pcaMethods/COPYING0000644000175200017520000004311014710217306014716 0ustar00biocbuildbiocbuild GNU GENERAL PUBLIC LICENSE Version 2, June 1991 Copyright (C) 1989, 1991 Free Software Foundation, Inc. 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed. Preamble The licenses for most software are designed to take away your freedom to share and change it. By contrast, the GNU General Public License is intended to guarantee your freedom to share and change free software--to make sure the software is free for all its users. This General Public License applies to most of the Free Software Foundation's software and to any other program whose authors commit to using it. (Some other Free Software Foundation software is covered by the GNU Library General Public License instead.) You can apply it to your programs, too. When we speak of free software, we are referring to freedom, not price. Our General Public Licenses are designed to make sure that you have the freedom to distribute copies of free software (and charge for this service if you wish), that you receive source code or can get it if you want it, that you can change the software or use pieces of it in new free programs; and that you know you can do these things. To protect your rights, we need to make restrictions that forbid anyone to deny you these rights or to ask you to surrender the rights. These restrictions translate to certain responsibilities for you if you distribute copies of the software, or if you modify it. For example, if you distribute copies of such a program, whether gratis or for a fee, you must give the recipients all the rights that you have. You must make sure that they, too, receive or can get the source code. And you must show them these terms so they know their rights. We protect your rights with two steps: (1) copyright the software, and (2) offer you this license which gives you legal permission to copy, distribute and/or modify the software. Also, for each author's protection and ours, we want to make certain that everyone understands that there is no warranty for this free software. If the software is modified by someone else and passed on, we want its recipients to know that what they have is not the original, so that any problems introduced by others will not reflect on the original authors' reputations. Finally, any free program is threatened constantly by software patents. We wish to avoid the danger that redistributors of a free program will individually obtain patent licenses, in effect making the program proprietary. To prevent this, we have made it clear that any patent must be licensed for everyone's free use or not licensed at all. The precise terms and conditions for copying, distribution and modification follow. GNU GENERAL PUBLIC LICENSE TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION 0. This License applies to any program or other work which contains a notice placed by the copyright holder saying it may be distributed under the terms of this General Public License. The "Program", below, refers to any such program or work, and a "work based on the Program" means either the Program or any derivative work under copyright law: that is to say, a work containing the Program or a portion of it, either verbatim or with modifications and/or translated into another language. (Hereinafter, translation is included without limitation in the term "modification".) Each licensee is addressed as "you". Activities other than copying, distribution and modification are not covered by this License; they are outside its scope. The act of running the Program is not restricted, and the output from the Program is covered only if its contents constitute a work based on the Program (independent of having been made by running the Program). Whether that is true depends on what the Program does. 1. You may copy and distribute verbatim copies of the Program's source code as you receive it, in any medium, provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice and disclaimer of warranty; keep intact all the notices that refer to this License and to the absence of any warranty; and give any other recipients of the Program a copy of this License along with the Program. You may charge a fee for the physical act of transferring a copy, and you may at your option offer warranty protection in exchange for a fee. 2. You may modify your copy or copies of the Program or any portion of it, thus forming a work based on the Program, and copy and distribute such modifications or work under the terms of Section 1 above, provided that you also meet all of these conditions: a) You must cause the modified files to carry prominent notices stating that you changed the files and the date of any change. b) You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License. c) If the modified program normally reads commands interactively when run, you must cause it, when started running for such interactive use in the most ordinary way, to print or display an announcement including an appropriate copyright notice and a notice that there is no warranty (or else, saying that you provide a warranty) and that users may redistribute the program under these conditions, and telling the user how to view a copy of this License. (Exception: if the Program itself is interactive but does not normally print such an announcement, your work based on the Program is not required to print an announcement.) These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it. Thus, it is not the intent of this section to claim rights or contest your rights to work written entirely by you; rather, the intent is to exercise the right to control the distribution of derivative or collective works based on the Program. In addition, mere aggregation of another work not based on the Program with the Program (or with a work based on the Program) on a volume of a storage or distribution medium does not bring the other work under the scope of this License. 3. You may copy and distribute the Program (or a work based on it, under Section 2) in object code or executable form under the terms of Sections 1 and 2 above provided that you also do one of the following: a) Accompany it with the complete corresponding machine-readable source code, which must be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or, b) Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or, c) Accompany it with the information you received as to the offer to distribute corresponding source code. (This alternative is allowed only for noncommercial distribution and only if you received the program in object code or executable form with such an offer, in accord with Subsection b above.) The source code for a work means the preferred form of the work for making modifications to it. For an executable work, complete source code means all the source code for all modules it contains, plus any associated interface definition files, plus the scripts used to control compilation and installation of the executable. However, as a special exception, the source code distributed need not include anything that is normally distributed (in either source or binary form) with the major components (compiler, kernel, and so on) of the operating system on which the executable runs, unless that component itself accompanies the executable. If distribution of executable or object code is made by offering access to copy from a designated place, then offering equivalent access to copy the source code from the same place counts as distribution of the source code, even though third parties are not compelled to copy the source along with the object code. 4. You may not copy, modify, sublicense, or distribute the Program except as expressly provided under this License. Any attempt otherwise to copy, modify, sublicense or distribute the Program is void, and will automatically terminate your rights under this License. However, parties who have received copies, or rights, from you under this License will not have their licenses terminated so long as such parties remain in full compliance. 5. You are not required to accept this License, since you have not signed it. However, nothing else grants you permission to modify or distribute the Program or its derivative works. These actions are prohibited by law if you do not accept this License. Therefore, by modifying or distributing the Program (or any work based on the Program), you indicate your acceptance of this License to do so, and all its terms and conditions for copying, distributing or modifying the Program or works based on it. 6. Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions. You may not impose any further restrictions on the recipients' exercise of the rights granted herein. You are not responsible for enforcing compliance by third parties to this License. 7. If, as a consequence of a court judgment or allegation of patent infringement or for any other reason (not limited to patent issues), conditions are imposed on you (whether by court order, agreement or otherwise) that contradict the conditions of this License, they do not excuse you from the conditions of this License. If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Program at all. For example, if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you, then the only way you could satisfy both it and this License would be to refrain entirely from distribution of the Program. If any portion of this section is held invalid or unenforceable under any particular circumstance, the balance of the section is intended to apply and the section as a whole is intended to apply in other circumstances. It is not the purpose of this section to induce you to infringe any patents or other property right claims or to contest validity of any such claims; this section has the sole purpose of protecting the integrity of the free software distribution system, which is implemented by public license practices. Many people have made generous contributions to the wide range of software distributed through that system in reliance on consistent application of that system; it is up to the author/donor to decide if he or she is willing to distribute software through any other system and a licensee cannot impose that choice. This section is intended to make thoroughly clear what is believed to be a consequence of the rest of this License. 8. If the distribution and/or use of the Program is restricted in certain countries either by patents or by copyrighted interfaces, the original copyright holder who places the Program under this License may add an explicit geographical distribution limitation excluding those countries, so that distribution is permitted only in or among countries not thus excluded. In such case, this License incorporates the limitation as if written in the body of this License. 9. The Free Software Foundation may publish revised and/or new versions of the General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns. Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and "any later version", you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation. If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation. 10. If you wish to incorporate parts of the Program into other free programs whose distribution conditions are different, write to the author to ask for permission. For software which is copyrighted by the Free Software Foundation, write to the Free Software Foundation; we sometimes make exceptions for this. Our decision will be guided by the two goals of preserving the free status of all derivatives of our free software and of promoting the sharing and reuse of software generally. NO WARRANTY 11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION. 12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. END OF TERMS AND CONDITIONS How to Apply These Terms to Your New Programs If you develop a new program, and you want it to be of the greatest possible use to the public, the best way to achieve this is to make it free software which everyone can redistribute and change under these terms. To do so, attach the following notices to the program. It is safest to attach them to the start of each source file to most effectively convey the exclusion of warranty; and each file should have at least the "copyright" line and a pointer to where the full notice is found. Copyright (C) This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA Also add information on how to contact you by electronic and paper mail. If the program is interactive, make it output a short notice like this when it starts in an interactive mode: Gnomovision version 69, Copyright (C) year name of author Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'. This is free software, and you are welcome to redistribute it under certain conditions; type `show c' for details. The hypothetical commands `show w' and `show c' should show the appropriate parts of the General Public License. Of course, the commands you use may be called something other than `show w' and `show c'; they could even be mouse-clicks or menu items--whatever suits your program. You should also get your employer (if you work as a programmer) or your school, if any, to sign a "copyright disclaimer" for the program, if necessary. Here is a sample; alter the names: Yoyodyne, Inc., hereby disclaims all copyright interest in the program `Gnomovision' (which makes passes at compilers) written by James Hacker. , 1 April 1989 Ty Coon, President of Vice This General Public License does not permit incorporating your program into proprietary programs. If your program is a subroutine library, you may consider it more useful to permit linking proprietary applications with the library. If this is what you want to do, use the GNU Library General Public License instead of this License. pcaMethods/DESCRIPTION0000644000175200017520000000445714710312421015376 0ustar00biocbuildbiocbuildPackage: pcaMethods Maintainer: Henning Redestig License: GPL (>= 3) Title: A collection of PCA methods LinkingTo: Rcpp LazyLoad: Yes Author: Wolfram Stacklies, Henning Redestig, Kevin Wright Authors@R: c(person(given="Wolfram", family="Stacklies", role=c("aut")), person(given="Henning", family="Redestig", email="henning.red@gmail.com", comment=c(ORCID="0000-0003-2130-9288"), role=c("aut","cre")), person(given="Kevin", family="Wright", comment=c(ORCID="0000-0002-0617-8673"), role="aut")) SystemRequirements: Rcpp Description: Provides Bayesian PCA, Probabilistic PCA, Nipals PCA, Inverse Non-Linear PCA and the conventional SVD PCA. A cluster based method for missing value estimation is included for comparison. BPCA, PPCA and NipalsPCA may be used to perform PCA on incomplete data as well as for accurate missing value estimation. A set of methods for printing and plotting the results is also provided. All PCA methods make use of the same data structure (pcaRes) to provide a common interface to the PCA results. Initiated at the Max-Planck Institute for Molecular Plant Physiology, Golm, Germany. Version: 1.98.0 URL: https://github.com/hredestig/pcamethods BugReports: https://github.com/hredestig/pcamethods/issues Encoding: UTF-8 Depends: Biobase, methods Imports: BiocGenerics, Rcpp (>= 0.11.3), MASS Suggests: matrixStats, lattice, ggplot2 Collate: 'derrorHierarchic.R' 'errorHierarchic.R' 'AllClasses.R' 'AllGenerics.R' 'BPCA_dostep.R' 'BPCA_initmodel.R' 'bpca.R' 'checkData.R' 'forkNlpcaNet.R' 'kEstimate.R' 'kEstimateFast.R' 'lineSearch.R' 'llsImpute.R' 'methods-ExpressionSet.R' 'methods-nniRes.R' 'methods-pcaRes.R' 'nipalsPca.R' 'nlpca.R' 'optiAlgCgd.R' 'orth.R' 'pca.R' 'pcaMethods-package.R' 'ppca.R' 'prep.R' 'repmat.R' 'robustPca.R' 'sortFeatures.R' 'svdImpute.R' 'vector2matrices.R' 'xval.R' Packaged: 2024-10-30 02:08:17 UTC; biocbuild biocViews: Bayesian RoxygenNote: 6.1.1 git_url: https://git.bioconductor.org/packages/pcaMethods git_branch: RELEASE_3_20 git_last_commit: 19b67d5 git_last_commit_date: 2024-10-29 Repository: Bioconductor 3.20 Date/Publication: 2024-10-29 NeedsCompilation: yes pcaMethods/MD50000644000175200017520000001575014710312421014176 0ustar00biocbuildbiocbuild94d55d512a9ba36caa9b7df079bae19f *COPYING f2ac2df24f180105baec51634df1bedd *DESCRIPTION 53ce21062eec00c7ac30ea834b3aee36 *NAMESPACE 2708cf17b2ee5018ba4013b72cc964cc *R/AllClasses.R 7f73df2dc3860da3929060be7c2164a1 *R/AllGenerics.R 3eaba9f52b5a6bd671dae7fe0a047a81 *R/BPCA_dostep.R cc6e82ca2e70d1dd963b36c04f139813 *R/BPCA_initmodel.R 7bdcfd79a58bb957fc2317200116d011 *R/bpca.R 74b878b140ac4f74c6e51ed01783f098 *R/checkData.R a2281bff4b4e90590b9ab6345c4c5241 *R/derrorHierarchic.R 26b181d3cda1f691dcf5a1ec7132b3e0 *R/errorHierarchic.R e632dcc95e3233b756a2d87becb824e9 *R/forkNlpcaNet.R a14e6635cd95fbd5836231b59838d039 *R/kEstimate.R 3a1e7f50f4fafa9736c2e45d94bf3381 *R/kEstimateFast.R ee2a6a4892b3e843ac1ed4ad09b3fd4f *R/lineSearch.R 1d89b891b08341fb449e097d61867b0c *R/llsImpute.R ed55566abc804fc6ab5e7caa5758ac52 *R/methods-ExpressionSet.R ac332abe1c596273687f247945d47c47 *R/methods-nniRes.R a6d847aaf6136e4ebd8be97d8937242d *R/methods-pcaRes.R c1af456f102f069f1324af18c79098bc *R/nipalsPca.R a6593f4d471a821017f4fcc8584e101e *R/nlpca.R d652199d1f27f0120332a2e5212eb51b *R/optiAlgCgd.R 61c314d080cb0df7ad0d7e429acb17ed *R/orth.R 6541d85885e9c9da309db508abcc23a8 *R/pca.R cae83d443e82efecefbd134e524ddfdb *R/pcaMethods-package.R d166d2fe17520c3cb365b33f359326f3 *R/ppca.R 49026eb29f7c9631e11905bba0544b93 *R/prep.R 40f437c3a59bb777aa67e595b7e82aa2 *R/repmat.R b9e049c0dbd1a90acfff8eeef11a97a9 *R/robustPca.R 8e8da524394917159dc5d5361d260a2a *R/sortFeatures.R 1208b9d05c9864ab9c3d8b62d625d316 *R/svdImpute.R d4701286a3a40c06f2fc9f3dd63a1336 *R/vector2matrices.R ab79a52a829939be97ff127ef31bac10 *R/xval.R 60ef7e1f4a90c81152f1faf53b13c63b *README.md 0c3ec6f9f0f2abc6726e947c6444bce6 *build/vignette.rds 1db4664839bb67f16355796fceacc465 *data/helix.RData 4cfc94cebc1f6c1c0b2fe22597cbe9ad *data/metaboliteData.RData 92a0c0b491e446ef83685176756c9239 *data/metaboliteDataComplete.RData f2ccf3d4bfd491d62f4f8ff68d2344e2 *inst/CITATION 60b40c6b5f239b430d681182884a8e42 *inst/doc/missingValues.R cd2ac72edc90b082feaa2412edb98d5a *inst/doc/missingValues.Rnw 95e3b97f955a9db26f1b79761ce27053 *inst/doc/missingValues.pdf a615f67a838f0d0440c619c69292afc6 *inst/doc/outliers.R cff153d5a7c7a15b9793dc40d61ec393 *inst/doc/outliers.Rnw e93f20baaf5629ab395e89292f891af0 *inst/doc/outliers.pdf 09d46c930268562f5f147468cbfaf060 *inst/doc/pcaMethods.R 53d2dd38f864f4496afc03ed629d5147 *inst/doc/pcaMethods.Rnw 1c9dff46d3dec2b3ef7ff7ed7dc4cda6 *inst/doc/pcaMethods.pdf fbf104068e163dc309e70b531bf28669 *man/BPCA_dostep.Rd 5ee917e5f7477cbf1cd3d00b1e3c8494 *man/BPCA_initmodel.Rd e61a0b81e19ff4f457aa37a073921240 *man/DModX-pcaRes-method.Rd fd6077b0c7fd1702048446af10776095 *man/Q2.Rd d28eef0e62c8e2b49a718d0191bb2a18 *man/R2VX-pcaRes-method.Rd fbafb65b6f3bcdb3d2c4a9d46d106dc8 *man/R2cum-pcaRes-method.Rd 9a9e5765fe09f86718a5686f5e31e836 *man/RnipalsPca.Rd 379cad260ea1253423b9a2d947be91ee *man/asExprSet.Rd 2f02f0f86d30097ffd6892f218305eba *man/biplot-methods.Rd 363ab2dd40d0ced60d7b997e74334a8e *man/bpca.Rd d0c31541f099585d51fdefd425f687c2 *man/center-pcaRes-method.Rd d1d93ac69074e86b8edfc01fb4539d9b *man/centered-pcaRes-method.Rd dee0e11f7565fa72baab68d313f94c7c *man/checkData.Rd a7a186e7b97f4f88f7cc7fa00a920b5c *man/completeObs-nniRes-method.Rd eb92f044a1ccc405bea537ee1375dffd *man/cvseg.Rd 181df4e1eb378c09f05d4064105ec614 *man/cvstat-pcaRes-method.Rd 716b33aeaa0df62e8680373ae37856bf *man/deletediagonals.Rd b9f12919f36f070cad62288f0cfe580d *man/derrorHierarchic.Rd 5e8a94f8d00287e2a3233a99f058a6d5 *man/dim.pcaRes.Rd e711282bb4132a42617220d28453f066 *man/errorHierarchic.Rd 1480a51b55068d1d43bd8ceaee7ff321 *man/fitted-methods.Rd 13fbf18fe74e2200be4924016f88d00d *man/forkNlpcaNet.Rd d300547edd993fb7c5968578c946bfe6 *man/getHierarchicIdx.Rd 003d8cf7c7618226d791aa316b01af4b *man/helix.Rd 7add4c9d6a52339d899d4d5c1dc7509e *man/kEstimate.Rd 5f4ec1afa6ad30fab9b4f0baa0df6081 *man/kEstimateFast.Rd e2d920339298ef50b84009d984a9b3b0 *man/leverage-pcaRes-method.Rd f01710fb7a3f8ac99ba270d6df2d61c8 *man/lineSearch.Rd 35a544855f2a637a9f5ffaafbfc2e6de *man/linr.Rd 07f865580d383ce508717ad6379f7072 *man/listPcaMethods.Rd ab8fbc9974b1f3dbbf37d7848f3f29cf *man/llsImpute.Rd b24c49ba4d9912ece2f3a961706c2366 *man/loadings-ANY-method.Rd ddf6de7c2a05b343eaf3dacab115190c *man/loadings-pcaRes-method.Rd 81fdb14fee2ae7768d532642f94f0271 *man/loadings.pcaRes.Rd 30478a240d34fdff44ac7ef5e1f9fbaf *man/metaboliteData.Rd dfad283bb21ffedda98fea0e04bea69a *man/metaboliteDataComplete.Rd 6f3091f579b66a61c1c5707ca54aed9c *man/method-pcaRes-method.Rd 3fc29289d16f3cf62adef9b36f31b42d *man/nObs-pcaRes-method.Rd f809f089c43dd1259baa4bfec9a61341 *man/nP-pcaRes-method.Rd 21b62e613167fc9a458cff800fd3d11c *man/nPcs-pcaRes-method.Rd 7f7915d84795ce2cbf4142247a378c61 *man/nVar-pcaRes-method.Rd 702316eaf30b9e2132f10c819a55cd06 *man/nipalsPca.Rd 0ed131e99ae3a5ff4de1d1f04876de81 *man/nlpca.Rd 27f2a7ea8d1543ac7ab56cf09c498886 *man/nmissing-pcaRes-method.Rd c0f13dc78997b0475f8270483e43bb9b *man/nni.Rd aab3be94bfa858f5578486529ddd0cbe *man/nniRes.Rd c189a8fdbae3b8b20d8b5575e249e307 *man/optiAlgCgd.Rd d4faad4116e81649e6b1d001c0943a58 *man/orth.Rd bc0afb9eec7861283199b9edd0bc95af *man/pca.Rd 9c7c8fee79d7a921d252ed0799671b8a *man/pcaMethods-deprecated.Rd a6f3ce318b18e597f86e2aa94adbac57 *man/pcaMethods.Rd 6ad7ee2d19b51ab21b6737484bdf1395 *man/pcaNet.Rd 32e40773afc9084cf5ad3e8e480679a9 *man/pcaRes.Rd f343c3c1113908a1297c8132f55be309 *man/plot.pcaRes.Rd 0fade2faec56ae9e6457a64920aa7b31 *man/plotPcs.Rd efaaa42bdd7a9f2d998f135425c1a14e *man/ppca.Rd 19660ae127a290127872f03dd8d363f5 *man/predict-methods.Rd cb2143d9734cc6e0a9bb8952580e5338 *man/prep.Rd 946ec312bd2c3671e9e26da6438ec8d2 *man/rediduals-methods.Rd 6fe738c58d27eacc07bd9ccb20e74015 *man/repmat.Rd 018ed24ade7b5f6dd6e6cb23e577f9e7 *man/robustPca.Rd f48dcb6c1f47865c6d77e2acc901854d *man/robustSvd.Rd 7bda9f57f6fd5f441d56dee9b6edfd5a *man/sDev-pcaRes-method.Rd d7cba51a0dbc7065afbcfcf0e5ad131e *man/scaled-pcaRes-method.Rd f8168ef06fbd7fcdeae3cecce9949cf4 *man/scl-pcaRes-method.Rd 435b6b831310e9f38517299e36abbe84 *man/scores-pcaRes-method.Rd cc6b256cfe302bb7cb4229111b8d583b *man/scores.pcaRes.Rd 92f3b6fb4681790b612f154b20e29e85 *man/show-methods.Rd e13fbbe877d76a6e4ab8859899557d4b *man/showNniRes.Rd da6fd6b66ef78e885e00a7a029c1073e *man/simpleEllipse.Rd 1fa75fbc0968f32355c18807e7b84f20 *man/slplot-pcaRes-method.Rd 2d37befcf418b2456b393a8a56339dad *man/sortFeatures.Rd f0518717d0e26007af298b05242fc7d6 *man/summary.Rd 50e5bffe335aa551e56af7506b689aa2 *man/svdImpute.Rd c13a64edf435d05e0a823f78c286ec07 *man/svdPca.Rd 665d2d1f27194d4223184182744dc145 *man/tempFixNas.Rd 7247bc86e24eae698bb8ea5d8faf0517 *man/vector2matrices-matrix-method.Rd 61ef3da94611444d00729d3ac2802771 *man/vector2matrices-nlpcaNet-method.Rd 0602263dde655042aaa20afe78c45686 *man/wasna-pcaRes-method.Rd b170b367f2786de7e1743ac7264e6ab6 *man/weightsAccount.Rd 9a742a3c7f3b0d45051fccfd148ef446 *src/RcppExports.cpp f9524c162711228f3a87c702919bc0b7 *src/nipals.cpp cd2ac72edc90b082feaa2412edb98d5a *vignettes/missingValues.Rnw cff153d5a7c7a15b9793dc40d61ec393 *vignettes/outliers.Rnw 53d2dd38f864f4496afc03ed629d5147 *vignettes/pcaMethods.Rnw pcaMethods/NAMESPACE0000644000175200017520000000236214710217306015106 0ustar00biocbuildbiocbuild# Generated by roxygen2: do not edit by hand S3method(biplot,pcaRes) S3method(dim,pcaRes) S3method(fitted,pcaRes) S3method(loadings,pcaRes) S3method(plot,pcaRes) S3method(predict,pcaRes) S3method(residuals,pcaRes) S3method(scores,pcaRes) S3method(summary,pcaRes) export(Q2) export(RnipalsPca) export(asExprSet) export(bpca) export(checkData) export(cvseg) export(kEstimate) export(kEstimateFast) export(listPcaMethods) export(llsImpute) export(nipalsPca) export(nlpca) export(nni) export(pca) export(plotPcs) export(ppca) export(prep) export(robustPca) export(robustSvd) export(showNniRes) export(showPcaRes) export(svdImpute) export(svdPca) exportClasses(nlpcaNet) exportClasses(nniRes) exportClasses(pcaRes) exportMethods(DModX) exportMethods(R2VX) exportMethods(R2cum) exportMethods(center) exportMethods(centered) exportMethods(completeObs) exportMethods(cvstat) exportMethods(leverage) exportMethods(loadings) exportMethods(method) exportMethods(nObs) exportMethods(nP) exportMethods(nPcs) exportMethods(nVar) exportMethods(nmissing) exportMethods(sDev) exportMethods(scaled) exportMethods(scl) exportMethods(scores) exportMethods(slplot) exportMethods(wasna) import(Biobase) import(BiocGenerics) import(methods) importFrom(Rcpp,evalCpp) useDynLib(pcaMethods) pcaMethods/R/0000755000175200017520000000000014710217306014065 5ustar00biocbuildbiocbuildpcaMethods/R/AllClasses.R0000644000175200017520000003074414710217306016246 0ustar00biocbuildbiocbuild##' @include errorHierarchic.R ##' @include derrorHierarchic.R NULL ##' This is a class representation of a non-linear PCA neural ##' network. The \code{nlpcaNet} class is not meant for user-level ##' usage. ##' ##' Creating Objects ##' ##' \code{new("nlpcaNet", net=[the network structure], ##' hierarchic=[hierarchic design], ##' fct=[the functions at each layer], fkt=[the functions used for ##' forward propagation], weightDecay=[incremental decrease of weight ##' changes over iterations (between 0 and 1)], featureSorting=[sort ##' features or not], dataDist=[represents the present values], ##' inverse=[net is inverse mode or not], fCount=[amount of times ##' features were sorted], componentLayer=[which layer is the ##' 'bottleneck' (principal components)], ##' erro=[the used error function], gradient=[the used gradient method], ##' weights=[the present weights], ##' maxIter=[the amount of iterations that was done], scalingFactor=[the ##' scale of the original matrix])} ##' ##' Slots ##' ##' \describe{ ##' \item{net}{"matrix", matrix showing the representation of the ##' neural network, e.g. (2,4,6) for a network with two features, a ##' hidden layer and six output neurons (original variables).} ##' \item{hierarchic}{"list", the hierarchic design of the network, ##' holds 'idx' (), 'var' () and layer (which layer is the principal ##' component layer).} ##' \item{fct}{"character", a vector naming the functions that will be ##' applied on each layer. "linr" is linear (i.e.) standard matrix ##' products and "tanh" means that the arcus tangens is applied on the ##' result of the matrix product (for non-linearity).} ##' \item{fkt}{"character", same as fct but the functions used during ##' back propagation.} ##' \item{weightDecay}{"numeric", the value that is used to ##' incrementally decrease the weight changes to ensure convergence.} ##' \item{featureSorting}{"logical", indicates if features will be ##' sorted or not. This is used to make the NLPCA assume properties ##' closer to those of standard PCA were the first component is more ##' important for reconstructing the data than the second component.} ##' \item{dataDist}{"matrix", a matrix of ones and zeroes indicating ##' which values will add to the errror.} ##' \item{inverse}{"logical", network is inverse mode (currently only ##' inverse is supported) or not. Eg. the case when we have truly ##' missing values and wish to impute them.} ##' \item{fCount}{"integer", Counter for the amount of times features ##' were really sorted.} ##' \item{componentLayer}{"numeric", the index of 'net' that is the ##' component layer.} ##' \item{error}{"function", the used error function. Currently only one ##' is provided \code{errorHierarchic}.} ##' \item{gradient}{"function", the used gradient function. Currently ##' only one is provided \code{derrorHierarchic}} ##' \item{weights}{"list", A list holding managements of the ##' weights. The list has two functions, weights$current() and ##' weights$set() which access a matrix in the local environment of ##' this object.} ##' \item{maxIter}{"integer", the amount of iterations used to train ##' this network.} ##' \item{scalingFactor}{"numeric", training the network is best made ##' with 'small' values so the original data is scaled down to a ##' suitable range by division with this number.}} ##' ##' Methods ##' ##' \describe{ \item{vector2matrices}{Returns the ##' weights in a matrix representation.} } ##' @title Class representation of the NLPCA neural net ##' @docType class ##' @aliases nlpcaNet nlpcaNet-class ##' @seealso \code{\link{nlpca}} ##' @aliases nFit nFit-class ##' @exportClass nlpcaNet ##' @keywords classes ##' @name pcaNet ##' @author Henning Redestig setClass("nlpcaNet", representation(net="matrix", hierarchic="list", fct="character", fkt="character", weightDecay="numeric", featureSorting="logical", dataDist="matrix", inverse="logical", fCount="integer", componentLayer="integer", error="function", gradient="function", weights="list", maxIter="integer", scalingFactor="numeric"), prototype(net=rbind(c(4,6,2,6,4)), hierarchic=list(var=rbind(c(1,1,0.01)), layer=3, idx=rbind(c(1,1,0),c(0,1,1))), fct=c("linr", "tanh", "linr", "tanh", "linr"), fkt=c("tanh", "linr", "tanh", "linr"), weightDecay=0.001, featureSorting=TRUE, inverse=FALSE, dataDist=NULL, fCount=as.integer(0), componentLayer=as.integer(3), error=errorHierarchic, gradient=derrorHierarchic, weights=NULL, maxIter=as.integer(1200), scalingFactor=NULL)) setAs("NULL", "nlpcaNet", function(from, to){ new(to) }) ##' This is a class representation of a PCA result ##' ##' \bold{Creating Objects}\cr ##' \code{new("pcaRes", scores=[the scores], loadings=[the loadings], ##' nPcs=[amount of PCs], R2cum=[cumulative R2], nObs=[amount of ##' observations], nVar=[amount of variables], R2=[R2 for each ##' individual PC], sDev=[stdev for each individual PC], ##' centered=[was data centered], center=[original means], ##' varLimit=[what variance limit was exceeded], method=[method used to ##' calculate PCA], missing=[amount of NAs], ##' completeObs=[estimated complete observations])} ##' ##' \bold{Slots}\cr ##' \describe{ ##' \item{scores}{"matrix", the calculated scores} ##' \item{loadings}{"matrix", the calculated loadings} ##' \item{R2cum}{"numeric", the cumulative R2 values} ##' \item{sDev}{"numeric", the individual standard ##' deviations of the score vectors} ##' \item{R2}{"numeric", the individual R2 values} ##' \item{cvstat}{"numeric", cross-validation statistics} ##' \item{nObs}{"numeric", number of observations} ##' \item{nVar}{"numeric", number of variables} ##' \item{centered}{"logical", data was centered or not} ##' \item{center}{"numeric", the original variable centers} ##' \item{scaled}{"logical", data was scaled or not} ##' \item{scl}{"numeric", the original variable scales} ##' \item{varLimit}{"numeric", the exceeded variance limit} ##' \item{nPcs,nP}{"numeric", the number of calculated PCs} ##' \item{method}{"character", the method used to perform PCA} ##' \item{missing}{"numeric", the total amount of missing values in ##' original data} ##' \item{completeObs}{"matrix", the estimated complete observations} ##' \item{network}{"nlpcaNet", the network used by non-linear PCA} ##' } ##' ##' \bold{Methods (not necessarily exhaustive)}\cr ##' \describe{ ##' \item{print}{Print function} ##' \item{summary}{Extract information about PC relevance} ##' \item{screeplot}{Plot a barplot of standard deviations for PCs} ##' \item{slplot}{Make a side by side score and loadings plot} ##' \item{nPcs}{Get the number of PCs} ##' \item{nObs}{Get the number of observations} ##' \item{cvstat}{Cross-validation statistics} ##' \item{nVar}{Get the number of variables} ##' \item{loadings}{Get the loadings} ##' \item{scores}{Get the scores} ##' \item{dim}{Get the dimensions (number of observations, number of ##' features)} ##' \item{centered}{Get a logical indicating if centering was done as ##' part of the model} ##' \item{center}{Get the averages of the original variables.} ##' \item{completeObs}{Get the imputed data set} ##' \item{method}{Get a string naming the used PCA method} ##' \item{sDev}{Get the standard deviations of the PCs} ##' \item{scaled}{Get a logical indicating if scaling was done as ##' part of the model} ##' \item{scl}{Get the scales of the original variablesb} ##' \item{R2cum}{Get the cumulative R2} ##' } ##' @title Class for representing a PCA result ##' @keywords classes ##' @exportClass pcaRes ##' @docType class ##' @name pcaRes ##' @aliases pcaRes pcaRes-class ##' @author Henning Redestig setClass("pcaRes", representation(completeObs="matrix", scores="matrix", loadings="matrix", R2cum="numeric", R2="numeric", # ditch, get from R2cum cvstat="numeric", # ditch, get from R2cum sDev="numeric", # ditch, get from scores nObs="numeric", # ditch, get from scores nVar="numeric", centered="logical", center="numeric", subset="numeric", scaled="character", scale="numeric", varLimit="numeric", # ditch, useless nPcs="numeric", # ditch, get from scores method="character", missing="matrix", network="nlpcaNet"), prototype(completeObs=NULL, scores=NULL, loadings=NULL, R2cum=NULL, R2=NULL, subset=NULL, cvstat=NULL, sDev=NULL, nObs=NULL, nVar=NULL, centered=NULL, center=NULL, scaled=NULL, scale=NULL, varLimit=NULL, nPcs=NULL, method=NULL, missing=NULL, network=NULL)) setAs("NULL", "pcaRes", function(from, to){ new(to) }) ##' This is a class representation of nearest neighbour imputation ##' (nni) result ##' ##' \bold{Creating Objects}\cr ##' \code{new("nniRes", completeObs=[the estimated complete ##' observations], k=[cluster size], nObs=[amount of observations], ##' nVar=[amount of variables], centered=[was the data centered befor ##' running LLSimpute], center=[original means], method=[method used ##' to perform clustering], missing=[amount of NAs])} ##' ##' \bold{Slots}\cr ##' \describe{ ##' \item{completeObs}{"matrix", the estimated complete observations} ##' \item{nObs}{"numeric", amount of observations} ##' \item{nVar}{"numeric", amount of variables} ##' \item{correlation}{"character", the correlation method used ##' (pearson, kendall or spearman)} ##' \item{centered}{"logical", data was centered or not} ##' \item{center}{"numeric", the original variable centers} ##' \item{k}{"numeric", cluster size} ##' \item{method}{"character", the method used to perform the clustering} ##' \item{missing}{"numeric", the total amount of missing values in ##' original data} ##' } ##' ##' \bold{Methods}\cr ##' \describe{ \item{print}{Print function} } ##' @title Class for representing a nearest neighbour imputation result ##' @docType class ##' @exportClass nniRes ##' @name nniRes ##' @keywords classes ##' @aliases nniRes nniRes-class ##' @author Wolfram Stacklies setClass("nniRes", representation(completeObs="matrix", nObs="numeric", nVar="numeric", centered="logical", center="numeric", k="numeric", method="character", correlation="character", missing="numeric"), prototype(completeObs=NULL, nObs=NULL, nVar=NULL, centered=NULL, center=NULL, k=NULL, method=NULL, correlation=NULL, missing=NULL)) setAs("NULL", "nniRes", function(from, to) { new(to) }) ##' Create an object that holds the weights for nlpcaNet. Holds and ##' sets weights in using an environment object. ##' @param w \code{matrix} -- New weights ##' @return A weightsAccound with \code{set} and \code{current} ##' functions. ##' @author Henning Redestig weightsAccount <- function(w) { list( set = function(newWeights) { if(!inherits(newWeights, "matrix")) stop("The weights must inherit from matrix") w <<- newWeights }, current = function() { w } ) } pcaMethods/R/AllGenerics.R0000644000175200017520000000456514710217306016412 0ustar00biocbuildbiocbuildsetGeneric("vector2matrices", function(object, ...) standardGeneric("vector2matrices")) ##' @exportMethod leverage setGeneric("leverage", function(object, ...) standardGeneric("leverage")) ##' @exportMethod DModX setGeneric("DModX", function(object, dat, newdata=FALSE, type=c("normalized","absolute"), ...) standardGeneric("DModX")) ##' @exportMethod nP setGeneric("nP", function(object, ...) standardGeneric("nP")) ##' @exportMethod cvstat setGeneric("cvstat", function(object, ...) standardGeneric("cvstat")) ##' @exportMethod nPcs setGeneric("nPcs", function(object, ...) standardGeneric("nPcs")) ##' @exportMethod nObs setGeneric("nObs", function(object, ...) standardGeneric("nObs")) ##' @exportMethod nVar setGeneric("nVar", function(object, ...) standardGeneric("nVar")) ##' @exportMethod centered setGeneric("centered", function(object, ...) standardGeneric("centered")) ##' @exportMethod center setGeneric("center", function(object, ...) standardGeneric("center")) ##' @exportMethod completeObs setGeneric("completeObs", function(object, ...) standardGeneric("completeObs")) ##' @exportMethod method setGeneric("method", function(object, ...) standardGeneric("method")) ##' @exportMethod nmissing setGeneric("nmissing", function(object, ...) standardGeneric("nmissing")) ##' @exportMethod wasna setGeneric("wasna", function(object, ...) standardGeneric("wasna")) ##' @exportMethod sDev setGeneric("sDev", function(object, ...) standardGeneric("sDev")) ##' @exportMethod scaled setGeneric("scaled", function(object, ...) standardGeneric("scaled")) ##' @exportMethod scl setGeneric("scl", function(object, ...) standardGeneric("scl")) ##' @exportMethod R2cum setGeneric("R2cum", function(object, ...) standardGeneric("R2cum")) ##' @exportMethod slplot setGeneric("slplot", function(object, pcs=c(1,2), scoresLoadings=c(TRUE, TRUE), sl="def", ll="def", hotelling=0.95, rug=TRUE, sub=NULL,...) standardGeneric("slplot")) ##' @exportMethod scores setGeneric("scores", function(object, ...) standardGeneric("scores")) ##' @exportMethod loadings setGeneric("loadings", function(object, ...) standardGeneric("loadings")) ##' @exportMethod R2VX setGeneric("R2VX", function(object, ...) standardGeneric("R2VX")) ## @exportMethod prep #setGeneric("prep", function(object, ...) standardGeneric("prep")) pcaMethods/R/BPCA_dostep.R0000644000175200017520000000543714710217306016304 0ustar00biocbuildbiocbuild##' The function contains the actual implementation of the BPCA ##' component estimation. It performs one step of the BPCA EM ##' algorithm. It is called 'maxStep' times from within the main loop ##' in BPCAestimate. ##' ##' This function is NOT intended to be run standalone. ##' @title Do BPCA estimation step ##' @param M Data structure containing all needed information. See the ##' source documentation of BPCA_initmodel for details ##' @param y Numeric original data matrix ##' @return Updated version of the data structure ##' @author Wolfram Stacklies BPCA_dostep <- function(M,y) { ## Empty matrix in which the scores are copied M$scores <- matrix(NA, M$rows, M$comps) ## Expectation step for data without missing values Rx <- diag(M$comps) + M$tau * t(M$PA) %*% M$PA + M$SigW Rxinv <- solve(Rx) idx <- M$row_nomiss if (length(idx) == 0) { trS <- 0 T <- 0 } else { dy <- y[idx,, drop=FALSE] - repmat(M$mean, length(idx), 1) x <- M$tau * Rxinv %*% t(M$PA) %*% t(dy) T <- t(dy) %*% t(x) trS <- sum(sum(dy * dy)) ## Assign the scores for complete rows xTranspose <- t(x) for (i in 1:length(idx)) { M$scores[idx[i],] <- xTranspose[i,] } } ## Expectation step for incomplete data if( length(M$row_miss) > 0) { for(n in 1:length(M$row_miss)) { i <- M$row_miss[n] dyo <- y[ i, !M$nans[i,], drop=FALSE] - M$mean[ !M$nans[i,], drop=FALSE] Wm <- M$PA[ M$nans[i,],, drop=FALSE] Wo <- M$PA[ !M$nans[i,],, drop=FALSE] Rxinv <- solve( (Rx - M$tau * t(Wm) %*% Wm)) ex <- M$tau * t(Wo) %*% t(dyo) x <- Rxinv %*% ex dym <- Wm %*% x dy <- y[i,, drop=FALSE] dy[ !M$nans[i,] ] <- t(dyo) dy[ M$nans[i,] ] <- t(dym) M$yest[i,] <- dy + M$mean T <- T + t(dy) %*% t(x) T[ M$nans[i,], ] <- T[ M$nans[i,],, drop=FALSE] + Wm %*% Rxinv trS <- trS + dy %*% t(dy) + sum(M$nans[i,]) / M$tau + sum( diag(Wm %*% Rxinv %*% t(Wm)) ) trS <- trS[1,1] ## Assign the scores for rows containing missing values M$scores[M$row_miss[n],] <- t(x) } } T <- T / M$rows trS <- trS / M$rows ## Maximation step Rxinv <- solve(Rx) Dw <- Rxinv + M$tau * t(T) %*% M$PA %*% Rxinv + diag(M$alpha, nrow = length(M$alpha)) / M$rows Dwinv <- solve(Dw) M$PA <- T %*% Dwinv ## The new estimate of the principal axes (loadings) M$tau <- (M$cols + 2 * M$gtau0 / M$rows) / (trS - sum(diag(t(T) %*% M$PA)) + (M$mean %*% t(M$mean) * M$gmu0 + 2 * M$gtau0 / M$btau0) / M$rows) M$tau <- M$tau[1,1] ## convert to scalar M$SigW <- Dwinv * (M$cols / M$rows) M$alpha <- (2 * M$galpha0 + M$cols) / (M$tau * diag(t(M$PA) %*% M$PA) + diag(M$SigW) + 2 * M$galpha0 / M$balpha0) return(M) } pcaMethods/R/BPCA_initmodel.R0000644000175200017520000000510314710217306016760 0ustar00biocbuildbiocbuild##' Model initialization for Bayesian PCA. This function is NOT ##' inteded to be run separately! ##' ##' The function calculates the initial Eigenvectors by use of SVD ##' from the complete rows. The data structure M is created and ##' initial values are assigned. ##' @title Initialize BPCA model ##' @param y numeric matrix containing missing values. Missing values ##' are denoted as 'NA' ##' @param components Number of components used for estimation ##' @return List containing ##' \item{rows}{Row number of input matrix} ##' \item{cols}{Column number of input matrix} ##' \item{comps}{Number of components to use} ##' \item{yest}{(working variable) current estimate of complete data} ##' \item{row_miss}{(Array) Indizes of rows containing missing values} ##' \item{row_nomiss}{(Array) Indices of complete rows (such with no ##' missing values)} ##' \item{nans}{Matrix of same size as input data. TRUE if \code{input == NA}, ##' false otherwise} ##' \item{mean}{Column wise data mean} ##' \item{PA}{ (d x k) Estimated principal axes (eigenvectors, ##' loadings) The matrix ROWS are the vectors} ##' \item{tau}{Estimated precision of the residual error} ##' \item{scores}{ Estimated scores} ##' Further elements are: galpha0, balpha0, alpha, gmu0, btau0, gtau0, ##' SigW. These are working variables or constants. ##' @author Wolfram Stacklies BPCA_initmodel <- function(y, components) { ## Initialization, write static parameters to the central M <- NULL M$rows <- nrow(y) M$cols <- ncol(y) M$comps <- components ## Column number M$yest <- y ## Original data, NAs are set to 0 later on ## Find rows with missing values, etc... M$nans <- is.na(y) temp <- apply(M$nans, 1, sum) M$row_nomiss <- which(temp == 0) M$row_miss <- which(temp != 0) M$yest[M$nans] <- 0 M$scores <- NULL ## Get the SVD of the complete rows covy <- cov(M$yest) values <- svd(covy, components, components) U <- values[[2]] S <- diag( values[[1]][1:components], nrow = components, ncol = components) V <- values[[3]] ## M$mean: column wise mean of the original data M$mean <- matrix(0, 1, M$cols) for(j in 1:M$cols) { idx <- which(!is.na(y[,j])) M$mean[j] <- mean(y[idx,j]) } M$PA <- U %*% sqrt(S) M$tau <- 1 / ( sum(diag(covy)) - sum(diag(S)) ) ## Constants etc taumax <- 1e10 taumin <- 1e-10 M$tau <- max( min(M$tau, taumax), taumin ) M$galpha0 <- 1e-10 M$balpha0 <- 1 M$alpha <- (2 * M$galpha0 + M$cols) / (M$tau * diag(t(M$PA) %*% M$PA) + 2 * M$galpha0 / M$balpha0) M$gmu0 <- 0.001 M$btau0 <- 1 M$gtau0 <- 1e-10 M$SigW <- diag(components) return(M) } pcaMethods/R/bpca.R0000644000175200017520000001615614710217306015126 0ustar00biocbuildbiocbuild##' Implements a Bayesian PCA missing value estimator. The script ##' is a port of the Matlab version provided by Shigeyuki OBA. See ##' also \url{http://ishiilab.jp/member/oba/tools/BPCAFill.html}. ##' BPCA combines an EM approach for PCA with a Bayesian model. In ##' standard PCA data far from the training set but close to the ##' principal subspace may have the same reconstruction error. BPCA ##' defines a likelihood function such that the likelihood for data ##' far from the training set is much lower, even if they are close to ##' the principal subspace. ##' ##' Scores and loadings obtained with Bayesian PCA slightly differ ##' from those obtained with conventional PCA. This is because BPCA ##' was developed especially for missing value estimation. The ##' algorithm does not force orthogonality between factor loadings, as ##' a result factor loadings are not necessarily orthogonal. However, ##' the BPCA authors found that including an orthogonality criterion ##' made the predictions worse. ##' ##' The authors also state that the difference between real and ##' predicted Eigenvalues becomes larger when the number of ##' observation is smaller, because it reflects the lack of ##' information to accurately determine true factor loadings from the ##' limited and noisy data. As a result, weights of factors to ##' predict missing values are not the same as with conventional PCA, ##' but the missing value estimation is improved. ##' ##' BPCA works iteratively, the complexity is growing with ##' \eqn{O(n^3)}{O(n^3)} because several matrix inversions are ##' required. The size of the matrices to invert depends on the ##' number of components used for re-estimation. ##' ##' Finding the optimal number of components for estimation is not a ##' trivial task; the best choice depends on the internal structure of ##' the data. A method called \code{kEstimate} is provided to ##' estimate the optimal number of components via cross validation. ##' In general few components are sufficient for reasonable estimation ##' accuracy. See also the package documentation for further ##' discussion about on what data PCA-based missing value estimation ##' makes sense. ##' ##' It is not recommended to use this function directely but rather to ##' use the pca() wrapper function. ##' ##' There is a difference with respect the interpretation of rows ##' (observations) and columns (variables) compared to matlab ##' implementation. For estimation of missing values for microarray ##' data, the suggestion in the original bpca is to intepret genes as ##' observations and the samples as variables. In pcaMethods however, ##' genes are interpreted as variables and samples as observations ##' which arguably also is the more natural interpretation. For bpca ##' behavior like in the matlab implementation, simply transpose your ##' input matrix. ##' ##' Details about the probabilistic model underlying BPCA are found in ##' Oba et. al 2003. The algorithm uses an expectation maximation ##' approach together with a Bayesian model to approximate the ##' principal axes (eigenvectors of the covariance matrix in PCA). ##' The estimation is done iteratively, the algorithm terminates if ##' either the maximum number of iterations was reached or if the ##' estimated increase in precision falls below \eqn{1e^{-4}}{1e^-4}. ##' ##' \bold{Complexity:} The relatively high complexity of the method is ##' a result of several matrix inversions required in each step. ##' Considering the case that the maximum number of iteration steps is ##' needed, the approximate complexity is given by the term ##' \deqn{maxSteps \cdot row_{miss} \cdot O(n^3)}{maxSteps * row_miss ##' * O(n^3)} Where \eqn{row_{miss}}{row_miss} is the number of rows ##' containing missing values and \eqn{O(n^3)}{O(n^3)} is the ##' complexity for inverting a matrix of size ##' \eqn{components}{components}. Components is the number of ##' components used for re-estimation. ##' @title Bayesian PCA missing value estimation ##' @param Matrix \code{matrix} -- Pre-processed matrix (centered, ##' scaled) with variables in columns and observations in rows. The ##' data may contain missing values, denoted as \code{NA}. ##' @param nPcs \code{numeric} -- Number of components used for ##' re-estimation. Choosing few components may decrease the ##' estimation precision. ##' @param maxSteps \code{numeric} -- Maximum number of estimation ##' steps. ##' @param verbose \code{boolean} -- BPCA prints the number of steps ##' and the increase in precision if set to TRUE. Default is ##' interactive(). ##' @param threshold convergence threshold ##' @param ... Reserved for future use. Currently no further ##' parameters are used ##' @return Standard PCA result object used by all PCA-based methods ##' of this package. Contains scores, loadings, data mean and ##' more. See \code{\link{pcaRes}} for details. ##' @references Shigeyuki Oba, Masa-aki Sato, Ichiro Takemasa, Morito ##' Monden, Ken-ichi Matsubara and Shin Ishii. A Bayesian missing ##' value estimation method for gene expression profile ##' data. \emph{Bioinformatics, 19(16):2088-2096, Nov 2003}, ##' \url{https://doi.org/10.1093/bioinformatics/btg287}. ##' @seealso \code{\link{ppca}}, \code{\link{svdImpute}}, ##' \code{\link{prcomp}}, \code{\link{nipalsPca}}, ##' \code{\link{pca}}, ##' \code{\link{pcaRes}}. \code{\link{kEstimate}}. ##' @note Requires \code{MASS}. ##' @examples ##' ## Load a sample metabolite dataset with 5\% missig values (metaboliteData)e ##' data(metaboliteData) ##' ## Perform Bayesian PCA with 2 components ##' pc <- pca(t(metaboliteData), method="bpca", nPcs=2) ##' ## Get the estimated principal axes (loadings) ##' loadings <- loadings(pc) ##' ## Get the estimated scores ##' scores <- scores(pc) ##' ## Get the estimated complete observations ##' cObs <- completeObs(pc) ##' ## Now make a scores and loadings plot ##' slplot(pc) ##' \dontshow{stopifnot(sum((fitted(pc) - t(metaboliteData))^2, na.rm=TRUE) < 200)} ##' @keywords multivariate ##' @export ##' @author Wolfram Stacklies bpca <- function(Matrix, nPcs=2, maxSteps=100, verbose=interactive(), threshold=1e-4, ... ) { ## R implementation of a Bayesion PCA missing value estimator. ## After the Matlab script of Shigeyuki OBA (2002 May. 5th) ## See also: http://hawaii.aist-nara.ac.jp/%7Eshige-o/tools/ ## Great thanks to them! M <- BPCA_initmodel(Matrix, nPcs) tauold <- 1000 for( step in 1:maxSteps ) { M <- BPCA_dostep(M, Matrix) if( step %% 10 == 0 ) { tau <- M$tau dtau <- abs(log10(tau) - log10(tauold)) if ( verbose ) { cat("Step Number : ", step, '\n') cat("Increase in precision : ", dtau, '\n') cat("----------", '\n') } if (dtau < threshold) { break } tauold <- tau } } R2cum <- rep(NA, nPcs) TSS <- sum(Matrix^2, na.rm=TRUE) for (i in 1:nPcs) { difference <- Matrix - (M$scores[,1:i, drop=FALSE] %*% t(M$PA[,1:i, drop=FALSE]) ) R2cum[i] <- 1 - (sum(difference^2, na.rm=TRUE) / TSS) } result <- new("pcaRes") result@scores <- M$scores result@loadings <- M$PA result@R2cum <- R2cum result@method <- "bpca" return(result) } pcaMethods/R/checkData.R0000644000175200017520000000537314710217306016067 0ustar00biocbuildbiocbuild##' Check a given data matrix for consistency with the format ##' required for further analysis. ##' The data must be a numeric matrix and not contain: ##' \itemize{ ##' \item Inf values ##' \item NaN values ##' \item Rows or columns that consist of NA only ##' } ##' @title Do some basic checks on a given data matrix ##' @param data \code{matrix} -- Data to check. ##' @param verbose \code{boolean} -- If TRUE, the function prints ##' messages whenever an error in the data set is found. ##' @return \item{isValid}{\code{boolean} -- TRUE if no errors were ##' found, FALSE otherwise. isValid contains a set of attributes, ##' these are: \itemize{ \item isNumeric - TRUE if data is numeric, ##' false otherwise \item isInfinite - TRUE if data contains 'Inf' ##' values, false otherwise \item isNaN - TRUE if data contains 'NaN' ##' values, false otherwise \item isMatrix - TRUE if the data is in ##' matrix format, FALSE otherwise \item naRows - TRUE if data ##' contains rows in which all elements are 'NA', FALSE otherwise ##' \item naCols - TRUE if data contains columns in which all elements ##' are 'NA', FALSE otherwise }} ##' @keywords multivariate ##' @export ##' @author Wolfram Stacklies checkData <- function(data, verbose = FALSE) { isValid <- TRUE isNumeric <- TRUE isInfinite <- FALSE isNaN <- FALSE isMatrix <- TRUE naRows <- FALSE naCols <- FALSE if (!is.numeric(data)) { isNumeric <- FALSE isValid <- FALSE if (verbose) message("Error: Data is not numeric") } if ( sum(is.infinite(data) >= 1) ) { isInfinite <- TRUE isValid <- FALSE if (verbose) message("Error: Data contains 'Inf' values") } if (sum(is.nan(data) >= 1)) { isNaN <- TRUE isValid <- FALSE if (verbose) message("Error: Data contains 'NaN' values. Missing values must be denoted by 'NA'") } if (!is.matrix(data)) { isMatrix <- FALSE isValid <- FALSE if (verbose) message("Error: data is not a matrix. Try to use as.matrix(data)") } ## Check for entire rows that are NA only if (sum(apply(is.na(data), 1, sum) == ncol(data)) >= 1 ) { naRows <- TRUE isValid <- FALSE if (verbose) message("Error: Data contains rows in which all elements are 'NA'. Remove them first") } ## Check for entire columns that are NA only if (sum(apply(is.na(data), 2, sum) == nrow(data)) >= 1 ) { naCols <- TRUE isValid <- FALSE if (verbose) message("Error: Data contains columns in which all elements are 'NA'. Remove them first") } attr(isValid, "isNumeric") <- isNumeric attr(isValid, "isInfinite") <- isInfinite attr(isValid, "isNaN") <- isNaN attr(isValid, "isMatrix") <- isMatrix attr(isValid, "naRows") <- naRows attr(isValid, "naCols") <- naCols return(isValid) } pcaMethods/R/derrorHierarchic.R0000644000175200017520000001251114710217306017467 0ustar00biocbuildbiocbuild##' Later ##' @param nlnet the nlnet ##' @param trainIn training data ##' @param trainOut fitted data ##' @return derror ##' @author Henning Redestig, Matthias Scholz derrorHierarchic <- function(nlnet, trainIn, trainOut) { weights <- nlnet@weights$current() netDim <- dim(nlnet@net) if(nlnet@inverse) { numElements <- nlnet@net[1] * dim(trainOut)[2] trainIn <- matrix(weights[1:numElements], nrow=nlnet@net[1], ncol=dim(trainOut)[2]) wTrainIn <- weights[1:numElements,drop=FALSE] weights <- weights[(numElements + 1):length(weights), ,drop=FALSE] } weightMats <- vector2matrices(weights, nlnet@net) trainDim <- dim(trainIn) subnetNum <- length(nlnet@hierarchic$var) ## ****************************** Epattern <- array(0, dim=c(dim(trainOut), subnetNum)) nOut <- array(0, dim=c(sum(nlnet@net), trainDim[2], subnetNum)) for(subnet in 1:subnetNum) nOut[1:trainDim[1],,subnet] <- eval(parse(text=paste(nlnet@fct[1], "(trainIn)"))) if(nlnet@inverse) for(subnet in 1:subnetNum) nOut[nlnet@hierarchic$idx[,subnet]==0,,subnet] <- 0 ## forward propagation for(subnet in 1:subnetNum) { if(nlnet@hierarchic$var[subnet] != 0) { sBias <- array(1, dim=c(1, trainDim[2])) for(i in 1:(netDim[2] - 1)) { if(i == 1) nBegin <- 1 else nBegin <- sum(nlnet@net[1:(i-1)])+1 sIn <- rbind(sBias, nOut[nBegin:sum(nlnet@net[1:i]),,subnet]) sOut <- eval(parse(text=paste(nlnet@fct[i+1], "(weightMats[[i]] %*% sIn)"))) if(i == (nlnet@hierarchic$layer - 1)) sOut[nlnet@hierarchic$idx[,subnet]==0,] <- 0 nOut[(sum(nlnet@net[1:i])+1):sum(nlnet@net[1:(i+1)]),,subnet] <- sOut } output <- nOut[(sum(nlnet@net[1:(length(nlnet@net)-1)])+1):dim(nOut)[1],,subnet] Epattern[,,subnet] <- output - trainOut } } ## error function Epattern <- Epattern^2 Epattern[is.na(Epattern)] <- 0 #set the missing values to zero if(!is.null(nlnet@dataDist)) for(subnet in 1:subnetNum) Epattern[,,subnet] <- Epattern[,,subnet] * nlnet@dataDist Eitemize <- apply(Epattern, 3, sum) * 0.5 Etotal <- sum(nlnet@hierarchic$var * Eitemize) if(!is.null(nlnet@weightDecay)) Etotal <- Etotal + nlnet@weightDecay * 0.5 * sum(weights^2) if(nlnet@inverse) Etotal <- Etotal + 0.01 * nlnet@weightDecay * 0.5 * sum(wTrainIn^2) ## back propagation nError <- array(0, dim=c(sum(nlnet@net), trainDim[2], subnet)) dWeight <- vector(length=netDim[2] - 1, mode="list") wBp <- vector(length=netDim[2] - 1, mode="list") ## wBp is weights for back propagation for(u in 1:(netDim[2] - 1)) wBp[[u]] <- weightMats[[u]][,2:(nlnet@net[u] + 1)] # cats the weights which belong to bias dw <- array(0, dim=c(length(weights), subnet)) for(subnet in 1:subnetNum) { if(nlnet@hierarchic$var[subnet] != 0) { ## last layer sTmp <- nOut[(dim(nOut)[1]-nlnet@net[length(nlnet@net)]+1):dim(nOut)[1],,subnet] if(nlnet@fct[length(nlnet@fct)] == "tanh") eTmp <- (1 - sTmp^2) * (sTmp - trainOut) #prev trainOut - sTmp (fixed to get rid of sign change) else if(nlnet@fct[length(nlnet@fct)] == "linr") eTmp <- sTmp - trainOut #prev trainOut - sTmp (fixed to get rid of sign change) eTmp[is.na(eTmp)] <- 0 if(!is.null(nlnet@dataDist)) eTmp <- eTmp * nlnet@dataDist nError[(dim(nError)[1]-nlnet@net[length(nlnet@net)]+1):dim(nError)[1],,subnet] <- eTmp ## all other layers for(n in 1:(netDim[2] - 1)){ i <- netDim[2]-n ## the if clause is to avoid 1:0 difference in R ## Matlab (1:0 => Empty matrix), R (1:0 => [1,0]) if(i > 1) sTmp <- nOut[(sum(nlnet@net[1:(i-1)])+1):sum(nlnet@net[1:i]),,subnet] else sTmp <- nOut[1:sum(nlnet@net[1:i]),,subnet] if(i==(nlnet@hierarchic$layer-1)) eTmp[nlnet@hierarchic$idx[,subnet]==0,] <- 0 dWeight[[i]] <- tcrossprod(eTmp, rbind(sBias, sTmp)) #gradient if (nlnet@fct[i] == "tanh") eTmp <- (1 - sTmp^2) * crossprod(wBp[[i]],eTmp) else if (nlnet@fct[i] == "linr") eTmp <- crossprod(wBp[[i]], eTmp) ## the if clause is to avoid 1:0 difference in R if(i > 1) nError[(sum(nlnet@net[1:(i - 1)]) + 1):sum(nlnet@net[1:i]), ,subnet] <- eTmp else nError[1:sum(nlnet@net[1:i]), ,subnet] <- eTmp } dw[,subnet] <- unlist(dWeight) #fixed sign change } } if(nlnet@inverse) { dw <- rbind(array(0, dim=c(numElements, subnetNum)), dw) for(subnet in 1:subnetNum) { eTmp <- array(nError[1:nlnet@net[1],,subnet], dim=c(nlnet@net[1], dim(nError)[2])) eTmp[nlnet@hierarchic$idx[,subnet] == 0,] <- 0 dim(eTmp) <- NULL #a bit unsure if this is correct but seems to work dw[1:numElements,subnet] <- unlist(eTmp) #fixed sign change } ## weights <- rbind(cbind(rep(0, numElements)), cbind(weights)) #old: only weight decay for real weights weights <- rbind(cbind(0.01 * wTrainIn), cbind(weights)) #new } dwTotal <- array(0, dim=dim(weights)) for (subnet in 1:subnetNum) { dwTotal <- dwTotal + nlnet@hierarchic$var[subnet] * dw[, subnet] } if(!is.null(nlnet@weightDecay)) dwTotal <- dwTotal + nlnet@weightDecay * weights return(list(dwTotal=dwTotal, Etotal=Etotal, nError=nError, nOut=nOut)) } pcaMethods/R/errorHierarchic.R0000644000175200017520000000440014710217306017321 0ustar00biocbuildbiocbuild##' Later ##' @param nlnet The nlnet ##' @param trainIn training data ##' @param trainOut fitted data ##' @return error ##' @author Henning Redestig, Matthias Scholz errorHierarchic <- function(nlnet, trainIn, trainOut) { weights <- nlnet@weights$current() if(nlnet@inverse) { numElements <- nlnet@net[1] * dim(trainOut)[2] trainIn <- array(weights[1:numElements], dim=c(nlnet@net[1], dim(trainOut)[2])) wTrainIn <- weights[1:numElements, drop=FALSE] weights <- weights[(numElements + 1):length(weights),,drop=FALSE] } netDim <- dim(nlnet@net) trainDim <- dim(trainOut) weightMats <- vector2matrices(weights, nlnet@net) hierarchicIdx <- nlnet@hierarchic$idx[,nlnet@hierarchic$var != 0, drop=FALSE] hierarchicVar <- nlnet@hierarchic$var[,colSums(nlnet@hierarchic$var) != 0, drop=FALSE] subnetNum <- length(hierarchicVar) out <- array(0, dim=c(trainDim[1], trainDim[2], subnetNum)) sBias <- array(1, dim=c(1, trainDim[2])) sExtract <- eval(parse(text=paste(nlnet@fct[1], "(trainIn)"))) Eitemize <- NULL if(nlnet@hierarchic$layer > 1) { #this should not be executed at all if sequence is 1:0 for(layer in 1:(nlnet@hierarchic$layer - 1)) { sExtract <- rbind(sBias, sExtract) sExtract <- eval(parse(text=paste(nlnet@fct[layer + 1], "(weightMats[[layer]] %*% sExtract)"))) } } for(subnet in 1:subnetNum) { sRecon <- sExtract sRecon[hierarchicIdx[,subnet]==0,] <- 0 for(layer in nlnet@hierarchic$layer:(netDim[2] - 1)) { sRecon <- rbind(sBias, sRecon) sRecon <- eval(parse(text=paste(nlnet@fct[layer+1], "(weightMats[[layer]] %*% sRecon)"))) } out[,,subnet] <- sRecon ## error function eTmp <- (sRecon - trainOut)^2 eTmp[is.na(eTmp)] <- 0 Eitemize[subnet] <- sum(eTmp) * 0.5 if(!is.null(nlnet@dataDist)) Eitemize[subnet] <- 0.5 * sum(nlnet@dataDist * eTmp) else Eitemize[subnet] <- 0.5 * sum(eTmp) } error <- tcrossprod(hierarchicVar, rbind(Eitemize)) if(!is.null(nlnet@weightDecay)) error <- error + nlnet@weightDecay * 0.5 * sum(weights^2) ## smooth (0.01) weight decay also for input values if(nlnet@inverse) error <- error + 0.01 * nlnet@weightDecay * 0.5 * sum(wTrainIn^2) return(list(error=error, out=out)) } pcaMethods/R/forkNlpcaNet.R0000644000175200017520000000131414710217306016575 0ustar00biocbuildbiocbuild##' Complete copy of nlpca net object ##' @param nlnet a nlnet ##' @return A copy of the input nlnet ##' @author Henning Redestig forkNlpcaNet <- function(nlnet) { res <- new("nlpcaNet") res@net <- nlnet@net res@hierarchic <- nlnet@hierarchic res@fct <- nlnet@fct res@fkt <- nlnet@fkt res@weightDecay <- nlnet@weightDecay res@featureSorting <- nlnet@featureSorting res@dataDist <- nlnet@dataDist res@inverse <- nlnet@inverse res@fCount <- nlnet@fCount res@componentLayer <- nlnet@componentLayer res@error <- nlnet@error res@gradient <- nlnet@gradient res@weights <- weightsAccount(nlnet@weights$current()) res@maxIter <- nlnet@maxIter res@scalingFactor <- nlnet@scalingFactor res } pcaMethods/R/kEstimate.R0000644000175200017520000003072014710217306016140 0ustar00biocbuildbiocbuild##' Perform cross validation to estimate the optimal number of ##' components for missing value estimation. Cross validation is ##' done for the complete subset of a variable. ##' ##' The assumption hereby is that variables that are highly correlated ##' in a distinct region (here the non-missing observations) are also ##' correlated in another (here the missing observations). This also ##' implies that the complete subset must be large enough to be ##' representative. For each incomplete variable, the available ##' values are divided into a user defined number of cv-segments. The ##' segments have equal size, but are chosen from a random equal ##' distribution. The non-missing values of the variable are covered ##' completely. PPCA, BPCA, SVDimpute, Nipals PCA, llsImpute an NLPCA ##' may be used for imputation. ##' ##' The whole cross validation is repeated several times so, depending ##' on the parameters, the calculations can take very long time. As ##' error measure the NRMSEP (see Feten et. al, 2005) or the Q2 ##' distance is used. The NRMSEP basically normalises the RMSD ##' between original data and estimate by the variable-wise ##' variance. The reason for this is that a higher variance will ##' generally lead to a higher estimation error. If the number of ##' samples is small, the variable - wise variance may become an ##' unstable criterion and the Q2 distance should be used ##' instead. Also if variance normalisation was applied previously. ##' ##' The method proceeds variable - wise, the NRMSEP / Q2 distance is ##' calculated for each incomplete variable and averaged ##' afterwards. This allows to easily see for wich set of variables ##' missing value imputation makes senes and for wich set no ##' imputation or something like mean-imputation should be used. Use ##' \code{kEstimateFast} or \code{Q2} if you are not interested in ##' variable wise CV performance estimates. ##' ##' Run time may be very high on large data sets. Especially when used ##' with complex methods like BPCA or Nipals PCA. For PPCA, BPCA, ##' Nipals PCA and NLPCA the estimation method is called ##' \eqn{(v_{miss} \cdot segs \cdot nruncv \cdot)}{(v\_miss * segs * ##' nruncv)} times as the error for all numbers of principal ##' components can be calculated at once. For LLSimpute and SVDimpute ##' this is not possible, and the method is called \eqn{(v_{miss} ##' \cdot segs \cdot nruncv \cdot length(evalPcs))}{(v\_miss * segs * ##' nruncv * length(evalPcs))} times. This should still be fast for ##' LLSimpute because the method allows to choose to only do the ##' estimation for one particular variable. This saves a lot of ##' iterations. Here, \eqn{v_{miss}}{v\_miss} is the number of ##' variables showing missing values. ##' ##' As cross validation is done variable-wise, in this function Q2 is ##' defined on single variables, not on the entire data set. This is ##' Q2 is calculated as as \eqn{\frac{\sum(x - ##' xe)^2}{\sum(x^2)}}{sum(x - xe)^2 \ sum(x^2)}, where x is the ##' currently used variable and xe it's estimate. The values are then ##' averaged over all variables. The NRMSEP is already defined ##' variable-wise. For a single variable it is then ##' \eqn{\sqrt(\frac{\sum(x - xe)^2}{(n \cdot var(x))})}{sqrt(sum(x - ##' xe)^2 \ (n * var(x)))}, where x is the variable and xe it's ##' estimate, n is the length of x. The variable wise estimation ##' errors are returned in parameter variableWiseError. ##' @title Estimate best number of Components for missing value ##' estimation ##' @param Matrix \code{matrix} -- numeric matrix containing ##' observations in rows and variables in columns ##' @param method \code{character} -- of the methods found with ##' pcaMethods() The option llsImputeAll calls llsImpute with the ##' allVariables = TRUE parameter. ##' @param evalPcs \code{numeric} -- The principal components to use ##' for cross validation or the number of neighbour variables if used ##' with llsImpute. Should be an array containing integer values, ##' eg. \code{evalPcs = 1:10} or \code{evalPcs = c(2,5,8)}. The NRMSEP ##' or Q2 is calculated for each component. ##' @param segs \code{numeric} -- number of segments for cross validation ##' @param nruncv \code{numeric} -- Times the whole cross validation ##' is repeated ##' @param em \code{character} -- The error measure. This can be nrmsep or q2 ##' @param allVariables \code{boolean} -- If TRUE, the NRMSEP is ##' calculated for all variables, If FALSE, only the incomplete ones ##' are included. You maybe want to do this to compare several methods ##' on a complete data set. ##' @param verbose \code{boolean} -- If TRUE, some output like the ##' variable indexes are printed to the console each iteration. ##' @param ... Further arguments to \code{pca} or \code{nni} ##' @return A list with: ##' \item{bestNPcs}{number of PCs or k for which the minimal average ##' NRMSEP or the maximal Q2 was obtained.} ##' \item{eError}{an array of of size length(evalPcs). Contains the ##' average error of the cross validation runs for each number of ##' components.} ##' \item{variableWiseError}{Matrix of size ##' \code{incomplete_variables} x length(evalPcs). Contains the ##' NRMSEP or Q2 distance for each variable and each number of PCs. ##' This allows to easily see for wich variables imputation makes ##' sense and for which one it should not be done or mean imputation ##' should be used.} ##' \item{evalPcs}{The evaluated numbers of components or number of ##' neighbours (the same as the evalPcs input parameter).} ##' \item{variableIx}{Index of the incomplete variables. This can be ##' used to map the variable wise error to the original data.} ##' @seealso \code{\link{kEstimateFast}, \link{Q2}, \link{pca}, \link{nni}}. ##' @examples ##' ## Load a sample metabolite dataset with 5\% missing values (metaboliteData) ##' data(metaboliteData) ##' # Do cross validation with ppca for component 2:4 ##' esti <- kEstimate(metaboliteData, method = "ppca", evalPcs = 2:4, nruncv=1, em="nrmsep") ##' # Plot the average NRMSEP ##' barplot(drop(esti$eError), xlab = "Components",ylab = "NRMSEP (1 iterations)") ##' # The best result was obtained for this number of PCs: ##' print(esti$bestNPcs) ##' # Now have a look at the variable wise estimation error ##' barplot(drop(esti$variableWiseError[, which(esti$evalPcs == esti$bestNPcs)]), ##' xlab = "Incomplete variable Index", ylab = "NRMSEP") ##' @keywords multivariate ##' @export ##' @author Wolfram Stacklies kEstimate <- function(Matrix, method="ppca", evalPcs=1:3, segs=3, nruncv=5, em="q2", allVariables=FALSE, verbose=interactive(), ...) { fastKE <- FALSE if (method == "ppca" | method == "bpca" | method == "nipals" | method == "nlpca") fastKE <- TRUE method <- match.arg(method, listPcaMethods()) em <- match.arg(em, c("nrmsep", "q2")) maxPcs <- max(evalPcs) lengthPcs <- length(evalPcs) ## If the data is a data frame, convert it into a matrix Matrix <- as.matrix(Matrix, rownames.force=TRUE) if(maxPcs > (ncol(Matrix) - 1)) stop("maxPcs exceeds matrix size, choose a lower value!") ## And now check if everything is right... if( !checkData(Matrix, verbose=interactive()) ) stop("Invalid data format! Use checkData(Matrix, verbose = TRUE) for details.\n") if( (sum(is.na(Matrix)) == 0) && (allVariables == FALSE) ) stop("No missing values. Maybe you want to set allVariables = TRUE. Exiting\n") missing <- apply(is.na(Matrix), 2, sum) > 0 missIx <- which(missing == TRUE) if (allVariables) missIx <- 1:ncol(Matrix) complete <- !missing compIx <- which(complete == TRUE) error <- matrix(0, length(missIx), length(evalPcs)) iteration <- 0 for(nPcs in evalPcs) { ## If the estimated observations are just scores %*% t(loadings) ## we can calculate all we need at once, this saves many ## iterations... if (fastKE) nPcs = maxPcs iteration = iteration + 1 if (verbose && !fastKE) { cat("Doing CV for ", nPcs, " component(s) \n") } else if (verbose && fastKE) {cat("Doing CV ... \n")} for(cviter in 1:nruncv) { pos <- 0 if (verbose) cat("Incomplete variable index: ") for (index in missIx) { pos <- pos + 1 cat(pos, ":", sep="") target <- Matrix[, index, drop = FALSE] compObs <- !is.na(target) missObs <- is.na(target) nObs <- sum(compObs) ## Remove all observations that are missing in the target genes, ## as additional missing values may tamper the results set <- Matrix[compObs,] if (nObs >= (2 * segs)) { segments <- segs } else segments <- ceiling(nObs / 2) ## We assume uniformly distributed missing values when ## choosing the segments tt <- gl(segments, ceiling(nObs / segments))[1:nObs] cvsegs <- split(sample(nObs), tt) set <- Matrix[compObs,] if (fastKE) { nrmsep <- array(0, length(evalPcs)) q2 <- array(0, length(evalPcs)) } else { nrmsep <- 0; q2 <- 0 } for (i in 1:length(cvsegs)) { n <- length(cvsegs[[i]]) # n is the number of created # missing values ## Impute values using the given regression method testSet <- set testSet[cvsegs[[i]], index] <- NA if (method == "llsImpute") { estimate <- llsImpute(testSet, k = nPcs, verbose = FALSE, allVariables = FALSE, center = FALSE, xval = index) } else if (method == "llsImputeAll") { estimate <- llsImpute(testSet, k = nPcs, verbose = FALSE, allVariables = TRUE, center = FALSE, xval = index) } else { estimate <- pca(testSet, nPcs = nPcs, verbose = FALSE, method = method, center = TRUE,...) } if (fastKE) { for (np in evalPcs) { estiFitted <- fitted(estimate, data = NULL, nPcs = np) estimateVec <- estiFitted[, index] original <- target[compObs, ] estimateVec[-cvsegs[[i]]] <- testSet[-cvsegs[[i]], index] ## Error of prediction, error is calculated for removed ## elements only nIx <- which(evalPcs == np) if (em == "nrmsep") { nrmsep[nIx] <- nrmsep[nIx] + sum( (original - estimateVec)^2) } else { q2[nIx] <- q2[nIx] + sum( (original - estimateVec)^2 ) } } } else { estimate <- estimate@completeObs[, index] original <- target[compObs, ] ## Error of prediction, error is calculated for removed ## elements only if (em == "nrmsep") { nrmsep <- nrmsep + sum( (original - estimate)^2) } else { q2 <- q2 + sum( (original - estimate)^2 ) } } } ## iteration over cv segments if (fastKE) { if (em == "nrmsep") { error[pos, ] <- error[pos, ] + nrmsep / (nrow(set) * var(set[,index])) } else error[pos, ] <- error[pos, ] + (1 - (q2 / sum(set[, index]^2))) } else { if (em == "nrmsep") { error[pos, iteration] <- error[pos, iteration] + nrmsep / (nrow(set) * var(set[,index])) } else error[pos, iteration] <- error[pos, iteration] + (1 - (q2 / sum(set[, index]^2))) } } # iteration over variables if (verbose) cat("\n") } #iteration over nruncv ## The error is the sum over the independent cross validation runs error <- error / nruncv if (verbose && !fastKE) cat("The average", em, "for k =", iteration, "is", sum(error[,iteration]) / nrow(error), "\n") ## if nlpca, ppca, bpca, nipals we do not need to iterate over the ## number of components... if (fastKE) break } # iteration over number components if (em == "nrmsep") avgError <- sqrt(apply(error, 2, sum) / nrow(error)) else avgError <- apply(error, 2, sum) / nrow(error) ret <- list() if (em == "nrmsep") ret$bestNPcs <- evalPcs[which(avgError == min(avgError))] else ret$bestNPcs <- evalPcs[which(avgError == max(avgError))] ret$eError <- avgError if(em == "nrmsep") ret$variableWiseError <- sqrt(error) else ret$variableWiseError <- error ret$evalPcs <- evalPcs ret$variableIx <- missIx return(ret) } pcaMethods/R/kEstimateFast.R0000644000175200017520000001215414710217306016757 0ustar00biocbuildbiocbuild##' This is a simple estimator for the optimal number of componets ##' when applying PCA or LLSimpute for missing value estimation. No ##' cross validation is performed, instead the estimation quality is ##' defined as Matrix[!missing] - Estimate[!missing]. This will give a ##' relatively rough estimate, but the number of iterations equals the ##' length of the parameter evalPcs.\cr Does not work with LLSimpute!! ##' As error measure the NRMSEP (see Feten et. al, 2005) or the Q2 ##' distance is used. The NRMSEP basically normalises the RMSD ##' between original data and estimate by the variable-wise ##' variance. The reason for this is that a higher variance will ##' generally lead to a higher estimation error. If the number of ##' samples is small, the gene - wise variance may become an unstable ##' criterion and the Q2 distance should be used instead. Also if ##' variance normalisation was applied previously. ##' @title Estimate best number of Components for missing value estimation ##' @param Matrix \code{matrix} -- numeric matrix containing ##' observations in rows and variables in columns ##' @param method \code{character} -- a valid pca method (see ##' \code{\link{pca}}). ##' @param evalPcs \code{numeric} -- The principal components to use ##' for cross validation or cluster sizes if used with ##' llsImpute. Should be an array containing integer values, ##' eg. evalPcs = 1:10 or evalPcs = C(2,5,8).The NRMSEP is calculated ##' for each component. ##' @param em \code{character} -- The error measure. This can be ##' nrmsep or q2 ##' @param allVariables \code{boolean} -- If TRUE, the NRMSEP is ##' calculated for all variables, If FALSE, only the incomplete ones ##' are included. You maybe want to do this to compare several methods ##' on a complete data set. ##' @param verbose \code{boolean} -- If TRUE, the NRMSEP and the ##' variance are printed to the console each iteration. ##' @param ... Further arguments to \code{pca} ##' @return \item{list}{Returns a list with the elements: ##' \itemize{ ##' \item minNPcs - number of PCs for which the minimal average NRMSEP ##' was obtained ##' \item eError - an array of of size length(evalPcs). Contains the ##' estimation error for each number of ##' components. ##' \item evalPcs - The evaluated numbers of components or ##' cluster sizes (the same as the evalPcs input parameter). }} ##' @seealso \code{\link{kEstimate}}. ##' @export ##' @examples ##' data(metaboliteData) ##' # Estimate best number of PCs with ppca for component 2:4 ##' esti <- kEstimateFast(t(metaboliteData), method = "ppca", evalPcs = 2:4, em="nrmsep") ##' barplot(drop(esti$eError), xlab = "Components",ylab = "NRMSEP (1 iterations)") ##' # The best k value is: ##' print(esti$minNPcs) ##' @keywords multivariate ##' @author Wolfram Stacklies kEstimateFast <- function(Matrix, method = "ppca", evalPcs = 1:3, em = "nrmsep", allVariables = FALSE, verbose = interactive(), ...) { method <- match.arg(method, c("ppca", "bpca", "svdImpute", "nipals", "nlpca")) em <- match.arg(em, c("nrmsep", "q2")) maxPcs <- max(evalPcs) lengthPcs <- length(evalPcs) missing <- is.na(Matrix) error <- array(0, lengthPcs) ## If the data is a data frame, convert it into a matrix Matrix <- as.matrix(Matrix, rownames.force=TRUE) if(maxPcs > (ncol(Matrix) - 1)) stop("maxPcs exceeds matrix size, choose a lower value!") ## And now check if everything is right... if( !checkData(Matrix, verbose=interactive()) ) stop("Invalid data format! Use checkData(Matrix, verbose = TRUE) for details.\n") if( (sum(is.na(Matrix)) == 0) && (allVariables == FALSE) ) stop("No missing values. Maybe you want to set allVariables = TRUE. Exiting\n") iteration = 0 for(nPcs in evalPcs) { iteration = iteration + 1 if (method == "nlpca") { estimate <- fitted(pca(Matrix, nPcs = nPcs, verbose = FALSE, method = method, center = TRUE,...), Matrix, nPcs = nPcs) } else { estimate <- fitted(pca(Matrix, nPcs = nPcs, verbose = FALSE, method = method, center = TRUE,...), nPcs = nPcs) } if (em == "q2") { # The Q2 distance q2 <- 1 - sum((Matrix[!missing] - estimate[!missing])^2) / sum(Matrix[!missing]^2) error[iteration] <- q2 } else { nrmsep <- 0 for(i in 1:ncol(Matrix)) { nrmsep <- nrmsep + ( sum((Matrix[!missing[,i], i] - estimate[!missing[,i], i])^2) / (sum(!missing[,i]) * var(Matrix[,i], na.rm = TRUE)) ) } nrmsep <- nrmsep / sum(apply(missing, 2, sum) > 0) error[iteration] <- nrmsep } if(verbose) cat("The", em, "for", evalPcs[iteration], "components is:", error[iteration], "\n") } ret <- list() if (em == "nrmsep") ret$bestNPcs <- evalPcs[which(error == min(error))] else ret$bestNPcs <- evalPcs[which(error == max(error))] ret$eError <- error ret$evalPcs <- evalPcs return(ret) } pcaMethods/R/lineSearch.R0000644000175200017520000000615214710217306016271 0ustar00biocbuildbiocbuild##' Line search for conjugate gradient ##' @param nlnet The nlnet ##' @param dw .. ##' @param e0 .. ##' @param ttGuess .. ##' @param trainIn Training data ##' @param trainOut Fitted data ##' @param verbose logical, print messages ##' @return ... ##' @author Henning Redestig, Matthias Scholz lineSearch <- function(nlnet, dw, e0, ttGuess, trainIn, trainOut, verbose) { iterGoldenSectionSearch <- 6 alpha <- 0.618034 tt <- rep(0, 4) e <- rep(0, 4) tmpnlnet <- forkNlpcaNet(nlnet) tt[1] <- 0 e[1] <- e0 tt[4] <- ttGuess tmpnlnet@weights$set(nlnet@weights$current() + tt[4] * dw) e[4] <- nlnet@error(tmpnlnet, trainIn, trainOut)$error if(e[4] > e[1]) { #got final interval calculate tt[2] and tt[3] tt[2] <- tt[1] + (1 - alpha) * (tt[4] - tt[1]) tmpnlnet@weights$set(nlnet@weights$current() + tt[2] * dw) e[2] <- nlnet@error(tmpnlnet, trainIn, trainOut)$error tt[3] <- tt[1] + alpha * (tt[4] - tt[1]) tmpnlnet@weights$set(nlnet@weights$current() + tt[3] * dw) e[3] <- nlnet@error(tmpnlnet, trainIn, trainOut)$error } else { #expand, add new tt[4] tt[3] <- tt[4] e[3] <- e[4] tt[4] <- (1 + alpha) * tt[4] tmpnlnet@weights$set(nlnet@weights$current() + tt[4] * dw) e[4] <- nlnet@error(tmpnlnet, trainIn, trainOut)$error if(e[4] > e[3]) { #got final interval, calculate tt[2] tt[2] <- tt[1] + (1 - alpha) * (tt[4] - tt[1]) tmpnlnet@weights$set(nlnet@weights$current() + tt[2] * dw) e[2] <- nlnet@error(tmpnlnet, trainIn, trainOut)$error } else { #expand: add new tt[4] i <- 1 while(e[4] < e[3] && i < 50) { tt[2] <- tt[3] e[2] <- e[3] tt[3] <- tt[4] e[3] <- e[4] tt[4] <- (1 + alpha) * tt[4] tmpnlnet@weights$set(nlnet@weights$current() + tt[4] * dw) e[4] <- nlnet@error(tmpnlnet, trainIn, trainOut)$error i <- i + 1 if(verbose && i == 50) cat("^") } } } ## golden section search for(i in 1:iterGoldenSectionSearch) { if(e[3] > e[2]) { tt[4] <- tt[3] #remove right value tt[4] e[4] <- e[3] tt[3] <- tt[2] e[3] <- e[2] tt[2] <- tt[1] + (1 - alpha) * (tt[4] - tt[1]) #split left interval tmpnlnet@weights$set(nlnet@weights$current() + tt[2] * dw) e[2] <- nlnet@error(tmpnlnet, trainIn, trainOut)$error } else { tt[1] <- tt[2] #remove left t value tt[1] e[1] <- e[2] tt[2] <- tt[3] e[2] <- e[3] tt[3] <- tt[1] + alpha * (tt[4] - tt[1]) #split right interval tmpnlnet@weights$set(nlnet@weights$current() + tt[3] * dw) e[3] <- nlnet@error(tmpnlnet, trainIn, trainOut)$error } } if(e[2] < e[3]) { eBest <- e[2] ttBest <- tt[2] } else { eBest <- e[3] ttBest <- tt[3] } wBest <- nlnet@weights$current() + ttBest * dw return(list(wBest=wBest, eBest=eBest, ttBest=ttBest)) } ##' Linear kernel ##' @param x datum ##' @return Input value ##' @author Henning Redestig, Matthias Scholz linr <- function(x) x pcaMethods/R/llsImpute.R0000644000175200017520000002204114710217306016165 0ustar00biocbuildbiocbuild##' Missing value estimation using local least squares (LLS). First, ##' k variables (for Microarrya data usually the genes) are selected ##' by pearson, spearman or kendall correlation coefficients. Then ##' missing values are imputed by a linear combination of the k ##' selected variables. The optimal combination is found by LLS ##' regression. The method was first described by Kim et al, ##' Bioinformatics, 21(2),2005. ##' ##' Missing values are denoted as \code{NA}\cr It is not recommended ##' to use this function directely but rather to use the nni() wrapper ##' function. The methods provides two ways for missing value ##' estimation, selected by the \code{allVariables} option. The first ##' one is to use only complete variables for the regression. This is ##' preferable when the number of incomplete variables is relatively ##' small. ##' ##' The second way is to consider all variables as candidates for the ##' regression. Hereby missing values are initially replaced by the ##' columns wise mean. The method then iterates, using the current ##' estimate as input for the regression until the change between new ##' and old estimate falls below a threshold (0.001). ##' ##' @title LLSimpute algorithm ##' @param Matrix \code{matrix} -- Data containing the variables ##' (genes) in columns and observations (samples) in rows. The data ##' may contain missing values, denoted as \code{NA}. ##' @param k \code{numeric} -- Cluster size, this is the number of ##' similar genes used for regression. ##' @param center \code{boolean} -- Mean center the data if TRUE ##' @param completeObs \code{boolean} -- Return the estimated complete ##' observations if TRUE. This is the input data with NA values ##' replaced by the estimated values. ##' @param correlation \code{character} -- How to calculate the ##' distance between genes. One out of pearson | kendall | spearman , ##' see also help("cor"). ##' @param allVariables \code{boolean} -- Use only complete genes to ##' do the regression if TRUE, all genes if FALSE. ##' @param maxSteps \code{numeric} -- Maximum number of iteration ##' steps if allGenes = TRUE. ##' @param xval \code{numeric} Use LLSimpute for cross ##' validation. xval is the index of the gene to estimate, all other ##' incomplete genes will be ignored if this parameter is set. We do ##' not consider them in the cross-validation. ##' @param verbose \code{boolean} -- Print step number and relative ##' change if TRUE and allVariables = TRUE ##' @param ... Reserved for parameters used in future version of the ##' algorithm ##' @note Each step the generalized inverse of a \code{miss} x k ##' matrix is calculated. Where \code{miss} is the number of missing ##' values in variable j and \code{k} the number of neighbours. This ##' may be slow for large values of k and / or many missing ##' values. See also help("ginv"). ##' @return \item{nniRes}{Standard nni (nearest neighbour ##' imputation) result object of this package. See ##' \code{\link{nniRes}} for details.} ##' @seealso \code{\link{pca}, \link{nniRes}, \link{nni}}. ##' @examples ##' ## Load a sample metabolite dataset (metaboliteData) with already 5\% of ##' ## data missing ##' data(metaboliteData) ##' ## Perform llsImpute using k = 10 ##' ## Set allVariables TRUE because there are very few complete variables ##' result <- llsImpute(metaboliteData, k = 10, correlation="pearson", allVariables=TRUE) ##' ## Get the estimated complete observations ##' cObs <- completeObs(result) ##' @keywords multivariate ##' @export ##' @references Kim, H. and Golub, G.H. and Park, H. - Missing value ##' estimation for DNA microarray gene expression data: local least ##' squares imputation. \emph{Bioinformatics, 2005; 21(2):187-198.} ##' ##' Troyanskaya O. and Cantor M. and Sherlock G. and Brown P. and ##' Hastie T. and Tibshirani R. and Botstein D. and Altman RB. - ##' Missing value estimation methods for DNA microarrays. ##' \emph{Bioinformatics. 2001 Jun;17(6):520-525.} ##' @author Wolfram Stacklies llsImpute <- function(Matrix, k=10, center=FALSE, completeObs=TRUE, correlation="pearson", allVariables=FALSE, maxSteps=100, xval=NULL, verbose=FALSE, ...) { threshold <- 0.001 correlation <- match.arg(correlation, c("pearson", "kendall", "spearman")) ## If the data is a data frame, convert it into a matrix Matrix <- as.matrix(Matrix, rownames.force=TRUE) ## And now check if everything is right... if ( !checkData(Matrix, verbose = interactive()) ) { stop("Invalid data format! Use checkData(Matrix, verbose = TRUE) for details.\n") } ## Exit if number of neighbours exceeds number of columns if (k > ncol(Matrix)) stop("Cluster size larger than the number of columns, choose a k < ncol(Matrix)!") ## Set allVariables TRUE if k exceeds number of complete genes ## Print warning messages in the first case and when less than 50% of all genes are complete ## and allVariables == FALSE cg <- sum( apply(is.na(Matrix), 2, sum) == 0) if ( (k > cg) && (!allVariables) ) { warning("Cluster size larger than number of complete genes, using allVariables = TRUE") allVariables <- TRUE } else if ( (cg < (ncol(Matrix) / 2)) && (!allVariables) ) { warning("Less than 50% of the genes are complete, consider using allVariables = TRUE") } else if (sum(is.na(Matrix)) == 0) stop("No missing values, no need for missing value imputation :))") ## Find all genes with missing values missing <- apply(is.na(Matrix), 2, sum) > 0 missIx <- which(missing == TRUE) # For cross validation we want to only estimate one variable, the others # are not considered in the cross validation anyway if (!is.null(xval)) missIx = xval obs <- Matrix ## working copy of the data Ye <- Matrix ## Estimated complete observations ## Center the data column wise if (center) { obs <- scale(Matrix, center = TRUE, scale = FALSE) Ye <- obs means <- attr(Ye, "scaled:center") } if (allVariables) { compIx <- 1:ncol(obs) ## Impute the row average rowMeans <- apply(obs, 1, mean, na.rm = TRUE) for (i in 1:nrow(obs)) { obs[i, is.na(Matrix[i,])] <- rowMeans[i] } ## distances between all genes, ignore the diagonal (correlation to itself) distance = abs(cor(obs, obs, method = correlation)) } else { compIx <- which(missing == FALSE) ## missing genes are the rows, complete genes the columns distance = abs(cor(obs[,missIx, drop=FALSE], obs[,compIx, drop=FALSE], use="pairwise.complete.obs", method = correlation)) } change <- Inf step <- 0 while ( (change > threshold) && (step < maxSteps) ) { step <- step + 1 iteration <- 0 ## Do the regression and imputation for (index in missIx) { iteration <- iteration + 1 if (allVariables) { similar <- sort(distance[iteration,], index.return = TRUE, decreasing = TRUE) simIx <- compIx[ similar$ix[similar$ix != iteration][1:k] ] } else { similar <- sort(distance[iteration,], index.return = TRUE, decreasing = TRUE) simIx <- compIx[ similar$ix[1:k] ] } ## ## Do a regression against the k most similar genes ## See Kim et. al 2005 for details ## target <- obs[, index, drop = FALSE] tMiss <- is.na(Matrix[, index, drop = FALSE]) Apart <- obs[!tMiss, simIx, drop = FALSE] Bpart <- obs[tMiss, simIx, drop = FALSE] targetComplete <- target[!tMiss, , drop = FALSE] X <- MASS::ginv(Apart) %*% targetComplete estimate <- Bpart %*% X ## Impute the estimate Ye[tMiss, index] <- estimate } ## We do not want to iterate if allVariables == FALSE if (!allVariables || !is.null(xval)) { break } else { ## relative change in estimation change <- sqrt(sum( (obs - Ye)^2 ) / sum(obs^2)) obs <- Ye if (verbose) { cat("Step number : ", step, '\n') cat("Relative change : ", change, '\n') cat("---------------", '\n') } } } ## Add the original mean if (center) { for(i in 1:ncol(Ye)) { Ye[,i] <- Ye[,i] + means[i] } } ## Build the nniRes object ## result <- new("nniRes") if(completeObs) { Ye[!is.na(Matrix)] <- Matrix[!is.na(Matrix)] result@completeObs <- Ye } result@centered <- center result@center <- attr(scale(Matrix, center = TRUE, scale = FALSE), "scaled:center") result@nObs <- nrow(Matrix) result@nVar <- ncol(Matrix) result@method <- "llsImpute" result@correlation <- correlation result@k <- k result@missing <- sum(is.na(Matrix)) return(result) } pcaMethods/R/methods-ExpressionSet.R0000644000175200017520000000272614710217306020473 0ustar00biocbuildbiocbuild##' This function can be used to conveniently replace the expression ##' matrix in an \code{ExpressionSet} with the completed data from a ##' \code{pcaRes} object. ##' ##' This is not a standard \code{as} function as \code{pcaRes} ##' object alone not can be converted to an \code{ExpressionSet} (the ##' \code{pcaRes} object does not hold any \code{phenoData} for ##' example). ##' @title Convert pcaRes object to an expression set ##' @param object \code{pcaRes} -- The object containing the completed ##' data. ##' @param exprSet \code{ExpressionSet} -- The object passed on to ##' \code{pca} for missing value estimation. ##' @return An object without missing values of class \code{ExpressionSet}. ##' @export ##' @author Wolfram Stacklies \cr CAS-MPG Partner Institute for ##' Computational Biology, Shanghai, China ##' @keywords multivariate asExprSet <- function(object, exprSet) { if(!inherits(exprSet, "ExpressionSet")) stop("Parameter exprSet must be of type ExpressionSet") if(!inherits(object, "pcaRes") & !inherits(object, "nniRes")) stop("Parameter object must be either of type pcaRes or nniRes") if (is.null(completeObs(object))) stop("completeObs(object) is NULL, exiting") if(!all(dim(exprs(exprSet)) == dim(t(completeObs(object))))) stop("Dimensions of exprs(exprSet) and completeObs(object) do not match. Did you really do missing value estimation using this ExpressionSet object?") exprs(exprSet) <- t(completeObs(object)) return(exprSet) } pcaMethods/R/methods-nniRes.R0000644000175200017520000000161114710217306017106 0ustar00biocbuildbiocbuild##' Print a brief description of nniRes model ##' @title Print a nniRes model ##' @param x An \code{nniRes} object ##' @param ... Not used ##' @return Nothing, used for side-effect ##' @export ##' @author Henning Redestig showNniRes <- function(x, ...) { summary(x) cat(dim(x)["nVar"], "\tVariables\n") cat(dim(x)["nObs"],"\tSamples\n") cat(nmissing(x), "\tNAs (", round(100 * nmissing(x) / (nObs(x) * nVar(x)), getOption("str")$digits.d), "%)\n") cat("k was set to", x@k, "\n") if(centered(x)) cat("Data was mean centered before running LLSimpute \n") else cat("Data was NOT mean centered before running LLSimpute \n") if(scaled(x)) cat("Data was scaled before running LLSimpute \n") else cat("Data was NOT scaled before running LLSimpute \n") } setMethod("print", "nniRes", showNniRes) setMethod("show", "nniRes", function(object) showNniRes(object)) pcaMethods/R/methods-pcaRes.R0000644000175200017520000007727514710217306017110 0ustar00biocbuildbiocbuild##' The leverages of PCA model indicate how much influence each ##' observation has on the PCA model. Observations with high leverage ##' has caused the principal components to rotate towards them. It can ##' be used to extract both "unimportant" observations as well as ##' picking potential outliers. ##' ##' Defined as \eqn{Tr(T(T'T)^{-1}T')}{Tr(T(T'T)^(-1)T')} ##' @title Extract leverages of a PCA model ##' @param object a \code{pcaRes} object ##' @return The observation leverages as a numeric vector ##' @references Introduction to Multi- and Megavariate Data Analysis ##' using Projection Methods (PCA and PLS), L. Eriksson, E. Johansson, ##' N. Kettaneh-Wold and S. Wold, Umetrics 1999, p. 466 ##' @examples ##' data(iris) ##' pcIr <- pca(iris[,1:4]) ##' ## versicolor has the lowest leverage ##' with(iris, plot(leverage(pcIr)~Species)) ##' @keywords multivariate ##' @aliases leverage leverage,pcaRes-method ##' @author Henning Redestig setMethod("leverage", "pcaRes", function(object) { diag(scores(object) %*% solve(crossprod(scores(object))) %*% t(scores(object))) }) ##' Distance to the model of X-space. ##' ##' Measures how well described the observations are, i.e. how well ##' they fit in the mode. High DModX indicate a poor fit. Defined as: ##' ##' \eqn{\frac{\sqrt{\frac{SSE_i}{K-A}}}{\sqrt{\frac{SSE}{(N-A-A_0)(K-A)}}}} ##' ##' For observation \eqn{i}, in a model with \eqn{A} components, ##' \eqn{K} variables and \eqn{N} obserations. SSE is the squared sum ##' of the residuals. \eqn{A_0} is 1 if model was centered and 0 ##' otherwise. DModX is claimed to be approximately F-distributed and ##' can therefore be used to check if an observation is significantly ##' far away from the PCA model assuming normally distributed data. ##' ##' Pass original data as an argument if the model was calculated with ##' \code{completeObs=FALSE}. ##' @title DModX ##' @usage DModX(object, dat, newdata=FALSE, type=c("normalized","absolute"), ...) ##' @param object a pcaRes object ##' @param dat the original data, taken from \code{completeObs} if ##' left missing. ##' @param newdata logical indicating if this data was part of the ##' training data or not. If it was, it is adjusted by a near one factor ##' \eqn{v=(N/ (N-A-A0))^-1} ##' @param type if absolute or normalized values should be ##' given. Normalized values are adjusted to the the total RSD of the ##' model. ##' @param ... Not used ##' @return A vector with distances from observations to the PCA model ##' @aliases DModX DModX,pcaRes-method ##' @examples ##' data(iris) ##' pcIr <- pca(iris[,1:4]) ##' with(iris, plot(DModX(pcIr)~Species)) ##' @references Introduction to Multi- and Megavariate Data Analysis ##' using Projection Methods (PCA and PLS), L. Eriksson, E. Johansson, ##' N. Kettaneh-Wold and S. Wold, Umetrics 1999, p. 468 ##' @author Henning Redestig setMethod("DModX", "pcaRes", function(object, dat, newdata=FALSE, type=c("normalized","absolute"), ...) { type <- match.arg(type) if(missing(dat)) { if(!is.null(completeObs(object))) dat <- completeObs(object) else stop("missing data when calculating DModX") } A0 <- as.integer(centered(object)) ny <- ifelse(newdata, 1, sqrt(nObs(object) / (nObs(object) - nP(object) - A0))) E2 <- resid(object, dat)^2 s <- sqrt(rowSums(E2) / (nVar(object) - nP(object))) * ny if(type == "absolute") return(s) s0 <- sqrt(sum(E2) / ((nObs(object) - nP(object) - A0) * (nVar(object) - nP(object)))) s / s0 }) ##' Get number of PCs ##' @param object pcaRes object ##' @param ... not used ##' @return Number of PCs ##' @aliases nP nP,pcaRes-method ##' @usage nP(object, ...) ##' @author Henning Redestig setMethod("nP", "pcaRes", function(object, ...) { if(is.null(object@nPcs) & !is.null(scores(object))) return(ncol(scores(object))) object@nPcs }) ##' Get cross-validation statistics (e.g. \eqn{Q^2}). ##' @param object pcaRes object ##' @param ... not used ##' @return vector CV statistics ##' @aliases cvstat cvstat,pcaRes-method ##' @usage cvstat(object, ...) ##' @author Henning Redestig setMethod("cvstat", "pcaRes", function(object, ...) { object@cvstat }) ##' Get number of PCs. ##' @param object pcaRes object ##' @param ... not used ##' @note Try to use \code{link{nP}} instead since \code{nPcs} tend to ##' clash with argument names. ##' @return Number of PCs ##' @usage nPcs(object, ...) ##' @aliases nPcs nPcs,pcaRes-method ##' @author Henning Redestig setMethod("nPcs", "pcaRes", function(object, ...) { nP(object) }) ##' Get the number of observations used to build the PCA model. ##' @param object pcaRes object ##' @param ... Not used ##' @usage nObs(object, ...) ##' @aliases nObs nObs,pcaRes-method ##' @return Number of observations ##' @author Henning Redestig setMethod("nObs", "pcaRes", function(object, ...) { object@nObs }) ##' Get the number of variables used to build the PCA model. ##' @param object pcaRes object ##' @param ... Not used ##' @usage nVar(object, ...) ##' @aliases nVar nVar,pcaRes-method ##' @return Number of variables ##' @author Henning Redestig setMethod("nVar", "pcaRes", function(object, ...) { object@nVar }) ##' Check centering was part of the model ##' @param object pcaRes object ##' @param ... Not used ##' @usage centered(object, ...) ##' @aliases centered centered,pcaRes-method ##' @return TRUE if model was centered ##' @author Henning Redestig setMethod("centered", "pcaRes", function(object, ...) { if(is.null(object@centered)) return(FALSE) object@centered }) ##' Get the centers of the original variables ##' @param object pcaRes object ##' @param ... Not used ##' @usage center(object, ...) ##' @aliases center center,pcaRes-method ##' @return Vector with the centers ##' @author Henning Redestig setMethod("center", "pcaRes", function(object, ...) { object@center }) setMethod("completeObs", "pcaRes", function(object, ...) { object@completeObs }) ##' Get the original data with missing values replaced with predicted ##' values. ##' @param object object to fetch complete data from ##' @param ... Not used ##' @usage completeObs(object, ...) ##' @aliases completeObs completeObs,nniRes-method ##' completeObs,pcaRes-method ##' @return Completed data (matrix) ##' @author Henning Redestig setMethod("completeObs", "nniRes", function(object, ...) { object@completeObs }) ##' Get the used PCA method ##' @param object pcaRes object ##' @param ... Not used ##' @usage method(object, ...) ##' @aliases method method,pcaRes-method ##' @return The used pca method ##' @author Henning Redestig setMethod("method", "pcaRes", function(object, ...) { object@method }) setMethod("nmissing", "nniRes", function(object, ...) { sum(object@missing) }) ##' Missing values ##' @param object pcaRes object ##' @param ... Not used ##' @usage nmissing(object, ...) ##' @aliases nmissing nmissing,pcaRes-method nmissing,nniRes-method ##' @return Get the number of missing values ##' @author Henning Redestig setMethod("nmissing", "pcaRes", function(object, ...) { sum(object@missing) }) ##' Get a matrix with indicating the elements that were missing in the ##' input data. Convenient for estimating imputation performance. ##' @param object pcaRes object ##' @param ... Not used ##' @usage wasna(object, ...) ##' @aliases wasna wasna,pcaRes-method ##' @return A matrix with logicals ##' @examples ##' data(metaboliteData) ##' data(metaboliteDataComplete) ##' result <- pca(metaboliteData, nPcs=2) ##' plot(completeObs(result)[wasna(result)], metaboliteDataComplete[wasna(result)]) ##' @author Henning Redestig setMethod("wasna", "pcaRes", function(object, ...) { object@missing }) ##' Get the standard deviations of the scores (indicates their ##' relevance) ##' @param object pcaRes object ##' @param ... Not used ##' @usage sDev(object, ...) ##' @aliases sDev sDev,pcaRes-method ##' @return Standard devations of the scores ##' @author Henning Redestig setMethod("sDev", "pcaRes", function(object, ...) { object@sDev }) ##' Check if scaling was part of the PCA model ##' @param object pcaRes object ##' @param ... Not used ##' @usage scaled(object, ...) ##' @aliases scaled scaled,pcaRes-method ##' @return TRUE if scaling was part of the PCA model ##' @author Henning Redestig setMethod("scaled", "pcaRes", function(object, ...) { if(is.null(object@scaled)) return(FALSE) object@scaled != "none" }) ##' Get the scales (e.g. standard deviations) of the original ##' variables ##' @param object pcaRes object ##' @param ... Not used ##' @usage scl(object, ...) ##' @aliases scl scl,pcaRes-method ##' @return Vector with the scales ##' @seealso \code{\link{prep}} ##' @author Henning Redestig setMethod("scl", "pcaRes", function(object, ...) { object@scale }) ##' Cumulative R2 is the total ratio of variance that is being ##' explained by the model ##' @param object a \code{pcaRes} model ##' @param ... Not used ##' @return Get the cumulative R2 ##' @aliases R2cum R2cum,pcaRes-method ##' @author Henning Redestig setMethod("R2cum", "pcaRes", function(object, ...) { object@R2cum }) ##' Get scores from a pcaRes object ##' @param object a pcaRes object ##' @param ... not used ##' @return The scores as a matrix ##' @export ##' @author Henning Redestig ##' @method scores pcaRes scores.pcaRes <- function(object, ...) object@scores ##' Get scores from a pcaRes object ##' @param object a pcaRes object ##' @param ... not used ##' @return The scores as a matrix ##' @seealso \code{\link{scores.pcaRes}} ##' @aliases scores scores,pcaRes-method ##' @author Henning Redestig setMethod("scores", "pcaRes", scores.pcaRes) ##' Get loadings from a pcaRes object ##' @param object a pcaRes object ##' @param ... not used ##' @return The loadings as a matrix ##' @export ##' @author Henning Redestig ##' @method loadings pcaRes loadings.pcaRes <- function(object, ...) object@loadings ##' Get loadings from a pcaRes object ##' @param object a pcaRes object ##' @param ... not used ##' @return The loadings as a matrix ##' @seealso \code{\link{loadings.pcaRes}} ##' @author Henning Redestig ##' @aliases loadings,pcaRes-method setMethod("loadings", "pcaRes", loadings.pcaRes) ##' Crude way to unmask the function with the same name from ##' \code{stats} ##' @param object any object ##' @param ... not used ##' @return The loadings ##' @author Henning Redestig ##' @aliases loadings loadings,ANY-method setMethod("loadings", "ANY", function(object,...) { stats::loadings(object) }) ##' Dimensions of a PCA model ##' @param x a pcaRes object ##' @return Get the dimensions of this PCA model ##' @method dim pcaRes ##' @export ##' @author Henning Redestig dim.pcaRes <- function(x) { res <- c(nObs(x), nVar(x), nP(x)) names(res) <- c("nObs", "nVar", "nPcs") res } ##' Print basic information about pcaRes object ##' @title Print/Show for pcaRes ##' @param x a pcaRes object ##' @param ... not used ##' @return nothing, used for its side effect ##' @name show-methods ##' @export ##' @author Henning Redestig showPcaRes <- function(x, ...) { summary(x) cat(nVar(x), "\tVariables\n") cat(nObs(x),"\tSamples\n") cat(nmissing(x), "\tNAs (", round(100 * nmissing(x) / (nObs(x) * nVar(x)), getOption("str")$digits.d), "%)\n") cat(nP(x), "\tCalculated component(s)\n") if(centered(x)) cat("Data was mean centered before running PCA \n") else cat("Data was NOT mean centered before running PCA \n") if(scaled(x)) cat("Data was scaled before running PCA \n") else cat("Data was NOT scaled before running PCA \n") cat("Scores structure:\n") print(dim(scores(x))) cat("Loadings structure:\n") if(method(x) == "nlpca") { cat("Inverse hierarchical neural network architecture\n") cat(drop(x@network@net), "\n") cat("Functions in layers\n") cat(x@network@fct, "\n") cat("hierarchic layer:", x@network@hierarchic$layer, "\n") cat("hierarchic coefficients:", x@network@hierarchic$var, "\n") cat("scaling factor:", x@network@scalingFactor, "\n") } else{ print(dim(loadings(x))) } } ##' @aliases print,pcaRes-method print,nniRes-method ##' @name show-methods setMethod("print", "pcaRes", showPcaRes) ## @importFrom methods show ##' @aliases show,pcaRes-method show,nniRes-method ##' @param object the object to print information about ##' @name show-methods setMethod("show", "pcaRes", function(object) showPcaRes(object)) ##' Visualize two-components simultaneously ##' ##' This is a method for the generic function 'biplot'. There is ##' considerable confusion over the precise definitions: those of the ##' original paper, Gabriel (1971), are followed here. Gabriel and ##' Odoroff (1990) use the same definitions, but their plots actually ##' correspond to \code{pc.biplot = TRUE}. ##' @title Plot a overlaid scores and loadings plot ##' @param x a pcaRes object ##' @param choices which two pcs to plot ##' @param scale The variables are scaled by ##' \eqn{\lambda^{scale}}{lambda^scale} and the observations are ##' scaled by \eqn{\lambda^{scale}}{lambda ^ (1-scale)} where ##' \code{lambda} are the singular values as computed by ##' \code{princomp}. Normally \eqn{0\le{}scale\le{}1}{0 <= scale <= ##' 1}, and a warning will be issued if the specified 'scale' is ##' outside this range. ##' @param pc.biplot If true, use what Gabriel (1971) refers to as a ##' "principal component biplot", with \eqn{\lambda=1}{lambda = 1} and ##' observations scaled up by sqrt(n) and variables scaled down by ##' sqrt(n). Then the inner products between variables approximate ##' covariances and distances between observations approximate ##' Mahalanobis distance. ##' @param ... optional arguments to be passed to ##' \code{biplot.default}. ##' @return a plot is produced on the current graphics device. ##' @method biplot pcaRes ##' @export ##' @examples ##' data(iris) ##' pcIr <- pca(iris[,1:4]) ##' biplot(pcIr) ##' @seealso \code{prcomp}, \code{pca}, \code{princomp} ##' @author Kevin Wright, Adapted from \code{biplot.prcomp} ##' @keywords multivariate ##' @name biplot-methods biplot.pcaRes <- function(x, choices=1:2, scale=1, pc.biplot=FALSE, ...) { if(length(choices)!=2) stop("length of choices must be 2") scores <- scores(x) n <- nrow(scores) lam <- sDev(x)[choices] * sqrt(n) if(scale < 0 || scale > 1) warning("'scale' is outside [0,1]") if(scale != 0) lam <- lam^scale else lam <- 1 if(pc.biplot) lam <- lam/sqrt(n) biplot(t(t(scores[,choices])/lam), t(t(loadings(x)[, choices]) * lam), , ...) invisible() } ##' @aliases biplot,pcaRes-method ## @importFrom stats biplot ##' @name biplot-methods setMethod("biplot", "pcaRes", biplot.pcaRes) ##' Flexible calculation of R2 goodness of fit. ##' @title R2 goodness of fit ##' @param object a PCA model object ##' @param direction choose between calculating R2 per variable, per ##' observation or for the entire data with 'variables', ##' 'observations' or 'complete'. ##' @param data the data used to fit the model ##' @param pcs the number of PCs to use to calculate R2 ##' @aliases R2VX R2VX,pcaRes-method ##' @examples ##' R2VX(pca(iris)) ##' @return A vector with R2 values ##' @author Henning Redestig setMethod('R2VX', 'pcaRes', function(object, direction=c('variables', 'observations', 'complete'), data=completeObs(object), pcs=nP(object)) { direction <- match.arg(direction) if(is.null(data)) stop('missing input when calculating R2') if(any(is.na(data))) stop('missing values not allowed for calculating R2') dat <- prep(data, scale=scl(object), center=center(object)) xhat <- resid(object, pcs=pcs, data=dat, pre=FALSE, post=FALSE) switch(direction, variables={ 1 - colSums(xhat^2) / colSums(dat^2) }, observations={ 1 - rowSums(xhat^2) / rowSums(dat^2) }, complete={ 1 - sum(xhat^2) / sum(dat^2) }) }) setAs('pcaRes', 'data.frame', function(from) { tt <- scores(from) pp <- loadings(from) if(is.null(rownames(tt))) rownames(tt) <- 1:nrow(tt) if(is.null(rownames(pp))) rownames(pp) <- 1:nrow(pp) dfs <- as.data.frame(tt) dfs$names <- rownames(tt) dfs$type <- 'scores' dfl <- as.data.frame(pp) dfl$names <- rownames(pp) dfl$type <- 'loadings' rownames(dfl) <- rownames(dfs) <- NULL rbind(dfl, dfs) }) ##' Print a brief description of the PCA model ##' @title Summary of PCA model ##' @param object a pcaRes object ##' @param ... Not used ##' @return Nothing, used for side-effect ##' @aliases summary summary.pcaRes summary,pcaRes-method ##' @author Henning Redestig ##' @export ##' @name summary ##' @method summary pcaRes summary.pcaRes <- function(object, ...){ cat(method(object), "calculated PCA\n") cat("Importance of component(s):\n") prop <- vector(length=length(R2cum(object)), mode="numeric") prop[1] <- R2cum(object)[1] if (length(R2cum(object)) > 1) { for (i in 2:length(prop)) { prop[i] <- R2cum(object)[i] - R2cum(object)[i-1] } } r <- rbind(prop, R2cum(object)) rownames(r) <- c("R2", "Cumulative R2") colnames(r) <- paste("PC", 1:nP(object), sep="") print(r, digits=4) invisible(r) } setMethod("summary", "pcaRes", summary.pcaRes) ##' Predict data using PCA model ##' ##' This function extracts the predict values from a pcaRes object for ##' the PCA methods SVD, Nipals, PPCA and BPCA. Newdata is first ##' centered if the PCA model was and then scores (\eqn{T}) and data ##' (\eqn{X}) is 'predicted' according to : ##' \eqn{\hat{T}=X_{new}P}{That=XnewP} ##' \eqn{\hat{X}_{new}=\hat{T}P'}{Xhat=ThatP'}. Missing values are ##' set to zero before matrix multiplication to achieve NIPALS like ##' treatment of missing values. ##' @title Predict values from PCA. ##' @param object \code{pcaRes} the \code{pcaRes} object of interest. ##' @param newdata \code{matrix} new data with same number of columns ##' as the used to compute \code{object}. ##' @param pcs \code{numeric} The number of PC's to consider ##' @param pre pre-process \code{newdata} based on the pre-processing ##' chosen for the PCA model ##' @param post unpre-process the final data (add the center back etc) ##' @param ... Not passed on anywhere, included for S3 consistency. ##' @return A list with the following components: \item{scores}{The ##' predicted scores} \item{x}{The predicted data} ##' @method predict pcaRes ##' @keywords multivariate ##' @examples ##' data(iris) ##' hidden <- sample(nrow(iris), 50) ##' pcIr <- pca(iris[-hidden,1:4]) ##' pcFull <- pca(iris[,1:4]) ##' irisHat <- predict(pcIr, iris[hidden,1:4]) ##' cor(irisHat$scores[,1], scores(pcFull)[hidden,1]) ##' @export ##' @name predict-methods ##' @author Henning Redestig predict.pcaRes <- function(object, newdata, pcs=nP(object), pre=TRUE, post=TRUE, ...) { if(!method(object) %in% listPcaMethods("linear")) stop("predict method not implemented for that type of PCA") if(pre) newdata <- prep(newdata, scl(object), center(object)) ## set na's to zero to achieve NIPALS like prediction newdata[is.na(newdata)] <- 0 tnew <- newdata %*% loadings(object)[,1:pcs,drop=FALSE] xhat <- tcrossprod(tnew, loadings(object)[,1:pcs,drop=FALSE]) if(post) xhat <- prep(xhat, scl(object), center(object), reverse=TRUE) list(scores=tnew, x=xhat) } ## @importFrom stats predict ##' @name predict-methods ##' @aliases predict,pcaRes-method setMethod("predict", "pcaRes", predict.pcaRes) ##' This function extracts the residuals values from a pcaRes object ##' for the PCA methods SVD, Nipals, PPCA and BPCA ##' @title Residuals values from a PCA model. ##' @param object \code{pcaRes} the \code{pcaRes} object of interest. ##' @param data \code{matrix} The data that was used to calculate the ##' PCA model (or a different dataset to e.g. adress its proximity to ##' the model). ##' @param ... Passed on to \code{\link{predict.pcaRes}}. E.g. setting ##' the number of used components. ##' @return A \code{matrix} with the residuals ##' @method residuals pcaRes ##' @keywords multivariate ##' @export ##' @name rediduals-methods ##' @examples ##' data(iris) ##' pcIr <- pca(iris[,1:4]) ##' head(residuals(pcIr, iris[,1:4])) ##' @author Henning Redestig residuals.pcaRes <- function(object, data=completeObs(object), ...) { if(is.null(data)) stop("data missing when calculating residuals") data - predict(object, data, ...)$x } ##' @aliases residuals,pcaRes-method ##' @name rediduals-methods setMethod("residuals", "pcaRes", residuals.pcaRes) ##' @name rediduals-methods ##' @aliases resid,pcaRes-method setMethod("resid", "pcaRes", residuals.pcaRes) ##' Fitted values of a PCA model ##' ##' This function extracts the fitted values from a pcaResobject. For ##' PCA methods like SVD, Nipals, PPCA etc this is basically just the ##' scores multipled by the loadings and adjusted for pre-processing. ##' for non-linear PCA the original data is propagated through the ##' network to obtain the approximated data. ##' @title Extract fitted values from PCA. ##' @param object the \code{pcaRes} object of interest. ##' @param data For standard PCA methods this can safely be left null ##' to get scores x loadings but if set, then the scores are obtained ##' by projecting provided data onto the loadings. If data contains ##' missing values the result will be all NA. Non-linear PCA is an ##' exception, here if data is NULL then data is set to the ##' completeObs and propaged through the network. ##' @param nPcs The number of PC's to consider ##' @param pre pre-process \code{data} based on the pre-processing ##' chosen for the PCA model ##' @param post unpre-process the final data (add the center back etc ##' to get the final estimate) ##' @param ... Not used ##' @return A matrix representing the fitted data ##' @keywords multivariate ##' @method fitted pcaRes ##' @examples ##' pc <- pca(iris[,1:4], nPcs=4, center=TRUE, scale="uv") ##' sum( (fitted(pc) - iris[,1:4])^2 ) ##' @export ##' @name fitted-methods ##' @author Henning Redestig fitted.pcaRes <- function(object, data=NULL, nPcs=nP(object), pre=TRUE, post=TRUE, ...) { if(method(object) %in% listPcaMethods("nonlinear")) { if(is.null(data) & is.null(completeObs(object))) stop("completeObs slot is empty -- provide the training data") if(is.null(data) & !is.null(completeObs(object))) data <- completeObs(object) if(is.null(data)) stop("nlpca requires original data to be provide") if(pre) data <- prep(data, scl(object), center(object)) recData <- errorHierarchic(object@network, t(scores(object)), t(data))$out[,,nPcs] recData <- t(recData / object@network@scalingFactor) } else { if(!is.null(data)) { if(pre) data <- prep(data, scl(object), center(object)) tt <- data %*% loadings(object)[,1:nPcs, drop=FALSE] } if(is.null(data)) tt <- scores(object)[,1:nPcs, drop=FALSE] recData <- tcrossprod(tt, loadings(object)[,1:nPcs, drop=FALSE]) } if(post) recData <- prep(recData, scl(object), center(object), reverse=TRUE) return(recData) } ## @importFrom stats fitted ##' @name fitted-methods ##' @aliases fitted,pcaRes-method setMethod("fitted", "pcaRes", fitted.pcaRes) ##' Plot the computed diagnostics of PCA model to get an idea of their ##' importance. Note though that the standard screeplot shows the ##' standard deviations for the PCs this method shows the R2 values ##' which empirically shows the importance of the P's and is thus ##' applicable for any PCA method rather than just SVD based PCA. ##' ##' If cross-validation was done for the PCA the plot will also show ##' the CV based statistics. A common rule-of-thumb for determining ##' the optimal number of PCs is the PC where the CV diagnostic is at ##' its maximum but not very far from \eqn{R^2}. ##' @title Plot diagnostics (screeplot) ##' @param x \code{pcaRes} The pcaRes object. ##' @param y not used ##' @param main title of the plot ##' @param col Colors of the bars ##' @param ... further arguments to barplot ##' @return None, used for side effect. ##' @seealso \link{screeplot} ##' @examples ##' data(metaboliteData) ##' pc <- pca(t(metaboliteData), nPcs=5, cv="q2", scale="uv") ##' plot(pc) ##' @method plot pcaRes ##' @aliases plot.pcaRes plot,pcaRes-method ##' @export ##' @author Henning Redestig plot.pcaRes <- function(x, y=NULL, main=deparse(substitute(object)), col=gray(c(0.9, 0.5)), ...) { y <- NULL ## the deparse(subsitute(object)) later fails otherwise main <- main if(!is.null(cvstat(x))) { cvs <- cvstat(x) if(length(cvs) != nP(x)) cvs <- c(cvs, rep(NA, nP(x) - length(cvs))) xx <- rbind(R2cum(x), cvs) barplot(xx, beside=TRUE, ylim=c(0,1.1), col=col, main=main, names.arg=paste("PC", 1:nP(x), sep=""), ...) legend(x="topleft", fill=col, legend=c(expression(R^2), expression(Q^2))) } else barplot(R2cum(x), ylim=c(0,1.1), ylab=expression(R^2), main=main, names.arg=paste("PC", 1:nP(x), sep=""), col=col[1], ...) } setMethod("plot", "pcaRes", plot.pcaRes) ##' A common way of visualizing two principal components ##' ##' This method is meant to be used as a quick way to visualize ##' results, if you want a more specific plot you probably want to ##' get the scores, loadings with \code{scores(object)}, ##' \code{loadings(object)} and then design your own plotting method. ##' @title Side by side scores and loadings plot ##' @usage slplot(object, pcs=c(1,2), scoresLoadings=c(TRUE, TRUE), ##' sl="def", ll="def", hotelling=0.95, rug=TRUE, sub=NULL,...) ##' @param object a pcaRes object ##' @param pcs which two pcs to plot ##' @param scoresLoadings Which should be shown scores and or loadings ##' @param sl labels to plot in the scores plot ##' @param ll labels to plot in the loadings plot ##' @param hotelling confidence interval for ellipse in the score plot ##' @param rug logical, rug x axis in score plot or not ##' @param sub Subtitle, defaults to annotate with amount of explained ##' variance. ##' @param ... Further arguments to plot functions. Prefix arguments ##' to \code{par()} with 's' for the scores plot and 'l' for the ##' loadings plot. I.e. cex become scex for setting character ##' expansion in the score plot and lcex for the loadings plot. ##' @return None, used for side effect. ##' @note Uses layout instead of par to provide side-by-side so it ##' works with Sweave (but can not be combined with ##' \code{par(mfrow=..))} ##' @author Henning Redestig ##' @seealso \code{\link{pca}}, \code{\link{biplot}} ##' @aliases slplot slplot,pcaRes-method ##' @examples ##' data(iris) ##' pcIr <- pca(iris[,1:4], scale="uv") ##' slplot(pcIr, sl=NULL, spch=5) ##' slplot(pcIr, sl=NULL, lcex=1.3, scol=as.integer(iris[,5])) ##' @keywords multivariate setMethod("slplot", "pcaRes", function(object, pcs=c(1,2), scoresLoadings=c(TRUE, TRUE), sl=rownames(scores(object)), ll=rownames(loadings(object)), hotelling=0.95, rug=FALSE, sub=NULL,...) { opar <- par(no.readonly=TRUE) cl <- match.call() mainArgs <- c(1,match(c("ll", "sl", "scoresLoadings", "sub"), names(cl), 0)) scoreArgs <- grep("^s", names(cl)[-mainArgs]) loadingArgs <- grep("^l", names(cl)[-mainArgs]) if(!is.null(ll) & length(ll) != nVar(object)) stop("Loading labels do not match the object dimensions") if(!is.null(sl) & length(sl) != nObs(object)) stop("Score labels do not match the object dimensions") if(is.null(sl)) sl <- NA if(is.null(ll)) ll <- NA ## no loadings for non-linear pca if(method(object) %in% listPcaMethods("nonlinear") && scoresLoadings[2]) scoresLoadings[2] <- FALSE if(length(pcs) > 2) plotPcs(object, pcs, scoresLoadings=scoresLoadings,...) else { if(is.null(sub)) sub <- paste(sprintf("%.2f", R2cum(object)[max(pcs)] * 100), "% of the variance explained", sep="") if(sum(scoresLoadings) == 2) layout(matrix(c(1,2), 1, 2, TRUE), respect=matrix(c(1,1), 1, 2)) ## exception plot if one dimensional if (length(pcs) == 1 | nP(object) == 1) { pcs <- 1 ## score plot if(scoresLoadings[1]) { newCall <- call("barplot", height=scores(object)[,pcs], main="Scores", las=3, ylab=paste("PC", pcs), sub=sub, names.arg=sl) tmp <- cl[-mainArgs][scoreArgs] names(tmp) <- gsub("^s", "", names(tmp)) for(i in 1:length(tmp)) { newCall[[length(newCall) + 1]] <- tmp[[i]] names(newCall)[length(newCall)] <- names(tmp)[i] } eval(newCall) } ## loadingplot if(scoresLoadings[2]) { newCall <- call("barplot", height=loadings(object)[,pcs], main="Loadings", las=3, ylab=paste("PC", pcs), names.arg=ll) if(length(loadingArgs) > 0) { tmp <- cl[-mainArgs][loadingArgs] names(tmp) <- gsub("^l", "", names(tmp)) for(i in 1:length(tmp)) { newCall[[length(newCall) + 1]] <- tmp[[i]] names(newCall)[length(newCall)] <- names(tmp)[i] } } eval(newCall) } return(invisible(TRUE)) } ## the score plot if(scoresLoadings[1]) { ## setup plot plotCall <- call("plot", x=scores(object)[,pcs], main="Scores", ylab=paste("PC", pcs[2]), sub=sub, xlab=paste("PC", pcs[1])) if(length(scoreArgs) > 0) { tmp <- cl[-mainArgs][scoreArgs] names(tmp) <- gsub("^s", "", names(tmp)) for(i in 1:length(tmp)) { plotCall[[length(plotCall) + 1]] <- tmp[[i]] names(plotCall)[length(plotCall)] <- names(tmp)[i] } } ## add text if (!is.null(sl) & !all(is.na(sl))) { plotCall[[length(plotCall) + 1]] <- "n" names(plotCall)[length(plotCall)] <- "type" textCall <- call("text", x=scores(object)[,pcs], labels=sl) if(length(scoreArgs) > 0) { tmp <- cl[-mainArgs][scoreArgs] names(tmp) <- gsub("^s", "", names(tmp)) for(i in 1:length(tmp)) { textCall[[length(textCall) + 1]] <- tmp[[i]] names(textCall)[length(textCall)] <- names(tmp)[i] } } } eval(plotCall) if (!is.null(sl) & !all(is.na(sl))) eval(textCall) if(rug) rug(scores(object)[,1]) abline(h=0, v=0) if(!is.null(hotelling)) { A <- length(pcs) el <- simpleEllipse(scores(object)[,pcs[1]], scores(object)[,pcs[2]], alfa=hotelling) lines(el) } } ## the loading plot if(scoresLoadings[2]) { ## setup plot plotCall <- call("plot", x=loadings(object)[,pcs], main="Loadings", ylab=paste("PC", pcs[2]), xlab=paste("PC", pcs[1])) if(length(loadingArgs) > 0) { tmp <- cl[-mainArgs][loadingArgs] names(tmp) <- gsub("^l", "", names(tmp)) for(i in 1:length(tmp)) { plotCall[[length(plotCall) + 1]] <- tmp[[i]] names(plotCall)[length(plotCall)] <- names(tmp)[i] } } ## add text if (!is.null(ll) & !all(is.na(ll))) { plotCall[[length(plotCall) + 1]] <- "n" names(plotCall)[length(plotCall)] <- "type" textCall <- call("text", x=loadings(object)[,pcs], labels=ll) if(length(loadingArgs) > 0) { tmp <- cl[-mainArgs][loadingArgs] names(tmp) <- gsub("^l", "", names(tmp)) for(i in 1:length(tmp)) { textCall[[length(textCall) + 1]] <- tmp[[i]] names(textCall)[length(textCall)] <- names(tmp)[i] } } } eval(plotCall) if (!is.null(ll) & !all(is.na(ll))) eval(textCall) abline(h=0, v=0) } } par(opar) }) pcaMethods/R/nipalsPca.R0000644000175200017520000001513614710217306016130 0ustar00biocbuildbiocbuild##' PCA by non-linear iterative partial least squares ##' ##' Can be used for computing PCA on a numeric matrix using either the ##' NIPALS algorithm which is an iterative approach for estimating the ##' principal components extracting them one at a time. NIPALS can ##' handle a small amount of missing values. It is not recommended to ##' use this function directely but rather to use the pca() wrapper ##' function. There is a C++ implementation given as \code{nipalsPca} ##' which is faster. ##' @title NIPALS PCA implemented in R ##' @param Matrix Pre-processed (centered, scaled) numerical matrix ##' samples in rows and variables as columns. ##' @param nPcs Number of components that should be extracted. ##' @param varLimit Optionally the ratio of variance that should be ##' explained. \code{nPcs} is ignored if varLimit < 1 ##' @param maxSteps Defines how many iterations can be done before ##' algorithm should abort (happens almost exclusively when there were ##' some wrong in the input data). ##' @param threshold The limit condition for judging if the algorithm ##' has converged or not, specifically if a new iteration is done if ##' \eqn{(T_{old} - T)^T(T_{old} - T) > \code{limit}}. ##' @param verbose Show simple progress information. ##' @param ... Only used for passing through arguments. ##' @return A \code{pcaRes} object. ##' @references Wold, H. (1966) Estimation of principal components and ##' related models by iterative least squares. In Multivariate ##' Analysis (Ed., P.R. Krishnaiah), Academic Press, NY, 391-420. ##' @author Henning Redestig ##' @seealso \code{prcomp}, \code{princomp}, \code{pca} ##' @examples ##' data(metaboliteData) ##' mat <- prep(t(metaboliteData)) ##' ## c++ version is faster ##' system.time(pc <- RnipalsPca(mat, method="rnipals", nPcs=2)) ##' system.time(pc <- nipalsPca(mat, nPcs=2)) ##' ## better use pca() ##' pc <- pca(t(metaboliteData), method="rnipals", nPcs=2) ##' \dontshow{stopifnot(sum((fitted(pc) - t(metaboliteData))^2, na.rm=TRUE) < 200)} ##' @keywords multivariate ##' @export RnipalsPca <- function(Matrix, nPcs=2, varLimit=1, maxSteps=5000, threshold=1e-6, verbose=interactive(), ...) { nVar <- ncol(Matrix) ##Find a good? starting column -- better way? startingColumn <- 1 ## sum(c(NA, NA), na.rm=TRUE) is 0, but we want NA sum.na <- function(x){ ifelse(all(is.na(x)), NA, sum(x, na.rm=TRUE))} TotalSS <- sum(Matrix*Matrix, na.rm=TRUE) ph <- rep(0, nVar) R2cum <- rep(NA, nPcs) scores <- NULL loadings <- NULL anotherPc <- TRUE l <- 1 while(anotherPc) { count <- 0 #number of iterations done th <- Matrix[,startingColumn] #first column is starting vector for th continue <- TRUE if(verbose) cat(paste("Calculating PC", l, ": ", sep="")) while(continue) { count <- count+1 ph <- rep(0, nVar) ##Calculate loadings through LS regression ##Note: Matrix*th is column-wise multiplication tsize <- sum(th * th, na.rm=TRUE) ph <- apply(Matrix * (th / tsize), 2, sum.na) ##normalize ph based on the available values. psize <- sum(ph*ph, na.rm=TRUE) ph <- ph / sqrt(psize) ##Calculate scores through LS regression ##Trick: To get row-wise multiplication, use t(Matrix)*ph, then ##be sure to use apply(,2,) and NOT apply(,1,)! th.old <- th th <- apply(t(Matrix) * ph, 2, sum.na) ##Round up by calculating if convergence condition is met and ##checking if it seems to be an neverending loop. if (count > maxSteps) { stop("Too many iterations, quitting") } if (t(na.omit(th.old - th)) %*% (na.omit(th.old - th)) <= threshold) { continue = FALSE } if (verbose)cat("*") } if (verbose) cat(" Done\n") Matrix <- Matrix - (th %*% t(ph)) scores <- cbind(scores, th) loadings <- cbind(loadings, ph) ##cumulative proportion of variance R2cum[l] <- 1 - (sum(Matrix*Matrix,na.rm=TRUE) / TotalSS) l <- l + 1 if((!abs(varLimit - 1) < 1e-4 & R2cum[l - 1] >= varLimit) | l > nPcs) { anotherPc <- FALSE nPcs <- l - 1 } } r <- new("pcaRes") r@scores <- scores r@loadings <- loadings r@R2cum <- R2cum r@varLimit <- varLimit r@method <- "rnipals" return(r) } ##' PCA by non-linear iterative partial least squares ##' ##' Can be used for computing PCA on a numeric matrix using either the ##' NIPALS algorithm which is an iterative approach for estimating the ##' principal components extracting them one at a time. NIPALS can ##' handle a small amount of missing values. It is not recommended to ##' use this function directely but rather to use the pca() wrapper ##' function. ##' @title NIPALS PCA ##' @param Matrix Pre-processed (centered, scaled) numerical matrix ##' samples in rows and variables as columns. ##' @param nPcs Number of components that should be extracted. ##' @param varLimit Optionally the ratio of variance that should be ##' explained. \code{nPcs} is ignored if varLimit < 1 ##' @param maxSteps Defines how many iterations can be done before ##' algorithm should abort (happens almost exclusively when there were ##' some wrong in the input data). ##' @param threshold The limit condition for judging if the algorithm ##' has converged or not, specifically if a new iteration is done if ##' \eqn{(T_{old} - T)^T(T_{old} - T) > \code{limit}}. ##' @param ... Only used for passing through arguments. ##' @return A \code{pcaRes} object. ##' @references Wold, H. (1966) Estimation of principal components and ##' related models by iterative least squares. In Multivariate ##' Analysis (Ed., P.R. Krishnaiah), Academic Press, NY, 391-420. ##' @author Henning Redestig ##' @seealso \code{prcomp}, \code{princomp}, \code{pca} ##' @examples ##' data(metaboliteData) ##' mat <- prep(t(metaboliteData)) ##' pc <- nipalsPca(mat, nPcs=2) ##' ## better use pca() ##' pc <- pca(t(metaboliteData), method="nipals", nPcs=2) ##' \dontshow{stopifnot(sum((fitted(pc) - t(metaboliteData))^2, na.rm=TRUE) < 200)} ##' @keywords multivariate ##' @export nipalsPca <- function(Matrix, nPcs=2, varLimit=1, maxSteps=5000, threshold=1e-6, ...) { nipRes <- .Call("pcaMethods_Nipals", Matrix, params=list(nPcs=nPcs, varLimit=varLimit, threshold=threshold, maxSteps=maxSteps), PACKAGE="pcaMethods") r <- new("pcaRes") r@scores <- nipRes$scores r@loadings <- nipRes$loadings r@R2cum <- nipRes$R2cum r@varLimit <- varLimit r@method <- "nipals" return(r) } pcaMethods/R/nlpca.R0000644000175200017520000001664414710217306015320 0ustar00biocbuildbiocbuild##' Neural network based non-linear PCA ##' ##' Artificial Neural Network (MLP) for performing non-linear ##' PCA. Non-linear PCA is conceptually similar to classical PCA but ##' theoretically quite different. Instead of simply decomposing our ##' matrix (X) to scores (T) loadings (P) and an error (E) we train a ##' neural network (our loadings) to find a curve through the ##' multidimensional space of X that describes a much variance as ##' possible. Classical ways of interpreting PCA results are thus not ##' applicable to NLPCA since the loadings are hidden in the network. ##' However, the scores of components that lead to low ##' cross-validation errors can still be interpreted via the score ##' plot. Unfortunately this method depend on slow iterations which ##' currently are implemented in R only making this method extremely ##' slow. Furthermore, the algorithm does not by itself decide when it ##' has converged but simply does 'maxSteps' iterations. ##' @title Non-linear PCA ##' @param Matrix \code{matrix} --- Preprocessed data with the ##' variables in columns and observations in rows. The data may ##' contain missing values, denoted as \code{NA} ##' @param nPcs \code{numeric} -- Number of components to ##' estimate. The preciseness of the missing value estimation depends ##' on thenumber of components, which should resemble the internal ##' structure of the data. ##' @param maxSteps \code{numeric} -- Number of estimation ##' steps. Default is based on a generous rule of thumb. ##' @param unitsPerLayer The network units, example: c(2,4,6) for two ##' input units 2feature units (principal components), one hidden ##' layer fornon-linearity and three output units (original amount ##' ofvariables). ##' @param functionsPerLayer The function to apply at each layer ##' eg. c("linr", "tanh", "linr") ##' @param weightDecay Value between 0 and 1. ##' @param weights Starting weights for the network. Defaults to ##' uniform random values but can be set specifically to make ##' algorithm deterministic. ##' @param verbose \code{boolean} -- nlpca prints the number of steps ##' and warning messages if set to TRUE. Default is interactive(). ##' @param ... Reserved for future use. Not passed on anywhere. ##' @return Standard PCA result object used by all PCA-basedmethods of ##' this package. Contains scores, loadings, data meanand more. See ##' \code{\link{pcaRes}} for details. ##' @author Based on a matlab script by Matthias Scholz and ported to ##' R by Henning Redestig ##' @references Matthias Scholz, Fatma Kaplan, Charles L Guy, Joachim ##' Kopkaand Joachim Selbig. Non-linear PCA: a missing ##' data approach. \emph{Bioinformatics, 21(20):3887-3895, Oct 2005} ##' @examples ##' ## Data set with three variables where data points constitute a helix ##' data(helix) ##' helixNA <- helix ##' ## not a single complete observation ##' helixNA <- t(apply(helix, 1, function(x) { x[sample(1:3, 1)] <- NA; x})) ##' ## 50 steps is not enough, for good estimation use 1000 ##' helixNlPca <- pca(helixNA, nPcs=1, method="nlpca", maxSteps=50) ##' fittedData <- fitted(helixNlPca, helixNA) ##' plot(fittedData[which(is.na(helixNA))], helix[which(is.na(helixNA))]) ##' ## compared to solution by Nipals PCA which cannot extract non-linear patterns ##' helixNipPca <- pca(helixNA, nPcs=2) ##' fittedData <- fitted(helixNipPca) ##' plot(fittedData[which(is.na(helixNA))], helix[which(is.na(helixNA))]) ##' @export nlpca <- function(Matrix, nPcs=2, maxSteps=2 * prod(dim(Matrix)), unitsPerLayer=NULL, functionsPerLayer=NULL, weightDecay=0.001, weights=NULL, verbose=interactive(),...) { ## do some basic checks object <- Matrix trainIn <- NULL trainOut <- t(object) stds <- apply(trainOut, 2, sd, na.rm=TRUE) scalingFactor <- 0.1 / max(stds) trainOut <- trainOut * scalingFactor ## now setup the initial nlpcaNet object numNaN <- sum(is.na(object)) ## always inverse in this version, bottleneck is not fully implemented inverse <- TRUE ## DATADIST (nlnet@dataDist) is given by weightOut dataDist <- apply(!is.na(trainOut), 2, as.integer) #0 for NA, 1 for everything else if(!inverse) dataDist <- NULL ## setup the network architecture if(is.null(unitsPerLayer)) { ld <- dim(trainOut)[1] lh <- nPcs if(nPcs < 10) lh <- 2 + 2 * nPcs unitsPerLayer <- c(ld, lh, nPcs, lh, ld) if(inverse) unitsPerLayer <- c(nPcs, lh, ld) } featureLayer <- ceiling(length(unitsPerLayer) / 2) if(inverse) featureLayer <- 1 if(is.null(functionsPerLayer)) { functionsPerLayer <- rep("tanh", length(unitsPerLayer)) functionsPerLayer[1] <- "linr" functionsPerLayer[featureLayer] <- "linr" functionsPerLayer[length(unitsPerLayer)] <- "linr" } hierarchic <- list(layer=featureLayer, var=rbind(c(rep(1, nPcs), 0.01)), idx=getHierarchicIdx(unitsPerLayer[featureLayer])) ## set up the weights wNum <- sum(sapply(2:length(unitsPerLayer), function(i) (1 + unitsPerLayer[i - 1]) * unitsPerLayer[i])) if(!is.null(weights) && length(weights) != wNum) { warning("Weight vector not expected length (", wNum, "), using random weights", sep="") weights <- NULL } if(is.null(weights)) weights <- cbind(0.2 * (runif(wNum, 0, 1) - 0.1)) if(inverse) { numPattern <- dim(trainOut)[2] tmpTrainIn <- cbind(rnorm(unitsPerLayer[1] * numPattern,0,1) * 0.1) weights <- rbind(tmpTrainIn, weights) } if(nPcs == 1) featureSorting <- FALSE if(nPcs > 1) featureSorting <- TRUE nlnet <- new("nlpcaNet") nlnet@net <- rbind(unitsPerLayer) nlnet@hierarchic <- hierarchic nlnet@fct <- functionsPerLayer nlnet@fkt <- functionsPerLayer[2:length(functionsPerLayer)] nlnet@weightDecay <- weightDecay nlnet@featureSorting <- featureSorting nlnet@dataDist <- dataDist nlnet@inverse <- inverse nlnet@fCount <- as.integer(0) nlnet@componentLayer <- as.integer(featureLayer) nlnet@error <- errorHierarchic nlnet@gradient <- derrorHierarchic nlnet@maxIter <- as.integer(maxSteps) nlnet@weights <- weightsAccount(weights) nlnet@scalingFactor <- scalingFactor ## ****************************** if(verbose) cat("Training network with", nlnet@maxIter, "iterations...\n!:\tSquare error is NA -- accuracy in line-search might be too small\n:\tComponents were sorted at iteration n\n^:\tToo many iterations while expanding\n") newnet <- optiAlgCgd(nlnet, trainIn, trainOut, verbose) if(verbose) cat("\nDone\n") if(inverse) { nObs <- unitsPerLayer[1] * dim(trainOut)[2] we <- newnet@weights$current() scores <- t(matrix(we[1:nObs], nrow=unitsPerLayer[1], dim(trainOut)[2])) newnet@weights$set(we[(nObs + 1):length(we),,drop=FALSE]) } ## for further applications newnet must not be inverse anymore newnet@inverse <- FALSE res <- new("pcaRes") res@scores <- scores res@loadings <- matrix() res@network <- newnet res@method <- "nlpca" R2cum <- rep(NA, nPcs) TSS <- sum(Matrix^2, na.rm=TRUE) for(i in 1:nPcs) R2cum[i] <- 1 - sum((Matrix - fitted(res, Matrix, nPcs=i))^2, na.rm=TRUE) / TSS res@R2cum <- R2cum res } ##' Index in hiearchy ##' @param hierarchicNum A number ##' @return ... ##' @author Henning Redestig, Matthias Scholz getHierarchicIdx <- function(hierarchicNum) { res <- matrix(1, ncol=hierarchicNum, nrow=hierarchicNum) res[lower.tri(res)] <- 0 cbind(res, c(0, rep(1, hierarchicNum - 1))) } pcaMethods/R/optiAlgCgd.R0000644000175200017520000000414314710217306016227 0ustar00biocbuildbiocbuild##' Conjugate gradient optimization ##' @param nlnet The nlnet ##' @param trainIn Training data ##' @param trainOut fitted data ##' @param verbose logical, print messages ##' @return ... ##' @author Henning Redestig, Matthias Scholz optiAlgCgd <- function(nlnet, trainIn, trainOut, verbose=FALSE) { tmpnet <- forkNlpcaNet(nlnet) derr <- tmpnet@gradient(tmpnet, trainIn, trainOut) dw <- derr$dwTotal e <- derr$Etotal dv <- -dw if(tmpnet@featureSorting) eSortLast <- e eHist <- rep(0, tmpnet@maxIter) ttLast <- rep(0.0001, 6) for(i in 1:tmpnet@maxIter) { if(verbose) { if(i %% 10 == 0) cat("*") if(i %% 100 == 0) cat(" [", i, "]\n") } eHist[i] <- e eLast <- e # line search in direction dv (downhill) ttGuess <- max(min(ttLast), 0.00001) linSe <- lineSearch(tmpnet, dv, e, ttGuess, trainIn, trainOut, verbose) tmpnet@weights$set(cbind(linSe$wBest)) e <- linSe$eBest tt <- linSe$ttBest ttLast <- c(ttLast[2:length(ttLast)], tt) #shift and add new tt gradRes <- tmpnet@gradient(tmpnet, trainIn, trainOut) dwNew <- gradRes$dwTotal e <- gradRes$Etotal ## define new search direction dv (conjugate direction) ## b1=dw_new'*dw_new; # Fletcher-Reeves b1 <- crossprod(dwNew, (dwNew - dw))#Polak-Ribiere b2 <- crossprod(dw) beta <- b1 / b2 dv <- -dwNew + dv %*% beta dw <- dwNew if(e > eLast) { dv <- -dwNew if(verbose) cat("!", sep="") } if(is.na(e)) stop("Square error is NA (critical) - accuracy in line-search might be too small") if(tmpnet@featureSorting) if(e / eSortLast < 0.90 || i == tmpnet@maxIter || i == tmpnet@maxIter - 1 || i == tmpnet@maxIter - 2) { eSortLast <- e if(verbose) cat("<", i, ">", sep="") ## somewhat secret method, sortFeatures calls ## nlnet@weights$set(x) so the weights are updated here ## 'behind the scenes' sortFeatures(tmpnet, trainIn, trainOut) } } tmpnet } pcaMethods/R/orth.R0000644000175200017520000000354014710217306015166 0ustar00biocbuildbiocbuild##' ONB = orth(mat) is an orthonormal basis for the range of matrix ##' mat. That is, ONB' * ONB = I, the columns of ONB span the same ##' space as the columns of mat, and the number of columns of ONB is ##' the rank of mat. ##' @title Calculate an orthonormal basis ##' @param mat matrix to calculate orthonormal base ##' @param skipInac do not include components with precision below ##' .Machine$double.eps if TRUE ##' @return orthonormal basis for the range of matrix ##' @author Wolfram Stacklies orth <- function(mat, skipInac = FALSE) { if(nrow(mat) > ncol(mat)) { leftSVs <- ncol(mat) } else { leftSVs <- nrow(mat) } result <- svd(mat, nu = leftSVs, nv = ncol(mat)) U <- result[[2]] S <- result[[1]] V <- result[[3]] m <- nrow(mat) n <- ncol(mat) if(m > 1) { s <- diag(S, nrow = length(S)) } else if(m == 1) { s <- S[1] } else { s <- 0 } tol <- max(m,n) * max(s) * .Machine$double.eps r <- sum(s > tol) if ( r < ncol(U) ) { if (skipInac) { warning("Precision for components ", r + 1 , " - ", ncol(U), " is below .Machine$double.eps. \n", "Results for those components are likely to be inaccurate!!\n", "These component(s) are not included in the returned solution!!\n") } else { warning("Precision for components ", r + 1 , " - ", ncol(U), " is below .Machine$double.eps. \n", "Results for those components are likely to be inaccurate!!\n") } } if (skipInac) { ONB <- U[, 1:r, drop=FALSE] ## Assing correct row and colnames rownames(ONB) <- labels(mat[, 1:r, drop=FALSE])[[1]]; colnames(ONB) <- labels(mat[, 1:r, drop=FALSE])[[2]]; } else { ONB<-U ## Assing correct row and colnames rownames(ONB) <- labels(mat)[[1]]; colnames(ONB) <- labels(mat)[[2]]; } return(ONB) } pcaMethods/R/pca.R0000644000175200017520000003477014710217306014766 0ustar00biocbuildbiocbuild##' Vector with current valid PCA methods ##' @title List PCA methods ##' @param which the type of methods to get. E.g. only get the PCA ##' methods based on the classical model where the fitted data is a ##' direct multiplication of scores and loadings. ##' @return A character vector with the current methods for doing PCA ##' @export ##' @author Henning Redestig listPcaMethods <- function(which=c("all", "linear", "nonlinear")) { switch(match.arg(which), all={ return(c("svd", "nipals", "rnipals", "bpca", "ppca", "svdImpute", "robustPca", "nlpca", "llsImpute", "llsImputeAll")) }, linear={ return(c("svd", "nipals", "rnipals", "bpca", "ppca", "svdImpute", "robustPca")) }, nonlinear={ return("nlpca") }) } ##' Perform PCA on a numeric matrix for visualisation, information ##' extraction and missing value imputation. ##' ##' This method is wrapper function for the following set of pca ##' methods: ##' ##' \describe{\item{svd:}{Uses classical \code{prcomp}. See ##' documentation for \code{\link{svdPca}}.} ##' ##' \item{nipals:}{An iterative method capable of handling small ##' amounts of missing values. See documentation for ##' \code{\link{nipalsPca}}.} ##' ##' \item{rnipals:}{Same as nipals but implemented in R.} ##' ##' \item{bpca:}{An iterative method using a Bayesian model to handle ##' missing values. See documentation for \code{\link{bpca}}.} ##' ##' \item{ppca:}{An iterative method using a probabilistic model to ##' handle missing values. See documentation for \code{\link{ppca}}.} ##' ##' \item{svdImpute:}{Uses expectation maximation to perform SVD PCA ##' on incomplete data. See documentation for ##' \code{\link{svdImpute}}.}} ##' ##' Scaling and centering is part of the PCA model and handled by ##' \code{\link{prep}}. ##' @title Perform principal component analysis ##' @param object Numerical matrix with (or an object coercible to ##' such) with samples in rows and variables as columns. Also takes ##' \code{ExpressionSet} in which case the transposed expression ##' matrix is used. Can also be a data frame in which case all ##' numberic variables are used to fit the PCA. ##' @param method One of the methods reported by ##' \code{listPcaMethods()}. Can be left missing in which case the ##' \code{svd} PCA is chosen for data wihout missing values and ##' \code{nipalsPca} for data with missing values ##' @param nPcs Number of principal components to calculate. ##' @param scale Scaling, see \code{\link{prep}}. ##' @param center Centering, see \code{\link{prep}}. ##' @param completeObs Sets the \code{completeObs} slot on the ##' resulting \code{pcaRes} object containing the original data with ##' but with all NAs replaced with the estimates. ##' @param subset A subset of variables to use for calculating the ##' model. Can be column names or indices. ##' @param cv character naming a the type of cross-validation ##' to be performed. ##' @param ... Arguments to \code{\link{prep}}, the chosen pca ##' method and \code{\link{Q2}}. ##' @return A \code{pcaRes} object. ##' @references ##' Wold, H. (1966) Estimation of principal components and ##' related models by iterative least squares. In Multivariate ##' Analysis (Ed., P.R. Krishnaiah), Academic Press, NY, 391-420. ##' ##' Shigeyuki Oba, Masa-aki Sato, Ichiro Takemasa, Morito Monden, ##' Ken-ichi Matsubara and Shin Ishii. A Bayesian missing value ##' estimation method for gene expression profile ##' data. \emph{Bioinformatics, 19(16):2088-2096, Nov 2003}. ##' ##' Troyanskaya O. and Cantor M. and Sherlock G. and Brown P. and ##' Hastie T. and Tibshirani R. and Botstein D. and Altman RB. - ##' Missing value estimation methods for DNA microarrays. ##' \emph{Bioinformatics. 2001 Jun;17(6):520-5}. ##' @seealso \code{\link{prcomp}}, \code{\link{princomp}}, ##' \code{\link{nipalsPca}}, \code{\link{svdPca}} ##' @examples ##' data(iris) ##' ## Usually some kind of scaling is appropriate ##' pcIr <- pca(iris, method="svd", nPcs=2) ##' pcIr <- pca(iris, method="nipals", nPcs=3, cv="q2") ##' ## Get a short summary on the calculated model ##' summary(pcIr) ##' plot(pcIr) ##' ## Scores and loadings plot ##' slplot(pcIr, sl=as.character(iris[,5])) ##' ##' ## use an expressionset and ggplot ##' data(sample.ExpressionSet) ##' pc <- pca(sample.ExpressionSet) ##' df <- merge(scores(pc), pData(sample.ExpressionSet), by=0) ##' library(ggplot2) ##' ggplot(df, aes(PC1, PC2, shape=sex, color=type)) + ##' geom_point() + ##' xlab(paste("PC1", pc@R2[1] * 100, "% of the variance")) + ##' ylab(paste("PC2", pc@R2[2] * 100, "% of the variance")) ##' @export ##' @keywords multivariate ##' @author Wolfram Stacklies, Henning Redestig pca <- function(object, method, nPcs=2, scale=c("none", "pareto", "vector", "uv"), center=TRUE, completeObs=TRUE, subset=NULL, cv=c("none","q2"), ...) { if(inherits(object, 'data.frame')) { num <- vapply(object, is.numeric, logical(1)) if(sum(num) < 2) stop('no numeric data in supplied data.frame') Matrix <- as.matrix(object[,num]) } else if(inherits(object, "ExpressionSet")) { Matrix <- t(exprs(object)) } else Matrix <- as.matrix(object, rownames.force=TRUE) if(!is.null(subset)) Matrix <- Matrix[,subset] cv <- match.arg(cv) scale <- match.arg(scale) if (nPcs > ncol(Matrix)) { warning("more components than matrix columns requested") nPcs <- min(dim(Matrix)) } if (nPcs > nrow(Matrix)) { warning("more components than matrix rows requested") nPcs <- min(dim(Matrix)) } if (!checkData(Matrix, verbose=interactive())) stop("Invalid data format.", "Run checkData(data, verbose=TRUE) for details") missing <- is.na(Matrix) if(missing(method)) { if(any(missing)) method <- 'nipals' else method <- 'svd' } if(any(missing) & method == 'svd') { warning('data has missing values using nipals instead of user requested svd') method <- 'nipals' } method <- match.arg(method, choices=listPcaMethods()) prepres <- prep(Matrix, scale=scale, center=center, simple=FALSE, ...) switch(method, svd={ res <- svdPca(prepres$data, nPcs=nPcs,...) }, nipals={ res <- nipalsPca(prepres$data, nPcs=nPcs, ...) }, rnipals={ res <- RnipalsPca(prepres$data, nPcs=nPcs, ...) }, bpca={ res <- bpca(prepres$data, nPcs=nPcs, ...) }, ppca={ res <- ppca(prepres$data, nPcs=nPcs, ...) }, svdImpute={ res <- svdImpute(prepres$data, nPcs=nPcs, ...) }, robustPca={ res <- robustPca(prepres$data, nPcs=nPcs, ...) }, nlpca={ res <- nlpca(prepres$data, nPcs=nPcs, ...) }) nPcs <- ncol(res@scores) if(is.null(scores(res)) | is.null(loadings(res)) | is.null(R2cum(res)) | is.null(method(res))) stop(paste("bad result from pca method", method)) colnames(res@scores) <- paste("PC", 1:nPcs, sep="") rownames(res@scores) <- rownames(Matrix) if(all(dim(loadings(res)) == c(ncol(Matrix), nPcs))) { colnames(res@loadings) <- paste("PC", 1:nPcs, sep="") rownames(res@loadings) <- colnames(Matrix) } if(!is.null(subset)) res@subset <- subset res@missing <- missing res@nPcs <- nPcs res@nObs <- nrow(Matrix) res@nVar <- ncol(Matrix) res@sDev <- apply(scores(res), 2, sd) res@center <- prepres$center res@centered <- center res@scale <- prepres$scale res@scaled <- scale res@R2 <- res@R2cum[1] if(length(res@R2cum) > 1) res@R2 <- c(res@R2, diff(res@R2cum)) if (completeObs) { cObs <- Matrix if(method %in% listPcaMethods("nonlinear")) cObs[missing] <- fitted(res, Matrix, pre=TRUE, post=TRUE)[missing] else cObs[missing] <- fitted(res, post=TRUE)[missing] res@completeObs <- cObs } if(cv == "q2") res@cvstat <- Q2(res, Matrix, nruncv=1, ...) return(res) } ##' Wrapper function for imputation methods based on nearest neighbour ##' clustering. Currently llsImpute only. ##' ##' This method is wrapper function to llsImpute, See documentation ##' for \code{link{llsImpute}}. ##' @title Nearest neighbour imputation ##' @param object Numerical matrix with (or an object coercible to ##' such) with samples in rows and variables as columns. Also takes ##' \code{ExpressionSet} in which case the transposed expression ##' matrix is used. ##' @param method For convenience one can pass a large matrix but only ##' use the variable specified as subset. Can be colnames or indices. ##' @param subset Currently "llsImpute" only. ##' @param ... Further arguments to the chosen method. ##' @return A \code{clusterRes} object. Or a list containing a ##' clusterRes object as first and an ExpressionSet object as second ##' entry if the input was of type ExpressionSet. ##' @export ##' @seealso \code{\link{llsImpute}}, \code{\link{pca}} ##' @keywords multivariate ##' @examples ##' data(metaboliteData) ##' llsRes <- nni(metaboliteData, k=6, method="llsImpute", allGenes=TRUE) ##' @author Wolfram Stacklies nni <- function(object, method=c("llsImpute"), subset=numeric(), ...) { isExprSet <- FALSE if(inherits(object, "ExpressionSet")) { set <- object isExprSet <- TRUE object <- t(exprs(object)) } method <- match.arg(method) if ( !checkData(as.matrix(object), verbose=interactive()) ) stop("Invalid data format, exiting...\n", "Run checkData(data, verbose=TRUE) for details\n") missing <- sum(is.na(object)) if(length(subset) > 0) object <- object[,subset] res <- llsImpute(object, ...) return(res) } ##' A function that can be used to visualise many PCs plotted against ##' each other ##' ##' Uses \code{\link{pairs}} to provide side-by-side plots. Note that ##' this function only plots scores or loadings but not both in the ##' same plot. ##' @title Plot many side by side scores XOR loadings plots ##' @param object \code{pcaRes} a pcaRes object ##' @param pcs \code{numeric} which pcs to plot ##' @param type \code{character} Either "scores" or "loadings" for ##' scores or loadings plot respectively ##' @param sl \code{character} Text labels to plot instead of a point, ##' if NULL points are plotted instead of text ##' @param hotelling \code{numeric} Significance level for the ##' confidence ellipse. NULL means that no ellipse is drawn. ##' @param ... Further arguments to \code{\link{pairs}} on which this ##' function is based. ##' @return None, used for side effect. ##' @seealso \code{prcomp}, \code{pca}, \code{princomp}, \code{slplot} ##' @export ##' @examples ##' data(iris) ##' pcIr <- pca(iris[,1:4], nPcs=3, method="svd") ##' plotPcs(pcIr, col=as.integer(iris[,4]) + 1) ##' @keywords multivariate ##' @author Henning Redestig plotPcs <- function(object, pcs=1:nP(object), type=c("scores", "loadings"), sl=NULL, hotelling=0.95, ...) { type <- match.arg(type) panel <- function(x,y, ...) { abline(h=0, v=0, col="black") if(!is.null(hotelling)) { A <- length(pcs) el <- simpleEllipse(x, y, alfa=hotelling) lines(el) } if(is.null(sl)) points(x, y, ...) else text(x, y, labels=sl,...) } switch(type, scores={ labels <- paste("PC", pcs, "\n", "R^2 =", round(object@R2[pcs], 2)) pairs(scores(object)[,pcs], labels=labels, panel=panel, upper.panel=NULL,...) }, loadings={ if(method(object) == "nlpca") stop("Loadings plot not applicable for non-linear PCA") labels <- paste("PC", pcs, "\n", "R^2 =", round(object@R2[pcs], 2)) pairs(loadings(object)[,pcs], labels=labels, panel=panel, upper.panel=NULL, ...) }) } ##' A wrapper function for \code{prcomp} to deliver the result as a ##' \code{pcaRes} method. Supplied for compatibility with the rest ##' of the pcaMethods package. It is not recommended to use this ##' function directely but rather to use the \code{pca()} wrapper ##' function. ##' @title Perform principal component analysis using singular value ##' decomposition ##' @param Matrix Pre-processed (centered and possibly scaled) ##' numerical matrix samples in rows and variables as columns. No ##' missing values allowed. ##' @param nPcs Number of components that should be extracted. ##' @param varLimit Optionally the ratio of variance that should be ##' explained. \code{nPcs} is ignored if varLimit < 1 ##' @param verbose Verbose complaints to matrix structure ##' @param ... Only used for passing through arguments. ##' @return A \code{pcaRes} object. ##' @seealso \code{prcomp}, \code{princomp}, \code{pca} ##' @examples ##' data(metaboliteDataComplete) ##' mat <- prep(t(metaboliteDataComplete)) ##' pc <- svdPca(mat, nPcs=2) ##' ## better use pca() ##' pc <- pca(t(metaboliteDataComplete), method="svd", nPcs=2) ##' \dontshow{stopifnot(sum((fitted(pc) - t(metaboliteDataComplete))^2, na.rm=TRUE) < 200)} ##' @export ##' @keywords multivariate ##' @author Henning Redestig svdPca <- function(Matrix, nPcs=2, varLimit=1, verbose=interactive(), ...) { pcs <- prcomp(Matrix, center=FALSE, scale.=FALSE) imp <- summary(pcs)$importance if(varLimit < 1) nPcs <- sum(imp[3,] < varLimit) + 1 res <- new("pcaRes") res@scores <- cbind(pcs$x[,1:nPcs]) res@loadings <- cbind(pcs$rotation[,1:nPcs]) res@R2cum <- imp[3,1:nPcs] res@varLimit <- varLimit res@method <- "svd" return(res) } ##' Get a confidence ellipse for uncorrelated bivariate data ##' ##' As described in 'Introduction to multi and megavariate data analysis ##' using PCA and ##' PLS' by Eriksson et al. This produces very similar ellipse as ##' compared to the ellipse function the ellipse package except that ##' this function assumes that and y are uncorrelated (which they of ##' are if they are scores or loadings from a PCA). ##' @title Hotelling's T^2 Ellipse ##' @param x first variable ##' @param y second variable ##' @param alfa confidence level of the circle ##' @param len Number of points in the circle ##' @seealso ellipse ##' @author Henning Redestig ##' @return A matrix with X and Y coordinates for the circle simpleEllipse <- function(x, y, alfa=0.95, len=200) { N <- length(x) A <- 2 mypi <- seq(0, 2 * pi, length=len) r1 <- sqrt(var(x) * qf(alfa, 2, N - 2) * (2*(N^2 - 1)/(N * (N - 2)))) r2 <- sqrt(var(y) * qf(alfa, 2, N - 2) * (2*(N^2 - 1)/(N * (N - 2)))) cbind(r1 * cos(mypi) + mean(x), r2 * sin(mypi) + mean(y)) } # .onLoad <- function(libname, pkgname) { # require("methods") # } pcaMethods/R/pcaMethods-package.R0000644000175200017520000000772014710217306017676 0ustar00biocbuildbiocbuild##' Simulated data set looking like a helix ##' ##' ##' A matrix containing 1000 observations (rows) and three variables ##' (columns). ##' @title A helix structured toy data set ##' @name helix ##' @aliases helix ##' @usage data(helix) ##' @docType data ##' @references Matthias Scholz, Fatma Kaplan, Charles L. Guy, Joachim ##' Kopka and Joachim Selbig. - Non-linear PCA: a missing data ##' approach. \emph{Bioinformatics 2005 21(20):3887-3895} ##' @keywords datasets ##' @author Henning Redestig NULL ##' A complete subset from a larger metabolite data set. This is the ##' original, complete data set and can be used to compare estimation ##' results created with the also provided incomplete data (called ##' metaboliteData). The data was created during an in house ##' Arabidopsis coldstress experiment. ##' ##' A matrix containing 154 observations (rows) and 52 metabolites ##' (columns). ##' @name metaboliteDataComplete ##' @docType data ##' @aliases metaboliteDataComplete ##' @title A complete metabolite data set from an Arabidopsis ##' coldstress experiment ##' @keywords datasets ##' @seealso \code{\link{metaboliteData}} ##' @references Matthias Scholz, Fatma Kaplan, Charles L. Guy, Joachim ##' Kopka and Joachim Selbig. - Non-linear PCA: a missing data ##' approach.\emph{Bioinformatics 2005 21(20):3887-3895} ##' @author Wolfram Stacklies NULL ##' A incomplete subset from a larger metabolite data set. This is the ##' original, complete data set and can be used to compare estimation ##' results created with the also provided incomplete data (called ##' metaboliteData). ##' ##' A matrix containing 154 observations (rows) and 52 metabolites ##' (columns). The data contains 5\% of artificially created uniformly ##' distributed misssing values. The data was created during an in ##' house Arabidopsis coldstress experiment. ##' @name metaboliteData ##' @docType data ##' @aliases metaboliteData ##' @title A incomplete metabolite data set from an Arabidopsis ##' coldstress experiment ##' @keywords datasets ##' @seealso \code{\link{metaboliteDataComplete}} ##' @references Matthias Scholz, Fatma Kaplan, Charles L. Guy, Joachim ##' Kopka and Joachim Selbig. - Non-linear PCA: a missing data ##' approach.\emph{Bioinformatics 2005 21(20):3887-3895} ##' @author Wolfram Stacklies NULL ##' Principal Component Analysis in R ##' ##' \tabular{ll}{ ##' Package: \tab pcaMethods \cr ##' Type: \tab Package \cr ##' Developed since: \tab 2006 \cr ##' License: \tab GPL (>=3) \cr ##' LazyLoad: \tab yes \cr ##' } ##' ##' Provides Bayesian PCA, Probabilistic PCA, Nipals PCA, Inverse ##' Non-Linear PCA and the conventional SVD PCA. A cluster based ##' method for missing value estimation is included for comparison. ##' BPCA, PPCA and NipalsPCA may be used to perform PCA on incomplete ##' data as well as for accurate missing value estimation. A set of ##' methods for printing and plotting the results is also provided. ##' All PCA methods make use of the same data structure (pcaRes) to ##' provide a unique interface to the PCA results. Developed at the ##' Max-Planck Institute for Molecular Plant Physiology, Golm, ##' Germany, RIKEN Plant Science Center Yokohama, Japan, and CAS-MPG ##' Partner Institute for Computational Biology (PICB) Shanghai, ##' P.R. China ##' ##' @name pcaMethods ##' @aliases pcaMethods ##' @docType package ##' @importFrom Rcpp evalCpp ##' @import Biobase ##' @import BiocGenerics ##' @import methods ##' @title pcaMethods ##' @useDynLib pcaMethods ##' @author Wolfram Stacklies, Henning Redestig NULL ##' \describe{ ##' \item{plotR2}{Lack of relevance for this plot and the fact that it ##' can not show cross-validation based diagnostics in the same plot ##' makes it redundant with the introduction of a dedicated ##' \code{plot} function for \code{pcaRes}. The new plot only shows ##' R2cum but the result is pretty much the same.}} ##' @name pcaMethods-deprecated ##' @aliases pcaMethods-deprecated ##' @title Deprecated methods for pcaMethods ##' @author Henning Redestig NULL pcaMethods/R/ppca.R0000644000175200017520000001463414710217306015143 0ustar00biocbuildbiocbuild##' Implementation of probabilistic PCA (PPCA). PPCA allows to perform ##' PCA on incomplete data and may be used for missing value ##' estimation. This script was implemented after the Matlab version ##' provided by Jakob Verbeek ( see ##' \url{http://lear.inrialpes.fr/~verbeek/}) and the draft \emph{``EM ##' Algorithms for PCA and Sensible PCA''} written by Sam Roweis. ##' ##' Probabilistic PCA combines an EM approach for PCA with a ##' probabilistic model. The EM approach is based on the assumption ##' that the latent variables as well as the noise are normal ##' distributed. ##' ##' In standard PCA data which is far from the training set but close ##' to the principal subspace may have the same reconstruction error. ##' PPCA defines a likelihood function such that the likelihood for ##' data far from the training set is much lower, even if they are ##' close to the principal subspace. This allows to improve the ##' estimation accuracy. ##' ##' A method called \code{kEstimate} is provided to estimate the ##' optimal number of components via cross validation. In general few ##' components are sufficient for reasonable estimation accuracy. See ##' also the package documentation for further discussion on what kind ##' of data PCA-based missing value estimation is advisable. ##' ##' \bold{Complexity:}\cr Runtime is linear in the number of data, ##' number of data dimensions and number of principal components. ##' ##' \bold{Convergence:} The threshold indicating convergence was ##' changed from 1e-3 in 1.2.x to 1e-5 in the current version leading ##' to more stable results. For reproducability you can set the seed ##' (parameter seed) of the random number generator. If used for ##' missing value estimation, results may be checked by simply running ##' the algorithm several times with changing seed, if the estimated ##' values show little variance the algorithm converged well. ##' @title Probabilistic PCA ##' @param Matrix \code{matrix} -- Data containing the variables in ##' columns and observations in rows. The data may contain missing ##' values, denoted as \code{NA}. ##' @param nPcs \code{numeric} -- Number of components to ##' estimate. The preciseness of the missing value estimation depends ##' on the number of components, which should resemble the internal ##' structure of the data. ##' @param seed \code{numeric} Set the seed for the random number ##' generator. PPCA creates fills the initial loading matrix with ##' random numbers chosen from a normal distribution. Thus results may ##' vary slightly. Set the seed for exact reproduction of your ##' results. ##' @param threshold Convergence threshold. ##' @param maxIterations the maximum number of allowed iterations ##' @param ... Reserved for future use. Currently no further ##' parameters are used. ##' @note Requires \code{MASS}. It is not recommended to use this ##' function directely but rather to use the pca() wrapper function. ##' @return Standard PCA result object used by all PCA-based methods ##' of this package. Contains scores, loadings, data mean and ##' more. See \code{\link{pcaRes}} for details. ##' @seealso \code{\link{bpca}, \link{svdImpute}, \link{prcomp}, ##' \link{nipalsPca}, \link{pca}, \link{pcaRes}}. ##' @examples ##' ## Load a sample metabolite dataset with 5\% missing values (metaboliteData) ##' data(metaboliteData) ##' ## Perform probabilistic PCA using the 3 largest components ##' result <- pca(t(metaboliteData), method="ppca", nPcs=3, seed=123) ##' ## Get the estimated complete observations ##' cObs <- completeObs(result) ##' ## Plot the scores ##' plotPcs(result, type = "scores") ##' \dontshow{ ##' stopifnot(sum((fitted(result) - t(metaboliteData))^2, na.rm=TRUE) < 200) ##' } ##' @keywords multivariate ##' @author Wolfram Stacklies ##' @export ppca <- function(Matrix, nPcs=2, seed=NA, threshold=1e-5, maxIterations=1000, ...) { ## Set the seed to the user defined value. This affects the generation ## of random values for the initial setup of the loading matrix if (!is.na(seed)) set.seed(seed) N <- nrow(Matrix) D <- ncol(Matrix) Obs <- !is.na(Matrix) hidden <- which(is.na(Matrix)) missing <- length(hidden) if(missing) { Matrix[hidden] <- 0 } ## ------- Initialization r <- sample(N) C <- t(Matrix[r[1:nPcs], ,drop = FALSE]) ## Random matrix with the same dimnames as Matrix C <- matrix(rnorm(C), nrow(C), ncol(C), dimnames = labels(C) ) CtC <- t(C) %*% C ## inv(C'C) C' X is the solution to the EM problem X <- Matrix %*% C %*% solve(CtC) recon <- X %*% t(C) recon[hidden] <- 0 ss <- sum(sum((recon - Matrix)^2)) / (N * D - missing) count <- 1 old <- Inf ## ------ EM iterations while (count > 0) { ## E-step, (co)variances Sx <- solve(diag(nPcs) + CtC/ss) ss_old <- ss if(missing) { proj <- X %*% t(C) Matrix[hidden] <- proj[hidden] } ## E step: expected values X <- Matrix %*% C %*% Sx / ss ## M-step SumXtX <- t(X) %*% X ## Replace the right matrix division from matlab C <- (t(Matrix) %*% X) %*% solve( (SumXtX + N * Sx) ) CtC <- t(C) %*% C ss <- ( sum(sum( (C %*% t(X) - t(Matrix))^2 )) + N * sum(sum(CtC %*% Sx)) + missing * ss_old ) / (N * D) objective <- N * (D * log(ss) + sum(diag(Sx)) - log(det(Sx)) ) + sum(diag(SumXtX)) - missing * log(ss_old) rel_ch <- abs( 1 - objective / old ) old <- objective count <- count + 1 if( rel_ch < threshold & count > 5 ) { count <- 0 } else if (count > maxIterations) { count <- 0 warning("stopped after max iterations, but rel_ch was > threshold") } } ## End EM iteration C <- orth(C) evs <- eigen( cov(Matrix %*% C) ) vals <- evs[[1]] vecs <- evs[[2]] C <- C %*% vecs X <- Matrix %*% C ## Paramters in original Matlab implementation were: ## C (D by d) - C has the approximate loadings (eigenvectors of ## the covariance matrix) ## as columns. ## X - The approximate scores ## Matrix (N by D) - Expected complete observations. ## M (D by 1) - Column wise data mean ## ss (scalar) - isotropic variance outside subspace R2cum <- rep(NA, nPcs) TSS <- sum(Matrix^2, na.rm=TRUE) for (i in 1:ncol(C)) { difference <- Matrix - (X[,1:i, drop=FALSE] %*% t(C[,1:i, drop=FALSE])) R2cum[i] <- 1 - (sum(difference^2, na.rm=TRUE) / TSS) } res <- new("pcaRes") res@scores <- X res@loadings <- C res@R2cum <- R2cum res@method <- "ppca" return(res) } pcaMethods/R/prep.R0000644000175200017520000000732214710217306015162 0ustar00biocbuildbiocbuild##' Scaling and centering a matrix. ##' ##' Does basically the same as \code{\link{scale}} but adds some ##' alternative scaling options and functionality for treating ##' pre-processing as part of a model. ##' @title Pre-process a matrix for PCA ##' @param object Numerical matrix (or an object coercible to such) ##' with samples in rows and variables as columns. Also takes ##' \code{ExpressionSet} in which case the transposed expression ##' matrix is used. ##' @param scale One of "UV" (unit variance \eqn{a=a/\sigma_{a}}) ##' "vector" (vector normalisation \eqn{b=b/||b||}), "pareto" (sqrt ##' UV) or "none" to indicate which scaling should be used to scale ##' the matrix with \eqn{a} variables and \eqn{b} samples. Can also be ##' a vector of scales which should be used to scale the ##' matrix. \code{NULL} value is interpreted as \code{"none"}. ##' @param center Either a logical which indicates if the matrix ##' should be mean centred or not, or a vector with averages which ##' should be suntracted from the matrix. \code{NULL} value is ##' interpreted as \code{FALSE} ##' @param eps Minimum variance, variable with lower variance are not ##' scaled and warning is issued instead. ##' @param simple Logical indicating if only the data should be ##' returned or a list with the pre-processing statistics as well. ##' @param reverse Logical indicating if matrix should be ##' 'post-processed' instead by multiplying each column with its scale ##' and adding the center. In this case, center and scale should be ##' vectors with the statistics (no warning is issued if not, instead ##' output becomes the same as input). ##' @param ... Only used for passing through arguments. ##' @return A pre-processed matrix or a list with ##' \item{center}{a vector with the estimated centers} ##' \item{scale}{a vector with the estimated scales} ##' \item{data}{the pre (or post) processed data} ##' @examples ##' object <- matrix(rnorm(50), nrow=10) ##' res <- prep(object, scale="uv", center=TRUE, simple=FALSE) ##' obj <- prep(object, scale=res$scale, center=res$center) ##' ## same as original ##' sum((object - prep(obj, scale=res$scale, center=res$center, rev=TRUE))^2) ##' @export ##' @author Henning Redestig prep <- function(object, scale=c("none", "pareto", "vector", "uv"), center=TRUE, eps=1e-12, simple=TRUE, reverse=FALSE, ...) { if(inherits(object, "ExpressionSet")) obj <- t(exprs(object)) else obj <- as.matrix(object) if(is.null(center)) center <- FALSE if(is.null(scale)) scale <- "none" if(is.logical(center[1])) { if(center[1]) center <- colMeans(obj, na.rm=TRUE) else center <- rep(0, ncol(obj)) } if(length(center) != ncol(obj)) stop("center do not match matrix dimensions") if(!reverse) obj <- sweep(obj, 2, center, "-") if(is.character(scale[1])) { scale <- match.arg(scale) if(scale == "uv") scale <- apply(obj, 2, sd, na.rm=TRUE) else if(scale == "none") scale <- rep(1, ncol(obj)) else if(scale == "pareto") scale <- sqrt(apply(obj, 2, sd, na.rm=TRUE)) else if(scale == "vector") scale <- apply(obj, 2, function(x) sqrt(sum(x^2, na.rm=TRUE))) } if(length(scale) != ncol(obj)) stop("scale vector do not match matrix dimensions") if (any(scale < eps)) warning(paste("Variance is below eps for", sum(scale < eps), "variables. Not scaling them.")) scale[scale < eps] <- 1 if(!reverse) obj <- sweep(obj, 2, scale, "/") if(reverse) { obj <- sweep(obj, 2, scale, "*") obj <- sweep(obj, 2, center, "+") } if(inherits(object, "ExpressionSet")) exprs(object) <- t(obj) else object <- obj if (simple) object else list(data=object, center=center, scale=scale) } pcaMethods/R/repmat.R0000644000175200017520000000166714710217306015512 0ustar00biocbuildbiocbuild##' Creates a large matrix B consisting of an M-by-N tiling of copies ##' of A ##' @title Replicate and tile an array. ##' @param mat numeric matrix ##' @param M number of copies in vertical direction ##' @param N number of copies in horizontal direction ##' @return Matrix consiting of M-by-N tiling copies of input matrix ##' @author Wolfram Stacklies repmat <- function(mat, M, N) { ## Check if all input parameters are correct if( !all(M > 0, N > 0) ) { stop("M and N must be > 0") } ## Convert array to matrix ma <- mat if(!is.matrix(mat)) { ma <- matrix(mat, nrow=1) } rows <- nrow(ma) cols <- ncol(ma) replicate <- matrix(0, rows * M, cols * N) for (i in 1:M) { for(j in 1:N) { start_row <- (i - 1) * rows + 1 end_row <- i * rows start_col <- (j - 1) * cols + 1 end_col <- j * cols replicate[start_row:end_row, start_col:end_col] <- ma } } return(replicate) } pcaMethods/R/robustPca.R0000644000175200017520000002237014710217306016156 0ustar00biocbuildbiocbuild##' This is a PCA implementation robust to outliers in a data set. It ##' can also handle missing values, it is however NOT intended to be ##' used for missing value estimation. As it is based on robustSVD we ##' will get an accurate estimation for the loadings also for ##' incomplete data or for data with outliers. The returned scores ##' are, however, affected by the outliers as they are calculated ##' inputData X loadings. This also implies that you should look at ##' the returned R2/R2cum values with caution. If the data show ##' missing values, scores are caluclated by just setting all NA - ##' values to zero. This is not expected to produce accurate results. ##' Please have also a look at the manual page for \code{robustSvd}. ##' Thus this method should mainly be seen as an attempt to integrate ##' \code{robustSvd()} into the framework of this package. Use one of ##' the other methods coming with this package (like PPCA or BPCA) if ##' you want to do missing value estimation. It is not recommended to ##' use this function directely but rather to use the pca() wrapper ##' function. ##' ##' The method is very similar to the standard \code{prcomp()} ##' function. The main difference is that \code{robustSvd()} is used ##' instead of the conventional \code{svd()} method. ##' @title PCA implementation based on robustSvd ##' @param Matrix \code{matrix} -- Data containing the variables in ##' columns and observations in rows. The data may contain missing ##' values, denoted as \code{NA}. ##' @param nPcs \code{numeric} -- Number of components to ##' estimate. The preciseness of the missing value estimation depends ##' on the number of components, which should resemble the internal ##' structure of the data. ##' @param verbose \code{boolean} Print some output to the command ##' line if TRUE ##' @param ... Reserved for future use. Currently no further ##' parameters are used ##' @return Standard PCA result object used by all PCA-based methods ##' of this package. Contains scores, loadings, data mean and ##' more. See \code{\link{pcaRes}} for details. are used. ##' @seealso \code{\link{robustSvd}, \link{svd}, \link{prcomp}, ##' \link{pcaRes}}. ##' @examples ##' ## Load a complete sample metabolite data set and mean center the data ##' data(metaboliteDataComplete) ##' mdc <- scale(metaboliteDataComplete, center=TRUE, scale=FALSE) ##' ## Now create 5\% of outliers. ##' cond <- runif(length(mdc)) < 0.05; ##' mdcOut <- mdc ##' mdcOut[cond] <- 10 ##' ## Now we do a conventional PCA and robustPca on the original and the data ##' ## with outliers. ##' ## We use center=FALSE here because the large artificial outliers would ##' ## affect the means and not allow to objectively compare the results. ##' resSvd <- pca(mdc, method="svd", nPcs=10, center=FALSE) ##' resSvdOut <- pca(mdcOut, method="svd", nPcs=10, center=FALSE) ##' resRobPca <- pca(mdcOut, method="robustPca", nPcs=10, center=FALSE) ##' ## Now we plot the results for the original data against those with outliers ##' ## We can see that robustPca is hardly effected by the outliers. ##' plot(loadings(resSvd)[,1], loadings(resSvdOut)[,1]) ##' plot(loadings(resSvd)[,1], loadings(resRobPca)[,1]) ##' @keywords multivariate ##' @export ##' @author Wolfram Stacklies robustPca <- function(Matrix, nPcs=2, verbose=interactive(), ... ) { nas <- is.na(Matrix) if (sum(nas) != 0) warning("Data is incomplete, it is not recommended to use robustPca for missing value estimation") svdSol <- robustSvd(Matrix) ## Sort the eigenvalues and eigenvectors loadings <- svdSol$v[, 1:nPcs, drop=FALSE] sDev <- svdSol$d[1:nPcs] / sqrt(max(1, nrow(Matrix) - 1)) ## We estimate the scores by just setting all NA values to 0 This is ## a bad approximation, I know... Use ppca / bpca or other missing ## value estimation methods included in this package compMat <- Matrix compMat[is.na(compMat)] <- 0 scores <- compMat %*% loadings ## Calculate R2cum (on the complete observations only) R2cum <- rep(NA, nPcs) TSS <- sum(Matrix^2, na.rm=TRUE) for (i in 1:nPcs) { difference <- Matrix - (scores[,1:i, drop=FALSE] %*% t(loadings[,1:i, drop=FALSE])) R2cum[i] <- 1 - (sum(difference^2) / TSS) } result <- new("pcaRes") result@loadings <- loadings result@scores <- scores result@R2cum <- R2cum result@method <- "robustPca" return(result) } ##' A robust approximation to the singular value decomposition of a ##' rectangular matrix is computed using an alternating L1 norm ##' (instead of the more usual least squares L2 norm). As the SVD is ##' a least-squares procedure, it is highly susceptible to outliers ##' and in the extreme case, an individual cell (if sufficiently ##' outlying) can draw even the leading principal component toward ##' itself. ##' ##' See Hawkins et al (2001) for details on the robust SVD algorithm. ##' Briefly, the idea is to sequentially estimate the left and right ##' eigenvectors using an L1 (absolute value) norm minimization. ##' ##' Note that the robust SVD is able to accomodate missing values in ##' the matrix \code{x}, unlike the usual \code{svd} function. ##' ##' Also note that the eigenvectors returned by the robust SVD ##' algorithm are NOT (in general) orthogonal and the eigenvalues need ##' not be descending in order. ##' @title Alternating L1 Singular Value Decomposition ##' @param x A matrix whose SVD decomposition is to be ##' computed. Missing values are allowed. ##' @return The robust SVD of the matrix is x=u d v'. \item{d}{A ##' vector containing the singular values of \code{x}.} \item{u}{A ##' matrix whose columns are the left singular vectors of \code{x}.} ##' \item{v}{A matrix whose columns are the right singular vectors of ##' \code{x}.} ##' @note Two differences from the usual SVD may be noted. One relates ##' to orthogonality. In the conventional SVD, all the eigenvectors ##' are orthogonal even if not explicitly imposed. Those returned by ##' the AL1 algorithm (used here) are (in general) not orthogonal. ##' Another difference is that, in the L2 analysis of the conventional ##' SVD, the successive eigen triples (eigenvalue, left eigenvector, ##' right eigenvector) are found in descending order of ##' eigenvalue. This is not necessarily the case with the AL1 ##' algorithm. Hawkins et al (2001) note that a larger eigen value ##' may follow a smaller one. ##' @references Hawkins, Douglas M, Li Liu, and S Stanley Young (2001) ##' Robust Singular Value Decomposition, National Institute of ##' Statistical Sciences, Technical Report Number ##' 122. \url{http://www.niss.org/technicalreports/tr122.pdf} ##' @author Kevin Wright, modifications by Wolfram Stacklies ##' @seealso \code{\link{svd}}, \code{\link[ade4:nipals]{nipals}} for ##' an alternating L2 norm method that also accommodates missing data. ##' @examples ##' ## Load a complete sample metabolite data set and mean center the data ##' data(metaboliteDataComplete) ##' mdc <- prep(metaboliteDataComplete, center=TRUE, scale="none") ##' ## Now create 5% of outliers. ##' cond <- runif(length(mdc)) < 0.05; ##' mdcOut <- mdc ##' mdcOut[cond] <- 10 ##' ## Now we do a conventional SVD and a robustSvd on both, the original and the ##' ## data with outliers. ##' resSvd <- svd(mdc) ##' resSvdOut <- svd(mdcOut) ##' resRobSvd <- robustSvd(mdc) ##' resRobSvdOut <- robustSvd(mdcOut) ##' ## Now we plot the results for the original data against those with outliers ##' ## We can see that robustSvd is hardly affected by the outliers. ##' plot(resSvd$v[,1], resSvdOut$v[,1]) ##' plot(resRobSvd$v[,1], resRobSvdOut$v[,1]) ##' @keywords algebra ##' @export robustSvd <- function(x) { ## We need the weightedMedian function provided by the aroma.light ## package. However we do not want to make the whole package dependant ## on aroma.light if (!requireNamespace("matrixStats", quietly=TRUE)) stop("package matrixStats required but not available") L1RegCoef <- function(x, a){ keep <- (abs(a) > .Machine$double.eps) & (!is.na(x)) if(!any(keep)) { warning("No non-missing data for l1 regression, unstable results") return(0.) } a <- a[keep] return(matrixStats::weightedMedian(x[keep] / a, abs(a), na.rm=TRUE, interpolate=FALSE)) } L1Eigen <- function(x, a, b){ x <- as.vector(x) # Convert from matrix to vector ab <- as.vector(outer(a, b)) keep <- (abs(ab) > .Machine$double.eps) & (!is.na(x)) ab <- ab[keep] return(matrixStats::weightedMedian(x[keep] / ab, abs(ab), na.rm=TRUE, interpolate=FALSE)) } ## Initialize outputs svdu <- matrix(NA, nrow=nrow(x), ncol=ncol(x)) svdv <- matrix(NA, nrow=ncol(x), ncol=ncol(x)) svdd <- rep(NA, ncol(x)) for(k in 1:ncol(x)) { ak <- apply(abs(x), 1, median, na.rm=TRUE) converged <- FALSE while(!converged) { akprev <- ak c <- apply(x, 2, L1RegCoef, ak) bk <- c / sqrt(sum(c^2)) d <- apply(x, 1, L1RegCoef, bk) ak <- d / sqrt(sum(d^2)) if(sum((ak - akprev)^2) < 1e-10) { converged <- TRUE } } eigenk <- L1Eigen(x,ak,bk) ## Deflate the x matrix x <- x - eigenk * ak %*% t(bk) ## Store eigen triple for output svdu[,k] <- ak svdv[,k] <- bk svdd[k] <- eigenk } ## Create the result object ret <- list() ret$d <- svdd ret$u <- svdu ret$v <- svdv return(ret) } pcaMethods/R/sortFeatures.R0000644000175200017520000000502114710217306016674 0ustar00biocbuildbiocbuild##' Sort the features of NLPCA object ##' @param nlnet The nlnet ##' @param trainIn Training data in ##' @param trainOut Training data after it passed through the net ##' @return ... ##' @author Henning Redestig sortFeatures <- function(nlnet, trainIn, trainOut) { weightsAll <- nlnet@weights$current() weights <- weightsAll if(nlnet@inverse) { numElements <- nlnet@net[1] * dim(trainOut)[2] trainIn <- array(unlist(weightsAll), dim=c(nlnet@net[1], dim(trainOut)[2])) weights <- weightsAll[(numElements + 1):length(weightsAll),,drop=FALSE] } netDim <- dim(nlnet@net) trainDim <- dim(trainIn) bneckNum <- nlnet@net[nlnet@componentLayer] weightMats <- vector2matrices(weights, nlnet@net) bneckNum <- nlnet@net[nlnet@componentLayer] ## ****************************** nOut <- array(0, dim=c(sum(nlnet@net), trainDim[2], 2)) for(subnet in 1:2) nOut[1:trainDim[1],,subnet] <- trainIn ## forward propagation for(n in 0:(bneckNum - 2)) { E <- c(0,0) for(choice in 1:2) { sBias <- rep(1, trainDim[2]) for(i in 1:(netDim[2] - 1)) { if(i == 1) nBegin <- 1 else nBegin <- sum(nlnet@net[1:(i - 1)]) + 1 sIn <- rbind(sBias, nOut[nBegin:sum(nlnet@net[1:i]),, choice]) sOut <- eval(parse(text=paste(nlnet@fkt[i], "(weightMats[[i]] %*% sIn)"))) if(i == nlnet@componentLayer - 1) { idx <- rep(0, bneckNum) idx[1:(n + choice)] <- 1 if(choice == 2) idx[n+choice-1] <- 0 sOut[idx == 0,] <- 0 } nOut[(sum(nlnet@net[1:i]) + 1):sum(nlnet@net[1:(i+1)]),,choice] <- sOut } output <- nOut[(sum(nlnet@net[1:(dim(nlnet@net)[2]-1)])+1):dim(nOut)[1], ,choice] Epattern <- (output - trainOut)^2 Epattern[is.na(Epattern)] <- 0 if(!is.null(nlnet@dataDist)) Epattern <- Epattern * nlnet@dataDist E <- mean(Epattern) E[choice] <- E } if(E[1]>E[2]) { #change features changeIdx <- 1:bneckNum changeIdx[(n+1):(n+2)] <- c(n+2, n+1) weightMats[[nlnet@componentLayer - 1]] <- weightMats[[nlnet@componentLayer - 1]][changeIdx,] weightMats[[nlnet@componentLayer]] <- weightMats[[nlnet@componentLayer]][,c(1,changeIdx+1)] switching <- c(n+1, n+2) nlnet@fCount <- as.integer(nlnet@fCount + 1) } } weights <- cbind(unlist(weightMats)) if(nlnet@inverse) nlnet@weights$set(rbind(matrix(trainIn, nrow=numElements, ncol=1), weights)) } pcaMethods/R/svdImpute.R0000644000175200017520000001132714710217306016174 0ustar00biocbuildbiocbuild##' This implements the SVDimpute algorithm as proposed by Troyanskaya ##' et al, 2001. The idea behind the algorithm is to estimate the ##' missing values as a linear combination of the \code{k} most ##' significant eigengenes. ##' ##' Missing values are denoted as \code{NA}. It is not recommended ##' to use this function directely but rather to use the pca() wrapper ##' function. ##' ##' As SVD can only be performed on complete matrices, all missing ##' values are initially replaced by 0 (what is in fact the mean on ##' centred data). The algorithm works iteratively until the change ##' in the estimated solution falls below a certain threshold. Each ##' step the eigengenes of the current estimate are calculated and ##' used to determine a new estimate. Eigengenes denote the loadings ##' if pca is performed considering variable (for Microarray data ##' genes) as observations. ##' ##' An optimal linear combination is found by regressing the ##' incomplete variable against the \code{k} most significant ##' eigengenes. If the value at position \code{j} is missing, the ##' \eqn{j^th}{j^th} value of the eigengenes is not used when ##' determining the regression coefficients. ##' @title SVDimpute algorithm ##' @param Matrix \code{matrix} -- Pre-processed (centered, scaled) ##' data with variables in columns and observations in rows. The data ##' may contain missing values, denoted as \code{NA}. ##' @param nPcs \code{numeric} -- Number of components to ##' estimate. The preciseness of the missing value estimation depends ##' on the number of components, which should resemble the internal ##' structure of the data. ##' @param threshold The iteration stops if the change in the matrix ##' falls below this threshold. ##' @param maxSteps Maximum number of iteration steps. ##' @param verbose Print some output if TRUE. ##' @param ... Reserved for parameters used in future version of the ##' algorithm ##' @note Each iteration, standard PCA (\code{prcomp}) needs to be ##' done for each incomplete variable to get the eigengenes. This is ##' usually fast for small data sets, but complexity may rise if the ##' data sets become very large. ##' @return Standard PCA result object used by all PCA-based methods ##' of this package. Contains scores, loadings, data mean and ##' more. See \code{\link{pcaRes}} for details. ##' @examples ##' ## Load a sample metabolite dataset with 5\% missing values ##' data(metaboliteData) ##' ## Perform svdImpute using the 3 largest components ##' result <- pca(metaboliteData, method="svdImpute", nPcs=3, center = TRUE) ##' ## Get the estimated complete observations ##' cObs <- completeObs(result) ##' ## Now plot the scores ##' plotPcs(result, type = "scores") ##' @keywords multivariate ##' @references Troyanskaya O. and Cantor M. and Sherlock G. and Brown ##' P. and Hastie T. and Tibshirani R. and Botstein D. and Altman ##' RB. - Missing value estimation methods for DNA ##' microarrays. \emph{Bioinformatics. 2001 Jun;17(6):520-5.} ##' @author Wolfram Stacklies ##' @export svdImpute <- function(Matrix, nPcs=2, threshold=0.01, maxSteps=100, verbose=interactive(), ...) { missing <- is.na(Matrix) temp <- apply(missing, 2, sum) missIx <- which(temp != 0) ## Initially set estimates to 0 Matrix[missing] <- 0 ## Now do the regression count <- 0 error <- Inf while ( (error > threshold) && (count < maxSteps) ) { res <- prcomp(t(Matrix), center = FALSE, scale = FALSE, retx = TRUE) loadings <- res$rotation[,1:nPcs, drop = FALSE] sDev <- res$sdev ## Estimate missing values as a linear combination of the eigenvectors ## The optimal solution is found by regression against the k eigengenes for (index in missIx) { target <- Matrix[!missing[,index],index, drop = FALSE] Apart <- loadings[!missing[,index], , drop = FALSE] Bpart <- loadings[missing[,index], , drop = FALSE] X <- MASS::ginv(Apart) %*% target estimate <- Bpart %*% X Matrix[missing[,index], index] <- estimate } count <- count + 1 if (count > 5) { error <- sqrt(sum( (MatrixOld - Matrix)^2 ) / sum(MatrixOld^2)) if (verbose) { cat("change in estimate: ", error, "\n") } } MatrixOld <- Matrix } tmp <- prcomp(Matrix, center = FALSE, scale = FALSE, retx = TRUE) loadings <- cbind(tmp$rotation[,1:nPcs]) scores <- cbind(tmp$x[,1:nPcs]) ## Calculate R2cum R2cum <- rep(NA, nPcs) TSS <- sum(Matrix^2, na.rm=TRUE) for (i in 1:nPcs) { difference <- Matrix - (scores[,1:i, drop=FALSE] %*% t(loadings[,1:i, drop=FALSE])) R2cum[i] <- 1 - (sum(difference^2) / TSS) } result <- new("pcaRes") result@scores <- scores result@loadings <- loadings result@R2cum <- R2cum result@method <- "svdImpute" return(result) } pcaMethods/R/vector2matrices.R0000644000175200017520000000252314710217306017326 0ustar00biocbuildbiocbuild##' Tranform the vectors of weights to matrix structure ##' @param object an nlpcaNet ##' @return weights in matrix structure ##' @author Henning Redestig ##' @aliases vector2matrices,nlpcaNet-method setMethod("vector2matrices", "nlpcaNet", function(object) { netDim <- dim(object@net) posBegin <- 1 posEnd <- 0 result <- list() for(i in 1:(netDim[2] - 1)) { wSize <- c(object@net[i + 1], object@net[i] + 1) posEnd <- posEnd + prod(wSize) result[[i]] <- matrix(object@weights$current()[posBegin:posEnd], wSize[1], wSize[2]) posBegin <- posEnd + 1 } if(posEnd < length(object@weights$current())) stop("weight vector has too many elements\n") result }) ##' Tranform the vectors of weights to matrix structure ##' @param object an nlpcaNet ##' @param net the neural network ##' @return weights in matrix structure ##' @author Henning Redestig ##' @aliases vector2matrices,matrix-method setMethod("vector2matrices", "matrix", function(object, net) { netDim <- dim(net) posBegin <- 1 posEnd <- 0 result <- list() for(i in 1:(netDim[2] - 1)) { wSize <- c(net[i + 1], net[i] + 1) posEnd <- posEnd + prod(wSize) result[[i]] <- matrix(object[posBegin:posEnd], wSize[1], wSize[2]) posBegin <- posEnd + 1 } if(posEnd < length(object)) stop("weight vector has too many elements\n") result }) pcaMethods/R/xval.R0000644000175200017520000002764114710217306015174 0ustar00biocbuildbiocbuild##' Internal cross-validation can be used for estimating the level of ##' structure in a data set and to optimise the choice of number of ##' principal components. ##' ##' This method calculates \eqn{Q^2} for a PCA model. This is the ##' cross-validated version of \eqn{R^2} and can be interpreted as the ##' ratio of variance that can be predicted independently by the PCA ##' model. Poor (low) \eqn{Q^2} indicates that the PCA model only ##' describes noise and that the model is unrelated to the true data ##' structure. The definition of \eqn{Q^2} is: \deqn{Q^2=1 - ##' \frac{\sum_{i}^{k}\sum_{j}^{n}(x - ##' \hat{x})^2}{\sum_{i}^{k}\sum_{j}^{n}x^2}}{Q^2=1 - sum_i^k sum_j^n ##' (x - \hat{x})^2 / \sum_i^k \sum_j^n(x^2)} for the matrix \eqn{x} ##' which has \eqn{n} rows and \eqn{k} columns. For a given number of ##' PC's x is estimated as \eqn{\hat{x}=TP'} (T are scores and P are ##' loadings). Although this defines the leave-one-out ##' cross-validation this is not what is performed if fold is less ##' than the number of rows and/or columns. In 'impute' type CV, ##' diagonal rows of elements in the matrix are deleted and the ##' re-estimated. In 'krzanowski' type CV, rows are sequentially left ##' out to build fold PCA models which give the loadings. Then, ##' columns are sequentially left out to build fold models for ##' scores. By combining scores and loadings from different models, we ##' can estimate completely left out values. The two types may seem ##' similar but can give very different results, krzanowski typically ##' yields more stable and reliable result for estimating data ##' structure whereas impute is better for evaluating missing value ##' imputation performance. Note that since Krzanowski CV operates on ##' a reduced matrix, it is not possible estimate Q2 for all ##' components and the result vector may therefore be shorter than ##' \code{nPcs(object)}. ##' @title Cross-validation for PCA ##' @param object A \code{pcaRes} object (result from previous PCA ##' analysis.) ##' @param originalData The matrix (or ExpressionSet) that used to ##' obtain the pcaRes object. ##' @param fold The number of groups to divide the data in. ##' @param nruncv The number of times to repeat the whole ##' cross-validation. The deletion of diagnols in 'impute' is ##' deterministic so result will alsways be the same but in ##' krzanowski where cv-split is obtained by sampling it can be ##' informative to examine the spread of the Q2 values over several ##' CV runs. ##' @param type krzanowski or imputation type cross-validation ##' @param verbose \code{boolean} If TRUE Q2 outputs a primitive ##' progress bar. ##' @param variables indices of the variables to use during ##' cross-validation calculation. Other variables are kept as they ##' are and do not contribute to the total sum-of-squares. ##' @param ... Further arguments passed to the \code{\link{pca}} ##' function called within Q2. ##' @return A matrix or vector with \eqn{Q^2} estimates. ##' @export ##' @references Krzanowski, WJ. Cross-validation in principal ##' component analysis. Biometrics. 1987(43):3,575-584 ##' @examples ##' data(iris) ##' x <- iris[,1:4] ##' pcIr <- pca(x, nPcs=3) ##' q2 <- Q2(pcIr, x) ##' barplot(q2, main="Krzanowski CV", xlab="Number of PCs", ylab=expression(Q^2)) ##' ## q2 for a single variable ##' Q2(pcIr, x, variables=2) ##' pcIr <- pca(x, nPcs=3, method="nipals") ##' q2 <- Q2(pcIr, x, type="impute") ##' barplot(q2, main="Imputation CV", xlab="Number of PCs", ylab=expression(Q^2)) ##' @author Henning Redestig, Ondrej Mikula ##' @keywords multivariate Q2 <- function (object, originalData=completeObs(object), fold=5, nruncv=1, type=c("krzanowski", "impute"), verbose=interactive(), variables=1:nVar(object), ...) { type <- match.arg(type) if (inherits(originalData, "ExpressionSet")) { set <- originalData originalData <- t(exprs(originalData)) } if (is.null(originalData)) stop("missing data when estimating Q2") originalData <- as.matrix(originalData) originalData <- prep(originalData, scale=scl(object), center=center(object)) nR <- nObs(object) nC <- nVar(object) if (nR != nrow(originalData) | nC != ncol(originalData)) stop("data and model dimensions do not match") if (fold > max(nR, nC)) stop("fold must be equal or less to max dimension of original data") if (method(object) %in% c("svd") & type != "krzanowski") stop("Chosen PCA method must use krzanowski type cv") if (method(object) %in% c("llsImpute") & type != "impute") stop("Chosen PCA method must use impute type cv") if (is.logical(variables)) variables <- which(variables) ssx <- sum(originalData[, variables]^2, na.rm=TRUE) if(type == "impute") nP <- nPcs(object) if(type == "krzanowski") { rseg <- split(sample(1:nR), rep(1:fold, ceiling(nR/fold))[1:nR]) cseg <- split(sample(1:nC), rep(1:fold, ceiling(nC/fold))[1:nC]) foldC <- length(cseg) foldR <- length(rseg) nP <- min(nR - max(sapply(rseg, length)), nC - max(sapply(cseg, length)), nPcs(object)) } q2 <- matrix(NA, nP, ncol=nruncv) for (nr in 1:nruncv) { press <- rep(0, nP) if (type == "impute") { seg <- list() nDiag <- max(nR, nC) diagPerFold <- floor(nDiag / fold) suppressWarnings(diags <- matrix(1:nDiag, nrow=diagPerFold, ncol=fold, byrow=TRUE)) if (diagPerFold == 0 || diagPerFold > (nDiag/2)) stop("Matrix could not be safely divided into ", fold, " segments. Choose a different fold or provide the desired segments") if (nDiag%%fold > 0) warning("Validation incomplete: ", (nDiag %% fold) * min(dim(originalData)), " values were left out of from cross validation, Q2 estimate will be biased.") for (i in 1:ncol(diags)) seg[[i]] <- which(is.na(deletediagonals(originalData, diags[, i]))) if (verbose) { pb <- txtProgressBar(0, length(seg), style=3, width=20) } j <- 0 for (i in seg) { j <- j + 1 if (verbose) setTxtProgressBar(pb, j) test <- originalData test[i] <- NA test <- tempFixNas(test) if (method(object) != "llsImpute") { pc <- pca(test, nPcs=nP, method=method(object), verbose=FALSE, center=centered(object), scale=object@scaled, ...) } for (np in 1:nP) { if (method(object) == "llsImpute") { fittedData <- completeObs(llsImpute(test, k=np, allVariables=TRUE, center=FALSE)) } else { if (method(object) == "nlpca") fittedData <- fitted(pc, data=test, nPcs=np) else fittedData <- fitted(pc, data=NULL, nPcs=np) } ii <- i[ceiling(i / nR) %in% variables] press[np] <- press[np] + sum((originalData[ii] - fittedData[ii])^2, na.rm=TRUE) } } } if (type == "krzanowski") { rseg <- split(sample(1:nR), rep(1:fold, ceiling(nR/fold))[1:nR]) cseg <- split(sample(1:nC), rep(1:fold, ceiling(nC/fold))[1:nC]) tcv <- array(0, dim=c(foldC, nR, nP)) pcv <- array(0, dim=c(foldR, nC, nP)) for (f in 1:foldC) { test <- tempFixNas(originalData[, -cseg[[f]]]) tcv[f, , ] <- scores(pca(test, nPcs=nP, method=method(object), verbose=FALSE, center=centered(object), scale=object@scaled, ...)) for (p in 1:nP) { if (cor(tcv[f, , p], scores(object)[, p]) < 0) tcv[f, , p] <- tcv[f, , p] * -1 } } for (f in 1:foldR) { test <- tempFixNas(originalData[-rseg[[f]], ]) pcv[f, , ] <- loadings(pca(test, nPcs=nP, method=method(object), verbose=FALSE, center=centered(object), scale=object@scaled, ...)) for (p in 1:nP) { if (cor(pcv[f, , p], loadings(object)[, p]) < 0) pcv[f, , p] <- pcv[f, , p] * -1 } } press <- rep(0, nP) for (p in 1:nP) for (fr in 1:foldR) for (fc in 1:foldC) press[p] <- press[p] + sum(( originalData[rseg[[fr]], cseg[[fc]]] - (tcv[fc, , ][, 1:p, drop=FALSE] %*% t(pcv[fr, , ][,1:p, drop=FALSE])) [rseg[[fr]], intersect(cseg[[fc]], variables)])^2, na.rm=TRUE) } q2[, nr] <- 1 - press/ssx } if (verbose) message("\n") rownames(q2) <- paste("PC", 1:nrow(q2)) drop(q2) } ##' Simply replace completely missing rows or cols with zeroes. ##' @title Temporary fix for missing values ##' @param mat a matrix ##' @return The original matrix with completely missing rows/cols ##' filled with zeroes. ##' @author Henning Redestig tempFixNas <- function(mat) { badRows <- apply(mat, 1, function(x) all(is.na(x))) badCols <- apply(mat, 2, function(x) all(is.na(x))) mat[ badRows,] <- 0 mat[,badCols ] <- 0 mat } ##' Replace a diagonal of elements of a matrix with NA ##' ##' Used for creating artifical missing values in matrices without ##' causing any full row or column to be completely missing ##' @title Delete diagonals ##' @param x The matrix ##' @param diagonals The diagonal to be replaced, i.e. the first, ##' second and so on when looking at the fat version of the matrix ##' (transposed or not) counting from the bottom. ##' Can be a vector to delete more than one diagonal. ##' @return The original matrix with some values missing ##' @author Henning Redestig deletediagonals <- function(x, diagonals=1) { wastransposed <- FALSE if (dim(x)[1] > dim(x)[2]) { # matrix must be lying down x <- t(x) wastransposed <- TRUE } nr <- nrow(x) nc <- ncol(x) if (!all(diagonals <= nc)) { stop(paste("Order of diagonal number", max(diagonals), "is out of bound")) } indexmatrix <- matrix(1 : (nr * nc), ncol=nc, nrow=nr) finalmatrix <- matrix(ncol=(nr - 1 + nc), nrow=nr) finalmatrix[,1 : (nr - 1)] <- indexmatrix[,rev((nc : 1)[1 : (nr - 1)])] finalmatrix[,nr : (nr - 1 + nc)] <- indexmatrix dia <- 1 + 0:(nr - 1) * (nr + 1) finalIndices <- NULL for (i in 1:length(diagonals)) { indicestodelete <- finalmatrix[dia + (diagonals[i] - 1) * nr] x[indicestodelete] <- NA finalIndices <- c(finalIndices, indicestodelete) } if (wastransposed) x <- t(x) return(x) } ##' Get cross-validation segments that have (as far as possible) the ##' same ratio of all classes (if classes are present) ##' @title Get CV segments ##' @param x a factor, character or numeric vector that describes ##' class membership of a set of items, or, a numeric vector ##' indicating unique indices of items, or, a numeric of length 1 that ##' describes the number of items to segment (without any classes) ##' @param fold the desired number of segments ##' @param seed randomization seed for reproducibility ##' @return a list where each element is a set of indices that defines ##' the CV segment. ##' @examples ##' seg <- cvseg(iris$Species, 10) ##' sapply(seg, function(s) table(iris$Species[s])) ##' cvseg(20, 10) ##' @seealso the \code{cvsegments} function in the \code{pls} package ##' @export ##' @author Henning Redestig cvseg <- function(x, fold=7, seed=NULL) { if(any(table(x) > 1)) { if(any(table(x) < fold)) { fold <- min(table(x)) } if(fold < 2) stop("too few observations in the smallest class") res <- sapply(unique(x), function(z) { if(!is.null(seed)) set.seed(seed) tmp <- sample(which(x == z)) seg <- matrix(c(tmp, rep(NA, ifelse(length(tmp) %% fold ==0, 0, fold - (length(tmp) %% fold)))), nrow=fold) },simplify=FALSE) res <- do.call("cbind", res) } else { if(length(x) == 1) x <- 1:x res <- matrix(sample(c(x, rep(NA, ifelse(length(x) %% fold ==0, 0, fold - (length(x) %% fold))))), nrow=fold) } res <- res[!apply(is.na(res), 1, all),,drop=FALSE] res lapply(as.data.frame(t(res)), function(x) c(na.omit(x))) } pcaMethods/README.md0000644000175200017520000000277314710217306015154 0ustar00biocbuildbiocbuild# pcaMethods R package for performing [principal component analysis PCA](https://en.wikipedia.org/wiki/Principal_component_analysis) with applications to missing value imputation. Provides a single interface to performing PCA using - **SVD:** a fast method which is also the standard method in R but which is not applicable for data with missing values. - **NIPALS:** an iterative fast method which is applicable also to data with missing values. - **PPCA:** Probabilistic PCA which is applicable also on data with missing values. Missing value estimation is typically better than NIPALS but also slower to compute and uses more memory. A port to R of the [implementation by Jakob Verbeek](http://lear.inrialpes.fr/~verbeek/software.php). - **BPCA:** Bayesian PCA which performs very well in the presence of missing values but is slower than PPCA. A port of the [matlab implementation by Shigeyuki Oba](http://ishiilab.jp/member/oba/tools/BPCAFill.html). - **NLPCA:** Non-linear PCA which can find curves in data and in presence of such can perform accurate missing value estimation. [Matlab port of the implementation by Mathias Scholz](http://www.nlpca.org/). [pcaMethods is a Bioconductor package](http://www.bioconductor.org/packages/release/bioc/html/pcaMethods.html) and you can install it by ```R if (!requireNamespace("BiocManager", quietly=TRUE)) install.packages("BiocManager") BiocManager::install("pcaMethods") ``` ## Documentation ```R browseVignettes("pcaMethods") ? ``` pcaMethods/build/0000755000175200017520000000000014710312420014754 5ustar00biocbuildbiocbuildpcaMethods/build/vignette.rds0000644000175200017520000000042214710312420017311 0ustar00biocbuildbiocbuild‹…R]Oà ¥-n®™s/>öôW4K̲Äc|%…9’ 6{Û×z»r];>páœ çž ¼Æ„P0‚e´„0qÛBÉ æE)­•êí…µ°é“j|b®kWHa†Üu•³p;Í{v,u¿é¥’÷N+‘eU;æ¤V>—1Ç’Fº]‚ÚXëA9£ywÿi°âÛOÜÐ`ÇŽ¥nÎzõt|ꥇ}^Ez1òËüOž†¯}¡X)¬ON¿;tÄ LpcaMethods/data/0000755000175200017520000000000014710217306014575 5ustar00biocbuildbiocbuildpcaMethods/data/helix.RData0000644000175200017520000004656614710217306016644 0ustar00biocbuildbiocbuild‹]œxTUÓÇÓ‘"ÒŽ l_Tæ*JG¤AD@é½Jé)Ò›ôÞçÒ;„@:i›ÝlOH„@Þ{Šfž×çy?öƒdg÷Þs~gæ?ÿ¹]¾î¥{§×;AAÁÚÿ´—!AÚÿ (SBû3tèàáÃ&kÿ^:  ä\Ww÷œS ZÁ¬Õw Zä=ƒ½{®h]ç]Þ{ où7øÿœc˹_ñ¯~ "ƒÒô¸ëj²}öÜWx¡ÿñò—jgâÉsà çlÆË£óϦ x‹'óï,,‰Û÷ï²ñPk\ñÅ_k¿ª‚§²ŠJ{õ3^ >±Ý³—x´w©w"¿ØˆwÔ¹åFþ¹ï©5®Ç“}ÂwÙû5^?5– ˜…ê{å®L4ïÅ­ù~¸y žÍ}§Á$ï·xáó-n<ú=^ß÷Ú>|±öó‹æ?ýÐTOëËîj™}Õy'oÀûx½l›7?…wó§¾PmÆës;÷:S¿ï ìÒhtÅkëί¿™6ï•ßרiw^‹s´øª9w¬Òøhð›Eñgó0ê£AÇz]ÆÇ%S’£<ÀÇ ¹ž­ƒßùbôŽ§ïࣻ«š7¼9£½§ƒ|¼7<6+û㽧ul¨í9Þií~»qZ>ùiCؽí Ó«Ì£r0úf™» ûoÁGí¶•ªõFZâíü¹Ÿã“ [§ÍÃèÈ!Kº¦cœ²ýß2Ò0ÆðdÛÀŒ||ìý¶^ô¬1fÏü+õfmÆù½â«õªû}p_Œ]ãµ~Z#ãÊú¯v¨²é˜ï®„1A¥ÚžëoħG’æ¼™†Q®U´ýCŒÿ£ÛóIßÏÆ']Óž|´ «˜Ò\ÕZbü7½˜ÙñÆ<0?t臘Ô~Ó©y‹]ûUì•Ý’1¶àÉݮߕÀäÍ7ë^MvàÓ„Ú‘¶_á³N‡|Tç&,÷æ¹gÝľœÞÅ'Ú§Ytý2&¬v¼¸ˆIß.?Ròn&&Ô<¼Ýßg >ûì§>ÕD`úÅŸëOÿæ>¾¸åÚÙa˜2oà+Õù˜ú×—?|“rÓn è1Ûƒ)ONõùkV[LÑVÇì‡iâ>ø£q¦¦¼ÿòäBL¯Ýµì‡¥Œ˜¿³pg#f´œµà»˜úÛ+öI1½ä’×…ƒ¿Äô–ïš:y8f¨.OAË·˜1Îy|ÈÖW˜¡-†A.ÂŒ¸_~ ¬û¦Éû”öàpÝj•?Ãô¬÷^mý f^L}qZM̨ðlp÷럢­J©ìYïLÇŒÚçÝûÏìÖÞÇŽ1íÄôêö «23Jžò ™íC•_Ó mQý“Bµß;qúãÎëÑVp€­DÌLd7¦/¦ Q{× ÁŒ!•V¿€6W³ÏÒ>íöß¾¨v9ÀŽöùMz/|VÓÿž¹doååè`»¤^ ¦³»øñ.Ì|±¾s½“.Û×ÍÀôZÿ´{½2Ú²™áZÜQ«Qm3Ц/ñå­ZÃÑi:óêh³ŸÑ±£Jû„­»0‹­ò/Æ M®ÛÐ÷µKö÷¿ó%núíò¾ÛòæîúlY´ßmѳë7fNªöç{Ó¶ {ÆþNw¾:ŒÛ„æ-Ô&è8þÅá¥ùÚûÊuë¿jJyÇû˜9òɤÑS2ÑycX茊sÑÅ>U§­˜5àË…æö7ÑYÑu¸ÌG?¡«^Öäè7¹˜u£BíÝÏó0«Ì6óó%ÝЉ›~Üvòzxßûü¸ ³–ž9§m1tv2O8S>ìÿkøf½I¾¸¢’½Yì”D—¶©G7žw«Ž™°¬6ºÛ ;½¿zµ¿ümeº7~|u`yt?)Ív:o-b;Ý•öŽóüGÌ’ûÖ•õmúªTz‡L ßÒû9fÝØ@Cºf1¢¡#añ1¥Ž£KòÒÉñØݽ¦|7szö7´´®=¿¦ÏȲºÐ“?­¤µ2ºÏLYä;®ý{¾¶£<èk<úþ­ èûñ£ÔùëÚõýr|‹u-Лöi¿Ô¾{Ð|:« zg š°½÷·Ìͨ{}|[@ψ6ùCÑÛƱ§zæFÌ®Ìo º#ª}ª÷å OÛœ}®EßÒN 9è{½wÎ/{]èŸðâM­¢OÑ/¹êÿõ»#9ÓŽ£èØ#5¯mB¶Ù->#úµMøËoèÉ;¸n÷Ör蛲ûdÉ÷h÷£cTö¾Btïhòõʆ Ñ_{h‰èµ•Ð_±WZåÁkï[=ó¯èÑ{áÒÀ‡o ïƒ6J꘿ѧQíDj ôë7•¾ôl'f/ù­A’“1»¿¦{Vbö®²›R&ä£Ï^Ô«y׳è;½¿nìÓÍè½S‰‘Rûu=‘]V£?(9àÑéŠèedz-}=b½³³§F­6¡_Û5ÏOµB¯<ï|óîÄL»™Þ?Su'ûç §p¹Î¼è(ú¯ÞöÞ}ýú%½æÛeÕÉÁlÉCi¢ èÿùá—9åžbvNˆ=„Þ²Zëº\D/£ÒXŸ7¸Ub¿ŸÐÿÑöM“þú½9¸ÑWjófC¯ûè¯þù&îžè«Ìlô<ÅȆ^§qõWO¢mVê‘…[Ñ'¹çöõKõßÐkpýUqDú%/½ç -£­¯í†Ÿ`ô³¥èmµ`êÒŸbÑWîàìƒÎ…è=uhFá¹èÕ¨rc}yë7gWl‚ž¿ÙAkC_D301ZÜJ64½„~y¾{6/ølÇ×_ ÷«pvr ·ÙàÇFíE¿v(~³Xû}Á_ïÕ¿ß Èê„ž“Ç¢£ƒÚ£'„-ÔDô¶ëظñ›“èIϼvýOm¥öÝ}'ø:z#<¶öm¢Ñs¯nxÍß¡§eÅ×|Úú)É/ztµW¾†=è ,蕯­·o[o¨zÝcNLþªéô$<[Þº©‚à`@¿] ´û–¼zËÁ'èìÛÅþɶЭÁiœz=üëˆ^Ž±éè½ø2w\H:Ûû7ô9<=ìÉEè\^½EL}zÞë|௷5Ð}fEÌ®³§Ž_¹ê§ªè˜tÀ›’«½ïŽÐ®õ*¡ó«o´¡}N¹Þ]'¤õÏCWÓ_W )Òîg»%:¬Gç¿œëÿCSmIcÖõ–¿.ߨñäïKo°¢c~•Ñ=>E—•ä1zÇ¿¿3ꘊÎ+ßø¾z>7•5„8ÑŶúÊtì^}èMÌŠc ôcV핆Í팮/î~÷tüž0ñøùyèHz’Ò!t:¥~ùçªëKÖAûߎ¡ú—Œ…Ÿi© Úo_‹_ìCûìªÕW~‹™S¤ÜŽ{Œ™Ýnÿî5 3Ù·×-ƬР;F6Á,Ž÷˜)÷™ý©áÓ„“[ÐÞ?›4è8½j|ç*è¨Ì.híÜ>—dú r8XÖÝzÚµ¤nB™Ž˜õó+}¿:[ÐQïËã NÅ¢cÉ°&#ç¬ÅŒKì†BG[ø#ÑÞöçi?Tõ¡£æhFT̘þÍÝ µŸct¿mü¯»a†\ǶÎ-#]£1#¸š3§ó´iÙÞÒWÐÞjó¸¯<çÑöIž¶ÂzcFÄÕç)æw1óæO‹æ|Œ<=ª¡->ùÑé÷ù}¦h‡sç–ˆéZv«¥˜ÆÀ‹é χ¸1Cæ-©òNc«yÐL]Pöh\ãY˜–Ø旨ģ˜Æo3¦¼{¨ÅðZ|¶|Gcª<¯ÓVßb;S%ÿRã.7wEKLš¸PËDÞ†ý&Ï®ZSY68y6¦k›}¢ÑŽ©Ú*É83Ód‘p“SÒøAÉ,+ÜÉüXí„i‹t¹mËmÀ”_ËÞx”Z:¬ÿ!ìÅ|vmäa˜ÒùÏímÖ\äÉ=Y¦‰žóZfåÀ'³ÝÑûˤ`R=üc˜ûLbØ~ñm,pŒmÊ`Œûå˜ð ?ßÃÄÊŸ5zô/|ʪƒçññЫ^8b<û±â06žwôf+ŸZ+k+£!ÆH<íáÔ2â9uèu£C Çc̪eŒ¬Se‹Ùß¼‡OµCÛªº1‘e==WcBÓ5Õæ|p£:6׶.Þ¯Îâ#m—WßsoiIY•R9øHæéwg?Õ*ž7ø°ãú6³ðVÅ?ïY‹žàc-ëøâÚY‚<øHÞטé£.Öÿ¢/>l²~\Ñêxÿ·ò»ûjbÔüÞírý±xGî›ë2O¿;†]ñÆ`–°4Å3;։܊*«Jš~€w_}áUxmço,ÃkZÕ©¥xýŸç–1Öâ5–ÝüXÏ_â •Øxrr5¼üûoõx6“tŸàÅΟ޸àèŠ;ÒV‡®éÄ«!Á¡þ¼"ëÄÃÇcSæ[ð°všø6Ã}1ý½Që ž ­ä½y÷5«}úœ;Wí¹Øæn;níSÒVfÅ82tbjòÛ°öæý…‹?Ý ë–NØâXúwêCJ¾ý Ï}ؾFZËž8=óEËKáüÓž/O÷™€m6]>ì:üÍËã°OæÑN}<©g§±¸ïU¨õõÎΰ&¯cuW áoèRX¯ÇØó½-.£Ô<¸T6áκ•{àöF8UuDËHë08üûÓ¯ó÷€+öv²Ã‰Ï4è5γãhÁY¸!ø WÙic=7D~ gÇvÕÙ½îÞÿaÃÕÅpmSÇ}-œp¥`ŽcÆÇÓAÛDŒ(p…}ÜáÏàÖ1v€ÂíعìĆÈZö+Üû°Ö'“š­€«Úé”SwDÞ7îü³¨¯í<¨8‚e²pãÅšVãfÇBO“‚á Û-mw'j@¤üûûbÿCô™]¾ÿª<âiR{x¼ìö÷E!=á¾Ýùƒ–ÁŽéÊ%êqx"ÎaˆaeÊN<Ý•Þ5ÿA$¤à'ñãO¶ÕÖJ_ˆ~þÑäJ-„Ø¿>þyEÏG%ö?<Ü€Ø1ÉÃ~u ž Ý‹¼’&Žíû…ž2U¡Â3ˆÎºw:ºâ×7†fpõ]vbBÂR¶€CÜ4^¨A¬%þóZ•~ƒ¸Áûö†¬ ‰{YA» âÙ©VÆ1Zu7ýÐHíVê沿…$¡@RÕìDgó¿îCò ‘Qv|&$É÷Ü€-k]°|>$l`ÕJˆ›YúqlïÛò]ºóroBÊW؃–õ« i,Ûþr!¤jEñ({eÈù(¤Zúy–Þ’D¾I~º½ßRH9:ZÛ¡U 󃱺s†´ØƒÿüÝ}#dˆ|R—óƒRŸtd‘!-§³V™ì´ðv¬’„ Á H:¤Íý|@Ðß!=⯠ u[A:©üX ‹çÌܶõ!¤f?ötù­ ¤ý6=,uãJ°=«÷ëÎÕ·!½–úuƒŽo ]K.ëwé6ýÑðñC@fÈWóž=héIL¨j¶÷ÎJÔ…Ìf¼ð‡t‘Ç‚m‘ïØþ´?ÁƲ¾ê|ÙׂL¦’|ÿ Ød‰æÈÈÉX¶hL2Ø7¾ºÑ¨j_ÈZ7ädö¥<°s,}™Z5ñ*á#°}sêêÄ’/ S~nGZËël‹À!×uæÈ9kþêÑ ì⼇Ìž¸€£p°ro²,ëþÜQ—CÖvVHWÇ^žxhŸ‹'Ä`—û6KÔÝà`ßÂ:œl[ÿó=Y42{ñƒì¢.‚,‘߀Ëûudù}õòÝ«æZ1à”ë×ác• œå˜Ñ\ÐNŠæÅ囑àêóû«“ÖƒGèK%ôÈzsBÛqµÀ͇oÀåË^Ü,ìƒnv:Ùö8µ,·ðeip²»ò]¸ÄyN.G}YÝ}&¹âÀÝyS(ÀÉÔ¯ÏKoáÀ/´ n¶ío€Kœ»à~Î.Äeð·|Ûwå#í:½—1eå…à¾ã?cVp—~Ã2@ðÈuîyÈØžàdrήà”ýeÍïUðü9¨ÿø†[ÀÍVGõnàÎh¾gã¹DðÉï噹{bMd ܼ|h Þkó~r÷}p³múËð‰ú ¼BG÷¡ ÊÒe·Á+×…ë2O<Á›RûýzoÇ‚—¥ñ—߀ËþÏ„Ÿš6›ßiðÊïçYÃ1ððcQ{?F™Á=ÀÏÔÌÞ'ÀWÚñè½ÎÁϲà´¿gx”ö{¿÷Û>v>øµ‹òÙ}xãË598ü$øu<ÏÔÔÖ‡ÞŸÞ†<_­Ù}ÂwŸÐGÀ·Ýßgô²dð.¾<ìÊgSÀ'òPðÊõàæò#€w!æÀ_í«´À+ÎSðhUDÅË—€};­”ï&HÖ/›¿?Ǻvý,¿—ÀMà¶%콌©àÿ¤øªð |Í/¬ý“ü5¯mÜ•Þ üò¼ð²ÝQÖ^‘G‚Ë!£ÁÏeæàûùóñ¯»m—ÓÆA¶Ü/ÞÏÊLjµ`øOÏ2Œ¸þüBGßYuøÃX¢[üwت~ùû¾ú0¼qÝÇ´ßt¤.>ŽO3xµÝÏŽƒ•¢ú'_è àzøyš¦]gyß}ëÊ?®ºqøÔ'y_Ÿ¯ ~VUý9ò|–º ø¿±° ¾#?=zM~VÍ—^¾M&‚ÿ ;` à°ÿ§+/-à¿Çþ»¯}~&̯ïw×YÅ Ù¬¼]”^VæU¼^m‘4é½ü}yá ž½ ØÑà¯7ksÛÎs!gK€Oä½àg²Èk;øØvªT¾¼&Wÿ¢Ý碖_ni >¦¦6^yŸ=‹ÿÔ2ëãà™Óø’vôƒ[[,†ËûÀÏd Œåàõ8øêñ‚ \,}øexù±Õ¼¼|ú¼lU*4îh«²ÛIÈîÙâÓÔ2àºÕ×Í[}«hÁÃ>MàwàÝR•(Ú÷á:#ødžäyË pkŠO·ƒ›©ñƒß‚§Õ;»;4™ É/w—€Ñ_Á­%¯m”TpœñÓ†°ûà–ëÍóGc¶‚À½—olðVb‚cðj§ìÌxp›x" ®—ö¼0|³9ÜZUñqÆ`ðñô¾78™¬ñ4œ2OñJN»DŸ¼ÃFöÚÑDã©ëZD¬{¸D=¦ý)Îßfp~øwà™Ÿ3ÀÅT·nwÀíÙÍ2ðj»Èµœ=ß?{wh[°kÕÿK.‚ëjí†/€¬ê¯W/Ø™:Óë'°u\®Aàhú8lûÑÑ`_ìÝ^âÈdÇuV>Ø®œe?öÓk?ùèÛs`û½Æ™Î®ß!Eæõé·Ú÷»7(Ò2[e÷¨\ÒµÍó]›Š`‹Y¡ Ž‡t¹®lr¥_;²e}’–ŸXY'Òå:Ìð4Xb>¶Ë™‡ï¥A:o_4‡4ÞNøl2ÏI•y{òšk­µÒz´ :ñÞH]Poÿ²ƒ eUc̪^ƒ«àÙÆòô³þ I×úž¿rö[HVeÂý’§µ<ï{¶Ó •oß~*¯{ºÐÁ´ïÇ0ð¤Éýš,óæ´ó:.W”†x¡3C ìN.œÚRXº½`'<ã²G;HfYß;»á™ßóÅï›–Bb!KHjC’–å·»Sqr¿Æ‹~$0yyT$²«Òm"<åÇç—ÿ=®þ„'ZÓ2ë'H’çO© ó¶f”…x¡³C< ³d5Äup™T¥6Ä Ýb°78ÉÕg`›™÷ V«¶þIVàá Ö`L‡(Öõ¬}¢yzðÄj‹¥W™á +&v?ƒ'³n² bò;‡Ät~ 1WO¼ÐRdˆj°íÙYËFx,ú/ð„ËÁC!JËz?8Z ¢˜ŠTw4<âÛ²<ÖVñ¾½¡#Ïÿ»U}möpÁ=¡¯ÀÃmüƒCT˜5&cÇ\¸%úSp[ô}àÖŠORÞ­:î} ®°¯_æ\f*wï¿!úóö%²–>€+—¹€7µäpZ¢î¯oWúË’ƒà*oÕ‡»Z£¥Êp£çëaÚ?ÁÕ›;Ö׬önŠúneÂéh¸-ô#¸/×óÕd8ÿôX0œªcdä‚“ç-¢v7Uô£áJVœú‘ áz-_„¾ÇË÷žR©rœn±pï¢uá„õ¹¦å*ªö_<ý± ì©=#hQõBXYoЗ_~ Ge ‡kŸÁ™-ç~¹3Å»~õöXô9 lkvgƽ׳ìâÇí°£‰»GÌÃ`uÁÑ’»ëÆÕÚ·ˆyµ w©e©r³»óÄý\&/~¶º>|Žþº‡2›½Ü†î™Ý[šv½¼½ññ}ï¿ú¶ì¯dkÉdS:zcêÛjõ†Ù_þ÷}³µt}EÑŸèÙ´mñ“'£oö•ôE•Ñç½ûºÂÉñèùúµjóêmú˜œ°é.Êü}üöD¯Ô¯¼<=Ó~þ]†ô±öæú™ªÚn z¥þdzêÀeèarßkÐÇåÛÐ{˜7rÐ;† ÞUпñ‹ð[íûkï'úg¾fËY$ô ~¡ûV©qÿÜn¾¼¡>Öûizx]}L¿v=?óÿг„ ÙÐÍ–MLºEŽÞÛ¬q6=¬ }/GW-3ÓÞW^/¯Ô{½\ŽŒ@öá'Liˆžé\Gß'¼q^–éî K+–[dž¡Wè ècryÌJôp[À—è©¿ÛlºŸƒ¾Ø‘lÅ£×r¤Ý©°3è*Á„Ûèæe`gô±tæbúÊŽõ½c3¡[êƒ~‘'k߇çèïÎ/zJò†úÙ_‡å£—Uѽ—7Ñý[— ÞAwäy¦l¢›©ÞÝ£ëGÞ¨@7k_9¡ýyþk[©Б;ŽeÐèaex¥½èÔŠÙûcÖ1&ôvAÏÊ®S þLE'_Fu0KƒX͸5è²²Fƒ]¼MSÚ·×v"ºyzÚ]¬½¶ø: x¡„Î¥ü`Ä,¶‹?)YÌ…Ñ0óáǯú}»í­ç}ôÞÃÙ˜Ån˜µ¯‚ùï}¿¡=©‰–™OB§àfõZfZÑ m²/éz8eúÕ­ÿéöÎO&i+(ã†k'ð.t±toÊ´Ëþ@ë*‡•A;/óÇ`V‡§m믴Ï]Áœ èýu?ÆCûX³Ñ.}!Î}\DÇ0&œ5@»Üw™v.p`Sý#£CË^Ú®™Lþ¸aGû§DÝ<³mZ–w©#fæõ>9ýw´}ÚƱ§ÚD=‰¶x…öf/·jHE›¶ ¤ÜAÛeÓGcºÔïmrŸ¦3ùÚö7¦¿á Lç¦1÷C$LexHsa†80“·%J`š¨Û0UÔChë7hÁà5‡1}L©ck³Ò0“—³?`ú¼àÄ´¹™t˜*ûãéoYâ» Óe?1 †ä&<[ŽiM/idXŠiƒ/m=÷a{Lã6—/0ÙÞsÕ²R[0MûöÚ%ÇÔ©Ö9÷`ªÈ1íý—'¤ÃDvÜ÷`jC¾@1e30åß÷ g`ø “e*Žƒ&S¸ô>ceÝ”U˜°§/Û˜$õáÄŽ…éÞ=„ lU•oŒ‰3Þê¶þü9ÆU¸òýåñû†öw1NòáYÁå¨Í7ëa¼–%û=-ð©ã›>:ƒO¥(Aò8^ú‰¥_#^öažðmXSê2eqÆj§:œÅ¡ àcÙo"ûpѲ+ò*ŒaêeiÄ[²Ïþ˜Ë¨“1Ž—‘V¼?‰„›ñ+{>ÿ£Ú®~ç“S³ž VÉÛêu0ºýG`´ôçD4ÌþµìM¼?‚`É®rÿãÿéí‘\>ëƒ÷YW§w2F¶?z¸ùûñ.·×4ÃÛkwzw^G¼)ÎG¼ÇÚƒµ_ãÃYVgRo¼'t,¼ÉÛd5ð¾Ð‘ð*³5äœÇ›²w‰·Çá5.#uÆk²/~fvãŒE½Nà-?3îTûZ6OW‹oܧf2^õ.^õâñ.LÉÄ[¼Ü™Œg¶»Û}g\8øWl<ÏÚé Và¶Z­¶í x‚X{λç^r[§lé‚«ì­•³á=<øîÃ{ß/½Ž§fï;swYlÜU% o«.¸ÞÁï¸üTùéÉaãø¤—#?k KyšRÿY=´~Óc`µ¬‡wžÚªâgóq¿ñÉ´íæ8>}Ûë!λ°iä‘ÓW¿ëǺ˜ÐbôopbûD¦ÈÃÉu5‡vïꃳK"N÷ß=Nï)Ô¹lÂÙ6“†~0ì!ŒHìÖ°cÀO0îá"ϧÃëÀÆ~ÁeºŸÙ»S{M(UõO¸ üWpIìO¸Âèøáß0»þ®“¬Õádhøî‹#ÃÉ*ëÀÁVU4;·4˜ïºŽ_Òˆ95 d¿®K½êžVåôJ« 7e}ýXôwážÈ#à»ý¯×@$?žÃàá‹™/¾Ì…k¼Ý7TŽëoá¡ÌçjÙ]½µë fî«?ª‘â<…(© =fmÅËk z{‰ý;V†H©C?<ÍR¥Ußï­ø1ª5© O¸}§)<þ/ˆjTµïôØ pëÃJ{n@ oßt‚;Í™¨%<(hº' 1—ðã2î_5å}`ŤîÝ Ée»ˆ–õV,Ëþ¿Šƒhá#„¨u „ŸB+ó;-†'"_ƒ'e¸` 1â„-éúõ숕õa²à3Ä'–˜uì«pˆ—:gÜßÿSkvˆãiù8ˆ=;é÷±@‚¬Óž-g‚þ8HÎæ†Häv¯i,óÛ¤àðœÏo%C·)Œ‚TÞV{ž1[cËuÆÐvʪœM!)ÿ¥VAç@²8w!…ÛšóƒÃOVCªð9Aês&°ü ©'7'Ö˜2Òǘv¼ÞiRwIÞø0K¬ƒd‘/AŠà$Ëú?•• Î[’¼—)œ|€­¨òwúÍ!SêØ¢ÏC¸P\î €LýØê]ï'AÆûÜ`鲎I‘ú’á`ÇzHûªZÈWóRÀ¶„z ý/ ]øcÀöÖ×ÿV)->OW¾ÛéµÌª-dÊú<3¦ó“%Ú@ª–Ív‰Òê\^n| ™“+ÁÊÌÖ`ýaí}"X‡ÌÝ×wØÙ1pz-¸ØÕot¬]íº™ÂöµÃßÕ­Ø˹}ð»2Ù2€U)|V`gånÇ5`?’3íØÛë`cå\€¬üL‘;oç~ vÙÿ°1y»æ|p•ÞÇ:Šà~+p0whéY`¿ŸÄ”È\Ò蛎v'8…ÿ œÂN.? s Fp²O&ŒØ.©Ç9’+ܾõspØþn¼ïÑ ð°ö}cW·E쎃«+´Ë‚KôÁûÎK;Kê±nfûêèç ^`€ƒ·Ýꀋ57‚û¯YÌqÎ땧ÖYØÜZ1d9Òþ?×!úËàŠä#pî˜[HÇÀ+üŸà^Þ·AÞ•åàýfðÈõçf¶®¯~7/#K‚KrË5éðOaÖ'àÔNKËüÞàçx¯°7º îD&X¸ÃR™BžQ,X^žÀ#ü³à­Ì6ÎprÃ%øرúÑ}ð Ÿ-¸¹L£?OwÇ€aqõ"ðÊÏï—ûÏÍå†ÿéÿÿéq˙ҼýÛYt'€›c!â?}ÝóäÇI×ú^ϧc™¢^y²Yzp)¾ø¥®šýœÀÍí#WÔàùøú4ÛyñÜ;àgí…ƒÀ'×›OêþûF­âh~᳟ԹÜÌ:µ.xð‚|Ÿ^á³Ow)–X»?d粆Ðûàý}u)xúžüÖ÷ø¥Nãgéò¼Sà?Ëçà™XÓ»´Lx¥Þä‘Üòóôi&dsùúGðÁFð3F«ñ-õv_w¶òÁ/ü)àå²F}ð·™ywUóFàåí9+ä°U×·!øGqãøt'ûg¿y~©3úã¹ñ|L&;|¼ÂG>áK—äpvUžˆþ·³ÿ½ÿöäºó3ùê]ø7²ÆñCð3Y|Î~ð¾—z}NðË~‰_ê…^–v˜ƒGöý¢]¾Çß,‰z§|¼]Û|ÂǾm¹G½uŒÝøÖ©_Þ×ç«<ßÙµxá>ŽçÚ:f·yøyþ%xžñ|"ïŸ<}F^x‚Ÿµ›j?ðwƒŸ·é>ßH~¡À+êIð _¸n°ÄVãÒZô3)¼?p! ¼Â'žïîÑ=û\L.8ò7xx[¾"¸eØ#ûênáKëj–¨áÿ/Çvdðÿ´8kßÆÿ\<šê}|n¸dÆ#|ÜàéËa%û&nVæ.?®&\@·ô¸…?|͘á¾*¸kÅ|žô$ÜjïZÁ¿ß;¶³£Á-ûšÎò¬C›¤O|á€SÔs%÷S®3'—G¿§› „àfÕ]§Òàà6÷R%|Åàâmÿ6à,|À&À)¹à>Hpðq à:8M¬¡ŸYòû;3ƒMupËþCø|Á.û~ümƒÁ)|·àdêFë«`gòJƒTpH_F—¡ÇB&;V-ƒLÖ^/L»ð¡‚ƒÛÏêAfö¥ÜMy£À!ó7ûœîGŒ;ƒCø)ÁÁÚ¢«Ÿƒ-—BtZ¾Áì uÁ&Ïw{û6 íñ$ÉÔÀÏò!“ÙuAúKÖ@j£º±Ê2+pcØ…n6~ÛHó#y™ßp° +Ø»°B82…ñÒåz³ =Ò¤Ni~HÈäédoHmΞ(H“¾Œtá[…´ŽÜ` é-x ©’÷iL¥¡‡4nC®©\3Cb¹Þ[žÞ(„tÙM“ý®¤è_ëá¿@:;RBŠ\ILÝÑ„TÖ®Ù~2¸­®•}¡šðìô¯Ÿùü9¤”bƒ;A:›&ÑÒßgF¾ùÄl1íîBJ[nƒÔsm?_˜û¤ó²¤)$s[HHbß"x#ıl®K$1»B^,ijéšþí %hÀÇ+f_…xéãHámîŸá™ôW$ìaƒ!™ÿ¸¢ùrœɬË3>âëG²Lb…♽7þ!<•z|ÒÌð`€X.w ‚hù9“åúNdåÇQ²ûoÃ9ˆ{t¯Lé_Þ@dÊó»-zvƒH&W×[™ÝèÕ-ô=ˆœWºbåüW+û½QÌ•úHÿRÜbv ÷ƒG븀¤O"Rø×àž¬?Î ¿%ÜŸ2|b{xôËêœÿœ€¨Ü¶åÖ/x ×d>{¹#8ïÂ=Q§Â]™·E‚ívõ÷3àºð±Ám¹oUჅGòzÜõÖNw¾:ôÈ ×Ä< \å¶ÖFpFæ×·e鞘¯Ó?MÛ¿±íf¸rš%Èá·ö;B‡€ã”è¡>ÑâI?ÌU1ï7™ »_*œš°ÐûË ¸xªê‘]€ý[Y½;þkIòV½N™Ú»‡lº ²å— «xÙ\öG^êÒú.(½¹ó‰·SñЂMU»Uîó=žC×v–ÀãŸ>³­.“:î? î´˜Q¾3®¨ØíÈ÷Óßâvvµ§%Á|Óˆ©=’õx<³S7&íj’°<ùÓ<ܺ)cÏ ['\]cƆQÛqI®cü­¢³¨æ°–‚WV2aiž¬kly蟩xFúá÷¾Êíøu1¸ Ÿý6˲÷þÞ!aªm î]øÛ†AåòP­t;¢Ú§xU;„ú½·Õ±¼@ÄE_ýz²ªà-=?ÀPç.îZ:¬k@/¼9™2xYê©{ó aö‡x›Û£ÄË¢ÿgº—y¼$~$ÞéÁ'x—ɘá[ð®è ⎧x——W#ñæ¡÷Yåw¤îy—Û«`ÔÙ»CÛ\:…ë0£ŠÑ¢nÀHQà•Û‡Ùd FVd‰ñ|*}¼ïsáˆ|Šz £¤þXêè°ÄÔ‰Qrþã±È§1u‘›ïÅ'R§Žbîw}ÆídÞøXœƒø˜ËUïã-«ùÞY ٸϜhŒf.ºÉŸ`¬V¼³²+>fÝåð$Œ“sbñïn`)Œæc7ƒð©80^ú¹g0îb}­2¯)¢¾Á˜Fó,¿¿ƒ‰ÜÎbÄxÞ~lƒ‰Gxƒ˜ ½y7Læ2H'Lf] 1¹7n`üž1·GçŸÃ8é»Mú >“zyŠð‘aŠôѦqÛÆxŒå2N>û‡eŠçÕ^>›úíÔÓ˜.çDRÚ¥`—·{a·ùMǦBv¯Ži]XC`¦Ký9G[6`F}.@ ½L^•Û‡Ó0U®‡´QÇ0¡#¦Éõ™ÆíwSÑ&tL•óxiÜÖÜ 3E_Ó.ÌdfÊù¿T9wbc*rS¦‹sÓx;f:f^â $L_\yg…Üþh‹]öò31]Î9ÚøñÔNû9᯵ñ´ ?f,:rkQkÚ:ó‚mrfhYr÷#&̾9Ìäem{̾PÌäáJ¢]ûÔ%fC»Ô-íƒ s.-¾Œ6©K;¥Û>õ½ëÎoÀ,©ß:E>ŒYr>ÐÉd—Òk1«z&kÿ^"kéýãèÐ6]3]ÐÙ,];¹Þ g¼D'O¿‡¡Éak‹0‹Ïd¿E—ÔݳD~‹®ïyƒÚiþ:j<º¥¾én=Š9 ÐÉí-Ññ|I7Ó™tÊ>”‹ÛV[¢Kø¨0‹·uF`ÃèÃuè>t ß ºüÎ*ôp9³ºÄ¹‹Y‡Ž[¹¢‹.á×@7K¿FD7¯ëŒîfSï}ûr :™MÜåEï*~  Kö?Üryùêt‰zÝükŒC·èï¢[~¿ÈOÐÉìzñè‘ó•>9?âu::™=åë•ÿÍÿ¹E_ƒgžÚ0oú¥_ßǪÛ÷"ÑÏËé‰èåÇS5ísŠy&k´|Š>^–ÖCÏfœJCß 6`„y¼ýk¶ºùÓ!ôð±È©èËOgz˜ `)^Ù—‘~%ô48;òòLô²¶ü“%èþ=ôÊy°lÞö˜†~^Æõ@ï].iŸ×…èç6¢ºèû®ÄÇMGðù WømÐ+çB³YÕÞã úD}¼üéŽÒŸˆÞ^.ÿ‚Þ§¼ C//¦£oÏú(ë—ÿæõü"ÄlÙדu8Jú˜8³ùe™Œ>ÉÏç¸þý~Þ†ª†~ÙGv úd?Ç/çþ×ʺ°¶¸]e½¾Ÿ˜a§Z¢¯gÕÑ?V#„n¡F]M(äu°!t5âÛÈÔásU#xÛ»ž!üfjDG>¨Fˆ|ZXøëÙ%¿5P#Ä:R#ÄçW#ðýr¡‚Ôˆó¼WÃ7p_ˆ!t75BÇ„íÊjøÖ௦†Ÿæõ§!t íç’þ|ÑøW5Bèõj¸ÐGÕðÕò÷yy¯¨5¹¡B z‡žÂ}vjøs®ë©ábÞ\¾X5\èæ /5LøÕðº¼VƳBív‘¢v…'j˜ðÑ©ab>Q ¾5ÔÁ÷ÆÝj˜˜³TÃ:íg‘ÕP1W«†MarúojØ f¯¯ª†ŠsJ :‰šÃç*ÕÐeÚ»kUã¿qBE>üßÏ…Ua™ÓÕСòïÙéÝõ¥*êb5Dìk5T^§Pá'TCDݪ†šÄú •÷;Tè jè׺;=÷ü¡†°±”O©!̶õŽ*úxj(»=¾PCÅÜŠ*æþÕþ65ÕPqN¨¡¢nWCY{£ÊV5dï¨!K¹/B :‘,ò5ä[^À©!"¯SCÊñ>·ÌdøÛóÕáwUCJ/b“jˆà¨ÌÇßÚ«ÁíµU0Î¥†¼ßJÛ‰jðDfÿk«GsC§Ìþ:!O îÏ j°ð]©ÁÅç 'ÖCp*[øûÔà?x?D :ÌÓÎÕà£ÜÏ¡ ΩA§Dü å|nS .¸©Á#¸ÏL jÌûý®$¯WSéfhi´\/AÝØ‚UÕ ù¹ƒD¿¥ø÷÷rÝN Êå†i50/D5h÷y«©<1Tƒø˜@I5Hô«ÕÀ ¼ï«rÜkïSyƒöß_jÜ·A¡â¾Š¹ 5P«Æ´ŠY <ÁD5p1×[ÕÀl¼c´8Ÿç¨7ؘÛÏjàÎ=5P̨AMׄŸ[ s¯j ðÕªL&Ù2O HáF5°++¨Â'£~/8ð–0j·3”Sdü¶M¿Ý¥ˆºU ý5@ÌÕ¨|Ÿ`‘ÐMÔ~™Û`‘˜SZhÛÏv‹Ä¼— tX|+yÐVþ9ûHÔÖ%,£ôùtÃ÷Кmpíÿ|Æ"9ÇY$ÏÍ·BoÅ7²ÿ_$ý6Eb½ã9ß]$ç ß }‹ø×Â7ò¹oäœî¡oãF±þÀ7B·ÄB™¿¾‘ói¯Ùm¼ … åœ|¡œ~#çßHŸL!·Ó¶ÅBÑŸÃBÑwÀBÁ ,äXjˆ…‡yƒ ëpa Eÿ xúú Šû‹…òº¼–yÃë†|câkÁ|­*¿Øc°P~97] üøøZœ?X ¿Oô]¼æÛ ¾–óï/…þŠ¯äóE^ 8¿:ðq¤ÑXÀlSíÀí·ûÄ9ùJœ XÀ.oô[|ÉÑ:øêž ã+éKyÅíbð•ð×ã+±ð…|žÂË |?áKá§Æ—‚CXp†_ÎÙ·ºdÊ3Ì—þ–W[¹q _ÊëÿR£›óøPÌ—¿ŸÏ¨f>‰/żæÏä¾’s”/Øñ×ì¾:æË|$_pó™M§º_ˆF¾Ì˜/æ]ð…Ìwþý3OúûòاíìÆ<±¾1OÞß\é;{!ö+æÉç—ä‰ó_Èõž'óÍ—×÷9+wÖßÂ\áOÇçlw~à×>¯xþDî~^`®ô¡åÈçwäŠ}Ž9òûæˆóŸ³¯±q7>—ï—-ó»n# Äf{½» ŸË}÷\ø^1[Ö£ÙÒ'÷œ—·e0û5¿±˜-ïkö%Ál¡·cŽÌWŸËçbäˆs³Ùc%²÷c6O‹>ÁléËýAÌýÌæcÆ?£OÎ{Ÿ§˜K%þ²/~¡£_>Ÿ"[®¿ÐSѧUU @Ïsž£_a@½^ásFé‹G_r¡Vi¡ì+¡[Ì} WÎ-{„¾^.·¡sâøGƒ sÑ+ü«è‘y¼/ŸÈ ×ö¼ô̼yèÝÍpè‰z‡9Ð)tB´3›az!zfSúÍíëè•Ï‹q_2úD¿N«WD=•É ñÐÅä¾»Ðññ´…ïöé‹RÇ,Ñ_G¯ÔEœâÜÀ,Ž£ÆèfíÌñÐÅÛ<3ÐÁÜM– ‡¹ÅÞ‚©?dqùïKÌü–ÆÐÁºQÜè>¾,©8y¾!fñ1Ý è”ëÀö 2Ñ–óMc&ÿ8hgn‹’KÑ!ý¢6©ÏØN±Ö´Ã éÌ>£L1oŠ™Lv3Ù¸°{6Ú„éÒ¯˜¡oÓØq¡"Ú¤Ÿ3C>O(íOþ0Cr"ËPSµ÷¯:dËý˜.Ÿÿ“šÍ<1UøÀ1]œ¯˜ÊÇ“yÓ›³ ÓgTœóxÞ$L•sÇ)ì]¦\ÀTù¬déÏM‘çRj+>pißžÝ61é ¦ˆ<“­l`y ¦ÓäsH’XÙ‰Ïnp#¦Êçr$ñÇŒDé¿ÇD¹Oãå|t¼ðg`¼|þV¢è‡`JõGÛkìèÉrÎ>Qê˜1q¬Yãåó¨¢ä9;” žó’çÎb¬<—b¹ŒÔ£EŽQ˜0³ŸÈ9þhÑÇÃhù<„Gl¹ï#|*uÒ¨÷xƒ£¥îU2ï邲Ç0’© _yñªÈãñ—=»ãSùÜ°{ìéRðç…;FŠy<|$ŸkpCø0R>¿æ–0LáU1‡Š7¥ÏôÖ¡®ƒÓza¤˜?Åbo°t MÞaŸ®Ì6<݇ ¿xWú ¯Šþ+^>¼ÈÓ/¯æL<= ²ÎèZ𜘃ÅÓýŒƒ?ÿë<ÞsAx•Ù]w®Ás|,d .üvë¤àá›ðøR?f<ÚÇÖ“8÷#tžbÝsyêB<>ÙÛ`Hýžx¦vÏ‹r þ¹'`Ñ°©^øëÅ}ÀÄøÀ¡'ïÁ–G85úØx\ä\WrÔVX[µÔ]µ_W˜²êÖõ€NGaß²3kÃmÇ­ß¼-[nâkÜÜh×õüá8—Éjâ±¹U¢Ê÷ûN×½ºL¹Íprl¯RM—»àTѣħa¥àOgÚÁÑUÍn´ß×N–»c ïfOl‚Ë«ÙôBì³öà´'þšê0ûëzp©Û‰¥Í3çÃ%‘×Áu9Oq ( .l(}IÿαŸ¹úY×ÕuK 9ßyQÎy œKF9¿z¶ã\e¥ú\ûð›Ð•=–`f*<àãÉ­à›–î;nIܽŒ2§šî«W–2!õ¯O¥ìá…Ækâ”)ð@úÈbvò|Š¼žlήø©ÿ<óp_ÎgÅ>àºÏÚ8†ƒð˜Ï›!FèQÅܸ£wB´Ð—!F<÷žVbΛ ²;b!^<bŒ¿ )ýqyH4VXd¡÷‚ô'Ã3^®@ÚÎnÉ¿íz ©Ü&<Òe?fpCšœÍs™iÒ?”!ò,È”ó¨ÜÞßÒäÜ}{àË)ÏGt9'”ÁQa+Øä\l†ôßeqyy؆ʹ!SøhÁÀO`¯ÀkùxÞÄœ];Á.ïc¦ôAföÊV.8¹Ýa2Øy¸Z`—}ÊLÙ¶ËyÛL¶¬‚çƒSø¯Á!ž‡Y‚¯àþ<§œÿrþÌ©ç‚í»~ïíZ>ìB—‡,é7sŠº\Ò·æ•ýH÷ß|À¼òó»k”a“à’>9×&và†ƒGÔ—àaÓÁcnƒ[èrà–>·,ñ\p˹}/,Jx½¼qûŸŸÅ/êðìç |ÿ·*|P¼|yLúo¾ØÇTû‘½ÁÏvkÙÍ Ÿë^ñ¼ðkIrØÒ³à—ß#[¾v9.ü@¶x~ø¶r}üLNm´üZõª¥ˆà—s 9׸ž~¡£A¶8—![οf‹~d‹¹È–ónþ¶µŠ¨dó²­*äH?OŽôãæÈÏ‘ó!kŒŒ©ïBŽÐàyÿ€#ê2ÈÏ·„\Ñw„çr^9ç)/\à¹ü>9¢î„\þ5 ð\ÎÉåJl.“Ooì…\¹žrÄ 3<ú%äJnçŠùÈ•×å9/;Â!OÎGæŽçú äJÿW®ô×äJ?]®|^F^0/¨!?nékÈ}È“sÕy¼ j/qáž‹üòäçÉõ4¼óL/tÈõ1¼àã·ý!Oú_H_j¾ôS½zäK_T¾ÈÏ _öã_ðr¥:¼àå[?x!ôkx!êÈ“þÝ|éw|)ç}óåúx!ç1_ŠçµÁK¡@¾x¼”þ–Ò§—/êHx%tVxÕŽïàK Oᥘû—¼¼ÿ ò….ÒõJú¯ ä\ø+ù|ŠWr®û•ô ?ˆçiÂ+1W¯DŸ^Êç¼ú ¼–~ª×rN±@è¤P æ PÎuˆçíA¡ô Ê÷+Ï9‚¹o^kYïØ#µàõO|@ ¾â åùùZ®÷BQ?@!_¶#á5s+U~oäû¿æËz¼•óó…r¾‘>¿7BŸ·lÛ•L…·¢_¯E^Ëõ\˜Í'(”ëäô¯ Þò4ú;x#ÏÏB‘WC¡\ÿoKs]Þ†q}Þ6çƒaP$9ôFÌWÁy.¾‘×÷­xî” ¹]$×E‘¨Çá­ô５ùWÿúZ<éã+’×»Hî§"¾}g@‘ø¼J€Â0‰óB ×­HúÛ‹D¿Š¸œ3S ó}J€à‚Ї¯O%àg¾n”€íü”€¼ÎWÅÜ› æz”q]•€‡\¯S˜ü“üX srJ@>Ts«(ø¦–æ %àµXÏE‚J@$¯ãµ?¸—)ÚîÓJ{% ¯k%𾑕À‘ÜW§Šy)%@ô” ÑçS×ò9<%ðΤ(ôi%PèMJàu®¿(B_V¤þ¬ ÝP þ %PpJ ý?%¨-׋•@±¯• ¡£)A—¸¾¡ }K zÌõ>%¨Ï” Íü P…Ž¤åç†$ê>%踞ý§‰çÑ)A—£ñ¯?] þšƒM þN\—`ÞNQ” Ë|nI žÇç°•à?ùúS‚W”à™|žD }%%8œsK º“”Ë,%xŸ{S‚ÿá!%¸!×c•àg¼¿¡‰sO ©"?ÏT¾¾•nìXh¦„Tc_¬¶²‡ï%XèYJ°xî„"úÛJˆðí*!òþ…lãƒvJ0o³uWBÊr]T -Íõ/%ä-¼"öŸ"×eÈÎI%ä"”"ž_§„~(Á6îgÕ~Ÿ-,í}ZóX YØJè®7*¡¢ï§„ØDƒ/Tô •P17­„ÄòçÖ)!…`J¨Ð¿•Pq¾)¡¢ß „•d‰E]%TôE”ÐR\ðVBÉ÷óñºç¿ýÎ.Û°åý¿Ô’±’ìYžîúG¿í†«‡ýEÇ7}oíÌk’_2Þ÷ô(—Ük÷¼99à«|õCƒ=+3Ëš¿üwÙ3{·—ýýç¾ûµöå~×”÷6O~½—ÿ ½ëX“ùÁ/Öíô†w ÞÙüÜq&9»Ï}ן™ôÙ¹ ÿÓuïçv_yÓ—ýñÃ7¾ã·/»ÐZ¿ô’7go¡ßÉ}œÍý¨­¹qøª‹¿âËÞÿÖ§··Òº“_ùZ‡ýhãÜó÷m}úµoù·ÿæñ îøçùÚŸN?ïæ®ßÛCߨ¿êèëÚ¼ýß?üåÃ7^lË/ù§WÜyáÇý·á¾[]rÚË–ú¬ê÷üÏSo{™¯O?~»­øÏ?è¯+>mϬÿÙU?÷½.¹óþÄ%'Šþø•÷ÄçuÏO_úO;ßÿ6{ò‘“ïýV‹?zׇ:~|Ã[þó=ÿ¼|ß8»/ü®ýéɘïó†/^ûõÅ7{Û'~\ý§þ‹þ߯½å+g–v Äç1=ëá-–÷þÌ׆õi+ï8ùÜ?¼ð¯ñøÛúªëÿ˼ô“_øÙ¢þÖ ¾öÒ“~l¦?Ñ~Ðçömmúò=¿öY¬Ûî+®{Í W7YM¸nÛ–ìÂÿ^kå/ÿÐmÏ´½Ü&ÉþOýþ ¾â‘JV¼ÏONsËc^ðó×ÝúȽŸ·•ÿ9nòÞÖé]éÛW[SX¾ðÈøÖÙ“÷Zç[Òo™p|ëa}®[¾oìȬû¼è¯(8pèŒu§Ûòs¦õÿ½Âmó^ûD·ÏJo¤­ûÇ:~3ñÓe †¼™ý·6]v·÷Õç¿6=a/èúÝÒïüí&]“lÓ«OÏøø\ás³3¯ÏÖeÍ;o~¹¯ä>ìÏß³ï]sª®ÃŠÚnsÖdض¬Ê.YŸ›]E#—[qÁ[“aõS?úÍËßþ#{,k<¦7^lϼ950¶õGé²?a‡~ö©ÍO·Éëì±OÊ*>¯?°~›ìÊ—Ï÷EoüÕ{/øÛOÙMw÷||ùEß×s²5ÉíþÜn›Å:½ú‡éBöUá~ØÚ‡RÃc¿ûÁVsÅ—þÁæ…uzŽ=cÿÚ½ér‹yòØöœ²uɶxÿÏmÛ±ä…ñó•ÙÝš5!~÷‡ó´™<‡Vîÿ’äö¯~ØþŒ}½‹ûÌÀ7möø‡‡¿¸ÍŠ²«ø™½öÝä²?¹ÙÏ:mM–Ù£ö‹°^ã~›µ÷ëWýfâglñÿ\fYj7¿ïßž:Xà¿á÷×ûn¿ëÞbÿ¿Ëú©~õ»³þ¿|^ÖJ]r¢ØõyÙ£Ç~|C²õ|ÑcÉ?Ûª¤Ú„ýà?áy”,,¼6»eü_°“+ŸL7œÝ¾è_>ÝÖ”üÜØÓ^ú÷‰~‡Mb­L¬Çw¾›³Ù›ö¯«JìAöeuø¾?ì®ÍÃ>mc?‡Çü*[]õ_ ~Yônùó`Mç=wÅç^ù‹«Ï·UÁú†ìÝyÇѽ¶¤Úï>òÐOíÉ=åÁŸ³yÁþ[S¸/öpòô§1°À<ù™/µÚ†°µÞmvɸìÿÊëصÉjù×Õ¶rÃOÆßRÚjs“Ç諬²ü®ä¶üm©£2ù©•¯OoŒìµcßmñ³?ºï…GþIÏÅW횺ìÃÅUÖ=6q¨Ï[&» ìµ™¬Ó¹ì«šp<ÿÇéH¬ÐGž·ÕצÝgýþêçKo±½¥/ÉžÁT¿w_²Òž¶jöS!ñÉî7¬»}ÿš›mvr˜žUÞ•ÝGž\íE7¿¬ù¡7¿7îûeø¿_|ìöŸ¶u°•ç¥ÿxé’§³óo´}mâŦ.³Š~ß›ð³àÇæ„øÂWã—îụ̀Ɋ·2ü|&Ø!ÿÉ ‰áþ£•à—VgÒ dmÿ0æ{cÿXj…écû{+HÍŵv_Öˆ½ôÞM6?øGŸ›}8ÙÌ qŠ-~Oú€l6÷yûyAv5¾ë'OÚïB|as²7ï+g–ÙïC\æó“}£Þu^“6üÄ×ñüe'a­Ê^TÖ3ºöË’›ÒÊ7§Ëè"×ï$aâP»ßü“ýÿÿ-ö㬰Ï|Éñ_ýeÓ?>¿·5Äqÿ-Âÿ.MË öd°«¶ffê`}Uº=/õ<ׇêÓ€ ®›g2©c÷‡ˆs “§}룦õÃýóöÙ/°“îó“ìkíƒù& d‡?”ºÕ/Øo³Þ²bíq_kòo´7u°¶8 ýû‡¯·ƒðk°C³²7õC·åâ =Ç_`æ¥îúlë÷ÿöçWµ‰t¼ùìÿGBœm‹·ù²—½ßò‘ôÊ-„™_ŒÇy‚û•«ì)ìë[y>‹~Ÿ.Pßt¸úÙ7¬›i SŸËÜä±ÍyÂîH¶×7gñ€ö­¸2]v_° .¢uôp6xÊ.q›ü”¯'., þÉ*N3[÷Ú42²gç·ÃºS< ø×þ3œ§¯b_oÇnÿ2\Ÿ×§Ëì ÿyXßöxêžoóZò–®ŠO5›J˜âèÒW¤ŽÁ—¦G»ÈÖâ÷ðÿÎïšüé†Ô@}Û’(4»d¼,‰òÆt{ùÈ’ø/Yb÷•ÈÞh}Í öØ*³‹ãò#—ÙÁŸÛ–÷¤Hôë"NÞìœ+ûëæâ¥íiØù¯ŠƒìnÖãÓØÃg^ѽ«ñ×u¾ˆã±¾âý›ƒßM\¼=uwy1Ïÿb¿$ß[òda^cëÈW¿çWÉòÂקØçñ¼ óñÖY®ý?' ÷®ô5ÄÃÊó¶g­Ñ¸ªë|}øòò2â…¸ÿ%î_™˜õkF¢íHÝÑt_žX•ë^cºOaÏ}MrcûÂ/ûfìÊýY+õŠîÝŠß¼9õ¾ûäß}ýÉäÄÛí©ØÉ—üÿWkìAüÞ,üŒìó}Øù2ò òOûbØ_ž¦ãõK|vºŒ^)»èÕÄË’trýr[–æ&×úx»™Â %Oø—Âó‹÷çñl–³&óC[†=.×»]ßû*~€xÇ‹‰›*V¦5Æ÷qõ!N°eÁžÇõ½1ä!ö0~ùQüŸÖx†éùÎN+_FF¦ã»Öùæ÷Ø ü^CsbøÇÚÃ?ñ%ä)˳Þí÷ý/ñ9|oc’ÍüøFßÎó_pƒègÖŽKx¼OºúOŠßlUbæ¿pÐæs=åÁ¾y ~þ©‡E;µéeßÞ”5õö}ò€Š4Ìúª—¥·õ¾1yÜÛá™GƸ£"Ù…·]i÷s˜&¾qÁÛ“'㵉¹uÚKðk›‚?ó€­MÂó‘Ù¾*økÿÞu‰#m–¿0å!SsxU¼žºp?my‚ð˜~ççÄï+C~ì›Èç–rK«ûÎëHÂìç~ç?ÃŽ¶†¸Ê7I?ò‚$Jå/¿(¿ëÿï÷²îô¼—NH>øi'NóÕ‰y.ú£k½pݶ(Ľ.ük5÷±ä“›³Ö•ÇõPøÁtzIðƒ®üqmš6~Õ'mþ»ôFÇ8}hßö0Ú7å¿öÄW§nð«ž™—&èÂi\ñÊ÷‰/¼)y`ïó‡ÉÁEì#ÜÙqây[›DcïóÆÔ¾Ý篫9¬£ï•…xEvÃ7†|•þ‚<ÄWñ<N\;çÒ'^-ûá›8§aƒ=’>æ7Ù†å%ߌ8ÊêÔœ½Õ‡|×Ã6¼ÎW&g=q]<¿âïê¯(8íàÖ_).ñ’ךâÓ¹)<òF[’º§÷5évx™ìbü]ð)Wº€øàfp‹Ò`WíIâMâ2÷€‹ÉFQx yª&0ÉÝ汇|=Ÿ¼W¸…oM~ý{+Ž÷Õ›.D»'Ýïv]vÆ BÜä³ÁOWcd_ñcÖ“]ä_¹¿Á„{­NžÊ›þÛÊC¾êÂ1¶§é×Glë™Ôñùæ€#ñZÎ?_—6&†û¯Ã|Â? 'U¾Ú• 6²Ãwßš:tÅkÞÇý"Žô?a:wõƒ{|g}ꄯMÍæWmqrÚw÷x÷ç1ìÜ—‰kÀ=í§øÁÖ˶V~»Íw¨| Ù0¿±Š«’Dt¡ßpÿ]ÀÃ|3Ï|ʵß+oõ‘‡}Q#6uáï<¯¸ïºÓÛÿY߆ß$Þ‹vu¸~ÉïK è¸ü xÏ"Þ® 8|Ö!Þ“¿úsêÖ_ãó“Óß¼ø¼/Wf37ºüߟñ›Äå¾,à_v'ëhq¸Ž˜Ï“gåpët{Üi›ŸË~sŽÿ}³"1cÿùg¯Î`?VÜùådA=j‹ƒÝ´p€‡Y?ì[ÿ}ð#þõ4Ÿ>5úáOà¯àñÞ›nÿiþãsÁ}ï溵n¾ÉýÛ’:Ô9?ž t_ÿ°oOÜI埕¦é“…{ùìÊßØúàßãýþqͱ‹k*ÓÄ/ú[ì»/g_/Å`·ü7¡nb§Ó€Â–‡€Õ—€¿(¾œŸ.Ÿâq!~>Xp*_“ýôñ_=çkÈß±[¾…õBœï ÒÛý[þ•ÔàúŠ¬ó¸¨o›ýœ~Ñh<×?}) x±W¤ýC¾"ÙηÿÌWœÈî!¯;”&pZGîi!Äç;fËk1óZŸö‘Íy]zÃl÷Gù—âšôn}{s<êHV”À»o~À7&Ëá’“þ öŽ}×øˆ7ÿc+sã¯õ¾êVõD²WÙƒàâÕéǾa›S÷ÿ!+âýòl•]‰Žëå»ÓÄ=‡'¤…'ùO«<š^`<òzWœŽ_sáÛ¡.bª—)Ÿ%/RýºÒåu¥).Çÿ+OyÉCà!åØ•ˆ#X?þ#pY'_“–/>êþë+ópå_"þYE\ô§ðÚ„û€‡ÄúûFq¤ÏqQ뻩¸ìðÂ_ƒƒkß(¯^²ˆýõË’õK[,\zå}ä=O…ýoøjão ëÚÖâÇ•W­Æ®6±^šÀM¿Ãóx‚x˜ú¦ÝüF\_Ã.>B\¾/äŶ$uŸ3}u°'ñü×QßÀÎvÛ…·ÕÏJ GŒsÖÀ΋B¼«ëº{tœd…©»>ÏÀY|Ãd÷çß¿! Lí‰?Øõà[ˆ[— ÌW‘O—€®Hàù=ôÅøÙ œ/ö[ëÓçà_…sV†xÔ—¿&Yñ¾ Ø;ßü˜«žI}È7§áû›my­pËGƒ2Ù¡©ýlÄ–&îï¥îî¥)œñïÑ®Éë÷©‹ø6>·‘x’ú™oʳ›Ý™Ô±Ë_ gö}!N‹ŸSnCºÝ>é Be›©+=M„¬Œý ^çÅÁ¯y!ñ…êeµØå]ªËmHÓõññøu Lóôv[’„ã{+Ï£"ø[/}[šàø6ê+ÅÄÝ›“pùŸW8uÕƒcœWòOQæ1¯²ÊÔ¼ÛŸÁÿ Øö$Zü¯…Öž¤í×GµÙØñ ´îOáÍwZ¸uŸ”Çü^€ê²µiØ÷&Õ%mKš_lß RŸz…Wo­Åo$^’Ý{:¬oßú¶ìdW¬­KÝà ¶ƒz²êQŠ›U¿|.jS~ܘ–ÇÿÁ¶Ÿ&¤¶8ØÜ~¥NUK½¿jÄýž¤[å}^Çï”$n¤s³55½A¶*äçŽìë¾i{È«áA4‡ë9Ì öÄŠ’]ôƒ¿³àÌÜGp6/ ‰£}Šõz¸3½k¹&’ Ü˨÷˜êö+¯O7P\g© Á+;ÁuTï…Ïâ«°_ !^·%ü>xK¼oaÇÚá/¨Þ(>Â2Öâüg°¿U@,î—MÊ+BÜæ¿ öÜB¾nlÛ³AÙÛ÷ýÙÄÓ©LÓ‡/Zú©Im˦ôû#ða~ >©xHõ‹ê¬³šôÙy¶†xdUˆâõ Ÿ¬N/ãUÊK »kEü<¯ûÌûžM½­™xbCK²@_oIxö£û½~uS\Ý®7Þ‡µàÚ™°oýÏ!°Æ4-¹ÛV¤pÒ-^åªûGÄü¶¹# ˜´¾¼Œø¯€ºøŠÀ+ŠyºòΞ4ì©Á3Ø^×þ"\­ƒ¼Pv¡‰û-¿³úê4ѵx+Bì—²~N¼¹õ¹>Yͯ}Ò;‰ËT?mHÃÄïxGX_ñù¨^´>¥|ßþD>ߥøÿ¿#Ü—˜7ÀÛð ®Ø ¼:Q8qWöÍìÒ¶§Søç VÀû•Ô¡fQSÜRü©mÅήM¾vÑ#_0ù;}¾05ßoŠë§…:[÷KŸ[Éq΄øÝ B\¯?ÃmÀ ï-"Ok`½’ÿjŸV¾…ðã}Ö}oj`­6àVð?“Ý-ø` DX_r÷>ø€5SGÌÀS³Ÿ¯¥§ùÛO<ÚOoý7Ò|yO.íŒqô¡´lñukùpzáF¾e!ܾ;òàoØö4½¿ùž—Ãßûü${þÄ ¯L–ÙµUÖ”l÷¹‹rü@ü ö—x ¡¾b C vÄJ‰›W“‡?Ÿó¼;ÀÈC#O+õ^7ï°ûá]´Âïôäé¾½?Æß{7¤‰“ýLû&ì?/Kîò;ŽYoà9|$á^“ÂnŸˆñ$8¯ü•x.>'…¾å«°Ç5ä¥ÁžYAÈ›½:à¶*à3ösÕãáËa­)àÂ~þEyCIvŒsáC]!Þ´öOg¨;¸¾=IÜ®7êÄÏZì©WQ·”ý^ðgßl!y¾pÈ äŸ[À[‡”„õnkÂúsòƒIPWݼ)äŶ5Ô„cØBÖi&]îW[òî²dÛ^°ËÿÈõ<W¼ð#/QñuSB›{h¿ö±³Þãy®IÝÛ Ëàÿ‘÷ó»¶6MO¾gÛƒãÔ÷¾ÎûŠWŽO’ˆn…ŸYâñ¦m¸b98XëdSØøªD>ô ââH_ þrÀË…¿k]ÚÆp?]õ‘¥Yã‘ÝR—*"ß,xerâïŒ8ÖšôöžoàÓ¶¾ u/÷ÓKØïà Ѿ—$Ëï3ÆûÏÜw%náø¯ÿÙÓÂ7Bü?/|K|öŸ²‹‰ÄÇÔ}( ~7_|uá>Ô‡mÛ%É‚Ý(~£ËŽ=œÇ—qSØ÷N¼e›Bâ%ÔÉcsõâTÕéG’—³ÿ]oŸ­›EœO:þÞÏ8ÿúwÅ8)€iS¦Šûû4¸iEàoY{P›†“TW¶âÀ‡6ê`NÂ6–±?çbG„©>·”õ³Þøøn‘çJ]Åž„¼™zK-qëÊé)ñÚ—²tÝâ娞IÞgmðgTçþ@¾ïkAº¬?ù>ÂÅæ³W]ž£¯ ühÏV÷P?+'ÏXöÃ׶'éÓ¨ÂN–rþZ߬ q†Ù­àXòËŸcgâ~>+¾ÛötùO0Åçâ©.ƯG.å¼V†üÝU¿«7jLàÊ]MFýÂu¿V§ðØ­±ÿœßꈷòñÕ&ì2ý%G¨£îZü¬ËïnëN ¿/I~æ+Ïzi¨ãƺè,ø­âÿµ…85W_ ^vðÿÕÈ°ûøÞº4í¸ÝzÈU‡ÌÌIO·ÕÛ)6“º­7Ç<‚ü6òÞTÿW_“plÅk¨+×{ ñúÂp_ŃSœ¯'ò¢Rsù9WýŒçî‹þêê'( õÄI<üå:xŒ-Ôã×7 ¤Ä™È;oH¶¶ò€w¸ø›­Ø·ò€ÃyERyO¹—…ü-Ú•ç£ú‚§Õ$gsg‹µ†óëvsàÊλð½|~ú7¼\Iýƒéí¼ÍJCžå 5òõÅ£÷Rü'üÞô»ÚßÔ#Ä_óBìf9u{ð‡/೉Wd·Þ“«ÿ€þ©>IÿÖ±ìZ†8r{G8øzð§o€‡Ê~Ãëõõ©ýbŒÿXGq]‰/O?‹¿7ÿRÈõ¯‚W2/ì‡È+Ú’Ò¶?ì[‚?ŽçOÿo&¿,"/R=R8_u[êp½~Œ«Ÿ {Zü¨o ø¥·‡:ƒí·DvÏÓåvwÄE2iºù:+¢žµŒúÕö”fs‘¯…çO}Ê3àý¡.þ^\‡ô˹î/vÖ oϨ‹xKà YS°ó¾0É\c-à-Äꇪ€‡ÙøíVJ} >¼‰¿¤zvÑ•TQW­GžÒ!/$<Ö~”ÇÝŠ½Ϩ<êqââ‡_åâ9Tw«ÿ ^«ÑOf%Ô…V„}ïð<‹ÓÛüN[ OöYu«àS÷E_yuâw›ê+cσ mû^º!­ü\8%¸”m ~ÃÚá“«®Qž»ídß6RßÇ~Xæ ‰¡Z8¯ þÑ·¯ ‡·óyð滓,ókmÝW§…*_OýLqSSàG¨/ÙÊÓåñ?ÖÂúR=¡,yÜSŠ×ÿ$ùéÜÑýqVš„Ÿ.·äÓê(Ia×8üK[CdSš~3Ö [áŸ?siúú\¼MÓ=iC›?Ãþ]>·uò«x>Êsï%ÚÖ‹«o¬>¦ì'üýØÎõöÅt_‹¨›lj z®ú\ÔYÁ:¨ .\@][y£x‹Õìõ#Ò7æ…ð«YoÊÃý;Ü7xêwtâ8[Î× ©*OW‹þÃI~3à°[Àùšéó7ˆý)ÚÔ›­œ~Í"âÏ&âØJü}5×UâYW}¶ð΄ ö±X[žCŒw÷7ïχW¾2Üo/ñV.ÎNh¯è±¨³ü¿¢xsyTM¨óǾ-úŸãúŽ"Þ¨òMÅs%ð^?¨¿§·çkN[®¯þ¦ðWñÄUoU_ŽìḊÀÊxžŠ´>À]|0îƒmƒG¯þ·¹á>;q•Q³"ø¾yu×È«Z(ž;vmaàÚRx|ð£c¾¨u/¤ò6ùuê•ÞH_¸ô ÄŸXø|QètÕ×Ó·G]6Þ_ü­oŠG¾Ž~WøXñó ðˆUßFýo.Q°Þ ý9ñ{sNj…ðÙá¥;ûÍ‹àA<@Ýö7ð?ÙñúÏ­ÈÃOÑ]0pžx\ðýIò¼Uç¨Áÿ(NÜú,ÖÙÂó4üÀXת'îÛú¥ÐxÚÔ§ŽÞ„êòôË8|_õuymàeÛ.ðÓ ý/ëéS/lK›KG ¼ýx½7²¯ï øj¼núËâçÁÕc?@-:'Ô âçàø3aG¾V9¸êððcÀbÿ’úé×>CìÛR=…ü8òÁþH=Eù¼ø» iºúOñzÖÂÇjæzéÿôáùzsàý¶)ä³¾-…Ƹú3éû‰ù¨ì"qAìGÁ¯Ûòká(â¥ÊÞ>xÒ¾ƒ7Sß Ÿ‰×GýjÒ)lýÖÏ«¬›¸¦<ðÉŒ~kÇ?·½3u¨`Š—ZùM'uÙåàm=§SÄÚˆàƒÚBðòä#Gè›ïÿÝG~ û­ûU†Ýžù¡­u+_ê"VIŸð1ê´Ãä¥[‰ÛÅÿŠýîØ©²$ úÙbKÏêŽS‘¢>8åïmØÁ-èt‚–œ)Þgñ"ék³nâTå‘â]ü„¼½ \–¾Ùgí_ú¥¬ ðÝ­–}ÛN}]}£ëR3û帴náÚø ¥§èsñêGŸ ^#¾ýDV A:=ðaè׉×]ÿ| üíäC…ô[ÔPo¯$Ž¥^jÛȧOJül›|/õøNÂÁ"o_ýåmÄ#íØ9úyÄ3õØWáÎEÜ—=Á^Æë¨ö8ׯêN¯®†w[òZïÄ®Kg>Ÿ5+Î'¾‘®¼S[O©ƒü@}( áÑâ_UQ`.nÅT_“n|+w® q®Í'N¨¥»2ukSrÏ+ôÇ{_à¡™êóØßêoT=»Ü¼½?‚Ϩ¯¨:؇È×Øœ>ÎOy9þSø6u^ŽŽ¾RÄ WçñX?±îF¿‡£³2é‚—…ûÖôɬ'Ô}­œ°œº|«ÿÐBž§zJ8'úOQ_ë`è#¶Jp«5¡ÏÍf±N…—Ãç½'ðãÔ‡û± Cÿ›!Ž:ÀýVüJ?†•œÇ†áï £¿"ÿ|yŽâ¦µäòÂÅß)ÀLhg*fý4‡~ aÒx#>1õŠ÷пW‰üÕ[zòøgŠï比„†~[.^Œîü5ÿ-<ºõ'ßÀÐ׈ü_ôXÄ|:ኊËÄ3Q}ˆ}ù¸½?ÎåIøñJt„«>"{›ÏgÙD>"˜‚—¥@‹K{^¶úÄÇ«J~þÃ%±?¦>yúe¥ôÑ—„õèÒË .ž¿ö_7Ÿ“¾”î-ý~káì'Þâ§Üú…¸9Äy¶”<‰~bë þÆÔ¸™üo+uxé?»pò­Wç+áQo€Ï›¡_¥|1o¥žº4ð òyv^½;ôYÇ|^ÔQSݸ;Ô­›>‘^ü®âÌè®Æ©Ÿ¾ à»®>ás›¹ßeôËÂÓÐ~sì Üœ(¼ÒæÃ?WžL]Uü¨I\êX‰\Ôï–ÚΠsa ÄôyÆþ´ÖcS|Ñýµàt­!ð®¯> p¶•äK¨Ï ÏìL5Ûx,ù¼z¹ ~ûaúÚ‚g ?½¯â> Ã7fÿn¥>ÏßÔ¯-û[FÛÈsï=‡x^Øs¯äxíè0Àó”þ›©e|Ÿ&xÆâAKçD8íFòˆUðË×P* ~VÌ÷èDZbøFè‚EÝ€Ía;õ©.~µ|¦[upñÒÀ÷c]G:côwGéa¬ èâ}*Î+û%^ôUà·G=Áðž*úEJàoOèÏ•¿T¿ŽôóÄß³^úÂÙ—^DBõŸMaý{õ õ£…ïIË»àkPséUñû»À/v£¥º¿úФ¢û¢¿ ØÆ}¦ÒJÉ¿¥VD]¾Vºw¡nb­—¥ú t:¢?hGe}ú6úó‰×#ÎÝ•>îŸ8¯£î•p;êÀÒ[´Íàžð"b<³µ&ù§Ötå×· ÅÅ­!ÏÂïxk¬7Ág™ôÒkC½¡ü®{IÜzm%¼ûê“¥äåô“µ€Ó ïWÓ ÿNþJñžêKÕðè{©3ö†üZ~ÉÙVJ¼|ºÝ°Ÿs}@äyy‚ ãÇdoÄo¿|™ê±ðé ÿÕÁW"²™:[ÿÞf-â\9êì}—|Bõ…nòõ#/R]ª™¾õ³im…o!> p3é\­Ih …Û£šøuªo®¤¯ƈÇWsòPé÷¨oȩ׊·-}½¨? :=}¼V Þ] /¯ ܨ†tDÔïYK†úˆjàã«MúvÒ+ê"O”_RŸ³øÂêó/ý+[GßQ1õk'~Q}¢ýHœhE§è‹©oÌ ºÁN‹«ß»<\öl 8LïË^ÒŸkÕàA•i:ññº¼–ûR >4öH¸ê/ÕŸM]³œ:Žô‹ôý*âXéh4*>d]ƒ§Dþ|k…×ØIü‚N²°^Å÷è ýqÖ†®vä~+uõG>¬ú ~É}o\5—Wûâ+À-ToÜžö¾G"ÿœÄѽ›ôWµ)b½Ä3½Ô'Ðs±RêÊ;Ñ‹k†w·!à‚ÖÍ}ÜAKñ`+y_ß/`Ý)^V½b¼¨7…mßb¡O*êÊ ƒ!žªÀDSúŠÂ½3ð†¨%þU-‚® ­JÊúluÀ‹r¼ÞgG½ròå³¾z$-$XSh¶ÒÀ{ñգ󰈇´ÂÇÜ:ê2gÈç¤Ç±‰úôýZ|0tb|&õ°þ}ËAˆ~tÌë„ÏÃ7vê,‘oŒNQ_‹qP]ЇŽû»ŒþÕ‹6é˜ÇjfЋ©$oG¿Ûf÷—à/J>ã™Ñýx¾=äí.ýtô¢ny3ýz½ø‰ è'‚WÆßë"½!¯ñÂÄyÊ*ð¯Q7‰þxë^ÁsÝD¼ÞÈ>àï^øeVŽ>‰ð³Jê[›é'(ÆN´ã7T·Dg0ê#„ëñê§õàÛÒËiÅo¢#y¤â㶱Þáá{5Ç­‚Ó¾ >ûµàS+п’þÃΠk½ôñ¨½X8wà?Æ>-õ%•ÇK'¹œ:cõté] wu¶w×V‡ø&×oôg•/Lz)û½z@7ø^ûTxªú~;¨O´‡8'þn7}²ÛKÁ9๱î­|Tϵ†}s˜¾ÀüÈ|¯…äݪ«K?ó@ÀÁì°tìÀóTÇj ¼~Â¥BüáGêo܈ýÕy6Ç´€_ãWáUÆý \§ú¨ò4ô±¬|_ë |iéY[3‡ï&ž¤x> ô‹—ÄÁC¥¯ìÔ½½xåŠ#/^qø‚mVÿ+û@8g=õÍ%x˜º¢êg™€Fý˜ àÍÒwÎ\8GoÐa× ËÊñ«ètX;º{{ñËÅI˜µú« ü''}…¡ÎŸ«xŒèäö:ïð~¼¼‡¾L—~s?ñÒ,æ%¨Ÿ·žy;ÑM|8ô7xy¨yÍhÝãx\ú7"¿˜¾)ø*V‡î€x˵èÞ¯ ~ÁTïØ_²ˆ>3ñ¥/'þÀ†”~ò!+"Ÿ]ð$/WÙ•nÏ ¬'”VßS}¸»‚>m¬«eBŸ³—‚wƒïZEð§‘GÚ‚Žq;}DšcÒÔ.ý¿Çû.j/Àã>BêõƒÃìÏNð¤ìpë|Wès´vòäè·w¼)ç(}ÉûÐïj&>F×J}™6Bœ»—úÎÑÔ üUÔŽ-¼ê8üªSÔÇTO†× ®9BÞ;Bž\]DwÖè'òaúvŠð;‚=Ž×Ñ~#ÿ³‹¸\,÷¹ ßc³6ôZkˆ“múÃVìýlñû¨ß`Eýñb<)—/®3úa¬]·µÔiêÁ4Ä©;5ckèË. ®­÷VÿRA°›®ë¾.}à ù«ú쎯¡ë"•Ä»ÍÄ›tù—ƒÓ©:XÔ1Ç.ùòÊRt¥±Þ²ðù‡ÉÃvÒ?¦tÅ냽qOºH®(´íüî)øŸµào ºÁé{åŸB_un]Sw¯£.pˆzË0ü€íÄÔ§m'öÞ´U†ý¬¼ZzÃ&¾õUk¢þØ„î©ú‘;YÕ¡ÎvN]C:ÏÃó?ê"¶Ÿ:Âö§ ûÀ¼ëÆ϶cßZàÂÎì&Ÿ©"<î×ÎÜZôvTï×\¡ôGR7;IýR:câ7‚§.sŠu[À¾‚§ÑƒŸ8E]h„õQI²gÈgÅ×VŸ^ ñ˜pº¦Ð÷óO¸Ž*öôr4Fzõú8\¼TéKïB¶ð‘›Éëè`Ýø2âÀB?›ÁŸROëÃ6«Öþë"ŽiÁ¯)Þ·bž˜ Ø1ê êçpÝ×…×ØjìÔpþ⇓ìg鲈¯"þàñ€cÛItëÄ»ÆÞôý´ø|FÀÏá Y)ñ8<ÈýN"~Z%ö=«¡®ÏºµÎðxrý:ô¯‰Õ¯ý]+Ãþ€ ¸ú…·a×á9IßË*ÈË/Ð'©úw½´VìMeàÿºù1ŸÎçOÓT€n¿ô ‹©wÐgû©K¯Þ¶³?;X›É«Úƒ‰õòºœ¤,¹·NaÔe.üîðܢߗÎS øùÚ<}Vê`uð"Ñ×´â%ú¾¥ƒæáú¢=‘NœøÄÌ»‰}VÔï¤ÿk«Ñë_¨}š6ú¾¥—¤y,[Òmp“æõ–„:D|½/à]9ž6vý¯á~÷¥´£÷Ú>êð{¨G*åw£.ÚêeõèyVRGTݾܾRø ×ßLœ"ü~'ý' ØñJÐëò*t:Ÿc½Á¾ bº¸è¨ÞF\ÓÂqZÉ¿÷sü]Ü_õ Å¿ˆ?ÆEéÅ‹?gÀç¸o'YÏ໑´øö8yö)tºá»Û0ëO:'óú¥«èO-#ÿGïÄr]»B=ßÚX/»¨#Pç·ìúêùâËÓ'j-ÔÇ·¦åÔVÿ‘mH ïnûá7€“ǹqõăà]škWF¼Çüº¨_X VsVzÐÛÒ<Ó2Ö5:ÜÞúy-ÎÙÏ™¸ß·S÷€Wå¥Aï˘[cÛ‰73œïfúfºáëyJ÷ùŽ5ƒ¿k]k>hy€úì—ÐÏ)½XÍ ßH}øÅè»*îéÃ? ?câãÕ©ŽÞò‚ø= Ô±ËB}6®âÌh¿úÐ3jF_b…pλ^ÅÈWv€—70Ç£þU'ù¹æÑ/íõð53àò“šo¹ »,]rÍKcž”tW}˳iâ›ãÁ¯“îÏÃëñÓðkí(ø1u¶ÕIøtK™õƒt¡/~ë¼ÎºÚËýÙö­ÜIx'zTÖî*=ú+½%äÒú®k—ëS£>N}ÈÚÁ¥¤ëC>ïÐ5‡_;éÂW~ûðôôtºY—M£û~rù4:¤ô÷ÁÓéà<œzѼôÔ#Á £žýGv\®ŸzÂPx®¼_8«æÈ Ño|"ðæb?Ý üÜàáGÁuGÈS—Ï ŸQ? ó#ÿL<ø:ÒK7Í&/Õ\.뇬$nÑ+õ5ªï­~?ñ.¨çDý‰è"/D_zl^ø/Æ\ͯózì¥ôÑ5ïp!|‹•àÒÔaï8ø}ÔÙ€ïáŸÛà‰+¯Yôt"ow€zˆú|«ŸÜ®:ÜoÍõ2Í_k /¡^=õ-é#yaðþ‰õßC?\/ó}4/Jýbê[P¿[7ùÐaÖOzðe\ú1e¡'ÚKærY'úHÒgûu^µLq(: MaŽ•À[T?p©ò>pÌê0Ψëúbüu+ÏCó³5/à ú6‹à{Õ‚Ÿ¡‡ê›áµµ2¿S󈱯g<‘§k$¿ÕŠ{7u(õ9£SiêÍÒ]ìß b÷4ïA:iíÄ7š£¶žˆòEé¯)OS¼ þ‹&òZéû6<ÅŠÙ‡è0ªŸßÑE˜tá×®{€úÝ û¬'øÅœnõôŽ¯x¾4@¼#ôiG}ø£¡žc­ä­åßM@NgÜäýćȟ‡Cþ¢y¾šë§.ÏëñÜpþäm;‰_±ŽFèû‚¨ú°k.f%øC†:ˆâ˜vòÞBê¿ÂÚ‰k4—¹‡û¬ùpè3Æy¤ì÷}ÔO°¿.=éq÷’¿©Az³å!ÉéÙÇ\þÀüÜbžŸ®SºáÂÿ5GIó~´Ÿ—Q—U¿ì6^S'sÍ“ŽÐbòOõ©IçI«&¿•½Ÿu6òü±Ã«OÂ)h{sÞªà£h>ˆæ Á£³Jp´¹ôQuÂ÷=ÎRŽÞ“ôˤŸ%´BtDéÏ°â 7lõÔo”_çõMÛú¦[‰TïOY¸“úE[Ð)“Ž‚ꛘÐHÝ>¢¯C/QzrâOH—\ýûðãÄK\I]§†ý%Þè)øªï֠ù;Ø¿x]ÚŸÒg øÖö‡ J¿ž ukÛ ž®¹»òÏ7S_ÏñÍ‚Ïñ¤ÀW™·b»ˆw¥ãßüDŽgÞøD\8¬‹øyñ–˜[ëQÃ蜀¯¶Œ|óùýNüâqâ©Ö‘ú†K‰ ˜ÛkÂiÅ?h‚÷ÓÎ~ßF<>ƒµ\À™‹aÝè1U® ¹âé*öy?8ŸêJ+À;5—¶¼®õ-=é’—Á§!°6ÍáÄ¿h®ëJÖ;¼?×¼ ÍB?*ê0jÞœúõÐÑ}ÅÅà…+©Û ž^n!¾V+80}EÒ#‰¼pô¨œ:°k¾Ò¦4l¹ÀÄ+“N®æ¥H¿QøãVô?¤kž¯G{€x«+ðå¨÷þ™«£Ð‹žIü»ðæòé#©ÿ ˜õRæ@Zx‚ôa ™S^B]ÙJ¸ŽmÔ+ÐR¬›¾gñ7¤iÜ.ýFñ‚ÐK_)ÖgÀƒÑ!ˆ×£ú¾ò“êêõàè†ÆϫϹuÐG_a/ë=òÈÃh…ÿ(#ñÁ5ŸD}«À©Å—ªE§©|žªíçV_¶tÉÔÿ½8à“. ‘ÿ^Ö“pÊê Ûàl…ŸÖˆn¡þÛËœ…ì²ôqóúôÐ;u鳊·Ôƒ=<ˆ>Ø‘Àr鼨o·œÔêÞBÿºÑFÿºÕ¡CÍÜW¬P:§Ì Ù‹VŸ…òzõÙw‚Ÿ«ŽH¾$= õA:ºj±¢µ(Ä/q}5SÖüMaŽ¦•áŸÑ»°Ãì£ ðàÔ¿Ððµ\™ôð9Oø´QG®>m+úÑ…ÌmRÿ}ü&âOËÏeð[ªûo¤¦Œ9#»á½eXßèoÙüL¸vsBÕ/*}üàöýà|ûÁÉ÷ƒ ‘¯[GÀ ãõòÚ2œÏúØšÓtsÎÁÝÑñÚì`Ô%<Å}Þâæh/E—Þ¡Ág<Š.ÊHžý9 Ïgøщ¼º_zéåÁÏÅ㣬\x2븃ºC q´êl§à«xý‚­£×…kîI9üÞCð>7ƒ3¡ãêìêA캤Qgr-ýªÌsŠx4<ûØ] O ý4ÕC·‡mMsêNë‚_T¿§KW²{QIžY„ýÒ<Ÿãàžô³ØNöQ³ú‰à}E¬ Öçzô]ŠÀ¥•WÕ1ǼÀÕyœ¼Gsb5¯¬e´žíƒGz„¸>/]óúÐïó¯O³_öÁ7ê¦^»1ð“4÷Õ¤÷©yÀ5îÒÓoqWÔËT¼VLýº\®ݦrðÏ>øO«èØòvÍð¹Ôÿóôñ·’ïÐáÂó¤s/ÞíIxE…ô> ν‡~œApå—âß·‚Зrü¼ð é [7y“ôêЪGG¨ƒü€>ü˜Gî÷j¤n†°]àÕäMÌk…—®¹‰“Îg~A?ûrý¬¼ýÅsìR'øy#×ÓoGXG9žuµ½ðÒØ7Û~‘~Ðjè£(ul†_¶3MW>m]äßâÍÑ,}·\ÜE\ÐO\!ü˜9Ï>'ü;~~§xéÒâ¼|IÐaò•Ì­ßCÝ|\p Ïg5ü¸âuá½Ägà9½Až‹æá2?ÌóñÈ>ž7};VD?uS¨OÆÏÕ`gÑ»ñïJ7?¨xFuèòKédgÀ …Ã5ÀëYÇ÷é‹óûˆ;5/±œ8„û#¾tÔ],~ëÞð0¼¢x:š/0ÿMzµêgV>§>õCèÁÇy˜:‘ê²êÿëa?H':¯•¹qªçäò|pänú«›ù^!ø›~k ñzœ¯!|÷yPÈ+sz'äÁÒ+ê ÿ@g·½špÔ~êÿ;° âUŽPéÅÎg˜ó\Àþ­Æžî‡Owˆ|é$8êløÎÊ·KáÙ#!Ž•ž÷)òÿ]¯ç¸q#<ƒ ò«Zæ¢cõ¿í€щŽ‹p¨-ø×SìßæÕí¯h'Êï¯m âù«?Xúëñ›ð‚©û½ø½üëFøwØ¿ÈKª¥®\CŸ¶òcòFéAǾÁ;Tÿ}åš«íè$y1óºá™€' GÔ¼Õq„st`oÚÑ“¡ÏÅ*XÏóÂ<1§~èÒ‹Ùø^|ìo×Äü–¹‹¾ý¦!ø[«ÁÍÅ·, :29>-ý’ûGÏë´uA?FõMûsꎔ‹ï°“ôaåê™ô¥©ßWkG£ŸúO3û1Ã:ë7‚c èŽ#ßhbž æµ6`ߥ·±ût˜<ä$}“ô¡û#Ø3õ¡cÈ<ÕÑNÒŸ¶ ÞÜ øw#àk™Ñú9±®¦|¶¾†êc舠§ÏÅgÜŸ}¬—–Pï‹¿O»µ¡ß¿¾–êÞÅá9Çd&ñ¹xÍê×?£þê…5ôIî‡gØB.½6æ¾EÜù(ó¤¨gtÓ§(}PÍuí ù—«ŸSsP¤ç£ú=s·¬;ÑJ|}OÞܧ qj9óB…»¢w6~1HÝl?ø[öMú†»¨‡Ã'ØMŸ[<øp«.øêï"¾ÛÏso翧 ‰<×ÃÔQO៥C¾Mº2ôsÓÇ8‚. ~?Ö£»‰ÃŽ3ì ø„tkà•bGÕ÷ÒH~Ñ ÿ¸»±xCscNb_Ë>œ0¬zL¼Ö-Ô3¤ƒF?Ž v8Îcþn..à}æz¸æ‘W’o/ zsñóê o¢I¼þxß4Wþ¼«ÿ]zçeàܕଛÈÃZÈÇÅ*ñZ®ÞpÇÜz£¯ÜZóD՟隣WÏzdÿúÆ'ˆïáïe˜û Ã!úºÑeðµè¢›-]«½¬Ë£ôÕ”ÐOÁü!_Á|%pWÿ³øOÍôÝö1¿£Hó”Ññ)fŽŽê„Uè!‰Vµ>󬘹ÈêŸT?¼ôW´TfþlÔaUý¯1à^†~B-}dM!>9`é9üé}uÓ2pÒ¨÷Î<(«ïíg=d½ ?h¾ÒòÙnêÂíÔœ ÷Ù™cmëÉÏ+¸Înð´Npä5àäýðWÈÏڨ˃£øö€ßO:½‘Aìþú/{àUÉ_í¦Ïä(qÀ.âqæ]X?øbuÕòŽðlúÙgµ!óÌÚè¿9Ξ£ÄŸ#Ì£v¡üyœf„ºzYv ]®nê­G©û!ncŽžÕ…ûèÒ½há>·ó¹ð_éßÌå¯ÜñâûÂ÷ãûÌ£Šs©cïðwñásÄï/“~C }8màEÔãç¥çÙK_!qDÔ'+…¤9Ëè•çøñÌw`Þ©sÝ^„Žtª¥×VJ]Z8æ¼éA×}+k—/Þ†xâ—Qÿ硹3Ìß0ôœ¼#àoÞ€žÔ }qêÇQ^F"þÞ^ú›‡Îÿ¾¼Mü3ôº¼)àu9œ Þà® ç¡y½1¾ÓœCéRÕP@ÝÛCHÄÕWU„n:¼×t#½ÿ¿„¾ˆrê,Ü_ï¤Ï¶šyêWbΠæ(õïÄeØ™Içã(‡‡‡ç_kNz5yyýæ½a_ÚAæ†/¿Œþ€ÔŸzÏáõJM}a¾ÿÏyë¹XõªêöëÑ£Ýï§ܪ:©xášü,<{õ à_vaÅÇÛAÿ÷x†;¸^ø̶¼°#Ä;9<>Å>â[ú¬\óZÁÿ‡É ‡é78â@—n/¼G{ <ÿ(ñå)âjéïŸÂîdýg½Ÿ·!} ôk½\»¿ß>oËè¼Ls/mcÞÜcÍËì#.ëËç]6/ †­×rtO Ñ/ZN‰®ˆïÇzÆ÷ÕÇ‹^ó"¾îéóÑ…gÞ /D_…8ͨ‡Ö.g~V#}|Â#×·Ñÿ$^Ø êsè™ôâ{ÐM¬€Ç+}ÙbúiT§?ë§oj_ˆKrø õÍpzæ[«úhéÑœÔ*t)2GÉÕcà™ÕÑ6@~Þ•WßAÔéC×>ú5ã熘ÔD>ÓCú9øûÝàBòÇ°=è {{+ý¡æÀ¯pñË À¿ž&¯¡.+ÿ©>ý:æ•œ€ÿw†øGõ…góüñiòÁÓÔ¡:ˆûÁ«ãs8¡9¯ð”öoî£ÎÒŠ=.€6Ì}ÿh?}á{á—6GæÇùûÕßúð¤ãõ†¨ ® :¯ž¡o$#;L¼ÙŠŸD÷Ìú‰ëZ¹eèè*žm#Ÿ_²¾æm5¡ëPp ¯$ŽþWNTpÆ\œ %žP}Öð>,N®z¿ø»EàuÌ›Q®¹ó+CÝÉ[éWh§.q, á×Ä‹èQGœ÷(xr+qâæ8Ç9VÕØÓ~â¡^ôA#‡»á·™oèÅêûƒÏA½&êï27Ś¼d«oÒ¼¦“ÿùnÔ÷:ÐëC§0Îó"ïj…ÇÖþ"¾Œîû–À;qésØ!}?ì ß _­Þ\¸¹ú4çPyå ôYŽã?O°OáóÆù&Ø¡úËÛ±s !¾ö8çóiÄK§¦z,úQÏ£ƒúY:wðŠL¼ÅAæõbÝÖùƒãg4ÿ\ðYúnÕßÓÿý ¸S¼¦}y8ËÁ k` àè4Y vøYÎçq׳ðš4oPø‚ôÄNža_Ò¿mgÐ7ÙËþ;¾ãiìÔžoö y¼ðž5ÄëUà—°üyŸë¹þ°nâqŸ©ï¬\  ÞLÞ€ôôÅç©W®ãz÷–Á/Üú¢>ÁqæzhžL_àÛ{?:Ô à-àŠ;àûƒô?£_*ÁÉÒK„ŸÐððx}š#¢þÀŽ¹æeôÒ÷~6‡ïmb>ýRúß…³;ºXûð{äqÝè7•çù¥EèUV箄7¶‘8@}èùŠPoúâà°þ+úô¤C^ >Åœ0ל/ê–~Lzzè1—ÐÚà[mƒç*=ížO98ûIöE3ù°æ¯®J¿Ÿy7¹üý3ô·ÇÏŸÏ„9êó³}è‰I÷¦Ü¢.ÔÛ¼ ìÃ/<£Iùe-¸øAêà.9\{õSêÈû¨ë’]#Þ’ŒêíÝðpÿ‚ßï†7Ö@Þ{€¼§‡~ ½Ô³ÿ¾UýuÏÀók ûÈN’gŸ@_ä4}õê{CçÓV†ëÊå³Ä ΀sö’猄x2‡g·|^ü{?Ésï ö0OuñWž/n%ð-¡´|‰9Ø>;¯î{Œz¾øeÚï…ø·!ò¹õø_ñava_ÔgØD^'|¢_ºõØ Õ)Jèã픾7uHô¨lóu[BýÒ›ÁkÑvé²i yJîþ³—vSoÉÀïh¤?@/»×@}¸\s3N³ß›ÁûÊÑ€§©~~ñœ¥û+½ÄÓ màp ƒç ±5/Yýû¨£– “Ù  ÿ^õÍ3°Fp8énõaÇÀÇ¥sÝÊ:êû<‡û`/û‰k#¯ô°ëàíw“ga?jþGñØ ‹lÝð ö„øÿœ:ËRøt»¨¯-§Ÿg<àu±ê—kÐ+žÝÖ€¬æøºàš‡|†<»¼k;wŠøvÇè~ÏñîÀ!ô}xQ~Óè>Xéìڡǿׅ¸<Çëà9ƒ;j\ÔŸ <«Qö~Êfâäú°Žâï÷çVƒÓTàw‹©;ˆ¬9Õè$Ãwpx7ŽN˜×ÀG׌·7œ$N?gø‡p8é–tß9Ä<ê~ÇŸ….…æ ¢«ãº^ôYè‡R?»t%_ÖN¼To§‘u-ÝBé}.‚¿T†¾«tø4ç ]öÞ¾1›9_ê«åz]ý‡eÔÁZÑÝÐ\bô½…ýTòœ.ï½8ÿ±¾PtÖ¬Z: àÏ2oövg¸°Bs¸X×{à“I×Oóq¤#']6ÕWk©kí?žO=õ…Cð›Ô7ÅzpæØŸ²‡8ù`žNv~¾ØÞóyp}ÔÁ±vcÏ÷kRˆ¿Ó‚ýd®í¤æ/äÏÃô…ž·"\@¸8öÿ¸‡øù³‚>Œ…9®ÞœÇ§ƒ‡¦º¹£ë¯Cs3þ‚N¤t¶€ÃôÒßÖÃ~”nnIˆl>œaìÀ0¯O€_TÒç"]˽ðxUG*¯GÖ7†>F“x†¾æ½ðò÷€_ÕäéNÖñ÷.øæ[˜O¦>‡~ìâüÉ~ðÛAô^Zñ—Ý!ÌñÉ‹Ä釽\\ý?gØçÃð؇èŸÇ¾ˆgYÃàT#øëû‰›¥Óó}(G¥£>ÀŸÉ·ÔÐËþ>‰NÐ pSôú­…¾ø•á¹[ü´Nâ²nx˜=ðK¤çV~}ˆ:[!vjür/ó·P§\Kž!üq{òmâ(ú:#4Äþ* q«/¦‰Í9q¯tuÅ7D/ƈC¬‘õJÿ{Ä—á5z9ö osd¼-ðwœ¾©¨·ªüò$ú“èÄý¥¹Yµðú–Ó?ñ ú™ B¼èÌ_¡¯Î Á5)³xÇ Þ¿'ðÛsñ5þ[}¦Êwà­ž¤¾[ðk‡ŸéÒg\ð|­ƒÜúfì÷”^¸TÔõ*g^|=“þýÁêÿuƒúXš©£ÀuÖ7SGCÌ5‡úvæeÓåðä}#ø^ý$ôyœ9œ^¿Ks¢žª¢¾F?¥tºâõ¯$~SE~ù xñu×£ììŽîCŽ¯ò§øº‚º¹ì´úЫÀú¹oûèãj O<^ÙA}¼+ünä«â¿£ïp>âáHaŸæž/¼Œ3ÌýíÁ®k®z?úÅ]Ø›½Ô³ÏP<Ìõƒ³ 7è¡_ñ~ö8uáÓôeˆ·´Ž}CÝÏŽ7ÐG#>Òø_qôIüá0þŽº°µÂ㿾ýÚêš¹üý3éÔˆ73ÿ¬úH?q…ÖýÀ³C,þÞ0xÿ\ú'kÉ4÷eœÿå{¨_ô¡ÇÂ|9ë@ßm-û|?~…9ö$yY }[-ôT3'–:ºÿ9ìKg®›÷†ûîÕè63ß×JyšC(Ôís}1ðb×g·ãß…”RÓ<_Å¥ôqºâ˜ZæØî£O||h?8 ü¸¸.j‰w¶¢ï±…ãàjêÛpv“ž÷ï°ëÍAOÌzé7ÛÍzÐÍk£^!]Ñ­ðÑ´FâPõg=ŸñRp©&ñ"Ð'ÛÎÔˆZ)uvüaœ? y‰ŠÃw`W¤ë®yƒêï®d+÷=ìIç„þî s¤s<@]y øĆ´¬1Þ‚ñ.ú•Fd7À‡ˆG*ˆ÷é×ÍÙEü)}\V¬¾ið{õ£W±ÿûáÃîã~ªßøxV#}ý;©/ƒëj®«¿íOh%#³m?¸aý=ô'¨!Úé¶òüöÁ3èGÜN4ÂzÞO|ÂÜj—NH7ñÊqúœŽcçGÈÓ™[oô;E½ñzO££ ¾ýùºô’NP?>C]±$é°çì6øÉnúòûð¯¥Ô‡Õ¯)ÝIú•¬ýŒ½àô%Á^KÏ=ǃÏòþ}¹Øo<ìn扗©^Gü!Sõ‘h^ú|—3ØÕ£ð):Á ©gYkèÛ›t~ `Åß{ܯ˜ü“>âÈ>a»ßú¬í ü˜Ýä}#àa'Ä?#ßm'­`þù qi{ð[ñ<Á_n?YK]÷|tpl>þ0zQ]à~ãÇÄÿ:ž¡ÕºØåØí“Ô/©C¯•Svv7xæN¢.ÜƼ¯:Åñàk=¿û×n^Çü8§O`¸D?ó^¤wÞÄyTIW\ýrÔ ô|ö¢#´ƒW£ã#Tóïçû{¨ŸhNÈ~xL'ÙŸì+pÿø|Ñ=tôã|™}Ôå‡CÞqN~¯¾_é2USw”ŽšpMé4|-~_uC]çRÖç¸ýgÑü ›xµKòô£ëàÓTÑ—Øp}¯"Nmb}=¿ØÃç4‡„þAï ¿¨ ùBüýgñÏ3G`¿t†ü¾9ôáy)vï÷¡’¸º“ºO5yŒæÅï$^jCü ë–>wk库vƒ‡÷‘¿·°¾¹žnÖë<ç!òÎcøõnxÁ>Nº€óì'Oìf~ÅøÒ>p¡PçštÞ…ïsb¬)|.ê&ï þaÒùy}‰GVÿLxi<îãÔGx>Íè¿ŸÝM>4Â<ùêyyó0ÈûýùÅ>üÅAöíAò½"xäÿ¶;¥>©aô‡¨;u²Ï4×}õîøÒUØ îZÆ\‰SÔ“ªÂy¸xo½<ߎ¼új=}´ÇBÝ,× < <¡ÜDüzé¥W‰‡ÏÜöP†ÿš×“oÐ7`à¼eÔyÊÉ›N«ßy/v]xQ;q¥úžÔßHßb.~B·˜y)VLüVòÅ?š}¡yEè÷¢‹÷$ëjzØIâp%õ kýIç&ƒNäÇåü÷±„ù”5ðUg ¿€Ÿ)}¸zæúÕ ›õÎÉjÈcÐwòG¨›®!nW}èOàuáï>@¾ZŒ¦ô+Á‹›±‡ô×`_z˜7ÕŒþÉóØé>t(Úá‹ üÈÑ…4ârËð»#áõ¤ó{C?ƒx§éh õfÏ^W®N®3¾.Eߤ-ØÇø÷6æɇ8€~Cös~æY¼‘&Ö¿ô¹÷`¿_Á':J¼ÙMœ×ð˜Içc7ú©·îb_&_ÚúÐuž“ÎqJŽ7G¼ü øÙžà'7}IöÍsÔMÊÏÑ¥O+^é1ìû)òzø§é“ì$Þ<Å:8E¼<¿ý8ôÓý88Ñ^ø‚™u ì[§)ÎÙŽ@UЯ¶!öéñVWx^®uØnsŠ<`ûJxÌ2xÒ'¹Ÿ5èË^îá¸]ÄC½è§•ó<™ÿÝH½ƒþ¾XoßÍyUæñq¥ßÈœL‡/ì-Ø)éhHïEû };u9u™=Ô»¥‡-=ÁfúP¤©ü«;§ç¼Šº9~Þë¨ 0?ÅNâ×2ìCñâzÐoXI¾¥¼ÿ8xc }6sÑ«BïÛV1gQuªøñCðÊö²®ÐŸõ|þ]s¸>_ NÒ„_ÔœÎñ'Ò«&ïo¤?Vucôü½ž:K^_šk¾“xmêgP½¾œp|z~Ç5Ç«iô^ëc=ŠO;Œ~ô4¯ºþÉoNºà‹áýSðµFàÕÂ[Œó¸÷ƒ;UÓ×ÛIße¥ô èê¢þ)=…ýäuÒ¯h¡0̺jƒ¼WóÄàé7î¡}ˆõ;D}f˜¾®ÝćýáþåøÄÑ;À·³n÷ƒ;êc“Î{˜¯4DßSú èZ:ƒ/à·˜?cἫ¸?êhÃÿGgýxxëð4º4;à‹ ‡¡ï¢ ^I[ȯ".{Œ~)tíì vý`ˆc}M˜ as©o$o®uÐ\ÞM߉î¸_Ž_‚N_qs;øa‘ú¼ð+õô­+ŸìÅþÂ/µ]Ü'ô캉âÄyæ¡îÓ¿€'ÝFý¦žù4ÌW^ãê7óá/ÖãKžãÃÐÏ'þVüL=NømÄ¿ä÷çœÍ[ØWuôIÇ#Cß#º ^Ëþ÷Y^ïúùû½ÔAöRo?ìZ´ šK >TtçÎÉ'°î{á{–°_4ou=}­ÝØÅJâðvìíNt^ÖÂã¯DsÒw€#ïßÚÂ:Õ:WŸ„柩à&p¿rêÇx1»èCOœ9➉8î/àß{Áñ•'gŸ‰WX¾Ûúgì@°›“.„çu é$zÈð´]º ƒàOÌ9ÈÕ7CþæšFý<âôá×$Ojc}bŽT+qF/û´»Y/»õ=D_àýnÃðÂw±Ñ ™tödè#Í=ò¦½è­G9nwš<½ <²<¹'䟓Îõ˜ø{σû® ¼3rtDí(ýn§ƒnaÎ^ì§:A=÷YöK üøvñÙا»©¿ïâþìÇ~ì'~~¿^|;ì›æ Q¢žÓE¼üxo/ó!OPÏb™uP·®?LU¾BõžpÉ6êÙMyvåó2ø7øOæðÜv“Ò/eØûuöÂÀ#wæ4ºúE¿¥¼*úˆãqã…ÑÇùŠ¿Oÿ±øÅ1aN‚wRO>prô#½!Ôß¼|c„¾BÍoe>«õRÇ¥?ÓŽ ‡~>RsÀ×¢î7ëÊ™KkûÁ1ÚˆÇæé l ýƒš£ãMà™[‰S6¡[Nàšû^žC·CÏd\²œõrPóàÝ \/ââ}æž;ö2æôUXyðk®~¡*êˆÏÃï¡E¼äßy+üöâ twíYôþ¾\ë€%ý³úÆvÑ_°óU¢Þ·úAö“wiÎÝüÄaüoGž_j“>58Þ<ò‡ö€OÅÏž†6ÀñÎ\'ÛG]§ ?Þ Þ\ïCs{ák[Yà¯ZxqøÓjxÜÌ5öò&ÍsjFgC¼ÿÇÑé…Ÿ‹ÏЙ–ND/ÇU½¶ ÝÍ5€Oêðzcý–ú¡+î©|0o ®§¼âÄ7Ò4ËŃØ_Œ9|Ò1±Ó¡=Æg›©·®GG>…k^)qˆuË?1@}¾ÂCqâ÷«ß›9Á¾VsnàMˆ—"æ=y1߯D'a6ù½ô°CýkÍAqpQ×\eé6ˆ/ÝL|ŒÓ‰_žÇÎ÷À›„lÐ}éÄ—bÈíYê·/Ð|Šû}†þpÍ?Šý€:¼æ%´‚Ç•‡ýïâ'·/6RÙÏþÞ¿V¼¤!>×E_8Uäsˆ‡¥8åûˆ:Óá Së§ ñ÷¤óá÷Hç^s£º‡~º#à6/Po8N\Ó_@ó<öQ÷|>ðÚsõúPJˆÅg’þã±ÑóÄrx18ðμúìáàÂuà{”ò¥c»Ý?ñœ?æ+À©¤Çt€9 àȹ8‰çG=ÔÃîöÀ>}š‡=Ý…^J Ç=Nmúàᶗ:NKÀ§sø*ý"Ãôª/ox\=õ“¥þùÒ« º˜1+Ë;Oæêúú°£Žm ÷Yüínæþ°ß—I?îyê-½àËúþQúx7€ÏjŽx·õýÈI'œÀNž¤OJu'Õ°ß*±›­ì§jp ôš¬•|° ¿q¼à0v-œî›â}è‡ê#ïÈ _ß·ݦ!úÕ†8Îæ7í îu¾IçSO„7ÜÃ>:¬>ú †î>é¼Æ€³§þÙFÿÍãäÁƒàš/€×ˆ¯÷pÿ2®WõŸFð•ãô½ÀC`ÿÛiø0»BÝ.gÁZ°[íô»oç¾õQÿÛKܲ»|(ä³¾~s+ì} àÆG°CG˜_ÞM_˜ž[/ç}Rü}ž;s¤/h‡Èkkˆ_47R:†;àçHO~½ ÑO‚Eœ—šA7®;,¾œê©áûïâ5?0¯Ó6P'“®ŸîÃqü³øg;ˆ[5OXû»»€Ž x ^I?'öÐ7£Ã×ögÑ›ˆoa7VÓ'ºœ²5Ä]¹º$÷aÞZ zVàO9 t{Ôg¸Ÿ¾;ôrù!sßÚÉ“ðÇqn|¼·ð wüÙJ”9™Þx ®¹öÒ‰j¡î×®Zø²ñøK˜‡ZêŽþ[|ÿ 8#}fÞÎs¨…ÏÛ̺yºU} ¥ð”9Í.å~4`ÔÏø¼åÌ-:ï¼—\=‘óÏ;ï5/Ëþû›¦Þ’ý×+³ÿKþüxöãó>ò7ÙÜ:ù–)·gÿûõ|î5|6yÿ­×Ý5múíS.¾eÊÌ›§ßy×ä[¦Þ:åâKÆ_{õ5ï᯾ö®üùÒ³þü¶™3&ßzû˜k¦Þ:ù¶©·N½ñâÉ7N½éâKÆõ‘K®½yÆ”ézoÌØ÷3mò3§'?•~î}‡áóúü賸èš;n¼ñ9ÎE×Ü<õÛSoyÑï½ášÉÓ¦Þx×´ýâ[¿8õ†ÿûâßüù»fLÿæ´;fN¾åEᵟ¿cæŒ)·ßxÎßøù›§ß~ÛÍÓgį}ó>û¹·Nysî¦^yñU3¾™½¼ì¾ÿáÏÿ0þõ¬ÏÑÏ:ÙÏÝ8sòMSnœ|ëôxÔ±g½ÿö[îš>æ“·N¿}êÌéÓÆÜ–žàä™ÙŸžxÖ‡^ö‡.¾dÂÙß¿zÊ´©7L=÷Ž]~öº¸zò´™ÿ÷G^{õäi/z[^ý™1_ʾׇ>ÿ™1Ù5˜=£ü5ø®ìaÉį\yñ¨¿ýéì;×L™1ú£úÃÙŸ{ÍgÆ|~ÆôsOãÙ¿ß<åÖ»¦Mž6ùÖsÞ}ýgÆ\½â©ÓÏ}+ûƒŸ™rÇ/ö•OÞ>}ÚÿòÖ'¦ß’½çž^ö¢?äY«gÔßÆ>é^¾—Ž¾³¹ëö™SÎ9HökWÝ~Ûä3_ôk¯çÝÉß<çxÙ_Œ+ô’ËÎzãÙ7Fݷ䤯z±;ùªO»ëÆsÎèÕÉ_§ÌHÖäÙ¾(üùEOó³W~ã3reô’<ûìÞ|4{»ÇL8k_ŒþøÙ;ám|üÿøÅ7~|r0vñÔÎþ׆w§Þš\ÏgßÚÝqËä/nN?6ãŽgþÿŸäßéóÿ×YþÇŒ»fÞ<ã1º¯ ïKpö—~tÊÍwÝ4cúäÛoœ>ã}7ëj¦ÌÈ>ÖKÞûžìú8ûÄV鈙š5ú·ü/à-7Ný¿}Ê›“\ucö„ÿ—¥yÔ™“sëjÔ[žrëÝÿ‹e¼jÆä²ãÿôƹ*û×é7Ü1ó®¼'tåÅ/úÞÙG÷¸1Ÿž23ø’/¾ ϾΗ^ÿÙ«¾6î«g¿ºtÔ«ñ£^]¡W~oÜè——Ž~9~ôËËF¿œ0úåÄÑ//ýrôqÇþðøÑ?ú×}`ô˱£_Ž¾„ËF_Âe£/á²Ñ—pÙèK¸lôY]6ú¬.}VFŸÕ„Ñg5aôO]1úÝ+FŸó£ÏùŠÑç|Åès¾bô9_‘w Ñç|E<ç¿I^ŽýÀò^Í{=.ïõ¥y¯Çç½¾,ïõ„¼×ó^_ž÷:ïüÆæßؼߟÿ:ïxòŽ7!ïxòŽ71ïxóîÇļû1qôýwÙèß7áy¯Çæ½—÷:ï÷&ŒÏ{}YÞë y¯'æ½¾<ïuÞùå]︼ë—w½ãò¯wbÞùMÌ;¿‰yç71ïü&æßļó»<ïü.Ï;¿ËGÿÒÜz{ÓõWŒåÅ×Θróä4O ¤Þ~óõWŒÍ¾}Ž3Õûo»þò c³ï'Þú¶l“Zøàõ‘w†|tê7ƒ¿¾ý®iߌ‘ƨ{ãõ¯¸üÊ‹£Wý;o¿~â¥ãÓ˜æEâ&}è-×O—=X6Žžr÷Ôð#cÇžý·^?ñÙc$yIÎÿ]6ú’'ŒÿÀ•Ÿ“wäÎrBzK^<|CöÝqÙoŠoõÕk¯¿lBö«c?‘ÿoÊÆ¥cƽïÒ17e#Ú)ï»xì˜KfN™1sLâà¦%Ž~æÍwM»}ê´»¦½gÌÄ1·ÍÈÞÞicÞwñ¥c.ùܘðæô;³Ÿ:ë6_6îŠ+/æ‡/3ùÆ)3GûÐä“If:žo\xÝnæ?ÿ*ûŸÿ:öìãÎ~qéÙ/ÆŸýâ²³_Lˆ?<6÷ÃcÏþá±gÿðسxìÙ?<öì›ûáôÅÄx”ñ¹£Œ?û(ãÏ>Êø³2þ죌Ïå%ק{iòßñ׫q£^]:êÕøQ¯.õj¨WãÑÆÏmÜø³6nüÙG7þì£öÑÆ?ûhÉ« ñ÷Ç_žûýìŸõûÉ«q£^]:êÕøQ¯.õj¨W¹«¹bBîhÙÿ>ëhÉ«q£^]:êÕøQ¯.õj¨WÉÑžOþ÷ÿ>[ýo pcaMethods/data/metaboliteDataComplete.RData0000644000175200017520000006350214710217306022130 0ustar00biocbuildbiocbuild‹Œx]W•ý‡a&ô: †0$NB`f¶ 5¡…aÊ€“bHbpHB !ô@z#ÅvÜ«lÙ–ûVïå==uÙ²zu‘lÉ’í”ÿ»÷üÖÅïÏ÷Í€¬§÷î»÷œ}ö^{­µ?ÿ‘/]ø¢/½èŒ3ΘuƬ³ÎÌþ_ö¾`VöÿyƯ8;û߯½aÞ¢¹W/¸~þ¢y™»hî‡ÜðÝëç-š—ýƒ—qÆ[?n_ºoÑîoÿä¤o_˜snñŸVxÉ=×eÿW‰­œ÷ÌoÛvLض÷ÿò?9>fK>ýÁš‹W}ɽ¤{äÍ>ß?Õü¢ÿøî2¿'ÿÂ׊þðŸ|宽䶗ßãׯû‡ Þ1t«ýüïvŸùÔﵶk §.ݲÔþ×GÞ·÷¥‡üá§?ûñ׎þ½=õ‡cW~ûµÿè}ø=Ëo»åzÿ]ÛçÞPû÷_·'öþá·ßë°¼]ï=øÅÆ”g¯òDû?ûw¯¾²~lõÓ¾é%›êÚdú»-‹®ŠÞÉôûøSλËî|}ô/óèÕ9¿Yl _~söÊÞé¥ÿø™ï½eƒß:ïÝ™¢¹¯ö²Å_ÀZsS?úÅÖýž~ÇÐ- ÏŒûSÑÕ}æûþÔœO/[Uðßþ«»>sàƒ×~Ù=|Í[{öYÖôż>{ ýîãîGMõ5c—ŸûU_ÿÞ7>¹÷ûÖ½Ë×[í'Û?7rç?n·mñŸ}Û¿óκù_>ê[n_pÆuí¿·¾YwùÑW5{Ë~èˇ¯9×6œ÷ößrÖÇü·Ü÷Úè²×¯óÅUïú¿'Þt¶o‹_~“müï?ýÇß—ÊžÚö_K*óïðç¼*ºóþØy…üê»üÎ+þú ßïwÞþ·ïï›ìñ‡Ž¿ûÛþð­ÿÑúÓ«ólýÿ²¡ÿB»;¼¯ÝûÖè,´5ñãø®å}þ ßXsm‘7ü§U÷~à_íæð~¶öá²îŸÏÙn57|õäºÖÁäy,ÏÏW4Ö­©yâ¶ù¢Ç^¶÷À2ïäzîØ]À ¯÷ø m[ö[­}ò‹¶õ_>ú¢W¼ÏrZ>’zõêµewŽ.˜ÿ/{|óù/ûâ ¬,ûnwô¾ÃïÝóï—¼rößùöÅwŸßùÒ7YÙÝîz¸¬'ù~Eá¹Ùº7Å ÊR/¼kûβWÚŠÏÇhwgŸò¾Róú3Òy¯ûR·,~gô í7¬›ªì*¾ûÁÝžùíÙ³:Ψ·m~üÅëÇßj…OýÝUïx]¡íïÌîv_}l÷Ï}KXŸ–sóñgþé¹óõý6¼çߟ´½}ÓWþǼäŸûùêû^bùÑ]ûè"¬3~¡/ëÙùò¿öŬÛÎ˾ôŠ«¯ÌXuøÞ¶'Ú…ÿ»Åʲ»-»ålîì ò÷¯ñ½=Zñ¾"ºÌ]xþ¯ºñ¡»>k9ÿ}áܽ©ÍÛã__i™°.|Õ‘ÙMKæöYÛâo©ðùÖÅúܺ¡ÿ‚©Åw{á?_–?rè¤uÆÛò3¦õÿƒ‚=Ë_ùX§¯o¤m ûÇZsé§JWŽzûoK¼ìn4xe|Ážßþ»uß{I®?/ƒ9¶ãå'~ì¿.ó¥QØYÞc[³‹æm×½Ès¸Ãáù{öwWMÖ¶ZaóM¯Ï† Û“ BÙ%ë˲«±pêýV”ÿÆhgXÝü|ëýoù‰=’]­ êϵ§^Ûý“xٽġ_‡}j+âmò*{ä“ã£Ù²ÕÑ·þð³ööÁo£]ù¢¾úµ¿z÷¬—|Ò®½­ëcÎù¡ž“åF·û3l1ëôÊÇ Ù7…ûa[ˆý>ìÛÌ÷,zsü¶<¬SÅ3ÛÔÜ>þ«§whÿÚ]ñrƒyôØz'mk´-Þ{‡í9½ ÀÊ®ÿ݇¸ý…V‘Ý­ÙâWs¸N[Äshâþ¯nÿæíOÄ×[¹O! |Ë–Üü‡zðó{¬0»Šß–ê³ïG_û;ý÷¬Ó¦h™=\n¿ëÕJâíþŸ¾¸ï—ÿæÒ+lÍÿ]lÙjѪâ`¾ÿ†÷Ï ñÝ~Ö½­"þŸõSõòwf#üÿøòlT8o¢ÈõzÅ£G~zu´õ|õ#ѵM?Š7´­ ûÁÆý+^Uð…ì–ñ%Næ<o8»iõ¿¾/»%,½Ý'¼ä£üV›Ã:ˉ¢Ç÷¾ï…sËÖeoÚ¿m*¶ûÙ—Uáïý‰wm9ñiû9<æ—ÙæÊÿYùËÂwêü1ñ×tÝË6~楿¸òLÛÎQÏËÞ·í³µñƒ´Ç¯ùõ¼û7~Æ–‡øo™p_ìÁèéÏ{Ö–\~îWòßxi²ï6DÇÃ[k«Wl}qö¥¶iõ©ì;}Äv†ukë³Aèãƒ+Í£·ùb“å…ý›¬÷%Åf¯øW^Ë>¨‰VË¿m¶œ¼ŸÍ¾¡¤É–EÑ7YEÙ­Ñ'؆7Å•éœÊyu|c¯ønkNýäîçz»ž‹o꘿þCE•ÖyAt >k©ïEw·-b.c_U‡Ïó?ð9­ìÇÍ_ˆ7º/>úÃÍÏ–Ü`}%/È^Á|¿‹øZÅ~* ?9ðš­7 ç^gKB~àíÙÝxäñÍ^xÝÙ ¼þÝɾ_Ïù÷‹Þt{sû„åœÿÇKÖ>™˜ÿ }ì[¢Slþz+ù‘qî{†sö>α¥!¿ðÍœKœ·VÊ9Ÿ qÈö\¸ÿhÅœK›Sñ²æ:ÿü±Ä âÇö–‡‹/ØÝÙ ö»vØŠp>ú²ìÃɦdöHÈSlÍ»âdK¸Ï+ÙÏ+³«ñ?{Ü~ò [š½y_=¹Þ~ò2_½ì›uþëèøÎû™oåù+N>ÄþÚ”ýRٓѵ_Ö^'P¾3^Fç¸Þ7?JG[üžp>ÙO9ÿ¿­|‘}¶6ŠrÿüTòw»CÞh…Ñ1—›Ãœ¿ëâÄò{<ÄUË]°¾)ÞžùJžëuqB¬›§RñÁîDOûƇMë‡ûçiöÙ/ˆ“)îóãìkíƒ÷G dŸ?«Ÿ³ßfOËò-—%ûZëÿö‡»ãÖÖDÑñC_±ûC€ð«ˆCäͶ2û°³+ÙõA|X×ÿg»ðèw>›sÔî'ÿÑç­`ÿ+^­‰ŽÍ³]ñ~ׇãon!Íü¼o‰Ð:ŒûÒ•Ëí âëy>«/Pßq¸êÔk¶.²•Cñm{ããà¶,zlK³›£íõ­%F> }c?/ »;Ä×y¢uô`6yÊ.q[Î)ßF^XÎ'«x0NÌlë+ãÌÈ>Jžñ$ù y“†ëõØ_[¸oZ·[Ãz°Ì;^—½ÂW%çÎCäC§Ëo°Â÷ð!Ÿõû¹9ái½$ëè¦ÿÛÞ(ÈõUxnØ÷¦õκS> ü×þ›¼{ûz/qû—áûy]¼Ì.ó;Âú¶GC¾í5Ô-íåŸúZ¶”0åÑ%/Ž_Ú9¶…sóßy_Óyš¨ïX”…f—Œ—FYÞù^I=²6äþKÖÅýÄ}ÕŠ7Z_KC<¶Šìâxÿ‘‹í–pžÛ®wʼnHr®ßKž¼;Ä9W>ö+ÖÍ äK{ã´óß’<è6Öã“Äç^ÜÙQÿëZ_Íç±¾’û·”sw yñÞø¸Õ‹xœ/öK꽵ߋæU¶•zåÑ;Ý!/xuü€}9Ï« zìo\ìÚÿKãtK>¬:oo6]Xù%ß'>‡<ŸºŒ|ÁëÃùí“÷çDaýª©$Ž¶ÆÇÑßE•/½Âtÿž ž=üŠèƬñU¡^öÄ•{²QêÅ”¿yC|úöë|÷mÇ£ o±'Býa¤>ü]ôÿ¿Vm÷sî-æœQ|¾›8_J]AýiŸûËãr¼n­/‰—ÑK½Š|xuTNnÛ`ëã¹Öµ>%nÆðBñcþEÖ)×ãÆáÿǦºiI¼\osýÝ×8Èw¼ˆ¼©<'Þ¨I>p7ߣ.ä ¶>Äód}ouˆ=ȹü0çŸÖx†éù.‹+_OF¦Ïw­ó!ﱜ{é†(ð_`üÄ×R§lÈžn¿x/åï¶GÕÌO¯ñ½<ÿõ7HΙ-Æ Ü'ãmó}Ï‹Ë„·+³MQ˜ÿÜA[Á÷) ñÍ«9矠S|ÚqöwvdC½ý: öÔs븎uQÔ}[ÚZ£múÌïüçÄѦWyÞ‘x0ð#ϲô—þBù‹êÛd½pþû]¬;=ïu—D/ü”“§ùæ(<þѵ^øÞªC]ø×fîcñ'vf+¬«þö{¹Î‚Ä+ЋÃ9誷Äeã7]uÒÎ7Ç7:É؇öÓø¦ºà÷!žøæøüš§–ǺpW¾òCòË•¯‹Ø{üAêFpû0÷_qœ|Þ¶DÙÀw{}|œ¾Å—s/ŽÓ͹Þ@Þ¢|¯4ä+Š¾=Ô®zðÔà ¾‰ç±êÒ-K/zì劾3‚sÒyöP|û^gyΉþ2ÁQ6Çáì¶&Ô»¶á—<'ºêK·&÷¿è‡ŸüçËò&qpwȯ”—x1y­òÓe1<òZ[Oÿé¹ñv8;‰‹¬wŸrå¡+É®·( qÕ'ß$/s¸˜ÎÁGÞBêLòc·åìÇ5ÔëeäÔ½Â-|wôî?xTy¼o~I¼íÎx;¼Óõ}gòÉ›–€Ÿn&.(¾rŽYWv‘õž´ ÷Ú=•×ý¯•…zÕ…cì˯Ûînøï 8²‘¯ù.ÎQã‚Û”ÔGû¯åü* ø„+Nªzµ|üÀñ®|Í{¸_ä‘ɾkŽ«Ýéû£ì蓾%›_³5ÑeßÖåiîÏ#Ĺ/“×€{Ú휃iÖËžV~‹-w¨¸/Ú0¿±òË£Bt•ßpÿ]ÀÃ|'Ï|ʵß+¢ÓêÃúêF¬?êÂßy^ɾëŒoÿ§}ç¦ò=å÷ËÁ58—üž7Ä€ŽëœïñÅäû”g€ÏÚCä{:¯þë¯ðÑå¿g¹+î+_.,ªÈV ¯uâÜ$/÷õà_·°ŽÖ„ï‘Ôóª³Ö”½"ÊÄ,/Þ·Øö†g²¹Ô̾Ù…±ÿþ“ƒ× g°Ÿ*ïür´ ¶5!nZ18Àƒ¬ö­ÿ>œ#~ý8_0?9G„? ‚¿ ÿõþãËÀ}oã{kÝ|‹û·+>PWºêáqYö!ß'G7…{ù’ê¼±mœï+ÈGï%¯ù1q1·".ü’ó–øîØ×ëˆÄ-ÿMè›Xþ‰8¡° !aõµà/Ê/WÄËçj­ ˆüEø`IÀ©<÷ö}Æs©ß‰[¾‹õBžï+ãÛý#ÛðÕ8àúÆìáqNÏû#8ýjò߇èÓ|–øRðb/ûøÆh;ßôó¤t'ë~Ùh\Ài¹×Ç_â˜mˆ¢Å¢/ø²°lé«âfû¸?ª¿”×Äwë;;]ç#}$+ŒàÝ×ßçÛ£åpÞq¿ŸxÇ>KÖøˆ7„óǶGQæš_ë÷êXåcÑBÜd÷ƒ‹WÅ/û¦íŒÿÿ°B~_–­¢²+ÑÁq½ì@\¸[úßãÅ*ãÆ“ÎO«8A¯aßP×'ý Î5¾Ýú"¦~™êYê"õ/¬=^^4ååœÿªS“ºäð2âÊ}ä¬ÿ ¸¬SÇçÆí‹x^À¿’¾E¸‹‡u’Ô_ Eþ³‰¼èÞð³ ÷¢þûFy¤¯yÑMÖw!ç¥ò²§À  ®}£ºNxÉjö×ωÔ/Mç4ûÎï¦îy"ìÛÁWÛwy&¬kÛÂ9®ºj3q5ÃzÉ€›~çñù0ýM»šzoÇ[¢ øŠ}¸øyy¨‹mm||.òÍ!ž$׿•þqƈÛ.¼­nq8’<'7-^òÍäùë{).êß âãú S_“zÃ÷”çß“o<{,äöðŽ]ä­P曨§‹Á7Fð|ï} çl×KüÖuøRÎWáœ!uÎ_âï 瘫ŸIÈwÆéûëmu­ð¦‡Cœ2Å¡íqýt‚¬‹Ž¿þA¸»«£¸¦8¬÷§/â{xÝvòIúg¾ƒ¸Y¬ü ì:¯„3{ÈÓüúêÃåÅÛí®i;é+ïY¼7.„¬”ý ^çEœkäê—o’ºK}¹¼¸\ŸmÕÄéÚ¦yr¯­Òñ‘ÃÞÄó(ç­—¼).p|ý•"òîQºü/>úÁIžWêQæó_fqxÿŽ?Åù.`{£lñVYKT¶%_8ª-!Žˆg ýpO o¾ÍšÁåèû$ùH1uÌŸà¨/[§}¯S_ÒvÅeð¹öMêõñéWxùÖÎáíäKŠ{O†õmœ§¾'û&Ùk[ããájÛG?Yý(åÍê_>µ©>®Ûãÿd{ÎŒ R[CÜQ^S@Ÿª†~çªÕ“÷;ýûZÞ§8:FÚvZÃ×âd›B}nàȾ5à›ÖK]U¢<\Ïa)ñ¤0ÚE?z³€3×sÁÙ¼0ŽöIÖëaúWÅ@²{ýSß>ç+ñJÖ™pªðÊ6põ{á³ø&âW:ä붖÷ÞBžmÇZà/¨ß(>ÂzÖòü§ˆ÷“¿•@Ìœ¾ÙÕ!oó_QwþŒzÜØöf“²·ôÿÉÄÓ©ˆË‡ÏÛ.ø»vÄ78ɳ~ >©|Hý‹ªìa5çÓË-—|d}òПHðɪøk¼Lu©w­÷ƒGãµW¼÷Žlém äyÑ}µ•GéÙOîñ4üú&Éúë$O½\v ¸v*ì[ÿSȬ>.Kn³1œtƒW¹ú¾äI}ÛÐ'LZ_^Jþ—O_|cà%uºêÎ.ú÷…Ô«ûÀëÁ_„«µR*.d¸ß:w6_ºÖï`#ùÍE¬ƒ;È7÷°>·E«ù•{y™ú§é8Müž·Š/¢~Ѷ˜~ðC»—z¾]ù9çÿ>ò+Õ ð6<ï¬yÙÔÉ[è5ƒ·g™]Úöd ÿ¼Æòù}}¨Åôã”·TRÏí&În‰þìœ^£^0wz}A¾_—¬ŸFúl…Ü/½.‡Ï9òwË×ÌЯH±@ÓœGÂ{ ÉGÒ¬×ê_íÓªÀ7°~¼Ç:ռÀ þgŠ»ùˆë‰îÞî³úˆ)Î9xjv;ùZa|™ï²aòÑzxzÛ¾ÿÇ7Dð亶$_:·-¾aŠ¿¸QoYH·oKøð7lo\Þ_'>ùÙáaqžûŠ¨zþøçDËì •–‰¶û²ÕVK_D8£ö—x ­¡¿b«B>cêÃïŽüßZcès%õò“á|´êUõ-Γ~×–¨Ì_R•ðÿ¢(ý«û:lŸçœ{{¨wÀem9xpòâ¤^^…óÙÍÊk\}0xL–‰—åm<|Îã ê¡TÜVûJr‰§&|/œ@|Ÿ²‹è€x†sbW„õ^/†W Ü&Ð5ÎKê”}Â:ø(Ä+!oÞLªüyuU<œVðõéÅÓŠ“ëöy 5Áïôèé¾e É¿ûòâÂÉ~®}öŸ—FóÖcÖxcI8„WÇ°ÛÇ“|œWç•x.¾4†¾í›ˆÇÕÔ%!žY>¸mUÀlSÀgìõãáË-paÿ.ç‹ê†â8í¸Ð…µ‡|ÓZÙm¡ï`à>V@ÝDŸ8Á³¶…xê•ô-¿·üÙÄ7[E/2úsx«òâ°Þ-7¬?§>˜3‹>¸úæ™PÛîÐGŽa«X§©x¹_iµð0Á‹,Ã÷­ùŠ—±ßZ9gUÏ©ÿ%Ü@8m†~ãþý¸5ü~ :L¬eâtûs¶9~ÌWx-ýLñ†ã·Ÿm-ý¡áÓz»/¬ÛŸŸGuwi´mguøù>OÀïüÈÅKT~‰hs k'uêòÜøx[h)οÕÔý¼¯m‰Ë“ØÞúøàNißà÷ÊsfÇ I‚Yòñ¦m)¸b8X9ëdGØøê$|è…ä 䑾Rüeúgâ3nS4ÜOWd]6xdq‚KRoæ¿4ºð·%8Vn|{Ï4ðiÛ _¾¿…ûéÅìwð_§ûç{q´ü®xØvÜZô‡mÏyÑ‚Ý.~£+Ž=8ƒ/!#x‰“oÙŽ‡x1}FÕ±ê*OUŸNyDui}ß øløÛbò ñ¤;_ÜÎõ×…¼+É“R˜6õi*¹¿O‚›–þ–¥ˆ5qº1G}e+ |hSŒ>…í/,e.#ŽÿRnëyy|ùÿîÛÁÃ…ÑϲŠÐOKâˆò“gBAc…ïdqö8ë%_Áž‚ï#¼°˜s½”z\ØÀ+ÊØï¥à¡žòòÕŸ«CÝ/ÉàfXWõob+à7‰?´•~K´M˜›ÄÑܸ]ñï^ OÛD>7ôÆø€°ø‘;£ãá·›øÖôA}D%øƒø®Ê#¶…¾Šê-ßÞ½'ÊÖÎʱFÖe!ý™íðöÀoÄwKx®ŠWÃGÞI¿¥†ý“³ &^û:ö¾·x9êgªîk†?#djitðtZ¼éNø™ÂaSñ±õú¤Ž ¾MxoêÿK×$[ùG.}òz/&__î«xpʳ}øK‹ŠÃåg\ý3ž»¯ ø«KOPú‰sf=ÎË­ðéÇooÊŒ‰3 ï<¥m?:leïpñ7›ˆoe‡óò¨ ò®2/¥^SÜÉ <õ‡¼<­:ºš[­‰uZK¿wgà*λð>þ~ô^®$ýÅP|;¿k%¡ÎòUðQÅ×Ü+ÉÿÀïMï«ýM?"á‹7ËèÛƒ8|_B¾¢¸µ:ðž\úôSÿ9Ñÿh+®¥È#÷†óÆ·?}Ͼˆ7žeBœ÷UQõøÓjk¤Ni$/ªf:ðÛ­„þ”ÖƒøKê·]u@%}ÕJpäåñÇüÔ ¨CÅG5ú”‰Þo7ñJ<£&ð¨GÉÈÿ~•‹çPNÞ-ý¼VCOfÅô…6’_ßÇó,ŠoóÛ,Ÿâ³úV÷Á?¦î«¿ áÕ‰WÞmêä ÆžÚóƒxCZø¹pJp)ÛÎ kO®s!Ÿç¾Ÿ}[OŸøa©×DrÒT×èÿqëh¼a­½J7õG&ô‹’>žÇ°ÛWm'뭗󶂸v7<“:¾ô@àf)xxà¿IÿTù–úHÂgÄkÙê_ßê)§ßiÔM^p>«#+§sÖ«x5›ÉC cùÙöú™âék¿mñ?áE±>mGæWû®€ûºâX:ü»•£§ÔúXK=·Œú© ]|•Ÿ»#‰ë:÷Ä;~–O],~ÛvøxÛ/Äu Àͯá­ûCèXÄW_ìâ£qá|ô]ÔëÂaÀí|9¼ùΨÊüúß[ç•q£Ê·Ñ?SÞ” üé’­,^ÿg¬/õJ£Ç=ïßúFþ8õ)}!k ú+‰Ò‹O•ÙJêiñp‹cäÿÒré“ìˆËÁo%}Â&øçO]ÿG:¯ƒG“¹3´ùSì_=÷x*ÔWIÞ¬:÷.ò¡=a½¸tcMð1?áï':p}?â‹é¾Ò7É'@ý\é\¤¿,g”ÓΧ¯­ºQ¼Å*ö‡ôˆèƼ~5ëMu¸û¯BzG'³­àÊôU§«…þpÎYß 8ì.p¾êláÒ§h?Òo¶2ôš…äŸòØ Îû*¾WqÈg]ýÙ‚["‚ÚG“þÛ†ð’ü¸¼?M¾¿^yN¸ß^D¾ÕC=_Ñ*^Üe÷Ñgù-çŠòÍ^ÖCuèó'º-éŸ×…¾G‚£ˆ7ªzSû¾ÞëôïñíùºKÇ&žö:ø›Â_ÅW¿Uºʼn¤ÿ)åy*ÐúwtñÁtöÀ£—þmY¸ÏN^eôǬ>—ú®y¬ñªV‰çN\[x„¶Ž¾üè¤^ÔºW¼Vݦs~¥×£ —_øâ Ÿ/ ú@W?qº=ú²¾7,\·¾ƒ8(ùVô® ‹þN±úã{èÿ-ƒG*½Tñ´—sBxˆpñÍj*bàËï#ŽQxCàÿ›úÅrt~^ +ï-fè:ÄWn®þ©ðHé/×ÁW/o|"à­¾¾ð.ô´Úûè#jÆðã×õœÝà%òÿx¼ñöï*òõ^ðÕŠè2/éIôeÚ§àꉠŸú > ÄŸ ë8ák•¨_?Fx€xvÒÓç>C¢ÛR?…ú8áƒý‘~ŠêyñwÓq¹úv'O°-ð±ø¾è?}{x¾Þø€†^Âv„zÖ÷€¯KŸ‰î'©GÉ = çºåQ_ G/Uñö¾À“ö}ÔÉZ¯ÄÑ?ãß¡âúÛ^úÆÏÀÑ!&ø|'õ·t]¿C§"¾ö¨§Òñ¶ºÙŠÀ£À·ÕpÎißx¬Vï½>ð2½p Ⱦ—óEúô€ÊW\zÁmôUëÐ)ÔË£¿ç;ÑcKRÃuЗþ/Ý©øq¾7Sg(Or‡VnŸpÛ‹¾I<¶=ìë*úPêK‘·JÿdÛðÿC{àK¸tvè¨à§Š”謫ÑŸ~Lø­ô›ÂI ¨r¨Wѹ%ü â«åÆúû­ˆøÜ@Cõ ú*£5gV [ÿ—µ‚wƒƒu’×àcèQ¬…ó¹ùmñj¦x¡uQß´q¾loë: ÖLþÔV—¯¤9‚n¾ ü·Ÿú@ñ[÷«”¸;J<ÝúV¾5ôE¬‚¼éçÍuénòvñ¿½;qª4*ƒ~¾Æ⫺y2áƒH§ú½™8¸ ?6pÂpWÕ#âEê¾w’§ªŽïâgÔípYérˆÏÚžôR–øîVþm¡¿.ÝèÖ8Ì~9ÙZ·âæÁo( xø艬ƒ|*ºàà×Ixœè ,þÎ^ê¡ôÕôÛ+Ècé—ÚΛúÀ“?Û–ÃßI8XÂÛ—¾¼™|¤…8‡žG×ç+‰¯Â ¹/½!^Z9¾Qåð©„k…¾S‚WWÁ»­ u­·×峟ϔç“ßÈWÞ©­‚§ÔJ} Ê*úªœ¯êÇ$ú@ñÚˆUWÅß ÁkCžk+ÈjèWäÄÇÚjêw†þ u¢éæÜUž¹s`3ñHzúü€ïºt.Âçvr¿KÑËÂÓÐ~sâ å_(¼ÔÔÿTL_Uü¨9³f…>Vkdõ»u¶?ø\Xš|g¢Okd=fBCK÷×öšB>àù|¿ø€ÂÙr¨;×r®ÏìŠ5ë‡ÇÒG§:Cxäøí‡ÑµÏ~ x_É}ƒo6ÆþÝMž¿I¯­ø[ N³ç&Þ¹üÄ%ž{Ÿ×‚<ÏÄÿM:”…ð}2ðŒÅƒ–ωp‚íÔ›à—çÒʧVÄߡDZ"øFø‚%¾;ÃþvñÓÄë@Ïl¹à3àØꃋ—¾Ÿôuä3†¾;ññ‘Æ–À•N4ѳ‘î£Îõ=hqñ_«©³8w ¼5é7Ág™óÂ/„~C õ]ñRykù‹øh9ðîÓô'K¨cÛГ5‚Ó ï‚WÓÿNç•ò=õ—ªàÑwÓgìõµÎ%gÿ%üžƒôíFØÏÂ{vPW¨NÅwkŒsLñFüFñË׫ Þ†.\üW_Iø“ ôÙBú—oâ\9ñÙãLú Ô«…©/Õ€n@ø°ÖÑnøâ 7“ÏUnDS(Ø›ø¡‰_§þfºZ|<¾ŠÏ¡•tCN¿V¼mùë%þʯÐñZxw ¼¼Jp£Bxò‘Þ³}†tDÕàÒÒ¯ÉßN~EíÔ‰:—¤s_X:_ñ"…wmEwTDÿÚÉ_ÔŸH³É­0ð} ý¥Á7ØÑ·¸ôÞÍÔáŠg»ÀaÒü^ñ}®UUÄåÄÄëòî‹t:â Wý¥ôÙô5ËèãÈ¿H_I+z凬kð”„ÿÁšà5¶‘¿È'9Ÿõ*¾G{ÐÇY3q¸œ>Ý*úëâ·Ò÷—<áÃJoðKî‹xûÂUÁ?­ Äßn¡~ãÞècï~(ៀ“8¾wsþ®&ø#w“ÏtÓŸÀÏÅJèCªNlÃ/®Þ]^À­“û¸>–òÁ&ê¾fþ>Ÿu§|YýŠ!ê¢î¶}ƒ Tâ+ƒ†xª6B<WþŠÂ½SÜßQúGÉ•Gë<ojý‚Í3ð¢úg'~åÔªg}óTÜH°L[Ià½8º2S_NxH|,ñÇÁ­_æõœü8vп“¿Ÿò1ðBˇ†o@’ŸÉ'@úÖ¿ïšoøÑI]'|¾±ÓgIøÆøýµ$ª þÐÉþ.Eÿ¢~ÑÎà#Ô±ÚŸ)üb*¨Ûåß½¼¿˜ó¢8à3žâó[Ø{CÝîòOÇ ñ-o@¯×Í9‘‡"x¥Ëï TÒéuDÇÑ‘'¬œó5ñMBÿoÝËy®;È×ëÙü»ç~™•áO"ü¬‚þÖNôEĉÎ õ-ñLü‘òÃ÷ñjú§uàÛòËiâÜÄG0á‘ŠÛÌz‡‡ïU|n%ü˜ð鶀Omä¾Éÿaðǵnt<Ò£ çüÇD§%]R1y¼|’Ëè3æÓO—ß~—‰ÏvxmuwzügU/Ìy!û½‡~@'ø^%ûTxªt#­Ô--ä9âot¢PÜ^Î1Ïuoåà£z®Õì›Ãè9Gá{÷“¾ºü3GЯ–xžúX™Àë·QÎÁ£è Ä/~$}ãv⯮3MÓ~]Ĺª9)øÀÂuZèªNÃË ¨wñùµ®€+ËÏÚà‰âƒiÊе¹ú@©ç=M=Q‰=ø´øn†îÒÄ“Ï'>X¼ø#*e§ïíøÅ«~Pž˜ð╇€/ØNé_ÙÂ9ëèwHoó ¸úg©€&þ1yàÍòwPÜéçè>¬¶–χeeœ«øtX ¾{}œËEQšµù9«ü'§Otô9j0á1ÊÇ^³íÄçÞ7‚÷ Ëtù7/-f^‚ô¼ũØoâƒAßàe¡äÕ3}‰—ðÐ>~¼ùï×â; Þr ü‰-á\0õ;úáKÂóQþrâäÅô“ÿ°BêÙMOò|p•p®PZ1|Dép;‚?mÒWKÛ•€wƒïZy8Oi#õT :"Í1IƒµÈ€ú]ú4pQ{÷úT#ô³?ÛÈ/ò‰Ã=¬óŽÀ›¶êäø°ð]êÑeE—Ü/Bù1¾VÒeÚùLý£ñ1ðwIŸ\8¶ðªqøU“ôÇÔOƒ× ®9EÝ;E\G\ÄwÖÐùºBÎ}ÄãNÞ·üFçOy‡p±êÎæàßc³½aSÈ“mˆú}7qLz¶"tò#ôï°ÿñbqI_Dü˜ò;üÀ餤–ø˜—<º²ßCùGôÁ[>/Éö£SÒÏ>9Ìõ4±ßñ qñ¶ÆÇõ'”ß[u\yD;üéâp>$~GòWU?¯žºn+"Ï‘/Uuø½5€ƒp> »´màƒè§6 ãWKtÀÒ!ÈŸ[~jÒ¡+¤/á©H.+°½¼ï$üÏð7áàôÝ:ŸØ¿Õô'šé»×Ò8D¿e ~À^ò úÓ¶ŸøoÚ*Â~V]-¿a߉þªeè?fð=•¹õQEŸ­™üByŸ|.ž…ç$ôEl˜> ñ'C|`^Žurζßá"Πž©$œ÷kaîG ~;ê÷k®ÐúÈ!úfÇé_ÊgL¼ñzðÁqú2“¬Û|öÍ(<.ΉIòÓ)ÖGùXQÐ… 7óqᥜۭܗýèð©´"öß>úïò}D¯ª~R)¸„|¸Zð ¯´›z¨œBþþƒð2ÕßÙEߤ„û¦¾Pq}²µÄe×O­Šü+pÞÄg²!–G}ÀJÉÇÕƒ¿â¥¡ã%ôÕÈç]8ŠêÇ*x9âÝ¥ŸÃZÙß)¾¯ð0ê#» üG> êÇ o•?M5óÀÊñmG¿cCœð‚“9+ðDé«Y|”!êÜÆ°ž}¸t¶H9<Ëž€C&þ>Ùð‘=E=+¾¶tzäcÂé2A–ìçŸñ=*ÙwòËÑ|ùÕ{Ðq¸x©ò9–ß…x®ò–\zF\,gÞU}ŒúóÇ©/ŠˆÓÒE“Ç:º'Û ÏtqI>éxîÃ<·ƒøôà+%ý˜5ÏuŸ«ù½üè5G?7«xŠ5rßȇ¥cÛ¹º.>@s4Ö“>ôlv„ó|ˆ~Zça=¼jí¿vò˜FÎ5åÛCàVÌ3p;FTzî×}Cxm&Nç$8Î~V__|ñÇŽmÇá;‰w?F¼éÆ?m’õ3~_ÈJÈÇááY†sl?8r+ú‹Ž§\ü]ô#Ûñ?{Š¼j€s£‘¼¤ˆ}/>(ñÈFèoî&¾Ðï³­ðʇ¸_ÂIÄO« Þ᧓œ›¬[kã¼ÑyŽ—ð±2ð:ðßµR⸀K×(¼µ¸ÏIþ^VN]®|q¤úß)øMÄ›ŠÀÿ5|ó“zZü)颊ÐåãÛ/ÿÂ"úè¬=uqàÕÛ^ög+ëc'uUKð!±n~Nƒ“”F·ñ¦ù&¾ ò…?ÀsK…úÉåóT ~®|[¾nMôÁjáEâ¯iäKè¾åƒæåáû%ñD>qâ3ï&ÑYÑ¿“ÿ¯mƯ-<¢Bò¨føqòKÒ<–]ñ6¸Ö ¨ŸŽržÓ‡ ßkýè}óÁ ñ×–?£Ws¿{bÚÑ»­Ÿ>|/ø½êWÞ7ñEË£n©Ãϳ‚>¢úö…àöÂ_øþ ä)Âï÷£?IÿÅ+Á¯Ë+ñé|†õ~„ø2DÔ£W?¼™¼¦‘Ïi¢þæó;¸¿Ò å|Œ9Šò‹ÎFÀç¹oÇYÏà» /h/ùí8uö$>ÝðÝmŒõ'Ÿ€ãø—L¯DŸZJý߉å{uð|›Y/ôôÜ‹‰ëûèç‹//h#ýñÝq;õ_Ôwµ¼¸ñî6 ¿œ<™WG0Þ¥¹v¥ä{̯Kü KàÑjÎJ~[šgZʺƇÛÓAÏk)pÎ hš|W:ûµè9å«yAâI‡_„¿«òžÎügL|¼Zõ!ðÑÛG]п'M»•t—Ê39‡¼?£ü%6 ‡àºàUüócxyš9âá¶QŸk^zi¯ƒ¯™Ð9©ù–ø°%¾äš—Æ<)ù®ú®Sqá«þ¢5¯“ïÏÃë8§á×ÚQðÿ"ò±ÍQútC© €´ã/E—ôyuÕÇýéÇ_´ÜIx'~T–w•Ÿ úJo u…üÁ×\p¸Rðï4ýqúCÖ.%_êyÏÃ×~휳^øí#àéƒøét².3ð’ñ ´fpöf|HSø1öÃÓiå:œ~Ñ øô~ú‘à†‰Ÿú#; .7@?a4<WÝ/œUsdFÑOÞ\¢§‚Ÿ»<ü(¸îuªæð Ÿ‘ž…y„ ÿL<ø:òK7ͦ.Õ\.뇬 oÑ+é¥{ëAï'ÞýœÄb?¾ ñÇ—›×þ‹1×%™_WG¼”?ºæ®‚o‘.-ßjéÊÁïŸ ø.ü¹ž¸êšÍÁO'áíÒ‘ÎWs3àvUá~k®—iþZ]B^=ý-ù#yA8‡}ë¿ =\7¾š%½˜t Ò»uRfýäãWQñÀåSŠO–â%s¹¬ $ù³É^sK•‡â’ s ¬Þ¢ôÀ%ªûÀ1«Â8£¯ëk8¯›xšŸ­y¡Û,„ïU~†ªï„×ÖÄüNÍ#&¾&8ã„ü ЭèÜj‡½“>”tÎøTZŠ~³|{8O†ˆ{š÷ Ÿ´òÍQÛODõ¢ü×T§)_þ"Ã9#ßð”"ö!>ŒÒó;¾sÎúFÀuGÀ׆Øg]Óqñ{­•ÏW>ßÀy4H¼/è´ø£¡ŸcMÜDzïÇ àÏ>“à&ÇТ~ õ‹æøf¾¿úR£¿ˆo°MÏ­ç¡nÛOþzŒu4…î ~ úà)ú ÊcZ¨{ èÿ oh!¯Ñ\æ.î³æÃáϘÌ#­g¿÷Ó?!þº|ôäÇÝMý&ý‚üf˘P-sÑþd1ós‹x~úžò þ¯9Jš÷£ý¼ž¾¬ô²{ø™>™k~œ|„ÖPJ§&Ÿ_|$­ŠúVñFütùlp;}G‡W)ž„ÓÐö|æ¼UÂGÑ|Í‚GgàhË8wÛÀÑŽ€³”á÷$ÿ2ùgÉ­]ŠôEø ‹©úZ:TõrÑM7‘¨ß-ž²p'éEñ)“‚ú;˜POß>¢oÅ/Q~râOÈ—\úýzøqâ%æÐשf‰7:I¿Vû¨ÎÄ?ù¾jÊDü2ò[ë#’9<ù:ïO×Ü]õ€§È›åÏÀü=>ZE|¬×œ%ðˆò]ùø§9çä/%¼ñYâö|"ÍGjÇ¿!E](©®Gü»&pƒü zÈ—à9Úyšæ³—°O¤[)#>D^˜u1î"Þsë’~Ô>gðÕð/µêûýœ‹ãÜ·)Ö‘tÃ%äÌí5á´âdàý´°ß÷ÏãÏ`ÍpæbX'~L•ÔÛòQV>]É>çS_i#x§æÒv‚×°¾å£'Ÿ€êòBø4äÖ¬9œœ/ÒÝç°Þáý¹æmhnþQ‰£æÍI¯‡~¢+V6‡¾ýùt.¸…øZMàÀzîø‘$¼pü¨œ>°k¾ÒŽ8m™eâ•É'WóRäß(üq7¾ò5W=Þïy„¸SÏørâ÷þéâó¥ÐÅÊGŸgÞœf}ËIúƒ"ÖKE˜i•à ò‡)`þM-x }e+æ{ì¡_ô±ò)“îÖó‚áªË¿Q¼ 4~éëá1%ýð`|lœþúûªOºè«×È7T<^éœ[X=è »YÇø‘'<Œ&øò9\óI¤ÞN-¾T >MàðTí8·tÙò%“þ{MÀ-æÌ¢‘ëI8e#}…=ðvÃO«'N7ÒÿífÎB#qYñwˆ:Fû ¿S—?«xK]ÄÃøƒáÜ–Ï‹t»eÔ1cø;MàÓ+_³òÕvæJ#_„¡>­tÉøds1­…:®^msƒäk¬ýÖòCñ}­“ó¿qÃïðÏY|N™ÒG–ÎBu½töm3úˆÔKò’ÒñUKtipÔBòõÇèk~ˆüÈK9Ÿñ»°Ãì£L¾9Ç8÷úÀ7ëññÚâ`âK8É}æùÇêa|yáZ!|Æ£ø¢L±ÿ—3á8<Ÿð£ â×Iê ~éeœs•äûÇЃ• Of·Òwh$VŸmþ¸ôÀ½è›Xªë5÷¤ Üñ¼ÏàLø¸zŠ¸zø/iâ3¹½*óœ<ž}¢‹®"Ž㟦~èÞp`[†çLÝgNßik8¥÷tù 6/*¨3 ‰_šç3î)ÿÁýì£éÙâ}É«œõ¹ —BpiÕUµøÂR¸ô—ãÔ=š«ye¼/¼`ë‡Gz„¼>ŸLóúðïMæ_+Ÿ(b¿ôÃ7ê¤_»=ð“4÷Õä÷©yÀÕîòÓo yWâ—©|­ˆþu=¸\ ¾MeàŸ=äO›Ðì u»æø2úÿõລàÉ»©wÐA¸ð<ùÜ‹w{^QºÀ'©ŸzÑã «¨¾ÿ¾ |xý|ü(¼ð ù ['u“üjñªÃG¨•ú~RG÷ª§oFœ°p‹*êÆ óZá¥knâœ3éÿ °/‡ØϪ;å¿Ø†nø øøy=ßgÞþ>ÖQ |°)új}àqiöÍð¯jt”Ä¿Ãàplj_ZßøT'úçcàSè¨À}ì8z¯ð ô®'З§Á1«ð¥ícðËöÇåʧ¬ú[¼9ôÀòw³I®[yÁy…ðcæ<ûRîóAl›9ÿŠÉ×…Ot“Ÿ ÇHSgð<’y¸Ìsá‘øº=Æ)x­ÌS?'ñÉH#w¢;jÿ1ø›ÞŸkBÇ¡ùÂw'©ƒxÞƒÔM½ÔÁò+j¥þÄg·¿šFpÔúÿûxÎâUNÑé&Χ˜óœÏþ­"žç;D½tu |gÕÛ%ðlQO‘ÇÊÏ{’ú¿¿Žãè‘'Áëá”S_Õ0õ먋ümŸp|\T/ïâ|dÿæ3¯®¼¢œHþ¯òOS§Žòü¥–Å6ÎÍAÎúã~ç^#çëvøwÄ¿„—TC_¹¶êcêFùA'ºÁ›ÕÿºrÍÕv|’¼ˆùðLóÁ“…#j^ú8Â9Z‰7-øÉ s±rÖ3y´Ó?tùÅì üGÏ^ö’ܤ¾eî¢çàß4 k3¸¹ø–¥øÈh½w¡—æºÒÔ“[ƒLÂwøsj™Ÿ½™þ`?qR:,p./E—&½¯xžðÓ?ytóÉ}’î·šú´‘ýW„OÛ|>þOÉ|žtà^ÄøüAä­ù*šï7Ò炯kÔ]Êå¯8ï\>/Ò§•€K0Fy9ÒÃäáÒaKWÐ /L¸kúfÍ!Óœ Í©©¥~èä¼o`_«ïÙÞ•âÜjÇ_m¸a }ÀIúqÔÿŠgûÙè`¬Œú? ìÇëlÜ>Ž¥ñ?F½‘až æµ¦‰ïòÛè#>¦9Žnº?D<“íó@¦à¨v}Z¼¹ øwSàk)ú)šs¨¾šêÙNù:ÓkÃG¬€þ8½MtúY/ôû4ÿIúöfüû{ák©ï]žsRGˆÇ&ßCñqÓø?j.º:¯ª›z£JsìÀ95§UóX+裨Nƒ'èॠFþ³ò—aNºm‚¡ùî£à¹š#%~µþN~“ÅÔÓòõ‚×ëµÌ–'¾¾Dd*ñ¹xÍÒ럔~…~a5:Éax†äáòkÓÜ7áÎG™'=H?£¢üA5×µKø2}*ÍAIËÏFý{ænYq¢‰üZϧYŠ<µŒy¡Â] ð;›¿¢o6 þÖE|“¿aýÐ1øйeàSR‡µÃÏh•¾‹ün˜ç^Î-~O%><×ÃôQ'9ŸåC¾G¾2è9ÇÐ1NáK"üLýèNò°qæä<’Ïc5ü â¨t/õÔmðÛˆ«È747æ8ñµôCqÃzèÇ4ÃkÝE?C>hÒã q8™?¦ù»âôó{æz¸æ‘WPo¯Ãoe¢ oBÇ$^rß4Wþ¼Kÿ.¿sxŒ^κƒ:¬‘z\ü¨ü¯¹ø1àŽ‰/¢Î pëdž(úL×½:Ö#û×·‡:A||×¼ËsÄc8D¾„/ƒoÁ‡´¾´|­úX—GÑÕ£§`þod^ ¸‡Kÿ,þS:€æwjž2>>EÌÑQŸ°Þ€øhµàQÊ÷˜w`EÌE–~RzxéεÔfþlâêþ_}À½ÿ„td™ŸÈ§Ñ«À5ô?‹ß#¿¯Nt!CørÊïyPV Þ;Àz>ÈzÐ|¥\êÙNúÂí¤N…ûì̱¶mœ«å|ÏNð´6pä\pòø«iê³fúòà(¾7à÷sÎÀodˆ¸?‚þ² ü\çÕt&GÉ:ÈÇ™waà‹íôUÓÔðlØg5¡ŽOæ™5£¿gÏQòÏ)æÑ'qŸóBõÄ(8Íu~Y6‰/W'ýÖ£ôýN’·1GÏjÃ}tù?(^4rŸ[xÝ~ø¯ÒoÊŸb’û"^|¿sŸyTÉ\ªÃÄû#ü»ø¿âsÈ'÷x™üŠÑá4ƒ©?XÎ/?Ïnt…ä‰?Y ü ÍYƯ\sm#ó˜wê|o/ÄÇF>Õòk+¡/-s^ü «ƒŸ5ÂËoC<ñË4ï¼;™;Ãü ÃÏÉ[þæiü¤†ÐÅI£ºL} Í-éCß<Î!àê6ñÏðëò xøõâ vàç¡y½òUÔœCùRUÓÀÝ[‚>$Á¥«*Ä7^‡k¾‘^Çù¿]D}ͳhCg[Å<õWbΠæ(u…óÞóÉˈ3s΄7&Q'ý»6ž?8¦iNzuy)zónpƃÌ-: _~=ú€¬Ë®Àßsx½òC“.Ì·rþsÝz.ÖE¿ª•¾ý6p„^êßzžg }RñÂ5ø<{é9_:ˆÇâãíCÿ}žá>¾/|fë/l%ßJ|Š~ò[tV®ù@M¬³1êÂ1ú ÇCèòí…÷hÌ=±£ä—“äÕòߟ$îgý³ÞOÂÛ¿þµ^®]Á¹ßŠ?o#uºHͽ4òˆDg¯y™=äemÔ¡â]jݤ©—†BÌ÷b=&óX WÔ\;æÿ$ýVñ+ùŸðüáÝiô®òáÍçüd^†çpŸá=¸|Ä_…é¥è™ƒ‘øRJ/qˆ8 OÉ5×° E~š­óZóèÊ©ñY^o›Ã÷wô‰.^wšûYÂz¾náòÉí!_ã¼W£9ªÒAì næ¡{íÛHW$8X~}Z¯eøžà_$´œ_'ßOú?”Ž¿æ-$ø¸§¯Àžyƒ¾ ò4OÓ­%/ÜÀú¬GÇ'Ô¿Ó ºkt´‰,ó ½Hº?øôkÿ]æ¦X>d%x“æ5ü„ï6H¯¿>| “y¾£Ô]MðØZÀ_Ä—Ñ}ßx'..p;¤¿ûÂ÷ÃW€7×n.’檮œÀŸeœós‚} Ÿ7™o2HÊG_ÞBœK‡øâÉœ®§žó[>5=ôcñ¯Hü•×nãû²nÊè«*?“î¬ \ #¼ùé‹ÏS®\Ë÷QÞ[ ¿p7sòäO0Î\Í“éo?€¿o¼±üAyüa¿ŸûU„•|çÊ/~B#xøsÏ4GDúÀÖ¹æet£{ƒ×ÓDßaóéסÎîøbõsî RÇuâßT&ò›ÕøUVÁ§® _Ø®9÷àÑšwº1ô[qpXÿ8¤|ÈkÀ§˜æšóEßÒÉO¿ æZ3|«=ð\å§ÝÊó)g?ξh ÿçqUþýšwS nÓ‡ÿÎIú]ãœóãàó©0ç@:?ëÇOL¾7ൡßæùìÃJpåzðŒp$Õ—5àâéw wѹ)¾îíô‘û9o)®‘oÉFýöNx¸Osîw¿JS÷ŽP÷t¡꣟ý4ûVý|Ö=ϯ9ì#;N=¿È tõÒ½áói9|¯qøË|/á'Á9»©s¦È'Çñùi#Ïn ÷Qü{?Îsï!öÒí§ ¾¢ü¬FåË ~ >ð%æ`û’}ßcÔ1â—i¿p¾RÏmãü†¹Ù&a†¼\øÄ€|ë‰êSÓ×k“¿7}Hü¨l=óuCÿÒÀkñvù²iŠêò:›`=¬$/í¤ß’‚ßQ>@´§àŸVqNá•ðä•'§Á·¤¿Ý§þ³úÕMôñÐë[#õ¼â^šþp\s3N°ßÀûÊð„§)=¿xÎòý•_â‰h[®¶1ðœQö¯æ%KOÑÏó*Æ'³óþ½ú)šg`õàpòÝê!Ž‰g#Ÿë&ÖQ'û\ø±øÁ䵇À‘q®t±káíwRgd?jþG†|l$ø"['up/ùÿ¸[3ý@Í#ï ¿¶¾ø xÀ}±)ú—¹øÈÏŽ¹b6N}0þ§yÈ'©³»Á»¦ˆs“ä·ûøÍèâ…‡OÀ³é‡ÐßÃÃH|àµ~FÁ3ñÙµCÔã½<¿Zô:â5ˆ‡:î¨9p‰/ø̘x`àYõŠ¿ðSv’'×±Ž„¯ ·UÓ”sîÑw/Xs:«ˆÇðÞãæÕðÑ5Ôù<ùvI?sŒý®9<½ÔSàŠèš‰S;Â~ÐœT¯$.ˆ×Î<âd¾8óa¼‰¼á8y‚ø9{8¯„ÃÉ·¤þÌ!ø‡êû Ç_Œ/…æ â«”äuÝø³¤ÑCIÏ.üRs¹[È—ÊáíÔ³®å[(¿ÏÕð—Já·Ê‡OsÎðewÍoŸb s¾¤«åûºô‡¥ôÁšðÝÐ\bü½‘ýTŒŽjŠz­Þ-~;ÉüÇ4ºP|Ö¬J> ৘7{ˆ¸3D^X®9\¬ë^ødòõÓ|ùÈÉ—MýÕâR¿øéð|Ô÷:¿Iº)Öƒ3Ç.É+…«¤_w¾êÅ&ð¨.ðžg¨ƒ{Ð}´‘c ž÷ƒkª0Î÷l$~j®Á~t4OS?¡ =I?A¸€pqÅó£ÔKSà¸ÒŸL¢ÇïA7þq ¾õ˜5†ß'u,¼„Ä_¹‹¾,}äïUD¾{˜>ójlˆó©W|uü8WÒ·ÜOÝ= þ£ù«øû¹ô¿ÃÄeæ¶zyÅ~öU õH#<‚Ôêßà‡Ô›š[›AG²Œº`1óH¶|ÝËð‘.ʳ7hݹð±máœut6WóQÚÉ·ã7%¢ê@Íù,ƒ·ÁQ̓qæk. 7 „8ŸT<6Âÿà—ÙçÎ$úd|Þ„æ´ËoÍ…oõ¿3ÔËGxîâ—2ÈwásÞÀógÞ”3ÏÞë˜wìáÜÔ|(+ Ÿ¥8·ª|ûü{àâç/þ0^æ¸:þ9~|Rxhê›;¾ÞòáHæfx2ÌG´nê"ñDàcï—þç$û| Rš}ˆÿy¢‹8Åú§šâ¼¾‡¼Y>=¡C9*müþD½%@7ûû8>Aà¦òëoDŸƒŸ{~ZyY'<Ì.ø%òs+¿>DŸ­€85~ÙÇ|Á]ô)·Pg#ž|‡< ]g‚²¿ CÞãÒÅts^âCcÕœ;»È{ä«+¾!~1Fbõ¬Wôï ¾ ¯Ñˈ_øŒx:Ì‘ñæÀßqtS‰ßªêËãøOÊ/€~ºknV ¼¾ è'à§z~Ƚœù+èê¼\“6‹gøüAúý½ðÛ—kÞ.ç·t¦ªwá­§¿[ðk‡ŸéògÌ x~² 7S~Þî)¿*p©Ä׫ŒyðõLþO胥ÿOúÅèXè£ÀuÖ7ÐGÃÌ5‡ú&æe£‡rxò¾|/…þ<Ž”—ÂïÒœ¨gÁ*鯡§”O—gèwæ¿IGÑù|¼x”ºà(ûFqG÷aû™Gý$\¿œ¾¹â´tè•àÜ·~t\iêÄ#à•­ôÇÛÃû&|•ÎïÃø;ŒÐߎÄ>/°nx'Ñu×5W}ÿâvâMýì“ôóý‹ÀÙ„t±~&9gÇé Ÿ`ÝŠ·´•}CßÏŽq~N¡£i ü¯‡<ú8çáçúÂMðxįof¿vÒ×ìâ:{ñ?“Ox3£ä³ôGÈ+´þÓè—€óaŸŽ÷/#Õ°Ž4÷eœóË{ÁM{ðca¾œµâﶅ}>̹¢9S—U£ÛjDwPÅœXúèþ§°/¹nÞî»Wá_˜?f¾¯•ð<4‡þ}{ù<»…ó]¸A }1ÍóU^ŠŽÓ•ÇÔ0Ƕøyþ0ç*ü¸d]ÔïìÆߣ…qp5év—SÏÉÏûwÄõò˜nôfXšC ùaÍô+ä+º~ þ‚VO*}Öð/—ʈ?Ù08“úG%ôÙ9¼Vó•‡ï#®È×]ó¥ï¯d÷?ì9gN=vŠ¹ò9¤¯,ßÞ¼¸­1ÛÒ!Žx;z¥)Å ðÅQò‘rò}éu{áLj„ŽËŠ¤›¿—½’ý?¶Ÿû)½ñ!ð¬ztýûé/ƒëj®«â|Žh%SKlÜ° }úé’8#ÿÏ&ž_?<ƒ6pÄ^p¢)Öó0ù s«]>!ä+ãèœÆ‰óSÔéÌ­7ôN‰_„x½'зˆo?E½.¿¤ úÇ'é+VC⃔ø°7)ï?9€.¿‡óµ„þ°ôšòD¯døgôÓ3ÏN~îZ_£ðYÄŸÅ_.Ñ»yâÒ%v“ÈçT:Í‹ÇïÐjè‹©Þ”üÜ àÙáO/? /'~ÖÂëóJp{å“òc~«…ÂÔ¥o#î—ÂwªÇl|Lé‡å?’ õ út¾ =¯ædˆ·â]£³‘îÁáí€Ãùø÷áÿrŒ¾‡æVà¯è¿€·îÌ¥³èáüÎpží#ÿï% ¿ZþïyaNx2G9âg8/Àuä;í!_HômÒ¯¥Ákˆg{áSlƒ÷ ?kù6êy§yÎMð¿Ê©G*˜ãQ‡Ï+u½Mq^Vâß$ßÀ^êÕøA ßrΛ¤þÒù9~Dÿ:Ñ/dÿKÑNYK°¼¨¾Ÿæ€„Ÿ|ç&è°^ïBçK#Ÿ—&/ì€Ï0 ÏHýíq^ßÏ|(¼~Î,ê×Ä—›û=Jþ1}üí$qõ(|Š6pBÍÙm º½9g–C'õ ¸_õ':â„7<aŠ¸Ÿ:k; ?æußxØ„øgÔ»-Ô¯åÌ?"/máÜ:ˆàx~ øÉúº#ðEðÁ±QîÓ~Qíà~rŽ‰ÿ5ɾAÇjíì‡2âöqúõô!„×ʇ©8{¼Gº…Cš—Ǽ¯Zåñàk]|~šû× n^Ëü8§/²\b€y/ò;Ïp•ò—^¼Iϧ¾ã†óžðCn<Á¿>nQÎz®áÜn½+ð¼ŠsG8¨æß=Êß÷Ò?Ñù4 é8û³•}%Ü_< |?Àd¾L?}Ô1êŽúm;¨ë¥û•/S}Gñò„kÊ? |z+éê{®c}Ž‚[ÈG_çÀÏÁÄ«•~ þ»]bkÀõ½’<5Ãúz–ó¢—×i úAo¥¾¨¦^¨'9Źñ,s¦8—NRß7ž—÷F¸äÕmô}ª¨c4/~?ùR3þãY·Ò¹7ñ½Sv‚‡÷P¿7²¾êù>¬×QxΣÔÇ8×;áU †ø8g×9@ØÉüŠQð¥~pÑÐçšsÆYï sb,^—ø&ïçÃœ3Ñ%Np=OÛ¨†&¼´ üœþïϧÿ·à³¨‡¦˜'?E?¯:¨‰<‚ºßQ_ôs^dߤÞ+„w õ8Lœ’Nj ½QúNmì3ÍuïÕœ)øòUØîZÊ\‰IúI•á:\¼·nžo+8£æ†Õ¡£=F߬¾HZxxB¸‰øõòK¯Ÿ¹í¡ ÿu¯£Þ@7`õ༥ôyʨ›&ɇ¥wî#® /j!¯”îIúFéå7_ƒo1óR¬ˆü­2ԋ‹ìûBó ñîÆïqÖÕý°ãä!­àJÒ kýÉç&…Oä8ÏAÍ™ä>3Ÿ²¾êáð3åWÇ\¿j|s¿sê‡jêü4ßÊsÉÛÕº¼n;üÝû¨W‹ðÁ”?pxqñ0žx5ñ¥‹yS øŒ¶Á'h†ÿ>1‚C†ø9®Á"ë7’aýËŸ»—ø}„úr>ÑQòÍNò¼€ÇÌ9“¸1@¿µƒ}}˜z©üj4àsÎ$O9FüU¾,]o8眱IöÍ3ôMÊÏÑ…7‹WzŒø>I=P¿ã:É6òÍIÖÁ$ë³ <¿ ÿ8üÓ}œ¨¾àA|ÅF¾uôKɼ‘a|*ñ…eŸŽ’oµ‡çåZ‡­à6“Ô{ØWÂcÖÓ>Îý¬ÆŸXñ²—Ïm'êÆ?­Œç9Äüïzúèû’~û®K~Å­Ì”#s2¾°7§Ä7‘ß‹öþ46u}™^úÝòÖŸ`:ùGªþ*$Îé9o¢oÎ9ïµô˜ŸbÇ9×RìCñâºðoÈ¡ÞRÝ?Þ؈Îf:mü¾m÷A}ªÎñCðÊúXWøϺøwòEhßÏ·€“d85§ó~ÎùÇUQ÷×£Uß?¯£Ï‚.Mþð®ùNâµIÏ ~}8áâ1ïãšã•_q ¼¤‡õ(>íþòмê4ü‡‰oΙõùðûIøZSðjá-&ó¸‡ÁªÐõ¶Á“¨¿ú vúŸòS¦®“E#}€1ÖU3¼à>̓§#ܸ‹û6Êú¥?3FŸãùá@¸sΤßO½|û0붼üHèe_æ+¢{ªÄÿQëÂé9Î-ÍŸyšë®äþHÑÌù7ŽÏú$xxšux_š}ðÅ„ÃMÁŸo‡WÒê«—=†^ _;;H\?òXÏ s!lý­ƒÔÍåôAÇÀF©ãÛÀCås¤¾Ã>}ùäÍ-à‡…Òyq®Ô¡[W=ÙMü…_jÜ'ùÿ‡ï ™gNß¡“ç‚'ÝLUùÊÂk Áëà}X'ÿ~}>úíLjkÍèi5—@:TùÎq¿ÝâÖ}7ü bîŸæ­nC×ÚI\¬ o!ÞîÇçe <>ñJ4'}8ryö.Ö©Ö¹tš&Áµà~eôSìÇtâ‰3GÜSâ#‘Ç=MýØŽ¯:aœ}&^aønówFBÜœs<¯Ip¤ãø!ÃÓvù‚ ?iÎA#<îòP¿¹æÑ?OpúƒðkR'5³¾1'¥‰<£›}ZKÜ,†—ÝÂúE8ŠÞm ^xëŸ9³ˆ'ÃèH»À­ŽP7õá·rå(8Ü êôvðÈöwW¨?çœ ^>NãYpß-øn?ˆO¹|D¢w;oa1øêyÁpª ú¹§Ø/º¯-â³±OÐïàþ ?†É„ßoߎø¦9@£ôcFéç´“/?ÞÛÍ|È úYÌ#³Vú–ÂõǨ£˜sîîG¸d3ýì q¥˜úçóRœoðŸÌá¹ >”^ªœx¿5àÏ^xäÎœF—^ôÛª›ñ©’ŽxõÜ8çÁÃð#z¸^ñ÷Ñ‹_œä!ÌIð6ú ò§NŽ¤§CÿÍ«À7¦Ðj~+óY­›>.úL;‚ú øHiæ"€¯%¾ß¬+g.­ ƒcÿì(ÿO|ßø¨š£ãðÌÝä);ð­'pÍ}¯ã9à/bøÛø¸dëå æ;Á»¸^ÄUäûÌ=wâeRw¢«°²p®¹ôB•ôŸ…ÿÞ.!^ò u}üöbò |wí~ O³/'ÁÍOÑ?MM÷5²!p‡bxÉèÏöiî4þ‡itúé¤nn&ÞhâûéœI‘÷Ëßy:‚Qx:£ôAŽ¢“é÷¼È9³V_þaž“|…Ž_ûÁ‡ŽÐwxŸ IòÀv|)ñ/¶°>æœÉûj>ð3ä½Ìa”A=qé(yÐ|ârpŒ)òŽýà½Ç‰/§ÈÃÑáZ+¼(ùŸu¡ë@_0 Šù*‰ß·ô ÃÔ]šs7 ?qŒóWxìÁ‡šåO Ž·œú¡{øÁQâkÎvæ:Y¿pÎñnðærxÿ£àšÛ _ÛJպċã<­‚ßÌ\c/¦nÒ<§|6ÄûŸ^ñï5mŸiùDtó¹ê×fðíÑ\ø¤¯×gøʺòžrø iúzª+&èC‹''D|1æðÉÇÄN C¯}ßI¿uzNø®y¥ä!Ö©ó ¿0é|…‡®æ{ˆß/½7s‚]ójŸâœ/E>Ì{ò"þ¾Ÿ„%Ô÷òÃ~½(þךƒâࢮ¹Êòm_º}|’§“¿‹Nh’ë?…¥˜üQ|&ù?Jǧ:Ró`O7ì'/œä{.\¾G ÏQ>¶½øþ‰ç4Î1ßN%?¦æ*GO;Âó£êw»Àñ'ˆOðA­ƒ<·šÏGG[Ž¿x¸õI>½þŽô"cè?¥ËÛWGÿdÏYóæ„Kçá« ùžåÔyô·’Gï Þ._ØcøÔÝ,Ÿ\ü¾ÙÿÖ Î)ü4¯Wû]y™ü㞥ßÒ ¾¬¿?ŠŽ7|VsÀ»­;øGΙE0Aœ<ŽNJ}'õFØoÄÍ&öS¸~MÖD=˜áÜ8^p˜¸ÖŒžá(}Óvp¨òÝþõðq{ðm%/åsŽÀßÙGÞ1ú|sΤŸ<o¸‹}tX:t£wŸsF}ÀÙÇé6£¿y”ðTà/z†8sŒ¸±è^pʦw%¾z‡¸SðÖñ³Jø×è*¥·aÖü†©·÷2÷­…:‰ó8™Ÿ†÷¶>á€?ûíôA™“é™Ðs͵—OT#}¿pÕ|xÜ:÷Ö2µ"ô-\þouè‚3¢3óžC |ÞÖÍsô­zÐ5”ÀSâ:4¸„û‘&HÏø¼ÕÌ-:ãŒÌʦÓgžqÆ+ÎÎþ÷Y×ο!û_/Íþ_ôÏfÿoöŒ—üCö%7νaÞMÙÿýj^÷ ^ýþ_ºõú7Í;÷†y‹®[pË­so˜ã¼sÏ›ý…+¯z¯xùn]È?_tÚ?¿iѹ7ÞtþUóoœûÝù7οæܹ×Ì¿öÜó.<í%ç}ẅóèwç_ðžÙç_?÷šE ¢·Š_÷žsÛðús¦½~úUœsÕÍ×\óW>眫®›ÿù7<ï߽檹×Ï¿æÖëŸ÷ßøùùWÿí/ÿúÏÞºpÁ·®¿yÑÜž÷^ùÙ›-œwÓ5ñ‡¯ýìu núîu &vúÍûôgÞ8Ñu¾©<÷ò…ßÊ~½ì?üðCŸýqò¯§½nöó¾nÚÅ~æšEs¯wÍÜ$ŸzÁi¿Ë ·.8ÿ7.¸iþ¢ןÿÝøç.ʾõ¥§½èÕ§¿èÜó.9ýﯜwýü«çÿå{ÿéëâʹ×/úÛ/yå•s¯ÞÛòò+ÎÿböWÉúÐë¯8?»³W4s ¾#û‹°d’?ùà¹ÓþíôWgsÕ¼…Ó_ª8ýu¯¸âüÏ.\ð——ñÚì¿_7ïÆ[¯Ÿ{ýÜÿâ·¯¾âü+³ßxþ‚¿üUö ¯˜wó5Ï÷'Ÿ¸iÁõåW_pCö6þååe¿ôÇ‚'üóó^æ?g¿ù57/üsD™¾$O¿ºwF/ÍÞîó/9m_Lùé;áM¼üo¼ãk?67»äÒNƒW†ßο1ú>—~k?zó s>8ý蛯Yôÿ‘oÖëÿÖUþ×Â[]·ð¯ÝW†ß†Hpú/.úȼën½vႹ7]³`áÕúÛìQ3oaö±ž÷îwe×Ç鿉b•>ñÃó³Aÿ†¿ÞpÍü¿}¦¼>zÁå×d/ø¯,Í«ç-šûçu5íWšwãm%2^¾pîÕÙ‡ñ7Ï‚KfŸyö_\}ó¢[g<¡žû¼¿;ýÓßyáùŸš·(œ% ŸAžþ=_ø•O_þõ ¿vúOMûiö´Ÿ.ÓOÿÝ…Ó¼hú³§ÿxñô/™þã¥Ó|ÿô§îìé/ž=ýų§¿øâ÷Mÿñ‚é?Nÿ Oÿ Oÿ Oÿ Oÿ O¿ª‹§_ÕÅÓ¯ê’éWuÉô«ºdú[]6ý·—M¿æ˦_óeÓ¯ù²é×|Ùôk¾lÆM¿æË’kþ‡èÇ Þ÷¾?_0ãç gü|ÑŒŸgÏøùâ?_2ãçKgüüþ?ϸ¾ f\ß3Þÿ’™?Ïø¼Kf|Þ%3>ï’ŸwéŒÏ»tÆý¸tÆý¸túý¸ðâéïwá%ï›ñó3~¾pÆÏ3Þï’Ù3~¾xÆÏ—ÌøùÒ?¿ÆÏ3®oÆ÷½pÆ÷½pÆ÷½pæ÷½tÆõ]:ãú.q}—θ¾Kg\ߥ3®ïý3®ïý3®ïýÓ?ÿ¢?¯·×}å² gðÜ/,œwÝܸN ‰¤~ýú¯\vÁÿk»ìYb0<ôã´Ð–‚é”B®X:Ý]BÖݺtx*I ††ÄCÓ_ª³ã;‰d0èAÒûÚ7X:NŸ Ó1ÿPÍ

Në-_cú?ü0Ç’§¡äµYózh×iÓPb·U˜ÏEšªZç± –úæÌÞ4ÝUÙŒ÷èú·DdÁ}Jöˆ÷’<ÿœþdOå¢8¹wä·ôý‘œ_o8‹Ü­öÛ±õ£rž[á÷ÿ不֬x£­§˜IWï:\}·9´û¦=´Ï&˜íŽ·5Óšɻ’ß?\%ŽÙá|Q…­ùüª;=Cce¼™Ò±ãbYnŽá‡/ %X $ÁIðI²0Ha …A ƒ†,ÜCH.”]Hºt!éBÒ…²Ëåp»ŽqRYE¤È)òŠBrCÊnHÒ Iº!I7$é†$Ý"ù¤O³¬Ï±Ð„Š¬"RäyEùkæ>»q,Ü"¡"«ˆ9E^Qtû‹Ï?zNP«# pcaMethods/inst/0000755000175200017520000000000014710312420014632 5ustar00biocbuildbiocbuildpcaMethods/inst/CITATION0000644000175200017520000000164614710217306016005 0ustar00biocbuildbiocbuildcitHeader("The pcaMethods package implement algorithms found in several different publication. Refer to function documentation for reference to the original articles.") citEntry(entry="article", title = "pcaMethods -- a Bioconductor package providing PCA methods for incomplete data", author = "Wolfram Stacklies and Henning Redestig and Matthias Scholz and Dirk Walther and Joachim Selbig", journal = "Bioinformatics", year = 2007, pages ="1164--1167", volume = "23", textVersion = paste("Stacklies, W., Redestig, H., Scholz, M., Walther, D. and Selbig, J. ", "pcaMethods -- a Bioconductor package providing PCA methods for incomplete", "data. Bioinformatics, 2007, 23, 1164-1167") ) citFooter("This free open-source software implements academic research by the authors and co-workers. If you use it, please support the project by citing the appropriate journal articles.") pcaMethods/inst/doc/0000755000175200017520000000000014710312420015377 5ustar00biocbuildbiocbuildpcaMethods/inst/doc/missingValues.R0000644000175200017520000000327414710312270020364 0ustar00biocbuildbiocbuild### R code from vignette source 'missingValues.Rnw' ################################################### ### code chunk number 1: missingValues.Rnw:43-44 ################################################### library(pcaMethods) ################################################### ### code chunk number 2: missingValues.Rnw:46-49 ################################################### data(metaboliteData) mD <- metaboliteData sum(is.na(mD)) ################################################### ### code chunk number 3: missingValues.Rnw:52-54 ################################################### pc <- pca(mD, nPcs=3, method="ppca") imputed <- completeObs(pc) ################################################### ### code chunk number 4: missingValues.Rnw:58-61 ################################################### data(metaboliteDataComplete) mdComp <- metaboliteDataComplete sum((mdComp[is.na(mD)] - imputed[is.na(mD)])^2) / sum(mdComp[is.na(mD)]^2) ################################################### ### code chunk number 5: missingValues.Rnw:64-66 ################################################### imputedNipals <- completeObs(pca(mD, nPcs=3, method="nipals")) sum((mdComp[is.na(mD)] - imputedNipals[is.na(mD)])^2) / sum(mdComp[is.na(mD)]^2) ################################################### ### code chunk number 6: missingValues.Rnw:71-80 ################################################### library(Biobase) data(sample.ExpressionSet) exSet <- sample.ExpressionSet exSetNa <- exSet exprs(exSetNa)[sample(13000, 200)] <- NA lost <- is.na(exprs(exSetNa)) pc <- pca(exSetNa, nPcs=2, method="ppca") impExSet <- asExprSet(pc, exSetNa) sum((exprs(exSet)[lost] - exprs(impExSet)[lost])^2) / sum(exprs(exSet)[lost]^2) pcaMethods/inst/doc/missingValues.Rnw0000644000175200017520000000565414710217306020741 0ustar00biocbuildbiocbuild\documentclass[a4paper]{article} %\VignetteIndexEntry{Missing value imputation} \usepackage{hyperref} \title{Imputing missing values using the pcaMethods package} \author{Wolfram Stacklies and Henning Redestig\\ CAS-MPG Partner Institute for Computational Biology (PICB)\\ Shanghai, P.R. China \\ and\\ Max Planck Institute for Molecular Plant Physiology\\ Potsdam, Germany\\ \url{http://bioinformatics.mpimp-golm.mpg.de/} } \date{\today} \begin{document} \setkeys{Gin}{width=1.0\textwidth} @ \maketitle \section{Missing value imputation} One application for missing value robust principal component analysis is that it effectively can be used to impute the missing values and thus obtain an estimated complete data set. The pcaMethods package was partly written with this application in mind. PCA is a way of creating a model of a matrix, $X$, by defining two parameter matrices, the scores, $T$, and the loadings, $P$, which together have less values than the original matrix but when multiplied with each other well reconstruct the original matrix. I.e.: $$X=1\times{}\bar{x} + TP' + E$$ where $E$ is the error matrix and $1\times{}\bar{x}$ denotes the original variable averages. Now if $X$ contains missing values but we still are able to get complete estimates of $P$ and $T$ than we can use: $$\hat{X}=1\times{}\bar{x} + TP'$$ as an estimate for $x_{i,j}$ if $x_{i,j}$ is missing. This is can be done as the following example illustrates. First we attach the metabolite data set with missing values. <>= library(pcaMethods) @ <<>>= data(metaboliteData) mD <- metaboliteData sum(is.na(mD)) @ Now we get the estimated data set by using PPCA and three principal components. <<>>= pc <- pca(mD, nPcs=3, method="ppca") imputed <- completeObs(pc) @ If we compare with the original values we see that the error is rather low. <<>>= data(metaboliteDataComplete) mdComp <- metaboliteDataComplete sum((mdComp[is.na(mD)] - imputed[is.na(mD)])^2) / sum(mdComp[is.na(mD)]^2) @ When using a different PCA algorithm, we get different performance. <<>>= imputedNipals <- completeObs(pca(mD, nPcs=3, method="nipals")) sum((mdComp[is.na(mD)] - imputedNipals[is.na(mD)])^2) / sum(mdComp[is.na(mD)]^2) @ If the data we are interested in was gene expression set of class 'ExpressionSet' we could simply do <<>>= library(Biobase) data(sample.ExpressionSet) exSet <- sample.ExpressionSet exSetNa <- exSet exprs(exSetNa)[sample(13000, 200)] <- NA lost <- is.na(exprs(exSetNa)) pc <- pca(exSetNa, nPcs=2, method="ppca") impExSet <- asExprSet(pc, exSetNa) sum((exprs(exSet)[lost] - exprs(impExSet)[lost])^2) / sum(exprs(exSet)[lost]^2) @ Different results will be obtained with different PCA algorithms. Which one to use depends on the general structure of the data set and the imputation performance can be estimated by cross-validation. Please see the 'introduction' vignette on further details on how to use the cross-validation capabilities of this package. \end{document} pcaMethods/inst/doc/missingValues.pdf0000644000175200017520000054242114710312270020736 0ustar00biocbuildbiocbuild%PDF-1.5 %ÐÔÅØ 10 0 obj << /Length 1833 /Filter /FlateDecode >> stream xÚ½XYÛ6~ϯð£ŒÆ /Idz¹Ú¦À¶‹dHR€–¹¶Y2$9›ý÷áP²ì›¢@ûbÎÅoêùÍ£'?*³9“"Õ‹›Û…Pœi•/²L3iÅâf³x—¼ÞŽ}Yo—+%T²/»n\|Z¦2qÕÑw´>žþêwž&‡åJfIá®|¿kÂ|Ó-?Üü 3"]¬¤`VvpÅH>Æ[„ ¤] IS‰J®òœqcáœ7ñÜïK#“¦ºmÝÄÊ/›ªÙ.ErOë÷<åׯ_<‡QK¸…»8—þvçêíΕùF$ì ‹BveíÂQ2£²x4§£hp°JÈŒ¥6?g|å>G–•« 4ë#­Ï,ÊO]5•/Ž•kOÇðT—;\ÜwÁÆû ×d,»4'ø¯é»ÛG‹~òíž8ÝÓuë‹ëVÌ<Ï,Ûõýáé“'ë²)kð'¨¸OÛÊýaµ\i‘l›jë-Ûø's8Ê-K•)¦ò”øþVôÍ@š'>Z)ícrƒäR›T,„fJg¤Ÿ fjŒ¢‚b¹‚óäj —,ƒpÉD\æI9,¨£†ù˜É 9óFب[çŒIÜáP•E@nØxCðÏ)8aÁ™Ò`Ù6ëc×ÓüЖuQ.á|P F7µ¯ráB;ZÑh!²]®ô#w’~e8%ÿž+勾ü„ñç«{Ú.\ÐV&Ñ¿´{ì`"’ ­ú†Æà™HA‰&ãÄɸŽÖÛ¥ãv³î]Y$Q¿n5§~_„|ä‚©ü ÂÆõg<é|‘—š<¹ÔBÏ™I~3˜ßqÈeUirY ¿Cõ\TüÓ»¶ì{u¼+ÁŽ9%wᔺÒI0Sa‚®7Œ²‰†€Ñç ®_ö´y· `†­=Ê?V} ˆ {àλAu\5¤w Ác¾ªhÕú¢ºÔ‹~>0<Á¢EÔ *&:*M^3ÏžÎx]@¤ÂºoYž‰3tJ3%„„ÌÕàþï1•™DDŽzJ×$Ä@øžs9ǃ·³‘Ìft[q–©Qäç69³|„Ä7sÀÅÙ ·PJí-µÈ å>õŒa²…fB("æ3âRÆ9¸jBõU™¯fp ÅM°T«ètI”›ÖÏe ©yˆ^ÑEÈ“6À°XÐÛP&a:âÕæCôÅÅ@äÇHž»àœeP™¿r¿‚qùo®÷"#f£&_7½¿´mz›Ç nK·®âÿ!±Sh¡u,F{ž²TªóHúµ¡d ‰—'åíŒÆðH°©~(‡#?9Ú‹,±wÄv,帘–r\‡ ‚“»˜²p)fœº6î‘}8þGH{±jayÀ¾j“XÊ‘[}¬ó1ßiÉLžŸ; ™³”Ö¥e¼k;b„Õ9ËTf* ˜äxÛ1óBZ£dHóbØ„æéé ꌆ¦oª‘ŽúçòÀø4Ç!zÌÿ)å¥ÿOÊSÿ4å¥!åé³”w‘«8¾etx-˜áµà†^,¶`¸h»ò™ú. Eø"ÜùD£”e»6¦mÔ¨üö¯9¨H–§)Ýí¬#Ä%®:ÎÍCRBa.’Ê õ­~à`wÜ£¼²c5 ~‰’i`ª¼0uRð߉ÄD ;ÿžZ8V'E?]% !õã„P‚t§'>Vè¦`Ö ¤kúèž÷øé 1¿ ]‡‡ůFȸõ‘õôýŒ?z?0f_–áQ çcÙVüîeO߯‚£½ºyô7˜gïO endstream endobj 28 0 obj << /Length 1056 /Filter /FlateDecode >> stream xÚVßoÛ8 ~ï_ì¥P»–ü#1°Эp÷Ð+°÷m€+‰pŽXÊÚþ÷#E9µ_Òë‹MIä'ŠüDêóãÅõ×,Q‘ó|ô¸1žFÓ˜&q1^ŒËÑ,øcf< ¶ ú ñŸÂX|³xs{EóõÃBJÜ`#ͺ)?}Ø‚ÖPcãâQÈXTd™¬6Û‘¥¾h6ÛJù÷\ã.ÛErý5™ô> stream xÚwT“[Ó.‚ Ui"ˆ€tBï½÷Þ¤† ”$$‘Ò;JïMºté J¤*¤HQ@zçGç;ÿùî]ëÞõ®•wÏÌ3³gö<³×æ‡:ú\²¶pˆ†æâåŠä55Uy ŸäÃgf6€¢!ëñ™ Hû_y$„¾Ñ)€Ð7@M8  öÔÀËàã|@ èß@8R  rƒÚ4¹jp…Ï,Gx"¡öè›}þ^XÁl^QQaÎßîY Áš ´ÄåfG0È C!hÏ…`•p@£b<<îîîÜ 7i/ÅÆ p‡¢zé±ü* rü)Ÿ`àEýeЇۡÝAHàFá C`¨—§0[p³;@_U €Àþküàü9/7ïÂýñþ ûí ƒá.Ì ³ØA!m% n´š‚Ùþ‚œQðê ²¹üNP’Õ€n*üS Œ„"Ð(nÔùW<¿Âܳ"ÌVîâ¡Qø¿òS€"!à›s÷äùÓ\'Üæý·d…ÙÚý*Ãö)‚Çu} QUøƒ¹Qáÿ£³‡ ‚@a~AÄñ;ðüÚÀÀùmäý¥¾©Á×GìnÊ€øBí 7/|oÈ @#ŸB|½ÿ·áß>//À Fl öPþ?ÑoÔ»¿ä›þ#¡3à ýxÀ_ÏV7 ³…Ãœ=ÿÿn1¡Žœ±¾ÇŸ’ÿc”“ƒ{¼¹øù\|‚@/P@ |³ðýwôOÿËWfü ö+ß›ƒú;g·?$`ý3!l€Ó‚ßP`ý‡éæ@A øæ‡÷ÿ›ï¿]þo4ÿåÿÉôÿÎH驳óo;ë_€ÿÃr:{þAÜP÷)úf 4á7Ãûo¨1ä¯ÙÕ„ØBŸºü·U ºY˜ý ¥¹x¸é¡(%¨ÄVŠ;üE›¿ô†¿Î ƒèÀQÐ_WÌø_¶›);Ý\#¨nþe¡nFý»‘¿dÈÍPý;Enûkúø… $ä‰ÓüIàÍ{3¦¶ßìðpÃàèÀM;;8ÿW£…D<Š¿T¿% €GçIÀcð$à1ùÄËwõø-þ+ðS$ò&áßDºIöoù÷Õx@ÀøS“p°x°cMpëIµ,;×êöNê‰ ×P%.º[qÔj)^?+sZ½Tiª‹WÉÒ±]KÎõ${þó¾÷J-C§ÀƒÒº=ƒMìäõÑ­ñïcZÆÉ{M/“åèÅŠ‘Ý:tQÄxÒ¶öÝúÌ÷|×t~Ipg!/QV iÕjë¨.ÉÔ  5úþUÝlÓg°*x@ù8ç«Ë‹ åhóˆ­tFu‰¶Süx÷*âÅÞþ¾Y’.ºwQjn«»L¡hEñ>Z;UF³yw/rŠÝ¶ÒI½m”Éd¶l»ËØ’Oj´öVö¶,²&v;X'ÚEó~÷:¥#AÌ’Ë¥"só×"mg²Ã°{¥ŽQFiP•÷ó_éœÔÂBÔ=§Ç;G/žª§¨š ÷³rjÆK^•¯|÷­g~ãfŠÙò¸:™dE‘ƒ"“mÓ×.(¯‡É9köˆ5dܘÏL$Õ¡Õ@ªhV4¸Åß„„ŸŸÈ šZix—§þ[ïBÓ#o/«¼ý¶69–iת(X i´×bí¬OzÁ˳^žuà´×Ñò^î.KG>UK½.ðBß ¯VÁ‹ŸIžGÃÅ Ù¯6Û#ƒc(¹["+f–»ÂéË1cm1J´rËÎ(>öÀ[1øñÝJ¢Ï¶¼.·0ÓZøˆ:/ìRûâF£íöB½yݘRSœ×?¥ZR´ýÕ\dƒ—zü “Íìä}Þmy-+¤¢$K!Ч«Ãš‰+Ú2÷|¹ÈfXŸ’ëömÄ¿ˆn•]Yª•~cH ö3_RÄ×è¦Ý¤%t>i…ßMã,àe´±¿ÓÛ_l§³ŠÈTÕ²czù$Ç%C¹EÙž÷¦eÈ…ê?_@ñ¿xá”ËimŸ¾ë5ÐŒ "°’ÃYÞ$Ã&=ï:ãÝá™&»uï»q} _?ž©pžYÕæ!Õo÷Y¥öÌŒÁÂ@ñècº^_¤Ç9ÁÊ Å& =•xÎYdÅËd6Qs“‘¥òÐèx6À¶áaƒ¿¸­@^¾Aá½)A1Îå37òG`õ쳓â^Ã÷ï˜DW‘z9èŽ-¸wK\{l„VycTÏš‡Ô‰YB¤!Ÿ&Žã$Ã7íš$vÕ/ž1959Í3[¸0Y§‰.suå®>pi <£ž=M5Öáõþ@FMËx>‘­4ímcðv°Õ'ôëÜ ïÑ-–ðŽü~“,Ñåþ¢Å³ãíú)é²é•:ƒÜ/}ðm=÷ô±%1ü0ú Hó>Ï^DË:í;J ÏóMw}ƒsÃÐkÓ&yA-ŒÂþ{®èàE»3ßÒµ‘D¶êP‘Ûúzc&IžH%2Qu|ŽIÚ–­KA-ÄÌ„/ ³ZB½¥ÆcxrË«^É»ÀVCÌå¾/3l¿y;PHE’FÞ±tW(Ú•èØÞíÂÚAmMG,eÚë¥q„P~qÚš¹îfY:Èù9}ïÁ!!Öc…:xcÍ ÿ@êâÃ…|D²€–¬éÃÀ™gÿÔ#Sšá-¨È¿Æ?KP—ï ¿eV#MŸ„©òˆ7ïÉ5ØH•ªKˆÕ ©Éx±õ÷š+ÕÚ å³CAKGû×’;t‘ioܼçÊy[Ê=~ÞdSâ«ŠiLuÏm7Ò·d¾þ+J\LŒ6äY뽜tõlÐâµ½w7 šº²œV´¨Ž*ÆpNÓjg¯eÙ3¬íGÃtK_×à±}¨Ö%2d–Gج”?é”Ç…l°(™±Fáç]„œÆžIQÆ 1"¤¶¶XT†Ü!µÓÆ5žÿ„¹\’Sòô8wãÄ_ÉR%xHøa´ä´~â ^Àk® S£'¥6ßL²ÍO)¤ÇjükštwhĶ5œ¼ì3L©>ÊXh†9Ÿ¨ÐöÁF¯<3º †==רÙÛ³™‘ð~†! @•EÚÊWsÀœV%Ku¢ò„!¯L}(7(¦xïˆê­Ía%¯D¦¢6Ã/<ê˜XΟí„<\séð<ñkÜ}_ª¨ðÔ²3ŸÍØ$9%T]˜¾ö„{|jp¬ÜNñȯÑ›ŠeîYÏ{ Ý’m´¨Ö@EŠÃÁÐGã É5£ðš´K}íOšòÙN•}ÎÌŽJS“”‚Œ¹õç«Jâ×ÒÊmVìvs®'£©è³·Ô:ñUØh‘ºG±õtº‚GùÏ (m4|y¤†H¿SKêºçi_ S^„?¢KúÌ™ghE¡¡#WßÛÿ@K¾rM_óhº3A4-•t—¤gø8|•põ'?ûgÇGÆY ›"ëQL{c>{Ù•"ú#›®¨7¦5Ä­žáÛê&«©25´$wšö±~¡¤pÑs]';ÎW3£Œ÷³›$ÆåL.ªýúÀŠ÷,ÌÏ@ðfë­>'¶4ï é \‚ýIÙñîOgÖÀe[ÍÃè7¹ X;†±¤Ûß×ñ:¨Ê\dFÙ)Îc ÈEcg›1 µí¹¦“'ßEß«VÙ@¬QÞ©àÐ3âš{\¾'æ£Vt4ÂöÎâØç 2xd¤q9*èi¬îãå?º¿œ²ë3Ý€¾Sà(g»b$Yé5Å?Ýa0©æ׺fVÓV5Áo}é–{…€!ƒ 5ŒK3Ž¸¿‘™r¡×ÓR7ª+jø¼²&~Æb„áKè ¡yë»æ¬ßؤ¯Ebä´`¨nQŒÐY±e=IÛ·‚+ÞÕG‚¿”_’‰ „w ÷zh(|l—MlxvÙ­UD÷I õ#Ó®ÇcO59p–[3=™9)y"êœj úzƜǭƢßW4{Å…U~ÙøýðîzPI€×ÛäÕÈJì{›‹£L…r¥˜<‘±ßÄšä(“óŸYü•i²G‡xŸ -PÖøñI>é- ¬¯)rŠY7våj.¹HœZß4g÷$;.œ¿Q§¯t‰š“™·*}¼ö¹ãIËX).'“Ïe\Á «"¦#á7°[¹A€c2› {)â"ë5uø-ó¶m›–8ãJš%ÎH|Ib¡õ(å~•¼rÑ]ÿ­O9ÕÏ»|Äêw¶…ˆ;dŽÇ{­­WÚÅx0`<©}á}eXÄÌ~&Dcþe¿Anm)a:N¢sU'š åXáßѳáÞ‰ဖÌC+ ÔJdz/Æo »vQë¦-YŽ!¾yøæïÑ*ò›wvúOéÓxð–Voúñ¼¡®eÂ4¬r‰—U4m›Ã—>ï¹½SÀÕaÎ'Àçùá¬ÅÓÏäqS9¶?Ò6ÑN=iºB0X~ƒŠïÕ1)–Æ+©Ä¸`°-<˜/àÇ.G/è­B3œ™€ÌN/åÓÄüE7¾Y q½ñuòÓ‘ùö£ü=Ñ€²d—e™Ã—M6§ªŒf(žˆÍX܉ ÊòÁ‰‡ò5w\(î½-JíãÐô'QÛŽUËJä³­Ÿˆ8™@;AÔz3Õg(»Äºü‚/æ.þ`'^økŽµg¸aƒk ·ýß)n†ÙìDqÊWºîw²„^šm©ægôŒ§ÞF¾5Ÿixøáúë¹ní¿Ô‚´}€þ><^Y•i7!?tçõ«Éfcì#Dèø¾N^ÐPŸl.1ë”…æ Õg’{xÄäG³±–#ÚI¯ç†?yJÒ”Xz *FPy1]€Ìœô#/ÞÊîb:#Q!oí±Õ>&\9Eµ6*+:Œ³¬Kß ä“-VÒX'e-ˈF=L…›Ëh$€ R,ôÐßïyµW=¦fwl(Ê¥8B½’c`óu®]p¥ª±ýâ-_b›TÙwuWþÎÂÅVèbºŠŒØO>#ÆŸ™Ãùƒ’kŠq2U-‰Ü;|–‹ÈÚB±V:©rÈú´Îà\í¸xÕñ—ÞÏ/ƒ}]Œ37¸~Ök.Éy¡¬Q6¾±ÙÑS—¨&r«T+fJö²Rx{ð=íxÅ÷e)ÎÛQóæÊÖ6·Ž@ô"éL¬ó{}–†üdU®éuã‡;ËûòØŒe×níQM‡ŽW$1Ò©9g³²tck‚Ü%(±6‹^q¾Ê·5ÒÎíu5`rÒ•+ ÂデÜü™M.4ïó/<Ô<½q6¡I ³—Nç<«Þ 8#‰ðŒ¡\²à{°–¿Š(3“D †Z?lõ4jþK:ƒ]¾gëûoÑ>,Ïn&E~– y˜Üà·¢€7öHDœðÙ˽ºë}ÔmºËÍjI{ùݼ¸º€9éÓõ­Ï^wÄc¬,Z‘~7ÿUª"xÐûëBò¯„•íŸ7§ž~ »w*')Êtxçgï%S{£PÜáYŸòÞáà <H¹øúÀ]ÿQãnî[¿cå+ bêÒFÁ¼ ïJCaŒ´:Ò¨Ô•0n°©ª,±‚xñ;@ë–»§Å¤yæ,ŠÏÕeûJ:›ûK¡zÃÙ‘¸÷á¡A³@ qhGÎûÅZy;ãÄ‹8mN[‹]¶‹ïT%{ƒŠÃ -â-ƒø_¢–mÕÀó|sT•ˆ Ãn§]mv›áU§àOăŸ^z6ê8-bÅÞ|‘Ô÷GД} h öi5ÑÏ5zÛͺá5?²éâíööy§š=gðg:˜C!íõ u<•8Úi|ž+#kS€Ž¹¸¸[©ät'ŠðdîG õ-,YíÖÊPÔ»×̬³çˆ!…F>ìT.D$úŽ_u„ÍòÿCɧŠûº‰)¸¬[ ⵎÕ]ýg Sª«§|(rº¼†…sÙ:ûPå„Zú½‘à¾ñ‹–ÏHRîйeá ùç–òr¹KQëÕä’sÈØëä=òLg× V>‹¹ö4-M–úz"°–î»tñ(‚7ùø'eLêÞ]òû$Ö:~ú× ’ês×ÒŠ§kê4,µûùHs&><&³ÏÑ<“þ6Œ­¯¦†/Ë]ƒ°ëÙ1­s‘€WDz †"=½ÝðXvŠö”H±ðÈÙ©ØF¦Ë±?(® ËŽà󭢞j𥳮¾ñEÝÈ èQ™õRtvBQ¬Í½?+{„BLJÓYÞ)©z¸Q¬Ök½û!Ñ„MÏcg>~#«ËOp„{Ú¡än‹‚ Õfΰ\ï÷®™B­p9´i' PÌõ6EOŸƒ§ú\Ý•’¸„éuΧ ¿NŸ½²Æ¼q²dûeCÕŽl—Ífko~ÝùŠs9³³×0ä‡ú–  #¯% nÞ0··&­øè§lô8‰qsÉéÑ\ÎA¼Œ~î™ÛõŠùmr_ÎÄQ7Ci 6ÿä‹‚ŒË‚í¤7»ñå_ç7”?õ’¼KÀÆhL7`ZWÕÁZÃát¸òQf·¾(5 ¤éþ>ùù£þˆÏyêMÕWú4B´F˜ÈÈk`J¸ÏðØf›6=Û‚ë§Ub¥$4^²vf¾¤GMaK7vÔ4Ñ—"«µnÉëâKŽ”ªQmÑÕPYßAÊ—æKž–÷8%=¬¢ß•b¿yѱðRH7ÌQnzYD:D:øÛªö›Œ8Z³çž ‰Síw\Œ‹¬ÅÛD®“}ÜÙHT­ïUtBA Pê-·Z–‚‘¿OV†eÍSæ튯c<88!˜Épž£èj‚¤ýCŽA vᆕa«]‰J^7”¶sÁ­Ú§ Á[ÔY²Êó×A¶NÆáUWO½Q‰^ú%É®L·ðÛÊÃä¾Õ¦ÍÛ^à´ñƒªa.6(æO£ÉÖrôäecû¨W·®C’<¯ïµ³ð*ø*Zülå¹P¬ÏLË¢Ùä“o*_ˆgsºÜTÃ'ë~_=ÿø-€¶"./_Qh…ÎÙ&#pu¢ÙoEmažæ÷8Õ­Šî\Z×ßq1äSvÁRâ³w©>\Õ` âÛÁu·Ir¶½üúHH03¯i4ΨC…Ç +”9žt äÈE¿^ˆlß”lÂeg 0.ñ{û(ut±§Ý9é½Ãý†šî·MjçEïrióXc®mVÙ)$}ÐÑ4œàÑ i7йq§ºî#J¨Å£¼—w4ñø%;à %ª ½®’ΙÚS’g^pÁ ¾ÂÙ²ZÛr,bÅg çý,ä_pzŠ Nži4sKëæ¸&–§õ­å 1…]!éÅ6…º¦q~´m`¬mJ;º+À­íÁäïñäçó“&ªÜò{uer«m`íÀIøW‘qÐîàT¬Ǹ•cïvŠzmkûŤÁm÷o¯P®aåmBÈéÍ=1cÓȮִ8À²}iA½†­TtD¦ú "™sþ óm}-1‘—ùíoo-àJϳ­ŒëÙ0œOãýS×¥"…ˆ–»zýØRñ´ãñË^õš¸ ~äaXòqJ+a…P1ñPΫ•µ)4-&Ð÷¾Ë¦jÂ9 F 6˜»öVoqéå²J²á–|ïã wh]ßiñQå|úÎZW'ÿx2Íúç«]ÝÞ† Ö‹;Ç ‚”Iµ–ÅAóÁäš Uðp:\AevkézJš%D‹þWƒöËÏ7vÓ ˜0¬+Œ6ÁÞ{¶åÍq?Ô»¶HÍ4ð=›\ñêä:˜:eS…÷ m<4ÞZÄŠ=S¼ßУڼ#tÂy®jðŽ),æjÈ3c*l;á–J2Ú¼vmvȶ tf™Yx©Ô9'ÖüžX»`ƒéQ_$îC±Œ(Œf6Ì;ä Êå M„˾^F+”3iÚ÷¶¤ìgba“À> aÖï-‚J‘VvZîÛn‚tÙ¨‰Ã絺¡Àö“õIŸ9e¤ Ü\ÓðÒ±ëWòT•T÷n'¿ìñwî'Þé1÷£Òã¸ÏäÖ OV(L›¶]À½ò 8ÝÌ]1„~SЂ!×Ùw*½r+N‚¦NÌXŠºYÂ:®×Æ/\¹+£úOèÚç>†=è*ÜêuN4òà?4È*%ïÏJ%Ï&Õ0ÒØÜ_ºâÈš$ë*MÂI/*}‘‰èRžÁÿÜòßïò©³pžk#Õ¬mù4®6çSA¥º7Nžæ|ïίù%’ËæM„M³ ˈ¾«˜}âZJi”ÿ&Ë«Ý”o}ºI”ø‚1ÂõÛñ{>s åñK ;“™Töø[`©0áp‚Ž˵Չª§xíè|EŸÁÇ}pˆ‘¢cëÈ“‰.ˆE;Ÿ˜Jil9ÅXâKsóu*ÑÕ·E¥FÞ?í†Þ_›Ï¬›ÇW—i f·ò-l3ð×6$ªžRI5ž\¤õË÷<±'GŸ ÆÔÛMÜ:Œ(öÕÛ×5°ó²Ä=mý¶î×^ˆéñÎElÓ+<ƒOe¥ò[<®I?‡j‡v=¥-Ød‚U~læ^_ÆrØŸõH &¿·ÙO—ârD4êì·4ü¤Ž^Xîón™Jw‘1Ä]®T |~þ øÒGq“5àý›þa.ì¸?ÆÎé;á…Ö‘ §œ–R¢Êœñ¡—ÍÒH£Çm2$TJ`Ág¼² dDDDgÒ àïGe·Tq,^e¶¬´=ú”§`Æy/VF‹2¸˜d±Õà·þÜb`<3A!ìc?ÇLˆcS¹Ã×¢T@~vº}ï Ķ$?_ô¥»‚¬A9ÚwÓ8–Úø ®ÑãÛ(¡E9jáe_øÝWñJ¾èá‰å_™/=Cˆ² <Ò9b¾?SgºÆ ß„ðXV}OúØÈ5š ·µ “M("ï­ŽßTï‘5Š3O+;ÚÍŠú SdÊx:n2é+ngšÒ· ¢5,Ø[âämš«= w1 ºÅÁÄe÷¸öhd+p•Ev woM71ËÆ>ÚÀo! LjrÔøo‹VGà$‡“}¢#yüì£É©ƒ¯›iÈ«s†ábÉÕ¥«KÚÊ램[tž|ú]bŒ†Y`OÐC›Ÿß)¶]KÀó*#`­ ÌASHÖ̦zÖIÒÞ» ‘šo‡·Ïfpš,î¾`]­øºt1ƒg¤½õ¬³q~¼²·¾%QHVÛ÷\÷±Tkòv(Òb ô0•«GPòý´Ö%CKêI­¦-Ii|r««–RŸdбá2U½aXéc÷î\AÚó3ÔÀÕOwûÀPžÕ~^ÍL`åþ“IH9pƒ0û¹'~¬Ó¦çNØ}ÃHn® ‡#uæÙϨº-æéÇ}V4aqæ¬:éW˜êótd8²?¹”V;—­Ð]O,®ûg®»²í»Ñ[wú3¸n§K´N-*ÒãôÜ3È€ËѼî î}l±]hÛ§ƒçê#ДŸ¢²Çnòd"¿È&ÅâÈlŒÅ¢½NwçÁ>ÜãÁÇÜ:qÂ$rã1®‚JOþÈ)®Ÿ½Ÿ¿¡~ð…WNÒð»´$-ÒÚÖº¤f1ô‘AÞJi˜*—áµ_+·æqÕ2wH&r¬Ï¨7Éþ\‰ô“­{B5³.Gùó±œ]‰òÀÈÄ{'èZªiDâÛšùWÛ;Ɔ°;ÎæñÇÄ.Ò™4ª…/j ‹¾]mJþõìü endstream endobj 45 0 obj << /Length1 1418 /Length2 6406 /Length3 0 /Length 7380 /Filter /FlateDecode >> stream xÚtT”íÖ6)ˆ")!C 3tIƒt§RÃ0À3ÄÐ ‚4H*!%]’ÒRÒ!C(ÒR’"Í7ê{Î{Þóÿk}ßšµžyöµãÞûÞ×õ™tôyålÖ0e$Å æI45UE $@ÀQN°¿`" ÌÍŽDHüG€‚ ‚Â`Š&N‰¨y8À‚°ˆXT€@âÿ DºI!žp€&@ ‰€¹.>np;{昽8 œ°¸¸(Ïït€œ3Ì … š”=Ìs"âÐGBá0”Ï?JpHÙ£P.üü^^^|gw>¤›4'À Ž²èÁÜanž0À¯ZgØŸÉøˆ€{¸û\i‹ò‚¸ÁÀ …!Ü1˜s8@_U íCü ÖøÀøën`>ð¿Ëý•ý«ñ;…"] 8Â` w‚´•5øPÞ(aó+âäŽÄäC‡ÿÞ/ÿ#ycS-eî?ÿÛ'/ôøñ‚¼ , ˆb^þYEÿ« Ðß©ª[$ úÓ-æšþÕ±ç_àøKœ€ÓBbX püMr30Šy€ÿÏTÿòÿcø¯*ÿÉÿ»!e'§ßnŽßþÿÇ q†;ùü€!­ #M$Fˆÿ5†ý­&Ìîáüß^U#9„†Ì¼`!>Ðî® ÷†ÙèÀQPû?”ùƒþ’šÓAºÃ}[0Y Ðù0ú‚:b¾î^þqAÜ1bCý^ã/†‘Ó?ûPB@‘6¿t' ,€¸¹A|ˆ0«ÇXÂ?0F 60ïßÌðó!(L 3sÀéFôkÍÂâ~Œx!¿àß$ à‡ÿ‡)àwømþã`¨‡›¦³ß|Átõ/û·úa0o”hf • q¨ i9©’£óâ]‘ÂßM=1àÉ· Dõ([.&êgf̪—*Ïtƒ•-Úµä]O²¾Lú­¼e¬õ:æeTþfÇh?u}Œ=‘ä÷“žyŠ´+Ç8Yþ¾D‘[–Cô÷7elìzô¤þì5]è$/6ŠbeQµÐ­¶÷UÅTôâFÛKz¨&ë>ƒUá#öì%ç¸gË1f‘)Z¯˜Õ¥ÚN‰½*ï|íè›'ëfèˆRó ^Ýc C‘)IöÑ+ TÍç’\dyn½ºëgý\^…óñ8gRqüjzñ^`xå€óÚó5<ÉΧ[¥rŸCjDH]2¬t‡>¬~[}4¯vÄx=@ƒt£%®BpO6ëþvþJ ˆ¾Ò©»Œ`¸Dÿ@ÿ –8zusšÑ×P –éÍô@%ÑêPC/vÁ6™Èü.ÛŽ “Í‹Ç¥O¢»ì’>WŸD¥Øm=º2°¡–© ࢉ”¿püTÜ”~úQa:Ä!—«©‰nλaùœ¢‰_\\?S¯ñ–Ó¾m^5ï‘y0ÉÍ7— åJ†=t—Ý–©ŸØ~fÌÜâ:søIQ~š@ü:ò‰U3‡c‚»ì7ÂWñ¦*õßJê"Úü6J¤ß»O ‹ó³Åß~ù9š¢ËþØ™osk´¤.’½üî*®k–åÍ ò ÀÐ*‚ø¾±ån«$χ| ©nKÕÄÒ¨û•açY qx¸OÉ·QF½| öOÝäÛ”R ‘‰’ÛI”­÷Iße¦kõú©žcAF©¾ë™Ý¢6©P殦ð'^p€å£uõØoCmÞ”š¥=«[ðSlt÷`çåæÞ¼ýC˶U+÷ …#?8oPœp ÷‚ÓÚØ—ÙýâÑ¡xeŸh<ÛO]‚zJZ6œ«µQOú“ ~Ë$¿szÁÙ§%;ƒŸ"ÃÈøØ»¬‘báÃÄÏvÚ¯Í8¥öïFù— (âI-Wuy±Î¹fêd»C:762˲JÅevôÚþß}Ü>3èN­|’ï‚žÙ¥Ž?ÈJ¼¾'»lv7ÖÎp=B¯$tˆõPœfMàÕ,ö“ܧÙãZ˜j2–¨!¶—°=ý ·¡•ÖíyÉ~¼Ô}W?ôʇӧE±A;üÉôw)ñâC(^ªŸ_n‰uû l¢™g?æ\ÝäNï}9+QÞ˜û˜Bë‚,ÿßYRjA“vÊ’°½KŒ´Ï J“0öªíŸg¾~SCÊéš3®Û²ñü¢d«îU¬)2Zò-ZY‡BD¥F“à¹óÃG&×⺞86®Wë¢ K„ú2²ŒÍ™s­“ÀQP­‹(ÕK~ ï÷,ú¬si’·MélƒÇõ-͇“Îq5ŠNpkM‡ûß3µ“jðµ‘(-Â,s²úÜ ‡ªÌ³,Â.Na¬$-+á,Æ]Tá[—GÓ=¹F\ä´¼TìÍãO/#4»lFÙª¼l‘ÏVê½µôN«×έûtòi”´„O‹ùZeQHõ†µ„‘ë¥îÝóãpåS×3ùÑ~AÉâöýeSçò‹Â„ÞÈäÔ‹ˆn X½÷VûäzöaHÎcFºV­Î[óÎT׿¥g¦n›C.Êúï+m‹nãüÌà )ª,¶ •h`ݧ§¦é$#ÂU t¥¢´ød…5‰oTëmÉ;Œ·âbÖ3[pý<ö,tlÏV¤oÒ±ô“RõsãÅÙ-œ¢fðˆ÷ö⑶kÒ›çI&S•Ÿ{³ø:õGó..Íž–W,MË¢(Jˆ×û…¶º?m§šJ p8M„Ïág¨ ¯ôr/GÚÉî;ÛÕŒWUánD脘°‚1gnLJpAÐÙAÄð´HšgÀE¡€|„4 NOñøСëàþ“e©òrïa!«9d×HŽSeûK=´ÖéLê{ßêºT.¹©:€¼…Ia)ïŽ2ÇÝH¢KrŒ¹lð%¨ŒÉòÅ5ZE‡PðWÜôÖ-·+š¹ââêò4ÄY‰fs 6áËJÖX5ý´’x-ûÜÅ‹Jâ63½¯pRŽûLx"ùYÉã÷iZoŠ+ téäVß þÞ(ZÔÜÿu¢¡Ú…dU6g±,u}‹{"¶(6²ìIû!ɲ…ñ;êÔæ–TS]ë‚AWÕ6V®j1ÑK·ÆÁ&Ú*šP£|ÙÆÍ gÔæ6@¶Ê^/ˆÞ¥UcÔÑNÖÎ’DRf~Ø4¶v½æ&ñþŒãcH5ã.-G± Í‡c]ß)Ïd£IÙÕª®>›ÃT«7uj–{‡÷X|‹*]’!´»W—ˆÔ«)8BèP)µh‹;åI–¤ß“Öy0PøphBHô…,RPõQš!9îä‰ÌöÃyjOÜYŽ÷ôÖnáOkÎj-ž“G‰Ÿ4‚5·9#X®¾Æ•ßí\™8Øä.sv_“(] Á½24¤Æmí ¥çkJŒòŸ“³hFP…b #Þ& v»‡¨@L?Äyƒ6¤ðFÕóÇC{¶ÙÔ±䃖 —U%³±+-ÔÁ vºµfôË $ñ¥º&$¢H%g‰àbQSš©¯Ïáìò?[EPïr¢?Y}m¸p䒱з~Ј“H•»TðŒ®àÝ•#§Xî]ÆO½ã·×k½,ðïiÅIŒ¯™>‰Ú°ØåIï„Í®fƒß>Ù=¬8Ò«ú$“±=ééêq‹Â_®aºÉ·¹WkM±ŠB˜’΂¢øüxXµ)çáëWë³°~®گϬ’+§SGi2?À¯£‚% ŽZç|Ep£„Û_¬Å”—Ž5Hâ8gOë‘öíIì¼KúQsxÐ:’‰V sþHg“>¤¢C8Qï }VhôóRwy?XŽ] ·çév»ƒÊh™[2ãÜN.îôLÛã«8ߎísØñâ ÎÐ ÎTsE:/p$’Äßh5íN0´ï—&â~_ v_Ù&‰fØö6ªŒah]âAß7;%Íè¬jkÝD£ *6ÑÇâ®sdºHÕï—uFÉxqšæî¸S@cùÅâÓKÿjÛþ%å8·ÒS¤zFž G‡x{7±‡ü»Kx_¸ ½î\îâ–‘>Ï$”a'X¶k÷ES9Õ'…Ÿìmò¼^¥yËÒXøªI®^A¹<-ÙÔ9+ÔâÖ¶l y!Ú}ª9~WšÎ–Éœ=­Fözü“‡—U4Í“ºT±ã> q-Æî3TËÄ]UKl‘5AªtdÊÃOÕvÚ`ƒ‡¤ßž³ñÖÀX^:©knvÏSí${:±³‹ì§ßŠ~+v’M¬²Ð&ŒMÏ3¯‚OG ÂL g½‡VV#ø}îÙšïmJŸ`»Z¬Z%Õtb‘­=a›€ku(x…×±q0Á±¶.›šc|Ò¹á;¡ç çuâQcAi T¹æìâ$) Îg½ƒýåå3Š}íŒ1mÊÜMÙ„òöèºwt+>d ~àåž²!‚ä®gœQÊêAOÔøNÆd„|IfÁm/–Yqïñö );ÒëT!Sbx9e÷¢N+ééÂ\t>ëºÕ)tÂߤõ¡Í]åŸ0²<£ˆ¨¸ÌzaÖ c`lÀÞ,Lòaó wV!Û73&µËÑô™ò=>½»cnþ7ä>A­8SÈ Z†Â;äcÊ’RÝ‹º;åÛ±»–³ŽÕ¤Ù¾2GÊïߤч¾šŽ:ºw PX1hÕYÑ0¹Å~bC=¬ÉéøaMÈ©sLäåw"óGmá7¦^A[Ï©“ n[Ú¿_°¯ÆÞîxÞÞ2ŸÀPÞ¯PWí8ú ã±9²ƒZwSëR¶oÇíVyšF ÅÙЂÇo)*„Ð&½bnÃæ"ë`.%jœdܼpÇÛ s½67¬xÝ—k¬6ðÌÆ£Ù?\m î´2-)×N~L³­“oœ² íž/âàÄz[™Æ¬ûü+Ðú£,M+w¢…NƒˆÜØŠciwÀW¼‰‡xøÏfU¸Ô ÌVz³–Ùu(’þžæ‡XGô…„[~š bŒŠ\5z<'+@Þ²« 2½©/S–£¹:¾=ºCŽ« x®7&49Pb€6‹uªÍŸÄ5fNêqÙÏqʃQ¢,æ+[¹¥w½µ^Ê)ê8 ÂuëäªâÈ fÄ& {¯F‰Kåi”¦šBW-¡øz±wÒ<µî~pRþ’nÈÈÞ¸âã¿6 „•wL>#HwÁPmªŒ|mÕiîâNEùØrÔ¿ê;§HÜ&‚žÿéMêö×pÞ›ùÕ4o)òºh÷KÅ~l ý6¦NݧËfx;¥¯êÿí¦IIŸü—y—Ýa´¾Ê©>ÀGë1»§vZ³™ï ¨Ïø4”x…­ê4Ç,›me§GS–íä$-ª9ñ—ôÉîŠõµ µÀH’áŸ=‡g+Ú¼óÜp6§¨ûpT+ºfä*ô&MÕODÀÚ~Ô׳ǣ+¬> ½Ù³%ÜÑœ‰øÞ® õ+AÝ—ó§[Q呯QéIñ…-ÌwøQ!üšAIåSw½ÞÍþ8%‚õ3ˆrÛàÑ3>ÉýaÍ!2Ú!3µqÈ=#2Ÿk5š´À›ÌqaÙkQÃ(aå7,Sx-ßnxN0MÀ–ÓŒ}cuþèxâê挸J·«¡ÀÁ)ï8Zø´ÅfO±þVþ=…Õ÷z'žêkJgp£/Lgë§MµÕ’®/Iï“®¡â/´¢Æ¾I‡);Oª;w íL][BVwÖãbõ;è“köŒ‡)¤l'×áÚ-C;Ò&ê±»Ì=Ò╬‘™ Ÿ÷Ê%—o¤¬â‰”„‰•«:°éiäœÒÉüƒ¦Ï¤J=jÛÔgévïm3?¸Àã椾|×l&ãÛûIm•qoí…‚oâ“‘K7h£&NÞÌ vâú±ö“6?9/Nƒ¯Nò ± ‡íöŽÔ³= uŽ¿9jPósŽ}/æãƒáÎâ²Ww2©,r—ˆÏ×¾*/ÙÕì7ú™œ:ÕÁ¦<ù½\zÂü~ìlY¾ x—ž*$XǨÜQÓ%È1Ç=L¢Se)“ÉÎ{Ù“«µù0 Dþ\'R—K‹uzúæ¢cv†˜¬S5}›7¾;ýéΡÎÇ!f|s‡ÔtöKµ&Ÿ—ëø,4S‰N­æ*&dã¾píóßÒº¾‹¶uôAiTl[»›T“æ€B-óU(÷û¸cïÇÓÜN+n íRjDñ?»ïêäUv-IuàÁËËƧr¢cÖErñÔ¿gzó¦à{VÆÜM²Û—*¯b›u%òú³(nM}oPü^ZR—è!à~pÁq„ŽØ 1ùy‡¸¬¶ZpÍkfv*¬ê†LÄÀÙ}5;ÓevO{Ì@îšwý^Ñ3ƒJ22'ub~qé\ŸC»¢Lˆk2 )R^¬Ñ‚ƒyÔ:J"´Ïôjß>o(hy*$“8 '¡9 i{ °Ê߸á€m$¸œSoê¼ÏâÁÒ¥a5ß ]$W*}“ b–s¼Ö¢Þ©/Ѧڣiv„j»žÄ0G‹Ñ±qªÎn&Ê&¾yS@äÌñ ø±kC ë£û« Kþ„™¢-™Yf¤¦ÐúÖ|IÙ˜Z@åÈ$z-;*9’ÙZ„C+ö:Àƈ ~L·H ä —Qäñð­®F%"mC¡|’ùcߣnÑCNù£+]€}ëNů¯çnry0ÓhÙŸŠ¤ò¬\ÒERˆô|)Еv°=Ò#+úAkÐñ]c}§§h–¤–|ÂЄ±§È!‹£c­¤,‰º^¸± ê æ»—`…|만Çy|¶çaG^ÿ*.Iž}‰úÌš)„S*9$œI¶Kp–?ÓBDÊ+|&EPéâëöþåº%M[Ÿ=öS¯a0¾Ô¢Üü÷Þòþ9±w–F¢Io¨`\›,Gû?uö\Ê­yU’,c‚¼kÛ®(Ä‚#öxXÒÍV ¡ÁÂDá«óXxIÖ+‚»8fvão4›!ÀŠÏò~KYÎùÒF@Ü—âÂ_¾ºâdÎüXX-)I¹zLžä/xUªÂçð´ví'Øïû…j€ì8Ö……>‰* ×™ÓÆ3ÕH&³p©U俹±.ï=^+ëÃ#4‚ÅÂÅÕ¢c*p³t)3­¢ïéȱϒÚ@ºìè„õäqyò]“$?EðŠrOF½ÂóP=èw•;Ý<îð%#íjƒ2œ¹_觰å‡sh×SªCÓŸ’¼k£Él)…ç¢e¨œŠN ßèîŽx9ÒJÓàæwjº»¬;¸]îÚfB1MT³›îuôqØÂxË«ûb¯TWÑœ&½ë‰¬ØŽÌWÆ/ 3)…;^©3c +\ÁSR=hݘœ”yö¤*Mhtm½–[^šHÜ:Ì~rzú²ÿë`ãÌÕÇmDÎ$"ºÍ¶ÆvÈ”M{·Åè¶++Á=ÒæFû{{bSSE_úžaQ5ÆMs¿Í<],Øò¹}¨}ÑûÔˆgä Øz¼ ¼3ƒk.wŽ¿x+Là˜j¢,Þ@Ž¼<¯ edàôöýl<„¯ŒF2®tØíGÏ"Þëkn,µC¾ŸœŠõÐ÷0;¦Ù)hLÄ}BŸ3&½Ž»?êDÁ’±™K N•$ÿTÓ¡Õ$è}à¦àxùlUßñBé^s •Á–{ñãºzòáÙì1z­ʅɘßâ·Á»EÄ·´uCF ÷² ú©6,ŵ§ÈÇæ;Ì{[‰uA] ø[¡»\;Tu> â­ö€ï©[;7®JÍy0™-Ë_S,”$v4MôEé{±° Ù*èÉWŸ¥ãn‹ÔÙ¯œtÚ>‹‘"îgôÓ±CÀŠ÷x >šNø¶±Óç©ébáBùXJ/½‹²+ê¶óJžJ ð&p›ù›l°Æwµ‘êxç bã4¿—b§œé•¬†2ä¾{ÓöÓ&-Ä~ŸÞb%˜ëðî865gÊeedéß ua~Ú&¨rƒà–¾q“®®GvÕe¸õ-=W,÷çcüM€Íþ¼k°)©QƒÒòŽÚâ±Ðá5rð©÷ÄtB<éÅPR{¹ÎªåÅ ü8î˜×£÷n56&utŒÊ´&®M²ÝÜ')õ 9†×¹¦UÆ®d€8E&½L.),<3;›ï ±Œm³0ÅÁÙOw›U†ë»÷õ#Ä”`Í­ŒûѶ%iãtLn¹±+œ`Rþµ¾ÜË!á’|ÕÇ  ´™îƒPSVBï¹&ò”MjÛÉ€iq¦¥¹_Íú]Bß³­’<ô£Í5oŶæƒ;Ò§¼UÇLk8?¼Q™­ æX©¢L”|çiÓÒ(j˜ô~˜iƒÖ¿_ÈJóÙÝêíL|>0®ï¨\>ù5™ÄÂB‰ —ŒêÁñ£æFÊy–T\¶’CU߉³¤Œƒ»Ú5ë÷” i],{/Ž+3xÝhäÍ‹67tk”_J:qeÑ[§Õ4Ë‘Õ‘®ë ÃÈ]Åvv!‚öM)|Ãä€@ÐØšTI›r õÈûñ°Ú¨àaƒ±+}>+Vé‰ßy~‡ÉÜr‚b,ÞåaÄܱ±3¼"G”÷è~ã-tŸ¤ü£À gGV@Dôú¤Ç†ÏÐÔ‚³»ä˜ìO5T¶®&˜¿J£óƒ‚Šp/þ¹ÅHÞ6±øÌ›æÄqÔZc{¤–ÏÞkùï)¦õ дñ|ÀGcæAù§ªX¥§µ>D¢©]š…–·âZýúD}Ÿe¹ð[l&âœçW¶= ȹgnƒ»ì2Ó®W£<“Oó F¦»²cN(bí_"E IN¬¤w–ҋ'¹î„¾g–¾Àºõü‡9«Q'þâÐÓ䬚´î¼œbfŸNÊ÷+YGµð@…°LñH5TªOǪþU¼[À6XÜQ†~îpþ„ôúü¬Ü‘=vMÇ£½›{žpïëÏ^àgd'þøCsS endstream endobj 47 0 obj << /Length1 1442 /Length2 6263 /Length3 0 /Length 7249 /Filter /FlateDecode >> stream xÚt4œ]×6Q"DDî&jÔÑ[ôÞE1ýá!ºHôŒ „¨IÔ Ñ{'º „¨Q¾IyÞ÷}Þÿ_ëûÖ¬uÏ9×¾ö>{Ÿ}íÃƬgÀ+ƒ¨ à(^Ÿ€$ ¨ý$ñ ²±BQ0Ș͂ô€"à’ÿAPDBÀ( ¦FaxÚ8 á @BHT$&)  HüMD %%°ÔÐæ4pˆ!›"ÂÍ urFaŽù{ pÚs 1ž_+ µÃm0ÊâŠ9Ñ  öPÊ÷!8¥Q(7I~~ooo>°«é$ÃÅxCQÎÀˆéq~ è€]!¿+ã#d ¡¿q„#ÊŒ„µ‡À=0žpÀ¨kºnøo²Öoðçnè_áþxÿ …ÿrÛÛ#\ÝÀp_(Ü p„ €®ŠÊÅ€á?‰`˜ãöCa`; áWæ`@E^c üSž‡=ê†òàó€Â~–Èÿ3 æ–•áŠWWåAø3?%(b¹v_þß}GxÃýÿl¡pÇŸE8xºñÁ¡îžu¥? DøoÌ ‚DÄ$„ˆ;ñ±wæÿÞÐ× òËú c*ôwC¸Ž˜" PGæÐßìPHOH ÿþ¹#¨= °ƒ8Aá„ÿŽŽ!Ž¿÷˜æ#¡>€…F{ @àçï_++Œ¼p˜ï¿é¿úËo¬¯­©*Ïý»âÙ>€?¯°À+("€~ŠL ³üg=0ôOÿá«wD¿³Å\Óß{ýçŸáàþKQ-àü·È-Dì1ÐÿYê¿\þ ÿåù'¤â ƒý2sþ²ÿ?f°+æû‡€­' 3ÚÌÀÿ›jù=´Ú¨§ë[ÕQ`Ì ÈÃ`ÿºF¨‡ Ôâ EÙ;ÿVËoÜèç”Á pˆÂúóYx1­ù/f´ìbžŒ$™ ˜Éùç‘Êp{„ÃÏÀH$Ø—ÓdÌNðafÑâóKÄ?¸˜òG’ðgG%„~{(ÒÞÓÕ†abl¿`QÀq÷Ä´ãoDPàwÛ#1‰þ Kü˜‡íß{!€ß æéñ øG–öžH$fPéSÂßû_¯â±'œGØK…»T‡7Ÿ¼’§óæ]í¿?¶jòœ‹× ùÖóìÖõ§\•™GòO{ÚnÏ.+sÊM3]øo5Ô\lLÖoúpn“ø`hµ‰prêÃ@á–üëN†ô¼†rkîÆ¡q°ßk°åº{ŠßÒË'?ñîPõy݉žé}<¾ª¿V)ªItŽæ5zbZ<Ê–g—5FÍ‚âe ¸G¶çC2zx4B–3pŤ‘ÈM¸+Täo¾(w:æ7÷ÒPУ…†•Æœšç¬wˆÝ_a#Uã΄IÑì‡IŸé"¦bžg³¼·ù63*¡bàuíU^½Sï@+¹Opºä+ÉU¥u6H]ÊÊ&›rT•ÐCÃÆ;G…¹æ#ãVG—gôÓ﯀Qb3ç­úVÿ5îs=?ð?~ÖxÒS÷q¹5ФX¶S–ÞÉtOÙÙ"zvF2ƒáÕw­Å¢Nq›gÚë©ø98BÚÿÑ7…3ZÐR¡€€¹8^;v/ýW‘Ü€oqåµ{:ïGI'`ÞdVaWa¢:WvïŸííÒ8 )Ûf>æÌËPný+o>žC_ªR;†æß[,§âS*õeÒJZ™üÖÅZ$¬®Ú¾:QÑmùÈ×åT¢€ï«q¹QS‚H>óÊv oßFmh·BXD‡Ä‘ÃÆÒQY™‚5å&Âøt¯Ü;J¡'XxÞ_ï„ä´x覩ÊÇ\{q¼<Ö~Ãî‚³k]<‰‰áý³å¾–çÞˆá/ýxKeÂÉ›EE$%Þû›Ak@åU(B_º½ÿÑmÐé,F@iAÚ•tÓ¦ënMÝédæ|óÞ|nøBùfAmí«W²=4¶²b©[TD6½¹xO߈ûèI ƒÆî£NÔ°¦Iˆ¥DÓê‘šž M¡éFŠXxÐtª5TyõÐHà‘/5ãøe=#›å‚ûnˆecãš°®°³E†Í±zcþ«Ýוç”ÖìchÊöµ5NµsÌ^öÆ{ù‹.P§Œ[ž´×9î=é,B4÷qR² ‡Í4ùf5©¿ôpµòÜQý¨éŽR4WPÍ~KÈÈà;Ãy+Z“ãq» ®Àˆñ³ Qu=VVØͳ—õ¼Éw-Z¨·F!}=1']´4š½ì‚lN)v•~‘ǽ4i<žˆ í—‚Ô~óžIi¹‘£¢÷Ò›î½ îl#.º’xß_² ÛQ ™BeÏEýiàø³åžSªœ]¶t9n ãû8¾á–÷ÛäµfÖ\KÅúŒÂþiÝ£ñ¹"A¦ qwf-‡áŠe:ÄVРȘÿI·Â}¹ªïg;ÌI­fy¿SÖÓÁ㮶Žò÷C=”ÞŸç? Z!Ðí @…Í…¬ºó'Ï<ׯ*øw$kA·&SJ»}EµLÏ *Ä%áÔZpƒ}oöRÞ­‘W•Ùª~'%_¨ïws$­pö^4«Ÿ;ëœÕŒû¯è’·žƒp´¼ìÇ\)ñ µ/ZØ»Y´·¿¸vÈÕ™Œ=.§.,wµc˜zŽwïÌ6}ëzG^òB2¿ãa{ì-›ž²`ž$£±<—‚ÿš4šê«9Z1[¾`á8 &Uáÿ©…ˆónã¤)\?‰… ë9V¸¸ÿƒIDik‡áΓ½kvò—Ášm"òñ‘g&¤ —ô‰o‰}W´7®DCÊr«¬ßƒQÙLrº ­<þ,,•oL)o„|ï¦-Û“hkO ªvš_ÌY฾¼jôx¢Œƒ †>-ciÓ©€àN+Ä›(Юs%àÎtÏ@ixòvCç‹wQÕS½mR‰íÂA¿ÕËâÛx«…;âs‚û‚ÚÆ¢ßÔ8ŠôN×ÁwýÜJbC½«C¯°ÜSBÈGž$·ËiÚ0d°sòY}²'Õ·„ð'¶Ü0Z‚cˆþñx‚œãQ!Zµ=°×Q¥Í«œ>Šæe Æ{è•®X²Êxmî.;·ÇRþsèÞ¡åúBTÆÆR“yÍ}Q#·ˆÉ*tùt³¡<ÍââeãŒaÛí#œrî[Þ_”cQY<ÓT91̳ûñ<ñqhŒæTÑš5m7\;Ç\¿Kûdùùâ,¬ºóª7(“Xr7.ª/î-’<¬îä{8ÄeÀoù%¶\//X/”§×Êö£±Š#ÝC#‰¢‹òÑa^:Jïb´UÅ$XÕµ —ìÉò“NtŸá0Q_ìHñ*ºµûÒš7V·ôú uãäå©?z=Nž†å³¯×ž˜­NÔpLõc™L®†±·ä „j—4”éüÐЩaÏvÙ´mŠ„â]~òjEc9BØ„¾|û¹> Ks(¼g9:릛üŸ…pÁ¤ w ·=€*-üª¡¥v¦Wê­ž#êdIcf Z)“âs³,ôö¡!3[R%´žQ®è›ÝH¦$£Ï±™mHU­=þ¼·ÂÉ(VG>ôd;+îÉwàÉýuyŠåì@Fj'د§fˆÑT¹ëƒÓ-ðÜãéx7 ™ HÅgàëèƳÀéÒ51¶ÖÇØ÷°všt(‹Åê®_õ³b7¯ýVâ,†Ç‰iG“%|H‰¿X)áõfŠåŠ ³ÚAîîq‰8ÖàCÈÑÍWع#R-_F“a¡*gœøD“Ot¢_`K-~.r¸|Ùöc;©üöù½æhé/òž.w’‡“£qÖ“˜Ø6õˆÄæ°].g$ÞFN›œJê¤mw{Œv¥Ô1Éá;†ÝÌ´f0™‹`ö&+…ù­±Ž šì’W{ÏE|ØÖ•°»*XeÑ}‘Cþ±Ï¤ýž ™Õÿ [_ôýªQ=)–úOÛèTî¡g¡†Tßb7¬u€#›mÛ°ÂÇl†¯u5Åâ+ ¸¤o[ ‹2λÍ;ÉwÌÜ!ÍU“ô7‹bŽ'Í"õξö«`)ço\>ÙÑ k#Q{ÂQ]+ƒm®/×O¢>µˆb+X¹JÍl±(áë`wžëÜϺYE³iÆ1:n{cØþÃ<ºÆ·?ÅÑÐÊ£""àL•#¬^Ïý4˜NX^«Ã௩R»0³gý0Ãn¾ þ%ˆjÇV«²ÀŠ`\h”‡ÝÐéEæµ´«µ0Û¸–/œüôœ5˜é*õÕã)EÛ‡ƒ}¸ùν´¦û¥Ç=÷vrþ‰¾ãDÌŠŠL1þwì øL‚Íßlï2® Ü-Ú¹E8yè=w`å²<ªìÍ`×’v{KVO#¿Ð”Q#¾qÆ×Ó‘ú†Û±¿ËÏ·Ä+Œé³Mµ‚_ÅÇÈœÇße½F› ð¯}9h>œˆl7Å0‹ÓŒº¦ÎDR;yO1_'$8[U¹a&çk`ô-"çZZù) ¡0¾ù2ùSj–³ÀéF]´JÊ£›s g_K•ÀQOô|„CÔ"r)¾\79¯‘šQe§¸mFÎ"¤ð]dêà|í䳚%ËH™þ (µ?mÇ^DçDí&#(¦Ñ¤pqPëÁó5úç÷RJ}ž[(‹Š%ñ¾dUÎúPI¦¬V¢ìQg®¾\ÐZšh¯žM—¶#“’­Ž Ö À°÷[>¯mr'̲$«E´‹+qÖ¾ÎPìF_'k*'©¥ .ùÁïR±Ø9•Ç8o›òuK‰Ý„>¤½2®oÚÀŠ¯¶P¼NÒ‹R/,o¾Ü=Æ7eFÿMnNY†L½mó}>ÚF@üN÷|%¯ƒÇ«E Oy&‹g3Oem1ãD‘T”Mõ!^*Ö–ñÕàN‰¦Bäüž(¤ãMë,Seb=¡LÊì½¾[âcÚ²üà®–Ço²Nk‹öùc¥®ÐpÜ´n8”\AájT¹È}¡ð){¦Ì) úQª¿‘rFßýZ\iWÊCCŽ†½ìÛRÙÒ îtðÉÏPLK”lÚˆôÑ>#tOœèp. •]‰ô)p·ÐèÚƒ!=× ûˆ„Ê4Rµ§?†?…q~©e%`l„ôè¯ò»ºéˆhŠa›\@¢¸Úg&çÁÑj©å‹æ #•Âž±Fï;;ÖÛ¿—àSCÌŸRÍžnÁ:eÅV'5[-É¡ÑdÛ°/R®¬ßfü‘q_aªWÝŠ.Ät Ü=i$u¾Õk¡ÿXËmÆ»ú(Èyé ´õÌTKÇ¥%pcé­pK‡î˜aQÅDȉ„$Í- qÒžd&Êeµ‡Õõ(©¹Dæß^ö DR>ò™M%\ãk“™¯¬Ìxø¬bÖ!ÅžZyWz}:ýÕ>$ËqhEÅ/ö®«N6YÝKÎ@=Uú ¯9é+.Ä Ž%Nš¥òˆÞrÙ¸ -Ôy.—göõm^–ü6ÂM‘mÏ\ೇ—úÈn@ûrü¯ÏüŽB ©[Ÿ^t-³¨–¦’–¯s8j{ ]e251ºıͪûš±ìvrv³\ÜÓw–¹Ÿ:wðÔ“Ð1ÐÞ½[¤Õ7ˆùbÎj,ëŒî¤eû–O9ö;T.×TRm?÷nDÇ g… D7t“AÅøÝ¥û >•ÌKÏÅ^UɨÏß:ü²#2ÔÌ-Ùâv¼&[°‰ý–­©?)Äú²…è.Õ¿ºO]aºµ(ÛìJ¡‘tBÅ;"~Ì]žP“›Mu»xXÿ€o+ãð‹#¤Ë•Œ£Ï`¹M©©w æ{ÎD-–L•Óìp\ñuy§ëf­õ›JUOÕŒë'µS ?HPüOF‰.c{—ö…ÿª•Ã£l~(ɺás¯9¢$œYÎJ:±­ävµOm'­‡T5¤DÊð ½»¶špWÔ‘î»ò%BŽòÌH¡ž/CÚæVÚ÷ßín“™!oo2(IÆÇN;0^P„3$Ë£Honñ¢x(¸–ÁèžFç¥ ‹ÑbÏ+ƒuj Í$špÒCÍR°)늷SOòÝ0‹¥k>1tî]a¾¸êœ©WÍ.·E´ŠÑê-öµÎŒ-¬‰~JKäFm ÞMf±Üo/^šKöï®w vf‚Þïׂ§ -t]ضž'r×hp_n3QŽÏFmÔ¤°ž¥ûÊŸí;ôü—Gö¢›8Ú¢²?ú¶íð»^åì‚6oÚV¼m0#ÝÍ<¿5è¾8üY ËÇWë=r? ¢W;ƒ}HÖvMÞà]O[…GzÄ=€ß»å«&k^êwóô#YŒLýºø“¯Ö{wÁœ]M[À¶¦„­Í65ÿëÜû´l£×§¦Ö–çÝK¢¿Ðÿø¹DÇÎ1‚{¦ùÝ|x6‚åFìT±Ñf¿¡§/N}à•¸Ï•Ü¡4évrÓ3ß¹zôËÁkg÷š¯u;] Ëš Zµ%—¨¥÷ågLì§[·ºkAÕ™Ïv‚n%% ŠÉls}(­±a8FÙ1¦Iµ]‘áªr$Fã·EÕ‰,™ÒÂ×gdÎ%{Œm~\ë‘{!qO] j¤Ö’Å’’`WÈhÚï÷Ï•LGœÞ©ÂˆC½…’\Ê Qâ$Ί2ϼ¹*öœÈ£?QºJ.:¸Ü-j¨œ¥Œr¾îÞET”Ùì æmK"ÔÄQ‹_f)ÐŒ¼¶†>[Ì÷ý<"thÕGI†8¢Ç\›‹Ûä…vãÇ÷ ˆ;îÓ$Ž[¼v¨›L>×È´œ}MÅXN¿Oe¿=Mr~J€ÿ õÎ4ô!C~yMYÑ2TqIÓ«áÕƣɀ}c¿&bà©ñ|BB&áóó ýð‹Ùyi£VSIzÜGÓ3²öÜGl‹U7vtšf D$Ÿ„µ¾ÆŽEg³´wd¾vÚ0Ülž[±¶+5é$ÀŸá¬¾Ç <ñÆy²ÁEý$y9šB‡YÞnæW$]ô˜ýò™”¼ƒº"-Th !¯ €-êgù ¶OÒ{©)L{¯/T“—fbΤÈ`U/¦¯B­ÊÖAÕ1{vçõàp¯ò®àä šžXëJƳ“Þj•Þrë¦H…fQïE×Îüc’]¾,QÛo vH?ñ+2Þ _AJI•Ei¾Áå }±£R Xehp ª:eã:gé›*y]<ˆáo×h hsîŽÔÞ8ìvNäaÒ»JðqƒÅZ¢Q¦Émié¤kýµÝ3¥muµó­Çƹúo¿µÇL.ºZXè„— ðé ݨ&bÙ¿HYĹÄj77ÇD+ý ÜxéÂ^B]$¬µ­—5 ͳ×j*ÇÆ$sÞ5lÌã~ÃLl.›’¶Ç‘ãd´$XA–•[Òd?¢Mæ»Þ]Òë—Ò¹^£8®Ï%7`:áÂ'8fTog¨š£id(vy£hcdøÙá`—¿€é: «¤¨•"$Áí¼H>Õ{µú×$ùº8)¥¨ÿ^°ªè“÷˜SM¡xWùdØDÄ|Èa‰ FܬŠæ¬/¬Œ M]Ê쟅óƒ_pb7«Â)“`jõ-w·wÄ=C‚Ϻ7´ÍI¨—£ˆzº4°rRùu7ØšÎ)´(¨ù¶…d}i¯Uø‘o·Å#¯£x;g”ƒÖ´ÉÏÄ@È|/¼¨¹hîL–û¯|¦Éb ûŠèûƒÆ0ú»ÍDššš9=Oq(MÃ3Ó‚.ç8l-Hñ²Ë`4cCjé·÷U¨†Ù§Ï°7™Æ> “’«fÚÒø|.ö8§ÛØy¥_OCm‘<üryïQ‡$ï«[ëê%Í=úì”ø‡oc©·‹Ž&× EžÕ~¬ªÏ÷blnˆ¤ ¿ËwÚy/ *€;k–[ZùDÔ_ýÓ»k‹íW v ö¤ “ïû½.äþÆ:™ôþ)¥%þ„8ÚÄ|T…Ý„Çbþ¬_»  OÒzþÿÚÖï endstream endobj 49 0 obj << /Length1 1382 /Length2 6009 /Length3 0 /Length 6959 /Filter /FlateDecode >> stream xÚxTTíÚ6RÒÝnΤ¥[ºA$†a€a™¡ARJºA:$¤QR@ZZAB) ô}}Ï9ïùÿµ¾oÍZ{ösß×]Ïs]{fmN6cA%¤=T‰@ Šd]cK‰ @¢Dœœ&04ú·ˆÓ ê‚!2ÿPñ€‚Ñ›*ê"À]O8 "ˆHȈHÊ€@€($ý7é!¨‚½`€®p‰€¢ˆ8Uî¾0'g4¦Îß·„‘––ø(¹A=`0У¡n˜Š00FB`P´ï?RðÜqF£Ýe„…½½½…Àn(!¤‡“<¯à C;FPÔà êüлAÿŒ&DÄ ˜8ÃP9Œ‘Žho°Àà0„x" ¦:`¬¥è»Cuþ6ùWº?Ñ¿Á¿ƒÁÒÍŒð…!œG è«ë¡}Ðá𠆣˜x°Ûc¿[êJ†3áŸùP˜;%„‚ÁÍ(ü+ f›Õ*H77("úÕŸ*Ì Á컯ðŸÃuE ½þ¯aÇ_c8xº ›"`<¡Zª0Ñ¿mNP4 ’– ¨ÄYøW_wèoço3f†@w¤;àˆs„b¾ˆüQ`/(€öð„úÿ§ãŸ+"ÀAöP'‚èßÙ1f¨ã_kÌù{À|+†~"è×ç_wÖ†9 pßñ°ª™Š®‘ÿŸ‘ÿåTVFúþ‚¢Ò€ ´‘$%ÅÀæ1Ãþôñ±ZG$ ýW»˜}ú»e¯?àù#^àŸ¹ôæBžý>HÁ\DþÏtÿòÿcù¯,ÿ+Ñÿ»#uO8ü·Ÿç/Àÿã»Áà¾æz¢1*ÐEb´€øo¨9ô/éêB`žnÿíÕBƒ1jPB8a-(r[tû/; ¥ó:ÀÐç¿Xó—Ýô—Þà0Ô‰‚ýzÂ`¢@ ÿòaDqÅPÑÂ,"îò<¼ã¼N‰É[ðã(ÞÊjWt¢e”8šk³~¶ÆÔe;‡ú‘ ƒùŠˆÉÙ&¾ýó1Ÿû¯rC–Ó5 ÐXjýëʇ‚…Cߎ_9AYŒNã’ÑëI†,dvV¸VÜ)æ³—${Ä6Ÿ}*¾êBmÞ“ÞÜ€ÒSr¶t&tœR~×.öcK6ŽUô°å}òxJ«LO-Óñ+ìÇ°ø3åŽX†o¾ëVñYÃïß¿ÎèÏ;*Òî­•¨‰ b ö»Ë–?’<È@“”{÷ ‹KYüT¥"-‘¡Ä-‚kÇNrùûtÑ@{Ò ˜g““žJÈ ý‡oš=â³rù=ËG̶9&ß„Iè[’£ê28”9k°N‡3jÅླྀ´¬^C×ÒOÝidYù™¥öbs,þÃ01›OôçvqÙI"¢Öirp©ÁO•% à¬î#Žæ‹—wÒ*FN*-=6—UÏ‘[B_Ÿöê,n¦ N¶Z^w ÷D"9öYh¦B-:³wûv’^h‰Ÿ¾¥0ãS“@×kÏ!»myò¦ƒŒ òq Ú»4dï)JÏO°'˜‰‘šPÌ™èlbeêZÊS ÔC;ÜÝž íH’óÓ s½›,FÙj©G¥{Y­ ÓØ'ß¹R×3á7ÕÒÞÒ¡À6ryƪžoÿ¥´fŸ~þ}“þ«Å[ô´¹lV÷›/™ƒžŠ<·*Ñ|oʨ¦o˜>ÉõXÿúÕ0,öQ½Ž“Öu,åƒ[Í„_<…NÖ4W*ï©Y“m`[4[žøA¢Ò6Lf†lÈXòjËÿœ¬ÐÓzS65·^nä¼@øÀzYU×}öªoÍiÙt“F{×½ŒTÒæKps U‰H§mòiGrÉ:])iãìá›JZç6Ø• YÞ ÙT¥ZCš~®<3±0Ÿï‰¬1C!Ÿ‡M ´ýµÇ*FQŒˆ÷G-k; ÷„?{½J·*lý>Ïڢé_¨¸›¥‘BygÙ€;þîÓ‘¥ ¯¤ŽFvgÓ2]…“†]…Õw¾ÙpI/Ò9:¹%TàºYŠb±^²¸Í¡XZ´ÉÇ)Ƕ€Ñ×°42¬UŸ7¼‘¤$9ýiŠ“ôa¸qñE²‘ßSªãcÇmÿÈ  Uçw'w6Šü³²Ö”ÓFvf/^,ä£D¹UŽ}—×êlŸM?S´†°J¥Ô#%p‡1|º¦uyÛU_ŽnïG•â¬)\þ.ðÞ©öx™î“+öÛ,>RI‚8V‹lxŸÏ.^ŒÍoMßG¶‡ÚÌq¼àŠýx|dM!OKØxjå %œ¸¶fÛÍfò±/LrŒ½Z½¿0Ñ°JÒiÌ^(›°•véiƒeM¤$~”%¡,G±Tä…XÊ2Y'¨ÉJ¨ï`4yVAe-Œ7*è590¤X0Îÿ˜9FzíªsG -­7N˜ñ¯$Ñ«Å:pæD}«Ü>Ì°ZçVõŠiC7>V›-nŒ †uƒ+²­O¶fˆ±÷ŽìÑ…ð”Lgr‰¤ŸúQ^=¶exFÛo¤ƒ=6C÷3ÂW‰Lg­g¯dKî¥oulxͳJ»ûRÄÓ6i&ûºË2ªÍ­). {"¹2‡FáÖÕïsìý4TØ9C ïK¼Ù¶êÓG%âFJŒ ¶8>½h•ùÙù–w¨3^íÖüV‰wuïÇn&ÁÖfâ×¢XÂLýç—øÀÈ…n¦ØùwtÄn#jï]2öéJ¿ˆÒ û$w¼~â´m\>TLæ ð¾Þ‚—Š'Ì2ìQß™«žy“åË=õù;[Ê éÛ® K½ø+F{<‹3ã6¢óœä l¥Ë«ÏnáíX±cdÛð·0äôëÏ· ?¶ÑÔ„úæQ¤$,zݤ<à˜X\ÀÚ—ªÖ‰û|'Yw`wN¸æ”…v=ÿçRš±` Ò¸í¥·¿É®!H\ú™d¿ ߺI¤P‰.ñel*ÀfÏF«ÙÌ—•òj¸d¨¶#“9µß²GížÜw vîÊÛ#º@)O7}oZ}ăŸð»†æ¿)Þ{á­Ñª×Xn ’}[©í¹7Ç0Þ3¾h9V\¬&ˆjx½Ê0ߢ´‘Úúª ïªÓ¼*´Â2 ñAé±<2ˆæk|…V$”:–vay».FÒ³žkŠðJ”'ÀØã&zB9@,ó?B«zÍ`Ý””Ñ~Ç›RýÝ%ò?]|M²ãBz›ÌÜ?e¬2<({ã¤2Ìtœt0ÿ²-ûš´ßõö&Q*ËAÂ}è‡mIý™×Ë,|{Ò±ã3Z ‘.µ¦{Ú§T>.…mB„x£çüö"u³Ï¿÷jåÀUu ïህœ5"'GÜw&,;Wð ‚ŸE„ÏIc $CbùÔÄoe8D~)FƧû[«TÊsüÀQ'ôsb›ˆâ{ŠlÚ¾îóVPï{ÈÞhÉh òH*¡_ÛÛÐé}œ{.Ÿˆ i¨ÐlLTÞXÏn=±ŽYôÓ×?/ÏÝH¨áà1kNêÏBì¥v‰í¶Ï _À˜ƒ7Ü_žödzã㣒±Å}…,Âp˜§FŸñ~ùÄ\dR¸‡Uzê ¢]PäZUõ&Æå}PUäGW©ñŒeõæƒè »à sÃmN¿ðiú»Ï[ç-Š‹zÄÐZ¾ ¾®§B¸Ó·¢û­Rn³„©ÞR{šìî^«ÎWUþµ~9ÓëÀCÚa !‰¸QL1…ã(ÆWËôÙBækצü`G¶ ß貋#öhM¯t28ÙE–ŸÅŸ¦IÈ;±¨”ìÔ[§Í·b¡Ýƒ½dÖÌí·pïøË"/!bÿtüFÃŒIG*EoVâ’ÎØ éÝ݃£mUXù•Ué¡ êN޻Ѷ/Ô˦‘9Xé·ð¾‰›…ï:NšÆì<0 <»œ û›ù?Ó#®`Ћ·…Ê–þRƸñ1…ˆ…ßÉ])X Nà•\€K•¨8…ÊP¨dÎ×âäƒôÚ¦Œyfò ¡maõr&¦¡ÖPQ:Ú(wµ-[ˆÙJìÌÈÐMjüº1ëæ~´7Ù¨XW¿–òÓöpGi*H¬lÍf»[™Õìä!Ç¡´ZQm¢‹Ñck j+Ïßûz­-™¸ØY‰ú¾ÓtwÃí<嬭œÏJåU”«žŽÝ\y¹öY`XÍ¢â 1[ÝtfDçžú¹r¸ºrXßÞ­¥Á=„8¶^Ô’·«àš§+Ë;Íp¿ 7¬}’bø)+×~µœFØýøZÓñf“”]ê“R¸¶?f”(’»»T†oñ¹iÇûsùÉÞ™éÎÇ q;Z·‚I¾§mÜÍ}ú…›öêE°É5ÄúÓL¬ë‘0B…¨oþȸãIkí^ÚêZѹï-žõZ;ÎEÿJ&ò2¶CÖ\ajŧõÄqS¹Š¿zÆãñÚÊzrÑI_›9Îð)ÇÇs9jÆ¡s¹ ;Šb^rk±JÆtÒæ,N>Ý@Mæ^Ƭí’|Y‡§«™×‡<Ü;êD¨]® Ô)É»£ P,è´= î³5•ûém5¼§x?rmofSñÖÏð´^ä”4ÝÐÕm#jj'y0àÅÕ\;|¦²òQL«YçªÁÛÄÀÓ?‰ÕÑ6^wnÿ˜_0¯¦¢ÐôqF½úWÓsöví±ˆþªŽ)Ë.¨$•¨Š'í”æÀÄ”­Ó(ëäŠ/ýŠ’ÇRK01n÷ìë¾¬Ý Íæ'¥ M™ÆgOV-5YìERšÖÚ5ý[¼Ãl,'³Hˆžp¨—¿M×T]82cTp?Â÷Ìh¥6XæT—kíPŽÄ;]8-ÔО7pnx¨×hÕ‚«5l»¬BphBë·<åH¸Ixg:=5ö2äšÝm̯²šÈì]Õ÷vF„xd¡ø‹A÷˃)ÖôD]‹ÀÇHÏ»’byAä;G^«;Ý'AÚKæþ¹ºÓËb s qºGÖßdÌd©)¼n&>"ϼn±&{=\ùf2at™U‚Áª–¬B›Iq tãŠö3Ω©õƳ°d¹I‹˜7®’Å_Æ{ë¢å}åj#øÖCìUpºÚt¢û™”<ê áÜÍçý¥i—Ç×Í 1ë„`pI¢Ñdاž~²ÚN·'ãXe_xò²o7È^¼3NKïñê9¼üªò°A¡ûnX»p²:»„©PH~û„ÿ¶át8ˆ}xºjRn7ä j{y/Ìy×zu®’í÷¥„þ{¡ÀÇùÁèÝ‘ qŠÃ"³µ‘ŠþjŠ’Õ5//Y$cLQ¡ÙÏf+|7?÷¥à›Īˆ_jx8­™âõ‘‹¿áHˆÌˆå ¨KhêŽ!x˜ö^£rc^*?c‡¤°O×ëã º,l} ù2^„ø²—c=fÝoˆýÆœþ‘þ|c”©PìÓøšCB#«.º‰L¤Þ—[Ÿ~p‡nöuˆÔPÍ“¾” ¼Ivøè¾nPd–GfšP¯pú(bê^ÖmVí(§HÜŠþ­âžŠM2*ÅØ\“êT`†|ÊÎ9 `g¨ø¡èõú‘§jDnJ4ià1WðUŸ”´ÐÕlYû¦=£—2„?H­›_xH¼4Š .}bÑj ?p:!ÔRß”QèA§Î'â›êAH]DUܬJ>TéKʇ¸š_¸–UtbXÔr‡=ßg)¥Í%£é¹cZö|NoIJé಄š##”'κÅÄø%}Ím‘æ@#ͯ&Y¨·Gå+åßo}ëÖš'Épcà y~û\2éµ£öœWN)Ê1y±cý ûñ'²Ê6† ú)bߺù”oˆ5„y"Úq8^¬h îü,°Õǃþ°{Kt3“ãåŠ]h  4¡p_ÉGÝhp¾Âü«Á¡©º¦k¬äŸ7p¬cJí» ¯¶ÛwªlwDïLž¦ðã’F3÷TÖYçÖG1Ü-üQ¡¯XÌ©›¤º¼M‚.mš÷ ‹˜¹ÏØ*l¿>Ù{¤k¦N3 ÿ9ÆïªõQÔ\ZÞ1=°â@>qÞ×|"†Žl®úÙVèRQÚÀ^Öøéd?q–¨`pÈŠfZ'ƹ;Þ’’¦w).éJ³ò#åg¾fò},jŸ±€T-Êg¯D3ÿŽº6ÅúFäüõÑ=ç²­÷$&èa¹ ,O: šß£L Å»KÏý¢îlX®|ã·ÞǼ0¼R&šã\_|Å`#Ü»×v"Z æ½,=1Ânóqò›ÓƒÒ zGµYwv¹ÃŽ9¡W ÔÞŒÊÈnÈz ¡þ·/»v”ÃeîÉ™þ-+åZt*éÚW*8ÂÞuynÔøUï†ßr<3üÎ:©¢sRtœÔBë“éMø¡«”å‘zËrÎu.’‡ìáj)ò£Ͳv¸dýÄ^ß9Ïo.ÖšXíAfñ¬B¡Ø1즌.q+Uxá¾1‡h/ŸOÃ$ZÏ{¾¶ÔÏMØS*ŸÚóo¾Ò°ž‡×ú|à–ð>Z—°x®BÞ­ªÍªÏð*Ò1‡ÇIÑÛÚ±Ô,k¤Çj'èán¬ˆú·mV2%1¯jÂÁ5ý¯©ZæfÑk^ÄD~MZ–/F@ ï©íoæ\³ÊñóOžáÀ…ŠT/¹Îˆg8)}¶ãé¤èŸ(wÝ×\¾û0jCþ"©vÓWñSÿÏVÓÕ ‚߯z5e!þé‡w|•í%+àl\> m‡×+"ߤ:â.uznìÜÇêÅþÄžtÆí¹?@Ã&¿$öÔG×ÛñÔfæY*=ïL– ÿ ý„Ö:ýÙ Qr£òžF2Ÿ×[»7£ û N¼•&’Ö:®Œ¯ÚósØç*á9ö‡§~²­õ™Z¾3÷Õ殿Ù‰v÷(wÕ%~¡ˆ§ØåzùD,-Þ;6>ûÔÇɲǻ%ìn4x¾¼ð²’^´É“(S š,~Îß·P‡*»wÄÍH@]ßZö5óæ}”ásÆ“Ã:c8eKÿ)ÈÀè‡ïÅ€E¾¼öSˆJÛï¾uaLøæåú­Ú¾À:ÔLtŠBN—IÅK-r2hö¡XfŠÑtAt²E~H$jîu ;Y·¡yÛ“0ý«¯“I·ùôý¾· AZŽ] ©U–\—+ïæçÇ©>ïÈíJR¦ZÒg5žž¿>ÜPG k­¨Ç¤ ç*\¾‰ÉZ%±£„õø6Ÿ›´5“Ò‰?G›äLX§?°o¹­¬ÕNמxTy9¼×óA÷mð8 ¨.çåP÷Ø–÷lyú~a¦ZÍ´ãß^0W@â:ÈW=cnæÊÇz]¥”§ÕkçJQ~‚ðõöÉaØëïïî‹+|; §Â©óÍû‘yA|nLyáÞ·íäÅA8üEþWzè§:Ê£¼Ói1¹Ê®ñËÂ=.8ªW{£ß¨¿FÕo‚È|RÙ³*b"+ª6 ç¿'J¸èÌ2Dÿ”DcÓDÜ& œ¨Ûé¹^|2^/9kΤ‡\ Hþ¾d:ú2³PJx¦³_¢]k^·l¤v&Ÿ\¶{ø2N<ñ ¿f†»lέ/âÅRsªÅs®ÍOS'—Ùd=µ/¼x²Hp#ý,ðUõ‰hN—4.öjkÌê±€hô µ†ï&Úhk‰+¾Ù9ÑVî§a¢|r˜±£¾G_Cyà °–T©P ´G¢8uŠ ³Aq×ß ²"ŠÇ-Í-’S“Yù&¯n{CÀS°OÛŒVw5*×ã h,¶²éÖüà¾Â1Eµhn½Íÿ ‚p endstream endobj 51 0 obj << /Length1 1372 /Length2 5926 /Length3 0 /Length 6871 /Filter /FlateDecode >> stream xÚtT”ïö.Ý(Ý1( ! C‹„tHƒ´à0 0Ä Cw7Ò%R* ”„t7HÒ)Ý!¡ÂŒsÎÿwî]ëÞ5k}ó½{?{¿ûyßçù8Ø´õød­–P%ÅâÈkè‹„ø‰88ôa(èß0‡éCÀ%þ@ £Ð10 Ó@Àj®$*“ <ø”(€Ý`V ~€u!âG8y"a6¶(ô6ÿzpA¸ Äîÿ.È:B‘0Уl¡Žè!`€ƒ¢<ÿÑ‚KÒ…r’ÝÝÝùÁŽ.ü¤4÷}€; e Ð…º@‘nP+À/ÂM°#ô3~"€¾-ÌåO\ar#¡tÀÂ]Юp+(€Þ §ªÐr‚Âÿ€Õÿîþž Äúw»¿Õ¿Áà¿‹ÁÂÑ ÷„ÁmÖ0(@KIåºí~Á.t=Ø s[¢¿'”du`4Á¿ô\ H˜Ê…ßæð‹"ðWô)+­äŽŽP8Ê…è×| 0$‚>vOàŸ›µ‡#ÜáÞÖ0¸•õ/V®NÀ'p˜³+TUá/"úOÌŠˆ<@Pˆ-ðW{}O'èï$èWÍÀ×Û á°F“€ú¬¡è?"o°€BºB}½ÿwâŸ+"`ƒ –Pœè?ÝÑa¨õŸ5úò‘0€©Z{ €À¯ß¿ßž¢åe…€;xþþû~šzjŠZê¼ÿ;''‡ðxóD|D@ ú!&.ðýgm0ìïÿ©U…[#þL‹>¦MìöW\ÍÁ øg/MZµP×Dn& "A?@ÿßRÿ]òSø¯.ÿ/‘ÿ÷@J®¿Ó\¿óÿGìsðü @‹Ö…6€møC ¡L«µ‚¹:þwVFAnƒ3H˜_@øOæ¢ó€ZiÃPÛ?’ùòËj08TáûõmAW üWí/ˆ=úûá‚ÖåïmŸî«‡ ¬~ùLPDF"ÁžDh9 Šˆ¼AhCZA=~+ä‡#P蚣/À$úu­h©04}t˜è­!®H$Úb¿€Þ÷_ëß~†B= ¢éIäaˆ]eHãE¹,£;ß×!¼ÅåæÈxã®çÔKo[uü å1g¹gVet©Ú_Þ„Nú0òì_ {˜µe.¤¨ä¢0»Våùòú.OÚl ,V›±ÔÈUòƒæÛÏLqLï%NþkÚx·^ð³ÙeÃäÁÀPôŒ-…+›Ó¼ƒmÌƾŒê76#‹£@š¦¹ê«>ù‰ ‹;—kŒ¡¿ô\5KïÍêîJ:~õ¸ã½hY%=•—[Î@b=uB–Z­ §œÈ7ù7Éñwè Ã8°+'@LLìqôŽ€b&»YXf"Rë¶ÚHuCãè  g¯$Û/O¼—RC·‘úÉZ`,p¯¬Œ ²åYܟмÞ|©\ýuq*Ùm8`lÌž¥×sH7°×Î%Ÿ»`í|¨rKSë*ji£-=Þ©A 6ü”UçˆÂ¨ÉŒÞ¬z%‰Žè<4¦à³íɺ™<ýs :ù;GFëSÊG;˜“ròøL;†™¸ïïa(n˜.kX\©hG tˆ¾á…P.[Î8å,šXÖ(~"oñÏÆ=0GÌ ¯Uo ÷¿$‰Q.×[˜3»6Î|:Vbdx¥žå‡â‡=d„|”Žã_ß“³lÖÁªð^øú6@³’Xª“¿ú-Ž?Eÿ<0ß½î˜1ByT’ánÍVÐt<]øKC±W“—óÒ(Ë}U¼äš ™|¯í÷Ëׄ˜–i|ß5UžSÐ&ÑôP€‹eÈåy £Éæý1~L¾†aUO&ß¿¾·]îšvœy#S̵“ïIA™PWµ8~ dwÚ ÍvéšCŽD°êÉ_"8ë¨Ç»‡ÊåLXºãð¢–ÔW Ïns«ôU]œ9ƒ“ÓdZ±ÛN²ÆÅÙÕaU¾*cL'o¯½Çå…”@ù1„?Â’û™S]Ö(]ÞÂ-9|7`ä&5Ù=ˆâIƒÔv²õ¬,ãU7_¿²½¿Xš>â™ÐÍqÛòÕe½ûòÅ'KH€ ?‘,êH$ÿëš³ã]- 5;¼ùÄìðNșɧaÿš™Sva•Fm·|¾CS.úM°XÏ‚ˆ54E}"v¶¢AÐÂÃffþZý³Ï‚¡˜d·)”·ØyÏñ2Žn"¨ë3á;fC¢¼* Àò±vË›Yë~Y­¹„g²Œ8‚^ã$ms¢/çB<|5ç*zrvfS n¦æaG¡üÇw°.¯!8n:ö…»é7™“3ÀäìBWMgÅ>±©Kg‹M$7o‰ŠCÈßpŦ½?‚ìéÎÂEl¿d˜Ð[{yËD=îg$¦š¦¿b0¾¬QC£œûÉZ^„áĬï󟀬ZÎ?†f'ªQÉ‘»ß®û(ó“ÌWá˜ÎéOÒk"eš’\´2ôvU¬VÑ_êícúí¸ô®hê¬,}0{†—”ŸV¥""Ü9#¼ÉK ˆõ.ŸœNu‚© ýø™ÓOó.J3t‰CŠÊÓµ„¹rI{¶ž|tįÿÌsKŒÀù1°ómìCâCœr®>ª0q! LQII`Hgz»1kÎåÏ&‰JZœªø™Jâ.nX׃úlÎÖ\¹Q‰>w"ýüú(ø=†nÜ7—øŠ"bêKv+–c/Í}Va2Ö£[/ØÊ@ãk‹Æ÷ hXèe&VnÕˆ|PèGt Ãÿ~ãÂå ¤忆ïQkd-ŸYRar¢B†€@³d•âÓQŽÙ‹­ü ¦¾’ç«b?IgbðxÔÄü \rs¤/ËûbcÞû,ôgÑ~û¢p;²o/÷Û.êœñf-°çíÎ,AèVgËà.Û¤ÑQÅ{’[,ÉÛ펜ÜûqjÍdCzî4ÂlBÃÕ»{¥ËC™J–rª†qGªBÃR ê·ÕR¾:ߎ{RfÒ2oš©ö¹Iç1=aLzªøû•Ðªñ|úG±å)½Ô߯gfmð}zË•*žé/µ?H¦|ž"˜–=–Ù0Û~Xv7Šu%»oËòëY°qà#Â’¾8Â;gbœ¦{ |îöÝ_jÏžÝKQÃǕ¿'ÿbù;Ž¶âžËílNé»ëƒF¢eŠƒlä *§"[œrÞ ~&Ð×ÇD†ŠÝ$Ú&¯õUcFe#ˆqT¨¸Ã½=ï®UfÐÍÝÚªyö®!‘)süõ›±Í—Ai›‡ãqúó4”e.¹8lÎLÙIà»+ßÁT0²©öøëdZl"¨së]æI6Áëyá ½ô¯wJ)=ô4Ž‹”¬óOƒÚx#¦o½¨³bà@fjÐÉÍØêx}jÆz(v¶¾‡a4L¿‹ýÁR9²'*Œ¸ò8;teº€s¯ :ù$D„Ä­$Fò*˜aÖA]Þ"%zhå_b7S ÏOŽ •×M£gVOü ú'H0…¬$;ÈÞ…+äÃhQfaå…7~Á-À7¹Q ƒEqû·r‹2üÁ\¹ø‚3š"Ò §êV´®šb¡8~=Ï}„¯¥tŸ¸Ù17U?§®ð#hMíý²9I$¾É’Q<Â*Zá`ãÜÍ}Ÿš,d-ž;épµeüCC7“oë¾À¥K83uÆ=ãoÊî£;®ûx<±7–k,UF¸ïÜˤd¥š=”ËŸî$«Ž'åÑvÙ8ß&lK©m.ñ´ßxiô¬§;5ʈ¹±²€Û;ÕÙÒÕê‘yƒo#J0F³ÜÊm<,-ú‰mLjtí©A¿ã7|\ËRO™ ÔíøÍÒ’v§à„O %N'éÖåaRŠ¬N¬"sò†-Ÿn‘é¿GTŸð8“ 7¨s&ú¤‹ ~²cº2¿*ãb²PŸšï˜L¤Ü_«.öŒw=oLÜ~ò­pB^I¡ðkø­ˆ¦O„¦aŽúä.K´ _UǤ{‹ô†Ö³;÷º"î8]JåÇR®®Õ{z±½ÏÃ$‘î`ÚDسׇʙñí<¡ˆ£Åós%—M³L®¬_f=pl¥Ÿcâ†;}¤OÚ²”JÌÌVù4ê,­øVo¿íÜþ$õFp­p$o¯®¥Ó¹®—"â}´òë tåf2\IÙú³)ÕMÇ?ÃÊÕ WfÁtHÝ;̲yï®×Ÿ’Ò”šý€Ï7BÉÙ8{7)"vj"9>™Á~œD |ËôG²¯Çß.›Î Œ˜ÜdP3N…Ë`Ë<çñäįŽYÆQ}dÇcòz}ô’°¯çûFWAJ#*mmP]ÔÉ mbºÉ!µo HXS`=ýò*CŠ©’AõµoO œY¨ãEÞÎÆä!ã½fÖ —ÕÙß`Äö™M‰k`|[óT•Æh9h¶dãs4ÉBá‹. ÝÅÈ|c¬jM‘¾•pÑÝq‹Ýæ"ëdìÒp¤)½“ž hn‡wxp”ÐÀå}I@gAëyÙ&°Ëê<Á;0™‹‰0©] :[r±Ç+¢ÆôUŠLFÌ$’X$8atWçzÖÑÚn1½îÔ$)ËI¢\%í°ŸO›g)U»aªØ"&ÙT·aZÒÒú©±m&XÕ˜MO¸bÃ2À+™èÛ)³jçSDðp%KË‹•©µÚü…mW÷JktzaÂЗvÑ4ñ(%? jñoWÞ§g®¶_<8Æ„V|ZðYäæ×ÃOñ–eamvuã:â¦]WQ75Cb$5Q6;úÞc"1N~ÏíQÊT³î”Ç‚Z%B„Çô…›Öt‰Fô>hœ-.~ÇKPÔ*À, {]E¾£=‹›T­‚iùðžóçó÷Ç)÷&HDÚe5®ïv,~ÂzÿIS϶}9-¿2±àÝà̧a2Éøµj¦&`õbä­Æg–€Æ—F?É*UG+ µœ2 EZóÀ¯IÃäv8º¿\ lNií [òTšÝݾ%=ÔN1ÛÙí8غ'1èkÔ¬Ô¶›S3g§÷›÷œ;B±T^¶¦amb‚C¿Jµu>]YíGµl®-‘‡5{Üý³“n8O¡Iù~þd÷únšÖÿî¡áBdiÖux•äõ€" a—„º”+;áºêâ™@srEà-yÖî¬5¯™ž–Í’I±‘cÿ˜™·U]ÓgõÕ,1?Îgº¸cy$|®ž¿3¦ *]âz9y‘ƽ°g[Ó^H¥o;eEsC•è·b\6¾ñcÓ4|•ïy..­œ©†+L·£:ÌGH#¦"ÈÖ1€Óž´´ ¤³÷ô+`å’Ưá@óÆgC5L !ovx{ÛT=¯¾¬&%F6Çpi,²cÇɆ© ëog¾½ÃÉÏÌlìýÍf×ô~T~æ.’³…aåVqèGuÄŽj)c£‚FÖ)^ÃÝŒ¯Žsá ¦äÿ¸mÉ †œ^€MŸ:£Ôµ‚cbR˜¶ÓßIIð“1ŸsjÀ±P-!‘Sðà×bQ±ÍÑ?øÏo±pcÐxYd†Ò}o“~ÝT%ñÃ3Q¶ UýÀ0¯tãá¤s–|¿¦¡ùúhãò2ƒêx€ÑÆ°j•òq¢$Ë5ÙQÓ«›í'eZHæͯ|Q|¯PD9nŽç†Øº2bÚ:Á_…¦ÐÅ&Ô¸ *hšf™ðÂ¥ ®š0Çá’i[ù I¯£t;øñ>n øAoýlqÍ7›àxŠìõ;¢]r¿õ„³{¹T>KµÉš"w(üHz‚ôÔ‰ŽUèy¦/óVã¾Iü€7–ÝRíÀ";´ž‚Í(×@$ÝÀ yá3è¡› ;êHJq,ûÕÙüùChÈì·Ù¾~Œ å¡#ûV6l7²Í“ªÓ-”šÚ]&•í˜»cƒàç—åökÙ®.§© Yˆ,¢© {ù¢íeŸªCÛAï¼èùÛÛ¯¢ŸÉ_å^·kØ[ë< '«·ôUç”.Âõ9ô6Çl¬ ´,Çõ}ÁMØ5ÐR‘ˆ®5±SËk‹"i³uOH 3Ú¡Rš¼,f%ÅL‡õ:¼hß›‘rYÃp‘ôÍñmêäö„ã=×+.Ü@‰]ºæ'Ô¥/Jöº¸ø¶Ã·¬ñ[‚ˆšz¶\Þ÷DÝÎ1œäµZd6™C]m£[ÚáÕ„×LŒ$¹,r¥Ym–ìin]%ÀÄÞ쉰ˆPe"åÖ:“Mf%=««¶Â“)¤“¯m‡þû7xîÚVUiÛ­7{¯=c&nQ?u· ´¥=­–; AU€ªY²oû¸a‡íöw¿×=!~Šý¬ŒAwÈ5h/Ô$žphŒÁMcÙ½•Š zŒÓD´ÜzT#ûɧ K™wA_.z+[wýp*˜THkòHvU®Î¡}Ô{ÝÅáåðè\ °:lß?Æ`‘]壸‰× Ó,UÒÉ øL™Iôb?å0–ÖÜ)µU‹<§$¾ËS™GµÄ!¥QÆèˆÖjZ<”÷K¬V%(¸5Þ‹WÂ}QL®…$$)®"±Ì„QŠÒSÔª‚ âóUú¤Ö_&OÂÕ‰Ïå“îŽÚf¹ËÜî*±JGŽ|ñ¹¨öÎÆèÕÿ*’§ƒ˜lÞkÞÉ–Ñ.Þp„F| É™åç)ýúÔ¡²»¥ï¸L<ïF8ƒšêZrX k‡O¹x´ó‘AËQãð@éá™y’óh%Dnr\]]=¿ã‰ÜjÛ´†N/ç·—ZŽ³÷qSæ”ê4‰Ïf³è2D=8éU°’7Ë$ªVêÚVñzCU°¿D0”¿ŠîfQÁ¡­¦C^:Ì”ôœ­Çd3ò’”A€¡ËkNÖ‚áBè~‚q»È\önžÕÄF…Ùš¹tD¢£1Aït‚ T‹këúgo+fŸ‚=å{$ÉB¤‚{ˆ(Æó¬5–¶æO9ì!K]Ð-¶zA¥¯pNýùë")’RüåÔG/$¦lr éË׸Sù…Ô>‹î„Uˆj±bFÕÐ.ƼTqè–Ô\ÆàÓ2I¾6ûØÆ6{ráoV[AŸÓ6‹µªÝÀ™²~GÇ«ûy/ 4eùÊ`R¥Pø¤ÉH³[¸èKŽdu §ºÒc›¾b6äëp¡f$ /©S´¦ó6”Û"£Þ@U%†³³ŸÈék+¡"Ùa¥ìý®ï¼pp4iË®F¿¯Þv±¬îß$öA:+ZU§E#Xvž.ž^üòR® endstream endobj 53 0 obj << /Length1 727 /Length2 11122 /Length3 0 /Length 11704 /Filter /FlateDecode >> stream xÚmvsp¥ý²u0±ídǶmgb[;æŽÍ‰Ll'Û¶í &6'˜è›óž{έ[õÕóOcuõêÕ]õ{(I¥\Õ½œ€¬Œ¬L,|5)1mV ;<%¥8hâjíè aâ ähÍj@'+€……ž îè䲶´rИÑþhÚ™˜[Û[ƒšŽvŽîÖfVwwww7&›Óß"5 àjXXÛâJÊ:²ŠÒiE €4Ð2±(»™ÚY›¬Í€.@Z€…#`÷o`æè`ný/N.Lÿpp‚\ÿ³9Ú¾Jª‹J))ª$Å™ÕÅ&晿å®.|Ñ@3SÏMÇpý¯egò_Ëô¿Ö³ Ï[Lð¬¬sk3W€)ÐÒÚžù_ºÉ:X8¸þ6wsúOê/!—¿ü45¤˜-þ¢ÝììMìqG{'7W ðÕÑrˆ9Ú™$=]æ@óÿ Mì­í¼þ¿øÿAȺšüÕCÔÁò¯‚,ÿY»HY{Í•­]ÿJnab÷WªâZÀïç_þ]ý·àï^Íì¼þ§áßaþiǬ©,§®¥Iÿß3ø'-é`æhní` Psý«¨ Èü¿ÒÊ&Öÿ¾¡ÿPùÏMXÿ×ÿjâ ²öè±0ýëHXþõýÇ2ø_”˜˜£§#+€‘• ÀÊÅÉàeåôû?tÍÜ@  ƒë?²ÿå?þ?çzÍà×WÍøCmÒZÂ*ü%‹æ*¡èçfŒ5èWXÁW§{®õЬ‡¬Ì.ñïÔñó=‰fƒLw/vdïšÉLâˆG“S1¤+øÏ%ä•–þ5?ÿ¨¼ÔG^}Ñij7Süryëš±±¬L7Å‘•{V#ç\&ˆ&Íc\ÞÑ…º°¶øà Mk@ž©ûüE_:gõ­ðÊTøyo"v_w;3¶£õ‡;®yoK¡¿ž!ÛÌf*Å:K[ÕÛýó ^Õ’øìe²—=פŸ?Zw˜m´ [ß7?Ùo›¥TÃý°Qõá.L ¯R¡/dW¡ðÁÐ’ i ߥ{§n“¯ÔQ<ú'ÃÈÚ'Ðs¤,¥z÷êt侑lyEÞ8à„}ÛïŒ>#€p·B6{¢™ÿo+[ÏŽ=èd]x뽎§7ŽÄ*cù£KŽ‰x1ͬçå[@ÿla¨\‘°ü kÓ$¯U¨Ò¼¡¦À*dÀúíK&©`Öï=Ç/†Æ‹*tR ÍL§–Ȳt`®Ý<°Ï’é⮌Ãô..·<Ðp'Ê¡Ü"d8¾) ÐzQu‡AR•I{nÐ!rF7e"Wõàý9³,X8«¦¯ÅáÅlâoÐÇ>Š÷%9c¡¼ÒSü¥ë³Ø¹f™³m·6^4àÔ!¬üßvŠ”GDôôƒéȼqˆù/Ib%¡ŠÃÍŽ¦ÈHõøV>•"ÉÚ'Kºý–ÊlI©Å¸ØBH÷(LÀ ^A²,'¦Òëç[ –tŠo0-°OBÞ£Ö´=hGßä÷7\ W§ç<º@-ªÆ}±Jr£ô¢n“áL|¤Ã»ù#LÏ¥'X=I[8œ²QXÒ¥9®Ê¹2¾N ‹¼Z 4jBÅ!áªK9]>4!æ†ÎîÁOºoë¦?±{­L}H‚SO§Yßù«AŒ!ä;³¨¿µ ¨zËUñM\ðÜ,Òe•'ßz·—7=X ¸’`øÚï•mùj™ÌjüTi‡Íˆ´Ÿám Ôü¯ì{Ñ­ÉîSEpdç|bÛ{¨Cd'mÄ ÷ö„ÁàÖ‰üø ¾Õ j·¡h9Í"T:ñz~ÀÕ&¨"SŠ©¿Ö;Ìåñ‚é$c9..ÓLÀó3GÍ£+ Ðu/k™t”»ÜΓQÿA/Žn«{ò fPÁ˜ #ÛNžâèñ2·Å«òÖÆKÀK9¢¼½CÖ¡>ºõ¡fAŸÎQÔýØf"é3IŸ)ßÉ÷¬;Бђu`52¡]˜”ÜBs>&w ÃÝîõ¹*S,A>·h(Ztò¥¼á²I9 ·M’ðXmòq~ÒAüc†·(EŸgfŠØ‚X\茱·ÙK¡QÅ·QµV™Ò #ÒVûö-êW{qf3:Á9í>U§+I4”¨c¬MnPú8â›fÉ ‰õ)§%¼xÔîhÀ²Aÿõ,îzr¨3bÔ:WåÃ55î}!xõì1+_œ¯–`“KŸ±]tðwÐ ‡Æè†R¼ªÕxa¼ÜsÍòoù÷BCÉÆÊX€“ú˜‡çš¨!g s¡à7Ú.ŽŽi•¡qÑ^X~qCôPµñ®$³—£t’z°Ž’U(ÒØŸrÕ¾le3Ø7 În2•Ž‚Y8q[—µ™Ôn%V:U¢Q¿Œ›‡¡Ç°ö7;¶(ÞÒ¡u07ì%¯¶¦tøZv춹V|›™1,üTUò'«ÜtÖfú7œ(F%–¾ŸG¹n‡Ø´M@Rƒ½4é<&íbéÜ&}ð™õ*Q7‡ÄÇ[HéôÎqÈñFoê,r0¦­r°­RvÉ+ó5ÍH» |§?¸×ΉAä7;шäY^KÈÆP[øJG¶œ`ů'åm4@|Ô¸úaÍC˜ÅàÎB3ÞØÃð?õž'>š+s0‹°–k²Å^“tÊx†‡D3¼Ì~.\o§°tJ*ÐãÀ ýÎOP3lÞöûÅ1°éæÁÎ5¹M–¸­á(’ìIöÄí÷áñìoH²®CCÁŸ‹ñaUóëw?"¶tyIš Ch½iLpuc3$K–×¢ k ÿEt!ë6#´¡›½evF‹)oȵäÝjÏ] Q©y–PÒD–µ g‡A3(RªaI;ˆs„¸æÆú­ õÿËà ××9f¿lÅ—4§-ÊÒïÇó_‹¾ãᇊ>´%Ü ß ºÇÝ‹m¸ÖÚ¤®ÈFYŒúÛÔ¥@PåZ?èUäŒj}'àQ[Où-ÌÖÐ'5à Ö»e}û{è¨ ßLõÍÞÓý,’ÕÀ¤n¨È²Ù„=É.ÌëuˆÍÏy’JìN¢h`Ka™溜¦ÚîHþ ëÅÑ^„deµúWÆð=—ºÊ«qòóÃõyÅC©!BOÊÛNÖÞP‘Gæ³7ÙjDzanÄdš‚yÑV˃“McÔ“Ñè(À±ß{¤Î«²Êþ:ïŠâ#´^–wôŽ3¸Lî^t¤²&ÛÚØ»EÐïß¿9@:ôì5GÀ8n]=%(.îê¶ó²$»éŒªBVö ¹ÜÚ!P|H£õv£zy+ÏÀ(ܳ4K´‹¾Œe¾­¤•¯};Ø›؃¡qN–T“á^-ÌÍü¦ÉBÐÏÃI}Κ´œ!²’ÛËßÙ6²¬(î0˜Á¦žÔÄ1Œ¨ž¿Ü8Ì?ë”$·ˆ¯¥“Rµ81j&Ì3nÚ¬f½ÐIïþSM—œí²3ðÃ÷ ,æПŒÊÖÇ^«j¥åúg%ó™ÄJˆcd©ø¥uwo“7'ë""±x©Á+¡°~Ú†£óh†ø9Íÿÿ»éì}ó¾GƒÈ7ïëv¸5d:i˜ýúþÑ`I’v]bWÉÖ¾È×l8èPºˆÏØ–“ô¬|¾wm°ýÛÌÔHèñÚ<;c„€¨í$³Mì1KÚ£áí–¯¿I>QçCïc<U"/MMå*²›sÅzŠZ'(a9õ5nk\’SÛ†ðŽ²ÊÏú8’Rh¸éàyIâDèXêÇ/èª9Ý)Y ðgàd‡¦:ÐôT¸f$è:ÊUç½mýW´¼+šH ]Ô¿°¶Á_85fEç…x¼„È‚gòUSKyXd)CÝÌö2ïÃpÌ9Ÿ0:ÙÒôϬ¹ÐŒùÏD]®ÁÌ–t”ËYc4…ñÏrB ßìEÙ÷ã̳B¨¾Ùþ`0GŸÙ#<¦ë-£¾ãÂËn=© “©Iàò{WÑðû…*›•ur›¦MbÙÑFw‡0© M÷ä/§Ë ž„5› :ýœD²x¡Ù? ΈNÓ¥øe¢„õq‚­ Íp=ëE‰1P3X*ðCƒ¿…þP¤ÿYåa„ŒvÔIbû|D®Ï`\¸}Ít5‡M>€È4* ;+:±\(oê,ֈɳ¤rÚ´vš‚Sz³nXiÒõÞ— ô–ܺ‡¤!Zr6×iíYXjrÍìäØ­“ Pü5n~]ó•¹À_ }Ÿ%ÜÎÌ‘µb·5¡Ë¦Ižú1°ô@ºåÛz;7{;µòfOrñ©Xÿ£ˆ‡‹ÀýŸ:Ý7(F´Ãæ×y:Kù)žÊPBz„_›ßv]ý]_îÄjB„ë,’âÏÉfêånr–ˆ~M×ÜkO¥–ùT·%Ã-Ò›&;M@Ìe ÇFè ïÄJ!$£ã«Là/¼që}‘ü®ßžˆù±[À@¸f™}×6ÎØúë=¦ò?óü£n•×íxteuÀ“azªÑtNß0±4õùå' ÜÑVvnÉ] q¾ØMlQÜiµ0©‡ÄåOïJUñ‰½×^ýrQ;µÉ”ƒ¸ýÑéÔ9®Ê[’j‰ûÍÇ\Å#øÁ_ƒqG&qç<[}…íÝCï;˜q$×L!ƒh6¥¿’Ò»µgiãݾÇæµ& {«f‡]-¤NˆZò…7ùó€yÝM·Á¹ug:ÇEâiüpwŽìá¤fè­5*Í!T\EÕ“F ,£†-MÁÞï‚Â%Lw¾S»njkÏ¿& v½ÉpïX«+­ð´ÂD÷&ødÃ.y²fÉO”­xô“UM-&µÇ #~w¿âu;)ÎÌÆÍ¢ ñY1ë¢=³Á²kÀ³Š#+£³Ûuy›é•¬ØðÀݦuwöÔðˆŠáxá°ÿbµM¶1$O½¸x}mcbŠV?%Øã)t܉ vðî F^¡nÙáÍ0F¦ ,Í£PŸ÷Ÿ{˜ýÂöBGZ‘Uý€s¼®¸¦{afëñclwõp‘Þ,awp¸&T¹®Å7TÆžÁ °}É~³B5tO诖 â`úº ØXµYP¬‚*Dõ§ô3›nÃÖ>yþ$”·³cÍy¥Ù^®½6Æoy::kb_Jا°Fgƒh`‡{ü‡}òHå3Juó>­j7<û Œ£c.Ô3³¬9Î?@8vŠs7n4¢!ã¹¼‰1Ääቋøzˆ¬õú;! <¸l½UD¸¡<Ç‹ô ukë“@iVÝoe+½˜! /Ò`*yžb€j†TÎpAZ«_•Ô— +Ø‘«ÓŽ(˜‡ ^À–m:»Çq•Ñ)²nj•+JCW¨`2ÜÜ2 Ó¦û‰ÛrW6RµV¨Ý·-ÇfäË8Ëb-mün0ñWs’Ì›WGÉáѧ‡¼^¢pK>|žve !T.Kçlž‡o„UUÛÙ·vè¼çê,{/éëgVwÓ¼7¨OÞ Ñ©˜Mt­kÜÈ‹;µÛn8ŠžÝ;8nT«|çTu.V¸Ý·?\1çpÇN‹É#v½¬3\î!y+¼(nS#:›m*[StËþÔ9q¤>ÖÊ4ÉyC×ÏÚöå×–§p“DFhŠ¿²&¥mÖ•Ü¿Ÿ¤áqö¢´|uEðQƒeåêÅJvݳfÔôOÃ-¸uÅ™v³ßBãDgÚˆÄb\xN z²vìß²=×ð ×ï›1b7P–*UÁû *ËÝ]ìËÜ6Ú2V:7Ðm££RûvleYž%Ëè˜I—©Ë”E'¿ÚÇ`<±Ï¤§×¼¡ßà›ÖF*l+5° Ù¨¯®#@z{Ö…„Of$3“W gw1ìï d)ûr6lCô×ôP\áUDY9-ãy âeµsZ°¢ÑÄ®$hÝ«T6,Úg³ֈis‰¢aߧÈV¸kÇå¬óY¥…$ašæ. {¥QSïT[º«$i‡Þ¶]ø ò/tšû´§ûXÔŠÐ_ÁêÅkISy!7&ÉM¼…FL+àk^”0Stk/åV /// Ÿ‘fî Î`7 ×}/Ë{(œ•¾ûgvd¤(Æó>ÌŽ—ÕîŸük"ÊCÒyaÑÕÙ‘|¡¯ÚEa°÷¬XoËŒC˜ýdÃ'eä'‚ Â@1bCK†™ ’•e…rq >¶àÜ/9ä ˆ©¬$ȉ›q¨áeï˜LT¸Ýlé/®(ŒiÖ´˜s&Š‹¨”eœ±Z½Ë’ˆ›=ƒYÞ÷=†cDql™{kb³ÏÇöÌTÐõÜl© j;ƒ®Š þž6AˆºÇi®KË–èôÛÍ]ŠŸ‚¾%”*?Q5­:kGÙS ºå±ì»q¸²¡º C¿òÕ¢Ï_¬âö(@«¤›~È©!w SgëP/ëã³Id™³é8hÒâ<¾p¢‹}Ô,Óö™[¯ 9EOMAžÁS¤(E`v²Óì)ZW½\q`Úéâ"ŠâÍûj"ÎÕ¬®Ä´hmC^ f9tïðjGƒù|Š„³>&O}—¦ž‚×úzõ\Óµ’×L¼Ç¸îõVÈéœ(%á(ÓF•S:™µ¥ ²M}Y¯ œ§ pN[h$“ºµƒ³ÜN,Uý¹"ž®§ló±MÙäê«r4ªM˜'yªî¯³šÁðkë¼ȶh™aâœUn¤Ç+w„‰¯Qí<6H?0 l:û»Ÿ§;ÛËÔ4.B2ö5Uyj×1âÒ òÿYø~iÛ›BH½ÖIþé-öÍõŽ¨|O!7 Tׄ\ëoÉo¿˜¹OZðøg0¡|1¬ë5aš9Žü•“19„:¶5æ‡^N}¯ºdêI ºŒ%[æ >•p}&Ú;¶l6ÑFPHéËž„Á²ùáâ%nËßÿXý¢$<®µ——QŠÉ@rŽ!¹|⮨¼=ŸM‘iKŒ0?픟ªènkÀƒÄé Qæ´ü-îÐ8bRA‘m›V!‡çNƒÙhÙ£1XMþÑV_ÈkžýwÊÚ =g('#wbø{<Þ‰¡^8#]øNæÇÌø¡=ké´”^9c¡ÉKT¼ÐÏ?O^5ÝV?ÆrÀP6u %¸Óâzõ ¢ÐDÑ‹@_뼩Qd@-õCìŒèö±§5ÿüå~ŽÞ ™(¤Â¸¦´à‹á¯¸,àQ„Ëkžº6*$>ÖxZð%¸×ãƒJcøä¼DÐ/…J>ÔÙhœä®KãʃRïqk Ôv×ðd>JÁu¨ ým8hÆþ’)“[†«ûÛ$˜ÔGKŪ Èî [¸Ã¤Ú¼!×…È-Ã"|GÔÈKjsRL$cÞS?&O$ÝTë2˜‚¾É™å[ wÇŽ7›1)MJ»ûáôggÍCný˜?ÿ“¢°¼Ö/ä¤ßŽhõ>.ÅÉÕ­/:ÁXé^Tc/ù·è(wf¬Án¦ñqûï­ÃvJX_ân¾e—á‰AW¬1­{}ãVûhi»ÐJx.PUˆ›¸esus ž¢ûHÞ±º[!µ`Qçw‰§´ú ¹Ðd/"¸)kJam†»¬Ÿ4×^=ZLòÄ‘ŠèÑpjnëö•?¶Õ¯-Ç~âÒ _.ÆŒÈBRa5zqg{p"L Z. Oï2 É«î8æ{½%P‘Šë@¿y?˜›2-q‡¶C>>C£®ñ„'ÞKÝ7iÁDçy$ æû,4ûD£PÕŸíMÂƸU±K/ ß%²óà Üå9.©ºó?š ¦o 霓Çõ=É%ëë)<KŽ£ºÔàêêNNÜdCó  öÙ‹ æYçh:›½0©ÍC8žìélŒ|Í 0àíðÝ®•½ñ£vž•à¬2$‘í«û*1þ§ÞÄŽ’õugR ÖzM>ùá‚CúX¡ä˜¬v—à„B5daTxùúï„Ó g%èö6¾Š.Þë¿2Œ¥¾ÉÑj›PíKŠîI×P*kwËèDntºXoä>žÜ³ÀGEróŽ˜^o¾G+èýÔÊ8–1›†OWƤËœšI]}_ÂyR²P¡P…Ä÷ `­ß‰m+ž¦yM ‚*”óÌ ÿñðÛkÀ¹°ßØÚÆçTé‚Ìà.­’³ñØS±æjz¹í&FÊü'„ð\»&Né,å…@ÔÔ«yF~mX9Û¥Å;H¾÷…Ϩ#E£æ§Î&”¡:B9FØleKz&—hfÀØàØé–Vf*µ8kqatÜ›vŒMºÄÀ¶)c¢33CÐÉÖ£ë^~†âfˆ pâÆש@|"Zï¸\¼t~†‹Ÿ¨÷j­hvl< ½ûÍîlºb²&še…gèü »2A‰T$F„v³fZ/Hó×F×)=˜¯ÔçÛË#x¹uBÿ*xqú8$>YçÌ™0³Ë¡iK‹µŠÁ*ÁÊ#q¢èKtʳžBé@ƒúÇӻ孾oÄ=±¸‰<ÁNˆêƒ Úouþöêph}6}uù1?ü+÷ãnƒê8¶uÙçûªÏ‚a|,°M?IÍîÄ~y2_*ÔÉÄçf5¡¾ÇEí q8†r&É^6@Ýb‰“qL„G >Ïûýºn8¯æíôy&Ñ“œ—SãõãJ·i‡¹ç¦¶eÝÕ«¨ŒŒš³™ŸBø í’KeJ!À77¬tÕxÛŇ ^!Z•O•àˆ¼ÑíKˆ¬J m1ù!Ï»µà™©LÁ©SèÑ/ÇìýU£k—õ¥WxßÈÂ&\9·¢ŒÓ/lw z|GÕÀ »É¹¢÷¨¼ˆZ `’6ýî™F§Â¥S8w÷æäýÌ¢Q™×'ÛAÜk-ÌËd³F¹¤-µÈG̾ߟ¶¤O_Øã9#+rf¹—SÚi¯hu_Tºƒ[¨ØþÔ6Ð˜çŽ cŸM/Ái:ê@–vœ™ÙP±‡¢Z/+Ÿ?¢&Q±¢u•Î±¿Üiˆ<ý¹cWÓ ÖRÉÓz‚OÑAméÈ`Ùºñ0/Ñ%ª“RŠemE=M£/QàE€‰bÛwÑS»o|àYñ¿ Ù‘J)¿ö“"0ªë0MñQæ«Ú¶†gÖê´!×Õsȯêª* tö¥U "‘rUí8f<‚/}ÿÄÅeîÙ0„ˈìè;F’«y•2ƒï'ߺú3ÊÛáÒnIÍWØ™:«ŽX¯s϶)ÆÜ"”idaЧÜý»ÀWIp3 ˜yu3²®ñO› H·)[ˆ/Z¸›—EFRgLù*ñ_<g¹&%¹‹%€ƒâ“#cÑÍ `ø@<kå«ÚN²Û«2{ÄRìúfšzyE(³"ÉÀᶞØÁai‹ïd*ÈËPw—H°of?J뛦_ÂZ2ºC!?3æÔÓa=»¤1U>€¡Ýxp¯YßÅ^6Ö…JéH*²?Já“z(|i<%)iè6/ÈáòRŒœ?èlëQxr62¢¤:òJ¶sPïl>!¿ÃfÁŠßm‹ÆxùFa—ºØ¥ÛÓ ‡-Ü›c«%a­Fš=ö´‡¯Fáù튴é¬M|UñØCb­Nò,ÎýÆxËqf Dì D¨U¸û[ýí+s‘á‹5”kŽ¨Óõ7ÂI[€ßÝZ—WKá›Å„lÌfês×WæÊxÖ`¹È¹áË/[ž*jèÕÏïQÝ=¢ûùp0ß¿·A†½èõ¶_à2ž+Í“ëÕÎRÆyßšý§;£¤Â[ öšû±ÈÌäR„†qUA.~b!ŒI“l¯hÑj.O}cî§I‰ '»?ç3@ •dÄ€Û %˜íݳÐzC›x’UÛAhm]ŒMe†Ôa{¤è‰ïžZÏä-´<$ˆ–s¿Ûöüõ5’_'•¶“O ƒeH? ‘uùáË “s~žJ²QϽEC&™©wzÊVߣ^°æa-ªæ'Íò 3n‰î’¯JÕ‘²éD°÷A¦ðßßIÂ)þ.Eð2ð¤0‘„­¶}pÉV{dø£¶]v ¤ræmõ.ŽE:sû2gÛö)™nŸÚšÏk‚‰“µPäRÆAåÉ®=€2©“w䶟G=Ùé¹%¸íÉE (Te^pǨ/iZü €Z†!…¹ê°nš§|ã÷|… >ÈèCíYgòzÕn-¢Õ6b?S®ñí"û!ªIÈŽw5&>s¡P8°^tÊ‘ÄŽp-nÇÖÒÔ8O¦hÓaZ&0ç«­F)=­êUm—~{õa”Ê œù`]aœ²Þª¼SO ÉÖ¾JŠ}ËŒìøåóVÅÔfSó‰ K*¦Çæ2óÓxß9 O¬]ÚTGƒqe:ßVôí“èµtÎfü}œïU§® ÛÒ)ñºøà*V¿ß°øzØQB5NÀhò É= …NŒH…N¶³ Y›7«Ífð‚ÎãHƒ>GÆDMyŒØÝÍ~r¨#K`ªÎ¨SŠ57×K»-Ô3i ô’ÊZ³»Lãš‘]ÒÕïNÙ‡óÔ¬gtž°:Õvx¬blg-Ésó«ˆ|…Iµ—„x•y¨÷fíñR™Ûœgû ‘I«Cålefz¼¾—&ŽÔçÄP¯'æÅ›ãÉfÙ×RÓZ#8Ø·ÁQ#l‰LfÖù–@>f>NøÌúu„Aé0_ÚÿŠ‡{€~cJo§/§+ÎDpµÙgCêÅjƒ…ˆ•`QÇ •*nòÊZ!>gCñ¬?ÑI!ÑôŸv‡í†OϯA„¨`û&_ØTÞ4µ#6ÕeiÍ;UJÖÝãïŸn`{#erzz“Ët7åÒß™P£^²?3h¾{„;’wÞ ˜\ÈpÄmê?åŒôNµÈ­Ý?ž>y³eŒSI^pKÊç2u!ƉzëÿƒOÖI»SbmSíáYöÿh·uŸ/4þ€a¡[-2ΚuÀìÏ7:«Ì„¥¤µ%L¦µ `$eÐÀšD$ŸÒ•Q£Œˆ­j]ƒ0ñ­ÚdÀßsé…$±ñ$rïœçŸšt{³=ùTRåù— šÃ:ZÔmÞ~¿}aX!ír&ϳZ+*×è\€¿EÊÔ ÆOcÙѬ­–0ÜZµ#”ÑUY®º£IÁ;Bq>€˜t½S]<ªå0VfPÔ“E—n罫8÷5m%Éò¸h‡ Ä1)Â¥¶Jmáx`½ëžþóÊÓ÷Ì›§Œ¨\ßgJü¨Ì¢Éï{Áº†_Ãĵ1ióvÞ¾¾½wÌÆÖÍ…‘ ¿¾8¦nñèë¦VKp‘/aYÂÒRX€Y³¤f!Œ9´2¿z>Š]?^^;¡Æ.1Ïr›šM‰åh®£÷vh Ï‹Ôµ$:Š&¯eµ¨Ž#¦©Lp¶’c‡Ü—‹,pÕð•„äáÀ\¨ÛœhwÐ,ë˜Ó×pQ*Fd>G˜9š'hg?,ÖQ¿5½{×éÆTAxHV}gg›fh÷4õ`^XqZá|µÎrz´y|xZŸ±ˆ”½I$U FÆm×ø)g­ªÔx^Ù~ö¬­‰ÀŽ-Tütkq ¢„FÕÔsDc»›ÊþãìÐmŽðÜ•gÏyLnusXÉDa sÙr{ªžè³ìœd jgÔ àP\†¼\„Ñû¹HçØK;ÙÌÙ4Àø¥Þýµ±£ãA„ý›âêІ_!G{ýœÞ™|qZî6âûá¶[ò¬^nŒw¹q´8É}m ƾŷÀØž}´æ×:dsö¹¡QÜΘh<ô Ÿ±¸*WºÞña>Ò°]zÙ žé—çÚ_¾Ó‹± M¥Üv‡×L†º­”¯Ðø?´C/5+$¢®@ZSÂ$ ˜¿bØžÉCŽõl>uÓÛŸ#Ëá–Ž÷gøU¡Õ­ PøººóyùÄP¥Sñ™ÔÀžÏHžsæ<ÞÇŸ³ °˜@:¬X®µì¨…Vâ>š”MéJ×Ì[¶ôfuÒkV9ÌBàñˆøþHx"_Ç‚ÃEÅJ†z79“B ¾Ô±ÑýÈ·KÉ¿$©ï{J`“ã?›Ðq¯½Þ\žÔ„Ô ;•NÏ­…ÐDN:åõKe·úWFÑÆŒ¬Ìš¯ßÜóœÙ2Í×<й_³C[œ~GÍjQKñ“¡¢¨X‘?rs’¹Õï‡I\:T;èÏ]Uº¤"þŽÁkŠQ«ÑâÈO+…BKëÞßš•ý오öó¡¹ãœÊqm±r“>ôBüøßCfL3ÝuTéTä–,Å êªÏ¥k)ý»*JðdyL~B»­ùõÜäòÀ÷n½Lœ|/ ¹>6ð¾Q6WÜÍ?hìÖ“a]¥"/–ó·SM†ÏŸhŧ6y$¬mX»Ü7à™Îž”î¥Éþ""°ÂBŒ7B'–“Á/)¼g&bŠá×ÔEåõ¢ãCô2ü§[>’ü¢¥}J ¿ JyÝo‚|¦?½Ÿ Ë¥’·Àzr­ ½¹%5(þ¤Õ;³ƒ„Ö©[Õo”u„{;‹^4ü‹X†RŽI´Épo|ÑN¼Â›3Æ1§˜·`è¦E#i‰(ݼãlGr£®§ÓÁj|ß’ùÉõ§¿GŠ×œ­òÞ@˜ˆóYÿÖ:ú,%AÖ%ƒa"r¨Ô~ Ò5ÒãŠüçÚu|{ N™È ‚…n‹M« ‡Ý•Í‡P#¢Aåð3ˆôúMØqßÚÒ6ËøilåK½¹‡‚jòL)_#>WP2Ö¦×͘,#ŠíÃØ©^¾Ìï`‰#QÄüÉÓÅA^ꨙ§(˜ªN0JS=ÒÒ™³‡º\e¾g&[N¸”ó€{ddÅÞ.Z½l5$¨ßOxëð0:Oî[Š³ DCm⩪·Óx“ÕH{ðûJy wO\]ÔÖ2/%B†Mví2º§LC%§ð“ZuU–ÌDZ½BÃ562J¦¡œqų`Ÿk³ÖX\¾ãÚM7êš zéÄ”.±¨F„H7bLõúÃV‹p¬ŽRpömsçjÙ]påFÀN›ž]\ÍÖa%­®ï@¨÷·Bl¦Cô§)ží¦zÍšš3ˆ÷m¥ï¿_߸m¹¯?©Vt%ï ¦Áá]œÔGrYá_ lØGúiÞ½ø9Ú~HU.¯ ÔóØ\€ÛäÜ*NÌiNV1t1h@œ]öaÏæ¼è‹09åÒÚìô ¶ž†´,6Üßóß¿C÷ä‹ýdÅ•9xCÙ¹0jàt¿fø™îÏ ¿äi„Æ }h`Ìæië7újŒ¹î. wÜE½“OÎBxSé÷I)¬mck4ߦæg¡ûÿb‡ý endstream endobj 55 0 obj << /Length1 721 /Length2 16749 /Length3 0 /Length 17347 /Filter /FlateDecode >> stream xÚl·sp¦ß-ÛvÞØÖĶ1±“7¶mÛ¶‰mÛ™ØN&v&ßïœsϹu«¾zþé^½ºöêµ»ê©MA"aoç¢æéda`adæ¨J¨È³033˜Ùà((D€F.–övbF.@€Ð  t°°X™™Yà(¢öžN–æ.jšƒu#SK[K'€º½½›¥‰€ÏÍÍMÈÍÙ•ÑÉU€ñŸ&U àb˜YÚ¢ŠJZÒ ’jI…ŸI ÐÉÈ äjlci³4Ú9iföN›ÿ${;SËirfü·;7 “Ë?ÂÌœìmòâjŠ jqQ&5Q€‘)@NêŸv;gžØ@'ÛMGpù_dcô¿ÈøÑÿªNÿ‰áXX¦–&.c ¹¥Ó¿|“¶3³pþ6uuøoéAÎÿèPÿã! ÀhöÛÕÆFÁÈ µ·upu:äíMNv{[#»ÿ²Œl-m<ÿyÿ‡!íbôÂvæÿ8ÇüÈÒYÂÒhªdéòÕfF6ÿXôo\øŸ{‘šZºÚþ§øÿܨ©½çÿ9òŸ1þ} “Š„œ†‚(Ýÿàßeq;{SK;s€ªË?^9™þøwYÉÈò?Ûó_1ÿÝ&ËÿÍå\œ,=:ÌŒÿZæ}ÿôþ/KDÄÞÛå7€•ÀÂÎÁ àfãòý䚸:9í\þmø?³ü7ÿ÷"@¸ß«ö&¼!Vé­a•~âÅóUtó³†?éVY@×fzÿè Z[˜]ãÝ«á¸#Îóï]íJß·ÅN¦¦cH"ôVñ^KɪÌýê¶ß• ß#ôo ÔqmM þDÞ¹dnüR¢fÏ먓q,çG•d‰1¬è€DÙIX_z´“¤Ñ£BËÒþ‹< «E™·-º1|Ýÿ ^¼‡ð‡×À¢¨2Ðá(\ëÞü/¼Æ®±]Ÿó¢{–ŒµíÉ1îqx$wIÄý_””rKŠòe ž¥9ö#J‡˜[;¿¾ÏÞ'-5;’ÈGœ¯”k+•¸H°*ä'ÚÃjhÖ[k“ƒèö š=‰Çï(ëž ]Z\FÊI-IŠÂàŠ,N)¿ªýú¤œ­chÒw-J"yÕÑêÝY$a¤5‘ð‚{P äØ4—òM=¾n§évò·ýÙ˜ºMžPÝÚGÇ—êß ·õÎTŠËÂݸeØ™¡v’˜ªÂ«Ê6“ƒu5ÛŠÏR' ¿°4[á]¦vÏsY¦ `UÊ• ¼—¡¦èF ;šË´)cƤpMð÷MUEÖ—¾˜˜‡è)&ô­¿ˆ~’½ë#õÖlkÁà Xöê«îÊ‚¬õ¿šÕåó²Ä’2±zŒgbòF$¼·ØO Án6EFãÛ&2êpgñw~-lÙÓ„½ŽFs~GÄ\œ‡YÄOj,ÞaYeÇðçærÀè ,G^Ö†ÓB;¾1Öµô ]M*ª1™Að%"ï91/‚+¸ÄØÍ7žLtG¬câ˜åú<¯Æ|äL« Û¢ÑÙŽ2\ z§iÍùé½¢í= Žâ*ݾðâ$eþø¸Ÿ‘ÎC0'¥€°ííÖJG9&ŸGy¶”Ÿ¶ˆr3¹ïÎùWn œ;§)>³(x³ðå†ùN+é<Å / &7dëÿôqî)ˆ®aý‚³2‘ˆ·â›Î`±ßÍR.'Ê,¨k= 9¢`K»=”š8‡d\+¨BÚ¾X7†«qKçÕ—GºIyŒpTÅ=éÂÕ‚2Bž`\ ‡Š[ìï,ð²9Èé¤㟤èOðAÜ`è.Fp0ˆ¬Ÿ}=_-ꇴgK)q WW8Û—qÍ’@ã{n"¬aTÁ‰ò’y«…"͆Ó&’¿Øàü9nB/z)âÛ§—snùɾ\;]J÷O*•Dª‹8Î ½Taq Xõ×ñ^/üÑôrÜ;Ý®HÙ®¢¶¡Fáÿ`©ÚÍ—TÆʳ½š«ºDÎ xùÌ36íÿRñ ç`ùŠ’ôC&«ÕŠ«È²’ñTõŸ·;=©VOºœ‡‹©Ü——jÚï&køƒ9¥8Eé߸wrŠå‡ë¯ZŠ›þG×nÉón¤ À+ñvRfè‹·ðUµL¾åf’ 7¸U·S9K(©àkýl‡)–O¶u‡ô‡â\q–C‚µÇŸ›ïßR×þŒÏ—èôPäp»ÁQ÷¬»Ù¶ÍöŠJž ¸û~4Ë?-Gây u£ò±·uì³N'Ù>xO‰ï¾^B Íu†O8Tšfçcß–Îs 2+ ²ÏdÈ,•¨e g÷[Á$ Ȥھ¨òùF™ZT‹äéû´ÍGÔ‘ŽÌëödˆŸ3Ź· …òM:–éèŸO9âÍxô3°˜P?zÒ)9ˆÔÑ{ñ¡9¾âÇU˜|GšêÖýþ9¦b•É‹ÌÔikÀQ¯Ãÿ7±à)`V5—Òé@Nÿt-¹7K?rÓK‰Æ’¶FqUk wñ`Z'plèóg„„%3Ĭ@ÅY¿|2u‘ÔhD¡Š‹,a-w3¨ -¥ÆÈtÿõ`®¶²«ëü{å7>Ío0v  Ú¯@qÆiÝeÍ8!ªÓ^w8]èúÊd"ADßïuÑ%8xv÷‹“JÆö”Êå@ë3çö4zžÏ¼AÑ)çÌ™5ªj‰þo\‹¨úÙ£¦uê˜Ïú«ç/Q` ñHüÀM¼Å3—nÎkxÐÅkl8NH¹Tmvó54Šô§ï~ԜǘYÉ˸«Išøp¨J53âà¹?݉4å‰ÊPð—KTŠ+Î(9.I æÆ<€ézõ&;£…ãQuÿ*Ò­r6´&›=^eÕÜÅž4+|ÂðųD¯ç„9´¤²‰«G‘ƒž™°mŠÊʤúþ^<†ó€ãlIîÞî1†³Õ†â6¿•Ûdƒaa²¨a®w4åC~m ´J¶3éOo×ýíù 5׊Ÿr9Ô¨M„¡P Kqä9ÞÌ’Ën~EôùøI(–6í¼â:Ñøa1'a¯Š_éO&ÿÁÿ&€÷Ær͆ž|˜ã:v;[ÛIˆ`Ÿä ðþ®HÅÚwÕÞÖM*ûxé½fõe>Öa:»U aÔ†|b=äŸYHxŽe¸ú{}ö .^›IÉŠGÑT!L¿kÙÍýÀêbƹ¾½d<ÎûÝ ‰ÆºÆrˆ:­x¾ùGŸù+LX#þ€ÉK± A’ß¾®×I®8ü0Lß$ëž éÈwJ!Ì9WUވ܋¿EýŽ“«ÝâÊã­÷ÈÏ_+ùŒO¼aBO±‘Š[~I_‚'&ž“lhÀøàºÍ«¦œ!b‘<íÅ;— ‰µeŠéº]¬è ^F©?.U·×F)j>µ¤£Ÿ‘AbÄŒZûâœ>kaŒrßl"ÂÙUFe=S€ÐYŸ¹Ô«;¥ï;@C÷‚¾)R¶hð<{¹¤%¯´“_ùó yÑ„î\4†GQ6I¹yË?f ‡b!ªWIHØÌvü•§%ò÷º$:ª™š7³­¿ÝìR‹|]awI¦àÈ)ƒ@wèÐ*äD$ëÉ¢˸kXæćvàÉÚ㫇ûÚŽ²Š!ÕrÀK“AÒuµÜŒ5»› x#µSõhd,ñ½’!”ê Ó®>lLq¬EIœ^"È/u"Ѷ§‹ñ°7Îü°›ŽÝÌUÖ‘žy¨6ÇwÒ¼SÞÙ]jÿ€¹íæx·O)4Ò‡8¼ÏÍLc„T“¶y4áÞ…=óSŒŸ¯ cCõ(lÆœÂÍøuŸ˜ˆsFkÇv™­“m”À/´èÿ ÉœJ‰3…à„þÊÆ®âs*å>¡à_©Görü¹HÇIŽÊhð¬ºî¤£Ãr¹V²ïrœ(^# q­ö“k%F2ƒ¹<ý çë°o¸ËdêÏ€dÕÇ2¿3:zW’¨áÛvåuø3œU"s¡;Aææ ·@ÍØüƒAmÒ8«a+k¶çΪZ#DèÁ‹”[¾xÇ[.–‚_ˆ8#5´'›2ÆØïö~>ÔÆÒ­—¶f³ «öÎËv]¯í'4™Fqþ7Œí¬€»Ÿ*'‰†V¶W¹zÀÃâݬ#šfsSº¹sY èÎ*<„êR<2Ù£¸±ô¤¹ÓàèE9CRôYò“—lä]Ž®?*óuìGÏ2Ô…Šý­»5Ð[ô³L9ÈБôJ¢”_VË‚`Gá:aðà™4ÿdAŽ»Mé:h¥j°öš¿J{u 0¶ÙëQs>r>²—E"àÆmyé\phÙEF[ªžW–Aß߀âˆÍP ÝÐ’.†`_?M»B±`®uó"†£ nûëd„¸—qxÎ{ïìG²žˆ|MÙi̾ËØ>ô¦9a©ï2x{ŽUÓ–Úw ÍòB–ÚŒNéD*3ï 'Eò‡{%oa»Ñ»ÒâÖ(ùø¢4¹X¸»ã‹ðó0üP<¼§;Òe°ÇûX¡èÁ’„ûÐÙ?O°ÂRìV0æ¸gH‚ØÛc”sÏÃA?ç÷Î’åÞ”)XËz/®u'…Iý‚Æ«%¥F§i;O« ‡½c›У‰cD™úÝ0+#ehÙ.Úœªöš±•«hj¹iwÆ€©kO2nW²~5Ê(õ꟭|<¦C¥…PcâP·¶•; ÉY6[úýsI†¥ôÚjd“Þ ðoÒg£2ä"N¥ï¯ 8A§>UPΧ,{±Á**Ê©MAT t& ʵ7p<Úš;Éêw¥Kj¥*éÏ9´?›/«òw›?ÚzùÁÒ¼Ç|t^etä%'„úpø19":‡±T}g?ûHÚ<‡6ÑRÏ9Nrô€U Ô9¦a<à¶=©ìsëGd5*wÌöQe]šQœ°a©W« qF ß«…›+ !F½t쯵ºìoåÄYÈü5G91jRØ—~iWG4q.´õ¼ÁåÅ/²’i}“ z€Rí”=‡VÄïð¹‹NE·Ø‚üŒ2NËvƒV©è„ 5ŸžeŠPÔZô%eÙ¶ªãNöFrºþ±ô„š8ú•5q¦Ý½,¹„ˆf'(‡[|ý³¼."Z¨ƒ»ÖäîZºµmk?§°–ŠÇéÆfHÆÒc¾θÇ×ÆZf®Þýïƒ:B¿¶ö¸È7Cãå^CùÙlMªß”ŠHM¶ù¤z¬¨˜Øs!§€#çm.þЕ·ì kEÊÔz^Bû¢k™r“ICá—ÙhûCÀ"~8¯LT)/`ue2‡ƒ€º´‘Œcì62?»ïN‡pç^ S=+ž6ùw¦?wöO4´î£dâhªîsQL7òUä4¦:fÙ™{žK$$<óXÈC¬œ–!þø%Óü=FÝPÆ(·Þ4M"¼ ¼CÌM󤊶d )Q¢Ã$ãT±ßËõmˆôLZÀv¹/xÎ`õ‚ ÓcÆe5Äeúé!i¿a„ÒQÕËü 3Z q>³±:jâËêK2™G=ã‘X?zöY‘möOZ Eʧ[“ç½dñzÙ3h§ ¥JYÖg[âš¿Á q-è]R¸<Øû.Eü~“ü¾mŒÜµ³­‘B[BøJ‚jŽSw Üù•s­wøŸ¼û>¡Þ_Äzç±³p»ujŸÑ€Þ‡~³äÍn±.í„äŸ0¿9ËC^ŠxØ¢íïxr"XIʼÍE¨½®áÆ´Œ{S`Öb­8ùX ?%³è½ýÝЙ²1ño’œ·Œéun8Š‘2ËÞ\¸ßÇ«©Tw°hG¢YÁ‰Œ38X0F?qýJP&I®åâ<¡c4&;ÝáTI‘pµn%ò³ÌãðÄy+Q¹åµ$Òõ£‘¼ìRõ[é}˜@#I'z_‡Õ‘O(Ÿ=4#ØèÉÐENÌ;Öí/ÁÈJ2f-/Ķ[ñds°ÙQìò3áóg2Åü”­(?8¤/}_Nò›ë²àç¾ FtÔ†`L],þ¸Aâ?Ý/-’.¶ÅÈ.æÆýÍNͳ¢¬Ó½¿:߯[þj}–›øY$ñ°ïƒô òK¶ßl]¡IšApÞܱðpè~ÂE+l¹$ú¤|‡ÇõÑnZ”Êr±M{€{$".‹°lP¨r ‚W¡‘keíJ¾F;kŠÉTÇ)¸?n—ô•Ýþb¡²úf‰¤À—lRìñÙ¨72dù+GBâ(fJOšvÂù¾Îúý^[ÿ½“öW  JÙuÈü4£ 3Š"]ùŒÚ;©œ3Áð÷²a¡ ¾·” ?(ä²÷þPp9±ÇQÂÚ"©¿&þÎ8š—½\%; â[fÈ}8K{Rï¢tŽM;®‘ö%}¿õΗÅžªçFÔƒ@¶6ØÉäðŸÌ;–CÙ^[6äƒ=®~(,,ÂÁšÎwIyŒÅ•¸~Å|{ÔÍÍw`¸Ùh¨?X ºfã”hW0=gSO‘íSØsËiI—¡ÜR4ÚNí½xé©,—ÌŠñu?}Ÿú ·&«¤©V´Ê÷ð )Ø·/”ÊWee‚Ãm5pŠZg´vá*-ÜÝ54–…S—ì¡jZú¥¾ PïôŽuHÿù™ËZŒ©­XÊ’%×êÍÖ5ö­YG0ÿ´nÞ´ÄBù#<_åò5Sìè$œ¹ )-;¼å½ì~‘~ZHrŽ£D Üoå”á[oÕ7£ÈÞ‚&ÑÆãÓÔ« ÏÏ?~}ÞªÛžPœÞ{CNžóì=€Ôé4aðªÚ¡Ž§/C˜‹OÄÑ]¯:1M K\仾]3xn¯…®E°³°Ñª¨ò >Ÿ1s”Ü_ßO !!»ùÌ´‘:ªÜÙײ¶Ó‘·Ü°¡Âúæu‡ŠÞh–¯árÎ+6ÕÓ}`g²øp)V³äW“Co‚±8ìR?” ëõN~5œõÍ{‰ú…­œØ{~ÖXZœhz¶œÝ¥¼¬ßz¬%gb Ž[š YyÅ!8+"v¨ÇÈx€=³: ™ÛLl$9¨vPsÎ…Êâ÷EW¾"Úß{½­Èð­•;ª_}Þϧéî‡*pVð5$z3¿_yéEÊ5 •£ÁxDzP·:QŸÊÁ‰èc>*'JdYÓäŒÇNl9Ÿ uÐ^­¶kœë„¿°¿'í#³Ô,+·ŸŠ2&Æ à­\)ôý¼—Æ1ö½ø,Ú û€ [7F„æÒñ°/Op!M™ÝAÆs NÒ1;„L+ñæôÒ½Âfé¿Ñœ^Ûtà:—B§äþº|ÉçûkoÄ]‘^u9âX໡Y|W ¾–qlàŒl>»ÿô$!øÔ¥ +[U‘þ’9´€%®¿=PƒÀ|ªTw$)ÔÒq²Yìh Qú±Ê+…Ö%ˆVt;-†ß±ĩ©»Z@Ö:¯ÑƒvÐþ{Ö+ÏP<}kÍy F_¯ž¤ˆ^T(½7ý(£cÜy¯ƒ¸gŲ-a¡ÊÐô¾ökÄ %Øš 5νl%}0Í1*Icx/œÿgÞTª¿ú79ªÙûuøÔýBž3¯kŽï³þ¬ÛGÔkCLŸ—6ìúñnKÂÇLµI¸Âåø¼æ¤õâXðŽGÚNûx—O¶èÁ¡ª¹ô¯O$`&­^u’,0Uy~?f2UÝ’:¹Ï4ø¡øT†-$ ·¯•‘Hxyš÷¯ÍAý…©k”ºó)u³ÄêiðWÆ]2i¨‚­oŒ8¿¶ËjÏ2(^IMˆq@P9¸OIy×>í“Þ³'žhäËaBø9md$O‰ñ,&Ž²…vùû®Ý³úké‹!÷¦ñÆ!õ1´Ï¡àGSÁz}•eØ£'ö¨%VrÎ Ptu!ÐŦnºYåmXxNëó jé€CJ]B^èpÀÈÁêo7ÚpË€+]ôß«dÙ«ˆõ(‘§l9œJ6;$šGK‰SÒ¢si²A5 ˆÇ Ý°ºó(j¡züæMnÙœ5ªŽ.n|Kß õÅ:õùÉêÊ«»ORõ ç¨ä¿Z~mF ,—E=`’–YÇåú3--=òK§îtÔa#Ç*r;µ·ú¤úÒ:^³›ÄTuEŠ8 í™b qÃÆJ¿¡æî#qQF­î²ÍF“5ïæÙ ˜LÅѨà9SÚ]ìYkþTæD“‚ƹQÑe¬ŠÓytë]·“·J6/òiè>ûÙXåŒey(o/FÒ<³r%$ëqé0yþÎ\OJ6‹KMÇ“ÏÉ:Sª»ÜêM©b ñ›Ñ¹Qÿæ%ÖPÂÜ<{®¾Bø =´Ø^‡ÉxMnŸ}¯í%½ÞŠ*O=>â&a;é~ùѨþ—ן#‰Ûÿó"Yx‰Vëãh‘/X›w”) ë\W,F_Úr£ÜH*6"Ó˜€rŽNÜ ÅD!Oð¶Ê–Äó­{ÍêÓØßø™xÝ76TœˆS‰ªžºeBIƒ{Šº—¥õGéÑy ’^«¢òN…‡üÅß¼`TŒ~ÒJFTœÛÃp®%‘Ò”z‹ÈË 7~D¸o‡8Éd;ËtJ±ó>3Ó ¾N]¡hЋG! Ø÷¦ÊUÈ¿ºÊŸÄ.Ï÷ÏbÈÜ«Ê¿W:zÍnµD ]"©‰µ(B¼¸Qʼ,5"1òˆrLz NÊç#‘]wS¯s=D…û +1“'EåÑPÎUS",íÔŠî”Îæ7÷¥ž\³j{¦B¿µ×?Àøt¸°J”,D ÙHt[òãÉB‰,Ɉ‡Þ’…"'ƒ³ÆE"æ?žˆzm×ÛÙþ#<–ÓÊï&ßPóAª üo¤“Ñ~÷jìý¦ ¹—øl®cî }¿]™éµŠk"$áŒ?É}Z5®ôžòz  çÐßÉdžÀ ŸMµ)ê¥ÍДS磑jÈÁ<\º±F2#Û†O975–lJ ,yoœä(¢Ü ÷Âf£®s-k ñIœ%ã}c´ ò¤k—nLO%ÄëǦkJýůºÊuj7ÈGv(nµ(Áv´s:¤§·)½¨QåÝv,k¯s¨¿§ˆUãÃ4”ë2ÉOÉ2é¡Ü¯``ñm4éLpîP]@Ìe°_ú)€¯õÅ–ÔÍBþAíB˜©Wn¢ìétÝ|"Ø/^ÌÀ-qJ4lë&Ô]c„½Eë.ÉÇÎÌêÜZ°ÛãËpɹݛqÒ§sÖw&/‰$»Á'{œ“£õ‘á>;! 3MT;zS¬ò4J8Ø*1°¦ÎŒ‡Î’«Øç;ÖÈ]‰—¼Õº[M³.›~Ó’15Ô°$ßyA½ý’@›ëwÚÁ©ÏµQüý>…~¦L˜Æe±$ãÿzq:¶ªÞ©ß®êºÀȦ†¼ɾi`/k_`áðì>¾ð ¼È¥gc-{n¿=QD Âk±o™“JÂÇ­®C–H%Çé%w™c0TS¬£Zþ*·;ŽÕ¹Í²ÿ5½~*òT¾›Ð4Ù¶J­!êzYHuʵîÆ]•?;¸ÝD+IA_ÉñIò‚ɶïL€hp—g‡OÁ€ó¶3™ÐAP3B„ªÊá–êD•6Î¥”C<›ôÐH}Ý\«ž—µ¡.³Ê­Pë·ð@IJ1W}¸¼ømД(z…sòé»4|¥tßZeHžíþinˆÊžO& f!ŠÓ+KNHu¥ù46šê…+4+ü·a£Wÿ 5•ÄɃI桉n“B„(Du'\MKNÚµHZøj… +{·sÚªÅU©aµŠ€f¶2›q#rÙÕµÓâ JRó ½Ñj8 ŸIeÍq²=ÅÑúÝÛ®>(ÆîZòúd#¿)þ*˜A5î5Rbý•Ãò¿o¾i6¸ÚãY6ãÈãZõŽT·SȲÚáH ÿ`Ñ ™‚SœÍ˜–CƒX›N׎áÆlœÇ‰Ü™ Qb ÚUúðä` ;ñ‚ÙÌMfpâR²@@SŠ“:!‡hÝß©e;iŸÁIeLýö(œœìXŸ#05´öó—’øà ãÝa}4ÇÕjÑòžÃ%¢Xô÷Åd# ÿpœBåµG¥HÉÀUÆØ””$,µ‡Ðˆø0ÅÕ«îJˆœ‹ Ñ(Üþóläe>uDÜ’îÙ2ÍUijÈç*á›BåË`SŽ:Ð+6ût òkù}¯BúÜ`9n€¾%zl«˜)‘vôC®‹bCȤ™-õèïíeŸø,×:ÝEçÛçò¸7·zˆ‚[(ò)˜è4‘hˆ$'4’ãC”ñá7ã›P“Cd\,¤Ubä÷pã[Æ°ðÖ9>× î¼E+7eéàÿmÑdŽ >šôa¶^9V&\:qð¨Ê‡úIš­1_o¥&Ð:ñ ˜5âQðŠL) ê’ÖkÞõÕ ã”:抔)&Ák­ô¸è(òL:ý;Bá*JÎKØz61ÛjéKå19z#`ççEÁŒ4ht¸\y/Û×Aìóú°º}…I•Õst @Ö^fDò›¥ù—7Ñ¿ÅýLHžó´ÎÀ'{y… l,XaÀþ‰ü 'j†cˆ‡u5K 5Ê™PŸÙOsÙUO¥:õqËÞY¶é{˜–í]<>P°j"Æœ3Û:úz±MNLæv+¤È?èÝ=o|GW¤“›”ï{‘>Æ)3þ|ñçkÛžNz´-°P9…æF“µýו¥= é9>wf™°¬p3²Ð뾫²—Ü}ÅJ@$Bí̦ +N`¨EɘM·Ê¢lêh“‘:ðyMD¬-ºB?o÷JÝæ·€?÷ ûŠ@Ông%N õå„u ¹,q ±ÚìÜ×õY„Ü)o$ '9;(0’+~5µå¥˜é.âa +AÝ)µxlฉƒ ™“uBß ÌIQ±| â´î‹¹ñÅšKŹÚAŒ Ãuƒ­À¦…E‰1΀EðTæ>ø.a$pÛL6bª´¿:#ªø¬¤FÁ”3ohØçb =ƒd¥F{ÚÙMžÌ±|•àuô^³ûõü’äì+?=¥dã,µhùiVK’ÓþX9逨‰`m)vËõÇYäõÜz°þs+t<×ï²[YÄ䲚u¿²¬îåÖ‰% .Ad5qïuÔ4$‡‘/*ÝGt0å ¦%-LËèïw¿S¼_8‡ÖÐ7_~jk¹§C€˜ûŒUCmV¥û7$¥P?I¯ª7±ï}y>ÄxÐÒfñu+ѵxóæÑýí#Ðk騾òéÂÿ÷@¥3V"ci‹ öĦR¦ˆÁa”½Ÿ‹>SâÒ‡r¡Ö*×:µ\>Åù¬jþ^ª1SκÒêæÊE^x¤†I×æDÍ]†Ô‹‹ÒlκÀYV-n;¬ÞÉ£C³dîºÕ£úg tÊ3ÄòÀ:‹‰ÜOfõé±3tMs†oÙïCAëߟ`;¯°­RP Jáù¡¿'¨‘Ÿ%ºùƒsƤ‰dXk>bQ½3øá ¿AµA2m㱯Úã!K&8Iè›lèn N$C<¼ŽOJ ËzGb„®›²ÒHX>Ìè^×'DÎ |x0òDObá\ÂÁ,)²ü¶‘p?eaý^ 8²Ûð¬±Þ(Ãwž¶ŠRèVújW½Ì—éˆþUº%ÁüÝ$Ú1¾ Ù¼ ßnêkˆtc:5[ Ö.ºõ Q²Öæ¹ȶ^’1ȸR%Õ­l€¨×È ÌLJ·“ÅK/³±:V¸â†ã)Ú íN…ö_ø%{|¶»u(4®vn¾èž™­Œ=…¢N5&£G­-}ô ÏõŽ5‰ Ó„ˆ ×ù3ÐfæþÀ‚#Ç¿²ÁWo­)íÕô²#ñ(”y†•¸4g3X«çiÁjßQü‹°ûOe‚ÙQ^<Ì°ï:ÅGÁ®Ð†ÕÕ´PêÜvÞ¯¥œïå ¢qT´7qgC!QÀz¯ZÅ9áÇ¢r>4UŠ%÷âþCÄ#Uªñ#”â¬lÍ*A8õâŒ^{NˆßX‹ù½pÛ€•ww%ÇR¹±°³‘%A±ó§È.Š¢I;I«‰/Ó#·7™oúê…,U]EM–ˆvL÷$áøÃ50ê%Wšõ¦R¶^uæ|5‰ß\J9+X´7/Š›RRúNÈNÏpe…'úbƒdÔŠ;¿¤,ìØûJÑ2J|Wœ@J ÆzäݥߥPì’Mì$%§»wwNSÒî,;æ,Âââõ˸~wMò1Då®; A8mϽbO›&ÃFCë+àÿnÙåÊy·eL›¥_nÓñZxÙI5!ÎåE'FÆN‹²þGò,x ÖnÓ<•ÊALR Z54uòØyjñ[¢ö=[º…ùbrS/Ã8‘K¦#,íƒ+ò $Ù_ÔÏHf“ 3@H#'ž â ’—±nÄU»$p%ÅzÏðäYÑ÷×^²ñεèwôûrHOcHVÍ`:m÷Ĥk£˜?œÜBoÃÛñ‰U#§;m£¬øüpü(˸­Ýê%ëÏ^Ã$'gD–æK:d óû?´î“Ù“?í‹ï4Ô?a¬oŠAmŽ¶Ÿ>ç×rE8wµv,£Òóy•›æ’ž~N³ê¿ ¬ó€a­fúíÆ`ç—oÑ”‹mˆ·.§Ø@ƒ|ŸEônDÏíZ”_þ‰üàdIIÙ1kûÝ“Ëþ\‰ýKÜpébò¢7š!3}"·8–3À²j/j&íOã¡( EŒï‹ˆÅn§%^ŸÖ€Î…ݬÁÙ±Çü3££ÉjNÁþ{5õÖຠ"MF º‹³t Q[áÙ‡0Úàg¦‘fç·&¨B=Ôÿ š4…1EðT‡rñ'ö~«¬ž|±ã-ŠíaV(•Ã!W§Ü_íÓ)†–Tëen½òxTOÚV¯B)¸¡¥'.«x‡Àà¸VHÓ© }a) Lû»×uù™ýú|]MX‘ß±˜?JÊ1Ž 3ýl•ÿ~äØcó{³ð¾çc®â…Z\ÖØ-%Éžûi…­b\ÍŸZ(¤^X»”W°c¯Ã)^ŹZL¡.gL_`}RÐáeWŠ6Éó*W•¯ìá§`8ø«Ic]xR™·ƒA%vyëÀ¬Ýç°â¿‘¥fÎØ×ίk`W;2ël•‡Bl94reï¬h`Pœ’½‚éÌñ&xµE,ñ!jšãÜà]”9™¶®‘+Çß$c8æþªŸoœÆ1Ùö¹Ô¬Ù"ùŽÉqøF•­”_"'F`žUz½áUhÂo¶·Žm Ë’†A˜W02fn,hðø ‰ïøïî…Ž¨ðüC$Ì‘ŸŠ-ª§Åfš.(“ëT~_ó\I¾ù‚PÏîk4÷fsrD‘!É‚-¬á0¸ ÎZrºCJþ;·aéÂ;¤ªóJT'Cµ3…û1µä¯ .OÚžtæ$'6ín†Ç¸årƒb›Êd[hâ÷I© ×Nœ–¿´u!»zü´Z…V‰ïzõȲBÃÄ1Åg’Ñ=#Þ£µ<…ýš~¢..¤ËøV»~¿¨‡AS;w:5Olö>^X›zA®UºÖ˜&\áC'ƒœÉÃÊ ac%àÛªš;Öõ”F"O9ÍŒZæY©ŒFPUâÑz1SÍwns‚:ë´ór=Ë&Šc*0í È(rwôÑ;‹©ƒi ©Ü–Aãè5g‰fKš&®|×+ÿÖ&Ðÿž¬¿ž¬™gëïa/‹€´/†Ê´Ã[,i+vÓYëì[Šwj!Ëá"ïËp¿üLØ:F`ÔâÖÛšûíÖüU“¶g·ŒW.ý•&Ñ}pŒ\D\DCcþÏ»í=2äÏ.Ê ¡ä]–2Ì5A5DâŸAüÆœ%tèwç#sE«_`#Д68Õàu´_Œ/ñÚ*›—=%±.Xþ&`bdkh7ø¦5)-7AŽyå¼›škªà}Wä¶àñ_0 ²`qØ…ãä7ÀVCwÜ®áÁ™l#_t˜,mÄ,M àÌCÙŸèr½í xUhœŠ€_£¦Wì€nñ͸Gh,¶uÆOI"íeE(?-=ÄÔI}ú"~S Ý‘ Ä•¦®Õƒ†½G´•ýdTÈvùÀ¾¯Íë’;+R¢dUï¦g sÉ($B° ñǨ‚L. lyÛÛ* ³ÈÐkšd$¡²Ð­Q…¶F¯C‰h°7+Å(üÓ£‹$RÝì­VË—B¿Ñ¨u¨¿j!†éBº>æ‚#?.‚ÑJ×0é«y{¥ÆZù›ÙïEÒ-ö™ÈÍT£8³sKRÈæù£ L[ ÝU[`²L õtó…Ä3³éÜ5t‰00ñäÚ$w?kFÆ; îE8õš©K­t2S­J},˜UùÝÏ;˜ë‰\ÅÖäª&Ú)åzó$½F) i¼#ûšâùng¸ rš;fOÒƒ|?œ AØ'zõäæ°­¹9»Ãa-ôœ™Ð²ÕDЈQÅaÚ7SÓ<#Ùw‡Ej íàŒrëu˜Òªd•Aâu‘Úÿ}ú: žæ§Û9<—Æ|1™3%4î ’¿YIŸÝúª~Q(à×o·÷ »ð‡tü^žÙR—¢9”lö‡2‹bmôÜLé1©g ¤äìËUÔ'/‹Q³õ¥?Eƒì„%Òpâ ®jqùç1EöÀQïåþ¾o4Å'ŽxÛH¼þü¤H™Ú[ª1¿Ò߸¦L:‰­3 ¡ã!DƒªJv¸‡w›)HBLkF"-Õ÷ªÍ„Ø(™lÀh·€3ú–´(G®z^Þ4¯8­ÂÃE9d„=@–§ƒèhgúñŽè/[Œ^ñ£íü‘@x{ópÍ‘Šé§ÎîuÁÊôÝ™ˆ]Í21×Ô·`¥l‘+£¥“·X>OÊ1 #w)_õ³÷ìÙa›µ'úæ/I²¾¯§›4±Ðª!vé¨ÕIŸ àl±³içâ~Í: x‡~)û×Â%}i¬kñIGiuš’J ¥ð·„3¥gäºó^;WÞ÷Ž±å,(‰>£6ˆ‡ ûS½y›œ{Æ+ù`+±‚… ðÒ†“³‡÷ñFp;¡ÿy÷««k¹kj+>>¢ü N½ùÅx•¡•€ÓáÏ2m­Ö¼¯¾‡Ò˜ÆñPcj˜¬J·Ý®ï{I*ÖI Ám¹Aâ‘D9s+ˆä{õ­¯¿2á4²‘J.ƒfc^¯ûæg¨ôÇXžµê²-jgƒ ²Æ$^G!ö«‡~‘ƒrÎÚÏ#t{3Tó#IAÆjAŸP”D%Œá±Å„[IwmY²x~Pªº %Ýî\©ùgQïOXÚñԾƿäHŠkaÍ7æt>3« ³E §ïã+¶aŠÍ8;K„•1=û»zwLó—£8’‹ºÖ¡ç$Bª_õÒíFþ0Ç­ð¸'Q"ÅÓ÷±gžHGÄ¢RÈðJ rôqw¢VRùñÏq'ƒ²4Ëx>NrÍܲ3ðNÖSm> Dë¡î÷ß°©t ì×·ƒÍ âi‘¨“ÍWø·„\µI¹u/§Ÿ†’YŽ³g©–Ìù ñ³ÂMmÞ÷þ-™†iôèp ±¾ yî½9ò•%F­Y>Ñ Øzú§ß=~ôŒ:pRíÛ†ˆcs‚ûÃû{åÁPû|±<ŽiŽƒ]‹U 8™x?ÁÔt½›[&ˆæJŽ‘uêEyMzÞ>QYAuÇ4‹¸ù~?6¾%¥…E¯©¡®4›Â…d‹“OÓI(ê¶PDsÎŒ×úw<ájÅÖO*ùÍ;5~â˜nqãÆW±½kôËJ4I!Ä8Àx(N³$&µOÏ—1‡’ŠR³p8Ùº`c‰ŠœPt¶Äa­×»™Þ檑Ñ!x èéâ4À›†¦Õ:»mŽWò¸ð㊠ öÐð ê´+ˆF™êýÉÐÜ  ;J†>t!Û4H¾*ûR.óÜ>òf½Ü¬o{úœî ;ž£µè^`àR"-á;É©s_!Ê#ð†I‰¦ZDû’ä•7*éïѳÿøÆåš«‹³_©m8DšÎJÔ¤/â8ïH>‚¯É#›¾‚´·.åNØãþ<(¤Ð:/ù}(‚؛ض½ûc’ý~ü#¢¶ü:¥Ÿ<4 ïèEúNÛ†‚ùéÛJò€á'u„ÓøsƒÏXÙa&U6†‰Éx¨Ížj9¥Å8N+Å0‚IÌàíVÈ0óžeªa^Ê¡÷º³ßðˆâlæÍ 4HÙXñ•ÎÈÒyŠ& “¿4Fìö⹚™.†Ú¢ˆzW4ÇVG±)ÉD]7fÄŒ¼ ³¹¼ë`s.ÁÛTÑb£4aëyÕhOýSuõT´2p'¼”Õ̲¸b/[)–³-˜ÛËp¡›¹þ´4}¢á vh#)¼ ï?1.Gõ##æІRœ~8‚òÃã:6BüȽï|¸I¨þÊv¥sªðšrzCd~ ’[Üœ§ c^¾=Š¶s·§°Ó ÕÕ§…/Ús³P¯q¢¤;qd/Q%"_ûÞ#P_U’ª…%ÀòuêÆü"Ýü'Ь8u([¢- ]µ‰;JdöˆèùoL¸ß"Þtðð‚6C>9 Hâÿä\ÿÿ³Ÿ?èdb:?ŠÝd¶Yûœj ½Ž„àYG‹a (ãÑ;õOâÒ¾ewN·Ù-) ˜iu¼ÍëvGþ·ô4n[Aä#ýÕ¥óN±‚ ðxù¹ÉD…î4M;)^¿+-¯lä?2|tÓøkÈ¥·v]ÚÇ z£3’´ÎŒñ¥J3ËI¥ËO8é¦||‹ÃâªÆ<´S”Ñeõ¢ ÑîÚKàÉW%õ0Š¿’ž£ú;õŒö‘ #ŸWÊÉÄϺ1á²î”ý3+w3i˜Úö½ÑbkÆK#âzïá~ ¥œóŒy­Šƒ±Ït‚6_>»¶@œÂ{*d¾÷r¢ü½Û3ÎÔÍòÎê”6lðO=(Äív`ã„Ñ“ -hÔ/vºµTµÇÀiêïáý[C5™ŒØe_úÐ+ú,ý¤Á dè¿$ ØœY RJJ_s¶¶¨` ü»é÷üZÌfÛcf{9‚º²ØïðIùعk>”áÿ'œk‰§•d—|½6Úf©8Èž|­IÄÈ÷ˆ‰ÖòHq)¿}Ô` š97ži~ƒ6 Іà;š”ÿ³ž^Ù"ÍPFÚùY©ÊÁ‡Ïûåú]ü~Ã`F¿ÁÞÑU(4;œ*D¼î$våu÷v’éWÇ„ŽŽ©ˆ*J%\hM‘Ùûþl”4^HÇó(©ßÍIHè!—h¦u9„©«Àçí/78§ ÿðÌVi¡À£Ug)×Dç󿈉$TÿÝ×&¨m}Éxž …)æL#©~΋Pj©|1„[¼š 'ñæ¤C¬w×N9é×íÙ9Ž°pQʘQ±…ß³?„¨'+…g©•ªñæ¯D6` +K,„ÇØ€²°ÎxRڣ׉-SUßXŽa/Èé0AkkBÁžñ\ÛS²§îÈmß4›Ž•ï ô™‘îw«+±«­ì?#Ö*Öµ6xZ—³ש¢w& u²pmÝ«}à2 endstream endobj 57 0 obj << /Length1 721 /Length2 16693 /Length3 0 /Length 17293 /Filter /FlateDecode >> stream xÚl¸Œ®M-ܶí~Û¶­Ó¶mÛ¶mÛ¶m»û´û´qÚ¶ï73wææOþTRÙX;µöªÔ“‡ŒHÌÞÎEÅÓÁ”‰Ž‰ž‘ ,¦$ËÄÌÈ`¤g!#q25t±´·ûeèbÊ P75(›:˜˜ÿ@˜`È"öžN–æ.Jcªj6†&–¶–N5{{7Kc ¯››› ›³+½“+?ý?Eʦ¦ S€™¥)@D^ASRN@).§ 7µ3u2´(¸ÙXd,MíœM©föN›ÿ8c{;Ëqr¦ÿ7;7S'—ˆ™9ÙÛdEU„ÄäåT¢" *"C;€ŒÄ?åv.ÎÜÿ MlÿÕ-Àå–áÿ,£ÿYÿË:yüÇ¢‡ab˜X»ŒLÍ-í`þ¥›¤™=€ý?aW‡ÿ¦þ!äü?å?RLLÍþA»ÚØÈÚš(Eìm\]L²ö&¦Nv%{[C»ÿ¢ m-m<ÿqÿ!ébøBvæÿ(ÇøŸ¥³˜¥‡©‰‚¥Ë?R›Úü#Ñ¿ãê¦ÿ¹YSKWÛÿÔ›þ£Ç?7jbogãùü§È £$¯®ªHó¿øwZÔÎØÞÄÒΠìò–†N&ÿ ü;­`hùŸéù/™ÿN€éÿù²†.N–mFúã¿Ö-Ýÿ‡¶÷ð¦câdб03˜XÿÙ¸X™}ÿ?t]œLí\þ-ø?½ü×ÿ÷ ™šz˜Ãl¬Úó„X¥·…Uú‰/TÓ,̨Ҭ2¯ÍöÝh#[ŽX˜è]áÜ«à¸ÃáÏñí]îJÞ·Æ›Âýž‰!ŠÐ]Åy-%©2÷«Û~W,|kŒÐ»Sö5–»‰¼sÉÜÜR žaÍÎ;¯“r,çCgŠ1¨èGÚIX_z´§Ò¥@ÉÒú†3= ©EZ°-º6xÝß /Þƒ»!ÃÖ·(ª t8 ×|`£ìàÚ]•¼·ír´)òQjG$ÄðÔZáÞ#ñkø£:¹Ë=à›5qÔ8Òô¼«ËŒ¾FY‹¹±YõÐÍtIÆ ;‰[’x¤iNE°ßUÊrRME»¯A§ƒ0\A雹´WÁFLVÏô&¢Î~ÌzÅñ~F8Æ©©<ÝêW“m‘ñ¶^²ƒ¹$‰Ô­¸EÆMfÿÁ4h˜r«øûsç€hòÚ—d^Ó«®ÏDìL‘´ï@|ÏC(öý‘v<Ô»'žݦ`뉣1ÊP’:¬³µi–NBíÅ×Ç{•cµ;'8_¯É¡=y8¹dËTÀÔ wM&Sý³ÃíÊsædùª.¡ÿ¹ÂºïîéwÆ ùú»ÚÄ/œØ­ÃhtÉàÿ{îוp÷dóvmêÓ&c.j‹sLZÖ÷hÔî…EE‚J(•e¼Œ=—1<Ÿ5U°e²‘U’OC_æÃhîŽÜ É­õD±ò!97(ú—û,î×ì¹%J@WH5ö-!ÿ/ºîÊZSãtgh[«åÀ4Éó>Ac#ÀƒÌ:ân† Åß#Jl+M>ߟß„'9ªÁ©²µ spü[ÄjO =“¡²¶3ÐË¥ÿ›Sê[ŠÓ¿,Õª^ùÃþ,uå‚nó/â \oäþãž?––,µ"U1^\¤8¬zÉîd¶aI'—y#,ܬ$‘éþšë–Ÿñ”ÑñhPÈ”ÕÖ·›4ÄóK8ËVau: {¢³‘9¾éOöŽSª›²wQH¼þ¥[Êàþžf¹w­M¾Wˆ1^éUÀ¬“U²Ášø¯c˜ÍƤ`¹€þüü Š¡ueÄ;Ý< ð+Ðh³±9ÎJßúµgŠ6-ld;/|Év2¦#M‹^ûžd¸e§{òÓ.4°—ýÙÅËfù… fItG Ž>tí!m‡4) )ò{(OODes,‡nçsÕ?†2JçZT|‚ RȪèþ(Ž…—£’‘V˜äE¨µjl[ÂYЕv’Ïæ^À˜±¢‹™©ä«Z±®Šì1Š8rû ¾¼_p°ÈWÞâÏÙÜ—F’ܬe±¶|.¹æ‡æÑ?ôñbÎÞ‰ôôYöú3›JÁIYf¶­ï bkn¥ž4ThàìÛzU–kQÞ›í÷êž'ík\‚’¦dÔ³¾OðÛKCW?î^NdåKÅÁìýTs豆¤ï+˃oúˆ<Û¯žÔ̼¬ZóØW¹V¬Šj9…€#>âP1s»p[»jù ³¹îN“Óé<¨Ã9³=à6jžz[û DóÅm,w¯žcç¦]ðï„šâ²W,ISý¯ó¤óë`s=A>ÊÊ,( )ÒûìtTïÏש è} [!Æ.­ˆ? 8„ƒÊ)ÉÀŽü~¾û|[Û‡ç+1À»jÑ#¨cÀOcÕWtr®J*Ö#ÏÕñÕ >ÒädTêç•8_î»,¸ÿð?M›#ÇÆتدùd’(€Ý×:ëÁ…çoÏC›8<âzlˆÁ%a‹|/zp·>UÃÜ]Ðå–E[x*54R)ã…Ø餎Ð]]-‘û" ëA»}ƒ@Bè6³Á“B±·3‰ˆŠ}ì g=¯E(œì©˜òSkíÉáåíßK¿ÉK¡Ø͸ܵ &,¾¯ã!•ûü¥£ã¡~€2.@»šö*ƒ-Ní•“IXÂ%Í­lë˜OêÕØÚ˜&X[ Åå+•V²‰ò§Xž0Ý,£ZN× rOqgOš¨_cð´NZ²ŒJ¬Äpg©- ýõù Goç#sÔšŽóóuþ ñ´\/†>ÍGV‚“öËWRa†ý+P¢Jœ'ñ‹Þ¢å²­sz˜ yÕâU3)2’\àìê¹'¨Q_ôÈx¾JÄdör's?û;ó[¸ô¹<øn iúh€*©}ƒŽçèÄ›óNÌ[ΚgÊV,Bo “ž¢Kw*Î:¬[ª¯ÓDœ© ^žb„*Ö {$½eTæ:9WVj Ý ̯ZÄ”H÷¼Ûe™Æ*1) _ôÓÂ#ºn€´µl—e#œ%_tL¯¡#_ÿ- |ÓC ·£áÔ×*l­œ['¬+rÑ¡û¥5ÚQ¡Ô×/ÁuÑñÛ{öâLÚ©‰n_2ÂM9>"ñbN7h~š~Oýòë|dão˜®ùþ¸>KÀö²KÑùöç¿•âÍB`á~)º>È°Ò«ÕOþÑ¡ú0èŠáIvTì!af—Á¯mÞ×̆ô¬|´&Ó+ZÝbJÞ*o1åa¬ª²VFöñw2 wNp•yYéò»ñ@ö^qý›G–ÔŠºJ?óNöú­ã—¯âŽ«Ã{ƒ°ÃKþºDÒÌ z_róéèY'™+jk²•lÄ–‘n<ñ¹JÇ/˜~ˆBNë<_mÆ«¡k•–À[ë];-Åi—^ ˆ û]ŒÕ,‰gŽËU¿fª|/CèɉÞäZ©É%Ùy¡q—·™‰ñ+²ÊÞäß ±)Ý2N˜‘—÷Êw½Ð¤Õ½»˜®´¡E‘™«GE„] ýÝ{~?íÓM ß˪±(Ì?‘Z´‚eüEÊ&lÖä€.Ùýº-üö+^:(Nú×ÛdwñÇ+Úú¥Ž©sËô:VHqÕû`´všpD•qÿW*CÍxH©€¯×X9"2ˆzÐÕ‹©à)eyl+›l=øüø”©Ô´Ùý³/Û8h¨ ¾"ãž(("&ié_hæ{õmTË€Tƒ ô©óÈ& N?dÎXã9w €Ìd'»×\'H”ݚìAè« ¹æÇHü[ƸänX˜ŠàÐ.G„ÀPÿÒØ!'UxŽ·š B#_IOh¬N‹Í&· ÿ´{…½äWs`~\ºÏ+Yï­< Ò`™btïÏ€ ^ø_=T­Ð;2Y‚€ƒªá×· RbÊa*®Ê¡h!(tÝì<{™ý0ç²>â®—¥ê?ÿ~ò‡¢}ÃcS¼¥Þ“ÔM†ùpo+gkücUÌ/ï‹L¹§î±;‚¬± 4Ê_p®ªÞ`º)ç¹V¿¸õ}WÃ>“6·õ„vÓÉ›´•øBo@ŸÏ€`zãÂ6k: ÔT’®îìЈ:'Ÿå@ Ž££uÐlXµ @‡sLZ²ÈV´~Aps0=qŒ—v„þu üf¨jÄDï Pº;z$E•€)ò.d¬–0¼L ¤áéëâàˆ2„òˆmãz iø |÷I´‘¬´šŽ<“f¤”áÎÌlÖ±ßVÒ®ºù£d®;°>ÁÅpî/ød9’aì+#dÂC£çŒ ²êßÆýýÑ þvIZ¨û@d+._AC5 ªj¥îŽb˜`*‡ñᄳ6éȇÂ,¦Pºž…ÄòØ‹$v?Æb_–4wm¡lŽIÍ ‚²Œ S½—UûRo_æšú#œGÏXw:ßïHÅt¤j>QÊEÆ‘ƒà d1§ ®à„Tü–V4A¸)q³4QY–!}[æ6¬ì’KÎi êZFyªÎªôp÷ÍYÞœÓÍÉP”BeŒ5gÛ×Wð‰QÌœPÁ1Q[ Ïá®ÉS¢ÞÉÈù4vï–Z©-Lü< DG·íð•-‡G“%(®ã²@@÷‹Ñô‹Öƒ’;ø„…4(x ä3ë0žµô?ƒŠ çâšþ3Ô^¿×Þ é7ÂÐÖp†ù"³f¯2³i×msr¹Ë˜W"©*÷ø µ Šjß)¡ê¾ö¾E†ðH•a–bÇ üj¬a­ý÷ÐfÁGÌoµÖ/µ Éñ í‹×‘)¨—ôJñ¶€“îŽßŽ{&Ë—÷‰„Ü?`ƶGc˜ H…¦QìÓSo› WŠÈ¸ú“’Ù£ëÈHÎ4®Î‰«ßX¾F;êÊÜuJzë#p5q\HŽ’uìkë·ˆH´ÛíhÄý´Æ(#ª„¤?Áà;ùÂ]ÛÂ,CO±}¸£°MÓÒ_v–E"ì¾Ó¢»†,' Ú\Ï|ºz<®Ðü–ógþ¼–*(é%uq5T¶ ávóÛ"jMt»ø¡Ÿ"Sk«öÌo…Éâ_h®·žÊOV«ÎÚîƒõ4ž1Ãê’,Îà³òFDc9cnûGj»Aü‰’Wð 3fû•÷¯ÛñœÊ]Þhº´Ë4$Y&\¾#nï<! ˜d²UGê+ä{D2Yx3ðÈõ­o&¨6â¡Ñšk&P™}6ƒˆ‚ëq”f9osði®ÂM.uøXmÒ8í_+%Çòˆ@"˜Ê‹?å"}×þ½ä¦ÔÁpW‡í•lD¨«ód'±ù”9@ãØíÂñ2ÈI¿Tû!ýÌ35ò¶[û¬«KíÚùÛe+¦W—ƒKL”š°M1¥.Bªª×Î 8· ä!‚Ù)úè:=wR¾þ¨eªi;Æ:§ÛÛð¶Ô¢*îU EÇ¥•G!©T[ZKúcNµFV„ ›ðöÇ'ÇIÙ ¤¤¸Âà´Pó­Þ㺠? £ÆYZžhß›í ‡ËÑÖÖ‘§ê¸"3奋Âzæ#´Š[ôÞ‚ž£[’øçƒ-ç÷ ¢ŽV1Û¨÷`¹ÖZchÍÖ®,eí.Æ ^˜:m$p1Ûœ~Tˆ\dt6±7o€©>’0r+fàççÊ0!î üª8»2Ù!’×BùIïËö¥!Î$VsŒXâýºÑH§xx`ÞÚdQ’-‘›™®”#ËCjÿ”|‡52³yÒ&_9í¼áæDt!}㵶:þÆ1s41Zö'‘ÁÖR¨îˆ˜*}ÓSÊwòöâÆU˜f=·‚s L—ç ]®·>ÿí±dí>LGÓýP´EzœG‹åáUþ)@¬œj½ ¶tE"v™WïyS ±ÛÓ(õݵ ½ ÜŒe¬'ü  yä&®i.tÝÐí“×ë2”Ý[Ä£“»ÉîD3}ç׊Šð‹w®†”úH% gº² _Ûƒá39]º¾Ô\~#¾‹Qz-på%Ó2ÍCÞ£bèŸIp%¡Q£óoW«1¶ñÖ¥µóäàÄ–ãtÑKú†#vƱœ‹¯ˆ™Óê2õ²Úï :Á=’R™²¥G§V<ÛN<ÜÉÃó£ZZ P4úKhº;c½z`¨à«¸?ý¤ª5ôЛ5K0aUÉA1QPß“9Ôª7Ò[íÙ3å!mQ,‡Û°‡1ö«×f$Å’ Ó›@àÙV à1’I&œ%lô“ÏvÎ1žuF4h‡ÉÃ6'‘PÉ´ÿ^t¹]gôÿm¹­¤ZË‚ÛB±\ƒ¹Žž>Ìé¥qÈ@Ÿin8‘ì‡íò"„mxŽó2Æ2örxí/µèIÖ„Fò»>¹Y9ØTgKM0fÿf®"P†T`‘:œjîÝEa§Õ ì÷Øôk¿·ÃÙö ,†ÞfBG# þ9b$‰_WÏ@“M®z¶v–>*wðŽ÷gÕ+ÂÒÏ,VÑB±Ä¶ê­y’5«ýÂqvz:±(µÖh¸ ç|Öaõ¿f÷òmtySa'îß—_²©°5ò6–%ÉÉŠFïòíkÄ€¬y÷£ÈÎ*(Ê*ìT¥hƒp{°¹š«(ðgÛþ Ö,$âƒê¸zÛÅâ¼½ó¾ÑWŸIfN›Éâˆ}v+sž¬¾ü¥Å؆máÀ[çµøÄZÞ»¥‰5\KÞ ãISæè˜Ë4mR.SÓWü²4–ÀF^ùBÛ¹ôhÉ·]è×Ù—%ñ°ãÇ}7Ó U…Õä•2›ltžÈ饔+ó³åHŽ‚ÓŒ=êhªrÇj,+ ‚Ðê¾Ôö …ëÏ;¡)ðéJ‚]o°ìLözGd÷H ?3[X©`¢QÛR]é±µ§êpîunŒWJõ4«²_Ÿ'Ýyõ…ñ"¬RŸQ8G¶2Uãu‚ tï²ø€ÿÅ•ƒ¦K‘ ÔÃ5;þ• gpò‹¢ádèÖÇ;@“xöÉû «´_ãutxs¾ŠnÍIS0É#¹!Í‚VmÙq’T½P˜Zã¯ÅÅ-é€MJv&( 2ôçk3äxˆ«².°ZÓÛ“63o^]É(y3å1;]²ÿq{"[3N´*r¸ó" ‹³ÐÁ ýqÃaÊè„Qð…¤×_Y"ш(¨eä@ÊPH«oÔtd]rý­NzטSújȆgÖŽa ¥¾fõð1¾ë”µ¬ø“JŸß¢×k”%lûŒŠy NϼÈzëå9Áã•0‡\„S ‹KM€nëdÂ`h„i×­“ñšáæ{r¸éà…z4YµÐØö…£‚?µ”j«Û2:^@³ôÝ“Åå„!¥ 4'ŠÆ.€v8¬œ‡B/,¬HÃçˬ¦X*¬ÂW®¯,†¯wÜH¥°Óã;h]Ð^/ÕS‘KFÿ¿aÃ1xaST¡4k$CV d‹.\b ðä4”žaž¦¢}€Ø懘-¦C¦Ó„q¨¨¹iêäq ‚Ó ;“³o o/Åzòz;k^¸µ©ÞìmÏ Êtë¯*ju‚Ù÷TÊ!9ôàéÛ,žéÕfS£í›¨P‚¤Ùºëqð»'$ÔlJ Ï›hÏ,Ž:3îПäcä­bE¿°ð¦j,n¥5¦†œ'ðrþ01ÅÚñ1!´1uÞè¾0ØSÄð–± ë<9Îa šJQ aËôe)z‘w¦„ÙV`I"d')»oç€xz W©­sX¬GI ¾¯¸!în‹âñÔŠ¸äEVp?`¨r1׌X¹™â“ïyIcpkK|xü—©tVº·jEg•‰QîÕL!œü˜ÑÜ“Lâºf*>‚(s/ß¹7±Ðe]9êlã•ÛFëq·©ßªýw‚Æ\Œ‘tŠäò·{²|¡`\{Û÷üzå…¿Í+¶o±'çoiW«¨#³]€JwVë½úÃppb¶[TîÍÛ/l0 ì¿1íEys;îêV\”;ÒïmÒ]X}Y¤Ýß%iwŒã Ó‡(\ج…nì´ªÐÑúl yãì.è§jηœüZ\†>ÙOÖt—6ÐvgV3º;e1›+¸Ùÿt¸€jSû2ž_ `2¶ß ¬%ª¾ M•G¦ÀgQ鱌\Çú5zõ¸ÄþViô}[ªõš1&Ȥ}ºYiB¦‡J¡¤¹ûî¼ÙV¸@×UÇ.eÒmƒ’§¨ *+»¢Q.»?ÚqƒGwMᴠȯœ]£žŽ»:èeê+Ò¥ƒd5Ý*v·ZÓËZ;ìºo‘䣹޽Š—C_™Y…Ôj]Å À¥ ˆ±Þ¸ HCÀ~k±îO¨0×U{EjkœK3Tö ™pMàÞþÍXZ¦º3™´°Ã ÜãÍ3ûêoà¾>"̇ôZ4ÂV’%'îébÌÒÚ±2‚>ò“x_¹…À=7„äY=#NÜáùkˆêäxË#ýð Ì5ÒÃÀ:‡ú’N'ìÌuÓä—3F Æõ4cì¥þ¢1âÚvßÔ¯P ½ÔÑyHÙÍ?ä ~y¢—A»l¢1¿ŒáNq>•õâ…ÉÕŸS¬¿Ô®¦zZEõ[ë·"&Æü[BŸO©ý±ƒiÈÚ%y}(‰ôÈu±x<ä#ôJ”™+ò_û^CÅpÎ*ª2úð…ùî!}@8G憛Ã>:ËèÌÆchAæx/ÂÍׯ¼€_M$°Üòè+€¤¬ŠÛ@ìS&"Ñ·–À™Aߊæl®\ ~"ÊœÒÑÜc_BµÂ‡Í;ÚX¨äߟO\W©®Yñ$©&á2Ó)Y¼>ØXYÑÓdKÀý–¼Qp"¨ª·ËKç @MþÀõö­i4íýA¬añ’;®T×ÿaqÈE¢˜DÉŸºÏ&û&i(ã #Ò¢çsÜó Ÿ”8´Â;€ØV„§õ…=ÛFwMzí`.Þ†Î-ì×5;¸Ó¿ZtñIž³âq2ø0‚QÅÑ» ÿH‹ûfM{qÎöŸ  žVG1Â&æL;{¡ ¨ÅZô 9¢³i³Yí—qfY¸NŠA»ïJ‰–ƒY•nWxO,©NÎ3•jó¿H»½Ôð¶Åô•ð?(*äcœ Î[åwè»éH,køUŸ¶Û|n-iq‘YÒæ‹Eäþx)ÕÒžµÜ<£ñüyUnµV#¿RœJ—‚ç[@/‡}K{ê^©<þÁ~ .6Ö&«ä¨AŽÕý†T”Ò¾q8±ÍzÑ¥²[]å™þÁ;±–†6 <¤<)ÙäO²¨anÖ?u¸„Üí°Êg–G³‹êÈþ^v«ÖÏ ÑÜ&sî@0O"GR¤‚s)D jÃ)’üÝÕ´÷áíÞ?5Ò­Ï«ÓmE5èß@“£z†uÀÿÁœjå‹÷â‘­ž˜pç^…¯à˜¹gkÅ/fŒï¡IªkۄΦ÷Ão~ùHDöz‚í#ÝúuÔ]ÒPKpyA û¹œÞCû+[™v’fýÓeo/Ë€on©ó"çï¦.|ÙØc˜ž¦Ÿ{£¨Š#e­…8«ü´²MÉ6˜‹ ™å}º|¿ÙˆáÉŒ¿…m¤ê¤[Ü”d¡ìsUµ}þ.Ÿ}Ï:]gCw…æ¿äYê‚Aa ½O©V#šˆÅúî1ÎÀèja‰d e(r™ê„ÜÍ}W:~Mƒ²‰~×Pé"G|WË%èó°ŠßXýµg,êžØåÌz0®Xu7¸/Üv°‰Ø—è~x#a>Ì­í8Hì¤Ì•Ârùi&ì‰wÞ%ñw%~çš3hÜ›¾Œµdë/×Ê ^uCÆè2ºú0Ušœ×hÉ|ÝøŽë+üù}J“ÐÐæDsXþCñxCÜçöPΉ§›ÔruCWçgKÝ£PU—ÆõÊ©é5ö±–6o8U‘Iîý&ˆ/,ª/?į&á·Û\Î*–É Âf#ÎGüjèÒ)Y>û°˜ŠÒõ,Fÿà™ìG²íŽ’‡ñ2öëœwl1„ƒèWËáÀv+4¾¬Ö(üÞ_µ5 n~8Š,úN¿ž÷è…Lt‡ÉqžGYÃÍ+j§¥Š#Á©úÄÈ¡­¼>õš^Ù—»Bï{S)à[ì#æ;ËP®£¼†Ÿ…ˆt/|U5$Í›1푯ÐáðŽrPÚ ék;šBgg§æ¨bt+ÅQFêZ\!öÅ™ÕÐX\ÑùnJþ“ƒJ?Ô"Fíx¿\+Nw–èSáÑJéʳ ›b'ù®“@à}Tóï¹Ø —m>«Rþ=’Œv .Þérà) gCé´1ã~1Íð€ØÎÆØXd$Õ¾fñU·û“<6ob%#ÃZ4šãZÊõvÇWðʃìíÙ{má¬ÿ™ –ÇÕX¢dâù„{9ª¥Z_ëVëÑ¿ýü,G93|2Õ—–E¼„æÏ(HÈ¥ê¯GPj–;ÄÚú¥Zð\žÚ—VIã®Ê½Í0ÈêõˆkæOÚ[Bä|%’¡*VùUéS7,¸ÙÍ©£·<–+æ6fôe¥è©æ†«£=õƯkÂÇd kkÎYþþÝ‚L,Má·%¶áafäqâ/0 e>$HBÇ«vç_¸¸;:u#~)ÔÒ”x,kÃeãEÊJ8å¼^¶ ó»vÞ9ÓvŠÇ­ª˜Ý÷·Cj×ОàV¨TPV1â¼Jœ^;¿*Á’2Iƒ´iígDF´qPŽãŽµB«Äß—¾Ý<•ÿNæÿnÙGéWMÉÀäJR7ö{*šyU!l|´‰)TŒ³$ã|Ty6w#€“’åêéR¼ºÇòi{û”ù^~âåLßÒÖMõU²xÔ){&ñm.–vµÏ\´ã÷®6hÚÊ]iLûãøA\™gO…†]Úî8L±K¬QµäeaÚýKŸ:ÜÝ“ÁjL˜¿Á‘iÇͲb+*‚ åº Ç¥‰Ö©ö,ª6_•j9ÈhÚcz«ëêåRs5ÞþTŸYL“Ãsߦ— há#Å=ŠÌÖò…Á )×u‡þî‡`â!òAaøÛÐMámëÇÿ¥ÒÄ’x’üRv™n3Q­4€>-«Õë@› ¯°ã]næ•Õ< }iTá:(“óvþ¶™,”glbËêÖÿ©>â12J®Éó÷aiäHææTLqô‹½³¹îÅõà~ÊŒ4gé%±6ÉPŽhtêÚOãT'˃Äççn—Ä"¬f V‡Þ…6š@ÏQ‰6Œ£=qAî`1~H€ÃT9—rÏAwÑmp6[<÷¦[ä£[%ñóx.ûëo¢ICÇBíÕý^ôú3*˜µ6+"[6l ºjNߨNöò›Ü- gW“PÚ¾¦ê5V¨ª;Ô°±|ºKúî ëxøå0° "/@ÛxU3+Òôš[ÑFPöJ¬áéÒÉ÷rcµ.T[E±%¶®Lë2Žyc ¡"—¢†ÚTŒtÏoÖ)ľ]»VÁã \£Ïá Ífå–¡ñ¦ã>ªî¤S“þÅS‚4ÐÆšêð²—–»æ[çš×õ°êÖ˜>¯xEc9ôy¿t"öÔ¸N|¢nj@Pr0Ïšù_'ž\±)/á†]2æ3àðÔ˜Ã'mÐ[1ô’¦Æ†"õ©Ó4^¨9¯ $ê‚d­Ð.ˆ–r Úl#EÝ/Æp_ZÁxʇÀ[qþâ˜/áNãxZbä€áuQ]®~Âsü]I ˆ,VûÖ3ó„ ,² ‹ Nf©z›wxyõ®}Žp¯ÈfqÂ?ä¢ÌÍ"áÌ#{ìïqmçèG¥"…‘ú™ðåï’’31gE툘{¼ Ma¨WZ¾5¥ö!Iá«.nèÁúìªÏŠ^ï¸õ‹lþ*É:k‡ rƒù {K'e gw¿ïÍêÎÇ+´”ÅUX,úïk×òI=AT l”´‘+.VgW¸i^ý˜á´â$Ò2S<­/ñŬ`cP1R‡µIýƧ*^V6pW’‹Çj¡ï– 1K‚¶wùÑ0>ó¶Å[Å­#áT;=Æ|Ç:95ŽCg5·%±%•Ð¾>4'¶I†b¯qÑRÓ #E_kW,±|ÞÆ}òˆÈµ{o¿4úŠ0q¾Ð‹üÝ•+Lâ3H½¥c”X  ËPqzÆÄXr—ÕW.Í5òîyTƒÓ”…ºt¯í„«ÜæôMPCQ'˜+}M1}(kj¡lþ&éqd)ö00ŽáÔËØVppV` „©bÅ#:ÌßGuGY¹nZÿF‘½[‡L (Hý„b§‡Ó"`rÕK¬˜Õ/íZH‚a ‡òÓbCUNc¨rP÷@AÁ“>¸7pì20¹ã-h•–N(Ä›X½¢ÉY¤Ð^À ,pï¶Õx;*W"¡õ}:|Lµ“WÊ(´÷V ¸6G_ÂI8.-ìÓ¤X*›YÅ/š¿É{‰ð{EÇë‹×=úñ'õl”ºÀ皩χ`‹&:¶ ÀloÑK¤àN:,\aL6`®á…šÛØä;•Oˆx=šýºñ,ÂSfŸN磲ešCqLl<¦¡©[–õk=Í«Ùã`ƒAäN‘!||“¬°³éœXm#ÁÇÈB­:ÖƱ?ÝþÁŠÀ‹úÅ»ûPG5u°ùƒXd=›ÌìÆz›Çm§‘$HRÜV׫J±©SÙä‡Âl³ëÂZPáÐfgJ»³è‰äß›c’žü£çƒýdÇK»^~Þ‰fcõ‘Æ3yacùÙû—¥/RÒ´ú˜È‚{?RE»¡ï¾ñýä*B4ƒ{föÙ‡ÍéIìy­Wgœn‘E!âÂDª4Í’:Ô0+’V£,ûÓ2À‹‚?ªÓä0´F<ú5ñ΃à5®ÈªÙ* RÌZ:")ø5ô@vb-nb,»¥†j¼3úû“Î^h™ËÜ–}ÒÔS¨ñ–´÷j+k¯×š‡9ª£_¥rævn‹„ŽU†Þi×J+ýÃõ"šÜf­ÄKéw ÇHx´ÕXt¤&ahRGe•¹"—l1U8tʘl{«ð=d@"Ì=¨RÍTSfêxÖïÚ}è ë\×ì"cqÏJ«›«¢M±á¸e! «qœ ˆVç¢`æw(¨)­“Wp¬í®céqPÒ?²DK‰ßºÖ‹ôËø'8qø€žãï©°aéÕâbxqtíC_°BD¯ØÆ3ùd†âj8Â\6?*PS—\Q×݃|àrîËDTI•²×R £(f³…(pB)ô…Ó«)†¢mlÐgß<€2㙵áúfÈÉ:Y…°™ ĉÍæLÿJ87/þªët ‰¸»‡µ5f'?ÞÀçÑ/ A3ýÖ·•ŒõÉПß:Åz5ÖÛ•’.· ù?ˮъÓ%¿L±1èìÿt¤¡Cú‡U1<9 –­Zú#vÄ“z»wíñÊá§pórð´íÊ…«#¥2\í†,Hz©ÿ~š!Ì‹xã ß”uÕ/ïì"cϵ^¸½…ƒiË eÁbŠ‰£ê=úÈžÓ×øTAšîÒ+û¢»ªlF˜D¢”®Eþñg 5ÇV¶“Ôëû®¹—Øœ£üÏëLn¬Ù • ©É¶#¯³Où}÷ù¡ vZ¡—h1l©]_Ÿï®?A߉άÕêþGTñìP%uú0Ôs‹X¦!(H“¡kË¥°yvIži 8´Ë5ršë¡h£›–;g¶`¿~Iõü™f¼ÔFq=AáiIå† ØîRo‘ÞCy’ˆ–0¥÷™!EØ­çÛ\߲ڌdzu½Tt1›¼9©’Œ‰èzÎ<µ¾pÏ-R¼T<Â] ç-»®¯ÂaÐs”ÑozÌçõu¢ïÆãŽùÐ2×`CÂA'ð>eÝ 3FWª¨™¿ýíš³¤ø¦ÌòõȆM× P¼a®#l%Þ ;ø¢Rb4ªˆÔ´×ùsõ¯ÔÅ¡ë§ÇèÅÛ8O£2XV)ô@›!Ö>F›gjM¤?¨Ÿ]æ©JI/¤Q¢¿E “¬–¯€ñ~…é–uÀµkáaž¥À!Ä€û’箊Fï)W¨£¥nfƒÉ*ŸåŽy™P©±)ܘDã×zk`123¢ß&xÝ' ñlZ!¢Zf†Q.rMxìr’Z ༠•6-&P”¢œµ„ÄvgÕ®ÇÜwŠôª¬EÅÓÜo¾Šºµ-^f8@rø©;Ž:AYíÕ[ÃGž¤¯N´®*>êY ¶ \hã¶&Ä(!UkCƒÏ}Ì£INH{F› þ^*Fl6|bqFO`ÔÙœvu!õœ§º/ξP’×ø ;‘û üêHGÂôV¼ç~ëm¬f…Ù”âÕGúù&½èÐ${SO…Üð†¾÷2QI¸¶µGÌÚvRÆfvIäñA"[Ê&hfïŸO^nÉN¾Xf•âøt äׇsicôŠúùCþgÜnnûtÏ­ô­°”òó(í/Ø´ÅX`9E÷ƶ93€À¢Æ}rHt²ÓsRÌ`ô oÞò2ƒ`‹[\@/FT5LŸ7%9Šb;³§¼nûe£‰Öb§!@â¼j¬Ó&Š6ËS9Y¹‚㩘äZ²Sø{;Îå—nÊCÿ…98~'¦‡­)ò‰|û‡c£…âC#”€[ùäú¡¯¢ñš]Ì8 Ôäq >+ÚÄÉ¿@ˆÝ CÖ–zkf$a‘çéH’Y\SàßHiEÀIþ à[ DØUç¾»­MPÛZª`‹ªgÓ&N´Gñ\`_)Å¿Q)HÄLj Xé#JlÆ°e…Û‹š!×ãYŸ1 ‘#ñ7ï+ü¯U*ð‚‡]†ML†Ð¤•¢ ­r÷KÀà)x¼*ˆœz".6±^q—›’Ú›pnþJúcxs^Üb9‹"rŸIb” í‰q±}O@wÛ–Á rp}ˆefÔÛ}GàˆWåžI.Xâ{™Áu›Ÿ´“ŠôXÔ:p X§ù®V…F¦àeÔ¤[-Ùé©bˆÛ¶´/¾MÂñQÜM;ÒãK ´Òm<¥Ó$gÜ«2ŒJYL‚îùW†ÞA…còiàÂýnJ°X„‰ 0Ä o±)+9êV¬šyuìôI‘o ³éŽØU¤ö‰ZøèNÌ£mUMª½5¸Kái„K}n ݸÏúXYà‡Êép|£o1è 9ÿì¼V–²õ+œHâËÍ"G'†Þ G7•'d¶öÚ)Xsë£Çì1[„q'Æ&¨˜£$tR_« Õ~¿ ž€6¼=poò:ð¨Ù­–G“WêBvzŒ„PfeÎõ›*‹/jrÑù²iHé…6/I¬E&žÃ »€FŹ…# Á³ã Ô"‘h¸¿å“^đ僗ûZh¤Æ„°ÖèDƒ€ ø©É `‚Fj«KóK°JïÍ?BÍmœ”ëÚ%·ò—Ln¼²1Bú…áhó±¢S’¦5jšá°Þð¸ÁdFUq6'ÒÀk‚8zï‡s°#\£ZEÝ \Ägƒã)¥ëÞqÎYÙÓ¢,–&œX÷ýAŸ«qÄó¯æ7eÓ(½‡® °E[˜–)ž64öð Oßy{„ä­aQAlü¾TJ;ˆ¨)(DÉ<Ðüºô‚þ¢Œ—ø·#ïƒÚææüSªj3–f›mÚ9C§bm{6­Ò•"ŽÉÀ• !˜ÑÔc×¾êÄìµy3“0«ÔãBgA‡JÁC’õŸc&fèó:¬UÈ\ÿå é±z¤iRßùŒQîŠF0ˆP:—6»,óÛ›€RÑšuÕ»!Î ³#û0„G;Yš5—ùª_(—Ø5Ô]”Ú»M´í/üxæÎtQGwB€Ms¾y=S±UwAr…ÅðV¹˜ø›¶¸+×åv4%ò›62·Çú c»p¤j=sÿÃaǤö‘܈b8Ù+h¬w+NËå‹jîæ—·S’³m%Y«|¤š¾õ^Ž&RùËžäÀ —IŒÁLÔ:† ·d†1›]G¡f¯¨ÀMÛeøÅøna¡^U¬$ÞéœRFo#öwéZÁâ8y« Ê[5¼ªqû½ÐFeƒ´ ë¯0ňYÄÐîùu:4ÍâVðS(ÍHKEVz¦a‡ÎNîPH‚Ý¡ïN¤&ePQÙÕ>U«äíhˆØ„{„ÿêýŒq',&Õíܳ`kiêâ-žº®áE1%…³¿Ë4D¿çsˆÎè;¢QÒËþôÌÏTÑDc}>ð”¨cQè{säxçµ@„l¸kµì"ÞÞüÝR¿m4¬¡Dq) Œ}mU¦ †ÆíÕ¤·RÙÏØ+X”ÚzÊ`ú6²̲Z0Iz ÆÅWögY‘†sØ¡,Í/GœbíE¾ƒgF/g@`ßGÉȦ Äâã±7ΆڔF„µ½ºŸ<øós,]–﨟Nšz»¹`àÇ›kã.cî¹±Ãü.½¿ "rT4:ÁŸR¨1ÞÆézÙ”…a0:½ÇŒº³àn:yá“tÌš·O×H’ÜM‘£pÉ®h+dØ;ˆ|dMqú™]1Ä”ùžPïüö:ýúHg—R»~<“½’è.´)01 òšŠ[Ò;¸8qFú>ûà- Q·hK9ñ ýåïÀ!¡FÜ!Ôˆš!bgõ\Ñ[¼6-•ë°9®aÞoúWp­n A,wžoÜèë¹ —•„Ü(Û–“4=¤××S‘`GØSwP5v ´}ú*ñ¯Þ§µÛÄc¼· e°r©Ö­¯ÕŠÄ¿ NfpýE‰ô¿Óvxù‚=|ÂÆ‹¨ãâæ˜÷вc¦—æ¼Òça ÅÏY3î)`\žYà0!äeí®ÍL ¯˜yÆu-”YPp¡fpÀ/}‹£;Ú^–%'@hÕ‚h³1½Uùýi¼fœ»×l¼FxM1ŠÀùPg=å‹ú{#Á)X¹{H“Žà* ‚ ƒp"<öæ D×j •Šœç5¦»¤9™)/Þ‹( è6Ã÷iãÃ…¡iÙøýÞ'`ûÄrU ´çÎÞJÿŽÿî^ÌY´Å‹]„1¢[ß«ëþXnò[jÂlßGé‘iöˆD>­² ôa% endstream endobj 59 0 obj << /Length1 721 /Length2 11008 /Length3 0 /Length 11607 /Filter /FlateDecode >> stream xÚmwcp%ܶml'c'étlÛ¶±ãdǶmtlÛ¶;¶Õ±Ñ±;Þ>ß¹çÜzU¯ÖŸ‰1kŽ9Öü±©„½‹š§ˆ™ž™‰ *¡"ÏÌÉÂ`b`E  u¹XÚÛ‰¹€xš S€*ÈÀÌ`abbF ˆÚ;x:Yš[¸¨L¨ÿ 4lŒL-m-ö6ön–&>777!7gW'W†¿Eª ÀÅ0³´D•´¥$T’ êIÈÉÈ äjlci³4Ù9ƒ¨föN›;{;SËqrfø‡€ÈÉå/13'{[€¼¸š°„¢‚@\”QM`dg “ú[nçâÌó 2q²ý×tt—ÿZ6FÿµŒÿký7ëäño‹™`jiâ0™[Ú!0þK7i;3{ǿæ®ÿIý%äü—€ê¯†ÔSÙ_´«‚‘-@%joëàêrÈÛ›‚œì*ö¶FvÿAÙZÚxþqÿ‹v1ú«ƒ°ù_å˜þ²t–°ô™*Yºü•ÚÌÈæ¯DÿÄ5Aÿ¾y©¥«í¿ëAõø{£¦öv6žÿÛòïÿ4d””ÒQСýïü“·3±7µ´3¨ºüÕÒÈÉô¿ÒJF–ÿÞžÿùÏ6˜ÿÏ—7rq²ô™þµLÿ:ÿ±ôþ%"bïáMÏÌÉ geæ0³²q¸Ù¸}ÿº&®NN ;—ÿ;Ëü ò™ ¬­Ø›ð†X¥µ†Uø‰ÍUBÓÎͪӮ0ƒ¯N÷^Ñ-‡,L /ñïÔðóÝ‘ˆfùw/v¤ïZÈŒâ@HÇ“S1¤z+øÏ%ä•æ~µ[/Ê"ô¯ 4ðlM ®#o]2Öׂ”h¦Ø²rÏjeËøÑ%™c Ë;û¡Ñ¶ã.>ØIRëQbdê| öikÐæ,B ¯ŒŸ÷Ö Âóƒw‘®)ð , +ÃuŠ7-óØ5 6ŽbøÏUTK7OÐíˆ)Š½ÜÏ3]àºú@Ò·´ÑÖV¿vïb9jÎJÁF÷.—ú¢}}óIL(hOƤ t<–œÓÎ+¨¶ÄIh­¨á˜þÒæ¶"jÊ»‘çøRÿ;Ä=Õų̂½g6‹_’ÄèÕ6Ó¦X‹´0ŠcçKu)fû¥W›\É“…ó³…b¦†½'lEÅ ¹†<"E9ƒ I‰ºƒVN²É†L¡wyÃñþ£ðñøº¥Ùá^J¸­î›g¸JC ­óöàQ5@ÆûìóGHØD i8(lªŸ.tlôO•ºÏ#´J¬m—9C­ŽâR nquÇIÉóo¬°°jGà€ÿ4¨0µ+AùÑ©õ„"•–RD·Û$ë–ZH*ŠÞåW¬jN™ÖƒuµÕàKŠ±~åª>pžŒ¦¤œ¬VoWŒm¦fØi¾èÙÁ'×GÅd\CÛ§Êt•H’$­+,‚ñLP/5  .ø\vŽw¨%˜Íeo•¤ áê¡Nn•àº¡TBR!ð=Äýà¢&í+o.° †.¤ÁBÆNRÉTt9ÛJJF{™¯D^CQòeÐÊÉcfµ{&{þhñ}° âi]ú&Ëщ(LÓæ=¶.wOýöši(É…Cæ.lƒûÞj¬ppÝme•÷rÞ2´‰Ù zöëÖ‡˜K?9šÞp`ÈN~Dù["躈¤„m,°ÑNÌe‰üÙ¨©_ªJìßÉUËxÉœG[7íUgQªvÉN‹·äs'õôŽ*žjQ5¥·¡§„tœà=¡…¢–Ú{?ÅCM¥WX6„  »è¼Ÿ3#Ê[øì"(LÜçËX´± Ä’Sq6Žü¦±9<®XxJ4v~Ü|èBê¼LÒ8¡x‰|Ô‹VšÍn²ãüIzs\±ÈŽº€sÊê¢úú#žB\rîLÉ î‘’ êð«æÙB]¡a0ÙþíضÈÜh#í b]HŠ˜+˜Ä$1Xg$[ Þè6¥GhHªfÕ)³ ï@‡œ rÂ`µ ï—a‡Î5—êhŒU´=Ý µPؤüwBç^-Ê›8)Ä‘oÌ#m$‘0ƒ*ÏV×>y‡šƒËRøŒÜÓ'5ζX÷nÂj >:§i`Hb3…¡]ie·êº”ɦ;zçŸ:OG‡ó‰%ü+¨Õ!%‹”\ßL¨®6óÝôŽ£WèLWÕ‚NàCçrÞíâË À••Dî-ƒ'š&€ðö6sr§Ùþ•A„ë0ˆ†¿ðëÄÙúéi9©dÔ}Œr=ª}¿g>4cN+lÒL—&+ÀÌT”`¸&j]ŸŸR„é"vO^½™¹Þ}-Øn)6½^{¸•¼ï£ëA÷åî2΢y­ƒ‰u?UhõÒR£ÿlNØqÇŽ7îjeEÜTªµ{Gh÷P-JHC‚îy3þ¥éYÀ@Ý[…tø%Ò0/¯è2ë:r0ž ÆÏÑ—ÈÝ×LøKÒ<‚ë×å¦O÷9¦›m&ëþ¤§mÜÀ`½f>÷ÀìsšüzØh9 ÉÆ #jN¨UÒÒaÅÑŸ±³¥JÍY/WBvÓd“K3›‘P¨[LðǛٹÕÚ²üZø‡†¬«ÀZ[À§\zŠ"NòÚz[ø¢àCNKw>¯çôÃû¸ãY£–²™LÆÞJ!û²Ù8ÿhÐI0*éÚÐ:ô§·®$»êTÒékÜÁIèÜ9ŽI êÒ¾¼z(æW ³+ÈeJù-þ!F©)°ç i‹Ægëj\F›·",Ë­o£eRoˆõ2aEœh­bo4Œ°s†Dd\{f„¸Ðn$bÈêó1ß„' 5ey“K›=–¿ÇÀk(–°¢‰«ã1Bâœt™%Þ/J !q–Í4uŽòš~ðÁ®kqBXÔjº5Œ,ßÅ•S’Sîå® ¿HÀ‘c–Ôì‹ß)º%‰LðNBÚxU»‚2yÙÏi_ÚßM“<)žC¥Ø«‰©tJÖMNÃï èæÞeÁK±Yí‹øS¢@!úq97nµõ›)9 qÅ ¸Ó lÆ;€ÀPHíú#"¨³®û~lÆÅ&0eZz¨i›¤f+Åàøm»mw\®.fþaH#ÔLìÞ\z BGw}Pè)¤ç,Hå(´8q r¡hæDšï7Sb„‡£Ñ3uƪ)ß©q<ÞÇmº}« §"±Òq„‚tàǯÁ(§VÕá‘–ÞAÉ԰1xP–9Òë»2Ïôõ{{'ñô‘~¶ð“›ŸViÂlE7çÙ±›Kº&`¢cwÊsWþÇ =#?ÒòC¶92ŽdC ‚ûÉó¤ÜÙêDˆÅКöÝŠ±.…F›ºPâ/g"‹Øu=…¾ª‘s ì–^Ó.íÓK$ù….fwH¯›z„ÉÀ}æ“ŽXKª4;_ë/“¦ë™”ÒçF«\ÜC|ìâà[ûuhÛ,?7ðôX¾F óôR„‡£¶ÝPa£b™šœ)øÁ•4kVÞú Æó©ÏÃ-3™‹¶µA×IþÎÏ\ó3a4½‹”(ož8˜÷iäøøP'ò£F…K ¦É·M±ÈÈ„k%Ó6*ÏÖ#P.Ù¸ó¾˜?I1Ú*¢ªÉµDïØMš÷Á¨1Ĥðbì¥è¦D}â7ÒOý¾ìÕɽnç­)¶ÕDÇð‡¢tq\Ù?Í¡W’wî4¹‘«ÈF9Ê|ûn²Ý¹t*ƒf­EÏCgái¼xÅJŽ0Ôd¹ŸÀ§7ªóH+a+ Ÿ{öºG–wS¢××='ÔgåG¹$xœä *ôn”N—x×B›TH’qꨘRR#ð÷¹úF ¨劒%QÁR¨Øè‘R?=³LëßõcVÛ ÒôNžõ,lüR8ض¶çL q,‡$ygô˜-ÈóÕun”* -€Kî¦"xµ9p‘f2›"i9G—="±Ý`8 ±ÓÜÙðs_}ˆ£Ìsk ¿– ‰8µurh›÷ù ~§9¤Ãd,Ý îÙÈ‘åѹ‡¢OØn$Œ•ÐÆ ÁÚ":e-oxF;´ÙÞÿ’HGdÒýbþ!ð+c˜Ì.x‘šØkØ÷^®>—uÞTî(èÜ¿9<¸*áŠJÿâÖ÷¢} â,A§ŽÆë¨ö¨~û%5ýùW7E!X§˜A0ÅXfCm±€žÕx¼£þÉöB9*N´‰óoàªä„÷ù¢ÛþÖ k¾mó7ˆvGb{§Ðeº*3úz=ìÂpÒÛ Rù®ÉùÄá aFè^”¯g6ÎüF” 6% ŸpÚÓ¾“FÈH]`²ñPJDBǘŸÕ™øŸÁ…Õ-¹ÉÉ€L±SJuù'Ëèüyþ>á 3ªU ÝþÙ÷”ç²%}ƒÉ E/Ô¸uža¼ÕÜñ¼ÍI n5~–î¤åÚvÆμ6—ùÓÚƒFŠ&–\æoöÛä´ûxÀÞ”E'sÈÅe—·¢™f+•U$zÞd‰ãÍn¬„ê·G‚cÖÆXv¦Mqxy ÿëí²¤n”òˆéS˜#¨Ø\d“Åd¸Œ?´7–GDƒ^ã#ô£ÂQdн:E¤„^bdÉ:þ:Ó]:•ŠÁ‹G¡5s²¹cåņ«J {™óPþl¸UrXÁ”Ð÷ÞNê^öMˉðY‘Š‹¾=N¹÷˜g2#AÊøCÁ%brH(P¿p¸W£¶v„€kžY>ÇŽ^´k!\r8m¹"³Zù7ÜŽFP Jl’½Vô ™\G†L¾–C'0ÃiÌ|7—¡ßâv"Sñôœ*pà0ßÞÍMsÌ‘…'Pcãøvz¡¡¤aKíxg»íÈÊ;Û܌ۣų‘\ù‡ãŸÛÀ©X^szXH?—i›£RZ[ãÂmì.ñJÒýKd…ÍK2Õ :IIL²Âò·Å¹§x7ÂM±ÞeBˆc[:‹ Óõn·vÞåÏЙˆóô`|ùSj¶äæSñf; ç5(nš3Y6™ËeérJ@" Zú!ÓU,úš¼ßFžåüéžw¨lXê&ÔX†w_kUÖ=òy™ç,"D-®‡"y)ù[ûÃs¤÷Òœ£þïöWtÏë¨Ñ¶—ÁëÉ…F˜@$ï{êÄŸ|£½†ÉÒ,U©*]È–9}Û@q³~Ù‚PÖ¯ñ¹à;‹¾”[æ½nòÇíR ‹ˆ„a¥¥È,»èhÃw¤~¾8^èðõð£C<}ZÖÃO‰Ã\jªéx·lÃWæ5Ö>yVÂSKçv•Ñ \©‡®·Û%©ÕSååvGQ¬‹4‘¤Ä.Ö¶±ÛFÕI)í–>¸*<ŒŽg8jÌ”9DxÛá"ïHˆ^ØßE÷ˆ†–©u++l#õßMóö¼„¿騤Å/šYÎ34e§gÛr›mÞ=èØÒL)Æ 6èºsLŸdÝÏéŽî¹O·ý/ØŠ£/¿ò.kÓˆŒî…jâ­@4óy©ŸŠ^0ùwX„I È,ðÞê柸­© $á½=r¥ÕJ½•Ö‹Mo¢n"„*@ºµŸ^ ¨µÊ‡î˜?‡ Ks€OI«å2v‹¨p“¸ƒD†„gµ™~£—ô;d£•?ïøj’´xgíF8*í`uúÜüÖ’Ÿé@[`£Ã, Äzx•U–åÆ­"œî: xøõ³·ÉR‡ß_'oÊÙ°~Ão &X Øå$Ùõ9²c&O°×)Ü=A´P ©ÑÇÏÛe1IÁ\öÑæ.f÷ý\¥KP2dŠA}ëØ©ô¤0 ‡£>‹änO§óÛtL%™’G \Lÿ[º%‡,¢yÑ÷×Ûj:nnñív‡òæ7›oäfˆ»üàoš¡Ú¥é¸ÂWŠI1÷‹FÛ9U*Üò´t…>rÛÕq\ÐoIÜ,}L×WõÑQP³åY2a/@¦ù´ˆI0ÂÂÍÃgqÄÈq:“` ¶À4 œÉ5.Ÿ]’ûÈs¦ÁÉi»ÒäC‹ý]¢YäàŠû—ãyl,»È¥ï&™Tˆqål2¿ÆÅÙ¤ª‘ñÜD§ÄJ’°±­×åûO±g—˜¹ÊKÍtÒt]"Ùƒ5®³™é~^éhñ*¤wl2Æ pM$ÄïøíkªOæI·ÑOÚ*Æweº“E†D¹gš°®‘ž-=äס< ¯Îõó(ÎøÛ¨ðªÂ·®c„uô#ölaÊŠ.,2® ë|]^Açcx¯’…Iƒe8œvâËë(Œ§À³Dâ‘7uxçz=´ÉrO÷í.ö0öÁ½ö ©Þ<–SÝöãâ¾4;žf÷©Eitß³ý*L+ÖFª*˜·°¢l‘¼O›e ê(~ƒ[Vô`áÊãm܃oŒI ù—(þäTùÉQûTX|mÜòÀŒ%S[59õÄb‚z¾ümÎkLô0 øiˆR+ŽÝ+~\†T,x]ÕI×tˆ3[œÇ´Î<^¶ôð~v˜øG»§Ye¼­2ál<&¦Å`x^¿“çs›þØ)ÆHð_’bÖhMRòâQ'ƾÐYŽnˆ†– ?hµ‰Òr ®Fð×P±†1à;¾)pÅ×É •‘„”0Òž(›¸Úðä3]¾]×û¹¤y£¯ WÐ ûœ=Å?É5gFB!»a2f}!¸–¨˜úª½Wf+ŽõËùâéͲOh\•öe³íàø–ÉÈ'7*ÖúZ"§šezªŽöÏb‡¦’ŽÙ¦Ç»ô;µûŒß=R}qâº(î-¤œTKéäÃV§)^gaj:LuÅÛÙ¬O¨(ÜÙú^¹¬þ†E,`GÚÈ6VÐ%Á;ꊴÌCÆS„Vÿ7jûúÂÕï44wUÛ(€@)Þ‚” ÍšrQpô¼·<¡h‡Ô½Ï˜Â•MdhÄ®(^áí åUš;š¿üx·ë{YVEhLX¨ôÎzšØÏ›»ð“v²æý] &6‰Û†!È­Œ¶ýaÁ–ðaƒ*Ü‘L”µ{á؉jü”˜¯‚DM²À¸ -ÉÅ㾨o†pDô!± ?Ú…÷‘"`CBçºÂµrc™>`”€¯øN&YªË£YÕæQÃ#Jý<,>°uÜGfüSœ”ß]¨!*Ÿ= óeyKªñA$1Ð!¿‡‡Ô'Í:Ø»•vª»³+'ºåÊù‚a??G¹ø$K|Àôo–©ìöËÍë+È3øHÏM«ü%“”ñ9§‘¢¢àaÿøI|ÉÄë¢ýóè®O5sþkG@5gÊi2@§¢Ì„Ë „f`Û,)ˆÄWƒOôŠ¶Û'S­\Mhe 5˜ŸqºÖ·›×T<ƒ¤Œ|øŽº4#B™ ©ÍGþêRù^DŸÒùÄtŠ#e!/Â찛ɶ0â ®« ÆœÕÖÀ›g?“¨\Ø÷qyhj±AìcßõÄ ÁC4T·]ÚGeçbï ]:³öÛÖß•[¨Cß*gBPÙˆâïÐãd”¶¢Hr¡eÇ7÷@7º#5'¢ŒÃ1øXI#`{":Í  “¦•ªDÚm°ÊÁã?JÖ§_ñ‘Hˆ¯1QL3'"ÕW¾ÂŽÀn E;ßÀòžÉV*îáœg™×pý¬kÊh²½—Ç¡:Š…bÚ9âmد0s‡#G Z‰#(hý¬$°~¿1ÄW ú}7·¹—dá`3ßµã$Ý%S¨ó§ îcµíŽÎ´Tnb®¹1íI>^ð{Á%Rv®"Á(fýN3½%ãsR>M’±Ñ·Â9XLf¿S½AÀÀÞI$ãŠçdXQ9õÝ@&MJ™|ÓLñ™Áµ„ýêÈÕF¯á!R+6®h½àØT ß QV‹‚9¡SåEÖeNætK"”I¿èðtl °=\í %\übq-Ú²K/Y/YÿñŠ~YqZœt¥:ÎkTÁu³Ò “n«¡Y[Áu)ö`¡¥1wÇÓ¡nçç‹ÛÅÈÉIiär’8h|Ã+˹ÿTÔU'«mƒÈáz?íÆû*zÏSYSvÚ’ÍÛ£Œy’¬ØL?¹dÓEâÂ'P¹4Û‰2+C_Ì3•¶–õr½uÐpMœ¾Ë©ˆ-Ãä½RŽž…¬¼3½n×½G‹ãõÛ¼_(ª•¢XêF›‰;]õàÏ;£!¼2zÇŸ¼iàز7m?]ý"Õ`Øž ʧ¯˜D´ÁòaþFM;•ÆÊ–œ®”Ø4C¯±©×&á¾Ïõô!tYõYøîö4½SÝDâ<ÚŽAªx¨NuÂåˆߟsVÏ>ÂW'²ëfýQ¸®é³Úyª@a `Æjý)Fùáïö7Ùr$dÃËrøÀ#Ħ\é$‰Æ½Ðb.«ÊK§¦·Œë½yÑĘÂûp…¢0´Èê8l²¯±â)ÇDÝQüÒ6êï&Å;ÚÑü¥}M •§³ÈŠ"ã•à3<Ä‚ÝLÄ}pö,T4¹*—:îS¤àîØóˆ. QÅã°¿j£Ð—Uº?Qür‚àa(ÔfóGÎaÅ9šÍ¥F!yÝ÷ùZNÝôö÷*>ø3^ÀcquÓ¥dµ&p(2{4ŠÙÑIžçð6¾_ 2Ç‘Måj¾½d\kvô¾B³A&EüþQ¼&þIÌ8À†xàmÎÏ[bŠdsYýèÄ"_•ém_˜†±g3ë=:\j+eÝo®7òš»àÔê®`¶ñó–ÝåÏßWm™½Ñµ[µ­[§ë+¬N„ÔßßûØåÛ7 ,#Hêž=ØÚþ×ríÈؾ†ˆÉ@Œ+1™¶SÈõÍ•iÕyȼÖH,ŠÊ¾mŸ¦×Ó7(42BI³=ö/ùT?Aª‹S’]_«§ni±Ä‘ßãÈ‹¥—Q—ãYl1y ëÔÓ´i¹°§Ý}ÔKt2M è–®OúÑœ<1¦Åj j´Ž8&â¼q_ú-÷]Ó–9*W^zü¯Žyº¯ësw˜o¯•§}ÝåšÎk*×Æ ]ÎDy׃ãäĵÎ…;œÅàþ¬MÖ+zƒ7‰Ý-j«ñï]öifÒB®¦öM| íËç“ýÝn¬Úx[JÕ—›tmïxcÝ/¼ž©‡ëŽ!œPÖoÈšš·ÕÜ!é™Ïm©ÑíƒçßÉÎΞᔇ£'~ì„Þ:Å>Ú6$Ÿçmùøø!¦!“ÊÔæ~ä¡ÚÀ‡ñù7GÚ½}té#ûÎzˆ3Ç\ÿ¢Ä¤½{xûÑ#bÁœŠï&¾¡‹¾&VÂËé#oø¬[c`ª>Q—¶¾kä¬o©—”åÔ«){\¯X(Î8ekißMùƒ°òY;º‚;ŸþOÎ jI²CÑf¤-Vò‘ºÜ»¹”ÈgÍ <þ©pkà ì9)ËÚÏ󞓸ܟK`Ä‚-kÏ4¡Œ’Wk?ø4Gè–OQ9¸øv&?ÀÒm9ïÍù÷NpmfX)Z ?‰Xj?’·+<]A­ô/šWðTOOQ0¤<³»ÅJ?¿ÎÖ¶&é±2ÍG"–c¾!Ʋ,×oZÜra‹?SfŒ˜Òš»¾Ì†¥ÆV.ßhåq/nVªÛ³õ†{(>!„eÞñW]j•ÆƒÇF7&h§ üô¹®š\¤óñrcîÁ³Ë Ó”HT¬š¶`ÄJ{943NÈjýœ þ¬è¼cÒ£ƒÕ˜8dÚÉl¹²¥vÚVÅ7—ƃyÝT-ïø8¬!EG7k߇@? 0þI2†+ ¬¶G3i&1WƒË ‹¥ÇÉ^¬¾Í‘:äWÍ£¤ÒO‹Q¤MÁ_Þ¢ª÷yc¹ÜDªc:—;>¥S™5ÓÖŸwïÉŠCˆ×·|C ´ÍŒY«\ö· ¶eíc¿ÛÛÊjüàh0’dk›°sÓÅo»µ’ìH!ï¢sæ± nP Ïó W‹¡ËÞ­É<®|Ì”ª¦3w\S|*¿[:Ô q–ömªpëj@¼Ò}µÔË?à5]§{Ä•iŒ$ÕÎÇ'—Ñ Ž'úÛ }”û º¦½÷qÜg à²c¾¿‡I7½¡ò2¾Q…Ò”ÄG¼·T‚Wêj m¾ùw^ˆL;^i ³‡]Aº\8pÖ†cÌõS{’žO¿\!°ÝJlTàˆ¼3Ùv••&|–(轈BÃ×›bl/o2^îíèWÀL¶‚:ð¬ÆsßVÿ7Ïå©ðR —MRÆo1áþäg†á–5õþMóS^Èfiï*ÆZŽ!Ñ!IÉ>_U SÕÝ°—7 /æƒp-\Å‚âVM¼Èü(׸ ºqÌ£¬U 4HÞèrŽ>„ËrVÞL7ÝýÄ!•¥é˜½¥KÉ’ŠØ¬€mîtf_³µÅêz(­Zw'›/%ZH<_•®ÏŒ)F®¹5šƒ³5ÛÏdýù8ºÕU»Å¼ŒöÝ2§t'æ“Èï‹o׳ÊiŸ6xIç`Ü"å‚ýXßêŠusl|3Åšo·Ô-* …tñÆ°¾ÓÏsJpý°hRµN¾F]EöŽövâøZöÆ ®‹LÇÌP!uuDZxŽ‘ÿ윌}ËÒ—¶/eÖ¶qÉIø1)/}L!h1AN>3RSˆ:Pb窤4ú<à%ëâø]ÿäµò§st#H¼Ilæå›Êüô5¹òÞ¦ó…—´Du’ñßgâ·Á̶ÞïÞÁ²&›h„LÁ†¤ñ˜ñ7G8Ã:¯ªËK>¹bÔžºïJ"Ã[Y„©ºüãàÃúOןe¡}û¬÷ëK²G2 ù~¹€0Hó^¥ñ'…Ú$ý3¤Ï0ŸÛ0Ùûv½Ü^hÌÁv,? ³›JJùÌrB““ ð5ï¾uÏ«‹²~.ìüº¯˜)I`4>cÓQj¿KçRœëûÙx ë=˜“7ztEl·šÛ ·›­Úû©¥ œð4lØqšÖBŽJ(½¥LžUºþº>kv–øLƒ÷F“•9 sa CÓÈVR¨ÞÕY„w³Bù3·ŠëǼö}ö+ÿxöŽç:ɽöq‘"±$ßpÕÆÕªõ"Uu]…{Ʊ-Y¬QLÓsw<·â;eve'ØIWàèån–ƒlùtP¨W¶–)ƒa„E¾ Ÿ!Øj&ɽ«9“ÔÛ c™Å£û>QɤCT‰%¸/Æ`qpÖ6† Gœˆa}9Ée hÚS™çàÔY»8ÄA˜_x¨oõ}ªtý‘Âø|0}}Ø'0à o]aDîhaáËËÄ:Úàl?ÿDˆ «`b)õ+m!9ÕŠ ßj‚ëJ“ˆô‡Dƒ®ý`˜Õ)šÀ¹>o ˜T_ÞÊè³…1¬Y.=xl $ æÏOepÞŠrïûìÛßû˜¶¨µ¯”qˆ7Tqá"‹(ÊUô‚Ï·ÐL¢®¸ ϵU-Q?¡î&Ùl¶K’NpØû€`{•÷P-;¥ªß»´c­y;U)´Më¿ì]Óœçáû ¸ Ø&PÖ›âw"õÎè7 KrvUåÌó”YZ:Ú_”çuøþ4P³TßB®LÔ«hÖÌc×:v™Ç·€ ¡ÿ@å8]¡ð•Î/K=cÿ*!÷U¯qÙ?…òæ‹âöÕÂPƒž“ò9m!Ñ@vO¿?Ë–móÝ݃½cꃓž7»äPoë$çdYÒJ‹Ž IhiºjËr"T£ê}¤Z(ñšè˜ý„ ‘šZËêü7b5õÅÝœmdD†¢Xp¼ÎTÛ ŠÝ›otj\*¿-Î?ù+MTƒjÅ€°æºõjõm’¦ìwÌr»ò(¤Od§:¶Þf D?jC¿DãÆIš‡,SW¥H‡Hî±Ntmw/ΔÎ~î }ë Zh_Wbák£Öô·£~¹W•~Âaoâ‚çJ*JÒë> stream xÚl{Œ'ÝÒ÷ضíÙ±mÛöüǶmÛ¶±Ã[;6vlÛúöÞç½÷Í›|é¤S<§êW•“êN79±¸½‹ª§€‰Ž‰ž‘›HE\E•‰‘‘‘ˆ‘ž†œ\Ä `äbio'jäà&Ò˜©ˆ˜˜‰˜™`ȉDì<,Í-\ˆ¨L¨ÿ-$R·12µ´µt"R··±w³4± âusstsv¥wrå§ÿ뤹XˆÌ,mD" ŠZRòDTòjD;€“‘ ‘¢«±¥ ‘¬¥ ÀÎ@MdfïDdóCdbogjù¯˜œéÿ€ÀÉåo`fNö¶DrbªBâ òªDb" ª"DFv¦D²’Ýí\œ¹ÿZLœ]þ•Ý"—ÿR6Fÿ¥ŒÿKýWëäñEÃÄDdjiâBd 0·´ƒaønRvföDìÿˆM]þ£úóßøˆ¨þbHMd 0ûkíjc#od ¢±·upu8ÉÙ›œìˆþ…½»“å¿$*6Fv3ù‹‘­¥çÿ×é,¤\Œþ‚"dgþF:.zVöÄ–Îâ–SEK—¿Ø»8¹þkþ©“ÀÔÒÕöŸ%ñù[aS{;ÏÿÙõoZÿÞ“AHLTDUö¿ ñoµ˜‰½©¥9‘ŠË_lœLÿ+ø·ZÑÈòŸn"büßåþÍ3ý//gäâdéA¤ÃHÿ¯vaü×õJï­„…í=¼é˜Ø‰èXþÞ˜X8Ù‰8Y8}ÿO¸&®NN;—ào.ÿáÿÝX€ÀfmÙÞ„'Ä*½5¬ÒO¬x¶ œvvÚPv™ xeªçZÙrÈÂÔààçN§À&Иoçb[ê®…Ä(wô{2†8Bo祔´ÊܯnóM©ðµ1Bÿ LÛÖDì:òÖ%s}-H‘f’5;ï¬NÚ±œY‚)Æ°¢£i+auáÁN‚Z%Kû °G[‹4kZte,ð²»^¼wMŽm`QTèp®]Øxi%¢£¥8ÇÔ%‰™¯³-ºF±åVõ7ô'Hèî’e"ã»ÐòîcÃÈÖªNßȃ·g¿i ]JǨ—iJ½-éy1˜ vAØïQÛ]î¼u™%YÇ Aï ®ñU üJË£ÙDƒ?$ÕÑíQÉóWN§EoÎØ|6ÛäÂÊ.¦òçìE v`¦ùvÇp²"'aÐc'ííÒIp•,o@„ ƒcæAúŒ°vñ‰ ×€i÷½®Îg¶½TJ‰1K–¬æ¸Ao!6¡äö½3U³Ÿ`ä ç¾t«È’ç ð´•i‘§U#5)Døn™ê饚žØåýUCÉ{1Ùv6hÂy™¶Cd>“¹B¼à|ILu†ßŸFýÈÒ ìèꙕ œ%EËGpÚÊ×Þc¥o·»ˆÎ‹Þ÷Ý×úDKì_u¨›»Ò, ô—XïÒ?»÷ S3Pý ”W¨ó¸ COØ:‘{^üÕÁÝ|Ä q;šzl㹘S„D…†.ƒ¶ÒŒ|¢ äDƆx[NÇ«½’NʯœPx—Èú$ßs\v£šbXÊ-~rq$°™qÅpCaÏŽèÿyÚ“+ླྀ °ž±:…•y¼˜/éîvEïàÛÈ5ám$xtóaÔ,æ/ɦÒrד"Ü‘ :Èå®PigDÑB—·æÔÒþ9·b7ÅCgÂ6Š˜‰Ñã± Þ)Ý`ñÎüÂ6Þ 6¤-…ïu‹ðÃÓ-TÕÔp¦aÈÅrBùˆ Èæ("Õ¾ g´œHÐíõ«É M`ôGÅiMIè—™hÁX#õò•Š!ñŸô¢Ùñ q‹¡'4ì ˜_¸X”9$°ŸöùwÇéxŽ­VyeÈ*„¢(°Ó…Ñ„|‡+o­©²)Ri axIì%Ú"…?‰<;Ùñ1É~2NH='^id {]ðÑÅh»Òžƒa!lªøpa/¡@?‹-É7¨¾°¨ÍHÈ[™æ°Ö½ÏéÌgn7–ù‡œhw=³ê*-' (~Òøj±ËZ†JÉkΩˆMö7ÌM2!úØVUÄaù ëdHÙûøHÈIÆÁ#ƒÝiiØj‘"6SM“À([ùÂ@»Zñ¡ðYé,ÁvL åÜ4À¼yˆØL‡k' †Æ@ÉÓ÷ >j†í&к`ˆƒ»³KDy£Shwñ!ef¯J~’AÊð~È} SSÙ_x©ŠFÏ7r"UÀL®ÆÈ\ÿàÛ@q±%+0‚ÙÇb§¹žˆŠb°ðÉA ~öÒš!jLÍØîѤÑΠ)jÕ:ŸÙ)ó¡.7#½\¾D6CÎTXDÏ ¦üC¹Ç(˜øžHâN9æëG èÞ#½=ý^>†1N{T‹ÒbtÏö•mâc‰y@“3ZãV4»ÏjœxÃüe‹ÅÆο¹ÍS||·ŸÅ¢û;ƒÊ¦投o®‹{Z —Zù- z~¡I*¦ºbåƘH@Ü8M|«*oð›ò&þ‘ýX5ÔçxÏÒ;J:×Åxöøhg¤‡ÆkÚ‹½«c5WD•_Ä}õ™iž79°/ÚñaŒ …*bbˆ˜Ù »¶Õ¹Ý,Eq”6ר´°ëOS÷B ýñò¬½/µžâ´!yV@Ù 6º°žPvÉôóN Ø7vðÕIå d½þM¹œtƒ]ŒXÕ³¸s’~G)àØ Ä |V€Èÿ>Ÿ‹ -@Aµ2±y ¬ScNÿõÒ|ø‘^Ú˜Ñ\_ÆU¶žxIõ3š„>ì?aÏ"Ÿœ¹—Ds ® -´“òú¼…^‹V+°U»gN •€°ôOœ¨›Ò&OÅöÐÞ–W#.\W`4Mð€”PTqqÜÓ7™ óÇ’+àB—:K¾Î…¶™ž:K3_Ÿúl3}‡Æ±µkfí›Ùw0P•ÄX¹,d§ä)8f&âôþ‡Äæ ,0f‚4у*®E·õýã[ÜpO8äí¨ç)*¡‚~þÇ·`€wWÍ8#0 Ü ¸çˆø$ôÈyoN®ýɸ ¢u²BÖ“·þW ÇŒÃ/óá÷þpkqûëiïV×cl-îêtU‹6ŸnàÛ9§œ6·aÖe>ÏÈÆi= ÂÚ—dvu¯EÖÊàîåÒ)Iñccy¥y¸ÑY€6ÅúÊåúƒù¸ÅB a°<Ó…Èг\‰v3­fè™!„^Oë¤rÐ/†ëLM¡×t(6àP•sV²«,"Þo5µ¢†™rx|Ð}}=³ƒï°Wª ÞXØ“Ùø+”;˜3©+Èl¾DOÈ€Â+|§BAX½ßÉßçR;õš •~/T¿º¬ªv7I·­s/Mf¨aBžŒ)¼ U_éÊ¥EÓpë´âÐmó“`A¬4Yñ¶ðžÜkMï‚i2¨d [¼öùû°ß‚ˆƒPÜ©Ã- æ£Åß&†p,“‹¿XÑ$OqVðBü°Û³ÚŽ°Ò_ÙúÉJë!€@¥Ä7i¤NU¾lˆŽ· ÍÏŠdëÏ3E Äd[7{CøÌ<2\›­úøæË­^ÑÊ®!ß/€¬K¦ÒìÜWÔ *mÎ5e±·ŸêÂ2JnwB|RŠë®ʱ~ÉÉqÐY´ŒÅ)¥'«¹²Ë8¿“™‡ßj¡-üL¼Ä^$y"+ÍßÛȬ ÔQç ÛÙ@÷äµßÚ Ì“+©7ð”d¹D$œç=^ÀÖÝ<Ac½Œ¦F+|»ÌÕ¢¸™HŸÝˆ—¨èû~x­™f<†I/1ÌJvçJ4Bij ‚vHØl¯æ¯o|:D§ý9Ð[eÞ ž=ueã©:ÚÍ!}[üª¾ÞðnÍ_!\FªÏÊ \)£@ñÃÈbqoÚ……½H˜ï—†=9gÈšÕ^Rô¯%$ ¥ûó"'ÑËXc ½ÐOŽH=¢4•ÓÓ,æÉÀûIµ•`êî‰BFe[NS»r»>+Y…rdE¨Š½¦§6‹Ü¬9ŠËzh­ËØõ>L¢„ZÄ¿Ë<Ê«m¼.ª¬*¯ Ô^}5EdIv¬ÜpŒËÇeSž±‘÷d™EØJEKNÔß”&jg8;B 5ö­Hë™TÔ˜=Fkœ/Ÿdäó}”¯x„0ÉÚÚ:z x—(³“¢_Ñ÷–ø“øp¼s9ظNjTx`%]h?G–O ¬ó‘‹ÜCÆïÚ1«´d´ÑV‘TÖ¶¢äv`:9ëK~k±½Ö4—{ 4Á/ü ñä|oóhVºã 7)¿¿3éx+àg>wä/ñ¦ÍB³D+’0ýÄæÐ(Nšø[¹‘ÔžÂc)YÿÏyÜ5®{7èc;à˜cWŠÞs¸ÏáÊ‹x`¬Yz9¼ÇÓe*Ö´Æ_ªVÀ¨2Ö0}+•@Û)ýÖ“^âåOíîZTàYD±ä˜¢ d¬”¼|¡ÄìoO²±H-eáÞ©¹H;~ºÀpq`#q4°¤bÑl—P/R†­?Çô;cèèóïîÒn§¨ËáÒcíù4rà ¡Â×IïM+tTÞ6”U«S¶²ø(†aIGÒ}ö>HPâW~oï¥é1€,‹[. C_»ÖéMŸÅ ùÉZ ­Õ âjõw'Ô(#$T˽Vfâ¬ï ]{Ò,]|.¯èµÅÁNyU®Àoä±jjÎçš:…êejßÃ9 ÑkºYÇ_U^XGÊ·ö$ Œ¼Lé$¼%ì¨N«Õ >{ØOú#£$¹U汶$bÞâM¡h"¡¡õ‡”èvÖþ!ù]r†n÷ˆÖ[kc:zð¿xt¦Œ@¹â½Ñ‹²gÌ:Þ .ógí–7Ô”çCX¾wk¯|ÿCÝE s´ã‚Ðÿ¬Ýê‘”y½y¬ÖjIñ½ûßF!8&p¡Å'¢JÁ…¢¿–X¤•y²è}8¿AȦþ fp¨OÃkTóÔ›äu®Œå™oV™KçÌ^ §wªß±¯` 0’ÿ`òM$Ù¡vï¢ÐõAŽßëª_èšd’RE1ÿ.Ï“ÜÁ¾å-íõs©©‚ܦ‘…\úR Úñe,ê%Ì—ç{›ŒfýêÇåz_Hœ3ÿ5rpÑr*+7@.¨Àešô×% oùÏ;7f6Ya-ÝÿîÒý0wÒVÉ%ÂZØÐ)[º?µ¾ nó¬ä©çÝÏ>'íƒlc¢¦ˆ-¥=qýѺ÷ Ž¯ Ô…„³ÖŒÁJÄ_BEó춴‹5À©h«ŸEC¥(=³ä&¶,éµbâ”çdU±j£>÷:à†Ï^ÑF<³¿JùUa1ÐM Iê0m¯iÞÂ>ü­yô‰4Ü”àˆùçS2T4[8ÀÞvSZ3hž6(æl¡@ƒ»Ñ›ä†džE`[Í}¯ŸÕkaƒþ¢7®È»S~ÂÜev>ø6¸Þ™V—ÝO.º`Ðå(ô@S^3ŒŠ©(6’§Éž€ÎÔæ©œÇÝ_¼K‡QÖu–™kFè ¨©’¤\®LeQþ&²çÙ~š‡÷EÅÑ–Øø]_rʺúî7]æ†RÑ•ísç½c㥠%=à†Y겊º(?Qaé&Áþž6⾤)&6ǵ5c#Ï\²$N;=9˜ztÚCr&äØœ’m6Y&\LÅŒ+uÚˆƒ:0{C¾b>ÒÛ¸š2èavÚ³ÞŒ¾¦˜‘‚#»W¼ kõÉÖ·ŠDJù”Y¦˜qaŽä­IؽØâánM©œ¹•. Xœ=¨xÊP¾·o¶fPÂB#Ý¿^Ûÿλ Ë@ÃÉwU³~6Áæ×a~ú ¦&ösÀñGÞkd~–ÕÑTIzs#EKolBÛÊBúqŽ_–‘ƒ`ÑFÀ£ ¬úÌøÇ)¶ãº^Ù„ý,ýªŸ)äïDÌkÑ—U檴âðÒUÚøÉúízÜûjfô ä´¿d^…,M/ß‚œ#ç g@¼,›ã):e€VK‡m×Vñe ‹6hÁ ‘mwK})p)-r2-Ëãrµ¹ÐpêÚäO»!Ö*K9û&ß®È]Œ”/Sº”¤š‘f‘×ýI—Aë(jžø×γãvÜF DF”¯*ëÐ%} 8LËŠ23kC÷Ûí[2Âí%ØDðPxñŽ$C ˆz^ §{S|¬9·aµÄÛíL8|\íB‰]âëj|ðì•÷HRNûû<ϷϪ—dœknŸg5+ÊD¨è§¡¬D; ºLYì—©çÈW\—1^^ì—Ž$Æ ‹WYFBâõ>+|i˜´¡ÛcxnDW2Q­l“u¸Ð³+ÏÆ*=0œÕìú_ ÎQg2ÙÂó : ïçP oq ióÔÀiÝиŗäA™5c´ËFã‡<cçš3L{ÇÇ?«âÃ'`!B«›öÀÃn®5Å•/o$L50á¢êúÁT•2àæAvp´¡àé/¬†oû$ºëwH$)i0MÜ©¥Uïþ¿²a‹é¹Žó˜û¹p T@u¨>™Ûí[J÷#Šúäq>¿¬"‚r*…|òQ|‚í™ =n̵ŠãcÜÐñ¡›inµ¨³y<؉ƒ´pÄó‚’.Nôê/T*p>”‡ìÞž¸m°ÐÙoÀ„©Ç튔èU7@í^wÍeÛYǺզ_Ѿ1È…û36äCÑ îkz;Þꮱϵõü•ºÅâ´ö“ö­U0¬—ÃamnΞВðêŒÈ-·*vÙrÐ ©×²Úç‰"^…ô 4AW¯ Lqú|ð‚dŽÅÝËÁhxµ«Ê.Y—þMn]ÕÖú–µ.m‡×=up 3I¤‰6‘XÑãRXÖoDBÊè×Áq¼ìÕ½ƒ>Cà ÷«¨Ú4Ùl’Í×æé=e3t«”Ð:Ä©Ñ ªšQ§5éÓˆ³"›¯Äë,—¨idƒ žù][uV*~-?xÛñª“¿ÖTXŸ¸¨j­õO-.sÿ{‚Y¬o˦±k¿N†ïdBåQí²óB“R*áÌlßuíø|ñ¯~6·k¢^5mØìUþö:¦éÝH¹­^ñ. ZWùÅF|v§Q@-:¼Á$•—­™ýgF,‹´ðP‰IYøb_“8|õ{‡Ul' a7’áÙÆyõy)Ú½w6[•(c-ã¬JKU+Sµ„ ¡ˆf™îf4*º>Â`Èó Æ2hîx”ä8Ê;[ÕJ±TÏq6âü™š½ó?vÿШ¿ídŒZ›I^—le†ºp€›€¾Œ~&ðÔTèê’ëÆŠª5ÄoEÁFˆdjr‚ÃÎüZ $qOLô¨3“Êû`°û9ŸPn¬¤("—“×D  ‡.’]šW ijµåÞCë[²Ä÷κçb8.‰wÉ[lúŠ×°4Å÷äqîú„/¤Ùcm¸(mKL°Vß¹¾'æŠ6t<úz%¸:b]¡oón-ˆxô—ÍÖÓU%ç5u¿G_=Á#Ë#cîâH«{fÕÐÅ*©dù6’oiã{$±êBàOòË4^ö vr^Ä­cºPjãö&V"Ñ÷®Fn«VØø(|™:}7U…ü²t´4*h4Ô- ”=+ëú@1 ¸eË#ä§cÀš&®¸=†í:¿üðßá$Œ'H¡‚®ú±9 Ò¨2iQÚ6ñ£kV¦Í!ã~ÔÅ™j’/ÜÏó‡;ÁÑññ ËǨì\@œvºÜû+<≠ºmêûÏTo‹¡0]íåQ¢Èc-¥’Ôp¾â8•yôr·=ôŸc0¡•¬Nn®G‡Êä<°6<çàoì57 ,‚©“);ÀdL_¬o®³O9I\Ò\Ç`nðQN©s©°.Gh¥œiÞÚ`Ë¥Å$«ëxD¼A…©Ÿ^?Ú»|0ùÁªÂÐ_Ì'Dá$[îøfiƒÙh’–±@ò,™ÃåbCÒ›foÊý§­º EC»ý ­}é_çؾWÏùòÃçÖø7ÕÙÄt–¿Ñf±g¦tu¢ç^ÅGÖ7Á|?)Ÿ á‚Åþ5Âp>'çZ”IóÍM€·¦úù ¯ƒÜŒo*Ø xòB ŧþÙ:I²ô&¬‹­¨‡Ï²´oš’ "ª,…ÔÊIic݃Heæ¨~ç\NéIE» ‚Ïêê‰hyÃ1´WÊNPƒWþöºI‹`ƒ§T/û“&HÖúýhµ@!‹SkŠóžŸÃEþÉÒ…Çؘ2ñ„6ùRxî1.§é€4øð@Dî>\7^LP“ã;ÞƒçPAéáyÄH[ ¼Ï+êþ2\˜Ë¤V¯ƒî{–Y‹·O qœñ»ø]¦®RSñ·ìѼ †ñUØ\ s¨ÆÆBd5WrÙåž Ë\·ØekŽ¦f8ÙÜ—Å ˜ó{mš#ðúl<—¥¡\ýÇ8r¾¾—?H kjqWç_Zé}b‘'oHã%¢f‡ŽtàÙ¿>–„ÚeKÇGPí¦#ªeBJú…MÔ™Øðä ó8÷€h u–÷ìMòÝU€¿§“èÏ ƒd#¸ô§§ú_yºùÖªæm¬ )¬üjÞ'ðß%(ˆ¾ãjY/Ýï‚ú3}èÙH{@v ßè‹$«à…„Q=ôór¬u¼‰ÄA"(ØÅ©¨ zuØyd>Œb-9¢LÉ)¢ÂáW²A‡Ï±jÉfñÊu f™g•U‘dôd„’˜ ©"X\M¼¥å7zÙ€h廡’0!™ïZ!-Ÿk ½/ï…׃’æW‰ùê½g—ƒ;y?OعèôC°Ò_ì2ƒcž2[³!K¹ô™£4®"r²ߌÉÐPËA賆d% è‚sàÐÑ"?<…&á]¹çÌéIxX)„ó r p{ ªpïØùÙ &à=zˆ]í¿BG®åyÝ<ìÄDj÷¨×'ž6Ûé k "/xÙ¥P6>Æ„õÃ52|&@ü;&›HwÒž^¡ý!)L®o{ :’ÎããREË°MÄÌZ,} )<÷wâ‡Ðò_GÚ-ÆÞJ°Ù²;P¸›ÒºeiÍXfc–̤yÜ‚ê‰&ï<ÀnD¢úWh“¶µr–3qiOÁÒQ¬ÊK^”ÛõôžÆ~c)f['áGŸ•Zgxú~sHéB'_ž9ßøœDÜêûªâsb‚0%[:/Ý&ö%âã-1*k*O®þÀÖ/&Ú¬iÑ!p:ƒC<‘=Á¹}™SÊô‹~qX´‡iJ¡KÎA=H’¶ŠÅµN1…`“!”¼ÜÁþúÈ´8´¿Ù²r˜—Mzd/² @§…é#DzÇ…~ÃÒ&-.× ˜Œ†LC‰?HE3µÛ3†ŠÂ ºÆCS4לèµ%ŽëŸQ8b<ÙW†”çÓj1}i¤îîç Îp?û¨ÍíÀÊrïyLG vvøP„q _@7ÄcѶÇô*UÝ”ÊÙ¯±³–¶få'æmS×:L:Eì‚Ò½þÈÝiý"ĉõŽ4Fû52S1åÒ$“¼tÞ5u¸JSÖÍ\ZªžÔ¥º-gq†§—«‰ß^õú® éÈ¥ñøJ̤ªHl‘$ÁÇ'4»¬3Ù‡®ƒ»PU€¬|~Uƒ’ìîsÞ?3BÈËл)mž 3Ûåo-¬mKîvìj<ûÑÃó…'—]CºK º¾`â—¬34 wü¢Î¢¬'Å|±°a®z;;ÀðÅúPX"§ý¢„Bè^W0!ÚSóK5⻚S‹Œˆ]DÙŠ¢ TY4àÀJмF觟¢èÚØñŽ˜+Qå%ØÂØEwÚÚ'×ô ;¡è’hMízàh+^·Ç ÷b¿Å=d9 ç«BÒ†ýWûïv…Útëo?eI_aÖQ®€ÃÌìÍph@ph8°¼ëî®4Â¥ÅtºÝ©v˜ÏO^g@¬èÊm7>E®àÆ3ÞCçp¯„ò*:åS22ËÝíѾš£Ç­+C÷v Õ÷P—«–O5ó:pd±hå—~92¶áÓ‚wfzËö UsØ8èï/WNà$ãÙÉxÚúuŽ€Û“ˆ0H»àãypÇÓ^AÄ:r\ ‘©e­œwP • 6pã@µDˆ1kaŽO¾Ìy¿ÔGI\ÖÃàwüÅ‘&îï@"ëÈ—Ö4}ƒ¬Ì–‘joéfGF™/ %£áÐjs‹Y:‹PæÖ†)X%½J¨ï'ø†™Õ±‡èìí-=àm4 U¦Fg¶Œ‡èÜš)ûlä‘è:e‚´¯“'õ~ æ¥ÒÒ’ô{º˜qº`ýyø'ØÕ/*r;ýàÐÄGyþB>è—øÛÚ'‹Bœ|ñf1 ·Ÿaž ¥V‹Tsb=ER2\î¤z¿Còó.E[îd°õz¾›ßSoŽ£½ò—Â`Å»_ ­hHâM“X®½Ød}MY¯ØËB¹;ÀÜ¡é18ä. uœŠ}ÔÜše(Ã@e¢º­QGªAµøRGš¤ZS ¸G0Se)ب“á'‰ËÁ4¾’¾r4O…> |Ëha€¬«›¹“5H½b /ÆÍ-‘qÚ²Fgšˆ#h3Xœ%ºà×ØCU‹r6°|G¼À]þ/†k¼àåô»øÝ„cI3uÙˆüɪ”ŠdQèdY¥"Èl‚+GÃøå*‹ÓÓEÿëùRlÿÍÞ‚Õ­.±€p¡ Ô1{ Ú8ª§ŸÊé!²8f<½Ø®zË,v͹1;[f}ÏÛx?«qiئ¶„ûŠßvEz`Gy0 ¶ù”þ ŽfUßÝî&!’\ʆz,úpkÄE+ý¼º¡m‘¡%¯W,ØSž¸` ý0ÿî(®ÉÙÑ3oµLu9Tw¤P¤YÄ°ý!ì›WëÞAÌ0Lh!¬Öp“5"?ïòöÕÏ­Qéšnè~&ƲUö2[3s=KÅùðJÙ×ßIó§e°SIge5…WGE«“|ë Ÿ“{%ŸÚ©X¤–ɯtœ ?O’ÁšôÂržÃ`­Íè»l¶F³P̈œÅÄèŠ 8«ûïÊ°º³ÉžÌk¥~æw:XŽSŽŠg?ÍN$—ë[D¨È_GÒ?Ós‹`šª+,[Û'Ì‚ uN¬ðm¯Z?ßC.¦—­,ãŒd1žo™œ^°%\|ºüìæ]tOø¹¶"w÷f3±øͶCýœØäW\p!^`{\¦7ulX¤âRPB¨í̃E"b¢?š;¤w9Ü ¨áV"›µr~¬n/Tbl]>ÒÅ^-J9/vëÎòH©%Û/5<:íjxÙÁ6© ëO $£0xõiÎœð˜ò kšÍÕAùsc ¦üŒ•­ôÖíkàM¼(m¬Ž›ªeȳ¾[§•o¬£Ã/'³·„¨ÞluŸÎX1ƒ)Ä6šé [x´è&V Ûý][‚j][ð‡ClN¦­„Q‡ÏÔ$Ú6•òHÛ®"r€—.QpZ¾Gòñ-¢D™áRƒ6Ì/!Ë–öéé4Ÿ¸Ù2¹~Ú´ ÿ.}ë@?¬ªkDr2ù¹ù¢"ظÏPÿ‰Ãj©jXì m9á+üÛJÝó;&LÝšý9#½¥Ø0¸Üd=Sç!"SOf=xÝÕ<4Õµƒ+«r ÄÈ0|=Ö?YdiFômþó£Ãè£o;±)÷-ÝÖÈ0ãUÌwã…‰!x÷î¥×ª-DÑÀÑ#QÃIÓ¨°ü~(ܘ©â¡ÅT™oé±BzÑ“E‹Œ‚ÁÎXpêRïV!¾À²EgK_ªÙÏéxy›uA®r±*‡6óä¦Ðˆ˜úÏDœHY$܉ÂâzåìWÝæ{[f¨ƒî©Ç¤ƒoAzÎ0,ÂÁ¨Ž@ÌZFö帰£WkµKš[Ø N˜©ÝÙ îuw¯qqv¯zÙ›&ã/±°¨”½Â豌WÍ ‰"Žz9øã= î¾±ûõò`¤‡ IØê-¸&Ö‡\œÎ6÷ä–˜u°GÉ?¾Ø`•ë§>kвë‡9“ëß뢧êi6ù‹éƒª¡5ñ¤kŠ-Rù3P|Ø#R8Ç%Èb™¡”ï:UšÐd€0Eö“U‘ ”4+–Dî:YÃU\M8ú¶RXΤ-Mújåæ8¸¹¸¶a\±‡éȧÓQ9™ÃëùnÞ‰«ŽdxÆd~@Mä£6uâûÊP–1ƒ ÖÚ¯®»(*Îø÷¹ÍbáÝ^rCÙl˜š´'oªÚÍ(:W='H;\ ¡ÿh¬½Øæ㻤èa~É—‘§ðQ|AÆ+×·çjõ¤-•ˆ]®ôìÖ#Û!¥rÌà]a†ÀÎÙ|ÁMÍzIwµ˜¤yÞB ´¿w¸Übÿv$q¿ 6XÎhJ ižÀ*Z݃dM÷*ž\F^ÿ‘ý¤ŠÚr'(EÎþŠB|Šºkî¬ÝQdbÀζ®‚m‰×ê9.h×Êtu*Éj(\W…·WÌg“È®X1Õ6£ÙWkKþŒ=䟧94›d5Í o›Hܵ÷ªŸ4þ1/ˆýYŽä—Õ’ç®Ñ¨Në³}ç§f>,•†!’Í¥ \@-{¥Ð@mÑæ„ø'Û$ØJgX§FEí¨êú~ÂeIœöNjDÖcßTPFpÈL/!‚’Ò\ZPÆ´î7#Ót_ *ÿ –hr@Øâ•ÌÁH‡^|"KN™¾‡®ýÒ£ÕÚò˜–Ç~³ÎÚ^3/O‚ëíè‡mÞV' Ô°Ê gÞÚZðŒjžôY+ Å…I#‡¶[ÜO+Ò0•º·_z©ÄN1Ü£ù€°˜Ò>^}òÒa²ÓhôÎWýE1ñSøCžÅÈ…Â9UM#[?‹>[Õ—‡g®«Í]Z}`'}û.²3î4=;¶ ØŸ~rÀ*€„¯ÎžVÞìBõöNÍÑïsÇórù¾¾Ü‰f_;Wí l@XíÈnæ€k[“i[ Ú!Nxl»ß.½ ñ}#ǺfêEzv'íÌ룻&¼1aw¿ôy5È(ÎI&*çá@!¦³o‚¦·~*©’w=>F×¹pê«b÷ÏuîøYTœ‡|ÆøWP:5`EnzEwàŽÆÇì&øx}Xz‹€­Ș˳è]°i ƒ*º­&tYôQ©Ye9Z¶\ïB½9"Rh(´lvu˜5Ѷ”xÁö ixªØžLzn[9¡e#C‹v[8BZ)Y¾(W!çÁ-¢e×ã?_Ÿ¶1R›»=ë„.ÅUAE'êÏZlßÎ"\ƒÂ9J{¶SŽó6oûo—¨Bè¿°Ž6ŠòžÀÑVZƒm0Œ³^„iý6¤5JÚóÝ(e9K¢¢Üå§ÈCˆBÉ|YVÄbçnÞ.ú¼4‚êMÁ£Ü ë$ýö—½¹ïÞÛ•?5Í_üðŒZ5N+)"vD$Óõ}X/¾«÷ƒž¸Ÿá8t;YÙJˆ` Ÿ`ù8s+ÁÅ5¬9a˜-îe¯ö¤+(¾êÅ´DçvD£?d¶tïf2¢âò~v?S"ã¯F‡Á¼±“A¶W4ñ !@¦L¡yo ÿº`Lq :7©éåLÝgm˜Œƒ_,¦[0Üíg¯Òà;7‘NÃÞË>ÿˆ21“@Ösú"GO) 8‹A“G½57Šé@ä»IQÌYˆæ©üù«Õ$g¼^•8®—´ü´¦ÆWÙ‘é˜RþªŠê`%OknðçYõê¾ýœ1ÔÞ2AŠ‰Š†ž–8az² ûCí=Ù42q¾¡$ß‘ýƒ^W|Íí¯H#ñÄóûŸó ~G Íp×@-›K'òq5×73V cx.3^HF5!â4:ëCrÑPYZ S붓k}õœf…ø< í¡Üü“Yèn^Pб>ÆÑ«û}N»>…qt‚ÏIŸ½ÝL‡–4oõ¢#›¶ ºÀP¾q®¯3A±õ«ç›Š)ãuyê¬Þ9Û4ÖÕs®Í>ïöÿ¨Z‚•Å}wÃ!Eþßs\>Fö`øEaúö «†2K;Nýï QŽtêj,Ùèùñ’ göÇ[¦H–BGê[Äa½åÎÝlª™T J;•¬¯R¾Â·B}µvå h§4}8[í ‡p8œqnðgù˜e†”K‰‡ùiÒ ÞìDÇ ¬-öÆ#£ŸžK£ »³:,œú^ôP욨[ :¬ñ¢–žBé¾ù=8ï©y4••³»áÔs8©÷Hæ9®äfÅ9î$`Ó®`ð‡SpzF…jÚT™·_uÁJ ÅÆÂ{ƤÊ"Z_¶U:E¢ö<ÉRÛR0ö (ñŠhýÚ»_¥Gœäæ6þN´éî^Ô#5ïê+¡¶®,b‰®~ûÐØê­‚KÆs{"Å»ýï«>Ž>ÉÀ8ÝZ•\lòéû„‹ތݷ²ø.ƒw`ÍRfq?™â=þN)!´±ï‹…8€ÓXèŠ:pB’t„¯Ðw(€þi2¾é4Á4—Ç++Œ½9 '5 j®»„qA¾]ëuËlL7©–Øõˆøø—Å÷fÃÔÛ{ªJSG²ó´ôÛ5ûÃc²™ Çö ÄÇ[>ì>%¡+¯³”iÖ¤&9Z~JIM€<ó’4àá¶ýd¾ÿج¦ß“è¶Þ±@â-xj!îó ¦Ù¾«¾çýÝY/ܽ ¾šËó½}õAéöCp&@Bþ|¡¹Ç*$c–÷®‹d`t6^¤Ù.CoAÀëÌúrÀHüYõÃ]©Åúãï#à–ßì"ãŠy (ÛVc§Å´’{ -CÑGør™_Ž2wF¤u‚‚³NµýmmÿbÏþOVÿªiò¦Bé¥i_Áh\Ó†×päw´ˆã©f¡S㬈yáÈÉfapÔÂúu¦G­´vX¾ù‡‰ÿA°$®^B« jÖ2=B1¥Ï¤wÔùRô õ¸™ØýŠœ¶ª ºg™\ËþT„͆«—tw|&+1XC–ÓêÏÈ…‘Ø?Xg€¸Eo!7ñý©?Wg† ·™kÐCU[*.Ú[!W¿ò”ká+É`/a¥™Ý < Õ*·½tº5°Ysk”y/Õ7ˆ ›ñ»|ï52±³©*ØP>Àͧ+§×UcÌiïñ€Ì~ÒÁ ºª¡5²à šyé­_•Õ%îí}Q_[²4ÂåûìA t½ER²õrÔ܆g€}˜Éè¥Dæ Û3¢l>Súíe40$†šV3É績§zŠBðqÁ^Ö˜-6ûZ8{APν¼Θ;}GŒR)ñ 3ôç³PÚ"ë²/I9Znm¶ ÑjŸklˆ.¨-Ÿ$·[Èpdóƒ&&Åò°Ü9Mšj zwö:?S@˜£› ¤lÃ;ùéÒÂÕ^ìÔ©ìMí¼þàíù\O€—×lS[^Òp÷x‡€m…Y3#^·µêz6n%ü¸#\rÇ9lŸ[?`» ¬ê/è¹k ›µîð—1kN y™TÑêi†¹‚¦¹ÎÃà­8ïúTDÜ.Îœ6EÔ£åÓŒOžÇ5sÂPšìŸ¤QlðÑ+$Áb¹û%eÍ]W%Iʼnµ];8Z¡ý‘ÓÒU1IŸJ3ÍRßà]qýøi)qäÕ‰‡Dòq²KÒçˆJC)º‹ðæ_á´e–†ø½ùL8¿ÍÁ{u‹úMe-:§ü@áÈÕæUX@Prk"wáï…&º©î‰2Aa¾ì&õ­‰«Ê?j L6àîØ…ŽóîàUÈo÷¾ÿÈŸ¥#m•f“]²"j–&!!úÝ<ÂÉ™§;@·2¹:b¨'oƒU4~Ö>`QQ¬á†Eœ­ìÐävV[[Q¯D0üÅpH‘¬÷Çbˇ½ð(yKk-azÄ­ëi)_ô÷±4(LöçQø·îS"Ÿá$©¡Y¾ó+däÜ”›•õ%ÓËÃM@Æ·-—o4,+ Ý™<†ãÁŽ>ÄhóGäwÕ Â)«³¿Ð”ž’açñ¶ˆ8M»9öqÝSl|â¤(Á¾™û9ü?¿E:ð]³Ô¦°h¢î=ÅC¤¿ƒOF‰ºXšXb9»å}´ŒÏÆØ÷ÓJŸøÄ~ ÀiË“î>,.U ôFcYÿÚÓ; {I ŽC`¨.µ¼ñ»…êà t¯VÀ²€ÍFX¥kˆúh«^O¯þp»‘Pë ùší þ˜Xûu-]þÁS {b•¶äãugVd¯NÅ"…ŽÇ b– Sš‰JμU güšk_¹2óå@´Äµ\MîÈ]}ùžã–ì–ª)Üg ªn Últ0¢1Ý3ø8>ÔqNÐqò4?bÜ«ázx¹•‹ÝzàƒÕß}+Öí‰K¨ÿTÑ'_`j QÝrðÖn„‰·å7_f—m%ûûÌúñ&_ßoÜ’žç¨%»1y'Åù?°0N–î,?&MZÀ&Gwä8€¬ã°6Qt}·J¹ìEI¹ë4©ˆ]û¢ÊT‘çN+ùÂpw˜Ó³®f¤I….ˆKLtpX\>å5½ÇhrÒwËh\ûi>7ü=Fy/\¤ô)åIÀöƒáУÝ»eZƒ"¹¨JŸ òªhwUcQ§•é+¢úNǧâààæTU$¥ŽsŠå™+gÃC‰íéjý{B"'¨ä.¨k÷¤Hëjü4óÕjuŸnÝÆ?_Z)÷Î(C¯¢HgO;[òÚ!xàöF³Ú*©v‰‘f'S32 od6Ì»ªGh÷5SJ-Àëþœ{èÿ*}BzD>˜©QaJÝ«j— ŒgD±]ßm}¹ggé­{rðçã¦tP¯Ü‰"ÿ–v}ìÜ(hrr0'dµ÷]?-§LÂ2ˆþ6æö×a$%¸¥¼nPØé§HaRT›mwÆìÿ¿‡'xØ6»9îJyá6‰wˆ²²1pôÞ­‰+:„£F‚õR=ËI¤ÑܱcŸâ×ÂÙÝù9¤s¸mé$2+¿ÌÎušKgam×öOºm.+ÃÄ»Ä<'åcé/¦«ÙHÅ|Uv9ò„V=\yþæs~Ø$äÿþD…§rG¸ò1šîD›Øz©(-ÿV±&ÑœƒÃï<ž.(ÃþmF úé¿Oú¡P„!ºÓFø °×Úd¾U2P¯Atœ—i5¤"².!Ì€ÑSl„âŸocœ^îw5Žb Ìié–s|#ütÌÉ[Ù ¹¼k Xéuí”+ÏF=ž{¹ÉZ± ò5ãè¼›sÏÁ^!Ó­„JkHξy˜Ì,7ö|©$sf<›<ÈÍt÷oyÛ.Ý–Ï*Óù$”c“r•%ÓmTÊüÓÁ"©˜¨³ÖO„&bZàö-’šlGZµ `n—‚íÍ°{—–ÚXuo ×/&„Ç’kË£®´ò÷Ï+‚£Š›§0\ø x¸å3=λ…±8ƒœÔð¶èü0’!ÌîâO–ßIB|³vö$J¡a®O ø3îçìDMo]C`u\ `oißFš…—éò…3O~2 rûY¼ï€–móï"ÌA^ºÞ"d2¿ÙJñAT}´>³•$Ȧé%ç(æ#ÓRdW±±°^×óà0&ja’…}ij åÝå‘Ë,y¼º@Ú§}¡i&&1öÓ­B É•ÛeðàŠ\“Âb‡§tÍc Iö6sQN(eÁ÷7ªÀ4œ%ÉÃøœ$7HŒñk&‚1ù3«ZDª ØQx.vǦÕi‘îÕ1ÆÝÂJNéKÊKsdH¸©Kçž$]K Sà^êk9½ Ñjºž üé¿Þ¼õ­Ø ªµZ<ªØâÚW5•b6þ2ªS  Ý1X¡câàu:fÜèPDÝ1ñMÜN¢Ÿq<[§‘ÙB9j²ƒg/Öl5øJg“ð3„‚†‘•Ö~­÷÷”ÛBm„ÊŽMû \C¥rªP³Z)R,É9˜&Ü9Ç2žÃïTµ|öœ:ûôi(Þ0…±ÂÈy˜ž$úez\M¶Ñäصĉ¶» àEµÂ=Â]¥#×Înhs¹®WŠ&‘Š Èûg¯¨» ÄâˆòüêÝAÿ•åF†,;1üI1ùÕÕó iO TºÞ ‰éÿRvÙšZFÕ³ôÅ ¨Q‡8tMˆþ êOuMótì»E¥¬¢ÂlÜ°\º$”rÝlÐŒ%å6GÕ²“>ÔS„cF„dêÝôRéì† ó©T@OŠÄjOYþöó«-·>)låÛJ¯ÛÝ»†)Ÿy·ÈÈ)TZe»–ÕQ—‚'+ÄC`Ö÷:[FþÿZÜŷܲš:ä³_%ÎK0Ôl-¯€ ºì³(ª-ÎE¨«þ¹Ïì´3ŒÔXLJ±ÚƒñŠµ'|-WB^Öëu8¿Æ¡(’DóhÒ¬W ýRxƒpÊ x’tÁ’š@ZD.šï ‘…·öE›¤<»l¶lÖ¿L~ÜôÙLŽWí5±X'ôs¶8XùÜ/ûGÙh‚L(À èãÓ £¢n¶ë„•ÍïÄ1 ¦…î§B©°¼Ž)/.b©6hJ—3Å«îÜâ±fr0V2„æ^šS­U÷?¶/¹¸0ýÃýî3ˆ¶º¤bK|ÑþÄnÆ[Ù:É‹{\+s~»üïIJN€ÿEY]{><¿…ÜÁxȇ#â·)F®Ñ?£]‡4|§…L[ÑŽyB±us÷ögUÿ õ>Ûl¹5¿kT™¹¬Wp¥#7‚iªFg)ózù‡^¿¬êH;C@¨ Cö| 7Tè(8NBÒÿ ˜L¯³Iý3Ž”ùØ|qÆ®ôß?”+*dgxÄÇõc[iƒo……òf|õød ¥Ѭ6ü½rƒ-S\ÐND¤A,Lm°¼Êú6æ`àÃwq«Û-Œ)¥H¿™,=–¸!d LfÅ߈¢Çk°äœ|^wªs¸Mú{……‹½V=HLF–QE½Ûäœi x*’T@ò¨«ÃýÔqg¥^/˜Óè&ê·‡ÿ*uäÀ³y¸§iÅoq*¼Ôìã=NŒ\+õS—I¼Q¥Py ¾&=Aì´×›Ø!Ë8‘öŸz–Ý«’cÕÐË+b? ¡SþØžGkÇ5·Æ. BâmÌ¿¥yÁ e4?'[ŠÙE:Z`œPØvÖÅ÷·¸?±ó÷N rHÖ;¯µÁ‡û'ÿÄol§>ªîßÝFrØ:Ò»b< “-·'rµËΚ= >€ éÍ_ìß'”b¿«ìþöC\:e„ýÍÛCƒÓÌf˜´é¸µÊ6Иß ø*Z†+–RgÐä53b@H/ñ\ú"Õíx¬êKÜ¿û4Vz9±è~¨EéËâT=³ß;¸Ð”T’Âr=3éÝ—""»º¡rî¦ >"Ò9W.›—G¸imŠM+Šö",à\Ì–ê³’¡`°‘×H[×qšì!÷êp@‚ò_°ÀbvpåB?Ez?ðIO7€ÀFù1,. ÇÍ•Ü.Yúp+e{ÊNL{>ÚH{DÔÚª^îÑ.L„8g€e¬L}ìŸôWøu]04Ç|cÕ^UÔ\’nLd ÁRê[l°[S]G=˜{s{«g¦ú5¸Áù˜[ó´ëv÷t-€ÑéãÔRí ¬rÚâ¦?šgû˜Çžß˜;Ãë¼(v î7ÿØ®‚J©Þcæ¼²ÄüÅž{ড়g¿ò»kÌ’=bÉ/Z$¤½ñ¶z ’%|ÕüÆ2ÿLh×SžàD¦šÄóÝ"—‚Ûûê=}zûø½¥Vß4¬…)<_Z3äŸ.yš!$‹_Ù.±” }J±àÛÒLëÁš¹”sœ–#©¼~IôIé£ÓŠ8¥ç¬êÔç…áŒÂÎ -_!Ïá9Lm6a쯘°®„ž°òxUMu¨„ @’Ì4SLOפåóµJ¬þ®›ÈìŸû§dóNÀŸÚí¤LRâk–JSšSö¢ñAá–õ%þï“uû`Tta6ÿïýn×Ó’.–žà/PáØ(R]lL“#‰.ôÓèïDôbÛ¥¡¡§r»-©¡ç&Œ®<­¼‰ë·.ÔÅAN›6¾/²Ú*n§´óû¶æfHâ¡-bä–šü,y±;Á*5×’qRkl²ãÁ¬’Þ؆ÝÓñáÓgåVŒš—6P`V…ÇŠs¡7*ÕàWµà)!òçùÒ6!X÷H³¡»zvdÚúümOTѼ;Ö`¿)ÄÏ£Q8œ+m¦¼=ôYÐ{ÌUž¼+®-22œ²(‡Býæ£4ßLu¸3U\ÿ¶“ðÖì’µº3.ì‚Øi„–I¹ZØ“ÿ¥"«æ«{;bYX~eꤙ:¸ÍŸ\Z‰WD&ówPeì÷>Qð¯PÊ8·L4Þ5¡3Î’Ò¶ävŠðêÌ[ ³ž”ÖèÈz=VÀjë[²›Ñzé KõšNqm<Øo ª¤"A«J·À.?¦öÛߤ“Ç(ö@G´ÙݨÝÏÍZºÐ ¿œk3ˆRËjñ–5 -qn À[>€ÿô›úD›0SŠšqm!²q‰fA®pô&/kç\ 9’œuD¦©q(PÊÌÊÉ >oÿüÉ_ ÷Ê8Ü É€ÉŽ8\“2Ãü&D>Fçš}•ŸÄT´tõn®¥ƒrÄ@ë—Ü=ŠL¬²>ÚKk5)1î:Â,?zÄò%K%èòrÙ“’aóTC¶Š=îç´?®H1Ùpß u‡‚îsÁ+K¦›°§Ôý•òÛÐ0d90…„LHðì«É:‘\ãÑýYoÙÚÀ0€ ‡Wù £O•5Óvz­Ç¾}{@-Ôñ©íú±®Çθ£º¤Ñf-ÅêDŸè)æø¥ƒ0Ï–¡Ú镲êB(Òjv.óžɳñ–ø,ªfH>£ðËeõÜ•C,Ê”@¤hÁšªÀ¾¿‹ëy€Ÿ£g?¥Èƒ’O³IÒòHZ ãףЦEÃ=¡×X!­o•ÜÐÑ,³cšï6…¬‘ùß÷>R.¹Ó»-«-]Ò‚¾Ï…UPSŠ]S A€“D§šNº+ÁÔZf§ÕŠk ³Á˜Ð 'à6–KÞ5öWþâ<à¯!qE–åÐ_Pì0{xÿ@(N~]–¶¯Çò ¦‰«@˜Wvÿ ˜ÊƒýLÈ®B-p#ÚpQ­oM/÷;Ž‡l»RÄ»ýdyÒl`%SÝëĽ<>f´ƒ69@x›ê‚¿<,Š»ÉׇR$â­Ó’aòÙÄ7ú2ñ…Z!i8W^\Bˆ‹K‹ ÁØÜœõ‹Ð$~Ùø?b®Ôwä ¿?ƒöWhµòÞrg;‰íVºÆ•ÀMÜd1¬1ƒŒhÓÍ„]3{U/ÎXÐsAEÔÙøy¸€%ЩÆFÙÄ}P­eÌ ”d(z¶LÄ¿t3¢£‡¯ýüçõõ5ǵ0ˆËW‡åX‹7ÅÏFày^¡¥öNf`_J"ŸÎÎÞjr UÙS¯íQǪ…H.šI‰Ù¹Äȱ4sð5b‡‚5fÙ¿·÷$F6cFÂò)˜qœâ¼çóyN9„ëi-°,:/ E&žªB>ÊôKx‘ÌyôÒÀ >Xƒ c Ø(fÿ–c¼Q¾®ànúŠ¼õ"Ð ¨WB›¶êA¯8]ßÛ…i!'ˆÐí?wMà˜ sNƒ$5f(ÎÎ3Ìpek„K|¨¡Í{Ôy³¨f|b–NZ§Iˆ7Ásúý¦'Z‹ÿ#ðͪãû—LÁÙO~:c"UbÐ1Ç‹Ë\Ï8æ «ÎPØñ¿õ‘9­Q©kŠü0›c’ä6°y"Îו›¡2e‹È{Å bvðŒR]¶$)±ñ…ÛÓJ­=z)Nm2¬ŸëO3j‡Of¬{¹øÛmg;-|ÎPvZÙšÈÈO³†ÀÁV’Žßl1Ÿž5Ðn²’@ìTd~Â;õ¬Vœ«™~?´Z1¼iŠBn!õÁ¼Øäôé]èXã¿÷©U»ñÆØç¡}¶BŽ fuõ_`n4Êöv4Á™[r¿¼fhŽ˜%¶Ð®¬v'+í'"(oaÈI~…40¬·UÌE`]€pÑÛxÌv”3¥û]HT[$pG›åÌ÷ƒ¹®sDÿCiíìQ^ Þæ]Ík‘7l­ßË#tµ;çGÆÓ8_P£Ì³â„e³,?é1™áÊ`ïóî´à) )5ä Â~€÷áØ0zj2@¨uGO6&“²µ=mË0Á­ŒÕŸ£ÞÿD=#ÜÅÚ§>½™w]Vª©÷?8D £௪djäÊpë i\‘Mº  ŸRÎVúÜ8mïS|zj÷A,•C‰Bx„3²P?D”|!G¯…Ø îâ (½=jvË`­úS}Tô!¬»±ìÔB û\œÑCkfAÌ>лþ*³ ˜#©ßo-„y Ý v“Jt`K¥L«¶Ø¾úD€ “lP ɨrÜZ¦$×eE éÿø?iÆxèñƒ¦á“VFõÛÄ **óežTœÜ·'+Knð‹Œ¹~¤T†ô äFúÈ辇›ñˆÉIÆw„ï­kð¿‰óÙ%—v 9!A’5¾eé³Î‰_£âWg‘w¸Õ~f#£hš¯iºÚ êŒWÖÈS}´¤A]Š_§ä¯®O:¹¦úZOºõÏzcªÌ"1´ˆßÈ„è‚d룈Yb‚Ø'蔾PÈ h~;ºÂ­2úšš¹« ¬”¤s?×£ÅF›Œr‹ƒªõ 5#=÷ˆ „!ßát$ܳÁFE.¤¯žŒlZLiÒ¬d¦Óãñßnmù¬×]Ÿ) ±Ô§Ñd *¸ËfÎvBv¥ˆ~ȱ±¦ˆ³M8B¶8Š0íÑ®EÀ3x-ìÌVÖ»wºÔŽÈÖ•Žò‹öc‘6kY³,OÅU„ò¶ë‹þ€µ<ªgd¥nêeÔÀTIÖP) ¤¸Ç”×I@!eybG3MÝr{X Ü%oÚtä¬Íóá[;5Ú'?¢?mv ÚÚøëäƒDéí~–¿|Ä5Á4±í)k¿°Ë¶è4èqs|Pv­ï”Je$K6Ú©tÞþVf¨ÍŽk7k——•({ÛÂl?ME˜:›câ 6%M@‡ö+• f|!黡þˆú¼ST'EXxPÕ+ª£ 2©7m6¼¹*¢sÔ¢Rn¥@­¼ÎHɪÍÍÞÕ½„ù1ÕQá’ɳêÑ“¾–Ž±ÉŸ¨0Ý‚7OA$ ñ pN£ð 'î7T¼o’­˜j ¢sìþ1›ésA‹e,Úײ·Ðüm!ÞðŠ{Ùr÷¥)WÙ"9°díÛÞbÖ—óNʆ éâEËaßÿcG ¿†Qªëƒ‘µ(”–.¤m_íÖ¹ÄÚ–Çmȇ´ÆµÉ1Ú”uŒ»”]Áv‹œ˜~ï'6Fó¦e†Q•Û¢Á™ C& hc} x~|É=²ÃÌÂNIÒHþ[¡òŽ¾¸Y@ß8}<È6GášvwBfq,êýÁe@©hâ6÷ w`g‚à[[`¹$œ@˘óWÏ™ïo3dïlwBcŽ¢ôwkäl9»ð?;yjPD…ðëÄg³Qš(¬²·BMãÓ¸SÁ'“ëgˆçßߌú•(_Á äÑô±‘ÅË\â~ŠÿNi?µçÿѨP„z]—>…­GƒJö»œþøfTÕÅ’m0C4Ìí‰ã…dëîÆl($ä¨Â›µ””צÏ)lJXÚÈÁ|“ÉÕŽ‡w=Ž @²²CBÅa®è•®À£ëNâ[š˜Hý¦ßN¿aK?H™]¼´²ã¡»%¼ÿW_1 GíÄs*ÖäÌüÚÕÍË!fÝIàÅ Ü4ŒØ ¢lêZs·™Ã¯324ô‘¬Ú‹«*‹Ã2œ·qDåx/è´p‹’£ÜÏ>±Tüš8 |"de—0Äç©Œ±Ý­nTõ{|EX^–A¥+^Ç 0â¾û¯¶nö$`’·3©TOñh®ƒ˜ÑØá¹—ŸÃÞY‡ïsW§+h;P÷ÆBäúåÒfLàÞ:Óð&«…Ý/èÒÀßEa*ÏeLd.E8¹ùº«’%ªB«ˆj —JIºC7&ƒ•¶†Häaé¶Iô­U#öÙȹ6w4\\ËW9>¼ Hq¨ä~y̧¼U Û"¿ÎÓ+†–7f£Ü¿ ¬ÂÍÔ·°6Â88†f6 8ÅDsµílðDè5';Z»(ŽúgÅÈÝcm=ÈÿêLúõŸÏfðAAö:Ò}?mBvC-}Ï4Ô–’¸«–íçºÞ)Túmæ×Ú›'õ«)n-çLÇBò=Ç-`ý&ÚÅüƒ6.´ÉQòço >=Ñ+²¯×¿ÔÓRq7ïõ2©›ÖM¾“ý Be#ïiÙ*Jд¤Å¼ÔXÒ#™šõ.Cÿo‚fÐÆàajmAñ*6VÜ\›Ë2^),ψÖçŽ †"\\d†ÓÄ«ªcI´âðÈù äU^üWÕk¬²¢¿’äVÊn¹…Þ˜7HäÀÑ/¢Uj[Ïæy®½´<Ë÷~=Sÿƒƒù¦.…¦\M¶{­äëÐÍC'[Äôñ´:ÙtõJó4,·8ž¸ËH¬òkq¯ pýB¸ô üLX<‡Þf:õF…‰wO.ùÀ”ÓêE.#ùe‚VÀ u¼hGÀ/,,i € ‡eÿ³àY‡Cõf2Ä™¼ØÎ8o»~u0®<ªÏqÂ|ÌHÛ2©ºkrL%gq%”§µlÿ éQã“œ$S½^§è=UEçN„~†^7õeï¼Í¡’1­º<%ƒÓÿú&H6>hè Ô®Û'4Ñ;‘€ò‘¹q˜ÿñ€D1Ô"͇?÷ÓðQŸG ¿–Þ [òe3Ï‹6<î&£le ûÖÑïhÛz±ûªAÂ[·Ö´áKÅv[À2òf°ÞËê&6Ã\9s»RNTL¾@KenÂ4½é&LŸ-½a¾¥‡N¯QPRdA'¹ ´µ»ÚµãÿÞáåhm~ËXT¨VU¾’D †uÒªÔìˆ ]u…Üæ\y¯ –ý92¡h}¿ã`8Õ»ŃŒg ¹àö¯äÿÕÒooðæ–©˜ÉÜAû€ÏAÐR•Åz'ëf»òðˆ5-Šx^©«HÖvª*°¼·È•¹Áõ“ÜtChî¨Qlô¾ÛS Q¨š»˜ï& aPë™h (BØ£é®a¡žlf0•LëCi>ŒUR¦%çM¢iØ|¾´ªÝÌÁNs˜I‚zÅ8uήôÏÞÖ£%û»Ï;¹zNµ!a‹ŸÉâ4ùb,|^\Zæ#lj®þüs&êÕµ@µòfôê.ªMÛÜý\ã«\fŠRÔç`ŸœØÞ“½ü#ãÝêËL©ŒË+^yÂM1oéº;JíAÓ)¥7Á£ç)˜éTNdQX»¯«Oš£jF¨k Ù spÖ¼™pF²ªóÔ%BÒ1VƒÀ,ßîúúøƒ½¶ÍÞ¬. P{ro‡¤oÅÜ[4zh†Q\Ž©îTï»íY¡B­Á_Ï^îZÓ`•´7Øõ‰×æÍ-Ü™€‚Éöj÷øHýœQÉ59Ü— :g@Ó=ýe87î%îõ’F×nU­øU©±:7¥ÃºʘòÇN@ð+ÏPŒÃˆ³ú»‘¾Î]C.šJzqÊeGãtjÂÚ•°úRJºÒOÈTÉÂüŒWÃ^%ˆnj÷|%/È«×8Ó¶KàòäêW}Bå6Nä"wnÁö¤?ÊØaj¶·˜„߶³#Pð†–œr·­½øá»ÿÈÌ "³…eúÓ9qõùßm2gŽŸ–ßKˆ‹Ú)®iý|–Å(kXOûHH'øbc$CºÞ®0ù~ñ< ¬k£/—$ï†øyÓpd¹"bÎ{Ášýr/A.O‘¢ª×Ùf+â6ò³Â»…FŸfðQïh ØOnMÍN*¼*æ¹pÐP Ó¿Œ›V°òàïd'ÐÑ’óKîxk·ÓM /훨áGö.‘HÅ/¢`ÖT¼:aÑ^îÙžµk¤ð’ÇS0ø\;ZA¡ïˆ˜Æ´†c¾ ŸÄš+Oû435õУﾧ´»}:u—²¦d¤Š§t¨Ã½ù¯Úù(_O·ð—Ó oi×}¤&Žl’)L­„­ÜÿK´¶ùÏ|0˜ÉðFá­žøÎó4&8<-&ÐåY¯¸¢0[<¨Ë_Š…§ö:7” %Õ<ÔµrƧ^…Ò8…Äl*VnÙ o¿;•8ÕKö\Ê0]MÍõ[®e2ÿ e_)ÞÝœÇ4ý®sY’z$˜æ–²dQËè÷\ëŒ(šp1­‹Ø{ýù­êÊ`nêÏÏÏ«E3×[ûJú±\_R_·‚ô½t;ňÆ~Œ OÕMÀ©@ÄýË¢¦¿3YÝ“²å&Ëj¶V"‚·‡éÁÄRÈ©àH ö]»û“Ï“×#ûº¡êI Ñw<ïL/ªÁ!÷"ñúÝøàv ×&6’«R%r˜ú¤T9ÝzÕ§¦?¼• ý&Ìh9);zàd½^æᣮ×6ÙáO~‰Š4X˜›b0陓®;<оãk^M#WÙ÷ò-f^©ÀÇÏó’ýÿÞPN÷’{¦–Èu ZëËëhø¼*~[¹Ÿâ±7ü£™ŸÜ{¸u…st¼5¿•³ÖþN±L,F"KolÒ1 £º?êà XºÎsõ»ç—1ÕùÊ ¦Ë.놆 ËAü’qÝñ”)Ê÷P^¬ô‡QiuéÌ9ËëpŠ`JXn¨¹²ÀÂzõ>«.þÀpÈr^öå´²w"üÑiĸ¨¶ˆ q[ +sÜr0Sô(K³Žÿžð«B·mxw݃Í3¸÷wRã\›¤©ÿ‡f™0þ¡,‚yèOg7kü—ZÝD¢C=. C†1Ý5‡t:3$OŒÞ"s2¤”;Fe!1  %|jH4%º0í‘«€M@råLl¢NH>+1k †\QÐQ¹ D¨à=Ëë¤.Mi&3Ò?w¡ý~oå!CéuCìÅ.ëFÄó vC)ßvnɤØyz³@ý×2]Ï6˜&`·?ÁÁ;Ç–B_l|(»KÉb(@< ‚VÉ3`@8óg(·^BàyC?ê¦jâ‘ÿ.T ìÔ‡\=%71›õp|œ¡é‹QBí‘«DI€}g mÊ¡•ƒ§¡š[ñU ‚î‚l$z½?+}£ñ dt ivhÖ#ŠØL{ÚéÝIµQ¥$јšë~3/UáËêͯÓì—t8ØmðîO½Ï¸ÂÎnlƒ,â»XµY\[Ó–bëÉSh ™ÇÙæ7>ð¼ÎÌ—ø°A[ÐRQcˆé†G€‘îkuYÄÐÙËù Õ|3Ÿ“se4,“F¬wóŽM ú D9ÁWÜ÷kj}øZlÍ¥ƒ.˜OP/È!n7@"g—ÏU%,W'é©oö ¡òÛñ)ÝWaÁ…+Æ)¯:ºÿë)ü #ÏxJ }ìÃo‰‹˜»çKú7–B8«ÁŸô¯"¹oKŽ¥Øcë})¾@ŽÈ)'{?!¿áùçZ'…Ŷ¾ö ‚(O²Ç3|R0@Îéʤ眇fÅVwŸK(XˆM%Œ»úk¡‹ é¾ïA|Â$ºJ5¨O‰äÍÚ‹öÔîÊ¡9âÏtÓ· ñ0Aû„~²hç1ÏKÙ+|P<蘥-MQŠÑ$+šÅP€þó ¬•=z׆c¶•ÎÄÁ¡)2Xy#Ú«k¾Û~ªùÿð@¤Z“&¤xn©Z¾{¢õÅlÏ„…çË]}`KÓnìlœêª=²*+ÃѤ³€¹t«tmnäédÆôq±jã1«¡|ÖT¿ì»s@H±5½ŠŠJôj8w»×û&o£(2> stream xÚl¸ctfݶ-Ûvž8©Ø¶mÛ¶mÛ¶S±m;ÛNÅæ÷ž½ïÙ·íÛ¾¶þ ¶ÙGc͹Ö$#³·sQñt0e¤e¤cà(‹©¨0200è˜aÈÈ„L ],ííD ]L¹ê¦&eS#€‰† lïàédiná 4¦ú— fchbikéP³·±w³4¶ð¸¹¹ ¸9»Ò9¹òÑý“¤lj p±0˜YÚ˜„å4%åÄ”ârªqS;S'C€‚«‘¥1@ÆÒØÔÎÙ” `fï°ù·0¶·3±üLÎtÿ`çfêäò03'{[€¬¨Š ˜¼œ @T˜^E`hg‘ø'ÝÎÅ™ëŸhSc—ÿ©ŽàòÉÆð?’Ѥÿx<þ-ÑÁ02L,]F¦æ–v0ôÿÛ¤™=€íßfW‡ÿuýÈù|Ê8¤˜˜šýíjc#ghk  ¶·upu1uÈÚ›˜:Ùþ‡{w'Ë,ÿjhkiãùÿü"$] ÿ!CÐÎüúþm²t³ô05Q°tù‡o'WÓ›ÕMÿÝYSKWÛ§›þÃÉ?]5±·³ñü?+þSʿ֣וV—W¡þÏüË-jglobigPvù‡OC'“ÿþåV0´ü÷ý/–ÿ(ãÿÕe ]œ,=Ú tÿ3" ÿóü¯¤û£„„ì=¼i™8´Ìl Ffv3‡ïÁ5vur2µsùéÿÔò¿ú¿†ÉÔÔÃÔf}ÅÞ˜;Ä*½5¬ÒO´x¾ œz~Ö@•z…xu¦÷FÙrØÂDÿè/ν N;þ\ ïÞÕ®ä} ±a¼)ÜÉÔt Q„î Îk)I•¹_íö»bá[C„Þ5˜¶­±ØMäKæÆz¯i–켋Z)Çr^dqƃŠÎp¤„µÅG;q*] ”,­o8Óê¤y‹Ð¢k#þ×ýuð‚à=¸2l}‹¢Ê@‡£p­Â†¿VÂÚš ŒÝ˜ùÚ‡ØØ"Ûaä;nU_qÃ[A‚÷™÷!2¶ÜWŽtž èGÿ@DßðÈoíðËâËßrAæ,}JÿæQŠß~âe©êò<Ú›s{ñ›UWסpŒ p™¥ÐÝ‘ú#J Ä&{ä=f»Ï•·!½ì"ã¸)à=ɹ9±¦Žò ¿Òòd>Q ƒøwt{TòŸk§ó¢wgl^›]2!¥ —M¹K¶"P;°M“ü@»S8á³0èñ³öv©$¸Jæw GBú¡qó =X»x þD„ÅÓY÷ƒήgÖƒT ñqaKæ¬æ¸!oAV~ÁäU¶ƒ ³F02úK_Ú5d‰KyxêÊ´Hóª£ÑêǾ[¦Zúf©†'6yù@AÕp2·àš ám~ÓzItBoÊ:äÔü‰ žÅ;glçÉÝwÏË´ 4?XÿÊü¥¢‰5ƒUNÈè±Ò‰R%rU1×<'ƹ.¶ýË»U]&“¹W—ý&w^Ûï¥ð6íE ¶fH™ÂÌ7¸{p äÑ’ç~ÙQr\ýúÞºg9`£¼”éìì ¬sqþ$ѹ«Âð.I@Ÿv݃î¼ï cñ75Ýc„Ç-ç¹þÞ±;u¼°TÿžëÈTê¯Qñð_³C²ÙÇèè5*Ý} À0¦Àû«Ëîˆ0¨ åáó¯ï {èdÝía­³¹¿2¦õÇžŽ–¢©Ûix(9™®^S®ÞíÃGÁÁÑg¬+Š™ŠWZÄ×UÞ£w³Rv=r—j›Õ<¼dDýru³€nì8óF*r îb’“x€}0;æP«ì¯p­å‡hIu‹trPûÿé‹o¨À73®–‰`N²¤ÑFS^d¿2ƒ†¨S‰«‚h楇ÉU§BªÍxÏÃqSü¢ã➫à€€¾1Ÿ±ºVü˜ ¼çh ¶€sLBØèûö³øcªÍæ`à㯺  >(.Xú4ýÀé›qBr£à‡“ˆ¦^->o‡<Í–qU]‚3Û•o­ëËCðÖÇÚ›«þx "ODY&—$~cÜĨü]WT´\=ºb+F–ñ¤loK±[ŒŠ.@LÖ~ÌI.òäUeà W|‚êšS}ƽɈ6ßbº~MԪѼUï¹@Þ^é„Œ#bÞM´M'2!»FÐKšéHuÉ·¢ËØö8ø{ëqžÂHŒv]gü:ßØß‘vY:Ž@ü¦tÞ9 ‰Rý•°½_Íœ ëä˜Í*gŠ5o|"Fwo?oGòl1ë,Y¤#ƒÎ׎%&¾.·¨ƒèÚ8õ«æü­I5‚€Ò,‘r¸§Ô…‚B‰§cE3´d‚»yÅPa°.÷kÆÔBìÅö,J}w5hZ¯çX¢‹pá̪ãàbã¯Ç¤í íÇþŸ£Ú“lú%¿þ¡ÞÒŒ3¥©fÃt{±dJÃÅj:ïA’¬ðQÝ·Ui§)*e᭘¨~yNe¹ˆ.¥ÞMlÂÁ¶6ö“ßSÚz5_E™¯y×rgÔHJѪ³]™Ì9 ¥"²=&¦5–Œ ]@hhäbBÌÀŸ¥k×^G4(^æ¦_×ÏoˆÈfûÏ,?¶2ú»]&lúá´Æiî£Ã°g#bcÃò;º¢}–U.)æsA^Ãhb™Ó— Æ2¾»°âQþ£T7¦êLvh=­ö×YoxÇÉÑ(Xκ?®ã®CØ;¤•¾<¿~ Ø.Ø°lµlx ½Ø÷BÄãàÎû¤»BŒâß]í ¸5€´½Ã3ãêB5ó6“ª)ÛMíešÄÚ¿x£ª<<ô®BPq—xÏ4^ÚÉO_K•î\ í(¿ò¥^ý “°"KæݪSY†I@áôæz‘ñÑþQ„k#á¤ãÏ«?Ðêÿ*S%•:pnc]®†ºDÜüȺ¼(…»&M¯Oî²9n¬v«¾Qñ2-ŠÓ)é?úU™WSmpèKŒeZ³yþW FØ^ÿÌ ®ƒÿ4²¬‹•óŸšבuü ѸdŠ0º3œÖÀÚª,s¿ê­Y¾Ê6"h*Ü6Þ êøÃFUÞ)áB`üE³_¢ói“^¶]sb;çÉ ƒ¼Ðl¬ú€³§3™9ýVÀ‹qhxWÑAFwЮÀ|*k"Dªº°¥IàÖG:4Þb •”.\oø/Anâ¿û YÚèæmËÔÚösÿ.ŸºÈnG>D+L‘SúÌHÉ­ý±fX­­æ‡•¦Å°sùl*I?ÌÀ^%dsúÓKðÔäú™1ˆýë•Ašš.¡BÇéø@±„’´ Š^†úîæQ<èŒ×ÄŸ ÕˆXÔCy£! Y¨ŒUE¶£F¥¬n;uƯ©é®%úûodÓRš}3Öʳ0’â–©¾\´ÞÛÛI-H5òöf®¯É-Mªc.«•#£F—vÎfËegîWd'üíu0í˜ZTåç“ø\Mg0a† w £ß8C×hX}©óг…óCC ˆ´Ö\~V"ž«QJUq»îüVSü”ùÐ¥Ni¢uªvÒ²[î.K(V‡j‘g(q ú--wþ.z¤I6Yž°«´d`ºcvYÙË‹ûPâ°5ñ­3‡úzr±D{Q²8F!jB}òç3R³8= 5åÓ,y°qCzx0~¡Þ·~²Î#œo][•pÝó&«(ÜÀÍnv¿þì›Ƀpðו\,M}ÀÙgÁn¼òü\¨RrÿAíµpôIû&™<@Y—­™²Ï٠ǵ۰éÃÛ~Ím£›c/xTl›²²(ÀÎl]´þX˜ã‰Hó2Ç ±‘vnµ‡¶XËOº‘ H#‰‘Ë XŒ!÷*i8(c°Ý×G’øž§~Ý8…:FµÉ¹}®ó9®©ÎÉØ#|ý"À¤¶°K"÷â¥:³M0àÏï×v+ñ-˜$à"síå¤{“ã+’µeï§)(¸"³ôóºÌqWÊ¡ÿí„P“É& ›‚Å“ñÅ=ç7ß¼ÁX;ƒé€Ô²%´!ˆì’„y5láKI~¡¯¶$O‡¾Iî{l‚¾YìX´kâ Î~¹\9b½yöš¦5”x±Å?Iš¼\Š—ë4”R'~ðMÞ]Ìì8Ïï)ïÀ@õ>;ádêT;ÌßF¤L˜bp5íë¦4Iˆ>3èRŒ;–&äºÍ³LZŸ4¸½Út.¨4Û9&ÉDîã…€uÎÌøq`HS¡Ïöm'1dQWÕ®Ä$5‘ÙŠ]1_ŽY¨çr³d„°B$<Ù=Ï)bï¨lïïC£[±áLÀ¥Î C|f ’6Xa‘˜¸}“R-ÐÃã^¶n§a!þÜù2 üÆœë--ŒY%,ÑYÝ'*®0j—´C~ ¯1÷±J¾'½]hgÕÇÕÔ¬µ“ŒYẅccꥆ}Ñiˆ1=)îeÊ’-G”™æqá1-¤AëSö_Vz\€Ç´¦L¶]±¦:î<æmâ„Ø>DÎ!ì¯JäÑcЋíðo)_À ]µli×Î'‰!ÁXç懤ðv9 €[²T£qÍÓ"+3_ˆ6Ñ!þžûÑqwÙºL‚õÉ IiùâÀ=à€ÚêÊg8üd:‰§»nï¿Ú·ˆî®¹iÜŽüBaû¤q$戩ªÁ•ÿÉ¢m‰ºdhéZSÁwê8œt `tN¢GÎ/ƒÞzò{|>"¬Õz]û«§eïùµ¾c,ç"â‹íK–çÙ^Ú×Vˆ=—¾ª@óÂÜw|-|nÙ÷‰û:¬ÒéD8ɳÅ;öu¹›3/n$ŠnüGÁmž~ ú½SI|µHq܆¯O­a§uòÌ":u×þb ÓÌe”òõ#{«ý¤Š»·ÿÀF$„jqïëRª†â"êСù|ü›ˆK]mF( ªõt”3ÎH†”½…ÁªF%—PQ‘U£(DBuBriuÂ,%D7{ò€¢GŸ¦ÜÔþ5a«ÂhGïþ½+÷'pþ5.¾PþdSUi¬Ep9{ËÜ­ÝÿZ¶Œ‰£iÃÜ*~ÁL ɤWPa*uà•«$2cÆêcJ™  Èb8]H_[Ç* 6‡3ë‚©Á)ORlB9#Š’±4º®~31É,«÷=þ@¸‡l§@⟙ÜÝ8!¼$9΃îÖœszDËÚÇëjÿ›!åï¸j{Ùa6öÇDßz¸A¾Ü¯çñR+&{ñQI×pU ”ìj4iÕË.O”ŸåܬÐtn“܉¶†Îä?áDç?‰vÖ½úØäå-µ@x› †¢—±Be8绀W˜,ù°éóøwh0ŠÊ“:‹¿– ±Ö½®ª™UñÑ…ºÇŽJ•dq±/±ó°²"ë²¢^Í f¾}aå“æ<#`céM瞃­GDÇm0}fÓ0òÖÓð±†Ô†¢ü?ZŸ¦A9×;Í`C×úÍzXC¶ñ ìã,W6 t+ÖÖl­[$NCPh®d Am#A:Z·æ×LÂÁ7`Ìæ«OÅu*ÝÓóº ²Ø£("e\'a™tD«xÛE¼a°G¡¤¥¨ï`cn u}p—–Jå_ºžƒYÛ®"jÐæ  ¨´ù$ÂLª­¯u˜›}Ö·if(¯ÏÂÖ›bè@­¬eœ±Â¶ÞŽÇt‚hUÆTˆHQéjEêqÂt_køùUÑ%ö“(GWoÝxîûQ¤¸ ðf¿‡.Hèc¤/Œþ°0@ Öð¸×ã–ÈF¬µë3#›yjMëNQ\Ï!}å1’ÏÓa„B&‡«¥†ªs¯cM†yΞ¼äÛ¬*!q¾UqŠKÕ1'~Lj%õPƒÏfĬØC>ó>b® ≩ Ê¢²öyhº'†û´VØákÂýF~ÍÚ|h‡´¦‰>r½±± ñ¡7÷²%¨#†(™õ9'ÞÓÍÄ*S:P÷?ˆð6ŽŸÛq¸-¦J8îwÍ‹Ï:1Òw½Ç×ÿœV©I=ÑËíu*]<¾gåË´ÖylæÆæ’cwLd“E|½lEÇåÖßê®@™ §!È׌í 9Lvƒð·ÞΦð*IµJ³Á|b€ EãznUô¢ÿÚn*÷ˆˆæ3 ϬéªtÚ±5H;¬>õõóvÂÔnõ¶¼¥bÈ©âRÙ,ë§Æ»8 ~€>ªV¦È _z{×FÝU{£%eºNYÍŠc±ä5âY0¤1>¤@ÚxàîݶÁW6M<|ð— ‹\`£% (Ã)µi+ý³7™s#îp½Ì?öYHð𒹊ÌÓA׶%Uì‰.úhøîíó$9áálýÍ_†r§ON^àŽWË×qPþ¨+†&;ã!ˆD+³›°‰Þt=‰•.©­Âüm±ë®›ÇF!ЪàoW¥ÃãÔ=`‹.(}kÚßúìeSŽÜ~¯zí®ÆAßMßßd&<Ãz]ªíý½4ÒS§2Ýí©© ÷ÿª¾¶V|»+€”Ùøù÷úáA+§»‰j*‰@€Ý½ø„IÊJ±º/ j°Å-Mt Â9$抴ռ§–¹¯œšÈ%¡=ëMʇú+ð“äã(çX7 DÛ× D‘Bo% Ì£÷»xONø¡þä׋¡Ò>R¹¤WV.›å>E»‘‰ ª±Ò?`lIYv&ùXóô%û@Xc‡—Ã5ƒëîŽ]uD´]pÔ'P{•[šÇFF|ÎãA ž,Ö™Ôå׎d½´˜•›' ¯<-êÖ1¡'IT+¦.k\Oé? ÏÚJÑPqÍ°äJƒ<¢Ðýd œ­vžE „C4qÏxZñgÃqÊŠÔ¬¢¯84ÖÚïU|CžÇ,&ï·Y%§HY8=!¶[£”™oÆD“±•µiuf&¤>ƒ±¤‘ºWé„×Z$z½‹¦ÈŒ” {pòÑõóg¾t{'q‘yàÂ'+Õ‰º&5æÆǃšû¾uŒ…Ë Ë°š½¯ÃÒ^†(©ª¹šu ¨]é.}žišðqp4ô3ÑL9ã°Y óÄ6:ŸÃ¹Š“ûY-ƒÑÁŽæ…ýõŽÒ‡ðØ ã¥«fÜL,+]slÔ©Áë9F^xè'fàœ'KŒªï™L²zYêCûQ _­,_o׎‘¬^}(ÃTg˜´¯ôÿ-áec7Î#¿«ÆÊÔš×’t¤úAÔ*þA¡ç]®/RÓÝ‘(¦û®÷ØÞníùM Ô¸½À|/³êÆø±¬.$Õ:èâÊ"ò~áÅÒä)n"gZ7ì,…ÅtI_zÜÐâ_X¥Œzò‡Ðüã†Éúò7«è³ØíA*ñÂýHà6âjZæòºÛ.-±ÎSD©‰F³´o3¸.R“ËPOþÓFÚûïPTNM‰U%Iþuáù @ÝÑTGl‡^ãFœF"Ë(§ü«ù”n+5ËmHËþö4v/¶2;±à¥ªíÃ9‡?YÛ/òU”_´-³™€l¿Z}îÖ‚¹9&ƒ=ôu÷ô—¦³$¼¦«´F^'LS÷È3 8p[SE¼coAì^òÉîöè²ÈrÉ<±Y¸&Ge§âóº$.P¿ÙKýÊù¸¬§ò•‰µB´òÿ¨„EÚ•D& ýHVx†žõ^ÚlQˆF]ÖÔ?˜Ê8Ç=»ë¼á}Ò™Éo¥—’پقOè¡©i¤!…k4¾ï\xPd¿fåÝc1…‘î>h¨F܉,ÌïºN°üuÓÙ g8M£í²‘: ÅStÅ\óÀ²ådÌM5ÅÌZ}«[Éö¢.Ù`gAGùܨ”"ƒcÆÝGø«l¬`ø…ò®,4TJ‘Þ3yÂêz%Šêº—wÃ[¥×6Pî2Ÿkpç2ƒVPuT:êuë‘Ëøeé ã‡ôûï8#³!Ó|É Sì㺠<^ŸÜäçëG¸CÐ÷Ù}„ܵMî˜à4¡WW€È‹Ì‘ÐmÁ¹u[<ªßUVïÂõý`‚ÿÂ@G”NtT ä]P%„‹¹aÇ7^,@—ÂèuLÅqý<º8ÔèVÕÑÔ͆ë‚IiW`ð§ G3Z àGa{‡Ž×ˆ›8 æ+]]EB3ä¶ –ŸÏº4MK'³)ÿÅ©ä|sö0 K+7\í‰+ÑñÔ¤YRƒîV@aú¶H’Ä'Gvåk}ue5 <ÉBî½€RmÍTMy1sZ¼‹´æçw¦iÀ-Ïœ -Þ•±;f)pRÂl=}ì\ªm¾Ý‰á ÔIZB¾0¸Í×èýÒ‹E‘©º–R*¨ Ø.x·ÙA±çgeÌÛvÁg@€$³³¢I~7í|™«¾K‰]«ÕíÂò©'hF™Œ xÐÓý,uùü wbÆTRÁSª`u,½ì‚†%dîvûÇíT2*^NXçVJZ›Vãxƈøæ·ëkëó‹S‹}³\ª1=jóhKÖ €ä¦¿}é’x¹òÙF>ÈŠ6Ê›—:F3–ÉSìí/‹§{[¬ŠýG·k’Ô¦£ðåÊX­Ææ/Öâïò‰hö| üIRŠŒ’.öp 7}\ÒLØ5c óºs‚ìrX¬¡ßÑ´ÓÙ¼ ž÷8ãþ•,¨¤é“Ò 9 iì•+—v“uâ²0ô†6üxðÍOuY3«/‰sºc6úZÈy•×1Û%ÔÞ_Ы@@ S¶FØv†Œèñè‹/hn2,z1Î&> í½:™ŠòG5*…s„þâ4ßØ vo.¾ð:ÞÛ¢ªiîq™˜ß7Úâñ{n1˜Án JçéH‡¤Â kŒÂhì"³Éz¸­ì ~ž\âö0;¿ºãmŽ\!삹U5v<ƒòcÅÛ€Vƒ Øñ 蜧 9åÒY²€æéqº#˜˜K#Ê=ƒ6µ<$Ô¸G££ßôÒìçÐò.&Œ¨¡&¥Õ†¸SüSÄÑü… dôðÖfê”]èÐøPå{zÔ¶êû›8éýzrŒˆ¤ˆRèñþâÉRÛÓiQÓW;0iO]K{òóÓÃ+¼Ôhö·¨.=OEh[U±2ž9‚¸š£.Ô[öe(ú|ž»o˜Lö;þ>q.q²×}Do ¼2“âöoÝN`ïÃø¨2Ñ6Ìš 1éi¹9˽ZýBb ZŒ“AÅë/ Û(8ÌÞ‘¯ˆŒ‡›™qö?¼ÂL¿¡Ú£×èîL…„;WQÌá VÐ( ¿*?2i[Ú/D·Î ‰á&}ýE‰å†É)[ë`Rm…UGº²Á¹ ÁyIó¸*´˜˜Ìi¥Ìª¨SÅoÝíNtÝ¿+ÒâµÒ“ƒ©Æf=$æbA~€ „:¯ÇóPí`<šìtßÌ•4áÂó·ç Mœ q=ÖÅp/¡‹,êÌ ÅÓ§'å÷‹Ò)Òðr÷¯ej’=Ù븴wgU(]Kà;)æ®Ü”²ÙfÏ£®îäí2ðCq3"NB|Ÿ¦ßsÙ ªñ¶Ö;¨X/5›öZuaE`¯/Àå^;:ìh‰^Ð8Pwä·±JH»RÓúQˆ¤éÍ]*xô'‚YüýnõŸê}4Zàõ[~‚šË¹IÒ S†Yâ³l×(¡dÔ‘V\ôÖž[R_*â$‚©ûûN6E>ÐRšÔ‡¹ë9êVï¨È…æÆ°ÂýÊo•Ã’r黕¶r×3úçMs‘‘Ãä9sÅ%èeRÔ ŸrZÿ˜¨GQ0I)nTI+äûgî2Wì#¡",­òå7Ž `çV44-3;!Dy,áhÒnî&Óž¤|ùLÊŒ«ç#½|Ú¹.Ñ•÷î}i±k ñ®¹r­Æ«r5j­¶<`'Ò¦MwGe©Ú—y‰1]9“k½šá$$CW 8»"çþ:"!†³ÈÑŒPIA)œ¥YvŸÕ˜£_3ý¬ûÂCmr|É·ØŸ@¶_¹Pk€b î\ŸUï÷ô‚#,_~O6ÞõÅÚµ^_ ³x„ïêbÓ¸¸¬DRŽ°¬ðcó~8·-Ž" &ÀYInúK¬"d4Ë8ÒT¿Q“Û[ãqäÝäFP´}}Yƒ“‡TÌÜe¨«œÜISØ̬î"–ô]-KBÆÖqÂy){ìóïAí2¡A†–8/æRê2;|#ñH¦œÌз°/š­=^½ªL EüͪqM}Gé‡|oW'Þ‡cégs)…;1Îòæ4ƒÂ¬–|Á}.4´Ö)‡ý®"0¶“úVô`‘WÃT^«jq½÷6ç_0©Ç¼²Šš¯ÉKÜÒ‡V‰ƒ@¿Dj4Û꟡,%$l;IæDœÒìo$o3veµÒÝèŸ2e쪹¬»}%måK‰cô’¡ d\๪§5©@D|¥uŒžøS¹6Ë"X 7­$ÉDKí7<‡ý„ãv½Qá™ügæˆLÕì«èÕjI݃øD*Ýp¾ õÚ±wjo|™Ÿd“ÜaÇW†íÒšnD °`U߬ÍL(ŸQP×X¿€Í&g’º[ÿÅ7fTÞÙ[14b;—,NöÕfŒg…d@>ùGX5„_,}#n9Œ X°“m%Rh‚䵧®šX¡MÎò'Æ—üìç{9ë =Öy’Ä° TAÇ>*Á[±'–[p¯ä£ýáyº[CÊ~­­8=†M}5š&_™;8mÓ?¬5½“žTî( nq«Ž¨InŽ]bé§y+3zͽoƒ3í¬}Îÿ!A¸ô·PpÐ#£ÉØ8ã¦×D$– V}ô¥c «qÀq4ŽŸ‹”¾ÍsÔú-•Ûnä’¡¨'G'ºÿ«x©Vyê‚¿`Þ£—ÈÕþi0tôFŽ—nXƒ/Š{af4¥¿ ÅÙvÅFdëÇy˜Ñ­q|¾gÖ1«ª¼ Ô¦Ž{:Å°fÝD.à”ï]ŠLùž9Åø‚°4H!h Q%2+v¾Æ·ÈÉÙ)ž%}dèÜjÞ¢õ<Ò†§5‰á>õîz0KB Üüüþ,; ~óù|U×*7ÞU¥P™‚OåÎ!{«Ìq2‚€,#(æºw²J6˜D9†Y1 ”t›Q)t;_”gYž·ð*¶eD-5ü¢”kó°!Û”®òЮý-[Oæf)¹hZFšöÃŒò«dÀïO™%ŸòE§@I3¯—¹ \nF”³Ž’ÌI÷­`-.8.jy_Èsd&[ T‡«ßü[+Pâ]Wi!®Í‚ᔧhÅY·*tÊ°ì?Œ,„!Ãù˜Ýš‡ÿøÁEÈ)ïY½:l‹Ø©è îò V±Pn%Ï þ@‹@õû!£ÜSü$' _>4Š‹«–}ÀPœ¼ÚM¢ô+¿·½ùMš€ïMËz–â¦JäÅD{YeSnnØHs´)›d¬hª*àézÄò¬ÄóÒÊ„U¨¨ä‹¸±*¿§ MòB{-l­äwÏv\‡ªä6Ž¯J¡±$åaNqÓjÐÆ?9£&ð:d4ÔˆöøGÙúvVVs#ÏH·œS‰xà¼O"Eäªá‘¸y°%BÊ7;ôn»U>“<š]T{ö÷’[UmL™åŽ@q‹E.wŒ,-"z@Ÿ×|"P3K5ÇšaÄAr5 ÖOº yíÉwNÌãñ àQ~Û=-†o×ôvIžî·mÉG-ŸÄ–…²ØÝÁêÕ”6]µ~:q‚aÈ*ˆ9ÿ±Ntþš8Ù I¹‰‹t I½W^1ÇÍéJÚI±&ÎðväZΕõÙé‚Ù’ î=Ç}}|V@5š§qÿ¹Æ ªõr˜jÖC™d'Ö|‚ªˆ9èAJä²vTÙˆÛhEÏoÿÂl«‹¼%Ô=‡‹…Ù­UùŒô>u‰äQðçv’| Щº:î*DÒ+ ,κ’¨ª™ì˜çüÓc…EDž ñäXÙø_¼ß2½õ¼°éQD'éûœU‘^ôqëªN~Ç.‡b< %ªëñ‘E¾Á˜‚ŠœÀ§}‡RJ³Jžaqé^ºòµf‚ã?Ï^LöQH¬¿u‘Ī¾E–­Hñ¤ÃwŽ~¡G©{–ÐQ£°ËA9Ü^>Õô…M­QE5d>¼l/¸3Œ\ê©9ÁÛµ14ÏÙµ”`O jÝuî×ÄââáÝ»™PCax6ùoñ›¿ÕÈÀ¹IJ+"Þg¥Jž!¤â,¨ùZâ«# ì{¿Ú9ÙÁ©æñ®ëß%žQK/Å “jº™˜ —°ŧ÷±XR$ªwVÂÏ*ßH’\fð¸SžÏ1$iVÒs3ÌËÎ!$M’ƒÒÅôdìk³ÒýjvÄÇ`A¼ç8ÚJ r—Ã…üÕýúƒ>ª$%³ï'ö”rœ Z¦³Ãâ7 ªÅL÷.0jÉ ´©$ ®î±Yª;¤ûÛ`5¯È¥~Üž¯E +}}ïÆ’¹.ßç‚ÃAF ¯ºÊZ6èvÔ êi¸H‚æÊm?ê’e:NQzvÞ£7 Yö'Bo†˜#…ˆú ¡]œ¯ýÑ€ RŽ¸'sã ‰atQ]b F'ÓuAä>-öÎ\Š;¦ÕãQqÃ$¼Âîéñß–t©T„·Ä£”ΰ§eŸ¨ÃçqF’Ÿ&Ï* ˆ 8Ò“¾ÅU/ MØÂ`Ó‘·yÔ9ãûDÅ&!äÃTQCU¿åUXa>YŸ«Ï*DV‚5@¦ÉUŠ†.Ö©‡Ÿ¹ mšÎlVáh,]žÖD t¯Îæ³³¬2Xeô)’xrÄ爺A•HREHà´[™ÖÀX³ ã! AQ©(ÜácÔÀŒÏ I]ánô–k$$_£ò½áÿ}Õ)qè0aèz¹ÛC×àõËÞ{“o‘À”e~ñÔD 3àqs’Eèiyt›N"ŠŸ˜Ó<a†ÿ7 ¿6-pmŸ†Ìܦ„Ñ=Â^b:–XW½JéøF0%¡ìëêøF8þU2/ÌÞi~Ígo2â±Tœy—ÙÂ^f6ahº<Ðu¡Ïc%Ÿ\ÖLüû›í¤:ÄŠ<{‹¤eïù‰…¥“tP+~ÏÀè½!z™j‚ \tŒóÎdgQW¾#>Ô…d#sч‰ÚÖ¶ÌûüºÌn¬³Ù¼–oõÌmæºLšf Žj‘0ì–|K³tºgž:V÷#jéhÁÖM/M<½X”šZ5Á%Â*ˆ\GiWökOÃô ©eÌ%ŸcMuiê?* É‘Êk7%&PëkqÔqÈéE‘Àyñ«c›yž °ÎµZ±äë®pƒ»æû»S¶(óüÜ.¹•¬—äwÒ¼åß«7×–˜­ºÚÁyedWu•›èëIŒOM$›(¥ažj½€.Œò¶žxËv'ÕoDMé=ñ–Ñíd:³B&E‡Þ¿‰õZ›÷°(lßëQÆLwÃb|ðºË:«wN-臧`¢/¾ìž¦uhõÃÙ­O”îZòjus9aÏr­Æh?€ŽôÙë¨RNVãõS£S:6Hõ§ÃN}Z– $h}“j:æ—<.L¤V%\U(rÁYXä2 C’w$ž•ËwL¢ïmóX~5çá2Ãuùï”0pÚ‹pÖjP¹ö‡ÄV‡©‚8–ëžø¢‚bYQÒJ T Sö|Ðá$ÁåÎ0¿N4—àG³iljAÿžÈ³sc(Øç*IÅ)†Abp¾@¬Ñô4ƒš^¨¿þž6¦9¤:(§ ÆäK.ÐÏѼWÝ´^³ñ·4LÍ?b¦©.ÁÔ2¼ŸROÛ]ƒ›w–ï¥>3fÞ––Jõ]_ <†^¬Ä[¦ôwLÿÓD,vgß+ÝRFF²kWxc®åônE=G«À>ÄÞ…ômwïî`¤8Ž©i+*´/”lð…Všƒ“ > ¦QL"L>ÿ)C¯×æ†\&2öýœ..Ó«¨5;‡å‚å¾|4Üø7ó°&A²u}s &y7B·cØH.vþ)®£ºZ²©Ø‘ŽP«n5â³á^tí¬ o#Ll,¾’ohkÕƒçpƒï¯ûE÷±õæ‘tH*Ž“%ådaó(84lÌL;mŠ4¤ù@|ìÕ…†ço÷fw¿ÿíbh‹ð1@„a9Dq#ì‹Û<ÌŒa•Õ?K±>èÉ5 Ý¿ý΢Ëܵž…y]6VF²)jbúæê5«í-æŽe`ÈÜqYÄâlJ/²ûÜræ̶{µKõã&ÒÄË«h™dãÈxj¨ÊÝÛx¤¹°öß|–Yßò2©4ܤÔ\ÎDI¸zbÐ+nl©. èíþ qRuÃé SEtD,'¶¼®S † zxZæõ§Ì.éå€gIÇ )Ÿñù÷¬;ùþ6Âk–&ãÔ””Ìbú!@øçohR H®dgu#Op6Ð|ŒàFe±fôˆQû$˜mS”ÅK”­ºŒ’Óµç1ha?*éÁªßÕ÷yê§ÒJL€š´Å°¼Ç0·`3 BY 8pµÍl»èÜàud¤‰â³½®½%ŠE*áþb· ÜRZ<Û|ÝÑÆÁýIÓ²£wH?ë3¨s ?À (šd[Ühë3ƒ¾taĺJ¨ÎȹíÇÆ‚_–|~uÅÅ®÷Æi‡xAøõäF¸4-¯GøøÛ‹Ü„ùf9ÍÝ—‰ñ ³caê«T¾zÁFþ{É›”A}âtTL þp‘Ö\bù‰ƒˆ¤5cfJ~‰1V’;7r‹ƒš{þí:ïæºV8£;zP°w}ê'ù`K&33±'ØÚ"Î4|–Þn×iû 0¦ÅݤÊÈ ÙÈàT‹è­£¨ßá`ÎÁW¿ß+ïž+Pü¥:V(…. ¸[!”¨.y³F>øxWÿ€P')éöGúÏey ò‡éŽH~}ÑûT:‹Ö˽Áš¢Š!Šõ(âvΔ˜Ô½›×Tå w+vë‘V HvŒ—ü?îÏm%‡ÿd£Q©)´\¾Xòº»ç"ŽRåBŠ’äYÔ­» ¡ñ'“ž€X«¡'åTHhY€ª©o⊃Õ*DLy)y»}uªqó£‹ÿ¿©b%÷0#Qã_ƒ"^±e/}iJ6¢µÍŒñÛYÜ—@D~ÊäZÄNùZ pµsì+—ûå?Mªø ?q½¨D1Äļc=!Z«_v¡¨1»›…éœ_Ê¥4H¹µv²ïsHioïåm¡TëÝî9ñáz3c&ª_ gô {Þó"E\cmH9×NPš§#~2§ßG*ŸàQÂ÷>îCµ_qðÞ=”ˆUBÔ¥ÉæØ~­=D:³ÕÅÃØb¨yC‹32/F>‘A6SGó$ÖÉ1»¦TæywÈ Âl‚ÈA)’UGnq+¥ · #&î¸(ã˜HÞ» ÀÛuOÊ+»ü¢µÖ)ŸŽŸ¦N>×þnu0A˜ôÌK%/›w‚œ}Ž=RÐ&áÂѬ µµ]´"³ôçÄÔûgçN®CV˜è÷Ý¥¾h¨¾èÝJš_Qº!L˜Q_õ<²¨ì3}x<‡ð4jÊ›Ðz(üñ<Ö†e”Þí…ø$"9“®Fе^Âß HR÷ûÀŒâ¯3,KçI!°PKP+Nƒ*ožA[|qËÒ_Ëdñ„‹uÃQâD4`ƒ½›hÜŒÞNk,_ëTÚGfЛ Ørù®ïYÕPÀð]óF¥˜ë¯çwò´¥‰2f+Ò»b“]¯c’0µgkúɺéî# b ï;\À=ÕSïǘ>i;^‡h³°g‚[Œ3Ë¢y]]{RÅ,ã­,ÛxÊ,Åà;Üßl˜²ã^)ÇH/.¹J=ÔÃÈ‹²‰7Þ7gÕ h`òœ¿ÌÃY˜«£1MÕH׸åðGÄ1\Ž/ùY *¼ãv›sjpôrµµCaÙd`­„ÿ¤G§] ÎvtîÌÉ8·]JIµÑ¿ ySRÍ?¯Yo¿=ÛlÑñ/Òx¤„u94phQ“  DÎÑ&߯)!Œãk;©‰½`[³Œ7pGðePHrÃòÏÆwÎý¨ò•íîr©µÌä“)R²ÃYn!gdª*gKùm´ä.®?Mƒ2 ³!㎊-u8«?í¡ñÙáAºäÇj! íQ)=)¿d…;ØÙreo-z˦3à"º1ðEÁE êšÍŽû`ÊÇt Va¸ªÊúÁåQ!ÅaßO•cèËcÓNü0´”Â_ÑZ¼qä-M ÇlÂ0CªK‚¯î„Û>—ŒZè&¥£c_Ãp'“DÅ ê²8h/ÁØY· àùf'u`›öÅžoF!\²Îÿ~!3NŸ‡‘Ê 4aÌÄéØв â­ècÐo)%{ˆg¥«c|B^nûQ¿gÌäôu¼Ï¡¾ H^O…Mú90Å>Ð@Ü„<4‚>{KWÑÃ_š“ÔšŸ*$7ËëÒa“þûÇ»J?Ä8j4“kÈ뜊<ª°ÁDõë¬'-ºV´‡õ7Ÿ]NÿdÉÚ˜_³+šµ•Ä‡ê¸¨\–ļ5<®–8„Ÿ'±$…´àaÍEâ-PÅ++Åp÷9:êwr ï—ŒŠÛD¦Q¸‘G1eÙÑKI&«ÛS¾=6[­ãñ!PûUÛgkŒ†ôet$ÄäºGhÕ*¸ríhõp>Ì ,Õ1ËÄ0èb¿?¨_÷ó'Šº•–þ§VªNk:sy·G“äGj½ªlY„Y}¿—ýšÍýæíúFn×Ë:¤ZŠÑ=¿²ý•vqâ«…´Åô‰«&T× …4„VˆRWn{›”™Í(šÔ°÷ˆež¢þÚ¸)Qž™‰FÂÆíô×UéâÌ÷ZsU«–¦}"”©¡jîŽo…u9O(÷ëŠÉàì·ÏŠ[¶r‚ð_$Xúùn*\ŽŠäˆ³ž?Ó«XëÝr»F¬4 v¤*ÏqíM­fˆÂ³/‹{ˆ£v„>Uº° I®ÃÔ9Æ’½œ²”?ž<Ê…EŸ]‰-ékˆú7åa«–¤(}µ¿„o©>–6yIpCÆUh8[i€‹Aó2-§Á;ÅçŒ==9µ9"DsÏ×Z$¯8ôöV‡f UÇ^Ä–>½å† ?sÆ—PbQ6ßÁ` †¡ÌŽ}¢MÈŸ:i$´,Ëõ'áºËþÀ›zÔ›ïÙ‘š¥ÝkamÏ|àéÑBå^ñÓÒ¸–ÿä±ÎL2‘’#E>íéÛåK”šš¢aŸ}îÐgåm:*ëh˜»±Zlô ÎëkÈÌ"‡²ÜI ¦§È+oë¼ÓÊq¡'Þ {Çþlà éûN«>+ÐrïñTšs*æyiX>¸ÎgfÔÈôŸïœ0Ô ‹OÓÔpˆ¹M•hC·umûéR»õ©oogyÙ’CÍwÇ2j¾–0˜êMÊ@ž‡…’ùº¢€ÅÆUMÇi@à# X‘™éƒõcÝù‚"„Åû+„üƒmwlyÞÏv(Ìó o×!]¶É‰¬üÆyGêÃ)¿³sþêæh¯Ô6qô…嶅µ‚M™MHÔAÒs…¡!pãU]70S•Ú2râ©aðzide#J»­›ØdM#$Ïêeþ‡qµ??Ûs býp³s‹—‡ô3BçÅ- ûqÇ)8Úô)ÇZ)&‰.ÃDªVöégvÀsàŸ‘TŒ>m#úªV«cô`×¹O­A±Åq¸Í-Ke"*õ·„=‰e(S= ×Á™¡Ñ;3‡ï³ßè™NÎÃÛa¥}sAhþlÍù‚Ꮇ/¤_ÞŽ$Œ•rÓéä‚ßzÄßÊÌÈ5Û´U'†jîúÕ‹¤G¦ÐRÄ!q}]ùA˜£6ÛwQ{—šhê4ª^ú,”Ë*Xxû§]àé.¾º¶:4J`Ï$¨¸¯ù’Ùj+tÂÏ7Þùè±ÓòÞ~ªÄ‡–ª˜óËÓüòïÊðÊòW¢¦)Æhûº³9Ê@ó&ÁãÙ@8‡)\5%»18ªvEË}Ïí9hš`Îñ§U6êTOÙÞ'ŒsËù&Ë•ˆ7ÇW¾Š% ê• û2šJ'{s÷}—ªcÐ ²ï¸c×ä53Íó¬¥Áÿ¿ ûôÈN²±ü‡‰#«?Èzà6-v½ÉD,'Ÿp’¾~–6­Æ·sÝvˆ÷ÝÀ5ÞhQ £TÇ¿VWlØWæŽ1\VZóÙ]Ì„lfvñlß2Üûu½µ•â{·š*c&ÕDæˆ}:B:ûrw¿Æñ<ðjã ”1»|KšqvX,I¹3ÛÝÿ…¸8› fŽLç¤ óŠœ‚½ TàÌà)oì-‰2×-uýÃo6}û¶_9)¥qªª <ö€»×ê‚öÐi}xXŠ$÷XÐþsj•[ÍTú(UnYH—˜ú¿]ŒòÓç߯ïycØ838Vjó°tIúWIݹÜ×j|vLÁoêE!Ä×b|:–ìòµ'T{|6;Ý”ÈfB(…_Ý{-:AíM¸|¬‰Ñgý =¯#o³btº°áþV t¬P­,e›‰çàC׬^¨¥l•öºIÈ.[-ØE:æ½rè=rÖÊ>Z Î!Â,Ï›Ÿ´†U¢…CLõײŸ¢Èaßæð¡C´v•*YˆGëÏFí2ûq‰øé¾ùnº“1]Ÿ a†T[oÓXÄ9ŸY›PÇšu¦¯²sÇWÏ•@^ʶ›—ì)u¼ûI¤•(JýÁuBsÕ-|á.JOUSÒ+ü™‘<1ïƒlÞ#,V; Z]%=ÅAF@¨ñ?ò­‡þ:_\¨›C7rf^ ÕŠ ’ÚžOHí=CºW.2ú l–Þt_%ë—]ãªþ—Æ\4ÌǸÐ,]«€{¬ø¶s—s‰àFažÚO¾ ¡£%îH<0ŒÀ™tÙȾ\[ê¾ÔSýà©Ú‡”¥¸é~¯ph÷‚è3”’æ=H¨,„†¼#Œ]Xí05î–r‰ÎàW„¯hýjÕ—æ–‡å©èEpÒ1SscN:’²)¿~n}zïú·%€³•ãN;,Þ€62ǬDÜ3báy÷$F"ÿ…‹â¥–ØfÏ©_"É®>•¹qÄS–9ã„¿Ê”sœCðOÎQZ9Ðr²Î@#)(ý\ëlÿ{[¯ P»mãOz–$DcD!„&3ãÞMGü»»Æc1´œ#œç`hDÌóÐ\—“Pª]ÍhZø¢ [§¸d"ðçÙâvaB‹>­ð$'½ŠÊzyô 7òs÷¯ñ~¤+E…™e;‡ ¿ÌB\\!37-WDÀ剎#§T…yÀ¯sY…¼ëàÿ;‹òVeÅg4‘uQõn(‹ö©G9¤—ü‡îxü×`§m%¼ ^%B݇g#n·;Á»ÐO†¯‡µÇ2ŸžŽ#ÑZ›Ð=`ûn+JVÍد‡zå5qíwaÏU[¡üO”XÔV>.Èïuâ‹I.Žd ¼ )ùÒ§ëˆÇø‹¦ Áè«äUV Ç0cRƒÁ×®øEãÞÁ-¿íSB³„™M€Ñw'x (LÓ^fêcùx€LL98NrZxMY„&„šÆçcÍÞÿÑ"5-ß §‘Q hË® ’÷|È¢7:#„Ç“2¨Q'!'•-ÉJéüŠ=h- ½3ø2@òù«’Èï p¹~uAŽUyî(ØKÇ÷ûÅÁè ø-‡ña’®*x/RG»mÕq=vÌðºþ8Ýþ¶¸sï6<ª)bá#°6r¨ïOämÕ ›£{å\ÚgÎÓl$rÁŒ£3X)_Ž‘]ð™`¡‘¿–ã-f{àÎb“­ï±º‘ýXœ…e‡g÷“:Çir3r0¤33T¦bçˆü‘¶)bÿàì“a>3± ƒ¬ËñšÎ·)c¶çÚâb}¶¬Ë¸`ÿ‚h·H'.æÚa C0=¶ÂƸ- %x<è,¡t“óÇ÷VíUï!Äìãl´o¾ ‰îͦM¼q=ÚKÎh 7εv¡`~ßÌúˆÿEï~ U4Ü¿ÉÄL‚ê:±{>ð2¸¯4ÉëÚnÖ.(”ݶ S1„±ëà„€µÍ‘†Cf‘Øž©$¹ÖTšžg 9¬=-“ ¶Ùô·ÐXr…eßgo±ú4Ãël”žFn8׉ÕÅ–ð²˜uH„Uw{ í­HÒ­ t««V ×Ë„°1캂ï¯ÉÕ+š'ÑáX’ÿ¬™¯ñXxõís)„ÀªwGœà¥E$½:1/‡àؾO뻂‡ CÛ&E_èk4£*¯VÇL{à‘Óú¥iK‰!€(›Tv±i!Öæv'¦í˜‚+ÝÛ=f©w.7gl`£žE*Šÿ…s+ê_™4àh¾4ü쨟^m™\ƒ<f=ðòt.weÄäWýgéóy€U„<»'eìM ’ø‡jዶ¦L¡Àð;C“XÈŽPlc0Ù®:{ÏýŽ!æCb  ^¦EšÛ~¸z\—¶#ak­iÞc?bÁQ95KN˜dE¢2C«³â<ÃC¡{ì>K¹Xò›®Œj3‚’oH<u^‰[éT“ÈýåVáv” endstream endobj 65 0 obj << /Length1 725 /Length2 20783 /Length3 0 /Length 21336 /Filter /FlateDecode >> stream xÚlºS”fͶ-š¶mÛ¶mÛúÒ¶íÌJTÚ¶]iÛ¶Ui[÷_kµÏÝ·Ý6_úhÑG3ZK8Ø»ªy9˜é™˜xˆT%ÔÔ˜Y˜˜ˆ˜XaÈÉEÆ®VöbÆ®"M€‘*À‘ˆ™…è3 9‘¨ƒ£—³•…¥+•)õ¿ƒD¶ÆfVvVÎD¶îV¦–D|îîîBî.n În ÿ©D®–"s+[‘¨¢’¶´‚$•¤‚:‘$ÀàllK¤äfbkeJ$ge °wP™;8ÙþÇ!2u°7³ú'†°w8»þCÌÜÙÁŽH^\MXBQAH\”QM”ÈØÞŒHNêŸr{WžÐSW×uGGäú?–­ñÿX&ÿcýOÖÙó? 33‘™•©+‘ ÀÂʆñ_ºIÛ›;qü'lææøßÔ?„\þáGDõ†ÔDfóÐn¶¶ Æv"*Q;G7W€3‘¼ƒÀÙžè_Ú{8[ýù/ÔØÎÊÖëÿüÒ®Æÿˆ!loñ|Lÿ Y¹HXyÌ”¬\ÿÑÛÕÙ ðŸ°&à?{#0³r³ûO9àMþÙU3{[¯ÿ³â?­ü{=FmIE1Úÿ‚§ÅíM̬ì-ˆT]ÿÑÓØÙìÿN+[ýg‚þËå¿EÄü}ycWg+O"]&†Ó¿¾ÿZúÿ%"âàéCÏÂÄBDÏÊÁDÄÌÊÆLÄÅÂí÷¿èšº9;ì]ÿ-ú?½ü×ÿ÷0žS˜õSÞPë´Öð ñ¢¹JpÚ¹#uÚfàÕéž]d«!K3ã+œ{5œ|8üÙ þ½Ë]éûãxÜÉäT,q¤þ Îk i¥…íö»rÁ[C¤Á5˜¶©ØMÔkÆÆz°Í[Vîy­ŒS?²$s¬QyG?8ÒNÂÚ⣽$µ>%J¦Î7à€¶iÎ2¬ðÚDðu$"?dî†Ûв°"Èñ(B'ÿàѸù´ý5ô5Á­Q—<™†î²¢¬Ås‘ *I÷´½Šg QE)ûjd\óRÃÝqö£BŒÑ€ÜÆâZ›H"½Ê± (Íø¬Án-«þ¡qk×7µ\/I¦]Â’rᨚ‘ 7v(?ú>ùÀ¹+8GÖüûϛɼZTnE[`㜅èLu{™7©¨Ú©»£rVîQ(o ÏrV¿Oè`;5…ÁL±»¸xr¿\tÖéì&ÄÄ~CT[c>éý³´^§«¯Ô¯©ÃWφx™?ݦ?+lSæSÇæ÷/ÝÔ^FÅÈß6Uàfb±a·Ÿ¯ø²‹ÀØŸûÑ0e€m üäíû÷bê%G~ ´JøxDõèÕÞ¨(»AÔ¥îêó öæ¿j+ße¯#X˾=Ü<íqäJî£b Ï]ª¾òzh`˜“ üTYðô8ΖÆ.C¿’`E2 ÙJÒ¹Ï €&ƒ“”¡„®ìu¸Vr×66»¦Sìs*T”B…!»Ô«6ý"¹ªC·WGSæ³ÐT>zEɈ‘†>¥P~ßßŒæ ±rbB¢X¥äK¨V éw†á,½”¬2’f¶SI%¿ö’H ·« ®ãÝ6„¹é¿ÀP}màúäÆî±*l3wáð A鹬nx¸4¶8Û Šâ1ù¦vÏl0e.Þ?Š=ÌÝõ»Ê×S²ûVc0–;J–ˆ³®•_Ârù`έ7è„[ôÅr÷A?™ÂÚ½w¿Àxà>Ƹì–ÅÖ¶Ò¾½&?cÙ^èAÂHÆ{^‡4ÀOÏ|U!PtK:&ûAy°´w1£-ò{Ê©ÅnvÏS'Y×Hh²"ß9 ’ˆ¥“üFød®8”ûŒ¯ˆãlbWsÈÂBÇ´þè|ìrÞ‚ÁÚ³D )y‘"⻚!1ýBj;£Å}¶r&?ñBSÉ®'²%[»bO zQp¥º™V\‰h$ýzOmiÐqZ+úx)# bo5Ì© ¢P-ÄR±,T2:å;xnÜA‘ì…ÎÀ¡ ôM@×eŽ‘ƒÊroá“ID:òôÜ}$¸“KK7pRѾØS”s¥…m-=V·b’o¾8Y SìYü(^M\Z·n Œ¢%Èdd~»y럻š(ZžÐ"‰à'þIéÚ™’gšla Dø aýÖÅó,XA[•xã@GÔ3 í=¶w@iT§Óþ–œübÖÆÓ/Ü¢ñUiï­ú_êõvQ(T3èmŸ8vù€†]ÿ'«ñNþ|cÂ=fÿn¬D*^r¢t¡@ÔáºYÏk‡2Cn~©wºžÄMvÙ”8?öP£Ë åNx朊&¢ÒQ[ƒç^C¨N>¦–H'¿Ÿf‚"Ÿ/=W]ùˆôè‚>$‹7›Éùˆ ¹ªÂ\ké/-%‡q%¿m`9jòO´ZyßÌ:È(e{±õ#,o@a¤Äàϳ#CžÏá<ìŒezðùíPý3þ·7GTä×\ƒYP‘Ô7uëÃqÐÒd`ÉŠ †LyA¹íŽ-D>¶n öÉÉÇ ÕE `à<ÑþØf­Û¹ÐìT°Õ÷ôä§O¼z]!*HÆ#ê+›q9þ·ºjåZîCl­8½f]J¶Ú+Z쾊í§sÂŽðw‡ØÂ$¹%=íp¸®;y'½Gåz3M‚?ÛÖ$É/ï#¨A’–$Qöë-ð#–Ô|µ2bpˬ €†¯E¯Î"U _Ý{§Ùéõœ(?aæåÖAΚŽò›ÿ+“r61—-¢Ã"÷Ç<³+§üÜ“F0´"ç$U%tÌPÉ-uºNØŠìNU2«êQÞÒ©Y…ãæC‘p:„zŠU> ¦ƒFn„Ë—]ÄQßù ·i½ˆm–pážûƒ[ˆúcñ]ž¸¢Õ§˜ ¦¶¾V›=ʾ-6E’²eÜ6áÛ§ Ü©_c2n09M}ÑxÉF)ì·Æî*PHh@v9.èÂîR"¾Þ> u¿É'¿“HÉw· ~YÅoQ³œÍ«‘†ƒZ¸¬¿'Çøž¯Z£×ÄÚÑÌl×¼®. „¤æuh½Ö{’~CS8Å/ún‘‰K±9ÞG?ROBGµWØ6ùÿŒ ³9¸Eœ×ü"¾3PË°]׸ n æÛRJ»Î!¨ÈÙ[̼ÖÒ¬ë«Ö½É~È&UU§×·†Þ…°CéôÕw¶Ù¬ƒÄEÇc•gÁ{äÕ;Í⃉DTT7웪ê€yЈ@™ý”s²¯W,íÐ}…tóôE€ÈÛЦ’^ÉÊìBQ3/líqôlëoˆ²ø{²™Èzh#íÇì5E Ý®FÁE.2¤FÑ«!$:‰íÿR§@Ä.6&x‘´JTôÌÚ „4ýdÜk,²Št²òìmê`‘*@]–í!˜Q(_×”ÓY¿ ü y¢üˆláãáÑy…fì\ðl&F_db"þZㆇu\Ít>(õ=Êu÷aÂJÍû9UU•BÑtŒ­ÐÏVšö¼ÃÁHoIºÜS0ÐHQé“( .n8+Î^&²T6ÔÔ|‡”¼?"=DZÁõ«Ž6Ú¬ntús_ÏYÿ>e£Üð¤#¨r³2;Za²{³+jêT2¦ ·Ö†.êØ|ØOÃŽE뢆½±IUÏ.ªƒe)ýǃ¡˜*.£¯?O‹ b»ºr\BÌÝï ã`Û»R!i§ÕÔ8Îl5=mfU¡eº:7¥¸m¯× Ï“ ±.¾ÞIyL‰¬‡ñ›··Uí–PÝípxcxåÀ”Ë-æÊ•ÀXgÝ°jx —ÞvÃî¯ò³1šì„´ýIYÍ­—zòƒ.÷²©©p—¿TvMâØ´õûóI¬~5ãÈÀæ ç>ê¯lØÆ•xº%×ÉJðU$Ý™ýŒÒ (äôN`¤JM5Ùþ„Ãw¯¨¢1‘¨‘žÕÃC½¿š1¡ô³áïÏ—Î#|àA R ™Í“ý%d©â؇j4)Þ ¨‘—(lqºDî°O)õif´A.Žrã(™†µDPã¤}%[0‘ÖÙ¦8%sÄ¡­ÖaH \EëU"Ýeö T‡æ[Ì.¾¡ÙòÛƒ ”¾çpÓäAK€Á‡Í@bµ¿8<>GÙ¾><å‹#»è?¯e±„¥†@¿Ã“%Áú=þ¾:ïçwüÉœžÄ3rr¾öž~íý®²Hˆ€âHÒÕ8T•·¦p@ŸÉ~5f­åL©×ôNì>‹_߸"Õ³àgé=™ˆ†ÃÈœZ¯uͼòxŽ®8x6«édè4Ÿ $æû©#^­Q~ÅÓQ#lÚz.Éÿ¢Ãú®ÃŒrÖûžÙ©¯hU¤™Œþ>¥ÞÒ}[õ°"Ä÷¬îù‚S€ø`A‡lsxQ¸ÊJüåc.živ“žºÊ²¦e*ñ S_ì7CNÔ \8ê•<âÁô=¢_‘DÔPÚ%·?Òº_JßlÚ>‰#dörŽêR\@/)5Öî†ÓÓRúѤLÛxÜB7:Õv Q„½ÐÈ;n-ë·Ö‰ïæ`¹8Dæ×UÖïªc¹Æ&!‹zÎÉ#Ñ=ãÍÌþÔi$õKÆˉU¿Ä´Ü4ÜSÿ:)9r óÄëÉÄ1ÞÚÁ ì–7ó!¹ß;/ÁV±Ô¯ãº6öHhœŠf{Qîåå!à MmQ:™þ.´°«^ÏOµ@9\ŸJ6º³Þ/jŠù„×ëÚI$Î!Ö2%D;†ÐÇbªn‹NÉßk–öNB.ìþ0æögµàÌõ™Ü÷ÓzpxyÀ‰Ë á¦òÓÉ0(x¯¿è´¬šä±.ÂE$'Ãßý–·¾)4$žÊ\¤à 13 Ê'²Â»øÌ•[Zg\Y…ë,õ-m)´Î É†ˆTf“¡È嘢žº¢0T7,rýl? ŒhÝiRŒÊãC±gÝL1#'tñx&×€Óa—t¢ù¤¢eœüh ΜÁ=¼â(L_ˆ«/U} C\ëB%  ˜DŠ12Að¾³%%bðìaYñÉÞ¼3m5ªÊþíC(* c?ø‘„¶’-”dË(‡\²›Ð]¦»=óŠž‘Ï•ãü_®µÔ±0äövà+Ó#óòÜ ¡éU5jŽÀÑ‚Œ´–L¾¾/ÇÎ4wf^ZO´9|^ä3ydÙ‘ û[†NkCpô”©Âï3k.ǧF!owMNÞ¨ #¾ÔsNMÚ&zqcúþ_1M„ìkåþ¬¶‰;§æûƒ(ŠšæHŒžßvÙµÁñ¨YÌÞ{T…]ÛÍú™eðÛ@µd9p†jÆÍQ¤O9Ô-™Q’‘(7Õ²L*îæõ%JUmdì'5Iù©ä1$>ÎëÃûuX{Ð7u5PV@8Ï Š?*ƒµç»†…axVEòxC{¾»6Ž¬³âuLGß¹"ÖðÅG¨Ç¥ÙQDY£ØôâMàª,>P;ƒ2ý…_÷p<“kn}ÌG‘š»aâ ¶,»¸o}&5´&-—ëËä[·9‘œ¹“ûE!WÐÀ¹Ôh{?>ø§ƒ¯kž^Ö…fÈ%©(ìÆúYúVí_\6q÷‡y=nò‚ð†Ot•uÛ¥²çªp߉…ªûZáêÚàJH¿¯|ípç–Â& ù‡Èò+_¬k«ƒºZZ¿ÎL£Gu=ÒçhüK¥¼ÎKl}[Åe¾!ê#q<‰9Œï—•žGD-È)<MõxK§«E½ÀÄý(B=¬Vv¸UC…+r^ímæP™IU!0"Ù¶ëúçù›]tѧŠâÙº½Ú²Ä .ˆĹ ɉ4-ï3~ )x­kŽUƒñà —+>B ýÐH§"_ów•yTËïj³Bl2W–É”hßoÉP{¼O ¨_IXÒý®›™Ù.xØŽÄÌMA/ïG®Od ê…³é%Ê„™Ú=zªJûÛ–ƒŸfÇtÊ«õºù˜hÅ}¿*h­ó ™ˆÄŽT“˜vuÃh;ÐmòIꀪ=ry#ÅBI1[IKÑ"Oï#„æõw~}°{í`ë3!¤…pˆgÝQ™¹œI‘vÚ{…QB†qô1{1î‰ ]úÊ–8ŠÔ—¦½£JQfÌ܀ î$¿²Íïib`ê”qᶴ 1H6ÝÏ8¦«.ÃQ¢"ªù¹™ã¬… ³ßƒMLYƘ@EÄÄ=ˆå"¢À;êGû*šdr¥¯ NäÎBKc˜ý•u¼6'gF/,AÙh¾×ˆÅ´´Ø¯¯ŸŒ‰VžöxZ´÷ÞÃùj(•ÌfQbž£:äÈå¬E;ºìÑ•7;~)uïö­zºm4t{ud³Å¨(\Ù˃R«ÒñÒÿ‚¤œï˜©G LqÈÏ-#rûÁjòFãë²Rj;’¤°vC¬™Ï‹KŒC¡3å\ò±ä@»–ߢ>µÇ°ÛP¾ã+ 'ä V¬{t:|/ry˜*º\MøÄZ¬{‡ ~­×#?e!A3•«m&fãä{-†ãw«÷jR<ì¸K¯sl¶!W¤ž™“nqâp äi^‡«I/þl‚£'ß¼ÜÒíÑ'xÍî1¤³ÄɈ8¥r{•Ò°ýÔÌÿžYóà^éý„Q:–’ÅùìӸкϼ,½[Æ´ßÇÿ‚‡á-Ofó|o‹ÒO¨öš¦!UÒóäE)ÑÒQ17SOWµ=/êðäVveEo` µºÈüª¾áJÙdÇ%åïKFïäuf|d¢³5L>ƪ±âÏDºš®÷IKÝÐrsÈ_e¤ëŠ&<ôþ­t¶©CSZí'Û×(>ÍOƒ(¥ßt=ãµ)E–Bò3úµ¾ªŽ²[á»@ÿÕCmä¾äÀÁUh¢É »ûÚ}¡/.¾¡ßÜ´Z$’5!*™g;L^ÙFH¯¤˜ÃWŸú Ky´úò4\C‘ÖJE’Ç0s¦ð£éÚ/ÎXFЋãÈ…™›à;…Ú—\‡š\lätq ¿ÃÞ?°Üörq4^Ž3Ñ%{E«…´ÎºVt:ÝH©´¯êY÷+äsk”q•kOè2ÖÕ#©—õ*65“4F­çTÕI#“Ý2Ïi‘¡®Å«aFéO5è»K83]ï¹åp|bïÌqxÅà·˜Ím_clÒg%ÿ3¹e_?No™™›Sä¾ÒY~$l%Ÿ]Ï#ï/þ¤B#:“QXc1Å(o^Ÿ"Á¬ïÉ3ߥ©—ൾžš¯§Â~@ÛÇJ:wñÇÕVuúïã1L£2^UI‰iÖÉ·rÛË„’2hHï^uÑISŒpÓEUÀë"-Ë|–¦°øLê³{~ ÔZ) ~+|´–²ô ¥eba6²=§*îhïq~¦%cÈïl†UªF»Ê^À¡c²¬øÊã£C¦u€ô,>º=±¼ˆÙZµ4ˆnÜÁ³¥³B• bøM(ƒ~}¥¬AR~ƒàCmåÇk¬h¨å‘¡"„,^ml>ˆ3ïC‘f~»Lu­ .‘/©@7˜Ÿ„åv—|>šk;¨×ç{àYáôXµ4Aâù~f"Žýü»pŠ@HC=wÿþ˜'Cê±Æ°{îöWd[¦Dèý(3ëS¹uyÆA‚.ô/´‘Ù¢Ö¤–!,ãôæ^KX;§Þ“Kõ¬|ãôhóhDÍwîVéºäÖëïÊ3 [íX›Ðöé MŒO¨ ‡'wã\dK{„xÏpÑî4¯3íÆ?ûü Ã\Ø›[ÄmÔNƒ/!}QÈw°V£ÊUôð3H˜L\Ð8càYn1Ê~¥VTƒ¡ÂÄȪn†õÃÉUŒŽÃ{7tþ^°óótÃJy|[3Ÿv.c‡7ö|o ª€h*Å+ÌTƦvIȨï xÓNvÿ”©®ªó:[ úá¤jbg,N§ÆGº¬èêQ'å´p5é·úïÚÍQ'ÓÁÆkäû²‘= \X« šœhª0!|µ§5ÃWÆÚY©‡¨ @´Í[ÈÛB#ÃsœC-ŒàÊâÒ9*J³N`¬’ÐÒiJ~ð|Îì![‹(šK{ׄ ÏLfÖÿû2öðC·ÔýZ­)N‡#"r.Ã’î n/+ß¿åẄb}¨ujhñ,Åd§ˆ¨&Ô³ýM\Ñ»fw=Ñœ²xûKéõ†}ÉÝRÌ£dV¶lQ5Æ`Š›^Iÿd¬ü¸,b'Ãw?‡Rûç,¡—2ÃSº;qj»«-:œ(´5µéfœò•àC¤±Y–ÃäbUá¬Þ†Ê<02üeØ‹ÂkNí»ôŸXÓ*Åɵ|JçC`?펩OtÕ:g~I›-7Ùž?ïÊá+åÄ‚½1X6í‡ÏcáGŒI×N§ÝjÀ±#'%Æ[ý×¹‘-¿¾¤q~¦:‰¥þt õ~W¼Y˘{Ü7Ó]<èDc &?! ¢aÐ’ÕT– ®ìßUnà–ÚD-Ù Ù»3&òœ˜Ù¨Ø'!hòË뤽”¬Ö5r»¥`Ú¤UYb†ü$> ÷ëÇƆ;±Ò4ö{à¢?l mÕõ>r:¦%ò}& Å©âh5ûDkf{¼ãÐ:m o² øSÛO»ò&¸á°s°Z?gûûà†ìÀ<2Kåpö]\¢Ã~O×^‚ß}Œ4ÍIÓ•”À fø0=/EÚlBr»°4^Ï;×6µ‰EL<¯¤÷á·¸&ІpÛ’ýÌÀÛhÁ-ö Π´™À§Ÿ³q¯n#úÝ¥ä±e¡tÊcŠ£áÇÉ—²dÌpséyü„­wrÚ›÷èÚs:50û{iQÖx— K†óŠ“%ú¸)**§çªR…DJ~†´Ä…=«ý~Q‡ Š¨v¨1Ó;QïCs ÌW³xG£)±ÚývÈJŠ›¼Ý#\j…{[e‹$§í3MÆ@ÞŽà`1ûá;±]pø UÃ+¿ñÊÎbr…Ú—òÒƳfÆh?ØPê9¤=”Ü‹’¼òu¸Ê–1z-Ãõ€øu÷ÒNûÎÕÿšŒWуý·ˆ->—Úee€zظë·xõB«˜×SœŸ\–Ö}а«£LÜØïäšãVä.ˆó‚ó<*`ÑvUçpô¬å¬+Ùßz¨L7%¿`- J¶0¡äª&ò»ÖÿV×”‚Õ•âÜN(Hñã*ìÂï `²Z:go/rGȧөr¢kžð,ßV&¿ÀÉf±EÉ  ¿gl<ÊIì7<’B;‡Ñ—‘˜ZSÊöŸ»‚ ż€g+EÍfå°OHýgÖœ›ï×É6nõb*d¤„<ËÔ\hƒ¾Ý”Ü¿¾tuz1 {°í Pͯºµ½úÀñî ÆŒ3 nSâa~_¾.uÝ"í)Œ­D; ’û%7=˜–\TzUO^Ÿ‰¥T 8Ù’ ‡ÛÜ¡wQ6†);^V #¸›ýiJu3¡Â’¥[ŒGM¼ŸÅÑ•«-³±±cƹ QÕåò«€É¤i¹}kÍqÖW·p—ƒ:Œ¯vm9—L(?µ(#Íà•™{5Mœò7v–’Áe±á’æó²aì[ù6#E#¸ÙÏÊÂîñ›¿†^Zÿ*NlÀ9N;HµŸ¿HÑ—PÐM…áhkÃÖh2Ïm]~´‘y5Ÿ6К@bÔ½¯¨óY_‚T‡^È#4 7!+·ˆÉÑ;]ßcõdýéÃYî;è_¯µ4q„Û¥«°&ű4qÂΞÖ~G’áA2æ ýœýôRG_ü;¿­¬‚x|3è( (”~ùâÞÚG:×cIM;Ó2ÀüÚ£TªSß­«ø‹b /ô ¾§ @D€ þì}uÕG±¦®³ŽEw–¢~cÀ0±ª²–ëûÿjº»òq”ú¸¬Öx÷£Pk‚1Òó\‰Ù\Üîšàom.KUX5]l[×éAÆâq£¶´ãg'dOy¾=ÙK_©ŸË)ÚQ£No¶‚‡èÒš!÷bì™ ždRi,]""Ë2œÙë8Ù†Ûdö|ì j¾I™ ËîªVÏ7ÛíþÜʘ¡Žª¼”sšÑsÑÒW…ö‘ÕÉö^MQä#(¶ž{Œ‚àäëåŠuíOž¤+öã¢3åÚÀ•ÒN3ÆrxæÒ@ñ9 °+2€¿ŠØ*@'Èãˆ4ß‘ aÈJn* ¬ÛÎÓ¨÷<Èì"$€˜ØQ›º*?×€ó›KØUeKÄî¥`·œE¦†ºi¨+¾Öõ,¡Ì¿3èëŒgÄ94c<廉“R,q#šñ\žÔ¼Q¾ÁqN½¢°¡¥“ Ëiñ©RDº#sê"+^ v©Èô²I5½}dd™3d-À°ãÏÝ:á_Û·6„ôpÌž ê Ö/sX$†Ù¶š*@7®lFá°ÙmàÚ“¯DqkÀÂÏ8ȃ‹þG±€Ê§ß–fKó–üÚ톔ÿVzMű¿L³kƨÁÎɯÛmRnZΕÁËYªšc $߆G DáÀð¥ ÄÓÀ¯† ÕO•ŠnrbP…Sè½í­ì9Ý "ý¡uë €¸ô‚¸^íÀ¾hÇ+òÁã±Hv—e»¢èÎÿm.˜¾ÛÆøZ„¥œv¯H¤œå抇vžø‚YU,Z^¨µ”ƒ¼¡äÊ^*ßÔ_ò|¶-bÂ]ˆ “‹XO@˼ ò¥þ’ä ƒù•”]€H|ô½žYܶÔ<á,ʪùòL¹CSÊΑ (ˆ+dHT³‘âß.x?Ú¢Ó? ìÜ•ÿº)ê&D˜’7Ó9–ov®7Ð.R¸ÑàáEˆ.5ŸÔû‡…°†¦†r>Rr:Q€Y "!³ƒ·{¥ü³?A¬Að²‘ÑYHKtÔz‚;E_¤ ¤âüx_¼>¶Ä¤§ÑªAƒyж•GÚÖÄ2m¨A{qºô¢ŸŠl|&#ÍÁiÛåJˆg*qº"ÃæÎ0x²×ú·ë+~÷RÛcrÇl±d4ÞvŒÅ‹@ÁSŽâ †rѯ¤‘"RHþH$T !9¡UFè4]t’Ç ô¥våy1û}”x̲ɦ*c’)¬@ÏuW$¢ ¦ºå>p¼h¹–щ̫׆xWÞ8‰’Â5íu¬Qÿñ%²Yž”7éǃGrfƒß&\õ&bÌú ìvtM¤Oà+|©oÃå÷«dÖÎ.~ÇŠ-…¿W3Æ ŠµÊ\mÕ `]?whQÉ''’DÚ¤Ž R ñ[([VW³¡ääL¾•=ût y«‹*ñíô\³ã å‰BZÊÙác“(3(:jžoÄΡ=ü¾~^â9ËÅùã¼ÁP¿‰ñÍêj¿´f¸’ÉVÜ\=ߢCõÁú¼Â«‚Ñ{‹kµBÇÒ?—­©Pm "— ]Þ§§[%ûÝ#Çø5ñàÒÃAeS¡¥nEæe.NXÐEÇo¡~fýœ]ˆzÿy1Õve“÷R€&B æ,“v‚”U ¹Œ/í0ÞªåM*ÆJMr±­®T¯çZ¦1ÉìMY˜Uë‡í×Á'ý9b¶¯™‰°XY@}…‰ß-^6”·˜œ;mpÔß7?ƒÞzß·9HE¼ShØÒ}Xkû8ßÒÁÏ^˜.ä…f„p•dŽ;™ÅËD~+u5 uE ƨñ_eq$Õ÷˜A«e& "S  ¶f‰-C¥Šz‘Í‹Àœ1y’ŸÔdÖ—p?"ÉÄZ£œÄZ!‡…¡Uà$âÚ:ø¤U×½×—Ã*¢yg·8­ íVÞ­²ávîÄOû‘éç˱À»1C¨–߶6Ña"½O×u`Z]ßÁ°bj.*¡ iM3M ‘¦$Äz\Á2¯Ù¡ã%†No(‹e=Ö!’Ôý0W‚U.†c²Q®j+a rêÊHŽó$9¯4|°R®´_#+¡lò ]$­bT%|pÖ¹ {¦P¿¬BR[¶éf2°IP£>p•Éc8*QeQÚØDHß´(Á0dNὕ™DŸb\ Íô¤ÔýŠö¶Á²u‹|±¹Õ‡viŽG¨p€!ÍÃC|d’J¾±FïÒi˜ Äí:™„í²¿ê$݃À°s· ªiÛaÐrž–d•"f9sÚ•¾ÿ4°någï$um1ÝÂݵ¡áݼ8Àâ35±Âé¤É¼Á#þÒ° Eݙ֎¢ßä0#ðgÄ;† ‚ÉØÛ° ik9” •˜êû$*GÊû­?ˆÄP<]œ^yƒÛN^õ`ÀÏX™ÒQb>dþ›2‘…ì9e¢ iU±P½]VñʧO<…oE@OÍŠŸaù /™½š?¿{SSÐTß*î"*õãaÔ…\ˆGq§V¯xél¯oÐÀ_ †sÙ—ÂÝqÓŒ–G¸‰±6½Zæ“^Š^I‘Æ‘ú·ˆïß?ü)ðã}þDÑÎ@xŠ|œ×r òeç$*‹!¢ËDæ–ßpñ£a&}×é³8Ôæ3>¤AíaSW7k¿^ª°ëµ5Œâw|$£ª¿õ^ú(Í2lbÿ½ÏÓwZIwZ”^ʺ9º”¸ë•ÇªT«¿éªÔ7±;¦ÝC‚! “ö[D /|õ4™lÓÕþô“îÐDð8Kwƒ áÏè:6߬=Fú;¦ÎSJpäLä5b¡c´›TÝ2ª3ß,AËúsí Þ2MÖ„»¥·o÷9œ«éâ+ØÚHÆž!E]ð–wK)èf¤c]´?;t¬ö//‰·+6ßø¸k.æô-pµ.fI.ï`õkWÆì’w®¨°cöǃ†×Êkç_•é@e“[Æœ]¸¸ÈЦ¦®Ñ›Bª[9çZ«•WÛ!ܽ?YjRñ‹OÄïKR/Ãv !ÆJeïGin¾=òÝü'styŽé®UØ3‚úˆlü“â¼ì,c‡¬ìÓXJ®±ís )~eשߴ6ä`®ié hÊ !r¸“£{ŠêMZWYjE^Z‡C:mi{ÜúšœnðÁËC˜ñpIx_Æ…¶O1Hý½{!`=A„ˆõ¾M’C·ªã­pl/ë§[„?„»`óŠõ q¸Ž}Pwï1ºŽé+ø o¥ŠWwÐÚê©AŠzf«DlKJ 3Ý"£0Ùm8Ú¹,c«h€„cE@ ³Õg;aj[°‘\40'GXëLÊö([3ÍÙ²Ö¿g< qÏ>ÑE—ãE•$q,‡Š–žX,–Óò5ºõÁ®ãÈ:ÄæÂ:·âöÎYýŠ M W2OÃ/Ý5håÑcPefZ./žx>t(6#Œõãg±7:°î’ŸÆÁT$òÉUŠµ9Š¶?™!xZdÿ–ÆH|ižÿÁV(\Uà•tP/éÔCwZM,s®,z¿&˜n ÿãc“¨€þ¿Jµ¿êÇÚFgI*ŒÑ Zéõ”õ3 –ò„£Š-Ï8)Ã…_r壣XZjx?ÃL‘¨ªVbsvÌ©[ƒÚ3Šr9»M‡R“Ç;Š(Â’/‘ÕÍô€Öü`?=ýàééÝa´ýVù7Má-…ßy·h8hök®‚¶ú„ÒžÉ$Kn ×ïÙ®#=†]rÒ& ¾ˆ?|„\¤»y¯ ¨L6ðvjÇ^̳¶ôŽ9[*Ç”C™ žs˜žzB~ÌmÁ“YÏ[nUGî9Ð:F{WáoN5wÏ ñŸH«ì™ÌhÑ=GÜa²á©[`c>½­U9.ªüKñ½¼Ôvq¸È+Œ.*~Htë¿ &C: Ý›ÇÑt‚¼3ïÐm #ù¸‡Ûj´JeÜDæfÈi-:åÐ:™›sÔ)3}Sjˆ°À -´õ±ixó¬§f‰KôÚúcãì 2À—ÒéžÚûlÒèVåíN¤øƒ;Ø͆R€þ'Jk¨.NTñ2F ´x)TÇf÷á7h0׫ !@}Ùø:1\Ð)ĨMç_§*ü®çZkzß1NèÓ†[Ñ­÷Ý. t|ó(fªß ±Ãò2d©(óà˜ní­vTÇyLHË_$îLÁ˜ÂǪ$‚®\KÊTk|4¸VœÄ\vù@€øŠY)IPèh …¼¥õbj¢‡õ ~hUÇX£ôõKäˆ)+L¾³™êm’Ǹ"Bc÷9dÐ3êL®ö&¦OzcwÏWÊIv½Y!Ô8Q×6#Ï.'Y£Ã)¥P3X'‹e³¥÷0öKúEØ\9ZzŠd€Vëçã>yóË·é|êo(^­µ÷Y·çŠßDs³c `Uñ õóËn«£Ë 2Â)ÓÃ0ÎÓSd€ ) 2ב‡•'49?úi¯¦Ž:Üóïî“ùp-®4f7#±Wôæ ®ðïšuÈý9þ$ª¶K€ aÍGÝÁ¦“ ±_}z‡÷=ÀÃ8cƉ *÷ú!x¹FÁ$¶Å:½P0óPìV§ ‰éÇ›¿9@<ôê5‡K?nU30Kš;j)åˆü "bÏ—"ÑôG'9¦sœ”…6ÑnSl ^Fäíïó˜QS²3۰˘Nµµ1cjÁR}J¨š¸?ÉXKF2Å’žÌK⥋9æèooxvéžµfÅø$4~ÝŽüåërb{_Ë¡˜b€ÌôϪñÁ©åò6mÅ è³âÊòU> ³Þ˜Y³f*#Kxû·232•ƒÑK°¡­Rm$)àz©?? [Óµòƒ”®’káƒ9SåúêöéQë ˆüI¥«HK%W£ìÏ.V#>9ÕM¹wHŽ#°Á{À{-¹§Õo<7Êÿ,âG–wŽRl¾!žÜ¾p§ûž“ñ2@>$"‘qŒInÃÇ5É[ñˆPc5Š Êè†zN…šÞ^0Ã)Ž ¨ó4 &êEÇeÀçm˜‚žôªùƒîœºx½ŒRq>Ì\;sfA…œÙ-‚ËJ,W,¸¾ Ec/ÝÜù|%_)[<á_ÛâH‚%zRÒÛѤcCZÞŒzß_›>Í jÙjk>­ìÌJ¨€Ù\‡\oj ¨’”!&â3ú n¶¹h#N¬×>Ê·ï U!NÈzjš Δ÷ÎÜ¥ ˆ0D<z­ÛÑ—Æ$éÿ¼¡®uŽÎ‚)XºÈ<á·3/Ò¿žâŸ¢ŽsÙž²ŸÁ:3ª5ÍÉ?‘w±aM_lM¼x-‚¡$'¿-s†+Ç”w9ÿIšð§ÐÝ©ƒŒ¼:Ê{×ñ™O.þ#—‡tFVǼ¹ñ¯MÉÙ†`Höë¸.äÔqª#]AÁÁv ¨Ý ô3,)²X÷|(]'o2ÐpMvØ0°ö½ãØ£á¦Îi‚~µc~Ö¡ªý>pfIXÁ*€Ö;—*ÇíŸn£ß6ßa.ŹUÈ^®´nçý¯ÃØܔ›'ts`½ÙM@×Äê öJèš_B§¦ëúÉö©ìõ%Ð<Ý´·èJèƒRv×-+²ÆGtcS&ÐNî}Ÿ&/§Þ‹šÎÖ=–œÂ<Ý<ÿïiWE0…ÝYPé‹ox„këáãǬš?¥ß²Ÿ„®Ç â®þÀ&Å_Ào^òc ÝP·Ô'Xµkx¨¿ZÞõ ‹¹J\§ÝYãKø¸Â:Y‹Jmù×t—’ˆSîßÀF¬CÖKªlÇ”4Ûwën̾ù¼ ¾S—m„VÚÙ÷§ñ™çZoаâå¥.JþŽ‹5 ÿUümú¾TŠÛw3Ö&C)„d,|]/;/îL×ÕdùÄÊžÀ—RÑx¸¦¡‘²Aéoã*Æ%’~ªï~iF åVôm¯³³/û\²>øÓ¾Œ$©üpQè,vêã­!šìÜý]¦~Ôܧº¸v=Ýø{ùEôëXBhm¬ÛtMþÛ¢ÊÈ’ÙÔŒ;"ÁÁ9:w~YÙ²‘Á‹ž3™ÔZÖÅ 8¸@Í÷ïNªˆ²(®¶³¸&ßmͺ¶°P6 l´¹œœU¾Xqy€e'HÃ:ìÐÜbòè—/–îQ´ "‰€5Üøl,+ý¹†ÓÑ­®Ähó&óýh#Óf®×¡–ØUp¢ˆ¤Ý”Û2ô²!&R·Q}ùhŠw;lÚú¬ÒíÍrƉôX‘¤Oý|\ÚèEè™»¨œÌáÆÖ¯( §†èJK“ä7(Hõw°¶Yf­›]2Ãݶê5rü¥)20é‹[”Ä,íИ§Þ Ѧ c(©M>Á¹ð㽶) ìxÎa¬ô›/½=5PðêóªÚb뮥;~·R%¶¶=E«¯(ÍÁÄšIÝÝ ~Ö_ËõĆ #Þžr¹ÁP(?®nËÓÇ…³4áìlâà!ÏU-Db÷±œó‰*oíyéÛV>9먥ßïµ1ÅX8ö÷ †J&T¬¶ëëÐW‰ìNœ÷ä%Üë­)°Ö­ëGÁÿ(©6Åé’w»Î¿¹®Àì"â:ƒ&.›B öC^õµÀ{·Ye2Éþä@4&%Ï2·Ž.ˆÀ'>"£½ >*üÖG!¾ÊÍP¢¥•(ûŠ€~8­=‰ý• ½Å¬ÎÒì+CJyIIJ9¤‚äô©‰{lÓ¾âi‡2vcˆÌžb·ÝGyëà*öR@æ-YJè(WÉ\£—9ã•3SLf î&ø)ß÷'†øaýh/‹SG«º3×L˜Ãß^')Hµ¹<#–°uÔKžy$Œ‚¸G,Ãð²K«÷ÖÉi[jÉÔÉ#½^‹Ƶ4æ9Œ}À¨MÏÒ¡Õ%FäÕñ®r:à]h›n·ÉW~k¸–wYkÄ®uþ`ÖJîÜ#+õ4ƸÑü•õýSÌHCG¢ØtÑíÓ´1H<“{9‹iÜoßN8^Wæ±[WéOhT«,ôJàk5à›|}è…gò­EI±h"˜oÆpv (ð<™Tù7SÀú £Þâñ€.ßR§ÌIÖ™ ”gé ƒ|™ôÊ ¨¶)j¾äÝËѲ*V©¤,Évû53‰Mãÿ ªUí³´Á—‹¡4s©Ñ¹¯21«IÞ°®„˜Jó"t¬ d…vc4ë•ìôx|óÂÜQRY<™|ùYàiñ’¢ˆÀs4íˆqÿ ‹=«gÍO‘pªuNuGÛ¥3nnïQï¯^Pr‡oZë·WåÑÇN®YEnÁîNù»|ÕÏ*1w½º# Æé:w~yòj@³†©æ†äÛäÑ\û?» 1¯ÐT0 tËÖ*ùŽ¤¯ˆ=ɶ³ÈË}ÉËúh è¹C£­Ýe; w±Ó}V,ë >Åñ±êt=ê½67D @Ì{¬Ô:@ È>[ÍF*Èâ°ñöiáÙ³ýø8VVú_ñ§u ¿+ót*ëüÐÉï0¬à¹ì•Ì7ªÒ¬íåÛÔÉvát’‹bQÔÛóžÃ_éûÈ\¤÷j >…¹Ø… Ëšø ¶VöDÖÁrC¬I’åpV¤Jm…Ë*M¢)"vK55™ðcÜÇÐ è“*dYˆüÞøXyФ'Ò¤SjœL·)Ò€™p[éÏ¡à”+÷S•¡j³¹£4|ŒQF Øð— ò5îöZ=Lº¶ ÷g’ý­J9"nÄñZÕµÜÒ·0îÐìXþóy}:D¦`OäÞ .§Çx„Ä•g¾ŒÚEâçfÀÍ3!~&D+g…çÅ¡ï B’‡ÒÌÝa(-D¸la¹ß<§^ýd¡­Õ˜À¡“‹J#UšWé¼>5ßõUê¬3¼vN;ºŠ-XFänkX-ÊD ñ<³¥n²ÁÅ9óäÈ»âH$•Â탠 ,nò6Q¿9€ÑÚ›1ŠÅ±~ø~Tû°NT÷ñÜ/¸'A!ô°.`Î OSŒÈ‹ÙŠõŒhþëî[\@™bê—eœE%›Ih¢™P€>R’ÕËô—¶köŒcn"‡ x¯å6ãªFÅ£›ì"Ð\$ L4âÙ‡ÚÞ¡1Ýw#Ì´5N€Ñóäóƒ·XlßïÈÜH)𢸙qª+"íJÉ)Þ6„Óv&¶ÎÚ<¬‘âSÞT#Ü:~$s.K7"Yr#½›-¶ø[2™6ñk¦øÛ‘àbI±4h·rË&Vç{t̑ЬAÀÉž#í‹)œ—‘€$3n«1x™‚ÕfN¤]Ä¡x&ÙË¥yî*wºƒ­Wû_PÓ3íIËçÄ×›7,ÛÝ`ÕÉhï&¬‹W^ܱbÇýú‘j“;¾¬ø ½¯žæݸlKÕdÝrݪµÛÞ»ýþn‡OpµèÀ{ [ÄËVöûšúø5ö2(íìï…r« ŠàûPC@ ˜>GÉ™x”Ýë³ ]ÝjK˶š+ÜŸþD¦ÈŒ),Õ•¿ t Jª@ûFc ɼ)õ—ß™+:9¼T©X½78{Ò9#óÎ\  8ÞsydëYµzśŒáÖ5ÑqCJ7Dp÷pRÓHvœ¡óØÎHHe£Ù[à¬_7öû+Žm#_¸h®Bˆ¤È’à³9S÷¥š[káí64 ÕEï(YtÌc„I¿Y}[<Ϲ׶@e¡$§¯ünàÏíòxu…ÛÞØá[Wó§¥&Ê a4#+@õúë|üK!G«€×}ÕDøèì£gX Ü9¿¶¹# w×e=$ E©§ÜLm†»÷ ..†“ÂÑ£æ…{ª{#…šXMµ<Òà0¨6mþãðõ?{ÑFÙëâþ‹l†Ô µî:ÎÐçRwñ¨9ÐýÆÓ“zôÇ ”ÒÕgº¨á†‚výIJ*¨Ž]¡Žy­ó{q§WlÒ èÔ »çcrÉ%2^Ò ¢ŸÇ{êy³ñl rIÒS#vAdñºÜeQ“(à¾.@N騀Üã²6×)uDY´¿ d%)öÐTæZ¬ÜóÚVŒ @A½ö‡ü([êP‹¾Ì©Ölz~úû%ç~M/àÒ§2H~=Ö¾[E† C.Š÷ æ¥Q&MŒ`eþU¯‰'¬.¶Ìhêk›Ã(iÞm² <‚WŒ¨Ô 7 誽™•ÍH&ú5ƒ´Ä‰Gv^“ê,°d8U#¢^{Â`’O¹aR3t¾:â¡8¯›£;t©¸&°,BßÔ³ ú¸êy·`]’;‚B{+µâhöC§ˆPT ‹ƺ:ϪV*O ï«)v–ç„•ÿ~³«§²:"“Ñ»ô/\9æ8±/[‹×ºy½Š-R؇Ç?,ı¢‚h7ÙãWE‹9·Qí Å ƒP½ùÏd!*Oæ—4ÕДµý»ƶÍÀB Ž»ñEÁhï`Ä6 I¢rµÂ4FÞF1gïÆåD¶ù³Ã^MS´«íEý/•ãëïÖqiyc7ø¢ò|n]™,ù÷ˆ•Ó óZ8UÌ°aïÅ¿›T¨É‚5$¸éÄtäθî÷ü…àþqÝ5ˆ£tšAÁâ hɉB+îìÂÄqÃ…÷Až@E9L,(|wQ3:܈'׆ôú ×WskµzØ_[λL“«}Ø Ê‹ÙqQØÐr¤Ú¶íÁ¼¿f“¤ nQžH.»¾Qôp„å ØØ$ùcºž'_j‹‡ˆ¿„n#\Uœ·>ó2ãb4,O9DÝNdW€V¤zEU¤z¾ ç˜ž× âÅ«³ Y_tÇ[ÁôHv¿Ÿüù”¨5RØbÍö"ôVqÓÁ¦å>Ýæ ¡é¬xZ¿9N†1«=áz|ð®˜:{å€ì[R Š¤‰èšá”9a|ç`Àˆ#kUÛÌ;mÚ¹ÎÌË\ö™++?Z]ð$8/3¦ª8{uOGø'Pc®¡÷ûeÍþoÇj¼9Ld©CkõÕª¶C©ã±¢'ÅÆÔ¡ÇTþNÊJ\'B'noÿÆEÀ„éô­`ô7JˆrÑià°~#Ïöµ”_žKªÔúŒ*£ÔÌuavñ eß‚º˜Jå숨n2Þ~¥.•´H±vyn¤ù9¸1$rô~¼lõ‡zëåÑñ7kíàjÙm§n$vÀ:MµP &¸œD2Q‘an^´«\…®eóœ6üÞÚ—qHØIø Çbâ* §Š¦¹kŠè„­\QeFco)Îày§Æ\ý×X4ËAFß\† ÜUÅÀf`èS1m–KÂ<(À‘›QL#rbübnyôm™hL* 0_Xäï?ÑfR{éoì •¹ôˆÏ…Ólþ¥ïIaTã5 M ]G‚­Ã4öž*fÿEO›ÛJ¬!•%n <ýtèÍ!Ù$ ¾RsXÎz­ñQ¸c4!EС ç:<Žðû¡‡á].ŸÈßÄÙ®™¡Íí[s…¥µ.‹þP,EÏlñÔ DлÛíšuGIé²ã°æKTcewá­9G÷»rýÜÞçݦ®Hn°”»º…W£ìdôãvÕ"Óàó®eZR‰þ"àß+¾RÖ¥þtj 4ûÅ:ØHýÔ ã©ÐMB\Œ%P§ž×å…®m ü,2Ôò#"§4®h 2îEÊ!“£ÝŽ0‡(2×ÂÌ4î|Ä– Á€"}:؃*˜¿TÌÙ¼©é*yBîÑhk?lÔ&:(rRŸØ”e¸5éŽ~qH)ç©‘$zlÂœ8ÔÛA²\9†©8´À­6ä¶eÊyˆt:§´ . JôúA¦oùìfÁ07â”3_»>÷Ö5 Ùš:ÎËœ¯sĉåvb|øú_òÇ©möãžòŸm„!ú+X=¸„ß ¾f¹©#ÇÄ`y÷B+$¿«ÏkÌ*Ì´Ü<½bq >Y|eZ£)íEPó±?Y¦ä×ñàY/b‹dc…Óœ¼„†«íQsÈø53Q²»Š%7DZõ¤â9ƒ‡¾ªŽå’ߦ*®µññ•ì4Á©¤Ccê ¥v¹“rðnt\Ù™:;B²DƒÔí &‰à³áèú$[÷ÖwüÛÕãt\Ì´æèÔ-ë%B‡²,D„Ð0Õž¹z±àÖ¨Ù\ ˆ¨&ÿÒ¯`9Æ ïšøð€wx‡ðö3[†½°@©…+÷ù\„XÎÎÂÃ.œÐ0#éÝr ùirº1¦ ÀúmX—qÄ1ÓTBà³·wo•ŠÑÔÃÜ„"î"‡pÿ6@À\·íÈI‘u_úägk«etÚÛ%®Œ07¾ÎžFµã¡OŽ)«oË:D>`£±ù¬œÈžó0·¿îÑŒWÕÇôñ}VáRݱßxïyáYÂI4’ ëkj0Ov[ß¼’šL[•SâÌ óñ“”ØÍÊsÁëªè™?¿/sBRÒ8ù¡æ²RÅA% $+²Kyâ 1µ:/Eᜑ%U}ª¶J“róL¹ïV·L¯T#] ÔÜ%0Ѧ=/Ò Ûä'‡l²óþ])jþx«†8™_ÞÅÎÑ"îË]ð§ûf·€ ~sÂÎêµÎ’f¥m7]ö¨ZEÜù q?Å`'ÁUµ omàØÀ¾<¢w<µƒ Ú윖¤_ˆ¾@[RBâã1!êÞq-æ-Ú‹Ph¯K’‰ûMུ3k?¬õ…$¬ŸËïUÄ–×Mâ’Iºm.ì-Na4ŸâÙF ]XçQ´á6¡Ô¡"gwšRe ]!CÕ¬>¨æ€:–„-}ç5>Ó>šÞðÏanRÑ%3漬—­àsŽ¨ƒ&ñr z½qÓsŠ´»úÚ[ H{"‡ætBa‹3ÝWØÒßž}€å¦äþVî¹÷ö5‹öšß¢ ¤I”‚Õ‘gõ0ÐQvA&ĸ‹¿D?Þæäò·ñëîU+Á´V•ß¼’a §2[`Ÿ%ù·bp£ )¬¸Æªªý-T`‰Á8iÁ´ùÒè<ê°&íí×èL„q·Åɇ"i”žDz’lÕYô ß3(vÝZÆìròaFÒÂj›H QA98-­³iåÆÍfþ&Hƒû„Ѧt®­ ˆ¦î9É/îRrRO Ôù;)›žöÒ ÷ƒ‚)¶ ½¸sÔ.¯à°J_§‰ÑÁ[P…^ùÚ_Àbë7¬9IÀTˆ’*o§üwÈi¾? ;‡¶‹¦“ýjv _ÍÃy¢ÓGƒ7;“8{T'·Õ{iW€éçòÊ»·]âß .ýWß®¤Ã· ¼ë#'@#Lâê£2·7ŠX²ûCœ¹ÐþÆ%E)¡„4üû…õcMúσðc ª„gÑg3I¿`ƒ8ïçBLuK+¥ÿíÔ‚rJïYLÞZ«O¾FÙ B\ œ+þ 8PjÝ÷Aï ²ƒË¶xåúÜ5*¶%O•t›5 Uq‚BÃŒÏI ïê ‚ê)¾Ôfµ9‰jŸÌŒ§ x9oÔ)bIÒ9‘QðTÕ©øY®BTˆqgæÓ‰EâËTÈc ¸}}3€>©BLé)á`Y)U0ª61!"S–fµ4ÔÛÞÖ Yé„;‹-h©vêLÕÏôro'²C*OMCj‡ÿS8Ú³Ü2žš©æÉ¥§Š[xt‚þŠ9¸ qè¸1@µCÜéA[õ+Ò”ýúrH JÑR¬}³ªÑ'Ú†ÛÇ\W?R¾Ö¨È­Ây¥^ËB](8r•ÎÚ¹ïÏA¢ÀÛN4:šÅ8èé@ ,4žrR÷Rqb΀Ë2Îíl`ÛcJ–î/Õôsǽ“GÿìÅüa`΋ª#k5!Xuª¦~±öRŒ¾ŸNuß¾©'ìDD@tË.ñà4úNí–˜$ÀJ–¦–&»ü´c‘I ¯|V®H•–”±ïÅÈos,‘R)’MÉ# ÆœÏy…kû6x5íÖ-~(ÍZÿ“ÜlD‚ ¦V#´~xr× î¼&÷ðû?4’·b\)•‚…ùJÕ*>…‘æuž†É‰cšMîø0˜¢‚ ÷Æ+WZìÒ/eeíѽßR–Ï%~i³!¸³ þaZ(årz ÂX®RƇfŒDe¶èŒCÄ6¼åéc q؇tm"Ðç«ô±„Ô'n=ïÏ[1·Ò%ÙœV‚®[ï!q͈—÷s3˜ñL+,Oµ9‚Ó®rëbG’© ñ¬wA Ùoo´Î¥nÇy¡ð Šþb/;ÍCt”DC}ãýö¼IÑ»`Î gÍa µ`¯·ÃÈó¦×"ŽØ±’g¨»AÙÅhEèa ÌwõœbÞá¨È?Gùøª†b®¨d2‚ ÅæâRt¹æ™éü¦Dñ<ûíâF]š‰ôùåÕ(‡Ø“š‰âo%¢Þ‰làø#¢~ÏT?ë€Lšr°ÇŸókè4Œ®wx“€J° endstream endobj 67 0 obj << /Length 696 /Filter /FlateDecode >> stream xÚmTMoâ0½çWx•ÚÅ$ !Ù ‘8l[•jµWHL7IP‡þûõ¬V=Mžß̼ñ s÷ëu;ÑU··õÈÙ›=w—¾´“ì÷îÝÝå]yil;<[[Ùj<=?±×¾+·v`÷Ù&ß´õðàÈ›¶<^*;²~&ûQ·‚>ìþÝþ”MS >Ù_êãP·ò{=éÇsæ@öd”ôÇöçºkŸ˜xäœ;`ÝVY×`Œs4½JaÓQÜ¡n«þª‡í¡.’Uu9\ßèY6î>¼ý<¶Ù´‡.Z.ÙôÍž‡þ“4>DÓ—¾²}Ý~°û¯ÒÜÑör:-d0­V¬²WÑÍÿ¼k,›þ8ãóþy²LÒ»ðºÊ®²çÓ®´ý®ý°Ñ’ó[Å*²mõíLrŸ²?ŒÜÔqù¥ã• â5F8@ šˆ=@Šð)&°  È8Ô¹€ÂÅRx u€Dº\j2H—†ª¡ÐVÁ¹0CzL]ø Âb°ct‘I ©g$`htÑ‹0œÆ\F„áŒ0ä†sê‡á jd< —Iê6œ»õñzgóñºË»þê W ¤qÈ’£+—Ÿ#ö•ñÌÇkÄÞ .‰bªsré…¤šáæÄç†bïmŽXú¾„Kß7ǵHß7Géû„û¾nb§>&jÊصäuœ¯¼ú•ñ1ÜV™÷•âÜãâµÇ‰Ou$ÕŸqWèS/%1{\øxB!€§ÔK(hH©—TЖ枃»J©Ïϯv×ÜëÁ=küÒ2ø¥UðKÏ‚_:~é$ø¥Óà—ÖÁ/¿Œ ~™Eð+7¿èË¢/ ÿlì¡ÛÒ(/}ïö -+ZXukoûìÔE?Z„ãæÅÛKýqíƒÄ endstream endobj 68 0 obj << /Length 695 /Filter /FlateDecode >> stream xÚmTMoâ0½çWx•ÚÅ$ !Ù ‘8l[•jµWHL7IP‡þûõ¬V=Mžß̼ñ s÷ëu;ÑU··õÈÙ›=w—¾´“ì÷îÝÝå]yil;<[[Ùj<=?±×¾+·v`÷Ù&ß´õðàÈ›¶<^*;²~&ûQ·‚>ìþÝþ”MS§“ý¥>u;áà¾×ÃÑq~:fc_0F)l®»ö‰‰GιÖm•u f8GÓ«6•ê¶ê¯bØÒ"!YU—Ãõžeã.ÉÛÏó`›M{è¢å’MßÜáyè?IáC4}é+Û×í»ÿ¢Ìl/§ÓÑBãÑjÅ*{pÝìϻƲéOÞ(ïŸ'Ë$½ ¯ªì*{>íJÛïÚ-9_±eQ¬"ÛVßÎ$÷)ûÃÈM—ÏñP:^9À ^`„ª‰Ø ¤Ÿbr š€Œ@ ‘{@(\,…RH¤Ë¡&€ti  mœ+3¤ÇÔ…Ï ,;F™$Б€‘zF†F½ÃiÌeDÎ(ó0œAº1a8§ÎyΠFÆÃp™ nù[¯w6¯»ü·ë¯Îpµ@‡ )9ºréñ9b_iaÏ|¼Fì-ÐÐà’(¦:×ù(—nQHªY^`nA|n(öÞ戥ïK¸ô}s\‹ô}sÔ‘¾oA¸ïë&vqêcâ ¦Œ YK^ÇøÊ›!¡_Ãm•y_)Î=^ ^{œøTGRý÷w…¾1õR³Ç…'ÄxJ½„‚†”zImiî9¸«”êðøüj'pͽܳÁ/-ƒ_Z¿ô,ø¥ãà—N‚_: ~iüÒyðËÈà—Y¿2qó‹¾,ú’ðÏƺíŒòÒ÷nЪ¢5Q·ö¶ÍNÝ Yô£58.]¼½Ñ»á‚ò endstream endobj 69 0 obj << /Length 739 /Filter /FlateDecode >> stream xÚmUMoâ0¼çWx•ÚÅvHU„dçCâ°mUªÕ^!1ÝH ý÷ëñ#xÙö?ŸgìÁÜýx]OTÝmÍ$|äìÍœºs_™Iöss îîò®:L;<S›zœ==±×¾«Öf`÷Ù*_µÍð`É«¶ÚŸk3²¾'ióÑ´ž‚}Øý»ù=©½à“í¹ÙM;áà¾7ÃÞr¾›f¶ÆnjÌ-ùeúSÓµOLg~¼À8÷ã ãâþÈ)okà çA„8 ö$`I\èÎ×3`çAfŽã<ÈZ]ƒÂ!‹„ê xNkÇyã¹ãÐð"œ7Á¿ _¥ã“§Ìq âH`òáö•‚nú¥¤kÌÂðRONH=CpB:# =Ñ%8“ˆ88QA~¡!*ÉzÆœøÐäT?!~Ž> étw©8éÄy*ás£¤Ï }nÔÌçFE>7*ö¹Q‰ÏR>7Š²¢ G]¼;~îó¤ŠÛ<©ò6OšßæI‹¯yÒòkžtèó¤g>O:òyұϓN|žôÜçI/|ž´òyÒÚçIg>O:÷yÒ…Ï“.}ž2îó” Ÿ§Lú> stream xÚmUMoÛ:¼ëW°‡éÁ5?$R. ¤d9ôMðð®ŽÄä ˆeC¶ù÷³k›m‘CŒÕp¹;;†wŸ~>Î|¿Ž3óEŠ_ñ¸?O]œ5߶‡âî®Ýwç]Oßcìc]=~?§}÷Oâ¾yhÆáô9%?ŒÝ۹׬“B|Æœ‚>âþ)þ;ëvÇw%gÏçáí4Œ3‰ä§áô–’>\ ‚‚6ý§ã°¿ õEJ™€õØ7ûÆ8ó 1¿’{Æ~ºðÏ`W(-ú¡;]¾è·Û%=°ùñýxŠ»‡ñe_,—bþ+-OÓ;qü\ÌL}œ†ñUÜÿI--=ž‡·B«•èãKª˜æÿ¾ÝE1ÿpÆ[ÎÓû! Mߊyuû>Û.NÛñ5K)Wb¹Ù¬Š8ö­iÇ[ž_®¹uÊ•MúÑzQ­Š¥Ò)V†€Ú(TØ€àx¿àÞ¢ žjy‹°°!ÀÐÔ•µZÔÀ2àP="¦ZdÔ0\ÃG©R\¡·”).–2*ÎШa!„U¼Ä,†³ÔÛHð° `+jÐÃ.¸5Nα@èâ°èÐVK-àxŸ%ô˜Ü3š% A°YÓ€z¡ÎšÔ>kP#¬³¦õ™5m0W£oš¦Ã¾žj­®§Üý·.†ÐZ¡ŽT$X/©)n)æ#W—„o(æ“oÀRZÞ $K¢p4’ŽZ¶-bâ\­1¦Ü°Jä æP"Gñ‘XÔQ¬‚i/8ºkÉ^€ÂZqŒ:ZsŒ½š9”d š­Bù Ž)ßsLù-ï7½æx˜ÏJ›¡¾Ò`¯ažÉ½)f¥É$†µ’1™¸ dÑŠcªCZCù<£7Ã3JÊgózÌnøþHÈ°íáÌYÉšäTœ¯a…Šï¯Æ,_»œ-Ÿ—Oë87Ë}êÛKÔ´Ü—Ll¹oKñšò+Êg­JÌâ.¾GZyóº‹Vðc­48¸’ï¼äØWtù]Í:P~`áŒñ±–rZŽq.nÍ1]Ç ÇàSÿæ/©ßP•ýïuö¿7Ùÿ¾Ìþ÷Uö¿·ÙÿÞeÿû:û?Èìÿ ²ÿƒÎþ&û?”Ùÿ!dÿ‡&û¿1y–¦¼ÍH·œn5þ¹ã)º½ÝyšÒ“Bï½x#†1Þž´Ãþ€]ôGoáõñÅ×Mñ?®Xê endstream endobj 71 0 obj << /Length 900 /Filter /FlateDecode >> stream xÚmUMoÛ:¼ëW°‡éÁ5?$R. ¤d9ôMðð®ŽÄä ˆeC¶ù÷³k›m‘CŒÕp¹;;†wŸ~>Î|¿Ž3óEŠ_ñ¸?O]œ5߶‡âî®Ýwç]Oßcìc]=~?§}÷Oâ¾yhÆáô9%?ŒÝ۹׬“B|Æœ‚>âþ)þ;ëvÇw7{>o§aœIä> §·”óѲH˜ø´åŸ8‡ýøU¨/RʬǾÙï0ñ˜_xˆù•ÙË0öÓ…ŒxµBiÑÝéòE¿Ý.‰ÍïÇSÜ=Œ/ûb¹ó_iñxšÞ‰áçbþcêã4Œ¯âþfiåñ|8¼E°²X­D_RÁ4û÷í.ŠùGÞRžÞQhúV̪Û÷ñxØvqÚŽ¯±XJ¹ËÍfUıÿkM;ÞòürÍ­S®lÒÖ‹jU,•N±2Ô@  "À–,Àû  ð õTË[<€5€ €¦¨¬Õ –€ê1Õ"à†á›×cvÃ÷GÂ@†m¯gÎ üKÖÄ §â| +T|5f©øÚÕàlù¼xZÇ1¸YîëPß^ê ¦å¾dbË}[Š×”_Q>kUbwñ88Òʘ×]´‚k¥ÁÁ•|'à%Ǿ¢ËïjÖò{ g䈵”ÓrŒsqkŽé:n8Ÿú7ÏxIuø†ªì¯³ÿ½Éþ÷eö¿¯²ÿ½Íþ÷.ûß×ÙÿAfÿ•ýtö0Ùÿ¡Ìþ!û?4Ùÿɳ4åmFºåt«ñÏÑíÙèÎÓ”^z­è¥À1Œñö öì¢?z ¯ï.¾~lŠÿP}éL endstream endobj 72 0 obj << /Length 868 /Filter /FlateDecode >> stream xÚ}UMkã0½ûWh…öF²ã¯²CÛ–¦,{Ml¥hìà$°ý÷«73nw—ÒƒÍÓèÍèͳ,]}{\OlÛoÝ$ºÕêÉúËиIù}s ®®ª¾¹\w¾w®uí8{ºSC߬ÝY]—«jÕíÏ7ž¼êš×KëFÖç¤Â½ì» ÖQ×ÏîçÄ5Ûßf6Ó“æ09]Žn˜œÍD#åy~õÔ/XÊO©Ï¦øá†Ó¾ïÕZûÀ²kËþ€ÆNÁTÄ©é(w·ïÚAª-ô&Tí¾9ˈÞÍÁ;„äõÛéì«n×ó¹š>ùÉÓyx#½7ÁôahÝ°ï^Ôõg=a}9_Ä(,ªu;_×ûr¿985ý¢ëwæóÛѩƆ56}ëNÇMã†M÷â‚¹Ö 5¯ëEàºö¿9qÊv7r—ž«kÿ u”/‚¹A² )`JbD>`´öØ2ãš™$`¤TY'`ä`ä9&£Ä*×ð8˜W`TœR±¤&4–`Ô(ZsJ¢5Rê’H©+¸†Çhì3Ÿ}7¿6ƒXäÅ¡°6„‹0×a™G„+˜ gıXFǯ€Æ8åÜ8ã8ñsέ-Çá´.x]â”/€+ö5MĹˆÖPN‰£é±¯<1\?‚Ãõ“Æåǘ¾tõ1¦:Ë¿ø#§þ7^ïÂYH{upÍý¤²àQF^o­Ç†1¼ÎBÆЖEŒQ?›1^“׆¼ÈƨŸ¥Œ‰“qÿЉ´.yafð%+CsV2†GYŘêSÿ&ÆžÈjÆЙ³þëæ¬??gý´‡rÖCOÎúÊeý zÏYJ|ÖŸŸu¦Ðœ³Îß8gå²ÎˆrY§¡‹}¶èÅŠÏèÑŠÏ1°øLñkYñûÒŠÏX׊ÏðÓŠÏÄŸÑ»Ÿ¡ßŠÏÐiÅgônÅgøfÅgª/>C¿Ÿ¡³Ÿ±n!>ƒ_ˆÏàâ3ôâ3åŠÏ轟‰/>?ã“°œEè¥ÿÑK!þcâ?ÕäÅRMþO ª#þƒSñZ1|¨8Gxp4Éaž«j9¡èDÂQKæý.h.Ã௠º‰èÜlj¿ïÜûeuìÈ¢‡n¹ñ¢Åè¡þðç( endstream endobj 73 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œ.RIý÷ë73´»«ªÐóøÍøÍÃØW߶3ÛÏnÝjõèNÃelܬü¾;WWÕÐ\®?ÿp®uí4{ºSãÐlÝY]—›jÓwçOÞôÍë¥uësRá^ºþƒ‚uÔõ“û5sÍx0ZëYs˜.G7ÎÎf¦‘òÔ_=õ –òSê³)E~ºñÔ ý2·~ÞÖ}[4v æ"NÍ'¹û®oGQ¨ž¡70¡j»æ,#z7ï’·o§³;lúý,—jþè'OçñôÞóû±uc׿¨ëÏzÂör<¾:ˆQ:X­Tëö¾®÷åÇîàÔü‹®ß™OoG§BÖØ ­;wwý‹ –Z¯Ô²®WëÛÿæLÄ)Ïû‰»ö\]ûW¨£|, ’MHS"#p òÐ# µÇ>y×È| #¥Ê:##Ï0)%V©¸†ÇÁ²£â” Œ55¡)°£FÑšSj­‘R—@J]!À5£w+>Ã7+>S}ñú­ø …øŒu ñüB|¿Ÿ¡§Ÿ)W|Fï…øL|ñ™øŸ „å,B/…ø^ ñû°ÿ©&ÿ+–jòRPñœŠ×ŠáCÅñ8ƒ£Iæó\UË E'Žj\2ïwAsGMÐMDç>Nü®wï—Õq8"‹ºå¦‹£û:ø‰9ç endstream endobj 74 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œ.RIý÷ë73´»«ªÐóøÍøÍÃØW߶3ÛÏnÝjõèNÃelܬü¾;WWÕÐ\®?ÿp®uí4{ºSãÐlÝY]—›jÓwçOÞôÍë¥uësRá^ºþƒ‚uÔõ“û5sÍx0¡Ö³æ0;]ŽnœÍL#å©;¿zê,å§ÔgSŠ ütã©ú;enµÖ>°îÛr8 ±S0qj>ÉÝw};ŠBõ ½ UÛ5gÑ»9x‡¼};ÝaÓï‡`¹TóG?y:o¤÷&˜ß­»þE]&ж—ãñÕAŒÒÁj¥Z·÷u½/?v§æ_týÎ|z;:ÒØ°ÆfhÝé¸kܸë_\°Ôz¥–u½ \ßþ7g"NyÞOܵçêÚ¿Bå«`ilB ˜„‘„­=öÌã¸æ@æ )UÖ 9yŽ€IÁ(±JÅ5<–§T`¬© M55ŠÖœR£h”ºäRê ®á1ÚŸúÌSßÍïÝ(yq(¬ ábŒÆuX&Àá &èq,–Ñ1Ç+à„±N97Î8Nüœsk`Ëq8­ ^—8%Ç àŠ½FMq.â†5„Sâhú@ìkO × Ápý$Áƒqù1¦/]}Œ©Îú/þÄ©ÿ»pÒ^`ÜD3F?©ìx”‘×Æ[ë±a ¯³1´ecÔÏŒ×Àäµ!/²„1êg)câdÜ?4dâ­K^˜|É ÆÐœ•ŒáQV1¦úÔ¿‰±'²š1tæ¬?ƺ9ëÁÏY?í¡œõÇГ³þ„rY‚ÞsÖŸŸõ'Äg)4ç¬3Å7ÎYgD¹¬3¢\ÖièÇbŸ-z±â3z´âs ,>G|ÆZV|ƾ´â3Öµâ3ü´â3qÄgônÅgè·â3tZñ½[ñ¾Yñ™ê‹ÏÐoÅgè,Äg¬[ˆÏàâ3ø…ø =…øL¹â3z/Ägâ‹ÏÄÏød ,gz)ÄôRˆÿ؇…øO5ù_±T“ÿ“‚êˆÿàT¼V *ŽÇM2G˜çªZN(:‘pTã’y¿ šË8úk‚n":÷qâw½{¿¬ŽÃYôÐ-7]´Ý×ÁÎ%ç# endstream endobj 75 0 obj << /Length 867 /Filter /FlateDecode >> stream xÚ}UÁn£0½óÞC¥öƆ¦Š"R»­šjµ×œn¤"’ú÷ë73iwWU çñ›ñ›‡±¯¾=¬'®žý$¹ÕêчóØúIõ}sˆ®®ê¡=ï}úá}ç»ËìñN=ŒC»ö'u]­êU¿;Ýòªo_Ï¿°>'•þe×P°Žº~ò¿&¾÷&í¤ÝOŽçƒ''3ÑHyÚ^õ – Sê³)E~úñ¸ú;enµÖ!°ì»jØ£±c4qjz‘»ÝõÝ( Õ3ôF&VÝ®=ɈÞí>8„äõÛñä÷«~;Dó¹š>†Éãi|#½7Ñô~ìü¸ë_Ôõga}>^=Ä(-ªóÛP7øòc³÷júE×ï̧·ƒW1 kl‡Î›Ö›þÅGs­jÞ4‹È÷Ýs&á”çí…» \Ý„W¬“bÍ ’MLS!"q Ø# uÀ!`NØÈÀÈ©²ÎÀ(À( LF…Uj®p4¯Á¨9¥cIMh ,ÁhP´á”E¤4ÒÔp€Ñþ¥Ïbvé»ý½Å¢ …µ!\΀ѸŽ« 8!\Ã=#ŽÃ2:åx œ1vÀ9çæÀ–ãÄ/8·v‡Óºäu‰Sq¼®ÙkÔ4 ç"nXCI8'Ž¦ľ&ðÄpý ×Ï2<WcúÒõǘê,ÿâ_8Í¿1ðbxÏbÚ ¬Ã€›iÆè'—Ý,ym‚µÆðÚÆŒ¡Í&ŒQßÎ/ÉkC^ØŒ1êÛœ1q,÷ V¼ uÉ 3ƒ/¶d Ͷb l͘êSÿ&Åž° cè,XŠu ÖŸ‚_°~ÚCëO¡§`ýå²þ ½¬?'>ëψÏ:sh.XgŽo\°Î„rYgB¹¬ÓÐÅ>;ôâÄgôèÄçX|&ŽøŒµœøŒ}éÄg¬ëÄgøéÄgâˆÏè݉ÏÐïÄgètâ3zwâ3|sâ3ÕŸ¡ß‰ÏÐYŠÏX·ŸÁ/ÅgðKñzJñ™rÅgô^ŠÏÄŸ‰oùd ,gz)ÅôRŠÿ؇¥øO5ù_qT“ÿ“’êˆÿàÔ¼V jŽ§ M2G˜çêFN(:‘pTã’y¿ Úó8†k‚n":÷qâïzÿ~Y†²è¡[îrÑbtßDÒÓçn endstream endobj 76 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œn¤" ‡þûõ›™´»«ªÐóøÍøÍÃØW߶3ÛöÏnÝjõèÆþ<4nV~ß‚««ªoÎG×M?œk]{™ïÔÃÐ7[7©ërSmºÃtãÉ›®y=·îÂúœT¸—C÷AÁ:êúÉýš¹fœŒÖzÖgãùä†Ùdf)O‡éÕS¿`)?¥>›RTà§ÆCßÝ)sëç}`ݵeDcc0qj~‘»?tí Õ3ô&Tí¡™dDïæèBòömœÜqÓíû`¹TóG?9NÃé½ æ÷Cë†C÷¢®?è Ûóéôê Fé`µR­ÛûºÞ—»£Só/º~g>½œ ilXcÓ·n<í7ìº,µ^©e]¯×µÿÍ™ˆSž÷îÚsuí_¡ŽòU°4H6!L‰@Œ@ÂÈBŒ€Öû@æq\s óŒ”*댌C¿Ÿ¡³Ÿ±n!>ƒ_ˆÏàâ3ôâ3åŠÏ轟‰/>?ã“°œEè¥ÿÑK!þcâ?ÕäÅRMþO ª#þƒSñZ1|¨8Gxp4Éaž«j9¡èDÂQKæý.hÎÃ௠º‰èÜljèÜûeuêOÈ¢‡n¹ËE‹Ñ}üMçA endstream endobj 77 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œn¤"Bý÷ë73´»«ªÐóøÍøÍÃØW߶3ÛöÏnÝjõèÎýehܬü¾;WWUß\Ž®8׺vš=ß©‡¡o¶nT×å¦Út‡ñÆ“7]óziÝÄúœT¸—C÷AÁ:êúÉýš¹fÖzÖgçËÉ ³ÑÌ4Rž㫧~ÁR~J}6¥¨ÀO7œ}w§Ì­Ÿ÷u×–ýƒ¹ˆSóIîþеƒ(TÏИPµ‡f”½›£wÉÛ·ó莛nßË¥š?úÉó8¼‘Þ›`~?´n8t/êú3ž°½œN¯b”V+Õº½¯ë}ù±;:5ÿ¢ëwæÓÛɩƆ56}ëΧ]ã†]÷â‚¥Ö+µ¬ëUàºö¿9qÊó~â®=W×þê(_KƒdRÀ”Ä$ˆ| 4Àhí±dÇ52HÀH©²NÀÈÁÈsL F‰U*®áq°¬À¨8¥cMMh ¬Á¨Q´æ”Ek¤Ô%RWp ÑþÔg¾˜ún~ï±È‹Cam `4®Ã2ŽW0A/ˆc±ŒŽ9^'Œ-pʹ)pÆqâçœ[[ŽÃi]ðºÄ)9^Wì5jšˆs7¬¡ œGÓb_#xb¸~ †ë' ŒË1}éêcLuÖñ'Nýo ¼Þ…‹öë0à&š1úIe7À£Œ¼6ÞZ cx…Œ¡-‹£~¶`¼&¯ y‘%ŒQ?K'ãþ¡!/h]òÂ,àKV0†æ¬d ²Š1Õ§þMŒ=‘ÕŒ¡3gý1ÖÍY ~Îúiå¬?†žœõ'”Ëúôž³þ”ø¬?!>ëL¡9g)¾qÎ:#Êeå²NC?ûlÑ‹ŸÑ£Ÿc`ñ™8â3Ö²â3ö¥Ÿ±®Ÿá§Ÿ‰#>£w+>C¿Ÿ¡ÓŠÏèÝŠÏðÍŠÏT_|†~+>Cg!>cÝB|¿ŸÁ/Ägè)ÄgÊŸÑ{!>_|&~Æ'a9‹ÐK!þ£—BüÇ>,ĪÉÿŠ¥šüŸTGü§âµbøPq<Žðàh’9Â> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œn¤"Bý÷ë73´»«ªÐóøÍøÍÃØW߶3ÛöÏnÝjõèÎýehܬü¾;WWUß\Ž®8׺vš=ß©‡¡o¶nT×å¦Út‡ñÆ“7]óziÝÄúœT¸—C÷AÁ:êúÉýš¹fM¨õ¬9ÎΗ“f£™i¤<ÆWOý‚¥ü”úlJQŸn8úîN™[­µ¬»¶ìhìÌEœšOr÷‡®D¡z†ÞÀ„ª=4£ŒèݽCHÞ¾GwÜtû>X.ÕüÑOžÇáôÞóû¡uá{Qן ô„íåtzu£t°Z©Öí}]ïËÝÑ©ù]¿3ŸÞNN…46¬±é[w>í7ìº,µ^©e]¯×µÿÍ™ˆSž÷wí¹ºö¯PGù*X$›¦D F á@ä¡F@k} ó8®9ù@FJ•uFFž#`R0J¬Rq ƒeFÅ)kjBS` F¢5§Ô(Z#¥.9€”ºB€kxŒö§>óÅÔwó{7ˆE^ kC¸X£q– pD¸‚ zA‹etÌñ 8alSÎM3Ž?çÜØrNë‚×%NÉñ¸b¯QÓDœ‹¸a á”8š>ûÁÃõ#h0\?Ið`\~ŒéKWcª³þ‹?qêcà…ð.\„´X‡7ÑŒÑO*»eäµñÖzlÃë,d mYÄõ³ã50ymÈ‹,aŒúYʘ8÷ ™xAë’f_²‚14g%cx”UŒ©>õob쉬f 9ë±nÎúcðsÖO{(gý1ôä¬?¡\ÖŸ ÷œõ§Ägý ñYg Í9ëLñsÖQ.ëŒ(—uú±Øg‹^¬øŒ­ø‹ÏÄŸ±–Ÿ±/­øŒu­ø ?­øLñ½[ñú­ø V|FïV|†oV|¦úâ3ô[ñ: ñëâ3ø…ø ~!>CO!>S®øŒÞ ñ™øâ3ñ3>ËY„^ ñ½â?öa!þSMþW,Õäÿ¤ :â?8¯ÇŠãq„G“Ì湪–ŠN$Õ¸dÞï‚æ2 þš ›ˆÎ}œø‡Î½_V§þ„,zè–›.ZŒîëàÍçP endstream endobj 83 0 obj << /Producer (pdfTeX-1.40.25) /Author()/Title()/Subject()/Creator(LaTeX with hyperref)/Keywords() /CreationDate (D:20241029220648-04'00') /ModDate (D:20241029220648-04'00') /Trapped /False /PTEX.Fullbanner (This is pdfTeX, Version 3.141592653-2.6-1.40.25 (TeX Live 2023/Debian) kpathsea version 6.3.5) >> endobj 2 0 obj << /Type /ObjStm /N 55 /First 424 /Length 3119 /Filter /FlateDecode >> stream xÚíZ[o·~ׯàc‚ÂâòN|‰Ó´±ÚNœKý Èk{Ou«$§iýù†¤¤Ý••XIûp€6Å%gÉ™o¾ÎJ+XÁ S³LXÇÁ3Ï”ÐL¦½aB2#1ÂŒ,0‹é˜/ŠŽ ̇À¤eA+ÜÂD!S’ …{•bÂk|BLú LY|jÇ”cÒ@Hy&½ šÒEÁt)([µÄbN(¦5ÓÁáÓ2# —ǫ̈`˜è鱘ÑÌÚtŒeVzLzfm Ø „•Ì ‡kÍœÉÖ2ç±0ðz دq³”̘'`¥Ó¢#1ｄò,˜0[XéXP8iÀà bÁjÀ‚O,È$ÆIo‘BÔ‘Áv/”†)øÔ…¨Ô¡ž°„U û¡ãeçÉÆÏ6>3~ÀÍÊþ¼ºâ1ûñÇΣߔ³øw¿EqDMEÍlÙÄË57Ôü†íèó#5=jÔÜQSÖâ}Cj&Kùò¦ùRd×ÜVó½½N?¬æ—¤ç“'~þפdüeï¦ìðýñh^Žæ3I‚~ZÎÆwÓ~9ÓâÀQyUõöÆŸØ{’0Át%|^u}¸ì`™)î'H£ôîh4ÆjïÁ`Ú÷²ÓØ3ÎvøÞxzUNãzâ’ÿÄŸó}\ñKÚ¿?gï…“]]hPJw-ø eÑ p¥ÑE×¹³»s¬É_T£ßùî“'q¾}ÂÏø«Óçôÿèv>Ÿü‡óÕ¸]§ÃÞ¼êϺÃI5œìÜŒCtoºW% =³ª+ÐÞ¼}‡àS]_€±FuƒQlt7\nÕITØnäCDµ] ª7DáF~ž¦gúãÞ3¯^î]œýüÃþÑÑsQ`bл™1$öâ· ;xLØ‘±Ô±ü»ìðÝYŸ¾T°’û½ÉOeus‹K¯:œ¶¡¹A“Ïç½AÕßÝ JŒ`àl^_ƒ‹þ&ߥ•À"·½)}iðˆ?å/ù9Ã?=NJV¸U«ú·_²éÝÞÅÛãÃh“»ß$‘-NüCy±Ñ¢þx8ìñŠÿ·a‘ÙƢ׿ýòlnr’.N*·˜T, ²a³AÕ´7¼”ŸxùÇ]oÀ‡½þt<âãQÉ'ƒ»YÃN·¯aâkØyöv“¡2°`Éw‚ŠL³2Vït±ÑÔáÝ`^M5L Û˜t|öóÓ“ѤMd4°ˆ¾€”†—ä?aÑfçM¦Õ°¬›cÄVL|ùóùÅëÎ÷Þ­7ùHx¶£ýc$Š*SgcÓ$º|MÝEZ× :â=^òˆÈÁÂÄC>ãs~Ç?6LÝ*œ¾¸8Þ‡©§GôܸÁy|TÞ£[‚r5S}hz/loª*ê¦îò}~€ yÈŸÃìc~‚\y†ly>ð>2Ía˜òÍ@¹æ×ø«Î-¿ýkr[Ž€Òï-œ&å´_ñ?îÆsô°+ŸFìæŽ ?þ'ÿÄü7[¥®§'¯~M(ÊÍ(BQ­Ñ-ÿ]÷€ã³ø“ÉQÄð(®0¬¡7¾›ÞÞ¨vcB¾^”×óÔKfHë8Cþw97€Ü*7>ûébïø]ÒI¿H§)òèdޠÿ äó{ý¡TçÖýñ·UæÜ}z°N©æìü3ñ‡£`GÑy€Ò¾ð¬,DÃ`º\,¥leÏð`êP®Â¯Ç{³~U¥Cú0íõ/瑹ŸXQçV:7o¦eAW'Ø€ÊÙ¬ Ÿ¡XŒÚ«ƒÈµÙ 7»%ÆÝNËò3¼³[eüwO¹89‡Î?ãYø…œþ7<„wŽÒHx˺)J Ä„Ö¬üWð¬³ÕùðöÙññÉAFnF.€ÑôÅGø7©Ÿ)ƒÝ4¹µˆÍZ‚ʤiÀpO‚:ê¯ªÑ T¬®¯Kð~ƒ|/]:b˜ÒK.2ê§É}ôÅžƒ!+·pPT†ÜÝF>JÞŠî$ÇâôÉnLÎŽ.Ž3[¤ØiÄŒ©EíaJû¨ —Äôμã¨UêìÂyÞ$X-ÒYpïÒý RO‚¨Ru>–Géå:¿–?ij8ìõfeüáuí­AGz›$þ|XMgs"B¬Û_ôòxÙáÕÕüvF/ªDÑóñ«QgÂÏΟϩÓz¶jicÛÚ˜†6…]i£Ö´ñÛkÓz.jiãÛÚPL-µ¡o¡Œ]S&l¯Lûᥩ)Ö´©)#Wº˜¶.®Ø^—ÖSGKÙVEûš.t±PF¯)#¾ÂKkÏ -}ÖH¬´5…òû4‹tD¯;µ5”Ûk¸^ê·4\#¶t0+ ý—5TÛk¸^F·4\#;}¯†Z|YC½½†ëõiSC»Nmð²–_ÖÐl¯ázAÙÒp-.”Þäåhh¿"»®ÕZ- ×#ÅÖSšZ)ø'Eú_¯yZ ®Š6 '¯N]|YÃ{Žz/jF/FÝÑ[MþKu5[¼–Þij°'wóŠ„YVŠeI§E?-&–÷UszXÉïⱌd~‘Ê5Nª±ý£ƒqçlÞ›ÎÓ‹¦¤Ë£ ¥÷…¨]ËÇL¦wÈê¯>ª¤6Q «yk¹šÜê5¸r1Ϲ°ßCÉ8¾é$˜–/.0XhŸõöKÓnÊ#Î_ÍÊ\'“r”^dK*ÿ AĤ endstream endobj 84 0 obj << /Type /XRef /Index [0 85] /Size 85 /W [1 3 1] /Root 82 0 R /Info 83 0 R /ID [ ] /Length 254 /Filter /FlateDecode >> stream xÚ%Ð=/ÃQÇñßïz(ªþõ\ÏTiQõÜÁ;0›¼«ˆH‡Š„D"MšH b0CŒ$F»w`àØ8¿Óå“oroν9ð€†ç2T1±"(‚h¢…@ä÷šD³˜y±(fÅœ˜Y‘3bZÄ‚XqÑJÄN}r›Èˆv‘"IÑ)ºD·è½¢Oôñ_•"òû^ÄÆ›× ±]hˆ8.x W^#Äǃ×(}y‘›ß^ãäÑ‹×ù³ç•f(ÞxM2Tvl“Õ’qvhœ×Œ‹ãr˸N·¯ÆÝ®qŸ1?ZýCËbU¬‰"ÃÓº¾§ð¯% endstream endobj startxref 181016 %%EOF pcaMethods/inst/doc/outliers.R0000644000175200017520000000340414710312276017402 0ustar00biocbuildbiocbuild### R code from vignette source 'outliers.Rnw' ################################################### ### code chunk number 1: outliers.Rnw:59-60 ################################################### library(pcaMethods) ################################################### ### code chunk number 2: outliers.Rnw:62-67 ################################################### data(metaboliteDataComplete) mdc <- scale(metaboliteDataComplete, center=TRUE, scale=FALSE) cond <- runif(length(mdc)) < 0.05 mdcOut <- mdc mdcOut[cond] <- 10 ################################################### ### code chunk number 3: outliers.Rnw:70-73 ################################################### resSvd <- pca(mdc, method="svd", nPcs=5, center=FALSE) resSvdOut <- pca(mdcOut, method="svd", nPcs=5, center=FALSE) resRobSvd <- pca(mdcOut, method="robustPca", nPcs=5, center=FALSE) ################################################### ### code chunk number 4: outliers.Rnw:77-80 ################################################### mdcNa <- mdc mdcNa[cond] <- NA resPPCA <- pca(mdcNa, method="ppca", nPcs=5, center=FALSE) ################################################### ### code chunk number 5: outliers.Rnw:88-97 ################################################### par(mfrow=c(2,2)) plot(loadings(resSvd)[,1], loadings(resSvdOut)[,1], xlab="Loading 1 SVD", ylab="Loading 1 SVD with outliers") plot(loadings(resSvd)[,1], loadings(resRobSvd)[,1], xlab="Loading 1 SVD", ylab="Loading 1 robustSVD with outliers") plot(loadings(resSvd)[,1], loadings(resPPCA)[,1], xlab="Loading 1 SVD", ylab="Loading 1 PPCA with outliers=NA") plot(loadings(resRobSvd)[,1], loadings(resPPCA)[,1], xlab="Loading 1 robust SVD with outliers", ylab="Loading 1 svdImpute with outliers=NA") pcaMethods/inst/doc/outliers.Rnw0000644000175200017520000001203114710217306017741 0ustar00biocbuildbiocbuild\documentclass[a4paper]{article} %\VignetteIndexEntry{Data with outliers} \usepackage{hyperref} \title{Handling of data containing outliers} \author{Wolfram Stacklies and Henning Redestig\\ CAS-MPG Partner Institute for Computational Biology (PICB)\\ Shanghai, P.R. China \\ and\\ Max Planck Institute for Molecular Plant Physiology\\ Potsdam, Germany\\ \url{http://bioinformatics.mpimp-golm.mpg.de/} } \date{\today} \begin{document} \setkeys{Gin}{width=1.0\textwidth} @ \maketitle \section{PCA robust to outliers} Away from often showing missing values, Microarray or Metabolite data are often corrupted with extreme values (outliers). Standard SVD is highly susceptible to outliers. In the extreme case, an individual data point, if sufficiently outlying, can draw even the leading principal component toward itself. This problem can be addressed by using a robust analysis method. Hereto we provide \texttt{robustSvd}, a singular value decomposition robust to outliers. \texttt{robustPca} is a PCA implementation that resembles the original \texttt{R} \texttt{prcomp} method, with the difference that it uses \texttt{robustSvd} instead of the standard \texttt{svd} function.\\ Robust SVD and its application to microarray data were proposed in \cite{hawkins01} and \cite{liu03}. The algorithm is based on the idea to use a sequential estimation of the eigenvalues and left and right eigenvectors that ignores missing values and is resistant to outliers. \\ The \texttt{robustSvd} script included here was contributed by Kevin Wright. Thanks a lot to him! \section{Outliers and missing value imputation} The problem of outliers is similar to the missing data problem in the sense that extreme values provide no or wrong information. They are generally artifacts of the experiment and provide no information about the underlying biological processes. \\ Most of the PCA methods coming with the package were not designed to be robust to outliers in the sense that they will converge to the standard PCA solution on a complete data set. Yet, an applicable solution is to remove obvious outliers from the data first (by setting them NA) and to then estimate the PCA solution on the incomplete data. This is likely to produce accurate results if the number of missing data does not exceed a certain amount, less than 10\% should be a good number. The following example illustrates the effect of outliers and the use of robust methods. First, we attach the complete metabolite data set and create 5\% outliers. We mean center the data before we create outliers because these large artificial outliers will strongly shift the original means. This would not allow for objective comparison between the differnt results obtained, e.g. when doing scatterplots. <>= library(pcaMethods) @ <<>>= data(metaboliteDataComplete) mdc <- scale(metaboliteDataComplete, center=TRUE, scale=FALSE) cond <- runif(length(mdc)) < 0.05 mdcOut <- mdc mdcOut[cond] <- 10 @ Then we calculate a PCA solution using standard SVD and robust SVD. <>= resSvd <- pca(mdc, method="svd", nPcs=5, center=FALSE) resSvdOut <- pca(mdcOut, method="svd", nPcs=5, center=FALSE) resRobSvd <- pca(mdcOut, method="robustPca", nPcs=5, center=FALSE) @ Now we use \texttt{PPCA} to estimate the PCA solution, but set the outliers NA before. <>= mdcNa <- mdc mdcNa[cond] <- NA resPPCA <- pca(mdcNa, method="ppca", nPcs=5, center=FALSE) @ To check the robustness to outliers we can just do a scatterplot comparing the results to the optimal PCA solution for the complete data set (which is \texttt{resSvd}). In Figure \ref{fig:svdPlot} we plot the estimated and original loadings against each other. \begin{figure}[!ht] \centering <>= par(mfrow=c(2,2)) plot(loadings(resSvd)[,1], loadings(resSvdOut)[,1], xlab="Loading 1 SVD", ylab="Loading 1 SVD with outliers") plot(loadings(resSvd)[,1], loadings(resRobSvd)[,1], xlab="Loading 1 SVD", ylab="Loading 1 robustSVD with outliers") plot(loadings(resSvd)[,1], loadings(resPPCA)[,1], xlab="Loading 1 SVD", ylab="Loading 1 PPCA with outliers=NA") plot(loadings(resRobSvd)[,1], loadings(resPPCA)[,1], xlab="Loading 1 robust SVD with outliers", ylab="Loading 1 svdImpute with outliers=NA") @ \caption{Figures show (from left to right): \newline Original PCA solution vs. solution on data with outliers; \newline Original PCA solution vs. robust PCA solution on data with outliers; \newline Original PCA solution vs. PPCA solution on data where outliers=NA; \newline Robust PCA solution vs. PPCA solution on data with outliers / outliers=NA. \label{fig:svdPlot} } \end{figure} \begin{thebibliography}{2006} \bibitem{hawkins01} Hawkins, D.M., Liu, L. and Young, S.S. {\sl Robust Singular Value Decomposition.} National Institute of Statistical Sciences, 2001, Tech Report 122. \bibitem{liu03} Liu, L., Hawkins, D.M., Ghosh, S. and Young, S.S. {\sl Robust singular value decomposition analysis of microarray data.} PNAS, 2003;100:13167--13172. \end{thebibliography} \end{document} pcaMethods/inst/doc/outliers.pdf0000644000175200017520000056141414710312276017764 0ustar00biocbuildbiocbuild%PDF-1.5 %ÐÔÅØ 16 0 obj << /Length 2093 /Filter /FlateDecode >> stream xÚXKÜ6¾ûWôPiYÔ[{³l’ &x,8»›°$öŠÒŒçߧ^R?VÆæÒ"‹¬‹U_ûýã›·ÿLÊ*ÂXeéæq¿Qy–Q±Éó4Œ+µyl6Ÿ‚Ÿtß´¶?lw‰J·ço£GÍ£ÚõÛ8Fmûó®il­üöóã¿@IµQ*¬²,F%q&Y¾ÙÅ «,YÇïÛ2\»tâ"xu½Ëà ˆñL3p?™^4õ£iŒí5mvYVEµÙ©$¬R‘ýáÝÃîîþGæ½ß¢¤aìÍÀü?÷À=N£áõ½xðÁu§iÔ£u½nyë{ëZwxåõ?¢,ºÿùÃ{ø*Ò­¢ üx­úá¨ûÃQÛïfå¥ Â¡h8Ú^kš…e’ kÁ¬xZ:’Šó0«ŠkÁwú«ˆlu_㙾ðüê8Åù8w®5õÔêá̆\£L8yõr@Ô[æa~{rž}£;9ÑfèXÒ+_uzuÕ»" K­+)Çq<ýãíÛ'ël¿Uhbn®}ØlwÚmw© ®í`~óv-†Š*Ì’bb©HYîoõ螶;°ÅÈ)ãê;vCÅ)‹ÉÔF¥a’ælŸ‚ T 3Ã*ÊXÚ¢àþÃ;`Ïó`pO“GWåà1Ç´›/6yyœ£ÔìRa©*÷n[&Á ³ÆˆŸ4Ë‚ýà:¹ýhzú£Ã-/ÝHè¬÷Ëäy ?ºŒ‡c¥eÜÙzpz.åâ…ã÷ÎŒ½QBZY $rÒâH†®le·²­bK¥ÖÃ0FƒIe`ÓxäóuLg˜ 6ålÏ1-f×`j@¬§I†éÜ7zhXÂÿ¿ç®£=ÛW¦ùÉ×Â`źÑ>µxàÀkÀï¬ ôäyá/ËGÙ·›FyPkoÐuqéÅë¶oì³m&Ìqœ‹‡`tbçAŒÂg6»çE?ý%ym gQûºêL°íny3p(éL¢ á {á™yƉ‘56­Ñ ßK< %Ì–çicj<_€¡·X!xm95Œé4YrŸ+æøDe~mï}­WìÝÅàqZ‰ræ@§¦Qhþ^á`nSbS³Hzä\%\7\!\»Ÿ Ü`‹ÞŠÅIæÉbñGf8 +^NÃ"KæÝák/ƒ†²©dL¡"—Êeu“îhQ’€Ó-d[1Ò×DIå$¸få;yãW,V€öi”þ•˜È"0a‰ …R…SÓCjÅ"?ãÙŠÆ4¦‹F¿ªKUa}–ìÙO}·ÎÎ(Ã8)®ñQ‚+ÆOPK”"Ž^(§SkëåÎÊ*ÉM­ˆ’ìRÆ p0Ó!í_(ß‘dEÜ'õù¬z%3?ÅŸŠ‹ °¥©’ç!5;žRÐÂ÷I³p’­j?ÜÓ-$Ç_¸^‘Æoþ; öY‚E aØÍGG±û³ØS=°„s]Å°ŒÄ«± Þ<R@n¥qÂé… D-è@"¹¼öлe«‹‚{®ƒh‡z„Q¶Ö Ëbü ðS\:‰ÓK´Yø-«˜ QÅ1?â\}=ØjW9ÄFÝN U˜)ŒpDH®=Oè•öiç½Kñ/æƒ ‡¿cÇÄžj!*0_÷_Ø!»" 3_»E³ Z'€À>I¡·èþ¶ÖûwhúÒ˜Ùciú~›û:êòøÍmßroH}Þæj® °f——Âÿo)+°å9—\H?—‹næº;ò¶³\"¹ê¹ÃA[†Qs`‚¼Ì`Ä:pIôƒ/™½é½™i°ki²´{eqÝîÁ4Pôc=&BïøKïŽb`pä5 Ù^Z}º]D¼òš¦˜ýNfÐm»ÐG»×õè×›UÜ$°ó•aË v®~¼À×XF·öFb¯º4NX¤y†”œ¨e¹Ñ5õ¨½ãù½zÛo4X$®Æê&=eË#x—'vªâ)ì@¹(¡ž— ò2œÀ*·èHÎ88iyâµDÜØÖމZxGT1&KÝ€å¹ \¢¹$¥ü,JÓò"‚‘JUWÉX–xc$G ¤½Ø¶åÁÆ™z0×ZD^q®Ã4C?­ŸwíÄCÞò— ~¸ÛRŠ¨ú&|+á§rxC€üŸm ï| ÄÔì\ã)»”mÄr¡’KRÑ|œC>¹sÑ`š{âÀt“ÿÖ[‚ýCZðãGä ˆ¹0‚^)ð^x’EOü>§9œe¤€fóë;úƒÆ”,´ì–m=¤ÂšÝ·~ÏþÇïeÉ~/«™«Ä²áºÙá0Ç„üh¡w o’ok¿P(´¯"Å1}N¬fªEŒ®ëiгP(–S;zfÂ׿ÑPš:ôÖù„@ÉßÅk'³Ãó áÇŒñ<¥„ÂækmLÃDÍ¤Ú ø?Otç&A)…o¸ÊjÇ!~xäˆìUÑßy@ËG7µ«í×ùÉP*âÁ ’‘Ÿ˜¶rT%•Wa’g¸†0ÍÔü‡,½ùáñÍŸ{eÍ endstream endobj 32 0 obj << /Length 1527 /Filter /FlateDecode >> stream xÚ•W[oÛ6~ï¯  X)ê¶5²¦ÙiÄ^‡!í#ѶZE2$*YþýÎá!cÉq/{Ò!yîßÇ‹~Y¾xu¥Gy'"9Z®Žx”©LŽÒ0 ¸È–åÑ [nôlñ”­Úºng"cU³¦)ý¯ºÛÖv=cU]½é”Ñ=­o©?„Q¤ C£v徃©+Ý9eÕ”{VC¯§]{ f—͹ r)á˃<Ž)Ó;m6íl–}0›gqÎΫ®7dzy¦ìs—2åL£Š™HÙ†Æ6¨L+Z,ȸxT·àr¨+?Y*£Hêµqî0ykÞiåõâlªá4I_5$(£œý=Ë"fëÌ!šj¨•…n0Y£;Zq=É\l”nm¡zÕvnéËq²Oåq›sFÕÀg…rÎпkÕ­õ$ñ„W©GYTª†Üñg=þ$ Úf]?â(eý¦ZZ .£eW­«ÆûÁº±I’Û*ï;ÐuIæuN”gá¡îB;°Œ„µPh³OÀºêþ©3‰ÅWuUß64vý0¨BÑnÅõ<à¸`·v¿à¼638¯¡Ò-ŸîX½tnFWÖ!ÎÓ4Hcéñl—^â¥ËG¨ ¿¾O/•ß¦<”;ÓËÓ¯³àŠ6—‡8‰p¶Gþíö‹èdÿþÃË¿fFXâËAJàYaŸ*º@R|¦)³Á6kÝcÇ“ˆÀ…ïèî­'¡GûÔ•Ot<€Tº`Š†£›É'à;à±BÑÝlÓ7™Ã®|™ínL`~R¦»Õv‹,¬i`›ÂîÔѽÛ'V£Çšôo#kgße b›ÊömCSð¼x¾öˆàξç8E<ˆÒØo!3 ˜·ÝäðÞ\öA–„ŒÓÇ?;qÙ73t;¿ÿJÒ ƒæwO$Õ­*áÌïi¤ÖðÞ Ø€Ê•g1pNŸœ¬ó;Ž#Ù"™Ww$4AŠ×z¥;ݺÿGíU–q÷kpÃñþ Cö›²ÇÕgÈ Ïˆ8dgÁ»À‰Õà%èV$ý[„f™’6È¢5í*Ô[à8Xø”å8‹< 2ñÁµ¿ÞÀÙW(>eÂÞãöQõ iñLY控iDÑ´G[{6µ}…| ”/Eîâ^*Ô³„M$  šÁwÝŠ¦´`¥ðz‹¢²Ý…2ež0†Ü‰v³ë'ž‚îµv)uîÇ%0Éš ¸šâ@æñô¹B †ß‹À SÄpÆ!†â¯¿Õ{6nzÏÚÜÁ'ó áÃ3¦Y?)ÄLÄY ä>dè ÷Ù”˜Êxz€ÜÏâÄáë¬D'`Ѽ0{üw’[pö®*ºVu-öÑy€câÀ‚‡A¼ÛÜW—§‹c²œ¢ŸxþÈá)MI?.eg4E~V.<Ê‚©-€ß=Z¸ôâíòÅ„’ 9 endstream endobj 40 0 obj << /Length 334 /Filter /FlateDecode >> stream xÚ¥’=OÃ0†÷ü É×ßA ¥P ((0š6‘ÒZ$ýûØ=Z(By°ýÞ{w/¡ˆ¸Eeg„!–1,ƽ¬£×K­ÜÙ;öŽ» yA]­)º0Ñ[C(j¦{EÏóh4ãc¤±VL¡|ùÙY,%Gù=ijzÕ·e’rÎcz’¤BÈ u v•IXoýMÄD’e›ÐجAhÊ¥£5°·õªòÖy]ŧüÚA¦”b-%ô¼u–zS4àŸO'Pª3Mok³ù­ÃžæPñ^¶€Ó¶¶œLo›ºl»Óÿ7mÍsßY°ÌÝs§IJãI˜ÈAÊW ñ+ú ÐüÑyT%|PñÑÿìfrt&÷Ã+‰$Û¯ïdàjøfvàæTÀ%ŒÀ‡¤ŽEÇ…w\TiÌ•ô2ÃL Àã>]æÑ;˜bÖò endstream endobj 29 0 obj << /Type /XObject /Subtype /Form /FormType 1 /PTEX.FileName (/home/biocbuild/bbs-3.20-bioc/tmpdir/RtmpCCJHlc/Rbuild86c017e3fbf7/pcaMethods/vignettes/outliers-005.pdf) /PTEX.PageNumber 1 /PTEX.InfoDict 42 0 R /BBox [0 0 576 576] /Resources << /ProcSet [ /PDF /Text ] /Font << /F2 43 0 R>> /ExtGState << >>/ColorSpace << /sRGB 44 0 R >>>> /Length 13233 /Filter /FlateDecode >> stream xœµ}K³æÈqݾÅ·”ºB½ðXh!Ñ0ÃVHž k¡ÐÂnÒ¼¤Å!­¿ïÊóHzäLÏ,ºod&Nf~@¡ªP8•(¯Ÿ¿Êëׯûð¯{õëmÛ^­Ÿo×þª¥½]Û«çÛÖ_¿ÿåëŸ^¿ýð×_ý÷ÿüw¯Ÿ}ña›Gn¯õÿ/~ö÷S{Œ×¿øçym¯_|(¯ŸÏ¿þPâ€×ûðzû[o¯~\ou¼ÞoŹ½õóuo½O¡¼õë5ÃqèéH1W1N½Æ¡VÌÌ«qÓùV야2¦¬J¸5×¾ˆƒç‰™èyšŽÂô©3^G s åí¯™f-SE YÛ[ÆI W*ʼnG\ŸyöJÁ'å´Âé÷ÈUéooeŸª-<¾/Š'å(o%õmŸèu³€@ímïiíoûHœxÅ¡T´+r®ou³W 8”1ee6­¹2ýÑ#v?Gü°÷EqzŒ·ýšÂ豿]ÍB'ÜÉz!ªpà‡JÑߎž¸öv´ô ‡*&­Ê†¸5W¦¿o¥¾ÆÖßÎé[±¿•òÚg écƒn¿ÞÚná#/Ô¸¬˜'mߌ³¯:¥Pà¼WC°Wq¨cÒªlÚ*È«ϼv£ 7x_ór¡ñôybëöVÑx¦ j<ç‘V85N¼ªñ@±¿Å¸2ÞÎÍ^)è—FLY• qk®yö¯ )LËû¢(:5³‡µê¤ÕfÁŽ´V eà‡RQІ„+hCòJgŸ1©P6Ä=reúeCŸMmÛ‘jú[›½ôvávŸ ®•W)%š¶¤VJeG+{cOK¬%øæÑÍ=ñÙnìövÖÛ7$íØ´;/byë·”­«·=î‘÷UsD¿]æí2z…+âíÑl%1»#šxÚOJÂJÚÝq¥fD§šØ?0}w÷][vå%ìš·¯ËõvÎ.`¯‘Ëûªéq]ËûbôÙ¥ ï”|¦ÛqÛç9o¬¤zÿriJô;‰Ý"Ÿô½Ý¿\±iw^Ä>òæo¹ztÑD6ŽnVì1^'.:Í)ìqq%D¬ëŒëkëlÙGâ$ ¸VÌNóL\};Žô ‡*&­Ê†¸5×ûRì 'oßòRì<ÝhƆ¢}FéhÆçaÉ—bv¨içÅ6V|\šìŽkØí)}Cò¥@lÚ±¼}‹Tü¾Ù4ÝîÖ 5GÏíÌÎXKðÍ£¥i¼…E–¾!±ëRlÙ•±¼u‹ÌöÆ WØĬ˜÷e‹ö»uœ˜Ö_×ã›Ý"£¤õ¢@œxÕ-š‘q³wÝÓ+ª˜´*âÖ\u)úˆÛ½Ï)ÎP³’fâÙ+ôR‰!µtN~(átÍÙd»Ò>0@+‰¾q´5ç[«‰%ú¦Dߌm»2=’|ó·t¨{ñ¾jæ5ÀÔŽñubÂ{?-1Þ3ã´ŸoG½±’ö;;iæI½nlÛÎÛ7¤:«ˆ-»òvÍ[¿e7'ÇLß"ÖT4ó9ì3!qI%!Þ [»íW4±ÄJ*y‹Xƒ¹µ±óI"nù¦Ä£Ûvå%ìš·~K=1ýŸÇœ(ZOe³»«³O"f>¹”¶a¢Äx«í £š°’è›GK3¯â–Øù4‰¾)áhÅ–Ý™žI¾¿P{ s‹¡ßbÍ£P´ðÑñÈ©Ä8%‰w@‹9^Ú{ŒV‰•ße;¶ìÊ‹ØGÞú-Z›¦S,Ö\qËÑb‰£Ï°ór—XXë–ov‘Ñ×ؾs,VÒ‘³7kvÌx„=‘}Ÿ÷/WlÛ•—°kÞ_æuŽµÐ]ýûª™Ol“·Xí1•˜°Kˆ7Ÿ¢MÛÞbân¬%øæÑÒŒ~Ænoõº}Câø¢Ø´;/by«OŽGœÎ¹(û±Ôœá¨Î‰]´ðp;{Â9±;NKè7 ®uÚÑ*+éðƒAj° —Ø]sú~6Èز+/a×¼ý[¬ÞÌ^T÷~j..óŸXëi˜aÕraœâU®"È^ –Ö…µtúÞOͱTØŠ ²}Sò/ض+/by{¬1AÀºù¦±R,~Ô9è­föãuº×°ÄñìŠéŠísø/Wb-m1¬éHÛØù °ïé›}3¶íÌËØ5oÏajô½àÆ{_5¸1Ëì!í”:F/Jì™F8Û÷ +©{ 35¸©‹Û8}W¯bflÙ•—°kÞnc;æDsf»µ1i®XL‰kÙøŽ f%ñÈuY♾âÁÜöÙjbÉAXKðíyrvÛŽÄ6,åÙ7%ß‹ÈDvå%ìš·~KÅË£vž^§·æÂT-|öñs`˜µ¶xŽ•ÄìzLÒ>â&–}óhiŽ8›ÆÎ^vN»ì›RÞ]õ¶+Sa×¼ÝÆ®hÚíÄÎûªÁf¯¶ƒ3ÛÃ’g¾ûvÛ[œ¦ÄJ‚o-M Pb±"”¾kö‘Ž Mæì3o]—Î%´˜‹ò95{„­ñ6jx>Zçd(F·eæ;Ÿ»ï¹píX½J¬$øæÑÒtôÆâa3}CâÑŠ-»òvÍÛ¿¥Eclsˆ-jcÖô¸Ç"^œ­9Dwæg’ãa¥4íGÜ(‰•ßOˆ-»ò*I¾ñ[æxœýŠqû}ÑLÌ|€­­Sªx^š·!žÐ«Ÿ—jãýiû‰yc)Ñ7–æŒÑÚØýˆÉ³}SâÑŒm»2vÍÛ÷>×f·_º÷¥iñê¯òÇ|RG.”|7Ç‚µí;¥öà›GKSÞú~c·x×¾!ñhÅ–]y »æíëRÐÇÏÑMsËÔô˜,ÕV1"ÌyÄœHÅÙ¸NKòvŸ|áeþœ¿åø" ^¢ÔÙX m6FŒàq4Kˆ7W´eïdS+éÌñÅš#†°}§Dß”x4cË‡t.ãËàý5/ðÁõ±Ô”xµñŽ˜÷`‰1r‰û“ãí¸m?po+ ¾y45³ñD/+lÅk û¦äó±mg^Æ®yû·lÔíž'§‹iq¶Îáfwð[îÖÌ –Û>˜½°’vÏ“S30–Ûc’–¾»çÉ[vå%ìš÷ý[â©5¦¬gþ–ÐÌæ[تb­“ÝyÄ«—%Ÿ»X%²½cUÕXJôí_ ¶-œ-Ê7%ÿ–ˆm»2vÍÛ÷ –5:9cï‹f€é3‹Æ^½rÖ1vK9+ÙnûIIXJôÍ£¥9b´6v:íľ)yöÖÚmW¦Â®yß÷K犖Ÿ÷­Ù1& 2².¼-Œów%·šx?e;yBÆJÊ©Áâkb%ùÎw[våÕÒXž÷ùÃÐkïÙmì¹÷‹×²`Ä86\õÚ-}ô/?n{‹¡ ±’à;³ãˆ×ÍØ7Qún^ÍØ´;/byû·à=Z¬3µ¡ß" ×R:(±â½z°àNKn5ñÜmû…'tc%Ý#¹5\K1vP’oHžƒ#¶ìÊkHòýE°+?;y€o$jæ;"ÌŸ$ù7¯×Ÿa}·¥ÞG¤¼2ð$|bù>$žëʶdby9ä‘Ü’ì}Èß}©Óññ«o8_}ü퇿þOõU^_þïÏÀýåëL­‚¾ùåûë/Bû—¯/ýá?~ ïß /ÛÀ“ö?¾¾c8yàË÷‰?§êñ”õÀø1=÷'¾f|ñxÉd|­7Þ?ϧÿÓZ¬K\¸ø‚ÍP\è|ó×È¢Í(.(+°&¹ž…å§ÿÕrFæÀ»Ï³Ðñò$ÎÂ_ÍÓð§ñ»Ì›x6s;øÆëøâYö–rðÍò»ì 5ÙÁ÷Ï XœsF#wS~Pºß?Ô |¥üËc¶£yØo|Q,þë ]Õx»ñéñÂ?û~.sN3‡å¶a;~ÆýÝÿüů~û&è‹ÿñ¾Çù(˜<¶“‚¯9zýû¯þð¯¯ßýñ¿ùÕ/ÿ•ÝÆolíø‰Ùï-ž«Å¡MÍoøZŒ}|\¬ñŒ¹ÇÚ±¤8úÀųýDh¬¤Ã/‡R3âmHb{ÌœÒw÷|Æ–]y »æÍ¡9<\jÓo‘8⸠]?¢XB¼ Ûƒ*;k ¾y´45A‹’ô )ÏÓyÛ•W}HòßB–7Úùè—<¾©cŒ°1î 5Ib<Žµ¶×˜3k)ýRƒ—ÍÆ’?gß{>ú9¶ì΋ØGÞú-dè/¯GSƒWœmö³{¾þl+UËëÑèKË~ÛD7VR¾$L æÔ‰ÅëÏô¯G3¶ìÊKØ5o_NPö\ÂJ –¡"­ÂV ¤CIucÞa».HH¹„•,C%KTé;—°2¶ìÊKØ5ﻩÕh)>5à·Dô^°´ºp»wKncçqÛ±m¬¥ÍKñÖìù“NvúI{·ä_ÞKÚ•—Ûþš÷}].ž×sÏë Øcq¶¢'ýj³”ñŽ´G›?ki˜à’š/=Ûb?}7/óglÙ•—°kÞú-ÁiØù…K‹©9ùžç@1!¢1_Wß%E¼X­Š¾‘šXÉŠž‰Ø”/ù¦fÇs›± —§ïá%ߌ-»òÚÒq/ùöËÀÓíÁå…Ô`q½o\Bžâ}óÆ%fJÌŽKжsyÚXI›ÉmÒDÚñ>šØ˜"Æhú–Ä£;íÈ+±kÞ¾.GâÅÙ¥ûÅšÊ7n'Ƴx,>…wH>Ó1Á¶½‡HØ”¶û*R3¯â±%v^Åx—oJ¼.ŒMó"ö™·~ ßqŃ×Åš L½õŠW†gõ[/IˆÇ7niÇÛ¸ÄR¢o-͉5NaOPìûLòcÛ®L…]ó¾¯ËNÚˆ’š3V”âÜDï21ã<,ùºÄMOÛ±)^ NÍÀ;^c;΃}÷•c ÉpDÉ÷Ë(·ý $¬¤’³k6Þ6Ä%ú¦Ä£Ûvåµ=$ùF‹—_K§U}²4uÇo>¾A*Xm˜AIB;ˆGêë¶c4N,¥’O©Á1¶`Þnß%ŸÛve*ìš·KÑ6^skn™š–ÒÄT¾ê>.®ŠŸ–4ÛI˜1VR^ÅÔàÁ0±8·é;¯bÆ–]y »æ­ßRIvýiº.ÒÌçµÊD‚ª»ÝÈ ˆ—_”/X°í¶£ûL,%úæÑÒ|÷AìFNŽ|SâÑŒm»2vÍ[¿¥ì l ²¿¯š‹<Ÿk …o$æ %ÆÛ0þÊïZb-&O§†[Q„¹‚Dß”p´bÛ®¼ˆ}äí6vð¥Æá=Ö©9ƒ‚-|ð…ÑE~QÐâ)!^sÀöÎÉ‹°–ÏSZbV±í{ø™5cË®¼ˆ}ä­ßÒ7¼Ri›Ÿ÷SSÌè…Tã‹\bH¤Äìiü²wþ…•”dPkªÞ”´ÌçJß5ŸóÛvæeìš·~ËÉm Ü3ü¾j°­8V¾AóØ°'ñÄÔ^â|1,ûÖ@ܸÌr€tú©75Éh…$Æ*}S¢oÆÞ“Ñê}×'yû·œ ®Çz†‹4XàŒx;÷ÓUæzî–˜]å«pÙ›_ØÖ|óè–ùn7¶ã¼ÛwÏërænÀãºó"ö‘·~Ë ÔõqzKWj.`‚{ÀMZ!ÍÓ4Kˆ‹$ûm¯,hp=¤Ü4•šƒ{†ˆØÊdß#7M9¶ì΋ØGÞú-ÓÃFêÇ®ßbMãNžbŠðÅìâô¤ø÷ùp<Êmï2•TMØLMaa· T¥ïÍ”«Œ-»òvÍÛ¿êñÛµ185³%DßH.ÜëÒŠ%ÆãÛ’~…•´{cpjXÀX!°ïîÁ[vå%ìš·Ë %ù4 5WòÖu©'7œ]¦EI3b*êÞ\Ù!Î6Û­OK8w±»ÜöÆ¢BÄR’o- ¦÷¢ Øì[ŽflÛ)±¼õ[æc“ª©\@j°*ñÚæ9磑‹Ê*iæ;VNÓ>‚ŠšXIÕ3ßÔð0v‹iWúÞ\. cË®¼„]óöu)<—Û˜4áád¥"]‰øS;Ï{q÷gÚwF–%qè›GKs²ÄïíyÞ£Ä}KâÑÕW ve*ìš·~KÙL°÷Xi j àŽ$}¥ *s9V–FÚ¼ìÝüs¿¥–c¥5¸é[0ÚwɱұeßL²vÍÛ¿K&ƒŒ¨÷UÓ¢%FE§¨ Æ.0ÎÆuZâ¹k(Of{g—#¬¤zßÒèº5‚Ø.ß’x4wdw^Ä>ò¾ÛØÁ;Ý¿Åš=w\çc°]vDÇo©y7×Áß"ûÎ\…Ýyþn¸NþnˆW)»}Kr;Æm/üe®yûºìxæž×Êã‹5cÄÁ¾fpü8±VG飮bÌeeÅ×3±–ZŽ/ÖTŽ/ÂŽ/òëˆ[våU]#äÎÛc%·¼\à6¼¯šæJ›R0B]¾ªïnž“½`G„±–’I‘pjŒ7]#}Sò,¡í·]yûÈÛ}òÁÓ¿y[Gjð¦kĆ3ºÏ†W¬’¯c´¶½×xÕg¬¥Ío¹­ÙXOØ oºì›ûdƶýðD$°¼]œ÷véùüb v%E¼8)Ãì R³Õ<í#›C½¥žÏ/Ö4<ƒ bIú®ùüâز+/a×¼}]Nê0¯~_4Ä›8[;°N¯æ,>2¨å¶³q +©ä,>5 í[@ӱ[vgJì#ïû·¨”û1kد¶Ë¤BêØ!.ÉçnÔÛŽ]âÆZºûdkد »±Ï•ïíî“[vçEì#oÿL‚#¹ýi°3‹ÆH'g1U¡”³’í¶ï”„•4²ÕXƒm…‰m”ä’go­ÝvåÕ’|û~‰ÆQNs{¬©XL‹s×ÙÆ6žéý²äVsÕÛ®Qb)Ñ·ï®K¥¡ŽÄPí›’¯y·]™ »æ}÷cÏdÙm<Óírc©©~᪫0U÷ݼ¹PdØ#ú™XK»Kí¥‹‰íhßÝko[öäÛP>òöoÁ{´¥Pfj@T‹+_‡‹aF«ˆ\îZ™Ñj¢4–íÊf+)ËeZú—Ʋ$¦}ß3Ûvæeìš÷B¹!laÔGˆO8÷Ÿ’òÍtñeaݧ¼Ò÷ë>ååk{°îS¾y¦·¤û#²î[¿>‹väˆÏ¡ÝÇrâçÐîãÁýsh÷1ývÚ½Ä;˜ÑïfÇIÕøú!”Êë!A Ç}ˆäÇ!u–òÊCŽ‚÷¨yåõ¥—‡¬éþ9ôóŠ·`âßÙ$þ„ƒ9ç=¿»M|·Ô )ßÙ(þ„ƒMßÕ*þ„ƒØñ5¾µY˜ƒ?X$9ø’“tïëò) ÿù÷?2 ?8? ¿^xëLÔíâŸW¿ÿÝÿúËWÝ^ñǯþðuRþ¿¼¾ü¹YùÜ°o?mEúÊj®ï‹Ë¬_Y&˜µã%ÄxÄÊòR°æ¼pºwnZÅã°b¯¹:㘴*›º òzW¤{¯KEú8SsªÅê𘭓µã% §Ê@´b—¡pªß~ZQbuÞ¸-çíuó 2Ç„ÂÙ÷Èu­H.ûR‘>vso—ªÃ£¸kÇK@ Æ²'´¢æ¼qJN´­8c­F¸‚: òZ²Lˆbʪôˆ[s]+ÒÇ{Ï•¤@¾Pú!`ϧLf0‰±sã$d9+šqÈÆ^·œ")&­Ê†¸5×µ"}œA=~[ÑâyšŒ‰(D|%|üàŠôR°æ¼p²Ò‹XÈ4ü{ÝüÔ혴*๮équ²ñ@Û‹ÕáUŠµã%¨ñÄþoYQ%Ä8 YØÉ ”£nÞò½Ú+ýÒˆ)ëæ¦ÕÛ3×µ"=~ë±T¤W%V‡ß»ÕŽ—à@GZ+à,$ñFŠ˜%wã N³¼–dÝ(&ʆ¸G®ŠôQ¾O/ÒSsF{SUø(’BY5^RSUù´³æ²°–v¿HO X€‰Å®»ôݽ`—±iw^Ä>ò~T¤GYŲV¤GéÅîªðá¡W—ÄìPU>í¬³m¬¤–B­)¨9h솺wö½e]<Ç–]y »æý¨H¹äu‘æà¹ÝX“•[Y5^’ÏtT%·½³ø¦°’Æ}]º‹såÆ6^ùÎu›vçEì#ïµ"}T+;ûR‘>ʨn‡ªÃG½ÕíTíx ‹•åm[×8 Ãåa¬À»ãðjÀ^› «;&­Ê¦¯‚¼.éãŠ3—šÁ:ñ؉M5äñº_’/E)·][Øl º{SSq-”ä»x-(cÓî¼êCê7gN•ÝUWõ}ՠ«&«ªÆKb<ìCH;®Db%Õ,!j ªo&Õ7Ó÷–%D[vå%ìš÷£"=ªuúºHƒŠžª Õ>ëáªñ’|vVœ7ÖÒiþ_j°œfì ´}ŸÉ‘vlÛ•±¼×ŠôhÙ}©Hí¾žªåÔBéV‚n‘QÒzQ8W!©âV`?¨qØ*j¯¹“Ô1iU6Ä­¹>*ÒGI謰- _ª*|Q}VV—„ÓŪò¶³â¼±–ú]a[šæ*ÙÀVVЖ[vå%ìš÷£"}‡_3¥ïT>Š‘BoXã,-ûÉò×çC:üš)5;B‹÷é{d5bÇ–]y »æý¨H_bÛé¹V¤/s*0'7ª _‚’’Uã%!«Ê§oâ+iÏ1Çš ŸÄ¶ ÒwË1DZeW^®y?*Ò—cËnØìóRUø(j«ÆKb˜±iw^Ä>ò~T¤˜ª‡¥¦¢dUøüÃû+E¬*Ÿö=£[º¿R$MG-c;jÁØ7¥õ!åÆÔ€übìv±â(}o×]‘”±mW^®y?*Ò—zxij@¬ö,!˜s" çN3Û»?aQî¹ }óhivôÙÂV<“Û7%Íض+/a×¼éq|Y+Ò&>YĪð9‡«ÆKbÒ±eW^ý!»Ê®*»ÇñÝUÜ¥Á:…ªÂç¹AÕøǹÃVœ´cqb%5¿NMe-~a kñËwñ éŒ-»òvÍûQ‘>Žï×Z‘>0GuUøÀ-ç‰5¿"¨ªò¶_jÄZÊÏ`¦fó¦°_©Ì9h~ Ó±mW^ÛC’ï¥"} £¥¬écÐ!šUáKCádU—„x¬*ŸöÞ°’º ¤Ë›‰å³Š}'-:cË®¼„]ó~T¤éÍö¨HÉÐáªð1YšS^U—„x¬*ŸvTœ7ÖRÏi¾5(úØ‚i¼}—œæ;6í΋ØGÞŠôeÃÂõûª)\ÅAUø2;̨¦Çªñ’Ðo²ª|ÚÙ*Œ•´ù¥±4\ÁôÂöCy¦ÀCØVädÜšò£ýu˜€m…ÁApz`V‹§€H\¬“•uæ…“p˜zm*Q7b‘È^‡y׎)+³!nÍõQ>žÃÚ¹ÖŸ§¶¨*ÍðñT EY%qôâb—ì¬/o¬¥=¿¾b͈M?‰íKm{Iô­Ø²+/a×¼õçãI»ø:Òœ¬úÞÜ—ôâñ’Ø¡†|Ú±W5±’ŽüΡ5#ƳÄbÿqúîùCÇ–]y »æý¨?ÐÍŠ ½qõìÒ‹ 9¡À³Œúñ²²´¼psgãXW^‡{iÇ”•Ù·äú(9Ïç¥â|ª¿ï—G¥àV?Õ_? ‹ £<;bõv[QØÝ8ôÊC©Àgè…Ûñzy¥ÀCSV¥GÜšë£àû~˜GaÅ…ËÅ’ëÇæ Ó³•¹÷BÛmE-wã(äÆ +øùâö!‚^)ø¤Ì˜¶2á–\5ÞwìàXJ¼;'Ðו9‘‚Oi¤/+Ö·ƒ@¯Êºï¹¥\Šƒ7> «•ê®¹Ÿ\eÙmEÅvã ì¹™Ü ìÎUºáu-☶2=á–\•ÜáªqVà¹OµÔ<ªÔ:…œ li)Ú8 ïj¬Àó¢qxD´×æ÷4Ž)+³nÉõQ¼}?\”È |fUåÓã+/—««Sp‹¸jZÁ¦1n×ýèUM)8¨ ‡Ê ö:rIS1ee6Â-¹>êµ£¬åÚ¹qYÓJÕ)|ô/=ÒŠRìÆQØrÎCÅŽ²AÂí(/(¯Üó̘²Š€GÜšë£Dû½hK…öƒó0IâÚîêÜ"f“´Õףгr¼¨áj\aN‡~¶èSÖ¡SÜ’ëX•ìĬsNq©Én;%ö”kEvÛ%®°þz@q9@Õ×3ŠëkN™ârÀU‰ýú¼Bì×#„»ûõÃøà.Ã~ý0:¸«°_ßÄ×½Jk°ÇÌcôµþyÚñšŒÒbÆËŒ»<ºÄõ€ŽoyÅõ€“»}ÅååäÖÿìêã;žÂxýt®ãýðòé±î{ôÏ(Ÿ/YÏòõò鶃Xž;Ë×n€•¾œÕ×ÕAX¬í"Ñ[×Ãâ£øúöÓq¾Uz=*ïÿ•×ð“ž~þá~ö·O’÷ßüýß~CñõŸŽçõš±—DïÔ€±uƒ½:ÇÝ G[¹Þ±³çÚÓ¾ëk©ÄZJºwjX†ÙØBI¾“ñ±eW^õ!-¤ï(`dØ›õšÆÒâüN™Ýø2Ù¾¿c·I0rmç7ªŒ•”ÜïÔ€ÄXð»ÓwÒ¿3¶ìÊKØ5ï».öv­pkÈåFýóä}Çymme£äö~Û±IÆXI7<5`tK²·}ß\pÇ–Ý™ûÈûQ{}ყ¦q[X[᪾PÂU?=íÇB O)Y᩽;±`~§ï$†glÙëB æíëRÌ·9<5`y·˜BÀ£¤»¾_ê äsŠ†µìxµ—XIIOMáÝ×ÍÎÆÙLÞÖKÆŒ-»òvÍÛ÷ 6‰.LñÔ\ü(Bõ×ê.–¯ùíb¶TŠK;Ö +é¼ï.i~4XrÂíû¦Œ;¶íÊKØ5ï»öúv®´ñÔð x³¿Ø.SÃQ·¾®ÌqÔµ·ýâÝSüƒ±¯äqkÈ7–qû¾ùãŽm;ó2vÍÛu~ ¾Ç}SÈSx”ìmIG9ß¾RÈ£ïqÜöáï¿—ã–’Bž¼ K,(âé;)ä[vå%ìš÷]{{PÈS8j$'…µÞrÔtN 9ª?'…<¥¤§4ðÄ‚"ž¾“Bž±eW^À>óöuÁ*ôB!O ?÷pî$ÜŸ<ïÉøç}]Ð`ÓŽW®‰•4îë"M繶ñ¼Ëw»¯‹bË®¼„]ó¾kId{¿¯•Iï”ÀG±ð²ÒÈQ0¼ßö“Ìß,±)™ä©éùÝän¶xúN2yÆ–]yõ‡´ðɣ̴Z©ˆŒ©Áœ+Úåæ‹G ÖЖÚÕå¶^ya‡[«ïfk*ËÛ [(ÉwòÉ3¶ìÊ«>¤…Oç&X7Ÿ<5ÍŸ_¨É'G úŸ¼±¬ˆíɈÂZJ>yjpÛ$Ü™ô|òŒ-»ò"ö‘·~KlÙ;V>yjp+¶¨a™|ò(/ão>yœŸ+ùäq&¯äò¤”|òÔ€n,ùâö}óÉÛvåEì#oÿ4Zå©3¼‰–CÒx|°ç|pÊçޭŽ‚\‰•t£È~gÈX~ƒÈ¾ïo9¶ìÊKØ5o×Å&ëé&—§CAÔ®®I'ïäA-äò.–íäg+)Éå©A<±*¼,ßI.ÏØ´;/by»^97ÚÜäòÔ€ Žï‹_&ã»æu%—w±þl?A¬4VR’ËS³³pº°ü6ª}'¹òv-iÔÒ]Èå©aÕîè’†ÉãQ¾u¿Vry”Ž ~›ì>d¬¥$—§fguhbI€¶ï› íض+/a×¼õ[Þ‰/äòÔ€ õ§£åP¥2‚vu“Ë£šõÙn;^Ó'VR’ËSrPb%ùNryƦÝyõ‡´Ëñýú¹<5¬x¾£²ÐrnŽòé¹»úzn¯qc%%¹<5 ˆ'äñôäòŒ-»òvÍûncû¹’ËSSYüøà•Wyá“­"ÉåÑjÊHûªwÆZJryj@7–Óû¾ÉåŽm»òÚÒB.ï\XÈå©A<ʶÇÝJòx”mm`7¹¼s5Âö…dµ”äòÔ° ©± §ï$—glÙ•—°kÞ®½^ÙÇ'¹<5ƒßçhv»ãÎG±›\õ gcµµEc5ŽB’Ë­IÜ8ÐÇí5ÉåŽI«²!nÍõ>û³—¸ÉåV€$Ž2¨›èãqÒf ¼Éå(){Ø3“Ò-$¹Ü ÄÛ(Ðk’ËSVfSVa!—G%âÙfor¹ ‰G ÙqŠ>Ž²e!—GÙ­§Õ„“är+ +q·Q ×$—;&­Ê¦¬ÂB.OþõæÆCÅî=JÒÇQÓu[ÈåÑæˆk+ } '!ÉåV4V'+ öšärǤUÙ·æz—oc!—[’ø(Þ{°=äorùàN!Y¹‡H8 I.·¢òBWx¡è5ÉåŽ)+³nÉõn<íXÈåV€$Ž²Ö¦ÇeÞËB.GAí‘Vœ.á$Œ»“¢$qã@·×$—;&­Ê†¸5W§¿ksÄð·Nv•ç9ô"ú8¾TÒrùàó»¬Q‹å4NB’Ë­IܸƢÍôšärÇ”ÕÛ5ÚöÌõþŽÆ5r¹$‰Ç—,.ÓÇãsåZÈåQ;9Š‰ÉÚTLl?-Üär+@Žôqy½ÉåŠ)«Ò#nÍÕ%Á=¹ÉåV\ü@·2müNÇ®"o.ˆ_|ºÒzº´Ù•Â]Õ_ P®…#}\^or¹bÚÊl„[ruúøÔðM.—B9õ-QÒÇQ°»/äò¨ }œiŇ¡Œƒp“Ë­8ÈñN9ÑëM.WLgÌôˆ[suúxk{“Ë¥ I<®ìqˆ> €da÷­‹ý ë/önr¹`Ö Gú¸¼ÞärÅ´•é ·äêôññ&—[q°õÚ#èãqÍç`y“Ë£EŒ-­(Ad…$—[Ñù=âðNß^“\2›îï¸lOry¸:ËB.·$ñˆzš>>øEÍ›\ŽRöWZ›®là$$¹Ü jÂ>n¯I.wLZ• qk®÷Ùç>Š,äO œ7|ì™ôñ8O±±¢ÞÝ >y$kG±À£¬B’Ë¥ I\8ÒÇåõ&—+¦­C{)®ó™ë㣠7¹Ü ÎÃøYÒÇõÕƒ›\®"ØÊ}!ÂQHr¹ ‰Wü5‚k%—;¦¬ÌF¸%×­þøy&u;¾‘²••]þÉnwÐDúM—¸Ð1EË(®œøÐr@q9à‘U&ùcòËÛÀ*ä%˜Ç«ÿýüÁs|Wºþ`‚y|jó?˜`ŽÒüç·ÌÅ™5ƒ\âÊßþÚå’¹H–¹ŽXiæ>Äd¡š?“ËC¾/Ùœu¶?ƒmNŸA7§ƒÏá›ËÃgÎéá[ç&u7å\brÌ}]>國bþ²ÎA7ëüÛʃ¿¾½>øŸ_…|Ž–Wù„“þÕÿûÅyÿ¿üÃ/¿˜þáÿài endstream endobj 46 0 obj << /Alternate /DeviceRGB /N 3 /Length 2596 /Filter /FlateDecode >> stream xœ–wTSهϽ7½P’Š”ÐkhRH ½H‘.*1 JÀ"6DTpDQ‘¦2(à€£C‘±"Š…Q±ëDÔqp–Id­ß¼yïÍ›ß÷~kŸ½ÏÝgï}ÖºüƒÂLX € ¡Xáçň‹g` ðlàp³³BøF™|ØŒl™ø½º ùû*Ó?ŒÁÿŸ”¹Y"1P˜ŒçòøÙ\É8=Wœ%·Oɘ¶4MÎ0JÎ"Y‚2V“sò,[|ö™e9ó2„<ËsÎâeðäÜ'ã9¾Œ‘`çø¹2¾&cƒtI†@Æoä±|N6(’Ü.æsSdl-c’(2‚-ãyàHÉ_ðÒ/XÌÏËÅÎÌZ.$§ˆ&\S†“‹áÏÏMç‹ÅÌ07#â1Ø™YárfÏüYym²";Ø8980m-m¾(Ô]ü›’÷v–^„îDøÃöW~™ °¦eµÙú‡mi]ëP»ý‡Í`/Š²¾u}qº|^RÄâ,g+«ÜÜ\KŸk)/èïúŸC_|ÏR¾Ýïåaxó“8’t1C^7nfz¦DÄÈÎâpù 柇øþuü$¾ˆ/”ED˦L L–µ[Ȉ™B†@øŸšøÃþ¤Ù¹–‰ÚøЖX¥!@~(* {d+Ðï} ÆGùÍ‹Ñ™˜ûÏ‚þ}W¸LþÈ$ŽcGD2¸QÎìšüZ4 E@ê@èÀ¶À¸àA(ˆq`1à‚D €µ ”‚­`'¨u 4ƒ6ptcà48.Ë`ÜR0ž€)ð Ì@„…ÈR‡t CȲ…XäCP”%CBH@ë R¨ª†ê¡fè[è(tº C· Qhúz#0 ¦ÁZ°l³`O8Ž„ÁÉð28.‚·À•p|î„O×àX ?§€:¢‹0ÂFB‘x$ !«¤i@Ú¤¹ŠH‘§È[EE1PL” Ê…⢖¡V¡6£ªQP¨>ÔUÔ(j õMFk¢ÍÑÎèt,:‹.FW ›Ðè³èô8úƒ¡cŒ1ŽL&³³³ÓŽ9…ÆŒa¦±X¬:ÖëŠ År°bl1¶ {{{;Ž}ƒ#âtp¶8_\¡8áú"ãEy‹.,ÖXœ¾øøÅ%œ%Gщ1‰-‰ï9¡œÎôÒ€¥µK§¸lî.îžoo’ïÊ/çO$¹&•'=JvMÞž<™âžR‘òTÀT ž§ú§Ö¥¾N MÛŸö)=&½=—‘˜qTH¦ û2µ3ó2‡³Ì³Š³¤Ëœ—í\6% 5eCÙ‹²»Å4ÙÏÔ€ÄD²^2šã–S“ó&7:÷Hžrž0o`¹ÙòMË'ò}ó¿^ZÁ]Ñ[ [°¶`t¥çÊúUЪ¥«zWë¯.Z=¾Æo͵„µik(´.,/|¹.f]O‘VÑš¢±õ~ë[‹ŠEÅ76¸l¨ÛˆÚ(Ø8¸iMKx%K­K+Jßoæn¾ø•ÍW•_}Ú’´e°Ì¡lÏVÌVáÖëÛÜ·(W.Ï/Û²½scGÉŽ—;—ì¼PaWQ·‹°K²KZ\Ù]ePµµê}uJõHWM{­fí¦Ú×»y»¯ìñØÓV§UWZ÷n¯`ïÍz¿úΣ†Š}˜}9û6F7öÍúº¹I£©´éÃ~á~éˆ}ÍŽÍÍ-š-e­p«¤uò`ÂÁËßxÓÝÆl«o§·—‡$‡›øíõÃA‡{°Ž´}gø]mµ£¤ê\Þ9Õ•Ò%íŽë>x´·Ç¥§ã{Ëï÷Ó=Vs\åx٠‰¢ŸN柜>•uêééäÓc½Kz=s­/¼oðlÐÙóç|Ïé÷ì?yÞõü± ÎŽ^d]ìºäp©sÀ~ ãû:;‡‡º/;]îž7|âŠû•ÓW½¯ž»píÒÈü‘áëQ×oÞH¸!½É»ùèVú­ç·snÏÜYs}·äžÒ½Šûš÷~4ý±]ê =>ê=:ð`Áƒ;cܱ'?eÿô~¼è!ùaÅ„ÎDó#ÛGÇ&}'/?^øxüIÖ“™§Å?+ÿ\ûÌäÙw¿xü20;5þ\ôüÓ¯›_¨¿ØÿÒîeïtØôýW¯f^—¼Qsà-ëmÿ»˜w3¹ï±ï+?˜~èùôñOŸ~÷„óû endstream endobj 56 0 obj << /Length1 727 /Length2 12774 /Length3 0 /Length 13360 /Filter /FlateDecode >> stream xÚm¸cp&\-ÛvžØ¶mÛÎÛê ã¤cÛ¶Ù±m'tl[ß;3wæÖ­úêüÙX»öÚëì:?‰¤£ƒ›º·……‘™ &)ªÍÂÎÎ `fdƒ£ sš¸Y;:ˆ›¸yZ@s€Ð À `eff£ˆ9:y»X[Z¹¨ÍhþhÚ™˜[Û[»4í=¬Í¬ü®îŒ.î‚Œÿ©7+ ÀÂÚSRÖ‘Q”PK)j¤€@;€²»©µ@ÞÚ èà ¤X8ºìþãÌÌ­ÿÅÉ•ñß<€.nÿ³pq´(H¨‹H*)ª$ĘÔÅ&æyéÊÜ\yÿAÍL½þ5=Àí–Éÿ,ÓÿYÿ˺xýÇb„ca˜[›¹L–ÖpLÿÒMÆÁÂÀùŸ°¹»ÓSÿrý‡€ú iæ@‹ÐîvvŠ&ö@µ˜£½“»Ð àhtqˆ:Ú™$¼Ü€æ@óÿ¢Mì­í¼ÿñÿ!ãfò"–ÿ(ÈüŸµ«¤µÐ\ÙÚíÉ-Lìþ‘êßq-àîç_þSýO+À?÷jîè`çýþ3Ì¿Û1IH‰*Ê*Òýo þ–p0s4·v°¨¹ý£¨‰‹ùÿÿN+›Xÿg‡þKå¿;`ù¿¾‚‰›‹µ@™ñ_KÂü¯ó_Ëàÿ¢DE½|XY¸ l,œNV Çÿ‡®™»‹ ÐÁíß²ÿ3Ëý¯è4ƒ[_q4ã ±Ik «ð—(š«„¤››1Ö [a]î¹ÖCµ²27:¸Ä»SÇË÷D œ 2ؽø#s×BjD8šœúEa°‚÷RBVié_»ý¦RðÚax¡‰ko¦qyë–±±þS™vŠ=+÷¬VÖ¹LUŠå—qyg?$ÊNüÚ⃃Z¦îð/] ÊœUhá•©ÐËÞ:Xx~ð.Â5®‘UaEÓA¸nþß“N¬³MÖ:{Q‡¡C;®(gœzí,ÁCQu­Z¸wnšpî’<íu`ŸS&²Ò|z¸^çŸ1ð2 ‹ä„õÓŸ¦ø©ì0Š±t/¸ó1ûº;™1mU8æ½­…þz¢¬3š©äëÌQ¬Í”oúçìâ<ü«%qÙˤ®{n‰ÛUm»ðL¶Ú†mŸ›ßl·-‡’ªáÎ?0Qôî³a/L ®R¡.dV!ñ@P“ © ?¥z§n’®Ô=û'ÃH;&Ðr$-%{÷êudC‰·¼#o°ÃB÷»¢ÏðÁ<¬ÌžiFæ¿Â›DËÖ³cºX>:e®ãèŒ#1˘ßtÉ0°.¦™ô¼ý è^, õ‘*â—C1÷1LòÚYY+Ík ¬‚¬? 2I²÷! Uh%]43Z#˪hAÀÜ~süH¤‹»1 Ó¹ºÞrCÁž(‡p “bû%ÇCéEÕIvV&î¹CËÝ” _5€öçÌ2cb¯š¾‡³Š}¼¡}ïKpÄ@.x§'ûK5d±qÎ2eÛnm¼jÀª €Yùnw —Ø U¢þK˜‰(b°³£ÉÒ’=~•Ï¥ˆ2v'dIî’™­ #uè[ðéž…ñè„B+ˆ–åøR”zý¼‹?%œâM ì‘vƨ4-]´£oòû¯…jÒsž\! Uc!¬\Ƚ(;Ä„Ø_ép®FþðÓséñ€/’ÃVv§ldæ´F)ö«rÎ … $áwkþ&MÈXDuI§Ë‡f„ÜÙ=¸IÝôg6ï•©/ ñbªéâ4ë;5°1ø|gfõVxUÙjÞ‰ EÚ³¬ò¤[ŸŽò– £k~7bt?û½²-?-“Ym5ᘌHûž¶ ͯñʾÏ_ºµù/}j£ðŽl¯C¬{õl"M8 >^Ðè\:‘_ߢ?ouƒ:lÈ[ÏF³”N¼_p´ñ«I•~5\ëær{Cw‘2—iÆãz™£äÑáèz”µN:ÊÞîäɽªWщ„£Ùêÿ…^ 7&E϶“#?zºÌmõ®¼µñæ÷VŽ(í¨A—1€oˆn{¨]Ðg†uñ86‡žHüNFÔgÌwò;ûˆïÌhÍ:°“.LLj%ƒ>“½„æêðþÞ ‘Â.'›[4”;-:(o¼,FTNÄi— 8V›|šŸtûšá)JÖçž™"¶":£ïÆnö’kT3ÂnT/Â…U&7J ·×}„Fýí(ÎlAÃ?§Ù§ì’g!Ž†´sék—€”:ŽÕ,$¶>åР€‹Ú X6ò¿ž¥ßUO ñoAˆZçl§¶|¸¦Â¹/ÄE­™=fáõÓhví3¶‹þ™ârî1º¡‡§j5^'ûR»Aˆô(7âñ@`(Ñ´QpRóôZ'àøÅT(JÓ­Â>Ð9­24.’ÌÃ'fˆ¢6ÞŒÈoöz”NÜÒY²ê I³.[ãÇZæ9ƒu#ïì.]é(…»uYwIå^b¥S-Åþ׸9aj s³sK€Ÿü#JcÃ^âjÇaJ‡·õÝçŠ/C ºÅU•üÉjwµ™þ 'òQñ%ˆ£ñó(·`›¶¢ p*×Æ ‡ÀÄ]LÛxÂ/^³^%ªâ–àƒ¸8 IÞEE$àî@DM™]xÃØówHB¢Š«€ƒqy5$0¶e‡Qm„¤þŽd|IŸ”uéêé}@ê`ÖPœy›¦$×è-º¨év™,1?hÊû¡j;.-ûþ,þ/ýäk¹ÝE|I,¸PÕ#Y$­Ç´ÇbÇJ½É~Œþé¹áKcઊShâeù üï°,"·zÀº`×/Ï6‰n¤Øhì ´ ÷1Þ[òäÞyÉ7{¯^T™™ß„Ô §¼#dûk£èÇ]F›ô!ZVš~±\"ÚfÁ<óçX_GÀíÝ£€u»¸"ÁX±¦_µfÐÏÖQSÕ«µ5%©_‘;†öÉFê²Ràú}_qBÈc—RårÒNÛ!v»R!.?“Q¤L×H&UÈÜA)½–CQ WøPâ_ÙªtÅ+pÖT¼˜x‘!}—c|Y(J£0¹¬Ò‚1Ls  †½Ó¨@¼Ù¤\qãÐÖÄ…¦¾s1Ë. ¡q‰»2Ôás~ µ³GôA,+ðu^¢Ž(‰ôF òZ¦àÛQ"WŠZµ_Ãwö—Œì‰‰ÌmÑ¡ Üñ«ë‹1«ãÚ‘m]i©€ùjÐôõ<悈FpÐn^ëâÓŠ‰°\q,Л”‚r)°ôƒQGà›&åhbvDßGóÑ ˆJ&çŸ0p4eä_‡x°KE!?Ýú„ta¿… Ø ¥ÈÚ&!™¶‚0ÇÆ“³¿)ô‰Ó‘ZÌ¥ÄþVØ×Lñ¡oÆêÔžðë§I¼·8^â t˜*)÷8çíq‡IÍ{9ÓØ g–œˆÇP¸l¿L=á¾—‚÷òMÉ'XÊ…Pnɦ®Ìòõ¤” )Ì7|Æ$Ø ¤êòVh3â<'» í¤‰K ¤²Œ«—¾.å<áGh¶ð]ŸöÑ0´¶j®iÉäD5¸:ÌûTÖ÷m̬«¶»¹2ªó&H°Plä1+E0º¸É‚zçVZº‰C X”­ÁIå`ݧ³M@‹MØY]1½x³>¢3¡a曆ðpVþ)ÆÜ[·Z[À„¹þ±g:ò]å½32SéVb24”ÓEuKr;½2 ©X¯g)kj'’²£kÓ|ªÇw™Áqâõ ¦Ð¤,øm^u'YaŽrQrûvž³cå}¢^Hñ~UGų*²ÃÄÑÇ:pš KŒ(™dâKža;¿‚Ôô Mø_ç75ìj,´ZjÒ[7™ŽÄhÕ²vš¸p؈d]ì?³ ~•d S°îT!˜¾æX8,3㵬bŽn¹ØìÒ) ³3öhÕâá*EtA~\-Õƒ±ntdï²»³ê:j­p®°Á„3þø÷S:p+ó!æN¦t¢IÂB7/riÒËÎX{”.ô? ƒ¹57ŒÇàì8üÈ€¥U±î…$ ‹—ôŒcžÂ1¸¼Ab‚#Ýk?óÅyíg¢:¸¸ŒÉYÜ$ÑØ8¹‘Y@²º&MãÏL(zqL³Œ ½|vx–Vt1dã¾gT‚¾ã“Ä>·0p¸o€ïl1³Dd Ï[že衱þÊÝé-3öéèè ‚*Þ8•V‡BöàË‘1 —§Z“Ü8Røç¡Ø—ÜyLúv¿„âX—å¾îlƒF~rŠ“m©×W#è~ïHÉq6 î(‚=ÇÌR÷Ï@“ H Ä?õû‰ô;ì'÷¤Œ4·Œïc[‹*Ü¢¡‘ñÖs6pyRw0Øß­wD˜œ”Jƒ²d\Î_ò8¨Ä~²Åyðª_òz S;v¯Ø™ElW¬}-« ™¿vñ½™èjS´ßmm{e²ªìMû(¢Ài Jy3‡…‡Øø¥¹sŸcçÃ6I÷¹q]¬6= ÇH j•¬î”;¿W‹D7ð­ñG]ÛFB_¦ãÌ“B1¾²ê‰³YO$Ç7„~ü­¨{">ôš"¡÷gøÔÓŸŒNiQd88ÖJ+htÇ*©Ø«C¦~À„a\fYïì§{B´‹Á‘.nÉÃU-wÿöñù"+¾ä¯ÓÜÜÓêm¸ÄëžÆôMjV¬ùnš¶L[`ÐÌMàß’ðaÏœíIÖÈÖ§‘Y‡I,ˆÔ&ÐäpCË„†8–J<É+κWèð^Ö_øUÉ°wqL®aÉçP¤È…é£ üñpÁ›'˜B.lüìp÷Uì^;ÄËŠâ“Ü/½‚ñB‰[®÷ {Ê›¼Ó»U°«ƧæÙ³VϻڙŽá¦ßÉCHÅÓ’B8°Ø!`ö+FŠ8¹§DúKß63)ëÁEUYÍ=.kV±ótâ9û΃wË_j¡RJ3ÑaŒ%œƒ*:Ï"°1N‰ß?ö.caèõ[óÜÁ+'®M<ÉõÒþ€\ ™Äò9ݧÀãÓö àb\ÛQ¤ðD!U'3}bv7‘Çvù'×Iš¤×rãV»˜åÒ­W‚Ìo ?%x’ãj›ë w|ÝPA(PBIÝq K~¶DŽ±Óºu´ Uî9~fLJKbƒ9ð„K *Ú†1Bk!ªð)›áóJ‡Å.:$ ’ÜÜ_oš7 Îâÿþî? Ú”AF¦ëš/§æÈJŒÖŸ_2‹±üô÷¨#^®„ñ”ÐÕÆiŠwD›×Gf´¬Ä =¿|öI%¬°ÐÉ/çËÛ¡ï.Ö2#×ãmV¯›ÓäüÉ3A=_ÿ&çź·,/7lô+ŒH–Öª‘åÉ„eÑU¤{¶á*qÓÁ‘Ôw¯ýÌûœf’òGkA”T¨œýX¼ÜxÓ‹Ę°Ü<«)Û•GôøêIAÝs=÷ ÝmR‹¢}¡â7Ôë1`²"ûæ(7Ž°,ê¦\e$lI §´cäÓÄü˜ó÷6µ2çmg(B¯ÉsÔ@v ªÑ’bÙ+_\Æ…yFê/“[áÀ5«ºu%8Ä•Á+Ë8…Þ߯OIŸ¥¡ö“sWýÒßÒ!üÀZ"D½añºïÒbŽVÙ–ß Å+ûV&—‘GJ–÷Ò'—`Ô%?ŸŽ¡:&r³ŒŒIxÑ_ žÑ•6o-!#è18vãœ[<ß’à糶ÉÔÖÛxÐElq…¢ ‹e¤(z#–z…צéi0Ö ?œß­ì,¸H‘^JŠºÊ9Ã×ÈšG6Ƀú=TtóWÓôª<‰½w±rðJmïq°8sÙ:ßÀ– Œ²ëðžA›˜×2H?5òq?eKyßIÔ¨ˆ­1®R&×4J#•½ú¯…ã! ¾V ¨•òÀ´íJâ•ÝÀ9‚ }W3üÓð@¸˜oõ¼aþF³ûÁ÷îC¾‰XÁäÌ ¶1FVöË­³ýÑ#‘k¦³_ÌÌHŽ­×8HSgk¢¥-6‰?Ö½ö$ëK(FDð“Ú-q†?Þµ¥éX!ïä%Ô ;l×ó–/M!EˆÉËXÂ'^]°üJ:§ˆÂ´—l/¾ÕìÜAªg†”†gë0I²G@ÖßdxÛD¯’i*u@Ó‡Sœêpõm¿žŸ°í5>â¼1ƒ´þ¼Çf~jéÖ;<ñBQÿ`÷æLM¶XÏ´5´Wµ0í_Îþù ãí™<›BSnЂ¬‚wzO ̺T¦þá…yÇZ‚ nNÝe_U(iݽ™öMæ»vHõtC2Z úñN<ÌìT¶uãÑD`AäÇ9 Bó½ß‡_ði>2Y—¡“K,ïŽf–˜Q }Ég½Ò>2®n9j+Ùñf1Õé×ù‡Ê˜Ç^öˆ43D­èÈoÂÜx¿)ô¾ÌÐ:,ðäb:n]u39Ê ´¾Qšë¦jòœ& Û¢ SÚœGY²¾1Ñ H¾r Ç©‰t¢ÌàçìùüzÉ Üþ×l¯æÐß“£‡wMâ Í0ÁH¾/©ÕIŠMžvï+ÉáÎìpgRÏz÷ÐÓÎÇcÅÒ&X´ÄÞ§ ø£8èË„Â…)uÞ'€zåœÿO(aôTmb<ÏÍï¸5§?>wA4@ó–*ª®ñ#¦Åíµ8'M…éð—x½J|ÛÎÑŸ9ªe~>§ ¿VGýéÝïÁâ¹ã• k½Ä>±ï ¦½Þr‚÷,e 膰Žk®óòŸˆ‡^³/ðd¬ÃVE˜a5Éà $¡,æœæž–†5úšÉÝ0a…ði;‡g6§r„¯S3 ´FUùÏB$:gÚ1Žâ¬Ìò¾Ár“æááU^zµU=”Œ9-†‚@«ew‘/- "5²/3tVæô„‘l»þà‡.PÈÉa«ìòÄÉ/hÞ60ѽ¸/Œ EXÝ ñùÎÎa‡Í'ͧöÉ7—¹0ø#"…Xú4o{¶ªøŽ%Kilîoúløë÷tBêÅ»»ª¦mœúÃ`§°Ö¢8ÑJ’^,_S¥dʧ†áYÓý2´¹ë·&/å;5\P‘Á%ñZ“›Ô Ðùô¥v“„EGåƒoVÛç6˜ÂþÞ%i®¦œÄ RIL'ïKò¹ï–A$¥·xH©ÿÎå£×#ùUö6BòZ•7d,E‘Þ |<ŃnfwÿE þ¿o—T‚lšPÜm_*þí çÙyÉ)V›»n”W®ËÉ>Y¤ÎD˜TŽÃ‹›Ð75¸]àÌPíè‰ñ'‡„õƒ¶BjÈ#جñg˜<2[†Y'ó |ò›S†’îÆnÜ7 âí$Ç7Þ_‹­ôZEál£à¿á=Pј®Ç™ôHðd0!C»þ§½Ó‰2¨§¹E§ÃJ¿Ð Âelƒ/΋J_S•eª?Ÿ2xR‹RààŒ–4ÓÄ6’5~wä k ô¡ÅŽ>0,'m¹¥¯.xt~©zâz“@n¶|ghÖÕÁä×ïxþZ{NSÊŸÏÔSybÜã. <›äïLg%cNÁ­v̠¨Èyþ{ÇŽ°{s(Ø´ÞFØ|0桦¢+ë”uû˜4œÙþÀn0ƒ¿Ž)vÞXIÍTãÚ"É-Ç‘”))²˜þ97€‘£»»4¾Št•J/™Ä…“Ez é?¸€(®¨Fð‘ uì"~¨éS¶ÁA6ø»'¢%ï¥ãwŒº…&[‘OªHÒ¯öOÝ…Ó":¢)\¦S=l1¨, л„Uó´Ïtˆke&‡Ä?«©É=L™éGƒª;‚—üYÒ,Ú1ŒÞØ0p-=tãTˆ¶ÓgP"tácw®ÙAÂÄ.Xè)nZkH);}ˆ¿7äQr%Š&J M¤Ѩƒ™!¬’]Q,u¹°é[Ä(_ãÞím,ìWPIø•Ã²Qp)®\wOåÀŸü¦r ?ô„AK ÚjÑuIžÇ‰”’Mš ZܳrÏÊÌî÷+²¸ jwVòäHîßÛÕ)ó­±Ë³à 1C9bïýÎ<¹y/¦­T-dÁx[°"tV™'™#‘Oqü°{˜ÐÀŠ¥®íŠ2›î]”2±³DÙ:oîùئ¨åviȶÏU¤©ï0øM“]HݳbIÆnQ£Þ±Ã@/õ ˽KOHÞ&sÁm¸»:d:m¢1ú,¹Æ Þ¸o/ÉcШ¢Ì.¦Æ ڇܙ‹ ǶˆÊíoy÷ì‡î4Ö»ÌøÖ$µ¯"l\Ìzù7¯å+•Ø[õ®‡ÌaïG(%fŠuDŽ?™¢ ˆi4x1füóvFáûÝDò„uãoóh …_mÝ7à€¨Š\Ïà 1ÖA¸;G¯ñçr©Rb'ùJ·ðO°÷w!‰:ðˆV=0X@ö†+¥EôÁЋÍe¾ß(Ñ¥°\ªÕ<8Ó‡3B…èÃ;"îî5*ñp£_mÃ('<§ ÒR_LôÐû¯B Š›ù¸’ój8Ó.5-• 'j®!dÛ©±[Ыþj‡; ‰KªÝ,X„ V{:åî Û¨Ée¢ d3o[h8:.¼¯ç‰`¼ð]xãÆ™ˆŸYÀÇïG]ÿf¦¤¢¹fW´Ë„pRb°V/I$Œáó`“NÅÆãï_ÙXRZBëA§ô:"ùo§½JoÊ!$‹’/ 'Ç”I[V;~ÄœGI;:ˆi~Zk:ZùËÉu/ÙñÈW°ÓÇñçB±§½/M  sL&y‚;¼Í¾ggýÓŠ’†üQ,ƒ#¦«Õšn>JƒüíårOÒ•[Òîó@#AKðX|¸8뽟3­¬ÜÉ–•ïr[» Ô8ÐfNªÿ`kCrQ„z3ÅÐnæX¸µ “Æ!12u…¢PµU ²M˜ÉñòýD° ú,xP}âç“Ÿb¶O³=Ï%…³ôH/n†ƒ"K:ë@¡Ò‚5‹16á’ƈv¢—T%Á©gààz>MìPMÉÈâÑ ^DEk5ïÔ†$ËÛà~¢ C TÓz¸¸žÚB9Y¿ò,–£¡F­á’¸bU÷Ã/+ònl5¬Ú¨–§%7@ßÞÊáZaÑ|Ùõ®JÔ3‡å‘@èï‚;¨X> ‹C™vÏ–¦Ìö£@,v|èH]Æúç+ø€…+Õâ…uz•3¥>¬¯"¡tᤙ`Æ Sº#‹mW'Å„Œ„ÚKT¢“­± ¬DpÜÖZ;]²#BŒ`c"z:×ûÃ-8&>l`|–Ð{. ð€ÉŽ6_•6Ø#­K2ïT hé|§dY"ôè>ÿ¬=.d™c™ºœ¡©ò~(øÉï¨æ»˜p£+¡+ˆF—@Ù9­ˆ¨æ:½ÎÝ ôäðUìèk²){F|}aòºË«¯õŧ¦ÓM"=H•ïOðŽ7æ6§ðvjyRbˆ|ȕ⎆ª«Gª€×I ¼žów(¤$±a÷ú,r”Š®«>ÙÜVF6Šµ«"J¡ñÀûlÕåD*Š±_FRŒÅ?åͳO­½kíåùG ý®G³Ì"Ia“ *ã–¼þâ“ÅäCõ¸Ë‚k+Ƙ%õ2?›YÞiþÄ£NÂCbzÌž¸€[‘åø1ÆÓË>u 'Ô„øÒÚú8 ÂüéÔ =ã¾]ÖÂOÚÿ\ºc‹UD­xÇÑVQb×·í6CÂÙ×ÑnbešãÙÄ©}éËkÞ}!¡¬k ¡††jÇ4;vóD4ä‡öãð,ăRŸ(ÜšY®×“iuµÎ" ä§WX¥Yô ³ül¨`–¤|®¹Ùã´ÌdãóÛpøFñ94¬½é¹c »8húˆoaéUî¿áÒqr2ÌÄQØ¡ÌÓ¨$h¥yóŽA‰ÒÁ‹(V°[HzŸŠ}êp54µKCŸ#ÑQø›î €%¿ ³Ë8Ä&z(wrNØÒ¯qƒëÃÀ§wÑ&Æ㞸¿±gÛ!æ–x9­šM…=DÐ ²OÌ_à«ReÈò´wDA\?MVàñ׉«¨ß›4:¼˜FE^„ÉiHžâë‡Ú,øÒÿ¶Çu&E ¹I.õ/\{ÿ©¬©v?_¬­©ßQº¨X¡ÉYð Ê_VdtÔSHØ£èV·.Jõ»Ñç—eÞßöÝmì¢D¼‘@º5 ¾”·)—F›(Û)×^æ5F¦>]†lž Óö´wKGè¿wî ]ªdn=²läóª½ïÌ‹aŽ9@ > †2Ë<#éûü5ä»Ö$šbLfdצ~xä©Ïaøu±<‚uòSýM[ 8"I÷vBû£Î¼x eõ·—7tÊVŽBV6«Ö>^øyhž/f .›Ê5¥¢Å§ å7¬á¸&ʼnÎ-Êc˜)ˆa Œ¡zÝùHdCSvwÜxl)"ÕˇGJ,ã j¼œå;áÍpÿ¬UÉÛÑI€ôaÿÜy6b‡Ã熉zMu¤ {:_§šY×°ÅôªÝÑÚ ap¢í†V›{L¶ØþzÈ fò¸¿ü”J>u\6ƒ¹^·›"`X4!òQ§!?u›©®nóln¨T½ÎýždVw6ÝeAÑÎTéÇIÚˆ˜À…¦DÐ ·ƒ\Òb_AVL¨JÍË/‡Ï%çÃð¢øüJÍMV÷Ç;\ ÜÝœpnëÂí°áªÃììî¸ø¦á!úÁè®'ÆÁ6Ú_8µqÆìÈ©~0ª c &žoŸƒ„—’"ê%ý$Ï„]ÛÀÈ6=@òŠÈpÚ$'±Ú’¡§M$V8krDó‚ÏéŒ$!v …#"ômI‚;ß9Ø{àTL¦K4¤}Š²ž«qFôÍbµêA$ÐE’aŒ>S²† a ë—ó»ýWbSU× ÈgËɧðî` ^²âk^-¯o%”½éyd.bY»%„ÀeŽ`DXHì{§À÷†fWìéb©±$LYÉ‘«‹^Ké@¢Ôw%J¼#ŽYwÔ-¹ŸÞ‡ŒQç Vp³w`­zÔ)-Bå*ÌË3`ÓDZǂ§Ylâ]¸)Ab±cÀ‚qÈÕZH?­’‡{}áêçËÞûÙ‹ÍBö9•‹¬™±ñÏ}Jq÷#Í£ë8y9–yÄcÖ9Œ¡ó§™¨ç$ ©¥O!- ÜÀÌE7¼DçòkYе°HøÝâ ôwÅŠI[ ‹i@P¸jø 2‚(1}2R(Š/¥ú˜°Çª,Ã:B41xvß"þiÿ#mØÕŽå@š1…´BõôW8‚®RAi- ù¡oluk×Yƒù³GÞÁWå#ÏçÓÖsþóÑØ>–£vÑq Æ Ü,f8T:˜¤/Ôïnűΰ؋Z–‰&·'ç©Ól1Ë€¸ª"æl‡:Úèûü=‘»’æᔲóR4akPÐé™±\ŠPÜ4ƒœŠkH(Ç_ÇEEøZ_RªoŽvPèqŸØêiÄìҨAËÀse·³¤Y-í&Ä0¨¯øÄaÖ<Œ Äøsëº=u³¸#ŸøV”T-½“6$tͯ;$Ë#gòŠ> z¸CìS í‹àIžþ(Ê ‘!'Ô7¾ÛìöDæÐZ&N¯ëû)_ýO5ÐÇ%ÎCÛWVs•–[[FmˆäåGz_±©ç˦Ւ±š{p1‡ß {"+îÍú6Öö¯¾ô_dšqÊ\m|Øté4÷Ÿ–ó4‹ õ$)…µçÄR¸îöÍi÷ 4Œ†ló/xŒ‡Jë_ÓSç9x½¸öÑ ¤ï4ƒæüB¶ñ•8(ó»b“wÎ\hýÏmDË-%^i€ExUrö¿m-AëÒPhûfÞ¸B³Ûfævá‰9Xfåãø`çÀKšå.œN?àvK—tìD§ÔŸ«6…W¡×†ßþ¾ Žà —í#¾”ÇÁe¶oƒ1×ÕAkÈXÇb™Ç¼¡7 ^ }Ö¡Øi¹å“îì ×ä'y~£9âé×6ÙÍÊß_Eû–‚ Ëðù1ãêÛEÆÉru|¤“†¿ÂR:úeS!ýîãg.ßX$ˆôÔp&Š4+¬.²WëÁON´Óô%d¹øõ÷‚žtëﵨ3DºB@}rüÀI<·« è,ø‹{¾)µ3H9mþß  b€Òèw\):6NO^pýkI´Z74Ac–8Ê‚±ghCH*d¯ß‡9§™³9Q¢ÌѪ-‡ø4Ö 2Ò Ý—•,ÏÔî‘0Q o‚l©ç«gYÜEI åb@øîëeéúZÙ³­É¨æí4]ßB-fð¢‰|b–ãÎˉÊ.î/ï_ pk’пÒd~½Zy×–´Q ýlíÞuáI©¡I¦\¤ÄP¨¤;¥ÂxuüÿµK¦6É(ç«Žl `ôz=à?ÃÕ¡ÅŸbÒE‘ŸËî`Qžßgên¤h¿kw"QžßO‡íç²Æ %q\µ©@=8^§ÞÓÍâ ¶sé‰,Ø·6ðŠËØMˆÏë‚ÔUpUÕ…/Ö©¡š]’£Û_Š—=7,¿YûF&-n„6Åà!Ï€†£>§ÍA¼/ÙA‹ T‚ Âp# ›Äãzò§'"Ô&Õ̮⠂ƒ<ùVË¡Ïæ *øýi®¹Éæ癢|VWì0tÌŸ—É?¸mQ~ï¡÷/!̼uÅÒ¡[¯hXæ+I¾ÞáH/F@ˆ¸b!PZŽ'AÒ“ˆè]H'|3ågÅäØ’f–p¦ç©]õ ¯u ^ÛH–ìÞŒßç±U¾†4Çõ+½·nÃ45“;Y$¶BÒáȨˆùF"%f”l;ÙÁ¼æªI³|çšê*àV4™·xï`²º#Ø@A$ˆ>èQ1õ*Í2´~uÑKä~h0þš²É±KÁ&Œ[&M¨ ;KV¤Þ9¦®…§Ë¡³h3Ú 8V®&µ±ã±»‰ à”B-äÌ=Ò=@ÉŒ&X¹f,®æxb2óµË­0"K5­rûêWò[-5+ŧO§ui-gAá+Rm n‰”ZY%ì#±†›‰;'å๺¥ßYïLæ¤eŽÙÃ6òb7¦âïE2õ9MÛç!Ê¢)CÉ$­áB-ÕV!BË‘|hâÝxä9åÀKiص\ÚaåwBSZàÙ­›&;ªzpäÁk.­¦° ÐbæqÅ.(›É¦ö7Ò^ì#xλº‚x'O=˜ë0mS ù}ÐûA¡ÕN¢ò¤_r@aHP+ZÅ®à臲¾,¾Ô~lþÿã„ë endstream endobj 58 0 obj << /Length1 721 /Length2 22225 /Length3 0 /Length 22805 /Filter /FlateDecode >> stream xÚlxsp¯ÍÒmlÛùŶmÛ;¶mÛ¶mÛNvìÛܱû~çÜsn}U·žzºW׬^ÓõL׉ÛÛ¹¨z:˜2Ñ1Ñ3rTĕ嘌ô,0dd"N¦†.–öv¢†.¦Ü S€Š©€‰ÀÌÈÈC±wðt²4·pPSýË P·14±´µt¨ÛÛØ»Y[xÝÜÜÝœ]é\ùéÿIR15¸X˜Ì,mL" ŠšRòJ y5€„©©“¡ @ÑÕÈÆÒ kiljçlJ0³wØü{0¶·3±üNÎôÿ"`çfêäò13'{[€œ˜ª¸‚¼*@L„AU`hg•ü'ÝÎÅ™û´©±“íÿTG pù¯ecø_Ëè¿Ö£Nÿ¶èa˜˜&–Æ.#SsK;†ÿÑMÊÎÌÀþo·‰«ÃBÿrþ‡€ò ©&¦fÿ ]mlä mM”"ö¶®.¦N9{S';€²½­¡ÝP†¶–6žÿ_ÜÿEH¹þ£ƒù?Ê1þÛeé,néaj¢héòÔf†6ÿHô/¿†é¿ÏEÎÔÄÒÕößù¦ÿèñωšØÛÙxþß-ÿ)ã_2HjHi( Óü·þ³3¶7±´3¨¸ü£¥¡“Éÿ +Zþ»{þCæ?Ý`úk9C'K€6#ýÿ´ãÿ|ÿ±tÿJXØÞÛŽ‰“ @ÇÂÌ`becp±pøþ/ºÆ®NN¦v.ÿüŸZþ³þW#™šz˜ì¯Øó„X¥·…Uú‰ÿ©§ù3k F³Â¼:Ów£l9la¢t…s¯ŠSà‡?hÄ·w¹+ußJlo w25C¡»‚óZJReîW·ý®TøÖ¡w ¦Žmk,vyç’¹±¤H=Íšw^'íX·,ÁcPÑ5Ž´“°¶øh'A¥K’¥õ gz@S‹ôÇ"´èÚHàu$¼ xî† [ߢ¨2Ðá(\óÞü›Vc×È®ßyÁ=KÚZœúäû8<’«$âþ)¥Ü’l†tcIƒ{qŽõˆÜ!æÖίÿ³ïIS ÉŽ(òkÞkéJÛJ9.¤Š[ñ‰úp§’ùÖÚø º#ˆjOüñ'ŠÍºw'h—ÛŸž|hYS‚¬0¸"‹]Ò¯j¿†‰6)gë’øÝFÓ†œ@Îce´úÅ_gAaU8¼à 16Í¥¼æ—.oÓtéÛþl/TÝ&w °Ní£ãKõzÃm½s…Â’PvDff¨„ºŠÐŠ’Íäﺚm…çÉ‚“„& …6X—©Ý¿‡¹LSÐÊåJÎÞKTCà_KÔ)cFÄ0Í°÷ÍUEÖ¾è臨)Æ´mMj$ïz}5Ûš°0ƒ–}z*»2@«¯fuùF¾8ÅÌ&[Á‚4ñùàÍ OJÇÔíílzþ¸l›]b¯‚ÍaqÖ DÆä]m&$Ø} EÇoøJiKéå¯ÖŽÁ’•g;ÏÍEjÙþ&’iùØÎÄUÎRêe6µNè”_hNЯo„ÙÅÉT’h~©J5]^E¢Þéöÿa•æŠÞ:Ö\d¥KÑ_óI}®ž=rïKïö±áÙ±üĺB 쉻t‚‰x!Lÿ*°Áº¡Ö4¼—Ây…µV@^$X5ˆÊ'ï*ˆ¾Ôèë™C‰gÕgŽm'|T0Q¨ð_Zœ®GìæwV‹˜xò`¿óœØáH· ˜‹Â”G uÜÇgm94Ã!Qîinõ@¿16lœ¤¿<älÀ&Y˜Ó>dŽÞoI‘>óÿ{gŒ‘ÑKÏ¥V…ÝÛƒ L³µ¾”.œ B]o.8s·H2<5òyýšØ&6Å€½`ƒq)‹~eûmЮcaßæKÐB@ý†Oçàg¯#˜»j€aeô²j˜ƒË%ˆP•©ò³±Hx Å=˜ãOûÅpùM^Ax”åÝ“ïPãúéBJ1ä$=ç¶Oµ¾££“ a‹ *½½½tóýfih´¼“—ËDƒ%<ŸÊx8°6“0|§—(nL·5ŽÂp°TÀ£l(ˆíðW'ÞŸ,ºDï£ UÛ8£*&ÈV„gïêrKð ng05BÜ]2ªv˜.ƒµÝ—ŸXÃÀS kC”w®•©;œZ-MgÍ:M:+ໂŠ,LùÜ}ßU)0Ä—˜¨}ã¼€Ò É—¹<øDÆZ¤£¦ûDk<Ë›£>lK^þ(Õ»–ÔáP¥2N_ØÐÖÆç9¹q2ÒŽÍ@ ¯\Z²ï”˜Ú÷ÎꦬRª\·À„¨…âUfÀ r!5 쮄ogÇH<å3´ gßÉ~ùã¼ÐI¯êí„¿ë†H{¡ZÜô”c_ob¹æ3º±&“Ð d6hxTH¤ Ï¡ËXTg8e+¯ ‰2Z‰úªG­³Uì6±XUšÅ?E2ß ð5”[¼€?CÔ^C´_Ñ·ë_gLûÓIt⦿N|9 q8íõ\sc›øI)š‘8¸–É#n#16¯}ói; ù‡´]í¸[´pÊoœñø\3²—ðý¶®­ŽàäO<ðæLwù ™÷ø¸à¦±à@µÐ ƒ<ŠŽmP‚Ûàܨ²z‰u8:uðÚ,Ý©¦)¶9^?eAÆ><ɘ„OB?”åd®s„÷˜ÏÖ4â`xá÷J!¥B>ô )¾Q¡•Û³’6ûMºœäKÙ·?b2¶cÝMŸñ 8tÄë2é–âyÊ_¦9IÍr³P¤àª*R}: ÑèŸæVÿw{A› <°pä0„…†í×,ø™çÀô%̵˜ué“CÚ‰Â͸>µêMÆ)>ÆÍ´¿Û)ŸýÜ×™ûç|ÁSôRR[tt8;˜G±%ݱÛh/ó㑪yú-‘Iuæ¢&Ê]Š»Í¢ô¶ÅJã’$mý,…t4§§ÅXóË”äôϹ`öŒµ b߯êèmÄ8Û:O•Ü:à+ËÃÑ,#æÓÑh‡s: òRÐd摣̴Ža…;¬i…Ò‹:ˆ^! MSÔ-0ßšõŸ'°ßv~[YäÈÓX±'{¦ŠûvD=%ù”¬&JC y`uPÃãšÁEs¼ö[ú­ä³Btˆ\Ÿ8ŠÇ ú5ÕË*&í½…r¨ìáQdŒR„¸÷àƒ~Åô䡈¶6æu–õ)j`b8ê‹®Wþð4gø»¸®¸ÕÔž^Mi\*àø£p`ðf7;UoÜ»BÿuBÇMµ¤NÎѬm³5H!h¤…U#2+v¾ÆFòˆ#«]þ‹?2ì¨ddIxMÊ\-—Q’¡NjFª)ØìU²Ðªp_mx\ÇþLveaÚ»“x¸`Š ro^Ð\²×¯ƒ^€÷Y# QôÉù'FE^OãS ‡â¶bw~Ç»=ƒ[Ñï^¶>©¥P嘦ù(÷[;”®µëu¸£©˜Œ@k"/\ fhõÚÂ(¢>ø+TéK`3c±ƒ*[}Æ¡øLs$0œ Æ’¸|oU¸…O#ó$à b¼jNFøô#íWÝÊPX‡+ff‹uí…li¥B0#ŽÝd<\¢­FÍôQvâ³lP ƒ±]˜Tû‹Ã'+ápÌÀEEÔ ÐE«¾æ6Û$ÂWØûèa¯µ×»â jÍñ#’ $1IöÛ1p­ ìu`@ ãþZ-e|CG~ª—F•tí´x¼[>╸0M‰Ž ÅõiãJÎ$RVSg-,L:øgâ''¨MÝô~GE…6SZÏÉvpõ¤{e;/6+$´q5GhÜ+)i—ö¼ãE{Ry–.ïåãœÃ;³tîP Â’|ìþ\Ñ,°tÍj“D zƒûà Ž@K›MUíÿMG)¬—{ÝyAî͸:×VÓZ•6vF†OÇ)ÉflþŽB …KŒUàqõüÛQ9|ÇE~)`£cI",>«‘ë´+‰FW¢ˆ\…›µëºlT|ñÁÞÔ# êD2;ÿu|8un®Ò„½ ˬ’<é²C]J à q’SM†Bä›sFä§ÚTω‘½é6ôÏw1Åî*]g‚õéTŽòrÞú°Ø–§š:š¼rœØ%V5Epe[í¸µìdâÜGIC¸àjóNow3ÑØÖª*’–É“A„Ym%Äê9˜ö×"ð†AÄ[ºïtBåIíŠób“RJ{½W+>ýÀcŒï¶ ¬váeÇû·¿Y@@`Aè³%tÔÜÔÃP|ÂC] ñC¦•C¹Æ\”6ÇžîÙKDÆ)q—Í̽l#aï‰ TA-âÛ÷6P‚;V5™žÖu~ÀÂî ÔÙ·s}¯pg*CbgR^øm‡ÑòJgHØÉÄxë|ùÜÓ[²J”ê “²èˆÃÍ„_8ˆÊ?#!©s;ó7qE(—tNAU‘…(EèÎçÈo$ˆJ»^:sã$&úø¦œÉJIn¬ôõ½K–F¸|ŸÜ.Xé7ñšˆý"íšÎÁâÔñ@´w"”ŠXú)ŒߟŮ&6·ú~ƒüd_¨¿Vð@žË„;ë÷FiÞ¹E ³ ÑSÉÊkØÒô Š¦=Eª&ò¥²J©5sÃ!ëö?FŒø#p j»,›Ü›üF“ÆæÄ`3•`Ý2Lx§ñŸ˜:©óÜ°*¶:«¡›óéN¬ÌpOÉLy›Ó~eÕå&ðsŠ'æ*)<`Ù³ÒÉ%/)ˆHçl]–¡Jì|„lë%z㌤¦*  âÉF&°¯Æ:ØÈë( ¨·ûKbÈ$ñx ©äÙRùöÒŠ3˜T±A—_ åeE2© ˜E>2œaÈw;C£aY»m~'X I"²ÙlvkSɇtç\î:ùøVV¬á5H4×…ˆÏÿ´l>ªmþ3Z§ÉýJ5&‚|/ä}‘gþS)[R!I¹T&æ³H=ÆÄ>‚HÒZ¢ Î î?VÿÌc¦ƒÔIª“B5$„8—Q½‘ëÝä!¨C•œÿryŒ·’›¼(¼Ûj|¶ ƒ!Mã¯ùõÀ¢ªÝ-dž;hÄùúÝêÈÓu°ºTì}:¬4€L<Ma]qÖÒ±$m†Ü^·¨Ü÷£Ž zï?ͲÊûBªål ÏF`cïö9 &Žù0LÙ¶×y‘üÍŠÐZ¥fî³[¬®÷Ž§ ¼äöÌ”€ëØ•d“[LÄ_bÊög뉗Zs`ÏO+ø@_ŠÄp‡¡ÃäõøHyAÁ»>ùwnþq·Œ`;+¦û©fôNˆŠn³éN_C,‡GÍúÃÝ­z–¥_?Ö5âéÍç­ýÞÓ»ÌÓL¤öñ™oRi»SÒ&8äÑ 0B­Yj‹ ÖroÁÅmûüÙ|&V¤tñèãQív-7ê;%zb8_R—Øé@¦XÛH06¥H9ªaQÌ.n–Ï VTÁ c``aÁ§*Ï&RAÓ¨ÞÏ“#É&ÎÐÁ-î¬ü«ÁápxÁŸ(‰OQÄ`@æ¾Ìœ› âšáa'tÀ7¼;v̉)=56«vµÈ¾ÎrOú:_Ÿ Wz8‘÷œæJVWlLŒ;!Ë@Íøö9®ågÆk.#Žû¡r¬SÍý"ýNºvy[0”yL6H#쌔ÿË¥þ6$Sk«Qjë›±üQÊihnàá3ßÜse=Õv]^jßÆÁ’ÿMG'5ÆSúÅqF5U}é†÷âtìjææ2®þø$ã8i„”WœjþV#6 QL'6ŠEÝßûð¬¸*Üñ'òÇËÝFœÀÇŠ«Ô¤ú»í5kdêK'’¤=²ð?ÌÐ^ÊÑõg‹¨ûçûùñ›ç$p,P.Þ•žÓûJS»ÒV f-,Eêç)JaŸ¹ÄŸ™6™!Ý÷ÁZ {Ú⎫=q/®„rcÂ{ˆÖ@òï.(¯xRã>í…d|PZÍ!êÁ/¦__cÕÖ9²\õœHÞg骺 SôÁ)4–¢47Ïq©¹Ý’qÊÖåA`óÇY,æ…›|ßbp1»nïÏGzGì}àìXwpúñí¬cÐn^¤ôss kûÇBÏèV¸>¯ |?æS~>$Q«Øøºš;øCWL”9ÂN(OH•M4¡ôV.Vm¯k½÷ J±yÕ–"W§Ê†À™ÚPEi~Íèá—ž©l\’ ëGCqá_ÏŒ˜äe8âµc:þ%)!ËÂ`Ìï&û*fÕ:(â\ªÁλjeÆ2b·Ûø[‘Ô×Á©Ø÷"¯BÄŒØ_3Lb„(Xñ£X"XŒ»­Þo«x_<þQ×?¦O ™nnC~…œƒà`fÍ&˜N†q–÷YUK亞ÀÅlsöQA"r‘ÑÙÄÞÚ‘kF÷VóËp ž$4ÏÁ êƒÝ“l« o%Üc³+cö3–‚šºíŠiÒ"õË-nórµÛ‚*h1bÓ V*òŠI572QY±7µfT¼xægŽS&Ùx dq‡¥r˜‰æ®38Û¢Ùã,Î웿`[%È“œÝª 6> D¦NÏÍGUêð¥ßä®è|þ‚¾&NÑE%0 þÆW0bñŠNìò†4 fz÷è©(ý¬=?Ý–YïåoÔC‡uB¶1Í\ªÀ±Ë’€³âÖ ¡?B¸AgÝ%Y‡Õ@¶¤<.uï?Ê8Rö¾Ô¥nµk 7ï*Þõ,\’³­qÅñÖŠÛÈ£@gŒÇÏo@á´<‡‹oßUt2 '…Õû¿0B2+¨%–tb·'Ï1ÿ”k9·H0 Øúpü\±7G4ÙÊM²GÀúš|k¥?É…“·ø}_´ l¶(µ*¡/©¾IÞnÑ£.9Æ‚]¥RÒõü‘âôÛpÍzíÿÝʾü—ñ=ñKú‰œ¼WäëyÞZ rüs6@““¦…z#þB\¥jbìÉöÁ Û<*(ÒA9µ0ïµ@bPV-‹W¢f¤Ëˆ¶áM8ÎËjÎÚ-çËñ¡õÁØ~Æ-SîÔbܲ¡û†Ò R›çºb?òeA‚ìX¸»¡uñÞWâ2øô%ã“¥ÒH*ék]¥Œ`|ïãsQMáA4!BŒÄàûa° .C{½- ÉW‘–Rtõâ€:Ê ÜÉóm>Ì/bò¬ÿõÁæPÛçE+ʘBÛÝNwSa ‚áv»æ‚ nB›íe&DO¼{¶ÌQœñ¾ +¹ø*=C¦M»Cœþx½·t-Ÿþ…qáÊÛMçÆNð“è–ŠÌÂþÚíóyqÄŒúõûamâyÖ?‚§÷òjGr¡bÈi(zvÛFãö,çiɲ<|›N zõ¨˜Gþ¬2Ë ûhÛ>×ÔžQ.]„ËK:%RõIZqNÐVk­ê+³Z–,O!ÏeÆÝåLjåX˜¹Ê·ïæÀ¸õ#ø¬ëMD¯G*.Æ-ª®Ç{$àˆÀ©‰ á蝹ãv:i€Z$W0vˆ~¯±f쇤;…öésþZZɆµÞ`òùÀÎ9ƳÎð/Ú!Š.³‹.]}P¤B• GüÑ‹å“©ó¤FC 0lDÁˆ¯öVDhʉ5¬§×Î7•œÑµÉQÄ`þ(p­¿Fy P"Ó‰žFpèWS|a±ÞmÏáˆÔ™Nþ26­Í{ýõ±¦Ø¬¾ûe‚¼•š%0¬ýÖ*ɳž,lÕ±tºYq¯5ä]ÔÄÃÆŸ=¤ Üg²ù 2øðOH SÃõ=R厡©©Qûþ#råhnc0¯é`?_î æ;^61‰=›@pWGcs¸É 5ßó¦¤É›!ïØZ'C^ë‰Fìb ©¸G@f8MK©"ŽnT€qº&Áçÿ`â½9š)ÌÓIµ,x•è(ÜCYÛé¡š]õê™sO°8NÙÊ®ºuÕñÔPd:I<>1](ËùH$ëJ%P³½¨RÇf†¬ßc(þáóõO^i:DÔ2…™8uŸ»¢ï"T,ÇGª“ôµÓ>Úo|áš úM˜ š¬#ʉ¨õº%Š9A™`$ly¿.•Ä¦q&1…†) w9¾ˆeßÕ!ÎØ6·e2¥®Íðqk¢ÜÑ÷Ë^Ì&#`’ÐHY#øæJ¡É…Ö&›è§Q`LÆ© }SÎ×?Ãîi˜R³Ÿ4úe°GkmU.h5Ô䎑ªìMm°N•ñLËÜ‹ë=És‰û…(²Ò‚êÚ fóQPˆhÉ›Ô JlUÊûùY%÷–¹ýëä—¾Âkù:cv$4õ~+éã ”Ðo¡Õ˜:æ hPªÑK˜4)$°7¦üûÇ5o«ÉØ$ éÇ÷d«Í·7Ÿ\G,_ a`X'ÍL@n¸ßÌMÁº­•Bv4 ‡ïEÀNÏ>×ÕqãMÊ5`¿íÓòmìU–†¾À~l0j\÷ûQಱÿý6º>¶Õ«÷Hlƒ¬¸ œ™êFq¢‘š‰kÿHý“®ÊS5©Ùè¤wf-³ºåâˆ'IÄMÏ~ÆÄuèÌöµˆçç—Gœñl™ q¹ øorÖØ{^e‚¼}"ãÂ5—²FhÝx\è¡EÞ²Ð郿ÐоøL@òýC®×tô9Ë";-èF ô‚/dw˜¥ƒ† ç,['ã.ZÖN‡ 7¾{š²êo*ÙöÛR¾ `çd.¼ßî°u™k~Œ˜léwŒöã(drê|ìHÒÒQ×# ç]¹PÎý—ÈŠýXÇ(¹ôÁp…–œíñŠ_G„? O¢J|ßèO¬ÜHÃ[î)/™áÞ™W²œBà/x†kt×ÅñÂÈÙ*?y˜Ov.CìNm7ìãÜ€…ÒoÚ®Ykjbšk7·µê‘”í¯ÕíßzhP _«,¢})ü-»Pt×íbWÐ2rX5ù)Û2ïä©Tÿ$‰ÜJuªòï8dñTó‘µénJl9Ë_sr¶à4E¢F°Žg±°p»ëuGÓbH™‹â?nD%BIîf<»Päš?+WÝ÷áVgè´ÐÅ<‹¤žSÒ à¥í¼)]ÒŽ’áëjí>j À®7d¤QÁ:Y¡ª×Ñú¸oP‘¯'ýc>1ëAÚ–_¬;‡Àž¾&ƒ+^]3)à(,iRê•lúœº óG5<ß3‰i#W-3›™´”’Ï)ÒàçïO¶”.þôãŽl0o!ÞŸK;‸ç\¿ÂòmÀ ŠÈè:Aß72‚ÇD~p)ön2ì/2œHþù †2U¾”_¶Œ©ò«‹(Ê|˜zTeƒÃx-ÌskKv¡íÚ‰ôÄÖHhõàQ?Ë/递™F0ì ˆqUOêchuŽÒÂ]V+BUÿÀt35®½V‹3MzýÓQ fûWL˜J¿iü‹·«i¤Yø6¥P9©ó«^²´dãÃsä'#Q‰SÉ-)„{$h_ð~⾞˫ïhå\#Å\j®ûW“åå0c`uGΙð¹/]œi6a~ y£Ö8FDð7¤„rjdgÉ*½ƒ\ùý_kK@ëDÎ¥òÉE ôF1åEÀi¡ã¤qn©»(JG¼ð³Oo6 #Þ’x¸™qcºtC9FC¡òl±üW¢%8t*yÝ‹¼|õQ2¿'27’ߺZsçú]Ǫhð”·˜s#Îû*¼ÆÀ>.·´û£¯m›—ÝÄ$üègvôZ>|ì;t ·Î®ÖŽDR=¡ZÇØv" _Ú?é-"•&~·ž=î=ê\ÄÉþ.O àJäK5±l!ç=e-±g+{†_œªm€yÀtÊ«•…[hní–o]Å ‰„=Q›v´»ÞB§>cãàóRÀ0+„åL#äòžÛÎsºšXgÐV8£xægùÉš5PõZFÃíôÈùê §K€½ÖàÙ<…ŒõÀ‰,ÁQRÍ&ÔÓº«Ë.ÁxjÜæ'®)>aç ¾”Lzì€Z‡ºÞ 7Cj"vä´‚‰‹Þ8+YÄ€»\´è‡3#Ÿd [×u‹Xm@b ôÂI]$盹Mûó €­<Ž<áܯf #Vxƒ£ñgÚi¢ aѦ$¼tpGY`' ×ÿ.åhÍìTa¬ކ̆ ô’õ\`I?çÁ‰š6šõúOG;î EúÑZ9ÈïYöÌÚ÷}=¼þ½Ž»Q¶}nÍk€Ÿ½âåîÖ¶®º[,Ì8ƒ:VµléÁÍê«vpæÞÜÞ«.m½î`é¼cíðúÍ€‡ÁÚ³ ŸâPºˆ›H¯PÉKXÛ¡SžúÅyLàÒ¡€¾Î¤ž×bÇ—á¯d~ñ⊔H«òKGõ«n|KÖ¾ƒöF³•ç·þï}fªqs’À#wüÕ¼/Qå²™{_ÍøœÓzeÙnÖU‹¹ ÍÜ_qÈnEh~¼u¬&cþ8šƒ^Þ•ÏÞýxÁÅ ‰)B¤:,ês=Ê4•Fšiž& ùWŠþp ¶ºw[…k¡}bö2A{(#4WæÍ ˆŒ"º¼¬yvn³*ÀĶÆ…†ÅúR°¾?c5:PôVÛF$¹¤øÑ7œ¯°ZãtU+¶Q„´ê¶×J’ã& ¨ 7æ´ß£­!(j×`fúju¿¤úèÊ“ÑAm«›ã9EÔˆ#‘2•Ú­º¶?÷ÛRìÌ'z=“_H¬pþ®-¼T9lEy]«c‡pÔ·^¯¹.†34ï—ã]iW6R}¬lc«q4=8Yß*´)Ä ŒýcFcÓ¦`Ë!¤#å´z:Y`)qY•…xsöù‘¸`Wú¦1º¼xÚ_mŸäúPZ"iKy™Å°ˆ(WÜ >P[êZ‹RêoáVüÏ!$øwjdyâ–”Øã õœËw°µIÂ&^z™fOfQ,5þÀcLœÛý˜s^Be¯"þçWßÖì“U“¤|!žÕLÓùè,!ض6‰Ô«pПNâ&e€ÂED½N)Æ#û*1ªì¡Õ¤2ËcÙd)WTX<D¿ôGü™Oë—K½zãëNŒˆXÁ(8²'ëⓨûaN ûâà·I<Ó’=Ñ ø~.P¸þaý Æu¾^¾òS_Ï#ª#8mM Ì{Fhá;Û…ÝköÈœ‹ ¯îì[&ýÔ±9$êºü*|VfYOäéXÞüÝ1ðºpOã•Ñ¸‰Ž ß/œ#–Uí:[ÀÃwË( ¤~À1zt¤¬ÉµbúœèE’®”œv:GS>(PÕÖ÷·¾]Ù×M8!ÏÒ8sî]÷¦+Í Bö:â=KPûr-˜ÒÊeÃu‘üNcã„Nq6ûÚÛÛéT ³¹ 5ÖÉ/w=F’ßüïc«AÔß ¶¸žÂÂZoêÊsÑ|ÞŠ¨Æ`N´rI*âÉ‘4,õ |eÈͳ‚–Ò§ÌiªQ‡•H¯þîbm§‡îQÌ Ø t™k‹à§õÉ-ÂJ.φ±–ÂòÇWl\+¯•˜VY Òv±ƒ‰? zÕžSO`ä÷‘V=žÑã.ÄDV]̱Q°½A¡Çtx£ÍI£Ó¢kP¨qvYéÏE¹ýËzlÌLk[ÆæÀHŸ‚¸\ ÏÕ­²ûOnÏj½èX˜ÞàÞY«„Yfý=>GncQVZïJõ^“ÆÌ#¥žLÿ÷S¹ŠOÄÉf-w¯—<ŒC7*ÑWç±6—ÀðµDIc¤NYûèøIxï‘w¾¥Ì¨ãÑr•„,l_6 ‡O0}H,7n¬Y2ñ ,m¤q“sêRL<¦‰”sÃKädÃæk·S°ŽÄëbIs¼ˆ¸Ö ÿ‡Þ¨°?PNE’Žõ¨âòp¶¼\×ÜÞÒž}Rú[Ú´7ÕŸ1.¨*b:‘«%…_`E¡ÑL<iærJM‚¼eË(I[¥™b-ï!¦NŒ—N£>51€üÙ‰H[S¸xN‰ŒpÀŸ|hdkéNPNÝýáåˉå1‹ÅNvïRÛÐÂì^Cinë þâSxÉŽø´Gâ\ŒôvLRE8­÷´È×í{ =¾,9QÔ?yŽëW¦ÛŸfsr¶yký ^À¼Ï’¸j%‰´t«øE§‹{b*‘0 tlssJ8–žÆƒ¼•¥TÆ—Úº¬™îåbÔ>ÇÆÀ…[}l"ÿú8d±ÕR„뱌œQU‡Â.Â[Ž$~5h„ÄÞ®Õ’ bÖŽŽÇ÷ažO1Ô¡áMâj·õN²³­D<•¨Ö‰WO¥^ßÞÿk£¡tíú€ë@ - Z×%ÙæE…J¿d†í)Ô>aÖ03q\ûQ¨®³hP5¢¦ÕÀ –Øê&q.UfÓy{0Æõ>M‚ï¡sD\ä"ßÙì´^̼‹Úå¦Vðýû×P´E·eð…G”Ì戽bÿÈ}lÀì‚2GFH“†…b5î×±Æïé:ȼ`} ¨¯:1šÂJ?;H,ãHÇMyÆï°LÍ¿h²Çê¦<”(n*2!\F@¯hAá!tTïÖÄk¿/¿at'•;nú."½}P7ë[€¬d¼E×eò#­†~#¼Ü³B3¦úó õüfaíÃ?hNå¶ö0§-'j”‡ÓW” ÚGû;X©K(˜]}žý»rÒ–¹U‘¢ <ÁUÿ¡#¾/pCìöš¬1kÞ8K6ËÜÐ(l6ÿ,¥ç•A; ’Ü Í¦k²†›¦혺£µK¯=1HT¸Õ²C(~yzåhò˜m²ëêÌž“Qc{¡- ëU¿Š•FQ¼!£3<ë§D'í8=\O rœuæšÜœ'»I×Ä3,N}(eçGµ\/StK7û)"ò@å€"[{ÍCê¯ÞµպóÈm&Ÿv='v‡  Ö2ÈshðAóW­Ðô2Ãn䧗¤ÿä«C!*Œ_"s:V©Bªk1fd$1Þï†6¾ºÖ›.Sµ>{Ä¢”aùΘ°ÏŒ#­9‡Å?•¹ ‹NG´@øÕ¯TVXá܈MÊÄn/Oúf[?ŽúÑmþòPBLdÈ#YÕ-³óE7¹>¦?~ï6vóIXs\øDÆ(³zFÅ쌉ÎÓ³P·ú^Ê ´oèg¢IuNÕzeÀ”ïïÊÅãÿå'£~=GÀ øÚxÞà­…bçAzÀ›x ÿfLò•ºÁ¦G>æZCRfS ¿èØ ·‡Y!†Cœ¾IÝ!è*1’fHqA V éÁ¼Êœø؇jÔ•†Wõ~F×JÏikP‹šž‡!?ÝôÂÇþ{Üx˜0t=™5šaÏíÜŽ³R7~øwâ‡ÿŒ³[‚r¦cx<-¼Æ¨ÛÑÙ<¾ìôÞÎéº4ÓÕ¶×fûý»°éè—ÿh$¯Üíi1· DÐIȦúw[ï é›Ç¥œÚd¦½°©N¬Råí]ïCÿQŽ,˜¶†úçØk)²b§þX1œˆeÁÖB NR#|ˆä/d„eÄxëq %Ýo%å‹rYŽB~ŸC †>üBjh½çåìd)ôùì-Ÿ»\.Ž…ý6âÐ, ¨±IÛ°ŽR²ÜpSžönƒr˜ñš<ݦMÝé±pP"óˆ$ r¢­ê“«z¤^&YDYòŸI¯=ý @‹Q³ÜŠ8híðBÅï®*Í2Ûy†§†¢C )º2Å!àÀ»Çă‡Âð„iß{ã,fÕ¯­ß ËÅ)KyÅ›ù÷{NÍlïð@R ;Jn6gZ,ƒb{%Æë!%˜ª'~pHŸhÈ9ðM Ž‘VðÍ”§$mtá‡R¨Ü”ˆŽk9{ÿÌ=_pxÒƒÉGmþ"ƒ 6²d«±ª­†zs.Ó%|¤UTƒ"M~L{ïå"=šÈÁXœfzÊ2µ+ýuîʪұÝ;°R˜°HŽŸ¡æd;¿këŽRöÃÞ Õå~ IÓ&ÑÜÅh) ð%<îzˆ1ŒI…¶Ì 0‰×-{Döï6ÿêISߟ@^q–t”¥Iб¢;+TÜÑŠéÀ;Nõ•}UJß½©ï£u»ËºŠ]N Àú›Ù™d:Ço#ª„ƒ ºrJx÷*×>Å^§³ˆ­õ'-º0g¡œ™ƒ,";`Í~ºžÞæ9ú §c¨ŸU ÁX—ˆñwŒID­éWÜÀt¯>_?î`»6(Åaÿ0nl>1¨ëîF  ‚Ø&¼háÀãÕ5"í9g8Åj‘³àâÓ³†á7ƒAkaÑ`ûKNK·9û,Ò¼ ÒÊ`hц8p/ŸÎ­‘ /Ên z‘ü}Ð\d#É8CåaaÈ+{Î;Zìf¨gm~d!Ó‡KŒsG)ý+ü=÷˜0>ˆÆI&·kð¹%Òݦ÷‹CêÙQà¡0毛±²@ÎÄê¶#¢—¨¾~YÕ'ÑÚLXS:¶à$0tò€–ÖOÒg‹ï¼†§_£±é¦4ÖÓ®&Òu“ËÖÉNEíg[¯šãÛâåTDÆ÷RY×Ïp,ó6_±œ½(¶©ÜúLüÞR :Tù¬J¶{”&¼£›·Y£—×'á¿ûÿáæ~¶€N–úÆ€Pçãx`uã%¢ÄVŸRÔØÿUvô`§L¨?žèc*÷!{ª›U¨šƒaE(¯´­¶}¡EøgGKK}ÊUM˜ç6/ûĪžI(]~u%*3À·ÓËGßOÁÎ[ýÌÈsõ—klŒ^_À>gAZØÌjÑ3Û`rˆúÚØ0ÙFdüV®aâ+AfèŽÀ¡%~ šfžAB73 ‹½ÕSO™;CàZk½ŠJÁd¯V.êÕÄÛË{uÝ}'®RöéÔ Mn±+Ðt̼.–Êçþê$U‘Kœõ >÷#Fj{œ¬6°‹“kªêIŒà¥Öãæ‰pS%eE f(½èW±ò'Í.H°—ᇰ¶íj>4Ý1Q¡j.ѲŸ!4f{“Ó^àžì·“ò_z{ktΪN`è_ý$âx…êé_ÎÔàRGh³š|äS1 Óƒ?ëb¢g¬vèé!Ü.Á‘F¸NŽG9ȶBüÂ,öHÕr'¶B (Ž7È’‰«3.¨Ë1Á+Û'–ýmw5u´íCumw’þ芙â,–™¯2oùtUÅÊ#—Ø’oõV:EŒUoÞJfk¤æž]‹ÓFOij‡kéŸÛÎýæ¸ ®@Ðé’ ;Ãd[¸ü~´=Èœ 5nEaà‰¥.|§°÷ŸR9$´mÐB¸±å¹¾Å`OÚ·±õEOœÜÓ¢õï·õ6«—)í„ЭÅxñE_9w­˜Uð+ÍÏ€ \”õ),D¾©ÓlíÉsc…æ°«ñ—øY‡ˆËü-õ½7‹tš®¸˜*šBµhÐòöÁ–m<§OÆ{¼¦tÇÅÌÏÕ.hÜ×Ék‚ˆº‹’YC2Ïî¸öú«ü4-Õ¼prôZý.¢Õ-9˜ƒ»š÷DÎû~­Æ‰FE¯‡µ6¯Ý5>G–`îX8i{a ¡–¿7%w×;¶¼ß©qq¥£sT¨hï¹tneÍyܾ˜¾ëãggu¸£´:þ™hÜW“¤ck·YŽvRé úqØ”ïÝ`tºlú¦ÞzÛ³(pãå¸,# Z»ìME¦ÝvºYBJÝž p0»µŠšvxõŸ¹I°lHu9ž6›…–Î…Þp“â?áÓóÔGecp¦¤â?Z¦g+JiM¹Ò<“…â[äRÃýŽwçô0Šåê8«:°EêZq]×~!rGeŸádüDs2ƒ=S‘$ Üqâ7°ïJ†»ÚVÑʽyÆ“AzP<¹îãûÓ¶nðÇ*(_T-êò¾•ÍÚ[cJrØÔ”ÔBuê×\Ô}X2Y–°ØáL•Ÿ©¢;¾8[þ†ÓžkÜ6çób©Ã&-sU¯=+®[GðJ5yCÑ›¢`Y`weìê>ÔëŠwñA¸Ïz4Ý}ž+ž;>pæX'?Î?ωB\v4ähûd#êo»RvÕ>û"© Xðj¾÷GŠòçWU±ƒ6Å»ú:$³+Ô[ÏÂH‚©I;¯…Žuº /eqº#çíëÔ)~LÅœ·›)k-…ñØ+©’§DÒÔu&ƒ¥&D{.u!¡àˆVZg:’l¤!ª+\t…Ç!GVÈGP“Ì^7±4úó+a‰ÜÍY瀇ÜÃÇèv¤ }55 /”šæÝÃPbÀ&ÉÿÎümîòâALiAq ÜY.£`t¥›ûDœm3¿·â$ ´h Š5CÓÅ<7Ý·¼§G¤…¯Œìîè>?+4b÷ƒÎÈ=GN«ªq`dJ“Æ)ø±ÞE[Õ<ʬEpÕCÖƒ4nщc¶^*¸/9 g¥¿Iõ†@Nº¦Ar› 5lœ†%X©a©¬¸¡Rã±¹yOŒ¦NAP’àëüÅÕ«É#NAÙ2'üx7[U5ð'嶚 dÂÁ÷Ä!¡Q(œ¡lùKöÌÈ`€e‹N„mÆœ@;ºÓ(g•Ä–· e55Ä8p”áƒtÜ3ÃKƹ¯ýâx9µR¨£Y‘÷[5¹Ú÷+¾øuŠÈ%¨h¿'jSúo—´Ð÷˜¢Xàশü¶ûJRÐbjªuáŒëá‰ÂKîí4‡¾š»ÍŠú_³êú7@SöÓ*ÈëY×Æ‹õ²É=6üL\e𘲛Àœ”0J§èxnUÜöm6Ô?W`V1Ü9P}}’¡ƒñ3E­Š³KYý± “vñ’;ojŸå&Ù‰­+êr‚©ØAy†Ž®$Ë9óÜÈHK÷Ð-ÿ¥ª å°„ù¡Wå¼Þ…¬%{Àû8Î3‚ß& PýU*ô€—-‰ï„ÀqÅß÷2.ˆÁïñ)Ç®ƒ+©<‡-zÿ þM%|Ù- ŒàIšO´?€er¼8øãµÅ”-ß;¸–35-=°"PHÔÄŽ©3زJÈ^Hßæl¨x³}*eƒR‡¢¢3ÁÅ¡¨gÇ̆7•pŠp‹rÛ«;È ³ÅZÿâÌ[]CÒ7y.ñü¹l³óÂfäP_̕샶ÓàOƒóÌj E%¢ÓTÔS¬[¤}€X»@Š¦x›/Þ¤ü¢Á´É Û¨`ªå.¬|¬ášÕ`K ¾üŽ2—ßÃd`1¼ù³ýÔ!J“ð34†‡œ8SšºµA ºœH^‡]“ï´‰·ËaS‹¾KœÐv§´?KMluòþáC4ê™ÇÆÍsÐIíu°ˆ¡k5¥lÁ¬£,q—EÚPÖuÑEÉëÕF®ðKT§ù b…'ßÐ6èhäKÄwR¼3 ïÝϱÕý4v–ÆÇ,_Æ÷éùö3Gk¯¯Ö¢%Ü´×zr:¾,äo¼ƒgÖR6×Ö ~íøC0|&~M¸cÐ@YÞY:M˜è¤ÚBÁÚ5êˆYàï‹Œþ$‹ëÂ)e…V á¾Þ ¹{mf¦M |¤m Ÿ’˜ÁNÓÃj9胆«µKZÅ+‰†âYáì=~[L_ ÄàF©[h·Lë,Ä0ijã“XK‚Ÿ¤Rφ½ûý z,¯>`“­ÞYÿÉŽ¡15ˆeRÈ rÜG°×©ã¹JŒRôÍ™w|p•z¢®ËÃï8ï³yza>˜ÃQÊLL¸r@.Ås”V~5y³kl÷|[i`ý}N>0Ê’Q ó[Ÿ¬&énu„hnevÃ!:vþÜî× Pø IÓzhžë"'”‹†X fSü1šÔ1Å)pm’™7“±Ñ^Ü0°ÛõXêǯÖåâZ­`ףà Èö–Ù,ˆƒãç¾~ÅÊ :)‰ˆÙщb‡¿Z УŠ V»ÝꥫšÂ´DüO­iEgs¼çèa±n¦,1$–Ê}l'ùµ˜×À%íÐŒJ…ܼè}b©sQtU™ê²—d$­ˆ*<Îô‚ŸãWÉ™¦Ũ$)éåßÍÇ©7'´LJê–6±L%Ë}1_Õ@jA‰)wÄtGä»F|Û%IDw»ƒš] ²·‰N:–å­/ýJÚîGø³oÐí.%Ø«oe0Þç1#óIVúÚ¥ÎöשüUÑs݇ÅÞ9Á¦å€;Y¹§$/ p`j¨Ä<¹$,c›‹Ã5<2ܪ¬þ ¢é Jë,4x ÝÍž|NP¬°Ê\ïÎ\bãm«Ë»ÙŽ»hçé$–ðy°ÎiX4Í€‹¢ŸhE@óÉάùÐ ¡1TU§^º’¥›ýÇd†ÿ2P” ùãQøsìã·›" #¦zjOå1íÙÅhé=ïb~­ÉÄá}ú˜ûè1T™°"R$›†£ü«\o‡©,­œ_Î{@ú¼ŒÜZ“ú8_y Y8îµ"ÕâQ‡*´’FNí©—ÀŸèJ3h̃Mr µèþ>Ó¤³jY´/Û®wk×°O¢Õþýú ÃyR%®+úë”Þ›ò ‘`r¥ì×/+„d*/Ý8½þbzBÈ›Ï8Ç4p 1Ù3:oßÄÆè'gK\üÈ_½pr½›ßIY QŠ9cìÐÐØ]EYøù òœë›[¯YEí]»‘p‰†m#yH/ä–íBy¬“4“”[±"e4ã#ÿ'ïÔÛû¬å®·Ø·¢¹ÖR3h§Õ£fÌu?¹UFqøÄY?U|E¤nIÛKn×ÿk”ç@òË ÞñCý³Ñ+EKÊu€‰ÇSâ DpˆÑMæ=õIJ;žw Aw–K‰qDMíhÑI%½Í&‰ñÍê¬%Tàòì̺—ý]§ÀkNâåtµlÑ _ü"ÀV•ÁT¤Ž”°.»@:HIzÓ‚ågãA‰_qÖÂÌKÎ(Fã3¾%Õ: …à0Ät}ÿ‹6¬4ìuÜi­;ksÓ]ž/v¹mƒ·|kû"¨/‹µ/õm¯éð\lø ’ƒõYt+`å”$_%G.fSÈç¸]þ*Qþ þ@ø·ðÅð`¢öÍ`¤&`[yæOæœ3™Ö8:?4œât‘[ÒÚ¹æÀUyIqrs4Æ»°^Ó¯á—p¥wî÷ ç‡ì/š!v³5#tç\YlE³¡7žÙ¦Y™ívÆÔo E‹×¾Å?KZ@f>% «A ¶cILjl‹ >붺‰áa_&žàÊÓr•ÙEªºw†¢wtê9ÝÅOÖþ+…×µ™ï²t»vÄ]™Än8>x~ý¡®,Ëÿ/W —ùIp%¾‹,)™Ò7—Éässû!0cFg¼q‚2d­©1ƒŒbwà*w™ø¡ítÕÅ€£¡Üƒ);¶fô'‡_Éß#Úš€Ë?kºv­Áè7 ƒc×ÏPÁa´à|ÿºßX«ÍŸežÍ¹¶¬çBˆ|QíAÓïƒÆ¶ÃБÃTl·jðõ›ï¸šž“Qb ¹M­®®URxÔ¥ƒ[,cŸãmis+¨Â²¿åÝMÖ è‰ Ë­*`Û‡Jt/dí‘ZóÖekx#½>ÕB#†„R›ÄÚ Èüy ~“¸NÉxÑ/ò¸Fn+-ª'þÀ¦/ "w«!æÅ Ì@ðæž„;à¼3¯S Ìéí€^I²¶u'R£wFÒúNv[rt=>ÍwëD~56£m®ÞÖ—8w™4“¥æ]ëÔ¼ þ†EñF‹ú©DivD7™’Å[rûW†cCPrŽ¥_ÃÑöÚ/÷ø3ã-Nt®¯#ÛòLU©ØT–‹­> “,¬†U±8ùäÊž7~ÙÈÇB…›©ŒA0*ñ¢å(ÝsáabŽË³#ŠPæ 9…SµƒV¦’,˜Ó¼D¡ªšG“Úc)î\­£'×C6Ñ­âÚð2ó—Ж®à—QºGÁ%,É_ÐCÊŒ—ÕÈ&m.Œ Ï.°fâ{yí&ˆÙ¼xcׯ´É³»ñIvŠF•Lµûr©ü]ÉCÈü¨|ìijDîÞhr3h™‹Yàߊè¯ÅÖíE`Ý€RI¼JlÎP6L»J-ÄRRjÌEe­ÑC凡#µL•6s¦™fë§ÄW‹:$ÿô@“8 g[ÙÞ°ŸTùžT°"u çIßoñèM/ ¸Á Ö.)„Ý ±š!=®Sº„êF]–û«ç¶þA…ØþGȶÒ}ò¸R¥Ñqw°„¨s_.$ô£‘'’”ÛOGîúü3Q,NO†æ{¢óØ©€­ÒG®ªç‡s®²Üþ‰\GiÈDñ×:Sþ(Ä×tº>3ý{žhÛÅ/”ÎTŽãåý0¼rŸãºñÄ"“b{Ž¶¨e$Œ7ña®§ÈïCì²=„þnšÏäJFA6ð=tùI2^Ö& ªkE’Õ|=¿M¿[Ÿ=Ã"wÛ­u n°åõ°^šTÊ”–hbš‰g¾8‰‘`ý[P’6k;T½ý8`%ru×>Ÿ9<¹€dªé¥9y¶@« 5²L‰ñ ‚Ü[ÞrU—³Ûðfýy*Q'†F´”‡¥ãC£i³óÓ4ß,R‰¼2Œ¾¤[ç}[\Œtõfür¯Wr‹Udø‚²œ–ɦå䡋ȘT çòž3£Ã¬ª’ñ*Ki´ÂhSïYÓñ%µü¤f`¨ÞZˆ] öÖêƒ;¸•ƒŸT' œéîrlL#È2= Á%Pj1¬‡¹é‘4[‡l}.– >K¬Tµik¼>Žx=K7É^3ÚÙRmß«RM~œµA*h,¤DZˆ|š¯ÑÀœÝ‹á±Dæk…À f …I'.Û/ÀÀ òë\”èw1: YIA <÷G-ÕJ¹k«TÒj,xÑPô¦¯[†…Uamj·¥®ñua‚Î ‹Ï,ƒ” à†/#Gf7±U3Lc§åPöA¾î‚“¹±'ˆ•à­.\梙£ÖÑ1²m‚LK‚²[1ݳ Ìi@„kmm)VîR¹9–ƒT&Ì–?›1æ‘<ù ­ ˜4.—t)9êú~¿9–3A yëfóÌøç&ê(YR3yV~Ÿˆó „:È÷Š ‡³«±¦ºÎ7–Ç6ª¤)•w†´ãL+±­v¦Î÷ŸGìo‚ÖÉ1F£oÍ+åa(vܧ$²‰Ïf"¹) ›¥*pßÉ”ä´9?Q¦¡΄ ƫϳÐÍËwΗïô(~¥×"ÍYüVaüˆ,ýXSeÆF+úì»±DxUªŒ…7c{„çžó&B•n Ñòàq–9+Š/B§´\u5YkˆiÄ}æû›œ>G³‡èÌy¡Z s®³:—’ ®Ó‘tIÒiR¤¦¹ÆåQß‘!~œ±Ùç&&³G8››ÙñÓ†ÇØÚñàŸRÏ—#ÆTBMWõþ ït¶¶… ŠŒñÎTHGûE¡ú‰‚¦zþ…®€°Ó~ÎÎ^³ïAn #øžÃ%<¬ÎÔ±—Ajþ>~:oø i]ˆz‰Uýy†ÎÍl4leµûü¹5ua×imV„z?¡ª‰¸Út¾˜½yH  _˜¹YÚ<««uþ”ìǽõá*isˆKÛÚÛüïî‡dî*H®ªŸ$œÓií2׋¾Eó%\$Orúˆ†î %ì~7ÝŠ"¯ÿŠ3Üög/0Dpu³ò¬6kõ‘,'SÎÉ;á:ÝîOXÿî¥M—©‰+pÝÒÙfYÕ–aøUÇU7b£Q"LÙ•­Ë™FBCm°kŽ=y»Äâ[›C‰…ä†0K¢Ã…Rä×U:_«Zþ†ðsÞ÷«˜þ–‘Ò}ÒË®ßÁÏar‡™Ûp3¶l‚WÁDÖϱµY&Ô¤¸Áµ9û°UFœeÀÐyÓúvb¢$ƒ½'Š’„·nFŸ”ç@瑹ۈ:I0N#-×]©\ˆ ¤éëpÒ2&K÷yX ¢QyfƒåŠZ¿¯æ±Y·lýÆœ‹ÎEòœaÑ:,ùî©t\L8ºg®(k½Xõ¸k·™D¾ qJØöH±”˼$ñ– PÑ÷ø,d0-8^‡ÀTU—I“s±˜˜*àÕ[ÔŒ»¸â…Åh0Ä24Y4„⻩Úß.‰FÒgâ½ÔÜaºAª;íÓtš ‰Ü~1»rÁ!¡èUþ4Êxf®tGUf\R4³›ú h¥ê#Ÿ{ÆWóº"ð8àAyDïHZ4ŒaìO#¾PXDé¯xeCà8ÈV¼± ",§ŒR-©§­Ý½ÜëAZÛ jƒNYtBÍ0;‹¹šHªDç K$E¨•BEl¿YEmb³[ lQâ…?óÖ†üpÁCxì—ö¢ÇïÀì Æ5(né8¬Ÿ(3R§4ëËÕ€½0§,ÊÎÀ@·=¢Á1ýmÔ’ób!jŸþÍ×5šÌÛnï8s¡: …Ÿ/ÑSd‘užyÊ÷è°j)í4¶Ñi§"£2‰PlŽ)gXf¨Í®mú21©CçÛYð?2…»;\mÚ6w ¹ë •$u´ êârÿ¢Mây´Çý ?‘ˆfP]Þï.cÕ/*ŠòÖ?B ….Ö.îód¹=FèC~$ð‚7!E—{Uˆx€0énJKi°?ò}"éNÐÄÛ7©šÇ’€Üeì­«G,ž¡-)Nv]N»gSÖLFžžÿÈëÁ p&„m{…Ð6ïXG gåQ—W e­o¬Ã5Å-«’`ƒ0$Œ»8 ‹hôCX|’ Ìw– À¬7~ùRr/o£¦ G†FéÎ͸J‰Žàu$/)Ë”QÁurÜófª‹Kœm©¡“ƒþJÍâT ÁÿX™žç©w‰æ ÌÞð`¬ògRßKmsÕ‰+scöÜëvg„Á{Q‘¥‡Š_ _üÈ0ËP>꥕fƒÒdÓüÖ™kœ÷ò¥‚er$:Ï1xÀœïÐö¸ÍÉQ¬WŠé”E{úÜÉŽsfÜF9ë<Éz\ÙÌè˜ £¦ˆ¾RIfû{TYê=öñ"ÍØÆläÅîÇò,`ó©‰!ÆúÎO…3~‘åøõù[˜ ~gÍhµž›Œ%ÖRJ8+(lZa*SZçí»í¼ûVnUÜ>ùѽǬg 7)û‘ t¦ *¿ÛÚ«¶E·~7ãq0,6û7Òñ'$5‚!s¤ù™¹Ùæc”á‡O®Ql+¦”±Öëœö¡W"›ô½w‰}Á³Åš+·Mîr»~¹š©òĺWp¬¹O Pk»c!·Èk?ÔOÙûhâî)b–³F”‡73Þ¼Pˆ»Ϩp——êO›Ñ6±²ø¤ú¼ ,‚¼È±Ë ìCÛ±\–s:gjž"ØCG{®:Æ)¦Ûz˜]6ƒ«.VÖu˜'ý¾úÚüZ‚ë¶"ü«eµ¼ÜÁ·)vºp…G”ÉC{5êtiÄ€q¿HJÈR ûÏèÌÈ«0Bjð }«×· ÐH¥ ã&Ï;£¥‘É!º¨¾ºÁ¥úpÐV*×Ñqé)‡¿…ŠÏ<ŒÆ1¼p?Ú^µ0v|YÍÑßq3 Í\¶Û& Evµš6tDóŒGל¹9Wšƒ½d^n}&@ eÊá CS|æJ™Ž;ôyõt¥rQ¹—NeÔr¡¡Ûÿ‡àm„'M\q{ iÈ䢷«Q҇؄ÙÈ¿âÚBÊ2·×—v_!)riîs…lN Û[<¨Zv‡>Çý˵Tr´­Š4MÏÃ߸Å>HrƒØ¡%LëD“íÃ#mV4w§8»Ùª.…#òó€Qkf-[«~eL+åÀ¤HzÀ±€Z§å¶Ã¿qH•Ï·HZÏׄ»ï 9»÷W]…γE“hwãEé„ó ©v ªSo¼{AQeÏR ¿È µ£%VáOHA²kâ̦ʼ¿ÎÀ½ûE%0Å÷ðï&–¡tÝ?L–Þ|–ÑÁ•ÚoÔœhÞÿ0û¡ rÍ‹€T<Ðú%ºgtGÿr)"RÉVBŽ¹V”Fô€$†GYOiëi–h»>zdT¯@ûÊÑOµó"¶˜>é÷ýÒÇÈ£ž‚%´}€Á-N6ø|k‰ 9%åÖu758wЧÔvüÝ“QG†Ÿe'C’£ìeë«œÆáQtÔæÓœÌúÓŽÉ5ÍÎ#TZºu6S@C[íCÐî«÷`kÛ ¿bPûüÚr•‡åßÏ•zk·.ÿE"C-%›=Í¢LD| _lð‹j¡‰',¦*²æ~FéyÀƒfÄž”z‚°û,xãOñÍ“3智^""ñw+ ›°ó±O¨Ì&¾Ò¬¦HPOˆ³<åh7©¯”¸BJ_~çLJī ñÏ0ì,#%,{¯;v’¹N?¿5ã ‰¥eßêE44X¤ áÌ ­ýjt+0ÜðŸGY{ãS2¯”&ÂÃöa§_Ô[{Í™–ìp, ¾OâÛÒõ¬á!K“¸~§CWH07CS£º;â7s¢}($¿¯úøˆp _/–vq&¡ÔŸÌíHw;Ème¾¿¤%fÕºa`­’*«_MÛ_Kr"uù~y,cÖ¹¿†FÐÅÙGñrÒ.¦úz^a›¸B¬kHê™n9?ŽØ—aGêv…+»©ÝuF †¨ smpŠÝþòqC~!™ –†U–¾ç'!˜Ã<¤$Vp¾\©í´|µ™FòÞ°k²«éïC¯ u˜Xã˜æe¬P¢>À›=JXvT_ØÈd@jEMK($Ú÷Qh$fI½£z¹¬5¬;«Šôâ¼øÊÃôVÐÉòõÔ·fÅ`¥™»Ž!….ëçú™´ç=Mâ*§®€×W/l(‚%^£ôqßèîrP+µý¨ŠW\YÝäÌSÕŸf>ŽˆŸF– .-äö>ÑåÀý²*Çãá+ À’µÍÿ¶Ÿ]L“º€ßƺWX^ÜÈmè(nøáã H3Êœ»f$…4ë_¦fÈÉöØo±÷yRò"íßrÛlÞÔoOp¢'ÒH? 'ÎÞZ¶³rs+¿÷¨ƒ´“‘Ž» /Гä›åŒï[ª§¬«ù¢Íy^¤G;œIjIJ<ô ÕW‚לŽNš·G×d‚¥m;Gø0]Û†Ô$»o%¶ˆÊ>Û9p¹V ð*4-¾%ï»_ÍÕh‘¤ˆp…S^žþ¯ªËê,jÚv*QÜöIÏxùQÕ”¡'õVJ—}:%ßöLd:«Oy„d-õ;‰_° Ï¡!»é]­™ií#[ endstream endobj 60 0 obj << /Length1 721 /Length2 16693 /Length3 0 /Length 17293 /Filter /FlateDecode >> stream xÚl¸Œ®M-ܶí~Û¶­Ó¶mÛ¶mÛ¶m»û´û´qÚ¶ï73wææOþTRÙX;µöªÔ“‡ŒHÌÞÎEÅÓÁ”‰Ž‰ž‘ ,¦$ËÄÌÈ`¤g!#q25t±´·ûeèbÊ P75(›:˜˜ÿ@˜`È"öžN–æ.Jcªj6†&–¶–N5{{7Kc ¯››› ›³+½“+?ý?Eʦ¦ S€™¥)@D^ASRN@).§ 7µ3u2´(¸ÙXd,MíœM©föN›ÿ8c{;Ëqr¦ÿ7;7S'—ˆ™9ÙÛdEU„ÄäåT¢" *"C;€ŒÄ?åv.ÎÜÿ MlÿÕ-Àå–áÿ,£ÿYÿË:yüÇ¢‡ab˜X»ŒLÍ-í`þ¥›¤™=€ý?aW‡ÿ¦þ!äü?å?RLLÍþA»ÚØÈÚš(Eìm\]L²ö&¦Nv%{[C»ÿ¢ m-m<ÿqÿ!ébøBvæÿ(ÇøŸ¥³˜¥‡©‰‚¥Ë?R›Úü#Ñ¿ãê¦ÿ¹YSKWÛÿÔ›þ£Ç?7jbogãùü§È £$¯®ªHó¿øwZÔÎØÞÄÒΠìò–†N&ÿ ü;­`hùŸéù/™ÿN€éÿù²†.N–mFúã¿Ö-Ýÿ‡¶÷ð¦câdб03˜XÿÙ¸X™}ÿ?t]œLí\þ-ø?½ü×ÿ÷ ™šz˜Ãl¬Úó„X¥·…Uú‰/TÓ,̨Ҭ2¯ÍöÝh#[ŽX˜è]áÜ«à¸ÃáÏñí]îJÞ·Æ›Âýž‰!ŠÐ]Åy-%©2÷«Û~W,|kŒÐ»Sö5–»‰¼sÉÜÜR žaÍÎ;¯“r,çCgŠ1¨èGÚIX_z´§Ò¥@ÉÒú†3= ©EZ°-º6xÝß /Þƒ»!ÃÖ·(ª t8 ×|`£ìàÚ]•¼·ír´)òQjG$ÄðÔZáÞ#ñkø£:¹Ë=à›5qÔ8Òô¼«ËŒ¾FY‹¹±YõÐÍtIÆ ;‰[’x¤iNE°ßUÊrRME»¯A§ƒ0\A雹´WÁFLVÏô&¢Î~ÌzÅñ~F8Æ©©<ÝêW“m‘ñ¶^²ƒ¹$‰Ô­¸EÆMfÿÁ4h˜r«øûsç€hòÚ—d^Ó«®ÏDìL‘´ï@|ÏC(öý‘v<Ô»'žݦ`뉣1ÊP’:¬³µi–NBíÅ×Ç{•cµ;'8_¯É¡=y8¹dËTÀÔ wM&Sý³ÃíÊsædùª.¡ÿ¹ÂºïîéwÆ ùú»ÚÄ/œØ­ÃhtÉàÿ{îוp÷dóvmêÓ&c.j‹sLZÖ÷hÔî…EE‚J(•e¼Œ=—1<Ÿ5U°e²‘U’OC_æÃhîŽÜ É­õD±ò!97(ú—û,î×ì¹%J@WH5ö-!ÿ/ºîÊZSãtgh[«åÀ4Éó>Ac#ÀƒÌ:ân† Åß#Jl+M>ߟß„'9ªÁ©²µ spü[ÄjO =“¡²¶3ÐË¥ÿ›Sê[ŠÓ¿,Õª^ùÃþ,uå‚nó/â \oäþãž?––,µ"U1^\¤8¬zÉîd¶aI'—y#,ܬ$‘éþšë–Ÿñ”ÑñhPÈ”ÕÖ·›4ÄóK8ËVau: {¢³‘9¾éOöŽSª›²wQH¼þ¥[Êàþžf¹w­M¾Wˆ1^éUÀ¬“U²Ášø¯c˜ÍƤ`¹€þüü Š¡ueÄ;Ý< ð+Ðh³±9ÎJßúµgŠ6-ld;/|Év2¦#M‹^ûžd¸e§{òÓ.4°—ýÙÅËfù… fItG Ž>tí!m‡4) )ò{(OODes,‡nçsÕ?†2JçZT|‚ RȪèþ(Ž…—£’‘V˜äE¨µjl[ÂYЕv’Ïæ^À˜±¢‹™©ä«Z±®Šì1Š8rû ¾¼_p°ÈWÞâÏÙÜ—F’ܬe±¶|.¹æ‡æÑ?ôñbÎÞ‰ôôYöú3›JÁIYf¶­ï bkn¥ž4ThàìÛzU–kQÞ›í÷êž'ík\‚’¦dÔ³¾OðÛKCW?î^NdåKÅÁìýTs豆¤ï+˃oúˆ<Û¯žÔ̼¬ZóØW¹V¬Šj9…€#>âP1s»p[»jù ³¹îN“Óé<¨Ã9³=à6jžz[û DóÅm,w¯žcç¦]ðï„šâ²W,ISý¯ó¤óë`s=A>ÊÊ,( )ÒûìtTïÏש è} [!Æ.­ˆ? 8„ƒÊ)ÉÀŽü~¾û|[Û‡ç+1À»jÑ#¨cÀOcÕWtr®J*Ö#ÏÕñÕ >ÒädTêç•8_î»,¸ÿð?M›#ÇÆتدùd’(€Ý×:ëÁ…çoÏC›8<âzlˆÁ%a‹|/zp·>UÃÜ]Ðå–E[x*54R)ã…Ø餎Ð]]-‘û" ëA»}ƒ@Bè6³Á“B±·3‰ˆŠ}ì g=¯E(œì©˜òSkíÉáåíßK¿ÉK¡Ø͸ܵ &,¾¯ã!•ûü¥£ã¡~€2.@»šö*ƒ-Ní•“IXÂ%Í­lë˜OêÕØÚ˜&X[ Åå+•V²‰ò§Xž0Ý,£ZN× rOqgOš¨_cð´NZ²ŒJ¬Äpg©- ýõù Goç#sÔšŽóóuþ ñ´\/†>ÍGV‚“öËWRa†ý+P¢Jœ'ñ‹Þ¢å²­sz˜ yÕâU3)2’\àìê¹'¨Q_ôÈx¾JÄdör's?û;ó[¸ô¹<øn iúh€*©}ƒŽçèÄ›óNÌ[ΚgÊV,Bo “ž¢Kw*Î:¬[ª¯ÓDœ© ^žb„*Ö {$½eTæ:9WVj Ý ̯ZÄ”H÷¼Ûe™Æ*1) _ôÓÂ#ºn€´µl—e#œ%_tL¯¡#_ÿ- |ÓC ·£áÔ×*l­œ['¬+rÑ¡û¥5ÚQ¡Ô×/ÁuÑñÛ{öâLÚ©‰n_2ÂM9>"ñbN7h~š~Oýòë|dão˜®ùþ¸>KÀö²KÑùöç¿•âÍB`á~)º>È°Ò«ÕOþÑ¡ú0èŠáIvTì!af—Á¯mÞ×̆ô¬|´&Ó+ZÝbJÞ*o1åa¬ª²VFöñw2 wNp•yYéò»ñ@ö^qý›G–ÔŠºJ?óNöú­ã—¯âŽ«Ã{ƒ°ÃKþºDÒÌ z_róéèY'™+jk²•lÄ–‘n<ñ¹JÇ/˜~ˆBNë<_mÆ«¡k•–À[ë];-Åi—^ ˆ û]ŒÕ,‰gŽËU¿fª|/CèɉÞäZ©É%Ùy¡q—·™‰ñ+²ÊÞäß ±)Ý2N˜‘—÷Êw½Ð¤Õ½»˜®´¡E‘™«GE„] ýÝ{~?íÓM ß˪±(Ì?‘Z´‚eüEÊ&lÖä€.Ùýº-üö+^:(Nú×ÛdwñÇ+Úú¥Ž©sËô:VHqÕû`´všpD•qÿW*CÍxH©€¯×X9"2ˆzÐÕ‹©à)eyl+›l=øüø”©Ô´Ùý³/Û8h¨ ¾"ãž(("&ié_hæ{õmTË€Tƒ ô©óÈ& N?dÎXã9w €Ìd'»×\'H”ݚìAè« ¹æÇHü[ƸänX˜ŠàÐ.G„ÀPÿÒØ!'UxŽ·š B#_IOh¬N‹Í&· ÿ´{…½äWs`~\ºÏ+Yï­< Ò`™btïÏ€ ^ø_=T­Ð;2Y‚€ƒªá×· RbÊa*®Ê¡h!(tÝì<{™ý0ç²>â®—¥ê?ÿ~ò‡¢}ÃcS¼¥Þ“ÔM†ùpo+gkücUÌ/ï‹L¹§î±;‚¬± 4Ê_p®ªÞ`º)ç¹V¿¸õ}WÃ>“6·õ„vÓÉ›´•øBo@ŸÏ€`zãÂ6k: ÔT’®îìЈ:'Ÿå@ Ž££uÐlXµ @‡sLZ²ÈV´~Aps0=qŒ—v„þu üf¨jÄDï Pº;z$E•€)ò.d¬–0¼L ¤áéëâàˆ2„òˆmãz iø |÷I´‘¬´šŽ<“f¤”áÎÌlÖ±ßVÒ®ºù£d®;°>ÁÅpî/ød9’aì+#dÂC£çŒ ²êßÆýýÑ þvIZ¨û@d+._AC5 ªj¥îŽb˜`*‡ñᄳ6éȇÂ,¦Pºž…ÄòØ‹$v?Æb_–4wm¡lŽIÍ ‚²Œ S½—UûRo_æšú#œGÏXw:ßïHÅt¤j>QÊEÆ‘ƒà d1§ ®à„Tü–V4A¸)q³4QY–!}[æ6¬ì’KÎi êZFyªÎªôp÷ÍYÞœÓÍÉP”BeŒ5gÛ×Wð‰QÌœPÁ1Q[ Ïá®ÉS¢ÞÉÈù4vï–Z©-Lü< DG·íð•-‡G“%(®ã²@@÷‹Ñô‹Öƒ’;ø„…4(x ä3ë0žµô?ƒŠ çâšþ3Ô^¿×Þ é7ÂÐÖp†ù"³f¯2³i×msr¹Ë˜W"©*÷ø µ Šjß)¡ê¾ö¾E†ðH•a–bÇ üj¬a­ý÷ÐfÁGÌoµÖ/µ Éñ í‹×‘)¨—ôJñ¶€“îŽßŽ{&Ë—÷‰„Ü?`ƶGc˜ H…¦QìÓSo› WŠÈ¸ú“’Ù£ëÈHÎ4®Î‰«ßX¾F;êÊÜuJzë#p5q\HŽ’uìkë·ˆH´ÛíhÄý´Æ(#ª„¤?Áà;ùÂ]ÛÂ,CO±}¸£°MÓÒ_v–E"ì¾Ó¢»†,' Ú\Ï|ºz<®Ðü–ógþ¼–*(é%uq5T¶ ávóÛ"jMt»ø¡Ÿ"Sk«öÌo…Éâ_h®·žÊOV«ÎÚîƒõ4ž1Ãê’,Îà³òFDc9cnûGj»Aü‰’Wð 3fû•÷¯ÛñœÊ]Þhº´Ë4$Y&\¾#nï<! ˜d²UGê+ä{D2Yx3ðÈõ­o&¨6â¡Ñšk&P™}6ƒˆ‚ëq”f9osði®ÂM.uøXmÒ8í_+%Çòˆ@"˜Ê‹?å"}×þ½ä¦ÔÁpW‡í•lD¨«ód'±ù”9@ãØíÂñ2ÈI¿Tû!ýÌ35ò¶[û¬«KíÚùÛe+¦W—ƒKL”š°M1¥.Bªª×Î 8· ä!‚Ù)úè:=wR¾þ¨eªi;Æ:§ÛÛð¶Ô¢*îU EÇ¥•G!©T[ZKúcNµFV„ ›ðöÇ'ÇIÙ ¤¤¸Âà´Pó­Þ㺠? £ÆYZžhß›í ‡ËÑÖÖ‘§ê¸"3奋Âzæ#´Š[ôÞ‚ž£[’øçƒ-ç÷ ¢ŽV1Û¨÷`¹ÖZchÍÖ®,eí.Æ ^˜:m$p1Ûœ~Tˆ\dt6±7o€©>’0r+fàççÊ0!î üª8»2Ù!’×BùIïËö¥!Î$VsŒXâýºÑH§xx`ÞÚdQ’-‘›™®”#ËCjÿ”|‡52³yÒ&_9í¼áæDt!}㵶:þÆ1s41Zö'‘ÁÖR¨îˆ˜*}ÓSÊwòöâÆU˜f=·‚s L—ç ]®·>ÿí±dí>LGÓýP´EzœG‹åáUþ)@¬œj½ ¶tE"v™WïyS ±ÛÓ(õݵ ½ ÜŒe¬'ü  yä&®i.tÝÐí“×ë2”Ý[Ä£“»ÉîD3}ç׊Šð‹w®†”úH% gº² _Ûƒá39]º¾Ô\~#¾‹Qz-på%Ó2ÍCÞ£bèŸIp%¡Q£óoW«1¶ñÖ¥µóäàÄ–ãtÑKú†#vƱœ‹¯ˆ™Óê2õ²Úï :Á=’R™²¥G§V<ÛN<ÜÉÃó£ZZ P4úKhº;c½z`¨à«¸?ý¤ª5ôЛ5K0aUÉA1QPß“9Ôª7Ò[íÙ3å!mQ,‡Û°‡1ö«×f$Å’ Ó›@àÙV à1’I&œ%lô“ÏvÎ1žuF4h‡ÉÃ6'‘PÉ´ÿ^t¹]gôÿm¹­¤ZË‚ÛB±\ƒ¹Žž>Ìé¥qÈ@Ÿin8‘ì‡íò"„mxŽó2Æ2örxí/µèIÖ„Fò»>¹Y9ØTgKM0fÿf®"P†T`‘:œjîÝEa§Õ ì÷Øôk¿·ÃÙö ,†ÞfBG# þ9b$‰_WÏ@“M®z¶v–>*wðŽ÷gÕ+ÂÒÏ,VÑB±Ä¶ê­y’5«ýÂqvz:±(µÖh¸ ç|Öaõ¿f÷òmtySa'îß—_²©°5ò6–%ÉÉŠFïòíkÄ€¬y÷£ÈÎ*(Ê*ìT¥hƒp{°¹š«(ðgÛþ Ö,$âƒê¸zÛÅâ¼½ó¾ÑWŸIfN›Éâˆ}v+sž¬¾ü¥Å؆máÀ[çµøÄZÞ»¥‰5\KÞ ãISæè˜Ë4mR.SÓWü²4–ÀF^ùBÛ¹ôhÉ·]è×Ù—%ñ°ãÇ}7Ó U…Õä•2›ltžÈ饔+ó³åHŽ‚ÓŒ=êhªrÇj,+ ‚Ðê¾Ôö …ëÏ;¡)ðéJ‚]o°ìLözGd÷H ?3[X©`¢QÛR]é±µ§êpîunŒWJõ4«²_Ÿ'Ýyõ…ñ"¬RŸQ8G¶2Uãu‚ tï²ø€ÿÅ•ƒ¦K‘ ÔÃ5;þ• gpò‹¢ádèÖÇ;@“xöÉû «´_ãutxs¾ŠnÍIS0É#¹!Í‚VmÙq’T½P˜Zã¯ÅÅ-é€MJv&( 2ôçk3äxˆ«².°ZÓÛ“63o^]É(y3å1;]²ÿq{"[3N´*r¸ó" ‹³ÐÁ ýqÃaÊè„Qð…¤×_Y"ш(¨eä@ÊPH«oÔtd]rý­NzטSújȆgÖŽa ¥¾fõð1¾ë”µ¬ø“JŸß¢×k”%lûŒŠy NϼÈzëå9Áã•0‡\„S ‹KM€nëdÂ`h„i×­“ñšáæ{r¸éà…z4YµÐØö…£‚?µ”j«Û2:^@³ôÝ“Åå„!¥ 4'ŠÆ.€v8¬œ‡B/,¬HÃçˬ¦X*¬ÂW®¯,†¯wÜH¥°Óã;h]Ð^/ÕS‘KFÿ¿aÃ1xaST¡4k$CV d‹.\b ðä4”žaž¦¢}€Ø懘-¦C¦Ó„q¨¨¹iêäq ‚Ó ;“³o o/Åzòz;k^¸µ©ÞìmÏ Êtë¯*ju‚Ù÷TÊ!9ôàéÛ,žéÕfS£í›¨P‚¤Ùºëqð»'$ÔlJ Ï›hÏ,Ž:3îПäcä­bE¿°ð¦j,n¥5¦†œ'ðrþ01ÅÚñ1!´1uÞè¾0ØSÄð–± ë<9Îa šJQ aËôe)z‘w¦„ÙV`I"d')»oç€xz W©­sX¬GI ¾¯¸!în‹âñÔŠ¸äEVp?`¨r1׌X¹™â“ïyIcpkK|xü—©tVº·jEg•‰QîÕL!œü˜ÑÜ“Lâºf*>‚(s/ß¹7±Ðe]9êlã•ÛFëq·©ßªýw‚Æ\Œ‘tŠäò·{²|¡`\{Û÷üzå…¿Í+¶o±'çoiW«¨#³]€JwVë½úÃppb¶[TîÍÛ/l0 ì¿1íEys;îêV\”;ÒïmÒ]X}Y¤Ýß%iwŒã Ó‡(\ج…nì´ªÐÑúl yãì.è§jηœüZ\†>ÙOÖt—6ÐvgV3º;e1›+¸Ùÿt¸€jSû2ž_ `2¶ß ¬%ª¾ M•G¦ÀgQ鱌\Çú5zõ¸ÄþViô}[ªõš1&Ȥ}ºYiB¦‡J¡¤¹ûî¼ÙV¸@×UÇ.eÒmƒ’§¨ *+»¢Q.»?ÚqƒGwMᴠȯœ]£žŽ»:èeê+Ò¥ƒd5Ý*v·ZÓËZ;ìºo‘䣹޽Š—C_™Y…Ôj]Å À¥ ˆ±Þ¸ HCÀ~k±îO¨0×U{EjkœK3Tö ™pMàÞþÍXZ¦º3™´°Ã ÜãÍ3ûêoà¾>"̇ôZ4ÂV’%'îébÌÒÚ±2‚>ò“x_¹…À=7„äY=#NÜáùkˆêäxË#ýð Ì5ÒÃÀ:‡ú’N'ìÌuÓä—3F Æõ4cì¥þ¢1âÚvßÔ¯P ½ÔÑyHÙÍ?ä ~y¢—A»l¢1¿ŒáNq>•õâ…ÉÕŸS¬¿Ô®¦zZEõ[ë·"&Æü[BŸO©ý±ƒiÈÚ%y}(‰ôÈu±x<ä#ôJ”™+ò_û^CÅpÎ*ª2úð…ùî!}@8G憛Ã>:ËèÌÆchAæx/ÂÍׯ¼€_M$°Üòè+€¤¬ŠÛ@ìS&"Ñ·–À™Aߊæl®\ ~"ÊœÒÑÜc_BµÂ‡Í;ÚX¨äߟO\W©®Yñ$©&á2Ó)Y¼>ØXYÑÓdKÀý–¼Qp"¨ª·ËKç @MþÀõö­i4íýA¬añ’;®T×ÿaqÈE¢˜DÉŸºÏ&û&i(ã #Ò¢çsÜó Ÿ”8´Â;€ØV„§õ…=ÛFwMzí`.Þ†Î-ì×5;¸Ó¿ZtñIž³âq2ø0‚QÅÑ» ÿH‹ûfM{qÎöŸ  žVG1Â&æL;{¡ ¨ÅZô 9¢³i³Yí—qfY¸NŠA»ïJ‰–ƒY•nWxO,©NÎ3•jó¿H»½Ôð¶Åô•ð?(*äcœ Î[åwè»éH,køUŸ¶Û|n-iq‘YÒæ‹Eäþx)ÕÒžµÜ<£ñüyUnµV#¿RœJ—‚ç[@/‡}K{ê^©<þÁ~ .6Ö&«ä¨AŽÕý†T”Ò¾q8±ÍzÑ¥²[]å™þÁ;±–†6 <¤<)ÙäO²¨anÖ?u¸„Üí°Êg–G³‹êÈþ^v«ÖÏ ÑÜ&sî@0O"GR¤‚s)D jÃ)’üÝÕ´÷áíÞ?5Ò­Ï«ÓmE5èß@“£z†uÀÿÁœjå‹÷â‘­ž˜pç^…¯à˜¹gkÅ/fŒï¡IªkۄΦ÷Ão~ùHDöz‚í#ÝúuÔ]ÒPKpyA û¹œÞCû+[™v’fýÓeo/Ë€on©ó"çï¦.|ÙØc˜ž¦Ÿ{£¨Š#e­…8«ü´²MÉ6˜‹ ™å}º|¿ÙˆáÉŒ¿…m¤ê¤[Ü”d¡ìsUµ}þ.Ÿ}Ï:]gCw…æ¿äYê‚Aa ½O©V#šˆÅúî1ÎÀèja‰d e(r™ê„ÜÍ}W:~Mƒ²‰~×Pé"G|WË%èó°ŠßXýµg,êžØåÌz0®Xu7¸/Üv°‰Ø—è~x#a>Ì­í8Hì¤Ì•Ârùi&ì‰wÞ%ñw%~çš3hÜ›¾Œµdë/×Ê ^uCÆè2ºú0Ušœ×hÉ|ÝøŽë+üù}J“ÐÐæDsXþCñxCÜçöPΉ§›ÔruCWçgKÝ£PU—ÆõÊ©é5ö±–6o8U‘Iîý&ˆ/,ª/?į&á·Û\Î*–É Âf#ÎGüjèÒ)Y>û°˜ŠÒõ,Fÿà™ìG²íŽ’‡ñ2öëœwl1„ƒèWËáÀv+4¾¬Ö(üÞ_µ5 n~8Š,úN¿ž÷è…Lt‡ÉqžGYÃÍ+j§¥Š#Á©úÄÈ¡­¼>õš^Ù—»Bï{S)à[ì#æ;ËP®£¼†Ÿ…ˆt/|U5$Í›1푯ÐáðŽrPÚ ék;šBgg§æ¨bt+ÅQFêZ\!öÅ™ÕÐX\ÑùnJþ“ƒJ?Ô"Fíx¿\+Nw–èSáÑJéʳ ›b'ù®“@à}Tóï¹Ø —m>«Rþ=’Œv .Þérà) gCé´1ã~1Íð€ØÎÆØXd$Õ¾fñU·û“<6ob%#ÃZ4šãZÊõvÇWðʃìíÙ{má¬ÿ™ –ÇÕX¢dâù„{9ª¥Z_ëVëÑ¿ýü,G93|2Õ—–E¼„æÏ(HÈ¥ê¯GPj–;ÄÚú¥Zð\žÚ—VIã®Ê½Í0ÈêõˆkæOÚ[Bä|%’¡*VùUéS7,¸ÙÍ©£·<–+æ6fôe¥è©æ†«£=õƯkÂÇd kkÎYþþÝ‚L,Má·%¶áafäqâ/0 e>$HBÇ«vç_¸¸;:u#~)ÔÒ”x,kÃeãEÊJ8å¼^¶ ó»vÞ9ÓvŠÇ­ª˜Ý÷·Cj×ОàV¨TPV1â¼Jœ^;¿*Á’2Iƒ´iígDF´qPŽãŽµB«Äß—¾Ý<•ÿNæÿnÙGéWMÉÀäJR7ö{*šyU!l|´‰)TŒ³$ã|Ty6w#€“’åêéR¼ºÇòi{û”ù^~âåLßÒÖMõU²xÔ){&ñm.–vµÏ\´ã÷®6hÚÊ]iLûãøA\™gO…†]Úî8L±K¬QµäeaÚýKŸ:ÜÝ“ÁjL˜¿Á‘iÇͲb+*‚ åº Ç¥‰Ö©ö,ª6_•j9ÈhÚcz«ëêåRs5ÞþTŸYL“Ãsߦ— há#Å=ŠÌÖò…Á )×u‡þî‡`â!òAaøÛÐMámëÇÿ¥ÒÄ’x’üRv™n3Q­4€>-«Õë@› ¯°ã]næ•Õ< }iTá:(“óvþ¶™,”glbËêÖÿ©>â12J®Éó÷aiäHææTLqô‹½³¹îÅõà~ÊŒ4gé%±6ÉPŽhtêÚOãT'˃Äççn—Ä"¬f V‡Þ…6š@ÏQ‰6Œ£=qAî`1~H€ÃT9—rÏAwÑmp6[<÷¦[ä£[%ñóx.ûëo¢ICÇBíÕý^ôú3*˜µ6+"[6l ºjNߨNöò›Ü- gW“PÚ¾¦ê5V¨ª;Ô°±|ºKúî ëxøå0° "/@ÛxU3+Òôš[ÑFPöJ¬áéÒÉ÷rcµ.T[E±%¶®Lë2Žyc ¡"—¢†ÚTŒtÏoÖ)ľ]»VÁã \£Ïá Ífå–¡ñ¦ã>ªî¤S“þÅS‚4ÐÆšêð²—–»æ[çš×õ°êÖ˜>¯xEc9ôy¿t"öÔ¸N|¢nj@Pr0Ïšù_'ž\±)/á†]2æ3àðÔ˜Ã'mÐ[1ô’¦Æ†"õ©Ó4^¨9¯ $ê‚d­Ð.ˆ–r Úl#EÝ/Æp_ZÁxʇÀ[qþâ˜/áNãxZbä€áuQ]®~Âsü]I ˆ,VûÖ3ó„ ,² ‹ Nf©z›wxyõ®}Žp¯ÈfqÂ?ä¢ÌÍ"áÌ#{ìïqmçèG¥"…‘ú™ðåï’’31gE툘{¼ Ma¨WZ¾5¥ö!Iá«.nèÁúìªÏŠ^ï¸õ‹lþ*É:k‡ rƒù {K'e gw¿ïÍêÎÇ+´”ÅUX,úïk×òI=AT l”´‘+.VgW¸i^ý˜á´â$Ò2S<­/ñŬ`cP1R‡µIýƧ*^V6pW’‹Çj¡ï– 1K‚¶wùÑ0>ó¶Å[Å­#áT;=Æ|Ç:95ŽCg5·%±%•Ð¾>4'¶I†b¯qÑRÓ #E_kW,±|ÞÆ}òˆÈµ{o¿4úŠ0q¾Ð‹üÝ•+Lâ3H½¥c”X  ËPqzÆÄXr—ÕW.Í5òîyTƒÓ”…ºt¯í„«ÜæôMPCQ'˜+}M1}(kj¡lþ&éqd)ö00ŽáÔËØVppV` „©bÅ#:ÌßGuGY¹nZÿF‘½[‡L (Hý„b§‡Ó"`rÕK¬˜Õ/íZH‚a ‡òÓbCUNc¨rP÷@AÁ“>¸7pì20¹ã-h•–N(Ä›X½¢ÉY¤Ð^À ,pï¶Õx;*W"¡õ}:|Lµ“WÊ(´÷V ¸6G_ÂI8.-ìÓ¤X*›YÅ/š¿É{‰ð{EÇë‹×=úñ'õl”ºÀ皩χ`‹&:¶ ÀloÑK¤àN:,\aL6`®á…šÛØä;•Oˆx=šýºñ,ÂSfŸN磲ešCqLl<¦¡©[–õk=Í«Ùã`ƒAäN‘!||“¬°³éœXm#ÁÇÈB­:ÖƱ?ÝþÁŠÀ‹úÅ»ûPG5u°ùƒXd=›ÌìÆz›Çm§‘$HRÜV׫J±©SÙä‡Âl³ëÂZPáÐfgJ»³è‰äß›c’žü£çƒýdÇK»^~Þ‰fcõ‘Æ3yacùÙû—¥/RÒ´ú˜È‚{?RE»¡ï¾ñýä*B4ƒ{föÙ‡ÍéIìy­Wgœn‘E!âÂDª4Í’:Ô0+’V£,ûÓ2À‹‚?ªÓä0´F<ú5ñ΃à5®ÈªÙ* RÌZ:")ø5ô@vb-nb,»¥†j¼3úû“Î^h™ËÜ–}ÒÔS¨ñ–´÷j+k¯×š‡9ª£_¥rævn‹„ŽU†Þi×J+ýÃõ"šÜf­ÄKéw ÇHx´ÕXt¤&ahRGe•¹"—l1U8tʘl{«ð=d@"Ì=¨RÍTSfêxÖïÚ}è ë\×ì"cqÏJ«›«¢M±á¸e! «qœ ˆVç¢`æw(¨)­“Wp¬í®céqPÒ?²DK‰ßºÖ‹ôËø'8qø€žãï©°aéÕâbxqtíC_°BD¯ØÆ3ùd†âj8Â\6?*PS—\Q×݃|àrîËDTI•²×R £(f³…(pB)ô…Ó«)†¢mlÐgß<€2㙵áúfÈÉ:Y…°™ ĉÍæLÿJ87/þªët ‰¸»‡µ5f'?ÞÀçÑ/ A3ýÖ·•ŒõÉПß:Åz5ÖÛ•’.· ù?ˮъÓ%¿L±1èìÿt¤¡Cú‡U1<9 –­Zú#vÄ“z»wíñÊá§pórð´íÊ…«#¥2\í†,Hz©ÿ~š!Ì‹xã ß”uÕ/ïì"cϵ^¸½…ƒiË eÁbŠ‰£ê=úÈžÓ×øTAšîÒ+û¢»ªlF˜D¢”®Eþñg 5ÇV¶“Ôëû®¹—Øœ£üÏëLn¬Ù • ©É¶#¯³Où}÷ù¡ vZ¡—h1l©]_Ÿï®?A߉άÕêþGTñìP%uú0Ôs‹X¦!(H“¡kË¥°yvIži 8´Ë5ršë¡h£›–;g¶`¿~Iõü™f¼ÔFq=AáiIå† ØîRo‘ÞCy’ˆ–0¥÷™!EØ­çÛ\߲ڌdzu½Tt1›¼9©’Œ‰èzÎ<µ¾pÏ-R¼T<Â] ç-»®¯ÂaÐs”ÑozÌçõu¢ïÆãŽùÐ2×`CÂA'ð>eÝ 3FWª¨™¿ýíš³¤ø¦ÌòõȆM× P¼a®#l%Þ ;ø¢Rb4ªˆÔ´×ùsõ¯ÔÅ¡ë§ÇèÅÛ8O£2XV)ô@›!Ö>F›gjM¤?¨Ÿ]æ©JI/¤Q¢¿E “¬–¯€ñ~…é–uÀµkáaž¥À!Ä€û’箊Fï)W¨£¥nfƒÉ*ŸåŽy™P©±)ܘDã×zk`123¢ß&xÝ' ñlZ!¢Zf†Q.rMxìr’Z ༠•6-&P”¢œµ„ÄvgÕ®ÇÜwŠôª¬EÅÓÜo¾Šºµ-^f8@rø©;Ž:AYíÕ[ÃGž¤¯N´®*>êY ¶ \hã¶&Ä(!UkCƒÏ}Ì£INH{F› þ^*Fl6|bqFO`ÔÙœvu!õœ§º/ξP’×ø ;‘û üêHGÂôV¼ç~ëm¬f…Ù”âÕGúù&½èÐ${SO…Üð†¾÷2QI¸¶µGÌÚvRÆfvIäñA"[Ê&hfïŸO^nÉN¾Xf•âøt äׇsicôŠúùCþgÜnnûtÏ­ô­°”òó(í/Ø´ÅX`9E÷ƶ93€À¢Æ}rHt²ÓsRÌ`ô oÞò2ƒ`‹[\@/FT5LŸ7%9Šb;³§¼nûe£‰Öb§!@â¼j¬Ó&Š6ËS9Y¹‚㩘äZ²Sø{;Îå—nÊCÿ…98~'¦‡­)ò‰|û‡c£…âC#”€[ùäú¡¯¢ñš]Ì8 Ôäq >+ÚÄÉ¿@ˆÝ CÖ–zkf$a‘çéH’Y\SàßHiEÀIþ à[ DØUç¾»­MPÛZª`‹ªgÓ&N´Gñ\`_)Å¿Q)HÄLj Xé#JlÆ°e…Û‹š!×ãYŸ1 ‘#ñ7ï+ü¯U*ð‚‡]†ML†Ð¤•¢ ­r÷KÀà)x¼*ˆœz".6±^q—›’Ú›pnþJúcxs^Üb9‹"rŸIb” í‰q±}O@wÛ–Á rp}ˆefÔÛ}GàˆWåžI.Xâ{™Áu›Ÿ´“ŠôXÔ:p X§ù®V…F¦àeÔ¤[-Ùé©bˆÛ¶´/¾MÂñQÜM;ÒãK ´Òm<¥Ó$gÜ«2ŒJYL‚îùW†ÞA…còiàÂýnJ°X„‰ 0Ä o±)+9êV¬šyuìôI‘o ³éŽØU¤ö‰ZøèNÌ£mUMª½5¸Kái„K}n ݸÏúXYà‡Êép|£o1è 9ÿì¼V–²õ+œHâËÍ"G'†Þ G7•'d¶öÚ)Xsë£Çì1[„q'Æ&¨˜£$tR_« Õ~¿ ž€6¼=poò:ð¨Ù­–G“WêBvzŒ„PfeÎõ›*‹/jrÑù²iHé…6/I¬E&žÃ »€FŹ…# Á³ã Ô"‘h¸¿å“^đ僗ûZh¤Æ„°ÖèDƒ€ ø©É `‚Fj«KóK°JïÍ?BÍmœ”ëÚ%·ò—Ln¼²1Bú…áhó±¢S’¦5jšá°Þð¸ÁdFUq6'ÒÀk‚8zï‡s°#\£ZEÝ \Ägƒã)¥ëÞqÎYÙÓ¢,–&œX÷ýAŸ«qÄó¯æ7eÓ(½‡® °E[˜–)ž64öð Oßy{„ä­aQAlü¾TJ;ˆ¨)(DÉ<Ðüºô‚þ¢Œ—ø·#ïƒÚææüSªj3–f›mÚ9C§bm{6­Ò•"ŽÉÀ• !˜ÑÔc×¾êÄìµy3“0«ÔãBgA‡JÁC’õŸc&fèó:¬UÈ\ÿå é±z¤iRßùŒQîŠF0ˆP:—6»,óÛ›€RÑšuÕ»!Î ³#û0„G;Yš5—ùª_(—Ø5Ô]”Ú»M´í/üxæÎtQGwB€Ms¾y=S±UwAr…ÅðV¹˜ø›¶¸+×åv4%ò›62·Çú c»p¤j=sÿÃaǤö‘܈b8Ù+h¬w+NËå‹jîæ—·S’³m%Y«|¤š¾õ^Ž&RùËžäÀ —IŒÁLÔ:† ·d†1›]G¡f¯¨ÀMÛeøÅøna¡^U¬$ÞéœRFo#öwéZÁâ8y« Ê[5¼ªqû½ÐFeƒ´ ë¯0ňYÄÐîùu:4ÍâVðS(ÍHKEVz¦a‡ÎNîPH‚Ý¡ïN¤&ePQÙÕ>U«äíhˆØ„{„ÿêýŒq',&Õíܳ`kiêâ-žº®áE1%…³¿Ë4D¿çsˆÎè;¢QÒËþôÌÏTÑDc}>ð”¨cQè{säxçµ@„l¸kµì"ÞÞüÝR¿m4¬¡Dq) Œ}mU¦ †ÆíÕ¤·RÙÏØ+X”ÚzÊ`ú6²̲Z0Iz ÆÅWögY‘†sØ¡,Í/GœbíE¾ƒgF/g@`ßGÉȦ Äâã±7ΆڔF„µ½ºŸ<øós,]–﨟Nšz»¹`àÇ›kã.cî¹±Ãü.½¿ "rT4:ÁŸR¨1ÞÆézÙ”…a0:½ÇŒº³àn:yá“tÌš·O×H’ÜM‘£pÉ®h+dØ;ˆ|dMqú™]1Ä”ùžPïüö:ýúHg—R»~<“½’è.´)01 òšŠ[Ò;¸8qFú>ûà- Q·hK9ñ ýåïÀ!¡FÜ!Ôˆš!bgõ\Ñ[¼6-•ë°9®aÞoúWp­n A,wžoÜèë¹ —•„Ü(Û–“4=¤××S‘`GØSwP5v ´}ú*ñ¯Þ§µÛÄc¼· e°r©Ö­¯ÕŠÄ¿ NfpýE‰ô¿Óvxù‚=|ÂÆ‹¨ãâæ˜÷вc¦—æ¼Òça ÅÏY3î)`\žYà0!äeí®ÍL ¯˜yÆu-”YPp¡fpÀ/}‹£;Ú^–%'@hÕ‚h³1½Uùýi¼fœ»×l¼FxM1ŠÀùPg=å‹ú{#Á)X¹{H“Žà* ‚ ƒp"<öæ D×j •Šœç5¦»¤9™)/Þ‹( è6Ã÷iãÃ…¡iÙøýÞ'`ûÄrU ´çÎÞJÿŽÿî^ÌY´Å‹]„1¢[ß«ëþXnò[jÂlßGé‘iöˆD>­² ôa% endstream endobj 62 0 obj << /Length1 721 /Length2 10315 /Length3 0 /Length 10912 /Filter /FlateDecode >> stream xÚmweP]ݲ-îÁew×ànw6ÝXpwwwwwwHîn!¸»kЛó{έWõjýi³æèÑ]5WS}–²³uV÷°³1±1³~¨I©*°ñ°óX™9¨¨ÄÁFΖv¶FÎà/-°)@ l`c°³²²!QÄíì=-Í-œ´&tÿšÖF¦–6–ŽM;k;WK €€«««ˆ«“ ³£‹óßCj`0ÀÙ 0³´Ä•”ud¥´ÒŠi°-ØÑÈ ìblmiZš€mÀt3;G€õ¿€‰­©å¿891ÿCÀÖìèü—˜™£ @AR]TJIQ )΢.0²5eþ·uvúò 6q´ùWuŒçÿZÖFÿµŒÿký7ëèþo‹‰ `jiâ 0›[Ú"±üK7Y[3;÷¿Ã¦.öÿIý%äô—€ö¯†tS°Ù_´‹µµ¢‘ @+ngcïâ v(Ø™‚mªv6F¶ÿAÙXZ{üqÿ‹u6ú«ƒ¨­ù_åXÿ²t’²t›*[:ÿ•ÚÌÈú¯DÿĵÀÿî‹ØÔÒÅæßçÁõøÛQS;[kÿ½òoÿ\È¢­¢© +Ëðßø'-ikbgjikPsþ«¥‘£éÿ¤•,ÿ==ÿ!óŸi°ýŸ¯`äìhé±2ÿk¼‹Ž@Rô·ÞŠÁ=ÏápJÊéûrõFÅÈZj†­ñ³¾-br}dtÆ%¬]ªÜnW‰ì0YÒoÅyˆ/ct M}˜ÃsÞç£ê f3Ùëe`C„z˜£ýke„nÕà”~¨yEBwI_„Èq»Ê«3\â¹3YÈ_Ð1ã´rd]N6ÒÒQžæK—д†D @ìÌýöAdŸ-àw.QAü±Nåµ0¶DÒsK% ]ã(¢'›Éû#$·Z/V¿P•Ø»™¢žñœ#$<‹ñÛô‡»r¨svZ¼¥€ÛgoÑøÀâ‰5S&k&h‡1þC¦'™í·#ôT&ÅECƳÎÛ3’¼¹.¢bátÀUP¾œE‡HAA ] OãÐ#½õÞAÅÜC¢±ÓýÚ½PJçy’æ!ÕsĽ~”:Èlz½ƒßàOÒ«½ð’5Tvä‚cV-åÏx*Ié™eƒ¸'d¢ª=J­ò¹ºpbà òë‘ ±™áF†C4äºà ©hRˆpžÎ‚á ÷àT­ª#69þ¾  jBµ&ÿ§Aûžçê(¬eŒm{½Àõø¤ü7{b§Ú4›Wq2ÈCÔlCPwmdpýªOV—Þy{»ZýЋ2„,|“‡RH5Ng68·®¢êsÞºGi(þS…!]ie×z4ɦ›ú§ºû{³‰%‚KèÕÁ%óB4¼Ô&´Bkù®úQKŒ¦Ëꇈ!39¯,¶sñe*ÊâŠ7‰–AcÍFc D;ëàQ¶_%V ñï n8dÃG#ÂzáIÎ^¦Z9 o£\÷ªŸØÌ»fìIÅ5úÉÒP„™ª2ïX­ËÓCŠ¢(£PøÖá³Ç2ÿ)ˆoâP«9OJø#Ø~V]xaF½+ǽ:‡¾“6<¤:ÎM_SΧçù\—ø«­EåÍ‹™6 |åp€˜–6“ß_ŸÎPÓˆ¬h^žŒ‰2m¦×¢|–Ó]…~/)p 2Ý´²‹R÷ƒÁœ«7–Çq#¦¡"T«Åh”-중ØRèVúÍûœ—ç´º"ÆÔ”ÈÓe+Öþf&tQ¡z…£ŒIVi`ñÃ-a à΢q ?ÇS§jèG'MVó鮋‹x§!èVØIÿ–õ«ò\£{ÓÊ5ÞNþºwTëGÁ<ú£ìMšúàgNIVHB댨¢IgÖ¬³¯LóæV‘ɶs1OcMÚžuÎR¡I‚¥ÂÉjœ~…2Ë7„·Ù†÷iù_³xö>ž—„ß29vDùEÎq“/Y» šÈ/j²ç²Da[½Ë¬‹: _ÓCŽ‹¥Ê·bÒgí ÛDVÞf[Û ³×ô§VB·]ÕÇê›yú@f•bå>Áp‚¾ªöw–ÂHê&ËwÕ©X][*Pʯ‡ú¸Ao²‡°&[ˆ‚g®SÛ§¬Áigý©gàmønNývKy±Ã¿oËÝä,œÄóZûë~©2|ÝûÅ@ƒˆù«9aÓ 7Þ¸«•y+B¹Öö ©Ý]­(Q( ¶sèÕøS¤–G©?3Ý*”½O†yyEçY—ýñD5¾>$¶˜>f¢Ÿ’f‘\(›>ÜfX¯6ÜY¿õ&=làé7 ¸dŸÒço2Á‡jJ7ŽÑAñÀ,.Tþ3-UªÜœõ|!b+4I>¾0Žƒ¾ÎŠx°–[­#/¨M¸gˆÄ±Œªµ|ˆ!¤§!ã%Ÿ0cün ›¾ËiéÎç÷˜œce~u8iÔV1“‚ËØB+áž7çï÷; G&]ºÀ†ìCâuãוdÂI;RÆí‘…̜♴€`ÎíÊ«â>ºjŽ[ÈÙüZÁΣ0*¯ñwÑÊMßOXÇqá¼×/Fåtø+B³\{V[– õ§@8ÏcV¤YþßT팑6OP>מ!ϵ›ŠrØG¿Ï6ˆÃLXD\å2däo3óK”HXR„Æ×u"sJ:Ï’lD'…’:Éf8E{Iß}çÒ³8$.j5]ÄBUèâÍ)É)÷tÓAÄEœ'â„J6Kjö!ì_—F%z#£¬¾¨_ŠÀ˜??ïä´/ìl¥IÏ Óôm×DW:&ë%'@†vuóm±¤X/÷„ÿ)Q¤?ŸµZdMöo\²mö€š û °S»þˆ‰E!ëþÖ{;0ãå”ñŸ° )ÝÓ²IR·‘ôw ÞhÛÖåÂÍÞ" h†˜Iܚˮ@éªáoaö‹<? TÝÏú‹)Y`8~Î÷*± $ÀÓ슞8a ÑRèÔ¼% íá³ Ù8ˆQªJ-u죡ìú j²Õ«:ìÃÓÒ;èyPVûwË2‡~ø,ͲRÆru’žÉîÍ}põµ€À)M˜®èæ99puN×LBulÍÁxl)ü\bbDY¼Ë6GÅ“.bŽFr;|ž,[ Ì1cèÜÌ ù’êQi¶‰hˆ$ë>‘XÄüÖÿ¢ØS5tŠ…ÛòôKçèEa®‹Í Úóª i<`‡í°#ÆÒvSCèÆçÛ§ñ{Óß™4²§F˼|\’ë;uì¿Vý ôÙ)Ã1 z÷¨ÂB‚Ð[ö‹®¨ŒpIvƒÐqLMN}Jšµ*¯½‘ã4fáA–™lEÚÄàKÿ$?§'ÞÙ¿ÛLx(ýy´?óWÜl;ý{ô@ô±üÈaÑR¢IŠ SrrâZé´ÕÊ“ßáhçœ|y±Fþæ2Ì6Jèê@ˆ–¨MÛqó8uæèþ¬í½”ÈÈF¦1Ðï§í:à3àzöÕ†ºøá@T‚žË<×{£9ìRòcçf“+…ª|¤ƒu¬ÉFçÂ-¸ –£3“ýKãÙ N¢è`¸¡ûí!S˜Q{Z gYØÌ“ç-ª‚«ŠÓW}ñáS⯂hçD¯ý³<"…¾ ÕÒɯZXÓ‚ irÐ\-kJj8áoÏPm"°(Yª"…–“ %õÃ#k×´þíkôr»Ašþᓾ…µoê.7çúÆŒ)1žå€4ÿ”>›y@¾†î•re¡hÁÍßTŒÀ½6!ÂŒJ®`M, )Gsÿü»XŒE7^BÌ$_6â ¥7i¤ynma,qÉ€˜c['·ŽiEo 7r,Ø#6KéVÐ÷Õù/º·0t’ RÄ1ò†xxd'’¬ÅU(û6›Ûc©tdV½OæïBǃä¶Aót¤žƒ>·Àú\6ôYSà~à©_sXPUÂ=Ì×]¨kŸ[¨öu¨“Ý:zœýÚýúçôô}ÔãnÁŠBˆN_ƒ ª‘̆Úb!_«Ñx‡¯‡W:såèxQîdN eEè1¯Šy×>œ+Ž|›<6j¨vR;ÇEÆ*3¦z}Ü°ÏןºÆg3DY` -y>qðä7¢¶)cy‡1 6BGè’RÂ+:F|­N$ÿôÏ-¯ÇK¤ý3%Žh4,£òg{DWÍh—©ôz§ßRžÊ¾Œg(Éy¢Çýú>Hзœ;š·6®É§.ÈÞƒ²XÛÎÒ™·Ëé<ûlZ»ÛH5ÅÊžËFm·AÁ°C€ú‘ò¬äh=¿èüZ4Õl¥ºŒÂÄŸ,u°Ö“Pý:å@tÀÑÃź&‰¨`áw¹Q–Ôí‡V>y·¤“‹j2ŸŒñ‡áÊrŸ¤ßstˆiX4©‚öG€nÑgbÏ rnèd]?Ý»É.Ý»J¥ ùýšù\ª‘òbÃeef©íÌY?Nü¾* N ì­—£†§]Ób"bV„Ò¼ÏwÇÜ[칡@Bÿ  Ì1¯Òräú¦hÅ#Ëû ÜÁ“a%˜è×–+6­­™Åç`¦Á%ÛnÅ“Ëu`ÎhÙs„0œÄÎwux”´ƒšˆgâQE€„£~37Í1GCÙ‹ØüË# _jË?ÝmK^ÞÙæjÜ9Ï@vár@xjh¤jyÉãn!{g\¦c4jŠLin¹ ³¶='(I÷+‘™7/ÉÔÍé&%±Ê‹*\ç\‰6ÅPy•‰ ¬ëfÌ#~w»¶ó/~„L†Ÿ¦*|·7¥ãLn>’Üe³¥rZà§9‘gC;Ÿ—.¦ø'¡a¤Ï³^dÁc®(ø®&XÎm ñ…ȇ¦®ÁŒdxõ´VeÝ¢ž–yL#CÕâ»+Q”R¼fpÝ=EèƒñÏ͹ëw ®£¾¿ mxܽh’i† E𿥎ýÉ7Ún/ÍR“©Òƒn™ùj iÖ‹&_ÂAŸ« Ùì3ïC³nþÃUá ]l±:•0¨¬¼eUkø€þš/I2x9xoÏ”–u÷K*ÁðŽŸŽv2Þ5[Îð…m…£GƒøÈÒ©]u8_æ®ëuI÷zAfùHe±]÷^gÊ"M,)±‹£Í_âºQm\FQ§¥G¡Š@«ãÉî vÊ 2¢Í`‘W,ÍÞ; r€§?.µ¼u¡P5–ͺ†–­šDíš^-tÍmœ ›E|ú¨^`°$íçü*¯G”ÌÅ;R¶gNU÷=ÕTÅÓ®“UèLbY:Ou§Wjˆê\“ø Pl}Ê“ Ì÷#{4_븵{¼ÏÝz¸ò‡'ÃÜ—Q„û6aºGÁÊպŕ6Pzo&ù¿?‡½í몦Å/˜YÎ27e§gÛp‹š­ÝÜéÚÐL¨F‰„V»sLäÝæO÷oùŽ6üÎ8‹£Î)ùuèņ·ŒB´V¡š<5ŽÄÏXý:,Beúäæø¯õò]OVÔû’^ïyÓje^ŒJë%ÆÖÐ×B¡][Ž.„wÕ[†7ÍŸ‚1ey ‡¤år9Ûyt{Ñqü~’?¢Ó:¬˜%½öÙåO›>Zd-^Y[áŽÊ[…}8ÞWAÚ ãCsœŒØ¥8÷}/ò*ò|øUD“]{þwÇ¿Zñ›,uýtó&ÌQ ëW} új‚„z“]ž":¦ò„[ðÃÜÄ µQ‘½}½œçÃQÍåï­orà¶ÞNU»„¥ƒ'˜5Ö¼PJ Cñ¸ë³Èn¶u;©'£+É(îѽ#œé–ÜòÈæE |”×ÕŒBÝ|†#íöåͯÖÔfÈ[‚¯ZáD!»:¥éø¢JIÑ·óFX9Uª| Œ…ÞÀê8^Ø×$>öÖË‹ú¨H˜éò,¹ÕÐgëlFø8qáÚ”á“$rÄ(£Ig@Þø ¯÷ÙmÄ)kÿø¤mir¸¡ÅÔ×-’iÔ ŠÛçƒY\Ûˆ…ØîLZä¸rN¹ãQIN™jTWñÉ]‰’$\Üo¿zp§Øx‘ÊKÍtÓôœ#¸‚Œ5/³Ù]èjó+¦w¬±D#ñŽ%ÄoúîhiŒçɶ1ïÙ¨|üG·îåºAãE†$¹'Zð.-ß).¡C¾4€=;ŸFòÄ_G†U¾v ýÆÜçÊ¥©èÂ!ç ¼üN¨Ç/'ìt€èÙA67n°ˆ€×Nz~‰õp’H:ôªèT¯1ÞGáá¬Ó¥ÛÊÕ¿Ýž!ó#ýN¯ý xµ'ÃöK³ÛÄ»’,¦ÏÉN•{&™G#mÜëDQ ®XÞ‡õ¢]¤ Aß5‡{zhåÁþ.5KRBþ9šm~rä-ŽÀ{Ÿ(cÁΆ[¨‘XLT/¿Ás‰Š…8 UjŽu§(ˆÄœ ‡#ô¯÷EmÜ%êÄï>­3Ÿ3=¬— .þÞöaZ…`½ O4›€•u>‘ß÷ðéÔº7f‚%òXºBâF“Œ‚d䡱l–ƒ+²¡eôO’´ƒ‹!ÂtœA,ÄjEÞø:¹¾¢2²à†Cg@ B¦3õe½¯sšæŠhý ÷ÉCü°93Õ›%ëÑ¥TÅ¥ÎB™tαÓÙëeȨÃóZÛîÁµà‚‘wndÌ·K©œjöɉ:†?óZʺfkîo²otF\S¾·(õʼn+˜âø×Ð@™–ºÏãw릅©épÕT¯'ÓÞ!â'¿GÊå¿®ZÄ6elb„¼"/>—¹Ë}ÃcøúˆÞþ{îâ1 CèMÍ& Tö™`NFŽ~ŵ((jÖË‘X¼Cf—É{DñÂÚ:$|Kœ ðº_Äò"ÍÃOa´Ûå­,«"¤ ®"Dvs3Mâ×ÕMØa;yóΖ+§ÔuÃôzFçΠpKØ AþP&ÚÊ­è3ü¦TµaJ4¥0I“<(îL[zþ '’Ú„)8†ùg»ÈýJ8|pÈLW˜vn ë;|ƒòè…ÐÑ$Kmq8«Úøqp@Hcܬƒ«[y³º«?k®¿r¬X.´ÿñ1Ì+ ]â ñõj‘Öv§¿Ü¼¾‚|>Càó©i•Ÿj’ !Ï$Jd$"ü_ÙÝO™] ÓîÝ´i§N7…ÔÒñÆaÇýu+ÊLx JèûqÍ’É|4Ä/¸á[p½3±õ9ËÕE–ÖуY&k¿áwó›*Af•Q ÞЕ&c…«ð¢´y+\üA)ßïQ>›Lq ñ'æGš4â4ÙEžÂwqÆš±Zï{õèe†ÆŽ*ÀÒI´ãø$h$V ïa »n1Ü«8ñ Ôc4k¿n}¬\G@¢®œ Fçd&‰¿ÁŒ“S^$SÌ…•]Û_é ÕŠ³œ Fâ$ Al‹é67€›–ª6 *ûSŒÿ(¯† ?¢$DI #½ÄF3Í‹ÐX¢„‚ß"C‰rº‚ç?‘¯TÚ4Â;¤È2¯áýU×:“Ñâ|-ŽÂt‹ƒ$4©tr$Û¤p_àfö†öµ‡Ð0zw9Èà}±$— z}„×6øäïà3ß´‰…ã¤Ý¤Sèò'`nctl÷O´Õ殢/ù°íÈÞŸ @J¥l†^D@PMûezIÇç¤|˜$ãb®‡q³›LÇÒ¾BÁÁßH%ãKædXÑ:‘ô\A'˘ûÐO˜!´„wäê`ÖNŠ}!Q/6®h9ã^Sˆ^¡ÐT‹C8bÒäEîÕeŽçtK#•É>ë~éXEã¼»ØI8;fw)Z·M/è þ]2öóó¼$ü¨8éB'u”ߨ‚÷j©A.ÝFS«¶‚÷\<ãÎB[sæFp7§CÃÖ׿‹…‡#‚ÆÈù0±ßøŠ_ž=b硨 ¦N^Ç™ÛåvÒ•ÿEü,þKeMÙQK6ÿwìÃd¥f¦ñë.2gn¡Ê…‘˜N´x9¦â/i+YÏ—ë» —¤é¿áy”`åX½–Ê13"QU6'Ûv/Å1ày>š· ŠD¶RË\é° ð¥«íþpƒa1DTÁìø“7 Y´çcèe¬Ÿ§íÝF;ãñ‘Š2XÜË_­i§Õ\ZêÉxBL2ÿ06µõ\#Þñ¾œÜƒ-«Þ$ÛÚ˜dr¬ EžÅØ4(B— Ñ­N8òøÓaÎÌáÑCüâH~¹Éfðu¡kò¤v–6@T ”±\ÿI†E!Z4ÖîJ(HF>¸l$í#7åÊ&I5n‡óZUž;6%¸fl~ÛžO<„+\²ó‡T, ň¨ŽÃ%§Œ‘L9 鎔µÖx3)ެЉ,íiR¬<š¾GU²­„œúB*ÜÍJÎ݃`ÇNKŸ«z®ë6ñÒ ×`Ùy.²xþ¸6sQµûÍ7'ŽˆJ}:¿qè^’»ÙYÖhšßÍaG åÈUg»â}N0ãy2_/]ÙH^Çh *óþ»f1¦0ÙÓ Áênì¨ sÕXCýœq©Õñãù3,'TaRøãÏâÉR–>Nä]/sAþSëóê{Gv…ªL/»ÂŒP¬mëi¯áÁR™o½æúC/¹sŽÍ!nŠf«¿®i0ÿüý«M8±3ºt­¶qítùyÓ‰”Úâó¸ƒ[¾q•À>„2x'¥&åñWÇïØŽº‹ëãoˆœ ºk;‚þ½V£:©6 ×CUÙ³áÝôrô ƒAN,ím¶Íõ)ŸVóXSm>cBº‹²zâšG¢õ-Ž¢Xv}1žÝ›¿°N#M‡wÒÍ[£Ä^7ÓŠ±oáò°ÃÑ[qR¢¦v_{Ÿ{,Î ÿ¹×boÇ%m‘»réù»ŸÓÅC÷e}î&Ûõ%r¢Ê¤°é´ÆcªJ(xœ3žD€D‡^F´Ç:›™˜u<Âëø9ôb¯ÒJ.ï1bJíË4c`ÀѤ<=i “³‹ÆvÂÔð ä4Á‡™—IUí³U9òøQ~V- ÊÂܧ$Ñü}¸œª, 6›ŽúÐrøzz"izW™ÈØ‚[eÔÐQñ/WIšÛÇçж~Ù¨ÁKüª¢Ÿô®v» zßÉ[ß'ô¦†®üE][‚”ùV¨`•å&Ï2ºTÏèÝ·ÛóŠ5[hhðûFYïhöö'#éš4ª”úžg«—€F§’X¥Ü‹þ)¯—%W@ËÂZnÃZÎDsdz77ÆÐt(môIxæg™kekøYEÅ©^#þ DÜN¼¡õõ¡OümDøŠ³ ˜£Bz;}&?xþîj²Ší¨ÙšS&?MË‘s[|ê=Ú%«ŽH×^ë]ùB|ýž®k5´Ï^ì'Ê\µ1ˆkàÒS¾‘‚$ (¿= ‚®ò½%Î&ùÔ‰A¤‹¬)2U”ï¿é_\ù‹Ü¾Œ2l¹›îõH5vü$\3¹ ÎF…†°-¿Ä>%¹Jx-¹¹‘¨]Àós–[#'E'Mž•ðUže)jrŠ,›þ¬îLXh$4^Z·cX·,BÖ7(WTy¾wÈF¤Ž;>Ç—$AÒî7N«j‹H*r-Õ¶ž8h\­Ô#›—rzÀ55Œê›¡Dí"Í°4S·¸iþYÚm¶n£•óK­ox …†è¨kÉÖ+µ."­lôÄÒFb’YêmLjÌæ¦Æs\dsBóöä‚æðëË`KZë«VhÍýÒ½¤+¤E;ÊÉúÝQ® àìå=ZvýËÑ0eï—Bͧlj瀯UmãâO~ø/Òý;²ÒœþMS™'ge_§å1Ž·E(3F˜}¨¿Añ,Oq•Œ ÌvØú ‰Â¯kƒK'>j‰ÞàU™ï‚—/]WGÄûqÜgxKg=7$O6Xð™Ôx‚Þ7Ç–C.ÊI^­V=‘ÃÕVØs@!Íß̇ŒóŠG;’1ˆ˜/‹i8ukžÄÙ#Œ%Ó{7[<÷åÐFYq¬îší*€fÁ»¶î";û•Ú4áXÈòíýN²—üÒQ½"ÜŠÇd:G!¾¨ÒzŸjP”h€µ½qÝM²·h.½…;õg›ýs-f‡à|(ºc1Ë!‚Í£Ò3òÅšÒá@U§&ËìÎ;ü†¨[×ufæ¸!Tå9ËT@f¢A­ÿÂÁÇ_°ÔW| rÊùp©ïŽMÔÌ¢N‡Û'd)9½v¡OÙŸàSGÅߟVìž>´5ÄODÖ[£!÷4]]ÞŽ|?ƒlס}6­\º™ýähÉo7Ρ@1>¿œƒG‰mÒˆ¸Ve¹‘öä.h\ÖÇaÆ¡àRK©[–þàÅ‹ùÉ°7j?°=ÅSg5KtöøYouY 2•‰?«3»^ª…\ÞÊihƒƒ÷e#Ô§âµl¥<ä‹ÌìMð§o+¶Eh+s¾2LáߧvÃ·Û W ¡DÂÎßË#|ù#˜aûŠ0¾£ÆŒ¤Þ,"3¯‘8¼5†LFÊǵæc ]GÙZ0"Ä@ì< ’Àë;¡¼ÄÇF¬¸‘UþxЛmf¦ƒG-Oîð+°Ôwg©Ÿãʘ`;ÇJPѯ¿)éeGξVP©¾¼Œ‘ 9º{ﳫÄGÅ°“Ÿ€Ä•|¼¼ ËÆö¸'“¹ïÖ)ðÜ/@:æÆÅ ~â £lAª‡ÇClä$e¤Nz¾‚ H娮‹³æ8Ó·šÛMÛ%¢·B }É´n×`rA·ò8¾Î_…l’“OutÆr]ýpÙgPžàÕÈ=Qú…úiÛìuò”4‹K¤â†G~¿®Ë³RÃOž­ó9À¢ Å[¹? ÄG2趷Ùßx™L ¶‰ziœÛ €ƒ×ÌÌC4û¹o`E¿>•<‚Ûz0h&ÏyÃN$ÔtÏ° ¯À{uE#òÑ4e&YñÐÿ³O¡#Ö úË—£M}èìõap/¸¨ #|ù{ý+ií„ÍôâÒÅtÕLdôX"Jâ·è¦ëôB~ð|ƃ­PKÕ±g/ÁÒW°o5(¬%Ñq %pÜ“ÇšÈ Tºh õWºí'V"¥mnÅ—Z#àx² ›ºÓ˨`Bá°lÖô?óPQãð¶~Yð‹8Qÿ*?¿‚˜Qÿ@¥I¼µUGÁŠ,*ï…3•*UþFSEø£ËW4ˆ›¢2'ÊÊõÛ9žÓF!:Óé ©Zþ|¯²4–ªŒù¢8Þ¼•°Uó{RûÅÜÝŒ9À”ߊ0¢¥=wmMqKŠmM?Q{F ‘p âíboI!ó !v„Î%óî l)z*õe. ‰ÿCÀ¢¤™øʦ'§;ºÌ'ÎÇ, `Ê…¨ýÂN2>-ùhg!§’Mþ4%3͘Âk”UD9ÂNÔ‹bÐý­DS¥•^êdŽÝDÍm‚.8¢×Nj„4Q’|ïƃu.²Çw$B䢹(SE«o¥&âG #¨½´{‘Vî&úÏ¡þÉŒ é¬O‡bžS€ ßV4©ÓO$-?¡”'wîºuï¢ÙâíÔ§æ×ùàJ£¾Ftß±¾•ÐRPûÁ¢'±›<㌥™n\ÿ–´Ÿk’ ùÈš,Y%÷]ü÷çkö¸ [~2ó€v”R\Ãe§dr?ygéòA¶Ms‡w³ W˜esXi„eÂ6¡²FPNöÏà‡2\·Õš¾l‘-bé¹ü ,÷Å$ @âz 7€¸þoSCEÓW“åÊšµZd~*§õÓ¡Ë-&N„/ŸAGºÿ˨ãˆíùîRTüIjö€§-ñ8ÈØÛÞ ôØ2¡KºFûšOìêo˜½o³ÍäEqþËÁjöa™6îÑ ÕÝËÕ‘xN”¹8I«ÿ]jôûR,RY²>Ÿ×„=Ü*Y.Û™dãý¶áS*ì@Ë‘Ó€hæúçó£d~‹a2–EFÒŒ~K¦'ÅUW<ønk¡÷,’•ŒóÁñqRÚõ–÷¤Ð,°á‘‰‹Ìô&V!/%¥o˜ñkzÞá;x‡ú4¼*üj‚Þ`ïÅ-pucó`RÆÅ]èi²"¶‚"jéôWÙ‡)_öeKõ'K[U³ Óa™îŸ(b—Wµ­÷;ÓuSô0ÈÙ»Ôò ç®0¥?ójÃVù$¢ë, c±e!1&qÕÿpˆÛ§î¢=½<;/}ãlåRëŽqÆ´«œ°º è˜þûð´I R¼¿A üÓv TÈí­¯t,9ý¶Gˆ…r’\í¡«^ĽXÌbý¾pˆ endstream endobj 64 0 obj << /Length1 727 /Length2 16330 /Length3 0 /Length 16873 /Filter /FlateDecode >> stream xÚl¸ctnͶ-ÛY±žØv²bÛ¶mÛZ±mcŶmÛ¶ûž½ïÙ·í¯}mþªª>zY5Z‘ŠÚÙ:+{Ø›0Ò2Ò1p”D•¤ tÌ0¤¤BŽ&Îv¶ÂÎ&\5c€’‰=€‘ ÀÄÀÀC ²³÷p´03wPQþËPµ60¶°±p¨ÚYÛ¹Z™~»ººò»:¹Ð9ºðÒý3HÉÄàln0µ°6ÉÉkHÈŠ(ÄdUb&¶&ŽÖyCk #€´…‘‰­“ %ÀÔÎ`ýo`dgklñ?˜œèþÀÖÕÄÑù`¦Žv6eQ9Ye€ˆ½²ÀÀÖ -þÏp[g'®¢MŒœ¬ÿ';€ó$kƒÿH†ÿ‘þãutÿ·DÃÈ0¶0rš˜YØÂÐÿo¶¦v¶›]ìÿ×õ §ð(þá`lbúO´‹µµ¬ €BÈÎÆÞÅÙÄ cglâh P²6°ýþÿÆØXX{üÿFþß gƒ˜°5û‡;ZN:¶›-œD-ÜMŒå-œÿ!ÜÔÀú¢þeW3ù÷îȘ[¸Øü{“Xùg_íl­=þï²ÿ$ó¯EéE5„”Åä¨ÿSÿr‹ØÙ[Øš”œÿaÔÀÑø?†¹å ,þ]C†ÿ7Ý¿tÆÿ§Ë8;Z¸´èþ§Hþçû_IçÿE Ú¹{Ñ2²2h™™Œ,¬NfvŸÿ‚käâèhbëü/ÚÿÉåõ•“‰‰»‰ÌÚ²w°ejsh¹¯Hál8õì´¾ õ2#ðÊT÷²Å ¹±Þáö½2vžÞL€!ÏîåŽÄ}‘A¬ ÜñÄda¸Î2ök1q…™oõÖ»Bþ[]¸î5˜*–‘,ØMÄsúúZ <Õ$KfÎyµ¤C)²c”~Y{8ÒvÜ꣭¥9J†æ7œÉ>õ_¤Yó‚kC¾×½5°¼ ]¸R,=ó‚òûÃ0{³ofXµCÛG°®Ô!\Éæúó"eŸa®ÆòÀ3Æ÷ßõÇè£Á_¢AÑz¿Ó‰ndJ!~èûId!5ÃEËn´sïªÈÒŽ·8ôY`–g©ÑÎ|« ?ããÆgóÞ`±ÅX ³Æ‚v8Ýèn<—š7H »Èă®tàÓào{šðL·üùÚ­E |¾°ä‹´ç¿Á¢àΕ«—aœ1oKƒ]@këãÒÍt„“´a§çجÐÂZ;_‰kÔÝÍÖ™¯ª…6º¤<:«ÛŒ$bŒ®(¦â*;–‡ùMc”ü]¸I&+Köˆ^Άɲ`$¡†.D¿ž©7†éúv òâù`—½Ç7ăô7¹$û‘˜81½º¸8øA57Í}÷œm\G6>Ý3Fú1=˜¼J¨ÐÙx’"&¼”_¨Û²Þ ÚÞ-Þ°6¿ü¦n:fÆ¢+èr°º´ÿ’ª2ÎÐÅ19¡uA±Žž{‘>T@)Ιæ{D)¦¾sŸI‡™¡±¼ €¹òº¤£ òf‹<•¢· ßò¨¯6êÑpÊTö÷³ Tƒ²Á£é SÔ$± uð]Ü ›âa1¾Ìêàóá $;œÁÉ©¢ªèy–]ǘ@S•|³´õžÍ¸ˆcùmZ¥1¼ç^ȪŽÖ]L Ÿµ'ϱÞ?Ïâî^g\Hq=›c¥á· ìÊ}­ðaÎÐ [’f¨­¡‰Ÿ!ôåöàÄÚ7z›vðF›c`0Ï“+ÒÅí 7dªXÄ×Ëš¼ÈíîÌ@ª˜Ÿ'X£ô.Ä&! Ƨí½áõÞ0³Q¤ÜþxxG–M7ÿecCÂÄÀÔM.íW²;æÍð¥c úC•^û<¯Gµ«J«‚JßÅhÕNù˜‰Œw»’SO,!×RÚÔ¢ -ó÷ pÖbµ·Ð"OWÍ^é€`}¼È[X!7É£]]Ôx¹ÖÕ†ÏI˜•L@ ûBówŽq–¢£É ï,xèÒ»ðœvÔŸYQsã8©%ÛÉ™£ûØ{~kð¾ÂZ§>‹ÄØ樚Ÿi<ð4ñó—`íPõÍe“g²VŽAl}ä^›ÕNÀz·õ²­?;¹ <Çê’@—OšO0 x q¡)Ox¦³¸Âã„åa’`XØ)‡úŠ_äC¿Ü!÷Þ:¡b>!Y=<´Rz “vXÊ~Tà}wŒS!áSƒâ¦”wʶbœáñýi.%ËO:÷×Ú”K:¹ÅŒŸYx¤ ©…n¦Ÿ˜ø—ÐèˆnïØUŠ×r‘ǵ†e<(‘~7Ê9Ô씜wU²ä9ç2x$Ðv 8’¿Jàµgí³î¡×<:øËpþ5­ÐÇIE.ý”yƒkÛ7vùýs  µäNŸ6“á·ôC­C¿Š$ygt µŒÔEt£ ¼§ÌôŸè:"—-«˜Ñ™kÆ·ºÜ².÷ ÀóˆIû8T^<É¡…¶ìL£ËÈ!õT¦82Eí)ש&`—ÿ*/ˆÿÌ]UoÈ·íA—ªCñ_/•ƒGP%~O½o¯ö¡Tµ2¶²šß€:ÃÖ9 Ð ô(+éð¯&ƒÁõµªÂŽ„šÔ[»051‚=üÕÕ«½bjð×.ÏÓôfm{…UœæW,Å¿RÕã´¶¿£¡èÞ£`*p¥0ZÆísü7ö·¯Oõaõ!ì,øɈÄ\ËbRÖ‹2_!¿ÕV¿´_+â78>5!nÜPÝ ´ ¬V•Y¨)ˆø6ClïÑã회fnx`:Ü 5<úÙòC­`Ct ¨ÝppŸ(×vs,ø©3á1–Î7¹@A”ÖvžgQ";ü¬ÆTÝ‘·øpòZùÞ5v‡rÍÃb~g£/U¾f:X@¬#’è…«ãÏxoŠ4M¬$¶Òæp{ÁÓf‰§(ÜK+x§ôÝßê«… rñ591ж»> ƒçmÒÒÚ«Sß·…}“8äZsq“jôývUYaŽ¿-Š"ôž¨T u¦ÕöÃAX™®bL:”*Îä. ¸¯÷TØÙy…¬³ì„Ÿð4YÊ1x–bª]id ,<Ó§Øs‡Xf¶§âç§l˜,¡“ÙFk™¸Ô­K!âÁ¢æow;éUhôùTÆÃþОþR©ñºeüCË×ð¡E8ì;‘Rð>Ï FfÕ,µH¡þ?86 Ze$rrûœ3n¿5à(Á ”ÄÄÓlîǘ|—»)˜íõ!xïW mB:&¯àÚ!Ò í.õŒj¶>ƒmÆåŽ&Šñb¾æÜ& f›üÇìé’r3 ÊCÁlOoÙîx<µÚ5±Ö·4seՙ잱*rc|8-,ˆÁîxi.v3jA¿X ooû÷Ì«ï 5ê WÅ&eÕ3ád73…ù5¡ÈžÌ9gXþÊ@Fûȶ#e‚¹äˆ  t1.ºøº¿+Ñ0ðˆ¸eëöØýá›^š™Ø-£´‹íàjÕrA'ìb©±°‹N ð"1f·Õ VŽ§T7“/6rs H?/°ÝÉø’ªFXKå¡z'VŠÔEÁS†³ÃG}½/QÒGM­Ð´ŸŽ.Uª‹K¬Ï±QO-&s „´Ù ×Î-uW²â»u ìvõÑr‹¹“r"[Ò¦KX4yp2`TÓ²|ˆÊ÷6ýaF3,_sí ¹³÷Å&âöà‚i¦§›i/X3ÉÍ”>¹d?å+/.ó³à7,]<Öa±¿¸•ú÷±ÙÔ¶Õ[ÕÉÚ˜ òaSÐx ƒµKá÷»¾·}X­0º#}jþL®<ÛAkµÕiÛŸ^«°‚äP¿†QlO.Áôa¹5o¥ö3$Žn:jVy>±¿2xG?¶´ÅÒ„÷;?*™­S_ö € PMˆ”5ÌVÖE$øìéùªËÖ8D¦¨Êò'¦–é.;£Ylk¸N?ܳ\_Ÿ8šñnV˸Òüœ]–]«ªSØ+áñß ¸ ^61=Åá‰AëFÀr¹Ü`D–D= ³xyÕÎu)ÇS{­z®ö{·Í|S1ò’ôüý Àu&$õùnØò<§ò¢%…Fä šó0¼ÍÒ?¼±š;Ü\îB¨i4ÆziÊΪŠjGݼìºÔbM*ªÉ‡Ñ”^jíUòà,&ýÉÀg´ä‚k€Æ¯–îÉd"kÞÔVôjl26MªOîýI ô Û<»ö~O€Špšò÷ ƒae&’-ŒœÛ´ø(ä×w´r”Pr®üœc0urKQE­¿&Ê’£#Ž^’W?S«J¿Ûî+o3íKKÔé´l¸¸ä|E[Åë*_$hé{L1Œ¸ió†gö!maÀûÉP/úq(/M')C€ P¸;BØÇ`$ñ°w›2t€‹½±³§Üª¼x–ß“}+n¤/Î/à"ŸÜ-åôÔ A(ùÆË]ÁD“¢Kè¡•oPšŽçgúõ§ûòªèKžgþ…ÁÞRqA›nu¤j%a-ss_ÃJõ äy fz³EïúÜÁ[Žà%z;QAt›±Ë±yê–ܯ*Ê æÏv3ôª–xgÜ3„™i¬O˜þÉŽ•2qð»—™íû'JŠñèÊEά7‰ä8Ï‚¥H V¯ù]%ƒ‚Œ•E XzŽD~ï `ýG ûAHàb6Ä+æ­ôc3¦~ì1 n“>q64†žú™ÐÏ·âL ¸Æn雿‚9ê@ûv¾ZŸæ™ñ`µp–öfmTÛÂò×5íµ*ˆ+“aSòjm;57Æß¿=ò¿SŽn†½ÅY‰ùðú´v¥ízð¹]® -·ÉoPvSÂõùˆÌ08výL~æB‹º³ ŒdÒƒ &3†’ýi,É,¥=oRÃw2TT꘱ÆsÉÑl{®Ìä`¯|Zæ¥xfbÚS’ÕcÀ?Åðeá8ÐMR¢Ð~úmÚÄvvXa¹/øµnŽê¥Dn@>À€oÅþ}¢I‚ñšFçjÙ…è…—3ÿí› ÁÁbYp¯$.z?…éeÕHûÖ¼<ê;Ôæ'WбŸð"ÎÒê¢ÔÇn!ÏüŸ—97¤ÚµžEFË‘F´0¯µ AÀÙÙ_ƒå,¯3.d~×Ò ŠÓlÂXÜc?§¿.óM!CƒÆªm‰?².Ó_35PMÀ¹ÖUû0É6"Œ ä¿D9^/E!eãn9µ q±Ê#óæp.£LñçE8hü˜û¾æfmÁxq\R©}ÞÓjʘôñ´½Yvò7Íbáì´¸‘x$rú©íʧºÄÔ9¹M`æ)¬ÙcíÁúÈë’jûB\eåto|è¼ÖGy)†AqPãšÁŽ ì@˜Ë? O…_kŸ<¯[ZùP  µTep#hä­×æ’xxŽ ÒcþÈ}ä¤Vÿüܘ,[– ½mSa*Bc½µxwG#Í,P%íÝqõ±l´8YL>dã\Ÿá½îošŒ/ÒwkÖ,PÓ¢~Û&8¸ø^à7Øä—yR‡•»™à¼KY,p5þ¹è Ú­-™Uv»ú–¢¥ß|™D-R(Ó™>.#Œâ´ÒSÉŽq>ªQÒhê‹ÖÖ÷^TÇœõRèÞÑ@d³µ kÊÖcœ/!%„¤¨8o†¤>ö^‹qÝßèbÿéšWIGlT…Þ‡N:ÒˆAþB9±q=b³þV@=M¯@ÄoÐ4U‚9ÐñN³] JÄÂFcéVÁq±AÐõxêV¥ ÀôwLòt» ™ØíÄšð´slvÈÆM-cõ^N;q_Æ@yNdxþ~À¾Ü¬'kkh› YïE]+7*‹¯}LÖÃ×2?Ýúy½ê Ê›ãp]îÜêbBÁ`EÐ ½:- Å…žb_?¼°ÎG\ü"ÝÑEMŸÏ×ÔÛ}[ŠÁ1%Ýh©Kœ?|ˆD¼–pHä ¥< æÒM̱ð¯3ÒÑv‘_ØU=bŽh¦¥8Pî7ÕK²sofÊ!²‡GJžµÅ‘Ãí'Ñ´K‘4*ì¯G6³U°÷M!óH<WKÚ1_ÎΈ/ýô­Ò5Qôó ²ýh# v&nÇžqüËA™ÎEÕ}Iþ»1m¬ùKîsÌ5Ô>£RÉ~I_Ñ…3³×]¼çþº08%ƒmÖÙ:èÜAuîí6Ê{š Ö2Ñð éy_½ÝG’oªs)ÁüZ˜™ª9…AÇÈýZ… ´6Wí¼sµ>Å?¡`I¦©hŽ ×1så@€„Æ"K3D)Nœ*þ›o$…ÿ˜@XqqÌý"®ºtbÉôÜ¡àÈ}£2Ån‰ð׺£ÐúF1ÄcÛa„ÍÖ×P¯gÎò a`‚}xjÙ|S&u^•XÆê‘×Ï70ÕÌ LŽbÕ Xã‘)pÔ<ÆqÎej6_&ä'Ž™ÑÅ•ÔOþ«WÁEÝ&êfycÃuš ÆKü‘ÌꜞCz˜ä–5¿Ö:‘Ò¨9s&j¤þ´ËúÝ<Œµ!Ÿ´Zàý¥±=0Ëu«~ ÜTÇ’žæÉâ1`ÿ`Ù·t¯ÀÙ3çOÅkñ©*~¡ ,7êD\·ê—›ž«$Ah”x:°|›lîöFy¡CÕ&Y®ç—Uëð™ÓGöJ °.TÎ5jƒVÈËEÞÏ<}Y"-)% œ¹48éÚåý/¾yøÔå_" íjåëßÔ’•ß³|­#†ëÄÖ\_wÃ.Zà€Ò[ˆK1Éô¦Ú§í)‰C ÂäcýD×ÜŽ¢ì•a~l¸ üžža •— A4©òéH›¤œi?ÅC0káj}¬UЬFƒu® è­›Ø^ G±·³ºoˆý‚eyBkÈ5—Ü‘âÄlQÖxTÇ›£’HÔ|ClR‰øeÒsÀ°Ç?å ód©i“ªôÓÔñØ !3}°?«ßØR«6)—ÉÅ%@£½¾ÜÃJÂ@èÛ0÷:¥#Û'¥);kÐbý±ˆée@¿¹ÀfjÇ ¦v_â Ùï€ö@Šnª‰ØÚ$½X VÁÖžNøðæçșҪ’¤uœ>ÞpQLgÂï éÂ8ýÈÌ:úÂl•oTú“Sd$6ü≃>Já$/Á?£a h,C¯—±ñõÛÛYp ë³e¨ x[Jƒ ¶9Ö£2ýDÏ_¾"߉әkr×¹U ¦{¤(HàŠ¯ñþ‘zóMŸÐ–¸°}5*CIžJR›C86¹÷ìð#³e[Ç•#ÌùÒ£ìØ ÷y$DB>ËQ¡,q‹o×"8=™Î  ©\2 ÕH8šÀe–à½Ú„禯ým¾¼med_¨L×+kZCø¤ ùä”y˜t—ªNãÈ°£Žâ ôGHtNCblTéyÑÆP·´âÑKS4[vzŸÝ¿83Âɘ¼Í µ[#]VúÈN`V÷Ò˜yJèR-AýëA¾³o5²‚üçÝ@<‹>è“üŠwdX5DrCòŠ Û’ÉU,Òé.‰¬SiÍKÀ¹­1«„Ïæ>‹J³àÀW„MÓrü²â:ÀNl¿\EÅ"¶3½¡_\ 7? p' “ ¥P ù‡Gz;êËWÚ–¸%ˆù#Y“AÆbw݈®Wi* kwE Îï]W¢…½Ü„NüÖ+ËéÝDc¼í*d.5¼s¡ð‡ÞzÆ3&Ùæ&&Å ÝSÀà§ìu·xr¢Ð®RT$Ìû‚¸ŽcŠÆc¤„•SËE•Hшb~¹ò +Ïk #±¶£ ãõ,â†UÔx&˜„d—òè/á±¥ª*m Ñfnªj¥ô묺ßíãQ—Î5× «yKhµHÞ;Û$Åß2óKP•#ªË°0K‡l#ò^ ÿFt†ÖIG?ôé²c»—÷ˆ`ó¼${ˆ è[ô³ýL"Ú¸º”äœyå4ö⿉,‘:ÛÐ<¤ØèÞ›ùl9¶5#»à½21ÆÒS÷Óìhk^…€/ÜS—Î÷Beñ˜‡?AncÖBx$ó°§Û&ïíw8/êÍ•¼¥F9ž†&Š–h¨'0Þ@?B7׃W0]z;ï¬Àöpé4«Ã@ö¬)÷û†/ã^¢ŠŸ5iY…ïK&ÓÈñÆ' >ònÆÈéã½h¡E–©ÐÍ$ä—¤79ô\´ŒÓ5:z¢J /P„jÏ›‘;5‡!’M×óãL›pé÷qb¸¾‘¨÷YûÐ'y_z*„“xø´”ÌxéæŸðbérȃ"Ûï0á9^1a6ÕS5eé4ˆv¼áxZþ"ˆ±@lL/ÔÅò}„E Ã[åŸþÔ0–àÞ[äsc'Σ9ÅOï>Ú— Gç± 3[QNõÑqØ»Hy†‰õ‡­g2´‘/¯³Ù(Åtv²îJD×{/Hù9/­Ý~£§±Ñ©´Š™s#ûœI”"v¶@§rj¨ÛoÍ1›° °O¥}vPX2zí€ë2ŒÓô€¾ý%¬¿YÓÏ:ñœ•¾çPõGw5™wu6òÓ#Žcnüê²Nr4{Rò6ƒY‰QÊe jXg©è±F×¾ŽÏκ£¯”•`Ž%ôv™òþC±¤C«Ît´ÅVÐg- ãÝ°í˜ñÕ÷…´aåkCx¢ÿë°¯‡xZäªctNìµ¼ÂfÜáXíjÕY?)º*÷Øß©É`µ‰7ª=Áó­¤ÅÒœCè®d¹Î}J½g«¤"þðÞ©ÅQ9­hæ½Fˆ7Sµ ¯a,yvMí ES±‘-š€¬DÔù Ÿ'`ˆåƒrQ¼.uzE¶i*¯ÔŽ¤Ê’Pqšô¤¦`¬$dÝcJ(€8ÇŸIHÒúâã÷„”bênMËë½â¥½¢»‡0pÕ×ÓýI¤¹eÄÁà€ªsê$Ø刧ßèq©Ñ<¿´0›ŽÈ œ®Â·ì¦õNËì2|ôò×ÆŠ”R^ÀîšÎ¹søÛÓ‰9^¦h =X®Óy›6]ÿÖ/ÞhãÈÇ8wT9›Õlþä³%‰ç”ÝÈyÕ/<Û©ß>„Í*•æˆoóp§SŒîà]áå|*í©=§6ÆV9ÙB‹TúàDÙ¿¸Õù¬úS ÓnÚbSÈEø”­ˆ0 µ¦Âh`2 õÇüRÉÞŸ{‡óXÃ1ýÃmãþsá¡‘€ÊVÎþxú@¶w/è\3´*`™b¨d)ƒˆÐU2ì¾ Y½{9}‰4wß%íÝE¼•þØþ¤–0ÆlàïlDMm‹´dxíuÓ¶NW¬*±cúü}ani±ïxÀΤÔÄ®T-HEùlBËn°,>ëÈ©Å®å0Ëm¹%jʉõ¶¥…ÆÄF“3¤…Óçå7-Cn¤C['Fvîõ¼TXÓZ `·p‚el¹­åJËEËð¢'û?]* ¹œ.¨·*Áã~.¿€­ÿèß>ñ†’K°'Ñu¥eñ­‚8eæ€é¡å yM[Ý€7À`;f’# Ö7´Jëübšñ6A³‹XÙyHkÌ,A“ÁA:løEø.z tÀ°Ì˜H¨~»ç+ ‡"néª^¬qK©PiUÀž9ß“i™Ny—6©‡Ea½—Û‚8Z}ÞoíŽ}§.÷¦S ¾2A…4šŸÔÃ1 –¥ ÿHW‡n_ G¡*"Ç5Ñ›‚´Î€˜š¼¡–;UÐ\Ã…*"‘Ü 3Eåº>Õ%RZŒæh2Σ±è=äj ’<5÷©g¡ KÃÐÿÊ}Ý%}~eea* ’oéÛHýÇ­Ü–“×çð<…—Ü|0M_ÃV¸Cá’Þð}Cò¾öTÿ­æ‡°dŽçfßC/謭«Åñ²g«”ÐÝòê ý4c/¸‹5MsdMÚÌgLø›÷7ëœ)JÃùDñƒr±f‰ú1ºß—*h[yµuCR ý¢ã‡7SÉÁÁu Ûfù@•%J{ !ï7‡³ ?nå þÄÓ“$úUB•K¹>!éX„„pO¢KÝýR–mŒà8Rê¢Ì´ûe œáxŸ0Ä£ñiÂ×ãçºÃê€<ŸZâzüw$Û™6Øóÿn,úC™axU°ÂTŒ1 qÿ¶›íŸmF°ã|ÑÄ C"Žd™%¶üzvœC csÛij(šB-Óô÷t–÷JÀ@Ø åž– üÆšþ rª µ²É••ejÒ+Q¡Q`þSÈSQ Q,³¥xñËÓ…¡bêÖí<hÛ >Cfß3¶6On¡¸ Ý;U‡¥L™y 1ˆ+Àï¿Ÿdœ\Xpv­kïß„9§<8 ÊtÐÃ)†ª;þMõ,ˆXë‹ä={;þyáíþ‰õ°ŸÖ¸ûi©ÃÕáïuÅĽ»â‚é›i®%ÆÈ \û^¦€ áhaë&˜\s¶_áWœ5*Ò/íèÀÓ›ÓFË †ðw6ÐsM{ ‚ÅQÍÖ7t_OUWä˜JÖ8.<{m£ûþ†µ0jv¶Gž£=å©}`÷ãļЗ鬵%³6ÒKr”Æ‘=¿—…!vríëó%ÒË8ı?âš­í=ÕN9®›ÍKÛl9¨¹ß˜g ‚}KQfÐs„ 1âUbr›e,Oȼñ'ãîeÔI©>*ÒIVî£& »V8¿^‘hËœ6ƒû&ž¥Æu±Òľh™5x†Mix ¢xò„àñ$ª§ &ìƒâšC=‘•Ö¢šf¿Þ°&D׃WàÁT¯¡ÎXÄB*_k y[!ñ1½ÝR;Ò´QÌDúªI¤€#h΃-âñ×áû`¬½¡ÖéŠsøÀtÝv#]œ³(RMÇ>xÀ«¤,üû¦B>6k§† \@$Q_ðJIðª7mÃç\¬’I4@†É\#̈סŠ½±ÄÙÖkð?zª H?³·Ø«Xâ£6®Ú_ˆHvR×òÒå’gÍK5gðÖG `¢@ÝÞ ¢E¦À ßÞ<¡ÛEC›ÿ0á*n?ýí»(üb·3î.e•ëØdR#?Ä.Ý@’x9?ñè®Q ªT‹’˜¸TÓL-5`9™æ“ßyâr–É®óm„Ùпßb, ¼m(šôΑö‘ÈÈV‹.Yçhq£p3ÈLj¬§ïÚB-ÝâÑ*RyŠ¿ùY{'·€ô7æ|ÕÚF®×V­ó m,9[˜-‹HL­—¤©›Ò¬›X†“§“}­ ¢ùTZ× þã€frc¤«g†'þÅŠTÊl¹€€²‹ùXÞhÇ)WóŒTFQQì_]¢ŠêxCÒéW¯­ „„øBøzî¿¡ðÙUsÄýô”[/ŒÄD¤þ˜þ5}ž «ëoþ†I3‡f‘ey±r?˜+~^¤Yø‰M©jh¬¼=¸IgÅ×é.#§2(ê7ýÌªÖ [aÁIõ6z~ÕÌýŸM^W hPEƒ ¶« pÙ|º§tŸÇJãç…G²¦ ¢CÆØ,!w'ºùžýÌIYÙISçû®aa†\V°èUñÞœ˜okš ÎO!6µ´Ì67U.Ú»WÞ ‚F¤@öte¦‘ìÊêÙKt§£µêí|_ÆaÓ¯™zÂ>CŸ‡íèP¹bÜidºí–Lõ×ð“eé8A¾XÓŒÔÅÒ7 ìFýÇ›x¹ìcžL@7¥…e¾·aNäï2dWÛ'*³ ƒ:yy4è7îqÈÿ6Ê™*–4æe´Cb&n¯¬JÃ_ȪLªd6ÄGöð,#V/ð’sQðž’¯>_åáQ vlAtøê3ðÖo½ÒršA–óC»ñÍRø•ÑyÛø±f‘le-×_þø“:g‰z{­PyœJÝW\ÇJå‡ bÇɶm×ÖR8wƒA#Rsü$ô¨w—žQâ?…Þ¼ìyÚjE¿i*nÝøUº kcs7GNk!0<®Âoåz5,Ö(aµo§VØ:@@Ì—nj[7.®ñÃÛ®U3JZƒ˜€z™ÿÏø‰ÔÉkÔÖ,ëKJöúƒš “»Ã¹ÿËËW!8ozÚ)ça {ü/¥]2I]"IP9 PÉ»®Ä¶`”ÍüpXz+ÊÚ =öÐ,³ ½Ó—'ˆ×ÄŽuBy|ÎRó¢ý„u®‡0-©äËlª° X-!’-ᥢ£æZ™Û Ô©;j±Åš!ÿáËK&5‹EšÞ¯LÉôD Qö"9v×MG”šêôzÎ8‘’ú½ç®g]éB) ZÁ¸0B:çÉ÷êÉ-óí’iÔËPBÜí°=`b¿)åÁ÷7¢(N§8Ÿ¹Mÿ+@$ 3!·ùƒ¤2ÊÇíªµ­/aHN{^ž‚¢7,4âúŠÜ®¼ñsXË`1̘Ê[÷h­cY‰þÅP¤¯+ ϲTç#û†ÒÜ »~CÕ€è>–ùzÞM}"âaú§ú;w’9Aîî9”Jj&?ýˆÿÌ ù¼hŵÔ` t*·,1q>]Ñz>ºïÉj I‚Á|bD\Yýð ÜU߇%¶šýß5ÔT} žlMrNJ_jòÚa:eØÏ”ØÄYØ# Þ*”êîI¦‹/.[t®ääX?Ü”¥ŠÅJgp!ù߉$™áØØÆ©‚YÊô¯ö~h¥ÐGJþ]Ù„kÕQHûþ©Ã,…@W(î2bñî&ñìµ×/ã;äHC°Ô´RaÈS©Æ½lº"£›¯ëïIûÓëocÜ2Èz(î•Rˆ—{$ñsð‚%+k^Φc‘àN Ëï‡lnúF³iÖò@½ bÝ{rKü°¡üZ#>[_öo<…–„>”fEƒQI`4×o¹q–øC³“=ës`ä[%›hÍšÉ`kU+0R5µŸÐëfF$Y1ÄÏŸ1û  ^Ô2Ÿ¹+¥žAAÚHj®ñDñèêf Ìt“† † ¾RLåmÉÛêE{;´ šû¯Ì>ŽT1äÎ%Öç%™[xa¥¢+ÖbÎM¨½Zɧ° wcö [6›BÅdU7«®µž¼§…‡_O[µÜ (d™ËŠVéy¥\×ééf9“ýÂퟆâŠÁåÙ¯¶V³¨ŒÄJ’ î76¿š«vV.¥¬0(¿P¨ëéÑjˆ;ôYÎí—.´Ž§5Eó±xåüôi:¥J;RIrðê¡À{H?Æ4ô½3«ËµÅD4çm ƒjMÏ,zgÏ™^DS kÛÌïð= /ÿÔ)+§)æÝrmÆ2wag50)ço'Ö„“K¯Ü«#³ÔÏuåÏ,Q·[”UYùzñ§ÙfèñüÎ(têqŽâÄ÷öéœDÁ†ª™Œ˜U§¢‰ ¤`Ã[z%òo–ŸÌw3‡bè*…4Deþ5¡@ÚD¤ÊÆÙ›¿ofÏUnöu~»\¢eö`‹qˆ‘óé`‚°]~9%"-âñNoæDt9Ûvóf+y$ŠY[†5ëP‡«)x†ìâ؉| ŒóÒš{脪¦þg„ebŽ¢&(g¼Ûyu<|Ž¯tyn04›\.Ì8åìþH¢`!QÍp§'»§ú¹©šÛ&9†£6=5 EN…ÈÅ€¹åÍá ÙZ5t•‘ú²½°’VÆ>U6ùÜ\„ý3@Ës¶INÎS¨¤—o,*±ö=„T•NdÜÏË*¼Å;Ìùèñ«Ÿ¶gæaSÊ>½Ãæ¤"gøiâ%PÇ&éÄ&°œÈ‡é΀z Ø`Ô¯T¿ãaTzm²ZÈ9þ{÷ZT9ž{„<(|dŸ-[¬;ÍÁ­ûÓ•šXøé ép©M9ý>+ÈzìÌ"nA‹Ùä±/U×û­êlVeýƒ)Ä•°5LCÙ lEwdß=ò½©€¨Li;q篺€l+xºXü6«_ç®™„Íñë(‹T 1²ŒÝ{¥û˜aMØÂär¼•®ro׊½.ï1‘íþ䇗,χŧ?Ó‰Êðê4bn]¯°Õ³Ó–‘¢ŸJÁŠ) ˜ vŠUꊊ²Šì;îÅ…†³oaQ ¤QHªéwàà Ç;&®¯Øƒ ƒtáÏÄW 6ÐÓf+xí䛕&7týüž³UÎiô¨ò¯éQ'ñ‘{(àÈ:¬-¶ {ŽL˜Ú?­Êe²Ht˜nEA¼ÞUÓ‘X¬$:º§Å^Ã¾ß¤Ú ï¶ý×$ñ-;ôª€c•kAØ°Ä–´G·¸d¯ÛF'ZŸ† -í;Gí'dÀn#fÞ@˜“Dk°¦Ža¤ QT h·âÁv&‘ì=¼áKI%/·K ‹n62®¡V0irUp]ØÇá}IÀ?,Ø("¨VÒ!ô„ÑIª­¢ÇµaÀ^&{•m;I÷wä=¶ ç¯WŽ7F0U ç-¦«¨Ü^nZ³3EÕrÉ\†Øyµ¡‘òø=×JÑßÔÃdù錜öŒ#eÁá> ¨¯p€ 5¶ÛPø}ÖèÉãG}_Nà¹8Ôªi½T‹ÞÿT` 7•ñlå¦ ð¾»s±¦¨[4¯öÜéÚ£ÙÙ((åYRwçH(ý½f‚±s¹m]‰Þ1 6'+•Y äýãÅË#{”#¨ã,¿ûnB|‹$€pªÕ@ÄÕ23|¼cÔ°0x;±”7'Èä‰ü-ìÒæÚE4⛃ž˜aC=H­:¨–º0h’ç<æK]Ý€4°ä̆ÇRç®Ù´é)ëçÑœlj\uÐÎ2ý‘j lŽ|¾•û‰Š½\ ÷Yc]«¼W6¯)5ãโlÌ‹‘­ÔþÍÀ5÷>*–2•øl{­‰C[{ÅzhLÂ6EXŠPR¤9¾ÚuIýʧì„gB VÆͳÖé}üÝb×;[m ù4Ÿ~%[s«å¿U†ƒË7 ^ í|!P³ëƒ•á½A)SÎî Èo#¸Ýö³ 7‘C%Æ@òF>^ý1 Ù…¬˜×Ö£­cL³â‹~œ Wµ†¹†¨{vº˜‘ä&ÈK‰_³¦fÎÛŽš“D¼˜7‹1¬fN¦Œ}V_ÐÀô‡»Ê²,2LÉ)åµ&K§Q†öhÔðÛ1uøVúú*Ï'Ò×ò̵)ì 7j•×3Êg^åŒ+ý8 ž±]îq¸êâU[m<| Y+ÐP´üÅ8ðpüvsצ ‰^Ï|ý•yÂ*%#³?8V 4MnÍØ_êçnk†GúñÔ8×1:æqõjÒ ¾Ó†L@‡<ôð¡î¬’Uê°uoW„.üÑ¿º÷E±ÝÇ”8+ËŠC€Á1¿¥A¾>¾‡±NÜúWÚlíàBW£²´‰=pC˜@KJ#TkææÇêÑ2Wál)Ö#æ)ù¬¥mOŸë¯aC5»uör–BÃS’tß~euaLF,Zf©V˜·ÑŠbºÈÓ1-7÷ꎟæ]_,ëªJaÒ)a%Àà®õÂ(8ÊI"P'Æã"‘ó·_3Ù`•gÚdÌA¬M )å…„Q“WJþ¹O¥Ið‡£nðä¶éç­tŽXìÎDÿJËr»öñ° dïésÉÚðkÇx´šçqÜÁž£ê3;´=Òøc½ [@©¨-niâÇ–Z•šQØÚ–œÀ»?î äƒÀiZ^RŒÔÓ¼3R(ÑÊ‘GºÒž£Í~·úP.êˆý:€‘YÆxì÷Ì.T‹ímÖ?€œÂ+ã„OÜ4ƒ!ÊŸÏÊf"OÌq–MÀv@q_ >>àB»þVDÙzÌgA d€éVo®Ìá®5Ûþ™5ÆœS×òÄm2¤ìwùŒd#Ùh*Š|¨Å§±*¼ÐïJM÷æ¹ýeN?"z¥ŒSáždÐäœà s2ðœÏ.‰wtªÔ2søÁü!)æÖ™èJòîTÕžÈúõúçk¦x|†‹·`ùà·RΩý“êäúÓçlõu9þJbŸíÎ-Å‘ÑÇë+êÖý÷Äž|É#pHYÞ¸a¢á°èï[(Ó‡8½¹Õ •?ƒnÿÜN-ç*ÌÚAŒÛ ·Ë¸pf¨Ô3ž¥DUŸÃ'ŒˆÓ vÅx¥iKK bI‹#*pžÝÇ@êQßâðUÓÈ üæõbµÙŒn1sþhÂîKÑ)›op΃õ®”Üþ”™9¨ŠèßnS,Ñ4Û¶xï,¢jAèEG×øMBã#3GŒ²$]šÑmdiäÛ£­æÖyjúSHYm›aUûe÷€9V;«Pl*T^¦s }nh€Ð§úšE%yB/A#hI·øNŽ7tí4º†„ž¹DžI ;¨L=Ö!‹—±¾º'¦ÆÜ)Þ«“xÏÄÎØßÑåš/ Á‡ÄA“bèI8°:KüþFq™n(ƒèaÁãò‘Œ(è•ut¨Ä¢á¿µÀP¹ùKöž×wÍhŒ´]¢ÀhœèAC¯%÷>Ýî­ÑyxbR¨ü›ªß(šÆ‘D©³±çðF–Ï(3qcÐ"h»Nïc[þ²n˜]öƒ„À<#£œKšm4Éùl0©4ÑÀ2ÿi†± ¥@¼˜¹Ñù”Τ:‰h•o<Òjš´7T´Í¶Î¼jYDr.³ uë/§R±}ýh®O¨?PÉ7b¤¼­Ü%Ôj7úL`aéŸUdEi¼¶ø•„RÆÊÎX)—Úúúî•:·¦ ÿ^±Üqº²\ G¾MG-L?\>°Ø`^=b{ŠËÏm Xh´-2ùÞßL¨Ÿ€ø«~|ÅJ}< ½>{ýË7Y}`¹þüjQ¦óžè6»Ñ„„¡öƒ¢d‚ZÏËs/¾†ÿ*‹¾•＞µˆ«42Ñçž™†¤p°’ÊQ!/ª»ŠhW›¿ÕÃ9Õæ &þvž!xÿBÒÂ쫪¬¹ w]–Žl§týoIX¼ÈY–Á§QK¯ŒAE©lBU=…øþý­cõŸ–ÚQúF£ÊžÛç3â ©ú «ºå`â2ùNÖMdg ‚Áh¼¦›´õØHýŒ€kãߢ"ßØX«ÝEð1lLw˜êÊÀ†û’B4ÎZÍÒ ¦HÍ xGCR×~]£ íj1ÝsÚSÁï2ÂDÚÀÊ>ˆ }çŠf¡%a½ùÞ¿|6±y¸ä-AùÉ@x­lÀ (2ÁJÑ„WèsPúºV[Û/¿¶?² ÚûH‡¯ã§Â77ýIANJ÷ø„ „o‘¾³%_œAß–ôµÒéö[Ì<#Üq~~“™*;ÖÁ½cW¬¼œ•­ze¸uÊYxÄZø/H³MFF.yÉ¡ÇûE%ËkÕ«ä´ë ŒÉèè ¿¦"ÙTcP¡jMœP=UùªýŽÇ)fB © ÿƒÊ†¸ΠÁm '’~ÖÞPYÛ‰u¤R:P¹sÏL²¢É#骬's*+ÔŸ/»²¼eb2DJ·‚24“¦×ÑäÃæÚl»w.×Á¨Pn¢(—Ù¢¦ò@|ôûõžÍ,T+%¿ÂiÈÎG˜p6Maþ—gò&¯«5Á‘ö0B)wгAmây¦SèÈãG-Et]p¼ Ó'’óájo¢¤8‘tr&z¡˜ZÃ&fȈJBÿX+×;vØ“³ålTUíë—*1{â\)¹ ëò‚¢—jÌŸôa¢NÙ“¿ßçUÃgnõÞB¤9‘¨{ŠÜ™Õλjol“Ë[ß»ïPMé»sþhQؽl%©®rmZ/.†’|ÂKºÅI=ðÕô⩵„&ò‚[EGÖ;S¡Ÿˆú Ñ»ˆåÚN7yìxð+I­*¸6Cðg¨ ¶6—!;eDŒZúdÖkÛ×6HYÚ¡ŠJ¡.n̲žLÃA„DÿQNeF‚‡²”b5 Ó»-zÃ8wff}º–¸wµœ­Và)—Zóv$qlÔ ±ˆ,FØ}þ]àzk)=Ì°â­©€uËñ V]w °õýÿËŸNaæpÚ@`Ü•Ùtoù‰â¨~•†Á±yÉɸP†Žô"ŽWÏõ)+®\¨>È´XÂUA8ë˼É„‰‰1Cu¯ØŸÌßEƒy"nj)ÒUNc®rÉgOæUëSaL~Fa«¦Ÿ9¸+±OÎiù  @<È¢y_µD¦†ªR—§ÝG÷£Ÿ ØdȨ©ä[ËHê …s+àbÝA=&Ìt‹ë'XÏD[2"7- :5©‹E×zê˜(O |†ùP]àÖ¢VÛ‰Ý>’¨:±îÙአÁdÍ·ú1ëY¢aBèqa•Ô¤Zį6ÆéÀÜæd»>+`rÓ=º“»å¼|,BP)qCí˶ú®¿êôñ#t«*WwH±]Þše—Ú«…¬n Ò‘¶‚­ëê®SS€½?ñ>!‹§Ó =cð¥iû’²´ŸÇ—ÌìD:¾F{©’5Ù梥Šà‚ÔvÒE’¿L6 Ÿm‹.Þ¢öÎi†Î1¶Û{äi*]PnÑë#†ômA]š™:¬mà¸q.£Q# d9P€a°!cC¶Fm<¿ŽÌyr¨Ác) &þ–ö§9 1g@…MâWÔ‚,óÃ…Ãå|IúfYS6ºÜ D¬.‡Ñ%(·°yzJUÚMGåƒãñÓI¾MA«#Û CÓVÑeMãägB{èSéW|TeU;ø9"^{‹í%+¬uSlw³0µòÀè«IàYÏ¿@ŒÉŸôë<Ü ÒŠþ¸ê}î_ìñŠ6h4UÏ£×%nÖ/mÐ6z<ëXƒÙz¾öD»—­ÙëŠ%ÁC C“ài›ÛÏ8‡Hk8øQCªè§&¯Î€ ð4)=ËkƒUýÂÅÅ¥ŒPëͦ‚0øDħïq ..Ö³êùvÃm†ÉýÓ5sQûøügì0¨€IŽ{ç>Yho9ؙ̕M¢Dž;hE´þž“‡JHF;g9âù¾åG¡zTDe§ëòWš1RÒq*Élž¿þû¬î¨~¼…½X©À\a¢ËÏRÍâ˧ž~ •è¤(,ÉL¨2LÝ]‰J ûÛ7·Þâ>G<ã!ùöçÄÞþl…†˜“9SʺY›äG)ïà™ï@ûýÝZûܜȮ”Y”pˆ©ˆúóè—?T7> stream xÚlºp.ݶ5ÛÉŽñdǶmÛ¶mÛ¶mÛÞ±mÛ6wò¿çœïÞ[§ê¯®êš\=çX£W¯ênBQ;[ge{ZzN€’¨’2===€ž– †„DÈÑÄÀÙÂÎVØÀÙ„ fb P2±00éé`HBvöŽfæÎr#ŠªÖÆ6ŽU;k;W #s·««+¿«“ ­£ /í?IJ&&gs€©…µ @HN^CBV @.&«3±5q4°È»Z[¤-ŒLlL(¦vŽëÿ(#;[c‹ÕäDûïl]Mÿ)ÌÔÑÎ #¢, *'« ¢SؤÅÿI·uvâü'ÚÄÈÉù_ÝQœÿW²6ø_Éð¥ÿõ:ºÿG¢…a`[9 MÌ,laèþ…›„­©€õ?fcûÿqýSÓ?õÈÿÁ`lbúO´‹µµ¬ €\ÈÎÆÞÅÙÄ cglâh øönŽÿ²(YØþÓÉÿ¤ØXX{üÿ&ý¿ gƒ@°5ûFZfÖÿ˜-œD-ÜMŒå-œÿÁÞÙÑÅä?f5“ÿÌ“Œ‰±…‹Í†0ùŸfØØÎÖÚãÿ]õŸ¶þ}M:q1115uªÿ%Ä¿Ý"¶Fvƶf%ç°5p4þ_ÿÝòÿa€þÿ†û·Î𺌳£…;@‹žö_t¡ÿ×ñ?’ÎÿE Ú¹{Ñ0°Òh˜þ910±³Ø™Ø}þ«\#GG[çOÀ?½üþob™˜¸›Á¬¯Øq[¦µ†VøŠÍU‚SÍÍè«P­0¯N÷Þj![ ™ë]c?(cç»ÁáÍòì]íJ<´ü6ˆ3;™œŠ& ×YÁ~+!ª4ó­ÝþP(xo×½SŲ1’»¸wÎØX”§œbÎʽ¨•t(ãAcˆÖ/ïìGÚ‰_[|²£Ð!CÉÔü†39 ªAš3)¼1ä{Û_ ËÚƒ»%ÁÒ3/¬°? Ó,h¸¶ÒÒŸgèÇÈÓ:ÄÂÞ%Ýq­ü;´(ðpÍ´‘¾å&°r¤ýœO7²Y~Ë-·µãÏ'ƒ'wÇ ™½ô%UÅ­·ýÌÃ\Ùåq´7ëúê;£¦¦MæéïGW‹ËA ™Ø˜Sfsì — Ÿ@Ò*ëÁ…²i# ݥͲø¥¤tAÆ;ܸ9òHñË™R~½lºažºY`ÃWÜä©~¬¬@gÌs1¾D‘ÙxË‚bð. @ŸVݣΜτø0×&äùßðÕ5éªñ6VEª ˆ¿},ø{›”aÜ »ÁRþ“8Ž‚‘<á|~k ¾Œú§™Ô³þRͨ¨g/.DX9ÓT¼õ+ù¸ßGtZòz와¸Õ,±~×¢NlïëI2-Ò^c~¦K64ö¤¤£úå+®Rä-r„œ±u!÷¾ù©‚»ú{‹àt6õZÆq0&  ]î¤x×GBÛË q·œWy&ž•9Üø;¢p/ÿCüÌvÞlŠf*3oä`óg¹<12àˆæ„šÑÝz9Éç ¹º"µšµ™>‡•z¾š/îé›qùÕɳ™cÄÝ€ÿùìêM¯^Ä[œE®á¦##@°'x”ÃY®ÔN¢ñK2‹LC³q~Õvš‹Æˆe63 ½×} ¼K²>Þü“÷1è.t¼ lHSÏóÚÃ5DÕX¶~ÈÙbBñ„Èú$<Å.?{´ ÀïúþƒÙä‰Æ7J]~^]\?êi*œ?Ö@±r£¤O¸•V87^Ž"jm>ô‚†UÓƒI¦–ýö¯]ÞÃi®C«e®žZ)²0 ìLAÏñê‡XkŠt²zê"B(n"k±¹¦PA#À£‹ƒ¸‘~Bâ5áF-XÐóŠ‡&ZÓ…’à id[YÏ›kúUdY¶^ùIOeVLÖªÐ8›¹ös^k!c·¡Ì/øL`D³û•Y[a%‘_Oþ/¥«´Eˆ„¬ú¼’ÈTýü‘·Meyì ¦ Vº„·7¾˜Œøq,<2Ø}¡†šb3ùÌéoâEKßQhK 4K­eØÎ ´ì»z˜#!kãášÓ)½¡1P’4ú½¯êa~Û ´n ž¬baî¨dª}ÜgH©¹ÛúâFbH)î/YÀ×0¹Ý•§²pÔB;R9ÌÔZ´¬Þ-5Ï&Š³ q¾Ì8&+åí_d$¹w6á«'_ Æ, !%/\PÄÄæ02mŒjMá8@£Æ¡èÂV‘u¥éíFÿ-¢r¶Ü¼_­x¯…ºF`ùFuj—Š i þÎÍ1€\ìwŽz©¿Û ˬ+¶Þ‚6*3‘ó ‰4z¼œ6aQô*êØ<¶QÕ§Ÿu$£f[‚eÆ-MÌU];“’ú/ÞiwÒÎ9x™ª3©ó/¥ø8´Ì²*!©¢ÎK;AÖ%y ìÕÚ3Òå“”ñIÆ©®äjO©›ÏeùT Ù²ÑÂ-“™Ñ<Í•ŸÌ™/NC^Uó{®®€¬#WÇL-5cÒ„n`óEm¶¦ÌÊÃæ•p:ú¤ç3U†Ž$…Ρ0ÇMøóH@ {š}u#r³µ¸]qù~Ò‡œ´•?-§JÏåeXxšÑŸ®äu‰9ßCæeõÅý_ªØ$ðuÐp5æ¥#Šà]DQï7·55. 2ý©ËGÕ™­Kn_@[|pÊÓÞJep…Štrã$Ú¶æüšZaÀv5;víÏÌGK¯äg༱^£ó ù«x];.Ä€Lm%;à ðJbæ¦R÷bfwŒ\¼äMÃ<8ŒÂìYŒà´Rs"…–N}Gy.\äe";ºÈ¢ UZ‹y,†yz¨Î3¦Bí¤ ¢ÝǨBáqKü/”6ÿL\ãÍÚÚÓÍð²Â"ç²¹£ØjYØbºrEïÇm®KjbàB`‹xùµþbqb¸±žÌ¹Å|/ð¿éÈ$OÆ©Û·®½‚?¦]Mì‹Æ€ m¿HΈÀ‹4]%Ÿœ^g«µq­J“!•ø-ÕEÅ"_*£1ˆCù”ç:Á,;äð¶ªÜ8Fü”ÁfÛu€èäå% =àe06­ôä!ŠìÝ ¼«æ0&»´xÔ7qŒj;ßgèf·‹.r‡xà® e¹juñ©òs†yBçÜN˜«÷í T±å¨¹aho#;Ek²Y}º¦DYe¬h~»¬ŸTYÕåÀB±g$ì’»9‘i1ý¦µ ÌO&_¬RÜ$·å=¹±•tiÚºœ²Éˆ ÇiœSd°ÿ• … Ó”qÁD w¶CÀüÞÏλÍ~U?ôM7ž$rÄŽbH 7Škaz$ùO~;²Jšõ€:pé‹\•Îê ^ Ùakö³¹½©=,r_¼•—“'Á8VÀ¡ë%—rÈ¢pI?§æûÛ¸*¬Ó67™RI+‘ ± ´ WàŒ÷A~[S‘¿Äi[©†IJ¨s' :ßØ&îRg.a¾Î«¦ÛIªAÍgO]bJÚ\»Ë~á`Æ«»Ë=\Á=Âw=ŸÁj¥ÂžVëèê‚ßJŒ¬%”Ù²0Õ×Ò,bæy'žUfÄêaúTuq‡š ñD P?Ìœ?` \º®ÄÄÌÑØÛ€BKmî÷ª<ýbÕ êc¿ÑâëÕl¢êÀ êþää>e8w„‡¡wfܼ¶%˜… ï­…8®3ú¥8ååxÄËl”MIi}¢Ùq=†6ú„‚1¿øfJitŽ‡€šk#ÆPÍ+Ñ ‰bÚ›ßЩ14ßK‚àÑ€(¬IÊŽ§³oz™öž.ï{pÝQñβûr¡Ø|×]îcÚÒùi²tÜW%Ó;±œ¹ÌRF]r?¤·3À Ô­©â÷ோLÛåÕ²®«D\‡ãžÆª™/‰Ç¶Ëé2(5¿GÏÔW¡–|D> Ót÷ƒ Ðô„V ù”þ`ê¡F7º»ùeéY_e?N‡Žrðê¸ç=—k’­S(äfZÚ‚q‹z2m Qkç1[K„²P«kG/oyÂ*— ³[Âz%¬ ÆØÐgÞ²'ëYª;K íØ ŒØŒ²R„:Æ[ò1û…4bùD+‰„j_µxTgU®ÝN»*Ô+JÖ%çq(ö>6@0‰äø*èÂîé¿Û㸈·FîºR(&øÎY·^ ¡•n¶BÁë!¹°¹ÈÏW÷„’œ†ÏxH ÞãÚäu Œ‹¥4 £C»—rÑGÁ}uN:Ô© {m_Ë+N7\t¦/™jPΰØtH!mš i”s  |æªËï»ëš<¿º7RHZ0foß3(ðv2ŸðÖŠ@âHðQƒh˜”ªþžQžd/#A¾óvnííáRí×`gÝJ?tÊRKˆ¿w«Ä“wÈØ3Ûø5ØAxÓ! q>±>®;–#N.°¢)ÈÄ9Ç+¨ Â÷¸Ü½8^¬Z–rª¾‡¥qè z V"hGªø¡›õ:'Kþã’æIûÍBž°qÔÉ(P>E×vјxâ©ŸèìªzÒlYdú·•¥u '[E%ê0,ëéºð”®†$Ö$L©æ¢•Gx7x;EA,#× ô2Þ× äŽ®_¸àúV±'S¼›yH}x7e ;YÓl—(?PL·êdá»Ñå¶OÔèò÷òÏÜ—.a:@á×b¤H,ÒV +›…L¥*oï>|.™©²Õæi,šÃŽëµÈµ}+û§wÍËfY.<*ñõ^+mB¬‘¦n|%±Ñ•ï‰ú¬Y_ÙààCŸk}¶a#5){œ" ë$ùWlJϘEõ™]„|F5‘áå(+?„OÊ£çg—ŸbD»æ>˜¸ÿœdÐ/S$²‰T±ÚÍ“hù*š6ÅÆ'kËm’éW‘ ítˆÙˆ!3ý| ý‹@ÜáooÜ¿s…wSi¨±!Plž%~…*SÆÛäÓò䋨îìíxô©ݱu¯5ÏÈ’>%»dpóË6<¼ä}õ¦ ®P‚dm\HìB_³_*¿>ß,=­ IÅ7çϸ<ßìdK¥ë#X‹å´ž)¸š æ×Ñ‘ È)?yP¥|#¸Û½“jÿzŠDDðŒD@»‰cüBn·âCÌÖÏ É¬÷Ï°Â2kÁ‡Î¥èÀ7…k™¢Y‘Ú…bí«¹DAoé5îÒ8»FÜD˜uæÍyÆ*¾»¦œq§a±ô!¶Ê8Ãz)!ΠÀ!»¥?¯ø±b; f†}@êO¨XRnW åÊî:nÎXéØÚŽvâc4 ýDÐÐwú@&ýóîd¬®0» kßUF°âQîؽTÊ?¬–·‰^'zñcsCÿZº 3ͽ´&=ˆR8/ ²³¦,Ad÷ç ÔWodð¸eþÜè7%U>HÉôåéƒN î¥Ö¥½÷'6ŽŸJûÙD*\ø¸é+ôÜ)†÷¬zf…FƒJ*Ë^‚½ê VmŸQSȨÇp %ü!é¤Ü‚ ñ~0¢µ”å¥e',m`Î>[“G¶eÎ¥Ú®o-bíÉÍy‡—•o5Ö6Ë6B7È3õTY¹<ÒÃÝþW¿·Ñ­4wÍÀ¾xÑ›y»r£¹¾|Ü‹œ3,í™(³”ƒÖæ>è'»õ11KÍ;m‘ÜI•G/0‡ƒþ·9ü‘ŽÂÑ«òV;ØD{Ä‘ôêšÕ³æ„hëð-¤\ri)›ÔÁ,»ôã-V[®à¬ºjþ‡ƒ˜ Zg>ÒÑÏZ¶,&‡íªv"Êh‰¨‘Y!Q›Â$m?§¯I©áįpVš…­—Ø2-­’Á↘w$¦rõ€ÌȈ )釆³_ŽÓ×é¼–¨Ô!îÐÉ .¢žpÆ»X´W`áú.©(ø̬}‚sÜÁ6;Â&îø[…É(ߙ¬ÊU éü`¨5Äñ ÙÚÔ«eøq¶÷o+õ)¬ŠL¾W¥ÒšF  d"ûM¦øV«àa²·÷ß ª9½ó ò±ñ~®ÇéKS¨¨¶Ë®˜†Ø×”ÂpMʉnvܸÞÚ0|HßmÞ2~ºD?¬qf¿sÜbýñp˜&WH½s ›Ã Œæ„p#7iQ“ôǪC󼪕X,|®Í6#ž0Ø\Èeb=øjZž~ÂRWÇtöì-“j;+§O¸–”Ж_ß}Žà™ ½Ó‰ì¥¤ST¢ •ïѼCP¬ªdN‘%•ËÔoËÈT¸õ’Z,¨7„Ä,¬SÅùï.ûÜÔ¼%ÖA "‡Þ|µŠá îvx²N‰X«Ð.ÉÞc1†>YåI¢<¹o&Ew­ß´:'%Ú8݇3Â|«J4EVSPþ¾Qvg‹$Gw¦ð¶š«¾¹ŸØL#$GáN³™¿ÛµßLCŒ[÷&ˉ.#W×ÙÈ–mÕöbûYJzÞ6Ó~¡Ë_RwŠ4\_'rÛ\œ^¸OìÅ÷© su°1ë®S-.%uÿæ;í2×Ûxåéb¬Ï÷#™ÕªÁ’Ç߸,ü¸ø\¦Çt†³Œ Œ:-£ü¤&Èà7ÊYžk$Í-—FÛ³7΀éYócÀ© «ðÜIB¦>ßHž‡uçåeRÆ™éAd7¡XÊ™ÑKÿ¦D?Ö^/â›0¯#Â/žúMƒ(b„¼!¯T©ys€hLGyÙ~!ãÜ\xŒÖ¤‚s8DCy)ͦÍ÷®sZjŸ˜ñŒ;ô*¥Q™0PþmÛHBË)ÓêØ©›Ö5u¨êmwÇ>wPœžM+ïÀ^³å±È¹¦‘™«Ÿþ:Ý]jÃiŽ1s%¤•&¬äáÊ1‡ARíZiê@P4Â؇±A¶™@¢ðNÓ.a(™ÇÓÆQ­eeéÒƒ…"è¾ãwŸ£…Ûù=Òg;ŒUÓ»n wEggQù#‡š[ Bn$@±PS—$­\¨qÕû!#*ÀC Ó;ˆ¬IrY‹)Õ‹Š™|ôÍ´|X/€²½[dŃi2ñúʤA&»çSCÎü”§àhÁHÖÉDTG=tðf¸‘–áÏc0sö.ÜéAý›À´„çX ýnÏ3üxLOüN²¶˜0üÊ}l=Â^Ì烑Âiu0š¶ìyf*DåoU};à A®³Æê`ÊЩ =2¶G ÓuÚú£Eßn–å'è)Ã…N.È­Ù^݇ÉkjòÞ/ñùƒè†cÆâMæö³çÆÙ”~ƒ§7>Ì`¸÷/8~Q“ øÎoßÐÙCÝ玿¡ Q[GðÖH„4Ÿ¬¥{r¾-CZ,Jë¥úQœ»çø î6F‰ÄW‘–ÚƒäÇç&Üž†Òqçˆ|ˆËŒ‹«—ŽÁéÌ¿Îÿ.O)¹‰¿ùþ†œ.¯ø®£úÚÜå±@@èê°1#dœ‰>¢ÒµÈD$Ø”œäòqOÁ Pï`% X†¾»®ð ‚WU‚Z¼YÂXLÐ¥.I!¼äٻIJê`~@äÍÓ1YÖ~#Ø{œÎ9Måkk1·± ñß$åÈH®þrEó­2!¾5Rñf)x•é{ñèlÙÜó‚Ð<Äq'ÉZïÀê<ÚF—w•=Áœ×‘S²ñ9!‚Á¯‡L½¤híˆýI†Ý ç‡úÒ~ÉÈ0è½6qÐÀŠùÏÐüre˜µ€Ú¢öodºò ¡IVûýRœÔÛW9¼lç謠ڦ³¸Ï-àº|\n;Kš” {Sÿ6 é´„:Ù•Í |ºôÔ­ÀnÖr±G„i—‡x¾›ƺ:ÍÀjú¹Á*ÝÙ’‰òe)^2§‘p¥Ð.(0´LèÔPÁ¹g9Ÿ¾a¶±7™ F½c¹ŸìI=bS”MÆþyÛ´_€hîèW ¡IÔl¹þ|MzÕùÍMó2ª¡4s¢JŸà ÛV—«ãO‹»l~m¼¶V¢S¬gNŒŽ™p8]Âü1Îôþ õ·ÞÅeq¤ìyU¹÷Œ]0ñv´_³K4®r™ˆX¬û«˜¹ UÐêá·Â‰§Fë#ŒG¹IâoèÄxÌ}ÅóáÆÏÏåöúß·ýù¨  ãÁü;º×Hý“Úî©æ•-<Æ<Ä‘.¿ƒüO8ú&Ië!y¬ŸH+Ô5 Š}7 Ê<µ¨Ýuˆñ¦Òg¿f–B_S$ªt‚kóè%ê|:}AXªœV´¶G©z¼ú.}±\Ü´ A\?îOÆÇèÖ‘é”нŸ|¬qêÍÕìѬˆ°à#ê8GÓãH,­`®œNç´EV½yœ9\Dÿ‹…L¾Ô¦1ÏEô)´$aqÈ-›_ÖrÈ·q&ÇÎ.`WÂ7Õ'ý–È(rÊ/—‚Þ·øiFXÇüzêfŒR2ýSÞ¶Q D^·è˜}]úúA2ä~IÊêãü7ãΦ4œ “%ï’¸µsÙ„¹å@¸ôÊæA[­—¼7 ±Ê¹¦]×ØAÉËЋá ¶ xÂVKÎ]‰Á Ù,ÅiÉ`:²×‚uƒ¨lÐj¸¯‡‘ [|NpûkÏzx´f\üÜÓÏ•@œ¶¿ÿÉèY_ƒ$¼<˜ºVqì¦î¯;Ç ‹l9¶<#é!Y;6Âeh}M…Ÿ:q€²á‡^ûJö¹å7šÃRÈ3jó¼–惌Lnoc5\S,×_ 7²ÕÐý­äeŒ˜‡EzŸ2 ø}FhOŨŽº3ÈEÔÓ+=P,c`… Åßû²ƒfüD*Oíc4âö_jæKpí›cL?ðð·:M1A˜RÖWH ÀÕ%)¥ç¯6mÑò;ìè3fÊ[ý¶ˆE¦Â7Ò/hðˆû7kö]YÝYjÔT@8Àåµʼ٬Ïçž¿2ÀF%†‚öËÀ×­¬ÌÇ?‚Y¿Þýb„ÇByé³%à^Ë×ݘ™K37zïtÿž¿s_t¿1o\/7ððœ¾÷GtNé‘Š^IXËÜÜ×°J}@dôz_=Èm{ÓikÍÐ;↛½<µcaÙåÓðˆÌÓkX¤V×1:ja…ÔC³Õ}¤S—-¦ÜšÍK°CrTWT*r°ëîöéÙF[‚;ÌÄÓÁƒ)p¨Ø½@¿#öò* ¦èóØ›<,h» °Q{œzFBC°)l:)‰:~zá¾ÏÔµD#3?'å³Ù]Úú¦²1¢x5Œâo68¹kÄ#FÂ%?ƒÐ<ãW’ò·Á¸yzðÅÝPý Þ·kD áå²@F®Äã7u0qOµVš`~£Z‰#‡¢sªKà+Š>WYsC¡¸O\ &<3èÕo±9³´ácâ,±‡àu’©ÐÎÜŸ%+pB µH2FI…MÕAÄ_ÿ¢Í̈¼Ù2¤˜XÆ\kžâ÷ÈŽyŒJö]¶Èkô¨ ŸÌJàVw–[ƒA›lù­š/Ûͯx¦ò}$T$BNœß-Ú†j€"tÕ*á|üÉç÷„¶[ö¹ "ÒsÛ'ò ®+“@ðl0L‘›©/ìa‡OÔš¿ÆˆÚ'2È0ŸŒÑC³“ô`Ï…¢ÞÙÅ 9ØÁ´.¿ÌbÁòâ‰ÞkY¹^älδål€,ZÊ ÅO¶’Â…hè1æU»*Wh”® bÐEà¿°Ÿâ¦49’¸fgOµŽù²ga‘ýáè)¶4lD÷“Ì!ýÿ"„V#!¦·$hƈFçR-Æ?VQ_Ȉ–•ÝQ!9BæÞø§## \9û†3›8O-y=”µIÛ cá'ŽÒåxпªü?‚$„ßSð¿v5 dŠ I°ÕJZ}±Ñ·]ÕX¢È“º#ùï¶4; –©«•Iѧä €~òÁÌ\ðõªÕ¥¡FomöîÞóS¦ÅE<Ö{5‘0´»ÅâLy<‰u3ÐÒ“PKãc¥9 oÑ? 3Æ;6Ĥü@Dœá>A‹n _E).”é´±":¯"½>qžV×å¿’cµt‰ž$ûcFžg©ÿöÓªŽ¸8쪼;##Ÿáƈï õbœÕîÄ3ՔϴFðoñãHq«XöCð]¼…à!G\ ûf8‘Þü’»–R(L OI‘Y_]OãÛýDiâ¯çÊ †› £ÙÚËËƇ9… ‚œ¥€~ÉhùÒ =­ÀÔ„ V¤5îˆ>8Ãý÷qá²KC/åW¼ø¶g òN[ lx(¥ƒµàˆrå8Nšpr²øÇ';mÌõ]¶ÅjÜýO•à0 ì>̶ëƒ÷*Ϲ@ ®Pø^ŽæÞÍ8kuy*ž3:8•¢Â÷êÞtXV¶Õ;®—Î^ùz61O”óeûgølA*Æp#¢¹éÔæ-aœi¯|3jß'¸Æ뻤³´Q‡å”•÷%¤°Ð†Å´”zS]`+L—ÈÔ7OWÓȃ­má”jí+4~¼Ÿê¯]EE¹ôŽCLv¸Î_V0 íÚMúÊ•”íI‚Ñ´J-%é¢æÆ۱·Žs*¬#Ý™˜®ÄXªýeñ¹ñ9†ºQËI@C²wn«¬•5ðHÕæÉwÜá!ÓÐQ¢p´Wš¹)åfÔ“:½JÐÊÍi§[oeg¥(О‡*˜?YÊòlÅ­XrÃP‘3;P0Ãï) B-Ñ´„K$¨ÆEË ¾AÝÓd{Ãa†¾šìgï¾EJÊØä»w5™`eÓ;þøñlêæ'/ª&—%EuZî°¸cŒðú# Õr_ÆÌû”:Ž‡»ˆ (5ž˜¿ &à·|-ô,½²(Ô€¥ê»{Så9^UKt‚˜¤³ƒšÆyϵ½¥þ\35$Nô½Í®ϮŮ[„ ›1!Ú[.¶ “µªó ŒÔà½ûÒ©’’Õ×·£Ïi@Ë-Z^ùìÉX-J`gkN¶rH¯ßqã‚ÜšgGf0¦"ñõ;.óBÖ½$“Fv¡ä%D°êJ÷k=¢\œy<›}Ÿ 0m8rØÕS—¿ˆñrƒ,… Ÿð‹ÿºê¼Ê£’ºåXÎêÝÑÈ ®gµäV"À,¯+uÜ9´’ÍÑk#I“\·;H§ºª! âŒÏÎúvívCý.ÍMòò„̤١p {¡\(xÛ¼O0%ƒ6‰°ã™æ¥y¯X—Kú’Î ¤Û’âØ$Št4æ-s¦W7o¹ ^á ¶qe"øÆKáCâzqõ‹øÛªø ¶àî5°|™v,k˜A¯ËöáEÛ{Ö…Ð濉|ÿ¾F »?ÆÃAžÂ V ¢NÔ™ ]Ðܽû /G½ ‚0kÑ ž'Áqý<¥¡Äh¸1ç’ôYQÌÓ½µ…\t¯Ô·³†ÏξP_îƼÿ6{o ¤«;,c†L'-qVÿcOƒc ôï`¦ÊKœ«©MŽ¨U-”k¯è¬Ø±Â¡åM&ñ.VОÙ¾T¦%{+qæ-E¤:ñL§ÞC®Í4—'­¯K™¡Ãw(XíÛ4÷a{ºKA¹e¥@Y¥’F^Ù'ÀD&7ÍÑëX½÷éã®/ ÄÌÐéà ÷c—˪öa •aËÕúY™LëFÁéx# å¤ý{N2ÅæÍ_\;–Jo@|9áäÝC’ƒ4,õ˜ÔšH5ÄòðEUí l][äsÛJr6†µœ´âóÌ((üyˆÚWyùbõªa+N÷C™­rºñbd†€ l[ „HâƒgÃk–³äËtÎÔ™ìPØlŠqêûW*]~4Ø(»ƒYm>‰²iÎ ©#õ墻Cæ¦d—–FJo– ;:^!¼zGæ¯i1¢‹È„º't7¦‘€aá`j¥HèT5Ðá\Ǻ¥êú°ø ×ÖÖ¯”ßqÆÉžƒ¨˜½æý«‹eR*KÓå iÓ~”J|>§½Úiy÷Ú%F¡+ž FðÁ@ÊUÌZ$e§†¡å«SÔ&ÚîÞlÅBˆ;;qº¦S§CF÷‡‘¶ÛªŽ|;!á,»\O¶p/=6ˆw þë—‹:œã9Žp  Иú¨Qþ ÔÝl`©wµ?nXê5û¸e)Œ8Cï/pU*ëþc¼:½&N6âAØ¿ 63ˆ)?6#±F žÙC|/×æ£ Š¬ší©F'›î&/±@6ªï§Öéc>2Ó•lâ¸ÃóêO×DéÂ{bñ“Ì(áÝ E¬õÌy¿nVÐQÞ»¡˜ü" ·ZÄ#“qçiÃ]¢ûs6:TºåvõÌ£î-a¯÷R™·­v3IÅcÒÎ;‹8äËÚåGÇÎcIî,M¿8ÕOufÚ¦HaÑv@ÏNžh=„ p¡÷‡ÉqCrÛÜþ– îÖöw¹.‰ÏÐÄ4ÌÚí¡YõŽ‰câľ¶Ð°¸l€îlzÓ^³RNWë¬[PE@*žK!2¿²…¾Ê ¹ZY~ÎrM2k»ÙÒºt6ÓUè¹çÙZùN¨íŽd¼cÇ5ù¢œz¼q¥54nî_ßnÀL_íýb!\vùœ4Ô¤¨v”ÛÚyN¼ÊÉÅ=žîä*'– ì­ ‡šË¸Œ#Ç ýéŽ6^Ì î¨Í ]q7ž‹w´ÞzÑœA²ýZ˜?€‚PìBe¾üå?Ôj¼ j\`xgf|¼¼®ÑqSŸ Üå‡ó§O;íToMžû݈âdØlq„€ 6%!»-Í ÝÚÕÏzº¬RÒR¾n›uÐHR–RXZ‹*DGXÎ8_5fJ= ݧôùæ+:»ƒg¼à’úTƒ\ŠIÂHøY7tÏ+¿ß}<ªàlÇR S~Ù7»¿Cƒ9_oð›AQß‘¶óÑ¢5”·åËOA,-«Õbép»£°âãâ[vI¼Ò›óE+rŽËh6Në%†X°!¢l¸ú„P0.|Ðx¦¼4~)€Ÿü&«èýåRàÈò½VÞõ­þ˜'·†&¼]ùi<è㯗LYô ˜.«=Ÿ8Š0 ý§„ûZÝ.ŽüÏ5LäÄÈ­‡ðœ›VíÔÂwaï#SPÅ¢š d;耳 ³ðÇ…'s“÷rŠ 'ù8TeüiÔ‚Ÿ3ÐÆÙu:"ÇG¨ØSA½®—ý´Ùi£›¥¥5\«Þª…ß’.O£Wijar-W,§ðõ¾¹Ö0ï(Î+üŒéfÙN¸£"Ì’€½.8=Öè9õW7<ûëpi+¦E¼ÑÎA×[ôtŽ} în†{»sq–ÒÛÜâLG|³p8uy›ŠÞ^þ¤=Ae@Π€¨7 ŽÉõ¥±‡L²°#F–3oŒø'¡-qíè!Ù—íX²·^öš¡&#yJž`lZz‚—X|NÊQ2:òv+£©<ºÀp¾+ yÍâ„®ß-™Î„×æ‡oZ°îlìckÑ<·0=nÛÌXÆ¡`·DÎjEãGˆîlÂ*sΫšM±‰}ï̼õi¯.u^—d ãäßϼ®Ç—¯½Ø~>ycÙIÑþ-K7lv)øº)Ðd~µ4¿J Iënž:ª%Û1‰b#D9t^ÎHHõï•kúOÓI‘“Wdú%;& ™I΢6â&¥™«‰5]uЈީ±}ÛéjùÁ…Å.7﹋ŸjT¥×þ!|g;Ýw¢ä1K»¿]#…¾í¿õ›²¿[®åŠ‚h¥Ÿ6ƒ?±J3FêqS…ˆ=w©˜Z×äL hx !{ÀB÷,%¾Çm ·_ÂRC„’­hBøüðËóÌÖÒ^5w[ò…2ŸbÄõSª:0ÃÕ%NCµOÐ’emq 1‡”*#øNl´šø< 8±‚Æ ãoªx݇)ùœQ?ˆiðs©Ä#4»‘À¶ Õ}^)*²dº1³œ<Áw*·ì†)jì¤sÞ4³[cnoJñ|.È@ß ñ"ft(šÉ¯BÕµþÙœz°sò;J‰¨¯ÁÿáùYxTYD;°;¦ ­Õ‹J¹6ßT_[;pQèI¥ñ øÔÔOÐi4dö3Åý¥Ö÷Q‚i¶æ„¸F|ZÁhªYyÞ§8C.è„ÙøºÌz`4¹8ox‡4=¹8îŸè©™Tß;öçKžÃ§b?—ìá)ŠÉé)¹+¸÷ë/ÅÎh"'$Ù‡~ÊÑhm¼«¾ ®Ìð&¯:ï*àèušÛf¯ÌOx¦>Êõ`‘ L7¾þúqæD]q±¹sukDÀ,îXn O2§Ò‡á`~$míYYr•³ßjîS”ø¾5WŒÁÂä.ž¼zK”änN¥SL~j½ˆjÁ“ίaòg¡+ìih 2g8"3è¹Ù¡†þ…Rš“GÅŠa7Oh@—£æ7ql¦AÒX²öN=ºkú3ÛȘl}ÊØA`ì—sxΑêjóÝOÅ6<Êk@+â¬îQ19¤ G O «.gä‰Ëy*ö'5Ÿ’¿Å+ðÝág¦Æ²mëÔù5jN„ÓÝçsí5UµsE#Ž<¦ß9¥ÔŒ哤×,Z†Ä©]qìžu]j\èÞôt#S!€lnëÙdÎqð ÑæÀχ3ã–à|˜)Z3„Z(Ø æª‹°Ùe8‹¦[@Þ¸jA€ç,» {Ã\1õ‹ä²…xm²±í%tžMmÓ½Bõç[Š oHà~Ur`ò6w²NŸB}8ï—Áƒµòâ,åXà¤;<›˜Œ6Ä9ê®"Y7tGã§ìµYæyYØMæ¸Â³ä½ÕNRÝ/–cS‡EÑûWãóº°—LIfs¤MÒ@gäÆôù ÄÓ£eãÎûÙ{΂‘ÖªpHÄx RïÔñ¯é÷ƒÆÏWåmm0'L¡Ë@><µ›°ÝîÔuux e× rË ÎÞ.5 ÅéŠ l³ëH"ÖÞòKÔ&œ®K{Aq±SD©Œd%üé{9C.‡ö¹Œ$0ŽÛ“K‹ÚuЄÛTÌ 9ÁØ‘ÉOb«&úgµ¼…×_ô ŠŒŒÛ³"R‹ÀŸš,rY*Š •®žÚ»êeìMÆ°¬E¶WµÃòbó çŽsŒÌþÝO|E~!#ÀÈÄj…¡0‰¸‡–“7lŠ__;õunÝ[if*ÿƒˆÄbì°Ðçq=¬@åò;ÖÃ@ØÉ;uˆªmzù¿Û1&ômx9væð(6n‰Ÿû.v FÅãh8q×ĨíM¸Yñ×—lÍÍo¡'èïä—qwúw³} “ûX)N²ýòOÐbÊnµ7}<¸ÌÅ‘Z°$4¤§ë÷»Hÿ²ûááRpÊÀDK„·fN0F~ËuÅ,ñzÇ©¬nr™/KÎña){0‰$&7Vþj?’Ñkò9®oý;}N”xÀ÷áÆ2†4g^Fé«´ƒæ»!iÒ=†3¼àuef—4ï‡-=þŠPÿÚÊëÍ1sÉrªñ¡9m7ÁŒŽlæØ_p˜»{Lv[—´•“Lèlëš()nî©\ùijÕ(òø^ ¯•´fº[7}¯œYž{tëžôªÐ`ˆ±º·énEµôêåwô÷_ÊÍ÷ö¶Äê€ÚÀ[¢ÁRGhÂy´€%®!ZEnøž¬\ÒUažKçÙáÄãÂ>Ê ‰8c¥Çë¸Æ«œ6SoéDîà$«zÙƒBê|ÿŽ°`O®Yý÷ÛÅ}p[Ã(íK2Õ€N«ñ¸X‚Éæ­ÜB" ÌË(æñ¶mÐA16iJª¦>óö/ U•´üÒô@è$`t8fù…3g,㤉|SÛRû”×A†ËÞƒGx?9–UàløOøXð̬¿· ôäóbùÊØÖ2â„Êù-CøÂv~[3BËÔ5x7~F$Žš (:Öô; ô¹ñ_³pÎÆ_£Ôc¢SEð`Ì¢©{†vŸ¡ »½†DºáþÜS‰fŸ? QŸÂa˜ ¼W?üÆ^bØ_ 2 ø´ŽâORŒÈv7Ì™aôR 2&[þþû\Ï»‹Ä.¨ à­¾¦VðDDÏ©€Ÿ)ÅãBë{m¤âÿ ±!1ÒˆNýû‹5y§âÃ߀ÈEiã"OÎ.A!ÛØÿ‚[´Ó8M<÷@•?4öœ¯ßùfÅqû§²*! º& y$e]5¼@ùO¬÷.¾{‘íWv׊§â¤ñnµ&ÑìÔC)‰S¡‡,÷®0º´‚%qçrR].Åv¥ lˆV2i=’¢¹SRÏ•á :ûòð ¡‹v‘ß;c7Áty÷ŽV`Cœ¦CÒ}ªZuIº)KyñèL„å]lÿc*»“€­¬µrð¥w³2•97—0ž ˆ^Ù2FÀd‘Rë’1xÒ˜"îoŒîÙnƒÎ:öªô3ÅÉfƒ4;Ñ[drƒÅ2^ëÃá Ð ƒ nHÁ?ã ]ëöTÞ„ÏomtMaÚ@ø®§uHœ„BÑ‹wI°ÐãЮõ³y¡×]hÌéѼ!~ýÞ­}3Þ-¾`7sYâóÓ¿tÕ=Êñ"Ax¼–OØø¡UsR:ÀèG <³µUÚcˆ™žÁÑt;*Ãù_G׎xàYJpZ•z"Õ©¶içš´©•ø¡êwgŒC€0Ú£¯dA•È3¯NùÞ{RÞT5) §ý™pŠ«† ”iFüÌu¢Šô²&-¨8ý~N8Dš5ü° BIp4Fô!3n©1Â9¾ßŒGdqLÇè ¼l: ÑÛñc7™OÖvÛœŠU¾J‹´>™ à±Ežž€_ž‰¡ÑǪš¦Sê,XjÑìÅK*£vÂwªôã>Ž&©GÕn†Ò?á÷æ‹RN;I·{LºAy j²äþ½äpK±lÕƒð¦w¤ŠJÀ?[¨ùbB•$Š»VE+šãÔÞÒúVLÔlGïUzNJ!¡Ðw&ÈEÒarž{§—0¼gÑsŵϹ?“ÉìM¶qùºFé%ŽÉh¡8 oùé°Öv¤à:ê3–£ P°`J³ÖÖTLÚŒÓe€\þší²C¡)Ìó`{ßZenÈIã:Sý½ü "^šöaI5Èl•âç­DXi,e¿³ä8åè9Ÿª‚Š¹Eü,«Æ>šä鈟©ôõóìÞÿXšdó_%ß,jzn\Ç!»nOPðƒ»9´€È'UŒ‘že2H^Îd}¾G4ýd&¸EÑf¯å®J'ÕÚ'X8û‹ÊB‚%«®™¾*‚¼_0jQÓ£u?A*ñ«~NS£$ÍV&~ü4çö‡+ÓOÓ=9ÆìÂ’¹YÇ:r@Ï ¶L^Ó,Ä°UÓ©¹pkø…%…`-8Š%†ƒ€rk™qŠøJ ï¢ÿ=¸w—=Kˆçã’è )³ÛÓ«A‰É=²jHç* WKh+È1‚y?wV«Çâ‘‚3ƒÜ¸I’…@[ÚIÚ4 vÂööTúRðçå uMÑAËy1,o1`ÙÌÉ_÷=w½{ç85ƒïÞå%C>–ºhƒÀ6T«øâÉ _\ߪ›wžó«js:»”Õ} Ùé¨ ÃD[YŒåê%øwájD·Ô5•:Ò:ÓZÑétGÌU#ÀíÒ|J]ù8©ä ‹ÅE…Г‘æ²Ò¦ÌÌ ‘úhEtxcà‹V: Ÿcd«gQ‰m€Æ­ÆnÓ³m¢D9@Zõ$†vÜ㎯à“0 #úb7çÊNßV.ÉÂ)žn÷RÁ‹n…'b:ox¶v“xAð(‡ËìJ¥™2ü´ÔUU¢ŽùÁÜû/Å2Ðäÿ™*fÕQ¼ ^Ç úý““Ñ ¾_˜´õä+ªÚÕ.æ2ã-Ý ÞŽéùMÈNM}í $€Àþ…+”[gÌÜé}$‚D;éïðÖ¬H+3‡Œeº_äš´ÇèfÈœºH¹\‘n\ö¶¦Uܲ(÷®«†™®f ÖzÐ.³k׿ZC¹Yé’Ê!fý[–† ç)Áö˜Ü‡ELl·O Ð¤c4‚îÿ­??:L…T#€ÞƒËÖXN<'"»'ˆt™#X) ”籎F+VñÆ9þ÷ .<du¬”lžÑ êe,ˆ,¡-Êc@â(ÙÌnÕU4»{CcV}ø‚žf« ?\¯µnî±Çy‚«¸îa@¹ßò‚kci"äö„Š#†NU:+{8³Ó%¥‰ñëTS/y;¡Ý}\?⼜ԟiÎV*1³?ñ€™`,ií7„šqXétáÛ4<Öçþp‡{æZ·GšjÌWpIRÇôå‚Qô,·‰}Û§bµŠ§ž–e£5%¤8ß\÷Ð5Öâ`ùS&`s;¡~NecÌœÐó£½=“?)¨7bý¹s¦ŒÚÙç¼Í S>”9w­-,mæ?þ™ÞN¤cé¬üæ1|5…D@ÞÐ3>³zJ’Aѳ¡£¥„‰=Åð•Ý½¥ÍæR`‘•(rÛ¾‡é§LZFÑ'aŽ Á7% 3 Á~XÓWmp_„Æ3pù„BÆIˆ@91Wγ 3&a ¢pÀ;€g0F60ò¶“¨þ œ<ë²hS{ §´¢cžzbeX­’Ú<¦«œ€Íô”¸x‰j'Ñ:²?hz6nA(9¡í§:¢³C°þÃaíIh´M@I/Ahž»»'Ñl Hèz¥`’¹ÞÄ#ƒ™Ëq¹…ÝP*ZxÀèé¶õ œKÊ5f‚÷4geÙÇ|@éC“±®¢PaEï]xù©¯±uaÃU¼ÕƒÃžn‹MU"AúÜS©— ËÀÈoÓf(Íìrùÿˆù=;Ãs¿o·à!÷”­~n‚°Á¢i݆ûìµ µ0ÂÙ\=8Hd¯D…j囧 ‘T&öŽ§ùï¬ç3Ä·ôOdO¯óÔÔ@Îsö¾Q©Xrâ´àµÑ0m0¥dŒd9­S¼¾úØd¯©8‹Ø| ©¹îË}žY Ç–)×~ Ü蔈„áªøИ‚èHæ,&…MP—‹Û߈š< :°Š«ì“J‘Zl ‡;*üÑÀ’˜ú{V¿[Ã)…ùyVZ4udo']~Rh”9Ή†!ªoCêBè|1ƒíLä®Y¬—ØqÄhsßí†ÝcC¾l“§ç'—LOR}kÝkb£èNE©ÓDG›X‘KXwÁu» ƒº¶†hG4%Ãs4ÕîlÌ:¡‰ÿ0”Dìp` ä±çõ£¡±ôÅþ5ú½šÆnQÙ¤PNÿÍ+Àæâ(™¦Z^4‘£¬ú§‰ñ}÷³d…ùA2ŠÕjZ”³Þ„%bI½Æ?j¼ˆÅ;碘$º“¯=UP~ñëðO¦WL~iÖÿêÂü÷—ÌÃãsË«~ï´ü×EdÞð÷Àaœ$‰¦{N*Gµ«é7ÁÏZ¥šýTÈZ>]GÊvsù<†œÌ©bOOæ×$g™œŒ5ƒ\ÂÚ³¯—ž ‰I?wõË:àE²nAÈi¯ñ";¼/­rCQúË¿ñV(µçµäcø9nvæÀ Žò幺÷Ï]µûò( ׃ÉI`®§;®”jtnëæ2؇ú(PŸMµb7F3€¶ïîBmÆà.K¥ÓbNýÌ';ž4çq`ñ½€¬,uÀ~¢¼1I¬ëKê´š\HÜKËrM~H±­¿YVwôíÅÉuùÕðò€–t UÆS#c¼ÅºVEPöºkz©«9ÓÑc²z¢»B"–6Z‰Ö!iÚ#yÊIB‚‰Ò£Cì3 <âq³àÆ#¯»öÓFN$Ø«ß_IÕÓ~@šŠ¦5èq“g_¶=v˜@Âß‚Õ×â(ð Tè¿ÄeÁüN¼Ü”òZh7rí£žUâ]ÒX\Éõ ²bBíÙà‚}ù,?TtM¥Þïóê¢À¬pº’pi¤°ŽŸøxü|pÅ \¨×uãôõ³„ä0ŽŒ®%Hä aã 3Ú(?ã"‹~/¤/MéHe/£6ÂÙ¦DwùV“Ë  Í8YÌèy&ÎÞt; Éfg4Ÿù?„C•¼ˆh0–Äb%TéW]•a1Y”x)µ…yæþêXšÇíªM'0&ÆŽ“²WÃØ2H‚T¨­ÎÅ—€GÔ÷¢Ü2²µš/UcIÝ4%ð£µyÜ_ŒNþNX·FpƒlÕ´Dì¦ü8 A7%š‡Ù"ûðp² [ìæí?H:Îv¤*¼fíÅê|ágï×»ää"fÈf¹‰ñ2“£¯úu¸³e#©¬#0´ˆ-æ{(í ¹äÀ?'#O ˜7“•rÒOýzãÛ,1þzírvšÍ¤´!CíÞÅ„îJ,à¾h ‚Ô^ö¼ûÍßý“>ÄžI]ÉÌ…˜{£u‚æ³»¾ÕÖ7˜Ø6µÄ“Tôªk\§æså·Sàdëò (ÍbS0¾Ïpš¦P“†ÕÍÚ]š5DÇærŽÄ_i'”ƒ¥XfU¹'Ù—6»•¿!-2ù5¹bflÌ$è?¬‘N‡¬ÄRzÝ7çxŠœät/§!Y¥)3±;dÄŠ<ì5ÒCÄ“hZ8òLÒP*%&3-?ʹäÈítkO{?[ÇžN¨»ùæ§Æ8ªEuÃìêš±]ö°BU©µÈ•Hlùö6麒Ðb§"I½Ù¼¯ÕÆ8 Zò¨|YX\/ KV‚}¶&Ë[4r¾Úæ1›æþ‹›„“×óM'¨ãh.®dyFV&kÌ^ÕËç&?sø [¥çK‰õ7Š“Õ’k(¢ìZ@ï4 î5÷Çàâñ YxÎ!.AŠÝȦ¾4û#ö3Ô£±Ž‘èërÖÄ0*¹[xrj–~Á7 ²$r:ÍήNo;ª‰"WE‹ ã|6<¶Ï~9œA{ †;eáÞ÷"ð¶¨ì¨éž–B5Jåö”>œ„û[GRñ=‰uÏ­"k áÆybÖ»Çó#Êÿ2sM]÷h  Ða·4xÇ &œÃ”Óü°«VöÞq—'>øÞÎhÊÎ]`È–V \°ÙrWîþ”VÚvÁ†}ÕraôߦT;|p&mœ›V»0IË°’¤Ì•7rZÍRV¿‡àð¨K~ÚÚ“7ûÃxè=ÓR0›´ÚlT÷›÷˜ªüÍsy‹1ó¾ÊÐ"Y ’J\dLqž´_aHÁÞ1KÚåq¸ÈuêqâÀ–Þ%„âi‡‡2×J/žlö2 ò;ÁrÁ9—¦Œ$¹Ê®"þLE–ÆÕ ·]s;ÇMût|Åç0#ð¤H^ÿÎð_•Ž0kZ¤GDLjB&» Ýⶠ±×6ÀÎRü³œÕ\Æf~[Î6°hêõÄÜMsÌÓ.Òéä5-pê>.ÿÇÝêÖ­ä÷É°BÎRˆ:i3[eƒÜØÓ´]ê?)õ%Öè]"½!Ll Èz£¨z)Ž©2X<ë@üõ‚H2(/RF%Šô ôô”ö´zqéaÚþg_؉ ‘é'à vœÄÀ‚î-s²ÀÀÕ­›µdŽ6@°ö/ÒßÚýþ£¹ÙKEOr÷zñhüò=öq÷h}ôhØÕ×Ó£æ9ÜÛdÅ g?ñä‹e b'#×pϤ¤LÀ€!ÇÝaúïîŒÙº\%ÒRÎ~8[žÚ¦þýµìÝŽûÑÿÛBÒÒðî‘ä<)óÓñ2¤?v…. ëÛ)q$IÈpN­€D•#z‘þ‹ýhú’iFÓĆ£šœî”ÛÊŽ¿©Òkûk×ï3èë9Kl08åþ„Ž«y‰ÄÉ䤶EÉmêaêIßOô„*úmÙGHl§Iã¯áL™áHáô\!ÜoÖ®d|ئc™¢”Û'ù2EWAòB6ó;}CñXç/Š\\]2teA§úEòð|Œé/ºZ<૦!¡…ê:ÌÊg9çNù¯ ˽w¨0xÂÈÕìt5ˆ»¸ç¤k(ëvXR+þÒ/Æ@iÅq1a4†LSÎé )[ü°Ð¡óF ì„97éS—î¦ú³Œ³/d“ËœnR%¢b2Õ—«£fö¤`ÔÚefÄŸéIZ#¯uÞÂzsê©]Ii¿M $Ëxíl)*Ì9ýz„‚}m3ܵŒbvvêÏøÛ-ózfû7QË´®®­¼\§CʱeæH༶£Ä§JÚ!hžÔ>ЩQµ›¥©1àœFh ǵ‡H7Éá¡°k>Òœ¢†çf"–Ÿ¾©¡üF®µ–|ÂP#—,Îã‘ýÇAÓsÃàªéƾ“Ë8¾Óâìx›¡2Û±çRéÔ§Çö1Dî’¢#ƒàQ__Ö˜e†4ßØBÔŸÖÍÑ©ÚŒ›lîõ02ÑN¥ß©ÂœÒÉo÷f—¶ jìãAQRQ6DéH°´_VÓDÊJÜI¤y¨PDK)‚ȳÞœ´­*»V·2E3È©Q<}_A… cr"Šî© >û·“ö ô¯8BÑTߟ.Ûï[ò"DIùû–f-c¤èæ³™žÆi^f„c®šØ¾ýæôC?"j”/m&f¹3}ÜæY5€/c2öÌúG˜:Hu¶Ë VÂh‚[Å÷Hu;#P!ê‰cœMPµÉÕ{ÕÐnGÁ¦/)‡êw‰ó4£ˆ–ž³Rñmlþ—ÄïÁ½øÙÆÙ©}•i€µsŠïõ¦>ëS€YÇà0¿Ä¾Ô&CÞªÌ,'=ÛâóÎ…êÎDCÛ™„f|Î.ä¯ã_t˜^ÎA¨ô!«bÓì•W[´gxâ@*@”¥5Ú.´¤WÄŠZµŽÿ„¤LÙh˜ý.£" a‹\ë'¯ë÷6²ŒÜPÎX5½ Âß©W‡CžÃâ­{Ñ뽨éEÂÔe]qz1mA²ûÒ^‹zî‹ý‰axâÛ‚NÓ?W.ÓVMÛ½iJµæù–¯Ô7›ÐÑüu ªYI_³Å×°AÿꪖÈ@ŠDþÐ’ÚáGL/ÞÊícp•[œþ wêáPhÏþ׃Kõšm± ĘæFrdDq´:á홚6'¡õ7ü¬SÈòV5̲¶ŒÊÑѼ6 ëxõºXº"æ²'ùZ©y7•;E;£¼:¨7œ7¼®úc<ÚH†1„ªª¿V³g2”»~Ù(óÚõ%ÄP6³ï_ÎŽÉËó²5áLÔ $zÔÈ{u[] 9± Žp|Íå}"ÂFWãƒò7¤°ÏÊh\ŠÜ…qKo¡=mzª%äÚ~§ø[bì-"êͱâm7»Ó¨QºÈ3‰¥U( gÖ²F’šï˜mÒ€1ƒNy[›¤èUù<4iEFãA8°ùø¬ˆu0ûÝÞrÐú´XtÑ>ƒ°ÏûLeÖ‚¯6¤‡,øÑÓ4UÇ‚0ÌâÔ©Ù Ji[a| í¢´þeÙóãºþ6IÛ¬²ÈˆéG(Ÿ G\TßõÛM!Ó‘þ]ndIPy]¹*$ãyÊ+õÐ@­B5öð|US DB“ofY+:ðã\+W¼µmBýÕÏÛô1'#W–ëì¸b­Ÿ,•RD®ùQñÏ~p”‹3÷ £5rír.K!ˆ‘ÉõàsMÌIäí܉µ |Ž ÿŽJ¬5c¸Ë"+¦Y÷ú¦82KÞÄ <ÀLu9!qh“ícdñ˜ ¢7OŽ "˜‹_–·sãO|H²évæT@u´iJñø`@ ¯ A«UÒwÒoÂÛWÀ½‚yy)¬$”ü<;º2Bjä¼PqRYÀn‡ ­ýìð¨'—,†ÆZ Àú€é^¥ÊðêS÷ósb'v:¾Ñ¢^ 0©}ηí\\Ðí„k™$Òh<÷ šÜ Ç‚÷V$ Y@ÔÂê7–mXI§8÷›òÍ„)o2Á¿.CEjŽ¿$¡Aú^ù¬o™ªvD¢¸V]:ÜìYr‘}ˆPÌŸ¥=Ž8ЂÚDALDÂ;ܧŸ£Ê’ã3”™%h ­¹wÜ—O£ËÓ߶’º4™ÿ*™H«§³ØMöƒšøtSKçaÒŸÉxéªáôK{˜Ý¯Í7ÝÂá€Êk1h¸nüê5¥ÀsÄíÒå¨ãQmÅA¡ý¥2l FÉîe_„E׬ÿº:Vý‰èˆ<^Ÿ©l 5Ï“ a_ZîébBZƒ¹,Ö> vQX'á}ƒYRø¬oÝÐßfˆ…Ñ )o¬€ñ e­H¨56ŠÌï(‚*´Cª×5,À5­¾þsÿ°4îªj#/°#‰WËÞÛ¸3¹Ÿ…ký/ŸpæüBãÂç¤?Ù[~,‚ãu¦1ÎöYôŠQOEc6ÕvGh2%OÔzf—Ûš®i d¤÷ŠU[C älGö•ø•§e•ÙWèÁ<\‰XŠ1'Tý„k¯Õ ®©ý²_0_Oô 4€ŒÖ»æš~C`Ô £Wao|ÉBÀoÕÑ—nÏîiZÂP4z›PŸ‰Üe«®¢C´ìò¨H$”­}øÿl€¦[=—’&EWúç¤åøl¡kרæ*›šH4îÅÉû£cà+›ý|i´?þÐIúz<щÜQðýx¶¿UH'¶ÆNÈ*s¤78Ü}½'|m€}?&R²n$ /M(©_t ¨Q¨+žñòp™Žº»ªè~ÞWTô}û¼"\÷©m_â^¡xÇ~y9þ³ ] Ä<õÝG¹­26zS@Ï·Ö{ýuÉçǦ2Z$çLþùDÄÞA¸ @ªÞYü¡†¹÷ƒõÖô/mÙ)‡³î•mkV!Xc%„Ôã¿÷CZÖo¹,6žìô|ªÂ ¢æb æéÇù¡ j q—­c–2¿ì·Xgb˜Ù•¢Cl`< /H¯°ƒh›nLûø^ äîÖ_’*gª]%$BƒÄuƒO£Ó™Š¦VIA¥46$t WÙq/,Ò}#þ› ëÕ„v[Nþ-ÝŸPË°v-ü$Ë0ȼX‚ûþKÀ|à ÜRCr‡„hÕz\-­ &ÃŽô{f<¶Ì’OºuÖ+½vExB±Ó­q 4½)´ûRÑ 9Þbë_Y¦)è¼" ‰•ÅÆÍ­ÇÒgáŽìÚCL(¶)ûh‡T¨£o€KûTU-~ù‡E¥SŒÎiÌU-4q³\v¨¢)™Ùf åÕìrZ¥! »Wf½üé6œ ÆFf;™e0mZ"÷+¨ôïDxQˤ'Žâ´ü¸±Â…5·FlšÖþ[Of‰­VÔ·FË3I÷^E„œ»¦`ä'â‚Š1Í­(¿Ù\$9ƒ×ÇQáïYa±¾£1y= ‚rñšVôÿñäås©´®¤Jç6¸˜Û³à†âšÌÑwk¿º™‚úIgë6“Š:qªß¡qÅxwtNÛJ¾kZ§úŠ¶sŽïK¼)ŸY±v±3!”<@§ˆKÈßtþJ¦ ûá56ß~ùŽ¶Ä‡‹4 ûå ožfuswdïMûC‹¿`×@µÄ®Ô•ä)9ntñJZˆ?žjÑ‚®t·ƒÓ)q6C_h^Óé™o·½@ÔMÐ=ý@ìZTƒ¾Ô5Ÿmÿ'6^1/ì}—_uD)€“¼¢âÙBëÉ…»ÍP>ÑØŒ^HÎÒ³vÊ<¸à=þ\¶Ê/7gò#+­Åúº¿Z“ú°…†WóâU"~ZhJ0Ruâpˆo}õz|Ǹͅᱫ»é ­¯…ÿŽ’4  ˜²\w\ì-t4;øáþ^lƒ=S½.dâlâ…ùªkK„Eræ±E3ápû‰%È Zt{XvÖšF'7Ê‹8éûZÐÑÅ ‹MóFÌì-ÏßîS·‰œA©ˆÅ¶–S<éÆr`HTN…¨,µE—5jÔˆ¢—8]TìP\¢§0Ã<ÿ¸ŸðjüÞéÞP„k.+"<>­BÒ}®W uîšžyË'}€ª°2 H¿*;¥Zè;ž%Zߧ 4žñøh$m+=5p&æÕú! ƒrâ«Xs& ûí<¾`°¼žY÷Óµˆz®ÏtÇÃDìC®‰Û’˜uˆÍ´‘otï· ÇËÐêKzzÚ4œ Ô){ïEÙÙˆ?JâúSód”+Ž"WUÐþÜâ7-ž]drÇ; 53^ 7S•ÑñìR°e²éðª6t[ÿ=JRǹ5l@D©Ãð•C8Mškôæ+t–å H8„fŸ.á…©â~TàO÷ŸaRy‡Oøf&>iÿ˜#ÕM%ϹVÊÌh¤ö„úÏìÆeaŸnÆzµ7õDãà8=_G^'¯Ãf­b¬5Í{ÿ!¶¥¾©\…ä@u¥Y IrßVƒ>éãŒA²RÕ!£W1ÅÏ70 ÙJÌža¢“ïF²ñ­ uÛŽž« ë£B¾¦®‡¤ ÐjYn;¹ ª£Ôwa½ è’Q&È›NÉÒ¸߈÷·<ÊQ48"â`ËACžRj”K_8@,·€§<»Ž«o.ÕSwòaÀbžÍmºóXÐ"9Åq:ì'yãe.zWès5%2d Uý±Üdg¥$çÂ1=SŸHüîv¶;¹Àôÿ£sE܃ ý™¶à}i LeþLe2&4¯j&ŸŒÑÀ¹“ÅæÛž‹BLó– V¬ôp1¾~ÁiV…T3èmœËij¨=' f$«ïtÅ âÈ3V€©".÷Ì4êd)Ç,çjø‚T¯éQ§ìœ±š´Q ÿ/mö¡`¹ãÑUdÃHUÔ2”Ú²ËÅ»¾Í pwô}y5«ß7ë}ìKƒK³cùÔk¸ÿ‚~c^#µp]h«G·¼þ½~ªÞM¸Íü!üg¿9­u|íçŸ{ô‡|;CEPp/<3à_éÑ0±»™ê"×»¼Qˆ–ÇÿÛÜ abK4£*ðv•T#åÌ8Î1ÿ£éZ†šèJ2Ûw¾B«Á×vÙMG24Mºì„ÿ½è¼ÛL³ã#€Ss%!÷m…gt#‘:Uà`ÍUR˜ÄÉ©8îØX1¦¦ÀÑá{&´¥~öIu&d:á¿nèöGš?h‰Æ€Ø@òH÷2Ö=¡®5oc©¦=%¹É JC&»ùdK‹ƒ sòe­Úû¨’q:ÑﱡúmÛ BQ`ŸÆp<ÌÁþÛðÛ{gÚ„=£Œl4/±ÿŠ³p(Gn¸³HZ<±:ô‘¹”£ô7°7Ïz.”£™u•ÿ¤lrÕ0hÓaÕdâ¶T´ÄO2î{>y3žˆw!eË$ôu…hfs8³BQ>ßAkG¶=RÀ°®Ž;+ììlƒ™ lƒzø¶Ïó9µÈž‰»Î‹Y›ð`88æç¡4…¯‹Àµž >ã2Têg0Ac ¤˜Ž0?)Ü“ÝYüœ98Sò‹äª\õ‹·Kj§åó÷‚µÌ4‚þ+Ž•ÿOé´µZ3wtÞ‰«V¢Q¡•ñ×öz.JŽºQ»žÚ¼©®{¯#XìßH:C ê¤vCSæœ"üÅ\+DkJ·¥?òÌÙ‘ßǶ®k¥¢ÀYú±_,v…j!‘Y†ùØ…Vcá”,Üÿ¥tú°;šºðá¶äèÓuÛLŪcÆ °&&ÆÔ}£Õx“ä JÝÝ7GB…¡áÊ‹L³ÅeaÃ6¨x–”ßË2ô¾µÖù)äŸ"tyƒ:j0²ÀkzÀFÙ\¢¸(†û¥›"<°J@®mfè—0zýF+‹a$_ö‚¥d¥4f‚rÚOL`BЀDðŠkP(•BªD<•rÐ%ƒÕôcÓ—´9ÎBâþæîe¥H™Ç¶ VÈ’Í^5› Vÿ~K@°c 2#%ÓD&NS½äÕÊßbÉïÑš›ùA튂wtQIì›þ;Òu^ìÓI[«³TU3À©& ® endstream endobj 68 0 obj << /Length1 725 /Length2 25566 /Length3 0 /Length 26038 /Filter /FlateDecode >> stream xÚl»spfÏÚ.ÛÉÄx’L’‰mÛ¶ùĶmÛšØ6&¶m;™Øþ~{ïóîS﩯Ö?7«¯ûê»WõêêEF,fgë¬âad¤e¤cà(‹©¨0200è˜áÈÈ„†Îv¶"†Î@.€:Р ´02˜áÈÂvöŽfæ΀_Æ”ÿ6Ô¬ M,l,jvÖv®ÆæWWWW':G>º’”@€³9`ja Ë+hJʉ~‰Ë©Ķ@GCk€‚‹‘µ…1@ÆÂh뤘Ú9¬ÿ£ŒílM,þ…ɉîßl]ŽÎÿ3u´³ÈŠªŠÉË©D…éU„†¶&‰Òm¸þ‰;;ÿ«:€ó%kÃÿJFÿ•þëutÿDÇÈ0±0vÍ,láèÿÅ›¤­©€í?fûÿqýÈé|€_ÿpH 0šþíbm-ghü¶³±wq:díL€Ž¶€qïæhñåB m,¬=þƒÿO„¤³á?dÚšýCÃLNbî@ çøvvtþǬüÏÜÈM,\lþ“ü‡“fÕÄÎÖÚãÿŒøO)ÿ^DCSBVšú¿Mðo·¨­±‰…­@Ùù> Mþkø·[ÁÐâ?ô?Xþ§£ŒÿW—5tv´ph3Ðý«Eþõü¤û£„„ìܽh™8´Ìl Ffv3‡Ïÿ‚kìâè´uþ7éÿÔò?ú¿› tí-Ûs[¦µ„–ûŠÎV@RÏN¨R/3‚®Lý¹ÖFµ47Ñ?¼Ä½SÁÍwC ˜ 0âÝý»#y×LbD8ž˜Œ&×]Æ})&­0ó­ÙzSüýZ®w¡†cc,qq뜱¾¨@5É’•{^#åPÊ‹*ÎmPÖщ²¿ºð`+N©K–©õ…ܧ®F™5)¸2âÙ[ ËÚE¸&ÃÑ7/(°? Óú]i)¬­©0ÇØ%•§}€ƒ#²J¾íZñ;¸(xwɼ•¾é&¸|¨ó˜O?<YyÍ#¿¹íÏ/K ýø!]É£·õÈËRÑéq¸;ãúì;­®®CáéïžFG/®Î8m£…“AÄ ¬Tù½Ib +ÏWÊê|Ûã“À*ÄYdÔ >©3£Bëâ1)Ø Âͳ±¯”6²±‹‡"ÎM{¸ï;,_kµÄ½#“Ny=¯r¸¼mN» •â%(^øŽièqpc~2±^Æ IJžñô`ö›?Á¬²EM_ã–0n›J›[õ`eîÏZ­ö긚ÙÛJîÂìX/X]Ô »êÑ™ZýGÑšGώ勤—ƒW}â;Õ׎·N,àOaà§H2¸Ý¥ÄzšÈß½íŸ\Á’¤ÛÖÛPbõ?¡ ÆU-©6Ç* ïÛ4áveU9ÐJD%1xu¥/È€c¹šO§1ÑK@N?Ç9W×µŽ5î€44¬‰5žÂªc1hSÕÛ¶_úz£Å)"”U)²LlÛëà)å”.è'×+¯W¶œ½ ^íS7Ü»´å™lC2=Û9~AJ¾”Ã2Mú¨¤ïÅîÅnôüT­¢ƒ]¯Z€ ­HÜd¼éѬ–°xÌXåoë =è–ú%G8¨;wî ¯®g+CÛþ€añ†xÂYÍ=¾HHÏØ÷˜¾OÙ¤²OFÌgÙÜ©LWâz5íC4œ§OtÆ’d}ªv.<>/3it'¹±_ÚÇ@¤*êâÑ¢º{>6NøÈâ-˜t¾W’aqegó4Š³7˜ôöð(2;DE¶¾“™…THòèmSljR£ŸBëìƧمZÉqº¡ÊŹÅß'o·ÝT¤™üÓd)Xk_­ƒ¡±/ªÄë‘Æ9~x8Žý&Ïúâxnw—³gF"‰BIziž•€>³êöDý Сk9Oô8±9Åz`TKìu™Bò+»ª³8ü@µ_é¸y>¥’¸V%ï¸èÓ0p¦2`Z_?^~IPQg›=uehu–,hé¿íYÉûx1Ë==!côuTðåÞ¤îŒ[‚i#¦p-÷?ß¹•“Cë¸?yJà0Œû6`Îûlbe\eHeÛ¨}Z\Ì_a‘?¡L¸Ÿ$ø“‚ˆjG¾¸|Á•Ùçt_ßÖÁHžÏÓ´3‘¼Ì[ö¡é‚N÷¶GohÕÙ¹øE»ÉŠç@Jèôãg$È4L»Dk~ô[Kþõ‹ÙÆË”£´ÃZ» æ‰i^ÁN+;f÷ùýØb<ºhÜP’ÓÙQL§›‘¬BçµGåDí:ý¼øf÷»ýù¹Ù"îV¡k×/½!¦a çYø¸B«Éã‚Ý…¢Aq̯‰~:›Þ¹·gw€9Uh< \ý½ â ‚‘úôDí½c‹=ô+sÒ¡Ä$´o€}'JÑç©ÑØì·ƒÞ²*8?EuyÁTDGÛ4lùdÈá23O8ý‹_„ ÚëÆ;%ócª•ƒ‡)©nh\ޤƺü¯ÀÙŠLí«n°XÊ¡^qT“ö&á|ÜÒ˜»÷¢ƒ€ÔYXYç%X1#G-îÎ{¸=¸Khˆãª¸áÂ.Òˆ âŸ“'£Vù¶Ÿy‹ÖÂÜ铽U¡¬¨“´Ôx)F7Ú-.·H÷Í3Šù×G½/át ÅP‰vc`šae³ì^¸kR4üò¿ÌO¬òi ѾUp1ß·wújÜ  ß¢ê ‡Gû¹]m0³÷”°G§å`Ì# R–Å~HMe}±ïW]ßá½ÛW¾E?¬'Ÿ¯jˆ}Eˆ´–U^F¼mÑÿ¬SÁ®€€t€p@¹dñ5yË[9Mú³ùç±{ˆUÇÇL¶ýòÅê,zùìÔr³Ç ?T‡IôJ²$Ô}$räSó¡Ù(ÞCÏ×wþU-ÿ«¹µw}W³ÏÜ™}Å Ê š1£ymöV[O¿{wì2X~[£ÖG»$Þ˜±­÷þ’ÀÛî×è윷æ±Öo=Fºãb$…ÛÝø7a¤î¯!⪠*E^VØ=ðÄ¥}Æîveo×Ë‘ø! ;¾r®áàûˆ«¦m5Q÷%=ŸC+Sþ>ñ=P+[¥î¦»ïÙÁÒ]_g­ª;.ÿdøÎÛRIÄL „Qhʵ<]éú%{ˆ³§lÉ5q°Zñ&ê„*Aš¯OÅ¿£N w¨}Kн–ôùå™Ç³=&ló¢b:ˆ„*¯ÞÜ!¾ê_™Ò‚u™¯^ëˆÃÙ¹.ü³/êóE’ÅócTœ´2E!pïGÓ¼™¾¬”õûL”‰bö#¶öY­Òí]~Ù¶ˆsßmi" TÑågk¤PPû}Ú…ñèÇ‚žtaþÄ@=™`?æÓ¤ÛÒY±BV·¸±Îœ@²Kw‘LäË—pSÔÛ‚Ö ÎÊ«•k#`Y°”ýËdJÔj%V¢6ÀSøj: ÑÑæQnÖO?÷Ïñ‘ãˆlHÁ(RÂÜDÍ–NSUÖÙmäèò‰›y£ /+”7P›P;H}¢(l- lªqF2é÷$sñµÐá4^…ò’ íïNËài~7p8DðÇHµõ̱O-Û•n6ø}’"§9¿‡ŠÑÅÃ*}6ÝãuFÑ gãGRQC'zŽÈ cÅe˪fô$æú¹¤PŒ°˜ÂqMPÑ¿/GTô¿&X"ô!‘…ÎÖÔÂø°UH¸BtçÞà CRòßW>q ¡táÅ<éÊ…Údî©X^¤¥›6 ²àiÖº!aPÔÒ®iX#sQöP0‘9hH<8ÅêFÝ›oMöñyǬÿlG¹ø6èΙözÒ8µO ƒ£GZc3æUÒÙ‹«“"-QºÚ‰~¸o¯¿ëÓÐw‘„Í«øÓu%Ž!M¬µó–eÉH³—£‡'>µ×P^L¡÷.µm„ÕY;$#è¿£Gï×|Ø[V » ìóI/æM)ÀœŠDn¸/ò|‹¤yßÏ‘d™"}|»þ‚¼ô‰¤òàƒÔaÆ*hµºÛäU|q,ÒšnV HQfiØÔÀ˜ÊýskˆT“sçlüˆ«ýâ`”&ãUáØÒÕr¸ésÙ4ÐîÏ3 ‡éƒ÷„qN5À)3¾¶ ¾S¦`¢¼™ò9í7éÚ÷îU0¨¿ª¹€Æä­êþ%.ÓÉÝœO“^>о­R¬¦Ù¯JK_Î 0(R#LUŸiK¸Ô/¦™ª±36ˆNe+D9Žj, x;õõIà Õ•á]ÇVîd5¬òÙùïÈ ´Úyº~ƒ¨È©»rÌtCÃ߇ŸfÉÞ›þÊ‹ó}4¨–kÉ㿯JÔ/›7ü£nÁ7ñá—<ª,ç¢Áé’x€‚/¤eÕ!t©;ÑS¹ÖFá{ZqÍhï‚ôÇe!€x”’DôR^†¡‘3 ©—&"¬J÷ð°þ~Ô…É< »®¤á¹¯àøH~Lß“=]'+,d¢…¹V«§bNñÚxóM Ö›‰½o)œãF ¨(ÃÝÛ{¦€Çä`i-³6§4ùý |¡ÜE;^°S4þ›¸¿­•+$›»z‰XÜC† ¡Û´š¬ÉÝpZ0Õžœ²e£@Û©¬q³2â>_½7xÚãœñ­KӒܥ/3r T|»‘cÍj„Üãe"²ºf+…Ö‰©ƒ,o ^‚˜!‚ày^n¾I{ì÷SÁó’ _1ª˜Û‰²ÅÝ;téhlŠ b+ “‰qg>žß^‚,핆ÐpÔðõtv*þ¦sVKŽ|Gº35Êðj×"¬¢ÃlkeD£([W¤CUŸövÎ^8ICVœ 7ÒÄQ5R.›‡»oRÎòö hÙÕrµ ¬€&Ç!@+æff“.*0‚ÃMaU¢ ÷Ƭ¥M"b)^q‹b_E·÷9 pÃÖ=Ù(¹÷…7$Š*÷¼m;yòŽÿüy¾¥yŸ RºÝc « =(mkŸû4Ò–pv°¢ø­C¦†¨”¹;k°§¿Æ0>žUÕ2zŸ˜ŸFxLZh"ã Þï¼Ö…×æ‘ä¦ü;VBy´Ã=úãC¨3öÎO<¥ysÜÅ*©îË‘ue0˜õ/Ý÷ž·‹Ç»¨h«Ç¾×¨HM²emWëGû=òöfy3“Áä–v‡±*ã@ÄÈœ.be¬l+>N˜LÚUyÿôÕœ°XZ¶Ú/QqLlV3?‘/Ë{{,ÍDbÄWtû¢Q½rrì|1¹š¨·±e ²!'xTò1ÎË02skƒ**ó§$T\¤¼B@x¼á±y‡g£gÃÍXÓKýNŽš– Z½'ÒËŠ©èqiofëÈTÉ©GËðCœ²‰4¥'ªŒ®Æ»ÕW·­R —]*’ßó贆;·x&‡I9eMKpò$Öst’¶öŽºÛ%£úÕý~ZÕÆáOBCÕ\P,êß]¡‡.w®Ýå† 0³Z>xÅJ©ýãUn}ëi†Í‡ØÔé •,Œyêñäq‹†n“(CUÙèzñ¯Ç׊…<-\Ç&†kÝvçQ –Ýø]†¨cÉÒhŠÂƯ!úÙ Ñ ß½°tÏõ³ÝMª4¦ú$ò./˜Û„ÀO~‡Ó¤½Ìû^\y˜tô4&l7DºbŠêeˆÆ[’™á†©y(LskM]{%èZݼ®kb±ÌP1wüä„yÙé|†ªu¹æÚe©çÙXA~'-ŸùÖ/ÏE1§–9uŽXu‘(U_$æS¼Ìß«8/¼\¥ä¯?kWå<~cÝ5‹öëEÎί—UéXP‡Sb#Ò¿$'ã™LeUh2ÎÓÜìVÛUù#g wö,<=ô“Ç€dÅ3™ƒX¦³+yÝÇèÕ¢ì6½ëÑä_6®SjL‹YƒEm¸½‡ƒ<ͬ!¾ù Ñ’OÀ‹ÁLK\kˇU=oPœµq¶é÷XPÈšîÞ`ÿ)…æ’µºî{¬:ˆÆ<]æ­…‚à xýk]s¦•Gúó˜ñqÎ$”Sh"Óû+ì:ñÇKRÎqëÇ0W³~«q™ )ƒ·Aù^nõìÚ†_Ó5¹ò²õñq<»“…ð¬ïR./oG:m/Øùüí 3Ç-·Ý‹(>Ò{€&–!»j˜J{s KEgx†`RÄ›$CÆèNçýLà«"*rxˆJºäÉóÍ¿ÄØô6Rƒ7Ž@6²²³©4 ¬:rÇEÿж_벜<‹×¼u¶J™qaì ¤4šÙž*»e¾Áö|ߘvþq —lwg졶ü^ o¶ñîD‰10¿›*rµÕ$©Ë/.áýê ŸVQ+’ Û:XøÂèª*æÒtŠ*ÃÄrƒ„Ã~щÞÅ…Êõ²ñ•Z¨x*áþèuëïÏÛ¹1,ð“hC_êèp,-‚:"¹û³†ñÈcáÃ!ôD²x(n‘s¡\ô7‚Tž» É…ÈdÐ ê Á¨ÚPR/JÝm ¾ú_4ž­0¦ØgVn" p1àÊÐ; xÜÔžHÀ{–èÍÃz w"7¡§¥2¯@»rñX(ËIU­äß´ŒÆî%ö)–ÿîã¨p–‹ß[ßù3%‚«çsÆ1M™"w£×í¸ŒÈ@ ®â¸ wø†‹â­4›’O°¯ðZÊ]¦þ¬0cÚ»Ùã‡í¾Ð`¯‡Ìý«ÿPÍìÚܧ„À{VzP²¹ƒ n’nŒI=‹8Ì#­•­å‹A^Ìák—©5x57Í“Ejâ»îý+uV‚¿©Qµ† C¯„ú|9ÓlÑ$'›ÇOÞÏh×îžA ƒ’ǽ1)¨ÎüDËXp‚iæ`3œïéÐ7–ùç{Èüʆô˜¼± ZB] ÎuŒòÈòúòt×eLИLV™ü§ÖtR°Xlb™ˆ‰R‡ÂÎE£Y\$\uî®6Pªˆ3 ¶]z Æ²…¦´ëB¼ ž`cuaðé¹^LÀU`é2U%T(ÏäÆ€zKMÍÄÇq?S»ÌÚد±ª›|ün¨9ïö*‡æÈ êÝsSîŒN™Kû:aöžÆonùð퇃anpïâê67¸¢mÚ5?ièòŠ `ƒ4‰ÈF^\&QY»Cmü×h<ÆÖºe÷,§¤¸néä*Š‹•îEë¬á|jã= ½q–Y¥Wù;O¡2³]m¦¼šƒn4D3ëfóO¨‚T?òRd‰d;:úÛbxÇldYü'aŽš,Ÿ}œ_°Ä'NÇOÝŸˆ™0çP©ù©©3|eÚs9mÙåD?D黣Àö¬)÷šú†ÿÆ?D?iÑïÈ¡y«†ÒèxÁd˜oI™+©}‚[ΦK¤=•rO§w3sœÇÔR¹&‡J›á90–<'3ýY ä¥#u§b‹To~«›{K{RçÍ"píNì7b1(Â)|É,‡D!±‹eü}=°ÃS¤ŒFq®Dª‰#Jù­6‡¿W4Òù˜çÏXc´ÎÍ0#ôÇ—œZ¿öÎr߸¡£Átžì7$ïUºü:„pqš%ÖöùW÷ñÜ2¤G£‹HcpÆ ŒbýÓ`UC<Õ ^X¥GÑsÕÅÍè±æâpwÿLõö­‚IÞ.Œú3N]¤k× <ˆ¨œ³ª3™ØÔ¯ŽÜ°’Ý\žotì`9Vž^ˆ j3Ï;–nó§˜Ç?Ãù¥Pù®zÍàFgà0ýÆÌC®¢i6wQЛ¸eMXßâ-Þθ›æU†Ð3#ºP¶O2ä3E‰­-h'ÕËä±0©b›KÒ°ЙŸøüqânâ)”%ç ±öÞ¡á©âM¨;TY¼.ÆdŽZàhºÖ Ûõ ¹É#éóuûtXÖs›QANiÎ꽯½^+êí±²ù“‡t9l󶣣Ø]–HÄl.›Ü¨ÊbSצM¥gŸÊÑ4˜Éý‡Ìà ˜”-ýŠ'*1]æ^3=è5&zACB‘·Bë§ò{‡u¿šÜ¦ýñ€®ÀßNÓ xƒßÞ:}EÁxFZE¿8ž$gM¯BbXßí¶xj*ô I‚do“]õ‹'‚c(-ö”0Ö¨ÑryãWgõç›É }9BO>¯–¸V-Pl pJâ¡>ýægR!w#ä¦áN ÀJ¨ÕÓÆK6ÖxS_xq ~•0×X“éð«ùÝÄ -‚‰ŸfE[ ½Ê‰Ìêîaò;Brí!N ŽîÚêÌ|*:ZpWaè|SR7%Ée ¸¤<ÊR¸§„™híåvÜüC"ȼK¥Í¢ÊD=zY`ù±2UUÌ¿Ž ×:SPê)/Š[µŽÕøj¦ÀÓɨñm;è1××ò[èÁ)}2ö4V@Ž)9¡7¹|âh›ÉZq“NQ-¹ ž^óú$“K;è½¢-,Ÿ"Æ ðxdÁ âóè‡B¹MÌ 'úãaF‚E7–ž[Ÿ®±•©²£KGWmÔfçć PÀÚqðÏ6¦1ˆÄÃþ}îËv?-ÀŒ»vQaœEp‚bX`£x¨ôp./•„¿£^O ìÉ¸É &+¦ªÅÆŒ6Ü:RG…Åç”íGãGb&Dl¨/­—T¢Ùߤ½qn W‡b(Ìlð£Šq©§ÑQ.½€C¨z„çÏô¤ì ëpšÔÝ{¼^‡n‚Ï?•0t ˜èß¼]˜ÓŒÁΤà?=;¾´°_°Ì <[« ؉HKñÀ_³™0Vfñ‡Í”.5ч‹Zœ/ܱ>«b&<}1¡ûœ¶ž¾V»ˆPL{ƒÔÚû‰·¿m’LV!6#ÂùJÜ+y'gá–-ûl}Näeó7‡%‘¨¤ûŸ†>_á-_ªÅÖðQ˜§¥`r œ»ÞmŒÀÍmß/4D7þ‡~¯#ÆÎWÝÐÎyR~·dZåb9ˆŽ¯ÌFQ\Jû¼gB#²à8 šò„{±&áê胧'(%\=üó*8Å –”ƒ÷³²hÁuM í1î]&Ø“Q<ýjE]A&//3jë þð Ã7–ê¬0ZÉ]êmߟœáuÖ_}lzùöxIÂ%Û]‘Ûd7ÒŠÃli‹1!Ô@©g‡ÏÛ©NDrj`:šÝý•OW‡WÅg[f†‘É@·¹ùaËlËbô{À‘ånÛ=fŽ­¥¬9ÉÅOv¬Ëå³4«<¤ä9y»L¦"Ón÷õÊ7DV¹MüþÙs³pZ¿¬C›™7þš—Ö»š)šØÒ§~ŽØ$B÷EÓ{é½.”؆; ì!o×ê“{B Y¾°Ä@]ÖelÁïLAG–ÒÑBÉÛøæú®ïN_]_¿Å¹HyÇ€ŽwÈA}‚yøÄ=RŽ¤P dm½#öôÃÒ†ò ƒæ€~ |håóöÜ,J F)"ÛÝ(gšÉK-ô¨ëy€²_ØùYyû)0Ø£ú¦Ðò;­H+C ëƒÕý²åÃMl9Œ9þï9ÖQÔ âP¸ŒÒŒNlžÑAù#Ûý,-U¹$¥aWk[n•)¤w]HÃ~ƒB§ … £ÝÏƲÞ%aÜÐZäB¶¶©Ò?ívöéß`’ÔªJÊ:Çý¦OÝŸ.•ˆœLÒ}2)ƒ“ܯ eR‘ÒÅB'7³“7™Ú”8xåØ”ô…?Æ]¸q˜Uäf9”°wàšÉàc†lº} TànR˜³nû-''ÀSåËs¦&-£{ﵦB}ÿ÷™®@ác_OŠ1¸•ìPíBÐã̘®¾ †ò`Q§w’ŽˆÛ3ló–É) _.tkãór!×É„Ù*AÐÞ±\ÒK¢(ÊøAX"¨Ìñ÷ôè}†¨qªïëØ›Nm“›B­¥øo÷‘Ùžyk´ù¢Eè0É‹êÒôÖpb£ÛÞ6àL”϶©<“Ÿém`jö‘ÃRêÇ7‡¨Ÿèà’o0õü'- b\Yí»_üù*‚W›ÏÅ\Z,™.À ƒÓgŽwd®]BÏJ&¦ÙlX=¾ø-X7¬s Áò‰°BÐuY3$bŒ“ÓS–W6ÄÎ^ —€z‚nbÖ¨L4F±yu3Ô ­cô8\­/Ð~ߤԇ‹H~füm-OÔÔóm~tlSº°XÀI¾¯ÇÓóÎ~äQVO¡.¤€|ǽ?´m&w«Ÿ»ÏY¤Ž4ìè"Ò?gà¼p~‡VÕlŒ•PK§a±€@ó]&cgª+Wˆ €„àê7+Õ×\y¨Ÿ OÛ&FÂð‰un¼K±Èù)l£KÅöÎKÛž‰¿n6â¿éúûE¥³d'©Ÿ ¬!½B{²ÏôÒwºâj/²Ðõ¼1kA37_à•_Î ìpŽ Øíˆmth9†!Âq—%³¸¿µFèÆWyLHpK*¨i9o8r›æÞÙïÝ6 ÄȪ wCâ™ÂÞ¾*M us;LØþáöœÞæ¡u ž÷ZÚéxG ‘KDþë/öz?ïyÄ#i±¨&úw¸ý!:Öå²ÆWh³ÈÑÓ?4•$àŠ³ç¸ÆDÔB¢KFÈ°ÚqM}o‚+0.ÜsOPÛÝXâô€ãuégB—7oÕÒSê‡-õiŽ 4×{àYÇиÚm®£|ÿùxÉÑ5TK­”ŽLCãT>µeö©üŸç6ÊDIùqêVÍäM¹pì"Þ+b;ÏqCH˜­lKÖò¦À5*`¯‹]Ø0=ÁL"™ûð)ûÆÌÄ€ [VáBõZÌéqËå8¥ïñ\õ×q†SgÈZ¡XÏø,Üé¬{>i¬ýÊ@œføÄ™c}Ñ&ª:ÂÏùücãÂz9x®®Æ¬vÞJ·RôR?’õ2“JP#ñ0ñ¸*'¯T‹{­Ú:øu3ÂÛ¢¡µÆˆß©ASHuÙ%CóBy pž¥‡=Lš+$'3z¸îØÐõª6“t.7b+5 &º¹ üÍøü&0Ò¢ Ãi‹s‘­qáCŽTqÞÓ™ÈÖ‹öv$NìÅE- ·R=ðÆ”Y´¶.$´®Éöàïlÿ¿‹ÈMBØrÍV%h(.í\µ)\˜'Ú¹ÿ§Yò±¸vA¤áŸÐ)ÐX²I9 U!fÞ”í·ë\V>vm‘äþQùÈIºÂÙI¼ª,®êxyOt—¤–!oï±ôù4dîñæ9Œ™§ƒµ(ÈØN¤µ¸r³Ç0šU¢L­«ÑµBŸœLÅ!päÝì{#ƒ<í[2H¤4掗HíÅîà>‰h"Öï ]³Ì\òó)ð}n¥¹5,t¸õI|Fcµ6ÅxQœ e} j‹¯…›,s¸…ò¿Óùþ&àâ¯ü$œëµªx#¯qv ƒÂô˜«”»ÞX]-xaÎ÷榇u’‹A?}­~LK¬“þdÏ`=¶ïqIܯÁ·+ÞŠ€da&|J>Q‹Æ ϪqòQ©"K}üe6<ja‡‡ [Œû ¹\ ±DÐBcÆí0'~\âòì­â˜;H¤ºgwý Öï± ¦u•÷v¤Ï¨ºÚé,‹K§ðÞ±=âž´»{NÝê©nlŽù)y=!]¨íHOwƈþ¾pÐËõ{þ¢DoÔW.%¼)ÎDñ‰ ©LtÿYäzÞ‰u 9±^qÔ¥‹-åÉMÊUèÚÔÕà=èrU2 ÑÄçKäÓF0 wQëà}|ònSÉX4žôجO-H„‘ÿ³ëg+îTmKî­m‡#µkÒpa’^¥L¯7§\ žò(ŸW&®^’=ý´_¡…‹XHëS<)‹JqækwnÊJCü#=(SRÔp#¡ÔªGwé–‡ :Gyn›¨„b†û „+b¤ãß”¹xçcŒÒÜ) RÌé\þŽ¹UÁO=–Ä ó6‘ ¼éø8OkÎÈî42\Í[Ó†„_­+ŒJL¥>ìX)5*á ]Ku¹Ü^vETü@ ¿Ø˜ŒœšÏîd¹ÈW´‹†Š.† bMhHè¤\*—E¶JP‹®’W¼Iº[Ç‘{ G©m—$ ƒôÒˆÅ!^²ÏQg³Ü=ƒd"ðs0Ð̉?ZÛ'î&¹=Њñõ|ž:ŒIÏYÕ?(¯‚Ú¶o<^j•ÎÖÅ¢”¡oKb2²>Ä>ºCǶox6ÌníŸòç›l³EÃbž{Ð÷Û@E,'»ï”âÆ{ò—%v@عÏA¿ÅMeÆÌKDBç„»ŠËMOM]úÓ0O´=Kÿ#™YöCÜt¾XÐ7{þKˆ¯´âËóE¡Æªý {6A|uóẩX¡~²¤Ô(ÄÚi¯?$ÙnêTtï« ú#RzŽ;ó>^ÐR ‹í^ÖÏÃÎYaSe3Ù ú[þÓË*HÚÙ30ßwrüšÍU _ÁÞ¬†Œ“ÔH_'(‰M4 +¼\¼ &±²ˆž =/ùD]QK'",?»Œ«¦Ž*(‰Oz5>èÔ(Û'uKPÃ…sBZù@äµNã¯å¶( µ¤?`r½”ÎÁÀôÒV%Ø y²3úF ž‘YÊë#ü!ŠŒM²ÿ-=AImÕ0ÇAôL*ñ-FßG ¤‰‹aßÒÈVg©€œ!‘ ®›šdU~¤Ê€}ÆìExÐòcÄ%Ïx(çÁ\¹Q}øÍWç@ñ+Ü``ÒáX÷ÿy.=Wƒ7kt†iÖ^ Õ‘À 7ÀGÅB`˜´\X-×2Š=F>ì'Ϙ@„×Iƒú§ô_ÐTÃå¹ÓMÁ /k/ $¢w #,ûÝ£ÆÌSàoîê%ÿÐö¤?ýýò&šëtö-Éï ôâ? bL:¿ .˜,z›âtˆ«A…Ûp-ó?úUõ8o9À²AêyñSð³V”¦à^–Z%K8zÛ˜ƒW09`D.5ŒZÆ•ÿÝcèORä’÷Äò`7¾­‡6 ikuO5KB†üýëÇTRš¿â¤]Îö=„ŸRÈ059hžÕ,ŠÎ aêÕÁı}³“²œ_~<@n)îš[yó¬®)NþׯrÃÃ5®Û!õÝËÕÊxÛ䄘Ç&@ìT Aøw{RDP„?Q±R˜q5U\Lq??+Y5#eÕøL#±ø‚q«ªvëð{4ÇÝ<®è+ bÖc¾É­+jâ^à7+“5ä·[ÍdÞ_özéÿBŒ)Ë è]>+¯ÅìàûÑ–ô£V®N¼ Wd_ã1é›´ü£qÒ¿¦©s‘lq048Ç'iòiœ¬˜‚m ‡×¢"Ð?Æõ“Öžþ<«T_ÆèõßÌ{ýÆdžcÕaü,Žcï ¢äÇÖBk1†¡äBO™…‡µémÄ*€˜â§&zàLÎU¤¦eÀÐ/Ï(¹ ‚Iì&Œò\¡ze ·äë9˜ášÅ„ ïW´ 0â"½Þ¹î²´Ì{»hÜDoߪ;|IR÷ysZs?LÉjs²Ê­;j?ãÃh5Fã[è© s@L4MhÍ>Ž—½´W@ê‰<þ$Á«ÁÃu§Êò€L˜ÓÝö< {™v]Ë÷X&E{9½Nvx®ÐÒ¿½™àá ÖH÷äÕÕkÙOgÅÓ.¥ut3â#Ø°YÔ! oð r¾#(Û“¡É u¬kBªu¶(‚èÍí2à㺴e¾J¦j ÉHÈä­‘Ž¡B6.õƒ®<Ç+FvB s·™OOeDjL£§dÔ÷FEˆAÄö/{íÿJ¤ò÷ÆÍ÷ßWœë~dugΟ /¾¼Æ§q·¯ñ1zk(nÃvÌ]¨6W 9ÿ0†çöÀê ýš>‚53%‡ãA»±Gž‹à…sEÙ̇¡–ëô«É—¦·Dæ±ÐO)óÞ¬øºEÌ‘¨y–¯[ùJ„üÙ?¦-­Þ‚¹ThLf‰ü+ê ”¯_ˆSŸª#ZËrì Ñ«*wÛÚ~ý¥%4ÿ“p²ˆí…UI«Óóå…ãú…J½ÕÑ‘¥­¸jÓÙ[±V,|VGþyYÆã åÏçÎ.Ïd“®(Jíæü!eO¨ÅÉyÑ|̳¯¼n¼ùÌ[ +Ÿ’ýÒŸ1Qˆ¼zÖ>VƦöÖ ŠWõ•çÙ?3÷\›$|Sj'LS.ŒêÓ vƒ’V{§t‡×MJǯ‡ªƒåX_½wmÅ…œì&íM¨0ÇÞ± ã/\Š§=ì_Úÿ{ãI fŒîÍÚ#4Â8Ì€”Ôî¡Ån5ÚfQ°a°)|ª¤‰C2« ô®—¢ÜðË+ X÷à%ÙŠåÌç¶b"êh³Bí4lbŽ¿øûõ¶“…íåÈ1z(îr‹¸y²¶[gÌ'ÚÇ^©‹#töYo»Ý@´Þ‘07Ü’bŽ«ßÜ“ÍVðçï¿û\ïçó‰4s ít¶ÆuÉý—A A››µä©î§·\]qw^Ò|æ“‘SAè*σ•ªîuæÕjˆtß½s1‘ŽüS,JUëˆ=p•Åvºx……T ;3T‡´»š(4$Æ6±6„}^µ =Ƴ ãªîȇ è°®§¿$|žD&}¸BxR2å &ªï¬¢òóq•1®ºaÙô«åËJ³t h¬¯\öH'ñŸ”ê:Óúu÷öL`ÝúZkW0±ÆESÏ¢b¹´PAo‹@þôT -Ù!8:+èBGî&³›²NÙg¤“À$‚TÜS­Ú±1ÙÚ ä‹}ƒ¬"rgý>êtÞ™  Ñ7_ïÞ3;¥ª`%ë9Õ@…|"$¶ßV8‚”öÍÀšÎËP‰§82øß7åKºaȆÍî·$Or=†\4âwÒ\Ðœbµ½«Ô’¥{û.PÏá¼×Ňu†Ú±¸ØÈ §A­»Y‰uc‚ ÜUø£²jRƒ/~lXwä‚”šAžbÁÀ•ínÆ‘¾E‚‡… ph‚7 ¨Û)TZý÷¸¶A!«ÝªeÕÊö|Ûû=ÎsøqxFiG\ÐÜ$ׄqÿw“çï=òGi]²8îröïPöøÜü6ù}çzzéÇz+püj–Yf8ayÏ—iÍ~¨«â%ßÄ袹õýQšCú¾Œ(ÞOÆl¿©m“=*G'ZòÀ|w’¿'ð4Ž×¨2AM‹3Zyì ¥Ò|úiPÄeÕ'úÔ¾“Þ„p¢¢è’3zF5t8üï˜X™ûÈN^^b°ý^†3Óy|f òÑ-ì‹y'—Ø€b;³àòóDcBxmƯµ×åÃËV’`Ç9ÀÐZo“[Ç#‡ûá2sPeOh+UÄ^—ýà;û퇓–“Ùú^KtºÍš)Ë*óðÖ3Üß\ »ü–ð9£L¾[Xá+Hv×pL˜¥j¸YA úþC†ÙÑÖºQ{‚ø5Œìá0’ ôÓÇÊ d•¬‘ yWu¢Uñc'Ç`bX¥µ;~*щ›ƒ‹YDfΉB£yÓµU½ëaw¬çUÌPŒ^áˉÁ:…Ö( 7ª]óåÆz…3Û³exQ®gnßg¼>ï,rLàa…vÌÚO´ã¹û9wPw5ë ‡j·Ì™ÌLêLýq £æCgIìNnÃ2=¼Ÿte=µ|å ׺`Üíš’Wœ'AFŒm©¦SÐòOi~ž‰.ôÍ:;{³È'}nÆû˜hƒ;‘ðKPþÑX_÷ŠZnæð2)t:E‚|´y²„žöœüœ†fܺü” j›ÄŸ°–îO_—,)Ç]ÅXçˆ\!i5-"X*ð5+Œç2Î`£1áÀTô›aµ¸‡’‹úb^°Î¤ïÝ£<£0KÀVãŤ.— Jf7`*õ] A›&¶Oùi¿Š¿e8­òT‹ùÿë⛿*¸®³íɶÍÉv7»nn²­©&×ÔͶmÛÖdÛu'c¾÷{x×zžá¬sö>{­½wL¨í–¯³#$mu\Õ.å‡!@ëèO¡È|e'ƒÊø/é|‚ŠS‚¿B8$þ AÆø½f'Ï•tJÉúÑšð1 Ë„ø`)uÒDi½öyC)…w&ð€b9‰³>Î*‘9xØ€)€èíºrzÜvÉ÷ =:oAŒ9“0¨·©uú‰ üò½ F× ðéöIô†ª"Lí:»¨S;¥“çýQ4×[ ÚUü¾©¢ J †ÂLž0 4‹³ZûT)u6Êå*PEÉûÛ$‡(F2¡âÝþi秚Y>1êUæk‘fF]=7”ƒìá›$kÚ%jvpÎÞzÙY K5Çfš Âd]Š™½ªÂèrc{VÉ1]µ‡Þ¯ôPŠ‘ìÁ¬sæ}?ßÊxs=¢a™‡…-‰š{–x± K*KÊUØy¿xœ¿6ü½:²ÃÎÅ{£®¬Ó³‹ë¶è»eJ³¢`vèG›êGbVí?ÊLéçÇ=S–Sœ¥»H´T†ˆ­LpÓtì‘ðo ¬JàWʈ7âÃAó¡ŠönM¶è™sÍþ΂§‚´šíÜ~ ã%&ÌÞŽÇÛrF–Õ–¾EhǬ É/¤ÔLêsHŽòíŸóYσcÛwá¢Uâ±±ãÊòÊgñ>v®cÛÿ2++ê'å’Ï´ÚÛº‹mCü k¾É€fĆ\lT>#3!‘GÈ“s0”äæïbCöQ”‘§åF’=×¢}ñ¬œóª¥JClÉìÞdÿ%0-?+÷oÑxq+F»v‹§ ÓØyÎ?6äƒG8™Yem©¢Þ‰Œéƒ“†äg‘f¦@¥‚ߌ†gP­Í§ù`Îæñ§%.ƈÐäM2ë/‹![ðÆŠÂ1TVÞÛôÇ’ÑÎ<™ ÃMëOñ±¥ϘúG“LÜv2yvßëD« þ$vÔù{ë•þápt-l&óD¨î8-†2CßÍþ‰÷x»0ª¿J#Zøâ·ðææ/¦T›¦ë’ºð™»£¶ìD#Ösµð6‰øCþµì2Ì·ç+ÅK_@vµ-xdÄÄc¦ nß$ØõÇ©VÃíq$éïÙ* cÿ£ aÙuÓ.ÒM’7è=¨•u>Míô6tªÍtÕLÛÌÂ…0Hž}­säyóühØRY?sh”2€å [— 4Œic½T=tËA®D”óÄOwó'Ç1MpévðE#(KŽõÌÓ&ÿxýÆÓrÔÁº¸anÐÄ¡¯:c³@ÇöF| ©[¯z‘òÊNŽÿ‚‹\M"òJwpM½— •! 6ÊÚ©¢m}ÈÞbf 9nB ÿK²”ê4E„ö >A#_T\BœjQ“&G&™ñz‡ÃgBÌ*ÇóNàK¢ò>£BÖ–þÄÿMÝšôŽY™‘~Œ¹U¤ÙXE ¨q…P½†¾ÂýK ß;]Ï„1i¶>·ÃĘ%v¡³w©¡£Ž'z³ÚfJ€•³itœýæÓÙ“‚ÊnœVƆòÆ?6L»2›ËûÉÏZK×®#œG§ååÓ~Y£Ș€ÚEPžÈø±~Ž±3d¯g³åšéÅÏê€ÓŠsÍM“ÇX;¦ÙŒò¿¾ãwŸÏôCä1Ø%+ø·‹$iÈãm¢°%ºv“¤ðR§e›¢Ìz¦MßcìEH_ånMjXK>ë‡þ2ó1ñúÛÁ%2açæž~;÷Þ N%D¨Ò·œþm£«õú‹qrdM0>SÔàÝ ùÕ¬ÌnÌQŒ+_î% QÕyóJ¸Ïiý§ò(ù&P‰×Œ©#NH\¯a¬BÒqÐý^In¯AQ¬Ì‡CÉH:apaùÙZÀoev5š¬}¨¤Í²{Aàñ/¡RŸA3ŠÝ]ÚÖE¯ç*V××0²<ÿ‚®‚º½ã,æg§K.ÈUägmW ý³··{Ñ—ÔÐkraÄuKNÚú`r¼[Ø6ZÄô¹üÚ¦ÛjXn«¥f„îí7ozs“§¡è¡o¡OZ Hîømœ è\{X˜ÞÑDSÚ(úÛ4Åñ t”_R0žÒm?D=ñv¸“ô«]¸|ã|~*ýÜ w^YŠJMo78±õ^µp ’3®þÌ@d5™íÌx߆Tm§Ä fQk2T±ááH°®]ÉúçbBŒ‡=n©5ºäDÃteYž–dVʾh©ÇØØ«}QÎNµÌW42ŠoÿsâÇ©øo$c–@±æ3™ú. ³ÙoälðÆ#h›â[ë,Ï%¿ð›‹;Üí¬é +þ¥ùŠ*qŒ·Ìo:¯Òºòë®ûÍÄH+.‘]Y¼_²Üfá7(òb»¼=M›‘õéý—»lÑ%b7>êº ®¼©ñ'8Dìåj{ÿ9k šÑÃÓ±|Çí;½'Eå²hBòõkÏ#]¬ÁdÁ2-ºy7/‹ºþ²ÈÊdª¬áèîDZ[BÆÄ™’.OŒ'nÕÂVë~¤ÞŒ$ÕÛ4Íb¸·A÷ÏY«SGCj$¥q5ÿV¡Q¼NQÑKñš„±ÀŠ’RGCÀkÄnjڽòÐxDÁ½ý#:Þ–Í¡åmÅQû˃àêÖT9®IËâ4TPØWà—s£Ó|Ìp¤+€S1²ù {»K¸‰SþS_Ž™e(YÌP odº9V(Gû¾ÁÌ°KÑr‘¥Á­[îb¡jcí®ÉDP£®`‡‡XÑw¾"ôæG¤*Y¿¨ÞÓCxf³¤¹®TÞߨhÄ;Ú³Ó4-\œîGÁºÏùÊì—:?Ì'¢t;{1Jø¼4œÛ~ˆÿÚHE›çۧʡEã*‹ãOìâh?•§°èþÍç›ÃÓ;ýF%'Ú÷Rm–'…äDˆ*X&|›ŽÝ|ˆzÖÆR‚ÿéó(‚W,‚°kTzaÿ²(a©YÕ‡O²¹Ó)‰à}C"cw5ª{FJË–i-µ~g!Є—³y²7†y&Û‚•m·…]Àg–i‹øÁBõîÍÞ²°àõ8 ~<­ž^ÇÌpžQÏ !9;G ÃÁ/ÇÑìN¢ñ‘³ŒŽ.>xéý»ŸN+ËO1`­ü‘ѨK -ò ªpeä*w÷ÿMc:Ðó¬±ØZŸ°À¤y¤_ÁØO²øÄ¥SžŠØ1ïnX«y{>ö ®"®ÈCÄNX†°Â—Çþ½ÝAïã·ú«lPÑ;|oÖË}ŧ Û¸I`‹zþT†Úòº½ìej/ÎØ;ƾ ‰‚^º® §z¼:–¦•ü³Tr¸4ÙÑåqfšFÞ|tª9Ê€ ¤0%*W‚kû™.ð éR{û¶#¤ßî¢åþDñ.³Ùbßw3û;-ä[|ç’ñb°º¨‡w.2;G~ÅÅZö€Zbš¦ËŽ²â9Îsn ·d¸Õø¶½«Ÿ‰×½¸By ƒ´m¿,A‹—KûÍ/&…v)ÍkXC¾´&Þ¨¼Gxe¾¶‘|öX™Uáв½v?(Zwiãîd\IegKo7÷æf—:X ?ß³ƒ>gŸ?J ËE5êõ®,p zݘ6/X„6›s«Å™“©!#”Þuî z$àõ{u–Á£7ÛߟEí˼¿F‰ßMÝÍ«™žrxh—÷«ÿD ]ä+5ø)œzýÞKçù/Gt¸±Ÿ\Ša±Ÿ?ë—Ä=7izðd‘Þ öÏ“s‘ ÒÆHhoâY³WÌîBëÛÝHþ6ÔÊ7$xý…F5\8ƒŒáßJHk¼;Mv9öúF»#ì?þ|dæ=Ô—u½¾~><øBˆ¹­q³ýåB  ê÷ìV;L¶¹ Î4÷¬¯à7½KÖ 1aØ=Ï87üfÏX}Àø²_»(#œËíÎÊG˜DlwÉíÂqóçg—üN¡<&¾Ðm‹e?“æÅ_øEäQ3½_CmºDZ=)k•²JenßÓnÈ-ˆÐ\ôÜæ·P¿²l³Z3 ×6à\¹UD)Ú9è><Ý(ÆíR»ýKÌÞ|zBŽïíï'GŠ˜õ²áˆÙâw7ò@š,Æ2òE˜QFãkb|o¨ ð3K:zÖ_¦Aúz«h³<¢/x}>ÊŒAm‹jy.ð€‰jì'¬ÙC¶_-o)ô„—¶<¯Ñ¼× äÿê*;¸£Ô]é#GƒŒ˜ ÷%ß3† âš™Ÿö'œ(÷1wY+^–`ãÑ“ç­ÁE)¸šá¬Î‡‘ƒ´’èr©—ºÛÜæm®|ê¢ø*˜MrŠCëöóÏ¡­ˆ—Ï]`-w$¨¾—Ö„sÚüá¿ÇW‘$dž†ƒ}î]R{iª.Q2k¤(?öø ÒùûY¯UVq\Ñ ýÃér4ë|—¬,óYé’³ ¤Ù¥§Ö ÿê@ÐTüÍêÝ~Nû²YÉN­–Âl¡qr‚hlC}j"2¢°º»V­òX8Fwå5žs¦S!rCÑE++pY_ž˜à«4uËQ-óé’6+7hËpAæú•!ºñŒ;*ë6lftÉ)0³¢ Ï¿Ä‘}\†;´¶ÕIííÙEÿ d¥~·›Ä&Ó·ö1Pf pWžÂ¨,ñ.ˆ,;î5²»É`´}|û¦YzsŠ½Èˆø]F‚02Œí“Íùì|I…+‹ë‰iAO{½U†s=Ú¬È/ž¨d‚*–†ˆá7˵çŽlVÐ õŒOμ6Ô“Î¥‚°Nõ%dÕÇÙïsŽÝ]¬ g…?*aò0¢KÚºî—mª€ôwѨÚ}Vú G¢Ý™;îÄ÷T;ÇsO0Xß©3á%IëÜGxÐ${ \˜lÌ«<=úðßu7`ic1àž…’ ¯%l¥Ú}D6 .µÿ¶Ä¿è¤tÖõAŸ8‰Dj€‡,³µG2Qu//±ÛCE# r %:† …‹¼ª±þ—[b8(; _2^rÑ5ýqPº•BѤý¢ÝegvÃQ²z꘽ty`ÄEoP!Ëo‘O2=fQW?#G˜i9fkÐ]½5@…àAªêü¡B+êòkàuo®‚tu¼Vqðw¢•‘,§}ëé׉ç¹. ¸Šnžßàø…rb0/Τõ߇~ÿqÍ'.AnNTó‡Ê+‹/Rùh¤ëõà:˜0Üv‹$:ä‹5ä›­nž‚‹TH @ãàÏž,YßdM,ãü’’$ü㜠ÕÄ9RŠE†º²ÞŽö#Ùö."þW»IÝ`ò÷®©ó!B0¨óÉLòäkæïuêþ_ï¦Å4!v %ajÌÉŠ2‡GzrŒMÖ‘€-ʸ"V×áÔ“sö¾^{$&âDì\9îXªÃ¡érP“øT1©[qaëxœ'4"3!L}᫾ì¹ÁÂm"q]Û9lЊyeÇñÕq~ü6ÞFÀÚCX>^DÿR«‰ ø2Ký¦=„%;  ’i{ƒ®±ÁçX½¥Iã{B þÖà·L#†Á:›C#?]œ=:Qg± nt:¿p°.(ÕÞ­Û,Vó„Šï2ðŒ›ÆÁðh“3b«—4ºáu9Mæg "¢÷©°;r‹FK‡ÿåñÁ[:™4¾;>i÷é´Vñå-¯S'î“¿JÈüÛŒ=£²Å8SšdP%Š#Ñ /…¾Ï³Óªý O9cÔ i•ùÖ» ŽMJ–}GxZ§‚’ÓŸ?+viüòÇllt'à¤>/`±¹{k7íY‘cˆ-ÛxÊ€­óðž ²ÔE×û:²·È×$º“¼ŒŠðsÜ„2)Ÿ™Y¨ü¾ª~Rñ¹«<. ( …ÒT2|qx‡Â)žd§ 4Iˆ#ðÍ“hú1÷‰Wø ÁH~Ï[X‹ , 8¨iêZÿ܇–úq©û £³I.%0‹Þñ3³q±’,¦¿J9Áf¢³›+¦ ÇÔ_=H|Böpú%“_(˜w¬…Y‰’k Ù½wƒÞ¼ë‹0v««¨·“~Xï{,‚Ð&ÚÌhJ~;¦ X¤Œßñ·)3cE×>΃Éöåù0ßzw{Eƒ\íºLsÍ@J,î#Àð“­ ÆûÛŸpî-žúO´F#¹YUÙxlÅú»â^„²”ó©ã(M¤…Ùº…Ùº²YkB*›MÞDz2ênªOf·¼T›'ƒì7¯ÐŒ'ì²Sô°¸¥ùX·ÍèÒo¡E丌Uœí>]3ä©;³âðÙûôtÔ¹Îù«Um|v0ýÞÒ¥\‹:BÚ¼€œÛ¦ÜÖG„fdž’{¹)íBKtZ‘ÔP‚³L ¦–€ü+ŠîÐÊ0ÒáÓ(àë*¬c¨‘U¯×-ˆèAM°DÝV$ÈY×Æ©À¬"Ô6ù!#kõEwjñ£˜¦¦†MM±q°·TÝ|ôƒÛˆ˜ ƸfŒí¯<ÜÏký³ø³ôñ¢’ŒôÏ„vö‘ƒõÜ¿¡p²'Ffà_0h_ë8bd½¤mp~¿êÂâéè´R%1 KWn’›c|Í>HiëÇÛµ×ÑÁ8W"!ú°RƈKF)ËáÌn¦·‚Î ùt{W™³îúÐdfó¥ a…šhÑ[¹ÁóÆ/dôÅþˆŒÿ¸ÇV|0¦n>ÈÄð$dŠÓub}ù=%~ ××$g —-àJ–bT…¢äò|CÓ>yVk-GNG^Œ¥v (ccWê ¤ ägmªo¹hÀ²™î®¤¤íW©g%– éjFÕq¶}ïÿ{MzBRx©t¿F^¬n½G(ñmÒov~Œr¥ÕÓÞÓEË¿ãµú¡—àÑ£H…¸ ¥¹³ýÅÐÚ¬Ú·íò غ3iÄ O–gµi«™]›n8¢SR£OªwÁtRSëgjpOWpM Â(”d1±ÊtqºgÓ¥š±£ uõ$ö<²ÞïƒYˆ´»N{ˬÈõ·ë-yŽ×™kõ³J€Ƽ*<(/˯ß_›¸“D_ÖøQDñÞP>K4:ô•(ù+-Ž5Ôî*ÈTïlnr+çÕ7‚÷˜¨¼'@d®¼˜«u+Æ`aŠ0ÍÊ^©Pk…²ôíjÊÄsƒ.sÇ r˜€ï:Jõæm÷Šëz•W/ÊèÝæ™”ç&Ì!s»’1H-‡öa°óûÎþëÁÕ©ÁÐк«õ Looø S=‰’9:Ü®LÐ…¥¿¬,|D8ä²òGR,Šuu†´VÝ.NB+Ÿ¦Š~«­§êc ¡ÊÈg¶ùÆ4Nréx)E‹-Âð¿Hõn~­ *C&èt÷$@Š“¿/*œi>$¡2ÎfUÏ…§SÐÆÞ»úÊçãðxB [X™f¤EÕÿºAöG(öÎ=BÕ™„LJY§}$Ÿ§ûÈÑ£í¶ÛáÂ¥þ›…WÙöÒÇ“«ÛÝs‰óàk iˆ ˜*ßx«Êý²#Ÿæ¿ÚrçñÖ:áÓ¸’h„ÞuDZxI·ÿKŠ @ú½Smí5.׊óï5(5ÚÁâ(/ŽF˪Úv #Ô$±Mì>½ JÀ3D7ÎPR™¼ô¾ÖôdwÕwŠéK&ê>oÎPÉ• Áiü5Ý9Ë玉 M1ö·'+ Ž ¬ÂíÄÀY5 Ž*JáŠñMû„ß6o7 ]µxºÒÁ4}0ó%¡ƒg 0õGC N8è··¦IƒD¦…™Ö‘$qI}.9}h\ Y¾Ï—ù¾YàL\m=÷XÏ\ZH5JaÇ_M@ò·ê{àfìÉöÀ“ˆÙÏ{ô‡[Æ.ë)ö5{ýŒµz Þ< ®Ô)7[,ëȲD¡à™¶ zŠ!!f¨…x±›ÊN–ò¢!’܉å­*©ïÌž†ù÷àQ¤ÖaݗܱZ°lQ·³xn’üLâ,uŽ"›"’øþ*upýÝ¡:c ü©6ã…œ>áZ·6aóÔ^“2FË£Ø?œ|YæÇy¡ùeúUiµ's'ÿ^NÌ€®^¿wÌœ§ç„¡ô®yÕL­ü N 5ö5¦X´ff2´f¯ªìà…F’¾ø{)úî›»¾jd‹8A+¯uôtó ÈùE'B^¯k¹}ÃØóG)/dž¢—N×á7â¾’úó~ô}¼‡¸8Á_~¥«õdQ‰0Ù•˜Þˆ8WLs;ꈜ4}žLG [Îç!N£g…%cÿ =‘´ ¦§‹VØîOUvò®¨Uˆ¸eÏ“Ö%“ؽ„H/ ÎjDŠæ¶KO0ü¸ÐãÕMüM¶Vë>DÛ®Õz/T§áH²-3Ï¡+£äþO 6mà[~xÊã°Ò²” £*æ;ñ#—þ[½—÷†Ð›¾:}O#’±•5x-Ä*ñ{qHˆÄûV¥ƒ‹¦]‡êöOö•ð=ŽÚŠdRH§š-ú¥fléw#„rÁ¦êÞ[Gç®ç3HRˆ˜ÎC4T”BˆÁ©TÄ“„5ªn‚v~Ð’ÄUÛÔ'æi)ÑÿvƒZ~ëJ”‹ê‘PÕÁ‹H±ƒÎŸ.¡eäÿfçÐ@éê\öoá q‹¨ö½êü9l=Lo å'"©^¡s†ÝrO#ù¬÷{Ùh+!™§šD¯$cÅZpÎ48¢çl:M½ã_ÝÕ{…N&Ü/–çÈmþ³Þ§!ÞâJ, §ˆQŽ–õ8~QM@õÜžÛö´eÍÑSRÆSkŸ¸¿¨À=C½±€µ&óþÆ4óÓ£Øn›²q…IÈr `ëe™‰Âše¼¾Ed<\]@ä ŠÃš ¨JõO hSµDIý«AqáîkÙSuc…f%¶ì­í¬„ÃLEÁƨø%)Xž_+:å5s€EòEs+=}†Ñ‘¾Ðg?dváæÚXÊOÑn‰qø`NSµ¾íÜ#/lJ„¤²Z˜¡ç‘g¿ŽRæ^Pëþ)ò­¯ÅÚ»+š}íø-¶âGkš(ÌÀŸPʹ•FX<”¥Z°‚‘¨"ÐuüAqw }C 8ë™j­N? &!,¹–7ÍWZÎËå=—¸‘rÜå@㱜y+VI½Õ}ý•nÍD€Ï_Ôü[Ý:õ28ûgUÃhâåX :ßIŠC ÿ¬iuö‡×X38s‰ão° 0­‰wªºïŒ(n±gcxP‡¦±?͘‡6‹X£ätøü¥üÚÝ%fÜ,Qw¼:¾6 ©¿-¯"G±âÂÊצLÌ\@™§èNXĪ½æ±ø–ðrw‹Û&oV0ÃÇóGGóª(D®/¾½HFõ<5Oy(æGíªè¶WïœÞ×i—M™§}i¾&Á*²üœéŠJRÓÂÔöáv`ŠEÈ;SR7Ôcõ=¢úÐIÊ,r5 ö܆û©9çoôv9Ûgd gˆ t&Ëbc_ÚÉ°Ž‘3²<—dãY¸ð"õÑx"#¨Ø Sñ,©l:¾qáâ"ØýÚ"’ë‹ñ‹]ì¾#´d%¸Z>ôSi[}˜ˆXª¦—x½Xt×r#+ÈìøD¨Z@“(;áÚAš’ôÈ“YÏÃ;µû²àó\ê# ÏÖô„°‹ûw^[lÒš>ÝiM‡B—þg-´¼ð?úZŽÀªÑ¯rߧ÷ß?Á·:ªõÎç¡6ÈRJ/îsòë{NeòoM¼ö†öG-ö'8‹^ˆg)XXøM'AMØiknÓVRu;nIíÏÆÙæO5K³mˆk‚ÆýT7ÿôÉ/߶%Ó„Í{Žì…°c}hK' *áDäJˆ®bà•z±ß KÈð(z¥¥ç!©…•!T±‘sÇ^¸nˆa'È–ª->&ÖHij]k,AQRû­cœ Pú¹q¶Ÿi'®eãYY u³Ñu8ÞH¯C"Jºâº™¤”ÍàÞ¾]„E‘¾î«¸5Îø6Ðಠâo|úýwìZi£je‹©¤Å¡œ99ü fJÍV)ŠS®Ë˜ ýö:Ÿªúˆdô Äæ'”(Pä߀¤7fTÐ=ýï¹Ft<¦ÿ8·ö{ÔÛñl²·¦a¦Û…–·“¿¿áÊp´”ÄÙ%”® $Ù‘us -)ï`ůÔ¦¶ëÕ–dßêó€v¦DÙÍ[+O£á}{WaA!£´¡×Æ`è-Ÿ[¡Ð¸ &¹EŸÁÕç$—¹q@ö6¾ðAj›ðžÚ‘B¢hÒt9æHE£É˜6Dá-c¤%Ê‹‹`Gí-Çì.Ü*;rxÓLÊ¢ñòðâ‡~†‚â/iêS#ã:áKA´µN\ÛÄOi¨±ãEø¤Ô€oÀ®4Óþ}¨ÃBÒáöa;”ÙîÕžý,³ŒúÜ›ôíÊã9 f¤GßÙƒZ¿PD]h©N$0áo›x¦åˆ¯ƒøRµ79=‘•j—DíZÒãúz¿C½Yêïúî¨eüZÜ=2ììŸó8­â5±ÊvE ê“féðÕšfñ.f-,E ù},Ûb>=1zè"Pöwí;ŸÒø×7ÒSg¼Sîåï=ZÃJá2á ÆÑTŠ\.dä0é~ÇÄ\‹Qk¨ƒÒ Nx(é^ȯ„XµÕ+…ü£וœpó}R•†?zºÔ½ËBºß3áf‰ä»ß[KcqçëH’u’Ñ­†.Æ,¾Õ…SìƒíÛ¤þçT¹ endstream endobj 70 0 obj << /Length1 725 /Length2 20783 /Length3 0 /Length 21336 /Filter /FlateDecode >> stream xÚlºS”fͶ-š¶mÛ¶mÛúÒ¶íÌJTÚ¶]iÛ¶Ui[÷_kµÏÝ·Ý6_úhÑG3ZK8Ø»ªy9˜é™˜xˆT%ÔÔ˜Y˜˜ˆ˜XaÈÉEÆ®VöbÆ®"M€‘*À‘ˆ™…è3 9‘¨ƒ£—³•…¥+•)õ¿ƒD¶ÆfVvVÎD¶îV¦–D|îîîBî.n În ÿ©D®–"s+[‘¨¢’¶´‚$•¤‚:‘$ÀàllK¤äfbkeJ$ge °wP™;8ÙþÇ!2u°7³ú'†°w8»þCÌÜÙÁŽH^\MXBQAH\”QM”ÈØÞŒHNêŸr{WžÐSW×uGGäú?–­ñÿX&ÿcýOÖÙó? 33‘™•©+‘ ÀÂʆñ_ºIÛ›;qü'lææøßÔ?„\þáGDõ†ÔDfóÐn¶¶ Æv"*Q;G7W€3‘¼ƒÀÙžè_Ú{8[ýù/ÔØÎÊÖëÿüÒ®Æÿˆ!loñ|Lÿ Y¹HXyÌ”¬\ÿÑÛÕÙ ðŸ°&à?{#0³r³ûO9àMþÙU3{[¯ÿ³â?­ü{=FmIE1Úÿ‚§ÅíM̬ì-ˆT]ÿÑÓØÙìÿN+[ýg‚þËå¿EÄü}ycWg+O"]&†Ó¿¾ÿZúÿ%"âàéCÏÂÄBDÏÊÁDÄÌÊÆLÄÅÂí÷¿èšº9;ì]ÿ-ú?½ü×ÿ÷0žS˜õSÞPë´Öð ñ¢¹JpÚ¹#uÚfàÕéž]d«!K3ã+œ{5œ|8üÙ þ½Ë]éûãxÜÉäT,q¤þ Îk i¥…íö»rÁ[C¤Á5˜¶©ØMÔkÆÆz°Í[Vîy­ŒS?²$s¬QyG?8ÒNÂÚ⣽$µ>%J¦Î7à€¶iÎ2¬ðÚDðu$"?dî†Ûв°"Èñ(B'ÿàѸù´ý5ô5Á­Q—<™†î²¢¬Ås‘ *I÷´½Šg QE)ûjd\óRÃÝqö£BŒÑ€ÜÆâZ›H"½Ê± (Íø¬Án-«þ¡qk×7µ\/I¦]Â’rᨚ‘ 7v(?ú>ùÀ¹+8GÖüûϛɼZTnE[`㜅èLu{™7©¨Ú©»£rVîQ(o ÏrV¿Oè`;5…ÁL±»¸xr¿\tÖéì&ÄÄ~CT[c>éý³´^§«¯Ô¯©ÃWφx™?ݦ?+lSæSÇæ÷/ÝÔ^FÅÈß6Uàfb±a·Ÿ¯ø²‹ÀØŸûÑ0e€m üäíû÷bê%G~ ´JøxDõèÕÞ¨(»AÔ¥îêó öæ¿j+ße¯#X˾=Ü<íqäJî£b Ï]ª¾òzh`˜“ üTYðô8ΖÆ.C¿’`E2 ÙJÒ¹Ï €&ƒ“”¡„®ìu¸Vr×66»¦Sìs*T”B…!»Ô«6ý"¹ªC·WGSæ³ÐT>zEɈ‘†>¥P~ßßŒæ ±rbB¢X¥äK¨V éw†á,½”¬2’f¶SI%¿ö’H ·« ®ãÝ6„¹é¿ÀP}màúäÆî±*l3wáð A鹬nx¸4¶8Û Šâ1ù¦vÏl0e.Þ?Š=ÌÝõ»Ê×S²ûVc0–;J–ˆ³®•_Ârù`έ7è„[ôÅr÷A?™ÂÚ½w¿Àxà>Ƹì–ÅÖ¶Ò¾½&?cÙ^èAÂHÆ{^‡4ÀOÏ|U!PtK:&ûAy°´w1£-ò{Ê©ÅnvÏS'Y×Hh²"ß9 ’ˆ¥“üFød®8”ûŒ¯ˆãlbWsÈÂBÇ´þè|ìrÞ‚ÁÚ³D )y‘"⻚!1ýBj;£Å}¶r&?ñBSÉ®'²%[»bO zQp¥º™V\‰h$ýzOmiÐqZ+úx)# bo5Ì© ¢P-ÄR±,T2:å;xnÜA‘ì…ÎÀ¡ ôM@×eŽ‘ƒÊroá“ID:òôÜ}$¸“KK7pRѾØS”s¥…m-=V·b’o¾8Y SìYü(^M\Z·n Œ¢%Èdd~»y럻š(ZžÐ"‰à'þIéÚ™’gšla Dø aýÖÅó,XA[•xã@GÔ3 í=¶w@iT§Óþ–œübÖÆÓ/Ü¢ñUiï­ú_êõvQ(T3èmŸ8vù€†]ÿ'«ñNþ|cÂ=fÿn¬D*^r¢t¡@ÔáºYÏk‡2Cn~©wºžÄMvÙ”8?öP£Ë åNx朊&¢ÒQ[ƒç^C¨N>¦–H'¿Ÿf‚"Ÿ/=W]ùˆôè‚>$‹7›Éùˆ ¹ªÂ\ké/-%‡q%¿m`9jòO´ZyßÌ:È(e{±õ#,o@a¤Äàϳ#CžÏá<ìŒezðùíPý3þ·7GTä×\ƒYP‘Ô7uëÃqÐÒd`ÉŠ †LyA¹íŽ-D>¶n öÉÉÇ ÕE `à<ÑþØf­Û¹ÐìT°Õ÷ôä§O¼z]!*HÆ#ê+›q9þ·ºjåZîCl­8½f]J¶Ú+Z쾊í§sÂŽðw‡ØÂ$¹%=íp¸®;y'½Gåz3M‚?ÛÖ$É/ï#¨A’–$Qöë-ð#–Ô|µ2bpˬ €†¯E¯Î"U _Ý{§Ùéõœ(?aæåÖAΚŽò›ÿ+“r61—-¢Ã"÷Ç<³+§üÜ“F0´"ç$U%tÌPÉ-uºNØŠìNU2«êQÞÒ©Y…ãæC‘p:„zŠU> ¦ƒFn„Ë—]ÄQßù ·i½ˆm–pážûƒ[ˆúcñ]ž¸¢Õ§˜ ¦¶¾V›=ʾ-6E’²eÜ6áÛ§ Ü©_c2n09M}ÑxÉF)ì·Æî*PHh@v9.èÂîR"¾Þ> u¿É'¿“HÉw· ~YÅoQ³œÍ«‘†ƒZ¸¬¿'Çøž¯Z£×ÄÚÑÌl×¼®. „¤æuh½Ö{’~CS8Å/ún‘‰K±9ÞG?ROBGµWØ6ùÿŒ ³9¸Eœ×ü"¾3PË°]׸ n æÛRJ»Î!¨ÈÙ[̼ÖÒ¬ë«Ö½É~È&UU§×·†Þ…°CéôÕw¶Ù¬ƒÄEÇc•gÁ{äÕ;Í⃉DTT7웪ê€yЈ@™ý”s²¯W,íÐ}…tóôE€ÈÛЦ’^ÉÊìBQ3/líqôlëoˆ²ø{²™Èzh#íÇì5E Ý®FÁE.2¤FÑ«!$:‰íÿR§@Ä.6&x‘´JTôÌÚ „4ýdÜk,²Št²òìmê`‘*@]–í!˜Q(_×”ÓY¿ ü y¢üˆláãáÑy…fì\ðl&F_db"þZㆇu\Ít>(õ=Êu÷aÂJÍû9UU•BÑtŒ­ÐÏVšö¼ÃÁHoIºÜS0ÐHQé“( .n8+Î^&²T6ÔÔ|‡”¼?"=DZÁõ«Ž6Ú¬ntús_ÏYÿ>e£Üð¤#¨r³2;Za²{³+jêT2¦ ·Ö†.êØ|ØOÃŽE뢆½±IUÏ.ªƒe)ýǃ¡˜*.£¯?O‹ b»ºr\BÌÝï ã`Û»R!i§ÕÔ8Îl5=mfU¡eº:7¥¸m¯× Ï“ ±.¾ÞIyL‰¬‡ñ›··Uí–PÝípxcxåÀ”Ë-æÊ•ÀXgÝ°jx —ÞvÃî¯ò³1šì„´ýIYÍ­—zòƒ.÷²©©p—¿TvMâØ´õûóI¬~5ãÈÀæ ç>ê¯lØÆ•xº%×ÉJðU$Ý™ýŒÒ (äôN`¤JM5Ùþ„Ãw¯¨¢1‘¨‘žÕÃC½¿š1¡ô³áïÏ—Î#|àA R ™Í“ý%d©â؇j4)Þ ¨‘—(lqºDî°O)õif´A.Žrã(™†µDPã¤}%[0‘ÖÙ¦8%sÄ¡­ÖaH \EëU"Ýeö T‡æ[Ì.¾¡ÙòÛƒ ”¾çpÓäAK€Á‡Í@bµ¿8<>GÙ¾><å‹#»è?¯e±„¥†@¿Ã“%Áú=þ¾:ïçwüÉœžÄ3rr¾öž~íý®²Hˆ€âHÒÕ8T•·¦p@ŸÉ~5f­åL©×ôNì>‹_߸"Õ³àgé=™ˆ†ÃÈœZ¯uͼòxŽ®8x6«édè4Ÿ $æû©#^­Q~ÅÓQ#lÚz.Éÿ¢Ãú®ÃŒrÖûžÙ©¯hU¤™Œþ>¥ÞÒ}[õ°"Ä÷¬îù‚S€ø`A‡lsxQ¸ÊJüåc.živ“žºÊ²¦e*ñ S_ì7CNÔ \8ê•<âÁô=¢_‘DÔPÚ%·?Òº_JßlÚ>‰#dörŽêR\@/)5Öî†ÓÓRúѤLÛxÜB7:Õv Q„½ÐÈ;n-ë·Ö‰ïæ`¹8Dæ×UÖïªc¹Æ&!‹zÎÉ#Ñ=ãÍÌþÔi$õKÆˉU¿Ä´Ü4ÜSÿ:)9r óÄëÉÄ1ÞÚÁ ì–7ó!¹ß;/ÁV±Ô¯ãº6öHhœŠf{Qîåå!à MmQ:™þ.´°«^ÏOµ@9\ŸJ6º³Þ/jŠù„×ëÚI$Î!Ö2%D;†ÐÇbªn‹NÉßk–öNB.ìþ0æögµàÌõ™Ü÷ÓzpxyÀ‰Ë á¦òÓÉ0(x¯¿è´¬šä±.ÂE$'Ãßý–·¾)4$žÊ\¤à 13 Ê'²Â»øÌ•[Zg\Y…ë,õ-m)´Î É†ˆTf“¡È嘢žº¢0T7,rýl? ŒhÝiRŒÊãC±gÝL1#'tñx&×€Óa—t¢ù¤¢eœüh ΜÁ=¼â(L_ˆ«/U} C\ëB%  ˜DŠ12Að¾³%%bðìaYñÉÞ¼3m5ªÊþíC(* c?ø‘„¶’-”dË(‡\²›Ð]¦»=óŠž‘Ï•ãü_®µÔ±0äövà+Ó#óòÜ ¡éU5jŽÀÑ‚Œ´–L¾¾/ÇÎ4wf^ZO´9|^ä3ydÙ‘ û[†NkCpô”©Âï3k.ǧF!owMNÞ¨ #¾ÔsNMÚ&zqcúþ_1M„ìkåþ¬¶‰;§æûƒ(ŠšæHŒžßvÙµÁñ¨YÌÞ{T…]ÛÍú™eðÛ@µd9p†jÆÍQ¤O9Ô-™Q’‘(7Õ²L*îæõ%JUmdì'5Iù©ä1$>ÎëÃûuX{Ð7u5PV@8Ï Š?*ƒµç»†…axVEòxC{¾»6Ž¬³âuLGß¹"ÖðÅG¨Ç¥ÙQDY£ØôâMàª,>P;ƒ2ý…_÷p<“kn}ÌG‘š»aâ ¶,»¸o}&5´&-—ëËä[·9‘œ¹“ûE!WÐÀ¹Ôh{?>ø§ƒ¯kž^Ö…fÈ%©(ìÆúYúVí_\6q÷‡y=nò‚ð†Ot•uÛ¥²çªp߉…ªûZáêÚàJH¿¯|ípç–Â& ù‡Èò+_¬k«ƒºZZ¿ÎL£Gu=ÒçhüK¥¼ÎKl}[Åe¾!ê#q<‰9Œï—•žGD-È)<MõxK§«E½ÀÄý(B=¬Vv¸UC…+r^ímæP™IU!0"Ù¶ëúçù›]tѧŠâÙº½Ú²Ä .ˆĹ ɉ4-ï3~ )x­kŽUƒñà —+>B ýÐH§"_ów•yTËïj³Bl2W–É”hßoÉP{¼O ¨_IXÒý®›™Ù.xØŽÄÌMA/ïG®Od ê…³é%Ê„™Ú=zªJûÛ–ƒŸfÇtÊ«õºù˜hÅ}¿*h­ó ™ˆÄŽT“˜vuÃh;ÐmòIꀪ=ry#ÅBI1[IKÑ"Oï#„æõw~}°{í`ë3!¤…pˆgÝQ™¹œI‘vÚ{…QB†qô1{1î‰ ]úÊ–8ŠÔ—¦½£JQfÌ܀ î$¿²Íïib`ê”qᶴ 1H6ÝÏ8¦«.ÃQ¢"ªù¹™ã¬… ³ßƒMLYƘ@EÄÄ=ˆå"¢À;êGû*šdr¥¯ NäÎBKc˜ý•u¼6'gF/,AÙh¾×ˆÅ´´Ø¯¯ŸŒ‰VžöxZ´÷ÞÃùj(•ÌfQbž£:äÈå¬E;ºìÑ•7;~)uïö­zºm4t{ud³Å¨(\Ù˃R«ÒñÒÿ‚¤œï˜©G LqÈÏ-#rûÁjòFãë²Rj;’¤°vC¬™Ï‹KŒC¡3å\ò±ä@»–ߢ>µÇ°ÛP¾ã+ 'ä V¬{t:|/ry˜*º\MøÄZ¬{‡ ~­×#?e!A3•«m&fãä{-†ãw«÷jR<ì¸K¯sl¶!W¤ž™“nqâp äi^‡«I/þl‚£'ß¼ÜÒíÑ'xÍî1¤³ÄɈ8¥r{•Ò°ýÔÌÿžYóà^éý„Q:–’ÅùìӸкϼ,½[Æ´ßÇÿ‚‡á-Ofó|o‹ÒO¨öš¦!UÒóäE)ÑÒQ17SOWµ=/êðäVveEo` µºÈüª¾áJÙdÇ%åïKFïäuf|d¢³5L>ƪ±âÏDºš®÷IKÝÐrsÈ_e¤ëŠ&<ôþ­t¶©CSZí'Û×(>ÍOƒ(¥ßt=ãµ)E–Bò3úµ¾ªŽ²[á»@ÿÕCmä¾äÀÁUh¢É »ûÚ}¡/.¾¡ßÜ´Z$’5!*™g;L^ÙFH¯¤˜ÃWŸú Ky´úò4\C‘ÖJE’Ç0s¦ð£éÚ/ÎXFЋãÈ…™›à;…Ú—\‡š\lätq ¿ÃÞ?°Üörq4^Ž3Ñ%{E«…´ÎºVt:ÝH©´¯êY÷+äsk”q•kOè2ÖÕ#©—õ*65“4F­çTÕI#“Ý2Ïi‘¡®Å«aFéO5è»K83]ï¹åp|bïÌqxÅà·˜Ím_clÒg%ÿ3¹e_?No™™›Sä¾ÒY~$l%Ÿ]Ï#ï/þ¤B#:“QXc1Å(o^Ÿ"Á¬ïÉ3ߥ©—ൾžš¯§Â~@ÛÇJ:wñÇÕVuúïã1L£2^UI‰iÖÉ·rÛË„’2hHï^uÑISŒpÓEUÀë"-Ë|–¦°øLê³{~ ÔZ) ~+|´–²ô ¥eba6²=§*îhïq~¦%cÈïl†UªF»Ê^À¡c²¬øÊã£C¦u€ô,>º=±¼ˆÙZµ4ˆnÜÁ³¥³B• bøM(ƒ~}¥¬AR~ƒàCmåÇk¬h¨å‘¡"„,^ml>ˆ3ïC‘f~»Lu­ .‘/©@7˜Ÿ„åv—|>šk;¨×ç{àYáôXµ4Aâù~f"Žýü»pŠ@HC=wÿþ˜'Cê±Æ°{îöWd[¦Dèý(3ëS¹uyÆA‚.ô/´‘Ù¢Ö¤–!,ãôæ^KX;§Þ“Kõ¬|ãôhóhDÍwîVéºäÖëïÊ3 [íX›Ðöé MŒO¨ ‡'wã\dK{„xÏpÑî4¯3íÆ?ûü Ã\Ø›[ÄmÔNƒ/!}QÈw°V£ÊUôð3H˜L\Ð8càYn1Ê~¥VTƒ¡ÂÄȪn†õÃÉUŒŽÃ{7tþ^°óótÃJy|[3Ÿv.c‡7ö|o ª€h*Å+ÌTƦvIȨï xÓNvÿ”©®ªó:[ úá¤jbg,N§ÆGº¬èêQ'å´p5é·úïÚÍQ'ÓÁÆkäû²‘= \X« šœhª0!|µ§5ÃWÆÚY©‡¨ @´Í[ÈÛB#ÃsœC-ŒàÊâÒ9*J³N`¬’ÐÒiJ~ð|Îì![‹(šK{ׄ ÏLfÖÿû2öðC·ÔýZ­)N‡#"r.Ã’î n/+ß¿åẄb}¨ujhñ,Åd§ˆ¨&Ô³ýM\Ñ»fw=Ñœ²xûKéõ†}ÉÝRÌ£dV¶lQ5Æ`Š›^Iÿd¬ü¸,b'Ãw?‡Rûç,¡—2ÃSº;qj»«-:œ(´5µéfœò•àC¤±Y–ÃäbUá¬Þ†Ê<02üeØ‹ÂkNí»ôŸXÓ*Åɵ|JçC`?펩OtÕ:g~I›-7Ùž?ïÊá+åÄ‚½1X6í‡ÏcáGŒI×N§ÝjÀ±#'%Æ[ý×¹‘-¿¾¤q~¦:‰¥þt õ~W¼Y˘{Ü7Ó]<èDc &?! ¢aÐ’ÕT– ®ìßUnà–ÚD-Ù Ù»3&òœ˜Ù¨Ø'!hòË뤽”¬Ö5r»¥`Ú¤UYb†ü$> ÷ëÇƆ;±Ò4ö{à¢?l mÕõ>r:¦%ò}& Å©âh5ûDkf{¼ãÐ:m o² øSÛO»ò&¸á°s°Z?gûûà†ìÀ<2Kåpö]\¢Ã~O×^‚ß}Œ4ÍIÓ•”À fø0=/EÚlBr»°4^Ï;×6µ‰EL<¯¤÷á·¸&ІpÛ’ýÌÀÛhÁ-ö Π´™À§Ÿ³q¯n#úÝ¥ä±e¡tÊcŠ£áÇÉ—²dÌpséyü„­wrÚ›÷èÚs:50û{iQÖx— K†óŠ“%ú¸)**§çªR…DJ~†´Ä…=«ý~Q‡ Š¨v¨1Ó;QïCs ÌW³xG£)±ÚývÈJŠ›¼Ý#\j…{[e‹$§í3MÆ@ÞŽà`1ûá;±]pø UÃ+¿ñÊÎbr…Ú—òÒƳfÆh?ØPê9¤=”Ü‹’¼òu¸Ê–1z-Ãõ€øu÷ÒNûÎÕÿšŒWуý·ˆ->—Úee€zظë·xõB«˜×SœŸ\–Ö}а«£LÜØïäšãVä.ˆó‚ó<*`ÑvUçpô¬å¬+Ùßz¨L7%¿`- J¶0¡äª&ò»ÖÿV×”‚Õ•âÜN(Hñã*ìÂï `²Z:go/rGȧөr¢kžð,ßV&¿ÀÉf±EÉ  ¿gl<ÊIì7<’B;‡Ñ—‘˜ZSÊöŸ»‚ ż€g+EÍfå°OHýgÖœ›ï×É6nõb*d¤„<ËÔ\hƒ¾Ý”Ü¿¾tuz1 {°í Pͯºµ½úÀñî ÆŒ3 nSâa~_¾.uÝ"í)Œ­D; ’û%7=˜–\TzUO^Ÿ‰¥T 8Ù’ ‡ÛÜ¡wQ6†);^V #¸›ýiJu3¡Â’¥[ŒGM¼ŸÅÑ•«-³±±cƹ QÕåò«€É¤i¹}kÍqÖW·p—ƒ:Œ¯vm9—L(?µ(#Íà•™{5Mœò7v–’Áe±á’æó²aì[ù6#E#¸ÙÏÊÂîñ›¿†^Zÿ*NlÀ9N;HµŸ¿HÑ—PÐM…áhkÃÖh2Ïm]~´‘y5Ÿ6К@bÔ½¯¨óY_‚T‡^È#4 7!+·ˆÉÑ;]ßcõdýéÃYî;è_¯µ4q„Û¥«°&ű4qÂΞÖ~G’áA2æ ýœýôRG_ü;¿­¬‚x|3è( (”~ùâÞÚG:×cIM;Ó2ÀüÚ£TªSß­«ø‹b /ô ¾§ @D€ þì}uÕG±¦®³ŽEw–¢~cÀ0±ª²–ëûÿjº»òq”ú¸¬Öx÷£Pk‚1Òó\‰Ù\Üîšàom.KUX5]l[×éAÆâq£¶´ãg'dOy¾=ÙK_©ŸË)ÚQ£No¶‚‡èÒš!÷bì™ ždRi,]""Ë2œÙë8Ù†Ûdö|ì j¾I™ ËîªVÏ7ÛíþÜʘ¡Žª¼”sšÑsÑÒW…ö‘ÕÉö^MQä#(¶ž{Œ‚àäëåŠuíOž¤+öã¢3åÚÀ•ÒN3ÆrxæÒ@ñ9 °+2€¿ŠØ*@'Èãˆ4ß‘ aÈJn* ¬ÛÎÓ¨÷<Èì"$€˜ØQ›º*?×€ó›KØUeKÄî¥`·œE¦†ºi¨+¾Öõ,¡Ì¿3èëŒgÄ94c<廉“R,q#šñ\žÔ¼Q¾ÁqN½¢°¡¥“ Ëiñ©RDº#sê"+^ v©Èô²I5½}dd™3d-À°ãÏÝ:á_Û·6„ôpÌž ê Ö/sX$†Ù¶š*@7®lFá°ÙmàÚ“¯DqkÀÂÏ8ȃ‹þG±€Ê§ß–fKó–üÚ톔ÿVzMű¿L³kƨÁÎɯÛmRnZΕÁËYªšc $߆G DáÀð¥ ÄÓÀ¯† ÕO•ŠnrbP…Sè½í­ì9Ý "ý¡uë €¸ô‚¸^íÀ¾hÇ+òÁã±Hv—e»¢èÎÿm.˜¾ÛÆøZ„¥œv¯H¤œå抇vžø‚YU,Z^¨µ”ƒ¼¡äÊ^*ßÔ_ò|¶-bÂ]ˆ “‹XO@˼ ò¥þ’ä ƒù•”]€H|ô½žYܶÔ<á,ʪùòL¹CSÊΑ (ˆ+dHT³‘âß.x?Ú¢Ó? ìÜ•ÿº)ê&D˜’7Ó9–ov®7Ð.R¸ÑàáEˆ.5ŸÔû‡…°†¦†r>Rr:Q€Y "!³ƒ·{¥ü³?A¬Að²‘ÑYHKtÔz‚;E_¤ ¤âüx_¼>¶Ä¤§ÑªAƒyж•GÚÖÄ2m¨A{qºô¢ŸŠl|&#ÍÁiÛåJˆg*qº"ÃæÎ0x²×ú·ë+~÷RÛcrÇl±d4ÞvŒÅ‹@ÁSŽâ †rѯ¤‘"RHþH$T !9¡UFè4]t’Ç ô¥våy1û}”x̲ɦ*c’)¬@ÏuW$¢ ¦ºå>p¼h¹–щ̫׆xWÞ8‰’Â5íu¬Qÿñ%²Yž”7éǃGrfƒß&\õ&bÌú ìvtM¤Oà+|©oÃå÷«dÖÎ.~ÇŠ-…¿W3Æ ŠµÊ\mÕ `]?whQÉ''’DÚ¤Ž R ñ[([VW³¡ääL¾•=ût y«‹*ñíô\³ã å‰BZÊÙác“(3(:jžoÄΡ=ü¾~^â9ËÅùã¼ÁP¿‰ñÍêj¿´f¸’ÉVÜ\=ߢCõÁú¼Â«‚Ñ{‹kµBÇÒ?—­©Pm "— ]Þ§§[%ûÝ#Çø5ñàÒÃAeS¡¥nEæe.NXÐEÇo¡~fýœ]ˆzÿy1Õve“÷R€&B æ,“v‚”U ¹Œ/í0ÞªåM*ÆJMr±­®T¯çZ¦1ÉìMY˜Uë‡í×Á'ý9b¶¯™‰°XY@}…‰ß-^6”·˜œ;mpÔß7?ƒÞzß·9HE¼ShØÒ}Xkû8ßÒÁÏ^˜.ä…f„p•dŽ;™ÅËD~+u5 uE ƨñ_eq$Õ÷˜A«e& "S  ¶f‰-C¥Šz‘Í‹Àœ1y’ŸÔdÖ—p?"ÉÄZ£œÄZ!‡…¡Uà$âÚ:ø¤U×½×—Ã*¢yg·8­ íVÞ­²ávîÄOû‘éç˱À»1C¨–߶6Ña"½O×u`Z]ßÁ°bj.*¡ iM3M ‘¦$Äz\Á2¯Ù¡ã%†No(‹e=Ö!’Ôý0W‚U.†c²Q®j+a rêÊHŽó$9¯4|°R®´_#+¡lò ]$­bT%|pÖ¹ {¦P¿¬BR[¶éf2°IP£>p•Éc8*QeQÚØDHß´(Á0dNὕ™DŸb\ Íô¤ÔýŠö¶Á²u‹|±¹Õ‡viŽG¨p€!ÍÃC|d’J¾±FïÒi˜ Äí:™„í²¿ê$݃À°s· ªiÛaÐrž–d•"f9sÚ•¾ÿ4°någï$um1ÝÂݵ¡áݼ8Àâ35±Âé¤É¼Á#þÒ° Eݙ֎¢ßä0#ðgÄ;† ‚ÉØÛ° ik9” •˜êû$*GÊû­?ˆÄP<]œ^yƒÛN^õ`ÀÏX™ÒQb>dþ›2‘…ì9e¢ iU±P½]VñʧO<…oE@OÍŠŸaù /™½š?¿{SSÐTß*î"*õãaÔ…\ˆGq§V¯xél¯oÐÀ_ †sÙ—ÂÝqÓŒ–G¸‰±6½Zæ“^Š^I‘Æ‘ú·ˆïß?ü)ðã}þDÑÎ@xŠ|œ×r òeç$*‹!¢ËDæ–ßpñ£a&}×é³8Ôæ3>¤AíaSW7k¿^ª°ëµ5Œâw|$£ª¿õ^ú(Í2lbÿ½ÏÓwZIwZ”^ʺ9º”¸ë•ÇªT«¿éªÔ7±;¦ÝC‚! “ö[D /|õ4™lÓÕþô“îÐDð8Kwƒ áÏè:6߬=Fú;¦ÎSJpäLä5b¡c´›TÝ2ª3ß,AËúsí Þ2MÖ„»¥·o÷9œ«éâ+ØÚHÆž!E]ð–wK)èf¤c]´?;t¬ö//‰·+6ßø¸k.æô-pµ.fI.ï`õkWÆì’w®¨°cöǃ†×Êkç_•é@e“[Æœ]¸¸ÈЦ¦®Ñ›Bª[9çZ«•WÛ!ܽ?YjRñ‹OÄïKR/Ãv !ÆJeïGin¾=òÝü'styŽé®UØ3‚úˆlü“â¼ì,c‡¬ìÓXJ®±ís )~eשߴ6ä`®ié hÊ !r¸“£{ŠêMZWYjE^Z‡C:mi{ÜúšœnðÁËC˜ñpIx_Æ…¶O1Hý½{!`=A„ˆõ¾M’C·ªã­pl/ë§[„?„»`óŠõ q¸Ž}Pwï1ºŽé+ø o¥ŠWwÐÚê©AŠzf«DlKJ 3Ý"£0Ùm8Ú¹,c«h€„cE@ ³Õg;aj[°‘\40'GXëLÊö([3ÍÙ²Ö¿g< qÏ>ÑE—ãE•$q,‡Š–žX,–Óò5ºõÁ®ãÈ:ÄæÂ:·âöÎYýŠ M W2OÃ/Ý5håÑcPefZ./žx>t(6#Œõãg±7:°î’ŸÆÁT$òÉUŠµ9Š¶?™!xZdÿ–ÆH|ižÿÁV(\Uà•tP/éÔCwZM,s®,z¿&˜n ÿãc“¨€þ¿Jµ¿êÇÚFgI*ŒÑ Zéõ”õ3 –ò„£Š-Ï8)Ã…_r壣XZjx?ÃL‘¨ªVbsvÌ©[ƒÚ3Šr9»M‡R“Ç;Š(Â’/‘ÕÍô€Öü`?=ýàééÝa´ýVù7Má-…ßy·h8hök®‚¶ú„ÒžÉ$Kn ×ïÙ®#=†]rÒ& ¾ˆ?|„\¤»y¯ ¨L6ðvjÇ^̳¶ôŽ9[*Ç”C™ žs˜žzB~ÌmÁ“YÏ[nUGî9Ð:F{WáoN5wÏ ñŸH«ì™ÌhÑ=GÜa²á©[`c>½­U9.ªüKñ½¼Ôvq¸È+Œ.*~Htë¿ &C: Ý›ÇÑt‚¼3ïÐm #ù¸‡Ûj´JeÜDæfÈi-:åÐ:™›sÔ)3}Sjˆ°À -´õ±ixó¬§f‰KôÚúcãì 2À—ÒéžÚûlÒèVåíN¤øƒ;Ø͆R€þ'Jk¨.NTñ2F ´x)TÇf÷á7h0׫ !@}Ùø:1\Ð)ĨMç_§*ü®çZkzß1NèÓ†[Ñ­÷Ý. t|ó(fªß ±Ãò2d©(óà˜ní­vTÇyLHË_$îLÁ˜ÂǪ$‚®\KÊTk|4¸VœÄ\vù@€øŠY)IPèh …¼¥õbj¢‡õ ~hUÇX£ôõKäˆ)+L¾³™êm’Ǹ"Bc÷9dÐ3êL®ö&¦OzcwÏWÊIv½Y!Ô8Q×6#Ï.'Y£Ã)¥P3X'‹e³¥÷0öKúEØ\9ZzŠd€Vëçã>yóË·é|êo(^­µ÷Y·çŠßDs³c `Uñ õóËn«£Ë 2Â)ÓÃ0ÎÓSd€ ) 2ב‡•'49?úi¯¦Ž:Üóïî“ùp-®4f7#±Wôæ ®ðïšuÈý9þ$ª¶K€ aÍGÝÁ¦“ ±_}z‡÷=ÀÃ8cƉ *÷ú!x¹FÁ$¶Å:½P0óPìV§ ‰éÇ›¿9@<ôê5‡K?nU30Kš;j)åˆü "bÏ—"ÑôG'9¦sœ”…6ÑnSl ^Fäíïó˜QS²3۰˘Nµµ1cjÁR}J¨š¸?ÉXKF2Å’žÌK⥋9æèooxvéžµfÅø$4~ÝŽüåërb{_Ë¡˜b€ÌôϪñÁ©åò6mÅ è³âÊòU> ³Þ˜Y³f*#Kxû·232•ƒÑK°¡­Rm$)àz©?? [Óµòƒ”®’káƒ9SåúêöéQë ˆüI¥«HK%W£ìÏ.V#>9ÕM¹wHŽ#°Á{À{-¹§Õo<7Êÿ,âG–wŽRl¾!žÜ¾p§ûž“ñ2@>$"‘qŒInÃÇ5É[ñˆPc5Š Êè†zN…šÞ^0Ã)Ž ¨ó4 &êEÇeÀçm˜‚žôªùƒîœºx½ŒRq>Ì\;sfA…œÙ-‚ËJ,W,¸¾ Ec/ÝÜù|%_)[<á_ÛâH‚%zRÒÛѤcCZÞŒzß_›>Í jÙjk>­ìÌJ¨€Ù\‡\oj ¨’”!&â3ú n¶¹h#N¬×>Ê·ï U!NÈzjš Δ÷ÎÜ¥ ˆ0D<z­ÛÑ—Æ$éÿ¼¡®uŽÎ‚)XºÈ<á·3/Ò¿žâŸ¢ŽsÙž²ŸÁ:3ª5ÍÉ?‘w±aM_lM¼x-‚¡$'¿-s†+Ç”w9ÿIšð§ÐÝ©ƒŒ¼:Ê{×ñ™O.þ#—‡tFVǼ¹ñ¯MÉÙ†`Höë¸.äÔqª#]AÁÁv ¨Ý ô3,)²X÷|(]'o2ÐpMvØ0°ö½ãØ£á¦Îi‚~µc~Ö¡ªý>pfIXÁ*€Ö;—*ÇíŸn£ß6ßa.ŹUÈ^®´nçý¯ÃØܔ›'ts`½ÙM@×Äê öJèš_B§¦ëúÉö©ìõ%Ð<Ý´·èJèƒRv×-+²ÆGtcS&ÐNî}Ÿ&/§Þ‹šÎÖ=–œÂ<Ý<ÿïiWE0…ÝYPé‹ox„këáãǬš?¥ß²Ÿ„®Ç â®þÀ&Å_Ào^òc ÝP·Ô'Xµkx¨¿ZÞõ ‹¹J\§ÝYãKø¸Â:Y‹Jmù×t—’ˆSîßÀF¬CÖKªlÇ”4Ûwën̾ù¼ ¾S—m„VÚÙ÷§ñ™çZoаâå¥.JþŽ‹5 ÿUümú¾TŠÛw3Ö&C)„d,|]/;/îL×ÕdùÄÊžÀ—RÑx¸¦¡‘²Aéoã*Æ%’~ªï~iF åVôm¯³³/û\²>øÓ¾Œ$©üpQè,vêã­!šìÜý]¦~Ôܧº¸v=Ýø{ùEôëXBhm¬ÛtMþÛ¢ÊÈ’ÙÔŒ;"ÁÁ9:w~YÙ²‘Á‹ž3™ÔZÖÅ 8¸@Í÷ïNªˆ²(®¶³¸&ßmͺ¶°P6 l´¹œœU¾Xqy€e'HÃ:ìÐÜbòè—/–îQ´ "‰€5Üøl,+ý¹†ÓÑ­®Ähó&óýh#Óf®×¡–ØUp¢ˆ¤Ý”Û2ô²!&R·Q}ùhŠw;lÚú¬ÒíÍrƉôX‘¤Oý|\ÚèEè™»¨œÌáÆÖ¯( §†èJK“ä7(Hõw°¶Yf­›]2Ãݶê5rü¥)20é‹[”Ä,íИ§Þ Ѧ c(©M>Á¹ð㽶) ìxÎa¬ô›/½=5PðêóªÚb뮥;~·R%¶¶=E«¯(ÍÁÄšIÝÝ ~Ö_ËõĆ #Þžr¹ÁP(?®nËÓÇ…³4áìlâà!ÏU-Db÷±œó‰*oíyéÛV>9먥ßïµ1ÅX8ö÷ †J&T¬¶ëëÐW‰ìNœ÷ä%Üë­)°Ö­ëGÁÿ(©6Åé’w»Î¿¹®Àì"â:ƒ&.›B öC^õµÀ{·Ye2Éþä@4&%Ï2·Ž.ˆÀ'>"£½ >*üÖG!¾ÊÍP¢¥•(ûŠ€~8­=‰ý• ½Å¬ÎÒì+CJyIIJ9¤‚äô©‰{lÓ¾âi‡2vcˆÌžb·ÝGyëà*öR@æ-YJè(WÉ\£—9ã•3SLf î&ø)ß÷'†øaýh/‹SG«º3×L˜Ãß^')Hµ¹<#–°uÔKžy$Œ‚¸G,Ãð²K«÷ÖÉi[jÉÔÉ#½^‹Ƶ4æ9Œ}À¨MÏÒ¡Õ%FäÕñ®r:à]h›n·ÉW~k¸–wYkÄ®uþ`ÖJîÜ#+õ4ƸÑü•õýSÌHCG¢ØtÑíÓ´1H<“{9‹iÜoßN8^Wæ±[WéOhT«,ôJàk5à›|}è…gò­EI±h"˜oÆpv (ð<™Tù7SÀú £Þâñ€.ßR§ÌIÖ™ ”gé ƒ|™ôÊ ¨¶)j¾äÝËѲ*V©¤,Évû53‰Mãÿ ªUí³´Á—‹¡4s©Ñ¹¯21«IÞ°®„˜Jó"t¬ d…vc4ë•ìôx|óÂÜQRY<™|ùYàiñ’¢ˆÀs4íˆqÿ ‹=«gÍO‘pªuNuGÛ¥3nnïQï¯^Pr‡oZë·WåÑÇN®YEnÁîNù»|ÕÏ*1w½º# Æé:w~yòj@³†©æ†äÛäÑ\û?» 1¯ÐT0 tËÖ*ùŽ¤¯ˆ=ɶ³ÈË}ÉËúh è¹C£­Ýe; w±Ó}V,ë >Åñ±êt=ê½67D @Ì{¬Ô:@ È>[ÍF*Èâ°ñöiáÙ³ýø8VVú_ñ§u ¿+ót*ëüÐÉï0¬à¹ì•Ì7ªÒ¬íåÛÔÉvát’‹bQÔÛóžÃ_éûÈ\¤÷j >…¹Ø… Ëšø ¶VöDÖÁrC¬I’åpV¤Jm…Ë*M¢)"vK55™ðcÜÇÐ è“*dYˆüÞøXyФ'Ò¤SjœL·)Ò€™p[éÏ¡à”+÷S•¡j³¹£4|ŒQF Øð— ò5îöZ=Lº¶ ÷g’ý­J9"nÄñZÕµÜÒ·0îÐìXþóy}:D¦`OäÞ .§Çx„Ä•g¾ŒÚEâçfÀÍ3!~&D+g…çÅ¡ï B’‡ÒÌÝa(-D¸la¹ß<§^ýd¡­Õ˜À¡“‹J#UšWé¼>5ßõUê¬3¼vN;ºŠ-XFänkX-ÊD ñ<³¥n²ÁÅ9óäÈ»âH$•Â탠 ,nò6Q¿9€ÑÚ›1ŠÅ±~ø~Tû°NT÷ñÜ/¸'A!ô°.`Î OSŒÈ‹ÙŠõŒhþëî[\@™bê—eœE%›Ih¢™P€>R’ÕËô—¶köŒcn"‡ x¯å6ãªFÅ£›ì"Ð\$ L4âÙ‡ÚÞ¡1Ýw#Ì´5N€Ñóäóƒ·XlßïÈÜH)𢸙qª+"íJÉ)Þ6„Óv&¶ÎÚ<¬‘âSÞT#Ü:~$s.K7"Yr#½›-¶ø[2™6ñk¦øÛ‘àbI±4h·rË&Vç{t̑ЬAÀÉž#í‹)œ—‘€$3n«1x™‚ÕfN¤]Ä¡x&ÙË¥yî*wºƒ­Wû_PÓ3íIËçÄ×›7,ÛÝ`ÕÉhï&¬‹W^ܱbÇýú‘j“;¾¬ø ½¯žæݸlKÕdÝrݪµÛÞ»ýþn‡OpµèÀ{ [ÄËVöûšúø5ö2(íìï…r« ŠàûPC@ ˜>GÉ™x”Ýë³ ]ÝjK˶š+ÜŸþD¦ÈŒ),Õ•¿ t Jª@ûFc ɼ)õ—ß™+:9¼T©X½78{Ò9#óÎ\  8ÞsydëYµzśŒáÖ5ÑqCJ7Dp÷pRÓHvœ¡óØÎHHe£Ù[à¬_7öû+Žm#_¸h®Bˆ¤È’à³9S÷¥š[káí64 ÕEï(YtÌc„I¿Y}[<Ϲ׶@e¡$§¯ünàÏíòxu…ÛÞØá[Wó§¥&Ê a4#+@õúë|üK!G«€×}ÕDøèì£gX Ü9¿¶¹# w×e=$ E©§ÜLm†»÷ ..†“ÂÑ£æ…{ª{#…šXMµ<Òà0¨6mþãðõ?{ÑFÙëâþ‹l†Ô µî:ÎÐçRwñ¨9ÐýÆÓ“zôÇ ”ÒÕgº¨á†‚výIJ*¨Ž]¡Žy­ó{q§WlÒ èÔ »çcrÉ%2^Ò ¢ŸÇ{êy³ñl rIÒS#vAdñºÜeQ“(à¾.@N騀Üã²6×)uDY´¿ d%)öÐTæZ¬ÜóÚVŒ @A½ö‡ü([êP‹¾Ì©Ölz~úû%ç~M/àÒ§2H~=Ö¾[E† C.Š÷ æ¥Q&MŒ`eþU¯‰'¬.¶Ìhêk›Ã(iÞm² <‚WŒ¨Ô 7 誽™•ÍH&ú5ƒ´Ä‰Gv^“ê,°d8U#¢^{Â`’O¹aR3t¾:â¡8¯›£;t©¸&°,BßÔ³ ú¸êy·`]’;‚B{+µâhöC§ˆPT ‹ƺ:ϪV*O ï«)v–ç„•ÿ~³«§²:"“Ñ»ô/\9æ8±/[‹×ºy½Š-R؇Ç?,ı¢‚h7ÙãWE‹9·Qí Å ƒP½ùÏd!*Oæ—4ÕДµý»ƶÍÀB Ž»ñEÁhï`Ä6 I¢rµÂ4FÞF1gïÆåD¶ù³Ã^MS´«íEý/•ãëïÖqiyc7ø¢ò|n]™,ù÷ˆ•Ó óZ8UÌ°aïÅ¿›T¨É‚5$¸éÄtäθî÷ü…àþqÝ5ˆ£tšAÁâ hɉB+îìÂÄqÃ…÷Až@E9L,(|wQ3:܈'׆ôú ×WskµzØ_[λL“«}Ø Ê‹ÙqQØÐr¤Ú¶íÁ¼¿f“¤ nQžH.»¾Qôp„å ØØ$ùcºž'_j‹‡ˆ¿„n#\Uœ·>ó2ãb4,O9DÝNdW€V¤zEU¤z¾ ç˜ž× âÅ«³ Y_tÇ[ÁôHv¿Ÿüù”¨5RØbÍö"ôVqÓÁ¦å>Ýæ ¡é¬xZ¿9N†1«=áz|ð®˜:{å€ì[R Š¤‰èšá”9a|ç`Àˆ#kUÛÌ;mÚ¹ÎÌË\ö™++?Z]ð$8/3¦ª8{uOGø'Pc®¡÷ûeÍþoÇj¼9Ld©CkõÕª¶C©ã±¢'ÅÆÔ¡ÇTþNÊJ\'B'noÿÆEÀ„éô­`ô7JˆrÑià°~#Ïöµ”_žKªÔúŒ*£ÔÌuavñ eß‚º˜Jå숨n2Þ~¥.•´H±vyn¤ù9¸1$rô~¼lõ‡zëåÑñ7kíàjÙm§n$vÀ:MµP &¸œD2Q‘an^´«\…®eóœ6üÞÚ—qHØIø Çbâ* §Š¦¹kŠè„­\QeFco)Îày§Æ\ý×X4ËAFß\† ÜUÅÀf`èS1m–KÂ<(À‘›QL#rbübnyôm™hL* 0_Xäï?ÑfR{éoì •¹ôˆÏ…Ólþ¥ïIaTã5 M ]G‚­Ã4öž*fÿEO›ÛJ¬!•%n <ýtèÍ!Ù$ ¾RsXÎz­ñQ¸c4!EС ç:<Žðû¡‡á].ŸÈßÄÙ®™¡Íí[s…¥µ.‹þP,EÏlñÔ DлÛíšuGIé²ã°æKTcewá­9G÷»rýÜÞçݦ®Hn°”»º…W£ìdôãvÕ"Óàó®eZR‰þ"àß+¾RÖ¥þtj 4ûÅ:ØHýÔ ã©ÐMB\Œ%P§ž×å…®m ü,2Ôò#"§4®h 2îEÊ!“£ÝŽ0‡(2×ÂÌ4î|Ä– Á€"}:؃*˜¿TÌÙ¼©é*yBîÑhk?lÔ&:(rRŸØ”e¸5éŽ~qH)ç©‘$zlÂœ8ÔÛA²\9†©8´À­6ä¶eÊyˆt:§´ . JôúA¦oùìfÁ07â”3_»>÷Ö5 Ùš:ÎËœ¯sĉåvb|øú_òÇ©möãžòŸm„!ú+X=¸„ß ¾f¹©#ÇÄ`y÷B+$¿«ÏkÌ*Ì´Ü<½bq >Y|eZ£)íEPó±?Y¦ä×ñàY/b‹dc…Óœ¼„†«íQsÈø53Q²»Š%7DZõ¤â9ƒ‡¾ªŽå’ߦ*®µññ•ì4Á©¤Ccê ¥v¹“rðnt\Ù™:;B²DƒÔí &‰à³áèú$[÷ÖwüÛÕãt\Ì´æèÔ-ë%B‡²,D„Ð0Õž¹z±àÖ¨Ù\ ˆ¨&ÿÒ¯`9Æ ïšøð€wx‡ðö3[†½°@©…+÷ù\„XÎÎÂÃ.œÐ0#éÝr ùirº1¦ ÀúmX—qÄ1ÓTBà³·wo•ŠÑÔÃÜ„"î"‡pÿ6@À\·íÈI‘u_úägk«etÚÛ%®Œ07¾ÎžFµã¡OŽ)«oË:D>`£±ù¬œÈžó0·¿îÑŒWÕÇôñ}VáRݱßxïyáYÂI4’ ëkj0Ov[ß¼’šL[•SâÌ óñ“”ØÍÊsÁëªè™?¿/sBRÒ8ù¡æ²RÅA% $+²Kyâ 1µ:/Eᜑ%U}ª¶J“róL¹ïV·L¯T#] ÔÜ%0Ѧ=/Ò Ûä'‡l²óþ])jþx«†8™_ÞÅÎÑ"îË]ð§ûf·€ ~sÂÎêµÎ’f¥m7]ö¨ZEÜù q?Å`'ÁUµ omàØÀ¾<¢w<µƒ Ú윖¤_ˆ¾@[RBâã1!êÞq-æ-Ú‹Ph¯K’‰ûMུ3k?¬õ…$¬ŸËïUÄ–×Mâ’Iºm.ì-Na4ŸâÙF ]XçQ´á6¡Ô¡"gwšRe ]!CÕ¬>¨æ€:–„-}ç5>Ó>šÞðÏanRÑ%3漬—­àsŽ¨ƒ&ñr z½qÓsŠ´»úÚ[ H{"‡ætBa‹3ÝWØÒßž}€å¦äþVî¹÷ö5‹öšß¢ ¤I”‚Õ‘gõ0ÐQvA&ĸ‹¿D?Þæäò·ñëîU+Á´V•ß¼’a §2[`Ÿ%ù·bp£ )¬¸Æªªý-T`‰Á8iÁ´ùÒè<ê°&íí×èL„q·Åɇ"i”žDz’lÕYô ß3(vÝZÆìròaFÒÂj›H QA98-­³iåÆÍfþ&Hƒû„Ѧt®­ ˆ¦î9É/îRrRO Ôù;)›žöÒ ÷ƒ‚)¶ ½¸sÔ.¯à°J_§‰ÑÁ[P…^ùÚ_Àbë7¬9IÀTˆ’*o§üwÈi¾? ;‡¶‹¦“ýjv _ÍÃy¢ÓGƒ7;“8{T'·Õ{iW€éçòÊ»·]âß .ýWß®¤Ã· ¼ë#'@#Lâê£2·7ŠX²ûCœ¹ÐþÆ%E)¡„4üû…õcMúσðc ª„gÑg3I¿`ƒ8ïçBLuK+¥ÿíÔ‚rJïYLÞZ«O¾FÙ B\ œ+þ 8PjÝ÷Aï ²ƒË¶xåúÜ5*¶%O•t›5 Uq‚BÃŒÏI ïê ‚ê)¾Ôfµ9‰jŸÌŒ§ x9oÔ)bIÒ9‘QðTÕ©øY®BTˆqgæÓ‰EâËTÈc ¸}}3€>©BLé)á`Y)U0ª61!"S–fµ4ÔÛÞÖ Yé„;‹-h©vêLÕÏôro'²C*OMCj‡ÿS8Ú³Ü2žš©æÉ¥§Š[xt‚þŠ9¸ qè¸1@µCÜéA[õ+Ò”ýúrH JÑR¬}³ªÑ'Ú†ÛÇ\W?R¾Ö¨È­Ây¥^ËB](8r•ÎÚ¹ïÏA¢ÀÛN4:šÅ8èé@ ,4žrR÷Rqb΀Ë2Îíl`ÛcJ–î/Õôsǽ“GÿìÅüa`΋ª#k5!Xuª¦~±öRŒ¾ŸNuß¾©'ìDD@tË.ñà4úNí–˜$ÀJ–¦–&»ü´c‘I ¯|V®H•–”±ïÅÈos,‘R)’MÉ# ÆœÏy…kû6x5íÖ-~(ÍZÿ“ÜlD‚ ¦V#´~xr× î¼&÷ðû?4’·b\)•‚…ùJÕ*>…‘æuž†É‰cšMîø0˜¢‚ ÷Æ+WZìÒ/eeíѽßR–Ï%~i³!¸³ þaZ(årz ÂX®RƇfŒDe¶èŒCÄ6¼åéc q؇tm"Ðç«ô±„Ô'n=ïÏ[1·Ò%ÙœV‚®[ï!q͈—÷s3˜ñL+,Oµ9‚Ó®rëbG’© ñ¬wA Ùoo´Î¥nÇy¡ð Šþb/;ÍCt”DC}ãýö¼IÑ»`Î gÍa µ`¯·ÃÈó¦×"ŽØ±’g¨»AÙÅhEèa ÌwõœbÞá¨È?Gùøª†b®¨d2‚ ÅæâRt¹æ™éü¦Dñ<ûíâF]š‰ôùåÕ(‡Ø“š‰âo%¢Þ‰làø#¢~ÏT?ë€Lšr°ÇŸókè4Œ®wx“€J° endstream endobj 72 0 obj << /Length 868 /Filter /FlateDecode >> stream xÚ}UMkã0½ûWh…öF²ã¯²CÛ–¦,{Ml¥hìà$°ý÷«73nw—ÒƒÍÓèÍèͳ,]}{\OlÛoÝ$ºÕêÉúËиIù}s ®®ª¾¹\w¾w®uí8{ºSC߬ÝY]—«jÕíÏ7ž¼êš×KëFÖç¤Â½ì» ÖQ×ÏîçÄ5Ûßf6Ó“æ09]Žn˜œÍD#åy~õÔ/XÊO©Ï¦øá†Ó¾ïÕZûÀ²kËþ€ÆNÁTÄ©é(w·ïÚAª-ô&Tí¾9ˈÞÍÁ;„äõÛéì«n×ó¹š>ùÉÓyx#½7ÁôahÝ°ï^Ôõg=a}9_Ä(,ªu;_×ûr¿985ý¢ëwæóÛѩƆ56}ëNÇMã†M÷â‚¹Ö 5¯ëEàºö¿9qÊv7r—ž«kÿ u”/‚¹A² )`JbD>`´öØ2ãš™$`¤TY'`ä`ä9&£Ä*×ð8˜W`TœR±¤&4–`Ô(ZsJ¢5Rê’H©+¸†Çhì3Ÿ}7¿6ƒXäÅ¡°6„‹0×a™G„+˜ gıXFǯ€Æ8åÜ8ã8ñsέ-Çá´.x]â”/€+ö5MĹˆÖPN‰£é±¯<1\?‚Ãõ“Æåǘ¾tõ1¦:Ë¿ø#§þ7^ïÂYH{upÍý¤²àQF^o­Ç†1¼ÎBÆЖEŒQ?›1^“׆¼ÈƨŸ¥Œ‰“qÿЉ´.yafð%+CsV2†GYŘêSÿ&ÆžÈjÆЙ³þëæ¬??gý´‡rÖCOÎúÊeý zÏYJ|ÖŸŸu¦Ðœ³Îß8gå²ÎˆrY§¡‹}¶èÅŠÏèÑŠÏ1°øLñkYñûÒŠÏX׊ÏðÓŠÏÄŸÑ»Ÿ¡ßŠÏÐiÅgônÅgøfÅgª/>C¿Ÿ¡³Ÿ±n!>ƒ_ˆÏàâ3ôâ3åŠÏ轟‰/>?ã“°œEè¥ÿÑK!þcâ?ÕäÅRMþO ª#þƒSñZ1|¨8Gxp4Éaž«j9¡èDÂQKæý.h.Ã௠º‰èÜlj¿ïÜûeuìÈ¢‡n¹ñ¢Åè¡þðç( endstream endobj 73 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œ.RIý÷ë73´»«ªÐóøÍøÍÃØW߶3ÛÏnÝjõèNÃelܬü¾;WWÕÐ\®?ÿp®uí4{ºSãÐlÝY]—›jÓwçOÞôÍë¥uësRá^ºþƒ‚uÔõ“û5sÍx0ZëYs˜.G7ÎÎf¦‘òÔ_=õ –òSê³)E~ºñÔ ý2·~ÞÖ}[4v æ"NÍ'¹û®oGQ¨ž¡70¡j»æ,#z7ï’·o§³;lúý,—jþè'OçñôÞóû±uc׿¨ëÏzÂör<¾:ˆQ:X­Tëö¾®÷åÇîàÔü‹®ß™OoG§BÖØ ­;wwý‹ –Z¯Ô²®WëÛÿæLÄ)Ïû‰»ö\]ûW¨£|, ’MHS"#p òÐ# µÇ>y×È| #¥Ê:##Ï0)%V©¸†ÇÁ²£â” Œ55¡)°£FÑšSj­‘R—@J]!À5£w+>Ã7+>S}ñú­ø …øŒu ñüB|¿Ÿ¡§Ÿ)W|Fï…øL|ñ™øŸ „å,B/…ø^ ñû°ÿ©&ÿ+–jòRPñœŠ×ŠáCÅñ8ƒ£Iæó\UË E'Žj\2ïwAsGMÐMDç>Nü®wï—Õq8"‹ºå¦‹£û:ø‰9ç endstream endobj 74 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œ.RIý÷ë73´»«ªÐóøÍøÍÃØW߶3ÛÏnÝjõèNÃelܬü¾;WWÕÐ\®?ÿp®uí4{ºSãÐlÝY]—›jÓwçOÞôÍë¥uësRá^ºþƒ‚uÔõ“û5sÍx0¡Ö³æ0;]ŽnœÍL#å©;¿zê,å§ÔgSŠ ütã©ú;enµÖ>°îÛr8 ±S0qj>ÉÝw};ŠBõ ½ UÛ5gÑ»9x‡¼};ÝaÓï‡`¹TóG?y:o¤÷&˜ß­»þE]&ж—ãñÕAŒÒÁj¥Z·÷u½/?v§æ_týÎ|z;:ÒØ°ÆfhÝé¸kܸë_\°Ôz¥–u½ \ßþ7g"NyÞOܵçêÚ¿Bå«`ilB ˜„‘„­=öÌã¸æ@æ )UÖ 9yŽ€IÁ(±JÅ5<–§T`¬© M55ŠÖœR£h”ºäRê ®á1ÚŸúÌSßÍïÝ(yq(¬ ábŒÆuX&Àá &èq,–Ñ1Ç+à„±N97Î8Nüœsk`Ëq8­ ^—8%Ç àŠ½FMq.â†5„Sâhú@ìkO × Ápý$Áƒqù1¦/]}Œ©Îú/þÄ©ÿ»pÒ^`ÜD3F?©ìx”‘×Æ[ë±a ¯³1´ecÔÏŒ×Àäµ!/²„1êg)câdÜ?4dâ­K^˜|É ÆÐœ•ŒáQV1¦úÔ¿‰±'²š1tæ¬?ƺ9ëÁÏY?í¡œõÇГ³þ„rY‚ÞsÖŸŸõ'Äg)4ç¬3Å7ÎYgD¹¬3¢\ÖièÇbŸ-z±â3z´âs ,>G|ÆZV|ƾ´â3Öµâ3ü´â3qÄgônÅgè·â3tZñ½[ñ¾Yñ™ê‹ÏÐoÅgè,Äg¬[ˆÏàâ3ø…ø =…øL¹â3z/Ägâ‹ÏÄÏød ,gz)ÄôRˆÿ؇…øO5ù_±T“ÿ“‚êˆÿàT¼V *ŽÇM2G˜çªZN(:‘pTã’y¿ šË8úk‚n":÷qâw½{¿¬ŽÃYôÐ-7]´Ý×ÁÎ%ç# endstream endobj 75 0 obj << /Length 867 /Filter /FlateDecode >> stream xÚ}UÁn£0½óÞC¥öƆ¦Š"R»­šjµ×œn¤"’ú÷ë73iwWU çñ›ñ›‡±¯¾=¬'®žý$¹ÕêчóØúIõ}sˆ®®ê¡=ï}úá}ç»ËìñN=ŒC»ö'u]­êU¿;Ýòªo_Ï¿°>'•þe×P°Žº~ò¿&¾÷&í¤ÝOŽçƒ''3ÑHyÚ^õ – Sê³)E~úñ¸ú;enµÖ!°ì»jØ£±c4qjz‘»ÝõÝ( Õ3ôF&VÝ®=ɈÞí>8„äõÛñä÷«~;Dó¹š>†Éãi|#½7Ñô~ìü¸ë_Ôõga}>^=Ä(-ªóÛP7øòc³÷júE×ï̧·ƒW1 kl‡Î›Ö›þÅGs­jÞ4‹È÷Ýs&á”çí…» \Ý„W¬“bÍ ’MLS!"q Ø# uÀ!`NØÈÀÈ©²ÎÀ(À( LF…Uj®p4¯Á¨9¥cIMh ,ÁhP´á”E¤4ÒÔp€Ñþ¥Ïbvé»ý½Å¢ …µ!\΀ѸŽ« 8!\Ã=#ŽÃ2:åx œ1vÀ9çæÀ–ãÄ/8·v‡Óºäu‰Sq¼®ÙkÔ4 ç"nXCI8'Ž¦ľ&ðÄpý ×Ï2<WcúÒõǘê,ÿâ_8Í¿1ðbxÏbÚ ¬Ã€›iÆè'—Ý,ym‚µÆðÚÆŒ¡Í&ŒQßÎ/ÉkC^ØŒ1êÛœ1q,÷ V¼ uÉ 3ƒ/¶d Ͷb l͘êSÿ&Åž° cè,XŠu ÖŸ‚_°~ÚCëO¡§`ýå²þ ½¬?'>ëψÏ:sh.XgŽo\°Î„rYgB¹¬ÓÐÅ>;ôâÄgôèÄçX|&ŽøŒµœøŒ}éÄg¬ëÄgøéÄgâˆÏè݉ÏÐïÄgètâ3zwâ3|sâ3ÕŸ¡ß‰ÏÐYŠÏX·ŸÁ/ÅgðKñzJñ™rÅgô^ŠÏÄŸ‰oùd ,gz)ÅôRŠÿ؇¥øO5ù_qT“ÿ“’êˆÿàÔ¼V jŽ§ M2G˜çêFN(:‘pTã’y¿ Úó8†k‚n":÷qâïzÿ~Y†²è¡[îrÑbtßDÒÓçn endstream endobj 76 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­Újµ×œn¤"Hý÷ë73´»«ªÐóøÍøÍÃØWßžf¶í_Ü,ºÕêÑýehܬü¾=WWUß\Ž®;ÿp®uí4;Þ©‡¡ožÜY]—›jÓíÏ7ž¼éšÃ¥uësRá^÷Ýë¨ëg÷kæšñ`´Ö³æ8/'7ÌÎf¦‘ò¼?<õ –òSê³)E~ºaÜ÷Ý2·~ÞÖ][öG46s§æ“ÜݾkQ¨^ 70¡j÷ÍYFônŽÞ!$?½gwÜt»>X.ÕüÑOŽçáôÞóû¡uþ{Uן ô„§Ëétp£t°Z©Öí|]ïËíÑ©ù]¿3ŸßNN…46¬±é[7ž¶¶Ý« –Z¯Ô²®WëÚÿæLÄ)/»‰»ö\]ûW¨£|, ’MHS"#p òÐ# µÇ>y×È| #¥Ê:##Ï0)%V©¸†ÇÁ²£â” Œ55¡)°£FÑšSj­‘R—@J]!À5£w+>Ã7+>S}ñú­ø …øŒu ñüB|¿Ÿ¡§Ÿ)W|Fï…øL|ñ™øŸ „å,B/…ø^ ñû°ÿ©&ÿ+–jòRPñœŠ×ŠáCÅñ8ƒ£Iæó\UË E'Žj\2ïwAsMÐMDç>Nü}çÞ/«SB=tËM-F÷uð‰>ç endstream endobj 77 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œn¤" ‡þûõ›™´»«ªÐóøÍøÍÃØW߶3ÛöÏnÝjõèÆþ<4nV~ß‚««ªoÎG×M?œk]{™ïÔÃÐ7[7©ërSmºÃtãÉ›®y=·îÂúœT¸—C÷AÁ:êúÉýš¹fœŒÖzÖgãùä†Ùdf)O‡éÕS¿`)?¥>›RTà§ÆCßÝ)sëç}`ݵeDcc0qj~‘»?tí Õ3ô&Tí¡™dDïæèBòömœÜqÓíû`¹TóG?9NÃé½ æ÷Cë†C÷¢®?è Ûóéôê Fé`µR­ÛûºÞ—»£Só/º~g>½œ ilXcÓ·n<í7ìº,µ^©e]¯×µÿÍ™ˆSž÷îÚsuí_¡ŽòU°4H6!L‰@Œ@ÂÈBŒ€Öû@æq\s óŒ”*댌C¿Ÿ¡³Ÿ±n!>ƒ_ˆÏàâ3ôâ3åŠÏ轟‰/>?ã“°œEè¥ÿÑK!þcâ?ÕäÅRMþO ª#þƒSñZ1|¨8Gxp4Éaž«j9¡èDÂQKæý.hÎÃ௠º‰èÜljèÜûeuêOÈ¢‡n¹ËE‹Ñ}üMçA endstream endobj 78 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œn¤"Bý÷ë73´»«ªÐóøÍøÍÃØW߶3ÛöÏnÝjõèÎýehܬü¾;WWUß\Ž®8׺vš=ß©‡¡o¶nT×å¦Út‡ñÆ“7]óziÝÄúœT¸—C÷AÁ:êúÉýš¹fÖzÖgçËÉ ³ÑÌ4Rž㫧~ÁR~J}6¥¨ÀO7œ}w§Ì­Ÿ÷u×–ýƒ¹ˆSóIîþеƒ(TÏИPµ‡f”½›£wÉÛ·ó莛nßË¥š?úÉó8¼‘Þ›`~?´n8t/êú3ž°½œN¯b”V+Õº½¯ë}ù±;:5ÿ¢ëwæÓÛɩƆ56}ëΧ]ã†]÷â‚¥Ö+µ¬ëUàºö¿9qÊó~â®=W×þê(_KƒdRÀ”Ä$ˆ| 4Àhí±dÇ52HÀH©²NÀÈÁÈsL F‰U*®áq°¬À¨8¥cMMh ¬Á¨Q´æ”Ek¤Ô%RWp ÑþÔg¾˜ún~ï±È‹Cam `4®Ã2ŽW0A/ˆc±ŒŽ9^'Œ-pʹ)pÆqâçœ[[ŽÃi]ðºÄ)9^Wì5jšˆs7¬¡ œGÓb_#xb¸~ †ë' ŒË1}éêcLuÖñ'Nýo ¼Þ…‹öë0à&š1úIe7À£Œ¼6ÞZ cx…Œ¡-‹£~¶`¼&¯ y‘%ŒQ?K'ãþ¡!/h]òÂ,àKV0†æ¬d ²Š1Õ§þMŒ=‘ÕŒ¡3gý1ÖÍY ~Îúiå¬?†žœõ'”Ëúôž³þ”ø¬?!>ëL¡9g)¾qÎ:#Êeå²NC?ûlÑ‹ŸÑ£Ÿc`ñ™8â3Ö²â3ö¥Ÿ±®Ÿá§Ÿ‰#>£w+>C¿Ÿ¡ÓŠÏèÝŠÏðÍŠÏT_|†~+>Cg!>cÝB|¿ŸÁ/Ägè)ÄgÊŸÑ{!>_|&~Æ'a9‹ÐK!þ£—BüÇ>,ĪÉÿŠ¥šüŸTGü§âµbøPq<Žðàh’9Â> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œn¤"Bý÷ë73´»«ªÐóøÍøÍÃØW߶3ÛöÏnÝjõèÎýehܬü¾;WWUß\Ž®8׺vš=ß©‡¡o¶nT×å¦Út‡ñÆ“7]óziÝÄúœT¸—C÷AÁ:êúÉýš¹fM¨õ¬9ÎΗ“f£™i¤<ÆWOý‚¥ü”úlJQŸn8úîN™[­µ¬»¶ìhìÌEœšOr÷‡®D¡z†ÞÀ„ª=4£ŒèݽCHÞ¾GwÜtû>X.ÕüÑOžÇáôÞóû¡uá{Qן ô„íåtzu£t°Z©Öí}]ïËÝÑ©ù]¿3ŸÞNN…46¬±é[w>í7ìº,µ^©e]¯×µÿÍ™ˆSž÷wí¹ºö¯PGù*X$›¦D F á@ä¡F@k} ó8®9ù@FJ•uFFž#`R0J¬Rq ƒeFÅ)kjBS` F¢5§Ô(Z#¥.9€”ºB€kxŒö§>óÅÔwó{7ˆE^ kC¸X£q– pD¸‚ zA‹etÌñ 8alSÎM3Ž?çÜØrNë‚×%NÉñ¸b¯QÓDœ‹¸a á”8š>ûÁÃõ#h0\?Ið`\~ŒéKWcª³þ‹?qêcà…ð.\„´X‡7ÑŒÑO*»eäµñÖzlÃë,d mYÄõ³ã50ymÈ‹,aŒúYʘ8÷ ™xAë’f_²‚14g%cx”UŒ©>õob쉬f 9ë±nÎúcðsÖO{(gý1ôä¬?¡\ÖŸ ÷œõ§Ägý ñYg Í9ëLñsÖQ.ëŒ(—uú±Øg‹^¬øŒ­ø‹ÏÄŸ±–Ÿ±/­øŒu­ø ?­øLñ½[ñú­ø V|FïV|†oV|¦úâ3ô[ñ: ñëâ3ø…ø ~!>CO!>S®øŒÞ ñ™øâ3ñ3>ËY„^ ñ½â?öa!þSMþW,Õäÿ¤ :â?8¯ÇŠãq„G“Ì湪–ŠN$Õ¸dÞï‚æ2 þš ›ˆÎ}œø‡Î½_V§þ„,zè–›.ZŒîëàÍçP endstream endobj 86 0 obj << /Producer (pdfTeX-1.40.25) /Author()/Title()/Subject()/Creator(LaTeX with hyperref)/Keywords() /CreationDate (D:20241029220654-04'00') /ModDate (D:20241029220654-04'00') /Trapped /False /PTEX.Fullbanner (This is pdfTeX, Version 3.141592653-2.6-1.40.25 (TeX Live 2023/Debian) kpathsea version 6.3.5) >> endobj 2 0 obj << /Type /ObjStm /N 63 /First 493 /Length 3104 /Filter /FlateDecode >> stream xÚíZÛrÛ8}×Wà1Þ­Ä•àTjª{ìxsqÆvâd3~ %ZæD5$•Iöë÷4@Ñ”dÅv2ó°[©²› Ðh4Nß@Š‚ÅÌ03Ë„YÊD33•:&Ó’+™P,QŠ Í\‚K®8°ÅF1ôÄiÂÐ+´`Âàê M™¦L‚Kc Í7F0…q“&iU…q̳N hµ7èN¦LcÙ“µb26†iͤP Ó˜*ŒeXJ*'º¤†0åpMb¦qÅ:bk±f2 W,í°M#! ZãO;t Iè‹%Lb-³1ä™”Y鳂ÙDé1ˆ5,‘©E'K äÙ”%©ŽY"˜ ôN˜ÓP^JæДŠ¹4Á=ÐC¤,Åü²,ÕÆ1t¥ƒ×û‘‚¥)m8Å"¶Ì‘Q#¸%¦kº&°Ž Â,kçŽLaiñXë˜#)COž0~ÊøayV2¾ÏÕù°)ÊY$vØÏ?ý¦‹ÿä·8Ž_Ù#²Kä·XÇt­ˆ”D.‰,ˆÔDš>WÓq-{ÊŽÛMˆDòNj½³U?¹©ßñ=öȈ̈ŒúÝÓn^Ý‘¢c÷Y?ub&ÝÚyŸ¡èÎW•ËºVÑa1ÛØí„[ÌN?(š Úï“'~öež3þ:ç¾WΚ|ÖÔbð“¼.Õ0¯)Þ|ÏË|TdOËÏìCLqšH&ä"réÅr* ¯òÜ»³Y q(ÂieD¸¿h¹¬èà™üiYòÊ‹ü?â{¸^>ÃÛHd¤)æ´ \Ë8J±7äH'|§‹Ë2ù‹bö‘ï>yâWà»ÞÖü”¿99¢ÿG×M3ÿ‰óË¢,fWe5ÍšbXGÓy1?—“)šãh”óèy›ªí2̯³]ó¸Ó\«4Bx3em”"°4v`Í*1‘²|»lÝE‡E“G×ÙŸ‹Y‡8úKtè!c(c"ÊDB©È í)«€nº]—I±ˆU_žg½{ÿoØZE.–È[*J‘ÿf‹Éäb+¯¼ÂFq*ïÅkbÅÚÞ‹W¦*ëràæ~kÈæÈt!&¨HÄmY].cÉ[Ê%OB‰¶í—”L© ±üuUOsÀŠXÚ?`ü,ÿܬ;øZ)¹d*þ® CÛV÷ðÚñH!MTùÒÈÁŒR¸H£b¨jÄöVÿ¸*Æ‹*Äšw¨5Ë$(ýÛ¼CÝbEel„¹¯’‘±ú^¼Ò¹(jÛ¼£omE%tÝò«Þ¯ì÷x„Ž7<Â}ƒG´KìUyF o?kröhÿ'K’žJ[#w ´m;+š FNØa•Í¯‘Ùñ¢™/š¿±Ñb˜W4ª# k·k•Ôµ³ºCB³ç|Ô)üU6¥1 ?ÌêÜ#Οå“O9p6à¿Ì†å¨˜éÖnè?ÚÛ#æÓã‹Þ"Kþ ®›ÍÏ‹Ùî¬.n†÷‹««P´ O‹Ù¢ÆɈÿ±(›|’_5(NšJì¾® >®²O9φ‹&çâ.¦W“ü3oŠ ÊÂ4V¨'—ULɆCX ,Q5x£üŠWX›aÁÉ$ë:¯³qV-¦“lÑðr\Îò|˜‘¼zž ó×MïŸ\Wx-"ÕF¢^šønNnûîøòwŠ:šÂ»Ó{xô…â‘$ŽÎi¥ßEƒœ@P»B‘”"±ìÑHN81» /œ¬]i¡%q8™–Oá”m„¢~¶¥<€ó²w+)ôûùí Ê@#ƒ"©Ã ÏS: wµ¨B}B›zƒ¸þXŠœ èúmHíó„Y¡ßÈp`˹=þÀC™gÔËDªs+£´™ÐòôbðÁHøAP¢?<â/B1DfšR~ˆÛH\‰h§¾ÚØô4Äu ý<ÚýþÐîç– ?ÐÀ³2ú#Qÿ$jí%<­uÁÍzÔˆ«·=*&39A«{­œ–„œ$=m ×´¾×’x§5l‚¶³1a Jo„ô’ÅICÓ ¯öm'’H¸!Š^f™vY!`Ȥ½ ÔzGjQíí%ôû£$V{©ebæ7c-Ôç(Ü÷ò3:Î0NéŠiE©åÒOüYCãþ ÿ#S°%J¢Ú¥”¬ðÑ+È·¤õ)’Úaô{hhX¥Oµ•X«íñþeS*IŒÈÓ>ÅYü+òåÇyÑr)ãKF,ƒTòGŠz?îǃDšÛÊ–äÇFû±¸ãòTıén5)åé>¡úû`…Ñ X#pÅJ["Õ‰žßopIÛãéÐÎíñž¥Á,¥ßîÒ†BËS8AJBX{Ø´óÕEžÒ&”0,å¡ÀCýuEɯm#Sò¦´ã¼tÏçij Éêõ˜DR>íiK¹.´CØ%¹Z$í ¡%]Ò ܾ/5ÝìÄùBsnäøqP6Ú_hcìbí|?¯‡U1Ç#zxò à¿>}õ¯Wÿ<=xúO½1†&Ù¸f:ð<õïKáØc%èw#‘$ý[Êšžtæìeógy1¾noi%{,RÈ9j²I1Ü'9Ãài“OßÂmñ˜ÙÎA!…ˆë¬¢ËG|—ïñcþšŸðŒ_ò!ñœ_ñ1/ø„OùŒ—Ê|Î+^ó†7–|Á?í} ,blÿÈ]<;?:?y N^" ¶ \ ¤`BCzª’.]€n €Š×Øçüп“{Î_ð—üU É)?ãoù9ïÁ¹¬²áǼ¡wËvEÚ°œ”3Ðé4óæ³QV_óüE6áùçá$›Õ+ú+¸ÿû”äk~ýe~Ï€öïüãâszÅäó­°Ô<¯ü+\‹rÄÿðv©ói4¨óOVŸy=!`±ë*Ï—vãòÏü ÿO^•+&t1á‹“ãó7¿Êí&4dÂ&¤È´ü{MøF |Ù™î¼õè«þrQ­â¾D}Vôò+°Üý ¸ H?Èw¿¾}yt€L¤Ûd¢) h(Èôïò [MäŽa׋•Ê‡lôàýÞÙá16zúâ+AoÄ2èá;ßôÓ{ìtß»ÈÛµ¤×yÅ2ç«–§ú²‚ƒ~Pò;<<<G8œ}8Ñc"P}qw=„XÁn;¤”߀CÈ~¿ ÿ­ç½7wç¼.Ûù4wµÌjcz5Wý(›pz»{¯ôæÓ½]Nú°oÛƒ Ïþ»÷Ï^>ög_Á^Æn‰=ýÿíØß·ð°oœpê}®ÛõÊv”¤ß¾zu¼¶+·oW.·«é11ý;·»LÆT¦ºx»#âÚ"Ö‡!¹%Ånù¹AŠ¶ãÄëkoçNªÀ8§.‹òÒÓpœ[ÖV<^ßêœ89µ>ß*Ý*”$ïôžMÕ!]_l|Lø‚L¥§ÔPi–åzY¸}x„8jÇYÓ¶Ÿ¾¼á‰¾DͺhÅ9#ü=Œý¨E&] \–&ëIoíô1oOÍ͉áâ>?1Ýü°´yÊ]ñSúÇÿÆvPTuC‚ƒõ€¿ÈÚ!P Ï‹Qs]ÓW=žõæªvîYùfìFðˆ-µèknC×4L×5”¢¯!Ý-5wk¨®áæ)kUC+60Œ·i¨ïÖP?\ÃÍãËš†j]ÃD®X9¹ÑÐÜ­¡y¸†›çŽ5 ͆f †ÚÝ­¡ý?Ü8¬i¸)JoÓ0½[ÃäánÖÍ5 7"Åšm±ß­¡{¸†›¥nUÃD|ÝÊÂÞh¨îÖ0ÝÔ>*¨é«‚ôQþ¼ÕÝ×1ígágâµDz¼h&¨u«k•tc­Ã¡²›ç¿h1ßeíWcí×®øu•b«í§-?_aâÖi¯è×åÐY;y´_Ÿ6YÕì0>yÙøHJÚ^ûÁ’LBßÍG*ªí™1ºý{¹C“z”ø‹bZ4kK/¹.ÖÕóýXO¨—ßþïhVú|×j–MÖW]Jì1u«뺰–“_Sø–ÙpL°¹[*Â^†U9wê>6ZúÈ™ü®^Úpœ¿„cò7u~ãNÇó|>¾£Û¥þ j ú( endstream endobj 87 0 obj << /Type /XRef /Index [0 88] /Size 88 /W [1 3 1] /Root 85 0 R /Info 86 0 R /ID [<931FB91817A4BBB38327D042CDAE4DBB> <931FB91817A4BBB38327D042CDAE4DBB>] /Length 249 /Filter /FlateDecode >> stream xÚÐ9/P†áó]ÛØfÆ>†±ïËØ·j ‰¿ ÕM¥PIT"‘hüD¢PH$£–(D¡Ó*¸ïiž¼7÷ÜS\3³¿`LáíÔ¨zXA€X ¨„*¨†H@£ÌR¾¥ê`fa¦a` š¡R²tÉ´BR–8óc ÐóÐ Yh“%Ë>×íÐè”mßùíô@/ä ú!0(Û+ú‹!ÙÉ·×°ìýÐkDÊyJ»^cÒù׸ô{ë5¡°ÿå5©pu¿øæ rŸ<¼D#O…ÈógäõÒ‡—a Öa6aK¡¼ç>2öΤ#æ endstream endobj startxref 188696 %%EOF pcaMethods/inst/doc/pcaMethods.R0000644000175200017520000001310614710312420017612 0ustar00biocbuildbiocbuild### R code from vignette source 'pcaMethods.Rnw' ################################################### ### code chunk number 1: pcaMethods.Rnw:102-114 ################################################### library(pcaMethods) x <- c(-4,7); y <- c(-3,4) distX <- rnorm(100, sd=0.3)*3 distY <- rnorm(100, sd=0.3) + distX * 0.3 mat <- cbind(distX, distY) res <- pca(mat, nPcs=2, method="svd", center=F) loading <- loadings(res)[1,] grad <- loading[2] / loading[1] if (grad < 0) grad <- grad * -1 lx <- c(-4,7) ly <- c(grad * -4, grad * 7) ################################################### ### code chunk number 2: pcaMethods.Rnw:118-125 ################################################### par(mar=c(2, 3, 2, 2)) plot(x,y, type="n", xlab="", ylab="") abline(v=0, col="dark gray", lwd = 2); abline(h=0, col = "dark gray", lwd = 2) points(distX, distY, type = 'p', col = "blue") lines(lx,ly, lwd = 2) points(-1, -1 * grad + 0.5, pch = 19, col = "red", lwd=4) points(6, 6 * grad + 0.5, pch = 19, col = "red", lwd=4) ################################################### ### code chunk number 3: pcaMethods.Rnw:253-255 ################################################### library(lattice) library(pcaMethods) ################################################### ### code chunk number 4: pcaMethods.Rnw:258-261 ################################################### library(pcaMethods) data(metaboliteData) data(metaboliteDataComplete) ################################################### ### code chunk number 5: pcaMethods.Rnw:264-266 ################################################### md <- prep(metaboliteData, scale="none", center=TRUE) mdC <- prep(metaboliteDataComplete, scale="none", center=TRUE) ################################################### ### code chunk number 6: pcaMethods.Rnw:271-277 ################################################### resPCA <- pca(mdC, method="svd", center=FALSE, nPcs=5) resPPCA <- pca(md, method="ppca", center=FALSE, nPcs=5) resBPCA <- pca(md, method="bpca", center=FALSE, nPcs=5) resSVDI <- pca(md, method="svdImpute", center=FALSE, nPcs=5) resNipals <- pca(md, method="nipals", center=FALSE, nPcs=5) resNLPCA <- pca(md, method="nlpca", center=FALSE, nPcs=5, maxSteps=300) ################################################### ### code chunk number 7: pcaMethods.Rnw:293-296 ################################################### sDevs <- cbind(sDev(resPCA), sDev(resPPCA), sDev(resBPCA), sDev(resSVDI), sDev(resNipals), sDev(resNLPCA)) matplot(sDevs, type = 'l', xlab="Eigenvalues", ylab="Standard deviation of PC", lwd=3) legend(x="topright", legend=c("PCA", "PPCA", "BPCA", "SVDimpute","Nipals PCA","NLPCA"), lty=1:6, col=1:6, lwd=3) ################################################### ### code chunk number 8: pcaMethods.Rnw:308-311 ################################################### par(mfrow=c(1,2)) plot(loadings(resBPCA)[,1], loadings(resPCA)[,1], xlab="BPCA", ylab="classic PCA", main = "Loading 1") plot(loadings(resBPCA)[,2], loadings(resPCA)[,2], xlab="BPCA", ylab="classic PCA", main = "Loading 2") ################################################### ### code chunk number 9: pcaMethods.Rnw:335-337 ################################################### q2SVDI <- Q2(resSVDI, mdC, fold=10) q2PPCA <- Q2(resPPCA, mdC, fold=10) ################################################### ### code chunk number 10: pcaMethods.Rnw:339-349 ################################################### # PPCA does not converge / misestimate a value in very rare cases. # This is a workaround to avoid that such a case will break the # diagram displayed in the vignette. # From the 2.0 release of bioconductor on, the convergence threshold # for PPCA was lowert to 1e-5, this should make the method much more # stable. So this workaround might be obsolete now... # [nope it is not, ppca is unstable] while( sum((abs(q2PPCA)) > 1) >= 1 ) { q2PPCA <- Q2(resPPCA, mdC, fold=10) } ################################################### ### code chunk number 11: pcaMethods.Rnw:353-356 ################################################### q2 <- data.frame(Q2=c(drop(q2PPCA), drop(q2SVDI)), method=c("PPCA", "SVD-Impute")[gl(2, 5)], PC=rep(1:5, 2)) print(xyplot(Q2~PC|method, q2, ylab=expression(Q^2), type="h", lwd=4)) ################################################### ### code chunk number 12: pcaMethods.Rnw:389-390 ################################################### errEsti <- kEstimate(md, method = "ppca", evalPcs=1:5, nruncv=1, em="nrmsep") ################################################### ### code chunk number 13: pcaMethods.Rnw:396-397 ################################################### barplot(drop(errEsti$eError), xlab="Loadings", ylab="NRMSEP (Single iteration)") ################################################### ### code chunk number 14: pcaMethods.Rnw:420-421 ################################################### barplot(drop(errEsti$variableWiseError[, which(errEsti$evalPcs == errEsti$bestNPcs)]), xlab="Incomplete variable Index", ylab="NRMSEP") ################################################### ### code chunk number 15: pcaMethods.Rnw:445-446 ################################################### slplot(resPCA) ################################################### ### code chunk number 16: pcaMethods.Rnw:456-457 ################################################### plotPcs(resPPCA, pc=1:3, type="score") ################################################### ### code chunk number 17: pcaMethods.Rnw:469-475 ################################################### pc <- pca(iris) irdf <- merge(iris, scores(pc), by=0) library(ggplot2) ggplot(irdf, aes(PC1, PC2, colour=Species)) + geom_point() + stat_ellipse() pcaMethods/inst/doc/pcaMethods.Rnw0000644000175200017520000005554214710217306020200 0ustar00biocbuildbiocbuild\documentclass[a4paper]{article} %\VignetteIndexEntry{Introduction} \usepackage{hyperref} \title{The pcaMethods Package} \author{Wolfram Stacklies and Henning Redestig\\ CAS-MPG Partner Institute for Computational Biology (PICB)\\ Shanghai, P.R. China \\ and\\ Max Planck Institute for Molecular Plant Physiology\\ Potsdam, Germany\\ \url{http://bioinformatics.mpimp-golm.mpg.de/} } \date{\today} \begin{document} \setkeys{Gin}{width=1.0\textwidth} @ \maketitle \section*{Overview} The \texttt{pcaMethods} package \cite{stacklies07} provides a set of different PCA implementations, together with tools for cross validation and visualisation of the results. The methods basically allow to perform PCA on incomplete data and thus may also be used for missing value estimation. When doing PCA one assumes that the data is restricted to a subspace of lower dimensionality, e.g. correlation patterns between jointly regulated genes. PCA aims to extract these structures thereby filtering noise out. If only the most significant loadings (eigenvectors, also referred to as principal components) are used for projection this can be written as: \begin{equation} X = 1\times{}\bar{x}^T + TP^T + V \end{equation} Where the term $1\times{}\bar{x}^T$ represents the original variable averages, $X$ denotes the observations, $T={t_1, t_2,\ldots,t_k}$ the latent variables or scores, $P={p_1, p_2,\ldots,p_k}$ the transformation matrix (consisting of the most significant eigenvectors of the covariance matrix) and $V$ are the residuals. Missing values may be estimated by projecting the scores back into the original space using $\hat{X} = 1\times{}\bar{x}^T + TP^T$. Optimally, this produces an estimate of the missing data based on the underlying correlation structure, thereby ignoring noise. This will only produce reasonable results if the residuals $V$ are sufficiently small, implying that most of the important information is captured by the first $k$ components. In order to calculate the transformation matrix $P$ one needs to determine the covariance matrix between variables or alternatively calculate $P$ directly via SVD. In both cases, this can only be done on complete matrices. However, an approximation may be obtained by use of different regression methods. The PCA methods provided in this package implement algorithms to accurately estimate the PCA solution on incomplete data. Although the focus of this package is clearly to provide a collection of PCA methods we also provide a cluster based method for missing value imputation. This allows to better rate and compare the results. \section{Algorithms} All methods return a common class called \texttt{pcaRes} as a container for the results. This guarantees maximum flexibility for the user. A wrapper function called \texttt{pca()} is provided that receives the desired type of pca as a string. \subsection*{svdPca} This is a wrapper function for $R's$ standard \texttt{prcomp} function. It delivers the results as a \texttt{pcaRes} object for compatibility with the rest of the package. \subsection*{svdImpute} This implements the SVDimpute algorithm as proposed by Troyanskaya et~al \cite{troyanskaya01}. The idea behind the algorithm is to estimate the missing values as a linear combination of the $k$ most significant eigengenes\footnote{The term ``eigengenes'' denotes the loadings when PCA was applied considering variables (here the genes) as observations.}. The algorithm works iteratively until the change in the estimated solution falls below a certain threshold. Each step the eigengenes of the current estimate are calculated and used to determine a new estimate. An optimal linear combination is found by regressing an incomplete variable against the $k$ most significant eigengenes. If the value at position $j$ is missing, the $j^{th}$ value of the eigengenes is not used when determining the regression coefficients.\\ SVDimpute seems to be tolerant to relatively high amount of missing data (> 10\%). \subsection*{Probabilistic PCA (ppca)} Probabilistic PCA combines an EM approach for PCA with a probabilistic model. The EM approach is based on the assumption that the latent variables as well as the noise are normal distributed. In standard PCA data which is far from the training set but close to the principal subspace may have the same reconstruction error, see Figure \ref{fig:pcaSubspace} for explanation. <>= library(pcaMethods) x <- c(-4,7); y <- c(-3,4) distX <- rnorm(100, sd=0.3)*3 distY <- rnorm(100, sd=0.3) + distX * 0.3 mat <- cbind(distX, distY) res <- pca(mat, nPcs=2, method="svd", center=F) loading <- loadings(res)[1,] grad <- loading[2] / loading[1] if (grad < 0) grad <- grad * -1 lx <- c(-4,7) ly <- c(grad * -4, grad * 7) @ \begin{figure} \centering <>= par(mar=c(2, 3, 2, 2)) plot(x,y, type="n", xlab="", ylab="") abline(v=0, col="dark gray", lwd = 2); abline(h=0, col = "dark gray", lwd = 2) points(distX, distY, type = 'p', col = "blue") lines(lx,ly, lwd = 2) points(-1, -1 * grad + 0.5, pch = 19, col = "red", lwd=4) points(6, 6 * grad + 0.5, pch = 19, col = "red", lwd=4) @ \caption{Normal distributed data with the first loading plotted in black. The two red points have the same reconstruction error because PCA does not define a density model. Thus the only measure of how well new data fits the model is the distance from the principal subspace. Data points far from the bulk of data but still close to the principal subspace will have a low reconstruction error. \label{fig:pcaSubspace}} \end{figure} PPCA defines a likelihood function such that the likelihood for data far from the training set is much lower, even if they are close to the principal subspace. This allows to improve the estimation accuracy.\\ PPCA is tolerant to amounts of missing values between 10\% to 15\%. If more data is missing the algorithm is likely not to converge to a reasonable solution. The method was implemented after the draft ``\texttt{EM Algorithms for PCA and Sensible PCA}'' written by Sam Roweis and after the Matlab \texttt{ppca} script implemented by \emph{Jakob Verbeek}\footnote{\url{http://lear.inrialpes.fr/~verbeek/}}. Please check also the PPCA help file. \subsection*{Bayesian PCA (bpca)} Similar to probabilistic PCA, Bayesian PCA uses an EM approach together with a Bayesian model to calculate the likelihood for a reconstructed value.\\ The algorithm seems to be tolerant to relatively high amounts of missing data (> 10\%). Scores and loadings obtained with Bayesian PCA slightly differ from those obtained with conventional PCA. This is because BPCA was developed especially for missing value estimation and is based on a variational Bayesian framework (VBF), with automatic relevance determination (ARD). In BPCA, ARD leads to a different scaling of the scores, loadings and eigenvalues when compared to standard PCA or PPCA. The algorithm does not force orthogonality between loadings. However, the authors of the BPCA paper found that including an orthogonality criterion made the predictions worse. They also state that the difference between ``real'' and predicted Eigenvalues becomes larger when the number of observation is smaller, because it reflects the lack of information to accurately determine true loadings from the limited and noisy data. As a result, weights of factors to predict missing values are not the same as with conventional PCA, but the missing value estimation is improved. BPCA was proposed by Oba et~al \cite{oba03}. The method available in this package is a port of the \texttt{bpca} Matlab script also provided by the authors\footnote{ \url{http://hawaii.aist-nara.ac.jp/\%7Eshige-o/tools/}}. \subsection*{Inverse non-linear PCA (NLPCA)} NLPCA \cite{scholz05} is especially suitable for data from experiments where the studied response is non-linear. Examples of such experiments are ubiquitous in biology -- enzyme kinetics are inherently non-linear as are gene expression responses influenced by the cell cycle or diurnal oscillations. NLPCA is based on training an auto-associative neural network composed of a component layer which serves as the ``bottle-neck'', a hidden non-linear layer and an output layer corresponding to the reconstructed data. The loadings can be seen as hidden in the network. Missing values in the training data are simply ignored when calculating the error during back-propagation. Thus NLPCA can be used to impute missing values in the same way as for conventional PCA. The only difference is that the loadings $P$ are now represented by a neural network.\\ A shortcoming of the current implementation is that there is no reasonable stop criterion. The quality of the estimated solution depends on the number of iterations. This should in most cases be somewhat between 500 and 1500. We recommend to use \texttt{kEstimate} or \texttt{kEstimateFast} to determine this parameter. \subsection*{Nipals PCA} Nipals (Nonlinear Estimation by Iterative Partial Least Squares) \cite{wold66} is an algorithm at the root of PLS regression which can execute PCA with missing values by simply leaving those out from the appropriate inner products. It is tolerant to small amounts (generally not more than 5\%) of missing data. \subsection{Local least squares (LLS) imputation} The package provides an algorithm called \texttt{llsImpute} for missing value estimation based on a linear combination of the $k$ nearest neighbours of an incomplete variable (in Microarray experiments normally a gene). The distance between variables is defined as the absolute value of the Pearson, Spearman or Kendall correlation coefficient. The optimal linear combination is found by solving a local least squares problem as described in \cite{kim05}. In tests performed in the cited paper the llsImpute algorithm is able to outperform knnImpute\cite{troyanskaya01} and competes well with BPCA. In the current implementation two slightly different ways for missing value estimation are provided. The first one is to restrict the neighbour searching to the subset of complete variables. This is preferable when the number of incomplete variables is relatively small. The second way is to consider all variables as candidates. Here, missing values are initially replaced by the columns wise mean. The method then iterates, using the current estimate as input for the LLS regression until the change between new and old estimate falls below a certain threshold (0.001). \section{Getting started} \paragraph{Installing the package.} To install the package first download the appropriate file for your platform from the Bioconductor website (\url{http://www.bioconductor.org/}). For Windows, start \texttt{R} and select the \texttt{Packages} menu, then \texttt{Install package from local zip file}. Find and highlight the location of the zip file and click on \texttt{open}. For Linux/Unix, use the usual command \texttt{R CMD INSTALL} or set the option \texttt{CRAN} to your nearest mirror site and use the command \texttt{install.packages} from within an \texttt{R} session. \paragraph{Loading the package:} To load the \texttt{pcaMethods} package in your \texttt{R} session, type \texttt{library(pcaMethods)}. \paragraph{Help files:} Detailed information on \texttt{pcaMethods} package functions can be obtained from the help files. For example, to get a description of \texttt{bpca} type \texttt{help("bpca")}. \paragraph{Sample data:} Two sample data sets are coming with the package. \texttt{metaboliteDataComplete} contains a complete subset from a larger metabolite data set. \texttt{metaboliteData} is the same data set but with 10 \% values removed from an equal distribution. \section{Some examples} <>= library(lattice) library(pcaMethods) @ To load the package and the two sample data sets type: <>= library(pcaMethods) data(metaboliteData) data(metaboliteDataComplete) @ Now centre the data <<>>= md <- prep(metaboliteData, scale="none", center=TRUE) mdC <- prep(metaboliteDataComplete, scale="none", center=TRUE) @ Run SVD pca, PPCA, BPCA, SVDimpute and nipalsPCA on the data, using the \texttt{pca()} wrapper function. The result is always a \texttt{pcaRes} object. <>= resPCA <- pca(mdC, method="svd", center=FALSE, nPcs=5) resPPCA <- pca(md, method="ppca", center=FALSE, nPcs=5) resBPCA <- pca(md, method="bpca", center=FALSE, nPcs=5) resSVDI <- pca(md, method="svdImpute", center=FALSE, nPcs=5) resNipals <- pca(md, method="nipals", center=FALSE, nPcs=5) resNLPCA <- pca(md, method="nlpca", center=FALSE, nPcs=5, maxSteps=300) @ Figure \ref{fig:eigenvalues} shows a plot of the eigenvalue structure (\texttt{sDev(pcaRes)}). If most of the variance is captured with few loadings PCA is likely to produce good missing value estimation results. For the sample data all methods show similar eigenvalues. One can also see that most of the variance is already captured by the first loading, thus estimation is likely to work fine on this data. For BPCA, the eigenvalues are scaled differently for reasons discussed above, see Figure \ref{fig:loadingBPCA}. The order of the loadings remains the same. \begin{figure} \centering <>= sDevs <- cbind(sDev(resPCA), sDev(resPPCA), sDev(resBPCA), sDev(resSVDI), sDev(resNipals), sDev(resNLPCA)) matplot(sDevs, type = 'l', xlab="Eigenvalues", ylab="Standard deviation of PC", lwd=3) legend(x="topright", legend=c("PCA", "PPCA", "BPCA", "SVDimpute","Nipals PCA","NLPCA"), lty=1:6, col=1:6, lwd=3) @ \caption{Eigenvalue structure as obtained with different methods\label{fig:eigenvalues}} \end{figure} To get an impression of the correctness of the estimation it is a good idea to plot the scores / loadings obtained with classical PCA and one of the probabilistic methods against each other. This of course requires a complete data set from which data is randomly removed. Figure \ref{fig:loadingBPCA} shows this for BPCA on the sample data. \begin{figure} \centering <>= par(mfrow=c(1,2)) plot(loadings(resBPCA)[,1], loadings(resPCA)[,1], xlab="BPCA", ylab="classic PCA", main = "Loading 1") plot(loadings(resBPCA)[,2], loadings(resPCA)[,2], xlab="BPCA", ylab="classic PCA", main = "Loading 2") @ \caption{Loading 1 and 2 calculated with BPCA plotted against those calculated with standard PCA. \label{fig:loadingBPCA}} \end{figure} \section{Cross validation} \texttt{Q2} is the goodness measure used for internal cross validation. This allows to estimate the level of structure in a data set and to optimise the choice of number of loadings. Cross validation is performed by removing random elements of the data matrix, then estimating these using the PCA algorithm of choice and then calculating $Q^2$ accordingly. At the moment, cross-validation can only be performed with algorithms that allow missing values (i.e. not SVD). Missing value independent cross-validation is scheduled for implementation in later versions. $Q^2$ is defined as following for the mean centered data (and possibly scaled) matrix $X$. $$\mathrm{SSX}=\sum (x_{ij})^2$$ $$\mathrm{PRESS}=\sum (x_{ij} - \hat{x}_{ij})^2$$ $$Q^2=1 - \mathrm{PRESS}/\mathrm{SSX}$$ The maximum value for $Q^2$ is thus 1 which means that all variance in $X$ is represented in the predictions; $X=\hat{X}$. <>= q2SVDI <- Q2(resSVDI, mdC, fold=10) q2PPCA <- Q2(resPPCA, mdC, fold=10) @ <>= # PPCA does not converge / misestimate a value in very rare cases. # This is a workaround to avoid that such a case will break the # diagram displayed in the vignette. # From the 2.0 release of bioconductor on, the convergence threshold # for PPCA was lowert to 1e-5, this should make the method much more # stable. So this workaround might be obsolete now... # [nope it is not, ppca is unstable] while( sum((abs(q2PPCA)) > 1) >= 1 ) { q2PPCA <- Q2(resPPCA, mdC, fold=10) } @ \begin{figure}[!ht] \centering <>= q2 <- data.frame(Q2=c(drop(q2PPCA), drop(q2SVDI)), method=c("PPCA", "SVD-Impute")[gl(2, 5)], PC=rep(1:5, 2)) print(xyplot(Q2~PC|method, q2, ylab=expression(Q^2), type="h", lwd=4)) @ \caption{Boxplot of the \texttt{Q2} results for BPCA, Nipals PCA, SVDimpute and PPCA. PPCA and SVDimpute both deliver better results than BPCA and Nipals in this example.\label{fig:Q2}} \end{figure} The second method called \texttt{kEstimate} uses cross validation to estimate the optimal number of loadings for missing value estimation. The \texttt{NRMSEP} (normalised root mean square error of prediction) \cite{feten05} or Q2 can be used to define the average error of prediction. The NRMSEP normalises the square difference between real and estimated values for a certain variable by the variance within this variable. The idea behind this normalisation is that the error of prediction will automatically be higher if the variance is higher. The \texttt{NRMSEP} for mean imputation is $\sqrt{\frac{nObs}{nObs - 1}}$ when cross validation is used, where $nObs$ is the number of observations. The exact definition is: \begin{equation} NRMSEP_k = \sqrt{\frac{1}{g} \sum_{j \in G} \frac{\sum_{i \in O_j} (x_{ij} - \hat{x}_{ijk})^2}{o_j s_{x_j}^2}} \end{equation} where $s^2_{x_j} = \sum_{i=1}^n (x_{ij} - \overline{x}_j)^2 / (n - 1)$, this is the variance within a certain variable. Further, $G$ denotes the set of incomplete variables, $g$ is the number of incomplete varialbes. $O_j$ is the set of missing observations in variable $j$ and $o_j$ is the number of missing observations in variable $j$. $\hat{x}_{ijk}$ stands for the estimate of value $i$ of variable $j$ using $k$ loadings. See Figure \ref{fig:kEstimate} for an example. The NRMSEP should be the error measure of choice. But if the number of observations is small, the variance within a certain variable may become and unstable criterion. If so or if variance scaling was applied we recommend to use Q2 instead. <>= errEsti <- kEstimate(md, method = "ppca", evalPcs=1:5, nruncv=1, em="nrmsep") @ \begin{figure}[!ht] \centering \begin{minipage}[c]{0.6\textwidth} \centering <>= barplot(drop(errEsti$eError), xlab="Loadings", ylab="NRMSEP (Single iteration)") @ \end{minipage} \begin{minipage}[c]{0.3\textwidth} \caption{Boxplot showing the \texttt{NRMSEP} versus the number of loadings. In this example only 1 iteration of the whole cross validation were performed. It is normally advisable to do more than just one iteration. \label{fig:kEstimate}} \end{minipage} \end{figure} \texttt{kEstimate} also provides information about the estimation error for individual variables. The $Q^2$ distance or the NRMSEP are calculated separately for each variable. See the manpage for \texttt{kEstimate} and \texttt{kEstimateFast} for details. Plotting the variable - wise results gives information about for which variables missing value estimation makes sense, and for which no imputation or mean imputation is preferable, see Figure \ref{fig:variableWiseError}. If you are not interested in variable - wise information we recommend to use the faster \texttt{kEstimateFast} instead. \begin{figure}[!ht] \centering \begin{minipage}[c]{0.6\textwidth} \centering <>= barplot(drop(errEsti$variableWiseError[, which(errEsti$evalPcs == errEsti$bestNPcs)]), xlab="Incomplete variable Index", ylab="NRMSEP") @ \end{minipage} \begin{minipage}[c]{0.3\textwidth} \caption{Boxplot showing the \texttt{NRMSEP} for all incomplete variables in the data set. For the first 7 variables missing value imputation does not seem to make too much sense. \label{fig:variableWiseError}} \end{minipage} \end{figure} \newpage \section{Visualisation of the results} \subsection{Quick scores and loadings plot} Some methods for display of scores and loadings are also provided. The function \texttt{slplot()} aims to be a simple way to quickly visualise scores and loadings in an intuitive way, see Figure \ref{fig:slplot}. Barplots are provided when plotting only one PC and colours can be specified differently for the scores and loadings plots. For a more specific scatter plot it is however recommended to access scores and loadings slots and define own plot functions. \begin{figure}[!h] \centering <>= slplot(resPCA) @ \caption{\texttt{slplot} for scores and loadings obtained with classical SVD based PCA. \label{fig:slplot}} \end{figure} \noindent Another method called \texttt{plotPcs()} allows to visualise many PCs plotted against each other, see Figure \ref{fig:plotPcs}. \begin{figure}[!ht] \centering <>= plotPcs(resPPCA, pc=1:3, type="score") @ \caption{A plot of score 1:3 for PPCA created with \texttt{plotPcs()} \label{fig:plotPcs}} \end{figure} \subsection{Using ggplot2} For using ggplot, the scores and loadings should best be added to a data frame that add other relevant descriptive factors. For example, after doing PCA on the Iris dataset, we may add the scores back to the original data frame and use ggplot to visualise, see Figure \ref{fig:ggplot}. \begin{figure}[!ht] \centering <>= pc <- pca(iris) irdf <- merge(iris, scores(pc), by=0) library(ggplot2) ggplot(irdf, aes(PC1, PC2, colour=Species)) + geom_point() + stat_ellipse() @ \caption{Score plot using ggplot2} \label{fig:ggplot} \end{figure} \cleardoublepage \begin{thebibliography}{2006} \bibitem{stacklies07} Stacklies W., Redestig H., Scholz M., and Walther D., and Selbig J. {\sl pcaMethods -- a Bioconductor package providing PCA methods for incomplete data} Bioinformatics. 2007, 23, 1164-1167. {\sl Non-linear PCA: a missing data approach.} Bioinformatics. 2005, 21, 3887-3895. \bibitem{scholz05} Scholz, M. , Kaplan, F., Guy, C.L., Kopka, J. and Selbig, J. {\sl Non-linear pca: a missing data approach.} Bioinformatics. 2005, 21, 3887-3895. \bibitem{troyanskaya01} Troyanskaya O. and Cantor M. and Sherlock G. and Brown P. and Hastie T. and Tibshirani R. and Botstein D. and Altman RB. {\sl Missing value estimation methods for DNA microarrays.} Bioinformatics. 2001 Jun;17(6):520-525. \bibitem{feten05} Feten G. and Almoy T. and Aastveit A.H. {\sl Prediction of Missing Values in Microarray and Use of Mixed Models to Evaluate the Predictors.}, Stat. Appl. Genet. Mol. Biol. 2005;4(1):Article 10 \bibitem{oba03} Oba S. and Sato MA. and Takemasa I. and Monden M. and Matsubara K. and Ishii S. {\sl A Bayesian missing value estimation method for gene expression profile data.} Bioinformatics. 2003 Nov 1;19(16):2088-96. \bibitem{wold66} Wold H. {Estimation of principal components and related models by iterative least squares.} In Multivariate Analysis (Ed. P.R. Krishnaiah), Academic Press, NY, 391-420. \bibitem{kim05} Kim H. and Golub G.H. and Park H. {\sl Missing value estimation for DNA microarray gene expression data: local least squares imputation} Bioinformatics. 2005 21(2) :187-198 \end{thebibliography} \end{document} pcaMethods/inst/doc/pcaMethods.pdf0000644000175200017520000176300414710312420020174 0ustar00biocbuildbiocbuild%PDF-1.5 %ÐÔÅØ 39 0 obj << /Length 2634 /Filter /FlateDecode >> stream xÚµ]Û6ò½¿ÂZ\̈"¥÷äîÚ²hM¦h%®ÍF– QÞÍþû›áP²ä(Yôz÷b‘Ãáœï¿ºùîù¿d¾á† ž©ÍÍÝF¤šñ"ßh­˜(øæ¦Þü–ÜìíÕVr™¯¶B'UùÖû.ŒkO+×W0)+üýt•‰¤ÜÙ«ßo~òņsVd™@ò[®X*³ÍVä,Ïs"ÿá*I×Üõåˆ “¼Rž|jœõ*Û:ùѶ­kwýÙÖÖn‡Gm¶YÁ SÀ’*Ò~ýòýöíõ´. ”ú¡µ=íÓÂîá4XZ¿ëz¼îÇÓP®kˆP_¹®évWëÍ/2ç½í[š²?oÃ]CœîqòèÃù¹fúò9Ýàëò_ôƒíDé‘Ä­–â6’å X *ûa8¾xþüÖu®~ÂÀéʳÃÑŽÛ«­âÉ®k0ß±Ú>_Ó#S°LJP#ɤQD÷]5t· ”&±ñ•¢xFl©PD&ãPA©tTGÐC.á™2g&êþîþJò8È“þÞÙ‡x³ýÒBãÆŽæ,çÅÙ@ˆ¼˜!‰Ī7à£Ù€Å|I/Ë@uù„Õÿ*Ë‚!m¥.’ßøï48ö2üÞÕÁ8RÒÇ[§I“û˜Ji{Û"¹¸xýú%­³{°¤A×ý3R9£Yž‚‘ÃéNC·ƒÛ#g•,’7ìi4 ÈÁ|=̓Æá ê;a÷W8‡ÆÕႃ‹ÎŸ`ÍÏÖðþ6úôøRžœšÁƒñÍGŽ£ æ·éé¯kBì¬íµm»ÁÎnD% Є¥/ín½í)´…D„üÌ—€ß}ÖËu s¶Ë•G¤`ň1DfÆ~ ƘÃAkÕRætS’Ò3Ò÷pkmV)æ TaAQü Š/ð£écÆÏ÷VÔGdÌH³8îÓŠ¨ÀP ÏHW”N0N1å\[bT ŽP=D߯(‹ÒùÈçëuY™Œ?%«É W›±|Ê+ª5Š ÙÉ“’2RPÇ5AÆцŸÆQˆ6XÚ¦ð>“–È3L4ZË( …ƒqï>ãXOYQ9ÈCRë˜b‡½#ÈBœ§ÙJ®€à˜˜13:L .é­Dðª#ïÕ¨­bD¦kŽ €«Þ/´ÊÈÔ|;6`Q7b”½3‚˜'ƒŽºꌘÊ#S&Gg·|;&ÚªÈΉ¶§ù”Îã$æ”4‰98æ(´6âiŠ¦4…(cFœ¶4!û‰;Ǻ/ÌbnÚ›ì×óç³o˜wSÒ-Œ†Ì O] ÈàG ³,ã±*ÿ÷Z P½üLh-Ò,ƒ+;„‚ýÉâÿA—12µˆ kyI>×½§R¡ì©¨­Èeq&fô"KÞQ‹°PÍ%æØ"71¡E™E²öTÙ¿£ú‘ħ2K&.Jø`ë:;”8¡²Nh…¥2ê0ƒÑ£{€µÖ¸ ´=Üjaµ,¿pÓTâ<#UX–8°œK×Gż¬ªBµŠœ<0;!€ãxpMC£Xô€][L` Az[z(CZƒ¸`\Ø6 EwG@z­’[X‘»0žÔ7½¤íºXx¤êOà:uåB9U˜”ûä਄¼¨ÌPî¡U%BWæ168Ó± çUuaº;aÍÀU?œ+;@pí,(DOߪ<¢xjBý®M¢ïïý°Â– ϾžXäSо¨÷01%¬nì Þ³ÀTµ¦±ASà•›*–æ8¥Î®_Ä>\Œ±oE¢`¢B‰‹”ä2Ž˜l’(õF€dk-6”æ7ª-V®“Zõ‡ÊÖ,¬r-þ!| Ó0žb Fá©b[Z{ÑøH¹yâ£Å÷ô-±kÐ(8Á£'έ0$ÇDÆ|“! âbbHízˆd#é{WŽ=äa©^Zòû_þîL‘\³,ãÆ?Wào<ª½NÇÒ=K÷ÉÊt.â3ìa¡dä¤ÂŽs_.‹™O…=›meͳÖÒý¬ÁD} 8Ž[IXÊch³æÉgw˜:“z–Èì|-œt·C ÊPÞí íäG”;Zü¢-‹‹½Ý3 9X<-[t3±ï™¥_ë{†^”*ôET&‰ïÀvqM×.ñ<àL gM ç€6õ‡é’*›8îað‘BáUuêAÅ¢—»¸Ü9@-R2ê@ÄwÍéÜÄ¡¯Zö[Žqêë=Η ¼û´CÍ“ÃntàEz²bôŨn0úâé+2´{_A5¼ÃÆ؆G‹¬cB‡pl~â‘]C›ÇþÔtz^þÕ\­Î}QÇ®›ÊçrKô©š“Â? 0‰!‡3º ^®Lýe•/úË0uÓa«=ЛØq“±±n’‡8§6àÔ€èJè'M ¾¹$¾LÙå9ró$Ê]Ljüó°`ynfåÞwÿ¼ùî?Vða1 endstream endobj 65 0 obj << /Length 2590 /Filter /FlateDecode >> stream xÚËnÛHòî¯Ðe ˆi²Ù|íad72ÀÙĘK’C‹jKó¡mRvrÙoßz5E*ô$@bVW?ªºÞÕzss÷.7±©ÍýÃ&V:,"µÉ£,ŒU¹¹ßo>ñö6Ž£(xÝzWÇvØ~½ÿýî]’oÊ°ÌT†;£Í­ŠÃ"–=¯›f{«Ê$híxì,‚ýÀgdzë6ü©ú¶í ¥ƒª1ƒ¬¬LÓØ=KÕŒX\$aQ@”ˆ*óÑ®1•èP•©_f†+¢ÝØMÝY¸aàýÐ 0­gx87ã"¼eÆe²™„ešòÑ÷Ç×qÎÆ>ØZÁµæ[Ý"êÜ2âK”¤ö[½«›zDüwFiˆ´Žup¬ ··YQ¯yêÙ™Ó‰äi¯y8wÕX£øpþE™EKŽAf_¢4‚ÿñŠàTê,ó‚û%Iœ\Ü>Õ{  Œš‘眭lý„Óvðs–§öv¨ï©nU|Çäå…ýƒOת Œ•á0ºº;„r±dÃw!£½náÈPEšyžöà¨UC &E±Ô\ðW!A^ Z-Ív§ÇsËLJpœÄKî£,I6YXæIKò0Ój“„Yœñ¢håæ`¸·³Ukö]„ùäÃhº½qkª×*LòÉ NÜí´æ-i¨‹i™¿-™Ÿ Þ|í½mDÍNä†zfÏH²P-­L\g®Q˜Sªû%‡Î.Ýﶷiüe+1Br ð†f\8R~†è%fè8”Íl†3»=™ ÷>nÓ40»n{(6JÄö’ÉöÞ·§óhÕü’2 êöÔØÖrè³À§?ÿ]ó‰44>Ë–§èŸl±ý@îS;{€ï·Ü·g ü1Ýwƒíä’„á…väóLCŠ½ŠŸ“¯`Zk`ž$ C;ïØqp¬»=#G¿nÎ7nd¾ç/¨¢nÍh¯vµõ0€ïóà‰8nÎVöÿåOÁܸ5¶Á$ðŠ»º3ìÄI¦XçYìmøÊ A·Qš{k{\Ñ&8{ûmO¦Ç xÿúÐÕèu%逧l}°üóæêYhˆã<Ì3Š ©Ï½?Õí Ž ² ô‘ªRô‘-å ·|Fú½{x¶%ßÁÂ[ŸÔr°ÉL$&gGoÐ~¢28ó=ꇅèðä%G·âqÝ]-ðzÝóp蛳¨V=@ÂxBŒ§AîÈLŸy‰:Öa²ö‡ƒçûf®)û-».8»N1wØCœX˜©Æh‹ÕÙ9ëG;&ãÄ‘q~iªs×£‰N€óàQ`Þ+,îíh]KµÇ”ætÐÙg{r|=(ÑJ­—¼Æ‚©Èƒþ„+ˆýüƒÅ#}¿ý9EÐÇ„=€\ÙÝplhD§®"­`B «Í®‘ °Æ¦9€ÎÈ1Rý‚Ÿ¥°\ý½›EJ]¹œ6¬¸Ø\·Èñ€¾¥Áûžd£„ÍS á¡‘Í>€Ö$²•Ä ÅzŠ0ó×jVŽ/9Šbœ/Aì•ø[Y„©J®ümU8-tZ\Ó‹fQrbžÌ Hr?rcI‘/JŠ™t–²d’X8 Œ©B‚…]?2Bl ç£íÚ[ /`ÛdF‹#½}±l›T_\Gj.»@³YUûŒ¸êçó¼w¬m)§¤Å”"°ÙX7³•Ë2gÁ‡çñpÇúpdÈ´ýy¾‹Ò|/™ {3†°°þç»qôÜ«•D´¬_?¸~g°~8®õ[ðá_¯Àó±6…|GU|üó:ãê¸4Ò|\%³8ŠÆ)ôzü¾ýƒW@%ìú)¤â UZ8%§øÚŠV3â´ ¹"VúBÛ`=‘e’¿tDt_ÑÅr'˜Ù.qE‰DûzÐ?žÛ“dš 熙i ìË™P±FÛ5[ªd œà4}͹ïMž·”¯šeÇ2]_S%àC¼›ðŽ65G5v6;0áýË¡þ=õ٥Ƨ‰{>¶™–ýÅ`¨ 5ˆJåŸ0Uð‡³[¹¾ŠT˜S {:­¾™©$,ÔÔ•«O#LšðMdQ† úåe’Ÿi NîÃœ,ºÚ Ú‰‘ßÍ#¾î$iü¹[·n·M1“âŸÇ•Ö9šz^«¿ïœ3ß9sÄŒ¢P'é•û7`¿;*”õÑʃ¡¡_Dí Ž¶91„£yáñÈ»À¼ð{c¶à*ß·ôR›n¥öÛýbí—‹-ÖmÝP€•xÏßeuF( ñŠ ü7fz$d~*û¥#ðSè64ö[N%Ñ;€+Ûeöá_žã¤ûý‘.N^êÄy´×—Öû*ÎûJ\-adJ£³wBí—¹Èݹý;5W„©Is7oïoþ{ƒÖmâéW’l6õ¦jo>6{˜„;†IYlžii»‰“<Ôö®ÍæÓÍnÞÐo.ÙÚž<—ã4,JüF…Q¹xõCÌa]³íÆ`Ÿ*,âd+1ã;Cã.MG0kpÊ9Ú?*ÐôfAÐÜÁ!Ķ†¨mqU«sTmSsÍaøê½uRGÐ[@éßñ©4TXoñq>õ1.Àìå1€á6¥—–¹ß ÖñÁ˜5ïC éÅaT‘ z`½p|è™×„§xÑqOÿ¸»waÝßÍÉ჻ûZÍ„kïÖÚèÎRýèPøîJyù?‹ûœ endstream endobj 88 0 obj << /Length 2058 /Filter /FlateDecode >> stream xÚuXÝÛ8Ÿ¿"/ 8@ã±üÛûp@»mq{¸Ûëîö¥íƒâh&ºú#+ÙûëïGRv’Ö‡FIS$Å/Emü©Jó¸JÒM‘¨8-÷›¦»ûó..ê}©˜ã d’ ßÄýÏÚ¼î~ÅßLÚÍ2wWBß<ÜÝ¿ÏÊM×ût¿yx\NÎË:ΊzópÜ|ŒÞۧəí.+’Hý¸Ýåeý2¸N·‚;Z?:{˜Fs =jžíxh<‘„¼Ž>%Y¾Ý©ÈmUä ßúhû'á>·Ã¸ˆ³½¬‡V7Û´Œ¾ÄСΣˆüüðØ·S*®‹½¨;nÓ*z¦Ãv—ÖeäHg,ÀB–Ñ ò¤i÷•þÁ°ªxÝ1´‡ˆfèaåÔˆÆ*² zá5Î NÀc=ùðù‡ŸˆûµlŽƒÐÃùý0¼!&rOkW!v‘KSbë½e_щ¼£iÉ/y ¿uòBÇúvþÂh/7JèGYOƒø7ì>Ó¶²ëM@‡‹DjŽßžpÑDöÖ¯Ù!ìIÆq£û&ìÝÐ òèìlßØ3Çð~:ø³n ™™åÑ[Q„ïî•åiw+8¿œ|˜Ú/±õ¤ [QQ † ‡–݃›v ‹%ì8„•%g·*g•e÷¼ˆ¸ :¢i!´Ë-w‰ºÑ½à8H%ÏàâYãLÅU4~`UŠ4ÒíÓàl½1µ.Ô¨3²5N‹ ¯ÙR¨ÑêÑŠ²BÀìÓI„èn˜zJË1Hg—‚ÐYïCB§ÿ׿Ÿ’"ùY¦"•ü€â ΣߛÁQŠE÷GBèá0jÛ›@ ¥&K£7Z²ƒ“Ìê^è~z-€oIwR˜æhΙq²ãxY[¹öªº>¹ªç“Çu]îUœ‚‹ã`Ä=œ¬—oe­¾©„y# W\ËÊHÆ£ 1HÔC‰]ÔÆê–ͬÓè‘ —›Áæë¶Ø#V&#[ƒ@ïtˆ7ìÅ÷XÓšƒ72oö° ûµPgõb}u/e¸—Åj®oG)Z*‘?Þ¼§Øx%˜àðÇMã@Ú6B@¤9ZÊ PG3×Ù~6 ‘À׿½ ÁFL?÷²’Ï_­õ°“ èRF½€œu%&×K™%…€ônƒ§kI ú3”©ã0*WÅulƒÜ^GÆ>™TážÇóɬz°º³æ–—g¢$­TiÚ1¶ºô#ÚPTˉO¹†pUChË­«äÖE[n] ²šðÅà'O»5Í(–ÎU$ÉôsÆ`˜Ä”ÙЧ΋èïÃÒÓ)ü©ÛÂkE®Ä—i(Nƒóò9yšÖ@MB2tÖ!3œìQ½Žëé®Éº2ÃüÑ´“Ì&´¥‚’—976v¶«$»ˆ£ËŒãx£m§©µÄúpÆY®ê^„…÷Ôܪ=Ï6/«C@ë)ì’rž`ùÈ]­ -4ºê*(¥Í–W>/ƒCÉçDù”¤‰3ºÅª$©%Abué¼F†2 Þ]Åe±Äåå„¡ñk©Ôj÷Ä¥6M$†Ñ½æœŠeNý?,EÎpoœRZ…2 &¡´åàÈÒŒ”—¢ ºeu~)LÃ=ëúx1«èËræÊØþ‘¦_9=§¹bU7ÍäpTps*BíáB)(4s#Ì—|§ )ù<¤0ƒí,»›E÷GÐÖ¯FuXîûèµ—aBÏcÇÊÔŽ¯ÂÂ)Ä]0ÌK`fÿb}ÔÍȉt™oò9^esiØÜ”%:Ó…Iˆ‹ËlÄãôZLûy<’ž×G«K•‹4±LgËWºe7­ ´›Ö² çÚîìr‰œw”±Jåqç·š†ž¬IV|FPÏ顨´…ð¿<§Ö”uŒ`C°~,>sÁ-ÃÐTgPQ8lƒ }É1ÛêCøøM¤¨Â±~#¶ùI½[3b6;Lœg> eû: Ä£qÿ¾H¯Þ‡û*Þ×.P†ãs£…éæ‰7dšíg¦éz‡q•ñ •,£bÈåý«=Ê,G‰+Åôêj³¥Â‹Vùf×eVñ«µ*ã$É6Y¼WÁÈlE¯<Î÷5†å W„eqPJl¸Hp&{e\U#fïôÏ8.'%Åx¿k)·0 òs¸´ç0Á†F_þ‰¨¡acE%Ê„*Æ\'‡0+}¹>¦ŸàëÂúýH¤Ÿì‚;ðÊåF(yüÐ÷ÉM£1uæ2°ƒ‚*ì‚q6 ÑñÖgK7k‚æÅU3å·°/¸©èÝ_º;·\Ò[`óÙ“`E±ê[ÅÒ¹–˜öOX:LáŽ~àvh‡§—µ6ƒ.V@$ƒéÿûBïø¼@v@9L‹^(,ŸÛŸdp Â]l Ì^D,¡ÿÀ)^úAQÎ*#nï*2~½Pš¨5s®§WE# 7 ‘g¸Bó}iÚ€ä‹Æz´““ºHHßàÉ·º2” ó;¤´ŸÏ“^3T5¡åÝ»‡«_”æ߃T–Ä*ËéG£Ÿ“ÍDØcüÝ<3k–2Ϋp»ùýîWùq ·Q MKùq)+âª&¡Y\¨›<-ªM ¾DIž"?wi\©L˜Nãxþñþþ¤Ÿµµ±¶~Ümw{ôBít¬›ø?çûÊwoR#„á~†Ö߯dœ*  jfœÎ¿pe³åÿ7Ÿ'ç endstream endobj 60 0 obj << /Type /XObject /Subtype /Form /FormType 1 /PTEX.FileName (/home/biocbuild/bbs-3.20-bioc/tmpdir/RtmpCCJHlc/Rbuild86c017e3fbf7/pcaMethods/vignettes/pcaMethods-002.pdf) /PTEX.PageNumber 1 /PTEX.InfoDict 92 0 R /BBox [0 0 576 360] /Resources << /ProcSet [ /PDF /Text ] /Font << /F2 93 0 R>> /ExtGState << >>/ColorSpace << /sRGB 94 0 R >>>> /Length 6611 /Filter /FlateDecode >> stream xœ•\Oï-·mÝ¿Oq—í"7ú?šm‚¶@€pm ‹"«×E Í¢_¿"Ï9”Æ~6â…ŸArÅ™+QIýòë¯üúÓë¾|·þûí_ÿõŸ~÷úý÷_Ò;¥ô:ÿýþ÷ÿ²¸Wýß—ÿã+½þãK~ýaý÷§/Ùxýó—‘ßóz•ùžéõyµ9Þ9“üóëõýä¤ò{PœÛx_%ä"÷¥æwñ€Èý@ÍãÝr< r?ÐR~—­AäñÀiô~‡ýÀï~à7úú×o|£¿~ýË—ßþcYô‡ÿzåò¦ÿ€îý꫼ç|ýðyýÝoÚß¿~øÓ—øÁ5ÿ24×Û`[þvlY¯TŽqÓß­¹¾Ó1쯵Þ×û.ú+^¶Íò¾® ‚¶ú.Kì®_È©ZœÄOø×áÊ~,žñü±8—‡<ÿD}ÏÆ(éñÉó§…å'&–§å'FÖ§‘ç78?ïñMs|êÒí—©õ&SýÖoó‹àkƒ¿9'~œóFç_îã=ÆÏÏä_ßþõ#—à_ÿÒå ð¯ÿÜæÎnÏÅ@wõ¥·k“Y—Ÿ,FsÚˆüï5Q¾ƒßzzj>ŸRy·ôúßÿ|ýÛë/çÆ0F}ÿÚÆ°_ë‰dî÷srÖ÷[®qí ÕÆ[Žr¹çQl?%åãö®cË»KÊuãipÆ´½NØq½ëÖ OclÉa—°§Ý|—µÓÎõîmMïBNOïºìIÕ¶ÅÜ—u÷¢š}"R>^‚Ÿ•ü²ÏXPЧɹl »â‰¼uƒÂÓ[rZJìi7ßeMëÞÜ­_îsr°Çæl³2÷ûÝÖ/»ü\i¢|¼Ü|H¾÷%åºñ49¡ ±k‰ºÂÓ›rÚEìi7Þ%¯§à·êÕß%8ËýØBYŸÁæX5ëòšÁ÷%ÊÆËk–·r[+ÊuãipÆýN=°kÍ­”?ͱ%‡]Ä>ì滬Õèúš~—à ŸwÅè—}÷|wÛHÁºå_Û–¯a,©¦ß%8ëÍûÆûî¡»èw‰±)§]Äžvë]ú»ÙóÝâÉY1íÄxø6ךÚ÷Äús ãݶWcÑÇ lP®+€œúîycË»ºŠïto9í"ö´›ëeMŽkøo53Ö‹8å]ÖŠ»Úûž>‡–Û"9£åÖ-‡y÷-¿ýÛ KÊuãipÖ<©%°Ë3å­žÆØ’Ã.aO»ù.k+î¶ö99Ã^¼”a›A÷-ÊÇ+·9cÉkzß+ÊuûÓâTÛ»ÞüÐítslÊi±§Ý|—µ†la­Ñ—ýœœ5µ»ïW6Y×Mvl»m©‘Ò^™rÈ×,.+Êuk¯tNó3Œ°¾,B·SÚ+}lÊi°»ù.+–¨Ù×P⧛)µ¸×_+Àvòê‘—îð5”ûÿKÊAxšWX0t;…§96å´‹ØÓní•—M?Û÷r–†¼öˆ4ÝÛúƒ7g J{@Ï[^|Ç –tãir¦-a×qoÝ à)¦ö“ËR`vó]–¯èˆ§VÄ÷99Íüx1/jqv¨åylNƒ²ñêòLëHN]»ñØ \7ž&'[ä!욟së…‰±%§]Ž}Ú­½òò½u͑ѹW’³Þ¿ílyîn]”v?ÛÉ%_çŸ庵c8§»û¶º;“n§°WrlÈe°»åÇŠ¿ßŠyãdq.÷²¥º/i²š÷°oJ¾Æ|‘ä—waI¹nìVœŒ(|쳘88OåiŠÌOÙY+ß~6û,V²…à’¯_~l¬¨}#§ã‹qlÉyžöa7ߥWZŸ,Îú(Kà ±3Vˆù\ËNLQ>\dÈo;6–TŸ,N3Øâ>WºKødM9í"ö´›~̶µõîµÚ'úœœf§wkN¯M.'8¨*Ê}Ír`#‡Ü£åºýiq2~^b“Ç­Ò"nÕØ”Ó.bO»wœœ±†f89c}ÛÆz{×g°ƒ1;(Åɵ‰cqplP)V9 b^‚ºA)N¶±%‡]žvsŽµE4bâwgù"‹èÖLÏeÝrôZQ|ë8CS>p6$tãir¦¿+±6#·nPxºÈ³å´-%ö´[kŸ¹l¡Ñçäsuxæ¤ù™{jë-J§^ÏfP~½Ó%庿ү$L>Ë–»&×µuƒRÄs§-§]Äžvk¯œGX2Š1Œ8kè—û¢ØÆ`]»D)þó3å ç)`E¹n< ÎZô}vù©ºuƒÒyÂƦœvû°[qrÃœlvüœœåÀ‡Ç²Œ²~YhO ÖM¬ÊoAXR®O“Slûlö|”tçÈWilÊi±§Ýz—îçÔ~+w!ÎðÃ…Eé¿„eVÖèuˆ’u³…| ›6–tcãa»E¤”N6¶ä´”ØÓní/ÙcÑQã]ÄénÏrP ç³µ7Ï‚Òpß[Îô)±¤Âëdzu_º·×ÓØ”Ó.bO»õ»x`k—qrp†ŸA–¿°xEëö]yædœ\ø…¼álH,©¦898^Nl¶eº³âärÙìÃnå.Vdɧj?Ùçät‹Û-Û0‘Ž²õ·Â®ZEá[<(ys/K¬(קÉ)æ›|uI·SÊtøØË.`vëw©îÛš'á>'ǶÆÌ/6_j6ƒë- súò9-ùí>WXR9Nðâ$yÃV?¦H7(y=ÚEìi·öÊìy³êsrºe9(;¡¯Y³>aY[“çâÂœNEQ¾fñ½±¢\÷W®Eç N»œñ¡Û)Ì1ŽM9íöa·Þez°c>œ±¥8SÖßX\q-‰‚uaåÍÃ2bE];%§{ž†ØÅ[7(¼ #gÈe°»uF¾í'ó9¬8Ý—â:¹Vì/žÍ–Ð!¥So[Þl# ,©¦w NñÐKØì!¤tg½KŒ ¹ìöa÷ÃT ʵcø“u ¶8ÂÖŸ¢’æëTÄ0åµXR®;b˜Š¨dvøñQºA)†±±%‡]žvËuøƒ¹>q\‘eFm?3ïᾦE®Ï|M¹%I6VTOAÎ…~b/?PJ7(èÆØ’Ã.aO»w-Éã„55kÔ’ŒÓ±âRñ=iMíQ½ÞcùPˆâ;¢$ʇï~‚‚nUžnœìZ¬Õ'·nPxcKNK‰=íæz¹¼ršËй2885\ÈÎœ–Ûðî„;V€}Ÿ{ˇû)aIXâàð!lõÔ’tW+clÊi±§ÝxëñÌý­ØRìPuÍbÏV[jVÍöLEQlY× ð…IùÄÂ$tãir<µ$ìå©%é…§›².§¥ÄžvkíWû‘ì·º8ÇȱóTñõ™qú²\ˆUæ«(årÞòÛç¼°  O“3­$ls‡ Ý äWrÛrZJìi·bˉl¡O¿ÏÉAc%l/ÏA[EÅÅ]ò>µ-Ê+<°¢v¥Vœäb»×O¥»ïJ-Ǧ\vû°[{¥—3|µj¯$Y¢åÀmïµM»mI¢´7ÛÞ-ùôõ ,©•ZrlÅöÀ6d‰¨”"Ñ+o9ìö´[ñØePó{E5 rn[˜A¡'àÆ)Ûª  t"afÐåÙ9`E¹nEÕÌÞY–“ØåúÖ *¢·rÙìÃî]1_b•æõç ›‰~ŽE¬p#Ò´”Î~óÚrœç…%庉:§âDMlщºÞ¢âü¶œvUÖÇ¡û{ž¥,3²bçÌßEœæÝË,›—+ŠŸ8]™{¥9ÝJÈ×d+ªFRœì9Za“w.J·SÐͱ)§]Äžvë§ì»ïõryÜnÝ6gå[·¡ªÞŽÓÕ…l7å¨2Kê8]‘“¼žIìšJeëëåÞrÚEìi·~®¯=ÇÄñŽ?ÇÂSÚZß®eQÊÁÞ)äp2ŠÚsLœêmˆÂbI÷žc›rÚEìi·ò–—G§ëw¼ ó–äw+÷íQúú-=*Iç‚RœaÙÉ«%ùˆ *E…ƒœå<*v­ß´uƒRN4'qd°O»w½ÒÞÝzÕÛCNEέÅV@NˆU[ì+–ÍeË+b]`EåØ1ÄIÞŸCìZ6yë…7ÇØ”Ë.`v+××IzNèsrª*a;Ê‚Z]«Ü¢¾™AÉoÄ–Ä’Ú(q¼MØæy!én;ű%§]ÄžvïºØZ´v¦Îª#“s»¿H<Ï{úÖ¾”eF@©.æy]È­qcEÍØYÅñ!„µÒøÖ Ju1ÏPN»€}Ø-?†tþ I«ÎûäTû(î/ªïW6§[GwAªÝθœ'tbI•¨Ú‰ƒŒ±oNÝ ´³ÚØ’Ó.bO»Uãk ¿“j|Mñªí¶£«*g»ßŠs-Ž‰`ŸòÛ³ªÂ’ê-Šã)ýÀ[D¡Û)<ͱ)§]Äžvkß/žË´|šòÉäxPm¹æñ½ß¥£#±E?LåÙòé9VaIåÝCNR_ÆD=½lÝ ¾2»5¯-§]Iý ãÐí>Ù|¼w.í•ÁñN=KQ›K­Þ µ+ßxCÞ<),©Ø‚3Ô>eØâ{„t—Ø46å² Ø‡Ýêí©È=äèQÇ;õÊD½g}w[sx‡ (œà'â?Êo¯ˆ K*Gä+NÂí `­Soë…§1¶ä´‹ØÓîÝk¿£ÝŸ¢v¢±aýˆvÆ°hp kJ"PšÓ6k$G",)קÉiG [=*‘n§t6ò±)§]Äžvë\y£Ë6[÷sr<UJòÞëš^Ó.—(õF\}˺>ˆ%•£K—œµûÙéŠØ5]ÆÖ OclÊi±»÷¹’=QŸ:)²®=Ñ…b5oƒ6Qúv³l¹·¬ +ªD&R?B »†¸·nPŠÁg¹ìʪ§C÷÷ü®7ÎÊ)‰ã¡žw©d|›ê,~j˜‘S²”ËÜò[çÇ’ê;J «KX¬.é®;JàØ”Ó.bO»õ.·Ÿ¬ª¡w!§ùUÙ³ÛQ›Bg(ùÍV·¼!³B,©=$âdß»…Mž÷îy ¹ìöa7}rÁj´ôãdqn8ÌËÛa×ÔugŠó(÷›R”Wô+êÒNœ!GoØÜà^¡”?ͱ%§]À>ìÞ±%ó,sDl9Зk•.‹ïÐÓk—%–öz‰Ò›#ä˺´±¢Zt·‰S½C_Øâ•.év*bË{Ëi±§Ý»Æ×Ñ­ž+q.t ÷n ?lEAeˆRÌtµ-o–ä ,©Ý?&Ž—ú‹þ0éÞýcrÙìÃní/—ÚüM?&ŽÏD‹™ì;½6`]R¶ϨBXÕU·{bEE"8Þl%ì:kÝ[7(ÅzÞƒ ¹ìöa·ê/žb´Hº¨þBÎ¥‹/ŽÔ‰ÄBõ©[c<¡5£p¯‡Å»êû•5²wW÷¥Õlûä –ÔŒUµÖ–­žÆØ’Ó.bO»µ¿ Ï&Œñ˜8Õcªå’ôß0*V<¶~äÒB¾^©m¬¨ñ˜8 yͤë¥ìs¹¶œvû°[=ðϘcÁi~¹á/ŸCåæÉ,æ˜÷¢ç-¿°cK*æXp<±XdU¥;æXŒM9í"ö´{ûdv9¶>y¢¯Q7¦×9t›`ˆRöξäÍyaI¹nùdç TÐmÝ3÷Ô Jùª4C.»€}ؽëû~§ÐœŸ“Óíf“í·¼G¸¢ÛaìbôÝJ>Ð;A,©^VdV„Í~Ú‘î'›rÚEìi·Ö>î ô¸—äÖúd¯†•°ÖQ8ߢ°>Y‡ƒ|¹œ¹±¢â^RpPÔ¶ íLÝq/)Ʀœv{Ú½ûÆ-ÿŒEù99èpÎî(ÔáœodEÂ~–‘9…¼Ôá€åºñ48ËÏÙìÞ9%ÞŽÓ+q†gØÖœNè·´ ›Í÷*JY¢T¶¼ëæžc»níõ}R,Ø,ª6#U;ꪱ!—]À>ìÖ»4Ý=Õþ"NÅÚï^ûÇ…œoßûËTÎ)#›–6VT‰ýEœ„õ;•3º¶nPò÷sn9íJêþ‡ní/v†h-ú-Å°¾âFÀ…J—¤´cäºåê3#ÞY+êŠJ—8˜ƒÄ^ wš VÆ–œvû°[¿ òÓ¨Ñ}NÎðjjF6ÛöŠ¬h¥C»'òqmcEÁªH õEϹtƆ\vû°û¨½"v¬y×^Ù u£þ‰ü}éŒ^PQ{­[^ÕFvÝ›êq“U<[¼kVºŠÚë¹ìöa·ÎÈè€DÔñ99—w8—‰âÇíÎ冽#ò­È!P^‹nxùÝR#"_qî‹Š½t;¥Ü[é[N»€}ؽ{:êKÊ]ˆƒžÉŽ> ý”Ý]/)õ*\mËo?O Kj÷[’3Ð3I,.ðI÷Øý–[rØ%ìi·~—¤*¡âdqšßÿYsÚânë `RqòÚ1Rßò¡wu,©²³äxJ?°Éã`éN'klÊi±§Ý»ç*ã®NÔ+Éi¸ƒ?Ñ]ÝQ¼µÛ¨FÒpc’rk÷ØXQ%ê•âdä5‰M¨GRwÚõÊ©Îî2·]À>ìfî"Uô eý.Á©ömkjÑyT5¼S”çÖyË'nK*ëw N²o+¬Ý÷ܺAáiV(¯êjrìi·ê• QEŽýEœŠ=¼IµzM‘ù¯Ø_¦ç%çAì­®íØ_nuHûþìZsë¥zek[N»ˆ=íÞùd¿OS­/ïsrº•}<âÃíhpÜ¢”!Î×–WÜ_VT.@qŠï?ÂfÏxIwŽ.@ ¹ìöa·ö}Ü_¶sãâ ³rMôvï™ •¥GÉÄmiÉ ºU£‡Dœì¹ a‘Y‘î5j ¹ìöa÷îU°8ÂTu~!§ãF îñdZ³(õ*”{Ë/ü•bIÕˆùÅñ¿¢Øì–tçˆù56å´‹ØÓnõ%äò<œøœü¡™;{dÑp+×Ò„SNð¨éHŽÈVXREÖá ±Ëµ­žÆØ’Ó.bO»÷«èÐ= r.ô6­ß¹ã>Ūø;x¤4ÇÚÜòâK º5Çœsù‰šXûM·nPšc=‡\–û°[¹>oW÷Ü?÷q.ü• ¬O«ILϾñ& Nð7ûÊoýŠz`‡nê/ŽÿŠ­¾7Iw½KcSN»ˆ=íÞwý®Üˆ:²8úM vã©{ƒ©‹RœÑë–T8ˆ%5¢Ž,:M…õrrè®QGÖØ”Ó.bO»÷ß»¸ãFþçä°êÛ ÊST#ñoGùÄý|bIíÛþâ ׈Å|é>nûslÉi±§ÝêSâýælÓïsrŠ;Ä5]{Hp¬_¢|}ÚE±rÛ 6VTŽ¿öDN¿Ñ¬l÷¿ô&Ý üiŽ-9;³}Ømïòí?2ûü[‚ø3±9yAòó¥¯Oë5`ï(üóRŽ5óϱþ”kcT{õãïzý¢ùÊŽºéЮ±‚ÙÍ»cM”2¢£„Üîžm¬¨+ê¦âtŸòÂ6ÔE©»íºé¥†²Þ¶]À>ìþúåwëù6QæèC÷=‚3í4ÑfS«ùÚ’ô'nA}õ§/´¹S>ÑæN,©ØƒÓ-ùXß«BwìŠ16åUmìŽ=í¶wùîËÿD|£0 endstream endobj 96 0 obj << /Alternate /DeviceRGB /N 3 /Length 2596 /Filter /FlateDecode >> stream xœ–wTSهϽ7½P’Š”ÐkhRH ½H‘.*1 JÀ"6DTpDQ‘¦2(à€£C‘±"Š…Q±ëDÔqp–Id­ß¼yïÍ›ß÷~kŸ½ÏÝgï}ÖºüƒÂLX € ¡Xáçň‹g` ðlàp³³BøF™|ØŒl™ø½º ùû*Ó?ŒÁÿŸ”¹Y"1P˜ŒçòøÙ\É8=Wœ%·Oɘ¶4MÎ0JÎ"Y‚2V“sò,[|ö™e9ó2„<ËsÎâeðäÜ'ã9¾Œ‘`çø¹2¾&cƒtI†@Æoä±|N6(’Ü.æsSdl-c’(2‚-ãyàHÉ_ðÒ/XÌÏËÅÎÌZ.$§ˆ&\S†“‹áÏÏMç‹ÅÌ07#â1Ø™YárfÏüYym²";Ø8980m-m¾(Ô]ü›’÷v–^„îDøÃöW~™ °¦eµÙú‡mi]ëP»ý‡Í`/Š²¾u}qº|^RÄâ,g+«ÜÜ\KŸk)/èïúŸC_|ÏR¾Ýïåaxó“8’t1C^7nfz¦DÄÈÎâpù 柇øþuü$¾ˆ/”ED˦L L–µ[Ȉ™B†@øŸšøÃþ¤Ù¹–‰ÚøЖX¥!@~(* {d+Ðï} ÆGùÍ‹Ñ™˜ûÏ‚þ}W¸LþÈ$ŽcGD2¸QÎìšüZ4 E@ê@èÀ¶À¸àA(ˆq`1à‚D €µ ”‚­`'¨u 4ƒ6ptcà48.Ë`ÜR0ž€)ð Ì@„…ÈR‡t CȲ…XäCP”%CBH@ë R¨ª†ê¡fè[è(tº C· Qhúz#0 ¦ÁZ°l³`O8Ž„ÁÉð28.‚·À•p|î„O×àX ?§€:¢‹0ÂFB‘x$ !«¤i@Ú¤¹ŠH‘§È[EE1PL” Ê…⢖¡V¡6£ªQP¨>ÔUÔ(j õMFk¢ÍÑÎèt,:‹.FW ›Ðè³èô8úƒ¡cŒ1ŽL&³³³ÓŽ9…ÆŒa¦±X¬:ÖëŠ År°bl1¶ {{{;Ž}ƒ#âtp¶8_\¡8áú"ãEy‹.,ÖXœ¾øøÅ%œ%Gщ1‰-‰ï9¡œÎôÒ€¥µK§¸lî.îžoo’ïÊ/çO$¹&•'=JvMÞž<™âžR‘òTÀT ž§ú§Ö¥¾N MÛŸö)=&½=—‘˜qTH¦ û2µ3ó2‡³Ì³Š³¤Ëœ—í\6% 5eCÙ‹²»Å4ÙÏÔ€ÄD²^2šã–S“ó&7:÷Hžrž0o`¹ÙòMË'ò}ó¿^ZÁ]Ñ[ [°¶`t¥çÊúUЪ¥«zWë¯.Z=¾Æo͵„µik(´.,/|¹.f]O‘VÑš¢±õ~ë[‹ŠEÅ76¸l¨ÛˆÚ(Ø8¸iMKx%K­K+Jßoæn¾ø•ÍW•_}Ú’´e°Ì¡lÏVÌVáÖëÛÜ·(W.Ï/Û²½scGÉŽ—;—ì¼PaWQ·‹°K²KZ\Ù]ePµµê}uJõHWM{­fí¦Ú×»y»¯ìñØÓV§UWZ÷n¯`ïÍz¿úΣ†Š}˜}9û6F7öÍúº¹I£©´éÃ~á~éˆ}ÍŽÍÍ-š-e­p«¤uò`ÂÁËßxÓÝÆl«o§·—‡$‡›øíõÃA‡{°Ž´}gø]mµ£¤ê\Þ9Õ•Ò%íŽë>x´·Ç¥§ã{Ëï÷Ó=Vs\åx٠‰¢ŸN柜>•uêééäÓc½Kz=s­/¼oðlÐÙóç|Ïé÷ì?yÞõü± ÎŽ^d]ìºäp©sÀ~ ãû:;‡‡º/;]îž7|âŠû•ÓW½¯ž»píÒÈü‘áëQ×oÞH¸!½É»ùèVú­ç·snÏÜYs}·äžÒ½Šûš÷~4ý±]ê =>ê=:ð`Áƒ;cܱ'?eÿô~¼è!ùaÅ„ÎDó#ÛGÇ&}'/?^øxüIÖ“™§Å?+ÿ\ûÌäÙw¿xü20;5þ\ôüÓ¯›_¨¿ØÿÒîeïtØôýW¯f^—¼Qsà-ëmÿ»˜w3¹ï±ï+?˜~èùôñOŸ~÷„óû endstream endobj 105 0 obj << /Length 2852 /Filter /FlateDecode >> stream xÚ…YKsܸ¾ï¯Ð%UT•‡&>s³;q"»¶ÖJíÁöšF(ñ1!@ËÞ_Ÿ~ó½{ÐhôãkÌëÛ_^¾ÓõU›¶•ª®nï¯rU¤M¦®ê¬JsÕ^Ýî®>'f¸Þ¨&KÌÆñ~Än²u&¸o×в<<À7Oæé:OL‡¤Hžðgœ™o;öZaôvÇ|ã½ìpÉ1Øç¦w;?®U؉IOn‹ÄîzÁN,–¿þzûï«ìj“çi[–r`QE‘„»(“/™Êî Y'cÝ v‹<=ôJ‘X0åÁívvàö0›Î ÖL¼Ê"Zâ‹v.Š_Z”“qÃÂŽ'á–™„Á»þÐýÙûaœðÔØyz°²ØÖtÛ¹CS0 TÀ=òdÜhE­všF¸‡"íÍïŸ5É!ëzܦñ`ö`ñãš­ª 4‹çž=Oúxóë›WÜ$ÝÒd²*Ë™ŒÇñ‘瘃eZ/:Z‘ 4VEé,#©¶††7½´è&ĉ@Æß{<6ÀTpø¹»œˆÜFá)·nãâ〺Æ}wîK¦µì°•1ç£$&eZ9Âb{0'? 8ºI릅 Äø«pœ…¤<Íã8ÖÜOðÄÉÀqä8ä@¼cÿ¢¶‘Ivžð¨ÜF›8‹Nl¼›¢Mèìx³5øÇÃ8Nl@À¸…_¶Þ‚â >CŽ ±ÚIÀEæ#tgš°¬A&¾ 6xJêOÖx¸¥»Î®ù—# ­ ‚L.؉M´(rq~øßl:¢F”&Áq Dëƒë kºGGpžn&=‘Ü8qGz;°e;Ïìê:‹)r—“ {ÑFé2ÕesqÛâ › WmžD’áϯaâ,ZíLœ*U0 ê¸@y mU,1råhMZdUdÀͬ¼ä`ÝþᘽÆyòç{‘Õ X¬t–U‡×÷¨×¸&ðM‘³uò…ö·F3M·!Í~?HÖt½]\GïölýØ3üAŸ@KCdÚxCúÎù`®yq’똇5Ln+l!N~™îÁeÊÄ*û'¹rH,ò›ÄÏwÞnS ß%ŽRï"áÓ¦)) #~¡ö»DÀ±úñàd;º…™3H¬É!Mâ`ý³Šä×ɘøÝâá2Ç.©#,!³b|åI‚/ Øy|¼ÁXUÇYõŽ:!0‘ú ï<%ß¡3"… ²\¨ IÆË,ð$@Ç"2k²*ùØå |ÒgÏ«†vöø€ïB“¼:¹Á!ЦóÒ#Ô¡3[~ªŽµ÷ñµ¨黹d'ç…Ü[Cåj£"ìA€RF”Ë"òø„•âf-káñ¹®-èMrãäøÒה玄XôrK€ß üJ$~7iʸÔ7¸ö'îœÁzà´êº³I°19ˆöv©?«NŽeƒ•KÌ‘Ja]œˆŽä{¸ ß=hµnäåHóÐÖNÁ¸!¢dýAÖch¥Y–3Ž`À˜£ñ‚Ö#$Ö ,D±Ô q–%ÿ´!H誀SXðf{ŽpUž6y£/°vßQ)e6ñBú|¼†@ˆ7]}jÓ¦Ö·ÝbyŽÖ¢ÀcxYìÄ{WšQt(ºb­¨æ”hˆ;ÖÖ€U“OëvÙþìq5JtJ‰Ñ@ƒª)Rb$HÂC&ª°°%;©äµx ©¸ ›ÄÛt^¼®T¯ hQÙ¢š‡ùòéé)½s£¬ Â8É[Wä?;Ë8í_®¨]ei›/hYàf)à]ÀÄÏ«¿»A”‰ÞX´ ÛÄZ’UiÛ.%Èo+[‚¢— Éèý×v–ÒÓñ1÷båªN+U.oŠfû·îWÖ/*(mFIõ³ÈM1çùâe™ªbyy;fmê¬Nó¬¾ÄO´=£3¾õì« "ñwQ×­`yåçQ,¡ž~ç(&ä‚”tžC Þ?0© s‰£/0-ÐSpLÎ¥ -<$Žˆ5KVàÝÏÿD‰ñhÛ9ñ-O?ÑXÓ,äx°Ã_Z—˜g•§—;/ƦkÜ8¾±ï/ÿ;¸ïX?4üúF£|.hÌ~ŽÿPÕŽµô=ža­.Ö©Îò£UÒͼùð×?ݾ‚u ¬>³Um}ònG»ÐÁMy°êXÕàÝ–"ìYf{óÛ«kÓi­«ÓG=ª¡”8æ4ú¤¦¥dÏÿq`Ûs$i$µ`ƒôv’/õŸhIå:UGÏ‘°›~îj ¸Z³L ?`Ÿc——Œ>æ)³¦¨¢H•ÿYèhÒ²Xb‹ç$-x,/!î”ôÐß–s8ôËÛÛ_þô„²( endstream endobj 120 0 obj << /Length 2162 /Filter /FlateDecode >> stream xÚ­Y[sÛº~ϯÐx¦3Ԍůº3'vãRw0 3ÆEìoù2Á§eˆ²P¯º¶éõð½í@Ž¹t¯'”XȃHOÍ[—' ¤U2B()PVP nrD{XYÄ Ž ~£­miÖÚÉIä\ìÄ""ÅÝ£ûIµ²ÔŠ§ÕšÔ…œèÒ Š ªTŒŽÑDb ”s -®=QÈ8}G;¾9G´Í÷½íæ2:"¸&"[FÎxLÿëP¶Ì¶±½i€“×[ÌDª8•¨H„¨U΀TBEÁCçOS³SÎBnÌE~ú&!_z“Nlq./rM«¼MQ>¡–|•÷ +Z>9Žjž4:Sd¡’ì¤?BÞ–ÉËWp;~(ܱ‘oùJ€g„ Üïˆø_»e˜_H,gÂêÈ:+ãlœˆ.qnCWLÝûsåo½ÞÛ{£èuÉé¾YŒ«Ô£@Ran7”ï-MKZÜ·]ÏMÃ#Q¨ ‚nÖô( ¥_Ô›CÕ×£’SU”EžÙþÆï#Döw*ß!¸q¤˜Âv(úŒ^Uœr`È" .¦ I0EäKôsaÏ›Ó×ôØ1÷Ø·¦³–šìoóÜÛ65AôåË|Öjÿ"§~"ŽÃäø»WOø‡£G ¼ñ¶E¨0k\A²³j2µÙ! qSE‚ã"?},8šK¢Î˜Òç´šüÙÒ”ò48mvô-éã'þ×éi±xÓø5öœ_…¹vÓ5•>eè¬pØNÇÖÐkÅ‚Cüú`G«…ww,Š3«Ñî†×ö.iƒ?“/ |¡£}ÊéßHqQ@;u«·ü«€¥„›Ф%X}…yÎðIpéÍû囓cr‘ endstream endobj 130 0 obj << /Length 340 /Filter /FlateDecode >> stream xÚu’MO„0†ïüŠË¡ÝN¿(Þ\ÝM44á¦* K‚BÑ¿oÙÂJŒ›:}gú¶ó´€˜€€KjGR3j€£ü}F4I•NN«ð”‚yß,lîŽÝvÑ“KŠ,ždeºÍ¢Í^$(¥©æeï§ Ia¨IÊ ôŒ÷u5öeL„˜_ÅDJ…wuU¶17ø+VÛf<å%\?æî\n‡0woÎÖmY„¢ïÚ‚^Ô/Lˆ²^.ˆÇ˜.Ý¡‹ Op1įÙ}´Ëν,­€aTp5ãÑRðü†ðˆËxV¦ÿàYNV†JfþàÑ G±?L-t¶¨ÛjJH !oÛ"> /ExtGState << >>/ColorSpace << /sRGB 134 0 R >>>> /Length 874 /Filter /FlateDecode >> stream xœ­VKSÛ0¾ûW쑺ծ–Ž%¥aZ†6™öÀpÈ@hÄg ýû]Y²#JpÚÆ_´OŸÖ»"8‚K¸«¾ÀØ€Ê@­ÑІ±f kÑ2ÜÏá;\WoW_?îÃxR)TJAùœŒ*F¶ð»:9çÁ¡ü]V àsUÖ˜5²‚«Š,¡ª8 u°¬X+4 HYôk¢=)ƒ¡ŽØ‡häW8‰Œ·RXo%”¶{9~‚K€ÉÆzBÎ!çå¼v=àï¨5È°0È[ì$X”œ:Š…Áþ4Kp¶zF‚Õ™Ù9D˜^ˆÐ˜Ò#aç#'SÇCž^Á`zYL›È/»’x ݵ/¿Þ—¹Få _ýz_­\Ÿ³àëmŸ³m}s퇦8ó—ÀäQÛ¤ôÆ:¡oQ±LJ¶¥»õ Kmbuw –ŽÑ…µA‚¥AP‘{g`aI·åJ} QÞZ©rµŠ´¢>òãsê¾è|䜼í`ob×ãÉÉš¸ßä¾CvÏH>•óÙ9öÊ•¾âʨ¤qj4-döbß˧ØŸ:Ý?éj A6Ûº<Ù;Xü˜_@Jxï—ü³°7[>ÎW£S˜¾r×ò*,½AçSÌÉÃìú|vçó‘|©y1{XÜ\ÃÍ»ÐC§—>®Ps|îOp±Ë Ü6xÈn žº?xŒížº?xŒnñqÿž\$Etš¬â¢›”´ÅÀEv-7Ùµ(é‹ì‚ë°Îž[M“´–$VÂ) Šÿ’=ɲɣQÁ÷Upº¯‚5ã×öTp­ $ò‰ >ªà…|i´Oe`’R¡2½Gî¥oÖ‹ôuy©Îé½TìÖD1j1zzI9£Ö¥ Ô;vYõEÂÐüØåóµÉù¶_0Xb{n¿]’ˆÁÈç¢"W)ê*Þ zµ’öÎ*6N͹ÃMÏXûmDÉIeÄ„sÄ!å›RœEÀŒsÀÝ*² -ǤËÐ ·¡‡YѺxëˆ ·z¾P7м¬Em^û±¥S•g•ñÿ¹i銛)Âi†ß ¸Ý$w r…Îî;ù{Š0úïïäïäÚ“ü'ßÞ/®næÃX‘;9ZÜΖ+؉Š´Ù¬äѧ"À—êÿ‘™³ endstream endobj 136 0 obj << /Alternate /DeviceRGB /N 3 /Length 2596 /Filter /FlateDecode >> stream xœ–wTSهϽ7½P’Š”ÐkhRH ½H‘.*1 JÀ"6DTpDQ‘¦2(à€£C‘±"Š…Q±ëDÔqp–Id­ß¼yïÍ›ß÷~kŸ½ÏÝgï}ÖºüƒÂLX € ¡Xáçň‹g` ðlàp³³BøF™|ØŒl™ø½º ùû*Ó?ŒÁÿŸ”¹Y"1P˜ŒçòøÙ\É8=Wœ%·Oɘ¶4MÎ0JÎ"Y‚2V“sò,[|ö™e9ó2„<ËsÎâeðäÜ'ã9¾Œ‘`çø¹2¾&cƒtI†@Æoä±|N6(’Ü.æsSdl-c’(2‚-ãyàHÉ_ðÒ/XÌÏËÅÎÌZ.$§ˆ&\S†“‹áÏÏMç‹ÅÌ07#â1Ø™YárfÏüYym²";Ø8980m-m¾(Ô]ü›’÷v–^„îDøÃöW~™ °¦eµÙú‡mi]ëP»ý‡Í`/Š²¾u}qº|^RÄâ,g+«ÜÜ\KŸk)/èïúŸC_|ÏR¾Ýïåaxó“8’t1C^7nfz¦DÄÈÎâpù 柇øþuü$¾ˆ/”ED˦L L–µ[Ȉ™B†@øŸšøÃþ¤Ù¹–‰ÚøЖX¥!@~(* {d+Ðï} ÆGùÍ‹Ñ™˜ûÏ‚þ}W¸LþÈ$ŽcGD2¸QÎìšüZ4 E@ê@èÀ¶À¸àA(ˆq`1à‚D €µ ”‚­`'¨u 4ƒ6ptcà48.Ë`ÜR0ž€)ð Ì@„…ÈR‡t CȲ…XäCP”%CBH@ë R¨ª†ê¡fè[è(tº C· Qhúz#0 ¦ÁZ°l³`O8Ž„ÁÉð28.‚·À•p|î„O×àX ?§€:¢‹0ÂFB‘x$ !«¤i@Ú¤¹ŠH‘§È[EE1PL” Ê…⢖¡V¡6£ªQP¨>ÔUÔ(j õMFk¢ÍÑÎèt,:‹.FW ›Ðè³èô8úƒ¡cŒ1ŽL&³³³ÓŽ9…ÆŒa¦±X¬:ÖëŠ År°bl1¶ {{{;Ž}ƒ#âtp¶8_\¡8áú"ãEy‹.,ÖXœ¾øøÅ%œ%Gщ1‰-‰ï9¡œÎôÒ€¥µK§¸lî.îžoo’ïÊ/çO$¹&•'=JvMÞž<™âžR‘òTÀT ž§ú§Ö¥¾N MÛŸö)=&½=—‘˜qTH¦ û2µ3ó2‡³Ì³Š³¤Ëœ—í\6% 5eCÙ‹²»Å4ÙÏÔ€ÄD²^2šã–S“ó&7:÷Hžrž0o`¹ÙòMË'ò}ó¿^ZÁ]Ñ[ [°¶`t¥çÊúUЪ¥«zWë¯.Z=¾Æo͵„µik(´.,/|¹.f]O‘VÑš¢±õ~ë[‹ŠEÅ76¸l¨ÛˆÚ(Ø8¸iMKx%K­K+Jßoæn¾ø•ÍW•_}Ú’´e°Ì¡lÏVÌVáÖëÛÜ·(W.Ï/Û²½scGÉŽ—;—ì¼PaWQ·‹°K²KZ\Ù]ePµµê}uJõHWM{­fí¦Ú×»y»¯ìñØÓV§UWZ÷n¯`ïÍz¿úΣ†Š}˜}9û6F7öÍúº¹I£©´éÃ~á~éˆ}ÍŽÍÍ-š-e­p«¤uò`ÂÁËßxÓÝÆl«o§·—‡$‡›øíõÃA‡{°Ž´}gø]mµ£¤ê\Þ9Õ•Ò%íŽë>x´·Ç¥§ã{Ëï÷Ó=Vs\åx٠‰¢ŸN柜>•uêééäÓc½Kz=s­/¼oðlÐÙóç|Ïé÷ì?yÞõü± ÎŽ^d]ìºäp©sÀ~ ãû:;‡‡º/;]îž7|âŠû•ÓW½¯ž»píÒÈü‘áëQ×oÞH¸!½É»ùèVú­ç·snÏÜYs}·äžÒ½Šûš÷~4ý±]ê =>ê=:ð`Áƒ;cܱ'?eÿô~¼è!ùaÅ„ÎDó#ÛGÇ&}'/?^øxüIÖ“™§Å?+ÿ\ûÌäÙw¿xü20;5þ\ôüÓ¯›_¨¿ØÿÒîeïtØôýW¯f^—¼Qsà-ëmÿ»˜w3¹ï±ï+?˜~èùôñOŸ~÷„óû endstream endobj 117 0 obj << /Type /XObject /Subtype /Form /FormType 1 /PTEX.FileName (/home/biocbuild/bbs-3.20-bioc/tmpdir/RtmpCCJHlc/Rbuild86c017e3fbf7/pcaMethods/vignettes/pcaMethods-008.pdf) /PTEX.PageNumber 1 /PTEX.InfoDict 137 0 R /BBox [0 0 504 288] /Resources << /ProcSet [ /PDF /Text ] /Font << /F2 138 0 R/F3 139 0 R>> /ExtGState << >>/ColorSpace << /sRGB 140 0 R >>>> /Length 6918 /Filter /FlateDecode >> stream xœµ\KÏm·mß_q†í 'zí×°6ÚA[4µ Š‚/}Äðqázп_q=¸u ;ÍE’Á½HjQ”6%QuêãWúøæñý§_?¾l׳ŒÇÑŸc<êÞžG{Ôm{níñ?ÿþøÍã»O¿üá_þþ‹Ç—_}*ÏRÊcýÿ«/ÿiríñ¿Ÿþõßåñ»Oõñ«ùï›O5 <þñÓãQ¯óyÕǵ=K}¼Æþ×£•ò¬ýqÏ£N¢>÷*âcm¥?kIéxöGZY”Œ–ׂe­ X”u‚ak€{³õãÓW°öµç.óÉ8Ÿ}vØuAÕõÜvèŠ@°¢3$í08Ðê–²=ýHÜxÖ[+ˆÀ:Á°5À½ÙJó[Ùžu{\å96˜oF}^aÆþì{4~²Z9žç&‚6]ÏÞ-_tKœhEQ2ÎÙ}‡qçž!­$¨uZ kŒ[l•ùµ?K׳šOÆ9ûäœÄx¶‰ž6•IlÏ£‹ Mdzµ”žÏqã@P+‹’A?#îØŸ[j%Á¢¨ÓRš'Üb«{W·´¡Þ'ã ¥Ñ5Çíŧ¸B/ wé,$i­HœheQ2öP'Ü9‡pj%ÁÞG–ÒâV[mþlë&ì]æ“1ݳâsÍ>˜îÑ;Ð×%‚Í.=Rz=÷GZíg“1W‡qÓ=®ÔJâCÞ;ë´Ö·Øºôþ-:Ï»÷Ñá}gOœáÐÛÉ~ÚEdïo–Fÿ&ŽµfïáûeÜìÒšZIdïŸ)¥yÄ­¶jæ™sòdÕé›»¦qZ‰qRñ<æ o5̪Ó#Ûn óD¸å¸ågX™XRÔÍÒâa›±Óû¯[7)–fÝ–ËRaW»Õ–ýÂÂÒÚs4¶ÅœópóÚæ 0™Š‰úPËçÌ–rþ1V@,-N 71¶²Bé&ÅÒ¬[rÛEì›Ýþ.#.Ê÷¡ïBÎÄœµG‡OÌ\X£ïÎaÊ==ÇDʯŠ‰%EÝ,-Γº±uî·nRnyÔm¹,vµ[m9dz\:]ÓêlÎløìÛ9QÄ"9½:º0fŠËê‹éâ¼åW¬Í‰Ý,-Î3”±1¤oݤXšu[.»„]í¾ÇK,‚³}uËñœÙ·ê9˜c<œú®¦<^æ˜ò“£KXRÔí¯ÎñÜÄÎ~·nR/Q·å²TØÕn·¥ÄD€±[ÔqZÌ áÅÑþ6"̈;MÙÊ~Ëwô´±¢ ›¥É™œ5±1~oݤXšu[N»Œ]ív[¦>¶ýêj‹8sª¯¨} {çÌsšòhGÊç>n¬©Ý+Dr6´ÆÎÅaÑ Ê-GÝ’Ë.aW»=^èÓï6}q"àÅÙùåk¬¢á¥ðŠê•:¼<½åG¬Î‰%EÝ,-Îñ,[bçx8oݤXšu[.K…]íV[ÆXÿ軘sÀžÁ9NôMü9L¡¾qEÜ`ùV"†3Öt£´9öÛ0ëY7(êVÝ’Ë.aW»ï¶ÄwؼVÎFE;&‘°®¢¾¤Ü–ðɧ­Ç55¼Hδ®ÝXl3R7(·uK.»„]íV[¶AíGL–¯•3—軳xc®äaÝÑHŠÖìÉ/4ÂXQÐÍÒâìl9±}°åÔMŠ¥Y·å²KØÕn·A@í-j­œñ;jù¢o—ס )Ô77¤á·–ÌLš‚n–§b™1¶`r³nPn9ê¦Üvûf·}¬aÐÎ\™saM*Xmç'¨‚”ý`ô[~$¬¨3wyæìX#„î2nݤXšu[.»„]ív[Š½h}1§Ú™[ ãfG?ª)Ö7è\Íe,XQ%W#q¦ÄÎö[7)–fÝ–Y`W»Õ–ùïœz=öÍ)b9é¥sBìÍꛓâYRîÉx8Ô)©¥Í©¿ÆbžºKŽ}×-¹ìvµÛmi˜Õú¼VŽÂ£Ž¨c†N…¡Q¬¤Xߎ¸Ýò#ö…‰µe[Ìé´GØF[¥»ÝmQÝ’Ë.aW»Õ–Ê)g†Ó _+§a2C«Vî ­bˆ“B}3¼¯wx4C™¾`EA7K“3ö؃»Xt¤›K³nÉe—°ovßm‰Øù¨Ñ“¯•Ó:unÉböˆ`~˜b} žayGL/¬©š³¬9°°Óºrë&EdÝ’Û.bßìv[ZlªÃ}s6†N#‚êblÏ`h¿L±>mó$?1ë+êŽDÍi¿ÆV†]Ò}G¢®[rÙ%ìj÷Ým¹FͶ€³co8õ†ïµ¢¾­›r[b XÎPÆØÓAá¨Ù–¦ e¿±œÀ¬»ùð2ë–\v »Ú­¶ÌÍí±a³™£€}Çf{?èC'¶û½’·‹[VÊçBVo¬©{%7§ÓÇ„Õ¶Bº[ú˜ë–\v »Ú­¶Ì%:ÖœŠ-ÍÙ1¹w.T3ÀˆemX+AѺ“k¥ä\$Ý,-NÇ”el‹ã¸Ô Š¥U·ä²KØÕn¶%Zãœx„»¿Nqâäµãäuœ¼VQWœö–”^pcáD +3z ãZ¸ºµ6)×I©¬!nµUæW²Gùf\á½æ–8cn8é=« UÔ®”2ŒNÄáøÞ ìQˆ 3Nk¢œÏ$­îÌÀ­¶ÊüóÊyùú@Œ8%¾¢Óbš¸b¥br!õ¡Å3" Ë;>BMQ³8pÜ;Ï~ë&¥ïu3l%o&{žš»ŸÆí¨Î¿Ä™‘f‡ÊM[0kwœpéŒjÎ.Ñ3–ïŽ3€%EÝ,-NâUÎçÛ­›Ô‡fÌ:n¹,vµ[_¤Åm|±ƒ[`3NùP,µ³wü Vaò’}Kéñ~ø‰>úáã»O¿ü»¹µz|ýŠãŸ’ÿ‘Þ‘ 6pñõëñW¿˜kÂ_?¾þæÓß~ åSÌãÞ?Îô7x5œ‰¦3r˜+ßKi§Ølúä5z†Ô"Ž£ü9Ü-¹Øzxd ¹ˆx» \ È&XM\{`iö/–Þ5&þFÙÁü©üƒð¸›Ûâëçã#nΟķòþä|Ÿ°ªî@&[CØ·î “ÿõÀ^>yÛPm;±CS²ðã¿].4~äBÌU‰;´YïünsžŠüÃÿöw¿ÿî?õ3ü°ñ"øŒ˜:t|ñÏ_þÍgôblùw\¬Ì¨)ðßþö‡~ÿñXÔDgtüå’§GaÔ½Ã_+§Ç= ³Ùb“bB©”ÞpûbùŽÛ{cEUg$§Ä…’±ò+­›K³nËe—°«Ý\¼‚§kùãnË um‰UFܣ𖘔Û²–ï±Æ&VTóñar{%¶Ä÷JÝÅ'ˆY·ä²KØÕn¶¥_ØnGù“;°ä4ö-öæÑÓ[‡ö8%Åúº¿äGäÆŠ*>07g»Ø·ÄngÌlÖMŠ¥Y79¶‹Øw»Ý–ŽÁ2.ߘ3'À÷k0A¢E°Ð¯ 3©”f–’å'oW„%EÝ,-΃±c{¬›K³nËe©°«Ýö1 ¾¸õR¾{rŽ8{ºãÞ‘k7 ¦iQì;nl%×é·°¦2«/9#¢ÀÄöÔ©»;ªÊº%—]®v«-•_ö@¶ÀkåàÞmTäV#µpzM¥¢uÌm³üÀ±¢ ›¥Å)1ýýpë&ÅÒ¬ÛrÙ%ìj·}ì Gežirû^$䑆—†ì™gŠùd33Ob“Ê<Óäì1!»!Ôº·Ì3uÝ–Ë.`ßíV[Æ@ÏüøÊJNóÄ>yƒ-˜=‚¡ÝêØçZ¾æVkª8ÈœéNs3v.’ãÖMŠºY·å´ËØÕnµe6ú`îåÅñbÎŒFس3ïºÅýKßp@+ŠÖ]ÎJ ùœ0ëEÝ(mRü­˜¬›u³nË7§»Ú­¶ôŠ»ÓÌÀ}xï ¯ÎÀõŽ¨+î„”â Ñ8»/_ÍÀ¨2®Çh±Öîl×))­n±Õn…£š¸'ÑÔ•Ü`÷ÁDÜù1çÔÔÇÆdÏÓSW|úm»å8=J¬¨ÍSWrzL?‰m15¥îæ©+ë–\v »ÚíOÁê’‡4ÉÁiKôa$¸Äô~â&¾ä9Mï¼ —|àE‹±¦ò¨FlA/ÆHó•G5ªØRØdÜj²?ÉŽlÈšY?ÉAæ:ñB¢GxÊÈ\c]}ÄعFò­2ÙˆXS™õ“dî[pcÝ%³~\·å²‹Ø7»ýIžz¡õu3b<7ŸÑü€l¦ Y+vY«v_PE8·W’+´VTw@r°‰N,.”R7¨mP·ä²KØÕn¯Ý»3%}ÖcNÇyMGŽG½Û ù½Iì›Ý÷Ƶ †¨Wn\rĪ<ÆÎ$à+Š=`Æ6íH)~Ã8Ã9[f ݸÊ=®Ão\T§¤´F¸ÅVŸíœ˜³·ÍÙ(É9â€8N\vž¡ì§1‘mJʧ5:¯„¼óÞؤ6ç¤$)‰ÅÒºû}6©º{žM&öÝnµåÀ}Rrjm7gÃùé=h$•ÏP®|­@ õ<—°œ'Ɗ겚Óâê=±5Ò[S7(–VÝ’Ë.aW»Õ–óàiê•~%ÎÆpïÔ“À ߉Ès7žÃUbÊ;6'Š¢n–‡AŸ°.gݤPZuKnK‰}³ÛßZ‘Úßu `ÎîÐ)ÀƒÛˆ³Ð–êkê>ÃÝH ¶GoÆš‚n–‡{ aã â­›¿ ë–Üvûf÷ý]b½—“ÏÌÙ°TEßÄZÇì»X‹Iù»lõ–Ó+„EÝ,-Î×áÆÆÁ­›”¿Ë6RnK‰}³Ûßeà;v¼Óx­ü’NxxÕëÔŽÞÀW*¾«ès[€/EùìÛ~cMå«äl8Ô¶#±îž¯B\·å²KØÕn·…sIœHØÇÄáÔ|^ç‘Ï‹H¹¾ÈŽ¶œ‡ZÆŠÚò$ÊÜT&OVSwóodÝ’Ë.aW»uwÐy oMK&çˆmæèc±ÜÖIáWbDa–Õ!›äƒ£YXS»wɘÁí±Â§îî85ë–\vûf·ÛÒùˆ£9–L×€Ù xÖôÛÐÐS®)¼’_ˆðŒÕK&oY‹5>uÇ’Y·äÝÏM€]íV[pâ7k¸s7çBöíè¸> g|¥ŠWb¤>Ôr¼‘|àÍ™±¤¨ÛÖsÄ‹Mcç&wܺI±4ë–Ü–ûf·ÛÂ=Ù¿Q‘œáÄh#ª#ôíXÏÎ|8æ–/h$?ù"Šʗæx‚`ìß °î#£Âu[N»Œ]í¶1o ¾\—‰ƒìÿè»Æ7KqçÚŸ¼síœk,Ç{¯ÄŠªžÇÌ9q{aìœz[7)ó¨ÛrÚeìj·¿ vþñ冿‹8øu¡ø–qr|Ej ü ¦>ÔrŒOÉ;fJaM·×ˆ³áfNØ°î3¹¹nÉm±ov»-œKÎâç_ÉAl:gžÁi|ÙxùwæÌð¼²’|§ +ªx—kÎàÜع°ô[7)–fÝ–{°«Ý÷}îÉ{Qݵ%ù#qËZxŸ»ÇÝ3ö†”ohK¹å'}^ØÓïf/—OŒÝ1[7)–fÝ–7ÿ˜ °«ÝjËÎG)wtry4"¥-óœG¤2÷5 zìxPfyTtcMe&tr*îsE¶sêÎdè¬[rÙ%ìj·ÛÒ" ŽdHår˜¿b˜QGä‘F®Æ>"°åúúvËýKŠºYZD„ÆÆ3¦[7)–fÝ–ËRaW»Õ–mã¯õð‘×ÊÙ aÛc:GNÛÔ§—)Ô·]|xByœÜXS==ÒœXb f&ë.é‘®[rÙ%ìj·Û2økw¸{|­ÜOæÚ˜n:g&>ÅúîtÔ¶Ökªûß’ƒŸ^L,~Î(ußdÝ’Ë.aW»Ý–“¿Bˆpâµrðk\¨]>ý^cÙe?(×-NVÔáh19WŒ­øuEë&•-o)·]ľÙí¶àÉà…Êã2‡ë뎬¶‘Kµã×GD¹-G»åƒ?p(¬¨Ó§¨ÉÁÓBc+Z7©ôÈ-嶋Ø7»ï¶TŽ‡r·¥2!{c[:sû¾Ãë»3³ÌöΞ–œ¹èÆŠÚï–‹3ø…íüŠÕ/ûúÝ=ïë㶋Ø7»Õ–Áߤ,Ã÷ÉÙâÌi ýŒ*Þ*qñÐÁ­½å[õc¦Àš¾H,%ïûRwõEÖ-ùîW¶}³ûÏ’)wU3ôvšu̲q9u§Šÿ¨€È%;Î'—dr‘k?˜H®øÀ. \ ÄIV½n#I®òð³ÉµÀ@ÎE ¹XžýðçL7ïÓꈥ֌ïö“ Ë?ƒøñˆ÷ŒñÏÁ_¸d¹ñ?—oýÓðˆ^ó ~×®ìÞ §w/'û¶¸éjv£÷<›$¹Àj\w ÑoEv<È»‹~+ráG7ï"¤×"2.‹¬ÆþyÏmîÉãÝ{é¯ùyÓÂã¸Õé矓½nü‰_»Ò÷ø|<žîNÿÑçtúû8úšþ.2óݳC”ÿ®¬÷¿`|ÜOýtüO~ŠŸÇþ'eÀ·XvÿßøOÿÙ‘aƒ endstream endobj 142 0 obj << /Alternate /DeviceRGB /N 3 /Length 2596 /Filter /FlateDecode >> stream xœ–wTSهϽ7½P’Š”ÐkhRH ½H‘.*1 JÀ"6DTpDQ‘¦2(à€£C‘±"Š…Q±ëDÔqp–Id­ß¼yïÍ›ß÷~kŸ½ÏÝgï}ÖºüƒÂLX € ¡Xáçň‹g` ðlàp³³BøF™|ØŒl™ø½º ùû*Ó?ŒÁÿŸ”¹Y"1P˜ŒçòøÙ\É8=Wœ%·Oɘ¶4MÎ0JÎ"Y‚2V“sò,[|ö™e9ó2„<ËsÎâeðäÜ'ã9¾Œ‘`çø¹2¾&cƒtI†@Æoä±|N6(’Ü.æsSdl-c’(2‚-ãyàHÉ_ðÒ/XÌÏËÅÎÌZ.$§ˆ&\S†“‹áÏÏMç‹ÅÌ07#â1Ø™YárfÏüYym²";Ø8980m-m¾(Ô]ü›’÷v–^„îDøÃöW~™ °¦eµÙú‡mi]ëP»ý‡Í`/Š²¾u}qº|^RÄâ,g+«ÜÜ\KŸk)/èïúŸC_|ÏR¾Ýïåaxó“8’t1C^7nfz¦DÄÈÎâpù 柇øþuü$¾ˆ/”ED˦L L–µ[Ȉ™B†@øŸšøÃþ¤Ù¹–‰ÚøЖX¥!@~(* {d+Ðï} ÆGùÍ‹Ñ™˜ûÏ‚þ}W¸LþÈ$ŽcGD2¸QÎìšüZ4 E@ê@èÀ¶À¸àA(ˆq`1à‚D €µ ”‚­`'¨u 4ƒ6ptcà48.Ë`ÜR0ž€)ð Ì@„…ÈR‡t CȲ…XäCP”%CBH@ë R¨ª†ê¡fè[è(tº C· Qhúz#0 ¦ÁZ°l³`O8Ž„ÁÉð28.‚·À•p|î„O×àX ?§€:¢‹0ÂFB‘x$ !«¤i@Ú¤¹ŠH‘§È[EE1PL” Ê…⢖¡V¡6£ªQP¨>ÔUÔ(j õMFk¢ÍÑÎèt,:‹.FW ›Ðè³èô8úƒ¡cŒ1ŽL&³³³ÓŽ9…ÆŒa¦±X¬:ÖëŠ År°bl1¶ {{{;Ž}ƒ#âtp¶8_\¡8áú"ãEy‹.,ÖXœ¾øøÅ%œ%Gщ1‰-‰ï9¡œÎôÒ€¥µK§¸lî.îžoo’ïÊ/çO$¹&•'=JvMÞž<™âžR‘òTÀT ž§ú§Ö¥¾N MÛŸö)=&½=—‘˜qTH¦ û2µ3ó2‡³Ì³Š³¤Ëœ—í\6% 5eCÙ‹²»Å4ÙÏÔ€ÄD²^2šã–S“ó&7:÷Hžrž0o`¹ÙòMË'ò}ó¿^ZÁ]Ñ[ [°¶`t¥çÊúUЪ¥«zWë¯.Z=¾Æo͵„µik(´.,/|¹.f]O‘VÑš¢±õ~ë[‹ŠEÅ76¸l¨ÛˆÚ(Ø8¸iMKx%K­K+Jßoæn¾ø•ÍW•_}Ú’´e°Ì¡lÏVÌVáÖëÛÜ·(W.Ï/Û²½scGÉŽ—;—ì¼PaWQ·‹°K²KZ\Ù]ePµµê}uJõHWM{­fí¦Ú×»y»¯ìñØÓV§UWZ÷n¯`ïÍz¿úΣ†Š}˜}9û6F7öÍúº¹I£©´éÃ~á~éˆ}ÍŽÍÍ-š-e­p«¤uò`ÂÁËßxÓÝÆl«o§·—‡$‡›øíõÃA‡{°Ž´}gø]mµ£¤ê\Þ9Õ•Ò%íŽë>x´·Ç¥§ã{Ëï÷Ó=Vs\åx٠‰¢ŸN柜>•uêééäÓc½Kz=s­/¼oðlÐÙóç|Ïé÷ì?yÞõü± ÎŽ^d]ìºäp©sÀ~ ãû:;‡‡º/;]îž7|âŠû•ÓW½¯ž»píÒÈü‘áëQ×oÞH¸!½É»ùèVú­ç·snÏÜYs}·äžÒ½Šûš÷~4ý±]ê =>ê=:ð`Áƒ;cܱ'?eÿô~¼è!ùaÅ„ÎDó#ÛGÇ&}'/?^øxüIÖ“™§Å?+ÿ\ûÌäÙw¿xü20;5þ\ôüÓ¯›_¨¿ØÿÒîeïtØôýW¯f^—¼Qsà-ëmÿ»˜w3¹ï±ï+?˜~èùôñOŸ~÷„óû endstream endobj 147 0 obj << /Length 1961 /Filter /FlateDecode >> stream xÚ­XmoÜÆþ®_Á<ÀÜì—dˆH¤v$l Ž´Äöî(“w‘ýïûÌÎ.”(;h ºÝÙÙÙyáåõÅ7?™"©Då´K®?$J[QJÒ ¥«äºIÞ¦ý‡Mf•Lw--TÚÔÇšAûú8tŸ^`cüùÁíxìpÔng7U:¶¼=Ë¢ùú‡ïRïnû¡;ÞíyË«t»Ñez×wÛvóþúo‰L2¥D•çÌb}h6™ÑEàÁw[ﶧ3TÍ­´ÐÆ‚Œ¿þ†1t‘8Q¦$ŒB”…KŒp*¼¡™¹¾ Qd3¬z»í‡î>oJ“ pouú=qdƼÈÄë¾ßƒYbúE`yèÇ1ûc“çÐB-wýQ·uXôÐ%dJ‚C«—z¸Ùd Cô —ÞóføÐû–ôSé4Ë«IÏ#cïêc<ÙõÄïo÷ÝÈö¢ ˜s@8µ#ïßÉ\v¢…˜yaÓCH\ýãGœ(Á»_@![˜Í1»LÖº4È|jyÛš6pÕ!*'ϨˆîŒü;nI¥wmsÚy¡‚Îþ~×F¥Ïï†_¸K;¬ùׯÉ7F±âNÖ‹pùSîäžw§R˜Ò±;E ‘T•I›ö4öà%¶ö` É¢­X•8ba±`?ÃbßzïÁjDo‡HˆC™VdI#lXùPuwý-¯„«ÔR-#‚¬mÈÒ¸fLH+úq¹pùn¿¯ˆ^ i\D*Ö圄ye“L6<ÚÕÕïà³(Ó—þ¤Q¢Ô9ÖÙų ªÊ S%Ù ‹ô°"ý&¢ˆrf`HHì)a« î_+Bª"™ayí=õ~Ï ¥ÌÂ_IJ2‡à_ƒ,N<âëß^]]=¯› m ¯—ü+zQ¬˜üÿ¦»T ðÜ´IfH樂kŒI¸Áä(ÿ\a(“ÂóßqôÄTnÉÒ³¦Ê½©ì2´\U4Jþ› Uä_KO©=bM#z²ÖK¶¸ZѬB•R**ä½ÒjR»ÐŠ­Õëå +ªrzŠâói¤gZjas迪§G½ö}€©B>u{Êh§=¦ŠÃ[JoOSJ …Wþ×z^¢¤ªE=§Lï¢lSa§øçá®ã¾$ðŽD;¡S5¥ªé\¡«Û ÊΊ(•Å—s¥Bµ‘eć>Aû¨°C{?´c(¡>éËåÎÄ pšnKõpüv%“ [ºG,-ÃÙV.=ƒØ4H ZäªúB¸jQæ%ýLh¶>89§£œcë)u0ß_јè<ý¨Ñ‘üÌëï2úµéM Ú¡£|¶o~+TÖ楒1ä7Ý×¾q]¥KG_£{ñêú‚âE&jj¾5<@¢m÷/4Š’1fKïÀ7?ïmòcññ(‹4³ÑËgÛ~] ŒlpŸºÛÓºsû—Mæ¤J/}³ñé~GžUzjÏÉ•¼Mr½ð +´;·Ùk!”<Ÿ"j;íŽ#÷­ Q¿dMÒò×îžF‰z£ÒŸ)¦ñá_‚5ÑæŽsji¸_Öný¨!Q‹^å)¹ž‡ ÍIaËÍu¿!iÒ´».´†s”öxŒ€³< Œ¤p`ðåâ±ÿ‚˜õnäèäp¥ë]€´Ÿjêd‡r®åÆÈ;K§ÆU¹¶½oꥧã]ï toÔ#¯Ù­D4Nvû÷+žçÚóå#ÐÔùžF?mjÔyù¤[ìØóo)3”½ª(QÚªb©“þžp)›æez5ÿƒÒ>ð‰ß]_ÓFªÏ‹àJOÓ mfc!yf¢w¯pR•^›O•¤KÌÈný×ß~¹zõzECHžÕyÔ¤Üpè‡,äl€CÞ\È »Ñçê²H¶˜¬J—zö¸ŸÇnüxª)D ±ßùcI* ßsZ-:`oí{¾‘”þ—GL,Î3$6§Àˆc‹'‘µy.Œ¶âÉ.!2êÛŒªŠ9èÌ¡”ep\`uz,è‹u2òþ‘&ùl:ðe0ÙÖ?ÄÐäÉ––þK‘ˆQr*Z…Ïg·l¾˜z±ç x5ï·íp¬»Ãuþf×2䆞ûÌ7BÍ•Ú‚ÐlÎ-ÁcÞCä«è±L\p1©¬kÚš±‚¬ Ö0 ¨&]ÆXÄa¤¾¨È¤*Ï&㸢³ûÁá?çË­1üÐQ×c$4t:öô)ˆrÍg ›»îö®õOÈ´ûÀ0~^>n™<Ê8¿æ5àž‹Ï܈ÜگǧE7šG46­¬âÔŒU·ѪKt¬Æ-J¹UðyÔ‡éKBš±G#r¤+ ê‘´ ·>Æâ.ã Ãœ…ÊbRð…úí{™48ƒzQûËäÁcî© —@”É.¹B±çz>pT¥D!OYü"ú''¦7ãÚÓN8[aÞûÂÛ¨¢Î©/½i‘^Mµò6.˜Å̘ze<™wïNh=5 jµs§c4üÀ_a“XŠlyöᨎ ßÛóú—Kº?´kÍ°C«[M3IF5h\áIƒk7kÓý3ìÜv½‚qñCã\Yï ˆ¬WD«ýD(_§ endstream endobj 143 0 obj << /Type /XObject /Subtype /Form /FormType 1 /PTEX.FileName (/home/biocbuild/bbs-3.20-bioc/tmpdir/RtmpCCJHlc/Rbuild86c017e3fbf7/pcaMethods/vignettes/pcaMethods-011.pdf) /PTEX.PageNumber 1 /PTEX.InfoDict 152 0 R /BBox [0 0 864 504] /Resources << /ProcSet [ /PDF /Text ] /Font << /F2 153 0 R>> /ExtGState << >>/ColorSpace << /sRGB 154 0 R >>>> /Length 872 /Filter /FlateDecode >> stream xœÅ—Ko1 Çïó)r„CCœ8¯#]¢¨eWp@œJ9T´ âëã™ubïvÙvè"3•;?O2ñßsbÀ\šoÃ_Ç«áÉÍÛ—ÇæüfpÖ9gôýæüzxò“Ïê³oùÁú¶¶½Á@Õ•ytºxlV—ÃóÕÜ÷«—©µ|°±Ñz?-pÖÞ?¾{ÃyÚÃQ‘f›¼ñmŒ“«×[S—IÉæ`b²Å™PÐB0¡:›À|¿0ïÍu‡×³Xîø˜åâ ý7GókøðÑ8óisB×åÓn^ôÆØæÊ43[,æ‹1Ëò´ÕNt[‹ÍUn 0Xé¶ è³õYn r/^Ÿkq‘V¡Õ\bkZcý0Ûý9› ð€£À¦ Ø”X› XÆ? `SqŒ}Øœ€9)â¶Rdº‡hk0™ÄµV!îÒð>o VßÜÓ\wÐò“ÜËl÷LI•Ù½¸¹î’°¹û¹îHÕÀCsÇí>@J®¿ed„1ÿuB¶ç̀ӯ̀“¯Í€S¯ÍàÁú+y¬Ð¡X_§C„û2Øœ•ïŒàyú$Ê7Ì<:ë£ò‘0èÑbU¾qK1wWû[’BÌtOµL’ ã2wHªf[²H*õr JSB°ˆB b”¨„`…TmöJUB´‚NÝszdžÊ]ßwŸä ‘S§“´ ¦JŽnóÔ'•W  ¾O*—n?¾ïFw/>˜ãQV5þx:t¤0óüsºxzеë›g<>È iÕœÒm"‰6©9¥Û‚$:»ªæ”n ’I¶ æ”n R¨µ¡šªº-ȃ+(RL´lâÜ#u1ÐÎ3Šh¢BH^ÊyFÍT CÕÎ3Êhq° ·ëèAµÔû›ÒÿX9 h¦¬›4SVMš)k¦ÍT@@[kd E%mL² ôQ– =Ë2Ò‡YFô4Û6Î6DͳŒô–=Ñ2ÒGZFôL{ÿzñ VÈÁ¤sÊUÅVš3Œ?òT•PGÓSÖ¤¢+‡3ø8Ž*^!8ž!Óxß7ý‡ù‡JýŸÚáÞÕÒUŒº.ß=;zuõõ狃vÅ;¾`y«H ¿é‰l endstream endobj 156 0 obj << /Alternate /DeviceRGB /N 3 /Length 2596 /Filter /FlateDecode >> stream xœ–wTSهϽ7½P’Š”ÐkhRH ½H‘.*1 JÀ"6DTpDQ‘¦2(à€£C‘±"Š…Q±ëDÔqp–Id­ß¼yïÍ›ß÷~kŸ½ÏÝgï}ÖºüƒÂLX € ¡Xáçň‹g` ðlàp³³BøF™|ØŒl™ø½º ùû*Ó?ŒÁÿŸ”¹Y"1P˜ŒçòøÙ\É8=Wœ%·Oɘ¶4MÎ0JÎ"Y‚2V“sò,[|ö™e9ó2„<ËsÎâeðäÜ'ã9¾Œ‘`çø¹2¾&cƒtI†@Æoä±|N6(’Ü.æsSdl-c’(2‚-ãyàHÉ_ðÒ/XÌÏËÅÎÌZ.$§ˆ&\S†“‹áÏÏMç‹ÅÌ07#â1Ø™YárfÏüYym²";Ø8980m-m¾(Ô]ü›’÷v–^„îDøÃöW~™ °¦eµÙú‡mi]ëP»ý‡Í`/Š²¾u}qº|^RÄâ,g+«ÜÜ\KŸk)/èïúŸC_|ÏR¾Ýïåaxó“8’t1C^7nfz¦DÄÈÎâpù 柇øþuü$¾ˆ/”ED˦L L–µ[Ȉ™B†@øŸšøÃþ¤Ù¹–‰ÚøЖX¥!@~(* {d+Ðï} ÆGùÍ‹Ñ™˜ûÏ‚þ}W¸LþÈ$ŽcGD2¸QÎìšüZ4 E@ê@èÀ¶À¸àA(ˆq`1à‚D €µ ”‚­`'¨u 4ƒ6ptcà48.Ë`ÜR0ž€)ð Ì@„…ÈR‡t CȲ…XäCP”%CBH@ë R¨ª†ê¡fè[è(tº C· Qhúz#0 ¦ÁZ°l³`O8Ž„ÁÉð28.‚·À•p|î„O×àX ?§€:¢‹0ÂFB‘x$ !«¤i@Ú¤¹ŠH‘§È[EE1PL” Ê…⢖¡V¡6£ªQP¨>ÔUÔ(j õMFk¢ÍÑÎèt,:‹.FW ›Ðè³èô8úƒ¡cŒ1ŽL&³³³ÓŽ9…ÆŒa¦±X¬:ÖëŠ År°bl1¶ {{{;Ž}ƒ#âtp¶8_\¡8áú"ãEy‹.,ÖXœ¾øøÅ%œ%Gщ1‰-‰ï9¡œÎôÒ€¥µK§¸lî.îžoo’ïÊ/çO$¹&•'=JvMÞž<™âžR‘òTÀT ž§ú§Ö¥¾N MÛŸö)=&½=—‘˜qTH¦ û2µ3ó2‡³Ì³Š³¤Ëœ—í\6% 5eCÙ‹²»Å4ÙÏÔ€ÄD²^2šã–S“ó&7:÷Hžrž0o`¹ÙòMË'ò}ó¿^ZÁ]Ñ[ [°¶`t¥çÊúUЪ¥«zWë¯.Z=¾Æo͵„µik(´.,/|¹.f]O‘VÑš¢±õ~ë[‹ŠEÅ76¸l¨ÛˆÚ(Ø8¸iMKx%K­K+Jßoæn¾ø•ÍW•_}Ú’´e°Ì¡lÏVÌVáÖëÛÜ·(W.Ï/Û²½scGÉŽ—;—ì¼PaWQ·‹°K²KZ\Ù]ePµµê}uJõHWM{­fí¦Ú×»y»¯ìñØÓV§UWZ÷n¯`ïÍz¿úΣ†Š}˜}9û6F7öÍúº¹I£©´éÃ~á~éˆ}ÍŽÍÍ-š-e­p«¤uò`ÂÁËßxÓÝÆl«o§·—‡$‡›øíõÃA‡{°Ž´}gø]mµ£¤ê\Þ9Õ•Ò%íŽë>x´·Ç¥§ã{Ëï÷Ó=Vs\åx٠‰¢ŸN柜>•uêééäÓc½Kz=s­/¼oðlÐÙóç|Ïé÷ì?yÞõü± ÎŽ^d]ìºäp©sÀ~ ãû:;‡‡º/;]îž7|âŠû•ÓW½¯ž»píÒÈü‘áëQ×oÞH¸!½É»ùèVú­ç·snÏÜYs}·äžÒ½Šûš÷~4ý±]ê =>ê=:ð`Áƒ;cܱ'?eÿô~¼è!ùaÅ„ÎDó#ÛGÇ&}'/?^øxüIÖ“™§Å?+ÿ\ûÌäÙw¿xü20;5þ\ôüÓ¯›_¨¿ØÿÒîeïtØôýW¯f^—¼Qsà-ëmÿ»˜w3¹ï±ï+?˜~èùôñOŸ~÷„óû endstream endobj 2 0 obj << /Type /ObjStm /N 100 /First 817 /Length 2571 /Filter /FlateDecode >> stream xÚíZIsãÖ¾óW¼KªF‡yxûâR¹J£ñ,UvìŒÆ‰ Qð„ €³äßž¯A"¸ 2upÅ9Fw¿^¿~d‚Y¦s,J™TŽIÁ´¶Lj¦N†k˜ôÌœ³N3%™z sÁ3e™Çs¼2ØÎ E¬,šHÚ¢À/ÁúÆ“±bïÉvx!Jø$2+ …„ ¬ôÆ×d>¡˜5ÐAÂcðJHá`—K‘A‚Á²p‚ê²ñ—"àÖ–ð|“‘ÈVÊ“—ÈÈ‚î8æ¬Ivs:Ã=Œâ‚´R à*aª»»@€ÊEC0²0`ù^Ú/Vo‹½hÕ¶Vrš_Z9÷ë¾·|µ_˜ÙvÞª[´ëè¸áS×ÓÍúG¿Ã¾v[ößÛ—«nDLºvJZ×±¹s³~ç>Ö;æÛµy+¬~8#쮌ø[Ë&oÓâãV` »Ù¸ëi»¤ÑV:-ÁhGvØÜu_ªXÏŽÿ¹å¹'ÐÇíaSÔ–ÄKZÁÞ±äU^_‘¸ÓÓAòþë]Æ’ŸÒq6H΋YÍꊡì‚p¼Ëªb^³Špsç‡l”§/Š/ìRxˆÐM3ÉC¼€O ÌÚõÙlV€Ý%õ{’ŒO§«AGxC5H^å(+¾ò*y“¼MÎñMàŠÖìRzÅ úƒ5Ž[°4Jpê’Öá¾w »˜_×à™|ŸÏ>&g§§„䬱rr‘üüî-ý=»­ë»o’ä:/òÙMQNÓ:V|z—Oïž‹É—c>Ê’è¹KÕ¥ÖÈÙ¯¹h5WFð ©ë;Ð~”ñÜ*ð0 -ðŒm†Ç0¯3^Õéðã$Ï*á›Y*³æË_þù/jÈ< 9z«yD‡›Í'“«½´fA+À3}h­ðÜ¢üâ€ñŽíÒ¾B`5 |…8.ZD!¢ aq  ‡þ»¸¶d§ §_ݺ «ûðºX^£Ýã•å}DG˺âl`ü¡RòSY /2ø‘ÿòKÞg_êͨÜH g7SÂé£RÂ.–e+²MÆ~mNê@ž<:ø€à>‚‹†7 Js Dc­âiûƒ¯.‹¯é¬ú˜~M…\¿ÇªC‹ºWè•{ G«7@ŒÚhpõ(‹ÃN}ÞÜE qBE ʈ h¯"2×Z#U1cìTã&ÏË'ÕAiÛñI!\ÂÂFzšjÒ^?` µW‘ÅTG.hÎB!@ö©`xDÜEã]K'YZò|Væéä.«øM™üçSV^gÙÇõ*Š¢áL·ÀÐIJ¯pmÒ*\»hÊ,áú>´6DŽé¨­‘UŒ}h1ö¢YÙôˆ9A³¯ ÜãÜ‹Qëã¾:Û§†ÂànW=Å„æâŠCšÜQOQ½|{íàÀÕ5f¹V©¹!lÖÜàŽª¹aQWÃBó°P:˜n±=/³”büeZgìÙËo”@R …¡S` >ðb´÷Ùû¼žàÉ;öºLïn*ØóúnøG͇YIO G·ªš¦Ã%äºÌ@ƒWÒáÞJF9DTy•pxk”Ý$%d'Cxz2IÛ›·óÙ8-çÓI:¯“b\ÌP`†)ñ«îÒaöD-SEÉ0•Æ 5RÅÀµVøœSaË,®S¡Ÿ°C z@;n¨cpÄz“A·@M‘u‡£Ù€@u†öFÕnŽ½ê ¢†¨^˜h[¬Á9A¤Ü!%ôSnD`àÔp=,h!Њ¹²½• ¸•±?>Ø ^MµXy0ýˆ]HfדÖöUÃzêg²± öÚ >Ö’ëYËF` ©ìQÙ¨·³Qýžl|xôJìÏþ£ Ô»fA©ÿTÃàÿ‚\Ì,õN–Ï_î滋Mú?K|P…Sñh©C H'åŒ<ØhJT1? ®]L³={Ñm,g´·Sú÷TÅ\ÈéTLcS1åâÛË:€1ýwÔ¼rÒÏýqÒÏî,£öÿ{jýGD­8ý3mˆ;ú)ÌÎ ÁC#ðqÿˆx“!V»Ã*åÕ 7ôÂ\b$_ƒ6BFÈÇ÷Ÿ»ŸMuqÇ^ùú7É5Œ³ñA³ÝtßNyúÿ'Ý#Ëÿ ÇH.õ endstream endobj 164 0 obj << /Length 2328 /Filter /FlateDecode >> stream xÚ­YKoÜ8¾ûW9©4Ã÷#ïÁØd‘f&™ø°ÀdPZ²­¤[ê‘Ôvòï·Š¤ÔR›ng€¹XY"‹UõU}Õ¾º¾xõV˜Ì§¹Î®o2Æ%±”g†j¸ˮËì÷¼½Y­…yû¹¯ºû•Ry1ÔmÓ“ÕZJ™_ßUa½úVlʼ¬>Q!›åÂbÝ¿^ýqý_8ÍÎ3†H¦³µ`„+ŽûeµfÔå¿­Ö6ÿ9Œ?®ÖÊäoà¯ÍßÇmh¦‰3Ââ6Òéx¶fða—¯AŒÛùiDk—Í„.£Ð\%F‰•˜1ÄîQ,£ÙZe¢QÇ9glb΄¹ç.Þ\_üuÁ`‰f,ãöf,ÓNÂ~<Ûì.~ÿƒf%,Â×D8›=xÑ]Æ#Újo³.®Ðc‹;qK‰6°™Ie‰CQŽšLƒÉ…¶çUl¾újÏTáY7›v·ßVC˜+Ö§ô`Æ‚§‚}›P„™ì^ŸªÐ¬¸Í;üûyµ†¿U ÅኊÈÂP^O©—zJsÔs7ŽP– éö„],`b @Çz£žNïÀ‡e€‰š] ô ×…‚ea&hlò]Ý÷us'O‚ñRá9=u §¡¶LAú%ÅìêES¦ˆ†àÈêå‰}Ï—à3&/ qjiUÄŠÕIO{ 4 L&ò“L„³Yð|ÆDéªçL„s ©|ágØ?ÆÕTŠ«- tü0îð]¼ùMÛ+¡=#@ ¾A›ëuÕõ®ð°Öb„ñöP¥;IFøL(¦ç‰d¹å“€ÆÒ2v>F©˜v=xÇ' Ý4¤¾¦¶žjö-JØÈ7Ø`åU´ÃÛúöÐù±ÉU˜òöÄAÑ„yhÂ1£1M+bø¾_—’ùF;ÿíçoÞ‡÷þ®=lË0Ž¡/<Æ>ÌV]çýÃ]Uô^ð¡“DÅ][opg,FÛüê0©:JÅíøP Ø/²§ª#–ôKŒ€2š÷»b»…R'¹‹'ÀääÞPØ@|*l°Z„ÇXØ–_`@ø™] ~/QOHÛq5¦¥S] D½ßBP›oºz¨:Ð=*tþî&Ì÷mxz'³ŽóËzì%7Å6ä]xy@…Š>¼ûý¶®ÊiÅäñ“µÜUM\âY‡¾J)üÇßi@ñª(c*Ñt£ø¼ñˆÑA à câ  5¼ü´Æ§Ì¿¾‰F*´+_†Õ]5ܵe_Áûý¦x׫ûbû~Ó_²×*Î4Ý¡ÙÜ_²—AºÚ]¾hº]_í_ŒÄ)éHH™Ö êq)¡ûqHHÿº †iª½Älè—ÆïâÄ«w;•ý»‚úaZZ{®g›^=úÝ <¹“ÌÍ#áÊÁ'ê5ò–_µè»oûm;à‚FСÏÖ)§xkJÇlùê­âóS¡Zé)­ü ü8·@“*äDîWhý¡)Â@f'é×ãÇ:yD¨™!ñøÄTõ”꘿˜¶<Ø2žV Xq„Y*ùÙ!£[4Ûï8R9 ˆ "üž¨!I|­„´ÝÆá¦kû> c©Ë"yxüµhÎC0R˜Ì>^’í®*Ã@jßaŽ’W‡Ó8rB‡œô=¼å}݇”€¯C»ø4¾J¢õ©Õ!ïÀÌŽs_ýFm“„ðd!’ˆ•5ǶµÀbG:á46 äéT“ŠmHV*ßw>bïëÒqŠt-#(˜ƒÅä-f~||$c]ŸcAÁ¥PÉ(òDz†Íƒ‡ÖV8é3Å6Áë}Ça|yK4 úü‰l~Hô‚À©Nõ‚Ë2-Å¢,kd7›Øø+8sl "&ýdÑÅIHå›ÃŒ^†HÈJû¢[1¨l Š¡AB§Ðé“*^¡ØN4FÍ:¯kÊÀ`W4ûâ¶:‡ÇQ”ö½?ý¡ð€Rûˆ/·ÃºÅ£ÝÞýjGä©)à&fSVP·}d4ܾ c¼‡t:ĶÄÝ‹[Òk?³‡º]Õ¶Ʊ¢ùm’cÖ–a EÈ!¬á=¨ˆûÝ倿›Î=†e ©ÇA˜«õ¯sˆàû®ø:*…¯}ÕôÒ-¼Ñ=Ÿ¬Vƒ">,ðµióÞíô§ÎGq qM* -䙊Kô«È1p-Ô@(Ц c­C)H´Ö³\û¾òF ‹E¿h°âLí‹ ä.¼"¤T„Eœ÷‚§ŒLæÉä}ê)ÌÜoBÆl笖<+‚éÃøqD‹Ìo €cÔ(Î $ó7‚ÒþÔŸÌX^Ãáÿ> ÒCG?´#¹ù?µ¯ÆÁ endstream endobj 159 0 obj << /Type /XObject /Subtype /Form /FormType 1 /PTEX.FileName (/home/biocbuild/bbs-3.20-bioc/tmpdir/RtmpCCJHlc/Rbuild86c017e3fbf7/pcaMethods/vignettes/pcaMethods-013.pdf) /PTEX.PageNumber 1 /PTEX.InfoDict 170 0 R /BBox [0 0 288 360] /Resources << /ProcSet [ /PDF /Text ] /Font << /F2 171 0 R>> /ExtGState << >>/ColorSpace << /sRGB 172 0 R >>>> /Length 394 /Filter /FlateDecode >> stream xœTÉnÂ0½û+æHz¼&W­„ *$7à€€V úûµ±LÕVJ±òü–x<£ €` lìžÇóä¹ Ë3ãh•†t=/‘î•ŽæœCº–½‘7iødÓ9pX1‚{¶Œ¼†ÌT¬FS€$´„Ð(,œÖ ºN¦PÙ;ž ŽŠj^ ÔêžÏ’®ùœ£¡o¼Û‘”£ÉïykP×ùÝꇪ®E?9%To@#–€I*äÚ}Ë'V{輫ÍáýœAµeýÊ7‰ï!‰s¯…"ƒBùŒig4–ýW˜uJµ[Ãæ²>e@:‹Ëæx˜eÙª¿µíŸ}Ñòú^÷°…k:ìÊï> stream xœ–wTSهϽ7½P’Š”ÐkhRH ½H‘.*1 JÀ"6DTpDQ‘¦2(à€£C‘±"Š…Q±ëDÔqp–Id­ß¼yïÍ›ß÷~kŸ½ÏÝgï}ÖºüƒÂLX € ¡Xáçň‹g` ðlàp³³BøF™|ØŒl™ø½º ùû*Ó?ŒÁÿŸ”¹Y"1P˜ŒçòøÙ\É8=Wœ%·Oɘ¶4MÎ0JÎ"Y‚2V“sò,[|ö™e9ó2„<ËsÎâeðäÜ'ã9¾Œ‘`çø¹2¾&cƒtI†@Æoä±|N6(’Ü.æsSdl-c’(2‚-ãyàHÉ_ðÒ/XÌÏËÅÎÌZ.$§ˆ&\S†“‹áÏÏMç‹ÅÌ07#â1Ø™YárfÏüYym²";Ø8980m-m¾(Ô]ü›’÷v–^„îDøÃöW~™ °¦eµÙú‡mi]ëP»ý‡Í`/Š²¾u}qº|^RÄâ,g+«ÜÜ\KŸk)/èïúŸC_|ÏR¾Ýïåaxó“8’t1C^7nfz¦DÄÈÎâpù 柇øþuü$¾ˆ/”ED˦L L–µ[Ȉ™B†@øŸšøÃþ¤Ù¹–‰ÚøЖX¥!@~(* {d+Ðï} ÆGùÍ‹Ñ™˜ûÏ‚þ}W¸LþÈ$ŽcGD2¸QÎìšüZ4 E@ê@èÀ¶À¸àA(ˆq`1à‚D €µ ”‚­`'¨u 4ƒ6ptcà48.Ë`ÜR0ž€)ð Ì@„…ÈR‡t CȲ…XäCP”%CBH@ë R¨ª†ê¡fè[è(tº C· Qhúz#0 ¦ÁZ°l³`O8Ž„ÁÉð28.‚·À•p|î„O×àX ?§€:¢‹0ÂFB‘x$ !«¤i@Ú¤¹ŠH‘§È[EE1PL” Ê…⢖¡V¡6£ªQP¨>ÔUÔ(j õMFk¢ÍÑÎèt,:‹.FW ›Ðè³èô8úƒ¡cŒ1ŽL&³³³ÓŽ9…ÆŒa¦±X¬:ÖëŠ År°bl1¶ {{{;Ž}ƒ#âtp¶8_\¡8áú"ãEy‹.,ÖXœ¾øøÅ%œ%Gщ1‰-‰ï9¡œÎôÒ€¥µK§¸lî.îžoo’ïÊ/çO$¹&•'=JvMÞž<™âžR‘òTÀT ž§ú§Ö¥¾N MÛŸö)=&½=—‘˜qTH¦ û2µ3ó2‡³Ì³Š³¤Ëœ—í\6% 5eCÙ‹²»Å4ÙÏÔ€ÄD²^2šã–S“ó&7:÷Hžrž0o`¹ÙòMË'ò}ó¿^ZÁ]Ñ[ [°¶`t¥çÊúUЪ¥«zWë¯.Z=¾Æo͵„µik(´.,/|¹.f]O‘VÑš¢±õ~ë[‹ŠEÅ76¸l¨ÛˆÚ(Ø8¸iMKx%K­K+Jßoæn¾ø•ÍW•_}Ú’´e°Ì¡lÏVÌVáÖëÛÜ·(W.Ï/Û²½scGÉŽ—;—ì¼PaWQ·‹°K²KZ\Ù]ePµµê}uJõHWM{­fí¦Ú×»y»¯ìñØÓV§UWZ÷n¯`ïÍz¿úΣ†Š}˜}9û6F7öÍúº¹I£©´éÃ~á~éˆ}ÍŽÍÍ-š-e­p«¤uò`ÂÁËßxÓÝÆl«o§·—‡$‡›øíõÃA‡{°Ž´}gø]mµ£¤ê\Þ9Õ•Ò%íŽë>x´·Ç¥§ã{Ëï÷Ó=Vs\åx٠‰¢ŸN柜>•uêééäÓc½Kz=s­/¼oðlÐÙóç|Ïé÷ì?yÞõü± ÎŽ^d]ìºäp©sÀ~ ãû:;‡‡º/;]îž7|âŠû•ÓW½¯ž»píÒÈü‘áëQ×oÞH¸!½É»ùèVú­ç·snÏÜYs}·äžÒ½Šûš÷~4ý±]ê =>ê=:ð`Áƒ;cܱ'?eÿô~¼è!ùaÅ„ÎDó#ÛGÇ&}'/?^øxüIÖ“™§Å?+ÿ\ûÌäÙw¿xü20;5þ\ôüÓ¯›_¨¿ØÿÒîeïtØôýW¯f^—¼Qsà-ëmÿ»˜w3¹ï±ï+?˜~èùôñOŸ~÷„óû endstream endobj 184 0 obj << /Length 1260 /Filter /FlateDecode >> stream xÚ…VKsÛ6¾ûWèÍD0|ô§ñL:ÓŽÓ¨½4=À$$q̇Bvòï³Ð’5ÀîbŸß.¨V üÔJ¥‰4:[Ù¤”E¦WUwõåJÚ\Iœm‰¥â½H¸þÐe«_‡«ð[X›EçæLéÍöêúÖä«R–°¶»•I™ÛU–f2MŠÕ¶^ý#n›ý<úõFë\d¿¬7¦TâfXëB|=¶Ã„ŒL„Prñ´Yÿ»ý Ln”’¥µ¬¡é÷ëµ¹˜ù×·VŸ[-¤NJ¸DÂüùû§÷w,öÒ¹L–z‘Ú #«tmK&7¹•J—¯ WCwlýk› 76î¾õIMÿÂãŒ/¢£k2Q»Éá.ÁOr½Q‰MÅíºH9`žczóçĤc€ì[ˆœ0oÏÌ#©kBÀô\ÐÀζ3x’å©hºã<¹©z<QùB€¢ w{¬‹µ¼öï¦×Î=`íü‰ §á’’åæ ëzom}ð2–*M X‚â¬Lr³ÚZæi„Ž…„©$7avmÐýµÉr1ìx¥d›,£s;…¨ß¬ØÒ®i# ¥¢b©Pµ¶âãÜTk£Å¨È­Õ0Rnaïúš7íàjHr$h„_HSD×? b>ƒ„ùé€YÊE˜BàÓ™uŽ­Ã}c†„+*]Á3¹‚WO®yŒ6\&Gj­Ç¦ö5à-ÕJlžy»¹¯?öЫ҅£üœØþêB´©•E’.Ýäš“S–„SâžÁÅDÇ´ĸZÏç'tt‰ |¹_°&À{h¿1á1Ö?^|.ê¦"Áæ¬H Ûrý|œü¤IŠ2`{š›©yŒFòR¾aŸ¾Á$"t#{gibEÎ`ˆÜ¸ó˜êÆ(z^ f=|4Žò 6¤=†Šô¡¿8îÞ¡šÃÅ¢VC;Ìc`jE!Áæ9í(Ž|ª&äˆÔx2~G€9‘]1‰·¦ÖhQ3}fúƒ(BËK>Ø34æL€¥09F%¯¼«˜*7M>^á×e›‰)Mà5>T4.ÝÈ‚£‡ÝùžŽ’)4_Užš*,†"„†Ð…~(]\_cÒ9 Õ]ï=ŸØ§žÑwØ-Mw9GoaÒÐ{Craz|&„°8²t>Pø‘Ñé 9zAlΰ‚³Õ†¹­ù»’^s>ÍFTQ3VaKXqÇGþtÀÝntÝÅ©0ê, Tƒ›RD8!mô­çOnu¦Ö>Tcs¤™—“@ܹjƀ㠺èùûú¯?p &0vÓ¢¾8Sâá-Ó†þ²›Ôó©ø0R+KÁÁ&WÅ—À3§‹£—.pX°‰*Îh Þ3œ˜Cé;‰b;é]{²ÈtN'ë§wÌ`4’~X—Π1ù3ø–Üñ*+¥É,âNI« k.‘wõ~{õ)ˆ ã endstream endobj 161 0 obj << /Type /XObject /Subtype /Form /FormType 1 /PTEX.FileName (/home/biocbuild/bbs-3.20-bioc/tmpdir/RtmpCCJHlc/Rbuild86c017e3fbf7/pcaMethods/vignettes/pcaMethods-014.pdf) /PTEX.PageNumber 1 /PTEX.InfoDict 186 0 R /BBox [0 0 360 360] /Resources << /ProcSet [ /PDF /Text ] /Font << /F2 187 0 R>> /ExtGState << >>/ColorSpace << /sRGB 188 0 R >>>> /Length 816 /Filter /FlateDecode >> stream xœVMO1 ½Ï¯ðq÷€;ßW­@•²7ÄÂV*ÚBÕöç×Ù™dâQ[u9ìh’g{’÷ü±§@p߆ ù½yùðîn_ƒÑyèŸ/·O|t)°1úçåÑyqòðs¸ºwÁ©üî*p6„Œ&Bô284È’çí‡C´˜B³G›+šä MsE“E¯#‡„>U4¡£eöÈqB³Eö 5ÉW4¡I:r@[#“a̬`ÙI¡Á±ÒÃ,±m…‰Ñ玹œ´¢Õ±…Àz0bF:¶0è µÊ›0µƒ[¬ƒ“+;5¡ÆÎRÝ‚b4xÌ-¶ è”V"UºÉ²b4õ¾> ©È"mh|ƒYñéwòTÔcTt¦ŒTS„¢Á ¤ ’4MÙwŠMo152“AÖBE´ÍW.@v‰¶ûÊ!²¢Ò›9­);LJ¦@¥ *š1(&%­O(‡N1)¯L Íh“B|»‹Úähô ME’eŒ•g–s&ô*²ÏM£,=L1iÝ\GV›ÅØ64L©ÖPS^'Tz}Š‹¼jMÉ’T¨ŽKAèáæSe7tÞ²L¨Í'ép8ãc\‡’B’³‘aóW«“§Û/_¶ß·ðc ÒÝW7Ïk8„_}¾ù(;V[8yºÛ®ÁÊâ×ú6§Ãñ¦¢}®ûÆA÷iyÍAæ€0lË÷VçÎ.߯as_"ÌSô•cÒï2s$éÆË,”> stream xœ–wTSهϽ7½P’Š”ÐkhRH ½H‘.*1 JÀ"6DTpDQ‘¦2(à€£C‘±"Š…Q±ëDÔqp–Id­ß¼yïÍ›ß÷~kŸ½ÏÝgï}ÖºüƒÂLX € ¡Xáçň‹g` ðlàp³³BøF™|ØŒl™ø½º ùû*Ó?ŒÁÿŸ”¹Y"1P˜ŒçòøÙ\É8=Wœ%·Oɘ¶4MÎ0JÎ"Y‚2V“sò,[|ö™e9ó2„<ËsÎâeðäÜ'ã9¾Œ‘`çø¹2¾&cƒtI†@Æoä±|N6(’Ü.æsSdl-c’(2‚-ãyàHÉ_ðÒ/XÌÏËÅÎÌZ.$§ˆ&\S†“‹áÏÏMç‹ÅÌ07#â1Ø™YárfÏüYym²";Ø8980m-m¾(Ô]ü›’÷v–^„îDøÃöW~™ °¦eµÙú‡mi]ëP»ý‡Í`/Š²¾u}qº|^RÄâ,g+«ÜÜ\KŸk)/èïúŸC_|ÏR¾Ýïåaxó“8’t1C^7nfz¦DÄÈÎâpù 柇øþuü$¾ˆ/”ED˦L L–µ[Ȉ™B†@øŸšøÃþ¤Ù¹–‰ÚøЖX¥!@~(* {d+Ðï} ÆGùÍ‹Ñ™˜ûÏ‚þ}W¸LþÈ$ŽcGD2¸QÎìšüZ4 E@ê@èÀ¶À¸àA(ˆq`1à‚D €µ ”‚­`'¨u 4ƒ6ptcà48.Ë`ÜR0ž€)ð Ì@„…ÈR‡t CȲ…XäCP”%CBH@ë R¨ª†ê¡fè[è(tº C· Qhúz#0 ¦ÁZ°l³`O8Ž„ÁÉð28.‚·À•p|î„O×àX ?§€:¢‹0ÂFB‘x$ !«¤i@Ú¤¹ŠH‘§È[EE1PL” Ê…⢖¡V¡6£ªQP¨>ÔUÔ(j õMFk¢ÍÑÎèt,:‹.FW ›Ðè³èô8úƒ¡cŒ1ŽL&³³³ÓŽ9…ÆŒa¦±X¬:ÖëŠ År°bl1¶ {{{;Ž}ƒ#âtp¶8_\¡8áú"ãEy‹.,ÖXœ¾øøÅ%œ%Gщ1‰-‰ï9¡œÎôÒ€¥µK§¸lî.îžoo’ïÊ/çO$¹&•'=JvMÞž<™âžR‘òTÀT ž§ú§Ö¥¾N MÛŸö)=&½=—‘˜qTH¦ û2µ3ó2‡³Ì³Š³¤Ëœ—í\6% 5eCÙ‹²»Å4ÙÏÔ€ÄD²^2šã–S“ó&7:÷Hžrž0o`¹ÙòMË'ò}ó¿^ZÁ]Ñ[ [°¶`t¥çÊúUЪ¥«zWë¯.Z=¾Æo͵„µik(´.,/|¹.f]O‘VÑš¢±õ~ë[‹ŠEÅ76¸l¨ÛˆÚ(Ø8¸iMKx%K­K+Jßoæn¾ø•ÍW•_}Ú’´e°Ì¡lÏVÌVáÖëÛÜ·(W.Ï/Û²½scGÉŽ—;—ì¼PaWQ·‹°K²KZ\Ù]ePµµê}uJõHWM{­fí¦Ú×»y»¯ìñØÓV§UWZ÷n¯`ïÍz¿úΣ†Š}˜}9û6F7öÍúº¹I£©´éÃ~á~éˆ}ÍŽÍÍ-š-e­p«¤uò`ÂÁËßxÓÝÆl«o§·—‡$‡›øíõÃA‡{°Ž´}gø]mµ£¤ê\Þ9Õ•Ò%íŽë>x´·Ç¥§ã{Ëï÷Ó=Vs\åx٠‰¢ŸN柜>•uêééäÓc½Kz=s­/¼oðlÐÙóç|Ïé÷ì?yÞõü± ÎŽ^d]ìºäp©sÀ~ ãû:;‡‡º/;]îž7|âŠû•ÓW½¯ž»píÒÈü‘áëQ×oÞH¸!½É»ùèVú­ç·snÏÜYs}·äžÒ½Šûš÷~4ý±]ê =>ê=:ð`Áƒ;cܱ'?eÿô~¼è!ùaÅ„ÎDó#ÛGÇ&}'/?^øxüIÖ“™§Å?+ÿ\ûÌäÙw¿xü20;5þ\ôüÓ¯›_¨¿ØÿÒîeïtØôýW¯f^—¼Qsà-ëmÿ»˜w3¹ï±ï+?˜~èùôñOŸ~÷„óû endstream endobj 196 0 obj << /Length 247 /Filter /FlateDecode >> stream xÚUOËNÃ0¼û+ö˜ìzýJÌ (‘à"â‚8¸IZ"¥1ÄAü>nܨE{™Ý™ÙEà±P(VrRhV”ÐÉ7aJJkÁ\(<¯›Çc[O^b­]-éÅó®&›J`™5Â@½¿ärÉJ# ná=«úÃÏÔåT)ÌŠ›ü£~ÚTZ\m)ÃJ”1f‘‡ákðs’ý3—1×èU¶÷S² ŸºpÂ> /ExtGState << >>/ColorSpace << /sRGB 201 0 R >>>> /Length 5165 /Filter /FlateDecode >> stream xœµ\[s\Çq~ǯ8/©*Úã¹_ŠO”lÉRH‹1Y «@Uj ®‰UX“À©ü÷ô×=çœY„Ù ý K/fzfúÞ==Ç? vøuøëÑ¿Ò?¿»þóßß½<2£1fèÿýò»?ѯ9ÿ}túË`†wGvø‰þùõÈbÀðü¨¸±øÁº<†0\ Þ›1– Þ ÃË»#&ÐŒ® °Þ¾,#f¸RìX;$3¼ qÞŽ¡Ã2ÃÝše:,3¼ 9ÜwžeÈ·¯Åήï¡ØõÙåÑï¾wDÞW¡icûƒüKàÇZkìÝðêb8^…“áÕ¯GxŸžl]c8˜í³sãÁÚæó';WF{°´be—ó˜VVÚ;ÜÁÊišëhè/¡Ž!¿ôÙ) »î°c™Á~@ £5Ë»Î(Â4 ý€âG_—v<ÑÄmÐÀ~€l{Т§QG˜UG¯`Çœèd4© F¹ÂÏ.Dè$³“z6T®ÚDúáÙ91> ÒÎö.@¾>)ÒONuôáP¤›X4CpäSd5Ã6Á¾š±’ MœÀsâÙVY$⚺ÿ\m†.?˘ø¦áŽj0l -œèq B¼—gû«ÍµÆ¢ø±º!–ѳܜ—4:óOÃþ/ÃÍùfø¯“ÁÅáx}uB”£ÿÙ®/Ï6Ãædðf8þŸ»õöróîä—áÕOŸ¹^å“pg‘Óß VÁ2ú_"± q,iš?s 4?àœ7 ë¸äa:TØ„WXBïjÓµÓã×·»ýõf¸ØÜœïin‰pôßõhxs^=ùæDA¨RÆdO@1²È«x¢áø–xa;Ô^‹Ú"Ž#ïFž6 î› çåõêåörýa{¹=ÖgÛw´€Ó.à=,ìvMmïçW›ý„re¿ «ÝúìfÏûôß ý4§±~Ì…V#‘p°ØBûÏÖ RãLÛ’ÊJõñìì‚|6”F›˜—V„þåùö?·ß¡5¤¯0{ ác\ï¶g·»Çï‘l?äÍ‘~áÙŸ·o¿¶P³çõDZb™ðêÅíÕþýîãÍúâñ[·IL¬Ï3Ö7dÏvùùì'AÊ„-6Úž¿8ß_8ßÏ6pÞ©^õ2"WàNM«V˜ØËíÍy¯ÐO†§WïIêè§ÿýöÅÿu¿ß3%|bŠžC¦"ô8ÞáñÏg7ëw›³õå~9·UË+±&sГ۹/HÙéM¢ö«/÷×Û›ýnõ‰½¾Á±ò#¬fËcìÃëö¤Å)>Šl7mìÏ7»íÛíßQ“¢]Ç‘šZ?Y Zh½»ùêËX çâ)–b螯w¿îÏòI¸]%tÿá]?[ýÛɱÇݶ³õªxÁ‰,'ê'¤‹'$^>ÎÂ$CPìh%p"¬â6º}>ÕRÄš(ê^‘ҶͿÜÌväÎRwÿ¢>”CìFJ±´8vZîÅÕþöŸ¿ûwF&O»q¾¹lµ[ïÖ—e«kfÆO.ñÙê9‰öö á#RŠiÌË?[=Û|<{äÑ£øhFiØ~¼Þﾡüúdf§òlõÇý©ô—òœ(™=—*Bm[ýAê39üUp äBäœ÷¬òh{am•óŽ²&Óp~w{}³yœ²‘F{’}K¦sÖµ§×8s d,n¾Äm“;„¸"&wsÈúþ‘‘«…m&Yˆ4‰íìÃßG-Ia-+Ë­›„áé-…(ÜýMQŒ¡È³ˆálšñÃîöìq£\¸JP`T›+öÍZ\dòjá°%7Ç/Ⱦ>ýÓÓÿp¿hrÖ5ú¬™1x"Þ£¼1a ÖK¦QQ~›0T ÒÔÔH0$É:¨‘œâv>£PQQ¼•`>-(T¤ >phYQ-QD5ÅAPf‘:Z$ -Pöm9{'VYE‹0YœfEQ  o RØ8¥ŒA'Wž´TnvD4Uç y ÈÕÚ´èGÐœÃF«‹ì´ÝÏ1 ÕA`ñ•åV;dÑhPÔÈ.îûrF«Ôt“9®  –Á ×Ô—V…NËÒh+“³t(TZ†ÒFas‘-‹*-Cù:³|›E¾£FËX´ä¼,ÒuZf¸âãro{£F:q{š9Þ·Ý94ÂI’aÙì gÒÈfqcò’±/{HѬì¶vŠžt¤ÌBJ7ÆE&ªJ?P†rly}çÇtúA)vab[Q5úá²C Æ»s!UçË…¥i!gUéG6¸æv‘ŠªÓÄaI½Ý¬*¦FîoâòJÇ•õfªªÄEŪÊzÃlG–κȷ5©u,‘ÍM) ¥Ž”ÊŇ*ñ$3Å«—ÜáPÙoÓö{‹cN@#G$]Ùw84š%t'£SR‡B#¡2©fÛ¡ÐH¨‹,W2¡Ã¡QÚ@sg¾§¨NF™ó°xUkU!¥ö= ‡.zõo‘ÊtvÃ&R–Ö{a]‡C'¶:8Kí΢s)¡0omíó›T¼Åõ‹Ü阇Š·”ãѸ~YÅ[SD>|2Ù¬òFzbPìø’UˆÔe×.ŒµYe²g“î ÂŽ%ÉRÑ´X¾ sÙU²,MѺ"×5]ôæT,i¥$1éq¨hJ¼Nfë6µþ Óž¹ÜŠðP‹†f`¥djæBJvDžç›ÝæoÛ¶skµTÉIŠ+Wƒ‚ÕQÐÛÝœE-Új$+®í>øø4ód¸§oUY@GQˆ=e†‰¥û %g°’¹ÌÎ+YG;½ÓV¦Ø&eJA Ó¦5TœÆDÛ´\ýxùn¿Û¬Ü7~õn»¿ÜÈ%Ý7ƒ]½9¾ÙÌb²Zî±v¸‡½™niw×ÛÝíîÍÉ*¯>\Íš·[}3xÂðóª:ÝCÑ|¥0Wv~>˜¹8]}2´ÍûÕúlss÷ n¦Ï½·»ó”hz†Ò½%úÄ[¹éi ¿r#Má8@°0ìç÷ ;†YY-`ƒ˜é`. †˜áf¯]Pm ÒÔL+.ð¨hÇ&K8CÊ1ÌD†¥S‹ÐDù»´înìÇ(Mâ‰Û ž%£"–Ýñ+=ìÒÙ/*\ä1(ã4¼?J-&… 4€=—ã¢A °å”äÐ3 W˜;S'¦GŒ ¥å\¦H(6Ü’ùB‚`Ã¥B‹¼þñ†ûiÑë'ô3r’¬ÌwUš¡W?<»òÉ£”ØpÈ™ø:Œ`*•™]x|ñ|Ažøg‚³ôR¦ŒèpdGž `‰•R•ó!½?éçÄûI\÷ðÙB/ ¦ƒ ´“ìßabd˜ù‹îdþ»úãÆ ò„¢ï ã–CÛ/øKëÓ¶¿7Ò¥‘ ì8já®UÞ?u 7Á™Ï‡[0:OÎÈÂKA0^ñ¡õ3Ëß3ã3†é›³ìúÊ/ü¨`ƒü°}m‰Ü483l8ŒýØÌ·p G† ;bàÃú–C‚ÜL$`#púãÆ…Çg¡7zAÛy<¯¢ì‡[ØKù/âp~Ô‚™Þ©­Å´L¿Äãq'•fú Fçó‘/¾ Ì/–´è0}};Åã¸ÉMQªH‘å¡à<µŽ^ÄÃ`»µqÓ@é4œÇ“p9l–d•e»ˆm(NÀ̯!ˆDHPSj`dE%úy(vfÜ;ÌML?¨«=ýì3kQÅÜ$ÕíÄ7a»#RF´²“dš‘ׂˆãYæF®•BcÙ‘¥€BE¦;‰õ3²È×0†ÿjÙºŠÄÚ]¸0åIð‘çf°=¬«dùéPn¦Åq"DÑL\zyî7àãKÖ»'z+U&XcëÄR¦"Z m€m% ŒË[¹g¦‰Yñ ×Å[ðø¸xÆ—oƒõ¬_¼`6=Í[o+Þ °èˆx;œW[¼!`fuó– ›ÁæM§°x[ÐßÖÅ-o ~ñÓåæÍÁ_Ù¯x{ÀEÌGÁ/ÑäE¼­D./ÑG>'0G#M+V£ˆ®•è…ÃH¶ìW¢>kI‹† b%ZrµÉc‹¦0,uÑ´¬ÌѦÙ.ZC£½D«ÍA…e~‹öŒï;EƒÆÉù¦hÑ´èlŠ&ÑŽ'°D›&MѦD£´ Á×¢USÛÓñÍJk‰vIÍ‚À Ó6«Œoï¢èÇMÛ,&hŠ¶m[?EãΧhjÚGó®yÿ)Úw¡ÑoÊRÛïA¶°;’—íÁñÃϬ¿øã"¡p1§2™(±¤6œÊÜ2ÃË;bàÛãyÈ /C(¡…‰š‡Ìp7D‚©eÈ/C·×m÷+~<$döaý<ègEêíY+çGÅ¥/iÉá÷G Üš7:Vøp¾býŒºs<œïæõ'¡$Gñµ{˜žö-Žß ¡H­Ìp?v™dzÒàƒ!‘“W^†ÜvsªÞ†4ø`ˆloÒm·§HG†ƒoXxnäÂýsû¢ÂŠh¢ù† p¦L½„@óI† Åü­˜E?ëç÷ßðxÌú¾pØÓ„â®LL_òÈÉ|É£Áó§;&¦Üý”GûpÇ?òc‘B 2ÉÝÇ<ží×ﶗïŸóˆ5~Ñç5‚)çûŸaù¿ÜÌQŒP­<ï‘Bæks®¸A X!NÓÍM¿_¤>ef9(wy~‚Šï…ÈÆ .íix(“D}M¦@ñѤ€ì>qØU S(£§á_YŽpïg>åk<Èxºu漶 &FÏc¯Òæé&FÊ )вɶ5Œ@ÁĈŽ$=çé LÄH+¸å &‚„!áM`{â &F4š‘ÑÁk÷0#Ð0í0݋˲Aá&‹b¤Ü¾)@6á¹™=ÌÓUR@é ¾<¹yoB ‘‚Èi´…ã¶3 ÷ò<Î4R@¾ˆ4ýœÁÍ4Q*Õ¤ÊÖM_Èym‚ ‰H€Šô¡]‚a¾† ÉqvŒÕ²`P°!ymûn >$ÜWDn¢š•É©‘<—>¸½7.4œ°Ü¿oÑôÝÑA£5‰0”é…7cPhd*\4Cžß>×öÚiTóîáS„~ž¯’†Ä•hílXN)Sàìßy?óÒé´2¡ÞbxícÐHCáw¢øì]kÉa *½$«jùËŠí-=cPHC¶r!ˆêèdœŠ‚—Ùf)çÆ© ó5¼Ì†ŸU¢ÆëÜ‚AÃKäì…#I³œAÃËL1ZFð1‘î ^âíåwë¼e ^ViÐ2~zмL…ch.Œu§PhvöŽKG9Ï”¬I# x°Pù¦Ç›y¾JœtSâVÆ-ÒÉSD¹×ÉËTÒ€ëÇwNSà i@›Š\bMZ ª:E’Û©ù­cÐh¶“K`ºŽ³4|YeíNý¢%Áõ‹û+k}ÁòµEÊ5hP6œuB|º*àÑÿAäÁý endstream endobj 203 0 obj << /Alternate /DeviceRGB /N 3 /Length 2596 /Filter /FlateDecode >> stream xœ–wTSهϽ7½P’Š”ÐkhRH ½H‘.*1 JÀ"6DTpDQ‘¦2(à€£C‘±"Š…Q±ëDÔqp–Id­ß¼yïÍ›ß÷~kŸ½ÏÝgï}ÖºüƒÂLX € ¡Xáçň‹g` ðlàp³³BøF™|ØŒl™ø½º ùû*Ó?ŒÁÿŸ”¹Y"1P˜ŒçòøÙ\É8=Wœ%·Oɘ¶4MÎ0JÎ"Y‚2V“sò,[|ö™e9ó2„<ËsÎâeðäÜ'ã9¾Œ‘`çø¹2¾&cƒtI†@Æoä±|N6(’Ü.æsSdl-c’(2‚-ãyàHÉ_ðÒ/XÌÏËÅÎÌZ.$§ˆ&\S†“‹áÏÏMç‹ÅÌ07#â1Ø™YárfÏüYym²";Ø8980m-m¾(Ô]ü›’÷v–^„îDøÃöW~™ °¦eµÙú‡mi]ëP»ý‡Í`/Š²¾u}qº|^RÄâ,g+«ÜÜ\KŸk)/èïúŸC_|ÏR¾Ýïåaxó“8’t1C^7nfz¦DÄÈÎâpù 柇øþuü$¾ˆ/”ED˦L L–µ[Ȉ™B†@øŸšøÃþ¤Ù¹–‰ÚøЖX¥!@~(* {d+Ðï} ÆGùÍ‹Ñ™˜ûÏ‚þ}W¸LþÈ$ŽcGD2¸QÎìšüZ4 E@ê@èÀ¶À¸àA(ˆq`1à‚D €µ ”‚­`'¨u 4ƒ6ptcà48.Ë`ÜR0ž€)ð Ì@„…ÈR‡t CȲ…XäCP”%CBH@ë R¨ª†ê¡fè[è(tº C· Qhúz#0 ¦ÁZ°l³`O8Ž„ÁÉð28.‚·À•p|î„O×àX ?§€:¢‹0ÂFB‘x$ !«¤i@Ú¤¹ŠH‘§È[EE1PL” Ê…⢖¡V¡6£ªQP¨>ÔUÔ(j õMFk¢ÍÑÎèt,:‹.FW ›Ðè³èô8úƒ¡cŒ1ŽL&³³³ÓŽ9…ÆŒa¦±X¬:ÖëŠ År°bl1¶ {{{;Ž}ƒ#âtp¶8_\¡8áú"ãEy‹.,ÖXœ¾øøÅ%œ%Gщ1‰-‰ï9¡œÎôÒ€¥µK§¸lî.îžoo’ïÊ/çO$¹&•'=JvMÞž<™âžR‘òTÀT ž§ú§Ö¥¾N MÛŸö)=&½=—‘˜qTH¦ û2µ3ó2‡³Ì³Š³¤Ëœ—í\6% 5eCÙ‹²»Å4ÙÏÔ€ÄD²^2šã–S“ó&7:÷Hžrž0o`¹ÙòMË'ò}ó¿^ZÁ]Ñ[ [°¶`t¥çÊúUЪ¥«zWë¯.Z=¾Æo͵„µik(´.,/|¹.f]O‘VÑš¢±õ~ë[‹ŠEÅ76¸l¨ÛˆÚ(Ø8¸iMKx%K­K+Jßoæn¾ø•ÍW•_}Ú’´e°Ì¡lÏVÌVáÖëÛÜ·(W.Ï/Û²½scGÉŽ—;—ì¼PaWQ·‹°K²KZ\Ù]ePµµê}uJõHWM{­fí¦Ú×»y»¯ìñØÓV§UWZ÷n¯`ïÍz¿úΣ†Š}˜}9û6F7öÍúº¹I£©´éÃ~á~éˆ}ÍŽÍÍ-š-e­p«¤uò`ÂÁËßxÓÝÆl«o§·—‡$‡›øíõÃA‡{°Ž´}gø]mµ£¤ê\Þ9Õ•Ò%íŽë>x´·Ç¥§ã{Ëï÷Ó=Vs\åx٠‰¢ŸN柜>•uêééäÓc½Kz=s­/¼oðlÐÙóç|Ïé÷ì?yÞõü± ÎŽ^d]ìºäp©sÀ~ ãû:;‡‡º/;]îž7|âŠû•ÓW½¯ž»píÒÈü‘áëQ×oÞH¸!½É»ùèVú­ç·snÏÜYs}·äžÒ½Šûš÷~4ý±]ê =>ê=:ð`Áƒ;cܱ'?eÿô~¼è!ùaÅ„ÎDó#ÛGÇ&}'/?^øxüIÖ“™§Å?+ÿ\ûÌäÙw¿xü20;5þ\ôüÓ¯›_¨¿ØÿÒîeïtØôýW¯f^—¼Qsà-ëmÿ»˜w3¹ï±ï+?˜~èùôñOŸ~÷„óû endstream endobj 206 0 obj << /Length 220 /Filter /FlateDecode >> stream xÚUPÁnà ½ó>&ì&½uÝ"m§Tâ¶õ0¥ÉZi]ši¿?R)•…x¶ß³ èhœ*µC¥rÆ@{?BQÅ+œ 8µpÖÍ…âõ\Âs»÷–¼Ï”‹¡O^µ]A¥*6 ¾$RÌƲ"íÀà=«O_¿C—KkmV®sée›”^¾Ã˜PèÓ}mC⺠×6Õú0$Ð4ÛYØÝçØñï4ó½+j2 3Æ¢bm£ÿÉÆmWÓ^?4éx0 ÜËŠâ×ÜëXiæ¤Ã‰*^¼øj¥O endstream endobj 179 0 obj << /Type /XObject /Subtype /Form /FormType 1 /PTEX.FileName (/home/biocbuild/bbs-3.20-bioc/tmpdir/RtmpCCJHlc/Rbuild86c017e3fbf7/pcaMethods/vignettes/pcaMethods-016.pdf) /PTEX.PageNumber 1 /PTEX.InfoDict 208 0 R /BBox [0 0 576 576] /Resources << /ProcSet [ /PDF /Text ] /Font << /F2 209 0 R>> /ExtGState << >>/ColorSpace << /sRGB 210 0 R >>>> /Length 32026 /Filter /FlateDecode >> stream xœ´}ËŽmIRå<¾"†0¨(?=© !šÊ” zr¡Ò’ ü~o·µ–™Ÿ H2[ô 3®™s·½·?Ìíåùý¯ßóûïßÿííïÞÿí½®TÞë̵¾ç1>ÆПÿç÷¿ÿÃùÕÛ¯ÿø»¿úÍûoxK)¥÷ûÿ?üöoììïÿñöÿøžÞÿé-¿ÿõóßïßòùÁûß¼µò1çÓÅøHëýó-ïý±‹àï‚{­¥>0~î࿾ÿð6éõùž×üXÕòþŸ!p~셬瑮Œäœ?ö¸:|ýäùwé×O_?壷ë'‚¯Ÿ¼H=Müä7?òûãŸxãüö‡·_ÿey>âÿò¾>ˆÇÿ ìãcå÷ÞÚÇhï?~¾ÿٯƟ¿ÿøû·¿øÑZþ)ÖU>ò+kû¹¬9Í–^xËÏæ­ý Ë7ýlÖ‘>ê ëÏïuõñò´þ°³û¡ìçqL–ü‘>Ñ—p¼~Ðò>_U? xÿ õ2ãïÌôÑGüàýƒÝ?V¼~@±õƒû)î7ïåWñ²ê3‘Æó\é#ןJ?ÁÛrÿhå'ÆÒOñ¶geÉ?1–~Šw¬úcé§Xwýõ¿K?ÁÚó:‹ÙËX:Ëì˘ûüÈ]¸Ìþ‚9ŸÇë`œS>ÜúÒûþß¿}~ÿ§àO7°ÒG{¾î†wþþ»ÿSÞÿ×ó“9ÿó³”´O?òYbc)©…öÿçÆBA|c!ìK}:íË7ÿÕ'H)éžæµ×^îä?àC¼О &)Áû³~¬?x¯$éYÆ®•àýƒ2>vLs‚÷ ¶ÿàzŠ_0¶~bl—g˜hø¥ó¢<ëZªÿO³±<ºÉHÿO³±ìõ‘Çëlü¹K^iÿÒ¯öüQ¶±®_6oþ'fƒF¢fÃ5©¯hÂHÑŒ8?)ÏúõÈ~¸±>?,íyÈg8œO«û#5ƒ'zÁ¾fð³—–ž𳟕畜Mé팡GÙ)ýv®ùëYèÜ?fzàgÙ(„Ï|éý£WƒÛyœò¨8 ý×vÚ›G=p9£¾<‹`A{ù0–¾¡l–l‹Ü·ñ?ïâa,1íùžïQ Ñzä}žçQãÐ_zv³çýŒz^ËÛi¸Œv†÷Ÿþ·ÁËúåtø»íÚ¶&yŸW^¶Áóc>òŒù‘²ÁÞ÷'àçyülfP–ËÑlËàûÏÏóÀÏ2ûÀk}ôçý=ZB±ß?Šî‘÷èÍúæT«—£=xƒþìôg¹zž×øç<„ƒ>û™gý²÷‘Ï?:Ö3ÐM¹(³±ÿGð¹ †|σ?»K™öYÜÎè=pí?ßk< ~>TÆïm<åçÃ#ï8b¸Ù÷z`¼Ÿn‹Å»ÑÏ@OãoÃÆÇÛxÌ­-ÿÀp>Jl™úÏbzÆë^‡çÛ÷|´øÉßÛû|–€\ÑžÉSžç“Áø¾Gûþ¥ž}`<ï³µìn0Þÿ3ßZ3ïó™¤CžLþÎöžñžªÁÝäÆóÀûËFÆsÉö=ðü©žý=ãÕÞOÃ|y†kÁv5Îã>£ÓÞVÅìzçl68æyøglvŒ;¹¡ÙlhÍóèÏȬÝÀ|z^˜øG=ê—6¹Žœa¹ ´añŒÊçqÎ,°Y1×Ùh/á¼\Ì™z„|žõ<âà+}„=R=3p©ÎHßæÓÑÓ›6_Mg8óªØôgx¬fÏì··õ4y&ËÚä{Ú°¹Ÿñ6¦m)©ç‰S¶•¦[Ëù¬xga2©ž_-¬[óüø¬GXÖÖ˜‡­zû|µŽAÐ;>Êô°FÚxÀÒl -Gª§g›8+l-Ïkæ…M<ããYŸÛãiòHõÌ.xfW[ø(gÕm¶Øb(ðy±ëúqÇ×WS #‘µŠWG1žMjçò™m{Å#< }P>`³S‡?~µ±è/ç™™5Ç«{ö²–ãÅ>óÔ6 ¾öÊõ„¥6l?üdÏ$¶±Áú€¶Ùñs×b;¨Ã3ÃíÕq¨Ô„©ÁT6¾ ‡Y±åËá³ØØà-ÃŒø¨t-†÷–kð—~vŸ~b?æÄy@ûÜœVϲbë'ÝÚ¤ã”,å,’>a0͘Î%ãës²? ½ .G÷m×JñÀöU´’<'ÜU®•æ[¿V¢Ì7¢•*/­\XÉ+W:( ±æŃ+å£` ?®¤Gé×JûÀh+q6+E¬Ôl/N+y†¾ä+}ž˜VÚ ò¤<Ü)¼¯ä´×¯æôׯèÈsïTlÓG;ªØéŽÂU¯0si§Ì\P´“žh¿vÚç}¿ìÄ<êµS?¯ï“;ùC3àNÿÀËÞ5’¡iISx`¼jÏ`"šÆSó€&Rª4h*ÏÀEÿÔdòQÓy†=¾5¡"M€šÒó™û­I}jZÅÔóÐÄxB“ƒ¦öÌÐÍšÜ3Ÿ­?iz•k’4Á^ CS|VÓl¤I>èCóÄóÍ~Þ˜~Þб;flÙù@˜M€¬¿g«9ïOôboW¼„š?¹0õtá¼åÈãmrõ ºäï‹Ü|–Œ¹ó<û3å>oŒ-š™c«™Öža"dýeSg>±Ó“—µ_c#ÏyÓÙ‘¼mƒðköM:å"ï-·¾K=Bp\~Þ˜qV곎žyÜLcÏGg‚0j†­¢ÛAÀy YÛúõêþäâÅ“«mƒðköM:å"ï-7Ÿ¥™•çôÞø]„™çÏÙcò6},¨ØŠ cºb^mG/!k[¿Þxºc€¯´½mƒðköM:å"ï-·¾ U«~Ö×Ïcg¯óügÿÖ3›ÍÍV_@xwÝæ èv s^BÖ¶¾‹aÚY ׶`oÛ o!#=”ÀñEn>Ë£VhØÏyãóÆô£š½5'X!4í«Âü\62D_g w^BÖ¶~½0g ÞrtRoÛ üš}“N¹È{Ë­1Ölõh¦M~Þ˜yÎiÚ žuì}ã ¾SÐÍ0ἄ¬m­²†17†ó¶£{Û}£$Ö7锋¼·Üz–r6Ë#Ëñ™Ü˜}~u µítsæv3£!õwf³èv(p^BÖ6~MŒiâmv2WÛ€¾Q’µƒN¹È{Ë­5y˜þSí÷Ÿ7fÅᬢç[>û}9вÕÖÍ ýô£Æ çÒ ÓOÎÛŽ ám¤Üú&r‘÷–[ÏbJÉÑE*Ç1g¡¨Ö_¶ª?ºì‘¥7Ag Ñ¡I’—ÚƯ‰1MX¼gݘÞ6 ½§>œ.IÁû"÷7×a2ö«gú¼1ùhv6¦±ßüùÀ{Â8XÓ¤o;S‰—PòU˜G‚³““÷èxÝÛäsqr‰÷–›Ï’ìðpp›ë˜0ù Öí è8¶†áÝ%ŒiÐ öœKÉ+ÈÚƯyä9+yè|S¶ è%9}‹¹Èû"7Ÿå(ÊØý&ǘ0ÛÖ¾âl?ûÇQ›» ëç7Ñ7´_òŠ7-ŒÙgÄÛL7WÛ€¾Q®F§\ä½å泜CÖ½Yð,˜RŸÏ!kß™qϑô áÖ1I¤ ×c¬t^BÍ5,aêj΋ýCmÇþ¢¾A—\à}‘[û~;æ Óû¢#Ì8jàÙ›V ªMAØ›í€ìt?eÌ ²¶ñkbÚ9u:¯!½mƒ¤S´tÊEÞ[îx–Œ3Ò*þ,†©çWÚÔ’²µ¾» i%6ŸH·£½ó²¶õä“s±¯S½mƒ¤[ZߤS.òÞrãYžÆŸ­·ÇÔd!òù’Ïkœ@tÊ>o‚Àék˜M‹ˆ‡¡ ñ °Ví§@A¥4à»e¥øÍŒ_güaábÅyÓ©5†½q"ÀŽL Pq–"k•OʇkÕù,xA­ð ï/§Rðݲj㨦 43Þsåc,Á¶wŽM©ãXdw¼,ÑÝa¼„¬mýzañ?j yŸ5÷D°m@ø5úr‰÷–;6Ž‚¥>Uß8 œ”Ϻx–óhe[êí=¤­À^%éÆwòBÛúõ¢ºÜ·™1_Ñ|Oì[tJJÞ[nmÊ×ò£0ûœ‡ŽÑ³Òr~ ‰#§APŒ@¤Ã$)^BËŒÂL;ô‘·š]Rm¯ѷèÝ•ÙòEn.PÃN=ÙÀ‚îÁís·XZ’Ùx9‹e0N‡¬aüš˜~ ¯ÎÛŽ¦¯¦ àF³–S)’1¾¬ižÌÖß«ŸÙ…ivî.¦›C°Ù÷·Sow]Fw§W|/òª~f¦ØsŠ7Û‚ª¶³ŸÙÕ7锋¼·Üñ,Ç/sìößÅ0æî:=œ…äYOLÃ4g!Iw–AѬøä%ÔÝÒ#L;{¡ó6,wl»ÅrȾI§\ä½åæ’•`&.Ç·÷y!ª©4“²!èöç&ûó±Î6§Ò¶>ÖªýÔÏ#>s2«Uì§ìT/åWYµ«ãq«Ù>o ”ÊTàRù쥶Ïä;o : "â%dmë×S6&S“ûåCÙ6 ü}‹Î³5Cî0.,2½¸éTG¾dæËA¿w2̨‚d 9èÍœeℶe82Ì:±x›éïjL§4’’÷–[æÆ ï¿ÍžÏc¾×㘀ÎÁ{ØYŒL‚vszGy YÛø50 oòIõ¶É1a†Ò!—xo¹µ•t«Ì õycÌ_™a72óŽÒmÒ1½µ oô ²¶¥m˜OFvÞsZÏÞ6 ßxVÐ!—xo¹}ºõiÕ§{r=µ'©±ÅæþÊlbó¶’ZÍÕK>KÖb!ÌQM¾c÷ÎjÀ7Àà†™]ðݲBüã°í¶'…0ýy.„‰´³QM3ƒ8- ¼&b™ñŠ|ª6M!Ê1K‹Ïf•ZõI§>A¥4Æ÷"+ÄÇ ¡Öòy!LqéõoGzGƒøöÏ‚Aj…S|¬U¾}º‹ž¡C¾£k4µ €o¿!$ñ½È*[€e¥Ÿ£ùçŠüL€v7»FJ‚déØ+è>5ò²¶ñë­è›£MŠ×|»Þ¶A²<¤tÊEÞ[n-PIöë¦*Éî~|M#Éî~Zx–Zâ¤ë"€ÊéUv÷zAK‡ZÇL;.’·áXÁ¶[;Ø·èIvwã½åûî9ŸYsû®a¶yjáÚ°ˆÖbÙ’ÉbKÝ|þ»wð²¶õëM/ù^ †ò¶‡[¶Õ7锋¼·Üx–±íÑgÑ~.D={í´(u Í|€l:›‹Å^2©Å™(ÚÉ…Èg¦ˆÏLÕjÒ6®> !iŒïEV®¯Ù,ó˜û¼ù<>uã Ò‚PGb(Þ¡ví>I[ ݆ùž)ñ,ól×:œ iÈw˪ž`¿›g•þ¼1js"|:FÒHægìY¼’m½!ž‚¼MÑÚ¼„1[‰óZ”·=Ü ¡¾I§\ä½åæHjf">Á2˜BX¢™-¶í,Ô”€°r$@53®ø,L„°èYñ£•«Õ!ÿ–ú•Ò€ï–U‡ŒæÇÚx„±PÕmnaÙ¬¶\$ÛÅšAŸ¦ ˆ—µ_“ÍqXÜñÐz´Ö-õM:å"ï-7gES„u⬠¢ãá-~w³DÏhÁhÛ©Ð!k•ÖØÂù¸ÏÔjÖAW}‚JiÀwËÊmÛ’O0î‚Ò$Ä:oéX$²ÂÕÛB¬ü–iïl£Sˆžén|¦t^!Æy·âçeªÕ!…W}‚JiÀwË*ý•.Û?oÌ<ÊÄÑße£Ã)£Ar—ôêö¿ ²¶u 1L7Éx›NÔ¶AßxÂ)ÍéÕ]F¥}‘;žÅL'õlUŸ7¦ƒûi¯!¨cfëÏÜ'Ý+ætÚVÉK¨ÆÆHL±£¹xËÙ®¼í#û]r÷En-¶ ám–Aöyc,ŽMá‘ÍŽ÷Ư«ìŠ_ƒ±œWµ­_/(XÇè$ÞvÞ¶AR¾rvºäï‹Ü|– öÑÙø,ÄÈ#Ï®ÚÏΉq7 ødým<7z7YÅ më×´y—y»ù:Ô6 ü}‹^´]ÎõEn­¼SmÓ­ëÂ,­¥ Ђ ¹%A²1×åôšàį ÖubÌë¼k)ÛîaØbß]ëô(Á{Ë&ÝPƒ¹Ý¤»œóÖ,~-–Ø@Hf·ë¤[¸E“Ú–aË08’·ò­¡m@2»ÍâtI Þ¹õ,êbu³›0ssÂ8 ÎÍ>¨ôËìv‚BRзB÷E€•0ÅFx-¦ÌÛ.~XQߤS.òÞrG(&–úòPÌŒU¸#äqó@1l~î*H³yç „(WP8¤…¦–‰×ÜÄÞöåÞ ª3'ïVÐÒH_äŽ0¹™¤ô}^ìŽb×éw„Ø– wi ¤Ã%9×ËxU1‰Yv8"o›r€·%H iOA§¤ä½å¦©}k!¤Dˆ~¦åÉ?Äfwl«Ç:¾Éz¾FЋù^ÀêPsÿ‡0Õ|â-fºQÛE~xum eœ/"¿.6sÝ‹gÓÊK{¼œH¾jÙÈl€ý}’JiÀwË©/ 먑úÂÓUE ‡bF!M×9’&hºÎ 8ÜÁ;4j¤¾l«Wðvhšl»G'ûîÒb[ Þ[n>˱#apŠcÖÿ“žØÕwtÉ`™æxV¥9/V¼„ŠO`a,BÙym¬yÛÉÏóê›tÊEÞ[ní"–IïóÆX®îY­éî?îŠ6 ¬ô8…‹Þ¤½Ö ÊîKy«¥uªm@߸ŸåtÊEÞ[î0jgDqåâFí ÿdNm@Ç Ôa'ôMºó:BKÄK(’ä„é‘EzÔñ5«íæ¾hõM:å"ï-·¾Ër;’¾Ë’ýëxÊë–ýë|Ù–dÿZ*3ÎK¤Ù¿ÖUÓâ-æ)WÛÅÃßÕ7è’‹AÙ·Üñ]èÕ_Õ¿K‡e o†m¦"ô¸(-ÈC“‘~&zÃ2MÞæ‹võÙœ°}®¼H›QÛ‘V£¾IOŠ8Xý‹ÜÒìáæªÄóyc,õèâtfO¬¥,HÖ†½~2”“ó ²¶e9Y''ùK¼Ýœjj»Ë¯ä}“N¹È{Ë)I _K‘’4âtÎÃÏ90ÝÒVϵqAW/%xÝaœb«§&íàmJPbp~‰­¾ S.òÞrÇ8ñϘû§þ‰Ù~EÉt"¤1Vgй…’·)yQ_Qv‘·!°‹m·øŠì[tÊEÞ[î8Õw$»½˜˜{qEA ;¥Yj¤S=­¸ »#-¼ã‚â<(LÍ—¼öà¦UÖíÅì›tÊEÞ[n­cS²(ÖI&€P ²¬-‹9Þ‚dkgBJ…ÕŸ.ô~A=bˆiŠWJPÕg‹¶«Û Õ7é” ¼/rk¾L¬mÓ¿‹0Ž{œâ=I>ó¶^w¤Õ“~´ýἂf|b<*^3ˆ{ÛÝ¿‹ú&}Þ§/rëY¶ŽMMsËÑÚ¸öáØpRô:ei¯djŒÑ«¼­Ô=ðR˜fI¥âµèXo»úÞ¥¾A—\à}‘;Ž_L‡q+XW-Y’©-Lãq+ØRPoÁ1€i<í‚zXÁ¶ÒxZÞŠ¯È¶«ân¼oÒ»ÒxhA ¹¥[fû}/žŽ(Œ9üÆWàÿ=šÃ©f4I[,=èý,ÆÎK¨øaM«9ë¼V×ÈÛN~XSߤS.òÞrdzîŸÅŸ…¹'Çâ|±Ü÷“õ§´äz²²B’IÇ|—uAÙíLÂX,³ó&‹ðTÛÉíLê›tÊEÞ[n=‹¿YdmÍ kh‡GóÔðê‚ÐßD17Ò-$Ñy U·¶c¥Ìœ7ÛlVÛÙ­-ê›tÊEÞ[nYÁ¶'r¾3Öfç¬LÇw:³'fwô’ÝOHû‚†'å c'Kç…櫶{Œß$cú¡K.ð¾È{e¦U­û^É2á%›‰ =ÙIvrãÎI^B+Êvƒ„\ò6 QÛ€´FîtÊEÞ[î+(ѱ©»3aL8UtÌßÅÀ™*È} 9è Fò67|tr¦ó¶õ¼(ü§ÄnÔ{Ð)yo¹#11Ñb{:")…ï£4Áµº ­›,¶˜`Öb t¹ äÚ"1ÇZš·Y9&µ È¿ù zS*nÍ_äö`Ô# «C±Î6”Q;+Ì 9’4’•à€1Ä‚>Kêˆf'_5û7[p~Ê>I¥4Æ÷"kDs6|ú¡êĸÇ&’èàŸê n–ª(SsΠo,¬äÝÊ;ô okáÃyÛÆjɈ³º8zV¹È{Ë}eãèZå²ß:(T¸Ý —Œ á&H&Eˆfd—àu{¢alSs^› Þ¶›Œ½oÐ%x_äuÒªó¬"Ÿ7fá@5Žµím´… ³]AÄKý¼´î¼‚f(îÄ „À× rxÛÉ c}“N¹È{Ëá: *Ÿ²~„¨¶‘¤‚&„ÔŒ)H<‘¤ÐzðV…ò¸Û§*wÚªm·¢Û®Q©#É$ÑfÈÞ¹Ãᵹ쫲 1õ1º“îqV÷!È]X-è NòZ~ðfZÕ ò6ÔŽaÛ€¾X×N¹È{˪qölüÏ3UˆTX#ˆTJD[³‚JA@ÐLÁK(rû…±C“ó6ÕÁa•î»½«‚J/_ä–ÚÒtD’¹B$Ž.‹ûYK†ÅÊ‚}~ÂBúÔ&‘.hú$Ì@* y»™#Ôv÷ H}“ÞttKû‹Üz–åYT'…Áqpx¤¶™ˆ“VQ7!g˜ôIÏroP%…ÁqP¼]&bºú´R¨oÐ%x_äŽ16ÑŸ¢Ò…±2lçËw@£¢WA5=}ãàMÞ­@¹û ÍÇÈ%ÞŽúal»ßŽ€ÞƒN¹È{Ë­ïR_ÌÄÇXi£ãÄ¥ôg•M’Ë$6Xµt¦4“—ÐôœMa°ÊŠ× !yÛÝM!ê›ô¢ãzN_äŽ}¿²>Fìû;¿Döi{³ÏV½i†N|‰µ‚w+neÇwYÙ#x³¾RN‚´®ätÊEÞ[n¿¦ŽH2í c>„s`Ê€Î5·2NeÚ_|$}%Qç ªnÚžâÍ8Doù UÿN}“>utËí‹Ü|–U¥¬* ‚˜nõ,U}òF‹]ôAHÝ5d¼4úTXL mã×SÙ×g ’—f¶}™uØ·è””¼·ÜtÜ#>îÔ•Åg¢/ÔBÍ€s[Ì Ì€é»PØ µnO@– *„•€%_±*7l€µÊk@¥4à»eÓã`Ç8=,ÖìÜŽC‚éÎ:5¦;CX"2Uç´<ŽW˜©(F>ëìÞ6 WóGЇü }|‘;¬Æ™våáVãv¶ÊÒUŒ&_^ŠM–ÝíåÚ`žÅy•P¤=6Ö¢U0¼<óY…Åg%(ÕjÒóªOCHã{‘õJöÀ‡ÎÃO\ ƒ!c˜@› YN\y½()_Pòýš˜~±œ·™2¨¶iB1ʨ.—‹¼/rÇÒ5PGÂ^ÄL7«ã\,“û¤)ÒºÓG‚› „® b˜ä@^O€hS¦HÛA§\ä½å ²ãÜ«@|aPSíÑâ*ÞÝ1>Z^‚¤AÖô¡ÀÔyAÃ#¢„éˆPêr ¦mwOðQߤÅfï-÷•;Ë Ù¹³n"DjÂpÇ+}å 3™¼^ˆ`…t À*Ák•¼íáõÑÔ7锋¼·Ü_ d@¦˜Þ¦B†g*¨~îP¡Ò»êåÛ÷Òg¨PÁÆ7e¡‚ßœ… H÷Òu}‘[V0¦ö°‚c…´­ ª¿+WAÔ „ ²NôŽ° òên¦Á’EÞ +WÕæ$+˜ú&r‘÷–;’p-–‘„Óh^EËb-ª,/£ gAmÒk‚Ê ^AËË[3M!oEž:Û®‘ÇξE§\ä½åŽïbAM•ßÓõ% ±‘ØÉŽß…_bîà%Ô¢ò(1¶ñ8oÑW²¶ÃÕ§¾‹¾„ÉEÞ[î+ ‚îpF°ì´ÎÒQ¿„á0 "ËÁ ¶tAévFÐÂAgWÅ2½í«º­*H§}ËrÇ:F;Ù(¾ŽÑVÆD™ée®ÙØbeâŒm<™÷à%¬„ØÈÛ͵¨¶»'"¨oÒ“*'ŒöEîH(Z¬˜¥ïâeÛê0쥡ÌI_‚´Öôt„Š·*$|DÀ;ëÜÛUxäíÇ\ìmäÊgrºäï‹ÜWú “~®ô¬}ƒé+Xû«^6AîŽ+AÒ,Fv¨E8a÷*£;o›ð<4ÕLuÏDÕApõ4’YF´­€Š†•#…É ô)¬„Ïp"HÎÂ’‚^UÂq^—®¦mÔõÄÝ“†Ö=±w‡íŸ‰¿{¹\ä}‘[ß%«°’o…©,Ó¨±Ü+v·¨ÁÜìÀ¤£€žx •°ÿMT¼ÐDÕvû_QUâÜB.òÞr¿ñ»•s¼ø'&µÝ‚q¹2µ&Èø%èCõ2K (ŠG “á&%¯ÅîxÛÉÝüê›ô¢Ñ^ƹcîoV?ˆ{6í‘Ð{&Ç–¹®2:Š3Úc-|ÈŠ·Ø˜,g™/(<ßÂðÉÉÛ½pD¤1¶‡Ó%=ß·Ü‘ Â{¹tFá· y¯ÝÊè>¦UNôI z×T↦®úŠæ’!¯mÿÞvr[‚ú&r‘÷–;ÆX¦VQ|Œežýð-Yfã8x“& àUAVUù§€®ó~÷`û¼<ÏW/ÁqÝ–Vß/{@®Á{ËëXÇóéNa*œ6qg²RkäëX :ogïuW[X2VEÞÞÆä¨Ù¢í{×V2âl!x_äŽuŒeáu~!†¾ì‰{»Ý`í!ÚÙßôôSñÀÊ´rðVÙšõ¦…A,yïö¹®×‚î'fã½å=™éNr ƒr³ w¬6”›íIçõ|%® gékõ‚f¸Ž³gUÏàíÐJ’ qyVõöŒír÷EnÖ=cÐYñr¥DØʧµvl ªÚ¼\iÂÖÛ®b¦àP¼\)ÙÊ•’/ÝI“¼øêÔ©k¯²^Þb$ë{ðáÔ¥"ê0e?!CqwAî-®NoI—” Ê|HLR!¯:Gx¶ ȽÅ3èÓS³ç¹C\¬ÀU]¤#„ÙT†b›©·8ºöt¿´r\P“ØTÝïу·ÂäÕ¥(¯0ÒRYffåÊÁ{Ë­Š.Mù`r¹c<žû1è®Ê¸iƒylÚð’ç`})­^Põ O˜‚2‚·ãC¶ÝoÃ&3=uÁaÞ[î¢æ‚¤üuap=+*•&dU?šºit9½³Z²^;ÏgWöK2kðZuo»øìRߤS.ð¾È-÷QÆ5Àqu£0Èâ.0¿d?–jkd‰ðۇѮµ&½ãRcòŠ«…Áõ‹âåÕYl;®ÖRߤS.òÞr³(+,l¹Ëê*„߆ßò,¸¬J½RÞ„_TT8ŸO«Â"lÅgFUµê9UêÓ’Æø^de\[U@â€qR+°Ô4 ÐZ] ãÍ×ð¢]ó¼…n•é(„U'_µ`l€ý´ù%sÉ¥ß-+ÅÇOåç…0gåâmΖä´P~À7<)‹ý* ÷’ÀÔÂ#„å7‰Ï¬ jµë-õ *¥ß-ëÿïÞÿí7ó¾ã>Utüùóïÿüþ÷ï°áJ"üÿôÆПøÑÛ¯ÿø»¿úÍûox;JRz¿ÿÿÃoÿöÁ>½þÇÛ?üã{zÿ§·üþ×Ï¿˦QýÍÛ-Ƨ]Ëý¹~È?ßùcÿõý‡·a¹‘ä·PÌçüBþ÷ÿD„ç1ò2ϙȀ.ò¹´ð9âyóïT«íä?xÿפùÞ?¸Eö'¸~ð›ù†¿ýñO¼á?~ûÃÛ¯ÿòyGï?þË»g’ÿÏÀŽ*¶V‘ìÇÏ÷?ûÕøó÷ÿö?ZÃ?ÅùŒ°üÂÙ~.çQ ÌÖã¬åg³VÞD&Öô³9‹þ‹óç÷¹°c:§?(¾ l|Ÿï×ðËø6¯t|4@y[rŒÈ€.ò HA'xÿ^"ÿÀûÈp x?z<ð¯â-T zi–/÷_©Ÿ`}æ̳jkþ…¬è2ÿë¯üS¬°^Ö_Üë¶úÜ/#äg-ˆÿëœk­l×ÇþÓK‰»ó‹’­*÷‰‹±JwkdåòÀÇ®…˜›£èIMG±l¡½ùìp€íd÷ à«™âW\~òì¡æ¢V¯ðoÕÒ¡l›}F¥ÙS³²{.È¢²BÁ¶-ëX£Ø¼´¨JöÀiãœ?ŽŸçÀËï`}˜…§Ùu½î*`Эýd®[¼Š=_²¼ —ÿy[¦]Œ‰í¦!$}aû™R¡íù3.y´t×ep“úQšŒkià½90½ÎV°ç×Bé“ïqRÚ–õúÀM%0n >vH“gY"KÐ>ð´“ïYàk¸%ظ"Z³±ÿ™YWL¾ÛøùìFA†“3Ó nXXm 90Žrcb¼°²õ‡½¿Î[6&å=!GôµYǪ” îÆÏx¢a¾¯„Â[@øq#û°L—g؉-Ìõqñî‹ÉîDœ€ÿØ¿/&Å ÒÃœ±\j¶œ3ÒÑLR6Îà„$ùÅ÷Y.ÜrÌ$ú‰ñ{â¥ß®ùnÚDâ­í?ãyà}w{žg<ÀɾW‚àD:Zÿ)ãýXÚ÷·½áÈiç(õ€o«bt?ƒs6¹€°àY-8í“*÷ÃûŒÌŠ¸^“| Ŷï#ØbI­ç\¨=†„L m·Y=jŽ LÊ9äq7!gWÜzZžöµ ©Nª—Ø€™ 5-Bûû¾“ãuL•„±Ñ2†®Ü¶ÅgX%óŒZËhVb«‰d J[¡€é÷7Dˆ™üˆ1h·2Ûßß`çÏ(‹f`Æ*g£¨càÚç´†¢ñλ;;†hGzF­XûPÚ³-€($iyÏöcËD=˹-?§²AÛž0¦ÂÅP`§üqÃÒ¬¦,Ñ;*›b{Ý.dÖâòÄ*ÉÄEïþø¹ãMòåd= ^].X!øbsÆçækÏ ã%qúñ“¥‰Õ›4YùTÿÜÉ<²>’å›ùPyü=b ÂÐ×0K ëáÞøÚ¡›sÃw[HÆö¶c‚þîGwÒ¬ØÜ»0evÅIóis*b²í‚郙¸óYª5MO6Vò9¼8 1Á—A4û—Ö&¬ Ëlë¾t,K#ö•e ¼.ê+ Ë—êtz‰·è]¯¸gÎ0vgƒófè°l;ÇE˜ìtÉÞ¹£ }a=Êåé jNžxMŽ­‚øÉl1[„¬¿Ú¤÷Ô‡Ó%x_äŽ[Vy¥ºJ cGnÓXy:ªaí!HnßÝ‚¾qù;y ¥¨è´uKÕt{Ák7oÏÅtÈ%Þ[n…` 2EnјÓÐnÆ <™mžÞ]Â'=£ŽVïÙ ¸œ/Ä´»…]f™UŸ ‰[w}¶%ÉѷWƒm+(h ¶Ä\Î[qµÛ®qõû’’÷–ûªc×TÓàóÆ@W¨Ifs©ZÐÔ½ÛV‹.½âf5òŠË…±š Î[Q°mWwõ«oÐ%x_äŽh~«}‘u—–cNÓ»„ÌÈ™oS‚Z:Nâ%”=ÎT˜„œ?ðd4±íOì[tÊ•<ŸpEÛz–£=¸Ïb˜‚L»fFíŠ;G:Lª€¤•ìô©n¼„rh<Ä ‚Q¼IvüÝI·´¾I§\I¢…ܼŒÁ)È%Y«ˆRm,î²£– –€ÞÄȈ]3>KÕg…˜§=òUÛõÙ* DŸ¤Rã{‘•žtOx’Âü#tv«8þûí¸R4Õ÷ôºsÀ\ηtr¡BD¶”òeËX`«ÙÓ(Ù'©]“¹^e¥ø ðÚáûóBX à¦oÊ*Fnœˆ°£½œ:egm,¥( 1͆ÂiÐh/e(Ø„}’Út?ðá»eÕÆaV‡Œú)Ÿ7Æj¨+HBÊsFHu§ÇU¥¡ ¾ŒS„x u«.&ñVÜäÖUòºéÍú®º ®Œà½åŽƒfî4|ã0 NsU'+ì;ôSÒVRзlè邺'B ƒ“ xqöSÛÕ ,ªoÒ)yo¹µqtiC ?¾iAM㢸‘ªræE‡K¼iC¦`è7kì”%¿fߤS.òÞrsBrhõ 6B˜µ£O,%–ÒÛ—û†ñMÏIÄð+ar^»†ˆbf@ò«LÃV‹®aŸ¤Úòz•5®\5Šß”% k_`›©¬}÷ ×bWÐý†~A%.h%&ãhKÞd'µâ‚Vö zö·Õ÷¹ã29 ”)IÌÓ-â¨Å:‚Èc Â0æI‰tÜ ^@h[ÚRÑæd;’7£€8Û$íÅ´JÒ))yo¹¹ò.™6ÌvDT šß e»Ì¹³·œ¦GBx`Žix4q BÃøµcRvÞ3K²šÀf-§ºŒ)¿ ¬ižPôc»BLMÚÀ6Ïݘƒk ÒØZ3èvU‘óBÛ[†áEÇà-ŒøAÛ€4¶v:%%ï-wž†‹¶ë€ãÌt4B2ÒšòMºijÎK¨yeap¦oÑ% ’äԠS®ªËBnfÖlèzU§&!Lu˜ &@s«LÔÆÀtÛAEü%ùªÛàxÅV×@|¦u¨Õ¬mU}BÒß‹¬\_3ŒñMåe„°0Bè9g?xkjGœÿ‡ÚÍCN>M °¶ÌŠÏnŽR«EÚ¯ú•Ò€ï–õº£§bT¯¸£ªÀD‰›ŽÓ&3Öz$G$ ’¬ú\vðš~oÄ ËŽÚöíÇû&r‘÷–›#Éj.‹äù ØRX‡9ªÖ>+Ï°5ÿ¶5òv@›çÝÿžÊù!lWñ’ÇâvØžçà±/Ð x.ùt”hv¼ÈqeŠ0ô 68p% ½ð9®L)0Ó‹>¡¤¸÷u,µ_CæPlDNÑv\™¢¾I§\ä½åæD 1~«dznºÃWñÌœnus”ÏGMZQ—ûç@«´¹Âjà¯Z ¶ €s¦W§R<ðݲr§¶°ks@@}ÂÂO|‡g[iKf*?¬s  EOÀ¼”Ö†~V#ñ™½@­v=Ô'¨”|·¬Wùd#›TVb¦YNœ Ųå­Äq$goIA¯ØmÉ[ezÕ!J˜Žò6ÙÙv»ë‘—ætÉÕe…¹ãY\Íyù³ î4ãsDm Ò³´tXPÅ[udÎqˆ*px–¼°@©ípרoZ¨¶YKú"w”»¨´òHƒZºiA Ê´ußë}ã!*­ ãšRò *Q°2Ët&³xñGm‡ÛY}ƒ.¹Àû"7ŸeÃM“_ô LF*TÅ&euÃNh£y¤ãFàbÑq‹x ņÄ<‡™sÐ!ï >Þ6 ü}‹¹Ä{ËezŸ³Ù®sEÞ3þ±ià|.qèpUlÝ õ˜P—øT¶"¢``€/ËpvZÍ÷êÜŠS! ù.YÃVËŒ&•"WÈÃinÆÞpÖ—¨¹Wpíè܈—ÛVÛ²Xq£fˆÒ€šÞg´Õ:Ôwת¸‹ó¾È­gé0h÷súû¼1Cn`ÚËœÆTí5¬N´G úÖb•r@Ý£Z„i¦úŠ·šVmW/Ž¡¾I§\ä½å¾*(áôP4݉)Ö Ñïʱ*GU&ðÎA7ƒ˜xe¯,LÂ-M.É<§]iº³Të…åæ¼/rGüÛÄ»Yÿ6éŒAüóòÍ7Üd]tß1Vxч–áyAÕíMÂð²Gòù†Û$µórÆH.¿H2ä¦ }#îÇÏáBØyz[:OÚföös·ÌâkÝ ™’Õ¡¨C j ˆuÔ¶ÇÕµ!$”q¾ˆü%"±û±O SE§¨Á¸•®sÊ’ÎéL^B=®È$¦ÁpI^dY©íêÇ>õM:å"ï-wØqÙºn@¦cm/ˆÖìö‚ŽÝf ’e–ÑÓFoˆ%/¡æ7 ƒj÷âEµ{µ]¼£ú&=+V3·/rG42üröhä…ñ­aÆÑ{4AXe&¼xU® V‚—PñXia_'^\øª¶ãBXõM:åÊ G ¹¯èJ(Wn“núfWîZ_+ yi“‹Ä8Ñ‘Ô)^B-lÒÄ`ðˆ—ùb]QWü"#"ŸÌyo¹ÃäSè “aŠŒ3aî8oÞîà]‚äÑ<}‹N¹È{ËÍrVþÎDá-[DT™ôè[ÚÐìÌRÛUÞeZ&PoÎ7„Ér„B˜O]|H‚b«YÕËÔ'¨”|·¬qA`í¡Äýì6“Â’Ê°™$¤j¦,H–˽ƒŽ¼`ñšêMÌ€AÞnÇ&µÝýd¦¾I§\ä½åÖtçÏî¶æ˜Vᵨv@²8!Mɹ‚Ž8dñÊî¶&Ù„jžry¾Û.n—QߢWÅõþEŠ7~̸™ªb{´:¨]xc56X@²Â¦éô‘¤‰Ï êqU 1 ×Í·á›nYpý›³ï¦1Q³ó¾ÈM…¥(¯ˆ%ì„À©¡*‹¢X´“(XÂnwåP;¶ìû ³ýtßê&æ|VòL­zm5õ j‘êÐÚ«¬Wa~º(û¦¤$¬ž¼ÒºdAÆŠÁ„Ú;x‡¦×Ÿ%fé:¶Œ»»· È5è””¼·ÜQ*í¬`µÜU¦«<2†U¹H^UÏê):#Éëј^å1"Gðb³PÛÙÏ ê›tÊ•Uå1äÖ.’-Sv»¢ÄQLÍÜH)”»ÂBÏsw’‰~§à5MëÇÈgpn\%ˆ†|ãVvâH…Œä»“uÆSöá&ëŒ'e–ò;š‘i2ñjÛʱ$~O—›±&}9oÙXPxX(kö-zRVñÞrë{,U3“, r¶NÞ;+~•¼ômÝcMr ¨», QÄ‹÷ ¶ã=©oÐ%“Xn¹ã»t·-~Þ˜ÇzFÍo7E¤7m>"Ò‘ª!^BÙ –Â$|,ÞÖD¶]#êƒ}‹ž]PÚ¹¥ÑÉ•«WBU6<1ë€6‡,HV†½^²øÁ+(î⦠Ÿ¼YJ—µwq©oÒ)yo¹#ÇH®È¨„<âìÔªÆEP7dÝÖ:ÖÄ¡l:Z÷Å“ S‘ÑÑu–±ǶãÉÕwÕ=}ÌOÚí‹Ü×-N˜Cuúã*w¢q“Ræ=P¸i‰ÆXA÷Õ·^PŽ»<&½nç=®®ímÒc|Ñ)yo¹ã4ß±5׸-¨c綂 õ*xoÊ‚Òi¾æ EªÕÐò%V˜‰gíRŒz÷¶ù{êA§\ä½åŽ’µýå¸à´B®”n#È8mA2«§éô†´\ò Zq{c’sÞno$ïÐíŒÉm•ß¸¢ætÊÞ¹£'W`Ý1/Ìô;ïé*;¼-AÚÚp:¯¿!¯ ëŽù¬UÚÊ’—ÁP[ñ{ž]žIŸZ…{ù"wÜâĤÓüVoëUf~s¸^FŒÈ¶ƒ^á|+ºÕ¬lµí+8ô´œW#2Ëu¬£û&]r÷Eî8v hVžë†€4„1˜…q"/'v^§)úFÞ yò݈d?Š¹]j;r¿Ô7é]AÆÌå ¹¥Sfÿ½Êïfe„K•- š/Ò©ôniñ*~„Æ 3;o²GRÛÉÝê›ôì)[ý‹Üñ,qẑJ¤‘ 6¹›Õ-'$ýXIfЈç ^BÍo¤^;ñ›6j»x¹jõM:å"ï-·ž¥Á¾a†ÞtÄP¹Þ¡¿©£/­«é‚Â! ü â…¯Am‡/bº–_‘f’Ô¿È-ë’èõJ9ßÄà îÉä&\Á=Þ}÷›ዤWEÿ÷Põma`î¯åoyÛ‘)¦¾A—\¼§â–;öÊ Yvì•£jc­ÙŒÅΩ\,Íæ¦Û­gŽI^BÝ3¯u¥oÓ n; Ò¹KÐ)yo¹¯Øu#¦â Å *Y¸p€û>rЛ´€P‰S˜0¼Ò¼e¢Ô5Ú¤ÝH;} I—jþ‡Ü‘_H³ÇPémbº2™±Þ°2[½…V²jÐ'¢ÉK¨©Š¾cì*CçÅ‘LmG¶ú&½)hŒ/r{¨)kNÐ¥*„hŸ¨§Úís£¶—v_¸£ˆ=|¼®»VH|²­V½¨»úuªdF¯²F¬f£óG©žŽÙˆ¹ëªÝ :/¡äf$b ®'/çÛ¾æ¢,ã¤7¥3öõEîP3í#¹jœýÕá¼ÝžhæÒ"È=)èÈk/¡á×J ãË4C•ô]i3ežÓ*!yo¹¥¶ LØIpàö˜£ª«wb{¬œ*B˜Á.o?n¾×!)ÒÂÌ«X¨ð+Ñöuiûªjr‘÷–;.l´ÌKŽú¼1Ȉ¥n¨>í}ƒ.¹Àû"wŒ± µä:~ÃãWÇ!ƒÇ¯^iÔôô-µe¿@×ñ‹Š _º-½xÛWô"û½ë6t¿Bî¸PËbS—_>)̶—6*ÂU­ÓàÎÚîB^ûØèÈ/¡å× ƒÜqñ¡¶Ã\¡¾I§\ä½åŽ}ßbÝJ|Çl| ³âWÄM£ ½i‹¬#}#ê®è»º¾‹axñ9yÊP-yÜw9åA0:åʺz8äÖñkÊÈ%ÓËtŸ€]YJ§N¨“ñ‹ÅK|-wcúJJ#=¼‚’›^ˆ)H|!oA".Û¤ÃZ®A‡\ä}‘›Ï²*”Õí!zÄTKRýidyŸªøT7t¬X(. :€x§€2 ƒ‰K^z2ÙöåédߢW%ÿï-7öˆ‹~¯9ÓÔ×…ÄÙi ÑBxÝt} õT‰Øˆoá7š aj/ù†iºluøuæì“TŠ¾[Ö8=*þ&"ÙÆVáûˆH¶.H[#ýÉVƒ—Ðk$Ï%5x#’­¿‰dkQ’@ry$[È}©ù8Sp´0w!NT²t!–&äv­êô”;ò zILJð'”¼T>Ùv‰Z6ì›tÊEÞ[î뢳3gXÍ0ŸTæy#¥À†ÌKÑKx¨¬Ü5è q§ä%Ôâê2arúÁÛ&ôH/På‘»è[tJJÞ[î>`q¯¢ïB ¬M§’3F\ÂwYM/¬Åé§ðLu^AÅmÌ ½O¼ ïÙv,tê›tÊEÞ[ny¿  »„óóÆdø»*"2ìO•R••Srº§{ì Jnû'¦²z5x+K[{¦†lÿì[tÈ%Þ[nžêaÆ-¶Ù|^+wÖ|è È MU€-2Ñ5 2†|vIÃÉòÇÚœÏî(S«~šú4„¤éÌ5 Y¯¼î;ò:°šB—”m·‘0Îl¼­p¼¢l=£c[/¡ìUn„IÖ y ® eÛ%®eߤK.ð¾ÈK×@ ¥…Ó`–C:\!ËåâSD sA\w^AÕƒB„)Ê6e^‡™å–ò:®‹¨™×±{ÈUPr‡ÉàZÝk( ¯¨rÚ%M®C¿’µÃÁ¶u DîÁKhúÍk d/µÝ½ˆªú&r‘÷–ûÊŒE©Gf,䱓MRIY»"Ãì¶ü:[–’Ý ¤ šž{* ªÈ;4ØÛŽ dõݵ¦¼·Ü¯epåá]† !Œ¥p¢JúƳæY³DÇP/¡¢ÄàbñúEBÖvñàõM:å"ï-·¬`mÉ-GÂdÅz.øəð<Øaˆ¨½ãîbòJn9"F YŒô[Ú‡, pùˆdߢjí‹Ü‘oÃlÕß&KUe`.ÕØÒ}£ýÆ9ªÁŒþ«”nƒoY*‰ÙnTs™ çÏ‚¾E*Á×ê¹ã»°*¯ŒÞÂtÅqó€Ó¸£gAú.òW"’ÂôHò £·0U+/í`B9ci4÷Õw‘5¼Ôà½å¾*0wSqÆRÛj<1°f º*DgÔ y›J0ï0É+ŒoG¡Ä¢UË )²oÒ)yo¹cÛÈ®(Ù×1à vQ7` 3g# ­LVº”t›ŠÎK¨¸M[”`/Šs©í䎎¸}£Õ‹¼·Ü‘H´˜¸Ô=‘h1± <ƒIOX“Y”þZ“û:ÂÅK¨ßk2¯ãšÜ©¸Îh»FÚû]r÷Eî+mÅ®@há»[V4°e Î=tUÍhÜWœjn"ñh·/©Ù- ,“κî­ϪdŸ¤Bò]²F W5Õ{¦Ê/Æ1K¸Êÿ$¹|}Þ˜‹)oUFá”/?ƒ7áÖû¬ºŒ:ö«oÐ%x_äÖ§@P{V’¶¨Ð-µaU$ëÙÍ–WP8o7†d¸Ûzr>9ìÉ@$œïx·’Z ·3›S! ù.Y_òuGègQeO†o68ÐÕš ?È— ÃŒ+^@õªm@Ì’ ½Ñż¼ízÕ6¨^8t…¤Ka£!wÌp^M«u Žë`…ƒŽk­²©¹€4ÃüÓnÂÀ qAÕƒø†'ÌE¼8’©íìA|êtÉÞ¹C³OPEsØ TÕ„­Î*Å#hðG› Ù†~!: ú‰—P —61œBä-¸6‹m—;ðÍú&r‘÷–;ÆO.«ûã‰kaÊÚ‰ f²ŽZ£€4ÆXøp@%à‰k]PŽ;È»N\`Õ]½™Ñvò Ô7éE'®5¿È­¥k©p³Œ ^:=‚ µÔé-¤åÅꘂ~nò©Î+è |#µÖÉ[‡LåÔY•ì[tÊ5e<¹Ã—b9¸V^òóÆtèê¸%âÔ«hfb4ÃÕp{êÌÊ&6:’ÐÉ+(ŠY Sc¼¸ÖMmG1KõÍkß(x_ä¾æK²S†N)ÂÀµÑQä$!!WrÜ:Š¯ˆ—ÐÌI â‡ÄÛt÷¦µÝ"„}“N¹È{Ë! gÆ!%ÿóÆtUÃÍÓùyk! êHA·Dç%Ô¼*‰0HMoÑÍíÖvy¹Ù½rUÔ†ÜWÚ <­Ò ‰9a‰PËË ¯UWҞפb…Û$ÏKjÁ;•xEE$œDè)ðØ2TFa£ZWš®[áµ™ ;­_äŽý¥#¡ß_:C’“­ñ!ɼ’§,AÚ_çaÏà Æ[•¼¨7- F$yyíÛ¾®Ebߺi+ ¶¶/rG Ž”¥& l‡ÇºS¬½ [PK‚¤}3EÉè &ÌKmëÉ3¬R,Èô_+ȶ«g©©oÐ%WSE´;Ò‰˜¡©g¦›<ÏZÃêÿ ³ ÂÊTa %þCñrc£cÂ*^¤°ªíâ§9õMzÒE*½‘;¬`ô°ìV0ÖË:f¦QUÁ¿â 5¦ ÙOFúPm¿ k[V0žÅGuÞSšµzÛ€Üδƒ^äýõ‹ÜWé†aÄ Ÿ.7bT>ÍT4áy4¾vqªéIä#°\c'wwoá².´ @#kQ) ønY#ËCsEzÓm!o°g´Æ›“rtvÂJjБ.+^BÍùÂTebo@)GÛ‘Ù£¾I/ºµrå/rÇ°:ÚY³« ?o Àª)<ÐÂbÄÛFåÒ¸ØEtTj/¡æ8aª­Æâµ #½íâ>'õM:å"ï-w,]Kåؾtu,¥ffø5ŒÚùÒU‚ÎxÒ¤;Øxµu³T*¤bYÜÙÛ.Q‰Z7«.IÁû"w,]V©4û‘E”„™¼‘»c¯SÔˆ`|¶èâ%”ýÈ" ÜvàåMÞh€+wÍ©ŠŒ·Ð¡ó®ò? ŽOÛ*òÞàTµ+­ù°ØˆÛyE apè/ë,%÷Þ«¾A—\à}‘›…Ë,”lûnBØömó'¡¼çvUO¼báQ%<ü·ï „m yEØ^Ñ>Ⱦ@ƒ ÆsËw¹}=·øóÆ Ô3ÆMÛGÓÉpÍvAîö­No^p\ç-;†šüôjÇ‘¥² :©ïªqò Þù’-Ñj±eO¦(ñhnŽ{.w°@_AÇò.^BÙ„ーi+»ÿ´ H£](A:å"ï-·J²¸ESD&·zZAkøN1Éå.@Ö_‚z%:Ò®ÄK(ƹ0IW@©¶“k£ê›tÊEÞ[îKËš,ª÷y!Prïè9SÅø¦Y®aU§ÚÈ€å‘!D TøX­¶+5ÄúÒï’õ2™bÐï6S8á6r¯¨üL ²ã‡¸óøÖ!ŽFü}Aó>¶(;oXd« 7žö »¤)‘; »ç‚|ÞÜ#tn6E@¶i¸ÖìÊ,96Îåô΢v¼ƒÈKܭȧãí«o‚æÁ¶#ÌP}“N¹Àû"·\?V’ÅR¤˜N n{*Eé1{À9ßÌi>ç‚ôn½‹—ÐåÀ#Æ+*vìg8!poõPšù3ÔjÒhSŸ RðݲR|lDÝ/ò bØô_(v;l•X( àž´m§¢Š.ùš ¸q6âT›_·â2lµûì“T/åWYø÷þoïçFèghÕS %«eþýŸßÿþýü™ÅêDi‰çÔfw†ÚŸ—ŸYküóßµÆ?ÿUko¿þãïþê7ï¿ýáíèSéýþÿ¿ýÛ7»ï?ÞþáßÓû?½å÷¿~þûý[6åëoÞ^Eù|«ÃvVÂßóÏwýÞá}ÿáç>Òo~¤¨ßþø'Dýã·?¼ýú/Ö÷ÿå7 'ÿà2lm-ÛfôŸïö¿û^ÞÿüýÇß¿ýÅÖþÓÀ‰%ÎïÔiàýwÿ§¼ÿ¯ó›­f®§iE†ÂŸüÚ Èüÿøy ˆ€þq2”ÿ8ŸSŠ%‡°…÷zÒàÔà÷÷ÿü‚x(ü Y€ÿàýƒi1þ€×j²R.ÀûÇ™ÝãïÜbûS\?øƒk½ Ï ëÏrV¬bÜ¿*jXýIÖf™§Áš~6ç´õ28~ŸÛ²öƒ³ý\ÎZ,¬08ÇÏæ<.Ø}q®_6cþ'æï¬þÔúCø'†©æÆ)>íšûOÀ0 ¬3 †ýµØtšÈèA¾àµh'´ñ=ÇlÀ¦6µ£1¶Øx+08ãWÀßuïIÜŸ;7á‚ÐÆ„þ›ªƒÔB¸#³…ý!Íåľۼ4–%a®ÐîØ?TÖ£ NÂô¥w´oˆœ4ÂYº0ÃŽM LÏ>Þç£ßtèÑxþ±äê[‹0c1L7Ž9fŒŽnwç,§Î‘Ÿ Ãô‚3Ÿï”%„~O €ÓÞÇpKWl Ç–ð{T/:ÍZÿ¥žn‹É×ñ<üˆy`T¨;I/pFPã²ö;Î Ïcb¼°É°:–n0OÊKßÆXêåG†ÙŒ[v†¹µŒ[Ú†Et?0/$–ªrड़ó ÁÑÃbÛŒ¬ð±)ß)C9g îÆ :  3Ûµepú žw%ãÀ?cñ}£Ðý³õ_©zRìûUdóûxšO ©jpLH7©¸}j˜â ˜vOŒæ© $;p…KÇ^û‹ÅÆ" éÀ¸ ž'…•ÖN81LÈøË–{ òÇå/Ûä; Âð¼ãÙRž0¿KW.æ'k¼žaƒþš*ÒV´_U«ò%Áµ8†×£[d2`&Öb|”,SƯoŸ~É’n¸Kž©8{ºÛ³iý•ÌùOyŠÕ y³?/ž'O®§|^ÜÒï#w½_¼¯ìÏË,çÂõ„ï;gŽ~œ8?ø½’æ+¿gš\ïù½“Õ:ŒñÌAã%™9ÆÓÉ ×xK™ï‹ã1%®W¯{s8`8o­íÛnóÉ°íJ9Ÿ+»ŸȧÒÖ·™öœÔ¡c"nÍcÌÓ]8í0ŸSEÓ¬~Ä=Y)¥9Œ5bY ¾„,ÉŒæ¤hõX€–@òõi ¾,_kp4`u[ƒ‹)¿e)_W×ÒlK粋J|e]f°ò…wi]ú¼š–}[¶WÓªm«úªjÊýU¹§aOX•/[ƪÑØQ8bÃ9M]ûÑÒðÀvµÚËn¶4y°Ù­†£ ÷ÂÕ¹c«\§Rî¤ÈÕ¾þÚWm~@lóئ×ä.†]|MBlòÇ¡Ät€µ¤2Ø Y›3ĶŠÒߥ`<#‡ú‚é°Û|—zrB-šö²+”›mV™ïÒ}¶¦T£=8Ó¡9=s!ƒjŠÕ9˜Ì붶uªeH=ýîj[ʳTëRá²Mµ/YöíwW “Ô}wµ1iÙ£Z™L{þîjg²hŸï®–•]jéƲƒ¹JµöY–0¥¨öæÊg§Z|.k’Z±ì-Áf"×Û Úß,Õò¼¥Cm/™Ë,ÕúRôû©¸T y úñy|ÿrløn­ÂƒN骯í˜q¦@NÎ03ãX[†a g–±þŒÑü“ä%äÅ#c œ×,©Þ¶[Z½oÒ‹•^_ä†qNje)r~8ÆÖç33"}¬Q½I:ùln– øaÉK¨ÈùỲÁy-ÔÆÛöPï›tÊEÞ[n> 3³©¼Ÿ¦X ßY(; #ݱQ&AzwND:È mã×ÄXšœx³ò©m@ø5ú’’÷–[ÏBY¦\µŽ±t_—Îv†3*ʤQSZЧ›ÆyÖ§clò8¯]½íî_Q}“^u ðØ_äÖ³4ÝÖô,ÄXü`QµL‹•<½·$HÒ1©ÌèKéyé‚–‚EcÖçf¿PÛC©ó%ª|r‘÷–›ÏR+jõW•¸0Ç>ù(¬õÿ¬?gÓŸžUdý}¿ÝkŸ¯CôÓ8¦˜•I¼¦ÂzÛYe¼oÒ«nHõ‹Ü1÷7SÚ–ÏýM¥xØücmÂ1l~Î,Hsì Õ=\´ðè«›"Þ£º%oæ>/º3z‘ë{§/rdzڄVˆæóÆ”3ñl3«b„wAc o6:²øÄK(Ç NŒ‡Ä{¼ŽÃÛ¤1VfÐ)yo¹õ,Cõ/™·ïˉ²þ0¦û6YÊ„þ<¡öÐ;ÓÁ+¨ÉAëÛ×ò©¼í*W¼÷M:å"ï-·æ‹eAØþ͹/Œ~Î(æJyV&Ý0ì7QÝ2L:î¹ïRοVpaÌè伦òzÛMu;½oÒ)yo¹õ,SW?i¦ØQ—®'?3à\³×}ãnT«Ó.#!¯ û 1I{„é=vöPÛ€ôäur÷En= ¿ãÒªÏcôÓ÷׎޹õk?ËA/ÚÉç%eÆ “7tðf35¨m@xôM:å"ï‹Üz–­k]XîÂ-ˆÁ/¹C Bô =‹•ë®%%ç (%ÃXõ4çµBhÞ¶WHõ¾A—\à}‘;æþF¢Âœ>÷7 %{÷„Y¸c÷3£LVÐ{aYÒYk–¼ªzëñŽYxrVŸxr´ HsΠSRòÞrs=Yï=»žÌhKê –0[]GéImKãál>>ñZö™·íÙiÞ7锋¼·Ü±¿0õ¶%ß_X>¯Bç…!Z:/ ÉË¥cýV££Æ˜x ùu‹Ž±˜ñ‹PÛ€¤ñ£7O ._äÖ³ ^1"SSð-qF<‘'ÛÞŸÌ÷—ª§ëÔ’rðòÈÇXt‰ó&̶•ê}'Í.¦¬òEnUÞ4­/rë» ¯; }ß1¶ïOU"8£w Ò»ãvÔ òrÞ€|ßO uãÞ]Î\r´bä}“N¹È{ËÍgy¶Ö#çMŽ1ËUÉ(ÁX-9øD,°²¹t¦ð¾I§\ä½å¾Ô–ö~Ýüí˜óÉ„j1`ZRwKq"ý¨%ÉyùÝߎ©0Ÿ·À´Â¶K˜^Ø7锋¼·Üá ã­›3\aÌkŸp…ñ8ØáLâ}=\a¬2Øá^b^ݼ á&1ahj'¯EyÛÍ•œ¸~y·Ü×wA†åñ]ŠoÞmªBñÄaŒ J³ùwaÙ¸‰WÞÎ+È‹Þ:†¦Lò6˜ Øv û&}(£¯/rdzØEmMײ;ÆU,à9êš _7KÐ$Þ­ËedÆ»ÕîÜéYò:å"ï-7Ÿez²•Ž_ÂXPÏ ›H€Î›žU!š«`™Ð¢#öT¼Uæ4¿„±‚ Îk¼m¯*å}“N¹È{Ë­ulaÇhÊrŒÅBž˜3ž¹]h ßõ7P&JtO§ïÔ\‘žÖ~ÂÙÅ[̽©¶K(ÒìtÉÞ¹cŒ%·åcŒ‡U:Ûx¥#neAÒ3æ:Êm‰—Ð GÀöêr)x=}„µoÝ à—ør‘÷–[ºå8nY€ïç…°dÜ£–'+»p–¥sth\÷+NµºÇâ0UúV»X|J¤V‡¦Ø'©†|—¬V°=Ó†1P:®JËÞ`]$nÉAG Gñšá-n*[ÄͶâF\ª )l¥r‘÷–[ÓQQ0â|^„ÖÙ$„ºx2-΂0%qý˜èØÊÅ ¨î˜ÀM¡þ-9oµ>j~]¤nŽ’’÷–›ÏÂÀøã†×qyñ[Ð)yo¹Ãû•q²ÎZºˆi6jÎAÐ8ÃAR-xÁŒÑQ\¼„.ÿ1ª3yÍëãm_þ%öM:å"ï-w|—…MgEï—^PºxsÜ€R5† }—Ñ‚¾PD°*ž§5µí6:úyvðf؇nÐKñäcr‘÷–[ëîö:ÐcÂdØkpûPΰ×l”4ͺØçljæ9}àæ#ò ²¶µ™ÌshJÃyÓ‚Wc«œ™{=–JžÑë±—ó¾È­ïÒϺ·k‰áßá¬v«|ôg§"ëØ¿¨G|äHí|v¸Z;n³OR! ù.Yã$œ ß²øc¬&–Nœ¸· 5ŠÎƒs:g[ò š®×3`ïH †ü¿•]M¯%ÉQÝ¿_q—°°œßY¹`ã !À3 ›Á²Ü–`ü}^Å9'"ꎻ=,ºŸ"¢"3êVUfÆ÷mÏÐØÓÍÊš›tÊÞ‡ÜésÇx¥ÇçŽù íRÐè¥ï '.x:ñŒwʈça8XVfðš%ÒÇöš>7éMZL™orÇke¹BõV·¿dLà +;¬pVrëÎ :Ê툗P ³1!xàšéV(×L1·è”‹¼Yî°‚] I­ncˉùT«ìÄ p†] y`¤Ï*j'¨G01V÷ÒyáöÑØ-œÝWv I.ð>äŽwÌ,C#T¬Z}Oz¼mJ®Ãë­éWЗ™oÅKh䀊z…f>U£…¡ãËxz¾ËÕPr'§*ôìß¾ lкâ^¬…uT/Þ1^Ê6#¯ Ú#1 y;´CŽí›|nÐ%xrÇ;Æ :YŽ„™Þb#«ój«é¤/àjN¿3'šó a9*rÖ¼ «Çö~}>7é[Á~u¼É®,š^#À•íx&ÇÖõ&7UZÁÁ}©NCðØ&¨$¨yˆ·0ëæ¼ëæcÒÔܤS.òf¹# œ½2Ü^ÔxÄ,àUÁýΉ!HïtiAïzJ3Aãƒ;_5x,Ü»…¼ªѽǪor‡|Ò·Ü>á Ýxzr“Bóä+Ó :Â9È+¨Fp1Å<Òz.ÐÃ*u86ó>ä‹Þ‚ÓTÖaÞKä6¶›ØÄSonmK%¾i3cÇ“š PÖ„)²k 8]ûåc×PÖ8·è]ÎÞz½É/J}ûÕ“2 ‰g¹qgëëáQ½¥VÄ;Ðsš¯àA=g/¾mÛþgЫ'PŒ7¹“ãg¦Ö}[ᶗœ—>vŒ(af㉿Š”µmWÆâ¥×>z{fe}qƹÀû;öÊÍ°€8m†àûœ¬ÑYí—œ[~é9ƒN›!y U÷ Sdƒž Ÿ>v kçr‘7Ëß~s›ô—Œ1Û²…Ò0Åh#̦ rïW z•%k'hzˆ„0Æí¼Ï´¨­“ž¹æ]r÷!w Ú¬#áÜ¢S.òf¹õŽMùnª"¦ü;µt„-ž!ÈßšôK}k‚º[k„1‡‘óV3škìêÖÍMúTßúÞäe£¯³¹‰²תÚjÜ»ý3[N‰LÜ?×ؼ‰å CóŤ¿ŸkU;î\sw½‘L¡:ãMî”B`Ja¬“@èQdH¤b§§8žU×r„)Hïí1;>6 ÝùA§\äÍrÇùøÂÎÓ»Ÿ/ì>šÑÂŽYa[K¾ý5‚N¼„®Xb§Ü}:o[8YQC\€¹E§\äÍr‡ÞÂþõIoá¯Á'Ën÷Ô[f¤=™5þùÖ0bª&èÊz »Po™p_Po™M~§9‚N¹¶ô–;Ùøxï%l|¸úÌîŸÁØk rÛÕº[kF‚V¤‹å–Î;ªº½)qóßIý1[È5•Èr'}’çƒú$ÎúáƒãÈ ÐÐÞýápð üF÷ž|%AÕÓ…±@0ñ¦7r r}ò½+e¶Ö7¹Ó½0ì½Æ½0¶€V=ÄXlX)L—NØ·xîÝq¾Wvœ‡Â€yâi j¡OS¡O’×¼¿>vq÷‡æ&rUé¢!wÚ÷!K‹óqAJÂýkÜ{/ã[pZ¾š ß÷KÐ;ºN’—ÐP¹}Çt8‚ÈÛþ±›Z<ûܤS.òf¹#Òmá‹’¯B˜¥'Á”þ g”±ùé}}J/ _ÅtÇi™ÁÛõ”ÆÓW¡¹I¯Z Ê|“;Ö1¾Švf æc¢Uò„ oÙi¾€SƒŽ.â%äE™ÓÌ9/Þª˜Ói7:3èSoì*orǾ¿QÓêtß÷ Ó/yïHSh×s Ò*;»Ó'R†É+hºÅ^˜ ‘¼<ùrìgÎ=tÆÙ;x³Üº—£’RˆéP¤&ê§u(Y·•a Ò{À{£7¥°”€ZhaÂ@"o³Zjî¼l§KRªpYîd{… d°½Ò’€:­ ðú0=$K×zÃìSštÙÛïçǵ‚·IÛmGÛ^«Ó%xr‡ýż÷ð¹3¤)ZŠä„Ï NÉþ²›Óïâ漂zø\‰± dç­ò©ÚØ5|®œ›tÊEÞ,w¼c*Â^üS!vXªyj¡›'–I,ð±‘$Â2£ç•(IÁR!¥ïÔN¶† ÷Ò,§K®%O`Èý´#ß«vyØ‘;ã1}w<8ÑéÛó_¯A‡X¼„º[Ñ„áN>”ø3jŒ]sJ"í£†\äÍrkY°T·ñ ce§nè è’i½'ºn^Ò®Œ~äÙ ªnã¦@ëoµª]Дäž[tÊEÞ,w*ă˹ü<6¸ö±0־ݦMA:µôíq¢ ê9ŞƲË}+1v‹§È¹IwIKy“;í/^ÓöKÆT”¢òô%JoŒèãq9]MÀ-fœJX¢h^ÎËoc§o‘s7/' ½óz“;ôý¯qEBþÆ;“†„¤EiO‘Ô¨Ä=Ò/ïÕ– åAyÂp½ß^è¤ÅØ#b987é”k*=?äNa<Ìè~†!æ@¯¾íåèL_¶ Iwïp¢_(áF^‡JX ³UâÅx¾fŽ½âÌϹ—t“‹¼Yî§^ E(ë•P”¤ÛÁ)'ÝPÖ+E‡Þ(^BSéûŽÊc0Þ®‡µe½RtÊ5<’䊱u/'–x.Äð¹Tè†|.Ô?Òs©-è—*JŸ”žËaQ¬î¼ eܲý§çRWÐ!—x³Üé^ŠÕw­ñ\ s`EÂÚw‡Ù3ZìäïØ :VYñ²±ý[‘òNÞ¥råä÷R‚N¹È›åŽ½²"›ÆŸ‹cLú¹7­ËClöVû”t|‘âu¨„o˜*=×x­À¢]\ ÕܤS.òf¹ŸEzq;Œ0Ln®ç¢XóZ3¨%èC–ýqTÜCL³„<ñ6¤ sl@0‚Î2äÍr§âx {*.€7‘; s¹Â¼‡ök+©,'Ðœ—Æv­7VxòÒÇɱ“”s‹¾uæçr+Zÿ ÞôQ·.aîF.Õò3o=vãLÔQ=|Ç™©ÛqÅév”q^@Wƒ}ä]e®X™|{A<ÎMº$ïCîØ+ÙÙÞcàˆ²"uÆBbUëS["{Ð/ùHv‚zÎídaƱ±èUi1vŠ›*šUZÈEÞ,w:ÃLx"#Ä0¡ÑÛxêÄ y¦5r¢˜‹JúDÕcò Šœ(ap/N¢»=Êö¬ S.òf¹#>éÂ*;ô½3q°w P¬ÕKÖ±ûY‹ËŠxÙþ¸•H˜ 6Ú®}vm{¤’Uû´Ñ)饈¹;t1k—ëbÔEi‡bœK¡V{i­a=í0uà– æÖga,õÑy J¸pì´êqnÐ%xr§‚•Å}Þ£úØ€<:ãrºä*Š› ¹£µ^ÅŽ3`½›²uÚn»dM˜Ú›·,I cï¥øUÅŽÓÃNÞjg9]ý¬§¹I§\äÍrÇÙ’‰6+â` N ^ÆñxLÇ? Á8~þvŒã_ Ú¾W ³TÌ–qü,Ë8~/;ßÄD&­ù&wÊ8¤ÝgFÆ!ËÀÂGÂS‡¢´‹ ·_A?Š») Ú9vÒ×b‹}ìżÚVÈEÞ,wZ“·E~É«"­b9h;i'º%Hkò}ÆÝÜ×ÎKhD„)1QfämŠ2ëK¯É—Ó%W×é-䎂bÑ nS"ILÓªy$ÅDuˆ°•L4}«BÄHP ›0Š¤oGŽ H§j³‘¹Ä›åN6Xz£jØ`aa눓f?ò[ºí¤]Û`«ÓQÆB¼‚J|]EåsFwÞv)nwAnƒAg6üQ …[÷Ûà Û¸0°w Ázm6‹².AËéK¤è=™ÉK(lãÂÀ¾-Þ†s2ÇnqNæܤS.òf¹Sœ%V–1Ö «Ï¡Æ‚•iGD íÊMŽøò Z‘OÌTI½ÂX£+Æž‘àwä5`ö~ÛÎû;Ùap:ma»0LÇ[ãuN”¾äv˜tÖt ¯Wxh±û¥§Å[PÂŽc÷‡hnÒ)y³Üé^˜£4ã^˜[Ū€¸&3n3­ÉlæÊ5™åÿJ‚f^“Ykò‚~ÓvŒÝ#{‰s“N¹È›åNùH–¡7"ÄúB–v»eNè4¸£×çPE•MPF#»œ†åðÑ»½¬Òš>j‹äLÌÙ´^S|YÖøDõ§"L.£ûŒmÀ6˜½I®=‚ŽzšäÔ3ràgì— ÷„¬ /Y/çÃÓ¼Rv>ö»ß»öS÷±鸪¦J=ä"o–ûáù¾ã­X’­<ßÊϹOQàúçÉë-ª9µÅàŠ;[1ÕjUÍK£H÷+*¤Ùî/^>ª'³±ŸÃñ×êbOh¦€b4†rÕ˜Œá·¢³Øy H˜RŽ.KÅ™¡´T¬K:° S.òf¹ŸŠ<9dEÖK)Ó¨}!ePVäEG ¤x pÇÓU Þx› ú­&(+ò¢S.òf¹ã¹pÙ_ñ\¸-,<—9³ë{vAz.4uѽÝkð2úä:ÿ¬)½«ó¶íý㺠=öéc¸,·£]ßäŽc¾ÎÔPYˆiP;.¥Jr AÙ}M:Û9“WPÊàĸâíÍ94€|!*…*ÒuBè”ÅÊxÓˆ ™ŒI…MÞaÚèN…„èfŒ½Ã+A©–Бw˜Å„˜³ÎROÌW¿"b§ï Oµ[ZçMîðv}¾'¶"¦ >qÛO:JåÞ’‡ïêó3uÜvbñHŽ+ ºòŠ†u”e2÷h¤Òßs’ iÈ—dÄÕ5WŽa–†à±‡õ«§ ¹á´ô®’„+A‘î, US©ŽkÆØ=)E½À̘J¹Àû;ŽÂw˜Y­Q‰˜†TqcµRLêì‚|m©A·[r^B5÷%3LAQ#ð–ƒ 2Œ HÏoÍ S.òf¹SpîFͯ(k>YÚîói18w Ò|}¡ÿÕÓ‘êÐØ:T˜GO ^O²±KM97锋¼YîP?_E”åø’êš½›P^­ð~' µñó{Õ¾ñ¸¢nåÀo0_ÄF)ÄæÒ/É«Õf¦ûñÕj#–çÀ6q¢FÇŒ¼VƒÙ‚~ЪŒ¼„vè0Žaufݵc‡}Vs“î’–ú&w:1zÆÈ—Œa3¥Ž ¬ŽCÏ€ç?e£L¯ù[cì¥lbŠ*Œ±Á?F6àñ¨·®=¬^¶zðf¹/…<êÝ<~]´¼‚Ãø†nÕ„l¾†¬mÑS+^BáÍƲ™-¯4vxó47锋¼Yn5c±¬$ôrþ’Ö­ÊZž,+d׬7Jl®j›¨(G>¯•÷èÙòyP¾37ȇ5ˆ£÷IpNR! ù’¬±‹œÚ”%LGZ“U˜°¤ عïVvê¨LÉC3A-^BbªLÚ á-W‰±K¼„œtÉއܑ¢~+ú3z–³Š}²©Gw+®íÎK@pÂÁ):’œÄ hFÏ’ÀÜ/=y' Øpì=K8·è.i©orß÷ò/¯ÿ~õ»Îý¥?w·sOÚŸÿùÝë__âeõ>ñO]ˆÏÀŸÇe6L»{m·¿8Z+†âoÆÑI4kmòµ±>~ýÓoÿþ7¯ï¾ÿø|_?ÕŒüÿ÷ßýÓ‡u#üßû÷WyýçG}ýÃç¿?|ÜÛoyýãÇC/óìÿ(°âtûG]íð½¾ÿ蟊áç†Á^ÆÒ4ÂëçÄMÙ£YOF]@0_°Ì-ãÌÛÎü€é‚‡T.dºà7?ðwüñ§?ó;þôãŸ>~ýwŸÏàõÃï_fR*þßa‘—;ïgó××_ýªýõë‡?|üí6ò·XGe肳Ö_Ìúù1˜ÿZ¬åsž ‹âüÅsNZÓo”/Ë´Ê}üùîÔôò–ü0_ð¹ØÞ«.˜/`Ïh¿‚ðã’Õ-ÆÛ/ü¸¢Å%IÔü3ÄÍÿ*~‘9-RzX’Ýןø·x/kôõGþ-^KwÙ_èßä½Kv^_ìßä=û>U?ü/[Ðþ_X]Ïð@Ac–j¼%øçï^ýõçîà+TØÝËÒ¯ßþG{ýÍ}M÷›ùø?S±Š  endstream endobj 212 0 obj << /Alternate /DeviceRGB /N 3 /Length 2596 /Filter /FlateDecode >> stream xœ–wTSهϽ7½P’Š”ÐkhRH ½H‘.*1 JÀ"6DTpDQ‘¦2(à€£C‘±"Š…Q±ëDÔqp–Id­ß¼yïÍ›ß÷~kŸ½ÏÝgï}ÖºüƒÂLX € ¡Xáçň‹g` ðlàp³³BøF™|ØŒl™ø½º ùû*Ó?ŒÁÿŸ”¹Y"1P˜ŒçòøÙ\É8=Wœ%·Oɘ¶4MÎ0JÎ"Y‚2V“sò,[|ö™e9ó2„<ËsÎâeðäÜ'ã9¾Œ‘`çø¹2¾&cƒtI†@Æoä±|N6(’Ü.æsSdl-c’(2‚-ãyàHÉ_ðÒ/XÌÏËÅÎÌZ.$§ˆ&\S†“‹áÏÏMç‹ÅÌ07#â1Ø™YárfÏüYym²";Ø8980m-m¾(Ô]ü›’÷v–^„îDøÃöW~™ °¦eµÙú‡mi]ëP»ý‡Í`/Š²¾u}qº|^RÄâ,g+«ÜÜ\KŸk)/èïúŸC_|ÏR¾Ýïåaxó“8’t1C^7nfz¦DÄÈÎâpù 柇øþuü$¾ˆ/”ED˦L L–µ[Ȉ™B†@øŸšøÃþ¤Ù¹–‰ÚøЖX¥!@~(* {d+Ðï} ÆGùÍ‹Ñ™˜ûÏ‚þ}W¸LþÈ$ŽcGD2¸QÎìšüZ4 E@ê@èÀ¶À¸àA(ˆq`1à‚D €µ ”‚­`'¨u 4ƒ6ptcà48.Ë`ÜR0ž€)ð Ì@„…ÈR‡t CȲ…XäCP”%CBH@ë R¨ª†ê¡fè[è(tº C· Qhúz#0 ¦ÁZ°l³`O8Ž„ÁÉð28.‚·À•p|î„O×àX ?§€:¢‹0ÂFB‘x$ !«¤i@Ú¤¹ŠH‘§È[EE1PL” Ê…⢖¡V¡6£ªQP¨>ÔUÔ(j õMFk¢ÍÑÎèt,:‹.FW ›Ðè³èô8úƒ¡cŒ1ŽL&³³³ÓŽ9…ÆŒa¦±X¬:ÖëŠ År°bl1¶ {{{;Ž}ƒ#âtp¶8_\¡8áú"ãEy‹.,ÖXœ¾øøÅ%œ%Gщ1‰-‰ï9¡œÎôÒ€¥µK§¸lî.îžoo’ïÊ/çO$¹&•'=JvMÞž<™âžR‘òTÀT ž§ú§Ö¥¾N MÛŸö)=&½=—‘˜qTH¦ û2µ3ó2‡³Ì³Š³¤Ëœ—í\6% 5eCÙ‹²»Å4ÙÏÔ€ÄD²^2šã–S“ó&7:÷Hžrž0o`¹ÙòMË'ò}ó¿^ZÁ]Ñ[ [°¶`t¥çÊúUЪ¥«zWë¯.Z=¾Æo͵„µik(´.,/|¹.f]O‘VÑš¢±õ~ë[‹ŠEÅ76¸l¨ÛˆÚ(Ø8¸iMKx%K­K+Jßoæn¾ø•ÍW•_}Ú’´e°Ì¡lÏVÌVáÖëÛÜ·(W.Ï/Û²½scGÉŽ—;—ì¼PaWQ·‹°K²KZ\Ù]ePµµê}uJõHWM{­fí¦Ú×»y»¯ìñØÓV§UWZ÷n¯`ïÍz¿úΣ†Š}˜}9û6F7öÍúº¹I£©´éÃ~á~éˆ}ÍŽÍÍ-š-e­p«¤uò`ÂÁËßxÓÝÆl«o§·—‡$‡›øíõÃA‡{°Ž´}gø]mµ£¤ê\Þ9Õ•Ò%íŽë>x´·Ç¥§ã{Ëï÷Ó=Vs\åx٠‰¢ŸN柜>•uêééäÓc½Kz=s­/¼oðlÐÙóç|Ïé÷ì?yÞõü± ÎŽ^d]ìºäp©sÀ~ ãû:;‡‡º/;]îž7|âŠû•ÓW½¯ž»píÒÈü‘áëQ×oÞH¸!½É»ùèVú­ç·snÏÜYs}·äžÒ½Šûš÷~4ý±]ê =>ê=:ð`Áƒ;cܱ'?eÿô~¼è!ùaÅ„ÎDó#ÛGÇ&}'/?^øxüIÖ“™§Å?+ÿ\ûÌäÙw¿xü20;5þ\ôüÓ¯›_¨¿ØÿÒîeïtØôýW¯f^—¼Qsà-ëmÿ»˜w3¹ï±ï+?˜~èùôñOŸ~÷„óû endstream endobj 215 0 obj << /Length 345 /Filter /FlateDecode >> stream xÚRÍo ¿÷¯àØFA> Œeî ›Évr±·m1µbCRmGõà?(­ÑlÉïñ¿x̲h²(¤ ÛBSt‡)àR"É@¶ïñc9åqS„ø}LÝ>ÿÀkZIò™½ !Hq~ƒ4v»»Åîµ-õ€‡^[ÔV{*Üž°/oÎSüƒ_ÜðWfcs{öвlªú˜˜þá) ¶»q°•ýå|_ÎiŸuUŸìtÕèÂèÖ xþN£k~“%pé´t½wˆuS›C'ùO\{ÌýUÖºªLÓê ÒA¢ç,"‡¹ÌŒJŽå ØG_AX`Õ¹ÊC³Gö…ÉË^§:zskhÁ^ÑÎüaòê¿PJM œ!‰eð¿0åÉê2ÆbuŸÀ4åñÊO7”º uÙ©5‡Ò§i?Ú=‡P(eÞ‚`ˆ H îý tú¨ endstream endobj 181 0 obj << /Type /XObject /Subtype /Form /FormType 1 /PTEX.FileName (/home/biocbuild/bbs-3.20-bioc/tmpdir/RtmpCCJHlc/Rbuild86c017e3fbf7/pcaMethods/vignettes/pcaMethods-017.pdf) /PTEX.PageNumber 1 /PTEX.InfoDict 217 0 R /BBox [0 0 324 216] /Resources << /ProcSet [ /PDF /Text ] /Font << /F2 218 0 R>> /ExtGState << >>/ColorSpace << /sRGB 219 0 R >>>> /Length 10878 /Filter /FlateDecode >> stream xœÝ}Ë®eÉqÝü~Å’æ;÷ž6a CR7à qÝ4Hô%­nCþ}gÄZ+2O±$U•8ò nˆØ+2v>##;?~÷È?>þå펿ùåÿö»Çû/où™RzœyÿÅ¿ýþ3âïûߌ;ÿ÷íŸþù‘ÿó-?~·þýñ-§õÀãïÞü?ÿSK³ÿJößÏ?¾=¾“–Ÿ¥>j~Þù‘ïö¬õ‘çõy=÷øï?mÓó.åqþ5ÿ=üÛã÷ÿÑ+¤g¯ûÒ'¯åm>Ç||¼•ÚŸ}’üéí{ŠWj­l1È-Îé~æNúx ÞëmŽ@Ìôbéý@IùÅÒö@ïÏ™˜=¤Êz½qN«HZÈIžôñ¬s?òx ,ÈÌñÉSÃK%aq•g¹·½ ÷ûŒjV„äßõ94È#»VmëœôñÀʆ3yÒÇ×x1€´=p¥çµ³Ôù²KRv~“<¸²€<ós‰®kç'ÈãUõg]   ÿ[¦7ƒ¿ä¢v¯²xi ÇÈVrjÏ;­6£/QžýzX=¸ºeüè"ÞíÑöœ5¤ÝZ™p$\«?JF}^%påy·Ðê„?Ê4!¥5À¶¾¿}kïõpÍ(ãÍ(ÏQLÕ5QŸí¶„–B´)åöçuŽ„kå£Î°®ÜÏZAÐ|K“RZÜi«™Ÿ¼ßY¥²ª[z–’XvŸr­ì ¸ÚÒª}ͻ͓3ŸÉOÏÉEÝöÒ—g2¨wú~^)ämõW`E¹nZœ•ÙucWÏÝ·n§ ›iSN»ˆ=íÆ{×ËÞpärŒ.Éå{ß«&,¥¥?sÇ{‹3Vö,jUkçó™“wúÖ÷ƒrÛîîù.ù@ùKÊuãiršuT­Ï|oÝNái¦M9í"ö´Õ5¯ªYs«¹‰3’'´êÖ\˜UÜ1¬Ú“òôÆ4`ÈW9\ ºñ49—IlŸnuƒÂÓH[rZJìi7ßeåMF›®7Ë…œölÌ»Û{„ºôÝ«¢ QÈ»ë™Û–ßÏT6–”ëÆÓä”g-»F¶¶u;e~o9í"ö´›ï²†i{¿ÕÁ–‹8Ãפlæ°ïë…bE<sX½ÿ8Ó†«+ÓV‡Y­Â5++¨A«Yß!>u#„kUÕ4F·%\³qFZ›†!¥I)¬!î°UcÏlB8è瓽Zesöjgµ<$…q¢¹ƒ#ys§@XPЭ§sa lšö6Ò ê–XÚ’ÓRbO»õ.Í«U‰/§ä\6É+løZy®j½²¡ß¢^Tt—³I äÒ›“3Ü-¶{§)Ý=:U¥M9í"ö´[…÷Q–¯E9ð'.ïÑr®6¹Ë«e¬‘ôðÑ^òiXR®û]ÑðrË3°+o˺Aái¤-9í"ö´[þçrTí5=õ“ãÞùyÖµ®l°aÚf:—(øˆ«…Ž-Ÿ6,)×-oÕ9nA`Wû[·Sïô„=mÊi±§Ýz—Ûõ-u•ÜÇÉYõŠ,õå³²ÎZ õ ¸Ó.yõZ,,)קÉ^<Âv/év O3mÈe°/v«½Lë-ór=Jf{!çò1}ÕpÄ2 Ë#4ïêA2 äË£´á’XQ®[Þªs†gª°ÝýévJ­ËÓ¦œvûb÷ö¥­j—lóÌ“SQno‹º RŸ¢d{”Dq€•cV+Nòº°É«’t;%_ÚÓ†\vûb·üÏivOãã`¬ŸXÀÞ¬.˜¹<Þù¯”‹Q GµêQcd+=á’ž´:!¯ÖÒ¤ÖwÚJó×Lt¢“ŸhîÁ¹m¼-Ëófg³zÌRÐæí/_ÜZƒä—·aI]á…‰3­Ö"€#tƒÂÓH[rÚEìi÷7ø9Ò±:3"£Yÿ¹ú™Š9H÷¸Oë"ÞáõÒé>4q$bf&Fyö¸ln¦´fù³JRZÜië7DG,ª„î­Òk§<›Í܋ϧ׼Ìp(Íúm^-y÷¹¨°¤rÌÙÅñê'ìêÐ,¨žFÚ’Ó.bO»QÀ|q9îåbDœâcÂÍzõ1ÁfžY”¢1õÚò5H%•Õ’‚ƒîØ5Ó²%¨!ÞBÍ”Ò(O£ù"kLñüËæ1œœâ]ÍH* ‹Œ­ÇÌlPʸ6¶|™Ö6–T–wœd“[aW‚"‚nPxiS.»€}±{wKûšrdcšËaýS) ÇXäÛÝê×Vµ§tUþ5€GbhL£Ûè#œ´:¡Ùº¥I)¬î´uGÜ ¦[9GÄÍ9—áôC–§iÃ;}Pòa¸*ãrŸÉ–”ëVÄÍ9¾BØîJ¥»Ç¨­´)§]ÄžvËEƒÙîÇÉAC¼Ð!%4Ò5‘´P(E_ÌO¦|U\+lbEÕˆƒF,,š­t;%¿ÕÓ¦œv{Ú½£‡ë5W©GðÐp¥,*Q,vcËsÞK$9\…!éeSá@¤Lƒq1–帋¡,× "b†%¤np‡­ªV>b[Ñu?ÅAÐÏšÐZÓ½^†ÖY…WëÆ’Š1)8ú ‹€ tҦœv{Ú­wñ¨˜-LÌCˆsŠô–Â`H½ÑL¢duS’w_ –T–—œë€-˜èQwÙA¦-9í"ö´[M¤(ºÒÔDȹl­ŠÎXmµéS…jÜÝO‘|xHGXR3ÂÌâø2h`;¢FÔÝ#j¤´)§]Äžv½Ç³:îÌ^ óm2Z²fÀÝ]ÄíËP\ñ¡ ½²8ÂÏ&´¾ÃJ˜ßU¸E¬fA­ üQ¤ †ÌsÜ‹­ßàñ¬ž¨úHqO–0˜l®Vj#õôj¾zÇgÌ-l`¿BZmDßap0šºÄa†J­-fLRZÜi«â¨Ék¼-M †©È)¾økaËîeCœhQŠ)]iË›/Ó KÊuãir’Mщ]nLÏR ÏzÊ’Ò(O£å$ó_­¦hnF¶PRɾôÒ|ͳäb:A`¯ÖÿHêk>Âp­xÔÕûqâª÷ãÔ zš’º5¶î°Î@‹»s„uz ‹Œ­öÛÐS˜Ÿ´Úví×-ŸfK`I¹n =ÅU7Ö£¨¡»ÇRœÒ¦œv{Ú­¼ƒWÍÕý89z4׃úl†;¦Ò‹@Í%{*üy·•XRujÈñ}]yŸ·n§2j×–Ó.bO»wèÐzíÊ88Ãy \"|bNþ„·J¡CsÁ$ŸîKªGMœæ¾¼°‡¨»Æ4@iSN»ˆ=íV[Öú,DÁé7ÕÖÌÊšìvŸçÕ’g#´ÆÛ&”Zý½…#áZßÕð#[ô\8}HkROiR kˆ;lÝMÄÉs·nûãä¸ïdÕÔ#ÎÓ=U-ìJMÄBF’Oï÷…%åºÕDœÓ<.+luLºktÏJ›rÚEìi÷Ž°yä'[Ýý89ÅÄåÂJÃzoî¢"ÂVCny–+*«+ Nò&DìšÈôºAE„íÚò©UkŸØ­éååúj ƒMDœînR‡"s'áŽ*JâÌ[^!w‹T‹8©8Õ}a &›ÔíÔ;'ºž6ä² Ø»¿ÞÉi˜\u_Ûÿ8Ó:üExKöí =)~Ë.Õbd-¤Å§<À‰èšÊˆÑ“kN &çZx‡ì=CÖ8îÅÖoqrÐAbHþ89pVÏX?hˆ¥‹’[Ë–¯D’ʱ$NÂþ6ÇÞˆ‹Cõ½ÃæHYREàiô^j5o7ûjóÇÉÁŠ´­úOD—“ûK8%ߪ—Ûþ‚XQ5bÕâÏa3VG¨;ïÕ¦M9íöÅî݇ D®èB|ù%¹f[†Êþú>:ÜØÇ–côXÝÀ!Ç^?aAAµzc\¾»Èå©Úæ (Ë_9¤°¸Óàÿw Uï²Ö˜ÖD ç²¼á³nˆæzܱz&×ä>G•ë†Â’rÝxšœá5a—ï–¶n§]mËi±§Ý;ö™ÉÑŠ©8Ã_üÎ>l.Œe‘Å>.Qš®ä¾å +¸Ä’ê±b*NóAØ5;h[·SjU)m9í"ö´{¯e¬x¦+@δj°¯(Á›Z•›ã~¬éd¬HQ^°Z¬¨³/q²ï66ûNDéÎRÚË.`_ìV{Á’wó–ôqr.Ÿ±¯nã]Ck_®±•:(­Z6Q>³¥ÄŠšÚ$œá^´°êî;ð´)§]À¾Ø­rÁ^Û|‡ß@NÁôn¹°«kƜس[ö01ÝE ¹»’•õUr.àk›«Zè%׺×-§¥Äžvﾘ«šuFgÌUOúMæ.LEŸ™º(uÆVo)¿PÉåºcÅ4ùª§ÄvõÇ©‹ŠIÅÜò©ÕVþØý K[ÅòÀ\ôx¤«4ªÂÝOW ï~ðÛÃ1ݪeÃœbð»DØtâ.nÃ$3˜ú’¢¥L 2ظc<ìû–լ泶–cÆ!NõX&dQ6£°ˆú-*¢õsË/ïׄ%•cÆ!NòY±vÞb†nPrÙrÚEìi÷7”ª/K·"‰tµº9=×|9`úT ¿Í¦é‹½”Uó8‰Áï"¿ˆt6φ˜dµ‘ú’<"¦å4mpÌi7Þûk6m$ísˆáS³é³ Ë´‹¿Y¿–Ê|ÂG ~·X0íóbŠEW©¯h_Ó‚ 6sØG‹‹»Â¶ÝxDÓI>F6?†Ð0 V? Ð»¶žQC:àæG¢+:"Fó<$®šƒ!­Uþ‡Ò„”ÖwÚ ó§WéU@¬'mIe½æÊ• —Á¤ñÛ‹tXœˆ²i> 1øäؾ¶Ìj†t}øíÏyZ”¹ Äö1Ã\t[Fú8Ã\"¬1ZkyDë}¹}šîïÀ´žRÌê‰#áZYœÑ,Ò*œÏÒ¥5&ñJRZÜi+ÍoÖÜm?« hwœlá.=p`xšü­šÛ%»l*I ~—»ƒö=ÄdöΦ/=·¥•Ù«çÀö±Š è`3 ‰bd«xÇO’/íÏØË\´ €-¤Ü` \ì¶e.‚qû `·¯@'~¿#õ• en 1§™Ìë® =üEÒ>»¶ŽØ¡Ë“¾Ý¿™‡£‡ì¦)÷­ß~%=X•èÒ×£Ld°˜Ã>X¼<8[E_ƒz­ŠETx4«1/o/söêLŒEØz=¥Eµ–ƒš3ŠÑÍfáš5 imj2JÓ²Æq/¶FU)ô•¢ª8ÃWy'¦À%±ª”.‚U…»¦PUl^N‰ë¨*ŽêJÅË¡²xxoDmAš”ÒšÉú²me}¹}fP¼p>†—uBÍ,°·Ë"˜Ð}‡´ú‰8C1:»×{ˆØïf9=ÍÆñ½ØJó¹ Ú»~ŒfÝ>wMZ_tùÖé"BUùc" œˆYJF±i¦pÙÏPR«ì´R )­î´•æ#J¾Šªa!†;üu¸ÒÛ&@ЦÜÄ°ÃE8Cë¥btkÂù,AZ ù÷RZã¸[£cŸÃÐÇÃÝD²溗½6íÇó†Gýº¿ßð˜_¼m¤ÌãÄà÷îˆ@›2ãêÛS)¤læ°ow‹8õÍí/{S¶±L a·gËÕD0ŸïRDu‰#ÏÄH¾ý…8ß&­±ñLiBJk·Íl[Ù-N­rß«¾9¶cp+>oïúJìÁ2{ÇÉÈ>™1×"­ Ø-r7)ÇPǽØõÅŽzÍ#q€Ò³;ùÉJÆ*+%ó°1ø=X€ÑC±€üÀN+êëG,GÓÐãÈba3üÒ2ÈŠq[4¡ßZ6¸}ï”»·ö®ØèÒ‚AÞq".Žcø;â|z'­C{W”¦3dã^l¥ùˆ6u(| 7tÄ©º":ÂX X²e†0âH4mRÃÃÂydBZ‹Ö•&¤´¸ÓÖ ø†3ŸÀ~ŒnZ­–V\àSx«m-ò e)©¸p$\ë®ýkó=kÒZ´WiBJk€;ma4áÐD‹È—º\IÈÅÝ”W"äÃèo¡ozÏÀ‘HáÄ€‘|ñŽ8[›JÒ ‚oêÇ_/HÒÜi++ÏíöõpmÚ*Ø«VÁÌu‰âEák !}H}øŽ„kÕpÓÜÛ,C8,Q+¹kY Yã¸[£¯ì8Ô²úJgø9Šî×íØ¢uõ‘ˆ1Æ\¦v†™Ñ…± *kY¶¶Ày.Jk”Ò„”Ö8îÅVš_ýTLó›d>Æ°%ŽÞ´hÓ½óvÎrºO!%Xnj¡ˆó91ÝÚ !ôQBkÕ9y¥ )­î´5|0?ÝXbîD°È•Ccˆ 6[H €'¢hî$F¶™‡p¾¾ ­)fOL«´Æq/¶Fux\ˆë_¤½‡…t{?<;½8ŽTÓ7fSæ;܈™ô ¹æEzÐs3ŒFÔ#Ó‚ 6sØ·ƒ¨¼jQq}yÎÐÌN\4¶¹QGHá¡GwåŽÉ*—…o‰³†—´‚P5‡”æwÚ½êw2õ¡ÀvŒÆ°þÊ„î]a-üÍ9DK’Ýæjƒß5œZÐ…¡’æÒúò1û*]2ØÌa_ô/³8š,†~¿tM@w¯i6ì_üögð¶"àD\axŽýÞ·pÕ뵂`ÿâç™!¥5Ž{±uGRqÞaGRý†ŠÛýÙå/Ý¢¾­ìŽðÿð³¢”ù)Rbðû:#©Ã|©«s ^°ôá·"©U2ØÌa_Ôp¿­i‡¹Øpé ½Š` ׎|>ŽDÓö 1ª÷ýÄÏ}j-q1Ó,,›>wÚ7Pľî¢¨œ@4÷& B=[y‡¯áQ Š8±s] „£‰ëܱŸ§Òžfçnÿ’wÚú 7P̧M[‡2·¥ZÖÚÁ³iVy|{ } °zÚl|‘ÔwÅ¢ìcƒ``›qˆèPkŠ=9L“ÒÌm°µ¾Úªm“ÜKßâ`²8±Ñ—µéÆþ?9ï\ÖïØž$9öp Kjoba•‘EiS»þ¯û»µ-eØŒÃ:.­“ƒá‘`#n_×hXOƒ2Í`P:ýœq$ª<1 ò8KkD•&¥°†¸ÃVš¿&ض}l‹ô¡ƒÁ‰MŒ^E”'f{áï-ŸØ‚B,©.-8Íï)¶b;u×Ø»«´«ö½_ccO»U­PÐùÞÕ œ‚YEÇÖÿ‚%ƒÕ›ÛP(úÛ]Ê­‹i%ÝÈfrp±‹²Ô º‘¶ä´”ØÓn]X„Kƒ°¸ôqr°öÄ+†¸,µír‘.ÊÓËÕóVòæY(,©§¼ÅéXb$-@º[Ü̦´)§]Äžv«\:vz%ëå?NúŽõþ ~{ÃfPë¦@!ï°ëMòÛç™Â’J±“›¥–ÀÚL¹…nPê´[Þò3ãò‰Ý_¿Çgb2ãç‘>‚^CWòôôAîö}üæÜj¥Elæ°ï|ƒr£ÃºcÏ'9Ó÷ˆeÝÛLØ  Ã3¦q[^pO°¤ [Oß8`}·À\HÝ ÞiÉC.K§nîÜv«µ Üü´qà±ä©K¤*:_žÄ“¯c«á)äv˜ý ¬¨íEX¸4Ò½=¥M9í"ö´{_Aç ))ÎŒ‹Sp(”ëDœXó'v˵+øHMÕ©¸àðÓ3U¤}ËBÓjÍÂ+ŽíKŽ«Â÷gkú-Ýšó9'¾çAïN[wŽ›Ò¦œv{Ú½¯æöyYÒq¹àdeÌл‰B<µtÈñ¸±¤â 8¶3k+0Wè%·ñ[»„=íV÷Žïõ¦ÓQÁ᪠²¸sEghuJW³ò«D’ã‹EÂ’já<‰ƒ#EÂbü’îa=¥M9í"ö´{U\nЧaÄ©^ˆ÷ôdœîd†çØDi õ{Ëq°¤bep|O``“{ Òâ@¾Ò†\vûb÷¾—öö}L<õE†ßß‚Àè…}ó¼ö®¢ä±¤´å˜ëTëáËÿñn`;Ä8q¥+¡ÙâÕ%¥À®úçV¶ë€ÚÙáó ºˆ¼~át!ï-Ç¥•Â’Ú Íâ ­tN®6έ»Ç;+mF´ Ø»w„«qSÖŽp5Î2ÑUL¢x̤ޢájsËo$Tt¿bíÒ­ÉfE¤Û¿C,Ã^Ô J½]Ýa3Ú%ìi÷7D…pm¦í­Í:Ä»¥²ï4æv’¥œ v ¿á¥UÉwœõÀ‰HqŒî+<ÄaO3µ‚p­H“RXCÜië·D…êqßÇÉÁõ§"nž°ønQ*†6·«âÂ’ÑQ‹ƒÙ–°¸ZCºãÔ\¤M9í"ö´[M/aý¤ã‹‹g(ÜÏk~î÷›‚ö5­jeŽq`]147Õ#Î#î™wÐH÷¾£FiSžt…Ð,ŸØ½g¹‹cÄ,w0rƒ‘µ3&…ðì(¢4Ëõ«j)÷3W%µã.âø žÀ2ž@ÝMߘŠ´)§]Äžvïn$¡‡ážhql|@@úÆTǃrtò^éÂýs’_N‰Õö­IâÀ$¶Á¤nPr±|¿*å´”ØÓn­½fݵ«{ÄÁ’[FxÔNŠpª2EiïBï[Þ±wXR#îAÇo: ,6ÃJw‹:¦´)§]Äžv‘Çê…/×Aœ¡ÏÛðþºŽkHò-*"cË à€µ¿*'.z—@K÷¾$ZiW}–§ÔÀ¾Ø­w™ºvtpØçÒÖV^—ª¯_ Qïì)x]jÇâ¯Kåw´x]ªZ—8C9£ùýܺ÷=HJ›ò±ƒþ‰Ý»\¼[ð“o'§ê>è/µ#j;¦(YǬc!sÔÀŠ*ûÍ‹>ÖÓÆ&oôÒâÍ•6ä² Ø»[Åç¸m/ø5αàÇïó5D):È`4Ì¿kÝXR5VÄAüMX¾uÇÙÚH›rÚEìi7ßWªø¥y§à’±‰Ëœè†bj ÊÓK7.ï„<ãà ĊÊZ3gÜ:gÓDãA›z‹‚î©»qÓv {Úýywè/?m_—žÌ™ŒNNQ_CûÉ(ìšÌö1âɯ¥¸³ídBLm5âE"š›}­{‘U_*©N­ó\öÑà™±vsYWýÓß¡y§¶ÈKsåÃ,r¢ñíB‹ôÛ`Ìûʦj pŽ;oY\•Õw,ödÙÝ3®9éž\ÿ\õòµlI‡Z‰+F#±9q6J»>»4MsÇ}øÃ÷õ8™á…×kë!'Ë¥~ÓÉŒi‚[Er½~ûaô¡ªúHhUa¥kfXôƒ_´WìŒïAÝÜþUM]ÿDËöDÃÇ-<¿,RÐqýc#ísgà +Ú‘žŸß—úpvk§gÙ<{:kìí…x¾6í÷­tW~ÔÄç™_ÅoúÚùi÷ìÕ#¿Ó.Då›*¢¼îÉâ@qZûÌ»´mÃìµ+ƒy~e×»yrîªt%Ö Ô4|k,*â¼Ñ°žN?%Õøò;£–Û­»z¸¦¤G ACúÜEzöýœºÍ¿1²øåjÞ‡Ûuªï(õ¾Ô>šsûqk/vûâmò#QÞÅÙû1¡ pÒ:•—Š](î«þIŸ'(~DfYl—3a[‡÷ÍVÝø­ÇÛ³ýÎ8`ë+Ÿ‹¶#Û8}Q]~m„G­žSg ¦×’é`ù*‘÷¸ŽÞz´òæžäÅ1Éo¬Æ5¢—Û×=Îï_ë@Aã^qÜÉjáWBvêkø‚ñÕv-ò/ÆV”†ëîçÍ÷o]ÃÜx¡3úP=?B>|@+bzÝ}šmOë(Ù[/¦Ç÷±Ý ýxßâæ¹³•3¿Òdù ?W©Íº³{µŠ^vi\ô3XXóÆ8²œ•0ŠSÔ|3,* ®d‹zÔ©ŠÕ¬û—£¢vÿ2sÔQÛO9v ÆÞ¨¨ßƒ%ÂÚ?:*ÌKÛ°&ôùêßýðö›_þño¿{¼ÿbMp½Ðù×âA¿ù¯«?üþá_9IñçÆå÷þµˆ~øxüêoÖ@üëÇ|û/?¸êÿkç ²°éë°·/(96}yºÅ¿ÐëWK áôåéÛ‹^øëÓ½Š^øx_Êâ·Öï¿Ÿ÷»óãoÙÓø»·êà >HññÖüCo•>Æ÷”â»7’‚ÚÒ»")¨-õÉÔ“<äúÒå ù5ÎÄI~MÍýOe}~Z]Ͷdªft>ß¿ùƲb’‡ט‡ä–—êÞ™ä$·ü¯Ôà¦ûÙÅÏåzÅ/_ÜdŠ‡(6ôË›Ý |&úÅi–âó6²}ZῼoJÔ¬¿/Y•ókê =Š¹òÌzSKýï›?Mÿ+Ò8ÿÍ‘^Ƨs¼ãßÿ¶|É[F²ðvο–l±Û¯«yS·Ÿª±²÷™~þñíñû¿^•ŽÏ5nšýßÿïßÿðã/Ÿïš<™»X}ÝÝ^êÁGgí3ºÿ¾|ªùÌBËÑÂ?xHæÓF ÏŸ´ðÇz|¹µ&*ƒãÑæ¢;Ô<.]V¹@Yp¤ ¬„IŽÛk‰•¨gÕòUöÂZ ½…nPþ4Ó¦œvûb·^>Óå}Y$æ̲œ­Y#;–Q\nqú§Çãû¯-^~…ëKŠ÷s‹KÇÛ|nÞü¥Å›«¾ú?³§,Î~fOEê,^ÉQ|ÄŠªº:*8Å6×6›¯º³ÎJEÚË.`_ìþ7Š÷³„/*^fG/èo(^~úòKJ÷s—¡/ó¹ÉÜ–î=4ÃÏ2o÷™ý Ž’•ÇI€:öÂp©~]8?Ñ%­NeJ)­î´õß(ÐÏåÁ(s@å ò¥8¿¢Ûÿ츌»¯m|\óuú¿üøþüËÿøOŒŽÿ^2Ù?®dþéWÿúë•u_ýøó/xÿóOþù×ÿüøáwý$ïp²ÿõ?ÿ¯?üéïñnµoÿÂH­¸ endstream endobj 221 0 obj << /Alternate /DeviceRGB /N 3 /Length 2596 /Filter /FlateDecode >> stream xœ–wTSهϽ7½P’Š”ÐkhRH ½H‘.*1 JÀ"6DTpDQ‘¦2(à€£C‘±"Š…Q±ëDÔqp–Id­ß¼yïÍ›ß÷~kŸ½ÏÝgï}ÖºüƒÂLX € ¡Xáçň‹g` ðlàp³³BøF™|ØŒl™ø½º ùû*Ó?ŒÁÿŸ”¹Y"1P˜ŒçòøÙ\É8=Wœ%·Oɘ¶4MÎ0JÎ"Y‚2V“sò,[|ö™e9ó2„<ËsÎâeðäÜ'ã9¾Œ‘`çø¹2¾&cƒtI†@Æoä±|N6(’Ü.æsSdl-c’(2‚-ãyàHÉ_ðÒ/XÌÏËÅÎÌZ.$§ˆ&\S†“‹áÏÏMç‹ÅÌ07#â1Ø™YárfÏüYym²";Ø8980m-m¾(Ô]ü›’÷v–^„îDøÃöW~™ °¦eµÙú‡mi]ëP»ý‡Í`/Š²¾u}qº|^RÄâ,g+«ÜÜ\KŸk)/èïúŸC_|ÏR¾Ýïåaxó“8’t1C^7nfz¦DÄÈÎâpù 柇øþuü$¾ˆ/”ED˦L L–µ[Ȉ™B†@øŸšøÃþ¤Ù¹–‰ÚøЖX¥!@~(* {d+Ðï} ÆGùÍ‹Ñ™˜ûÏ‚þ}W¸LþÈ$ŽcGD2¸QÎìšüZ4 E@ê@èÀ¶À¸àA(ˆq`1à‚D €µ ”‚­`'¨u 4ƒ6ptcà48.Ë`ÜR0ž€)ð Ì@„…ÈR‡t CȲ…XäCP”%CBH@ë R¨ª†ê¡fè[è(tº C· Qhúz#0 ¦ÁZ°l³`O8Ž„ÁÉð28.‚·À•p|î„O×àX ?§€:¢‹0ÂFB‘x$ !«¤i@Ú¤¹ŠH‘§È[EE1PL” Ê…⢖¡V¡6£ªQP¨>ÔUÔ(j õMFk¢ÍÑÎèt,:‹.FW ›Ðè³èô8úƒ¡cŒ1ŽL&³³³ÓŽ9…ÆŒa¦±X¬:ÖëŠ År°bl1¶ {{{;Ž}ƒ#âtp¶8_\¡8áú"ãEy‹.,ÖXœ¾øøÅ%œ%Gщ1‰-‰ï9¡œÎôÒ€¥µK§¸lî.îžoo’ïÊ/çO$¹&•'=JvMÞž<™âžR‘òTÀT ž§ú§Ö¥¾N MÛŸö)=&½=—‘˜qTH¦ û2µ3ó2‡³Ì³Š³¤Ëœ—í\6% 5eCÙ‹²»Å4ÙÏÔ€ÄD²^2šã–S“ó&7:÷Hžrž0o`¹ÙòMË'ò}ó¿^ZÁ]Ñ[ [°¶`t¥çÊúUЪ¥«zWë¯.Z=¾Æo͵„µik(´.,/|¹.f]O‘VÑš¢±õ~ë[‹ŠEÅ76¸l¨ÛˆÚ(Ø8¸iMKx%K­K+Jßoæn¾ø•ÍW•_}Ú’´e°Ì¡lÏVÌVáÖëÛÜ·(W.Ï/Û²½scGÉŽ—;—ì¼PaWQ·‹°K²KZ\Ù]ePµµê}uJõHWM{­fí¦Ú×»y»¯ìñØÓV§UWZ÷n¯`ïÍz¿úΣ†Š}˜}9û6F7öÍúº¹I£©´éÃ~á~éˆ}ÍŽÍÍ-š-e­p«¤uò`ÂÁËßxÓÝÆl«o§·—‡$‡›øíõÃA‡{°Ž´}gø]mµ£¤ê\Þ9Õ•Ò%íŽë>x´·Ç¥§ã{Ëï÷Ó=Vs\åx٠‰¢ŸN柜>•uêééäÓc½Kz=s­/¼oðlÐÙóç|Ïé÷ì?yÞõü± ÎŽ^d]ìºäp©sÀ~ ãû:;‡‡º/;]îž7|âŠû•ÓW½¯ž»píÒÈü‘áëQ×oÞH¸!½É»ùèVú­ç·snÏÜYs}·äžÒ½Šûš÷~4ý±]ê =>ê=:ð`Áƒ;cܱ'?eÿô~¼è!ùaÅ„ÎDó#ÛGÇ&}'/?^øxüIÖ“™§Å?+ÿ\ûÌäÙw¿xü20;5þ\ôüÓ¯›_¨¿ØÿÒîeïtØôýW¯f^—¼Qsà-ëmÿ»˜w3¹ï±ï+?˜~èùôñOŸ~÷„óû endstream endobj 224 0 obj << /Length 1461 /Filter /FlateDecode >> stream xÚWKsÛ8 ¾÷Wø(ÏÄ,_zµ'§Í£Ifïîì¤=06s*K^Iζûë (Çr”í^b !߀9ž¿y{‹‘ÐLéDŽæ#!5˸¥#(Z{êBf0ÿX…/0P¿Æ@i–ðÿ‰Ž$$ý†g b÷9Ä@ æ˜K*›ŸøÇ” $0†+ì4°Q‰è Ã_I™DÅS¢ÙW$®g‡n¡k¡x©V½ú SÂHÞí=¦þ.ÉÇõ8ÓÑž?¼ÈP97ÐÖ|ñ‹hŽiÉy€î¾Y¹Ú”ŽÖ7‡Ž«¶i­+Iý‘‘vgžíÚãÍñ0U êô¥Y¨7âI°õâ}òÜ,¶ØSrYOÖ!¬UIº®O¥¾O¡Æ÷)>^M÷öÔ5Ô HJ •¿Þ˜À H„`Ãî^°Ûò½H¿ò˜'ðG¼‹%ŸÄò™VdU(ƒ;XuŠ¬‚Ö‰ÙSytæ›…Ô‚bZ¬åüÒC¿ÁžESÀõ 7X_?mвóA8ÌU™uW½®íÒ-(¡è´z Ó³~5fÑï4ܶ4ná&YÆ2¥úI@vHÀé]ÊS u>\~k, ©ÿíþaƒm†Ž-ÚÓVd8éÈ`ÚpÆ,™ÂªzÚ,aY¼ƒÊ^&ß-ë¨Gz™n6Ehg¶´-£n1«:-P¤Øë,ï5"/<òÓXSX2 >@€®­Ä_î±ÑièñŒ~ t¨Ópk0h™M»-}í{ôÐïÿµñ :õépë ^že(÷X„ Ó6Û{S›—ý#ö+!Öá5«èt OZÞñ-,š„ôOÉÃñsôöÚÝÆÅ~ñÃ_d»Â‡o­ñ»ÏÅO›¨öAxÌ(ïáó=píMmáSÄy…¡¯\i‚Lû¹óÚ¤ÏEö«)£HºÂé˜ð^äž$Ô$ϲIžì~-Ô?’Àÿ°¬  „sF¿'{)QÀ_º^Ç• ·1-ña拪*-Í † „<¬mU¼¯÷+÷]뮼:]l ±tmH§**,ô%›¿¶RÎhõ©¤ßÙ¶Àˆsí|]£vZšâgãZaÆN–á º›°º¬]³*3+LèÑУt:ö¥…@à^#òá-võg`~Á;“ÿW¯N—˜§¤$d7AqVÛû 2²ÊÎ* t ©©¿Ój°3 ˜ ׇóÑûÙ«\îW®éqŽqèÑ%’˜ ž>q÷'!ýãàKÅK½²€5Ö<Ñâ”G•ÙÂs LZ´´pëͶ¥¸^ëĺW?“ý¬ûѽàºG4ÒB"ì„ê;¯8‘gDN•0ÿüåðO`®È™Ph{s2ó/'÷›º endstream endobj 253 0 obj << /Length1 1510 /Length2 6478 /Length3 0 /Length 7479 /Filter /FlateDecode >> stream xÚt4lß¾Y{ïZAÕ±÷ž­½g B$DìÚ{kÍVmJm5ZJmŠRÔlQjÔ5J)_:žç}ž÷ÿ?çûNÎIr_×oß×ïæd30P# 8J((,PÕU· „…E……E99M (ä/œÓ ‚ô‚"à2ÿ°PEB@(4¦B¡ upÀo( JÈ%e„…"ÂÂÒ"25 ÐÜAÀ!^„œª$ÔÙ…Îó×_·#(--ÉÿË ìABAp€.åqGgtÁÆG(åÿ¯Ür.(”‡Œ¯¯¯ ÈÝKtVàáøBQ.#ˆé~¶ йCþ´&HÈ 0qzý&ŒN(_@0¨#î…vñ†ƒ!H:;ÀX[ ïÿ6ÖùmÀø3Pøw¸?Þ?Aá¿œAŽŽwÜ w8Aa€¾†Ž ÊÅÁÁ? A0/Úä‚Â@hƒ_¥ƒʆºÃ?ýy9"¡(/A/(ìgB?àǬ«"ÜÝ!p”áÏúÔ Hˆ#zîþB.× Žð…þur‚ÂÁN?Û{{™Â¡žÞmµ?6hˆð?˜3–‘”@<?G¡Ÿ Lü= ¿HàOÝCP Âà„nu‚ ½@> é ü'ñï!CQˆ3NøŸèhâôûŒ¾$Ô`-Œ– üóó÷?[´ÂÀ8Ìÿ?æ¿®XHÓRCÓʈïOË“**?@ €ˆ@@DZBHJŠ‚þÈýSÈ?œµáN€ôïzуú«fŸ?"àþ³!<€ÇÒC ¥ pÿGé6ÂâÂŽè/àÿYï¿\þ2ÿåUúW¤á ƒýâ¹ü?<È óÿc–®7 ½ºô2ÀÿÛÔò{wu!`¨·û³Ú(z”áΰ¿ õÒ€úAÀP”£ËoÅüÆMî ‡ ¼ ?_€PXø¿8ô‚9º¡_/´,Qôþü;¥:Üþ¹h"â ò'D_3ú$¢7 ñû%d€ B»ÐíœHŸw D?cBHøçðU Î?É?¸Ä߸ê°Ôß0Êã°äß0Z(tο()´ƒ—·»û¯ õò€üÿC ÿƒD{ýNô¯½‘Hô¦ÿ!ºû¿Î¿žÄâH87p”t}ÙvV§Ìè+°6*ÿžsÍ<›G pÙîý/ƒ§öIø'ä©rÆ`7ùüŠ:÷‰ÒÖÛ-x1­i†¯.î_Ú=0_{E8;FÛÿ®d[¹¡™€IÀDiýþÏûfanØ-˜w8 <½¥H Š¨Î|{5ýú*>¾ž^3\¯•¸KtY1!dšhV6ÉYè;EÏ~%ÀŒÏKyàG6yrúž2ÿÝ5ë|„A;I¢¥VK"ÉçS U&"^ ·¬è™±O(ߎßTÙȼC7X^ºX<É–ìäq­©h}j†ró~Â%N«Ol~±¬QTÉØQA§¼5CàhʆõÍßf<©RÕßÅcÂŽ¾mciÈáSEâ>h»¡..Ùªü©˜/\Ôº¯4ê±d¢h_.­Ìmo½B[•U‚; ¢’Ñ+Ÿ‹nnÊϱX’w0´I5ô…pݬxT±žº…¥T>â•,µðöñ}ÜK’X6}æ|{:¡ÊôršO­Ú\¹´oéÓdË.óêÀ”A®Ÿl1§-2—B©d ï:9Ÿ¢P/£f)tw-Øéû±Èm©Ùó@Ÿt Ý㢳³Ï/Øs‚«¾à>Üò}ìSõy`2æŠØ´}Õ»­Aì~ÇÂí&N2õÔfŸ8F-/)v‘¾ßzØá[¹Ç½»/K3á¾,gþCßbžœ œÖ#Ôò1 òÆGZß™e¥4lǬg\Ú4ã»éß;(‹ü_„Õ÷ñ,Ú{kS(gŒã9#v8ñNà™k)œ—ô<:µ5"´±/Êìvèm³àfü“ß¿Q1E´»ZOÒh=Ÿl0‹Ò}èê U:¢2Œ:±RY{Ì}¶].uz"ÐÖdm@ YÂÓz&cG¸Ÿ´`4þzkˆÕUYÔFãÆôñ°vÀ7­¨guþ·²Šn!q‰ç…êý\G¤DsÈñN>î-+D(Ü\o'“þ¦7þÄGàÃWŸâ%žK“b5^æWÚ:AùÍžLuŸË?:øÍ(|;eö& jO­¹½)3º-³bÙy4Žyñ¢1f«D=-¡ú±l×:Eù^滧$±ÖYBdkêF#²o6“dÓIÙû%¸ûž5ÜeTTeQuváy=E¬ jÙ'†>c]è” W¢[?ˆÎå¦b•û€›”ÃxŸ/š'hÊ 8†Zäö³z0K¦§»ãå'ŽmR—çÍ=k§í Xãæ¼9âHVøê—·÷KYcü}’Ä„k.º·m}ôc¬(ßÅ éqhïÕpaÚHÏû^ôDÔ`{¤]á¶ÚoµC«wp¸WE†õp0œJ‹¾Ö&*· =°V|r>t”ò^Rï#OŸê„§ä^¥º6÷½ù‡ºÖdŸí/·MœÊ­_ow0,lÌ!ÉU§Ej:,?«ö• à–öU±Tc;­ý ;÷†W›ì9›¨8ÏÝsjÔ§.1¯"Š"-zwÁùKÔû`c6 ¢Àxz®×•:“qFYõ”õ›g9pÖXpxºˆ/¥MøM\ˆÜò·2Ûæ·>y5ÓîxVßöJ ‰sjY/äÛÍÍ\,Én]V³‰„|œoÛàÚÖCJ0Ö§8$¸•ØEÐx6¾‚¿YwE›O×ùìæÈò6W§R,°ŸfY´)ô@÷îBj¨Ýé…Ý\¯&hÁ¤CÛs(~Dت\%pµ¤<½Ö~CäÎifͪÂ@Òè=~ÜèÝ*ÉÄeœì™”TüñYº.rÙ7u;Õ|Öb©*G¼LøžbÁ²’ jqì(sE,ão¯nψf”/Zë¬ížWžÜã L¼4à1{Û\¨ÂŸ¶²e6¹HšR¡†ÈÊ+½—\„¯• ¼Ï2Å2ϯNVÐô=œoègí‘© Éh«$7QôE”Í]‰Gé—žÝ=†?¨` 2t5Oí=w'ý÷¡ùGvúůq$7©?ëÞeë¯ãöÚ¢w’RÆ¥å»hµ­àÜXÖ;¢u£”ÖȪ— Ͷ&+E¤ôáÓ#™ÐwU··R¸Õ"„…_”8¹©2ãQ[‡å×dÎû1$.pÉûEw™‡h}}F &ô"-‹§¤ùQåsž”TÜ _Òj¶UÍ™¼­³Dÿþ;çà †À)ÿ…l, †KvšY õüUÏ~.™5 |Ú+ ’·l+ ¶&˜ù5û‰#ê¶]¦jß­¼•\m[µÙìþÊ ®¢”È£û§jš1Æ7z oM„z`2VLEvàIÈ]9« Æ;`WÛìpžÊR/Ñ"&q„ø¤Žj–æ-sÛñnZ<á«y7¶UW}:õ:îu—Œ]¥…n×üia87eY0êF°_bïƱc9_…½–ãƒUjö'|cØlëÎ+ÉØÌ0rzYN·ÑŽðÄÇ>;ìá/9?ê>ëÆQ~(1­õ†³5Ã6‰^í⹽캂öv[Ãzqÿšs3Û“BUcèÆlO6ÂilÀ™.Fï-ƒ-ðÏd6=X(š9°åãÒvÉâîp¥-Å9U*-ç<ò™ý\eõ£Ö•HOå¹mD)õ ÿK艹hÏx`lâ›[ìT²Œ›3–—¶•ÉölPàaàåóz«È:8Ùý#ŒªH3c~ŸZ‰ù¾Ï‚ ÷[÷W8©çÆ'>ïò˜„¨v|Éœtc/~ ÊJ½‚x®Äkd:ß<-»*ÇÈ!Ö-†oŸè§ÖЋ‰QŸF}d Ô”à‰ÉŠ®ylëûe1ï$GU¹"/ƒËÕ¶Pþââ®X,î1óžºr|Iý¨Ì°®þ$¨‚ÏÌbÐzå©ÿ³;hŽ·%žþሄt–U|íPÉ«¶Øˆûµö,Cï©iP>‡E¾³£Hžþ1fÖ8–i¡Ul}Ÿzqì>Š.%ÿg áÒÛVŽ±kÌ™ñ7çG˜~ÁÑûlÃŃrV'ˆdA”ÀŠSëÍÃæ¥à¼¡î¢:Ή*£Øû6Ä*л¼zœÁṲ̂•v·æòœrL«H¬ð[Œ '£Ô´ÞïÏ?Æß(Ìëq°&¿l9FŽÁÌ·CiÛZÍ…pȆÖ]ä5(j%éѽ\Ø]ÇvaÌï×Κç€A÷7üÒèæðo]#©>É5¢r⡘ ž‚ÝwÔ7Dòå~4d>„ ||:—¢½5Nòli†&1qÍöóúº.~l£o2uZ…³§a–$™ZÆ!{4P¦7nÆw2ÈdhR[®ÿˆlýªÙ€°‹hn…a ´^ím&°`/§ˆ4ù0DÓô]Üñ²ž Ò¢›>³[;T룃cBô”2£Ã†b54•Á_csþhËÌsa`åø9X<ˆmã‘ mkKƒ^¾úÊä)"3G©Û£îGÇ«ËÃ1IŒ«tê1“‰ÕÒ©S´õ†Â+É,”AüBEìéõæóc0‡6â’|J'#©Ü!)³DE]YS»®5¬¾"W[Ä\”3sQ/òÆÞ» Þ’ûgm®Ó‰Ãdó¡|ÜþNKT{.VjÄà°–ÌVáöÝõWÄÞŹ`þ^Êxœ÷”b±4õïœd†®®w¶Ý;H?™xÈÏj4唺Ó|¥Þm“Ss]Ŭc÷yfš5,+³¶omxhô~Mz–É ü:úÜ·3cÈö Ô—jšá³$РýG=8[ü¡i¸S±J2¼¸Nå~ÖåVæþä`¸o›²·÷…@“>5[ÉÖëíÁðòN`Úy5þÊ9Šæk„®UtôC7Uá7>~:+[©=E-Õ¾Ž0;õ›;³—/GC5±8Ïg&ß×wu7SnU?–·µ”¤ä}’<Œ0¼°™οDöTœ:c.u¥Y:ßÃoò¾2xh|Í”Œõd)Þá“hdå¦J«„œøT hä¢ïŠ{E‹unó-{DÅßÑ—4;€6—´È@ijzÿ^kF1µo“²‘ Ã4?˜Úcó­z8ÙÆêSÁ*·í´š%G1g¹éÍWt«U¶ZÏ­¶m:°ThX*£Ä]Âs‘ÍÑd êÅ‘4à{T}¾ÝñXsЦî€"Y¼bzvž-B[{»zÚgVz<)L›ÍÑcßì'õµsÛøR¤Éi\GMuÀ|±MiNÞémÃ¥½ÁõN,=^š‡º£c<[F–"˽ 5þ"à8yœ—£ ¯LÉ2q !»Û³—_[0¤¢•xåx›V×ÙNRRwüóm \’éô¦Ç*ÇTo=C{öב–TÎ{Ïët“> %RnuÛY ÎŽ®¾z½ôæRþäá ¡â¾io‰§à 0ÿj¡Ù&ÞýöÚŠü!IÇ!m%)„äÉUÌ5~xúÉá÷'&-gEƒ»ûv1QÝ®÷Öǹ¬á%p}«býÕ«IEÓäôŠ²Ô×¢²HÐr9NYPªIŽÀÒ˜Z>â9“þª!WÂP½Îè©R6ñ³YA6S¿±¯Ád#WySùÑR5Ïõú»O?=¼ç-½y3â™3ª&'ó ÐT›@±¬ãÐ.(Úvxé¸z…ËùÁ϶r&N§¦º«›ø€®} ´ÛœmäOòÑêƒÌTRìAòëÓ{açð™rR¶a!ä¥iâS¬©§{b<}.ú/êâé‚{GÈ[Ö†\MoÅN…,.â^xêè…Ù÷p”+·«šN ÈGËÍaÉØ*zdñÈ|©Üñ…:Ë/ÁHé—„>Å-G—¹¼L¶ï‘¬Ž„Ô–8_£ôÉÛ?~w,¡<™ò${-Ž›Ÿ•iÓü=“»ÕŠ6åÇÓ{Ÿç¶±¹uÊš‰€â¢s®>1Ç.ñLS›Ò›/Òkâ׳%§©¬NT-Oo`®%5„¶¼ÕïZªág{ñð‡)u”ê-Þ.ÖšaÝœ˜bIúÑǯ9øJüÓ2%ØÇŒŒ(d8-t8;,óhŠRÕâÀ•ÕüÖ‹7.«jjb4ŠL¶ JÂb‡-ÃÇê;ÚeJÒ16{‘Õ×m.œ>Ò'\(Z¼SêÒÆRþŒÛß5®=ú&K?NòK™'¬mg5Ÿnk´¼ÓÈËNë0,Q7oSS0.î•Ç!hrC Hn$¹_U/sqk!ˆ/É»Ymž’—sŠºXcêÐñÄÆV½ÜÇ ï~¢ŽâyI²~MÉät@#).Þ/c„[*.¸|NŽèù.å‹G ˜{óÀ!”ýIç]ü{Pñ¨é.Õûßl™ï|YÅ¢+ô@¹öÅ÷yˆ~CØ—ý “(^MÁX«h'9k`nŸŠY”kÓ¹ÒÝÏZbò*ð†©i6‰‡j‘[LÜî÷zù€Þ÷ ïáYVáÃîáUYÆ­U/îê$•¹×„>Š3þNÊ«‚zïqŒSÒEA™ËÃãÌØ$×·!+k^,^µê%ýk¢ ·¼® ¶r|ÒyKÝlñÅO1Ô.$Ö“«üÖE—?r¾«U?*o‚š]w\»ö¿/¯m)K7(Ž`Ñó³~|J´?ö ~Ba0óÈÌaŽ7æSš£,µØ!QH{™«³{ˆÒÜs7¸ÌõåÌmOáp~u£2*ñjLCf®sû@쇌7•ŠÙ7%“þZ†HiÀc8[‘½Ç+g_P¤æ¼äUú‚<‰ÃfÀ–×äXI{I¦¥ReRs¶p| ô¥ÊžÄ­¡¬É29¶øäÝæ•¥ùtÙ¥zã`ÇÜQ}åÕ‹OÕšú‡#ÙAf8e­òn> c •C´p1yZŽûå.?‚Ï3ľÝ°“¯X{KÀ4@ÇAtxùþüp)†¨\{'¥ã°qÂS‰:¶÷}àùé9Ô&¿Õ.T¡ x¿³»`ñìμ{ð¡(v²tÓÉùK¾aÌT;šþ$c¬7zû¾B^`Ds|…±Mh·<ûÝ$dÍÛäñ»¤nJ'oÄ/?®&=rfh,~P-‰=šÅÕ[Íç²4ŒÊ¶õ•q× o_$W‘uªè=Îmø|ãiÚ1Ž`úW¤}ƒr‚ ˜ùÚ½â~õå³#Ûyî› ´~—8®ºvRÜ·ðâ²A6-N"̆Ä×fµîu-Nq’QrÈýçá‡I"ë“1;'лdï7Dúª³8¿ÿ()3f8yÍ1,¶Š3Я¯É„"Š¢©_’yn~JÛò½ ©H¬2ƒ Z›úHû†wðRÂêrŸ‡«cyåÖù±û˜vBê¯>dÔ󘼴bk=:¬‹O€¾—œ™×ž¶y “;ìBßÆð®9{ªO§ú[Ó{ë¹bu•x”¼¡MüK-¯G@<’Ôæ.F7xovð6ñÅìn8×EÃéò¹&¥t£õ»GT†£³wíNj[‰E†ˆÈUx=É[a‚>_' üaàE·Ð ¿´ËUJö~å£[î\– YúV­ÄÛFä°É9>&`°÷Úš.ÙöÈIËÚæéŠ ¾‚Ç4$)~Œ¡,´BõBß²•ŽH÷£÷ゔɉº‚ñ{«úÏ`ÙœLÛÕ’ù¹9Tþ nŠÎò:ûÞ‰ºŒëWÄ ¡Žû½¾`dЕ…çfÝ°A(Œ€ãsâÓÔÕ™½VËJ˜—ˆ(š=ó;´÷ícó„š…ý3jðĽ"û™ùš´¬~‘æ÷M׈Ôfª•®ä7ˆkúÒÌoöÎ;…Uêt~ ]I>É49M;9Çï·üYGó‹LE~úf¥ÁÆ7;Hög±½õµz$åo.Å]/–—ªn9—ÖŠõ°²Ä¹ðubuvø³–ã÷œTfô®¨ÅÚ¥«2Z† §æx鞊qí5ãH3è%Û$ãý:v#ÉßlÅb Gš½øC®S&@ÍòýÙˆïûÄhÑDFAþñÿíD…Å endstream endobj 255 0 obj << /Length1 1766 /Length2 11433 /Length3 0 /Length 12568 /Filter /FlateDecode >> stream xÚ·Pœé-Š;AC€ »»»»CN5Ò¸w·àîÁÝ!@ ¸ ‚K€KfæÌÌ9ïUÝ[]Õý¯íûÛkõ7™ª“˜9Ø( ¶‡0±1³ò$””äØX¬¬̬¬ì(TTš ˆ-ð?r*m “3lÏÿ/ '  äE&iy1TÛä]ll6n~6~VV;++ß ÁNüIW9@‰ ¶:£PI€<œ@–V—<ÿyКÑØøøxÿpˆÙ@f&ö%ˆÐî%£™‰-@lB<þ+­ âÀÏÂâææÆlbçÌ v²¦c¸ Vu 3ÐÉhøÝ2@ÙÄøWkÌ(TM+óŸ °ÄÍÄ xØ‚Ì€öÎ/..öæ@'ÀKv€†œ"@Åhÿ§±âŸŒ€¿ÀÆÌöw¸¿¼Ùÿálbf¶s0±÷Ù[,@¶@€Š´"3ÄÂ0±7ÿmhbë ~ñ7q5Ùš˜¾üQº @ZL `òÒá_ý9›9 ÎÌÎ Ûß=²üórÌRöæ`;; =Äåw}’ ' Ù˹{°ü5\{°›½×ÈÞÜâwæ.,Zö G œä_6/"”d–@€‹•—‡ƒ— tÝͬX~'Ðôpþ¡dû-~éÁÇËì°xiè²¾ü x9›¸' ×¿ÿPØØæ 3Àh ²Gù'ú‹hñ'~™¿È ÏúB?6ëïÏßO/ 3ÛÛzücþLjYÄŵ5þjùo¥¸8ØàÅÄÁ`bçb°±ròx^|þ;Žª è¯:þå+goüö»Þ—ƒúOÍ®‘€ö¯ ¡üw0eð uÚ˜þž•‹Õìå‹íÿ™ï¸üÿÑüw”ÿ+Óÿ·"i[Û?ô´üô&v [¿,^¨ëyY%ðË2Øÿ¯©ðÏÝUšƒ\ìþW+1yY1{ËJ3±q2³rþ)9KƒÜæª ˆ™ÕŸ´ùS®õ{álAö@U°3è÷óâÅÊú?º—-3³y¹Fœ_¸ù§ÊÄùeå  ò7¾,Õ×!eo6ÿ½}ì\Ü''”—á¿ .€ÛËššÝÿ`7€…Ù yq¼ôì°;¡ü47€Eê·èÄÃ`‘ùñX”þA¼åЋŸÊ߈—À¢úz‰¢öb°¨ÿƒ8,ÿ N‹æ?ˆÀ¢ýzɧû7â{A¦#®—ì/Éß6Ö—À–ÿ‚\пàKhëÁ—Îlþ/Ìg±ÿ|©ü/øÒ€Ã?©_ qx¡3Øü_/Éœÿ†Ü/îζ&ÎVÿ2xIù²¿äsÿþ×,Í\œœ^†ýǾ ú?øktš¡,̓Í‚¬ëƒ:nkÅÞº1íLŸ¦Þê²3M"A¥¦64²2—Ê¥—ؤ ­»•Åo³×.¼¶?‘6xp^3‘JïY’šÆÍ?_CÏ&zÝ‘Ïc´@å餈“ð—8 B©Gbö ‹˜[jPaxÓÔ÷/&ºQã–JóÈw(wõÔ–f*âñi}W‡´šŽhîp]âÓä|·‹ ÜŠzþQ9\A°ë%Á­scxtdk€¸7Ò_Þ5`çŒ2‚%%0B$ɉX£½’ÿêWN‰ëa:¶—© Žè¡ù`Å:]Êm½ò¹Õöù¡mR•q’EÛ‘ç1–à ®xªSnÁn„ù·ìPRØæaáÄgíþ4lßZÕwbùÐ`åÙþ©_. åô'‡‚h•„ž*·|©:£]-B`ÚhjS°¶¥ð2é|,ó‡(m³V®iƒguØõyS­:4…‹Wø‚Úüt±88Ð]A©ÕZ¯X #é]{yr>å_tu‰S/;ÖDÚGcGyn|ZñN/Ì»?ecùÜ`–V%áéf·yí]ã¾i 8l.ò©òì°ã:ÞXÀŠ‰óH³Ëº#‚¢ñ™ÛÂÉ«¾m „]‘TÂÄ™C•*çVÜ_úã}w@q ¸–úGÝz>¤”Q<áËU_ hOw[rç¢Â/1>ÿnû¨â'Ïi'Æ:þ¨ÎÑ4ßJ;âÕÅ׶͇µ rÙ™TJÚVÅ°òaÂùFƒ€«6¸ŸÕ!¶qÙšI7S÷vö­£Tȉ™¯ìGæá%‡å[‹L¡W¤ú]F¤Ï¦Ï`£×gâDÚ…ë÷CcˆMç:ä¤Ç0ÛMY}&çûàwÒÎ*ñtAI…T§æìç­nQÊíU3²ÂW’>WgHß²âõ’_u7Ô±Y¿«´,}…ÐÍfÚÕš)_nâmÜe]bú®×M/ðÔLŸšÕ:é¬^ŠÇãÖ'ÕŠ@£98Ÿþƒ¶ÃÓôż!ÀäaÞeð­-§§öþÅž½úÅF7±Q4P´æ=Œ§wJ|²ˆ¹™oœ.f8-CÝ{ú-Q8à P¦iM™ûÕûeq©±¿ãàÇÏj~;"¨'3åÅvûl¦-GP ËMkä¤D~ 6÷ó \-W> a70™l^s^< qJkwõøVceHß+«°TçýþrQ/Ü §‘Pé°€ƒ‘¾q •ˇ¹¼ˆ¯øMm9·IGÖ+ Þ~ºO,w»B 뉑v-ÞD{â·q–îß±pÖsñ„°¡'Ûa·„5ÝÁï²zѵCpî®Ø8>~ÛÒL”䦸d´ž6’jê¦Oïö­ [²lP·• ®9«f8ïÒ.×?­<[‹Ñg[N‡ª•×Õ#Óÿó¯UC×" ’p0Ýû"*pÛ/Åܧ–Ö§DÉÿ|wW,2íqœÚÑftYŒˆ5m¡‚¤³öf«4§ÔåŽuõ&>Ú†%¡šº†ë íóté]ãÜ"²Óº¢žö»rÓ#NÝlýqV—qkºÔkt¦Â¾÷h¤³#Ò–­½ ßD@r‡À@:òÖÆ$8WÏseÜ"µÎûã‚…ýª¾Œ6Ç UÇϲjÌDÇòê&Zøî÷LïÈ¢H²åûQd鈜Ԯã‰Õ¸® >hâ›*ú°OpÙ`áß©¹å«<ñàqý £xËŸ¼À˜¯e„§¨*ß8<úFY¢zWCézt:‡¾,œü khò&lmç'ý‚5…N–äï^$åùŒ÷y6Y³tÄúèxDËáþ²b¢€Ñsµy9ʦ¶”~½ÚE|<;uÇ=œ›y}ü_‹yLX>Æ” îOßß±ÝÌøºPÛˈ4áEØbE‘P/æÅf×=li¶¡['!í¹¤°§ZqØ'G{È=v¢Óôxqš¸|qߤZaŠtT,™–Sæ{£0je÷vñ_™È"#½÷,|ìÃd™ž! ’²Ö†÷Êb ¿`Ô¬KÖV|œæòÐQðöô›¾Øúxæ½ÜA,´7ßÖªö\âXîÑœ—÷íØÕ¯ïj|šã0~tÍ}r°w ÔBSÒÖ)ϸfþ£ÇÙKKݯ­ªg÷ÌšûTŠ"¨Á ak´X5n7M߀Êiƒ’;Ä›"6¢Ëz—v1kÄ”z¥A/Ò2ã:ú"±)äÂvW”üÒ-–ÑôáqP¹˜ø«:@1ò8ÓbÈ}—]’6%`Ü>·~y>óh%º|rª,ʽ©#:ðû=³ÓR-,É…”~låcóÑÕ«½ÀRÏΔX§jxŒƒiÊ"ñrTE–ˆ¸ü-âxW€ùµj?™·ÙÓl.ܬÎÆþ( —ÉÞém‰´uxœ9ë%´ûgâõLX6í­ÙC)6E›—í $Õv‘«¢kF…b4» =ïÚfz°‘)ÕÈv³’ú°UÑ=ѾcgÕgᨺ+‰b¼vbžKW?2áÌ»”È·Ó|ñ¦J»Ö;®?¾n!)~å—¨ü5§6f`Ý[ €xÂÙ#¶~3;ll¼ÝeÍÏ‚ óêݧX¿ç'­b*ú{î·ï/šÄw7—ãGvT£(`­«üz†öÝú!Á  Ò5Pu¡|ùlvó¯Y@§VÑàØ™óž^[–u°O>Êû>ˆ¬ÄâéèI Šäæ }Ù ØBKƒ­Ì«yÛu0ùèÝÇìõѬ6Ôö–5¦ Œ¶oùƒR*ÃÉ¢¢¥w-O”Ꟙ&% ”Rå Ò²ÑvPtëoÖ 9à+!ëê;  [JV*w²G‰4~?¾ýFLí>6¾ŠS¢?Ž+ûÐÇd„ +¬èlj2ZAȼ¦3ñ·¢Î†onÝež™ãC0ñUnBÒßEžX×lIãj®t|Eg¤49 $TÿV{ïl‘ÔPP¸øÞÎÏ̆ \Ç°û)t|wίWê Ôô4’Q¢Úñ¢Ÿ:äQçP® ch6Ωóý·¦¾€ÏÏßÔ`cׄÍ6'ó7"G“³Õ5™s¯êÊæ[uà¯Bf/lò'FÄrI1i— ”¶ °01q¯Wâ ý«¦æT’ëlW'¿z]–zrIEùxR>Œ™¯±ŽÙhD8ÄvŠÁX•89£w†ÃËI|²‰ìh—‘²š¥Þy (ÀÙ¨~kœœµ%ÊI– ~/ª˜h¢ùÑ@‚z„áÙ]CCHß{£%Â$5E¸œn×4ý¾Ú͵]ÓÜý«“=©K¸âHÁÝ‘£¿h㳤„¸*c?î«÷”ÎUkÁ¸Ð®T¼hM[ó)»á†Ó§"þZbYáJàÞ²êøê§Yš›Åá…¼ ;Ì}¦ŸJ›bžÎÆΦ>qÙQKýÎ-¢¸F‰Î,0³·¤Ãºƒ0T¤ú*>ÚžD®Y½—16…¾6!á-Lo¦¤];1ÔâÀ©qLo˜½:ݺ€'¯xvíŽl¹²~‚ŠI͹_#žiÞåb.uæï2~+ Ë^ÜY/bÛÝPo†‹½ý¥ZtsID*øvÀa‹Yd!dŽ7 IN\y´y`ÙñZGØŸJgLäâÅ~‚Ý:*â•¡Â w˜èø|8Ô¬ô“&ùÞþ±ÒLÃïð-YyBvkrÄ÷eÑ`²”&ßmIä ^´yç ÏÎpÄ9†µB–gùñ þ«"w{‡ žˆÑF‰N¾/ÿ¯ekÂY ¿Pû/ö¸%Êxd,cZSï>¿åǸ⣼Bü9|ˆO™AßÌu?bå„qEˆL ¬Ø{å7­3È\jî{#óÄ*•º9 U¸&ʶÝTíªÐ.wD‹o©©HªÂœwøaY¨ íæa°>ÿ>sÅ™ÝÑîäI$›y±H¡éþZÀëêJ³•33¤'§/¶D9ÿts¡ËæpcÀ|ýWÎJJh8°$¬È ÁÀ  ™ÐãƒÞv"ðpœÁ¾JPM^ —öt0¨\ûe.ÁÌåÑ£UKÝßf6îå¤q4ümÅÿî ï]ÕÙ\íÎAÚ}ϵ©;/×Θ~e}ÎøÏt3L(É´ºqÂT#ˆÍìSFÖ'1UII¼téÝiÚíêq=!4¬˜Jmuˆóg‹:*Ú•‡ Éo²äìÔð©L¨ §¾µá¦[—dB.RjûI‘h%>Y׌6߯/Éíܱ;ãç7­?ˆ5X†È$~QÖŽ0bŸýÕ¶à„ͲºÅ3'c(!ž»¹W‹+´ê7eœrŽ›iëXEËn°Ú¦ |›¥°—Äú‰øH¤dÚ-åæ'ªatêù+Ü×Xƪ¾ϤB «Ï"Rw» o32¨?1\„8½§dG¦Ô_ˆbqŸwàè‚:ü®§•WéßHcœë(»C­JÔâÇÑ;8£Ý¡K]ÛÚ”˜ŠÚ«1\Œ ¨ÇGÙg'„³{L†Ôà-5ù´ï)ì/*hkRTðnFe';Hñw¹feOáI™a£&æôWö É]퇗ÈŸ}NÒ¥Sw?]K؃pGæšb^¶*¥@j“”%8ÈYb1kôímÅ“/BÛ i9õ/âw„ñòPXmx’Ô{ÎùüË·ßû¼¢yy çR“Å©û±˜´æTlÀô c¸ áaÛ¶’ÊÖó’5ÝÄ×ùG¢¤ª„2§Âû¦Õó])ŠŸêöÓ7Éä›6«9— ¢¹÷Þò¨tÏ—²ïáÞ"àú0&M»j‰H Òù¥œf<ž$·Ÿ%T~_Û—ù:ŒÕ›Õœ®I¹'§t½é •Î‹ÇÉ¢Dì]ÛÜM‘#dÌM8AÇ™àµ%åï!šÞ%gæ´B£låà‡1¡gÀÉ|ûðÁ““GÃÉ:¤ž2Üæjû‹èRcqj¤‰’_X_Ê\ŠÑz ÷Cê¥iÍw’4T4ˆb(ïTëÇÑ0ïÉ™ƒ.ºuǯ;˜ÒÉ䕘!÷ú¢¶AøÈeôÅb£ÝA¡çDÌG†5Ó*|;!b>ãøyïQ7= 1…Ü¢z+´àÛc{Öó¸UB­Å—·xE‚E‚~ìH·GCÅéÇ\zH&-u#ÚéQ$˜Fš<§x»ÑaÉo`TõƒLL¼M†œ;™˜å³4È9FÄD©w=D;ñ¾Ï° Ãô¤€Y®c!S„¨î¨ÉW ãxô¹@Š{›ØVÇ)TóéViÔˆÐ@3,Í—&Õµ†øADÍÅz#økváfÒ‡^ÔTo¦Z3EL¸ ËM8¬œOß,,˜Ìç·Š÷„!<³EU2 ïzÆrÄ£êÖ#º„Zè©Çý5y*x}?„ÆQ¤No uÛ&÷ñ’¾n"­ìl±’h.îÍ%ʧµŽ~6Ý¡Çò†D½e4›NdÅdµm>­møâ,¨ÝæU9Ì6t“GJO+Mxþ‘©´ÿÛ§;¬ž`®À‘¢•ŠOæ•°˜RF©ûq†J o?(¶2‹°6¬2Ím-kK Að85Ó ñJM‹P õÍkÓµMã¢é3NfwJ÷÷X?cnu[r+Ç,ydpNX?‘ùσ/=‹uÏÆÇ@üÃ: çp>utÿš×„sûQæìªYÙeÀí´|p®˜Hͯ£1Бز,/lT4Ž …Q@Ì@}à˜{¢¡ÌÏ›W€Gc ·_¼Ž¤%²F·=«nJ5õ° ŽjöKÝŽàFßö¥KEVI@©(ÖuãúÂB>¶ém“VJ $ dÁ_“¯è’AmÙH$¾êÍ&bUºSX9LiîS œï@3©çÒ ÑOÜ#O—°s…Yuì¶yË1s¹ô«©ÜÌÐÿ,û¨¦6Ü´Oû ñ†3‘ ?ù“aIàZP®Òz 8Œ‰K†ÞX䋺´R)úF¿ßÓܸåVÌCÅò:Œ=ìT׆yeakü±ÂÀ!¶¾"ŠG‹#rƒxOe¿( AØ(÷>™ÎáÊDܽÔë¦!¹ÖSî[Æ9Í^ÊÐè§ Œ¥Ð“DhÙPë™U¶¾-õ·õ<éþUþÖ>L•Â}JŠ‘$2IþR¨H,¨`*:D\ü­ %÷ÄÇ‘ar#Å÷"D}æØô÷ü¡ó¬§ì<î´Gmœ¨Õª¼ôDÌp®\ÄÙÎc¦‚W1Ÿ ÕBXOéo›Ž“šE…ÌLËàò™Ø½' ‚j ¸”¼!?ÛQÌÓ¡©÷¾ê ¯)]ûÁ)’EiË–!ëHOþw¹{&8ß¼ísé„ÙRòjIÊ î¼­\#%¹êÉ_«B(hÎCvºZÙ¡eò¥N4ä'uÆ’ÇË •ÑÅàÞ}Ÿr ¾û¾Eö|Ü‚wTHž¨û(=ÝÏt{N— Gi¨÷j³Î P/ƒãzÂR±<ø"aoÂëI_É> |â.\A4˜ Ét¸0£NŽ‚–“·êdÉWð|=õ9è(åÁm,‡ó–€Ë»wá\.¸š2}ì5eº]0=ÿ%5NqN“Ê[Ⴣò@•sB!î»Òc_›cW)?1x„x'«Š¶]&^&»ÜÇÝÞܤÙMÃ=Ûøi¹Æß_«†É]&Vªí{¾dE÷’pcˆ"Ç£¥BûõÝwQqä±®áý>Ö€i/úƒõ 0$Ù´õùdi@šMz\jX„zÉE ‡ bL žæ;ëqtåo¯‚çrƒ:Àýé>„‚S\[ ÒÛz…çöù¥·÷³ßÐQ1o”<«V4œ$m’b1ðñK j{÷‡6#zV‚¬³Þàö×D󤥱ÙlgnAïg( «qìBŽ°IÉ(‚NET†V~g˜Ç¾’“ X•4ˆÅV6Pþ«Ôßz¢ØËðÕ*S…z‹Ln´4ÙÇnC-…©Ð7üõÕ}ÑÐLssàÕ&T¡È]Ó›ÐÙý~1d:F› ]Ž£@¼ ßÀ~­›= ƒ#úy?ì¥#›ÁŠz=1¢Ø<>í Ï9üjX„«óTgDø+åwú¿‡ô¡ƒÄ7¥\‡ü~Vêv•b{Ôy´Çê˜_ReLJK:ëª(b§ßpÇpØÜe'Þ8iðuå=`ï ÒqJb‹‹åZ¡Ö\±âòyÕ²:Išaž¯*äCø@ï*k?Üyo)œP,Ü‘¾¹ÅÖbcv0áîN•¦ÁvÖ‰Ÿ|ëAœÃè>ëÓZòë~ܪRbN×5ªX›‹¯ã]¦þlÍÞØV…³šu{!2”ü?-¡Ï¢o^´·a»­||ÏjwÅ;?ÑM1òsá/·N¶‰9cziÐ8"…¯ZsOc[hÚw1µàª)Q¿0ØE)9ÿäÍ„C3ŽŸüBL&†„’¾z'´Š®¾Zñê°ew8.Ÿnî‘6C€ñQõÁLJm »i{ïFc‘úã87á3ÌÀhgßÜJ‡%‚¶;Ôöv9Ê žUÄbŽ™;#úúé`ÕßÚÐ÷ Š$._YäÃkÄü•açtSz¿ûvO².=^ÇêËï„Ñ$®"TY%6ºUŒ¾«ê½ÓØÜ’-…¶¶pßð*(‘ÅcŸøÔæXì„üM’æ ÎûŽê…ŠbÔΞ'gÏÖ°±œŠ-•[Gc~ºE”ûnÁ²øf w>Ä¢¦ŒM © X+8aå’%åb†üì‘îps¼B1$ÜëþÀœ³0–‰öóµ—ܼØ8E¶Žz¥?&àX@ç†kµt•c¤Ú³¥g¡ÚèÛ®'ñèpÓÔ¥¸Ó¬Hipââ{/_¡·ñ&õfo§&ò ØÉN¡QEÇZÝ3U^ ZÔUÈO÷Iî ®ö‚q ]Êw$çgÍ¥a‘Ð_zŽIÞ’î)¹e¦C“€Û=%‡µ?½¦6ÐÖž—õj–ž®s}EìY½ÞøzÅõEÞËÙ„«?]ì>)w"ÂÇp ÜÚŸ-ÉÆ£A‹4†žú+´…í4B.0«á~ïñÉãúËSÓ9\ƒôZ!øp`Ä(ŽŽidîù—Z3R”Á¯'ç®%¦e±ì{‘äïoÛEkå®ßáÁì5Ž|lŒÑשñØÀæ·ô1ü¹1쇙¤.¾Í•Ùʧ‹w@éK_€@bÿñ©z>”ÁTÊ>æ¤Q) ñ]åqærtylÏÐæÁÉ‚»°8Ù½]’]¦Ú+Ý×µ¼ðæz¡lÔn;L2ÇÇ)—¬èì _ ÑK0€Êcp/àv8Òl–àb\…›Š0°CbÊÌs?/¡ÆQGwp'5ÊŒMâ¾öC2¹©ü'”ä‡îX’E:™Õà¥"·xwã‘Ñ¢·’,Çl—n Xyƒ¸y†BˆJ¶õ?®ŽãM ^bÎifi;2çUŠ/DâMD—ùÐIyøÞ[úâò|{¢<_T£òõ"“ÚÍ­Ï£\Ì0ÌÍ`xÂq¶êÉ´}Ðrœ}ŒjØYÞ­7{«|BJûÁ#cæysÜB[§úkt“¿Ç†òKÃk±=ŽíñZ•œ‘Ž÷ëVÏyïFø]oQ6M­Ôɤ·#³Ÿ;;¿”‡9¢I-sCUš!Ëž|h“ü°h V—³xEõý–6V¥§0Ÿ¶MŽ§7uõQL¬oà`æÁÅ-¿_–Y‘ó%M¦ä™øa¶sZGÂj¢¡â.Š¦YÙ¯üI«rŒMWcáp8Ø2&³„F鵪¼Ÿ%ž›`Û½.3 \—Sã•Y¼¿ç«ã€ŠeŒÀw…§Ç“ࣄfÃgg¯¶‹GìñQAâd—T¿d¡U  ˆz³œô&x/ZÕÅmîuý®^¿ZÛUÍL bsùNÊ4rò×X"PßfljªGŸq†Zœøù¸a§­ûFÒ‹žüì‹þ’š7j=1&?la*¡”ì¸yV‰Ç»©m{$Ê?vÅw·æ#bâ3³©ì®¨ŸÀþHz¿W9}ƒ 먷‹á§v-YKRy9OÒç 4)~’íCÉB¿š'/LªüTíïh‚øžtÒmeˆ[îí2 kRKà~/‚eç± Ž‘‡_ò¹F°•×Ym¡×j!‰^•ŒíD¡ÑðûN”@Q~©¾ÌËÓô/Šêý¦¶Ù5‘b£Ì_„ëÚ=êŽÏ)áºÅŽœbUcÂê“ѳ܇îÃß݃^÷ÆŠ —ižIûŒeìßóáF‚ãn@4·#ÚºØ ð«ôOÂ%Ÿ†üëÖ4•×Q%Tîd q:úâëÂä_Þ=”¸‹[wúr¾;Md‰îîhòŒÜò'>4—B™—s~Ô„è¬K1f„ÒÛB'«Xáê¯E2ç3½Z¥iΫŸ‹Ò•¤¢QÅ¿Í)a_åƒgG¼‡QjÜ‘ÌŠ—‘÷QðéùJøcøt¢õ5oíMÊïqéGI”¹´£« Ú|=c¤!®ûºJâ®ÕÅ"£úfÞZhŒ@-}zïÞöŽ­ÎW”&/A)*Ãí-Œ®Rœ)¤/[Äd—õÕ´h»¹ÁU.QÅÎì¢É½ÿëµ+Ýé`B§+ Øï˜I¼;`N†OÙ÷më¡û+ÔεçOR.?¤"}M™ðÛáê·ÛÌÄH„@&pZ_q.Fº*ÛÇ‘­½ý`v¥¼°Z8é<¹z$¥¢Íy€¥,FÝg^\Çïž³¤ªñíÃMÇ™ƒC?Q$7jîå>¨†™Vh‰ÿPNáЛßbü2t˜½ÛýyÄRþ ;ž¬mB˼ýѦÌùÃ"fQ“ s@ÎÁQqzcvŽÆ)ðgÉ*ì@…hÄ{ù¯-ÀÄõý˜+äÌ å§.!ƒêü£¤+Ÿ._2—Uù¦èüË6HP¹¯^v»£ÿ†ƒÿ™Dácòèפ˜ñ¶/bí†Dh`öÉ!žWn>¢0Æ…Bë‹<;Ø4‹ ·ßÇ gˆqHAÿܘê‡\žÖ†ÀñÅósÏ %Í&¯M·ÖP&¸€Pu’ûš `vhû¦,ºåRðW…êeú,‚‹dÈT©Þ$‰ %ï¢q©ÂR’p‘kÕð}JZùÅOb?µ’¾GûÕ©‹Q{7ÕÌÕ+mÇ'V ¹ŸFШ·#Ê“é €© ‘B°àc¶fYõÇ¡OÆŠÞöfÙ‰o— nœ]ÄÊé˵wÀ–Mi7…‚Ó%(ÈÓ-·t;ÛD Dcüƒ†@mN½½-¶‘™Ä1¤‹æ«7´×¤uzCš÷|æZä'8Pïc¿óŽª«êÌí$¸Aà5Ãø³4CÁ/öñÈ HgtôY]ŒåÞÕ½¶QÍŒ·Ó­àyë£öL½# ÷¥\È‚¼\ž},ŒÈˆýRá¡”ÂoÑîõ˜ªÍšâнQ´.â‹Âû9éˆ4Š*`‡‡¯\ºÜ°ÂÍ»½,…O–jŒ:0¬Ý–,„9Þ¿O›÷£Zrü±Rºî3O±.£¿Œ~C”È·±“Q³‘a©î5_c¸ŠÛï%AØúÄ„="Ø{iG?(‹È«>¶×LŽ[Æö5J?%s4ïZzÖK :*äþºëݽ¡#ÕWoOò[¢±ÅæCgTAr¸Ü&1mSÒªòõüH¬¦³ÛÑÆ™UD6®;¿,Æú‹æsh+ Ö˜WÒ}#ëg-Që q}PU!TØvšÙÈÜå>»ºš/ëÒ;É…´F3(ÓÍSÚïNÕ†ˆ¥©žLïb›ð&ß¿º<Ó0p{WEøÁEöáJÍj=§Tü¦GUF”ôv¥¼¨í–‰O‰³jI‚·„Š˜'ÌGz¨¥½¯÷­úX骡rô˜\f£ëS°z¨ÙQÎ\ÇâûÚ:H¿hr8êÍ;$+JØQTEÕ¦sÆÈÇ2Ktu0Èéð LQŸÞ<ΩÔsƒ±HÜÖЧ öñún'Ú£Q´…—0ˆô(*ºE³]öýâQk¾6ÔŽv¶ýšü;«Miï½³²cPÇM¿ªø;»z`Í̇×ßD‘Ÿ¤s•hUöÇKòBb*ÑÈßä YÊ”= ˆ k¹ÅWdrƒLQÄ ²ú„ÅhCÊ.YÒTyȤŸcͨõ‘³À¹0q¹u"×^¨dwfØX¤ôî} [¨ÊEŽÔvNýr5 Õ÷ŒQµRP"í”ÆU©(«bK0`éMQnEø!+ëhzÊr¬îDÕß‹íe0ÿ×xKïü³dêQåû°,!ì ^ßWÑ qDì$ô¿áz½ÖtJs˜Íaƒý†e“}›[=2‘ßÚ…¨»¶©µRÏ+¬ J¯7úÍËd¬\:ÐÎƪÖp· Ì$?( vxÀ®ÀàÄ^÷ãt‘Ú÷YgÅbøbpUŠT @¼M×iáÏ£Eºü{3‰í¢¦µ×Œ|f8ßçx­³µ(càñÏLië=M“eص7…IyàâNx ˆGáj¢Î ÚKÍè¥ìÅç­žÍú·ÖØ­¬n‡.ùv)Y`Ðhî3Ý]`ÜòÐê.iÊ_§}¾G"70zÞLªï¼Ož*7(Âòi(¿Ú×Öµ¹ôCFÓ Ë¡Sk™Ñáɨ+ˆk­!âZû$žAWÄõ”d?Øá~’ùY¨ Å4…g+ùÌ€ÅÈ"»ãÞ¹»ªx@6°•6ëð a–§O¤eHs®„ŸÓÃŇ~²¦6Ö&»ù3¿ô`¹ƒ’%Gù®0"».=NÅöiœLI;±tô#š±Â¢ÙÇ[ŽÜàóë+q{OyóV¡':vb-NHsoy¯At.W’ŽÓ5Æé 3«V,iA²¥í^½‰°$ŬdÇÅrzÜz²‹4ƒü·²®û4o—*SÜ6”ÞóH³ª­8ž-èñÓæ£nãľfÇkŽL »R,1p—V9¢ÓJ•öoPòÌýÁ…t€Ê÷=yŽs!€§êV~ir§¼‚ˆ(¶­2Œ-LÍ·–6Ý(ýXô*’ƒ6FØøƒ5†AÞѺ³¤‹¶{AÕ³ÐUì¶{NA‘ßèg­`mBaÒïUÞ3mXo¼.&>Û"v ³àÂÐãk Ņ䜒¹Õ—s‚æv$ó®;DéJ{D©–Ô™·òJJ»ÙBô®úJÅ :»J/*6‰ë»dš®à–U?ÇžàµY÷/7ç}ý™mf᧥Kx›pò)ly¦ýàËú-ñªOÅl䆈B¥ä-©¯6t3Œ½,3SimÛ‘¼H] *óýN¡û»¦&s×ÂäŒ}ýBŘùÉCÃøÓ‘pBò‚ZFïÞäJI½[ØŽÇ™½|"Þ‡Àù\Þr­µÍ|Æ.ON/Yõ#çǤlÚWJ#úFªÚôš=í.ˆ¤yLMÏwòzÿQìiÌ endstream endobj 257 0 obj << /Length1 1385 /Length2 6006 /Length3 0 /Length 6956 /Filter /FlateDecode >> stream xÚtT”ïö.)*%9€t ÝHw§¤ ÌCÌCƒ Ý©€ " ‚´¤¨4’Ò!Ý] !üÇ8çüçÞµî]³Ö7ßÞûÙûÝûÝÏó±0êðÈá6e8 ÁÃÏ ”(hi© €@A^ PŸ…ÅŠp†üuã³CÜ= p˜Äÿ(¸C@”O„@á´à0€º§3€_À/"Á/*€@ñáîE Ðâ¨Ãa|¸«¯;ÔÞ:æ_¯v[¿¸¸(÷ït€œ Äj ‚´@ˆ êD[3Àn … |ÿQ‚]Êp•àãóööæ¹xðÂÝíe8¸ÞP„@âq÷‚€¿hƒ\ &ãÅg:@=þø àvo;€r8Cm!0T†' q ¨it\!°?`Í?nÀß»ðóòÿ»Üßì_… °ßÉ [[¸‹+æ …Ùì Î€Ž²&/ÂÁ ÁÀ¿€ g8*ä‚:ƒlP€ßƒÊrzjÀ¿ãyغC]¼Pç_#òý*ƒºe%Xîâ!<ðõ§u‡Ø¢®Ý—ïÏf`po˜ÿ_à ÛýìéÊgƒºyBÔÿBP.üÿøì!€0PLTP €¸ >¶|¿ÊúºB~ù¹Qú»Â]v¨! P;êßßä Ü=!þÿ;ðO ŸŸ†Ú"6{( ÿ?ÕQnˆÝµ|w¨Àˆâ?øë÷ï7K½Àp˜³ïà¿÷˧n¦£«¦ÌõgâÇäåá>AQ€0EW!q€(ê%ðŸUtAп]ÿ“ª³ƒøºE]Ó¿:öúKö¿âàü³˜6ÅZ€ý?$· mQþÿoªÿNù¿1üW•ÿÉÿ»!eOgçßaößñÿ# r:ûþ Hë‰@ @ Ž’ì¿¡&?¢Õ‚€¡ž.ÿUC€PBƒÙ£ÈÌÃ/Ä úã‡z(C} `](ÂÖáeþø~IÍ ƒèÂ= ¿¾-¨, ð¿b(}Ù:¡¾(^þ «äh¼yÖ‡¤°3x† ¬p]J£O–S ²_Íj¼UžéäW¶rü¤-ï†|½0õÝ­–¡ÎW茇AyËžÁ&iòö }<Õÿœ–i’¤-×$Mž^¢Ø½ M—.–ô3Þ#°}— I[MÇtª7+Y‰²¨zøퟫJ^iRЊï­è#šlz ×…P±å¬¸$†®ÆYD§kg2iH}¼ÀOñ®$]êéë»ÛI×ûLÝ+dýèaâ®’d/­¢N¥ñ\ñuN±×næ=•û²»à®²EŽ4dö‰ÃÚÉ®%#§[çëªNܽBM•”§™¹ })²ƒÉUß“-&Ë#a d>üCŸÁ°GF¯„$ïnha<Õó%hû6·.txÞ6î™{V6ur ñÖx~|áNtXXBCˆŽÊ »KÿÝDœ—µ%‚•ù§OïƤ»E`ÇŒEBö¶7ßáºÏ–¯:™H¯š8vI§ä XÏîíÄÛ ?/ìF~)ëèx/÷†ãË’èÕUÀ¨«…‰¤B€ÎÒ; ßwÀœwN Þü¤þsöP“£‚GsU«±QQÐ mÁcÒ—Ó¾˜Q¢ ݾlS¹·âj˜Ai)?“¨ž`c2Ÿ 5=½ÂË?‰çe%(먂ìQet‰u=ÀÉÿÑäURûi]Æà 胉ѴXÍGMÜìøaôÓZzñ@ì‰9þ¶ªs­çÝ .Ü¢”ÏÉN9›”X欤Cn¬À]ÙϦ?”†ªT0NzTbu&“ß÷×È †¯:Iç…™Q*¿¤ÚÉ~–=_›ù´Oš5dˆ™ŸíBÆ«íV9Þ &Ž°¡«cp-€f ;S·È5ø}«)Úñ,î’¨O=h:üü¤$qðÌpŸæƒ|VBÜ ‰1ÅW¥}gêá©ÊÞÚÄqM!||KgÃ3É3Ü?»—n«H_ªò«B1¢þ»oºâÌ9?PTi2:ãÐWÍþÈZ¤‘#1ú)[®9>9óÕJ²·õJÐKn‹„X)Â:DÅ‹íNVåà½!¥/i}‹ö‡¹€#†xÛ:o\ †S`ôýn ±œÛÛ1?¬8š-ׂÏL J›T+[×ìxF6 6J/)‘«ëÃƱT:Ãh(¢(S‡)³ü¸áÕ  ¶Ýö3)¼!åøÉ‚ç{Çê|2aJ =JlüÈï.‘rÒ`WKû}E]dG˜R…¦X}«æ1»’àu|§h 9ÀÀÖÒTÑ]åšJSNvÓ6šgˆV[’ïñWž¸Ng‡¶¦K¡03×±éG–Œ·sé•sÄÞ°Âú¯‹ö ÏÞãbË [ 9_ƨ·2º¬ÝÞÍ×­ ε~¾¹jF¨ †~E´2Û>8¸I~Ë€ÅÍsÜ°Ð)Ùü“æaa’ÂlWÇeûAþ´§‰¥¯z0º}‹±ù2#czo5K6…¡¥]?Ÿ£8„tõ¦p”€†[üaJù;HѺJŠ<Œì3 {ïQùŒ2’Fà ç…àˆAÄtèmLxÅëÀŠË} nt»êË2¡Ÿ!YLd%b„•Ó˜Väv{hm8Ókùu&/^?ÓIÜÓ{ÝYT·ÓD;m0ïfQ0“Œ>Ú):nŒô¾ÎýFÊ;ï‡'Ž½oˆ¼AÑ‹éL4·t,‰=ÊýŠ—:å?&õÄy­s’Y–û±Æ³¾\Všøjå¥Ú‚)ªccfê´@ÕW#Ð/­g:À2}làÊ­MRصôÊÛ÷( ô\œô w¶CDàD.óÙ[æ^“}QäÙÍF­×ñõÛRó¼Çä{÷öЖҩ[ž›ã ëäÔa},ÕT4ä?;²ó£9EYâ‡]Å¢¼=hÌå _jSJγö:oÅ8èÖ«Jf 0¢ …ñÔʦÛéwºã·V#­šò(_u†öt/õˆÐà²Ë ©‘!ò¤±jIŒ#áÑœ Œ²¡_¢ÍU"RpMÑNWJLø,+põ¼³ ?9r\Φ•¸é\©ñhg(ž+ý1F©Ñܳ`“sm,?j®³ Ö—÷Ó£Ûk·7G·Y!K?¿ NL©m©-ï›tgl·aæýdXÞ{J¨šÅ>¦0|:8‚ÇÃÅÙ°®fû¡Xlk&±ûà‘?x IP¢s«#øÿƒR·NÕZÚï[æ6_)õnËÑ ]1Ôh¨Žeç®ùÃ5ƒš¬ì0\ùîäºÓd¤Ý"›QN 7˜wWËP!Í¢èÁ+jT Ôs×óULÚPmk‰q­Ë^_L™ghèâ½ð"üÈá*ß.;o#wמÀU0 ¾qMÝ]×›¨·ø=Yš°íJŵ·g^#B\ûs}§ChŸÑsOÅ»õׂô§.“£*—k¼iÐöñ¤P<‡uµw††Ó‹§Òu£žèÆÉ3d -ÕR<¥Öéy邃5aòÈWáÅÅ ‚qÙc-g/Ö)d`ÔM’ˎѪ3šjhW.ùê*7/ZxÄîYŽå `Ò“ÈÝ"žÌÉóïÚ7cT4?"h¿H¯5žQ?é«ÉÔÿL+ ¿©w·£ë€°÷;úp’îxíò•‰þˆç Ái]V{>+eNUN¿=–] ¥PÅJ´ôÓˆfCÎr­Z³BÏé‚`òAâzB›~sÝw›“ßiøaˆ2JYs'˜~ÁyÁŒ©x-¥›Ç5®aTƒ¥’¢›£{ÚÎ`0s±}dûîãXÙË$³ ÎÖ¢kp²–£ûÑWfY©éö•eö†ì5çÐûnPPrÜAƒæ4½9¯¡*xDXõØÍ(Õ5 4Pø¦…iŒËœI[’fKý”ìýœË€(-8ôÇw¥¡o.ÓÆ, Î2,.@* Ú"£øw/uUš…ŒúMMN–ޜڈ(}{ße¬{7ƶaA"SB½á¯¾Vƒ¾Ùy… ¾18•+J ÕäÖ`Ç«5(ÜlöÿÞ0G¾ãJ0ÙÞú¨—] ªEùÅujÐéáëéÀ!—ÍÔöêÒ %Lý6oVò'LDF”¦Ëf5mi ú½¸„£ÁÛÙRÄÁ퇱Ÿf¬ÑóXoö=îÖÕÉkùEzøæš÷úŽ¼ÁBkªÍΠu2âüÉrg¤F„ê2scšLjêALXH¿œóMSŠP` ®™¬›dkòºùD‘0}ë âU»–úXNÜÞÔdïýtc|MÌÂXAÄP÷ã1të UÛüÔ˜hš÷¯>b´(ÑÐÔÔoÄÔÓí†_XI%eÅÚ­šŽC‚)hÙrHÏñ¤æHñ &äÜðÍBstT‰–“œF[aË·£\Ó¤!1ôA!Tmf¼éýÝx)ù\Á¼o¦vMlî[IÞ1žðLŸ¦Êë€ ÄŒš0’Ðëlˆf “H3}ª”+÷ŒÌ²Ý%N⛨¢3¿öËøÅý·? âRNab9çE Š6ª?j4ƒÊÇ‹Þ~»Î¶ FwÓÇëT9Ï Þ«ã®<ç*•ºãχ1Ò T0§ìãG̬»UÊî—^À»œ(‚%³4ZŽ¥‚Ë0÷4‰F-Û|DÓ¬¬Fn™ô„ L Ù'¿X&¡÷lHsɯaN4Y~´’h´ï˜R¬ÛÃB˜•¼’žfCÿô@T€ljöYM¤Y†‹`°ç­Ë2’ãHuÝ <ÓY¾M¯VHÊü\B \fl2I*aˆ}8><²[öP†O*ÍÞ׿΋²ÞÚð£“á¦Ó¢­yëq‘À£~:6Å#Ð%Å2(#R.ÿAµ±6V3#^šNF>k\Ùšîˆ]±í)»š¡±:mDD±ë´|qiêñ·3Úë‹›¸‰‹Žš—qÆʉ<ÈëîT3i&úŠ7mIà}·žù7U)˜¹TšÇòº0h£\Ni¡f½~ª’µ7ÂVû††ï³XÞLç6ö·ë€gÞ[Qt¢ß‡|¡Ç ö”Ìßÿ5qJjð•Éhù£ÈËXV¦ÏžY“NkŠòºs^íoÚM7ÉzÖ7¾ÀÜf©µ@ëßc÷SçÀe¾ò¥¬Ï Üo®¸ã«–Êƾ•uÐWJ·Z™uäÚôxp‡oÔ+÷–›<Ù"c#èÍ‹–Œ´@²ª¹ÊR^m^âO=ÓØáå‹ìÓ…íõ|›oÏø®’N¡Hu<Ì—.Gmî‚v"k¸‰2\MuAéÑØœnÃjIqst³4®¢É¦V§Æ§8ÉI&o§2Ùë¥~n¸­'ÊpþO–~êŠÈ{À–•:ã<Æ•sÎXSäMÉC š™ùŒdŽÚðœôÀ—=–õ™­›Ikû‘y;"gÂæšÝä¨?Rïe¸èÓ&ñºÝ1pe—"1R «=Ù<Ý@ƒG+Qj"áM'”ãßk´uþ܆æì Í€ÐSžÓ3Ài;Þª„æ'ú³HÅ”bùyÁCГﲢ“Ö ÷¿àú¸$1aÓ„è³>fË穲„hÕoŒ‚ê¶*ñß1 ÷è©hä^[wDx1–åÍÕÕî °.U3¥Ãv^&úéËÙ¬«,N|ÉTx¿mHŽì0ÿˆä\ï;Ù}¼ÄhE=ía‚kgj&f‹ä¼cw¾8ðïxò‰T«R„°Ýí~ÄœõbLm:ò –sÁT_Θd"ER)×ýÛ+v«c¥´Ý[Ìï°š¥âŽg¯Ï«Âõ9xÔTù°ÝÝn0x £À3Ó{LEq̸¯+n¤÷ÁaÓrôÈúŸRæ‹f½Â<ËôÀ¯ýnKIÛÃMeÜ•_I80A’·ÎÎA¨õÕ u;i¨j塵›/\£æ¶ èé å¦#iNÞ'pJñMæ†A±jÛHâ{–ÓÚ‘ÇmWN|vA+˜¬Üv6~‡7c)5b õ]æغý!¼ÙånóʶŒÁ‘:KùÁ Š–±'!ô¯Õî_(ÒÍ´nÄ{¿s×ûìԒɈ‰õ:Š­ïèÄéPð̸(‰Ùò ›y@/ó2dsÞ”ãC25ÍAT@e­~(òžã'•¼øæ6è ‚}4Aｑ²)>õÎxÃIuYœySÛ%æê£DÙ¬‡ÉÔŸ†a?JÖÈY¥|r¢ïȸ}N&jÇTQN n¸¦ßšú‘Z}¡¾.Ë·À¼ðiºÒ7é 09PÕÕ¶ØJ½‘2@tË}ëDÓ±4|Õµ?vÂärdxÏÕ~!‚*‡À4·‰döÃâH^o:êàç\7Ž4FÒn½UÚýoq,^…™ûàí'O¿ÊWTð=}VN(›`*ÝAÔ×!´¿}À6#ДˆE47òb’Üg –NÜÂXª©ßªC\9¾&]Ÿ-•¢d… S@ûÖáòõq.èË;˜FZ¶C^Óh?g†§s‚óävÏ÷/å|µó,6Ó[uÙ•µ¸ÇÍ B5Ôcš6»÷Ê÷¢žF&ø¥ž9”âY=ɽëx›Ø±Âä Ñâìžö¥ Ábž•i[x‹A|pÂN¨˜«wÀ÷¨ÍÛÁ·3î¶ù)*‡ý©ÒótêPk+9èxžÌf’ž>óbäClíÙìS_ªk‚n4= ó¹¬â×^RL_¼¯@ÊÕD¬7¹ÍÅÍ'÷ïÇS°'<äA¹}{Ì(Ó±V¼“-¼Jn òŠ¾0y|Å ÎëùŠ#{`Õ.ÚÁe©sn•UMMD·ùº n×N½kµë#n×OÄG«ùªŠSÞj :ՊŹ܉ð#¶’n’¢áSÖd^ÓâØÇ…IšVQ'Þx`@yrIĤöyzC­cæÝí}AÞ¼µ/ƒ"ÓŠ2½Glõ´‡)ÝE <ãåÍk¥ù§'¯(øEˆm_$ú™¤.ÐbUU¸ R5ö G“›úX²‰«Ëgkø±èC\}ÂðTsÝÞúI±ãTÃVœß¦C‚ibr§[+¤P~Û\½sœSw̬€ÛÇÓ cL½[öMÝ}‚nI.qùtjüY{Îülé•uðÏ€ä/^]ôUý<Š¥z-m>IÀ…å÷\†o<\æOÞÚuÚ°N,õ¸Ú ¥»êi{‘”§>¿bäæK×a˜ˆÃʧ?•ˆ©­p÷ÑÙ°/«âãà)8+¶ >è$DFU(§X¶¼ç"–ÐK´ ²gÎÓ²lªÙ=Äêµ®f¦o1“nH8—ÞwËØÎùë|‹“W« wc åZû"aA¼ÜÎ7¸v©èî<ÏEá†x!/‚›ÇbR²iúg_²ós¶ Ęè=?Ð÷§–U‘Èáë}“ ·+&™¥Œ®œ‰¨”ì—o 8ÿ@¦±ÐøœPT{‡P^qçƒnŠ`Ò: ¿ØòŽ%r˜±úsšÈãÎb'"-ô _î¾U´¬‡ZÇ x«ämºé¡²ÌéfiÃ'0š9çGâ¨û/ðqsƒÃ#KпÎ{ROo±%Rã˜nYÆà*&µT!à°·"b±¤÷ÁÒüÂx8Q*­¸›ðÂO2ÌCxÒd Rí ô¬_´Ìd÷Bq÷ªøÎç±öH±óÞÌ!Æ×#ñe¾«m±ŸåŒEÛð"&fk÷¿\¸â–!>Gb>?¤®||»ç*5f=È(tìñªnÅUèç:Gfœ¶aTRð î†^œ8©ç™âs“ïš ÷Š÷­©Z6ÙÖǽ£¥ÃŽMp4¶ÉÄ«>ëYŒ™»¿ûf:]é0µë6Fk!â²Å/ËæP®üIÀå®}tšú}á§éÒ“Ð}Þã$'ꇅ/»^5îlç…²ÎÛБ¿¾s[Gqäx7¼>5ÊŸÒ­{=5‚Ü÷Ûï¡€±> ¹‡\‡ß!âÁ§7»_2©mÄGNß»¹"(1÷)/Á N£ø¨ _ˆ€³)O;\Y\eRZ‚ÎÈ¢­TmAœî.ïÕÎü´¹Ãï"î}Ñþ=5ÛÕ%«çé¤fIÇæÞLbDë•f ÉJÀÖ"yðd­b0ÚåÖé¹~¾3ºMIØS Óºs®ù897’ø*éÆ)tašVñì,J‚ªÁן¡]|ûÇöbÏDÉmsÕHƒ=-Çbê_ªâãeÊíõuÈN«n íW¾]24éêÊÙY¡MÌhô]Hß/„aíGt­<¸ÈÐÈ?q JVs ß[¹g7|úi¬÷g¾Ëál,ZŒS0éȽæ;þn@' ÿ4êÚ mÜ~Ü"1AÊü„ÝÞ6‘+õƒs'MêöOñda»_f1ùмÙf›^JJˆ}É8×IBY›w–2=-Å8Ý“U³ŸÍ A®%5&ƒhD3Þ,M0Òˆ;–®áýL¢‰z endstream endobj 259 0 obj << /Length1 1546 /Length2 8420 /Length3 0 /Length 9447 /Filter /FlateDecode >> stream xÚ´T”[6L "!))Hw§tw‡ä 1ÔÐ ÝÒÝÝ) ÒÒJHHwׇzÞsÎûþÿZß·f­göu×¾ãº7%™²£ˆ©­1HÒedebáˆ)(ÈpXXØ™XXØP()ÕÁPkÐ_bJMƒ#ØÂ÷/1ú$BŸìl!Y'k+;€•‹•›…ÀÆÂÂûC[>€8Ðl P`ÈÚB@Ž(”b¶vn`s èÓ5ÿ9hLh¬¼¼Ü ¿Ý"6 ° PB-@6O7š­j¶&`Ôí¿BÐX@¡v|ÌÌ...L@G&[sAZ€ jP9‚œA¦€_6 ?•1¡PÔ-ÀŽäj¶fP ð$°›€ ŽONSàér€šŒ<@Éùc,ÿÇ€ðWo¬L¬‡ûËûW 0ä·3ÐÄÄÖÆqCÌf`k@IRž ê e!¦¿ ÖŽ¶Oþ@g Øhüdð;s @RD|*ð¯òMÀvPG&G°õ¯™…yê²ÄTÌÖÆ:¢üÊOì2yj»óŸÉZAl] 30ÄÔìW¦NvÌ°½HFü/“'Ê?2sÀÉÂÃÍÎÃÙ@®&̿«»Ù~+Y‰Ÿ*ðò°³µ˜=ò›žþP<Î ÔÁ äåñoÅ#VV€)Ø 0™ƒ!(ÿDƒÌþà§á;€]º,OÜc°üúý}Ò¢—©-ÄÚíóßóeVSÑV–•¢ÿSñß:QQ[W€#+€‘“ÀÊÊÍ à~:xýwe ø¯,Xþq•˜ÙXYþdûÔ¦ÿdìühþZZÀS´}b-@óÉõX8YLž>¬ÿÏTÿíòÿÇð_Qþo$ÿß„$¬­«i~ëÿ?j  ØÚí/ƒ'Ò:AŸ@Áöi ÿkªú³´ S°“Íÿje À§E˜?‘™‘•ƒ‰…ãì( v™*ƒ¡&(óG®ñkÕ¬Á²­#ø×ÛòäÅÂò?º§ý2±zz?ŸxùGt|Z6èï1þ §uúï<$ &¶¦¿öŽ“ tpº¡<þ qƒÑ“­• úš¯Ð¡F™$üåççB¦æ½j”žÔ5Ýs±.T8E’ܲ­ŠíŸ«ŠÒäñˆy5÷ÖT¡MÆêœgøÔ™k6Qþ?>è…&*¦Ë ´_£Ä¸T¾\éXÂì!é ÷•uöÛ8zÅ”à ç@ªÔ\ÊA¿Ë,tÞMÁò0–•¦Õ¤-ŠÞHíÕÚEð®²ÙŒØtÊåïò±Ü-¹KùPGÂ…a—f¤2Ò·qP[}ê»${Fú8„ïû¬5ª‚ýH8ãõ^Þ‰t+ q¥uO):@cø­Ú „•ï,åù,©»[$YÙìP%ÊÆHI?lþ&×Ò!Õ¾6™iœnÉûðnóØoU“×ùa‰æ»ïÔM_ ÕyÑᇊÞYM P ú|6 a ÉQP˜kÎãyfÓÄÌË«Œ˜®Ú@úÂúØ,·šñÌTßýyÙýr¹„F/Ñ}á—Wï;‘¦ïÒÝX^â”_DË }oÔLcõÑQø' rJ´Žô§ŸÅõ!íÛÅ‚Ÿ§çŠò2yk¿_Ž'ªPÇQvå|x1^\J]ŽµoŸ¡Žû|Zþ\ô#e`RøXË÷°Ÿ¡/O^ GÙ€7ý#ìueÐèu /8س·øg§67©²jõTö®ÀrWhÌÛ„tIãcôÕ2ŽLäTÅ~™[€¦.ÞЪ*Ç+2KáôÕ8žhËVÑpU*Û²ÆoÓ²½dÿúeñFGwlnÚw“œç¨çŠfmŠ9'­4y~¹Ã¼È›ðw´ÆZîäŽwïNy+¸£©.U>áRÁ=lŽ;_Ý1ƲÁyܳ/øïk Âd¢î6Š±å EóßïxÔ£8^Å ó,a?GàãXÓ¼­êvy³hŸ®œéhÒе½\˜I)QTjN¬äyàÎæðDef}J´ÛäÆsB1.BtN¼Üíý.OÛÐÎùBŸFöÃsúÔþø¾ò¶ï]!˜;Ì4ÏA~eæ…sàǽC4[‹4”d>-瑺Á%òDz ZˤìI•Ö툻âi—ú”l2PSh8-Òœ¤2—tR„Ô;íG^g8Sû‡-n1ÝbŽ´ -}òãXÖ°4Å»0™{fJƃ âe˜[AôÚ¦Tªá‹O­Íiº£±·ðò…Wðu:£ƒŸÉ:0ä™ÚÑ%VA†9b&§2ä œvÖAoÐ[×;ƒ)´îºñ‚wïÏf{+r4é° Ó ßI[èG_߇(t›ŽSU¹˜Ùú¯×¿«5tMþ¤”S?u%e«»J žåq7ÊÀèj ÂV[yÕäAc9Î$§ªóá¼ã—Â)è ¾©ëG\A"DuìëL\H<è“ënÇ×­Ìä€ì—Z¤DmlŠ]/–mðkþ:l;#v’jŸ+Í Qá.s¦iÐ +‹89ùÙš½Þ¿ÂïÂDõÇÃ5˜2‚ùŠ¨Yçjȸ䇰œh§×»ÿy8qdÆ5ðÕªcJ¢:Bkua®°™uÌuï r I>†×îϵ Ū2],[ÍJ:o>„Ç?âPt‰Äo]å†r0ºÆ)Éy2vÉJ'RŽ&s1Y^JƒJͨæ]ª C§=ƒ8Å´ˆhs>$úåûÞœ„ŒÎr%;{áp°‰†¢ÃõMŽœÚ—Q¿ÿ!P^î:Êa4$bÛ=–m]Ù¯:§øl6ýÁÏúné{z¼NJÆ‚Ø Ä– ÍôIôðâl-:SH14íkùê&!¾¸e ë |jÛ®ÃþbÑ[Zuy2d‚ ³X¡¹9¾òM¤¬Zrq´¢EÎê]%ZŸé| \âÅ€¶C(óì(RÒ#+ü6Ôq^IαJ"‘Ï+~߃;¼¹ ›W¦ªíÐ7„³WKƒ¡/è§# #CKßwœ¢ÿ0à@kax•4F@‘!«o;ÏV·—i{CW̓Î}ïpE:ÜD؀Dž¨™'Üè¥pgÃ"°K²£mP‚˽ eIØ••”2DÙmqÓûv´Œíé9Ñ]¿Á¹iàÍ; Šà,+1Áj %3UÀín““›…Ï&«S¹ÖËyÉxÁ»‹—ÊOX Û¯‚‚áÒä"OiY)°j?i%dHšå'>õS~;T 52Í!Âò3”]æ]²6ü×+¡=©%ÓWÎð 4Ÿ‰§‚}jnê "°Ãx¯A~ {´!+QåX]ëÓ';ô¥6Ž›|%›p,)¯àÛú‰™šbÂ<E š!x°œÚÃ=ŽÒÀe¾hW–m„q¹¼ÉÀÞ=*9ö°aÃÐ}UñBäzë+?’®†·û=6¤ŠójåH|ZˆÝÓ|a|¶71¬ÏxÂf’²nÁÔ¢çìm\Жìð)£•†;+:!5ã·p1x9kù?1‰ò[¬hyr°H§ú‡&Q·ê\ Å £ø&7uÞ‡Rš`¨E/0.^-l¤a³ÖúsžV\ñŽ¨VM ¥í}u¶wzã)Ò0ÛäÞܯ¸)^…ÉKd€St{1*Óä'••²µÄGî®[ñ7J¨œMÇOï?†ùñå“œµ-ºsÁ‡qvÄm~(F-™h8…³BËœUÅ8âÛo‰ý ”µ|Û8–>'†ió…È4uDZyú“‰æå½Êc?~0l Ø‚¡Çá%4­uqY@‡v{4¡¶òÂQ0k,¶ó¢è½ qøú.Ë»½9Šãäð»jsöË׊‰¯©}q:øÔÓÖzb‰ü3¿h@ï½g$É>¶Ã, =ÎÑe#íó¢gT~Az{¾½'¿È+¹ùº~ÍLó[äe^;Gßà @Û»#ºŠ¶vâ˜Éœaxx¾š.ÔfÕ‹&F=úYû¸NÏG ƒÉAAúÏù²¯%McñÇ)ÛkÃJIF¶øÞü4—P_‡©vzT‰±Óôų£âÖåµ_ÄT±•9¸¿«×L@ˆR`ÓöD]:i`HËïV}î=«Íæ™×$£J®‘ã0T5ï>œ"U<‡½ñì)fŒ³›s•Ú¿?„/ňHGò¢FÚ 7ïpŸÃ³®)Š ¾:ÚaÈÚ «¥h,Hiù$&Yžœ c“hðbOøvÁœãLsô¡ ‘™>ñܬ,fÖä”ØK©û*'üæí¯*x‘“nbQ­ZŽóx?к«Ö¨Bk|eÔ‰0%G}d÷ÛAç?#¨k@o4³Æ­(_Õ<ïYÂÛOp¶¦¸ æ:N}^ËsŽ&½Ü.â7Çš«_ž Ádé /¸¬o„0»˜é¡ì^ÁÚlÅœøÞÔtÁ`n¾§®M÷‚7:e@èÜ>™¦Ù\g-Y$}ßµí>­j)âråTä¡€*UÙg"~A/ôË{óö{¼?ÎG5¥´ %Üœáåáõ-Dën˜$¬û’½¥#H Ýþ´a’r¾>^²LWÂ$CîË ¬íq?ÄÙßÀY0ø²àv¦ÖËgx²K ®+‰‰?Ù)Sq¨ë—%ÌYéÛ‹¾'¥Ê÷Ç ©¸ÏˆÓkPV×Òó¢næDïÛy è_ËtO Ò4}ÃýŒHìh•“÷Óö©Ž—,àjNÇ5àKâšD}\ýËrÝØŒ‹FYAªòPQ¿ãçøj&)³ag¯‡¼¼9Å}7lÄ5Z-¦·å¼Yh­úŽæi•/)™­1=¡»ˆI&m·¯¼cWÕQ ->/[TÃGX$w´.}$ù<(V_m 8[‚¨¬¨YÚ“ê÷P£«œuBl­x©†æØ/jq*8Ùæ´ûyîF½·õ¹¶Xé$^Á%Àçk[ ~\ì7}fÄè¨õ£î£Ñ6‚Þd8ußÃöð~ÙšdÝ×/ÉfÕ<¢3fó=K…4´0µ•Éä*+”Æ_„ñ{íéc ”¸D&ÖHKz¼V”ù¤h-ý¤éd½€äFª Ó†™õPôÁÞf)˜3â´ä]…eRqºU†«uJÆÒ‡ LÕ™ï3Þ†ã9Êî½*#VËöϦh5¡<€vƒ-¼K´=k¹¨Ì¸r?(7…þúnN –«b¼ˆ¦¸²%4X¥^¤* S}žgx’2†ŒV"Š/1Ó¸ahR¨ù2•ÍY«ÏZò{ª)u㺛çæLîö<;ÑIµŽ¤m–Q—¾#JZh«^q:ËmJÜ€5¿“Ý4n]7ÕUóÛÇcļÆ@Þ„Fß)Qrk¹Çž&î[¾¯î:Ô0×±o ØØߊŠTë$N¨[>ÒÅ0ûºVjÙÌ×–‹”?d#ïä­|šNÂ÷í¨œÿ‡÷³d€Q´Š'šdU',1¾¨¥u*¦§¸ïì@‰S]»ÜÑ!ÁùÛ;Xz&öW÷-Ís˜L»—¯ŒÒ6ãÄÜcÞ ]{F6}U¶ÈÞïñf[£ùýmQ2xã+ÅhÐaÿØ'ª·„6ÑÏÇÕk.·Ž>|y;ÚUTš‚%–Žg³f‰v»¹"¹f^sÜè¡3tm]šqfvY¶ë ò8?Û5oñjIMâ`¯'•ì¬éf§™F`09%’!+ÊÌ} <^ ª3í’ öJPU¡S¥{3;mRv×¹0†Ù%<‡Êݓ곪üe„JÞ)0›/Ûä¿ÕÇd „rm´Xy6-õlœWKhßnf啯Ø3vÔ®ÆÈf 4Ì“Æ= |š\Ô€ÑO‰+þÝóæµ½µK~fÐ#?Þ‰#åèD&"ž5¹Š-­/CÝÅg²²²üƒ\˜´Å瘨÷Ò)‘ÃM§*|¹ƒ$8¨?YfÄ¿³®­Éñõ"Ñ¿½£9› ٠о|‰VZWÍþcŽQÌIn]PÕœÍÿ¦vWèJÚG¾¶êÕcqTÀò$¸„ÏOK ]ÕóxD¥Ò} ìÓF«I/ÀIŒÛ$dõcX–í:)QºQ­+;hÈo-c)Àœà;!FÇ?áCoÏåm?³„Õdÿ‘ýIÇæ˜Â‰¢)LÞh©A ¶P¤Dð ćB2*bõ¨8ÄñjÿS±â(Þ~³•‰’ýÕò0?¢B*Z™…ᘲ¬ù(64o‹t98ì¼íß½vÚ`Q6dþ8_¸+´@îÛ´ø©-_øC rìëÜffXB(¹1b䣗©&ø‚h•’Ž’³'—icË;¼hX3ÆÖì"Є‰?¯sâ 1ì1ðš9¼ÒŽr`ËÛº(ëqñ99¾ÂˆÅ5WÃú]Q(Wï÷|AK³3UÌÂsBõÎù­ýÞÂô:ìidGtÒ^¸BË(šÎÍâÒØWŸØ(·Yzý˜>ÙÖz(ú!Ð^€ŽœÌ±?ýÄ‹Š¥^{ucL@+L&<@Ç ¸É›gHDÁ`ä¼@ªôpwø¿eˆ@C5`ëã2ÊŠ(°j.²tÐ_>¸ÈÓb¨É[†7¢Û¡8;¾T>²+7f”Ž5üàëz2\×þ€ÃãrÄ@‘ªw"U12Çʉ¼±ô–‚}«Â¯›–m^ªq»MÂ*îŸ{9*`¸èNbÒฟò@Kžþ¥ šŸwã;Ö“ý¡D9€ÉÒ§nó’ÕãàN&€yiGA‹a"V^TBБOÀ4°«Uä•=Ûu¬vcàƒ¡ £kUÖa{^²–ž\1à3vá¶f>§.XzvõæJ·<K;ÖCœuŠ{$$Wá|¨öxÈ™mž´üž–ü°L>ÈAé!¶ëwjÞ[¢ŒßžáÇRÏV 7€c’VÔLš‹ >éé ‰_bQL–§g¶nÂú¡2¼Wnß®3‹Rs˜êröeÔ@k×¥çî¨DE\?µû½0Ï{š8ò‡w“÷鸜)rä°“œâê§(àÄ$'B² kIË#Jom|3W±­ñÚ|/NÓ(§¬}â[ì‡ç¾ìA²_Æ¢í´7vOL;z Æ÷ìß `47ZèöF&%qÇ{"ßÀà5FRêäüœ÷YÍßuC=Uºë÷ÑĆ;ñ3ž,eÝŸ‡×çõʽ F<ö~Äv7]­.‚]ž›ß:6tú:â.$Ÿ€¤"„úÎ?䳚ÂöZðàꚧ—¸—Ü*Ù\L~: ejî–¼h<6+êõ¸5EÚN/^¬¨‚e›¶ïgïöÉò…ª‹]ˆA³·‘ú®c‘ný'ìÑ…Ì bäÅÊ寉 Ë5&ñŸÃXN…h/Ø”TÆ€™g–&jI¦Eu׶ºúûäG»^1õ¾ÝùvéöñpêÝXÐv;¼’v÷Ÿ=‹ë3˜-…‡ðº§¾ Ø¢úÆ‚í¼1á¬:NÓ¶ 2âÙáV=qSµ i®ã0>âÆ¢)QQÿ¬íN,ÆKU¹³ê‡øUÓd/ Ÿ—(d=2žÒˆËS'e)05.æDlDm[?ö¬2rî?sÂUà ðö)Ÿó%|ÌÅŠì’¾ÇÿÐimÙj»zsÜÌDGT¾¶áî6}`˜pÍ<âƒpUSL6~âç2V;Wø.~‘ðÝ6Hö ®Ûæ’/­†“´‚7rβ¥¿’tÚ[7Àö,á‚„‚æª= †•údCŸvö]œ.[[0Z&©H!xŠÂ¾ŸŽŽÐ÷pxe5\åGYk)“µÐÁG®wÅç›ÑS‚hâBßÄ%ìæžKØg˜0 Gä ᱄ùäêÁ$úÑa£ºm7ÜãºÀºæ¼Žp#Yó¥ÔÅ<òž¸í@£&®V¸À<¼ˆOcuŽK& ËD:´ŠkfS Îújsµ*x;N¶yQo¦¬Ï6ÎEoÖ? ÁÙ8„PY_b3H>(±S¨ZU¾3[?YâD4êÖ%ÙsÌ…sý: ¾:«ßëÇáwj©q¦zÇwº96×ÿ˜ãî;KâFÅÅJÒûÈ÷3V‡·!‘9ÌöÓm2¢i„:¡†FùRœŸ=ïcdÑé,L!ñþ6ò–º£ Mä縄dN!øŠÈ+âžä†"Ü¡înCúµÎw-~èzÀÛ‰ ÊIâ»5”ܬÖNÝwn°ôÃ*Š”| ÛñæåÙ;£HçÊÈ°'ÉyS*.°<êÕÏ1¿}^»‹ÙØô¹aóL½ZÚ]â9©ªŒ}3l\ ©s%¼z2‡Jjcr™Ýü9Q¥¬4€š/¾&'žO¡¬ïGot^uﮃª™qý}æÏz¥-Cž™g5p[éJªmGóΟ`ã^6¦EÓ´ú32fÙ–arÓ±dFÃ.>Hžúr¥œÞ3£É“HY4 ?×„Ç ×Ií²&G¸{|»ZzÜ`ëHI¸«Z@µÓå°û1#är¨üXX|‡†Ç%éÚe+Õ¤åTÙÑël"½}­Q¤\9Úw„ÒésY$ó¾—6mÓá>¹§ X*%€¤¯Ãn ,’gêoàDûÁ‘Úáy³8 òG”k¢T.î8zšd{Î08ˆƒùµA ’ÀpS ÓÍ\aÖn’L–+˜ÚV¯å)t·—š„w,‹Òú>"¸S“ËT¼·8Í>e¬k*ÔòBü¾6üµÁ J9ôYádpú ª,¥¾z–Åoa÷5g;ïT¢6ܼ>Òq»DeNCÅÅ-jÚ¿¢-PÈ%¹l¯c‘X܆:¸¹üp¢äçÑÿVò Q_`ývz>Qì3];Œ3ÑÜ¿(³^@"çZç/ï>' ¥‹«Õƈ ËErÅFÊFø¿%þ܈l×·68Û_K!Kšä®5ô¢>¸L * /Õ®Xùäg~‡Kr¡=Ö–Ë^)I‡¥ˆ‚‰ßÆËÊT÷eød\² Çdlóü ꡻ɔ?ü4,Åk¸Y‰!”bAT¨ÑŽQù•íÛ)Ð-RƦÀwæ—…´ÖLƒzy€WL^}ÈÜÏÖW±ûK$øô°ˆOiiŠ}rh[ŒD××É Ž~¢ÅRR%Å †­×)ÙÒøll$’€DƒóC¶À$Éþ ¼Ÿ ô×"² ÇHë‘ïFXŠPùhsVÂ5 ^Š‘®2¥ Uk¬˜…hÛDEöÝ ÷<ú³¤RÔvx¥É%,³ßÍføm©¾á›èû6¿ØÚ¦m1cññ­g]‚0JØB'>ò·B.ù2_ pˆÚx•XŽÞiÿ¾n)³öÐæ ºäREÉTŸÈ}à2»²§ûz¿ý6£aÕ³~µ/kƒËÆþøõöp¡Qhs¹€€#åCmLíè§þt±‰pמ¦îêüžl¯+@}YñrÊ™ò>Rb„‡È*6¬7*O=²1Pȹȡ4¡mGÀoè5”¬)¦É²à‘tîBPâåX—·75o“ż×]†µ<'韓¡‘’½tåÉŽº”3ÛN•%W3ƒ8_!¤r'|Vo&ZSãt^É•¢¡û/WU|èinVÆ`<Ë´Ws‹FàbÔ§~.³ôs©¡X´Þë© õ=@ îžÿ9±@˜/£Uò§xÎ|ã<Óí)³^uÎS4 \¾™²ð´[ï0'S§k µrÓXˆŸY}…ßœ5ÇÞ1Û"§óÕ Þ.Ѩ֗djÈÝñµ&½“ ¦ér+˜|€gPgçåmD e•x rññ¹©-©ØeÆlFeÆfo £ðQ=ÜÍÍ)Rë§øø—a^V ‘ V±m{Mª“{(õTKOT ‡Âx1©ã Ì6 KGäNÐF hµYžM$BOж;¢e¢©ñ‡ ?™3*ÙÐŽx­ ÕødAÖ¡Mu§ôÛ¾SÈ_ûr-Öóýoóá£#,g‘¡tÕ s>êÞº…ÚŸÖé †f\ãIž/Þ:jC‰¨Zº02 “Ùƒ¥ š¬¾‹æ£[oŸ'¸L¶fÃJ»¿Ë¬ë¾rUÆÁ|g`JÛ[¦²0G z9wœ2 Ç«žHTOV–ãY)“Þ¦¯wõcGÖJëQH²~a¤tjð>ßÔ"ã[ƒôJ1¦”þqÖ¼Z·g \Jv¸‘LÅ]âÇóC/¦hÀyuãËFíŸ|7NIÔÀûa!°~´4z&ù-7Fƒ ªùû‚Œ™aÅ›ëT‘"ƒ”ÛØ7EyY•¹÷YÞ>§’ð›Ï&0ÙÂÙë K­9ì’ÐB¹–‹‰bâÈZ²Ûu—³5r9:媷'“7Vn×Hq‚Õ!2³ÏH(jnw‚Œü¿FµY†Àª¨Ã¦Œ6ä6²¢îO¹þlR>‹ñ& ^O„Zº[¼H¨g:Q¯°½'ÏÜ5fxgýuåj<äþnÛ¼ijt¨2ì›#ÅóbuåÏ@Õ­¦÷ "¹ŠA^ìäS…uõ_œÅ¾Õušj<§`]X·'ápoÿâ7ñšoN#Î(Ì&I=N ³ßéäÌ/³e£Áqõ’á1抃`ô®]+)|k݆ÜoG'gº!Y4ïp F—G“&|Z÷±ÒÔEË1„ü1AÖ"ã¹(Tí3“N=zX7ëy§ë»x¾Lwm»u8Ä…uU®àç 1ÑJ7æj:†¯sÆIíhè'åQ°H‘åYjÚÝÀÁ5bØpGaÁ­*Z¡þ6.f›9oÜ;44§ó’Üt‚”> stream xÚtTÔíÖ/ˆ„ RCƒäÐÝÝ’ 0: 5t7‚H HJK7’"]’Òtw燾ï9ïwνkÝ»f­ÿ<û·ãÙñÛ•º«„¥9XÖg²q¤T4.6Nt::m þF§Ó;:Aì`‚ÿË@Ê ‚?bÒ ø£Š  è ¹@^A Ÿ €“ƒCà_†vŽ‚i Ä ÂP´ƒÐé¤ììÝ!Ö6ðÇkþu0Z0€|,ܶ`GˆPÁmÀ¶7Z€ -; îþ!…màp{AvvWWW6­›£µ( À·h‚ÀŽ.`KÀ [ð_•±¡Ó´m NáZvVpW#ð@!`˜Ó£‡3Ììx¼ ¥  P³Ãþ2VþË€ðwo@6à¿Ãýíý;öÇdaagk‚¹C`Ö+ P“Ufƒ»ÁY ˜åoCÔÉîÑä‚@Aæ2d%4 Çÿ.ÏÉÂbwbs‚@—Èþ;Ìc—e`–Rv¶¶`Ü ýw~ÒG°ÅcÛÝÙÿšì[˜+ÌóoÁ ³´ú]„¥³=» âà VþÛäBÿ³Ã<|\°ìfaÃþ;¼¶»=øø~¬ÀÛÓÞÎ`õXØb~üC÷t¹€pGg°·çÿVü§„,!p€9ØCÿ'ú# ¶úK~¾#Ä `ÈñÈ= €ã÷ïß'ãGzYÚÁ îÿ˜ÿ™/»„ªš²’ó_ÿ[')içðdåæ°ròp€¿IÆ÷xðþÏ0ê Èßiü/_˜•@à¯lÛô¯Œ]þ&ãßËÁøÏXªv¬ÿ!¹‡ÅãøÿMõ?.ÿ7†ÿŽòÿ"ù'$ë …þQ3þÑÿj-êþ·Á#iá  b÷¸°ÿ6Õÿµ´*`Kˆ³íkà ÇE€YCÿÝFˆ“,Ä l©[ØüÅ–¿pß[…ÀÀêvNßÏ €õq4ÿ¥{\-‹·O‡Ó#%ÿ¨À›óŸWÊÀ,ì,¯'/äèrGò£Äð>î¢%Øí‰ìl0;ø£ à±ÿÈ9Ø?¾I0(Ø þ üý‹Oÿ†“²‡:;ýþ£1ÎŽÖð?Ô}ìÚ¿ä?ì¶@Ÿž°³ zSÔtY.Aêʺ6(2J·¦÷‰‰ÕsÚñ›ó5jܲ3>!?NwÁ^Ї%;iíJ-Ëó1ÊõH+ŒÖ«ìÁßNCxpc0cÄÌç«p£À -éA¸qËö ºöj*u¡×æp!#h±dË­¦¦¼tF‹~F ëÜjÜ.T†)¦ YåUe·]ë/9É‹ŽU¯T>p÷G¹xò.œ0ºlNbÊšô Ê뎪ø,JV­æXXéÙÚŽú@ŽÚ â>—8ß…ŸÐ,úšKƒµÉj†éA½™áµ@ú–,%À= A2;ädú§ó3óô—mRèüÝò¹ãŸ¹ÑÝÙŽ 3Üê—çá@ ç’´çöCl†ÜÙZSœæ®™-ð‚œ}UðKb2éþÊyŒNê$* ŸÕ"!f>{£W7\ƒe9t“y™Ò¹Í8\²Ï¦sõ r5¿·ü)$žCkñFöfsüÎl'ÎÌû+ÓÔ×'¢B‡,¾½Õ#tève‹nH=Ÿœp&zHpõ€²nCûc›‰‹ë{Û˜È7]÷³šœ‘4jcâÄ/ö.Ç­˜iä!1IçSùlŽ’ïh `uR%RÀëx¡¥8ÍE>?Âð[Õб¥rÄŒQ¡ì–í±8¨¿ì5# ÆÔ{Õ“ð(‹¨#“µ{rÿþpT)°e[þ=Ce(âk ñAl…é%8]öêCBj‹a>_}óm†Í ,íyÉ–ÃØ,¨í«vûƒ¾£ïØb;̤¨œIˆÆ˜ Ž^¹‘ñc8½¤F•ÜSI¶fqöÐämŠùBlÛ‡pÏL¹,Ûm”£“‰Xßéy†ë²| ¼8×°žÅw üDë‹çh+ñ°ïßÖur {½ˆô*©Q¤à¢ÏéìÐ.ß\Ü3Æ}JJŠ2“ÈMÏ÷õ×ÝŠu"Æ’¹}°µ“ú'KZÖ£ó²Á±zeaûïOqë°¡xϽê}Rªù7¯Ùz;¾2[ v{¸ç»Rþ2­”ô(í ýrÊ)æ2Ö„†òdû¤él2´]”MÅO2žc‹ž0Jlí:M…ŠŽvÉ·&ûÌ@Ü]ËØç(ø󓤒+¨äÄVqÜ3Ê„4*Ž«ÍÚpÙÏç¯÷ dA†aV—ênÜ~òÁ^Û¨z·ÕB³rø/ ð¨¹$Oy¦On×/ÉQ¢\“ö,xT/ íE9ù%ïNjœØ‘öÿZ3ž¸h°`ÁE K—ùâäœ:ëü)5iŒyñéîmÝK™ 5$á.½[oÚÎõò#”;4%MÈ5%AÐï;nU¦“i‚•<*_Êž.®ïÏâ„£â}h,Á®!õÍ¿a“¼Ô9Iå½`öqGš^̯§puBC¿žE~ñËv?\!§¤é™oÏ8Û´Ù‘øŒ È1á…éé.lο ³Yív.¡ŠÌªŒ KˆHC6] 41 žµÆ­ñI'슊4£(ÞËW’ ]8äw|m£,‹©Cý¸ ý®aò²(Ëw”©%äkÚUMî1{¤ª Ãs“ú3ÁUøSÅŠ7âÛønEU2Œr}Bƒ ³O>t|«ÍP˜þB(]8Ú6µêì‡*Kãô;</Ñb&;~æú‹­F {¥´˜[ȲÕí>àÞ'Zœ\EŠ *3}AñPÆ£ÑZ4Šp¯Æ»­½*¢Þ8Ü‹©}vj.ŸP²ôº~´ŒÛ9RÇû{gÇFûi> ¾`„êG‚Aüî6ÂÕ¢Ú˜ÁZþèg%ºM‹\™²º]Š›Éé~cR?-à€æd´Ç«ßP#D٦Ѐu-ÀGËñÞÄPQ7+¢xIfü´`„È0"XœÇ/v~q„ñ%úejŒ(FZ{`m¯´ü›Ð¬+NŽns èëlm¢3Ãee™CoKg Bô e4Â3ÉåÇàh&«‘UY³ÚîZ±¬ q-yúáä¦Øý§`?žšl™ÑŠnE!Wç ¨D‚ÊݽiEðh”?Ý‚*Ûí9!À|Hå~âÝ/v+o®Å„”yÝ+Ôr 8% V*N#¨c”úz:wÞ·ç5µu5‘&~c“toº©ä›"ÔäÐz>Pš^}Ú ­<öYV¬sÉĪô7ž£A§F-Äže²óçIÒ"¾#oÜûQÞz$WBœö,T\íÇŠŸ$ák²µÓC¬Va<ŽBqD¨ýB3k69 3ð|3kÆå}âB^jÖÀ Í¸ó]ª Z«¾Í<UoëÛõü˜…"Söé@dдi‚á?e³£öÉ,­v?eÅþòúy ¿³û­ãqÂüI¼3º•·…C;3O«»ÕݼñxÚ5éeË.–K b3<ƒi"(œÑ¬Ü1ËXj Vóß"—Õm¾wF!-vÞ\ê`• ûuÔᘜ"ÿÚˆ.JgÏeªsØcÅ„xé‡d¤;ŸªörÝW.âRý%±&[|ÅØKBÖQ®  rÈʤ¼ø\´¿ÕPàÝÛ:M¶”³;¿n[6\†­•6éÆþi¨û-%1_aÖsƒ³ ©ª’NÛ­“¯erÎ<¦r)¨—5ÓÀ¼l8ûû1Œû¥ÈþåCô%îwÀX˜é´¨·= óºU¡µ€ -ùu~'¢±øÊ•Ù”ÉÏz9¥s!ž"”×îƧ« Žù\V‹8çb¿¶ùÕöão#»?q _l‘KŸ FEÎXR jæ"Íâ$à8ÏwX £Yð™–@…9½ 6ËC†c_œд6ˆ”bI‚p:´• @ú´«®Ö½q_Õ —Ÿ¸E:tf»?U`Lxhzó‚GùK¡B‹E·— E mŒ‡ô2žÎP[´k£A$ó·»âׂƒC9©ÀïÇ5Pßip i7¢™ó¥ø“—^O‹w) &æÂ6«“»LfxI÷ÙÓÝGnÌ@‡™¸y=˜á†eƒá/Ì~6÷ˤg·¹¾lEÞQ>ÏôX‡+ƒ‚ˆp½ßô•oôŠßd‡¯ywúºa¯è}EFMZƒ…:}øÄ{…å‰U{]àq߇!Z·Á¿GiϪ§q[4w¾|ò@VÊäq²Ië°ƒ2Ø®2õ“ÎÞ‰Ïó‹ÜØØa>Ñ]¦®œ“$yXtú“†ü >ê-OéMavNøJg÷é•ÓZžßÍ[ýÓX«ì‚Ԥ凡»f¼é<}™Ý/ /zضQµ ·S¿RœÕâÃj™1 (C#ÛD>x3‹INÿsZbSñ9sÑzØ÷úF4Ü\‘³ÕÌ5Z^—J­—Ú8ïhf‰K•¸¿ìoˆ¥ 1œï^´_ùÊYêÊ)O°0o:ËÝç,™°×8T·?í=G´y߉]|uõš¢Ø#Hº.þœëŒ›œòySeJ\ô “°žÙªJS˜èÃdVïƒÏ/ðÒž®´â¸WbCƒ>¶÷È)Çð¢.퉞M +â»'Ÿ²c?ª{ !‡S6)Ò ÆhgÔbÉ^\m¼¼»0‹C"‰„!”ô¬Åk[ mÔ€)ï<Æû Ñy7ü‰!Mƒ«Ç\Ȧî%× ¥V¿`ásœ½Yº:¸b *»h_¦ÚÃ~*¹£î‚‰›qy=J“(ÓòÁçÖc?hÚ]ó ÃÌaY¡,ªôù x‚p£ UV…cù±ÓaÊ1mÁ=ñ)Ù¢ Õ T-ýÖyŽ áb¦F›%½ ÿoÜVõ²Òµ6Q³lSUÚO˜L´GX)‚a®Þw,a¯Uëd†¥OË.0 0"5‘Á5mcç‡\M?пNÁ÷îï ~ñ˜EÉú˜¦¸µF"é¤{E0èÇÖt‡ɉcëL¼àUÀ0¢L Ú˜;½ìÕ5½yÒ+:ž'ðJÛp´ÆˆÚsŠí€KgÆ£ë¨×yU6Õ ©:'øL}1¿I&:Œ;àV'7õÚ•©ôÐ/üØà’åšÜú²9‚0T‡nŒÜÔ&';LÖzñXt%$9¾¨ê/áÐ'ë…×k€×ÇN^ÁÞÙÊãÕZxṂ~V…ã¶M_Ú$¸¢&µ08;DHb& «,k§ânSæª)JÈŽ -vgˆýn¯°½P~À›ßGºR$6ïà J[FJïIú]ÚX”(½ŸWÅë.DG§¢ºMѪ6œ[ÖiÕ${03+fÁ|N·TñlOµqNR‹G ä2°µ 1²0º}¢#µÊzS{«i~Õļ@¯ e–±ò-à½+Jñå&Aøû¸•`|U*Ç’v\áÜz"ÇD! K)R3ö0otÖ7D^£Ðcìþ»ðT%î—¯ü•XI&çõrµÖÔ#J% å;;殉ª†öËpN Ë«óµ•Qì¬êõ²Vʼ 0ehnM¼ÆüäÕ/EÛ…È0xHûÎÿ5¸z/â‡ë*T½Uâƒ8 ¬$åãdK»÷. ’ÿòÏÙºz´£wQ7ña¼E¦\öHÍH±eÓr{&é+~qî×£ÄIrbÆÆ1”¨‰M\"ß]r2ÍhÞIbMò;~ ]µu«KäTËxv ù¡Å.®]Î.µÒ ·T·²¾î~à_kIâ•«8¿*‰(-b-Ýb´¡8gÈ*Ó*Ð;Ý™És*Á¿Ø@³ª²7|+kÝ‹~¼;T Gü¡YS]3L¡âs„£¼65V…4xkx r‹th‚LÎ1g}ìã3ÇqH€h¤°|*+­bvßrÃ;¿S‰Fü&ð»Jt¶_ƒjŽÞM›ý—¦¬;ñÉ‘«1÷«Öç‘t}Í«âЋ5 oÎC^µ°FäÛƒT^H ×t,,§ÎÒuè N‹×êJäÑ:“>+È6n¬0¿p Ú¦DœYŒä÷#~õ¦·à>e×êÁ—‰LGÌŒQ^åŽú6ÿøÌÓŽ~s!Žr—±ŠÀ³ýlÆ*‰©ÂAϵyãÞO&¹žá,„Ë›çÑŽãÈ|¯UDŒ¡3×~›ªOÓÇ·»ŸSx~ÛL¤ |C„`¸ƒngiØÅÎP‹’íš!ß,¡þ#¾êýÅ-žÏGûƒ¸ô{¿[“8tÔ†¯âæúö%9oµ ÷íÙ ^¡ž#bs:à!ÆÛŠ1ˆ­aÊœex¾s}=V†Ó©`ƒ)0 ˜µ,¸`‹ûd;¨×@¦œÝ„#›aQLÍ‘´?ÆÅ°Ä^zhï&ÔQD€9Ñ )7™¥U„å–ÖàÿCeÁ[1h>´·†9?tSwÖb`ëÝ·)£¶ 'ÒP£êFyušõºï.Ž> ¶\‚RÊUÆmYPÔÍÔl³ ÓÑL?Ÿ¡íÑ‚úRƒPw<—dg‘$õl‡–¶lÈï_Gkc™ŽÇ²k° &¥¢ÖfûÚGs§ :v^T»»± ìׇ†"ÊÆ7Ü¿º¦³©‘Q½ùµàwLÍ[DœN¦‰,Ëç¦*!$Çl!1ÁWL½ (¼€µFgØ‹5*vå5v ôÏ*àú¥ðŽ6:s™Ã¸½2™‚“ž‡^[ûn¦tCbGëƒ`÷J—°ü¥‰-å)ª!å¿÷§!Ÿ {51Òu®åCÆI¢è–î*Ï*g“xîí$œÑ…¹h4Å‹ÈàQ¬ ‘Ÿ9$Ý=Ha䳞ckëÈXm_7Ô|´N —î˺î¼óŠÊ'+šJ;¡íO’QúXp·¤îõ{SMGqë‰ið_bîu#8«a'ø|½³y]@ØŽhJ+±ŒÕæÒ©Û8.1ûiÔåÙ&××%é¡E4”Š"‡®_¬Ü˜Z{OM{£b¢¨ƒÙt‚&]ú¾EÄ)¬-.²`ÛεyHÜmÛöúŠøzj“ÖW+?øÄÆBù¢“ Óá[›Q­9žV®½OQQæ;êåU¦ó_™ óÇ;{tt2Ž3j2—º¸`ñEt;yyÏøöç[' 0æ´M–þÔ·ÿÿîùXÁ7‹ÃdÆ9ÒÜwwìâF”æ„®²Š[J+“‡¶Ò§h¡î+ñºîžÙ-³- 9dváAº@ <ýh+©ñ¨@s *.ÚKE¬@ê©ðö÷cÕõíT‹gnöK$<¢<ýëó|„C ±N·"6F.£1Rçm9cpétÈ–ùº;¢àg ¾š£&þ—̃þ²<9òýÐ ÈS|#’o_ºS9£jÛg¶N(ßul—+KZl¯Â՛Å­T‘Ï­}z§²7'tc›>‹ž¡9®€¯nÇ;»Lü4^ G¾˜t×bƒ;¤„!MÒê ùÌí¿—ùY¤ wð¨Æ§œljo?!5±)´'™í¤ŸÇŽî´Ÿ²È$¶Š¢ùˆ@ÌG¡Å[·GTi:&Y—ÔFµÏm›¾®Ý2–Ú“s»Ô~À ˜ïaûuD•hµ7Êq|:´œóxå.+»–únÑ-‚yvÚËoíãVÞä$•ßÜÆBi𜇵P…·.ç)•»Ü“„<\a]ª"‹ #ÓMq"&ñpU++¹ÑtÞ S1b’u<Ý´˜> ;–³€å]þÚQíaêáp挮ÄÕûeÃÓZŵä7(zÞu_ðŸ]ø0:9åú«kEâò8e\tAÌ1|ò2 èq%Œ¾‰d¿.zeŽP¾½©ˆÓÕW ³; +HÂZµÎT1Úî¢r'£œ}½CÚk€úUýY0â,—ßÆ/îú\Ž¸ñ|£^Î6K¥J¸XYª¸Î9Û‹ ‘å67¶ùB³êf×î›ïòµZ ©F|¢~è%©‰F•ÎÌ1Qˆ°í£?|¸ÏNV° ”hú0¹ët£ÆËãÐÏÂ\£ ofBQ°UA~I2âÚOÑOö)U²^Í"…ÃÑ‚n—;Ðc«ò¼Y¾ ÍtGœ!y•¬bìcáZ+ƒ»pïÛKuÝïR:ŒO£™6.ðGe’jBáúÏ£rS”}ÂùF:š½¨~~`B™ˆºhwÝ• Ö$JÔ8B™|ñÝx¹'Cu ¿yÝ,“Ý‹‰$“ˆÚs±;(å0Ç-¨ÐÂ$]5[ íjU=?ªåH¤SW_jÐÔçpHb§faSáMÇDï1®}ñÝH9ýÒ儘®Ã-_ìAL›hïg‘ɳ&Y™=Œ‹¡Süt¾ãè_@ž÷ó1Ž§” ÓnÍæ¬_a–p³B!·J6âØA_!OÕ¦‘»l*»³R%õtÆîÞn^SÞUÒ5²Ûv[SÛŠíÌf2—t:l;QGiÊHëPÁj—´fÀMþ‡M.RR¾nƒ¦Ð\«1u>Ã#,µxV:›~Éø¼¦r:jRyOy…»AKAƒÁtqGùó]l—xj2"ø]+“T (uôÊ$Ç×Ô‘°žÆñŠ&PÔ5ÄÂç)qƒtª‹¼Á„ÒVKlO_QoírvºF¡Ì„ó”N¹ãLÁ{µ¡DæúOƒÎ`€?±ò´^ЯɫÜøm^Õá·9¼^ÏæD©,42[߆Ñrîw ±ó.ó²‚ßj¶êéaû#½’ b©‰X⡼å®ÈIx&(Ý©Åä"‰Ó?‰ÉZ¨D3ª—TÅ”ÃÍy£”(Á4¢¦-‰ŽOJ±¹ì… ä2e9Ò‘F唘84A/‰¹»uÅú™g•±ïZ¸lšÜpOPªð/ýèS©2©š·89MÒÚ œ—äÍà€2ÑgÀ¯¿œâ7'Du†­T!”ø|LѵizÁþë]Ð5oúÇÌçÍešYß°Ó®e$DTtR,NƒØ»>˜/Ž·4©±àMÍ f•ÒÓ›1I<â=¢IH‘ÿv<ÈÉaÉyéeàªà’¶0óæÇÂÐU¢W*£ò÷É3~YèÇ#FÚé L9’†Y$Ô‹BoZÉÕ¢ë}Ì8ÜM²zü(ô¾K¨äÏ?ärâÎTÉ¡i Xi6ýôìþ‹J"ð endstream endobj 263 0 obj << /Length1 1391 /Length2 6186 /Length3 0 /Length 7136 /Filter /FlateDecode >> stream xÚvTlß7%1:¤‘I(½¥»;¥Æ0`Œ‘$”$$¥Q¤™€H(!!-¡¤"HóN½Ÿû~ïçûÎù¾³s¶ëßqý~×™àeS q57´+\ŠC$ÀŠ@ #s9 ,%K-Xø-@ÐŽñG QŠÿË®C±&Kp3B£€ú>@ˆ"«‘Sƒ’`°ÂÑE &4á4’ê£Qp€ Ú7ƒððĪüç‚ ! rb¿ÃjH8ƒ¢€FP¬'I¨ƒú-Ð0ò¯B×=±X_E(((HŠô—@c<”…Å€A¬'ÐîÇÂÝ€¿ÆC‘ð߃I–žÿ?j ´;6Š  Žò' Üà ¡6ÐBÏhâ Gýq6üã ük5@ˆäïtEÿJ„@ý†Â`h¤/‚@yÝ>p ‰¶¡6+„¢Ü~9B}üÑ„xh áu%8ün ÔV3B óý5? ƒðÅúKø#|~Mú•†°d-”›‰„£°þ€_ýi"0paë! ß×êB¡pÎî”›û¯Ü|AV(„_\Oó/‚ ðÎŽÊ€Á`90î„Ã9°å†Fù„üãþûrAFj¶º¶¢¿þÛ¤®ŽâÄ%å€â’2` ") ”#ÂÿÅŠø« ð?±z(w4PáO³„-ý§áÀ¿n_è/bÿËM@,(ôÀÀ2`á òÿ óß!ÿ7tÿÊòÿø÷£àãóÛ*ôËüX¡H„OÈ_v^°ì¡ @ý·« ü]ànˆä[õ°PÔP>/ᯆ»™"°0Ï?Pù£·úE0 nŠöGüzP€â0ø¿lVÁ¼ †?¿Mpiþ]R C»ýb—¤Œ,ŠÁ@C„+&H2@„@C7xðoA(4–$ŒtGc¿îSÁý['¨k¤€ ÂCõ·L¨ ¡Ëÿ* À`Äû B_ÿ‘³†Ã“ãhص[^Ïo½>¨Uã _T\¶ÉÇMbZŽè(2„ŸæFÏaöÕ2ú:>-i í©NñžâÖ_½ ˆkJ3k>;qN5¿Ü ˜f}3ôx]­ÏCÅ-n©ºvêfåMúŠ¸M_°À/@žÎ´ˆù ¨['¸_9ý.v|Ùl婬õIåñ»VIQe£‚…®ÆØùȱâ<”"L;Áô£{û#LùCç¼ú©¢€ð»R%8ûyÉäñЙjKIÿv{vÒ=¦wï¯àÔ×2õÙ>âÊKf‹GJ¤>µ9êDŒ/RÉ]A¤Ê?Sé%½¼â±ðs£º±%¦_3•ùxåT¬êf/‰ÌÊê x/¿÷&Ãnœ§’©á9_ÛÉc9ƒ..Ê&Dêó_æTRÚ»šÎ»ƒ¹3È«<Å«óƒ#³ær,ˆ–׌V›ÎèEÞ„¼Ä%Rˆ¸#”ÇwQÑ‘0ˆ”aV¦ÁŸP¹¥‰é;UÕßæâ ÛrŒv*„®²±q~á¢ú^/ïíY1Bâe´ÖòÉoƒ„jÇûŠXÚŠ³Jäƒwg}ñ°K|i(•ýÒÁ1›ø4ἴ̱;¼Ì휃<¤¥Tw½»2W3\6ïÚ?y™ÞúqÏ‚‰–:Sd@%£žü&,Œâsú² ÞÝšº•z:Æù¶æ¹kBÞîÐZkF8ºÕ×$-¬]˜zÈØøvÿa¤)õÖr9ëDJ»/Â;µÑÅk{lç³K{N}‡V••0šrëk½Íµ—º±ûŽéöÆÓ>ˤ ò¦ÒPgí/ Ohû4ÝXb'¤’Í(÷³Ô9ì-ÜC¤iÖòʃ;çK›æÞ^?ËO‚Lp}X,´]ðÎcÑËy¥,Èu²>r0quzÙòÁ¹pEÄE=ÁÁs¿ùû¦züuâFLiCkÖÁÞƒtÁø ÕÙ×£Ó¹Ìösú²ž‘ÛO§”<Õ¦jçß‘+2´ñ§ ]Ç›fYÀðV‡SÞô…›Ä…xpZßÚ…=˜éSõPž³ÈHiî Á ’øuFçáU\äc»ªÛ fDˆ-¦5C³ÆÄ>¥\«Ob(å“Òzš Sy HÒº½‘M¸±•óÑ×øžÏBs;›Èí&÷&Š LÜæÉ’`v—Ù„—–†È âýÏfÌɆ÷ßrGÓ†Lñ Ÿñ”<+ÉÝ9™ uúÒ8¨V÷€Öê?Ó æ­ äÞØ´}©¬FÒ·kó"ÿ1Eïјrð”FgeÛºÌ{²˜ôÛ—ƒ Gœ¾¤Ñ²Lt—L¦ ‘þx/-&^ ïüöz¤¹˜Rdž°´‘¥’¬Yç÷ëÈìd|t·Š›²N¼° œ2Xks (°÷)û.kJ‘ˆúõBMv´Ý¨('j,qtZ‰WŠ¿ªTõ²GOÿ˜» àîø.NƒÏß][  ¡ÂI)iLyÕ¬“HïÚ¿•=Ý3ˆ ŠjR9#þÁúxm‹?óEA,™B%çµ€dj„š†ÐQËŇrªGpioåNôH%¦W×G e;ɯ[;øò†Zß ùYó˜O«Dž^‡á pÿ¹ûew«z¼W ÝêëÄ.‹l—'zÎó‡Ç¢%ŽyP¼íUS ¯ô) 5¸(@ÁN€Ãõ¸²áãàpFö6}þÇunÝî'ô]¢J•Œ!ÛçQØåþw¹ŽiR_ΨX…£ U“¨voÐéÌ8÷Ê(<ØŽð-ƒþ 8.8ˆà{ÛœbvÃS1:;潇gàŠ,3œã§Ý¹9wRKe¯´ÛŠsnc–ž f(®¢ç•¾B©ƒWÁ”8„Éæ”Ó=è5†u–²§~êÊVÅŒª§ÀV§ó ¤nwr²æ/¿¨Vã8 —ºŠ†Õ.QL{…ãü¤lœÔ‰*®>š:ef·+!êõDÇmºÍ×îY,¨0MuÓœ’ldÙãyk2âï!RQ/–ÆÞ*M¶4Nôd ¯§áêýŸ*­!¢£¢3Ú}œt<{y'ÈU×ç…¸³RüiƒÉšd/¦Mr?#s=2Š‹ºÊ$ò•²á*ie)^vYåóò:ªM­à.7¯ö¯XóH<.Gk?~CbÜ<i,ÍÑä=·"ÄBÄœóÊG ltÈPIÚ–?½²à E”ÈQ‡œ¯‘¼˜£|°ç{º9€Ý ™¡âœCoÒ+ˆqegP笧/-öÛ;*74TIdnaßë1f³k´l÷S'8FÝuÇ­Z­K«šõ’³†:CûÚw÷øÄqü›#›l+e\âQGûŠ«Ü®&YÓ7¾ 3né>yM1ªfK2¥µÞœÞ}^öÓÊ#ÿºw½_‘!GWzÕÊø[¯ê‘&FOºq½¤«4ZU§‘± ¿,ÁJ_ z$ˆ‹ÖÁýËr€>ìRV AþòÎCÏô»ÁŽH’tšÍU>ÌÖAæ˶¤kòÛ2[Ë7yEu4"<×å3f pÞ\³ÏFrd©aU-†Â6ö‡ò‘Ó¢Jµ•ÖÝ̺ä6¦ÃóRßÃŽI |‘À|Í%­Vnr« [9zoXÛæmgH”b[žj¿m ÇŒ5„¼²‹((~Áî>Uß2<&x±7æ ”wﶎѼYĨR²iÿ¶æÑ熌–¾=vo|õ—öi8âa³f·æˆ¼Ò¾,•Ï)Ž«ò Öúdl­Þ½IÌsæ<ø·Ilüês·íËS—š#vô£FF 4H­m‡ðå%P®?÷âíYìéux0ŽKH¿’±Z™á8ÜBorPÊ9¤¿ïKi}¼{|v¢çhïV½š ®¯G·®¸Yë ¨Ö„mß¹ÔË–D-òµœ³5'$h‡õ”‘Æ Ÿ~èò÷¢1:pÍÚºÔ©_ Y–„v3­Zf¡vøþõ'†ð$åÑÉP“ ¿²²²›~µ1ÅI\|yÀÛg½¾ˆ¾O¼eÊþ+³›{ÅäÀµvªå‡M½å2Gk /qz-É$ÊÌïB.¬F?Gk¼˜ÙˆáU“TjW?TÙ*~ÝÁgÏX èðpî‹h8þé'¡D¡ß eYÊ^C¨Ø&º¾“Ž™ùŽý Ïk?#e':i8‰X¿’fèf?þC2Öf7‡Qb]º,9Fb«ÙIp@¤d˜§Žµ_s2lÛ^¥nó–Ü­dÈñZÍë¦K~/ÇÄyêPýV惌‚ò¯÷•y¾±hr櫘:Ÿoc‡Ñ…»‰)†‰Wa•‰{/yˆöf©>qÏÅÜA²³X+³tˆ?JšàKM¦¹vspñ†9„tyªèN†Ë@½ª*ºauàɽóZbo=>£-Ú$Å<“ŸõYØÜ–žŸ§±ô°H¡Ä?Tú›mßs~×Eú RC\Œ(V;Òø Ñ‚åÕœ7If‰£©ØWû܇’\-§P[›U1À[mý\a•ÎøKÌ;ìs)gÞCBO´Ï‘Ô’¼HÞ ·‘Ô!Ÿ„QÄ) õ‰êËYßügå_S‰.ÍÄÒff ²'¿hʽêcÕ¬Ùå_~º7ÝŽÖ:62€}+P•ïîh9DuOÖ‹NðÈe¾u¦n6ûIcÞ@;ÄŽÐD«ßÏU +(@ Jº^´îÊiµáý±]Ïb¾°óÐiã•ã 9IþØ'mq;Õ€`9–å—Z¬Í2µÅ&yÃÛVoUÐOühêlèÏÞ^ü¶Ì›ï—HRÊÆf:Gô{}ï¢'ñR}gÏÀ‚¹XàS¹;ôZ®DÑø•&{%Î=宜Ã,Š;øñ\–ƒbþ‹ÍžLæjuʒݳä1”&c*>ô|~³Bôï5¿šÄÆíxï‘J;‡óN÷äñWЕ¸·Å«(¢š5–«'Õ¯/*Uy¬˜ÐA@GЖ˛—ãFÓ„\ôûŸ|äPˆë1/î/PxQ”~óAeÖýt9r?b¬Œf21PÐœ¨ÿÌ+Hsˆ¢°Ókàd#gxN,Tÿ¡Í[wMƒ¿É¶¡Jbí!w¦¼]õåýÒefš–[±.àË®ß.4dR5/Z»¡-_9Wi Ä–Åß3Úk¬G¹~¥›g¼—Â.\Æp/^›™§X¹^Ä€$LJs‘Ÿ_ O¸sý¦ý‹—ºêŽr YK¨µÛ§+mtJ ² ûë°ˆKü.OÈÐF.zóŸïëh$rÀðózwÍ;…܉_Ð@èU3¾(‹RØPpiسܬL÷göý0ôs6xgÿ\ׄMÓü£ç-Ãc’„¤¸iÆÅŠ €¶Ø×}Õé—){º|ăé˜ëmæVß’´[fdÆËA dËï9zÏ\•Ñï“YËaws g€Œ©_ÃKæNw®‹ùRVðÒ<3Y¯û [ —ëiSK±7—í Åøµ¾žÆôŸ6Ê«DÒ«_ÿÂxŽT©Øîºõp±³ñt¤Ín+‹­]nÌÉ ×óÍÈ¥uÒVl²oMZúŠfÚØ1¶üìÓ19†ÅN‘¯ëxÿH[ï¹Cï.þÏ.)™+¶¼¢o¾­p…ê¤y—¿Þ:˜Ö~IÿøÖ—Åqô#öœP“A;ËL ÇÄhèèÍ×c ‡S}iœ9´W㢗D~xáqÅ-—²¸ÍkÕ:™þŸÞ ’vºó é):ëú“Ý{£iY&õÊ¢Q´Ü¢ÚÓ»Ô‚Ÿ6‹Î?ÚLÑÓM}™µ±¿@“ã6c(:3Of)MYµ| É« œQÞƬÅìçdn&&]Ó¾¤3^öíêi+âs|E±ÛÅ VÛ¸!T˜Þë=Çòâ`?Ð*;„2î¤ÉQj,àê›ùÁ÷åu+×=0_j…®öúÎaÝÕžOÄvl÷žö‹•õw³¾î´,½*ßÜç›ñ8èŸQ"ŠRèV"W;%S ½:®áj¢P(=K@]ðÆWA˜gýßß]M‹]ž¸ÄèËÒ»ì°ÙÎ;ìÙJüêÇ@–†…tæ"š–Š*À~Ú¨•,Pã¹sÑÈ2=Ì*n#…»už¹PÑ „ïÚ—NòÇwš/ˆgòć@Ú‡kggܯ÷ K–¹•ó¦Ì¶uh“'MfmìÜØÜiu™ÌK¶$]‹Áenp±Gð=XGŒþœ„Í®°ëz{,d›Võ/hÒ¢]qDWüDœ“×ê@w<§fX·ØWì=£(®QÔ5ð¿PÓ‘UWiàÅ÷€˜ÎáN—Y®”’žüMp®ØÊ¢›Tk!îÐ-íز¼IÿÈЭ]çò=º)£Æ8åËLwé®ÂK¬´6]B1q?s¬ºhÊ+h prI‚ñxr@f…úv¡90µáÓ¿`Uò$v‰ƒðýê2þÔÜRb™ƒÏt`+£\è×]XVŽØSµ÷DÜ÷BÙEkE ZizÓ ½¤óûÛ*j;®õ©‚m¯BàïdK"GÌ W /ø@Mµ]4“ã)Âó..ÚÿL}.\}OÞ7ºôF]ЗÆ¡*¾æWªŠûDHal‰P%*e—×óa4ÔQþc\û·vr¬‚RNÆ-r‰2='üÇ9z4: Ò–NKYÛŠUz%[~V^Íöýðñ] ûè€sZŠÝzØV·BPºVÔhZ—´ÉNê¸3F)~è¹Û=õ¡^|ê K•BáDËqÙ®C»à®ò‡5ÐØ{zÌÏ^¥K¬ L¤ž¯¨¬Þ{Q•è÷wWÜK×3Œ6?¹[è)¼vÓxp°|šu©Ò@™\-u^¢q1)£T”yM`´‘÷p)<5ì`ÙÞ¸Ùªçgòúeþ9Z³-ôµ™yzm€W+h®ŸÔÀ°<²ÄE3ª',X!·)~sLáñÅwÔ4¥Þõ›Èùžœ] |OåÄ]Á·)Ÿ".Ê`5ØËîAò•![bUÇ­hîN$/§šþêEAiªëÓÃå>#ãÌÎ’ý›é6õ^•ÍnF&]m¼Ÿ‰—-¦‘ð ®N¾®–9ê!¥"$q¥BZ"l×½Ï5ìvœÊÍrFÀŽÇŒ\µ=†Ä1g¤g è£Íz´°í·Àé¶4;@ß é÷IÑö´k(ƒÜ9…Fè椖*—&ñ»šHÑg¤ß³r¶Ùk¼ðÃï¦u¿ïÚ`’~n½þ®Ó®ãîñ±Üº®öà»ûwe¨Œ¦ƒ@…ì«E”ÖÓŒémP¡s6÷’€Ñ>\í\§¡Ô*ÞóöZZÿ„DyóÝçú(Sæµ’LŠ¨^r½soO«æ¶Ân<‡v&Ònºb5"ÌXÓ÷I–%·¥¬½à·ãa]*à ëUù=T*?wÊá"P$¡nâ.Q0繑ó¦Hªk 4û"=9^¹™Jd0{òÔ©î&õ‚”붠:§ª@¥‘[Æ€¼ÐºÚÓ·6°R»óEÉŒ÷ÆW¾û1’ |’ÎÞK(”[ÛÖu¾Ñm ÅfŒæÐÁP‹ªÜ«ªíû¬¼èÛ¹ò²¦B•Ö6”‘æóñsþÏÁLºº¢š‚Iö6&wU¬)¥ðR¤ÆÔÚ·G¢r†‰) 7øhøšë1Èjxn|Özv?nbôLÊ—SîÞªGº¬×ëæd.fòÏxv¶¦2"ÝÁž{Ï€‘×kÎåY oÓ†ÄL¼(Ä÷;¼6;ÜÐŒ}åô¤b8 ywËá¶C­F«¶ëö®‘qè] Þ¦þŽé§Dµòî4UⓃó$‚Sy%&Q#ê–Ö÷ÉØt+¦íØDÛ}õÃ-s}Ï‘¾­úR±å}v·O¡Eþ43ùdõ"fv†Mê¦Î7¢umÈÔ™ºÍ<ÜW€.;­Ü\ŒåÉ7‚ßM“Þ‘‘„ø.R ÔÊ9¡Áo™&z–†û}º[Uð _—^re»€ ¯ô}ÒyÛ8×ìþ£@+2^Õݸ`Þƒ®P‚ÇÅ_ÃÄ^^¡^\fœàôÇ4IËp$YˆwY±ô1°9?$™?wy§ÍO7·Ü–‚$ºwÝj¶ÜªÄ~S„wýÌ ëýt©·Kmäѳ—ÈÒZŃ¥0þ‚Ï.9}a@JV<ŽÙ§jT…-jg2Æñ. Í1þÅî•ÄxÈî«= £&©ðgÛlHjtzççÍ5þï¦Í ÒoUøJÜø¼eõŒß+‚Ÿ^d'¤Hê8>6'#pô)¿«* {co/öÏ~¿q$Ã<Ôb2Ž7M7õIs̸¦,ÊëA½Q¤ÂYl–!ÂË\Qw Ó·(ÿź6M}ý2C\3Ä„´<‘´$’äÕÙýýo”½-ªyYUžØ¿½½÷”£¯;cn®&j¸óniÃôMså×ÝÖÏ® IÙ²há˧ž©Pძ‹»óz¸}_ó·™ëÉ@Û&S0”×çdGº"ñê°Tvu)KØ%“Ó¸i§}ÅVÝ¥9‰Ü§LâýR2†ÄW,&}ã4÷NOÊ[å- ¨%ÓvV™MŸÖáÉ8ðäÐxž€/LΈ\­‹áC£-Nb%¸ ×'Ƙҟ}NÝ!^øU¦‚ÝÝg¢ñÙafØòw÷‘ÔÒy‹žêøäÔ–ħ îÖw÷â$R˜í f´‰±ëøG±}@ ÆE£éØçÜ‹rIÕ(¶i·]æy¬1÷– Ƽ¶¾˜\XÙ2W yËç~ßžšr7†ëÀeùÑòð[Õš>¸½Ùè>kP}t Çc×®—¸à4§ä÷Ó=휦FV­J4å¥ÏuNH(ì›g„äzãYÉ’~ß‚S-²;+&n§Kök ËÞÄ2YæÌ(©êeA'\+ÉO»|è½égžS¨î:\ˆ•““}¸"É9Óý:nÞ0¢£IåÌ.%™W@˜Š±ÚÓëÉwnò©¢v¡Ä+Î ògÿ®´±u endstream endobj 265 0 obj << /Length1 1399 /Length2 6072 /Length3 0 /Length 7032 /Filter /FlateDecode >> stream xÚxX“mß>"R’Òé@:7îîncÀ`lÈFw‡tI ‚tHH£¤€´”(HI‡€€úMŸ÷}Ÿ÷ÿ?Žï;vÛ}¿óW×uþî{;‹¾¿‚ÒªŠD ùA@I€’Ž‘…€@!Bvvcý'd7…z `H„ä0”< `4S£1D$ é €„ QI˜$%þ&"=$Ê`/˜@G ‰D@Q„ìJHw_˜“3“çïK„’ãûíPpƒzÀ `@Œv†ºa2BÀp€ƒ¢}ÿ‚KÚv—ôöö»¡N²Ü|oÚ`EA=¼ €_-tÁnÐ?­ ²Œa¨¿ FHG´7Ø ÀpŠ@a\<P&;ÀHC çEüEÖþ‹Àø³9è_áþxÿ CüvC H7w0†p8ÂàP€žª¶ÚÍ#~ÁpãöÃà`{ áwé`€ª‚ŒéðO(ˆÌ@Áà¿zü³Í*%¤›FþªOæ…`öÝWðÏáº"Þÿ¿WŽ0„ƒã¯6<ÝM°‡žP å? DøoÌ Šˆ%DEE@èCÔâ,ø+±¯;ô·ñ7Œé!ÐßépÄ´ „9B1„þ(°€öð„úÿ§áŸ+Bàƒ öP'‚ðßÑ10Ôñ¯5æü=`>+ F~ ð×ë_WÖ…9 pßÓ± ‘ž‘’¢)ÿeTTDúüù…$ü¢@ˆ‰‰ÿG ûSÇøj ‘‰¿ÊÅìÓß%{ýÑןáü3–.£\(€ëßBB0o ÿ³Ü»üÿTþ+Êÿ*ôÿ®HÕÿmçú‹ðÿØÁn0¸ïF¹žhÌè 1³€øoªô¯ÑÕ:À<ÝþÛªc¦Aá„Q4?è¾ðþ_8 ¥ ó:èÃÐç¿TónòkÞà0T‰‚ýºÃ`¼€Àÿ²a† ⊹‹ 0Òüm‚bfèŸyU¤Ã¯a€=<À¾„˜³Æ¬Dþ ÌT:@}~‹ (€@¢1.LG¤ᯃÝ`OÔ/ô7 „<áh˜;&Á/ìÙ ž˜Ñû- L)¯Ï9ê….Ì"!R./":.ê¼ù7Fq—Wºb“-úcDÐsOýµñrÔ¦*Ú9ÔÐŽ<ÖŸ¯ˆšœ `à9¸óyð:/t)S½¥Ò¿¦xÄ_8ôíäµ”ÙIï4.ö8vü0Å€‰ÔÎ ÇŠ3Ílö*R¬Gxóùçâë.Ô¦¥Ä.À PzFÆ’É€NPÈïÚÅ~dÁ¶‚¶x@–xÇÃ*ËÓXÃdüû,ñ\±#žî›ïšUböðÇoòú iõÖŠÖDQÐQúi²ä¤ÒQ¥äi¾Daq(Šœ)Ud$³Ò•¸Eq,Ý´[ú>]4О2(ìÙ䤫z[/èm³gRdV&¿gé˜Ñv=×ø› 1mKj¬3_Ýc6EöZ¬³á¬ÃZaxA/53·¡kñ§Î4²¬üÜBû}s<^P¸°Íç Ú »„œu† \|ðÓ3EQ}8³ûˆ£Ùû+é =;…¾ÆM]—Ï‘{|_ŸõêÌï¦!N·ZÞt ÷D!Ù˜¨¦B“Í;svûvR^jˆœ½#7åQE×kÍ!»m¹“ò“¦ƒ òqôÛ»Ô¤,å%æ'X“L…I É猵7±²t ,$#B¦ÉH(\*˜ž¿¯}ûT&¬E&C<´8qtCÙ`/‰¬~í¶AÏ·[6ÉöG‚Bø6Z]–ÎveºœšèûÉ°æÝä«-E‹©Ð#±Ù¥Å&Á¶åA ¾'~Ý{ËÔi9Șõjb¯·“3cÓG½éò çˆÂ~XMr5¢^Õœ$Ò¦ôq€*ÖÄj¬ÝqÚ‚ïÑóÖ±ùôÈê\êþz¨»ÍV“î¡û÷ʇŠÊ€;¤œó‘Âk5:!þ8D1¦ÁÊÄ_Åïƒ'šÂ„fXÎŽÍ„ëp9­÷ÔE»VlZâÈ.ÕÜÎz?R´Õ3 júAÎo–„©Š÷ÒþðIÐ\Oô¯²»þ hwÓÚì!ƒr˜N+nŒÅý岓¨ò"‹¤™·Eæú*GÚ×Bk¶B­·kZ+êþ„ Ÿ):Ê¡ÎîOº€v$ñÅÙ¥™î]&ÕôãÒýìÖ¤iìÓïékYð»ª?©k“c:ƒž3«æÛ)­ä†+]XJK{¸ð#½gἎã´y:»ûwÏárŒž¥Öƒ›`ÃêÏÖ•nQëdï`eZ•zCS‰,·ã_L¿ø¾IûÕü‡-zÚL*»ûí—¬AOy®{•hž·eÓ·Mžæù¬}ýjÞP-‡ßÌè:÷Á©fÀ+žB§ª›)”÷Ô¬J5°¼7]šøA¬Ô6LnŠlx¼èÕ–¿—*×Ó~W*=¯^fäƒ,_ÄÀZYU×ÖªoÍ94“†û·¼ •2æKpò ” éH¦åmùò©GòH;]ïP'Ø;Â74.ôm°+²½r(J4†Ôý\¹ çÂá>ß“™ã†B÷Œ†Mô5ýµÆ*FQô‘D£µ…û‚{^¯“Æ­ [¿Ï3·h³ëÊïf«‡G¥Ý‘^ÒçLÔ|6²ì•2Ã7ÀÖÈêlR¦#wÚ°+·’æÎ3!æE2G#³¨B®©‘¦É릊Ø gÄo4ˆ;¶Eè} J£Ã[õ¸#‰«bÓŸ§ØI" "ÔˆŠ/S ýžQœœ8nûGOØ+;8•^/òÏÎ^UÌÙ™ ¼|5°räT:ñ]Z­³|>ý\ÞÂ,žV {ÈN÷ù†VÌÕ}W=šýUò³&€Ôòp¾úH•GK4Ÿ]±ßeóˆá?¼iõžwïüòåØüÖôd{˜ÍÛKŽø“ckrqxFÒúKT+GDÄkÁäÕUîã¾0ñ öJõÁÂDà uh§k¡TÒVÆ•§ –Q,¡‚Èq¶¨¢ ùb‘b1Ûx ¿&;©~¼ƒ5rò¼‚ÂZw”?8Ð?fr`H¾`:“÷cŒäÆu<æŽZ\kœþ4ã_I¬[†uèÌŽúV¹}”Àj[Ñ-¦[ߨ6}¿>6Þ ®È±>Ý’›á#ÂÞ?¶GÂÓ²œÉDS~êÅxõØ–án¿mH6ØÜÔßg1p?'xÌpÞê!zÞñZª¤Á2s«{`ÝkžY•ȗ<‘ºI=Õ×]Š^yhn æB·Ê›<2Œ°®þ˜k罹$ÇÊýXâͲUŸ9*š0Rb¨¿Åöùe«äÏÎwÜC‰*÷æ·J8«{7ºø[›‰Þa 2ôØ¿$F/t3ÄÏ !r‰¬µ¼¢ï+Ò&uº9iGR.`WÇåj«—¼kÓúpzÒœÏõïa>ïÜ"ÄÂ7£´½¤ ÷!…­ÎÛøk;~ù<àSÜkÎÓ¼ñú‰³ ´QùP1©ƒÜÇzs* ®p‹ðȾsW]³&‹WûªóÒ[Š ™Û® ‹½xˆû\ïgÆm„æÙÉ€Ø W×{níXñc¤Ûð·päô›½ŠûtmT5…¾O(Ò’Þ{ݽsÈ”\\ÀÜ—®ÒI|÷báfªÎÀîœ`Íõ.z>òæ•:4'jA_©qÛK÷`“UM4¨tø¯ ߺ T¨@“ü*>7`³Žk½ÕtæËrœZy5\,LË‘ÁŒÒoÉ£Àv_滫?gåý‘(s@)W7moF}Ô!Ÿà‡†æ¿)í^ˈ€ÖXåÌ÷ùI¿­UöÝ›ãè-.[NW¢«ñcÞ¬ÐÍ·(¬§·¾îDûêÇÔ¯ ­°ŒC}Pc:EL‘ú±tP[JVÔ­2$Ïþc‰øñG…H—Z“}­3 —Â6>ÜÑ ^{PÝì‹/D½¹peíC˛¡çÍúüäiÇq'§ô„Eç2.”$xA_ Ï£"rï±Cô—bdb¦¿µR¥,Û›ª~N,1<ÏmXÓ·\âÞñë~Ý-­A‹'õky8}LpÏãÙP„ ÌÆÅ<ëÙ­'Ò6 :}óóê˜#¶f§Ü`-µKn¯°}nðÆL¿îþê¬?‘y•Š-â+`žt“qjô9÷—ÏŒEÆ…ûX¥g ­‚"תª·q.ƒ«Šð@?šJµçL+wƾLÚW˜l³ûELÓj¾h7/¦/êFkøûºž àLß‹í·J»WÌ®|Oå]Xª»{­*OUù×ú¥,¯C½i‡1„âv1ùÔMG‰pžZ†=sɯ]›²ƒ9’<£KÚLŽØ£ T½©àT)^&ªv`4÷Ä{…T§Þ:-žPs­ìEÓfN¿Ë“· Lï¹ gã·fŒ;ÒÉ{³“µÇ–I4w·m”aå×V¥G’(é'Z„zX¾P.™D_äbeÞÃý&b±ë8i·óP?4ðüj*@æg|æûLˆ^€~/î*GâKsbàúF!¯“»B_ßk™—*!r¥¡0±Ü¯3D©‡™µMçÉ‚¶UˆZG€E™hü•¶8lVFo(‘3=]7‰~ÈSèú¬›ûñf4²Q¾®~53ô§íÑŽÂT°pÙªÕv·"³éiÛ‘„JQu²‹ á#k9@ÕÖï[µ¶¤"Âç%ªNÓÝ ÷Ÿ(fo­çî)”WÝYé|øbüh„ôîò«¥ÉûMxqÌÑÅÉ çúŽ´ïªOøfú DO×:n1>õ(.³JùøÚËîQGSý÷šÓî`Ÿs5¶Ÿ %;ã…­c·ß¦¿,§ù0Þ¥:š©«ÏI “¡­yíÈBsã.ÏHg¥®à¦7n_¬¡x—ü£?bHK§|¯.‡†ÛãüÏdÍYÞ¸e@¿ ËÁ©IÌ~²N™*©o]<çJ3“rKÚ÷"OדœŽ·SRÞÙ/P:pˆÖm,é‹ ½ kqUаŽæ¦Ñ¹ n6*úüÓ"sÑD–€‹›4ÃQÎc“È4˜Ây‘Ç"]¥½âx9 ½<Ö_üø¼‹y’~¿îÎÏ…JÞtIMH³ta/§ÛJ:Ûd;TÂŶ »Ì ãý×uÎRÝ\µÆÕ[Y²$Øý壯\¬ï N„Íe5 ýä”÷Cdt/"otÚ#ˆ™üÒØ[pɲ.ˆól=êu-÷¾@®Æ]`}„œÄª¯õ'){¯¶^X‡u2$Sí:<õWI¥.é}äº@·’M.ÀFeº²éLÎ;ósesue³¾¿[K…sql)¼¬%kWÂ1Ë”?9ù.üäì±û]¸AíÓ4ƒÏÙ±8ö+åT‚î'7šNž2§=ìRDZý1£@žÚØ¥0x‡÷³ÈM+ÑŸÃOJz¦;7ÔíxÍ &övœºq7ïÙNêë‡æ!Æ7Çhi†™x×cA|¹ !7:îü‘qÇÓÖÚýŒ•Õ¢ ß{\kµvìïý+ÈÊX: ˜ó¨eŸÖSÇMÅ*ÞêG‡*ˇVèÉ÷NzZŒ ÏØ.Ù^ÈPÒ ›Ë“Ûq†p“Y WÒg’ìÌ0Î`±óèª3öÒgo—äK)9<[ɺ5äáÞIW¢t¹.P%§'ëŽD¢˜Ðûjœç«J2Ûܹqj¸ÏpäÙÞÍ!/⮟áj½Ì-iº­£ÓFØÔAüpÀ‹£¹vø\iyCX«YÛÕàm¬ïéŸBêh¯»ÐÏÄËÿ¤¦¢ÐäÑãzÕ¯&ú¬íZcQQ¼US]P1 9`WÆ>Ì!G»QÊÉOâõ. þäx¸Ýó¯Rv3T›Ÿ&ðÕ%Ÿ?]±Pg²¥5­¶«ú·x‡ÚXLæБð>eS-—©®¼plJ/ç~Œç;ù¸•R‰]U¦=¬gp@1÷lá¬PMk^ß¹!H·ÑªGcØv'&D ÿȘÚoiÊ‘`“@z:3=þ*ô†Ý}ÌSY4«ÉÙq½ò‘Í.З4 ï}ðƒòòU]!WÇ"ð ÒSSLøI0ÙNã±×ÊN÷i°Ö¢™žÎô’°Ü\RB…αõ7I3:)Jr¯»É‘dY·ÌW¥nE(ÞM%ˆ-3§HÐ_Ñ’+`1.ŽÜ~‰¢Þ»yfb½þ<³VŒS4NEÿVqOÅ&)…||žqu:`†|ÆÊ> `g ø©èÍÚ±§rT^Z,Ià GÈuŸ¸„ÀõlYû¦=½—"„7X¥›WŽhH¤4†.qjÞª/;p:%ÐPÝ””ëA§Ï'ã™èAH\„”ܬJ>Ué‰É†ºš]º–Utb˜×r†¿8`*¥Î#¥ê‘6){1§»(®tpYDÍ‘“EËeßc ÿ’9‰æ4Ïp ’àU«?Ò]ÇÅÏvˆ÷ï·>Žwk}"Mw{à Y~û\"U‰½Ý# ì(\‹Ì³w½‘î§ÛÙ,Ù*}úOÄTå´ã}.ú¤«Ç5¬¹¯ÒãžÄõ/è$žJ)Ú$éU¤ ëæQ¼-Üî‰h¿ÉörYc€dçg­.ô‡Ý;»Yl¯–íÂù%¨ ûJ6t !ùró¯‡v¦êšn0“í9¯ß´Ž+µ?ê&¸Þnß©²Ý’ž¼ÍàÅ!ÛC3öTÖYçÕÇÐiþ¨Ð“/f×IÑ[Ú&F—6Í{á†GÍ< o´_›ì=Ö1U¥ŠüãuÕØ2“uÌ ¬x ›¸èk>FG7Wýl+t©Œ*í `-kü|z-:Ii¡öÉò&øÔUÊ©Ú²”{‹ƒ8ˆ5b]%Mv4* Yʇ´ŸÈë;ûÅÝ÷µÆVûY\ëÃ0(v«Úcz—„å*Ü߸#´—ÏŒ¤§ƒA µ§¥¯-å cÖ´Êgö¼›¯õ¢¬çáµ>Ÿ8E½×DÍ_(‘u+k1ëáѽŽvÌår’÷¶v,5Íé±Ú Ú©~Ûf¶!Uöª&\ÕËõšªelºáEDèw¹aܲt9\øHiï,÷†U®Ÿïxê T½è:#òØÁPaÏŽ«?RÞ?Yæ–¯™l÷Q̺ìeJí¦¯üçþŸ­&+—ø¿ÿRèU—„øgIû*Ú‹UÀY8|’ÚŽnÉWD¿;ÌÀçw(ÄYìô\ßy€Õ3Šý™5+øœÓó`8€ŠEvQø;°1¦·ã™ÍÌótZî™l~ÿ—zI­uz³—Bd†å=¤>o¶vñïÆôãŸz+:M¤¬v\]·çç²ÎUÂsíÎü¤Zë³4|g¨ÌÝz=²ëîQîªCôRW¾Êñê©pF¼w8mtþ¹)tŸ{KÐÝpðbiáU%+­P“'a– 4Uä‚·o¡UfyÌIOYßZö5ëÓ¦}ŒÁ úÓ£:#ø—þ3@2½žÙRÈj»…–ß]ʸˆÍ«µsj•<¾5¨©Ð¹Œ ’‚›’?T2hú©XrŠÞdAh²Ev3÷&•´ÛÀ¬íi¸Þõ×É”ûï¶7XÉ‚ Kìå-¦U¤m°þã.®bÔ“4%ß×g¤B®{y‰![Íß< hcºB? GË3þâ4¥\ endstream endobj 267 0 obj << /Length1 1409 /Length2 6124 /Length3 0 /Length 7090 /Filter /FlateDecode >> stream xÚtT”ïö.Ý(Ýà €„0C‡„tHƒ´à0 0Ä Cw#% )¥R"‚HJwKw#]"(ðãœóÿ{׺wÍZß|ﳟ½ß½ß÷y>vVm=^Y+„%T Gñòó$òzÆ¢H bgׇ¡ a"v(Ò†€Kü/‚< F¡10 ÍÓ@Àj®~A¿ˆ¿¨ÄÿED % `7˜@ƒ †€C]ˆØåNžH˜- ½Í¿^œ.¿¸¸èƒßéYG(Ã`”-Ô½#ìÐC@`P”ç?JpJÚ¢PN@ »»;ØÑ…´‘æzp‡¡lºP(Ò jø50@ìý3;@ßæò×CX£ÜÁH( 8À P¸ :ÃnEЛôTÕZNPø²úÂÀß³ðóñÿ»Üßì_…`ðßÉ`áè†{Âà6k˜ ¥¤Î‡ò@=€áV¿ˆ`:ì†9€-ф߃J²:0zÀ¿ã¹@0'” Ÿ Ìá׈À_eЧ¬·’G8:Bá(¢_ý)ÀPúØ=nÖŽp‡{ÿ]XÃàVÖ¿†°ru>Ãœ]¡ª )hˆè?˜ ‰‹ˆ‰ Î¨Äø«¼¾§ôwÿŒžÀ×Û á°Fõ…YCÑDÞ.`7(…t…úzÿïÀ?WDüü+°„ÚÀàDÿ©Ž†¡ÖÖèËGÂ<¦ ´öø _¿¿=EËË wðüý÷ýMLdÕôžðü™øß199„À›—_À+.Ìà@?Džÿ,£ †ýmôŸ\U¸5 þ§[ô1ý«c·¿àük.À?ki"Ъ…8ÿ#r30‚~ðÿKýwÊÿMῪü¿Dþß )¹:8üsþŽÿa°#ÌÁó/-ZWÚ´ àÿM5„þ1­Ô æêøßQUmY¸ Z̼üB| ¡?8ÌE æµÒ†¡ ¶$óòËj08TáûõmAg@ÿCû bþ~¸ uù;EÛçŸû*Â!«_>€‘H°'-'aa€7?ÚVPßJùà:€žÑ`@ýºVt] ÔúËK¿¿14䃻ºü@ûè„„¡Ïèòý!®H$:÷·LÐÍýkýÛôP¨B4=€< ±« iø^.Ëèλ>€·¸Üô<θ#BÅ1ùÚÛVÿ•ò¨³Ü3«º¾í©Â°‘ FîýïƒfŸ3’U²QŠ«r‡¼oz.N>Û@Y¬ 6C¢)ÏWÉâu˜o?3Å1½ÿÒpâg¨h«àféFîU“˦‰øÀÿŒ5™#›Õ´ƒeÌʶŒê56#‹¥@š¦ºê«>ºÂŠ‚ÅžË5DÓ_x®šÆ¦õÎÏwgtv¤¿}ÜV&RFIOå¥ÆšÕ—ÐEOŸ¡Vë‚Á!'üM¾0)î.}žcǶcå8?[l;=ÈPÀd×' ‹ÀL@jÝVª®oéàè–dýéå‰÷š]jàVR?I Œî…•pB¶= zã›6šŽž+W¯/N&¹ ì€Ù2#õºéúöZ9峌¢œUnij]F.m~N‹sª—‚ >½£S*¬0b2£7«^Iâ£#2Îýb{²a&Oÿ‚‚‚Nþ®Ã‘ÑÆä+å£MÌÀ 9y">ÓŽa:nÙ} ÅMÓe ‹K㈾6‘Bå²åŒSÖ¢‰eâ0y³&îAŸ9bNh­z{°÷5I´r¹ÞhÎìÚ8w2|¬ÄÈðV=ÃÄ{(Àù$!Ê·±'gÙ¤Uá½°þ>@³’Xª¯ZœÇŸ¢wƒïþñ˜1ByD’á^ÍvÐt,Mhª¾ÀΫÑËyi„å*^R^M_ÐL××2åkBLËTÞš*/(h_ÒôP€ dÈå¹ó¢Èæý1~N¼ƒaUO$=¸¾ÿµÜ5õ8ýF¦€s'Ç“‚2þcÕâØÍi74Ó¥c9qGOþÁñ‘z¬s \Τ¥3/rI­a•Ñ ïì6—JOÕ÷3gpcªL vëÃIÂ;±qùv±*ß–°Š=O#o­½Ïé…”¡üÂaɉŽ^eU—4H—÷¢p‹KûŒÜ¤&ê3ûQÜ©‚Úvò£®•e¼ê¦ë·¶KBÓ†<ã;Ùo[¾½¨s_þ>l ð'’E 笯9;ÞÓ’P³Ã›OÈ o‡œ™ ú×Ìœ² ©4h»åð˜rÒoE»„­¡Éêã1³õ6›0ówê_|,Eå »üoÛï;^ÄÒu|!,e6$zS…X>Ön.œµîe‘Õš‹&ëÁˆ#à5Öç@òyNäõ\ˆ‡¯æ\EWÖÎL2ÄÍtÓ<ì(”ïø.ÖÅ5Ç ò‘má^ÚMúÄÁÃW`r6ÁËƳŸ ¡¥³ÅF’›>Œ÷D!ä…œù1©‰àG=ÝþY¸ˆ“íÔ+zk/¯c™ÈǽŒÄTÓô—lÆ5j`¤óqo?YsâANôÆÁñ>ß ¿õQóù§ÒÌ5*9ò1÷ûB?Ê\‘ù*Ó8üIBºM¤ R’“òWÞ¯ŠÖ*úKB½}L?ñCÛ.Ý+Z#:+ ùÌÞG„á½ÌI­RjŸÚâ¡Äx—OL§8ÁÔy~|Ìi§o¾;¿Ò%~ É/?LÓâÌ&íÚ~òÉ7®î ÷-QçÇÀö÷1‰q²È9ãy©ÂÄ%0E$%!íi­Æw².®%*iqªvàgJ(‰{¸½aâu™-Ù r#=î$Dú9u‘†±{ ¸…øŠÂ¢êKv+¯,G_›û¬Âd¬G¶ÙYKøÇÖ¾/ß?TÈ¥¹O/3¾r«FøƒB/¢s8 ÿÇ §ƒ N¤ÿ¾G­‘µ(|fI9hè΀Õ2„ŸÖ€¬Rl:Ò1s±…oÁÔWò|UôŠt¦>F[}AÔÏ ×%;»^Pñº¼'&ºÌ7`¡7ƒöÛ”Âíç={Ùß6qQçŒ7k]ïwf B·Û›ûwY'ŒÄEïKn³$}mu$ààÚUk"Ðs§b¬ÞÝ+^øÄT´”U5ˆÓ?T–bP÷U-yÝùApì““¶yÓtµ/:é £³¸ÕSÄÊVB«ÆrèbÊ“»©\ÏÌÚàût—+U<Ó_^·?H¢|‘,š9š^?Ó~Xr/òN?Kf϶‡edz`ãÀG„E=Yp„wÖøMçøÜí‡,®"Ôž-³›¢†+…_>qùŽ¶âžËíLé{¯4úDJûYÉãUN… ¶9ä <ßù˜\ø8?%0TD芓h›¼Ó[T‘ Æ©W¡>à ÷ö|0²VùŠnîÖvxji}SúØ»ÂÑ­×A©[‡c±úó4”%.Ù8â¶g¦Ì—à{+?ÀT0²É.¶¸ë$Zlv"¨sË=næ Vëy¡" ½´õ»Å”zÇùÊ}VŒ9§AŸy"&n%~´bà‹!Ó5 èäflu¼†›°Šžmìa Òïb°T~ÞB\zœº²=]À‰¾_x"LâV” #yxaÖF]Þ,%rhÜO»§´Çî?cug+½ªX‰uèÑ⾯ï»HgEšÛŒIÅž–¹a`Ã%VØ¥—KÓãorÀ2.mÚëz¦îš Nh×ã·CeýW©ÌÉF(^_¶-ÁZuÏÜ›ü/cF9ØcÔBóCHå Ó¨™ƒÕ?ãÞqLA+É6²Òp…-ÊÌ"¬<ïÆ/¸X˜ߟ»+;ÿ•¿˜3ÿÀ@`¦A“@XZáTý»mŸë+MÑP ?ö®>B×RºÀ/‹F \l˜[ª_RVx‰´& ö~™ìŠ¤ñ?dÉ(aå¯ ±³rìf”%‡& Z‹eO8\n[ÿÔÐMçÝ~*Tº€c1Sg`Ü7þ¦ì>²ãºß†gÁsc¹ÆRe„[ê^"%kì,}eöP.gº£Ÿ¬:Ž”{HÛeSâü+áçäÚ¦"Oû¸Í×FÏ p:S"˜*s¹¼Sœ-]­.é7ø6"Ô£4Ë¡\ƃÒ"ì;F¤kO} z¿áãZ{ò› HÝŽÛ*.ju Žf(zt:A·!‡“Rl¸ãdpGxŽ]¾ÂðUsÛð­#2ýBxDõ ·3‰€P½:G‚onš°A?àŠ ‹¿”¾®Œ‹ÉB}j¾c2žü`­ºÀ3Îõ¼!áë“oyãòJ yëá·ê!˜>š^„Yê»,Qv,¼UmîÍÒ›ZÏîÞët!•MC¹ºVçéÅúzì &‰tÓÂþ›½>TvÀŒwÿà E,-ž5+¹¼HªeReÝò‡þ:ð9&n¸Ó'ú2Òæ¥bfÖʧ‘g©ÀŸ°:û¯Î­ORnÖò†Þì}lnwþØMQ¥ü.(M¹I˜ WR¶îlgRµHÓqÁÏ°r5è­Y0R÷.³ì›Ò매4Åf?áóõPrVŽãî-ŠˆšçìÃf0…ú«€“¨£¾oéÞcH¶˜ãÛ%³QÂ2}[ jÆ)pÙ\l‚<žøÕÑË8*"üŸØð˜¼ÞF]½$ìëx¿ÑUÒˆH[T×ô„³BE™n²Híâ×î‹L½}%EƒTä—Aõ´~Y¨ã=¿-”‰ÉMÆsͬ.«³ÿÄ(€í3›[Ïø¾æ©*ÑrÐlÑæ—(’… nÂÄ ÝÅç9ÆìXÕšÂ=+á"»c»MùÖIØÅáHSz'=þ¹žÁþB—²¢€öÜ–ó’-`‡'Ô)x<‚g `2'2©] :]r±Ç˧Æô毞ˆ˜I ±ˆwÂè¬Îö:&üHkûõˆé]»&IIÖKÊåÒ6ûqùÔy–bµ¦Šmb’-u¦%Í­+¯fU ™ô„+6,}<’ ¾í2«v>ùW2´¼î0µT›'Úvt®´D¥åÅNÕ³‰¤ŠE*ùÙÈXP‹}»ô>=sµò`Š_ñ hÆg‘›ß?Å[–…}¶û8Æ®#fÚqyS3 JRi³£ï=*íä÷Â¥L5ë>Ay, U$HxLŸ·eM—`DÿáƒÆÙââ¼xE­\ÌâòØwUÄá ÑÚ³¸/‹‚¢T0-ÞfþrxVöiŒroœD¸UVãú^Ûâ0VÙ°¦žmërjNeB1À-ªÞ™W>Äd’óp½j¦&`õûÐ{/8, ¯,®È*UG*ó´œ^å ·¼¿# “Ûaïœ÷#çéWÖê©™‡·+ÿ×Yä8壄èµñ-G±†iØgó;¼ñ¸˜ ]ªSVœëTKΰ£Þ< ¶Kt¡ëñÞ?ß]ôžÝâ ·çQÑ";À-_¼2È¿‰<Öê—P 9ðŒ–.u.æÄ_8gÂÃeæ´o•;¹Œ4¯EèŽ~Û[è7úU0Éû Ì“Toaï±ÛjÔ/yÇÖ8OaÐ1…ñß~u¯f’¹`n8OŸÖòùXùp-°)¹¥+lÉSiv÷+ÿá{ÒCMà$ã±m¾‘]ÐŽƒ­ûK}š•ÚVsjæÌ´^ó®“cG(–Êë–T¬-LpèÔ#‡'w‡XM¦ºíÛB¤yi¤ñ_«öPªmðêÊrk?ªeum~~X³ÇÕ;Û6á†óú2çÃ՛׋ÐÔæ¸Ò‡†?óÅ×áU’ß©ûI;$Ô¥\iö7TÏ@MI?·åïtf¬yÍt5oMˆûGÏ´ÜVuLŸÕU³Dÿ<ŸéàŠá©—ð¹|Qg LFT2ºÄzaóø¹ùèe½ynÜM{¶=í…Tjó¶SV47T‰z/È™kã§1:=NcÀ[YÆýýÂÊ™j°Âôkd›ùiÄdÙZpÚ“–¶=žtö¾~B¬\Òøh>Ãøl †I!¤p‡§»õ@Õ#ürjõeÂó¦hN­ƒE6ìXÙ0#5!ýMãô÷w9ø˜‚™½¿9#ÃìËFägî!9šVn„~RGì¨36(hdœé5ÜÍx?a.œÁ”¼â^2ƒ!§ßÁ¦OQêZÁÑ1©ø™¾¦•JIð‘1ŸshÀ±PãÍ!Ï'áÁïD#cš¢~òßÎgá ñ²H¥ûñYú]c•ÄOÏurØW¨ê†y¥'³¤50 ÍwG›×ׯ¨Žûm «V)'H²\“5¾½ùú¤½D ɼµÎÉûýà-Š( ÒÉþ"Ð[WFT[‡"¸a]è*íƒ.n¡Æ…HPnã4˸F(]pÕ¸9§Ìç•šôj0J÷°ƒŸe±£Àzg‹kÖ¸™Ç“dïJ‰vÉý6âÏîgSù,Õ&i ߥ<ð#éÐS'|8V £æž¾x³»N$ñÞXvJµóìø¥ý¨ðlF8ûžÓõõ›ç=ƒº¹°Ð°¡Ž¤G3ߞ͟?„†ÌŽóušýäíŨP8"±oaåÆv#Û:É" ú4ÝL©™›«ÝaRùØþ§¹;ö7.p~Yn¿–õòbšz’È š|e//.¢ÑZ2\úù? Ô‹ž¯µõ2êÙ™üeöu«†½µŽ8=Y¥¯:‡tn®oôA G°9fCU e98¶g {$ ´k ¥"Ukm§öæs$I³¨ýˆ{|r˜AЕ‚ÔÄEÁRÌóVÕ¾÷>ƒ!ídµ=“!rí>GáŸ~t`æ%’½$m;Í/K~U›Û¬ã¡ÍÜ¡ rˆ3y£1z’7S%*+ôYÿ¥v†SyF–Æ`áÝ–òävüëÝüYžú¤ø¦hþõ¤æ.îz÷Á¢ >Ï[î“Ö÷p *ø–I|õá©¥YlêF«]Úþ˜MÆ/±³ÄóF°¹h¡ázш‚ô:c‡·é®¦>·@8vfMs’% ?2#$M+q1Òv‘ºsµ‹µ"2šçÈgš9¹öT¸l¤¶>{/~ù]÷TFÑOß¿¦˜óìúhÚ(‚â~uä3ýËæ•åÞ”w9wÛ»eуÅôØô6—yWŸä6´±í<‡±®µ\æhx¦˜PQ‘oãºLP&¸’ë.·Dú^\LsR2£â>ÑÍ—IŽ€YlÍé«Ê|"CJˆ¿ØÖІÎzÏÔ•ž©Ý:©äªÚ}öÒ£Â@ág1VÚ3§b&žÀÉ­öŠ;-^Máâ \Lƒ %å¦ÕÌƤæÝ낤Hnÿdz?ÚsÂͺì×øÊ÷ÎH—™rªô «>Èë¦OÉR÷èp„G±Î€ßv•¢J2ÎL‘¯3éÖYÕLwèŽ{ç¼d½3Ì.H1OÃGì»H4É+×ËÒ߶v³šø£6sÈâßæ÷%¸NRCôCÀtX·CbëÞø¼ Ë†‹D¿o–oc»`?—'ï…þ{1¡zJìâ5?Á}²w·¾eŒÝ@ÔÔ±fó”uZ8Gs·Õj‘Ù¤t|ÙÖ¯&¼fb$ Éf‘3(Îø,¾d—@sëòeL”¡pO˜E,€*)·Öžd2+éY]]¿žD!tm;ðß|¼ÉýàжªJÛn£É{íC0q³‚ü©»¥-íéG`¹“ T¨š!û¾‡ vØjïÇÇ'ÄO±Ÿ•p è9ûí¥ñÀã Ѹ©,»ßG¤ž=Æi$Zî >ª‘ö©ÂRæYЗ‹ÚÎÔÝ8œ &Ôš8’]•ûèÐ:â½áâp—rðˆÿ\°:hß;Ê`]᥸‰Ó Ó,VÒIøB™N”¸Ÿ|à kj—ڮŠžSÛå®|CµÄ.¥QÂèˆÖbZ0ðæ)–h­JP0p{¬¯ˆ šXLª…Ä¿T\Eb™ ¡¥'©UêÅæ«ôI­§&NÂÕ‰Ïå_Þ±}Eä.wBp»£È* 94åó½Ú;£[]øb^ i¯i'SF»`Ó1’5ËÇÛ7T¼þÔ¡²³¹ç¸DìÍÐ+ jªkÉA5¬^å‚‘öG^ÍG ƒ}ŇgæHÖ£• D¹Équuõü®'rûó–Í tz9§µØrŒ­‡«€B «8O§Ql6“E—!jýà¤[ÁJÞh4¨Z©ã«Š×@!UîþÁ@Î*~¸ÿ™uD»¶šyñl0Óˬ]&³˜Ï/H@ ^s² Ç€dB÷ŒÛ= D†ä²÷ÞXoV˜­™KçB$ÚâõNÇ Šu±8·¯pö¶£÷)ØbQ¾G’,D*¸—€ˆüA<ÏZcik¾äÃ.²„£BÝ®•ÐbªOwì }ßSóÅi—Ù)òåäWŒ7úHÿº¸ÏFì±2•Å>¨Š5pqê•oü“[è©G+ónûûMÌÏU„;Ó‚ŽG"oÏÈ-»7u©›®nï¥Ô Ýó˜´53וb\N ¯÷oœÎÇyµ“¨VÙæåJíª¶¨U3ÍŸèþÌhÉØ0Â¥å©a¤ÁÈ,e!WjŸÐå?õ*&Â#Uè–À¢]±©Œ:ö.7>ö º¡¾Ü9˜ß$!î|œr¨c{æ@R0Õqó2Ÿ%Ó´ ¾=ŒþùV6Ÿ¹9Ä—š‰Ï¥ (yÉ|"CYúFæ¼âz¿ÏP­õÌzJ<’<ŠíÍ}¼ïô õÚçÖ’L[1¸ñV¨Ûs¶¥\:¾ñx¶ßl/ôbk:«ÇÍù±ÅJfË…$ïRɬýGï €œ‡¼;íØù…Ö3Ù‡±†Ù´ŽígÝïú««é…†áZÜzä,äññÞ\Rs¥k¦)‰l„­ò:ý’uÆ>¾‹Í°Žú€ôПâtØò4z\Eyç¯_ð¢ˆ"0dpõwsNßòg~ºÁÅ ±‹ÆÞjžïî­¹M/% sq¥Ÿä Øe cž%H—X·ïä”&$¼d/ÜÔ–¯'Œ%¾JO…{å›uO‚¯~~ Ứj^M®ûr±.þ)Ó Æ(¶®©ø"Ïν”Þ:ƒº£¨YáVÞ6Npe1çÕ ”;j¼ýA¡,ŠŽÇ·]ôh¼$âìÖ (µ[|4×]*Fb%ñÖÎ$Iø““¢å/‡­Æ ÚKæÛ}ûwÈœr²âGl S› Û°þÄi ù»v1k?=ž5Èh†cì>›.øÔ§Ô"Y+#Ëš ?˜,M&J¬Jð6J²tè3SŒ&+ù¥\ù °·Hvvëͳ¬Òšr:µ“ì$Å0È_‰?™IbÕºþ¸ìÍ q‡É‡ÕFŒ­)”ruZzЯ?šá<šP˜ç#ÙÞÿ>Å" endstream endobj 269 0 obj << /Length1 727 /Length2 14637 /Length3 0 /Length 15223 /Filter /FlateDecode >> stream xÚm·spfo5ÛvÞX£cÛ¶m¾±Ù±ÕéضmÛ¶:¶q3óÍ|÷VÝ:ÿl¬§öÚëYUç q{ ª§£3=37@E\X“™‰‰ ÀÄÀ GA!âlf´r°5šq4ÌL*fŽf 3@ÄÁÑÓÙÊ 6¡ùw nkdjegå Pw°up³2±𺹹 º¹¸28»ò3üsHÅÌ ´4˜[ÙšDµ¤ä%Ôòj 3{3g#[€¢«±­• @ÖÊÄÌÞÅŒ`îà °ýO0q°7µú'†°w3sþCÌÜÙÁ '¦*$® ¯ aTÙ›d%ÿ9ntáþmfbìñ¯í~€ÿ‹lþÿ/ú_×Ùã?33ÀÔÊ06³°²‡cü—nRöæŽÿ”M]ÿÛú‡Ë?üÔÿhH053ÿíjk+odg q°stš9äLÍœí¶¦1 ™½©™éÑFvV¶žÿ¿øÿƒý£‡½Å? 2ý§då"nåafªhüGrs#Û¤úw]Ãì?÷ó¯Aÿ9ýÏ(À?÷jê`oëùþ³Ì¿Ç1ª« IÈ*ÒýÏÿn‹Ù›8˜ZÙ[T€ÿ(jälú¿Â¿ÛŠFVÿñЩü×Sæÿ›Ë­<:L ÿ2 Ó¿žÿFzÿ%,ìàáMÏ geæ0ÿdý àbåòýÿÐ5quv6³þ[övùoþo;™™y˜™À­-;˜ð„X§5‡•ù‰Ì–CÒÍNªÑ-3ƒ®Lu_ë Z Zš^âÝ©âåº#Îóí^ìHÝ5‘Å™!OLF“„ë-㽑•[øUo½)å½Ö…ë_A¨ãÚ™ÈC\Gܯ¯)ÒN²ýÉ>«–v*áC•`Ž6,mïƒDÙŽ_]x°— Ñ£BËÐþB0Û§«B™µ Í¿2xÙ[û•¼‹pMk`™_èxøK;¯îÒZDl@c‰ªG§:‡Ùä%YØâÖÁеœ–Ém®Î‚*»·ò1 òø¾°«)™˜a/‘ÆAȤOdz@µ¼XF'…™ëatýÉPe‡$ºrשžT¡'ô‘ÜÀm‰–z·/ÓYý×7lÛ 9Øc"ˆ³ICN–ÀyžŽmé‡äž§6õOÑ›í×ümLEdjâM©¸¥èËž 1üì·êÈL3xÇc¡^¢ÓiH7tÒÁà`=£¹ëe%Ô‘qlAwŃ½7£jØÀc7ïÆã®!G­Mbt îšeû<‡4­“u¨¾^ÑÄ_=$‘¦ŠNÖ`d¥$—æà–4é Í\c :’‡ÚT"g ²ÍX±BÈEø"-v‚¬#q¬Þ:Î,þŠE“A‡„ÙÑÖ„%\¨{w?ƒ5Hmñ#§Jœìŵüqa”Â(ðµ†mçNÔž¡ˆ(ò|ïܼl^pýãKݶÿàÒi|ZQ*‡ÛTµõ%Zâü kw²® èþ9z/¼ÍäÑçXsQrë®g„ª·ts #ÅÏâ—@ÄŒìKË`C\Ù´a¬vŽ:#Ž4Ùx&´¥çƒ¸öÕ½fk•k½ TPŽ”W²D½G±¡’ž=a1 ì݇Aµ±ÌZeMØøÎèpÈWM§xƒ`2á—,-V—¬ü­äS[èÚ ƒdßDÃV»s‰ÂÚÌ(b¶•Á²'‚)kX)l€(½ã"¬ÿ°¿:Ê/‚®)·ë·œ+-¶Ïa³€%1‘(ÏÎyœ5¶= ܇wù‘%Ä5«CyŽÿÔ>…<Øýí´ÞzV×.»øÈ\â}Ó`ço²Ü‰‰vÈ!TÒ 9mßÔâÀ‰V"¤¤Ír†Í×mŽmåSgþë.šFö[£лÉ-법`¹¼x·í×ÉîŠá=%HÞ+]Åï‹o¥àůñ Ú…Å”h¸½rÐß3áO^Ôç{†…èmL©yNärÐÛO«ì(ùŸ4b/Eq3çWt^ 'eÜᦢv3ÈœõñÜmA¾.†À“˜¶N´1g¦Íd(«n¿1R#ªñëâ*cÝW}¬úa>&JÏÚ$èuù‘©?"Á`@ÖØÔÀó—eˈ–ÑuESi¾#ðìH‰Ù»k 6ï%2Úâ²QAØâ9”?)i‹™ž5àzQIüd{¢ §WßóUã˜}íæ€9°GP`y»\d>$ÙඖtwMF•R}³\¬cqï0”…µk@Y´éÕT ¾îÔA¡S,œ”’õÅ{aìdø¸ }¢¶:Þº„·90 ÅY¾2¼5)Øm}›‰ø ¯O›î\¥„{°@lçG¾·Xµ‚$,‘!¾Hÿôã®GÄcýÓ]²4—µ*|õC•¼”Á5@q-èÜ«psš4-+Çc÷y‰ Ø }‰0™S ¡äßÏ…¦G|ZC$ïîÃÙÊõœÅßêw9aö²ˆä­ø#ù¸`Ãà.¨Ûíe"jÈY‰RãG$º_\9õÀ¦Ý×·)]Þ<¨|?ß”|ñ!Ê xƒ2L¾zÏG‡/T~ŸêZPTWâ y–óÁò= É©PÚ’d“ÏšËu›®3Bƒ]Èâä”~ _Îp–“y< Ÿ«Ê^)ú{”¸{4%ÏfÉ6'² SœÖW••Ió˜þúÙ þBk%ö„_zÜ Æ%-Á¦qÉ+,ý‰âË5Çtkp1 :Ô‹=ò!ïÑÝÐÜþÄzÿžY€Q´ 3Ýì¤!rS½ý%WØ%sôzß‘vóûW,òiu6̯=>ÄTí* )vÆ™¢!øüît5Ô*ýµ˜{nïÖÔNhTËä‘ö$¬@ ¶ëôinŸìæݯ d±9[È¢’9'ðWÔW iPØçaHç]\šü®ïC©çª?]i€)ƒOzŽ8K®¸Ã22ÓAª=0L¸¿.{ÔôøVžûÏÈE5~•±çœ¨Ýtc’e”ïUs-‹˜ÅÖà÷ ‚ÝïÁê³dÕûÞVëõ“ž[qñÌFÓ` T7`íÁ„©ï¦žuTgÉ)]‘sï«~QÚkÔ÷©þ†nÌãKÿõp¦°¤¡„ÏGÂNè){ûÙ ôŸ¬„vf »ûEB[º† OfŽ¿¿ËÔÐLä™qHy˜þÂTm¤µ©€wZž™Gbȱ»5NUAƒ ;¤+âô€U·¼÷Úo%Od¨4ËÛ¸$ظ €¹.ÁˆñØ%o®Ë5]wØ£æ?.=¥ø0n‘Op´TNËt¹j²§Lüx¹µbØ7ãÄt¬Íñ)ßv‰ÅÜژ˰X† ?àæÍúÈT`ƒ'Áâ+i讘K ¾}î-kòºÇD;Ú1ó³po÷çW«C¯#hÊ®[Ÿ*©Ønœ ’(‰×Yø,ûÀ¤n7Í2"ŒâŸ·~l¥â!ßú²/~þR6Ø“]ú‰Z"JjvgzOŠœþÀR|ëXôéÛ« ¼@2àŒ \¸J>Vo§_sô¨²¤èwÛäDÀ²çI½bCÅYû^Ÿ °«]yF„§‚.(&¶Îí{^°#yȦÀÈáü%<`óué0±Ë%tDÆâ $AéŠ+nT)c7¬J+ÌZ41ß_Jë}Qs@­¹^Çž,²?DË€^(íii}]¸i*õ¹~}æË?˜ó§bG “=HãEª7§R—e(?)b¥:T¼ÄˆÃª¶Å5\„ùxVã‚–<‘Ž1\£yHœ¥iÜí={%4ÑħÂÃêE0™óÝ98²µ•‘†ªó<+«Ž:ßÉ!Z%6ÆÅ­†pœ|KIF]ù çùÓcYy$+,2FÍ"Ðÿ|Ú9.âµ OÐ¥-¼¥–”UbÖ vèK|D§„4b.& B¢™Ùó©ÑÅ—Q¨.µ‚¶"Ö¶UNáDZPA5Â>}=G8Ñ…FÆZ ÞúP,,q S­ä òLÑÓ:{ˆ&ÄÐö|y,Þ=¿\ó”õ“6û[ðB£¼;ɼ\à\¼ÍA<°—8;ÍÓFqûí 4ÌÜ!Áß™ Σ”^¸XžyþYÅVˉ‰)ðxØ;ÏE¡‰§câ&L¾}c%›Z§‡~OÖïÐÚm,$áIå)K–ÏÍØx¾Ç,ËsÞÞ3‡rªÈ&ÉŒˆÛlbžû&ž–¤ƒµýbÖE§³ë„ê%܇FiöÌËíFÅ.F»f²ÔVÝ‚B…33‹7Ûü=/ûeóHKýdA¢8"M†¤ëÛYp•TˢߋQä®ô£Eà«ä¬ÖàöÏC²B)Áâ„N-øTÓÕ8,åÍ,þFf@'â—úö«ç÷ÂVw_… po.UrÇ¿ÁRùMµûóéü¨êAS¥ø蔟_o b÷*ÜWw„3D¡ævg²(-&laŸ'¶´Ã™j#“É4‰dëðøìú¦‡¡ Æ¿<¸G ZúAè@Ã]Á?x2¬žé6â¹ÿö/Ê=ëgú•'Ìg‹F~¥R€‚Vì†à¯ b8ôú³·‡š]N5‹ÜXã¹Ur5ŒúÀÖÅ’’¬ÄÇ£tε7r$;šR·Y¾•T¾¥þL³Ý«\ñvàçŸ&SúT'˜R‹Tõ*¹ˆÿj»¬·C1¦s]Лן–M¤‚hÉ@qèdÝ9é¨fŽ÷Î ù‘Cçn“m¾¦/¬#×õDlrç­©-¶NCi+±Í7Ü”º°ù«\ïjrW*ÍE",r‚H–á–¾¥gíÇ0z¥Ì¥Á˜çjw~0F)"nó·çïŸGõ̶´ v=ä©Š¨¬ ™(ZЈ(wËÍ”óa¯ziCXÔ˜HçÓˆ¦Aù¯â€Ahf¡^Ò-·álP3ËFw<íä]B‰Dé6ÌÏÎ_ôñðW<GNRo}šˆ±w±&;][¡`ÙØDÝÝÄž`¿Å 7>vÓ×ØÀ”DÜ£Ó±ˆƒÔ£eÀ›¿!üdl¶/e1H5ŽaÇC§§jÄèwæ•`A'±AˆöMšå¤´¿‰‰ï«ï¹^q]ÓÔsISº(L­ÔT qÏiª%xâ¬0÷¢ùªšÍƒª¹Ñ˽UHB†Á7wúW BÁ¹Áá²4hPªÓß²ñ\ ´¸¿1|— c,zç]|N9Í‚ªM˜ûX_Çi ¸ÞøÚhOÍãˆ"ˆ³‰,¹fP\éâ.>= Aûðè‘ÏV›Â–6…ç°fzø…ÂcxÇ wúnŠ|kÈÝò#ûf…Á‘?‰ÏÚÛ¼ @ð âÂá⊕¾m³dçñ©»nÅ“§öàÒ´9˜vc0CKA ì­câT×8ìón{d ‹õ‘nfÄ#ÌÖÒëFÛõ×ý !1¼>Ï-Û ž¥˜nÉö]ÞÏ·ׄ¶LzuÐÒlìš{9„´”=NÖ?‰˜úÇ=ŸÕû…¹ …À»EõôIW€Òœ(”4k×_2w€AêvÝ=!<øV‰ r5H—º$U븧%š¡›Ð®ƒ€æ€Ã±Ý‘©ÔSiŽò[?â*Þñû8ÜÁ»rOX½úLrT0S»£Ê‡…C–”ÞW'L&³Ð[ˆìŠµ.'‰!ãé˜'°ñc™zÇAh¾_÷ÝËÇÌ ”½ì„+²¹_¨©¯E& 5‹¯XÇ’ËQn®65ñt1¦ïþ&¥OŸ5¾‹àÐϘN¡0ΤÂ6#!-\ñ-½u²ÄZıޡf‡ä5‡‘æ¾” d] È\Œ£ ——…BÔÐev1uåC(ó€ÚËîê¾húQ|Ÿ+÷Õ² þÞ‹4¢#x½Üþ”påM’IUd.{Žª%ÝøƒÍÁ.”›©bm1ìÙï»Dç³ÞLzŒ*&Vs©œ@}µÚ÷%?gfÖZZs©ÿ‚Gãpô1#¥"H86f~0ÒÆáð©úDb[%Ü2×Ts#pDOÑü˜@SϘç–uÆúg&Yc"´L³Ÿ‰b×ø…ÎÔñ„ƒ’z®e»T*K˜ê_„ðXh—Õ˜·ÛÙz&hÉSÕK”|ƒ¥0ieãÆîØ®ªQ®×|tòüTÝMÑ»Åe-ðÖ¬S­ áJr\ÏÌ8{x—ñ®ìò€P(A£¨ïØÚÜPŒhXkSÏéï¡Ç=ÙNªáV‘3‚sT/ßf x‘Þšáûe*vKzN¯¾úÿ}5:bœ nX{ët´’ËDYtÆÚn±íhG"÷#îYž=Á»·ßþXR`P¦ Ý™ãNûÝ&Ï„è/ïì[ôwÎw ]wUXÉŠHz)ÉÞué 5Ãœ(ø7³5üs䀩Q½¡q‰ÊjWöH1küŸÁGɞݎ}ç~×Æ{˜ŠèàpIÈ(Šo–iMuä‰Óå¾ãLÕ‰†oVù§'dŒ¾Žƒrþì›xvÑöˆ}è›L’MHÒ jl¿„» ¡Þ*âÁÁÑÖO%Ñ´ÏL£‚,Ûƒ"Bhuó“ÄÆKpl)u¤ø©0þˆ×­7C xÃúY¥á‰I¢Ýž€ŽÎÑ8qT$ª3È°‡œ5|“,¡:¸à˜—6s°—§i-¤‰n±¦©¹½~ ÝCÖZù”²™Œ$újDqn ÞTïÙÞÈ"Æò±ºáS%_óïÉBN›lY_¦Ëþ+ÕÆÀ£ó >Ó5SÇ|’¼ÁÀÀ1šÉ¤‡~mÉåW_ ºrzŸÉ<«11ÒÃj2ýçñm~ÐçI˜¡¦+W~0}’ —Ø­—M~LÌç>§wwé§m+}VÒ_¦Ü{ë5}ÝS?«Ã çÐËÖ¸†—׉¸ApM9…9Ü!<Óë:®Í¢÷|ÿˆóXß³ÙÂÖí„$ôõªØhIï ¸…~'ð H}”Jé}[:…o$*eÄi [zµê’õ&†àwìCd#è¡6Ò£énŠo¸j6û/=~ÚãO/‰&q/öF@ÏÁÈjwˆŠØÕÚ±i˜ù Ù¯!­mœ‹ÚrD ~Gôª‡¸Aóù “_¦ÁfóÔ,EªÓè\`ÒÜPÇ’ÍBÄT(SoêMÔçá·7pTñä Í›økp©!@²¦êô¼V½—€`¿6? ¶æ\``Gžê_'¿ÊekŠýtø O·‚¾1r3ÐÏ÷Ç#Täóζ­lA0(êèËìñQñ()ö!í Ẍ&&~m–ºýšDåI‚1ÕX³ÊoB,Ÿm)9¹ÅIo8Ü0å²*¦+ qhºö¼žÿËðW)—ʧ‘é¬.x ý§WIɃs]þØÁÙaÕMõ¸Õó/åü$q¥u­Yϱ$I˜(ÞäZ)>"ÿ­‹Äò~¼b÷sH²>±ó\ðXÉçëO·™-û©Zè¾ýáŽKá–q”2 «ECvkŽ«„²ì–ë!¡Jbâ5Á’ÖlÓ‡üØú[<)kc:ÓÁ´ðþÒœåz%|Ô[ž6ppY:öDÏÉfÆ+VÚ_~r.C¥\¯›eš³æ îöè* ¢`´m‘3ÂH^Ï·5‰(Á^ íŸ);-Ù¨FLql0µ%¥']°Ý÷gòtÆø0–˜G¼.X' )×i¼è!Q•µÕ=¶I-16Pðõ/”ßY…?\5¬@pı ¶”øÂÿÝB;Ä` Ú)ɦx;lý#cÍ’6}Ÿ¸J8üY—ÑLŠA‡MDèHéÑ{™ƒ|Š§ù ˆ—ÊÍŒuè2áý•ç¾…±ãQ–”ôQâýô‹@ãM• —˜<˜ŒjPÞä$¼»%¦Âl709Z~¯ŸÉçdV|Ñ·ËÊ3²A‚¶ï˜?9uPȶ½«âAê n;éRÜЊ¸½ª±¹”^ûS_zü”¦Ê&M8÷²(K„w˜d¹2¥À+[Vi±æ+Œ1¹´š^f’•&Ä¢“;KVÚ¦ŽVi‹»XŠŠÔÓAú®”A á—ãþ(kʳ±àˆRÑ« lD FÐ/ ¤ê pƒXº¦ þ¡†8úØõêÖý 0ýµÈO{“>#«­FÆJißÄXEÔyâë?l@hûÕ (oíRàôÓ tÒ˜` Æ°Ê[ÊÒËÉ“5h°Ð#Ož¨-öü|Ë4ECÒû@>û÷SÍ°Úĺa 3Š‚pRk2w•ë* ùùT5JçÍFî ´öaú.•„ª< u¢—¡ :·áébsEëyŸ †;(ÔÈoô£GiâØ‘nªÒæj|$|-J8+û2{K†©Ì7 ,ŒÌF„u”#˜—qÂÁ0EÞÂH Ü6#wZ¡Ã’p{Íïä>ê× æu€%_[Zu÷4d$A˜â#älÎçb,¿Kg U¸K:V-o£ÇvÈn‹Jœ}æ<ïS”D*Ñ!€`~]­4º§î„¤(¡:êgµPêàfÃÎi6{Ø3dEïÂH Sjš¼ê°°Ün3wRÞ¬íì@>¾)„ˆ[¾Þ í ü®ÄdÌ£µgÛL¥]+ž˜SôâÀ~* @7lƒ„”N!rZlÀ„ÞÀªC ÚïbÙQnVˆ4þ †LÄÜ×:ç1ˆc’¦p_ø:-y\/ÂV–Qæ+|QAÎ'Úþfιõ6s´áÜdí|½‚k+œ‘k/gfôë­ kh‚«@Ì•l…òႨŽ¾AŸ(ß¡þ¾¡…c&☬²ÁÞAÿú_Š‚PØ3ÉT£pìMœšÅúØØHVÑÞp?;Ì‚ºÎéЗà¬q£Nœ´~jgø£;IÁ”p» ²nZ?Ê£g•ÏU¢´°%„MõŸ ù}©‡‡ÊFhÔ¥ÇrqÉœÎu¤W‡ ŸùjÕµ¡S¸¾uÇ âÉú¾ü:¡/;ú">§k_iœ8ã?C/ƒ{S¼§GÄ"¾¼Úï×nÝ~ô[·x%Á¥Â¤qy¢Ã”OK5I~¤dÏè?EöŒúoÔH®.×´¼šý¥éµ;’ŠÐ‰ˆG¬ ¹Y&þY/ŠÇPY¤2DºÛâ¦:Ó<^h0“ C!ÇÏñØ¥²n¾ÿû 1’sS{ùGš)ª+ÔÝ—¶Öï·…Ϙ•Cv³7&[•UC\bwâ0*{»¢œw“r®°Ù«æRM†#t3<¥ØÏ£Ñì?§ëëÈ8«pnR×ñ°å§@Ñ.6»GMÕÎ ÷àìPíÆQh|Òt‹˜‹æP£àHDP-°56ù}†ŸûK#IDЯXβ¤ÑjƒãºhÌ ˜·­í¦7årêº**:ïîÊ¿{ª5Vl$†eSoíÞ>ÛhñÅ8ð,qñ»µþžpƆåÑ<Ñû3+`ÐÊ/OÅ4Y5ÊËåPö3Áµù{Þ™ÊþÄíß®±.ÒIÞ€SÃyŽöåE”5­£Š_a 4>xMR,©eÐé /]_Ðô{ÖK(g»¸ôó£Šol‡‘Ñ(“UãŠ_ì‘õauø»]Î Qœ­”ŽøI;H˜B)…WâªÛ¡Siug„xìo˜ÑªÏ_¦ Q8+<“[Ü òà)ÝüÛ†kqƒ ƒ¯Íjõî†OîGzß?O'rL‰Ñäs^Z%[«“è­Wîbúº'¦P=˜àh¿\¥Öƒ4 ÉÚ^EÛǪ›Ü£'²Öu‚Úó¶ÎnÖíTnö¸“8>r…?õÞn‰—hG•¬Ÿ LþçŽÅÅ{«M£j'‹Ï·ò`}!ŒÔ>’R_°Ò ˆ-h¬toz¿Ë\½ÞüãñŒ<‰]žuVäÁǽÉÜöGð%É=½Æ…q¾Þ³Óª¿_šÌÕªƒx´?xcÛi÷‰(³Ï4¢3wü¢U d×]µI•äÓÅ Bò %jŒ:Ÿa½ÊnÍÉFÛ»b!à“KG”Ö2«èù­¯ÈP‘«Ì•,HTÕNŽ,‘éÙ´ ?¬n#:Ûl]Þë=U najÐÀ.´´:a!P)•IÓ½ÂqƒGþÃ2^åËê,®w;¥„†ØGë((å[# t ŒJÔH´,ÿ÷“ÿôE7E—òˆðx)±;ýc誛9uÙÙñ¡)öE#¹½ú°üØ/È=ÁxžzõM|þÅx2yú#KnU^c ×8Ë ø»Ø›c{ª·6ÄТAQé*Ò:€70±×¨¿îúQ§s¸Öè,ÌÄT‘y ÅΞo ¶~ uPDK÷³Ñœ*ó7OûÍ:ï˜dåêÊûï[jå÷ß^†»º¦®Ws=”x†Q…qt Ï!gCe´¿/n \½@õ-E¡O™­_~ÝXTóÑákLÅ~N3À‘*(/²~em>‰Çzðë‰ë. YϨ*ò‡Ót‡ÀªüPZÃêåÍ$xϨ™âP+9öZ|ì/[ Ò>©~Š žB¯ Ú1ÜŽr™°§×­öõHG²(úÂ_T´ £ü†º6ý<õuQÊVMÜ.¾”¨¶ †@œšt:’Ž.2Íšëox¾LÉ÷Ü`‘øÕD'1²dN¶8 U ’¬O¥Ì%‰àoc½\(;>áy ]ðʾË7àM{þ1;†¢6…h ¬'¹ºe Ýñ³ÎH^èß'×L‰ÙZ%ùt﫾Ü”R“Ì‘Ÿ„ÇÊA#ý{–¡&šÀ†tasN[WAR ‰.öž™Ìõ)¯=zùë­Üm›ÿ/*Î.(€£Ã'YÈ]¦®4 º¶îÌ~ËÊàw”[­Í3šhˆ-Z·#/¸óÒÐü[ õ±McÇ(¢¶,ÏÆa],€p(¯Xl£ðò«I¯QË'\8=÷@›iY¿œ²Vk™JîÇqŸß0í@1eE2&•‚âU#PxBØ|è¦ÀÀÈIó†{™ =—üY¼¹\oš$‹÷Ž¬ݦ4XÑœO0.£©¨ÉÎz:“§›oýÍ­^‡)|w¤šUãFùyi$ŒMo%Ä<¼ª(˜µû7 â—Cõƒ*Ï鬆øüÌ:l.ÉÇ ß‰ÝˆÒÑaŒƒPP•¤x”,ŸÎ?c¤Ñv!\©“ã/¸ÐU¼}§àkÙéyDšµz q8jaPýÙ'ôÇÎ/#ÌËkEÖãÀ«ë#VÖÿ^ù:”]¸Œ†)w’‹9]¹wþEïúx Ïkã™­\Ö°­Ê”ùA˜Zº‚Þu{m­ÄíQ°$‹!9ïþÉ‹ Û°ò|šl¦‹žU&gm6’‘\)™ÇƒˆUkjÂ+¶—&.šq©'wN6·=¯ÙM)²Šþ摵ÿ)hÓsW·áU'KRʼXqìŠ7¿¦^»¢‰Òú CÉvêQœHoÛc×fYa•±ÆäÆÜY­"—j'B¢Z@¶˜fñæÏpôÙ¬…û à]ÆY0ßÌðÅÉ×Ôeù€¿% )±aŒŠ!¤G·à¹Dbâ(š[,GÉ–ã÷+¢ß2êmÓsÏšÓyÿvÈ`Ú^J)µ q—·¶èã‰+$Ó!é 5U=¨ÇºÛã;'sUi>‰/ž ‚­'¼Øêß|½¢t†ZŽ"‘zGþ°‰üðàxÜZ›—ä7¤Ì CVöÆ…¦<8:íÎîM¡~BÏs´Äã.°]ÎÀ”Síæ髳¹n4•îQK%0ã“Ydñä“ ÃôÚ.è¾JkËï´b&ŠØ ¢è©m±˜Ì?ç°cê‹ðÈ¿›6¢œÿš=ÜJöHrEÓ-Ðh «zê¾g†J3÷íõìþ¹„Ð;ÓˆéÚí¬C[ÛÃlNXߊÀ65œ©Z+“ƒØG_,\@ÂO_y/¥E^¦TF<Â|ææ^vÍÞÐûO‡9x¯ÞùtßEäÅ$óþÜVHP'ÅàózÙ!ŸëÁtë$ í|äîë¤S ›Ó›×ÄÇ &J`I¨«U”b™§:QF?h4¤vˆ”›Å w“Ò¥ž§¯`±\s6lné±7Ô+†{"§È:üÓE=›ÃS}KùqYl*´TaD¼!£Šþ±?‚ƒPÎ¥¢°e'4Õœ8+¥Ý²}ç\&næ,©·s-™¼1.ÀØDv]Éüˆ¾ÂF"ˆ;ô9ËNèÖaYÞ‘ð؃R¾aB•$ÏǼ¦v§}öÈÐ+ô”ÆûŽ"φÀÎÛ \åŽUc¨Ô(´²oŠñÈÁwÙ4pëå†`”Q_lîÙB0®_¨‰¿VÉ™dSÃÎá!u ‘¹Cë5;AaIO7¦ãF‘ö“Þ½|½»-e¸æ\8¨Sºè Ðk.þ”'y¿¢¾½68 ƒbÝàWÙƒÛø‹³©¡k¬Öó¢ ÷ƒ4Ä;vXD›×m8CìÖ+ ”%MUaÞšåœy}1åw^Ò*¦+£ºªT”©| fÜZYK:Û,O¹ê2¹ÍË ‘'Þ0¯Ç#hvªñ€n퓪üŽàÙ«ìC;dkhŸØÀ¼ÖqwîS.‚Èz›ìÕ¾ŸOœèz†rExü‰Ž$=â–DŒL¸&_"Õ¤±NãAGú£Ú—Pi’èZ¾\¿Od½KcÊ5Y¡y2ó‚¸N¦L/õÞŸêwc"µHq¿iœ\UèàwÚô‡x(èúºâ›†[7ï´Ï; K­õUʱÿׯ œ¸¥–¦UwQz Íë’d*!ÕNMâÊEGlÜVŒr9ÌÁigڼܞáYµ‹Q'ç´7`o‡'L­Öõ^]ðÍ«À"Å`éÇ-€-¡ùÝÒª}˜I´ztž6¬t®¶ã\áb*0zÅ Ä-þ…‘d!-Ù#(An{µ>å*¹èj}Lµ—5¯ã¬wyÈdX“ãòê·JW¤£˜õ©¢[Ê;1Ò¾ú82ñkŠÃ—ê#g­ªJ°wu,ž3qùlUgF‘ŠvØ«³ª*“/‰/@?Ô¾£¸âæOìÍCé™@…y‰ðÃDJ*J®ÕÔ’q—N0(a‡QÓ= UW“5Ð¥–N‘‚¢YÒþ€ç‰Â&ú”Ò‹¡3]Ãi7Wáu5ÍšÍ1 B· ӽχOðß\d²NÐõÙ-c§K½—’ðã£'àSùóýk#¬ ¸¯Ÿ÷ •¡L2~¶nÕAêy% úD©©2ö•JàÓ}ÞbÛË6:¡sÎíÏÙèžFê”®K±]!lÖqR J€ç¤/Ÿ'd[aR i¢åú n¡Ž]Ùj~¤Ž^ô»ÅòÜR&ïÐDyæ¦Êëpám0UÊ‘ÈòÁÊ{­€ÅßÎÅŸc’¬ØœU—O‡QÂ<¹`B*ˆ1]"–k%$¥¢±G¸a±C ÛFuÀúû–YYŽ¦6^ž’u&‡_*͵J®—%·ù:-'–¢7YjðýèÝüVñÑõ}¡¥w Å]‰(?ÔB?ºÿR•“ï_“¿H-ÕÙJ~ö¼Kƒ†BW™Û» ;Uq<•'ùšS ="û…°s¬…R^¿{:ý-2&“½u¡ðÍ\s‚ЉÌÇ“÷L,¯Ø Y¡7¨üɘ‘UZÑq»dÓÞYÏ°öÂl½[W´Þnca6à ”®|/&#J9§ç«%M^)L°é¬S–j/Ò÷]è ¸h$ líÐ÷^<¶1”v}tïýöUL>^)BxŸÄf] °ÎoÀ+³Û!9J™—Úï~O|0»ä¥ž@^‰xÕ+gÐ> Mkt> –1©à؈YÛ%/”=¾46]F“VÔ«ÙD%Ðô]²d ÚR¦DøÄÇaÉPP™Bå-Ù×? pêÔíi.(C¬µ£V§·uÛ”öîü¡T€¥ˆÐKAAÁ`ñZ嚶øÝÊ_–sm{‘‚°½ÙCÑQ£‘ÿ>uãk„YHw®Yø¤ˆ. úu]É!p-tá‘73ÿ]_c9l]#6®ùœÉ¢6}ÅG?Ó§žE¡7ËJ$ÿ(Ë£i÷CPcÌ­:®8#«°gTÂåR+ò#XM19íÏ2nb»ÖLÙªÃ:– ]ãþce¿¤0:À1aêã#T‚LÁ’zù-½â ¶Šî/J5H`jK;2‚†BcøMéÎcË(Ó,-‡Ä8«öÞ˜'Ã[º½>¸b°~§Ã¯äCò­xÍÀz’DHê8­ê´ (N 'E¥³aB‡`$›7äÞM3ªÇéâðŸTmy)N‘$E×ÆÖQûÜŒ×÷ÔçWµÁ¿b‡\zNs…)Œ­-èÊóIÕi*íîs•‡Zr‚ÅÇÝÙêšTEÛBË[ðŽÉwêЈ Î*CPtsþ°>S˜oÆ¿ßuëý1k&"MúȈ½ÇÄN¯"™7,[ÈÈ.¸˜ªžQŸ²r÷ñfNOdÐÝ&] ˆ²•syÒºêŒt¯ÚêZU:ýYy¯Ä*;bÍhØEGHú!Y„aNt$n)ºØë†GS€åkØ»<1ؾD•Ö'4»ñçUK&¥1_«M,Œ“ì¹$ó-­b(üݱ¶­’ m•Ñê£×»ø¤™q‚¥L^hü©f·>}ªÉŒ}Ë&㌢›\³ÆøFD±Í«k”D"É$!š"Ö†ë|(µÐo<ñ¾&²‹ª¹Qß/kѺ³5xâ05•1zÑ/Öˆ…};„ ÍÆ~"Sœ€¾ÜŠü ìóÏUŒÊoÛÏg7zc§7îá—íÙ¯Oý„° 1êšîæo7r¤®2.Ò_¾Îl…ï~‘= ·FèQ4Šöö…ôT[&Þï[Cb.;iG-–1“¼SºÜ©ž¿»¹å¨Œ‘xáéħwÞh0"·ÑõR·±7ਰ®=TøÓ›)~’älª$Í&c6ÜmZq–#£N†5ÆrÉ V”µZo¬Ã¯ó ›ªÑÒ"êÉ«Œã¾ðʲ’dÁ¸©Áû<¦c%’‰ëI7VT–<¼/ç³ùBÔû /HÞ”F}¯…†Z¿›K ¨xI?àUËM«2AžœhâjX3½ºÛÈ…B§š½F2o=JfñcÂd ÝQ;vØ‘•£¢ûô* ô6Ñ3ÁDPŸ0/;ì•Ë=÷ø‹²£'="»á1!×<Þ^þ¨/Yþë-ýYø z5}ÿý-^)hØ«ôšÂbVƲù †i?Q™i|çˆïJ¡®Ë£ÏùÃóÞ3ÎC÷ä1’$è¸3 ®šàE8Á¿î†’.DïœÑÄMþ Ë#e‘) ÐüñÐU[ $8Ò[ôQ†½­F±’–7¹*¯jš=BÈ-Þ^çÜíîË¥’¹åäò}ê·oº~ÂK6øº}ÂX>s³pt²„W¯ÿìÁ6 6ð2üø]£”{•ù¡÷¨å«=ÈD5Twèwø‚:.fw8Òluq¬Ãtñi7lù¤¾B•7mÌ”NÅ¡ËÝ›:°…#x~“w(U2ŒØû)°U¨‚¾ÕeÖ:AžÆ1)‰T.WJõѯ!«!¡þ¹ºV¡DvåÄôA\Ãë5íÿýä¥<½1¼\¸OS~“Ñ(/y²¬t­þž4’ÍÂâ*^ìø_OIZÛûâ™à¤(Ø-(0& A'û'Ëç|‰ý^¡“ç± Øű[yìmEŽ´zZ-Ë®þDV•ÔK­OÓ«.!!yÇã$Ws¤‘ô†ê•À/Š‡Øuu®òÕÏÁÙMÿ(KÊâgt«ë*\ê› }¼vÏf´5!À§ ‚¿39µÚ›òÍÙŠ™u‡x •nð¾QHû,_fŒ.¾ýìú¡À£qxêNŸb—:'éT½tª¶)¦èCHáü˜çeF!@j?“Bw.ºYð‡t "Ù¾IÕ]?{àä¯.¼ }z’ÞÉ£¦TRôçþómzê‹ÚIѹÂgœiŒÛn•\ÉHõPùY™åj_ZY93ßÛZÎñIVøGk©ëÜ;wþŽÃ·£mûNfV¬"ZaÿñšHjhç¡‹i’Åw?¦r³á/éñ©ÊÕ»g(wp}‰{1EÛWÒ†}rrd wŠ 9Õt5Fg3ùP=ÚA/Šk»¿Ð¢V`râçmH¤üåŠã4 ÛÐ÷c )è…B÷ ²´æãCQýS‘qát¿À §¾)1 |lÊÌgzèÝñÉYîÜ3K£ûco…‚]Ⱦ|uöb`p‘Äý*=›nU¢ÆFõDg[ù¡vQð“eynEwÕ—ÔP$é÷/Ϫƒ}£ÆØ°– ;|”n$ÔÄ¡ßúù,^L´A¶ì˜Æ$ßò ­©¦öõˆwÌhÔu™IeæðPSžG",4lneæÅJ§íÆ û)ðͲ_ “p¾FY{USòsYÖv†["ÍH‚œKe–JŸ¢™¾éu(:ÒØ®[ C²×]‘2ââÒuDyš»5ô4‰Ž÷Y×WL›"s<ŸÀˆ¥üÝÂ=K#’d÷·»{á~FÒ ç/ó¬x„2ý(2™2P#v 7îÚ z"ÐwÒõ¸æúÅu#ozÕÞî·‚¾ŠÊ;XP_ÄÕê‰jlyô5¡ÌT-]šæ1=ˈ¼íIÞ@Àm%¬òa†ô¦‹sÔü] #]ˆ*û$nâ« v5œ’ýk7·¦çÁÔ×±Ãy>Ò¯É0Pº2ÚêŠW¨~ÜMÐ:Q9ô>¨é”Ú2–`Q¬Óø?¿Ž^½[u¢.›J†¸NÁöÿLSZ_%5M\+wNÐzž~Obm"\¦'rùŸ?ûª?j¿„”uo‚§?‘à+jS(’4ËqÄ5>üÌçðÍìIT7Yv·7¾ºp[î&$†Kå¿©¨¨ZÓîÂr¹üô!+}qj˜É ¾M„¥óîÀA‡Á]9_õ’égûÔÞšÿŽ@¢tlåHío”:Û¬Š/™F‚ãLN¸Qã÷Quw±RÊD¥PkbuÇí Îˆo'ͪsÝŽ”z²-_×ݸtWVXTÎèq_Š&éꪧ`×|wTB…DÊ*°xÎwb.fg¾% S*#E]c!yÿ–ª‘:k?Ax, Õ˜?W‚|È’éQÏ9Žú… { (±hljóN~J|vê,ÛÍÐ TÁJ>Ðeå¿Tþ‹§Ø¦2’aQÇmÔ‹Ãn…3ßö¬”Y¼¢üªÈýGÙxµ ]²…b¤'@Ðà8!U1W¿±Ú-˜ÃŠ¸]?úØmnsó„a×2át ae‘Ë&ƒ ßØéažO‰rF[o/Uf£ö`ª¯ÁíLÒ ¾©˜ß$“ Ö¹Ž3UçqSÚ#t2¥T[¸ëh±c#°è'¶ª'*Ô‹m<}åäao%C\uÛþ±Yª¢&{t” ê?Š0‚+¸b4áFÅê Ñ´‘ÿF …=Ec7û'S?dMЮ®tòyH )¾ð~­1)ô,"™>ˆv5ê.›¸¯ïRØ! -×ËhÑ4³Y¯ Òqm4QN}Â&šñ<[gBL:™üÅYÈ'7[þ“]¹¬ÌRå…Urš0å‹öÁ£ZU”ýÐ]ïA uDqR ÜI1ä4RÑȨŽ…åa¨eÝ:[½Æ:¹S™¨õ0½ˆíÉU:Ñ:åHê¹0z%µIól-8…;ù»™~‡øqü2i ¨ò&0;Ž}íÔ*âIÅ~$™tšìTL¿FF‹Ï!.B8ærç´ªÛÙ;cz "Sâ"¡wø @¥öœõ›¸Ìƒ'ù5kžŸ'‚µ¡â˜ìÍSùG$¨Ñq•têNqOíÆc’ºÑ03e°Fƒ*ª,/Þt¿`ƘŒR´¶!¿ÓÔï,”åÇœäNíMæ×’¥‚Ròýç︤+еZG‡o×!~•±y“1 °>÷KE2”vÄ¢– jôêŽ`­aôÕ kW«²Õ©+ªb>¶#¤U„6íOלö(§17"¦Pw»êÁ‘ŽA’´¯EØüÖ¤l®Þ¼}mÜš#]ovFwútú1i×iî0¡PJ²ž}}GQí¨ó0Âú“±MPÚµ^÷·Ñ Dg¼®ÑÞµ7P¯zpwëÝ…ìΛŠ½œå.€žâŽy<óP(œÔF?vŠ¯÷Þ¸5_qSo•\”¹S…}ÞnG7#U§´øçó(࣠Æu¿oIÐƲãs„\/ Y¨Õ3dÔû®â7ŸbcŽXPs ¤=nŸL^Ó˜ñ‰²_¼’ı¥2Æ‘ü’ uÔÓøïÿ˜*n™ endstream endobj 271 0 obj << /Length1 727 /Length2 17546 /Length3 0 /Length 18146 /Filter /FlateDecode >> stream xÚl¸spfo-Ûè8ycÛ6:¶¼±­Ží¤cÛVǶm;éØÎ÷›™;sëV}uþÙXOíµ×³ëÔÙ‡œXÂÞÎEÕÓÈLÏÌÀÄP‘Ñdfab01°Â‘“‹:\,ííÄŒ\€<  )@è`füa†#ˆÚ;x:Yš[¸¨L¨ÿ¨Û™ZÚZ:ÔímìÝ,M,|nnnBnή N® ÿR.@€™¥  ª ¨õS^@%)¯ÚŒlŠ®Æ6–&YK 3`fï°ù0±·3µü'g†°s:¹üCÌÌÉÞ '®*,¡ ¯ eTÙ™d¥þ9nçâÌóhbìñ¯îè.ÿ³lŒþgÿÏú_ÖÉã?33ÀÔÒÄ` 4·´ƒcü—n?íÌìÿ ›º:ü7õ!çø¨þÑ` 4ûíjc#od P‰ÚÛ:¸ºrö¦@';€ˆ½)@ÜÃhg 4ý/ÚÈÖÒÆóÿÿ?]ŒþÑCØÎü™þ²t–°ôš*Zºü#¹™‘Í?Rý;®üÏýü«ÐNÿS ðϽšÚÛÙxþŸ‚ÿ4óïrŒJ¢J?ÅÒþo þ·3±7µ´3¨¸ü£¨‘“éÿÿN+Yþg†þKå¿3`þ¿¾œ‘‹“¥@‡‰á_CÂô¯ç¿–ÞÿE‰ˆØ{xÓ³°°èY™9ÌÜlnö_ÿ]W'' Ë¿eÿ§—ÿúÿ' Ðh·¶loÂl•ÚZî+^8[I;;m¨F»Ì º2Õ}­ƒj9hajpx‰{§Š›çŽ@0`Ì¿{±óó®™Ä(ˆp<1M®·ŒûRLZaî[³õ¦”ÿZ®¡Žck"qqë’¾¾¨H3É–™s^#íXÊ*ÉmXÖÞ‰²¿ºð`'I­G‰–¡ý…ܧ­F™µ)¸2|Ù[ Ë ÚE¸&Ç1°((p8 ÓÎÛ0jÿq¾ÁLk 2bɪÕÞ–5U"P)ú¾#»»¶™ §FÁNœ2·q¹ußZ“§F¬‹Ks{U2¯/pF¦¥.ó#ç4AW–Ç}w™t[Æ#Ÿí"Ýy4L¹ÝÄ HMbY“Ì)9f¨“j™¯ï˜ãÏ!Ï+1úq Š6÷øÖ¡fµÊ&D`¬¨œËƒ×^jždÚªñàb£¹+ºYûN¼ËÅŒRŽúƒ€÷£ÄbµqHWzVøÆR®ò³6²º¾Ø«”×4Vñ9<<Ù¦Ÿzðžá‡ŠÁÐd›Ä+(ýiŠ³ÂVRLal‹@ç—‚™ °à¿«ƒMèìŸ\(ù&à/rÄ©:8rÙ•áØÃý(°ŒQC¡)^î UŽb;”ä#·„Ïë´·Ë%L#üJdâ¥É©ˆv•CÜ_« ‹ì‘ Ò£æõÌ¿†¦è´ýÌZ&µ¬×Çov3Ñ }¡›KçÓƒ°ÈëÛ¶ ÿÀÔéfW 9€’ÝK¿IݯÚõæ*ÇJM•Íf2é ·ÎáÅæ<ØZd¹iŒ”«wýÁzDæVÇb º¾Â\í´|“÷.ßÿªeÌuRâ©Â¯@p”,DºïöÐi‘wéÀ-ø'>‡Î1'šyÆäÃbÝœ‘•(¥™Ûw9ÚXÉ%†GWÉѬ5õWìñm«¤d•òML#ñY@)'`BRƒdÓ#õà\Í«š–ï%¼ JGî0k! ÷uWÇØ^sa;«óeMWðÊ•a±S˜ü⸷5þOvÙ°/N·Mç¡œéAß^z ù~G]Ƙ® ìQ]ŒûêüMs‡¢Hsg‚ "zl$¶ÉT%ˆq?ÑÐy7Gó½õ—ßû¦Q1‡$áoΩ³$>6ÅÛµ:ÐÄfÒiX ÐP©hq³ÅF\Û’HnßAH6ÇB­²ç÷:ÃX)s¦Ø u×+Ÿ½ëÂþˆo åÌOãÙKÉjïö4‘¼„s71…‘ ëê_×ã•*øÁ¹…äáB7d¡QçjP?º€t¦¤VI•­E׫åߣPÉþHû’‡><º®aæXyw%äJ¬ðu“1 æ&‚€ªº-0¬è`' öDÝþJÕè¢U„;ŒÕy|ÐT^=黢 ¾é‹&¦Ÿî[š{^!5Ë"ŸJª2:¬S3ŒV“ièû:Ùô”ù½r´q’¶,ÖŒd‹Ä>ö4ü£gpÄU×Å—$SvT3\ nâÈ_ “ëøŠ9ö7@}ÿèe(}G®|é6<¾LtüñRóIa,~Ïý"QÕê›iŠ“N%.~ŠþŠÙMåµ0aƒ0<ìêíH œ'h–Ñ=™õKM‹¢à® Ÿ¨ÃÅÚ—T…ûì2Ì2á2LÕì*íjŠ^©¸i‹þ™H·°àaF»Pc„Í9Š9åZLFÿc,=•y?mÀžÞqñ¡öE¢$Å8h¥»F]Ý8¥Ä!WàäVƒÊëš|v_¢ ¦ö‡xIœ¡V‹4åÓß<¤vL½Wö#M±¯§sÐ д‘2Ѩ€Îµ¾É¥VÒ)+±€›ž²þs¼×#?Îú^ìãòGYÑÅëhÖ‘ÜŒéuï~c¡GOê·åßp ŽùvyÅÅRªRåîh™~0´WRú‡V\3ÛÇ»BQF… ߆µq›•X‚&…³ÏÖøìëµÓâT¯è(îJ2ø›5Z-½’äN< zÈêÓyü!ùFu2þ3éïY.¬7=¢<¡’:ç)*C’HJAÊ®ÿµÂsS€Dùèm•u{4B¦eYÔÛ›ÙhpôSñû%œ´*Ñ~†ï"u…ºrWò]\òsüw>mš­•¼sº×Ú~º¡Dˆ”© TÎêüþÄO__ä¦|ƒúj­½ÐÞ.ÍGQÞ»Œo\±3^²„ýoTR¼ÍY   •6£K9ÆÛtT µ¾X%EÐ’åÈ5*tý ÄØuÓ,•  æÇ„Dò÷ÅNsá–4HçÜ©qÎçNôÚ‡Ýz¶}L2G@Ε¤^ií ÇÄÃ3*u×x›Ðf6ñIŠÆ#Ð =79IÉôïÄJ·fÌqIÿ wÈè†újf“lKÁ•I»¤‘0ŠyðdɽߵSâÕpª|Þå °â¼qL„6æW…ä›s”2†ÙñØ¿¹ÿÄÒÄ5p†m.úÆr<þ‰îê“RF÷¬¥šâ$ñÒæ³1ЃÕýj\³XTŽhFÓ­qb;éC\™ÓE(ß< Kô¦9Ô£K3™7( _tÎåôœƒÝP×Y]Ó^BxZÞê¢*&&úsü›yB«C~µRÉ‹ˆcç¯Ú5k±iù+—–×ð+ÔJÊŒQhÀxY¶i‹¾L ?/(ÓÃò.»Ai­Ç9ºmQÄ5©–©[ׄDæžž–Wô]?áÞ¹$;{…9Ïž:âÎMliJø/a Á¥QýÚ/Öä T?,Y³ÒûüùrÂ]ƒ”]©›È†ËQ »X5¹I„·âQ ä²>ÕÑÈô€zŒ¦;Öå+”nš”.a@I_áú­jÛuŽH\²ín'2 |?®_ÙõbûoGÛ£»ø±k0CRÐ…CZŸðpTÐ=K&Z€­ ½Õ^˜hˆäOU/’VÞŠðAp#=†öó÷Y=5DU˜ÊïÄs–è_¢×`•¬p†Ú~¯ÃŒºÕôºà£ÐšÏÁXÌ${1z_#Cu}[jW¶>Q;õ5ÏSþ°A¤g€ÿ¶Ù®välV ==ópa(Á7@™ èÚ]ÑAø…§¾)"{ï<°kÖ¼ÔàÇs(ú[ uÖKÆàþÓÓ’ >C­¾È­kîñ²ä+ñ"ëQˆ+ß”õ»…s#&{ÜÇdÛ{cUÊžx ežÍ¥tU£ø6ñÎÔúùñ}[õ•^¬ñe2²ö,Šë§ôŠ25q_?D‰Ê)×-xp+r€®ò"ý½3¸…[äÕ‹mS+³Ÿâ½(ÿM í×–XƒÞ€°‰‰ö!º`@â^Šrž¤Qú¥ð3áÌA®ãÎE‰”*,ê²íî{ó?Ùi~Å÷ Vª¥‰~žÔDV™qðR=„Ÿ 9;r„{'X;ÓçIVy³‘#†¯÷G'ß (*#ÙHO+û,ÌV`à8,+>žÿu×·øSGSúç]}·â$´lZ_ÄÛûs„›þø*éã×€Ò{öƳ»ìƒæ·G¥ó|îÒÄþa@jñ˜ý"%´öòú?Rh1?ƒÐ¬º‹º˜ 'Ôǫ̂ò/óŽA:ÄIÍâji$”ÍûO ºÊÔW ¿²“[_F›“C£iŸ¯>Ä>ZäÛ5>ñ_d=/ ƒKàà¾#¼qƒj2|ù½W2g–ÕJ£ƒÌ/ ‚ÅÇÔòƒÄFòÚkØîŒ\ šLá~>q8+ÅËRûž9 :@©4ÔŽû}ŠOÇIEs&ÿ ÛÇTöT¾|ûoiO!Ýò»z2CÎ+xÏç ·v"K×±%[w1˜àGAŸ¦˜¶˜ÕØ8"ŸÒs Iœ,9dǘòp2~²Y&†GáãyX‘ðºQ%ýVsåq~'3 »Õž´›ßÌuÇI ±ÆÐØ;¢¶’E ß ¶ò°²‚Q¸”{bÂ˳QÐòR«c®¢5ø8¾R½v óãänZ@!¬®«¾ütJÃìáoI~’ëCUµjž:ýd„³ô@8ÌÊ©¬'ûk¶~×ÕyîoŸcE=¼äe+Ìêh[š9Cý­ê—•%¶„Ië²B–BšüÂr‹¸Ú2¢GÔ9ÚJ¥•ëoõ]Ú“®€òšD3Æd^ý/ c°#3³l·ý€8?·‹`%Çö5ÂIö?SŸ5ÔΉ沽'HMÑCðLÚùÆEE:íf)ê+ÞR×ÊcÒAPKËHQSwd`#R÷óš,I²— ¯/ ¥Ù¥¿céöß»ÕÈ&¦óé€;0KI®c:A •Ê ܪ«õäy¬&\tÖéì:¡ —c=ðGÎßï×±ÇÔT­Í°zCG¯¿Ít†ù¢¢B_Û÷ËAŠ‚Ä A:ÿ³ró÷ÓÆqÒ‘.ý€cMщÿ[rØz×4›>¿ÊºtË¡À8^(ò5ÅKR:¼ërŒC™µsr2‡Lȃò¿»æ+½ùëÚ^_ÈcUeí7xèçã<%"—ÁBÇ,¤?Ýw…™õfƒSŠ™BýxzÌ.±-ˆ>Ûsw„ \¬dJÀ‘_)gÇ< lQo®%|d/*’B¯¤Kæ  ’h«üF6Ø'Êø@“÷šOŠ,'íE#ÿÚ¤á×$X-cò!S6°±ðzÛn$ý¯ÂD°zøB{Hßæ>ªe}´ò>ÆøýÎ|µ¯6v7qèqâ½´ÀæB¨µxˆÉyèT#m–-–$ž+µ÷rÇ7ÄH§‹ä¹dx—cŒB­|påÚÿ€ûM)QÖ?Ï|ÂBž;'«ß†ã’bGpsˆ9V§KÑò}‚xKÇÌÓ;¼ Ȳàï{ƒuu³ÔUN[V…?@Í?ÜÍ3)'Ë:Ù~ºúÇxwÈ›»Ì…2sF)†ê‡RšÜà=Zx”’Õô:[¶fZ<XqdÅÌ®Œ g“ª›ö,’Ïù…öM“Ty}G™:†CóÉý­güfj‹0øG½†i"ÐY…p†‰A«çÁÓÿÚS|2Š=ÓVOÈôT# Ÿ·ýÒp 0£d;;5’¶öˆbݲ¾aɳ¥rºê;œ¹¢ÒIH1ÂEõ9–Ú=(FÓH"òpPôh %Äî²_Š¦¹»0Ï º! ‘nE¼©ˆ~*7Ÿ¦u1!»ˆÿ¼G#?†.…qÝ´Xo¶{¹7ãß ž+-IßtdÚ÷k›îOÐë_@üÂQÔâºÿ:QT¿w|@8>ŸWñFZ QÇ© ¯éOο¥¸VsÒB¶°/–2Ö5ù0Ù„o+ÝI†3 P¤‰ÿï o39õ¥í¯àÛbz¯æAjá_#(Èp–Ô/±¸H§ oHÒÀî<â|¹ÉÉ-ü]¨Y€‰¨´y®`Í÷ÄãYè;úX•ŒèèíÍøSú5Ô ?Æ:ÿ.uŠã£‹ºÿȲ§™5ýcArÄLÒãêWWrîÍI™©– Ϲ&S5ì4Ð+gôn wàýŒ¿ÅE˜òÍÌwNäèUÈŸ.‘8Ÿ xšâB‚Ycªu_±e…~s¥TyœJÛW\ÏNÍÌÀ1úÚMaê ¶RÈÿR5°õñnX:írJ!ã<³]îdìó £>pñC¬–yUé]JE'‘ùx¾ð©Àâ§c¶ eïOÿ&5¼L­“âî&¡•qV*'A§ÓZ;ç#h¤?²·8Lס”= Ç ½V?_žÐgÁ'ʆ„Žt ŠOE,Nÿꦣ”z7ÓÑØBÿ óf§×¥Þº®ŠàÿîùÚÂÔ5*‚Ý”ó áÌ'~=rl”@9œó‘;5¨·é9–SrÑ©;5zŽr‡”¢•Õ¶TŽÑŠ–ŸüÅH ³©ý”ŠÚs’…{#d&d£,ɵv?=3ÖÞaz=œàúà䟿½-D«ëwÛ«lŸ$Âx¼¸Ó%ãU°)ʧÝâ­˜ø¡ƒ‚Тpïuòþâªë;,­‹®BˆIˆ’ …!~ÕãÐÄ€-Õ¾®H¸÷Ü …ˆ¼BF¹Ñ­÷áz‡Xªã¸GMÕÏŽ…âÂ!?¾JÓ½@‰ôºpóÄÁ¼?p–â¹Ö~åfF… gò÷8 þäYÁØÖæ¢ñœ¡tÂ:µ;¾ùP“èì²H¦w ¼ÍHБ¨k"ãý.¢ø²Å=P­Š¥a5gEa¤•bf ~‚^©…IRÐùýñÙ9§ã‰–Šc/I„¤oÅ*´¸_”ÄÕo¨îÖo*q™ý‹ô'(i"$ß²)€êÏÅ[ á6"é%OòÖóvΣ[꺵PÍÕ<CiÿG,“T4j".¦Á„Ø“¾jÛŽÃÙá¾ÅHÉ…â4‚ÁØȃ^t«Û5õ6ìWŠÌßè¹óßA¼/¾¨šŸÝº/}’>FáëÏˤç é,¾vní½IódªZî“Ä î? æà]æ#‹Æ OÛñ¶¨ß‡i‰9 VÌìôgþ›Tf77•¿4óg%>Å´…r"n_ ö€.9v÷’$ˆ:æ€] ž[!Š2=±B·a@t¯Qtš<©ü¯ƒqÀs©ƒP¼´Zy©»CævÔR,*qepOåLxÉnFwôƒ3Ï.©{¿Vu´Zpô@h…§ó=¡ºQH¹’ÛFõ±,I÷ÿßZêYôx¾b“ˆ›;CTÏðöŒ‹@‹Z ›–Ôá{3^ÊaÀìv(Võ¯x‚ýñ©ÁF /‹½§¬ªZXÑržßFû û3™:Dƒë¬~dL¯°¸_Õî;½o}­¾QŸCæŒ4uË‚ÎOí{¹‚fw£«Pt¼ÆyF *Ÿºôãè3ò{1çqˆ£{ß' í{:ÊgÃî©P‰è%½é—NÁ–Ü̼Š„ÖúNêÂÂ{£Óm>Í}XlæMÓLÁ»Š¡¢ 1éÏ%\¯èëV0yƒHTýˆ³g~.{=Ã>ì¡>)h;FÁB‰7í®`­ ýJ«ûeº¿4Êêeèò£"êöNÙÜÓ û!}¡õ‘ÛÈëh÷ã)næw±¨ÆÖu[´¼ÖZ )¶ÃÛcÑ.+:§[…5jj6ôœ ¯Á=<_—ÕOß–jB´ô&%²Éþœ[.3OlìÐ|xYÐÊ«ú Ôß‚º?×d±CÕ6¸¬nXvþYnOœ<òX4ñ]¶ÒÉrÆÚÎ`~oçõ=αh)¿4•ú}¨ï\íhrÜ^T>ɬÝÆZÉ@¼üPŒº£×±Å÷±OÌP{§X‰]nžý&çwÄ\³sG2ÁýÈu¢]Âá\ÐOM´`fÏWìó«©´vZ«hêïÞáhqÔxº duB°z8õow“™ÜÝRGyùºº•Öòr}pw÷ì qÈ8–Ý! _4Z‹v)?Rü…D‘¹¦Ÿj/”¾Ö«“¹=‚½©Õ†ö'îç’Hšäêm%›³Z¥R»S³½/HV‡ûJ·¯é“ì­®JJŽj8¬V¿ä“û›"Ìö¹ Z—³Î¬¾0CÑ3®]„º #ƒ08"‰¼ØÌQ/‚šÈÿÝÈÁ‘`7f#šbúP¬¾ulîuœ.ùk© JLh¨èäo¨¿Aõ”†Ò8^T3X§Íqõw’ØÚ$ù›ÍM\5Ùç!V’܇èÀò3mÆ" ë€]Á ïFŠçÆCND¼0û¹se–Œ”]ïOUßl'eå²!ÛŸÍ6£ÙÑD^Ÿ¬hŠ®Ô|åg‚Áž=Æ,Ờ®I‚Kgo J¾¶ N©ÅÕÝ/ßÅ+Ø x%û#©o´^mv@5&´Â.}-‰aX{¶¤ì‡@·€IièîÊ(1Fó“ÙÀr¡£¼\¨ÞkW$ókõ`³1päçÓ!¾8˜ÇE+Å>Š Ágõn“q¼³ånûózCíÈúð ŽÄ ð,ˆZ·]þÓ·Ñ<{'{%9ˆpS{ÖÌTë©ÄO‹Š 2º,îÕ×!ÌÊÉdÚ{g%ŽˆSIOù‚(?r‡”¸íoK:Þ(Vã÷Uør/ÏiRv“2Üíyã @}áxS}ð{Ùøe¬ñÝóH»×`LIMó*÷ažÅʆz4GŠôÉpd‚V=3# cýzÏw+4´øPæ®r7xîˆ,)ZHÿàg¥²o;´d`qoVì¥{ä´öeîHÖlÝ’MÔ¤±üÕCÜ÷×á-oÆÁYp àþ.«âÕÕšG´ÙFÝ–ámÛF“™ZXß/•ˆs–fm dó»°¿2xWœ;`×a g#fCî¡°¡½@Ê-CNe„'µm(•zAÜC««Ëj•‡¾Ø~²ohð³ò[²iy#ÁM=¨&NÔÙ)Ŕ胱D“jÀÏyY$¦˜¿W¹Ú‚WçEæ«ùå–€iÓä3·ȸÐH…ªfì0™Ôö6ï©“Ù±SH5´^h¤ô¿ä+ËÆŽßÒ…:‹TÐ;nË4ä3m)H?éoo|ÅvçÏ‚ÓÆ×µ8Ñä/ÏÂ#òß7+2ï_ìYÂV¿XCwcœ»tj‘äWd¬¼åíad§Jœ‹%Ï *‚b«5ÕhäSĦÐCùÈÛ/„±ŠÏ¿ŠïÇf¡l\–š@7€vÛ?îxâÈû}-ñ>ŸöCDÔ—ù5¹> ÐwüÄY$ÞfOêv&M&ÛŽRÙ#†‘¹¬.,Ýr*²‰Þñ©rw3¿­‚ 0ŸÆdöû„‚F§5+ƒÒš;Å¥³'hSá‘ÿñµ-#˜m­‚“RÊÙiËù†AŸE‰B}@eÀ'œ¨P‰«’eÝ(e_'-±ªU儽°e<®^n [×úIÿ#!&±÷¡à’™„årVŠ·wm—þN´‘½¹™‡Ô’V$(ºLaÿpÍ÷ÈÆ µSN¨5O4ô°jU‹ÍXpÙ¹dTT`¥§¤Z¬ÖFZgN’²È;õƒºÆËaÔ|ÁµÖø†ÊÕßìÃüK,ÌÁþ^_žJy(®§[<:ÉÕ$¿”(‡™&×2èL˜Ö×^Æh"ù»D7R9øº99mDbîûp!Ñþ|}dü8 <¯aJN³¯ˆà¶Ýïp×ÊÈn¥2m[—Ϻ߇‡ÓÀM‚þ rc…ZÌ?ögÓ³8·):AÊN²›!á>{mÑ­Úž®â!ØÔÛßâ'Ôù-Z.#8Öxå0¸'ãM¢õ_# òD(é‰U/æ…@òX·cý‘ðTÛì»…] a[»T.b^y™«ó5P©J¾Ïïž·ñ{Bç•6\­Çg”ã[ü¦ GÊćâòáÝâŠ?ˆÐJU0j áŽÿ¤BPvG,†T&cõüugôUAºp|½¸+~úì¡~3û+X:úEêú ù¿†Æ"è=°#YfëÏ7”dz·vX´—¼éan³¤ Ì"ç5.®>xrd‹¾JtvL¯«Ç>_Æí1©¸3ÆÇ º)7`h> 33^Š;1Y}¿ßöjaöµ4Œ{ ¬¬»SÏ°:6¯IÒkÂÄ Ýª2à˜c*ý,[´÷9ä»û Âì`º2¯-ºâóÆLù'óU”Œ)¯šh¹–Áypœhqr#\žšß“â”ÿüA°¡n%¡E7 õö,ý÷ž‘êJá² _j6Öó]Í¿fâ±÷â,’g’\ùSÌÕÒe»^¿Œv^ÒHÐ`í&qênŽØO¥¼îì75Ì`ÂH©!ý£)á/D­üÊhúYR/e”†”Äö’h—T‰fTÿ´ûë©zIÎy~Q–J˜Ö¨UJä5¥¦sœ‹Xù–LéÍ<©cŒgƒ™IÄm¥ õòÕÀÁ,{W)¥uÜ2KQ´ ‚id-ûÅ…Q[(iíÊãêxúAH~`ú_8úÍJ ìH³P‡r¨sŠ_'„8ß öЬúBYŸÑOë¶ðV¦WÞOàJU?Ëö6£?¤‹þú:©Æ—´¬Muà!0%4ú)4î·ëÔù€øþ‰­¯þ{Ó1B.Ï|«Ýü Š‡Ö.Žø2ƒ¼ÿÿ”s}ÑögÅß[2¼ÔËw/5ÈþÇ=¨Ew-–$y™_$B7 •n©Ã†Â/Õ;+‚¨Ì5Ȟ̚:YÍÛ³?#Úoöy—±IO}òBÞ§¦ÔÉp5°IïW:KÀ¦ÌÕnŠ{æ«ÔÊ) @PY€j×þÒ–èc:è‰×) sm²þ,ÀðÅq¹ö}gó1ÿŸeMV¿äa€%À^&È_bZá(Ô±zQ½YÙ×êœ;ª¦.ÏÖ(Ú.jU6ƉJž‰½‚¤”¬]ûÙ¥«ý¸¿JÖ t¦|Çx±•- ›·Y0ÛŸ2[°rá_95hãBÓ2è`~,ã§à¿?0Dê†j®ÿ†páiÖ<ôA7ò¡ X dbSR|þ¥°Èú»'*¿6²¥"ªÀqO¼¾ìQþ6 ßå] ‰Ž)²Ó_{îúµE}þCÂX±–  ƒ\„1×Ö ßÎlþi´ûFð‡„Q„°;¬õ†rô©àTr© ³”¶±Iò¹Z˜ýi xý ÖJ|XñQ‰É­Y'øZ½­‰ÃðƉT?YéÆ&ÄF9ÎÈßýýLBË¿Àö°8á@Ê­2ZÙ¯f¿BrÈ|¾Wƒö‰~â؉h¾ïâåæCK3òr­'ËLJnÞ¤oÌY”k",”ep1(– ž‹:­h‚eþIõ§ž‡ý>G ÿ€)bsò /ÍÅŠÖèà°¨³ÅÉšo‡V2 ¡ô Pš½†”9õ„‚4Ó© í3˜/¢è«ë9Š™zM“…½ª¶\Ô#ˆšYgßùh%†üAì‘n½¶Rïä–ŸÂV+mm§¤åYŸ\}vÄL#»Zy.È`åkÛ¼ ÷£ $ÕaMâAÎ÷Þ,ˆ&rëë3ò’E˜ˆ¢)¶×£Bb¾H/Íê—WÐ6¨­&}Ÿ$–’êT ,œ7Õ÷TÉÈP ‰(sñ9‘Ùù&Óí_P½C󉶂¸œ- Ö‡ÝfÒŨm"=9P)9ÖCÇó?-\Jy?^y›´™³ŒÔ3Äšæ'û¡¡)¥/š¤*¨µt›Dh¥äa  ŽióŒd ¤ke©þ4ë G)lHëW›Øô3#z‚×øc)¸ùòŸ[Ɯ֮ú±&!ñÔ³fgÀª1È•ËE\š=6Ä}wá!µÒÛJÚ]ÊPú\¸êQbÇñE½ÿu¥ªÁŸÎf8AæÓ5BÖlqþjç,zƒ1ò¨¦¾Ë—çgøw™‚ÿ).ÛŠB‘‰Õ+fÈí!¸cŠj©Ym\Ì ù¤]&×æËÞ‚_…rWøâÜBà°&ÜW&îÒ EçV³3Â%ßL6zd)û–-23_”Ž)á§ÆA å&†6&è!èðÄT—nQxb Ò§»”Ò¤Iˆÿ; <’pá-s£¤EEÑ~é-8à…6à•¥dœ„ Ww€ÍMŒã̱Y&·j”¨ƒª°ú%¤Q†=Cª`°ÄF,´©»òH¶Â³•`ýv§­ÿÖNÓP¨Ð  ‰°_ç ó‚lí«Œl§´L m¤¤9WMìäÙvZÖ‡Dñ`p>–€{?½œzÛz»_J¾²ŸkDVË8¿œVù½ZAÿüVããc í¤Õê"ö’‚7šçlÌ‚Ür,δ¿AhHâ/ DR¦¹Ž†Œ3 4}ßK-Cù½|B"§ƒÎ¿.Ùá>ÍK:´ÀðëÒD"°3 £ŠÝÅèjE®D”±¦)æòÍ\Ú¾ò#G㜸P¸‡ç.p‚ˆ%ª8Ϭóµ•ÁN ŒÚ®S‚‡;š&8ú¿¨*†×ãÆ7Â1û”þ!ÿr¿P®uyƒ°&7%8w j¯¾0vO’3èŒG·C/I¸æ:‰Úü=-™ÏÿÊÙLý©Åw \j£Æ¼[,©å©s§a-” ÊX:²ãuÙÆ¥÷U“4#ï/z”_}ëc9~—6Ùë³_]Ùƒ±#F~*¢ÿTŒœZ¶"²Q®«¯”]èP:—ÈùueýŠàÜ@˜PZÞú;¨SÄCÐM2’ƱàHð.>ð›ßQª Pp^.aÙ¬MçÕæ±ÍÀCðF9Ö¡{©»Béå¿ÆÒ¤Uý˓ż&±øúÝ)µÁiv6Êb™Ô©ZÇ}Þ=V^³‚p‹½Ã“–nr·ÀeïÇúÂý›‹X›ÄN’“£¨§#aÁQ7B‹¤âñÕ‹ŽúF\ |Iák³ùÊ?ïC”xO°Q˜åX9¬IDz‡£‚I÷6FYòj½Þ˜NKPž@5++Pß »~ù‹ËÔ÷±@³Þ£{"‰œVà/´#”˜ƒíÆ! sÔ–PÍ Cf"¬1Š:®ù§¹7 ¸5ƒYŽ¶"ë}¹Åð˜ib¾üçkN£É°î¬îrÁÓ þœ]‰¬Ë瑬˜œûÛ]o:}¹ F,Ö+Izv‚3kyüÌ&›Ó¿Þ=KM” »».G˜XKmrHYEÛr?ý+ËLI6Í,dŸôÁ=A$TTÑÿÚò )‹¢Eö[ô“‹Ù8H¨ÐoíÐÜÔuø9|Ò‡Is‚^ÕëJE ÷qb$î0›0'o0íúÙG5DØ‘FU$í…×%üþ¥;=¹J±ÈP"½IøBT… à”1-czDâ‹ð´4†è£c *4Àâ†2r ¹Ø‘ Æ=æ#ÐhçÃælL€Ä ¤Q>Š_¾ôÖ \q^ø4–«=¾^!Z¬£¹› ÍZ†î¡Y\0uàÆÌÉ6£à:ó¾R±œ.!p—6kÔº…Dœö‰_ç½@÷ ìÚø†k¤{€Œ°ª: ÂÍ,ürÞùô9. 3“(Õ‚X°w!e‘.nÜ~)œ„C~$‘MY.v‰D àp7™b/sKvÎ žRšEöªëǃÎF¦“jœP3:Ën¼8 Pv¾É#ì°$磟{^ päf\Ž¥9ž%kŸU§›üÂý4{CÆ.à‹g<Úb†¦^¿\œ½CøZG$ÔÓªTÌ¥G Gàþ˜l9·£5¿öߌˆR÷Al]½Î‰->7A³|Æ]¸"ͨ² uÿC¢±+QX÷áo`´Rö~ÆÀˆ!×+LS “µñ¼ÿvJû5ÜÃ6‡Ï‚÷˜”C—<1ÙF¬Ø¿È,<&|h3£è‘(öØkßvõ• 0e~ ü„j tÂaƒZ¾3°¥óÀr—~rb)Nt_aå&â Œ€½Ü$¸¥†PÁ%$”ìÑúÔ'ÕnË·‰ØˆÇ¼A‰ Ì鼡ñ1Ê!Ý7ôc|,ÞKãdH®#œ]Gîi¯oìNüÖ?Ægþ"úÿ%'QÅnP£’øûELt8K\7±Z?ŽÅ’ÆÉõ,ÆÇK]6}táiâ(÷¶c§0‘øÓÛ©‰ïå'œâý G—ãï,Lñ<¼µÙ@‚Ú¬±U]ö˜_ø§á™‰D<Ñ3ð—_6!ã•QÛ'6Å„ÒƒŠ+½0W|H²†ûeo fÞ`?x]Mô„V0åbÔ›œ—×ÁÆw¼ée~4»  IDù;…öÚWÄÞ«ßìô‘ÒtžøZ Õ3Wl¾«xi¢ÕkÈ>±BüŠa¸5>š“¨€lð}æ®m¦ºcõ€«a‘•¯Æ©FÃuÀ%n¾cŠ²TGzvy[RL¯AÅãóí)._wix#¤pbµRaÂ_„à‰ËŒ‰ËdH¨ß/Î ¬G_fë/#¶«:ríé,=5AÿðìsÓ#*0µƒúXoŽËãUÇOì4ú¾«æ¸€˜¶Ú/Ö¶¨ûfºõ&t6¡Ç°»¨²Õ«)ƒ±ñ‘µ[ðF£ãH³î¬ßt ‹ˆÓ¦Ø‚ˆî@èÜc1í4C¤¼?üÄWF¸!²Š]óÅ@ÁÅð¶"WæQC4ôyå‹$”}4?X_ԟ¢‰ã ÇÏè5'x¬Ìæ,ï#ÝÔR™Ç&ÐÉaAh¤¯=mQôZãú`1á¨ç°Õ“Uh×z aÞ¸].ûgÿøZјӓdE“ÜóŒæ‡«/6øÔ¾»JÙe¿L?•ã&ý}ÒU&œTé•L=óâóÑüusüìfçë•Y û¬±™P‚;5Ì׈¨¾æjk^5W)°„o 7®RÔ’¬º[/Fe7fqN ²÷!ÇÙþBú׺>.EHIäè hŒ,Y¼y„¸Ýé`ܨä ñ­Qm 0Ë”±ž-^öê_Hk#òAAúyxd„3Ý7[~ÔŸl+V½>¦\1ÈW…§Ý0€²Y˜#›X<kÍ4}F;*jØBRÁ<8/àÁê… ¡-)sWŽöÃWÒG8ö ¡ËwóáökÃáë„Jô±4ŽBîŤðfÿD3–sýmw±&¾ÓLŸàÞÒ*±æ‚‘9Ô-ÀÙïˆcéÎD†ÓÏÁ¦G¢ÈôfÔHÁÇQRae£ $ÆXG\&\‘]*•ñæߎŠ¨N¼,8‡óI¦lìžC$¡^˜ûð=ôÀÂœ/ëß7§LšÚs’,8f>J¨%Ô‘ªs^>šJRÒ•ù±ÝjzéÞÕcÊô›¼g‡Exzmi%¬wŠ/mª°ZsÍ¥¶h;ø¤¤ÛP$äK©´4ÅC¸—}ÊW¡áe7¶Ýö]ÆìÍJC>(fyÈtJJb«7xÇö2å:nÍÙ¼Ô^!‰âÀpIJEiYCD…ð:­-‰oµ³­$­ ç¼Ñ®8kçt-A‘r@+Úl~ܘäì…ô“M’Fi‡\2*†¨Ôp0é#‚ñÄ—ysëKIE6…z̃k)‡Ee±¢X¹œæñáfªkK³×ÛŸÛSrèSiß8Ë4u ˜Ô)i£ä… ”"òÖ$[>+ ÁêĘôÌê¹6d&½ ’Q+ ˜™Æ9zȯzšPDšEúBÝZE*÷݉®]b»Ç¥µ¯QCë,˜ÔÈΟ٠2䣷]‹Z•òú*ë¦PŠštâ(ø®–ï@žÚÐö­M‘ôLõ.ÀTnÄ/1û£œõû…SÍRU˜ß§?=‹Û{ç¦ê|Êm"a üs˜‹¨®¹k>Sâ®!è›0®göǘóÖj”3ûY! âQ¯¦™½ú1Cè %”„ŒWÁu(ç‚¥Uõt Üëy wrìv¥³Æ™þ2¸E;@ÅôC‰_a‹/p‚IÂìí)G0ÔDÓ+GÅôä¨7ÒûAćëÐVºL"‡;¨Ê‚AÝFù¶±ƒ/û=ùßÀ´Uq|äøJoe¨ Õ@õ…IÓóK7±—­;9‚ìcá¿Ã̈Vv4íŒKع/‘M‚ÏÑZêD.aydŒ½p:‡BÛ\ÿ¦6|ÓG1e#xì?Îm[×Ôwr(qÂöqêÕW4Ÿ~׳’c¶v Q»ÆÌ G;d-2´¢V¢OejŠ¼‘ñ¸K8iPü¥4¨S£:Î xÿ 0íZ4°ð0ž$Kn®Ðu#Ð""çJ–Ì-Iûhu½A!›Ü‡qF—Çý1'xM'¯Ø>N H´â”Ù̲Á*Ý! »ƒéŸh›‘µt¼ü‹õ 1brÀ»<çáÉ-7Ũ_‘¼7Ý‚Á¾ámé7jkH',ˆöZÄ}¹Ð4ïá]lÊ×9õ.5SA_ÒQÝõв¹/òýöyJ¼¹'’uëÝ£<;Ç0GI좽×0ejµ]Âd{(Iû2î½ñ€C;à‹ÿÚ¯Q²÷éj3ì› pÖ9_Aô¨VÖä9Hç(*äÄÈ; iRm­¿P׌#rV06…™àÌM•ú]^A°#Ìþ|&á4y¦à,Ê›O[†6•ëçŒÇÄqœƒ§:Ó©{ûì‹çÜ…*‘<ƨ4¯IÜŽU[ò%¨cQZ¨`Øx#&,(0|†¾|Æ‹×ñí·ŸÔS¾sdÖ®è(gç)v©‚Ší§£V-2€§½L‚ïwm~RÎV\Ä ¿ôãhìʾŒ*|Ö/?Ïö7‚dì%HÕÂg?ÙcnɯÃ0¸œØƒ¼ÎÜ‘À²ç¡'RÌ ø74ÊPº£%; G°bô÷QæžË¢ ºbY¢òÉm9ß蔣¸¤î‹]§o´© Pôˆê/ú  ÉKs>‰)<Ó.á=§ŽfG×Ì šOÌt¦TZVŽŠ“Œ{vÓ‹†!'wHmâ°ûñºEDûÈ<ÛilgqLópŠùÿ™3+ld\|ýÎ>éj‹½²S.ë&Ÿ¡ÿv^´o™á-«co h EWœ?ð™pºei\LI¤œbw«èv ?¬ËæµÓ4ïrzÒ½ò¤¸êߦþ<'¥e»ùX5,É©IªŸŸ¤BãRÑI¥°Tqê±jn#ávNtK˜ÄdG¤¾zêùý äˆÅ>2×ÈÁYêèY»= ë‰8uDnÅ<ý›‰J*_E,¨ÕåhÌ\|Maa4;FEP ð«@™nþDþ¢ên(*°ƒ9 em#FCÊÒ¨RÕ³Ø6©8Ä)~'£¾b-•(hÚbN?È÷ë¨Véûê ¤ƒDS‘Ÿ½ÝÔ@n*ž™»*Ž.cñØÓLáún%Ê€$©Ÿ½›BC“´ûõš¾*Ù!ß´Ž§£Ð¯D]ÒµBµ­7kT¼'Cltr7xP¥¦NiüŒ|ë~õ2\HŠ³@é:9=†ý[Ãókæìëg*‰ób‰ª$¦LâÖ鱗€ƒ¡äé~µ¸ˆ³Ê"Ù kåÆn¾ d½+¼çOp©á†´‚q&A fÌÁ1Sâõ¾Us[JÄ+kÆÄ?Euq©Ìßñô1¡ÈkS àß!u!“ÂöS›7Œ¢¹^Ÿ$DÞⶤÄ,À4aÐçQzߢÆ |Î&ñ¾Ò7ˆ×ܼýš{4Ç!Š~Ë;F7¥'Lÿú” åz*Y"Åfø£ŸÖ&6õõa ëwkª"2oûPìû#…•JMyÇ*q·‹‚ ™9U z¢šª.””²Rªì;i©V¿¹Gª³L"™Jò] ù=‹4'N¤~°U‡ a¶•taŠÝ³Y£Þ®‹~«çÏx  zÞŠ§j7±'²M4ˆWÜ_ƒaÖý1ÎùÁ¦Þº"L1UJ!ÌŒY±aõÜŠ?× •ò~$vÍ|Š§|øö–žÑiÓEF°víeùI3÷ 4‚=ÑO”BȈ:päOâ–ÕÎ,¨¡¤òÔËËž«zÎYƒý/3äÅ7a†”d’†0µÑ APšß;Þ˜;ýë[¢ý×qMôõýƨ¹Ó¸Q7”&¯øô¤,ˆ·é¸ºªZ3àBbç\Í@´+©}MìÛ\˜™™ÿ3>ОMðe>Ì~P³&‚ÂþÖH>Èg[·Ë¼…› d¾#&0‘ ¿"ÓJŽ7ÓJY¦u ÂW©”Þ™u„'—´Ðt5ö ŒªdŠð \¦Uœß,ýy`Ër­HFÅÝa·ñx° h[F"b»~ˆ»OeiàÈ7[$"«K¬JˆÍ;ûf"ü²¹ÇgZ7\Ê[ ðÒ®ò«ï¼ÞÒÝ<\Ä| u˜ö]VD ;„¥ãÓ“‚€¶žÉ‘µßyì›fó8Àpsv¢ú¼ 7(2 ãèò!ê—šOd-2|ÎZÇÒ*2‡}9c¦B{È4Ü*žï4TIóÍÜ–Hÿ™W¨Ö—(ÄìAŒ,ç¯d c±Ò³Š ºßÕš¥$Èa[ýãCrîDþ¿òŽÎ×V Ó]Ró›þt¢òNá’ŸÏAšÏÆ—§R¿­˜®?LPÑöú,Gã<à'f ÷[€kãq˜ÖâŽÿ bk“sý@C·ÈN)ºÃY0JMxèÿ¾ñ#ÉŒç1/\Â̹Ž&ä(ˆûÛ¡¢è9Øâ=ñ*0âTBšv ÐeáÆ tN¼T×ï·´œ;!CI/—ž<œ»Ü´ÙlÔãË-à›·Â»™YbÅ*Vþãr³ f†I™Ö7õ/:Ý ¬V¨Y¸¡O„ÕÜ[¤`«N)±íD)¹×ÿL(‘e‘Ng–èØ][åV‚%¸›zêa”kEàp’7ÍÙõ ÿÉ«¼A [妥±›ÕúnÕ*gé:‚É´ zNσÀ;LÁ.¯_ßYf¨/758‘EzÛ©…ZØ[ÌÐÏÊ7“vm.s†˜®Å5÷E²Ç²$ÇëÇ `jŸ´v;îßN½•ÑlæÇzƒ½“{šehkÿòÇz!׋N¶¹eå>  Œ£‚\{œxŠ»‹Õ™Åº«Ù;X)ëWƒ»`÷'•n½lH¤I+Öž2®RqYú½CŒ=°¨òÒz?áêB³éüˆ—k’FØL܆ù$bÊ°žª]spDpÔk>=¿=uØxéØ•B4o¤$'H„ËÐQó_ß¡Åôdû¿«’Šá1=ÈÈv™XQuô›xdÀØPzÞ¥ ñ–øõ%ºB°h*p%qŒAV<úpaeæPòWR’$;dÈ[2â…À<²¹ÿ%áYÀÝq‡®å〟ë¨È‰ª„‰æ“ ò‚bv!òÃ3PU]…ˆ¥IÅ º¡MÞP¬ÖÀ¤J:’­0¥yr-’?E¼¯{7æƒW˜7¡¾á@xëšw\]e«€VŸk¶jÓÓqUÓ7Þ/qt;›®ÉÓQQó$4pm›æ1<™bA+•z•ÜPçéQTí‹Hl©£©TÇiÂÖ¸Ÿ~`°E¦- tõ, endstream endobj 273 0 obj << /Length1 727 /Length2 14460 /Length3 0 /Length 15030 /Filter /FlateDecode >> stream xÚmycpfÝÖmlÛýĶmÛ¶m;éضmÛ6;¶í¤cÜ÷œsÏw뫺µÿL­šcŽ5ö^«j“ŠÙÛ¹¨x:˜2Ò2Ò1p”Å„4YX tÌ0¤¤ÂN¦†.–öv"†.¦\uS€²©€‘ ÀÄÀÀC ¶wðt²4·pPSþ;P³14±´µt¨ÙÛØ»Y[xÜÜÜÜœ]éœ\ùèþY¤lj p±0˜YÚ˜„å4%åÄârªqS;S'C€‚«‘¥1@ÆÒØÔÎÙ”`fï°ù0¶·3±ü&gº°s3urù˜™“½-@VTEPL^N *L¯" 0´3ÈHü³ÜÎÅ™ëŸjSc#MGpùËÆð,£ÿ±þ'ëäñ‹†‘`biì025·´ƒ¡ÿo’vfö¶ÿ„M\þ›úó?øÿpH 015û§ÚÕÆFÎÐÖ@!loëàêbêµ71u²ÙÛ˜D=\LíLLMþ[mhkiãùÿ­ÿ¿’.†ÿð!hgþƒ ÿ Y:‹Yz˜š(XºüC¹™¡Í?Tý;®núŸýùW£ÿ¬þ§àŸ}5±·³ñü¿ ÿæßíèÅ¥å”e©ÿGÿN‹ÚÛ›XÚ™”]þaÔÐÉäÿN+ZþGCÿ…ò_MÿŸ/kèâdéÐf û—Hþõü×ÒýUBBöÞ´LŒZfF6#+€“‘Õ÷Á5vur2µsù7íÿÌò_ÿßr25õ05†Y_±7æþm•ÖZá'Z4_ N=?k J½Â¼:Ó{«l9la¢tó ‚“ï‡?hÄ»wµ+ùÐBdg w25M®»‚óZB\iîW»ý®XðÖ®w¦†mk,vqï’±±¤@5Í’•{Q+åXÆ‹,ÎmPÞ9Ž´¿öçÉNœR—%SëÎô€ºiÞ"¤ðƈÿu$,?xî–[ߢ°"Ðá(L+ÿàÉ°ãb“é7µ-Иݧ!ÿ± {¤#V½Fß±Šz-#Ìe˜!GIžÆºéon‡LDù…•0íÎÝqÐ2U³ä„õó £ ÜT(¹Xê#»WðFì…˜C­Ì˜Î¶*7,“¾ÖB?m!<¦Ù#µT’u†H¦f²ÏwËpNžÕ’¸ìe¢Nç}—Äíª¶=Xzë# ½¶¯Íæû–c1¥0G_ô$íÇÄlè+#¼›Tˆ+ÉUp ä$½SŠÂ/ñ¾é;¸¤•D÷©P¢ŽI”1s±¾ýzM©_[žwv˜¡!‡]Q¸ nÆ/”£ ßaMBeëÙ1G]Œ‹Ÿ’·qÔèe ïZÄhpW3ôÚž>Ô¯fz:ñË!è‡h†yíLŒ|•&µÁƒ–Ÿ`™„¼Y÷íÁô žþ(R‰9©e:´F”UQ¸ôp@½Š¦ ƒºÐŽP;;ßs@@Ÿ)üf ÂôIŽ‡ÐŽ¬?ë¬LÜw…–ÒÕ¿+¸iÈ™ `@Ç\5ú(+fþ|Gÿ.>e_ôLOöoÈbf›£Ï¶ÞÚxS…Vá±ðûÚî(yé&–BpNüsÏj=icÏu>h‰/ȉ;±3=èC1}Œœ¦k ë,ç`Íc2TŒë[]4É”rÇºÓ Áab•Ú Á Õ¨ î­>š‚¢ÛÝ‹s6P¯£–‡—&?Ôâ'J”…`ŸEÉf‡Üdò›‚6Iå<±F,m˜ÒEámŒÔ2çO±‡ˆÌñ€h.ûoÍ[Íݦ¿v‰’åØ¿2šxië7W¨†“«$ö¯DƒÑÄ%ŸÁêâÂÃS´Ê½TØÂí¦9««_ ÚGà(ÎÍâ<ò -&ï]\¥.Ðá`£‹ù¥y[ƒõZ EE-RL-îxÆú©8‚A«êˆ‹héõ€È•0 - °2³RDã´¨SõÓôA¸2‘7BÜôÐþám&)UÛwa^wªu*¡æ§ìùêW) KãÕ÷=]ÆhU 3-ó³rÖï›ØˉÐïI˜˜ØmÄ´&>ÿ† ÷ !§ÿ¿-„`nG \*È-}éÝ®¿hC^Rƒª3~¾õªDÖÅçÛÚµ›ç¸tbø$5•“ÅÏ,!$>a¦2Ή"¤•hæZõ™PsјF~¥=mE‹‹›ü¥Xy§Îzè½7…z+^î$/'páé"y@1üÛ1É‘Îη<Ûü8²OØÕÀ*\°<×ÃÆÆ;ઠ{°J-˜üÑÛIÕ˾iÕ½gJ,k_WÓÎ+úa`D(€¯ò÷ZË…˜k«¥“Àõ!@ñcXûˆ-¤MêkÈOÚ¥Èmx\„b…é-œý-cN[Î$a{ÙMyaKy3HãQé.¸ù½¾BÒ…¹«–÷Þ… “vÌ£í—Éü„èìjÏ-‰þÀ“J`ÎâÇ< LRïR)dœŽ™¡@â.THR{KÓ¸)!£XZÕUð76ÌóHâdÂ#çPx‡­ñÚ¡|"8h+LL;sV­“t¤ªTp¤ÆÊŠ×u\y<«Ôg'8„'Fa«×ô‘Žºën€µâÍÕÒæajÿÅV1Í cU=^좤f‹C6"CZ#¯8ËM9[†T!o}¥ž ? V1±\n*ÆåIzE‰ëxn3+œ@Iz½îÀ®ô\ þ›}ÐÏ´ ¨¬,÷ˆ²ÍÍs4‘s±×.¸é‰+RÆÏÄã»Ás5¦‡N–F2ñw,‡‰9R÷_¬ÑúŽ4Tl|D§ÅëÈ÷ƒ5 .ÌäžÔ×uж±ËÞÀŠåÞ4Üõ+I÷‡;:l×$Ç›‰I_¹ðj» $Á8œ6Æз Ë¡ð̓ôh a˜ºdúÉ,Äs‰NN¯û¶ÂõÞÚÒ< ã´"fÕ‰“yýÅl¨é6áihLé,ã4£! ²zŠyI5,‹¡Ç8=NG5£Fž(~ößâãû]œò£ä«¨L°.£—qƒäó£Fî³q()¬¦:z¹i˜DÆùhP½¿$uP+Ÿ·Óuæs!…Ä‹q.U¯0/ßl1èwˆº÷yÑ€VÒ\)úíReòêŸSZ¥ï\ òM‚Ò¿Ø dzÛIjv‰À—¿¹–Ë´í;νˆÐʲ[J⺤Ì,˜©n‘1nø0¢³¤¾-¨½õ-`ÉÒ»»I“"zÞnÀxøÙÃ`J°Ül[gPùØÈYQ` ÊZíXuv¼_¬Nþ¶^¤CÒ|X¼@pïÜÊ2¿H’Ð4/ÇW‘( , ï)$w^5ò[Ue=ÅŽaŽâ¡ÒÃǾ¾V’—¢hMAðY20ZI`-gTÐG€ë;ˆ‰I½à¢},=﬋€ê ×!¾„ÉÃéÈ<èß™dv3ß»ÉB9.ñs-òTZ W»ã¸Ã n0ßkéz6˜ϘsÄúLÁa®ÔËÄ>„ohKsã·¸ÄÇó_ÿ[ƒw„ïƒÑhïõ’vEXKʶSHùUHïkȈ˜RЦý4„¡«‹€õ­nzŽoÌŒW°wˆU 67TtþڇšÒ¸ANñªËfþsŸ½ÚµFû¸1+RßNm»-;¨Òš¨»}¾JMjÊõ÷⼃é/8šîaGe8?âËØÚOp©Äig® +»]qÔ™ØÁ#Aœ±wj¦5‚h8 çzèáW œþA÷ŸE$ÞÝN Ñ! Ê_2I¹ûIWK ë—˜ç TÉ«áj¤Wù„ÞÈ›yBs ”ÜØÂü#=5´oÆvë mÌcÜÝ['«Šôû”?n‘”„û2WÃ^½9ϧg[$yW¦_šÜ½^°½üÈнŸÀFìêhrˆLÕµRAmÀvÁ»Íʘ4,BT˜çŒ'uˆÖÑò¤(ñ*úššgmÛ³‹ÓdLÒÕ½÷¾a¼èà5aQ+¦X9¢Ã$ÁµîPí7ÈÁ8Ì~ßÐ2iþej§ÏÁƒ„FšÐÍœÉ `2±Í·ú¿@Ï Ã¥p=.Z?8[G;±NåŽÉÛ4:‹Œã’22 ù—I+2*Â.øjãºX¶8Xf¦0¶,[&ÎY_±xÆߑ㙊¬(í…FYFgÔïd©«WoÔö´©O{åçYås9>ö«j4‘úìè€!Ìsœt÷\«M ŒvV4F«üñÏC8íYlÒµ =º'Ú´:Å¿òÀTI¤FÅ—O Ìz…O2þ®f U×°Á¿Þw­ÕÌz䫶7>˜a®1û³²hÅîÕߦÛßZQ5c).®E?Ü{;d›H£AÞ çVn„YÁ[ Œ^°†¾Ú8i=|Ró€×ühæ‹}ç¥èƒFv6²ä&LW†?ok‡ÆΣ ¥ïºßB­D3 ß ÿ‚0€ç{½mž‚ü…ûÐ×mØ5reA+’Rk¸ñ}ö= q ÄnüИ’#'£ çrVagÑ4Ë@6ºöIȶ dO—,H&ƒqaõ¢·ay¼­þµ.Ùv3h`Õz¨,¤ÒJ=Ÿ[8ú•ä²:}ºÊ`x½Q¤ÖŸb~IXQ9*…¾H%±p‰¯îYK@g3iKœ»ºË‘å¢M86ôò?…àRàÙzÆŽáö›½cxÿº­˜Ù.IHô4+c;4‚é¼þ 'I 6šŒõ[=à`éc>"ÛGÕ™µ¿ÞB[q/t@¯¡8@/yßèeŠæè–³ ¬ù~×é7&)-ZY-[zïæÀ†þÂh©#©PÜÛ ÜˆtøXYGéýú`ÇóB5¯Âô"Ñã±G•1`3ýl„$S[š{<:¡ÖŸë¦{¯?X®üÈÙÊmÄëE„w)‹:$ù„å-­2Y[5¯ÆEA;G;¾RßvFH¥¨|¤±âªpÁ!\Þ²ñº>³—D¦¦h@ ÆSMZÊ°e}B=ÖÂtÔÖ?%hŸFöpˆÖ ?ý°0÷yšG›>V¶ŽÅ£cGó§›åë‡N:äS¡úà­•}q1U;'Z_H4yzô@d7åÉÔŘ¼§•_–¶(ÑÞ^² ¶¼,Ȯű](ò[3àñ1ÓÊ~L5X%ZŒ½8Ž*RÜÚmjîËcèÎAû# komVØ ÚëàÑ6Ða~vòrš+P©ö¤öJã‚x”nž{xA©pƒŸ›2»¹ Åü´uVû¸4:P¾¹GÓRˆâiÁ ”Å{Úx:X;Ö´hHõø×d’| n çJ•öÎëlY!&Ûkyš¦©iRœÜøT¿zje8ÌSÚÂÇw¨QR”“3’v zTE¦L˜\”‚hêXÞXМǀÕÞªæý@-<:ÁùhqÇÛ<?´²t’W­Üæ4<ØzT]RÙã4†ØPª¼.¿n{x^|%ïO®I¾å›c`Îý‡È¹çõ‹ðØöç‚¡/®½òfãÏƪ¸I&vƒô_…F ²pï}*;kƒiëZ†k ÇÛuåmz¨üäÕ” UÒZ?×eÍÜñœT³ˆëý¥ ›5‚“{åûDÉéõ”ú•<Ž? ;OV“SÕ¢Yo\’øádʳŠ2vxJŽŽð3­Dltd'XcPrý÷y¡´C~ ”å|SIÆ1•:õŒ4J½ü-ɺ·/$x<+wéAײ(„‰ÅŸK…k‹g¬äýÌãÖkÃÙÖËìi]i¹P"M¡E鯯zrµ¾eGlYÀ316vbç®y.Qƒ“¢r9sì3ÞH,}‹gßgLa$+ùdn“Ã& 7Y+´Tò®ø˜3“z3Ü6Ò4¦|l†Ë Ï£Âæ¤d<‹Tä(R¯ŽwHDk@P¢”â[ºäŠOóñP“8^ü6”¼Ö²¸(x==æ;óÒßfÉ˼6 Ù,¶×;R¸ùë‚`.'oã—ÅØ9a€§˜h™¬6Ê|U0AkÒ1ä%H±c)ìGFТÀa‹žM%·_è,ŒêÞh¢zDVÞi2 ~Íô¿î›yê„…”>–â”ÛŸÐEË ç÷·¶lQ3~tÀPf2áõ¯áÏ?Q?@+8XÈ÷mˆ:+ˆµ9Ä…èiª’ª_DZAÓ{ ÁpoP!ÊÃØôuø¤?øíyš<ÖKÓqrŽ@\y&—0\¾ãÜò¥ê'ZrÂk ¶"ü^ý¦&µo™sðòו ÕûsÀ“,OJf2ês½BÚ žô:£$ÌK_½Â˃p¢Üó7÷vKNQ2P' …VÈçÐAQ_ ÐÏüCÁô¬==¢‡˜ìŽ{Ô· ®Ð3j¢Ú+¹”­zJ>öºhõÙï=‚ãñŒ‰“µ‘hsWÍOÉ{VûŠIˆDf%{dû5È`›Ÿê²fV_çtÇlôµ¿øs!êKŒoæPdQ„šQ†"ˆâyxœAo¡8‹9«ú –OŽr{Là_Î.Y6g5|¸ùÿÜ¢ˆÚ €[~9G§<»wŪ [ûÓbRRIŠkšP‹þ°§±Êj7Ñ¿‡RvÉ6b3n-U&òŽï-x¦'«9üy"üFìw€ýDÉ2Ëü‰ßòà=ÁÆ•Ò¿U·¿KÙ~x) 8ôë—ifH»(jdÞ•ãü­Þj;Ö·Q€à˜¬u}}N‘¤á ?Éy3¹ÐVe“7¤éŒq2 OjÃfÎØŠžtŸýi>ªÊ “F_‹x .º+63µÑı]( $¢>ö»PbüÌ–ñQG¡°pÏm®ñûÙ_:/üßWZ…èä«OfÀ;„Aò£àG!*P`éuÈÃmÕcå!kçµþꙢñÎêâK`™ÌôjÜõ‹ÄB¹S Z•;¶‰×ÓZi'ÖËù©(a® ØšFrS$v»FæÒó$Π³Ä¯Yp¨‘µl8T'²KŒ?ntò.™ê8,ŒwÉUíòù|V&{h³W.yòÂÖø¶ÕXÿyÐV¸Y[*e4 ‹*]z|é’ƒ°•·‰zFFzÞpªa «\l…%Ý ÚÝ—<ÎQ8aʼnôëA¢‰`ÚÈÝ&Îèš½IÄq2¥ü‘Œ lþJ¿zÅŸÅée^ÔâÔQ‘+êhmÝö{¿T#Oµ3&1V…UËØ÷OJŠnÔ6ØÇÚaòð ìXµñÙ½Ùß²‡HÍ~-­éßmãœs¶Ö«²¤~§öë–'(g fFî êQU§Ò‹f¡ ê&Å«§B&Ëï…²";k/y‰©2…’†OpÚÑÈs‘IÅEïßûñ} ²f' BGMý •ï>“øÞ.±áÓçƒÉ–اö‡Æ2AÒž—z_§@#—P§ºò¢y€O—¾`"}€ö.µÎüŠ|¹´×)¯ Ç{´x`)jxkUjcwÙFÞ&,\¾X°Z‚KtÕ´5c+ž£¡Aãn~JØ Q^Œ%úBc{¢Ï:Ô-zÄjßϤYs+«Œ™ñÚqÆnÌ!7('Ÿôˆsæø]>²ßU•ó±ôq»špþòbŽuU‹X÷§÷Þ|Xå:ÑO†öF·Êã"R›ÎxœÂõ Åò~Ђ8óGßKÊJþãþÈ/dÛ“±†}ê¤{æÞÖVöü"~lòDÏ£6Ÿ4xbé¨©à‘‰3&Ö†ÁÜ઺ë…õ È™ƒ¡:"V8eoÿ]ÖN’[õiF_¢a´íYì!!¾êµ?=os®jäüCüñmº¶ùá*,ÿ‡ËÕSö´8X‚¿kM°6ê5öéÇZÕ.ÔC¼UQÇ b¼` ånKù9þ|̃ݳóî£ è ü[%‹*´~¤&Á6¼xÒ'UÈPæúº îÜk»ÜŒi6žîé#¿k$1ª±a;JÈÊîhµ éï8iÜÖ™+%”ÆÄÆGCX”1ü;%÷Ñ’{ÓRó¢ÉZpÎyEúïÄFH ñ´;6ÃIlt`5®u¤}0²='6®Ü¼™nût,åqÙ*Ìâ`L¨B"uj†Ð¨Kjˆk ʃÂ_9…t&ÍæE‡|¡”'—Q4qG<óˆ*¼!€ñÛ‰ #s£Ÿð Ò‰Ž’,n:î‘߈æÁD—… ¡Ã,wÕàŒú/Ì{« ò•Ï^À«ÌZPí½â÷6P ɾq#ü@ÓIÈ—¯WQí4¡MÏb‘ŠbF3e®u%ó畃RY)’’6‘—´C{IüÄ¿„× ¨$ ÜåÚƒ8~kG웧ëN"ð1Hj-*å+´3IÝhB{R/ßiMz”ú`‘ã©#—äDU™&~QJÄdÁü)A·e“cÔó"?bÔ⥗ÜÀaü[¢xvqì,‘êqŒÆ(¹ïh]òye,Á I?r4äªú¹Z¾ ñR,JL\“ÑÖ³•7:(7ÁFŸB >dF]Âø7…WI7Îm”ò³¼>ÿJàáœÈˆ–¸z¤-‰'í³9(7P•0éÔÌ¢j½â¯TqÏ{¸,"¾©?`h¶¿.MGýL›½òž«hò“¹†Ï‰'™{c1ÏzØR§¬Àì×Þˆ^mˆ(q¶/øé„Æx[ÒZë YrCfõæ\“ýTä3Jå†næ œ^ÏHLâÙO«±ÈâÛê ›åÇÑÀµb¥û´øîeP ø¿ï½¼1R¨§x¾‘ÞÕÄìÊ&I¦‡›Y³§\¤Têusñ#H´P2ÔÒ•üا6²Üú½[Ú$ÜiÖ% QÙ¶5kV¬s6m嘓±XÊï‰ÍØ0Þ0Á½ž¼‘Ýxû Β̤1œ²5!ä!˜“CàÚܪf¹6 ä%paÇ_^̹órt= 5ZX¤} o_š4(‰”A,ÓºKZ2}‡¤aîwiÐø³ÉNÅÜú+µŠO*X×Äyr"³k^àÀÅE™t_ÄÞÎ_¢­½÷D‰@û÷où¿Z¡éFº¿x§E€õˆÌ>2½ÙQ"W’Ôjñ[›S#Å´ºXxŠD¼ ‡$ÿ9[™ªüá|(WYLä"4(h9àIݾmdQ=_/kûƒz’ö÷¨újèÉÄÛF§òÂPW_Ù‡Pìê–u³wèàð†xÑçý]UÛ¢Ÿ>0ôaó‚ï;ßU »õô­ `·‰j OÄNú&D64%9‹‘Àu&ƒ.ŸßHúA_0t¬þä“Â;»7ZÓt[fH©Û«„`–.²×ruüUìt íz –†—­‘óA(×Ý-ŸÝ<Üäø¼«»!(ÓÜp÷øÙbö¸Ï«’¢¿Î¾Ïp°®…>ßA¸ÐÆ&xYpŠBæŸÍ‹dÑy7øƒiÒ,twŸüIP”¨ÏÈ– ‹t-¥„ UΈDDâõœ½áëßçÌŽŽ6è™­°¥0?Šè*¤ ÛÌ)† ßÉã–Àt“rkwªió2Û­Îm;Ï€±Q[¦øŽÎp$çàêKŒï‹Àœ[¯'Aµñ )DøYˆ¥À¬KÀA-¢• 'ÿÀ§0œ ã–JâŒä_»»£ËBË%ó„fášl½éó¥ËëlWWVW£‰­Áù-tÝ]­Û`ÞÁ€No\OþËÂ\_#_˜š:¥Ù–=ŒÖìÞ‡ôYbT"sçŽNùd5l¿~î´ƒh“Uˆ}ÌIcþžN0–å×$Ãcúo‚˜lÒ|L^É$s*'ܱë èp ,:ÌpàLŸç§Mü€g+Ô3«fÖÉÅÌ®²¿Éé„­bDCZÑ›IÔˆ}t„—©f“̹ìúŒÝ,i7¢ú¤jÁAõšò=ÔÑhžš'ÌÀØ^ƱÍT4ƒC}䤹îʯ_Tßß Tì@ƒ:õ‚t¿>(« }}²£|Í¥iž[î]f_m›ï´‘V—ÂÁ|L[—b{Kÿ—IñUŽÞÑ‚:ãr ÝêNÑ\#ò+ÄéhoDbJûŽD"⹇ÄS°{¨…(rØ'¼˜€nËŠÊýîÂœÉØÄÔšÌViºvÐåú›·^ءľì'm™5GÞÖ 6TàcÞS¨žHÁ‡öÕU6üHPò6_€ZЀõyö`“ìjÂbfÜZ#Ñ‚×ps‡Öç…=†Ëãnò›ÎR³8ñ‚ö;F.­ÖPûéŸ,ÙŠÌ!¾£ÂCŠþsýÛ®×ùM0¬V.©â»µiz«û¸«`+{Ãø‹Þwký„õ„K:DøL{7Ïo–:ÝŽÿ…Ç2)ËKŽª±ãt Ó]Ñ­WàáI FýÛ4àçχñ ó@t$¾Ï7®ò¼¸¥Ò7«8ä ¸öZÕ §Ž-aSÛ騘T# ü|ç¨øD¥„&o«ï¦í ¯†DÃ^m :hRîlp¦ø£À_dy}m¨åµ#Ãö ÖþÒ¾û:ýË\ÐQvíû׳¹`¥Ç”j:/>Ðõ4}3ƒõ*=š¤JÄÅ.…¹ûs–Ç{£ŒÉ¤"„EHP êÈÁÑo1¤ñýViî~? 0ÁHÏÌã³ïÊI9Ì¿PÀBB 5‘è9u¢W¬Ð°Ñm…íݵgæoS¦wLµò9››‚ö–3­¯ƒ8¤`¾ ˜j÷†®:5lm¨y߬ÎBüæW”á¸Kc;ißw“ñ¹Š8í±%-à-d׎AJ w¯Eó©ëy‘mTðÀQ_Û,Oê6-d’ë~T‚‹_¼EÎT€Ú®/”Ì©\˜=n‡8DÉ¥~XMUºf¨Ô žSMõû…,ç+{¡Ã±4» suaùÙ›´mwf{› <°0ä[0„`ºUÇ1ÇŽ˜ËÙÈZ²ÏÄoï€tx‘nÊè™LÀi±Ÿd‡ÞÁ•ó ÂÙ.ÕQÓì<ø9ɘäBÛRŠ hg¸!~)@ÜÂLú^ÜüŽh¼&_=öíûÒ¬È04+¹Šðr!B×´‚ÈŽ´å ]ˆ#À\Ü=ÄŠçY\ût"wFûɸxæÀz`Ióô#¿ÞšŸì»ï½f—ºË^Æ@ü÷ÌÂDåyD=¬¼×(âAK²½éÄ@¡:°_EZ3éU¸Ìg%¦A6!ªì·MÄ–8ÂÏêjߙÕ•R¢[g@Èñüd½Jœëÿþ-:È{’éà®&/ò§Ä¢ë#º–à*Y6Í>„CÂ!’rÍU‡Z¿®}ÖòÖ:SÃM(wÞÐO~ª' ¡ÞÚØ„¢f]I®$Ùë„XùGÝLJ%îô®›åkõóþâ™K~Õð&½OË® Ù^BÕµÑxÊÓzÛ1ñL5±'Ä6Þ]ø1[ŽšÙUZY\‰ÀÚ¹¯nÕÍÎÙ(A«“&5UK¤m9^²C7_þ»0\–Q_M©½’Žû%ÞP»éäz¯nh}: Y`јÑñ4´T¸ …ËÉ´ü™ƒú¦Gæj®«[!ø˜â]ãÂúrWo.F«ŒÐk~g‚/µ¤Ä´Ê,€ï݆úI—g,ŽŽ$ˆ¿ØÏÁÏÉ‘-,/,ëº*Yj¦/jÞh<|ãqÉœÀ’Q}G…êáE—Ýe'ZQ…-²ãÞND“ß5X¹># f‰c@ÊPŽEïHÐó›Ñ*›šgKžj©ÜœÒå® ""#ºJÔ¯Š(˜j)GÛuÑRO4¼Ô×4ùã$i+*n:Ë€™g2³…íƲ• “ÀNâ’pûö€èµ*¶x¿‘q-CÞ@ò1_&MBŸ{`Ó5YNÔú@U0JrÙ×ÿúé)¶{íþ‡Ó¦'Áh°Â¿4—?yÀ•\ G{¦¦‘Ý–¦Ê ‡ê _}€—¡»L²™z\}€ ž~Ù´#b‚ÉÝõðWn©E¬ß{«æŸÚZÙlœs,)Ëïd£éåxn¦cg[£És{"TŸÛ‚_Ï3ÁÉýlx*Ú¿ Œ5ÂäuG‡&ø;À DvUëæ}òƒN‘bÒXC’†ªëøÑj^yM†Å‚†ðÑû#Û¡nçžáí$aS ÑíyÊ-óQÃjÊÒU¬Û:»i¸üH9:á¼7 @U\¥kVfåCˆqbBCv.õ9Qú¸dÚxIT¥†kÓ„ yTÔË|>8잉þ ¾>Àíú‚ï4{:+¢=»ÍèØ$ûì}hQ<òW3u·3^(M6ôâ_ö”fªTZd£ï³Eà8:ÚÕøˆPmÖnžùåœ7(ܹJ^kÎSçäǘÀ¯Á& ž×búÅÓ§p è Bp Ƀ»®´ê˜ªeˆ‘U­–d*ÿù=1`’à*盬ú2α:„v;¿œ4Ð+Ë@.äî^`ò+K¥y—§R¸£Üÿ|FܯŒeEÛœVÑôSˆó ëw%‚?‚wM9¶,Щï›ÞŽÚÌO>¥æ@ó†î>ož1éÍD‘ðnR96‡3_×—÷ªŠ¯ÐfCÈÄ!…MØ…ÿE°‘óUáË/BéîjB·”˜z ;ÄØÃ+ÚÚÆtºrÁ¨µ×a]#£¼üUm¾uUº‚œ²e²™S4àà ¿lzêW×ðUp¦o jʹ#ø”?aÓ.²dcŽ–d:©‚6?v0‘m¹¢»ÚX>`ˆ¥O¯Ëˆ*•¬¶®Ë¾>ú„]¯TX=i˜ò>Í•+¥äb¸ É*]+±l†š„LìC¢[LΘÔk¤ÄDæòk$Ï (K—Ñ¿]ûºlŠ·^{ÈÑ,¹ì]EÞ®å˜8•lUI`$Pl—0é|õÊ;Y»,öÃÅo걩üzl¬µ‰iöu<ä!4K߬òº²¨ó'<-³4ld,+Yð}‚îɨH4 6ÆÐf\zɨ<•úÓ6bI†ºóý9€ebáY“ª»¨¢ˆÁ¡ÍGÈ0^ü-ÃQˆTÜZèwѺy<:A×UG«:9¿è U=<`Ñ`ãKÌ|Ý oáñvÖC¢/¥ííÉB@.2µ8ßp"«$ëܨsú…&`’ È8ˆæ‚¦YÌ8{ ðžÙ†ª: ¢sSïد#\•wiªé·œ§²"P½¢¸=.ÙBÛ. zF¢üU%,ÄŠRþYR¾îˆX£}¹ÛµúBt ½â¸&:*¦U¹«W5ÒB¼¦Ã"Mziw)¡­nƵ‘ÚTƒÕ¥Lm;2ÈÄ3€’^qÍT ß9wHÛ°Áà9T¥%Ácp|ÓÆeÈꀧcÓãAÐÛ=ó1¯¢³Â!ðT.vJj'ÅÙ &åËáÏaF\>|a˜ÝP÷ØbŸ=6Òž' å䨨zuÂúaçÅØïc¿%âèÔÔŒÏãûG¡5Uû-²k,‰A ·§vÙmëâ6 ÄÛö€ ‘=ƒ¨¸Õ^Œ %b“f¸6¯ «ÎP›”S“ðñÀC¸7 ‹z­7šÞ¡~ )…é ø¢Ù@·Žö¨xýD·0à3SÅAmõ¤¶ìjµs´(f,M;’‰k.ƒÁúu5kxO‚Ô‹ Ät¯rìE^?è tÍå>˜1qv\íêWfTèú³"ž# ·H_Ù£½|gÚŒ]qáà  XÜ3 k¸¤*]†%ÈÖˆˆ€u@罇z½ïÌe]6¬ã[˼ {†µRLÝÁ8ëá/Ã÷NÖuê¿Ž/ +‰Õ,M/Ž‘”8//`Z ¸LÑ|ý{NÈl„6IüWÄÐØÂuÀ"áíBð‚8ꑽÜÙ;ÓUšùi­‡Y¹ Å0[Ü£HÇá’s“&^››;Ïxœ5}¹è‚×™Ì^gGRù\{{×Åzx„$~X)ðzö›wˆÖnnÞü¸`ŸÇËãìê,î¡ÊJW °ŸØeTIŠgê‹mæZ÷;#„VáÐ݆uשˆUÐB™_(ê[¤ú~´ é+V;ÌÄ küµ»aí­ê7ôÓ@1ÚÌшT·¦ÇK%#+˜I¢#báAš>ùÛ”\@ŸìiG?í #%ͤ0ÆÍ/ß[c8Ò…üÒVŸžðp$lè›_T¿µºªµ©Y1í)¥û>—fŒ,ýáçú>‘"£cÜ¡_8GÂÚÕµHÛ­Bõãz—Ñ«Æ™* S\“8êÈjÛ}…(µ@ª4q4îVù§^µ3ب¬Vk§·ð®¥Nå ð½@¾Eº&l?壶Ϝ\1\è %ºÀL׌ùÖL] Ð(•l*æ€uï™)¬o¥(ω¯5ÜÌFa5r­ z¿2Vt¿£¯ g‡<ø§[½ÅFˆ¹ìô¶,Î^2ÊúÇ7Š€z£)šúV)⺩µ°‘‰¿¼²’É»éé¼j->ÙNÅÀ_P£ØÔF«fEÖæߊëc-€´Üö}[’ “¨i½ea&¾ÐdÐÕ-¢5¬O$–Yø oO~‚&`ÐE5—×΂ø~˜ÈTøfTfƒ–ç8Cv@„2¿Ð¦Š¨ÈL¡Xn,Šyê´ÑvŠÜ§¬LêDYC>ÏäÕ–¯Ê¢ ;3vEPlæÚB¾ÆGà®ö Àk«VH\IÐ vD'•S»6™$PžÃM@­´ø؉é9³€ÑZS\þân$­‹|ÔRQõ9ø$Ýsóqå7V›Z;ˆÉ`Ƈúi'äfõ‹h‚Jžÿ9WJ¾ß “¼ë½‡ª\ï  8l?úõ’°‘êäç süA&j¾òú| ¹kJnó–ã¬YÞ™4ÏÆ´#Ë<Åñ{…G³Þ·Îä š|;뒣ɥZfÞ*Ýîþôs F˜=¡ ? @Ó5(çðT)CG¥R’·×f§âÀ m3–&ѷȱA`-ÑêýüíÐlßÞ …ýrƃ¥i®D;µ¸ x½¹÷ę̂Ì1òÎÉÉV—cUHâ“ò–žh¦BÖÜÛñ ÛYÜHúP¥ÀÏ£ë8FW„:¼}¼ŒDþO-a47 ~s ;—Çüª˜þÜ$@ïËÂ}uäAXH.wÅÎ%'Ѳ 3]HÐøáL]Ìí%kád/„È9ekiµE‘Èý74,U|Šc¢è}z-tïµî‚Ø!ÐòŒÑL™Š¼Jlµ…Š{VÍ{B¯<› gøéà«"|FçOQ£%¶ÎÞŠR£ë䈯WÀßz3H2x‡#$-‰xýÒPF~¢‚­?¡6¦»Aâ„•VaZ®*7~Ñ»ý¬cõ>„°¸Ç¥ m‰qVÐëV¹b|ö öÒü—ÿ°1ÔLR5‹ÞxE¦Iè´?=ô–ÃEw+t¶†¡9uŒ¢Û¾)Uzm¥úJÏ[!²YÚFùNYÙO[§‹ Cˬ7 L÷üo’cŽð ï( EQ¦^ìÅ8"ý´ÇE—º…d5¹ëQä¬Å?õšÀªÀ»¶Pf*1þuh¸PÍ¿å&î ÞºT«ø2åccg‘½±²²²Áñö²è¯oøÈÏ@± ¼B­ï@_ß¡” ¿'^+|*AØ ÿ¢+,‹¨¨ê¬ŒÄôã!(´ ÂûJ uã„}Žù  t²#ÍÛb=üvGïAIËe–ÔƒµêIeCTIªb”6y\q=Š‚¼ËNÙOÜp `(ZÉ.å¨CXà©°©³t/vVª+Èi0r2)Àä‚—'d²˜fݳV”ŠÖ)îšóâ_ ÏVûDo;Ð.|öt& à"kV‚«d‹%*&ñ÷(Âq{šÏ>ÕSƒ"¦bÈ¥ý®1¼é} Û?÷¦CR&•jDÖ½ ?‡ÛÕ+*ã½Mˆ0Éou—|EX k’+2meø–jwS?L$iN­Z„€oÜ`›!Ò!ºWNÍñ ;m#{XõÕ/²–ØrYfª&2îÅ°N Ö[‚I 5¦Ð|ª³ ¨‡f¢hÁî’l‚ïH[€ìh&xˇ; 8&[ÀŒ4N1Xm]7¹t^Â= ÎÇxÞñ.À2rœz6¯!Wšåó¹Ã,_Ú– ) ±·qw?qa}ݘrK…Oi¢oàœ$pQÀ lcp^‚ëºaFJàªTËGŸˆÞ¥ ¹ßþ™µÈÖâñ9)?cN‰†þp†!èáÏy2|÷z0kn {ïÙ‰»ö5ñ.ƒ-j¹þ¹À·ì  #p#EzžžGa‹Ùâ~*²"‘Ó†åÝ…â9&œ`?š^1Ax ¯†ÚÇþ4b@\ÓŠB,Xªg)jü¹»§Ùàµi¥$ æÜdDÉèl^œjú¯­ŒÄ…c“ +”ÀºäÂ1Єòè9_cóTdrcýÍ4U}dFÏ·–ØeMÌ£Þ¹­s££bû?ó)&ú¤Òl‘ͯè’¿3Ê©þYS+¦ÂÏ“ÉpŽ<eAöͽc°A§¨õalA~潊â}Þ†„О½“¬ãmjß·œ¯R2ýñ²lMÌEæN«*H4ª¬K¿·çxy<}«s¼cäЩ§{8_=ñ—ó¦^¥ÃéeÞÆB²î\†ñFK·ˆ-§D{^º¤ÝfUuà•ÈáèpÎ(¼zJ×[wYî€Ë_ š€ º³(p%/Ùz8ðÖPþµóSÕö„jɧÛn]ØöuÜ€G‹å)òÝ-Ù˳á¿^¶á/ûŠÃZaÒÞ’%©ô”H·95¡Ò¦ ”Ó%>’Ú=2C?\1óÀé ¬©;Ò:¬ÃÀJgo\že"¥°÷ûžèÜὃomªÓƒ›‚e¶‘“P4ÆȤ Åþ{fYË2b…»žvÎ~Ôåz}Y£™Jõ§Õ=¢~$k,l{EWæ%<ðÏ*Ç¥XÎïòØÐï*Jk=ÐéÚÀ%¢ˆWÄVÿÔÉ>þ5¡Ì:ý’‡iÞ½êÃ(¹Ì7§2à}›ônbå»´ ü½þÅŠ¥££2û>·‚4ƒ½Ø¨v÷EsÙ×_4â‘×epýEßAvƤÀ¦çû&Áçx,Æm-ËÔû'é]°¸x}—@n[%@ËõDÃ+5n"ßO¥$ým×¢¢BÛ”ü"òU°0ç-ܲÃÅtú&ökvµ•Ä[gyß_„@Kg¼*íZÁ Ÿi‘‹úŽŒ…ƒ»Q$Ûl]\f'í)O>ïXª#½—ÚwßCVØߣx@ré!F§så×ëĪÍÙ¸65*è.KÄw]ä<I'DÏmÑ¿ì²R(±yl0ZîäSg°i¨®u6‰†«ÞÏ?:ÏÓ*Íc­^EFÖd 3­lS/ž]M”‘À1rhP þ¤¶ðv¶/šèØqŸ†b#ÏB·LE.‘¥ÿ|Þ³.í=öˆñß|ÍÅÏÕžkho¢(?³?8Àïq½ëJ;÷“¤ ¾æØ'ð\”Ÿìó?Kæ¥S¯^x^íçµîlªØÚ‡×v.Y_ÙˆP”Á‹óÄ6¬º‘4¶·ï²²¶F„JqëZÔ8ªòªz< èM¸&]Ã+¢V6‹U¢8¥p“5DÝ(_pÕü”ßrrõðàà8*?c“ÊO’ %Óùƒ|4Û"¥AÖ­âfÏrr¸j‡9å5yðÈr1`6‡g×OÁÙ3§PU§ôvŽ£µQu¶2i•Q,S˳EŸ‰Ñp#ÅÁЃDlÒ Þ„©/Ú&JÆbælö‹*A •ÝTUk£ø}ã1j<«á©F B—œäì2;n&aܽMI5u2>|â°ìVwöa9ÐñßkU÷Ui$fÿÄ^ïz endstream endobj 275 0 obj << /Length1 721 /Length2 16080 /Length3 0 /Length 16641 /Filter /FlateDecode >> stream xÚl¹cfͺ-ÚeÛ~˶mÛ¶mÛvuÙ¶mÛv»lû~{­½Ö¹çÆùçaäxFŽÈ˜IJ(foç¢âé`ÊHËHÇÀPS’e`c`0Ð1C“’ ;™ºXÚÛ‰º˜rÔMMʦF&#4)@ØÞÁÓÉÒÜÂ@aLù¯ @ÍÆÐÄÒÖÒ  focïfilàqsspsv¥srå£û§IÙÔàba 0³´1Ë+hJʉ(ÄåTâ¦v¦N†6W#Kc€Œ¥±©³)%ÀÌÞ `óo`logbù?˜œéþÀÎÍÔÉå`fNö¶YQA1y9€¨0½Š0ÀÐÎ #ñO»‹3×?Õ¦ÆN¶ÿ3 Àå¿–á-£ÿZÿÍ:yüÛ¢ƒfd˜X»ŒLÍ-í éÿ‡7I;3{Û¿Ã&®ÿIýÈù|Š8¤˜˜šýSíjc#ghk  ¶·upu1uÈÚ›˜:Ù”ìm íþSehkiãùÿ[÷¿’.†ÿð hgþs ÿY:‹Yz˜š(XºüCµ™¡Í?ý+®núï}‘55±tµýw¿é?|ü³£&öv6žÿ»ä?cükAz M - êÿ à_iQ;c{K;s€²Ë?\:™ü7𯴂¡å¿Õó0ÿQ€ñÿø²†.N–mºÿ‘Ãÿ|ÿ±tÿO•½‡7-#€–™‰ÀÈÎÈ àdañý¿à»:9™Ú¹ü‹ðfùÿ/!™šz˜Co¬Ús‡X¥µ†Uø‰ÍW‚QÏϨR¯2­ÍôÞh#Y[˜è]aß«`ç»ÃâÍñî]þ‘¼o!2Œ7…ý;5C¡»ŠýZB\iîW»ó®XðÖ¡w ª†ek,zyç’±¹¤@5Í’•{^+åXÆ‹$ÎcPÞ9†¸›°¾ôh'N©KŽœ©õ kz@]ƒ8oZxmÄÿº¿ž¼{CŠ¥oQXèp®yoþÍ £þÇÈ®ÏU¦wª¾½ZH ò!(€c¡’Ë=UD¼ôkcaeur³=cܬ@ù‹¨áõTY¥Üòë÷Eåž±­1Íéx¡¸¡í9â'íD]>[»Â–(¬ÞùÁ£¼zé¼-›ëN.\fxi¡\$:çqÄ5oŠúì™ý¶rÇÈ°P𥢒6z;‹ß<·${X†Íî .Õ_½lvߎgpÙfÚÃ(¬îA†mËêOŒÀeøøÒ[…ÀõWÆw8¤4_â³³òÄ%êT9Ñ“¢Ü¹3é‡#í¸p’š¾@OÊ)Móx£Ú ´ÉDçí-r«PöKþ5ìËdÊßØ= ½¾~K}­çALašQÔŠ„Ò§};±f"ç]KŸÙah4Jo:–˜b&ÝÞnÄ¿ÖŽž§wáŒuBØU3¥¼vÔš±xY˜Ós·#MŒvMÔ)/œÃþ,u6R“…7!›%ƒ¹'žºÖ—‘Òr•3!$!UøÅùõ >Σ*ã|“µ‘ñ»~0K,Aî¯ë´7Ç–õ]¾è¡Ó&ë…ç+Ðjõv7àW}wdëÚ±,ö‘}d^ Wî85ؾvi„œÔL›"¨\I¼q8†Hk:Ôd÷cZ¼­ÐËO²¯¯³@z“ÍšHx”I®D’>^blD»[UÞad0ú5¸a%œqv—“u-ÕO:8s§íkÿ„½…Ý/o^³uÒ¹Ëb>£ÊŒßi ¤}‚FCñ_V4jœŒ.&3 îl–¸èP!sêƬ!#t¹ITDvÄ(¶îÞ%D!y²Ü!‰üÂÜ9‰š]5(Ó€4³Aèö Œó4ª’rHîÕ/›ÒxäJ¸ÂàV‡É^ôÝT:òudÝôM["â,ƒfgX$àS+CÝ{ Õ²:¾½Žã¶×ÈQCi:ÙŒë“ÄIk îuíor¼Á+w%EûPRmŠÞ+o2¸ý=$ÂÅfÞg5.E6u2Ë5ãw@¸œÈTk–@0¬Èé3‹¦u!UÚLØÄ%#P6Â-= ¸ôM LN˜cNvZ«^òõZ;®ˆÂÉ`WTí[¾b qKÇkòïͲÝ^º‚ø‰IÏAGÖåÚ¾œ:»ÇèÐίéÞK9ßø »§©güõ¤¹Ø›=Br̈ÐI çSÊ Á=˧oÖ¿~ì ‘Üù'”Ýœññ´¤»W2Ó] )Ë°»ä™Çéšò9jågm\ï’<¢_ÝãÜN"1¢<¡G8ÅÍnÖ²–¬J|ß#â“7“÷݆±¯©%…Œ Î!P0ùÂSß›œwP–ü—Ñb3R>Dõ žó1Ç…­¼DJÿ,ñdX·©ö¶%gÀ¥e¦\­ º-º&Ë\u×l(–é± »,ýM‹8<äörKžÿ4E‡†Ó®ôü€s»1ÞMGèAñxØ·?úŠÜŒÖRr?¢ÇŒXÙ[ÿuû»;Ô`è²ú™[ïs¥[y$&ö ì6Âÿ]S“¹Æ»Z¼Æ×¹ÁÎp'rØoÇ£÷§v‹¤A+š›/¥mD¾ÑY1Ï8rDJ¬òÜš–×aÄ-º‰•2œ*ÒãoUå[ý~@ œ®\˜ÞÒOŠ<)™É¨7Ô A9½è\Xp§‘ó û9ÃŒøK8Ò}vÂt³ t¬L?N"M¼¼*¾–I¶Œ§öw½ BÞ.‰ZA9ÍçÙ×aìà._okéòÉÄΕObäå£Zë«Ñb˜ï)ó&&4QÂxñ‚”}4“AÄ‚•­j¶šÍF“[*ÉÅÔ¸±îlž0ÒJY/Îpk°3á:~BûØÆ/,ž+;ôôwxä«Œ„Wƒï¡kBÔØXU²ªQ¨Ü3R‘sB'ûÂØÌÙõ‘³Îï\¼êTIŽÈLHõ}&½€9_u.ž"CÏàü mrÖÂà²Õ؇&(H1k‡Û·épavbeŽ’1Ðå '½DÙ·/FÆyòM/BôdÆ“¨ÝЄÁ=YÅ"<Ñ)ÔGÕ"§yÐ+¦ðçœiñx®¢ ½©q}sø²‹R>ÃÔX“$§…¿¡RÍ >·™=‚g>—Êy}¬uŽƒoLGk¡®¨Ñûó #:-…’¹¢ÔYc´¬¨_s³œðë¨ +â(þ’:³;;AŠèø8Ò DnÊÇæaK¬Áð¦vý•_S3 2y±qÛ|wqx-5À2èÁlÞ¶Á·»˜_†ùüçèsgwìY[G‘óŠ§&ÝÌ<‰<¢áÈÝ 1J~|2)Ö=©{Wº“Å6÷ËŠ'%xz¾þ Úó‡_¡ÃQHN•ô¦ÉÅ_2m"ñ ÃKc»…ÌïA퇽"Åàfûg[êø„™ú¨”ÃÅÊ•<)ÉTëjSÅòµ©º¨QÂw"€PãVäA¯‡Ux#ˆt°ï±e@F Ãþ7,dåŽp¡´ÞÄ.ëÛ0éŠpØýurêFÀFD÷ú0‰‘pòúhÏÜ)²¸‡«*¢ýûsÛ¸a‚™s9Ÿ¶½³ålaŸÉÓØáü¹ÿeÔŸ»û™Â;sÜb3mŠVÁÒkk\)¿À¥–dxkºdr®‰«|Á8µÖÖXÎk…»^Î}ÈTqj3PJ¹Ž…Ÿn)Ð,ú›ó?<«(™`ã-9Së£é8åŽUhÀëpeÌ¿»ËÙbäà¤*µð’õÚu¹UÍ胉ŠÕ€–dZ£F£C.”ËŽ%“{ÄOÉ(†f-ÙEBpq1ÅŠÂóÍo G.üMû8TíW'ª˜8ÁŪ—ʧ¬“Ôl?›ºØݹïCyÚšic!ã÷bØïüµ/lCp]1¯&úác~…ä[-F–€…·!Qš•êÄS†a²~(üöÀ-U:Ãøêþ”?[€øL„+gnÌ#j0)íøäÖѤ¬b9±¾¯¬øjÂ!C¤Öº)ª+É]ððå•Pg¤r¨½K¶ÚL¾g|1¹ ±ä ?¨Ü¡"]Ô[·ïØ¢Íødþ¸Ž#رš ŒJ’]×àEU¶GrYæãV´©WÅ«Ñ÷>Kº!<ìebä!úZù¤|tH<ëÖ õÚ£÷ÐôÓsnÑá6";ÕAþq‡Ÿ©îK§&2I(Ïz“²¡þº#“˜Ò†nòeÈÓ„ “qøx4ž‘©Êìmå·Ñâ»›VNïN§ÕVX/TÚ¿æϪ²Qbƒ‡7©Ö†lb¤)_¯êv’ Äoåþe$¸?`FE¤õýbÚ©wÊ}ÉíŒhxa·Kÿ‚vTÜOɨ½¹æèm ?à¶È]>º.QÖ[u8`Яl…bKXž3~OJ­ .N®2éb"Hriy2âœ#B¾Ø‚"b|ÁYÞ†õebb ŸÉœL} 7×õÑQ såYòÃü租¿[ÐPݹ¤#ÚïQ ºÛP~¥´)#¨uœyt®Ÿ3j&©/­ä2áè¼ïÃgÙ ‰qa¦mìÝX27Àæù€c*¿ïp Za`ü%J%ínš²ÉVq†¯L;FF5ÚžBl Ó?Ù1†u‹†!g½™+Q;È\È'ŠB :’¡0àîMd'8Æ/J:mÑÍÓ`%ÁëÙaßþ ßá¬4x`6‡  ,³c«]R+!ÀqØ«iêê·ï0ŒçµH´ô¤É»³Ú@o£DåœÐš7,èË(ŸªaJã ÁÊuøMª œ­ù¬|Ó´ÙËÈ¡¥Ë…ìµ\OŠ´]¯«ßî4{§¹$‘Ê5Œ7’mM|k‘p ±Í´òr‘<âÅ$㺠Ñ×ã-|¿Ô%|QkŒ÷÷ÂN õ€—N¦V^i°,$¡¡âzzš`¦ÜHÀ$ßR³ÎÐt¾I¿ý»e$Æ“Õ—¢ÓÕ^sÀ¼ù]{£ˆÇΉÜuQ£°¿±Ðõ£ûo¦&ܹyݳ½™ÕÜõãz¯g“Ò#Ðéã:—ôãÇ‚‰Æ´l ÌæxiQ ùàè2Ê‚&‹áà_V`3²D(#Bžt k€h‰Àþbè_ †ÊõßÜ‹Ÿê6ˆ˜‚2ÏÌAð‹Nr%6 =3m— ‘Àæw}ׄ¶w&:©´Ë‚•ü &Úà-ù A¤'ïft9h¿KÒ¿u I4ï@‡‰AÂFë°z fÓ”ìÞT@媫¾|„à[.y†­‡ñô),êžÆòÆKôim4dB&8C´‰õ»0A¿#ö˜'›¦Ž 3)ð1X±æo[þ篃‹&SjCQþ­OÓ ìkòÝf…ÝæÞ­V ¨Zé)äÔ5GU½ÝSЦ¼…~|öê±FÖæEëž-¯`]æ³C§þÖçÈ\… [~¦YYÒe»š¤¦«ùú™I8g<ü6eSÞĶNdù-p§PZ}À“Òé˜0(œJòý!ì·Ëu¢U*zÿ Ju:~NIò77Š?ßf@§7%VâþØ}Ù%«êƒÖ­ôA¦äñm`WÝ:߯ŸüpÏJö\&áÑas–Ÿl²,›'…ÔRÆÒ¡ÖÈ®æ;úÌUŸè§ ^2·|#áFAöQ{ª&Mp&$ö´éQMîÃœ+†VÖ¨íP§‰Œ¿œ×$\†jwr;*“…0Uój\4qst±˜ö¼‚) B›h.Ò}ÙË2’â¢^_äÏ˦ôáEþÕB°^\Њ• e}­7 ©[U*tDÖ¥Ô×Ox¾ê-4eDnCój`¾ÓjÆ8ãvÚô{9ê0<ÕèÌ ùÉåþò œ}°za¶z‹è?$›3¢Ì„QžˆT¾âûº´²5jö|¶Wiª`µ'yQWǼ.q™§ —ϱֲ›$»ujcÇŽË_o‘µ¤y5°  ™× µkmOºËŒÄPÛÌëT|œŒ"+Öû[Ü©ô?îûÙ‚–7Ö™n"þÊ?)W#A ½lÊÝxšP¥mÜ+Z$í±_`qÕö`ùÓšjGÖã„ö®ðŒÙÀÖ,:Û®;B>âèÂÎV—€]gþÖtW¤³ú÷ëÀ£šú-Q¡‚& Šd¹„jßÞe˜ßÖC·;3ÌhÀä ÏX‡Ü“BDs•-„ˆÞÙ¦.5lÏ™z´6ÖöçiÞ'«žýyBòÏ5s€Tó˜!€Aj|t*cåÐb«ñÍü«©š›ô»ºq~lwÞ¤Y¿‘7»Aº~m?·çz!¥·DR¸’†kO™5âà¹î'[ðà„lµÐ,”¤‰œ£L£€É€èÂüpHï}zͧ£œæ£e™Ùåï‰)¦é®”j/éë…äÀb›b­l{µ,LQïc¤¸ØÁ+v•‡ Þ¨Z:ÑiÍ F«‡(Þ”õ·Ò[‘éá$Cƒv%&ò)êM‡5  »s)ݯ}âÏ) Ð£f‹  ãÞÅïêã 1O?NCõ™PuÏ‘ä“pÓÍ&Š>Xª6$¾­NÌD×ïõõ&ª°¬hÞ;Hh5J¼¥„'C§†ImLà6*kfe",„@£cçR?œ–øŒÂJÐNóµ1âlALll¡£šuÏ8‰$ûõ‹U(Ù5$ÿ—ŸG«û¨0y%bj™éóò¦ï°’è*d[…m‹°ró3BðWÉ‹âì*%ÒõvªàP¬’ sfÙÄÅE5MÃÁ²7]ÂÌÃ: ɽ÷ó‡¨k‰×• º<€ P:¸êoË|dêØÜ#6MoN‡äYÞ½>•>ÉM“é qÂ,#ÝrØ'|ÝäÑýÚÍ$?–;µÄ_D*„ ¯5ãc)VåÃ.nÉzé”9·ÎŽ¼—±WÊ="ƒã(§”_k@&µÉëRX)zN¤ÀÌï<?tl ›aHÞ³»C¢nÝ7ý $•£®úÓAUüoïÚ´+u6jó¸ íHçW‰?½›‡S=O¶È¾‘PÐd™è yÜ0¯N¼P}8¸$•íãÚ'+Fë>û$‘gƒkôù z®îâþ}U&¹V² -02™å_^u+¿¢–CzüVö±øÔc]œ´Þì*“OõŽ?º†•3ãy¼ö„¨¢¼ERòë*¹Ø"q=pÇ99䕇î^å÷é€YÞ„¡èÌo3ý¹:ºÉ{‹œ@=õ&É×›åĦ® ZY5LÅ4'SmŠVƒË*‰Ÿ¸æaVë³é©ÛjÊì×[{ 0&]sÖ.ÒH”Ôãì=¨w¸’u±VÇ@qµ{ ¥c]ú w`×S rg Y †"Úa©…§Â³ÑŠŸ¸íhöeäx1&BOîÂ4^]´E´©n*J›P¸ØC6¿ÕÚ|<ÈöˆÖ ”/ÑÈu@¡ó+¹c;i·õûQ•ƒõÅ.ÕR»•q >NÀn ø=ÍžWþºYg(Ñ 3íäŸvk u†™ÒÐX›JH/OüürùyOçÆ÷Pp œÝO±%K3(äŸKWþ«ÒU³ä. Û8ágg T(¯ÍÞ{}eƓθòg¥K{Œµ¼Úftΰ7¹…ñîÀGkk-ä“$l}€ urPfŠáVè˜èûfbY«¤fRþ%ÀöC7-Éb嵪u¸NŠB0rñ>c…šˆrGØP¬1+ö•fú[×µ»·º–KðAÅW;ŽSØ6Ì…TFA‡[¸‡èÁï×€Ykû8pé¢Wºé·„Þf"Ù/ž2þ4>xP2•@$ çŽZõ(²Ä(¸ñ[ ^kYdÜžóÝyéo³Ô>Y‰—‡žår‰ŽÌ"§÷u†]•Ã—ôOæ #ɤ¯ äRü‰„?&‹¯pæ¤ä/e\YsÂV±‚­ãŽMwýUäzø{öfµ˜’!†âYu;Ò]fŒ È¿«Dn…•IÆ€ËwVõ‚(nˆÆ%¥ jU‡†É€`:+bz~i†¹\‚ìÒ*CQ¹HÂõî]zj$i°Œ£Ææa×5WóX ²ú-s®¾Z‘3Øhz …’³‹.3‡…í{NoÚQ ¡`æ(. ?±z0çÏETO8äx´}‘ùy20¢ŠoD –%Åœ°' ££”Ó_œi¾&\7< ‰ƒ¤^ÕêœxÅ‘­·]¥{ÛÞÃÂò+þÎEZKºÖÞ}ûu®ŠS¼Ü+¹aÅŸ{Gæà4ÂT)êݽ†ýø€ç.s‹÷Ô:i¤H~ý,„$¡ÔWÕ‘÷È}—ó¨VpÉíÑ6…*Aa2nà>Y ²R[‚} ႸáåTÛ‰ˆp7hr³Iš†¨_8ï£e„fä°Da«ñ^ 9®€ r¡åЉ´ Ñ1– ï8B’{áGš£Õoü¶,T´|#}nuö2ad¿†ï$»L¡Z•ðWÏ©b;S{Îë?A}öÀ€nc#ñgsÞ•ùìÆ ÉÃô(¿M5…Yä]–̉8«3ÓðE`ÆÂóEBÇEI:Ù¢3ESôæ`Ú¾þÃ0"äÄ.)2³“óŒ?4®ÝÜjî"°$ŒÑÄèÀõ~“WDFã¨5Á┪V)>×7àí,uR:2M3›t`Ë¡Ù®°»X›®ï‘æ{H¾‡_ù×Nï¡ŸÑ8±Ë˜ª®\IB*þ£¿‡‘\Üòè6›»ë†ž×:Ëgg½áè™ÝÕEªJ”H%Ýá|†ð™%3Íö¶ocÚ ÁŒV €e‡'û<¬|@2lS­}ÍÓBžÙÃBhj[)¥d,ˆÖM^Oþvφ  €øîº&Ù`Û˜x£"[M÷š¾\Ùë±xÆRŸa:ÉC}}@rcçÚlÖnÙx `‘‡¤æò›-l]hýTïúúEØü¾®.ŒtŸLÆ&=ƒsæá6v³‘°S’S)— ÊyJž†äY®v“(™ŸR­=͈õííb`µ”Î[=’£«B •Fs3éAëÚ=KµtæʬEg“ƒSx„R¡ÌêØ9ž"V‘wûK¨98^à yÂ)0DŽ:憿TbéÃwÛTÜ¢g Üñ†Ã_AJÚ‡ ¦N1˜{M‚¨…rG"ñuqe^¥%jÒ§s%0Îacî‡(c††w€!¥ˆ-ò¸3|À¯@V¥&íIvÝDÂr©y¨@iÁ§8æGÔ/y@hX¿Vƒtý{–fHib Œê¬ŸêXíœeê£}›úâR% ·† ¿”I.‚ÞvI‚Ì®£GK'öTšÓå¼½âkÇïàëjó}ëIˆñßíæ: 9"÷ú2ª¢Øù×[ÂV(ˆöp¬•¶¯YÈþU7öó¼§2÷U™–KUÁW‚ÌñÀ_4"¡+öÇ$áüß _aúO^i_‘èf>”òI åo]á+Ÿ0g„¯Õªû{6uý0f!àèiVâM„k$Èd“Ír ¦:} -²Æ±Ÿä–tÓû} U]HÅ{(íÙ>Tl·ö¼÷‘µÀ¥2|7Çs•‚FðÍ &Ä4÷psš?Ÿ¢¼4Ðx/A€¬õ‡ŒÈ nXõ¿“+ÀH#ð9È7ñHÉñœØHž{\G(ù$9ýøx±ß¡Ô)ÚH!ØœÔàðoÅ1äF6Uê?Kü1ºãÑ9bäýÃ^Ýl¥UL VÆNÎw-»ª{•k]¼b¹û‰Öc.‹§YHÕ¥u—~að”Î`ÙJãü Vz4Dÿ‘áÏEgf}krÊžZp½ýuV¬ÚdˆI€ÕøML† â4úÊiîÞ³šœ¨}J™ï:IzñË ñêD9>JYRÿ´0ͧٚáäN…ŠÈ1VîbÕõ7±Ä¦×S$ê“0Êœ ¹Å”[U ïçM%åÖf[íMQ±ÏÑ?·¹3áY»¤(n£R‹ûPajÝWª ÉåÚ‚Ÿßp7‚Åv9t趬lfŒô²´’«†±0oW|¾ÃP(£C~øA ¾™GâÕ«Zq“ 2fÀ´Ÿ&01’&ÖŽŒÉ“oj>‰ß±Ò1ò/ »…»yH¨ô`]ŒKÔ¼Ù«&Ù©žÁ§Ò&ó¼b¤ x#Ä“vi  ò…º jŸ4môè,q¼zzã±wäó¤6”醰®¿#Y%˽?™ÜQõ9ŽòíŽÝb÷“¾qǪ“ãÖYŦ¹’]«Yþ¨Z©¸+ÔÍ?߈ø[=6åmܾ^Båò"{N‡‘2‰¯ÁÐðÀVY%u¤JЊ;Ý$tOUBÃÏSàœªhÞA¸?ÐóO0Š¤cƒ&9£ EÚ¶Ù2{G–n¸{Æb£‚Ëਠ+Qrƒcüs Ì˦ÙÚÀRªÅ=x £#·ÌÝúñXý‹P²¥|L¥º§ËSŸø°ö™0›¯ðꉌäêª}f#üX?ɯ>؉|€*ìõï É/6éØw"ù´9ùv|ÃÓMm÷ÞòÊ1 mäK_‚;ÙQ˜=_Ñ7 ô#3!ðø;¡[o°–žö`Úá)Q-ˆ@ 2ÞéA– /C°åše _~ì1øÚ z‡åY–Pi@Í¥â™÷%`‹ýÊæ›ñbìFaòã¾— Œð³š‚ijKx(Ÿ²+ÔýÛÖ |à©Ý9vùw[RŒ™_Yu¬ç[{«~ª:Þ[)’ƒ¯Èn"hŒ~½pc—µ@üálÖp.Jf¬·u“VšFàáj=¦Kj­8TÓ= ƒ?ò˜y†ß+bæÁ3¥õÓ Y<§˜ÅDè[L|S}*—.¼¥ÃCRN#©fÔQ‚ñÔA¯úœVï‚¥àÍ)ˆdñg÷½ª¾>ë”¤ä» $Ú¨Àˆ}.Õÿš:ÈçÞÔÝ0cÖ‘2÷ú!h¥F&ç¸NPÏüdjú]¢„LÏmÖϾºË¹¼òçmËÃÖÏ…G¶ªÉ>¢(~~®?ÌÝÀŽæRo Ù*Eþ|‘ß©ØÑ‚`gcù1¹„”wSPƒÙë6TÜ©£¡gfö·L£+³K‡­;Õ÷-£“ƒÎiöï–«R~OhÆéGŽ‹³ÌÃIÀrÍàœëÈUîÃs·RÎîðyçêJ&K1ƒK‡±—¿K‚Šmê•Eî ñõ‰X¼Oo‹dù§ÁÙÎÚÈXî|I] X¿&¡+UXLn¦8ã´Z§WJM§¹PØW&’’‘‡] Yú«ƒÚÕúÀ;[Þr´ðgõRFž?ûP"xñé½æ^Ðl˜Ê;M<5´ÎCs€D·UžN4T^äT…úÀ°E'™8%Øا€]—†t7ƒ{#ä¾O¾ï1„˜-tU49@ˆ¬ Ó ßèœà8dIåcÆŽ’ãer½Š;û£g„›ö~ ªþƒÝ%P§fƒLŽò½;ÏÒÿoè(Sv“Ð/Âcì™k`-¦ ŸoÍXZü¶ÅÛnqXzH€ùÏDy™z‚Ìò5³#¦UÁÕ8»ÌÓÏèàvu’ÄóÄÍo[XògñV±ß*[–•à'w­x¶BÞlÓÙÜáú¾Ä|=ÃÞUjæ+„ K èÂ`nèyòÏ_N¥®‹q‡WŽ?—_@ŽdMÛ4ÞOªëúKA{ê…ý¡, ©¾¿„}~9ÖX6^ˆW¶'ë«Sœ®µ^¡=Ú™y f»B8âm[IL•œµ âû%’µV*º¹\ ¯Ã»¾H.g“g}!/qÿv¹ü™¤Ñ7œÜìÝ7Kœ_wA›Klos•½r·|Ú,¨5ôf4t„rEq$uÇM0[ÞzÉZv0ÜŸ½f–ƒýÞûHGêƒÔ˜à×V~'Oq ·6ˆlÿ@¢†{->Œ\"£eÉ,üE^9ù‚¹äéÙÒC|ÊÕ`êÕ¹ã—Ýèâ(/çpγkÈ“é3ÙƒlgD O["•oçtqØÑ´R›Ô)d¾/MZðÀjÝ~ÃyýôÕŒ¡‰|fù-Þ¦R5-‘Ó/PwÔçg—÷nÚ‚9j±‚hÖâŽv$5"-:C>2ä’¥ÍQé»ÂtŸ53†¶‘eù Ð…¯údÆ YÍ8ò *h´ŽöCêU±ÝRi«òÿ9Çø·†•@gäÊ;S̈…ÌDàÌyDÓýåzQ óñÀ„ ÿÞƒ¾P|¹ÑMéùqÿm¾)“K´ß71++FÜw¢“ÁÙ-Ÿ …ä2U_sÕƒ+ÀÈ0£1ìùÙÕ¥³¢'­Ÿ‰V¸~¡§\žIâž~þ¾ÅfÍ0ÇÅûú–Õ&©z4+$Š_ìw¦â›I2§Ž‡•5^%ÌâÊ êõž9ÎP½B^׈º;ä¦r_!W?Ó÷vÛÒ¹û á—ª/7~qf³tŠ©ªwFNPž%7¥ 1Ãsà™D!T-‘†”¾öú^Uôíq%óeþÙ—”Éà1f8Ð÷ùTÎwz"XÇÒËN¾§8SIOl{ ýoÖÖÎðÃ<ïŠRŽ¡~ö°rÏh‡ˆ¨V¸æ¥x]RV±£Š+½ÖqÔ—³I±[À)-å÷ÈW`ÿì+¸›á £ý8D:>7~[¤=U®úµîïÃ/áGý»Tö̺Ð`UvòLôc›Çô r\õ$ŠBI•‰K H• œºgeø«ŽÕ Úð/%«'Z(µ@Ùce›8ƒnZ1fAJ±^C”Olßz,Þ£™ÓõuL!Êö°ÐúøE׌·+»(].J¤1§‡%ªIb§iöCÔR&H”-‡ƒø<æ/vð$oNÒBˆ¾õjèoÈõmìâß³îü¶˜šd.£’´ü0âha â¶NoÉÜ9rCè__˜õ„[-÷ ªÖl/é&å1XvÒµ‚N##µP¬˜5 ¿gµ¤6„éREK€(¼0mð®uÝÁ«n„å5ô £ÐMG™à?„0ÓÚËwU»'βé|¸öD\»øàÜxŸ]gª‡õPœ‰IŠ…cí ?Èâ¶è¥ N®m¤C€Ø¦Ž{XŽ=g×.¬CçTö–WEéšâ¾ßÙkl"QI²¶UÜhplɤÝòívp©–ŽðxÅÃ]²‘€Œ”Ç É×'é~5L#C›ŒSÂÆ`³È‡"$ÿaWjw¥îs½‹ðÂY¿Ì*«™¾«óÙ†j¶ÍŒ°¿o—‘ñœwã_r,“æÓÇëL',¯9Q<¿æ¯¯„yk³$ÿko£ˆ4$÷x6©µ— ±b‘TY¤²hù6‘æ±½\í7Eщ@¤¾Ç™ÏgÆðˆœ‹s`^T€xn®<£ôYWze=ΙîP…z|ˆ7à!„QùüÃK]Ñýi'°ÍY(?ÿ€±PUÚ´“ãÝηäŽE Z. *Nc€}J¢§ 44pÙ°ñ{{ÅQm|ù,„Š”¢JóøÜ‚l{Z­ÜœäɧR'sÖ3/z¶˜ÄÌ•ÿSv. ÇcjÇ2rR÷p'EsÅÞõgUOlèÀÓ†ïô…5æäG'¦â=¨9º;Rvc—BÌçž «j?3ý¬óŽ7 y•”@Äu¿¯¶AIšº÷åtª¿ÞþðHÈÊôMÑ‘¾‹LÂ%e¥É¹ûÇãj4u­+ ¬ú«¨I¨!M—¡¹X+^úÒ.‰UnѨ«qG•ñ„õA'ÿAËtz•LÌLeņ蘎ڗ‚a~ÇöÎòˆk?­PY5­ŠÙÐ뒬窳Õ9a7Ir¤.VÚ׬LRRdk´Ç:>v¹(¹H¦¤)™”4 j>·–Y6ñEU~À2¸ê¹”uO‡ú“‹åÀ9ND:=)\í3ëïwx Îzl5sŒºvÚ䋆Tý @Á¦´4•x¿E‰b¾–‹•)åéɪDyùF'X(ý„O ¥‹¾rS8‡÷–yíPP,LvÉ<1bjW4T :††##êKŽl®8ýf$öðVþ Uõ ø‡E ôF@Ôüug÷ðRK²[µI­ã•¨h½ç8(Òžr õ~Èq¯k,]»1ù…ɲ3:D?›_*Ÿbú’Ï©vo6–4:O’E|Ü8†Â$‡ª*5¹ð7 ÆŒðO)9j’¾Îù4$U‹J|E¦EeZC?f4Å,¥åáCàâœrâŒ0D!y|?ßtüt@$²–âlM?i7Ý}ŽõÇœí,ømœÑªqí¶$1õ gT¦žc›º‰š¯´…eŠÐ{Ìó«š“ç¹Öi´óûÿëº>¨Ðô1w…|Œ€¢° ¬®¨R‹!PÞI,”F€˜áˆ«ï®_üGxá"FÛï#ÿu\á8ÖknïàRò3b))3ÃDòuÝÍFLÚŸlš®,$û´ûƒí‚pš÷ÉCÈÔ¿c7«¢6l>%²+µˆ`K»Æ=Ýø¿æ!TôÂþŒ‘÷ž£ÀnZí—˜cçð4¾ôž \åÆášçTàî ÅãY½Sh=… ‹R¼…yïâJ¸=iAßiòòš&õ;‡¿ß‹yYaHÎ%‘Ì—pÊ­cRÃNלiK¸6,¥™é» ÷X| û²—t»Xen£Ø¦³t½JqrèTÈç>‚iä¡aÿŽ¤„DÌãÞšTg~¦å?ª@¾qÿ+ g1cËÁ“|73Y¥Li/„Â4ÝjHšOgnÑÃb6¨5ñK››‰C|Ý4gÏlúÂOÉv?ÀDQ˜]& ÊÿE½£=˜Y^{Ó ´Ù1ýOI#)±ùÍÆ‹?D•þ³7ÄŒŸ«•Žà@n ­í²püªJÜe¬à8­\™Ânâḻ×.R‹­ŠÑ½rl¢!ã`%îbo6 1“Úk©ã¨ÝxS†ž<>µ÷µÞ-{&…m¥2ªE2Á@vyç^î|Ͳ*.j°úèTâ=5¼æþròqyx°ÇÝ(1ÕÞ6x}§ªUeM—xa`G‚-()™mÝ¡Vê;ÚüœQÞ7Ԗ΄­³Bz@4,!©ƒÀÃ*ðÈ!?²Gã/{6býáÙË+b‚Ü ’Rûn…Ãz¶£]lj¨•[~T3¥Œ?{ORÝÚSEx¹{†—¹ _€Ï+úyÂö)v1)ü6¢½"£cÒÙ£¡ß4+ÂÒwâT1|l?[~” ÙéàÇi :¬LdJÎÔà«ŸYÅÍSTÕÇy¾Z![<ö: ŸÃ7+Ñú~°ú3aBN?Ý9íÖ 8€Ž~ºØ™ŽQ+UeÃaÒUüžïK<ÇœK4Ö¾Á–d\q‰¢‹WW%úùíÜg_Ó@þ÷õ:–L ©›Ý^Tv@¥E;žè*uú1¤¯NƒÔ?¤/ÇŠê«bÂÚS•ZÄ\¡g+ ·a'Âw}ìèukÕ¿‡ˆ«fŠ°\JBºp­í¢!”ƒ¾ñ©”‹Y†>É[ý£?? ¡™_¡ï>Ê¢ž¡X@Œ3¹XÓ9Y'qÌ°ÓÏæÓS‰"‚÷3AÖvB-B²&øŸNy¶¾æ”Ü¥aãÑž>ý¢»W2ûAÇä~J.9"±°ÍB¾ðÊqL¹ûí‹Yÿv•ž“è¨_V¦ƒÚzðKŒµf²m;Ûë›E.ŸU$ÔýE* ökPÕ\îç6Ò$ i˜ù5Ot¯^þÁ¢½;Ã?%'+Í°â'ÂUBóRivmÍ/7ŠA\µlPdˆœÃ¥ÑS°æñV×0úä ¡tˆgO÷^€Ý’çÝH€[ìí0g‚Ûìœz%+º4Z´a&…BS³W:Q¹cµ‰3PCÿæ5#ª)¾d3èøôü*:PÉtîo$PIì+ÂÓSø ×8D}™áÜüc7c3 0Zh1íZ!/œê½ ØÝBŽŸéï€â9x$‰íòQOóÅWñ©Ú¢:/êVa 2KÁ æÿzYN.I:pôÜÒµô4iøœ<’«uv 3MäÜ8-8Ï÷Ÿp£‚š±9s¤] Í‘ïăóˆæTÕ¹ïbá$>p;†y/‹sÂ÷Jð)‹‘@ðÏO%äG>0á 4Ô‘ÆuH±ý™[Ê;³¤kg772x”Iíï©N€ ÿIõå9þQ«DYÙ’»"Ü—rŽ¢_€„jAÓ+m>URÌcs¼ïG‹_+~9'Ok£j“5@ôJ´`<åŠÎ¤ëÂC{ |THlÄ)?-­dö µ ºãÒ×Q"¢ßå~Á¥je×…Þ-]>ÎÞvå2F‰yZ¯9Bj’ZìÞŠ¾ö!.ÀÔ0]T´;à|ïg¼ä€>@TQ'@Bá\x»´!$ˆ>køãkyQJúüÙK”~DÛþÞ¾¦ùü+—Ïb!ÔïÎI7ä¤TìyÙz}ºº‹¢¥)ÍsC)s¼+bV&ÛÏâÀ‚ÜÜ\Ü !-sÒxIÁ0E“9_È3xG§†ë€H*•tÀz[!_ÕzuËÉöƒl¤Ìør”jl‘ñÌÔpÂú›X¬xÖ§½ôç9=(4gþ¯R§»f‡Šji?’Ž÷ ´9ºkì7q¹9\5+üF“ywo}.<åÿV©ÿÇçÁàˆ±8qÃèH½ž A+h~v‚WCé$:ñÆ:‹±””Ҕߣ"º{3y^ú©:Íá7Ôfð½ÏÙjÁ&æ Ôp¿ ÖÖ{§WÅÂÁiT ë endstream endobj 277 0 obj << /Length1 721 /Length2 13405 /Length3 0 /Length 13943 /Filter /FlateDecode >> stream xÚm¹cpfÝ-£c'=±mÛÛzbÛvGÛ¶m›[ÛÎíóž{έóÕWûÏÄXµÆs®]k×&#·³uVñ°7a¢c¢gä(‹+É2r02éY¾‘‘‰8š:[ØÙŠ:›pÔM€e{3€™‘‘é@ÄÎÞÃÑÂÌÜ@iLõO fm´°±p¨ÙYÛ¹Z›x]]]]\è]øéÿ.R618››L-¬M"ò šRrJ 9U€„‰­‰£¡5@ÁÅÈÚÂðÃÂØÄÖÉ„ `jç°þ·0¶³Zü‹“ý?l]Mÿ3u´³ÈŠ©‰ËË©ÄDTD†¶@ÀÉ¿Ëm¸ÿ¢MŒmþU-Àù¿–µá-£ÿZÿÍ:ºÿÛ¢ÿÆÄZ;ŒLÌ,l¿1üK7)[S;û¿Ã@ûÿ¤þrúË@ùWC*ÐÄô/ÚÅÚZÎÐÆ@)bgcïâlⵚ8Ú”ìl mÿƒ2´±°öøÿÅý_„”³á_„lÍþ*Çøï…“¸…» PÁÂù¯Ô¦†Ö%ú'®nòï¾Èš-\lþ½Þä¯; ´³µöø¿[þ-㟠Dee”Õ¥iþ;ÿ¤Ål퀶feç¿Z:ÿø'­`hñïéù™ÿL€éÿù²†ÎŽîmFúã¿žÿXºÿ%,lçîEÇÌÈ caf0±3s¸X™}þ‡®±‹££‰­ó?‚ÿ­å?þ?ƒdbânbümuÉΘ'Ø2¥9´ÌW¬`¶’fvÚ@•f‰ tyªûJÅbШpñýVå{®<þL€ßÎù¶Ôm±a¬ üáÄdQ¸îÒ÷ç"’r3ßêÍWż—ºp½K5c9ˆ«ˆç´µÕ@êIÖŒìÓji‡> ¦(ƒÒö>Hä­_+¿ïm%¨t)PÓµ>áMöhªgÍCò/žwWÁÂrƒvà¯ÈpôÍóËìÂ4ïØ([¹¶Úè®%IBhE7Œõø©T  Òoa.ºnóˆÛ\üI¦mÓÎÁk7ZÝÌ—Ç.ËF‹tkš¢Bü+úŽ,KÄ*EÐ|üª6úϜڞ“A­jøÆ«< R©Ó³ôŲ{(½ ƒ+ÁEA¡•û~šA<[ñ+ÞëÇ­|¨ÌøÉ•!Ÿ¥•[w%óÈé êu’žHö8R ³¯Z&ß.Ý$¯ŒÜMˆ¾ô/w}ü#†åȹGJ×ãçèe½ƒÛÙde=ëçéÖl½ }ƒ‹æ9ó=·µÖ¢~„Ÿã9ßAù>HÄ}تXH›°N¦©‚ncyÓcŽ`m)¯Ø ³ "QE/¬øz“2 {Sj[űºÛ.X0ýUSãl3ÒuGuiF»Õ¼¸5”1LµßœÁRÕ×H± ç>üµL#Ó­go Ú«.¦ìŠï° áüœ»­)cزCŸ½•££ò¦Á7û§pÈm±±öŠ[l)©ö 4ç A‰ŽBZÙ2q‘£¹J‚úν—t~&.M'œ75š¡ˆÁχŽþþX–¼3ñCc7.ù=0í®WQ6ÅÜe3ÃÑ@SÁíW9îÆù„ò‡iŽÕn5Î!û(sª§á;‚‡öÅã¼,ûœVKjwlZ@ýíG®ŒÝAËÔ¤•sœÈÿ¿ÛIaðÏ—èqX?ƒcëÛPRga¦l¾ÝC[P¢Y·QA~[·éçÔ>p‰Çî¡ bs{ _s3šû°^ç Ì? å6iªŽwdV™~‹æºà?¹HÁë‚T¤¶rçDú®QŠ¯Öõ…Ô¯Ìk•ØIg01jÙCÚê-q±Ë}—??©bÈýP)_Ó ¯0æ%ü¬À‡Â”ñ¾f«Áø¹þÙÃçΛU›z‡›%ëÎÖhÖj$¼&‰N!ÅEÏš²ç÷zÝ %Ç=’{¨“³Zª[_kV‹œÞ•tr&7s‘61tÀ­ávì¹y1 ßÆcüGÕ$•RE ne«*¤ Ûvܼpä ð»Œ:æ®S7ìÒmþ¨“ÀVßàu,ëÈjù{—²I.éj~¨xHJ µàÙLìÛ0úñU8÷‹»o÷M•y…OÔ4¢1>ærƒÆ¦À÷ ðàÚTŽL¡QªY/αUkbfs‰$=ÏÍÁ}Mõ¤ú®É©\˜_’rÙ;ó6@ükŒTrZG9óFá| 1ìf¥žk$$­½„<~F½QWHnµ•tÆ&õ@›ÓH=[‚.š·;ïà Ô[ooâE3/åqïªçw$RâOÁ?òè×KŠ:êöJéóõ…¿—~öxóüÑ÷Ò ÎV‘Â4'úvG³Þ(ÏûCl©¿ÿ†#$uÒ¾Ñz—ÿzu,êPe¯Ý´ì(B"ÇןÃ7“•Bˆ[pÀmµ§ ¯¾VO‘ ´„]ÊÑ oÚdXT—Úwð±Éž[Å-×ô&…?\ªƒ²‚ªÌ¡åákìˆ;]_Euiõq]Îá/W½3„KIƒº½Mž¦¬I¯µ„h€½L wÃ0ç ËI¤¾âoëP2´½©…+Tf(–^3Òßg¢EÎþsl~vYNÃ,ìîÞ6Aá6ÙñM/‡W,&Ñ«´ˆ˜;Î~0Ý*é‹ÕUL%Éû>F‚Ì›p7B|êS4ìú@¶ë`eÔ¸Ý×âôþö‡ÿ‹Î.ѸœØzK³ûÅÕf´FOìéàun°S*Ò Ž }0ˆÿ~Ã;²É«Póïùûž<žý¸úaÅnÂiäÈøMä lî;ÑÛç ÅJŠýŠŒÌ×Åœ Ë× î•}džÜÖyÙèö¬I‹Æ‹ØQ5t,¯IDG‹SH0¶?6^/ØÆÊ/•Š©‡b™™À!‰iÜa&‘;pb×üm&ëîúXC.¦­!'^åÛÃ8ö¦]uðüNò*q½(Ów~ßÏcý7uø)ĺҘäW2£c•,òöV¼TÓ> ê„iÔY|­¿˜cÆ€”B+×á­¯ë5ô#äÛê~·c÷?™^y]{pŒSÍ+Ý"òåè)XUw# ¼ §¯(4mê–ïâöåw?'Mñ`6• ¦äJÛMI„MEA–óiAaü½ˆ§ì4 ^ŠþüïyϽº˜¬¼i/$ˆ®¬PÌà1yá*»½žµM… Œp 11îÍÇ´¯NÃŒbf˜¶Ä8” ôHÅÑÔ¯" ‹J ÞkáhƒÕ¼vÎ]ïüJòÒ6ß ÕÑZD©˜sÔùXçÝ¥ Aa¤ŽJ.Ñqr$ZÕ/æ·Éxf=̘¥hB ¥´¡j°½lAÃ~ðɬ%NÈ÷pή]¥üäªJ7Ú¢mQ6ÐüŠc‘FMKd¾s¢÷”KV²ÓÛò\VÇ6¶Í]3ÖÇš¿Ëèâ«.Ö¹šÝ»ašàǼ¯úÄfìàþ -PÜ` Ñüü>Q•u§¬˜HŒx»‹Ÿäjâ%ø"˜‘±Œæõ¨ š¿m¸¯Qb[lµ wG פÞïÈ 7¯|ÕĶ·ôaÿˆ*»ü¶sHì»^žùº¡VQ:ÊÝØGàÜ«½±«Ñ²®T,ðR‡×‚¿iì’9/Ÿ¨T­þ{‡~lÓD3†Èu¤•WOˆ a1(çïhP=Ñ«YJMÙ¦ŽÙ´²)í(K¯Üi"·bGyêRŠÖ)`Ê‹aA¾~ëJ ¸ø° ¸ufñašýq¦TK7¾…mCÆç’øïÂêÌŒì,Él²Ç¹EÊt:Ú˜ØRZwÊ ’¤‡\Ù)ááxߧq'hÉ®òC5¾·–7w‹ã UöK@tj×Y‘ô³ÖÅ]cÞ˜Š*3^Ÿ&Ò}Â/å`¥Bâ…ŸSiZaߧWÓ…ÈVeJ:Ÿ"¶—ì[ìd•`ì¸3“º7Áô÷²7t{¼4âúá ,Ð]5@7Ö#HAâgô ÌqF_c†d›¨sÄÙò°¾v¸oï1ÖeÐÌ÷8eaífšÙìüáçåÉKh”ÝEQz‡ïÖµÎxfµÅ”69òk àf–÷44-ÒÚм¸ÌIQç”Æüßs…wÃÏp-ÇÙ¯rzõê_×=É®…·šxªÅžlŽ6ù ‘…ÔJ¥§-Këy€Ì£Ðþ ›øXû7ä–k‘7~GïðeM:„7ŠwêiS_ËÜžqäe#¦Swåµâ² 9ÚhXÐŒç¿F|YØ,Ûu&¥§ h—?¿cþø¥Áèw5ãA†Ö_=PrSþ~eO?™‘¬l@NùŠÂ¬S(“7jÄ“aìçE3 „’ßê25ì‘…4ú‚y7·ú…#E4Cü·ŸnÞ»W”èÓã´‘Tñ€ ¨ ¥}gOE"sÔݱ@Ç~Oâ:³ôýœ¬“£Ò¥œìŸ^’>Ž‘ýÄbÝDÐÑ6IˆòX¾ß×­ˆðÚãû|è‹L#žçx,“Fa6 ²ÃÌ3xkÝ »,øÓ?;Mƒ„\ʨ×AqÜË.ñÿûž“TIñ5¬cæ<%l8ôcÿx3AΘ‡!w8ëÅ¿)k÷ë&R“ßôÏ® Qêéáç“œƒ†¨ªŸ1fÌ…Jèòo*5(ƒUñËô¾C¹ÄzË,x¥r±¥:ºNº ]º¼ƒÚh©BS)%:JÝzÊêYY½:¾OÜEÏÿ*ÑUT‘»ý-NÈ“åQ™Þ’ˆՈ‰-ɶÅIZ6µÖ‹°zï9Lù¾Ÿ•¨¦Ðáöà…⧦7€Èû­ÁúByÚ/:¾ôÒd0š[:œŸ“îµ›®Òç'à{Çm9š³1O¥¢Ì9çˈ(b³¶Ûü|ÇÎÒ^BƒlnÇî Ù² !çÚrZUëÉ£å–ΡJ‘%ÞÖ‹ˆtqhYø‡Ö€ ù3¸TÜ)bÂ#H^ER Ci]hdù7f¯V÷K½íÄI”ç{û»æĈDï:<ÛœÅÚŠ‘ Ñ“ÈÅ#8oÐYšgýùœ>º>Æ„ƒg#éÏeo~ôŒ*%Õ"Ä-ˆ" QõÕxÆÐO?'+—+p¬“—CÂÊC)­òídRTöR‹Ïµ$(>ŠÒH Ü>åA½,›H‹l¼àÒj!—ÿÌ­ æ÷ªáE…îRe1š°9c¬ÎÍ þèËRuÞ ¼4h˜Ëmaü$éO¶Zê'©²n‰ ôq·1› :¢ó©‚…Ì+ìijñu¹”p4ýÖ®šåW®Ê3dñ£‚½„Ð þ”÷‘ÕéܶW/ij¾}NN…pŽmͼÒzfDÊöñùLaûŸùM,Q×qÞç¡ãÜël"…0+ˆw¢¯Î%´}y«ïU{iG‰z61ívƒÑ/–bµ4·¥›o‚—¿·ÙÓW†ZÊ8R*~µN>ö‘5jÆóŒ_IOÌgLÏ‚äý çŒfˆjI Z“›&ÏÚZK‰Õ¤!` ?Cµ…Uº¬/©ংљbjIs_†ÊÅú-ƒk뮑†ŽË.?8Gêž¡^ê¥8 {Ž{ô³( ®n‰Í¨½â²~ø¶¨š°è "p ÂÂMnYÙañh&âçÞDjlØ?n° 9\`Q]ßa#ÿQÞ‡„H+2¸øÓ\÷D"b*Œ”zÝ´^-?õF^÷AÈë¾ûÅ·/éjý¹oÌ4Œ©ãêlœ²:ÚÿMº“F¼×1ÑàdüÔ–Ãg]÷û›ë³"ÿ Â\$Œ]îÝ]÷2O¡×Tý™­¤\„ß‹ ¤PÞªÁJq„YJ)ÜlÚrì¼±åaÇ_ "Ý| µ’IG°\»‡kAo0‰¶171].€r×å¬ñ5jãC ½,ão 4<×(Ÿ³½‡´i½;txýf`‡±îêÊR'p-ixàôpq¸ù)7þ}Ì{v0¿×=ñ ¦Òï±_ÌႶåú·`«§Þi¾SLN@^%ÚÑÎNUƒ$„z!Q­Žp ǸÅniÔKBhAj'©p”{q9@÷WÔB êú>¸®‰ÅX¤x@c\¡%ŠÎâ-ÌP,¨÷b(ŽqJÿÊ)Ñ®Váà²ðÉï"4ÍBx¥âÙ3Z$ÅùäéJúEaƱëUåAìé¨ã„Úëorµ>Tòc ÁQõ§\&/;O åÍŽ™(±~$çÙ38÷ÌÑd—`v¿Yœ<(¶‰ÝÃk1ïäw}ç3Ku¢ÞÔz{o_¿tÝXpG­°Ä3ÈŒ–Øëøfòtã|¶õkî¢PŠú8 1­‹BŠÏŸ¤üO½RÈ7*j+žôÓptEÕ™ýñ¶.V[Sn±œ²ŸJO~Œ M]‚Šß®É#S½!{ÛPdX!Áw@¸¨5üh&38oÞ`Fõu;ÑØÎ;e’T¶Q–bìØÖ÷çèÁ퉄aµÚ…âì8çÛ·šÃBsÏ´ñ²<Õ1^móQD·VÈvñ4#ã̪ƒkÖæ¬Så³ZV3Ɇ•Vi¬°zpqDqpÌüš7]¡§½šãú#(”ø²ðPîKåÉÄÎR Íîï@X£®°ÓŒÜ[dü2ùÃ{jmÝߊð \ŒS9^7úg?¶¢«~¦ãXËÑÆØ®¯~æŸ,Æë_@.!„pè(®£rˆíF$Ûo8ži=0vT^ÛùÚ ÄrŒØ<=·*Ħž¶yÆ…[Ý—”wÜýUbÉåÑÏl ±C#ykäj´ÇòX!³Çˆ.ƒ‚ÃæòÓAŸçX] ­§®d<OxG®ª¾º0ÞgwÀÿ‘ûQÝÉd)P©Ù«-×BV`Æî¦<$W´ûîÿ–u_W½·¡zL&AhŸLÌg©VM¹òðäÄ9û¾ICp [ö°Îã~­ÛÎ6éÌZ4ŽMˆNÛš¼ˆÁM£ZkJI0ÑòþÝ&פþsT;W¤¢f-{±õ%øÐcšN´þt+e@ëºX<¬Ê‰UßìS¿áÎIÔÆY—€¢6¯û@Œ#¹¢`ëϯ«+`{oK-'Ì×áy¸\ohÕ3ˆR£DFÂRš§~ÔŸ`?gÜ“®ëìþÃA‚ó6äØ@×Ãóü NˆU‰ z²i©2`4+ 9X«¥öø#eoHoH¥¦|Ë<¡Õ Äùod “§¯;ý¹5蕦»°‡¸Èý~g¼^~EÒc–§ÖŽõV;§iš«äKÄI—m Sn¾pQ Ê ~T®ßnÃ0Šè*ÿ{ÇüœÎhà–= ªûA²8?*„‚ɪšÃ^ø ½Ëú`æM¡3œ)ì!hÌíá=×mŒÌ‡§*ïž|e(uúÉõÄåÂaº®«Àš(ª³ež)¯ûON‘δ ÎhÄD¬9þ™Ps>nà»›oªóÒ‡h!HÔ¥·Âèœ _ñrˆzELwA]©@m‹_'­sƒú^–^¼ß¥j¯× ó¤Q_­\Ç,uY’n‘Þý q”hÛ[ BPà"Ê :?V¬^jkI@s*nTI!ƒ…{Çè4¼ZÕ<´èµ±ú†Y-"É{’^ ?µUT©3 £°ßv©i]ÞAÛ:ί­.¥#â½÷X8øG)Ö»¤‰AŽrÊT†Î töˆ\3)¹¢¦ áPÊ]ŽÄâö¢šíÙ9E³ÙÇ<~°_†¹àu{þÖUÝc¸ïôG”‹ Ÿ ½ÈcõžMNöh€‡Ù·v`j •Xú¹*ƒ#Ö¨mE?0òõ§Ÿ I–¯øVtà )1#©ü '&.É“g«…‹v2Ú±JA­6[Ù³üò ܵ ¸Eþî1Î%KôB.ETÊøk ¹öp¾ss“J¡ƒÚ3/¤V_ˆ×Ê$1k¥ÏmV3è!½ŽY ÊmÏ­H6&Uë·1ÚWÈve)•bSFç3{r50@6LûËZ/O†Qïè">1=aí)šŽw5÷Å–] ƒùÓ¤šàëº'‘`xçÌ¡\1Ö@`®ùãó^ˆô¬îÞÞN˜Ëcñ+õÀCý§:oΡ¬aõï÷!‰ã@[QHnvƒ†ýÀ6Õ‹;Â0yÝšx¶ôGú±Îy‡ëŸ ŠŸ°vWeWÔ/ ÌÝ>tm®›1ð_Õ‹ò«±ê[¾pµÁOwØv¬t² ò"Ò™„þ0µÀ˜,IvÆyyQÛï=‹‘-êÃì¾:ûŠŸá“ûSÄÝ"y柖³¨ÔÉð9 >tXTÔRš@±/Ž¶)Î-¡´XhT,¢Í—f?ÛÙò"ç/ Ú îá—…>\©õ Ä¢œÃ¿9£kÂÌž.Û9Õ,Lùä ØI^îG?}yDìÔYA€ÿYô Äš‚A+˜Œf ÆUç¾¥†›RGæÈD“[]óy_Ι¿"N,^âL+d 8užÎ¢•³ñ¿Ÿ$2€_Û$#1T q*"úÃäå7™4ó™ù%냮ÆoÝ䊛É@R*Ì3ÈMÇ©°Ks Cg‡[-5õéh.ÆWdu‡yÀvIÖÐøå«zéÿC!ŒËâ*W>Y Ê¡huLµðŠƒ$}wχ…ƒ{Íð²´>¥[—Ä©y ÖGÛ¶ð µÀbÙš§CDèùŒ”‘LEñµ.&šO¡]6ÌÓžäuÚ‰®±ƒ0Q’B}±[†ˆ5Î8¶Üò¾ý¨SQWzЫk ¶¥k)F«&ÏL Y×O f­¡P‰.6 zº&A;£ óô´'KI?µþ5$àü”³p¢l­ìeÛ&3©ˆÓyúƒA‰"·Èõ FJF‡„ÌAqÇpç<_íñ˜sþEj˜{Þ¯Nˆ†i ßj‰>ˆFŠ-h­êu:f§;5‡Œ7¬ñ*Ü®X°Ë+©O;zlûÑfƒÚÚöéè×my“‰ð ù+”ø5ýLA_(xÈ{ÀM= ç‰vxi¶GÕ®¶j‡™ÉÑÝäÚ'[kl×±ÑØÔzŠÌ#ÂuM9îÏx‡n§‰ŽéþÜp°æ€:!§È«Hê­Noiñ†7XîE%¢råâ´º6@C ùѱw혇ý³¦=¯Ÿ Q±õŸ³Â‚MЖüdøui­^ïÁ¢JÉ<¤²qwŽÆ á¦IE;¡ÞSTÄþRÓxïËy½$G+*¶g=•%TÿtªÐk]Ì(„ãI]fQw\w,ö÷’Qu§ÌABGL´±¶‰hw¶ì·‚“ù¬W ¡èœ©b/_Za *|÷ À>4ÓÛiM±/ër‘™ø‚æ2}o"Í)U¸±<‘À!ñ#ÿŠ ßMè^vŽÔ=ˆŒ/{¾U!Øn+îÅ[Û—oÏH÷EB²F2I°Ä‹7ôÑ"ŽÄ¡>oEisûÕŠ¾›{håݶh”«mÜ» è ¬NŠU¤pe¹ÑjD?M/Çý ãìÎx„"á/IâYéêãFñw1#“çS€ìgù±”„œ=âU±·#õ飚fàÁº³“Å°‡tyÅ– Jbp®{Îk„ž :f|×!mºÐC3BzzÆãXä®JžŸA¼:ý¶PµAõ&NÄ™A:{å¬ÇªGi %¸¼ëÎz3¢ÿdS¡ÅÛʸ`Û]–å,rAd>$ç¬J=ß-iÖð ð¼6Ÿo™×«`ŽîYiÀ‚Ç¿7u¼¼ŒÊGQÛÅû56µ§ÚoBÏ»#]¯s°†ö¹ó581éŸèÛÒã6xÿ.¹ Âêi®2bÞæßþzÝÛcÝ©>/a¿7'DÅÆ‘‘Îíè6yïåSLµÊ‹~²Q²•Þ± YÐÀ"‰>Z]Oe[ ½Y¤Óˆ¯6‹=mWQuZ¯ ‹¾qëmqzzRQä•ógµ\5\Œ«.i–_n±vîˆhÉÐêSÛÇEkqRS{êæŠ!Ë÷¿çŒ<Á$UÕ#+Þ‘­¾ øã·`‹ß¡+¡¼= ó€2mðè¾?…ȺԂÒK‰^AgkÁÇz¤ÀÁêa–±ïò7‹Ï7á¶JRõ£ˆVÑ$V¿ýÂ'±uÚ¢…@ÊD 7'-W°ÞsðTzŠ0À´Dâ¬bZ Ñ,d¤þÙÊ¢öu²í@®wчmîèSêÈð.àï)Ò9•Ôë&¼iÿAvÆo뛽JnyÁûÛîÎߧª Þ `on°˜o•Y °hÎv.Ëuß~Uñç´Å#!B“«z¡{h“Ô¨•›ï¬pd÷8 ,|ø¸ÁEÉvÓÉö:jÓ[`ºÒ±Gu°Yé_ƒçù*×Ëà ‘ùÉ¿²U‚iŸ.KlðýÄW´NF¾ÕYÁüÒ¬f­m˜DŠVËû O‡mˆÓyÝOqÁÄI‘„Gï~Åè<™êÉ¡²LöÆk0ùæÃ}…mZ¸'©•9·úûx—>~tÔø,€C?a¬;©ÊÓYŽ;=C–½¿;¶vž’ƒŸjá°$–·±Ï^´0\§Qñ£X¥?‡cY5:¤Î9õv–¦ UGžÄWwBÎtÎ"Nן†>ŽÍà¾.'Ç€¦«VîïƒßSöÿ\·zü⃆»€“fvÓ÷0P-ÛöÔî^2"&oøxB½¡™ú‰<z\<-ݳ©|]Í7y–yTžŸï^kî¼d~ˆ=óÒB.¾ÂÅC?Aj¯øÓ$vûI Žˆ¨­mmàÆ6<¿BÌ}V¤PhJ_{Ë QG“-$1~Ë9zŸh}/… PûBßàB8!&sœ}ÆÁ"|*|" 9½››—E›Üú-ÓúŽ dqÚÉ¡ ³Ê»•{H.ýÁÜB³üM!IøIT‡Ë¨r7Ö3EIÀx *GŽ–7\Ôw}’ÝjE£/F…Ô]¥Ú/‚ÊÊjïRß9«ÔVof`k#¥~{v× ÊžZ¹ ËÑ÷œ‹œGÀkö'â%ßWÅüëÉ(yÝF3!°ä€)Ó rù²|„st»X7²\þ²R?„j>X‡Öå‹RŒ&¤–§˜ß¹u:YG}¯.Ç^ò½W0¢v¶“µ™Ã»Ìv¼ ‘bÑTRdÒÅ^÷o(=Á_~%Â97ŸÂ|ã>©ß:HžÁàõÔßN^”)îAU­Ëª,70ž3Ðy\G—Vz¢y©< N¬R„ðöõ¨°(l9%¤kX~í¿MýiñJÒƒºæ#è¾íu­0b­î™ÔÚØBäZ§Ûo›vÙâÇvÜ´S¾ßk o;%¯ö‰n“Ýk¬M·Hš°ÇÈK â³Ã4ϲr°aV€U2ù6ö—ã¾ó´ÈèÜ ½ÿÃ2§±–åx¸Ë§êI,Š¹ßU* \ÒÍLþWÙÝ#½(Ns>æ*ª*4$>”¡XþV=Ã@T®B"÷µÓi%Û…Á¼Ó*Tç~„í¨/ùVEÒжçñ¾œH¸ÆP–© Ê·Æu{â CZ­ðNe$Ðñpc¹· „¥¶ikPÊMˆ÷Åi[:]ž#azlà§V#ßú/uî%—&½ìKÐüÊá\ÌvHæ„Ò‡•<Ô_)%bÆâ#xõËWQG”ɱc¨¾BãŽvÂ0å³Î~æ¡8ƒ$«}./žSìPó þp‰VP[ºÙ¥cL'UÑõy8\z4a±½©‘ÓñPÊË¡ˆ5¤ G´†Û‰ÞgÙ„¬Í¯ÄH~Dq§³‡x} À6¼†IÅ ÄÝ…—ƒúÑOö[úóUÛÑÈ™ð`L4äjK]vf7eܲUðSÓ}K„‡ “‚ÑÍ@þ¨‰7­yf¿9[h.¾3GHµyáYBbã‡_Æ!baë%• W$Û[˜!cÖLI•àÙ JœMS öòaÇ‹UÞ!6‹zܯX3¢žê6ÄÊâJB*9‹<È´< ±ê r"tdë<¯·O(÷aGt†¬çwKB§MYá…}±ä.æòaˆÎ‚ùJüì,¾Ð°AKvH MJö»Ë؈ ÛŒîžkÝÁpó©ö€tÿy~{ud¹ZûkÔXèr¢qíç.wïQ9%üQAj1ËJ÷×emGï4ÎO¾ <å¥*¹iLXæÕîzT xåãM‹ücxX!ï4´HƯÔ0ò|o @猵”ËÂ’d ° Ç»õ O°h·ÊFlóMo¼…˜îTÓJUÓ?ˆaó¬4_t74|Ð!~oµñæ³ÈìÇZN ¤$!i<\»o7ïˆw[¹¢Í><%µš¾n}`¿@Z fß>Aco¿²œHöÞ m0- U5t‘³[z¾4þ¸’ ^_Ÿ(¬bkß<ä„!‰@‰î TŠ"œ´²YíVIDØ(¢ÑÞ^oŽrþúîÈ®r‡yÑ¥¥6“‡qãâÇX°PPjs§h¹Öò'‹çó›o%)º¡¦|Uf+à•uLsâÒ¢°—ð½¢ÒøZŽx˜ãWê ¥Œû9ÈÏü†¶d×ÓBÅoh£s±ry3Ü~tsªP¸ý%T ä$W„"=N2˜bnl6}ÓØ‚ÊÕímU¸ÈÓptärãkšbäoùD ±c-ÛŽs‚ùÚ­Dª-wDßéÄ„aÐ… Ÿ¤&ØF~Ä€áix_N嬈¬Le>WŸÎÓÝ2߯M¯ª‰&Ú¦GMÖÃ**•NZà™3¦2‹¶¢6¢‹àüâRCÔŽŽ±yfÅõWö«5²¯0¯}Ü¥šòÀâ17Ž)D’‡ïã˜=>­X‡Ïœ³Ãc¹½¡ÄÒªLTíÿ>bãËÆö*²4ÌKc¶€“i;KãÖI{k²»Àîo ±1ÓÏ„ùd†ìiv D&/F[\Ép&çí½óGP@ö{wl^½VØÈ̬=uÞ;Y-4û@èaf‹ñlŽMÍ…@{Ž›” ~3@ò7eÚ—Æü:ŽuûÙ±a‘n©ÝÕ™¥£~_'eçôU!k>ÖÞÉ™—eà­fFÆ[%$ˆ®ÞHߊ©êÅ$$À`”uFydW 0 j{ÏÞÁÈO )m™á†ü²Ùg=s @v\”¾GÁpËdS‚;æ{1´s0ʱ`^`ëvs†eb)Ó£qƒ¯#Ú-qµ—eÐY3áÓT°áæ‚2Ü_Äáð`âcËöÿ]ë;ëîCtúF˜µ– .(›¯À¡þ°±BŒõSÔ&ê\«?¢#‡„ùy_*<<¸{Æ%n%†Å¬¶NC•!UÓÿÖN»>Œ9ø%ö( *^m¬ų̂Ĥµ)Ef—‘ÿq#|lEÂæxÙÒ*íí®ÖÅb•W$ªŸãÇ7ÄÛÔm‘Qf™ìÂ]sçž÷çžF-¦LCµ´”ó©µgvGÊÿ¶»µÅ~+èW…$É—|~r\È|BÇ:°7óíÀ–¯‹Ÿ+Ô>õIìÚi|¢-÷_ÇŒ@ì)_Ì$äÄoé‚k¬dQÌü¢êMÊé¡´ "1T¡ç÷·Oyówà‘Œº‡w!vUŠÄgÃü§ºÁÁŠùbÛA…ÎKÆNhmèB ˆV7¿u¦“Ö¸0†e¸a"ñ­ªÐJ2ÓÎeF#˜YÀŒ ·AKgfÔÐ4ú*$‰[{¢Idê·’+žãÓ–K$Áe¨ã+s7ì`é;ü ŸÎÈÉjìä'a(hÂ8£C%ÂѱëaékôõœŽžOÛSx.‘ù„˜«·ï?"t#U´Mg6óY°õ_R˜OËWÐP\UƒÔ¤E«Õîy¢ñ3P˜×öknŸ«ç^ÀëÐcË©'ÖG…oŠ—‰xÓÒÂĤåãõåhô’¯a‹ôÊ!v‘h…Tˆ!± ¡Uº¼tl ™ñ*x.Œäx@KØxTµÍÎnÃùS?É7._G¸ÚÚÌ(¸«éÒ“‡t2zR궙œ¦ÁÓ3²nž®˜ÔU)ê<Ú†«Ž”Rç ¯6hÀ­A ¸i4o¤}Wõh íÈxûíܲX §ù×È¢ºDJ0¬úwK¯V‹6'Œ“ÛþôÅéR¡ÙƒÚ3_Þ²AŒ .Ì ÌL³~6“8 ,Ü°Œ®/èôý)ð2`Ý=YV`¸8ç>75U\—¤Cy:´ÓÙN:Ûè ÁŽ`¾™…œ’=ö<ø}éàþ=äÐ(n*ÆŒYO.=[¦þŽ ¦šELåô&âäòÑŒöp§@V)½p'~æ;º¼o)¤f…ù€àkµ3î­\ȵ:9 ÷ÐhÈ’õÚM3) ÖæØÑæÍ¢  Þ£ÇB)žÙÅ^à b öšW1æéú0á{À„`!k’kâ,Ê7vž¥„[æÛ¨ÙL–B0Á€èŠÔ"UZ¸›þ­_âQñÄk5€}¥¼ÒtÓ§iøÛy‹¬¤öâØ÷eƒî!qÇrW‘3‰Ø´h‡|·”ÚX«ÿÀë¼} ¿£ÿ%½ËÉE|OzOªâG¨6õsæ7˜‰Úù™n±r' ´!DC(y…ž²´Ú#,³$ȱEÊàÛLgÒ–Õ3x0+–Úßù˜6eû£r„³Tj‘¶(ñÆšxÙ%ä`„…gŒo³âãr{!»ŸñHgfXhn%‘t·ŸÌºïn½:¦¡I?JæÝ]“Ñb˜kPVЩzêÉ> stream xÚl»sp/ßò.Ûv>±mÛöŽm;Ù±mÛ¶mîØÎŽm;¹ßsÎ=¿ûÞ[oÍ?½ºŸ®õôÓ]3kjjȈÄíí\T=L™è˜è¹*âÊrŒœŒŒFz22'SCK{;QCSn€†© @ÅÔÀÄ `fdd‚!ˆØ;x:Yš[¸(©þí¨ÛšXÚZ:ÔímìÝ,-¼nnn‚nήôN®üôÿ$©˜š\,Lf–6¦EM)y ¥„¼@ÂÔÎÔÉРèjdci µ46µs6¥˜Ù;lþ³ÛÛ™Xþ‹“3ý¿ ع™:¹üCÌÌÉÞ '¦*$® ¯ aPÚ™d%ÿI·sqæþmjìdû¯êh.ÿcÙþeô?ÖÿD<þcÑÃ01L,]F¦æ–v0 ÿÒMÊÎÌÀþ·‰«ÃCÿrþ‡€ò ©&¦fÿ ]mlä mM”"ö¶®.¦N9{S';€²½­¡ÝQ†¶–6žÿ¿¸ÿr1üG!;ó”cüËÒYÜÒÃÔDÑÒå©Í mþ‘èß~ ÓÿôEÎÔÄÒÕö?ù¦ÿèñOGMìíl<ÿ÷–ÿ”ñï D$DDÅ5hþgþ³3¶7±´3¨¸ü£¥¡“Éÿ8þV4´üÏôü—̧ ÀôÖr†.N–mFúã¿®ÿZºÿ%,lïáEÇÌÈ caf0±±±¸X8}þ/ºÆ®NN¦v.ÿüŸZþ»þ÷ ™šz˜ì¯Øó[¥µ†VøŠÍU‚ÓÌͨѬ0¯N÷Þh#[[˜è^áÜ«âä»ÃáÏñí^îHÝ·Æ™ÂONE…ë®à¼–TšûÖn¿+¼5„ë]ƒ©cÛ˃ÝDܹdl¬*RO±fåž×J;–ñ!K0E”w€#ý_[|´“ Ò¥@ÉÔú†3ݧ©Aš³)¼6xÝ[ ËÚ…»!ÃÖ·(¬p8 Ó|`£lçÚYê »•$M ¡Cçöë³C“IOîDqKX2ŒDü0:0†MÚÆFèŽc´n‹PcMÑZ“xÓÆ&é€-vðrŒšTö‹Ñ7ôуIˆGpÊÌ ÇÙ¨¯wÒyú6埉և Y@”ÉBXLŠ ^ýÒå zä<´¢@\bv°ÈÔgløŠ^!ðn½õ;òqÌÅœŒüf5‚V¤â?e› C7Û+ÈèÇðù‰ :g¶ñ ¢µPÁ–¤†ÏÎf\¢2V°^r×+–_zÊBtó'›ïžž½½S•üF÷v¿ê%‹@ òBÖ£ÜØ1.¯ré]FÓ ŠüœÑvñ¶‰±´=úp›« ¸ÛÓ 0ÈKÎÓ«5Ö+T¥>Í;ݵᧃØõg ©»‰2+ʼ :f |ÄÃðÛ©¢r[ P\üÆînþæ„ i0r=op”ÔúRPú®û{gñ½Ï½èdH€'½c[×ÓÁ·>™X`Ž€ä=Â,õ}7ÒdyÃç«Œ™ÚŸ«&qáˆãÚÑ>ʤ|=µ|`Fú ‚[›ã^rjIôï·5Éj£Fd󼊛JQjÚè“þó’D{)Ã6#‡Ðàß‚¼“½C†8\i<ÓsgŨ‚èlz°C¬ÞÊ4ÎÉ‹¬9rô:§”¡>.ƒ’Árð&† d†ßWc8Mâ‡RÅÏGèkf/³‹äz%ùþèÔuÇ'lÿé Û“b `9Œ,Ôã(ô«Ël+;x²>M d&(°ßDÛ6Ž &CÍËFT.dQ&»¥†Él£†C!^d ·Dº,ì æÐÒ91š÷>¾w¢Ö²aá|n†Ë žÝƒ™„rû«•þÂx­¿ÜåR1‘΂"STLàj¹ŒZ72 u§úcéSɇˆÅªÅx]š¥ûN$1º¬+äxv„àþö#Ãv8±úëH׈°l Ÿém3˜C1´Å×ÖªÏáßO÷‹þc“Ã(¤¡{ÌRXÙ±ä¨ñ‘ïðšk~—´€ÄÚJÛ»¹"ÃÞùv±rðدé‹gPŠÓlƒÆú«ÇS¨m— V)%TB=ýø¶[]Lý3ló,MµŸ$Š·ËjmyU]ôŒ5gÃÞ™Y¡fAþÇ“i~/VÑŸ¼‰ÍG«Z÷"[&ÔÃ/JY(°6bI^îÕÐÈ1Ä05çÃÂáŒõG" QRŠ«)7. /Ýœþ©G冲S3¤Ë°ûÍ‚äL-lÏíà‘í—vÕá>öª/VËÄð(\#\+b>±û\aIÞ^IµÎüª©”-ØæÈs?Û¥MáL­Ð Œ þÊൖsq×VK'fÁ«Ýæ•pƤ§Su¦Ž$¥ÎßÉ{‡瑨—kusH-ªV]b†ÂäÊëÍrý‚.(ºD»^8šò·¤9“‘&;H÷4õÖ@TQ&qExŽÅˆu!‰ûhÀLkµ¨Q¥ LY¤ÖaG×bˆ“­ë¯,Ó+ø;]ö½=k!ï5áÚ|ЉCO³k+–­Ä~±¥‹±Ð:òþAVˆ!½*Lœ¥‹)m^ZhõðVÛ'rîˆ?êõTz ˜Õ!Í›š•WÈ©ù:¥!É¿(Ëû¬ü9̘õ+q¨ûŸñòæC#XC–ŸÇ^ÝÈÌÜ~„,¢õÜ6RÏÉÇ”/Ñù¦uîúMC‡:¼¨ì?vçE ³rxip™¾iH3©ªXĶwƒÐZ(¼ÿ&ú/>a†&õĨv÷î#¿,ªƒ¸¼q?…LqÒ*³=F€‘*‹nKŸ²Ì¡CGË‘ ÇqྌÆuý-{ê¤ vlúÇÞÔ¿òiSn^àçœ\:¾’˜"ÿ<Ð1Šýí$8ï–Õ½ž5™œ¡Š…-‹̉)O?¢¦fx ~žíÊD¥ü|Òõxí1¥5!±Jú´k\Ö)áuzæ-éâ·D~ Œ¾ š2´=Z‡‰ÎÑ–¿Ràv—Ré2ïd!:­f¾õ9»­<ÎsV%®~ƒcLÿûŠnÓk'†…o¹åñ%õEeÈ6ãûW_œŠ¤œ"„ï Š´Àiü΋|ó;2»¸£Ø GXDñ®Ã(§´>oM9YïÉ™ã0+Që¹áøôºkPÍúdWÚ‰ág­rkø1á/Fw*6·¬0CB;OL-”êl²¯§åÛQìRä6Gõ_… hÚ»ɤ1¿u"FF>Žyn¾Ü #ìjòTá>š¢ý0‡0Í)zÇ6èöég#ëÊž¡YŸ½ÑÔT^IÚa,ܶÌïú{§{pÐÃŽÚ¥™y"“Êb»¾¸:¹<#l(bt?/Ö,ÚiŽáy„(`¾RM³{`Ò RAŽÂÚhäÒ¢äê†+ôÖO7ŒgXç«ïkÀ‘˜¡?UTóè-'?³M®Ž½"ñǦ²7’õ ~1èˆXGêè|äÎ`W& ,\çÆè×ê}k´›< p2¯g—þ I{]#ÄqÐÙõ¸¾5†ËÁs£—Î]š[¸_Ëd¸VòäF¶'âeGáÝþ!þƒ‡FƒòD.%Êu±¢Â럪uÕU[öç´À7Úp …gnSëC!õX¨£–Ы|4N­ãAüèctiä Ĩ0§˜¤ãM´s-Yàv-!wÞ1(V]:§ †f 97"^ˆžíܹ*[FÒºu½´Y¤P´pÏ (#ø—W/Cp/Ò=T(¼èö ”,Aì¨º`:¶«·೶_Yª£.É+»X¶ÐŠµÔ—€Ù¡ ¹®^Îø+‘’g¿.¥Õ˜yBðWÊ /þñ’çaá;nr¸°WÈmºN™ªÀB, ,nÜ*ž¨´;Æ9Ó8óò[ó}÷w#“_a»®»þÊ„®ð© t!¶­OWp³p6°",ðqžÅɈ¾åïÃ"ï…¶³Ùi*ó÷ËwëÇ7ì³°Ò´].—?Ø #f‚ÐS)Ñž†£-S„fŠ¤¹Á­¢jiyV« åÇè?03Ú _P×Õ¿$ºt= .ÃlÈ1 Ϥ6ç[¾m07í|5… âùÑJ%&ŠƒßÜ™`laþj+ÅÍ‹lé%\ §ÿóh©é'릿G€ˆ1¼ï³OÚ ‘²„îtKè€Îï¨ôÖþ:Ðv³u_ê÷{_VhÐ!&½C\Û ¨ºÿ¢˜¶%càbä-ÎÔáÚ¿nÑs¨4\%ÕeÀëŽÂŽ‹oÙ"óÊC¯3PjZÝ7A{bgŽ»j›‘1«"t¯¶KTå«`aÆY ðùŒóßÖá§ëÉàðÉÑšÐ;è^FÑ¿NZ+aò õ,ïŸiŠ“‹cÇl3I$aCñwOÑŽB1KùNB»a¸œXõIHËMÃ=ñ­“”%ǨöGb6«…P%IÕÿZtº´…Zù&ßÝ|安ÁµUž§dóI1…50€›úŒ£¡[Îti%ô^Y*•HiHw&óèÂ9ë¦2Ÿ°‚‰·sF.'zö¤æ¾Nµ•= …WÒP;§òøN›ݹ´[¥Ë*Y‘“{¡-ûH>ç·”]‘MCITIF¬´ú—Ô£°—‡Br¦%¶(ÎZqy蔑Óì‹)A˜XŽvüƒúzƒõr(ŒÛ*HDUùe$hëÁ¡E‡zåhØÇ}†ÝóÝÞTÒ°Ém)”ÀßØ›*®Pâ>ÆW+7ŸöF¼ÀYµ u2úgk—…á“ÕïO¿W„é6Åèæ«ðE¯Y®+ÐfØVÔÅHEá³ù¡ü†‹ZYä1š Ç(—v!Ú /4éÒT³œAÝ@}½,/¢ŽïàÓTŒ]¬“ý›õëfôò˜ñ‰@öÇ+yJ0÷6Åz×]ŽQZaU#Ðçà—_ŸUç\X§ÔQÇl¤?D¿HŸã¨ÐNÙΦtóˆ5ëç6ðÅkCé9ó~[(Eè³ éL(5š$`ˆÉååÒ㺥»žÔ¬yÆJ&ÂÚ¯]¬@Ë­"ûêº1a·3ž~Ê°?–åTª6ø¸4NÉ:ÐäOµ™2ß…8‰½¦>òÊiEóÐ盹ËdÔêÈ¢ó·#B‰‹‚o2æNÁ­Ã‰ ËtßéÐ8PõÒu‘@SS”Ô†45EŒfdøJé¥âK{/l §—PO¶(ŠŒÃâO»?D>®œÔS*]+0ExùPr˜‘> 3ÆÚNÇ7Äò1ÉäK›6ˆÈÓÜ;Xrn©Kj%ÐÝGwÊ’bksž¡¿:Ì ¥÷uû¦J'4xÃà,q­úy4u=ÔdZû$ gÑi,¥xp3pÐÍÛ”ÅãFÛ¬{EЩ¿mcuX1!Ǭ_Á% ‹ûz‚Ñ àIÀCà \«L  5U» ˇÈfÐÂëŒÀp ›’Úg‹73|èšffçyxq x¶Ê bãlºì-×aï߇T¥œF–ûJ6W£Öæ§xÞË¥#eJ„O n禳×#E‚yŪý!®‹ØZzv=Ko ™€qv¢çʧËØÚ€•ŽÈK"o·rìˆP¦C˜x¢ ç’g±µ-þ:t‚øQlq?ÜP üdŒ]·{Ík9Ïxn*Y£‰ž)ZzsænÚn×}*?qŠÜ¾5³ze¼Ã¢XÑÁäv¾þ¨Q3 äh™h7ãJuMÔ¢¹Ä‡¯ˆàåzÛÁÃþX°0YxT. RŒÔO®í{r^:@F´ì ‚#(º3p(“Ž°ŒäX„q­Þ™WÉ›"E©è1¸‹œØeÄ¢o†éæ²Èe9u¢+s9 ’ª½þ8Ú%t… '™_‘Û&ùe•òc‘ud,östýž'›ë€®@»ÚÌ>¤yÌ·ÅÝ°ØêœT£!úgUùÁó7ÙÇN@ýé~ƒV#“G«9ߨ{MæÑg±¥iÞñÈ…®%†‚Í-YCúö9C!#ÖôÍ€ÄíÕÑâ—{„ʶ„A!d.Úv€”™ ðNWnxÐM#náx@Q ­ê%f»¹1£#4á½µIîY|, =®[\-*l>U«†ÌLŠa ü÷SíVz326Û{ÝoŸv¤¥ºcÔÅ¿¤úÉ˺Ūs¡«®XÚ=]w$4gë¿P¨Du zS¨é‡J5(7áƒà¡†s×Éü´º·(ºÉ\ ¬Â…Ês=Œal¼ü/+íS÷‘tæéj¢À¦¦ùû]>½‡„zfªK€¤Ð(1Årf¢f¶m4naO³‘–,Ë–vý)ÿVëÐó®aaç„zr\Bå‰áö ®ÚPªcÖ/±NºÄ¤ì¶YåyÙ¼ÛOæ¸Á3&(„NŽCƒQT×Y\Y&•»Œ{ÎD¶ÎýX’ù WÚgQñ4ü^ϧÌÕ‡³SÆD^(ÑÐöÓLká4Ù¢èa½Ú>øúDP–(°Á„äþ³B³„:Q_{yŽ@“Eg íDV‡;Áe`Ògëμ ¢3j¿§ž¨Y×Ra{ÔI.I5ÙíEXõ– „Gç£Ü÷ Ü\¿¿<6Ç l[gþçyÞw›qÍÔ Õ­Œ³kÝ8;w°o1XLš‡„ßw|"’y0x¬ÿ°}Ók³ I=톖-2i ]×>Kàê ~蜆ÞAÈymU$®¸«ØUqƒø¥3@òÜoo[Üän¾Ê`à¾ÖË_?'ÿª0/#Kš{N WÍžþ6SŽ‡Eƒ©§G›Î5w áÇáÁ«CNéÍlÝkö,“c«NiÈžúX”•þØ´sÞèìH§0Í> 1Ç&}M¯&z¥¤‰·§ÇN•Æ¹÷tëkÆüVã+å?…=îCpÑ ¶.”ûSûLðB˜P%|d[½;j?°W6Ò!P1†VÌn`Ø„]ÑÁŒ‡.… ПjJa: j ¹ŸªÊ»C©§ìÛó2®UO3n©}‚R>¤ˆ;ó¶4êC\ÛÈtÜq¿+R±¢oŽ€ã܉áö­f!Ò6Ô‹ 2dKñÇ?X1X9ßW½¡íàôÝþO%Àù4¨R¼7µK/„ùL=…Mu[A*®W\ä`M„ÞïÂÔž}céa±ºêí$\S3ª¹ÒÞ¦¡$²"è)û Ú¯å"×IƒîYh)k<]Œ Øø‘¿¬€Ã»ÝÚíN§¡!ïq¤·&¸°‡ÐLap5F£TÑî¯,‹¨2å†nPçñ¬VíŠ=G\û#óÙ=¾UÖäQI¨rñiFpn.[˜x)öî(ræû®=ˆ%Ž“(ØÜ‘HæÖÚø€Q1ŠHìο{ŽçWÀq¦1¹ˆví³?LÁd,ª}\K@¸œÝ Á« ìè9à¼]h¯n4Q ” K7 …“ °­á‚“ó’96“þ‘ug¥K¨^Ð_`†m|njøúŠ&ÎL"èk·;­VSÄu´ïÑŸ¦/ÊÝÙNÁQh»oFÝ}rø‚Çv2ÛÈLŠŽ•qhtà^AúÛý€)XM;bï2®”È€þ4rrFäê Ò­N­M¡`ÔêŸRu¢c²¶.v´îÀÃ-7žÚ¬Ëf&`‰•±#fîGº›×OܼSLDg`6:ŒxG±ôÏy©î¶`£Éá[ ?*y߀Y€ žø9›ûC+œ¹ £Êþ88.Áh-Y[>zÖ_‹À *R¸pµ¢,g±9Û%™~4Þ¢ê! “‚9ˤ™ aù…\ž'€{ M­ìºt6:ìz–¦¿ÿIQOÕ’ÔÞÒ4îAFÄ$žV£¯Xßûå¢ÐÍ°Ý{å·ö¢jïž@ûI‰r¡…¶¸†¤²þ7ŒšveìÈdešùP\$KóJ:z2ðFK³'ôÏ_³ÜÊ‹§]@²5<±É¦0SpD×ä1É}¹"òDs k>÷¦)}^æˆ^úËßúY€yÊ#nv:6‹H¿‹vÆud¸içÚŽªö œk&KÛ~.vhMÛy7¸ìx Ê'/z)¿Û¥MD3UÝ=¤ ׃DŽ±ªFË)ß„£yŠ×4éTè8©R\‰Ñz'§èˆ±¹•]ÿ’Úä€ïòL $â¦E>/8´«Å×í–;AšãÐöPºý7=ðjŸ‡å[­ÎÚ¼gU£¿:´X®ÈhSªëm:Ù‘ÐEJ¤¹;ÑÊÂz'®œ‘„=A:QXTËúÀò€ÂdžN!›=v)l„Àõ6¡H ²ÎOÒ¿…«M},ÔÃw´†ASòâ«&ãTÐò «ÃetXÑ?‹‚õ# D9±Ì[D$ÛÊñϱ±+'±Ï"z;SEìý·s/ž¹-9åüçÅw.ãÄBNûÙµå *[±™`)uL¿-bŒ„ÇŸfËJ3å,ahÈ‹ú{d¸í/°?ã%oûåšr¼ÝšÞÜÚ¡ûÅg"—Ñtá_j0 M©¡Þ)€Áä… PC âŽ"þ+÷äí²SëSçöTZîÏÜA‘YãÃHîʉ4/P;ËøTÖÉ&<ãn&²°=ø>X2_ùQ'…‹OÚxNöa_á.°$ÍLèŠÖ_]²Âj5iÒÖŨ¥â¼F Éáªs*X—åß#ø$/öv†2)}µq„2¦rüžï¦Þ`d:ŠúÅèz†íÕ¬ÅáÁ®«-ñ\†¾£%xZ_A)´¬[jÎ -[èpE—žíÿÖiù×ãW#ÁÁZÌ©áÜa¯ ­YµŒ¶"åÄ¥àï¿ãµÀßÖcæÝ¡0é†g—öü†î€hÛt9G^Óïþ¡ù-«p,î:m(–v¯²\±¨©Gä$êm%2HV³JçÛk™R•ÿJ{˜’Oà—‰ÂJüู˜"û"•{nûNA†̂½sdzŬóž çÛ%¨)gö×2»Ö‰õ×í0AÀ«37W›6A妫K¡ ³2ªðCiân—À„¦`'¿•èc[ÀSo3ðŸBs¬PÍ…a)wyÀD©íO?š(6®†ÍQ:æîÙ´ä+,¼Û„)W¤ØÚº¥ ¶sa›3¯gœ5‹9Í9›¶ìü›jÒ§GYN¹bP}÷É–¦êBjAàË'Æd\y¦•´§ŠßœÝÉèq•zâ9rOþ½ôí/Xr¢“B̯ßG(äeÞ¢A4«)I &rªîVÈÚm=Ö¡”uàè‚ Š[V!C7²¡í“c;÷¿â”ã}k²Þ½7¾é»ä†©¨&¾o½ …ó*_å¿}ŽLn$”>z²wÖPãÖ:Z„ž’æ“#é„ÜNpUd‹<kPÆõåF Thèfºäa?®µ9¢ôÔrIröÚ1]ì?á“UP3¤écQÁ…ª 4Ýs«]’e(½Ý+Eôî¤KyP/úÁxH®LúóŸAc›6}ÜuBìGK[8ì ™ý+fï r\ñ°Žªχ$¿Ç8ô–Npi;‹Ãj¸KjtBuÎM“þèxˆ54zЕœDB5oe[c³êÏÆ°‰†¸6©Œ×8aù§þ…³©öF5Õ†@jN…Íõ¾÷ eà7¯ŠîF Á¢´¡Ÿ_M¤ö÷`p$”à/bÃRIá]9 0”þçwŒÎ&‹‰)Ï–ºvwÛgrWx>A_«‰è¢Ú‚†~]ë[g°„sYªm0_y³Yänµª—¹zÐyß,ÅGs½†Ø4â?¨CݼùG-°áö˴㥊é=4ÚãÉu—Íq )7¶-V‹3UzýÓQ fûW`Jqùr8d¼+ä•MÌ`çºM*GéŠéáNµÑÒ™È_>8ŸÏ1_‘ø9ÎbKÖµMˆ Ek  iÕÄ{FP*M•0WÜWß r¸©8K$1×½ïÅD€17óÔ!¤=¾X”ª1x³%àðPÙ…²üX..Ž= %8žÛº¯5…ueÿ3#ÓyÒv_…Ü< 9â¹°<”a5Å(˜Gt6Öʆ(èBê¢$Ä.óUÖ5¯¦…Ãÿ y}AÝf—ž< ˜/Dà'yk§óÑ^–ŠvE=s[Ö ãGœM€ÀÔù)-šš¡•Üíúµ-”Q­Ñ7®¯¯ÊÍÀmÿ¥‘ÈÅ,ï½QÅΧ¥r‘uä›:ikY ›ðº3öèmO†‹4}óÝëŸðpÓE[]¿ºâÈÿ“[ìWëªj¤8>¬_@¯šI˜1ÎæÑ{ÔT׳TÈÊ0 ì¶b&óî&òŠ«¹ì{iê¿ô¬#}㨅?û ÂqÒÇí>Wê½É„Éî•Ýé˜[?Úm›ÊŽãÁ/©CU³;:8ß:©9lq6^ïG|í K R{èo·¸d‡"Ï>¿<*8øÝY.÷žÓ³ßô‡%ÛNŸ+ÆúR0IŽ¬µI;NÀé×—è^bÎÖm|ƒ²HÖ ›ØyOÖüz lÅð>Ã/¥å˜ªÆԶɘś:£C¦e=ìwÚÝÓÜ}Òcäò_ÆëIÖQÝ’¨ô ñcXfI*Œ^”ã_<×›0˜ü°0±ôRÊÆ´úë‹¢Çe©sàZJ*o`|õ:ç¹3³ dÓ‹Yìá»ñ½í±¬ƒgñ¿ºVGÆRòººqËæF%åú;âz“¨ì9]Þ×ÕxÌ4ãÜqòÍ»Ð}(ŒXLGJÇØMªž>™{Ó$ˆ$7š ÐŒ–g£ú7µB=’¿ÓãôϵžU  G£#¡¾ÓÿdS«Ýl¦µX(*‰CȲ½dÖŒæÌÒVº+µ€­[¼b²åÚ¯jBd–ª½Žb)“Ž° ÿÞwPñ¯Ðßí¹ê]”G€Æ%Ÿœ|ϳûæd8üè:¹î¡ƒ³pÍN8®‘ÂbŒOw åfŠñªí#~Mû| ‡eN àF3Ú; rÑIyëè)ð¿«Wv¹Dà¼Jý—Ϋ6ѧ.?=™‡êzi‰rT,ññwAwüê²…ÐÊb:yãN t3p:Œ:£é?’[Ó."Ö ”\¹Ga³¨³_Ý¥¼ÀÉcü˜y‰QZ"âû ä:3hdí·¨PÀ‡f%o­ã`)kX6y÷-Í ü Öžý”²d4Æ Ú¯#ªé_]be Á¸åí1Ž˜ëö-“„ìãUã—9늎lg…óØ('ÝÎœ`¸"i0Ù|Øt7דÆöfW¸Rða\ Iìv²nEy‹ìˆcd(•{ÌŒKË^_7,`­WµÉÊs+Ý}¡·'²V¤Ë©¶§»™°ûÛwƒáCÄÑ [3æ œ0‰<ò;¬l›CÑäø=%F?WßÈt˜éXw :~7g[KK§CãÝËÍ2ÖpKŸä°— ÝŠ¸å>©Â0öNå3ÅžNʺ2ƒ*ð°“cÌÑ> §\¤ï·Çò t@kÞºä/)4lÀXP,Ô æ„r¤¢ÞŽªÇ½^ÆeLû‘à))Ï$Ò¯‡¥{’êIÔê@Çô†5«þØ%Ý; %³ÿ‚2ÛÑ.,8x+|²Ԕ^ZNÄÄ{» :³è6âý2âÔgq3ÞfUE¿ûà&HàRð,ŒZ¶Eì‚„Œ4>™†ñOÜw6jìPºôtoç~«¿iD¬S‰ ÿ§eþERýß-Tw¦ºÕ$Ú%øj â©ùzRçùãaÖè>BÚVG¶Æd$dVŠE…lÔ"D4ŠýælÉá4RXA›]Þ˜„:Nó—(§•’YœaÕbÏ…E‰º)ÑSº²1@Ö&rî&Ñž¹5&ʇ¦› nèÐqoÆn q›áu¶£¡À”^å;æMâ ö¢óñ„ðšIɬËbÀ€ì⤘¥:W4¬‘˜ÌøIàa2-¾ƒÎ:r5‘k§…¹ñU‹*ú¬±§6ÞpF3á¹A}껕uÂq¿ÔQ÷G"–ºˆQ²Ë‹2}û}[ž„'wÞ¦´ýhòu,p éUB×[¸)ŠWBÛCÑ)V±²ßΊú¬>î‚r÷·“ÿ°låbðz ú±ªÛ(½J«(QœXø²6wl0ºœi²W‡èy,Úú©SUB3T ‘—,dœÊú£¾„@Û’íßØ ü®ð€› ’>Ü]í æEçžüë1ûÓÐùüÜHôêæ/? Á¾¢DOá r <Ò¢ñYŽä¿_IåGîŸPÄíc§KÁp¨J¶,aõŸ­×>Ò£C})A6å‹7kóåÙ…¿oG  0ÙkÆ«tª±×ílFÌDµ]´­¾Ç’8ŠýùΦÂ>¬.$Eòúíb8¼ÕÆÛ÷4ÑdɧáÜÓ$FÜv®Vé ’øƒÅ»kÀKáKæVyëþ£P¹ãj0Êd­OÞÛ]Qøÿœ©ƒ7~V–¯òM»ê ™4j¦22ÎgŠi(¦ÂSnƒ'£VUÆžÛßžoÞðÁ$¾©DÙ¤hxwîößG-…æ¿ÂÄàÛúÝ´;‹/ùsG4Óè§d ÷ßwþèÇA’ß ~„ç×`t~âYü½M eìʻև˜P Zx¯I\çdîó©Ä†{üüpÚ†ébNáèÅüþŠ¼b"$oŽj‰|Vô:\Z%y±¤§kkÊ›·î¦'Æ$Uõu“}Ê'þ"¢>j$Åxæ~ï3/aSåå9o‡® Ó5%rtÙ ·™4¢ÁZÓ”óãk÷èýîûÝã…&®s)ÛAÕ67zÈ%”Cg@w´uã+ÄÄ1Ò;{«YØtϨ_M¼#ý÷oۊׯ9E>¨Ç4Í0Dt©èßùïÄë@¯[¾tÎ m0&aaÆ>\vYú„뀎åî¹ÕŸóäÎ4žk$>açEæÚ`ÓbúD½û+Oyìož~ÆÑ‘{7êXç E²á, Á£4Bê¯*O°œ~áýHÛ÷ÎîK–Óå&à`Fùw¾õ»)ÉCì1êA=å‰&Âp#9§Ôiì‹KøõÀ%MITÆ–!}øñáWÚyœ^UÔ)jLr2â&TU@ÕxL½ÓaVAJ;rÁcZ0ÇPË5¡±Ôž×‰•ºxt•Ù [åÛV13• FT–Ü)OìÜØ6NÈ7QÅJØBq;JàëŸz…qÏRqÜW-QØt“¡h¦+0ò·zx¶" ÷ýy!J  ç£A05_ù)»$~¢Ûç$BdÁ·D5¡Êžš‚5íAÙƒ?XE¬5¿Ç«C×ÀÏAÿRÖ9Àˆkœ¹6d¶+~1‹³Žï}n:Ìœ©¶­ãNÝϤ>q]=däÃÜ¨Ç îQÔ±ù%Í·Æuë …–ð³ ½Iû@¹’ÑG_á@):çpíd£øYCˆ—ZÆà);®§b(í°í ™ÐãÐ_ÐÜd™¿û´ýY0/ýøöîô†µM nË‹v+Ikæ"?By’—˜Tе­y`ö&/;­)')Tkö´+íÛ¥=®•xë°·8œÞk€êjé=Ä+Ö²ðá<íÕ—IÄEœà¹}ÏÀA P¹¦{DK½ LÊ47„­ŒtR)¹ïKÕ'åBä–$â½%Š“8¶‹äó…6˜Êеt¾ŠQ´8Ï ¦N xõé"‘C¨ÇªÁÂnÅø­*!ƒ„‚H§,ÅÙm¸c4"Ugaòef‘­»nyÁ‡ÙA /œj¤ œ¾³!qB¤FLõÑZFz87CãWd=I  1É]²~•\<Ø f-¯$¸¸Xuù[/½7¤{“uH {ù¨—ȸ"s";ƒˆßðo I!„¯¹‘¸&á?‹)i’ù^Óð¦˜* ¬RøCH9C« %;#òŸ÷1c [‰£0L¤C’K½¦+ÀywVØlŒY€Tsà÷_tŠé]:eê­Èfõqï šóPäÇí2"qC€wr^èº)>QÛÜÝO=óX>=Š“éB{˜S’•øx”.¸öÆŽ$ý ©÷ŒuÀSŸŽ7be•ý¹É‹ÝíÇj›ô8¯°ÏèÉIâè¬gþ¿o2ã9˵ œ ž7ãÈ­ß,þ“ó˜ µþ©Û÷aG‹åSê¾›¶3ú°"µÂx:|f˳}ÜR4 ¶V Âh[ѧ¸r ÇÿÈÛxmxºWad”?.²¦#3îCOâ|€Êxøh‡»Ò:èPem |°Õï">G1²6ÞYlÓ^C0UÕãÅ.AjöF”ìn2¸À€#åÆM¾nìo8bøq¥J=‰¤¦DØ( rÄNièóíÅb*Ø*…CÎѨ­ë ¿‘cn¸ÜÏ€Ô©?@‚(Di%ošŠtýèdwƒËT[Zì=å–X¤!‹3KÝÅü‹ÃÌÁÂÕøn^îoɸúaÂB¾5ÉÄ\ÅíC³?’ kÍQ£m ï¥<¬wzmCXòÿÜúÅݺÏ =v0Ô*>¦©¶ªÔ]:]§.øD½íUã7þïOU¬® ƒdvÅÜ&{€±•äÒ‹Mª·ÄâR8tkæêÌkœÖÝ«•\ËTÔŸƒè.Šw™ˆ½mGJ©øs”¶õ¬¨kœG8§Ò„׈Öayúš2¼«a°"¤ïó×ÞnQ.ÂÄ9Õ`-`!d_>KúÂ"^’E¢å„pWØ-·”–% lC²† ò&¦vz«êà,Uj’ñÚ"6ùm㧠8£÷ï$†i‘áA‰¨¦c› 1›|ö:§¦)ÌI°d¿ŽHbÐß½(È»d™£Àmœ:ÇÛÅ`  oRY gÇ-¸q¬D>Ö6·O Ø<¬<(´@ÏÄ2ï·é)Ër_(¶«ôÆ Äc¸²;î¬Ç ý$1Ÿ™êâJ\ áDËV8¹Là ¹ F.‰ÁÜ.J  qïùÂàq{„åÀÃâ`4g@ekR9…l1¬Œ^V ³q¹ª#3”°Ú>j£¥MÚ¾ĆÀª£¦³£â'Ä^õsXZp¥²˜œ@R[Ãê8˜ú!iöYΙ°¶ö`4­W~Ô/ظîd]ž{õ®1ã«i˜¢ ÷’{{©ésEF»¢D ±ÓÏ-¶F ca+Ÿòå‡N;ªºDXlº$kÚøD*õ3mE~NÖÇÝTOhV'¼ö¢1Ë„,uîèëÔQâ…0¦U’»ñÌÓ‰¯ü62Açû5».H)¡;yD<«£ç£=3c;¹=q¢ÁZضZH! ?‘ýHУuã~àIgr׈û‚½ ¯Ý›@å¥2î_&¯=ãì“-…§ d\§+6(©µ—1}MS_%ß¡)PQ”èrœÏ¢Ûpóu­Å>p²¶cæ¤ø> †Ú˜ÔþbäÃgã›ÜÜ—Òñ4$ÚÙ!O¦_p©¢wú‡™JþÅ–Ü‚^(oWUy& Œ‡<~}xzSœ–q,‹šûú~Nì_©¾ÀÚ›9Š|øKÁ‰=rVȘ¶íŽÞ|ÝábæŒÑ^ç×úí”»R"vH¿#ü”žä¥ë%ߺÇøÏ Öš2|a§®HÑMŸûn4Œà×Vè5¢†·eèßPߧúÈräÌfTŒgzoM¦Œ°EþêNòê”VM—Þ¢Ÿ;qÈQGv«å´Eºž 8ý ³˜]gÕÀ})Öc¯˜”ÜNô}f]/͇øMš=CuÑ—/{‡NEïß,!©ÜK¯¦¿ØFé XrM,>×=FõFزÆXçm RËo2„ Å „’ùX¡Ô¼»^:t4)ÊG”7ÉÜ9hû•¾B‘˜9a]F/â):çG~Û§>X}¾ Xn™Mc]û& ÑMÿ¾åõ6é§ûb}ax—à å«&çz›kvfíä~j†»Úƒ¶]!íxß-æTßÐ#¨°àfÚ^/E:Nçjã¡úi$MÓ9+Y¾â#gùý· €·†HÓGUªEˆUr>Gîö6^éY\‹Z´JËÑe;D&y|¯úäî<ÜyÞÔ*¬|÷§‚r'mOÔÁs÷CwÖŽ¿uþðãˆ;w¹‹9°,„Ÿq£1Yª™Ž3Z7šœYÐ¥p Áî(ãòqÀg(«é y( 42æÁA¥Þ/uÚOX1X¯W IéH¯agûstESáQ[DÕ˜¡a¸0€ú4x×&£=Âv‰¡v~°^ñÎãõx\ ­1f° kãk;·OÆ°DM•'¤Ñ¬r-g)¶^þsÂÒô”5xãr±úI½S³[ÿ8¥¶qé.ó+óMZ«K¤ÐiÄ¿¶‚Ïk÷ÐÖ²âHJ»Šº–™þɺ•³¨-SW£ü÷«‘¶ó´m•µ¤ÁW!lѲ^ƬE–)[¡Có¸‘Ô˜Âo½Bÿ%¤‚”´C¯oý|/ofsôú^M¦ÀZÄ—˜‘QèÍNó$ß¿™qSÜAX¬æíý*ž–WÐñÔ~Ö…Îpê':éC¦ UÙÄ1‚‡¡¾·:EfS¡ÜÛJkÉœ!ιt ºk"Dæš”ÙŲk¾Ž™¨‚£’€ZʶaéI¸ì„Æî«—(®}tå-d÷ó42¯¥¤5XYæ³–UL ÅF…]˜œ"NÖE‚ï:mOS[“yœgSøäë/9D‘dÒf4{&¹KüÜst4ƒMMJ1råÏr‘,ò5ÏSJ8sTn1é<ã}Ãõgjé”\E¿º×›“è§zÀ £ + ËÀvìn´p PÜ-Dr¿X&…0JHï¬âÂÅ»µv²ªaÐÕI_ÁóTSöÛËÞºÇó¤*–‡PóضPߦ‘ÕCîu!ò9¸Lâ\ )ÚñÅ´h®èÿŽßV­URݬxvv♫+=9…sn,ͪ3<ð”z)¹K†5FiÕ‘9uÙȬôoÖgƈ‚Ø û¤¡8ôÉ°K½;b2w®ôjå}ƒÙŸ:ÒÉyécj“M¤(Z0šÊ:…65©”CI)Ž77º&ÃóL`Àå1V_AÙÐb;H½[4?Í̤^Ôꮺ6xO–ûÉlj ù»»Dü1çyñÀ¾sˆS~&ñÿ\ BŠþIvè((µŒŠ I%“OY·>@-¦ÒÊ6‹¢VlÊc÷Þ¶áÛ+6+Ù¤•+mä”w92´,…x+\p ¬b”V6 0¿‘¢·aDR é"V}'oçkų[i½}ºòù}p'+e›¶ í.¶oo$¨&³¶ÒN‰{|ÛXeù]sô°¥°‹Zf”kÂ7å_ëÿîÅO 6{ÛâpkCÕ—ñ Ü–d°^çG°»ªýZg1–(Ö™Ò: á"ÍOûSüo¢›0éRJúç˜R Ù.+éœáäÕPÉX› ì§.Ár¡Æoý ³bôž6§Ë¾í2"K~JòpFÓ‹y ¯îÁN”áÕŽ+ú1ÎTëÝ™·Ç穈´SVëÝj <"cöƒ^TbÅ[{gd 4-ssŒA« š›*’mOÙ4,º¢Y!—:œ.ƒÎÃ$¾óW²§áƒ #˜)¯n.c)lèk5㬯½¶v|•£s!„µfOÜwaQ*±¿P‡b²Àk!Jê9±Kŧ¸£öÛ»¼ ó¥P¶Ÿsëv´2öí…8b-»Ü¶:¢'h¾_ü‚¡¡ØJ F= t ˆ;.+¶3ËŒdÿBœÝE¦bœ¹ÐGYo¾ÁYnª…þý; ¤9é¨%ÛÎtÇ ¤Â~JG jò»³²˜º[ãÕ.sEù؇IH°«–ïÎ…“nÌ£ÐdnÂ3GÖè6×Yå·ÁN§ñ3’À¦Ì¸…ó¼çZ$LòB+gÄ\óœ†?¾ºð¬æÛÉçTr¯ò.Â;$ŲóàakˆŽ%WpÌÉóR;±–z\¤Ë½CÌ´™¾l±ûÇi¾E™áŽP3•Œu(\ átžÊÇP®£½ôáBØŠZná¸vj°¨£ Ï‚yZ±gÕ{Á˜u pV-íìZQh^zj˜)È–‹E*ŽnzneL|C°Ñ}ƒ³EÐÖëß W_þ|E¬,)1—{œäsƒ²"·°(çÚÆE÷EL <ü@ j²»R5"ÿé¥iwÍÔþKq ã]‹ÓAnf,*w½õz\ »®i3—lÜ×6ÔíeV‘/û2ŽUcŠa†²glÿ}¹ý±"–¤)„¬à«%|Áì*‰dû~bû‘„»¢jŒ"ëìØðUÌ\51\¾yPÇÉŸNtyH7†‚¬€¶Àiw'… {ËyckåÒ,È{gQAÆC=ˆ‘Æ K!õW÷QˆSÆUÿdŽè×÷åCì$úO¨æÔ‡shÿ6><'š¨Oà˜%R¢ê2ÔÝ8¯ŽÛçity(.le%³;NÒ¨JF©¤IŒ™º–.7MÉý¸ÂZª hü7 /oÆ ^"¨Q¶3båÊS¿­íGÝ” ”zo»`§t¬b¬Dy ¸¢¬veB¿›àÕñጹÁJ¬š0X2ÕÀéO·í-´(¹£·÷Ë‘õ‰ç¯'štf`šªÛ Zx™¹³3F/âa†MF×w'¥ l¾­Òˆž#çhÂl.ƒŒQZáöúÖS¸Òþ ¸Æ_`ê%z˜“Jw–^ÄþãRÙ¼ÃËv»Õ½}·ÙÉy DÐ].½@Áœ^Êýc†> 8½VÖ áõO˜àY%ûÃ9jzNº›ýg©& VµcÖJƒ´3ÖÔÄ4×nnkUáD} ýi6ü±**úJí'fïEϱrÓÄMåHº½g¢ã9ÚæëŠÔ( ߊ~{ÞµÖj-€3Êís÷=m82]QÇ)¿ÇòÍýK›ÜÂÒëF{·Úrw=ÛMC>µýB›“un7²Æbà?°™ËÖ|XGºÊ§ñ\¸UU†ìòœ{Η¢`ØP ‰B?À ¯¦y'ù;{úÚ¤;SH*þæÅ1Tþ§ÄÝÙÌçø)Ã5€ëþvû”òp¾TfÍ…ÅßDÚÉ>“üDZàþ'2Q®Ô’þº¡‘ÓZß‘Œõ‚õÑB*'î`oå‹þ²b#Ú$8fNÌÀ…”í>&:é×w¿Üs„Ô}tc_\É-æÜ« ‡îwû ¢dd}Ø>¬ë1zÊÎÃѶá»Ì: °k¡Í+×ð;žðú¹hµ÷Ø=¢§œ|ÍÓ4E|fÈIÅ"v¥:<áž÷6`冷 Ú1¤Jç“šsù@EcWº4k±Uí#“ŒŒΒĸNàÅš}]àÏ+N„M@s nÕ×õg€0µz/˜ø)טŠÞ‚ 5Ü4âÍÔmâÇoÝôqÀj‹Ù¦U‰no›±Üÿ©ûÄ䊫•‚d}Ò`*'ã3W>½ê‡qWýX0o¤Ä0o skt©/?ž›©ó€•Pl!ñË÷{"…F…J(ü}öQ2+×,Eï¤É–ÍòêoAš‹mÛ¤Š@o¯4…Õz]Õy\1wºŒúºw?å ?H„ªVIgº’Žn { 4…î¦UÄZžbÀÝ7´¿…Nä^¢Ž–‘ÐHÏÛ¤Hcêïõzl¡ßav-¨IGÑßà`Ë8WÔ˜ Ó‹ï~"([è¶nkjú¥è“ƒa¨ÜךÖ]úò®-OŽB/ ¨aÜ„*ö\áRoÙ—F,')_;Õ”ikñŒ¦Î©ÙÞ‡h|©”ãMŽ¤O˜ÇzPœ<XõÁÊáÁ#½ÿÒãuýN ²ÛÏ?“ûeæP×ßîŽ}ê :9bp@= àíÏ÷¨Òȱ€_n×~³©(ÿüª;N˾ BšÖÊR‘ôÒz¡MAl}<|¾,ö\P2N¾~Ë;8pHóL°ü¿ÿ¬¸:ZK¥Ž>¨“ê0³3fÔ´‹¸§œžµ7¼°Q a•TÛgû0/PŸSBêaÄÚ¢é}νêE÷†lbÖa¼Ûª›l_=e7X¬=¯x¹’‰o§+iÓ9‡EqÝvÿn‘SßÛ¤»‡Àjù±qEŒ´èÇ1eˆu¹ý»-9yåñ_9úíéóâ…ž/&ÿ§åL µ¾«¼K¨˜Ÿ¦š”7ªªew.ÊÊ 6}[H<+<öÖ©ÂÙ·£SªÕ¯³ÇÙ½» äæO‚DÇâZü,¥O¢=§¾å~I $Æʤ—æŽÕî`nó†ú¾`fQý-™JOøÂÁtåhE:ÛHøa÷Ø¿â–ÏÅÊzº ¤õ*M×’Qe¸{à­uÞeòÓÁ8ûóý®9&2ÙÃ/´H ¿N7"ƒÃ³îì2D5ÂÇ·—¤ç§[kÈ&w7ªÿîo‚UêjeRô $eFÖ<¿rKUkßwýŠŒ\¶^ÓÓç{šöln ’å7g£à<:7«ä"d?Õ’iõ˜4wÇÃka™¦!H6®^!ï°§Þ¤fÝTÿ¾pHÏ`+«~ä@nÀ—lâ;›ü/ûN´¡èæYEwí>Ís'¸{ÜÜq©ÿ'ð؇ˆ€ÑQòÕÇÒŸ5ŒõBI^"»NÆ;´sV=[‹ˆÀ»®aÞJ’¡¼‚» Úð“ø‘QÊ‚çƒ\ÿlEøp Õ90ÍM-4 ûØ­&ŒFgªŠ0×V·9Q£ßÕ 6W#É‘Áò‘1¼»ãTXëJ…e¿$Ã\së—jÁÅ/YnGà÷Ùóƒcã.èãCùŸ+C‡«1zn “{,àΆ¨hì üeÃ[¹Ä ½+ô¿áo`T„@u^+ŽÚj~ÓÉè"–4†´:Œþ¥ýî{êsûOÂÇÏH“þ™s ­„™¥T”$E3*䯞* ¿ál[(ºf “¥˜™TØBN5"ßàSl©ýêjífO™õ¡£Ò®'P‰-šÖ’œÈ6iÉZ 8!ÑM 2¯%¸¸ÀÇ0Ç!¡£^©ÍÑ°åËg@–‰œªŒÍkÅ@˜FbÕŠ$íeׄ$Ôœ¸iËЫ;þ ÑÑ_ÑÿS–ËÙ×íW™å<ÃU¬—ñ{Q ü5 ô]{>¯¯ ïC=1X¸< kŽ·w°3JàçYÑžQe®¬(0m8°qGÀˆ+6.ð ŠãT¤£J„ý–/ º ¹R¨)ÈÖdܶãÚfÕ†à5*˜W<’j]ÌðÊflpLŠhÀsƒ‘øTAé÷Ü™mzëw&¥± ¡uÖ'æÄ(wÏwý©ÛB±YŒH°¥qW ÂÒ¼ ŒKá|"v`V‘y²¼Ù-H¢X’ J1_%e$í(ìœ2/¾fy¢ì¤Qðø~!2=ž{¥ja¡™2i K¹åé×É _ˆ½Å! ¨\õ˜¾×â‘Ȇó¨Šh§/b¼vV;Á4àà0‹æI§Œ'£³ôß?^VþÔÒ¨HùZŽmýX7 *Ík{‡ë3)/Ù‰ã“Ú*Á°s7ÆÛ{ÖˆØt¹‹ýpANWÔ/ÍÀïó{êç>¿d üe'«+ØkÅ}1Π4º2mŠÃÂuçÕöžNÿohGtp§¥ˆa‰Žåî™…\T%Ò=@ Íõ°.ˆŒ•sqˆ—Kþ¿p­•C£­´ðEŽDOY+AÅö…¨Õ,!H§no<'òyê’§ ^6ƒÙC~b~‹·.ëOãs1FRÝ9 8N­ ¹PÐé_ÂáNXKæ© ‹y×lÓè×±4',Ø ,ߗåă øã 0‚л‡Xž0þ±PÒ[û9Ú ïBÙ«7@* J»|.Þ©Þíõ-îÔIrÔ½P ú{J¼Æßx8ö*lª¢ ËÔqÂ0%CœÑ~ƒpzÏ ÑöÖ¨ y™v•a”+0͇9²—‡cšÍ9¨·z€|h<½Æ2RgŒt¯gk¶ÏnHa9; Ò ÒéîlX,'M¼ñ5û\ÛJDÁóH–Gúÿ>ˆöžX®¦i$þdM©ì· ½‰È+ò’e§,Qäú#Þª­€ó©[ŸèdÃ$1As{w¯ ž†DóÏN;˯lÅÙH‚§Ö8: 1BÁžÒb{z¦”ùÚz0DM™&C Þ>mß«ê­º6ÜêPT” yz·ã˜þ`0G·;§5ÏÍ“&x†Ÿ“*)“ì^:.«+Ð?=dÞúÝ=¤6,Ž :1 §ÒÝÐçÄ‚^¿Írj“`ÊGDå Ú£oõyÚ—k±©çyœ]K#¸vp 䥑ãHv«ŒRh®îyþKÕÞ¸ƒlÍ"*\ëOáTrž4³¢U4Ô™0N“•'Ê%¿&^~ŽÏ¤¬nßlZª¯!Ò4"¹Ôâ(U{õí~,Èâ7ó¡ù OüÜu¸Š8›k\ADÑm` ÆE Ó-£s§-&x3XÏ‹çÍÀÞ-Aieª "Š‰àdT° yO>Ò $µ®…Wè;á}y¯ ŸŽƒ—½°åùvw²ëåL8"ñõ0x?Œ¶b7B˜âO¼M&ÎH©ýôjòŠ†OR#+LÌê¨Ï'Ç÷0\ËËîd˜²8/ül8¨Tœ…“<%6ç¦*¨Ö{»6"Î'WTKÉ/F…“áXqŸY Gû¤KâXªÒ7,Ü5YÚÐßg¡ hèÄÒí$ä4Ù…s3m_I|úkÿ“¯-Ã0_êÖ !¥zôWPó\‡]¨öY)#œñ›ë ÉVÔ ™µÛ‹réh ´ ©¿nBïU@@ Có!b¥†7ÕU!CùxÙ1µ¨¤F%$â·ˆ1¶³›òŽ0m¤Nœ½ãoˆjmL·#³Pa¼Í}ŽåUÊÈk›¤ežóI¶ÒGDšÄÔ 00ߦ¨nUJ,‹õÓ–ºh©‘Šiõ[®m þ’8AÛ©÷ºòn!$‡Û›‘/äKO6ÉnÙ±.y ±ØŠx£jÃ*-:¨Y^ûžÆó› §ïç”à3³ÉŸh`Þ£èvy*ýŽ ÙlüqšK°tÆ G`Ná`7þK® 6‡‡Õ`v#ÿ½ü%ûÚ Šj=­‹Þî‰û¤ð[Œ«vÀ>±µŽMLA65ú™Û&|æHJöîÆ–,àà€†>wœø¡[`ž0oþÙ-=•;×) ‰ ±Y|Œ½ì\×ôN>îø!Xq+-•»hïÀŠ¥M‹ —;0±FW_y¨ÇqýG} R» ‚‹òiì³Gånc°ˆç`/÷Ño«ß Æ…>}.i‘Ÿœg²÷Ûð[äµó ×Hò¥>Ð)œ>’Á›`”Ì{yÖ‰½yµ¿ÙÛ˜Ž×¢ "&ÂÙ¯—õùZD‚p"¸Ïiï*ï/Š½xuÜ¥I<Ö›C€Ó*r|- s^Ç$Kô½”û}™~#ûr£V…[ub›Å»Ï‚ÑJGù!¨äÎT†<²‰€ƒAU?ÔŸ#EÔЄ>BYB  šçºBOa f¢ +¤.p3æ õ¢*ï²PŸC$ Ÿ4ˆSË7«¶XúUnó?ñ÷SÉíM¢zÿ÷ñšÿ©ešä#½Ö(`ÜÆÓ _àõîA×Bééò›‘šxoŠßŸzlÿdö§wÜ ~¬%–)téWŸç=±æ¶zHÃ$Û¿Óˆ  Zé^q”©ïÈ›°¼…¦ cc’°œ¢C•ÜS¤ý×ØŸvÔußÆÅz”öi;Ø]ëPÍ¿o~ÍØ9‘¥‡ÜŸ‚¶¾Ÿ ƒ#µ6?‡›P©ˆ›1ª€J? ÇÔòEÛÁàØCHģ÷®óÁ"5íxIXMáQMEE÷«J|é†Ýs i²I¬ýñ2òçƒJhj‰ãÄÉÉHp¢l3-=é†:M;r_iz·Ú\n¤|4öäË…3•nÓnÓ%ÌÐů¼&La‰`àïyË"I1ÕW²bŽÑ¡§pwìku™­Í5fÿƒªhÎE5Ïž¬mS9N¬5SCCã‚{_È4¡‹Òí£¹ô ×Aýv*]Š)ÂÅ»šUå\ÏŒËt¿ZîÇSÈ-!b"øR92áúÊq½wZû½ýâ·øÖ3„~íR9.ùq”¥:ìоÁ$JŒŸEM7 ¨ðz1íùópp§‹S³S=ÊmSFs¦r·-È“M—›vRØ+R¹eHQØ¡„Ä s±y-¨ªT,ö®“ã qÚNþm Q7(?:cØ(a¶É™u¥8É6íªï˜†OÂ[¢­ýã#­Ð„-iMøµ­âgùRÁeLü‘Ä=¨ž¸î4~A™ÃÍ9§¼*¢¢À8úôz9?¸ÝëvÚWåÒiÆ{n âËð}y-R¥e}ê™~=ÉBÞ†¤Î³ÃÎŒ»ÑN¡ÌNé°‹Å#›ßº†@F\I½«2? Ë|n`uçZnËÅú£Éãá>9…K†–ÜÈ™{¿¶«›úìá 5_Øv¯êÁìj./­—\‰Þ«i‡”>&ˆþ’"šã¶ïO1Ñ1•×›é”à 0\Ç~÷§Ä(†Áë½<rú+,Ö±I·±Åi¶Ó£ê®ØâúùâÇÐÀŸ…µ?¦MÓ>Qw“Ž41ï¾Ò%€7ãÁ¶ÓÚiLUkE5ÓEÈw+|(Twd<çÅF”Faõ®ñO©Öa{rl»èÛ9&Cÿë“»Ü]ߨ >5ÚÄ5ëêÊ÷ÊEå”#Æ„21"@Œ3O Kž“¶Jj"ܸlÄÙ=‰ÐÓ¼eÌž&=‡Cuw˜´$–WüwA…Œ™4ŠV=ë@qÕA8¿¯WéDŒÌÁ{ÿë¶oxâÃeR¸–ycÇÌúŽÖO·|^½ÞO’hûÄF UþŠÝ±1ú&”té·Öy„œ¤¸çջݠ‘'WùmÂó”ദɌKÉÅ }¨ ¼]gk=^7ö¡Tˆ™6ø#ÙÐÔ[šÎ†ù‚ëÈÛâ˜ŒÄ 7ôãßõžfpAVxÍ ŒÔ˜£²Ú—^hë/ÑÊŒ×V£(uΧ€@:£dJUê^3òÈã€Z×#æjºÆ‡„QPÌÝ¢°º>B(¯ºv`Y|ÓÞsi.z·›æL"iþu-ìéØvÖÒ3‚úƒS0ù­Yù Ü¢ ¨>-zÇõ8ë…Ôn¡L¦—ç—jŠ]í¿å+;á¥xYŒg eTÃÅ ãc²ÜæBu^lDèÔž¬ /ȯ¹Â®#vF„[emÛ0õž±~„¶AU˜6$´¤b4.ÜC·U©{ÆAÐy€U9½X6à^âûÈ,7“äÁ‘¯0T¹•‚ z¯|~q~ãºh'ƒ±Ô$Öþ8›}Ø ¡ÝãyZXêAN¤¨"‘ €„ÄÊmãÓ –rÖ=xQ­.ÀÈÁBº$¤/§m}ý°2ªf¹êdói¬Pô扪Ðc5[÷tKÇîVg½‘P­A/R}kâV¥C\±Qhö¾ÇÓÆ jIÛH׿¤~Ô<0|º‰k¤»í L·Š2±fÙ×ú¾dî-_d.¢Jív°.0¿>ÃJs9hhÂ"^”`]X0 M±ÿž?e*$ÇiÊÌý¬ýÕ–X$·ýêEšÈß$´ðC%m¼ï:eè$â͘Óóúœ§Äv…¼„‚I.[rƒÀŽáÕô€§Éçìȳ°›ˆÆY¿ÙÇ!íYb¨0÷9 ±ÃÚ‹Ö‚ó¬­áä$.t4â.É6G–3îù¡=èê ’e8‰-ÎîBÄÏäývÜŠÖ(›"!°âã–"¢šú¸³çS|?Ž%½çœŠú ŠfØʉ#†üiëîŽhxrò¸¶ß„ïŃ yoCÛ¬ü‹±¬™z΂ÄÆADhØãgªd‡k2zIª¸c\chUŸ…Ïâ6‡”|oàÕ4£-7×Ý£I{Lª€á¿ëv¤þ={ËB$g±ÇSdó¼Vr×pYŠ¬œ\O2ˆ<^TŽsÎb3X¨Ÿ?WJOaLéã¥Uo¨ÙeRm_PdË5p²p¼‘—UîÕÜ5ŒŽ=Òex;M†Òû–œŽÓ9î=cê‚éFö©qðgý)»E’æôc™BøÑÂJ‚/¡©JÖ!‚[Ÿæ–43,Èî Þù˜ÀìÙØꆣ¥ @˜”!A.$ü|‘øcÛŠ®pŸ…JKýl‡ÔÝ^²#Ìþ€xµq<@5Q>¢A´Ÿ^Ì„³¹äwí+l1QžÙYIŠ ùV};ŒÇ—ñ»žUT«æŸÀ[Õ±éQ2Îð/¶ ˆžÛ”oWôþ“V´KÒôñé¹mG5>Âæ:ýµBºàšœN¢>˜Ä¼[óvðì¦çÏÇÏ çj¤n, Ç¿ïîîÉa—* ‘æ×hÓAé,g`RØýWïr·qü“$#Ù ž/½33,ìy?jN•Ó)JüO…5ÆŠ„ÿ×ÑèÈe‡þÁOäcüg¥ÇLÇeÔäß&Óõj᜷ ,Ó©ðÎn ™ád’ 6)!ºåU›–4xW( @ ÒE£&% abNÃ’'ÿwEõ&ˆ—O@'{ùÄûQeÞh*Zö§±T÷hÈeQ±¿îuXÕB§a…ÞœÇtE–ø¹¸YBïÈÍÕí“°x´›lŦQ–QT©”ňº·¡™á‹ŽÝªøÊÎ×…EÍ2 !9ª´ùœäí²& –š /->úW?-ÉÕ)ª"’}çÜCÔç @@ƒ<ʘ“.ztW:„h¨öû!r™¯y’§n"4nˆ?Ò+5YøgCÞñ:Ä3wŠ‘Éëii"jÙ¼›£Óñú)ÝÝç2>“âЪÖÝÌI‹[¤ÎõaðmpEzË HæSœÝU“ѶÏðŸ?rã⬽Îþ´ÕYá ÷²Eï/¸ .Gw*ôù H“Œñ뀶¨&T¾«{j^ž—ÍËÿÝ; ­‹nW¼ŽÐy½ÄŽ^êüè jG0JWI‹?[ÒR¼x™­º—{ý‚#$Ñ|¢<ŸÝg˜jÅèÿ]0sþ¬,·€ûk"­ŸòÙŽJIâÓÌÖXθïí\½ö²ƒÂÏÁÛu–rC,­Ûc‡`#ß?³ {0¿SR×®Oe¿Ÿ©_äÍÿÆC¹TÃU0†.jùh–aY_Ôæq Ë=d™ÙÅŽ1L»0Og“Ú2IÛu*Ñúg´£<ËŠc¼Ê8 {n}Âè"ûÄ -$˜…Þ¬™qšžk:¼tnpbãP²›öÿóਠY§i3›ï§V'¡Ó¢ƒgkO… °ç©Å~šž,W’häðézdý@Ë´‘¬ˆ_µz‹ß;ùÁl—)‘E£ŽˆŽ\îQ ÙáÁ‡q4­…°ÁW¢€RõÀj<°BŠ#Ĺ­§(qŸd}lœãâ^E ·¿™ø¥j  z—qñ,|ëÔ©-lZ…£´VqMB‡÷-ÿJPì‡ïW•e o¯›OX³¿på´±•g~µùaŸMaÚíCÓ*åÂèQÜú®PgT«Rc[O¢aá˜3GÛK¤Æ¥>%Í@J`À–Ïܶ=ãïƒ?ÙåY€Ò[¸£Úۥɩ[”jÎðÞ &fEñ®Æ«, ¤“ÿý,žGÜJýdK6¨¼ÒùP„JoŸàËuJÝë|}}ÉQ°~£b–,«ûŠÛÔ{½ ­ÐŸÞÁÀô]yÒ;O6.χ,KåÊSS2 hc4G—‘=T»ur2®ø"9aÎÿyIì–ÿ•ÓÄ–•Ý÷Qïj×ÊÐ<ïiÿ_',ÍÄ ÅdÙ6¸p\€ Y°À=P¥²qš|ò}IOô¥žy°âAÚ®Òßò']‹ÌâÌ[o"/:/̘ÛY:”§tqÛ¶!M‹ã°ž9ï !ʾ^à“€È»;Ož'‚Š÷s£g­-Àw.ªß>96^Nð×êxéœsUI®ù”•UÙ1ˆ7h•ÉN”¼ï(hDõ„JO:¨hØ!jD®‰&K˜øž¶5…øÉ­rê~*ƒ¿Ðž×÷¹¤ø÷p %³­ÖàÞyÆé5´[Ùö `dNz*¦ÛT1”‘ë½ ýŠ€¯§Y1çCŒ ”¹ÌÕ±Xqº0V®a('öõÔkp3cpj¬íߣþJ zµº}kŒˆD €Eqù¡`ü8/øò¼Œ‰‚-•°¬ÃÆ?‡euQÛ½‹±löQŠ§g»‘7Oá8br # ¦]ɲË:•bGxš=ÓŠ¬mDzӄùn¯ÑxlûH2¯­A.ÿIHiS)Sä1Ñ=vñ]ïïª2FG6qå{j³¯š§-±à/ª_g>öPIørËÕÜÛÈl˜½Ó&íPJC©³–`—:iwèí²|šŽ:øJn)ŸXg6‚]^SÀ}(ÊQ³ñ"Q>ítß:üy·Å|·~窴:Ü:÷N ‡ MÙÃ.oUZ?þ|ÒSä7:W4AwÓ,¿u“‚úÜ(YrSÄDº{’)”*kÖÜÒ ¿ÊqÙÊY<À< !ÕT5*¯N+L}÷&Ɉ*¸èÖé6½¡q¿söê½æNÞ÷sŠ/ å°òœ9=£/?‚Èö7"Û§cÛO ËæIÜñ FLpN}‹q•/w¹Ÿ‡Po*l¯ÓËk aÍ]ß&ß®ÍüjÇìÂv]Õÿ€ÓqâL—Íýwà¢$åú­*Á•Ïî€×ÃÐR—Q$ÐÑåÂ}ÕAFã‹®ÑaâÝ)½f9®ì­öcÃM%ˆ¯&}ae*EàÒ æëö?PÀÑÍ5´w&…% wg¹V‚ø0HÏ—Cé· ë_ÛðƒAVDé kœ9ÆÌâ›e—sêù6à«ýâó;ne¼§e;éå®WŠûX1$›bMiNÈE«FMO›EHû8‡ðRb»¨¹ Ó@{:‚›A§Ü`’¦°öüòËÄù›ºÒ¦W{œ–Ê'n§ç«q $-[Rú±6íöt„"ÞR¨«Ÿ‰<»qF¢ =«$Ð8Ç$içt#¹M4•'DãeG…‘TºÂjß.+N0nÑaƒ½š…ÙkE¦p9ÜËKàÖѹ6ÝÅax² JIfè„nRCÒkˆ™¡å‰¯ Ï©‡f|l‚Ä?âuØ7Î1‡,••¬e£éÓ¬òSœlÁZ1wç·³GÎ$éõàž;Ì|©5™ÒáDz1à3¥bfTNèuþÌÀNÖHž1#˜’i/köÆgø üºWKoúÂg•·qá¯Ìè°ñDĊיÛæîHÙÌЪîÝÑÄícÔ ɺŸEÉg~¬¥\¶ÓgÉ]‰˜@ZÍL§BlañZÉe7¹©º%2ÔÝ ‹-˜‚+: –‹Ñ=¹éÏ®'ôžâÖ+Z6ëÐÛETw»ˆ}Öa‡AÌþNI%b^ÆðœãÍ=R>$³Ò$gÚÁÓºN q£~Í6¨e(–œæ¨þÇhÅ×K>'Óü‰3ež7¹¯Ty«lµ?Á€™}ä“Ïñ*gg-)¦Ês?抟¶âÁv AØwò§]f¸dIGàN¨?v¬bœ„Åò6|7˜~˜Z{£Ï»®3pý–JUÚš úú+Öˆ8ÕX<‰-4ŸhpÃL/t#ÿË‚âÀ¤€¤½$ ט!™ã@u³^•ü´‹&+Œ½E×h|»õCÆJ Ö”{^àm?Í\¼;$p€^hHJö’R7Öz[õí~;ímDfg߸¿÷|ò¢X˜" jå`_ÿqälÈw¥™/é苇QC1¸,ÿ€e3<´DðF¦¸•Z÷o+eßô/úe¢½Þ ÷ÄÜo.‹è©& ´”l¿¢@_u%¨\ÞÜÅ¥ñeUÙUδHÀ, AÓ³r* M à°J^:³Nœ¤­a­Ùúc­Â¡®ÔäB¯]ÅdX„73=³Õ‰‡í¬k€Lìvq%êÜú‰êksÂ[#4.¹K'ðsVÅ€|ßôEÕ÷Áå¢ZsLáðxHæµ½ù–ÌÏÉ»h÷Èb‡†˜²í¿¼*S'¢®bDü:éû<8ëü@¥3¶yVÚ~»Žã3¬µHÕýÁÝp¢"•û#4g¶¬£%ð^ä/¿©ÛD^f.ç1õ.¤ƒÞň;-#‰6’—Ž3¦JCØ{×-‚`a築´¸—•OzðEø!´|a)‹ÄÌT1¼iûD3GÎ÷3œµÍ99Š~¸ƒÖ#°Ý(@Á÷áøã…ajª¢Wh.¾W2™îpA~´Ò1K8%`)•‰WòbWic}æšwÏ(ÿê ÿÂÑ®DLRýqñ¹œÆSäzs6*Å6û ý6¹ PúÚ°©ß(ØGwã^ ‰\¨”&Þƒ-¢ûibjM]B<×BÊÑZÔh1rŒ!-ç.HEŽ3é4-· Ÿpék×’ñY»qE°:¹Ž*Uï=ö„nãÍ[D–£&?°ª>¥[-U èΗ©ÉØÓiäbÈV;jb‹W3JHâ :,™v@HTKÜXl~ÕIð‰KF_žW °`Aðrw$±ûŠ=æn–±øVi*@óƒãùDŸå$ŒUù4½Â§ÎéÁÜ•¦‚ýäU‡eÐä¹ß¹–Èͨ¹‡ëw5éCÀÈ´“çV&Œ{omêì$ÐK_Ñâ—Ç:aÎýOXš¬ÕVsÎ[ƒÎÇî׳£fln‘“kå¯ãE»®Ç7(̸”:¸”‚ªìÎ÷ü7¢{ÉÊžÖáÔ­S@¦8>ö϶[-åÞZ­D͘¨`>TUo7HÒ¢ùÑóC¯&Ë9jâ—d>•<z>ËY ×<”L‡ã²UØKÿÿQ¯ qì^éÏ_•æ6 šŽ„¾†,•±­§€øræå4½_ØÖWÞ˜ìV§ô·ÂÔVá¢cÕ^òâ¢Çt-Oì,­šj@O)qh6ìH9FÊ5!{ÙÙG/ŸÑ˜SLds“ÌsðFÎ%^§xѸxcw˜'Íñ\YÀqÈE†Î$¹è —‘é1oÎùò ,ÔUàرó(` ò¨Éz óãQwOhÛ{ ÛB]a$ÿð“;4‚`Êì!÷]qÍý©£éCCæ1òYØ^ž¯x:Ã% 媞àÃ7¶55d'§ç¾9ôUxÍ”;Z×ÏxB®èudF¬bs pS`“Ê„ÔfÑsxü\ îÑ"]¨•Îk³—Œë­r<‚kA endstream endobj 281 0 obj << /Length1 721 /Length2 24495 /Length3 0 /Length 25085 /Filter /FlateDecode >> stream xÚl¹cf]-\¶Í®§lÛ¶mÛ¶mÛ¶m£K]î²m›ß;3wæÆøâüI¬Œ½rí<'"ãŠÙÛ¹¨x:˜2Ò2Ò1p”Å”d tÌ0¤¤ÂN¦†.–öv"†.¦\uS€²©€‘ ÀÄÀÀC ¶wðt²4·pPSþ;P³14±´µt¨ÙÛØ»Y[xÜÜÜÜœ]éœ\ùèþ)R65¸X˜Ì,mLÂò š’râ q9U€¸©©“¡ @ÁÕÈÆÒ ciljçlJ 0³wØüÇÛÛ™Xþ‹“3Ý¿ ع™:¹üCÌÌÉÞ +ª"(&/§¦WÚ™d$þ)·sqæúmjìdû¯îh.ÿ³l ÿgýÏú_ÖÉã? ##ÀÄÒØ`djniCÿ/Ý$íÌìlÿ ›¸:ü7õ!çø(þÑ`bjöÚÕÆFÎÐÖ@!loëàêbêµ71u²(ÙÛÚýehkiãùÿ‹û?IÃt´3ÿG9†ÿ„,Å,=LM,]þ‘ÚÌÐæ‰þW7ýϽȚšXºÚþ§Þô=þ¹Q{;Ïÿsä?müû@zU1EI)Iêÿ À¿Ó¢vÆö&–væe—´4t2ù_àßiCËÿLÏÉüwšŒÿ×—5tq²ôh3Ðýk<þõü×Òý¿(!!{oZFN-3#€‘…• ÀÉÌîûÿÐ5vur2µsù·àÿôò_ÿ߃djêaj ³¶loÌb•ÞVé'Z €C¤îync (Âj^®m!Ðôª<•¤}Œl6š—ì^ò–²†\Xw%˜gtùhÜ,NâªÇ„hÂ8 ¿š¥S½·ä£•îØÈt&‰ ©7(Õa3wÓþgITÕiE6v̹á*^œh·¹¹zË䢕MR<µÀêÛßÉÍæ×ÙÞ_ž‹Oò$qw¾Vu~ã^aŸ»qƒü &rSbƒ‰ÚV®Ípw/¾Òé¡ãn/÷iïgÈ?jn|õTö`âǯ=bµY˜˜¤,ZgžË^¸k#j‘ÅXå¼¾âYcæŒ OQ^´˜4Q#€^ø"¯–ÆÌÜkªèã.V’B ’çxJ´ƒ¼ß¬ÙUËøOý’pPù/–” %8>e0ln3:!aq?Ü‘UÃ,¡ááÀâ¢IíHj’CÊÉ‚š¨N]?Ä ÇEé‚´-®ïèd€ㆧÀ÷ŠÐŠˆ= ÖÔ¦ègýc,‰y«A9èeɽ®'Œc‚ö>¸ê®×ø Ê3‹^²ATìaÆ›`¼‡÷mM(ŠxA4 £xåymQn²0_î_1Õ‹~jø|®_»ÅCåXSÀ`«ßD9²mÞy†cðzö5­W!»kð== è×Üà’TgÿnwŠ[œz×ÔFòaf3µ˜>GÚ¦É~õ~÷Ãz†€ÿ¾f©™_‹LŠ°lâèê µÏ_ÑÁE‰†pŽ™4i®Š´cÞH{ý朠e§RΡ„Ð‘ªY¾¾–’T«D„*Îü-‚å-̈AšÜÄ%ÑY¿;FýÎ% ™íH€ˆž,¶NÝLJƒadׇ‚"¿³`¾°™?çŠj¦¾wDCQSþèÔ3•‘âm°iHõ™ 4;©kÆ+,1†&ô…«¦ÞŠkl}ÄÉáló@jÙ`ðÑÝó·He Z¾P“Y÷Mî’õ[¤d×$Â[œi¨ÅBþ¥eŒƒÃ;F]i­|5éäjmßb3fÅDšÇ€~Àþ~k¢Ê¦åjŒà­ÊÆðãÚvý½Fõd£[©ºqÇÈ•ýe2õϨ¥Dº^^Wãe¼ãO% A ‚Ë£?yÞ¹Ñ;×8fp˺bбùЮ¼øÞÆP¹L’\:bà|ä× wJk2ùlj¤1ØåÅDÎ;¸â>xÁÙBß×GÓÂg‡ZŸQym©MÇm® s‡XGyã4?òI2¥Œ^Y&h$ÆBß9ºz'uô¥@ó^µÎ'Dð!a4=vg{˜t!©Úšp®¶;pX;`já£ÍYØ‘¾uð@²‹GS<¯ŽYÊM¾½éà¹Ú‡ %TpŠÀ+ ÝY¤þu¥ÔNÖK‚OøM÷y—>~‡ö>|S`–_Ÿ»Ýý…Ð秉ï)΂?^”´Ñ[É˪´UÔÕ½Pu^ÎÎεnø ^єާKìÜb"Ui]QŽV8läÿ\Êr÷ìÎ)RÔF,Ÿ&Žî˜MÞßùÈAÉ‹]é$,k&œàãï!‰nš¶>¯™õÂVÐÞ Z~DÚVXë :Fûs¸JÍè½ Ìß8×\뿈Kæûw%í²Ñ„¿‘£ œ&ÞbŽÛL.»4he$g^6Œ=IAÓd’èÁ¢g4‰è¥¹Òë1µü¾YÒïèÝÈŽ'É2ÝÎØEÞ€®è;CŒ´^z.• ìî.N5"¥4l]F«n})(P#·³x'î ¸ªÌE_µ\׃Fî°àGx€¾¶U¯5Á¾x_KGà·¥säïÔdLÊ¢­Q$\m:³Çe„© n27lEL„§¢÷6 gQ}°žÌ–éãÛÑïÇÔµû¦Ú3Œˆ[=žºÌ:‰×«Óê"ÿLÏ-ÿ½o¡mãxÉãú8<÷âD%’íÖÎí¬Oê¿Ñ¡'ðL@©¼;Â5{úx¸æí.±‘H\™­ÝŸQ¸„Ü0 vzO„æF´´Õ€âè~¹Ñ\LÃÅÔlC¶_ÂPOîäªd:Õ†s)°Ô)šçgñóDÎ]Þû’1-òúÕù˜¡ŽÂ2}JŒá] WÖf¶ %…©¦éq’¯Ð$Ö…§ÑËš8Õ-U¤èÊM.¦8üI (²Ìž:G-FÐ4PuhüÝ.™ø± 9Gè,<’g?åu–m¤µ4+¼º§±íø€HwÈþž9[$-q_¤üé…ÎV¦ÒÈàö]z\QU)\ðvHƈ¤¤Ð_ àÏe‡:m¡kìÌä6Jí9ØIGƒê9Y˜÷L½Ê‘e –ƒ´‚5Á½)äâdé[%4ÅNÝ”"b(:eÂ]ž–aŒ9QÃq’~±‡mcõ;¤r ڂą봀͑ÚŸQbã:µ"Ÿ®œl…”[G‚!½×‹ª‚ šŽK…¶_„%[4ßînCBäïZpoÉ•j¶7-oŠYâó(ÝLN½€(é1…\*÷Ÿâ’8‘×K:ýg7ØéÙ ª)>W×Ô´M‹”ð“l~Û5Ô±™!Jp4n-`l§]ÍAf)íK„ŽÏQÀî…&›úüZýÔìúL/°Õ“ãP…|¡êƒïËÔ6¢Ja²ÇBýn³à+ÔXùñ‡žâ‚þ©Õ+”o—#¯Ë- 36’ö`Ç"7Œd'…Ág†tÛ'Qma”Žý[›ƒ ë]\®höÚ(ƒf!ÚUÓ8ËvÉãûѦÏIYJàXUÃó1rÌ2ö¥JõþÀâMîª{TÏæW·tæÆBƒçÛW±Ç¦•4’£ê:¼ÃaçÛ)sÞH‰å“!¢ßSf0i³ÄEFíÊFE½hX–Y /¨®ƒÜ{Š¡Å§N¯³ Ù®Ôÿ9¾µ¬%[ BéK墽 u­²ƒ¥ =©û~Ñ6\pû‡#nž¿É'Ës1P°dQæCt£ÛÇÉêgØKçû6÷W&R1÷4Ú˜mbýßµ Ví„ò*™—ç“l‰“ð‚Å݃ZaI¤yGòŠrµÎÕ-n=I=‡ïm§>–º3nTgÓ‰ÁQ:l¢ kâ¦Uò0¬·÷A°¿í¸×0~Èï¤n˜þŸNmk/|µ!‰9(Ð㨚k<9ûøóÊs¼¦ëýëT"ï 7HH˜áR¤{&ó[rá¼…A¦Y«Õ¥)ž¨ñ0rccun<%D¥¨Û5>Fô"/ - «NÂË“Ç3 ò3Áú0AÊÔö|}+è•Dê ¾÷57a˜ÖVo›õ¹”ê ·©Ðð¤ÐW)ʶv3oë›qŠ~R£öüb< Àhìø³®_¥ñc ÒÓg«]&Î ñ€q&c=‘Kƒ[èž“šEø#÷€"îì‰xÓ‰x2ŠT·ñdR ¬Œ‹1€¦tg)"XyÙìK¥#ÃÝÄY}¡%¼´L@›>^Ïh¹ËPv6|Ô.r·š‡ºþQëƒIxìEäËÏoíþX:iƒ¼ÄÚüØÑtÒÜ×.ÏßÏ,†+úüÂ&¯M­ˆ;‚6¾ÞÑÐUé~U§-`¡;²ÑßÐê'/ô +¡ ú|»/GãˆÊBoñ$Ø).Ëc0ñ>ƒª­Uà~;wæÁGba àÔ¢Ìúr•F ÎE}n š€Ý8ˆœ™O>"âïWÏàKçÐѬýò.° D±9A›!ƒÑíp:æÎÌ]ÏHˆMêF/ ä€ñÝ ò/þ•UâdGçÚ}É ì>ÞáHªŽ¿á®-°ÒÆKç(ú5ŒV—ÉÖRµŒ­T{ï6–íÖ±ç*n쎰±ì¶jH#íäDu`% 7‘^cƒg¶H.›À}êI’JXìÁ×½þ.ÊY&G6 Rêæ<ÛñevU{Îæm•Ý aPHª©aÊxu]%ÃßÌô¬™ªÔ®Y/i¼×|ÀǹíàEïlqhQW>˜ÁЊ”–Q«4Ú«"¹±7IäØŽSàuãÓ¶š9 ‘± ánxé- w "¡Z·ÎN1 Ë_½³ÈfÓÿGõó“ w?ê ÌÇá,:ñ)=̵’{³Û ŒNF*6.\Qg †‡ïax‘Ò§Ø,úvÔX§Ò`§[¦ÂC¬;ïe'Q…¼½Ãèé5OîŒÒ=ì™+sÆ°ïðÛ}G?»½„ÕŒyö™¨ÃÒ¤ÒD@ÉVÇd]fQ{ÕR3s*O#÷L¥§ž’ŒøüFÁa‚Nêþˆ´œÚnà—Wm›ï€Î˜øɧËi„Q4èÏx¬9bí* ·ŠËÈ!…ý»¸F$Åô®4Ï—ÓÔÖ¡Æ §ÙŸsH$4Çï¬)~µÛø¬M'0ƒ¦`8$Š]ÍÍ®ºàߢF ¬˜–X×ÂŽwWœ5Ä)( OÜhír^«‹Û™Ùžàù éHª„*¸Ù>60"×ü:ì„ɽE–…ÄõþáLääžB¸6Õ5vûª ´{¼sbæ䡾Ij¨`„;ÞTç¼Bâˆe)¼³{“5Me¦”º½êªdˆ4–i¸ø¤Zd9ãŸâ½,¢z9ñÐ!}VŠ£»é´cQÚ®={âï¢g"‘„o5²q•ðf’§¹=A`ÈcöE/æûÀÍó˜µQ²7êŠÆ¤¹çþõâ>æË#‚”ž»¡ »ûƃyï«€Vœ‘eñ†Ÿ ª~rcàI¹â·§SìpI…©á¢ÐØ4k]óøòä²7dò7Ý?x£ºv¾óÅ…žEê:^‡¨pÛ4s·*(Ê{"¥u9ðî\û¥@JŽLæ"LwÀ#h73­÷±8&‘o¢s ŽD UÝ4ÔÞÆ"‚r9ß¡ n¨I¢ÍL[ueó~âùßk„¾À¬‚|‰ÿpߢêH¯&-X¶‹Å5v…&/½ÐÖÆ2k–ÁÓ¥`ÙÁ”VǪyΞß6_q+Îc)Ï–wS­Šœºd$™µ ßŸ"lÃU²ÀÒMç!ÍöÒFÉ/Î4_®kPÄAR/ëç™Dxp›}¢Ù3W±¥ÖÀè΄Œ Ç®KŠ °À'EKçS)ü¸Ç/;ñÈ<ÈjwÝ¿ËmÆPCÝé·Þä©Ò÷÷¿)Nn,™áò}ö!°aoürÛq..(¿à7…¸;6šÆÞÇÅjUTŽ6vÖoñ´8<€ ‹Æ ½õêÄ©Í?oðšÖ°›rÀŽñÐ/*PFá¡ôë°õL¯ ÓWæ[ÓZâSå'¾W Í5B’±³RÉÿÂI-Ýô¹ïõ×5Bê•Ä¤|†àøŸN`—¸Ë(‰X)ùkýÆÀ7a‚媑1@òÁ¿¦ ÇΠ¯BUT¾Ñ ëÙJ¦ŒÛŒàm—™®!K¾‚˜¬öBGÜ+Ï´|hᎄ–o¿bø»qÙL/S?ÅHœ«Ú¡ PS55¬²T’¯ .U¡.ìAPÛð3~Ϫvb"ÇfÝ8A80ëM’‚¨å¯nàÆ/a¥Îì·©I×$k µP¢ýX^&N†ÌJÂA81/d˜û†µ’ýÍË/5Û•³UÑä!²l,ƒ£s‚G0?ÏÔ2+Ð[}‹hÆ!¤Ð=U.Jú.ì‘ÁÕVårÎ|½=sÌØ÷ŒMº»ªOtÂâYµÈ¥ eql×Ùýz§}Œà Þ= å’v#Ý™;ñ PmC÷¯®zAÖâ¿šÀrbMÍ;(礉¡ˆPæÁ>/f½œ¯oªÊæ¼7¢ú…#XsP%*[5ñYU ÔÉ)k”»=öv0êañ\ç´;"ŽžQs]àêuòÁ'ìâéñ°Ë¶=Ën$6ùþ¤ucˆ*2ü€™³¡›Ñ@ Â%ÌÊá*Ñ3%yD9³aèÑòþ ²ÃüëÖÈéñèrÜ:W*Òì …É[T>?œ¢•à\yÍKŒó‹L =‘áp¬Nß}´Í„ª’÷b8°Š ´ÒúÛÝÝ4ÂåIþ %t„[ Øá±0cÙÄJÓ5kÔ‹U…Çï÷¡z‚&wª£ÚÞD´¦8p·XŠÏㆠK¢ÈQ·­¢Yµ{B[¢È:|ô(žOUêí£lÎk¶_˜* nçCŽõ +5ÆÛÅWÆú ô¡'nGXŸ¸*¹­·<úv–€$YÓv6|É=q]jV'=3ÄHj¹1q— ÇöÆÎëb ‚u±-okœÈ_F梗eh'GY@Íéš‚C„iêà뎘[7~¬‚nN¡‚… T¶ˆ=‘yQ.‰x9ÄřʩÁ_üTlÐBÓIÊånò]«¼hˆÚ\6+Z8ƒU>I  ƒÕ2z_úÞÎŒ’Jϲ7Sª®j¾Ág+ŠÜð1ΰ¯ιúñ¯p'þ®mãdÀ‚Ù\NZ¯?F=?‘@§Ñ™›øš÷"ëh],èúÕÓ1𯇈d9+d·€u'»sÓvqªð“rÅëËR*3š‚×õêÖvH)–mÿ™Xñ×0rïN¬0-gCïäRUJsÏG6Œ3²î<”Ю/´úS3Þ Ä¶ò©ïbÕÅüqº›[å# ©¬ÃÒ•úċƵD@èî´5'dš…L¥ÖþÅdQCÞ?‹¯-—W— •‘ %{¹®5"2“¬‰ÛZPðó¸ý6ˆs¹•XJ)^Èç{{õjëX€âžWQôÂxÐu(”À4ºËî µ€¸G› i€v<ÙUXŠ–Ç¿<ÁpÅ-ä-¾¯&J7š‘ é»ç=0Ù)8eqfþºr?öÍÈs÷ŒúÏ“Ç|/|ÃÜ÷QÇb?éÿ¼Ç€Øz(@(DAgŽm5ø@†aÑ û¡––yÒc!‡ý5Á DÇ8öe´y/|ûä†ZíÂaéþú¾¨âð´wø}Ƨ+ÔìL,Ý)Mß=×QQ[Óƒ…(«J•c%Æ”Kn µÀFƒ"È•-qgøŠ¹¬"ÀùÌŒ†°ä‡/ÚZb@ÐR<Ô Òë®þðÆ7(ø#„™×Éúp‹|·ðá‹øÍŽ#›šŒí 4ëXš§cÕ ¿è²BÝ<™¾ÙÄOv¥ ç­­Ó^3½«k\‰€ºŠ^’FIJ؉Nš{ˆüê²4ÁQWŒæ€ÞÙ¼ —Úñ«ûÑ Ši‹èùÍËßãÏÈÁÎ4 [ó uâ÷‹ÔÝ]z&€£_zç—ß„<-5 á LêÁdãš7YÛ$~ÔT"oL.Ür•z-ˆ†BŠ©‘·¥ò„³?Èu÷FS×À„­;*x½!C(Kªàf—¿ëa˜ø¥¶“q>pSbúÑÎü·g­s°H±ÅìÈÐ7Ÿ{œ€+=‡¤Ëêo¹9Âa›Î×úR¹:–>_¸WøhZùÑ ö`äybõ Ý0N¸¦'QBd¢óÁud¨ü¬ÐÖ5d¶^«ºEö=7?ÃF( ÈŒ†&ŸˆGþÖ†÷¼¶2l’†½vÔ¢>÷TŽ;|¥dpÜ œÄÒ¡IÏ  y3}N@-ú‘1'¯šÅ¤§£Ù.¡m¿•·t‹LF­ù$-3¨ÕÙ"¹¿ªE\41¥êvëŦÓÀóip|è1~ ªnrÙÜ‘l;ôç®[ÿ"#0µòõsd”êhí îî…þuc_îKö<.¾–€–uïOáCù%j ôÌ(u›X‡eú<š¦>þ+>™i`)ÃRóD‡ù›´ÊÒŠ¶Š¼Ã»ÛoÒŽhˆ›êrEs|dLØÙâ‘hÍ_ƒºÝ4?œ²°g©*ã±øW®Ìš±‘µ¿»äÀß|¥f³ÑûIJƒŠ½ÉN?ÓøžýfØúú+Ú'Ïz%åbíb®¸÷Lîì#fS"3£äk2ÿQÌÓ}í¯Bù¼ÐvÆû åEö`»˜Å¯Ë»b|Œß×Rˆ_w ÍàÔŸúèöâ 2Ù"Þ²ã‰Û‰4f•×.G QM!çj¼ÞªÅÃw Ï2-Y‘`ðn¨ôÙè¸gõ¹P‘f¤R…›Bé0¹jGWÚg>õ¤Z/°Pì[h¨Y%æº+J³ŒS´¯÷nÔ=“÷`)éËéo†g‘R J{EPF¶Ñ#¤<êˆPï~žêf$y…—†t NËì"CÀA%0Hëô›ãâ.è‘ëM>ê–¥hOT©sÅcÓD'»3df$StZçûG(¶tsG‰ ‰û©ÃÆÀI÷lºyusÓ•\ø"À…Á[›Ø$&¶çv Ö Vàn49D6çÀiZ£Ú ID’Rn¼4l"’:ÚØœÐ뵂 | |d$ÍÂYXºTžW0r;–dèöRl:ïà ÌЀ(Þ5έ/vôfcµÊ¹íA>ÊÊÌ( )Ò{l´”MUÐZ±ÒÅ^yz[ÄZ½Råß¼EMhDÛv©{^•i)RÁÉ uvc2 ñîõÀPÖ— ¹4aIy¥ó¶/Ô‚°“¡qµ>uÖÌ$Þ¿½õhž>¡®ßÖùæMS¼Eò‰%ˆÔ/ù·°öK»}mîÕ"È™@ °À˜zIü¥OÏ™×Ò`$m˜¬Ú]²ÆwHZžd Ä·‡(07·$Rd#/èlÕ¦Âp—n~ùgJÚT.ŒÑ2a2£RJš©Z숞ˆO>â¼è@s•³ /?,2½Z EñK)£æτ߉sŽ¢‰Rþ>Άîæ¢ÆS‰ ¹25M{[G Nÿˆù…{s‰{Ko*bš;ÚŽxù$NdõýŒÚw¥ÖϘ B!âßç¡`ê…ŒmØgÓ¶@ß{å]†•ÂßD â½—pðÒó£Yb#‘Ïò˯ L» Õk§33X—ppá“a-HiÇ¢¿¦Ë÷ÃŒvùÀ6¼ Åyãfž©Ó‚Œ¯4MæžM¨ÚYx9¼öŒZô$kB#ù]ï«ÂsDµ!,õñÊ^ŽaØVó—/Ò…öšgÊjäÐùQrÍf Ͳwß#lÛœN#ê’P&Ëg=ÛYq)OæbÐÞZãÞ™HxWδw Ò>.k~’)@åŽ/ãƺx²I¸qÕ|0U›žP3moÈÀ 1[ñM#Iظ”á|¿5~¼Oæš`Ü0ÈçÔ•X³ÜþL í€îöuCьǰ²Ï45sÔ6ÿJ8“WªÃBvË£Û€ Ýd|GÓ=謊àG|ù½zÈÀñZ<_/,¯2"…‰fáÇŠ;ú'òMÍ þ` Ý2nX—Yß¼îúÿÝ/éO±CÙa„uôÚ£®?¾7ëЭPVŠM&²˜>(>¨göüùYáòJ$«¦T°bl°Ul?¹Gï·„†u’ótîlyðübìP|å­2¦ÂiÚo· Ù6KyçÈøȶš„J~—ncÙ½ZÇ$ ªsì<~¿(qÀÎ- D ×@eòž©v3»Ë+3dJgnx²èÊ!2Ó‹Å^ ³ Š(Y §‰É@5ÄûÛµt™j_ˆ¯Ý.LÐ&«Ù»(©©ÌWzÐwXŠwñκîÿÙX>.×qŸò<êšÏ 2¥%»Ò{’ºN?îm…àl¤¸}™õUsѺÁ4KÌ>&gêÈX-êA)¹HBj7o¤¦ÂϲwÊÿ9Ú7j‘K(œ¹yͬÎs@¨¬ÚK2[²‹B.7Vr½Çc±Ãˆ"àÔ §š‰p™›ì_<‹Õ¸ê9UÞŠâ ¯ àv³ ÅQ<\\,ôù%9ѺÚõ‹“® ™+it×'1t´[¬¤´~5ºoă†¹ùÃAÕQºK C±Sˆäƒ nµªÚ|d Ï#«§Pqš;ˆ=IÆú¥tâC?*Û‹Šh[že‚†gIžKŸÍË'¿ëòÀI-M^þíOknÖîÒÞ{DÏ.:_ããŠoŸ†x<ˆ”zNOEc´h9µGT†Í®à¶Ææ … ÑP‚TŒ¤ºÏ²â âÑ3{›ß“+U4!ÈÓRF¼ß šP›b«žãÀÔ´ñHVÝj3sšö^¸¢FÊTrYÑæYÙyRršì u¸[øí¢h‹OjU2|¦D¯¹P®Ø…¨Ý´& ´míÉ- 1æÞP‡$%%Ö\Šk©U¦0U‰[{MÈâ—²¦ 2W†ht)+q1s´ÚÑC+àé|”É!¸~þ–z¯·ÂÊ}0bÞpk;ÜÉ,äò9#YÔaåeÁ|’2dÿÏÇhXw´Ñ?â£êpšõâ*”ä刬ƒtkòÛ^¬Ü÷'r(ðä3<£=“. ‘øÄáóñŠú¾½kh*£ýòïl¹­ H௃JµDV|E§t×ÝÕÈuœºfraCÇñꌴÿ)À‚ÅhìnPÚãOç¡}BùüSM¡Qÿ†”.ÍbŠ|¡ý½û YqóÙZo« MS’ÒŽ Dó4I—äð^ñŠ-Hq‹²Jó½éK–‰þ¤€\i4:—Íã¡6«‘US/T"“7ÉÁHû+{EösS|×Õ ÏžG‡ê‡‚JòDÉÛ“*Â&ÆBmú›ÅD‡Ë›jÜéOépå5‘Wà~‡× øHoŽæðâ7-nÔFKùÏð1ÐÁäQ’ç7ú‹4Phà£íéá•>CŸ>vG|à¡ÏtéÂØ™Ý,ÀÊ?ÀšGàmýs0!rKt´{¥·óTMvéS•ÐS2×/ cÍÊ‘Qcò¼Çm¹å+”³””{?ŸÏ¬ô*íIõ „Cbg+Š§Äu‚ßËqò!*u°ÓJù„f%ð9@¸8›=›ÞÊ#'íÊ-:Žµ ¡×§îÀç"Ó«9ÿl”`¯¤Ü¦çº•Žtjúø¹ÝžÞiFÆvMXpÛvב€wi-Õ}Ï{¼»‚Ÿð†,Ï¥…}FWS2unDÂ;ýhÇÀþp©Ÿ@ÿ¾×_CE¾bYìqE|õV«P M{ýtìÐ÷öꚸ‹ï¦êá]}Z­V z>PÒ÷SAkÕ¼Žß¨ýgSOFsé©hyÄßtfÆg… Óº¤©âéÛe«Ôë•\w:ÙÒoº[ó£BpêîÇôÊO¸!)¼:\òcþ)!‘z+°,smDñ/ƒwçÞÚÂè\`b°"¼B]sÙÁ6@*7e 1à‰åì­ÂœcHÓþËÔø#hÆ Œ²ïk'Ä ¬M›G/sN]®ù|“¤ÑžvNÙ¸UÈ®yŒ¥“à¨ÂÌœ·Ýµn¤’³/n?uÚë_lþâg+Ÿ™¦%¦íÅ­±Ùh\ˆ;mª”š;=ŸƒÄÅÀ  œØþÝmþ\6/E:³î¼¾DPêþÓÌtµy•wõË~2&â†á5^³]í`ìUŸ†Æ5tÓsÓ½ê¢|©+È÷2L[(Š2Ëî§yýâMr†ÜY}óÕ`~PÀ2ˆç;Pe"^hgZÓ2ú™­»fíQhhÁ¦ W퓵­gÈFÜZæ!Ó³ND#‰%'‹o½o›ù*Ø#”}Ô¯Ñçö÷42õt'É ŒYøñÚ”FAÐkm2l_òƒu/9ÏvpyÁŸþ‡+±‡È”IQ‰Û×\N"XÔõòøµþYˆQ[Ug*!©“OÐ.òmý£(øNâ° IÚ’w`Ðc+ZTß+§t/•àꯉ¼¢¿éoÛs»P4eéuøa"éaò¡®U8‡’d2u,¡cIï* âXÛ2U«¿Ï¬%à™¤ZÔ-|=äï¦ ¯ïëÏÏ>ÏÆv6’@–Ë à¯Ãò¤•¯ÊŒB¥ûí2‚;‰L{aR¼B®  áw¿­çù âŨξö=4"qÿOÜ K ÷î]KºYèŸTY¦ªç—›‡"Šó¾ÈÛ.R<ë¥ÛÖtmõ»¦€EÖçX¾è¯[êj{*kÅ–dóж嫇©gäQΛiRÿövÈènì޿Ѹ\øC?rш¡n núL¨±ðêj3ê* \= Øùòéí1™Ær^»œ$<Ç`ŸÁ²l¸t%J²óD:Ü9<˜º9h'àöÆ´þNCQ`X¯3ŸÏQ8qŒØ«=C0áç} Û²Àçô6é+WiŽuþ!éf{A'•ñV=õ9MRuC†ŒXQô½¢ Ä/Ù~‚` Ô¡ðax.n»—†llËnè…;­—t‡ IÝ‹¿ræì•ã~hIÇh©&v=š|¦ÁüBÖˆ›³üˆ•ØV»»‰Ž_‡½ ˆ¨ —¾æM÷›öúd6_Kto‡D7YDå²îAþ+'½îkHÛtì'6‚îzÚ–v3>Ù éW»Ó2::”OP?Tìª6¦æ§Ô‡ÅîL+84—胙´ômì8‘ „Ç–Ó‡kêã^C¡ÈMWvêb¡S1dq¬ÖçÐÇJhúk⬠·ˆš¯™÷5Úõé0R¼DÕC/ƒèX… «šå|èh · ƒŸJ“-ץ郅ÇÜVûI)%|P!mÃ,dK}yÛÿSÐgý¦&r; ¸D¤âY×¼¥?V`vŸÇ‹A$Êq‹4GX2ïðro”G¸Ò7^½K¸ÔG§²½óŽÑÖÅñ¥›7äm*ΣÃ*Íâ‘’'abÞúg×}?`ÁûFå‚wmgÀ¼Á<²ï«5ÒŠu‡C‡oô‡¹/¦Ò«#„óê¶`âLj]|e`ÍO3Y Eæ^4Èx›@Ý£±Ž2@ƒèÝÑ5>}ËXüw 2Üi)úŒb!¾û•vž³³³2´åE`"³ÌèØÏ9­Q;«1~Pšåòßò+°ŽF!e(uSEìDûÃwŸ§×ä8ÚcÌSJŽ9ŽÑK%æBŠ,¨ÀêSËZ`Äìµ]«¡¸ÏYßÿˆ³óf9PA„p”°\§ÝU|îžÅ²aâúÏõpRíÏîMåüíO:è¼ÏÞúeáÄy ùB‹_‚ÁC~DÐß [–Œ4ƒÞâÊ;÷ÆÏÞ>7ªïÞ§Eõ:'»>‘G±œçnN´g{|Ç8럖råB·™ö|¹3áDÆÉôá[çëTû‡xnLÕt˜óõâ0NIÍ ’"&÷;ðR'» 0å©UË›üafÏîJÿº5³Jq­ý ½ Dc ´ Õ¦m&ο™!ÈÛ,‚¸§N‡KÄï,g·MDùße{_!WÙ±Ùpâ7øÇÂãG‹¾}óîÔ¹WêA˜à±œå[Øõ„û‚h”JäšFHºÔ>ƒ;ëö=üC„ªa#ÙŸ¯ ý·Çmo©d™ñ5RÐD»¥ÁïóštSî¦ÜíYö ‘ÜùÙK3,&ºŠ_ÅÑáH$º3¶s{¿2ìžîtî-ip~wBˆ.#MK¬Jó1ª³úÅu&ˆ™ÅËj_‰¦|z†²È. Ê%QŽhx¥«•qÑøt°ºIÿ N_¸ô««@?ÌsÄa¾¬È#4•{Kîd¦ìš¸Çüñ9wõ?“yyD<›Á–移•ÿõ္ðø«mþ[àæªä‰Fùè f­Óþ‘6žBþ´l›_­×j\Ö­v ñ;d6¿jT”H›È ,vìÐ2ãh5_ «¤o¨Çˆ«Qf‚sMúy.é»N4ô¼ ÑêÚ{­/’!Š¿ üAÒâû9(Úc_JÛY’)ñøWïYà`CÜ}Ø(6MõY"ØSkÚÂôe;›¦Îã³^N×ç_ Q Æ\«8·ÊkTÜBFúgsÂOv}0#RŽ²uNHÅKr¦˜äs0M.úêX(.êÝ”c]›,,ŒíÀœ‡â@lð¶ÓÆQä…Óf¹…úurxN˜çÔõ‹åNqO"Mα¹®þöÞÀ6¾†ãBÑK{=¬$Ect°'«c¶ø‚ŸSt)zÜ[ö,ƒé󚵕96 ZÿIPÛK¸(NNó×ÙDS‰€n–nëÂÓ'”¢j‡N®†SžQ¸G0Åj7¹XQ §.­Ü¨îŒ!Vm=¼%Ça¦ôÃÁD‚¸Æú˜ÔqEBSŒµßdó6ozäbé4±0?b¤%Àî6€9ö»Éo0gUÞ¸„|‚¶.+q·!ùÕ•X+æîæ>4›M„qÍ¡”‡’^4เ{÷xÊÉýöáX7lÄäb{¿)í.B/a{¬ë)h£:¢I{ôa3…î+'ñ¿¬ïÇ#IcÌV³Ò1DÃé½×¥šRêz#Ìh¯¾4ÄÑÌþ¦‘ÿÅ3ol“@bÅK|Ò\òÔð»•91û;–‹s±Ý0a®=lӤЦÔÎÌ»ú7 s}aâEáÀWr‹á O{l'Ô9Ž0èÝCøCN)!Î7ˆÚØñdÝ€¥VAå©¿Ò£~xÞêc.Â"˲óCÚr–tÉÜPÌñdƱgÔ—\N ðêì4„Š¢A„éi¤±ù“¤‹•ùýgP õ·ú¼ƒ"mI£áTæTÝ|Xa9 µ”¨B“Ký Ÿø¬Å1²ØÁ. b<š¡‰Æ‚ª“UÞ¦sÞŽNÊ›^e¶z.ƒÇA;°ŠÅ^…‘_LgLjôÓøô99K£oH­}4‚Ò65ccwLCæAËÎòHiˆþT¼y¿ùx Ñ×m´êÉà‘K §w+6B¼žÑ5'¬ª˜ #Ò¾p¶ +½Ka ч‹Ôæ=™ë7’4JdŽÉ­r¯–â,A¸V/ŒÁ„¡Ã¿mD]h{‰»PÒû[¿$цƒQÖQ"j€r`%Z$5ñÑã;»Ïu:w.€:S ר°/¬ÉÅ“šÛôæKÅ6¦_1ƒ?Á:3œ‚b˜8U§i¼ô‘¦¤ ÔÞÝ–~ Mü,ôÆÃHݾŸö 8õ4–pÄp á‰U˜Ÿ"îj/ÿ-ªm¡f :‰MߤMÑO¬§p‚ŒZ ­c>&þóçy#P_¥õo§Ø$|Þ†¦F0kÀå”!í0¦:û–Òçþ=y?¸¾š_+æ‹fÓûå2ž›š¤²ú±âÕ¸_«÷n ¿ý˜DwÓYKœýäÁQDPfjì~GòLÃJ ‚)ÕæÛ6j¸/õ÷G”ü[¬]•‰ã»]ûÊzne¢³µ)„RÙ|âlÌjHóɯúWz‡?Hñ-ò«¹5É= ð µÛjŒÇ¸ÊOžs‚¤U /ò*{C…@ŽG™²ì‡IÖ­p)${6³i± üqå8èjs¯xJg”Júòî ˆXã`©±çò¦Wqùº¼Í¥Ä |åNA¡ë+  °Vn1|4PwûŽ¯‘è½³kèü–7@¶&ô~î‰ÌJ'H f3”õWQúâî,® ¤$LÌ]Ð?x/u>îV,DÉfwyrGËŠß…vÆ8Í(f­¹Ki®£¿ÏZÞ˜Y“^•­›ÛÚ"²©¬ˆÚü©$'~å¸uŒ¯ÃÍÆ; ÁUCî<¾éÌε?Ö}„ÿ¶W"*¯‘@¨Djëß|á³™“aY¥É( ÝhºW½´aR¼ˆª[0•KŽ˜ß”ºÿÚìö\º:©U¥ÉçœVQfꙺ.ðôzâbÚ~“çžg¦xSTÎYM꽋ð!”ƒ„$±þQ—€`ðK›ìÜKö¿1÷6<÷—-&ÁLñ§NÐ Úµ û, kÏ\ŽÁ\;±~IE*«ßõ,0À”ãá7òûÿY!¦ÞQOìmÖä’6Ð_êé!J:»¥ƒªìí½w›áÊaby¨…8ùÎ]Àòjê2îJþïyP]#©i ;ÙÊ¿÷laë²õwñØPÇÜï °à¸|¯»qÔ=äëÙè¶èÑJjä³H0ÜF÷ëšÜ1kµß(×(Äã&"À§ÀŽjÒº‡Éqî…ýcft&ζüú Ý,7„D$éÁÒIÕ˜ž ð0Ó/7;~_%,ö )Vç( Žé-˜Õf?\‘#x«l¹åº:¯fˆ­S¿êÅ{êÁŠ‚ Éf^K?+Wº" 2ôÀx Ž'³lô[ ÎeŽ:¾¹ã¹¨Ÿqš²µ¡Ê,êí&´ÙñZvœušÀ@óíã4"³Kuv™•–鶴nÒÝUZf”ÅSQ,.ì ¾¹ì7\iÒŧØïQb.il,n‘÷Üfôiy ÍzÐs±LnðʺUD^"ßÈ×`æ|©0è?ü)o+prn'A§ihc` ³Ðd„Ô+þ´Hg­ ×/¾â!5Pô¼•-'}Jõùr×ë‡õþÁa¶üf©ã*w˜¬N"ë]·sN úl-»^žcRd·¾ÚÑî¾"|_}´+_¤ÔÃÞ—p¬\J,£™€*δf S—þ32{ø¿çïº>¯*ˆù=Bè×öœVìÙRD³:†rŠ§:‹JzyB+6'|± _-šcÚVž^˜¯ v½@–Š¢Ž•Œ°Â+ò a#YFä¶à·Ø>c~µ2~¬’*%ör&ÕFAקÅXÀ44µç+ùmÿÎ"mV×ö¸Â s40žÙ!CS͈êƒ3'ðgÆ;…ïØ[“5’*’xEÖð¨í<™j¢›‹ív6ehŽ‰¯Úî‘Ï–øV¹Úé×tÞÜ‘^-œž½wô!–U‚‹i CŠ­r3@.oææd¿’?Òƒy-D­‡”¿ý•åKOÔÄ܆Ÿ^‡ê²r)JwMÄsÂ21>*r…Ʋ–Ñ÷Wê ŸÑçn'Ù]Uø˜ZòXhâ³osË> Ì+¬ðCú—ÐBù¸OÊVÏ´ÿÚž¸¤~áu(¾ “¨ðñ9› R¬aùÁ•é#üºy¥‹¢~…èÔ(‘›ìÔÏ:‘¦»¾dliZÂ&ì…{‰=*ž±’3uäx½v€¹WápôsÙP \¸i]ÄÊÑ,ÖžBò5=Glv”—œ™k+>Í…ôkrˆuC«5\ä J„;–j{+Óf9Õ$hý‚2š‡•GÉ€í­œ³ÚdfÂçvO'ˆÚÊar¨¤†žÆu[B uëqÈ8êäî£Ê‚¾âdÓÖÛ燈$µÝ{zÉ ¦ˆ 7”¾}Qh Wöy˜;´+¢2g¼ ÊüéydHYFqö+…Ú5rYûê3óèd9ØÖ𣠺³µ€pR-¨‚ãOŠÌ€R/¥dŽŽøÂïÔr\B;N–t\&„݉Ë•òÄкï¨})dõõ”ùµŒOm‹žÔ¸6¨ÏU*•Ðçc”T7atmÎþvÖºeÀ¯À®½Ñ^/ÛYÈ)wÉ>»× T™ËcÑ% ‹+Ë<ÊÅA¬)(˜gŽÎ bØ®™W>YÇð~› lxÐÓ2FÑv¢,‚&ÓŸ"Ž?Ô˜-xe­‘‰¾}I0ÓUíDN”d¿¢a·\\Æ4féÆjéü‰e‡,½ì áŒKW×ÂÛ°U^b¶…Ìmî·Ê?Þ¾ÝàÌEõY÷™>òwšúÞþˆé“¦°ßŸÒõs¨áR³‡ênª)AG=¿ÁÛk„[Ÿt¸Ü¿XûÿvÖ †‚vƒî03ø;Ñ„[ÚÑM¡ï)" a‘r\'õŸLÍ•§pã[§ó=øWX™8b³»¶èÌdgï‘B19Íï$‡…Ý(F! G>„ŠÁbNéî¼õô´ŒÜìI$¡KÛØ…$œðµrÊ"È$óˆ«Êv“³U KçtÀŽq}ð;±~Þ,%C[¤«ïÑ!Io|{ô…<+Ô:Gù^½Yø÷Þ$2qƒ-Œ?3Tßa·Dïö¦ÙD`×äÖºòÊÅå(©%Oô§~%ì݇$€qð{ÕŽ0ÐÜl³³{1œÌ$±}†)õ½Vµ^#ÿãg™;>ü× 2_1¬Q£äzDUGã"'ÙEüfA5Ñ ÕEGãWsÑ¿KÞÞh9@¹™-aÿIQ¼ @|øñ+nÁÄ©Š}î˜r¸ß+€˜ç&M·ùDøûÖ½†RÏêyßöÕÐn-ÇvT§„ÿ+m x¶úatQÝŸ¡)‘U)÷ÝÉÃ>‘‹Ð>'ŠÄ 2e[莩‚Ú…<^ÜïoŸgQçØûÉi£"hG€Ž!M˜s~%vFÊŒQ¢1Ì^9Sœb£É÷  ?B%/Â$Ff¿c­¼”Úd‘eþ¹ˆïÆÊÍë‚+G*ߥùé1ðVì¸0•ÎLbþy6g³›®ÐÊ–L§\bÌäAµ0<‚×îo4Ön¡% _•".f7ÍI&ƒÎÙU€ÀÇüì“€ó‚ øÑ e2F¥Úø{Ù¶xÖÚÊ_ä§HyACVëf;y+IÁ“–xMdó è¡ßIh`RKhùEØV¼ β@Ö=MÚÙµJ½•)i”Ý“-\]n‚ÔÑÁÉ*äÕô5/ÙùïêOG…<ÎDmÀ̧‘Ò`ìÜd¼§aÔõ¶9“µx*¦™¡Æ8*n][8i–ãȲ¹{AÜOò@2¸á k©gBŽ4žØ^K$4ëoÔÂYÌQ¿–ÅÐR{ex¤ƒ)*)ì’të7u©œvëµZ:wåÔ„¬îµ1µ)¦¡s°Ø®X«UM \ €ó¢Xc­¸ÛÕÜSìîè/É=²xÞ뽇¶Ð¥?Ž £ª¡*ë æ“/r&ÿ–‰D=‹á&¼üÀ.ˆ{úÞ Ï°ü³±‰AfíxÎs¡‡H¯l+ŒS)¼uH³¤x†EºÐóÝ2cAhBÓƒÉÞÀý¾–¾6b¥yÒ,rZsüÄ7* ØR}—kÎA›¾þi"ð¤H‘03xªå†c|!ƒ­–/ bÓ·mÍŒ|ƒ¼b©µéÀC c/Aö‰›Vuð Á±þLjh¥\jœ+R ’ƒ§gK§EïÆqÑÔ¯Öª¡™Oð‹ä¥j£oW">Þ-! yÖ?pØEgÑñË)¯ŽBó ¶Fú¬¿±%ràYÓÈÔü!´huÔq}ø-îŒòs+-ªktž‚sÄÀ‚üV‹ ¡øè#y¸ã„ÚñrÉmÒFÐOƒƒñàüVWŠ6¥U´Go1¹œîº¦ÚܨÚ. E·#4©NÍÚ7o£ž>êü "JüXü^x!Yö|̾i½0ôÓî=3$š§¤Í¦rn·Ù»5Ý"XE|oÒ. :ÁïaƒÙž55f èøÙw/B‡ku \DÛnfb)ÛµÅ9ߢ(&z]_Gè1š *<£ßFNýk°žæŸfþœ<ôNä~,UõqövŠH@8¯:˜aÓséDþü#h–¦;¶'¾RƒFeÆz«‚Ä\-ÙrÐ%jÅÉ`ÿûë¨ïé‘žýÅ Mƒ e'·üNOª½-ãA9œ·úh¯œÂñíÛ‹˜Õñ Ì}rÖr¯Þí+†³Š?FÆÁ$JkßkMÑ ñ9D^¿Ñ¢àÒ»K<¸V! õäÚ–ž¼ZM@5Î[t¡jbj*D¾<}nÊÎÚÈ(••ŒõAÊ#,7×öó•¿—H`îS»ïâKA!ÄO~´ìó{1/ŠY{ÂëxRm²‚[†, } ³'¦ï†@îjPG£jhÎ…µÏ¹³M6Q ÎP\Š.×õy×MÛüƒ¯È­ç¾°dñú…ýtÎoõuÂdÔÌ1óëöu|š’q= Ì¨XÆé|•à»››rZãp²ñæB“0¡ä4(ºi÷î¯~j¸Œë¶+’鬛3ïT§"Ûë0”ï¶ìZ,íÔP‰Lû ¿â,Ç(à¾4¥cùpFYóÞ÷E¤/™F7sVJÚ•€Ž8§» ý¼V`8’Nƒó¢`¸ø™AÙ²€æL¾_x|Mäb(RÚãÂÔ‰ ÀTÀï—Öáç#:‡Ë^ Þ¯"‡‘»çÌûÓt»qJ »Rðz€ß°´€Nð¾iÖxŸ¡Å¤ÉdlÐE¢±tuLæœ`‚ž& ïxÐ{º·è,ÍnÛ´eçºTé}Æ3hËÂa5Â_É0²wÚÕH.g¿cÛqvÙaMlÅæ@z©Çž ÙâÇåT÷c lÓ1WÔOÈ"D“¥¡  4q/¹¢,էϩÛyUžÒü,d·ásAM‡ÿ•kBüdˆQl>3æÜ؈ÙÓ6Ms% %ò¿¶î$í† É¡ŠFêq_AÚ­ÀÆÁo¶ùnØq7’CŸÙ;ºªn£}^Pû“:¡êpMbÁ¨ÃOœ˜\}»?wÆ{¶ÝžRHË”´\úyˉ’xKŒh&à=¢ŽVé6ÔbšûÚ7Wñr‹ElãÖ­·þÙ]jtcý’:ç0ÌÐäØŸòè€JLZ2Ù'E0ºc‚D]Õ2º_Æ’‘ǃñ8ß3jGlKŠ¤—s aò^µó ëbF4¿]›‡óU_Ùœ^ÜdôÓ2ç âËG¬6k õ}rong-Ù)â¸A9Ošd†bm]šf¹ß}‰¯øJíZFßϪ'î©ioûâ'»Q ówcþ”.œd`N1I±ÜË¥>¡™º0zG³°N”•ºmZ€Ádji±<+òöÔ¦ôbuî bl¢NºŽp³[ôqŸ™K»±zx°g• ᥬˆ‹‰;‘)‚7õ¸Þy»³ÄÑç,a¡;ÀÓþ—ß `^1ŸMD'¾Õ?Ërl×TÑ•ŠÏ´ûK'¸BÈÓúöüÙ&X‰‰7Í?8µvË*Ö®Œ¿Ãç‰$éæÝãë"ãI{ß·ÁyqXl5¯‰%ÌIŠˆi(ô:b)¯/)ˆÚ@À(é*9ídr\æGXdžz¿ÃóR á(Ï]ts=6Pô™Ïí¦K0*þ#Eù?&Œl[‚kÆÄP¡O_¨•Ýþó  BDƒè3¤·“ƒÕü¼ê}gbˆ¡Ï/e°ARüì¶5!<÷…žß…¹Ó\¬ƒTE&¤U ³Ü i‘ž€l h ÿ ¤´x`tWàI®Ý¨Ã—b »ø“DD©…ÙÍE¤@3ÝCŽJ6åGby¾ F¯²ÞWLŸÍ¸‚˜…N£;Á™Aò€NÌ„wRw¼P>"*bø@·K9È™•W*¹}b}äOåØñš3¦C¶~ˆ›t¹ ZȆM’l²Ãµ?"ÙcÌëžÔPiíì¯43ÿ¨piâÄtª»ƒâ^•"ou„4ºÊR—Ô%áèi(ʯ\ÿú*sýæl€…/st=†Md©-øŽXõ,¾þàB@»æI?@ÕçÈÇjHíeœLó³T@¥tY//rc´t¼¡ýH>ÞiÙßÿ&$vÌ ¯‘žSÃ2†PÎäÏ&õ*f¯æÀ×Ö=–‡§aÇ߇¬ÔÝm\©šèåXy:þnTÿìåU±ÐE*V£àÅ.{”nøì¡Ñ¸i¯›”ø¬Â<!8¤cw¬»¤„xõ{€ÕÎ!c‹BevRU9¾{T@#ÖîðŸ´"©àw+²äNXnºŒ0÷d-CmAÂJ§%&7]¸w¹T/»õèa}– ¸B…ÞÎQ߇*¦ãg@8ÌñÒÿ.–j±OÀ™ƒÚ™}7“×ûî×ö¶­#ª±ÁÐÞÛ}é²Lƒ«D½¢«²Û(¢d÷`FàXóa”·6Ô¦ì2R×ÅŸ`=™ùV)g×n°€ØqØ„4 Ô"àOhàT;Ò´œ×bÍy(ÇÆe!óLY=LvZøùAœJÿZ*ŒÅÜ~yš·1ÿN|KÉmÕev·túš¿­¥s;‡IワQ¶¼L ¼Å»ýƵ´ÑqPtO¡_òü¯[m*ØBÈ–Ås~"»®“ÓåÉ`±˜Ë7Ïç Ið$MfLçšMÏ Ûçüý €‰Nó‰´™…ë}…-tuü¨ƒÄ/ÌÚ ÿ7ÀµÜOªî-mš´÷êJ*·S~¤k”OqtÆ0Ñ„i}¨J,Lù?R$¶‰¥¿YNG@±Ýb¥•‹œ~äɈ|Ö$]:.éÂ/^à›7ÚÿEíŽùQ‰î¤"‹ù0oóÓ5™x!kˆ2ûœ&ë&Rƒ]2zØ?"G„É ½fbŠ&Û«Tää‹U IÓö² \Ï’›ËÜÒ^ÕÃÀS­ê4¾þÝú¡g>+…wÿ}”æaG†]ûHrEÄ/ÿÄ<¹Ë i¼Û.)ÏôU(éÎN X Ìê¶èoö­j ¶£sªÕÉ¡ìäà &TÉz¯GÄz4C*â¸Ï”ãÚz™úkÑ{Ñ Ë™pMñ'Â{-€m¦°àñ}ùƒ#Ý£$”YGùój’rñÿ˜€HQÆÅFö 4ûR ÒŠÆp¨Ä%^š3üû÷Á2ßló3©¶bhüØ0x3›O^á­ºPœXyA\†ºž:aLsF§šÏdžA¬¹SÀ³¥12hc¸)Wx‘ã— #'u.Äjor8“ê&«/ì9­{æÎe¯iJU³‹*þÔò5ˆÀ=ê,Æ«tØVö5&ƒ“õåtÄÈcm|”Ÿa{•¨ãf3†t^ÒˆÇË?|Û; òbÓ†«§×Y é½<æÖÒÝýÐðr _õgkUÆ|›¦{¦ ÓqYK´Ó"{ˆsÔé7îœÜÙ»E¶à›>¹§YÈš1H¿òꘑ.¥¨ñˆtË-9m$¸—ʲC(}™NÏ2pIùcF£qÛÛ²aÁ_ý«Ëµ7P]Ú ÐmGC"%ÈÌáË™ÃKƒ²6*;KÅ—÷¬jÛ…~–ÕÆüᥗ¦|åòûäRšâá'0¶±ÕÆŠ±=¹]ÕAN0ò7ㆉlÆŠœ­ÖÐQ±—ûÅP'j4ì·eìDŠ#Ì4#I3˜Mõë;Þ(áÜLŸ­çM®VHâàÝÔ3©îž·§¡Ë~ä‘ù:=ÕTÙµ«_?OÅ£ã%í[+us¼S˜ˆ~Í@ÛÞ)ërñá¥6å(cáJ"Yí'ÙYÄ:ü¥;¤’+ìú¨á{5ËGK~´Å#LžŠ•d ¨nôT ´âcxöyԿʨõp)qÊŽXšÀ'.¡©ÿ)ÈàÝÐGæ %zMyúÑgÚ­|š´ê£EÎi˜ªáz’Ç=n)¼o¯Â‰ eˆdU¯h-Bƒ·5¨Úú ‹™©’ ‹|Z³gÛÝÅ‹^¨§ÊsÝl wb¨__Iµ„Ü!ÀÝ@Ñh#A`­9µ%\»¾«âèê ša7hˆBƒhQݱúìÅöçîûsyHžF¬™yØ­ÉkÎDÖ=kʉ$ÓR¸îlY€¦Çq6v—’ô™¦kðOÀT¦7kÇ£°ÞÂìûÓÕØÑXŽVƒ1$ÛHŒ–Ó;}‘tUlC¡íqéP ¢¢¾½íß3ñ«U¬öàCÙ‰Bãzê‹YÎHi©dm¡4*}Xœlà´!–Ä]µËåׯO8J,Iˆé˜ÓN©°QyP£ í¾„6ñ%E[ŠGóê¿»Æ%0é,\’3$Ól6iNt ·©@’£‹ÿ¼k„¦ÇRÉ•êÞðW;úQ6Š¶ßÔI¡³1šË9M~¾òøyG5«„ ³{²25äðáúú,žôpƒÄï›ß°ÆÎP¡s#i5Ý[Àé¦Þ» ª6ÎÏkä7Ÿ@dCÊ@á‹,<*}w¥°gŠØÞÖ§'§$–$õ3R"œh¤šñÝs÷’¼2Òz¬}·iˆŸÄŸVý€“Ç!÷j¥¯AÅ÷ÂÊUg,ú”˜ŒÛý®ÑzSéûÊÍ$ynPÄf[ûžY§©ÎI#Y?3BUé°¦ªøʯ\¤¥ Ü —yÓÌ ôÕûS£ž?ó€‡$Áˆ‹¿ÌìëÄZF§¢á¨åî.¾Œ endstream endobj 283 0 obj << /Length1 721 /Length2 16693 /Length3 0 /Length 17293 /Filter /FlateDecode >> stream xÚl¸Œ®M-ܶí~Û¶­Ó¶mÛ¶mÛ¶m»û´û´qÚ¶ï73wææOþTRÙX;µöªÔ“‡ŒHÌÞÎEÅÓÁ”‰Ž‰ž‘ ,¦$ËÄÌÈ`¤g!#q25t±´·ûeèbÊ P75(›:˜˜ÿ@˜`È"öžN–æ.Jcªj6†&–¶–N5{{7Kc ¯››› ›³+½“+?ý?Eʦ¦ S€™¥)@D^ASRN@).§ 7µ3u2´(¸ÙXd,MíœM©föN›ÿ8c{;Ëqr¦ÿ7;7S'—ˆ™9ÙÛdEU„ÄäåT¢" *"C;€ŒÄ?åv.ÎÜÿ MlÿÕ-Àå–áÿ,£ÿYÿË:yüÇ¢‡ab˜X»ŒLÍ-í`þ¥›¤™=€ý?aW‡ÿ¦þ!äü?å?RLLÍþA»ÚØÈÚš(Eìm\]L²ö&¦Nv%{[C»ÿ¢ m-m<ÿqÿ!ébøBvæÿ(ÇøŸ¥³˜¥‡©‰‚¥Ë?R›Úü#Ñ¿ãê¦ÿ¹YSKWÛÿÔ›þ£Ç?7jbogãùü§È £$¯®ªHó¿øwZÔÎØÞÄÒΠìò–†N&ÿ ü;­`hùŸéù/™ÿN€éÿù²†.N–mFúã¿Ö-Ýÿ‡¶÷ð¦câdб03˜XÿÙ¸X™}ÿ?t]œLí\þ-ø?½ü×ÿ÷ ™šz˜Ãl¬Úó„X¥·…Uú‰/TÓ,̨Ҭ2¯ÍöÝh#[ŽX˜è]áÜ«à¸ÃáÏñí]îJÞ·Æ›Âýž‰!ŠÐ]Åy-%©2÷«Û~W,|kŒÐ»Sö5–»‰¼sÉÜÜR žaÍÎ;¯“r,çCgŠ1¨èGÚIX_z´§Ò¥@ÉÒú†3= ©EZ°-º6xÝß /Þƒ»!ÃÖ·(ª t8 ×|`£ìàÚ]•¼·ír´)òQjG$ÄðÔZáÞ#ñkø£:¹Ë=à›5qÔ8Òô¼«ËŒ¾FY‹¹±YõÐÍtIÆ ;‰[’x¤iNE°ßUÊrRME»¯A§ƒ0\A雹´WÁFLVÏô&¢Î~ÌzÅñ~F8Æ©©<ÝêW“m‘ñ¶^²ƒ¹$‰Ô­¸EÆMfÿÁ4h˜r«øûsç€hòÚ—d^Ó«®ÏDìL‘´ï@|ÏC(öý‘v<Ô»'žݦ`뉣1ÊP’:¬³µi–NBíÅ×Ç{•cµ;'8_¯É¡=y8¹dËTÀÔ wM&Sý³ÃíÊsædùª.¡ÿ¹ÂºïîéwÆ ùú»ÚÄ/œØ­ÃhtÉàÿ{îוp÷dóvmêÓ&c.j‹sLZÖ÷hÔî…EE‚J(•e¼Œ=—1<Ÿ5U°e²‘U’OC_æÃhîŽÜ É­õD±ò!97(ú—û,î×ì¹%J@WH5ö-!ÿ/ºîÊZSãtgh[«åÀ4Éó>Ac#ÀƒÌ:ân† Åß#Jl+M>ߟß„'9ªÁ©²µ spü[ÄjO =“¡²¶3ÐË¥ÿ›Sê[ŠÓ¿,Õª^ùÃþ,uå‚nó/â \oäþãž?––,µ"U1^\¤8¬zÉîd¶aI'—y#,ܬ$‘éþšë–Ÿñ”ÑñhPÈ”ÕÖ·›4ÄóK8ËVau: {¢³‘9¾éOöŽSª›²wQH¼þ¥[Êàþžf¹w­M¾Wˆ1^éUÀ¬“U²Ášø¯c˜ÍƤ`¹€þüü Š¡ueÄ;Ý< ð+Ðh³±9ÎJßúµgŠ6-ld;/|Év2¦#M‹^ûžd¸e§{òÓ.4°—ýÙÅËfù… fItG Ž>tí!m‡4) )ò{(OODes,‡nçsÕ?†2JçZT|‚ RȪèþ(Ž…—£’‘V˜äE¨µjl[ÂYЕv’Ïæ^À˜±¢‹™©ä«Z±®Šì1Š8rû ¾¼_p°ÈWÞâÏÙÜ—F’ܬe±¶|.¹æ‡æÑ?ôñbÎÞ‰ôôYöú3›JÁIYf¶­ï bkn¥ž4ThàìÛzU–kQÞ›í÷êž'ík\‚’¦dÔ³¾OðÛKCW?î^NdåKÅÁìýTs豆¤ï+˃oúˆ<Û¯žÔ̼¬ZóØW¹V¬Šj9…€#>âP1s»p[»jù ³¹îN“Óé<¨Ã9³=à6jžz[û DóÅm,w¯žcç¦]ðï„šâ²W,ISý¯ó¤óë`s=A>ÊÊ,( )ÒûìtTïÏש è} [!Æ.­ˆ? 8„ƒÊ)ÉÀŽü~¾û|[Û‡ç+1À»jÑ#¨cÀOcÕWtr®J*Ö#ÏÕñÕ >ÒädTêç•8_î»,¸ÿð?M›#ÇÆتدùd’(€Ý×:ëÁ…çoÏC›8<âzlˆÁ%a‹|/zp·>UÃÜ]Ðå–E[x*54R)ã…Ø餎Ð]]-‘û" ëA»}ƒ@Bè6³Á“B±·3‰ˆŠ}ì g=¯E(œì©˜òSkíÉáåíßK¿ÉK¡Ø͸ܵ &,¾¯ã!•ûü¥£ã¡~€2.@»šö*ƒ-Ní•“IXÂ%Í­lë˜OêÕØÚ˜&X[ Åå+•V²‰ò§Xž0Ý,£ZN× rOqgOš¨_cð´NZ²ŒJ¬Äpg©- ýõù Goç#sÔšŽóóuþ ñ´\/†>ÍGV‚“öËWRa†ý+P¢Jœ'ñ‹Þ¢å²­sz˜ yÕâU3)2’\àìê¹'¨Q_ôÈx¾JÄdör's?û;ó[¸ô¹<øn iúh€*©}ƒŽçèÄ›óNÌ[ΚgÊV,Bo “ž¢Kw*Î:¬[ª¯ÓDœ© ^žb„*Ö {$½eTæ:9WVj Ý ̯ZÄ”H÷¼Ûe™Æ*1) _ôÓÂ#ºn€´µl—e#œ%_tL¯¡#_ÿ- |ÓC ·£áÔ×*l­œ['¬+rÑ¡û¥5ÚQ¡Ô×/ÁuÑñÛ{öâLÚ©‰n_2ÂM9>"ñbN7h~š~Oýòë|dão˜®ùþ¸>KÀö²KÑùöç¿•âÍB`á~)º>È°Ò«ÕOþÑ¡ú0èŠáIvTì!af—Á¯mÞ×̆ô¬|´&Ó+ZÝbJÞ*o1åa¬ª²VFöñw2 wNp•yYéò»ñ@ö^qý›G–ÔŠºJ?óNöú­ã—¯âŽ«Ã{ƒ°ÃKþºDÒÌ z_róéèY'™+jk²•lÄ–‘n<ñ¹JÇ/˜~ˆBNë<_mÆ«¡k•–À[ë];-Åi—^ ˆ û]ŒÕ,‰gŽËU¿fª|/CèɉÞäZ©É%Ùy¡q—·™‰ñ+²ÊÞäß ±)Ý2N˜‘—÷Êw½Ð¤Õ½»˜®´¡E‘™«GE„] ýÝ{~?íÓM ß˪±(Ì?‘Z´‚eüEÊ&lÖä€.Ùýº-üö+^:(Nú×ÛdwñÇ+Úú¥Ž©sËô:VHqÕû`´všpD•qÿW*CÍxH©€¯×X9"2ˆzÐÕ‹©à)eyl+›l=øüø”©Ô´Ùý³/Û8h¨ ¾"ãž(("&ié_hæ{õmTË€Tƒ ô©óÈ& N?dÎXã9w €Ìd'»×\'H”ݚìAè« ¹æÇHü[ƸänX˜ŠàÐ.G„ÀPÿÒØ!'UxŽ·š B#_IOh¬N‹Í&· ÿ´{…½äWs`~\ºÏ+Yï­< Ò`™btïÏ€ ^ø_=T­Ð;2Y‚€ƒªá×· RbÊa*®Ê¡h!(tÝì<{™ý0ç²>â®—¥ê?ÿ~ò‡¢}ÃcS¼¥Þ“ÔM†ùpo+gkücUÌ/ï‹L¹§î±;‚¬± 4Ê_p®ªÞ`º)ç¹V¿¸õ}WÃ>“6·õ„vÓÉ›´•øBo@ŸÏ€`zãÂ6k: ÔT’®îìЈ:'Ÿå@ Ž££uÐlXµ @‡sLZ²ÈV´~Aps0=qŒ—v„þu üf¨jÄDï Pº;z$E•€)ò.d¬–0¼L ¤áéëâàˆ2„òˆmãz iø |÷I´‘¬´šŽ<“f¤”áÎÌlÖ±ßVÒ®ºù£d®;°>ÁÅpî/ød9’aì+#dÂC£çŒ ²êßÆýýÑ þvIZ¨û@d+._AC5 ªj¥îŽb˜`*‡ñᄳ6éȇÂ,¦Pºž…ÄòØ‹$v?Æb_–4wm¡lŽIÍ ‚²Œ S½—UûRo_æšú#œGÏXw:ßïHÅt¤j>QÊEÆ‘ƒà d1§ ®à„Tü–V4A¸)q³4QY–!}[æ6¬ì’KÎi êZFyªÎªôp÷ÍYÞœÓÍÉP”BeŒ5gÛ×Wð‰QÌœPÁ1Q[ Ïá®ÉS¢ÞÉÈù4vï–Z©-Lü< DG·íð•-‡G“%(®ã²@@÷‹Ñô‹Öƒ’;ø„…4(x ä3ë0žµô?ƒŠ çâšþ3Ô^¿×Þ é7ÂÐÖp†ù"³f¯2³i×msr¹Ë˜W"©*÷ø µ Šjß)¡ê¾ö¾E†ðH•a–bÇ üj¬a­ý÷ÐfÁGÌoµÖ/µ Éñ í‹×‘)¨—ôJñ¶€“îŽßŽ{&Ë—÷‰„Ü?`ƶGc˜ H…¦QìÓSo› WŠÈ¸ú“’Ù£ëÈHÎ4®Î‰«ßX¾F;êÊÜuJzë#p5q\HŽ’uìkë·ˆH´ÛíhÄý´Æ(#ª„¤?Áà;ùÂ]ÛÂ,CO±}¸£°MÓÒ_v–E"ì¾Ó¢»†,' Ú\Ï|ºz<®Ðü–ógþ¼–*(é%uq5T¶ ávóÛ"jMt»ø¡Ÿ"Sk«öÌo…Éâ_h®·žÊOV«ÎÚîƒõ4ž1Ãê’,Îà³òFDc9cnûGj»Aü‰’Wð 3fû•÷¯ÛñœÊ]Þhº´Ë4$Y&\¾#nï<! ˜d²UGê+ä{D2Yx3ðÈõ­o&¨6â¡Ñšk&P™}6ƒˆ‚ëq”f9osði®ÂM.uøXmÒ8í_+%Çòˆ@"˜Ê‹?å"}×þ½ä¦ÔÁpW‡í•lD¨«ód'±ù”9@ãØíÂñ2ÈI¿Tû!ýÌ35ò¶[û¬«KíÚùÛe+¦W—ƒKL”š°M1¥.Bªª×Î 8· ä!‚Ù)úè:=wR¾þ¨eªi;Æ:§ÛÛð¶Ô¢*îU EÇ¥•G!©T[ZKúcNµFV„ ›ðöÇ'ÇIÙ ¤¤¸Âà´Pó­Þ㺠? £ÆYZžhß›í ‡ËÑÖÖ‘§ê¸"3奋Âzæ#´Š[ôÞ‚ž£[’øçƒ-ç÷ ¢ŽV1Û¨÷`¹ÖZchÍÖ®,eí.Æ ^˜:m$p1Ûœ~Tˆ\dt6±7o€©>’0r+fàççÊ0!î üª8»2Ù!’×BùIïËö¥!Î$VsŒXâýºÑH§xx`ÞÚdQ’-‘›™®”#ËCjÿ”|‡52³yÒ&_9í¼áæDt!}㵶:þÆ1s41Zö'‘ÁÖR¨îˆ˜*}ÓSÊwòöâÆU˜f=·‚s L—ç ]®·>ÿí±dí>LGÓýP´EzœG‹åáUþ)@¬œj½ ¶tE"v™WïyS ±ÛÓ(õݵ ½ ÜŒe¬'ü  yä&®i.tÝÐí“×ë2”Ý[Ä£“»ÉîD3}ç׊Šð‹w®†”úH% gº² _Ûƒá39]º¾Ô\~#¾‹Qz-på%Ó2ÍCÞ£bèŸIp%¡Q£óoW«1¶ñÖ¥µóäàÄ–ãtÑKú†#vƱœ‹¯ˆ™Óê2õ²Úï :Á=’R™²¥G§V<ÛN<ÜÉÃó£ZZ P4úKhº;c½z`¨à«¸?ý¤ª5ôЛ5K0aUÉA1QPß“9Ôª7Ò[íÙ3å!mQ,‡Û°‡1ö«×f$Å’ Ó›@àÙV à1’I&œ%lô“ÏvÎ1žuF4h‡ÉÃ6'‘PÉ´ÿ^t¹]gôÿm¹­¤ZË‚ÛB±\ƒ¹Žž>Ìé¥qÈ@Ÿin8‘ì‡íò"„mxŽó2Æ2örxí/µèIÖ„Fò»>¹Y9ØTgKM0fÿf®"P†T`‘:œjîÝEa§Õ ì÷Øôk¿·ÃÙö ,†ÞfBG# þ9b$‰_WÏ@“M®z¶v–>*wðŽ÷gÕ+ÂÒÏ,VÑB±Ä¶ê­y’5«ýÂqvz:±(µÖh¸ ç|Öaõ¿f÷òmtySa'îß—_²©°5ò6–%ÉÉŠFïòíkÄ€¬y÷£ÈÎ*(Ê*ìT¥hƒp{°¹š«(ðgÛþ Ö,$âƒê¸zÛÅâ¼½ó¾ÑWŸIfN›Éâˆ}v+sž¬¾ü¥Å؆máÀ[çµøÄZÞ»¥‰5\KÞ ãISæè˜Ë4mR.SÓWü²4–ÀF^ùBÛ¹ôhÉ·]è×Ù—%ñ°ãÇ}7Ó U…Õä•2›ltžÈ饔+ó³åHŽ‚ÓŒ=êhªrÇj,+ ‚Ðê¾Ôö …ëÏ;¡)ðéJ‚]o°ìLözGd÷H ?3[X©`¢QÛR]é±µ§êpîunŒWJõ4«²_Ÿ'Ýyõ…ñ"¬RŸQ8G¶2Uãu‚ tï²ø€ÿÅ•ƒ¦K‘ ÔÃ5;þ• gpò‹¢ádèÖÇ;@“xöÉû «´_ãutxs¾ŠnÍIS0É#¹!Í‚VmÙq’T½P˜Zã¯ÅÅ-é€MJv&( 2ôçk3äxˆ«².°ZÓÛ“63o^]É(y3å1;]²ÿq{"[3N´*r¸ó" ‹³ÐÁ ýqÃaÊè„Qð…¤×_Y"ш(¨eä@ÊPH«oÔtd]rý­NzטSújȆgÖŽa ¥¾fõð1¾ë”µ¬ø“JŸß¢×k”%lûŒŠy NϼÈzëå9Áã•0‡\„S ‹KM€nëdÂ`h„i×­“ñšáæ{r¸éà…z4YµÐØö…£‚?µ”j«Û2:^@³ôÝ“Åå„!¥ 4'ŠÆ.€v8¬œ‡B/,¬HÃçˬ¦X*¬ÂW®¯,†¯wÜH¥°Óã;h]Ð^/ÕS‘KFÿ¿aÃ1xaST¡4k$CV d‹.\b ðä4”žaž¦¢}€Ø懘-¦C¦Ó„q¨¨¹iêäq ‚Ó ;“³o o/Åzòz;k^¸µ©ÞìmÏ Êtë¯*ju‚Ù÷TÊ!9ôàéÛ,žéÕfS£í›¨P‚¤Ùºëqð»'$ÔlJ Ï›hÏ,Ž:3îПäcä­bE¿°ð¦j,n¥5¦†œ'ðrþ01ÅÚñ1!´1uÞè¾0ØSÄð–± ë<9Îa šJQ aËôe)z‘w¦„ÙV`I"d')»oç€xz W©­sX¬GI ¾¯¸!în‹âñÔŠ¸äEVp?`¨r1׌X¹™â“ïyIcpkK|xü—©tVº·jEg•‰QîÕL!œü˜ÑÜ“Lâºf*>‚(s/ß¹7±Ðe]9êlã•ÛFëq·©ßªýw‚Æ\Œ‘tŠäò·{²|¡`\{Û÷üzå…¿Í+¶o±'çoiW«¨#³]€JwVë½úÃppb¶[TîÍÛ/l0 ì¿1íEys;îêV\”;ÒïmÒ]X}Y¤Ýß%iwŒã Ó‡(\ج…nì´ªÐÑúl yãì.è§jηœüZ\†>ÙOÖt—6ÐvgV3º;e1›+¸Ùÿt¸€jSû2ž_ `2¶ß ¬%ª¾ M•G¦ÀgQ鱌\Çú5zõ¸ÄþViô}[ªõš1&Ȥ}ºYiB¦‡J¡¤¹ûî¼ÙV¸@×UÇ.eÒmƒ’§¨ *+»¢Q.»?ÚqƒGwMᴠȯœ]£žŽ»:èeê+Ò¥ƒd5Ý*v·ZÓËZ;ìºo‘䣹޽Š—C_™Y…Ôj]Å À¥ ˆ±Þ¸ HCÀ~k±îO¨0×U{EjkœK3Tö ™pMàÞþÍXZ¦º3™´°Ã ÜãÍ3ûêoà¾>"̇ôZ4ÂV’%'îébÌÒÚ±2‚>ò“x_¹…À=7„äY=#NÜáùkˆêäxË#ýð Ì5ÒÃÀ:‡ú’N'ìÌuÓä—3F Æõ4cì¥þ¢1âÚvßÔ¯P ½ÔÑyHÙÍ?ä ~y¢—A»l¢1¿ŒáNq>•õâ…ÉÕŸS¬¿Ô®¦zZEõ[ë·"&Æü[BŸO©ý±ƒiÈÚ%y}(‰ôÈu±x<ä#ôJ”™+ò_û^CÅpÎ*ª2úð…ùî!}@8G憛Ã>:ËèÌÆchAæx/ÂÍׯ¼€_M$°Üòè+€¤¬ŠÛ@ìS&"Ñ·–À™Aߊæl®\ ~"ÊœÒÑÜc_BµÂ‡Í;ÚX¨äߟO\W©®Yñ$©&á2Ó)Y¼>ØXYÑÓdKÀý–¼Qp"¨ª·ËKç @MþÀõö­i4íýA¬añ’;®T×ÿaqÈE¢˜DÉŸºÏ&û&i(ã #Ò¢çsÜó Ÿ”8´Â;€ØV„§õ…=ÛFwMzí`.Þ†Î-ì×5;¸Ó¿ZtñIž³âq2ø0‚QÅÑ» ÿH‹ûfM{qÎöŸ  žVG1Â&æL;{¡ ¨ÅZô 9¢³i³Yí—qfY¸NŠA»ïJ‰–ƒY•nWxO,©NÎ3•jó¿H»½Ôð¶Åô•ð?(*äcœ Î[åwè»éH,køUŸ¶Û|n-iq‘YÒæ‹Eäþx)ÕÒžµÜ<£ñüyUnµV#¿RœJ—‚ç[@/‡}K{ê^©<þÁ~ .6Ö&«ä¨AŽÕý†T”Ò¾q8±ÍzÑ¥²[]å™þÁ;±–†6 <¤<)ÙäO²¨anÖ?u¸„Üí°Êg–G³‹êÈþ^v«ÖÏ ÑÜ&sî@0O"GR¤‚s)D jÃ)’üÝÕ´÷áíÞ?5Ò­Ï«ÓmE5èß@“£z†uÀÿÁœjå‹÷â‘­ž˜pç^…¯à˜¹gkÅ/fŒï¡IªkۄΦ÷Ão~ùHDöz‚í#ÝúuÔ]ÒPKpyA û¹œÞCû+[™v’fýÓeo/Ë€on©ó"çï¦.|ÙØc˜ž¦Ÿ{£¨Š#e­…8«ü´²MÉ6˜‹ ™å}º|¿ÙˆáÉŒ¿…m¤ê¤[Ü”d¡ìsUµ}þ.Ÿ}Ï:]gCw…æ¿äYê‚Aa ½O©V#šˆÅúî1ÎÀèja‰d e(r™ê„ÜÍ}W:~Mƒ²‰~×Pé"G|WË%èó°ŠßXýµg,êžØåÌz0®Xu7¸/Üv°‰Ø—è~x#a>Ì­í8Hì¤Ì•Ârùi&ì‰wÞ%ñw%~çš3hÜ›¾Œµdë/×Ê ^uCÆè2ºú0Ušœ×hÉ|ÝøŽë+üù}J“ÐÐæDsXþCñxCÜçöPΉ§›ÔruCWçgKÝ£PU—ÆõÊ©é5ö±–6o8U‘Iîý&ˆ/,ª/?į&á·Û\Î*–É Âf#ÎGüjèÒ)Y>û°˜ŠÒõ,Fÿà™ìG²íŽ’‡ñ2öëœwl1„ƒèWËáÀv+4¾¬Ö(üÞ_µ5 n~8Š,úN¿ž÷è…Lt‡ÉqžGYÃÍ+j§¥Š#Á©úÄÈ¡­¼>õš^Ù—»Bï{S)à[ì#æ;ËP®£¼†Ÿ…ˆt/|U5$Í›1푯ÐáðŽrPÚ ék;šBgg§æ¨bt+ÅQFêZ\!öÅ™ÕÐX\ÑùnJþ“ƒJ?Ô"Fíx¿\+Nw–èSáÑJéʳ ›b'ù®“@à}Tóï¹Ø —m>«Rþ=’Œv .Þérà) gCé´1ã~1Íð€ØÎÆØXd$Õ¾fñU·û“<6ob%#ÃZ4šãZÊõvÇWðʃìíÙ{má¬ÿ™ –ÇÕX¢dâù„{9ª¥Z_ëVëÑ¿ýü,G93|2Õ—–E¼„æÏ(HÈ¥ê¯GPj–;ÄÚú¥Zð\žÚ—VIã®Ê½Í0ÈêõˆkæOÚ[Bä|%’¡*VùUéS7,¸ÙÍ©£·<–+æ6fôe¥è©æ†«£=õƯkÂÇd kkÎYþþÝ‚L,Má·%¶áafäqâ/0 e>$HBÇ«vç_¸¸;:u#~)ÔÒ”x,kÃeãEÊJ8å¼^¶ ó»vÞ9ÓvŠÇ­ª˜Ý÷·Cj×ОàV¨TPV1â¼Jœ^;¿*Á’2Iƒ´iígDF´qPŽãŽµB«Äß—¾Ý<•ÿNæÿnÙGéWMÉÀäJR7ö{*šyU!l|´‰)TŒ³$ã|Ty6w#€“’åêéR¼ºÇòi{û”ù^~âåLßÒÖMõU²xÔ){&ñm.–vµÏ\´ã÷®6hÚÊ]iLûãøA\™gO…†]Úî8L±K¬QµäeaÚýKŸ:ÜÝ“ÁjL˜¿Á‘iÇͲb+*‚ åº Ç¥‰Ö©ö,ª6_•j9ÈhÚcz«ëêåRs5ÞþTŸYL“Ãsߦ— há#Å=ŠÌÖò…Á )×u‡þî‡`â!òAaøÛÐMámëÇÿ¥ÒÄ’x’üRv™n3Q­4€>-«Õë@› ¯°ã]næ•Õ< }iTá:(“óvþ¶™,”glbËêÖÿ©>â12J®Éó÷aiäHææTLqô‹½³¹îÅõà~ÊŒ4gé%±6ÉPŽhtêÚOãT'˃Äççn—Ä"¬f V‡Þ…6š@ÏQ‰6Œ£=qAî`1~H€ÃT9—rÏAwÑmp6[<÷¦[ä£[%ñóx.ûëo¢ICÇBíÕý^ôú3*˜µ6+"[6l ºjNߨNöò›Ü- gW“PÚ¾¦ê5V¨ª;Ô°±|ºKúî ëxøå0° "/@ÛxU3+Òôš[ÑFPöJ¬áéÒÉ÷rcµ.T[E±%¶®Lë2Žyc ¡"—¢†ÚTŒtÏoÖ)ľ]»VÁã \£Ïá Ífå–¡ñ¦ã>ªî¤S“þÅS‚4ÐÆšêð²—–»æ[çš×õ°êÖ˜>¯xEc9ôy¿t"öÔ¸N|¢nj@Pr0Ïšù_'ž\±)/á†]2æ3àðÔ˜Ã'mÐ[1ô’¦Æ†"õ©Ó4^¨9¯ $ê‚d­Ð.ˆ–r Úl#EÝ/Æp_ZÁxʇÀ[qþâ˜/áNãxZbä€áuQ]®~Âsü]I ˆ,VûÖ3ó„ ,² ‹ Nf©z›wxyõ®}Žp¯ÈfqÂ?ä¢ÌÍ"áÌ#{ìïqmçèG¥"…‘ú™ðåï’’31gE툘{¼ Ma¨WZ¾5¥ö!Iá«.nèÁúìªÏŠ^ï¸õ‹lþ*É:k‡ rƒù {K'e gw¿ïÍêÎÇ+´”ÅUX,úïk×òI=AT l”´‘+.VgW¸i^ý˜á´â$Ò2S<­/ñŬ`cP1R‡µIýƧ*^V6pW’‹Çj¡ï– 1K‚¶wùÑ0>ó¶Å[Å­#áT;=Æ|Ç:95ŽCg5·%±%•Ð¾>4'¶I†b¯qÑRÓ #E_kW,±|ÞÆ}òˆÈµ{o¿4úŠ0q¾Ð‹üÝ•+Lâ3H½¥c”X  ËPqzÆÄXr—ÕW.Í5òîyTƒÓ”…ºt¯í„«ÜæôMPCQ'˜+}M1}(kj¡lþ&éqd)ö00ŽáÔËØVppV` „©bÅ#:ÌßGuGY¹nZÿF‘½[‡L (Hý„b§‡Ó"`rÕK¬˜Õ/íZH‚a ‡òÓbCUNc¨rP÷@AÁ“>¸7pì20¹ã-h•–N(Ä›X½¢ÉY¤Ð^À ,pï¶Õx;*W"¡õ}:|Lµ“WÊ(´÷V ¸6G_ÂI8.-ìÓ¤X*›YÅ/š¿É{‰ð{EÇë‹×=úñ'õl”ºÀ皩χ`‹&:¶ ÀloÑK¤àN:,\aL6`®á…šÛØä;•Oˆx=šýºñ,ÂSfŸN磲ešCqLl<¦¡©[–õk=Í«Ùã`ƒAäN‘!||“¬°³éœXm#ÁÇÈB­:ÖƱ?ÝþÁŠÀ‹úÅ»ûPG5u°ùƒXd=›ÌìÆz›Çm§‘$HRÜV׫J±©SÙä‡Âl³ëÂZPáÐfgJ»³è‰äß›c’žü£çƒýdÇK»^~Þ‰fcõ‘Æ3yacùÙû—¥/RÒ´ú˜È‚{?RE»¡ï¾ñýä*B4ƒ{föÙ‡ÍéIìy­Wgœn‘E!âÂDª4Í’:Ô0+’V£,ûÓ2À‹‚?ªÓä0´F<ú5ñ΃à5®ÈªÙ* RÌZ:")ø5ô@vb-nb,»¥†j¼3úû“Î^h™ËÜ–}ÒÔS¨ñ–´÷j+k¯×š‡9ª£_¥rævn‹„ŽU†Þi×J+ýÃõ"šÜf­ÄKéw ÇHx´ÕXt¤&ahRGe•¹"—l1U8tʘl{«ð=d@"Ì=¨RÍTSfêxÖïÚ}è ë\×ì"cqÏJ«›«¢M±á¸e! «qœ ˆVç¢`æw(¨)­“Wp¬í®céqPÒ?²DK‰ßºÖ‹ôËø'8qø€žãï©°aéÕâbxqtíC_°BD¯ØÆ3ùd†âj8Â\6?*PS—\Q×݃|àrîËDTI•²×R £(f³…(pB)ô…Ó«)†¢mlÐgß<€2㙵áúfÈÉ:Y…°™ ĉÍæLÿJ87/þªët ‰¸»‡µ5f'?ÞÀçÑ/ A3ýÖ·•ŒõÉПß:Åz5ÖÛ•’.· ù?ˮъÓ%¿L±1èìÿt¤¡Cú‡U1<9 –­Zú#vÄ“z»wíñÊá§pórð´íÊ…«#¥2\í†,Hz©ÿ~š!Ì‹xã ß”uÕ/ïì"cϵ^¸½…ƒiË eÁbŠ‰£ê=úÈžÓ×øTAšîÒ+û¢»ªlF˜D¢”®Eþñg 5ÇV¶“Ôëû®¹—Øœ£üÏëLn¬Ù • ©É¶#¯³Où}÷ù¡ vZ¡—h1l©]_Ÿï®?A߉άÕêþGTñìP%uú0Ôs‹X¦!(H“¡kË¥°yvIži 8´Ë5ršë¡h£›–;g¶`¿~Iõü™f¼ÔFq=AáiIå† ØîRo‘ÞCy’ˆ–0¥÷™!EØ­çÛ\߲ڌdzu½Tt1›¼9©’Œ‰èzÎ<µ¾pÏ-R¼T<Â] ç-»®¯ÂaÐs”ÑozÌçõu¢ïÆãŽùÐ2×`CÂA'ð>eÝ 3FWª¨™¿ýíš³¤ø¦ÌòõȆM× P¼a®#l%Þ ;ø¢Rb4ªˆÔ´×ùsõ¯ÔÅ¡ë§ÇèÅÛ8O£2XV)ô@›!Ö>F›gjM¤?¨Ÿ]æ©JI/¤Q¢¿E “¬–¯€ñ~…é–uÀµkáaž¥À!Ä€û’箊Fï)W¨£¥nfƒÉ*ŸåŽy™P©±)ܘDã×zk`123¢ß&xÝ' ñlZ!¢Zf†Q.rMxìr’Z ༠•6-&P”¢œµ„ÄvgÕ®ÇÜwŠôª¬EÅÓÜo¾Šºµ-^f8@rø©;Ž:AYíÕ[ÃGž¤¯N´®*>êY ¶ \hã¶&Ä(!UkCƒÏ}Ì£INH{F› þ^*Fl6|bqFO`ÔÙœvu!õœ§º/ξP’×ø ;‘û üêHGÂôV¼ç~ëm¬f…Ù”âÕGúù&½èÐ${SO…Üð†¾÷2QI¸¶µGÌÚvRÆfvIäñA"[Ê&hfïŸO^nÉN¾Xf•âøt äׇsicôŠúùCþgÜnnûtÏ­ô­°”òó(í/Ø´ÅX`9E÷ƶ93€À¢Æ}rHt²ÓsRÌ`ô oÞò2ƒ`‹[\@/FT5LŸ7%9Šb;³§¼nûe£‰Öb§!@â¼j¬Ó&Š6ËS9Y¹‚㩘äZ²Sø{;Îå—nÊCÿ…98~'¦‡­)ò‰|û‡c£…âC#”€[ùäú¡¯¢ñš]Ì8 Ôäq >+ÚÄÉ¿@ˆÝ CÖ–zkf$a‘çéH’Y\SàßHiEÀIþ à[ DØUç¾»­MPÛZª`‹ªgÓ&N´Gñ\`_)Å¿Q)HÄLj Xé#JlÆ°e…Û‹š!×ãYŸ1 ‘#ñ7ï+ü¯U*ð‚‡]†ML†Ð¤•¢ ­r÷KÀà)x¼*ˆœz".6±^q—›’Ú›pnþJúcxs^Üb9‹"rŸIb” í‰q±}O@wÛ–Á rp}ˆefÔÛ}GàˆWåžI.Xâ{™Áu›Ÿ´“ŠôXÔ:p X§ù®V…F¦àeÔ¤[-Ùé©bˆÛ¶´/¾MÂñQÜM;ÒãK ´Òm<¥Ó$gÜ«2ŒJYL‚îùW†ÞA…còiàÂýnJ°X„‰ 0Ä o±)+9êV¬šyuìôI‘o ³éŽØU¤ö‰ZøèNÌ£mUMª½5¸Kái„K}n ݸÏúXYà‡Êép|£o1è 9ÿì¼V–²õ+œHâËÍ"G'†Þ G7•'d¶öÚ)Xsë£Çì1[„q'Æ&¨˜£$tR_« Õ~¿ ž€6¼=poò:ð¨Ù­–G“WêBvzŒ„PfeÎõ›*‹/jrÑù²iHé…6/I¬E&žÃ »€FŹ…# Á³ã Ô"‘h¸¿å“^đ僗ûZh¤Æ„°ÖèDƒ€ ø©É `‚Fj«KóK°JïÍ?BÍmœ”ëÚ%·ò—Ln¼²1Bú…áhó±¢S’¦5jšá°Þð¸ÁdFUq6'ÒÀk‚8zï‡s°#\£ZEÝ \Ägƒã)¥ëÞqÎYÙÓ¢,–&œX÷ýAŸ«qÄó¯æ7eÓ(½‡® °E[˜–)ž64öð Oßy{„ä­aQAlü¾TJ;ˆ¨)(DÉ<Ðüºô‚þ¢Œ—ø·#ïƒÚææüSªj3–f›mÚ9C§bm{6­Ò•"ŽÉÀ• !˜ÑÔc×¾êÄìµy3“0«ÔãBgA‡JÁC’õŸc&fèó:¬UÈ\ÿå é±z¤iRßùŒQîŠF0ˆP:—6»,óÛ›€RÑšuÕ»!Î ³#û0„G;Yš5—ùª_(—Ø5Ô]”Ú»M´í/üxæÎtQGwB€Ms¾y=S±UwAr…ÅðV¹˜ø›¶¸+×åv4%ò›62·Çú c»p¤j=sÿÃaǤö‘܈b8Ù+h¬w+NËå‹jîæ—·S’³m%Y«|¤š¾õ^Ž&RùËžäÀ —IŒÁLÔ:† ·d†1›]G¡f¯¨ÀMÛeøÅøna¡^U¬$ÞéœRFo#öwéZÁâ8y« Ê[5¼ªqû½ÐFeƒ´ ë¯0ňYÄÐîùu:4ÍâVðS(ÍHKEVz¦a‡ÎNîPH‚Ý¡ïN¤&ePQÙÕ>U«äíhˆØ„{„ÿêýŒq',&Õíܳ`kiêâ-žº®áE1%…³¿Ë4D¿çsˆÎè;¢QÒËþôÌÏTÑDc}>ð”¨cQè{säxçµ@„l¸kµì"ÞÞüÝR¿m4¬¡Dq) Œ}mU¦ †ÆíÕ¤·RÙÏØ+X”ÚzÊ`ú6²̲Z0Iz ÆÅWögY‘†sØ¡,Í/GœbíE¾ƒgF/g@`ßGÉȦ Äâã±7ΆڔF„µ½ºŸ<øós,]–﨟Nšz»¹`àÇ›kã.cî¹±Ãü.½¿ "rT4:ÁŸR¨1ÞÆézÙ”…a0:½ÇŒº³àn:yá“tÌš·O×H’ÜM‘£pÉ®h+dØ;ˆ|dMqú™]1Ä”ùžPïüö:ýúHg—R»~<“½’è.´)01 òšŠ[Ò;¸8qFú>ûà- Q·hK9ñ ýåïÀ!¡FÜ!Ôˆš!bgõ\Ñ[¼6-•ë°9®aÞoúWp­n A,wžoÜèë¹ —•„Ü(Û–“4=¤××S‘`GØSwP5v ´}ú*ñ¯Þ§µÛÄc¼· e°r©Ö­¯ÕŠÄ¿ NfpýE‰ô¿Óvxù‚=|ÂÆ‹¨ãâæ˜÷вc¦—æ¼Òça ÅÏY3î)`\žYà0!äeí®ÍL ¯˜yÆu-”YPp¡fpÀ/}‹£;Ú^–%'@hÕ‚h³1½Uùýi¼fœ»×l¼FxM1ŠÀùPg=å‹ú{#Á)X¹{H“Žà* ‚ ƒp"<öæ D×j •Šœç5¦»¤9™)/Þ‹( è6Ã÷iãÃ…¡iÙøýÞ'`ûÄrU ´çÎÞJÿŽÿî^ÌY´Å‹]„1¢[ß«ëþXnò[jÂlßGé‘iöˆD>­² ôa% endstream endobj 169 0 obj << /Type /ObjStm /N 100 /First 903 /Length 4426 /Filter /FlateDecode >> stream xÚí\[sS¹²~÷¯X3µ ké.šÚU$™„p‡p P<g<8vÆvØ°ýþº[¶—“8Ä;PuÎ!U¤W[jµ¤îÖ§–¼Œ¶ª+ñgé©+ã#ž¦²Ž>ÛÊ<}ª| •uÆÓWÉ["#SéûV7ó¢pÑ^½ŒîŽ¦ƒeõÎàãÇ–$Ó¿«œW'ƒÑÙáªþ>ÏšaóqF!£ŽÆ˜ÿt:PÇ“Þ—Fõúg³Fõ“þÙÉÇaóUÍãFôú“ñH}˜4A“^¿'©£º˜¦ª 5Õ}«><ö…ŸÎFǽÉÙÉ°w6Sããñ¨ù¬ú=Ò7=íõ›÷«†d·,ùp0úŒÙŽ'GÍ„#F¿W÷Ô¾Ú~GR¿§ëϪwÀ¯®wø¹[cÚœºXH>çn¶bw«?þ¨ÔA¥öÆ/Æ•Ú©~û88>›4]ÿ{…üA˜Ð9íè•À.ÛnM‹3ÖÝhÝU£íQìTï4oqÏ+uøæ-€Áu#à)zÛÍØaFgÃáûuˆñ.4P ›€‚W “Ù´³u—0mE˜–Æ» |Žq«] ¬¯çF¥Ÿ|ø‹LM£Ú?¡M"/ê°`û Ü<ÚÙÅ"l¾Îªsáu¨ÒE J7*H1PÅXéÿ!`¥|`åú°®+.Sr€2”Àv=V¦3º½ *â,cº™¢×]dµÈÉ2@“²g üŠW "ýÀAøе‰Ò@·æü·KÙYÌÝTç«‘σfZM‹öf=h¦Kp0ø„¾¿#LpLš‘^ÁwñšÂØ(ƒ½–°µ©K)ï5àx‰à5òï9Ô]‡›œèÝ7s¸€›Ùý7¸yLtW`¢û_ƒ‰t,kÍè*í¥*ïl‡G—ëma-Å~!¬¥…’ýÑŠpÔÝ€Õdµï&7[UWùãMV©/¬:pÿÚ«GëË¢\›_,ÊéâæÚ[ت0Ž:†oR×!³¸a”'DùrkÚ¿ÎG¹¶¿x”›ú²(7úW‹rí®å+Â&Ã=?BX{^]9A.#þb”ã8šn” sáI·’D¹1©å4a³A{NXòLjgÌ5…ëºK××ÎðázÂ8ÜwMaï]× ®' Í.\S³ ¡Kw˜×†½­[›z·Sì6”Ò »¹NVðÎÑ•T•AÉï6%¢6Ìu²~¡"³R y»¢í&Tô•^ÚÔE³[t4ûX㔃XÅŽ|‘nç»t3$‘l!r8ŽÂá–Zpë@®ª².š¤œÛ—¬ÇˆÄ~܃È1ÅÆÊ…­WÌ <•¶Í%u¢Ô9Óµyq_[fÅ­fiqSiÛ’:‹PÖÉŽk×Òd„cú¾óÎCÆà©kˆ…*;²YÛpGv-ÕŽE5 J: ô%‹ÆØ®zäKZÕu=Ww±DÆ"”.¡k4*mÏ“¢¶õ ¹#`½[zK‰""‚MHL³O¯L}$)ëxµ1ß.>aÍ3¯éH°ÊTdÚµ–AZx©½ -:™J/mê±åÎ{O!Õ‘Ážà.%M¨P€Î‚æÚ²<áš³‰!I®5ZúšËWDcoL„—§`Ð@LE†ÑX:×xæ#ak‚ì¶5„—ò¶Ý<윋jt-vNºƒ´¥µ”«py¨m\"¾*-[ò"ã¢%ÈñLÉ{„–pë–ÞÒ Ña‘xêªM}®—”e5²é.¥õNÓÖmËÓ+Ž²eXi’°–L¡Ò\ä…o—Iº¦Èˆ~¡E¦U+Ãkk» =B¥—6õ5¥!ÂÇL»nÄRDþ"CPM¼1”±eSrO[$m2” 8Î^ é¡ 9@,2E3k(|"TJ/¥ß,{>*âJQš¯šCx)o®L›UK"ÚæÅ‘m™¶ƒ}¨)3‰óZ‘Qx–qü¶Lž»¯UÃӡ݉0OþÙK Œ^&ò83¥Ý%âCL >šL8‘ÖŒø"É4çÄ©õ²$hÞ¥€‹ÑV8)Ë:ç9‘gºü,õËRùÜ.5ž²[Ú‘Bɵ4-™Z±/(ÊŽµT&¼‡}/ã¯Cµ¶iÑlÓÆ×R]ëkÚ}¯Ö¬4]-£c-ôŠFÖÐæ7o &XÐÙH¹ S²¾£·–x§$Þ–½Ø,joB‹N½ì¥M XI‰Ë ŒIS2‘$f-Çr)¡€,çYZ Gp’­•=Ò ¹%œ#"/[øÄf“`L|Ö Ž%3›1èÅ8„†é<¥ÏÏGÊÛVóŽ"¬ZNgÂ{YMŽR$Ñ åÅ.Rž-gBT’-i.m[ò"C«ƒ²LæÙíZ±šðLË òÕ¦ÔrNئ)Ò©1%Î¥,á¨Çz¦„^lã+‚dù&"5'¿Ãΰ±`ÊØjÍ Å~`¨"žÐ::ÖON ‘„¨3æä²MWƒÝ‰M]øšý­ÆsKïˆÐÕçäž­øb£vå€IY û9PÖ¨k͹mÍûY°hn† Ë ´·”Â:«KèQ «Ëž§ …¤…ù´+g^æÉ!8—F_óñ­è“¡HYúœ1$ÃÈU[…YˆÈ¶¹äÚR–+”Õ·ëW7¢çºæÂ(Dkû.%4µ‚M2-6`5ê /<^–¾f{Ë%Í|žÁx¾¢‘úRÆm‹vD†.—E¿sòõLZ| |鰜肃.Nci&ž°å,…5% AÍØaèø$`ߘ¥À·"4Îr>çùÄS ƒ†#çv5,b éG Oɳ„OŽa…$è.Ïò… ÅSâÛ=6_âS˽B—÷ŽMç méÎH„²FC‹ÁÑõ7¿@÷jt M|Û¢u‘£¤%ñ‚ÇXé’—^ÁåeÅ9¦£øvžŽ…Ž—Qqb¤ÜM$)ÉȼüíeË-]så"PÔ¦íÛëÛûîÛûîwß-}ÉMÞ†TÆ&4I°µ¨×œ/– ¬Ë‘”ÎœŽóHÞó ' YsÂIÒ¹d è#D¾P«ùB»æ›67¡spâ3´.<š¬=ð¾ä|é¹g»Õ|‰_sZÅ'U±mk.R.cn×’ZÇÖ$Ž6CúÉò®X‚Ê8Œä³·XHJ=m´9RX:3©û0OžÑü5À-ý¿A ÀSÍÐI”’ Jp4¯gÍÏÀ I¼ÔÞ„Š¡ÒK›Ê‰ ”p¬…lË™1:½ø¾7ñqP)ëy©hå¤ë'ÇR/©mÑm8õw\W/¤˜"ÿò‹B}&än›@x)oKjE±³|õÔâM¦‹©¶L1—{Îù¤$Ñ»jó¶-y‘™;ŒRháÛµ4!á˜Òa3»X;6kÍWæ|Dž¿ðÆðeSœË1Íœi¶KèÊ~iD çè¢ðR.Ûµ¤×éXzΤتi.Ë~Ñš¡ZÚ,õp½ t¿b#,óïϽ ³ÓLû“Áél<‘¯îåõœ½7»{oŸÿcûÑŸ‡ºFÅ°w<­œHlñkwŒ«î˜Ì?¬ò´³zûxJïÀá;j»wz¯šUø@PŤֵ?ë ý»£ãaC•³æäR€Ž:,-pÀ„‚O½ ½ð›šôŽýÞpkp•9>“--²k›$$Û2Š¦Ê¶Qîh77K4kÍò§ÚSÔcõD=UÏÔsu ^¨WêP}Pý1¦¬ŽÕ@ý¥>«‘«SuÚLã#5UÓaoúIÍÔ×U[øMlqÿ퓧û»l ©)l,–°.ÿ Käz­%þZJÜd*ÏŸÞßã©Ä˽ªËTè·³?f*I¯ÊÜù.ü´pÞô¢¯ò&¼ûøÉÃÛ˜àóuaëêyØÖ7rÖb-‡|EÐ>åP=l¿P×ü}ÖÎ_¨uJ¯{ÑûxÂMH“:žMWìô&vxt÷ðÞƒC²Ã?›yÌjc~ˆâz3ÈŒiª³W'µ(<9ØÞz…I¼Y‹Ó¹ºÃ0­5·ürZ˜ã¿nýRäW*ÕÉÙp68~[ÙFóöíÝû/yfëÖ¥ÇÄø§Àžn¥Ì˜Ø6'ô~§Lðt28iVg·ê¼zqwïáÓìnÒÖϑ묦ÏO³Ííˆ\a®¯‘¾í8çÚó£_=ÄÂì©>ö!â‘jÔÇ¶Â¢Ï ôüV²º†`éÙö³ý?÷Åf½l±–`•ÿ™¸«¶Ô6[á1ï¨ê%¬Ý´XâKC¶øvú©­„k d‰¥þV†q¬uu¦¾¨Õu7‚±½ûÏ<»9·În:Ííæ ª?ÙnÛ¼m¬¿âøa{‰Å>ŽÏ&%„VíU,òiÒ4 »üëÜF7ÂÂÃ7‡÷ÞÞƒež?ªÃÚˆÒØì¬áÿ #Ómóz”'øÞÔ2¶n[†ñ}>ÃÕ™m„…;¼¾/3‹kgVÛ2³`è?ëháaHyæÌ6ÂÁí½íÝ×2³ôý™yO_“¥Ÿ93Šæ§HÁ–+5z¿³Òé‡G†\=ÿ ðòUsmš/wŸíßßgs­ß6èÇw,ÿ¬Üúfõçš‹@syÝ.@€6‘ûê ôüÁä% áµzÃfý0éõ?73¶QáÅDóGÎ/d{îS5£#:³Rèß@ñ?—á_Žé7(ÍžZô_ ˆ ­ßÃiþñƯ¿ãÇis2¡N›/èm:øº8W]€4õMý[ý»™¬®“´ê?|þäõËgâx³Þñ¾¬ò“@/.Ä®&G¿.ëféÁË-àÂÞYsõ’jolË –\ìÿ•\]c endstream endobj 285 0 obj << /Length1 721 /Length2 8540 /Length3 0 /Length 9136 /Filter /FlateDecode >> stream xÚmtePܲm&@Á‚ËÁ݃»ÜugÜ]ƒ;A‚;Á݃»»K€`Á]oÎwî9·^Õ«ý§{õêÚ«×îÚTäRvP'5w„‰™•¨*¥¢ÀÆÃþÈÊÌBE%î1v²´ƒJ;Aø€š0P²±ÙYYÙP¨€âv0wKs ' ­)Ý? PÃÆlikéÔ°³±s±4µ ¸¸¸ˆ¸8:3;8 1ÿmR…@€N ™¥ (®¤¬ýEQH+­¨”†@!Æ6@egKS ¼¥)ê¡šÙ9mþMí `ËirdþGÔâàôW˜™ƒ-PARMTJIQ ()΢&4†‚ò2Û¡NŽ|ÙSÛMÇtúodcüßÈä¿Ñ«nÿŽ˜QØØ€`KS'  ÄÜŠÂò/ß¾@Íì€Üÿ†Áΰÿ”þ rü«Hû×C: bö—ílc£hl ÒŠÛÙÂœ @;0Ä T±³5†þ‡elkiãþÿåý/ã‹“ñ_D¡æcý7dé(eé+[:ýµÚÌØæ¯Eÿàš¿‹lélûï~È_?þ¾(Øjãþ¿Wþ㟠Y¤eÄ5´•þ»ÿ”%¡¦v`K¨9PÕ鯗Æàÿÿ”•-ÿ½=ÿóŸm²ý_®`ìä`éÔeeþ×z°þëü'Òÿ?–˜˜›''‰ƒí3ƒ“ÈËÉëýÿÈ5uvp€@þ1üï,ÿÉÿY$Ä bŠ²4ogÊd•ÒRâ#™7YŠÀ09n¤Î0ÏXë8ÕÅ´ìµîü!¼P#ÌvE%™ð7Ü<ÞørQOaAÝ$ÓŸ'¼+ ,5÷©X{åÜW‡œÀkØš*Ÿ~=wJ]^ P¦åLÏ<¬µ/Ä”f‹4*néBÀX]œ¹‚JÓéÓ`¥é¼ B¶Ê1&-‚sOL„ﶖކfn¢žRZä–øÃvBuòW-³¸åÌ6ö„w0¬2ÚÙ³„œ¦÷IJ¼wDã¿Z-zº:…O) >A\žùòt<¥NÂ5®­NqLyÛæ#Øðî, ãÌR˜6ìß÷[NàãG´üàûÆß] Ö5æÑÑ\ôßÅú^¼¯æÖÇ®z™¾õÝ‘ÓâàZšó‡Íg†áÒàÑle’ÙváÉêçãÄ‘Ü)‹ ‹\ °QÇoμ•|æL#>±¼ï=D©ÊjÃ<Š‘êl‘g£¡®Ý—P?ÉØÊÙ‰¾¤µ¶³;¤R# ­£¦À5`êóD¯ÛKPÜ›aòPHÈhcðàÀ}™º×. ‚J”m«9s…ŽÒl9~þæßw·8!!?x Y§T±£’™´³.&«Ä "ÅfÖ•²LŽ°äTõû9ÈßU¶[í»~¢'Dxëg(:*-œïäVUṳ̀ÑOzÊzÌþf’ú¢{¤%>õ‘YÉðõ“Åô§7˜#ü9?ÒÅœ¤ñƒÞô(ŽÎFIѼ´€ÂoÐeR’ò7kå\’sÉ%!a}›‚JeMIcúˆÎ¶ÚB ‘ÖŸKúv€<¿’’Ž^~¬— ®&§B5‰F6Êæøá3Ošb¥ÜO¥žêp›>k ·m]ö8àè OJ >H:숵¨os|O¡ò•n«à-¦ÀM<í÷O‡€ºôf[7{»ÒvqK ½Ç¶ø«øo¢åUßÞ–`ä³Ù©[Ôç’j"šZÅꇜ+h¬M;=L ,h>³„æ¡°L JŒˆÄ}7݈Äõ[8¦t+®õ*;n|íQZÔ0Æv‰¬z½äT±°©Ý• Ìdƒ3òCpXÕdž?sú Í•öI>àÌǾ„¾ZàÈmtÌaôbðá1¶2Â÷ 6G%BÂbë˜âÔp”ü‰ŒíÇ»y©‘ä1â©Ú‡|Î 52ìlïâ6=Bj´’7…5*3Óy÷úgp‚T²KÇ®ÐoB¬K/g_"\óz¾›* Õ¼=G±¾X"®Ô3a>Ù|«?HNðŽvN÷',òÇqÄ’Ja“ØSì'‰u³°]=mƒ:ÜÅÔ¼jøôf82•¸‚Þ"ð1 5In­ñõÛ¹Vƒo¶¬]è¡óŽÏ—¶QÂ¥$~k€TóHwšÆãüÐäu úµb!%Àè(þ!ê—s/‚3TO›¨cªG‚Ê´_»1QôgÁ#uN3¦wB »ë>LôòwðÜÊu>ƒ¡þy½ì²”Rù¡vË™ ؤÿÐëýOÏM’æh…ìÑÈ—ÊZvžqÙž9½/qïgÍ4(øÔ‡¶•¸rëgVòÝŒ\؆ËÒæúSõ ìÇpe*‹›¼8³s1—˼§ß:Y0âÞøHuãÑj‰|—`bl¬º›d¡ŒMÊd(/Î<é¡l¯ž\šã=ž˜¢íxV¹^ñVšz- ⶰM®Þ,híLyÆkhÏj£J‹Žx/…3+§‹q_¶Ëëš¼K5—6ûeëök‰Ç ¶DÎ32Dç\È¿î·‰ßÐÇb*Zü´ÓXG…ßSºF0\ì#Á@ î<Íš#ý¦864;hèÚ¡ð³¥·“\öPAéÐ\xi@(¢]Õs<ð'ih„}´5ñ‡‡ÜÉãœrb³ ò>4¡Z}Z|`^Ðx–ØÄ-[Éf7äá53”»Â:Ž"ó©öhsŠšV–¯ö¾Fð³Èfök""œ«Û§p Ý=«†± Kif®)˦Ûz|/€0›ïmÁ´5¤ýª}`›áà‡Vê\NLx4edœzGÿP¶ Þs. j„ìUŠˆ¶¥ŒìXIez¢Ñ½Ò‡’$Ó'Þ>xf¼Sà~Ž£Œ÷Ä(¬¥`08õ—•¼y¹KËù‰îÞ6UŸú·NN>Â5Ìq ËV‰áœV[¬  ³.RxÔÁž¥™ê˜PýIêé žâ DžÃå;ä›N¹ÃƒûÚD¼ð¥TÈõ{(yV/NŽ‹T¹ä½F¹[¯nè‹ÎÒñ a·Ê#+Húþ½×¹·ÿxñ·~³Åø¯B+``ÇCíÐ yæܤ ¼">„v¹ZRëÔŒNð©{ÚFFî¹_v2ŽZ(Õ–¦y_™×ùÅ<üÚxóˆõa×ú$/ª^ùøÓ¯¤ p,Sæü¶ìšp8ÂÞù'Álæõ¶á/ùmòè«6‹Vº¨SãètíéÓ⵪W¿Ë•ó…Òu0»_ýµ(Þ%’l9Xösg ŒŠè«ð/ é¹† ¼³&‹p;¢*¹ÆMœ$1%f0Õ·û*OøýKQ`ËÁÇÍ Y”s?³×ÓJ³n é/šB}{?jèÚëž‚ƒ•-…åaß9ç’…Š©9$ÚÀ§˜ï àŠ(ûýS ÙÎu뤶Wb9ª~º Êk0L-zÆErm™˜yÉàÁ—Mæ_·°MñyDÑ¥“™…Å é~¬’èqo$Þ7 Ù¿dã&Œ•_[;÷YZ‡§{f‘ÑÊF@É@ µÁÞ~{í£o¼‹ŸyºŠ;‘`ð‹Éý²mo #MÎ7‰Í9Žþ ýçboz6ä¦H¶`÷Èn;ç®ûø‘uJ|ÀWòë –¾Ö7§ò| É.×_KH ¾gVem´:\I^Êtå1ÞJL‹†ÄÎÂëÁ‚½g1œÛo–JŸÚµ nØyb‰ôytYÜI>6ß`¹ËeÚ§†eMW¬Td†X¿å­ oߟ.œæÝÄ"`(ëê:t~Œv`E¡ý"†ƒ_ƒÌ\ihฉÏ^mëj¸¬X„.Hž4ÜÓ*C¾Ý‘©¤MY›˜¶s8ÍžNÐ ß¾ñ}þMš RÔÀõÀ`­=Íî%ì¢6¤D(˜Íšf^šÎòVa¸5ÒÚ®Õ²‰¬î ¼’¸f\®,Óÿ}.Hm«Ë¬¯›ê¹5ø£ú!´9Ù!›|Û„ô[úÇÄþWM¿âŒLž~ëFrû[qMaªQXÊÐPßÈŽÇïÛ;.dÐQ[ÿK\gXÞ`Œ2qrÕâ’H#[ì[>e®‹roý ñœG7ZÃõ›žð1~‚€õj§Úºíù<iF¢ÖWS+Aî9-ÉPAµù)è5yo(o„¤÷Œe´fºmô>$\•S.åƒGþÖ—WÄgšš]›â–·Y¾ÒèÞÂ,K½Nèôç Í1´.©í°AKŸ®C2+(ðƒvŠÇâœ=G‡ƒ(È> ©tß׶¶ã)b#±Gç5Õcx%5&íè—÷î» ü+Â7>`M׈›ŒeoßNGž8Ýî*¢Æ‚· Ìö=;¿Í£Ùõv‡rÒÞs7É1 4¢¢O§Ë-¼ ˆHã)©žÃ}êüyùÅ*Š¿¯‡_¡Ò÷b*5ÚP€Î.v|Á0Ny·ôÝûâhCb½’ÜÀN26LʘhçÕNš¶°x矦JOî¸Ìýô’OV¹ücv%!تô•ï8ùï[ä–ÓÞ`LÄLIè_ƧÓи&Ý&¤©D×ìRo³I¯øø,ËÈéØæSnÅk.Nå÷“¤«Gd»«ó5-@·`È3ê8_w.QkÖ óŸÏ”VÁÑ,E·ƒÝ(”ì}ŸZ¡^Æ€J]i ï_å9(U9n –ª@*4q¨@ª6Y9u3`:0¯iÇø3)-bä2P‰:XéÝë_xë÷!n UHæP=)nžsQJˆRÿ!½yræuè“9³h •9MÁk& ØÓ&Æ5:úyœ÷sœçÑÅúüyl\4 Q¹Xµ@ÄÃYdÈ\ÓyÍün:s;f²Y¹¡f~bœ°5öL+víóÔ´hɃ6ìäÕwmQÁ¡Áù~OA4( ˜í•ëƒ¿™d‹ý¡…ðN08‚>ó°iö㛽èúYbeÆ äHû‘á0¦)fIÑO?£ÃÛ5­¡oSzÇÌ„.Å œ"\‡ðýÝyƒM‘ÿ¾xß… ´ºƒÉT4Q È’l&|e¹`=#žÄÆæ-0§œ>x -eí=“élÈû ~J¡?k+︨á1jyÚFd_¼Àò<èŒÿ™”0<(¶#®N•Ø•™ŠÄ"±W7p'+±„l6Ê&ôÖ¨âT€~5š¸•;±EšZ³Lg8ÕtCd~ šé#5¤xƒ¢\Cûö§¥¨ÔDbNÛÂN×eãÌvá4‹&!ÉG£,Èh—L³s5¨Ä–¾Î}‘oç·wð{Œ¯ 0œÕ^~ÜÎœ i~zæíŽ+´âÃÜ<)/k’î…®åÕŽ Ú+cQqtKp-®´ìŨWŸŠ~l7aÞKÉ-Ì0)öa›ÞW¡l'jêÕw‡Ûg}+ßÙVß'®Ù/DX2*í.ò~%˜+ë/-Ñ&&ûPú+…gÙg-ˆ®hßØG/ R†ŸøÈNjûfß´Æûê¯ÕîÈCr Öî@}rÓ…ò¥‰‚…|bP}P¨/C‹kc=BM³¡¦îc`ŒÈ€&eð›uÐÜ®çÙo˜•Ðˆ&—(æ¶ÝM¥„*°*ÇD,H U‰¶ŸJºøÁ…¯(IÞáô©£ÏR:¸,ÎZB­˜ q Â$1Ý“»¾²Ô”¤7ÇÛ6:w+K?“í÷­ì^CÕÉ´·þõ…â‹s»—Ãůxþ=yJGNtKwßÑu‹_f0ªæbEKo ÇÅu Ðm ˇkKŠO?Ž ·óÏ´7ïó2ž%Ÿ|›$·’¿üXNÂüÆ$?q°0­Å¹ÛpBÍGý>&`!¾Pn¼¤f}ã[BVŽÅ³ÉÁ‹Ë…º7åbÒS³Éh_Ë”wp×”²•i1>VÑì”Øäþ¢‡^J¹BÂtªïGP-ÿO×ß]F5;¾¤2x7lU&1ÎqÒ‡Ó‘”ÅÇሧ÷püûªT˜<î^ȇ‹žgó¨Ò—=…†©¸ÅúÞ¯ ­e!‘”‡®ú8᱄/¸†=0G"´o:--WÁ¶Ø=n Õ%ŸÎT¨ é6Rôñ/Ø„‰‰(øo‡DÖ~ãRúx‹Za‡Œn÷l`p60v¬l¼‡ÀÊ>H}\zyè#µoß©EŸ¢`|=AïHÁÿ$X}q5ÖdñrãÄÏѽ}f˜ÜÜ(Î4ˆæâÏÓY€ èqL÷–"ÝöœdwB cUw!4ÉíÞ.“Ù[M|#Üó3è’gjથë1†1²c- {lþ£0jl„=u¬èÂàkÄö1¤¥m bZn²TIïb¦1nßI$oÀ[ÿ…F’¢£É/øÆ%àsŽà¸L÷ç‘oª¯¹$82@gÜØ0µ›H QÛ­O‘«V•ÔÏ9œï’¾ÿ[Š™ÑV\dãY 0¾¿*FQÔ0Ÿ`;£ÿidIÛÜÖɨNÄ»›êK6íEœÃ ¯€¶ÛÔçžèþ³Þ[žÒ+CaþyôãÂüBB¼=¢ÇâÚ¼VWäƒ] ‘õ§hd‹–aJS}{üe˜ßñDÏL3X‘Pƒ}âså‡ÙDÑÆ9îYMQw*ÊHe,š.D½—œJÔlïÿ€K5ûv@ÆÜ@F«¡<›º¢ö²)*òXêñ슆өOœf3uHmëe€Œû\þ˜uÐÊ/Ä›  E-GiIM½î ­Åh¢YB›äyÖàÐÒZë\;…ëõ€Xœ ÌT>Þ6µ­vžÏ0N¶Ð¸+«ÅeÊ02¤ 4]®\ä ‡rg’U½˜'N&·¯Q„HÆÌ «UÏÆ sžÃv¯1^ËnØa—¹d¸ÎÄ-ä£k4ãïqi#Юo ;•byÖÍÑ©9‚su@œ®ôÞE­–¢1"ÜY/ñÍÔXóÖbÿòó:Öyä3ÈÉÙŠÃÊÆf?•\66Ø¢a§ WØr„‰‡¢±º’ SŽZ®/²Y ˘à†-9mnÏÆ®¸gä•£™€;@QÃm—ú••Ó¿žî×^¨¼xײåçÇ»Èv o…&÷.%›z×ý=\ª)mh ÿë‘kK.#Š:XÍ# a®Å§%'…‘ø UrH)º‰µ]óŒ<‹þ×o*‡Óõ&}¶ޅƧÐù=ƒ¯báuiÐoÇ‹Ëw­ÊÍ5Ä Í#ÞDžK- öœÌ ~/° íþIäÇ×h/à äIq oßýÛ{×ÑÒüc.·¨®Õ®­)%• —@0]ˆ³»e Ô;$šÈû!ØLrÞó5GÄÑNS;K܇¸‡»'»a|/]’÷JˆeÊçVÂlêhÒD: Ê–éè°cf¸›7]· Ž++?¬QGÆY½ð²Na]îº2“)ç‚ÊWŠõ{“K‘e {™´JÉñçnFÕçÒûçEŽ-G[É×14¹Ÿˆú[±¶‰x˜M´ÕpÙƸ®Š¡pŒÎ6s~–9 \ Ù ò!è¨CÚ*Ö)‰Ô3zЃtj«¾öúaœ> d­­}BˆžOõ¯Æ f½à˜?Áè Û~‹5_ôU¹™ÅÙ›!kÌà郀x-^ù\éU„dÂßƨþöÀQäÛ Ð&S¦<`¥rÏP×zÜ¥!KÿE8^„U\H;ën j~ë'²972Äæ=¿K˜}}T.³ì[ÆLsŒåÝÑ¿ø›“ÀÑFì 9édp/á*ÔÅô²·z=$" á²F©Ü‰Š‘Ç+#ªÂ’gÙf"\ýmÊY>L6hÐfpi¾Æ‚Ü-°X‚ 4McÆž‘™>nÏϼÇ|¦Þ`1úóÌÔ t’ÄÜV2±{øÖ•ÀïRjV˜#ŒgxÕî–õgK‡`éXìÓ½®ŸvµÙà‡Ò§¾}¾©¦[aÞ_2,¢nŸqh®+fÜä›/}¿íŸLé4/?¦õfÅùnxUßSkeqŸëKñöu•6è`üºÞ.÷åú”Óÿâúw[–¾ìÒ@2?ª£Ìö‹m?ög±:½aA­‘Ô%–1Õ§7üã gÌ x=V~¶o ²]Ib™þ,wQSCY5ݸáõ‚¸käß<Äi±­Ü«bø2O˜÷®}+ØEfÁû^ŠN(FœÛ¼'×ÏuxÚX–/Òš/sQÓ ã±®$—ùËð…Be÷¸;/tµLVd’€þ?(r;ët\NŠnŸ#ñ*ô¨©uÅÎaöhÇEÂu”¹•[R,gKž—¡Vqm…¡%🉲³ˆÁò”Ùa8¹9nPÂí+w¥»™®õÉIæ¼ýaÄñ ipÏ:¦‘«o¬àBÌ­Ï+˜z‘ú÷±£ +7^Ûä³q)#¹µaïú+ú³´Þ®Ì<Ú‰]¶‚ûÏx϶0t0ñ÷D„ê+˾å$[åkäG˜xy:žŒCä ȃL½rü#ä²(ô½‚¦rþ÷³Œ8í×¾¼SWáÚ¼iÇ·~Rš ÐRÜ»›¤Ï^ ÒÚË+S(}+.FççúÍ0oo4rÚ~ƒÊ ï™…QæçŠÁ~'ó(ÔMp‘{¾ðèm}t Ýtñ¾}¿`Œgw &xùÀx¹´á…ž±‘‰Ñ7ÇDóÈŒV •š5^¾¼gzgVV…}Ô—×…ÆJ!ØêÞHåÔ!¿×Fö i6™Ú.7Smgó3Åõâ[Â#óªxàePü݈ÈÙ"à‰lígXeÄ5Õ©…”W{1§ùŸÎ O“ä¾\»P ïû¬ š²>R¸Ó¹ïÉÅd¶^¢°0_ð•dQ­1¯qÿØH‰$y"ç®5¶™ScJ•`ü,v»h),JXPÚ^ðvÓ~ÉÒ×Ͻý#F¾!>SíQ©O±S²Éå¹ =&Î6yìnqäŠÜàÔ´÷B¯hûªF[®ïoË·ÝÒŽsŠííòÅÆ’^‡écº#sjVëÒtÛ ÆjéHDò…]=;ZÓz°ìsáÇƬ7Bn Üå­Öøx¸?…,+V°èÖÌÂr¨C3O. `ÙÇ¥Åäð¹zU˸R h"– GÞJmrNTæz% ÷$Ó>"]:=Üü*-G^¿_ã7@ŸÇi$Ö±¹g›rJNÑ…¹<ýš@6‹„œ‰÷èúÎÈhvû"‡ØÑPpéÂÏø8"Hpz34ßtuSø€,Ò@âa[Gùʦû:$Dàö+¿¸¹Z¥¯˜;ªã¼Ïnº²H«Øý$êXYñHaŽ'׃ŒªºŠ,Ç".Ez-^÷ÞC”æ>§è8^¯§§J†Yz}'’¿@qu}F;+ PÕ²(lT$î¾|;çk±^¢–¼«»cä1wÏÀ/§áÕĬòöu´Æh‰Ø½v{ùÆ¢r›¥†îHœ™Æ ÷¼~1×…y%¼M|,ªEPÜP‘Œ¶ ³íœ kçﲈý-Rt„!iôƹCûìå;x´/LØ^áO/tþø.åä X0<êþšðeØC‹ÑœñÄŒ-U´Q§^E¹ž]Âd O2z#~® :+» ±ÊÀÿè·PåŠ)Ôw]*Ò¨nëÙ‘þDÓ‹ˆÃ¹Í”¨¾©®©^ËŽ²(7ˆNçΪ=øa/ñ%æ¸ñ cÌŒÃ` oÙ;‹r„_áœÉ.± #Fuà Ì¨{€XQqo1ÕqžÏ¯U-îyš¯*ò`gÁÔS‹N&ÿáò.—´~µwf¾¥²J¯c%ùY~ø~•ž}ʃ㸂;)2ûöHÕ1B§Ë6—I_B„6»»{4ν5^͈5œz˜äö èg¾l\qmïÕngc•¹¼†øÃ\ðW(Év Ö‚]Ÿx£\¤;~ä*Ø^á—LÑÝ€º@&¤ m‘3êéš'éGD#æûÞÑîɵéqšJàâøó»õ…%v´O©ÒœDuZê ]âÉë¦ Al¬RK.v!'®+Þd}u4ìrwÔü’1Uj^ÿk¥î£®·ë\ç¾ìÞàÆÖǸëߦX)ëðÎZHïýþ6bô›³–F("EO¾w5ÂùÎýL䥩RèþM£{åþóŠ7|kù¶Š‹Ônú¤#†²–tUã• úÚoB†‘xh¸qîPr°¾dóÅò¼×æÚíø~º†èÂßÕ'™ò¹¡ èôF¤8Èp*Œc©®+b¤EQu‡i>LSÂü{Gé5®veÑLÆÂà[-Š{;d'ß0NH034J„…Ùì hqî=±p ÂìºnazÛÚ÷ås.ªCltþ^O)$ךe£Ua…‘.CöA€]Ÿûv%VãËʧ7wfN¡š´Øb9âpªÛó×D_m¼«ÏÄÓÊú%•Žyï˜è¼â4bÕ—à³ÅV.5•Zâ¤Üˆs¤,¼nügóðÇ*Vo¯³YD–É>“ô4á¾€àÈlš˜U`¼qÛ×’¾šï‹(«Idƒ #³è‰0i B“Ãàã%fñ/=ž^ßއzLÿìÖ{>Èà1ºvZ}œÙ(~€óðÕU— Íb<&;XZ¿ŠVPšžêÕ‹Ù%qÊ^x_³`ÎÈw½`ÈN'5äµ~PWš$£P½bßüpë‚“ã\„ñ,ôå²û m(×ãç±QÂDUªýo†?"(~Ý©ã~1¨Ë3Àxœ¾GHÂá Ú¼ÿÞ“|5>‰‹ÓÝ9/WVŸˆô¬k{mÏ`­`¾¸RS! f£™ºw œþ…í3ßËðAº·Ìêþ]þ-×ꑤçfÔýóñ‚÷ ?Òðåá?) ìŒû¬©ë‚öµ¸ÀØÎÌŒÚx›à†„š÷þàuËeÙÊ=ÙŠ*„'Æ×~6?¯ºÙíÎ*ÝêÒɽLB÷é.?Á1ìWrÝáÚU}ÐLïžkß›®žµî¦CÚ‹¥Gjœ%‹õ eQ ´vÖ¿̾Çëí´²|V“Kü;a|š tÙî]µ\º†i×2ñл‚w\ZãžÐÕF_‘í5Ñ&‡SÍã-ùb{z¿ •ûŒº/ëÐgnM'³ùÆDЫ¸ÊwBý(Ë‹9/–6„"´_dFV«»v­a1˜@2J™E·‡ÍˆŠ =ÔTËeÄ6,¤û‹:m¬itôa×¢íHê*:¹uÕxªëoûŒÛ]SB1d‘w"• úâ>U Rˆ8 á"†%™}¡¡%N(ÃŒìal(ÊI\W¿4.mr›Ú¼ô¿²¬³XgÌЪ–ÝDâhÔõ!š ¡à- ÛßBü9©²î€Ð-ÍB qÛ uÍ5„â_çI¬¬¹ü¶Úðõý6ƒ8™à€÷°p“‘A°· Þl Ï)kxÓ[jO…Æ~n‰¬2}œiÿ~~xügBÄÐt]X*gƒ¯h{ßõÎÚÊ£¦Ÿe¹¦Ð×b8KÎ_ÖVΨyöØ€zzõïÚÕç9çøH³mü†«Éê¾³[ÎÄУП¥»mëBq\ù«*nrºià¤+û,¦Võ€¨á-Oþ\ Tuxcþ©¼ƒ­–Šóµ”î çYÔõ´HõfŽé·^Øiåß?éë¦^>½ñLZ‹õC‘Çäóÿ|ÍÅÆ endstream endobj 288 0 obj << /Length1 727 /Length2 19006 /Length3 0 /Length 19619 /Filter /FlateDecode >> stream xÚlzcp&ÌÖm̉í'¶mÛÖ$=±óęضm[“LŒ‰mÛÖĹï9ç~ç«[u«ÿl¬î^{í]Õš’TÊÁ¤áédedebá¨K©+°²°°X˜Ø()ÅÆ +{ c  4¨¬l6VJ€¸ƒ£§³•…%@cJûï @ËÖØÌÊÎÊ å`ëàfej pssqsqervbúg“:YæV¶@€¸²Š®¬’4€FZI ´:ÛT\Ml­L V¦@{ -ÀÜÁ`û`ê`ofõ/N.Lÿ&`ïtýCÌÜÙÁ (©!*¥¬¤gÖÛ›dþÙnráû 4u±ýWu Ð-[ãÿZ&ÿµþ›uöøÅ„ÀÊ 0³2L€VöÌÿÒMÖÞÜÀõŸ°™«ãÿ¤þ!äò?Í?ÒÌ€æÿ ]mm•Œí€q;GWРè`t¶¨ÛÛÿCÿpÆvV¶žÿ_äÿEÈ‚ŒÿQBÔÞâíy™8¸þ¶r‘²òš©XþÜÜØö¡þ×þ§;Š@3+W»ÿœüG•újæ`oëù¯ý§˜_ʬ-#-&§Hÿß1øwZÒÞÔÁÌÊÞ úGQcg³ÿþV1¶úÏ Xþ÷¸û¬ÿë+ƒœ­<ú,Lÿ–­ÿ± þ%&æàáÍÈÊÉ `dgc°rðpxÙ¹þ?tM]ö ËþO-ÿãÿ{œ€@ )Âê’ƒ)°uZkh…¯dÑL%4ýÌ”‘&ý+øòd÷µ>šÕ ¥ÙƒKü; ü|÷oDÓ&‚;Û²w-dÆqÀoGãѤáKøÏ%䕾µ›¯ª/ á†WPZxv¦JP×· ŒµÕ@º Ž¬Ü³Z9§2A4iÖh£òŽ>hÔ­ø•ù{iZjôL½ÏoÀ=úÔËÂ+áçÝUˆ°ü o×”x?, +Âtï,>Ùµ·Mì{œ¡ºÒ†åZÏŠ5~ó5Wžš4½ 4aHÅüÈ »V¨‡øa¿í%S…Ô ï/¹3ξj¡)8ß0gÃ@Y¿¦ÁÄ€„W ¾ÆÆÌNçÆ} âJð¬XfÌÖź±÷Ý/ô®Q›vÐÈÝa™ § _võ:‘Øî‰Äsõ› `ÄÂa)çéO5ÁRÐ «ª•‹0ÞØ—ÅÁ.°Õµ1šV¦qRÝ Æ°“³|Îø7 ýí¤UúîVÛ¬g­"d;CJAƒ•-VŠuiV7tsÍm냂–QZ‘.OÂdàò¢#Š7èÞØ$%P Š"ÔÄ• ó‰~}˜©oÇ8?Auøñð‚kœñ¢<ìø'6^úGC|<Ò€˜ª¶»ÞžÇ1Á¡3—°á½« óè„üBZ¸Ð™B$Šb6¢TLŒ-%ˆï>m>ˆv÷~“×Ó£1•L¹x]ßk(µX§™âÙ\°~ÃqŽœySÞW©͚xF«¥Š½òŸ*„Y` q<‹B‚¹ ¹f`‹ åH>–a·KÜê¬4ÿ`àU¬ê!ï稅äBÂ25£¡hHæ,ê>¿E³$ â|X²4  aŠ¦8~S"ÓÂ~×üáU~ „§+ýç&ùÕ{:Mâ*ƒç·a“Îòšw®X¤5FÚp>>|ú :e–#ðîiz_Œpç*ó\žïÜŸ,¸M\tGùcY—tšIš2SËtKæĹ/¯‡ αÙǼS(Ƈ}ŽZ)~W¢)K;À*¡Q ðWyg{VÓò,ѽw>.U,!íp÷…¨÷š‹&õæËÓ'ª|ªµ†‹  ×0¥¬_ÝáH(Ó—‰…T<’.£. ñiNÿÝŽMœ&z $s«EtíCÑÍrn¹¬r[YK›!¼â÷{ÕÓ6›Ýù¦†î*'äÛ£­$¡¢Jå5¯ke¼g®¥Š}I¦Èääúã¿<ÜšÌM‚;×D¶8+†LWaÈD*•úu¢÷Ñ$öTÍæÙ•Åó{)Ùx¨ ¹sdZè¨9œ¿m–˜·ˆéÈf’ѫȼo{äøO]Œ£bŽeóiʹÔy“Ëtc‘ƒ©d–hàzmămÚ°’Æ—è~‡@Z¢Í‹dáÌW lÂm½_2½9#®#ÍF²…_ð:Áè¾µï™ÜÅjïíó˜hVtÎÏÏ_âgX1Ë=Xô/#ˆ¿¥Eæ ‚Å~/ÎpQ+Àãj§ Y5BYJ*µŠå99@â±xê˜)bŽƒ½§¼tº[³mb:BxD1ÎNãºjY»†˜?©BúýìÏVýß#ó¨œûñF:®½7À{£ÁÎÑ¿•WÊ-¿ÝO‘ôýqØ/Èû;‰¥ª6·¿4†2êõGõ“ÊóÖl¢_½iˆ»/wòÂϨ?jp5™kO.©cÕÈ·‰X¿ãlØ  u¨Tüü1†ê¦=®ì3ÚQÅvŒôPÑlêÒðçËúÅ%™zúÌ^»ì‘¦.Ôäú£YÅKµ7’|^ïŸF|PÓœ àLg·¹Ù 4­ƒŽä MÅ :¡aè‹: q”$מ³ƒGK¤Æ~8½ Ï—ýŽó)ÿ´Œ¦QI_ô~¸‡h’C²\¦W°h×úó¸³_*qÃŽ™}ÕËð ؆Ǯ‰ŸŽ³ …ö©Á{0¯<†,r¹Ö·°–÷J OØóüÓ}4UÊ2 ç~±ëG9UõQ•†¿ºa“ ŸÖKè1ÑY¢Áèo_:è8Kcd6Ï5VòžÊgͯùДE…çÐ%š$›ž×(c˜A¬½¾žû$‡á ç½*y0î÷JköM0ú‘‹hâܸ$ëºa5‘ÜhÓ÷¼Ö Ùmï%ÿ»³©„Yœ–Ø £N€`ä”–ßźœµÜ†>G½~€i?7k­8ó~1…œëÇU•6^žDŽÒ--¡пÀh :*wÏ© Gž â)ïëÎÒщ›°ðWŽU?Ã×æ•ÕôâgC™i±UÓèn·WÓÕ#ݽ!Ðãiò,¡ õ¤ñRÉÅîRLö…Æ©FºvMŸ`Å!?ë«úÝä†rå±^8-[ïg î÷ýzHˆÓÊ5Y)Ƥ´ÄùPäºôUÚ#ŸîTL~™ý€à ´`ݨ<³¡¾%ƒz& 61ê4¬ C[Îdpd,ê˜&þñvšrˆ`,»3ñð‹Ellµ : [ƒR=V°‡—ArV‘x½uÍÍÇJIÕR¯Ü}¹þaï=É•uOË?$ÕåÊZA'‹ 2þ´¦"ÌmÜVþÅñ?R…:vdèAÈëwLª›ã¬MáÚø/J¥’$·ÞCcã‡Ó7+äwþhR«÷Ú…)ÅzR¼ÁͺvA§º?§,ÆErŽý IY¦ƒ<Ži(èŽöXû±1‚¨$V@·bÝ QTÑWˆÊl»£çyH=Å‘úÚ‚ßZ:a ¥bÈyÅør®ßÜ>t»üùúÛÇn-î—>“?˜Ý!@õ%’AüKévÛð±ÒRžaá;o¡"ý³ºK54;ä;óó‰ÎÍ]l(ì9ÉZ{é{H¢ÓRò·Ífà|;d¤Å‰Ü^²ìŽ¾<ü¦w%/Wú`} «)‹NZ# §Átä2HÃð“øßR ûSÛ—lgõø9¿ÉGÒ›¨¥ Ыc‹š%‘˜p¿XMÑý¬¬PŠqJÄüÈnÎ<ù0lê¡×Ä榓üÝm->;Ô¾4g±þŽbùäx2çM™¨m³F/Ô¿€ßvÿú6°O2‘ëTÀíŸ&뵭ļ€q4RB‚þ焬€|ñùú«Îþ­¤ó«¡A?èœü¯·›Ùà¢Î­ATœÕŠÀUéÓõÛ~3›q>X ø±Ì„7–®ÛVê YsÔ…óXK~S,zVXËkéݨS>IfQýeŸð WŽÙ&ôKôT¢–â—ôôGÃ:¸$Ø«]é¼{š|ò`Ì­3ÄILQÌ ÷sˆ†±gim Eê™iGôbòo£p§¼~tªlëIëž1³ï×…æÞ2^9ÓäoòcqOß×äzƒ[+{'Q³ï8Ñz~gWeŽ‡µˆb«czí&ÿ•+6Û…¾ìàcWVlœj>üóÞ`Àe)“ð„àˆ aUâW÷ÐLÛ߈ySÌó¹Mv¹ ÖÆýüžyEVj›ãÇ(|b*µ?ji[d¤`½ßƒ ‹^/³1@í¸ä|sm0Åû Ù¯$|—‘zžüÇŠD±•²ˆTà:XRéœJ èªäè>)Š¾Ñƒ>ž‰TÓá ë¸ ³“—gVõ:;ï !?bUZͶ­×^dY(¿¯¦Ý/ŸÝ¬;bRóÆJxXŸäcÎã¡DÊãWç®jÕyÕ´¿ç6>!'ùývƒïŒ ½åc+Ñ̆yF›hv?¡ê>äi2ª`{»W½F°UeåÛdÑïÒùòvhñÃa«]#FÊ[•è,ð^S¨6tÝÒPUMQ…{YRnpŽÄî„pów¤ç7BI¶.|ÕknD¨GD{3÷ms§ì—ÉÜÇôËNœÞxbΧ"k©Þ“õÉ({©/F_#È w‚$´*ýy²¼nÎàÉv³ÈµÐŒ[Jw¹×S.g„—§7ÓÏ9^˜  ?›­‚ãõØsšÚ Oîf­BhêNjšÂ®âP7RÆ{âT–¥£ãí¼ÙåÓ“vw† î.—ÙCÉf¿û%kÃW[)CɃ»]FK…õ¦#¦ÙµS2ÒdtO'Åp f¡Å±ð¢Ø¼B×î$×£#ZÂø3ü{àÕyƒ€-Ø÷)Å[ßrg0ô»ÞÒ@ÄñÄQüÅøåëCÉwr?ôº…àUѶ´ñ¼(ÑÉ° 振ƒ”Ú¥±×ôÐÑuŽ°:¹¹¶Ñé*o Á6šÙŠFÇaMž®½^‘»ØÇ {gDEÎâ4÷bò/Þ¦ìpk-úh'JcÕ5‹© eM§Pë¾g˜sŒétðЛåö™µæ°À®ÒæjêÛŠEº‘Oò/,ù&›€-ÊHkÚëmpµ[ÿ¨l€ =Ãì ãí•©ãNÒ†ñ.dHg…œ›^3 Tzƒó¦&²¡[¥ãEÔ¯uš´ÖÊl¸™úìI}eWÉ”ßMð™-ÌF·í€ðƒ¢gŒÍ’+bM#;µxe>6Æà"’t G«ðÍ“éåfk±)‚Ô¡*—(^ å²_Î_˜—HÒ%­ù`‘-±øwNb‚ÁŽ9®Jê‡Fˆ’røÓÖ“¯q-dëO‡D*9¦ÙÁbzjœÃ)È?ÓËzm©µÉFàz”ÌÊ#\e!û.© ë¾ïþ“ÈÙ¬'"Üä’Ïnü#5ˆ±êƒhL¨ŠßO?ÈCW3Ú3ôGÆ!k­öÔ‚úö¢¦NÍ; &FnÊ/ÿÓj¼òN?{©eÏJžˆñ<-4kÙ/nPhéÞ<´“öoô c©|[6ÄÔò‰Ö¦ü_lá•.ÒLû?SC ÆÃSÄR…Y&{a±¦è žVsÃë}i*uSÁ¢Òùô÷þ˜9Á#²ÂÒG2ù9]«V–¤úõƒ<¤7ò¨)¤J5¼Q,é\ gæÒLD÷Ðg¦+XךºÛ4ê1Ê*CßØi¼âjû--ìÛ³o΄×=ôé8cGÖIn5”ŒìÚÐÜ’:Ìcb‚Îœ«¸RŸ='ˆ6"~ÁzÈÈ$8<Óu¢žzŒH?ã×¹5ó?=` 1ɶîí÷aRŽñŸœÝ{ â‡on¬÷’ÑøÒ jŠØä¸ÐEÕk†%¥Ƕ7Cf)©5&‘—†šôóïSä‚’ÝZ Á:„‹.¹~nÖ>¦_vÀ¬ÍÕyÍ!…ýîsØ-ŠÌ66ŸEÁŸõ @ J3À ÿ›$kê«ÿë¶BüÚî®")¸v3Z]ìA%šµVŽÔÆÌÂß:îõ>·¸ÜíÚÉ· W÷œê Šš’Ñ *>¶ME˜¶©6µÞ×Vá" fWůÞÛnÚäimÛ819ÈþœÃ VÒd ’uø¿ÐDk¤û ÇÆäTxk¢}t[¤š<õxÄlFÏŠ¦¤|pä  |J‹üj ݋Ŷ©'5¥êð÷§·ƒm óædüN»y]7!øõ[‘2Â[ E&Ò<™æ9Ç&àt‰Ù^.ì^íDo"âbQ Á¡‡n߃vÇVÖDÔt[ ÝG3”á»7\†å¦œ¥v›œ¹%;Ý5ö•vt–Ü+Ž1õ"îÞÎ rœªû.óØž2Ô|Gulº;Kòž&r6åþYHQ}B(e8·ˆB'õ^lXK‡ð%·>l”"»)¬:®´|·Šaí?òXw~^åø†×ÂLo©±Ÿ©]²<ìrû!5^ë‹ÏǤ­ö=&¦õ…³§Ñüî‘XY˜ÿ¶€>ää™ÈC»…N|¤ü;}¨MúÀ±MŸÝÙŒw¼kí•îÔ¡Lk™@Ûëõlk¿²÷?Ø]Ùô4W0Ö §€§ÍKØXi:Ð2–—Mí‘"q•ßÚøRe{ᾜâ&¿ˆw%eT#Z™Xz„ä\x--ûÉ “M„Å¢°“*Ù× tœ†øÄnFÊ(½æ”DÀ øÀ{›PêRÍ€R{W;±>ŠäŸ¹O•¹ZM„bÚ o«tS"ŠèàÓ• @u½&zfÒ£/ÔÂÒ×»|)êx~ã)îI#1 Ôã„õÈKÔhõ·ó«ÓÍŽ­/¶p¸a"ìXg¶ Õº°EšzT4|²òî¾¼K4ÈÞìKIh%ÔøsÈno3!6è³±%Á—WUà,ð¶äÄ*ÐUdLÄÁLîE•&„'ŠŽoòLcÂ?g/foV‰ˆ!eJ§kÊÖÿŒo¦[>Ðyÿ”‚$N…Ž÷ƒÎÂ(|8ñ@;ƒ}FPH±…Р7|úŽ‚èÝYûu*þ°úá íÌÞt‹X¸Ø¿•½ßE0'ný«÷áÀx¾l¦R$Â`:iÖ«‚¤½î3dŠ`6M9¡8ˆÓ´õ4óÄ&wѯ©¬ŸBà /‹¹³‹‡UQØ 9Rìòὶ|åÒ’Üøõ*á>u¬½ ù³|ÚÎÐõ±'”©Æè J¢$-7ÌN+d¡‰t©ˆRb8ýDu.‘T¸¬`rדEr6óòu Xƒ“ö'×òh­^ë¼ja¼psÄ éÒɨè°Ì¦ÔÚœÂtëÄVü¨Á–ŵŒôøÅY¦v`Ùñ½V§SãÏsAW_^íâ;×'-i¡«'C¾j°ˆ«!}ìZŽµïüè ôP™ã£ñO$Ý‚ªhÆ ÚLÿyÓ§‹ÃPk:X6µ’Dšè™v¸Ç YNŒ ,[²x“³ê£Žcä7§ò‡"ˆòÍ÷`;$âµYä%£!¨Î8ñ2ò`c(Z yãÇ¥"¹yqIh£¨Ä,½Œøº¬•|ådÉOJðåO뼯§Œ%¦Sñe‰c+OUä}’³Q¤-Â.ÑüóT&É’_’ßsÒwÑoš¡é5lÉ ^¨tb« ;.ÓE"z¦Ó9Éǘ¸Qþt(ãœ3Š.?Ìü~äXÈ^Õߊýß¾X£oñû†¤YYKÚH²l*ªóÊ$•´¬ú{.‡³+ýD¶ÜäŽ5¬s?‰4£ý1œ•;’9•™üŽÅ@¢÷lœÍSþ ¿ FL®ÕE¸gƒW]„[d{¦/¾¹½ºŸ¾F–*pÉ,J—ÎYëúŸñ÷h¾Dv9„¨±ª!è:'?í|g±±¨gc:Bs¼|ñ,Új…EëWx¾·a‚m UÚL!¥O}샺-èé 3®üLó ªÄ‹v-ÙD|‚9…‚Í{+h®¥ÂXÞEn^qÉ~æDwJVýp“tÛILÎLZûÃ(°šÄnÏjˤfJ3·TuŠ—ï=ЙTÿ[?»ÿÞø&Ñ}DÊCmÛKt-“#­%–gk3»ò*‚»Ëæ* *q(Ø=‹Á¢’йɋ*ÏduÆžb³»þ¦ØŠpgÍ«’iöÅeb]ۙʈe@…ŽÃF¢LÒM*Çß@Ô󰽓˿–ŠÝrËüàØ'~4P aç¹»þpÂÇò`›®ì‹ W ¿¶"&-{ZA¶‰ýÁ–bm4롘½ýûcôBe7ï=ÎÎÌ°z¯-r¸CH~6q#Aÿz"txw½Ok[…Œ ³Ý ÁwòÁ©$°sX­ácNdYƒâ!ˆŒØ8Œ6Q,añJõ@²nÅ|¯øÂI”w@”f&Öôa›¯Ú?¶¸»_¬7ùê¤}–EG¯Ç¦"gþUÁr¼÷÷ éËïü#>$Èl\ Å×úµjl Ê¡æ‹dôÌ¥tÁù)Q ±cî-ÍÖçžèseknÎZ‹³®èIXd¢uô.ó×Öa:ž|¯Wª¢=±”žÏ„5=K·×—¼“î‚6Òëwé­ÅªàåHe"JoE&ˆÚŠÞÒñ›×º‰åðÄåŽ9½If}®³Jœ¯Ÿ3,å1õ§[Þ=ªr]Xôé@j^ðDHtD‚åÌÑË¥šXç¤Vù[½ÙMÔæÂJú˜ïm@ÀúÊBü[¸EB4Ö>/·Q±÷Xúdp£.ì&ä!<œ¢xËdN$«òût‹x Äªp[þÂXq*0¢ïQ·T„¶³UêÙÄ×±3™ªñÛg ]Ž¬¾)¼éäÓÂc¬©°NsåUÔâ…«¨wÖwenˆ·òv‡ƒ¡~._É·òç­ZÁ5onkÇáV,)¡%¨T’åçA)fˆ­bƶ¹+·ˆ$Œ>Û¿J†I Þè‡nk9JÁ¢¾Äˆh‡åAØúÉlñÚT/†´h T¶ø¿Ån Žû.sc - r*7±‚¤Åp)&a3(¤ê¤ø~qš"ÓA*èôŒÒ±8Ì uYX2”wíÔ f¸£'áTˆÄ¯èÿÆ#.Ò‰l…€¶ÕOèfà>©j*_ˆM-K>Q¸ËCt¢¼§m1€60rì+õ˜|˜Yú}Ã÷‹GH½Ÿ ù³â0¿÷¦/—àoÑ7«ÛonCëµ´1ƒ$¼aê­õª®—e·…úmÇV^hx ,>°6Å—ù²d/ó͇ÇGZtJ}br\šþd[¡ÍjÐM¯‘Hp¡Uáä!'̦%Í‚ú½g#P3½çf¦¦á_ñ¶o±Çqô»(³ ß{ ïkÒ'½ÆsíÐlÔˆd'–¦¤MµÀ¯­žñcÓëlW“RÝ‹i›í€~WL’eË;¶ ›SQÚ¶m¤žy‘×£ÀtæÛá†ðÄÙ\Ô'åI,¦:¨v «åVºzÞÄ®½1 ÈŠñ¦'åp‘9Ù«ƒܘ!uƒp¸ ô›ó†½ç÷÷!Å'2œdë)=Â솒£rŸ‘õª” O„ƒ)­ó‘]0#[ØY‡Üò±Œ™[%hv½êü¹ì„Ùp”{6÷NCŸÌòÄE¦F8)ãcLã×›þ°õeaaö®¥MúO9J“¿¡B, Ý­— 3—÷;îše鯶„XÿFÊÚÙ_eÖáwþüiÔǽ/¬¶2 ÚJ3íÎXBâ[lµÝ†³ÒÝHS{0nýVîn£eôdΆ³uB‘Þ e©`Y±'ˆG47„Uù>a˘o…üQÑ’s˜]=?h:ŸéüÕnž‹fr—3f úèºlmUTÓ“–‘N貋ûª™w¡äå'®_¯›_¿¥Ã÷2YSÅðŠuuØ8³h’QV…è‡&„>¾`ô'æ‚e,9ç©D"…Q™|G6$á¤ß“–‹ÀÓ r0³V„N‚´<÷)¤ñ3åüPM`l…y%”Ϻí~»S6SþÌÖð?ðˆ‚~:²Uš}Ãú®ü SýÛqÓbâ|·¥h» ªÜÑ9ØËñ€ÍíŽ+0"#PbITAö Ô|­ôH* ˆÉ‹be›;VÊÈ檋Õ1[ ¬ŠáeÕžjBb§2F(tìåVèsN½bžHõyX~Ã+Í|OÖÎRMþÛJ …!-Qž2 «ê†_‚Yxâ‘VutëI/ ¬'Ì–¢*f\1A¨ß5þ<ðnˤ3Å£Z”8 ÿ‰Ï§ ‰“öÛ9ì·,\– „;ÏC5ÕÛÞäTOùöNzLÀGràgê‹ÖN‡f›àY’u ˜Ÿ2†¬  ˆOl©ƒ9ÈÕMÊ>Ý6‘4¶"ow‚ÈYh"LÏîšì\õ˜zº‘å4W0ù³ãÙè¦exêT¾ãô1J7±´lƒn…zÔ¤ž•9­ ºJp-MBúÜ(w ›@ãOB&. åÕ3m ݲò¤{ˆÂçp{ª?\Ö‰3»ÙøXg &6ÍzsŽ•óU²ÁülÀ Y©-bò0UÄ‹ŽUs‡ï¢…dÁÛôk‹w¸¬jÏcí± ë+éªÎ"&\c ê“G{‘÷×K—sÑuØ&…ëæ…9ªÓú+ïö·Z 5º»Xè^åŽ.ƒÅAˆ'œHß_%Œ…  g~ƒÍ-Û³]ï|l³‡?0Û.ê󣬼Š£|µ³³{Ñ^J+à×±äC¾~‹7í _êE°­½NæŽf?ƒ6-; 0h¬O<¤¥ÝŽCa̪ú׶(ÈäÝÒ²"xç·Ìà;:°ø¥°bÅÇ&6çè!÷ jNÀ€›Ê]bÂòÅWAjFÓ^ždå2ïEkÁ¥ƒ¬Ù…YDÀ”©k ‡K/éÕK°;ê‚7eL2I7<£´[g¶}ë­Uj„0H¶*9Ù/#¸Z«ög23ÝÒî+4j’›¼ñü¼ç0Ú\t¡nSM|)0¤¾?ÅΔ¿PŸôua·~5ÿÒr4Qs˜¼Êâšñ¼Wö'oмð¢Ïò†‚¯.l~Ë1Ð 1kÈ=ß}&®3X/-iÅû˜s͘„â-ûMWXØËózCÀŽ8 ¤‰¦i†KXÓÄ2[Jo4•¶Wh#¯Û:Æ:1‘Ù©`~, L̯ÆäÝ–Fš…Ü$‹¨÷ff±ËÆ5|ʼwl:iƒZ&O°EócÜXwÖߊª2,§‘;*¿4OëoÌò‰PFtË |1‚…Û&Ýg¬Ü˜ÿnØ ¯E¦[2Çàp“Õ/ŒWã:-³ƒ,,„2кstnL2i½/ ’#ÞµbïVà ¼ÒýÀD {¾à¤–||Rve>þÄÀö% Ý0÷à`ÀÕ‚Qàü›Y3œ_0ÈHRE1ò7(Ÿ2K8ªy@ŒO¾'OÜÜ? ­«£tNšu§~xœ,­añnP ëÒxÌ-_•§¡]ëŽúÝŽXh¯3SʼnŠ³š@-ö†X6žÖ+c”e ÝjZè>! œŠ[õåI'a>óÉ€»ëΉ(ÆUûqAàÐMB »mkùr¯)©Ä>«À¬nè6¥Ÿµ³ŒÁöµLòú˜0s×/Bäž S‰ïƒÉK¸¹¥³ÊZ`eIy(ciú}»Ç[…ÁJyD£^µËY‡§‰„,Ùê}®m§¶à‡ú|ðÈ꧜-÷8íÒdì/q`ªŸ˜[²Ó* †ë%³Õ%ÜÌŸ5k"YóXàõåÄ–sÈs¼ÒJ»|§ïú³}r²ú1»kÍJ½ýØKC#û ÃÐz¶ˆþÞq÷ÚéQK‰Ù#\»ˆeýÞ´ð@~dv\y©Î½¥û] å4®`?­·ÆÀµu.–´çi2×SI¼§Ä b-•¤uøo3Š¼›Òm.R~èÆ¿¿Ý]ë@ýA¬hÎÛªAíØ‚€0'À<žƒîŠô½hIò2Ÿ¸/ ®Pþ0³V*ÈR„óâJ¢ögüB M6°ŽVî!Ò;'üs…‰I 1,ᆑ›ØÏLÍîÄAx™sQÇ‚s~C.Ý2\Íäé§Æ£ÛÙåuႻצþy²Ìx_H…Òmª2ó!BÊ°Ù«!$"ŽžÜ\’_ʺæ;-ŠÐt_å·ðô¯Ï­Ù3ЙyaÜV½âZð'°¬N•°1«}íi¶®*¹¯PüÊëƒ'r÷)u‹ÂøoY8wóU½¹µ6Dv$á&Ƈ8£N0@»_å®â0DÒ®Æ~,NtâeÇUîDx©‹·ÊÌò§YÊw/¤,kušSÆm¶Kw`O7×M0!Ÿ7=¸¦Ý*!d|Û(&?¤‰zo›G|l_¨8½ØÛ~ÀÞ\€d¿ä¢æÌ°™–IÚ­EƒFS¹°gI òã_4ltCfjgíçÉhE”è*½É‘OõSzú>H=• Éöpz’uõˆ‰ÂüºŒùë³ B®+w—/ÜTý¤îtÂâ¡¡¬ÉRcµùèˆÕ¤s5„ÛHhfßÿÂÀìE|òk6ßOU†¿èqH Ñ¿»$Ö ¾Ê7 cÂûYú¿~Ÿv^Äœ7q™°<-.?̼þâý€ûy|H£M:_O.hJ{ä8=¢ªs‚ú¶*X$ ™i<T÷Ôó—ŽË)ü­|ÓŠZ!ßÙy#GÚ¿í6u{+Ì6;ˆuöè¨d~ D±ÝéQ+o£2†IƒÖʪI˜–Bba‡ó<`œDE.6Õ¹ßn\ ñ¼Ê~¡ç†–ÕA7yLF1ö²çœq…2 ÑÓ *‡“è¾iÅIå4Sé68OÚBõ™Qœ¼ÿ¸¾8æ¿ó|Áb§Sl˹ÈÉõe´fæÌvÔzË› ¨6ògêx‹ÉÓ…x›œ@êÊÏ5¾N|ÂZcÝóü–ªvî 5=l­]*ç÷âYNž‹Qù!ã\e WŒ¥—¹z3QJEŸö±˜Ó~Kòš÷ÏS Þ½¢/cÖ¨W²H˜]ª¿€ õPûéŸú9sü•¥¢¥’ÏØ"뾘×%°ZÛÓ)h[7¡뙧×vâêˆ@˜¶´Ô ceZÈ£<æÝɱ®*žTÂXm!ÅŸ¹Ð.ò\@ŒüjôÚ °«6ž¾‰ufß."XbŒó‚¤$7m’ž pƒÆ\¤‡lÿ®sWˆ§,Ç^_Dö1GñSè|÷ÔŠêë^b$¿¶W©òs€ð2( Y½õtXaâ‘Ò“Ö¦m<{{òtôHÞ5‹©©'Û2Ó¯{öeâ~Ý=Ã:=ß œ£ö¶"gÖ˜Á¨Ÿûh3¥HUýdY18ü1"5ˆ¡\µ;¤_xë2:Y55 “·È-¿™ìc5­©¾UΞÂË­(Â}”ª•éľ‹µUÎúüÎ4³›,¿-P¤3@¹ÿ.„RŽmÞ®|bTÇÀÊ­¸ð„êùkBl÷ÙxñHõ[`‘fÚàØqS‡ûœòÙz”RŽMubŽí$½2“¶Î-¡”%%D]&k‹‚¯:É[z·ïQŠðUŒ8XvõÌ`ýÓ[v&Wß".&]qè#ÖÔÁ!*øuëß6¯ešqšÖ[à¤#žƒJ¶ŒÍñÒ†ñ˜nˆÕŠæËlQÓ)ÏAÍX\–c\”½æq4~’ZkyÓÑJâ\á‰òx%ªC û8ý‡Ÿ±ÈÞá~¶¼FÙU÷©žùüµ%õrØÎøˆÂB Üã2P5|·ãLñîå¥>¯7‡TÓ‘ãÑW_KÔ/¿—ÐEP3„ÁŒÃn¯!WLåövu÷ PyTš“d„G¡(]ZÊ?(õuÀæC_FWF7*³vÆ°¿Ù`áä20þš}Ù^b"£Ñ±b,eiÒ¬œ¹‚ä„Î&™ËXÔÔþ:Řa&FSì4½ ’ýZ”p£<ŒÀ!|Fb¡s{¨+µP¢">@§òí§zÊ Æ¦—ˆz÷»€‡ê´ˆ¼g7óŧ“¸àdߦAœùn¨}±·",yÁ,¼Mç²_•v%Væ\19’ê&Ó°ð-óµX†kÍEÈn"xdì#÷_XÆá÷Ù<;°ŠGxõÕd­²Á°¿› ÓÆÉËsÕWö^’â§)NS¨ºÕáÊÏÒ^Ü›DÝJ¶çþ 9ü6긄ñI>°á(aŸ#Ž'ˆ›èŠè Ïó °ò}s‹[N+¾ûbï¬ze!xG%sÅ(e($²ÆY§zÎ6lt À9ÆÓ@½Ò¹<¯céàRs±Ÿ¸ ³òÉÓÎx¬ÖD”ÀâÁWþ(\h¿FøÞ|5CÞ €¾¦Îû[w»™…Q~PÊa'¥õ«€ã¢üÌØüOWÀ–áªQÁtf#¢Â}n©4þ„c?´ž¼µþg¹â‹RïD·×^½¾=ÓX¨;VÙ?C[ì©¿JWPßüï‚+S¯Ãt—í5øÙK5òv(ŸŸ˜qz\^ð€Ä£]â’Ú¿àÅQª}^°*Þ$ìû Ÿõ’–Õ?Ä;g!v‡#§«=bÍg]ëÇÑmâÅæ–dC®3ï6•â!ûÓjÜá2 ¬ ךÖ5^qY;Ô™„›j¢tÖ¨Îõ-~Fß¹wŠI'ŠæíÑJÞÓ1¦„§ƒ†æÅ·hƒ«ëM8Z3&íÍVYøc’r‰ó½ðêÚZâ8º×.cd=ZeêjpcQÆ~mc*Ä#•û„Hù>´¸,R%ÁégNôµr¢¨òR0. ÷W}çŒO²î#ÖÛ;;Ý™õÚh¸.ñ+À-…ë<ë4úùzÕR@­Æ~~œácS#ªZͤ}¼íl—ÎÿRÛ/ÜÑÐ…dc;Š#—ÜØë¨"›Ú @ÛX_C“L'Õ/ö7Àü!Á63lC³¨üʦ©“L­ aè’9ÞpÉtqh·YaCŽ*0š£cå^ÃW\ÚÉ/ýb—«Ê\sk¹ØQFœv~kǦÅ}=‰œ±:ô;eÏüqóBÁu¼o%Ñ1=°@;„×Vô·¦÷‰r­dÆD¶z§¡\lfTR­Ö\)\ì›°Ç>g ØE8vµMöVf9Kî ä~T¦Ó<ÖK;·[«îyÊ¢ÚG(%,)ÂXe´¹NFkV°q*ÅÇ'Ú"©ü¬9ÁMóNÝÂS¬Hß¾äÈ$Kýzx~ÙŒE&žŒƒ®Ž1Œfò¤Óö8ä²ùÜÚ·\W] Ó?h¥Ù.}9бuF¦ÊʺOãÅ]çÓ…a&•âÏ'檔-¹ªBœå¢J¥ñ.yØdè™ZµuÈ„€‡rgµÀ¯žR•â…SÚ·ŠþrR ™ÇôÅëóZù¼ÅêRsó«Ø†Å=¹`¿þÆÖŸ“h˜A=ݘ°jBÃHÎ㟊1Šf Ï r ˜§‘`\˜,wÖ *!°œ­_œÜ6fŒ¨ˆ)Am7d!§05ÈG¼ Kg•bÑ óW6Ž™vG³&d/]ª‚sµjèìî7îÜ^îÓMºî* âºÑ+ªÇØ”ñðšW© Ð&*Î Hnzš®/˜°…»ŠÃ ——ΫÇíé“GÕÒ¹Ät´ƒC>1ï+¦Às¯ðe«¿XÖ“ã(<¨ îLaÞ‰ùZe÷,«@Yaì™ê ^Ÿi¿[³¸#Æ© ­•”Q\s½ä„òk¹Ÿ€™õD¸ªÂ#¡‘B™ º8¿KBrØI2°Æ»V˜F†M b©‹þ}Ù¤¤ûîƒÉ›˜DØb•S­^³ 'z¶KJ柠è8Ûv`§ƒÄŽÃMÍu ±4qLÐÛÊÂO(íÜËy7ú0¨÷eòü¹‡¡¢Qoši› î;€AM()ryþo+ñ¶Á‡×³$FB`B¨nsïMü7á^ß‹ðᬔ…:º†‚–1˜.š©½ÁDÿ““y6£/‘Ãó»ùZÖ­!êÔvû é©4Á¨€Áð/~AkË•ðÃÌîþÍZxXø¦ €@E²òöéNͤҕÝp ‡]YDé…)ý¦Ëë™ãgŠSl`>nñ¼Y–´©H“—Ð’ ©®ðõ¿B°Šc³'m%ð' É°>´³•cûÃCRˆ”,òuQ÷ŸEÕȼBá- ™ù,ÜЬÓõŒ¡¦+½!ô)éž}éÛ3À»Íq0Ë„V_eü§Zí±ôÓQ§KžFŸÏ¶1ǃ`i,&U:a£Ü´®~‰½û­ 1‘çióIÁß$a¨„üqM‘z;+ÃPo"8‘X£³$¯¯Á¹‘RªŸÌÖĽ'$*Âi °üÛv«nWônôFe{«äó„[bA%y_äçŽ …Wü¡ì‰&ðZ´7£v sÇ)§oå*_Ö+=dýy ãÉ{_sF,ð¼,Ÿ åx’L  °†G3ÈOâ-)ÔL?²™ªÈ< äÖyôOšß„s»±KÁ[ëÇîb‘Us½|•$gHuö¹nǾÎèÓ¯· ¼O+˜ræéõ]Ž\¼ >º%ë±ä=¤ã ”rÒ(KŸžÌ?a‹1ÒOIÿýá%µÅ«ø.¨ÂÚ÷ù÷Z1œÐ¶7.ˆpßm^?1JpçÚèŸÞ̨]5Ö!à7¿/3–Y(VXìR+©IkktŒL5jŽ”y7oLÛ1(.M|y RòªΤßÄP†Oìµýj„µ&W iaöÝ3á±z¶Ü¾ÌƒÖjüôí¼Š¸Q">9 qšHþ3Õ®ŽÁÍV®y»§³?¸Ô[b”éPJôã»áÁŽxs¯ÓµË!¡=‡ÀÂÜš§jšÄw®ä*^Ôpõw{tfìÆ/¯xÞ‹¿àåýò.}gź]°yahÌúòŸü©Çn ÿ¡½ 0±–‚*[Ýä/˜d9åùæ±³¸4¬F*cú%ä²Óõ·æŠÉYòÈ3B7ßó/×ÝIj­gG€I ·Ô•ÕçÊ)¶7ò“tõÔƒÓ„ Jxö¼Àó `yp8¾¬‡Ù½ƒT½î¡Ÿ SœÝ=áj2ìÕdÍhê=S4åÐèw4Ã×ÌêàžÐKçHŠÚñ«—F/~+^aó²S5ÖIðU341êõ)çO>ƒo+ÎÔ!&v.¤Æñ›x¢}†w-Ä 2 ñ… ä §¥¨XÜ[èâö!Ï’¾Ü…ŽUåK PjiÌM\¨·Ÿ¹¯°Üf/LÄO📇ؔû£UÌ3´¼¹×çÏÈ`ÉíC;SÌÚ0®ëü ýó ®ë÷Ót_w9•2U6 ÑÎÜØÝ’õ‰|Yö– dH0¤xÅEŒP4'» nÖwÒR¶A‚˜Ý:›W ÌÖ£&Ñ!¬e¸Û'»ë¦XÞHƒ!ýIÔ 'Dû€×ͤ?›îð•¬K×øIK@ŽHsÖŒÄïUÕ®–]4oÛÍX^J­VQ¢­ˆGva,KŸ-jËÝ•µ¯Í¦—X š"`–2ˆ<ñK²$íýï£=Ÿ|Úæ¡&o;på;t«–áJ¸®7”g;4LWÍ‹“²hW)9ëS¤þàâjí‚ÐŒ²°Ù¸Z‡ªõ›‹þ;ñ¹‘ïê½EÌ}Gµ¼g‹ˆ{Š/JÏ$ÁàîÞ||¹~Hè°Î5@¹×kŠ×±;Eð²âc@X<ð=5( 2 ݱ‡42Àr±åÊÐA›/×uØvXMZjë…LHcÓÙ£§ —×PªÓ3†ÔÞmH–æú?yd»ýqŒjÐÏÙø毧êtí¶‡äª¹ÐÓÑ1lcNIB·ÏÙø$©n©§WkOTãnãºÆè8E×,Ö⊠†ÑÎˈrïD¼ë­´°Ó)ù$Y¿Ù3ËÑÆRGìt ËxKÌ]÷&Ç&Ë«­„û|Þ%º§v`õóÿ8’wâ^)ê>eÿÅ@ÏYÿ\=¯FÎŽ*[§¯ï#eÝøåýï?wQüSEv`@K Q»³Ð­½„5b)rà ©†«Œ—T8Ò Ÿ[Hº¢ŒHE…Ä~É éÎÂðo­m³Ödèhª³ÊIh\ÕÚ}Vq&ºq [Ë-JyENo'³+«£vªq1Òæ˜x‘B:Æç:Œdëtij× *ê+=û )~‘D Ø_g0•Ûe§ÄzÅÁlÆ'Б®W¨Î:ÉÙPüœH¦õ¡üª[„Àñ ¦‚ÃÓúAo¦u¿eÌ,(èÖÔ¦·hÓJ£c2¤Š1¤L^¬wά·æÙ<Ã:ËjÉŒ½ð8S˜³ÒV·!° ¡H·i•p¦*gT‘ܽšEóêà é¨dÝbà%åÄ„åã§ÚPÔ‰†b…sŸŒI‘o¥›ŸÑ­®d¸¨Î£°Œ$*yVÜÑs\M¥õ`’žëm"¹·»ÛïaÓ-ãݬ·°úƒïz›KïŠÏ~°÷¯Â‹QÖQ$Àí’÷¨!#j•(_ÃåGŠ%'À?讃 É}xbýlØ»=ñ4å@#cQú»¨I™ b.™5x—TbT³aŸhí<IšΩKüQ..YÀ*Ür¨Üù9Û°ÉÖÎøâ‰ánÿ¿ðêpÓN/íWFmé scò·=ÇVšzG¶{nS™(ÃÁf›Ú…\Ú±žwΠw ø¯œ–/8®b<-ª*Ï°ŽF!•w{\²óÙßé÷€}ÛH]ïdð@…ÉÄôX}£Îû0945Љ(á­ÉAޯ䳴fG¤ø§hºN¸ßpã;Œ@Ìùìä™:KáxÛt:s)hZK¥açTÞ%®‚Át¡G]”l%·­ßôU R-Pe7\æ¬x’µK¶®LJÙc­’—ö<÷aݯr‰ÛDúúS+ĺdë^CçŽx‡[b¢^Ój‹Œv)îÀøs“]ÂÉ `D *±Fïϱ@D}ÁE4¥>âõc/4oQ,{G#͈¥h~å~ qpq!»º•ÈV=^ÁGWÃÇ%­ÆjÄ45¬Ð&R¨2Ƙ\3JgFÆžÛ.Š÷Aüd°±Yi4zÀeÊ¥¸m9ÃcË”t,Ü“èêÂ^raô Nègúù&nbgEô»ÙwZ7¼í(^¥ù&Wü®§Ib^¥Gâ"V( WñÖãH\&#b²å2ÉsR[üàÂXå ¦mmjìË]c«ªÎݸ<½ƒôîÓ©íVÇM1T#ö1ñl)-G£ªOÿCJÝ3þ[6ʬ·ýC+ÙßwþmOpz’Õ Ÿ† ãÒ»UG O-6‘Ày*¾³€u&ªDÙÁÜ ï†¶®¢iûôb8ËíQå››Ï ®ç`¬!Y_»sR(kPI°ñ2ÔP†,À‰¹4¶º¤¦†Ù{K…YÇ)ß‹ÛÐÖùLг'Ú%µu¿ˆ^Ð;KóãˆQ“ÛMÄâ™0 ½ÐfƒÇ¶]Ü£‘|ÔÏŸlD9tÃlhGÎÄòÏ9ÏèŒ?\£¤ßŸ:>´Z|ÏriPnBx¯Sóô‹ê6†…2>EŽõ\ßíybm¨eÒGîF“Âœ&ÏÒ_V„§8+iæq‹çJÙÎS*6Ù÷OgSv<¦ç”#åÕÍSóÌ“ª°ÊN«ƒÃ’Ä슿#)F9ËEH€YÜ8ÚŽB,,ÛEK×b÷!¥,«Í\¸嚆‰šo5¤‘WÙ¦v‚YŽA†*àÏD»<¹…'ݘ9ÂLZÁyuŸ¢çkËŽiiFB|Qmu8`oÏR_Z;²ç`)Ö4‰iøâø‚Þ¹[PK ¸ud <õÜ078ú'»Y¢ò_ŸoÊö…êÊ/Æl™VÏ(¦ÏŨ½°™K‹¦\鵟ì8tª(š€“Á\—¦0HÞçÃÐ~»†˜_~Üå3É2tìFüÓTÜ™ àœÕq´u3¶få¿“Éðk@Ž@ro‹”³1î@k9ž\DkGðA«Æ™Ô´¢¦['êÝèÙרF’è “­häÊüúw§~šéÔ¿jH…ý‡ÃÔKL%ä¿žî/9¤s~ð*it¢à¹Pa8˜Å>üþë›2ü_±#)âìOðû[O'ÌUÍÁä'˜gŸ(¥·GC9BQID7i’²{gûgD¢—‹díì©%-‘ÉÀÅ"=ÙÖK hÅ¢N¢•°±Èoû¶’³»ã"e±%>r³5¢‰˜úK0$ÏLçSMŒ¼1éa…¤{4¹Â…Når˜ÈÑz‚ifSäröMǬÄðäíþ0ï£S{? ÖîˆáÂtÀ¥¸¼,/€»á¡ìÓ„‘¨Æ‡ûþ“Q.°›á:†LGq}Cr,±eÿÛI¸î­NRÙ ÒÈ*~á0ÓÒ>ÀA¯ÌŒO³µW2¸äˆÀó|ˆb-bï4,Ò9»Ó¤ +R±Çm}ì1q3· &Nv€×·þFbñ&ÉIÖ~…è[^ø@IE„‚Èg¸„':GVc9bp¹ò2`Ë•´F(Ë0¦n$h‘á™™þg…Ö_‘ªø·*cR ý5¾'ÖØ ?á!™â ÿgß²ÌJ³ñø««ì[ \šžD^ýtsÀC%1 Ú~ÍÈ—ä.äeÇßÝ|\(féWc$ÍÛ,ÜMòÛ÷‰¨{fPèá*ˆï%’„h§Xª⼈kÁ¶,ɹR_rKÓW¶T,Ô|»³›4‰Ó*óÀINˆÕ—hùh€<É{Ùç¶qÄÏŒ¤ÞbCÁ;ù5Dò1fwù‚7†@…_)L]¶£Ó}jS5"Ñ›Š ÕûúÒçÑÕMüìãs öÙŠ7 ŒS½tV1ª”ŽÉô3s´]ÓÀKÚ~¾tb|6> *ßîªI 3)dþz¶óbKøûÌéçQß«1ž.L¾õŽQŠo ¡»ÚîN6ÖÍ-y’We³Udˆ•Í—•Ñ ´ìG»2Ù”D ”‘Bã ìwcç¥î4ñÈ6ˆB! endstream endobj 290 0 obj << /Length1 737 /Length2 28631 /Length3 0 /Length 29219 /Filter /FlateDecode >> stream xÚl»Œnݲ6Ú¶»Wk5Þ¶mÛ¶m۶ݫmÛXmÛ¶m[÷Û{ÿgŸœäf&3Å1ªžªdÔ˜É$!µ³uVö°7a a ¥ç(‰*)3ÐÓÓèi™`HH„M œ-ìl… œM8j&Æ%{#€‘žž† dgïáhafî 7¢ø· jm`lacáPµ³¶sµ02p»ººò»:¹Ð:ºðÒþã¤dbp67˜ZX›„ää5$dÅäb²*1[Gk€¼‹¡µ…@ÚÂÈÄÖÉ„`jç°þ0²³5¶øWLN´ÿÀÖÕÄÑùŸÀLíl2"Ê¢r²Ê!:e!€­1@Züw[g'άMŒœœÿ•5Àù¿”µÁ)ÃÿRÿÕ:ºÿ‡¢…a`[9 MÌ,laèþ…›„­©€õ?bcûÿQýÓ?ñÈÿÁ`lbúµ‹µµ¬ €\ÈÎÆÞÅÙÄ cglâh øönŽÿ’(YØþ“Éÿ¸ØXX{üÿ:ý? gƒ@°5ûFZfÖÿˆ-œD-ÜMŒå-œÿÁÞÙÑÅä?b5“ÿÔIÆÄØÂÅæ?K˜üƒÏ?6¶³µöø»þ“Ö¿÷¤Ó’—Ò£úoCü[-bkdglakPrþ[Gãÿ þ­–7°øO7èÿw¹ó ÿËË8;Z¸´èiÿÕ.ôÿzþ‡Òù_+AA;w/Vz Ó?/&vV;»Ïÿ ×ÈÅÑÑÄÖùßø'—ÿáÿÝX&&î&F0kËvF\Á–©-¡å¾"…³àT³Óú*TË À+SÝ7ZȃæÆz‡WØ÷ÊØynp¸3†<»—;÷Í„q&pÇ“Ñá:ËدÅDf¾5[ï ùoõáº×`ªX6F²`7wÎéëkò”“Ì™9ç5’¥<Èb Ñúeí}àHÛñ« ¶b:d(šßp&ûTÕH³æ!׆|¯{k ayA»p7$Xzæåö‡ašùõW–BZòs â¹ZXXÂ[¡¤Û®_±ƒ›÷WL{i›nˇÚOytÃó‘à7Ür›Ûþ|2¸r·œY‹ŸR•Ü*‘q[O<Ì>‡»3®/¾ÓjjÚd‘þÎÓd:Û’ó"„0¬ü°‡^#6{œ9ëRKÎÒü^ãc«j(”¸ådz z›è„•Q#“æ¯Ï Þ°x¬wHÓ œ7Œe/X @mÁ6ŒslOअNC¡GOÿþ•L„+gzzħ5 Ô¥‡µ3çK@X¸1™vÛÆãèhvbÙÿC&6*dÁ”Ñ;à%ÀÂ'´Âº®lÚFBwáC³Š,~!OUžá~Vq8\õ˜ ÀuMWMÝ(V÷À"-íË«LâXUÆ¿ÍmÜçV+*ŽïNPƒœ8…?Ð5çˆi?¾ûîz™€æë]ž½P0¶¢·Ì ¾3Ræ´æÓwêKÀü‚O†5;_˜µ–ùõä¿(}4X¥-B$dÕç”D&ûêæ&ñˆ¼m*Êb1}µÒ$ì¼½ñÄdÄbá‘Áî 4Ôl4ˆ›È§Oaˆ‡-}G` ],y€Ð@x,µ–`ÛÇѲnë`Þ=…¬MŒ‡ªO&õGAIRé÷õ>«†ømÇÑ:a‚º2‹„¹£’©öpž ¥foꊈ!¥¸?eŸCäv—žÊÂQóõìHe0“«Ñ²z7Ô<(Î6Äy0»à˜¬”7³|‘‘tæÞY/ž|3€ú?¹á‚"&6”©£T3h GÕ…綊<¨ËMH¯×ú¯M3eæQsÉm"(èyã?ã‰ìÂÂCÉgÀl2}hþ]{£Dwg?Ë'0†)¯*±ƒôn™>Òœ`LÑh2k‚ò¦Wcõãïþ6˜t¡¬ Àqð[;Œ#—S·ôeu`# ¡ºÆ1äœC×ã´Þ ‚¯ øº7x…q_Ä{>/2¢ûÖ©þóÂêÀ·d¬dºªxQZ÷ɽ¯–:‰ ®ò;Õ’®¿—W/i/÷ûêÓ(|]ü§ª}ô¡3„¹À ` î«rí b XûÌëîZRîñx!£ëÄ"å-Ë3C±ê ÑF~¿Ï3ˆÎžÜd0iO¤‘[7Ã\3³á ÃurN‹éõP(¾‡BÏ 4ê@EW–­Ü¤Lð‚êÇ£î”e¢M|§¼†¨3Ÿ*{˜zFˆæ:éÏŸžl tи»~cíiYÎU"‘çp^?¦§›åLö·"H  飘Ìl‡Þب\ŠŽl£8Hš©Õò›Ûö¥¨z¢…P?GO€ýíýSKzÖ€4˧¨ Ý`²_rÅÐð§ì3ð愤ôŽ²^û®XJ´Á*B ìQ ÜqI¿«àìjÈ|ž‡Èû1Ÿ ÍGJ¾2¾q¬Um Në¨ûýwÚtô™ZZŸÖT[ÂQ²žpEÕ MHúƒ°ožKÂØC¨> W‚ÒAvsÑL«A¥Ð"ŠÕ=§€Š‡Ÿz‹# èÂ"³ÎQ²9²³áV‹ “ãæIá?$ÂVG\óðIâ2ÑÚ´àð¿Ô¦È­qF¡jb‚§ÈPÏե؄m¢mW;±rI¯ž‚`1ý ª0‡-•†ì?ÇHGœ>8g›Ã˜Eƈ—<*ÿ"í²º¤}zï…ꃼñ8CÅ—ÓMAÂýüá÷÷ꬣg†TA÷„¾èÉʶ;3BÔ!¦H’Ëx¶µäÔ½çhf›±?äf<ú9j)üûvÖ³ÝùSý{µº¢Y“G;àý‚]F“S?ã*—Œ…‹:¢¾NR(¿ú5‰UÕs‘9˜,¨ëT±xJ\ôÄÐ[VanäSÚÄÈ‹t}åj}áÑlÌ|!?H–áRhðE&ÏH³‰J½[ø\B§»eR1°ÍÝZƒy¦*ßs:ËäH™}V ²³ ,<‡Öw5µ¢Š‘lhlÀm}=½çËd¿XNµ>¿;½¾-„3 ˜=±3Ðt¾DG@Ô$C\ =!Z?—Þ*ÆÐi‡«ÕŸX¶t¤`ÓäGf$Wsù'¾<G‰B”±$4Á;⇥y¢ü¼O¤-a5\ldƒõk”¹&Ì¡p%Æ<öV´-îWB¿ËNoÝNñ˜‰Œ{»’SO$!×Dn­åŸ]º„RNá=æpáî­ò(ÿÌNôñzµ€CYeOù;ãÞÞˆmö µ…þ9tT<'p«®Öƒó©fòͽ5 ÕœåÃ??¶æd|.›rƒ¤*£åYÙr¶%«¥é®ë»Têž»Ÿ"×Zֳдów}&)§ÙËŸU‡¶}I²?*ëË_YÆyŒð%“r¸cÓ×okÒ“ â´ñ;ëÆ´ ±XWßÞFˆ'ü‰Ø4X¸ž±§cmI.n8ÜÊd Ì»Me]ñ€‰7é² ‘›7Zš&¿&ôL§*S8l/È8(Å"܃K¶nØ4Šôñ“qÑ©óåìÛA =‰7gô~›ý)ýnÃökd t/´¨Ñ¡a{!sèœaRíñ0å¨[. ttzvÆz w;ç\$±?Èt±qËñ/åýp1Gèsþ¨?¥ÙŒtæúdS5ë3+iðO¹Ué2ã ’ ­>‹“„GèC£ïñø´ùLE‡N¥ ¨®A ¯¬9›ÛtÝ8‚ j¦’Êq?”ï½î¼…õ<3/!m"‹AþõÝÜ^2ª­“ÎÄî|)Ð5”æA¤¶Z Üï¡rs0ôgPôØT’¥ct”á%‡X`÷`¶\ï½WxÎ1z~Ø’~.Ô$vk!ÔãËI}_ÉÅÚ:-+ö¶ˆv¬²é/ëÛ&ÓT0 OGå>*¯µeRQ”œZ-8þz4;¼„˜+M–Üͼ§ÏSí<ÈPI3â\—ã™ÇÌ¡º×·Àrµ³S2§u Ç3Õ©­Pz§ZVKé Ü$j»íÔûwQvª¸”R¦× ©"‡aƒ™‚ì9ˆ[`º)> ÒÅ5TïEó±xåüô©;¥J;èã…íÏ8é±M…I땸㓠Qí"&\Vêaâ«yó÷ÂAaau’ojÀ/é¦9c×Æ’ Oæ¯6 ·ùSƒ<>†ô=ðª.û†BääGÖ¨kæ ,X]HEî¾juÛûd$õeÎî× ©‘?!pçXQÞäÓO5×l2,9•FL8±G®Í ש Âp ^]ºÞIç¥å&‡ãçó©³Y_&óª;d”†z‘ü]Qƒ”þÂl·1øÍHj`LA*~X0ñÇ•~Ö“Á«8ôä¼±yëÙaµ¡VH0û$ûÅYÜŸ_µ¹bK3ï‰þSr^ÂÞ3†qR»©fê?žc쪛и¿÷™O|—†àÉÚV›&g,E¡ „VªlÝs¹!iÕr.è—mšèiï¿$\ú4“BoZ;ÿ(ßï»CÂâ—­$Õo30s®u˜äÝ Àh}¤ «XÛ&X܉ϕ9 UºaÉ!”7ñ‡µ¸À øôåf~:#IVÏx8¯8Yâé÷S¹"RÉþÑ/‹ ËK–‹úFë`"ß“µ('z¿) \ìÈßYÍgã2ð…@†?ÂN¼Â3fœí~à=Ô‘v g¾Ýö‡…ä!jK‡2sDžv÷”†f ºm7é¬ ‹ÇŠ¨Å'ofüižÞ.RøÉ> 12]²c8$)«µÊãr9ÇM Õ¢j†a‰ôiÐujþ17Å–åaŠþ5Äßÿ¸êðk…QãƒÍçO(éz“ûµÍ ~7¨5wܼ`Ôx om#À[`Ó€¬´Ê棧Bñ­†Ü_Ý”I]„ƒâ§ý8A%‰Ý)†`¾£m‡¦yŒÚ [ÛÇJf(™þOfol[¥º/n|ÆðJI6¹HÓ̘«*fAEt0„¸ÈÆékõDì:³b»‡=P›˜ )BS3EÁH Mk8Þ¥ZÆZ-öú(î;JY íú‡ÏÏB{‹ŽK¤÷ÐWoßö’'fÈܳ'I]oªuD:o¬¤’žB7/v³Ô'ÁB;÷Öªa=vSdøà ÆWOD–w+ÂÔ6€äi ÙÖïÅ¥¿š†æ‰u%þŠåö[ô·”XtêÓÕ¬›õ3!ffH}V…ôšà}Õxº£Ü0]™ñg£6ƒew6á7|‚ÔaS¹ÏÂrs¢X«£p’2Œy^ñ}ÆjF]Êó¹Kú÷¨ízŒKyàß1u·ÄàýN[ÆG §\+aðÅ“D¹®ŒÒ9ÁÍ£TŸ@ûÑÅ‘#h3SRn” {ņU€—£·`tÇÉÀª†}_»ÁY6-kÈ2 ,Ê›8áä_]žäO¦Á+Á§œ}pÂAûÕ¶äkÉÖ„¾ž>­q^èbI9G(®.ïëÈ9]I_áÒÖæ±!?ƒ¹+%ÉRêl¡¸+WuÓ ‚%ÉMåO•¨°:9TF†7cŒ7­øиZ̲>ŒD«‘RAÕ–##Úøm2 Sîµ|WÄJ'mUŸ ¡däØ<î‚æw•×r»õ½º…¶= I‘Ý®,Ú¼•^v"[è–q}`À¨Ž™&mz= ¶S<ðDQ»Î4:Ù„†nxõÜÕëÙ[§37\{>1.OU ÏQÖ™c:ÐSVاD&LÕGM0'8f2“â<ûaBÔ춒ò‡˜M¦@lµV³3“íY¶Ë‹¶ÚÇ…èjñÆfêRN6œ '‹Îgô ˆðð–ší'®ë p6Ê8„iIó¯ i­[r€Ú;!MyIDøº;ࣦB»N·nÕ¡ •þÚ,™”¸(¥¡4º¹¿+6‰¨ä÷í¼p(æ0_@ÓR uyæ¬IongÖhË=âÙªÇè4°½^Ã¥÷,2þ˨¦ÇêM³€ÃÈ3` ßÎÑ‚c»Ürûtå#äyp'-ý†)3Ý¡¨üæMï›ÿ}Š/?z"hu|ˆ#]©K›pÖL§#ñÆkUö•Å"¶]÷øZXœò±ý¨k“¬§,)ªJét ‘J&l\×-Wµ]n>Ra¦4þ]aºt!9L—¸^Ž¬Žqð!5|‘×ø—C¢~2¶dG‡°ÝÄxˆ6î÷õMØÅW³I¨5ô;ˆ¢oP,¼iüÝ:p¹±F¥R¼¦µj>ºŠ0c;†•”ΙëòÔìFEÄRÉy¿Vç€QȦ4H€~²Í”ç-··q/Íð U¾WôM™ Ôj´áÅOæ<'²/Ÿ~>N ™i®KLt87sóq³7Â]lCäÍžhe6V¦^‘[ ÏéQÞ•3u?Ù—q•±P{ƒ0ºÆC ëÎ|£±Ð_©kÕ ƒŸPØï$K íÛã¾sÉŒK†=M×ZYc²¥ï¯§:Þ$vA×¢ý¿{|é1#";2@ÌdL8'¦W«Ý¹H@€/šÔÒS7²Žèé›y"r‰÷z ]ÄcÏXu73É)›âÔýR†¢Ç_ð»?»2ÔšUãä֔€TÜ°› ²W•q£¿„…ðs'®ºßKmÖo¬…¢%‰csçR®Ï×C®&M†,ÕìÓRvs7¢1œ‡åhäÑINLñ¸Ó#øç;\¤ÑkÛ>ÿœC‘¡lZ;üðÕ'sm…Âó3ú‰¨ ”M›Î!jèÈuºê;/%€Ý¾N׌õ"ÖÔÿBðL³4ÍL0‚rZ§¹àØѺä‡o:è§E70Í)H’t¸\ºßuÂ9| Ð8TÇŽºWZ9ç 5lf^øÀÖ'×:âËÙ)Ä7áµh (5nœìq3Ða)áÛB)>·éÿñó]¥ÅÉë B¦Úý–‡±^ÆÊ ÑjEØx ¶yYvÿ™ßI‚‘:–À)ã#€£øn½Ùmá†ß6`¸‹X»í´ówÀbŠ|‡ï}aàÜ( µÖ%ºnLÓŠ«'~kY n¤ƒèCêèÓ;IT7¬ï>Xú[&|0ÁÎ^uSNG^™³å¬¨5"=£eK‚ PÆèqÓ=Á!¢ë.Õ“‚S§W 6µiN°¿*ŠjÊ€ýð„é/(xéÌ7*JÍÇ ÎÌo¼¢TYÐDhôš”Ž"ÿaGR8ÿØ¢‚==@/§ÜŽÏ°µ,ÇTVAc8q…“íOJhþD„Д6Þ­@SÕ ,KàåGÐ/ÂEHå½=y­v Ãf2b£+´ve‚}ßFó4zLFÜø¢ALoóSÑuÔŠu(í­$šäõ-äЦ*Ú'†l„˜Aš kƒÝ†Vi•N!ø™@Ì:´–G”“êeÒX¨”FÑÏÍ Pk™‰Ö>Q:/^=š"=V‰`•ï%Új_¸ŽPì¼É|úøÚ'Ǒɯl¶D<«v‹2¶$ĈO MèÀ[ƒNùUØòùÂ+³› tý’Û‰&ÏúK1®\{5¡¤åTZiÆû1ìÁá c~&°êµƒÖÂ’Xè‡}ËÒØS <ýM§qˆú„'ñ)¦îäè'ÒJJ`v­Óî+Ó£ ƒþ‚z¹ÿÅøo ‚o¥©9ÍV iFd¶ÔÉa ®ƒì+OÇa; H+ë¢;[V³kþVR¬.ˆr1.±É»m7e±›Ø²`%"[&B‚tËçë%28ëÉn]·g¼Áô‹¯æ{2›Ñ†‡bÞGœC¨8^r,Þ,ó r œ?¶såUÑ8ÒÓží‡ÇðA8ñŽ ”¬ìI'œ P)û!L4cÄ-´Ô€mãÿûMˆ‘+ùô{b„æôkLÎe;ï¢ɨŽh^'ìwüoCÞl¿ %hÎ,ÉÊš‹èÅ–('VU3Uàž"ZX‰H'iúÒ«B•ßÞ¬\5¼ÝÌa¯¶ÿ°‹T¬¤P?$K2î]1º£ÍW ÆpÈavYXÑé)ܲ”ð,#‘&k×ë4· µ»TU¥%ªówz!ЮÀ¢ü@ž“!#b˱ývþ9 5c‘Y’¥P(Á\ö>%K¹ˆP*¢¡ à•.’èÕ`Ú†5³Ï‡,Ìà Ic¨H¯jý6-Í!¸EOÇœÂ,’vB½g‹¨8°«¯Ë{×Ò¸×çµb4™;•¯˜TÊЭi×þçЋão4BGÍ$’ÖóJB<¥½ú@—mÑãK%R25ómA¬ý}<{VíÎgXpkã¢Í0 [ò•á•7k³Ð´¦ …Öº89Ä‘Ó¨÷ Ád¨c.å{ðÖ F½Ö=Ü*ÑÓʘÓY3Ûä$—¶au†/_Žj"µÓ…ÿ®R­ñö¢Ý«d×FvŸ¢-]vD|çù«Îûߦ§v°¦g9Ég±)\ö-j„kGŒØó×ö?ô(µ_…åt°SÂ|Cý¦­½ÿÙÞ¥ÓŒh%ÕC£šKK9Nõ®ÝHŒÁþ–¼vlà´G¤ïí„Ø€.ÊùÝÕ¿bM®ŠQ·È<.çU ƒ›‹”ÚUwŽú æÉ ^bþ“¥–L„5¹Ië[x âOP‚§õs%Q$ŠÌ𞃙#Ó™"/g*g ‚ 8ù•R Öùr–›ôŠ¤Û˜xÍñ€€ušñ‚œxS4°Oæ·sÈžŽ`zS£Êlr—¸·È¨Øyù^‰ÙíN•¥½ú¬Žˆõp]÷ƒÅ âJ}ÉÓí]EÝ%Ú›\ÀT£’zãáÉŽu½E„X±E{+å„C+8˜#³Ì 7&}Fšp=‹ó%áãÖöM¢hb¼èÈW4¹À·^£Q“¡hÙר—ð¨i– ê Kªa»'loü¡[’I=Š’C€d×à™šJ$þ¹"k¼é×Ò?Ä`‹ôÕýeÜ|çÀΈ=jôy¯7cÊsÓ«Ù‹2Ó•• GÜ‘˜G–M—1È‚;tÿž¢¼¹>©e‰tÃôš|'IþÔÎÛ{'AVö÷‚ÌêÇp× ÅÁ÷8½cám6$ÓIjqw-“”g6PéþX¡r°×&–^·†n1Dµ˜™v´;”:«fÔûˆ.O âOëe—ÓÇú«Wg?Ð ÌÅ’ÙÃ$Wb VGŒè±±¨"·à|»7rµýÏD·©[÷ulβ6ü7’0hÓ£DTÊ}1ÒdE«[!ÊTàOÆ7N.A­‹¿üZà…gÑæ4ajẨ¬ªPW€Á™úc¹†HzNÀ¢‹Q_÷ ¦6çŠÊ‰ø GiÖíP@FFôæãTxà’º„rþœuCÁ6cGüÝH«o°òL2¹Ébtüر¹à…«¦ä]&,!9C_Lu#­7tÓ.üáäƒ`d£‡ƒ¾5$Ñktå!BäîˆÏ–d÷A´¤ÖTuxìb{¹œ†êHÊ<³sé²Zñ]â½±ÕËíKVnO¡Kó/¿?Q×>{[oó” Ëci^‹E~ÚŠd¨ -qnDLC¬¦ºÍ-‹Cøô¹bXÎÆî±e¢2¿ZèËøS(ºn#þ²úoÛÆuíÚòžsÓæ;Vt°€0[šÓ MÁ‚½´h=}•­ÑAëÙŸ˜|b„––ÆøöŒn`Ýи9x s9èûÔ‘:m̱Ӿ\î,ɸV`d5£¯P…ð_è«2Èš€–úÙu&HÚ=ú¾v¼ÑÊ0«ÂA_³ ä­~w†0Gh"Åq×l'¶44!I¤àa¦|9ò9Ë‚"áÜ»SA¡{4úmò™½U%¸>ù¿{àa콟ÃÝÖ!°XøAþ¾Ò¦×+—YñY/Ý;F0Dósa܃H+ Éɘcåm¸Z[;ìý@‰¼¥¶ðÛ,A8Ii¡s9[r¤Tq“h©¿Ž³ü×4éç¾@ap€‰ÎRÓ6xIŸ0dFÄÜFáÞ—Rd–gXH5ÈÚ¡Y»@ÓbI¿ÖÝø/:?#‰Ç'š#ú_àN·l9Ø¡‹B;'ƒGD%*â¦*· œqfRŒÊzÎLŬÊãdâ§=ª<7îÂáe³Ç0HªqeYM%íâ’@9/l©ªl€ñHÖ D=a!ƒ^„8pTü•ÿ&çfï6½®çbA8þ ½öÄĵqZ;ö'Nÿf¢LQu"´þÃß{s©Åš 2QÁ]SeˆÁ+ÕpáA›/32äÜ)%šÓ>ý:!kî0JSpﺆýùe2úªx#»O*Ju)•˜p´¬à-Â|â¨LÏà£W!ჹ¿‘Ïò^‡ËP+$B/´4ÍÕqîÌe¢!Œ¨K.! çãÜP; õ¢ã(08 o±‡¦ˆ°ý#̉´–„=:…~R¼—.D Á»$FœEî7>lÅEþ qhpuÐ’þd@lcÍG¨,STçòÒ©¶É*(…qþ)ýRÏäCKL—a|õÖÒðg|Ó¬Ž†·í`A©¦#àþ\Ç%“ªúFÐé‘×+gG…ê(”¤úrJH|ø«uXÊô„8×~#{ö# )¥eέÕŒ•j=ø?HsÁÑîv §Þ~Vò¹¤Êz~Na}$•²UŠ*‘Lãå…Éèh-Ø’<Å°˜@òûìç—“îrèT‰ØÂcîœñ_g;f?cñdÁdPí½-t ¸ú;IÇÖ¥Žà hÁŠyS/#¡VµåXÙŠ—¤P¥á‘@Üä;æÕû.ÔL'0®d¨íÑéUðä·åÛfAÕFæÌÑJ•Ðj aŽ[SÁȺ#C*“Ú_Ûx­¥‚Î÷?ªw‘Ru]æO‰º~ÿÔqŸ=eˆ~4ú ƨÌq©Ó¶ƒ„÷ à ݘg*¤ÄŸÍ|ã•Ü^ùI]¬2o@åÇZavH¹±ÁGL5¦QFÔ·5 Œ «PñjâƒoE¹%‰†ÀJº.Œ‰ðMu.{Ö˜ž #oA8¿2æ*1÷êå°™Ýüëc܉¶ˆ*6½§/W¯—tµò¶¾ÊëHh/&e݇õ÷£÷M³:ÑÁÅYwÞ+&¦;‡þ²À+ÅPØ@ÇòMÃ%È}:ÔÀ—kÞ¦/”s‘–03Kæ¡èدÉ÷¡çfjךy vÙþëVãÛ¨­ÅÕ.”qß»€€ÔhÈ3%G•LºT¼ìU%4Ãç¤ôøèX†(Û.×loÂm'«@øN\o;L«àܯŽ•ÁE?ÍK%š¡z­ˆ›á£'‡bþý[KÚö.«¾Ýé·ÇçÉ(G¨®@:ç—åCÕX1[7a§÷ ^ÊBh™fÉò’ÁU[@/¨*‹‹ B×ãªÍÀZ¥;ºÙŸQ­ wÅo·©H¼5=¨7c¹ßžâ²å;ŒÅÂԂʆ ÏÁ/ꟕç<ªN”g»°ìr´1¹S£fÈfiÅ9. ä!T&Æ<¥{·u5 qˆ³ (¹k ÁÕî~¹ëäüË™ÈãeÊ1þ ……e^Uá!òë¶qºe¦Þß—³Dt Ä=lÓ^¥×)+%3 w{‚šÃå4_+märÔô¹1C½%jžaå9˜*ApÄJ)—A%8ÚNd–!sƘ´u8Jøô‡Ânñ`ÒìçäWÅa$ú©‰Ö⼋¾+ú½ ÅÃQñ ×>ìþÓt±A_… é¹Ø¼=#!.¹Lw ËvUókV°Ç}B#êÃI.DÍ»cöHnÅBÛ**Ì'ŠµÛ]¥JPq¯–š}£v¤­*n Ë‘ÒFFz™Z{š»çÁõº–qø+Vêš(t?—è+OGx׉"ö,à½÷ûÅIž¥Áaõ kd¢üIs-qðJ¡Å¼Lho2qZ÷¾í€PözGjÝÁÈT²z\|½ŽþºÂÙ¤0Â-_PôP §·/¼¾þ`8®G \>£¿ÎñL¡‰{§Ó1áýí··;Z¶ÔlÖRöwp2ůnþ¾wH6©k3û5˜–}¨ž0Æm܆=®y¥:z$çw¸)|Ù æÙÏ5šg:MÝî~çA¾‹À÷&V`“ìBvÂåKC¤ 7»)ð3ó0½_b‰-ßÃØÓŽcM꺬þ¹gsjþ0ð&R1ì@‹ëÉ”Â:Ånñ‚Ÿq»ØÉÃé—h»,ñfØ“r š+ŠÛ¤D!·1shÑP:¾¯ĉë @„Fѧ¹Ýj5—÷)˜ÉhÄùü¶oSZ*[`ä öäOÃ¥·>ÚbOð´\•HF´'Š³(š¶1q é ”À¡Õr êÝß7Ë’9ÜÉÄ2éÇKåå ^n8§µø¨ÈsÆ-V×R~ï•.Ï™+©GþÏEˆóqFYif(¹¢™Ò'÷ÇOù ä ¤ËÀò‰@ÏtQ+‡mšècÖ¡üDÒo\x*Óû<¤‹Ä%¢ËBÚRjË}!?N‰í2ÂU¸í Œ<ð)ª¸ø÷=÷útäl¨ÔÐ %›?Œ‚ÜŸ¹NIƒúéêúmwÎ$ÙK8—0¬0‰y„ˆcìPÕK%Æz<°¦CtͶ“¦íÌ$B/Éå‹®5p@¿ÞC/ý êi‰F¡"¾´§å¦>ec¿R™‚Kæ0ß³?M³J0AÁ·¨ñ-¥%”}•+€kÎÞ}çœTžë³Xmîž ,)f¢1µÁ1Kç—„q‹÷Ý\þ¬Ÿ>³r3¯zpR¼‹†N³­Kx›(’çE#8u°Ó—qwHˉ})é·N‰ šQ@Ú=¤Š|27¯œ{Ö.Š÷t0øºw~˜ˆ‹ ê§ž’£qï,Djt×Å @æ3›¾}oËï(?÷f‹JaÜEÂÙÙ_PÑÚf0ye*P#Â7UÐMt •XHØ6ÉxùõÄ8¶Í8‹F›¼óféÖà#¦Ã5âj²ŒY:JÔDº•]Ë ^Â`ø·r6=&9£=J)œxdÀå¶Ï•C6+JÌ92%1JfIÕ˜{e?Ó<6û§í*IO´9ˆHÅœ’—‹ðð‚“%¢ÞO1Št~È¿*©oL½#—î_q(ÐmyÿÔ…ýäpQkD—¾Ï°1¢ËÚ-„1‰Ì¿‰ìÇfvÅ…F‚b¤v:†uJ@eZ¸1®Aýz[m,ö=C\Ýú™œÑ…o¯#²WþrrÌtfR¨@ÁÚÇc¼fl?ð› 0 `Œ–s¤ÿê]òG#Kçš5Óøئ÷H®l|Js3òðæ+ÖsWÖÉQÌ{M¸Ø¸½Ä;­Žò°6‡²°˜ ºÈ+Ãå·@zÛŸ7¢V25.¶‹‘:4dôŒ‰œ¨^úÁ(Ä>²Æ„Áì“tšF õn¼H3OŸéÏ%úÿ{¬/ŸÝÇ›E´ÀQëUµ-ì~§(¢»À g‡+³£Àà7»wFÕBÚ¹§pèù,~Fšz§çéé½âen ÌMñß3ÜýuİDzO‚0 kªÅaʉmÙÆE Ÿµ ÄŒ> œ¢Ÿv_ ¶òÈΈDã)õõì6ðuâ|UŠ¬´ÚÆo2_ÕÐFУ–æñÆN•Š¦g=—Õék­ËÕ^˜efxGmd£re1Sþò¼­¾VDEgØÙ+§uîd°§ÁâìÖ®õϲƒ6WÙ’IjÕ|&ëe±Ù@Ǩº>k2J m>¬DîuoÝtÕ*¡Ÿ*¾a¹‰S6Qxã[n(|©â©³aÎXc¦ÐYg[_¸Üú,‚V A`àü‘z¾ö¬ÙŒgk!yöcˆ'‰Œaƒù`7¹©þ%¡®ÊÇøþžpµãXv¯Šå¹z/\HÐÿ¨òè`KS§ƒ^FpWN Ó11÷ÇÂÁq­ÕšT÷7ΰ—OâÞÁVþyF>”s¶’M:¸/#̾] ³y«£wkZÓc ²€%,[1cœ˜òÉÁ©vE[ӯʵc¢´|‡ÅMžd­¦D“¯ãÉòó‚LùbE*69Í—c[ËNUÞÐn™J7èšïMàÝFœ/"Ôʯ]öݷ¬/‘Ið÷ êš­?²LÜoõÀ\q'¿ ¥Tä++ö¤PØ„$зd;…8Ü{Æ-ˆŽb?5¾÷.¡¯S¨¯¬oˆ Õ Ë'« Øï-W•^gPO§ (ãâÏmä ½ »ä%ë°¹”SY8tžÉ±l5iñÝ_ÖÆdì¿J{ORZAˆîmØŽAcR­—NFûXpA”DÐn¯\u!pÖ!²ðøXj{Ig…)tÕ9æçD),Kj+µ¡ä2š­üð°ð µ)üU·åzŸ»Ébdpö4xŒ¶_\*Œ˜D9Yà0{;Ó’º#ØÚ¾«®ì½›âQ¬…»À¤i ã`¥V'~i‚ëjK%&ea?u['1Y‡J Úµôç†JÁþ1ò€ pÜÖRLaGÒo•-3rÉ[¸ó<›æXIœ—­Š[e0)·Xãêè§.ëärÄGI‡ôq>ѲÙ×ö.ec†¤‚tÿkméIõ»L—vG}4–zÀ';,ˆ U,eåº8[Èdzú2ÇÕ€¸¢Éôg‰^¯pV¿D¤=ü‚¢`BMNªBN¶ii”­85æ’MÀS~Eíåµ<­3*9(×ë¬zñ™QÔÆ:¾nËcS1¬¯Í/ÈptpÝao‘‰ë1ítßÐ{ðý8vŸDÈ•›þÙ™à;å• FÆʘ ÅI… g„ÿBÍý Ëg€–Æ@ZöUÿ`0,o6RìSXIoü&¹È†€zvª()Ô vóXŒÖß´Ñryì0wâåD —”8Ÿï1-S[[F kCÀ¥•ƒ÷¨=Û‡ïdåBcE¥1ðD¤Iø¸ÈÉÃLr7®Ú7 ñåT9 BëÝJæ¸îq¨4·£…*¡'?¹Ü#ñ;Œ°öý/.eþe–¾atúß9/åõ8ùŸˆ­ôžÓ¿R^\FÛ™ëÊ3Ôó÷ns¥V[îø’–æG6qŸDæ¿©*óâ4€‡6üZsÇRA¸…ˈýª/ÈöÎ ÔfƒSÁ7D‡ïd()Õ1ãŒçþÄ°î9Þ-ŠÄýu¹BIÛÈ”Yºã éy‘ÅtdÓ™@Æ!#ºXbK ÂÁèJ[„¤|I:ž¹ã¡õŠÐíK¼<Õ•«=ëÍA‰ed°Aƒ»N¸0Øhþ:sÂœ¹ô›’p“ÈÔ…I1ERK´G\•±’!ž,¼êõîãB…ˆ(gE¬:ñŒ§Ä@ãÚãÅúbCIOF€uÙ-Õçç]j4÷ªOõÚÒ+¢|×ê••â»XÆShªàˆu¶r¨xYÌÕF¸ß §n­¢PJÌ“-`îïÞþBˆSpÛ£ûÑ–XuÊzs–á ÞF#žDœ¢Õâ„'›2Ý 4 µ‘,ÏjÎì.l'bÑVB5KÄæ‘ÖHŸ¾úJ÷Ù @˜Ú¤'”«Òuû’%ŽØ!¿ï^±û£œ}21‡ÝÉäm‘ ¬VÝ@¢„Ć5çn ­wÎùÉ£ˆ‘¬•§WÂs»ØÃœá¸I¿‚©Š²û¡¦è.¦†Ú™§nR§ö¢~¼;<‹Aðbjõ½M;’>Á ?ÈNpxÿpêD±®é±Èw[ñÜ›œ½³>S1t cÕœ¨¼ çõ.)Ñ l»Ì5F%U1——·ý €Üioz²^yºEÙ—‘`–½U³_Õà\®8ÚÇ—ÉfæÏSò_„B˳ýŠA!cM(›Ð£V¹ýdQQ«k Å#¦ð3€+¶3¶qBͺ‹VÀS^¡ÈQ)P€Ý)¢±K„ªæ’Y„xeáPÐX*Áþ½dbŸ6Ý(ÝéÑý¿¿Œu„©…F™å¯{_¼ø túXðÎ%/ûÕ{ä;N ”ÉÁ€Ìu9”¸ì[_žjßô.ïÜÆ«àSŠ}HÕáZ RKÙþ-ÀiÞå´Ñf©yed))¿À=*¢A‘2ßèw¸n²"ðÔêÌšu¹WŠIØTœo7âe^ŒäH°£®yáÞGìÌû×¥›JšÜY>í%ƒk~ÈSœ"y¨ 8Z±=yUÅS¯*¾-°~øËì• ¨êƒ„19¹é£4\ˆí cl¡;Úžuk _(þ©ƒÊfKfR8–Kžÿoœ?ua$«-r£',Êri„|‹*UÊì÷ÊXç¦Ût-ÿë1¢Òøß*-ö ùÔSŽŒpþè"l7ée22¬®³¤ñéÈO0i¨[ (/|5’BÜH²Ü²Ð—-ƒƒ$èPö¾(`±®f‹‚ÁüÃpààb¸2¯òãÞÍ]õ×Xȼl$Ä: ¯•–Óç²%j?:î!3Aþß®rí¦Ý_ 9§i%ñòtb°sod‰ØÈš|WØPÌÅ™4KÅO,SÍÈ»ö}зXÖZí›Dð?1¯aŠ1æí9O&à~bÉHt}` “eý^è:à Tˆê³ €8û¥xÊÇfŠ˜ö<¦y[\=Ä[¿Ê¦þÁø¬LŒë\r„K&Gge^‰ºQßÒ(»4ç£ÍšøŒkü¡D¡}+RUKfOÌ¿$K-WWrcÚHŽmDË›–Ö¾©Ø+‡ÕÅeäùm™oå€èí°$Ãs{(ʃ^ôî ×joŸ.¶u'˵2=™/?îèï=úÁºÂéj»‡ÕY½Éßv„\ª˜Ì·¾G\+ÙãÙñ¶o3ö¬ /9¼åÚõëÁ¦©ôlÿá+)ÁÂ<=³ý=Q= q¾Ýf`ß),äMët, óqz5õ¾õþ°”*EÓ‡’&M;[ˆô‚L[Sunèw&™ÁJ“”ËÉT sÿA¶•Ø·ª;]Ž`~Ò3À%qür‹¯Óý³Zq¨Pg«S¢ ÂM¤;çiEñ’YÔAzP’ÖGRc=ø°_¥ÉŒÞypf²ü]x=‰tØúÃ%¾ýx0Ìê½S\Ý™ÆMŒ5•£­ÑY€P ïZ^xM’èÎâ±€²°òMT—-˜á:¤ 1ÖÖqÔÆë¯hâ7äN×À¯&mž*ü{°<Û‚e­ö7`iV” á¤^c¤nV¼õ|boUb”ϺD¥‡Øòñr¸Pw WÔÜ)‰*n€9]pCúè*‘Tµ÷–œ„mçŒö\FøìIÅmb3k”ˆW:X©Þ†Æ^>5ˆ¡ Ùr€ŸGæWÔžÜä×1”T.ˆ±¼LédVŽ$är*;Ê$HL•Ú¥ÛHýø«˜š° Yãéí!ÏTàµß@õÓËÿt<]#¶¾Ó³[PVõàªzŒ‹¡vÊyN ™h?Ù E0`è`°œ2(ª5ö=6:ëa@xF‹Û™ ¿‚Ù{E˜RÇÁæj5 ÚùÀéxˆ7èž^¡7„F¸'¹Ÿóˆ¥™`…|*_M0ÝoÓ†÷…IN¹ƒ¨xVSq¿†ßf#òad˜;Ôä)ð–‡RÈéJh P|6ÒÍã™HU6iÆ:wä±à p(Õ`‰žèF^ªXÌ6µ®}åk¿ÙQ™¹Â»½U#„?³˜Ð žþí©ˆÌJQš¿T ³¸J!,Éøg jÞ¤Ïrpºª/Ú ^zæTNÌÚ“¼¥·{áv²+ÛåE›šËNz$7yóÑ@See–ˆç$RÏ,ØŸ{#.EíÛŒ=l#Ôk–Ÿ7ºí°3y”6²°—Ãqš­ž‡qÄ!lR4æÎA.çmߦDä˜á›ÆìºÝš„ê.Uª ÄáFG•ÊùÛSn@ÕûàrÆbŽU¯úN‰aâ(Ö'¾ÑWo4”˜ý3×â·Ê;éW•ƒV2ÒjÜ"ÇA¶Ž¨1Ë`ß%u±*Y5JߘûŽ^KµKÖ_¾Ú›D·, Y²þ(GO¼A»Ž<¯dÆ“N Þ\D­MÆù Œâ¬èƒ®s+ñÆ°ñàP u·p6Ígó%µm‡ª“-NÊóžÙÉ:³nˆ ñ#§‘,ÿsÇÂüé8©G‘ùf‡­ê£YãÍÊh:émÓì#Ѩ[8wâmF' ¶ÂÇéë0¿Ï ZÖôÞCD[9!fÖ̤odºCG6WÕ…ÀEéd…¥Ô.Î ¨¾#&ŽGÝù->oúî[í`¨ùN‘ˆ±@î’«~ ™H dAªƒ9Îé}«æ/8i(uc i”VSjòÈQ¹}K ÎZC¨£@R8}ñƒTPò~ŸgÈæ#ñg¤–©)|òçÑÍ;± &Á=lB€ò¹ ºÒ£¯LåOÉádÔâ—1“Ê:?¦l%Ë€¡ÐIŠò[1:Ù_Â0oÙå¥ÑFÄ?]Wƒ¿?ÎÂJRt8ˆœÇ°@†ìB©þÌpÑÈï‚ mîê²»¢AûhÙ=q Z¬Xsuˆ7u©óÛÑ‘ÂÓ½¤c½r܃X^{áÆ|Úž*m“Û Èû‹áL'$Ö-Ïkþ(V¼F¹¦ç<_yááT¬r–c‡”.„¤å×~eLš0;°ã ‘»’;:Á_¾Ó‘œSL“ây\ìœ-T.I9Uݽlx&%¿ú»ŸÕ ü`žð:t®cl…k€"o¾¢ð•tLV¶Üc⬻˜Jõ:…€æ>/ÎÛ"Çü®…>>b*9næ<¹lüºKõ¦•ô±ÃL‰ ¢¤¢°F¢±aÄ×¼¤e÷ôÏkœ :Ù§½{;\Qº—"Eÿ_n1‘΢*9¨:;ÞzÕ¶4¯¥U`F– ^Iê ÈliläÝ8/ÔhØäþ’òoãrÿ¬Q#dß›· Ù·ù÷ ­ÆjzÀ캄°‰HWïŴ߃Ä&Õé*?{ˆðÖ•OWå*Ïÿå¼?>'“õ&¥ý”ÆS£Nμö3¼j3ÓŠû[Dì »Ê¦>ƒ‹·ŽúŒ ŒœóÎ( "K‰²¨ {™E’{²f¹lø]“Ö_ ÷dîšg!žcq0*BVZ=[í>¤;3Q½“œt#Pšƒ7Qí‚ÁÔYˆW‘SK?ݱŸZŸÉÖAüõ^l·h”U_–úc$ŠFw…óEFíwö×l€¦ƒµ5™àÉñ2˜ž7h¼NgÈ0?ˆ“àî¸!¨íá²Ýg¦™ ­òY­!;@µÏæâKš—o•UärÅ{¦¨‡•K¼-C9ï•|ü°LŸé$V¡¨,üéï³Ë9{^8SL-¿Y+‚ɱÝ4ö½zC›·èX¯±ã‘“ShäÎ{DäYΨŒÎW+: ËŽbp€¬Îª?Ã× ì½®A²ÜP*B¼ñ\'k‚înÿ‹¥2˜M7-«VCz‡9æ}' §<„È°³¸ŸäVªR4 3#ÚÌÚ 9åËØzn쟄¤ÊóÆÆ'*õI·+&1äYhâ=½S·ÊžAÒ9&%=º§QeÃqv“@rŒî^w}® ~ ÖŶÌ,òþ>!®R@¾S± öùÄ fŸ ÁÁ«úT#i?•·cg „.žé«Œ¢•ðÕϹ¹§tÌÜk«“rϵVõ$Õ¦ÃÂpÈò̽`Ljš7]G-“)ÑEk7UÐÞ• ñÄé±{^S´*¾¡’^.£€çíóPuöÙUÏHJ›¹q.$×ðU¤VÑÂhËê¸*Lô³BûÁ­ü‡Òrê?ï9¸ºD)…]5™Hb䔘•ÖrÛ'ÎõÊM:¦;s{ Wåî‘+–OmŠX'˜.,¬$/°í˜J.›H ³Ä¶Ë(0³LZ³­ó'ƒŒ ÀO›ö®Äé]Ÿü6wºö|QÁ'>dï-`æʺ¡¥ž?ÖÃTtbOò#FY~K û9IäMŸIþ§­¨e±>ØÉ’&Œ;ó0Xîv³e¶ÊÍüþ£qQnú,Ä[*dyMŸ€…rFgü–t5þ-.pœÉÖU„ðæLëa×89Ä<5EÝâ¯Ö¸R1Ó¿é¦ÙøÅà߀Š qõmó0oA•Ó]ߌýág¤37‡¨íÆ´èc…ø´S ÀÓ§+‘JJùBŽ o™ÈIR¨‹øŽÈ¼ÁA[Hô'É‹†’erEy_M- WÕ‘C-ëϦ„‡±ô˜0Ln®øß¼²jTÁÃŒ¯bŒ¶KÇ»Úë?á Oò‹ïé t› æ&˜‚ “ 4PÒ暥ºA ‚Òƒ[Vw>zxÌÉvõÔݽæ^„Å$óêä!RñÛt®é øâB£½³[I¬ÿ7áSmê^µ¡ûÓ`¢{™ÇÁ~X•d’ÀÄW}¨ÑŠµÐÿÖ¡ee á§ÏÖ;’è^«P0uÙxoµ"æ—ý»n 5«Ë} Ml]ß[yäÓ+¸`6±¯ÿàøႇË,Ìijޚq†w»Õ$÷“:T[ƒÜ”ͯz¢•®«ŽäàôÑε™Ä}éÄ‹ eŒè'n6'¼þ¦r½ÐplM÷¨8(Õ4`Ï!µn›¶g$ê=šx³n_¡¸K êÛY1Jàó§§X…¥KV0uò¤Å£—GN™"éÞ)­Ó®›ŸDùP…ù V K½ áÙ5áÎvûü­º^’¬p#Ì°ø¥_4ŠªptœR ¢ýÒ…/Ðgªü‚êäcÄ4ØT\—Qƒžœ0e䘃w¥çŠ´5ò¿ta…²€æÑö$ 6ÿœ\B2§V¯uìüµ*ÕYZ¶Ú)”/ñSñÊ/<ŽSÕǪ ±€ÑÄJXzQ °þ²øù,Ø‚$T3óŽ›ïÅuX3'Ð’þ”;yà`I"PR쬈جñxè³pˆrr¯wkpáˆd]äÙtˆa·_N9Ö¶ôólw²'**¶í1Êh•ý}f×;ñ\ çvêÙT“üõ3¤ °¢6ü|þñC!Ùp L;ší¹w2tF¸8-ïa#ǘÿO$cò™Ãýv¬ ¸Pd *›ÖîlÓð¸iÉ·§×ºE,øåöØ”t®ät ÐûãwiHøÜ8™ è¾Ë§#FAc†¢tÓ/A³šûh J¨É:\Cáy^‘H'=NŠÈŒ¢§Æ)tÔÐëæQÿl„µ-¸o«ÃízFÕrv¸ þ¦¢,ŠÍW?–½_¦8`eQ®§ôW¤ñy ”ú´Ò3†à³÷¬"/‰ó¸¸£=ÙË¥lGY9Õúo'÷cfÓÏêüG¨àpqR²ê#O×F Úß5Æ’bÎ΂7áRÂc«šc°$UP¾TÅ ôNàßiÒ£s²!<à2IfZSŒs6à-M·ÍÂÇñ˜Ø=í®9ª%»˜tÿeAª &ˆ&®Tÿ܃ЫÚë7DŽ-Fd§TU»iêlU“ì»—lôÝ€EŸf§Ö@n‹Ù“ù1ŠÍËÔ9¸ÞCÞ:ùb~cmiª€_tÞ0&”*E ]hv^´u™»Ûwǃùßê%×üqJ˜!ŠàO­ø·œûÊ #•¾œÙþü"ÖóSVÃËI— žçŸ¨úÑ ÌûGœ“Žƒ¦@ !øtõ§Ñm˜’_ßqHÊjUÛÀñÁŒ©âsVuVâÒÉüÖÁkU>ýTm^†„Ë/ü©Ý•-1¶iué)¬íyÌ¡˜×ý‘òþˆ”øp˜ú'×di^4”éʬ!3ĸzµ«CWçˆpè^9f%ª]t?„æ‡X.ˆ«B86RŽâìiF7°Ê™+¨·e¹úà 6ã9Ä 3‚Þõ–'K:J#cßÞ;Ó‡Aò¦ÃBï´!Df]×pZ?ÄBé)ÉdÜ«M‰êíÚð$8N§©àŽzF¿xŽ½ö ÍáB¶þ  5­¡>=úÀÓë÷›eLÁÅšiÄ[¨Y jã8:…y)ôî4Fu´×ÐAit¬y¢I ‰•ˆ(õÔ Þ0;(|Ù éšž¾Ð~©äÐGßô ª“³*†ø=<´Éê–å vñaæQâŽÉªìL ¾î1±µj˜å/¬Š‰ŠÊ\:n ÂØ>ô†©äŸ4bKTRÑ+Š.¦ø$þÿ©µ5w…Z¿8×–²œ“€‚v1Bµ\L«(‚ŽuHšsŠ‡u…•ÀbÁðD †ÃD€&ï*‚6 8ßEƒù Ï Ñi77­Ûä7º«àwe»©©ˆÑhjÛt‰:ÓÁuÅè“ròuäÈ^ ¾áb‚²çÙQ'«íŽXzàs×-•¹¿æf7 öºPPÙ>´â×"Rú—ˆËŒ…׎ˆí~0üTävæª%?ž=E]Âvת”†•a&¿ MäÑXPFœoÈéŸýŽòõ ]6Ç“ÉØôðÔˆ­¢Ç†Nª•Câ{½9ì(OLí‚'ŒšºÖÚ|§Ò ²pÜ —N?ÿº·¼¬wŽ‘—;zßò‚ƒPGt=kRߣ<šÙrÆ’H7L¢A™Û~I,í*ªAç+Ž‚h½S·“BÙN…HëÐHÆG'/ ÎK‡`ú,?ÌÏ®ON#±‹Šð±eø‹ Éc0ò[¥êo4G‹ÚýQüSÁ«ø¶ÕñD•¼¶U˜ôj·‰¸Q°0Ì/‡‰d,Ãüi‹YOÑW/Bݵ_oµ ܜ֚tca¿º~9è3Ž4ŸÔf©d˜Ì,©­›(Ð:oKªWhòÈsbÿ·×¬6¬%OQÀã7<‡_UMê>—»-8 -Ʀu G‡2äís'ÛÏuÔ¼¸0½#ŸÉ[9½7¤[Òs[i6›žçï9,$x›4Ü=õ}_’Eo3ÜsŸ{21×è ’<·§©Õ8q5TŸìÿó“X5Ò8ƸqÑzà}¼T«­¬—/GbUŒŽÝ(†ÑwIrå p2f ¡Â.EA®´zÚ…|Ѳlìšâ!§'ꦲcÑ ¶ý²oIås0Îl\š˜^¨û¯¹¡…ÄÔ@p%&48µ«-R·ßhÛ™¼wÛXÅöô:ã—C’Ò¦Ú‰­qèk—ÄrqÜö‰ÙÔçöÖñ‡û¼R×42°'”Fô¿Á¥1 ü÷!Æ&/Ðå0Îq¾Ti»ÿ攑;Mø¿—¢~Ð=Të̪} 7±?Ã>Ðã耯Éíù$2d3ÔÉ= l £9€zß¼.À2Ø„G¾cf[-PÁnoEgª¿…âžÛï° Ïìž 1Sv„‘œD=¤ËËÔÕ s"€âí¼hŸ¿*ña¶tì䢬ü`3p‡ÑQd4?ø¯²Ø|€îÆ‚ ;VáÑ¿0õO3+Öƒüy`ÿ0”ÔjÊس¾ »ÖGù²ØÓx.u•±–«ÃÊQ/Eo¡ Öܲ„"ƒm¼¬VÌ´¯©ìcIßçNñ© y»ÆE¬/ù*68ÌþKä`áªò½j»™[Ð|@E–Çä½¢©oÖžµÖH?ðO«¥ÿ–© óûbȇ/RþXï¢1 }Ù¿Á:e×t%àÙ†wøÉ%ñò3ú¤åm¯5#KOû½ÝŒ\*É‹m®#¹2…By‚hƒòðWÖG7˜úæ›]³ˆcsîf°ÝÈ%Ü€ i®A'ßu‚î±ûýþéçXéjƒ—±$ŠL½ôv/¯y] vËÛ9£æÂ;ëF»„Ëò£h‚Ó¥u¸<׉]K_Ù‹ùìòk|Å œëC êz–g­‹N&á Fh·î'ÃNzQàŸOlŸëC (ð0ç¸v|/µxïÄz ?hUié&¬ ¥òýñáiçwwPÎZ—XHì 7óæ4`äQú°"=æq½ü¦ŠhÑæcD ¿ô¥/ 1ìIg¬SÞ†ý¦´XÏÐâg 4kôoèXG)‡Àô5øK©¾Cófc:†ý½Ñ䧈„¡C‰.mÈã!šŠø‘©½vhg%ïd›^Wi[­ªXÙ‹RŸSv¼¤ß±ˆ¼Ã1,ϵµryxs2”ôç ›ÃÓ¸˜2ü‚‡JYeè“åeôr±ÓMvð”ÒšmþcwRfóµ‹>yòEc+=}¹>ÒÐÐKì/u¹Œ# OÕOÖšꜜ}wÓzQüÊjšÄû(ð òZæÃ7!é@ª¾ ÁuüÚ¿}ÔkW32bRðGpÒ¤‹ =¦Õ™ÆÝy|ÂSζº '¾ß¼“cƵ€O§í¿5Hr-‰ ¯y /¹?4úP4ÐL> ¤]/Œ¬á±àË–VŒLF¬<•²5MëC³³ÊúÓÀÇz–ömf2H‹×‚‚(ïmþ±×>¡1(¼±û/3Æ:G½šs§ öÑ6ñ ôÓAKÑðxP—^ lýßv‘(»[*‰£®pÖR.´c­@>Ÿõð pcu^¿\dÊ›¥“ç¡.orÙÂĽ®'„ 1·Åƒ‹åÝ«üƒéàE¿z"ˆÝ‡Ùb>ÈõΗÐWÞ§–O E¬EA¦ ËJºÆc³&Q©á¦øûÆ‘§À…À-d7‚¥¯²`?ýÄÓg§g—GÍöžô-àµ)™ÎÞ:ãÀEð¨¹ð£„àHE†ß­6e¶Ž†‹·îÇœê|¦"Uw:ÖO@N›ì¨nÇÊû)S´0É÷­ó–¸W $;ÐO¶ÇIf&z¶n2æè;Ã¥-ÓÂZÚG”87ÅÓí®†€©ålU£öˆ$þ7ïƶÛ#l]}r#¨7XzB_Jqÿ,¼ù?ã‰Ã„à À@æ­SçÆøWJL!Sk3žÐh«KOê®x~‘vç̹î{šØÉn¼Û5[€QŸ€A{$ÌÒ×AÙïÊ-Kaà-K©Öó*݇دÉZA³óÁ勽»[(ê²ÀA1›fdÝÃ÷Òç‚!º÷WÑœ7Þ\’=μcá1s,͉ì`Œ:wtŸ–¡Ì€ß÷-¥¦Òö£3O3}X»j=IÁ£L"eüª(zÊ탻 7+(¡^ð ;rxž ü¼aOItü±*‡üë>`xd˜b~Õ×a£ÇÞ|WL6¡˜ýbMŽ6ëAÓ±H,¸?|Ô7µŽI¸R cø>Å;Í™ZÉÍdÆ+noĘ}ª”Â#'Ôs÷×L14Ì×ú8láobÒnLß5…8W·(ÏšèUíÀåÒÎïù=!ð¦gëµã×pöú.e‡^9¸—óåÂÛ(K4qKøÜt0å}¡ ‚¾Q2’@ãX|Fæ(þà†M33vÊ>bÅzönüÞ tiìœ!9ÿÃÅ‚3Gb* q¥ZàôÚ…ŽEÈý„*Û`U;¥±Tº>o§Ñ…M‚ÿҤ󤕬4—ÖGÝÌQ¬¥PrLé.äo‘öœ;þaÖ¯i+—ªàLä¡Ž]U²@h“N –sß –tþà?ÆÜ¢52äÉú*E¼01¨àŸÜZr«_ÙFÊcž¢e9…=b’ûÁ3†·M?®äŸ }s⨛åˆMR™ööÇ AýºÆÔÁñú&i6ó}M"ØYï—×ÈWSú6N¥´ª$—ÒÕ!9æ¡BÿŠrýLx‘rw™ÒÄI5Víî?žèAúoéÑ ]¹]HÖ­e‹H ðËÍÇåpÏ„ h‡’Ë×4‘k“ÊE„æÃ(ÅXÎóG“ꥥr É ¢ =5D®çè{/NDv¡HB%ô)_‡ê¦EŸfžÍÀƯîÍ=Yl'Ô×Y•Ìenu2ê”S"“Óöœ‹bs–þQÙfv+ h˜)ð=t&ª×góý¾v.ì=2áBûé¦P­À~)±=~qG»k¤†@z0ñJ…Š:*ͽ»Ä"PWôÓLþ½@?µüƒhƒñGËH½Óõ«‘V0ÀšüŸΩ¶â%€R¾+:®3(Édf§6${É,Î;V9²OƒWÜi”Ë9‰ñu¿r¾=Ì– –¿]euù›¡ž h’†s‰eŽ ¾Õ ‡‰”*0:`ÐG©z)&/ºßƒÛ±vò¡“wßuÝ*ÊJEââ(&Qæ <þ¯sϹÜ÷mÝ6ÄÆrõ±÷®P&ØÈŽô²ä™ßÅ­7Îô†ÄÍð¨väuÇSÀ<’Þ „Ó”€Ïµ€ïH.Ùš´LI¬ÜC9ßÜuh˜ë¯ÜAóiï‚Ix'¤AÝdz×Øë‚rUZ,á<IéÜ¡Jec—±G9¥¸ sX‡ïú7cõSŽ‹0OpðzQD½y#wþ”ºÊÖ€GC};.;®°S‡‚ýZÀ»™'mnv†a£,žÎ‹þ°‘¡½J™,;Y»)PaÍßsJk饪·ÔúmRͽª¿ÿp™õtþîºø`¤J©>O‰»ÓÛŠ>jeÕ™èŸÄWª2¢ƒSÎ"“!ö  7M A€w5Ó7KfKOb¹’ñ[Ù–³zêáý¬öÀKÒ±w9)YCÀyj»(Ê-M7\¢í7Å:Ƙ(Àw&IÊŒT³Ý ´èSFo¨kÍô~í­÷9·iëÞ슉©·àòXèÊÔúTFtïyþA}wqƒOÞñéëcS) z•èù²V «}0̬ζ¼üâBºN½e«&^áïÄÊ1‡nMßàOá± U×ÖóèŽð(rò­24é{‘ëíà|x˹PRü„½o…Óûèš®QšWØ9\½¢3#&3ÓGMH •8øþÙr˜áø¾`ïß¡ß ÿ-Nª<²Aêù{ÂÆýOV7,oã±Eáöeëd[õ¥£F29ÇCTéõ½õ÷ÝIÊ ÃXðmdWe÷V¾¦ 6›ÂrCñÂÏhj-æ\ßòËl\£‹i¾ëRVO"ßÐ9Ê-÷’ÔÇ${smÊd6ø‚ÁíyI¡G÷eê{*».Á©Àl÷Ý|×Iܼ`æ2L)/ºåpÜü³»Îåoøb¬8웨ß<º}ç‘øµÍ{]7Œ`(ÈØn3nT~<*ñ%ÃöÛŽ` ­N†’ÀwYôà?UG³T(_eª€j>4å0 é´ö§Ñæ£ÉdÝ:ž7Ù('OB¾oú+z~Š“±¢¦óOý”`3µªnúMõ60±@Ó5Á¯:©Yñ`˜ à™¬é¨I„õ ÿ-ŸWªÑ…QZ„lþT³M©‡Ž´€îÈÕä.ÜÇç3àgx{ÑE]ÏÂ+(oŽMA²ÁÆž»ð4Ug«Œò!­§·¼RüﾬÔè`¿ ² –"»§%„›Àºr? n³ŒêŽ§Ø“é4­?m¶íŽÄoN»tÌâd¼Ë—4øFærZ¶7>¦¬= ì_QІ£Â²D¡ ­’›ÆÂT;AÖ˜X_ä,»ùw»xÖ›Çôlš Úö5H « É@{‚IáÁüQl¶ÂgŸÃ~¿F¾È Ÿ}—AÉ¢p¨ÑIªž+ü[•ûõÍÖß4´)vaÔ¤ÓVt#ݨLÕA‘Ä›ZN Më€:Uêšhâ“mü¸ÇOàÉUýöƒs«SºÐA‚n«J³¾ÍËbÑh°=¹Î@'jLEbëyX½î°4¾§ç[[Ú‚e˜Â͉‚šé2 hw³¬–Q_j•gSÝx‰¹×ø qx`€>wª¶x_1¬/y† ‚’M)ŒË¦‡ Îm¡W‰2Ô:Ž ’ Ú:=n¥&õ!ƒý ˆ€–n/¤Ãn"ÉÎC‘Ìm€´×ÆG­ çÄŒP$cÈ^ã¼Õèy¤$ƒÙ‡k†ƒjà T HŒëTëÎÐ×ïö)B1YjU#¡U##U¥Å¶ ¼ï¦1ESÌ1ôPJµ«ß‰UýßÎTÇœÉXÊ-××`Uò}Må”/ꑯ%Ê0OJ0HÛp§aY ¡•€gq‰úÀU—äþ?]_®ûýišÐD у÷–œ•ù/†>Jw“®ëѹñ²3ϦWéLí¶ÇaßXt´š\Õ7Ë"åÁ¯JÄ1§æË+”U˜/ܳÄÏõ4‚(Ÿ†3‡vÔkª v­6Äß…Ü¢Âw¦ƒLKd«ã‹ÌF Î"“±zBÑ›a˜I+å˜ ´èÇ·DZI2z]¦£&¸Š­fCH ›Cfœ†k%ÕyŸMgÙ5ZŸoÉÑ ¡”‹NrùO%åi•:Ñ£¯)øþK|µÈûù# ¸ƒ`„6¯‡¨VYž¹ è_ Ña¨n&2Ï¿é©ó' :…ZB70C%Æ»9èŠ[J»Ý;\…Ü·Ï ŸQÅO Å.*7ääy(²GV[úÐF‰¬Üã(é"ü4ÕZSù¡ëÝŒí^‹õ ”9û°k"„TÑá.ÞS”TMXÕ¢ ª:}=“uô¸~¹ƒrÃ?ÇTd{ä‚;y]3ªXe‹›«ú\AËì© ô?IÈIS&Ýû,úŸÔŸ# 2æ¹”Û/^ê6>kv³Ñ›¯Ìísœ½lÄ·#AÁ.ú*6ß‘ïõ¿”e»Ï5¡DiæÆM37E×M¯CÛ<é´ÏFÌnQ!µ‚(Sõ¤Ëü,ÌÝnër½3Z EVëDßÆ2'À[ú2chgJhˆM(¦e¿Í׺Z:–­È¨Ê¡“ÍÈ@"eÞ*ö%€2 ˜PŠÔ–ĈQJ=G`KKm’gl I‚æÔÀ¦…ørztæp©;+BˆúH4Þ=ÁV:Aƒ#Â_D²ý2Ê,›@Â>_eHÍd? - ìGƒÓ¢ŸAx4XGÆŸÏëBpÅ3é.Ô@VK”­ 8D½×á­™f†kv[©SKòˆg»‰Ú/€´ú´Ý¿Š¯£­b¦áKà#ye!˜á\ƒx¤ª7N‹Å4wc^)g‹~4"blxë˜_ì­­Ì#„.ŸÕM5L']¬Ì3”£Ìó*;SêäßÉaieçSæ¸V <`s¶ù¼B«­"-Obí‚t±µ­¸w«•NÐ âä€8e«¸üdðýÚNCS;÷W€ð¾˜™*©`gcÞÉT!(ûûŠüE @ÙVr9…ÆÛ?cSQÇý7s|Õ´c}ŸÖO\ûW´!ÔCÜj >)ìâ8“‹´õ«©†×þzÄ& œ`+€“{’cZR¬‚WH%ãÁ9x'°šÇ䊞ÆØGþx™Àø”þö©ªâ¿ž §Æ\û—XýðõP³d0àÀt9§(åɘöAû‚ì|ë™arvùò¤Xja90Ub^¿(b@ì]Û-j²¨b^dØy‹;š`Ü,€Z+¡Æ„iE÷ëFpÉnF—ÖB’Ù£FŽméܵ—É›òØç®>?Xe‰êi$2Â(Û(Pôò:ÍûªEr;âÆ#Ð?ȇ7B¤–Õu´C×™þŽ®ÉñÂLtÆr¶ýƒ| ~$»“RAkQ³öcMCXa²’¤¸žU%9ø›Nn6rêzD";æ’3/»6  ˆò›¨|‘–­“4ü!üø ÙÇ‘fÐFÕÅuLÅáìÝñO› (=äèðŒƒÛt…ÄKËCä{ÌÐ}c0OgyJÂ[eiÓ)Ô‡¨)úŒR2ò JcÁ,9?LìÑjgIbÖvÒ"9‚F=%!¹ÞKÝÚ¦u.Ç=JÊF5ÊêŽÕF"0,ÏADIßÇ=s-ù–S ÈõÕëó;‚<›O¤xš½k·nXëÚx¯‰‚ Ææ~ Ò@D›ˆ.w9³AsÿkBV”¶ÉŒ«L§nSžgH×vçËŸþðIØJ– ÷0Ž=á~ºk§‡}qÖQ›ÅÅÄ·ÇÃ… ãZ|O9µtM:è§/e‚r†õ\[hŒÙÁög1õy²©B¼êàt Õé´UïýMM€áyAQ5k$½kÄ"{A=aF˜²=’Y0t²8¼.’¯ýHÝÌ_ꋲÇèéÎHoxÑOÜB„g-t’?4Ü?ðÒÑ*8¯Ò®ÀÞ#Gר7ô®Z"ác½o-ÜäÈõÄp±j'‘¨ïº=/˜D»=#Fg;*‹K mnÁ¼¬zMã<æÏœÏoÜÆÞÌú(ÒŽƒò´¢ ÇS<óƪNz쫸VävP};ÿ¦ß¼câç;ùó€Šyœ…†ËãQN®¼$qrä–…?ÖG²ñsvñÄt‚.»-*™vÅt Áò-‹ÙmWiß²aô$ÔUj»Ö´ª[Q襕ʅ—âgæmòÙçRœkÕö J° …Ýˬd"^H‡O®±’ò¦'Ñ -‰ÇoðJ™jY"ÏöH¶]ÔÎ8p®¥\°b¥R¸GAò1ÅPFš*hs Óì Ä5x„G£¥kr k¦š‡µ{U©I³-mšê9ŽÏlݽÂ|l°Rý7KÊ§¹«{¤a-ÆÒx(GZ¦¦AßÞšGHkaNvão’ƒ§îlTÓÑ%¼m$"†Ö•­`u}µÈ7S¨kq¥V¦³‡R ~¡óè~Û¶«3'C,A¯BŠèVûžÄ©ÊÅåÇ54uÕÜì7DÈCPXÈÎ¥l»ÜÓõ`‘ IYðß³çDêÅÏÈì ÄŒAxªW÷辟î×äÖL„šŒbœÇŠRG«A5†OõSŠ’úsÖ?;  Œ‚ Y03ötm©Fý¿³*#É5sS‡¬™¶‡™]*ÌŬ"…?‘lŠFÎ:ö˜¢¬æ÷¿¸øÇ”xN—+C¢Ò·Þ4\[obXÄàäÑávÝK%­êq\¢§ôƹäÈ”´åGäeluÉXæË}IÃÛiØ"s×ì HJ]0Øl‹p1S¯±ç7‡ti#˜S†n§þçPRªD&&Ù4È ¬ƒ“œÕˆ!7ɹµl=ÃÃÎ~ñÝ­°>¿)N'$çm'xñŠ¡Oý‡8¢2lïL¾¸)‡Ï[lyø‡¬É²±ZVØøâþzc½ÓLCH£ê˜«Xí—ûm•©ëvz÷1‚k²¸]ﬖ¹Û4vß~ d>¯<ë$ÚQgÆKdww&;‘ÀFã´–v@Èo /„!ׂjο ¤iŠ†Ñþ¹×E4yÄa|¬ÇÃÀ‹Ê‡„?¯8”Ù…JÖ¥®|áà±ÔÛ8P™¯i‰ ½j©’¿`P»MÔ Êa‚šö³n¼Ÿkg‡8®s Чúœ¨jSàûK¢@£Ùä½…ÚžúV¡X&°!,Ht3$S$Å•­ç'°'ÿ%©—ê3ÃÿTõßàC|l÷× ~¾;î«òò »içÚ PÍEÅ_µ¼nÅLiÝ_ Ö©lMÏ]ã»Ó•ÿò&dîdÄE†<Ò™ÛêCëæH-]¨bÇû“䚟£ÁCeà?Þ<æî¤$œoÌZ¬Ò‡Ã¢°ÑeqAÐ;¡Îÿ»‰FD9ÍË¨Ú CÞϬÏâK4ô²Zˆm)a¾'X€³huÖy_ b èUÓk ¦ð}¦Z’qs'¤›C¢{p”K9qLomð²…‡äÿYñåîò÷Âl‹*äg)n«`ÉÓU0Žé2¢ñ¼æݯñÜc®¤¢(ÏžòëD’/I~“³n+òCéWƒ¨T ˆK±S¶ÁÂ>ñqÚyu‹A$¶/ã'¼î>Ÿù›Åì¾™„uNSk± œö/ƒ°ÊjUpØŽÛÌGÅâo™ˆ¸#.eáÎ7r‘Ãe\y´vä"±¬j«çï¦#ȃ,ÂïÆx¨MÍÀÀÿ\U÷kŒcŒ®$Ë€pTIéðÀ±Á‘(Þ$“¹;ÑÛ'|«ƒcŸö­ŸA²¥}Ž¿{#Ýô¹¼ÒºÍÇ¿{ª¹~C_û}3@ªtX•ZAÔü¦¿j?ßùlUˆÓ¹éãÆI¹îÕ“æÿI…ºù™Úç­eʆKí®UelèaÂÖÉÐ%`Ï´ò6™`¤A:ü]g±FEã}õ¡Ýã[f†ãÁ€çSJ-•3OûìÙ‡ ×oŽ:}·(ÙoÒT´GÀLUJ„i¬˜¹ÂtÅ?¢¬Î˜´©LC0|õZú†bëÇ||ïp~à6QhÉƼ[9Üà„€úѼ0êÓvBVÕÖmZÁŸÓ™ü{#ë‡ÅâŠA˜´fìGòÃ%ïgÍ@Û³l®;ŒP’”aÅ'¯Q—º®ò~³‡aÿ”£§ !ô9$ZEtb ˜5—êx1!5€+Þ±ÈÛ±úÈœ;¹ü@Ù3ÇqrÉÈÇÜr#6ö,û‡=9&®ÜU(蕉ãß1ØYÇ𔘠™ e{£ßá¨ÏØ—•“åp0yJfŒRóJ¤r™¢–' endstream endobj 292 0 obj << /Length1 726 /Length2 11729 /Length3 0 /Length 12309 /Filter /FlateDecode >> stream xÚmysp¦ÏÖm¬‰í¼±mÛ¶7¶m“dbÛ¶mkbÛNf2wÎï|ç|uëÞê6Vw¯½zwÕÓõK8Ø»ªy9™é™˜xªjÒÌLLL&V8rrQg ±«•ƒ½˜±+  4¨Ì,&&f8r€¨ƒ£—³•…¥+€Ê”úŸ @ÃÖØÌÊÎÊ á`ëànej àswwrwqcpv`ø;I¸ZæV¶@€¨¢’¶´‚$€JRA ´:Û”ÜLl­LrV¦@{ 5ÀÜÁ`ûo`ê`ofõ/N. ÿ°w:»þ%fîì`W–PTPˆ‹2ª‰ŒíÍrR§Û»ºðüEM]­þUÀõ¿–­ñ-“ÿZÿÍ:{þÛb€cf˜Y™ºL€VöpŒÿÒMÚÞÜÀñï°™›ãR ¹üå ú«!5À hþífk«`lP‰:Ø9º¹òf@g{€´«ñßÊþ3¶³²õúÿÿño¸°½Å_éè™Ù˜Øþ·r‘°òš)Y¹þÜÜØö¯PÿÄ5ÿ>y ™•›Ý¿þUå﹚9ØÛzýϾ‹ùgWFIe-EeQÚÿ¶Á?iq{S3+{ €ªë_EÍþø'­dlõï0ýïrÿøÌÿëË»:[yt™þÕ$Lÿÿ±ôÿ%"âàéCÏÍ gea0³1³¸YYüþ/¶¦nÎÎ@{×Tÿ[Êüº ôšÂm¬:˜ò†Z§·†Wø‹ÍWBÒÎÏ©Ó®2ƒ®ÍôÞê¢X [š]ã>¨áæ{ÀÌ™ðï]íJ?´'áO¦¦c‰#õWqßJH+-ük·?” Þ" n 4pìL n£î]3~n+ÑL³eå^ÔÊ8•ñ£H2Ç•w@"ï$®/=ÙKRëS¢fê|Áhkç-à oLßö7À"òCöàoÉq - +‚"´ïµ7á/E@ÝX»š N,ã¶o¢ym²X›3½wNC%–×eöåÉÏ\‡G^ÐyW«”ÞwÌÀŸî¼§É_WçâôѓCˆy¶ÓñM’Uo4¿¼üFê¥Ö¢m-^ö»ãĉÙc„Þ~ÕËlß«óôMŽ‡‰å ¹1P‡÷• ˜kÎl Á6$&1 Üÿ,1J«ì!¯¦¥1,EÇ€©T•iƒ®#žËÔ20ø qk @z:nºK ¿Iôii¦  |nKûÊòðL~‰šXmh§–¥p§_Þ¯6Ê7°fwòÃ{S©;¤ì0Âßÿ°”¸&Öì¡b+%•‹L&2 1#æóeH Œ—n¥|º•Õ¥ý†÷¦a­ÕZÉŒ¸0óþYO^(ú–¹ùØÔTE…ôü©R¼N6ÑhY,LÇåJ‰dM4ÉŽãȲŒn™§*¥if4š¶”H†¨ÒËiÞ¸‘3«ÎH±6PÇ> Yüº±D*žI³÷XS‡ª°ìkØÚ¥lH*]Ò{ÓØê j’t…!R¸«ý¨ån-¾ÈÑNg˜–-<øØé@mñ:‡MÒ†ë*Ó±½åô´”~2)Ó6ž°ÐNí³‘;dé’§#è™+¿o†„tãáñ} 0Ô„Ê!±¦ ›v2Ý}ûþÁêØõ%R£ï@\H '_7,dã оŸ™@&ùšÒóár²ºJÈC_Dß>j4ÝÈŸóÓžÒxùDÎüxcái… ´Ë:Ÿö.˜Ù…–  €mn\{ÚñeóÏ,R3Z#ƒœšY—⬵s¯N—Я˜m‘:V( gàJë.­sŽ¦âßÕNcF‡´Òp0e¯Iù±Heí"Tb¤iØzØüô-G‰³j4aËKhïDÕ¤Ôua€VG]jü€!7 — :êL³£÷ƒ’ ¬0y=¢x¦Ž[LÉÍå’O\]ÑöbØAVNˆ‡„ (F²6fÓ¦Zg³Bkû“ lé>X,±‚4Mhy–î—ôÞÛ2òO³^u¥:a×ìôD+k_c¹’Èî =RÒ AkÒ´¸Xhн®Ù®µC{ãsÒ´TûöÌíÁ˜[Åž%ûfîlsCv`Ÿ-uÈÈ1 +‰öÒUCDLVOù°y¿í÷lœ zjBqR8Ø‘P#–†ûèW‘o½üÇhbaaɃ‚ û,bog¿–ðUþt<~R'[˜ÖQ‹F†5Ð3¶®=õÚyì«dVæ#®„˜~‡HÍÁªŠ$Zëæñ‰±À¨ßHµ*T¶z'ƒxi¹ÄG…o‚è4iÐ.Z·´ß"oƒ3–ÍB“ÁÇMFP5-C|‡‘¿Äú©3ü÷Žà»s]Ú)J5 õÏ’nÕH¯˜=]Èq]a6Η‰ãž±èÞðÕ*‘cµ€?WMâÞÜ5§2|N\Q¢lw7%Ü}Å>0:åvÿL•ŠÂgîf$¡!L`%£Êƒ4×…pÓB­CRp1²óÄozkÆk#à¬'bŸ›ý‰åçË‚~p.' «Êÿà_NÔÄ*¸Aü=„Q›ÁÑ\¹Š£R4Í ½~ôhÞð&ÔÄs²-‰YÝÚêå{ÉìVò§¥ÎëéT½Dç.>¡p€Äðîµ÷2G~¡³Œ²šª‹\¶ž¯ñ §ôÊnÒöÛ!ÅÊ¡b‹Ù=/Ö° Yyb‡¾¹Ðæk.½}3ERj{Þk‚òB#ÿÛŽKÜœïóõšÀI¿NgòþÖ ºªØ﹚‰»¡>¶[Ë舆…ߣrE숙P·_)ýÒû³ VñÜi‘N¹&!‹Ú¿|†Šs*Ôg¼1k 3sLl!÷+P‰uüŸ²Ç9³Šy$ñ6!Ä3ô%•w/¯yº]‰+º«Éd>ñÎØ^»ÐÛ×wÕt_áÝ’°X³ÐÙR—Y½¤¨wYGônÄù"uûä ™(Ñ %q[ñ¶ggª7Ô¾óÝvŸ±¼û. £9%ù³«ØÚÅ¿ï™5Núƒá´‹wz¯>ct’YXýô¹òuKR’C µœND•µTîvbdJV¡äÞ¼^â–Žð{¶Ç â–_°> 3v;aC|E"°O½eb·µ~eªûrí2¥;ï·¢tëŒB¥šÿ,H4oÁ'¼=6›eÆñŠ\mcß±.4»õqÆ»w“3\hÞÿxÛ!nv+øð µ Û}“~!KqôÿE××?¿}ΊGs÷cNNG Œ•Â¾…¼¾n¢äxO‰èÚy‘üàqö£F‚°8cd¼‹K§<‘6ÎÕÐ5pŒd ÂcùŠ2Sç‘A^=«I3\ÓiEŒÛÜàD…@¸r¢¡¢zäŽ6į\zùÙ£þ•ãòR1Cv,G·çääó•ê²8xñýñØJUû¨¯]u bæƒkaÉb嬿ÉÊõ«…(×)|¢.ÛçÜL½ž\›ÛþqÑ(Ï ý˜{#ü\ ãî"*üËÂ\ÊÚ5@ÅøпXé\`tTˆÿ×!­´`ðN>gN&n‚1ýšèÝÜ\qSúÊ Ï—fˆæëØcxürÉØO¦==>·4äõ¯m0[.RÃÙ ¡¥äÄ4v28Êñw“VÖoéÔZÌ;ë`óÓøg–ƒ$9DâSBÜÔ=%ºÛ^:¼Rz 1­†b¼˜+hùR⨬ß>#ËtLÓ6ñ¬œPnðçy"„•ˆLgGiz–(þî«ÅØ:Hûþá¶~Bá¹åF~¨Ÿ“ŽabØu$¤³Å-JÆÍ%é{U(Z •ðŽäÈ°ø¸V¤û.¨õ[ƒ£î½w ÿPOöÄâIMX-™®–¢͇wy»‡í÷Ù°_JÕï­œ+F¥g2Ê!¶7œH!ðTuÈGÈöÝJÝp"ŽZÈq ™#|š(£A9N? ÓH\ ,½Ä`]\mb±7ž›qeï×ÐÝv«ST¨ËS%=õ5«d³EÊ<½×äüÐÍ1{ÎÐãUa½ïú‹J²Tü»­K£®±9¯Œ×£hM ¨º»öÿ¼e8PX—/~¹¨lSàTn’ПbF{‡>Hb-äâS%G…ob¹mW€+ qÁ\ „:ìÔXð…ãßœ,(ôÕ‡œ4 5,ÊQ)'MõxV‰À¦®#Õä+¦Ól(ãá§ÝÍZCW“èùÇ•ZC‚(lδÍÀiõ×Øä?¢GÐÔ:4¼ ,ˆ½ÛÐÄ«•ßÀb{2à"_e†ûçžo¸¼TÅ~M¯u ¶Í8š¢x=!¾ßâ)EŽèÐK÷Bz6sdÙU|ü†Iõü×=BPÜCçŽ,mqwR\vªoˆ)× ˆyDXPÁª]—§"åÏaíý¤XzãXѹ§Í]>¥÷ÈŠ·¾ÀT,µ`rpó‚«)ôêî°bŒ·¯Œ•‚XŠh©Zr; Üºê¿85Û½€§Æœþ°°¾Y†ˆ_†P¤‡€Ë9hqä”%hD™÷ôFoVøA¨¦{dƒ´Æe° Š«ûLs•Í0ØáÙéPQ?ÈêS ¶ØXõ Õ$Ž´‡Œ¦Ÿ¶ÉKVJbž9y\õ`ŸªcØB×ìvBÂƈÒy O#¸ñ,XÐfrƒ›#Í׉5£(üÀy“ØŽÑ|ŽþÂH‹ä)‚^|åhòä½ë…çOãxú±£XèÎR­+]ÕšÝÃ'¯òô‰,=4!O´QÅL0hS¯Ô‹5lt3š2`")M›ðM6xh} C¼Ä@7Â4?*,Ÿ ÿ débu´þ7$n`Ä+èMǯEAå8È«ÉòJ\¹4K<ÏcJó*Gý½h” ¶¼.–2ÛkW1"8!pÓÅÁé ç.Ý{…ÂÙ€“w=,ü_„5ûȽ £Îkã—ØAƒöìkøÞÐé»Ã£.@¶%Ë„Å©㩃;ÇQM&;ysã­Oy(¬ýrnïÕ-ù9¦N”2 ¢ zûÔŒ*=æf¤½dìØ…ó›î’-Î:ŒèM8/¿Gcl6]šCypˆúÒÌÚµ—”–ÄżF]-xi>+Ú’§ëë˜PøžêîÖ›ÅtøaÒyI'·¥4í _0R=Çæž9ç»"ѶÌ)Kç´ e¦u¸èç«\Ê@s»©dþ¦ÒlCæÒ^šïžØjµQ®t¬F3µK/xð¦4pP‚ª'üßó×Ú\ˆ¬ÒOEÁ:_3’rq䛇äÅhR$zN‰ÃЀàÇ²Ë õÝo-àGÑÜĹèä¼F(k=ø¤AÆÜIÒÞb¦ÝN$Xwˆ¢yµI¯BíÐK˜ß­£2¡š”ÊóæRÃf›Æ£Wo=d¬K΢e"’ì¯>â—VK¢­T¦d,6ͤ^Pl?hMfÅ·nÔœjâv— †t$ó}²>ƒ åЊ¼· ugq8oE…ßbØÎŒ Ì:ž.FD3K%¥7–½Ø¢Äž!r“×ÄÎy: ðµsœ«—ûešTËÞÆe ó¸¸{®‹‰ïF˜5ʦ2å:jë9ZsäŸ`hÔÜ7ééÉôEKõã83€bú#0”e*ò±Û›Ž’~ÙóØÙöÝÚÈn™å¾[•ÌÑñ«a~¾ŒÚéyŦt:Éx¯f>(-Žªç_šªUÏ¢|2¿!/Ó„š”}°!©€i¿Ï.!>ºB»o©"Òš$ðž{w=Y¥¡n߆èÇ`GÕäÊD)Å©H¦Ž¥þd·Õ¢€ÎƒÒ›ÁYAÛ\1yy²ÌUÄGÊGÞZ—Ý‘©œÐugôïÑãZ€$x*Å‚¸â÷ ð*“±¾¶÷B¨ÀØ»"¹Ê™gžëH)í—¶úÄŸ+³\ƒ€ýŒŒÛãÇ8¾*LùÔu ÓšÕ¯¾Ú,Û‚¡åc6:k~Ý­Þ0!µo“0+(÷£œ Iß‹™Ä!­}a=&[s}€-™Åú¯tmç'QªÙ‹V»i°NJ»bÿð¤ çãšÆÚ÷‡/ˆS5×m× óaæ–¤®•áý“g3wî•”â׸NŒJéò4ðMÍ9׊%B>v(4½iþ®Ìë(Œ x Ÿ¼¿Ì¢Ü}õ•Qi~ö×ñbû‡"ecõÑ•€z—º¯4!Ó6Z Þï/^¨wÓwr -ÉýSûH ‡¢õ€ ¦¾½øß}ðK‰åt/ó`µµNndpöX†keÀv‰õ²o ½‚¬<ŸD˜fsñ‹ÒÓ«¿9’F¯„˜©³2BŸ,#TMÓÈêCŽåÓ£úa ¯ÈŽ–çåØŸ„•·Añãj{|T^|ÁQ¡©´„ Z2ÁìÄ•\}ÖÔâô~TWÉ28~ÅÙñ”Dâ ¡²üVü™Ô}3Ol!…š_xlm1îôË”ZóÑ#s8/)`×ñ¶óCfá…õHI\_@*^Nà燡…'xhù]Øžÿ~©Ó•¶?³tx´[m²|.: fÀ#—Z;ÊOšF¶n·3²³ÚåÞÙÆÍï„Â]Y=gZ÷SRW9˜)Ní®Ó3zpå7RRVÔN°ðeü—\åj¬]}Ī]PsV˜*#McÆJÊJ7 ”qYså¦o²‹1•ô2Ìêe½à絶Ûc.þ‡á'ƒðjIƒ› üò¹ˆ “ ‰õþkµWø–ë Üô¬èú[¹eú ùHµnI*ÌÔÿ}Íj™Žx£‰çѤ!aÙ°ãåg† àtØš21® õa—ÚŒaŽ/$(ÌñdÐ/¨CáÁ•qUl"·€Â~ê¹ n–zBu©LÙ²îÍ@BDaN ¬Û8zý Èp©nêì)Ø.Gä¢/Åjä¸(üf‹eÛftˆpß³·l´A¡!i‚Ä®ÿ'–à*]ÇT1ç# [ßP 7¼j¦!¼j{4Ìs® ag#a_n‹ù„¬§:à• ¤Ùñ%…¼¾R¤ø³¡-ý ­ÅÁã°Ôìœ} H6‰¥ñ{4{FŽÒ+‰©­nT²é5Bõ;g)Jö•-‚¢/˜ ”±M¢Ö/Qi pÌKÁ†Œê¼IbLä'k!£Š©hŸŸ…•!ùÇÊÅ "ÛÓ§zùü„_õß4z'ø̘éi;Ë,Ÿ+Þ9/jЪWÊçߑײCüãOŠÖRkt»CÛ!=±öÍœ³Qq^F‰€í l]a €?©Ó*m,Ü¢¯¢@<1 T Xë¿m´‚uïOÆ×ûVn8`ç§øÐí͆#”[^!€„öHÞDv²°ó‘t™‰¿›à)è_àüAÙ5qØX±¨êuwmѤzA6 ““)DZ"×iÑÀýUŠ2œ8)ãN5Hî]ÐdY ´ŽB?\ Es}kd-µœ¹Å®*ê§qw8Ô¥ñŸùŒ;˜7ÚÀÌÞ:±ÿÁ?ÿ!vm+n3žËã%üdJTÑèiàóþ'ì sº£ã¨À–¿ ?Í.‡ká&‡È™ê|èexR4”Ø Ž±óÁÛ‘ *âe¬Š¡Jñsˆ¶LîXº.a±æ·‚ <ê¨:ÊOvÉ@“ËÕ%Sû‡ ë±l2U:ün4_4£íì­öH½˜aâ ’—\·2³=Z„qº18Zä%X]ñ5ÐÑšY[¦{iáôTBE’m½zæ«jˆËñ?”¯Pµ|m~Œ 6éߥûl^òç4î-.ð;bõqG²H±$êY´X­òL–Ø7ՇƫqøÃ’3$VÂa«¤®”ÁéÊãùʨE½Åx!²bäÜZeÞt6*y–;¥ï™Œt 鿦£‹ÂžG¡²ž¤ÞÄÁìðßTÑ“vE1ÎÃìJÙJ‹B.‡|Þë|”ƒŸÛ§þîä™}V“7¬ Œ"×Üœd(àæ@éÖPät+é @-ÍrüŽU'fÓ•"mÃí¶ŒqÅÆFlR`ÆØ&®(µíÖ“÷¿¦ÍñÁYÃÛÈG½0Gt¡¼…H…(ÑD”­E WRìÝVnP2-{Âî˧PsÀ¸f¼Ž5Z“6ÕóA« Ð1,„–‰-Œº3vJPìúáTysYñ³a°ëél“ë´¥ ´šC­ hµ~uÝåÆ8˜V¯$•´\¡£†Ó‰î¤“–_:÷÷ dZy Ái [Ö—&4ßøSr¨D✉ê›0^"1U-\*ÎÁZÏ(ìÞ#œ¥ÝÀˆŽ®]$Ñi‘B³Ê1=¸„’Hxùó+T’’;·£0 ¢ØÂoÞ< D.Ðúèô™®„CÖø½]ïÛÇ“¶i+n4ÛÝz3p Åd#‘DOÆ%Éd)lâ€WtEŸß¾“ò›ç© x”߸ٯ’»ÍÌ&‰8—?é3ÿ©çKº*i¨Š_®a̵3ø ø)æ÷têmzØçÆèGå¯ÇOM Ö¶Îl*?ÑfêM ؃î3}£ÌO‘ER€eÙäyoþ%¿­ìÅË“…Ql#…> Ó¼½÷Ò‡ñº?³pÕ¤8©Æ"µ,ÑQ·7JûÁ# åD„„‘>æ6V?ÒИn±Ã…5ܽ™®¿ ž¡…J,Îpbij+~KúübŸü3éê2d¤ÔÎàù±C›¡Û¡Xý;Ç\ŒØŠ~ø¼•é‰â'Ûàêö¢²Êrl|mý…]3…ÄÀ’$’vë’ ê:ÚšBÒ¹*Q£¸Q‘B÷"ºF8í¥Q–Gh9YIû1ï¾w)ñF…ଂj(8¬véK2Ìó‡¢öLðä±”‹0¤ý„Œ*å)"kW)hwÉKŒiZª¸ß’¿@d–r±[kUñªÜÈͨÀ‚Ákôóo/à”?0G½ç‚+‡†€£iÂýÖÁ¯þ/i\h\¶ß"’Q 4Ùp¢L)Å]k?ÚÛ²lÈ:Z詤l¥0—¶1§úƒq4㖎𹫾¢ÔzÙ™¡0ÒjUáì>åudó¦‘H x¿]¢"²ÀÑßËô¢X·§¾ApD¦G4áò^ÚæXŸ7¬xÎc{¡ép^ví¾É±!þ~I¬{n¬ipSü¶Éò. ]q¬®«~eÖF3”‡‹­¬2…+Í3†óá!ßu‡CÖMï1ƒBÏ3c°’Öú®©Õu7: ýÔQØ¥ß=T:.»ª‚ËŽ/¨^yX(£‹h9&¹MÛyÌ 6ÿ+,PK·{¦ /K·½Ô6­¶Ž»+´Ä±9+Uóg«þjâ'Y€ã&újË¡u~–Ü^…N(A!Ýù%2¬. |À¢¸©x¶o=åwÔL§ËŒȣθ€þÀ‰â9ùå#Âb÷°tä”},8‡Çk]á3·clŸLJØ&kBõü¯rF°‚çiœ •8áFg^¢ùXܽ¡ ³I#@ÔnÖ *™5¡ïõsŽÆñu»!yfû@Ô¶,Ø¥h¯'–BæévÌATfH¹~‡õ–4ºk;Z?­–þ<»ý uŠ èÎ2þºÃ˜-¤¨ í"$XÁ¡ÓäÆhêEqý$¢Ž¬¶UM…`¬²Å®#l±”_¶KGŒ]¹w ™sÚÜHóJôñ'ßêš?r“ Ñ‹‘WÐĽ­¬ßö¢”jíõ{ª%!AàŒ?Êg‰ÃºÁ­€ö¼\N£¸ÌTñí,µö rD@Ò÷A•‹´Ë=Ò°½ÉÊ6o™f!vˆá°/$†kª¤E¤ê¬÷Uªç‘Ö5`a.Œ'ΔÓP»€°ÙZû8eQk‚ÊÖ±t÷¦)_Å¡ŒýöÅÖ=ŠD"°†y´Êe­Cþµõ’‹Å¬BÌÿøü½>Ó4Ú±[€}”Ñ0?ì¿‹Å+é‚dW©á¨áFI9üg Í·2W½êÛl`vF“–ãÐ.õ»½ Â(Í€\˜ôqê8 "ÕlÝ^3ç ·—sºC½§'óåxMCê“ ‘šÏÙæ "êÑí=(å•Y×j"R¿ÀŽßuœ5,Ó]7•Yê×ÞœÉ|Ò ù£:²Í|ÁÝ>Ú…+Ì%Ìÿó#ªßtj‹žÚY½vß„wˆƒ ñ3…”æc_kªÕ„J ¸º\·T.*$%5ç ŠïÊĹG{Å99QÒuƒ_c²å(š¬=?õec/˜Î—ïш¼Ãdýƒ¶…ϸ¥| ‹´Y÷óu27bQãldpdð TÉW›³ æ›Ú’9>á“ÔÍâ“K†Óu¡¥hRþð¡/wçÃ.ÂG›°,GWÕ·[•)¯ˆÎl’ï~±²š’{Åúx¥&ÝXw¶çx!uþ½Yû:ÙË„h:ãY›sä÷æÓ3)0ª,z`೟7ã)¤8ƒM–Øö `Ú££ù"Õ8E+SŸ<} õr97A1Ø˶¶w(6(w=&|ZœbâÜ$®æ‰I¾U‚)r‘aQ0 pYÞÞŽ5J?„îÐIT(/B}±Í®€ûñæÊó¢(¢¶"¯¸/nºl,Û7Ñà²U.3¤Èx»»M™ÁH¤‘&Ëã,.}„q¥ä;¸×¥Ã×h2"› ¬§x›1*ìR—ÌÙj¦³¼DB3!D½¥7ÿ‘Jl®ñÄKÓàÑŠ_»P<þ(œÞº›ÚIÍYm6¿bÛŸÛê‰$¶sT?ž!ï}ñ8”n»Ï^;‡¡&ª¬Þõ49Ý‚søúÊZ*TS!-slw¼ÉFÜ*3E™yݵÞ| 2Mã„âç®Ì\G9EÅ ö™1±<Ù˜^+ä¥,.Ænð”aà˜¨x’îÞýE¦W' ó§dó«÷RšÍ‰hñÕwG*¥qüQ$éîÊÂ,šD½6O“gÞGL4OÂÏ?Ú£ ›1`!"+?±wI/Ò‘ömø¥X=¨<†ƒåÖ]U~4×aÌøGÇt)ôŠïn²:Á%ÐL±9…e°ÛæTX\yüR0ÍŽf‹ ¬ìÿ«S›ÿÖ¯:öÍ•ïLÄ>?BšN ¹ „4;ãd>wzNÚÃر\΢3bLã¿<ä«P[{­Ì=õA:MjÓî"¢-ó£q¨'€‰êàO$êÛx³HÞ¼¡Úb‘PµËÑY½|V" –§Æ^KY×L¼×_çù£ºÔŸäÞäÐW¨IŸ>¸ $)9¦‡Ðù ØG˜PÛÊt~ßη1|Ü"xôv˜< ¹væ -]¬«ïnç¾ê¸ß'®’ï·[_Ö´ó;؅Ѷ;çñŠ´,Ûáyì®$g§°µ„Tk;;jp:ωÅ0¹¥’¤ ‚E´‡¦8‘Á¢}Aö.uÎü‹üxt7¨/²[ݤ±4±¿HX†ØmuL&å’^dUQCðšç«è}î®;¯žfzg»tÊ€©Uùš6‚—Mz,òÐ\ÇÇ™+ZÎ ÔÎÔ÷Š¥J ¥ú¦«mÁܾ)ßß¿ÜCOMk‡þ`Ÿ (™g¼b¨íò‘æ÷z"0Ÿô–oÑ:jÑñíšÞDˆe 3A}Ç4¬È„Â/üè]G¾@‰Qįùæÿ£O½÷6c4Gý v>gמuáyµœŽqò)Ùçâ}6C£›• VeìQıßð³×ÕÄ(y®|Ó-(c¯•KŸŸj¡ÌÁ5=ŠÖÔ©É»ñn¢ù­a7GCÇ µÕd…ìU¼ c¶yèaÑøÖæseh{¤úf·t?O_滦!<7Û;H¶BîÞà“é”\H›äåƯ€) +J0Å y®×N/-Æ­×2G㚥Á·íÓ®áðݦÇ2E=²¼ EFƒ²¡Ä s«ìl²}Co¨Ç¼Ç‚;¥Ñˆ& i¤H¨²ž'3 Ðë…þôÆT]D »¦²6\m¶àø©ˆcp?Â~ÝU-Þð2Š]d}^²€PBúÑëd*’Š55¶Ô^5@þj9¹`1~ô‚zHòkŽ±aÓQÏ¹à‘µ°¸ sHÓ#ÑŽÀJŸ¨JHà ö%FÐôè"æÄãü{Àpˬo•ɤ›!ü¶žªàµ|"æ=§Û¸»bÈí‚…ëÖCýb\²šëOÙ jIMö« i¡>…qŸWG‹ž?àœáB^ìüYÝcƒz+{$‹#L ,ùýb²b3NüþŠ=‹¸,gb„ûáÊSj­¿b˼Ý[ïºÆT±9T)dk šÛ…¥ñúxÁ¹¶©M,bªì…ïÀÙ!šÒšò4¦]b@d¹p¯³kÒ]&KBTË?Á`ùÙŒM«]=Ù•aÇA*¶¹€?S¨áÙÃWƒ`r×ù'‰:‹$oþ`JÛÕûr³2c_ý¼ ‡@üS…ixDœ¡äêšA¦fæ`ó{~æ Ç~5ksØÜÿg8zCâË·nzåørÇØÔ›3Ô×IJ ˜Ê/5všk;2Äî ­Ï©°òºoQ‘×hò*P˜Ï fÿúÚš„ò|€R0-û1~:mB^FQÆ`éɃs»·´ÁMáfÐôEŽ§‚(ñBÉÞ}«žýĤÝò0üä}2žØeøîŸî;dzÚ$JzáâÆí¸â+Š.U¸¸˜hóÜyxÀ¤ÙsÚÓCªHzðãØQê!à8P"¸ ¹À†%òŸazÐEFÅÀ†®M¶â0œÏ±7A@×Ïò·n[z×N#“ÇÅn0Ë(Ä*ÄÍdûíâitšd©sFRæ"/ON&°ÿU áü®Ïš7ïoƒ”½¦¥d.Ò¬-(‚[0Jäœà}iÊ¡AlChQN;ÚŠÌ6Ål³Ó%pG2†@á²]FÊí|_‹ë¥„ÎHç!ÕoVÒQ'Ÿû%«\ÓÑ°¥BÚß6ëq•ÕbÇ–}Ú3ŽI®|LJßµÄoì]5;Èl>Øúüå ¾î‘Ò%4§Æ^ó\F¬!UVìsûo€š†ÔYyK“³_6K^?x»FûØWm|‹0ÆôŒ¯>fKWxábæk•Ã…ªÒ?j‚ª±ä þ]» ï`h{™_-[¨>2¡x''(Cž3úS<÷ÍômÄÓç³}Í8t¤õÑË’hW`zi#·×&ÒjZž¥“K0[ÒJ’úÌJ*cp$Rœ_ÉUIåÝ–sYÐ[Ï26ÉlË8¥;¹ÉmßÄÙýÖâ%öº…¡o‹°àÄKvIš‡ÖÖ̓RrVgÄ—Írôó'bt4 q’*¯hn{„´ ä2”v~AuˆF¥î™ o‘6}(³>õ¸³›ó”)·º¹äÍí[ëR‘Î’NæM½ó½uö¹OwèŠÚ 7ÉßC1pW”ÐV« EðZßyLf,ŽKôÇ—B?Îé~óg¢ÆåÕò¯®|ª/Ù“7{•(4ŒO'©kÕ8í7þp“m¾rDz‹ò¬uX¿òhʘýψöhÝ[lžà¤ Ô¨Urã,ÙÓuyž=¥ø¥rÎœ¹Ž÷\D¿A¿Ú·ê­´=®º0< )Â;8 *랪Œ¬‰ÿu}âߊQ-’(Õ&Ð.–±¨$ß¹ÚaÍÚ6˜ìvÐXsÓ[€=DÝi½ÐÎ]}$r ›}`Eògíű~^^Ê<¾P‰¨ìH Äb°±]‘Šß3‰,3Ñþ9(7Dœè®1Ê:¹cð0h 8ÉxŽ¿4®Ö¨e˜Iºíï«·XÐpTŸ«‰TK;l |%Õuº$-ôj-‹ö:6=^…ضæß$ìš’y ƒubmÆ1K­P½J`tÕ“zÄ^¬£75 ¡}æ¸e?#»g«JÍU¦tVv^æk}yª s[GõÿµñlD,‘ÕÚÔ>DXèÆû å­)ëØþ £ý¬ð°+¤s×èh7P©½ÿfyøÈ[{oŽ"MåIp+ åk.1VseÕ¿à”«Œñ vFDqaåÝ-« Eqàt± ÈV~^MZx‘ÜÈV{Ù‚q½¥™P…òPÝsM­ëEKÞÀ[z¾Qu³_š{¶ MmAžÑÿÈ&°A +©/ ´rC+ôUdˆú®ï˜zŒS}`Ýá’*h°u €Û\±Š÷V‹okL{ÆÖt\êj¡ Q„9ÖØ•v­^ßÁ2±æW9Y¤y¥h2Ù ö ØïZ'2wdzä)d¼VŸÄ:åËSæW³%G´ÊÍ{æ°—¡U±–B­ y: þ0ËÝíQ,6¬vžÄ‚ <þµp¶?~µg‹´W×E%cðÏßüiz+æ«-6J½ôtI “åIÂú$¯SÌûë Ñ<æü·J}^ãÚRƒæ}þ÷/gRE V’g%FO&“x9$ã@‰Æs¹7Bnçï1™‘'áŒKRš’ï = vöÓç_„mº²yå=ê ºz¼!Ìù)¶Ò7}í^?[@¬µ8e*fITÊ|z«ºß¢|7¨ÌŠY®üḆD!QÀ½cE†ÑÑÇ<{±½¶ÌÍàJ|—¯#ïêòf?8\˜ê9mËRõE { µàÀfÁÒ”3¥m“æó!}4fƒXB…!¬ÕØÕ1œˆ,cÚØqŸû¨‘Èx`õáJm’Ú6ƒ‡(.ªk»>Ú^ƒÚ»Éû¨‹Bg~5ÀyˆFŽ6gs!K»{ù?+Æ’D endstream endobj 294 0 obj << /Length1 725 /Length2 29292 /Length3 0 /Length 29864 /Filter /FlateDecode >> stream xÚlºc.]5Ú}Ú¶õ´mÛ¶mÛ>mÛ¶mÛ6OÛ6O»û¾3óÍ|qoܨ?¹W®Œ½r팪¨¨"%³·sQñt0e¤e¤cà(‹©¨0p00蘡II…L ],ííD ]L¹ê¦&eS#€‰š lïàédiná 0¦ü7P³14±´µt¨ÙÛØ»Y[xÜÜÜÜœ]éœ\ùèþ)R65¸X˜Ì,mLÂò š’râ q9U€¸©©“¡ @ÁÕÈÆÒ ciljçlJ 0³wØüg0¶·3±ü—&gº °s3urùG˜™“½-@VTEPL^N *L¯" 0´3ÈHüSnçâÌõÛÔØÅå_ÝÑ\þÙþ/2ú_ô¿¬“Ç":hFF€‰¥± ÀÈÔÜÒšþ_¾IÚ™ÙØþ›¸:ü7õ çô(þñ`bjöÛÕÆFÎÐÖ@!loëàêbêµ71u²üË{w'ËÿR m-m<ÿÉÿ‡!ébø‚væÿØÇðÈÒYÌÒÃÔDÁÒå¿]œ\Mÿ«›þçldMM,]mÿSnú'ÿœª‰½çÿÙñŸVþ½½¤†¸–õÿ†àßiQ;c{K;s€²Ë?~:™üøwZÁÐò?ô_-ÿ(ãÿ]˺8Yz´èþ5" ÿºþéþ_–½‡7-#3€–™•ÀÈÌÉ à`eôùÉ5vur2µsù·éÿôòßõ¿‡ÉÔÔÃÔz}ÕÞ˜;Ø*­5´ÂW´h¾Œz~Ö@•z•øÏLï­6’å°…‰þÑ5öƒ v¾;,Þ\€ïÞÕ®äC ‘aœ)ìÉÔt4a¸î*ök q¥¹oíö»bÁ[C¸Þ ¨–­±èmĽKÆÆz Õ4KVîE­”c/’8c´AyçâNüÚÒ“8¥.9r¦Ö7¬éu â¼EHáÿëþú¯°ü =Ø[R,}‹ÂŠ‡£0%¾µ¤<ÉRþF"uÁ›©âöÔœ5%ï«mú‡äÚÐ Ùh\„Øk¦êþ׌Â;Q›ôßÌÜ}=#uú\ü-VÌY^Ì¿­q=Sáé5;ÈòçߨuL{ï²˜ÓªÇ dð7âFK€1úîVe¹‹/^!N'ó,ÉA@WˆçhIqŒ>9%õã8ñ\µu”þê¯+ :QÝJF×@ª¶S­¹ÞšÍ¾Š1Þ·Uîå yYôÿ%³§Œ0#Uÿ1¥yÂIzÖ7TeY& ¶W¬IIQßlñŒÄã7Ï8½‘FAæóo+¸jñûLÚõåÃî$÷>P¨¼¶5_ !=}¶[“aa…¨`ÜÕØ°ßKþžÔ@û«HºrAŒºà•yŒ­~‡‚¹ã}ÝÛ<+QQ#Q ŽœxÜ­‘¹¿ýXv‘0ßaï>OÙÇj“äÔãݪÁ–aK[ô·-§Å’i¨("p¨©¼á¾ûÕ’ÜøšE#œ!òŒšÌ5ôpÕ3–‹^f¢ X»´GÒ?¡Š EýK¬b¯¸'¤«t4~ýNƒøÆd•…î{zÚäڈř™d'ü·êÁÝݳêxKöX!¢©%ݶ­úò—v;KT£<å)†òyC1ÕM`ÓQ+íÂâÑr´)넽ÈØož›4Ä H‰Iaÿ•íé¶aûb©–Qà[(4㨞.^ut†I˜y .©]ðÆáë 8´D#ÌçrjÂC ~¢ï›X>OóN½Î 6n$C—6±„ v;£„ÖFÿ§Üïð³wĉb³cGZ°£¬Ž»·&ð`“`è¤ýÖXÏF êÙaù‘¹:4ÂìQ *1‚v! î3¼‚Ò«€)\'FÛ™z&C'6)X^»±Á1ɵ^q\¸M!vÚÓµ])'?ìÞ»(GÉ©N¡tÑrÙÛ¾ûµ‡ð1EŸüúR“+Š‚â0K;|iÿÀmÐ+§iP@ î468ï-|¢¦Óùéj&RheMi%ÊËùyâÝ›4{ΑU3SJ¦îâæÕÊÉßè¦Z·u`bMƹš»]Áûj„ „ÞõðÄƯÓ[ñJ'‡Ž®J?äÑ_›ApÿýOÌ$gFÈoÈk…XN%+RG•/›×J|D êÇbZ¶Ç9Û¶$µÐá>©Ö><‡Æeù¬Ý:{ûÓüñ ëQuXeÓÀÖg—€ó$º’¿†§-²•3ÎÒ6ÃÒÆ-9²>Ìâi/öÇÄȆNi?.nz‹}dnK0Ì1ëo!sÒ%=~=yÉŸ”¢ÏE ÀÙêØú£+ôõ¾‡:÷!hN†…3nžæÈÆŒS*äld—ÄÝÅNZåÓPI¶¾1ÄS{¸ß˜c;…¼;Ù†³¹ÃÍC‰/•6ZÛ ¹ôËÁ¸õ¿v:=ÒgG/ ‘ï…±Ah@ÄWè0% ®àR%BI¯õä¸ÊÔ ™ë´a‚Háã®EW—F×±<˜kfdÑ ¡®ò^ßJh‹<’Þm7ó«ˆÛ['FŒnT¤+å(«š¨òª|«ÃûÀåŸxö”üI›naäZ×MÔáÆ Ê …³¤ó>ÂévFµ!ËÁ°,û½39ÈØ›f£EHWªe‡: èXy ± Dì§I¹XǦ¿Ò ñ#ŽÊ[IÓÅz}vB·ªPÂx´¬å/¸lÊb6zÁ1ŒþΓÅq/oó†´âyÀ´2]¾RDýålZ;ÍdV|“C’Y¬7—7‚§Wç E ^PÚ;ŒÊ¦3`é9Äà–šõüÂÅ@’V±fi#‰9oæ\Bd y‡ÝóÊð­/¼˜¿&KµýõÎ'œ°gdî¹­&‚Õt¢u–w¨@ñëTÈç>iä¡¡~­q«=w=¨X˘>à@ž;§¢ÊÔûaF†‘^´i I_‰iŠÍ#ý†ºô óZ NkžA°ÿx[ÿ%MÈêó ‡ï–Å¥ÙÊÁcºúL"¡è™ÙÜÙ&ïV˜/jšÿû*õ„ R ÈÜqj(ë4ÍiQ}Š½6tŒ€]u“5ÖÖ¬î±\‚ò¨Ô(O·r(ßbrÏ1ÌORß¡krˆŠ‘þÆùkQvÚD‘§vr\&¤ã XâÜäLÃ'¹*yû+B"Æãïßó÷?\ú¿¼‚î™ÇU­ÁqdHŠ^¤ÈÅ®Þ\ÿÈo¡¿ ûWüÆuäã÷´1‰×ô)y˜ëæ}|ßJIgV¤_$=^n”©ä‘EÂB”ŽÎ¦T }J"êw×!ܯØ,ôþB*ˆÁŠL3êñÝýam½õ€YÉ|CÀÂK& \è´<äó \½áÎTEEí×½ÞL0ðœŠl‘ÿ\Ârx\’»´¸°mФ¯4ªk&pß»×_{iÿîi ôW产øÅ>µç²€TRjÌ,ÆÅ ‹9Üßá˜Ô;Y‚á¯d|€dWá$Ì|C¿Ô¾¨ÎÎÅ/ÚCZ²T…ƒ$ÒÒ÷¡hC96î=ïN$ÞÑmU¹¿M¼kÃpÕÐЂÉæß±%ý Xã ŸY­] -äP¨p}ÜHoÿz4ö7½H÷õqÀ6¶"P½óNØô¦K%1–«ú¶B8H­ZÏõõÄsÍ»·›oD õÚ~„"\‹g7ÞßùñÔl‘på½#Ü㉊é-F}–»Í(GÊ~9ß`¶çŽt2º˜ÚdI ã&O9ÓŸ9|Ó&—á£RûÙÛP&®hêV…PÎO¾Z²*x_?Yôk½ZfO«t= pNgÁƒ¡Mj~¯ŸêŒ¶wk{­Ý¯ø9§ê ùnf{l’ú–Æ5D±v*ÁÐøy¼A¥:Ÿ¥$ÕHü‰›Ñ¶ž}ÑÕ>¬TÚÍ)(‘û}%d¯¥_Ó2î2g1"vq_š|gEŒ!b=ó´l3@5¹Mþ¨MÈ¹Æ lz2½ÀÚÈW­vPwŽ׿áyØ4ÍêhàQìÁ5IZß/Ìûл¨4|ï9^z±OØßc)A‚éi'Ò†(æ+f‡ò|>áK/7‚ñ\#㥪?jxZ”™ È+“”ÔD¯¸‘_Â`¹N¸¢¬/PªÖ·+i‘æ’†B³Sê!5M‡ò~r«ûÓYJ»ì6Ý’™:wþÑÔˆâÿùÈ1ïë*(* ŽÈá¹KÙî!(ѼŒ­xõ(Y”;‹L‰î#×’4!Û'H›3=‡°¼ƒ1¨ÆÂbÕŒ¨“ÿ;>ä_‘ë4Óæe2—4ÇÉÞKº#¶Áiž¼><Ý#½y¡ d,ý×áЄ^oßB:ÏCa— ŸP”ˆ2ÔCu‚—~„NB~pÒ_+v|à-ªò ܶÞk·û1iJˆxz§…\R)9m:³§` éÅõº·ø¸1+RßNm»-;¹çCR2àsàmD¦v5—JåI3 a²—E|@Ù¢Áˆ›ÖŸ18p¨ÚÙOèoE„:1¥w ,Œ¼]‡¤š‡­mC±ÃŶ‘d“ =ª³?!²‡Ñ"O|ªÒ¤>ƒc<>ÊÏp¡‚ƒ,ø}öbǦ"£šXŒSÏ“¿Ã~‹‚,3²Ç‹!gñÎóálJP´{WªiÌÖõö!ò¨~K¬ÊŒ´²ÑRfˆÓ\í¢”~,V˜óWxpŸD6Mem’¤äs¤ý¤M(Q…•Ÿíéä‹”a”ù-Õ]ÏCžðXª;ȹü·oA9O·a =Ù91¸~j8úøÎ'Dá™ÀèkP©;v9&ºIs¼í%‡]¾“>ëXd‚½E9¬Ž,½K’ÿµ%Ž¨ªXcxbb¾R9‹ªZfˆaŽKÊ@w ßœx,{| ê!ç»Á˜.: îµÒ–ݱ]Äy?Ûq¤#…GKÆq?ôWPùZ¨p¨³åjä¢ïºŠÅ–y©–û0¥ªñv•Ößm·Š}“éÓCžÚJ=¢ Ùûd7ý²¶Ï÷:¸‰Ò‰WŠOÇRÏëY/>yûyn—ŒG8~á%µ%E!»URÝ­T›‡]`¨¬0uSœ{uδ½È[hQ}>\*È£B=ŒËþÞï*4 <"&P Ïr¼ˆÆ_¼ÞËJ§Œì’"jªÜ3óÏaçË]j¾+ŽÕçÄ÷üµ[26 &]Ja wmh~Ùû–’ßKÛ°Dz”ûyfŽ2…°x/ž².醢w‚ß4bÔ=c†iÿÜ@5v»æ×¼­ûÉ<{àJ¾:Ø´2ÞÌN îéj>¿³uë(¸7:Ú\,÷÷ÎÉ–âÕîGüûà?87¨®X%UE)íVW¹§~pO­¥¡Ì†á/ y¸gÜý0ô‚ ZëFt»>q>Œ3 .ì&ñs}…‹0~ýqÙáW¾ŸÍ_µ¿ÙthÌÙàí«Zúüañ€Â‘ˆëã Ç€¸â}PV&8|ÈþÓÊ'Xw¼!‚'û(iRŒTƹÀdM¨;½º™îâ/cݽµºñÂ̶Q¥›s‘©¶T|iÑÜ‘°H!>,oKoÍhëí5ö9gR­=íA£Ôš”ÔSè ½üÚŸ©Ô(AÚ‘X¤iJ 8ØÆ9˜×íÎhž>¹ù~JYh¬Í:ZöÐU¼FÙ¾Þ_WJJEHŒÓ~é³×Q¥˜¬Æí'7·–÷¸K%°—Jíðñhx:€Õhóÿ{ñPemË·iÚs²ÇMÆøzŒ©`ˆ¹Æ)Ë#—Âî  lsíD:!¯êJs‡’o7ÀD{™¡ÅÏ5¹Y³HFÆëÉ»M­†ÌÄÜv7èµ/XÞ¬U  © šê›âøt”DÇæØÛåúÕéJi\g¥Á8÷Á¼xñ–qÇôªý40ÍHï9²º(ÞóÊ Þ°` Rá|3«3"ká ù#g4.¦Ž­G;ˆ°Jõñ» »ÑõK¦‹jVØØ”"‰Ù³.)>Ž6—èjv*S¶ŠxT§5A³â¾CéMìúD„8¡ ƈ!€YÒŸìxg,ø-ÇD«ïÈûªs˜¯kð©Œ@ßS.ìHÿ80ûh°T­ô‰Ë ‘/OÂÿ,›Þ@–Ò|Êû‡­¯”\@Á øœL£m@]÷ýÛFÝ>"ÙTù”Þâ™R7ö5±Ýx›T pïìñ™Š ˜7nÿé5PÔï~ê¤}}Ð1NoÅËuY‰ûõtÄJŽbtçg•®ƒØ ¢S'þØ•Ú§x(ä84Õ†æó&‰¿©¥a3äå G·…*>ÁO{÷ÌåéăÂs¸ùiʦ@tz§}ºðb¯]×ÏÏDk6" «<[FÐï­ú¯›Ø~Š©þ þ˜ÓÃ.Q0ÉìÔßRÌìšd{2’ê4pØC!¹Ÿ5/Ç•©ûæö·ôjÒÞ^º§Ñ¡$ƒ+‰F\¿ðñ2@ÓIó@»,ï £¶ðó Ïaö¹r 4ÎH!žó¬^Ù<®ÜÌ·N-©êÈ}=ÜKwà;A€Ä+ †Fõ20|ºÓµßÐ^ñÊ¡þ{œžú ¥îòW5:ˆs!1º*G›Hµ¯¿£±C¹;È㬀u>I :LTÙŒ3eÖbþ¥‘;z9ºqs%TÝW1‰WO–òˆr>WA³½ôI›!:¿Î>úÞET”5óûäÌ=g;ßâ8m³Wè3 F(QwïÂ-€8þÖ§·î}º/«—h&;;ŒÖ[8¹¬eµ‘·ˆ°Áôfâ!€V$»Ä˜îÖ0ÖªžXår_{G¥~‹xŽÎc‰ùj}µ"–y7UÐ6Єô^Bb"šé(Fá`ŸÈ€Mø:ù²•ä·UÈzí 3'‘éî×¥B8ætóþEPñÃT3µãcþÌ/ ž)ˆÃhÓØW ×°jýø9 GR³nÑZ´Aë°à¯e$úDœ w¦¬I¶þC ­uE¨¥0ß°a6'Éêd6–@©c˜rµU»-,½J*çUȱNQáa“æuˆG k{Ä‘Ì49Êšõ³Ö$¸¸Ír!ý»ü>kÖBŸšc‰ÏJ7Y\Leâ^ˆôñ9k™áÄx]¹iªD1áô‚ësü-y©p`[zº=Íœ3 M÷ORz^z²Dž3S¢a`r ”Ƹ Œ ÓHeAR0ÖÀ[s ÷+Csà¨D AâJ]Qm êˆ3*x©´ ,E2œÙ¯.SÒ`ÄWLN¼— ÁZ×S;ì8F¬=f9 ‡§±Ùpd#H¯ÏôÈïجmTg ¥¥Ð†ž<ÈW±»8$k!AÞ“ð‰%•ë;ñžHu7\LõHá½bBá½ÏÄx·ÝÑV–@èuöïÞïšÃG8?yÖ?ÀÙpp±`™y3QW÷ÙÎùÜDÖOw*ÂÌí,ýy çõy#žT«Özp&©†G™(àžS2À·4>Ý—Š§ÌÏÒa¶kž>”Ô;ølÁKnd­³<Õ ÊËfhŽ;)‹y]Á~½Ø:˜²'z/_ȬF+™ºWÔÊÊUtÚˆTH¢´Åó rYP½¾…Yã˜OrKºéý>†jÚŒ“.0I*r¥×àÀ9&V w„,Ò™à@)чê^ÚÑ|?7<ª|L$¡”Þ®ó!#i8«ŸE>‚À» iÅÄD_žÑ!ÞL™›1íN÷Õü.ÊëJ¼ðqý•EJu"mIy]ó]Y\˜]ø!ªÆ‘ÅØíe*X(áÞ2‚ºØ–­)¹IFT0Zƒ=È^ù ë)Õ=T#ã‹õ7L$'q|¾ö$£iy¢ /ï‡+Ðp;ñ.ÛE÷â5ïëþBǹòªõM¬‘ò cpn‹b>½è®&ÿø$¤å¦áœúÖIÈt¦ÒÖá}ï‰{¿(Îzy0²Eý¬Db1öù=œ³øLâ˜ÓYú³Ô¦ØÁ/…3°¨õçpíÖ—uGHëràÞ»öIý*>êaJjR 1€¤•„ ’-p-)as¢1Æ;@k¤;ïOõ^Òÿ(T¦ûZà ȫŒ*¶#ÈŽ”•–Ø¥‡ÐßÜÚüjæÈ-7~dø†±hÝBTÕ%%u¤v>‹+ßQì2ÔGAqÞs<F 0—dŠ*t,å­›eD€é4BIq1æÄM(m•þlÚrE––ŽtÐôÁ‰Äqÿá{,ˆ+8)|Sm íüS9Í‘5ë'¾ýR¨¹ç*Ýtˆû$ÓjDtR0vTL2ŠïžAáþ(N©ËLÜaÔ˜«hU¶´Íî¿(ËÑì³P”Oç,¨ØSi¹jyŸowí\º4›„c€¤õ*âAb‘·ý×1gyÏ97ùÏ´Nøªbÿè Ɔ“×£RÒÓèĈ§õ9ì©xèwÄC) V]1|«KÀÁ¯)íûÆ3†:l%µçc0ØFŠ·â¨À^Tw…5‰rì;óÖN¢$ÐßÖ<}I¿Á-‰°ÂÊ`[øAç u6 r+/²É¢QüÀQ©{¯º¢r5÷ûtºJEV§­lý[œ®»œ\>¸ÎæÕÅç6 òq ?_y÷üù›3s©¹I³5pJ%„:ýò»$*)Ý"\UkŸ2yå‰CYvÅçè2||²Á¿±mQf‚¤òÚ>Œ)´¥W$ZÀ¶uOû÷³¼ò]ÌDÐÚ›õòì=Âýr4õû”±ã‘%9T‰á|âÛäáO–˜÷ÝEü•ï`LêoÅ©¶YçžÃÉ>7D{¢Ú~,¬8_ eÇ&,ʯ¸éÒp¹„¬ƒ‡‡˜|£ßã»èŽ*U)I]1‚7Á=LÙY¢¿RaàtÑxj1…’’¯6Š—pöáò(ÒñY×͢܀.IáÛ‡9F…pzûyg¿÷¢&Õ2/~ª[[Âjð3q½—Dɪc0/«˜â£.Â7LýÆå¾úÁ “Ï&û²ÁyH^ãç!ÝèòuÙ¿PÝêF½ngé“ÊÏz[õÃÜö„=ñJï 7$´!Úú»§p< ¨†Få £Ý\–ˆ‘cÏàs€øâÜ›MÁú·ä2×GÞ(&jƒLÒ Ä2”bâÙájE|/ÐçjZ3ý%£HwuÆZ[e¼ÚÝŽê p)ÏÔzY†µ \˃1!Y?ð ;½gÜjtÎRjš¥Úû|A1öSâÌêÔÁŸÕWÉ)±¹ðëˆJX®ŠÒýÃKÀÂêcä…Àýˆ`w)Ÿ'–dkIn»àIBgº’‡')#âç>¨¾‚¨È’$7SŸÌ# Xù?Ï«æmÞ¨ó†ñ®DF7“{Ä×·‰²²Âlœn„²ÄzâQMÐYLWäy«íÁOÏa™XD¨â¿‹-,뮪ã?8°Û&3äz5ýY* ìÅ­X’ÄÓMKÀ¢YÉ´^¡S?Ôs^zÀç:ÿ,ù4 ´j*›Ì@5­·ÏQƒõ§ÌŽ$I@Fñ$×Kòâûm_%VN)èGpf»Ì„¸€æ9&gbblW%þ\~[?5ïGHYnØ£7Æ…zÝ#¡4˜…&üY6¥Ý¨Ÿo2öšiL'›* üšº‹1ž;&ÀFG¼O²e¹õ\Ý ¾»“õߪ…˜dÑãUÞjÝÇç›Ä E8‡³~‘»ÈJˆñ6Í49á‡iˆjllh†Š'ßùtš—ádÙ¨ ¼Mi¸mwÃm_•—Ðw$ì}Ô],´Æ‹ÌŠD‡‡BzËBi»ów›Ë¼K*ê>g¶QãL2ƒ{¬ Ö&Ü7%»±ì|G+2VövÏ9rÜa òB0/ZþNuWüTOeAÛJˆUÖKˆ¯ŸªsPj:ÝÕDQ!hÀ&µ?süUV©no›ÿ-°S@çoži.ÿ… çlh\y»ÊV·Sb«øþß<܇¨?[1›lét*y s¼«lrÄ*bßÔâ­Ã#4†$2‘®rŒMAfìÀÃ_<Ñ+Á0èyvbÝZÖÜ}ñÛ«Z°¨°cÝkÀ ÁžÍ÷AÕÞMu:07e}f!V¤2ÙìÆ tŽ9å]cŸR%Inýdž†Oç7ðŸ|HÁ³LÕ Šf×k^ËpËðán©WÊ×#öj³¢²?IÆÞ)3naaF\Ö‡0èÕñi 翘ÚWs³—{d ~JàÊÃq Z†¤Å ü ú¼ÓÖLYpƒsIXùz"aÄ¿‘W+Mèeº¡uM¼%aÑöçÒ»`õ—8ÍbÜ—Ptª¾5x¯êîòí¸Šú“=:*vrêH3èÄ5PáA[£4Ìê<Ç‘ïðš­2pÉø tÀË•(ù Àîî…]Ô‹™¨Š ôRŽ=•›årIe}Þ­»¹å¤'Û?—Y -qˉì6”°H{ä·aÁP•Á¦½`„œùf$¤k;Lž»ZeƬï/À¾1S– äŠûçóÛV3¤bVç –~¥¶lÈ5{'úø‡3ÆŒE±„? ¾&q¨N"Thd?Ãz€å©¢Ê;ÃØDR1võÝ(ï-kqr¼rÇú¬‰™ð Ä„^JÖýætG+îönü à¾ú*és+ÜæDbHé¯ë´Þ)Éü.¸±y¸­zŒ¢˜…òî%T“ßöÏmQêëããÉH¢ %ĔԬ5ŠÈBà,{4"AXP&äð∵øéeO‚R)Ô¼êáÛ ä€}âMC€)~)1%“ÊP[¿;xfèé¡ù$\ ® ®jàö}n“VÛç1ôÛfd M°o€Y¹1]n3,Wr{„CØ|ÍÊw!ÕýþÍ&Âü‚«o_‡óMl¦<³·1Þ¹ïŠnv„Ï»6+š¬2¦A /ß® Cá-† i–Lñšcë›uPÏ>Τ_%/³ÇFΦ¤ÉJŸª6@|-'³³!(Ò0Â8:Dq?c£F½óV¶²«q/|N÷ä¹®fïSVo“K!R~0ô†È÷¥l à+#ÄZ”•{Ì5Î׉³—ŒŒª{# ‰ß5£Ž–òï,œ}ó–7ÝÈe"rtÐØHo5lWÖ¬N[#C¡Üê%Q w­AÊfþ<îst`·à§ß0Ë%í¡út ºÐJòª‘nÝÊñd²UgN¬[7hAÔâ÷ F¦ðÀ']E R¸ ¶œHçïI®peœ]±ìØ9Ö­ø-­5ü 4÷Ö—þ¼=ÏddžÝüÔ¶Vb ÷Ô3k‰ˆ ¯¿h%ïÀfD@ª†ò"»cRb5î`d¾!ˆäJ¢?ƧXgU·Å%Ù*ù蘓ò(rœ&Þþ˜±•1úv’+âÈÜýØ*®lfBã““ p&ïqZÁ†I›‡šü´ýý@F8¶¾)Ši›Ti2Ä$Àjü&&Ãq}åè[2~s†t‘=-ʽ²jË ÅÊf.Ð>”³·e¢|¶tTÜÅIä|ˆ)™|¿¸Á ´*®ÄSòO×møúèí¿ê)gfÌÕÁ£èNùkÁ¤ùÌ'€’%8„¼M6ÀZ'º`l…Á¨õ¤ºhvC½ƒ†årÍ©Ö­Ö¥;"=ŠµUY´»Z^¦…èï‰ÓÒd¢úÇ°:ݹñ) Í»þÂEK¿ÓQÅ\Íb®:Âä¢Æ!(ÂØïå?cµ …M|# øâwÜÜ ¯@ÇîFÜ„Vþ RX“ëÒb­+öœÂf©‚Þâ(Wâ-û2u>ÏÝ7L*s&Vín cÓ0æXÕ~c(ÐZî6èÌ)ùbf»`FNâJ]ž·fä@ó)E5ºœî%k—ÿÐç]( œ•„XS':™>9¶a¯•­½‹ùE—OxŸŸ-N”^P5Uc¾£ èà¡d ˆ‹üö\w²—¦óÁ2Þµ ’:Í ¹È80ÌƯ¿K‹tãÔª“ˆÏfuä®ôé½=0nù:ë|ÖëJŽŒÿe1µ$ÿ’ËZ4ÆŠŒµ´É;¢‰ Y)ÉóYo=sù]<Å#.ß2›¹\¦Js\O¾ì![ PϽm4ûCÓz7G?J&&¾²OSäöá“LÌk–§ŸüF3Ò'ŠÊí‰êåâ¶qð*¦ú¤üTê{FMãJ9ü‚Æ,°W"ÙÓ ±ar\|Mfª×Ÿ¿æX è )‚YÓÀPX¸ü ¶»õ÷ÑÔo¸ñ¢än!ôkÒ‰l{‰Å¨¸ÿÅMÂzTûVÞBGbŽùNh«-HàÆaìUh gëZ)^[ ç…z¿ªöœ} ·Þ7™Ñ¯c3ÃÈ~#ë1ÄÄ4,£yPGè0ˆ˜Ü½Î@L0¤Î캥ê)åú?Tî(»hAìßtä*ÕÀ•ûZ Lr¨ªR“ Ó`Ìÿ”-f8’÷1àøëw ÖÍ'§æpƒÐE y4ù„v‰@rÝЬœL4r«Ë]¥T‹>×æP Y}Wå—{V¢`ôºÛ&F4ÑdëÈ_~DQ(hÜåºý­8ýgœ#YyíDÕz.m-/ûÊ”„wO膈0e¸ï¸úeа-[r-WG4IV£•¡4 ÃýQ+ÿ–P>F ËÁé°@ýŠš»cvÃã²ø»Âçšlõúzr8¶]Ëz½È7_K/A’Ö„yµùôuäºó´]·Õùë¢]j‘qh9Ÿ4\*±ÞRÛÐ6^ZêLª·7­íZm‘%æ×øÓ¸>nûÈyÌ{Ì­9øOÆãÂÕög1”z<#×ô߯ZÝï'øÄŒ=«P^f4VaaG2á…â¸-*dâI[s'#È”·§PÞ«6sKÔR—ÞÜyœMPÑ ]qÅÝËk®n—/d´±›E‚ @°š,çĽí¥—ªÊ‘¬œ²WðèÜ™$m{˜ö Ò“ü‹/¥½¯›M̪ä‰JŽQU²ê`"ixœÖ™#ge¾ñþÇ{x—×h$™ ·)NËõ¸dB[d+HrH_Ž¸.‡„II_ó57ÌÕ™ß ²Ñ¥ ê™ÜÙZNAß“n2ïJÝ…T3zh€œ‰Í×Ãá–2±EhiìO,R…Àg[A=%Sp=iWs*ô–WµÊdbkOýê\»OFÆNºÃ_ŠƒöÊWø|ùÈI±ˆäw¢UY×Q¯Ækøª¾o8±s`é]á(øë}‚¬Ú¯€]ÒH+ÊÛ]p¥{ÿ¸Uè’ņX®"ÁØîlŽL"×=¾2ïÄ["ÉvÔ#½Ä1ûšÝ1ÎÑX÷¸F P-0íT c›ª E‰ n{5”¦ÓØ>Iü‘ø³¢ÍµÊ:uÔM¬5ëCÕZ ±½¦¿¨gl*“)1„ßOóÚè ÐãÝACÉ`S”G0ì&5þ2=Ï›)ˆ*»ø‘_ Ö•p`‘¿wèp»¼Ò‚ìòOW3Þdu ÌOðKâ>R™̑I½ë£lzhnéCç (?˜Ñÿ¨-BÓ¸ÀNÎY•ìÛä_[$¹Óu,’N< ¸:øsÆWâ?ÆYŒz|ÃÁh>†x4ìƒJX\ã4µG ÄUÂ62jv;Å}˜lϪþ¤[ [¬º”†S–0,ê·¥CenËl³õ@nJ²^ÁÃô÷Š¿c˜½’Åõˬÿ¨ãô@°úà8ãiÔÈãü,»²ÛSˆ˜í5¯¥²×Và¿xxæû$µNèWdûÑzp¿¹„ þù=·(u%¾áGý÷KÙd³Ü<´€‰øƒ`ßÓQ¦&¿p;\|E¤ô·³B’8Õ‡ÞùÄCPS–€/Õ‚Xp_Ì ò‚™¢Gz4«¬þY²õAïHŽyˆ?È ‚ùb JN({*ÊݨZÜ»lNÊÁ§øëëæD’“à0N‘•E8çŒgl þH9'Ó ç`r¶$ÖZw¬°*o°lFHJŽîoÜfb qÑôæÙ¥¾óÒõWô‘½]ÿÈ_z©ìº'Öå ÂVî‚Þâ°úô…Þ¼A}_ù€Ð÷´ïY°nлD ¦|lj.4#ºÜmä¶Ùù‹Äe³ì–ÌÈ&¸`—íŠ{ýE_”ï» …ò{…­Lr, •œSQ¨§¹Åz,>Ð}ØKº]¬2·PlScçt“"xïóƒö»±JNs©zýO}¯ïXìæÛäJ¿j ô$ùa'vªäìxÇ7¨DÞo1ž…*Íh7ŒÊ91¯¸±?ìr 1}ñ¸Áæqz#©P«-gpݶ%5–Û¢’½§ (4}Ã?Æcò1]š@ÎÏÅY”œ†%^ñ«{»—„R³H4öz)''ÞˆòŠ'B¤5þ‚c]ɇCO"wì(z¼³=¦fïÊ¿lFÆ~4 ›‹]?Ñ5øʺ]œ—ôŠ»'Wì¬=;Ú/7Mo鶺ó;¡.g ž´¯Jì•œY±ÜùfÌù|é5'ÙÊ …åNÒWû§Õ¹[è¡Ô$Ú ®KÏwø8¾±0<½Œqæº2Dz+šæÎç6üòðHúbò ‰ðé“YO¬'$¢ÓC•ÉwLw( ‘ð’Χ£ àGÓ ›vðž–õ"ŒÔi1•k‡B–>Pt&ÇL+G]Æ\ƒnó\Áןuù{Eß·µ‡Ðå&2@¢ä¤>‹üÀÂú¦N†XSnŠ¤ïƱE—wïHáz._IK]¹^Óf:ýš–úŒ3£5îJ—H¦;Á<ÍoSdä[ßïÃŒ3`v‡Oe¬0¹>GŽÜwy ˜ñ—ÖQ¿ñ¥®B®«UiÉ/3BÆ?΃Ig|zGÔ[)ù‰×&öœÈ óнM­ ™?cÐÛ#Oó`uq´¼˜y„þ´Š§¾í)¡ôÌË”61Jç®>ù-~ËŠP†å8“î´TìÔÚ}`wôÊ“ùÒ|ƒ°;‰QcJ‘ÞðÆÁ<ëŸÊ™Äû Э>Z[³µn; A¢º†Ùå3j Úläy–$<²à×ÌÉþ.ô›÷ÿµ1ø¡‰tþƒ¬²¾˺ ƒK¾%S€nìo^Ù=úùE4‚†™Š¥B {Q«Bs­÷Ž‡%p:JXu N/’ •Í€%Mñ{¿©Ü þ\‘+¦›`a®t‹Iu‡GÛ%›Ãå¢B°òtÆÅ¥ûéV—ÜZIbk®EÙ‚o|Ø9J !žTƒÓÁ{gZ¤úeXõ™–žˆÙ¡”§ ”ÊS£)GÖ{?ô"X»oaê!³œá¼±ª¯ê•èç7é¶ÖTm™&–w‡c/¤Ë-0ÎÞüÙç¦|4Ùì{&ø¾YÛi‹Ã”$ Z _)‡ÌCxj=‚%kxؘºý¸t~¸<<‡1Nž „Ы½ƒÌX¶ÈÄè^¼/³P0O¢½¾U“³à\è­Ë#+£  ƒ1-ï8÷¥ ü†é~Hêlôxª/öe<:Ðʹ šÖÇž+—Jõ›ó8€°Ï¶ —90úǽ<èqŽ2ÙªH±W´ÊÍÒ:âÝ°ói rÑF¤ÆpíÛîèTZkï<î· Ãĉ+ŠóÔm3Ú¶$–$$B`§ÜQ9v%0=y–¤žãŽ„4þþ¶hE1ºJd>ÄÉ)ÓîØ© }RQòüÞt¢§ôR{¢rŠ‚˜ëRôh\xWd¥qª1ÛTBQ<-èÏöîù+ýÿü»ˆÝG¼²?Ùxßc×zsÄâÜ‘öÚG}"¦ P»o7{ýjŸ9O~Ñ–P­7—Í€â &Kpömæ?UVÏ ·ŽJ¹‰bëxvÿ-zâL—t¥Í,ö“ç6É" 3ìøúG¥)pœ´Jˆ¥\ý÷-e¶N“bîX[ï2´º»ocþz$„Ð4lÂ*˜‰¶.ˆØ\¹T1\æBÈïvÐÜ·-¸å«¥mæ×& !G'5]5ô› X'ÜÛ>êtŒþïˆkwÖm˜+wcš>˜ýÒ¯¿âºWŠjÛn]S+v4(ˆõáù·…H N«›ú(Kb³Z_`÷ T?o¸â–YñÐE›-–½˜¦ã3yœ輆4HÇ(q¤þ³iA“n6%q}Êu6ŠNg¦Æëã§Ú¦6‘°Ñž†Ú«K„40”ê,Z[(÷KÛét”íZò‚~Yê8ñŽ_Š¸é[2(oïÓAm _ ø笅BHÞzìzÊõv×´kÔ9y\uè˜`Öh÷¸÷²½’ç¶ Ë‚m©f%½áŒì]{Ôí(ÂÞê]:{ý±qÔ:Ïþ¿¤“Ž¿&}¹Äx\b–5%Pxíd¾Ô,hüÂÏ€i€Û Þ†ïÜõÛÚÖ‚¬ÁŒŸ × U7Û³ä3»þTÉĪÕÇ~”æù?ZI²¬OÚˆƒ-Q «bÙv—mlŠ$»Á*ëžEÀ¨mþÚ!û ú+œóŽ °´÷ÑÉùÀI†MKHOTye4´?¢ V‚ꯒíƒÜzXw¯8Õ—æ èì š`AeB´ŽÓ† د;„¬©ïRR`C˜…yK%UůŽ‹ÀÎW]âàɱ(- A•ššØüò7‹{Ì-.w»õ ’òÙ|e¤Úxí³j VPšñK…ÕžO´šåCf³ÍSݦeý| » YÒƒ%?0ÐnOìSãÖÊ¡)¥e$k}˜P½xfs¿+éԩܸ ÉŸ“kŒþ…$×û‚h£ð 6ð`À@ë(¯ ³ý¡rÄÌ*¶^ô°Â»“ÖURíL7¦¡“ô›)6Ñ@l´œj㸅®^=<¸¼sŸ¯§7f$ˆÞ5uü9k X-†4®W¬ºr7«pß Áù9„èÊ~‰³ÓNJŸ¼†²š3‘ÃêVQÓH½×4–sö"¡Å"@€BìÓÔ $Ò–ÃA|È„Ý<Û²¯•U((H,Pp»ŠžIä‚ýªßÂŒ”қ*¼DCß±³Ñ Êpý²]AYæÆpuÈ ðö«&'µ4Ab/Æx@Àzè7ò¨°¼æZYŽs¢ +ÿÀ飼ÑöÆ=úÝ»/ÕÖõ¹s{¯úc”Ç M¨OKù/#y«+Ê©Ì–zTa²"ñ0y®­G9¦4]Íbh–>ëŒý‹gÓ’™—ò«ÏðÓ>ïhøVÄáY™²¸°%OˆÓ<š/º—¸—k¿È®Ç@’õÏ€FHQãÊN Ñe¢xG Ƙ|ø¸Mìá™s‚ðÔ»²›aÞ ‡3Ôpcß•Q![øžêô'ã·–îã•P‚'Œ§´úŒ!„’óÊèrJ R„©ýš4Ö SS"I[VokѲnéaÀ•Ofã•ççB•PÇ%´˜#aÓÓ9{òKnÖØŽßÝ)Íœ”>1ŠÄö¼ÞKHU2J×v|-H©Ò¶x!écú QA[Û©-Ö‰š¼å…Ú¬\sá ›ÆéBp|bc7â¯F-‚E-Ï+¸EhV“\†}§r‡8ÊÿSÆ95WÂH4¶mÛ¶‰mÛ¶íäƶmgbÛÖĶ¹»_íèç®:§›ëår€Ü¡!8˜î²Ò›¦ø'øz¨“³üøÝKnÐõêŽú$«´ºÔÌæšêíéd°tîþ‘Ðp}ñ÷®y´Š&§›¯þÅ®Ê:ö<`—¥9h§ Øæ®ã1Q.­Ð“k;êµ£E;ðøž@•I·~D×'w”"ëï±ÁäÂ+5²¶.ƒ;Ø£_ç0ÓýóS5¿›]½õË ³¬—SmjNšàö;ûë>£ÏÛÒ5¦×Ûª2x+l+Nus¬86I ¢ã¡Çç˜ßÍl÷u:F§c¼3ú-Šøtú V‚HgMßí¤PØw™&ÒÇ•ãÅÏvªL6k)žä*I«._xåó{q|~©Ûú°­~ áÑ86®†ñÀäãÑ–ðF¾†Læá_¥±L\&Ô)ýe6Ÿ5;Ô’«"¼Þ‘eÐz,L&S=»>Q¸²à)†R<£ ªë¹RKž›85ÐB£³[P„”vcMg[ü3>U¢uE‚]TúAÍôs(3ÜsÏE¤þì*ýñ ÛÈç>ò^H£6WG àú3GÎ6L>/{±€xlÚjÓ8×Ïã¾þNvÒ{¹y4­oÜ­¼!:ÏòËB¦Sf‹ôÔ_Xĸ\ÝÃÆöÛäŒP‰¨ª¢Ù½‚†iªš§Jµ´Á[e®°ŽËmþ1ÂãÑZ;•Kº 5¹c¤Å‚ P-Ê!áàœh¾×6nçP½cÁ#—BÜ8ÆýšÆéH_öäþ,ùÙþmi 7Õ'ü» Ëâ¼RyL¥ƒýÑn¯,ˆ1/PåÐÔ!ÿËÖéiëZ0pŽ¶BžrçâþC»ßÇƳžjL-ì†ZÒ€EdVîU0ü­ÌmzBó·Ù"ævý¤"íâÖ4åßµT²eg¯j0!Löæ< I8cȣʹ&¥k@]£ ÛbõýñXöÈŒÂ|#íôæ vMÇÌ:ã»$+õA¬·Û¡šG_û¦ü“+«lÞ=rCÇ ,ãbä×ÓÁ#ÿ2c"õ("2²þz9gDEªTß±||~¨ìJÊw!Cx噜€,ˆ†ËŽ™KŃÂâ€ÊeƒŒ<¥yy'ÚÏ1c»snS5Gp¦=2ÉT­Pβû5á"ãhHm¦¹°Î"^(¿nljw á”á%ÿÆcÀ[àÝãÁ`Õ#M-=Êì ]ÄÒ ÷;í”—u=Ö0ât¿¾œoPe”†. Ë£á.sÅÁÆÞøï-àƒ¦â ,ÛW||?§sŠz𦦷3ƘǡRÒ=GÞü"­$aL5SîVÿY:´‹:š¯¸W­©Žð WÕ9Rî‡÷íµhqÞZ Ç€R:íGž‡xCE¹l¤äfd¨(½ñSDÝöús7ff"¨7“pd$æ‚ýîôƒŠòýÁº۷“`Bª-¥~ëQi%DÜ{ÐÎ –? ‰Ò× ¢_¬a3c8–#| ¢ÛöMvI°6Á0ƒ¨r›n–‘…ÏžÂC| «‚º‰Ìë–Ru:xwR@ò!‹3ø…ñíĵ€G0 ˜ª´< „…õ‚¯Hâ‡~¨2’Üå;\5;,¡˜yFQ4Wä%&‚B©»•,e®a¬ÿ©A|ÜÆGK˜ò/®×K=¼‹%0£E*Âùº8þBA’‰‹Üõõ\üÖܵ¼ôS'`Î隸ØÌáªd=ÿº®”«˜ï°ƒ¶< ´Rø‚(£‡œð€2ñcˆY®ÊŸÑ¨÷F"'Öl'õ:Äåµ¼6T‹§ú%Êú8»:X•â%¾üý” ¹’ûߧ¯2+G¬˜deB[4EóØ çÛSÙKaÿË¡–©w$ã 8zšî(‡äŽÞ²è@m(È) ²g4î”:œöJŠ‡ÛBm4íò`Ž×ѽâ²(¯¢ðéäÀ½Ø^JxóÜGr""[ÞXY;}Ѐ±œ”k¶ ËAšzß(ÜŠè„Ià´ÁEDU…Ñæ̨5ìŸÈŽØ­mñĔܖÀùjzŸ;È3§Ó¿ªÀ±O£'¥Ff¶»†:ƒ™æ‘P°ÞP®i†Èø:}+¤ax5øå.º„b#ó`˜OŒ³Ð(ƒxØRJþ‹ÀX ²QŸÖ×>g?ÌÅàxõwý“c’üªëÅóŸj¿gX'”§bsÆÂ^^Ü:ˆ\Q@bN¹g45kx¹„p7ä£V•LörÕ ýà¿mx´¬>îý‘‰%‹ï;c²”¥ò6Xÿ¦íÅ,lÖÝPé6¦)•Î€ö®yêÿqñŽc>$MïïÄq:m!ð¼fOdÕ¶KˆþK³£4´þ™xØ•):`gþû­Ã ]§Î­TP'¶*_à}Å%ѫȨòAóÛR1&qmw¿ü…eã§wÃÖªÀr#2-ÏÁ)®yûQ¥µ‡~ÌHZ’6ø ”¯Œ'=ßTv|FË*fá{ê<ª ¨M¬÷Éd£^}eVÆN“œžðe!}¢¥ñ¢`ŸM<Ëšd¿ºÙeˆ{§­ðc Í‹Á¯Y/lµÜ ®HOðÃ.ö‡[½®ãŸÝ¨îA/iÈ"Ëö“u3ý¡…2«-阄âç1,½ªc‰äΟƾÓ3­»ÐBArÍsà9ã ÔÙÜ)aâô?[é•ÞÃþ äºþë!Èî¡sû”RÛFT&èÏPz2å¹PÂ<å™–ã¤9ƒ—} «^b}·û¶†‰¤Ûÿp²]ç„â¸ÆO*ÀKiùÝbü;ól@»!þ%!VcFÂöa¦¯r6)î½Vaöÿb3|ßø§O]*øpƒWÙ%tÙR½¿Lw}Ïç;§ _QÔ\>o»ª»ÕUPñÈhÙ,-ƒZ¤î|q«ª0¹ê#빜Ô9dw4ý©<ÑOˆ¨û*÷HéLK#¹Þ€d^ìŸ/Áà4‹,–]$Æ’ïc7?A±( |÷œFŒM;|`¢­RQKÞ@tË^xhPùY[mºi¼ü“É$q²þ CØ>•­áä·Åݽ©Ý´+~–áÙUT>G(Á¨ÆŸ½ ß6-—Qd“—p¦QÆ̃³Z²ÄkxJ[Ýü$†-L0¯¬v·D|‚—ªº‘c LÜà^ÿ Ì–åu¼ãuòõá<½:¤(?ɨôw¬zÓéáŽ5’=æØÌc]óèÉYv¤æI³÷ç²…¾ *…K×ZÇOy7ºŒ  ™ùϧÓéÂV—ì›}gij¶oÀóöw¹.ŒJ Q…A\Š, ~“¬Bäêi?ÎJ¨2Ë•4ƒªú•Ɖ=Ým—™uÖ]Bþ!Ñ7–ØØ|ùà· wã(Ú€N~A ŸmXûètwÁŠÆXôa¨ rêó@4újuôEò/¦nP~z§˜ÇÅ´.àú¡Té&“`ú¢Úêø 64¤Í\GÒ7Ë#Edw@£Á?ŠðûM‹‡Ä"ÝŒèO’ ‚ê¹>= Àèìù·†'`QFLÐü Ǩ`7îÍ”)lõ¡y]T9YØÉÁÿp õ`-2Ñèe-æl˜žrÖKV¤\þ±b» Z³ÓdâìUŒë¯ÔLaƒ'έŽ*ô1k:fdkørºmýظåfü¼¾Ïc#ÚNSź8¥t39I#Fº¼9Mæo¤:†¥’3-ŠK6”À”¨UÆAB@]oL¤ŽÁ?¡XN<×ój© ïHF© \`O˾P‡o^gYN©nÛDL{΋ˆöî^&t¾Ÿµv±‘ ´Ø&a™Î²Ã”ÂãÃUtæ›èÉôDËôâ83ÍÄôF`›}h„]ë "»ÛŸÍíÏë ~´DKÚ&$IÍLž. ß`dS&÷„â ÀfÕ[¿æ‰*óÃæpù:­9ÍY ªphúM^ðòB\Mç…êûTÚN®#òº¹¨ÊvxYý!Á¡)CV[•iWËã ýjï ±©´E¥_§cŸŸíæÛŽ EŸ®çV„Aü~‡ëG¢>]¸;°»äÑ,µg¾oøÃ~Œß7í†Óé( 3ÄÎ cIQÀ½®€übEî†ÔÿˆOÝSpJ÷«#ãõðéDS{'i°SצVÄC›Üìiê©v¿©uV1ŠøH©%A|e“ëd8Ë_\-”깘é/ePw£…Ò'ŽGö—°;w ÉIÒù³Ë¾x8ŸÌµt§{íŠJV7ê rõh«|6W,šó¢Š¹˜M Œ0`“ßXÿÙ¯‚eÀiUÈñ¨#Rû8¬,4OÄ™Ój™Ïl‘?¦¤@ËsG;Õ¨†ðãn:6^Ä|'æ  CÇá“d¯ ó6áƒK‰ß@‰D°€z¡A'ªiÄŽéÃrx¹ë`FY0¯bU–H^‹Á)díÔÁsÐ*ÇÛM’²'Sü×ìþγX×)¥»…i}8ó¡Ë&Ÿä…=¹Ð(Ÿë–ZÅ„ªÐf ‘°0×+7cO·‡ÙÁŽ÷D?t£•&Ó«³='Ïã½s’­§ÉÊÂfY2²¢‚Uæ­ÐyÐŘ({°Ó.á*;^Õ@áÐ`frA‰yÿ$Ô‰sXEG*¸A¤ïâ +¾$­¦±<¼oE¤ØösCL“<‰=ÓmÖWëÞóSØiå`ÎDZ±WËÄ« 3ƒÅ6à=tËq™˜ã< <"fDËE-Ž–|ÐäD€™ãŸ*ŸŠ™Ù.ù@ÎÓ>ßü‘ð¤K)[©ÃÒ"µ[{r°|ÒŽåy„]Ù†ŠWæÛRïé3ÉëÓâ4p}³†ŠïÙ-°jä<¥ã6ìöŒ†{Ï?fÖŠË:GcŒu͈Š’ž:~Ü©Ú¬I*q@ÅͺP;Í…ÈÄl$mRN­è°‚bºFƒJÅß¿OU1qé¶e\êaNò„v#›„ØëTÖ®š¥“8Ç‹¯÷ž)ˆû#:~Ȥ,ÜV€žï²Þý§`ù)ðòÆ>ÑÛ±ðßMóê`oï1m%šçÚÉÓ™+Š´‡=nÅg  ¨ËÀF=ÇÛØ.ñëüï€(¯ „îΖH’¥HPÿÝÝï\“¹1Ù%Ú&dw_¯Làx*˜ŠoÀ6Óh4ÜDL®éh3ë4ꈎ+o7Ô8ùéÙîäΩ¾Êsúçè¯={¬x1s€¥ˆ0×WQTÚ‚¶ªâŠž˜ \r[°Êå›'˜„!Зì+®›Ãa%úœ³7)v£ª£B¢Yã-ÚR6\µ·í P0r¡ ÜmÜ_ò N«rzäY%ó:Å3jݶêѶ#âãv˜>óiM›iø°­’¦4\³°¡ž¥Î^I1NzQhÛÑ ÐM£L>H‘þ‰Ä&G¶i°mø'†F »Ä8*>—oÕ»\Þ‡ñ´\™ûh­‹u×›R ,@+&&iâØ¡~ ‘âeˆÙø±¦×ñÄoƒ­ß–Lù9b‡è¤òÆ.’SÖæë`¤„ÃDP8 y:´MÎ,PmÖº¯Ö'K(¸Ï"º ñè䣶 —‡u«¦¶§Å÷†¸ŸåxuÇÃí#’JŽ˜²>Xhœ@[K ßÅöõýÈ{ûoQö xEã£pØ*Ò™„®ß¡’h-Ë—ƒ/kR‡:ÌõnK jç!Õ5ä Æ{ìÖéŠ 89ÚSmܲw|x ·–9ñ[¹ï ¿ƒ`/È_¥™ Xj¼IÏüåævëÁ…:½vGg3 à} ÿýn¸ilÑ1nWp ù˜«Pá–ÞU¿àåjddL™£ Ñ©ñ}’Ó¹õ]s‰ÕrD‡äDE„tF‰¾¥Z›Œ$9‹)¼¨íÇŒ×Ä,¸Ý“è744×ù5ê™8ˆ ªŒ™óD`ÔÒV™é˜ó˜·|÷” A‡®mL\ßÓž‰±n a“Ø¿>êÐ8mFë÷€¼OògŒ[¯EŽ<ÖZe/u¾2;>õr¢H³xí“6ñ©‹šû$o«—’tK×DÑe¾HëùfŒè¸;Nq ™ãÄŸ…¯ 3víì#ÑΠJ¯˜aAt¶0—ì<0Œ Š<2$ŸY› \5¬¿»œßð¿wœ7ŽF©ãê¦ò^KÄl*ì¼,ç< ÈÑ[CBÇB@[Ù9ÙJ4ZøV{žäŠ´,ƒ¸±!ź@÷‚—ÛV‚Ñ«.ÆßT»<íK‹ú˜ €Î¡koŒ<6öó„c,G»°x ÝÄà`°«^ÕÁªƒOP±­ÖE‚YWf%nõòJ^·Ä’¸8êÙÕ¿Èê„R]‡u€oL Ž5Ñ’%AØ•ÖÑäW¶€i·à‘±¯.B×c€¸«?ÔŸk¼©ÑPÆ_ºßSdzí'ðÏf°®±µ·ÄÌœ“º×s ­u½]9¤µ›Ð}_àì’cä ¡¬à_¯$ùÌ\¢›^ œ… ¶ùXŒk$xòdæoM»“A´Yx-;jƒÐ8Ûô·¦JúG‹ŒH3p—û‰×Z߸NO^˜%Lv“æßb13ä´WÑ—jP°F*A» ®ñ\°GÆGt³6㟠˜@ßhQÚÜñy[©CR—OÌ]Ù¨<ï½Õ«Ÿ‰J”0þkþW;Ý¿B¯Ýù}†F¥±éG4 †<Å[Qÿ6q<5Šˆhî"l@—%.+F£ñÜ; nRº*ª µEãdŠ½óôwHµ;$lcJ¦–´:—,îY†–ðŸ[´J þ­ëîÝÐÇ ¯ Å '”t—¸)n’¼vI]$ 4ç߶tJO–PœÃ~ø/ë „×tse]©[–øsÉü宿àÌvÀÕhÉF0Sî±)•ÎXGÛÉçÝi¢;03¥õM«>ƒ¼Òó²~ýÞN¥GPŸp*ë™è53 «gW•ë¬U-«ÁNñoî†Øv³Â+- á]rˆîXÙsÅ5`²©47aˆÒçU’# ô ÇŒ7Á”R•õy³¯žŸD¼¤k°JÄy°øEÂ¥¯½™´åw‘2iÝÉñƒ·l~ŽŠh¨x©Ê⺸1æß0½eG²B]9Îq£|1X½ws·ôr?_ŽÄùGœ¯JÏŸa;¹ð1žžSÑŽêѱ=Wj­)Pv™t7*D²í¿õ}¿ «]]DÖœ„_ãöt+³åŽICjë=*O—Á¾¥:¸¬v,DþÐ gÉ´˜âÛÐ÷vG ÿ[Ó•Ú‚sêãR+ªP“Â}u¯ÐoZÔ!¨ÅãŽFòêšÉ©Ø Óå2¸×é^W(ÓÆ?÷4@°³²MVÕ(‹œÄAA‰ã7\0 lá‰\²NcÛ]ìYH¿8™Þq WË<~|ý€¿ý[ö×Åjiù w¨!ð>J*ãóTñ€ª5% hªí_Â’‚Špü»–@»`Ò8—ˆÐ ÑýZú;º–(¯$—Ç3hÓosua9Ð÷–ôIêXšÃÜŽjöTåBµT$Ž ÎŽ°#ñÁ=VU& p§J1€ÕoalÔuPöª®*ÆEÜ4I©Øª±q}i1Fx÷'ìá¦uH®öX >Ëm¶»7»K¼‘+ݽ!@ÚUy4N ¼KZ 78DHîf„GÐ"¦³¤ÍqŸ·u¨ôÙ.OÒ¾Hæ6*$Ëäbg¯²àùg ¿w_^ƒ; ,„¤«›(/O¥vF¢á9)ÏUƉp}ôä´is×Ê`ꦶ‘ÑÅêÚÁÈùïoªÚñ̲º›õdµd Ñr:ZsÕæËB2êÔ„EQÍžyo%Ø õ>8?GêÂÀÿK®ŒÖ—lo qŠ‡{©±lZ–;ìdä0 •«O#þðÄOF„œÉ™»ÅfµAíÀåàñ‹Ôºmé…3ˆqµãŒã*<С‰aàˆ>>˜ã¥CÓF¢­($8Ýí;Ñ.Lóî*Õ}ŽM’±‘Þ@óöˆŸ\sýã®NÌ“¢‘Þ`'ÂñÆD³V?Šñx*` ±üPN»v@‹Q²Ô¤ä4AžX£–tô@Û¦õF)5K>7½"Bi'Á߯ìwF«l[AM·9Ä¢©_±ϪOÀyôØ}´†m®©ÀàîZÈqßÓ_»µ´íh½1¢/Óq*pð>]å–?9WOÝñ_° šî.0×l‚!{›YÖÞ¸Ã=Óî0 ÈgxqR{¢½˜gu›:+.;ÙæAžªÙ,ÎJp›ƒg´S•Ø1o€øqˆða)ä¯6š"ŒWuSé‡[–ç]ivLm ¸e!=ä"ŸXѨjÄÇóIö¨Ìþ gtþwRýú…«Ÿúeý'èùÜ={ÖøçÕ :^˜‡J5¿\zˆR³SŽ©;ͲáÍ}Ac Á/uBA¾ÄËÁ‘C4–šÂV„0ƒe'‘Wã`Vè'Ž|óÆżп҂ÊÂcÛfÕ…¦ûAÄœmç|{ÚéÙÅN,‘ŒßN÷`šžãýïà3ßõ›GWˆ^{P(puH«u^ÇäMO)(:¯$ÉAôa ‰ÜºÃx–¥YSñ"ƒäÉi±LÂeºR¼»êŽ™>ÎSÅ5x„î$T‡8.P-ÀìUFïQ6”€ñ¨oÔ„ï«òàRg|®Ã——Þy¼Ã*̱Ç…8H4ŒG H­(ú-SqF¼c ¹ëÞMǧ# ËLž-øÂàm_"=~â£p°ì¾ ɪífÚgT°óçÖOZþËá Š³þþ]l¿x2D+œ¨$v0Ú1^Dð9¾¨ñÿÖÚàƒXZ–,ÁCH"ÉeýºÃÙÊÛ8˜FóÄŒ5-É¢à{ö9SiV³^™‰[~,SòÒbº‚>uMKTa\yú²âW¬]D |%—H„vøáèžDÆgËqdÊC@9©ªá=°cOÌi±Ùtr¨Óð}£ `cO\ XiïP¯jlÿKäùªy4~ÀO'C€Mª)¿}G9¶*Í£ýªä“°Äí4àÄÒ‹ñ®ÌÖ(·ÒážT_èaÖ-j¿c f1M¨ž¤#Kpk5hÝú e½«`ÌdJ§ao¹!¶åÃB ÇÄ \úo?T|j*ùNR;¶-:Yj âô›BÀ¡Áœ·ð¾¿ûÁ žúçÍ¡;A_ý<(ê‹Ç̲ºÏ$ñZ+âè®úê“Û5=¨÷VÔL¦hÌIeÌA®˜—0^JOBÕG7/ì'³s ÇϙܤØý Ä¥Ñ_b¡†-ÑòøÆ·‘1öi±ÁVÜg’vË<plõ¥©p¨ÃZ[6öÈGQ—úu@¡Ó§WMÿî9ÆjÈÓN#„cqdCâ×ËÂ~ kŒ[Ö=2àDIg+º–@Ý×ø£È¼Á…h*hªü;[8Ú{›z[J|lÀ6²J›ýÇk/™/o÷` ¼°ùmu>§$Í’“àÔÄÿæÉG½¿¯Š ª——ˆÅ û=JÞa¤Ÿ°üZ žræq.·Vs[ÿ[‰º#ÆÙ Cy”qÔ)Þůª¶žd-ÃÂüù–4¥r¦¯ >@½›Jg†9ËF0–°èòœ Ä’ñœ…smÑSW`4*38ù‚‚È»÷LÚäÞÓ]qÿ%xÙI1ˤÊÂPiNž– \· Úœ‡OŸ¥s¸’n”js=¬˸.TƒÊJ’Çžh«ðËmÏlM(ÓSÌýžEóf®-̹셖äç¤/¬$k>£P[ì¡p,ÝÉÿŒ0.œÃß'÷Þl³†z!dNzy¢¼&z~­,Ê÷•½¼.¯Úˆzý•©=LüŠŸZÎ-‚†ôi.“}´j]É¢‚y]7úª2’5Î ’}—1é¢Â,-D™›âBLJ¥(p3ôQÇóù3;kˆþÊ'ΪøÜ“©1Ðâai¢[ çw^¶ì,N2¤­çú⊎ô¿´‚Óh"sÒOÞ ÈnÛ:¿b!þ!r÷•}ÚWãle›3 ªÒ’¹ùŠtCô~®.cc‡¦$­SG.ëÝ –lÝD¼WŽSÒÞ 8z¶Næt߆^Â|ÈzLDD}°Ò0ú˜ê|xO…;Ö"öÖÎÚ„ï¢Y;ÈQ[„¨ÔÏ‹1I) æ šn3" {— l¼Zwq¶•ú¼â+â_³ðˆÿ)üûp””Ñüþº\ÍZ§–ŽÚ Öæ:õQèÖù| õ}‚i+ÿñàM{<ÑÓ®IƒeØμqø-øí‹%°ýq”[9¶p">²©Âqš¬)Œblï|ŽÃ>–Mqõt…4—ü½N.™  ÓÆ&!¯WtÇ{q\!¦\\ä GêÈ“Sš¨Á{ÓÝpÙ1Ê\aŠÏtywg¾A²P^¬Wd ݨæ$\Oò\5%°-|å|^´e¤SÍRsÀæ«Øþz#·¼~eWÞaÜ(ÇCßîÄÌúç›îFš° !"´BH*%õŽäö3›qvæد)z€J5üNàä_sK6êMìÁ›yýÆ¡Å–)Dפl €f? hªmlz?bA6ê€Çíüm™;‡Œ­ÊAQ‡ènŸš –ªcìwª¾‚2‰y ·V†¬ÃOXin·FCÏ7ÊÐdŠõp@鿱YĬٺ¨|Efcƒ:+íu DQ_uN³>ì!;r½$=hgOl ¹8í}úù¦GŽ¬9 J¿í€Ÿr¥¼TÏš „fׯ7 ¶hQh,QÔVAto¤å¤}?±­ñ¥ƒç0𧽤ûß¿C–œ%( HH­&F äå s‚‡Iš­ePPG;(nèÁ'æß|;[…zàÁè9«xlº{8ÌÂU¿Iõö­È¦3—ÈS“s•ÞuF:Zñ³@ê<ÒV>'ÊÑðú‚p VãU©9é’Я,Šk4Ï÷$äéxƒ¼ÖE¤AÛ•Kd·Š¢Q”!z Ûúøñ·¤œäÇÍÛ ÚëD¢<8g?ÜÛàëÉM¡9Q[K´TÎn<(=¿G~û:Ññþ¡-êÃu[ñµk|»¯gôìÞ­ú«bgØÊYð4Ãz‡¼$6D³òÙÎ-é;×Ù¶Ž÷É*)nºMsËÀÉÇ%¶*ò¹–FϘ¶usEF¬®±5ø9/ hу©øÆÏÅ)e%&ÆÞíÑJäÏû„-–Ë»`2žÊJ_±oŽ%G3U.‘èø);´ü¶˜YIîÂÅdÓÝJ ÉãIû#ê# [=Ñù«¤Ù…ÌÙÅôD@é£Zë­÷"Ëè‹€àÛþþ+[eþ=§ßª1å|P‚-!JE¥ägÍLâ%ž·?ëPK0¯ªèˆ,Ê@à:-™ä>OÜî$ìic PÖ Ïyeæ¡ä¢Óo—̲ömÜKòç|à¡å|ËöÉj#Ð Q!T«nNµêR)ÁÐ…•äÐ0M?ðßJ°Š6©[4”"1C!)æqPbóWTò¸a’½åfq´ÿ`óèøúå1„ƒ™p¸sj0o:QÙŠ·wÎú ‚Šá¼Úî³²@(sôª&ñ¥x¨¤&2;O°îù6uÐy~îrDç»Éß?Cãµö—4g«/7#(q¹ž +¡+¹ ¡÷92­ñú€ ¶tOZ`ÿåÚý÷ä'‡9—žÔ¿}™€A€¬Œú e³š°ú9ï´kÖ[†Á4fE‚˜FÜ®ò]Š¢‡:ž¨tÿ©Í/VÍ8k»Ë¯—¹l’¬¢êfÌ*•ãÓ4Tf8ùºGV½ÊWÄÂdÒ=ê²QXæÓí¸¢I꽶öŸ©i¼6Gñ¾g,¯Ö æ6<6ã¼bÒÏ°á«÷ºA|Z…h£‡Ç0aìô0jÝjÕçÎif—1¥…2·@‡[’€r¹nìWsW¨X_¤å'0d/LI²Ä56O½R¿*• <ë,‹VNGXô'É3A-xõ^¬üùûV „Uçè§^dùÏñ¯Q¾öøùë¡.Œ4ŒÕyè¥ù©vBÄ'ÅXŸ>Ùó8ƒ!é˜@”QDJ‰¹ŽÉ¯ú¾ïtÅ!ªV… ¤ìÛk*ðí˜]…¶øæ}›i [)㸙 g4<ÿ…mÙô‘Ñò”Ö‰­§jf'óíi£Â–P|%1šyï6ÚÀ$¼÷]Ht˜nÍÄM?Ü<0 3:„ßQA§hÂÆor¥Vxªw¶ø‡&$Á@ÑϺ×Å9®îƃŽ@eQql¸{’ .çJâo!tƒõ,wõ¡^Œ^žÍqS–‘î>Ár”'"¿eõ—óÏ¿1 UÚB8$’èeql³D\¯$MF>í+™¯g‰€ÄpëœT¿¯ùÞ>p«óÅ´GW1¦F ¯’¯”®5Œ QN9`¢Îí×i³7Tj芌Y›P½Æ£lè!Ýž1ƒ“Ø"(ÜtFæl8D“U® Â/ðmçý¦O“¦¦×=0-º3NÅØZ¶>>>—í‚@ñà…SͼWI£?JÞÓ–áiÐQ…±U5cù&ôU &Ñâ¬4Cä`_†\%]æbÅý,^‰ÆÍÃ÷°!¡·Ú‰Éáí7]k½qhL»àjõPD\_ áñF”´}ÓÝ·uèäx>9ù3`^ Œ3r H2ãØT ¾A³ï龓öÔc,-ˆâ÷²@ÁuO`Žf=ŒH¹ÂE§ïFçîÛu« ™Cº¢½¤2G…ñábÄÞŒ*X›žéüí$ŽZ*¦(u¨ª@2»c@ >Âñ*!ý>VÊEHÔ.yPQ­ÃíÚDMï_ÝbO\—÷aúY¯Kf:`%¥HcÏ7>¦©²ŠTV™û:!›V¨*’cÖ0}Aî€vk]ˆ”â^N·p…ïÊ̪ï×1²S >cì´’¾ëã!Qöüh¯•š+wlû°šbÏ…¾Oû6³õ×R¹õ-£ˆ¿í²7áÄî8™Ü;ª¼"²Ý‰ªŠf= Frœ?»VÆEE&mw‰»v¿eþbF…·Ám[ŽÝC×\Êæ”ñ|昀øùö'ˆéòlfW2ƒÂ×Y2sù„ÞÃî뇎À?dÔ(È;ÙGL ‡þÊBæ>Ä+˜÷u1øÌóªµþAKìÇùz¥Öo‘ÓȆïŠÑH@{\b1漢SRåÍŠ7/)£ÐõÌÛ½ãy™ —‡CS~ó'±\÷xZ‚B;€åÞ¿€§³ÔåÍjÐæ(¨>~ª‘Hë¤9t*H6Gf*Ê’…+4Î<æû¤ˆCº}<$!öó¿LÀ=\R5Ô £§{£ICsD÷p©±±—LÎ%Á†AV6#l2‹v7Î&ô‚àU½É=ßu~ÝSîŽIþàá² 1CÁÄP=¤[R°¦V3óµÞ‰á×Qù¬‚¡î”‚ȸ$Øt%µƒiÃÒOÉððú˜CiEמUÓzM ’¥õ²yÑk÷7¡|zóg,nÚ>ôìšÓlëìÞ©¸õèÅ'û:«·ðêå©w![Ξ¤±?+ÌÓ^÷*[h4¯2tÚäDè‹»À²Ô¸UŽŒC²ئ–kÔ§,£†t«‚ˆeÖ CN!¥`T'ðd˨^ ô¸ûøEɸŽÚù0Ÿ±ï›\k Ýh"û*?ÕÍVü~D®E”‚wøŠ;èÏÜ¡9eÒ´Ÿøi38úâòÂbzÀ)aÒhÑÎD¢gæ÷¸Ôê9^wÈh¥"nׯß)Q‘e ©eÿl9Äð»enÜãäåìWìå[Cf`lüêâ0蜇-EMÙã³YšþŸ£NŒl…Sê×\‰vÆñykKª–ý~+¨*=û{Œó:mH=Áô[c:eãá.ÅÒ”V‰NœáÄòÍEøAËID¬ø øâÃ20ZÅy·ßÉ¿A !ÇÒ¶WÂÛF&ÛÜjEù¯ovˆyýTqïÓîO²ä¶°€l½ÿR>Ïþ endstream endobj 296 0 obj << /Length1 725 /Length2 31251 /Length3 0 /Length 31768 /Filter /FlateDecode >> stream xÚlºcpfݶ6œŽmóŽmÛ6:¶;¶tlÛêX;éضmÏÞûœóÖ®újýš5¯qͱjÕs‘K8Ø»¨y:š1Ó330ñT%ÔÔ˜™˜˜L ¬°ää¢@3#+{1#3€†™)@ÕÌÀÌ`abb†%ˆ:8z­,,]T&Ôÿ6ÔmL­ì¬€u[7+KŸ››››³+ÐU€áŸEªffK3€¹•­@TQIKZA@%©ð ifo4²(¹ÛZ™ä¬LÌìͨæ@€í€‰ƒ½©Õ¿093ü€½›Ðå`æ@;€¼¸š°„¢‚@\”QM`do “úg¹½‹3Ï?Ñf&..ÿÊŽàò’­ÑÿIÆÿ'ýŸèñ‰–™`jeâ06³°²‡eüoÒöæŽÿ˜M]ÿ×õ çð¨þá`jfþO´«­­‚‘€JÔÁÎÑÕÅ w05ÚþŽ;ÐêËÿ†ÙYÙzþÿÿO„´‹Ñ?dÛ[üCÓLVÎVf¦JV.ÿðít5ûYÃì?g#ofjåj÷ŸåfÿpòÏ©š:ØÛzþÏŽÿ¤òïý•¤e•%iÿ¯þí·7q0µ²·¨ºüçÐôÿ ÿv+Yý§‚þËÿV€ùÿéòF.@+€ÿJ„é_ÏÿJzÿ/JDÄÁÛž…‰ @ÏÊÁ`fådp±rùþ\W ÐÌÞåߤÿ“Ëÿêÿ.&333ØÕ%Þë´–° ?ñ¢™JÚ™¿†?i—˜,Oý¹ÖA±´458¸Ä½SÃÍw‡'˜4æ߹ؖ¾k&1Š3ƒ?š˜Œ!ŽÐ[Â})!­´ð«Ý|S.xmˆÐ¿WDZ3Q¿Ž¼uÉX[ R¢™dËÊ=«•q*ãG‘dŽ1,ïèƒ@ÞŠ_™°—¤Ö£DÍÔþ‚7Û£­Až± -¼2|Ù] ÏÞ¿&Ç1°,¬t<×.h¸´ÕÑRšeî’ÂÊÓÙÇÁÛ £Ør«üü5¸$|wɺ ™¾á.¼t û˜Ï8<Uqͧ¸± (O xÕ½ð![Å÷3*nó‘Ÿ­²Ó7ò`gÚíÙ﯆†.¥ST€Ë_J½-™9qØ !¸ï»]žÜ5ÙE9§u!ïqîõ± T‚ «£™ƒ L’ªè¶¨¤¹+àiá›3¿í6¹ˆJ:¤Ëº©Â9G!˜=øºi^ ý1¼œèIÌèI[›L"|ëÈãÀ¨E>œ}œ¥`âüµÙ_÷½BîÎfgö½JÉQQ+Ö̦_ÞÂì‚ÂIË{gjæ¿ÁÉÏ}éWP¤Îh+R#=N+†«’nêië%šž8e}ù•ƒI¼Â+jD7y{|Å¥0ñ’4 &NŽ„õ-߸c;Žn¿ºŸ'…aÁ{—fΕMm˜¬³C†oMTŽT*P*‹@yfÀ¹1Oõpž¯Ý­ë2X,¼êxÖys[«Â[ÅPuæ19òY¡ä 2^áï ,Q†‹Ÿzå‡)ð„ ²‡øëž~?ã'Oöád¹`ŸJ &ŠOÇ[CtJztêîõf|ÇeŠS@H/`>*µM·Nþ4ય? ¬¥`EF±{î÷¡§Ñø$ûY¬—wµ†À’¹ò,¤©RhÃ] ÷Ä @kYqëo¼à¹/+UY$§Ïeå(ºÑd¡ßz**‡¤ Ð>áï ^ÇcrnWù oïÁ}wïÆ y\ŠŽŒ¤2ß!O˜¼3Ó¤nðu±Ë6ŽZæXdß \ã¬z(Û#•¶Ë´(¾^:ŒšÉ©­ë=¦zu¦†ˆO‡ì¢~ei-á°ü¨üÝÐzª+‡Ž ï“NšÜüÉ…·Ç½Ñ™Ç,tÐXB¥¿GÈ_p´[߯\×{oße÷=»[CÎnÓZµbä:!:ô+Ú,nœ¬GÏ>Ö Z.-¾2ñ åÔ²@>ŸªA´X8¿Ìvuew5Û*(†ç~Òý`%»à) à_ñU¹Â<— Ñ_½iSbàé¶&Ó:×4dŸTRêtÝ&`u±M¹GCøuSlâÑ áº"kÙ×X8÷1•…òxTx Jm«„ˆm:†?=¾àà¢BxKL Ÿ÷d§µâÈeXùŵ¥l®ôÙu:p.8Á¸ÉîEç˜l€æ2îÎñ4¹¼‡d)‘Cnâz´cÙ¤cKWlPÁñ ֌ӭY£€éÒyŠT Û…Ir )ëë°Âm¬µ2ªÜ}‹wßjã3I;ö•Ù‡ƒ$z z,Ìæûo·w&@Ô|íŒk1Jχ³«¦V1éáÒt(8¥s8”Ù“Õ§B"U‹†²ÚÅe™Ôñ·ï` ÂaÈ ËOð1 œâ¶cÜd¼DSíçi«ù®;N_ïÙ'v­œ»Dœ 0ìˆz/Ï ÀÐAL}n%^ú?g¦É Ç~c_68ÈþÔ´%šñúðX‰peÒ©n¸iI>ñc'Ö{ëÛ•\íId úüi–.ÿ0œÀw»ØvBŽïÕ`‚Ô ‰µ£”,ÜQj|ë-µ[X'€x.)KÏ¢Wl¬Ü/±U´#û¹›·`0äð×è„®˜3gæì­ óœ;Vm"3[û.J>bó?¯g²2­Ó9†±hC[Úᛧ1µ…²˜É•2?®b CîŒ$s–qßrÄO¹2v,Õö“â㸒MæÁd ‡VŸ_=£ùœvò+=Wž_f>JÑ nÛhB'¡<ÖÁ¢›ßmB{ð tpŒþû¡Y–j¸ÁB‰ßhSYy£t²,å+xÜÑ. “3ßÓõ‡Í-Ü‚c?tð¥´)EϤª¿ø{¸éÄgå ×[ÖܼۚW…˜Õô6ÄŸ×qðU”¹¼ÝmUdÖç8ƒ5Á¿¹4ã)×ÆÝ•–¯ ƒ‹©×JÂuU">§™Å¡ÊÛè/¿äû¨ý‡¾´{L,ëróBÏ  ‹Ë—{wKÀ©îã+Ž»’_íÎ-Š½è?wÌ1yk*(ý?š¸ÆXÌ’¬øq£Fß7u§¨c¯ÛŒ_«¯>+Ƨ /2õ°sÂÂl_Zôu2»tÈd$¢ûc+Ï™Íuý€Žôò©ÅzÚ‚`'i>ÔzP89ò̳‚ô@,õt½FY 2¤Æ ‰xêú)jññJn°/±rêñ>Æ’˜—vT¡$)ä €sÛ¬ù®0ü&]»)/Ìm0Éàƒ‰òUêƒ9a¹hcëÎ7×®šÚôe§½|ØsU¨¾º¾ÇGYð}ÔÇËògá”:j-x1À¬Šè´ò`&7íþ; :¹CKì¶ØLû¶“ùÙÌ¥}‚¿®ë׎7éÏEjN¢<­¨Ô|b™’? È‘A60^ß$¯ë«Ü’wÅÉà˜Î<¬Œ±Ž.ÎØÂ-ÕˆÝyî:M¬‡¹]ÁL¸0`Ž£õæö¬ªmÇ©nد;ôEÜ¡•s®ߟ©¬èÆlÇIiF†âÄ6ÙxôÀ>brcrÉnÎqY.$ß#å-ô*P¡Jê&w&É<óû?Å4ýÒÒ! «î2 ®á¢— V~‰ µs€*fFE)eÓúÎdðocüüí`¼®´Nê‹h÷Éò•½Uô—ì ox£uƒÞÞÕÑSz–I*â¾Õ¤·Œ"õKôæmûôU_P¬zlj9=õÒüÝþbá) t˜ëý'(’ø®È=ˆ ot-KSŸ,dER“ÅŽ£8â[È^ ú4 ÑÓSašEJK 7+3bsÕ¯\>Ê‘ݬò)ð3”­„´ù(º·d"²³y÷Ø_'8¾p÷¦»úñ÷<Æ4ߤCÕâ) ÙÃùݪÞ&±ÄoÇVòÀ\JÛZŸ ÌC›ŒÊ½zwVÍP×èî4H…Å…Ê‚âÆày)¹Ê+ý€º¿ùzkm#âÏ]ó‰}`´ípCLå繊4óU´¬)o@ŽõÁ GY(4‹Ç‡ÿ’æ=?Cs^¹_ÐäóšÀ£xrײBÏ7·x»æ/‘¹ïð®|§2œƒ«†²æÁvNBö4~ê~2B´ô“ a›ôÙ¡c\ ÓØÙÛ4ÜËè‰ýìnZÇÜ^Ä‘ÉÆú’_<®š9 §£Õëyü¿òÌËæç†êë¨ðÚ!øc[Ÿ«î°û­Ïð¥×»¦j|”ËÇ´_ýÄŠ‡í@™TŸˆ1[ø¦Ú¶H=6Ï[”é5°?,Ý+÷/ñ•‰¤,­Á@‚° "®‰‘‰°°õSÊú5:+"¥êÐMÆœ±?y4¢†ÒŒËd#û;¤ F¹ôá|®ÙžOÞ®ûJÚ(Z𯙥©Ã(r *,“lF½ÎgÑ‚¥0oÎ9mØÍÈEBlÃ-ê_)ÍuLE¯ÉHúU;¤ù3Wâ€à FºõñðUþúËRáqÃP"…®t"ºnFX­ùA=|<Æ<ÒÕ£-bö)üÈ»&lö“4´»vùAé6èéIâ[a¤²+!p66ÐB ·ÏQÆ#CºTu7$úb‹a˜”¬³·lsï@Ê”¥´ßOkd£ê›J¯ûþ½ìžQd9¢D»ihž{¼Åʹ›V/4¥[X±>²U#Beè 6Ùbý¤ž¤ê..·UÄ' Ò»³^©  ó.WE¨*Jºä ¾)8#aíát¸|ù™C¨¨áò°¯Û»þ?r"ßß`‹o½¡¶+M5âE`]߆¼•0ÈYا‚ ¬t  Äï-f‡³‰PPNˆ8êà?Š¾=A²J‰—+'ÊáiäÊ©(¸ØN!T¶í÷~® âƒ6pÜÌcjÏH𿉛6ág§±o9¶dbqoTîe ‹ø¤ó]lˆx“ ®ÉÊ~#ò–¶Œ‰?›x,øKZ£â[¥CZ÷tÇ×p9@l:w½Žc†ÉvÁ½náTÙ}?d)ÀV€˜¶’Ê6`-aÀ°«yþ·Ü`äN¶ .¶© ÌÙ «æ±SÖ/£XÙbKÔpÙâk› F¸b#C6÷ IÐ#l £Uü¹J`@áÛÆŠVP%pb+™Y¿ ,¬‰á¼ µ8IÊ‹ÞèòtÖ|Ð{Û­‡Ÿßä;e¼}2*W#žª(ž"Ëö®øAD‰œê$©>›îMRb zlÃýEòÆ”so¬Av}}YfÏ´«cËA |Äžº9cœÙ;>ØÚØ„êè}ˆáÚöas$hópzýXÑ“U„KNUÁ/©Þ¾®½5Ò}›&bPš p åÑØëë#d×r½*µØ(.¹ ëpýQ2͈‹xòýGNÙ 'Òï{?ÿ6Q zv±]q± ¢ÜÊóZ¶x¦U‚)Ä°@æ®>R’ƒ±ŒLcÂÓþ†énU íÚ°ƒ™Îeœqv$’²xó)VEðs™u¿¹¯Œ§kó¦ ]ül}>®iæÈÕ%êõ‘ÔÛÿÔÿÄÓy9½ œ aÇÎlÃÃâuO¯ýó8¦ Œ`]*ÄÑlíaÁ/ó%£ZFÏ¢q£?ãÆ6Üs˜rdè ²¥x}˜nùœgÚ' Ÿñâm·Òÿ`¾¾œÜ \²þ+nú„¯Í’Uì4†pBƶõG¡ÖC臅wäÃU9œÂ z`³±†îýÓÍþeÓ‘k®z܃Á—Ÿ²$Âl¾•Çº. Š©cZ(þ‹ui‡¢¸y96r›òlzÚ@îιæ ô—ê¢WDÈpµŒjIèèGÓÏõ>n´}Ô×>Àùè~ÉO95”*ú¥_5¿j©„9ÐÍÈ)Š*dܶQúœÏeîÀ™"á¿Ñš–ZÞÎk×êeº#w/˜š ŸÂÓ ÀK'4I“fäb‰¶p/¤žàÑ‹ˆhÒYÇg{ïÞ‹_Qº²0jîÂUˉ ÿ´Koƒ¬½U”•Î‰Ñd¶é"øÑIl¶Hï½kõO_Ö¥¶ö|°#»Ž™lIÌܶWÌÝsB»u3–£%–ï•2‹ÝÇz0â§à²ñÔ˜ùø~wxŽËùúɉ‡+Œûx<‚ äýÓt„¤6ë¶ãx~êéÓëŠ àâÑúj3³»WÕ2ljt–dP1;/ Ä‚Xœä±ŒÜÑ 2šp¶áj•uk<µwx9Ò¹´ïkjD|²kZ¿µu„G—úaÉŒð'õ{`Ï®¡…ô£sØä¦ýp踮‘leäĉeN–7HZ8…!;öà_N­{‘àJþ¦AÄ¿‘í=¢?ÔÒܹ=‚Ê×Ê=(rTfÂ×”ÈlÂ÷ÈO*ž?©H+ºéÛç€É$ãÈïçu °Ëüu›©ÒfàÝÑÚÄC»ä¿)’¡PužI’Þéà µUÞ®‘!ÌGâe9/ƒtcÂÓ—DL¸¨×})TPšæ|dÚJ•@-3k&þ¢ÐÍ+žz°÷˜sxK_Hjã Ä KµKK8ª•ÆfŠOw ãaŽõ… EùôÁÖ9ˆD"°†À\úô_Ø`SeÎܶF tI™;t“M-FíGBÇj§7þ89†ý¼)yeµ´Û\…Y[bÈï =ÊìÁ‘I\c~0”˜œMc’ÑÔV˜ÐsR–@ÂÕ:šH<I3ÀIŽç¸!å¥ÔÞRg:a±bAâl|%$_Âîžìî†þ@â;bJór1nUÑd„¡\«ê%°ÁE]OŒ(ʬø»áJ‡ÈRÓ¸«Ö$ žopSDns*¼+ùvpÁ/“ íЭʑAož¡•ŽæÎ*z?ºh(Ó·ÀíÃu8òѨÇÞùJÖþ2w€ú§a\iGˆ¢’”ž:fmŠ å?ö#ì_û‚›%¬‘,¸Ü›Efóó÷'&§õï]Pü(¯‹óø31DEyt¬pò]ÓÙM…pþ‚ ³ òøjÉN%ù|¢œBÓ&ó'T™æ/äö{[ r<‰cˆ#³`mù*—øMyÁÄÔ¢WWZë¥ú2[ì2ÀhhÒüÅÊ‚pœ]±wŠr¬|ÜÑ %Aù¥ ®lSô%½<îøh¹Ð>žGVÁ<Íg<8°_;íž i3471ê¡5ĶtâêÐÁÍE1º_Ý7ø©ˆªñÍ›î=«÷ÝÌìØ©^H°?&Î À4è(R¤k8¢ §Ì÷+Ú±œO'ñ±[ˆp7÷áÅâ(˜`Á'ÈärØÅÓbÙ*j3ñ{žÓ }„ÚƒšhŤ>{raH£?}ðä¾(üt];à(ó`Ìóu1ô¯]ìŽ{å’»•ý •áÁD8²ºìëQ?-—î[Åk§Ïù\#vÈüíÿËà)Ä+0€áìÒ®ÀSê숱‚$–4ȉ¾.)ÎIB7†í½ƒ §&zdG–²ñ¤:A¶Ï~ð ¶÷gLë%ƇºMV¨• ËýX+ú7Ä^UßÞö#Õ…ý‹žÁ1{lé'TÈ{X´–ô{'¡L‘§3µF5û.úèØê “‰I6ŽÓ;)ãÄþ`"‡RÂE왘 -\u8w1‹›ÜýqÀBmÿ´ë)ØeîŸÜB‹¤Ç‡)ŒqwøYVþ2Z»=d ¯, vÈ´íõæ<"ÚÌúcÄ5Ïd(çÁ6Ù>óEÎ5¯¦™3`?yuN-û8Å<Çœå!}6tŒ«Jþ¹ ¶Ä)ÃO",NÚ9çÁä. aZ*måM†XUßY¦£[.Ȩ’cìð'¦d""k¶Ö¨yÊNMsÚèµè[„cpv·à, R‹Äâ×çÂ¬ç– ¢áH0VßãX”®v <_"¹Ö;jå³À“>¼€_ 71whDÈdÒU&Y¤K¹8ˆÓø•Æ‰pq¾‚CÜ/'¦Kwª§ÉÈäÛ/¢%ÅVýîÍÈUÓœN#¬‚VÅMÉnE) UˆÃÕ%„£ß•HO„ËR•db±À>ÞÆóÖiÉró,),½$A<%«¯;QFd2IG˜ zeèäýܱÅ~9j‡gý‹ÜÃA²]j2YœAŠÌ¸„‚Ž¢"@7ÓY¶Û,’ÚË‘øªC`dFb‡_(RNîªê']3ÓàR oÅŠ±bë“ü4G¨¬ò#à9d‚[Ÿª 1ƒ³2&êôK æÃ]††niqÈÇ=³ð¡“àµa…®5y³$ÁñáÕDSgÝeˆ dÿ“Ñã ÞŠh*È¥g£²7°MNÅ2ÎëUU|­|&b>Êi5Úš‡A¥ûÚµo¬à·ÂtKêx3ž¯)ßp ñ*5јnCZìÛa-ŒGÊŒùÞ|“íÜ5ÂöT;’aJîpí½ÅHâ›_”ÀÄÏà–CúØIºÙ“/®ò «É¸R‘¾ƒ}º~’›GE|°dÆð¡´÷Þ^èpãŽÍXp GÖ7hß#­Ï©œR˜òùÖ,t!ä”ïÄ{¼µ„©L™‡«Èã¬íw¥0 øÝé‘ÔõìÖVý UÝs´~(¤d7´ÑIöÖäŠÜ‹ä˜NµÂÂ~òVs«ÿ´L¡Ç2$þ“¿ …–ËF" ,GTÂ:D^{à €ŒØ§B¦!òG§í±t­Z¹L`$Ú²ú þZ7÷z.–nÜÚ–zHI.{íNÎÜÚžþ4äâ^þ†P, 'p÷ά¡páÜ A¢­ÁJ ’ꈼ­êN¶ø#üžÏAÉè‘Çg[q²c\Ç‚Fy\\`ˆ .?0ÈÄuä=ø.ŽÇœ€i?“Ù(׃ôhkÕg÷Î7º ’Esd§:ë:ùZ ›öð<§×Qˆ’çðÞû5Z¢—ouT×Sj2D.»“»aˆÓ•Hk~×çJXHš¬]¿Å×2qˆ¨_–¥à}wñavýËwîî®Ü ôxÑw"õõ¾ð´|÷‚Ëæ·ô´¶){q]Z®š"ÈæOåŠ8Ö’*hÞüõ~âÁÏ<ú1«ša(x:';Dd›‰ú ›ôNßyš¥õ âYуÀIχªÍ¸ÄZ›Ï;fè9Û3ƒZ|ÂK:ñÿ†³FCºÜº´ˆU\íZ–HÚ–=ݤRc` ;E?_©[@iÓB9ˆ²ä˸à{ðÄÑží²wš‡‚n!r„ëí ´~J\®îHø¡ù>»|‡<ßÔ;E³g‚®–‘8üA‚9}ûµ·½Ï!‡LfÄ¿2ÙVí¨ÙÔ >xœ`ŠÍÏd}Ì‘sŽúóíh¸ xiäÑí¡7ý# ¨¤îŠz’ö!¯ºÝ\ð«á3ÌàÑ+#š¬XJ¤ä£ "÷š‡æp&é¨`Q ©à' ‰;¦ž»Vþ úZsÇ“Ðêh¥þ@é J«”0çÚ²Šˆ…ÓíQ”™M ”wóJƒ&ÁîÍ‚7"¤Â ³‚­ÉŠ´ÒðŨ˳ڇpËI =Z«ÕW~Jõ›\/\ؤ±ò¯Žònb H5=ùív"Ž®éŒ\k¡8;„sÄ[…’ƒ%|cL A·œDÀ{ »ÐÕ{éTËËI¸5”¼#ÏÎìÿ˜ŒDï÷IO´*À®Ù‰ÈÖ5¶3õ–ƒa“ïPR®îVÒ§7f ”|Ë~–úKȨ {5€¦Õ^ûe éÈ¢Šg2 „÷ú©!) 1º,›ôüm*rF2#WôÉ€<ç{ x©¢ò§¦SùïFßl!Gš*/¯Ä&ÞWAڦ駵Ål®â÷ù8uÅËZ½‰R’ˆJGN ÚÂטCùÓ+sŒÕ”[ ¤*{µ±|–þ@ãö¦°CŒçæŸ –Bbz{"¦£[˜j¿Ì>gºL:s`»ˆ¯™<;\F+\Äoxâ¯E°xUá²ç|×TŽæ(9|Aý<¥4ÌLj•çˆ£%Xh›p¡<ÉŒ‡QT0W*ÙQ’.ÙëëI'#SçßÔ)Ó„Úß¾é´c×`5°¬ù͆VN±~nÏã[Zª&í…]ï!;Suè03j‡ªë•ù¯Hô’ïé°(Ä¿¬ œ1jÛFyņû‰†ŸfU¸×è GªDƒZíÆy!ª9ù«ŒÀá_Ü=É^%`/GH‘9EÊÇò`ß’Wdö¹@³Ðè. ˆöZ²!CÁ©%ä¸ô¢ìz0WÚµ¤‡žK”ï™e«ìqú¨vô7„óµ T&-¦¸Ãjð§n§®ý¶ þH;å6ôÒ/ l›úXÝÐÕלÔe+p»ØÍ–uŸâ¦+e³Í §“‡®pQ}ð¡v¡ß§cwšX“UG©KÊ, ØYA°üyâüë!’¿%+ºŠ¶û6ÿ"Ž˜ ~?wÊBra­3€»ˆzuñ54{Ð,Óáäu’´›yßK‡K Ž–ÆöºáK~DáÀ„ª\ a¿ #ü+µvL$¢$™u·Ç&ðF|þ­qΕÑÝŒˆ„$Š¢ž)‡èࢊîv·5]ZGûm˜äÕ… ~,r?VºŽïU=g¸U{¡ç˜$Ê2;8ÉÛ(§ìn4¾É kKl`vÀ&Ò`µa°#Çæ@Hüc'­Ý¦üôåÃÏÅÄš‹Ä¼¥&š é•HSŸÇÛbõ±w²H1$?³h±Zä9}§éF`i‘çatÄ—4‹ÖŒ™f‘åNa.¼åbÝæ¨:DaE¸š‚Ýa*—šÿ²œÅ*N¬!¶—Þëã iÕßÆ]ŒŸ'²Kg±€,•íu¯˜Ñ³xŸêœŠô¹Ã9Bb"Â#±˜%@s´ï…ÊXªX½‹µ¸E‘ýÀ!‘l*¬°¶*1>Q`ó mjVAÅ|€n—t2 Àùºÿžš²Úõ+Å "’nè}­SkM6VšÂh"äèËáç†B —|¶á¾f+ô訶‡ÊNõDÊø¸»õ ÓÛ/c’2æû=Ê„ë蜙3w‡|96„ìæ\VË{ÞÆû tÁ8CÓžˆô…â™æ&Ì5íS‘«ôÓ*wž©Õ4}¨åUŽ0€Sô“þÐ §Hã]suÖ6ÁÓ²ÙI²„ øD «Fá2“‡vZä™ã«´¬vÇõÖ6¡iVÈèÇv¬.çáXos\¡ý~áºÃY¬¦Ô¥L¯VäÌ;¡ý1ÑG”Ó8™)÷ùëÙ»ÚÎÑ•?RÖ¸ãqd>ÈÔU|©îL˜»5¶Ë±çgê`ÃEcÈßkCãX6íVuà–ó®m¸sÖáð 4æ¸Ð6sWXö¼©3{&æÜ¿Æ÷šDYò%îh­áEJõÃF#xRV£§”¤©ð÷:ª±ÐVú\¯™ý+(ˆP÷~¤Ý­ ž“:¼’¾YoñðÃ` – /õWõø_Užà‰œÍŠ×i´å®T3;yßrô›B~l6&Ûø„W”6@”+讆ùpŽcÑŒG?Ê‘Z­„M\"ëŒY¨†¬2½f+–xôê$¨yŠŸÔ†¶«)‡uÉ8ztzà>RòbTjîåÅÚ?%½>¾Ò=“U¶Uѹ QÔÈ*…>Z[_¿˜0žW!{M‰¹<¸ìSÌ䊥ÓâûÃꆚ="»’Gm¼[Ô«àûÝuÖ¹~µg٘˳:–o¡Õ<öRçÞŒYïX^ ™ Á‰fvìð©18ÛrÞͽd Àï¯?«vEtö•·ÕVù…Ñc‡‚ÔÈ×½¥p(@Û}£Ý4æfѸÅÚiÇÒà~6ˈµØRöüVH›-ä2[B*ß!³rŦ.º|1ì ²€§¬ðË}º•Õ â„ÓY¹?£æôzÔ@.®PL€Î­ƒ"3çO&ÆìMðùi³êxëâ­Ùåä½$‰Dܪ›—ч£UHœhÏëçF =¢ú˜Si×ÌÿCù†£Š}”ˆÁ6€ë0lÍîŠE âºéó¶ [ »jc¾É$9ĸÜvbEÚ(°À´­¸šbV"¨Å"¿sêÇ@í9¦kŠcº O’¾ bEòƒfOÛ¦Ê £Fø0ÿUòrfºK"ÖMuyµKwRiCî†t¥\Ì€M=ÄÙ½÷¿¯Ÿ©+­yh˜±å©´‘%úÍ';ä—Þp`‹ÖEe°×Û†ÓAêç«Vìà¾7ó~—Š%çS7ŽCí =ZÀî~·%oüžhìùªœÂ?YJà*„ö6ú´EPøï>³ßKÇD™¦—˜©ò]’¥]\“³/E*øÀ¨óþºþNqáS>åù§«²øÍK­élÂQoÎ]ÇZO:_Eu‰Í‡ÒƒK/è=Ï.…¼.PÇ ÂÑ€Ë Ûƒ¸(fmtO פêÝß6š[|µÙ‰Öë$ד³nÕ?Àò“”'ÙÊKÀl˜h|jÞ^ˆ´ŒQ‚ÃÕ§"#˜’¡Šýñh—ǧ,ÈZpñ$H³V¶G‚ç« r4l&{® žE§žQ "i' ÝCö ™ C…ÒwÖ!F™¬{œB-Ä”ç„é<ÚÁFi»³M¬ ͆t}†{¡HáWUT£Î½*ý)&ÑYzœÍw -K}-×פÐ`2ê)»ÀCrõçäø„ÿÇ4ôs™ßZÔØ¿ ZB4¤ÉÃçxcÉ »¬º¬¬H(æ¾oôÄ+ˆ9=Û¾nL¶%®¿r»«ïÌ’AîÉՋŒž6‰KáYà1%hä,[UðƯ9ÆÈ™ÂOÁšé¹ŒY€Æ¿kÕâôGì@˜=(fM›¿Üb9¥zJÂÌlµð“*Ò‚Èô´~ct$wÁËKiˆ Îæ•XÊÝ:ÚþÈ=;ÐFRy&ÙZ–í“ÂŽ.Ø>i*t?_­>Æ­!ŒK~tÖ*ã0Èèus;PÍWc«èÙeçñÁ•‰jD|îRt)už¾¹Tb$þ8ç"qY‡;ÍÌý8ÆÆ\ô)6Ò‘å“Dú5_>ׂ½Ð“¢Òˆ,Sn§°bŒ—, ´™ (̺èÆ®ü½¡D#BØu«Kƒ.,šf=ù,aA]ùMZÆj)WÙÌW‰àô‚lan˜@º±‹›å2-Ë5zTCe)-cÞŸ‚+TÆJÏU÷Ó¡>@–¹ð Ûîhß­T vgFÑ)bDÍï¾{ƒ ”àŒ×&³NÞ`)¸ó]¤«’ƒ‚*&†À‘0~ÐZn2¼ÛÎkd³KÜK‡Áüa qúØ­ u‹íà%Èž!©ª«ëïY,‘_ƒW—K윒\ƒï#D ƒdGûa­¹ÔoSD3 qs3Â߇©åäyľ?róÈejúSæwq`…~gZ~ì1üˆ¹eÔvUí:„R–³MF ©f²áªî—E¿RТ]ðø=U´["½¢¯ 6 ùgrt×~ªZUù³eâÓn¬„ØkÐn]Pìj€×äîz8%50:1hÔ 8éj:kBê1Ò~ÿœË¨.ÑŒ³[ êQîÓ«“;f9u ·#oû•Ü&ñÐ(Ú‰<믉ê #WmUÙ+JΞW›5Šà{‰óÚ–²×Éeã ¾‘r¼Èê=% ˜ údX± ûñ`PIËm¶i0Äe˜^6IcžöÕÃâùøáÜ,½ÿ„êZáa`ÅÍ–Ýßä¹´1HÆ—g¦APú©®°Nìü<ìP:ø§2_µ#V°†jYê¶MËáÑCëBA€ÉÏy•VxDûmTžú“–Ú÷Ëx–ZùÈ!ÓV%è؇ⵦ6™tG·{h$÷Goéê´E µSeÜÕXjÉœnVžê‹TÖ–êJfh³—ÉDEí¯M¶BþU-j!¨©%‘?jZ¿m”>€ŒGqºaþ/¬œö›²¢Ç?6Y7·×†3éexÇK—sxNñug­=8³Nƒù©Š{³üÙËè5ɃÏTd)•94ç0ÃÆÄ›&ýä›sT>#H'¨¶¼ˆç–Ø7 Ù´ý,Зr KŸ„>@¡¯•Ç›mÞ«ó|å …–<Ùœ·€¾mK£_ô‘cùsŽæ¹žt÷mY½ˆôGJ[6ï\?|4•K¡ Uäc6×%}³-f°³í;kÜH¥7…d_$W>דß›¯±Œ¸Ê³ø‹½Îr½¼êï+k%©°†Ïè¥W“ÖÅÆ¡š½¾€ur œmÄu÷¾›/ù '„NÍ€) {ÜóŽc WtOùaùz\mšîa ºŸå5•ËI`ãB|ŒãÐ/¼¼¼/ÜÂ:zùã·Y”Œ ¡fMñ+ä³ÜñßS1kX‰§z?$‚µ£Û1¯ý05¯&˜b/ þb"™ - Jòº@Ê3ÒO\NñéFð*íN\Ê_"¶—x@€ù&¢JÞ%Ë6‹[Õ%†G‹nÆÄiUJ ë¬D'•Ô2½a¯ÉPrš1,Ê9xˆmºlw–´~ÂýJ¥ŒG#vkü¾uàKm‹"WneŽIÁA»D6s¶PäŒõˆ¾ø×PkÝí×—»7@è„ãÆ—9ÇQ Aubäõ<Ø^þÖ>àðÓÝWþŠ0˜t:¬õkRaÌ*­0UWiZíVÒ–ËIVŸ—¦§`‘ z[2§®Hº®‡|sz3] ÐÍ¡Wæ|†8:­æAW¬¶e?"áTm°¡ïÀ቟”NÆ ¬Ä¯]ÂÙÒØà Š‹ðÚÃ6Øò0‘õÚ¸1‹7ÓæjÀb<§¦?pjã¹_b¸b WæZ:¨ª–ÔWȱš)ÄItÚW‘ÚÂ'³va1»v5w(ñ{líLñ f樕Ҁá½Z酠ʳ­}Ö‚ìóÊÄÕŠsXEE(ºBÆ{Yõ¾`Låédy›Ôt/í›ßd§?ò7ÓˆíÀÿŠF×[ëgQÛ·´±ßüV~Ì $Uàiµ¹¦O¦®Ôú‘>”â²ßtÂ?îd‘â=M66J¦†Ë;õW:¹ðzˆEáâ¯Ù*ó…ãÿ¬$E_äÖýMÈlïA(}`h¼‹¤ «Ç{Q…½TZç b„jÓ —ˆ°;¾ùEAÝJ›o’aùê¶kÚ¡jx"³J³õ|èŸ2ï˜m9)ÕÚÓ†µxÈ×=ß>Öê“£Ë: æ™ ³Ä½V2Ϧãƈæ°L’êóx?¸D ž«›À Î>ÉÞjŸ=+Ðwue·Ûþò(a¬¤ ]y‰&d}&Š±²"!à«xÏXr!F2‹zßÑ¢=Þ¡µ8oŽ[–RpÄf’uh‹q—r.XàäTb›¡¡ð¥N6ñ_ á4Ù#'¡6þÆ# ¸=Wݨkù‰¨H–Þ]–Øò>—Ѫ>Ýï¾õ¯jPõ¹ËÝ·:!›c¼°yÒ¶5>÷õ›”ù¦ —¿x¥í˜m!¥Ü\¤Ï”ù.-²˜º"Æ/x•ÅçéUšˆg˜D,¡Í¬û#ä–Ö/ˆZk½•|`)˜B› ¥$’‡s%©N¾ü²'"ø’jüeÆæå§u…G\žíž†G²¡óò†ëÀ÷ðÒ¿‡Gï:§ÝÞv«Þ§$Uⓨ³‹:[!w;nC¸ÑìĪ\]¿U»'¥iÈ*ºPÇTº—ˆîB¬  ûÌÏb˜¾Cq¸Ã•WŽíz¹—PhÍŸ~ßëÎ]1Œ4kO¶dŒ¨Ï(ú´È,ýª%‹v“F­ñÄWPØšZΈëz¾Ýÿf8/°S£òEΕôó.Âu­Ì($v²v’s¹)öxs?6—‹j1…ß‘ Ø?òõª@|^CýéY-ý³~•÷˜è‘þŸ?.0qãñ4-‡`MèE\6=§Ç§ÈCÛ’ÍU¯»šä>báUƎ߃8_xAˆ¾Ô£~=CY?}lÔsœ–S[ŠÁFËJhŒ:¡Æ2]éy6?¼¢¯:þçî\,WŽØl·ÝuƒDb¯0lûÓ>Ë9Ǹü§ŸÂ¤}’ñÍãÁÊ&£ŒLø½Ê‰öd è©"Öû dÏj¸ažMßÓh‰è¹—:=(2µM)ñ$‘§øÇ0ÞPDvIP·L;¤`ëºÓÐ5þ>r!¬D‰I×ÇÞv6 äTÞÐtí:ÙœÈA»œ¨6¼LwñdÞ*¥kÝ ¬`:ùHðÑÀö^ÐåÖ<§Ðao‡ Êx‘-ëÎx$–ÁÂF¾.zDÒa¿s„yõj¡3‘4H£‘.ä8êN "“µžî»\{åmx{莣>×±:vW™¢übo´ñàGŸU“ùÚ^âú4†¸%†÷¨Fäd ߤY¤k#}uv¸Ðb5¨ Ÿí¿ñ:Sµ§ñ”¬yÕha”¤…58”õÀº,2‰sõQ!‘ Š_³î>Äá»…ðé’<9?@[ßXñô ›¾ø[e¶bOŽZ.Ç©ýŽfk¾Ž2¹k6*r©Ãúq¯uTnTºhÜKºHjÄšBýZPµ£Aχ»G…íHººk ë¹}ú ÛkA/ùëŠNÓ)lD\â ç]1h{”W‡ <•|ͧŸ€åp:*¹'H N€BÙ0Lg0ÛJ)õÓK ‹”>›*ÊÒ-)º«_MÊPAÎÇ%fí#£Â‹¤Ž…xæ¥ ÓÎݦ¡Ày³Æ…réZøfÔÒk„£_8­¹¤<~dÎ ²œ$Z@ÒîïÃŒÐÜ—:Îúª•Lî¶&2žNÒ Q†uvè_n‚ìC2ƒBv7LšÙUªèkfG«ºUR|ÉñdÈ5eƒÈô½È æ»Òˆây+)îiOð¥P‘×–< #²Ó5€˜»ÌºN>PÚw‰ëÔ÷g¸ãOØ(Áõ—ÛŠ¬?¾C™U à÷ÕÍ{‹DŒs5d®V'ówÁ‹P÷²]Ó†ëë]ý˜ (Û|º¬¯=:éW¥x®šJÈ·²pJÑ]€Ÿy¯ˆŽKV®¤ÝRzÕ·_Û”­nZÖ¡Ø·…¡MXµáußÈ!†e;ø#£Çh‡^ìb~–Þ}àÎøÉÜ·ÒÊÖiÐ%<ñ¹9¥‚õá/~¼€êÔ-‚[:WñÞûÛúý¦´›áé²íaŸ’rõ7úÝÏœº¨ŽÔlL‘¼oÛE ê(~ƒþ[Ž¨Á8Mjhr¤Fñ >MÖhnŒLN_aëC;rRr«¢ [^‚ŽK>zh.†—›¢ü?e } ´¦Mž©q‰r©š³ µB[ÕÖ¶]wü™ˆAÙ= ÇÉ7äïýɉ/ì bm=ÕFº9en‹^5 Öê¿žŸâ À&¾§¸ëg †a ëˆJ·a^>L:C1æņXT¥üeü¤‹Mßä”ÓQÏ>"[R1«éc —ÿîã¢x‚ÎÆ‘O2%Q=~†z;(ÎvÊ ¼HúˆùѦ\ïâlÊHÃ&OÂ`Ð;º3)ÇUϹ©*[Å·éhÓ§_çctòm<ºå[ææ[p³t{†›‰Š¾?øË"É\ß,`³èugÌuB¿dó!Ï^åî7®óÜYQ´ôÆ—ÃSNÚ1àåíYÄ®þúõܳÑ«í›ãJeÅXÝ)"øuÃ6(JašM 3Êis“*¹×W–@#nªFE 耨ýÉ0¯zÓ]ASÍ ¨>ųÂ#"º ’Îj_§ÈëXB0p¬ ߯l|Z9» I.žØ ïÐg@-·åô)Â’œ¢4lÇz :+%‡Âhrºs}}/@/ÊÚÝᬋ bbÅîÑ8Ö >€q/wþl ÌûUˆv-£ß»ŸûÃy›­ÞÎ;O-Ô=ô´ãñX¡´†5±çéôVq®÷²6˜'8U›öŒëR±øU†›K—¯¶Ž„ï§4ã6³¢ƒ&Dý¤êiû»'µÿø6žŒæœ‰…øv"Ê ‰4—´ô+ˆÆºæB%­0©—š³%T½SÚyø©êQ¬\Šë*Ö ëÚýçÙ¨d5×wÙ"jI·•¸AbUuZÑþËßi¨’·ÏÞAß ½^ïï\ È¨b©{‰oHže]á’éÓ=ŠRYüöA—Šã•¬y£ãHÏ ±Ò%Ñ“ÅNø þÎFƒ^^T„-E¯\ok¸%8ÿ·•<ï1Œÿ¤ÖûòUzs;AõWÕƒ#êù=ºm‹-\ ÉüÆLs%ñæÚEP5ækSô^wjÅráózÀѨ¹]=ÞÂArõ;RÜ[[Øڽܾ_³™à ꥺ<À Of šM66Þ™‘©øSw_¼ÕØ·)4øO³Î¯]jÄX¹r“x’à!åf]ÁdÙ¯óqû( hÞ!JòJóñ6#òº¦AÞ°áÈ:›Èžnü^>ü`+awïFƒðÔ¨IuÊŽ‹}Ë(ÿdòƒ-°…ø†ÖÖ ÃÙ­=oµ¦‡û^1â±-7B.Ü01X ¿%AøÒŒí¬mdz›o7Óa¸µ>Å#ç·âÔ|·«!k…¢²öá´Ó˜Ý  îæøpu0'®‹„ùY:¶w”¦ZæÜR“?|¨¹Ñ€êQ±¹pG¡á|3 …³„¸Qý {ÊX¬ºJ¤T‹o Î>ï]棛A—6†‘Ï<áýÌSY÷áJÇa9ËhÈþJ»Þ¿7B >éqq6¤‡`…ôdG] Ü£K2è Ï0›jŸ »­(ôaYÓŠÁ(¸õþƒiü=½ÈÈfk¿Y‚Ž„\Buã·œ¤ 9.Ý=Ä®§ýQ”æãy R~ )‹×c©€˜É?³U±‡¥n*dÝB¡/×Ò;)¯Ï›Ôs ;âX¡«Õ5$DÛ9I#NgG~k•XŒ¹š'eÊÝõmIcÞUˆ?ωzŽù°y‡¯°-µÓŠõ&‚ ð5‹1˜{ðüT›&‘WPõ Ùï+øŽj¶a¥²îÙÀîÒ§š 9* \ä•jA!JTNV£y=Ü ’jÅÖTük—~N‰¯‘H1ì§mx3ô7¢”¦!7ú¥W+»0Sõ¹ ì)ÑôãW‡æ ²Ë;uê.7(¨£|›9J¬ Ÿ½¾¨Jiê™é ­.?"ªT-ο&†Œ*O›0„bFPê_IJH-Êè¯v”xöñòþü‡µN|Þ«ÆîHœ[iM­7JÑݲP:QÄt#T”1òV6 I4ˆ{¬M˜÷v-ƒËž2¶‚üÓÆeØÙn0©-1 Ìùñf½Ï¢KF)»úBº€íºP®‘ ®“"ßöõf@Gÿã;v‘u"ÃÁs†oXÓ?Á:²%m¤Îd:Á»r§u\˜gÃÿØ󤚱´“XR…þ$Õ ýé ñØßõ¦×²÷u*¸f²-˜Ì>€gæHT…÷8èuósŸpÂÕ½ÙvÞM—å(‰4æIsäš š^Ï&)rnì/€á ìü@Yn…z?^ÿfgÄÁæÜ\o‘.ýŸwÞåLus©„f´ÆÛ¶0~ª±¹ILÌd®²Ø©s«P˜D“q®Ôö¬7¦—uÇ'gƒ§P«s#[§K›L ,Syéeׯ‚¼}#òz¸+Öká?‡W«î-ZâióéÕu‹wÁæ!dýåÉñŠ(Öj,æè>¼4—–¯b˜ì· á¯u U°”½2fzÕAÚ1\@ÒhNÀŇ èXI¤L^jyíbF rêVPŸŠø`kÿá¶Ë—ì0Ÿ.GuÂ'?¹ ž‹ä9¹!âç”@z_¨L¯Œ’Z -ðnï«ŽÆ¡¢¦!±¯)iåKîñp¢” _ýÙên÷UZ1æÿŒ6¼­´+=×=ݬUD£I;X½J¢»)™?g@¾TºäñqX‘îÈ柙E8òR†üà…¬Šo2øÛq}4?G=Ÿò.€¨­«1xôFÏ\¤6êÀü§·­€eò.ýceÜwú,öž¶o"‡•.ŒrãͼøaÿÌ^/æ7E±­ŽDŒÅ^qïYhÁB†ÙþÒ¤Óm~]á+ç3qÎ'Ÿø"¥p6éÛÙ}\?0Æ;Zœ¿ -]Xp&ç½}5­ãþôÁä’¼Ù­6×fbtìÉ·½«!ò‰þaÖ©ÅÎú•Œá­†K¸­3ä§ÎÇìñu>|´ún|j‘WØ Õ³óêû·Ôƒ55åI^˜Ü]¹&ÑŒ6à¢ä´‚º8$© .dóáQÚÌrùÿëâ›3”m[o2±mÛÛ¶mÛ¶mÛ¶mÛN&¶ïÞŸnýºêéÕÝkµNG_^—hŽÉ +®f¹c¨)Ç‹a1Ê{Ò¿Vë´±¹¦Ü%ýÏ0ÖaÛ`ßÕ.˜Ärµ¶!þS|ÝÊ›M{Ü謑ÜG8¡q`,iìæûtЃ\F d›µϲ¼¸{Ö!°èSÓ†¨rÈv²œûFÑÄ£nÊs¬Ac³¸ZÑ”´–Ñ㉛£ài`€ÕfáEõ.7'”• ¬ÛèYÒ Œj_M„C3wt–[0°E Z·¢k$?èþt9‡Yï}*-ÔeÜã h 9”ãÔ‘;}þÇCÚ°§ÇY‡—¡7põˆItB†+Î"Žî´Fc^ñt(èãrÈä8u·|Óî;B1½O ì1Œ’ê·ï®WèÓÞÿÀ¬†$ã@«`8š¬ç¬ç ;#ßÚ<ô¥úáî_ýdR ©Ý» ÷lÀÆiš.Ùšÿ \ÉáÖŸŠÁkkÕcÿ € e±Žý4g§ª&†Oã€xü¯™ 6›ù8Ò2–Rkoo‰ŒüÍeŠ fhwü^yüÔ[«Ôž–=Äϧ‘Ì–ÐÐœ>1p¦ÓÄ!$·Rêpç¿úþ:åó#Æ §5O°9HXùþ mò3¾c2H]:æÿ¢Êì1Ã/~Ëéðž™i ÙìThHtR«®×à LœËn"ZnwÜ'"ÅE0·ñ†#ž?@á¥Z†Ì@tEq輤‰1¯;,<~SÄkù·Ÿ”§Õíqdµ3m«`üÿhZèÏ ¼°ïH­÷·;ò†~IMÚÒq¹ƒÞøų@gdŸa ©«7üÈš¹ w*!®iWÀ2D€±¬™(hëR¦i‡sDB|@Æ '¥+…c;G¯VP_f—ÄSqA(ºo·åų¾zêV8pfn4Vn¦ ð¾ëŸUT+S¾ãT©kO‡üê‰e½°ëÛ÷·bGù¹·XT‰ yйì„i΂µJQ|ë5M‚à ºÏV^ЋBUÇýÆóXÓëCd|>¶bþÁ!‡v±XÇÝÙpCe»fT»y¯KB¡ü(×%³#T;HHe©(Áb´LÜ㟡l1éвHi@3Ù¿„`¾W4bêšø2C0DÃܹÞôPèwYÈÿH"‘ÝÍrBéÜêC ©‡lšÆw (ñ!õ* <7ÙÆ IUY¿ƒwºrgi]ù.ör%üJ†Œ1ãNË(_."ÿ†B¿ç¢mœð]ꎌ‡ûâdUŒ1K&ÎÊ7žh~Œýï3ë´Z§PÓÿt,8¶ï‰«£œA¦´>‹»’ï ï±m?4Ÿƒ¾LafÿvxŸ—{]‰a²€§X}Bñ&u¡"·¥‘3úNOy¬·Š|µ8Æ Ë àÁ¹£Lþë-®abÂÿbd­Ó,齶žØÈÞ/LœÚKÍ°«p±|^úo^0 ¹G>)§«‡Êé5€s<_9ƒrÍM÷H5$e;Ëg±+\,k„«ÞVËé{ÒWÄ#‘:ð<­JÿUm"M¬Ø1VH }í=€[i§ñž›A?ZÀi(ÖðæW #„}0GBņ€*ebbÍŸ„wï%t±ûA ½B~éù¸Þn*5ÎÚq ‰~„µuE“mÓG«$mìÆdUíäqM`ºI¹µŒŒ;Ó°þuQÚ°çp®òÛ Yñ…½ˆ×q¡fg­L¸«Ïãë Ò5Ÿ~ßUsã¡­£ˆ:(:í‚PLìiË0.mÁθ,½3,@rÆõQaø$‚¸¨—Ôm”C¯Aý€{ØA‹1…*råüüúè¶k[’`5w|Ý}L1q©H1Ö‡±8>ƽ`)ÿï¬ÎÌ37àpú„…왃UÂ]ÓwìYè¬Mp6·Ì­ðLO•Ô.”³G.ª­bÿ¢ËkTñ¾T5Ü s%/:2Ýð)U´* /¯ñà8Ÿ(µ gÊÀ˜zsô¯¢pôºYùGĈRà` 3X˜ç#ÌV°KGmƒP¨ ¶U¸`ïZÒAªª9ç!«YŠíë+Ææ;<èÕ7 çV… éBXŸ€ô² ¾ÆÛ›X ÕD¸²}Á©‡õMY`©Àaþ"!+^8§©€­B³ÃdYõž€®O”VEßbE»nY.ö³-¦òvÂzh)»óBÎw•³Ò“©QÈ¿¸a´.YÎzáPË‹T²†-Ä‹¨À°J¼½æl2•àËt3±¦LÈIBýRÓ¨Šy0AÛ¯|§Å<_A=I‡ýªü©R ]'?çyFåPúïȧø€¤¾×þ9ä7±áµÉ×J^`˜¿…»¥?^?Š“UʇdÚWüÃ4'œ-†3\†°ã$¨3dÌÓ¥xÒP÷eÜmÕñ¨P^‡qà'ÁYK?„æVâ?Ei§_a$ü !´sĹ0³Í'G~{—¬"áÈOWõŠå)õÒTìh—UgKéeÛåw§%I6‘Ñ—¦`èöXìáàÂwñµgþʺ)=z‚v2oiw•^œÝ¦*!ƒ‡S—p¸€Ôö%8‚¾a¿¸F'ëðòï‹‘÷ÅBß +•7R…øùõ™¶å®…md‚oуY›~¦â›GÅüÇáqTºÛ×é‹„q¢®ï:14g­Òûùm¬žÝ1_ÿ^W²ŒßR¿ahâ[±b­“ C{6""#¸Ã|~¹Õ(ê_•3¼ï…Á`çÇŽm%  Ó¦BÃ>pVM¹ßóyÖ•žótL³·å;F÷åç¥úå°qú§Ú~ŠÞT"lfzÒM£a1¯*rV åäT뫪%t¦¾Ä&ÍšïrÍsÊxDi{&pohçóû,äß©/$âH Í¿­¿_O„µb¯Ñ‹Ö—õRÖ=¿1ß.Юj»ò'a=­IÊO%"r¿¼{¯aåè›È„j²·3æ±—æ_üv™Âè‰ẩWz Æ ÿôZ·`´äzŸ¼Ò¿”שbÜP_H[’Þdê7ÙïuðvMvùSN£hç–ýþb…׎:\ï+œ§šãw_|×雡Þ0ýn-¡{(¿ÞfÒRpU¤M}u]g¤º]DÓCî¦MÉ7@N ®Î Ū(få,F`KÌ–@c‹ÐoßÜÞÁ‡J<„ÆŽÐí´HQÁxß%×zhÈhízÅ ¬˜‹B~Ûï£R «ú–Êà t…rÄ„Š…ÿÁh„¦«ÓÓuÚÇ`6Õò5a8n^jP‹BØш¯ô\~¥ÊË•À~Iš… HÈ…T4"š!£·IUwÊ3CnŒ-PÈ¢auôÅ€C‚*ÕpØÌ–öVGŒ¼ràÛ]ãƒU …bZ¢>0toPpµÅF'yλ!e¦1ÉÚ!Þñ~«(ÿ£ž3B·øé¡6Q)Zõ…kGÉUësJq—™v’NL¼–r.½+›•¦³ š¡L(Bwj9¥qÓýG¨»¹wt[m¬›ÿÃ&å•y6=½ö ùË°ñ1éß/ïÐy½…‡Òîñ8ø]}òk‰?fv"1§åy ‰ØvñàkîMN$.ÊÓWíA䣙%B­­ýt+DÕÍe*/{œH :š­<—/z´c~ NKû••39,Äå‘Û7[h8¥;2Ö@²‘Y[èÙ.™k5nóب‰qk€#D'SÄîÝÆ~tvÕp»ó«ÑÚEÆ{f~ g,põG^“xüÁÂ_Ý“I1eP©v¦µ¥êOï·( ?})m–Ë¿³®ò¼p”¦ö7ù4‹”Fd(l)'(v!Y”i‡Y澇ô@)¢ú¹á¥n¨…O¦JZf_uœ‘ËØ^)€× ÷Tæ‘ð<ögô Çõè뀘ԉ[>qò•ŽµÍa!'1A‡›ôÒÔ ì¥Õ©b5m‹ÄÔµu¸…~¢SYÐÁ_uŸ»§¦Ç<Ž@þ¿'ì’øIoNëýK @±Ãó<«Êž£3å½\©|C©™ŽÑêF[ o34q¯PÕÿ–±ô~iª¦ñJA ¿g0í¾x7Éæe‹„%ÏÿÚòíž<–Àw ´Å¨ù¹Ñ÷òÔ \MKóîOj]kžgZ¤,¼ $ÇÉI抿ºcÔ‚/åž °p—¹Чrfb€IöÃ¥_-“iȹ·¬3Œ¾e<ÌÚµ1 jM;gºªàêIÆsáàÊÎÛ/!¬5u笩â "cú¡°¸Ã7¬pð3GÙ³©'ų+Dø±°yãí]x?˜ÍM1'®µ^Y`xåO Fµ"—'fuѨzÍ+$jvѸûÅú©ýgKRxàQ¥3ïÍ)L“ÄkCxP)ŒG‚&IšN83—ÌáTIŸp;­p~¢YÔÚ|ø'Oj„U½˜ò¯µö³?ˆTÔÆ3/BÃÁ…V 1W$¨?¡p!¿¦^Ž÷â¸/ºë_ÙÇèH‰‘ÊþïhÍÚü·~¥1X;f¶ÞõŸ{O5“ñŒ*Þ!8„T)°Hlß­WÚ’R¶I[¤!IÜ ;‹æ³AèÞIi#¥|Ü{ê~ëÙÁnB™DÄU‘;ˆ>¯‹zWL̹?zÆ] Â(\㋈Ê"G*×ø,Š`‘ewׯŽx:Ó^›nUhåOK,é>ó¾!ôPthdŒÖ0=£ Obà[©Sm«ÓN‚ÎÏ挡Z¹7#ê-ãEšJ]§RneÚÊâ\=ˆ›­)"qáº÷J?ÎìÛlÖs •(”#JEqýDrœ-P‘«ÁR¬‹=‚7˜D^7PCþˆˆ›Î×ô6:mÈ:ÅC¢ßçkd2\žq!LÆuÎ…5<㸇Ø^ûdÝ"^(ç})jÙ}0šó¦¸y‹¿Ó‹¯&ŒÝ¹Ià´Tá†ô”°eiÏD˜}AZì´ðˆƒñÆ?"HSõšÃÃ|ع ±ä #ê‚Xªr¤„ÍÏäí};­ê%žÕ8ìº\ Þíz}Ἃ)d¬ ~†ð¬ôªv ¼n÷ògrÝÛÂV—»Zïâ¹ ‰ÌÃ#¿>|<“_³N³<R5‘ŽæŸïÝY%ß½=¦ZUæà9\Cþ™£@¥¶ý¥óPnBSi2°ÿɧHNB¬ðËŒ2Íálê÷-À3㎯H¿Õnâ"áèŠ4íBÛE†­,‡ÂŒˆgÓwãŸ+ÊŲj—n!Q_6œz8^/ìØÛŽ€Ü ^^4³³¦´3úÊÊÿˆZ=E‘H@+t<ŒJ™ ÇBJöz`n,ókÃÊÕ`î#ÝV.lÓÒ¥LNMUÙñ Ÿÿ=v§ÐåBìõ3¡)ˆßvž`¤ÌŽº{Å(•,úUKrÝmeé ½Î^çÝ÷n î<ž+ñÌüÔ´ÿ‰ù>{Ýf·üåígðü ½¡jÚ'Cü±ÏÓÄÞ"Ñå7±0h†ººSæè{ñ·«ÑŒ¼SoTZYTnjâö9ÉfxƒÊá¬Ù8â8µv*© èÓÀg·ÿÊ;^øêo+¡:CŽÇZ稣 R9ñay{–˜­ÙžÙ$>oZßÌ $/¹ŽH¦x†Šç¢í<‡NTLá+iÝ¿tÀ4ÌÆRIûõ]Mþ4_MýùýäHáÎ2!@t>êqà_:ÃP8$Œ€‰î`Ï×ûlß2k%„o–Ž¿GMäáVX©\›©Á A4-Ïp3¥×dµ¬ \X/ÎMÄRæKÁ ±/Y yxs/> H‡–q^&ú+|̹´fòÈÒ¹çŲ̀øýg§ŠwAÞI…ä³IƒÁ4Å<›þ·œ¢˜kÿJ ²9fßìþßáf[чPÇÈÃ6Ÿ:›C¥57œéƒnò¹¸V,`LÔgKl§Á(ïpmpcç³Û)¾¡3—2 ~š° Lÿ'`â#±¬x½›®Ÿ&‡çŠÂÓÞžåþ§ñ<¾´'DÕé& R9/hRíVÓÕ­e=‘]¼Q2ÆF ©Ì(°ªQÿœL­È%ª®4óÇ~'qéHžæSÜRµ6gÓç ¼Óêø‡éXYéò‹„>Æm!Uzk1w’È¡ +ßóÖ»u=óÃiÈH18e(ðÀU ÆO’ˆIÒÞJ¶‹ògDÆ>[>TIIˆ£êÿdPæ¡kôÞ™»o"6†öµÌ9N¿7þïÊ®ã?³T7°3—ey¨=u$éMoѵ©7‡úèRMXc¬ó÷׬Ì]ú¥¿ n¾nbDÊf¥ „tm@|ÔI`½ü9ÌÄØ—`«@ŽVŽà]]#Á¾6,/Ê`ŸðCl¦m÷°ø÷°fæ6‚öIˆø/¨»U· ì›¤scCN_Ô¹*ÑÊ'ˆxýY®€zM½Þé~ºhŸcáÂ:×z{©µ½V Ìì×¾û@’ÅÝ»Æ^²lf¦´Øt¯ë¡÷õ“NŒ#Ò ÿÿô ±9TµÓB½x‡wOKDµ>;õ&þpjðâñħ–þ“>¿cïHŽd¹.E†Ï‘lï fÆ_P=®^nïÞ ¸lÛcÄ‹Ob¦%Ý­§§m¨9 Á¨dTL¡ðcx<‘7”MB8ëöùþL̬›¶D•ç7ÜKl§‘´qì¶3Í.ød" 2'»PûTÈð‰oÄ–/äÿ¿^½ô§/õ@¶púý…¬÷ÙÌà„º ±yüÀ­¢,Ò<,—“±½2 (µ4#À>¯âÖõ§ÐW¿² möî ÇÅQfªœxCŸ<œt\§uóæÛ]AtEØ|G\gà‡ Q“é,±œímqXàÅÍ)©tÉ×ÿ5èœù<È7ÕÝÚmXf%Í5)"Ü7Uö€¼V hNNGlV…>„N>ш|Ø÷SV˜nÀ½û£‚Ëü6là¦×#;©ìEŒ‘i­º>jÆ,cè±ÛPò0¶ëL‘ }¨²Áux™ºs~%Œ­ßN÷ÉyP€ñÜÁ»u•ÉL4ÒèÉË›()—´‹’òTœP¡“” ,ÞïLŒ¬7„m–Cê%=³0\Æ1’[§Vø!€[J¸ËVŸA‰F‹/ø l°íì°Ð%XO«…š+¸ D°P*›ÿÜDaÔ3ß%å1º„zÛ§5·w8·h÷&;f]v¹úM¶Ž² ¥EçsºÊgîs´^‹×nÖÚkS¥E(.6ë@€Œƒ oÒ‚Žaº‹È©Ý9í¹6QQŒ_{¶Dcí¯†  ©qðõ?±æ)í¡Ž£ÄûŒºR$f`º¿‚½h¦§Ü*2’¯ ±ÄåºY¨eIÀæ£{œ®Rc©ðï\gé UÞì:XYôB\j_8<˜æ?ÂÎ~5ÇW­¿(­LÈ{÷€À³Ø2eÊóü?sí¸ï‚í>\*›ŠUw/KÕûV“8œœrñ¡WF¶ÔKEË°ÊB?þ:ðPí‹V »e ¹jóRRþ j€"mÁÏŠ9ð"z• ñàÎÎí+œž†ƒ )ÿk—äR;J¤)9ùÐõŠ¢®sÚ7i×Wž:>îRŠ-™|&ø*__‘4n¶G¸bð|? æUi¥y|ÞÌ•î›Ø`ƺ"ú WÃr/¥Ã nνŠ@ĨÃÙ‚êj^Ôœ ?š%M,á‡öH" #±’óB¬g¤B1µ”Ôµµ tm$©³Æˆ!PµÎéZÛÿ­ÞžÇe†ówôYv[P”ÙrÏO¬kŽÝ¦T{ôʽjgO!ö Ø©º=¾ÞÓB-ÑâòGáÈ6Lo÷ë1¯«ÿÍQw)q ¬Ar·ûÌŽÚˆ h†RÑ[Øâ.™ïðC³”t¨?ägž÷DØlˆ8”(œ*¡á4 ®X ¯‡f䎉Òø­?Bíùw`â;v¯O“<¥muR³D£Š|È¢ŠY×É~{ž±è.vÜ´ háh|Íc³ŒpêÞÿmð䆧8ÝÍÅ÷å·t ÝŽå¹# ÝY³šxøœA4Š« ˆDgÎzÀk;Sª¾ŠÏÌ*;ïl,Y¨¿Zo¶¦óïPÕTâ EÜ«V.¯‚}:ª­Åzm û(½ž5»)ÑGYÈ䎲¸ä56Ô0YÂ:/©.ÞÎߺ—å‰ Þhp5¬YaÊß~ým¡è¼J)Î ×zÞf|‰¦Æ ¼TÂé¾²Žžaü¼4Ý]þ½ÔŒùÙæ^k@{«yRÓ¬Ä|áûùôrø~§è¡ Æk<‘'ë¶]·¸5Øùkö½ÉµÑäˆÜbŽä2°‰#=.ÙÖÚ¼³Ë]W•Ý§–j›OÝ`òƒÊf³ µc;3¦uTó"¶%±êe`Òv±Iœ ŠhìXH—DÞqòOùÜÞ.ÝïÕƒÁOÉcvñÙùiÿÊ7‘—mXE œû†mtãÐ[†3DýNoÓòƒÃ»Í4úAæVóEOÔdͳ¢¼å_P3óäL¶˜‘ªwmé;[SjïR£DDÓiºO­h¡EÓu}»0×ä ™üS—~ %ÓCÒ®ºœ®ä£[-nn lnþRWß̬låCžW£5‘0Bj53È`±Ž†yLîá-/E8:¡Œ^YÙg:ÂüŒ¬IŠ¦^8‡¸­ØØÑ„Ï.–^œ>Ùf8ÏVœJ©áoñ› ?Öp¼V²ß#™1ðK4_%*e·W(à+#]£ßðm¨×Ct‰GãX(%ØíP÷",ð?øöÛöþ5ïOÒ´T8;>°[Í«M?NgòµŸG :ÛèÝùíaMÃð"\%_Ž¯ì+ßc ›†‘2¼/äÅùÚ‚øM@å)¶ˆ²Ai[™©G÷d¦¢(¼#ó¡àŽž¨ Þ°¥Äi ÇåÈ£3q˜u}ÆâÔ£‰`ãÏ*˜‡y›²ó:/Ëp¤Ô–&?šßÚq'$ì´.hqh¥94HD·ùU·O¨0N™MŠTùä$Â#_0-v ”ò›¹O¼óìºð=¦º”a´™wOù’6“‚áµå«Á°D¦#²9:SlÂ-&N9µy‰Q¢Çžµ7‰Ûß<2©ÍÍòb®Œô=O]Sbÿ2£82ø±Xú¸WÃwÈÇ”J!-N£„ˆÐ$S¸£lÕ‘-€t]+ 5¹™`°ÄZ“_®=æS*1uúÙÞr â Y‰Õ5ê÷Œ«'[€²Ò°×¿}O3ÕnÓmn›B8ˆüüÖîe~,9jô[Î5ïì8mô&µ|~w"d€¤‚ö#­ WÔÔË6‡5NɾDyÝMün"³ïÂán­q®'Ùâü—Þo! Hð EX…»ˆ×±ö-†øsç6ÁÏÌmm•,{H5^ÏT0×L‰¤g¤…\Wœ'Ïq àÙ ŠßfÜ1Ðãe¬:UègeR.a7D€Õ"m·á{Å]ŒHJ°:jV(„éýIBX³3~˜°',…Ãr €£ÊÙ&£@;òöûÇf…åÞ§‰B Ö  ©@ç%ËaÊŠôÞÓ>:¾KO¡V¢ösçÇY“í4Ô\X€`zt†ÏZÆä±¹~wdOÞŠah¼„%N.ùÒ¹i`‡AeùfyWk¼6¶‚5V wYhä(ÆŠ¹(äCÔr´ ‚›Ï÷›Ù :œª,¸©ò¡±û^¥**%}÷^ËâJ`§ÚMs)„«ÝAöC¥.·`LÃùE*".ÇUe:w,ŠÒÍÌ!K4–sÁhAn¤çЋwÊ¡«oƒ9‘”Cú_õaFY^µ³˜é©Aó]¹õÉv@Ùz¢bÌÁ°?¦›ê~Me7ï—rF¾n ã›*êÙ®ÞAÏhÿ ÙÊ¿|‘::e„ö ¦ƒ7à]É›aï3&¬u% uµ5ýÓ¯ È•ÐÓeÚè?ðit Ž &fÎKkK ‹ ³y)f‰à;3T„«Í®ï'-7?î< ³°–KÉ5t1Óƒ9âfúz I²0¯¹]+8¿TþžÅ7Å^0•µ6Ž0ÕëÝZŽúMâ~²^WSçÑ ò*Eòç:píD5ÃN´v¬5RÎVП!Ãl§Qz'ž·[kxdÊtÉÞ~û¨>rfš0k{'9 ÊÜeØô•·஫€w~qn_T *ü|ê‘¢wÖ× [8øÕid¾oäxFúu”Ú•MH|d>§- D_E=qÙìcî”à¨PBD^²Ä~ŒaI1ß”çŽn»¸µOq4×ÞÞÑ´:¶ÉK0áµôö ä>_·ZXe‰f ÂK’áX‡oæ< &¸ýß°-ÚIZ†ãya'¸†ÓJ!u&_–9÷ç³¼Aø¹µ²•"Örô­Þüpnr-LIµptEVF'‚m£ø…â+¡}±¤¾7èU1A¬x¬¨rJ½’ucð¼¹–†ÂuaÉ:Ê`ÆI!uÕâP>ÀO‰½ƒ-ð¤K¶0–•  «³ÑLí±Õ%t6Ç™íŠèvÉ#Án ¨’d^ÄÙÑdϼD_̽¯Ã¦ÇøT.ùâßq±âÏÌÉÌ6¿·QG­š@šÂm5Îß+±6}ƒ`kæÃCêÒžµrÙâ*f²‘‚q6˜›ø¢öê}yÐÉþý¡b‡Dé°«Xóúó¾k"¸Õ3V6ÔÏš9PÖáB˜ò=THJºÙD`õó!—EÕ€!QaNáŽ5¬Ðë0RÂܘv°ã…H¢Ç0”¿fNWnEEœ³9*¼>Ùñ Îñ£ Uh……Ó;ãVÏ5•öÏ :~PÑ{÷Œ©ª„ …ÈÇßÇtûɾ(›M_©û'¦lÉô,K#Wͧ–.¡6mÜÄó w÷m~4MÄ/,~ñ£©éY¤°uŸÐ ïÎþI%€L1ƒ¼ƒÚSã,ÉÌ‚O™øQlQQ»j8:åCMÓÔPd&¹›6;øWƒ¿‰d¢u‹i:-£Î!L¶q¤§I3zõ΢›Î:/Æ{üî/oñî’q‘K@W}¥¿¤úê¤|W$u êßT^f Ã2èESu®ƒ¬¸0èþýsïg%9Ðzîݺ¡–û®Èk“üóÝZ2ý+Ëñ´Hþ+) ­Äa Þ¨'¼|×¢‰–\RŸãÒ­õšŽt+{Kñ9‡(îPz«Á<È×ô¯‹æ0ï°gX=b=éKuNç]» ùxÕ<6zíÞª}ÊÂi'°4‹k.r¥­?ïq&ŸÚâºjpÂèöî+Òð]Œ®#Uþ®*……ø n³½ìWëÆn“yÿ0Eÿe`1‹Úý“Ùñ95êËØ,4ÂLéÅ–ÓA¬¹óÑgˆlÄ:6h @Ó}¯ÉCmJÕE«ˆËôê é ¼!K†Ï¤Ýlšç 8ÞÌÊPa¨¤n!Žocg=οœßó'øóé\­äk‚n.U(b¦ó(‡˜dé×TÛι€LÀvckFK Ø"H ŽD¤é ÖJ6)š!èÝÃÂdȦû •%iÀû~¡ñ&ø·=Tñ±Ì—ÜÀÂÒˆ=ç2Š @¢’¯½Wºå9òg°Yå KdÛ›àô0ýXéý×ßõøX& 6,‰Û )JÍóÚ^•¼lNp­…¶'Ý0ÛNk­ë«¡M¿‰ ÛâÔÁPßݨBç¾Wˆ‘Œë¡Á„ì‘ô¢[Ó@8¤ÂŒ¡„ª|T4jæ’½;6\†uº—4 D•nËB*Õ¼MÚ^ùõQ¸=‹‹9Àßrë$LÊ“.±•*V´ºÜšÊØྣ×ÞYé¶5œ‹õtäÃMXé,Æ÷-VoP×L•¿ÎÆ'R—?@¢kj õߤU€rJŠtçYáô• H TÈñ%«ab) ZR!*¾Ã“y˜£_bV˜Œ€àþ‹,S«-o…­Äû§ßÝ?R*$¸ä«¹S…"(xò…øšøaÎ ¯ Z‚Géx=µèçöt¾Ùì×9»í.…†çúÍÂ*Öª™HHÊ%cD>²?ûV%þ9Q /)GZ÷”©ÖiÙ0q‘ J‰ÆÍ­Ö¡Y<+u|«>ÕóK«¹Iåï•Ù Ükí .…e“ÛÇêííŒ1³¸Á`§íw3i(Z,?©Ÿ±çÅb²èîãê°G»*îd£´-ŽÃdk0“©ñIlS«ÐF7–XnsæóRéÄtFòØ›ßôç‡7D56ç…”>-Š-7ÝN\ ¤wÊo¥¤&H‡£ý­}¸:£­ƒ¬£.àÕU×%džõ®+$‚Ç0|3ÔGöVFû•b)ë|Ó©^¶]‘Ö‰q‡žêÈí0zIœ³Ó¿šy L†Çz6ÜK¬ 6БD’/7X!’ %¥Ùx:M¶¶ðË3ŠTP’ð˜4¡Ú‰FÌ‘A×@×ÝÇôX¥ŠŠö“BÒênî¬tp¹ø7ïMÔËAñ¥k<ÇÑEQy{dUsZû„Ó‚‚´ Œ±1Ü÷Ld¹¿£œ¤­è޺ϸ¨òa/H÷Ý#Vh '=w ûV©²Aó¦]ì‡L­²Û¬¿}ù:±öÿ­ÏîÅ?õðœI#N&;c¨ V„+9 Ä à‚9Fþ"C‘ÝûZ«)ð#W{;­ßm…pM’Å2ãjŠDñûÒþáMxë€÷ÃZo`¹cÍÏ<þ\ÒJÆ\ã÷²‡„§DÏö{ßvåC)¾ML1¨Dë1U(=E <˜zPµNFŽ˜“þ™ŸòÊëÿþ§náª/]ãá„ý©ªmVI.âÛ»lÛ§¾6ˆª“óðKi©g?aÓÄþ#ôƒÉ Ì Z†$»9:9é3°¸…‡¬•ÇìÎAHC—¨ü—æ°†ærôÍ¢"n=b[Ìl 7œ:Ü]{÷?×£öŠÐ‚é}œeU±îÔ—Ñ~²‡íÁ[JŽÃÓ4?ÈKÛq½\ÿ͸¡FŽp J,šG 'AgH;)ù0úfÝ!‰‡/o^¹åýmF…0¡/J’Žˆ·ÔKúˆè?¸Ô_ xŸ$½9ß0=i"ªÒ²o›dý\´ò¢ÃNŠ Z¤Ú8&›qÃp×/¾*FýP>ö_Íu½ÒZnš$,*€Tæ=‡Í[]AÙ ÎÃOæ¨/0!Úø0Žxk.óépBŒ pmÔþJ˜¡Ä£°9¯‘ô 3w$\LÀöW™»K–¤$ê'»é”'!§s©?µ]lñ£l RëÍ胩²ýø‡ Ò*?ÏÖá\_×'m\Ù¼ñj¢SéM†Þn·”8—º#±²ÛTäÄ=™) ’×!ÝmÅÿˆH¥±ºP©ºkÏR_!¦ÞµÝ|ê%©qéí„ü¼A´ï² ‡½M¶Õ]€šÇPV€ °{ò,~²\bH^ÌC±šE†*›Ó—Kcß´;!þ$]I\Ø Ñê›°ç'jæ‹2Å‹‹”ÖG™/ö9UH¨é§ [üK‹FÕÙÛ=ÉÞ<€XÉÚ©#Ü&ücíDD…t߬øصÞ­ë*¾uA¶­ÂT„Ú$¿9E³£—qkWE†_?µÓPÛû’3Òµÿ¹*?üÇ>—{Ù>ˆÕ|½LšbË8Þê"å3d郡‡ ’WC…¿¼VÃgãpÉúÎcéunŽÁ1îßÔW ª³…X²¹‚ Q»P9A×ö&¤:X¼W}¹ˆ!犎Q+Ê!ï±ÖxH1”kèœÃcÊÒâýIC´Ø' €_ %;¿Nér:õe®HéugsÂDö±NÕïZ-ämaTüóã„+CH/Àßh*–î}×ïôF!Ŧûþú!B1£j}õ'"$(ûü ‹žJîlØô“$œa×c¤DZóÄ…éGæijyŽãÈÈ)N3ÈÁn÷ÀX£êÄ]Š8|˜—”1£ÏQªԈè3â.»hf¿‡O c¨´b~ e8Œ÷]½£»ßu¯p |'Nªu&MÞIœK•Ksi“O¼5Èþ+ÉGË—]:V©UVZ¹»”Ý}<…Åõ–Q'ù3H*e3C¥Ó =Â|ØØFî—½—ãÐô–ŒŠ‹Á¹/¯‡4p"bÐÓ%Ö¿]dž6ãO>{†ó†b•™°ËW¼}|5 á‡8 ¨›S¢ŒµæŸ}ºŸ 'Á‰=½IˆÜל’¸Ä3àoóòLDíÁÕö£~Eœã¼ò¦â;° -:eŠuR»‹Ðî+…èS8„û@µï‡Æá/Íý7\@©žt…í°`PG*jæ¹ %¤ãê*óRÜËö󀊙O¤G§ÌoeIè³iò^ƒïçT|_ÝsÑ.ÚüyS;f…óýÈ_DΪ GÂçzlêš#&mÿÚµº¥«SóÑA Gƒ‰×ª ª$­ï´e ‡+¥EWuÑóÚ}®ýy^¼–º#t±úSò&­ýßñY£õ›ÜŒjW«'‹;­àŸ¢ñ€¢š‘µ¡o¡È?FuØC[n^zs „ Â'D©]qMÛŒä«"cM¶ú¯BÞ¤tá­Ñ×WŒLêšòôÙç§ú+‘vrr@­}s­åFÓx»82H‚gA§hã?zÌß L{²(Š¥ äs“›*_*º_TÁ½Þ‹_r·?Š¨‹xòÅš£WxZFõƒ¦hfˆ¨ÀuOšèËÇNR3øÎ]ÂmH=µ—ëüj‡‘ÈÖ¹ç{lSû¸0Ò®aL|M&?£wð+ÿ}‚÷EgIX|3R±+‰‚/¸&Di‚Ô™õSÑê¬Ó›9O¦œìû•rÇà(P ÞÞ?ƒ³LÛ2¬ÈjáG'reõQ7 Ød Y÷_+ø²ýdæÒWkãØVÿ]O#æb’fÈÂsÖ/u܉öEQ‡ðàDÛiA¾ŽdeԤݦAß!‹ê»ž»Ã gï¯(-#z˜]?±·_;¾ü7™KÏ +#J¨oqÏlŒ¡ËåªÕúü:ûlKͽ1b¬ŽBâF,ÆîK©ØH3sOŽ~ØŒ+nŸš¬RúÜavd{Rrö£2È2Ž  ¸&†¿¾É»Z ŠPþmwUÐûÇΫVôˆè“¸qû[ÁˆPÙ„ ù`…„.sµžËën ÷J·#Æ÷eÞ‡5G?/åæ“:þlŠù’ÀY©z¡½gÜ9ÕÍŠ@E–xÇ‘k-0îÞRÀ0"`oÂBÄ4Æ«‡°ONÊ—ûèM;¤(çÖE*`™«¦|c𨀀>w…Z¾me}Ù`RàUôý‹ý] ùè‘ [–}Üõ‡±§Y4¥§–¹5©&ŒEï¸æ”+)I%Æ¿|ŸÙÙD_à-y“¡ákõ¤Ú׺äÂ'™ÕåŸr;ñ*47Æ퇿1P-\s¾Ø ¦Wˆk•¤­H¸…¯hRIQ2X~°Š]ê^vj`?ËÃþQÍ,ÕüáEû ^1w§Leý;—Ð7K=S² ü§*ŽµÈ‹ªÞøwwç˜Mš¬ˆPÚÏÂé·elÞú|ˆ\:‘þí[šs3*ðÊnd‚þjo~ àû XÙª¶ šŽù“³1hž²0v2ò\ŸN¼5ßhª”†ÉÐaDÓ´w“K,»f°Ñ²˜;@s:*8ŸyAƒV•>4¨„ âZà ^4Šü‡O Q†m )wuä´{ÍîþªÞ­[j9àÈÊWz·sx5¤Ä+çš$×.ز²ŽžBÕdXPjJ .¾r¸Š6ªÍAP„xŒ€¿Í·€JÒps¼€£ò€VP*ÊøÕ²b:ÀMF—÷Ê¡õ¥Uþˆˆömu€xÄ·ÍŽžmk=¹R…6Ó©kþÛÓSU„¬Žž5vþî*[[·—')˜-eµ# rfª—pt°.‘è¾5ÄÅΡëÐvFĦ4/«ƒF1¢b¶tòÓ ˜ŠrÏL–ˆÅ|–åÜÓ×ú18&Žúc©êFÉÑ <—ZŽþßF|—²‚  ean\ñ†–ª¼` ®ñBèÛñ_ÿ3(Å¿[ý}¹™ûM)ïpIºáLI›{e˜Ï#ß›Œ;Ó;¨#h Ùwà ñ'Ó ¿šð¡.àK*$Åj=ý·‚Aþ¤„úøùš6r3÷FnÙH‘X?ÙKê6&ÁƒEìÖ”¯ïêc }Xr±*Ëëõ!I×;˜O*N9;½¶c'Ê“Œ~Ia§& ›gªÃ©Âóíû…Bt¤îÿnèq„ÄóSXw*Ø ³D¾gÉ:5ĨÔJ³kìêå ~Y"„1Øõug~å?:„ üHÞ÷íÎ&¢£AË°¹›Çáò¦Ä[cuž½œvõo¨õËäµVÆ~GBßÂìH ¨8¢jtס?,º”ýá§7Ñ·–HÑæm¾jXÂœž.ÄlW(;·3 “ÚCЯ]µBµ“¸ÿwTŒ¿x¾|³¹gºŒÞßq…m®‡ o¡=»ýº¦å¦ ©^å]ÐjôPd%!i™±ýµµþ8SÉñ*Ô@ãr´ÑØ–> !­ÌÙŸ,KÓ6û{VâÿÜê¨hÅ€A³€™ê¾™·x²‹ƒMÏÒëtë³,‘â%Ò²Rj£Ì݃cØSJ'Dˆ“Æøv»\#8Uá¸BXNŠ£"NVx]?žš0Ùƨî—û˜äºhçâ 7½*ÇtùgÚÅ[$6ñ× Î²Z!±êÂcêΓ­ ªé\*pÂh“ôº¦%¢T_dÛ>gœs/â:ötéyDÖ?Ô××_úœc‰+i¶—tìæ¥ÓÊ“xci‹ò˜s\üvúõÃ&Ÿ=¢8›ò þ÷CYw ÔŠC‰^¥rçL!λ=½µ`ÂßNl¦ >aëq ??÷\Œ“«ñà•Í|¢„­R ±(Œ7 ß‚‰;5…mbú;·O¸§˜šB.µÑ™5úŸ_eš0Uë|óÒáÍý‚L3íË\ˆ³YM˜”ªy)Þ%åz°S<‹^ϳS’Ç!YI5Z”-!~93ŒKÑBWˆAɾî¼Ieÿ¶³8­’Q‰ä[ê|´·8"š"íî›ûœt/fÁÏòà .«è綮'À‘Œ¿DÜY&€B˜`.)`ßS|³Ê'qlG›X8ê7W™ Ü:55“ƒÿ ¾»âÌÁ‚.YOÊ(„8†Æ~†Áã=z’:¾eÚÞ=êÚR<ýÃ)Öl¹ëŠ ù<…ãR€Ã€¢»/Ð.½ÛÍDÿZÏ]ëHçQŠœt>ëqü°ä$©_¾Só÷Žþ)ƒeaMÑ&¼%,˜Ì á‚îÿNõß- endstream endobj 298 0 obj << /Length1 725 /Length2 20783 /Length3 0 /Length 21336 /Filter /FlateDecode >> stream xÚlºS”fͶ-š¶mÛ¶mÛúÒ¶íÌJTÚ¶]iÛ¶Ui[÷_kµÏÝ·Ý6_úhÑG3ZK8Ø»ªy9˜é™˜xˆT%ÔÔ˜Y˜˜ˆ˜XaÈÉEÆ®VöbÆ®"M€‘*À‘ˆ™…è3 9‘¨ƒ£—³•…¥+•)õ¿ƒD¶ÆfVvVÎD¶îV¦–D|îîîBî.n În ÿ©D®–"s+[‘¨¢’¶´‚$•¤‚:‘$ÀàllK¤äfbkeJ$ge °wP™;8ÙþÇ!2u°7³ú'†°w8»þCÌÜÙÁŽH^\MXBQAH\”QM”ÈØÞŒHNêŸr{WžÐSW×uGGäú?–­ñÿX&ÿcýOÖÙó? 33‘™•©+‘ ÀÂʆñ_ºIÛ›;qü'lææøßÔ?„\þáGDõ†ÔDfóÐn¶¶ Æv"*Q;G7W€3‘¼ƒÀÙžè_Ú{8[ýù/ÔØÎÊÖëÿüÒ®Æÿˆ!loñ|Lÿ Y¹HXyÌ”¬\ÿÑÛÕÙ ðŸ°&à?{#0³r³ûO9àMþÙU3{[¯ÿ³â?­ü{=FmIE1Úÿ‚§ÅíM̬ì-ˆT]ÿÑÓØÙìÿN+[ýg‚þËå¿EÄü}ycWg+O"]&†Ó¿¾ÿZúÿ%"âàéCÏÂÄBDÏÊÁDÄÌÊÆLÄÅÂí÷¿èšº9;ì]ÿ-ú?½ü×ÿ÷0žS˜õSÞPë´Öð ñ¢¹JpÚ¹#uÚfàÕéž]d«!K3ã+œ{5œ|8üÙ þ½Ë]éûãxÜÉäT,q¤þ Îk i¥…íö»rÁ[C¤Á5˜¶©ØMÔkÆÆz°Í[Vîy­ŒS?²$s¬QyG?8ÒNÂÚ⣽$µ>%J¦Î7à€¶iÎ2¬ðÚDðu$"?dî†Ûв°"Èñ(B'ÿàѸù´ý5ô5Á­Q—<™†î²¢¬Ås‘ *I÷´½Šg QE)ûjd\óRÃÝqö£BŒÑ€ÜÆâZ›H"½Ê± (Íø¬Án-«þ¡qk×7µ\/I¦]Â’rᨚ‘ 7v(?ú>ùÀ¹+8GÖüûϛɼZTnE[`㜅èLu{™7©¨Ú©»£rVîQ(o ÏrV¿Oè`;5…ÁL±»¸xr¿\tÖéì&ÄÄ~CT[c>éý³´^§«¯Ô¯©ÃWφx™?ݦ?+lSæSÇæ÷/ÝÔ^FÅÈß6Uàfb±a·Ÿ¯ø²‹ÀØŸûÑ0e€m üäíû÷bê%G~ ´JøxDõèÕÞ¨(»AÔ¥îêó öæ¿j+ße¯#X˾=Ü<íqäJî£b Ï]ª¾òzh`˜“ üTYðô8ΖÆ.C¿’`E2 ÙJÒ¹Ï €&ƒ“”¡„®ìu¸Vr×66»¦Sìs*T”B…!»Ô«6ý"¹ªC·WGSæ³ÐT>zEɈ‘†>¥P~ßßŒæ ±rbB¢X¥äK¨V éw†á,½”¬2’f¶SI%¿ö’H ·« ®ãÝ6„¹é¿ÀP}màúäÆî±*l3wáð A鹬nx¸4¶8Û Šâ1ù¦vÏl0e.Þ?Š=ÌÝõ»Ê×S²ûVc0–;J–ˆ³®•_Ârù`έ7è„[ôÅr÷A?™ÂÚ½w¿Àxà>Ƹì–ÅÖ¶Ò¾½&?cÙ^èAÂHÆ{^‡4ÀOÏ|U!PtK:&ûAy°´w1£-ò{Ê©ÅnvÏS'Y×Hh²"ß9 ’ˆ¥“üFød®8”ûŒ¯ˆãlbWsÈÂBÇ´þè|ìrÞ‚ÁÚ³D )y‘"⻚!1ýBj;£Å}¶r&?ñBSÉ®'²%[»bO zQp¥º™V\‰h$ýzOmiÐqZ+úx)# bo5Ì© ¢P-ÄR±,T2:å;xnÜA‘ì…ÎÀ¡ ôM@×eŽ‘ƒÊroá“ID:òôÜ}$¸“KK7pRѾØS”s¥…m-=V·b’o¾8Y SìYü(^M\Z·n Œ¢%Èdd~»y럻š(ZžÐ"‰à'þIéÚ™’gšla Dø aýÖÅó,XA[•xã@GÔ3 í=¶w@iT§Óþ–œübÖÆÓ/Ü¢ñUiï­ú_êõvQ(T3èmŸ8vù€†]ÿ'«ñNþ|cÂ=fÿn¬D*^r¢t¡@ÔáºYÏk‡2Cn~©wºžÄMvÙ”8?öP£Ë åNx朊&¢ÒQ[ƒç^C¨N>¦–H'¿Ÿf‚"Ÿ/=W]ùˆôè‚>$‹7›Éùˆ ¹ªÂ\ké/-%‡q%¿m`9jòO´ZyßÌ:È(e{±õ#,o@a¤Äàϳ#CžÏá<ìŒezðùíPý3þ·7GTä×\ƒYP‘Ô7uëÃqÐÒd`ÉŠ †LyA¹íŽ-D>¶n öÉÉÇ ÕE `à<ÑþØf­Û¹ÐìT°Õ÷ôä§O¼z]!*HÆ#ê+›q9þ·ºjåZîCl­8½f]J¶Ú+Z쾊í§sÂŽðw‡ØÂ$¹%=íp¸®;y'½Gåz3M‚?ÛÖ$É/ï#¨A’–$Qöë-ð#–Ô|µ2bpˬ €†¯E¯Î"U _Ý{§Ùéõœ(?aæåÖAΚŽò›ÿ+“r61—-¢Ã"÷Ç<³+§üÜ“F0´"ç$U%tÌPÉ-uºNØŠìNU2«êQÞÒ©Y…ãæC‘p:„zŠU> ¦ƒFn„Ë—]ÄQßù ·i½ˆm–pážûƒ[ˆúcñ]ž¸¢Õ§˜ ¦¶¾V›=ʾ-6E’²eÜ6áÛ§ Ü©_c2n09M}ÑxÉF)ì·Æî*PHh@v9.èÂîR"¾Þ> u¿É'¿“HÉw· ~YÅoQ³œÍ«‘†ƒZ¸¬¿'Çøž¯Z£×ÄÚÑÌl×¼®. „¤æuh½Ö{’~CS8Å/ún‘‰K±9ÞG?ROBGµWØ6ùÿŒ ³9¸Eœ×ü"¾3PË°]׸ n æÛRJ»Î!¨ÈÙ[̼ÖÒ¬ë«Ö½É~È&UU§×·†Þ…°CéôÕw¶Ù¬ƒÄEÇc•gÁ{äÕ;Í⃉DTT7웪ê€yЈ@™ý”s²¯W,íÐ}…tóôE€ÈÛЦ’^ÉÊìBQ3/líqôlëoˆ²ø{²™Èzh#íÇì5E Ý®FÁE.2¤FÑ«!$:‰íÿR§@Ä.6&x‘´JTôÌÚ „4ýdÜk,²Št²òìmê`‘*@]–í!˜Q(_×”ÓY¿ ü y¢üˆláãáÑy…fì\ðl&F_db"þZㆇu\Ít>(õ=Êu÷aÂJÍû9UU•BÑtŒ­ÐÏVšö¼ÃÁHoIºÜS0ÐHQé“( .n8+Î^&²T6ÔÔ|‡”¼?"=DZÁõ«Ž6Ú¬ntús_ÏYÿ>e£Üð¤#¨r³2;Za²{³+jêT2¦ ·Ö†.êØ|ØOÃŽE뢆½±IUÏ.ªƒe)ýǃ¡˜*.£¯?O‹ b»ºr\BÌÝï ã`Û»R!i§ÕÔ8Îl5=mfU¡eº:7¥¸m¯× Ï“ ±.¾ÞIyL‰¬‡ñ›··Uí–PÝípxcxåÀ”Ë-æÊ•ÀXgÝ°jx —ÞvÃî¯ò³1šì„´ýIYÍ­—zòƒ.÷²©©p—¿TvMâØ´õûóI¬~5ãÈÀæ ç>ê¯lØÆ•xº%×ÉJðU$Ý™ýŒÒ (äôN`¤JM5Ùþ„Ãw¯¨¢1‘¨‘žÕÃC½¿š1¡ô³áïÏ—Î#|àA R ™Í“ý%d©â؇j4)Þ ¨‘—(lqºDî°O)õif´A.Žrã(™†µDPã¤}%[0‘ÖÙ¦8%sÄ¡­ÖaH \EëU"Ýeö T‡æ[Ì.¾¡ÙòÛƒ ”¾çpÓäAK€Á‡Í@bµ¿8<>GÙ¾><å‹#»è?¯e±„¥†@¿Ã“%Áú=þ¾:ïçwüÉœžÄ3rr¾öž~íý®²Hˆ€âHÒÕ8T•·¦p@ŸÉ~5f­åL©×ôNì>‹_߸"Õ³àgé=™ˆ†ÃÈœZ¯uͼòxŽ®8x6«édè4Ÿ $æû©#^­Q~ÅÓQ#lÚz.Éÿ¢Ãú®ÃŒrÖûžÙ©¯hU¤™Œþ>¥ÞÒ}[õ°"Ä÷¬îù‚S€ø`A‡lsxQ¸ÊJüåc.živ“žºÊ²¦e*ñ S_ì7CNÔ \8ê•<âÁô=¢_‘DÔPÚ%·?Òº_JßlÚ>‰#dörŽêR\@/)5Öî†ÓÓRúѤLÛxÜB7:Õv Q„½ÐÈ;n-ë·Ö‰ïæ`¹8Dæ×UÖïªc¹Æ&!‹zÎÉ#Ñ=ãÍÌþÔi$õKÆˉU¿Ä´Ü4ÜSÿ:)9r óÄëÉÄ1ÞÚÁ ì–7ó!¹ß;/ÁV±Ô¯ãº6öHhœŠf{Qîåå!à MmQ:™þ.´°«^ÏOµ@9\ŸJ6º³Þ/jŠù„×ëÚI$Î!Ö2%D;†ÐÇbªn‹NÉßk–öNB.ìþ0æögµàÌõ™Ü÷ÓzpxyÀ‰Ë á¦òÓÉ0(x¯¿è´¬šä±.ÂE$'Ãßý–·¾)4$žÊ\¤à 13 Ê'²Â»øÌ•[Zg\Y…ë,õ-m)´Î É†ˆTf“¡È嘢žº¢0T7,rýl? ŒhÝiRŒÊãC±gÝL1#'tñx&×€Óa—t¢ù¤¢eœüh ΜÁ=¼â(L_ˆ«/U} C\ëB%  ˜DŠ12Að¾³%%bðìaYñÉÞ¼3m5ªÊþíC(* c?ø‘„¶’-”dË(‡\²›Ð]¦»=óŠž‘Ï•ãü_®µÔ±0äövà+Ó#óòÜ ¡éU5jŽÀÑ‚Œ´–L¾¾/ÇÎ4wf^ZO´9|^ä3ydÙ‘ û[†NkCpô”©Âï3k.ǧF!owMNÞ¨ #¾ÔsNMÚ&zqcúþ_1M„ìkåþ¬¶‰;§æûƒ(ŠšæHŒžßvÙµÁñ¨YÌÞ{T…]ÛÍú™eðÛ@µd9p†jÆÍQ¤O9Ô-™Q’‘(7Õ²L*îæõ%JUmdì'5Iù©ä1$>ÎëÃûuX{Ð7u5PV@8Ï Š?*ƒµç»†…axVEòxC{¾»6Ž¬³âuLGß¹"ÖðÅG¨Ç¥ÙQDY£ØôâMàª,>P;ƒ2ý…_÷p<“kn}ÌG‘š»aâ ¶,»¸o}&5´&-—ëËä[·9‘œ¹“ûE!WÐÀ¹Ôh{?>ø§ƒ¯kž^Ö…fÈ%©(ìÆúYúVí_\6q÷‡y=nò‚ð†Ot•uÛ¥²çªp߉…ªûZáêÚàJH¿¯|ípç–Â& ù‡Èò+_¬k«ƒºZZ¿ÎL£Gu=ÒçhüK¥¼ÎKl}[Åe¾!ê#q<‰9Œï—•žGD-È)<MõxK§«E½ÀÄý(B=¬Vv¸UC…+r^ímæP™IU!0"Ù¶ëúçù›]tѧŠâÙº½Ú²Ä .ˆĹ ɉ4-ï3~ )x­kŽUƒñà —+>B ýÐH§"_ów•yTËïj³Bl2W–É”hßoÉP{¼O ¨_IXÒý®›™Ù.xØŽÄÌMA/ïG®Od ê…³é%Ê„™Ú=zªJûÛ–ƒŸfÇtÊ«õºù˜hÅ}¿*h­ó ™ˆÄŽT“˜vuÃh;ÐmòIꀪ=ry#ÅBI1[IKÑ"Oï#„æõw~}°{í`ë3!¤…pˆgÝQ™¹œI‘vÚ{…QB†qô1{1î‰ ]úÊ–8ŠÔ—¦½£JQfÌ܀ î$¿²Íïib`ê”qᶴ 1H6ÝÏ8¦«.ÃQ¢"ªù¹™ã¬… ³ßƒMLYƘ@EÄÄ=ˆå"¢À;êGû*šdr¥¯ NäÎBKc˜ý•u¼6'gF/,AÙh¾×ˆÅ´´Ø¯¯ŸŒ‰VžöxZ´÷ÞÃùj(•ÌfQbž£:äÈå¬E;ºìÑ•7;~)uïö­zºm4t{ud³Å¨(\Ù˃R«ÒñÒÿ‚¤œï˜©G LqÈÏ-#rûÁjòFãë²Rj;’¤°vC¬™Ï‹KŒC¡3å\ò±ä@»–ߢ>µÇ°ÛP¾ã+ 'ä V¬{t:|/ry˜*º\MøÄZ¬{‡ ~­×#?e!A3•«m&fãä{-†ãw«÷jR<ì¸K¯sl¶!W¤ž™“nqâp äi^‡«I/þl‚£'ß¼ÜÒíÑ'xÍî1¤³ÄɈ8¥r{•Ò°ýÔÌÿžYóà^éý„Q:–’ÅùìӸкϼ,½[Æ´ßÇÿ‚‡á-Ofó|o‹ÒO¨öš¦!UÒóäE)ÑÒQ17SOWµ=/êðäVveEo` µºÈüª¾áJÙdÇ%åïKFïäuf|d¢³5L>ƪ±âÏDºš®÷IKÝÐrsÈ_e¤ëŠ&<ôþ­t¶©CSZí'Û×(>ÍOƒ(¥ßt=ãµ)E–Bò3úµ¾ªŽ²[á»@ÿÕCmä¾äÀÁUh¢É »ûÚ}¡/.¾¡ßÜ´Z$’5!*™g;L^ÙFH¯¤˜ÃWŸú Ky´úò4\C‘ÖJE’Ç0s¦ð£éÚ/ÎXFЋãÈ…™›à;…Ú—\‡š\lätq ¿ÃÞ?°Üörq4^Ž3Ñ%{E«…´ÎºVt:ÝH©´¯êY÷+äsk”q•kOè2ÖÕ#©—õ*65“4F­çTÕI#“Ý2Ïi‘¡®Å«aFéO5è»K83]ï¹åp|bïÌqxÅà·˜Ím_clÒg%ÿ3¹e_?No™™›Sä¾ÒY~$l%Ÿ]Ï#ï/þ¤B#:“QXc1Å(o^Ÿ"Á¬ïÉ3ߥ©—ൾžš¯§Â~@ÛÇJ:wñÇÕVuúïã1L£2^UI‰iÖÉ·rÛË„’2hHï^uÑISŒpÓEUÀë"-Ë|–¦°øLê³{~ ÔZ) ~+|´–²ô ¥eba6²=§*îhïq~¦%cÈïl†UªF»Ê^À¡c²¬øÊã£C¦u€ô,>º=±¼ˆÙZµ4ˆnÜÁ³¥³B• bøM(ƒ~}¥¬AR~ƒàCmåÇk¬h¨å‘¡"„,^ml>ˆ3ïC‘f~»Lu­ .‘/©@7˜Ÿ„åv—|>šk;¨×ç{àYáôXµ4Aâù~f"Žýü»pŠ@HC=wÿþ˜'Cê±Æ°{îöWd[¦Dèý(3ëS¹uyÆA‚.ô/´‘Ù¢Ö¤–!,ãôæ^KX;§Þ“Kõ¬|ãôhóhDÍwîVéºäÖëïÊ3 [íX›Ðöé MŒO¨ ‡'wã\dK{„xÏpÑî4¯3íÆ?ûü Ã\Ø›[ÄmÔNƒ/!}QÈw°V£ÊUôð3H˜L\Ð8càYn1Ê~¥VTƒ¡ÂÄȪn†õÃÉUŒŽÃ{7tþ^°óótÃJy|[3Ÿv.c‡7ö|o ª€h*Å+ÌTƦvIȨï xÓNvÿ”©®ªó:[ úá¤jbg,N§ÆGº¬èêQ'å´p5é·úïÚÍQ'ÓÁÆkäû²‘= \X« šœhª0!|µ§5ÃWÆÚY©‡¨ @´Í[ÈÛB#ÃsœC-ŒàÊâÒ9*J³N`¬’ÐÒiJ~ð|Îì![‹(šK{ׄ ÏLfÖÿû2öðC·ÔýZ­)N‡#"r.Ã’î n/+ß¿åẄb}¨ujhñ,Åd§ˆ¨&Ô³ýM\Ñ»fw=Ñœ²xûKéõ†}ÉÝRÌ£dV¶lQ5Æ`Š›^Iÿd¬ü¸,b'Ãw?‡Rûç,¡—2ÃSº;qj»«-:œ(´5µéfœò•àC¤±Y–ÃäbUá¬Þ†Ê<02üeØ‹ÂkNí»ôŸXÓ*Åɵ|JçC`?펩OtÕ:g~I›-7Ùž?ïÊá+åÄ‚½1X6í‡ÏcáGŒI×N§ÝjÀ±#'%Æ[ý×¹‘-¿¾¤q~¦:‰¥þt õ~W¼Y˘{Ü7Ó]<èDc &?! ¢aÐ’ÕT– ®ìßUnà–ÚD-Ù Ù»3&òœ˜Ù¨Ø'!hòË뤽”¬Ö5r»¥`Ú¤UYb†ü$> ÷ëÇƆ;±Ò4ö{à¢?l mÕõ>r:¦%ò}& Å©âh5ûDkf{¼ãÐ:m o² øSÛO»ò&¸á°s°Z?gûûà†ìÀ<2Kåpö]\¢Ã~O×^‚ß}Œ4ÍIÓ•”À fø0=/EÚlBr»°4^Ï;×6µ‰EL<¯¤÷á·¸&ІpÛ’ýÌÀÛhÁ-ö Π´™À§Ÿ³q¯n#úÝ¥ä±e¡tÊcŠ£áÇÉ—²dÌpséyü„­wrÚ›÷èÚs:50û{iQÖx— K†óŠ“%ú¸)**§çªR…DJ~†´Ä…=«ý~Q‡ Š¨v¨1Ó;QïCs ÌW³xG£)±ÚývÈJŠ›¼Ý#\j…{[e‹$§í3MÆ@ÞŽà`1ûá;±]pø UÃ+¿ñÊÎbr…Ú—òÒƳfÆh?ØPê9¤=”Ü‹’¼òu¸Ê–1z-Ãõ€øu÷ÒNûÎÕÿšŒWуý·ˆ->—Úee€zظë·xõB«˜×SœŸ\–Ö}а«£LÜØïäšãVä.ˆó‚ó<*`ÑvUçpô¬å¬+Ùßz¨L7%¿`- J¶0¡äª&ò»ÖÿV×”‚Õ•âÜN(Hñã*ìÂï `²Z:go/rGȧөr¢kžð,ßV&¿ÀÉf±EÉ  ¿gl<ÊIì7<’B;‡Ñ—‘˜ZSÊöŸ»‚ ż€g+EÍfå°OHýgÖœ›ï×É6nõb*d¤„<ËÔ\hƒ¾Ý”Ü¿¾tuz1 {°í Pͯºµ½úÀñî ÆŒ3 nSâa~_¾.uÝ"í)Œ­D; ’û%7=˜–\TzUO^Ÿ‰¥T 8Ù’ ‡ÛÜ¡wQ6†);^V #¸›ýiJu3¡Â’¥[ŒGM¼ŸÅÑ•«-³±±cƹ QÕåò«€É¤i¹}kÍqÖW·p—ƒ:Œ¯vm9—L(?µ(#Íà•™{5Mœò7v–’Áe±á’æó²aì[ù6#E#¸ÙÏÊÂîñ›¿†^Zÿ*NlÀ9N;HµŸ¿HÑ—PÐM…áhkÃÖh2Ïm]~´‘y5Ÿ6К@bÔ½¯¨óY_‚T‡^È#4 7!+·ˆÉÑ;]ßcõdýéÃYî;è_¯µ4q„Û¥«°&ű4qÂΞÖ~G’áA2æ ýœýôRG_ü;¿­¬‚x|3è( (”~ùâÞÚG:×cIM;Ó2ÀüÚ£TªSß­«ø‹b /ô ¾§ @D€ þì}uÕG±¦®³ŽEw–¢~cÀ0±ª²–ëûÿjº»òq”ú¸¬Öx÷£Pk‚1Òó\‰Ù\Üîšàom.KUX5]l[×éAÆâq£¶´ãg'dOy¾=ÙK_©ŸË)ÚQ£No¶‚‡èÒš!÷bì™ ždRi,]""Ë2œÙë8Ù†Ûdö|ì j¾I™ ËîªVÏ7ÛíþÜʘ¡Žª¼”sšÑsÑÒW…ö‘ÕÉö^MQä#(¶ž{Œ‚àäëåŠuíOž¤+öã¢3åÚÀ•ÒN3ÆrxæÒ@ñ9 °+2€¿ŠØ*@'Èãˆ4ß‘ aÈJn* ¬ÛÎÓ¨÷<Èì"$€˜ØQ›º*?×€ó›KØUeKÄî¥`·œE¦†ºi¨+¾Öõ,¡Ì¿3èëŒgÄ94c<廉“R,q#šñ\žÔ¼Q¾ÁqN½¢°¡¥“ Ëiñ©RDº#sê"+^ v©Èô²I5½}dd™3d-À°ãÏÝ:á_Û·6„ôpÌž ê Ö/sX$†Ù¶š*@7®lFá°ÙmàÚ“¯DqkÀÂÏ8ȃ‹þG±€Ê§ß–fKó–üÚ톔ÿVzMű¿L³kƨÁÎɯÛmRnZΕÁËYªšc $߆G DáÀð¥ ÄÓÀ¯† ÕO•ŠnrbP…Sè½í­ì9Ý "ý¡uë €¸ô‚¸^íÀ¾hÇ+òÁã±Hv—e»¢èÎÿm.˜¾ÛÆøZ„¥œv¯H¤œå抇vžø‚YU,Z^¨µ”ƒ¼¡äÊ^*ßÔ_ò|¶-bÂ]ˆ “‹XO@˼ ò¥þ’ä ƒù•”]€H|ô½žYܶÔ<á,ʪùòL¹CSÊΑ (ˆ+dHT³‘âß.x?Ú¢Ó? ìÜ•ÿº)ê&D˜’7Ó9–ov®7Ð.R¸ÑàáEˆ.5ŸÔû‡…°†¦†r>Rr:Q€Y "!³ƒ·{¥ü³?A¬Að²‘ÑYHKtÔz‚;E_¤ ¤âüx_¼>¶Ä¤§ÑªAƒyж•GÚÖÄ2m¨A{qºô¢ŸŠl|&#ÍÁiÛåJˆg*qº"ÃæÎ0x²×ú·ë+~÷RÛcrÇl±d4ÞvŒÅ‹@ÁSŽâ †rѯ¤‘"RHþH$T !9¡UFè4]t’Ç ô¥våy1û}”x̲ɦ*c’)¬@ÏuW$¢ ¦ºå>p¼h¹–щ̫׆xWÞ8‰’Â5íu¬Qÿñ%²Yž”7éǃGrfƒß&\õ&bÌú ìvtM¤Oà+|©oÃå÷«dÖÎ.~ÇŠ-…¿W3Æ ŠµÊ\mÕ `]?whQÉ''’DÚ¤Ž R ñ[([VW³¡ääL¾•=ût y«‹*ñíô\³ã å‰BZÊÙác“(3(:jžoÄΡ=ü¾~^â9ËÅùã¼ÁP¿‰ñÍêj¿´f¸’ÉVÜ\=ߢCõÁú¼Â«‚Ñ{‹kµBÇÒ?—­©Pm "— ]Þ§§[%ûÝ#Çø5ñàÒÃAeS¡¥nEæe.NXÐEÇo¡~fýœ]ˆzÿy1Õve“÷R€&B æ,“v‚”U ¹Œ/í0ÞªåM*ÆJMr±­®T¯çZ¦1ÉìMY˜Uë‡í×Á'ý9b¶¯™‰°XY@}…‰ß-^6”·˜œ;mpÔß7?ƒÞzß·9HE¼ShØÒ}Xkû8ßÒÁÏ^˜.ä…f„p•dŽ;™ÅËD~+u5 uE ƨñ_eq$Õ÷˜A«e& "S  ¶f‰-C¥Šz‘Í‹Àœ1y’ŸÔdÖ—p?"ÉÄZ£œÄZ!‡…¡Uà$âÚ:ø¤U×½×—Ã*¢yg·8­ íVÞ­²ávîÄOû‘éç˱À»1C¨–߶6Ña"½O×u`Z]ßÁ°bj.*¡ iM3M ‘¦$Äz\Á2¯Ù¡ã%†No(‹e=Ö!’Ôý0W‚U.†c²Q®j+a rêÊHŽó$9¯4|°R®´_#+¡lò ]$­bT%|pÖ¹ {¦P¿¬BR[¶éf2°IP£>p•Éc8*QeQÚØDHß´(Á0dNὕ™DŸb\ Íô¤ÔýŠö¶Á²u‹|±¹Õ‡viŽG¨p€!ÍÃC|d’J¾±FïÒi˜ Äí:™„í²¿ê$݃À°s· ªiÛaÐrž–d•"f9sÚ•¾ÿ4°någï$um1ÝÂݵ¡áݼ8Àâ35±Âé¤É¼Á#þÒ° Eݙ֎¢ßä0#ðgÄ;† ‚ÉØÛ° ik9” •˜êû$*GÊû­?ˆÄP<]œ^yƒÛN^õ`ÀÏX™ÒQb>dþ›2‘…ì9e¢ iU±P½]VñʧO<…oE@OÍŠŸaù /™½š?¿{SSÐTß*î"*õãaÔ…\ˆGq§V¯xél¯oÐÀ_ †sÙ—ÂÝqÓŒ–G¸‰±6½Zæ“^Š^I‘Æ‘ú·ˆïß?ü)ðã}þDÑÎ@xŠ|œ×r òeç$*‹!¢ËDæ–ßpñ£a&}×é³8Ôæ3>¤AíaSW7k¿^ª°ëµ5Œâw|$£ª¿õ^ú(Í2lbÿ½ÏÓwZIwZ”^ʺ9º”¸ë•ÇªT«¿éªÔ7±;¦ÝC‚! “ö[D /|õ4™lÓÕþô“îÐDð8Kwƒ áÏè:6߬=Fú;¦ÎSJpäLä5b¡c´›TÝ2ª3ß,AËúsí Þ2MÖ„»¥·o÷9œ«éâ+ØÚHÆž!E]ð–wK)èf¤c]´?;t¬ö//‰·+6ßø¸k.æô-pµ.fI.ï`õkWÆì’w®¨°cöǃ†×Êkç_•é@e“[Æœ]¸¸ÈЦ¦®Ñ›Bª[9çZ«•WÛ!ܽ?YjRñ‹OÄïKR/Ãv !ÆJeïGin¾=òÝü'styŽé®UØ3‚úˆlü“â¼ì,c‡¬ìÓXJ®±ís )~eשߴ6ä`®ié hÊ !r¸“£{ŠêMZWYjE^Z‡C:mi{ÜúšœnðÁËC˜ñpIx_Æ…¶O1Hý½{!`=A„ˆõ¾M’C·ªã­pl/ë§[„?„»`óŠõ q¸Ž}Pwï1ºŽé+ø o¥ŠWwÐÚê©AŠzf«DlKJ 3Ý"£0Ùm8Ú¹,c«h€„cE@ ³Õg;aj[°‘\40'GXëLÊö([3ÍÙ²Ö¿g< qÏ>ÑE—ãE•$q,‡Š–žX,–Óò5ºõÁ®ãÈ:ÄæÂ:·âöÎYýŠ M W2OÃ/Ý5håÑcPefZ./žx>t(6#Œõãg±7:°î’ŸÆÁT$òÉUŠµ9Š¶?™!xZdÿ–ÆH|ižÿÁV(\Uà•tP/éÔCwZM,s®,z¿&˜n ÿãc“¨€þ¿Jµ¿êÇÚFgI*ŒÑ Zéõ”õ3 –ò„£Š-Ï8)Ã…_r壣XZjx?ÃL‘¨ªVbsvÌ©[ƒÚ3Šr9»M‡R“Ç;Š(Â’/‘ÕÍô€Öü`?=ýàééÝa´ýVù7Má-…ßy·h8hök®‚¶ú„ÒžÉ$Kn ×ïÙ®#=†]rÒ& ¾ˆ?|„\¤»y¯ ¨L6ðvjÇ^̳¶ôŽ9[*Ç”C™ žs˜žzB~ÌmÁ“YÏ[nUGî9Ð:F{WáoN5wÏ ñŸH«ì™ÌhÑ=GÜa²á©[`c>½­U9.ªüKñ½¼Ôvq¸È+Œ.*~Htë¿ &C: Ý›ÇÑt‚¼3ïÐm #ù¸‡Ûj´JeÜDæfÈi-:åÐ:™›sÔ)3}Sjˆ°À -´õ±ixó¬§f‰KôÚúcãì 2À—ÒéžÚûlÒèVåíN¤øƒ;Ø͆R€þ'Jk¨.NTñ2F ´x)TÇf÷á7h0׫ !@}Ùø:1\Ð)ĨMç_§*ü®çZkzß1NèÓ†[Ñ­÷Ý. t|ó(fªß ±Ãò2d©(óà˜ní­vTÇyLHË_$îLÁ˜ÂǪ$‚®\KÊTk|4¸VœÄ\vù@€øŠY)IPèh …¼¥õbj¢‡õ ~hUÇX£ôõKäˆ)+L¾³™êm’Ǹ"Bc÷9dÐ3êL®ö&¦OzcwÏWÊIv½Y!Ô8Q×6#Ï.'Y£Ã)¥P3X'‹e³¥÷0öKúEØ\9ZzŠd€Vëçã>yóË·é|êo(^­µ÷Y·çŠßDs³c `Uñ õóËn«£Ë 2Â)ÓÃ0ÎÓSd€ ) 2ב‡•'49?úi¯¦Ž:Üóïî“ùp-®4f7#±Wôæ ®ðïšuÈý9þ$ª¶K€ aÍGÝÁ¦“ ±_}z‡÷=ÀÃ8cƉ *÷ú!x¹FÁ$¶Å:½P0óPìV§ ‰éÇ›¿9@<ôê5‡K?nU30Kš;j)åˆü "bÏ—"ÑôG'9¦sœ”…6ÑnSl ^Fäíïó˜QS²3۰˘Nµµ1cjÁR}J¨š¸?ÉXKF2Å’žÌK⥋9æèooxvéžµfÅø$4~ÝŽüåërb{_Ë¡˜b€ÌôϪñÁ©åò6mÅ è³âÊòU> ³Þ˜Y³f*#Kxû·232•ƒÑK°¡­Rm$)àz©?? [Óµòƒ”®’káƒ9SåúêöéQë ˆüI¥«HK%W£ìÏ.V#>9ÕM¹wHŽ#°Á{À{-¹§Õo<7Êÿ,âG–wŽRl¾!žÜ¾p§ûž“ñ2@>$"‘qŒInÃÇ5É[ñˆPc5Š Êè†zN…šÞ^0Ã)Ž ¨ó4 &êEÇeÀçm˜‚žôªùƒîœºx½ŒRq>Ì\;sfA…œÙ-‚ËJ,W,¸¾ Ec/ÝÜù|%_)[<á_ÛâH‚%zRÒÛѤcCZÞŒzß_›>Í jÙjk>­ìÌJ¨€Ù\‡\oj ¨’”!&â3ú n¶¹h#N¬×>Ê·ï U!NÈzjš Δ÷ÎÜ¥ ˆ0D<z­ÛÑ—Æ$éÿ¼¡®uŽÎ‚)XºÈ<á·3/Ò¿žâŸ¢ŽsÙž²ŸÁ:3ª5ÍÉ?‘w±aM_lM¼x-‚¡$'¿-s†+Ç”w9ÿIšð§ÐÝ©ƒŒ¼:Ê{×ñ™O.þ#—‡tFVǼ¹ñ¯MÉÙ†`Höë¸.äÔqª#]AÁÁv ¨Ý ô3,)²X÷|(]'o2ÐpMvØ0°ö½ãØ£á¦Îi‚~µc~Ö¡ªý>pfIXÁ*€Ö;—*ÇíŸn£ß6ßa.ŹUÈ^®´nçý¯ÃØܔ›'ts`½ÙM@×Äê öJèš_B§¦ëúÉö©ìõ%Ð<Ý´·èJèƒRv×-+²ÆGtcS&ÐNî}Ÿ&/§Þ‹šÎÖ=–œÂ<Ý<ÿïiWE0…ÝYPé‹ox„këáãǬš?¥ß²Ÿ„®Ç â®þÀ&Å_Ào^òc ÝP·Ô'Xµkx¨¿ZÞõ ‹¹J\§ÝYãKø¸Â:Y‹Jmù×t—’ˆSîßÀF¬CÖKªlÇ”4Ûwën̾ù¼ ¾S—m„VÚÙ÷§ñ™çZoаâå¥.JþŽ‹5 ÿUümú¾TŠÛw3Ö&C)„d,|]/;/îL×ÕdùÄÊžÀ—RÑx¸¦¡‘²Aéoã*Æ%’~ªï~iF åVôm¯³³/û\²>øÓ¾Œ$©üpQè,vêã­!šìÜý]¦~Ôܧº¸v=Ýø{ùEôëXBhm¬ÛtMþÛ¢ÊÈ’ÙÔŒ;"ÁÁ9:w~YÙ²‘Á‹ž3™ÔZÖÅ 8¸@Í÷ïNªˆ²(®¶³¸&ßmͺ¶°P6 l´¹œœU¾Xqy€e'HÃ:ìÐÜbòè—/–îQ´ "‰€5Üøl,+ý¹†ÓÑ­®Ähó&óýh#Óf®×¡–ØUp¢ˆ¤Ý”Û2ô²!&R·Q}ùhŠw;lÚú¬ÒíÍrƉôX‘¤Oý|\ÚèEè™»¨œÌáÆÖ¯( §†èJK“ä7(Hõw°¶Yf­›]2Ãݶê5rü¥)20é‹[”Ä,íИ§Þ Ѧ c(©M>Á¹ð㽶) ìxÎa¬ô›/½=5PðêóªÚb뮥;~·R%¶¶=E«¯(ÍÁÄšIÝÝ ~Ö_ËõĆ #Þžr¹ÁP(?®nËÓÇ…³4áìlâà!ÏU-Db÷±œó‰*oíyéÛV>9먥ßïµ1ÅX8ö÷ †J&T¬¶ëëÐW‰ìNœ÷ä%Üë­)°Ö­ëGÁÿ(©6Åé’w»Î¿¹®Àì"â:ƒ&.›B öC^õµÀ{·Ye2Éþä@4&%Ï2·Ž.ˆÀ'>"£½ >*üÖG!¾ÊÍP¢¥•(ûŠ€~8­=‰ý• ½Å¬ÎÒì+CJyIIJ9¤‚äô©‰{lÓ¾âi‡2vcˆÌžb·ÝGyëà*öR@æ-YJè(WÉ\£—9ã•3SLf î&ø)ß÷'†øaýh/‹SG«º3×L˜Ãß^')Hµ¹<#–°uÔKžy$Œ‚¸G,Ãð²K«÷ÖÉi[jÉÔÉ#½^‹Ƶ4æ9Œ}À¨MÏÒ¡Õ%FäÕñ®r:à]h›n·ÉW~k¸–wYkÄ®uþ`ÖJîÜ#+õ4ƸÑü•õýSÌHCG¢ØtÑíÓ´1H<“{9‹iÜoßN8^Wæ±[WéOhT«,ôJàk5à›|}è…gò­EI±h"˜oÆpv (ð<™Tù7SÀú £Þâñ€.ßR§ÌIÖ™ ”gé ƒ|™ôÊ ¨¶)j¾äÝËѲ*V©¤,Évû53‰Mãÿ ªUí³´Á—‹¡4s©Ñ¹¯21«IÞ°®„˜Jó"t¬ d…vc4ë•ìôx|óÂÜQRY<™|ùYàiñ’¢ˆÀs4íˆqÿ ‹=«gÍO‘pªuNuGÛ¥3nnïQï¯^Pr‡oZë·WåÑÇN®YEnÁîNù»|ÕÏ*1w½º# Æé:w~yòj@³†©æ†äÛäÑ\û?» 1¯ÐT0 tËÖ*ùŽ¤¯ˆ=ɶ³ÈË}ÉËúh è¹C£­Ýe; w±Ó}V,ë >Åñ±êt=ê½67D @Ì{¬Ô:@ È>[ÍF*Èâ°ñöiáÙ³ýø8VVú_ñ§u ¿+ót*ëüÐÉï0¬à¹ì•Ì7ªÒ¬íåÛÔÉvát’‹bQÔÛóžÃ_éûÈ\¤÷j >…¹Ø… Ëšø ¶VöDÖÁrC¬I’åpV¤Jm…Ë*M¢)"vK55™ðcÜÇÐ è“*dYˆüÞøXyФ'Ò¤SjœL·)Ò€™p[éÏ¡à”+÷S•¡j³¹£4|ŒQF Øð— ò5îöZ=Lº¶ ÷g’ý­J9"nÄñZÕµÜÒ·0îÐìXþóy}:D¦`OäÞ .§Çx„Ä•g¾ŒÚEâçfÀÍ3!~&D+g…çÅ¡ï B’‡ÒÌÝa(-D¸la¹ß<§^ýd¡­Õ˜À¡“‹J#UšWé¼>5ßõUê¬3¼vN;ºŠ-XFänkX-ÊD ñ<³¥n²ÁÅ9óäÈ»âH$•Â탠 ,nò6Q¿9€ÑÚ›1ŠÅ±~ø~Tû°NT÷ñÜ/¸'A!ô°.`Î OSŒÈ‹ÙŠõŒhþëî[\@™bê—eœE%›Ih¢™P€>R’ÕËô—¶köŒcn"‡ x¯å6ãªFÅ£›ì"Ð\$ L4âÙ‡ÚÞ¡1Ýw#Ì´5N€Ñóäóƒ·XlßïÈÜH)𢸙qª+"íJÉ)Þ6„Óv&¶ÎÚ<¬‘âSÞT#Ü:~$s.K7"Yr#½›-¶ø[2™6ñk¦øÛ‘àbI±4h·rË&Vç{t̑ЬAÀÉž#í‹)œ—‘€$3n«1x™‚ÕfN¤]Ä¡x&ÙË¥yî*wºƒ­Wû_PÓ3íIËçÄ×›7,ÛÝ`ÕÉhï&¬‹W^ܱbÇýú‘j“;¾¬ø ½¯žæݸlKÕdÝrݪµÛÞ»ýþn‡OpµèÀ{ [ÄËVöûšúø5ö2(íìï…r« ŠàûPC@ ˜>GÉ™x”Ýë³ ]ÝjK˶š+ÜŸþD¦ÈŒ),Õ•¿ t Jª@ûFc ɼ)õ—ß™+:9¼T©X½78{Ò9#óÎ\  8ÞsydëYµzśŒáÖ5ÑqCJ7Dp÷pRÓHvœ¡óØÎHHe£Ù[à¬_7öû+Žm#_¸h®Bˆ¤È’à³9S÷¥š[káí64 ÕEï(YtÌc„I¿Y}[<Ϲ׶@e¡$§¯ünàÏíòxu…ÛÞØá[Wó§¥&Ê a4#+@õúë|üK!G«€×}ÕDøèì£gX Ü9¿¶¹# w×e=$ E©§ÜLm†»÷ ..†“ÂÑ£æ…{ª{#…šXMµ<Òà0¨6mþãðõ?{ÑFÙëâþ‹l†Ô µî:ÎÐçRwñ¨9ÐýÆÓ“zôÇ ”ÒÕgº¨á†‚výIJ*¨Ž]¡Žy­ó{q§WlÒ èÔ »çcrÉ%2^Ò ¢ŸÇ{êy³ñl rIÒS#vAdñºÜeQ“(à¾.@N騀Üã²6×)uDY´¿ d%)öÐTæZ¬ÜóÚVŒ @A½ö‡ü([êP‹¾Ì©Ölz~úû%ç~M/àÒ§2H~=Ö¾[E† C.Š÷ æ¥Q&MŒ`eþU¯‰'¬.¶Ìhêk›Ã(iÞm² <‚WŒ¨Ô 7 誽™•ÍH&ú5ƒ´Ä‰Gv^“ê,°d8U#¢^{Â`’O¹aR3t¾:â¡8¯›£;t©¸&°,BßÔ³ ú¸êy·`]’;‚B{+µâhöC§ˆPT ‹ƺ:ϪV*O ï«)v–ç„•ÿ~³«§²:"“Ñ»ô/\9æ8±/[‹×ºy½Š-R؇Ç?,ı¢‚h7ÙãWE‹9·Qí Å ƒP½ùÏd!*Oæ—4ÕДµý»ƶÍÀB Ž»ñEÁhï`Ä6 I¢rµÂ4FÞF1gïÆåD¶ù³Ã^MS´«íEý/•ãëïÖqiyc7ø¢ò|n]™,ù÷ˆ•Ó óZ8UÌ°aïÅ¿›T¨É‚5$¸éÄtäθî÷ü…àþqÝ5ˆ£tšAÁâ hɉB+îìÂÄqÃ…÷Až@E9L,(|wQ3:܈'׆ôú ×WskµzØ_[λL“«}Ø Ê‹ÙqQØÐr¤Ú¶íÁ¼¿f“¤ nQžH.»¾Qôp„å ØØ$ùcºž'_j‹‡ˆ¿„n#\Uœ·>ó2ãb4,O9DÝNdW€V¤zEU¤z¾ ç˜ž× âÅ«³ Y_tÇ[ÁôHv¿Ÿüù”¨5RØbÍö"ôVqÓÁ¦å>Ýæ ¡é¬xZ¿9N†1«=áz|ð®˜:{å€ì[R Š¤‰èšá”9a|ç`Àˆ#kUÛÌ;mÚ¹ÎÌË\ö™++?Z]ð$8/3¦ª8{uOGø'Pc®¡÷ûeÍþoÇj¼9Ld©CkõÕª¶C©ã±¢'ÅÆÔ¡ÇTþNÊJ\'B'noÿÆEÀ„éô­`ô7JˆrÑià°~#Ïöµ”_žKªÔúŒ*£ÔÌuavñ eß‚º˜Jå숨n2Þ~¥.•´H±vyn¤ù9¸1$rô~¼lõ‡zëåÑñ7kíàjÙm§n$vÀ:MµP &¸œD2Q‘an^´«\…®eóœ6üÞÚ—qHØIø Çbâ* §Š¦¹kŠè„­\QeFco)Îày§Æ\ý×X4ËAFß\† ÜUÅÀf`èS1m–KÂ<(À‘›QL#rbübnyôm™hL* 0_Xäï?ÑfR{éoì •¹ôˆÏ…Ólþ¥ïIaTã5 M ]G‚­Ã4öž*fÿEO›ÛJ¬!•%n <ýtèÍ!Ù$ ¾RsXÎz­ñQ¸c4!EС ç:<Žðû¡‡á].ŸÈßÄÙ®™¡Íí[s…¥µ.‹þP,EÏlñÔ DлÛíšuGIé²ã°æKTcewá­9G÷»rýÜÞçݦ®Hn°”»º…W£ìdôãvÕ"Óàó®eZR‰þ"àß+¾RÖ¥þtj 4ûÅ:ØHýÔ ã©ÐMB\Œ%P§ž×å…®m ü,2Ôò#"§4®h 2îEÊ!“£ÝŽ0‡(2×ÂÌ4î|Ä– Á€"}:؃*˜¿TÌÙ¼©é*yBîÑhk?lÔ&:(rRŸØ”e¸5éŽ~qH)ç©‘$zlÂœ8ÔÛA²\9†©8´À­6ä¶eÊyˆt:§´ . JôúA¦oùìfÁ07â”3_»>÷Ö5 Ùš:ÎËœ¯sĉåvb|øú_òÇ©möãžòŸm„!ú+X=¸„ß ¾f¹©#ÇÄ`y÷B+$¿«ÏkÌ*Ì´Ü<½bq >Y|eZ£)íEPó±?Y¦ä×ñàY/b‹dc…Óœ¼„†«íQsÈø53Q²»Š%7DZõ¤â9ƒ‡¾ªŽå’ߦ*®µññ•ì4Á©¤Ccê ¥v¹“rðnt\Ù™:;B²DƒÔí &‰à³áèú$[÷ÖwüÛÕãt\Ì´æèÔ-ë%B‡²,D„Ð0Õž¹z±àÖ¨Ù\ ˆ¨&ÿÒ¯`9Æ ïšøð€wx‡ðö3[†½°@©…+÷ù\„XÎÎÂÃ.œÐ0#éÝr ùirº1¦ ÀúmX—qÄ1ÓTBà³·wo•ŠÑÔÃÜ„"î"‡pÿ6@À\·íÈI‘u_úägk«etÚÛ%®Œ07¾ÎžFµã¡OŽ)«oË:D>`£±ù¬œÈžó0·¿îÑŒWÕÇôñ}VáRݱßxïyáYÂI4’ ëkj0Ov[ß¼’šL[•SâÌ óñ“”ØÍÊsÁëªè™?¿/sBRÒ8ù¡æ²RÅA% $+²Kyâ 1µ:/Eᜑ%U}ª¶J“róL¹ïV·L¯T#] ÔÜ%0Ѧ=/Ò Ûä'‡l²óþ])jþx«†8™_ÞÅÎÑ"îË]ð§ûf·€ ~sÂÎêµÎ’f¥m7]ö¨ZEÜù q?Å`'ÁUµ omàØÀ¾<¢w<µƒ Ú윖¤_ˆ¾@[RBâã1!êÞq-æ-Ú‹Ph¯K’‰ûMུ3k?¬õ…$¬ŸËïUÄ–×Mâ’Iºm.ì-Na4ŸâÙF ]XçQ´á6¡Ô¡"gwšRe ]!CÕ¬>¨æ€:–„-}ç5>Ó>šÞðÏanRÑ%3漬—­àsŽ¨ƒ&ñr z½qÓsŠ´»úÚ[ H{"‡ætBa‹3ÝWØÒßž}€å¦äþVî¹÷ö5‹öšß¢ ¤I”‚Õ‘gõ0ÐQvA&ĸ‹¿D?Þæäò·ñëîU+Á´V•ß¼’a §2[`Ÿ%ù·bp£ )¬¸Æªªý-T`‰Á8iÁ´ùÒè<ê°&íí×èL„q·Åɇ"i”žDz’lÕYô ß3(vÝZÆìròaFÒÂj›H QA98-­³iåÆÍfþ&Hƒû„Ѧt®­ ˆ¦î9É/îRrRO Ôù;)›žöÒ ÷ƒ‚)¶ ½¸sÔ.¯à°J_§‰ÑÁ[P…^ùÚ_Àbë7¬9IÀTˆ’*o§üwÈi¾? ;‡¶‹¦“ýjv _ÍÃy¢ÓGƒ7;“8{T'·Õ{iW€éçòÊ»·]âß .ýWß®¤Ã· ¼ë#'@#Lâê£2·7ŠX²ûCœ¹ÐþÆ%E)¡„4üû…õcMúσðc ª„gÑg3I¿`ƒ8ïçBLuK+¥ÿíÔ‚rJïYLÞZ«O¾FÙ B\ œ+þ 8PjÝ÷Aï ²ƒË¶xåúÜ5*¶%O•t›5 Uq‚BÃŒÏI ïê ‚ê)¾Ôfµ9‰jŸÌŒ§ x9oÔ)bIÒ9‘QðTÕ©øY®BTˆqgæÓ‰EâËTÈc ¸}}3€>©BLé)á`Y)U0ª61!"S–fµ4ÔÛÞÖ Yé„;‹-h©vêLÕÏôro'²C*OMCj‡ÿS8Ú³Ü2žš©æÉ¥§Š[xt‚þŠ9¸ qè¸1@µCÜéA[õ+Ò”ýúrH JÑR¬}³ªÑ'Ú†ÛÇ\W?R¾Ö¨È­Ây¥^ËB](8r•ÎÚ¹ïÏA¢ÀÛN4:šÅ8èé@ ,4žrR÷Rqb΀Ë2Îíl`ÛcJ–î/Õôsǽ“GÿìÅüa`΋ª#k5!Xuª¦~±öRŒ¾ŸNuß¾©'ìDD@tË.ñà4úNí–˜$ÀJ–¦–&»ü´c‘I ¯|V®H•–”±ïÅÈos,‘R)’MÉ# ÆœÏy…kû6x5íÖ-~(ÍZÿ“ÜlD‚ ¦V#´~xr× î¼&÷ðû?4’·b\)•‚…ùJÕ*>…‘æuž†É‰cšMîø0˜¢‚ ÷Æ+WZìÒ/eeíѽßR–Ï%~i³!¸³ þaZ(årz ÂX®RƇfŒDe¶èŒCÄ6¼åéc q؇tm"Ðç«ô±„Ô'n=ïÏ[1·Ò%ÙœV‚®[ï!q͈—÷s3˜ñL+,Oµ9‚Ó®rëbG’© ñ¬wA Ùoo´Î¥nÇy¡ð Šþb/;ÍCt”DC}ãýö¼IÑ»`Î gÍa µ`¯·ÃÈó¦×"ŽØ±’g¨»AÙÅhEèa ÌwõœbÞá¨È?Gùøª†b®¨d2‚ ÅæâRt¹æ™éü¦Dñ<ûíâF]š‰ôùåÕ(‡Ø“š‰âo%¢Þ‰làø#¢~ÏT?ë€Lšr°ÇŸókè4Œ®wx“€J° endstream endobj 300 0 obj << /Length 494 /Filter /FlateDecode >> stream xÚm“Mo£0†ïü ï!Rz ˜|U‰ÄAÊaÛª‰V½&ö$E 6ÿ~=HÕUAgÞ¿“ɯ÷½Ÿ«ú~üÌÙ´uo$ø›ßÇÆ›LD-û t÷  @Ùö…½›Zî¡cÓÍNìtÙ=YñNËk¯`T=­áRêo ÞæøôeCîŸúòÚ•Úç(>”ÝÕŠæ™ ²ŸAæŠþ€iËZ¿°ð™sn[­6u…c´^0XaÁhî\je?ì„î¼0bª”ÝprOYÙ÷Åû[ÛAµÓçÚKS|ØdÛ™›óøäoF)õ…MZ³©}ß4W@Œ{YÆœmG;ÿë±<œñ®9Ü`‘;‡äKÖ Úæ(Áõ¼”óŒ¥E‘y Õ¹¡ât¤ba¥bi<Îg®bÌÅw­ü:/]×åvYsäˆâ[ä˜â+䄘#ψ]íœôò‚â9ò’8D^osâyMìîÚGÈ‚X o‰ä‚îBŸÉà5Éà‰<øÇ»’ÁÿÂò k£(Do9Örá,Âq¼B?"tŽýEDqì)bbœW$ÄèYÌèM»>sb×gEìjqÞ(ŒæÃ×po¿$îÝ}IdoŒÝ·œn-p!J ÷ýmê«ÜÏ-þøOÃÓ[áýL‡ endstream endobj 301 0 obj << /Length 696 /Filter /FlateDecode >> stream xÚmTMoâ0½çWx•ÚÅ$ !Ù ‘8l[•jµWHL7IP‡þûõ¬V=Mžß̼ñ s÷ëu;ÑU··õÈÙ›=w—¾´“ì÷îÝÝå]yil;<[[Ùj<=?±×¾+·v`÷Ù&ß´õðàÈ›¶<^*;²~&ûQ·‚>ìþÝþ”MS >Ù_êãP·ò{=éÇsæ@öd”ôÇöçºkŸ˜xäœ;`ÝVY×`Œs4½JaÓQÜ¡n«þª‡í¡.’Uu9\ßèY6î>¼ý<¶Ù´‡.Z.ÙôÍž‡þ“4>DÓ—¾²}Ý~°û¯ÒÜÑör:-d0­V¬²WÑÍÿ¼k,›þ8ãóþy²LÒ»ðºÊ®²çÓ®´ý®ý°Ñ’ó[Å*²mõíLrŸ²?ŒÜÔqù¥ã• â5F8@ šˆ=@Šð)&°  È8Ô¹€ÂÅRx u€Dº\j2H—†ª¡ÐVÁ¹0CzL]ø Âb°ct‘I ©g$`htÑ‹0œÆ\F„áŒ0ä†sê‡á jd< —Iê6œ»õñzgóñºË»þê W ¤qÈ’£+—Ÿ#ö•ñÌÇkÄÞ .‰bªsré…¤šáæÄç†bïmŽXú¾„Kß7ǵHß7Géû„û¾nb§>&jÊصäuœ¯¼ú•ñ1ÜV™÷•âÜãâµÇ‰Ou$ÕŸqWèS/%1{\øxB!€§ÔK(hH©—TЖ枃»J©Ïϯv×ÜëÁ=küÒ2ø¥UðKÏ‚_:~é$ø¥Óà—ÖÁ/¿Œ ~™Eð+7¿èË¢/ ÿlì¡ÛÒ(/}ïö -+ZXukoûìÔE?Z„ãæÅÛKýqíƒÄ endstream endobj 302 0 obj << /Length 695 /Filter /FlateDecode >> stream xÚmTMoâ0½çWx•ÚÅ$ !Ù ‘8l[•jµWHL7IP‡þûõ¬V=Mžß̼ñ s÷ëu;ÑU··õÈÙ›=w—¾´“ì÷îÝÝå]yil;<[[Ùj<=?±×¾+·v`÷Ù&ß´õðàÈ›¶<^*;²~&ûQ·‚>ìþÝþ”MSÇ“ý¥>u;áà¾×ÃÑq~:fc_0F)l®»ö‰‰GιÖm•u f8GÓ«6•ê¶ê¯bØÒ"!YU—Ãõžeã.ÉÛÏó`›M{è¢å’MßÜáyè?IáC4}é+Û×í»ÿ¢Ìl/§ÓÑBãÑjÅ*{pÝìϻƲéOÞ(ïŸ'Ë$½ ¯ªì*{>íJÛïÚ-9_±eQ¬"ÛVßÎ$÷)ûÃÈM—ÏñP:^9À ^`„ª‰Ø ¤Ÿbr š€Œ@ ‘{@(\,…RH¤Ë¡&€ti  mœ+3¤ÇÔ…Ï ,;F™$Б€‘zF†F½ÃiÌeDÎ(ó0œAº1a8§ÎyΠFÆÃp™ nù[¯w6¯»ü·ë¯Îpµ@‡ )9ºréñ9b_iaÏ|¼Fì-ÐÐà’(¦:×ù(—nQHªY^`nA|n(öÞ戥ïK¸ô}s\‹ô}sÔ‘¾oA¸ïë&vqêcâ ¦Œ YK^ÇøÊ›!¡_Ãm•y_)Î=^ ^{œøTGRý÷w…¾1õR³Ç…'ÄxJ½„‚†”zImiî9¸«”êðøüj'pͽܳÁ/-ƒ_Z¿ô,ø¥ãà—N‚_: ~iüÒyðËÈà—Y¿2qó‹¾,ú’ðÏƺíŒòÒ÷nЪ¢5Q·ö¶ÍNÝ Yô£58.]¼½Ñ‰ç‚è endstream endobj 303 0 obj << /Length 695 /Filter /FlateDecode >> stream xÚmTMoâ0½çWx•ÚÅ$ !Ù ‘8l[•jµWHL7IP‡þûõ¬V=Mžß̼ñ s÷ëu;ÑU··õÈÙ›=w—¾´“ì÷îÝÝå]yil;<[[Ùj<=?±×¾+·v`÷Ù&ß´õðàÈ›¶<^*;²~&ûQ·‚>ìþÝþ”MS§“ý¥>u;áà¾×ÃÑq~:fc_0F)l®»ö‰‰GιÖm•u f8GÓ«6•ê¶ê¯bØÒ"!YU—Ãõžeã.ÉÛÏó`›M{è¢å’MßÜáyè?IáC4}é+Û×í»ÿ¢Ìl/§ÓÑBãÑjÅ*{pÝìϻƲéOÞ(ïŸ'Ë$½ ¯ªì*{>íJÛïÚ-9_±eQ¬"ÛVßÎ$÷)ûÃÈM—ÏñP:^9À ^`„ª‰Ø ¤Ÿbr š€Œ@ ‘{@(\,…RH¤Ë¡&€ti  mœ+3¤ÇÔ…Ï ,;F™$Б€‘zF†F½ÃiÌeDÎ(ó0œAº1a8§ÎyΠFÆÃp™ nù[¯w6¯»ü·ë¯Îpµ@‡ )9ºréñ9b_iaÏ|¼Fì-ÐÐà’(¦:×ù(—nQHªY^`nA|n(öÞ戥ïK¸ô}s\‹ô}sÔ‘¾oA¸ïë&vqêcâ ¦Œ YK^ÇøÊ›!¡_Ãm•y_)Î=^ ^{œøTGRý÷w…¾1õR³Ç…'ÄxJ½„‚†”zImiî9¸«”êðøüj'pͽܳÁ/-ƒ_Z¿ô,ø¥ãà—N‚_: ~iüÒyðËÈà—Y¿2qó‹¾,ú’ðÏƺíŒòÒ÷nЪ¢5Q·ö¶ÍNÝ Yô£58.]¼½Ñ»á‚ò endstream endobj 304 0 obj << /Length 739 /Filter /FlateDecode >> stream xÚmUMoâ0¼çWx•ÚÅvHU„dçCâ°mUªÕ^!1ÝH ý÷ëñ#xÙö?ŸgìÁÜýx]OTÝmÍ$|äìÍœºs_™Iöss îîò®:L;<S›zœ==±×¾«Öf`÷Ù*_µÍð`É«¶ÚŸk3²¾'ióÑ´ž‚}Øý»ù=©½à“í¹ÙM;áà¾7ÃÞr¾›f¶ÆnjÌ-ùeúSÓµOLg~¼À8÷ã ãâþÈ)okà çA„8 ö$`I\èÎ×3`çAfŽã<ÈZ]ƒÂ!‹„ê xNkÇyã¹ãÐð"œ7Á¿ _¥ã“§Ìq âH`òáö•‚nú¥¤kÌÂðRONH=CpB:# =Ñ%8“ˆ88QA~¡!*ÉzÆœøÐäT?!~Ž> étw©8éÄy*ás£¤Ï }nÔÌçFE>7*ö¹Q‰ÏR>7Š²¢ G]¼;~îó¤ŠÛ<©ò6OšßæI‹¯yÒòkžtèó¤g>O:òyұϓN|žôÜçI/|ž´òyÒÚçIg>O:÷yÒ…Ï“.}ž2îó” Ÿ§Lú> stream xÚmUMoâ0¼çWx•ÚÅvH U„dçCâ°ÛªT«½Bbº‘ ‰B8ô߯ß{ .Û@ãçñóŒ=˜»/Û™®Ú½…œ½Ús{éK;Kîºàî.kËËÉ6Ã/k+[M³ç'öÒ·åÖì>Ýd›¦yÓ”ÇKe'Ö÷$cßëÆS`vÿfÿÌÊS¯fûK}êfÆúVGGùf–¹û\b¸à·íÏuÛ<1ñÈ9w…¼©ÒöÎÁ|Á擬CÝTý¨„íAW $«êrGø]žÜIÀâíÇy°§Mshƒ$aóW7yúÔ÷ÌŸûÊöuóÎî? sÛK×-ˆ`ãθtJ!±'™ˆcøÀ8õãŒ3?NaœâOœâ¶<Dg!Àƒ;IXô ôÀÍ0z)rЃÌ@« kÐpÈBQ]^ÒZä 7ž!‡î /½‰ü òU Ÿ<¥Èɉ#“ÜW ºmÐ/%]cXß!õÔÀ ©gœÎÈ€žhŒœIDœ8QN~ACT/ès⃕QøŠøôQ¤ïRsÒ ç©…Ï–>7:ô¹Ñ ŸùÜèØçF+Ÿ­}n4eEƒ=zG~æó¤óÛ<éâ6O†ßæɈ¯y2òkžLèód>O&òy2±Ï“Q>OféódV>OFû<ãódRŸ'“ù<™ÜçÉ>O)÷yJ…ÏS*}žÒÅõÎð—¿tx›à½>zå¥ïÝ{ˆO->tðÄÕ½¾Æ]ÛÁ*üà3>ýcÀè¹þ¤C§~ endstream endobj 306 0 obj << /Length 900 /Filter /FlateDecode >> stream xÚmUMoÛ:¼ëW°‡éÁ5?$R. ¤d9ôMðð®ŽÄä ˆeC¶ù÷³k›m‘CŒÕp¹;;†wŸ~>Î|¿Ž3óEŠ_ñ¸?O]œ5߶‡âî®Ýwç]Oßcìc]=~?§}÷Oâ¾yhÆáô9%?ŒÝ۹׬“B|Æœ‚>âþ)þ;ëvÇw%gÏçáí4Œ3‰ä§áô–’>\ ‚‚6ý§ã°¿ õEJ™€õØ7ûÆ8ó 1¿’{Æ~ºðÏ`W(-ú¡;]¾è·Û%=°ùñýxŠ»‡ñe_,—bþ+-OÓ;qü\ÌL}œ†ñUÜÿI--=ž‡·B«•èãKª˜æÿ¾ÝE1ÿpÆ[ÎÓû! Mߊyuû>Û.NÛñ5K)Wb¹Ù¬Š8ö­iÇ[ž_®¹uÊ•MúÑzQ­Š¥Ò)V†€Ú(TØ€àx¿àÞ¢ žjy‹°°!ÀÐÔ•µZÔÀ2àP="¦ZdÔ0\ÃG©R\¡·”).–2*ÎШa!„U¼Ä,†³ÔÛHð° `+jÐÃ.¸5Nα@èâ°èÐVK-àxŸ%ô˜Ü3š% A°YÓ€z¡ÎšÔ>kP#¬³¦õ™5m0W£oš¦Ã¾žj­®§Üý·.†ÐZ¡ŽT$X/©)n)æ#W—„o(æ“oÀRZÞ $K¢p4’ŽZ¶-bâ\­1¦Ü°Jä æP"Gñ‘XÔQ¬‚i/8ºkÉ^€ÂZqŒ:ZsŒ½š9”d š­Bù Ž)ßsLù-ï7½æx˜ÏJ›¡¾Ò`¯ažÉ½)f¥É$†µ’1™¸ dÑŠcªCZCù<£7Ã3JÊgózÌnøþHÈ°íáÌYÉšäTœ¯a…Šï¯Æ,_»œ-Ÿ—Oë87Ë}êÛKÔ´Ü—Ll¹oKñšò+Êg­JÌâ.¾GZyóº‹Vðc­48¸’ï¼äØWtù]Í:P~`áŒñ±–rZŽq.nÍ1]Ç ÇàSÿæ/©ßP•ýïuö¿7Ùÿ¾Ìþ÷Uö¿·ÙÿÞeÿû:û?Èìÿ ²ÿƒÎþ&û?”Ùÿ!dÿ‡&û¿1y–¦¼ÍH·œn5þ¹ã)º½ÝyšÒ“Bï½x#†1Þž´Ãþ€]ôGoáõñÅ×Mñ?®Xê endstream endobj 307 0 obj << /Length 900 /Filter /FlateDecode >> stream xÚmUMoÛ:¼ëW°‡éÁ5?$R. ¤d9ôMðð®ŽÄä ˆeC¶ù÷³k›m‘CŒÕp¹;;†wŸ~>Î|¿Ž3óEŠ_ñ¸?O]œ5߶‡âî®Ýwç]Oßcìc]=~?§}÷Oâ¾yhÆáô9%?ŒÝ۹׬“B|Æœ‚>âþ)þ;ëvÇw7{>o§aœIä> §·”óѲH˜ø´åŸ8‡ýøU¨/RʬǾÙï0ñ˜_xˆù•ÙË0öÓ…ŒxµBiÑÝéòE¿Ý.‰ÍïÇSÜ=Œ/ûb¹ó_iñxšÞ‰áçbþcêã4Œ¯âþfiåñ|8¼E°²X­D_RÁ4û÷í.ŠùGÞRžÞQhúV̪Û÷ñxØvqÚŽ¯±XJ¹ËÍfUıÿkM;ÞòürÍ­S®lÒÖ‹jU,•N±2Ô@  "À–,Àû  ð õTË[<€5€ €¦¨¬Õ –€ê1Õ"à†á›×cvÃ÷GÂ@†m¯gÎ üKÖÄ §â| +T|5f©øÚÕàlù¼xZÇ1¸YîëPß^ê ¦å¾dbË}[Š×”_Q>kUbwñ88Òʘ×]´‚k¥ÁÁ•|'à%Ǿ¢ËïjÖò{ g䈵”ÓrŒsqkŽé:n8Ÿú7ÏxIuø†ªì¯³ÿ½Éþ÷eö¿¯²ÿ½Íþ÷.ûß×ÙÿAfÿ•ýtö0Ùÿ¡Ìþ!û?4Ùÿɳ4åmFºåt«ñÏÑíÙèÎÓ”^z­è¥À1Œñö öì¢?z ¯ï.¾~lŠÿP}éL endstream endobj 308 0 obj << /Length 867 /Filter /FlateDecode >> stream xÚ}UMkã0½ûWh…öF²ã¯²CÛ–¦,{Ml¥hìà$°ý÷«73nw—ÒƒÍÓèÍèͳ,]}{\OlÛoÝ$ºÕêÉúËиIù}s ®®ª¾¹\w¾w®uí8{ºSC߬ÝY]—«jÕíÏ7ž¼êš×KëFÖç¤Â½ì» ÖQ×ÏîçÄ5ÛßFk=i“Óåè†ÉÙL4Rž÷çWOý‚¥ü”úlJQn8íûîN™[?ïË®-û;S§¦£ÜݾkQ¨¶Ð˜Pµûæ,#z7ï’×o§³;¬º]Ìçjúä'OçáôÞÓ‡¡uþ{Qן ô„õåx|u£t°X¨Öí|]ïËýæàÔô‹®ß™ÏoG§BÖØô­;76Ý‹ æZ/Ô¼®ëÚÿæLÄ)ÛÝÈ]z®®ý+ÔQ¾æÉ&¤€)ˆH8ù@h€ÐÚcÈ<Žkd>€‘Re€‘ƒ‘瘌«T\Ãã`^QqJÆ’šÐX‚Q£hÍ)5ŠÖH©K ¥®à£ý±Ï|6öÝüÚ b‘‡ÂÚ.fÀh\‡e®`‚žÇbs¼N[à”sSàŒãÄÏ9·¶‡Óºàu‰Sr¼®ØkÔ4ç"nXCA8%Ž¦ľFðÄpý ×O<—cúÒÕǘê,ÿâœúßx!¼ g!íÖaÀM4cô“Ên€Gym¼µÆð: C[1FýlÆx L^ò"K£~–2&NÆýCC&^к䅙Á—¬` ÍYÉecªOý›{"«CgÎúc¬›³þüœõÓÊY =9ëO(—õ'è=gý)ñYB|Ö™BsÎ:S|ãœuF”Ë:#Êe†~,öÙ¢+>£G+>ÇÀâ3qÄg¬eÅgìK+>c]+>ÃO+>G|FïV|†~+>C§ŸÑ»Ÿá›Ÿ©¾ø ýV|†ÎB|ƺ…ø ~!>ƒ_ˆÏÐSˆÏ”+>£÷B|&¾øLüŒOÂr¡—BüG/…ø}XˆÿT“ÿK5ù?)¨ŽøNÅkÅð¡âxáÁÑ$s„y®ªå„¢ G5.™÷» ¹ ƒ¿&è&¢s'þ¾sï—Õ±?"‹ºåÆ‹£‡:øÜZç endstream endobj 309 0 obj << /Length 867 /Filter /FlateDecode >> stream xÚ}UMkã0½ûWh…öF²ã¯²CÛ–¦,{Ml¥hìà$°ý÷«73nw—ÒƒÍÓèÍèͳ,]}{\OlÛoÝ$ºÕêÉúËиIù}s ®®ª¾¹\w¾w®uí8{ºSC߬ÝY]—«jÕíÏ7ž¼êš×KëFÖç¤Â½ì» ÖQ×ÏîçÄ5Ûß&ÔzÒ&§ËÑ “³™h¤<ïϯžúKù)õÙ”¢?ÜpÚ÷Ý2·ZkXvmÙÐØ)˜Š85åîö];ˆBµ…ÞÀ„ªÝ7gÑ»9x‡¼~;ÝaÕíú`>WÓ'?y:o¤÷&˜> ­öÝ‹ºþL '¬/Ç㫃¥ƒÅBµnçëz_î7§¦_týÎ|~;:ÒذƦoÝé¸iÜ°é^\0×z¡æu½\×þ7g"NÙîFîÒsuí_¡ŽòE07H6!L‰@Œ@ÂÈBŒ€Öû@æq\s óŒ”*댌ÆTgùäÔÿÆÀ á]8 i/°n¢£ŸTv<ÊÈkã­õØ0†×YÈÚ²ˆ1êg3ÆK`òÚYÂõ³”1q2î2ñ‚Ö%/Ì ¾dchÎJÆð(«S}êßÄØYÍ:sÖcÝœõÇà笟öPÎúcèÉYB¹¬?Aï9ëO‰Ïúâ³ÎšsÖ™âç¬3¢\ÖQ.ë4ôc±Ï½Xñ=Zñ9Ÿ‰#>c-+>c_ZñëZñ~Zñ™8â3z·â3ô[ñ:­øŒÞ­ø ߬øLõÅgè·â3tâ3Ö-Ägð ñüB|†žB|¦\ñ½â3ñÅgâg|2–³½â?z)ÄìÃBü§šü¯XªÉÿIAuÄp*^+†ÇãŽ&™#ÌsU-'H8ªqɼßÍeü5A7û8ñ÷{¿¬ŽýYôÐ-7^´=ÔÁ!Uç endstream endobj 310 0 obj << /Length 868 /Filter /FlateDecode >> stream xÚ}UMkã0½ûWh…öF²ã¯²CÛ–¦,{Ml¥hìà$°ý÷«73nw—ÒƒÍÓèÍèͳ,]}{\OlÛoÝ$ºÕêÉúËиIù}s ®®ª¾¹\w¾w®uí8{ºSC߬ÝY]—«jÕíÏ7ž¼êš×KëFÖç¤Â½ì» ÖQ×ÏîçÄ5Ûßf6Ó“æ09]Žn˜œÍD#åy~õÔ/XÊO©Ï¦øá†Ó¾ïÕZûÀ²kËþ€ÆNÁTÄ©é(w·ïÚAª-ô&Tí¾9ˈÞÍÁ;„äõÛéì«n×ó¹š>ùÉÓyx#½7ÁôahÝ°ï^Ôõg=a}9_Ä(,ªu;_×ûr¿985ý¢ëwæóÛѩƆ56}ëNÇMã†M÷â‚¹Ö 5¯ëEàºö¿9qÊv7r—ž«kÿ u”/‚¹A² )`JbD>`´öØ2ãš™$`¤TY'`ä`ä9&£Ä*×ð8˜W`TœR±¤&4–`Ô(ZsJ¢5Rê’H©+¸†Çhì3Ÿ}7¿6ƒXäÅ¡°6„‹0×a™G„+˜ gıXFǯ€Æ8åÜ8ã8ñsέ-Çá´.x]â”/€+ö5MĹˆÖPN‰£é±¯<1\?‚Ãõ“Æåǘ¾tõ1¦:Ë¿ø#§þ7^ïÂYH{upÍý¤²àQF^o­Ç†1¼ÎBÆЖEŒQ?›1^“׆¼ÈƨŸ¥Œ‰“qÿЉ´.yafð%+CsV2†GYŘêSÿ&ÆžÈjÆЙ³þëæ¬??gý´‡rÖCOÎúÊeý zÏYJ|ÖŸŸu¦Ðœ³Îß8gå²ÎˆrY§¡‹}¶èÅŠÏèÑŠÏ1°øLñkYñûÒŠÏX׊ÏðÓŠÏÄŸÑ»Ÿ¡ßŠÏÐiÅgônÅgøfÅgª/>C¿Ÿ¡³Ÿ±n!>ƒ_ˆÏàâ3ôâ3åŠÏ轟‰/>?ã“°œEè¥ÿÑK!þcâ?ÕäÅRMþO ª#þƒSñZ1|¨8Gxp4Éaž«j9¡èDÂQKæý.h.Ã௠º‰èÜlj¿ïÜûeuìÈ¢‡n¹ñ¢Åè¡þðç( endstream endobj 311 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œ.RIý÷ë73´»«ªÐóøÍøÍÃØW߶3ÛÏnÝjõèNÃelܬü¾;WWÕÐ\®?ÿp®uí4{ºSãÐlÝY]—›jÓwçOÞôÍë¥uësRá^ºþƒ‚uÔõ“û5sÍxЉֳæ0;]ŽnœÍL#å©;¿zê,å§ÔgSŠ ütã©ú;enµÖ>°îÛr8 ±S0qj>ÉÝw};ŠBõ ½ UÛ5gÑ»9x‡¼};ÝaÓï‡`¹TóG?y:o¤÷&˜ß­»þE]&ж—ãñÕAŒÒÁj¥Z·÷u½/?v§æ_týÎ|z;:ÒØ°ÆfhÝé¸kܸë_\°Ôz¥–u½ \ßþ7g"NyÞOܵçêÚ¿Bå«`ilB ˜„‘„­=öÌã¸æ@æ )UÖ 9yŽ€IÁ(±JÅ5<–§T`¬© M55ŠÖœR£h”ºäRê ®á1ÚŸúÌSßÍïÝ(yq(¬ ábŒÆuX&Àá &èq,–Ñ1Ç+à„±N97Î8Nüœsk`Ëq8­ ^—8%Ç àŠ½FMq.â†5„Sâhú@ìkO × Ápý$Áƒqù1¦/]}Œ©Îú/þÄ©ÿ»pÒ^`ÜD3F?©ìx”‘×Æ[ë±a ¯³1´ecÔÏŒ×Àäµ!/²„1êg)câdÜ?4dâ­K^˜|É ÆÐœ•ŒáQV1¦úÔ¿‰±'²š1tæ¬?ƺ9ëÁÏY?í¡œõÇГ³þ„rY‚ÞsÖŸŸõ'Äg)4ç¬3Å7ÎYgD¹¬3¢\ÖièÇbŸ-z±â3z´âs ,>G|ÆZV|ƾ´â3Öµâ3ü´â3qÄgônÅgè·â3tZñ½[ñ¾Yñ™ê‹ÏÐoÅgè,Äg¬[ˆÏàâ3ø…ø =…øL¹â3z/Ägâ‹ÏÄÏød ,gz)ÄôRˆÿ؇…øO5ù_±T“ÿ“‚êˆÿàT¼V *ŽÇM2G˜çªZN(:‘pTã’y¿ šË8úk‚n":÷qâw½{¿¬ŽÃYôÐ-7]´Ý×Á5‘ç2 endstream endobj 312 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œ.RIý÷ë73´»«ªÐóøÍøÍÃØW߶3ÛÏnÝjõèNÃelܬü¾;WWÕÐ\®?ÿp®uí4{ºSãÐlÝY]—›jÓwçOÞôÍë¥uësRá^ºþƒ‚uÔõ“û5sÍxЩֳæ0;]ŽnœÍL#å©;¿zê,å§ÔgSŠ ütã©ú;enµÖ>°îÛr8 ±S0qj>ÉÝw};ŠBõ ½ UÛ5gÑ»9x‡¼};ÝaÓï‡`¹TóG?y:o¤÷&˜ß­»þE]&ж—ãñÕAŒÒÁj¥Z·÷u½/?v§æ_týÎ|z;:ÒØ°ÆfhÝé¸kܸë_\°Ôz¥–u½ \ßþ7g"NyÞOܵçêÚ¿Bå«`ilB ˜„‘„­=öÌã¸æ@æ )UÖ 9yŽ€IÁ(±JÅ5<–§T`¬© M55ŠÖœR£h”ºäRê ®á1ÚŸúÌSßÍïÝ(yq(¬ ábŒÆuX&Àá &èq,–Ñ1Ç+à„±N97Î8Nüœsk`Ëq8­ ^—8%Ç àŠ½FMq.â†5„Sâhú@ìkO × Ápý$Áƒqù1¦/]}Œ©Îú/þÄ©ÿ»pÒ^`ÜD3F?©ìx”‘×Æ[ë±a ¯³1´ecÔÏŒ×Àäµ!/²„1êg)câdÜ?4dâ­K^˜|É ÆÐœ•ŒáQV1¦úÔ¿‰±'²š1tæ¬?ƺ9ëÁÏY?í¡œõÇГ³þ„rY‚ÞsÖŸŸõ'Äg)4ç¬3Å7ÎYgD¹¬3¢\ÖièÇbŸ-z±â3z´âs ,>G|ÆZV|ƾ´â3Öµâ3ü´â3qÄgônÅgè·â3tZñ½[ñ¾Yñ™ê‹ÏÐoÅgè,Äg¬[ˆÏàâ3ø…ø =…øL¹â3z/Ägâ‹ÏÄÏød ,gz)ÄôRˆÿ؇…øO5ù_±T“ÿ“‚êˆÿàT¼V *ŽÇM2G˜çªZN(:‘pTã’y¿ šË8úk‚n":÷qâw½{¿¬ŽÃYôÐ-7]´Ý×ÁXç7 endstream endobj 313 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UÁn£0½óÞC¥öƆ¦Š"R»­šjµ×œn¤"’ú÷ë73iwWU çñ›ñ›‡±¯¾=¬'®žý$¹ÕêчóØúIõ}sˆ®®ê¡=ï}úá}ç»ËìñN=ŒC»ö'u]­êU¿;Ýòªo_Ï¿°>'•þe×P°Žº~ò¿&¾÷Új=i÷“ãùàÇÉÉL4Ržv§×@ý‚¥Â”úlJQŸ~<î†þN™[­u,û®öhìMEœš^änw}7ŠBõ ½‘‰U·kO2¢w»!yýv<ùýªßÑ|®¦aòxßHïM4½;?îúuý™À@XŸ‡W1JG‹…êü6Ô ¾üØ콚~Ñõ;óéíàULcÃÛ¡óÇæõã¦ñÑ\ë…š7Í"ò}÷ßœI8åy{á.W7áë¤XDsƒdSÀT¤dHB 6ÀhpØ€Ó†620rª¬30 0Š“ƒQa•škÍk0jN©ÁXRšK0m8¥AÑ)MŤ45\#`´鳘]únoF±(ˆCam—3`4®ã*N×0Aψ㰌N9^gŒpι9°å8ñ Îm€Çá´.y]âT/kö5M¹ˆÖPΉ£é±¯ <1\?Ãõ³ ÆÕǘ¾tý1¦:Ë¿øNóo ¼Þų˜öë0àfš1úÉe7À#K^›`mÀ†1¼¶1ch³ cÔ·3ÆK`òÚ6cŒú6gLËýCƒ/h]òÂÌà‹-C³­Ã#[3¦úÔ¿I±'lÃ: ÖŸbÝ‚õ§à¬ŸöPÁúSè)XF¹¬?CïëωÏú3â³Îš Ö™ã¬3¡\Ö™P.ë4ôc±Ï½8ñ=:ñ9Ÿ‰#>c-'>c_:ñë:ñ~:ñ™8â3zwâ3ô;ñ:øŒÞø ßœøLõÅgèwâ3t–â3Ö-ÅgðKñüR|†žR|¦\ñ½—â3ñÅgâ[>ËY„^Jñ½”â?öa)þSMþWÕäÿ¤¤:â?85¯•Â‡šãi‚G“Ì湺‘ŠN$Õ¸dÞï‚ö<Žáš ›ˆÎ}œø»Þ¿_V‡á€,zè–»\´Ý7Ñz}ç< endstream endobj 314 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œ.RIý÷ë73´»«ªÐóøÍøÍÃØW߶3ÛÏnÝjõèNÃelܬü¾;WWÕÐ\®?ÿp®uí4{ºSãÐlÝY]—›jÓwçOÞôÍë¥uësRá^ºþƒ‚uÔõ“û5sÍx0ZëYs˜.G7ÎÎf¦‘òÔ_=õ –òSê³)E~ºñÔ ý2·~ÞÖ}[4v æ"NÍ'¹û®oGQ¨ž¡70¡j»æ,#z7ï’·o§³;lúý,—jþè'OçñôÞóû±uc׿¨ëÏzÂör<¾:ˆQ:X­Tëö¾®÷åÇîàÔü‹®ß™OoG§BÖØ ­;wwý‹ –Z¯Ô²®WëÛÿæLÄ)Ïû‰»ö\]ûW¨£|, ’MHS"#p òÐ# µÇ>y×È| #¥Ê:##Ï0)%V©¸†ÇÁ²£â” Œ55¡)°£FÑšSj­‘R—@J]!À5£w+>Ã7+>S}ñú­ø …øŒu ñüB|¿Ÿ¡§Ÿ)W|Fï…øL|ñ™øŸ „å,B/…ø^ ñû°ÿ©&ÿ+–jòRPñœŠ×ŠáCÅñ8ƒ£Iæó\UË E'Žj\2ïwAsGMÐMDç>Nü®wï—Õq8"‹ºå¦‹£û:ø‰9ç endstream endobj 315 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œ.RIý÷ë73´»«ªÐóøÍøÍÃØW߶3ÛÏnÝjõèNÃelܬü¾;WWÕÐ\®?ÿp®uí4{ºSãÐlÝY]—›jÓwçOÞôÍë¥uësRá^ºþƒ‚uÔõ“û5sÍx0¡Ö³æ0;]ŽnœÍL#å©;¿zê,å§ÔgSŠ ütã©ú;enµÖ>°îÛr8 ±S0qj>ÉÝw};ŠBõ ½ UÛ5gÑ»9x‡¼};ÝaÓï‡`¹TóG?y:o¤÷&˜ß­»þE]&ж—ãñÕAŒÒÁj¥Z·÷u½/?v§æ_týÎ|z;:ÒØ°ÆfhÝé¸kܸë_\°Ôz¥–u½ \ßþ7g"NyÞOܵçêÚ¿Bå«`ilB ˜„‘„­=öÌã¸æ@æ )UÖ 9yŽ€IÁ(±JÅ5<–§T`¬© M55ŠÖœR£h”ºäRê ®á1ÚŸúÌSßÍïÝ(yq(¬ ábŒÆuX&Àá &èq,–Ñ1Ç+à„±N97Î8Nüœsk`Ëq8­ ^—8%Ç àŠ½FMq.â†5„Sâhú@ìkO × Ápý$Áƒqù1¦/]}Œ©Îú/þÄ©ÿ»pÒ^`ÜD3F?©ìx”‘×Æ[ë±a ¯³1´ecÔÏŒ×Àäµ!/²„1êg)câdÜ?4dâ­K^˜|É ÆÐœ•ŒáQV1¦úÔ¿‰±'²š1tæ¬?ƺ9ëÁÏY?í¡œõÇГ³þ„rY‚ÞsÖŸŸõ'Äg)4ç¬3Å7ÎYgD¹¬3¢\ÖièÇbŸ-z±â3z´âs ,>G|ÆZV|ƾ´â3Öµâ3ü´â3qÄgônÅgè·â3tZñ½[ñ¾Yñ™ê‹ÏÐoÅgè,Äg¬[ˆÏàâ3ø…ø =…øL¹â3z/Ägâ‹ÏÄÏød ,gz)ÄôRˆÿ؇…øO5ù_±T“ÿ“‚êˆÿàT¼V *ŽÇM2G˜çªZN(:‘pTã’y¿ šË8úk‚n":÷qâw½{¿¬ŽÃYôÐ-7]´Ý×ÁÎ%ç# endstream endobj 316 0 obj << /Length 867 /Filter /FlateDecode >> stream xÚ}UÁn£0½óÞC¥öƆ¦Š"R»­šjµ×œn¤"’ú÷ë73iwWU çñ›ñ›‡±¯¾=¬'®žý$¹ÕêчóØúIõ}sˆ®®ê¡=ï}úá}ç»ËìñN=ŒC»ö'u]­êU¿;Ýòªo_Ï¿°>'•þe×P°Žº~ò¿&¾÷&í¤ÝOŽçƒ''3ÑHyÚ^õ – Sê³)E~úñ¸ú;enµÖ!°ì»jØ£±c4qjz‘»ÝõÝ( Õ3ôF&VÝ®=ɈÞí>8„äõÛñä÷«~;Dó¹š>†Éãi|#½7Ñô~ìü¸ë_Ôõga}>^=Ä(-ªóÛP7øòc³÷júE×ï̧·ƒW1 kl‡Î›Ö›þÅGs­jÞ4‹È÷Ýs&á”çí…» \Ý„W¬“bÍ ’MLS!"q Ø# uÀ!`NØÈÀÈ©²ÎÀ(À( LF…Uj®p4¯Á¨9¥cIMh ,ÁhP´á”E¤4ÒÔp€Ñþ¥Ïbvé»ý½Å¢ …µ!\΀ѸŽ« 8!\Ã=#ŽÃ2:åx œ1vÀ9çæÀ–ãÄ/8·v‡Óºäu‰Sq¼®ÙkÔ4 ç"nXCI8'Ž¦ľ&ðÄpý ×Ï2<WcúÒõǘê,ÿâ_8Í¿1ðbxÏbÚ ¬Ã€›iÆè'—Ý,ym‚µÆðÚÆŒ¡Í&ŒQßÎ/ÉkC^ØŒ1êÛœ1q,÷ V¼ uÉ 3ƒ/¶d Ͷb l͘êSÿ&Åž° cè,XŠu ÖŸ‚_°~ÚCëO¡§`ýå²þ ½¬?'>ëψÏ:sh.XgŽo\°Î„rYgB¹¬ÓÐÅ>;ôâÄgôèÄçX|&ŽøŒµœøŒ}éÄg¬ëÄgøéÄgâˆÏè݉ÏÐïÄgètâ3zwâ3|sâ3ÕŸ¡ß‰ÏÐYŠÏX·ŸÁ/ÅgðKñzJñ™rÅgô^ŠÏÄŸ‰oùd ,gz)ÅôRŠÿ؇¥øO5ù_qT“ÿ“’êˆÿàÔ¼V jŽ§ M2G˜çêFN(:‘pTã’y¿ Úó8†k‚n":÷qâïzÿ~Y†²è¡[îrÑbtßDÒÓçn endstream endobj 317 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­Újµ×œn¤"Hý÷ë73´»«ªÐóøÍøÍÃØWßžf¶í_Ü,ºÕêÑýehܬü¾=WWUß\Ž®;ÿp®uí4;Þ©‡¡ožÜY]—›jÓíÏ7ž¼éšÃ¥uësRá^÷Ýë¨ëg÷kæšñ`´Ö³æ8/'7ÌÎf¦‘ò¼?<õ –òSê³)E~ºaÜ÷Ý2·~ÞÖ][öG46s§æ“ÜݾkQ¨^ 70¡j÷ÍYFônŽÞ!$?½gwÜt»>X.ÕüÑOŽçáôÞóû¡uþ{Uן ô„§Ëétp£t°Z©Öí|]ïËíÑ©ù]¿3ŸßNN…46¬±é[7ž¶¶Ý« –Z¯Ô²®WëÚÿæLÄ)/»‰»ö\]ûW¨£|, ’MHS"#p òÐ# µÇ>y×È| #¥Ê:##Ï0)%V©¸†ÇÁ²£â” Œ55¡)°£FÑšSj­‘R—@J]!À5£w+>Ã7+>S}ñú­ø …øŒu ñüB|¿Ÿ¡§Ÿ)W|Fï…øL|ñ™øŸ „å,B/…ø^ ñû°ÿ©&ÿ+–jòRPñœŠ×ŠáCÅñ8ƒ£Iæó\UË E'Žj\2ïwAsMÐMDç>Nü}çÞ/«SB=tËM-F÷uð‰>ç endstream endobj 318 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œn¤" ‡þûõ›™´»«ªÐóøÍøÍÃØW߶3ÛöÏnÝjõèÆþ<4nV~ß‚««ªoÎG×M?œk]{™ïÔÃÐ7[7©ërSmºÃtãÉ›®y=·îÂúœT¸—C÷AÁ:êúÉýš¹fœŒÖzÖgãùä†Ùdf)O‡éÕS¿`)?¥>›RTà§ÆCßÝ)sëç}`ݵeDcc0qj~‘»?tí Õ3ô&Tí¡™dDïæèBòömœÜqÓíû`¹TóG?9NÃé½ æ÷Cë†C÷¢®?è Ûóéôê Fé`µR­ÛûºÞ—»£Só/º~g>½œ ilXcÓ·n<í7ìº,µ^©e]¯×µÿÍ™ˆSž÷îÚsuí_¡ŽòU°4H6!L‰@Œ@ÂÈBŒ€Öû@æq\s óŒ”*댌C¿Ÿ¡³Ÿ±n!>ƒ_ˆÏàâ3ôâ3åŠÏ轟‰/>?ã“°œEè¥ÿÑK!þcâ?ÕäÅRMþO ª#þƒSñZ1|¨8Gxp4Éaž«j9¡èDÂQKæý.hÎÃ௠º‰èÜljèÜûeuêOÈ¢‡n¹ËE‹Ñ}üMçA endstream endobj 319 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œ.R!‡þûõ›™´»«ªÐóøÍøÍÃØW߶3ÛÏnÝjõèNÃylܬü¾;WWÕÐœ®Ÿ~8׺ö2{ºSãÐlݤ®ËMµé»éÆ“7}óznÝ…õ9©p/]ÿAÁ:êúÉýš¹fêŒÖzÖf§óѳÉÌ4RžºéÕS¿`)?¥>›RTà§OÝÐß)sëç}`Ý·åp@c§`.âÔü"wßõí( Õ3ô&Tm×L2¢wsð!yûvšÜaÓï‡`¹TóG?yšÆ7Ò{ÌïÇÖ]ÿ¢®?è Ûóñøê Fé`µR­ÛûºÞ—»ƒSó/º~g>½ ilXc3´îtÜ5nÜõ/.Xj½R˺^®oÿ›3§<ï/ܵçêÚ¿Bå«`ilB ˜„‘„­=öÌã¸æ@æ )UÖ 9yŽ€IÁ(±JÅ5<–§T`¬© M55ŠÖœR£h”ºäRê ®á1Ú¿ô™/.}7¿w£XäÅ¡°6„‹0×a™G„+˜ ıXFǯ€Æ8åÜ8ã8ñsέ-Çá´.x]â”/€+ö5MĹˆÖPN‰£é±¯<1\?‚Ãõ“Æåǘ¾tõ1¦:ë¿øNýo ¼Þ…‹öë0à&š1úIe7À£Œ¼6ÞZ cx…Œ¡-‹£~¶`¼&¯ y‘%ŒQ?K'ãþ¡!/h]òÂ,àKV0†æ¬d ²Š1Õ§þMŒ=‘ÕŒ¡3gý1ÖÍY ~Îúiå¬?†žœõ'”Ëúôž³þ”ø¬?!>ëL¡9g)¾qÎ:#Êeå²NC?ûlÑ‹ŸÑ£Ÿc`ñ™8â3Ö²â3ö¥Ÿ±®Ÿá§Ÿ‰#>£w+>C¿Ÿ¡ÓŠÏèÝŠÏðÍŠÏT_|†~+>Cg!>cÝB|¿ŸÁ/Ägè)ÄgÊŸÑ{!>_|&~Æ'a9‹ÐK!þ£—BüÇ>,ĪÉÿŠ¥šüŸTGü§âµbøPq<Žðàh’9Â> stream xÚ}UÁn£0½óÞC¥öƆ¦Š"R»­Újµ×œn¤"Býûõ›™´»«ªÐóøÍøÍÃØßîg®žý,¹ÖêÁ‡ÓØúYõ}sˆ..ê¡=í}?ýð¾óÝyöx£îÇ¡}ô“º¬nëÛ~7]òmß¾ž:f}N*ýË®ÿ `uùäÍ|;MÚj=k÷³ãéàÇÙdf)O»é5P¿`©0¥>›RT৻¡¿QæZk뾫†=;Fs§æg¹Û]ߢP=CodbÕíÚIFôn÷Á!$?¾'¿¿í·C´\ªùC˜õoRì Û0†Î‚õ§X·`ý)øë§=T°þz ÖŸQ.ëÏÐ{Áúsâ³þŒø¬3‡æ‚uæøÆëL(—u&”Ë: ýXì³C/N|FN|NÅgâˆÏXˉÏØ—N|ƺN|†ŸN|&ŽøŒÞø ýN|†N'>£w'>Ã7'>S}ñúø ¥øŒuKñüR|¿Ÿ¡§Ÿ)W|Fï¥øL|ñ™ø–OÂr¡—RüG/¥ø}XŠÿT“ÿG5ù?)©ŽøNÍk¥ð¡æxšàÁÑ$s„y®nä„¢ G5.™÷» =c¸&è&¢s'þ®÷ï—Õa8 ‹ºåÎ-FwMô±çi endstream endobj 321 0 obj << /Length 866 /Filter /FlateDecode >> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œn¤"Bý÷ë73´»«ªÐóøÍøÍÃØW߶3ÛöÏnÝjõèÎýehܬü¾;WWUß\Ž®8׺vš=ß©‡¡o¶nT×å¦Út‡ñÆ“7]óziÝÄúœT¸—C÷AÁ:êúÉýš¹fÖzÖgçËÉ ³ÑÌ4Rž㫧~ÁR~J}6¥¨ÀO7œ}w§Ì­Ÿ÷u×–ýƒ¹ˆSóIîþеƒ(TÏИPµ‡f”½›£wÉÛ·ó莛nßË¥š?úÉó8¼‘Þ›`~?´n8t/êú3ž°½œN¯b”V+Õº½¯ë}ù±;:5ÿ¢ëwæÓÛɩƆ56}ëΧ]ã†]÷â‚¥Ö+µ¬ëUàºö¿9qÊó~â®=W×þê(_KƒdRÀ”Ä$ˆ| 4Àhí±dÇ52HÀH©²NÀÈÁÈsL F‰U*®áq°¬À¨8¥cMMh ¬Á¨Q´æ”Ek¤Ô%RWp ÑþÔg¾˜ún~ï±È‹Cam `4®Ã2ŽW0A/ˆc±ŒŽ9^'Œ-pʹ)pÆqâçœ[[ŽÃi]ðºÄ)9^Wì5jšˆs7¬¡ œGÓb_#xb¸~ †ë' ŒË1}éêcLuÖñ'Nýo ¼Þ…‹öë0à&š1úIe7À£Œ¼6ÞZ cx…Œ¡-‹£~¶`¼&¯ y‘%ŒQ?K'ãþ¡!/h]òÂ,àKV0†æ¬d ²Š1Õ§þMŒ=‘ÕŒ¡3gý1ÖÍY ~Îúiå¬?†žœõ'”Ëúôž³þ”ø¬?!>ëL¡9g)¾qÎ:#Êeå²NC?ûlÑ‹ŸÑ£Ÿc`ñ™8â3Ö²â3ö¥Ÿ±®Ÿá§Ÿ‰#>£w+>C¿Ÿ¡ÓŠÏèÝŠÏðÍŠÏT_|†~+>Cg!>cÝB|¿ŸÁ/Ägè)ÄgÊŸÑ{!>_|&~Æ'a9‹ÐK!þ£—BüÇ>,ĪÉÿŠ¥šüŸTGü§âµbøPq<Žðàh’9Â> stream xÚ}UMo£0½ó+¼‡Jí! á«Š"R»­šjµ×œn¤"Bý÷ë73´»«ªÐóøÍøÍÃØW߶3ÛöÏnÝjõèÎýehܬü¾;WWUß\Ž®8׺vš=ß©‡¡o¶nT×å¦Út‡ñÆ“7]óziÝÄúœT¸—C÷AÁ:êúÉýš¹fM¨õ¬9ÎΗ“f£™i¤<ÆWOý‚¥ü”úlJQŸn8úîN™[­µ¬»¶ìhìÌEœšOr÷‡®D¡z†ÞÀ„ª=4£ŒèݽCHÞ¾GwÜtû>X.ÕüÑOžÇáôÞóû¡uá{Qן ô„íåtzu£t°Z©Öí}]ïËÝÑ©ù]¿3ŸÞNN…46¬±é[w>í7ìº,µ^©e]¯×µÿÍ™ˆSž÷wí¹ºö¯PGù*X$›¦D F á@ä¡F@k} ó8®9ù@FJ•uFFž#`R0J¬Rq ƒeFÅ)kjBS` F¢5§Ô(Z#¥.9€”ºB€kxŒö§>óÅÔwó{7ˆE^ kC¸X£q– pD¸‚ zA‹etÌñ 8alSÎM3Ž?çÜØrNë‚×%NÉñ¸b¯QÓDœ‹¸a á”8š>ûÁÃõ#h0\?Ið`\~ŒéKWcª³þ‹?qêcà…ð.\„´X‡7ÑŒÑO*»eäµñÖzlÃë,d mYÄõ³ã50ymÈ‹,aŒúYʘ8÷ ™xAë’f_²‚14g%cx”UŒ©>õob쉬f 9ë±nÎúcðsÖO{(gý1ôä¬?¡\ÖŸ ÷œõ§Ägý ñYg Í9ëLñsÖQ.ëŒ(—uú±Øg‹^¬øŒ­ø‹ÏÄŸ±–Ÿ±/­øŒu­ø ?­øLñ½[ñú­ø V|FïV|†oV|¦úâ3ô[ñ: ñëâ3ø…ø ~!>CO!>S®øŒÞ ñ™øâ3ñ3>ËY„^ ñ½â?öa!þSMþW,Õäÿ¤ :â?8¯ÇŠãq„G“Ì湪–ŠN$Õ¸dÞï‚æ2 þš ›ˆÎ}œø‡Î½_V§þ„,zè–›.ZŒîëàÍçP endstream endobj 340 0 obj << /Producer (pdfTeX-1.40.25) /Author()/Title()/Subject()/Creator(LaTeX with hyperref)/Keywords() /CreationDate (D:20241029220815-04'00') /ModDate (D:20241029220815-04'00') /Trapped /False /PTEX.Fullbanner (This is pdfTeX, Version 3.141592653-2.6-1.40.25 (TeX Live 2023/Debian) kpathsea version 6.3.5) >> endobj 287 0 obj << /Type /ObjStm /N 59 /First 496 /Length 2494 /Filter /FlateDecode >> stream xÚµZYoã¶~÷¯àãä)W‘PHœ‰ã쉳ó Ør¢Žc¥–á3|‰op„‡x„cü‚_ñWœâwœá|Ç¡:L`)¾CÐuç÷ßײóþ¨·|vN)!d‰à»œO X=äj[;wÃ5 ÝÃû¸‹ðg0ùŒ¾ÅwÖìa:I§Öøx:Š²W<ÆãÄñúóý5žâ™à7<µ¼¼Ç³$á¿ðÌP„çøþV©Ò›PõxryòØ3Tݬ J  þ£² 8©G¥ªÌcIcì?Quˆûø´ ì _ãøŠ#n˜¾½EŽÀ¿æÑÄò÷͸ÒË,ŽòxVgrgYAg: ¥Ñ,žNâqîîf~ŸÌ3`ù¯yšÇ£çIAwþ:‹cœOøùtϲa:‹Ýà¿ãYZ9HˆDïêá⪠qÓ_~µ —¥ÆeÙÖ.KÅq\PÛ œUc›·ßè=ãnˆ^fœI‚»<€252 ÝÖËÖI<ŸÁ²(&IžLF±çO1xP%üþ´nS†ŸÙs†Æ+²ø¨e»9¸Â÷*Mbš.û'W‡–¦å>À ôÁ™ú?#Y+—G¢…£’·Z¾j ®"‰ùÁµ ÐÐh"ªJ¥Ü(¯õÎÏ/œÇQ¶œJ¶ Š«fáÿ“ÊvºJ7käx•*-ùýót˜Ž’é `LÆã˜ÆúƒÊ’d{‹{ ˆÑ¢æ@S‡Ç¶îŒ!ŽhÞJUèÛ^i=O0ÔåK—Š ‹ +œ&7ÚtkŽ×eÑq:Ÿ¹T%?Š(²¼áiŠŽ§,~KÜMÛ.Éz´J©èá#pÒc|R8êE­hÜCŸ„QHñó,~sc$ÑÅ£³% ¼ÌŽBU›ïßV<Ü/Åß]!€FQzÉåKÓgÁ]æϹ}4Bð­ý(‹Í‚ÇÃÞÓõ/ݳÏ”tj.ŽXà2 >LfYn¼ ÚÈ>Š ­ ¾OFùkf«{“ÞNÇQ íú’¢´ ÑÉþþÝÍ :ë·"’uDö1òHÈ&"º9¢ã§‹Ëþ¡E´Òu@”H‘y*9¢MDlsDƒ«‡ËãžE¤šˆ$©#Rt)E¼ ˆohïüâô¤ €®ÛÎL²: “ÒJ@¦_(ñM£(€í¡EÚ+ /f]ÎâoˆUóˆ¹8Ö=èá[gžÇ?Œå-3ƒr&«ÌäÕMý^bÁ™´xùw&í²Úút±~‘ÿÚ×/ ÃÉDg‘ãÍ:bÕ:Ì_‡´¬SrLÉŠu|ÚjEÐ.Sò]¥»xÉ/mßÎmá7AÈgWWÈ¥‹µÍÇGðÇOépwG³|¹ ôéhœ¦ù4Íã_érÝ´/d Ôu!ßAa1}˜Àó8Îã) vÀÓ•'þš¼Y¡ƒû &oI^Ãá«~©ãµcésDÌŽþÒÙð5ümVµ/Σá×IgDí @z#ù,ýM³¯Ñψ€¡®Ù-ƾ§“‘”ha”ù —'éôWbHa¾«ªÜ° †J* Lãäe>³”kZ‘%VUâF¤*"a¨&‘!Zê:Òêöº œÅˆaA•…IVAj#bQhD.Ý~z‡lh‰#þ31ªðcßR­Í À‚Œ(E°ˆQéK€*ÆO*Ò××OφH÷aÄUn¸Pj‡d(gÔ[Í°Ç}áNhO`˜“>@Û.Y<4gõ?Ë]Ð ¶ôÕ¸>Æ ÝŒ×¦F*êRCäÂéJ©¡NÕd@Ÿ"5P¨^Ááë.G ¦©:2 4¬ïg8%u`†W¢kBÃ-c¡¡–/ë±»‚\£´º¬‰Ñê)Z× «"X¿<’ùs¹“ͼ q`ÄL-kù¦ÝìÃ盕ÕjuA¹dÑÂÁÅ oáE^â®ÄÛ?(±Wב,Kùm\lP„wm‚ýëwÿâb«q½'çr%®%‹@C ºœ·¼8v£<š¤/×~t‹‹®Î«ÜöP8/ûš—ø :j|›Å-àÅ{<ݳ/»ƒriTŠ endstream endobj 341 0 obj << /Type /XRef /Index [0 342] /Size 342 /W [1 3 1] /Root 339 0 R /Info 340 0 R /ID [ ] /Length 896 /Filter /FlateDecode >> stream xÚ%ÔKlUUÅñï;÷Õmi¥¶E ±ÕíøU¥ûa³Û®_4Ü÷ÁvØÁ~z§c0`Ømô=}bvÁnxö¸M iõAØ û`?Œ¸<‘ üV™.9 Çà Û©jÍMr_ÕqŠ´tÃq{·J·/ŸÒ°èv帒º›nôY­†t<ÊØO å^hz^ œààÇ%ònß¼¯z³Ýn–4¬u·×”êÜ_~N©*Ý~hÓ–„Χ"t*jÜfz´Ê÷Î[PX©Ž7.jU-æe­ •‹îÆ·nèú‹Ýné‡úK€CœÆhqŸÖ*]‹î/^×:5‹v ¬Ñá~á¬VéPt¹Ûª!5¼«&Ð!Ñû£f±`#PÇØ Ô,a+Ц MA›bì„! z1 Ô'¨OPŸØ4'öº_îÐísÿóö;ØïYãJ#žu×)z6¦ÃcžM?«4îÙ;×”xvá{¥ƒžÍü­tÈs57•{n´_iÂs§o¿Ø#ž›ùMé¨ç+~RšôüjlLyþê3JǼжžãfóÒõóÂî‡4t/œ™VʼX3W)çÅsJy/ž¿¤TðRç¨RÑK'—§]^x8ñÒc‰WžLœšN¼úyâ¿o8{:ñÁ¶Ä‡ÿ&â“ÄGS‰Oëç/'¾x:ñUGââ‰ï^O\Úžøñ¿Ä•?Oê¾úÃÓ];¬„X«¡º T‹è…µÐý°ÎK×æ¤í?aÿÍìÐ, endstream endobj startxref 516484 %%EOF pcaMethods/man/0000755000175200017520000000000014710217306014437 5ustar00biocbuildbiocbuildpcaMethods/man/BPCA_dostep.Rd0000644000175200017520000000132414710217306017011 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/BPCA_dostep.R \name{BPCA_dostep} \alias{BPCA_dostep} \title{Do BPCA estimation step} \usage{ BPCA_dostep(M, y) } \arguments{ \item{M}{Data structure containing all needed information. See the source documentation of BPCA_initmodel for details} \item{y}{Numeric original data matrix} } \value{ Updated version of the data structure } \description{ The function contains the actual implementation of the BPCA component estimation. It performs one step of the BPCA EM algorithm. It is called 'maxStep' times from within the main loop in BPCAestimate. } \details{ This function is NOT intended to be run standalone. } \author{ Wolfram Stacklies } pcaMethods/man/BPCA_initmodel.Rd0000644000175200017520000000267614710217306017512 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/BPCA_initmodel.R \name{BPCA_initmodel} \alias{BPCA_initmodel} \title{Initialize BPCA model} \usage{ BPCA_initmodel(y, components) } \arguments{ \item{y}{numeric matrix containing missing values. Missing values are denoted as 'NA'} \item{components}{Number of components used for estimation} } \value{ List containing \item{rows}{Row number of input matrix} \item{cols}{Column number of input matrix} \item{comps}{Number of components to use} \item{yest}{(working variable) current estimate of complete data} \item{row_miss}{(Array) Indizes of rows containing missing values} \item{row_nomiss}{(Array) Indices of complete rows (such with no missing values)} \item{nans}{Matrix of same size as input data. TRUE if \code{input == NA}, false otherwise} \item{mean}{Column wise data mean} \item{PA}{ (d x k) Estimated principal axes (eigenvectors, loadings) The matrix ROWS are the vectors} \item{tau}{Estimated precision of the residual error} \item{scores}{ Estimated scores} Further elements are: galpha0, balpha0, alpha, gmu0, btau0, gtau0, SigW. These are working variables or constants. } \description{ Model initialization for Bayesian PCA. This function is NOT inteded to be run separately! } \details{ The function calculates the initial Eigenvectors by use of SVD from the complete rows. The data structure M is created and initial values are assigned. } \author{ Wolfram Stacklies } pcaMethods/man/DModX-pcaRes-method.Rd0000644000175200017520000000344414710217306020377 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{DModX,pcaRes-method} \alias{DModX,pcaRes-method} \alias{DModX} \title{DModX} \usage{ DModX(object, dat, newdata=FALSE, type=c("normalized","absolute"), ...) } \arguments{ \item{object}{a pcaRes object} \item{dat}{the original data, taken from \code{completeObs} if left missing.} \item{newdata}{logical indicating if this data was part of the training data or not. If it was, it is adjusted by a near one factor \eqn{v=(N/ (N-A-A0))^-1}} \item{type}{if absolute or normalized values should be given. Normalized values are adjusted to the the total RSD of the model.} \item{...}{Not used} } \value{ A vector with distances from observations to the PCA model } \description{ Distance to the model of X-space. } \details{ Measures how well described the observations are, i.e. how well they fit in the mode. High DModX indicate a poor fit. Defined as: \eqn{\frac{\sqrt{\frac{SSE_i}{K-A}}}{\sqrt{\frac{SSE}{(N-A-A_0)(K-A)}}}} For observation \eqn{i}, in a model with \eqn{A} components, \eqn{K} variables and \eqn{N} obserations. SSE is the squared sum of the residuals. \eqn{A_0} is 1 if model was centered and 0 otherwise. DModX is claimed to be approximately F-distributed and can therefore be used to check if an observation is significantly far away from the PCA model assuming normally distributed data. Pass original data as an argument if the model was calculated with \code{completeObs=FALSE}. } \examples{ data(iris) pcIr <- pca(iris[,1:4]) with(iris, plot(DModX(pcIr)~Species)) } \references{ Introduction to Multi- and Megavariate Data Analysis using Projection Methods (PCA and PLS), L. Eriksson, E. Johansson, N. Kettaneh-Wold and S. Wold, Umetrics 1999, p. 468 } \author{ Henning Redestig } pcaMethods/man/Q2.Rd0000644000175200017520000000654714710217306015224 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/xval.R \name{Q2} \alias{Q2} \title{Cross-validation for PCA} \usage{ Q2(object, originalData = completeObs(object), fold = 5, nruncv = 1, type = c("krzanowski", "impute"), verbose = interactive(), variables = 1:nVar(object), ...) } \arguments{ \item{object}{A \code{pcaRes} object (result from previous PCA analysis.)} \item{originalData}{The matrix (or ExpressionSet) that used to obtain the pcaRes object.} \item{fold}{The number of groups to divide the data in.} \item{nruncv}{The number of times to repeat the whole cross-validation} \item{type}{krzanowski or imputation type cross-validation} \item{verbose}{\code{boolean} If TRUE Q2 outputs a primitive progress bar.} \item{variables}{indices of the variables to use during cross-validation calculation. Other variables are kept as they are and do not contribute to the total sum-of-squares.} \item{...}{Further arguments passed to the \code{\link{pca}} function called within Q2.} } \value{ A matrix or vector with \eqn{Q^2} estimates. } \description{ Internal cross-validation can be used for estimating the level of structure in a data set and to optimise the choice of number of principal components. } \details{ This method calculates \eqn{Q^2} for a PCA model. This is the cross-validated version of \eqn{R^2} and can be interpreted as the ratio of variance that can be predicted independently by the PCA model. Poor (low) \eqn{Q^2} indicates that the PCA model only describes noise and that the model is unrelated to the true data structure. The definition of \eqn{Q^2} is: \deqn{Q^2=1 - \frac{\sum_{i}^{k}\sum_{j}^{n}(x - \hat{x})^2}{\sum_{i}^{k}\sum_{j}^{n}x^2}}{Q^2=1 - sum_i^k sum_j^n (x - \hat{x})^2 / \sum_i^k \sum_j^n(x^2)} for the matrix \eqn{x} which has \eqn{n} rows and \eqn{k} columns. For a given number of PC's x is estimated as \eqn{\hat{x}=TP'} (T are scores and P are loadings). Although this defines the leave-one-out cross-validation this is not what is performed if fold is less than the number of rows and/or columns. In 'impute' type CV, diagonal rows of elements in the matrix are deleted and the re-estimated. In 'krzanowski' type CV, rows are sequentially left out to build fold PCA models which give the loadings. Then, columns are sequentially left out to build fold models for scores. By combining scores and loadings from different models, we can estimate completely left out values. The two types may seem similar but can give very different results, krzanowski typically yields more stable and reliable result for estimating data structure whereas impute is better for evaluating missing value imputation performance. Note that since Krzanowski CV operates on a reduced matrix, it is not possible estimate Q2 for all components and the result vector may therefore be shorter than \code{nPcs(object)}. } \examples{ data(iris) x <- iris[,1:4] pcIr <- pca(x, nPcs=3) q2 <- Q2(pcIr, x) barplot(q2, main="Krzanowski CV", xlab="Number of PCs", ylab=expression(Q^2)) ## q2 for a single variable Q2(pcIr, x, variables=2) pcIr <- pca(x, nPcs=3, method="nipals") q2 <- Q2(pcIr, x, type="impute") barplot(q2, main="Imputation CV", xlab="Number of PCs", ylab=expression(Q^2)) } \references{ Krzanowski, WJ. Cross-validation in principal component analysis. Biometrics. 1987(43):3,575-584 } \author{ Henning Redestig, Ondrej Mikula } \keyword{multivariate} pcaMethods/man/R2VX-pcaRes-method.Rd0000644000175200017520000000144214710217306020161 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{R2VX,pcaRes-method} \alias{R2VX,pcaRes-method} \alias{R2VX} \title{R2 goodness of fit} \usage{ \S4method{R2VX}{pcaRes}(object, direction = c("variables", "observations", "complete"), data = completeObs(object), pcs = nP(object)) } \arguments{ \item{object}{a PCA model object} \item{direction}{choose between calculating R2 per variable, per observation or for the entire data with 'variables', 'observations' or 'complete'.} \item{data}{the data used to fit the model} \item{pcs}{the number of PCs to use to calculate R2} } \value{ A vector with R2 values } \description{ Flexible calculation of R2 goodness of fit. } \examples{ R2VX(pca(iris)) } \author{ Henning Redestig } pcaMethods/man/R2cum-pcaRes-method.Rd0000644000175200017520000000104714710217306020411 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{R2cum,pcaRes-method} \alias{R2cum,pcaRes-method} \alias{R2cum} \title{Cumulative R2 is the total ratio of variance that is being explained by the model} \usage{ \S4method{R2cum}{pcaRes}(object, ...) } \arguments{ \item{object}{a \code{pcaRes} model} \item{...}{Not used} } \value{ Get the cumulative R2 } \description{ Cumulative R2 is the total ratio of variance that is being explained by the model } \author{ Henning Redestig } pcaMethods/man/RnipalsPca.Rd0000644000175200017520000000423114710217306016762 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/nipalsPca.R \name{RnipalsPca} \alias{RnipalsPca} \title{NIPALS PCA implemented in R} \usage{ RnipalsPca(Matrix, nPcs = 2, varLimit = 1, maxSteps = 5000, threshold = 1e-06, verbose = interactive(), ...) } \arguments{ \item{Matrix}{Pre-processed (centered, scaled) numerical matrix samples in rows and variables as columns.} \item{nPcs}{Number of components that should be extracted.} \item{varLimit}{Optionally the ratio of variance that should be explained. \code{nPcs} is ignored if varLimit < 1} \item{maxSteps}{Defines how many iterations can be done before algorithm should abort (happens almost exclusively when there were some wrong in the input data).} \item{threshold}{The limit condition for judging if the algorithm has converged or not, specifically if a new iteration is done if \eqn{(T_{old} - T)^T(T_{old} - T) > \code{limit}}.} \item{verbose}{Show simple progress information.} \item{...}{Only used for passing through arguments.} } \value{ A \code{pcaRes} object. } \description{ PCA by non-linear iterative partial least squares } \details{ Can be used for computing PCA on a numeric matrix using either the NIPALS algorithm which is an iterative approach for estimating the principal components extracting them one at a time. NIPALS can handle a small amount of missing values. It is not recommended to use this function directely but rather to use the pca() wrapper function. There is a C++ implementation given as \code{nipalsPca} which is faster. } \examples{ data(metaboliteData) mat <- prep(t(metaboliteData)) ## c++ version is faster system.time(pc <- RnipalsPca(mat, method="rnipals", nPcs=2)) system.time(pc <- nipalsPca(mat, nPcs=2)) ## better use pca() pc <- pca(t(metaboliteData), method="rnipals", nPcs=2) \dontshow{stopifnot(sum((fitted(pc) - t(metaboliteData))^2, na.rm=TRUE) < 200)} } \references{ Wold, H. (1966) Estimation of principal components and related models by iterative least squares. In Multivariate Analysis (Ed., P.R. Krishnaiah), Academic Press, NY, 391-420. } \seealso{ \code{prcomp}, \code{princomp}, \code{pca} } \author{ Henning Redestig } \keyword{multivariate} pcaMethods/man/asExprSet.Rd0000644000175200017520000000174114710217306016647 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-ExpressionSet.R \name{asExprSet} \alias{asExprSet} \title{Convert pcaRes object to an expression set} \usage{ asExprSet(object, exprSet) } \arguments{ \item{object}{\code{pcaRes} -- The object containing the completed data.} \item{exprSet}{\code{ExpressionSet} -- The object passed on to \code{pca} for missing value estimation.} } \value{ An object without missing values of class \code{ExpressionSet}. } \description{ This function can be used to conveniently replace the expression matrix in an \code{ExpressionSet} with the completed data from a \code{pcaRes} object. } \details{ This is not a standard \code{as} function as \code{pcaRes} object alone not can be converted to an \code{ExpressionSet} (the \code{pcaRes} object does not hold any \code{phenoData} for example). } \author{ Wolfram Stacklies \cr CAS-MPG Partner Institute for Computational Biology, Shanghai, China } \keyword{multivariate} pcaMethods/man/biplot-methods.Rd0000644000175200017520000000356614710217306017672 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{biplot-methods} \alias{biplot-methods} \alias{biplot.pcaRes} \alias{biplot,pcaRes-method} \title{Plot a overlaid scores and loadings plot} \usage{ \method{biplot}{pcaRes}(x, choices = 1:2, scale = 1, pc.biplot = FALSE, ...) \S4method{biplot}{pcaRes}(x, choices = 1:2, scale = 1, pc.biplot = FALSE, ...) } \arguments{ \item{x}{a pcaRes object} \item{choices}{which two pcs to plot} \item{scale}{The variables are scaled by \eqn{\lambda^{scale}}{lambda^scale} and the observations are scaled by \eqn{\lambda^{scale}}{lambda ^ (1-scale)} where \code{lambda} are the singular values as computed by \code{princomp}. Normally \eqn{0\le{}scale\le{}1}{0 <= scale <= 1}, and a warning will be issued if the specified 'scale' is outside this range.} \item{pc.biplot}{If true, use what Gabriel (1971) refers to as a "principal component biplot", with \eqn{\lambda=1}{lambda = 1} and observations scaled up by sqrt(n) and variables scaled down by sqrt(n). Then the inner products between variables approximate covariances and distances between observations approximate Mahalanobis distance.} \item{...}{optional arguments to be passed to \code{biplot.default}.} } \value{ a plot is produced on the current graphics device. } \description{ Visualize two-components simultaneously } \details{ This is a method for the generic function 'biplot'. There is considerable confusion over the precise definitions: those of the original paper, Gabriel (1971), are followed here. Gabriel and Odoroff (1990) use the same definitions, but their plots actually correspond to \code{pc.biplot = TRUE}. } \examples{ data(iris) pcIr <- pca(iris[,1:4]) biplot(pcIr) } \seealso{ \code{prcomp}, \code{pca}, \code{princomp} } \author{ Kevin Wright, Adapted from \code{biplot.prcomp} } \keyword{multivariate} pcaMethods/man/bpca.Rd0000644000175200017520000001332014710217306015632 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/bpca.R \name{bpca} \alias{bpca} \title{Bayesian PCA missing value estimation} \usage{ bpca(Matrix, nPcs = 2, maxSteps = 100, verbose = interactive(), threshold = 1e-04, ...) } \arguments{ \item{Matrix}{\code{matrix} -- Pre-processed matrix (centered, scaled) with variables in columns and observations in rows. The data may contain missing values, denoted as \code{NA}.} \item{nPcs}{\code{numeric} -- Number of components used for re-estimation. Choosing few components may decrease the estimation precision.} \item{maxSteps}{\code{numeric} -- Maximum number of estimation steps.} \item{verbose}{\code{boolean} -- BPCA prints the number of steps and the increase in precision if set to TRUE. Default is interactive().} \item{threshold}{convergence threshold} \item{...}{Reserved for future use. Currently no further parameters are used} } \value{ Standard PCA result object used by all PCA-based methods of this package. Contains scores, loadings, data mean and more. See \code{\link{pcaRes}} for details. } \description{ Implements a Bayesian PCA missing value estimator. The script is a port of the Matlab version provided by Shigeyuki OBA. See also \url{http://ishiilab.jp/member/oba/tools/BPCAFill.html}. BPCA combines an EM approach for PCA with a Bayesian model. In standard PCA data far from the training set but close to the principal subspace may have the same reconstruction error. BPCA defines a likelihood function such that the likelihood for data far from the training set is much lower, even if they are close to the principal subspace. } \details{ Scores and loadings obtained with Bayesian PCA slightly differ from those obtained with conventional PCA. This is because BPCA was developed especially for missing value estimation. The algorithm does not force orthogonality between factor loadings, as a result factor loadings are not necessarily orthogonal. However, the BPCA authors found that including an orthogonality criterion made the predictions worse. The authors also state that the difference between real and predicted Eigenvalues becomes larger when the number of observation is smaller, because it reflects the lack of information to accurately determine true factor loadings from the limited and noisy data. As a result, weights of factors to predict missing values are not the same as with conventional PCA, but the missing value estimation is improved. BPCA works iteratively, the complexity is growing with \eqn{O(n^3)}{O(n^3)} because several matrix inversions are required. The size of the matrices to invert depends on the number of components used for re-estimation. Finding the optimal number of components for estimation is not a trivial task; the best choice depends on the internal structure of the data. A method called \code{kEstimate} is provided to estimate the optimal number of components via cross validation. In general few components are sufficient for reasonable estimation accuracy. See also the package documentation for further discussion about on what data PCA-based missing value estimation makes sense. It is not recommended to use this function directely but rather to use the pca() wrapper function. There is a difference with respect the interpretation of rows (observations) and columns (variables) compared to matlab implementation. For estimation of missing values for microarray data, the suggestion in the original bpca is to intepret genes as observations and the samples as variables. In pcaMethods however, genes are interpreted as variables and samples as observations which arguably also is the more natural interpretation. For bpca behavior like in the matlab implementation, simply transpose your input matrix. Details about the probabilistic model underlying BPCA are found in Oba et. al 2003. The algorithm uses an expectation maximation approach together with a Bayesian model to approximate the principal axes (eigenvectors of the covariance matrix in PCA). The estimation is done iteratively, the algorithm terminates if either the maximum number of iterations was reached or if the estimated increase in precision falls below \eqn{1e^{-4}}{1e^-4}. \bold{Complexity:} The relatively high complexity of the method is a result of several matrix inversions required in each step. Considering the case that the maximum number of iteration steps is needed, the approximate complexity is given by the term \deqn{maxSteps \cdot row_{miss} \cdot O(n^3)}{maxSteps * row_miss * O(n^3)} Where \eqn{row_{miss}}{row_miss} is the number of rows containing missing values and \eqn{O(n^3)}{O(n^3)} is the complexity for inverting a matrix of size \eqn{components}{components}. Components is the number of components used for re-estimation. } \note{ Requires \code{MASS}. } \examples{ ## Load a sample metabolite dataset with 5\\\% missig values (metaboliteData)e data(metaboliteData) ## Perform Bayesian PCA with 2 components pc <- pca(t(metaboliteData), method="bpca", nPcs=2) ## Get the estimated principal axes (loadings) loadings <- loadings(pc) ## Get the estimated scores scores <- scores(pc) ## Get the estimated complete observations cObs <- completeObs(pc) ## Now make a scores and loadings plot slplot(pc) \dontshow{stopifnot(sum((fitted(pc) - t(metaboliteData))^2, na.rm=TRUE) < 200)} } \references{ Shigeyuki Oba, Masa-aki Sato, Ichiro Takemasa, Morito Monden, Ken-ichi Matsubara and Shin Ishii. A Bayesian missing value estimation method for gene expression profile data. \emph{Bioinformatics, 19(16):2088-2096, Nov 2003}. } \seealso{ \code{\link{ppca}}, \code{\link{svdImpute}}, \code{\link{prcomp}}, \code{\link{nipalsPca}}, \code{\link{pca}}, \code{\link{pcaRes}}. \code{\link{kEstimate}}. } \author{ Wolfram Stacklies } \keyword{multivariate} pcaMethods/man/center-pcaRes-method.Rd0000644000175200017520000000070214710217306020676 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{center,pcaRes-method} \alias{center,pcaRes-method} \alias{center} \title{Get the centers of the original variables} \usage{ center(object, ...) } \arguments{ \item{object}{pcaRes object} \item{...}{Not used} } \value{ Vector with the centers } \description{ Get the centers of the original variables } \author{ Henning Redestig } pcaMethods/man/centered-pcaRes-method.Rd0000644000175200017520000000070514710217306021212 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{centered,pcaRes-method} \alias{centered,pcaRes-method} \alias{centered} \title{Check centering was part of the model} \usage{ centered(object, ...) } \arguments{ \item{object}{pcaRes object} \item{...}{Not used} } \value{ TRUE if model was centered } \description{ Check centering was part of the model } \author{ Henning Redestig } pcaMethods/man/checkData.Rd0000644000175200017520000000242514710217306016600 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/checkData.R \name{checkData} \alias{checkData} \title{Do some basic checks on a given data matrix} \usage{ checkData(data, verbose = FALSE) } \arguments{ \item{data}{\code{matrix} -- Data to check.} \item{verbose}{\code{boolean} -- If TRUE, the function prints messages whenever an error in the data set is found.} } \value{ \item{isValid}{\code{boolean} -- TRUE if no errors were found, FALSE otherwise. isValid contains a set of attributes, these are: \itemize{ \item isNumeric - TRUE if data is numeric, false otherwise \item isInfinite - TRUE if data contains 'Inf' values, false otherwise \item isNaN - TRUE if data contains 'NaN' values, false otherwise \item isMatrix - TRUE if the data is in matrix format, FALSE otherwise \item naRows - TRUE if data contains rows in which all elements are 'NA', FALSE otherwise \item naCols - TRUE if data contains columns in which all elements are 'NA', FALSE otherwise }} } \description{ Check a given data matrix for consistency with the format required for further analysis. The data must be a numeric matrix and not contain: \itemize{ \item Inf values \item NaN values \item Rows or columns that consist of NA only } } \author{ Wolfram Stacklies } \keyword{multivariate} pcaMethods/man/completeObs-nniRes-method.Rd0000644000175200017520000000111514710217306021712 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{completeObs,nniRes-method} \alias{completeObs,nniRes-method} \alias{completeObs} \alias{completeObs,pcaRes-method} \title{Get the original data with missing values replaced with predicted values.} \usage{ completeObs(object, ...) } \arguments{ \item{object}{object to fetch complete data from} \item{...}{Not used} } \value{ Completed data (matrix) } \description{ Get the original data with missing values replaced with predicted values. } \author{ Henning Redestig } pcaMethods/man/cvseg.Rd0000644000175200017520000000170014710217306016033 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/xval.R \name{cvseg} \alias{cvseg} \title{Get CV segments} \usage{ cvseg(x, fold = 7, seed = NULL) } \arguments{ \item{x}{a factor, character or numeric vector that describes class membership of a set of items, or, a numeric vector indicating unique indices of items, or, a numeric of length 1 that describes the number of items to segment (without any classes)} \item{fold}{the desired number of segments} \item{seed}{randomization seed for reproducibility} } \value{ a list where each element is a set of indices that defines the CV segment. } \description{ Get cross-validation segments that have (as far as possible) the same ratio of all classes (if classes are present) } \examples{ seg <- cvseg(iris$Species, 10) sapply(seg, function(s) table(iris$Species[s])) cvseg(20, 10) } \seealso{ the \code{cvsegments} function in the \code{pls} package } \author{ Henning Redestig } pcaMethods/man/cvstat-pcaRes-method.Rd0000644000175200017520000000071714710217306020730 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{cvstat,pcaRes-method} \alias{cvstat,pcaRes-method} \alias{cvstat} \title{Get cross-validation statistics (e.g. \eqn{Q^2}).} \usage{ cvstat(object, ...) } \arguments{ \item{object}{pcaRes object} \item{...}{not used} } \value{ vector CV statistics } \description{ Get cross-validation statistics (e.g. \eqn{Q^2}). } \author{ Henning Redestig } pcaMethods/man/deletediagonals.Rd0000644000175200017520000000134414710217306020054 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/xval.R \name{deletediagonals} \alias{deletediagonals} \title{Delete diagonals} \usage{ deletediagonals(x, diagonals = 1) } \arguments{ \item{x}{The matrix} \item{diagonals}{The diagonal to be replaced, i.e. the first, second and so on when looking at the fat version of the matrix (transposed or not) counting from the bottom. Can be a vector to delete more than one diagonal.} } \value{ The original matrix with some values missing } \description{ Replace a diagonal of elements of a matrix with NA } \details{ Used for creating artifical missing values in matrices without causing any full row or column to be completely missing } \author{ Henning Redestig } pcaMethods/man/derrorHierarchic.Rd0000644000175200017520000000061414710217306020206 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/derrorHierarchic.R \name{derrorHierarchic} \alias{derrorHierarchic} \title{Later} \usage{ derrorHierarchic(nlnet, trainIn, trainOut) } \arguments{ \item{nlnet}{the nlnet} \item{trainIn}{training data} \item{trainOut}{fitted data} } \value{ derror } \description{ Later } \author{ Henning Redestig, Matthias Scholz } pcaMethods/man/dim.pcaRes.Rd0000644000175200017520000000054514710217306016717 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \name{dim.pcaRes} \alias{dim.pcaRes} \title{Dimensions of a PCA model} \usage{ \method{dim}{pcaRes}(x) } \arguments{ \item{x}{a pcaRes object} } \value{ Get the dimensions of this PCA model } \description{ Dimensions of a PCA model } \author{ Henning Redestig } pcaMethods/man/errorHierarchic.Rd0000644000175200017520000000060714710217306020044 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/errorHierarchic.R \name{errorHierarchic} \alias{errorHierarchic} \title{Later} \usage{ errorHierarchic(nlnet, trainIn, trainOut) } \arguments{ \item{nlnet}{The nlnet} \item{trainIn}{training data} \item{trainOut}{fitted data} } \value{ error } \description{ Later } \author{ Henning Redestig, Matthias Scholz } pcaMethods/man/fitted-methods.Rd0000644000175200017520000000324214710217306017647 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{fitted-methods} \alias{fitted-methods} \alias{fitted.pcaRes} \alias{fitted,pcaRes-method} \title{Extract fitted values from PCA.} \usage{ \method{fitted}{pcaRes}(object, data = NULL, nPcs = nP(object), pre = TRUE, post = TRUE, ...) \S4method{fitted}{pcaRes}(object, data = NULL, nPcs = nP(object), pre = TRUE, post = TRUE, ...) } \arguments{ \item{object}{the \code{pcaRes} object of interest.} \item{data}{For standard PCA methods this can safely be left null to get scores x loadings but if set, then the scores are obtained by projecting provided data onto the loadings. If data contains missing values the result will be all NA. Non-linear PCA is an exception, here if data is NULL then data is set to the completeObs and propaged through the network.} \item{nPcs}{The number of PC's to consider} \item{pre}{pre-process \code{data} based on the pre-processing chosen for the PCA model} \item{post}{unpre-process the final data (add the center back etc to get the final estimate)} \item{...}{Not used} } \value{ A matrix representing the fitted data } \description{ Fitted values of a PCA model } \details{ This function extracts the fitted values from a pcaResobject. For PCA methods like SVD, Nipals, PPCA etc this is basically just the scores multipled by the loadings and adjusted for pre-processing. for non-linear PCA the original data is propagated through the network to obtain the approximated data. } \examples{ pc <- pca(iris[,1:4], nPcs=4, center=TRUE, scale="uv") sum( (fitted(pc) - iris[,1:4])^2 ) } \author{ Henning Redestig } \keyword{multivariate} pcaMethods/man/forkNlpcaNet.Rd0000644000175200017520000000054414710217306017317 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/forkNlpcaNet.R \name{forkNlpcaNet} \alias{forkNlpcaNet} \title{Complete copy of nlpca net object} \usage{ forkNlpcaNet(nlnet) } \arguments{ \item{nlnet}{a nlnet} } \value{ A copy of the input nlnet } \description{ Complete copy of nlpca net object } \author{ Henning Redestig } pcaMethods/man/getHierarchicIdx.Rd0000644000175200017520000000052514710217306020136 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/nlpca.R \name{getHierarchicIdx} \alias{getHierarchicIdx} \title{Index in hiearchy} \usage{ getHierarchicIdx(hierarchicNum) } \arguments{ \item{hierarchicNum}{A number} } \value{ ... } \description{ Index in hiearchy } \author{ Henning Redestig, Matthias Scholz } pcaMethods/man/helix.Rd0000644000175200017520000000110214710217306016031 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/pcaMethods-package.R \docType{data} \name{helix} \alias{helix} \title{A helix structured toy data set} \usage{ data(helix) } \description{ Simulated data set looking like a helix } \details{ A matrix containing 1000 observations (rows) and three variables (columns). } \references{ Matthias Scholz, Fatma Kaplan, Charles L. Guy, Joachim Kopka and Joachim Selbig. - Non-linear PCA: a missing data approach. \emph{Bioinformatics 2005 21(20):3887-3895} } \author{ Henning Redestig } \keyword{datasets} pcaMethods/man/kEstimate.Rd0000644000175200017520000001435514710217306016664 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/kEstimate.R \name{kEstimate} \alias{kEstimate} \title{Estimate best number of Components for missing value estimation} \usage{ kEstimate(Matrix, method = "ppca", evalPcs = 1:3, segs = 3, nruncv = 5, em = "q2", allVariables = FALSE, verbose = interactive(), ...) } \arguments{ \item{Matrix}{\code{matrix} -- numeric matrix containing observations in rows and variables in columns} \item{method}{\code{character} -- of the methods found with pcaMethods() The option llsImputeAll calls llsImpute with the allVariables = TRUE parameter.} \item{evalPcs}{\code{numeric} -- The principal components to use for cross validation or the number of neighbour variables if used with llsImpute. Should be an array containing integer values, eg. \code{evalPcs = 1:10} or \code{evalPcs = c(2,5,8)}. The NRMSEP or Q2 is calculated for each component.} \item{segs}{\code{numeric} -- number of segments for cross validation} \item{nruncv}{\code{numeric} -- Times the whole cross validation is repeated} \item{em}{\code{character} -- The error measure. This can be nrmsep or q2} \item{allVariables}{\code{boolean} -- If TRUE, the NRMSEP is calculated for all variables, If FALSE, only the incomplete ones are included. You maybe want to do this to compare several methods on a complete data set.} \item{verbose}{\code{boolean} -- If TRUE, some output like the variable indexes are printed to the console each iteration.} \item{...}{Further arguments to \code{pca} or \code{nni}} } \value{ A list with: \item{bestNPcs}{number of PCs or k for which the minimal average NRMSEP or the maximal Q2 was obtained.} \item{eError}{an array of of size length(evalPcs). Contains the average error of the cross validation runs for each number of components.} \item{variableWiseError}{Matrix of size \code{incomplete_variables} x length(evalPcs). Contains the NRMSEP or Q2 distance for each variable and each number of PCs. This allows to easily see for wich variables imputation makes sense and for which one it should not be done or mean imputation should be used.} \item{evalPcs}{The evaluated numbers of components or number of neighbours (the same as the evalPcs input parameter).} \item{variableIx}{Index of the incomplete variables. This can be used to map the variable wise error to the original data.} } \description{ Perform cross validation to estimate the optimal number of components for missing value estimation. Cross validation is done for the complete subset of a variable. } \details{ The assumption hereby is that variables that are highly correlated in a distinct region (here the non-missing observations) are also correlated in another (here the missing observations). This also implies that the complete subset must be large enough to be representative. For each incomplete variable, the available values are divided into a user defined number of cv-segments. The segments have equal size, but are chosen from a random equal distribution. The non-missing values of the variable are covered completely. PPCA, BPCA, SVDimpute, Nipals PCA, llsImpute an NLPCA may be used for imputation. The whole cross validation is repeated several times so, depending on the parameters, the calculations can take very long time. As error measure the NRMSEP (see Feten et. al, 2005) or the Q2 distance is used. The NRMSEP basically normalises the RMSD between original data and estimate by the variable-wise variance. The reason for this is that a higher variance will generally lead to a higher estimation error. If the number of samples is small, the variable - wise variance may become an unstable criterion and the Q2 distance should be used instead. Also if variance normalisation was applied previously. The method proceeds variable - wise, the NRMSEP / Q2 distance is calculated for each incomplete variable and averaged afterwards. This allows to easily see for wich set of variables missing value imputation makes senes and for wich set no imputation or something like mean-imputation should be used. Use \code{kEstimateFast} or \code{Q2} if you are not interested in variable wise CV performance estimates. Run time may be very high on large data sets. Especially when used with complex methods like BPCA or Nipals PCA. For PPCA, BPCA, Nipals PCA and NLPCA the estimation method is called \eqn{(v_{miss} \cdot segs \cdot nruncv \cdot)}{(v\_miss * segs * nruncv)} times as the error for all numbers of principal components can be calculated at once. For LLSimpute and SVDimpute this is not possible, and the method is called \eqn{(v_{miss} \cdot segs \cdot nruncv \cdot length(evalPcs))}{(v\_miss * segs * nruncv * length(evalPcs))} times. This should still be fast for LLSimpute because the method allows to choose to only do the estimation for one particular variable. This saves a lot of iterations. Here, \eqn{v_{miss}}{v\_miss} is the number of variables showing missing values. As cross validation is done variable-wise, in this function Q2 is defined on single variables, not on the entire data set. This is Q2 is calculated as as \eqn{\frac{\sum(x - xe)^2}{\sum(x^2)}}{sum(x - xe)^2 \ sum(x^2)}, where x is the currently used variable and xe it's estimate. The values are then averaged over all variables. The NRMSEP is already defined variable-wise. For a single variable it is then \eqn{\sqrt(\frac{\sum(x - xe)^2}{(n \cdot var(x))})}{sqrt(sum(x - xe)^2 \ (n * var(x)))}, where x is the variable and xe it's estimate, n is the length of x. The variable wise estimation errors are returned in parameter variableWiseError. } \examples{ ## Load a sample metabolite dataset with 5\\\% missing values (metaboliteData) data(metaboliteData) # Do cross validation with ppca for component 2:4 esti <- kEstimate(metaboliteData, method = "ppca", evalPcs = 2:4, nruncv=1, em="nrmsep") # Plot the average NRMSEP barplot(drop(esti$eError), xlab = "Components",ylab = "NRMSEP (1 iterations)") # The best result was obtained for this number of PCs: print(esti$bestNPcs) # Now have a look at the variable wise estimation error barplot(drop(esti$variableWiseError[, which(esti$evalPcs == esti$bestNPcs)]), xlab = "Incomplete variable Index", ylab = "NRMSEP") } \seealso{ \code{\link{kEstimateFast}, \link{Q2}, \link{pca}, \link{nni}}. } \author{ Wolfram Stacklies } \keyword{multivariate} pcaMethods/man/kEstimateFast.Rd0000644000175200017520000000553514710217306017502 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/kEstimateFast.R \name{kEstimateFast} \alias{kEstimateFast} \title{Estimate best number of Components for missing value estimation} \usage{ kEstimateFast(Matrix, method = "ppca", evalPcs = 1:3, em = "nrmsep", allVariables = FALSE, verbose = interactive(), ...) } \arguments{ \item{Matrix}{\code{matrix} -- numeric matrix containing observations in rows and variables in columns} \item{method}{\code{character} -- a valid pca method (see \code{\link{pca}}).} \item{evalPcs}{\code{numeric} -- The principal components to use for cross validation or cluster sizes if used with llsImpute. Should be an array containing integer values, eg. evalPcs = 1:10 or evalPcs = C(2,5,8).The NRMSEP is calculated for each component.} \item{em}{\code{character} -- The error measure. This can be nrmsep or q2} \item{allVariables}{\code{boolean} -- If TRUE, the NRMSEP is calculated for all variables, If FALSE, only the incomplete ones are included. You maybe want to do this to compare several methods on a complete data set.} \item{verbose}{\code{boolean} -- If TRUE, the NRMSEP and the variance are printed to the console each iteration.} \item{...}{Further arguments to \code{pca}} } \value{ \item{list}{Returns a list with the elements: \itemize{ \item minNPcs - number of PCs for which the minimal average NRMSEP was obtained \item eError - an array of of size length(evalPcs). Contains the estimation error for each number of components. \item evalPcs - The evaluated numbers of components or cluster sizes (the same as the evalPcs input parameter). }} } \description{ This is a simple estimator for the optimal number of componets when applying PCA or LLSimpute for missing value estimation. No cross validation is performed, instead the estimation quality is defined as Matrix[!missing] - Estimate[!missing]. This will give a relatively rough estimate, but the number of iterations equals the length of the parameter evalPcs.\cr Does not work with LLSimpute!! As error measure the NRMSEP (see Feten et. al, 2005) or the Q2 distance is used. The NRMSEP basically normalises the RMSD between original data and estimate by the variable-wise variance. The reason for this is that a higher variance will generally lead to a higher estimation error. If the number of samples is small, the gene - wise variance may become an unstable criterion and the Q2 distance should be used instead. Also if variance normalisation was applied previously. } \examples{ data(metaboliteData) # Estimate best number of PCs with ppca for component 2:4 esti <- kEstimateFast(t(metaboliteData), method = "ppca", evalPcs = 2:4, em="nrmsep") barplot(drop(esti$eError), xlab = "Components",ylab = "NRMSEP (1 iterations)") # The best k value is: print(esti$minNPcs) } \seealso{ \code{\link{kEstimate}}. } \author{ Wolfram Stacklies } \keyword{multivariate} pcaMethods/man/leverage-pcaRes-method.Rd0000644000175200017520000000213414710217306021211 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{leverage,pcaRes-method} \alias{leverage,pcaRes-method} \alias{leverage} \title{Extract leverages of a PCA model} \usage{ \S4method{leverage}{pcaRes}(object) } \arguments{ \item{object}{a \code{pcaRes} object} } \value{ The observation leverages as a numeric vector } \description{ The leverages of PCA model indicate how much influence each observation has on the PCA model. Observations with high leverage has caused the principal components to rotate towards them. It can be used to extract both "unimportant" observations as well as picking potential outliers. } \details{ Defined as \eqn{Tr(T(T'T)^{-1}T')}{Tr(T(T'T)^(-1)T')} } \examples{ data(iris) pcIr <- pca(iris[,1:4]) ## versicolor has the lowest leverage with(iris, plot(leverage(pcIr)~Species)) } \references{ Introduction to Multi- and Megavariate Data Analysis using Projection Methods (PCA and PLS), L. Eriksson, E. Johansson, N. Kettaneh-Wold and S. Wold, Umetrics 1999, p. 466 } \author{ Henning Redestig } \keyword{multivariate} pcaMethods/man/lineSearch.Rd0000644000175200017520000000104014710217306016776 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/lineSearch.R \name{lineSearch} \alias{lineSearch} \title{Line search for conjugate gradient} \usage{ lineSearch(nlnet, dw, e0, ttGuess, trainIn, trainOut, verbose) } \arguments{ \item{nlnet}{The nlnet} \item{dw}{..} \item{e0}{..} \item{ttGuess}{..} \item{trainIn}{Training data} \item{trainOut}{Fitted data} \item{verbose}{logical, print messages} } \value{ ... } \description{ Line search for conjugate gradient } \author{ Henning Redestig, Matthias Scholz } pcaMethods/man/linr.Rd0000644000175200017520000000043314710217306015672 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/lineSearch.R \name{linr} \alias{linr} \title{Linear kernel} \usage{ linr(x) } \arguments{ \item{x}{datum} } \value{ Input value } \description{ Linear kernel } \author{ Henning Redestig, Matthias Scholz } pcaMethods/man/listPcaMethods.Rd0000644000175200017520000000105714710217306017654 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/pca.R \name{listPcaMethods} \alias{listPcaMethods} \title{List PCA methods} \usage{ listPcaMethods(which = c("all", "linear", "nonlinear")) } \arguments{ \item{which}{the type of methods to get. E.g. only get the PCA methods based on the classical model where the fitted data is a direct multiplication of scores and loadings.} } \value{ A character vector with the current methods for doing PCA } \description{ Vector with current valid PCA methods } \author{ Henning Redestig } pcaMethods/man/llsImpute.Rd0000644000175200017520000000776614710217306016724 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/llsImpute.R \name{llsImpute} \alias{llsImpute} \title{LLSimpute algorithm} \usage{ llsImpute(Matrix, k = 10, center = FALSE, completeObs = TRUE, correlation = "pearson", allVariables = FALSE, maxSteps = 100, xval = NULL, verbose = FALSE, ...) } \arguments{ \item{Matrix}{\code{matrix} -- Data containing the variables (genes) in columns and observations (samples) in rows. The data may contain missing values, denoted as \code{NA}.} \item{k}{\code{numeric} -- Cluster size, this is the number of similar genes used for regression.} \item{center}{\code{boolean} -- Mean center the data if TRUE} \item{completeObs}{\code{boolean} -- Return the estimated complete observations if TRUE. This is the input data with NA values replaced by the estimated values.} \item{correlation}{\code{character} -- How to calculate the distance between genes. One out of pearson | kendall | spearman , see also help("cor").} \item{allVariables}{\code{boolean} -- Use only complete genes to do the regression if TRUE, all genes if FALSE.} \item{maxSteps}{\code{numeric} -- Maximum number of iteration steps if allGenes = TRUE.} \item{xval}{\code{numeric} Use LLSimpute for cross validation. xval is the index of the gene to estimate, all other incomplete genes will be ignored if this parameter is set. We do not consider them in the cross-validation.} \item{verbose}{\code{boolean} -- Print step number and relative change if TRUE and allVariables = TRUE} \item{...}{Reserved for parameters used in future version of the algorithm} } \value{ \item{nniRes}{Standard nni (nearest neighbour imputation) result object of this package. See \code{\link{nniRes}} for details.} } \description{ Missing value estimation using local least squares (LLS). First, k variables (for Microarrya data usually the genes) are selected by pearson, spearman or kendall correlation coefficients. Then missing values are imputed by a linear combination of the k selected variables. The optimal combination is found by LLS regression. The method was first described by Kim et al, Bioinformatics, 21(2),2005. } \details{ Missing values are denoted as \code{NA}\cr It is not recommended to use this function directely but rather to use the nni() wrapper function. The methods provides two ways for missing value estimation, selected by the \code{allVariables} option. The first one is to use only complete variables for the regression. This is preferable when the number of incomplete variables is relatively small. The second way is to consider all variables as candidates for the regression. Hereby missing values are initially replaced by the columns wise mean. The method then iterates, using the current estimate as input for the regression until the change between new and old estimate falls below a threshold (0.001). } \note{ Each step the generalized inverse of a \code{miss} x k matrix is calculated. Where \code{miss} is the number of missing values in variable j and \code{k} the number of neighbours. This may be slow for large values of k and / or many missing values. See also help("ginv"). } \examples{ ## Load a sample metabolite dataset (metaboliteData) with already 5\\\% of ## data missing data(metaboliteData) ## Perform llsImpute using k = 10 ## Set allVariables TRUE because there are very few complete variables result <- llsImpute(metaboliteData, k = 10, correlation="pearson", allVariables=TRUE) ## Get the estimated complete observations cObs <- completeObs(result) } \references{ Kim, H. and Golub, G.H. and Park, H. - Missing value estimation for DNA microarray gene expression data: local least squares imputation. \emph{Bioinformatics, 2005; 21(2):187-198.} Troyanskaya O. and Cantor M. and Sherlock G. and Brown P. and Hastie T. and Tibshirani R. and Botstein D. and Altman RB. - Missing value estimation methods for DNA microarrays. \emph{Bioinformatics. 2001 Jun;17(6):520-525.} } \seealso{ \code{\link{pca}, \link{nniRes}, \link{nni}}. } \author{ Wolfram Stacklies } \keyword{multivariate} pcaMethods/man/loadings-ANY-method.Rd0000644000175200017520000000077614710217306020443 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{loadings,ANY-method} \alias{loadings,ANY-method} \alias{loadings} \title{Crude way to unmask the function with the same name from \code{stats}} \usage{ \S4method{loadings}{ANY}(object, ...) } \arguments{ \item{object}{any object} \item{...}{not used} } \value{ The loadings } \description{ Crude way to unmask the function with the same name from \code{stats} } \author{ Henning Redestig } pcaMethods/man/loadings-pcaRes-method.Rd0000644000175200017520000000075114710217306021222 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{loadings,pcaRes-method} \alias{loadings,pcaRes-method} \title{Get loadings from a pcaRes object} \usage{ \S4method{loadings}{pcaRes}(object, ...) } \arguments{ \item{object}{a pcaRes object} \item{...}{not used} } \value{ The loadings as a matrix } \description{ Get loadings from a pcaRes object } \seealso{ \code{\link{loadings.pcaRes}} } \author{ Henning Redestig } pcaMethods/man/loadings.pcaRes.Rd0000644000175200017520000000063514710217306017746 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \name{loadings.pcaRes} \alias{loadings.pcaRes} \title{Get loadings from a pcaRes object} \usage{ \method{loadings}{pcaRes}(object, ...) } \arguments{ \item{object}{a pcaRes object} \item{...}{not used} } \value{ The loadings as a matrix } \description{ Get loadings from a pcaRes object } \author{ Henning Redestig } pcaMethods/man/metaboliteData.Rd0000644000175200017520000000174514710217306017654 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/pcaMethods-package.R \docType{data} \name{metaboliteData} \alias{metaboliteData} \title{A incomplete metabolite data set from an Arabidopsis coldstress experiment} \description{ A incomplete subset from a larger metabolite data set. This is the original, complete data set and can be used to compare estimation results created with the also provided incomplete data (called metaboliteData). } \details{ A matrix containing 154 observations (rows) and 52 metabolites (columns). The data contains 5\% of artificially created uniformly distributed misssing values. The data was created during an in house Arabidopsis coldstress experiment. } \references{ Matthias Scholz, Fatma Kaplan, Charles L. Guy, Joachim Kopka and Joachim Selbig. - Non-linear PCA: a missing data approach.\emph{Bioinformatics 2005 21(20):3887-3895} } \seealso{ \code{\link{metaboliteDataComplete}} } \author{ Wolfram Stacklies } \keyword{datasets} pcaMethods/man/metaboliteDataComplete.Rd0000644000175200017520000000162514710217306021342 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/pcaMethods-package.R \docType{data} \name{metaboliteDataComplete} \alias{metaboliteDataComplete} \title{A complete metabolite data set from an Arabidopsis coldstress experiment} \description{ A complete subset from a larger metabolite data set. This is the original, complete data set and can be used to compare estimation results created with the also provided incomplete data (called metaboliteData). The data was created during an in house Arabidopsis coldstress experiment. } \details{ A matrix containing 154 observations (rows) and 52 metabolites (columns). } \references{ Matthias Scholz, Fatma Kaplan, Charles L. Guy, Joachim Kopka and Joachim Selbig. - Non-linear PCA: a missing data approach.\emph{Bioinformatics 2005 21(20):3887-3895} } \seealso{ \code{\link{metaboliteData}} } \author{ Wolfram Stacklies } \keyword{datasets} pcaMethods/man/method-pcaRes-method.Rd0000644000175200017520000000063214710217306020700 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{method,pcaRes-method} \alias{method,pcaRes-method} \alias{method} \title{Get the used PCA method} \usage{ method(object, ...) } \arguments{ \item{object}{pcaRes object} \item{...}{Not used} } \value{ The used pca method } \description{ Get the used PCA method } \author{ Henning Redestig } pcaMethods/man/nObs-pcaRes-method.Rd0000644000175200017520000000073514710217306020325 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{nObs,pcaRes-method} \alias{nObs,pcaRes-method} \alias{nObs} \title{Get the number of observations used to build the PCA model.} \usage{ nObs(object, ...) } \arguments{ \item{object}{pcaRes object} \item{...}{Not used} } \value{ Number of observations } \description{ Get the number of observations used to build the PCA model. } \author{ Henning Redestig } pcaMethods/man/nP-pcaRes-method.Rd0000644000175200017520000000057014710217306017776 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{nP,pcaRes-method} \alias{nP,pcaRes-method} \alias{nP} \title{Get number of PCs} \usage{ nP(object, ...) } \arguments{ \item{object}{pcaRes object} \item{...}{not used} } \value{ Number of PCs } \description{ Get number of PCs } \author{ Henning Redestig } pcaMethods/man/nPcs-pcaRes-method.Rd0000644000175200017520000000074314710217306020326 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{nPcs,pcaRes-method} \alias{nPcs,pcaRes-method} \alias{nPcs} \title{Get number of PCs.} \usage{ nPcs(object, ...) } \arguments{ \item{object}{pcaRes object} \item{...}{not used} } \value{ Number of PCs } \description{ Get number of PCs. } \note{ Try to use \code{link{nP}} instead since \code{nPcs} tend to clash with argument names. } \author{ Henning Redestig } pcaMethods/man/nVar-pcaRes-method.Rd0000644000175200017520000000072414710217306020330 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{nVar,pcaRes-method} \alias{nVar,pcaRes-method} \alias{nVar} \title{Get the number of variables used to build the PCA model.} \usage{ nVar(object, ...) } \arguments{ \item{object}{pcaRes object} \item{...}{Not used} } \value{ Number of variables } \description{ Get the number of variables used to build the PCA model. } \author{ Henning Redestig } pcaMethods/man/nipalsPca.Rd0000644000175200017520000000361414710217306016644 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/nipalsPca.R \name{nipalsPca} \alias{nipalsPca} \title{NIPALS PCA} \usage{ nipalsPca(Matrix, nPcs = 2, varLimit = 1, maxSteps = 5000, threshold = 1e-06, ...) } \arguments{ \item{Matrix}{Pre-processed (centered, scaled) numerical matrix samples in rows and variables as columns.} \item{nPcs}{Number of components that should be extracted.} \item{varLimit}{Optionally the ratio of variance that should be explained. \code{nPcs} is ignored if varLimit < 1} \item{maxSteps}{Defines how many iterations can be done before algorithm should abort (happens almost exclusively when there were some wrong in the input data).} \item{threshold}{The limit condition for judging if the algorithm has converged or not, specifically if a new iteration is done if \eqn{(T_{old} - T)^T(T_{old} - T) > \code{limit}}.} \item{...}{Only used for passing through arguments.} } \value{ A \code{pcaRes} object. } \description{ PCA by non-linear iterative partial least squares } \details{ Can be used for computing PCA on a numeric matrix using either the NIPALS algorithm which is an iterative approach for estimating the principal components extracting them one at a time. NIPALS can handle a small amount of missing values. It is not recommended to use this function directely but rather to use the pca() wrapper function. } \examples{ data(metaboliteData) mat <- prep(t(metaboliteData)) pc <- nipalsPca(mat, nPcs=2) ## better use pca() pc <- pca(t(metaboliteData), method="nipals", nPcs=2) \dontshow{stopifnot(sum((fitted(pc) - t(metaboliteData))^2, na.rm=TRUE) < 200)} } \references{ Wold, H. (1966) Estimation of principal components and related models by iterative least squares. In Multivariate Analysis (Ed., P.R. Krishnaiah), Academic Press, NY, 391-420. } \seealso{ \code{prcomp}, \code{princomp}, \code{pca} } \author{ Henning Redestig } \keyword{multivariate} pcaMethods/man/nlpca.Rd0000644000175200017520000000674114710217306016033 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/nlpca.R \name{nlpca} \alias{nlpca} \title{Non-linear PCA} \usage{ nlpca(Matrix, nPcs = 2, maxSteps = 2 * prod(dim(Matrix)), unitsPerLayer = NULL, functionsPerLayer = NULL, weightDecay = 0.001, weights = NULL, verbose = interactive(), ...) } \arguments{ \item{Matrix}{\code{matrix} --- Preprocessed data with the variables in columns and observations in rows. The data may contain missing values, denoted as \code{NA}} \item{nPcs}{\code{numeric} -- Number of components to estimate. The preciseness of the missing value estimation depends on thenumber of components, which should resemble the internal structure of the data.} \item{maxSteps}{\code{numeric} -- Number of estimation steps. Default is based on a generous rule of thumb.} \item{unitsPerLayer}{The network units, example: c(2,4,6) for two input units 2feature units (principal components), one hidden layer fornon-linearity and three output units (original amount ofvariables).} \item{functionsPerLayer}{The function to apply at each layer eg. c("linr", "tanh", "linr")} \item{weightDecay}{Value between 0 and 1.} \item{weights}{Starting weights for the network. Defaults to uniform random values but can be set specifically to make algorithm deterministic.} \item{verbose}{\code{boolean} -- nlpca prints the number of steps and warning messages if set to TRUE. Default is interactive().} \item{...}{Reserved for future use. Not passed on anywhere.} } \value{ Standard PCA result object used by all PCA-basedmethods of this package. Contains scores, loadings, data meanand more. See \code{\link{pcaRes}} for details. } \description{ Neural network based non-linear PCA } \details{ Artificial Neural Network (MLP) for performing non-linear PCA. Non-linear PCA is conceptually similar to classical PCA but theoretically quite different. Instead of simply decomposing our matrix (X) to scores (T) loadings (P) and an error (E) we train a neural network (our loadings) to find a curve through the multidimensional space of X that describes a much variance as possible. Classical ways of interpreting PCA results are thus not applicable to NLPCA since the loadings are hidden in the network. However, the scores of components that lead to low cross-validation errors can still be interpreted via the score plot. Unfortunately this method depend on slow iterations which currently are implemented in R only making this method extremely slow. Furthermore, the algorithm does not by itself decide when it has converged but simply does 'maxSteps' iterations. } \examples{ ## Data set with three variables where data points constitute a helix data(helix) helixNA <- helix ## not a single complete observation helixNA <- t(apply(helix, 1, function(x) { x[sample(1:3, 1)] <- NA; x})) ## 50 steps is not enough, for good estimation use 1000 helixNlPca <- pca(helixNA, nPcs=1, method="nlpca", maxSteps=50) fittedData <- fitted(helixNlPca, helixNA) plot(fittedData[which(is.na(helixNA))], helix[which(is.na(helixNA))]) ## compared to solution by Nipals PCA which cannot extract non-linear patterns helixNipPca <- pca(helixNA, nPcs=2) fittedData <- fitted(helixNipPca) plot(fittedData[which(is.na(helixNA))], helix[which(is.na(helixNA))]) } \references{ Matthias Scholz, Fatma Kaplan, Charles L Guy, Joachim Kopkaand Joachim Selbig. Non-linear PCA: a missing data approach. \emph{Bioinformatics, 21(20):3887-3895, Oct 2005} } \author{ Based on a matlab script by Matthias Scholz and ported to R by Henning Redestig } pcaMethods/man/nmissing-pcaRes-method.Rd0000644000175200017520000000067414710217306021255 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{nmissing,pcaRes-method} \alias{nmissing,pcaRes-method} \alias{nmissing} \alias{nmissing,nniRes-method} \title{Missing values} \usage{ nmissing(object, ...) } \arguments{ \item{object}{pcaRes object} \item{...}{Not used} } \value{ Get the number of missing values } \description{ Missing values } \author{ Henning Redestig } pcaMethods/man/nni.Rd0000644000175200017520000000237314710217306015517 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/pca.R \name{nni} \alias{nni} \title{Nearest neighbour imputation} \usage{ nni(object, method = c("llsImpute"), subset = numeric(), ...) } \arguments{ \item{object}{Numerical matrix with (or an object coercible to such) with samples in rows and variables as columns. Also takes \code{ExpressionSet} in which case the transposed expression matrix is used.} \item{method}{For convenience one can pass a large matrix but only use the variable specified as subset. Can be colnames or indices.} \item{subset}{Currently "llsImpute" only.} \item{...}{Further arguments to the chosen method.} } \value{ A \code{clusterRes} object. Or a list containing a clusterRes object as first and an ExpressionSet object as second entry if the input was of type ExpressionSet. } \description{ Wrapper function for imputation methods based on nearest neighbour clustering. Currently llsImpute only. } \details{ This method is wrapper function to llsImpute, See documentation for \code{link{llsImpute}}. } \examples{ data(metaboliteData) llsRes <- nni(metaboliteData, k=6, method="llsImpute", allGenes=TRUE) } \seealso{ \code{\link{llsImpute}}, \code{\link{pca}} } \author{ Wolfram Stacklies } \keyword{multivariate} pcaMethods/man/nniRes.Rd0000644000175200017520000000252714710217306016172 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/AllClasses.R \docType{class} \name{nniRes} \alias{nniRes} \alias{nniRes-class} \title{Class for representing a nearest neighbour imputation result} \description{ This is a class representation of nearest neighbour imputation (nni) result } \details{ \bold{Creating Objects}\cr \code{new("nniRes", completeObs=[the estimated complete observations], k=[cluster size], nObs=[amount of observations], nVar=[amount of variables], centered=[was the data centered befor running LLSimpute], center=[original means], method=[method used to perform clustering], missing=[amount of NAs])} \bold{Slots}\cr \describe{ \item{completeObs}{"matrix", the estimated complete observations} \item{nObs}{"numeric", amount of observations} \item{nVar}{"numeric", amount of variables} \item{correlation}{"character", the correlation method used (pearson, kendall or spearman)} \item{centered}{"logical", data was centered or not} \item{center}{"numeric", the original variable centers} \item{k}{"numeric", cluster size} \item{method}{"character", the method used to perform the clustering} \item{missing}{"numeric", the total amount of missing values in original data} } \bold{Methods}\cr \describe{ \item{print}{Print function} } } \author{ Wolfram Stacklies } \keyword{classes} pcaMethods/man/optiAlgCgd.Rd0000644000175200017520000000073714710217306016752 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/optiAlgCgd.R \name{optiAlgCgd} \alias{optiAlgCgd} \title{Conjugate gradient optimization} \usage{ optiAlgCgd(nlnet, trainIn, trainOut, verbose = FALSE) } \arguments{ \item{nlnet}{The nlnet} \item{trainIn}{Training data} \item{trainOut}{fitted data} \item{verbose}{logical, print messages} } \value{ ... } \description{ Conjugate gradient optimization } \author{ Henning Redestig, Matthias Scholz } pcaMethods/man/orth.Rd0000644000175200017520000000121214710217306015676 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/orth.R \name{orth} \alias{orth} \title{Calculate an orthonormal basis} \usage{ orth(mat, skipInac = FALSE) } \arguments{ \item{mat}{matrix to calculate orthonormal base} \item{skipInac}{do not include components with precision below .Machine$double.eps if TRUE} } \value{ orthonormal basis for the range of matrix } \description{ ONB = orth(mat) is an orthonormal basis for the range of matrix mat. That is, ONB' * ONB = I, the columns of ONB span the same space as the columns of mat, and the number of columns of ONB is the rank of mat. } \author{ Wolfram Stacklies } pcaMethods/man/pca.Rd0000644000175200017520000000752314710217306015500 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/pca.R \name{pca} \alias{pca} \title{Perform principal component analysis} \usage{ pca(object, method, nPcs = 2, scale = c("none", "pareto", "vector", "uv"), center = TRUE, completeObs = TRUE, subset = NULL, cv = c("none", "q2"), ...) } \arguments{ \item{object}{Numerical matrix with (or an object coercible to such) with samples in rows and variables as columns. Also takes \code{ExpressionSet} in which case the transposed expression matrix is used. Can also be a data frame in which case all numberic variables are used to fit the PCA.} \item{method}{One of the methods reported by \code{listPcaMethods()}. Can be left missing in which case the \code{svd} PCA is chosen for data wihout missing values and \code{nipalsPca} for data with missing values} \item{nPcs}{Number of principal components to calculate.} \item{scale}{Scaling, see \code{\link{prep}}.} \item{center}{Centering, see \code{\link{prep}}.} \item{completeObs}{Sets the \code{completeObs} slot on the resulting \code{pcaRes} object containing the original data with but with all NAs replaced with the estimates.} \item{subset}{A subset of variables to use for calculating the model. Can be column names or indices.} \item{cv}{character naming a the type of cross-validation to be performed.} \item{...}{Arguments to \code{\link{prep}}, the chosen pca method and \code{\link{Q2}}.} } \value{ A \code{pcaRes} object. } \description{ Perform PCA on a numeric matrix for visualisation, information extraction and missing value imputation. } \details{ This method is wrapper function for the following set of pca methods: \describe{\item{svd:}{Uses classical \code{prcomp}. See documentation for \code{\link{svdPca}}.} \item{nipals:}{An iterative method capable of handling small amounts of missing values. See documentation for \code{\link{nipalsPca}}.} \item{rnipals:}{Same as nipals but implemented in R.} \item{bpca:}{An iterative method using a Bayesian model to handle missing values. See documentation for \code{\link{bpca}}.} \item{ppca:}{An iterative method using a probabilistic model to handle missing values. See documentation for \code{\link{ppca}}.} \item{svdImpute:}{Uses expectation maximation to perform SVD PCA on incomplete data. See documentation for \code{\link{svdImpute}}.}} Scaling and centering is part of the PCA model and handled by \code{\link{prep}}. } \examples{ data(iris) ## Usually some kind of scaling is appropriate pcIr <- pca(iris, method="svd", nPcs=2) pcIr <- pca(iris, method="nipals", nPcs=3, cv="q2") ## Get a short summary on the calculated model summary(pcIr) plot(pcIr) ## Scores and loadings plot slplot(pcIr, sl=as.character(iris[,5])) ## use an expressionset and ggplot data(sample.ExpressionSet) pc <- pca(sample.ExpressionSet) df <- merge(scores(pc), pData(sample.ExpressionSet), by=0) library(ggplot2) ggplot(df, aes(PC1, PC2, shape=sex, color=type)) + geom_point() + xlab(paste("PC1", pc@R2[1] * 100, "\% of the variance")) + ylab(paste("PC2", pc@R2[2] * 100, "\% of the variance")) } \references{ Wold, H. (1966) Estimation of principal components and related models by iterative least squares. In Multivariate Analysis (Ed., P.R. Krishnaiah), Academic Press, NY, 391-420. Shigeyuki Oba, Masa-aki Sato, Ichiro Takemasa, Morito Monden, Ken-ichi Matsubara and Shin Ishii. A Bayesian missing value estimation method for gene expression profile data. \emph{Bioinformatics, 19(16):2088-2096, Nov 2003}. Troyanskaya O. and Cantor M. and Sherlock G. and Brown P. and Hastie T. and Tibshirani R. and Botstein D. and Altman RB. - Missing value estimation methods for DNA microarrays. \emph{Bioinformatics. 2001 Jun;17(6):520-5}. } \seealso{ \code{\link{prcomp}}, \code{\link{princomp}}, \code{\link{nipalsPca}}, \code{\link{svdPca}} } \author{ Wolfram Stacklies, Henning Redestig } \keyword{multivariate} pcaMethods/man/pcaMethods-deprecated.Rd0000644000175200017520000000105414710217306021113 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/pcaMethods-package.R \name{pcaMethods-deprecated} \alias{pcaMethods-deprecated} \title{Deprecated methods for pcaMethods} \description{ \describe{ \item{plotR2}{Lack of relevance for this plot and the fact that it can not show cross-validation based diagnostics in the same plot makes it redundant with the introduction of a dedicated \code{plot} function for \code{pcaRes}. The new plot only shows R2cum but the result is pretty much the same.}} } \author{ Henning Redestig } pcaMethods/man/pcaMethods.Rd0000644000175200017520000000221714710217306017017 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/pcaMethods-package.R \docType{package} \name{pcaMethods} \alias{pcaMethods} \alias{pcaMethods-package} \title{pcaMethods} \description{ Principal Component Analysis in R } \details{ \tabular{ll}{ Package: \tab pcaMethods \cr Type: \tab Package \cr Developed since: \tab 2006 \cr License: \tab GPL (>=3) \cr LazyLoad: \tab yes \cr } Provides Bayesian PCA, Probabilistic PCA, Nipals PCA, Inverse Non-Linear PCA and the conventional SVD PCA. A cluster based method for missing value estimation is included for comparison. BPCA, PPCA and NipalsPCA may be used to perform PCA on incomplete data as well as for accurate missing value estimation. A set of methods for printing and plotting the results is also provided. All PCA methods make use of the same data structure (pcaRes) to provide a unique interface to the PCA results. Developed at the Max-Planck Institute for Molecular Plant Physiology, Golm, Germany, RIKEN Plant Science Center Yokohama, Japan, and CAS-MPG Partner Institute for Computational Biology (PICB) Shanghai, P.R. China } \author{ Wolfram Stacklies, Henning Redestig } pcaMethods/man/pcaNet.Rd0000644000175200017520000000657214710217306016152 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/AllClasses.R \docType{class} \name{pcaNet} \alias{pcaNet} \alias{nlpcaNet} \alias{nlpcaNet-class} \title{Class representation of the NLPCA neural net} \description{ This is a class representation of a non-linear PCA neural network. The \code{nlpcaNet} class is not meant for user-level usage. } \details{ Creating Objects \code{new("nlpcaNet", net=[the network structure], hierarchic=[hierarchic design], fct=[the functions at each layer], fkt=[the functions used for forward propagation], weightDecay=[incremental decrease of weight changes over iterations (between 0 and 1)], featureSorting=[sort features or not], dataDist=[represents the present values], inverse=[net is inverse mode or not], fCount=[amount of times features were sorted], componentLayer=[which layer is the 'bottleneck' (principal components)], erro=[the used error function], gradient=[the used gradient method], weights=[the present weights], maxIter=[the amount of iterations that was done], scalingFactor=[the scale of the original matrix])} Slots \describe{ \item{net}{"matrix", matrix showing the representation of the neural network, e.g. (2,4,6) for a network with two features, a hidden layer and six output neurons (original variables).} \item{hierarchic}{"list", the hierarchic design of the network, holds 'idx' (), 'var' () and layer (which layer is the principal component layer).} \item{fct}{"character", a vector naming the functions that will be applied on each layer. "linr" is linear (i.e.) standard matrix products and "tanh" means that the arcus tangens is applied on the result of the matrix product (for non-linearity).} \item{fkt}{"character", same as fct but the functions used during back propagation.} \item{weightDecay}{"numeric", the value that is used to incrementally decrease the weight changes to ensure convergence.} \item{featureSorting}{"logical", indicates if features will be sorted or not. This is used to make the NLPCA assume properties closer to those of standard PCA were the first component is more important for reconstructing the data than the second component.} \item{dataDist}{"matrix", a matrix of ones and zeroes indicating which values will add to the errror.} \item{inverse}{"logical", network is inverse mode (currently only inverse is supported) or not. Eg. the case when we have truly missing values and wish to impute them.} \item{fCount}{"integer", Counter for the amount of times features were really sorted.} \item{componentLayer}{"numeric", the index of 'net' that is the component layer.} \item{error}{"function", the used error function. Currently only one is provided \code{errorHierarchic}.} \item{gradient}{"function", the used gradient function. Currently only one is provided \code{derrorHierarchic}} \item{weights}{"list", A list holding managements of the weights. The list has two functions, weights$current() and weights$set() which access a matrix in the local environment of this object.} \item{maxIter}{"integer", the amount of iterations used to train this network.} \item{scalingFactor}{"numeric", training the network is best made with 'small' values so the original data is scaled down to a suitable range by division with this number.}} Methods \describe{ \item{vector2matrices}{Returns the weights in a matrix representation.} } } \seealso{ \code{\link{nlpca}} } \author{ Henning Redestig } \keyword{classes} pcaMethods/man/pcaRes.Rd0000644000175200017520000000577214710217306016156 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/AllClasses.R \docType{class} \name{pcaRes} \alias{pcaRes} \alias{pcaRes-class} \title{Class for representing a PCA result} \description{ This is a class representation of a PCA result } \details{ \bold{Creating Objects}\cr \code{new("pcaRes", scores=[the scores], loadings=[the loadings], nPcs=[amount of PCs], R2cum=[cumulative R2], nObs=[amount of observations], nVar=[amount of variables], R2=[R2 for each individual PC], sDev=[stdev for each individual PC], centered=[was data centered], center=[original means], varLimit=[what variance limit was exceeded], method=[method used to calculate PCA], missing=[amount of NAs], completeObs=[estimated complete observations])} \bold{Slots}\cr \describe{ \item{scores}{"matrix", the calculated scores} \item{loadings}{"matrix", the calculated loadings} \item{R2cum}{"numeric", the cumulative R2 values} \item{sDev}{"numeric", the individual standard deviations of the score vectors} \item{R2}{"numeric", the individual R2 values} \item{cvstat}{"numeric", cross-validation statistics} \item{nObs}{"numeric", number of observations} \item{nVar}{"numeric", number of variables} \item{centered}{"logical", data was centered or not} \item{center}{"numeric", the original variable centers} \item{scaled}{"logical", data was scaled or not} \item{scl}{"numeric", the original variable scales} \item{varLimit}{"numeric", the exceeded variance limit} \item{nPcs,nP}{"numeric", the number of calculated PCs} \item{method}{"character", the method used to perform PCA} \item{missing}{"numeric", the total amount of missing values in original data} \item{completeObs}{"matrix", the estimated complete observations} \item{network}{"nlpcaNet", the network used by non-linear PCA} } \bold{Methods (not necessarily exhaustive)}\cr \describe{ \item{print}{Print function} \item{summary}{Extract information about PC relevance} \item{screeplot}{Plot a barplot of standard deviations for PCs} \item{slplot}{Make a side by side score and loadings plot} \item{nPcs}{Get the number of PCs} \item{nObs}{Get the number of observations} \item{cvstat}{Cross-validation statistics} \item{nVar}{Get the number of variables} \item{loadings}{Get the loadings} \item{scores}{Get the scores} \item{dim}{Get the dimensions (number of observations, number of features)} \item{centered}{Get a logical indicating if centering was done as part of the model} \item{center}{Get the averages of the original variables.} \item{completeObs}{Get the imputed data set} \item{method}{Get a string naming the used PCA method} \item{sDev}{Get the standard deviations of the PCs} \item{scaled}{Get a logical indicating if scaling was done as part of the model} \item{scl}{Get the scales of the original variablesb} \item{R2cum}{Get the cumulative R2} } } \author{ Henning Redestig } \keyword{classes} pcaMethods/man/plot.pcaRes.Rd0000644000175200017520000000237414710217306017126 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \name{plot.pcaRes} \alias{plot.pcaRes} \alias{plot,pcaRes-method} \title{Plot diagnostics (screeplot)} \usage{ \method{plot}{pcaRes}(x, y = NULL, main = deparse(substitute(object)), col = gray(c(0.9, 0.5)), ...) } \arguments{ \item{x}{\code{pcaRes} The pcaRes object.} \item{y}{not used} \item{main}{title of the plot} \item{col}{Colors of the bars} \item{...}{further arguments to barplot} } \value{ None, used for side effect. } \description{ Plot the computed diagnostics of PCA model to get an idea of their importance. Note though that the standard screeplot shows the standard deviations for the PCs this method shows the R2 values which empirically shows the importance of the P's and is thus applicable for any PCA method rather than just SVD based PCA. } \details{ If cross-validation was done for the PCA the plot will also show the CV based statistics. A common rule-of-thumb for determining the optimal number of PCs is the PC where the CV diagnostic is at its maximum but not very far from \eqn{R^2}. } \examples{ data(metaboliteData) pc <- pca(t(metaboliteData), nPcs=5, cv="q2", scale="uv") plot(pc) } \seealso{ \link{screeplot} } \author{ Henning Redestig } pcaMethods/man/plotPcs.Rd0000644000175200017520000000246214710217306016356 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/pca.R \name{plotPcs} \alias{plotPcs} \title{Plot many side by side scores XOR loadings plots} \usage{ plotPcs(object, pcs = 1:nP(object), type = c("scores", "loadings"), sl = NULL, hotelling = 0.95, ...) } \arguments{ \item{object}{\code{pcaRes} a pcaRes object} \item{pcs}{\code{numeric} which pcs to plot} \item{type}{\code{character} Either "scores" or "loadings" for scores or loadings plot respectively} \item{sl}{\code{character} Text labels to plot instead of a point, if NULL points are plotted instead of text} \item{hotelling}{\code{numeric} Significance level for the confidence ellipse. NULL means that no ellipse is drawn.} \item{...}{Further arguments to \code{\link{pairs}} on which this function is based.} } \value{ None, used for side effect. } \description{ A function that can be used to visualise many PCs plotted against each other } \details{ Uses \code{\link{pairs}} to provide side-by-side plots. Note that this function only plots scores or loadings but not both in the same plot. } \examples{ data(iris) pcIr <- pca(iris[,1:4], nPcs=3, method="svd") plotPcs(pcIr, col=as.integer(iris[,4]) + 1) } \seealso{ \code{prcomp}, \code{pca}, \code{princomp}, \code{slplot} } \author{ Henning Redestig } \keyword{multivariate} pcaMethods/man/ppca.Rd0000644000175200017520000000727414710217306015663 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/ppca.R \name{ppca} \alias{ppca} \title{Probabilistic PCA} \usage{ ppca(Matrix, nPcs = 2, seed = NA, threshold = 1e-05, maxIterations = 1000, ...) } \arguments{ \item{Matrix}{\code{matrix} -- Data containing the variables in columns and observations in rows. The data may contain missing values, denoted as \code{NA}.} \item{nPcs}{\code{numeric} -- Number of components to estimate. The preciseness of the missing value estimation depends on the number of components, which should resemble the internal structure of the data.} \item{seed}{\code{numeric} Set the seed for the random number generator. PPCA creates fills the initial loading matrix with random numbers chosen from a normal distribution. Thus results may vary slightly. Set the seed for exact reproduction of your results.} \item{threshold}{Convergence threshold.} \item{maxIterations}{the maximum number of allowed iterations} \item{...}{Reserved for future use. Currently no further parameters are used.} } \value{ Standard PCA result object used by all PCA-based methods of this package. Contains scores, loadings, data mean and more. See \code{\link{pcaRes}} for details. } \description{ Implementation of probabilistic PCA (PPCA). PPCA allows to perform PCA on incomplete data and may be used for missing value estimation. This script was implemented after the Matlab version provided by Jakob Verbeek ( see \url{http://lear.inrialpes.fr/~verbeek/}) and the draft \emph{``EM Algorithms for PCA and Sensible PCA''} written by Sam Roweis. } \details{ Probabilistic PCA combines an EM approach for PCA with a probabilistic model. The EM approach is based on the assumption that the latent variables as well as the noise are normal distributed. In standard PCA data which is far from the training set but close to the principal subspace may have the same reconstruction error. PPCA defines a likelihood function such that the likelihood for data far from the training set is much lower, even if they are close to the principal subspace. This allows to improve the estimation accuracy. A method called \code{kEstimate} is provided to estimate the optimal number of components via cross validation. In general few components are sufficient for reasonable estimation accuracy. See also the package documentation for further discussion on what kind of data PCA-based missing value estimation is advisable. \bold{Complexity:}\cr Runtime is linear in the number of data, number of data dimensions and number of principal components. \bold{Convergence:} The threshold indicating convergence was changed from 1e-3 in 1.2.x to 1e-5 in the current version leading to more stable results. For reproducability you can set the seed (parameter seed) of the random number generator. If used for missing value estimation, results may be checked by simply running the algorithm several times with changing seed, if the estimated values show little variance the algorithm converged well. } \note{ Requires \code{MASS}. It is not recommended to use this function directely but rather to use the pca() wrapper function. } \examples{ ## Load a sample metabolite dataset with 5\\\% missing values (metaboliteData) data(metaboliteData) ## Perform probabilistic PCA using the 3 largest components result <- pca(t(metaboliteData), method="ppca", nPcs=3, seed=123) ## Get the estimated complete observations cObs <- completeObs(result) ## Plot the scores plotPcs(result, type = "scores") \dontshow{ stopifnot(sum((fitted(result) - t(metaboliteData))^2, na.rm=TRUE) < 200) } } \seealso{ \code{\link{bpca}, \link{svdImpute}, \link{prcomp}, \link{nipalsPca}, \link{pca}, \link{pcaRes}}. } \author{ Wolfram Stacklies } \keyword{multivariate} pcaMethods/man/predict-methods.Rd0000644000175200017520000000335314710217306020025 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{predict-methods} \alias{predict-methods} \alias{predict.pcaRes} \alias{predict,pcaRes-method} \title{Predict values from PCA.} \usage{ \method{predict}{pcaRes}(object, newdata, pcs = nP(object), pre = TRUE, post = TRUE, ...) \S4method{predict}{pcaRes}(object, newdata, pcs = nP(object), pre = TRUE, post = TRUE, ...) } \arguments{ \item{object}{\code{pcaRes} the \code{pcaRes} object of interest.} \item{newdata}{\code{matrix} new data with same number of columns as the used to compute \code{object}.} \item{pcs}{\code{numeric} The number of PC's to consider} \item{pre}{pre-process \code{newdata} based on the pre-processing chosen for the PCA model} \item{post}{unpre-process the final data (add the center back etc)} \item{...}{Not passed on anywhere, included for S3 consistency.} } \value{ A list with the following components: \item{scores}{The predicted scores} \item{x}{The predicted data} } \description{ Predict data using PCA model } \details{ This function extracts the predict values from a pcaRes object for the PCA methods SVD, Nipals, PPCA and BPCA. Newdata is first centered if the PCA model was and then scores (\eqn{T}) and data (\eqn{X}) is 'predicted' according to : \eqn{\hat{T}=X_{new}P}{That=XnewP} \eqn{\hat{X}_{new}=\hat{T}P'}{Xhat=ThatP'}. Missing values are set to zero before matrix multiplication to achieve NIPALS like treatment of missing values. } \examples{ data(iris) hidden <- sample(nrow(iris), 50) pcIr <- pca(iris[-hidden,1:4]) pcFull <- pca(iris[,1:4]) irisHat <- predict(pcIr, iris[hidden,1:4]) cor(irisHat$scores[,1], scores(pcFull)[hidden,1]) } \author{ Henning Redestig } \keyword{multivariate} pcaMethods/man/prep.Rd0000644000175200017520000000445614710217306015705 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/prep.R \name{prep} \alias{prep} \title{Pre-process a matrix for PCA} \usage{ prep(object, scale = c("none", "pareto", "vector", "uv"), center = TRUE, eps = 1e-12, simple = TRUE, reverse = FALSE, ...) } \arguments{ \item{object}{Numerical matrix (or an object coercible to such) with samples in rows and variables as columns. Also takes \code{ExpressionSet} in which case the transposed expression matrix is used.} \item{scale}{One of "UV" (unit variance \eqn{a=a/\sigma_{a}}) "vector" (vector normalisation \eqn{b=b/||b||}), "pareto" (sqrt UV) or "none" to indicate which scaling should be used to scale the matrix with \eqn{a} variables and \eqn{b} samples. Can also be a vector of scales which should be used to scale the matrix. \code{NULL} value is interpreted as \code{"none"}.} \item{center}{Either a logical which indicates if the matrix should be mean centred or not, or a vector with averages which should be suntracted from the matrix. \code{NULL} value is interpreted as \code{FALSE}} \item{eps}{Minimum variance, variable with lower variance are not scaled and warning is issued instead.} \item{simple}{Logical indicating if only the data should be returned or a list with the pre-processing statistics as well.} \item{reverse}{Logical indicating if matrix should be 'post-processed' instead by multiplying each column with its scale and adding the center. In this case, center and scale should be vectors with the statistics (no warning is issued if not, instead output becomes the same as input).} \item{...}{Only used for passing through arguments.} } \value{ A pre-processed matrix or a list with \item{center}{a vector with the estimated centers} \item{scale}{a vector with the estimated scales} \item{data}{the pre (or post) processed data} } \description{ Scaling and centering a matrix. } \details{ Does basically the same as \code{\link{scale}} but adds some alternative scaling options and functionality for treating pre-processing as part of a model. } \examples{ object <- matrix(rnorm(50), nrow=10) res <- prep(object, scale="uv", center=TRUE, simple=FALSE) obj <- prep(object, scale=res$scale, center=res$center) ## same as original sum((object - prep(obj, scale=res$scale, center=res$center, rev=TRUE))^2) } \author{ Henning Redestig } pcaMethods/man/rediduals-methods.Rd0000644000175200017520000000216514710217306020347 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{rediduals-methods} \alias{rediduals-methods} \alias{residuals.pcaRes} \alias{residuals,pcaRes-method} \alias{resid,pcaRes-method} \title{Residuals values from a PCA model.} \usage{ \method{residuals}{pcaRes}(object, data = completeObs(object), ...) \S4method{residuals}{pcaRes}(object, data = completeObs(object), ...) \S4method{resid}{pcaRes}(object, data = completeObs(object), ...) } \arguments{ \item{object}{\code{pcaRes} the \code{pcaRes} object of interest.} \item{data}{\code{matrix} The data that was used to calculate the PCA model (or a different dataset to e.g. adress its proximity to the model).} \item{...}{Passed on to \code{\link{predict.pcaRes}}. E.g. setting the number of used components.} } \value{ A \code{matrix} with the residuals } \description{ This function extracts the residuals values from a pcaRes object for the PCA methods SVD, Nipals, PPCA and BPCA } \examples{ data(iris) pcIr <- pca(iris[,1:4]) head(residuals(pcIr, iris[,1:4])) } \author{ Henning Redestig } \keyword{multivariate} pcaMethods/man/repmat.Rd0000644000175200017520000000077314710217306016225 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/repmat.R \name{repmat} \alias{repmat} \title{Replicate and tile an array.} \usage{ repmat(mat, M, N) } \arguments{ \item{mat}{numeric matrix} \item{M}{number of copies in vertical direction} \item{N}{number of copies in horizontal direction} } \value{ Matrix consiting of M-by-N tiling copies of input matrix } \description{ Creates a large matrix B consisting of an M-by-N tiling of copies of A } \author{ Wolfram Stacklies } pcaMethods/man/robustPca.Rd0000644000175200017520000000632014710217306016671 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/robustPca.R \name{robustPca} \alias{robustPca} \title{PCA implementation based on robustSvd} \usage{ robustPca(Matrix, nPcs = 2, verbose = interactive(), ...) } \arguments{ \item{Matrix}{\code{matrix} -- Data containing the variables in columns and observations in rows. The data may contain missing values, denoted as \code{NA}.} \item{nPcs}{\code{numeric} -- Number of components to estimate. The preciseness of the missing value estimation depends on the number of components, which should resemble the internal structure of the data.} \item{verbose}{\code{boolean} Print some output to the command line if TRUE} \item{...}{Reserved for future use. Currently no further parameters are used} } \value{ Standard PCA result object used by all PCA-based methods of this package. Contains scores, loadings, data mean and more. See \code{\link{pcaRes}} for details. are used. } \description{ This is a PCA implementation robust to outliers in a data set. It can also handle missing values, it is however NOT intended to be used for missing value estimation. As it is based on robustSVD we will get an accurate estimation for the loadings also for incomplete data or for data with outliers. The returned scores are, however, affected by the outliers as they are calculated inputData X loadings. This also implies that you should look at the returned R2/R2cum values with caution. If the data show missing values, scores are caluclated by just setting all NA - values to zero. This is not expected to produce accurate results. Please have also a look at the manual page for \code{robustSvd}. Thus this method should mainly be seen as an attempt to integrate \code{robustSvd()} into the framework of this package. Use one of the other methods coming with this package (like PPCA or BPCA) if you want to do missing value estimation. It is not recommended to use this function directely but rather to use the pca() wrapper function. } \details{ The method is very similar to the standard \code{prcomp()} function. The main difference is that \code{robustSvd()} is used instead of the conventional \code{svd()} method. } \examples{ ## Load a complete sample metabolite data set and mean center the data data(metaboliteDataComplete) mdc <- scale(metaboliteDataComplete, center=TRUE, scale=FALSE) ## Now create 5\\\% of outliers. cond <- runif(length(mdc)) < 0.05; mdcOut <- mdc mdcOut[cond] <- 10 ## Now we do a conventional PCA and robustPca on the original and the data ## with outliers. ## We use center=FALSE here because the large artificial outliers would ## affect the means and not allow to objectively compare the results. resSvd <- pca(mdc, method="svd", nPcs=10, center=FALSE) resSvdOut <- pca(mdcOut, method="svd", nPcs=10, center=FALSE) resRobPca <- pca(mdcOut, method="robustPca", nPcs=10, center=FALSE) ## Now we plot the results for the original data against those with outliers ## We can see that robustPca is hardly effected by the outliers. plot(loadings(resSvd)[,1], loadings(resSvdOut)[,1]) plot(loadings(resSvd)[,1], loadings(resRobPca)[,1]) } \seealso{ \code{\link{robustSvd}, \link{svd}, \link{prcomp}, \link{pcaRes}}. } \author{ Wolfram Stacklies } \keyword{multivariate} pcaMethods/man/robustSvd.Rd0000644000175200017520000000620414710217306016723 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/robustPca.R \name{robustSvd} \alias{robustSvd} \title{Alternating L1 Singular Value Decomposition} \usage{ robustSvd(x) } \arguments{ \item{x}{A matrix whose SVD decomposition is to be computed. Missing values are allowed.} } \value{ The robust SVD of the matrix is x=u d v'. \item{d}{A vector containing the singular values of \code{x}.} \item{u}{A matrix whose columns are the left singular vectors of \code{x}.} \item{v}{A matrix whose columns are the right singular vectors of \code{x}.} } \description{ A robust approximation to the singular value decomposition of a rectangular matrix is computed using an alternating L1 norm (instead of the more usual least squares L2 norm). As the SVD is a least-squares procedure, it is highly susceptible to outliers and in the extreme case, an individual cell (if sufficiently outlying) can draw even the leading principal component toward itself. } \details{ See Hawkins et al (2001) for details on the robust SVD algorithm. Briefly, the idea is to sequentially estimate the left and right eigenvectors using an L1 (absolute value) norm minimization. Note that the robust SVD is able to accomodate missing values in the matrix \code{x}, unlike the usual \code{svd} function. Also note that the eigenvectors returned by the robust SVD algorithm are NOT (in general) orthogonal and the eigenvalues need not be descending in order. } \note{ Two differences from the usual SVD may be noted. One relates to orthogonality. In the conventional SVD, all the eigenvectors are orthogonal even if not explicitly imposed. Those returned by the AL1 algorithm (used here) are (in general) not orthogonal. Another difference is that, in the L2 analysis of the conventional SVD, the successive eigen triples (eigenvalue, left eigenvector, right eigenvector) are found in descending order of eigenvalue. This is not necessarily the case with the AL1 algorithm. Hawkins et al (2001) note that a larger eigen value may follow a smaller one. } \examples{ ## Load a complete sample metabolite data set and mean center the data data(metaboliteDataComplete) mdc <- prep(metaboliteDataComplete, center=TRUE, scale="none") ## Now create 5\% of outliers. cond <- runif(length(mdc)) < 0.05; mdcOut <- mdc mdcOut[cond] <- 10 ## Now we do a conventional SVD and a robustSvd on both, the original and the ## data with outliers. resSvd <- svd(mdc) resSvdOut <- svd(mdcOut) resRobSvd <- robustSvd(mdc) resRobSvdOut <- robustSvd(mdcOut) ## Now we plot the results for the original data against those with outliers ## We can see that robustSvd is hardly affected by the outliers. plot(resSvd$v[,1], resSvdOut$v[,1]) plot(resRobSvd$v[,1], resRobSvdOut$v[,1]) } \references{ Hawkins, Douglas M, Li Liu, and S Stanley Young (2001) Robust Singular Value Decomposition, National Institute of Statistical Sciences, Technical Report Number 122. \url{http://www.niss.org/technicalreports/tr122.pdf} } \seealso{ \code{\link{svd}}, \code{\link[ade4:nipals]{nipals}} for an alternating L2 norm method that also accommodates missing data. } \author{ Kevin Wright, modifications by Wolfram Stacklies } \keyword{algebra} pcaMethods/man/sDev-pcaRes-method.Rd0000644000175200017520000000077314710217306020327 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{sDev,pcaRes-method} \alias{sDev,pcaRes-method} \alias{sDev} \title{Get the standard deviations of the scores (indicates their relevance)} \usage{ sDev(object, ...) } \arguments{ \item{object}{pcaRes object} \item{...}{Not used} } \value{ Standard devations of the scores } \description{ Get the standard deviations of the scores (indicates their relevance) } \author{ Henning Redestig } pcaMethods/man/scaled-pcaRes-method.Rd0000644000175200017520000000072614710217306020657 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{scaled,pcaRes-method} \alias{scaled,pcaRes-method} \alias{scaled} \title{Check if scaling was part of the PCA model} \usage{ scaled(object, ...) } \arguments{ \item{object}{pcaRes object} \item{...}{Not used} } \value{ TRUE if scaling was part of the PCA model } \description{ Check if scaling was part of the PCA model } \author{ Henning Redestig } pcaMethods/man/scl-pcaRes-method.Rd0000644000175200017520000000101014710217306020170 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{scl,pcaRes-method} \alias{scl,pcaRes-method} \alias{scl} \title{Get the scales (e.g. standard deviations) of the original variables} \usage{ scl(object, ...) } \arguments{ \item{object}{pcaRes object} \item{...}{Not used} } \value{ Vector with the scales } \description{ Get the scales (e.g. standard deviations) of the original variables } \seealso{ \code{\link{prep}} } \author{ Henning Redestig } pcaMethods/man/scores-pcaRes-method.Rd0000644000175200017520000000075214710217306020721 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{scores,pcaRes-method} \alias{scores,pcaRes-method} \alias{scores} \title{Get scores from a pcaRes object} \usage{ \S4method{scores}{pcaRes}(object, ...) } \arguments{ \item{object}{a pcaRes object} \item{...}{not used} } \value{ The scores as a matrix } \description{ Get scores from a pcaRes object } \seealso{ \code{\link{scores.pcaRes}} } \author{ Henning Redestig } pcaMethods/man/scores.pcaRes.Rd0000644000175200017520000000062114710217306017437 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \name{scores.pcaRes} \alias{scores.pcaRes} \title{Get scores from a pcaRes object} \usage{ \method{scores}{pcaRes}(object, ...) } \arguments{ \item{object}{a pcaRes object} \item{...}{not used} } \value{ The scores as a matrix } \description{ Get scores from a pcaRes object } \author{ Henning Redestig } pcaMethods/man/show-methods.Rd0000644000175200017520000000122114710217306017343 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{show-methods} \alias{show-methods} \alias{showPcaRes} \alias{print,pcaRes-method} \alias{print,nniRes-method} \alias{show,pcaRes-method} \alias{show,nniRes-method} \title{Print/Show for pcaRes} \usage{ showPcaRes(x, ...) \S4method{print}{pcaRes}(x, ...) \S4method{show}{pcaRes}(object) } \arguments{ \item{x}{a pcaRes object} \item{...}{not used} \item{object}{the object to print information about} } \value{ nothing, used for its side effect } \description{ Print basic information about pcaRes object } \author{ Henning Redestig } pcaMethods/man/showNniRes.Rd0000644000175200017520000000060214710217306017023 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-nniRes.R \name{showNniRes} \alias{showNniRes} \title{Print a nniRes model} \usage{ showNniRes(x, ...) } \arguments{ \item{x}{An \code{nniRes} object} \item{...}{Not used} } \value{ Nothing, used for side-effect } \description{ Print a brief description of nniRes model } \author{ Henning Redestig } pcaMethods/man/simpleEllipse.Rd0000644000175200017520000000156314710217306017542 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/pca.R \name{simpleEllipse} \alias{simpleEllipse} \title{Hotelling's T^2 Ellipse} \usage{ simpleEllipse(x, y, alfa = 0.95, len = 200) } \arguments{ \item{x}{first variable} \item{y}{second variable} \item{alfa}{confidence level of the circle} \item{len}{Number of points in the circle} } \value{ A matrix with X and Y coordinates for the circle } \description{ Get a confidence ellipse for uncorrelated bivariate data } \details{ As described in 'Introduction to multi and megavariate data analysis using PCA and PLS' by Eriksson et al. This produces very similar ellipse as compared to the ellipse function the ellipse package except that this function assumes that and y are uncorrelated (which they of are if they are scores or loadings from a PCA). } \seealso{ ellipse } \author{ Henning Redestig } pcaMethods/man/slplot-pcaRes-method.Rd0000644000175200017520000000340614710217306020737 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{slplot,pcaRes-method} \alias{slplot,pcaRes-method} \alias{slplot} \title{Side by side scores and loadings plot} \usage{ slplot(object, pcs=c(1,2), scoresLoadings=c(TRUE, TRUE), sl="def", ll="def", hotelling=0.95, rug=TRUE, sub=NULL,...) } \arguments{ \item{object}{a pcaRes object} \item{pcs}{which two pcs to plot} \item{scoresLoadings}{Which should be shown scores and or loadings} \item{sl}{labels to plot in the scores plot} \item{ll}{labels to plot in the loadings plot} \item{hotelling}{confidence interval for ellipse in the score plot} \item{rug}{logical, rug x axis in score plot or not} \item{sub}{Subtitle, defaults to annotate with amount of explained variance.} \item{...}{Further arguments to plot functions. Prefix arguments to \code{par()} with 's' for the scores plot and 'l' for the loadings plot. I.e. cex become scex for setting character expansion in the score plot and lcex for the loadings plot.} } \value{ None, used for side effect. } \description{ A common way of visualizing two principal components } \details{ This method is meant to be used as a quick way to visualize results, if you want a more specific plot you probably want to get the scores, loadings with \code{scores(object)}, \code{loadings(object)} and then design your own plotting method. } \note{ Uses layout instead of par to provide side-by-side so it works with Sweave (but can not be combined with \code{par(mfrow=..))} } \examples{ data(iris) pcIr <- pca(iris[,1:4], scale="uv") slplot(pcIr, sl=NULL, spch=5) slplot(pcIr, sl=NULL, lcex=1.3, scol=as.integer(iris[,5])) } \seealso{ \code{\link{pca}}, \code{\link{biplot}} } \author{ Henning Redestig } \keyword{multivariate} pcaMethods/man/sortFeatures.Rd0000644000175200017520000000070514710217306017416 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/sortFeatures.R \name{sortFeatures} \alias{sortFeatures} \title{Sort the features of NLPCA object} \usage{ sortFeatures(nlnet, trainIn, trainOut) } \arguments{ \item{nlnet}{The nlnet} \item{trainIn}{Training data in} \item{trainOut}{Training data after it passed through the net} } \value{ ... } \description{ Sort the features of NLPCA object } \author{ Henning Redestig } pcaMethods/man/summary.Rd0000644000175200017520000000070214710217306016422 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \name{summary} \alias{summary} \alias{summary.pcaRes} \alias{summary,pcaRes-method} \title{Summary of PCA model} \usage{ \method{summary}{pcaRes}(object, ...) } \arguments{ \item{object}{a pcaRes object} \item{...}{Not used} } \value{ Nothing, used for side-effect } \description{ Print a brief description of the PCA model } \author{ Henning Redestig } pcaMethods/man/svdImpute.Rd0000644000175200017520000000602414710217306016710 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/svdImpute.R \name{svdImpute} \alias{svdImpute} \title{SVDimpute algorithm} \usage{ svdImpute(Matrix, nPcs = 2, threshold = 0.01, maxSteps = 100, verbose = interactive(), ...) } \arguments{ \item{Matrix}{\code{matrix} -- Pre-processed (centered, scaled) data with variables in columns and observations in rows. The data may contain missing values, denoted as \code{NA}.} \item{nPcs}{\code{numeric} -- Number of components to estimate. The preciseness of the missing value estimation depends on the number of components, which should resemble the internal structure of the data.} \item{threshold}{The iteration stops if the change in the matrix falls below this threshold.} \item{maxSteps}{Maximum number of iteration steps.} \item{verbose}{Print some output if TRUE.} \item{...}{Reserved for parameters used in future version of the algorithm} } \value{ Standard PCA result object used by all PCA-based methods of this package. Contains scores, loadings, data mean and more. See \code{\link{pcaRes}} for details. } \description{ This implements the SVDimpute algorithm as proposed by Troyanskaya et al, 2001. The idea behind the algorithm is to estimate the missing values as a linear combination of the \code{k} most significant eigengenes. } \details{ Missing values are denoted as \code{NA}. It is not recommended to use this function directely but rather to use the pca() wrapper function. As SVD can only be performed on complete matrices, all missing values are initially replaced by 0 (what is in fact the mean on centred data). The algorithm works iteratively until the change in the estimated solution falls below a certain threshold. Each step the eigengenes of the current estimate are calculated and used to determine a new estimate. Eigengenes denote the loadings if pca is performed considering variable (for Microarray data genes) as observations. An optimal linear combination is found by regressing the incomplete variable against the \code{k} most significant eigengenes. If the value at position \code{j} is missing, the \eqn{j^th}{j^th} value of the eigengenes is not used when determining the regression coefficients. } \note{ Each iteration, standard PCA (\code{prcomp}) needs to be done for each incomplete variable to get the eigengenes. This is usually fast for small data sets, but complexity may rise if the data sets become very large. } \examples{ ## Load a sample metabolite dataset with 5\\\% missing values data(metaboliteData) ## Perform svdImpute using the 3 largest components result <- pca(metaboliteData, method="svdImpute", nPcs=3, center = TRUE) ## Get the estimated complete observations cObs <- completeObs(result) ## Now plot the scores plotPcs(result, type = "scores") } \references{ Troyanskaya O. and Cantor M. and Sherlock G. and Brown P. and Hastie T. and Tibshirani R. and Botstein D. and Altman RB. - Missing value estimation methods for DNA microarrays. \emph{Bioinformatics. 2001 Jun;17(6):520-5.} } \author{ Wolfram Stacklies } \keyword{multivariate} pcaMethods/man/svdPca.Rd0000644000175200017520000000260514710217306016151 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/pca.R \name{svdPca} \alias{svdPca} \title{Perform principal component analysis using singular value decomposition} \usage{ svdPca(Matrix, nPcs = 2, varLimit = 1, verbose = interactive(), ...) } \arguments{ \item{Matrix}{Pre-processed (centered and possibly scaled) numerical matrix samples in rows and variables as columns. No missing values allowed.} \item{nPcs}{Number of components that should be extracted.} \item{varLimit}{Optionally the ratio of variance that should be explained. \code{nPcs} is ignored if varLimit < 1} \item{verbose}{Verbose complaints to matrix structure} \item{...}{Only used for passing through arguments.} } \value{ A \code{pcaRes} object. } \description{ A wrapper function for \code{prcomp} to deliver the result as a \code{pcaRes} method. Supplied for compatibility with the rest of the pcaMethods package. It is not recommended to use this function directely but rather to use the \code{pca()} wrapper function. } \examples{ data(metaboliteDataComplete) mat <- prep(t(metaboliteDataComplete)) pc <- svdPca(mat, nPcs=2) ## better use pca() pc <- pca(t(metaboliteDataComplete), method="svd", nPcs=2) \dontshow{stopifnot(sum((fitted(pc) - t(metaboliteDataComplete))^2, na.rm=TRUE) < 200)} } \seealso{ \code{prcomp}, \code{princomp}, \code{pca} } \author{ Henning Redestig } \keyword{multivariate} pcaMethods/man/tempFixNas.Rd0000644000175200017520000000063414710217306017007 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/xval.R \name{tempFixNas} \alias{tempFixNas} \title{Temporary fix for missing values} \usage{ tempFixNas(mat) } \arguments{ \item{mat}{a matrix} } \value{ The original matrix with completely missing rows/cols filled with zeroes. } \description{ Simply replace completely missing rows or cols with zeroes. } \author{ Henning Redestig } pcaMethods/man/vector2matrices-matrix-method.Rd0000644000175200017520000000100214710217306022613 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/vector2matrices.R \docType{methods} \name{vector2matrices,matrix-method} \alias{vector2matrices,matrix-method} \title{Tranform the vectors of weights to matrix structure} \usage{ \S4method{vector2matrices}{matrix}(object, net) } \arguments{ \item{object}{an nlpcaNet} \item{net}{the neural network} } \value{ weights in matrix structure } \description{ Tranform the vectors of weights to matrix structure } \author{ Henning Redestig } pcaMethods/man/vector2matrices-nlpcaNet-method.Rd0000644000175200017520000000074314710217306023066 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/vector2matrices.R \docType{methods} \name{vector2matrices,nlpcaNet-method} \alias{vector2matrices,nlpcaNet-method} \title{Tranform the vectors of weights to matrix structure} \usage{ \S4method{vector2matrices}{nlpcaNet}(object) } \arguments{ \item{object}{an nlpcaNet} } \value{ weights in matrix structure } \description{ Tranform the vectors of weights to matrix structure } \author{ Henning Redestig } pcaMethods/man/wasna-pcaRes-method.Rd0000644000175200017520000000144014710217306020527 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/methods-pcaRes.R \docType{methods} \name{wasna,pcaRes-method} \alias{wasna,pcaRes-method} \alias{wasna} \title{Get a matrix with indicating the elements that were missing in the input data. Convenient for estimating imputation performance.} \usage{ wasna(object, ...) } \arguments{ \item{object}{pcaRes object} \item{...}{Not used} } \value{ A matrix with logicals } \description{ Get a matrix with indicating the elements that were missing in the input data. Convenient for estimating imputation performance. } \examples{ data(metaboliteData) data(metaboliteDataComplete) result <- pca(metaboliteData, nPcs=2) plot(completeObs(result)[wasna(result)], metaboliteDataComplete[wasna(result)]) } \author{ Henning Redestig } pcaMethods/man/weightsAccount.Rd0000644000175200017520000000106014710217306017712 0ustar00biocbuildbiocbuild% Generated by roxygen2: do not edit by hand % Please edit documentation in R/AllClasses.R \name{weightsAccount} \alias{weightsAccount} \title{Create an object that holds the weights for nlpcaNet. Holds and sets weights in using an environment object.} \usage{ weightsAccount(w) } \arguments{ \item{w}{\code{matrix} -- New weights} } \value{ A weightsAccound with \code{set} and \code{current} functions. } \description{ Create an object that holds the weights for nlpcaNet. Holds and sets weights in using an environment object. } \author{ Henning Redestig } pcaMethods/src/0000755000175200017520000000000014710312420014444 5ustar00biocbuildbiocbuildpcaMethods/src/RcppExports.cpp0000644000175200017520000000102414710217306017445 0ustar00biocbuildbiocbuild#include using namespace Rcpp; // Nipals List Nipals(SEXP Mat, SEXP params); RcppExport SEXP pcaMethods_Nipals(SEXP MatSEXP, SEXP paramsSEXP) { BEGIN_RCPP SEXP __sexp_result; { Rcpp::RNGScope __rngScope; Rcpp::traits::input_parameter< SEXP >::type Mat(MatSEXP ); Rcpp::traits::input_parameter< SEXP >::type params(paramsSEXP ); List __result = Nipals(Mat, params); PROTECT(__sexp_result = Rcpp::wrap(__result)); } UNPROTECT(1); return __sexp_result; END_RCPP } pcaMethods/src/nipals.cpp0000644000175200017520000000731114710217306016447 0ustar00biocbuildbiocbuild#include #include #include using namespace std; using namespace Rcpp; double difference(vector& vec1, vector& vec2) { double diff = 0; double a; int len = vec1.size(); for(int i = 0; i < len; i++) { a = vec1[i] - vec2[i]; diff += a * a; } return(diff); } void norm(vector& vec) { double siz = 0; int len = vec.size(); for(int i = 0; i < len; i++) { siz += vec[i] * vec[i]; } siz = sqrt(siz); for(int i = 0; i < len; i++) { vec[i] = vec[i] / siz; } } // [[Rcpp::export]] List Nipals(SEXP Mat, SEXP params) { try{ bool cnt; int count = 0; double tsize; Rcpp::List rl = R_NilValue; Rcpp::List rparams(params); int maxSteps = Rcpp::as(rparams["maxSteps"]); double eps = Rcpp::as(rparams["threshold"]); int nPcs = Rcpp::as(rparams["nPcs"]); double varLimit = Rcpp::as(rparams["varLimit"]); Rcpp::NumericMatrix mat(Mat); Rcpp::NumericMatrix omat = Rcpp::clone( Mat ); int nr = mat.nrow(); int nc = mat.ncol(); Rcpp::NumericMatrix est_mat(nr, nc); Rcpp::NumericMatrix tt(nr, nPcs); Rcpp::NumericMatrix pp(nc, nPcs); vector r2cum; vector thold(nr); vector th(nr); vector phold(nc); vector ph(nc); double tss = 0; double sse = 0; int np = 0; double anotherPc = true; for (int r = 0; r < nr; r++) { for (int c = 0; c < nc; c++) { if(!ISNAN(mat(r,c))) { tss += mat(r,c) * mat(r,c); } } } while(anotherPc) { for(int r = 0; r < nr; r++) { th[r] = 0; if(!ISNAN(mat(r,0))) { th[r] = mat(r,0); } } cnt = true; count = 0; while(cnt) { count++; for(int c = 0; c < nc; c++) { ph[c] = 0; } tsize = 0; for(int r = 0; r < nr; r++) { tsize += th[r] * th[r]; } for(int r = 0; r < nr; r++) { double ti = th[r] / tsize; for(int c = 0; c < nc; c++) { if(!ISNAN(mat(r,c))) { ph[c] += mat(r,c) * ti; } } } norm(ph); thold = th; for(int r = 0; r < nr; r++) { th[r] = 0; for(int c = 0; c < nc; c++) { if(!ISNAN(mat(r,c))) { th[r] += mat(r,c) * ph[c]; } } } if(count > maxSteps) { throw 1; } if(difference(thold, th) <= eps) { cnt = false; } } //deflate mat sse = 0; double mathat = 0; double err = 0; for(int r = 0; r < nr; r++) { for(int c = 0; c < nc; c++) { if(!ISNAN(mat(r,c))) { mathat = th[r] * ph[c]; est_mat(r, c) += mathat; err = omat(r,c) - est_mat(r, c); sse += err * err; mat(r,c) -= mathat; } } } r2cum.push_back(1 - (sse / tss)); for(int r = 0; r < nr; r++) { tt(r,np) = th[r]; } for(int c = 0; c < nc; c++) { pp(c,np) = ph[c]; } if(fabs(varLimit - 1) > 1e-4) { if(r2cum[np] >= varLimit) { anotherPc = false; } } if (np + 1 >= nPcs){ anotherPc = false; } np++; } if(np != nPcs) { Rcpp::NumericMatrix ttt(nr, np); Rcpp::NumericMatrix ppp(nc, np); for(int r = 0; r < nr; r++) { for(int p = 0; p < np; p++) { ttt(r,p) = tt(r,p); } } for(int c = 0; c < nc; c++) { for(int p = 0; p < np; p++) { ppp(c,p) = pp(c,p); } } rl["scores"] = ttt; rl["loadings"] = ppp; } else { rl["scores"] = tt; rl["loadings"] = pp; } rl["R2cum"] = r2cum; return rl; }catch(int e) { if(e == 1) { ::Rf_error("Too many iterations, quitting"); }else { ::Rf_error("unknown error"); } } catch(std::exception& ex) { forward_exception_to_r(ex); } catch(...) { ::Rf_error("unknown error"); } return R_NilValue; } pcaMethods/vignettes/0000755000175200017520000000000014710312420015665 5ustar00biocbuildbiocbuildpcaMethods/vignettes/missingValues.Rnw0000644000175200017520000000565414710217306021227 0ustar00biocbuildbiocbuild\documentclass[a4paper]{article} %\VignetteIndexEntry{Missing value imputation} \usepackage{hyperref} \title{Imputing missing values using the pcaMethods package} \author{Wolfram Stacklies and Henning Redestig\\ CAS-MPG Partner Institute for Computational Biology (PICB)\\ Shanghai, P.R. China \\ and\\ Max Planck Institute for Molecular Plant Physiology\\ Potsdam, Germany\\ \url{http://bioinformatics.mpimp-golm.mpg.de/} } \date{\today} \begin{document} \setkeys{Gin}{width=1.0\textwidth} @ \maketitle \section{Missing value imputation} One application for missing value robust principal component analysis is that it effectively can be used to impute the missing values and thus obtain an estimated complete data set. The pcaMethods package was partly written with this application in mind. PCA is a way of creating a model of a matrix, $X$, by defining two parameter matrices, the scores, $T$, and the loadings, $P$, which together have less values than the original matrix but when multiplied with each other well reconstruct the original matrix. I.e.: $$X=1\times{}\bar{x} + TP' + E$$ where $E$ is the error matrix and $1\times{}\bar{x}$ denotes the original variable averages. Now if $X$ contains missing values but we still are able to get complete estimates of $P$ and $T$ than we can use: $$\hat{X}=1\times{}\bar{x} + TP'$$ as an estimate for $x_{i,j}$ if $x_{i,j}$ is missing. This is can be done as the following example illustrates. First we attach the metabolite data set with missing values. <>= library(pcaMethods) @ <<>>= data(metaboliteData) mD <- metaboliteData sum(is.na(mD)) @ Now we get the estimated data set by using PPCA and three principal components. <<>>= pc <- pca(mD, nPcs=3, method="ppca") imputed <- completeObs(pc) @ If we compare with the original values we see that the error is rather low. <<>>= data(metaboliteDataComplete) mdComp <- metaboliteDataComplete sum((mdComp[is.na(mD)] - imputed[is.na(mD)])^2) / sum(mdComp[is.na(mD)]^2) @ When using a different PCA algorithm, we get different performance. <<>>= imputedNipals <- completeObs(pca(mD, nPcs=3, method="nipals")) sum((mdComp[is.na(mD)] - imputedNipals[is.na(mD)])^2) / sum(mdComp[is.na(mD)]^2) @ If the data we are interested in was gene expression set of class 'ExpressionSet' we could simply do <<>>= library(Biobase) data(sample.ExpressionSet) exSet <- sample.ExpressionSet exSetNa <- exSet exprs(exSetNa)[sample(13000, 200)] <- NA lost <- is.na(exprs(exSetNa)) pc <- pca(exSetNa, nPcs=2, method="ppca") impExSet <- asExprSet(pc, exSetNa) sum((exprs(exSet)[lost] - exprs(impExSet)[lost])^2) / sum(exprs(exSet)[lost]^2) @ Different results will be obtained with different PCA algorithms. Which one to use depends on the general structure of the data set and the imputation performance can be estimated by cross-validation. Please see the 'introduction' vignette on further details on how to use the cross-validation capabilities of this package. \end{document} pcaMethods/vignettes/outliers.Rnw0000644000175200017520000001203114710217306020227 0ustar00biocbuildbiocbuild\documentclass[a4paper]{article} %\VignetteIndexEntry{Data with outliers} \usepackage{hyperref} \title{Handling of data containing outliers} \author{Wolfram Stacklies and Henning Redestig\\ CAS-MPG Partner Institute for Computational Biology (PICB)\\ Shanghai, P.R. China \\ and\\ Max Planck Institute for Molecular Plant Physiology\\ Potsdam, Germany\\ \url{http://bioinformatics.mpimp-golm.mpg.de/} } \date{\today} \begin{document} \setkeys{Gin}{width=1.0\textwidth} @ \maketitle \section{PCA robust to outliers} Away from often showing missing values, Microarray or Metabolite data are often corrupted with extreme values (outliers). Standard SVD is highly susceptible to outliers. In the extreme case, an individual data point, if sufficiently outlying, can draw even the leading principal component toward itself. This problem can be addressed by using a robust analysis method. Hereto we provide \texttt{robustSvd}, a singular value decomposition robust to outliers. \texttt{robustPca} is a PCA implementation that resembles the original \texttt{R} \texttt{prcomp} method, with the difference that it uses \texttt{robustSvd} instead of the standard \texttt{svd} function.\\ Robust SVD and its application to microarray data were proposed in \cite{hawkins01} and \cite{liu03}. The algorithm is based on the idea to use a sequential estimation of the eigenvalues and left and right eigenvectors that ignores missing values and is resistant to outliers. \\ The \texttt{robustSvd} script included here was contributed by Kevin Wright. Thanks a lot to him! \section{Outliers and missing value imputation} The problem of outliers is similar to the missing data problem in the sense that extreme values provide no or wrong information. They are generally artifacts of the experiment and provide no information about the underlying biological processes. \\ Most of the PCA methods coming with the package were not designed to be robust to outliers in the sense that they will converge to the standard PCA solution on a complete data set. Yet, an applicable solution is to remove obvious outliers from the data first (by setting them NA) and to then estimate the PCA solution on the incomplete data. This is likely to produce accurate results if the number of missing data does not exceed a certain amount, less than 10\% should be a good number. The following example illustrates the effect of outliers and the use of robust methods. First, we attach the complete metabolite data set and create 5\% outliers. We mean center the data before we create outliers because these large artificial outliers will strongly shift the original means. This would not allow for objective comparison between the differnt results obtained, e.g. when doing scatterplots. <>= library(pcaMethods) @ <<>>= data(metaboliteDataComplete) mdc <- scale(metaboliteDataComplete, center=TRUE, scale=FALSE) cond <- runif(length(mdc)) < 0.05 mdcOut <- mdc mdcOut[cond] <- 10 @ Then we calculate a PCA solution using standard SVD and robust SVD. <>= resSvd <- pca(mdc, method="svd", nPcs=5, center=FALSE) resSvdOut <- pca(mdcOut, method="svd", nPcs=5, center=FALSE) resRobSvd <- pca(mdcOut, method="robustPca", nPcs=5, center=FALSE) @ Now we use \texttt{PPCA} to estimate the PCA solution, but set the outliers NA before. <>= mdcNa <- mdc mdcNa[cond] <- NA resPPCA <- pca(mdcNa, method="ppca", nPcs=5, center=FALSE) @ To check the robustness to outliers we can just do a scatterplot comparing the results to the optimal PCA solution for the complete data set (which is \texttt{resSvd}). In Figure \ref{fig:svdPlot} we plot the estimated and original loadings against each other. \begin{figure}[!ht] \centering <>= par(mfrow=c(2,2)) plot(loadings(resSvd)[,1], loadings(resSvdOut)[,1], xlab="Loading 1 SVD", ylab="Loading 1 SVD with outliers") plot(loadings(resSvd)[,1], loadings(resRobSvd)[,1], xlab="Loading 1 SVD", ylab="Loading 1 robustSVD with outliers") plot(loadings(resSvd)[,1], loadings(resPPCA)[,1], xlab="Loading 1 SVD", ylab="Loading 1 PPCA with outliers=NA") plot(loadings(resRobSvd)[,1], loadings(resPPCA)[,1], xlab="Loading 1 robust SVD with outliers", ylab="Loading 1 svdImpute with outliers=NA") @ \caption{Figures show (from left to right): \newline Original PCA solution vs. solution on data with outliers; \newline Original PCA solution vs. robust PCA solution on data with outliers; \newline Original PCA solution vs. PPCA solution on data where outliers=NA; \newline Robust PCA solution vs. PPCA solution on data with outliers / outliers=NA. \label{fig:svdPlot} } \end{figure} \begin{thebibliography}{2006} \bibitem{hawkins01} Hawkins, D.M., Liu, L. and Young, S.S. {\sl Robust Singular Value Decomposition.} National Institute of Statistical Sciences, 2001, Tech Report 122. \bibitem{liu03} Liu, L., Hawkins, D.M., Ghosh, S. and Young, S.S. {\sl Robust singular value decomposition analysis of microarray data.} PNAS, 2003;100:13167--13172. \end{thebibliography} \end{document} pcaMethods/vignettes/pcaMethods.Rnw0000644000175200017520000005554214710217306020466 0ustar00biocbuildbiocbuild\documentclass[a4paper]{article} %\VignetteIndexEntry{Introduction} \usepackage{hyperref} \title{The pcaMethods Package} \author{Wolfram Stacklies and Henning Redestig\\ CAS-MPG Partner Institute for Computational Biology (PICB)\\ Shanghai, P.R. China \\ and\\ Max Planck Institute for Molecular Plant Physiology\\ Potsdam, Germany\\ \url{http://bioinformatics.mpimp-golm.mpg.de/} } \date{\today} \begin{document} \setkeys{Gin}{width=1.0\textwidth} @ \maketitle \section*{Overview} The \texttt{pcaMethods} package \cite{stacklies07} provides a set of different PCA implementations, together with tools for cross validation and visualisation of the results. The methods basically allow to perform PCA on incomplete data and thus may also be used for missing value estimation. When doing PCA one assumes that the data is restricted to a subspace of lower dimensionality, e.g. correlation patterns between jointly regulated genes. PCA aims to extract these structures thereby filtering noise out. If only the most significant loadings (eigenvectors, also referred to as principal components) are used for projection this can be written as: \begin{equation} X = 1\times{}\bar{x}^T + TP^T + V \end{equation} Where the term $1\times{}\bar{x}^T$ represents the original variable averages, $X$ denotes the observations, $T={t_1, t_2,\ldots,t_k}$ the latent variables or scores, $P={p_1, p_2,\ldots,p_k}$ the transformation matrix (consisting of the most significant eigenvectors of the covariance matrix) and $V$ are the residuals. Missing values may be estimated by projecting the scores back into the original space using $\hat{X} = 1\times{}\bar{x}^T + TP^T$. Optimally, this produces an estimate of the missing data based on the underlying correlation structure, thereby ignoring noise. This will only produce reasonable results if the residuals $V$ are sufficiently small, implying that most of the important information is captured by the first $k$ components. In order to calculate the transformation matrix $P$ one needs to determine the covariance matrix between variables or alternatively calculate $P$ directly via SVD. In both cases, this can only be done on complete matrices. However, an approximation may be obtained by use of different regression methods. The PCA methods provided in this package implement algorithms to accurately estimate the PCA solution on incomplete data. Although the focus of this package is clearly to provide a collection of PCA methods we also provide a cluster based method for missing value imputation. This allows to better rate and compare the results. \section{Algorithms} All methods return a common class called \texttt{pcaRes} as a container for the results. This guarantees maximum flexibility for the user. A wrapper function called \texttt{pca()} is provided that receives the desired type of pca as a string. \subsection*{svdPca} This is a wrapper function for $R's$ standard \texttt{prcomp} function. It delivers the results as a \texttt{pcaRes} object for compatibility with the rest of the package. \subsection*{svdImpute} This implements the SVDimpute algorithm as proposed by Troyanskaya et~al \cite{troyanskaya01}. The idea behind the algorithm is to estimate the missing values as a linear combination of the $k$ most significant eigengenes\footnote{The term ``eigengenes'' denotes the loadings when PCA was applied considering variables (here the genes) as observations.}. The algorithm works iteratively until the change in the estimated solution falls below a certain threshold. Each step the eigengenes of the current estimate are calculated and used to determine a new estimate. An optimal linear combination is found by regressing an incomplete variable against the $k$ most significant eigengenes. If the value at position $j$ is missing, the $j^{th}$ value of the eigengenes is not used when determining the regression coefficients.\\ SVDimpute seems to be tolerant to relatively high amount of missing data (> 10\%). \subsection*{Probabilistic PCA (ppca)} Probabilistic PCA combines an EM approach for PCA with a probabilistic model. The EM approach is based on the assumption that the latent variables as well as the noise are normal distributed. In standard PCA data which is far from the training set but close to the principal subspace may have the same reconstruction error, see Figure \ref{fig:pcaSubspace} for explanation. <>= library(pcaMethods) x <- c(-4,7); y <- c(-3,4) distX <- rnorm(100, sd=0.3)*3 distY <- rnorm(100, sd=0.3) + distX * 0.3 mat <- cbind(distX, distY) res <- pca(mat, nPcs=2, method="svd", center=F) loading <- loadings(res)[1,] grad <- loading[2] / loading[1] if (grad < 0) grad <- grad * -1 lx <- c(-4,7) ly <- c(grad * -4, grad * 7) @ \begin{figure} \centering <>= par(mar=c(2, 3, 2, 2)) plot(x,y, type="n", xlab="", ylab="") abline(v=0, col="dark gray", lwd = 2); abline(h=0, col = "dark gray", lwd = 2) points(distX, distY, type = 'p', col = "blue") lines(lx,ly, lwd = 2) points(-1, -1 * grad + 0.5, pch = 19, col = "red", lwd=4) points(6, 6 * grad + 0.5, pch = 19, col = "red", lwd=4) @ \caption{Normal distributed data with the first loading plotted in black. The two red points have the same reconstruction error because PCA does not define a density model. Thus the only measure of how well new data fits the model is the distance from the principal subspace. Data points far from the bulk of data but still close to the principal subspace will have a low reconstruction error. \label{fig:pcaSubspace}} \end{figure} PPCA defines a likelihood function such that the likelihood for data far from the training set is much lower, even if they are close to the principal subspace. This allows to improve the estimation accuracy.\\ PPCA is tolerant to amounts of missing values between 10\% to 15\%. If more data is missing the algorithm is likely not to converge to a reasonable solution. The method was implemented after the draft ``\texttt{EM Algorithms for PCA and Sensible PCA}'' written by Sam Roweis and after the Matlab \texttt{ppca} script implemented by \emph{Jakob Verbeek}\footnote{\url{http://lear.inrialpes.fr/~verbeek/}}. Please check also the PPCA help file. \subsection*{Bayesian PCA (bpca)} Similar to probabilistic PCA, Bayesian PCA uses an EM approach together with a Bayesian model to calculate the likelihood for a reconstructed value.\\ The algorithm seems to be tolerant to relatively high amounts of missing data (> 10\%). Scores and loadings obtained with Bayesian PCA slightly differ from those obtained with conventional PCA. This is because BPCA was developed especially for missing value estimation and is based on a variational Bayesian framework (VBF), with automatic relevance determination (ARD). In BPCA, ARD leads to a different scaling of the scores, loadings and eigenvalues when compared to standard PCA or PPCA. The algorithm does not force orthogonality between loadings. However, the authors of the BPCA paper found that including an orthogonality criterion made the predictions worse. They also state that the difference between ``real'' and predicted Eigenvalues becomes larger when the number of observation is smaller, because it reflects the lack of information to accurately determine true loadings from the limited and noisy data. As a result, weights of factors to predict missing values are not the same as with conventional PCA, but the missing value estimation is improved. BPCA was proposed by Oba et~al \cite{oba03}. The method available in this package is a port of the \texttt{bpca} Matlab script also provided by the authors\footnote{ \url{http://hawaii.aist-nara.ac.jp/\%7Eshige-o/tools/}}. \subsection*{Inverse non-linear PCA (NLPCA)} NLPCA \cite{scholz05} is especially suitable for data from experiments where the studied response is non-linear. Examples of such experiments are ubiquitous in biology -- enzyme kinetics are inherently non-linear as are gene expression responses influenced by the cell cycle or diurnal oscillations. NLPCA is based on training an auto-associative neural network composed of a component layer which serves as the ``bottle-neck'', a hidden non-linear layer and an output layer corresponding to the reconstructed data. The loadings can be seen as hidden in the network. Missing values in the training data are simply ignored when calculating the error during back-propagation. Thus NLPCA can be used to impute missing values in the same way as for conventional PCA. The only difference is that the loadings $P$ are now represented by a neural network.\\ A shortcoming of the current implementation is that there is no reasonable stop criterion. The quality of the estimated solution depends on the number of iterations. This should in most cases be somewhat between 500 and 1500. We recommend to use \texttt{kEstimate} or \texttt{kEstimateFast} to determine this parameter. \subsection*{Nipals PCA} Nipals (Nonlinear Estimation by Iterative Partial Least Squares) \cite{wold66} is an algorithm at the root of PLS regression which can execute PCA with missing values by simply leaving those out from the appropriate inner products. It is tolerant to small amounts (generally not more than 5\%) of missing data. \subsection{Local least squares (LLS) imputation} The package provides an algorithm called \texttt{llsImpute} for missing value estimation based on a linear combination of the $k$ nearest neighbours of an incomplete variable (in Microarray experiments normally a gene). The distance between variables is defined as the absolute value of the Pearson, Spearman or Kendall correlation coefficient. The optimal linear combination is found by solving a local least squares problem as described in \cite{kim05}. In tests performed in the cited paper the llsImpute algorithm is able to outperform knnImpute\cite{troyanskaya01} and competes well with BPCA. In the current implementation two slightly different ways for missing value estimation are provided. The first one is to restrict the neighbour searching to the subset of complete variables. This is preferable when the number of incomplete variables is relatively small. The second way is to consider all variables as candidates. Here, missing values are initially replaced by the columns wise mean. The method then iterates, using the current estimate as input for the LLS regression until the change between new and old estimate falls below a certain threshold (0.001). \section{Getting started} \paragraph{Installing the package.} To install the package first download the appropriate file for your platform from the Bioconductor website (\url{http://www.bioconductor.org/}). For Windows, start \texttt{R} and select the \texttt{Packages} menu, then \texttt{Install package from local zip file}. Find and highlight the location of the zip file and click on \texttt{open}. For Linux/Unix, use the usual command \texttt{R CMD INSTALL} or set the option \texttt{CRAN} to your nearest mirror site and use the command \texttt{install.packages} from within an \texttt{R} session. \paragraph{Loading the package:} To load the \texttt{pcaMethods} package in your \texttt{R} session, type \texttt{library(pcaMethods)}. \paragraph{Help files:} Detailed information on \texttt{pcaMethods} package functions can be obtained from the help files. For example, to get a description of \texttt{bpca} type \texttt{help("bpca")}. \paragraph{Sample data:} Two sample data sets are coming with the package. \texttt{metaboliteDataComplete} contains a complete subset from a larger metabolite data set. \texttt{metaboliteData} is the same data set but with 10 \% values removed from an equal distribution. \section{Some examples} <>= library(lattice) library(pcaMethods) @ To load the package and the two sample data sets type: <>= library(pcaMethods) data(metaboliteData) data(metaboliteDataComplete) @ Now centre the data <<>>= md <- prep(metaboliteData, scale="none", center=TRUE) mdC <- prep(metaboliteDataComplete, scale="none", center=TRUE) @ Run SVD pca, PPCA, BPCA, SVDimpute and nipalsPCA on the data, using the \texttt{pca()} wrapper function. The result is always a \texttt{pcaRes} object. <>= resPCA <- pca(mdC, method="svd", center=FALSE, nPcs=5) resPPCA <- pca(md, method="ppca", center=FALSE, nPcs=5) resBPCA <- pca(md, method="bpca", center=FALSE, nPcs=5) resSVDI <- pca(md, method="svdImpute", center=FALSE, nPcs=5) resNipals <- pca(md, method="nipals", center=FALSE, nPcs=5) resNLPCA <- pca(md, method="nlpca", center=FALSE, nPcs=5, maxSteps=300) @ Figure \ref{fig:eigenvalues} shows a plot of the eigenvalue structure (\texttt{sDev(pcaRes)}). If most of the variance is captured with few loadings PCA is likely to produce good missing value estimation results. For the sample data all methods show similar eigenvalues. One can also see that most of the variance is already captured by the first loading, thus estimation is likely to work fine on this data. For BPCA, the eigenvalues are scaled differently for reasons discussed above, see Figure \ref{fig:loadingBPCA}. The order of the loadings remains the same. \begin{figure} \centering <>= sDevs <- cbind(sDev(resPCA), sDev(resPPCA), sDev(resBPCA), sDev(resSVDI), sDev(resNipals), sDev(resNLPCA)) matplot(sDevs, type = 'l', xlab="Eigenvalues", ylab="Standard deviation of PC", lwd=3) legend(x="topright", legend=c("PCA", "PPCA", "BPCA", "SVDimpute","Nipals PCA","NLPCA"), lty=1:6, col=1:6, lwd=3) @ \caption{Eigenvalue structure as obtained with different methods\label{fig:eigenvalues}} \end{figure} To get an impression of the correctness of the estimation it is a good idea to plot the scores / loadings obtained with classical PCA and one of the probabilistic methods against each other. This of course requires a complete data set from which data is randomly removed. Figure \ref{fig:loadingBPCA} shows this for BPCA on the sample data. \begin{figure} \centering <>= par(mfrow=c(1,2)) plot(loadings(resBPCA)[,1], loadings(resPCA)[,1], xlab="BPCA", ylab="classic PCA", main = "Loading 1") plot(loadings(resBPCA)[,2], loadings(resPCA)[,2], xlab="BPCA", ylab="classic PCA", main = "Loading 2") @ \caption{Loading 1 and 2 calculated with BPCA plotted against those calculated with standard PCA. \label{fig:loadingBPCA}} \end{figure} \section{Cross validation} \texttt{Q2} is the goodness measure used for internal cross validation. This allows to estimate the level of structure in a data set and to optimise the choice of number of loadings. Cross validation is performed by removing random elements of the data matrix, then estimating these using the PCA algorithm of choice and then calculating $Q^2$ accordingly. At the moment, cross-validation can only be performed with algorithms that allow missing values (i.e. not SVD). Missing value independent cross-validation is scheduled for implementation in later versions. $Q^2$ is defined as following for the mean centered data (and possibly scaled) matrix $X$. $$\mathrm{SSX}=\sum (x_{ij})^2$$ $$\mathrm{PRESS}=\sum (x_{ij} - \hat{x}_{ij})^2$$ $$Q^2=1 - \mathrm{PRESS}/\mathrm{SSX}$$ The maximum value for $Q^2$ is thus 1 which means that all variance in $X$ is represented in the predictions; $X=\hat{X}$. <>= q2SVDI <- Q2(resSVDI, mdC, fold=10) q2PPCA <- Q2(resPPCA, mdC, fold=10) @ <>= # PPCA does not converge / misestimate a value in very rare cases. # This is a workaround to avoid that such a case will break the # diagram displayed in the vignette. # From the 2.0 release of bioconductor on, the convergence threshold # for PPCA was lowert to 1e-5, this should make the method much more # stable. So this workaround might be obsolete now... # [nope it is not, ppca is unstable] while( sum((abs(q2PPCA)) > 1) >= 1 ) { q2PPCA <- Q2(resPPCA, mdC, fold=10) } @ \begin{figure}[!ht] \centering <>= q2 <- data.frame(Q2=c(drop(q2PPCA), drop(q2SVDI)), method=c("PPCA", "SVD-Impute")[gl(2, 5)], PC=rep(1:5, 2)) print(xyplot(Q2~PC|method, q2, ylab=expression(Q^2), type="h", lwd=4)) @ \caption{Boxplot of the \texttt{Q2} results for BPCA, Nipals PCA, SVDimpute and PPCA. PPCA and SVDimpute both deliver better results than BPCA and Nipals in this example.\label{fig:Q2}} \end{figure} The second method called \texttt{kEstimate} uses cross validation to estimate the optimal number of loadings for missing value estimation. The \texttt{NRMSEP} (normalised root mean square error of prediction) \cite{feten05} or Q2 can be used to define the average error of prediction. The NRMSEP normalises the square difference between real and estimated values for a certain variable by the variance within this variable. The idea behind this normalisation is that the error of prediction will automatically be higher if the variance is higher. The \texttt{NRMSEP} for mean imputation is $\sqrt{\frac{nObs}{nObs - 1}}$ when cross validation is used, where $nObs$ is the number of observations. The exact definition is: \begin{equation} NRMSEP_k = \sqrt{\frac{1}{g} \sum_{j \in G} \frac{\sum_{i \in O_j} (x_{ij} - \hat{x}_{ijk})^2}{o_j s_{x_j}^2}} \end{equation} where $s^2_{x_j} = \sum_{i=1}^n (x_{ij} - \overline{x}_j)^2 / (n - 1)$, this is the variance within a certain variable. Further, $G$ denotes the set of incomplete variables, $g$ is the number of incomplete varialbes. $O_j$ is the set of missing observations in variable $j$ and $o_j$ is the number of missing observations in variable $j$. $\hat{x}_{ijk}$ stands for the estimate of value $i$ of variable $j$ using $k$ loadings. See Figure \ref{fig:kEstimate} for an example. The NRMSEP should be the error measure of choice. But if the number of observations is small, the variance within a certain variable may become and unstable criterion. If so or if variance scaling was applied we recommend to use Q2 instead. <>= errEsti <- kEstimate(md, method = "ppca", evalPcs=1:5, nruncv=1, em="nrmsep") @ \begin{figure}[!ht] \centering \begin{minipage}[c]{0.6\textwidth} \centering <>= barplot(drop(errEsti$eError), xlab="Loadings", ylab="NRMSEP (Single iteration)") @ \end{minipage} \begin{minipage}[c]{0.3\textwidth} \caption{Boxplot showing the \texttt{NRMSEP} versus the number of loadings. In this example only 1 iteration of the whole cross validation were performed. It is normally advisable to do more than just one iteration. \label{fig:kEstimate}} \end{minipage} \end{figure} \texttt{kEstimate} also provides information about the estimation error for individual variables. The $Q^2$ distance or the NRMSEP are calculated separately for each variable. See the manpage for \texttt{kEstimate} and \texttt{kEstimateFast} for details. Plotting the variable - wise results gives information about for which variables missing value estimation makes sense, and for which no imputation or mean imputation is preferable, see Figure \ref{fig:variableWiseError}. If you are not interested in variable - wise information we recommend to use the faster \texttt{kEstimateFast} instead. \begin{figure}[!ht] \centering \begin{minipage}[c]{0.6\textwidth} \centering <>= barplot(drop(errEsti$variableWiseError[, which(errEsti$evalPcs == errEsti$bestNPcs)]), xlab="Incomplete variable Index", ylab="NRMSEP") @ \end{minipage} \begin{minipage}[c]{0.3\textwidth} \caption{Boxplot showing the \texttt{NRMSEP} for all incomplete variables in the data set. For the first 7 variables missing value imputation does not seem to make too much sense. \label{fig:variableWiseError}} \end{minipage} \end{figure} \newpage \section{Visualisation of the results} \subsection{Quick scores and loadings plot} Some methods for display of scores and loadings are also provided. The function \texttt{slplot()} aims to be a simple way to quickly visualise scores and loadings in an intuitive way, see Figure \ref{fig:slplot}. Barplots are provided when plotting only one PC and colours can be specified differently for the scores and loadings plots. For a more specific scatter plot it is however recommended to access scores and loadings slots and define own plot functions. \begin{figure}[!h] \centering <>= slplot(resPCA) @ \caption{\texttt{slplot} for scores and loadings obtained with classical SVD based PCA. \label{fig:slplot}} \end{figure} \noindent Another method called \texttt{plotPcs()} allows to visualise many PCs plotted against each other, see Figure \ref{fig:plotPcs}. \begin{figure}[!ht] \centering <>= plotPcs(resPPCA, pc=1:3, type="score") @ \caption{A plot of score 1:3 for PPCA created with \texttt{plotPcs()} \label{fig:plotPcs}} \end{figure} \subsection{Using ggplot2} For using ggplot, the scores and loadings should best be added to a data frame that add other relevant descriptive factors. For example, after doing PCA on the Iris dataset, we may add the scores back to the original data frame and use ggplot to visualise, see Figure \ref{fig:ggplot}. \begin{figure}[!ht] \centering <>= pc <- pca(iris) irdf <- merge(iris, scores(pc), by=0) library(ggplot2) ggplot(irdf, aes(PC1, PC2, colour=Species)) + geom_point() + stat_ellipse() @ \caption{Score plot using ggplot2} \label{fig:ggplot} \end{figure} \cleardoublepage \begin{thebibliography}{2006} \bibitem{stacklies07} Stacklies W., Redestig H., Scholz M., and Walther D., and Selbig J. {\sl pcaMethods -- a Bioconductor package providing PCA methods for incomplete data} Bioinformatics. 2007, 23, 1164-1167. {\sl Non-linear PCA: a missing data approach.} Bioinformatics. 2005, 21, 3887-3895. \bibitem{scholz05} Scholz, M. , Kaplan, F., Guy, C.L., Kopka, J. and Selbig, J. {\sl Non-linear pca: a missing data approach.} Bioinformatics. 2005, 21, 3887-3895. \bibitem{troyanskaya01} Troyanskaya O. and Cantor M. and Sherlock G. and Brown P. and Hastie T. and Tibshirani R. and Botstein D. and Altman RB. {\sl Missing value estimation methods for DNA microarrays.} Bioinformatics. 2001 Jun;17(6):520-525. \bibitem{feten05} Feten G. and Almoy T. and Aastveit A.H. {\sl Prediction of Missing Values in Microarray and Use of Mixed Models to Evaluate the Predictors.}, Stat. Appl. Genet. Mol. Biol. 2005;4(1):Article 10 \bibitem{oba03} Oba S. and Sato MA. and Takemasa I. and Monden M. and Matsubara K. and Ishii S. {\sl A Bayesian missing value estimation method for gene expression profile data.} Bioinformatics. 2003 Nov 1;19(16):2088-96. \bibitem{wold66} Wold H. {Estimation of principal components and related models by iterative least squares.} In Multivariate Analysis (Ed. P.R. Krishnaiah), Academic Press, NY, 391-420. \bibitem{kim05} Kim H. and Golub G.H. and Park H. {\sl Missing value estimation for DNA microarray gene expression data: local least squares imputation} Bioinformatics. 2005 21(2) :187-198 \end{thebibliography} \end{document}