| geom_bar {ggplot2} | R Documentation |
The bar geom is used to produce 1d area plots: bar charts for categorical
x, and histograms for continuous y. stat_bin explains the details of
these summaries in more detail. In particular, you can use the
weight aesthetic to create weighted histograms and barcharts where
the height of the bar no longer represent a count of observations, but a
sum over some other variable. See the examples for a practical
example.
geom_bar(mapping = NULL, data = NULL, stat = "bin", position = "stack", ...)
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 overlapping points on this layer |
... |
other arguments passed on to |
The heights of the bars commonly represent one of two things: either a
count of cases in each group, or the values in a column of the data frame.
By default, geom_bar uses stat="bin". This makes the height
of each bar equal to the number of cases in each group, and it is
incompatible with mapping values to the y aesthetic. If you want
the heights of the bars to represent values in the data, use
stat="identity" and map a value to the y aesthetic.
By default, multiple x's occuring in the same place will be stacked a top
one another by position_stack. If you want them to be dodged from
side-to-side, see position_dodge. Finally,
position_fill shows relative propotions at each x by stacking
the bars and then stretching or squashing to the same height.
Sometimes, bar charts are used not as a distributional summary, but
instead of a dotplot. Generally, it's preferable to use a dotplot (see
geom_point) as it has a better data-ink ratio. However, if you do
want to create this type of plot, you can set y to the value you have
calculated, and use stat='identity'
A bar chart maps the height of the bar to a variable, and so the base of the bar must always been shown to produce a valid visual comparison. Naomi Robbins has a nice article on this topic. This is the reason it doesn't make sense to use a log-scaled y axis with a bar chart
geom_bar understands the following aesthetics (required aesthetics are in bold):
x
alpha
colour
fill
linetype
size
weight
stat_bin for more details of the binning alogirithm,
position_dodge for creating side-by-side barcharts,
position_stack for more info on stacking,
# Generate data
c <- ggplot(mtcars, aes(factor(cyl)))
# By default, uses stat="bin", which gives the count in each category
c + geom_bar()
c + geom_bar(width=.5)
c + geom_bar() + coord_flip()
c + geom_bar(fill="white", colour="darkgreen")
# Use qplot
qplot(factor(cyl), data=mtcars, geom="bar")
qplot(factor(cyl), data=mtcars, geom="bar", fill=factor(cyl))
# When the data contains y values in a column, use stat="identity"
library(plyr)
# Calculate the mean mpg for each level of cyl
mm <- ddply(mtcars, "cyl", summarise, mmpg = mean(mpg))
ggplot(mm, aes(x = factor(cyl), y = mmpg)) + geom_bar(stat = "identity")
# Stacked bar charts
qplot(factor(cyl), data=mtcars, geom="bar", fill=factor(vs))
qplot(factor(cyl), data=mtcars, geom="bar", fill=factor(gear))
# Stacked bar charts are easy in ggplot2, but not effective visually,
# particularly when there are many different things being stacked
ggplot(diamonds, aes(clarity, fill=cut)) + geom_bar()
ggplot(diamonds, aes(color, fill=cut)) + geom_bar() + coord_flip()
# Faceting is a good alternative:
ggplot(diamonds, aes(clarity)) + geom_bar() +
facet_wrap(~ cut)
# If the x axis is ordered, using a line instead of bars is another
# possibility:
ggplot(diamonds, aes(clarity)) +
geom_freqpoly(aes(group = cut, colour = cut))
# Dodged bar charts
ggplot(diamonds, aes(clarity, fill=cut)) + geom_bar(position="dodge")
# compare with
ggplot(diamonds, aes(cut, fill=cut)) + geom_bar() +
facet_grid(. ~ clarity)
# But again, probably better to use frequency polygons instead:
ggplot(diamonds, aes(clarity, colour=cut)) +
geom_freqpoly(aes(group = cut))
# Often we don't want the height of the bar to represent the
# count of observations, but the sum of some other variable.
# For example, the following plot shows the number of diamonds
# of each colour
qplot(color, data=diamonds, geom="bar")
# If, however, we want to see the total number of carats in each colour
# we need to weight by the carat variable
qplot(color, data=diamonds, geom="bar", weight=carat, ylab="carat")
# A bar chart used to display means
meanprice <- tapply(diamonds$price, diamonds$cut, mean)
cut <- factor(levels(diamonds$cut), levels = levels(diamonds$cut))
qplot(cut, meanprice)
qplot(cut, meanprice, geom="bar", stat="identity")
qplot(cut, meanprice, geom="bar", stat="identity", fill = I("grey50"))
# Another stacked bar chart example
k <- ggplot(mpg, aes(manufacturer, fill=class))
k + geom_bar()
# Use scales to change aesthetics defaults
k + geom_bar() + scale_fill_brewer()
k + geom_bar() + scale_fill_grey()
# To change plot order of class varible
# use factor() to change order of levels
mpg$class <- factor(mpg$class, levels = c("midsize", "minivan",
"suv", "compact", "2seater", "subcompact", "pickup"))
m <- ggplot(mpg, aes(manufacturer, fill=class))
m + geom_bar()