Boxplots

Boxplots can be another useful tool for visualizing a dataset, showing both the spread of the data (range of most of the data, presence of outliers) and how the data is concentrated in the middle (mean, median).

Boxplots are often most useful to compare data across categories. For example, you can construct a boxplot for penguin weight (body mass in g) across all islands:

ggplot(data = penguins, mapping  = aes(y = body_mass_g)) +
  geom_boxplot()

But it may be more informative to break that data apart by island:

ggplot(data = penguins, mapping  = aes(x = island, y = body_mass_g)) +
  geom_boxplot()

You can create multiple boxplots on the same graph by setting x to a variable with multiple categories. As seen in the previous examples, you can also use color to illustrate additional variables within one visual:

ggplot(data = penguins, mapping  = aes(x = island, y = body_mass_g, color = sex)) +
  geom_boxplot()

(The dots you see as part of the boxplots are outliers, which geom_boxplot defines as being outside of 150% the value of the interquartile range [IQR], the difference between the 1st [25%] and the 3rd [75%] quartile of your data.)

A common problem when creating multiple boxplots on the same graph is that even though you supply an x-variable, your plot only has one boxplot. This usually happens when the x-variable is not being properly treated as a factor. One option for a fix is to encode the variable as a factor within the ggplot call:

ggplot(data = penguins, mapping  = aes(x = factor(island), y = body_mass_g)) +
  geom_boxplot()

## Warning: Removed 2 rows containing non-finite values (stat_boxplot).