FLXMRglm") model }) setMethod("FLXmstep", signature(model = "FLXMRrobglm"), function(model, weights, ...) { if(model@bgw){ w <- weights[,1] } else{ w <- rep(1, nrow(weights)) } if(model@family=="gaussian") { cwt <- cov.wt(model@y, w) coef <- c(cwt$center, rep(0, ncol(model@x)-1)) names(coef) <- colnames(model@x) comp.1 <- model@defineComponent(list(coef = coef, df = 0, offset = NULL, sigma=sqrt(cwt$cov), family = model@family)) } else if(model@family=="poisson") { cwt <- cov.wt(model@y, w) coef <- c(log(3*cwt$center), rep(0, ncol(model@x)-1)) names(coef) <- colnames(model@x) comp.1 <- model@defineComponent(list(coef = coef, df = 0, offset = NULL, family = model@family)) } else{ stop("Other families not implemented yet!") } c(list(comp.1), FLXmstep(as(model, "FLXMRglm"), weights[, -1, drop=FALSE], ...)) }) flexmix/R/lmer.R0000644000176200001440000001506413431371013013213 0ustar liggesuserssigmaMethod <- getExportedValue(if(getRversion() >= "3.3.0") "stats" else "lme4", "sigma") setClass("FLXMRlmer", representation(random = "formula", lmod = "list", control = "ANY", preproc.z = "function"), prototype(preproc.z = function(x, ...) x), contains = "FLXMRglm") defineComponent_lmer <- function(para) { predict <- function(x, ...) x%*%para$coef logLik <- function(x, y, lmod, ...) { z <- as.matrix(lmod$reTrms$Zt) grouping <- lmod$reTrms$flist[[1]] llh <- vector(length=nrow(x)) for (i in seq_len(nlevels(grouping))) { index1 <- which(grouping == levels(grouping)[i]) index2 <- rownames(z) %in% levels(grouping)[i] V <- crossprod(z[index2,index1,drop=FALSE], para$sigma2$Random) %*% z[index2, index1, drop=FALSE] + diag(length(index1)) * para$sigma2$Residual llh[index1] <- mvtnorm::dmvnorm(y[index1,], mean=predict(x[index1,,drop=FALSE], ...), sigma = V, log=TRUE)/length(index1) } llh } new("FLXcomponent", parameters=list(coef=para$coef, sigma2=para$sigma2), logLik=logLik, predict=predict, df=para$df) } FLXMRlmer <- function(formula = . ~ ., random, weighted = TRUE, control = list(), eps = .Machine$double.eps) { random <- if (length(random) == 3) random else formula(paste(".", paste(deparse(random), collapse = ""))) missCtrl <- missing(control) if (missCtrl || !inherits(control, "lmerControl")) { if (!is.list(control)) stop("'control' is not a list; use lmerControl()") control <- do.call(lme4::lmerControl, control) } object <- new("FLXMRlmer", formula = formula, random = random, control = control, family = "gaussian", weighted = weighted, name = "FLXMRlmer:gaussian") if (weighted) object@preproc.z <- function(lmod) { if (length(unique(names(lmod[["reTrms"]][["flist"]]))) != 1) stop("only a single variable for random effects is allowed") for (i in seq_along(lmod[["reTrms"]][["flist"]])) { DIFF <- t(sapply(levels(lmod[["reTrms"]]$flist[[i]]), function(id) { index1 <- which(lmod[["reTrms"]]$flist[[i]] == id) index2 <- rownames(lmod[["reTrms"]]$Zt) == id sort(apply(lmod[["reTrms"]]$Zt[index2, index1, drop=FALSE], 1, paste, collapse = "")) })) if (length(unique(table(lmod[["reTrms"]][["flist"]][[i]]))) != 1 || nrow(unique(DIFF)) != 1) stop("FLXMRlmer does only work correctly if the covariates of the random effects are the same for all observations") } lmod } lmer.wfit <- function(x, y, w, lmod) { zero.weights <- any(w < eps) if (zero.weights) { ok <- w >= eps w <- w[ok] lmod[["fr"]] <- lmod[["fr"]][ok, , drop = FALSE] lmod[["X"]] <- lmod[["X"]][ok, , drop = FALSE] lmod[["reTrms"]][["Zt"]] <- lmod[["reTrms"]][["Zt"]][, ok, drop = FALSE] for (i in seq_along(lmod[["reTrms"]][["flist"]])) { lmod[["reTrms"]][["flist"]][[i]] <- lmod[["reTrms"]][["flist"]][[i]][ok] } } wts <- sqrt(w) lmod$X <- lmod$X * wts lmod$fr[[1]] <- lmod$fr[[1]] * wts devfun <- do.call(lme4::mkLmerDevfun, c(lmod, list(start = NULL, verbose = FALSE, control = control))) opt <- lme4::optimizeLmer(devfun, optimizer = control$optimizer, restart_edge = control$restart_edge, control = control$optCtrl, verbose = FALSE, start = NULL) mer <- lme4::mkMerMod(environment(devfun), opt, lmod$reTrms, fr = lmod$fr) sigma_res <- sigmaMethod(mer) / sqrt(mean(w)) vc <- lme4::VarCorr(mer) n <- c(0, cumsum(sapply(vc, ncol))) Random <- matrix(0, max(n), max(n)) for (i in seq_along(vc)) { index <- (n[i]+1):n[i+1] Random[index, index] <- vc[[i]] } Random <- Random / mean(w) list(coefficients = lme4::fixef(mer), sigma2 = list(Random = Random, Residual = sigma_res^2), df = length(lme4::fixef(mer)) + 1 + length(mer@theta)) } object@defineComponent <- defineComponent_lmer object@fit <- function(x, y, w, lmod){ fit <- lmer.wfit(x, y, w, lmod) object@defineComponent( list(coef = coef(fit), df = fit$df, sigma2 = fit$sigma2)) } object } setMethod("FLXgetModelmatrix", signature(model="FLXMRlmer"), function(model, data, formula, lhs=TRUE, contrasts = NULL, ...) { formula_nogrouping <- RemoveGrouping(formula) if (identical(paste(deparse(formula_nogrouping), collapse = ""), paste(deparse(formula), collapse = ""))) stop("please specify a grouping variable") model <- callNextMethod(model, data, formula, lhs) random_formula <- update(model@random, paste(".~. |", .FLXgetGroupingVar(formula))) fullformula <- model@fullformula if (!lhs) fullformula <- fullformula[c(1,3)] fullformula <- update(fullformula, paste(ifelse(lhs, ".", ""), "~. + ", paste(deparse(random_formula[[3]]), collapse = ""))) model@fullformula <- update(model@fullformula, paste(ifelse(lhs, ".", ""), "~. |", .FLXgetGroupingVar(formula))) model@lmod <- lme4::lFormula(fullformula, data, REML = FALSE, control = model@control) model@lmod <- model@preproc.z(model@lmod) model }) setMethod("FLXmstep", signature(model = "FLXMRlmer"), function(model, weights, ...) { apply(weights, 2, function(w) model@fit(model@x, model@y, w, model@lmod)) }) setMethod("FLXdeterminePostunscaled", signature(model = "FLXMRlmer"), function(model, components, ...) { sapply(components, function(x) x@logLik(model@x, model@y, model@lmod)) }) setMethod("rFLXM", signature(model = "FLXMRlmer", components="FLXcomponent"), function(model, components, ...) { sigma2 <- components@parameters$sigma2 z <- as.matrix(model@lmod$reTrms$Zt) grouping <- model@lmod$reTrms$flist[[1]] y <- matrix(0, nrow=nrow(model@x), ncol = 1) for (i in seq_len(nlevels(grouping))) { index1 <- which(grouping == levels(grouping)[i]) index2 <- rownames(z) %in% levels(grouping)[i] V <- crossprod(z[index2,index1,drop=FALSE], sigma2$Random) %*% z[index2, index1, drop=FALSE] + diag(length(index1)) * sigma2$Residual y[index1, 1] <- mvtnorm::rmvnorm(1, mean=components@predict(model@x[index1,,drop=FALSE], ...), sigma = V) } y }) flexmix/R/stepFlexmix.R0000644000176200001440000001351313425024235014565 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: stepFlexmix.R 5079 2016-01-31 12:21:12Z gruen $ # setClass("stepFlexmix", representation(models="list", k="integer", nrep="integer", logLiks="matrix", call="call")) stepFlexmix <- function(..., k=NULL, nrep=3, verbose=TRUE, drop=TRUE, unique=FALSE) { MYCALL <- match.call() MYCALL1 <- MYCALL bestFlexmix <- function(...) { z = new("flexmix", logLik=-Inf) logLiks = rep(NA, length.out = nrep) for(m in seq_len(nrep)){ if(verbose) cat(" *") x = try(flexmix(...)) if (!is(x, "try-error")) { logLiks[m] <- logLik(x) if(logLik(x) > logLik(z)) z = x } } return(list(z = z, logLiks = logLiks)) } z = list() if(is.null(k)){ RET = bestFlexmix(...) z[[1]] <- RET$z logLiks <- as.matrix(RET$logLiks) z[[1]]@call <- MYCALL z[[1]]@control@nrep <- nrep names(z) <- as.character(z[[1]]@k) if(verbose) cat("\n") } else{ k = as.integer(k) logLiks <- matrix(nrow = length(k), ncol = nrep) for(n in seq_along(k)){ ns <- as.character(k[n]) if(verbose) cat(k[n], ":") RET <- bestFlexmix(..., k=k[n]) z[[ns]] = RET$z logLiks[n,] <- RET$logLiks MYCALL1[["k"]] <- as.numeric(k[n]) z[[ns]]@call <- MYCALL1 z[[ns]]@control@nrep <- nrep if(verbose) cat("\n") } } logLiks <- logLiks[is.finite(sapply(z, logLik)),,drop=FALSE] z <- z[is.finite(sapply(z, logLik))] rownames(logLiks) <- names(z) if (!length(z)) stop("no convergence to a suitable mixture") if(drop & (length(z)==1)){ return(z[[1]]) } else{ z <- return(new("stepFlexmix", models=z, k=as.integer(names(z)), nrep=as.integer(nrep), logLiks=logLiks, call=MYCALL)) if(unique) z <- unique(z) return(z) } } ###********************************************************** setMethod("unique", "stepFlexmix", function(x, incomparables=FALSE, ...) { z <- list() K <- sapply(x@models, function(x) x@k) logLiks <- x@logLiks keep <- rep(TRUE, nrow(logLiks)) for(k in sort(unique(K))){ n <- which(k==K) if(length(n)>1){ l <- sapply(x@models[n], logLik) z[as.character(k)] <- x@models[n][which.max(l)] keep[n[-which.max(l)]] <- FALSE } else z[as.character(k)] <- x@models[n] } logLiks <- logLiks[keep,,drop=FALSE] rownames(logLiks) <- names(z) attr(logLiks, "na.action") <- NULL mycall <- x@call mycall["unique"] <- TRUE return(new("stepFlexmix", models=z, k=as.integer(names(z)), nrep=x@nrep, logLiks=logLiks, call=mycall)) }) ###********************************************************** setMethod("show", "stepFlexmix", function(object) { cat("\nCall:", deparse(object@call,0.75*getOption("width")), sep="\n") cat("\n") z <- data.frame(iter = sapply(object@models, function(x) x@iter), converged = sapply(object@models, function(x) x@converged), k = sapply(object@models, function(x) x@k), k0 = sapply(object@models, function(x) x@k0), logLik = sapply(object@models, function(x) logLik(x)), AIC = AIC(object), BIC = BIC(object), ICL = ICL(object)) print(z, na.string="") }) setMethod("nobs", signature(object="stepFlexmix"), function(object, ...) { sapply(object@models, nobs) }) setMethod("logLik", "stepFlexmix", function(object, ..., k = 2) { ll <- lapply(object@models, function(x) logLik(x)) df <- sapply(ll, attr, "df") nobs <- sapply(ll, attr, "nobs") ll <- unlist(ll) attr(ll, "df") <- df attr(ll, "nobs") <- nobs class(ll) <- "logLik" ll }) setMethod("ICL", "stepFlexmix", function(object, ...) { sapply(object@models, function(x) ICL(x, ...)) }) setMethod("EIC", "stepFlexmix", function(object, ...) { sapply(object@models, function(x) EIC(x, ...)) }) ###********************************************************** setMethod("getModel", "stepFlexmix", function(object, which="BIC") { if(which=="AIC") which <- which.min(sapply(object@models, function(x) AIC(x))) if(which=="BIC") which <- which.min(sapply(object@models, function(x) BIC(x))) if(which=="ICL") which <- which.min(sapply(object@models, function(x) ICL(x))) object@models[[which]] } ) ###********************************************************** setMethod("plot", signature(x="stepFlexmix", y="missing"), function(x, y, what=c("AIC", "BIC", "ICL"), xlab=NULL, ylab=NULL, legend="topright", ...) { X <- x@k Y <- NULL for(w in what){ Y <- cbind(Y, do.call(w, list(object=x))) } if(is.null(xlab)) xlab <- "number of components" if(is.null(ylab)){ if(length(what)==1) ylab <- what else ylab <- "" } matplot(X, Y, xlab=xlab, ylab=ylab, type="b", lty=1, pch=seq_along(what), ...) if(legend!=FALSE && length(what)>1) legend(x=legend, legend=what, pch=seq_along(what), col=seq_along(what)) for(n in seq_len(ncol(Y))){ m <- which.min(Y[,n]) points(X[m], Y[m,n], pch=16, cex=1.5, col=n) } }) flexmix/R/factanal.R0000644000176200001440000000345213425024235014027 0ustar liggesuserssetClass("FLXMCfactanal", contains = "FLXMC") ###********************************************************** FLXMCfactanal <- function(formula=.~., factors = 1, ...) { z <- new("FLXMCfactanal", weighted=TRUE, formula=formula, dist = "mvnorm", name="mixtures of factor analyzers") z@fit <- function(x, y, w, ...){ cov.weighted <- cov.wt(y, wt = w)[c("center","cov")] cov <- cov.weighted$cov; center <- cov.weighted$center fa <- factanal(covmat = cov, factors = factors, ...) Sigma <- fa$loadings %*% t(fa$loadings) + diag(fa$uniquenesses) df <- (factors + 2) * ncol(y) predict <- function(x) matrix(center, nrow=nrow(x), ncol=length(center), byrow=TRUE) logLik <- function(x, y){ sds <- sqrt(diag(cov)) mvtnorm::dmvnorm(y, mean = center, sigma = Sigma * (sds %o% sds), log = TRUE) } new("FLXcomponent", parameters=list(mu = center, variance = diag(cov), loadings = fa$loadings, uniquenesses = fa$uniquenesses), df=df, logLik=logLik, predict=predict) } z } ###********************************************************** setMethod("rFLXM", signature(model = "FLXMCfactanal", components = "FLXcomponent"), function(model, components, class, ...) { FUN <- paste("r", model@dist, sep = "") Sigma <- components@parameters$loadings %*% t(components@parameters$loadings) + diag(components@parameters$uniquenesses) sds <- sqrt(components@parameters$variance) args <- list(n = nrow(model@x), mean = components@parameters$mu, sigma = Sigma * (sds %o% sds)) return(do.call(FUN, args)) }) flexmix/R/refit.R0000644000176200001440000005270513425024235013374 0ustar liggesusers# # Copyright (C) 2004-2016 Friedrich Leisch and Bettina Gruen # $Id: refit.R 5079 2016-01-31 12:21:12Z gruen $ # ###********************************************************* setMethod("FLXgetParameters", signature(object="FLXdist"), function(object, model) { if (missing(model)) model <- seq_along(object@model) coefficients <- unlist(lapply(model, function(m) { Model <- unlist(FLXgetParameters(object@model[[m]], lapply(object@components, "[[", m))) names(Model) <- paste("model.", m, "_", names(Model), sep = "") Model })) c(coefficients, FLXgetParameters(object@concomitant)) }) setMethod("FLXgetParameters", signature(object="FLXM"), function(object, components, ...) { lapply(components, function(x) unlist(slot(x, "parameters"))) }) setMethod("FLXgetParameters", signature(object="FLXMC"), function(object, components, ...) { if (object@dist == "mvnorm") { return(lapply(components, function(x) { pars <- x@parameters if (identical(pars$cov, diag(diag(pars$cov)))) return(c(pars$center, diag(pars$cov))) else return(c(pars$center, pars$cov[lower.tri(pars$cov, diag = TRUE)])) })) } else return(lapply(components, function(x) unlist(slot(x, "parameters")))) }) setMethod("FLXgetParameters", signature(object="FLXMRglm"), function(object, components, ...) { parms <- lapply(components, function(x) unlist(slot(x, "parameters"))) Design <- FLXgetDesign(object, components) if (object@family == "gaussian") { parms <- lapply(parms, function(x) { x["sigma"] <- log(x["sigma"]) x }) colnames(Design) <- gsub("sigma$", "log(sigma)", colnames(Design)) } parms_unique <- vector(length = ncol(Design)) names(parms_unique) <- colnames(Design) for (k in seq_along(parms)) parms_unique[as.logical(Design[k,])] <- parms[[k]] parms_unique }) setMethod("FLXgetParameters", signature(object="FLXP"), function(object, ...) { if (length(object@coef) == 1) return(NULL) alpha <- log(object@coef[-1]) - log(object@coef[1]) names(alpha) <- paste("concomitant", paste("Comp", seq_along(object@coef)[-1], "alpha", sep = "."), sep = "_") return(alpha) }) setMethod("FLXgetParameters", signature(object="FLXPmultinom"), function(object, ...) { coefficients <- object@coef[,-1,drop=FALSE] if (ncol(coefficients) > 0) { Names <- paste("Comp", rep(seq_len(ncol(coefficients)+1)[-1], each = nrow(coefficients)), rownames(coefficients), sep = ".") coefficients <- as.vector(coefficients) names(coefficients) <- paste("concomitant", Names, sep = "_") return(coefficients) }else return(NULL) }) setMethod("VarianceCovariance", signature(object="flexmix"), function(object, model = TRUE, gradient, optim_control = list(), ...) { if (object@control@classify != "weighted") stop("Only for weighted ML estimation possible.") if (length(FLXgetParameters(object)) != object@df) stop("not implemented yet for restricted parameters.") if (missing(gradient)) gradient <- FLXgradlogLikfun(object) optim_control$fnscale <- -1 fit <- optim(fn = FLXlogLikfun(object), par = FLXgetParameters(object), gr = gradient, hessian = TRUE, method = "BFGS", control = optim_control, ...) list(coef = fit$par, vcov = -solve(as.matrix(fit$hessian))) }) setMethod("logLikfun_comp", signature(object="flexmix"), function(object) { postunscaled <- matrix(0, nrow = FLXgetObs(object@model[[1]]), ncol = object@k) for (m in seq_along(object@model)) postunscaled <- postunscaled + FLXdeterminePostunscaled(object@model[[m]], lapply(object@components, "[[", m)) if(length(object@group)>0) postunscaled <- groupPosteriors(postunscaled, object@group) postunscaled }) setMethod("FLXlogLikfun", signature(object="flexmix"), function(object, ...) function(parms) { object <- FLXreplaceParameters(object, parms) groupfirst <- if (length(object@group) > 1) groupFirst(object@group) else rep(TRUE, FLXgetObs(object@model[[1]])) logpostunscaled <- logLikfun_comp(object) + log(getPriors(object@concomitant, object@group, groupfirst)) if (is.null(object@weights)) sum(log_row_sums(logpostunscaled[groupfirst,,drop=FALSE])) else sum(log_row_sums(logpostunscaled[groupfirst,,drop=FALSE])*object@weights[groupfirst]) }) setMethod("getPriors", signature(object="FLXP"), function(object, group, groupfirst) { priors <- matrix(apply(object@coef, 2, function(x) object@x %*% x), nrow = nrow(object@x)) ungroupPriors(priors/rowSums(priors), group, groupfirst) }) setMethod("getPriors", signature(object="FLXPmultinom"), function(object, group, groupfirst) { priors <- matrix(apply(object@coef, 2, function(x) exp(object@x %*% x)), nrow = nrow(object@x)) ungroupPriors(priors/rowSums(priors), group, groupfirst) }) setMethod("FLXreplaceParameters", signature(object="FLXdist"), function(object, parms) { comp_names <- names(object@components) components <- list() for (m in seq_along(object@model)) { indices <- grep(paste("^model.", m, sep = ""), names(parms)) components[[m]] <- FLXreplaceParameters(object@model[[m]], lapply(object@components, "[[", m), parms[indices]) } object@components <- lapply(seq_along(object@components), function(k) lapply(components, "[[", k)) names(object@components) <- comp_names if (object@k > 1) { indices <- grep("^concomitant_", names(parms)) object@concomitant <- FLXreplaceParameters(object@concomitant, parms[indices]) } object }) setMethod("FLXreplaceParameters", signature(object="FLXM"), function(object, components, parms) { Design <- FLXgetDesign(object, components) lapply(seq_along(components), function(k) { Parameters <- list() parms_k <- parms[as.logical(Design[k,])] for (i in seq_along(components[[k]]@parameters)) { Parameters[[i]] <- parms_k[seq_along(components[[k]]@parameters[[i]])] attributes(Parameters[[i]]) <- attributes(components[[k]]@parameters[[i]]) parms_k <- parms_k[-seq_along(components[[k]]@parameters[[i]])] } names(Parameters) <- names(components[[k]]@parameters) Parameters$df <- components[[k]]@df variables <- c("x", "y") for (var in variables) assign(var, slot(object, var)) if (is(object@defineComponent, "expression")) eval(object@defineComponent, Parameters) else object@defineComponent(Parameters) }) }) setMethod("FLXreplaceParameters", signature(object="FLXMC"), function(object, components, parms) { Design <- FLXgetDesign(object, components) if (object@dist == "mvnorm") { p <- sqrt(1/4+ncol(Design)/nrow(Design)) - 1/2 diagonal <- get("diagonal", environment(object@fit)) if (diagonal) { cov <- diag(seq_len(p)) parms_comp <- as.vector(sapply(seq_len(nrow(Design)), function(i) c(parms[(i-1) * 2 * p + seq_len(p)], as.vector(diag(diag(parms[(i-1) * 2 * p + p + seq_len(p)])))))) parms <- c(parms_comp, parms[(nrow(Design) * 2 * p + 1):length(parms)]) } else { cov <- matrix(NA, nrow = p, ncol = p) cov[lower.tri(cov, diag = TRUE)] <- seq_len(sum(lower.tri(cov, diag = TRUE))) cov[upper.tri(cov)] <- t(cov)[upper.tri(cov)] parms <- parms[c(as.vector(sapply(seq_len(nrow(Design)), function(i) (i-1)*(max(cov)+p) + c(seq_len(p), as.vector(cov) + p))), (nrow(Design) * (max(cov) + p)+1):length(parms))] } } callNextMethod(object = object, components = components, parms = parms) }) setMethod("FLXreplaceParameters", signature(object="FLXMRglm"), function(object, components, parms) { Design <- FLXgetDesign(object, components) lapply(seq_along(components), function(k) { Parameters <- list() parms_k <- parms[as.logical(Design[k,])] for (i in seq_along(components[[k]]@parameters)) { Parameters[[i]] <- parms_k[seq_along(components[[k]]@parameters[[i]])] attributes(Parameters[[i]]) <- attributes(components[[k]]@parameters[[i]]) parms_k <- parms_k[-seq_along(components[[k]]@parameters[[i]])] } names(Parameters) <- names(components[[k]]@parameters) if (object@family == "gaussian") { Parameters[["sigma"]] <- exp(Parameters[["sigma"]]) } Parameters$df <- components[[k]]@df variables <- c("x", "y", "offset", "family") for (var in variables) { assign(var, slot(object, var)) } if (is(object@defineComponent, "expression")) eval(object@defineComponent, Parameters) else object@defineComponent(Parameters) }) }) setMethod("FLXreplaceParameters", signature(object="FLXP"), function(object, parms) { parms <- exp(c(0, parms)) parms <- parms/sum(parms) attributes(parms) <- attributes(object@coef) object@coef <- parms object }) setMethod("FLXreplaceParameters", signature(object="FLXPmultinom"), function(object, parms) { parms <- cbind(0, matrix(parms, nrow = nrow(object@coef))) attributes(parms) <- attributes(object@coef) object@coef <- parms object }) setMethod("FLXgradlogLikfun", signature(object="flexmix"), function(object, ...) { existFunction <- all(sapply(object@model, existGradient)) if (object@k > 1) existFunction <- c(existFunction, existGradient(object@concomitant)) if (any(!existFunction)) return(NULL) function(parms) { object <- FLXreplaceParameters(object, parms) groupfirst <- if (length(object@group) > 1) groupFirst(object@group) else rep(TRUE, FLXgetObs(object@model[[1]])) logLik_comp <- logLikfun_comp(object) Priors <- getPriors(object@concomitant, object@group, groupfirst) Priors_Lik_comp <- logLik_comp + log(Priors) weights <- exp(Priors_Lik_comp - log_row_sums(Priors_Lik_comp)) if (object@k > 1) { ConcomitantScores <- FLXgradlogLikfun(object@concomitant, Priors[groupfirst,,drop=FALSE], weights[groupfirst,,drop=FALSE]) if (!is.null(object@weights)) ConcomitantScores <- lapply(ConcomitantScores, "*", object@weights[groupfirst]) } else ConcomitantScores <- list() ModelScores <- lapply(seq_along(object@model), function(m) FLXgradlogLikfun(object@model[[m]], lapply(object@components, "[[", m), weights)) ModelScores <- lapply(ModelScores, lapply, groupPosteriors, object@group) if (!is.null(object@weights)) ModelScores <- lapply(ModelScores, lapply, "*", object@weights) colSums(cbind(do.call("cbind", lapply(ModelScores, function(x) do.call("cbind", x)))[groupfirst,,drop=FALSE], do.call("cbind", ConcomitantScores))) } }) setMethod("existGradient", signature(object = "FLXM"), function(object) FALSE) setMethod("existGradient", signature(object = "FLXMRglm"), function(object) { if (object@family == "Gamma") return(FALSE) TRUE }) setMethod("existGradient", signature(object = "FLXMRglmfix"), function(object) FALSE) setMethod("existGradient", signature(object = "FLXP"), function(object) TRUE) setMethod("FLXgradlogLikfun", signature(object="FLXMRglm"), function(object, components, weights, ...) { lapply(seq_along(components), function(k) { res <- if (object@family == "binomial") as.vector(object@y[,1] - rowSums(object@y)*components[[k]]@predict(object@x)) else as.vector(object@y - components[[k]]@predict(object@x)) Scores <- weights[,k] * res * object@x if (object@family == "gaussian") { Scores <- cbind(Scores/components[[k]]@parameters$sigma^2, weights[,k] * (-1 + res^2/components[[k]]@parameters$sigma^2)) } Scores }) }) setMethod("FLXgradlogLikfun", signature(object="FLXP"), function(object, fitted, weights, ...) { Pi <- lapply(seq_len(ncol(fitted))[-1], function(i) - fitted[,i] + weights[,i]) lapply(Pi, function(p) apply(object@x, 2, "*", p)) }) setMethod("refit", signature(object = "flexmix"), function(object, newdata, method = c("optim", "mstep"), ...) { method <- match.arg(method) if (method == "optim") { VarCov <- VarianceCovariance(object, ...) z <- new("FLXRoptim", call=sys.call(-1), k = object@k, coef = VarCov$coef, vcov = VarCov$vcov) z@components <- lapply(seq_along(object@model), function(m) { begin_name <- paste("^model", m, sep = ".") indices <- grep(begin_name, names(z@coef)) refit_optim(object@model[[m]], components = lapply(object@components, "[[", m), coef = z@coef[indices], se = sqrt(diag(z@vcov)[indices])) }) z@concomitant <- if (object@k > 1) { indices <- grep("^concomitant_", names(z@coef)) refit_optim(object@concomitant, coef = z@coef[indices], se = sqrt(diag(z@vcov)[indices])) } else NULL } else { z <- new("FLXRmstep", call=sys.call(-1), k = object@k) z@components <- lapply(object@model, function(x) { x <- refit_mstep(x, weights=object@posterior$scaled) names(x) <- paste("Comp", seq_len(object@k), sep=".") x }) z@concomitant <- if (object@k > 1) refit_mstep(object@concomitant, posterior = object@posterior$scaled, group = object@group, w = object@weights) else NULL } z }) setMethod("refit_optim", signature(object = "FLXM"), function(object, components, coef, se) { Design <- FLXgetDesign(object, components) x <- lapply(seq_len(nrow(Design)), function(k) { rval <- cbind(Estimate = coef[as.logical(Design[k,])], "Std. Error" = se[as.logical(Design[k,])]) pars <- components[[k]]@parameters[[1]] rval <- rval[seq_along(pars),,drop=FALSE] rownames(rval) <- names(pars) zval <- rval[,1]/rval[,2] new("Coefmat", cbind(rval, "z value" = zval, "Pr(>|z|)" = 2 * pnorm(abs(zval), lower.tail = FALSE))) }) names(x) <- paste("Comp", seq_along(x), sep = ".") x }) setMethod("refit_optim", signature(object = "FLXMC"), function(object, components, coef, se) { Design <- FLXgetDesign(object, components) if (object@dist == "mvnorm") { p <- length(grep("Comp.1_center", colnames(Design), fixed = TRUE)) diagonal <- get("diagonal", environment(object@fit)) if (diagonal) { cov <- diag(seq_len(p)) coef_comp <- as.vector(sapply(seq_len(nrow(Design)), function(i) c(coef[(i-1) * 2 * p + seq_len(p)], as.vector(diag(diag(coef[(i-1) * 2 * p + p + seq_len(p)])))))) coef <- c(coef_comp, coef[(nrow(Design) * 2 * p + 1):length(coef)]) se_comp <- as.vector(sapply(seq_len(nrow(Design)), function(i) c(se[(i-1) * 2 * p + seq_len(p)], as.vector(diag(diag(se[(i-1) * 2 * p + p + seq_len(p)])))))) se <- c(se_comp, se[(nrow(Design) * 2 * p + 1):length(se)]) } else { cov <- matrix(NA, nrow = p, ncol = p) cov[lower.tri(cov, diag = TRUE)] <- seq_len(sum(lower.tri(cov, diag = TRUE))) cov[upper.tri(cov)] <- t(cov)[upper.tri(cov)] coef <- coef[c(as.vector(sapply(seq_len(nrow(Design)), function(i) (i-1)*(max(cov)+p) + c(seq_len(p), as.vector(cov) + p))), (nrow(Design) * (max(cov) + p)+1):length(coef))] se <- se[c(as.vector(sapply(seq_len(nrow(Design)), function(i) (i-1)*(max(cov)+p) + c(seq_len(p), as.vector(cov) + p))), (nrow(Design) * (max(cov) + p)+1):length(se))] } } callNextMethod(object = object, components = components, coef = coef, se = se) }) setMethod("refit_optim", signature(object = "FLXP"), function(object, coef, se) { x <- lapply(seq_len(ncol(object@coef))[-1], function(k) { indices <- grep(paste("Comp", k, sep = "."), names(coef)) rval <- cbind(Estimate = coef[indices], "Std. Error" = se[indices]) rval <- rval[seq_len(nrow(object@coef)),,drop=FALSE] rownames(rval) <- rownames(object@coef) zval <- rval[,1]/rval[,2] new("Coefmat", cbind(rval, "z value" = zval, "Pr(>|z|)" = 2 * pnorm(abs(zval), lower.tail = FALSE))) }) names(x) <- paste("Comp", 1 + seq_along(x), sep = ".") x }) setMethod("FLXgetDesign", signature(object = "FLXM"), function(object, components, ...) { parms <- lapply(components, function(x) unlist(slot(x, "parameters"))) nr_parms <- sapply(parms, length) cumSum <- cumsum(c(0, nr_parms)) Design <- t(sapply(seq_len(length(cumSum)-1), function(i) rep(c(0, 1, 0), c(cumSum[i], nr_parms[i], max(cumSum) - cumSum[i] - nr_parms[i])))) colnames(Design) <- paste(rep(paste("Comp", seq_len(nrow(Design)), sep = "."), nr_parms), unlist(lapply(parms, names)), sep = "_") Design }) setMethod("FLXgetDesign", signature(object = "FLXMRglmfix"), function(object, components, ...) { if (length(components) == 1) return(callNextMethod(object, components, ...)) Design <- object@design if (object@family == "gaussian") { cumSum <- cumsum(c(0, object@variance)) variance <- matrix(sapply(seq_len(length(cumSum)-1), function(i) rep(c(0, 1, 0), c(cumSum[i], object@variance[i], length(components) - cumSum[i] - object@variance[i]))), nrow = length(components)) colnames(variance) <- paste("Comp", apply(variance, 2, function(x) which(x == 1)[1]), "sigma", sep= ".") Design <- cbind(Design, variance) } Design }) ###********************************************************* setMethod("refit_mstep", signature(object="FLXM"), function(object, newdata, weights, ...) { lapply(seq_len(ncol(weights)), function(k) object@fit(object@x, object@y, weights[,k], ...)@parameters) }) setMethod("refit_mstep", signature(object="FLXMRglm"), function(object, newdata, weights, ...) { lapply(seq_len(ncol(weights)), function(k) { fit <- object@refit(object@x, object@y, weights[,k], ...) fit <- c(fit, list(formula = object@fullformula, terms = object@terms, contrasts = object@contrasts, xlevels = object@xlevels)) class(fit) <- c("glm", "lm") fit }) }) ###********************************************************** setMethod("fitted", signature(object="flexmix"), function(object, drop=TRUE, aggregate = FALSE, ...) { x<- list() for(m in seq_along(object@model)) { comp <- lapply(object@components, "[[", m) x[[m]] <- fitted(object@model[[m]], comp, ...) } if (aggregate) { group <- group(object) prior_weights <- determinePrior(object@prior, object@concomitant, group)[as.integer(group),] z <- lapply(x, function(z) matrix(rowSums(do.call("cbind", z) * prior_weights), nrow = nrow(z[[1]]))) if(drop && all(lapply(z, ncol)==1)){ z <- sapply(z, unlist) } } else { z <- list() for (k in seq_len(object@k)) { z[[k]] <- do.call("cbind", lapply(x, "[[", k)) } names(z) <- paste("Comp", seq_len(object@k), sep=".") if(drop && all(lapply(z, ncol)==1)){ z <- sapply(z, unlist) } } z }) setMethod("fitted", signature(object="FLXM"), function(object, components, ...) { lapply(components, function(z) z@predict(object@x)) }) setMethod("predict", signature(object="FLXM"), function(object, newdata, components, ...) { object <- FLXgetModelmatrix(object, newdata, formula = object@fullformula, lhs = FALSE) z <- list() for(k in seq_along(components)) z[[k]] <- components[[k]]@predict(object@x, ...) z }) ###********************************************************** setMethod("Lapply", signature(object="FLXRmstep"), function(object, FUN, model = 1, component = TRUE, ...) { X <- object@components[[model]] lapply(X[component], FUN, ...) }) ###********************************************************* setMethod("refit_mstep", signature(object="flexmix", newdata="listOrdata.frame"), function(object, newdata, ...) { z <- new("FLXR", call=sys.call(-1), k = object@k) z@components <- lapply(object@model, function(x) { x <- refit_mstep(x, newdata = newdata, weights=posterior(object, newdata = newdata)) names(x) <- paste("Comp", seq_len(object@k), sep=".") x }) z@concomitant <- if (object@k > 1) refit_mstep(object@concomitant, newdata, object@posterior$scaled, object@group, w = object@weights) else NULL z }) setMethod("refit_mstep", signature(object="FLXMRglm", newdata="listOrdata.frame"), function(object, newdata, weights, ...) { w <- weights lapply(seq_len(ncol(w)), function(k) { newdata$weights <- weights <- w[,k] weighted.glm(formula = object@fullformula, data = newdata, family = object@family, weights = weights, ...) }) }) weighted.glm <- function(weights, ...) { fit <- eval(as.call(c(as.symbol("glm"), c(list(...), list(weights = weights, x = TRUE))))) fit$df.null <- sum(weights) + fit$df.null - fit$df.residual - fit$rank fit$df.residual <- sum(weights) - fit$rank fit$method <- "weighted.glm.fit" fit } weighted.glm.fit <- function(x, y, weights, offset = NULL, family = "gaussian", ...) { if (!is.function(family) & !is(family, "family")) family <- get(family, mode = "function", envir = parent.frame()) fit <- c(glm.fit(x, y, weights = weights, offset=offset, family=family), list(call = sys.call(), offset = offset, control = eval(formals(glm.fit)$control), method = "weighted.glm.fit")) fit$df.null <- sum(weights) + fit$df.null - fit$df.residual - fit$rank fit$df.residual <- sum(weights) - fit$rank fit$x <- x fit } flexmix/R/mgcv.R0000644000176200001440000001373113430265374013222 0ustar liggesuserssetOldClass("gam.prefit") setClassUnion("listOrgam.prefit", c("list", "gam.prefit")) setClass("FLXMRmgcv", representation(G = "listOrgam.prefit", control = "list"), contains="FLXMRglm") FLXMRmgcv <- function(formula = .~., family = c("gaussian", "binomial", "poisson"), offset = NULL, control = NULL, optimizer = c("outer", "newton"), in.out = NULL, eps = .Machine$double.eps, ...) { if (is.null(control)) control <- mgcv::gam.control() family <- match.arg(family) am <- if (family == "gaussian" && get(family)()$link == "identity") TRUE else FALSE z <- new("FLXMRmgcv", FLXMRglm(formula = formula, family = family, offset = offset), name=paste("FLXMRmgcv", family, sep=":"), control = control) scale <- if (family %in% c("binomial", "poisson")) 1 else -1 gam_fit <- function(G, w) { G$family <- get(family)() G$am <- am G$w <- w G$conv.tol <- control$mgcv.tol G$max.half <- control$mgcv.half zero_weights <- any(w < eps) if (zero_weights) { ok <- w >= eps w <- w[ok] G$X <- G$X[ok,,drop=FALSE] if (is.matrix(G$y)) G$y <- G$y[ok,,drop=FALSE] else G$y <- G$y[ok] G$mf <- G$mf[ok,,drop=FALSE] G$w <- G$w[ok] G$offset <- G$offset[ok] if (G$n.paraPen > 0) { OMIT <- which(colSums(abs(G$X)) == 0) if (length(OMIT) > 0) { Ncol <- ncol(G$X) Assign <- unique(G$assign[OMIT]) G$assign <- G$assign[-OMIT] G$nsdf <- G$nsdf - length(OMIT) G$X <- G$X[,-OMIT,drop=FALSE] G$mf$Grouping <- G$mf$Grouping[,-which(colSums(abs(G$mf$Grouping))==0),drop=FALSE] if (length(G$off) > 1) G$off[2] <- G$off[2] - length(OMIT) for (i in seq_along(G$smooth)) { G$smooth[[i]]$first.para <- G$smooth[[i]]$first.para - length(OMIT) G$smooth[[i]]$last.para <- G$smooth[[i]]$last.para - length(OMIT) } G$S[[1]] <- G$S[[1]][-c(OMIT-sum(G$assign != Assign)), -c(OMIT-sum(G$assign != Assign))] } } } z <- mgcv::gam(G = G, method = "ML", optimizer = optimizer, control = control, scale = scale, in.out = in.out, ...) if (zero_weights) { residuals <- z$residuals z$residuals <- rep(0, length(ok)) z$residuals[ok] <- residuals if (G$n.paraPen > 0 && length(OMIT) > 0) { coefficients <- z$coefficients z$coefficients <- rep(0, Ncol) z$coefficients[-OMIT] <- coefficients } } z } if (family=="gaussian"){ z@fit <- function(x, y, w, G){ gam.fit <- gam_fit(G, w) z@defineComponent(list(coef = gam.fit$coefficients, df = sum(gam.fit$edf)+1, sigma = sqrt(sum(w * gam.fit$residuals^2 / mean(w))/ (nrow(x)-sum(gam.fit$edf))))) } } else if(family %in% c("binomial", "poisson")){ z@fit <- function(x, y, w, G){ gam.fit <- gam_fit(G, w) z@defineComponent( list(coef = gam.fit$coefficients, df = sum(gam.fit$edf))) } } else stop(paste("Unknown family", family)) z } setMethod("FLXmstep", signature(model = "FLXMRmgcv"), function(model, weights, ...) { apply(weights, 2, function(w) model@fit(model@x, model@y, w, model@G)) }) setMethod("FLXgetModelmatrix", signature(model="FLXMRmgcv"), function(model, data, formula, lhs=TRUE, paraPen = list(), ...) { formula <- RemoveGrouping(formula) if (length(grep("\\|", deparse(model@formula)))) stop("no grouping variable allowed in the model") if(is.null(model@formula)) model@formula <- formula model@fullformula <- update(terms(formula, data=data), model@formula) gp <- mgcv::interpret.gam(model@fullformula) if (lhs) { model@terms <- terms(gp$fake.formula, data = data) mf <- model.frame(model@terms, data=data, na.action = NULL, drop.unused.levels = TRUE) response <- as.matrix(model.response(mf, "numeric")) model@y <- model@preproc.y(response) } else { model@terms <- terms(gp$fake.formula, data = data) mf <- model.frame(delete.response(model@terms), data=data, na.action = NULL, drop.unused.levels = TRUE) } model@G <- mgcv::gam(model@fullformula, data = data, fit = FALSE) model@x <- model@G$X model@contrasts <- attr(model@x, "contrasts") model@x <- model@preproc.x(model@x) model@xlevels <- .getXlevels(delete.response(model@terms), mf) model }) setMethod("predict", signature(object="FLXMRmgcv"), function(object, newdata, components, ...) { predict_gam <- function (object, newdata, ...) { nn <- names(newdata) mn <- colnames(object$model) for (i in 1:length(newdata)) if (nn[i] %in% mn && is.factor(object$model[, nn[i]])) { newdata[[i]] <- factor(newdata[[i]], levels = levels(object$model[, nn[i]])) } if (length(newdata) == 1) newdata[[2]] <- newdata[[1]] n.smooth <- length(object$smooth) Terms <- delete.response(object$pterms) X <- matrix(0, nrow(newdata), length(object$coefficients)) Xoff <- matrix(0, nrow(newdata), n.smooth) mf <- model.frame(Terms, newdata, xlev = object$xlevels) if (!is.null(cl <- attr(object$pterms, "dataClasses"))) .checkMFClasses(cl, mf) Xp <- model.matrix(Terms, mf, contrasts = object$contrasts) if (object$nsdf) X[, 1:object$nsdf] <- Xp if (n.smooth) for (k in 1:n.smooth) { Xfrag <- mgcv::PredictMat(object$smooth[[k]], newdata) X[, object$smooth[[k]]$first.para:object$smooth[[k]]$last.para] <- Xfrag Xfrag.off <- attr(Xfrag, "offset") if (!is.null(Xfrag.off)) { Xoff[, k] <- Xfrag.off } } X } object@G$model <- object@G$mf z <- list() for(k in seq_along(components)) { object@G$coefficients <- components[[k]]@parameters$coef X <- predict_gam(object@G, newdata) z[[k]] <- components[[k]]@predict(X, ...) } z }) flexmix/vignettes/0000755000176200001440000000000013432516322013737 5ustar liggesusersflexmix/vignettes/mymclust.R0000644000176200001440000000205313425024236015737 0ustar liggesusersmymclust <- function (formula = .~., diagonal = TRUE) { retval <- new("FLXMC", weighted = TRUE, formula = formula, dist = "mvnorm", name = "my model-based clustering") retval@defineComponent <- function(para) { logLik <- function(x, y) { mvtnorm::dmvnorm(y, mean = para$center, sigma = para$cov, log = TRUE) } predict <- function(x) { matrix(para$center, nrow = nrow(x), ncol = length(para$center), byrow = TRUE) } new("FLXcomponent", parameters = list(center = para$center, cov = para$cov), df = para$df, logLik = logLik, predict = predict) } retval@fit <- function(x, y, w, ...) { para <- cov.wt(y, wt = w)[c("center", "cov")] df <- (3 * ncol(y) + ncol(y)^2)/2 if (diagonal) { para$cov <- diag(diag(para$cov)) df <- 2 * ncol(y) } retval@defineComponent(c(para, df = df)) } retval } flexmix/vignettes/mixture-regressions.Rnw0000644000176200001440000022021513425024236020467 0ustar liggesusers% % Copyright (C) 2008 Bettina Gruen and Friedrich Leisch % $Id: mixture-regressions.Rnw $ % \documentclass[nojss]{jss} \usepackage{amsfonts} \title{FlexMix Version 2: Finite Mixtures with Concomitant Variables and Varying and Constant Parameters} \Plaintitle{FlexMix Version 2: Finite Mixtures with Concomitant Variables and Varying and Constant Parameters} \Shorttitle{FlexMix Version 2} \author{Bettina Gr{\"u}n\\ Johannes Kepler Universit{\"at} Linz \And Friedrich Leisch\\ Universit\"at f\"ur Bodenkultur Wien} \Plainauthor{Bettina Gr{\"u}n, Friedrich Leisch} \Address{ Bettina Gr\"un\\ Institut f\"ur Angewandte Statistik / IFAS\\ Johannes Kepler Universit{\"at} Linz\\ Freist\"adter Stra\ss{}e 315\\ 4040 Linz, Austria\\ E-mail: \email{Bettina.Gruen@jku.at}\\ Friedrich Leisch\\ Institut f\"ur Angewandte Statistik und EDV\\ Universit\"at f\"ur Bodenkultur Wien\\ Peter Jordan Stra\ss{}e 82\\ 1190 Wien, Austria\\ E-mail: \email{Friedrich.Leisch@boku.ac.at}\\ URL: \url{http://www.statistik.lmu.de/~leisch/} } \Abstract{ This article is a (slightly) modified version of \cite{mixtures:Gruen+Leisch:2008a}, published in the \emph{Journal of Statistical Software}. \pkg{flexmix} provides infrastructure for flexible fitting of finite mixture models in \proglang{R} using the expectation-maximization (EM) algorithm or one of its variants. The functionality of the package was enhanced. Now concomitant variable models as well as varying and constant parameters for the component specific generalized linear regression models can be fitted. The application of the package is demonstrated on several examples, the implementation described and examples given to illustrate how new drivers for the component specific models and the concomitant variable models can be defined. } \Keywords{\proglang{R}, finite mixture models, generalized linear models, concomitant variables} \Plainkeywords{R, finite mixture models, generalized linear models, concomitant variables} \usepackage{amsmath, listings} \def\argmax{\mathop{\rm arg\,max}} %% \usepackage{Sweave} prevent automatic inclusion \SweaveOpts{width=9, height=4.5, eps=FALSE, keep.source=TRUE} <>= options(width=60, prompt = "R> ", continue = "+ ", useFancyQuotes = FALSE) library("graphics") library("stats") library("flexmix") library("lattice") ltheme <- canonical.theme("postscript", FALSE) lattice.options(default.theme=ltheme) data("NPreg", package = "flexmix") data("dmft", package = "flexmix") source("myConcomitant.R") @ %%\VignetteIndexEntry{FlexMix Version 2: Finite Mixtures with Concomitant Variables and Varying and Constant Parameters} %%\VignetteDepends{flexmix} %%\VignetteKeywords{R, finite mixture models, model based clustering, latent class regression} %%\VignettePackage{flexmix} \begin{document} %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \section{Introduction}\label{sec:introduction} Finite mixture models are a popular technique for modelling unobserved heterogeneity or to approximate general distribution functions in a semi-parametric way. They are used in a lot of different areas such as astronomy, biology, economics, marketing or medicine. An overview on mixture models is given in \cite{mixtures:Everitt+Hand:1981}, \cite{mixtures:Titterington+Smith+Makov:1985}, \cite{mixtures:McLachlan+Basford:1988}, \cite{mixtures:Boehning:1999}, \cite{mixtures:McLachlan+Peel:2000} and \cite{mixtures:Fruehwirth-Schnatter:2006}. Version 1 of \proglang{R} package \pkg{flexmix} was introduced in \cite{mixtures:Leisch:2004}. The main design principles of the package are extensibility and fast prototyping for new types of mixture models. It uses \proglang{S}4 classes and methods \citep{mixtures:Chambers:1998} as implemented in the \proglang{R} package \pkg{methods} and exploits advanced features of \proglang{R} such as lexical scoping \citep{mixtures:Gentleman+Ihaka:2000}. The package implements a framework for maximum likelihood estimation with the expectation-maximization (EM) algorithm \citep{mixtures:Dempster+Laird+Rubin:1977}. The main focus is on finite mixtures of regression models and it allows for multiple independent responses and repeated measurements. The EM algorithm can be controlled through arguments such as the maximum number of iterations or a minimum improvement in the likelihood to continue. Newly introduced features in the current package version are concomitant variable models \citep{mixtures:Dayton+Macready:1988} and varying and constant parameters in the component specific regressions. Varying parameters follow a finite mixture, i.e., several groups exist in the population which have different parameters. Constant parameters are fixed for the whole population. This model is similar to mixed-effects models \citep{mixtures:Pinheiro+Bates:2000}. The main difference is that in this application the distribution of the varying parameters is unknown and has to be estimated. Thus the model is actually closer to the varying-coefficients modelling framework \citep{mixtures:Hastie+Tibshirani:1993}, using convex combinations of discrete points as functional form for the varying coefficients. The extension to constant and varying parameters allows for example to fit varying intercept models as given in \cite{mixtures:Follmann+Lambert:1989} and \cite{mixtures:Aitkin:1999}. These models are frequently applied to account for overdispersion in the data where the components follow either a binomial or Poisson distribution. The model was also extended to include nested varying parameters, i.e.~this allows to have groups of components with the same parameters \citep{mixtures:Gruen+Leisch:2006, mixtures:Gruen:2006}. In Section~\ref{sec:model-spec-estim} the extended model class is presented together with the parameter estimation using the EM algorithm. In Section~\ref{sec:using-new-funct} examples are given to demonstrate how the new functionality can be used. An overview on the implementational details is given in Section~\ref{sec:implementation}. The new model drivers are presented and changes made to improve the flexibility of the software and to enable the implementation of the new features are discussed. Examples for writing new drivers for the component specific models and the concomitant variable models are given in Section~\ref{sec:writing-your-own}. This paper gives a short overview on finite mixtures and the package in order to be self-contained. A more detailed introduction to finite mixtures and the package \pkg{flexmix} can be found in \cite{mixtures:Leisch:2004}. All computations and graphics in this paper have been done with \pkg{flexmix} version \Sexpr{packageDescription("flexmix",fields="Version")} and \proglang{R} version \Sexpr{getRversion()} using Sweave \citep{mixtures:Leisch:2002}. The newest release version of \pkg{flexmix} is always available from the Comprehensive \proglang{R} Archive Network at \url{http://CRAN.R-project.org/package=flexmix}. An up-to-date version of this paper is contained in the package as a vignette, giving full access to the \proglang{R} code behind all examples shown below. See \code{help("vignette")} or \cite{mixtures:Leisch:2003} for details on handling package vignettes. %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \section{Model specification and estimation}\label{sec:model-spec-estim} A general model class of finite mixtures of regression models is considered in the following. The mixture is assumed to consist of $K$ components where each component follows a parametric distribution. Each component has a weight assigned which indicates the a-priori probability for an observation to come from this component and the mixture distribution is given by the weighted sum over the $K$ components. If the weights depend on further variables, these are referred to as concomitant variables. In marketing choice behaviour is often modelled in dependence of marketing mix variables such as price, promotion and display. Under the assumption that groups of respondents with different price, promotion and display elasticities exist mixtures of regressions are fitted to model consumer heterogeneity and segment the market. Socio-demographic variables such as age and gender have often been shown to be related to the different market segments even though they generally do not perform well when used to a-priori segment the market. The relationships between the behavioural and the socio-demographic variables is then modelled through concomitant variable models where the group sizes (i.e.~the weights of the mixture) depend on the socio-demographic variables. The model class is given by \begin{align*} h(y|x, w, \psi) &= \sum_{k = 1}^K \pi_k(w, \alpha) f_k(y|x,\theta_{k})\\ &= \sum_{k = 1}^K \pi_k(w, \alpha) \prod_{d=1}^D f_{kd}(y_d|x_d,\theta_{kd}), \end{align*} where $\psi$ denotes the vector of all parameters for the mixture density $h()$ and is given by $(\alpha, (\theta_k)_{k=1,\ldots,K})$. $y$ denotes the response, $x$ the predictor and $w$ the concomitant variables. $f_k$ is the component specific density function. Multivariate variables $y$ are assumed to be dividable into $D$ subsets where the component densities are independent between the subsets, i.e.~the component density $f_k$ is given by a product over $D$ densities which are defined for the subset variables $y_d$ and $x_d$ for $d=1,\ldots,D$. The component specific parameters are given by $\theta_k = (\theta_{kd})_{d=1,\ldots,D}$. Under the assumption that $N$ observations are available the dimensions of the variables are given by $y = (y_d)_{d=1,\ldots,D} \in \mathbb{R}^{N \times \sum_{d=1}^D k_{yd}}$, $x = (x_d)_{d=1,\ldots,D} \in \mathbb{R}^{N \times \sum_{d=1}^D k_{xd}}$ for all $d = 1,\ldots,D$ and $w \in \mathbb{R}^{N \times k_w}$. In this notation $k_{yd}$ denotes the dimension of the $d^{\textrm{th}}$ response, $k_{xd}$ the dimension of the $d^{\textrm{th}}$ predictors and $k_w$ the dimension of the concomitant variables. For mixtures of GLMs each of the $d$ responses will in general be univariate, i.e.~multivariate responses will be conditionally independent given the segment memberships. For the component weights $\pi_k$ it holds $\forall w$ that \begin{equation}\label{eq:prior} \sum_{k=1}^K \pi_k(w,\alpha) = 1 \quad \textrm{and} \quad \pi_k(w, \alpha) > 0, \, \forall k, \end{equation} where $\alpha$ are the parameters of the concomitant variable model. For the moment focus is given to finite mixtures where the component specific densities are from the same parametric family, i.e.~$f_{kd} \equiv f_d$ for notational simplicity. If $f_d$ is from the exponential family of distributions and for each component a generalized linear model is fitted \citep[GLMs;][]{mixtures:McCullagh+Nelder:1989} these models are also called GLIMMIX models \citep{mixtures:Wedel+DeSarbo:1995}. In this case the component specific parameters are given by $\theta_{kd} = (\beta'_{kd}, \phi_{kd})$ where $\beta_{kd}$ are the regression coefficients and $\phi_{kd}$ is the dispersion parameter. The component specific parameters $\theta_{kd}$ are either restricted to be equal over all components, to vary between groups of components or to vary between all components. The varying between groups is referred to as varying parameters with one level of nesting. A disjoint partition $K_c$, $c = 1,\ldots,C$ of the set $\tilde{K} := \{1\ldots,K\}$ is defined for the regression coefficients. $C$ is the number of groups of the regression coefficients at the nesting level. The regression coefficients are accordingly split into three groups: \begin{align*} \beta_{kd} &= (\beta'_{1d}, \beta'_{2,c(k)d}, \beta'_{3,kd})', \end{align*} where $c(k) = \{c = 1,\ldots, C: k \in K_c\}$. Similar a disjoint partition $K_v$, $v = 1,\ldots,V$, of $\tilde{K}$ can be defined for the dispersion parameters if nested varying parameters are present. $V$ denotes the number of groups of the dispersion parameters at the nesting level. This gives: \begin{align*} \phi_{kd} &= \left\{\begin{array}{ll} \phi_{d} & \textrm{for constant parameters}\\ \phi_{kd} & \textrm{for varying parameters}\\ \phi_{v(k)d} & \textrm{for nested varying parameters} \end{array}\right. \end{align*} where $v(k) = \{v = 1,\ldots,V: k \in K_v\}$. The nesting structure of the component specific parameters is also described in \cite{mixtures:Gruen+Leisch:2006}. Different concomitant variable models are possible to determine the component weights \citep{mixtures:Dayton+Macready:1988}. The mapping function only has to fulfill condition \eqref{eq:prior}. In the following a multinomial logit model is assumed for the $\pi_k$ given by \begin{equation*} \pi_k(w,\alpha) = \frac{e^{w'\alpha_k}}{\sum_{u = 1}^K e^{w'\alpha_u}} \quad \forall k, \end{equation*} with $\alpha = (\alpha'_k)'_{k=1,\ldots,K}$ and $\alpha_1 \equiv 0$. %%------------------------------------------------------------------------- \subsection{Parameter estimation}\label{sec:estimation} The EM algorithm \citep{mixtures:Dempster+Laird+Rubin:1977} is the most common method for maximum likelihood estimation of finite mixture models where the number of components $K$ is fixed. The EM algorithm applies a missing data augmentation scheme. It is assumed that a latent variable $z_n \in \{0,1\}^K$ exists for each observation $n$ which indicates the component membership, i.e.~$z_{nk}$ equals 1 if observation $n$ comes from component $k$ and 0 otherwise. Furthermore it holds that $\sum_{k=1}^K z_{nk}=1$ for all $n$. In the EM algorithm these unobserved component memberships $z_{nk}$ of the observations are treated as missing values and the data is augmented by estimates of the component membership, i.e.~the estimated a-posteriori probabilities $\hat{p}_{nk}$. For a sample of $N$ observations $\{(y_1, x_1, w_1), \ldots, (y_N, x_N, w_N)\}$ the EM algorithm is given by: \begin{description} \item[E-step:] Given the current parameter estimates $\psi^{(i)}$ in the $i$-th iteration, replace the missing data $z_{nk}$ by the estimated a-posteriori probabilities \begin{align*} \hat{p}_{nk} & = \frac{\displaystyle \pi_k(w_n, \alpha^{(i)}) f(y_n| x_n, \theta_k^{(i)}) }{\displaystyle \sum_{u = 1}^K \pi_u(w_n, \alpha^{(i)}) f(y_n |x_n, \theta_u^{(i)}) }. \end{align*} \item[M-step:] Given the estimates for the a-posteriori probabilities $\hat{p}_{nk}$ (which are functions of $\psi^{(i)}$), obtain new estimates $\psi^{(i+1)}$ of the parameters by maximizing \begin{align*} Q(\psi^{(i+1)}|\psi^{(i)}) &= Q_1(\theta^{(i+1)} | \psi^{(i)}) + Q_2(\alpha^{(i+1)} | \psi^{(i)}), \end{align*} where \begin{align*} Q_1(\theta^{(i+1)} | \psi^{(i)}) &= \sum_{n = 1}^N \sum_{k = 1}^K \hat{p}_{nk} \log(f(y_n | x_n, \theta_k^{(i+1)})) \end{align*} and \begin{align*} Q_2(\alpha^{(i+1)}| \psi^{(i)}) &= \sum_{n = 1}^N \sum_{k = 1}^K \hat{p}_{nk} \log(\pi_k(w_n, \alpha^{(i+1)})). \end{align*} $Q_1$ and $Q_2$ can be maximized separately. The maximization of $Q_1$ gives new estimates $\theta^{(i+1)}$ and the maximization of $Q_2$ gives $\alpha^{(i+1)}$. $Q_1$ is maximized separately for each $d=1,\ldots,D$ using weighted ML estimation of GLMs and $Q_2$ using weighted ML estimation of multinomial logit models. \end{description} Different variants of the EM algorithm exist such as the stochastic EM \citep[SEM;][]{mixtures:Diebolt+Ip:1996} or the classification EM \citep[CEM;][]{mixtures:Celeux+Govaert:1992}. These two variants are also implemented in package \pkg{flexmix}. For both variants an additional step is made between the expectation and maximization steps. This step uses the estimated a-posteriori probabilities and assigns each observation to only one component, i.e.~classifies it into one component. For SEM this assignment is determined in a stochastic way while it is a deterministic assignment for CEM. For the SEM algorithm the additional step is given by: \begin{description} \item[S-step:] Given the a-posteriori probabilities draw \begin{align*} \hat{z}_n &\sim \textrm{Mult}((\hat{p}_{nk})_{k=1,\ldots,K}, 1) \end{align*} where $\textrm{Mult}(\theta, T)$ denotes the multinomial distribution with success probabilities $\theta$ and number of trials $T$. \end{description} Afterwards, the $\hat{z}_{nk}$ are used instead of the $\hat{p}_{nk}$ in the M-step. For the CEM the additional step is given by: \begin{description} \item[C-step:] Given the a-posteriori probabilities define \begin{align*} \hat{z}_{nk} &= \left\{\begin{array}{ll} 1&\textrm{if } k = \min\{ l : \hat{p}_{nl} \geq \hat{p}_{nk}\, \forall k=1,\ldots,K\}\\ 0&\textrm{otherwise}. \end{array}\right. \end{align*} \end{description} Please note that in this step the observation is assigned to the component with the smallest index if the same maximum a-posteriori probability is observed for several components. Both of these variants have been proposed to improve the performance of the EM algorithm, because the ordinary EM algorithm tends to converge rather slowly and only to a local optimum. The convergence behavior can be expected to be better for the CEM than ordinary EM algorithm, while SEM can escape convergence to a local optimum. However, the CEM algorithm does not give ML estimates because it maximizes the complete likelihood. For SEM good approximations of the ML estimator are obtained if the parameters where the maximum likelihood was encountered are used as estimates. Another possibility for determining parameter estimates from the SEM algorithm could be the mean after discarding a suitable number of burn-ins. An implementational advantage of both variants is that no weighted maximization is necessary in the M-step. It has been shown that the values of the likelihood are monotonically increased during the EM algorithm. On the one hand this ensures the convergence of the EM algorithm if the likelihood is bounded, but on the other hand only the detection of a local maximum can be guaranteed. Therefore, it is recommended to repeat the EM algorithm with different initializations and choose as final solution the one with the maximum likelihood. Different initialization strategies for the EM algorithm have been proposed, as its convergence to the optimal solution depends on the initialization \citep{mixtures:Biernacki+Celeux+Govaert:2003,mixtures:Karlis+Xekalaki:2003}. Proposed strategies are for example to first make several runs of the SEM or CEM algorithm with different random initializations and then start the EM at the best solution encountered. The component specific parameter estimates can be determined separately for each $d=1,\ldots,D$. For simplicity of presentation the following description assumes $D=1$. If all parameter estimates vary between the component distributions they can be determined separately for each component in the M-step. However, if also constant or nested varying parameters are specified, the component specific estimation problems are not independent from each other any more. Parameters have to be estimated which occur in several or all components and hence, the parameters of the different components have to be determined simultaneously for all components. The estimation problem for all component specific parameters is then obtained by replicating the vector of observations $y = (y_n)_{n=1,\ldots,N}$ $K$ times and defining the covariate matrix $X = (X_{\textrm{constant}}, X_{\textrm{nested}}, X_{\textrm{varying}})$ by \begin{align*} &X_{\textrm{constant}} = \mathbf{1}_K \otimes (x'_{1,n})_{n=1,\ldots,N}\\ &X_{\textrm{nested}} = \mathbf{J} \odot (x'_{2,n})_{n=1,\ldots,N}\\ &X_{\textrm{varying}} = \mathbf{I}_K \otimes(x'_{3,n})_{n=1,\ldots,N}, \end{align*} where $\mathbf{1}_K$ is a vector of 1s of length $K$, $\mathbf{J}$ is the incidence matrix for each component $k=1,\ldots,K$ and each nesting group $c \in C$ and hence is of dimension $K \times |C|$, and $\mathbf{I}_K$ is the identity matrix of dimension $K \times K$. $\otimes$ denotes the Kronecker product and $\odot$ the Khatri-Rao product (i.e., the column-wise Kronecker product). $x_{m,n}$ are the covariates of the corresponding coefficients $\beta_{m,.}$ for $m=1,2,3$. Please note that the weights used for the estimation are the a-posteriori probabilities which are stacked for all components, i.e.~a vector of length $N K$ is obtained. Due to the replication of data in the case of constant or nested varying parameters the amount of memory needed for fitting the mixture model to large datasets is substantially increased and it might be easier to fit only varying coefficients to these datasets. To overcome this problem it could be considered to implement special data structures in order to avoid storing the same data multiple times for large datasets. Before each M-step the average component sizes (over the given data points) are checked and components which are smaller than a given (relative) minimum size are omitted in order to avoid too small components where fitting problems might arise. This strategy has already been recommended for the SEM algorithm \citep{mixtures:Celeux+Diebolt:1988} because it allows to determine the suitable number of components in an automatic way given that the a-priori specified number of components is large enough. This recommendation is based on the assumption that the redundent components will be omitted during the estimation process if the algorithm is started with too many components. If omission of small components is not desired the minimum size required can be set to zero. All components will be then retained throughout the EM algorithm and a mixture with the number of components specified in the initialization will be returned. The algorithm is stopped if the relative change in the log-likelihood is smaller than a pre-specified $\epsilon$ or the maximum number of iterations is reached. For model selection different information criteria are available: AIC, BIC and ICL \citep[Integrated Complete Likelihood;][]{mixtures:Biernacki+Celeux+Govaert:2000}. They are of the form twice the negative loglikelihood plus number of parameters times $k$ where $k=2$ for the AIC and $k$ equals the logarithm of the number of observations for the BIC. The ICL is the same as the BIC except that the complete likelihood (where the missing class memberships are replaced by the assignments induced by the maximum a-posteriori probabilities) instead of the likelihood is used. %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \section{Using the new functionality} \label{sec:using-new-funct} In the following model fitting and model selection in \proglang{R} is illustrated on several examples including mixtures of Gaussian, binomial and Poisson regression models, see also \cite{mixtures:Gruen:2006} and \cite{mixtures:Gruen+Leisch:2007a}. More examples for mixtures of GLMs are provided as part of the software package through a collection of artificial and real world datasets, most of which have been previously used in the literature (see references in the online help pages). Each dataset can be loaded to \proglang{R} with \code{data("}\textit{name}\code{")} and the fitting of the proposed models can be replayed using \code{example("}\textit{name}\code{")}. Further details on these examples are given in a user guide which can be accessed using \code{vignette("regression-examples", package="flexmix")} from within \proglang{R}. %%----------------------------------------------------------------------- \subsection{Artificial example}\label{sec:artificial-example} In the following the artificial dataset \code{NPreg} is used which has already been used in \cite{mixtures:Leisch:2004} to illustrate the application of package \pkg{flexmix}. The data comes from two latent classes of size \Sexpr{nrow(NPreg)/2} each and for each of the classes the data is drawn with respect to the following structure: \begin{center} \begin{tabular}{ll} Class~1: & $ \mathit{yn} = 5x+\epsilon$\\ Class~2: & $ \mathit{yn} = 15+10x-x^2+\epsilon$ \end{tabular} \end{center} with $\epsilon\sim N(0,9)$, see the left panel of Figure~\ref{fig:npreg}. The dataset \code{NPreg} also includes a response $\mathit{yp}$ which is given by a generalized linear model following a Poisson distribution and using the logarithm as link function. The parameters of the mean are given for the two classes by: \begin{center} \begin{tabular}{ll} Class~1: & $ \mu_1 = 2 - 0.2x$\\ Class~2: & $ \mu_2 = 1 + 0.1x$. \end{tabular} \end{center} This signifies that given $x$ the response $\mathit{yp}$ in group $k$ follows a Poisson distribution with mean $e^{\mu_k}$, see the right panel of Figure~\ref{fig:npreg}. \setkeys{Gin}{width=\textwidth} \begin{figure} \centering <>= par(mfrow=c(1,2)) plot(yn~x, col=class, pch=class, data=NPreg) plot(yp~x, col=class, pch=class, data=NPreg) @ \caption{Standard regression example (left) and Poisson regression (right).} \label{fig:npreg} \end{figure} This model can be fitted in \proglang{R} using the commands: <<>>= set.seed(1802) library("flexmix") data("NPreg", package = "flexmix") Model_n <- FLXMRglm(yn ~ . + I(x^2)) Model_p <- FLXMRglm(yp ~ ., family = "poisson") m1 <- flexmix(. ~ x, data = NPreg, k = 2, model = list(Model_n, Model_p), control = list(verbose = 10)) @ If the dimensions are independent the component specific model for multivariate observations can be specified as a list of models for each dimension. The estimation can be controlled with the \code{control} argument which is specified with an object of class \code{"FLXcontrol"}. For convenience also a named list can be provided which is used to construct and set the respective slots of the \code{"FLXcontrol"} object. Elements of the control object are \code{classify} to select ordinary EM, CEM or SEM, \code{minprior} for the minimum relative size of components, \code{iter.max} for the maximum number of iterations and \code{verbose} for monitoring. If \code{verbose} is a positive integer the log-likelihood is reported every \code{verbose} iterations and at convergence together with the number of iterations made. The default is to not report any log-likelihood information during the fitting process. The estimated model \code{m1} is of class \code{"flexmix"} and the result of the default plot method for this class is given in Figure~\ref{fig:root1}. This plot method uses package \pkg{lattice} \citep{mixtures:Sarkar:2008} and the usual parameters can be specified to alter the plot, e.g.~the argument \code{layout} determines the arrangement of the panels. The returned object is of class \code{"trellis"} and the plotting can also be influenced by the arguments of its show method. The default plot prints rootograms (i.e., a histogram of the square root of counts) of the a-posteriori probabilities of each observation separately for each component. For each component the observations with a-posteriori probabilities less than a pre-specified $\epsilon$ (default is $10^{-4}$) for this component are omitted in order to avoid that the bar at zero dominates the plot \citep{mixtures:Leisch:2004a}. Please note that the labels of the y-axis report the number of observations in each bar, i.e.~the squared values used for the rootograms. \begin{figure} \centering <>= print(plot(m1)) @ \caption{The plot method for \code{"flexmix"} objects, here obtained by \code{plot(m1)}, shows rootograms of the posterior class probabilities.} \label{fig:root1} \end{figure} More detailed information on the estimated parameters with respect to standard deviations and significance tests can be obtained with function \code{refit()}. This function determines the variance-covariance matrix of the estimated parameters by using the inverted negative Hesse matrix as computed by the general purpose optimizer \code{optim()} on the full likelihood of the model. \code{optim()} is initialized in the solution obtained with the EM algorithm. For mixtures of GLMs we also implemented the gradient, which speeds up convergence and gives more precise estimates of the Hessian. Naturally, function \code{refit()} will also work for models which have been determined by applying some model selection strategy depending on the data (AIC, BIC, \ldots). The same caution is necessary as when using \code{summary()} on standard linear models selected using \code{step()}: The p-values shown are not correct because they have not been adjusted for the fact that the same data are used to select the model and compute the p-values. So use them only in an exploratory manner in this context, see also \cite{mixtures:Harrell:2001} for more details on the general problem. The returned object can be inspected using \code{summary()} with arguments \code{which} to specify if information for the component model or the concomitant variable model should be shown and \code{model} to indicate for which dimension of the component models this should be done. Selecting \code{model=1} gives the parameter estimates for the dimension where the response variable follows a Gaussian distribution. <<>>= m1.refit <- refit(m1) summary(m1.refit, which = "model", model = 1) @ \begin{figure} \centering <>= print(plot(m1.refit, layout = c(1,3), bycluster = FALSE, main = expression(paste(yn *tilde(" ")* x + x^2))), split= c(1,1,2,1), more = TRUE) print(plot(m1.refit, model = 2, main = expression(paste(yp *tilde(" ")* x)), layout = c(1,2), bycluster = FALSE), split = c(2,1,2,1)) @ \caption{The default plot for refitted \code{"flexmix"} objects, here obtained by \code{plot(refit(m1), model = 1)} and \code{plot(refit(m1), model = 2)}, shows the coefficient estimates and their confidence intervals.} \label{fig:refit} \end{figure} The default plot method for the refitted \code{"flexmix"} object depicts the estimated coefficients with corresponding confidence intervals and is given in Figure~\ref{fig:refit}. It can be seen that for the first model the confidence intervals of the coefficients of the intercept and the quadratic term of \code{x} overlap with zero. A model where these coefficients are set to zero can be estimated with the model driver function \code{FLXMRglmfix()} and the following commands for specifying the nesting structure. The argument \code{nested} needs input for the number of components in each group (given by \code{k}) and the formula which determines the model matrix for the nesting (given by \code{formula}). This information can be provided in a named list. For the restricted model the element \code{k} is a vector with two 1s because each of the components has different parameters. The formulas specifying the model matrices of these coefficients are \verb/~ 1 + I(x^2)/ for an intercept and a quadratic term of $x$ for component 1 and \code{~ 0} for no additional coefficients for component 2. The EM algorithm is initialized in the previously fitted model by passing the posterior probabilities in the argument \code{cluster}. <<>>= Model_n2 <- FLXMRglmfix(yn ~ . + 0, nested = list(k = c(1, 1), formula = c(~ 1 + I(x^2), ~ 0))) m2 <- flexmix(. ~ x, data = NPreg, cluster = posterior(m1), model = list(Model_n2, Model_p)) m2 @ Model selection based on the BIC would suggest the smaller model which also corresponds to the true underlying model. <<>>= c(BIC(m1), BIC(m2)) @ %%----------------------------------------------------------------------- \subsection{Beta-blockers dataset} \label{sec:beta-blockers} The dataset is analyzed in \cite{mixtures:Aitkin:1999, mixtures:Aitkin:1999a} using a finite mixture of binomial regression models. Furthermore, it is described in \citet[p.~165]{mixtures:McLachlan+Peel:2000}. The dataset is from a 22-center clinical trial of beta-blockers for reducing mortality after myocardial infarction. A two-level model is assumed to represent the data, where centers are at the upper level and patients at the lower level. The data is illustrated in Figure~\ref{fig:beta}. First, the center information is ignored and a binomial logit regression model with treatment as covariate is fitted using \code{glm}, i.e.~$K=1$ and it is assumed that the different centers are comparable: <<>>= data("betablocker", package = "flexmix") betaGlm <- glm(cbind(Deaths, Total - Deaths) ~ Treatment, family = "binomial", data = betablocker) betaGlm @ The residual deviance suggests that overdispersion is present in the data. In the next step the intercept is allowed to follow a mixture distribution given the centers. This signifies that the component membership is fixed for each center. This grouping is specified in \proglang{R} by adding \code{| Center} to the formula similar to the notation used in \pkg{nlme} \citep{mixtures:Pinheiro+Bates:2000}. Under the assumption of homogeneity within centers identifiability of the model class can be ensured as induced by the sufficient conditions for identifability given in \cite{mixtures:Follmann+Lambert:1991} for binomial logit models with varying intercepts and \cite{mixtures:Gruen+Leisch:2008} for multinomial logit models with varying and constant parameters. In order to determine the suitable number of components, the mixture is fitted with different numbers of components. <<>>= betaMixFix <- stepFlexmix(cbind(Deaths, Total - Deaths) ~ 1 | Center, model = FLXMRglmfix(family = "binomial", fixed = ~ Treatment), k = 2:4, nrep = 5, data = betablocker) @ The returned object is of class \code{"stepFlexmix"} and printing the object gives the information on the number of iterations until termination of the EM algorithm, a logical indicating if the EM algorithm has converged, the log-likelihood and some model information criteria. The plot method compares the fitted models using the different model information criteria. <<>>= betaMixFix @ A specific \code{"flexmix"} model contained in the \code{"stepFlexmix"} object can be selected using \code{getModel()} with argument \code{which} to specify the selection criterion. The best model with respect to the BIC is selected with: <<>>= betaMixFix_3 <- getModel(betaMixFix, which = "BIC") betaMixFix_3 <- relabel(betaMixFix_3, "model", "Intercept") @ The components of the selected model are ordered with respect to the estimated intercept values. In this case a model with three components is selected with respect to the BIC. The fitted values for the model with three components are given in Figure~\ref{fig:beta} separately for each component and the treatment and control groups. The fitted parameters of the component specific models can be accessed with: <<>>= parameters(betaMixFix_3) @ Please note that the coefficients of variable \code{Treatment} are the same for all three components. \begin{figure} \centering <>= library("grid") betablocker$Center <- with(betablocker, factor(Center, levels = Center[order((Deaths/Total)[1:22])])) clusters <- factor(clusters(betaMixFix_3), labels = paste("Cluster", 1:3)) print(dotplot(Deaths/Total ~ Center | clusters, groups = Treatment, as.table = TRUE, data = betablocker, xlab = "Center", layout = c(3, 1), scales = list(x = list(cex = 0.7, tck = c(1, 0))), key = simpleKey(levels(betablocker$Treatment), lines = TRUE, corner = c(1,0)))) betaMixFix.fitted <- fitted(betaMixFix_3) for (i in 1:3) { seekViewport(trellis.vpname("panel", i, 1)) grid.lines(unit(1:22, "native"), unit(betaMixFix.fitted[1:22, i], "native"), gp = gpar(lty = 1)) grid.lines(unit(1:22, "native"), unit(betaMixFix.fitted[23:44, i], "native"), gp = gpar(lty = 2)) } @ \setkeys{Gin}{width=0.8\textwidth} \caption{Relative number of deaths for the treatment and the control group for each center in the beta-blocker dataset. The centers are sorted by the relative number of deaths in the control group. The lines indicate the fitted values for each component of the 3-component mixture model with varying intercept and constant parameters for treatment.} \label{fig:beta} \end{figure} The variable \code{Treatment} can also be included in the varying part of the model. This signifies that a mixture distribution is assumed where for each component different values are allowed for the intercept and the treatment coefficient. This mixture distribution can be specified using function \code{FLXMRglm()}. Again it is assumed that the heterogeneity is only between centers and therefore the aggregated data for each center can be used. <<>>= betaMix <- stepFlexmix(cbind(Deaths, Total - Deaths) ~ Treatment | Center, model = FLXMRglm(family = "binomial"), k = 3, nrep = 5, data = betablocker) betaMix <- relabel(betaMix, "model", "Treatment") parameters(betaMix) c(BIC(betaMixFix_3), BIC(betaMix)) @ The difference between model \code{betaMix} and \code{betaMixFix\_3} is that the treatment coefficients are the same for all three components for \code{betaMixFix\_3} while they have different values for \code{betaMix} which can easily be seen when comparing the fitted component specific parameters. The larger model \code{betaMix} which also allows varying parameters for treatment has a higher BIC and therefore the smaller model \code{betaMixFix\_3} would be preferred. The default plot for \code{"flexmix"} objects gives a rootogram of the posterior probabilities for each component. Argument \code{mark} can be used to inspect with which components the specified component overlaps as all observations are coloured in the different panels which are assigned to this component based on the maximum a-posteriori probabilities. \begin{figure} \centering <>= print(plot(betaMixFix_3, nint = 10, mark = 1, col = "grey", layout = c(3, 1))) @ \caption{Default plot of \code{"flexmix"} objects where the observations assigned to the first component are marked.}\label{fig:default} \end{figure} \begin{figure} \centering <>= print(plot(betaMixFix_3, nint = 10, mark = 2, col = "grey", layout = c(3, 1))) @ \caption{Default plot of \code{"flexmix"} objects where the observations assigned to the third component are marked.}\label{fig:default-2} \end{figure} The rootogram indicates that the components are well separated. In Figure~\ref{fig:default} it can be seen that component 1 is completely separated from the other two components, while Figure~\ref{fig:default-2} shows that component 2 has a slight overlap with both other components. The cluster assignments using the maximum a-posteriori probabilities are obtained with: <<>>= table(clusters(betaMix)) @ The estimated probabilities of death for each component for the treated patients and those in the control group can be obtained with: <<>>= predict(betaMix, newdata = data.frame(Treatment = c("Control", "Treated"))) @ or by obtaining the fitted values for two observations (e.g.~rows 1 and 23) with the desired levels of the predictor \code{Treatment} <<>>= betablocker[c(1, 23), ] fitted(betaMix)[c(1, 23), ] @ A further analysis of the model is possible with function \code{refit()} which returns the estimated coefficients together with the standard deviations, z-values and corresponding p-values. Please note that the p-values are only approximate in the sense that they have not been corrected for the fact that the data has already been used to determine the specific fitted model. <<>>= summary(refit(betaMix)) @ Given the estimated treatment coefficients we now also compare this model to a model where the treatment coefficient is assumed to be the same for components 1 and 2. Such a model is specified using the model driver \code{FLXMRglmfix()}. As the first two components are assumed to have the same coeffcients for treatment and for the third component the coefficient for treatment shall be set to zero the argument \code{nested} has \code{k = c(2,1)} and \code{formula = c(\~{}Treatment, \~{})}. <<>>= ModelNested <- FLXMRglmfix(family = "binomial", nested = list(k = c(2, 1), formula = c(~ Treatment, ~ 0))) betaMixNested <- flexmix(cbind(Deaths, Total - Deaths) ~ 1 | Center, model = ModelNested, k = 3, data = betablocker, cluster = posterior(betaMix)) parameters(betaMixNested) c(BIC(betaMix), BIC(betaMixNested), BIC(betaMixFix_3)) @ The comparison of the BIC values suggests that the nested model with the same treatment effect for two components and no treatment effect for the third component is the best. %%----------------------------------------------------------------------- \subsection[Productivity of Ph.D. students in biochemistry]{Productivity of Ph.D.~students in biochemistry} \label{sec:bioChemists} <>= data("bioChemists", package = "flexmix") @ This dataset is taken from \cite{mixtures:Long:1990}. It contains \Sexpr{nrow(bioChemists)} observations from academics who obtained their Ph.D.~degree in biochemistry in the 1950s and 60s. It includes \Sexpr{sum(bioChemists$fem=="Women")} women and \Sexpr{sum(bioChemists$fem=="Men")} men. The productivity was measured by counting the number of publications in scientific journals during the three years period ending the year after the Ph.D.~was received. In addition data on the productivity and the prestige of the mentor and the Ph.D.~department was collected. Two measures of family characteristics were recorded: marriage status and number of children of age 5 and lower by the year of the Ph.D. First, mixtures with one, two and three components and only varying parameters are fitted, and the model minimizing the BIC is selected. This is based on the assumption that unobserved heterogeneity is present in the data due to latent differences between the students in order to be productive and achieve publications. Starting with the most general model to determine the number of components using information criteria and checking for possible model restrictions after having the number of components fixed is a common strategy in finite mixture modelling \citep[see][]{mixtures:Wang+Puterman+Cockburn:1996}. Function \code{refit()} is used to determine confidence intervals for the parameters in order to choose suitable alternative models. However, it has to be noted that in the course of the procedure these confidence intervals will not be correct any more because the specific fitted models have already been determined using the same data. <<>>= data("bioChemists", package = "flexmix") Model1 <- FLXMRglm(family = "poisson") ff_1 <- stepFlexmix(art ~ ., data = bioChemists, k = 1:3, model = Model1) ff_1 <- getModel(ff_1, "BIC") @ The selected model has \Sexpr{ff_1@k} components. The estimated coefficients of the components are given in Figure~\ref{fig:coefficients-1} together with the corresponding 95\% confidence intervals using the plot method for objects returned by \code{refit()}. The plot shows that the confidence intervals of the parameters for \code{kid5}, \code{mar}, \code{ment} and \code{phd} overlap for the two components. In a next step a mixture with two components is therefore fitted where only a varying intercept and a varying coefficient for \code{fem} is specified and all other coefficients are constant. The EM algorithm is initialized with the fitted mixture model using \code{posterior()}. \begin{figure} \centering <>= print(plot(refit(ff_1), bycluster = FALSE, scales = list(x = list(relation = "free")))) @ \caption{Coefficient estimates and confidence intervals for the model with only varying parameters.}\label{fig:coefficients-1} \end{figure} <<>>= Model2 <- FLXMRglmfix(family = "poisson", fixed = ~ kid5 + mar + ment) ff_2 <- flexmix(art ~ fem + phd, data = bioChemists, cluster = posterior(ff_1), model = Model2) c(BIC(ff_1), BIC(ff_2)) @ If the BIC is used for model comparison the smaller model including only varying coefficients for the intercept and \code{fem} is preferred. The coefficients of the fitted model can be obtained using \code{refit()}: <<>>= summary(refit(ff_2)) @ It can be seen that the coefficient of \code{phd} does for both components not differ significantly from zero and might be omitted. This again improves the BIC. <<>>= Model3 <- FLXMRglmfix(family = "poisson", fixed = ~ kid5 + mar + ment) ff_3 <- flexmix(art ~ fem, data = bioChemists, cluster = posterior(ff_2), model = Model3) c(BIC(ff_2), BIC(ff_3)) @ The coefficients of the restricted model without \code{phd} are given in Figure~\ref{fig:coefficients-2}. \begin{figure}[t] \centering <>= print(plot(refit(ff_3), bycluster = FALSE, scales = list(x = list(relation = "free")))) @ \caption{Coefficient estimates and confidence intervals for the model with varying and constant parameters where the variable \code{phd} is not used in the regression.}\label{fig:coefficients-2} \end{figure} An alternative model would be to assume that gender does not directly influence the number of articles but has an impact on the segment sizes. <<>>= Model4 <- FLXMRglmfix(family = "poisson", fixed = ~ kid5 + mar + ment) ff_4 <- flexmix(art ~ 1, data = bioChemists, cluster = posterior(ff_2), concomitant = FLXPmultinom(~ fem), model = Model4) parameters(ff_4) summary(refit(ff_4), which = "concomitant") BIC(ff_4) @ This suggests that the proportion of women is lower in the second component which is the more productive segment. The alternative modelling strategy where homogeneity is assumed at the beginning and a varying interept is added if overdispersion is observed leads to the following model which is the best with respect to the BIC. <<>>= Model5 <- FLXMRglmfix(family = "poisson", fixed = ~ kid5 + ment + fem) ff_5 <- flexmix(art ~ 1, data = bioChemists, cluster = posterior(ff_2), model = Model5) BIC(ff_5) @ \begin{figure} \centering \setkeys{Gin}{width=0.8\textwidth} <>= pp <- predict(ff_5, newdata = data.frame(kid5 = 0, mar = factor("Married", levels = c("Single", "Married")), fem = c("Men", "Women"), ment = mean(bioChemists$ment))) matplot(0:12, sapply(unlist(pp), function(x) dpois(0:12, x)), type = "b", lty = 1, xlab = "Number of articles", ylab = "Probability") legend("topright", paste("Comp.", rep(1:2, each = 2), ":", c("Men", "Women")), lty = 1, col = 1:4, pch = paste(1:4), bty = "n") @ \caption{The estimated productivity for each compoment for men and women.} \label{fig:estimated} \end{figure} \setkeys{Gin}{width=0.98\textwidth} In Figure~\ref{fig:estimated} the estimated distribution of productivity for model \code{ff\_5} are given separately for men and women as well as for each component where for all other variables the mean values are used for the numeric variables and the most frequent category for the categorical variables. The two components differ in that component 1 contains the students who publish no article or only a single article, while the students in component 2 write on average several articles. With a constant coefficient for gender women publish less articles than men in both components. This example shows that different optimal models are chosen for different modelling procedures. However, the distributions induced by the different variants of the model class may be similar and therefore it is not suprising that they then will have similar BIC values. %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \section{Implementation}\label{sec:implementation} The new features extend the available model class described in \cite{mixtures:Leisch:2004} by providing infrastructure for concomitant variable models and for fitting mixtures of GLMs with varying and constant parameters for the component specific parameters. The implementation of the extensions of the model class made it necessary to define a better class structure for the component specific models and to modify the fit functions \code{flexmix()} and \code{FLXfit()}. An overview on the \proglang{S}4 class structure of the package is given in Figure~\ref{fig:class structure}. There is a class for unfitted finite mixture distributions given by \code{"FLXdist"} which contains a list of \code{"FLXM"} objects which determine the component specific models, a list of \code{"FLXcomponent"} objects which specify functions to determine the component specific log-likelihoods and predictions and which contain the component specific parameters, and an object of class \code{"FLXP"} which specifies the concomitant variable model. Class \code{"flexmix"} extends \code{"FLXdist"}. It represents a fitted finite mixture distribution and it contains the information about the fitting with the EM algorithm in the object of class \code{"FLXcontrol"}. Repeated fitting with the EM algorithm with different number of components is provided by function \code{stepFlexmix()} which returns an object of class \code{"stepFlexmix"}. Objects of class \code{"stepFlexmix"} contain the list of the fitted mixture models for each number of components in the slot \code{"models"}. \setkeys{Gin}{width=.9\textwidth} \begin{figure}[t] \centering \includegraphics{flexmix} \caption{UML class diagram \citep[see][]{mixtures:Fowler:2004} of the \pkg{flexmix} package.} \label{fig:class structure} \end{figure} \setkeys{Gin}{width=\textwidth} For the component specific model a virtual class \code{"FLXM"} is introduced which (currently) has two subclasses: \code{"FLXMC"} for model-based clustering and \code{"FLXMR"} for clusterwise regression, where predictor variables are given. Additional slots have been introduced to allow for data preprocessing and the construction of the components was separated from the fit and is implemented using lexical scoping \citep{mixtures:Gentleman+Ihaka:2000} in the slot \code{defineComponent}. \code{"FLXMC"} has an additional slot \code{dist} to specify the name of the distribution of the variable. In the future functionality shall be provided for sampling from a fitted or unfitted finite mixture. Using this slot observations can be generated by using the function which results from adding an \code{r} at the beginnning of the distribution name. This allows to only implement the (missing) random number generator functions and otherwise use the same method for sampling from mixtures with component specific models of class \code{"FLXMC"}. For \code{flexmix()} and \code{FLXfit()} code blocks which are model dependent have been identified and different methods implemented. Finite mixtures of regressions with varying, nested and constant parameters were a suitable model class for this identification task as they are different from models previously implemented. The main differences are: \begin{itemize} \item The number of components is related to the component specific model and the omission of small components during the EM algorithm impacts on the model. \item The parameters of the component specific models can not be determined separately in the M-step and a joint model matrix is needed. \end{itemize} This makes it also necessary to have different model dependent methods for \code{fitted()} which extracts the fitted values from a \code{"flexmix"} object, \code{predict()} which predicts new values for a \code{"flexmix"} object and \code{refit()} which refits an estimated model to obtain additional information for a \code{"flexmix"} object. %%----------------------------------------------------------------------- \subsection{Component specific models with varying and constant parameters}\label{sec:comp-models-with} A new M-step driver is provided which fits finite mixtures of GLMs with constant and nested varying parameters for the coefficients and the dispersion parameters. The class \code{"FLXMRglmfix"} returned by the driver \code{FLXMRglmfix()} has the following additional slots with respect to \code{"FLXMRglm"}: \begin{description} \item[\code{design}:] An incidence matrix indicating which columns of the model matrix are used for which component, i.e.~$\mathbf{D}=(\mathbf{1}_K,\mathbf{J}, \mathbf{I}_K)$. \item[\code{nestedformula}:] An object of class \code{"FLXnested"} containing the formula for the nested regression coefficients and the number of components in each $K_c$, $c \in C$. \item[\code{fixed}:] The formula for the constant regression coefficients. \item[\code{variance}:] A logical indicating if different variances shall be estimated for the components following a Gaussian distribution or a vector specifying the nested structure for estimating these variances. \end{description} The difference between estimating finite mixtures including only varying parameters using models specified with \code{FLXMRglm()} and those with varying and constant parameters using function \code{FLXMRglmfix()} is hidden from the user, as only the specified model is different. The fitted model is also of class \code{"flexmix"} and can be analyzed using the same functions as for any model fitted using package \pkg{flexmix}. The methods used are the same except if the slot containing the model is accessed and method dispatching is made via the model class. New methods are provided for models of class \code{"FLXMRglmfix"} for functions \code{refit()}, \code{fitted()} and \code{predict()} which can be used for analyzing the fitted model. The implementation allows repeated measurements by specifying a grouping variable in the formula argument of \code{flexmix()}. Furthermore, it has to be noticed that the model matrix is determined by updating the formula of the varying parameters successively with the formula of the constant and then of the nested varying parameters. This ensures that if a mixture distribution is fitted for the intercept, the model matrix of a categorical variable includes only the remaining columns for the constant parameters to have full column rank. However, this updating scheme makes it impossible to estimate a constant intercept while allowing varying parameters for a categorical variable. For this model one big model matrix is constructed where the observations are repeated $K$ times and suitable columns of zero added. The coefficients of all $K$ components are determined simultaneously in the M-step, while if only varying parameters are specified the maximization of the likelihood is made separately for all components. For large datasets the estimation of a combination of constant and varying parameters might therefore be more challenging than only varying parameters. %% ----------------------------------------------------------------------- \subsection{Concomitant variable models}\label{sec:conc-vari-models} For representing concomitant variable models the class \code{"FLXP"} is defined. It specifies how the concomitant variable model is fitted using the concomitant variable model matrix as predictor variables and the current a-posteriori probability estimates as response variables. The object has the following slots: \begin{description} \item[\code{fit}:] A \code{function (x, y, ...)} returning the fitted values for the component weights during the EM algorithm. \item[\code{refit}:] A \code{function (x, y, ...)} used for refitting the model. \item[\code{df}:] A \code{function (x, k, ...)} returning the degrees of freedom used for estimating the concomitant variable model given the model matrix \code{x} and the number of components \code{k}. \item[\code{x}:] A matrix containing the model matrix of the concomitant variables. \item[\code{formula}:] The formula for determining the model matrix \code{x}. \item[\code{name}:] A character string describing the model, which is only used for print output. \end{description} Two constructor functions for concomitant variable models are provided at the moment. \code{FLXPconstant()} is for constant component weights without concomitant variables and for multinomial logit models \code{FLXPmultinom()} can be used. \code{FLXPmultinom()} has its own class \code{"FLXPmultinom"} which extends \code{"FLXP"} and has an additional slot \code{coef} for the fitted coefficients. The multinomial logit models are fitted using package \pkg{nnet} \citep{mixtures:Venables+Ripley:2002}. %%----------------------------------------------------------------------- \subsection{Further changes} The estimation of the model with the EM algorithm was improved by adapting the variants to correspond to the CEM and SEM variants as outlined in the literature. To make this more explicit it is now also possible to use \code{"CEM"} or \code{"SEM"} to specify an EM variant in the \code{classify} argument of the \code{"FLXcontrol"} object. Even though the SEM algorithm can in general not be expected to converge the fitting procedure is also terminated for the SEM algorithm if the change in the relative log-likelhood is smaller than the pre-specified threshold. This is motivated by the fact that for well separated clusters the posteriors might converge to an indicator function with all weight concentrated in one component. The fitted model with the maximum likelihood encountered during the SEM algorithm is returned. For discrete data in general multiple observations with the same values are given in a dataset. A \code{weights} argument was added to the fitting function \code{flexmix()} in order to avoid repeating these observations in the provided dataset. The specification is through a \code{formula} in order to allow selecting a column of the data frame given in the \code{data} argument. The weights argument allows to avoid replicating the same observations and hence enables more efficient memory use in these applications. This possibitliy is especially useful in the context of model-based clustering for mixtures of Poisson distributions or latent class analysis with multivariate binary observations. In order to be able to apply different initialization strategies such as for example first running several different random initializations with CEM and then switching to ordinary EM using the best solution found by CEM for initialization a \code{posterior()} function was implemented. \code{posterior()} also takes a \code{newdata} argument and hence, it is possible to apply subset strategies for large datasets as suggested in \cite{mixtures:Wehrens+Buydens+Fraley:2004}. The returned matrix of the posterior probabilities can be used to specify the \code{cluster} argument for \code{flexmix()} and the posteriors are then used as weights in the first M-step. The default plot methods now use trellis graphics as implemented in package \pkg{lattice} \citep{mixtures:Sarkar:2008}. Users familiar with the syntax of these graphics and with the plotting and printing arguments will find the application intuitive as a lot of plotting arguments are passed to functions from \pkg{lattice} as for example \code{xyplot()} and \code{histogram()}. In fact only new panel, pre-panel and group-panel functions were implemented. The returned object is of class \code{"trellis"} and the show method for this class is used to create the plot. Function \code{refit()} was modified and has now two different estimation methods: \code{"optim"} and \code{"mstep"}. The default method \code{"optim"} determines the variance-covariance matrix of the parameters from the inverse Hessian of the full log-likelihood. The general purpose optimizer \code{optim()} is used to maximize the log-likelihood and initialized in the solution obtained with the EM algorithm. For mixtures of GLMs there are also functions implemented to determine the gradient which can be used to speed up convergence. The second method \code{"mstep"} is only a raw approximation. It performs an M-step where the a-posteriori probabilities are treated as given instead of estimated and returns for the component specific models nearly complete \code{"glm"} objects which can be further analyzed. The advantage of this method is that the return value is basically a list of standard \code{"glm"} objects, such that the regular methods for this class can be used. %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \section{Writing your own drivers}\label{sec:writing-your-own} Two examples are given in the following to demonstrate how new drivers can be provided for concomitant variable models and for component specific models. Easy extensibility is one of the main implementation aims of the package and it can be seen that writing new drivers requires only a few lines of code for providing the constructor functions which include the fit functions. %%----------------------------------------------------------------------- \subsection{Component specific models: Zero-inflated models}\label{sec:component-models} \lstset{frame=trbl,basicstyle=\small\tt,stepnumber=5,numbers=left} In Poisson or binomial regression models it can be often encountered that the observed number of zeros is higher than expected. A mixture with two components where one has mean zero can be used to model such data. These models are also referred to as zero-inflated models \citep[see for example][]{mixtures:Boehning+Dietz+Schlattmann:1999}. A generalization of this model class would be to fit mixtures with more than two components where one component has a mean fixed at zero. So this model class is a special case of a mixture of generalized linear models where (a) the family is restricted to Poisson and binomial and (b) the parameters of one component are fixed. For simplicity the implementation assumes that the component with mean zero is the first component. In addition we assume that the model matrix contains an intercept and to have the first component absorbing the access zeros the coefficient of the intercept is set to $-\infty$ and all other coefficients are set to zero. Hence, to implement this model using package \pkg{flexmix} an appropriate model class is needed with a corresponding convenience function for construction. During the fitting of the EM algorithm using \code{flexmix()} different methods for this model class are needed when determining the model matrix (to check the presence of an intercept), to check the model after a component is removed and for the M-step to account for the fact that the coefficients of the first component are fixed. For all other methods those available for \code{"FLXMRglm"} can be re-used. The code is given in Figure~\ref{fig:ziglm.R}. \begin{figure} \centering \begin{minipage}{0.98\textwidth} \lstinputlisting{ziglm.R} \end{minipage} \caption{Driver for a zero-inflated component specific model.} \label{fig:ziglm.R} \end{figure} The model class \code{"FLXMRziglm"} is defined as extending \code{"FLXMRglm"} in order to be able to inherit methods from this model class. For construction of a \code{"FLXMRziglm"} class the convenicence function \code{FLXMRziglm()} is used which calls \code{FLXMRglm()}. The only differences are that the family is restricted to binomial or Poisson, that a different name is assigned and that an object of the correct class is returned. The presence of the intercept in the model matrix is checked in \code{FLXgetModelmatrix()} after using the method available for \code{"FLXMRglm"} models as indicated by the call to \code{callNextMethod()}. During the EM algorithm \code{FLXremoveComponent()} is called if one component is removed. For this model class it checks if the first component has been removed and if this is the case the model class is changed to \code{"FLXMRglm"}. In the M-step the coefficients of the first component are fixed and not estimated, while for the remaining components the M-step of \code{"FLXMRglm"} objects can be used. During the EM algorithm \code{FLXmstep()} is called to perform the M-step and returns a list of \code{"FLXcomponent"} objects with the fitted parameters. A new method for this function is needed for \code{"FLXMRziglm"} objects in order to account for the fixed coefficients in the first component, i.e.~for the first component the \code{"FLXcomponent"} object is constructed and concatenated with the list of \code{"FLXcomponent"} objects returned by using the \code{FLXmstep()} method for \code{"FLXMRglm"} models for the remaining components. Similar modifications are necessary in order to be able to use \code{refit()} for this model class. The code for implementing the \code{refit()} method using \code{optim()} for \code{"FLXMRziglm"} is not shown, but can be inspected in the source code of the package. \subsubsection{Example: Using the driver} This new M-step driver can be used to estimate a zero-inflated Poisson model to the data given in \cite{mixtures:Boehning+Dietz+Schlattmann:1999}. The dataset \code{dmft} consists of count data from a dental epidemiological study for evaluation of various programs for reducing caries collected among school children from an urban area of Belo Horizonte (Brazil). The variables included are the number of decayed, missing or filled teeth (DMFT index) at the beginning and at the end of the observation period, the gender, the ethnic background and the specific treatment for \Sexpr{nrow(dmft)} children. The model can be fitted with the new driver function using the following commands: <<>>= data("dmft", package = "flexmix") Model <- FLXMRziglm(family = "poisson") Fitted <- flexmix(End ~ log(Begin + 0.5) + Gender + Ethnic + Treatment, model = Model, k = 2 , data = dmft, control = list(minprior = 0.01)) summary(refit(Fitted)) @ Please note that \cite{mixtures:Boehning+Dietz+Schlattmann:1999} added the predictor \code{log(Begin + 0.5)} to serve as an offset in order to be able to analyse the improvement in the DMFT index from the beginning to the end of the study. The linear predictor with the offset subtracted is intended to be an estimate for $\log(\mathbb{E}(\textrm{End})) - \log(\mathbb{E}(\textrm{Begin}))$. This is justified by the fact that for a Poisson distributed variable $Y$ with mean between 1 and 10 it holds that $\mathbb{E}(\log(Y + 0.5))$ is approximately equal to $\log(\mathbb{E}(Y))$. $\log(\textrm{Begin} + 0.5)$ can therefore be seen as an estimate for $\log(\mathbb{E}(\textrm{Begin}))$. The estimated coefficients with corresponding confidence intervals are given in Figure~\ref{fig:dmft}. As the coefficients of the first component are restricted a-priori to minus infinity for the intercept and to zero for the other variables, they are of no interest and only the second component is plotted. The box ratio can be modified as for \code{barchart()} in package \pkg{lattice}. The code to produce this plot is given by: <>= print(plot(refit(Fitted), components = 2, box.ratio = 3)) @ \begin{figure} \centering \setkeys{Gin}{width=0.9\textwidth} <>= <> @ \caption{The estimated coefficients of the zero-inflated model for the \code{dmft} dataset. The first component is not plotted as this component captures the inflated zeros and its coefficients are fixed a-priori.} \label{fig:dmft} \end{figure} %%----------------------------------------------------------------------- \subsection{Concomitant variable models}\label{sec:concomitant-models} If the concomitant variable is a categorical variable, the multinomial logit model is equivalent to a model where the component weights for each level of the concomitant variable are determined by the mean values of the a-posteriori probabilities. The driver which implements this \code{"FLXP"} model is given in Figure~\ref{fig:myConcomitant.R}. A name for the driver has to be specified and a \code{fit()} function. In the \code{fit()} function the mean posterior probability for all observations with the same covariate points is determined, assigned to the corresponding observations and the full new a-posteriori probability matrix returned. By contrast \code{refit()} only returns the new a-posteriori probability matrix for the number of unique covariate points. \lstset{frame=trbl,basicstyle=\small\tt,stepnumber=5,numbers=left} \begin{figure} \centering \begin{minipage}{0.98\textwidth} \lstinputlisting{myConcomitant.R} \end{minipage} \caption{Driver for a concomitant variable model where the component weights are determined by averaging over the a-posteriori probabilities for each level of the concomitant variable.} \label{fig:myConcomitant.R} \end{figure} \subsubsection{Example: Using the driver} If the concomitant variable model returned by \code{myConcomitant()} is used for the artificial example in Section~\ref{sec:using-new-funct} the same fitted model is returned as if a multinomial logit model is specified. An advantage is that in this case no problems occur if the fitted probabilities are close to zero or one. <>= Concomitant <- FLXPmultinom(~ yb) MyConcomitant <- myConcomitant(~ yb) m2 <- flexmix(. ~ x, data = NPreg, k = 2, model = list(Model_n, Model_p), concomitant = Concomitant) m3 <- flexmix(. ~ x, data = NPreg, k = 2, model = list(Model_n, Model_p), cluster = posterior(m2), concomitant = MyConcomitant) @ <<>>= summary(m2) summary(m3) @ For comparing the estimated component weights for each value of $\mathit{yb}$ the following function can be used: <<>>= determinePrior <- function(object) { object@concomitant@fit(object@concomitant@x, posterior(object))[!duplicated(object@concomitant@x), ] } @ <<>>= determinePrior(m2) determinePrior(m3) @ Obviously the fitted values of the two models correspond to each other. %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \section{Summary and outlook}\label{sec:summary-outlook} Package \pkg{flexmix} was extended to cover finite mixtures of GLMs with (nested) varying and constant parameters. This allows for example the estimation of varying intercept models. In order to be able to characterize the components given some variables concomitant variable models can be estimated for the component weights. The implementation of these extensions have triggered some modifications in the class structure and in the fit functions \code{flexmix()} and \code{FLXfit()}. For certain steps, as e.g.~the M-step, methods which depend on the component specific models are defined in order to enable the estimation of finite mixtures of GLMs with only varying parameters and those with (nested) varying and constant parameters with the same fit function. The flexibility of this modified implementation is demonstrated by illustrating how a driver for zero-inflated models can be defined. In the future diagnostic tools based on resampling methods shall be implemented as bootstrap results can give valuable insights into the model fit \citep{mixtures:Gruen+Leisch:2004}. A function which conveniently allows to test linear hypotheses about the parameters using the variance-covariance matrix returned by \code{refit()} would be a further valuable diagnostic tool. The implementation of zero-inflated Poisson and binomial regression models are a first step towards relaxing the assumption that all component specific distributions are from the same parametric family. As mixtures with components which follow distributions from different parametric families can be useful for example to model outliers \citep{mixtures:Dasgupta+Raftery:1998,mixtures:Leisch:2008}, it is intended to also make this functionality available in \pkg{flexmix} in the future. %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \section*{Computational details} <>= SI <- sessionInfo() pkgs <- paste(sapply(c(SI$otherPkgs, SI$loadedOnly), function(x) paste("\\\\pkg{", x$Package, "} ", x$Version, sep = "")), collapse = ", ") @ All computations and graphics in this paper have been done using \proglang{R} version \Sexpr{getRversion()} with the packages \Sexpr{pkgs}. %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \section*{Acknowledgments} This research was supported by the the Austrian Science Foundation (FWF) under grants P17382 and T351. Thanks also to Achim Zeileis for helpful discussions on implementation details and an anonymous referee for asking a good question about parameter significance which initiated the new version of function \code{refit()}. %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \bibliography{mixture} %%----------------------------------------------------------------------- %%----------------------------------------------------------------------- \end{document} flexmix/vignettes/.install_extras0000644000176200001440000000004613425024236016774 0ustar liggesusersmyConcomitant.R$ mymclust.R$ ziglm.R$ flexmix/vignettes/flexmix.png0000644000176200001440000022746113425024236016135 0ustar liggesusers‰PNG  IHDRݲêÛsBITÛáOà pHYs   jŒw IDATxœìÝi@WÛ7ðIØqW6w©UqƒZ¨­¨Xµ¨`Å]¬Õ‚€hk«µRën¥7ŠâŠ÷ÝÞÔ­®ÕªuC±*¶¨ˆ²+*BBï‡yïy¦3É0Ùføÿ>M®9sÎ5I€äâÌIuu5‚“Š@…º €8P—‡¹Ø €±¨¬¬ÌÉÉ; 0;;»† Šü ê2ðÿ=~üØÍÍMì,ÀPÃÃÃÅÎþ×1ˆóeÀdŒ?þåË—bgâøòË/ýýýÅÎBÏP—“qüøñ‚‚±³qŒ1BìôuP­Y³fݺu; ÐÞ£G²²²Äθ .ªuïÞýرcbgÚ[¹rå¬Y³Äθ`Ý_q`¾ ˜$[[Û«W¯ŠPbbbBB‚ØYê2`’¤R©‡‡‡ØY€ÙÛÛ‹‚Áá:&q . Ôeĺ €8P—ƒ“Ëåëׯ3fŒØ‰ÔeÀ€ÈŠŒ››ÛÔ©SŸlØ0F0>>þíÛ·*Û_¿~}ìØ±J¥’\½zõ€¸S"UWW_¸p#Ÿ,]ºTå.ö‰  .Ñ®.Ó½{÷ûéÚµ+ÿç̙ÈmÞ¼™Ý277700°¬¬Œœ1cÆÔ©SkL‰rþüyŽd"""ÔÍôÑè¤ô«Wd´˜µ'‘HøôÃ?‡Ñ£G³5j÷Q#GŽdEŸQ¨2ó°°0>)}ñÅYµj»Oíê2Û¶m;uꔺ½¨Ë€(jqEÆx¬[·ÎÅÅ…ܵk×öíÛÕ²mÛ¶Ý»w3‚®®®ëÖ­ãkûöí/_¾änóâÅ Ž¡¹)•Ê={öx{{{yyMŸ>}ïÞ½OŸ>-//—Édùùù×®]Û¼yóĉƒ‚‚.^¼¨Ý@µê2 ëׯkzHÛ¶m7n\c?š^òÞ2“““CÿZ¢T*ÇŽ˨¶téÒeÇŽR©ŠÐ§výúuƌҋ/fΜɑ$ê2 0TdÓ¤I“äädöl”ÈÈÈœœvû'OžDEE1‚‰$99™ý’¡¬¬lëÖ­Üm’’’Ô]ËÉíéÓ§~øaHHÈÕ«Wkl¬P(<èëë«Å@µê2 „—/_fggÓ#‰äÕ«WÕœþþûo>ýtéÒE£düýý}||ÁåË—SëÈ|óÍ7‡¢ïmÞ¼ùÑ£GëׯÏçÔè Å¥K—ØñÙ³g¨;ªqãÆmÛ¶å8=BEFxýúõ‹ŽŽfKKKCCC«««éA¥RZZZÊh<}úô¾}ûò+11‘Ñ'£ÿÄÄD~YÿË… ¼¼¼¸—Ð>P—!ܼy“iß¾}ƒ tï§]»v 6Ô´ö”™{÷î8p€ ˆ7ÆÅÅÑwÙÚÚ9rÄÕÕ•gJŒƒ³/eºxñ"}Eö¿Í½¼¼j8}@EFDË–-óðð`OŸ>ÍøýwæÌF3OOOuK†³eee1n!G—žžÎ®€×èÞ½{C† ©ñ~sÀê2\´Xó TÒqQŽ~´»oÑ!CØßˆ–.]zòäÉÈÈHzP"‘lÛ¶½ü0GJü±™™õQ—©¬¬ £þ}maaÁ^ï1¡¡"CážµG2ĸÖÖÖÛ·o·°°`ÄçÎ{çÎrûöíÛ111Œ–––Û¶m³¶¶æ?VBB‚º]kÖ¬áß©¢¢â“O>yñâ#nff6zôè]»vegg¿yóF&“ܼy399yòäÉöööšTG . »ƒ €ŽØ+°hWzÐW?‰äÛo¿e¯]»¨P(èÁeË–i”RïÞ½éբ˗/ÓŽ‹‹£¾tñõ×_3–&P—CBEÆxxyyÅÆÆ2‚2™lìØ±r¹\.—7Ž}Ë¶ØØXM§Ô¥§§«¼Ü2++ëøñãuEÄš5kØSlºtér÷îÝ´´´Ö­[׫WÏÒÒÒÞÞ¾K—.¡¡¡›6mzþüùÞÿ}M‡«õP—ø?|Š/(Ðh‡=©D»Kuô5ï† ˆÑ£G·nÝš¬¨¨ ?üüóÏgÍš¥EJ~ø!õðíÛ·×®]#·?~¼páBjW›6mbbbØWBi7 €›B¡@EÆØÌž=ÛÏÏÌÈȈ‰‰™;wnFFc—ŸŸ_¿”(ÞÞÞä†R©Ty禵k×R³Ø«n©TVV¶lÙ2FÐËËë÷ßçw84336lØï¿ÿÎg”:u-K-(ÐðW^^~ïÞ=FpàÀNŒ{T«ëG놹¹ù7ß|ÃÑÀßß¿Æå0U¦Ä¨Ë´K™"##é÷=Y·n]^^ã.¶666;v¬1MÍŸ?c#•JSRRØ‹m­ZµŠ±Ð A 4HMMUyW8•èwqJJJ*//§ï-++KNNVÙ˜Ão¿ýVRRBXXXìÝ»·^½z<³s±¾J*T?ºö ¸}ûvUU•F‡4oÞÜÙÙ¹Æ~\\\tY³`Ò¤I .ÌÏÏgïêСþ}ûØK?ðIÉÁÁÁÏÏO*•R7x:wîÜ7ß|³oß¾_~ù…j9räÈ?þxß¾}Œ>;uêD_ž@__§)‰dãÆ£FàN<;v¬.=¼zõJ_ɉ6mÚÄÇÇOš4‰¤~{Э^½š=ËCHHÈÌ™3 ‚())IKKûüóÏ©½©©©ÔmžBBBÆ_cŸ¿þú+#2f̘víÚñÏ P—€:Çp8ÈžQ`c_éS#•ëìêñ"&’µµõøñãüñGö®¤¤¤Æרƒº”7nÜ¥Kjïï¿ÿ^ZZ:}útªY£F~úé'Žôîý÷ß¿zõ*{=£êêê1cÆ,_¾|Þ¼yAAA­Î¦¥¥®-Ôx¾|ºûì³ÏŽ9BÞN   ‰'jÔ­••Õ”)S–,YB>\»v-½.C_ xÊ”)–––|ú¼zõ*#2dȲ\Çu…`—áú&6öʸ5RY—a÷£ã:,7nT¹ëðáÃ|zàX‡¸wïÞTðÅ‹#FŒÈÍÍ¥"K–,iÞ¼9º hüøñ÷îÝ›4i’ʉ`·nÝ>|x×®]÷ïßÿ3 oãÆŽŽŽêö:99©ûeÅ-<<œšwãÆ‹/’Û§OŸÎÌÌ$·ÍÍÍÃÃÃyvXPPÀˆôìÙS‹Ä€‚º Ôr"VIP  ç|™ÜÜÜÁƒ«»&"11‘šáÏ#%Æ3'Nœ ¶}||¨oAúºó7m۶ݲe ª3FÈÎÎnõêÕêöÆÇÇk·îO‹-†J=¤æÈÐ'Ë :ÔÕÕ•g‡äUQtXHG¨Ë@ídT5£J@x …âöíÛŒà£Gª9òéGëºÌëׯüôéSu ^½zµvíZîN¸SêÝ»·Ê_þæææ6l WîÌÏÏþü9co§Nøœ€ÖP1NGŽQ·ëèÑ£ZwImïÛ·/???''çСC*€ð°¾ ˜€þù‡~Û†† ¶lÙ’Ü6òÚ•ûC<ÏÙÙ¹iÓ¦bgP'Ü»wqóé¦M›¶jÕJ¬~‚P(!!!·nÝ¢% ãûg||ü—_~iccÃ?¥&MšP«r6kÖÌÓÓóÎ;Œ£¦OŸÞ¥Kr›}TÇŽ­­­58m‘Õ™¹sç.^¼855•½î YéÒ¥‹×éܹó† téaРA/^¼Ð=csøðámÛ¶©Û›ššúé§Ÿj·’‹¿¿¿‡‡ÇÝ»w ‚Ëå7n,//§,÷ôôìÓ§ÿÞìíí?~LµhÑB‹Ä€„º ˜€ &à.÷À!88xß¾}F^‘aÀ?BABBþ; ŽXDé‡ ˆˆˆÆ-EœãããGŒAlÙ²…ãÞ±5¦ôá‡2ê2-[¶Œ¥bqÀÕ™ ¼÷Þ{ºôPã]Ò4e ‚JJJ¸ۄ……ùùùi÷_¥ÈÈHêcφ är9}—F]9880ê2W¯^E]F¸Ž LÞþýûM«(P×è«ô ¯~–,Y²iÓ&z¤~ýúGýôÓOÙÿ4^µj•B¡Ð:%Æ3A$$$Ô«W£,.¢À•Mâš6mZ^^w›¼¼¼èèhíúŸ0aBÆ ÉíÜÜ\j˜Fñ¹76{•_ú%Q ÔeÀ° W—Ñ¢„±cÇŽ˜˜zÄÌÌl÷îÝdJsæÌa´ôèÑŽ;ø§Ä85ú-™‚ b¬›ƒù2`TPÅÁƒÙ÷o×®]Û¶mÁíÛ·kW©_¿~hh(;Z¿~}º0`#’––öðáC-²ê2`@ÕÕÕ7oÞdµ(=襟sçÎ}öÙgŒ/“k×®8p ¹Ý¿ö} –/_®òû'Ÿ”;vìHnׯ_Ÿq³•ÒÒÒþù‡‘H$^^^|ÏÀ0PRqq1ûÕ‰$))iëÖ­ìYáááá%%%Z ÉèM"‘hqMwß¾}›4iB(Š#F”••i‘X_LÑ€¬¬¬ÄÎÄôêÕ«3gÎPÉõe£_ô—M»•Û@kÿýwff¦ØYÔ9ÿüóãnÓ¶¶¶:tн‚ ÜÝݹ277ýú5¹Œî½{÷† &“Éè fÍšÅXÓaΜ9Ÿ~ú)=’™™yøðaúfÕ¥dkkKUa(ýõ—ºônÞ¼ÉøNÛ¦M›FqŸ€0DY¸ŠŠŠÊÏÏg£££ÉÙvÓ¦McÔsóòò¢¢¢ØókjÔ¡C‡~ýú8q‚ŠôïßÿwÞÑ´ŸzõêÍž=ûÛo¿¥oÞ¼éëë»{÷nŽ«ªªŽ=ºbÅ ,Ê€º ˜ž­[·6oÞ\ì,@L7oÞTùR꣭‘,À¿•Ä7sæL±³¨sØ×étîÜ™¼E´ŽýðÑ¡C²(SPPÀ¾‡KHHȲeˇuèÐáþýûôàÒ¥KÙuÝO ‹Ë€ñCuÆ öïß¿sçNF°}ûöK–,!·—.]úË/¿üý÷ßô;vì1bDPP¦ÃEEEÑë2‹šs‹ŽŽÞ°aCvv6=xëÖ-OOÏaÆõìÙÓÁÁÁ¢´´477÷Ö­[çÏŸ?tèµ® Ðá:&¨…ªÿGìDþÅ8³04Ã-.ÃyMPyyy`` ã+„¯¯oJJ û;¤T*5k#xùòåÓ§Oט’¦§†ÅeÀTàÊ&>$<ÐÛM:•щT*ݺu«­­-ùÐÖÖVåÕLS§N-..Ö4Ã!C†TÓ0–»âÏÆÆæèÑ£7fÄ EZZZHHH›6mêÕ«giiiooïååºyófeÔA]j3c(…C"½.£T*ÇŒsåÊzÜÍÍíСCꮌ7nœ««+#¸téÒSB]j7Tgô+22²  €ŒŽŽöóó£G>øàöm˜òóóµXF<<<<ÈXh´ƒº Ô ÂGPŽ ‰^—ùòË/üðÛ7oöêÕKÄjÔe n1t¹å†çÏŸWÿ[=ôÒýúõ‹g ÛµkÇ=ÜŒ3ؽ1nÕ¤ã©YYYUVV2zprrÒâÉÏꌧ§gzzºðé¹ÂˆˆF¼‚ÉÆÆ†ÝÞÆÆ&))‰½|UDDDQQ‘¡²ä¡eË–çÏŸß±cûNvlfffC‡½pႉ™Ôe ŽÒoõ娃j¬Îüõ×_Ÿþ¹ð‰¹ÈÈHöb+3fÌðõõUwˆŸŸßôéÓÁ‚‚v}G`fff£FúóÏ?oܸñÓO?¿ûî»ÎÎÎÖÖÖVVV]»v8qâæÍ›óòò<ÈqŽuîÇuZµn·pB! Æ{6}}~Ó¥öL|ÄÅÅÅÅÅq4ÐýÔtéÁËËËËËkÆŒ:æPa¾ Ah8á³cjœ;*¡.ð/5”c¸¡: )ÔeTc/÷(vF¦ÕþP—ýCu€ÔeÀPPà†º YyôèQ`` Ø¹Ôe@ÎÎγgÏ; ズ €8P—ê2â@]@¨ËˆÃ\ìÀà$ #R]]­QƒºŒzrðœ€Þ¡.µ»âÀû¸.•‹Ñ£Gïܹ“9r$;Èh°{÷nvWiii<Ç]°_qh .`¼½«ªªÞ¼y#V2 ;¹\Nˆ1Fuƒ[·nÝùóçssséÁ]»vŽ;Vå!Û¶mce\]]×­[g¨, Ï$fåð)ä¡@µØ“'OèO:Õ A±’½»páÂÔ©SÅÎþëË\“&M’““Ù_ø###srrØíŸ>ž,-- =uêý»½R© ---eô0}úô¾}û ‘k]¢—ª ¦ÏˆE&“EDDˆПþ)v ‡º €@–-[vâĉ»wïÒƒ§OŸŽ‹‹›9s&‰‹‹;sæ ãXOOÏ¥K— da I.(Ь²²211Qì,t‚º Ô9b}g¶¶¶Þ¾}»··wee%=>wîܼûî»Aܾ};&&†q ¥¥å¶mÛ¬­­ —[©#vÍ 4`º,,,èëÕ«×¢E ±’Ý•””PmllDLTB]@8^^^±±±ß}÷=(“ÉÆŽ{õêU‚ Æ'“ÉGÅÆÆzyy —e­#â0(ЀÉqvv¦?ôõõ=vì˜XÉ€îV®\9kÖ,êa=DLTB]@P³gÏþïÿ{áÂz0###&&¦ºº:##ƒÑÞÏÏþ§T¥âââM›6=z4++ëåË—vvvíÛ·2dHhh¨Ÿ¬xÞ)©ªªê—_~ùå—_nܸñøñã—/_* kkëF9;;·mÛ¶cÇŽ]»võöövqqáè\Óqµf<‹ò¢@*¡. (©Tš’’Ò¥K—ׯ_Óã«V­b7nРAjjªTÊuß´ÔÔÔéÓ§¿xñ‚Š<{öìÙ³gçÎ[²dI\\\hh¨^2ÿóÏ?Çwÿþ}FüíÛ·oß¾}þüùµkר ¸Õã)ǰQ¹=~üXÜLŒ\³fÍêÕ«'v`ŒvíÚ%—ËÅÎÄA®ÿPË . ´6mÚÄÇÇOš4‰T*•ì–«W¯nݺ5GW«V­úúë¯Õí-))™8qbnn®¶™þŸ¬¬,F-ÉØs9†­U«Vb§`ÔRRRÆ/v`Œ>úè#±SÐ'®ÿÃÔJ’šÃgŸ}ÄÝ&((hâĉ ~ûí·/q"bîܹå¦ÒÂ… ¹(³ÿ~Ó*ÊP—ÇÆÕíurrÚ¸q#ÇáJ¥2<<œ1˦iÓ¦ 999r¹üÉ“'kÖ¬iÚ´©^²=qâ#2|øðk×®•••)ŠçÏŸŸ8qbΜ9nnnìc«ÿ‡cG€Z uqØÙÙ­^½ZÝÞøøxî%{9’••EØØØœ9s&22ÒÕÕÕ¢E‹QQQgΜÑË ¶‹ŠŠ‘-[¶tëÖÍÖÖÖÌÌÌÉÉ©_¿~K–,yðàÁ©S§üýýu .Àú2¢9r䈺]G ÑèØ¨¨¨N:1‚:uŠŠŠúñǵN’Ô´iÓÂÂBzäøñã#FŒ`·ô÷÷¾.¼oß>ÂÔ–˜!pu4Í£G=z$vBC]@‡Þ¶m›º½©©©Ÿ~úé!CÔ5¸té#¬²epp°îu™Þ½{“…JHHH›6mÜÝÝÛ´iãææÖ¥K—ž={Š~êJ(#/ÐàŠ-¶ ÄÆÆŠ€ÐP—:Ǿ—””„……q· óóóS·@ÌóçÏwww•-ÕÅ5òÝwß9r„q?Âìììììlê¡……EŸ>}"""†ªûˆ:2Î1¼÷À¨`}L›6-//»M^^^tt´º½ì»#Õ¯__eKuqtëÖíðáÃÍ›7çhSYYyüøñaÆMœ8Ñx ư¢°1äÆ u¡#Ý»wïÞ½{O›6±+///**ŠÝO¯^½‘¨Q]\GM›6íÙ³gHHÈòåËÏž=ËØ{ûömö!쪓R©4Dn|è½z‚r huáìß¿çÎŒ`ûöí—,YBn/]º´]»vŒ;vì`×V>ùäFdÍš5wïÞe333tJš ‚èß¿ÿ™3gÔímÒ¤ #¢òâ)öÝšrssuNMW:ÖSPŽ] . % ôöEEES§Net"•J·nÝjkkK>´µµUy5ÓÔ©S‹‹‹é‘!C†´oßžyûöí‡~¸nݺgÏžUVVæææ®]»öÃ?|ûö­î'{òäÉ>}útèÐaöìÙû÷ïøðaii©R©|õêÕÙ³g?ýôSF{•+³W¥ùî»ïž{öLìD¨žÕö IDAT´qàÀŸþùÊ•+………–––mÚ´éÝ»÷„ PѨ­P—€ZeÑ¢EdQfüøñ+W®Ô׆öâÅ‹‘#Gž8q‚Š(ŠÌÌÌÌÌÌÄÄDüo_˜LÂú¿P«üúë¯äÆâÅ‹ùeŒá†kr¹|àÀdQ&88øÊ•+2™ìåË—èÖ­›ˆ‰Ô2ÆðóÀ€ù2P«äææ’...âfÂߢE‹._¾LDXXØúõëÉ ¥¥å°aÃgÍš%jv`@˜/µŠ\.'7¤RÓøœóòåKr¡w{{ûU«V1öš™™±ƒPk˜Æç0uoÞ¼‰õôô´²²ªW¯ž¯¯ïž={¸¹~ýúgŸ}Ö¶m[KKË tíÚuáÂ…¯^½b·”Ш Òãì]ìêú'âèÑ£ï¿ÿ¾M£F†úðáCóß¿YYAcÆŒ©W¯÷sBQ(›6mêÛ·¯½½½¥¥¥ƒƒC@@ÀÖ­[«ªªTfž™™ùá‡ÚØØ´nÝzíÚµA$''wìØÑÊÊÊÍÍ-99YÝù8pàƒ>¨_¿¾¥¥¥‡‡Ç‚ Þ¼y£u>ìþõþ|jÚ?Ÿ7÷ÛÃ@xžoHH™aË–-©½¥¥¥-[¶$ã!!!ŒC´x}Gedd 6¬I“&VVVÞÞÞ{÷îÕ:REEE\\œŸŸ_Ó¦M-,,5jÔ½{÷¯¾úêÏ?ÿÔ½½¦?ï„ ïg€ÿS `ô|}}éoÚgÏž‰ˆìÆô·Dpp°áÆÂ¯JaÌŸ?Ÿþ𦤤bƤƒ„„CŒbÒ²²²èOQ@@€¾z~öìY‡ØŸC.\¨î3ɼyóTNxiݺõƒ5ýÌScuý3êA¸¸¸”””°áŸÿèÑ£É]äù|ùøø¨< ??¿/^°3wvv¦7ûâ‹/è%ÉÙ³gÙG-Y²„=D‡‹ùç#Àó©iÿ*ÓfSùBpïÕÿó-((°··'÷~þùçdpÒ¤IdÄÁÁ¡°°P]Ú<__ÆQ¿ÿþ»µµ5÷ó£ÑëUXXøî»ïòæ5mϳE€÷³V¬XAO‰\GŒ ¾l€ @]¨Ëðÿ´zºŒ‘0\]¦oß¾dŸŸ|òÉ;w***222ú÷ïOÿæFoO½X~~~—.]ª¨¨xöìYLL ôôô¬¨¨P7–¦?°|ÚSmš6mº~ýú²²²»wïvìØ‘ .Z´ˆÑ^£ü===É8ûûª:ä!¼}ûvEEÅíÛ· DÙ™OŸ>ýíÛ·ô¥j¢¢¢Þ¾};wî\òáĉÙGI¥Ò€€€ŒŒŒŠŠŠÌÌÌÀÀ@2Þ¯_?íòàùÔ¢ö<_MÛó¤éùÒ§ž¥§§§§§S÷îÝË‘6Ï×—q”§§gûöí8PZZZ^^~ñâÅ   ]òŸ2e ¹ËßßÿêÕ«ååå<ظqc¯^½Ø™hÚ^åYp´1ôûY`¨Ë?|Ç€º 0®.C𦯄ºŒ‘0P]æøñãd‡:u’ËåT¼¼¼üwÞaÿX=~üØÂ‚l_^^Nï*,,Œl¼iÓ&uÃiúsʧ=Õ&22’ îÛ· ¾ÿþûôÆšæß¸qc2øúõk> Sϧ‡‡‡L&£â2™ÌÃÃÜuúôiFæ™™™ÕÕÕýõ#rÿþ}òaçÎÙçëîîNÿ -—Ë;uêDî:uê”ùð|jÚ¿ÊÕ5б=Ú½ÿ©‹•\]]]]]Éí‘#Gr§ÍçõeÕ¬Y3ŽcZäOÝ7-77Wý£}{•g¡®¡ßÏÂC]Àøa}‚çÕæ:Pg¥¥¥‘QQQä6’µµutt4»ýúõë+++ ‚ˆ‰‰a\.1yòdrCåz ®9"‚ú·Áƒèm4Íÿõë×䆭­-Ÿ¨ç3::ÚÒÒ’Š[ZZN›6ÜNMMeE~W§_ÍDFœœœÈ‡Ô}¬è¢££­¬¬¨‡QQQŒ4´Ë‡dˆçSÓþvç»víZ‚ ž>}úôéS‚ ¸Çâóú²}ùå—Í›7×cþ%%%äÏÕX4m¯C¿ŸØP—€ºK/µhjDÞš ˆ>ø€±ë£>b·?yò$¹áïïÏØEý¿úæÍ›zÌ?êš#‚ ìììÈ—/_ÒÛhšµ&“õ‚¸rå ¹Ñ»woÆ.*B=çrEaz里P+ «\•=õ"Rih—Éϧ¦ýë‚úW§¾:$´=_;;»ÄÄDzdýúõÔ)«Ãçõe£®èQI‹ü©ys}ûö4hЊ+Ξ=[QQ¡nMÛkÄÐïg6Ôe Î1P%uÈÿÞѲeKÆ®V­Z±ÛSÿwrr277733333“J¥R©”*"Pÿ0XÆ ©m333rC¡PÐÛhšƒ È·oßòÉãù¤"TF¶æææŒuäö D ‘““£K>$C<Ÿšöol´>ßàààQ£F‘Û£G ªq,>¯/[ûöíõ›?u)«B¡HOOŸ={öG}Ô¸qãO?ý”qq¥ví5bè÷3ê2P‡S7AŒŠ1¼©rƒ c;BЮ멪ªªªªR*•J¥’1+AeA*o1àiþ...äF^^ŸÈ›jœÏ§º[kŠ=ue 5„.ùâùÔ´c£Ëùzyy‘]ºtá3Ÿ×—­~ýú}j‘ÿˆ#ÒÓÓ½½½éA™L¶oß>j $­ÛkÄÐïg6üî€ÚO¬B 4$ê[\yy9c;Bµ·³³ã^$ÏÐikMÓü©;þfffò韚tÀñ|rsæ=uµuI”¡ó1õ÷ƒ¦´>ß«W¯~ÿý÷äö¼yóþüóÏÇâóú²qÿ]Ó.ÿ?þøòåËOŸ>ݵk×_|A­ìûâÅ‹o¿ý–=Цíùòç €„º ÔZÆS1žL .øþûï,X —ËÙ»ärù‚ ¨on‚¡îóäÉÆ®G±Û·mÛ– ˆ’’êצEÓü©e8Μ9ç=5¿æñãÇŒ]T„j£#öÔ‹H½¬†ÎÇÔßšÒî|KKKGŽIÍC‘Ëå#GŽ,--å>ŠÏë«)]^/— 6º~ýºR©üõ×_‡.L’z¤iþÇŸ1cFEEEZZÚâÅ‹küçííMÞîúܹsÔBª¤sçΑô—^gÏže qþüy* aòòý`iiIÖ4år9ý^åiÚžƒ?_ÿ÷ Cc@Ýh@¥àààêZ4ƒÝ8µnÝÚ ?æÔ"ޤ””CŒ²jÕ*ú( †Å\¸p¡{÷îR©ôË/¿$OvÆŒR©´{÷î.\à8±df@@€^ò9qâÙa§När9///§î«BÐ>“Ü»w,b¾óÎ;/_¾dôVRR2nÜ8ŽáØrãÓ^]•q-òÿꫯÈ~¦L™ÂØUUU5sæLzä×_%{xxÈd2*.“ɨ¯‘'NœP—¡Fww÷ŠŠ *.—Ë;uêDî:~ü¸ùpù$33S&“eddôïߟºe ñï¯.Ô îîî{÷î-..®¨¨¸}ûvlll³fÍ­j(º´W×F]\ÓüËÊÊ:wîL2tèÐK—.•••½zõêðáÃ=zô`´W*•Ôó9hР;wîÈd²ÌÌLêîÅ àÈP£ˆT*0`@FF†L&»{÷n`` ïÛ·¯vùó|jÚ?eÊ”)dƒ=z\¾|™^³PIÓ7Oï­[·¨Åz§NJÃÃÃɈµµuFF†º´ù¼¾Zœ¬¦¯—§§ç÷ßêÔ©¼¼¼ÊÊÊW¯^=zÔÍÍì$&&FÇö…ïg¡.`üP—€º €èP—19J¥2))‰^—IJJ¢îŠ¢Žáê2¹¹¹:t`¿µ-ZDmÓÛWUU͘1Cå»ÑÞÞ~ÇŽc©ìPen|úT×"ÿüü|???>ÉTWWõ6Ÿ¢¢"Ž 5Š,\¸=„‡‡G~~¾vùó|jÚ?åÑ£GÔO ƒÊö5v¨þçûæÍê'«K—.åååd¼¼¼œº%SÇŽß¼y£2mž¯¯¦'«õëÅ6hРꤴkÏјŽ~ˆ¡ßÏC]Àø¡.&uÑ¡.cZ.]ºÔ³gO33³¯¿þš<Ù™3gš™™õìÙóÒ¥K®.S]]ýêÕ«ùóç»»»[XXØØØôêÕkÏž=Õœ_]®\¹2iÒ¤víÚÙØØX[[÷èÑcåÊ•¯^½âˆ£CFn|úäKÓü«ªªvíÚ5xð`GGG333[[[ððp•¯š\._¿~}Ÿ>}š6mjnnnggçïï¿iÓ¦ÊÊJî 5ìܹ³W¯^666666ï¾ûî?üPVV¦u>ÜÏ›¾žOíú'ååå}ûí·Ý»woÔ¨‘ºù\u¨5>ç;aÂ2úõëß¿Ÿ¾ëÞ½{Ômƒ&L˜ .mž¯oµæ'ËÿõÊÌÌ\¼xq@@€‹‹‹¥¥¥¥¥¥‹‹ËСCwïÞ­²š¬Q{u?à Œ£x? uã'©Æ¢ `ô&MšD]3 Ož<)**rrrrvv;cáïï¿råJ¸WüU©[·nb§ÀåúõëÔvëÖ­É% dÁ‚±±±ÔÔ””ñãÇë}”¸¸¸™3gR"##õ>Šè¾øâ‹Í›7{zz&%%õìÙ“ü¹¨®®¾rå ¹îï”)S6lØ ò؇R×pìØ1òcBý:ÅÅZ ¯o²råÊY³fQ©ËÜÀHà~L`’’’ÄNÁXdgg“Ó•ß¾}{âĉ¦M›Š‘q¡>_yÆT>›™™)•J±³988Ì›7oîܹô¸··÷µk×–,YRUU%VnÀ€º €)Y²dIee%A¯^½Š‹‹£¯ËtdáÃØª3¦RŽSÇñ›ÁÒÒrÁ‚æ5Šð•ýóÏ?S׬YSRR"b>ƺbi€qB]ÀdP“eHä”ó1!¢TFPŽ#÷!€1C]À40&Ë0eFSÔJPŽþP—0 ŒÉ2$L™Ñš!ª'(Ç@­sí†×À¨ .`TN–!aÊŒŽtÿlŠO· 5ÔeL€ÊÉ2$L™ÑMË+(Ç€îP—0vìÉ2  ?Ä”ýâ.¸ z„º €±cL–éÔ©Ó\]]©¦Ì½ƒr ê2F=YfÞ¼y666ß~û-=ˆ)3…r ê2F=Y&88˜ ˆÉ“'cÊ €©C]Àx©œ,#•J ‚°²²Â”S‡º €ñR7Y†„)3¦u#Å1Y†„)3¦u#Å=Y†„)3&Í\ì@…'ËÈ)3QQQTdÍš5_}õUÓ¦M…ÈÀÄ­Zµ*--Mì,ŒKEE…Ø)Ô-¨Ë#>“eH“'O^¶lÙÓ§Oɇ䔙E‹ ‘%€‰ËÎÎÎÎÎ;‹º«°°pùòåééé=zûö-ä¸-½D"©± Ñôôk“ZöRèu£Ãs² SfÀåçç{{{?yòDìDÄQÇOè°¾ €Ñá?Y†„UfÀä,Z´ˆ¬JŒ?>//¯úô8„äôا¾púÂ0æ'ÀT .`\4š,C™À䤧§“‹/vttäsˆé/Ø´8ýÚ¤6½”ºÃuLÆEÓÉ2$¬2ÀGxxøØ±cÅΨegg÷êÕK€rrrÈ †36uüôô¨¬¬¬]»vbg¢9tèØYè u#¢ÅdV™àÃÖÖÖÖÖVì,ŒÚëׯ…H.—“|~ÅÕ>uüôô¨ºº:??_ì,@4ôi›.|0"ÚM–!a•0r•Aö2%UtB¬ÅPt9}ŽÞ8â×®]>|¸……E«V­¢££¹/q}öìYlll¯^½š6mjnnîììzçÎ-ND÷—’¤P(6oÞܯ_?KKK‡þýûoÙ²¥ªªJe{]N@,¨Ë ­'Ë°Ê €îÄ­ÝèÑ¡C‡zõêµÿþââb…BñäÉ“5kÖøûûËd2•í7mÚäææ¶`Á‚?þøãÅ‹UUUÏŸ?OIIéÖ­ÛêÕ«NžT\\ìçç7eÊ”S§NVVVžêó‰ç(ÍÔí¢âÆ £Ç7oÞLÆûôéÃ8äñãÇÖÖÖA˜™™:uбW©T~÷Ýw:žˆG?~œlãáá!“ɨxEE…‡‡¹‹­§Pë­X±‚þ‹=11QìŒôŒ±2ZçÎÅÎ k„ ôWüüùóbg¤¸Ž À(è>Y†„Uf€ ˆ©S§Ò0€ÜÈÌÌd´\¿~}EEA“'Oö÷÷gì•H$‹/6Xšj¥¥¥‘ÑÑÑ–––TÜÊÊŠšð²}ûvu‡ó?}Ñ¡. >W–¡Ã*3º þs%v"ºêÞ½;ý¡““¹ñâÅ FËcÇŽ‘cÆŒ 1ž®\¹BnôîÝ›±‹ŠW9©ÄÿôD‡º €øô5Y†„)3ФIúCsóÿ¿¤ û~¢ÔÅk:u 1ž¨µÒZ¶lÉØEE¨6lüO@t¨ËˆL“eH˜2üÿ޼yó†ÜhРÁÒÑXYY¹Á^¿“ŠP™³éòg@`ø£ 2ýN–!aÊ @-¦T*õØ[ýúõÉ ÆÚ™âªW¯¹Q^^ÎØEE¨ÌLê2bÒûd¦ÌÔÔÕ7UUUôxqq±Gqss#7îܹ£ÇnuäââBn<~ü˜±‹ŠPmLê2b2Äd¦Ì@GÝÄG.—‹›‰Ö¨ERòòòèñk×®éq”€€rcÇŽšk¸'ÙÛÛ›Ü8wîc¡Ú˜4ÔeDc É2$L™€:ÎÙÙ™ÜÈÈÈà”ä “”f<<<È ê–I¤µk×êq”ððp+++‚ 6mÚtöìYvƒ~øAݱÚ=É|Œ5ŠÜX³f ½è#“ÉÈm£ºðtÿþýùóç‹€qA]@4†›,C”¨Ëú÷ïOnL:õòåË2™LÜ|´DnÌ™3çäÉ“r¹<'''""ââÅ‹z¥U«Vä_…BññÇÏŸ?ÿÞ½{r¹üåË—Ç}úØÙÙ‘Ï@Ÿ>}6lØpñâE{{{=ނРæb'P 0Y†DN™‰ŠŠ¢"kÖ¬ùꫯš6mª÷±øãóé\÷OðŽŽŽK–,Y²d ÿC„ùÚÀ;;»µkײ”Q×GÏ5êèè8oÞ<Ž«–ÔU㓬õ³jaaƳ½.§úuóæÍ… ê¬בéÝ»÷‚ úôé#pb& uA½óÎ;º.‘H¨m OOO3àCÀuLÀ…çUKgÏž5¹¢ŒD+|úáŸÃèÑ£Ù‡5Šû¨‘#G²3f ÷©………ñIé‹/¾àyâê¢XXX4nܸeË–}úô™6mÚÿû_™LÆÿÉ© P—ÕjqEÆx¬[·ÎÅÅ…ܵk×öíÛÕ²mÛ¶Ý»w3‚®®®ëÖ­ãkûöí/_¾änóâÅ Ž¡5¥P(JKKsrrΜ9“0xðàöíÛ§¥¥é«ÿZu`BEF0Mš4INNfÏF‰ŒŒÌÉÉa·òäITT#(‘H’““7nÌ=VYYÙÖ­[¹Û$%%½}û¶¦¬µ÷ôéÓ±cÇ.[¶ÌpC˜Ôeàÿ "#¼~ýúEGG3‚¥¥¥¡¡¡Œ—@©T†††–––2OŸ>½oß¾|ÆJLLT·BÙbb"¿¬usãÆ 2~¨ËA "#ªeË–yxx0‚§OŸŽ‹‹£GâââΜ9Ãhæéé¹téRžeee;vLÝÞôôô¿ÿþ›gWº¨ªª¦düP—¨ëP‘¡Tó`ˆq­­­·oßnaaÁˆÏ;÷Î;äöíÛ·cbb ,--·mÛfmmͬ„„u»Ö¬Yÿu¨'ªªªª´´ôúõëóçÏggxúôiÝǪP—¨»P‘1^^^±±±Œ L&;v¬\.—ËåãÆcßÌ(66ÖËËK£ÒÓÓ³³³Ùñ¬¬¬ãÇkÔ7©TÚ°aî]».X°`ýúõŒ½ÏŸ?×ãX¦ u€º#4{öl???F0###&&fîܹŒ]~~~³fÍâÙ¹··7¹¡T*UÞ¹iíÚµÔ;ÁÇÇGƒ¼yfD …~‡0Q¨ËÔ-¨È-©Tš’’Ò AF|ÕªUŒ…f‚hРAjjªTÊ÷{=ý.NIIIåååô½eeeÉÉÉ*ëûæìì¬ß!L”¹Ø €@nÞ¼¹pც[$¥wïÞ ,0\9æúõë:tÐWoJ¥R_]6mÚÄÇÇOš4‰Ty¦«W¯nݺ5ÿžCBBfΜYXXHDIIIZZÚçŸNíMMM¥nóäàà2~üx-òWgïÞ½Œˆ¯¯¯û7]¨Ë€±+,,\¾|yzzú£GÞ¾}K ´ô¦±‘H$äF9_0¨ 6DDDˆU‘!•——?xðÀ Cèˆú¡SG€ÆÏ>ûìÈ‘#àh4qâDºµ²²š2eÊ’%Kȇk×®¥×eè‹O™2ÅÒÒR£ÎUª®®~ýúõÇ8°råJÆÞ°°0݇¨P—£–ŸŸïííýäɱ0y7oÞT—H$2dˆÀù‡7^¼x1??_å^''§7jÑmxxøòåË«ªª‚¸qãÆÅ‹ßÿ}‚ NŸ>™™I¶177×6q‚àQÛ"bÚ´iì•tê&¬/FmÑ¢EdQfüøñyyyhÁØÑ IDAT½O­AIþGìD îjÞ¼¹Êxuuõˆ#ÂÃÃ?~,pJ ŽÝêÕ«Õí·³³Ó¢Û-Z :”zHÍ‘¡O–:t¨«««ódiio¸!L ê2`ÔÒÓÓÉÅ‹;::Š›ŒðL·Fèûï¿ß¹s§‡‡{—\.ß°aÃ;#êŒñ8r䈺]GÕºÛÈÈHj{ß¾}ùùù999‡RÙ@œ.^¼8oÞ<©)¨Ë€QËÉÉ!7\\\ÄÍÀÔI$’‘#GÞ¾}[ÄêÌû￯П—/_"IcpøðámÛ¶©Û›ššzøðaízö÷÷§^}¹\¾qãÆÄÄDòÊ&‚ <== ºÆP^^ž¿¿ÿéÓ§ 7„ÉA]Œš\.'7øß 8H¥R«3‰ÄL¯ô›A›¤¦ŽÞGd+))©qMܰ°°’’íú§ÏˆÙ°aÃæÍ›Uî2W¯^ùòÏBÂß6}e}ûöùùùÕ«WÏÊÊÊÓÓ366öÍ›7|¬ªªJHHðòò²¶¶nÔ¨Q@@À/¿üÂ>$//oÁ‚ÞÞÞ7677·³³ÜÎÎ΢U«VÑÑÑÜ_xx¾.|N„#1Ž6lš¾Çt9}¨ĭηiÓ¦åååq·ÉËË‹ŽŽÖ®ÿ &4lØÜÎÍÍ%ïœMD£FôòkŸ,`UTTüóÏ?©©©]»ve4(++ d2j¬€ñ ß¬®E‹b§µýÿÒ­[·6èXóçϧÿmJII1èp NVVý…ÐcçT·?üðû‰»»ûóçϹT(ô¥ Õ}˜IMM­_¿¾Ê=ãÆS(¢$VTTäãã£2«÷ß¿¸¸XÝó¶qãF[[[öQñññ5?ïœ^½zÕ®];*ÊÊÊêêêÊÊÊ^½z‘ÁöíÛ¿zõŠ£u¯‚îT>WÏ0w25rðàA Fÿ]ºt©¨¨P™ÿׅωp$ÆÑ†A‹÷˜Ö§&dÅŠô—511Qe³ªª*uÕ’¥¥eXXØ£G´ÈáÙ³gô®|}}u;§yýú5½óÎ;kÚƒ¦?k†î§ººZåí±Ûµk×¶m[vüàÁƒÚ¥4mÚ4vƒèèh-NªÆf2™làÀìfgΜÑôÉ™0a½‡óçÏkÚƒB]À” .BB]¦¦.#•J?þøãŒŒŒŠŠŠÌÌÌO>ù„Œ0€ûÀ¸¸8++«ï¿ÿþÁƒYYYQQQŒ©©©dãzõêýç?ÿÉËË+//§nJÄìÙ³EI, €lþœœ)#•JOž|(pbǧ†ÉdT¼¢¢‚úþsêÔ)F?¶¶¶&sfïU*•ß}÷;+-,_¾œÌA*•þðÃÔÏþŠ+j<–ã#¸ñ…£™º]T|ذaô8µÔBŸ>}‡èòºh÷tñ9J»÷˜§&G£º IïÕÔe42jÔ(vWÓ§O'÷ª¼piôèÑÚ¥Ô¿ú^öß}ž'ųYbb"£»T£ZY—Áú2 ¨èèh+++ê¡……9»„ ˆíÛ·sèããCÍa¡Ð§a'&&’ËÁŒ=ºoß¾ôf¶¶¶3gÎ$¢ªªjË–-'–––F A¯°[YYQÍØC¬_¿¾¢¢‚ ˆÉ“'ûûû3öJ$’Å‹sdÅß7ß|Ó¯_?‚ ”Jå÷߯T* ‚èß¿ÿ×_­—þMÂÔ©Sé @ndff2Z öºhD»÷…ÿéC]€ugD´ÿþ;w2‚íÛ·_²d ¹½téRêúSÊŽ;T^úT#êϜʇzÚ¬Y3z¤ººú?ÿùA5 ¨Ë€ z÷î͈|ðÁ䯕+W87nwÏéééä†Ê5 ©»~ž;wNàĨÃÙCP‘Ë—/3v;vŒÜ3f wÿ:’H$©©©öööTÄÁÁ!%%…ÏB³Ôÿú ™ ºwïNèääDn¼xñ‚ÑR°×E#Ú½Ç(üOêTgt§roŽõ¼‹ŠŠER‚ ¤RéÖ­[©­lmm·nÝÊþýÔM¨Î`ïÞ½»wïfÝÜܨ+˜è–.]êææÆîÚµkß¾}†ÊOO\\\>ýôSF0!!A¡Pˆ’‘À¯`Tyy9#B®ÖAYP‡ª’¨CÝ»ªªªªªJ©T*•JöU6ê>ü.1êpöT„}oo*ÂXÕÒÊË˃ƒƒÉïí¯_¿f§jºÈsôEÈ×…?íÞcAuÆp #""Aò &v±• ›¤¤$vE5""¢¨¨ÈPYêÉôéÓ‘œœã¯(ê2 (ö‡õ'Ož...ºôLþó°I“&5Þø@àĨÃÙCPöÔÿBïܹ£Ëè|L™2åæÍ›ÔÃ7n°—0~ÔÜøªª*z\‹58ùºð§Ý{ @ ¨ÎBdddaa!#8cÆ ___u‡øùù± ìúޱyï½÷|||Á:¾ú/ê2 ¨³gÏ2"çÏŸ'7¼½½ué9 € ˆ/^¨[AF¬Ä¨ÃÙKSöäé±cÇ]F¯Q||}ºÊõJŸ>}:bÄá5j5ý ¹L&£†`ß܇:M›6±kFAüðê¼ÙÛ¹sç¨ûaO™2åàÁƒ“'O&Μ9Sݽ«4EÔ—Cê–I¤µk×êq]^gggr###C)Ú¾Çtij:Ã1éÔe@`÷ïß:tè;wärù_ý5|øðÛ·oѧOj&‚vZµjGį¿þêãã³sçÎÜÜ\…BQRRròäÉððp77·½{÷ ŸØ€È»tß½{7(((33S.—gffݽ{— ˆþýû÷ë×OÝé(Š?þxþüù÷îÝ“Ëå/_¾<~üøàÁƒçÍ›§KVÄÿ Uä‚;>>>äø„„rb…B¡ ÉÍÍÕqÁ‘sæÌ9yò¤\.ÏÉɉˆˆ¸xñ¢GÑåuéß¿?¹1uêÔË—/Óë€:Òî= 5Vg—¯ÉU`¨ùÏA´hÑBìt –£/&׺ukƒŽ5þ|úߦ””ƒêP7&è±sª[ÆËMêØ±ãóçϹä9Pbb¢º…x]]]÷ìÙ#Jbùùùên§Ý³gÏ‚‚unÙ²Eå¢fff«V­â±Æô***¨ [Ÿ>}JíÊÉÉ¡n2åãã#“É´E/xŽR^^îééÉx¢,,,>¬®Žž¹Õîuyôè‘£££Ê·Ê¡¹1:×â=¦õéƒ Y±bý͘˜hÐ᪪ªÔUg(¾¾¾z‘±wçÎõØ9¡ &Ð_ñóçÏ‹‘`¾ jÁ‚iiiï½÷žÍ»ï¾»páÂk×®999é¥ÿðð𬬬ï¿ÿÞÛÛ»Q£FffföööüñƳ²²Ø·ç&1‡?þøcýúõ}úô±³³377·³³ëӧφ .^¼hoo¯îÀI“&eggÇÆÆúùùÙÙÙYXX8993æêÕ«_}õ•.)EEE]¹r… ssóÝ»wÓ…% Xä2º—/_ŽŠŠÒe ÁX[[Ÿ9s&""ÂÕÕ•|†‡~ùòåÀÀ@½¥ÝëÒªU«[·nÍ™3§Gä›S)iýУçΛ¤ZÛe@xVVVÔEã-Z´0òEÔÀÔ™™™Q7—mݺuvv¶áÆZ°`All,õ0%%eüøñ†Ôyøð!u¯‚ ëtè‚Z‚ÄØ>{mbº[¹rå¬Y³¨‡‰‰‰ááá ­T*÷ìÙ³páBòb:Н¯ï… ô5Ê›7o4h@=ìܹó­[·ôÕ9¡ÐÐД”êáùóçýüüDÌG/0_ô sgxB] Õ€¡.D¯Î¸»»‹€q1ÏËË;SÕ¸qcu·ü:²:3räH±0.æÍ›7;Sµÿþ   ±³Se.vP'íÝŽŒ61¨ °¾ €8P—Ç¿®c²¶¶îÖ­›X©¿G={öLì, –øW]ÆÙÙù÷ß+ã÷õ×_¯ZµJì, –ÀuLâÀý˜ 6ÈÊÊêÙ³§ØY€egg‹‚þ¡.ðÿØ»ó¸¨ÊþñÿgmÀ EÅÜEÍ¥ÀDÍ}i5-µLsÉ-Á½Û>&*˜[Y–†k)†hV†ZhÞfiÝj.Yn€Z¦*ŠK˜ì2óûãüîùžûœ™afæÌ ¯ç=Î\ç:×õ>3„ðæ}® àòòò~ýõWµ£J‡ç˜ÔA^@äeÔÁú2—äïïçε£PÇÎ;_ýuA¦M›6wî\µÃQGåÊ•ÕÁÈËÀÙeee½ûî»ß}÷Ý•+WrssÅF½^¯nTŽ¡Ñhărr¿@©h4šjÕª©… t:ÝòåËÅã 6DEEU¯^]Ý`5ò2pj7oÞ KKKS;pÛ·oONNÿùçŸ>ø`ñâÅꆫ±¾ œÚÂ… Ť̈#233õÿ¥v\¥¦ù/µàÚt:Ý‚ ¤-±±±åöy.7@^Ní»ï¾-ZT«V-uƒq<×ÍC°i±ŒH,™Q+”y8µôôtñ nݺêFªSˈ(™q]äeàÔ Å~vPÞ)‹eD”̸.þmŽ ]Yã믿îÒ¥‹¿¿¿O«V­bbb}úlذ¡¸¸Øü[wýúõ˜˜˜N:U«VÍÓÓ³N:#GŽ0`€òZYÿÍ›7W¬XÑè:Çøð¡*ݾ}»cÇŽF£ ¿s玩÷mýúõ~~~Ê«¼¼¼V¬XQòûnÚŠ+Ä¡BCC•g;uê$ž]¹r¥™AL} egô½2ó›¦ÄKvîÜéåå%¿mÛ¶ùùùFóüs±äFÌf¦Œ_cVß>\È{ï½'ýX׬Y£vD(“/¾øBúV¨P¡N:Ò–Ù³g«#JzàPóçÏïß¿ÿ™3gòóó“““Ÿ}öYARSSGeþ•+WîÝ»wîܹ/^ÌÏÏÿã?"##e}FŒñàÁÿ>ú(333//ïÈ‘#áááâÙ9sæ¨ذaÃŽ;&ÂSO=uöìÙüüü³gÏ>õÔS‚ 9rdĈFG^·nÝøñãsss}}}£¢¢RRRòóóïܹ“””Ô¹sç©S§š̼W_}UL÷Ÿ8qBVþüù_~ùEŸW_}µ,³XMú«ÑF½M×Á=zôøñã¯^½šŸŸ¿oß¾š5k ‚púôéõë×+;—ês±äF”÷bÅmZ÷5fÅíP‹²Xæõ×_ŠŠ’¶P2ãò¨—Ì£^å õ2åcêeBBB¤„/,,lÓ¦xêÇ4sa½zõ¾ýö[ÙYi5Ç7ÄJýû÷K»åää4iÒD„ *üùçŸlß¾}†) íùùù!!!â©~øA6ÂÕ«W}}}Ř•gu:ÝÛo¿­ŒªT,Î>}útiû¬Y³ÄöÁƒ›Áð&”1›Ìb¦›©S†ö^xAÚþé§ŸŠí=zô]R–Ïź·Ë’«¬û³âöár¨—q'²booïË—/çççׯ__ÚNɌˡ^8Ô”)S||| /½¼¼ Õ%[¶l1saǎũɓ'Ž×¬Y#.óÊ+¯ôêÕKÚÍÏÏOÌ­oذÁÁmݺÕ0…4Ãîããc覜bíÚµùùù‚ Œ;¶gÏž²³fÑ¢Ef¢²Ä˜1că„„„¢¢"ñ¸¸¸xóæÍ²nï7Þ¾ìׯŸx \\ÓŸ‹¬û3°üö¨EY,3jÔ¨àà`ŸÙ³gKÛ)™q9äeœˆ«»•¼9à6ž|òIYK×®]ŃãÇ›¹pøðáæGþî»ïÄ£lôèÑC<øù矘árå†ñ ©ÿûßâÁ°aÃÌoµ¾}ûŠ;Ogee%%%‰ß}÷Ý7A¨[·nß¾}Í`ø[Ÿ"t˜Ç{Lú2((H<¸w§>+X÷5f`ùíP‹l&oooC:f̘1Ò’6fr9äelÀè‚ù%R;jÔÑ AS-éééf.ìСƒù‘/\¸ <þøãfæ½xñ¢ƒËÈÈ(q CÃce†Ç©lÎÃÃcäÈ‘âñÆѸ¸8ñ`äÈ‘åg_ꀀéKOOOñÀPFdà€ÏÅ Ö}X~ûTaªXF<¦dÆÕ©ÿo- Ê­V+kWëÁ̦Ԃ ˆk‘šñÏ?ÿˆµjÕòôô¬P¡B… <<<<<<4a÷S%ö ,''ÇԆ冖J•*™¿,F-ìÝ»733óöíÛß~û­ØRâ‚ÇîÄò ”c>—Ò²îkÌ ü$àe¦XFDÉŒKã[0p¨¼¼QæŒAxçwÌLjùÊ}þþþ²,Ì!CÌ?ºeÅ,"¶L­ZµÊ†³”ås©S§Žx æ¿lȺ¯1ÎÏÂb%3.м p¨ . 0àܹs………©©©/¾øâÙ³gAèÑ£G‰»2›×°aÃåË— ‚°wïÞŽ;nÛ¶íÚµk>¼{÷îþýû'NœØ´iÓíÛ·;>°~ýú‰»t§¤¤ 8099¹°°099yàÀ)))‚ ôéÓ§wïÞ¦nçáÇýû÷Ÿ?þùóç ÿþûï}ûö=óÌ3óæÍ+KTRcÆŒ1óÒU 8P<˜={öþýû ÓÓÓ'MštäÈÎR–Ï¥OŸ>âÁo¼qìØ1i°Œ¬ûàä,/–Q2ã5jdI¡‡ÌÇwíÚ5~üøÐÐК5kz{{{xxøùùÕ®]û±Ç}zΜ9~øa‰wT­Zµèèè^x!((èÖ­[;wîœ7oÞÝ»w t:Ý„ RRRŒ>ªã5¨dí IDAT¶ºý^½z-^¼ø­·Þ’ÿÆoˆ?–ɼûî»âbretìØ1ót:]|||‰µKF©uSà®RRRÆoóaí1¦M8m`ööðáCÖÏz½¾Ü¾€[²ùæ_pŒRmä4f̘%K–¤§§‹/Å™/^lã(a¥}ŽIY›ð÷ßíùÃ?ôìÙÓè)effüùçŸe½½½>,ë–ŸŸ&íæååUâÂÃZ­öÌ™3²¡Îœ9£\;p×®]æ‡*‘uo… o_ôâ‹/ZíСC- X„^½z:u*77÷äÉ“†í*eºwïþûï¿çææ?~\¹>Q÷îÝ­xsluSæñÊžc*‡”G¸žcr V¯,#Å*3.¤LEÕªU“µìÛ·ÏhÏž={þðÃe™K´eËYKdddxx¸¬ÑÇÇgÑ¢EÒ–¢¢"åµÊ¡Ú´i#klÓ¦Mdd¤¬1))ÉÒˆmÊæ·×¢E ó“¶nÝzÆ F²{÷î¶mÛjµÚ:,]ºTÙ§Y³f{öìi×®V« ]µj•¬Ãùóç-œÎ(›ß8L‹eDcÆŒ‘>—*–ÌØ&>ØZ™ò2O>ù¤¬eÈ!5zæ™g"##W¬XqðàÁœœœ²L!søðaYËË/¿l´g§Nd-%&† daû‘#GÌe'6¿ýJ•*íØ±ÃÌâÍUªTÙ±c‡¿¿¿…Θ1CºKT»ví”}¦M›&ÝWò‰'žuÈÎζp:£l~SàeYYFŠUf\H™ò2o¿ý¶ô© ÑåË—÷ìÙ³jÕªiÓ¦õèÑ#  _¿~»ví*ËDW®\‘µ„……]X·bÅŠ²ž)))æoÙ²¥…í²^”J¬S2¹)ö¸ý-ZlÚ´Éè)F³yóæ&MšXa·nݤ/kÔ¨¡ìÓ½{wéK///iK^^žå3eÛ›ǰI±Œˆ’WQ¦u;tèðÍ7ߌ=úƦúíÛ·oß¾}#GŽŒ‹‹+ËVD‚ X²­’)YYYæ;(s¦ÚËXÐa5;Ý~ÿþýëÕ«—‘‘!k¯[·nÏž=K5‹lo#iíŒAÆ e-ÅÅÅ¥šÅ6¼)(?êÖ­k“çŽÎI¹¶#œŠ­ŠeDbÉ̤I“ -±±±3gά^½zY‚„Í•)/#B¿~ý.]ºôõ×_óÍ7‡¾~ýº©žŸ}öY·nÝF]–é´Z­ÕF˜ïðàÁƒ*Uªm—µíævºýˆˆeþB„ŒŒŒ7Þx#>>ÞòYdk$MÃùùùÉZŒn–TF6¼)(?´Z-ÉkÔbÃb3¹lö¬Õj‡þå—_^»víÎ;Çÿâ‹/fÍš$ëWƹ¬ÎZ"55Õh»rZå­9†=n?>>ÞÔ#?‚ lÞ¼yãÆ6ŸÔÞÜò¦¸1ÛˈXeÆ%Ø /#U­ZµÐÐÐ!C†¼ûî»?ýô“ììÙ³g•—(K*ÌTO(—³½pá‚å»O™>11ÑhûŽ;d-Ê-Ãæ·Ÿšš*­j3jòäÉÒ”­ósË›àÞl^,#b•çW¦¼LŸ>}µålrû)))Êö?þ¸uëÖ-Z´X³fòldddJJJYæµ+·¼)nÏ®Å2"Jfœ™ƒò2}úô9tèP5”§:wîctCeSZ·n}úôéiÓ¦iµZÛÅ(DEEmܸQ¹,Ž( `óæÍS§NµáŒÖ)ãí‹K«ÈÎaĈ¯¿þºáX¹£¹© [Þ·g¿•e¤XeÆ™ýO6¤Q£F—.]²üâk×®8qâôéÓçÏŸ¿|ùò7îÞ½›——§ÓéüüüjԨѴiÓŽ;4¨}ûöæ‡:zôè§Ÿ~úË/¿¤§§?xð@ÿ¿Ï­èMú¨ô!¦ñãÇÛiºÝ»wÿý÷†—•*Uº|ù²©ú8Œ§ÚPÉV–!%%eÚ´iŽ™]\e†’Õ9nÝ_ R®,ãx¬2ã ÈËàhÊbÇcc&g@^‡r†b%3ªc}êÎ;6Yßwîܹ÷ïß½½½—-[fÅ 7oÞdõ_‘—À¡jÔ¨1eÊ”²³dÉC^ÆÓÓÓ&cÂÁÊo^†]±€ºX_@äeÔA^@äeÔA^@äeÔA^@äeÔA^@žÒ—/_öööV+Àù«À}üO^F¯×© @¹ÂsLê / Ï ¨ŒKKKøŒœ–ŸŸŸÚ!\˜çÕ«WÕŽÆi4ñ€Ï·ÄsL.À Œ“"€Û#/ ò2®òÜygD€ò€¼ €:È˸ Šhp3äeœùÊ ò2ê /ãJ(¥À—q.d^(?<ÕànòòòvíÚ¥vîïÑGmÚ´©xìéé¹mÛ6uã)ž~úéÊ•+Ûp@ò2.F£Ñèõzµ£sîÝ»÷Ê+¯¨E¹sàÀµCp©©©¶ÍËð“á!&Êò2ê /ã,,/–¡¬÷Àú2ûª_¿þˆ#ÔŽ°Æ—_~ùçŸÚo|ò2.‰Õ¸àààE‹©`S§NÙ5/ÃsLNG“(‡È˨ƒ¼Œ«¢ÄWG^F}dX(ŸÈ˨ƒ¼Œ £Ð—F^FeäV(·È˨ƒ¼Œk£Ü×E^FMdU(ÏÈ˨ƒ¼ŒË£èE^@äeÔA^ÆåéõzµCÖ / ò2j¢Ô€òÌSíP&dv6‘°ÿ~µ£àï¿ÿ~`` ÚQþäe€püøñÏ>ûLí(8Btt4yªÓh4²–r[vÀsL*+·_y`†Æ*–Œcy ¯¼òŠòò—_~ÙüUC‡U^5lØ0ó·6aÂKB?~¼…7nj"//¯ªU«6hРG“'OÞ³gOAAåol…¼Œ #§ö³zõêºuëÊ¿øâ‹-[¶˜º$!!áË/¿”5Ö«WoõêÕæçÚ²eËßÿm¾Ï½{÷ÌL]Z>ÌÎÎNOO?xð`llì3Ï<Ó¤I“­[·Új|Xˆ¼ FlÚ´IY‘žž®ìŸ––)kÔh4›6mªZµªù¹rrrâââÌ÷Ù¸qcnnnIQ[/##ãÕW_]ºt©ý¦€ë˨O¯×—ª”{{ï½÷žxâ µ£`Ó§O?yò¤ÚQ®ªwïÞS¦LY±b…´1;;{äÈ‘?üðƒôW9N7räÈììlÙS§NíÕ«—%s­Y³fÚ´i¦~=ÔétkÖ¬)eøÖˆŠŠêׯ_ûöí0ò2®‹‡˜öÓªU«®]»ªÛ¨R¥ŠÚ!®méÒ¥ßÿ}JJŠ´ñÀË—/Ÿ9s¦¡eùòå”]ÛªU«%K–X8Ñüñïÿ»ÿþFÏ~÷Ýw—.]*EÜÖ*..^³fÍúõë0žc¸½ì1¯¯¯ï–-[¼¼¼dísæÌ9wîœx|öìÙ¨¨(Yooï„„___ËçŠ5uêã?¶|S oTqqqvvöo¿ý6þ|e„(û\°y§@ñ 8­víÚÅÄÄÈ ^}õÕÂÂÂÂÂÂáÇ+73Љ‰i×®]©&úî»ï._¾¬lÿã?öíÛWª¡Ìóðð¨\¹rûöí£££×®]+;{ãÆ ÎóÈ˸$ò8àHo½õV—.]dgΜ‰ŠŠš3gΙ3gd§ºté2kÖ,  t:Ñ›V­Zeø5°cÇŽ¥ˆÛƒ ’µ<|øÐü%÷îÝ[¶lY×®]kÕªåããS§N®]»¾ûî»7oÞ;”e{rSƒœ8qb̘17Öjµõë×ê©§víÚ%ý5Y¯×ïÙ³gÀ€ 6ôññ©]»ösÏ=·}ûv+bpÖ— 8zôèÝ»wÕŽÂ)èt:Ãq~~þž={T ÆyuèÐAí(àÎ<<<âããÛ¶mûÏ?ÿHÛ?øàeçJ•*mÞ¼ÙÃÃÒJˆÈÈÈ×^{M<Þ¸qã‚ ´Z­álNNΦM›¤;VÚøÍPþá¿N:fúoÛ¶-22òÎ;†–7nܸqãСCK–,Y¶lÙ¸qãlž”^¯Ÿ={ö²eË ¿edddddìÝ»wРA Z­öï¿ÿ5jÔ®]» Weff&%%%%% 8pÛ¶mÞÞÞv Ï:äeœ…å»2•Ãb™7ß|óðáÃjGát²²²žyæµ£p ƒ úúë¯ÕŽnî‘GY±bŘ1c¤Ò?¬\¹288Øò‘‡ 2sæÌ¬¬,AîÞ½»uëÖ×_ÝpvóæÍ†mžjÖ¬9dÈ#FX¿)ÊB’Î;›ê;eÊS¿–fgg?>33Ó†áI½ùæ›Ë—/7z*11qìØ±Ÿ~úiŸ>}~ýõW£}vìØ1cÆ 3‹ø¨‚ç˜.@ùTKÙŸ”)­Ñ£G8Ð|ŸŽ5ªTÃúøøHkLV­Z%=+Í#Œ7Î&åz½þþýû¿ýöÛܹs'Mš$;;a£WýüóÏÓ¦M+±V`Þ¼yeÐ(SIÑÖ­[{ôèa*)#Z³fÍÅ‹mW™—q"å°\ËúõëkÕªeêlPPu;LOœ8±B… âñï¿ÿ~äÈñøÀÉÉÉâ±§§çĉ­ÜÀÃòðð¨R¥Êc=¶páBÙŠÅ“'OV®¤#‚^¯Ÿ0aBqq±´±råÊË—/OKK+((¸zõê|P¹rå²DXF%>á¥Óéâã㌅È˸r7 ¢ÀÀÀ•+Wš:»bÅŠÀÀ@+†­_¿þ€ / 52Òb™Ô«WÏŠÁ-äíí³bÅ £g“’’Ο?/ë¿ÿþéӧׯ_ßÛÛ»Aƒ3fÌØ¿¿]péիשS§rssOž<b´O÷îÝÿý÷ÜÜÜãÇ7oÞ\vÖÙVÉ`}¸žÑ£Gûûû«ÔtûöímÛ¶©Ê©o¿ýÖÔ©¤¤¤!C†X7lDDDbb¢xüõ×_ß¼y³°°Pº~mDD„u#["((())é±Ç3Õá›o¾‘µLœ8144TÖ:qâD3©«² Ù½{· :tXºtéóÏ?/ëÓ¬Y³={öˆ '‡††®ZµªwïÞÒ²ì’êÈË8ËWÿ-Ï-ZT»vmµ£€šN:E^ªøæ›oLݼyóK/½¤LX¢gÏž!!!)))‚ ®_¿>//ÏðÜP«V­zôèa]Ì–ÈÌÌìÙ³çÎ;MÍòË/¿ÈZLe † b§¼ÌŒ3ĤŒ¨]»vÊ>Ó¦M“îfõÄOÈ:Qv<ÇäJxˆ @¹¥/‰b¸{÷®©5q &L˜p÷î]ëÆ—VĬ[·îÓO?5zÊNîß¿ÿÜsÏ™Z÷Ʋ–V­Zíiª½ìºuë&}Y£F eŸîÝ»K_úûû{yyI[òòòìšõÈË`‘É“'—¸ tffæ”)S¬ÿµ×^3¬›{íÚ5qçlAªT©b“½±ÅV~~þ_ýµyóæöíÛË:äää˜JÝ¿_ÖR©R%£=+V¬XöP’-¯#­1hذ¡¬E¶V±³!/ãt(Š'´sçέ[·Ê7nܨQ#Yã–-[¤ëÂX®bÅŠ#GŽT¶9҆ɟGydøðáG}ê©§dg÷ïßÿÓO?)¯Rn´”““ctüØ$N%___éK£Ë€øùùÉZt:â± ò2.Ãò5âÆljG–ºsçŽrjF³qãÆ¸¸8å/8'N´îi¦ˆˆÙhÆN1y{{'$$(·ý^¶l™²s:ud-¦Ð=wîœMÂ+'ÈËP‚ÈÈÈ›7oʧL™òä“O>ùä““'O–ÊÌÌŒŒŒ´b¢æÍ›ËöêÓ§O³fͬÊÕªU‹ŽŽ–5îÙ³G¹ÊLxx¸¬ÅTMЗ_~i£èÊò2ÎÈJcDsçÎŽŽ.,,Tž*,,ŒŽŽž;w®ã£K$&&*÷mÒ¤ÉâÅ‹Åã%K–4nÜXÖáóÏ?ß±c‡ÓÉ:Öåw,7räÈêÕ«K[ôzýG}$ë¦ÜdjåÊ•Êô͉'Ö­[gó Ýy×PâÊRNîúõë111:t8v옴ýرc:tˆ‰‰¹~ýºZ±(W4ö¿}ûöo¼!ÄÃÃ#..ΰ”‰ŸŸŸÑ§™Þxã;wî”6ÂçŸ^ºÏÔsÏ=WÚJE«ÕŽ?^ÖïÞ=iËÓO?ݲeKi˃ºvíºnݺëׯ¥§§/_¾¼wïÞFÿ$SÈËÀ6lØpèÐ!__ßððð3fˆÓ§O÷õõ=tèІ ÔŒŠˆˆ¸uë–¬qÊ”)]ºt‘¶tíÚU¹ ÓÍ›7°¿uÙEDDxzzJ[rrr>ùäi‹F£Y·n]… ¤·nÝš8qbݺu½½½4h0sæLå¶M0¼Œ“R>¹çê:wî|üøñõë×0ÿüóÏׯ_üøñÎ;«ÀV,ù ¤ù¿IšÇò^yååå/¿ü²ù«†ª¼jذa¥»€ÛÙ¾}»rµ”¦M›ž`’Z²dIÓ¦Me_|ñÅ×_m¯øl¤nݺ/½ô’¬166öáÇҖ®]»~ôÑG%þ»üÎ;ïØ8>·F^Ž£Ñh<<þçKÎÃãT?jP¢Õ«W×­[WÖøÅ_lÙ²ÅÔ% ÊŸ¹ëÕ«·zõjÛǘV–Œ${ÈÊÊš4i’¬Q|‚I«Õ*ûkµÚ7Ê~ëaÒ¤I·oß¶W”62uêTYKzzº2£¹eËÙz4UªTùä“OfÍš%k÷ññ±Uœî‡¼ äèÑ£;v7n܈#Ä–áÇ7®cÇŽGU76€; Ø´i“òÚˆˆˆôôteÿ´´4åzŠfÓ¦MU«VµW”W‘••%kœ6mš™’ÿ.]º(·nÝRæwœÍO<ѱcGY£rõ_A^yå•?þøãÝwß ¯Q£†——WPPPçΗ,YráÂ…±cÇ^½zUvIPP½âv}äeàãÇÏËËûå—_–-[&6¾ÿþûGŽÉËË Ÿ0a‚ºÜIïÞ½•Oøggg9R¶é¡N§9rdvv¶¬óÔ©S{õêeß(fél5Ž%Äk¿üòKå©>øÀüŒË—/W^%-Ì,û­Y8Bi':zô¨¬ÿ/¿üb´g@@À¬Y³>|ëÖ­ÂÂÂ7n:tèÿþïÿjÕª%ÂW_}%ë/[0ØÂØÙGEäeà5kÖœ7oÞo¿ý*m ;yòä¼yójÖ¬©Vl·´téÒYã–/_.mY¾|ùÁƒeÝZµjµdÉ»†€‹êիב#GÌt8tèrñAƒÙ3(׿Yr Ì.\hê”··wtt´c8”Z’òõõݲeKXXXQQ‘´}Μ9ýúõkݺµ gÏžŠŠ’]èííàëëë¸Xp?þøã?þøè£¾òÊ+;wnÞ¼y@@ÀÇoß¾}úôéíÛ·'$$K/ ~õÕWÕ Øù‘—î©]»v111o¿ý¶´±  àÕW_=qâ„ Ç/((]Ó®];ÇE € :sæÌ™3g,ééãã³aÃ???{‡äºxŽ *pÂ'úné­·ÞêÒ¥‹¬ñÌ™3QQQsæÌQþ@Ù¥KåJÙÙÙŸ}öÙk¯½Ö¾}ûÀÀ@ooo//¯jÕª‰<\±bÅÅ‹Í\þ×_-]ºôé§ŸnÔ¨‘¿¿… üüüyä‘§žzjñâÅþù§™kMí×sâĉ1cÆ4nÜX«ÕÖ¯_ÿ©§žÚµk—ô_[½^¿gÏž4lØÐÇǧvíÚÏ=÷ÜöíÛ­˜+++ëý÷ßòÉ'ëÖ­ëããS«V­.]º,]ºôÎ;%¾uvºýÌÌÌùó燆†V®\ÙÓÓ³V­ZÏ=÷\bb¢%ñ‚pûöíuëÖ :4$$¤jÕªžžž¾¾¾ 4èׯß{ï½wíÚ5{‡df÷%vhàÒvïÞݳgOµqnÊõoà ÂÃÃ¥Sff¦Ú©I¶Úùõë×ÕŽ*ûý÷ߥ_ƒ R;"÷$Ýâ188Ø®sÍŸ?_ú™ÆÇÇÛu:(Mž]|'í’™ÙŽ_Z²ßˆJü:lB–ÓìÚµ«ÚÁ6,üfåïï?qâĬ¬,µãµ§Ÿ~Zzk©©©¶ŸzàÎyä‘+VÈu:N§“5®\¹288ØÌP.\hß¾ý{ï½—““cE$‡nÛ¶­¸µ‡™nz½þ«¯¾jÛ¶­ùU Þ|óÍwß}Wy;‚ $&&Ž;6//¯OŸ>»ví2zùŽ;f̘aÉD‚ ,Z´hôèÑ÷îÝ3zöîÝ»#FŒP¾Û";Ýþ¼yóÆŽûàÁ£gÍ@?¾}ûö+V¬P>Î&UTTôá‡vêÔ)33ÓÞ!€ó»|ùògŸ}6yòäž={6iÒD¬ ôôô hذa÷îÝ'Ož¼mÛ¶›7o®Y³&00Píx]mÓ<°êe¤lX/sïÞ½>ø wïÞµk×VþqLìÃÿ&f8É›C½ŒcP/S®Ø¯^¦DŽS†›ÓëõúšcàÀæGHOO ²ú¾RRR”e;æU®\Yù¹R êØ±£ù.\PÞ²s‰£&L¸{÷®©³W®\‘µ”¸˜®ÔÕ«We--[¶4ÚSÙž––f~ðnݺI_Ö¨QCÙ§{÷îÒ—þþþ^^^Ò–¼¼<ó³˜ ÏT»ôÁa»ÞþsÏ='kQ>’¦¼;å¦1F¹ÝuJJŠ=B”[äeP^Ü¿_ÖR·n]U"qi¬>ÀMž<¹ÄõÅ233§L™bêlnn®¬¥jÕª– ü%\ùÛ¾©ö÷ä®W¯žô¥r·AA6l(k)..6?¬Q–‡--À±ëí7iÒDÖbôQ~ –ËÊʲGH€r‹¼ Ê eAº0À]íܹS¹:LãÆ5j$kܲeË®]»Œâçç'k)Õs(ÒgˆD<0ÚSÙîïïo~pÙZ¹FÙG¿l¿$ Yv•*U Çv½}åà*T0‰Ñ«,Wâ~ŽÖ…(·ø½nÎP‡l攩º}ûöºuë†RµjUOOO__ß ôë×ï½÷Þ3µkøqã”ÅÒAAA²þiii5kÖTö7nœ©1Üˉ'ƌӸqc­V[¿~ý§žzj×®]Òü”^¯ß³gÏ€6lèããS»víçž{nûöífnÖÔ\6¿)°‰;wîLœ8QÖ¨Ñh6nܧüæ?qâD£O3ËZJÜ|ZJY®’ššj´§r9XÙJ%ê²<ì   ñÞ¾ò@-žj‹?~¼ZµjjG¡å#HNâþýûóæÍ[»v­ìOgÅÅÅééééééûöí‹ŠŠŠŒŒ\´h‘ì¯g±±±§Núõ×_¥7oÞ|饗~úé'oooAòóó ¤¬— ‹5˜^¯Ÿ={ö²eË ÍÈÈÈÈÈØ»wï Aƒ´Zíßÿ=jÔ(é_†333“’’’’’¸mÛ61†R±ëM€"##oÞ¼)kœ2eÊ“O>)ÂäÉ“e›ggffFFF*ëk:uêtîÜ9iKllì3ÏZ­öÌ™3–„-Vhªrû–tûñÇež~úéÂÂB££‰²³³7lØZÚ¹JÕM¹ÂNZZš™¨,$ÛêòåËe(‘©²qÀÕ¥¦¦Úöò2NмŒΗIOO—Vh[¢mÛ¶†ß öïßoô™ó¸¸¸Õ«W+Û+T¨ðã?Z}%n#âááqáÂKÞOe›Ü”yäeƒ¼L¹âüyKHGÈÊʪY³¦¬ƒ‡‡Çþói·ŸþYù4S­Zµnß¾- ©GÊ{õêµcÇŽ7nýóÏ?ÇŠŠK\¥×êt:e6¼zõê±±±ééé………×®][½zuõêÕe}¬x‡í×G„jÕª}üñÇÒ°•éøfÍš«uûv“íc%B«V­Ö®]›’’òàÁƒâââÜÜÜË—/ïÛ·oñâÅ}úô1”‘Ú/$½^¯¬Ù>|øÕ«W¥ï§ÈË@äeà®ÈË”äeÌ(U^Æ@9NYzöîÝ[ÖÇ××wΜ9ÉÉÉùùùÙÙÙ{öìyì±Çd}¦M›¦êÝwßUÎèëëkô1¢÷ßß’€mbΜ9V¿e¿)óÈË8y™rÅýò2C† Qv0ú}xêÔ©ÊžC‡•u»zõª2Ñca0z½þܹs¦6!2¥bÅŠÉÉÉV¼Ãöëc ¨xûv»|ù²ÑÏÙ/$½^?räÈÒ` ò2Py¸+ò2åy3TÏËüüóϲÞÞÞ‡–uËÏÏ “vóòòÊÈÈPÎøâ‹/ZrãÊ_L,B¯^½N:•››{òäÉ£}ºwïþûï¿çææ?~¼yóæÊ³eyËxSæ‘—q ò2力åe¾úê+åÙ¦M›æææ*çÊÍÍmÚ´©²ÿöíÛe=“““-‚U9ÑÏ?ÿ`áå?ÿü³uï°ýú¼óÎ;%FþÑG)‡räí[ÞíìÙ³Ê5‰Kd×:TÚ,A^ª /weó¼ ëþ:©þýû7iÒDí(œÅÞ½{oݺ¥vÿÏ–-[d-‘‘‘ÊTšÏ¢E‹úôéch)**Ú²eˬY³d=ãââ’““•›PHµnÝzÆ F²{÷nA:tè°téÒçŸ^Ö§Y³f{öìW# ]µj•¬È|<%²ùM€…²²²&Mš$kôððˆ‹‹3º;²V«Ý¸qc·nÝd[GOš4©[·nÒ’Šßÿ=&&fíÚµ¹¹¹¥ ¬k×®§OŸž>}º¸~©næ…^øè£œj'&QTTTݺugΜitÄ€€€•+W>ÜèµNxû­[·>}útttôºuëòòòì=%:wîmæ-\…V«}íµ×ÔŽ¢\8|øð¥K—4hн{wµc)ªT©bãm›æì¡sçÎÒ/ZÕëeZ·n-ëpüøq£C=xð@Ö³oß¾F{¦¦¦VªTÉÔÿ§UªTùã?,øÓO?•vHKKSöY½zµùPµZ­oŽ­nÊ<êeƒz™rÅêe¬<5cÆ ó3NŸ>]yÕàÁƒv¾wïÞÆ‡þè£V«V­B… žžžU«VmݺõСC?üðC£Kt\¼xqÑ¢Eýúõkذ¡V«Õh4¾¾¾ 4èÛ·ï;ï¼cþZKÞa{÷¹uëÖ{ï½×µk×Úµk{yyÕ¨Q#<<|É’%ÊEyT¹}Ë»ܽ{Wü@Û¶m[­Z5OOO­V[½zõæÍ›wëÖm̘1K–,IJJRþb~ùå—×_=$$¤R¥JÊõÌß‹QÔËnìÏ?ÿôôôA£Ñœ>}Zíp` ò2pΖ—)íòRõêÕ35¯Ñ}CÄï°ß|óM©nM–ï0ú7À””Ù8^^^æo¼TocoÊ<ò2ŽA^¦\±U^ßóá´ÈËnlÔ¨Q†ÿ»_|ñEµÃ5<äÿÞ(‰µëYYY¦Nõïß¿^½zÊöºuëÊ~œ*‘lñ&åõÅÅÅ¥šÅ6¼)R—.]JHH0¼LLL}zQQ‘™Kîß¿¿qãÆ°°0å©ÜÜÜÁƒçä䘟ÔÂnNÂ-o pF‹eD”̸ò2@©õèÑ£[·nÒ–={ö´oß~ݺu©©©999:.//ïÊ•+ßÿý’%Kúöí[£F×_ýĉÊÑ"""”5#-Z´hÓ¦¬Ñ’ 'á–78 £Å2"Jf\ާÚ.iÓ¦M¡¡¡·oß6´$''Oœ8ÑŠq”+°hµÚ¯¾úÊÓÓóñÇ—Õ’ÄÇÇwïÞ}ôèÑVEí nySP}ö :e±Ldddll¬á¥X2óè£:<4XƒzÀÁÁÁPn5]*)))Êö?þ¸uëÖ-Z´X³fòldddJJJYæµ+·¼)ÀyÈŠeÚµk·råÊ.]ºZ(™q-äe+µnÝúôéÓÓ¦M³nÛlqi•ÜÜ\Yûˆ#^ýuñ²ŠÄÔ…ÎÀ-o pÊb™yóæi4šèèhi#«Ì¸ò2€õªT©òá‡^»vmãÆÃ‡oÛ¶mµjÕ<===<<´ZmõêÕ›7oÞ­[·1cÆ,Y²$))éúõë†k'Mš¤¬Q–“ÄÆÆ†„„Ⱥ¥¤¤8çš,nyS€óP˼ð ‚ ôêÕ‹’Åú2(/,H¾´ÓŒ=º´Ë£]„EÉÏϯĤ- Ø1}lxSdLˈÇÑÑѽ{÷6œb•WA½ .ÀT±Œˆ’E^gg¾XFÄ*3®ˆ¼ ÎÎ|±Œˆ’WD^§fI±Œˆ’—C^§fI±Œˆ’—Ã~L¸!ñh–o0wåÊéËY³f-Y²ÄæQPŹsç¤/ïß¿¯V$¬cy±Œˆ™\ y \½zUú’Íì7F^p9–ˈĒ™C‡‰/Å’™íÛ·Û7JX‹ç˜psçÎŽŽ.,,Tž*,,ŒŽŽž;w®ã£eTÚb«Ì¸ò2¸ƒëׯÇÄÄtèÐáØ±cÒöcÇŽuèÐ!&&æúõëjŬVÚb«Ì¸ò2¸ƒ 6:tÈ××7<<|ÆŒbãôéÓÃÃÃ}}}:´aÃu#¥e]±Œˆ’WÁú2¸‰Î;?~<..nΜ9bË矾~ýúÑ£G{x”ð—˜–-[JVû׿þÕ®];;Æ À/^,]4ªzõê* T¬+–±ÊŒ« /€ûÐh4²Œ‡‡‡%U«Y³¦ôe÷îÝŸ~úi@%6læe´Z­ŠÁ°\YŠeDlÌäxŽ 7qôèÑŽ;Ž7nĈbËðáÃÇ×±cÇ£Gª(­²ˈXeÆ%—ÀŒ?><<}éÒ%µ£ø{÷î-ZT«V-uƒ)‡øgÇ /PßСC‡ªvåÚ”)S>þøcµ£ø×®]êÖ­«n$öÃsLÀŠü¸Ü?èNÍò0Ø2€szðàALLL«V­|||üýý;wîlf»„ÑF¯“Nš””®Õj«T©2`À€?ÿüÓ|ó톖ääänݺiµÚàààU«V ‚°iÓ¦-Zøøø4mÚtÓ¦MÊY~ûí·Ñ£G7jÔÈÛÛ»R¥JíÛ·_°`Áýû÷-¹‹3gμð >>>aaaÛ·o7Ú³´oø‘#G^{í51¤Š+>ýôÓGµäB è81þw†Š¤E ÁÁÁj‡ûšûì¹sçòóóÏœ9Ó§OŸèèh£ý¥?£mÔ«´íŒ3Ö®]›“““’’Ò¢E A®]»¶zõê²K„G}T¼äêÕ«^^^âû™——'„ bçO>ùÄÌ]T¯^ýúõ륽w3}.\¸ ~£ð÷÷¿pá‚ì,õ2΃zÀQ/㨗v±uëVñ 22RL(ˆ|}}§L™¢RPVzå•W Ç;w.^¼Xö‘ëÕ«'üïÓLb‹áçlþTk×®-**!**Jö8ÒØ±cÅÙz12Ó§O¯]»vÙc6X·nN§aÊ”)Ò\›è_ÿú— çÀ]±O6°‹cÇŽ‰]»v•êÞ½»mçÒÛùù¦V­ZŽŃ¿ÿþ»ì#ûûû ‚àçç'kÿ+‚aAßýû÷‹²ÚABBBăS§N™™ëé§Ÿ.{ÀR^xáÛŽ @ùA^pRÖí9¢ÑhìýË X(##Cr䈡Ѱ_x?^¼Á·ß~[YƒsïÞ½#FØd"+BZ¹r¥aYbƒ?üÐÁñàŠÈËnHõßL@„Þ½{÷êÕK„³gÏ4(%%¥°°ðìÙ³Ï?ÿü¥K—ÔŽÎ.úõë'Ì™3çÂ… yyyëׯÿòË/m2xóæÍßzë-A.^¼Ø©S§¯¿þúîÝ»çÎ[°`AÓ¦Ml2‘!=xðàÉ'Ÿüì³Ïnß¾ŸŸêÔ©1cÆÌ˜1ÃÁñàŠØ ØK|||Ïž=/\¸”””””dh_¸paTT”ŠÙIddäÚµkoÞ¼™’’"n­ÑhâããmUɲhÑ¢üüü>ú(55õ¥—^’žªQ£ÆçŸ^–ÁMåôeí²¥… !ݸqcÔ¨QÒ«/^\–x('¨—œ ¥.ÜI:uNœ81þü–-[zyyiµÚN:}õÕWsæÌQ;4» \õ2€{Òh4æwKǨT©Rttttt´¬½ÄïQ¥ú&f§ïx¦†53]‹-¤•A¦ú[×" 5€’…oNYÞÃÒ† ¨—œ‹­~µ )μ €:È˨ƒ¼ àtÊþ1€K / ò2ê`ŸlÀéõzFcõµ¶ åÁW_}uÿþ}µ£@éäååŽ>|¸aÃ|}}É˄ٳg_ºtIí(`½Â±cǪJ§fÍš<Ç ò2€“²îq$bB^@¬/[´h‘Ú!îiÁ‚†—äeçUÚ]™xˆ ¶2{öl«w`ÆÒ¥K¥yžcPyu—œšå&ñ¸ò2ê /ìBc‚ÚqÙ†…·ã*w••õæ›o¶jÕÊßßßUb¶•òv¿gÃ~L€³³dW&b€Ò2|kå[èÍ›7ÃÂÂÒÒÒÔ€òˆzüìÝy\Gþ7ðnE‘(ÊDPhÔâ…FÀ#š5/ûì3jPfÆŒÅÅÅ÷-ÃÌhŒ0_@cqØBß~ûJ|þùçíÛ·7ãkJo%4F˜/ÐðdÄJÐEAA•èØ±£¸‘4C—hÖjjj¨„™>þ÷hÜø'Ë<~ü899yÕªUï¼óN‡ðà=4µµµÛ¶m6l˜³³³•••³³óðá÷oß^WW§î”ªªªU«Vy{{[YYÙÛÛ<øçŸ&z]sDÛK=zÔÚÚšªÁÓÓóÑ£Gº‡§•]àïO—…œrõêÕñãÇ;99YZZº¹¹EDD”••ñDøèÑ£U«V 0 M›6:t˜9sæ7´èO`½•šÞ–ºtš‘zh$´øÆ>è‚ù—s©T*v8`XáááÌß)))zoBø¯ §OŸöë×O寯>{öŒ{JIII÷îݹå¿ûî;uíjñ+Q—ßÃÌTéý IDATÌ£GZYYQù]»v-**Òèt½PÙþ®ñÓà)?ÿü³¥¥%«þž={Êår•áÅÇÇ·hÑ‚’¥¥å?ü iGxã)âÅm©u÷õ. €Ùtnn®qÚ…FÇÝÝy«(•J±#hšZ¶lIÿ 9;;c¾ @SöÊ+¯üóŸÿ\¹råÑ£G‹ŠŠÄ@©S§^¹r…2jÔ¨¬¬,¹\ž••5jÔ(BHZZÚŒ3¸§Ìš5‹šIñæ›o^½zU.—ߺuk̘1ü±c~¢Òº’_ýu„ Ô£C]»vMMMíСƒþbJe_XŸõØÜ¬Y³BBBòòòärù‰'œ !ׯ_çÞ²eKHHHuuµMTTÔÍ›7årù³gÏ’““ ´páBM;Âí‹ÝÔâ¶Ô®ûÐìè4ÈF¤ûÏ/~ðA#˜/Ó¬˜Î|™'NPż½½ /—˽½½©C§NbžröìY*ßÃ㪪ŠÎW(>>>êÚÕñW¢ÀÓYÅ~ýõW3eô¤@ÚõEÈ!:ÿÝwßeæoÛ¶Ê÷÷÷g’——gccC177g½ÑõõõJ¥rÙ²e:vD‹³´¸-ëµê¾`¾ „ù2Æù2U=ö]€f`ïÞ½T"""‚ „X[[ÓƒG{öìaž²{÷n*±páBæÃ/VVV¬ñ&q¥¤¤Œ?žš)ãááqæÌQfʈbþüùÌ—#GޤÙÙÙ¬’qqqr¹œ2gÎÖh!D"‘|þùç S--nK&áÝ€fÈBìþ+==J 2„uˆÎ¡'¡]ºt‰Jøûû³N•J¥¬SÜÜÜô¡vèABÈ®]»„ Ê4%ŽŽŽÌ—ÿ÷§Á—/_²JÞ½{—JôèÑà ¤ÅmÉ$¼ûÐ a\ 1aÎh’ªªª¨„­­-ëSYY)ðj¥S@ÊB–/_Þd&ÂÄ\¯Šýæ2Ÿ½·%“ðî@3„ÿ$À„ØÙÙQ ™LÆ:DçØÛÛkzŠ)èÔ©“¹¹9!äôéÓëÖ­;½Q*•z¬~s_¼x¡Çju¤Åm Ð$I8ÄŽèÿüý÷ß111ÇïСƒÊ M6xŒË€1üôC?à“——Ç:Dç°¢$ÉÍÍUwŠèÜÝÝÓÒÒ–.]J½\¶lµ±w#B?}SWWÇÌöì™[éÚµ+•0©ë£Åm Ðlq¿ú !¤á1L™2…{úäÉ“ùÏ æžÅ\ëJeäsçÎRHHˆÀŽ«kˆfiiÙºuëÎ;ûûû‡‡‡;vL¡P¿8MÒÅ‹_{íµÈÈÈ“'O>~ü¸Ñ]ŒË€1´jÕŠJ<þœ§˜¯¯/•8wîëC—¡ôïߟJœ9sFÝ)¢;s挫«ë§Ÿ~úüƒ¢P(¦M›Ö¸>8Ò‹¤3ó¯^½ªÇVFŒA%öíÛ§é¹ôӾ̧ÆôB‹ÛD´iÓ&îPéxöMÛ½{÷ÁƒY™®®®›6mâokÏž=ÿý7™òòrž¦5U[[[QQQPPššºaÆ·ß~ÛÃÃÞ6®zòäÉØ±cYÿ75.—h¬*++7mÚD?ñ`âºwïN%èí“T¢ÿžËüv­P(6lØ@¥Y;õLŸ>J¬_¿¾ººšÎ¯©©‰Õ9pý k·´´Ü½{7µ"IffæòåËùÏÒâ¯Ä†ãííM%è-“(7nÔc+óæÍ³¶¶&„lݺõìÙ³ÜkÖ¬Qw.½ïxff¦C"ZÝ– "GGÇ;vpy†††pËçç燅…±2%ÉŽ;Z·nÍßVUUUBB™ü‘ùß“ÞN›6í«¯¾2\¦lÆ ú¹)‚zh<¸ëþ ??ø æB•R©TìpÀ°ÂÃÙ¿[RRRôÞÄæÍ›©Ê»wï~ëÖ-uÅ”J%½ÝõèÑ£oܸ¡P(nܸ1jÔ(*søðáJ¥’uÖ[o½E}óÍ7¯^½ªP(rrrÆŒCOÒáþêÓñW¢ÀÓU£G‹$ÉéÓ§uoEG[ùþûï©bÎÎÎÿïÿý?…B‘ŸŸ?þü6mÚhq‘yÑ=666ÑÑÑ·nÝR(ååå¿ÿþûèÑ£yâüàƒ¨ûôésùòe¹\. ÷‚º¯Ým©]÷ ! €0äææ§]htÜÝÝ™· ÷®‚hEH=šF²páBn%þþþ¬~ÕÕÕ :”[rÑ¢E»ÖµkWžkUWWǺ° vJÐ%ã077ÿóÏ?5½JšØcêÓ§+¤3f\}ûö-))ážR\\LÏæ I$’;wªkHeý<áiZžu3S©TÒ—¢S§Nåååê.šºjõK`+2™ÌÇLJÕ_KKË£GjÔ}!nß¾»ó!ÄÜÜü»ï¾SáÇÛ·oßàû"ä}ä¦Åm©u÷õã2 P—‘ÉdÜÿ!ß~û-³ØÚµk¹e|||d2™ð®ýöÛoêÂHNNÖ¨ãZ_CBÈ| éUÒ”À.“ƒƒ+¤üü|•%M'xÖ¸ žcc077ÿõ×_¿ûî»Þ½{«üÊMsvv¾|ùr\\œ¿¿¿“““………“““¿¿ÿ–-[ÒÒÒÚµkÇ=¥}ûöW®\Y¹r¥§§§………Ý›o¾yìØ±iÓ¦Qø[42‰D’жm[BHAAÁ‚ ÄŽH›ÔÔÔ ¸ººRoÊøñã¯\¹2f̽·5{öìÜÜÜU«Vùùù999YZZº¸¸L:5##cñâÅêÎrss»~ýú'Ÿ|Ò§Oj÷+}Ñâ¶Н©†hׯÆfÏž=–––¬üåË—Ó+‹geeEEE± XYYíÞ½ÛÆÆFx[ô#\zy¢–¾PuuuþùçÊ•+¹r×Yk¸ëÖ©[ˆÝ87ž$¦ 4ÈÚÚšµ–aQQý8}ƒè‡lñƒB˜››ÓÛßJ¥RîN7ДDDD0?8¦¤¤PŒ4vOŸ>¥¾0wêÔ)??_ìpÄxúôiúenn®T*/0]÷ïß§_*•J-ַ➢Ý'O}Õóå—_.[¶Œ•ùúë¯gddBúöíË]‘êË/¿üøã…„D333»wïÞ«¯¾ÊÊ¿{÷®§§'Oä* ìûÎ;ßÿ}fŽ]ee¥º¶ôB_ï‹™`H jÕªÕ‹/¨4æË42_}õ÷×=¨sýúu*Ñ»woq#Q,]ºÔÏÏ•™™™µ|ùrî ŒŸŸß’%KVNoĦT*UîÜ´qãFzŒ _¿~Ä-À¸qãX9µµµü§”——¯]»vðàÁíÛ··¶¶îСÃàÁƒ¿þúë'OžPîäÍO]%³gÏvww·µµíÔ©Ó¨Q£~ùå—úÿ}´óرccÇŽuss³¶¶~å•WÆŒsøðaþV  ÁÀ(|𷀋‹KQQ³X~~¾³³3·$½Æ™6tz( ŒnðàÁÌa¬/†ƒõeš#¬/#Љ'R=Ú±c‡Ø±ˆëË€@Z_F»`ôøÕõÁƒÌå<(fffÌ:”–-[òü€pCúé§Ÿèt›6mª««™å+++™KŸìÚµK`§«¨¨`{õÕWy®Ã¾}û¨Gh¹âãã…4-$6n¥R¹téRî'„Œ7ŽºnåååcÇŽU^PPB¡h°u/—˹KBú÷ïO·.“ÉÞxã n___‹ÍS°¾ @óÒàP±ˆ±èQ¯^½¾üòËŒŒŒ/^¼|ùòöíÛ!!!‡"„øøøÐû@sóꫯþðìL¥RI?¯M[¿~½FOùMš4‰^^ª¬¬lï޽̣»ví¢‡Nœ'Mš¤QØ âN$4hºÂ6l˜:uªºý¤+**BBBÖ¬Y£Ïø>ú裯¿þš{Á !‰‰‰sæÌ‘ÉdÇÿå—_Tžž””ij¸˜¾X[[9rÄÉɉ•ùòezo¯ \½z•U ]»vGޱ¶¶Ö¾má#:` 4/ƒ_ 5Ì—iVšÀ|u¿åÜÝÝïÞ½+vtbÂ|H¬ý˜Ö£[çꃂ‚øÃ Ò´kõõõÌÅkz÷îÍ,ÏÜÃnùòåÂ;ÅSL©TVTT\½z5**Š;pþüy•ž={Vë¥Ð…ÇÆS¦A >äeff–““£]+òäI•×*!!Aå£jæææ§OŸVyÙy`¾ @óÒà/±Ð?ÿüséÒ¥ýû÷wtt477wttôóóûöÛo¯_¿îáá!vt ²øøøöíÛ«;êââ¯EµóæÍ£¿Æÿõ×_iiiTúÌ™3ÙÙÙTÚÂÂbÞ¼yZTN£§º›™™988¼ñÆŸ}ö™B¡`– 箤C©¯¯Ÿ;wn]]3³U«V111ùùù …"//ï»ï¾kÕª•.êèÊ•+ü”J%ó©1¢Éh‹F‘~ñÅÜüùóç/Z´ˆ›ÿõ×_ûûûkÔÆe )èÝ»÷W_}uéÒ¥²²²ÚÚÚ²²²óçÏGFFÚÙÙ‰ˆÏÉÉiýúõêŽþðÃÜX„èÔ©sUzÃlæÎÙcÇŽuuuÕ¢r¬¬¬V­ZÅ}V‹’œœ|ûömVù“'O~øá‡:u²²²êܹóâÅ‹Ožw:¸ 4Y?ÿüóÞ½{Y™îîî]ºtaeîÙ³‡¹.Œpööö3gÎäæÏœ9SƒÖÖÖ¯¾úêôéÓ/_¾}úÐ/;f¸…Ó¡ ðòòêÕ«—ØQ[hhhII +3""ÂÏÏ™3xðàˆˆˆ~ø™ùäÉ“ÐÐPî\SºvíÚÚÚZ:§ªªjëÖ­K–,¡s$É–-[üýý™Û•””Ì›7»òìØ±ƒ»¬Œ­­í¡C‡,,,úôéÚ óÓO? :tÖ¬Y:µÚàfò`Lééé:½ÿ+22RëßHffÿS)•Jõø»Lëû””±#½aíçš››+vD`¢XOô(•J-*á~Ñ.í>ùЧ:tˆ{´k×®ÕÕÕܶª««»víÊ-øðaývM` 54yòdVáN:½|ù’U,66Vå.HLkÖ¬i°i!±³ŒF—«Á’ÙÙÙÜm¡!Û¶m£ üôÓOÜ£-Z´ÈÎÎVרJÌ­Êñ˜!O(¬ÁpA€É*--å>]B=ÁdkkË-okkûã?2ÿ"EY°`ÁÓ§O ¥ž,\¸•SPPpäÈVfXXØž={XëÑÐX³l(ÖÖÖúŠÓôQëÅTWW³òg̘ñ¯ý‹Ns§Æ¨;Q8ŒË@ÓZZZÊÊ\´hÑ AƒÔâççÇà())ÑËê!Õ¿ÿ~ýú±2¹«ÿB¦L™r÷îݯ¿þzàÀíÚµ³´´tqq4hЗ_~™““3gΜ¼¼<Ö)...†ŠÛô,X°àæÍ›¬L//¯Í›73s6lØàííÍ*vóæM]n¬/ 2zZ‡ÊYÖÍë:˜Ú´¼_<ôõs¡K=Úí÷ÃS@÷® ¬AÓ†._¾,°¤££ã’%K¸ób(܇¿X  ŒÍ˜e„k°¤Ê•e¸¨§–¶(ÆeLš··÷»ï¾+vИܹsçðáÃbG 2Œ˜0®Y³fàÀê \¸pÞ>œ6nÜ8Ç„`\ÆÄõèÑãóÏ?; hL’’’0.L§OŸ>}úô믿>eÊ”Aƒyzz:::ÖÖÖ>}úôúõë‡Þ½{7sÃ&BˆT*6mšX7+—hú233333…”´¶¶Þ¾}»Ê͉@ï°î/ sOœ¤¤¤ÁƒÛÛÛ[YYy{{úé§•••¬ò999111ãÆëÖ­›½½½¹¹yëÖ­}}}£££‹‹‹…´’™™ùî»ï:::Z[[ûúú²¦ ¯Ÿ®3;;ûÍ7ß´µµ•J¥7n$„ìØ±ÃËËËÚÚºk×®*DÿóÏ?gÍšÕ¥K++«–-[öîÝ{õêÕÏŸ?W6sñ!; ¬_‹ëC‘Ëå111~~~mÚ´±´´tppxã7/^üǨ«ßth±…“ðþjý~4:ŽŽŽ)))bÒlh´É6Zzz:óÝ ;"hd™¿F"##µ®Š¹[¤T*Õc`‚ÂÃÙwNJJŠ^ª¥+ä>²Nñôô|ôè‘Êò\­ZµJNNæoåâÅ‹666<u„×OçwèÐY,$$„ùR"‘œ={–ÙDtt4w§UBˆT*½sçŽÀ`ÔůQýZ\ŸÒÒÒîÝ» ŒDe+W®¤µµµ¿ýöÛÒ¥K‡Úºuë &ܽ{WëÈMVsë/˜&ŒË€ÁÙÚÚ²rè'qèݲ•Jehhhß¾}cccÿøã²²²ÚÚZÖYü32ìííyŽê^?¿/^P‰ººººº:¥RI/Í@—¡qD©ŸÿúB&NœøÛo¿ùúú23 Å‘#GúõëG/©Ód4·þ€i¸ œL&cåÐO‹Ð­_¿~Ó¦M„6mÚÄÆÆÞ»wO¡PPãuuuBZ‘H$•™ššjèvÕ2¢¤ Së/4+—ƒ;{ö,+‡Þùˆ~ŠäþýûT‚»Hð™3gtÁÐõ:”¢T*?®é¹ôQMM!êך••Õ7ß|C¥é'ÎŒ¯U«VT‚^eÙ@öWÈû ÆeÀàbcc ýòåË—6l ÒS§N¥ŽŽŽTâæÍ›Ìs+++W¯^­{ †®?$$„šÖ±lÙ²ŠŠ ÖÑòòò3f¨;—Þ‡;33ÓõëâÁƒT‚žÙÄ%ùC@éÞ½;•¸té’š  é¯÷ @ ‹†‹è&''gìØ±k×®õôô¼ÿþÒ¥K³²²!ǧʼóÎ;[·n%„L›6mÓ¦Mýû÷ùòåùóç?ùä“ÒÒRÝc0týžžžK—.ýꫯîܹ3`À€5kÖøûûÛÙÙݽ{711qýúõÏž=Ûµk—Ês‡NÅ6þü7öìÙÓÚÚZõ Ô½{÷qãÆ :ÔÇǧmÛ¶2™ìܹs~ø!utúôéºT®‹3f\¼x‘òÑG¹¹¹yyyé¥Z­û+äýhNœ8aè§š'öÎüKÇ‘¥§§3ßàà`±#€F&11‘ùk$22RëªÌÌþ;§R*•ê1H0AáááÌ;'%%E/ÕÒªœâííýäɺpii©ÊïÛööö§Nâùè"ðSFõsë’SWW·hÑ"n„víÚíÛ·O]l>¤WœeaÓ®~×§žwÍàÑ£GËd²Ol° íÔÕÕ=šÿâð¯Å)üýø~™Ž€€f¹¹¹bG&ÊÝÝ]à苳³3žcƒ[±bÅþýû `kkkkkÛ½{÷5kÖddd8;;Óeœœœ222V®\Ù£G ©TúÁ\»võ­R;†®Ÿbff¶nݺôôôÙ³g»»»ÛÚÚÚØØôéÓgíÚµ÷ïߟM¿ÌÍÍ•J¥â…¦ËÃÃ^"ŒóeDƒu Éâ®T‡¿Ø¨3uêÔ§OŸŠ@óbooqŸv˼s¿`ëò%|Ê”)ÜGÆ|Ž,88øàÁƒÜªöîÝ+°]¡rv04<Ç@!›6mêØ±#+óÀ{öìQwÊîÝ»¹ƒ2®®®›6mÒ|¦DÂ!vDÆe!ÄÑÑqÇŽÜ!†ÐÐЂ‚nùüüü°°0V¦D"Ù±cGëÖ­ %4-—ø?Æ ‹ˆˆ`eVTTÌœ9“õ<”R©œ9sfEE«ðÂ…  %@£Rÿb`¢0.ð__}õ•··7+óÌ™3111Ìœ˜˜˜ÔÔTV1Ÿ/¿üÒ á@ƒq0Eõ¢]›={öXZZ²ò—/_~ãÆ *••Å*`eeµ{÷nCDMÆeþG¯^½V­ZÅÊT(Ó¦M«©©©©©™>}ºB¡`XµjU¯^½Œ#4Ø'€méҥǎ»pá33333**ª¾¾>33“UÞÏÏoÉ’% V[QQñóÏ?Ÿ:u*++«  àùóçõõõ-[¶tuuõññéß¿ÿ¨Q£ºuë¦îô>¾}ûöꎺ¸¸ÄÇÇó×0kÖ¬ââbVf@@@bbâ£G^¾|YQQqþüùE‹µjÕŠ{úûï¿ÿâÅ fN›6mÖ¯_ŸŸŸ_SSSXX¸aÆ6mÚ0 <þœ¨H ¥R¹sçNf=ËŽ[˜;¿ò£Gò±ãÆ„qµœœœÖ¯_¯îè?üàääÄsú¹sçNž<ÉÊ\µjÕ©S§‚‚‚^yå ‹V­Zùùù­[·îþýûS§Ne–¿þú«ºCÉÉÉüçîÝ»—•3räÈèèh•…œœöìÙÃÌ9pà«LXXX=X™=zô ceúè£Î;3sX;5BþþûÔÊ´iÓÞ~ûmuG Úq!„,gfÌ%ÏLSiiéG}äããcgg×Ü®Fsë/”…بÐà%FPVV6wî\þ2sçÎõóóc-tB{øð!+§_¿~ÂÈËËcå¼öÚk*Króóóóù+3f +‡ûL–L&k DmMŸ>ç¨A;zñäÉ___üÖÐæË¨Î]²—¥¸¸8""BÝÑêêjVNëÖ­…À±··WY’›ßàžÜ¬kkká±é¨wïÞ}ªn+Æ>{öŒç²hÚ}~Ìzrssµ¨ ¤Y|KiD0.#äSŽðz„`Ö r¶Å¢E‹¸m-\¸[’û~åååqzS__ãÆ uß1Ô±··ÏÎÎÖî ,6sæL:"°Z;ÞÄ`\´c„q33³·Þz+33S.—gggÿóŸÿ¤òGŽÉbLLŒµµõŠ+îܹ#—ËïÞ½K==Ê,¼k×.ª°Ý÷ß_\\,“ÉÒÒÒHå/]º”UÿСC©CcÆŒÉÎΖËå™™™#FŒ`ÞùBúÕàˆ‹‹£JÚØØDEEݼyS.—?{ö,99™ŠA÷&t?KS[á)¦îïèèš——'—ËOœ8Aÿº~ýznm¦y‘GŒA5jTVV–\.ÏÊÊ5j•9zôhžš5ê~ƒ0.`ʚŷ”Fã2º|”äÖ#}ú¡C‡¸G»víZ]]Ím«ºººk×®Üò‡f•ÌÎÎæîâÔ`0´sçÎ9:: <ÝÑÑñܹsZ_aÅ.\¸ QGV+bÇ›ŒË€vŒ0.ãííÍüSMMM=¨C§OŸæ9ÑÕÕõ×_ee~;}üø15˜kffvòäIf±ªª*Bˆ¹¹ù½{÷èüãÇS•÷èÑ£¦¦†Î—ÉdžžžB~iüÅ’——gccCpêÔ)ÖQ¥R¹lÙ2›ÐËYšØ O1u‡èüwß}—™¿mÛ6*ßßߟuŠi^ä'NÐw¾B¡ óår¹··7uˆ­Ýã2¦ ëËBHiié‚ X™fff Üm› !¶¶¶?þø#w±Ã <}ú”™ãííý×_-^¼Xåäê <øúõëãÇçˆ]"‘]»vmðàÁZ´¢‘Aƒ­ZµÊЫ˜`Ç@GÖÖÖôKKKKzÑô={öðœØ¯_?zr 9´yófjš)S¦2‹µhÑ"22’RWW·}ûv:ß¾}T",,Œù´ˆMDD„&Ýj@\\œ\.'„Ì™3‡õݘ"‘H>ÿüs=6ה̟?ŸùräÈ‘T‚»â˜i^ä½{÷R‰ˆˆ+++:ßÚÚš¾{yî|á݀Ǝo·Z€æ#44´´´”•¹hÑ¢Aƒ©;ÅÏÏoáÂ…ëÖ­cf–””,X°ààÁƒÌÌV­Z}÷Ýw+V¬HJJ:}útfffaaaEE…D"±··wuuõññéß¿ÿèÑ£U6Ô©S§Ã‡ß½{÷СCçλ}ûvII‰\.·¶¶vvvöòòJ­ BáÙ-›ÒàêÔwrBHûöíɾµ2ÿ¥0‡~ª««ÕE¥òñU­Ñ]£Çz@ îÃÂê˜æE®ªª¢<÷Ï/¼ûÐØá7>ƒL&cåPk‚BX³ XèáuèÜêêêêêê”J¥R©¤Òc«­­åžÂŠ›£ º!zð!ô\<½0Í‹LßÕ<÷˜¦›@“„q0†¼¼•èØ±£.5Sä9::6¸ß}J§NX1Ð>|¨K0*c#„ܸqCÕ6ÿ75»®®Ž™ÿìÙ3=¶b𙾫¹w>£ãMÆeÀΞ=ËÊ9þ<•`®­¦…#FBÊËËïܹ#ðºÅ .°¥¦¦ê aì%½‰OMMC2&zµ—ââbfþÕ«WõØŠi^dúc.8ÍÊÑñ΀¦ã2` ±±± …‚~ùòåË 6Pé)S¦èRó¼yó¨¸.\øòåKn‰'2sè{bcc™§ÈåræúǺ£cÛºu+wdвfÍuçvèÐJdff oQòFjÞÞÞT‚Þ2‰²qãF=¶bü‹,ÄäÉ“©Dll,sÐG¡PÐw¾Im bÁ¸ CNNÎØ±coܸQSSsëÖ­ñãÇgeeBüýýéùÚqss‹‰‰!„?~¼_¿~û÷ï/**ª­­-++;yòä¼yóºvízøðaæ)„¬¬¬ñãÇߺu«¦¦&++ëwÞ¹wïž.Á¨‹­¶¶ö­·ÞZ¹råíÛ·kjjþþûï'N¼ýöÛÑÑÑêÎ>|8•˜?þ•+W˜£ZEPP•øä“ONž‰ Æ”žžÎ|w‚ƒƒÅŽ™ÄÄD毑ÈÈH­«bî"•Jõ$˜ ððpæ“’’¢¯šé:W®\Éýâååõøñcþ6´yófu+»ºº:tˆUþÑ£GžžžÜÂ̹*ãá§2¶íÛ·«ÜæÉÜÜü»ï¾S×£‡R;L lE»ë¦­Èd2Vð–––GUWOÍ"^d•|_ž}¨é=0C³±±IMM]°`«««………““Óøñã¯\¹2f̽·e‚ÙÙÙùòåËqqqþþþNNNÔð÷÷ß²eKZZZ»víôØ4^’zÁ#ÈÈÈ`®¼ÿ~ã€F'))iܸqôËÈÈÈo¿ýV»ªÌÍÍé½l¥Rinn®âSÁ\W%%%eôèÑz©™^è9Äxúôiúenn®T*/ø˜/  ±“SVVöøñcýÖ©÷ @ Æ»Õ:@s€q`›1c†Þë¤7$žcÆeÄqq`}Ó"“ɘ/ïß¿'V0Ð]»vùòùóçbEK—.] $v` ÖÖÖb‡ÿ%©¯¯;ø¯;w¾ÿþûbGMGŸ>}222´;×ÜÜ\©TRi©Tš››«¿¸€<Ç ŒËˆã2âÀº¿¦¥uëÖÌ—R©tÈ!bQ~~~jj*ý²}ûöâÅ À¸ŒiéСóe¿~ývîÜ)V0Ð%%%1Çe¼¼¼Ä‹€ç˜Äqq`\@—ÆeÄqq`\@—ÆeÄqq`\@—ÆeÄqq`\@—ÆeÄqq`\@b"H$T¢¾¾^ÜHt‡û/Ì—æË@0À@0.Óa„šÜÏÐxá9&q`\@—i.$ª)ÌS÷Ä?ÿüsÖ¬Y]ºt±²²jÙ²eïÞ½W¯^ýüùó"„dff¾ûî»ŽŽŽÖÖÖ¾¾¾‡Ö±§ÙÙÙo¾ù¦­­­T*ݸq#!dÇŽ^^^ÖÖÖ]»vݱc󬜜œ˜˜˜qãÆuëÖÍÞÞÞÜܼuëÖ¾¾¾ÑÑÑÅÅŬ&&MšD5ѹsgºw;w¦ò'Mš¤]ðZc^ÉäääÚÚÚ:88Œ;öÞ½{üåù󵻞ÃÝÝÏLiiiï½÷’½½ýèÑ£/_¾,äDƒ¨S’žžÎ|w‚ƒƒõU³Fï¾v7Ott´™™Š‘>©TzçÎþ†.^¼hcc£—›“>½C‡ÌÚBBB˜/%ÉÙ³g…t¹U«VÉÉÉÌ&JJJÚµkGý׿þEeΞ=›Êqvv.--m0<íz×`µÜñ‘Ž;–•• ƒ›¯Ýõ¬7ðýÀ­¶ÁKZWWÎ=ËÌÌì›o¾á?·qILLdv022R몘ï T*Õc@Á|™æ‚ù®kW˜{÷0O‰‰‰Y½zµR©ôóó»té’\.ôèQTT!äáÇAAA …‚§ÅWWפ¤¤ŠŠ ™L–––¤m_ÿÏĉ«««—,YB½Œ «®®^¾|9Õ„„º°ÏŠ+~ÿý÷üü|™LVSSsÿþý5kÖXYY=þ|„ ·nÝ¢ ·k×nÓ¦MTzûöíÇ?~üø?þHålÚ´ÉÉÉIÇ൶xñ⸸¸ªªª›7ozyyBŠŠŠèhu¡Ñõ4ôý ÑýL‰ŽŽŽ%„¸¸¸$$$”””Èd²«W¯N›6é¨t`¸ù2L½ûB çååYZZBzôè!“ɘ‡æÎK¾uëVžÊÛ¶mûèÑ#zÑ`ÀÙÙÙõõõÌñ*'''‡zùúë¯7XÛÚµk©Â³fÍb¢Vruuuuuø–èG®644”Îjjj‚ƒƒ+** ­^Ð1žô¼ ÞTH„Û·o‹ Æe ôEMMº2C‡%„(•ÊãÇ'*ýºÿ>•2dëЙ3gÔ5gΜÜÜ\Bˆƒƒƒƒƒ!äÁƒ|ðÁÂÔ'*`BȳgÏèÌììl½Tn‚÷!ä—_~5€ÿqh½/rff¦º2!!!Ôƒ'Ë–-ãι(//Ÿ1c†¾â‘ü‡¾*$„8::R‰›7o2ó+++W¯^­ò”Í›7Ó;=oÛ¶mÛ¶mTúСCqqq•Ëå×®]›={6ýnÿœ`déééÌw'88X/Õêr3<|ø°}ûö –¯««[´h‘ÊbíÚµÛ·o`ZôExsJKKUnŸlooêÔ)VáÊÊJOOO*§gÏž2™ŒÊ—Édô–L^^^•••úê²êªU—_ZZÊz[%É®]»´¸z*s z?¨¬–‹u–º$É_|Áßbã’˜˜Èì`dd¤ÖU1·÷’J¥z (˜/ pss»~ýúÇüÆo888ÐëŲ˜™™­[·.==}öìÙîîî¶¶¶666}úôY»víýû÷'Ožlä°5âä䔑‘±råÊ=zØØØXXXH¥Ò>øàÚµk¬Â ,ÈÉÉ!„ØÛÛúÈÞÞž¿•ÌÌÌèèè³gÏVWW÷ìÙsÉ’%&LÐ:~BHmmmBBÂþýû333+**Z·nÝ«W¯)S¦¼÷Þ{êž'¡érýùóµ»(þùgllìÙ³g ­­­=<<‚‚‚-ZÔªU«{¡÷÷—’––wáÂ…ÂÂB++«!C†DGG÷ïß¿Áš›V­ZYYY‰€˜ÊËËëêêÄŽšz0%éééÌw'88X÷:éÚ:tèÀ¬<$$„ùR"‘œ={–{–À»….³cÇÖY;v,++SWþ‹/¾à6äééùèÑ#žV.^¼hccÞ¦ñ?}ú´_¿~*Ïòóó+//çïuƒñpOá F tµš^ÿóéîŸúúúèèh333ò¤Ré;wø{¡÷÷·¾¾¾®®.<<œ{–™™Ù7ß|Ãnã’˜˜Èì`dd¤ÖU1ßA©TªÇ Á±~@RRRÄŽ@dݺucþPÔÖÖŠ4MxŽ©™8qbuuõ’%K¨—ñññaaaÕÕÕË—/'„Ô××'$$Ð…™w‰F­,^¼8..®ªªêæÍ›^^^„¢¢¢M›6©+5bĈÌÌL¹\ž=fÌBHNNÎ{ï½ÇÓJHHˆ««kRRREE…L&KKK bÐ4þ©S§^¹r…2jÔ¨¬¬,¹\ž••5zôhBÈ… øƒÑhzý…Óèþ‰‰‰Y½zµR©ôóó»té’\.ôèQTT!äáÇAAA …‚§-½¿¿„èèèØØXBˆ‹‹KBBBII‰L&»zõê´iÓèN›á‡~@/“]__ëÖ-VNNNõòõ×_ç¯AH+¡¡¡tæ‘#G¨Ìª+ÿÚk¯Éår:¿¦¦¦GÔ¡S§N©;«mÛ¶*'Ôhÿ‰'¨ÞÞÞ …‚ÎW(ÞÞÞÔ¡3gÎðÔ,<ýèi}ýÌ×âþÉË˳´´$„ôèÑC&“1ëŸ;w.UxëÖ­<½ÐïûKIMý°³³ËÉÉaÅ|u0_¦YÁ|Ì—ãÀ|™fÄÕÕ•üïÓ(TŽ‹‹ õ²¨¨H÷V¦L™B§ D%îܹ£®|DD„µµ5ýÒÒÒ2,,ŒJïÝ»WÝY~øá+¯¼¢k¬ÿA7Á\OÁÊÊŠþ¢²k×.žô.4½þ ¿âââ^¾|I‰ŠŠb=Ž4gÎ*qøðaž¶ô~=·lÙ¢T* !¬ÏÙ„ÿûßzl @8¬ûÛŒØÙÙBZ´hÁÊ¡þ%„<þ\÷V|||è´““•øûï¿Õ•2d+gðàÁT‚5{ˆ‰zÂH_膸ÁÐ9ÔSNºÇS¯ásašÒôú 'üþ9yò$•`UBÏ?ºvíO[ú} !©©©TâÝwßÕoͺÀ¸L3Bm*daaÁÊ¡7¢æ8船Õ]smm­ºònnn¬œÎ;S‰‚‚ugq7ÒEaa!«in0t#Ä£ M¯¿pÂïŸ{÷îQ j* 5Åü—RVVÆÓ–ޯ烨µæ€‰À¸ è™Ê-xxØÚÚ²rè'_*++Õ¥rm­UUU© †Îá FïñèBÓëo/^¼ <Û‹òêýzÒ!ѳ{Lø_á ™“Éd¬¹\N%˜Ì°H$=Æ@WçCçðè7±P+°èŽºVNNNü«[ñÔ ÷ëÙ²eK*AÁ˜ŒË€ÈòòòX9ùùùT‚ZVÖ:vì¨.:‡.Ó4Ð1§´TTTè¥ò.]ºBÊÊÊLg„ ‰rûömq#`¸ ˆììÙ³¬œóçÏS ___ãÄ@7tîÜ9Ö!:§_¿~Æ Æ8¨Ä³gÏèÌììl½T>tèPBˆR©<~ü¸^*Ô!ä—_~5€ÿqYll¬B¡ _¾|ùrÆ TzêԩƉÞ[:66¶¦¦†Î¯©©Ñ{0’ÿÐKmZ£÷ŠNKK£3y6&×HHHÕÁeË–qçà”——Ϙ1C/ iÒúõëée‰iëÖ­3r<ŒË€ÈrrrÆŽ›••USSsëÖ­ñãÇgeeB‡nœFŒH¹yófPPPvvvMM •¾yó&!däȑÆ 3N0Æ1räH*±|ùòœœ™LðàA½Tîéé¹téRBÈ;w päÈ‘²²2…BqãÆÕ«WwíÚu÷îÝziH‹*++‡ ²sçΧOŸÊåòk׮͞={ñâÅFŽ€‚ý˜@u³9XùüK· ôé§ŸFGGÿþûïÌLooo]ænh¿D"Ù·oßèÑ£ÿøãcÇŽ;vŒY¬_¿~{öìÑ:Ó÷äÉ“›7oR[GK$’Ÿ~úI_3Y>ÿüs¹\þý÷ßߺuk„ ÌCíÚµÛ·oŸ.•kwÒ!=~üøý÷ßgžõÅ_è€Ö0_D¶bÅŠýû÷0ÀÖÖÖÖÖ¶{÷îkÖ¬ÉÈÈpvv6fíÚµKKK‹‹‹ó÷÷oÓ¦………““S@@ÀÖ­[/\¸Ð¶m[ccNNN©©©o¿ývË–-mll˜œœ<}út}Õoff¶nݺôôôÙ³g»»»ÛÚÚÚØØôéÓgíÚµ÷ïߟ½²²’U &&æõ×_×´-‘Häââ2nܸ .Lš4‰uô§Ÿ~Ò´Bà ã2fjÙ²eÞÞެ̌ŒŒ¨¨¨ÈÈÈŒŒ Ö!ooï?üP—…BáêÕ«Y™ùùùºÔ jX™:PN(îÞ½»gÏž/^¼`æoܸ‘[ØÞÞ~Ïž=B¡®ßºµoßž•SVV¦c  æË˜¯×^{í«¯¾beÖÔÔÔÔÔ°27oÞÜ®];Ý[üûï¿Y9ºì· êa\@%A]ŒCHHHPPú2AAA³fÍÒ½­šššèèhVfÛ¶mu¯”¸ €¹Û±cG‹-TuqqÙ±c‡.õWWW?~üøÄ‰¾¾¾‡b>|¸.•€X_t¥ê«Bí  #úŽÅ- õˆ““ÓæÍ›ƒƒƒ•ýꫯœœœ4­“çd‘H4sæLM+ž0_ 8yò¤ªCIII†k7,,¬k×®†«ÀÂa\ÆâèýyøÚÓWµ@ûñÇ÷îÝ«êèž={~üñGC´kˆš€À¸ XŒ!€Fjëb„JKKçÍ›§¾Ì¼yóJKKõØh“&M¾úê«#GŽˆD"=V ,X_À¬-\¸°°°P}™Âˆˆ5sjê$‹\\\z÷îíëë;iÒ$­kž0.`¾Ž?þý÷ß³2;tèP[[û÷ß33÷íÛ7iÒ¤±cÇò¬3Ìžc²¥™JWœQ( ~~~Í›7—H$ÎÎÎþþþ‰‰‰ÕÕÕú ì?þ iß¾½D"±··ã7V¯^ýüùó:;BQTFFƸqãš6m*•J½¼¼>Ì,™››;~üøN:ÙÙÙ‰D¢&MšxyyEGG+ýÊ‘YsRRÒÀmllÆŽ{÷î]5]HMM}ûí·Iìììÿýw}õ×p4í¯ª›DÍ}•••5tèP›víÚmݺ•¢¨]»vuéÒE*•º»»ïÚµ‹ÛŠáîîÝÎs‰%®/€Þ•””„††²2ÁÎ;¹¿ÊBCCõû4\OMƒ1¥¥¥1¯Npp°^ªÕâf(..îׯŸÒbÞÞÞOŸ>­³­:£ŠŽŽ •Œ ¶k×îöíÛê+ÿí·ß¬­­Õᦛ7NJJRU3w¼ÀÕÕµ´´”LuuõÂ… ¹õ …Âõë×륿½ŸÑ´¿ªÂàæÓ9­ZµbV;wî\æK@péÒ%fUƿԿEš^_óqôèQfÀK—.Õº*æi×®ƒ3ĺᓓ“M‘ihú»ÂÐõ¼õÖ[ܪ-ZDŽFDDpN™2Å !YŽN:1ß.…BaꈠaÂÉæÅ@ã2L<ÿó÷÷'Å233e2Yfff`` É3fŒŽõoܸ‘óöö¾råŠL&{ôèQTTÉôôô”Édj*÷ôôìØ±ã±cÇž={VQQ‘ššÄ,ééé¹råʳgÏæççWTTÈåò¿þúkÍš5‰„¢(kkëììl¥5;::ÆÇÇ¿zõ*;;»K—.$ó“O>áIŽº¸¸$&&>yò¤¢¢âúõë3fÌàv_»þò?5¥iU…Áͧs-ZT^^þá‡Ò9áááåååôû6kÖ,ú,Cß|ú¢Ñõ5+—í`\† 8LXÏ‘#G¸õtìØñÕ«W¤À«W¯:tèÀ-sôèQÃuÍr`\Œÿ%›3—III!e<<<*++éüÊÊJrèÂ… Zן——'‹)ŠêÞ½{EEó½ßDBB‚šÊ›5köèÑ#5M¨òÅ_BBB”ÖFgÒ8UOnn.ùÈjkk›››Ë:ÊšO¡uk ?.ó¿ªÂàæÓ9YYYµµµ·nÝbåäææ’—=zô §ù~àó–jt}Í Æe@;—!(}Õ󆢢"gggV¡PøË/¿0‹]¾|™û4S‹-Š‹‹ Ô5Ëq0¬/JÐë FDD &„D"¡ÿpß³gÖõÇÇÇWUUQÅzüdΜ9$ÁZ„eñâÅ-[¶Ô¢é)S¦Ä/¿ü¢¾EQƒ "‰Û·o³Šmß¾½¦¦†¢¨ˆˆÖßmE}ðÁÌ—º÷×pxöW ­[·¦þý4Éqqq!/>|H&¼TÑèúBXXØ“'OX™ÞÞÞÌœÁƒsŸfzüøqXX˜aã=Á¸ (AOÛ2dësõêU­ë?wîIøúú²ÑóqþüóO55ÐOTiŠþô^PP ´€§§'vrr"‰²²2V±‹/’ĸqãêlT—þÒc¨u¶¢žýÕ‚­­-EQ5bå)Š¢ô5áý ŠF×@ï>|ðàAV¦»»ûºu븅?ýôSwwwVæ?ü ô1(07—%!·oßž¢¨ÒÒR}èݱcIüðÃááá:t W/殡¨ÒŠ¢rrrx6\€~øˆ9°õìÙ3½Tn†ïF×@k—±8ô…\.WUÆËË‹$._¾Ì:Dçôë×Oé¹ôz%ôª®\Æ £(ª¦¦æÌ™3|bÖÔ_ýEÜu‹/\¸ —&H(Š:qâφë¯888DII ™••¥—ÊÍðýÑèúh ã2‡Þ·8##CUzï丸8æð\.ß²e IO:Ué¹Ýºu#‰+W®¨ªîܹdZÍŠ+¸s.ž>}:cÆ õ½P¯iÓ¦$‘ÍÌùòåêÕ«u©™FwaóæÍô²µ´/¿üRia-ú+ø½„­5z¯èÔÔT:“ÞO]G†¾t ‰ÏõÐÆe,Έ#Hbþüùiii•••Ü2þþþ~~~EeggeeeÉår’&##GŽ>|¸ÒúéÐï¿ÿ¾ªg@:wî¼lÙ2Š¢nß¾=`À€#GŽ”––VVVÞ¼ysõêÕîîî{÷îÕ¥o¾ù&IL›6íÒ¥K•••/_¾<}ú´··7w%EíÐ]xùòå!C¾ûî»ââb™LöçŸΞ={É’%J ¨¿F0räH’ˆŒŒÌÍÍ­¨¨Ø±cwWí˜áû£ÑõКUÝE a‰ŒŒüñÇ?~œžžÎz‰^P ìß¿?000==ýÔ©S§Nbëׯ߾}ûTÕÿî»ïþøã§Nºyóf×®]¹•k×®•Éd›6mºuëÖĉ™‡š7o¾ÿ~­;HQÔºuë~ùå—œœœììlúŠ¢ìììNœ8A†œtGw¡  `Ö¬Yt¾@ X·nªÂ†è¯„‡‡ÇÇÇ?~ü8;;›l-vïÞ­¯™,}TÍ6b嫺Eù\_€z'++ëêÕ«¦Žt5eÊî.ºPÏh·_HZZóꢕÂÂÂ>ú¨wïÞôz®Ü›A.—ÇÇÇûøø8::ZYY999ùúú&$$TUU©¯¿ªªêË/¿ìÕ«W£FÔßiiii³gÏîСƒµµuŸ>}¾øâ‹çÏŸ«ª™ÿ}ûâÅ‹U«VuïÞÝÚÚÚÊʪ]»vï¾ûîÝ»wUU¢ªæ:[LKK !]hÔ¨‘¿¿ÿ¥K—ÔÖ¨¿uY#Zô÷Ö­[£G¶···¶¶8p`rr²ÒòÚ应ÿîãPz®F××L=z”Ù¯¥K—j]sõvíÚé1H0C .dÞ9ägÀ’ÑÏð …ÂÔéYll,Ïÿ"Àœ˜ú× èJP‹]Íɵk×è5w)Š >pà€ ã€zçØ±cãǧ_.]ºtÆ ÚU%‰èÝÐÛµkwïÞ==Äæ*"""..Ž~™œœhÂxL®sçηoߦ_* æ·Y À—_~‰'s€‚‚SG:Áú2¦qÓÀº¿–n̘1ýû÷7uP·íÛ·ççç›: Ð'ŒËX:ÿððpSGuKNNƸLƒç˜Lã2¦qÓÀ¸ €i`\À4°@CPPPpýÒÓÓ H~mm-ŸÓÓÓÓ™/Ç'â˰hr¹œù²ººZ$™*hÀ0.дjÕJ—Ó_½zÅ|YUU¥[8À Æe‚–-[öfpuu5uDP7ŒË4=2u`@åååyyy]»v5u  gxtÀ|ôÑG-Z´HKK3u, ˜/”··wFFýòèÑ££F2a<&×£G»wïÒ/M²èoqqñ† ¶nÝúòåKã·Æq`æ”J¥666¦ À˜vK2ŒÈXŒË˜ ŒÈXŒË€i’¨­­5m$`npo€eˆ €e¸ €)aDÀ’a\À40"—ó‚Ç—ÀÔ9"ãèèØ«W¯sçÎ9002S®3`iŠ‹‹?úè£×^{íóÏ?W:(Ó¶mÛmÛ¶ 2Äøá€‘a¾ €1Ô9G¦mÛ¶Ë—/Ÿ={¶D"1rl,ôV 4½Lj6PµõæË4|=zô`÷îÝÜ£»wï&G{ôèÁ=úÇ„„„´oß^"‘ØÛÛ¿ñÆ«W¯~þü¹ª¶ÿCQTFFƸqãš6m*•J½¼¼>¬ê¬C‡y{{ÛÚÚJ$U«V½xñ‚[L¡P$$$øùù5oÞ\"‘8;;ûûû'&&VWW+­V&“ÅÆÆz{{;::ŠÅb‡Þ½{/Y²$==]U$šö×p˜ïdRRÒÀmllÆŽ{÷î]õåÕçÓ9YYYC‡µ±±i×®ÝÖ­[)ŠÚµkW—.]¤R©»»û®]»¸­î~(ÃçJMM}ûí·IHvvv¿ÿþ;ŸŒ‰ç™;w†Ö—A¥ÿ}׉O=üc˜2e ÷ô·ÞzKýYÁÁÁܳ¦NÊ¿kR©´E‹={öœ={öÞ½{e2ÿ˜ØjÁœ¤¥¥1¯Npp°îu~þù礶#Fp>œýüóÏY‡¢££…B%#wíÚµ»}û¶Ò¶è2¿ýö›µµµš›Î\½z5·‰Î;?|øY¾¸¸¸_¿~Jïaooï§OŸ²")**êÖ­›F·½ýevDUíÐÕrÇG\]]KKKy†ÁͧsZµjŬvîܹ̗àÒ¥K̪Œs?¨*ÀU]]½páBîYB¡pýúõêÏmØŽ=Ê|C–.]ªuUÌ+Þ®];= FÀóG‰ÆúJNN6hxæ¯S§NÌ …B¡]=EEEË–-³³³SúŸõ¿™ÊÊJ?“víÚ¥[Ÿþ%66–Yy\\œ¦5¨ê‘z|êáCii©««+·†½{÷ª:eÏž=Üò­[·fþ=©i§·nÝZSS£Ñhæx^M/Ÿ.—ˆ2ßÀ‚‚SGºÂy1ĸÌýû÷ɇ+‘HôèÑ#æ¡G‘CB¡ðþýûÌC7–¸·Y IDATn$1x{{_¹rE&“=zô(**ŠdzzzÊd2n[täžžž;vcÆ ½_Žúã2@húÛ ã2,ºËè2"C`\†Ÿ~ú‰;ÅÆÁÁ!??Ÿ[8//ÏÁÁUX œ;wN÷®Mš4Iëñ;3Äóºhzùt¼ÜP‹q™†?æÅã2µµµ>>>¤Â72ó7lØ@ò}||˜ùyyyb±˜¢¨îÝ»WTT0Í›7œ’ÀmˆŽ¼Y³f¬1 U%»víÊüH_YYéááAýôÓO$3%%…äxxx0ÿva¾pá³þ-Z|Ö¼¥´îo­áÇeÂÂÂèÌ#GŽÌò ƒ›OçdeeÕÖÖÞºu‹•“››K^öèуœb„ûO_˜rssÉ­­mnn.ë(æË0`\ÆbiúÛ ã2,ºŒËè>"C`\†§E‹q+ñññaM`©®®6l·ä{ï½§—®QµxñbMƒ7[<¯‹¦—OÓòÀ…q™†ëËX„éÓ§“ÄÞ½{™ùôKº_UUEQTTTëñ“9s愚õb(ŠZ¼xqË–-ùÄ!•Jé—‰„þl°oß>’øþûïéÂÌ'®™…YSRKKKIBéj,,º÷×p¦L™B§ D·oßÖ½æÖ­[Sÿ~š‰ä¸¸¸—>$ cÞ£¿«£]¸p5ö{ñâEV1OOÏO?ýT_‘ÄÅÅåääè«6°—±'N$§ÿûßÿÒó#²³³ÿüóOŠ¢¬­­'NœÈ,îÜ9’ðõõeUEÿŸGÎU…~¨NÜ ézöPSøêÕ«Ì|ú³ºŸŸ_``àúõë/]º¤jI6]úKq*=ª;OOO:íääDeeeº×lkkKQT£FX9ä_Š¢è}y?ðDÿQ5nÜ8ýÖ  ŒÈPü&>¢]kkë}ûö‘é½L‘‘‘7oÞ$éÌÌLúlšD"Ù»w/w<.f EAAÁ±cǸš*Šï¾ûNÛ~XãÜõ Æe,BãÆÇŒCÒô¼:1f̘Æ3ËÓsL\\\¬¬¬D"‘H$ …B¡þÐNOHQªcÇŽ iZ9Ÿ¦U•),,\µjUß¾}7nleeÕ¢E‹1cưVÖSãï¿ÿþì³ÏÛ·ookk+‰5jôÚk¯¬[·NýêƼˆ`¬ê. B@@@³fÍJJJòòò._¾\[[KþÏhÖ¬Y@@«°]YY™““SQQ‘vÍñÿå^QQÁúh]QQA‡A¶¶¶ä™>…i£F5jÔÇûí·óçÏŸ8qâñãÇE=}úô£>:vì]R÷þš'²‹îŒy?ðdooOæzõêk¶€oذaëÖ­J‡c(ŠjÛ¶íòåËgÏž­Çá˜ôôt=~!qãÆ }Ue&„BáîÝ»{öìICÐÛJ2ÙÛÛïÙ³G÷/ÀÚ·oÏÊáó¼ymmíòåË¿øâ ú¯µ|¸dÉ’C‡±¦ŽVTTüóÏ?ÿüóÏ™3g¢¢¢&Nœøå—_*Ýy ô‰ÏS `4Ú‰˜?>©vΜ9ï¼óIÏŸ?Ÿ[²W¯^E …—/_jÔÿûŠ.IöbÊÊÊ"‡ºuëFrºvíªª0ý̰‡‡‡ú+++éw I“&ÌCZ÷×pT½“ªò™8Ñ™Ìg‹ÔÔPgŽîMÏ"!QuõêU*·Ø ´ƒý˜XTíǤ¯½–øàÎÅ0½ìǤ]7õU±sçN>ýMLLÔKHÌ­-‰æÍ›×YÛ’%KÔÄ6uêÔòòò>}ú¨)ÃܲS‹ÈùS_FMlšžÈ§é•+WªobÑ¢Eª®Ñ¯¿þÚ¬Y3ž¡6kÖì·ß~ã’á.b°SÃù2dúôéÛ¶m£þ½u÷!&Š¢† öÇÔÔÔœ9sf„ êÒ¥K¬õó/_¾LôSH^^^äÿ¼Ë—/«*ܯ_?õ I$’õëדw€õÕ–1ûk d¦¤¤ÄÙÙ™dÒ#\:2Ã÷‡„DQÔ‰'XO«”••ÑË6AƒÄZ=¸L2G¦~©s†l­áº 9yò$sf4WPPЬY³to«¦¦&::š•É]‘‹µQË÷ßÿ×_¥§§«)³mÛ6¥»R²4˜øÖ¬Y£¾@\\\xx8wMÃ[·n°¦P©QRRpõêÕ:14ÚE‹`ê!øƒÎ—©­­eÍ´lß¾½Òb999ä—x§NÊÊÊXGKKK§OŸ®ôDþ÷]²k×®2™ŒÎ¯¬¬¤G^Î;G2Ïœ9Cr<<<˜ß>1 ÿôÓOu6JÏ×e}ó¯u5ê²FTU«*¿ÿþ$ÿرct&=?ˆÒm¾ŒîMÏ¢C²³³»sçëhll¬F-60š/–Æpóež0Φz7_¦N<ëÑñ}+**jÑ¢…ª\\\ŠŠŠ4íó¨B¡(,,<~üøÐ¡C¹%—-[VgmzYgä"‘H‹ò/ÿ˜ŽMóÅ­Šûž££ãæÍ›óóóårùƒ¶lÙâèèÈ*Ó¯_?½„T'¥±N˜/ÓðÔ›ÿ-„¡ÇeX3W®\©ªäG}DÊtíÚõðáÃ%%%2™,333&&†ÌTz]s‘Ð%…BáÈ‘#333+++³²²FMò‡N®©©ñóó#ù7oÞ$…éÝ—GŽɪßÓÓsåÊ•çÏŸ/,,¬ªªzþüyRR’»»;)Ïý­­]5ê²FTU«*ŸžÈàáá‘““S^^¾}ûvæŒM55ðÉ1ôý ÅYtH-[¶ÜµkWQQQEEÅÿûß½_Žúã2 —),,TúM²šSø”©/´è~CRïúËóKìV­ZbD†À¸ŒÖ~øáU1üðÆè!‰²³³ùÔæçç÷矖——_¿~5œ6lذÿþ÷¿åååiii;wæ­3r+++žÔ® ÿb:6MQT›6m8ðôéÓ—/_8p iÓ¦¬>>>¬zΞ=Ë*ccc“‘‘Á*–‘‘ÁÝ+ýìÙ³u†d ‹X'ŒË49 PéCLÄÚµke2Ù¦M›nݺ5qâDæ¡æÍ›ïß¿__!ÅÄĬ\¹’õK³sçλwï¦_ ‚ýû÷¦§§Ÿ:uêÔ©SÌÂýúõÛ·o«Ú¬¬¬¬¬,¥3###Y™F믄‡‡ÇÇÇ?~ü8;;›Ìº»wïž1c†^ê7èû£f§æËÚÿH‡TPPÀœŠ,Ö­[§K<`8Ÿ|ò Yu~ÆŒ_|ñ…š/´µFÿê¨5Ø×›Z3B÷Üßdã+++»sçÎÓ§OpA»u릯ڮ]»váÂ}ÕfnNž<©êPRRÒäÉ“ ÔnXX½0¢ÉÉÉR©”¢¨^½z}öÙgo¾ù&«L§NN:EöíÛ·ïÖ­[‡Î,““SgC æ+–¦M›þúë¯ôÐvpppEEù6ŽÆ}C¸ÃsáááÝ»wgevïÞ=<<|Æ Ì̃úûû« Éh,‚‰Ç…àß =_¦¶¶¶oß¾¤ò¾}ûò‰‡ÞøÍÚÚºOŸ>_|ñÅóçÏU•ç_1K8p`À€666b±ØÃÃcÕªU/^¼àž"—Ëããã}||­¬¬œœœ|}}ªªª¸…³²²Ö®]ëïïïêê*‘H$‰««ëرcõ{xxxxx¬X±‚g<„¦ý¥ öU¡ªjÕ4×¥K—¤¤¤:Ëk—Cè~Ðå=Ôâ’»wï6ÐfhÏž=)))†nåþýû$ÁsÒÆÂ»_ß…„„Ìœ9s÷îÝk×®½wïëhyyylll|||hhè‡~X'CéÈþ×(--­óÏÅyóæy{{sWÑZ“&Mbbb.\Ès]ÖÂ4Í›7ç–6ló¥­­­X,®ªª¢s***´‰µ~3f +ÇÉɉ•Ã}C¸c7ªf3q󕎳0á"‚a\”›6mš©CãIKK3¸Œ\.' ËœeáÝoÄbñ;ï¼óöÛoctÆœ-\¸°°°P}™Âˆˆˆ½{÷jÝŠX,vpppqqéÝ»·¯¯ï¤I“ȳ*<µnÝšù’< ÃâææÆÊ©®®Ö4Κšnfyy¹¦õ˜w£%¥o wÔCÕêÝÜüW¯^©¯Üh,þ,0(Íä~½,PFÇ&LµY¬.ÝWS›šüëׯO˜0ÁÉÉI,»¹¹EDD”––ª‰ðÑ£G111 ·jÕjæÌ™7oÞÔ¢#º_JB¡P|óÍ7Çwvv–H$ÎÎÎ#FŒøöÛoU}žÑ¥ûZ £3¹¹¹ß|óÒ'›ÈèLûöí—.]úøñcý¶ê?~üûï¿gevèе')EQûöí«óq'&ÖCr¹¼¨¨(33s×®]o¿ý¶Fƒ2E±V™UúsÁܵP:È¢^UU÷¬¿þúKÓzLŽû‹D"-ÎRµÉ=7ßÖÖV}åF»ˆ` 0.–δc7ztâĉ=z´¤¤D¡PäççÇÅÅùúúVVV*-ŸàîîþñÇÿþûïOŸ>­®®.((ؽ{w¯^½6oÞläà‰’’ooïwß}÷üùóEEEUUUEEEçΛ3gÎ!CÔ²hÚ}]`tÆ •””„††²2ÁÎ;¹?à¡¡¡z¶39¢"V÷É÷†Š;]åÖ­[JKr×ßUº}€`\ ˆù ³ÒL¥«\©:¤u¦ZöB‹îë"$$dîܹyyy2™,%%ÅÙÙ™¢¨7nìØ±ƒ[xûöísçÎ-//·¶¶ŽŠŠÊÎΖÉd%%%IIIƒ Z´h‘¦ÑýRR5uêÔ«W¯R™™)“É233(ŠJMMU¿Ï FÝ× ŒÎ˜•ððpî›1dÈ!C†,\¸u¨°°0<<ÜXÑI“&MX9¬]·ïóðáC=Ö¯ îÖ"[¶lÉÊÊbefeemݺ••i¸ÍÔ¸0.Ð@ÌŸ?ŸùräÈ‘$Áý/“É(Šš3gޝ¯/ë¨@ X»v­ÁÂT‰^$""B"‘ÐùR©”žì°oß>U§óï¾!`tF”®L¤f¡¢ââbÖÕ§(J(&&&Ò |4jÔHéÓLóçÏ/))1\_ŒlðàÁ¬œ’’’I“&ÙÚÚ6iÒdòäÉÜ-ŠtÔªU+VΊ+òóóÍaÖŒ¿¿Ÿ>}˜9åååC‡ݺu냪ªª=z´mÛ¶¡C‡²ÖBöòòò÷÷7n°`Ñ0.–δÏ:éQïÞ½™/]\\HâéÓ§¬’gÏž%‰©S§!0žÒÒÒHbÈ!¬CtyÊI)þÝ7ŒÎ˜DXXØ“'OX™ÞÞÞÌœÁƒGDD°Š=~ü8,,̰ñÑÛo¿mäß ÀÊÙ»w¯›››H$2ùÒ]`×®]¬½–JJJÂÃÃÛ´i#‘H\]],XÀ˜³³³KLL4n¤`é0.Ð@4mÚ”ùÒÊÊŠ$ªªªX%ïܹCÝ»w7B`<=xð€$¸+nÒ9t.þÝ74ŒÎÓáÇ<ÈÊtww§Ÿ`búôÓOÝÝÝY™?üðÑ#G ŸquìØ‘µ8Wtt´[|÷ÝwõX›Þyzzž:uŠõËA¦M›ž:uŠ<8 `4—h øONo kooo°p4öêÕ+’àînKç¨Úæ–Ò¤ûÆÑ#(**Z°`+“<Á¤tïj›;wro• *Jãúì³Ï&Nœ¨ô½½ýÎ;cbbôØÜ AƒbbbÌy?»Áƒ߸qc„ êƒAAAþù'÷Y0C3¯ÿ½@)ý.Ö@Oìñâ…«Õ½€hEEëÃz$ÁüatƠ¸û@¿÷Þ{ƒ RuŠ··7wFÉ“'O¸ã;õ”D"9tèÐÑ£GÿóŸÿ8;; …B[[ÛÞ½{GGGß¹s'$$Dï-FGG§¦¦¾óÎ;öööf8FÓ¦M›Ã‡çææ®]»väÈ‘nnn666ÀÚÚºmÛ¶þþþkÖ¬ÉÉÉ9zô(¶Ç“°2u𬬬 EQÕÕÕ"‘ˆÎ×ïÒ¤îîîüñEQ7oÞ4Ÿ/‡]]]Ÿ?NQT^^ë9z±RWWWD¦32:óöÛoïÞ½{íÚµ÷îÝc £3ñññ¡¡¡~øa‹-L§Fôµ$“.õpŸ`â#66666VMý®6ŧ6}•¡­¹þýûóÜXZMkz™ÜÝÝW¬X¡Ýfs&¹ˆ`90_z¹\nÚH´F¯ƒPXXÈÌ¿~ýº[¡·Ù¿¿¦çîMöòò"‰Ë—/³Ñ9t™úsg€ ã2ÐpÐ;¶fddð?Ë䛆0ÑóDè-“ˆ­[·ê±•ÐÐP©TJQTBBÂ¥K—¸Ö¬Y£ê\íÞd>Þzë-’ˆ‹‹cúTVVnÙ²…¤Íj)íðiӦ͉'ŒÆe á1bIÌŸ?ÿêÕ«•••¦G ô£Ë—/?wîœ\.¿ÿþ‚ RSSõØŠ››yŽC¡PŒ5jÕªU999r¹¼¬¬,%%eôèÑjvl1Ü›}úlܸ1;;»¼¼ÜÔá@C†ÑËqP®ößLŽé}òÉ'ùùùE͘1£°°ïŒåÀ¼!0ŒÎXŒË€%ÒbT…ìAQÔÚµkÉ&F@FgîÞ½ëççgêX@ÿ0.ÀËýû÷IÂÕÕÕ´‘€ …­[·6u —àE.—“„PˆŸÐ|´D=Љ‰0`€£££••U«V­fΜyóæM½Tž››»qãÆ   N:ÙÙÙ‰D¢Æ÷ìÙsáÂ…ªÎ’Éd±±±ÞÞÞŽŽŽb±ØÁÁ¡W¯^‹/NOO××)Zl/¥´¤ªÞxã ’Ï­*!!}ã7Ô7jhÚ]ýëׯO˜0ÁÉÉI,»¹¹EDD”––r‹iwõ©¯áR]]½eË–×_ÝÚÚÚÁÁÁßß?99Y/­ðï>ŸK¯êþ),,üøã½¼¼š4ibeeåää4zôè'Nè«ûРԂ9IKKc^àà`½7±cÇŽFqï±XüÕW_©:‹ÿ £æf+V¬àžRTTÔ­[7nQ-NÑâþWÓn _}õÉéÛ·/·ª£›7oæÓ¢úÀ´¦ÑÕ§?~\,³NéÙ³§L&Su —ª«Ï:Q¡PŒ;VÍû¬K+ü»¯¦ UQ{öì±³³SZxúôé …B÷î×éèÑ£Ìs—.]ªÑé`±.\ȼs’““M€‰uêÔ‰ùC¡êwxýËì`\\œ©#^Èüá-((0uD +Ì—±,Û·oŸ;wnyy¹µµuTTTvv¶L&+))IJJ4hТE‹toÂÓÓsåÊ•gÏžÍÏϯ¨¨¨ªªzðàAbb¢‹‹KmmíºuëvíÚÅ:eÅŠd‚¯¯ozzzEE…L&»}ûöŽ;èá ÝOaÞ÷<û¢ôÖ?mÚ4‰DBQÔµkײ²²˜õäää\¹r…¢(©T:mÚ4ž­ëÖW?$$dîܹyyy2™,%%ÅÙÙ™¢¨7nìØ±ƒUR‹«Ï²yóæ3gά\¹òöíÛ2™ìÎ;áááziE£îó¹ôÜiïÞ½3fÌxùò¥­­í¦M› +**RSSÉÿ{÷ԽûÐøF#¸víš——ý288øÀúª|8EQýû÷'ƒ4—ÇSõðáÃV­Zñ©M‹S˜´èŸS&Ož|èÐ!Š¢/^ÌüjÙ²eëׯ§(jÒ¤IÔo`>>?ÿü3ŸÖÕ\}V[­[·Þ¶mÛþóæÑ¸¸8Ö7ùš¶b„›¿°°ÐÝÝýåË—B¡0%%…¹eFyyyÏž=ïÞ½+‰rss;tè ª ]ºO;vìØøñãé—K—.ݰaÿÓÁbEDDÄÅÅÑ/“““M€ÉuîÜùöíÛôK…B!‰LÞ}ùå—K–,¡_†……½ùæ›&ŒxZ¼xqvv6ý²  ÀÅÅÅ„ñ€è0×ôÏ Ï1-_¾œT;oÞ}Zïñ¤ÅÕ§ƒ9{ö,3ŸÞšÊÙÙ™gUj®>«­ &ð¬S£VŒpóGGG“bÓ¦Mãݶm9º|ùr5MèÒ}žcíà9&K{Ž ê)<ÇÔà9& rö쨥lhIDATY’˜:uªAºwï^TT”···³³³D"! š’g|(Šzþü9«|çÎIbøðáÇ_³fMJJÊË—/Õ4¡Å)FàïïOvÑ.**JJJ"™§OŸ.(( (ÊÕÕÕßß_} ôO¦ÞcÓåê÷îÝ›ù’úô)·°¦WŸeúôé|BÒ´#Üüô¨ÜŒ3¸G}||HâòåËj*áÙ}h0¬LÏ;wH¢{÷î†keË–-ï¿ÿ~ee¥ª555¬œ?þxÒ¤IµµµÕÕÕçÏŸ?þK¯^½êŒG‹VŒpóçææ’DŸ>}¸GÛ¶mKÌ)ñ\|º æËXz:‰½½½š8}úôÂ… +++ÅbqddäÍ›7+**È5È'L˜pæÌ™~ýú13«ªªNž<Ù¿ÿ[·néåã !‰3gΟüí·ß.^¼xüøqòìϳgÏ–/_~üøq½œb;v2dÈåË— Åwß}'•Jɤ’Áƒ»»»›$$œ¯>“µµµ!Z1B÷é&ª««ÕS(jŽÖÙ}}éÑ£Ghh¨©£]ÑßÿAý…q âîîþÇPuóæÍÁƒkt®@Àkë®ôôt’à®âÁg‹««ëäÉ“'Ož¼iÓ¦÷Þ{,•zñâEýžbh!!!d ‘ÄÄD©TJgš0$J·«Ï“ŽWß ­¡û¤‰¦M›–––¢~ýòóócç˜,½èìþýû5=·qãÆ$¡~ÝV™LFôx¹Ñr$ ÙXš¢¨W¯^è™4iyX&777##ƒ¢(;;»É“'›0$J·«Ï“¾®¾!ZÑ¥ûôrÂô~Ojšxúô©úd˜0.cABCCÉGÙ„„„K—.q ¬Y³Fչݺu#‰+W®¨i¢C‡$Aâ¥Ý¹sç믿Ö(Ú¿ÿþ›$Z·nm¸S ÁÖÖ–5 3yòd[[[>ç þGïQérõyÒãÕ×{+ºt¿U«V$AFÙêlbÑ¢EÜ‘)ŠzðàÁ¤I“ÔÔ`xþÒ+**zÿý÷===mmm ÷{R f˜XZ ã2ÄÍÍ-66–¢(…B1jÔ¨U«VåääÈåò²²²”””Ñ£GGGG«:—Þ¾÷ý÷ßÏÉÉ©³ØÜ¹s“““e2YYYÙþýû‡ ¦fQnݺEGGÿüóÏ……… …âÅ‹ÉÉÉ'NdÕ©ã)Æ4{öl5/MB—«Ï“vWß8­èÒý#FÄüùó¯^½ªj(º ²"õ>|¨P(JKKÏ;êîî~øða­;`±?~ܧOŸ7fgg———›:œÿc¶@=S æ$--yu‚ƒƒõÞÄ·ß~kccýD"ÑÆU¥P(FŽYçýSYY©ôAå¦M›^½zUÕYjîÏÀÀ@z·]NQS^Mw”Ö îÍeèܹ3)ïîîÎó-ZÑ”FW_M0Jiwõël‹E—V´»ùÿùçŸ-ZpÏRÚʶmÛT­ÝÛºuëC‡éØ}>Ž=ÊlwéÒ¥z©¼… 2ïœääd#4Êçæ'ef̘QXXh„¨x2ÛÀ4¥Ý¯ ýþâ2O:ubþP( SG æËXœÙ³gß»w/&&ÆÛÛÛÉÉI,»¸¸L:õÚµkK–,Qu–H$:yòäÆßxã ¥Ÿl ‰DræÌ™M›6õîÝ»Q£Fb±¸C‡aaa7nÜðòòRuVVVÖºuëFŽÙºuk‰D"‘H\]]ÇŽ{ðàÁ¤¤$¥Ÿrµ8ÅÈÚ´iC3gÎ4m$LÚ]}ž´»úÆlE»î»¹¹Ý¸qcùòå}úôqpp‰Djš ½sçÎÊ•+½¼¼HáæÍ›5jÇŽwîÜ¡§t§OŸ&‰µkת$5 ³ Ì8è?%M@½Çk“0šk×®1?[8pÀ„ñ€vî߿߮]»šš¡PøÏ?ÿÐc4FpìØ±ñãÇÓ/—.]ºaÃÆõEDDD\\ý299900ÐÐÒ«“¨ùkD*•’U·«««…B3ú>ÉlÓŸ«`™:wîÌ\Ê]¡P¨šÐN=þ3Àl%$$ÔÔÔPåçç‡A]Ð[¡™Û؇Ùõ‹•©u~üñGÓn-Z¨­­-,,$éëׯ㠂‘Ñ[‰˜òòòØØØC‡ݽ{—¢¨7ÞxcñâÅ&LPU^é.?¬L“LîÐ"05³QTbæ_¿~}ݺu—.]zöìY«V­ÆŽûñÇ;::ªŠðÑ£G gΜÉÍÍ}þü¹³³óˆ#>øàz_EþQ˜š2\ …b×®]ÈÈÈ(++kÒ¤IÏž=ßzë­Y³f)~¢K÷ê%#¯gê±ÖýÐÖýž ·îoAAA×®]¹7çêÕ«é4ë>÷¶šù”ÑŽ© ¦ÎSŽ?.‹Yõ÷ìÙS&“) oÇŽ5â†$‹¿úê+M;¢&05eXŠ‹‹ûõë§ôÄ–””¨y[4í¾ÞaÝ_0̼š2eÊ­[·(Š3fLVV–L&ËÈÈð÷÷ÿøãUÂüKEif­‰VB1r`!!!sçÎÍËË“Éd)))ÎÎÎEݸqcÇŽÜÂÛ·oŸ;wnyy¹µµuTTTvv¶L&+))IJJ4hТE‹4í·/ZtsêÔ©d³¼€€€ÌÌL™L–™™@QTjjêŒ3ôÕ}€zLÓ0(Ì—ýÂ|àÉ@óeΜ9C*ìÞ½»\.§ó+**:wîL7§¦>et?E <[QSLÕ!:ܸqÌüo¾ù†äûøø°NÉËË#{ŠD¢óçϳŽÖÔÔ¬X±BÇŽhqVJJ )ãááQYYIçËd2rˆ­Ý7Ì—ãÀú2æE(J¥RSG ‡•~σ)íß¿Ÿ$ÂÃÙϤX[[GDD„……™(®úaþüùÌ—#GŽ$‰¬¬,VÉøøx²¶Ôœ9s|}}YGÁÚµk ¦Jßÿ=IDDDH$:_*•.\¸ônß¾}Ü€ þݨ×ð÷ºyéÝ»7Ö쀃ž:xð`Ö¡¡C‡¨ÑÚ†²ßsïÞ½™/]\\HâéÓ§¬’gÏž%‰©S§!0žè«?dÈÖ!:‡<å¤ÿîÔkX_ åÁƒ$Ѷm[Ö!777£‡SÏ4mÚ”ù’žþVUUÅ*yçÎ’èÞ½»ãIÍÕ§sè2\ü»P¯a\ ¥¼¼œ$lllX‡¸9À"òý;íåË—$aooo°p4öêÕ+’Psõéȹøw ^Ãx`(vvv$QQQÁ:Äͱ(555z¬~Ÿ_¼x¡ÇjudkkKj®>9€Å¸ J›6mH"??ŸuèŸþ1v4¦@?}S]]ÍÌ/))Ñc+îîî$qóæM=V«#WWW’ÈËËc¢sè2 ã2`(^^^$ñ믿²]¼xÑØÑ˜½HJaa!3ÿúõëzlÅßߟ$è °ø£wJ’Ëåz ‰b\ýË—/³Ñ9t‹…q0z{ ¸¸8ær­2™,..Î@ þÇ@õkÄÃÃ$è-“ˆ­[·ê±•ÐÐP©TJQTBBÂ¥K—¸Ö¬Y£êÜV­Z‘DFF†C¢(ê­·Þ"‰¸¸8æ Oeeå–-[HÚ¬60 ŒË€¡øùùùùùQ•™™9a„[·nÉåòÌÌÌ7ß|óîÝ»¦ŽÎ‚‚‚HbùòåçΓËå÷ïß_°`Ajjª[qss‹¥(J¡PŒ5jÕªU999r¹¼¬¬,%%eôèÑÑÑѪÎ1bIÌŸ?ÿêÕ«•••úŠjäÈ‘>>>EeggeeeÉåò¬¬¬   ììlÒôðáÃõÕ@=%¨­­5u `bÌ ,ÉÉÉz©¹  ÀÇÇ'77—•¿fÍš•+W’´š¿Fèi/üÿbÑâ-ðlE&“õéÓ'++‹™)‹9òæ›o*­AMÍêݹsgxx8w‘]‘H´~ýú%K–(0//¯_¿~?æb¶Âsþ+°'OžüñÇÜ’}ûöMNNnÞ¼9+_ëîë]çÎoß¾M¿T("‘È퀥Á|0 –-[¦§§¯Zµªk×®b±ØÆÆfÀ€‡ŠŠŠ2uhÆ`mm}ñâÅ ´nÝÚÊÊÊÉÉi„ W¯^3fŒÞÛš={ö½{÷bbb¼½½œœÄb±‹‹ËÔ©S¯]»¦jP†¢(77·7n,_¾¼OŸ>úzpvvþý÷ßããã}||œœœÈ;àãã³}ûöÔÔTî  €Â|0à|€z óeÀ80_À4¬L˜7nX[[›: S*//7u`0.l+V¬0uÏ1˜ÆeLã2¦ñÿïªz»ËM0CIEND®B`‚flexmix/vignettes/flexmix-intro.Rnw0000644000176200001440000010260313430477461017246 0ustar liggesusers% % Copyright (C) 2004-2005 Friedrich Leisch % $Id: flexmix-intro.Rnw 5156 2019-02-12 08:11:16Z gruen $ % \documentclass[nojss]{jss} \title{FlexMix: A General Framework for Finite Mixture Models and Latent Class Regression in \proglang{R}} \Plaintitle{FlexMix: A General Framework for Finite Mixture Models and Latent Class Regression in R} \Shorttitle{FlexMix: Finite Mixture Models in \proglang{R}} \author{Friedrich Leisch\\Universit\"at f\"ur Bodenkultur Wien} \Plainauthor{Friedrich Leisch} \Address{ Friedrich Leisch\\ Institut f\"ur Angewandte Statistik und EDV\\ Universit\"at f\"ur Bodenkultur Wien\\ Peter Jordan Stra\ss{}e 82\\ 1190 Wien, Austria\\ E-mail: \email{Friedrich.Leisch@boku.ac.at}\\ URL: \url{http://www.statistik.lmu.de/~leisch/} } \usepackage[utf8]{inputenc} \usepackage{listings} \newcommand{\R}{\proglang{R}} <>= set.seed(1504) options(width=70, prompt = "R> ", continue = "+ ", useFancyQuotes = FALSE) grDevices::ps.options(family="Times") library("graphics") library("flexmix") data("NPreg") @ \Abstract{ This article was originally published as \cite{flexmix:Leisch:2004a} in the \emph{Journal of Statistical Software}. FlexMix implements a general framework for fitting discrete mixtures of regression models in the \R{} statistical computing environment: three variants of the EM algorithm can be used for parameter estimation, regressors and responses may be multivariate with arbitrary dimension, data may be grouped, e.g., to account for multiple observations per individual, the usual formula interface of the \proglang{S} language is used for convenient model specification, and a modular concept of driver functions allows to interface many different types of regression models. Existing drivers implement mixtures of standard linear models, generalized linear models and model-based clustering. FlexMix provides the E-step and all data handling, while the M-step can be supplied by the user to easily define new models. } \Keywords{\proglang{R}, finite mixture models, model based clustering, latent class regression} \Plainkeywords{R, finite mixture models, model based clustering, latent class regression} \Volume{11} \Issue{8} \Month{October} \Year{2004} \Submitdate{2004-04-19} \Acceptdate{2004-10-18} %%\usepackage{Sweave} %% already provided by jss.cls %%\VignetteIndexEntry{FlexMix: A General Framework for Finite Mixture Models and Latent Class Regression in R} %%\VignetteDepends{flexmix} %%\VignetteKeywords{R, finite mixture models, model based clustering, latent class regression} %%\VignettePackage{flexmix} \begin{document} \section{Introduction} \label{sec:introduction} Finite mixture models have been used for more than 100 years, but have seen a real boost in popularity over the last decade due to the tremendous increase in available computing power. The areas of application of mixture models range from biology and medicine to physics, economics and marketing. On the one hand these models can be applied to data where observations originate from various groups and the group affiliations are not known, and on the other hand to provide approximations for multi-modal distributions \citep{flexmix:Everitt+Hand:1981,flexmix:Titterington+Smith+Makov:1985,flexmix:McLachlan+Peel:2000}. In the 1990s finite mixture models have been extended by mixing standard linear regression models as well as generalized linear models \citep{flexmix:Wedel+DeSarbo:1995}. An important area of application of mixture models is market segmentation \citep{flexmix:Wedel+Kamakura:2001}, where finite mixture models replace more traditional cluster analysis and cluster-wise regression techniques as state of the art. Finite mixture models with a fixed number of components are usually estimated with the expectation-maximization (EM) algorithm within a maximum likelihood framework \citep{flexmix:Dempster+Laird+Rubin:1977} and with MCMC sampling \citep{flexmix:Diebolt+Robert:1994} within a Bayesian framework. \newpage The \R{} environment for statistical computing \citep{flexmix:R-Core:2004} features several packages for finite mixture models, including \pkg{mclust} for mixtures of multivariate Gaussian distributions \citep{flexmix:Fraley+Raftery:2002,flexmix:Fraley+Raftery:2002a}, \pkg{fpc} for mixtures of linear regression models \citep{flexmix:Hennig:2000} and \pkg{mmlcr} for mixed-mode latent class regression \citep{flexmix:Buyske:2003}. There are three main reasons why we have chosen to write yet another software package for EM estimation of mixture models: \begin{itemize} \item The existing implementations did not cover all cases we needed for our own research (mainly marketing applications). \item While all \R{} packages mentioned above are open source and hence can be extended by the user by modifying the source code, we wanted an implementation where extensibility is a main design principle to enable rapid prototyping of new mixture models. \item We include a sampling-based variant of the EM-algorithm for models where weighted maximum likelihood estimation is not available. FlexMix has a clean interface between E- and M-step such that variations of both are easy to combine. \end{itemize} This paper is organized as follows: First we introduce the mathematical models for latent class regression in Section~\ref{sec:latent-class-regr} and shortly discuss parameter estimation and identifiability. Section~\ref{sec:using-flexmix} demonstrates how to use FlexMix to fit models with the standard driver for generalized linear models. Finally, Section~\ref{sec:extending-flexmix} shows how to extend FlexMix by writing new drivers using the well-known model-based clustering procedure as an example. \section{Latent class regression} \label{sec:latent-class-regr} Consider finite mixture models with $K$ components of form \begin{equation}\label{eq:1} h(y|x,\psi) = \sum_{k = 1}^K \pi_k f(y|x,\theta_k) \end{equation} \begin{displaymath} \pi_k \geq 0, \quad \sum_{k = 1}^K \pi_k = 1 \end{displaymath} where $y$ is a (possibly multivariate) dependent variable with conditional density $h$, $x$ is a vector of independent variables, $\pi_k$ is the prior probability of component $k$, $\theta_k$ is the component specific parameter vector for the density function $f$, and $\psi=(\pi_1,,\ldots,\pi_K,\theta_1',\ldots,\theta_K')'$ is the vector of all parameters. If $f$ is a univariate normal density with component-specific mean $\beta_k'x$ and variance $\sigma^2_k$, we have $\theta_k = (\beta_k', \sigma_k^2)'$ and Equation~(\ref{eq:1}) describes a mixture of standard linear regression models, also called \emph{latent class regression} or \emph{cluster-wise regression} \citep{flexmix:DeSarbo+Cron:1988}. If $f$ is a member of the exponential family, we get a mixture of generalized linear models \citep{flexmix:Wedel+DeSarbo:1995}, known as \emph{GLIMMIX} models in the marketing literature \citep{flexmix:Wedel+Kamakura:2001}. For multivariate normal $f$ and $x\equiv1$ we get a mixture of Gaussians without a regression part, also known as \emph{model-based clustering}. The posterior probability that observation $(x,y)$ belongs to class $j$ is given by \begin{equation}\label{eq:3} \Prob(j|x, y, \psi) = \frac{\pi_j f(y | x, \theta_j)}{\sum_k \pi_k f(y | x, \theta_k)} \end{equation} The posterior probabilities can be used to segment data by assigning each observation to the class with maximum posterior probability. In the following we will refer to $f(\cdot|\cdot, \theta_k)$ as \emph{mixture components} or \emph{classes}, and the groups in the data induced by these components as \emph{clusters}. \subsection{Parameter estimation} \label{sec:parameter-estimation} The log-likelihood of a sample of $N$ observations $\{(x_1,y_1),\ldots,(x_N,y_N)\}$ is given by \begin{equation}\label{eq:4} \log L = \sum_{n=1}^N \log h(y_n|x_n,\psi) = \sum_{n=1}^N \log\left(\sum_{k = 1}^K \pi_kf(y_n|x_n,\theta_k) \right) \end{equation} and can usually not be maximized directly. The most popular method for maximum likelihood estimation of the parameter vector $\psi$ is the iterative EM algorithm \citep{flexmix:Dempster+Laird+Rubin:1977}: \begin{description} \item[Estimate] the posterior class probabilities for each observation \begin{displaymath} \hat p_{nk} = \Prob(k|x_n, y_n, \hat \psi) \end{displaymath} using Equation~(\ref{eq:3}) and derive the prior class probabilities as \begin{displaymath} \hat\pi_k = \frac1N \sum_{n=1}^N \hat p_{nk} \end{displaymath} \item[Maximize] the log-likelihood for each component separately using the posterior probabilities as weights \begin{equation}\label{eq:2} \max_{\theta_k} \sum_{n=1}^N \hat p_{nk} \log f(y_n | x_n, \theta_k) \end{equation} \end{description} The E- and M-steps are repeated until the likelihood improvement falls under a pre-specified threshold or a maximum number of iterations is reached. The EM algorithm cannot be used for mixture models only, but rather provides a general framework for fitting models on incomplete data. Suppose we augment each observation $(x_n,y_n)$ with an unobserved multinomial variable $z_n = (z_{n1},\ldots,z_{nK})$, where $z_{nk}=1$ if $(x_n,y_n)$ belongs to class $k$ and $z_{nk}=0$ otherwise. The EM algorithm can be shown to maximize the likelihood on the ``complete data'' $(x_n,y_n,z_n)$; the $z_n$ encode the missing class information. If the $z_n$ were known, maximum likelihood estimation of all parameters would be easy, as we could separate the data set into the $K$ classes and estimate the parameters $\theta_k$ for each class independently from the other classes. If the weighted likelihood estimation in Equation~(\ref{eq:2}) is infeasible for analytical, computational, or other reasons, then we have to resort to approximations of the true EM procedure by assigning the observations to disjoint classes and do unweighted estimation within the groups: \begin{displaymath} \max_{\theta_k} \sum_{n: z_{nk=1}} \log f(y_n | x_n, \theta_k) \end{displaymath} This corresponds to allow only 0 and 1 as weights. Possible ways of assigning the data into the $K$ classes are \begin{itemize} \item \textbf{hard} \label{hard} assignment to the class with maximum posterior probability $p_{nk}$, the resulting procedure is called maximizing the \emph{classification likelihood} by \cite{flexmix:Fraley+Raftery:2002}. Another idea is to do \item \textbf{random} assignment to classes with probabilities $p_{nk}$, which is similar to the sampling techniques used in Bayesian estimation (although for the $z_n$ only). \end{itemize} Well known limitations of the EM algorithm include that convergence can be slow and is to a local maximum of the likelihood surface only. There can also be numerical instabilities at the margin of parameter space, and if a component gets to contain only a few observations during the iterations, parameter estimation in the respective component may be problematic. E.g., the likelihood of Gaussians increases without bounds for $\sigma^2\to 0$. As a result, numerous variations of the basic EM algorithm described above exist, most of them exploiting features of special cases for $f$. \subsection{Identifiability} \label{sec:identifiability} An open question is still identifiability of many mixture models. A comprehensive overview of this topic is beyond the scope of this paper, however, users of mixture models should be aware of the problem: \begin{description} \item[Relabelling of components:] Mixture models are only identifiable up to a permutation of the component labels. For EM-based approaches this only affects interpretation of results, but is no problem for parameter estimation itself. \item[Overfitting:] If a component is empty or two or more components have the same parameters, the data generating process can be represented by a smaller model with fewer components. This kind of unidentifiability can be avoided by requiring that the prior weights $\pi_k$ are not equal to zero and that the component specific parameters are different. \item[Generic unidentifiability:] It has been shown that mixtures of univariate normal, gamma, exponential, Cauchy and Poisson distributions are identifiable, while mixtures of discrete or continuous uniform distributions are not identifiable. A special case is the class of mixtures of binomial and multinomial distributions which are only identifiable if the number of components is limited with respect to, e.g., the number of observations per person. See \cite{flexmix:Everitt+Hand:1981}, \cite{flexmix:Titterington+Smith+Makov:1985}, \cite{flexmix:Grun:2002} and references therein for details. \end{description} FlexMix tries to avoid overfitting because of vanishing prior probabilities by automatically removing components where the prior $\pi_k$ falls below a user-specified threshold. Automated diagnostics for generic identifiability are currently under investigation. Relabelling of components is in some cases more of a nuisance than a real problem (``component 2 of the first run may be component 3 in the second run''), more serious are interactions of component relabelling and categorical predictor variables, see \cite{flexmix:Grun+Leisch:2004} for a discussion and how bootstrapping can be used to assess identifiability of mixture models. \pagebreak[4] \section{Using FlexMix} \label{sec:using-flexmix} \SweaveOpts{width=12,height=8,eps=FALSE,keep.source=TRUE} The standard M-step \texttt{FLXMRglm()} of FlexMix is an interface to R's generalized linear modelling facilities (the \texttt{glm()} function). As a simple example we use artificial data with two latent classes of size \Sexpr{nrow(NPreg)/2} each: \begin{center} \begin{tabular}{ll} Class~1: & $ y = 5x+\epsilon$\\ Class~2: & $ y = 15+10x-x^2+\epsilon$\\ \end{tabular} \end{center} with $\epsilon\sim N(0,9)$ and prior class probabilities $\pi_1=\pi_2=0.5$, see the left panel of Figure~\ref{fig:npreg}. We can fit this model in \R{} using the commands <<>>= library("flexmix") data("NPreg") m1 <- flexmix(yn ~ x + I(x^2), data = NPreg, k = 2) m1 @ and get a first look at the estimated parameters of mixture component~1 by <<>>= parameters(m1, component = 1) @ and <<>>= parameters(m1, component = 2) @ for component~2. The paramter estimates of both components are close to the true values. A cross-tabulation of true classes and cluster memberships can be obtained by <<>>= table(NPreg$class, clusters(m1)) @ The summary method <<>>= summary(m1) @ gives the estimated prior probabilities $\hat\pi_k$, the number of observations assigned to the corresponding clusters, the number of observations where $p_{nk}>\delta$ (with a default of $\delta=10^{-4}$), and the ratio of the latter two numbers. For well-seperated components, a large proportion of observations with non-vanishing posteriors $p_{nk}$ should also be assigned to the corresponding cluster, giving a ratio close to 1. For our example data the ratios of both components are approximately 0.7, indicating the overlap of the classes at the cross-section of line and parabola. \begin{figure}[htbp] \centering <>= par(mfrow=c(1,2)) plot(yn~x, col=class, pch=class, data=NPreg) plot(yp~x, col=class, pch=class, data=NPreg) @ \caption{Standard regression example (left) and Poisson regression (right).} \label{fig:npreg} \end{figure} Histograms or rootograms of the posterior class probabilities can be used to visually assess the cluster structure \citep{flexmix:Tantrum+Murua+Stuetzle:2003}, this is now the default plot method for \texttt{"flexmix"} objects \citep{flexmix:Leisch:2004}. Rootograms are very similar to histograms, the only difference is that the height of the bars correspond to square roots of counts rather than the counts themselves, hence low counts are more visible and peaks less emphasized. \begin{figure}[htbp] \centering <>= print(plot(m1)) @ \caption{The plot method for \texttt{"flexmix"} objects, here obtained by \texttt{plot(m1)}, shows rootograms of the posterior class probabilities.} \label{fig:root1} \end{figure} Usually in each component a lot of observations have posteriors close to zero, resulting in a high count for the corresponing bin in the rootogram which obscures the information in the other bins. To avoid this problem, all probabilities with a posterior below a threshold are ignored (we again use $10^{-4}$). A peak at probability 1 indicates that a mixture component is well seperated from the other components, while no peak at 1 and/or significant mass in the middle of the unit interval indicates overlap with other components. In our simple example the components are medium well separated, see Figure~\ref{fig:root1}. Tests for significance of regression coefficients can be obtained by <<>>= rm1 <- refit(m1) summary(rm1) @ Function \texttt{refit()} fits weighted generalized linear models to each component using the standard \R{} function \texttt{glm()} and the posterior probabilities as weights, see \texttt{help("refit")} for details. The data set \texttt{NPreg} also includes a response from a generalized linear model with a Poisson distribution and exponential link function. The two classes of size \Sexpr{nrow(NPreg)/2} each have parameters \begin{center} \begin{tabular}{ll} Class~1: & $ \mu_1 = 2 - 0.2x$\\ Class~2: & $ \mu_2 = 1 + 0.1x$\\ \end{tabular} \end{center} and given $x$ the response $y$ in group $k$ has a Poisson distribution with mean $e^{\mu_k}$, see the right panel of Figure~\ref{fig:npreg}. The model can be estimated using <>= options(width=55) @ <<>>= m2 <- flexmix(yp ~ x, data = NPreg, k = 2, model = FLXMRglm(family = "poisson")) summary(m2) @ <>= options(width=65) @ \begin{figure}[htbp] \centering <>= print(plot(m2)) @ \caption{\texttt{plot(m2)}} \label{fig:root2} \end{figure} Both the summary table and the rootograms in Figure~\ref{fig:root2} clearly show that the clusters of the Poisson response have much more overlap. For our simple low-dimensional example data the overlap of the classes is obvious by looking at scatterplots of the data. For data in higher dimensions this is not an option. The rootograms and summary tables for \texttt{"flexmix"} objects work off the densities or posterior probabilities of the observations and thus do not depend on the dimensionality of the input space. While we use simple 2-dimensional examples to demonstrate the techniques, they can easily be used on high-dimensional data sets or models with complicated covariate structures. \subsection{Multiple independent responses} \label{sec:mult-indep-resp} If the response $y=(y_1,\ldots,y_D)'$ is $D$-dimensional and the $y_d$ are mutually independent the mixture density in Equation~(\ref{eq:1}) can be written as \begin{eqnarray*} h(y|x,\psi) &=& \sum_{k = 1}^K \pi_k f(y|x,\theta_k)\\ &=& \sum_{k = 1}^K \pi_k \prod_{d=1}^D f_d(y|x,\theta_{kd}) \end{eqnarray*} To specify such models in FlexMix we pass it a list of models, where each list element corresponds to one $f_d$, and each can have a different set of dependent and independent variables. To use the Gaussian and Poisson responses of data \texttt{NPreg} simultaneously, we use the model specification \begin{Sinput} > m3 = flexmix(~x, data=NPreg, k=2, + model=list(FLXMRglm(yn~.+I(x^2)), + FLXMRglm(yp~., family="poisson"))) \end{Sinput} <>= m3 <- flexmix(~ x, data = NPreg, k = 2, model=list(FLXMRglm(yn ~ . + I(x^2)), FLXMRglm(yp ~ ., family = "poisson"))) @ Note that now three model formulas are involved: An overall formula as first argument to function \texttt{flexmix()} and one formula per response. The latter ones are interpreted relative to the overall formula such that common predictors have to be specified only once, see \texttt{help("update.formula")} for details on the syntax. The basic principle is that the dots get replaced by the respective terms from the overall formula. The rootograms show that the posteriors of the two-response model are shifted towards 0 and 1 (compared with either of the two univariate models), the clusters are now well-separated. \begin{figure}[htbp] \centering <>= print(plot(m3)) @ \caption{\texttt{plot(m3)}} \label{fig:root3} \end{figure} \subsection{Repeated measurements} \label{sec:repe-meas} If the data are repeated measurements on $M$ individuals, and we have $N_m$ observations from individual $m$, then the log-likelihood in Equation~(\ref{eq:4}) can be written as \begin{displaymath} \log L = \sum_{m=1}^M \sum_{n=1}^{N_m} \log h(y_{mn}|x_{mn},\psi), \qquad \sum_{m=1}^M N_m = N \end{displaymath} and the posterior probability that individual $m$ belongs to class $j$ is given by \begin{displaymath} \Prob(j|m) = \frac{\pi_j \prod_{n=1}^{N_m} f(y_{mn} | x_{mn}, \theta_j)}{\sum_k \pi_k \prod_{n=1}^{N_m} f(y_{mn} | x_{mn}, \theta_k)} \end{displaymath} where $(x_{mn}, y_{mn})$ is the $n$-th observation from individual $m$. As an example, assume that the data in \texttt{NPreg} are not 200 independent observations, but 4 measurements each from 50 persons such that $\forall m: N_m=4$. Column \texttt{id2} of the data frame encodes such a grouping and can easily be used in FlexMix: <<>>= m4 <- flexmix(yn ~ x + I(x^2) | id2, data = NPreg, k = 2) summary(m4) @ Note that convergence of the EM algorithm is much faster with grouping and the two clusters are now perfectly separated. \subsection{Control of the EM algorithm} \label{sec:control-em-algorithm} Details of the EM algorithm can be tuned using the \texttt{control} argument of function \texttt{flexmix()}. E.g., to use a maximum number of 15 iterations, report the log-likelihood at every 3rd step and use hard assignment of observations to clusters (cf. page~\pageref{hard}) the call is <<>>= m5 <- flexmix(yn ~ x + I(x^2), data = NPreg, k = 2, control = list(iter.max = 15, verbose = 3, classify = "hard")) @ Another control parameter (\texttt{minprior}, see below for an example) is the minimum prior probability components are enforced to have, components falling below this threshold (the current default is 0.05) are removed during EM iteration to avoid numerical instabilities for components containing only a few observations. Using a minimum prior of 0 disables component removal. \subsection{Automated model search} In real applications the number of components is unknown and has to be estimated. Tuning the minimum prior parameter allows for simplistic model selection, which works surprisingly well in some situations: <<>>= m6 <- flexmix(yp ~ x + I(x^2), data = NPreg, k = 4, control = list(minprior = 0.2)) m6 @ Although we started with four components, the algorithm converged at the correct two component solution. A better approach is to fit models with an increasing number of components and compare them using AIC or BIC. As the EM algorithm converges only to the next local maximum of the likelihood, it should be run repeatedly using different starting values. The function \texttt{stepFlexmix()} can be used to repeatedly fit models, e.g., <<>>= m7 <- stepFlexmix(yp ~ x + I(x^2), data = NPreg, control = list(verbose = 0), k = 1:5, nrep = 5) @ runs \texttt{flexmix()} 5 times for $k=1,2,\ldots,5$ components, totalling in 25 runs. It returns a list with the best solution found for each number of components, each list element is simply an object of class \texttt{"flexmix"}. To find the best model we can use <<>>= getModel(m7, "BIC") @ and choose the number of components minimizing the BIC. \section{Extending FlexMix} \label{sec:extending-flexmix} One of the main design principles of FlexMix was extensibility, users can provide their own M-step for rapid prototyping of new mixture models. FlexMix was written using S4 classes and methods \citep{flexmix:Chambers:1998} as implemented in \R{} package \pkg{methods}. The central classes for writing M-steps are \texttt{"FLXM"} and \texttt{"FLXcomponent"}. Class \texttt{"FLXM"} specifies how the model is fitted using the following slots: \begin{description} \item[fit:] A \texttt{function(x,y,w)} returning an object of class \texttt{"FLXcomponent"}. \item[defineComponent:] Expression or function constructing the object of class \texttt{"FLXcomponent"}. \item[weighted:] Logical, specifies if the model may be fitted using weighted likelihoods. If \texttt{FALSE}, only hard and random classification are allowed (and hard classification becomes the default). \item[formula:] Formula relative to the overall model formula, default is \verb|.~.| \item[name:] A character string describing the model, this is only used for print output. \end{description} The remaining slots of class \texttt{"FLXM"} are used internally by FlexMix to hold data, etc. and omitted here, because they are not needed to write an M-step driver. The most important slot doing all the work is \texttt{fit} holding a function performing the maximum likelihood estimation described in Equation~(\ref{eq:2}). The \texttt{fit()} function returns an object of class \texttt{"FLXcomponent"} which holds a fitted component using the slots: \begin{description} \item[logLik:] A \texttt{function(x,y)} returning the log-likelihood for observations in matrices \texttt{x} and \texttt{y}. \item[predict:] A \texttt{function(x)} predicting \texttt{y} given \texttt{x}. \item[df:] The degrees of freedom used by the component, i.e., the number of estimated parameters. \item[parameters:] An optional list containing model parameters. \end{description} In a nutshell class \texttt{"FLXM"} describes an \emph{unfitted} model, whereas class \texttt{"FLXcomponent"} holds a \emph{fitted} model. \lstset{frame=trbl,basicstyle=\small\tt,stepnumber=5,numbers=left} \begin{figure}[tb] \centering \begin{minipage}{0.94\textwidth} \lstinputlisting{mymclust.R} \end{minipage} \caption{M-step for model-based clustering: \texttt{mymclust} is a simplified version of the standard FlexMix driver \texttt{FLXmclust}.} \label{fig:mymclust.R} \end{figure} \subsection{Writing an M-step driver} \label{sec:writing-an-m} Figure~\ref{fig:mymclust.R} shows an example driver for model-based clustering. We use function \texttt{dmvnorm()} from package \pkg{mvtnorm} for calculation of multivariate Gaussian densities. In line~5 we create a new \texttt{"FLXMC"} object named \texttt{retval}, which is also the return value of the driver. Class \texttt{"FLXMC"} extends \texttt{"FLXM"} and is used for model-based clustering. It contains an additional slot with the name of the distribution used. All drivers should take a formula as their first argument, this formula is directly passed on to \texttt{retval}. In most cases authors of new FlexMix drivers need not worry about formula parsing etc., this is done by \texttt{flexmix} itself. In addition we have to declare whether the driver can do weighted ML estimation (\texttt{weighted=TRUE}) and give a name to our model. The remainder of the driver creates a \texttt{fit()} function, which takes regressors \texttt{x}, response \texttt{y} and weights \texttt{w}. For multivariate Gaussians the maximum likelihood estimates correspond to mean and covariance matrix, the standard R function \texttt{cov.wt()} returns a list containing estimates of the weighted covariance matrix and the mean for given data. Our simple example performs clustering without a regression part, hence $x$ is ignored. If \texttt{y} has $D$ columns, we estimate $D$ parameters for the mean and $D(D-1)/2$ parameters for the covariance matrix, giving a total of $(3D+D^2)/2$ parameters (line~11). As an additional feature we allow the user to specify whether the covariance matrix is assumed to be diagonal or a full matrix. For \texttt{diagonal=TRUE} we use only the main diagonal of the covariance matrix (line~14) and the number of parameters is reduced to $2D$. In addition to parameter estimates, \texttt{flexmix()} needs a function calculating the log-likelihood of given data $x$ and $y$, which in our example is the log-density of a multivariate Gaussian. In addition we have to provide a function predicting $y$ given $x$, in our example simply the mean of the Gaussian. Finally we create a new \texttt{"FLXcomponent"} as return value of function \texttt{fit()}. Note that our internal functions \texttt{fit()}, \texttt{logLik()} and \texttt{predict()} take only \texttt{x}, \texttt{y} and \texttt{w} as arguments, but none of the model-specific parameters like means and covariances, although they use them of course. \R{} uses \emph{lexical scoping} rules for finding free variables \citep{flexmix:Gentleman+Ihaka:2000}, hence it searches for them first in the environment where a function is defined. E.g., the \texttt{fit()} function uses the variable \texttt{diagonal} in line~24, and finds it in the environment where the function itself was defined, which is the body of function \texttt{mymclust()}. Function \texttt{logLik()} uses the list \texttt{para} in lines~8 and 9, and uses the one found in the body of \texttt{defineComponent()}. Function \texttt{flexmix()} on the other hand never sees the model parameters, all it uses are function calls of form \texttt{fit(x,y,w)} or \texttt{logLik(x,y)}, which are exactly the same for all kinds of mixture models. In fact, it would not be necessary to even store the component parameters in the \texttt{"FLXcomponent"} object, they are there only for convenience such that users can easily extract and use them after \texttt{flexmix()} has finished. Lexical scope allows to write clean interfaces in a very elegant way, the driver abstracts all model details from the FlexMix main engine. \subsection{Example: Using the driver} \label{sec:example:-model-based} \SweaveOpts{width=12,height=6,eps=FALSE} <>= library("flexmix") set.seed(1504) options(width=60) grDevices::ps.options(family="Times") suppressMessages(require("ellipse")) suppressMessages(require("mvtnorm")) source("mymclust.R") @ As a simple example we use the four 2-dimensional Gaussian clusters from data set \texttt{Nclus}. Fitting a wrong model with diagonal covariance matrix is done by <<>>= data("Nclus") m1 <- flexmix(Nclus ~ 1, k = 4, model = mymclust()) summary(m1) @ The result can be seen in the left panel of Figure~\ref{fig:ell}, the result is ``wrong'' because we forced the ellipses to be parallel to the axes. The overlap between three of the four clusters can also be inferred from the low ratio statistics in the summary table (around 0.5 for components 1, 3 and 4), while the much better separated upper left cluster has a much higher ratio of 0.85. Using the correct model with a full covariance matrix can be done by setting \texttt{diagonal=FALSE} in the call to our driver \texttt{mymclust()}: <<>>= m2 <- flexmix(Nclus ~ 1, k = 4, model = mymclust(diagonal = FALSE)) summary(m2) @ \begin{figure}[htbp] \centering <>= par(mfrow=1:2) plotEll(m1, Nclus) plotEll(m2, Nclus) @ \caption{Fitting a mixture model with diagonal covariance matrix (left) and full covariance matrix (right).} \label{fig:ell} \end{figure} \pagebreak[4] \section{Summary and outlook} \label{sec:summary} The primary goal of FlexMix is extensibility, this makes the package ideal for rapid development of new mixture models. There is no intent to replace packages implementing more specialized mixture models like \pkg{mclust} for mixtures of Gaussians, FlexMix should rather be seen as a complement to those. By interfacing R's facilities for generalized linear models, FlexMix allows the user to estimate complex latent class regression models. Using lexical scope to resolve model-specific parameters hides all model details from the programming interface, FlexMix can in principle fit almost arbitrary finite mixture models for which the EM algorithm is applicable. The downside of this is that FlexMix can in principle fit almost arbitrary finite mixture models, even models where no proper theoretical results for model identification etc.\ are available. We are currently working on a toolset for diagnostic checks on mixture models to test necessary identifiability conditions for those cases where results are available. We also want to implement newer variations of the classic EM algorithm, especially for faster convergence. Another plan is to have an interactive version of the rootograms using \texttt{iPlots} \citep{flexmix:Urbanek+Theus:2003} such that the user can explore the relations between mixture components, possibly linked to background variables. Other planned extensions include covariates for the prior probabilities and to allow to mix different distributions for components, e.g., to include a Poisson point process for background noise. \section*{Computational details} <>= SI <- sessionInfo() pkgs <- paste(sapply(c(SI$otherPkgs, SI$loadedOnly), function(x) paste("\\\\pkg{", x$Package, "} ", x$Version, sep = "")), collapse = ", ") @ All computations and graphics in this paper have been done using \proglang{R} version \Sexpr{getRversion()} with the packages \Sexpr{pkgs}. \section*{Acknowledgments} This research was supported by the Austrian Science Foundation (FWF) under grant SFB\#010 (`Adaptive Information Systems and Modeling in Economics and Management Science'). Bettina Gr\"un has modified the original version to include and reflect the changes of the package. \bibliography{flexmix} \end{document} %%% Local Variables: %%% mode: latex %%% TeX-master: t %%% End: flexmix/vignettes/myConcomitant.R0000644000176200001440000000130313425024236016703 0ustar liggesusersmyConcomitant <- function(formula = ~ 1) { z <- new("FLXP", name = "myConcomitant", formula = formula) z@fit <- function(x, y, w, ...) { if (missing(w) || is.null(w)) w <- rep(1, length(x)) f <- as.integer(factor(apply(x, 1, paste, collapse = ""))) AVG <- apply(w*y, 2, tapply, f, mean) (AVG/rowSums(AVG))[f,,drop=FALSE] } z@refit <- function(x, y, w, ...) { if (missing(w) || is.null(w)) w <- rep(1, length(x)) f <- as.integer(factor(apply(x, 1, paste, collapse = ""))) AVG <- apply(w*y, 2, tapply, f, mean) (AVG/rowSums(AVG)) } z } flexmix/vignettes/flexmix.bib0000644000176200001440000002336513425024236016102 0ustar liggesusers@STRING{jcgs = {Journal of Computational and Graphical Statistics} } @STRING{tuwien = {Technische Universit{\"a}t Wien, Vienna, Austria} } @STRING{jasa = {Journal of the American Statistical Association} } @Article{ flexmix:Aitkin:1996, author = {Murray Aitkin}, title = {A General Maximum Likelihood Analysis of Overdispersion in Generalized Linear Models}, journal = {Statistics and Computing}, year = 1996, volume = 6, pages = {251--262} } @Article{ flexmix:Aitkin:1999, author = {Murray Aitkin}, title = {A General Maximum Likelihood Analysis of Variance Components in Generalized Linear Models}, journal = {Biometrics}, year = 1999, volume = 55, pages = {117--128} } @Article{ flexmix:Aitkin:1999a, author = {Murray Aitkin}, title = {Meta-Analysis by Random Effect Modelling in Generalized Linear Models}, journal = {Statistics in Medicine}, year = 1999, volume = 18, number = {17--18}, month = {September}, pages = {2343--2351} } @Manual{ flexmix:Buyske:2003, title = {{R} Package \texttt{mmlcr}: Mixed-Mode Latent Class Regression}, author = {Steve Buyske}, year = 2003, note = {version 1.3.2}, url = {http://www.stat.rutgers.edu/~buyske/software.html} } @Book{ flexmix:Chambers:1998, author = {John M. Chambers}, title = {Programming with Data: A Guide to the {S} Language}, publisher = {Springer Verlag}, year = 1998, address = {Berlin, Germany} } @Article{ flexmix:DeSarbo+Cron:1988, author = {Wayne S. DeSarbo and W. L. Cron}, title = {A Maximum Likelihood Methodology for Clusterwise Linear Regression}, journal = {Journal of Classification}, year = 1988, volume = 5, pages = {249--282} } @Article{ flexmix:Dempster+Laird+Rubin:1977, author = {A.P. Dempster and N.M. Laird and D.B. Rubin}, title = {Maximum Likelihood from Incomplete Data via the {EM}-Alogrithm}, journal = {Journal of the Royal Statistical Society, B}, volume = 39, pages = {1--38}, year = 1977 } @Article{ flexmix:Diebolt+Robert:1994, author = {J. Diebolt and C. P. Robert}, title = {Estimation of Finite Mixture Distributions Through {B}ayesian Sampling}, journal = {Journal of the Royal Statistical Society, Series B}, year = 1994, volume = 56, pages = {363--375} } @Book{ flexmix:Everitt+Hand:1981, author = {Brian S. Everitt and David J. Hand}, title = {Finite Mixture Distributions}, publisher = {Chapman and Hall}, address = {London}, year = 1981 } @Article{ flexmix:Follmann+Lambert:1989, author = {Dean A. Follmann and Diane Lambert}, title = {Generalizing Logistic Regression by Non-Parametric Mixing}, journal = jasa, volume = 84, number = 405, month = {March}, pages = {295--300}, year = 1989 } @Article{ flexmix:Fraley+Raftery:2002, author = {Chris Fraley and Adrian E. Raftery}, title = {Model-Based Clustering, Discriminant Analysis and dDnsity Estimation}, journal = jasa, year = 2002, volume = 97, pages = {611-631} } @TechReport{ flexmix:Fraley+Raftery:2002a, author = {Chris Fraley and Adrian E. Raftery}, title = {{MCLUST}: Software for Model-Based Clustering, Discriminant Analysis and Density Estimation}, institution = {Department of Statistics, University of Washington}, year = 2002, number = 415, address = {Seattle, WA, USA}, url = {http://www.stat.washington.edu/raftery} } @Article{ flexmix:Gentleman+Ihaka:2000, author = {Robert Gentleman and Ross Ihaka}, title = {Lexical Scope and Statistical Computing}, journal = jcgs, year = 2000, volume = 9, number = 3, pages = {491--508}, keywords = {statistical computing, function closure, lexical scope, random number generators} } @InProceedings{ flexmix:Gruen+Leisch:2006, author = {Bettina Gr{\" u}n and Friedrich Leisch}, title = {Fitting Finite Mixtures of Linear Regression Models with Varying \& Fixed Effects in \textsf{R}}, booktitle = {Compstat 2006---Proceedings in Computational Statistics}, pages = {853--860}, editor = {Alfredo Rizzi and Maurizio Vichi}, publisher = {Physica Verlag}, address = {Heidelberg, Germany}, isbn = {3-7908-1708-2}, year = 2006 } @InProceedings{ flexmix:Grun+Leisch:2004, author = {Bettina Gr{\" u}n and Friedrich Leisch}, title = {Bootstrapping Finite Mixture Models}, booktitle = {Compstat 2004---Proceedings in Computational Statistics}, year = 2004, editor = {Jaromir Antoch}, publisher = {Physica Verlag}, address = {Heidelberg, Germany}, isbn = {3-7908-1554-3}, pages = {1115--1122}, pdf = {http://www.stat.uni-muenchen.de/~leisch/papers/Grun+Leisch-2004.pdf} } @MastersThesis{ flexmix:Grun:2002, author = {Bettina Gr{\"u}n}, title = {{I}dentifizierbarkeit von multinomialen {M}ischmodellen}, school = tuwien, year = 2002, note = {Kurt Hornik and Friedrich Leisch, advisors} } @Article{ flexmix:Hennig:2000, author = {Christian Hennig}, title = {Identifiability of Models for Clusterwise Linear Regression}, journal = {Journal of Classification}, volume = 17, pages = {273--296}, year = 2000 } @InProceedings{ flexmix:Leisch:2004, author = {Friedrich Leisch}, title = {Exploring the Structure of Mixture Model Components}, booktitle = {Compstat 2004---Proceedings in Computational Statistics}, year = 2004, editor = {Jaromir Antoch}, publisher = {Physica Verlag}, address = {Heidelberg, Germany}, isbn = {3-7908-1554-3}, pages = {1405--1412}, pdf = {http://www.stat.uni-muenchen.de/~leisch/papers/Leisch-2004.pdf} } @Article{ flexmix:Leisch:2004a, author = {Friedrich Leisch}, title = {{FlexMix}: A General Framework for Finite Mixture Models and Latent Class Regression in {R}}, journal = {Journal of Statistical Software}, year = 2004, volume = 11, number = 8, url = {http://www.jstatsoft.org/v11/i08/} } @Book{ flexmix:McLachlan+Peel:2000, author = {Geoffrey McLachlan and David Peel}, title = {Finite Mixture Models}, publisher = {John Wiley and Sons Inc.}, year = 2000 } @Manual{ flexmix:R-Core:2004, title = {R: A Language and Environment for Statistical Computing}, author = {{R Development Core Team}}, organization = {R Foundation for Statistical Computing}, address = {Vienna, Austria}, year = 2004, isbn = {3-900051-07-0}, url = {http://www.R-project.org} } @InProceedings{ flexmix:Tantrum+Murua+Stuetzle:2003, author = {Jeremy Tantrum and Alejandro Murua and Werner Stuetzle}, title = {Assessment and Pruning of Hierarchical Model Based Clustering}, booktitle = {Proceedings of the ninth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining}, pages = {197--205}, year = 2003, publisher = {ACM Press}, address = {New York, NY, USA}, isbn = {1-58113-737-0}, } @Book{ flexmix:Titterington+Smith+Makov:1985, author = {D.M. Titterington and A.F.M. Smith and U.E. Makov}, title = {Statistical Analysis of Finite Mixture Distributions}, publisher = {John Wiley and Sons Inc.}, year = 1985 } @InProceedings{ flexmix:Urbanek+Theus:2003, author = {Simon Urbanek and Martin Theus}, title = {{iPlots}---High Interaction Graphics for {R}}, booktitle = {Proceedings of the 3rd International Workshop on Distributed Statistical Computing, Vienna, Austria}, editor = {Kurt Hornik and Friedrich Leisch and Achim Zeileis}, year = 2003, url = {http://www.ci.tuwien.ac.at/Conferences/DSC-2003/Proceedings/}, note = {{ISSN 1609-395X}} } @Book{ flexmix:Venables+Ripley:2002, title = {Modern Applied Statistics with S}, author = {William N. Venables and Brian D. Ripley}, publisher = {Springer Verlag}, edition = {Fourth}, address = {New York}, year = 2002, isbn = {0-387-95457-0} } @Article{ flexmix:Wang+Cockburn+Puterman:1998, author = {Peiming Wang and Iain M. Cockburn and Martin L. Puterman}, title = {Analysis of Patent Data---{A} Mixed-{P}oisson-Regression-Model Approach}, journal = {Journal of Business \& Economic Statistics}, year = 1998, volume = 16, number = 1, pages = {27--41} } @Article{ flexmix:Wang+Puterman+Cockburn:1996, author = {Peiming Wang and Martin L. Puterman and Iain M. Cockburn and Nhu D. Le}, title = {Mixed {P}oisson Regression Models with Covariate Dependent Rates}, journal = {Biometrics}, year = 1996, volume = 52, pages = {381--400} } @Article{ flexmix:Wang+Puterman:1998, author = {Peiming Wang and Martin L. Puterman}, title = {Mixed Logistic Regression Models}, journal = {Journal of Agricultural, Biological, and Environmental Statistics}, year = 1998, volume = 3, number = 2, pages = {175--200} } @Article{ flexmix:Wedel+DeSarbo:1995, author = {Michel Wedel and Wayne S. DeSarbo}, title = {A Mixture Likelihood Approach for Generalized Linear Models}, journal = {Journal of Classification}, year = 1995, volume = 12, pages = {21--55} } @Book{ flexmix:Wedel+Kamakura:2001, author = {Michel Wedel and Wagner A. Kamakura}, title = {Market Segmentation -- Conceptual and Methodological Foundations}, publisher = {Kluwer Academic Publishers}, year = 2001, address = {Boston, MA, USA}, edition = {2nd} } flexmix/vignettes/bootstrapping.Rnw0000644000176200001440000004740313425024236017332 0ustar liggesusers\documentclass[nojss]{jss} \usepackage{amsfonts,bm,amsmath,amssymb} %%\usepackage{Sweave} %% already provided by jss.cls %%%\VignetteIndexEntry{Finite Mixture Model Diagnostics Using Resampling Methods} %%\VignetteDepends{flexmix} %%\VignetteKeywords{R, finite mixture model, resampling, bootstrap} %%\VignettePackage{flexmix} \title{Finite Mixture Model Diagnostics Using Resampling Methods} <>= options(useFancyQuotes = FALSE) digits <- 3 Colors <- c("#E495A5", "#39BEB1") critical_values <- function(n, p = "0.95") { data("qDiptab", package = "diptest") if (n %in% rownames(qDiptab)) { return(qDiptab[as.character(n), p]) }else { n.approx <- as.numeric(rownames(qDiptab)[which.min(abs(n - as.numeric(rownames(qDiptab))))]) return(sqrt(n.approx)/sqrt(n) * qDiptab[as.character(n.approx), p]) } } library("graphics") library("flexmix") combine <- function(x, sep, width) { cs <- cumsum(nchar(x)) remaining <- if (any(cs[-1] > width)) combine(x[c(FALSE, cs[-1] > width)], sep, width) c(paste(x[c(TRUE, cs[-1] <= width)], collapse= sep), remaining) } prettyPrint <- function(x, sep = " ", linebreak = "\n\t", width = getOption("width")) { x <- strsplit(x, sep)[[1]] paste(combine(x, sep, width), collapse = paste(sep, linebreak, collapse = "")) } @ \author{Bettina Gr{\"u}n\\ Johannes Kepler Universit{\"a}t Linz \And Friedrich Leisch\\ Universit\"at f\"ur Bodenkultur Wien} \Plainauthor{Bettina Gr{\"u}n, Friedrich Leisch} \Address{ Bettina Gr\"un\\ Institut f\"ur Angewandte Statistik\\ Johannes Kepler Universit{\"a}t Linz\\ Altenbergerstra\ss{}e 69\\ 4040 Linz, Austria\\ E-mail: \email{Bettina.Gruen@jku.at} Friedrich Leisch\\ Institut f\"ur Angewandte Statistik und EDV\\ Universit\"at f\"ur Bodenkultur Wien\\ Peter Jordan Stra\ss{}e 82\\ 1190 Wien, Austria\\ E-mail: \email{Friedrich.Leisch@boku.ac.at}\\ URL: \url{http://www.statistik.lmu.de/~leisch/} } \Abstract{ This paper illustrates the implementation of resampling methods in \pkg{flexmix} as well as the application of resampling methods for model diagnostics of fitted finite mixture models. Convenience functions to perform these methods are available in package \pkg{flexmix}. The use of the methods is illustrated with an artificial example and the \code{seizure} data set. } \Keywords{\proglang{R}, finite mixture models, resampling, bootstrap} \Plainkeywords{R, finite mixture models, resampling, bootstrap} %%------------------------------------------------------------------------- %%------------------------------------------------------------------------- \begin{document} \SweaveOpts{engine=R, echo=true, height=5, width=8, eps=FALSE, keep.source=TRUE} \setkeys{Gin}{width=0.95\textwidth} \section{Implementation of resampling methods}\label{sec:implementation} The proposed framework for model diagnostics using resampling \citep{mixtures:gruen+leisch:2004} equally allows to investigate model fit for all kinds of mixture models. The procedure is applicable to mixture models with different component specific models and does not impose any limitation such as for example on the dimension of the parameter space of the component specific model. In addition to the fitting step different component specific models only require different random number generators for the parametric bootstrap. The \code{boot()} function in \pkg{flexmix} is a generic \proglang{S4} function with a method for fitted finite mixtures of class \code{"flexmix"} and is applicable to general finite mixture models. The function with arguments and their defaults is given by: <>= cat(prettyPrint(gsub("boot_flexmix", "boot", prompt(flexmix:::boot_flexmix, filename = NA)$usage[[2]]), sep = ", ", linebreak = paste("\n", paste(rep(" ", 2), collapse = ""), sep= ""), width = 70)) @ The interface is similar to the \code{boot()} function in package \pkg{boot} \citep{mixtures:Davison+Hinkley:1997, mixtures:Canty+Ripley:2010}. The \code{object} is a fitted finite mixture of class \code{"flexmix"} and \code{R} denotes the number of resamples. The possible bootstrapping method are \code{"empirical"} (also available as \code{"ordinary"}) and \code{"parametric"}. For the parametric bootstrap sampling from the fitted mixture is performed using \code{rflexmix()}. For mixture models with different component specific models \code{rflexmix()} requires a sampling method for the component specific model. Argument \code{initialize\_solution} allows to select if the EM algorithm is started in the original finite mixture solution or if random initialization is performed. The fitted mixture model might contain weights and group indicators. The weights are case weights and allow to reduce the amount of data if observations are identical. This is useful for example for latent class analysis of multivariate binary data. The argument \code{keep\_weights} allows to indicate if they should be kept for the bootstrapping. Group indicators allow to specify that the component membership is identical over several observations, e.g., for repeated measurements of the same individual. Argument \code{keep\_groups} allows to indicate if the grouping information should also be used in the bootstrapping. \code{verbose} indicates if information on the progress should be printed. The \code{control} argument allows to control the EM algorithm for fitting the model to each of the bootstrap samples. By default the \code{control} argument is extracted from the fitted model provided by \code{object}. \code{k} allows to specify the number of components and by default this is also taken from the fitted model provided. The \code{model} argument determines if also the model and the weights slot for each sample are stored and returned. The returned object is of class \code{"FLXboot"} and otherwise only contains the fitted parameters, the fitted priors, the log likelihoods, the number of components of the fitted mixtures and the information if the EM algorithm has converged. The likelihood ratio test is implemented based on \code{boot()} in function \code{LR\_test()} and returns an object of class \code{"htest"} containing the number of valid bootstrap replicates, the p-value, the double negative log likelihood ratio test statistics for the original data and the bootstrap replicates. The \code{plot} method for \code{"FLXboot"} objects returns a parallel coordinate plot with the fitted parameters separately for each of the components. \section{Artificial data set} In the following a finite mixture model is used as the underlying data generating process which is theoretically not identifiable. We are assuming a finite mixture of linear regression models with two components of equal size where the coverage condition is not fulfilled \citep{mixtures:Hennig:2000}. Hence, intra-component label switching is possible, i.e., there exist two parameterizations implying the same mixture distribution which differ how the components between the covariate points are combined. We assume that one measurement per object and a single categorical regressor with two levels are given. The usual design matrix for a model with intercept uses the two covariate points $\mathbf{x}_1 = (1, 0)'$ and $\mathbf{x}_2 = (1, 1)'$. The mixture distribution is given by \begin{eqnarray*} H(y|\mathbf{x}, \Theta) &=& \frac{1}{2} N(\mu_1, 0.1) + \frac{1}{2} N(\mu_2, 0.1), \end{eqnarray*} where $\mu_k(\mathbf{x}) = \mathbf{x}'\bm{\alpha}_k$ and $N(\mu, \sigma^2)$ is the normal distribution. Now let $\mu_1(\mathbf{x}_1) = 1$, $\mu_2(\mathbf{x}_1) = 2$, $\mu_1(\mathbf{x}_2) = -1$ and $\mu_2(\mathbf{x}_2) = 4$. As Gaussian mixture distributions are generically identifiable the means, variances and component weights are uniquely determined in each covariate point given the mixture distribution. However, as the coverage condition is not fulfilled, the two possible solutions for $\bm{\alpha}$ are: \begin{description} \item[Solution 1:] $\bm{\alpha}_1^{(1)} = (2,\phantom{-}2)'$, $\bm{\alpha}_2^{(1)} = (1,-2)'$, \item[Solution 2:] $\bm{\alpha}_1^{(2)} = (2,-3)'$, $\bm{\alpha}_2^{(2)} = (1,\phantom{-}3)'$. \end{description} We specify this artificial mixture distribution using \code{FLXdist()}. \code{FLXdist()} returns an unfitted finite mixture of class \code{"FLXdist"}. The class of fitted finite mixture models \code{"flexmix"} extends class \code{"FLXdist"}. Each component follows a normal distribution. The parameters specified in a named list therefore consist of the regression coefficients and the standard deviation. Function \code{FLXdist()} has an argument \code{formula} for specifying the regression in each of the components, an argument \code{k} for the component weights and \code{components} for the parameters of each of the components. <<>>= library("flexmix") Component_1 <- list(Model_1 = list(coef = c(1, -2), sigma = sqrt(0.1))) Component_2 <- list(Model_1 = list(coef = c(2, 2), sigma = sqrt(0.1))) ArtEx.mix <- FLXdist(y ~ x, k = rep(0.5, 2), components = list(Component_1, Component_2)) @ We draw a balanced sample with 50 observations in each covariate point from the mixture model using \code{rflexmix()} after defining the data points for the covariates. \code{rflexmix()} can either have an unfitted or a fitted finite mixture as input. For unfitted mixtures data has to be provided using the \code{newdata} argument. For already fitted mixtures data can be optionally provided, otherwise the data used for fitting the mixture is used. <<>>= ArtEx.data <- data.frame(x = rep(0:1, each = 100/2)) set.seed(123) ArtEx.sim <- rflexmix(ArtEx.mix, newdata = ArtEx.data) ArtEx.data$y <- ArtEx.sim$y[[1]] ArtEx.data$class <- ArtEx.sim$class @ In Figure~\ref{fig:art} the sample is plotted together with the two solutions for combining $x_1$ and $x_2$, i.e., this illustrates intra-component label switching. \begin{figure} \centering <>= par(mar = c(5, 4, 2, 0) + 0.1) plot(y ~ x, data = ArtEx.data, pch = with(ArtEx.data, 2*class + x)) pars <- list(matrix(c(1, -2, 2, 2), ncol = 2), matrix(c(1, 3, 2, -3), ncol = 2)) for (i in 1:2) apply(pars[[i]], 2, abline, col = Colors[i]) @ \caption{Balanced sample from the artificial example with the two theoretical solutions.} \label{fig:art} \end{figure} We fit a finite mixture to the sample using \code{stepFlexmix()}. <<>>= set.seed(123) ArtEx.fit <- stepFlexmix(y ~ x, data = ArtEx.data, k = 2, nrep = 5, control = list(iter = 1000, tol = 1e-8, verbose = 0)) @ The fitted mixture can be inspected using \code{summary()} and \code{parameters()}. <<>>= summary(ArtEx.fit) parameters(ArtEx.fit) @ Obviously the fitted mixture parameters correspond to the parameterization we used to specify the mixture distribution. Using standard asymptotic theory to analyze the fitted mixture model gives the following estimates for the standard deviations. <<>>= ArtEx.refit <- refit(ArtEx.fit) summary(ArtEx.refit) @ The fitted mixture can also be analyzed using resampling techniques. For analyzing the stability of the parameter estimates where the possibility of identifiability problems is also taken into account the parametric bootstrap is used with random initialization. Function \code{boot()} can be used for empirical or parametric bootstrap (specified by the argument \code{sim}). The logical argument \code{initialize_solution} specifies if the initialization is in the original solution or random. By default random initialization is made. The number of bootstrap samples is set by the argument \code{R}. Please note that the arguments are chosen to correspond to those for function \code{boot} in package \pkg{boot} \citep{mixtures:Davison+Hinkley:1997}. Only a few number of bootstrap samples are drawn to keep the amount of time needed to run the vignette within reasonable limits. However, for a sensible application of the bootstrap methods at least \code{R} equal to 100 should be used. If the output for this setting has been saved, it is loaded and used in the further analysis. Please see the appendix for the code for generating the saved \proglang{R} output. <<>>= set.seed(123) ArtEx.bs <- boot(ArtEx.fit, R = 15, sim = "parametric") if ("boot-output.rda" %in% list.files()) load("boot-output.rda") ArtEx.bs @ Function \code{boot()} returns an object of class \code{"\Sexpr{class(ArtEx.bs)}"}. The default plot compares the bootstrap parameter estimates to the confidence intervals derived using standard asymptotic theory in a parallel coordinate plot (see Figure~\ref{fig:plot.FLXboot-art}). Clearly two groups of parameter estimates can be distinguished which are about of equal size. One subset of the parameter estimates stays within the confidence intervals induced by standard asymptotic theory, while the second group corresponds to the second solution and clusters around these parameter values. \begin{figure}[h!] \centering <>= print(plot(ArtEx.bs, ordering = "coef.x", col = Colors)) @ \caption{Diagnostic plot of the bootstrap results for the artificial example.} \label{fig:plot.FLXboot-art} \end{figure} In the following the DIP-test is applied to check if the parameter estimates follow a unimodal distribution. This is done for the aggregated parameter esimates where unimodality implies that this parameter is not suitable for imposing an ordering constraint which induces a unique labelling. For the separate component analysis which is made after imposing an ordering constraint on the coefficient of $x$ rejection the null hypothesis of unimodality implies that identifiability problems are present, e.g.~due to intra-component label switching. <<>>= require("diptest") parameters <- parameters(ArtEx.bs) Ordering <- factor(as.vector(apply(matrix(parameters[,"coef.x"], nrow = 2), 2, order))) Comp1 <- parameters[Ordering == 1,] Comp2 <- parameters[Ordering == 2,] dip.values.art <- matrix(nrow = ncol(parameters), ncol = 3, dimnames=list(colnames(parameters), c("Aggregated", "Comp 1", "Comp 2"))) dip.values.art[,"Aggregated"] <- apply(parameters, 2, dip) dip.values.art[,"Comp 1"] <- apply(Comp1, 2, dip) dip.values.art[,"Comp 2"] <- apply(Comp2, 2, dip) dip.values.art @ The critical value for column \code{Aggregated} is \Sexpr{round(critical_values(nrow(parameters)), digits = digits)} and for the columns of the separate components \Sexpr{round(critical_values(nrow(Comp1)), digits = digits)}. The component sizes as well as the standard deviations follow a unimodal distribution for the aggregated data as well as for each of the components. The regression coefficients are multimodal for the aggregate data as well as for each of the components. While from the aggregated case it might be concluded that imposing an ordering constraint on the intercept or the coefficient of $x$ is suitable, the component-specific analyses reveal that a unique labelling was not achieved. \section{Seizure} In \cite{mixtures:Wang+Puterman+Cockburn:1996} a Poisson mixture regression is fitted to data from a clinical trial where the effect of intravenous gammaglobulin on suppression of epileptic seizures is investigated. The data used were 140 observations from one treated patient, where treatment started on the $28^\textrm{th}$ day. In the regression model three independent variables were included: treatment, trend and interaction treatment-trend. Treatment is a dummy variable indicating if the treatment period has already started. Furthermore, the number of parental observation hours per day were available and it is assumed that the number of epileptic seizures per observation hour follows a Poisson mixture distribution. The number of epileptic seizures per parental observation hour for each day are plotted in Figure~\ref{fig:seizure}. The fitted mixture distribution consists of two components which can be interpreted as representing 'good' and 'bad' days of the patients. The mixture model can be formulated by \begin{equation*} H(y|\mathbf{x}, \Theta) = \pi_1 P(\lambda_1) + \pi_2 P(\lambda_2), \end{equation*} where $\lambda_k = e^{\mathbf{x}'\bm{\alpha}_k}$ for $k = 1,2$ and $P(\lambda)$ is the Poisson distribution. The data is loaded and the mixture fitted with two components. <<>>= data("seizure", package = "flexmix") model <- FLXMRglm(family = "poisson", offset = log(seizure$Hours)) control <- list(iter = 1000, tol = 1e-10, verbose = 0) set.seed(123) seizMix <- stepFlexmix(Seizures ~ Treatment * log(Day), data = seizure, k = 2, nrep = 5, model = model, control = control) @ The fitted regression lines for each of the two components are shown in Figure~\ref{fig:seizure}. \begin{figure}[h!] \begin{center} <>= par(mar = c(5, 4, 2, 0) + 0.1) plot(Seizures/Hours~Day, data=seizure, pch = as.integer(seizure$Treatment)) abline(v = 27.5, lty = 2, col = "grey") matplot(seizure$Day, fitted(seizMix)/seizure$Hours, type="l", add = TRUE, col = 1, lty = 1, lwd = 2) @ \caption{Seizure data with the fitted values for the \citeauthor{mixtures:Wang+Puterman+Cockburn:1996} model. The plotting character for the observed values in the base period is a circle and for those in the treatment period a triangle.} \label{fig:seizure} \end{center} \end{figure} The parameteric bootstrap with random initialization is used to investigate identifiability problems and parameter stability. The diagnostic plot is given in Figure~\ref{fig:plot.FLXboot-seiz}. The coloring is according to an ordering constraint on the intercept. Clearly the parameter estimates corresponding to the solution where the bad days from the base period are combined with the good days from the treatement period and vice versa for the good days of the base period can be distinguished and indicate the slight identifiability problems of the fitted mixture. \begin{figure}[h!] \centering <>= set.seed(123) seizMix.bs <- boot(seizMix, R = 15, sim = "parametric") if ("boot-output.rda" %in% list.files()) load("boot-output.rda") print(plot(seizMix.bs, ordering = "coef.(Intercept)", col = Colors)) @ \label{fig:plot.FLXboot-seiz} \caption{Diagnostic plot of the bootstrap results for the \code{seizure} data.} \end{figure} <<>>= parameters <- parameters(seizMix.bs) Ordering <- factor(as.vector(apply(matrix(parameters[,"coef.(Intercept)"], nrow = 2), 2, order))) Comp1 <- parameters[Ordering == 1,] Comp2 <- parameters[Ordering == 2,] @ For applying the DIP test also an ordering constraint on the intercept is used. The critical value for column \code{Aggregated} is \Sexpr{round(critical_values(nrow(parameters)), digits = digits)} and for the columns of the separate components \Sexpr{round(critical_values(nrow(Comp1)), digits = digits)}. <<>>= dip.values.art <- matrix(nrow = ncol(parameters), ncol = 3, dimnames = list(colnames(parameters), c("Aggregated", "Comp 1", "Comp 2"))) dip.values.art[,"Aggregated"] <- apply(parameters, 2, dip) dip.values.art[,"Comp 1"] <- apply(Comp1, 2, dip) dip.values.art[,"Comp 2"] <- apply(Comp2, 2, dip) dip.values.art @ For the aggregate results the hypothesis of unimodality cannot be rejected for the trend. For the component-specific analyses unimodality cannot be rejected only for the intercept (where the ordering condition was imposed on) and again the trend. For all other parameter estimates unimodality is rejected which indicates that the ordering constraint was able to impose a unique labelling only for the own parameter and not for the other parameters. This suggests identifiability problems. %%------------------------------------------------------------------------- %%------------------------------------------------------------------------- \appendix \section[Generation of saved R output]{Generation of saved \proglang{R} output} <>= set.seed(123) ArtEx.bs <- boot(ArtEx.fit, R = 200, sim = "parametric") set.seed(123) seizMix.bs <- boot(seizMix, R = 200, sim = "parametric") save(ArtEx.bs, seizMix.bs, file = "boot-output.rda") @ \bibliography{mixture} \end{document} flexmix/vignettes/ziglm.R0000644000176200001440000000220513425024236015203 0ustar liggesuserssetClass("FLXMRziglm", contains = "FLXMRglm") FLXMRziglm <- function(formula = . ~ ., family = c("binomial", "poisson"), ...) { family <- match.arg(family) new("FLXMRziglm", FLXMRglm(formula, family, ...), name = paste("FLXMRziglm", family, sep=":")) } setMethod("FLXgetModelmatrix", signature(model="FLXMRziglm"), function(model, data, formula, lhs=TRUE, ...) { model <- callNextMethod(model, data, formula, lhs) if (attr(terms(model@fullformula), "intercept") == 0) stop("please include an intercept") model }) setMethod("FLXremoveComponent", signature(model = "FLXMRziglm"), function(model, nok, ...) { if (1 %in% nok) as(model, "FLXMRglm") else model }) setMethod("FLXmstep", signature(model = "FLXMRziglm"), function(model, weights, components, ...) { coef <- c(-Inf, rep(0, ncol(model@x)-1)) names(coef) <- colnames(model@x) comp.1 <- with(list(coef = coef, df = 0, offset = NULL, family = model@family), eval(model@defineComponent)) c(list(comp.1), FLXmstep(as(model, "FLXMRglm"), weights[, -1, drop=FALSE], components[-1])) }) flexmix/vignettes/mixture.bib0000644000176200001440000003623713425024236016125 0ustar liggesusers@STRING{csda = {Computational Statistics \& Data Analysis} } @STRING{jasa = {Journal of the American Statistical Association} } @STRING{jcgs = {Journal of Computational and Graphical Statistics} } @STRING{jrssa = {Journal of the Royal Statistical Society A} } @STRING{jrssb = {Journal of the Royal Statistical Society B} } @Article{ mixtures:aitkin:1999, author = {Murray Aitkin}, title = {A General Maximum Likelihood Analysis of Variance Components in Generalized Linear Models}, journal = {Biometrics}, year = 1999, volume = 55, pages = {117--128} } @Article{ mixtures:aitkin:1999a, author = {Murray Aitkin}, title = {Meta-Analysis by Random Effect Modelling in Generalized Linear Models}, journal = {Statistics in Medicine}, year = 1999, volume = 18, number = {17--18}, month = {September}, pages = {2343--2351} } @Article{ mixtures:biernacki+celeux+govaert:2000, author = {Christophe Biernacki and Gilles Celeux and G{\'e}rard Govaert}, title = {Assessing a Mixture Model for Clustering with the Integrated Completed Likelihood}, journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence}, year = 2000, volume = 22, number = 7, pages = {719--725}, month = {July} } @Article{ mixtures:biernacki+celeux+govaert:2003, author = {Christophe Biernacki and Gilles Celeux and G{\'e}rard Govaert}, title = {Choosing Starting Values for the {EM} Algorithm for Getting the Highest Likelihood in Multivariate {G}aussian Mixture Models}, journal = csda, year = 2003, volume = 41, pages = {561--575} } @Article{ mixtures:boehning+dietz+schlattmann:1999, author = {Dankmar B{\"o}hning and Ekkehart Dietz and Peter Schlattmann and Lisette Mendon{\c c}a and Ursula Kirchner}, title = {The Zero-Inflated {P}oisson Model and the Decayed, Missing and Filled Teeth Index in Dental Epidemiology}, journal = jrssa, year = 1999, volume = 162, number = 2, pages = {195--209}, month = {August} } @Book{ mixtures:boehning:1999, author = {Dankmar B{\"o}hning}, title = {Computer Assisted Analysis of Mixtures and Applications: Meta-Analysis, Disease Mapping, and Others}, publisher = {Chapman \& Hall/CRC}, year = 1999, volume = 81, series = {Monographs on Statistics and Applied Probability}, address = {London} } @Manual{ mixtures:canty+ripley:2010, title = {boot: Bootstrap R (S-Plus) Functions}, author = {Angelo Canty and Brian Ripley}, year = 2010, note = {R package version 1.2-43}, url = {http://CRAN.R-project.org/package=boot} } @TechReport{ mixtures:celeux+diebolt:1988, author = {Gilles Celeux and Jean Diebolt}, title = {A Random Imputation Principle: The Stochastic {EM} Algorithm}, institution = {INRIA}, year = 1988, type = {Rapports de Recherche}, number = 901, month = {September} } @Article{ mixtures:celeux+govaert:1992, author = {Gilles Celeux and G{\'e}rard Govaert}, title = {A {C}lassification {EM} Algorithm for Clustering and Two Stochastic Versions}, journal = {Computational Statistics \& Data Analysis}, year = 1992, volume = 14, number = 3, pages = {315--332}, month = {October} } @Book{ mixtures:chambers:1998, author = {John M. Chambers}, title = {Programming with Data}, publisher = {Springer-Verlag}, year = 1998, address = {New York}, isbn = {0-387-98503-4} } @Article{ mixtures:dasgupta+raftery:1998, author = {Abhijit Dasgupta and Adrian E. Raftery}, title = {Detecting Features in Spatial Point Processes with Clutter Via Model-Based Clustering}, journal = jasa, year = 1998, volume = 93, number = 441, pages = {294--302}, month = {March} } @Book{mixtures:davison+hinkley:1997, author = {A. C. Davison and D. V. Hinkley}, title = {Bootstrap Methods and Their Application}, publisher = {Cambridge University Press}, address = {Cambridge, UK}, year = 1997, isbn = {0-521-57391-2 (hardcover), 0-521-57471-4 (paperback)}, series = {Cambridge Series on Statistical and Probabilistic Mathematics}, } @Article{ mixtures:dayton+macready:1988, author = {C. Mitchell Dayton and George B. Macready}, title = {Concomitant-Variable Latent-Class Models}, journal = jasa, year = 1988, volume = 83, number = 401, pages = {173--178}, month = {March} } @Article{ mixtures:dempster+laird+rubin:1977, author = {A. P. Dempster and N. M. Laird and D. B. Rubin}, title = {Maximum Likelihood from Incomplete Data Via the {EM}-Algorithm}, journal = jrssb, volume = 39, pages = {1--38}, year = 1977 } @InCollection{ mixtures:diebolt+ip:1996, author = {Jean Diebolt and Eddie H. S. Ip}, editor = {W. R. Gilks and S. Richardson and D. J. Spiegelhalter}, booktitle = {Markov Chain {M}onte {C}arlo in Practice}, title = {Stochastic {EM}: Method and Application}, publisher = {Chapman and Hall}, year = 1996, pages = {259--273} } @Book{ mixtures:everitt+hand:1981, author = {B. S. Everitt and D. J. Hand}, title = {Finite Mixture Distributions}, publisher = {Chapman and Hall}, address = {London}, year = 1981 } @Article{ mixtures:follmann+lambert:1989, author = {Dean A. Follmann and Diane Lambert}, title = {Generalizing Logistic Regression by Non-Parametric Mixing}, journal = jasa, volume = 84, number = 405, month = {March}, pages = {295--300}, year = 1989 } @Article{ mixtures:follmann+lambert:1991, author = {Dean A. Follmann and Diane Lambert}, title = {Identifiability of Finite Mixtures of Logistic Regression Models}, journal = {Journal of Statistical Planning and Inference}, volume = 27, number = 3, month = {March}, pages = {375--381}, year = 1991 } @Book{ mixtures:fowler:2004, author = {Martin Fowler}, title = {{UML} Distilled: A Brief Guide to the Standard Object Modeling Language}, publisher = {Addison-Wesley Professional}, year = 2004, edition = {3rd} } @Book{ mixtures:fruehwirth-schnatter:2006, author = {Sylvia Fr{\"u}hwirth-Schnatter}, title = {Finite Mixture and Markov Switching Models}, publisher = {Springer-Verlag}, address = {New York}, series = {Springer-Verlag Series in Statistics}, isbn = {0-387-32909-9}, year = 2006 } @Article{ mixtures:gentleman+ihaka:2000, author = {Robert Gentleman and Ross Ihaka}, title = {Lexical Scope and Statistical Computing}, journal = jcgs, year = 2000, volume = 9, number = 3, pages = {491--508} } @InProceedings{ mixtures:gruen+leisch:2004, author = {Bettina Gr{\" u}n and Friedrich Leisch}, title = {Bootstrapping Finite Mixture Models}, booktitle = {COMPSTAT 2004 -- Proceedings in Computational Statistics}, year = 2004, editor = {Jaromir Antoch}, publisher = {Physica Verlag, Heidelberg}, isbn = {3-7908-1554-3}, pages = {1115--1122} } @InProceedings{ mixtures:gruen+leisch:2006, author = {Bettina Gr{\" u}n and Friedrich Leisch}, title = {Fitting Finite Mixtures of Linear Regression Models with Varying \& Fixed Effects in \proglang{R}}, booktitle = {COMPSTAT 2006 -- Proceedings in Computational Statistics}, pages = {853--860}, editor = {Alfredo Rizzi and Maurizio Vichi}, publisher = {Physica Verlag, Heidelberg, Germany}, isbn = {3-7908-1708-2}, year = 2006 } @Article{ mixtures:gruen+leisch:2007a, author = {Bettina Gr{\" u}n and Friedrich Leisch}, title = {Fitting Finite Mixtures of Generalized Linear Regressions in \proglang{R}}, journal = csda, year = 2007, volume = 51, number = 11, month = {July}, pages = {5247--5252} } @Article{ mixtures:gruen+leisch:2008, author = {Bettina Gr\"un and Friedrich Leisch}, title = {Identifiability of Finite Mixtures of Multinomial Logit Models with Varying and Fixed Effects}, journal = {Journal of Classification}, year = 2008, note = {Accepted for publication on 2008-03-28} } @Article{ mixtures:gruen+leisch:2008a, author = {Bettina Gr\"un and Friedrich Leisch}, title = {{F}lex{M}ix Version 2: Finite Mixtures with Concomitant Variables and Varying and Constant Parameters}, journal = {Journal of Statistical Software}, year = 2008, volume = 28, number = 4, pages = {1--35}, url = {http://www.jstatsoft.org/v28/i04/} } @Unpublished{mixtures:gruen+leisch:2010, author = {Bettina Gr\"un and Friedrich Leisch}, title = {Finite Mixture Model Diagnostics Using Resampling Methods}, note = {Unpublished manuscript}, year = 2010, } @PhDThesis{ mixtures:gruen:2006, author = {Bettina Gr{\"u}n}, title = {Identification and Estimation of Finite Mixture Models}, school = {Institut f{\"u}r Statistik und Wahrscheinlichkeitstheorie, Technische Universit{\"a}t Wien}, year = 2006, month = {September}, note = {{Friedrich Leisch, advisor}} } @Book{ mixtures:harrell:2001, author = {Frank E. Harrell}, title = {Regression Modeling Strategies}, publisher = {Springer-Verlag}, address = {New York}, year = 2001 } @Article{ mixtures:hastie+tibshirani:1993, author = {Trevor Hastie and Robert Tibshirani}, title = {Varying-Coefficient Models}, journal = jrssb, year = 1993, volume = 55, number = 4, pages = {757--796} } @Article{mixtures:hennig:2000, author = {Christian Hennig}, title = {Identifiability of Models for Clusterwise Linear Regression}, journal = {Journal of Classification}, volume = 17, number = 2, month = {July}, pages = {273--296}, year = 2000 } @Article{ mixtures:karlis+xekalaki:2003, author = {Dimitris Karlis and Evdokia Xekalaki}, title = {Choosing Initial Values for the {EM} Algorithm for Finite Mixtures}, journal = csda, year = 2003, volume = 41, pages = {577--590} } @InProceedings{ mixtures:leisch:2002, author = {Friedrich Leisch}, title = {Sweave: Dynamic Generation of Statistical Reports Using Literate Data Analysis}, booktitle = {COMPSTAT 2002 -- Proceedings in Computational Statistics}, pages = {575--580}, year = 2002, editor = {Wolfgang H{\"a}rdle and Bernd R{\"o}nz}, publisher = {Physica Verlag, Heidelberg}, note = {ISBN 3-7908-1517-9} } @Article{ mixtures:leisch:2003, author = {Friedrich Leisch}, title = {Sweave, Part {II}: Package Vignettes}, journal = {\proglang{R} News}, year = 2003, volume = 3, number = 2, pages = {21--24}, month = {October}, url = {http://CRAN.R-project.org/doc/Rnews/} } @Article{ mixtures:leisch:2004, author = {Friedrich Leisch}, title = {\pkg{FlexMix}: A General Framework for Finite Mixture Models and Latent Class Regression in \proglang{R}}, journal = {Journal of Statistical Software}, year = 2004, volume = 11, number = 8, pages = {1--18}, url = {http://www.jstatsoft.org/v11/i08/} } @InProceedings{ mixtures:leisch:2004a, author = {Friedrich Leisch}, title = {Exploring the Structure of Mixture Model Components}, booktitle = {COMPSTAT 2004 -- Proceedings in Computational Statistics}, year = 2004, editor = {Jaromir Antoch}, publisher = {Physica Verlag, Heidelberg}, isbn = {3-7908-1554-3}, pages = {1405--1412} } @InProceedings{ mixtures:leisch:2008, author = {Friedrich Leisch}, title = {Modelling Background Noise in Finite Mixtures of Generalized Linear Regression Models}, booktitle = {COMPSTAT 2008 -- Proceedings in Computational Statistics}, volume = {I}, pages = {385-396}, editor = {Paula Brito}, publisher = {Physica Verlag, Heidelberg, Germany}, isbn = {978-3-7908-2083-6}, year = 2008 } @Article{ mixtures:long:1990, author = {J. Scott Long}, title = {The Origins of Sex Differences in Science}, journal = {Social Forces}, year = 1990, volume = 68, number = 4, pages = {1297--1315}, month = {June} } @Book{ mixtures:mccullagh+nelder:1989, author = {Peter McCullagh and John A. Nelder}, title = {Generalized Linear Models}, edition = {2nd}, publisher = {Chapman and Hall}, year = 1989, location = {London} } @Book{ mixtures:mclachlan+basford:1988, author = {Geoffrey J. McLachlan and Kaye E. Basford}, title = {Mixture Models: Inference and Applications to Clustering}, publisher = {Marcel Dekker}, year = 1988, address = {New York} } @Book{ mixtures:mclachlan+peel:2000, author = {Geoffrey J. McLachlan and David Peel}, title = {Finite Mixture Models}, publisher = {John Wiley \& Sons}, year = 2000 } @Book{ mixtures:pinheiro+bates:2000, author = {Jose C. Pinheiro and Douglas M. Bates}, title = {Mixed-Effects Models in \proglang{S} and \proglang{S-Plus}}, publisher = {Springer-Verlag}, year = 2000, isbn = {0-387-98957-0} } @Book{ mixtures:sarkar:2008, title = {\pkg{lattice}: Multivariate Data Visualization with \proglang{R}}, author = {Deepayan Sarkar}, year = 2008, publisher = {Springer-Verlag}, address = {New York}, isbn = {978-0-387-75968-5} } @Book{ mixtures:titterington+smith+makov:1985, author = {D. M. Titterington and A. F. M. Smith and U. E. Makov}, title = {Statistical Analysis of Finite Mixture Distributions}, publisher = {John Wiley \& Sons}, year = 1985 } @Book{ mixtures:venables+ripley:2002, title = {Modern Applied Statistics with \proglang{S}}, author = {William N. Venables and Brian D. Ripley}, publisher = {Springer-Verlag}, edition = {4th}, address = {New York}, year = 2002, isbn = {0-387-95457-0} } @Article{ mixtures:wang+puterman+cockburn:1996, author = {Peiming Wang and Martin L. Puterman and Iain M. Cockburn and Nhu D. Le}, title = {Mixed {P}oisson Regression Models with Covariate Dependent Rates}, journal = {Biometrics}, year = 1996, volume = 52, pages = {381--400} } @Article{ mixtures:wedel+desarbo:1995, author = {Michel Wedel and Wagner S. DeSarbo}, title = {A Mixture Likelihood Approach for Generalized Linear Models}, journal = {Journal of Classification}, year = 1995, volume = 12, number = 1, month = {March}, pages = {21--55} } @Article{ mixtures:wehrens+buydens+fraley:2004, author = {Ron Wehrens and Lutgarde M.C. Buydens and Chris Fraley and Adrian E. Raftery}, title = {Model-Based Clustering for Image Segmentation and Large Datasets Via Sampling}, journal = {Journal of Classification}, year = 2004, volume = 21, number = 2, pages = {231--253} } flexmix/vignettes/regression-examples.Rnw0000644000176200001440000012412713425024236020432 0ustar liggesusers\documentclass[nojss]{jss} \usepackage{amsfonts,bm,amsmath,amssymb} %%\usepackage{Sweave} %% already provided by jss.cls %%%\VignetteIndexEntry{Applications of finite mixtures of regression models} %%\VignetteDepends{flexmix} %%\VignetteKeywords{R, finite mixture model, generalized linear model, latent class regression} %%\VignettePackage{flexmix} \title{Applications of finite mixtures of regression models} <>= library("stats") library("graphics") library("flexmix") @ \author{Bettina Gr{\"u}n\\ Johannes Kepler Universit{\"at} Linz \And Friedrich Leisch\\ Universit\"at f\"ur Bodenkultur Wien} \Plainauthor{Bettina Gr{\"u}n, Friedrich Leisch} \Address{ Bettina Gr\"un\\ Institut f\"ur Angewandte Statistik / IFAS\\ Johannes Kepler Universit{\"at} Linz\\ Freist\"adter Stra\ss{}e 315\\ 4040 Linz, Austria\\ E-mail: \email{Bettina.Gruen@jku.at}\\ Friedrich Leisch\\ Institut f\"ur Angewandte Statistik und EDV\\ Universit\"at f\"ur Bodenkultur Wien\\ Peter Jordan Stra\ss{}e 82\\ 1190 Wien, Austria\\ E-mail: \email{Friedrich.Leisch@boku.ac.at}\\ URL: \url{http://www.statistik.lmu.de/~leisch/} } \Abstract{ Package \pkg{flexmix} provides functionality for fitting finite mixtures of regression models. The available model class includes generalized linear models with varying and fixed effects for the component specific models and multinomial logit models for the concomitant variable models. This model class includes random intercept models where the random part is modelled by a finite mixture instead of a-priori selecting a suitable distribution. The application of the package is illustrated on various datasets which have been previously used in the literature to fit finite mixtures of Gaussian, binomial or Poisson regression models. The \proglang{R} commands are given to fit the proposed models and additional insights are gained by visualizing the data and the fitted models as well as by fitting slightly modified models. } \Keywords{\proglang{R}, finite mixture models, generalized linear models, concomitant variables} \Plainkeywords{R, finite mixture models, generalized linear models, concomitant variables} %%------------------------------------------------------------------------- %%------------------------------------------------------------------------- \begin{document} \SweaveOpts{engine=R, echo=true, height=5, width=8, eps=FALSE, keep.source=TRUE} \setkeys{Gin}{width=0.8\textwidth} <>= options(width=70, prompt = "R> ", continue = "+ ", useFancyQuotes = FALSE) set.seed(1802) library("lattice") ltheme <- canonical.theme("postscript", FALSE) lattice.options(default.theme=ltheme) @ %%------------------------------------------------------------------------- %%------------------------------------------------------------------------- \section{Introduction} Package \pkg{flexmix} provides infrastructure for flexible fitting of finite mixtures models. The design principles of the package allow easy extensibility and rapid prototyping. In addition, the main focus of the available functionality is on fitting finite mixtures of regression models, as other packages in \proglang{R} exist which have specialized functionality for model-based clustering, such as e.g.~\pkg{mclust} \citep{flexmix:Fraley+Raftery:2002a} for finite mixtures of Gaussian distributions. \cite{flexmix:Leisch:2004a} gives a general introduction into the package outlining the main implementational principles and illustrating the use of the package. The paper is also contained as a vignette in the package. An example for fitting mixtures of Gaussian regression models is given in \cite{flexmix:Gruen+Leisch:2006}. This paper focuses on examples of finite mixtures of binomial logit and Poisson regression models. Several datasets which have been previously used in the literature to demonstrate the use of finite mixtures of regression models have been selected to illustrate the application of the package. The model class covered are finite mixtures of generalized linear model with focus on binomial logit and Poisson regressions. The regression coefficients as well as the dispersion parameters of the component specific models are assumed to vary for all components, vary between groups of components, i.e.~to have a nesting, or to be fixed over all components. In addition it is possible to specify concomitant variable models in order to be able to characterize the components. Random intercept models are a special case of finite mixtures with varying and fixed effects as fixed effects are assumed for the coefficients of all covariates and varying effects for the intercept. These models are often used to capture overdispersion in the data which can occur for example if important covariates are omitted in the regression. It is then assumed that the influence of these covariates can be captured by allowing a random distribution for the intercept. This illustration does not only show how the package \pkg{flexmix} can be used for fitting finite mixtures of regression models but also indicates the advantages of using an extension package of an environment for statistical computing and graphics instead of a stand-alone package as available visualization techniques can be used for inspecting the data and the fitted models. In addition users already familiar with \proglang{R} and its formula interface should find the model specification and a lot of commands for exploring the fitted model intuitive. %%------------------------------------------------------------------------- %%------------------------------------------------------------------------- \section{Model specification} Finite mixtures of Gaussian regressions with concomitant variable models are given by: \begin{align*} H(y\,|\,\bm{x}, \bm{w}, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s(\bm{w}, \bm{\alpha}) \textrm{N}(y\,|\, \mu_s(\bm{x}), \sigma^2_s), \end{align*} where $\textrm{N}(\cdot\,|\, \mu_s(\bm{x}), \sigma^2_s)$ is the Gaussian distribution with mean $\mu_s(\bm{x}) = \bm{x}' \bm{\beta}^s$ and variance $\sigma^2_s$. $\Theta$ denotes the vector of all parameters of the mixture distribution and the dependent variables are $y$, the independent $\bm{x}$ and the concomitant $\bm{w}$. Finite mixtures of binomial regressions with concomitant variable models are given by: \begin{align*} H(y\,|\,T, \bm{x}, \bm{w}, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s(\bm{w}, \bm{\alpha}) \textrm{Bi}(y\,|\,T, \theta_s(\bm{x})), \end{align*} where $\textrm{Bi}(\cdot\,|\,T, \theta_s(\bm{x}))$ is the binomial distribution with number of trials equal to $T$ and success probability $\theta_s(\bm{x}) \in (0,1)$ given by $\textrm{logit}(\theta_s(\bm{x})) = \bm{x}' \bm{\beta}^s$. Finite mixtures of Poisson regressions are given by: \begin{align*} H(y \,|\, \bm{x}, \bm{w}, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s(\bm{w}, \bm{\alpha}) \textrm{Poi} (y \,|\, \lambda_s(\bm{x})), \end{align*} where $\textrm{Poi}(\cdot\,|\,\lambda_s(\bm{x}))$ denotes the Poisson distribution and $\log(\lambda_s(\bm{x})) = \bm{x}'\bm{\beta}^s$. For all these mixture distributions the coefficients are split into three different groups depending on if fixed, nested or varying effects are specified: \begin{align*} \bm{\beta}^s &= (\bm{\beta}_1, \bm{\beta}^{c(s)}_{2}, \bm{\beta}^{s}_3) \end{align*} where the first group represents the fixed, the second the nested and the third the varying effects. For the nested effects a partition $\mathcal{C} = \{c_s \,|\, s = 1,\ldots S\}$ of the $S$ components is determined where $c_s = \{s^* = 1,\ldots,S \,|\, c(s^*) = c(s)\}$. A similar splitting is possible for the variance of mixtures of Gaussian regression models. The function for maximum likelihood (ML) estimation with the Expectation-Maximization (EM) algorithm is \code{flexmix()} which is described in detail in \cite{flexmix:Leisch:2004a}. It takes as arguments a specification of the component specific model and of the concomitant variable model. The component specific model with varying, nested and fixed effects can be specified with the M-step driver \code{FLXMRglmfix()} which has arguments \code{formula} for the varying, \code{nested} for the nested and \code{fixed} for the fixed effects. \code{formula} and \code{fixed} take an argument of class \code{"formula"}, whereas \code{nested} expects an object of class \code{"FLXnested"} or a named list specifying the nested structure with a component \code{k} which is a vector of the number of components in each group of the partition and a component \code{formula} which is a vector of formulas for each group of the partition. In addition there is an argument \code{family} which has to be one of \code{gaussian}, \code{binomial}, \code{poisson} or \code{Gamma} and determines the component specific distribution function as well as an \code{offset} argument. The argument \code{varFix} can be used to determine the structure of the dispersion parameters. If only varying effects are specified the M-step driver \code{FLXMRglm()} can be used which only has an argument \code{formula} for the varying effects and also a \code{family} and an \code{offset} argument. This driver has the advantage that in the M-step the weighted ML estimation is made separately for each component which signifies that smaller model matrices are used. If a mixture model with a lot of components $S$ is fitted to a large data set with $N$ observations and the model matrix used in the M-step of \code{FLXMRglm()} has $N$ rows and $K$ columns, the model matrix used in the M-step of \code{FLXMRglmfix()} has $S N$ rows and up to $S K$ columns. In general the concomitant variable model is assumed to be a multinomial logit model, i.e.~: \begin{align*} \pi_s(\bm{w},\bm{\alpha}) &= \frac{e^{\bm{w}'\bm{\alpha}_s}}{\sum_{u = 1}^S e^{\bm{w}'\bm{\alpha}_u}} \quad \forall s, \end{align*} with $\bm{\alpha} = (\bm{\alpha}'_s)_{s=1,\ldots,S}$ and $\bm{\alpha}_1 \equiv \bm{0}$. This model can be fitted in \pkg{flexmix} with \code{FLXPmultinom()} which takes as argument \code{formula} the formula specification of the multinomial logit part. For fitting the function \code{nnet()} is used from package \pkg{MASS} \citep{flexmix:Venables+Ripley:2002} with the independent variables specified by the formula argument and the dependent variables are given by the a-posteriori probability estimates. %%------------------------------------------------------------------------- %%------------------------------------------------------------------------- \section[Using package flexmix]{Using package \pkg{flexmix}} In the following datasets from different areas such as medicine, biology and economics are used. There are three subsections: for finite mixtures of Gaussian regressions, for finite mixtures of binomial regression models and for finite mixtures of Poisson regression models. %%------------------------------------------------------------------------- \subsection{Finite mixtures of Gaussian regressions} This artificial dataset with 200 observations is given in \cite{flexmix:Gruen+Leisch:2006}. The data is generated from a mixture of Gaussian regression models with three components. There is an intercept with varying effects, an independent variable $x1$, which is a numeric variable, with fixed effects and another independent variable $x2$, which is a categorical variable with two levels, with nested effects. The prior probabilities depend on a concomitant variable $w$, which is also a categorical variable with two levels. Fixed effects are also assumed for the variance. The data is illustrated in Figure~\ref{fig:artificialData} and the true underlying model is given by: \begin{align*} H(y\,|\,(x1, x2), w, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s(w, \bm{\alpha}) \textrm{N}(y\,|\, \mu_s, \sigma^2), \end{align*} with $\bm{\beta}^s = (\beta^s_{\textrm{Intercept}}, \beta^{c(s)}_{\textrm{x1}}, \beta_{\textrm{x2}})$. The nesting signifies that $c(1) = c(2)$ and $\beta^{c(3)}_{\textrm{x1}} = 0$. The mixture model is fitted by first loading the package and the dataset and then specifying the component specific model. In a first step a component specific model with only varying effects is specified. Then the fitting function \code{flexmix()} is called repeatedly using \code{stepFlexmix()}. Finally, we order the components such that they are in ascending order with respect to the coefficients of the variable \code{x1}. <>= set.seed(2807) library("flexmix") data("NregFix", package = "flexmix") Model <- FLXMRglm(~ x2 + x1) fittedModel <- stepFlexmix(y ~ 1, model = Model, nrep = 3, k = 3, data = NregFix, concomitant = FLXPmultinom(~ w)) fittedModel <- relabel(fittedModel, "model", "x1") summary(refit(fittedModel)) @ The estimated coefficients indicate that the components differ for the intercept, but that they are not significantly different for the coefficients of $x2$. For $x1$ the coefficient of the first component is not significantly different from zero and the confidence intervals for the other two components overlap. Therefore we fit a modified model, which is equivalent to the true underlying model. The previously fitted model is used for initializing the EM algorithm: <>= Model2 <- FLXMRglmfix(fixed = ~ x2, nested = list(k = c(1, 2), formula = c(~ 0, ~ x1)), varFix = TRUE) fittedModel2 <- flexmix(y ~ 1, model = Model2, cluster = posterior(fittedModel), data = NregFix, concomitant = FLXPmultinom(~ w)) BIC(fittedModel) BIC(fittedModel2) @ The BIC suggests that the restricted model should be preferred. \begin{figure}[tb] \centering \setkeys{Gin}{width=0.95\textwidth} <>= plotNregFix <- NregFix plotNregFix$w <- factor(NregFix$w, levels = 0:1, labels = paste("w =", 0:1)) plotNregFix$x2 <- factor(NregFix$x2, levels = 0:1, labels = paste("x2 =", 0:1)) plotNregFix$class <- factor(NregFix$class, levels = 1:3, labels = paste("Class", 1:3)) print(xyplot(y ~ x1 | x2*w, groups = class, data = plotNregFix, cex = 0.6, auto.key = list(space="right"), layout = c(2,2))) @ \setkeys{Gin}{width=0.8\textwidth} \caption{Sample with 200 observations from the artificial example.} \label{fig:artificialData} \end{figure} <>= summary(refit(fittedModel2)) @ The coefficients are ordered such that the fixed coefficients are first, the nested varying coefficients second and the varying coefficients last. %%------------------------------------------------------------------------- \subsection{Finite mixtures of binomial logit regressions} %%------------------------------------------------------------------------- \subsubsection{Beta blockers} The dataset is analyzed in \cite{flexmix:Aitkin:1999, flexmix:Aitkin:1999a} using a finite mixture of binomial regression models. Furthermore, it is described in \cite{flexmix:McLachlan+Peel:2000} on page 165. The dataset is from a 22-center clinical trial of beta-blockers for reducing mortality after myocardial infarction. A two-level model is assumed to represent the data, where centers are at the upper level and patients at the lower level. The data is illustrated in Figure~\ref{fig:beta} and the model is given by: \begin{align*} H(\textrm{Deaths} \,|\, \textrm{Total}, \textrm{Treatment}, \textrm{Center}, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s \textrm{Bi}( \textrm{Deaths} \,|\, \textrm{Total}, \theta_s). \end{align*} First, the center classification is ignored and a binomial logit regression model with treatment as covariate is fitted using \code{glm}, i.e.~$S=1$: <>= data("betablocker", package = "flexmix") betaGlm <- glm(cbind(Deaths, Total - Deaths) ~ Treatment, family = "binomial", data = betablocker) betaGlm @ In the next step the center classification is included by allowing a random effect for the intercept given the centers, i.e.~the coefficients $\bm{\beta}^s$ are given by $(\beta^s_{\textrm{Intercept|Center}}, \beta_{\textrm{Treatment}})$. This signifies that the component membership is fixed for each center. In order to determine the suitable number of components, the mixture is fitted with different numbers of components and the BIC information criterion is used to select an appropriate model. In this case a model with three components is selected. The fitted values for the model with three components are given in Figure~\ref{fig:beta}. <>= betaMixFix <- stepFlexmix(cbind(Deaths, Total - Deaths) ~ 1 | Center, model = FLXMRglmfix(family = "binomial", fixed = ~ Treatment), k = 2:4, nrep = 3, data = betablocker) betaMixFix @ \begin{figure} \centering <>= library("grid") betaMixFix_3 <- getModel(betaMixFix, "3") betaMixFix_3 <- relabel(betaMixFix_3, "model", "Intercept") betablocker$Center <- with(betablocker, factor(Center, levels = Center[order((Deaths/Total)[1:22])])) clusters <- factor(clusters(betaMixFix_3), labels = paste("Cluster", 1:3)) print(dotplot(Deaths/Total ~ Center | clusters, groups = Treatment, as.table = TRUE, data = betablocker, xlab = "Center", layout = c(3, 1), scales = list(x = list(draw = FALSE)), key = simpleKey(levels(betablocker$Treatment), lines = TRUE, corner = c(1,0)))) betaMixFix.fitted <- fitted(betaMixFix_3) for (i in 1:3) { seekViewport(trellis.vpname("panel", i, 1)) grid.lines(unit(1:22, "native"), unit(betaMixFix.fitted[1:22, i], "native"), gp = gpar(lty = 1)) grid.lines(unit(1:22, "native"), unit(betaMixFix.fitted[23:44, i], "native"), gp = gpar(lty = 2)) } @ \caption{Relative number of deaths for the treatment and the control group for each center in the beta blocker dataset. The centers are sorted by the relative number of deaths in the control group. The lines indicate the fitted values for each component of the 3-component mixture model with random intercept and fixed effect for treatment.} \label{fig:beta} \end{figure} In addition the treatment effect can also be included in the random part of the model. As then all coefficients for the covariates and the intercept follow a mixture distribution the component specific model can be specified using \code{FLXMRglm()}. The coefficients are $\bm{\beta}^s=(\beta^s_{\textrm{Intercept|Center}}, \beta^s_{\textrm{Treatment|Center}})$, i.e.~it is assumed that the heterogeneity is only between centers and therefore the aggregated data for each center can be used. <>= betaMix <- stepFlexmix(cbind(Deaths, Total - Deaths) ~ Treatment | Center, model = FLXMRglm(family = "binomial"), k = 3, nrep = 3, data = betablocker) summary(betaMix) @ The full model with a random effect for treatment has a higher BIC and therefore the smaller would be preferred. The default plot of the returned \code{flexmix} object is a rootogramm of the a-posteriori probabilities where observations with a-posteriori probabilities smaller than \code{eps} are omitted. With argument \code{mark} the component is specified to have those observations marked which are assigned to this component based on the maximum a-posteriori probabilities. This indicates which components overlap. <>= print(plot(betaMixFix_3, mark = 1, col = "grey", markcol = 1)) @ The default plot of the fitted model indicates that the components are well separated. In addition component 1 has a slight overlap with component 2 but none with component 3. The fitted parameters of the component specific models can be accessed with: <>= parameters(betaMix) @ The cluster assignments using the maximum a-posteriori probabilities are obtained with: <>= table(clusters(betaMix)) @ The estimated probabilities for each component for the treated patients and those in the control group can be obtained with: <>= predict(betaMix, newdata = data.frame(Treatment = c("Control", "Treated"))) @ or <>= fitted(betaMix)[c(1, 23), ] @ A further analysis of the model is possible with function \code{refit()} which returns the estimated coefficients together with the standard deviations, z-values and corresponding p-values: <>= summary(refit(getModel(betaMixFix, "3"))) @ The printed coefficients are ordered to have the fixed effects before the varying effects. %%----------------------------------------------------------------------- \subsubsection{Mehta et al. trial} This dataset is similar to the beta blocker dataset and is also analyzed in \cite{flexmix:Aitkin:1999a}. The dataset is visualized in Figure~\ref{fig:mehta}. The observation for the control group in center 15 is slightly conspicuous and might classify as an outlier. The model is given by: \begin{align*} H(\textrm{Response} \,|\, \textrm{Total}, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s \textrm{Bi}( \textrm{Response} \,|\, \textrm{Total}, \theta_s), \end{align*} with $\bm{\beta}^s = (\beta^s_{\textrm{Intercept|Site}}, \beta_{\textrm{Drug}})$. This model is fitted with: <>= data("Mehta", package = "flexmix") mehtaMix <- stepFlexmix(cbind(Response, Total - Response)~ 1 | Site, model = FLXMRglmfix(family = "binomial", fixed = ~ Drug), control = list(minprior = 0.04), nrep = 3, k = 3, data = Mehta) summary(mehtaMix) @ One component only contains the observations for center 15 and in order to be able to fit a mixture with such a small component it is necessary to modify the default argument for \code{minprior} which is 0.05. The fitted values for this model are given separately for each component in Figure~\ref{fig:mehta}. \begin{figure} \centering <>= Mehta$Site <- with(Mehta, factor(Site, levels = Site[order((Response/Total)[1:22])])) clusters <- factor(clusters(mehtaMix), labels = paste("Cluster", 1:3)) print(dotplot(Response/Total ~ Site | clusters, groups = Drug, layout = c(3,1), data = Mehta, xlab = "Site", scales = list(x = list(draw = FALSE)), key = simpleKey(levels(Mehta$Drug), lines = TRUE, corner = c(1,0)))) mehtaMix.fitted <- fitted(mehtaMix) for (i in 1:3) { seekViewport(trellis.vpname("panel", i, 1)) sapply(1:nlevels(Mehta$Drug), function(j) grid.lines(unit(1:22, "native"), unit(mehtaMix.fitted[Mehta$Drug == levels(Mehta$Drug)[j], i], "native"), gp = gpar(lty = j))) } @ \caption{Relative number of responses for the treatment and the control group for each site in the Mehta et al.~trial dataset together with the fitted values. The sites are sorted by the relative number of responses in the control group.} \label{fig:mehta} \end{figure} If also a random effect for the coefficient of $\textrm{Drug}$ is fitted, i.e.~$\bm{\beta}^s = (\beta^s_{\textrm{Intercept|Site}}, \beta^s_{\textrm{Drug|Site}})$, this is estimated by: <>= mehtaMix <- stepFlexmix(cbind(Response, Total - Response) ~ Drug | Site, model = FLXMRglm(family = "binomial"), k = 3, data = Mehta, nrep = 3, control = list(minprior = 0.04)) summary(mehtaMix) @ The BIC is smaller for the larger model and this indicates that the assumption of an equal drug effect for all centers is not confirmed by the data. Given Figure~\ref{fig:mehta} a two-component model with fixed treatment is also fitted to the data where site 15 is omitted: <>= Mehta.sub <- subset(Mehta, Site != 15) mehtaMix <- stepFlexmix(cbind(Response, Total - Response) ~ 1 | Site, model = FLXMRglmfix(family = "binomial", fixed = ~ Drug), data = Mehta.sub, k = 2, nrep = 3) summary(mehtaMix) @ %%----------------------------------------------------------------------- \subsubsection{Tribolium} A finite mixture of binomial regressions is fitted to the Tribolium dataset given in \cite{flexmix:Wang+Puterman:1998}. The data was collected to investigate whether the adult Tribolium species Castaneum has developed an evolutionary advantage to recognize and avoid eggs of its own species while foraging, as beetles of the genus Tribolium are cannibalistic in the sense that adults eat the eggs of their own species as well as those of closely related species. The experiment isolated a number of adult beetles of the same species and presented them with a vial of 150 eggs (50 of each type), the eggs being thoroughly mixed to ensure uniformity throughout the vial. The data gives the consumption data for adult Castaneum species. It reports the number of Castaneum, Confusum and Madens eggs, respectively, that remain uneaten after two day exposure to the adult beetles. Replicates 1, 2, and 3 correspond to different occasions on which the experiment was conducted. The data is visualized in Figure~\ref{fig:tribolium} and the model is given by: \begin{align*} H(\textrm{Remaining} \,|\, \textrm{Total}, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s(\textrm{Replicate}, \bm{\alpha}) \textrm{Bi}( \textrm{Remaining} \,|\, \textrm{Total}, \theta_s), \end{align*} with $\bm{\beta}^s = (\beta^s_{\textrm{Intercept}}, \bm{\beta}_{\textrm{Species}})$. This model is fitted with: <>= data("tribolium", package = "flexmix") TribMix <- stepFlexmix(cbind(Remaining, Total - Remaining) ~ 1, k = 2:3, model = FLXMRglmfix(fixed = ~ Species, family = "binomial"), concomitant = FLXPmultinom(~ Replicate), data = tribolium) @ The model which is selected as the best in \cite{flexmix:Wang+Puterman:1998} can be estimated with: <>= modelWang <- FLXMRglmfix(fixed = ~ I(Species == "Confusum"), family = "binomial") concomitantWang <- FLXPmultinom(~ I(Replicate == 3)) TribMixWang <- stepFlexmix(cbind(Remaining, Total - Remaining) ~ 1, data = tribolium, model = modelWang, concomitant = concomitantWang, k = 2) summary(refit(TribMixWang)) @ \begin{figure} \centering <>= clusters <- factor(clusters(TribMixWang), labels = paste("Cluster", 1:TribMixWang@k)) print(dotplot(Remaining/Total ~ factor(Replicate) | clusters, groups = Species, data = tribolium[rep(1:9, each = 3) + c(0:2)*9,], xlab = "Replicate", auto.key = list(corner = c(1,0)))) @ \caption{Relative number of remaining beetles for the number of replicate. The different panels are according to the cluster assignemnts based on the a-posteriori probabilities of the model suggested in \cite{flexmix:Wang+Puterman:1998}.} \label{fig:tribolium} \end{figure} \cite{flexmix:Wang+Puterman:1998} also considered a model where they omit one conspicuous observation. This model can be estimated with: <>= TribMixWangSub <- stepFlexmix(cbind(Remaining, Total - Remaining) ~ 1, k = 2, data = tribolium[-7,], model = modelWang, concomitant = concomitantWang) @ %%----------------------------------------------------------------------- \subsubsection{Trypanosome} The data is used in \cite{flexmix:Follmann+Lambert:1989}. It is from a dosage-response analysis where the proportion of organisms belonging to different populations shall be assessed. It is assumed that organisms belonging to different populations are indistinguishable other than in terms of their reaction to the stimulus. The experimental technique involved inspection under the microscope of a representative aliquot of a suspension, all organisms appearing within two fields of view being classified either alive or dead. Hence the total numbers of organisms present at each dose and the number showing the quantal response were both random variables. The data is illustrated in Figure~\ref{fig:trypanosome}. The model which is proposed in \cite{flexmix:Follmann+Lambert:1989} is given by: \begin{align*} H(\textrm{Dead} \,|\,\bm{\Theta}) &= \sum_{s = 1}^S \pi_s \textrm{Bi}( \textrm{Dead} \,|\, \theta_s), \end{align*} where $\textrm{Dead} \in \{0,1\}$ and with $\bm{\beta}^s = (\beta^s_{\textrm{Intercept}}, \bm{\beta}_{\textrm{log(Dose)}})$. This model is fitted with: <>= data("trypanosome", package = "flexmix") TrypMix <- stepFlexmix(cbind(Dead, 1-Dead) ~ 1, k = 2, nrep = 3, data = trypanosome, model = FLXMRglmfix(family = "binomial", fixed = ~ log(Dose))) summary(refit(TrypMix)) @ The fitted values are given in Figure~\ref{fig:trypanosome} together with the fitted values of a generalized linear model in order to facilitate comparison of the two models. \begin{figure} \centering <>= tab <- with(trypanosome, table(Dead, Dose)) Tryp.dat <- data.frame(Dead = tab["1",], Alive = tab["0",], Dose = as.numeric(colnames(tab))) plot(Dead/(Dead+Alive) ~ Dose, data = Tryp.dat) Tryp.pred <- predict(glm(cbind(Dead, 1-Dead) ~ log(Dose), family = "binomial", data = trypanosome), newdata=Tryp.dat, type = "response") TrypMix.pred <- predict(TrypMix, newdata = Tryp.dat, aggregate = TRUE)[[1]] lines(Tryp.dat$Dose, Tryp.pred, lty = 2) lines(Tryp.dat$Dose, TrypMix.pred, lty = 3) legend(4.7, 1, c("GLM", "Mixture model"), lty=c(2, 3), xjust=0, yjust=1) @ \caption{Relative number of deaths for each dose level together with the fitted values for the generalized linear model (``GLM'') and the random intercept model (``Mixture model'').} \label{fig:trypanosome} \end{figure} %%------------------------------------------------------------------------- \subsection{Finite mixtures of Poisson regressions} % %%----------------------------------------------------------------------- \subsubsection{Fabric faults} The dataset is analyzed using a finite mixture of Poisson regression models in \cite{flexmix:Aitkin:1996}. Furthermore, it is described in \cite{flexmix:McLachlan+Peel:2000} on page 155. It contains 32 observations on the number of faults in rolls of a textile fabric. A random intercept model is used where a fixed effect is assumed for the logarithm of length: <>= data("fabricfault", package = "flexmix") fabricMix <- stepFlexmix(Faults ~ 1, model = FLXMRglmfix(family="poisson", fixed = ~ log(Length)), data = fabricfault, k = 2, nrep = 3) summary(fabricMix) summary(refit(fabricMix)) Lnew <- seq(0, 1000, by = 50) fabricMix.pred <- predict(fabricMix, newdata = data.frame(Length = Lnew)) @ The intercept of the first component is not significantly different from zero for a signficance level of 0.05. We therefore also fit a modified model where the intercept is a-priori set to zero for the first component. This nested structure is given as part of the model specification with argument \code{nested}. <>= fabricMix2 <- flexmix(Faults ~ 0, data = fabricfault, cluster = posterior(fabricMix), model = FLXMRglmfix(family = "poisson", fixed = ~ log(Length), nested = list(k=c(1,1), formula=list(~0,~1)))) summary(refit(fabricMix2)) fabricMix2.pred <- predict(fabricMix2, newdata = data.frame(Length = Lnew)) @ The data and the fitted values for each of the components for both models are given in Figure~\ref{fig:fabric}. \begin{figure} \centering <>= plot(Faults ~ Length, data = fabricfault) sapply(fabricMix.pred, function(y) lines(Lnew, y, lty = 1)) sapply(fabricMix2.pred, function(y) lines(Lnew, y, lty = 2)) legend(190, 25, paste("Model", 1:2), lty=c(1, 2), xjust=0, yjust=1) @ \caption{Observed values of the fabric faults dataset together with the fitted values for the components of each of the two fitted models.} \label{fig:fabric} \end{figure} %%----------------------------------------------------------------------- \subsubsection{Patent} The patent data given in \cite{flexmix:Wang+Cockburn+Puterman:1998} consist of 70 observations on patent applications, R\&D spending and sales in millions of dollar from pharmaceutical and biomedical companies in 1976 taken from the National Bureau of Economic Research R\&D Masterfile. The observations are displayed in Figure~\ref{fig:patent}. The model which is chosen as the best in \cite{flexmix:Wang+Cockburn+Puterman:1998} is given by: \begin{align*} H(\textrm{Patents} \,|\, \textrm{lgRD}, \textrm{RDS}, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s(\textrm{RDS}, \bm{\alpha}) \textrm{Poi} ( \textrm{Patents} \,|\, \lambda_s), \end{align*} and $\bm{\beta}^s = (\beta^s_{\textrm{Intercept}}, \beta^s_{\textrm{lgRD}})$. The model is fitted with: <>= data("patent", package = "flexmix") ModelPat <- FLXMRglm(family = "poisson") FittedPat <- stepFlexmix(Patents ~ lgRD, k = 3, nrep = 3, model = ModelPat, data = patent, concomitant = FLXPmultinom(~ RDS)) summary(FittedPat) @ The fitted values for the component specific models and the concomitant variable model are given in Figure~\ref{fig:patent}. The plotting symbol of the observations corresponds to the induced clustering given by \code{clusters(FittedPat)}. This model is modified to have fixed effects for the logarithmized R\&D spendings, i.e.~$\bm(\beta)^s = (\beta^s_{\textrm{Intercept}}, \beta_{\textrm{lgRD}})$. The already fitted model is used for initialization, i.e.~the EM algorithm is started with an M-step given the a-posteriori probabilities. <>= ModelFixed <- FLXMRglmfix(family = "poisson", fixed = ~ lgRD) FittedPatFixed <- flexmix(Patents ~ 1, model = ModelFixed, cluster = posterior(FittedPat), concomitant = FLXPmultinom(~ RDS), data = patent) summary(FittedPatFixed) @ The fitted values for the component specific models and the concomitant variable model of this model are also given in Figure~\ref{fig:patent}. \begin{figure} \centering \setkeys{Gin}{width=0.95\textwidth} <>= lgRDv <- seq(-3, 5, by = 0.05) newdata <- data.frame(lgRD = lgRDv) plotData <- function(fitted) { with(patent, data.frame(Patents = c(Patents, unlist(predict(fitted, newdata = newdata))), lgRD = c(lgRD, rep(lgRDv, 3)), class = c(clusters(fitted), rep(1:3, each = nrow(newdata))), type = rep(c("data", "fit"), c(nrow(patent), nrow(newdata)*3)))) } plotPatents <- cbind(plotData(FittedPat), which = "Wang et al.") plotPatentsFixed <- cbind(plotData(FittedPatFixed), which = "Fixed effects") plotP <- rbind(plotPatents, plotPatentsFixed) rds <- seq(0, 3, by = 0.02) x <- model.matrix(FittedPat@concomitant@formula, data = data.frame(RDS = rds)) plotConc <- function(fitted) { E <- exp(x%*%fitted@concomitant@coef) data.frame(Probability = as.vector(E/rowSums(E)), class = rep(1:3, each = nrow(x)), RDS = rep(rds, 3)) } plotConc1 <- cbind(plotConc(FittedPat), which = "Wang et al.") plotConc2 <- cbind(plotConc(FittedPatFixed), which = "Fixed effects") plotC <- rbind(plotConc1, plotConc2) print(xyplot(Patents ~ lgRD | which, data = plotP, groups=class, xlab = "log(R&D)", panel = "panel.superpose", type = plotP$type, panel.groups = function(x, y, type = "p", subscripts, ...) { ind <- plotP$type[subscripts] == "data" panel.xyplot(x[ind], y[ind], ...) panel.xyplot(x[!ind], y[!ind], type = "l", ...) }, scales = list(alternating=FALSE), layout=c(1,2), as.table=TRUE), more=TRUE, position=c(0,0,0.6, 1)) print(xyplot(Probability ~ RDS | which, groups = class, data = plotC, type = "l", scales = list(alternating=FALSE), layout=c(1,2), as.table=TRUE), position=c(0.6, 0.01, 1, 0.99)) @ \caption{Patent data with the fitted values of the component specific models (left) and the concomitant variable model (right) for the model in \citeauthor{flexmix:Wang+Cockburn+Puterman:1998} and with fixed effects for $\log(\textrm{R\&D})$. The plotting symbol for each observation is determined by the component with the maximum a-posteriori probability.} \label{fig:patent} \end{figure} \setkeys{Gin}{width=0.8\textwidth} With respect to the BIC the full model is better than the model with the fixed effects. However, fixed effects have the advantage that the different components differ only in their baseline and the relation between the components in return of investment for each additional unit of R\&D spending is constant. Due to a-priori domain knowledge this model might seem more plausible. The fitted values for the constrained model are also given in Figure~\ref{fig:patent}. %%----------------------------------------------------------------------- \subsubsection{Seizure} The data is used in \cite{flexmix:Wang+Puterman+Cockburn:1996} and is from a clinical trial where the effect of intravenous gamma-globulin on suppression of epileptic seizures is studied. There are daily observations for a period of 140 days on one patient, where the first 27 days are a baseline period without treatment, the remaining 113 days are the treatment period. The model proposed in \cite{flexmix:Wang+Puterman+Cockburn:1996} is given by: \begin{align*} H(\textrm{Seizures} \,|\, (\textrm{Treatment}, \textrm{log(Day)}, \textrm{log(Hours)}), \bm{\Theta}) &= \sum_{s = 1}^S \pi_s \textrm{Poi} ( \textrm{Seizures} \,|\, \lambda_s), \end{align*} where $\bm(\beta)^s = (\beta^s_{\textrm{Intercept}}, \beta^s_{\textrm{Treatment}}, \beta^s_{\textrm{log(Day)}}, \beta^s_{\textrm{Treatment:log(Day)}})$ and $\textrm{log(Hours)}$ is used as offset. This model is fitted with: <>= data("seizure", package = "flexmix") seizMix <- stepFlexmix(Seizures ~ Treatment * log(Day), data = seizure, k = 2, nrep = 3, model = FLXMRglm(family = "poisson", offset = log(seizure$Hours))) summary(seizMix) summary(refit(seizMix)) @ A different model with different contrasts to directly estimate the coefficients for the jump when changing between base and treatment period is given by: <>= seizMix2 <- flexmix(Seizures ~ Treatment * log(Day/27), data = seizure, cluster = posterior(seizMix), model = FLXMRglm(family = "poisson", offset = log(seizure$Hours))) summary(seizMix2) summary(refit(seizMix2)) @ A different model which allows no jump at the change between base and treatment period is fitted with: <>= seizMix3 <- flexmix(Seizures ~ log(Day/27)/Treatment, data = seizure, cluster = posterior(seizMix), model = FLXMRglm(family = "poisson", offset = log(seizure$Hours))) summary(seizMix3) summary(refit(seizMix3)) @ With respect to the BIC criterion the smaller model with no jump is preferred. This is also the more intuitive model from a practitioner's point of view, as it does not seem to be plausible that starting the treatment already gives a significant improvement, but improvement develops over time. The data points together with the fitted values for each component of the two models are given in Figure~\ref{fig:seizure}. It can clearly be seen that the fitted values are nearly equal which also supports the smaller model. \begin{figure} \centering <>= plot(Seizures/Hours~Day, pch = c(1,3)[as.integer(Treatment)], data=seizure) abline(v=27.5, lty=2, col="grey") legend(140, 9, c("Baseline", "Treatment"), pch=c(1, 3), xjust=1, yjust=1) matplot(seizure$Day, fitted(seizMix)/seizure$Hours, type="l", add=TRUE, lty = 1, col = 1) matplot(seizure$Day, fitted(seizMix3)/seizure$Hours, type="l", add=TRUE, lty = 3, col = 1) legend(140, 7, paste("Model", c(1,3)), lty=c(1, 3), xjust=1, yjust=1) @ \caption{Observed values for the seizure dataset together with the fitted values for the components of the two different models.} \label{fig:seizure} \end{figure} %%----------------------------------------------------------------------- \subsubsection{Ames salmonella assay data} The ames salomnella assay dataset was used in \cite{flexmix:Wang+Puterman+Cockburn:1996}. They propose a model given by: \begin{align*} H(\textrm{y} \,|\, \textrm{x}, \bm{\Theta}) &= \sum_{s = 1}^S \pi_s \textrm{Poi} ( \textrm{y} \,|\, \lambda_s), \end{align*} where $\bm{\beta}^s = (\beta^s_{\textrm{Intercept}}, \beta_{\textrm{x}}, \beta_{\textrm{log(x+10)}})$. The model is fitted with: <>= data("salmonellaTA98", package = "flexmix") salmonMix <- stepFlexmix(y ~ 1, data = salmonellaTA98, k = 2, nrep = 3, model = FLXMRglmfix(family = "poisson", fixed = ~ x + log(x + 10))) @ \begin{figure} \centering <>= salmonMix.pr <- predict(salmonMix, newdata=salmonellaTA98) plot(y~x, data=salmonellaTA98, pch=as.character(clusters(salmonMix)), xlab="Dose of quinoline", ylab="Number of revertant colonies of salmonella", ylim=range(c(salmonellaTA98$y, unlist(salmonMix.pr)))) for (i in 1:2) lines(salmonellaTA98$x, salmonMix.pr[[i]], lty=i) @ \caption{Means and classification for assay data according to the estimated posterior probabilities based on the fitted model.} \label{fig:almes} \end{figure} %%----------------------------------------------------------------------- \section{Conclusions and future work} Package \pkg{flexmix} can be used to fit finite mixtures of regressions to datasets used in the literature to illustrate these models. The results can be reproduced and additional insights can be gained using visualization methods available in \proglang{R}. The fitted model is an object in \proglang{R} which can be explored using \code{show()}, \code{summary()} or \code{plot()}, as suitable methods have been implemented for objects of class \code{"flexmix"} which are returned by \code{flexmix()}. In the future it would be desirable to have more diagnostic tools available to analyze the model fit and compare different models. The use of resampling methods would be convenient as they can be applied to all kinds of mixtures models and would therefore suit well the purpose of the package which is flexible modelling of various finite mixture models. Furthermore, an additional visualization method for the fitted coefficients of the mixture would facilitate the comparison of the components. %%----------------------------------------------------------------------- \section*{Computational details} <>= SI <- sessionInfo() pkgs <- paste(sapply(c(SI$otherPkgs, SI$loadedOnly), function(x) paste("\\\\pkg{", x$Package, "} ", x$Version, sep = "")), collapse = ", ") @ All computations and graphics in this paper have been done using \proglang{R} version \Sexpr{getRversion()} with the packages \Sexpr{pkgs}. %%----------------------------------------------------------------------- \section*{Acknowledgments} This research was supported by the the Austrian Science Foundation (FWF) under grant P17382 and the Austrian Academy of Sciences ({\"O}AW) through a DOC-FFORTE scholarship for Bettina Gr{\"u}n. %%----------------------------------------------------------------------- \bibliography{flexmix} \end{document} flexmix/MD50000644000176200001440000001662113432625350012247 0ustar liggesusers75a923fbeb25f4f6313ce04e595879f2 *DESCRIPTION f053171213d6fb0ad87de6e9c811f414 *NAMESPACE 88a1e6330ac4ebbddc726522a46dfc34 *R/FLXMCdist1.R 0b00b7105930913d1c755a346dd4b048 *R/allClasses.R ebb73a41e988babd784db3f85a6b966e *R/allGenerics.R b03125231df0214b1395c7ef7f9bf25d *R/boot.R 889f52aae8df7342a12c941702964959 *R/concomitant.R ffe49b2c72bb7df7681bfd5f450468de *R/condlogit.R 3495256b72a55590ac6bd775f3782028 *R/examples.R 3ebb141d9c261a101e884f66f81a9120 *R/factanal.R 915eca9767df3030a613318f897e3ce1 *R/flexmix.R ac7d510575121174e050def18e1d8f68 *R/flexmixFix.R 3eba12e52c7117b1dc31709f780bace7 *R/flxdist.R e46cc7864775903c177d0fbebf62562f *R/flxmcmvpois.R 94cecec609f15f14b8eed8002bb5b792 *R/glmFix.R d49134e08be219e48ebaa680b76e647e *R/glmnet.R c5c5f8a325c65778e15792e456459ff2 *R/group.R f929b19e094c157d3556617cafd46a31 *R/infocrit.R 296476baa3f3e730cc906d1d1663c0a0 *R/initFlexmix.R d509a6bcd55ec42162950011f0e3fedb *R/kldiv.R 2ad7b23c815889ec724736b3b5b0f017 *R/lattice.R 49e5b8eef79f6f8884e4e0323329b78f *R/lmer.R dc418c5edbfb3d5371d0962534a1448e *R/lmm.R bd71d0d6bb0867c1957058595ff5800e *R/lmmc.R 5a7ccfa40918a1075c31b79b9b713c64 *R/mgcv.R 2d40a708cbab9a061aa4f2ab7e8c978a *R/models.R a1543ac30b4c4faca4ebdd23e5fca5f3 *R/multinom.R 7a1d8c2b82999b1bd1e0419b00117498 *R/plot-FLXboot.R 621a422936c28bdebd8fbbe1d9fe7970 *R/plot-flexmix.R 4a0e5ec24c4b8b048afd7473bbbf0cb4 *R/plot-refit.R 2f3409e7007c2e61fc5ee28aaef440d5 *R/plot.R ce8c2d2018c456648e26145ea1cbc038 *R/rFLXmodel.R 2dce928976d164087b04d9edba0771eb *R/refit.R 89f17e9ffcdd456af71b16580a64ab7c *R/relabel.R 942a774a6195a4cf5da06bfefc18adb6 *R/rflexmix.R aee09f6209b75aaf4f4aaea863fb9352 *R/robust.R 6c4c0f1663db73d9c2a8fd48ab78eb2e *R/stepFlexmix.R 06a4c9417ad23fc25ba52d1be38dee42 *R/utils.R 90d0c43f34243c3adbb022664bd0ddf5 *R/z.R 16c7cb035f39de0d66d086ecdb22fbe5 *R/ziglm.R 463111e9e25501b817418eafde3f5199 *build/vignette.rds 52f6bcc1bca4a592e3351b49431d9065 *data/BregFix.RData c1b30314976ba5d8428baa72c2c7472a *data/Mehta.RData 90b92021bfed03931e116746bd331d76 *data/NPreg.RData 33ab45d7bc0754237ecd01ce0254a604 *data/Nclus.RData cce71227ec0d604bcb2c78fcb6249384 *data/NregFix.RData b30604eae02b77396b87d0defb8d5c98 *data/betablocker.RData 45576be271f3423ba91c02ebfaca00b9 *data/bioChemists.RData 43669f632f40def80fbb61ca25c657fd *data/candy.RData a8bd5b28da6a6a50d872c0ff711d70f5 *data/dmft.RData 6556322b93e01104ed3485ee02993d6d *data/fabricfault.RData ca4de9fc56b9223d9612d44d3df1420e *data/patent.RData f53cbfa94cbb22c8ba6a8c5d54895e88 *data/salmonellaTA98.txt.gz aadfe3a142210960be9cfaa393a420ab *data/seizure.RData 309d62c6fbd66165dda517923e184109 *data/tribolium.RData c444a25f85d0973c69e2063fa5349b7d *data/trypanosome.RData 84ee26fbac9ffa0d64a1ed89ad17f751 *data/whiskey.RData 594304b06ad475bbc7c4b0cce89e0f96 *inst/CITATION 327f448a6881025040669de7df7a8bec *inst/NEWS.Rd efb89b5446ba221ffea93c9d9f10e1be *inst/doc/bootstrapping.R 842d21bfc3dfc1282005d1ae529efc3b *inst/doc/bootstrapping.Rnw 893bdbd8896e833b6ec27beb589a2ff5 *inst/doc/bootstrapping.pdf 02a13cd5fc2b4b13b805f54faef012a3 *inst/doc/flexmix-intro.R 21adc9c3f8293aa96460361171a050af *inst/doc/flexmix-intro.Rnw adcbdce9a64e98fd8d8327415940e75c *inst/doc/flexmix-intro.pdf 41c0fbb1eb98227cdc4ebacbd4208eb0 *inst/doc/mixture-regressions.R c82c2f7a526b7b2169d69f431f0848d1 *inst/doc/mixture-regressions.Rnw 986fa938c1fd22fa60ce69ab6f8f8707 *inst/doc/mixture-regressions.pdf 6dca73aa5a96343db67a43c9835bb4bf *inst/doc/myConcomitant.R fb9ae81acbfcada421bbd89681834714 *inst/doc/mymclust.R 7ae4d179976bf5f74cd2aea099ab4899 *inst/doc/regression-examples.R 525294b2c27e94ae1dadc683238b8e5f *inst/doc/regression-examples.Rnw d6f53882eb40529411d3e0427e824316 *inst/doc/regression-examples.pdf 76a28867c7670eb60951e3b1d2d77d33 *inst/doc/ziglm.R 1d39a7829ea8ce6aa54c11a0a06780eb *man/AIC-methods.Rd 00c18dd702febe081d951aa73ebb95d0 *man/BIC-methods.Rd cad198606d9e736399f5c5fc540786a1 *man/BregFix.Rd 85c040578c4b3ca0594b016b90336792 *man/EIC.Rd 85d7f2bae2833443d209050759a2e525 *man/ExLinear.Rd 7ed1926717bdc73c8e6b117c5075eb0e *man/ExNPreg.Rd cf4a61f3fbf44d92894525cbff4ef003 *man/ExNclus.Rd 4abb23a945f334576b2da3458d68784b *man/FLXMCdist1.Rd 4c28bd39db43ab9080cbfb2b2d495054 *man/FLXMCfactanal.Rd 692b3facd60fa4125aff97003555a621 *man/FLXMCmvcombi.Rd ba41986c3ff634c81b8f5b44955efe92 *man/FLXMCmvpois.Rd cd2d92661ee530052fd637930c3a1cdd *man/FLXMRcondlogit.Rd 4628f562cd45f409dd16ea2cc7d3eca4 *man/FLXMRglmnet.Rd 9986fd3a37b6a2f810d7200324a92524 *man/FLXMRlmer.Rd fcba2b7fc629a4f86356903bb1c38691 *man/FLXMRlmmc.Rd 94f4aa1616d758b3252d3f0b4b17842e *man/FLXMRmgcv.Rd 55014e2f3276e6b59c0ea5c5d7206bc6 *man/FLXMRmultinom.Rd 9683eb48c27e0a90738af42b16aa0820 *man/FLXMRrobglm.Rd a9623e0383c948ca9803a18b5528f090 *man/FLXMRziglm.Rd 1be70d781f7affbf5b51663f75fa87e0 *man/FLXP-class.Rd e5014d435eb288059e6daa8b9163d8e7 *man/FLXbclust.Rd c87d4220e578ac71869feed74a3d7c47 *man/FLXcomponent-class.Rd 1348b9444773021d770f7463579961ff *man/FLXconcomitant.Rd eff1590dbfbd4e3f1783fe61b20037d6 *man/FLXcontrol-class.Rd f8a1b4148b65b193d78ca1159e888a0d *man/FLXdist-class.Rd 28a563506d19c0836af092f0c3f15775 *man/FLXdist.Rd 47b0ea1622e710755d45534208ad95ea *man/FLXfit.Rd 2c91e46fd8c6120a6fd8bc4ca30a0651 *man/FLXglm.Rd b2fb5d0cee6a2e15ce31819dbd5f540b *man/FLXglmFix.Rd 24a39ce85331b9ddbb500d499932fee1 *man/FLXmclust.Rd ac520f781fbde7d2f327b842490b2aaf *man/FLXmodel-class.Rd afbfbeb61e91eda57fbaa568a39b38e2 *man/FLXnested-class.Rd 0768bed00f7a8554c33329522dfaa169 *man/ICL.Rd 028fc57ce85d724790cc7da759a24dd7 *man/KLdiv.Rd f6736b5d249d5d684f1464180b8a3b59 *man/Lapply-methods.Rd edab8de219b6630db7a71d22366f70d3 *man/Mehta.Rd 8f55051592bec6c88e23c7d81f57d38b *man/NregFix.Rd fb40ad5bde9b66ddc38d3572ff7429cb *man/betablocker.Rd 4c7ac0b259ec58b72764a20f3a2dff01 *man/bioChemists.Rd fdb60dd3f31d2281e7fa21e38a3c8b76 *man/boot.Rd b4db12be29b542f9bea083d3bd0c980e *man/candy.Rd 5a69cda63c913ef33f104c6d5adb3729 *man/dmft.Rd 78db5196994b9a665e4c17fb6653c93f *man/fabricfault.Rd cd0b0ac588e0bd163a930ddbd0d8ae4f *man/fitted.Rd b53ba6892c3f8d98ea7ee048c854a47d *man/flexmix-class.Rd 67e00b3127b226f15febe7a54124769f *man/flexmix-internal.Rd 57ab6985b4c34528329c50c5bd67b30c *man/flexmix.Rd 83812639242cedc8bf5bc226f2747c88 *man/group.Rd a187b88ae5b3121c29579aceef270951 *man/logLik-methods.Rd 2b624ed0faa1b3345438a9da35f2f78d *man/patent.Rd 3aae6929d9c9dbe26b4be487330edada *man/plot-methods.Rd b3773bddc048e6fbe106eed18469702d *man/plotEll.Rd 2195e1c45f80a83a7b3baf8d449e5b05 *man/posterior.Rd f8ae3162f10350e983e30a5926fa6025 *man/refit.Rd d627472edfd9f5b2c3ec8e0f56296ae4 *man/relabel.Rd cf67fee780ab0ea895b55f07989aa76c *man/rflexmix.Rd 2aa3384e40b6d464a252378e17f08b1c *man/salmonellaTA98.Rd 4d165b442083a5ba599fa8e1c5e5f28a *man/seizure.Rd 4d457ddbe589788191b3ee5021e0ac6d *man/stepFlexmix.Rd 168fda1db5cdaf8e9d0a5410c145e825 *man/tribolium.Rd b5fad3c3973f6658af8b3663f20f4ac3 *man/trypanosome.Rd 4381e5944c15dc03c92a2151c049fca3 *man/whiskey.Rd 842d21bfc3dfc1282005d1ae529efc3b *vignettes/bootstrapping.Rnw 21adc9c3f8293aa96460361171a050af *vignettes/flexmix-intro.Rnw 509bb7f1d5c2cd61b6b5088e912bd5ef *vignettes/flexmix.bib 0a928bdb28d680f31098902b157e9204 *vignettes/flexmix.png c82c2f7a526b7b2169d69f431f0848d1 *vignettes/mixture-regressions.Rnw b1012b8287285f2b470111025699cce5 *vignettes/mixture.bib 6dca73aa5a96343db67a43c9835bb4bf *vignettes/myConcomitant.R fb9ae81acbfcada421bbd89681834714 *vignettes/mymclust.R 525294b2c27e94ae1dadc683238b8e5f *vignettes/regression-examples.Rnw 76a28867c7670eb60951e3b1d2d77d33 *vignettes/ziglm.R flexmix/build/0000755000176200001440000000000013432516316013031 5ustar liggesusersflexmix/build/vignette.rds0000644000176200001440000000077313432516316015377 0ustar liggesusers‹ÍT=oÛ0Ul'© m‘¦éxÀŠ.ÍÄP†Ö@`4I׳trˆJ¤@2°Ó©?<¨{’HëJ¦…!‰¼G¿w¼{äIƒ`x8COøuÄÏ[~Füó÷ÝJ)k¬Æ<r=[Ê’”¶™ØN…´Z5€3ÚMSMkMÆ%M®ÃSÚb–§TÁmÝ/¡Â,*2X¨˜R˜ \Ke¬ˆ ÜN–d Žb¸ {¯bãîBÎÿ}pE’4¦jÌh£ôOH”†>(cø†–¤…ËaŸ/ KGŽnI—اó¡°÷p©d¤2a‘ oQ \ñ†Kž=yc^eÊ%×Xäh™Ô }¾Èy{Ú¢Ž H*•Ì«p¨.)då.^nb'ýM¬Þ&6à¾&p[÷MÇ<>Ü±Ž Ÿöǃ½¶aà¤Ô«tÜâc'ð¯ÂM ¿·ŸÙûvCª8l¢÷öt‘ñ¾$Ïñöñy7|(g°BC1Déƒa§Ôägieܨ4n]±ÿ@ê¥*}\WçSüb%®¡náÏk†J>9ÞúG.8 EJ~Áwa÷“áõ<ì&y<§œäþ yõ•ù¶ày[h¬ÕfæÅ^—èo~ív»§nFeÆ_Mb´8KŠcγ?eft£†flexmix/DESCRIPTION0000644000176200001440000000342413432625347013450 0ustar liggesusersPackage: flexmix Type: Package Title: Flexible Mixture Modeling Version: 2.3-15 Authors@R: c(person("Bettina", "Gruen", role = c("aut", "cre"), email = "Bettina.Gruen@jku.at", comment = c(ORCID = "0000-0001-7265-4773")), person("Friedrich", "Leisch", role = "aut", comment = c(ORCID = "0000-0001-7278-1983")), person("Deepayan", "Sarkar", role = "ctb", comment = c(ORCID = "0000-0003-4107-1553")), person("Frederic", "Mortier", role = "ctb"), person("Nicolas", "Picard", role = "ctb", comment = c(ORCID = "0000-0001-5548-9171"))) Description: A general framework for finite mixtures of regression models using the EM algorithm is implemented. The E-step and all data handling are provided, while the M-step can be supplied by the user to easily define new models. Existing drivers implement mixtures of standard linear models, generalized linear models and model-based clustering. Depends: R (>= 2.15.0), lattice Imports: graphics, grid, grDevices, methods, modeltools (>= 0.2-16), nnet, stats, stats4, utils Suggests: actuar, codetools, diptest, Ecdat, ellipse, gclus, glmnet, lme4 (>= 1.1), MASS, mgcv (>= 1.8-0), mlbench, multcomp, mvtnorm, SuppDists, survival License: GPL (>= 2) LazyLoad: yes NeedsCompilation: no Packaged: 2019-02-18 11:53:54 UTC; gruen Author: Bettina Gruen [aut, cre] (), Friedrich Leisch [aut] (), Deepayan Sarkar [ctb] (), Frederic Mortier [ctb], Nicolas Picard [ctb] () Maintainer: Bettina Gruen Repository: CRAN Date/Publication: 2019-02-18 22:00:07 UTC flexmix/man/0000755000176200001440000000000013431367112012502 5ustar liggesusersflexmix/man/FLXMRcondlogit.Rd0000644000176200001440000000435613425024340015570 0ustar liggesusers\name{FLXMRcondlogit} \alias{FLXMRcondlogit} \title{FlexMix Interface to Conditional Logit Models} \description{ Model driver for fitting mixtures of conditional logit models. } \usage{ FLXMRcondlogit(formula = . ~ ., strata) } \arguments{ \item{formula}{A formula which is interpreted relative to the formula specified in the call to \code{\link{flexmix}} using \code{\link{update.formula}}. Default is to use the original \code{\link{flexmix}} model formula.} \item{strata}{A formula which is interpreted such that no intercept is fitted and which allows to determine the variable indicating which observations are from the same stratum.} } \details{ The M-step is performed using \code{coxph.fit}. } \value{ Returns an object of class \code{FLXMRcondlogit}. } \references{ Bettina Gruen and Friedrich Leisch. Identifiability of finite mixtures of multinomial logit models with varying and fixed effects. \emph{Journal of Classification}, \bold{25}, 225--247. 2008. } \author{ Bettina Gruen } \section{Warning}{ To ensure identifiability repeated measurements are necessary. Sufficient conditions are given in Gruen and Leisch (2008). } \seealso{\code{\link{FLXMRmultinom}}} \examples{ if (require("Ecdat")) { data("Catsup", package = "Ecdat") ## To reduce the time needed for the example only a subset is used Catsup <- subset(Catsup, id \%in\% 1:100) Catsup$experiment <- seq_len(nrow(Catsup)) vnames <- c("display", "feature", "price") Catsup_long <- reshape(Catsup, idvar = c("id", "experiment"), times = c(paste("heinz", c(41, 32, 28), sep = ""), "hunts32"), timevar = "brand", varying = matrix(colnames(Catsup)[2:13], nrow = 3, byrow = TRUE), v.names = vnames, direction = "long") Catsup_long$selected <- with(Catsup_long, choice == brand) Catsup_long <- Catsup_long[, c("id", "selected", "experiment", vnames, "brand")] Catsup_long$brand <- relevel(factor(Catsup_long$brand), "hunts32") set.seed(0808) flx1 <- flexmix(selected ~ display + feature + price + brand | id, model = FLXMRcondlogit(strata = ~ experiment), data = Catsup_long, k = 1) } } \keyword{regression} \keyword{models} flexmix/man/stepFlexmix.Rd0000644000176200001440000001451413425024236015306 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: stepFlexmix.Rd 5115 2017-04-07 08:18:13Z gruen $ % \name{stepFlexmix} \alias{stepFlexmix} \alias{initFlexmix} \alias{initMethod} \alias{stepFlexmix-class} \alias{initMethod-class} \alias{plot,stepFlexmix,missing-method} \alias{show,stepFlexmix-method} \alias{getModel,stepFlexmix-method} \alias{unique,stepFlexmix-method} \title{Run FlexMix Repeatedly} \description{ Runs flexmix repeatedly for different numbers of components and returns the maximum likelihood solution for each. } \usage{ initFlexmix(..., k, init = list(), control = list(), nrep = 3L, verbose = TRUE, drop = TRUE, unique = FALSE) initMethod(name = c("tol.em", "cem.em", "sem.em"), step1 = list(tolerance = 10^-2), step2 = list(), control = list(), nrep = 3L) stepFlexmix(..., k = NULL, nrep = 3, verbose = TRUE, drop = TRUE, unique = FALSE) \S4method{plot}{stepFlexmix,missing}(x, y, what = c("AIC", "BIC", "ICL"), xlab = NULL, ylab = NULL, legend = "topright", ...) \S4method{getModel}{stepFlexmix}(object, which = "BIC") \S4method{unique}{stepFlexmix}(x, incomparables = FALSE, ...) } \arguments{ \item{\dots}{Passed to \code{\link{flexmix}} (or \code{\link{matplot}} in the \code{plot} method).} \item{k}{A vector of integers passed in turn to the \code{k} argument of \code{\link{flexmix}}.} \item{init}{An object of class \code{"initMethod"} or a named list where \code{initMethod} is called with it as arguments in addition to the \code{control} argument.} \item{name}{A character string indication which initialization strategy should be employed: short runs of EM followed by a long (\code{"tol.em"}), short runs of CEM followed by a long EM run (\code{"cem.em"}), short runs of SEM followed by a long EM run (\code{"sem.em"}).} \item{step1}{A named list which combined with the \code{control} argument is coercable to a \code{"FLXcontrol"} object. This control setting is used for the short runs.} \item{step2}{A named list which combined with the \code{control} argument is coercable to a \code{"FLXcontrol"} object. This control setting is used for the long run.} \item{control}{A named list which combined with the \code{step1} or the \code{step2} argument is coercable to a \code{"FLXcontrol"} object.} \item{nrep}{For each value of \code{k} run \code{\link{flexmix}} \code{nrep} times and keep only the solution with maximum likelihood. If \code{nrep} is set for the long run, it is ignored, because the EM algorithm is deterministic using the best solution discovered in the short runs for initialization.} \item{verbose}{If \code{TRUE}, show progress information during computations.} \item{drop}{If \code{TRUE} and \code{k} is of length 1, then a single flexmix object is returned instead of a \code{"stepFlexmix"} object.} \item{unique}{If \code{TRUE}, then \code{unique()} is called on the result, see below.} \item{x, object}{An object of class \code{"stepFlexmix"}.} \item{y}{Not used.} \item{what}{Character vector naming information criteria to plot. Functions of the same name must exist, which take a \code{stepFlexmix} object as input and return a numeric vector like \code{AIC,stepFlexmix-method} (see examples below).} \item{xlab,ylab}{Graphical parameters.} \item{legend}{If not \code{FALSE} and \code{what} contains more than 1 element, a legend is placed at the specified location, see \code{\link{legend}} for details.} \item{which}{Number of model to get. If character, interpreted as number of components or name of an information criterion.} \item{incomparables}{A vector of values that cannot be compared. Currently, \code{FALSE} is the only possible value, meaning that all values can be compared.} } \value{ An object of class \code{"stepFlexmix"} containing the best models with respect to the log likelihood for the different number of components in a slot if \code{length(k)>1}, else directly an object of class \code{"flexmix"}. If \code{unique = FALSE}, then the resulting object contains one model per element of \code{k} (which is the number of clusters the EM algorithm started with). If \code{unique = TRUE}, then the result is resorted according to the number of clusters contained in the fitted models (which may be less than the number with which the EM algorithm started), and only the maximum likelihood solution for each number of fitted clusters is kept. This operation can also be done manually by calling \code{unique()} on objects of class \code{"stepFlexmix"}. } \author{Friedrich Leisch and Bettina Gruen} \references{ Friedrich Leisch. FlexMix: A general framework for finite mixture models and latent class regression in R. \emph{Journal of Statistical Software}, \bold{11}(8), 2004. doi:10.18637/jss.v011.i08 Christophe Biernacki, Gilles Celeux and Gerard Govaert. Choosing starting values for the EM algorithm for getting the highest likelihood in multivariate Gaussian mixture models. \emph{Computational Statistics & Data Analysis}, \bold{41}(3--4), 561--575, 2003. Theresa Scharl, Bettina Gruen and Friedrch Leisch. Mixtures of regression models for time-course gene expression data: Evaluation of initialization and random effects. \emph{Bioinformatics}, \bold{26}(3), 370--377, 2010. } \examples{ data("Nclus", package = "flexmix") ## try 2 times for k = 4 set.seed(511) ex1 <- initFlexmix(Nclus~1, k = 4, model = FLXMCmvnorm(diagonal = FALSE), nrep = 2) ex1 ## now 2 times each for k = 2:5, specify control parameter ex2 <- initFlexmix(Nclus~1, k = 2:5, model = FLXMCmvnorm(diagonal = FALSE), control = list(minprior = 0), nrep = 2) ex2 plot(ex2) ## get BIC values BIC(ex2) ## get smallest model getModel(ex2, which = 1) ## get model with 3 components getModel(ex2, which = "3") ## get model with smallest ICL (here same as for AIC and BIC: true k = 4) getModel(ex2, which = "ICL") ## now 1 time each for k = 2:5, with larger minimum prior ex3 <- initFlexmix(Nclus~1, k = 2:5, model = FLXMCmvnorm(diagonal = FALSE), control = list(minprior = 0.1), nrep = 1) ex3 ## keep only maximum likelihood solution for each unique number of ## fitted clusters: unique(ex3) } \keyword{cluster} \keyword{regression} flexmix/man/ExLinear.Rd0000644000176200001440000000560513425024236014506 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: ExNclus.Rd 3912 2008-03-13 15:10:24Z gruen $ % \name{ExLinear} \alias{ExLinear} \title{Artificial Data from a Generalized Linear Regression Mixture} \description{ Generate random data mixed from k generalized linear regressions (GLMs). } \usage{ ExLinear(beta, n, xdist = "runif", xdist.args = NULL, family = c("gaussian","poisson"), sd = 1, ...) } \arguments{ \item{beta}{A matrix of regression coefficients. Each row corresponds to a variable, each column to a mixture component. The first row is used as intercept.} \item{n}{Integer, the number of observations per component.} \item{xdist}{Character, name of a random number function for the explanatory variables.} \item{xdist.args}{List, arguments for the random number functions.} \item{family}{A character string naming a GLM family.Only \code{"gaussian"} and \code{"poisson"} are implemented at the moment.} \item{sd}{Numeric, the error standard deviation for each component for Gaussian responses.} \item{\dots}{Used as default for \code{xdist.args} if that is not specified.} } \details{ First, arguments \code{n} (and \code{sd} for Gaussian response) are recycled to the number of mixture components \code{ncol(beta)}, and arguments \code{xdist} and \code{xdist.args} are recycled to the number of explanatory variables \code{nrow(beta)-1}. Then a design matrix is created for each mixture component by drawing random numbers from \code{xdist}. For each component, the design matrix is multiplied by the regression coefficients to form the linear predictor. For Gaussian responses the identity link is used, for Poisson responses the log link. The true cluster memberships are returned as attribute \code{"clusters"}. } \examples{ ## simple example in 2d beta <- matrix(c(1, 2, 3, -1), ncol = 2) sigma <- c(.5, 1) df1 <- ExLinear(beta, 100, sd = sigma, min = -1, max = 2) plot(y~x1, df1, col = attr(df1, "clusters")) ## add a second explanatory variable with exponential distribution beta2 <- rbind(beta, c(-2, 2)) df2 <- ExLinear(beta2, 100, sd = c(.5, 1), xdist = c("runif", "rexp"), xdist.args = list(list(min = -1, max = 2), list(rate = 3))) summary(df2) opar = par("mfrow") par(mfrow = 1:2) hist(df2$x1) hist(df2$x2) par(opar) f1 <- flexmix(y ~ ., data = df2, k = 2) ## sort components on Intercept f1 <- relabel(f1, "model", "Intercept") ## the parameters should be close to the true beta and sigma round(parameters(f1), 3) rbind(beta2, sigma) ### A simple Poisson GLM df3 <- ExLinear(beta/2, 100, min = -1, max = 2, family = "poisson") plot(y ~ x1, df3, col = attr(df3, "clusters")) f3 <- flexmix(y ~ ., data = df3, k = 2, model = FLXMRglm(family = "poisson")) round(parameters(relabel(f3, "model", "Intercept")), 3) beta/2 } \keyword{datasets} flexmix/man/FLXMRziglm.Rd0000644000176200001440000000276113425024236014732 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: FLXMRziglm.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{FLXMRziglm} \alias{FLXMRziglm} \alias{FLXMRziglm-class} \alias{refit,FLXMRziglm-method} \alias{FLXreplaceParameters,FLXMRziglm-method} \alias{FLXgradlogLikfun,FLXMRziglm-method} \title{FlexMix Interface to Zero Inflated Generalized Linear Models} \description{ This is a driver which allows fitting of zero inflated poisson and binomial models. } \usage{ FLXMRziglm(formula = . ~ ., family = c("binomial", "poisson"), ...) } \arguments{ \item{formula}{A formula which is interpreted relative to the formula specified in the call to \code{flexmix} using \code{\link{update.formula}}. Default is to use the original \code{flexmix} model formula.} \item{family}{A character string naming a \code{\link{glm}} family function.} \item{\dots}{passed to \code{FLXMRglm}} } \value{ Returns an object of class \code{FLXMRziglm}. } \author{Friedrich Leisch and Bettina Gruen} \note{ In fact this only approximates zero inflated models by fixing the coefficient of the intercept at -Inf and the other coefficients at zero for the first component. } \examples{ data("dmft", package = "flexmix") Model <- FLXMRziglm(family = "poisson") Fitted <- flexmix(End ~ log(Begin + 0.5) + Gender + Ethnic + Treatment, model = Model, k = 2 , data = dmft, control = list(minprior = 0.01)) summary(refit(Fitted)) } \keyword{models} flexmix/man/rflexmix.Rd0000644000176200001440000000315613431370572014640 0ustar liggesusers\name{rflexmix} \alias{rflexmix} \alias{rflexmix,flexmix,missing-method} \alias{rflexmix,FLXdist,numeric-method} \alias{rflexmix,FLXdist,listOrdata.frame-method} \alias{rFLXM} \alias{rFLXM,FLXM,list-method} \alias{rFLXM,FLXMC,FLXcomponent-method} \alias{rFLXM,FLXMCmultinom,FLXcomponent-method} \alias{rFLXM,FLXMCbinom,FLXcomponent-method} \alias{rFLXM,FLXMRglm,list-method} \alias{rFLXM,FLXMRglmfix,list-method} \alias{rFLXM,FLXM,FLXcomponent-method} \alias{rFLXM,FLXMRglm,FLXcomponent-method} \title{Random Number Generator for Finite Mixtures} \description{ Given a finite mixture model generate random numbers from it. } \usage{ rflexmix(object, newdata, ...) } \arguments{ \item{object}{A fitted finite mixture model of class \code{flexmix} or an unfitted of class \code{FLXdist}.} \item{newdata}{Optionally, a data frame in which to look for variables with which to predict or an integer specifying the number of random draws for model-based clustering. If omitted, the data to which the model was fitted is used.} \item{\dots}{Further arguments to be passed to or from methods.} } \details{ \code{rflexmix} provides the creation of the model matrix for new data and the sampling of the cluster memberships. The sampling of the component distributions given the classification is done by calling \code{rFLXM}. This step has to be provided for the different model classes. } \value{ A list with components \item{y}{Random sample} \item{group}{Grouping factor} \item{class}{Class membership} } \author{Bettina Gruen} \examples{ example(flexmix) sample <- rflexmix(ex1) } \keyword{distribution} flexmix/man/AIC-methods.Rd0000644000176200001440000000104613425024236015027 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: AIC-methods.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{AIC-methods} \docType{methods} \title{Methods for Function AIC} \alias{AIC,flexmix-method} \alias{AIC,stepFlexmix-method} \description{Compute the Akaike Information Criterion.} \section{Methods}{ \describe{ \item{object = flexmix:}{Compute the AIC of a \code{flexmix} object} \item{object = stepFlexmix:}{Compute the AIC of all models contained in the \code{stepFlexmix} object.} } } \keyword{methods} flexmix/man/ExNclus.Rd0000644000176200001440000000151313425024236014352 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: ExNclus.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{ExNclus} \alias{ExNclus} \alias{Nclus} \title{Artificial Example with 4 Gaussians} \description{ A simple artificial example for normal clustering with 4 latent classes, all of them having a Gaussian distribution. See the function definition for true means and covariances. } \usage{ ExNclus(n) data("Nclus") } \arguments{ \item{n}{Number of observations in the two small latent classes.} } \details{ The \code{Nclus} data set can be re-created by \code{ExNclus(100)} using \code{set.seed(2602)}, it has been saved as a data set for simplicity of examples only. } \examples{ data("Nclus", package = "flexmix") require("MASS") eqscplot(Nclus, col = rep(1:4, c(100, 100, 150, 200))) } \keyword{datasets} flexmix/man/FLXMCmvcombi.Rd0000644000176200001440000000407613425024236015226 0ustar liggesusers% % Copyright (C) 2009 Friedrich Leisch and Bettina Gruen % $Id: FLXMCmvcombi.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{FLXMCmvcombi} \alias{FLXMCmvcombi} \title{FlexMix Binary and Gaussian Clustering Driver} \description{ This is a model driver for \code{\link{flexmix}} implementing model-based clustering of a combination of binary and Gaussian data. } \usage{ FLXMCmvcombi(formula = . ~ .) } \arguments{ \item{formula}{A formula which is interpreted relative to the formula specified in the call to \code{\link{flexmix}} using \code{\link{update.formula}}. Only the left-hand side (response) of the formula is used. Default is to use the original \code{\link{flexmix}} model formula.} } \details{ This model driver can be used to cluster mixed-mode binary and Gaussian data. It checks which columns of a matrix contain only zero and ones, and does the same as \code{\link{FLXMCmvbinary}} for them. For the remaining columns of the data matrix independent Gaussian distributions are used (same as \code{\link{FLXMCmvnorm}} with \code{diagonal = FALSE}. The same could be obtained by creating a corresponding list of two models for the respective columns, but \code{FLXMCmvcombi} does a better job in reporting parameters. } \value{ \code{FLXMCmvcombi} returns an object of class \code{FLXMC}. } \author{Friedrich Leisch} \seealso{\code{\link{flexmix}}, \code{\link{FLXMCmvbinary}}, \code{\link{FLXMCmvnorm}}} \keyword{cluster} \examples{ ## create some artificial data x1 <- cbind(rnorm(300), sample(0:1, 300, replace = TRUE, prob = c(0.25, 0.75))) x2 <- cbind(rnorm(300, mean = 2, sd = 0.5), sample(0:1, 300, replace = TRUE, prob = c(0.75, 0.25))) x <- rbind(x1, x2) ## fit the model f1 <- flexmix(x ~ 1, k = 2, model = FLXMCmvcombi()) ## should be similar to the original parameters parameters(f1) table(clusters(f1), rep(1:2, c(300,300))) ## a column with noise should not hurt too much x <- cbind(x, rnorm(600)) f2 <- flexmix(x ~ 1, k = 2, model = FLXMCmvcombi()) parameters(f2) table(clusters(f2), rep(1:2, c(300,300))) } flexmix/man/FLXmclust.Rd0000644000176200001440000000465313425024236014662 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: FLXmclust.Rd 5115 2017-04-07 08:18:13Z gruen $ % \name{FLXMCmvnorm} \alias{FLXMCmvnorm} \alias{FLXMCnorm1} \alias{FLXmclust} \title{FlexMix Clustering Demo Driver} \description{ These are demo drivers for \code{\link{flexmix}} implementing model-based clustering of Gaussian data. } \usage{ FLXMCmvnorm(formula = . ~ ., diagonal = TRUE) FLXMCnorm1(formula = . ~ .) } \arguments{ \item{formula}{A formula which is interpreted relative to the formula specified in the call to \code{\link{flexmix}} using \code{\link{update.formula}}. Only the left-hand side (response) of the formula is used. Default is to use the original \code{\link{flexmix}} model formula.} \item{diagonal}{If \code{TRUE}, then the covariance matrix of the components is restricted to diagonal matrices.} } \details{ This is mostly meant as a demo for FlexMix driver programming, you should also look at package \pkg{mclust} for real applications. \code{FLXMCmvnorm} clusters multivariate data, \code{FLXMCnorm1} univariate data. In the latter case smart initialization is important, see the example below. } \value{ \code{FLXMCmvnorm} returns an object of class \code{FLXMC}. } \author{Friedrich Leisch and Bettina Gruen} \references{ Friedrich Leisch. FlexMix: A general framework for finite mixture models and latent class regression in R. \emph{Journal of Statistical Software}, \bold{11}(8), 2004. doi:10.18637/jss.v011.i08 } \seealso{\code{\link{flexmix}}} \keyword{cluster} \examples{ data("Nclus", package = "flexmix") require("MASS") eqscplot(Nclus) ## This model is wrong (one component has a non-diagonal cov matrix) ex1 <- flexmix(Nclus ~ 1, k = 4, model = FLXMCmvnorm()) print(ex1) plotEll(ex1, Nclus) ## True model, wrong number of components ex2 <- flexmix(Nclus ~ 1, k = 6, model = FLXMCmvnorm(diagonal = FALSE)) print(ex2) plotEll(ex2, Nclus) ## Get parameters of first component parameters(ex2, component = 1) ## Have a look at the posterior probabilies of 10 random observations ok <- sample(1:nrow(Nclus), 10) p <- posterior(ex2)[ok, ] p ## The following two should be the same max.col(p) clusters(ex2)[ok] \testonly{ stopifnot(all.equal(max.col(p), clusters(ex2)[ok])) } ## Now try the univariate case plot(density(Nclus[, 1])) ex3 <- flexmix(Nclus[, 1] ~ 1, cluster = cut(Nclus[, 1], 3), model = FLXMCnorm1()) ex3 parameters(ex3) } flexmix/man/FLXmodel-class.Rd0000644000176200001440000000320313425024236015544 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: FLXmodel-class.Rd 5079 2016-01-31 12:21:12Z gruen $ % \name{FLXM-class} \docType{class} \alias{FLXM-class} \alias{FLXMC-class} \alias{FLXMR-class} \alias{show,FLXM-method} \title{Class "FLXM"} \description{FlexMix model specification.} \section{Objects from the Class}{ Objects can be created by calls of the form \code{new("FLXM", ...)}, typically inside driver functions like \code{\link{FLXMRglm}} or \code{\link{FLXMCmvnorm}}. } \section{Slots}{ \describe{ \item{\code{fit}:}{Function returning an \code{FLXcomponent} object.} \item{\code{defineComponent}:}{Function or expression to determine the \code{FLXcomponent} object given the parameters.} \item{\code{weighted}:}{Logical indicating whether \code{fit} can do weighted likelihood maximization.} \item{\code{name}:}{Character string used in print methods.} \item{\code{formula}:}{Formula describing the model.} \item{\code{fullformula}:}{Resulting formula from updating the model formula with the formula specified in the call to \code{flexmix}.} \item{\code{x}:}{Model matrix.} \item{\code{y}:}{Model response.} \item{\code{terms}, \code{xlevels}, \code{contrasts}:}{Additional information for model matrix.} \item{\code{preproc.x}:}{Function for preprocessing matrix \code{x} before the EM algorithm starts, by default the identity function.} \item{\code{preproc.y}:}{Function for preprocessing matrix \code{y} before the EM algorithm starts, by default the identity function.} } } \author{Friedrich Leisch and Bettina Gruen} \keyword{classes} flexmix/man/BIC-methods.Rd0000644000176200001440000000105013425024236015023 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: BIC-methods.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{BIC-methods} \docType{methods} \title{Methods for Function BIC} \alias{BIC,flexmix-method} \alias{BIC,stepFlexmix-method} \description{Compute the Bayesian Information Criterion.} \section{Methods}{ \describe{ \item{object = flexmix:}{Compute the BIC of a \code{flexmix} object} \item{object = stepFlexmix:}{Compute the BIC of all models contained in the \code{stepFlexmix} object.} } } \keyword{methods} flexmix/man/plot-methods.Rd0000644000176200001440000000727613425024236015424 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: plot-methods.Rd 5115 2017-04-07 08:18:13Z gruen $ % \name{plot-methods} \docType{methods} \alias{plot-methods} \alias{plot,flexmix,missing-method} \title{Rootogram of Posterior Probabilities} \description{ The \code{plot} method for \code{\link{flexmix-class}} objects gives a rootogram or histogram of the posterior probabilities. } \usage{ \S4method{plot}{flexmix,missing}(x, y, mark = NULL, markcol = NULL, col = NULL, eps = 1e-4, root = TRUE, ylim = TRUE, main = NULL, xlab = "", ylab = "", as.table = TRUE, endpoints = c(-0.04, 1.04), ...) } \arguments{ \item{x}{An object of class \code{"flexmix"}.} \item{y}{Not used.} \item{mark}{Integer: mark posteriors of this component.} \item{markcol}{Color used for marking components.} \item{col}{Color used for the bars.} \item{eps}{Posteriors smaller than \code{eps} are ignored.} \item{root}{If \code{TRUE}, a rootogram of the posterior probabilities is drawn, otherwise a standard histogram.} \item{ylim}{A logical value or a numeric vector of length 2. If \code{TRUE}, the y axes of all rootograms are aligned to have the same limits, if \code{FALSE} each y axis is scaled separately. If a numeric vector is specified it is used as usual.} \item{main}{Main title of the plot.} \item{xlab}{Label of x-axis.} \item{ylab}{Label of y-axis.} \item{as.table}{Logical that controls the order in which panels should be plotted: if \code{FALSE} (the default), panels are drawn left to right, bottom to top (as in a graph); if \code{TRUE}, left to right, top to bottom.} \item{endpoints}{Vector of length 2 indicating the range of x-values that is to be covered by the histogram. This applies only when \code{breaks} is unspecified. In \code{do.breaks}, this specifies the interval that is to be divided up.} \item{...}{Further graphical parameters for the lattice function histogram.} } \details{ For each mixture component a rootogram or histogram of the posterior probabilities of all observations is drawn. Rootograms are very similar to histograms, the only difference is that the height of the bars correspond to square roots of counts rather than the counts themselves, hence low counts are more visible and peaks less emphasized. Please note that the y-axis denotes the number of observations in each bar in any case. Usually in each component a lot of observations have posteriors close to zero, resulting in a high count for the corresponding bin in the rootogram which obscures the information in the other bins. To avoid this problem, all probabilities with a posterior below \code{eps} are ignored. A peak at probability one indicates that a mixture component is well seperated from the other components, while no peak at one and/or significant mass in the middle of the unit interval indicates overlap with other components. } \references{ Friedrich Leisch. FlexMix: A general framework for finite mixture models and latent class regression in R. \emph{Journal of Statistical Software}, \bold{11}(8), 2004. doi:10.18637/jss.v011.i08 Jeremy Tantrum, Alejandro Murua and Werner Stuetzle. Assessment and pruning of hierarchical model based clustering. Proceedings of the 9th ACM SIGKDD international conference on Knowledge Discovery and Data Mining, 197--205. ACM Press, New York, NY, USA, 2003. Friedrich Leisch. Exploring the structure of mixture model components. In Jaromir Antoch, editor, Compstat 2004--Proceedings in Computational Statistics, 1405--1412. Physika Verlag, Heidelberg, Germany, 2004. ISBN 3-7908-1554-3. } \author{Friedrich Leisch and Bettina Gruen} \keyword{methods} \keyword{hplot} flexmix/man/FLXdist.Rd0000644000176200001440000000255613425024236014316 0ustar liggesusers\name{FLXdist} \alias{FLXdist} \alias{simulate,FLXdist-method} \alias{show,FLXdist-method} \title{Finite Mixtures of Distributions} \description{ Constructs objects of class \code{FLXdist} which represent unfitted finite mixture models. } \usage{ FLXdist(formula, k = NULL, model = FLXMRglm(), components, concomitant = FLXPconstant()) } \arguments{ \item{formula}{A symbolic description of the model to be fit. The general form is \code{~x|g} where \code{x} is the set of predictors and \code{g} an optional grouping factor for repeated measurements.} \item{k}{Integer specifying the number of cluster or a numeric vector of length equal to the length of components, specifying the prior probabilities of clusters.} \item{model}{Object of class \code{FLXM} or a list of \code{FLXM} objects. Default is the object returned by calling \code{FLXMRglm()}.} \item{components}{A list of length equal to the number of components containing a list of length equal to the number of models which again contains a list of named elements for defining the parameters of the component-specific model.} \item{concomitant}{Object of class \code{FLXconcomitant} specifying the model for concomitant variables.} } \value{ Returns an object of class \code{FLXdist}. } \author{Bettina Gruen} \seealso{\code{FLXdist-class}} \keyword{utilities} flexmix/man/FLXbclust.Rd0000644000176200001440000000233313425024236014640 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: FLXbclust.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{FLXMCmvbinary} \alias{FLXMCmvbinary} \alias{FLXbclust} \title{FlexMix Binary Clustering Driver} \description{ This is a model driver for \code{\link{flexmix}} implementing model-based clustering of binary data. } \usage{ FLXMCmvbinary(formula = . ~ ., truncated = FALSE) } \arguments{ \item{formula}{A formula which is interpreted relative to the formula specified in the call to \code{\link{flexmix}} using \code{\link{update.formula}}. Only the left-hand side (response) of the formula is used. Default is to use the original \code{\link{flexmix}} model formula.} \item{truncated}{logical, if \code{TRUE} the observations for the pattern with only zeros are missing and the truncated likelihood is optimized using an EM-algorithm.} } \details{ This model driver can be used to cluster binary data. The only parameter is the column-wise mean of the data, which equals the probability of observing a 1. } \value{ \code{FLXMCmvbinary} returns an object of class \code{FLXMC}. } \author{Friedrich Leisch and Bettina Gruen} \seealso{\code{\link{flexmix}}} \keyword{cluster} flexmix/man/EIC.Rd0000644000176200001440000000241413425024236013372 0ustar liggesusers% % Copyright (C) 2004-2009 Friedrich Leisch and Bettina Gruen % $Id: EIC.Rd 3912 2008-03-13 15:10:24Z gruen $ % \name{EIC} \alias{EIC} \alias{EIC,flexmix-method} \alias{EIC,stepFlexmix-method} \title{Entropic Measure Information Criterion} \description{ Compute the Entropic measure Information criterion for model selection. } \usage{ \S4method{EIC}{flexmix}(object, \dots) \S4method{EIC}{stepFlexmix}(object, \dots) } \arguments{ \item{object}{see Methods section below} \item{\dots}{Some methods for this generic function may take additional, optional arguments. At present none do.} } \section{Methods}{ \describe{ \item{object = "flexmix":}{Compute the EIC of a \code{flexmix} object.} \item{object = "stepFlexmix":}{Compute the EIC of all models contained in the \code{stepFlexmix} object.} }} \value{ Returns a numeric vector with the corresponding EIC value(s). } \keyword{methods} \author{Bettina Gruen} \references{ V. Ramaswamy, W. S. DeSarbo, D. J. Reibstein, and W. T. Robinson. An empirical pooling approach for estimating marketing mix elasticities with PIMS data. \emph{Marketing Science}, \bold{12}(1), 103--124, 1993. } \examples{ data("NPreg", package = "flexmix") ex1 <- flexmix(yn ~ x + I(x^2), data = NPreg, k = 2) EIC(ex1) } flexmix/man/group.Rd0000644000176200001440000000073313425024236014130 0ustar liggesusers\name{group} \docType{methods} \alias{group} \alias{group-methods} \alias{group,flexmix-method} \alias{group,FLXM-method} \alias{group,FLXMRglmfix-method} \title{Extract Grouping Variable} \description{ Extract grouping variable for all observations. } \usage{ \S4method{group}{flexmix}(object) \S4method{group}{FLXM}(object) \S4method{group}{FLXMRglmfix}(object) } \arguments{ \item{object}{an object of class \code{flexmix}.} } \keyword{methods} \author{Bettina Gruen} flexmix/man/seizure.Rd0000644000176200001440000000414513425024236014463 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: seizure.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{seizure} \alias{seizure} \docType{data} \title{Epileptic Seizure Data} \description{ Data from a clinical trial where the effect of intravenous gamma-globulin on suppression of epileptic seizures is studied. Daily observations for a period of 140 days on one patient are given, where the first 27 days are a baseline period without treatment, the remaining 113 days are the treatment period. } \usage{data("seizure")} \format{ A data frame with 140 observations on the following 4 variables. \describe{ \item{Seizures}{A numeric vector, daily counts of epileptic seizures.} \item{Hours}{A numeric vector, hours of daily parental observation.} \item{Treatment}{A factor with levels \code{No} and \code{Yes}.} \item{Day}{A numeric vector.} } } \source{ P. Wang, M. Puterman, I. Cockburn, and N. Le. Mixed poisson regression models with covariate dependent rates. \emph{Biometrics}, \bold{52}, 381--400, 1996. } \references{ B. Gruen and F. Leisch. Bootstrapping finite mixture models. In J. Antoch, editor, Compstat 2004--Proceedings in Computational Statistics, 1115--1122. Physika Verlag, Heidelberg, Germany, 2004. ISBN 3-7908-1554-3. } \examples{ data("seizure", package = "flexmix") plot(Seizures/Hours ~ Day, col = as.integer(Treatment), pch = as.integer(Treatment), data = seizure) abline(v = 27.5, lty = 2, col = "grey") legend(140, 9, c("Baseline", "Treatment"), pch = 1:2, col = 1:2, xjust = 1, yjust = 1) set.seed(123) ## The model presented in the Wang et al paper: two components for ## "good" and "bad" days, respectively, each a Poisson GLM with hours of ## parental observation as offset seizMix <- flexmix(Seizures ~ Treatment * log(Day), data = seizure, k = 2, model = FLXMRglm(family = "poisson", offset = log(seizure$Hours))) summary(seizMix) summary(refit(seizMix)) matplot(seizure$Day, fitted(seizMix)/seizure$Hours, type = "l", add = TRUE, col = 3:4) } \keyword{datasets} flexmix/man/FLXMRlmer.Rd0000644000176200001440000000757513431371744014565 0ustar liggesusers\name{FLXMRlmer} \alias{FLXMRlmer} \alias{FLXMRlmer-class} \alias{FLXMRlmm-class} \alias{FLXMRlmmfix-class} \alias{FLXdeterminePostunscaled,FLXMRlmer-method} \alias{FLXdeterminePostunscaled,FLXMRlmm-method} \alias{FLXmstep,FLXMRlmer-method} \alias{FLXmstep,FLXMRlmm-method} \alias{FLXgetModelmatrix,FLXMRlmer-method} \alias{FLXgetModelmatrix,FLXMRlmm-method} \alias{FLXMRlmm} \alias{FLXgetObs,FLXMRlmm-method} \alias{FLXmstep,FLXMRlmmfix-method} \alias{predict,FLXMRlmm-method} \alias{rFLXM,FLXMRlmm,FLXcomponent-method} \alias{rFLXM,FLXMRlmm,list-method} \alias{rFLXM,FLXMRlmc,FLXcomponent-method} \alias{rFLXM,FLXMRlmer,FLXcomponent-method} \title{FlexMix Interface to Linear Mixed Models} \description{ This is a driver which allows fitting of mixtures of linear models with random effects. } \usage{ FLXMRlmm(formula = . ~ ., random, lm.fit = c("lm.wfit", "smooth.spline"), varFix = c(Random = FALSE, Residual = FALSE), \dots) FLXMRlmer(formula = . ~ ., random, weighted = TRUE, control = list(), eps = .Machine$double.eps) } \arguments{ \item{formula}{A formula which is interpreted relative to the formula specified in the call to \code{flexmix} using \code{\link{update.formula}}. Default is to use the original \code{flexmix} model formula.} \item{random}{A formula for specifying the random effects.} \item{weighted}{A logical indicating if the model should be estimated with weighted ML.} \item{control}{A list of control parameters. See \code{\link[lme4]{lmer}} for details.} \item{eps}{Observations with a component-specific posterior smaller than \code{eps} are omitted in the M-step for this component.} \item{lm.fit}{A character string indicating if the coefficients should be fitted using either a linear model or the function \code{smooth.spline}} \item{varFix}{Named logical vector of length 2 indicating if the variance of the random effects and the residuals are fixed over the components.} \item{\dots}{Additional arguments to be passed to \code{smooth.spline}.} } \details{ \code{FLXMRlmm} allows only one random effect. \code{FLXMRlmer} allows an arbitrary number of random effects if \code{weighted = FALSE}; a certain structure of the model matrix of the random effects has to be given for weighted ML estimation, i.e. where \code{weighted = TRUE}. } \value{ Returns an object of class \code{FLXMRlmer} and \code{FLXMRlmm}. } \section{Warning}{ For \code{FLXMRlmer} the weighted ML estimation is only correct if the covariate matrix of the random effects is the same for each observation. By default weighted ML estimation is made and the condition on the covariate matrix of the random effects is checked. If this fails, only estimation with \code{weighted = FALSE} is possible which will maximize the classification likelihood. } \author{Bettina Gruen} \examples{ id <- rep(1:50, each = 10) x <- rep(1:10, 50) sample <- data.frame(y = rep(rnorm(unique(id)/2, 0, c(5, 2)), each = 10) + rnorm(length(id), rep(c(3, 8), each = 10)) + rep(c(0, 3), each = 10) * x, x = x, id = factor(id)) fitted <- flexmix(.~.|id, k = 2, model = FLXMRlmm(y ~ x, random = ~ 1), data = sample, control = list(tolerance = 10^-3), cluster = rep(rep(1:2, each = 10), 25)) parameters(fitted) fitted1 <- flexmix(.~.|id, k = 2, model = FLXMRlmer(y ~ x, random = ~ 1), data = sample, control = list(tolerance = 10^-3), cluster = rep(rep(1:2, each = 10), 25)) parameters(fitted1) fitted2 <- flexmix(.~.|id, k = 2, model = FLXMRlmm(y ~ 0 + x, random = ~ 1, lm.fit = "smooth.spline"), data = sample, control = list(tolerance = 10^-3), cluster = rep(rep(1:2, each = 10), 25)) parameters(fitted2) } \keyword{models} flexmix/man/plotEll.Rd0000644000176200001440000000253513425024236014411 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: plotEll.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{plotEll} \alias{plotEll} \title{Plot Confidence Ellipses for FLXMCmvnorm Results} \description{ Plot 50\% and 95\% confidence ellipses for mixtures of Gaussians fitted using \code{\link{FLXMCmvnorm}}. } \usage{ plotEll(object, data, which = 1:2, model = 1, project = NULL, points = TRUE, eqscale = TRUE, col = NULL, number = TRUE, cex = 1.5, numcol = "black", pch = NULL, ...) } \arguments{ \item{object}{An object of class \code{flexmix} with a fitted \code{FLXMCmvnorm} model.} \item{data}{The response variable in a data frame or as a matrix.} \item{which}{Index numbers of dimensions of (projected) input space to plot.} \item{model}{The model (for a multivariate response) that shall be plotted.} \item{project}{Projection object, currently only the result of \code{\link[stats]{prcomp}} is supported.} \item{points}{Logical, shall data points be plotted?} \item{eqscale}{Logical, plot using \code{\link[MASS]{eqscplot}}?} \item{number}{Logical, plot number labels at cluster centers?} \item{cex, numcol}{Size and color of number labels.} \item{pch, col, \dots}{Graphical parameters.} } \author{Friedrich Leisch and Bettina Gruen} \seealso{\code{\link{FLXMCmvnorm}}} \keyword{cluster} flexmix/man/FLXglmFix.Rd0000644000176200001440000000423613425024236014576 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: FLXglmFix.Rd 5015 2015-01-17 11:12:29Z gruen $ % \name{FLXMRglmfix} \alias{FLXMRglmfix} \alias{FLXglmFix} \title{FlexMix Interface to GLMs with Fixed Coefficients} \description{ This implements a driver for FlexMix which interfaces the \code{glm} family of models and where it is possible to specify fixed (constant) or nested varying coefficients or to ensure that in the Gaussian case the variance estimate is equal for all components. } \usage{ FLXMRglmfix(formula = . ~ ., fixed = ~0, varFix = FALSE, nested = NULL, family = c("gaussian", "binomial", "poisson", "Gamma"), offset = NULL) } \arguments{ \item{formula}{A formula which is interpreted relative to the formula specified in the call to \code{flexmix} using \code{update.formula}. Default is to use the original \code{flexmix} model formula.} \item{fixed}{A formula which specifies the additional regressors for the fixed (constant) coefficients.} \item{varFix}{A logical indicating if the variance estimate for Gaussian components should be constrained to be equal for all components. It can be also a vector specifying the number of components with equal variance.} \item{nested}{An object of class \code{FLXnested} or a list specifying the nested structure.} \item{family}{A character string naming a \code{glm} family function.} \item{offset}{This can be used to specify an \emph{a priori} known component to be included in the linear predictor during fitting.} } \value{ Returns an object of class \code{FLXMRglmfix}. } \author{Friedrich Leisch and Bettina Gruen} \seealso{\code{FLXMRglm}} \examples{ data("NPreg", package = "flexmix") ex <- flexmix(yn ~ x | id2, data = NPreg, k = 2, cluster = NPreg$class, model = FLXMRglm(yn ~ . + I(x^2))) ex.fix <- flexmix(yn ~ x | id2, data = NPreg, cluster = posterior(ex), model = FLXMRglmfix(nested = list(k = c(1, 1), formula = c(~0, ~I(x^2))))) summary(refit(ex)) \dontrun{ summary(refit(ex.fix)) } } \keyword{regression} \keyword{models} flexmix/man/dmft.Rd0000644000176200001440000000370213425024236013725 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: dmft.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{dmft} \alias{dmft} \docType{data} \title{Dental Data} \description{ Count data from a dental epidemiological study for evaluation of various programs for reducing caries collected among school children from an urban area of Belo Horizonte (Brazil). } \usage{data("dmft")} \format{ A data frame with 797 observations on the following 5 variables. \describe{ \item{End}{Number of decayed, missing or filled teeth at the end of the study.} \item{Begin}{Number of decayed, missing or filled teeth at the beginning of the study.} \item{Gender}{A factor with levels \code{male} and \code{female}.} \item{Ethnic}{A factor with levels \code{brown}, \code{white} and \code{black}.} \item{Treatment}{A factor with levels \code{control}, \code{educ}, \code{enrich}, \code{rinse}, \code{hygiene} and \code{all}.} } } \details{ The aim of the caries prevention study was to compare four methods to prevent dental caries. Interventions were carried out according to the following scheme: \describe{ \item{control}{Control group} \item{educ}{Oral health education} \item{enrich}{Enrichment of the school diet with rice bran} \item{rinse}{Mouthwash with 0.2\% sodium floride (NaF) solution} \item{hygiene}{Oral hygiene} \item{all}{All four methods together} } } \source{ D. Boehning, E. Dietz, P. Schlattmann, L. Mendonca and U. Kirchner. The zero-inflated Poisson model and the decayed, missing and filled teeth index in dental epidemiology. \emph{Journal of the Royal Statistical Society A}, \bold{162}(2), 195--209, 1999. } \examples{ data("dmft", package = "flexmix") dmft_flx <- initFlexmix(End ~ 1, data = dmft, k = 2, model = FLXMRglmfix(family = "poisson", fixed = ~ Gender + Ethnic + Treatment)) } \keyword{datasets} flexmix/man/FLXnested-class.Rd0000644000176200001440000000200213425024236015722 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: FLXnested-class.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{FLXnested-class} \docType{class} \alias{FLXnested-class} \alias{coerce,list,FLXnested-method} \alias{coerce,NULL,FLXnested-method} \alias{coerce,numeric,FLXnested-method} \alias{initialize,FLXnested-method} \title{Class "FLXnested"} \description{Specification of nesting structure for regression coefficients.} \section{Objects from the Class}{ Objects can be created by calls of the form \code{new("FLXnested", formula, k, ...)}. In addition, named lists can be coerced to \code{FLXnested} objects, names are completed if unique. } \section{Slots}{ \describe{ \item{\code{formula}:}{Object of class \code{"list"} containing the formula for determining the model matrix for each nested parameter.} \item{\code{k}:}{Object of class \code{"numeric"} specifying the number of components in each group.} } } \author{Friedrich Leisch and Bettina Gruen} \keyword{classes} flexmix/man/FLXglm.Rd0000644000176200001440000000255413425024236014130 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: FLXglm.Rd 5115 2017-04-07 08:18:13Z gruen $ % \name{FLXMRglm} \alias{FLXMRglm} \alias{FLXglm} \title{FlexMix Interface to Generalized Linear Models} \description{ This is the main driver for FlexMix interfacing the \code{\link{glm}} family of models. } \usage{ FLXMRglm(formula = . ~ ., family = c("gaussian", "binomial", "poisson", "Gamma"), offset = NULL) } \arguments{ \item{formula}{A formula which is interpreted relative to the formula specified in the call to \code{\link{flexmix}} using \code{\link{update.formula}}. Default is to use the original \code{\link{flexmix}} model formula.} \item{family}{A character string naming a \code{\link{glm}} family function.} \item{offset}{This can be used to specify an \emph{a priori} known component to be included in the linear predictor during fitting.} } \details{ See \code{\link{flexmix}} for examples. } \value{ Returns an object of class \code{FLXMRglm}. } \author{Friedrich Leisch and Bettina Gruen} \references{ Friedrich Leisch. FlexMix: A general framework for finite mixture models and latent class regression in R. \emph{Journal of Statistical Software}, \bold{11}(8), 2004. doi:10.18637/jss.v011.i08 } \seealso{\code{\link{flexmix}}, \code{\link{glm}}} \keyword{regression} \keyword{models} flexmix/man/FLXfit.Rd0000644000176200001440000000227013425024236014126 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: FLXfit.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{FLXfit} \alias{FLXfit} \alias{FLXfit,list-method} \title{Fitter Function for FlexMix Models} \description{ This is the basic computing engine called by \code{\link{flexmix}}, it should usually not be used directly. } \usage{ FLXfit(model, concomitant, control, postunscaled = NULL, groups, weights) } \arguments{ \item{model}{List of \code{FLXM} objects.} \item{concomitant}{Object of class \code{FLXP}.} \item{control}{Object of class \code{FLXcontrol}.} \item{weights}{A numeric vector of weights to be used in the fitting process.} \item{postunscaled}{Initial a-posteriori probabilities of the observations at the start of the EM algorithm.} \item{groups}{List with components \code{group} which is a factor with optional grouping of observations and \code{groupfirst} which is a logical vector for the first observation of each group.} } \value{ Returns an object of class \code{flexmix}. } \author{Friedrich Leisch and Bettina Gruen} \seealso{\code{\link{flexmix}}, \code{\link{flexmix-class}}} \keyword{regression} \keyword{cluster} flexmix/man/ICL.Rd0000644000176200001440000000245413425024236013405 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: ICL.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{ICL} \alias{ICL,flexmix-method} \alias{ICL,stepFlexmix-method} \title{Integrated Completed Likelihood Criterion} \description{ Compute the Integrated Completed Likelihood criterion for model selection. } \usage{ \S4method{ICL}{flexmix}(object, \dots) \S4method{ICL}{stepFlexmix}(object, \dots) } \arguments{ \item{object}{see Methods section below} \item{\dots}{Some methods for this generic function may take additional, optional arguments. At present none do.} } \section{Methods}{ \describe{ \item{object = "flexmix":}{Compute the ICL of a \code{flexmix} object.} \item{object = "stepFlexmix":}{Compute the ICL of all models contained in the \code{stepFlexmix} object.} }} \value{ Returns a numeric vector with the corresponding ICL value(s). } \keyword{methods} \author{Friedrich Leisch and Bettina Gruen} \references{ C. Biernacki, G. Celeux, and G. Govaert. Assessing a mixture model for clustering with the integrated completed likelihood. \emph{IEEE Transactions on Pattern Analysis and Machine Intelligence}, \emph{22}(7), 719--725, 2000. } \examples{ data("NPreg", package = "flexmix") ex1 <- flexmix(yn ~ x + I(x^2), data = NPreg, k = 2) ICL(ex1) } flexmix/man/Mehta.Rd0000644000176200001440000000265513425024236014037 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: Mehta.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{Mehta} \alias{Mehta} \docType{data} \title{Mehta Trial} \description{ For a 22-centre trial the number of responses and the total number of patients is reported for the control group and the group receiving a new drug. } \usage{data("Mehta")} \format{ A data frame with 44 observations on the following 4 variables. \describe{ \item{Response}{Number of responses.} \item{Total}{Total number of observations.} \item{Drug}{A factor indicating treatment with levels \code{New} and \code{Control}.} \item{Site}{A factor indicating the site/centre.} } } \source{ M. Aitkin. Meta-analysis by random effect modelling in generalized linear models. \emph{Statistics in Medicine}, \bold{18}, 2343--2351, 1999. } \references{ C.R. Mehta, N.R. Patel and P. Senchaudhuri. Importance sampling for estimating exact probabilities in permutational inference. \emph{Journal of the American Statistical Association}, \emph{83}, 999--1005, 1988. } \examples{ data("Mehta", package = "flexmix") mehtaMix <- initFlexmix(cbind(Response, Total-Response) ~ 1|Site, data = Mehta, nrep = 5, k = 3, model = FLXMRglmfix(family = "binomial", fixed = ~ Drug), control = list(minprior = 0.04)) } \keyword{datasets} flexmix/man/candy.Rd0000644000176200001440000000250313425024236014067 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: candy.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{candy} \alias{candy} \docType{data} \title{Candy Packs Purchased} \description{ The data is from a new product and concept test where the number of individual packs of hard candy purchased within the past 7 days is recorded. } \usage{data("candy")} \format{ A data frame with 21 observations on the following 2 variables. \describe{ \item{\code{Packages}}{a numeric vector} \item{\code{Freq}}{a numeric vector} } } \source{ D. Boehning, E. Dietz and P. Schlattmann. Recent Developments in Computer-Assisted Analysis of Mixtures. Biometrics 54(2), 525--536, 1998. } \references{ J. Magidson and J. K. Vermunt. Latent Class Models. In D. W. Kaplan (ed.), The Sage Handbook of Quantitative Methodology for the Social Sciences, 175--198, 2004. Thousand Oakes: Sage Publications. D. Boehning, E. Dietz and P. Schlattmann. Recent Developments in Computer-Assisted Analysis of Mixtures. \emph{Biometrics}, \bold{54}(2), 525--536, 1998. W. R. Dillon and A. Kumar. Latent structure and other mixture models in marketing: An integrative survey and overview. In R. P. Bagozzi (ed.), Advanced methods of marketing research, 352--388, 1994. Cambridge, UK: Blackwell. } \keyword{datasets} flexmix/man/ExNPreg.Rd0000644000176200001440000000165113425024236014304 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: ExNPreg.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{ExNPreg} \alias{ExNPreg} \alias{NPreg} \title{Artificial Example for Normal, Poisson and Binomial Regression} \description{ A simple artificial regression example with 2 latent classes, one independent variable (uniform on \eqn{[0,10]}), and three dependent variables with Gaussian, Poisson and Binomial distribution, respectively. } \usage{ ExNPreg(n) data("NPreg") } \arguments{ \item{n}{Number of observations per latent class.} } \details{ The \code{NPreg} data frame can be re-created by \code{ExNPreg(100)} using \code{set.seed(2602)}, it has been saved as a data set for simplicity of examples only. } \examples{ data("NPreg", package = "flexmix") plot(yn ~ x, data = NPreg, col = class) plot(yp ~ x, data = NPreg, col = class) plot(yb ~ x, data = NPreg, col = class) } \keyword{datasets} flexmix/man/fabricfault.Rd0000644000176200001440000000227313425024236015257 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: fabricfault.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{fabricfault} \alias{fabricfault} \docType{data} \title{Fabric Faults} \description{ Number of faults in rolls of a textile fabric. } \usage{data("fabricfault")} \format{ A data frame with 32 observations on the following 2 variables. \describe{ \item{Length}{Length of role (m).} \item{Faults}{Number of faults.} } } \source{ G. McLachlan and D. Peel. \emph{Finite Mixture Models}, 2000, John Wiley and Sons Inc. \url{http://www.maths.uq.edu.au/~gjm/DATA/mmdata.html} } \references{ A. F. Bissell. A Negative Binomial Model with Varying Element Sizes \emph{Biometrika}, \bold{59}, 435--441, 1972. M. Aitkin. A general maximum likelihood analysis of overdispersion in generalized linear models. \emph{Statistics and Computing}, \bold{6}, 251--262, 1996. } \examples{ data("fabricfault", package = "flexmix") fabricMix <- initFlexmix(Faults ~ 1, data = fabricfault, k = 2, model = FLXMRglmfix(family = "poisson", fixed = ~ log(Length)), nrep = 5) } \keyword{datasets} flexmix/man/FLXMRmultinom.Rd0000644000176200001440000000216613425024236015453 0ustar liggesusers\name{FLXMRmultinom} \alias{FLXMRmultinom} \title{FlexMix Interface to Multiomial Logit Models} \description{ Model driver for fitting mixtures of multinomial logit models. } \usage{ FLXMRmultinom(formula = . ~ ., ...) } \arguments{ \item{formula}{A formula which is interpreted relative to the formula specified in the call to \code{\link{flexmix}} using \code{\link{update.formula}}. Default is to use the original \code{\link{flexmix}} model formula.} \item{\dots}{Additional arguments to be passed to \code{nnet.default}.} } \details{ The M-step is performed using \code{nnet.default}. } \value{ Returns an object of class \code{FLXMRmultinom}. } \references{ Bettina Gruen and Friedrich Leisch. Identifiability of finite mixtures of multinomial logit models with varying and fixed effects. \emph{Journal of Classification}, \bold{25}, 225--247. 2008. } \author{ Bettina Gruen } \section{Warning}{ To ensure identifiability repeated measurements are necessary. Sufficient conditions are given in Gruen and Leisch (2008). } \seealso{\code{\link{FLXMRcondlogit}}} \keyword{regression} \keyword{models} flexmix/man/FLXMCmvpois.Rd0000644000176200001440000000175713425024236015112 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: FLXMCmvpois.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{FLXMCmvpois} \alias{FLXMCmvpois} \title{FlexMix Poisson Clustering Driver} \description{ This is a model driver for \code{\link{flexmix}} implementing model-based clustering of Poisson distributed data. } \usage{ FLXMCmvpois(formula = . ~ .) } \arguments{ \item{formula}{A formula which is interpreted relative to the formula specified in the call to \code{\link{flexmix}} using \code{\link{update.formula}}. Only the left-hand side (response) of the formula is used. Default is to use the original \code{\link{flexmix}} model formula.} } \details{ This can be used to cluster Poisson distributed data where given the component membership the variables are mutually independent. } \value{ \code{FLXMCmvpois} returns an object of class \code{FLXMC}. } \author{Friedrich Leisch and Bettina Gruen} \seealso{\code{\link{flexmix}}} \keyword{cluster} flexmix/man/FLXdist-class.Rd0000644000176200001440000000464313425024236015420 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: FLXdist-class.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{FLXdist-class} \docType{class} \alias{FLXdist-class} \alias{predict,FLXdist-method} \alias{predict,FLXM-method} \alias{predict,FLXMRglm-method} \alias{predict,FLXMRmgcv-method} \alias{parameters,FLXdist-method} \alias{prior} \alias{prior,FLXdist-method} \title{Class "FLXdist"} \description{ Objects of class \code{FLXdist} represent unfitted finite mixture models. } \usage{ \S4method{parameters}{FLXdist}(object, component = NULL, model = NULL, which = c("model", "concomitant"), simplify = TRUE, drop = TRUE) \S4method{predict}{FLXdist}(object, newdata = list(), aggregate = FALSE, ...) } \arguments{ \item{object}{An object of class "FLXdist".} \item{component}{Number of component(s), if \code{NULL} all components are returned.} \item{model}{Number of model(s), if \code{NULL} all models are returned.} \item{which}{Specifies if the parameters of the component specific model or the concomitant variable model are returned.} \item{simplify}{Logical, if \code{TRUE} the returned values are simplified to a vector or matrix if possible.} \item{drop}{Logical, if \code{TRUE} the function tries to simplify the return object by omitting lists of length one.} \item{newdata}{Dataframe containing new data.} \item{aggregate}{Logical, if \code{TRUE} then the predicted values for each model aggregated over the components are returned.} \item{\dots}{Passed to the method of the model class.} } \section{Slots}{ \describe{ \item{model}{List of \code{FLXM} objects.} \item{prior}{Numeric vector with prior probabilities of clusters.} \item{components}{List describing the components using \code{FLXcomponent} objects.} \item{\code{concomitant}:}{Object of class \code{"FLXP"}.} \item{formula}{Object of class \code{"formula"}.} \item{call}{The function call used to create the object.} \item{k}{Number of clusters.} } } \section{Accessor Functions}{ The following functions should be used for accessing the corresponding slots: \describe{ \item{\code{parameters}:}{The parameters for each model and component, return value depends on the model.} \item{\code{prior}:}{Numeric vector of prior class probabilities/component weights} } } \author{Friedrich Leisch and Bettina Gruen} \seealso{\code{FLXdist}} \keyword{classes} flexmix/man/FLXMRlmmc.Rd0000644000176200001440000000274613425024236014543 0ustar liggesusers\name{FLXMRlmmc} \alias{FLXMRlmmc} \alias{FLXMRlmmc-class} \alias{FLXMRlmmcfix-class} \alias{FLXMRlmc-class} \alias{FLXMRlmcfix-class} \alias{predict,FLXMRlmc-method} \title{FlexMix Interface to Linear Mixed Models with Left-Censoring} \description{ This is a driver which allows fitting of mixtures of linear models with random effects and left-censored observations. } \usage{ FLXMRlmmc(formula = . ~ ., random, censored, varFix, eps = 10^-6, ...) } \arguments{ \item{formula}{A formula which is interpreted relative to the formula specified in the call to \code{flexmix} using \code{\link{update.formula}}. Default is to use the original \code{flexmix} model formula.} \item{random}{A formula for specifying the random effects. If missing no random effects are fitted.} \item{varFix}{If random effects are specified a named logical vector of length 2 indicating if the variance of the random effects and the residuals are fixed over the components. Otherwise a logical indicating if the variance of the residuals are fixed over the components.} \item{censored}{A formula for specifying the censoring variable.} \item{eps}{Observations with an a-posteriori probability smaller or equal to \code{eps} are omitted in the M-step.} \item{\dots}{Additional arguments to be passed to \code{lm.wfit}.} } \value{ Returns an object of class \code{FLXMRlmmc}, \code{FLXMRlmmcfix}, \code{FLXMRlmc} or \code{FLXMRlmcfix}. } \author{Bettina Gruen} \keyword{models} flexmix/man/flexmix.Rd0000644000176200001440000001266613425024236014460 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: flexmix.Rd 5115 2017-04-07 08:18:13Z gruen $ % \name{flexmix} \alias{flexmix} \alias{flexmix,formula,ANY,ANY,ANY,missing-method} \alias{flexmix,formula,ANY,ANY,ANY,list-method} \alias{flexmix,formula,ANY,ANY,ANY,FLXM-method} \alias{prior,flexmix-method} \alias{show,flexmix-method} \alias{summary,flexmix-method} \alias{show,summary.flexmix-method} \title{Flexible Mixture Modeling} \description{ FlexMix implements a general framework for finite mixtures of regression models. Parameter estimation is performed using the EM algorithm: the E-step is implemented by \code{flexmix}, while the user can specify the M-step. } \usage{ flexmix(formula, data = list(), k = NULL, cluster = NULL, model = NULL, concomitant = NULL, control = NULL, weights = NULL) \S4method{summary}{flexmix}(object, eps = 1e-4, ...) } \arguments{ \item{formula}{A symbolic description of the model to be fit. The general form is \code{y~x|g} where \code{y} is the response, \code{x} the set of predictors and \code{g} an optional grouping factor for repeated measurements.} \item{data}{An optional data frame containing the variables in the model.} \item{k}{Number of clusters (not needed if \code{cluster} is specified).} \item{cluster}{Either a matrix with \code{k} columns of initial cluster membership probabilities for each observation; or a factor or integer vector with the initial cluster assignments of observations at the start of the EM algorithm. Default is random assignment into \code{k} clusters.} \item{weights}{An optional vector of replication weights to be used in the fitting process. Should be \code{NULL}, an integer vector or a formula.} \item{model}{Object of class \code{FLXM} or list of \code{FLXM} objects. Default is the object returned by calling \code{\link{FLXMRglm}()}.} \item{concomitant}{Object of class \code{FLXP}. Default is the object returned by calling \code{\link{FLXPconstant}}.} \item{control}{Object of class \code{FLXcontrol} or a named list.} \item{object}{Object of class \code{flexmix}.} \item{eps}{Probabilities below this threshold are treated as zero in the summary method.} \item{\dots}{Currently not used.} } \details{ FlexMix models are described by objects of class \code{FLXM}, which in turn are created by driver functions like \code{\link{FLXMRglm}} or \code{\link{FLXMCmvnorm}}. Multivariate responses with independent components can be specified using a list of \code{FLXM} objects. The \code{summary} method lists for each component the prior probability, the number of observations assigned to the corresponding cluster, the number of observations with a posterior probability larger than \code{eps} and the ratio of the latter two numbers (which indicates how separated the cluster is from the others). } \value{ Returns an object of class \code{flexmix}. } \author{Friedrich Leisch and Bettina Gruen} \seealso{\code{\link[flexmix]{plot-methods}}} \references{ Friedrich Leisch. FlexMix: A general framework for finite mixture models and latent class regression in R. \emph{Journal of Statistical Software}, \bold{11}(8), 2004. doi:10.18637/jss.v011.i08 Bettina Gruen and Friedrich Leisch. Fitting finite mixtures of generalized linear regressions in R. Computational Statistics & Data Analysis, 51(11), 5247-5252, 2007. doi:10.1016/j.csda.2006.08.014 Bettina Gruen and Friedrich Leisch. FlexMix Version 2: Finite mixtures with concomitant variables and varying and constant parameters Journal of Statistical Software, 28(4), 1-35, 2008. doi:10.18637/jss.v028.i04 } \keyword{regression} \keyword{cluster} \examples{ data("NPreg", package = "flexmix") ## mixture of two linear regression models. Note that control parameters ## can be specified as named list and abbreviated if unique. ex1 <- flexmix(yn ~ x + I(x^2), data = NPreg, k = 2, control = list(verb = 5, iter = 100)) ex1 summary(ex1) plot(ex1) ## now we fit a model with one Gaussian response and one Poisson ## response. Note that the formulas inside the call to FLXMRglm are ## relative to the overall model formula. ex2 <- flexmix(yn ~ x, data = NPreg, k = 2, model = list(FLXMRglm(yn ~ . + I(x^2)), FLXMRglm(yp ~ ., family = "poisson"))) plot(ex2) ex2 table(ex2@cluster, NPreg$class) ## for Gaussian responses we get coefficients and standard deviation parameters(ex2, component = 1, model = 1) ## for Poisson response we get only coefficients parameters(ex2, component = 1, model = 2) ## fitting a model only to the Poisson response is of course ## done like this ex3 <- flexmix(yp ~ x, data = NPreg, k = 2, model = FLXMRglm(family = "poisson")) ## if observations are grouped, i.e., we have several observations per ## individual, fitting is usually much faster: ex4 <- flexmix(yp~x|id1, data = NPreg, k = 2, model = FLXMRglm(family = "poisson")) ## And now a binomial example. Mixtures of binomials are not generically ## identified, here the grouping variable is necessary: set.seed(1234) ex5 <- initFlexmix(cbind(yb,1 - yb) ~ x, data = NPreg, k = 2, model = FLXMRglm(family = "binomial"), nrep = 5) table(NPreg$class, clusters(ex5)) ex6 <- initFlexmix(cbind(yb, 1 - yb) ~ x | id2, data = NPreg, k = 2, model = FLXMRglm(family = "binomial"), nrep = 5) table(NPreg$class, clusters(ex6)) } flexmix/man/flexmix-class.Rd0000644000176200001440000000370213425024236015552 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: flexmix-class.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{flexmix-class} \docType{class} \alias{flexmix-class} \title{Class "flexmix"} \description{A fitted \code{\link{flexmix}} model.} \section{Slots}{ \describe{ \item{\code{model}:}{List of \code{FLXM} objects.} \item{\code{prior}:}{Numeric vector with prior probabilities of clusters.} \item{\code{posterior}:}{Named list with elements \code{scaled} and \code{unscaled}, both matrices with one row per observation and one column per cluster.} \item{\code{iter}:}{Number of EM iterations.} \item{\code{k}:}{Number of clusters after EM.} \item{\code{k0}:}{Number of clusters at start of EM.} \item{\code{cluster}:}{Cluster assignments of observations.} \item{\code{size}:}{Cluster sizes.} \item{\code{logLik}:}{Log-likelihood at EM convergence.} \item{\code{df}:}{Total number of parameters of the model.} \item{\code{components}:}{List describing the fitted components using \code{FLXcomponent} objects.} \item{\code{formula}:}{Object of class \code{"formula"}.} \item{\code{control}:}{Object of class \code{"FLXcontrol"}.} \item{\code{call}:}{The function call used to create the object.} \item{\code{group}:}{Object of class \code{"factor"}.} \item{\code{converged}:}{Logical, \code{TRUE} if EM algorithm converged.} \item{\code{concomitant}:}{Object of class \code{"FLXP"}..} \item{\code{weights}:}{Optional weights of the observations.} } } \section{Extends}{ Class \code{FLXdist}, directly. } \section{Accessor Functions}{ The following functions should be used for accessing the corresponding slots: \describe{ \item{\code{cluster}:}{Cluster assignments of observations.} \item{\code{posterior}:}{A matrix of posterior probabilities for each observation.} } } \author{Friedrich Leisch and Bettina Gruen} \keyword{classes} flexmix/man/Lapply-methods.Rd0000644000176200001440000000313013425024236015670 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: Lapply-methods.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{Lapply-methods} \docType{methods} \title{Methods for Function Lapply} \alias{Lapply,FLXRmstep-method} \description{Apply a function to each component of a finite mixture} \usage{ \S4method{Lapply}{FLXRmstep}(object, FUN, model = 1, component = TRUE, ...) } \arguments{ \item{object}{S4 class object.} \item{FUN}{The function to be applied.} \item{model}{The model (for a multivariate response) that shall be used.} \item{component}{Index vector for selecting the components.} \item{\dots}{Optional arguments to \code{FUN}.} } \section{Methods}{ \describe{ \item{object = FLXRmstep:}{Apply a function to each component of a refitted \code{flexmix} object using method = \code{"mstep"}.} } } \details{ \code{FUN} is found by a call to \code{match.fun} and typically is specified as a function or a symbol (e.g. a backquoted name) or a character string specifying a function to be searched for from the environment of the call to \code{Lapply}. } \value{ A list of the length equal to the number of components specified is returned, each element of which is the result of applying \code{FUN} to the specified component of the refitted mixture model. } \keyword{methods} \author{Friedrich Leisch and Bettina Gruen} \examples{ data("NPreg", package = "flexmix") ex2 <- flexmix(yn ~ x, data = NPreg, k = 2, model = list(FLXMRglm(yn ~ . + I(x^2)), FLXMRglm(yp ~ ., family = "poisson"))) ex2r <- refit(ex2, method = "mstep") Lapply(ex2r, "vcov", 2) } flexmix/man/FLXMRrobglm.Rd0000644000176200001440000000525013425024236015066 0ustar liggesusers% % Copyright (C) 2008 Friedrich Leisch and Bettina Gruen % $Id: FLXMRrobglm.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{FLXMRrobglm} \alias{FLXMRrobglm} \alias{FLXMRrobglm-class} \title{FlexMix Driver for Robust Estimation of Generalized Linear Models} \description{ This driver adds a noise component to the mixture model which can be used to model background noise in the data. See the Compstat paper Leisch (2008) cited below for details. } \usage{ FLXMRrobglm(formula = . ~ ., family = c("gaussian", "poisson"), bgw = FALSE, ...) } \arguments{ \item{formula}{A formula which is interpreted relative to the formula specified in the call to \code{flexmix} using \code{\link{update.formula}}. Default is to use the original \code{flexmix} model formula.} \item{family}{A character string naming a \code{\link{glm}} family function.} \item{bgw}{Logical, controls whether the parameters of the background component are fixed to multiples of location and scale of the complete data (the default), or estimated by EM with normal weights for the background (\code{bgw = TRUE}).} \item{\dots}{passed to \code{FLXMRglm}} } \value{ Returns an object of class \code{FLXMRrobglm}. } \author{Friedrich Leisch and Bettina Gruen} \note{ The implementation of this model class is currently under development, and some methods like \code{refit} are still missing. } \references{ Friedrich Leisch. Modelling background noise in finite mixtures of generalized linear regression models. In Paula Brito, editor, Compstat 2008--Proceedings in Computational Statistics, 385--396. Physica Verlag, Heidelberg, Germany, 2008.\cr Preprint available at http://epub.ub.uni-muenchen.de/6332/. } \examples{ ## Example from Compstat paper, see paper for detailed explanation: data("NPreg", package = "flexmix") DATA <- NPreg[, 1:2] set.seed(3) DATA2 <- rbind(DATA, cbind(x = -runif(3), yn = 50 + runif(3))) ## Estimation without (f2) and with (f3) background component f2 <- flexmix(yn ~ x + I(x^2), data = DATA2, k = 2) f3 <- flexmix(yn ~ x + I(x^2), data = DATA2, k = 3, model = FLXMRrobglm(), control = list(minprior = 0)) ## Predict on new data for plots x <- seq(-5,15, by = .1) y2 <- predict(f2, newdata = data.frame(x = x)) y3 <- predict(f3, newdata = data.frame(x = x)) ## f2 was estimated without background component: plot(yn ~ x, data = DATA2, pch = clusters(f2), col = clusters(f2)) lines(x, y2$Comp.1, col = 1) lines(x, y2$Comp.2, col = 2) ## f3 is with background component: plot(yn ~ x, data = DATA2, pch = 4 - clusters(f3), col = 4 - clusters(f3)) lines(x, y3$Comp.2, col = 2) lines(x, y3$Comp.3, col = 1) } \keyword{models} flexmix/man/FLXMRmgcv.Rd0000644000176200001440000000376213425024236014546 0ustar liggesusers\name{FLXMRmgcv} \alias{FLXMRmgcv} \alias{FLXMRmgcv-class} \title{FlexMix Interface to GAMs} \description{ This is a driver which allows fitting of mixtures of GAMs. } \usage{ FLXMRmgcv(formula = . ~ ., family = c("gaussian", "binomial", "poisson"), offset = NULL, control = NULL, optimizer = c("outer", "newton"), in.out = NULL, eps = .Machine$double.eps, ...) } \arguments{ \item{formula}{A formula which is interpreted relative to the formula specified in the call to \code{\link{flexmix}} using \code{\link{update.formula}}. Default is to use the original \code{\link{flexmix}} model formula.} \item{family}{A character string naming a \code{\link{glm}} family function.} \item{offset}{This can be used to specify an \emph{a priori} known component to be included in the linear predictor during fitting.} \item{control}{A list of fit control parameters returned by \code{gam.control}.} \item{optimizer}{An array specifying the numerical optimization method to use to optimize the smoothing parameter estimation criterion; for more details see \code{\link[mgcv]{gam}}.} \item{in.out}{Optional list for initializing outer iteration; for more details see \code{\link[mgcv]{gam}}.} \item{eps}{Observations with an a-posteriori probability smaller or equal to \code{eps} are omitted in the M-step.} \item{\dots}{Additional arguments to be pased to the GAM fitter.} } \value{ Returns an object of class \code{FLXMRmgcv}. } \author{ Bettina Gruen } \seealso{ \code{\link{FLXMRglm}} } \examples{ set.seed(2012) x <- seq(0, 1, length.out = 100) z <- sample(0:1, length(x), replace = TRUE) y <- rnorm(length(x), ifelse(z, 5 * sin(x * 2 * pi), 10 * x - 5)) fitted_model <- flexmix(y ~ s(x), model = FLXMRmgcv(), cluster = z + 1, control = list(tolerance = 10^-3)) plot(y ~ x, col = clusters(fitted_model)) matplot(x, fitted(fitted_model), type = "l", add = TRUE) } \keyword{regression} \keyword{cluster} flexmix/man/boot.Rd0000644000176200001440000000607313425024236013742 0ustar liggesusers\name{boot} \alias{boot} \alias{boot,flexmix-method} \alias{LR_test} \alias{LR_test,flexmix-method} \alias{boot,flexmix-method} \alias{show,FLXboot-method} \alias{FLXboot-class} \alias{plot,FLXboot,missing-method} \alias{parameters,FLXboot-method} \alias{clusters,FLXboot,listOrdata.frame-method} \alias{predict,FLXboot-method} \alias{posterior,FLXboot,listOrdata.frame-method} \title{Bootstrap a flexmix Object} \description{ Given a \code{flexmix} object perform parametric or empirical bootstrap. } \usage{ boot(object, ...) \S4method{boot}{flexmix}(object, R, sim = c("ordinary", "empirical", "parametric"), initialize_solution = FALSE, keep_weights = FALSE, keep_groups = TRUE, verbose = 0, control, k, model = FALSE, ...) LR_test(object, ...) \S4method{LR_test}{flexmix}(object, R, alternative = c("greater", "less"), control, ...) } \arguments{ \item{object}{A fitted finite mixture model of class \code{flexmix}.} \item{R}{The number of bootstrap replicates.} \item{sim}{A character string indicating the type of simulation required. Possible values are \code{"ordinary"} (the default), \code{"parametric"}, or \code{"empirical"}.} \item{initialize_solution}{A logical. If \code{TRUE} the EM algorithm is initialized in the given solution.} \item{keep_weights}{A logical. If \code{TRUE} the weights are kept.} \item{keep_groups}{A logical. If \code{TRUE} the groups are kept.} \item{verbose}{If a positive integer, then progress information is reported every \code{verbose} iterations. If 0, no output is generated during the bootstrap replications.} \item{control}{Object of class \code{FLXcontrol} or a named list. If missing the control of the fitted \code{object} is taken.} \item{k}{Vector of integers specifying for which number of components finite mixtures are fitted to the bootstrap samples. If missing the number of components of the fitted \code{object} are taken.} \item{alternative}{A character string specifying the alternative hypothesis, must be either \code{"greater"} (default) or \code{"less"} indicating if the alternative hypothesis is that the mixture has one more component or one less.} \item{model}{A logical. If \code{TRUE} the model and the weights slot for each sample are stored and returned.} \item{\dots}{Further arguments to be passed to or from methods.} } \value{ \code{boot} returns an object of class \code{FLXboot} which contains the fitted parameters, the fitted priors, the log likelihoods, the number of components of the fitted mixtures and the information if the EM algorithm has converged. \code{LR_test} returns an object of class \code{htest} containing the number of valid bootstrap replicates, the p-value, the - twice log likelihood ratio test statistics for the original data and the bootstrap replicates. } \author{Bettina Gruen} \examples{ data("NPreg", package = "flexmix") fitted <- initFlexmix(yn ~ x + I(x^2) | id2, data = NPreg, k = 2) \dontrun{ lrtest <- LR_test(fitted, alternative = "greater", R = 20, verbose = 1) } } \keyword{methods} flexmix/man/FLXcontrol-class.Rd0000644000176200001440000000333413425024236016131 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: FLXcontrol-class.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{FLXcontrol-class} \docType{class} \alias{FLXcontrol-class} \alias{coerce,list,FLXcontrol-method} \alias{coerce,NULL,FLXcontrol-method} \title{Class "FLXcontrol"} \description{Hyperparameters for the EM algorithm.} \section{Objects from the Class}{ Objects can be created by calls of the form \code{new("FLXcontrol", ...)}. In addition, named lists can be coerced to \code{FLXcontrol} objects, names are completed if unique (see examples). } \section{Slots}{ \describe{ \item{\code{iter.max}:}{Maximum number of iterations.} \item{\code{minprior}:}{Minimum prior probability of clusters, components falling below this threshold are removed during the iteration.} \item{\code{tolerance}:}{The EM algorithm is stopped when the (relative) change of log-likelihood is smaller than \code{tolerance}.} \item{\code{verbose}:}{If a positive integer, then the log-likelihood is reported every \code{verbose} iterations. If 0, no output is generated during model fitting.} \item{\code{classify}:}{Character string, one of \code{"auto"}, \code{"weighted"}, \code{"hard"} (or \code{"CEM"}), \code{"random"} or (\code{"SEM"}).} \item{\code{nrep}:}{Reports the number of random initializations used in \code{\link{stepFlexmix}()} to determine the mixture.} } Run \code{new("FLXcontrol")} to see the default settings of all slots. } \author{Friedrich Leisch and Bettina Gruen} \keyword{classes} \examples{ ## have a look at the defaults new("FLXcontrol") ## corce a list mycont <- list(iter = 200, tol = 0.001, class = "r") as(mycont, "FLXcontrol") } flexmix/man/FLXconcomitant.Rd0000644000176200001440000000177513425024236015673 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: FLXconcomitant.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{FLXP} \docType{class} \alias{FLXPconstant} \alias{FLXPmultinom} \alias{FLXconstant} \alias{FLXmultinom} \alias{show,FLXP-method} \title{Creates the Concomitant Variable Model} \description{ Creator functions for the concomitant variable model. \code{FLXPconstant} specifies constant priors and \code{FLXPmultinom} multinomial logit models for the priors. } \usage{ FLXPconstant() FLXPmultinom(formula = ~1) } \arguments{ \item{formula}{A formula for determining the model matrix of the concomitant variables.} } \details{ \code{FLXPmultinom} uses \code{nnet.default} from \pkg{nnet} to fit the multinomial logit model. } \value{ Object of class \code{FLXP}. \code{FLXPmultinom} returns an object of class \code{FLXPmultinom} which extends class \code{FLXP} directly and is used for method dispatching. } \author{Friedrich Leisch and Bettina Gruen} \keyword{models} flexmix/man/trypanosome.Rd0000644000176200001440000000321513425024236015352 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: trypanosome.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{trypanosome} \alias{trypanosome} \docType{data} \title{Trypanosome} \description{ Trypanosome data from a dosage-response analysis to assess the proportion of organisms belonging to different populations. It is assumed that organisms belonging to different populations are indistinguishable other than in terms of their reaction to the stimulus. } \usage{data("trypanosome")} \format{ A data frame with 426 observations on the following 2 variables. \describe{ \item{\code{Dead}}{A logical vector.} \item{\code{Dose}}{A numeric vector.} } } \details{ The experimental technique involved inspection under the microscope of a representative aliquot of a suspension, all organisms appearing within two fields of view being classified either alive or dead. Hence the total numbers of organisms present at each dose and the number showing the quantal response were both random variables. } \source{ R. Ashford and P.J. Walker. Quantal Response Analysis for a Mixture of Populations. \emph{Biometrics}, \bold{28}, 981--988, 1972. } \references{ D.A. Follmann and D. Lambert. Generalizing Logistic Regression by Nonparametric Mixing. \emph{Journal of the American Statistical Association}, \bold{84}(405), 195--300, 1989. } \examples{ data("trypanosome", package = "flexmix") trypMix <- initFlexmix(cbind(Dead, 1-Dead) ~ 1, k = 2, nrep = 5, data = trypanosome, model = FLXMRglmfix(family = "binomial", fixed = ~log(Dose))) } \keyword{datasets} flexmix/man/betablocker.Rd0000644000176200001440000000276613425024236015261 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: betablocker.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{betablocker} \alias{betablocker} \docType{data} \title{Clinical Trial of Beta-Blockers} \description{ 22-centre clinical trial of beta-blockers for reducing mortality after myocardial infarction. } \usage{data("betablocker")} \format{ A data frame with 44 observations on the following 4 variables. \describe{ \item{Deaths}{Number of deaths.} \item{Total}{Total number of patients.} \item{Center}{Number of clinical centre.} \item{Treatment}{A factor with levels \code{Control} and \code{Treated}.} } } \source{ G. McLachlan and D. Peel. \emph{Finite Mixture Models}, 2000. John Wiley and Sons Inc. \url{http://www.maths.uq.edu.au/~gjm/DATA/mmdata.html} } \references{ M. Aitkin. Meta-analysis by random effect modelling in generalized linear models. \emph{Statistics in Medicine}, \bold{18}, 2343--2351, 1999. S. Yusuf, R. Peto, J. Lewis, R. Collins and P. Sleight. Beta blockade during and after myocardial infarction: an overview of the randomized trials. \emph{Progress in Cardiovascular Diseases}, \bold{27}, 335--371, 1985. } \examples{ data("betablocker", package = "flexmix") betaMix <- initFlexmix(cbind(Deaths, Total - Deaths) ~ 1 | Center, data = betablocker, k = 3, nrep = 5, model = FLXMRglmfix(family = "binomial", fixed = ~Treatment)) } \keyword{datasets} flexmix/man/refit.Rd0000644000176200001440000001204613425024236014105 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: refit.Rd 5115 2017-04-07 08:18:13Z gruen $ % \name{refit-methods} \alias{refit,flexmix-method} \alias{FLXRmstep-class} \alias{FLXRoptim-class} \alias{show,FLXR-method} \alias{show,Coefmat-method} \alias{summary,FLXRoptim-method} \alias{summary,FLXRmstep-method} \alias{plot,FLXRoptim,missing-method} \title{Refit a Fitted Model} \description{ Refits an estimated flexmix model to obtain additional information like coefficient significance p-values for GLM regression. } \usage{ \S4method{refit}{flexmix}(object, newdata, method = c("optim", "mstep"), ...) \S4method{summary}{FLXRoptim}(object, model = 1, which = c("model", "concomitant"), ...) \S4method{summary}{FLXRmstep}(object, model = 1, which = c("model", "concomitant"), ...) \S4method{plot}{FLXRoptim,missing}(x, y, model = 1, which = c("model", "concomitant"), bycluster = TRUE, alpha = 0.05, components, labels = NULL, significance = FALSE, xlab = NULL, ylab = NULL, ci = TRUE, scales = list(), as.table = TRUE, horizontal = TRUE, ...) } \arguments{ \item{object}{An object of class \code{"flexmix"}} \item{newdata}{Optional new data.} \item{method}{Specifies if the variance covariance matrix is determined using \code{\link{optim}} or if the posteriors are assumed as given and an M-step is performed.} \item{model}{The model (for a multivariate response) that shall be used.} \item{which}{Specifies if a component specific model or the concomitant variable model is used.} \item{x}{An object of class \code{"FLXRoptim"}} \item{y}{Missing object.} \item{bycluster}{A logical if the parameters should be group by cluster or by variable.} \item{alpha}{Numeric indicating the significance level.} \item{components}{Numeric vector specifying which components are plotted. The default is to plot all components.} \item{labels}{Character vector specifying the variable names used.} \item{significance}{A logical indicating if non-significant coefficients are shaded in a lighter grey.} \item{xlab}{String for the x-axis label.} \item{ylab}{String for the y-axis label.} \item{ci}{A logical indicating if significant and insignificant parameter estimates are shaded differently.} \item{scales}{See argument of the same name for function \code{\link[lattice]{xyplot}}.} \item{as.table}{See arguments of the same name for function \code{\link[lattice]{xyplot}}.} \item{horizontal}{See arguments of the same name for function \code{\link[lattice]{xyplot}}.} \item{\dots}{Currently not used} } \value{ An object inheriting form class \code{FLXR} is returned. For the method using \code{optim} the object has class \code{FLXRoptim} and for the M-step method it has class \code{FLXRmstep}. Both classes give similar results for their \code{summary} methods. Objects of class \code{FLXRoptim} have their own \code{plot} method. \code{Lapply} can be used to further analyse the refitted component specific models of objects of class \code{FLXRmstep}. } \details{ The \code{refit} method for \code{FLXMRglm} models in combination with the \code{summary} method can be used to obtain the usual tests for significance of coefficients. Note that the tests are valid only if \code{flexmix} returned the maximum likelihood estimator of the parameters. If \code{refit} is used with \code{method = "mstep"} for these component specific models the returned object contains a \code{glm} object for each component where the elements \code{model} which is the model frame and \code{data} which contains the original dataset are missing. } \keyword{methods} \author{Friedrich Leisch and Bettina Gruen} \references{ Friedrich Leisch. FlexMix: A general framework for finite mixture models and latent class regression in R. \emph{Journal of Statistical Software}, \bold{11}(8), 2004. doi:10.18637/jss.v011.i08 } \section{Warning}{ For \code{method = "mstep"} the standard deviations are determined separately for each of the components using the a-posteriori probabilities as weights without accounting for the fact that the components have been simultaneously estimated. The derived standard deviations are hence approximative and should only be used in an exploratory way, as they are underestimating the uncertainty given that the missing information of the component memberships are replaced by the expected values. The \code{newdata} argument can only be specified when using \code{method = "mstep"} for refitting \code{FLXMRglm} components. A variant of \code{glm} for weighted ML estimation is used for fitting the components and full \code{glm} objects are returned. Please note that in this case the data and the model frame are stored for each component which can significantly increase the object size. } \examples{ data("NPreg", package = "flexmix") ex1 <- flexmix(yn ~ x + I(x^2), data = NPreg, k = 2) ex1r <- refit(ex1) ## in one component all coefficients should be highly significant, ## in the other component only the linear term summary(ex1r) } flexmix/man/FLXMCfactanal.Rd0000644000176200001440000000452613425024236015343 0ustar liggesusers\name{FLXMCfactanal} \alias{FLXMCfactanal} \alias{rFLXM,FLXMCfactanal,FLXcomponent-method} \title{Driver for Mixtures of Factor Analyzers} \description{ This driver for \code{\link{flexmix}} implements estimation of mixtures of factor analyzers using ML estimation of factor analysis implemented in \code{factanal} in each M-step. } \usage{ FLXMCfactanal(formula = . ~ ., factors = 1, ...) } \arguments{ \item{formula}{A formula which is interpreted relative to the formula specified in the call to \code{\link{flexmix}} using \code{\link{update.formula}}. Only the left-hand side (response) of the formula is used. Default is to use the original \code{\link{flexmix}} model formula.} \item{factors}{Integer specifying the number of factors in each component.} \item{\dots}{Passed to \code{factanal}} } \value{ \code{FLXMCfactanal} returns an object of class \code{FLXM}. } \references{ G. McLachlan and D. Peel. \emph{Finite Mixture Models}, 2000. John Wiley and Sons Inc. } \author{Bettina Gruen} \section{Warning}{ This does not implement the AECM framework presented in McLachlan and Peel (2000, p.245), but uses the available functionality in R for ML estimation of factor analyzers. The implementation therefore is only experimental and has not been well tested. Please note that in general a good initialization is crucial for the EM algorithm to converge to a suitable solution for this model class. } \seealso{\code{\link{flexmix}}} \examples{ ## Reproduce (partly) Table 8.1. p.255 (McLachlan and Peel, 2000) if (require("gclus")) { data("wine", package = "gclus") wine_data <- as.matrix(wine[,-1]) set.seed(123) wine_fl_diag <- initFlexmix(wine_data ~ 1, k = 3, nrep = 10, model = FLXMCmvnorm(diagonal = TRUE)) wine_fl_fact <- lapply(1:4, function(q) flexmix(wine_data ~ 1, model = FLXMCfactanal(factors = q, nstart = 3), cluster = posterior(wine_fl_diag))) sapply(wine_fl_fact, logLik) ## FULL set.seed(123) wine_full <- initFlexmix(wine_data ~ 1, k = 3, nrep = 10, model = FLXMCmvnorm(diagonal = FALSE)) logLik(wine_full) ## TRUE wine_true <- flexmix(wine_data ~ 1, cluster = wine$Class, model = FLXMCmvnorm(diagonal = FALSE)) logLik(wine_true) } } \keyword{models} flexmix/man/FLXP-class.Rd0000644000176200001440000000211413425024236014643 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: FLXP-class.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{FLXP-class} \docType{class} \alias{FLXP-class} \title{Class "FLXP"} \alias{initialize,FLXP-method} \alias{FLXPconstant-class} \alias{FLXPmultinom-class} \description{ Concomitant model class. } \section{Objects from the Class}{ Objects can be created by calls of the form \code{new("FLXP", ...)}, typically inside driver functions like \code{\link{FLXPconstant}} or \code{\link{FLXPmultinom}}. } \section{Slots}{ \describe{ \item{\code{name}:}{Character string used in print methods.} \item{\code{formula}:}{Formula describing the model.} \item{\code{x}:}{Model matrix.} \item{\code{fit}:}{Function returning the fitted prior probabilities.} \item{\code{refit}:}{Function returning the fitted concomitant model.} \item{\code{coef}:}{Matrix containing the fitted parameters.} \item{\code{df}:}{Function for determining the number of degrees of freedom used.} } } \author{Friedrich Leisch and Bettina Gruen} \keyword{classes} flexmix/man/FLXMRglmnet.Rd0000644000176200001440000000657213425024236015102 0ustar liggesusers\name{FLXMRglmnet} \alias{FLXMRglmnet} \alias{FLXMRglmnet-class} \title{FlexMix Interface for Adaptive Lasso / Elastic Net with GLMs} \description{ This is a driver which allows fitting of mixtures of GLMs where the coefficients are penalized using the (adaptive) lasso or the elastic net by reusing functionality from package \pkg{glmnet}. } \usage{ FLXMRglmnet(formula = . ~ ., family = c("gaussian", "binomial", "poisson"), adaptive = TRUE, select = TRUE, offset = NULL, ...) } \arguments{ \item{formula}{A formula which is interpreted relative to the formula specified in the call to \code{\link{flexmix}} using \code{\link{update.formula}}. Default is to use the original \code{\link{flexmix}} model formula.} \item{family}{A character string naming a \code{\link{glm}} family function.} \item{adaptive}{A logical indicating if the adaptive lasso should be used. By default equal to \code{TRUE}.} \item{select}{A logical vector indicating which variables in the model matrix should be included in the penalized part. By default equal to \code{TRUE} implying that all variables are penalized.} \item{offset}{This can be used to specify an \emph{a priori} known component to be included in the linear predictor during fitting.} \item{\dots}{Additional arguments to be passed to \code{\link[glmnet]{cv.glmnet}} fitter.} } \details{ Some care is needed to ensure convergence of the algorithm, which is computationally more challenging than a standard EM. In the proposed method, not only are cluster allocations identified and component parameters estimated as commonly done in mixture models, but there is also variable selection via penalized regression using $k$-fold cross-validation to choose the penalty parameter. For the algorithm to converge, it is necessary that the same cross-validation partitioning be used across the EM iterations, i.e., the subsamples for cross-validation must be defined at the beginning This is accomplished using the \code{foldid} option as an additional parameter to be passed to \code{\link[glmnet]{cv.glmnet}} (see \pkg{glmnet} package documentation). } \value{ Returns an object of class \code{FLXMRglm}. } \author{ Frederic Mortier and Nicolas Picard. } \seealso{ \code{\link{FLXMRglm}} } \references{ Frederic Mortier, Dakis-Yaoba Ouedraogo, Florian Claeys, Mahlet G. Tadesse, Guillaume Cornu, Fidele Baya, Fabrice Benedet, Vincent Freycon, Sylvie Gourlet-Fleury and Nicolas Picard. Mixture of inhomogeneous matrix models for species-rich ecosystems. \emph{Environmetrics}, \bold{26}(1), 39-51, 2015. doi:10.1002/env.2320 } \examples{ set.seed(12) p <- 10 beta <- matrix(0, nrow = p + 1, ncol = 2) beta[1,] <- c(-1, 1) beta[cbind(c(5, 10), c(1, 2))] <- 1 nobs <- 100 X <- matrix(rnorm(nobs * p), nobs, p) mu <- cbind(1, X) \%*\% beta z <- sample(1:ncol(beta), nobs, replace = TRUE) y <- mu[cbind(1:nobs, z)] + rnorm(nobs) data <- data.frame(y, X) ## The maximum number of iterations is reduced to ## avoid a long running time. fo <- sample(rep(seq(10), length = nrow(data))) ex1 <- flexmix(y ~ ., data = data, k = 2, cluster = z, model = FLXMRglmnet(foldid = fo), control = list(iter.max = 2)) parameters(ex1) } \keyword{regression} \keyword{cluster} flexmix/man/patent.Rd0000644000176200001440000000312713425024236014267 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: patent.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{patent} \alias{patent} \docType{data} \title{Patents and R&D Spending} \description{ Number of patents, R&D spending and sales in millions of dollar for 70 pharmaceutical and biomedical companies in 1976. } \usage{data("patent")} \format{ A data frame with 70 observations on the following 4 variables. \describe{ \item{Company}{Name of company.} \item{Patents}{Number of patents.} \item{RDS}{R&D spending per sales.} \item{lgRD}{Logarithmized R&D spendings (in millions of dollars).} } } \details{ The data is taken from the National Bureau of Economic Research R\&D Masterfile. } \source{ P. Wang, I.M. Cockburn and M.L. Puterman. Analysis of Patent Data -- A Mixed-Poisson-Regression-Model Approach. \emph{Journal of Business & Economic Statistics}, \bold{16}(1), 27--41, 1998. } \references{ B.H. Hall, C. Cummins, E. Laderman and J. Mundy. The R&D Master File Documentation. Technical Working Paper 72, National Bureau of Economic Research, 1988. Cambridge, MA. } \examples{ data("patent", package = "flexmix") patentMix <- initFlexmix(Patents ~ lgRD, k = 3, model = FLXMRglm(family = "poisson"), concomitant = FLXPmultinom(~RDS), nrep = 5, data = patent) plot(Patents ~ lgRD, data = patent, pch = as.character(clusters(patentMix))) ordering <- order(patent$lgRD) apply(fitted(patentMix), 2, function(y) lines(sort(patent$lgRD), y[ordering])) } \keyword{datasets} flexmix/man/FLXcomponent-class.Rd0000644000176200001440000000147013425024236016452 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: FLXcomponent-class.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{FLXcomponent-class} \docType{class} \alias{FLXcomponent-class} \alias{show,FLXcomponent-method} \title{Class "FLXcomponent"} \description{A fitted component of a \code{\link{flexmix}} model.} \section{Objects from the Class}{ Objects can be created by calls of the form \code{new("FLXcomponent", ...)}. } \section{Slots}{ \describe{ \item{\code{df}:}{Number of parameters used by the component.} \item{\code{logLik}:}{Function computing the log-likelihood of observations.} \item{\code{parameters}:}{List with model parameters.} \item{\code{predict}:}{Function predicting response for new data.} } } \author{Friedrich Leisch and Bettina Gruen} \keyword{classes} flexmix/man/salmonellaTA98.Rd0000644000176200001440000000342413425024236015531 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: salmonellaTA98.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{salmonellaTA98} \alias{salmonellaTA98} \title{Salmonella Reverse Mutagenicity Assay} \usage{data("salmonellaTA98")} \description{ Data on Ames Salmonella reverse mutagenicity assay. } \format{ This data frame contains the following columns: \describe{ \item{x}{Dose levels of quinoline.} \item{y}{Numbers of revertant colonies of TA98 Salmonella observed on each of three replicate plates tested at each of six dose levels of quinoline diameter.} } } \details{ This data set is taken from package \pkg{dispmod} provided by Luca Scrucca. } \source{ Margolin, B.J., Kaplan, N. and Zeiger, E. Statistical analysis of the Ames Salmonella/microsome test, \emph{Proc. Natl. Acad. Sci. USA}, \bold{76}, 3779--3783, 1981. } \references{ Breslow, N.E. Extra-Poisson variation in log-linear models, \emph{Applied Statistics}, \bold{33}, 38--44, 1984. Wang, P., Puterman, M.L., Cockburn, I.M., and Le, N.D. Mixed Poisson regression models with covariate dependent rates, \emph{Biometrics}, \bold{52}, 381--400, 1996. } \examples{ data("salmonellaTA98", package = "flexmix") salmonMix <- initFlexmix(y ~ 1, data = salmonellaTA98, model = FLXMRglmfix(family = "poisson", fixed = ~ x + log(x + 10)), k = 2, nrep = 5) salmonMix.pr <- predict(salmonMix, newdata = salmonellaTA98) plot(y ~ x, data = salmonellaTA98, pch = as.character(clusters(salmonMix)), ylim = range(c(salmonellaTA98$y, unlist(salmonMix.pr)))) for (i in 1:2) lines(salmonellaTA98$x, salmonMix.pr[[i]], lty = i) } \keyword{datasets} flexmix/man/fitted.Rd0000644000176200001440000000213513425024236014251 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: fitted.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{fitted-methods} \docType{methods} \alias{fitted,flexmix-method} \alias{fitted,FLXM-method} \alias{fitted,FLXR-method} \alias{fitted,FLXRMRglm-method} \title{Extract Model Fitted Values} \description{ Extract fitted values for each component from a flexmix object. } \usage{ \S4method{fitted}{flexmix}(object, drop = TRUE, aggregate = FALSE, ...) } \arguments{ \item{object}{an object of class \code{"flexmix"} or \code{"FLXR"}} \item{drop}{logical, if \code{TRUE} then the function tries to simplify the return object by combining lists of length 1 into matrices.} \item{aggregate}{logical, if \code{TRUE} then the fitted values for each model aggregated over the components are returned.} \item{\dots}{currently not used} } \keyword{methods} \author{Friedrich Leisch and Bettina Gruen} \examples{ data("NPreg", package = "flexmix") ex1 <- flexmix(yn ~ x + I(x^2), data = NPreg, k = 2) matplot(NPreg$x, fitted(ex1), pch = 16, type = "p") points(NPreg$x, NPreg$yn) } flexmix/man/whiskey.Rd0000644000176200001440000000300313425024236014450 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: whiskey.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{whiskey} \alias{whiskey} \alias{whiskey_brands} \docType{data} \title{Survey Data on Brands of Scotch whiskey Consumed} \description{ The data set is from Simmons Study of Media and Markets and contains the incidence matrix for scotch brands used in last year for those households who report consuming scotch. } \usage{data("whiskey")} \format{ A data frame \code{whiskey} with 484 observations on the following 2 variables. \describe{ \item{\code{Freq}}{a numeric vector} \item{\code{Incidence}}{a matrix with 21 columns} } Additional information on the brands is contained in the data frame \code{whiskey_brands} which is simultaneously loaded. This data frame contains 21 observations on the following 3 variables. \describe{ \item{\code{Brand}}{a character vector} \item{\code{Type}}{a factor with levels \code{Blend} \code{Single Malt}} \item{\code{Bottled}}{a factor with levels \code{Domestic} \code{Foreign}} } } \details{ The dataset is taken from the \pkg{bayesm} package. } \source{ Peter Rossi and Rob McCulloch. bayesm: Bayesian Inference for Marketing/Micro-econometrics. R package version 2.0-8, 2006. http://gsbwww.uchicago.edu/fac/peter.rossi/research/bsm.html } \references{ Edwards, Y. and G. Allenby. Multivariate Analysis of Multiple Response Data, \emph{Journal of Marketing Research}, \bold{40}, 321--334, 2003. } \keyword{datasets} flexmix/man/logLik-methods.Rd0000644000176200001440000000105413425024236015653 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: logLik-methods.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{logLik-methods} \docType{methods} \title{Methods for Function logLik in Package \pkg{flexmix}} \alias{logLik,flexmix-method} \alias{logLik,stepFlexmix-method} \description{Evaluate the log-likelihood. This function is defined as an S4 generic in the \code{stats4} package.} \section{Methods}{ \describe{ \item{object = flexmix}{Evaluate the log-likelihood of an \code{flexmix} object} } } \keyword{methods} flexmix/man/tribolium.Rd0000644000176200001440000000356613425024236015011 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: tribolium.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{tribolium} \alias{tribolium} \docType{data} \title{Tribolium Beetles} \description{ The data investigates whether the adult Tribolium species Castaneum has developed an evolutionary advantage to recognize and avoid eggs of their own species while foraging. } \usage{data("tribolium")} \format{ A data frame with 27 observations on the following 4 variables. \describe{ \item{\code{Remaining}}{A numeric vector.} \item{\code{Total}}{A numeric vector.} \item{\code{Replicate}}{A factor with levels \code{1}, \code{2}, \code{3}.} \item{\code{Species}}{A factor with levels \code{Castaneum} \code{Confusum} \code{Madens}.} } } \details{ Beetles of the genus Tribolium are cannibalistic in the sense that adults eat the eggs of their own species as well as those of closely related species. The experiment isolated a number of adult beetles of the same species and presented them with a vial of 150 eggs (50 of each type), the eggs being thoroughly mixed to ensure uniformity throughout the vial. The data gives the consumption data for adult Castaneum species. It reports the number of Castaneum, Confusum and Madens eggs, respectively, that remain uneaten after two day exposure to the adult beetles. Replicates 1, 2, and 3 correspond to different occasions on which the experiment was conducted. } \source{ P. Wang and M.L. Puterman. Mixed Logistic Regression Models. \emph{Journal of Agricultural, Biological, and Environmental Statistics}, \bold{3} (2), 175--200, 1998. } \examples{ data("tribolium", package = "flexmix") tribMix <- initFlexmix(cbind(Remaining, Total - Remaining) ~ Species, k = 2, nrep = 5, data = tribolium, model = FLXMRglm(family = "binomial")) } \keyword{datasets} flexmix/man/FLXMCdist1.Rd0000644000176200001440000000371313425024236014613 0ustar liggesusers% % Copyright (C) 2016 Bettina Gruen % $Id: FLXMCdist1.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{FLXMCdist1} \alias{FLXMCdist1} \title{FlexMix Clustering of Univariate Distributions} \description{ These are drivers for \code{\link{flexmix}} implementing model-based clustering of univariate data using different distributions for the component-specific models. } \usage{ FLXMCdist1(formula = . ~ ., dist, ...) } \arguments{ \item{formula}{A formula which is interpreted relative to the formula specified in the call to \code{\link{flexmix}} using \code{\link{update.formula}}. Only the left-hand side (response) of the formula is used. Default is to use the original \code{\link{flexmix}} model formula.} \item{dist}{Character string indicating the component-specific univariate distribution.} \item{...}{Arguments for the specific model drivers.} } \details{ Currently drivers for the following distributions are available: \enumerate{ \item Lognormal (\code{"lnorm"}) \item inverse Gaussian (\code{"invGauss"} using \code{\link[SuppDists]{dinvGauss}}) \item gamma (\code{"gamma"}) \item exponential (\code{"exp"}) \item Weibull (\code{"weibull"}) \item Burr (\code{"burr"} using \code{\link[actuar]{dburr}}) \item Inverse Burr (\code{"invburr"} using \code{\link[actuar]{dinvburr}}) } } \value{ \code{FLXMCdist1} returns an object of class \code{FLXMC}. } \author{Friedrich Leisch and Bettina Gruen} \references{ Tatjana Miljkovic and Bettina Gruen. Modeling loss data using mixtures of distributions. \emph{Insurance: Mathematics and Economics}, \bold{70}, 387-396, 2016. doi:10.1016/j.insmatheco.2016.06.019 } \seealso{\code{\link{flexmix}}} \keyword{cluster} \examples{ if (require("actuar")) { set.seed(123) y <- c(rexp(100, 10), rexp(100, 1)) ex <- flexmix(y ~ 1, cluster = rep(1:2, each = 100), model = FLXMCdist1(dist = "exp")) parameters(ex) } } flexmix/man/posterior.Rd0000644000176200001440000000225513425024236015023 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: posterior.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{posterior} \alias{clusters,flexmix,missing-method} \alias{clusters,FLXdist,ANY-method} \alias{posterior,flexmix,missing-method} \alias{posterior,FLXdist,listOrdata.frame-method} \title{Determine Cluster Membership and Posterior Probabilities} \description{Determine posterior probabilities or cluster memberships for a fitted \code{flexmix} or unfitted \code{FLXdist} model.} \usage{ \S4method{posterior}{flexmix,missing}(object, newdata, unscaled = FALSE, ...) \S4method{posterior}{FLXdist,listOrdata.frame}(object, newdata, unscaled = FALSE, ...) \S4method{clusters}{flexmix,missing}(object, newdata, ...) \S4method{clusters}{FLXdist,ANY}(object, newdata, ...) } \arguments{ \item{object}{An object of class "flexmix" or "FLXdist".} \item{newdata}{Data frame or list containing new data. If missing the posteriors of the original observations are returned.} \item{unscaled}{Logical, if \code{TRUE} the component-specific likelihoods are returned.} \item{\dots}{Currently not used.} } \author{Friedrich Leisch and Bettina Gruen} \keyword{methods} flexmix/man/relabel.Rd0000644000176200001440000000357313425024236014407 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: posterior.Rd 3937 2008-03-28 14:56:01Z leisch $ % \name{relabel} \alias{relabel} \alias{relabel,FLXdist,missing-method} \alias{relabel,FLXdist,character-method} \alias{relabel,FLXdist,integer-method} \title{Relabel the Components} \description{ The components are sorted by the value of one of the parameters or according to an integer vector containing the permutation of the numbers from 1 to the number of components. } \usage{ relabel(object, by, ...) \S4method{relabel}{FLXdist,character}(object, by, which = NULL, ...) } \arguments{ \item{object}{An object of class \code{"flexmix"}.} \item{by}{If a character vector, it needs to be one of \code{"prior"}, \code{"model"}, \code{"concomitant"} indicating if the parameter should be from the component-specific or the concomitant variable model. If an integer vector it indicates how the components should be sorted. If missing, the components are sorted by component size.} \item{which}{Name (or unique substring) of a parameter if \code{by} is equal to "model" or "concomitant".} \item{\dots}{Currently not used.} } \author{Friedrich Leisch and Bettina Gruen} \keyword{methods} \examples{ set.seed(123) beta <- matrix(1:16, ncol = 4) beta df1 <- ExLinear(beta, n = 100, sd = .5) f1 <- flexmix(y~., data = df1, k = 4) ## There was label switching, parameters are not in the same order ## as in beta: round(parameters(f1)) betas <- rbind(beta, .5) betas ## This makes no sense: summary(abs(as.vector(betas-parameters(f1)))) ## We relabel the components by sorting the coefficients of x1: r1 <- relabel(f1, by = "model", which = "x1") round(parameters(r1)) ## Now we can easily compare the fit with the true parameters: summary(abs(as.vector(betas-parameters(r1)))) } flexmix/man/KLdiv.Rd0000644000176200001440000000532413425024236014006 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: KLdiv.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{KLdiv} \alias{KLdiv,matrix-method} \alias{KLdiv,flexmix-method} \alias{KLdiv,FLXMRglm-method} \alias{KLdiv,FLXMC-method} \title{Kullback-Leibler Divergence} \description{ Estimate the Kullback-Leibler divergence of several distributions.} \usage{ \S4method{KLdiv}{matrix}(object, eps = 10^-4, overlap = TRUE,...) \S4method{KLdiv}{flexmix}(object, method = c("continuous", "discrete"), ...) } \arguments{ \item{object}{See Methods section below.} \item{method}{The method to be used; "continuous" determines the Kullback-Leibler divergence between the unweighted theoretical component distributions and the unweighted posterior probabilities at the observed points are used by "discrete".} \item{eps}{Probabilities below this threshold are replaced by this threshold for numerical stability.} \item{overlap}{Logical, do not determine the KL divergence for those pairs where for each point at least one of the densities has a value smaller than \code{eps}.} \item{...}{Passed to the matrix method.} } \section{Methods}{ \describe{ \item{object = "matrix":}{Takes as input a matrix of density values with one row per observation and one column per distribution.} \item{object = "flexmix":}{Returns the Kullback-Leibler divergence of the mixture components.} }} \details{ Estimates \deqn{\int f(x) (\log f(x) - \log g(x)) dx} for distributions with densities \eqn{f()} and \eqn{g()}. } \value{ A matrix of KL divergences where the rows correspond to using the respective distribution as \eqn{f()} in the formula above. } \note{ The density functions are modified to have equal support. A weight of at least \code{eps} is given to each observation point for the modified densities. } \keyword{methods} \author{Friedrich Leisch and Bettina Gruen} \references{ S. Kullback and R. A. Leibler. On information and sufficiency.\emph{The Annals of Mathematical Statistics}, \bold{22}(1), 79--86, 1951. Friedrich Leisch. Exploring the structure of mixture model components. In Jaromir Antoch, editor, Compstat 2004--Proceedings in Computational Statistics, 1405--1412. Physika Verlag, Heidelberg, Germany, 2004. ISBN 3-7908-1554-3. } \examples{ ## Gaussian and Student t are much closer to each other than ## to the uniform: x <- seq(-3, 3, length = 200) y <- cbind(u = dunif(x), n = dnorm(x), t = dt(x, df = 10)) matplot(x, y, type = "l") KLdiv(y) if (require("mlbench")) { set.seed(2606) x <- mlbench.smiley()$x model1 <- flexmix(x ~ 1, k = 9, model = FLXmclust(diag = FALSE), control = list(minprior = 0)) plotEll(model1, x) KLdiv(model1) } }flexmix/man/NregFix.Rd0000644000176200001440000000243213425024236014334 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: NregFix.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{NregFix} \alias{NregFix} \title{Artificial Example for Normal Regression} \description{ A simple artificial regression example with 3 latent classes, two independent variables, one concomitant variable and a dependent variable which follows a Gaussian distribution. } \usage{ data("NregFix") } \format{ A data frame with 200 observations on the following 5 variables. \describe{ \item{\code{x1}}{Independent variable: numeric variable.} \item{\code{x2}}{Independent variable: a factor with two levels: \code{0} and \code{1}.} \item{\code{w}}{Concomitant variable: a factor with two levels: \code{0} and \code{1}.} \item{\code{y}}{Dependent variable.} \item{\code{class}}{Latent class memberships.} } } \examples{ data("NregFix", package = "flexmix") library("lattice") xyplot(y ~ x1 | x2 * w, data = NregFix, groups = class) Model <- FLXMRglmfix(~ 1, fixed = ~ x2, nested = list(k = c(2, 1), formula = c(~x1, ~0))) fittedModel <- initFlexmix(y ~ 1, model = Model, data = NregFix, k = 3, concomitant = FLXPmultinom(~ w), nrep = 5) fittedModel } \keyword{datasets} flexmix/man/flexmix-internal.Rd0000644000176200001440000000724513425024236016267 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: flexmix-internal.Rd 5079 2016-01-31 12:21:12Z gruen $ % \name{flexmix-internal} \alias{FLXgetModelmatrix} \alias{FLXcheckComponent} \alias{FLXcheckComponent,FLXM-method} \alias{FLXcheckComponent,FLXMRfix-method} \alias{FLXdeterminePostunscaled} \alias{FLXdeterminePostunscaled,FLXM-method} \alias{FLXdeterminePostunscaled,FLXMRfix-method} \alias{FLXdeterminePostunscaled,FLXMRlmc-method} \alias{FLXdeterminePostunscaled,FLXMRlmmc-method} \alias{FLXdeterminePostunscaled,FLXMRcondlogit-method} \alias{FLXgetK} \alias{FLXgetK,FLXM-method} \alias{FLXgetK,FLXMRfix-method} \alias{FLXgetModelmatrix} \alias{FLXgetModelmatrix,FLXM-method} \alias{FLXgetModelmatrix,FLXMCmvcombi-method} \alias{FLXgetModelmatrix,FLXMRcondlogit-method} \alias{FLXgetModelmatrix,FLXMRfix-method} \alias{FLXgetModelmatrix,FLXMRlmc-method} \alias{FLXgetModelmatrix,FLXMRlmmc-method} \alias{FLXgetModelmatrix,FLXMRmgcv-method} \alias{FLXgetModelmatrix,FLXMRrobglm-method} \alias{FLXgetModelmatrix,FLXMRziglm-method} \alias{FLXgetModelmatrix,FLXP-method} \alias{FLXgetObs} \alias{FLXgetObs,FLXM-method} \alias{FLXgetObs,FLXMRfix-method} \alias{FLXgetObs,FLXMRlmc-method} \alias{FLXmstep} \alias{FLXmstep,FLXM-method} \alias{FLXmstep,FLXMCmvcombi-method} \alias{FLXmstep,FLXMRcondlogit-method} \alias{FLXmstep,FLXMRfix-method} \alias{FLXmstep,FLXMRlmc-method} \alias{FLXmstep,FLXMRlmcfix-method} \alias{FLXmstep,FLXMRlmmc-method} \alias{FLXmstep,FLXMRlmmcfix-method} \alias{FLXmstep,FLXMRmgcv-method} \alias{FLXmstep,FLXMRrobglm-method} \alias{FLXmstep,FLXMRziglm-method} \alias{FLXremoveComponent} \alias{FLXremoveComponent,FLXM-method} \alias{FLXremoveComponent,FLXMRfix-method} \alias{FLXremoveComponent,FLXMRrobglm-method} \alias{FLXremoveComponent,FLXMRziglm-method} \alias{FLXMRglm-class} \alias{FLXR-class} \alias{FLXRMRglm-class} \alias{FLXRPmultinom-class} \alias{summary.flexmix-class} \alias{posterior,FLXM,listOrdata.frame-method} \alias{FLXMRfix-class} \alias{FLXMRglmfix-class} \alias{FLXRMRglmfix-class} \alias{predict,FLXMRglmfix-method} \alias{fitted,FLXMRglmfix-method} \alias{summary,FLXRMRglmfix-method} \alias{listOrdata.frame-class} \alias{refit_optim} \alias{refit_optim,FLXM-method} \alias{refit_optim,FLXMC-method} \alias{refit_optim,FLXMRglm-method} \alias{refit_optim,FLXMRziglm-method} \alias{refit_optim,FLXP-method} \alias{FLXgetDesign} \alias{FLXgetDesign,FLXM-method} \alias{FLXgetDesign,FLXMRglmfix-method} \alias{FLXgetDesign,FLXMRziglm-method} \alias{FLXgetParameters} \alias{FLXgetParameters,FLXdist-method} \alias{FLXgetParameters,FLXM-method} \alias{FLXgetParameters,FLXMC-method} \alias{FLXgetParameters,FLXMRglm-method} \alias{FLXgetParameters,FLXP-method} \alias{FLXgetParameters,FLXPmultinom-method} \alias{FLXreplaceParameters} \alias{FLXreplaceParameters,FLXdist-method} \alias{FLXreplaceParameters,FLXM-method} \alias{FLXreplaceParameters,FLXMC-method} \alias{FLXreplaceParameters,FLXMRglm-method} \alias{FLXreplaceParameters,FLXP-method} \alias{FLXreplaceParameters,FLXPmultinom-method} \alias{FLXlogLikfun} \alias{FLXlogLikfun,flexmix-method} \alias{FLXgradlogLikfun} \alias{FLXgradlogLikfun,flexmix-method} \alias{FLXgradlogLikfun,FLXM-method} \alias{FLXgradlogLikfun,FLXMRglm-method} \alias{FLXgradlogLikfun,FLXP-method} \alias{existGradient} \alias{existGradient,FLXM-method} \alias{existGradient,FLXMRglm-method} \alias{existGradient,FLXMRcondlogit-method} \alias{existGradient,FLXMRglmfix-method} \alias{existGradient,FLXMRmultinom-method} \alias{existGradient,FLXP-method} \title{Internal FlexMix Functions} \description{ Internal flexmix functions, methods and classes. } \details{ These are not to be called by the user. } \keyword{internal} flexmix/man/BregFix.Rd0000644000176200001440000000212113425024236014313 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: BregFix.Rd 5008 2015-01-13 20:30:25Z gruen $ % \name{BregFix} \alias{BregFix} \docType{data} \title{Artificial Example for Binomial Regression} \description{ A simple artificial regression example data set with 3 latent classes, one independent variable \code{x} and a concomitant variable \code{w}. } \usage{data("BregFix")} \format{ A data frame with 200 observations on the following 5 variables. \describe{ \item{\code{yes}}{number of successes} \item{\code{no}}{number of failures} \item{\code{x}}{independent variable} \item{\code{w}}{concomitant variable, a factor with levels \code{0} \code{1}} \item{\code{class}}{latent class memberships} } } \examples{ data("BregFix", package = "flexmix") Model <- FLXMRglmfix(family="binomial", nested = list(formula = c(~x, ~0), k = c(2, 1))) Conc <- FLXPmultinom(~w) FittedBin <- initFlexmix(cbind(yes, no) ~ 1, data = BregFix, k = 3, model = Model, concomitant = Conc) summary(FittedBin) } \keyword{datasets} flexmix/man/bioChemists.Rd0000644000176200001440000000240013425024236015236 0ustar liggesusers% % Copyright (C) 2004-2015 Friedrich Leisch and Bettina Gruen % $Id: bioChemists.Rd 5035 2015-05-06 19:59:02Z gruen $ % \name{bioChemists} \alias{bioChemists} \docType{data} \title{Articles by Graduate Students in Biochemistry Ph.D. Programs} \description{ A sample of 915 biochemistry graduate students. } \usage{ data("bioChemists") } \format{ \describe{ \item{art}{count of articles produced during last 3 years of Ph.D.} \item{fem}{factor indicating gender of student, with levels Men and Women} \item{mar}{factor indicating marital status of student, with levels Single and Married} \item{kid5}{number of children aged 5 or younger} \item{phd}{prestige of Ph.D. department} \item{ment}{count of articles produced by Ph.D. mentor during last 3 years} } } \details{ This data set is taken from package \pkg{pscl} provided by Simon Jackman. } \source{ found in Stata format at \url{http://www.indiana.edu/~jslsoc/stata/spex_data/couart2.dta} } \references{ Long, J. Scott. The origins of sex difference in science. \emph{Social Forces}, \bold{68}, 1297--1315, 1990. Long, J. Scott. \emph{Regression Models for Categorical and Limited Dependent Variables}, 1997. Thousand Oaks, California: Sage. } \keyword{datasets}