geom_boxplot {ggplot2} | R Documentation |
The upper and lower "hinges" correspond to the first and
third quartiles (the 25th and 75th percentiles). This
differs slightly from the method used by the
boxplot
function, and may be apparent with small
samples. See boxplot.stats
for for more
information on how hinge positions are calculated for
boxplot
.
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.colour |
colour for outlying points |
outlier.shape |
shape of outlying points |
outlier.size |
size of outlying points |
notch |
if |
notchwidth |
for a notched box plot, width of the notch relative to the body (default 0.5) |
mapping |
The aesthetic mapping, usually constructed
with |
data |
A layer specific dataset - only needed if you want to override the plot defaults. |
stat |
The statistical transformation to use on the data for this layer. |
position |
The position adjustment to use for overlappling points on this layer |
... |
other arguments passed on to
|
The upper whisker extends from the hinge to the highest value that is within 1.5 * IQR of the hinge, where IQR is the inter-quartile range, or distance between the first and third quartiles. The lower whisker extends from the hinge to the lowest value within 1.5 * IQR of the hinge. Data beyond the end of the whiskers are outliers and plotted as points (as specified by Tukey).
In a notched box plot, the notches extend 1.58 *
IQR / sqrt(n)
. This gives a roughly 95
interval for comparing medians. See McGill et al. (1978)
for more details.
geom_boxplot
understands the following aesthetics (required aesthetics are in bold):
lower
middle
upper
x
ymax
ymin
alpha
colour
fill
linetype
shape
size
weight
McGill, R., Tukey, J. W. and Larsen, W. A. (1978) Variations of box plots. The American Statistician 32, 12-16.
stat_quantile
to view quantiles conditioned
on a continuous variable, geom_jitter
for
another way to look at conditional distributions"
p <- ggplot(mtcars, aes(factor(cyl), mpg)) p + geom_boxplot() qplot(factor(cyl), mpg, data = mtcars, geom = "boxplot") p + geom_boxplot() + geom_jitter() p + geom_boxplot() + coord_flip() qplot(factor(cyl), mpg, data = mtcars, geom = "boxplot") + coord_flip() p + geom_boxplot(notch = TRUE) p + geom_boxplot(notch = TRUE, notchwidth = .3) p + geom_boxplot(outlier.colour = "green", outlier.size = 3) # Add aesthetic mappings # Note that boxplots are automatically dodged when any aesthetic is # a factor p + geom_boxplot(aes(fill = cyl)) p + geom_boxplot(aes(fill = factor(cyl))) p + geom_boxplot(aes(fill = factor(vs))) p + geom_boxplot(aes(fill = factor(am))) # Set aesthetics to fixed value p + geom_boxplot(fill = "grey80", colour = "#3366FF") qplot(factor(cyl), mpg, data = mtcars, geom = "boxplot", colour = I("#3366FF")) # Scales vs. coordinate transforms ------- # Scale transformations occur before the boxplot statistics are computed. # Coordinate transformations occur afterwards. Observe the effect on the # number of outliers. library(plyr) # to access round_any m <- ggplot(movies, aes(y = votes, x = rating, group = round_any(rating, 0.5))) m + geom_boxplot() m + geom_boxplot() + scale_y_log10() m + geom_boxplot() + coord_trans(y = "log10") m + geom_boxplot() + scale_y_log10() + coord_trans(y = "log10") # Boxplots with continuous x: # Use the group aesthetic to group observations in boxplots qplot(year, budget, data = movies, geom = "boxplot") qplot(year, budget, data = movies, geom = "boxplot", group = round_any(year, 10, floor)) # Using precomputed statistics # generate sample data abc <- adply(matrix(rnorm(100), ncol = 5), 2, quantile, c(0, .25, .5, .75, 1)) b <- ggplot(abc, aes(x = X1, ymin = `0%`, lower = `25%`, middle = `50%`, upper = `75%`, ymax = `100%`)) b + geom_boxplot(stat = "identity") b + geom_boxplot(stat = "identity") + coord_flip() b + geom_boxplot(aes(fill = X1), stat = "identity")