ggbeeswarm/ 0000755 0001762 0000144 00000000000 14423307262 012400 5 ustar ligges users ggbeeswarm/NAMESPACE 0000644 0001762 0000144 00000000344 14346724045 013626 0 ustar ligges users # Generated by roxygen2: do not edit by hand
export(geom_beeswarm)
export(geom_quasirandom)
export(position_beeswarm)
export(position_quasirandom)
import(ggplot2)
importFrom(beeswarm,swarmx)
importFrom(vipor,offsetSingleGroup)
ggbeeswarm/man/ 0000755 0001762 0000144 00000000000 14423030634 013147 5 ustar ligges users ggbeeswarm/man/geom_beeswarm.Rd 0000644 0001762 0000144 00000012070 14347131302 016251 0 ustar ligges users % Generated by roxygen2: do not edit by hand
% Please edit documentation in R/geom-beeswarm.R
\name{geom_beeswarm}
\alias{geom_beeswarm}
\title{Points, jittered to reduce overplotting using the beeswarm package}
\usage{
geom_beeswarm(
mapping = NULL,
data = NULL,
stat = "identity",
...,
method = "swarm",
cex = 1,
side = 0L,
priority = "ascending",
fast = TRUE,
dodge.width = NULL,
corral = "none",
corral.width = 0.9,
groupOnX = NULL,
beeswarmArgs = list(),
na.rm = FALSE,
show.legend = NA,
inherit.aes = TRUE
)
}
\arguments{
\item{mapping}{Set of aesthetic mappings created by \code{\link[ggplot2:aes]{aes()}}. If specified and
\code{inherit.aes = TRUE} (the default), it is combined with the default mapping
at the top level of the plot. You must supply \code{mapping} if there is no plot
mapping.}
\item{data}{The data to be displayed in this layer. There are three
options:
If \code{NULL}, the default, the data is inherited from the plot
data as specified in the call to \code{\link[ggplot2:ggplot]{ggplot()}}.
A \code{data.frame}, or other object, will override the plot
data. All objects will be fortified to produce a data frame. See
\code{\link[ggplot2:fortify]{fortify()}} for which variables will be created.
A \code{function} will be called with a single argument,
the plot data. The return value must be a \code{data.frame}, and
will be used as the layer data. A \code{function} can be created
from a \code{formula} (e.g. \code{~ head(.x, 10)}).}
\item{stat}{The statistical transformation to use on the data for this
layer, either as a \code{ggproto} \code{Geom} subclass or as a string naming the
stat stripped of the \code{stat_} prefix (e.g. \code{"count"} rather than
\code{"stat_count"})}
\item{...}{Other arguments passed on to \code{\link[ggplot2:layer]{layer()}}. These are
often aesthetics, used to set an aesthetic to a fixed value, like
\code{colour = "red"} or \code{size = 3}. They may also be parameters
to the paired geom/stat.}
\item{method}{Method for arranging points (see Details below)}
\item{cex}{Scaling for adjusting point spacing (see \code{\link[beeswarm:swarmx]{beeswarm::swarmx()}}).
Values between 1 (default) and 3 tend to work best.}
\item{side}{Direction to perform jittering: 0: both directions;
1: to the right or upwards; -1: to the left or downwards.}
\item{priority}{Method used to perform point layout (see Details below)}
\item{fast}{Use compiled version of swarm algorithm? This option is ignored
for all methods expect \code{"swarm"} and \code{"compactswarm"}.}
\item{dodge.width}{Amount by which points from different aesthetic groups
will be dodged. This requires that one of the aesthetics is a factor.}
\item{corral}{\code{string}. Method used to adjust points that would be placed to
wide horizontally, default is \code{"none"}. See details below.}
\item{corral.width}{\code{numeric}. Width of the corral, default is \code{0.9}.}
\item{groupOnX}{\ifelse{html}{\href{https://lifecycle.r-lib.org/articles/stages.html#deprecated}{\figure{lifecycle-deprecated.svg}{options: alt='[Deprecated]'}}}{\strong{[Deprecated]}} No longer needed.}
\item{beeswarmArgs}{\ifelse{html}{\href{https://lifecycle.r-lib.org/articles/stages.html#deprecated}{\figure{lifecycle-deprecated.svg}{options: alt='[Deprecated]'}}}{\strong{[Deprecated]}} No longer used.}
\item{na.rm}{If \code{FALSE}, the default, missing values are removed with
a warning. If \code{TRUE}, missing values are silently removed.}
\item{show.legend}{logical. Should this layer be included in the legends?
\code{NA}, the default, includes if any aesthetics are mapped.
\code{FALSE} never includes, and \code{TRUE} always includes.
It can also be a named logical vector to finely select the aesthetics to
display.}
\item{inherit.aes}{If \code{FALSE}, overrides the default aesthetics,
rather than combining with them. This is most useful for helper functions
that define both data and aesthetics and shouldn't inherit behaviour from
the default plot specification, e.g. \code{\link[ggplot2:borders]{borders()}}.}
}
\description{
The beeswarm geom is a convenient means to offset points within categories to
reduce overplotting. Uses the beeswarm package
}
\section{Aesthetics}{
\Sexpr[results=rd,stage=build]{ggplot2:::rd_aesthetics("geom", "point")}
}
\examples{
ggplot2::qplot(class, hwy, data = ggplot2::mpg, geom='beeswarm')
# Generate fake data
distro <- data.frame(
'variable'=rep(c('runif','rnorm'),each=100),
'value'=c(runif(100, min=-3, max=3), rnorm(100))
)
ggplot2::qplot(variable, value, data = distro, geom='beeswarm')
ggplot2::ggplot(distro,aes(variable, value)) +
geom_beeswarm(priority='density',size=2.5)
}
\seealso{
\code{\link[=geom_quasirandom]{geom_quasirandom()}} an alternative method,
\code{\link[beeswarm:swarmx]{beeswarm::swarmx()}} how spacing is determined,
\code{\link[ggplot2:geom_point]{ggplot2::geom_point()}} for regular, unjittered points,
\code{\link[ggplot2:geom_jitter]{ggplot2::geom_jitter()}} for jittered points,
\code{\link[ggplot2:geom_boxplot]{ggplot2::geom_boxplot()}} for another way of looking at the conditional
distribution of a variable
}
ggbeeswarm/man/geom_quasirandom.Rd 0000644 0001762 0000144 00000011462 14423120633 016773 0 ustar ligges users % Generated by roxygen2: do not edit by hand
% Please edit documentation in R/geom-quasirandom.R
\name{geom_quasirandom}
\alias{geom_quasirandom}
\title{Points, jittered to reduce overplotting using the vipor package}
\usage{
geom_quasirandom(
mapping = NULL,
data = NULL,
stat = "identity",
...,
method = "quasirandom",
width = NULL,
varwidth = FALSE,
bandwidth = 0.5,
nbins = NULL,
dodge.width = NULL,
groupOnX = NULL,
orientation = NULL,
na.rm = FALSE,
show.legend = NA,
inherit.aes = TRUE
)
}
\arguments{
\item{mapping}{Set of aesthetic mappings created by \code{\link[ggplot2:aes]{aes()}}. If specified and
\code{inherit.aes = TRUE} (the default), it is combined with the default mapping
at the top level of the plot. You must supply \code{mapping} if there is no plot
mapping.}
\item{data}{The data to be displayed in this layer. There are three
options:
If \code{NULL}, the default, the data is inherited from the plot
data as specified in the call to \code{\link[ggplot2:ggplot]{ggplot()}}.
A \code{data.frame}, or other object, will override the plot
data. All objects will be fortified to produce a data frame. See
\code{\link[ggplot2:fortify]{fortify()}} for which variables will be created.
A \code{function} will be called with a single argument,
the plot data. The return value must be a \code{data.frame}, and
will be used as the layer data. A \code{function} can be created
from a \code{formula} (e.g. \code{~ head(.x, 10)}).}
\item{stat}{The statistical transformation to use on the data for this
layer, either as a \code{ggproto} \code{Geom} subclass or as a string naming the
stat stripped of the \code{stat_} prefix (e.g. \code{"count"} rather than
\code{"stat_count"})}
\item{...}{Other arguments passed on to \code{\link[ggplot2:layer]{layer()}}. These are
often aesthetics, used to set an aesthetic to a fixed value, like
\code{colour = "red"} or \code{size = 3}. They may also be parameters
to the paired geom/stat.}
\item{method}{the method used for distributing points
(quasirandom, pseudorandom, smiley, maxout, frowney, minout, tukey, tukeyDense).
See \code{\link[vipor:offsetX]{vipor::offsetSingleGroup()}} for the details of each method.}
\item{width}{the maximum amount of spread (default: 0.4)}
\item{varwidth}{vary the width by the relative size of each group}
\item{bandwidth}{the bandwidth adjustment to use when calculating density
Smaller numbers (< 1) produce a tighter "fit". (default: 0.5)}
\item{nbins}{the number of bins used when calculating density
(has little effect with quasirandom/random distribution)}
\item{dodge.width}{Amount by which points from different aesthetic groups
will be dodged. This requires that one of the aesthetics is a factor.
To disable dodging between groups, set this to NULL.}
\item{groupOnX}{\ifelse{html}{\href{https://lifecycle.r-lib.org/articles/stages.html#superseded}{\figure{lifecycle-superseded.svg}{options: alt='[Superseded]'}}}{\strong{[Superseded]}} See \code{orientation}.}
\item{orientation}{The orientation (i.e., which axis to group on) is inferred from the data.
This can be overridden by setting \code{orientation} to either \code{"x"} or \code{"y"}.}
\item{na.rm}{If \code{FALSE}, the default, missing values are removed with
a warning. If \code{TRUE}, missing values are silently removed.}
\item{show.legend}{logical. Should this layer be included in the legends?
\code{NA}, the default, includes if any aesthetics are mapped.
\code{FALSE} never includes, and \code{TRUE} always includes.
It can also be a named logical vector to finely select the aesthetics to
display.}
\item{inherit.aes}{If \code{FALSE}, overrides the default aesthetics,
rather than combining with them. This is most useful for helper functions
that define both data and aesthetics and shouldn't inherit behaviour from
the default plot specification, e.g. \code{\link[ggplot2:borders]{borders()}}.}
}
\description{
The quasirandom geom is a convenient means to offset points within categories
to reduce overplotting. Uses the vipor package
}
\section{Aesthetics}{
\Sexpr[results=rd,stage=build]{ggplot2:::rd_aesthetics("geom", "point")}
}
\examples{
ggplot2::qplot(class, hwy, data = ggplot2::mpg, geom='quasirandom')
# Generate fake data
distro <- data.frame(
'variable'=rep(c('runif','rnorm'),each=100),
'value'=c(runif(100, min=-3, max=3), rnorm(100))
)
ggplot2::qplot(variable, value, data = distro, geom = 'quasirandom')
ggplot2::ggplot(distro,aes(variable, value)) + geom_quasirandom(width=0.1)
}
\seealso{
\code{\link[vipor:offsetX]{vipor::offsetSingleGroup()}} how spacing is determined,
\code{\link[ggplot2:geom_point]{ggplot2::geom_point()}} for regular, unjittered points,
\code{\link[ggplot2:geom_jitter]{ggplot2::geom_jitter()}} for jittered points,
\code{\link[=geom_boxplot]{geom_boxplot()}} for another way of looking at the conditional
distribution of a variable
}
ggbeeswarm/man/offset_beeswarm.Rd 0000644 0001762 0000144 00000007622 14346724045 016632 0 ustar ligges users % Generated by roxygen2: do not edit by hand
% Please edit documentation in R/position-beeswarm.R
\name{offset_beeswarm}
\alias{offset_beeswarm}
\title{An internal function to calculate new positions for geom_beeswarm}
\usage{
offset_beeswarm(
data,
yLim.expand,
xRange,
yRange,
method = "swarm",
cex = 1,
side = 0L,
priority = "ascending",
fast = TRUE,
corral = "none",
corral.width = 0.2
)
}
\arguments{
\item{data}{A data.frame containing plotting data in columns x and y.
Usually obtained from data processed by ggplot2.}
\item{yLim.expand}{y data limits plus a small expansion using \code{grDevices::extendrange}}
\item{xRange}{x axis scale range}
\item{yRange}{y axis scale range}
\item{method}{Method for arranging points (see Details below)}
\item{cex}{Scaling for adjusting point spacing (see \code{\link[beeswarm:swarmx]{beeswarm::swarmx()}}).
Values between 1 (default) and 3 tend to work best.}
\item{side}{Direction to perform jittering: 0: both directions;
1: to the right or upwards; -1: to the left or downwards.}
\item{priority}{Method used to perform point layout (see Details below)}
\item{fast}{Use compiled version of swarm algorithm? This option is ignored
for all methods expect \code{"swarm"} and \code{"compactswarm"}.}
\item{corral}{\code{string}. Method used to adjust points that would be placed to
wide horizontally, default is \code{"none"}. See details below.}
\item{corral.width}{\code{numeric}. Width of the corral, default is \code{0.9}.}
}
\description{
An internal function to calculate new positions for geom_beeswarm
}
\details{
\strong{method:} specifies the algorithm used to avoid overlapping points. The
default \code{"swarm"} method places points in increasing order. If a point would
overlap with an existing point, it is shifted sideways (along the group axis)
by a minimal amount sufficient to avoid overlap.
While the \code{"swarm"} method places points in a predetermined
order, the \code{"compactswarm"} method uses a greedy strategy to determine which
point will be placed next. This often leads to a more tightly-packed layout.
The strategy is very simple: on each iteration, a point that can be placed as
close as possible to the non-data axis is chosen and placed. If there are two
or more equally good points, \code{priority} is used to break ties.
The other 3 methods first discretise the values along the data axis, in order
to create more efficient packing. The \code{"square"} method places points on a
square grid, whereas \code{"hex"} uses a hexagonal grid. \code{"centre"}/\code{"center"}
uses a square grid to produce a symmetric swarm. The number of break points
for discretisation is determined by a combination of the available plotting
area and the \code{cex} argument.
\strong{priority:} controls the order in which points are placed, which generally
has a noticeable effect on the plot appearance. \code{"ascending"} gives the
'traditional' beeswarm plot. \code{"descending"} is the opposite. \code{"density"}
prioritizes points with higher local density. \code{"random"} places points in a
random order. \code{"none"} places points in the order provided.
\strong{corral:} By default, swarms from different groups are not prevented from
overlapping, i.e. \verb{"corral = "none"}. Thus, datasets that are very large or
unevenly distributed may produce ugly overlapping beeswarms. To control
runaway points one can use the following methods. \code{"gutter"} collects runaway
points along the boundary between groups. \code{"wrap"} implement periodic boundaries.
\code{"random"} places runaway points randomly in the region. \code{"omit"} omits runaway
points.
}
\seealso{
\code{\link[=geom_beeswarm]{geom_beeswarm()}}, \code{\link[=position_quasirandom]{position_quasirandom()}},
\code{\link[beeswarm:swarmx]{beeswarm::swarmx()}}
Other position adjustments:
\code{\link{position_beeswarm}()},
\code{\link{position_quasirandom}()}
}
\concept{position adjustments}
\keyword{internal}
ggbeeswarm/man/position_beeswarm.Rd 0000644 0001762 0000144 00000007656 14346724045 017217 0 ustar ligges users % Generated by roxygen2: do not edit by hand
% Please edit documentation in R/position-beeswarm.R
\name{position_beeswarm}
\alias{position_beeswarm}
\title{Arrange points using the \verb{\link[beeswarm]} package.}
\usage{
position_beeswarm(
method = "swarm",
cex = 1,
side = 0L,
priority = "ascending",
fast = TRUE,
groupOnX = NULL,
dodge.width = 0,
corral = "none",
corral.width = 0.2
)
}
\arguments{
\item{method}{Method for arranging points (see Details below)}
\item{cex}{Scaling for adjusting point spacing (see \code{\link[beeswarm:swarmx]{beeswarm::swarmx()}}).
Values between 1 (default) and 3 tend to work best.}
\item{side}{Direction to perform jittering: 0: both directions;
1: to the right or upwards; -1: to the left or downwards.}
\item{priority}{Method used to perform point layout (see Details below)}
\item{fast}{Use compiled version of swarm algorithm? This option is ignored
for all methods expect \code{"swarm"} and \code{"compactswarm"}.}
\item{groupOnX}{\ifelse{html}{\href{https://lifecycle.r-lib.org/articles/stages.html#deprecated}{\figure{lifecycle-deprecated.svg}{options: alt='[Deprecated]'}}}{\strong{[Deprecated]}} No longer needed.}
\item{dodge.width}{Amount by which points from different aesthetic groups
will be dodged. This requires that one of the aesthetics is a factor.}
\item{corral}{\code{string}. Method used to adjust points that would be placed to
wide horizontally, default is \code{"none"}. See details below.}
\item{corral.width}{\code{numeric}. Width of the corral, default is \code{0.9}.}
}
\description{
Arrange points using the \verb{\link[beeswarm]} package.
}
\details{
\strong{method:} specifies the algorithm used to avoid overlapping points. The
default \code{"swarm"} method places points in increasing order. If a point would
overlap with an existing point, it is shifted sideways (along the group axis)
by a minimal amount sufficient to avoid overlap.
While the \code{"swarm"} method places points in a predetermined
order, the \code{"compactswarm"} method uses a greedy strategy to determine which
point will be placed next. This often leads to a more tightly-packed layout.
The strategy is very simple: on each iteration, a point that can be placed as
close as possible to the non-data axis is chosen and placed. If there are two
or more equally good points, \code{priority} is used to break ties.
The other 3 methods first discretise the values along the data axis, in order
to create more efficient packing. The \code{"square"} method places points on a
square grid, whereas \code{"hex"} uses a hexagonal grid. \code{"centre"}/\code{"center"}
uses a square grid to produce a symmetric swarm. The number of break points
for discretisation is determined by a combination of the available plotting
area and the \code{cex} argument.
\strong{priority:} controls the order in which points are placed, which generally
has a noticeable effect on the plot appearance. \code{"ascending"} gives the
'traditional' beeswarm plot. \code{"descending"} is the opposite. \code{"density"}
prioritizes points with higher local density. \code{"random"} places points in a
random order. \code{"none"} places points in the order provided.
\strong{corral:} By default, swarms from different groups are not prevented from
overlapping, i.e. \verb{"corral = "none"}. Thus, datasets that are very large or
unevenly distributed may produce ugly overlapping beeswarms. To control
runaway points one can use the following methods. \code{"gutter"} collects runaway
points along the boundary between groups. \code{"wrap"} implement periodic boundaries.
\code{"random"} places runaway points randomly in the region. \code{"omit"} omits runaway
points.
}
\seealso{
\code{\link[=geom_beeswarm]{geom_beeswarm()}}, \code{\link[=position_quasirandom]{position_quasirandom()}},
\code{\link[beeswarm:swarmx]{beeswarm::swarmx()}}
Other position adjustments:
\code{\link{offset_beeswarm}()},
\code{\link{position_quasirandom}()}
}
\concept{position adjustments}
ggbeeswarm/man/position_quasirandom.Rd 0000644 0001762 0000144 00000004075 14423120633 017712 0 ustar ligges users % Generated by roxygen2: do not edit by hand
% Please edit documentation in R/position-quasirandom.R
\name{position_quasirandom}
\alias{position_quasirandom}
\title{Arrange points using quasirandom noise to avoid overplotting}
\usage{
position_quasirandom(
method = "quasirandom",
width = NULL,
varwidth = FALSE,
bandwidth = 0.5,
nbins = NULL,
dodge.width = 0,
orientation = NULL,
groupOnX = NULL,
na.rm = FALSE
)
}
\arguments{
\item{method}{the method used for distributing points
(quasirandom, pseudorandom, smiley, maxout, frowney, minout, tukey, tukeyDense).
See \code{\link[vipor:offsetX]{vipor::offsetSingleGroup()}} for the details of each method.}
\item{width}{the maximum amount of spread (default: 0.4)}
\item{varwidth}{vary the width by the relative size of each group}
\item{bandwidth}{the bandwidth adjustment to use when calculating density
Smaller numbers (< 1) produce a tighter "fit". (default: 0.5)}
\item{nbins}{the number of bins used when calculating density
(has little effect with quasirandom/random distribution)}
\item{dodge.width}{Amount by which points from different aesthetic groups
will be dodged. This requires that one of the aesthetics is a factor.
To disable dodging between groups, set this to NULL.}
\item{orientation}{The orientation (i.e., which axis to group on) is inferred from the data.
This can be overridden by setting \code{orientation} to either \code{"x"} or \code{"y"}.}
\item{groupOnX}{\ifelse{html}{\href{https://lifecycle.r-lib.org/articles/stages.html#superseded}{\figure{lifecycle-superseded.svg}{options: alt='[Superseded]'}}}{\strong{[Superseded]}} See \code{orientation}.}
\item{na.rm}{if FALSE, the default, missing values are removed with a warning.
If TRUE, missing values are silently removed.}
}
\description{
Arrange points using quasirandom noise to avoid overplotting
}
\seealso{
\code{\link[vipor:offsetX]{vipor::offsetSingleGroup()}}, \code{\link[=geom_quasirandom]{geom_quasirandom()}}
Other position adjustments:
\code{\link{offset_beeswarm}()},
\code{\link{position_beeswarm}()}
}
\concept{position adjustments}
ggbeeswarm/man/figures/ 0000755 0001762 0000144 00000000000 14346724045 014625 5 ustar ligges users ggbeeswarm/man/figures/lifecycle-defunct.svg 0000644 0001762 0000144 00000001704 14346724045 020735 0 ustar ligges users ggbeeswarm/man/figures/lifecycle-maturing.svg 0000644 0001762 0000144 00000001706 14346724045 021135 0 ustar ligges users ggbeeswarm/man/figures/lifecycle-archived.svg 0000644 0001762 0000144 00000001707 14346724045 021075 0 ustar ligges users ggbeeswarm/man/figures/lifecycle-questioning.svg 0000644 0001762 0000144 00000001714 14346724045 021653 0 ustar ligges users ggbeeswarm/man/figures/lifecycle-superseded.svg 0000644 0001762 0000144 00000001713 14346724045 021450 0 ustar ligges users ggbeeswarm/man/figures/lifecycle-stable.svg 0000644 0001762 0000144 00000001674 14346724045 020565 0 ustar ligges users ggbeeswarm/man/figures/lifecycle-experimental.svg 0000644 0001762 0000144 00000001716 14346724045 022005 0 ustar ligges users ggbeeswarm/man/figures/lifecycle-deprecated.svg 0000644 0001762 0000144 00000001712 14346724045 021404 0 ustar ligges users ggbeeswarm/man/ggbeeswarm.Rd 0000644 0001762 0000144 00000001767 14347131302 015573 0 ustar ligges users % Generated by roxygen2: do not edit by hand
% Please edit documentation in R/ggbeeswarm-package.R
\docType{package}
\name{ggbeeswarm}
\alias{ggbeeswarm}
\title{ggbeeswarm extends ggplot2 with violin point/beeswarm plots}
\description{
This package allows plotting of several groups of one dimensional data as a violin point/beeswarm plot in ggplot2 by arranging data points to resemble the underlying distribution. The development version of this package is on \url{https://github.com/eclarke/ggbeeswarm}.
}
\examples{
ggplot2::ggplot(ggplot2::mpg,aes(class, hwy)) + geom_quasirandom()
# Generate fake data
distro <- data.frame(
'variable'=rep(c('runif','rnorm'),each=100),
'value'=c(runif(100, min=-3, max=3), rnorm(100))
)
ggplot2::ggplot(distro,aes(variable, value)) + geom_quasirandom()
ggplot2::ggplot(distro,aes(variable, value)) + geom_quasirandom(width=.1)
}
\seealso{
\code{\link[=position_quasirandom]{position_quasirandom()}}
}
\author{
Erik Clarke, \email{erikclarke@gmail.com}
}
ggbeeswarm/DESCRIPTION 0000644 0001762 0000144 00000002144 14423307262 014107 0 ustar ligges users Package: ggbeeswarm
Type: Package
Title: Categorical Scatter (Violin Point) Plots
Version: 0.7.2
Date: 2023-04-28
Authors@R: c(
person(given="Erik", family="Clarke", role=c("aut", "cre"), email="erikclarke@gmail.com"),
person(given="Scott", family="Sherrill-Mix", role=c("aut"), email="sherrillmix@gmail.com"),
person(given="Charlotte", family="Dawson", role=c("aut"), email="csdaw@outlook.com"))
Description: Provides two methods of plotting categorical scatter plots such
that the arrangement of points within a category reflects the density of
data at that region, and avoids over-plotting.
URL: https://github.com/eclarke/ggbeeswarm
BugReports: https://github.com/eclarke/ggbeeswarm/issues
Encoding: UTF-8
License: GPL (>= 3)
Depends: R (>= 3.5.0), ggplot2 (>= 3.3.0)
Imports: beeswarm, lifecycle, vipor, cli
Suggests: gridExtra
RoxygenNote: 7.2.2
NeedsCompilation: no
Packaged: 2023-04-29 20:56:12 UTC; erik
Author: Erik Clarke [aut, cre],
Scott Sherrill-Mix [aut],
Charlotte Dawson [aut]
Maintainer: Erik Clarke
Repository: CRAN
Date/Publication: 2023-04-29 21:40:02 UTC
ggbeeswarm/build/ 0000755 0001762 0000144 00000000000 14423302153 013471 5 ustar ligges users ggbeeswarm/build/vignette.rds 0000644 0001762 0000144 00000000503 14423302153 016026 0 ustar ligges users }R ic[U/__ڋ6l3Hb=VBQf o^DŽdIBmә[r7o 1PԊDwUF4 PP9c
AgTtѓY}Ȓ_*t O=2iiOLGQ )Fa Ҩ<4kܲs#h S0hr, ggbeeswarm/build/partial.rdb 0000644 0001762 0000144 00000034440 14423302150 015620 0 ustar ligges users }Is Gf $}ߜ( R2N q$V U@fd"ȈdD$iʬlOTu 0sӜzSmnskt?/2=7@S{@?wT*ՓIl'9(qܖM
pL_`L̳5湫kEӹ4.s܀y+ '_*(rIƾ{/={|eyf'']úUl߿5ko4o:97o9ņh(~?~C` F
?km7D|bnu.JH
5P[ɖ=kWbvΛ]
Ko
&__/W\1|v6({rvvb+7^y%`fEW -Nh7o2ѳo yba1EE5F}PW}Qlٶ?jž(! zo;Z4/ifhy7W)N`0ar.*
z#҄cW.?J~BB<43X}`Ԩ= |PZTw,2mCub(.T CJe; 26n W$j?"!jӗ~C%g2wʶ_d ,,FV6r B3 Vg`v+Cہ*$| q"! 2:`^b9 q=T3 jTdO$B lq~ɾT-nYv @sqOrw=ϰF 15*SjSp2x:9PVJZB+b;ΪW9;1u~R?:xl~s wsV*բl{']fjTԎTw1,gn0}nϱT.FϘ2I6:jrϤnmN f_m.ǥs2:SѤCi5ÐBܩTBcR̀ƍ+
||:ZРxfFb<:A9plbcw_߸emE54iF6O kM[9s Pͷj(y] )0l欂Upm X\[;5ծt*lA>
; F`r
6J3y`T`,ͬY)rK+p^FZ*>|Ul#lsٴe_g
+U̯˶rzLILwj~g` QrYmȷ2-QaIZ縢Tu(M_A["BiN2Ԧٌ& hR$m %۽KSÊΥ
O<54a |H_5ސ[3nca Ki/ އ|?v5^?JgU< IG`o;| ez^zR> mٕ Uu[\/pP*}"@V
&7疪ވ.iӱ,
_rE`ah(Ҽ "<\,MkqMyI>> A
>P4 Qxx ͣ v)a#ȏbsOFլ(π3gS|S0g)Aڄ={(vQk)xU. WSJn; # 19]|kxbxKC)SNzcPCܡ,^({؎d Sw'GzӐ:zr71Xэ썺p`>< d"}c pH
t 1V51CkdزYƢmt }͘:Ȕv~GkJh`EԦ~Q_oȼ^ ބ|3~u{,ph3p++f8yH "_h $k[uGx x vuxm4#H+>;¥]P.J>M2 yx+3 VU%4'fn i(֖EE)Uvm ANF[}!'x3:5sx
jfgl44o7X!wN' 712J2 fzY4YI^eFT9J4(;^FE㯗
60E.'.3 <
x.ujS4R4" rЃI3KQ&P[[v[bsWU$6' 2|H !MN%ēP>2ݒNjT5r@+Ƭ_YAhfVX xt1J8Nj0L_A
8Ee>&EjB%ov}D>S@u^(#s+U= ɵ9:t
x֎/
NA1+sozhGc
Q h,(/ǐ6V`)kn, d ou)1SOc ݠOL_C~ ѧn!0)? s̀G;>&Vb(lr>"Sȟ&\ۭxt3}Tv ܅(z^p)18<xb[_S̃:#;
-q<ڎ ]M| ？D]-1}D \/hC}DbNqӮ,vSL++ISf:ݓܖĥ)L4/ Iz)"3Z{]ϝ\
_6=ZG0TދLIK, R-[b> n*910RR@#JyX6 ⳼OU.ԁPr;mwd%ۆAߒOiL"}GA}
$ ".QLz
9/ 3Y[5N`nJuj IJ:ms T"m3(;DFVFJhGwg+Y%R' k-^IZ kf Q#&bŚm[FOgSՀ1ɟY K|_KWCiIgsvU\ ~YTS6J,c ǮC"~ "iB Fi ZZXRvі# mOQv#ȐblݮĶ S;Y9U&V' USrÐckp})Zއ|?xxN=Jr7*9]ͥ_Hm %JB<ݦ %aBKpّJ/6,{ꘂ hre2[fƻh[+f
mӧEPnͲg(NG(2teX5Ck3rۋ@as?Z~dv05* h/ "Ưo`+^Ɵ\/luK-= _V&N
K o &ύ5`ܲl*bwAs< s;֨-c"T(#S/!nxF w)2$0 瀘29 sv9 O!w9 g%X ; PrϿS1/C}7J?
_u9_I-
oYsTs
^$ CdJ %m(;*RK~ at+%-i\BD .2Wvӎ($dvOA(mץd99Uw{B:իs"Q9 .^qc·-ȷl HT.xxw]s܅|}%?-&-1~Cʤ ~
;-]"̰%v mİ].
sQ-e"oSYy><}$쇜m/a_U!JCt/ɻݢ;15n λw k5W :.Qu;j|#Ys1EDNR 7 *!kmQa`Vӆl^1j-)'y" Š FΖMAK
cr7m:;ҰV}3gbg T?.^lr3COi|C>t "TUO
C#lrJ\NA1J(ppl ]ϗ%9> fXqKG- ptWrL [jY22N ?FPx,vkk#G_@ ~cKƲEQ萋j~~<8\G{kxsⶁ4 /@;zMWC@N2TFNw
旀w!eiSZG|՚GJC3Z;DmդK~ւd'/ 较CɿOBߵPG"ORqXE@HG
g#T2ц2죥}KF [e sAc<̣]|bݖlNL_(d.:簒X|SlQEx>02VdaYH;eeStfn7擲 ecC۠/ D\' rԂ_Q#Io M#J=vg!OjPؽQ:m[TO-w(E cN94[ZhtddF\LS#цՑ1`Ox99ytLLV0ޙ2 |46c-r틂W%sTJ Mo#ˆg92WqHzs\i䖦& jW y_oIFQN8S/s4V.}4e%
9mSy {!ƦiS+B:Rm
B%N@Sȟ&Ռ(s1Sccce>qXlt4tj$o:>F42u)sU0%B&
,FlGOJJRv+( r+ĥ
Gd|mkJN-) A\ISWLWی9Aiȧu7}s K J
̇2 &"~E`zX&9U>M @жJG
䤀Ue^M@f!'J>|xvoC BB LO%we#GXO7R wMu&f ?K\לo?
ԎZ* Ax"5g IDנ΄ _jA C [ωY؆fp68 D5zI++ש04(a"UWu
`?< l"5k__Մ7!ߌGUk4 |Ej}^U1ToRaoPu R]52A D(7;@ oA[B
aV5PvwpkE+D6-䷱I'bE
M(I(FriŤ |-R^Fj?n %#! h֔fgvJ'Frvy.͈"F7V'
c7JoSrPWzL''ssyҎsX 2wz
9ouv+7zJ1lͪ/7a A>ꢉT^j.VO Z//HݨUbu%m. N3wA~[= ߢh }˯AcRpd=9\} _V* \(stj]B8S_lMv.yp魽:o^ u-6
y:lT3s5[\x3;?L6/ߍ~_ D\)1eZe Җ''!;k 2DCݍ_-{2 ܍xGypmb< z_ggRɻw/ggEL\Yy+3[,F+uj4Pek^iؾW>j: =?Rώ@oC]ӳ܆g(OycOa7߄)?&ZE8 y2<[
ϖj9?A9o)l6Cf; ^7ѕݙًq1M7Q|2f>,,sVdTD:I*WǑqq%< N1%Qb} x
y'nWF3-'QBwK
o_Fi !8I=7;r~dBo8/Qa iQWCQ9Lcj("^T#DŽQ D^>
8,<"gw!ߍ]:~fX 9fB|OKU
KA;x0SMwYMlV#=̇2TbLu_S?hzbBJ-1]]+ΥIw,klh{e߽Ϟ=xjQZ 1dOf fi'5
mMc[&(!H뭸_^ڦᛡ)ݜX бYQ+jEPz&kMEH_V;mtkf?mn{*܍Cx v-SEZ&
l^ꆒ?&!HPK<
YZZ]l6`ݪ Jx;/
Z\*>'ezѽ.ف#|au]%-|Q뒒LBLufScjR)OՏL< 0yolVc1aun0ñ1do(YL6'QYIQ\=O&+ ,ik
/6CFԆ@C dlamNMl7x)](Ω~@r\٣ hA3 9ؓ3ŀ+%Mua?lGأ}_K~{$ģI2jL'Y y+}Aȃ:R@ ؉"E_G@7)^Sl 8 y61[@Qb=
|.o<`y9;G]CdL!}tM X\[0՞A֚>S0($f`47 F
(C--RqO3ވ'Z=%|Ul 2eӦ~Afa?p+TWe[wd goڪ)+-ȷ:~\xxYJX98U"JSWV?'44PTK~D>!%+͚uۯwI.[4Π"'{inSTcE`t9_
ijǢIQ ˶rvm~2l^ ] ,yqb?~j _Pls 0vQ)FRlB=W/v[ahˮ JۢzCbW:ˏ&I-mDKC)SNzcPCܡ,^({Oc Sw'?NCEBkH* Iэ썺p`>l4F?BNP:5S?!VR5lVM1
1sϻȇk6Cڄ5# AyEC~VU%4'fn To--(&Yn7V_PBCNbw쨙)ftjSؒiG3i61npnC NN@o=ce%eZǸdzTX@!#SW%mѮ{Ef4^6, ?Ү%NmFZę0y ǟ=(\43-1 n+Φnmme
\V 2Ȭ"%wxzF=ɰxGfÊV0
9ݝ
9b
E6ffPt< PJD\oR6 *x[n%oT2ȬD |VS2 ,?x8V`XoC^A3UØ4u?O\U4dߜ>}d.%7|
il#^|T3kȯc3-S;Y}nW[PZ lzi%q'sV<:Cy>xrfp ziX;8N?N0.lNmBhXXct-ʔ
Z LOy B~1nbذ(Ecɒ\6
6^=YOOkЙttjƝ- Jܤ9+֗<,8hܤA:*_)}Ni( Z}5*}IoVn_ls-[Ui#~a9A (6PWϫΙhRD-a/C69
ۙ;w Qeǝx2M7万Z~iȧqw%]6%x r7 kQtE\LCæ[B- n1m[%# ttm ""5҇>`flvHc+VRbFfCIgy6VlPާH.Ԛp/*@&)t-[Ь,JUH 'χVZD 1d8өmS)-Fo1S`p
3uUhA(5ֻRvLa"6
ꤣ%"Z*R$h5ZYj爈b") %CZ<]2|f[A`,\n'!H)j l6{R,Rn %agsUC%~Mt{pB0(
.@Z>!?| 9~W0:#J7O
}3\vmeQ \zʳR{m$V ̻ԘyP j`-eXi$)BpG"N|v58dn[:%MB<ݦ1C(V";2>`i/Xnoòw)( OA:aSe^9ފn!-l3]Eq
ĦK6?C9Wܫ7fۥ6P. @~ӱ fbq_nv o?XNFqaS#_UhD 5~} XZMS_BexegԴT
K o &ឹElJBm Ix?w 4V_hwCMB; !cf匒oOU 鴏8D QSB<ݦ%an$Ҟ>`AatVQRcʘ4Fj8 & _HѓCXO9Kb\*>C*1pw$Fc^ٱT۴y9&=ÁS
Nb5}xuXf ?usz@]3:kc$'V%kZU2ݩ/!}W[諈ӮT t(!nxSCt"б 8^ՃM!>k@vTrxvÍ7!~)zqfO
ѦK/]|3K܃
ý:=
|Y0SC(0uz {1!չ ܅k_(㐏Ǯ[jo"uR ZޝZBRvϠ%mPvh,]SHp6vZGK[H>6Vt3-'gWa;L-*ݞӪ J rδiasaUy _kPk>d8k1[tSF|^ xVs2#ݩU.)Q|ns]2
ݩa`V8 {XlG"ȟ530AxÍ-O
c4ihӞ=%ϡ3O*&U 0 ubRɵ %KB<ݦ%J TݔiʃƤ 9<&4 dD٦!C :jN wdXEi,ZA_ڊ\hi F7JRj y]R-~)#g$.ؒlQ,ZS & h ,'L'WG(]2[q_BM@N2Fntw
旀w!eiSZхԚGJC3zO.s tz- /P/%ēPw-TH8 wb%#! qG2 g1$!HfD p}uMv8i8b3I!:}@=E'H{! v[a,' yDԸbs EFQT
D ƞ<\UҤO68~0q 횮OT864ͻRqb-.qDB<ݮE(Lʳը]~9U9"$4NC|u%W"3f?l m+4F-8obk63qM3ئKJL1(_sj~Le :lq~qmgj8s.]^2oe [1t2-2۪
.Ėyln Q6c>죥{Ec<̣RmTITBV[?qix+ɌET᥋RSeFǗɚسzvco|,|o'e3 5bعN#QcԎM"J=vkf"OjPi(M }n ,E'llie- ˘Q*bԈvdL93E(Ccg(c#1=#0YxkSfFG! 0\n Ug1bӛ`ȲY{d3ˣX 4ݰ351>>pT[opDlWۮpҹQwW(N],V5:? z"[iNؒaǻ+3|䘘 AK\Zo?Ў
{s}
3$ j{Pg/Q>'Ν{fb^UVu ~{iא_'G;$C
|n+:k |6
~fNk& fl
?"4
%\VignetteIndexEntry{ggbeeswarm package usage examples}
%\VignetteDepends{}
%\VignetteKeywords{visualization, display, one dimensional, grouped, groups, violin, scatter, points, quasirandom, beeswarm, van der Corput, beeswarm, ggplot, ggplot2}
%\VignettePackage{ggbeeswarm}
%\VignetteEngine{knitr::rmarkdown}
editor_options:
markdown:
wrap: sentence
---
## The basics
This is the simplest example of using `geom_quasirandom` to generate violin scatter plots:
```{r echo=TRUE, height=3.5, width=5}
library(ggbeeswarm)
set.seed(12345)
n <- 100
dat <- data.frame(
data = rnorm(n*2),
class = rep(c('a', 'b'), n)
)
ggplot(dat, aes(x = class, y = data)) +
geom_quasirandom()
```
The usual `ggplot2` options can be used:
```{r}
ggplot(dat, aes(x = class, y = data)) +
geom_quasirandom(aes(color = class))
```
Additional factors can be shown on the categorical axis by setting `dodge.width`, which creates smaller violin plots at each category (akin to `ggplot2::geom_jitterdodge`) and allows data to be compared within groups:
```{r}
ggplot(dat, aes(x = group, y = data, color = data > 0)) +
geom_quasirandom(dodge.width = 0.8)
```
This also works on the y-axis automatically:
```{r}
ggplot(dat, aes(y = group, x = data, color = data > 0)) +
geom_quasirandom(dodge.width = 0.8)
```
These examples also hold for `geom_beeswarm`.
Note that for the beeswarm methods, the `cex` argument should be specified in order to create well-spaced swarms:
```{r}
ggplot(dat, aes(x = group, y = data, color = data > 0)) +
geom_beeswarm(dodge.width = 0.8, cex=2)
```
Both `geom_beeswarm` and `geom_quasirandom` also work with facets:
```{r}
dat2 <- dat
dat2$group <- rnorm(n*2) > 0
ggplot(dat2, aes(x = class, y = data)) +
geom_quasirandom(dodge.width = 0.8) +
facet_wrap(facets = c("group"))
```
```{r}
ggplot(dat2, aes(x = class, y = data)) +
geom_beeswarm(dodge.width = 0.8, cex=3) +
facet_wrap(facets = c("group"))
```
## Options
There are several ways to plot grouped one-dimensional data combining points and density estimation including:
- **pseudorandom**: The kernel density is estimated then points are distributed uniform randomly within the density estimate for a given bin.
Selection of an appropriate number of bins does not greatly affect appearance but coincidental clumpiness is common.
- **alternating within bins**: The kernel density is estimated then points are distributed within the density estimate for a given bin evenly spaced with extreme values alternating from right to left e.g. max, 3rd max, ..., 4th max, 2nd max.
If maximums are placed on the outside then these plots often form consecutive "smiley" patterns.
If minimums are placed on the outside then "frowny" patterns are generated.
Selection of the number of bins can have large effects on appearance important.
- **tukey**: An algorithm described by Tukey and Tukey in "Strips displaying empirical distributions: I. textured dot strips" using constrained permutations of offsets to distribute the data.
- **quasirandom**: The kernel density is estimated then points are distributed quasi-randomly using the von der Corput sequence within the density estimate for a given bin.
Selection of an appropriate number of bins does not greatly affect appearance and position does not depend on plotting parameters.
- **beeswarm**: The package `beeswarm` provides methods for generating a "beeswarm" plot where points are distributed so that no points overlap.
Kernel density is not calculated although the resulting plot does provide an approximate density estimate.
Selection of an appropriate number of bins affects appearance and plot and point sizes must be known in advance.
The first four options are included within `geom_quasirandom` using the `method=` argument and beeswarm plots are generated using `geom_beeswarm`:
```{r height=8, width=6.5}
library(gridExtra)
dat <- list(
'Normal'=rnorm(50),
'Dense normal'= rnorm(500),
'Bimodal'=c(rnorm(100), rnorm(100,5)),
'Trimodal'=c(rnorm(100), rnorm(100,5),rnorm(100,-3))
)
labs<-rep(names(dat),sapply(dat,length))
labs<-factor(labs,levels=unique(labs))
dat<-unlist(dat)
p1<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(alpha=.2) +
ggtitle('quasirandom') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p2<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='pseudorandom',alpha=.2) +
ggtitle('pseudorandom') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p3<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',alpha=.2) +
ggtitle('smiley') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p4<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='frowney',alpha=.2) +
ggtitle('frowney') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p5<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='tukey',alpha=.2) +
ggtitle('tukey') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p6<-ggplot(mapping=aes(labs, dat)) +
geom_beeswarm(alpha=.2,size=.75) +
ggtitle('geom_beeswarm') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
grid.arrange(p1, p2, p3, p4, p5, p6, ncol=3)
```
`quasirandom` calls `vipor::offsetX` which calls `stats::density` to compute kernel density estimates. The tightness of the fit can be adjusted with the `bandwidth` option and the width of the offset with `width`. `nbins` to adjust the number of bins used in the kernel density is also provided; this can usually be left at its default when using quasirandom offsets but is useful for non-quasirandom methods:
```{r height=8, width=6}
library(gridExtra)
p1<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(bandwidth=2,alpha=.2) +
ggtitle('bandwidth=2') + labs(x='')
p2<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(bandwidth=.1,alpha=.2) +
ggtitle('bandwidth=.1') + labs(x='')
p3<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(width=.1,alpha=.2) +
ggtitle('width=.1') + labs(x='')
p4<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(nbins=100,alpha=.2) +
ggtitle('nbins=100') + labs(x='')
grid.arrange(p1, p2, p3, p4, ncol=1)
```
The `frowney` or `smiley` methods are sensitive to the number of bins, so the
argument `nbins` is more useful/necessary with them:
```{r height=8, width=6}
p1<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',alpha=.2) +
ggtitle('Default (n/5)') + labs(x='')
p2<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',nbins=50,alpha=.2) +
ggtitle('nbins=50') + labs(x='')
p3<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',nbins=100,alpha=.2) +
ggtitle('nbins=100') + labs(x='')
p4<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',nbins=250,alpha=.2) +
ggtitle('nbins=250') + labs(x='')
grid.arrange(p1, p2, p3, p4, ncol=1)
```
The `varwidth` argument scales the width of a group by the square root of the
number of observations in that group (as in the function `boxplot`):
```{r height=4, width=6}
dat <- list(
'10 points'=rnorm(10),
'50 points'=rnorm(50,2),
'200 points'=c(rnorm(400), rnorm(100,5)),
'5000 points'= rnorm(5000,1)
)
labs<-rep(names(dat),sapply(dat,length))
labs<-factor(labs,levels=unique(labs))
dat<-unlist(dat)
ggplot(mapping=aes(labs, dat)) + geom_quasirandom(alpha=.3,varwidth=TRUE)
```
## Real data
```{r}
library(dplyr)
set.seed(1234)
diamonds2 <- diamonds %>%
group_by(cut) %>%
sample_n(size = 100)
ggplot(diamonds2, aes(x = clarity, y = carat, color = price)) +
geom_quasirandom(size=1, varwidth = TRUE, width=0.7) +
facet_grid(rows=vars(cut)) +
scale_color_viridis_b(option = "A", end = 0.8)
# facet_wrap(facets=vars(carat_gt_1), scales = "free_y")
```
```{r}
ggplot(diamonds2, aes(x = clarity, y = carat, color = price)) +
geom_beeswarm(size=1, cex=1.3) +
facet_grid(rows=vars(cut)) +
scale_color_viridis_b(option = "A", end = 0.8)
```
An example using the `beaver1` and `beaver2` data from the `datasets` package:
```{r}
beaver<-data.frame(
'Temperature'=c(beaver1$temp,beaver2$temp),
'Beaver'=rep(
c('Beaver 1','Beaver 2'),
c(nrow(beaver1),nrow(beaver2))
)
)
ggplot(beaver,mapping=aes(Beaver, Temperature)) + geom_quasirandom()
```
ggbeeswarm/vignettes/usageExamples.Rnw 0000644 0001762 0000144 00000026151 14423030634 017704 0 ustar ligges users \documentclass[article,nojss]{jss}
<>=
options(keep.source = TRUE, width = 60)
packageInfo <- packageDescription("ggbeeswarm")
library(ggbeeswarm)
packageKeywords<-"visualization, display, one dimensional, grouped, groups, violin, scatter, points, quasirandom, beeswarm, van der Corput, beeswarm, ggplot, ggplot2"
@
%\VignetteIndexEntry{ggbeeswarm package usage examples}
%\VignetteDepends{}
%\VignetteKeywords{visualization, display, one dimensional, grouped, groups, violin, scatter, points, quasirandom, beeswarm, van der Corput, beeswarm, ggplot, ggplot2}
%\VignettePackage{ggbeeswarm}
\title{\pkg{\Sexpr{packageInfo$Package}} package usage example (version \Sexpr{packageInfo$Version})}
\author{\Sexpr{packageInfo$Author}}
\Plainauthor{\Sexpr{packageInfo$Author}}
\Address{ Github: \url{http://github.com/eclarke/ggbeeswarm}\\
Cran: \url{https://cran.r-project.org/package=ggbeeswarm}
}
\Keywords{\Sexpr{packageKeywords}}
\Abstract{
This is a collection of examples of usage for the \pkg{\Sexpr{packageInfo$Package}} package.
}
\begin{document}
\SweaveOpts{concordance=TRUE}
\SweaveOpts{engine=R,eps=FALSE}
\section{The basics}
This is the simplest example of using \code{geom_quasirandom} to generate violin scatter plots:
<>=
library(ggbeeswarm)
set.seed(12345)
n<-100
dat<-rnorm(n*2)
labs<-rep(c('a','b'),n)
ggplot(mapping=aes(labs, dat)) + geom_quasirandom()
@
\begin{center}
<>=
<>
@
\end{center}
Normal \code{ggplot} options can be used:
<>=
ggplot(mapping=aes(labs, dat)) + geom_quasirandom(aes(color=labs))
@
\begin{center}
<>=
<>
@
\end{center}
Factors can be used to generate custom group orderings:
<>=
labs2<-factor(labs,levels=c('b','a'))
ggplot(mapping=aes(labs2, dat)) + geom_quasirandom(aes(color=labs))
@
\begin{center}
<>=
<>
@
\end{center}
The axes can also be switched with a categorical y-axis:
<>=
ggplot(mapping=aes(dat,labs)) + geom_quasirandom(aes(color=labs))
@
\begin{center}
<>=
<>
@
\end{center}
And dodging can be used to compare within groups:
<>=
labs2<-factor(rep(1:2,each=n))
ggplot(mapping=aes(labs,dat,color=labs2)) + geom_quasirandom(dodge.width=.8)
@
\begin{center}
<>=
<>
@
\end{center}
Or on the y-axis:
<>=
labs2<-factor(rep(1:2,each=n))
ggplot(mapping=aes(dat,labs,color=labs2)) + geom_quasirandom(dodge.width=.8)
@
\begin{center}
<>=
<>
@
\end{center}
And with \code{geom_beeswarm}:
<>= labs2<-factor(rep(1:2,each=n))
ggplot(mapping=aes(labs,dat,color=labs2)) +
geom_beeswarm(dodge.width=.8,cex=2)
@
\begin{center}
<>=
<>
@
\end{center}
<>=
ggplot(mapping=aes(dat,labs,color=labs2)) +
geom_beeswarm(dodge.width=.8,cex=2)
@
\begin{center}
<>=
<>
@
\end{center}
Both \code{geom_beeswarm} and \code{geom_quasirandom} also work with facets:
<>= labs2<-factor(rep(1:2,each=n))
df<-data.frame(labs,dat,labs2)
ggplot(df,aes(labs,dat,color=labs2)) +
geom_quasirandom() +
facet_grid(.~labs2)
@
\begin{center}
<>=
<>
@
\end{center}
<>= labs2<-factor(rep(1:2,each=n))
ggplot(df,aes(labs,dat,color=labs2)) +
geom_beeswarm(cex=3) +
facet_grid(.~labs2)
@
\begin{center}
<>=
<>
@
\end{center}
\section{Options}
There are several ways to plot grouped one-dimensional data combining points and density estimation including:
\begin{description}
\item[pseudorandom] The kernel density is estimated then points are distributed uniform randomly within the density estimate for a given bin. Selection of an appropriate number of bins does not greatly affect appearance but coincidental clumpiness is common.
\item[alternating within bins] The kernel density is estimated then points are distributed within the density estimate for a given bin evenly spaced with extreme values alternating from right to left e.g. max, 3rd max, ..., 4th max, 2nd max. If maximums are placed on the outside then these plots often form consecutive ``smiley'' patterns. If minimums are placed on the outside then ``frowny'' patterns are generated. Selection of the number of bins can have large effects on appearance important.
\item[tukey] An algorithm described by Tukey and Tukey in ``Strips displaying empirical distributions: I. textured dot strips'' using constrained permutations of offsets to distrbute the data.
\item[quasirandom] The kernel density is estimated then points are distributed quasirandomly using the von der Corput sequence within the density estimate for a given bin. Selection of an appropriate number of bins does not greatly affect appearance and position does not depend on plotting parameters.
\item[beeswarm] The package \pkg{beeswarm} provides methods for generating a ``beeswarm'' plot where points are distibuted so that no points overlap. Kernel density is not calculated although the resulting plot does provide an approximate density estimate. Selection of an appropriate number of bins affects appearance and plot and point sizes must be known in advance.
\end{description}
The first four options are included within \code{geom_quasirandom} using the \code{method=} argument and beeswarm plots are generated with \code{geom_beeswarm}:
<>=
library(gridExtra)
dat <- list(
'Normal'=rnorm(50),
'Dense normal'= rnorm(500),
'Bimodal'=c(rnorm(100), rnorm(100,5)),
'Trimodal'=c(rnorm(100), rnorm(100,5),rnorm(100,-3))
)
labs<-rep(names(dat),sapply(dat,length))
labs<-factor(labs,levels=unique(labs))
dat<-unlist(dat)
p1<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(alpha=.2) +
ggtitle('quasirandom') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p2<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='pseudorandom',alpha=.2) +
ggtitle('pseudorandom') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p3<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',alpha=.2) +
ggtitle('smiley') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p4<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='frowney',alpha=.2) +
ggtitle('frowney') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p5<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='tukey',alpha=.2) +
ggtitle('tukey') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p6<-ggplot(mapping=aes(labs, dat)) +
geom_beeswarm(alpha=.2,size=.75) +
ggtitle('geom_beeswarm') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
grid.arrange(p1, p2, p3, p4, p5, p6, ncol=3)
@
\begin{center}
<>=
<>
@
\end{center}
\code{quasirandom} calls \code{vipor::offsetX} which calls \code{stats::density} to compute kernel density estimates. The tightness of the fit can be adjusted with the \code{bandwidth} option and the width of the offset with \code{width}. \code{nbins} to adjust the number of bins used in the kernel density is also provided but this can usually be left at its default when using quasirandom offsets but is useful for non-quasirandom methods:
<>=
library(gridExtra)
p1<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(bandwidth=2,alpha=.2) +
ggtitle('bandwidth=2') + labs(x='')
p2<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(bandwidth=.1,alpha=.2) +
ggtitle('bandwidth=.1') + labs(x='')
p3<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(width=.1,alpha=.2) +
ggtitle('width=.1') + labs(x='')
p4<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(nbins=100,alpha=.2) +
ggtitle('nbins=100') + labs(x='')
grid.arrange(p1, p2, p3, p4, ncol=1)
@
\begin{center}
<>=
<>
@
\end{center}
The \code{frowney} or \code{smiley} methods are sensitive to the number of bins so the argument \code{nbins} is more useful/necessary with them:
<>=
p1<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',alpha=.2) +
ggtitle('Default (n/5)') + labs(x='')
p2<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',nbins=50,alpha=.2) +
ggtitle('nbins=50') + labs(x='')
p3<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',nbins=100,alpha=.2) +
ggtitle('nbins=100') + labs(x='')
p4<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',nbins=250,alpha=.2) +
ggtitle('nbins=250') + labs(x='')
grid.arrange(p1, p2, p3, p4, ncol=1)
@
\begin{center}
<>=
<>
@
\end{center}
The \code{varwidth} argument scales the width of a group by the square root of the number of observations in that group (as in the function \code{boxplot}):
<>=
dat <- list(
'10 points'=rnorm(10),
'50 points'=rnorm(50,2),
'200 points'=c(rnorm(400), rnorm(100,5)),
'5000 points'= rnorm(5000,1)
)
labs<-rep(names(dat),sapply(dat,length))
labs<-factor(labs,levels=unique(labs))
dat<-unlist(dat)
ggplot(mapping=aes(labs, dat)) + geom_quasirandom(alpha=.3,varwidth=TRUE)
@
\begin{center}
<>=
<>
@
\end{center}
\section{Real data}
An example using the \code{beaver1} and \code{beaver2} data from the \pkg{datasets} package:
<>=
beaver<-data.frame(
'Temperature'=c(beaver1$temp,beaver2$temp),
'Beaver'=rep(
c('Beaver 1','Beaver 2'),
c(nrow(beaver1),nrow(beaver2))
)
)
ggplot(beaver,mapping=aes(Beaver, Temperature)) + geom_quasirandom()
@
\begin{center}
<>=
<>
@
\end{center}
An example using the \code{integrations} data from the \pkg{vipor} package and the argument \code{dodge.width}:
<>=
library(vipor)
ints<-integrations[integrations$nearestGene>0,]
ints$logGeneDist<-log(ints$nearestGene)
ggplot(ints,mapping=aes(study, logGeneDist,color=latent)) +
geom_quasirandom(dodge.width=.9,alpha=.4)
@
\begin{center}
<>=
<>
@
\end{center}
\end{document}
ggbeeswarm/R/ 0000755 0001762 0000144 00000000000 14423033016 012572 5 ustar ligges users ggbeeswarm/R/geom-quasirandom.R 0000644 0001762 0000144 00000003557 14423053314 016202 0 ustar ligges users #' Points, jittered to reduce overplotting using the vipor package
#'
#' The quasirandom geom is a convenient means to offset points within categories
#' to reduce overplotting. Uses the vipor package
#'
#' @section Aesthetics:
#' \Sexpr[results=rd,stage=build]{ggplot2:::rd_aesthetics("geom", "point")}
#'
#' @inheritParams ggplot2::geom_point
#' @inheritParams position_quasirandom
#' @import ggplot2
#' @seealso
#' [vipor::offsetSingleGroup()] how spacing is determined,
#' [ggplot2::geom_point()] for regular, unjittered points,
#' [ggplot2::geom_jitter()] for jittered points,
#' [geom_boxplot()] for another way of looking at the conditional
#' distribution of a variable
#' @examples
#' ggplot2::qplot(class, hwy, data = ggplot2::mpg, geom='quasirandom')
#' # Generate fake data
#' distro <- data.frame(
#' 'variable'=rep(c('runif','rnorm'),each=100),
#' 'value'=c(runif(100, min=-3, max=3), rnorm(100))
#' )
#' ggplot2::qplot(variable, value, data = distro, geom = 'quasirandom')
#' ggplot2::ggplot(distro,aes(variable, value)) + geom_quasirandom(width=0.1)
#' @export
geom_quasirandom <- function(
mapping = NULL,
data = NULL,
stat = 'identity',
...,
method = 'quasirandom',
width = NULL,
varwidth = FALSE,
bandwidth = .5,
nbins = NULL,
dodge.width = NULL,
groupOnX = NULL,
orientation = NULL,
na.rm = FALSE,
show.legend = NA,
inherit.aes = TRUE
) {
position <- position_quasirandom(
method = method,
width = width,
varwidth = varwidth,
bandwidth = bandwidth,
nbins = nbins,
dodge.width = dodge.width,
orientation = orientation,
na.rm = na.rm
)
ggplot2::layer(
data = data,
mapping = mapping,
stat = stat,
geom = GeomPoint,
position = position,
show.legend = show.legend,
inherit.aes = inherit.aes,
params = list(
na.rm = na.rm,
...
)
)
}
ggbeeswarm/R/position-beeswarm.R 0000644 0001762 0000144 00000040552 14423131630 016373 0 ustar ligges users #' An internal function to calculate new positions for geom_beeswarm
#'
#' @family position adjustments
#' @param data A data.frame containing plotting data in columns x and y.
#' Usually obtained from data processed by ggplot2.
#' @param yLim.expand y data limits plus a small expansion using `grDevices::extendrange`
#' @param xRange x axis scale range
#' @param yRange y axis scale range
#' @param method Method for arranging points (see Details below)
#' @param cex Scaling for adjusting point spacing (see [beeswarm::swarmx()]).
#' Values between 1 (default) and 3 tend to work best.
#' @param side Direction to perform jittering: 0: both directions;
#' 1: to the right or upwards; -1: to the left or downwards.
#' @param priority Method used to perform point layout (see Details below)
#' @param fast Use compiled version of swarm algorithm? This option is ignored
#' for all methods expect `"swarm"` and `"compactswarm"`.
#' @param corral `string`. Method used to adjust points that would be placed to
#' wide horizontally, default is `"none"`. See details below.
#' @param corral.width `numeric`. Width of the corral, default is `0.9`.
#'
#' @details
#' **method:** specifies the algorithm used to avoid overlapping points. The
#' default `"swarm"` method places points in increasing order. If a point would
#' overlap with an existing point, it is shifted sideways (along the group axis)
#' by a minimal amount sufficient to avoid overlap.
#'
#' While the `"swarm"` method places points in a predetermined
#' order, the `"compactswarm"` method uses a greedy strategy to determine which
#' point will be placed next. This often leads to a more tightly-packed layout.
#' The strategy is very simple: on each iteration, a point that can be placed as
#' close as possible to the non-data axis is chosen and placed. If there are two
#' or more equally good points, `priority` is used to break ties.
#'
#' The other 3 methods first discretise the values along the data axis, in order
#' to create more efficient packing. The `"square"` method places points on a
#' square grid, whereas `"hex"` uses a hexagonal grid. `"centre"`/`"center"`
#' uses a square grid to produce a symmetric swarm. The number of break points
#' for discretisation is determined by a combination of the available plotting
#' area and the `cex` argument.
#'
#' **priority:** controls the order in which points are placed, which generally
#' has a noticeable effect on the plot appearance. `"ascending"` gives the
#' 'traditional' beeswarm plot. `"descending"` is the opposite. `"density"`
#' prioritizes points with higher local density. `"random"` places points in a
#' random order. `"none"` places points in the order provided.
#'
#' **corral:** By default, swarms from different groups are not prevented from
#' overlapping, i.e. `"corral = "none"`. Thus, datasets that are very large or
#' unevenly distributed may produce ugly overlapping beeswarms. To control
#' runaway points one can use the following methods. `"gutter"` collects runaway
#' points along the boundary between groups. `"wrap"` implement periodic boundaries.
#' `"random"` places runaway points randomly in the region. `"omit"` omits runaway
#' points.
#'
#' @keywords internal
#' @importFrom beeswarm swarmx
#' @seealso [geom_beeswarm()], [position_quasirandom()],
#' [beeswarm::swarmx()]
offset_beeswarm <- function(
data,
yLim.expand,
xRange,
yRange,
method = "swarm",
cex = 1,
side = 0L,
priority = "ascending",
fast = TRUE,
corral = "none",
corral.width = 0.2
) {
if (method %in% c("swarm", "compactswarm")) {
## SWARM METHODS
# Determine point size as per `ggbeeswarm` CRAN version 0.6.0
# divisor is a magic number to get a reasonable baseline
# better option would be to figure out point size in user coords
x.size <- xRange / 100
y.size <- yRange / 100
compact <- method == "compactswarm"
x.offset <- beeswarm::swarmx(
x = rep(0, length(data$y)),
y = data$y,
xsize = x.size,
ysize = y.size,
cex = cex, side = side, priority = priority,
fast = fast, compact = compact
)$x
} else {
## NON-SWARM METHODS
# Determine point size as per `ggbeeswarm` CRAN version 0.6.0
# divisor is a magic number to get a reasonable baseline
# better option would be to figure out point size in user coords
x.size <- xRange / 100 * cex
y.size <- yRange / 100 * cex
# Hex method specific step
if (method == "hex") y.size <- y.size * sqrt(3) / 2
# Determine positions along the y axis
breaks <- seq(yLim.expand[1], yLim.expand[2] + y.size, by = y.size)
mids <- (utils::head(breaks, -1) + utils::tail(breaks, -1)) / 2
# include.lowest = T to account for cases where all y values are the same,
# which otherwise would result in NAs. Fixes issue #85.
y.index <- sapply(data$y, cut, breaks = breaks, include.lowest=T, labels = FALSE)
y.pos <- sapply(y.index, function(a) mids[a])
if (any(data$y != y.pos)) {
cli::cli_warn(c(
"In `position_beeswarm`, method `{method}` discretizes the data axis (a.k.a the continuous or non-grouped axis).",
"This may result in changes to the position of the points along that axis, proportional to the value of `cex`."
), .frequency = "once", .frequency_id = "beeswarm_method_data_axis_warn")
}
data$y <- y.pos
# Determine positions along the x axis
x.index <- determine_pos(y.index, method, side)
x.offset <- x.index * x.size
}
## CORRAL RUNAWAY POINTS
if (corral != "none") {
corral.low <- (side - 1) * corral.width / 2
corral.high <- (side + 1) * corral.width / 2
if (corral == "gutter") {
x.offset <- sapply(
x.offset,
function(zz) pmin(corral.high, pmax(corral.low, zz))
)
}
if (corral == "wrap") {
if (side == -1L) {
# special case with side=-1: reverse the corral to avoid artefacts at zero
x.offset <- sapply(
x.offset,
function(zz) corral.high - ((corral.high - zz) %% corral.width)
)
} else {
x.offset <- sapply(
x.offset,
function(zz) ((zz - corral.low) %% corral.width) + corral.low
)
}
}
if (corral == 'random') {
x.offset <- sapply(
x.offset,
function(zz) ifelse(
zz > corral.high | zz < corral.low,
yes = stats::runif(length(zz), corral.low, corral.high),
no = zz
)
)
}
if (corral == 'omit') {
x.offset <- sapply(
x.offset,
function(zz) ifelse(
zz > corral.high | zz < corral.low,
yes = NA,
no = zz
)
)
}
}
data$x <- data$x + x.offset
return(data)
}
#' Arrange points using the `\link[beeswarm]` package.
#'
#' @family position adjustments
#' @param method Method for arranging points (see Details below)
#' @param cex Scaling for adjusting point spacing (see [beeswarm::swarmx()]).
#' Values between 1 (default) and 3 tend to work best.
#' @param side Direction to perform jittering: 0: both directions;
#' 1: to the right or upwards; -1: to the left or downwards.
#' @param priority Method used to perform point layout (see Details below)
#' @param fast Use compiled version of swarm algorithm? This option is ignored
#' for all methods expect `"swarm"` and `"compactswarm"`.
#' @param dodge.width Amount by which points from different aesthetic groups
#' will be dodged. This requires that one of the aesthetics is a factor.
#' @param corral `string`. Method used to adjust points that would be placed to
#' wide horizontally, default is `"none"`. See details below.
#' @param corral.width `numeric`. Width of the corral, default is `0.9`.
#' @param groupOnX `r lifecycle::badge("deprecated")` No longer needed.
#' @details
#' **method:** specifies the algorithm used to avoid overlapping points. The
#' default `"swarm"` method places points in increasing order. If a point would
#' overlap with an existing point, it is shifted sideways (along the group axis)
#' by a minimal amount sufficient to avoid overlap.
#'
#' While the `"swarm"` method places points in a predetermined
#' order, the `"compactswarm"` method uses a greedy strategy to determine which
#' point will be placed next. This often leads to a more tightly-packed layout.
#' The strategy is very simple: on each iteration, a point that can be placed as
#' close as possible to the non-data axis is chosen and placed. If there are two
#' or more equally good points, `priority` is used to break ties.
#'
#' The other 3 methods first discretise the values along the data axis, in order
#' to create more efficient packing. The `"square"` method places points on a
#' square grid, whereas `"hex"` uses a hexagonal grid. `"centre"`/`"center"`
#' uses a square grid to produce a symmetric swarm. The number of break points
#' for discretisation is determined by a combination of the available plotting
#' area and the `cex` argument.
#'
#' **priority:** controls the order in which points are placed, which generally
#' has a noticeable effect on the plot appearance. `"ascending"` gives the
#' 'traditional' beeswarm plot. `"descending"` is the opposite. `"density"`
#' prioritizes points with higher local density. `"random"` places points in a
#' random order. `"none"` places points in the order provided.
#'
#' **corral:** By default, swarms from different groups are not prevented from
#' overlapping, i.e. `"corral = "none"`. Thus, datasets that are very large or
#' unevenly distributed may produce ugly overlapping beeswarms. To control
#' runaway points one can use the following methods. `"gutter"` collects runaway
#' points along the boundary between groups. `"wrap"` implement periodic boundaries.
#' `"random"` places runaway points randomly in the region. `"omit"` omits runaway
#' points.
#'
#'
#' @export
#' @importFrom beeswarm swarmx
#' @seealso [geom_beeswarm()], [position_quasirandom()],
#' [beeswarm::swarmx()]
position_beeswarm <- function(
method = "swarm",
cex = 1,
side = 0L,
priority = "ascending",
fast = TRUE,
groupOnX = NULL,
dodge.width = 0,
corral = "none",
corral.width = 0.2
) {
if (!missing(groupOnX)) {
lifecycle::deprecate_soft(
when = "0.7.1", what = "position_beeswarm(groupOnX)",
details='ggplot2 now handles this case automatically.'
)
}
if (method == "centre") method <- "center"
ggproto(NULL, PositionBeeswarm,
method = method,
cex = cex,
side = side,
priority = priority,
fast = fast,
dodge.width = dodge.width,
corral = corral,
corral.width = corral.width
)
}
PositionBeeswarm <- ggplot2::ggproto("PositionBeeswarm", Position,
required_aes = c('x', 'y'),
setup_params = function(self, data) {
flipped_aes <- has_flipped_aes(data)
data <- flip_data(data, flipped_aes)
# get y range of data and extend it a little
yLim.expand <- grDevices::extendrange(data$y, f = 0.01)
list(
# groupOnX = self$groupOnX, deprecated
method = self$method,
cex = self$cex,
side = self$side,
priority = self$priority,
fast = self$fast,
dodge.width = self$dodge.width,
corral = self$corral,
corral.width = self$corral.width,
yLim.expand = yLim.expand,
flipped_aes = flipped_aes
)
},
compute_panel = function(data, params, scales) {
data <- flip_data(data, params$flipped_aes)
# get plot limits
if (params$flipped_aes) {
xRange <- get_range(scales$y)
yRange <- get_range(scales$x)
} else {
xRange <- get_range(scales$x)
yRange <- get_range(scales$y)
}
data <- ggplot2:::collide(
data,
params$dodge.width,
name = "position_beeswarm",
strategy = ggplot2:::pos_dodge,
check.width = FALSE
)
# split data.frame into list of data.frames
if(!is.null(params$dodge.width)) {
data <- split(data, data$group)
} else {
data <- split(data, data$x)
}
# perform swarming separately for each data.frame
data <- lapply(
data,
offset_beeswarm,
yLim.expand = params$yLim.expand,
xRange = xRange,
yRange = yRange,
method = params$method,
cex = params$cex,
side = params$side,
priority = params$priority,
fast = params$fast,
corral = params$corral,
corral.width = params$corral.width
)
# recombine list of data.frames into one
data <- Reduce(rbind, data)
flip_data(data, params$flipped_aes)
}
)
get_range <- function(scales) {
if (is.null(scales$limits)) lim <- scales$range$range
else lim <- scales$get_limits()
if (inherits(scales, "ScaleContinuous")) {
out <- diff(lim)
} else if (inherits(scales, "ScaleDiscrete")) {
out <- length(unique(lim))
} else {
stop("Unknown scale type")
}
if (out == 0) out <- 1
out
}
determine_pos <- function(v, method, side) {
# if(length(stats::na.omit(v)) == 0)
# return(v)
v.s <- lapply(split(v, v), seq_along)
if(method %in% c("center", "square") && side == -1)
v.s <- lapply(v.s, function(a) a - max(a))
else if(method %in% c("center", "square") && side == 1)
v.s <- lapply(v.s, function(a) a - 1)
else if(method == "center")
v.s <- lapply(v.s, function(a) a - mean(a))
else if(method == "square")
v.s <- lapply(v.s, function(a) a - floor(mean(a)))
else if(method == "hex") {
odd.row <- (as.numeric(names(v.s)) %% 2) == 1
if(side == 0) {
v.s[ odd.row] <- lapply(v.s[ odd.row], function(a) a - floor(mean(a)) - 0.25)
v.s[!odd.row] <- lapply(v.s[!odd.row], function(a) a - ceiling(mean(a)) + 0.25)
} else if(side == -1) {
v.s[ odd.row] <- lapply(v.s[ odd.row], function(a) a - max(a))
v.s[!odd.row] <- lapply(v.s[!odd.row], function(a) a - max(a) - 0.5)
} else if(side == 1) {
v.s[ odd.row] <- lapply(v.s[ odd.row], function(a) a - 1)
v.s[!odd.row] <- lapply(v.s[!odd.row], function(a) a - 0.5)
}
}
unsplit(v.s, v)
}
ggbeeswarm/R/position-quasirandom.R 0000644 0001762 0000144 00000023263 14423123510 017107 0 ustar ligges users #' Arrange points using quasirandom noise to avoid overplotting
#'
#' @family position adjustments
#' @param method the method used for distributing points
#' (quasirandom, pseudorandom, smiley, maxout, frowney, minout, tukey, tukeyDense).
#' See [vipor::offsetSingleGroup()] for the details of each method.
#' @param width the maximum amount of spread (default: 0.4)
#' @param varwidth vary the width by the relative size of each group
#' @param bandwidth the bandwidth adjustment to use when calculating density
#' Smaller numbers (< 1) produce a tighter "fit". (default: 0.5)
#' @param nbins the number of bins used when calculating density
#' (has little effect with quasirandom/random distribution)
#' @param dodge.width Amount by which points from different aesthetic groups
#' will be dodged. This requires that one of the aesthetics is a factor.
#' To disable dodging between groups, set this to NULL.
#' @param na.rm if FALSE, the default, missing values are removed with a warning.
#' If TRUE, missing values are silently removed.
#' @param orientation The orientation (i.e., which axis to group on) is inferred from the data.
#' This can be overridden by setting `orientation` to either `"x"` or `"y"`.
#' @param groupOnX `r lifecycle::badge("superseded")` See `orientation`.
#' @importFrom vipor offsetSingleGroup
#' @export
#' @seealso [vipor::offsetSingleGroup()], [geom_quasirandom()]
position_quasirandom <- function(
method = 'quasirandom',
width = NULL,
varwidth = FALSE,
bandwidth = .5,
nbins = NULL,
dodge.width = 0,
orientation = NULL,
groupOnX = NULL,
na.rm = FALSE
) {
if (!missing(groupOnX)) {
lifecycle::deprecate_soft(
when = "0.7.1", what = "position_quasirandom(groupOnX)",
details='The axis to group on is now guessed from the data. To override, specify orientation="x" or "y".'
)
if (groupOnX) {
orientation = "x"
} else {
orientation = "y"
}
}
if (!method %in% c("quasirandom", "pseudorandom", "smiley", "maxout", "frowney", "minout", "tukey", "tukeyDense")) {
cli::cli_abort("{.fn method} must be one of: quasirandom, pseudorandom, smiley, maxout, frowney, minout, tukey, or tukeyDense.")
}
if (!is.null(orientation) && !(orientation %in% c("x", "y"))) {
cli::cli_abort("{.fn orientation} must be 'x', 'y', or NULL.")
}
ggproto(NULL, PositionQuasirandom,
width = width,
varwidth = varwidth,
bandwidth = bandwidth,
nbins = nbins,
method = method,
dodge.width = dodge.width,
na.rm = na.rm,
orientation = orientation
)
}
PositionQuasirandom <- ggplot2::ggproto("PositionQuasirandom", Position,
required_aes = c('x', 'y'),
setup_params = function(self, data) {
params <- list(
width = self$width,
varwidth = self$varwidth,
bandwidth = self$bandwidth,
nbins = self$nbins,
method = self$method,
dodge.width = self$dodge.width,
na.rm = self$na.rm,
orientation = self$orientation
)
if (!is.null(params$orientation)) {
flipped_aes <- has_flipped_aes(data, params, ambiguous = TRUE)
} else {
flipped_aes <- has_flipped_aes(data, group_has_equal = TRUE)
if (flipped_aes) {
cli::cli_inform("Orientation inferred to be along y-axis; override with `position_quasirandom(orientation = 'x')`")
}
}
params$flipped_aes <- flipped_aes
data <- flip_data(data, params$flipped_aes)
# get the number of points in each x axis group
# and find the largest group
params$max_length <- max(data.frame(table(data$x))$Freq)
# check that the number of groups < number of data points
if (!anyDuplicated(data$group)) {
if (!is.null(params$dodge.width)) {
# Warn if dodge.width was set to something besides default
if (params$dodge.width != 0) {
cli::cli_inform(
"Each group consists of only one observation; resetting dodge.width to NULL.",
"Disable this message by explicitly setting `dodge.width=NULL`, or by adjusting the group aesthetic."
)
}
params$dodge.width = NULL
}
}
params
},
setup_data = function(self, data, params) {
data <- flip_data(data, params$flipped_aes)
data <- remove_missing(
data,
na.rm = params$na.rm,
vars = "y",
name = "position_quasirandom"
)
flip_data(data, params$flipped_aes)
},
compute_panel = function(data, params, scales) {
data <- flip_data(data, params$flipped_aes)
# perform dodging if necessary
data <- ggplot2:::collide(
data,
params$dodge.width,
name = "position_quasirandom",
strategy = ggplot2:::pos_dodge,
check.width = FALSE
)
# ggplot2::resolution needs to be after dodge
# set width if not specified
if (is.null(params$width)) {
params$width <- ggplot2::resolution(
data$x, zero = FALSE) * 0.4
}
# split data.frame into list of data.frames
if (!is.null(params$dodge.width)) {
data <- split(data, data$group)
} else {
data <- split(data, data$x)
}
# perform transformation separately for each data.frame
data <- lapply(
data,
offset_quasirandom,
method = params$method,
width = params$width,
vary.width = params$varwidth,
adjust = params$bandwidth,
nbins = params$nbins,
max.length = params$max.length
)
# recombine list of data.frame into one
data <- Reduce(rbind, data)
flip_data(data, params$flipped_aes)
}
)
offset_quasirandom <- function(
data,
width = 0.4,
vary.width = FALSE,
max.length = NULL,
# na.rm = FALSE,
...
) {
# if (any(is.na(data$y))) {
# if (na.rm) {
#
# }
# }
x.offset <- vipor::aveWithArgs(
data$y, data$x,
FUN = vipor::offsetSingleGroup,
maxLength = if (vary.width) {max.length} else {NULL},
...
)
x.offset <- x.offset * width
data$x <- data$x + x.offset
data
}
ggbeeswarm/R/ggbeeswarm-package.R 0000644 0001762 0000144 00000001565 14347131222 016443 0 ustar ligges users #' ggbeeswarm extends ggplot2 with violin point/beeswarm plots
#'
#' This package allows plotting of several groups of one dimensional data as a violin point/beeswarm plot in ggplot2 by arranging data points to resemble the underlying distribution. The development version of this package is on .
#' @docType package
#' @name ggbeeswarm
#' @author Erik Clarke, \email{erikclarke@@gmail.com}
#' @seealso [position_quasirandom()]
#' @examples
#'
#' ggplot2::ggplot(ggplot2::mpg,aes(class, hwy)) + geom_quasirandom()
#' # Generate fake data
#' distro <- data.frame(
#' 'variable'=rep(c('runif','rnorm'),each=100),
#' 'value'=c(runif(100, min=-3, max=3), rnorm(100))
#' )
#' ggplot2::ggplot(distro,aes(variable, value)) + geom_quasirandom()
#' ggplot2::ggplot(distro,aes(variable, value)) + geom_quasirandom(width=.1)
#'
NULL
ggbeeswarm/R/geom-beeswarm.R 0000644 0001762 0000144 00000005610 14347131222 015454 0 ustar ligges users #' Points, jittered to reduce overplotting using the beeswarm package
#'
#' The beeswarm geom is a convenient means to offset points within categories to
#' reduce overplotting. Uses the beeswarm package
#'
#' @section Aesthetics:
#' \Sexpr[results=rd,stage=build]{ggplot2:::rd_aesthetics("geom", "point")}
#'
#' @inheritParams ggplot2::geom_point
#' @inheritParams offset_beeswarm
#' @param dodge.width Amount by which points from different aesthetic groups
#' will be dodged. This requires that one of the aesthetics is a factor.
#' @param groupOnX `r lifecycle::badge("deprecated")` No longer needed.
#' @param beeswarmArgs `r lifecycle::badge("deprecated")` No longer used.
#' @import ggplot2
#' @seealso
#' [geom_quasirandom()] an alternative method,
#' [beeswarm::swarmx()] how spacing is determined,
#' [ggplot2::geom_point()] for regular, unjittered points,
#' [ggplot2::geom_jitter()] for jittered points,
#' [ggplot2::geom_boxplot()] for another way of looking at the conditional
#' distribution of a variable
#' @export
#' @examples
#'
#' ggplot2::qplot(class, hwy, data = ggplot2::mpg, geom='beeswarm')
#' # Generate fake data
#' distro <- data.frame(
#' 'variable'=rep(c('runif','rnorm'),each=100),
#' 'value'=c(runif(100, min=-3, max=3), rnorm(100))
#' )
#' ggplot2::qplot(variable, value, data = distro, geom='beeswarm')
#' ggplot2::ggplot(distro,aes(variable, value)) +
#' geom_beeswarm(priority='density',size=2.5)
geom_beeswarm <- function(
mapping = NULL,
data = NULL,
stat = 'identity',
...,
method = "swarm",
cex = 1,
side = 0L,
priority = "ascending",
fast = TRUE,
dodge.width = NULL,
corral = "none",
corral.width = 0.9,
groupOnX = NULL,
beeswarmArgs = list(),
na.rm = FALSE,
show.legend = NA,
inherit.aes = TRUE
) {
if (!missing(groupOnX)) {
lifecycle::deprecate_soft(
when = "0.7.1", what = "geom_beeswarm(groupOnX)",
details='ggplot2 now handles this case automatically.'
)
}
if (!missing(beeswarmArgs)) {
lifecycle::deprecate_soft(
when = "0.7.1", what = "geom_beeswarm(beeswarmArgs)"
)
}
if (!method %in% c("swarm", "compactswarm", "hex", "square", "centre", "center")) {
stop(sprintf("The method must be one of: swarm, compactswarm, hex, square, center, or centre."))
}
if (!corral %in% c("none", "gutter", "wrap", "random", "omit")) {
stop(sprintf("The corral argument must be one of: none, gutter, wrap, random, or omit."))
}
position <- position_beeswarm(
method = method,
cex = cex,
side = side,
priority = priority,
fast = fast,
dodge.width = dodge.width,
corral = corral,
corral.width = corral.width
)
layer(
data = data,
mapping = mapping,
stat = stat,
geom = GeomPoint,
position = position,
show.legend = show.legend,
inherit.aes = inherit.aes,
params = list(
na.rm = na.rm,
...
)
)
}
ggbeeswarm/NEWS.md 0000644 0001762 0000144 00000002534 14423030634 013476 0 ustar ligges users ---
editor_options:
markdown:
wrap: 80
---
# ggbeeswarm
## Development version
**Bugfixes:**
- `position_quasirandom()` default `dodge.width` is now `NULL` instead of 0 (#79)
- A few stray references to the deprecated `groupOnX` argument are removed.
## v0.7.1
This release incorporates all the incredible work done by @csdaw to refactor the
various geom and position functions, fix outstanding issues, and remove
dependencies on deprecated ggplot2 functions like `qplot`.
This version will be on CRAN.
**New features:**
- Can now use new features from the `beeswarm` package, including compact
swarms and 'corralling' points in overlapping groups.
- Updated examples and documentation.
**Changes:**
- The `groupOnX` argument is deprecated. `ggplot2` now has logic to determine
when a plot is 'flipped' automatically so this argument is no longer
necessary.
- The `beeswarmArgs` argument is deprecated as its arguments are now called in
`geom_beeswarm` directly.
- `position_beeswarm` is once again an exported function.
- Examples and documentation no longer use `ggplot2::qplot` as it has been
hard- deprecated in `ggplot2`
- The license has been clarified to be GPL-3, not GPL \>= 2.
- Minimum R and `ggplot2` versions have changed to 3.5.0 and 3.3.0,
respectively.
**Bugfixes:**
Too many to summarize here.
ggbeeswarm/MD5 0000644 0001762 0000144 00000003275 14423307262 012717 0 ustar ligges users b2b3f6b172c2484ec7bbb1681df6e52d *DESCRIPTION
fa0c2110d6223ad501ab214205e757ad *NAMESPACE
549110535286a26c8f27a61018fa1dc1 *NEWS.md
28fe8ff8b2bb0d44be6479ba6e7cd5d2 *R/geom-beeswarm.R
45f1779037b7f09bf97bd7d9e976a67d *R/geom-quasirandom.R
55896b76da3aca64446dc2c5e0d12f7a *R/ggbeeswarm-package.R
8ee9967c6d1ec1110f5be051673b5c6e *R/position-beeswarm.R
845acf35d8b26d2b19eeee2250d93c70 *R/position-quasirandom.R
a213c2703ce4cb6ae51e3989cc8cb84e *build/partial.rdb
fe74dcc56945630ba183c6ca8e94530f *build/vignette.rds
6bf3f22998936407b07868038bad5c41 *inst/doc/usageExamples.R
da4f297363bc3761bfde15cbe0b9b455 *inst/doc/usageExamples.Rnw
9757240c4dd5d85edf050fd78fe03aa2 *inst/doc/usageExamples.pdf
cb1e46f469cfbbbde29c8b5113e1d789 *man/figures/lifecycle-archived.svg
c0d2e5a54f1fa4ff02bf9533079dd1f7 *man/figures/lifecycle-defunct.svg
a1b8c987c676c16af790f563f96cbb1f *man/figures/lifecycle-deprecated.svg
c3978703d8f40f2679795335715e98f4 *man/figures/lifecycle-experimental.svg
952b59dc07b171b97d5d982924244f61 *man/figures/lifecycle-maturing.svg
27b879bf3677ea76e3991d56ab324081 *man/figures/lifecycle-questioning.svg
53b3f893324260b737b3c46ed2a0e643 *man/figures/lifecycle-stable.svg
1c1fe7a759b86dc6dbcbe7797ab8246c *man/figures/lifecycle-superseded.svg
e1293b7f5a64040145e14a6b3502cb8e *man/geom_beeswarm.Rd
2b12543aefbdf0eaa807429e0faec958 *man/geom_quasirandom.Rd
9d24ba5fb2e5c239380ee6347de274b2 *man/ggbeeswarm.Rd
c80348541465be6428aabfb41b90c670 *man/offset_beeswarm.Rd
aa16283bf1765399564d068f30aef5f9 *man/position_beeswarm.Rd
76cba2b82c1745b24da04f405e02230f *man/position_quasirandom.Rd
da4f297363bc3761bfde15cbe0b9b455 *vignettes/usageExamples.Rnw
54e19302c156f569ff153f2a31fe5f99 *vignettes/usage_examples.Rmd
ggbeeswarm/inst/ 0000755 0001762 0000144 00000000000 14423302153 013347 5 ustar ligges users ggbeeswarm/inst/doc/ 0000755 0001762 0000144 00000000000 14423302153 014114 5 ustar ligges users ggbeeswarm/inst/doc/usageExamples.R 0000644 0001762 0000144 00000034141 14423302153 017045 0 ustar ligges users ### R code from vignette source 'usageExamples.Rnw'
###################################################
### code chunk number 1: package
###################################################
options(keep.source = TRUE, width = 60)
packageInfo <- packageDescription("ggbeeswarm")
library(ggbeeswarm)
packageKeywords<-"visualization, display, one dimensional, grouped, groups, violin, scatter, points, quasirandom, beeswarm, van der Corput, beeswarm, ggplot, ggplot2"
###################################################
### code chunk number 2: ggPlot (eval = FALSE)
###################################################
## library(ggbeeswarm)
## set.seed(12345)
## n<-100
## dat<-rnorm(n*2)
## labs<-rep(c('a','b'),n)
## ggplot(mapping=aes(labs, dat)) + geom_quasirandom()
###################################################
### code chunk number 3: showGgPlot
###################################################
library(ggbeeswarm)
set.seed(12345)
n<-100
dat<-rnorm(n*2)
labs<-rep(c('a','b'),n)
ggplot(mapping=aes(labs, dat)) + geom_quasirandom()
###################################################
### code chunk number 4: ggOpts (eval = FALSE)
###################################################
## ggplot(mapping=aes(labs, dat)) + geom_quasirandom(aes(color=labs))
###################################################
### code chunk number 5: showGgOpts
###################################################
ggplot(mapping=aes(labs, dat)) + geom_quasirandom(aes(color=labs))
###################################################
### code chunk number 6: ggFactors (eval = FALSE)
###################################################
## labs2<-factor(labs,levels=c('b','a'))
## ggplot(mapping=aes(labs2, dat)) + geom_quasirandom(aes(color=labs))
###################################################
### code chunk number 7: showGgFactors
###################################################
labs2<-factor(labs,levels=c('b','a'))
ggplot(mapping=aes(labs2, dat)) + geom_quasirandom(aes(color=labs))
###################################################
### code chunk number 8: yaxis (eval = FALSE)
###################################################
## ggplot(mapping=aes(dat,labs)) + geom_quasirandom(aes(color=labs))
###################################################
### code chunk number 9: showYaxis
###################################################
ggplot(mapping=aes(dat,labs)) + geom_quasirandom(aes(color=labs))
###################################################
### code chunk number 10: dodge (eval = FALSE)
###################################################
## labs2<-factor(rep(1:2,each=n))
## ggplot(mapping=aes(labs,dat,color=labs2)) + geom_quasirandom(dodge.width=.8)
###################################################
### code chunk number 11: showDodge
###################################################
labs2<-factor(rep(1:2,each=n))
ggplot(mapping=aes(labs,dat,color=labs2)) + geom_quasirandom(dodge.width=.8)
###################################################
### code chunk number 12: dodgey (eval = FALSE)
###################################################
## labs2<-factor(rep(1:2,each=n))
## ggplot(mapping=aes(dat,labs,color=labs2)) + geom_quasirandom(dodge.width=.8)
###################################################
### code chunk number 13: showDodgey
###################################################
labs2<-factor(rep(1:2,each=n))
ggplot(mapping=aes(dat,labs,color=labs2)) + geom_quasirandom(dodge.width=.8)
###################################################
### code chunk number 14: dodgeBee (eval = FALSE)
###################################################
## ggplot(mapping=aes(labs,dat,color=labs2)) +
## geom_beeswarm(dodge.width=.8,cex=2)
###################################################
### code chunk number 15: showDodgeBee
###################################################
ggplot(mapping=aes(labs,dat,color=labs2)) +
geom_beeswarm(dodge.width=.8,cex=2)
###################################################
### code chunk number 16: dodgeYBee (eval = FALSE)
###################################################
## ggplot(mapping=aes(dat,labs,color=labs2)) +
## geom_beeswarm(dodge.width=.8,cex=2)
###################################################
### code chunk number 17: showDodgeYBee
###################################################
ggplot(mapping=aes(dat,labs,color=labs2)) +
geom_beeswarm(dodge.width=.8,cex=2)
###################################################
### code chunk number 18: facetQuasi (eval = FALSE)
###################################################
## df<-data.frame(labs,dat,labs2)
## ggplot(df,aes(labs,dat,color=labs2)) +
## geom_quasirandom() +
## facet_grid(.~labs2)
###################################################
### code chunk number 19: showFacetQuasi
###################################################
df<-data.frame(labs,dat,labs2)
ggplot(df,aes(labs,dat,color=labs2)) +
geom_quasirandom() +
facet_grid(.~labs2)
###################################################
### code chunk number 20: facetBee (eval = FALSE)
###################################################
## ggplot(df,aes(labs,dat,color=labs2)) +
## geom_beeswarm(cex=3) +
## facet_grid(.~labs2)
###################################################
### code chunk number 21: showFacetBee
###################################################
ggplot(df,aes(labs,dat,color=labs2)) +
geom_beeswarm(cex=3) +
facet_grid(.~labs2)
###################################################
### code chunk number 22: methods (eval = FALSE)
###################################################
## library(gridExtra)
## dat <- list(
## 'Normal'=rnorm(50),
## 'Dense normal'= rnorm(500),
## 'Bimodal'=c(rnorm(100), rnorm(100,5)),
## 'Trimodal'=c(rnorm(100), rnorm(100,5),rnorm(100,-3))
## )
## labs<-rep(names(dat),sapply(dat,length))
## labs<-factor(labs,levels=unique(labs))
## dat<-unlist(dat)
## p1<-ggplot(mapping=aes(labs, dat)) +
## geom_quasirandom(alpha=.2) +
## ggtitle('quasirandom') + labs(x='') +
## theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
## p2<-ggplot(mapping=aes(labs, dat)) +
## geom_quasirandom(method='pseudorandom',alpha=.2) +
## ggtitle('pseudorandom') + labs(x='') +
## theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
## p3<-ggplot(mapping=aes(labs, dat)) +
## geom_quasirandom(method='smiley',alpha=.2) +
## ggtitle('smiley') + labs(x='') +
## theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
## p4<-ggplot(mapping=aes(labs, dat)) +
## geom_quasirandom(method='frowney',alpha=.2) +
## ggtitle('frowney') + labs(x='') +
## theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
## p5<-ggplot(mapping=aes(labs, dat)) +
## geom_quasirandom(method='tukey',alpha=.2) +
## ggtitle('tukey') + labs(x='') +
## theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
## p6<-ggplot(mapping=aes(labs, dat)) +
## geom_beeswarm(alpha=.2,size=.75) +
## ggtitle('geom_beeswarm') + labs(x='') +
## theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
## grid.arrange(p1, p2, p3, p4, p5, p6, ncol=3)
###################################################
### code chunk number 23: showMethods
###################################################
library(gridExtra)
dat <- list(
'Normal'=rnorm(50),
'Dense normal'= rnorm(500),
'Bimodal'=c(rnorm(100), rnorm(100,5)),
'Trimodal'=c(rnorm(100), rnorm(100,5),rnorm(100,-3))
)
labs<-rep(names(dat),sapply(dat,length))
labs<-factor(labs,levels=unique(labs))
dat<-unlist(dat)
p1<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(alpha=.2) +
ggtitle('quasirandom') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p2<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='pseudorandom',alpha=.2) +
ggtitle('pseudorandom') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p3<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',alpha=.2) +
ggtitle('smiley') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p4<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='frowney',alpha=.2) +
ggtitle('frowney') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p5<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='tukey',alpha=.2) +
ggtitle('tukey') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p6<-ggplot(mapping=aes(labs, dat)) +
geom_beeswarm(alpha=.2,size=.75) +
ggtitle('geom_beeswarm') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
grid.arrange(p1, p2, p3, p4, p5, p6, ncol=3)
###################################################
### code chunk number 24: distAdjust (eval = FALSE)
###################################################
## library(gridExtra)
## p1<-ggplot(mapping=aes(labs, dat)) +
## geom_quasirandom(bandwidth=2,alpha=.2) +
## ggtitle('bandwidth=2') + labs(x='')
## p2<-ggplot(mapping=aes(labs, dat)) +
## geom_quasirandom(bandwidth=.1,alpha=.2) +
## ggtitle('bandwidth=.1') + labs(x='')
## p3<-ggplot(mapping=aes(labs, dat)) +
## geom_quasirandom(width=.1,alpha=.2) +
## ggtitle('width=.1') + labs(x='')
## p4<-ggplot(mapping=aes(labs, dat)) +
## geom_quasirandom(nbins=100,alpha=.2) +
## ggtitle('nbins=100') + labs(x='')
## grid.arrange(p1, p2, p3, p4, ncol=1)
###################################################
### code chunk number 25: showDistAdjust
###################################################
library(gridExtra)
p1<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(bandwidth=2,alpha=.2) +
ggtitle('bandwidth=2') + labs(x='')
p2<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(bandwidth=.1,alpha=.2) +
ggtitle('bandwidth=.1') + labs(x='')
p3<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(width=.1,alpha=.2) +
ggtitle('width=.1') + labs(x='')
p4<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(nbins=100,alpha=.2) +
ggtitle('nbins=100') + labs(x='')
grid.arrange(p1, p2, p3, p4, ncol=1)
###################################################
### code chunk number 26: nbins (eval = FALSE)
###################################################
## p1<-ggplot(mapping=aes(labs, dat)) +
## geom_quasirandom(method='smiley',alpha=.2) +
## ggtitle('Default (n/5)') + labs(x='')
## p2<-ggplot(mapping=aes(labs, dat)) +
## geom_quasirandom(method='smiley',nbins=50,alpha=.2) +
## ggtitle('nbins=50') + labs(x='')
## p3<-ggplot(mapping=aes(labs, dat)) +
## geom_quasirandom(method='smiley',nbins=100,alpha=.2) +
## ggtitle('nbins=100') + labs(x='')
## p4<-ggplot(mapping=aes(labs, dat)) +
## geom_quasirandom(method='smiley',nbins=250,alpha=.2) +
## ggtitle('nbins=250') + labs(x='')
## grid.arrange(p1, p2, p3, p4, ncol=1)
###################################################
### code chunk number 27: showNBins
###################################################
p1<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',alpha=.2) +
ggtitle('Default (n/5)') + labs(x='')
p2<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',nbins=50,alpha=.2) +
ggtitle('nbins=50') + labs(x='')
p3<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',nbins=100,alpha=.2) +
ggtitle('nbins=100') + labs(x='')
p4<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',nbins=250,alpha=.2) +
ggtitle('nbins=250') + labs(x='')
grid.arrange(p1, p2, p3, p4, ncol=1)
###################################################
### code chunk number 28: varwidth (eval = FALSE)
###################################################
## dat <- list(
## '10 points'=rnorm(10),
## '50 points'=rnorm(50,2),
## '200 points'=c(rnorm(400), rnorm(100,5)),
## '5000 points'= rnorm(5000,1)
## )
## labs<-rep(names(dat),sapply(dat,length))
## labs<-factor(labs,levels=unique(labs))
## dat<-unlist(dat)
## ggplot(mapping=aes(labs, dat)) + geom_quasirandom(alpha=.3,varwidth=TRUE)
###################################################
### code chunk number 29: showVarwidth
###################################################
dat <- list(
'10 points'=rnorm(10),
'50 points'=rnorm(50,2),
'200 points'=c(rnorm(400), rnorm(100,5)),
'5000 points'= rnorm(5000,1)
)
labs<-rep(names(dat),sapply(dat,length))
labs<-factor(labs,levels=unique(labs))
dat<-unlist(dat)
ggplot(mapping=aes(labs, dat)) + geom_quasirandom(alpha=.3,varwidth=TRUE)
###################################################
### code chunk number 30: vpBeaver (eval = FALSE)
###################################################
## beaver<-data.frame(
## 'Temperature'=c(beaver1$temp,beaver2$temp),
## 'Beaver'=rep(
## c('Beaver 1','Beaver 2'),
## c(nrow(beaver1),nrow(beaver2))
## )
## )
## ggplot(beaver,mapping=aes(Beaver, Temperature)) + geom_quasirandom()
###################################################
### code chunk number 31: showBeaver
###################################################
beaver<-data.frame(
'Temperature'=c(beaver1$temp,beaver2$temp),
'Beaver'=rep(
c('Beaver 1','Beaver 2'),
c(nrow(beaver1),nrow(beaver2))
)
)
ggplot(beaver,mapping=aes(Beaver, Temperature)) + geom_quasirandom()
###################################################
### code chunk number 32: vpGene (eval = FALSE)
###################################################
## library(vipor)
## ints<-integrations[integrations$nearestGene>0,]
## ints$logGeneDist<-log(ints$nearestGene)
## ggplot(ints,mapping=aes(study, logGeneDist,color=latent)) +
## geom_quasirandom(dodge.width=.9,alpha=.4)
###################################################
### code chunk number 33: showGene
###################################################
library(vipor)
ints<-integrations[integrations$nearestGene>0,]
ints$logGeneDist<-log(ints$nearestGene)
ggplot(ints,mapping=aes(study, logGeneDist,color=latent)) +
geom_quasirandom(dodge.width=.9,alpha=.4)
ggbeeswarm/inst/doc/usageExamples.Rnw 0000644 0001762 0000144 00000026151 14423030634 017416 0 ustar ligges users \documentclass[article,nojss]{jss}
<>=
options(keep.source = TRUE, width = 60)
packageInfo <- packageDescription("ggbeeswarm")
library(ggbeeswarm)
packageKeywords<-"visualization, display, one dimensional, grouped, groups, violin, scatter, points, quasirandom, beeswarm, van der Corput, beeswarm, ggplot, ggplot2"
@
%\VignetteIndexEntry{ggbeeswarm package usage examples}
%\VignetteDepends{}
%\VignetteKeywords{visualization, display, one dimensional, grouped, groups, violin, scatter, points, quasirandom, beeswarm, van der Corput, beeswarm, ggplot, ggplot2}
%\VignettePackage{ggbeeswarm}
\title{\pkg{\Sexpr{packageInfo$Package}} package usage example (version \Sexpr{packageInfo$Version})}
\author{\Sexpr{packageInfo$Author}}
\Plainauthor{\Sexpr{packageInfo$Author}}
\Address{ Github: \url{http://github.com/eclarke/ggbeeswarm}\\
Cran: \url{https://cran.r-project.org/package=ggbeeswarm}
}
\Keywords{\Sexpr{packageKeywords}}
\Abstract{
This is a collection of examples of usage for the \pkg{\Sexpr{packageInfo$Package}} package.
}
\begin{document}
\SweaveOpts{concordance=TRUE}
\SweaveOpts{engine=R,eps=FALSE}
\section{The basics}
This is the simplest example of using \code{geom_quasirandom} to generate violin scatter plots:
<>=
library(ggbeeswarm)
set.seed(12345)
n<-100
dat<-rnorm(n*2)
labs<-rep(c('a','b'),n)
ggplot(mapping=aes(labs, dat)) + geom_quasirandom()
@
\begin{center}
<>=
<>
@
\end{center}
Normal \code{ggplot} options can be used:
<>=
ggplot(mapping=aes(labs, dat)) + geom_quasirandom(aes(color=labs))
@
\begin{center}
<>=
<>
@
\end{center}
Factors can be used to generate custom group orderings:
<>=
labs2<-factor(labs,levels=c('b','a'))
ggplot(mapping=aes(labs2, dat)) + geom_quasirandom(aes(color=labs))
@
\begin{center}
<>=
<>
@
\end{center}
The axes can also be switched with a categorical y-axis:
<>=
ggplot(mapping=aes(dat,labs)) + geom_quasirandom(aes(color=labs))
@
\begin{center}
<>=
<>
@
\end{center}
And dodging can be used to compare within groups:
<>=
labs2<-factor(rep(1:2,each=n))
ggplot(mapping=aes(labs,dat,color=labs2)) + geom_quasirandom(dodge.width=.8)
@
\begin{center}
<>=
<>
@
\end{center}
Or on the y-axis:
<>=
labs2<-factor(rep(1:2,each=n))
ggplot(mapping=aes(dat,labs,color=labs2)) + geom_quasirandom(dodge.width=.8)
@
\begin{center}
<>=
<>
@
\end{center}
And with \code{geom_beeswarm}:
<>= labs2<-factor(rep(1:2,each=n))
ggplot(mapping=aes(labs,dat,color=labs2)) +
geom_beeswarm(dodge.width=.8,cex=2)
@
\begin{center}
<>=
<>
@
\end{center}
<>=
ggplot(mapping=aes(dat,labs,color=labs2)) +
geom_beeswarm(dodge.width=.8,cex=2)
@
\begin{center}
<>=
<>
@
\end{center}
Both \code{geom_beeswarm} and \code{geom_quasirandom} also work with facets:
<>= labs2<-factor(rep(1:2,each=n))
df<-data.frame(labs,dat,labs2)
ggplot(df,aes(labs,dat,color=labs2)) +
geom_quasirandom() +
facet_grid(.~labs2)
@
\begin{center}
<>=
<>
@
\end{center}
<>= labs2<-factor(rep(1:2,each=n))
ggplot(df,aes(labs,dat,color=labs2)) +
geom_beeswarm(cex=3) +
facet_grid(.~labs2)
@
\begin{center}
<>=
<>
@
\end{center}
\section{Options}
There are several ways to plot grouped one-dimensional data combining points and density estimation including:
\begin{description}
\item[pseudorandom] The kernel density is estimated then points are distributed uniform randomly within the density estimate for a given bin. Selection of an appropriate number of bins does not greatly affect appearance but coincidental clumpiness is common.
\item[alternating within bins] The kernel density is estimated then points are distributed within the density estimate for a given bin evenly spaced with extreme values alternating from right to left e.g. max, 3rd max, ..., 4th max, 2nd max. If maximums are placed on the outside then these plots often form consecutive ``smiley'' patterns. If minimums are placed on the outside then ``frowny'' patterns are generated. Selection of the number of bins can have large effects on appearance important.
\item[tukey] An algorithm described by Tukey and Tukey in ``Strips displaying empirical distributions: I. textured dot strips'' using constrained permutations of offsets to distrbute the data.
\item[quasirandom] The kernel density is estimated then points are distributed quasirandomly using the von der Corput sequence within the density estimate for a given bin. Selection of an appropriate number of bins does not greatly affect appearance and position does not depend on plotting parameters.
\item[beeswarm] The package \pkg{beeswarm} provides methods for generating a ``beeswarm'' plot where points are distibuted so that no points overlap. Kernel density is not calculated although the resulting plot does provide an approximate density estimate. Selection of an appropriate number of bins affects appearance and plot and point sizes must be known in advance.
\end{description}
The first four options are included within \code{geom_quasirandom} using the \code{method=} argument and beeswarm plots are generated with \code{geom_beeswarm}:
<>=
library(gridExtra)
dat <- list(
'Normal'=rnorm(50),
'Dense normal'= rnorm(500),
'Bimodal'=c(rnorm(100), rnorm(100,5)),
'Trimodal'=c(rnorm(100), rnorm(100,5),rnorm(100,-3))
)
labs<-rep(names(dat),sapply(dat,length))
labs<-factor(labs,levels=unique(labs))
dat<-unlist(dat)
p1<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(alpha=.2) +
ggtitle('quasirandom') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p2<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='pseudorandom',alpha=.2) +
ggtitle('pseudorandom') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p3<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',alpha=.2) +
ggtitle('smiley') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p4<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='frowney',alpha=.2) +
ggtitle('frowney') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p5<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='tukey',alpha=.2) +
ggtitle('tukey') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
p6<-ggplot(mapping=aes(labs, dat)) +
geom_beeswarm(alpha=.2,size=.75) +
ggtitle('geom_beeswarm') + labs(x='') +
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
grid.arrange(p1, p2, p3, p4, p5, p6, ncol=3)
@
\begin{center}
<>=
<>
@
\end{center}
\code{quasirandom} calls \code{vipor::offsetX} which calls \code{stats::density} to compute kernel density estimates. The tightness of the fit can be adjusted with the \code{bandwidth} option and the width of the offset with \code{width}. \code{nbins} to adjust the number of bins used in the kernel density is also provided but this can usually be left at its default when using quasirandom offsets but is useful for non-quasirandom methods:
<>=
library(gridExtra)
p1<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(bandwidth=2,alpha=.2) +
ggtitle('bandwidth=2') + labs(x='')
p2<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(bandwidth=.1,alpha=.2) +
ggtitle('bandwidth=.1') + labs(x='')
p3<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(width=.1,alpha=.2) +
ggtitle('width=.1') + labs(x='')
p4<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(nbins=100,alpha=.2) +
ggtitle('nbins=100') + labs(x='')
grid.arrange(p1, p2, p3, p4, ncol=1)
@
\begin{center}
<>=
<>
@
\end{center}
The \code{frowney} or \code{smiley} methods are sensitive to the number of bins so the argument \code{nbins} is more useful/necessary with them:
<>=
p1<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',alpha=.2) +
ggtitle('Default (n/5)') + labs(x='')
p2<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',nbins=50,alpha=.2) +
ggtitle('nbins=50') + labs(x='')
p3<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',nbins=100,alpha=.2) +
ggtitle('nbins=100') + labs(x='')
p4<-ggplot(mapping=aes(labs, dat)) +
geom_quasirandom(method='smiley',nbins=250,alpha=.2) +
ggtitle('nbins=250') + labs(x='')
grid.arrange(p1, p2, p3, p4, ncol=1)
@
\begin{center}
<>=
<>
@
\end{center}
The \code{varwidth} argument scales the width of a group by the square root of the number of observations in that group (as in the function \code{boxplot}):
<>=
dat <- list(
'10 points'=rnorm(10),
'50 points'=rnorm(50,2),
'200 points'=c(rnorm(400), rnorm(100,5)),
'5000 points'= rnorm(5000,1)
)
labs<-rep(names(dat),sapply(dat,length))
labs<-factor(labs,levels=unique(labs))
dat<-unlist(dat)
ggplot(mapping=aes(labs, dat)) + geom_quasirandom(alpha=.3,varwidth=TRUE)
@
\begin{center}
<>=
<>
@
\end{center}
\section{Real data}
An example using the \code{beaver1} and \code{beaver2} data from the \pkg{datasets} package:
<>=
beaver<-data.frame(
'Temperature'=c(beaver1$temp,beaver2$temp),
'Beaver'=rep(
c('Beaver 1','Beaver 2'),
c(nrow(beaver1),nrow(beaver2))
)
)
ggplot(beaver,mapping=aes(Beaver, Temperature)) + geom_quasirandom()
@
\begin{center}
<>=
<>
@
\end{center}
An example using the \code{integrations} data from the \pkg{vipor} package and the argument \code{dodge.width}:
<>=
library(vipor)
ints<-integrations[integrations$nearestGene>0,]
ints$logGeneDist<-log(ints$nearestGene)
ggplot(ints,mapping=aes(study, logGeneDist,color=latent)) +
geom_quasirandom(dodge.width=.9,alpha=.4)
@
\begin{center}
<>=
<>
@
\end{center}
\end{document}
ggbeeswarm/inst/doc/usageExamples.pdf 0000644 0001762 0000144 00006554400 14423302153 017430 0 ustar ligges users %PDF-1.5
%
16 0 obj
<<
/Length 679
>>
stream
concordance:usageExamples.tex:usageExamples.Rnw:1 2 1 1 7 18 1 1 0 6 1 1 2 1 0 5 1 3 0 1 2 1 1 1 -3 1 7 3 1 1 2 4 0 1 2 1 1 1 -3 1 7 2 1 1 2 1 0 1 1 3 0 1 2 1 -2 1 6 2 1 1 2 4 0 1 2 1 -2 1 6 2 1 1 2 1 0 1 1 3 0 1 2 1 -2 1 6 1 1 1 2 1 0 1 1 3 0 1 2 1 -2 1 6 2 1 1 3 5 0 1 2 1 -2 1 6 1 3 5 0 1 2 1 -2 1 6 2 1 1 2 1 0 1 3 5 0 1 2 1 -2 1 6 1 4 6 0 1 2 1 -2 1 6 17 1 1 2 1 0 1 6 5 0 3 1 1 4 3 0 1 4 3 0 1 4 3 0 1 4 3 0 1 4 3 0 1 4 3 0 1 1 3 0 1 2 1 -2 1 6 4 1 1 2 1 0 1 3 2 0 1 3 2 0 1 3 2 0 1 3 2 0 1 1 3 0 1 2 1 -2 1 6 3 1 1 4 3 0 1 3 2 0 1 3 2 0 1 3 2 0 1 1 3 0 1 2 1 -2 1 6 2 1 1 7 6 0 4 1 3 0 1 2 1 -2 1 6 3 1 1 8 7 0 1 1 3 0 1 2 1 -2 1 6 2 1 1 2 1 0 2 1 1 2 4 0 1 2 1 -2 1 6 3 1
endstream
endobj
19 0 obj
<<
/Length 991
/Filter /FlateDecode
>>
stream
xڭVMo6WVl m-Z`AAK#D Qv6 W*mppmAȳ\"cF@[ ιF9#RNM"=~ެ? TA #hoʔlm\&wfoO"1,K3]. ~qFygX2Ы}qmEؖOe7 q{gnȤxugtwƛ~tSr~zXQHA)I
A? .㨵CBx}Z}\HcfƲ`H3
)1& _
s%aS`sΓi5~I5-Pwats
4 Emׇ
K(2]^Yb)jXN4\Ik Pxr'73X(6L3Rr4!NxJ "Z Uuīm չ&.
um< x
Юn <@7 Kp0M[uP|
uLJLU}h 6@ݻ!
Ë
!m07R`4\FSf rl
<\,mFs>D/wY;
\=R}8c~Йv(h'nmkV,bbmk{3D' 8"A']`Y0"v`< 2ԛO~C/RC ᜑk9;`gmQq!~6 oFߤW8@ؽ/O|:*
A7KI,m\d#E{#d Ӹ3={c?>
/ExtGState <<
>>/ColorSpace <<
/sRGB 34 0 R
>>>>
/Length 12422
/Filter /FlateDecode
>>
stream
x}ˎ-Kn|E
J̜v @nA4T6wp=Q
$bpdD( Gۿ(u}Jo~?O?_~\ OSG?-hn |~?j音}&^o a p} >~?G/?~#=>{InqY墏YjO~@t|#4ɭ o1ÚykH H a.i[z|_&k@.<%7.mX`) 7㭟5_}hlOJg㳿CVڤTwǯև}BSU>=ىOwn%E8߉+?Ͷ-v7~4]y>{$g #϶ԎYnSl@Jy,mcEAONlAjrE7]ZZ/9˺ϹZ_3e4ftMQho:+
iqn"첮Mfۖ.aO\jRs빈 po0ǽه)1$VtXnPږ\v {ڭrEk8χ\^>G7
&qm,)y0ƅ]MfےRb٭ﲈ`_ϪԲKoSf[~+
iq29?Ͷ%]~e-Kta:9VjF&'\iXQS .`Ʈ~2nPږ\v {U}~mF.>+(4IL5XPTaUOYڵ6
wGur <ğ~MyŘ6Vty31;0?YmK.=wY]̥g?/y1cfgg;~@cEA7?-V,6
Vے.aO] ."b\ȚN87z7;!q" Of`j/$EN[=4jHn9D]~"ގμF% [:c2d\5waI+~mJJk;l>mNTO9Jny AA'p֨ v
ږ\v {ˠC4(xbj\8 ckxLD[kTisa|g"g 37b>VЪ6)5ĝ Fak0ߌJ -T ynKCD fpߍk[kb65}sCi}Y|bk;8l7zkbMaNJolJo%h-" 7+=8 Õ>(H> 6n~Z&-tyGږ\vU Χ5 qԬK XLi{
{i%QmJ*k;mգ(Tu34P3U NQFM))ub'T۔.bv,S.9j4ɇryͱs
E0Vԝas&b.J9Ueے.aOc=5wKb)$gp%װ[ [ڢG_+Fa ^^ԪMJe
qc|@~Xٗ tHҊ8Њ1b5nl˶ Vdx.n)%]?J5z? |94zގ |p26;ژ Xd.NLݖwsƊn(}$=e27[vdžcu͗! 3V p-b
gr ƆH14D@+>*Fh^&Vu8}A`4N6
sF5Į-_œXSîur:F
ڦv]^9o~W GV{e彷\cb{7 8˺Q@Ԕ.b!rW o@ ikTE7^۬_ W75ݞ{x]R7)C~.a5mMdP`VPXQ9&gxvH-mm[ot7v_g=X@=AL7+`IQ?D0[؈M,y-r\e6GF_c<}תV~}v豌-fmR itKTAc|(`9 ?J\^L.NMݤ .aOwӏ; |R|e$!6ɰy}Փ͂ 34E⸪c)_ĚnoUـ&啳_[~(v eK$yˌ-O츍#13)`#Q+7FV7Q(P`0mfܶK9HwЕiLOc#_rg_!cE,q*1a;6uQm[N=x}.g5 _韁 37a9 ƒ!s*km m6oٶT \0Z0nU R3f-lwsgnR[.=햻$yj{O6%5 zb<
}ƅ-KG#hyeGXSmO .,S7(o ѶKn
fn_'`OS5VnۗKޔ"t{Tk>]J+u|1 {n\"h12L1SsD\GlYV Ϋli8%U1.Vx >| jSRYCi TElb95\ RΧ,D6ySG
=v߁` h_0)#rnԍ48
3L}X7(KC_ɣa'k
rw;_LƊ{rfwۖ\v { ~KTe5N* , /2DO F n֚ڔ4',xl1/FA2
N/7'"#f ,\CYZ[զ֝\7}2^:-wg5˕R췍Ah0N.a$<6-y; }VB?)̺U:ŕ~DcXYE|ЙVN&s#U&¼v{ڽ+b
~ OLylS+'i$Z2d6%5 zN ys'' f AʓM2sc'ʘèSOmwP`ݻ{GȨ_vE`5 'Zۃ
&+u )!˶pLƃ{Z vf"XSe!q./9YgṶ<'{Mvk_Qx>uF oqn=MhF#AȽ
WQt$5kԦ2]T Sܻ rZ٬(.S"GXbXPO= ѩf9&bR4%~yed,97S
Vvfe_S-G~"f>19axt=ܶK n:9ng¸VSiתVTϳNX*'-jC7ݏ xokX?~()&,P$eО8K0m'Omakf>"N K:go
u:H}b\p:Lmof 橹|Hkr{ uOŢ^ފ\Pr+ɩٜ]HAEvoԶK \Sy pu9u* jRz>31ĚYce3+O+me9*S:5CqAť"4*'瑸~5^p
Ƽ;4Dd
e`=Ն U g?1c~2ȭ9J6)5ĝAkQY.|2f+) '|׳?q$Z@+/7|Eh݅jSRZ#aQҖ(^
oҠi~l/Qq]BbE]Yy$2WMlMʁhreiR1Y%s1Q
Q(-:$0}W2VTܒ9(N u_qے.aO5**vO^C+:J aC Oʓ=Ɗ껰J捡ٺkBݶKn?:z _'n@/g}r3w
~2`9ƊZ C؎i=ۦvn?/ȴߗ l,zMS9-oX553f`
u,u۔.bR;\Vur*ΪǑc.k R}XQb c?Ͷ-]vL ܘ'YqQP&#ߌkjF(M9Sdے.aO]x!R(ѣT {ii0 Ϟ;q$7fx%k+ܦFV ak%%=ފw@>w] [|"ff,95wb{Rwn-ۖ\v {2o%1U6F-_Z2* K55$Ģt#ugeG-{OX:q#kpE2_SbOS 4vQK7)O]:\r=i.M'Gξ \# U(i\s6V9 D<`hXi-ys:~.~ :yR75jFZΫ%uO
ͬ:4-vVz<ٻ9c/7sz$L >e)GƑhQ}5Ěa) )) jbrSwryVVET^S~R^ {ˑK,"MN}Ƣ3ugi-K`}K"וia=Kl@#"$p/ki5g{TF-";KUէKw)zE~Z{ R\Ee 8wiŇ;ݦFV,rT#aOopPSFEqwr+0i] ζ%]v^cVcwɬX`KE{mw r($Vt"ۜ
̺kkn[r%i'ۋ5
KISᆣq/NSޡF
.W*6nR>.cO#4\H~ w/ (x-pɍU:s~Mb8r'%]v;/H]÷]L߸R]L#bvHnք]SÞĖ5ul[r%i6<ѵ,+|īZZ7M9kZ4dE=P~iLM"G.K }2po9ǯyrr cyۖܖn/)%Gms`a#+jdv*Tۖ\v
A /y5nOMhUl#{ا"dy,XQyjrpUpbK|;i-]]8c_y3?Z+]t)gYF1s{SwXۖ\v {}s}<`jQ<Ti"kI} 4h~J
KJf]e_'ea5.QV)NJ|Sk_ObM LƖɭ%u&[.f ñxc6BdgH9LBXK!E_?Ͷ-]ƞv/.3h5D%@e+
ڻrT__;#vX[͖ۖ|t`O=T^께0 ܋1+/xr+
<^;"Aj[jg~a8WDZɽtt m#AY+7k%R_\V?+
\/ " <?U)IG"/35os[icF5dwM &lXg^$wrtv)XcMA7?-.dO,N-"ƑW L6[mbxmr^f Է867[ms5ƞvI(T夈sMGZ W>(O*)8gV$ƚ:pKww]j-" HoE'G\=A⧶ewO)oƑT!RcqnRYCiN&wAV) qa& Tʖv4cEa0ac,ԝqme;ޫdzW;^&~~yqvXQ=.iR.Uے.aO=uM᾿J7<8_)8i< xIM&Q2`aNu^95 bLQRp;xeʩ~o9+D88elő-"Vm߅KəYu\]g.|\r9XQ$YX2uݶKnޖ)ws
_y Ի27c')wsXi,+{WfmmsAϮu"eRǥ/ћ"2- mX8ϮrCJQBqъТ]< d?upwXQyVr1ƻwGme{ȋS-b &ŭ"r-C8wnĘ#p\<զ.OU )mYز˳E^H5Fs/wQ Fʳc''Fsn;~唙Ilr%iWD>w
1^xYI9'YD9*`dE?j9*Ԧֽpu+3k
$<ݬ%RT}G"_LjW
K;5_L6 FVuޛtq&1XyAf1#x匵.sxmצJmeݻA/kv ܁: rື|uFw" a@˺Gn ݶͯ;t.u0:˂D])ŹNHfB8\Sdy3۔wغUt8w<q8novN[ ʣ.}7gfvm NjזgS}efmK.=xzɹ9J_wl /,xr6Pmrޗ jv
]b} #~q{J窦22sm9oVT~9؆v})':9YJ7`#mK.fc7[9e>}}VS5ʻ5X4^[w"ۦKv; |Q
:joRnTr\ SpZ?n[r%7.{H×n(W#WY~zטsޱ0l~M,y.aOdk{<]qy3UU o_@UuU+[XScWә*q&moԶ䲋ovsBr ^17(ܙxF+jds:Ƣ9u*g-vtmzmt},/mge >26~ʵmZΨl[So`_a42 =*[Y÷0İ;bFn N[kmJJk;l݁R+[el Ԯ- Fz9Q fP}*d[l[^V+zl>ޞ5Ďƻ
Dq)AĊ7g flgNEιLTE9%iQf:SPG1x{n8woT6|%VȓtL2a-{L^m
crBȬ0o{/
crJ-5qin,8Rw^mK.=[U[UՕ/BGfr^
" %#z NX O5me;0TDxea0:}sNeQlyV5 #dږ[bgݻv~aiHΠSL;m/m~2+J{q0[TIwݕj[r%iw|?ů>~G?^O>_?~ P #㾧BdcJGDN%bco@<L͏(- 7 `~\ |GA+!_<&{dCdԿ'Ԗ
'"]gJȐ)=?#-g2Kݤ`/VHn7,"s 莢;#V`L,U(Zv4Zxg_ b
endstream
endobj
36 0 obj
<<
/Alternate /DeviceRGB
/N 3
/Length 2596
/Filter /FlateDecode
>>
stream
xwTSϽ7PkhRH
H.*1 J "6DTpDQ2(C"QDqpIdy͛ ~kg}ֺ LX Xňg` l pBF|،l *? Y"1 P\8=W%Oɘ4M0J"Y2Vs,[|e92<se'9`2&ctI@o|N6 (.sSdl-c(2-y H_/XZ.$&\SM07#1ؙY r fYym";8980m-m(]v^D W~
emi ]P`/ u}q |^R,g+\Kk)/C_|Rax8t1C^7nfzDp 柇 u $/ED˦L L[B@ٹЖX!@~ (* {d+} G͋љς}WL$cGD2QZ4 E@@ A(q`1D `'u46ptc48.`R0)
@Rt C XCP %CBH@R f[(t
CQhz#0 Zl`O828 .p|O×X
?:0FBx$ !i@ڐ H[EE1PL ⢖V6QP>U(j
MFk t,:.FW 8c1L &ӎ9ƌaX:
rbl1
{{{;}#tp8_\8"Ey.,X%%Gщ1-9ҀKl.oo/O$&'=JvMޞ