Merge branch 'boxplot_varwidth3' of github.com:tsieger/ggplot2 into tsieger-boxplot_varwidth3

Conflicts:
	man/geom_boxplot.Rd

Author: hadley

NEWS

@@ -27,6 +27,9 @@ ggplot2 0.9.3.1.99
   fledged mapping projection as `coord_map()` does and therefore is much
   faster. Yet, it gives a correct approximation for small scale maps (a few
   degrees in either direction) at medium to low latitudes (@jiho, #922).
+* `geom_boxplot` gain new `varwidth` argument for controlling whether or not 
+  the width of boxplots should be proportional to the size of the groups 
+  (@tsieger, #927).
 
 * Allow specifying only one of the limits in a scale and use the automatic
   calculation of the other limit by passing NA to to the limit function,

R/geom-boxplot.r

@@ -32,6 +32,10 @@
 #'    the medians differ.
 #' @param notchwidth for a notched box plot, width of the notch relative to
 #'    the body (default 0.5)
+#' @param varwidth if \code{FALSE} (default) make a standard box plot. If
+#'    \code{TRUE}, boxes are drawn with widths proportional to the
+#'    square-roots of the number of observations in the groups (possibly
+#'    weighted, using the \code{weight} aesthetic).
 #' @export
 #'
 #' @references McGill, R., Tukey, J. W. and Larsen, W. A. (1978) Variations of
@@ -92,13 +96,17 @@
 #' b + geom_boxplot(stat = "identity")
 #' b + geom_boxplot(stat = "identity") + coord_flip()
 #' b + geom_boxplot(aes(fill = X1), stat = "identity")
+#'
+#' # Using varwidth
+#' p + geom_boxplot(varwidth = TRUE)
+#' qplot(factor(cyl), mpg, data = mtcars, geom = "boxplot", varwidth = TRUE)
 #' }
 geom_boxplot <- function (mapping = NULL, data = NULL, stat = "boxplot", position = "dodge",
 outlier.colour = "black", outlier.shape = 16, outlier.size = 2,
-notch = FALSE, notchwidth = .5, ...) {
+notch = FALSE, notchwidth = .5, varwidth = FALSE, ...) {
   GeomBoxplot$new(mapping = mapping, data = data, stat = stat,
   position = position, outlier.colour = outlier.colour, outlier.shape = outlier.shape,
-  outlier.size = outlier.size, notch = notch, notchwidth = notchwidth, ...)
+  outlier.size = outlier.size, notch = notch, notchwidth = notchwidth, varwidth = varwidth, ...)
 }
 
 GeomBoxplot <- proto(Geom, {
@@ -118,14 +126,24 @@ GeomBoxplot <- proto(Geom, {
       df$ymax_final <- pmax(out_max, df$ymax)
     }
 
-    transform(df,
-      xmin = x - width / 2, xmax = x + width / 2, width = NULL
-    )
+    # if `varwidth` not requested or not available, don't use it
+    if (is.null(params) || is.null(params$varwidth) || !params$varwidth || is.null(df$relvarwidth)) {
+      df$xmin <- df$x - df$width / 2
+      df$xmax <- df$x + df$width / 2
+    } else {
+      # make `relvarwidth` relative to the size of the largest group
+      df$relvarwidth <- df$relvarwidth / max(df$relvarwidth)
+      df$xmin <- df$x - df$relvarwidth * df$width / 2
+      df$xmax <- df$x + df$relvarwidth * df$width / 2
+    }
+    df$width <- NULL
+    if (!is.null(df$relvarwidth)) df$relvarwidth <- NULL
 
+    df
   }
 
   draw <- function(., data, ..., fatten = 2, outlier.colour = NULL, outlier.shape = NULL, outlier.size = 2,
-                   notch = FALSE, notchwidth = .5) {
+                   notch = FALSE, notchwidth = .5, varwidth = FALSE) {
     common <- data.frame(
       colour = data$colour,
       size = data$size,

R/stat-boxplot.r

@@ -76,7 +76,8 @@ StatBoxplot <- proto(Stat, {
 
       transform(df,
         x = if (is.factor(x)) x[1] else mean(range(x)),
-        width = width
+        width = width,
+        relvarwidth = sqrt(n)
       )
     })
   }

man/geom_boxplot.Rd

@@ -5,39 +5,49 @@
 \usage{
 geom_boxplot(mapping = NULL, data = NULL, stat = "boxplot",
   position = "dodge", outlier.colour = "black", outlier.shape = 16,
-  outlier.size = 2, notch = FALSE, notchwidth = 0.5, ...)
+  outlier.size = 2, notch = FALSE, notchwidth = 0.5, varwidth = FALSE,
+  ...)
 }
 \arguments{
-\item{outlier.colour}{colour for outlying points}
+  \item{outlier.colour}{colour for outlying points}
 
-\item{outlier.shape}{shape of outlying points}
+  \item{outlier.shape}{shape of outlying points}
 
-\item{outlier.size}{size of outlying points}
+  \item{outlier.size}{size of outlying points}
 
-\item{notch}{if \code{FALSE} (default) make a standard box plot. If
-\code{TRUE}, make a notched box plot. Notches are used to compare groups;
-if the notches of two boxes do not overlap, this is strong evidence that
-the medians differ.}
+  \item{notch}{if \code{FALSE} (default) make a standard
+  box plot. If \code{TRUE}, make a notched box plot.
+  Notches are used to compare groups; if the notches of two
+  boxes do not overlap, this is strong evidence that the
+  medians differ.}
 
-\item{notchwidth}{for a notched box plot, width of the notch relative to
-the body (default 0.5)}
+  \item{notchwidth}{for a notched box plot, width of the
+  notch relative to the body (default 0.5)}
 
-\item{mapping}{The aesthetic mapping, usually constructed with
-\code{\link{aes}} or \code{\link{aes_string}}. Only needs to be set
-at the layer level if you are overriding the plot defaults.}
+  \item{varwidth}{if \code{FALSE} (default) make a standard
+  box plot. If \code{TRUE}, boxes are drawn with widths
+  proportional to the square-roots of the number of
+  observations in the groups (possibly weighted, using the
+  \code{weight} aesthetic).}
 
-\item{data}{A layer specific dataset - only needed if you want to override
-the plot defaults.}
+  \item{mapping}{The aesthetic mapping, usually constructed
+  with \code{\link{aes}} or \code{\link{aes_string}}. Only
+  needs to be set at the layer level if you are overriding
+  the plot defaults.}
 
-\item{stat}{The statistical transformation to use on the data for this
-layer.}
+  \item{data}{A layer specific dataset - only needed if you
+  want to override the plot defaults.}
 
-\item{position}{The position adjustment to use for overlapping points
-on this layer}
+  \item{stat}{The statistical transformation to use on the
+  data for this layer.}
 
-\item{...}{other arguments passed on to \code{\link{layer}}. This can
-include aesthetics whose values you want to set, not map. See
-\code{\link{layer}} for more details.}
+  \item{position}{The position adjustment to use for
+  overlapping points on this layer}
+
+  \item{...}{other arguments passed on to
+  \code{\link{layer}}. This can include aesthetics whose
+  values you want to set, not map. See \code{\link{layer}}
+  for more details.}
 }
 \description{
 The upper and lower "hinges" correspond to the first and third quartiles
@@ -58,8 +68,8 @@ This gives a roughly 95% confidence interval for comparing medians.
 See McGill et al. (1978) for more details.
 }
 \section{Aesthetics}{
-
-\Sexpr[results=rd,stage=build]{ggplot2:::rd_aesthetics("geom", "boxplot")}
+  \Sexpr[results=rd,stage=build]{ggplot2:::rd_aesthetics("geom",
+  "boxplot")}
 }
 \examples{
 \donttest{
@@ -116,6 +126,10 @@ b <- ggplot(abc, aes(x = X1, ymin = `0\%`, lower = `25\%`, middle = `50\%`, uppe
 b + geom_boxplot(stat = "identity")
 b + geom_boxplot(stat = "identity") + coord_flip()
 b + geom_boxplot(aes(fill = X1), stat = "identity")
+
+# Using varwidth
+p + geom_boxplot(varwidth = TRUE)
+qplot(factor(cyl), mpg, data = mtcars, geom = "boxplot", varwidth = TRUE)
 }
 }
 \references{