geom_histogram is an alias for geom_bar plus
stat_bin so you will need to look at the documentation for
those objects to get more information about the parameters.geom_histogram(mapping = NULL, data = NULL, stat = "bin",
position = "stack", ...)aes or aes_string. Only needs to be set
at the layer level if you are overriding the plot defaults.# More complex m <- ggplot(movies, aes(x=rating)) m + geom_histogram() m + geom_histogram(aes(y = ..density..)) + geom_density()
m + geom_histogram(binwidth = 1) m + geom_histogram(binwidth = 0.5) m + geom_histogram(binwidth = 0.1)
# Add aesthetic mappings m + geom_histogram(aes(weight = votes)) m + geom_histogram(aes(y = ..count..)) m + geom_histogram(aes(fill = ..count..))
# Change scales m + geom_histogram(aes(fill = ..count..)) + scale_fill_gradient("Count", low = "green", high = "red")
# 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 movies # in each rating. qplot(rating, data=movies, geom="bar", binwidth = 0.1) # If, however, we want to see the number of votes cast in each # category, we need to weight by the votes variable qplot(rating, data=movies, geom="bar", binwidth = 0.1, weight=votes, ylab = "votes")
m <- ggplot(movies, aes(x = votes)) # For transformed scales, binwidth applies to the transformed data. # The bins have constant width on the transformed scale. m + geom_histogram() + scale_x_log10() m + geom_histogram(binwidth = 1) + scale_x_log10() m + geom_histogram() + scale_x_sqrt() m + geom_histogram(binwidth = 10) + scale_x_sqrt()
# For transformed coordinate systems, the binwidth applies to the # raw data. The bins have constant width on the original scale.
# Using log scales does not work here, because the first # bar is anchored at zero, and so when transformed becomes negative # infinity. This is not a problem when transforming the scales, because # no observations have 0 ratings. m + geom_histogram(origin = 0) + coord_trans(x = "log10") # Use origin = 0, to make sure we don't take sqrt of negative values m + geom_histogram(origin = 0) + coord_trans(x = "sqrt") m + geom_histogram(origin = 0, binwidth = 1000) + coord_trans(x = "sqrt")
# You can also transform the y axis. Remember that the base of the bars # has value 0, so log transformations are not appropriate m <- ggplot(movies, aes(x = rating)) m + geom_histogram(binwidth = 0.5) + scale_y_sqrt() m + geom_histogram(binwidth = 0.5) + scale_y_reverse()
# Set aesthetics to fixed value m + geom_histogram(colour = "darkgreen", fill = "white", binwidth = 0.5)
# Use facets m <- m + geom_histogram(binwidth = 0.5) m + facet_grid(Action ~ Comedy)
# Often more useful to use density on the y axis when facetting m <- m + aes(y = ..density..) m + facet_grid(Action ~ Comedy) m + facet_wrap(~ mpaa)
# Multiple histograms on the same graph # see ?position, ?position_fill, etc for more details. set.seed(6298) diamonds_small <- diamonds[sample(nrow(diamonds), 1000), ] ggplot(diamonds_small, aes(x=price)) + geom_bar() hist_cut <- ggplot(diamonds_small, aes(x=price, fill=cut)) hist_cut + geom_bar() # defaults to stacking hist_cut + geom_bar(position="fill") hist_cut + geom_bar(position="dodge")
# This is easy in ggplot2, but not visually effective. It's better # to use a frequency polygon or density plot. Like this: ggplot(diamonds_small, aes(price, ..density.., colour = cut)) + geom_freqpoly(binwidth = 1000) # Or this: ggplot(diamonds_small, aes(price, colour = cut)) + geom_density() # Or if you want to be fancy, maybe even this: ggplot(diamonds_small, aes(price, fill = cut)) + geom_density(alpha = 0.2) # Which looks better when the distributions are more distinct ggplot(diamonds_small, aes(depth, fill = cut)) + geom_density(alpha = 0.2) + xlim(55, 70) rm(movies)
stat_bin uses 30 bins - this is not a good default,
but the idea is to get you experimenting with different binwidths. You
may need to look at a few to uncover the full story behind your data.