Controlling Legends in ggplot2

Legends are crucial for interpreting visualizations, but sometimes you need fine-grained control over which legends are displayed in your ggplot2 plots. This tutorial covers several techniques for removing, suppressing, or customizing legends in your plots.

Understanding Legend Behavior

By default, ggplot2 automatically creates legends for aesthetics (like color, fill, shape, linetype) that vary within your data. When the same variable is mapped to multiple aesthetics, ggplot2 often combines those legends into a single entry. However, you can override this default behavior.

Removing All Legends

The simplest approach is to remove all legends from a plot. This is useful when the plot is self-explanatory or when the legends clutter the visualization. Use the theme() function to set legend.position = "none".

library(ggplot2)

# Sample data
data <- mtcars
ggplot(data, aes(x = mpg, y = disp, group = gear)) +
  geom_point(aes(color = vs)) +
  geom_point(aes(shape = factor(cyl))) +
  geom_line(aes(linetype = factor(gear))) +
  geom_smooth(aes(fill = factor(gear), color = gear)) + 
  theme_bw() +
  theme(legend.position = "none")

This code creates a plot identical to the original, but without any legends.

Removing Specific Legends

Often, you’ll want to remove only certain legends, leaving others intact. There are multiple ways to achieve this:

1. Using guides()

The guides() function provides a flexible way to control individual legends. You specify the aesthetic (color, fill, linetype, shape, etc.) and set its guide to "none".

ggplot(data, aes(x = mpg, y = disp, group = gear)) +
  geom_point(aes(color = vs)) +
  geom_point(aes(shape = factor(cyl))) +
  geom_line(aes(linetype = factor(gear))) +
  geom_smooth(aes(fill = factor(gear), color = gear)) + 
  theme_bw() +
  guides(fill = "none", color = "none", linetype = "none", shape = "none")

This removes the legends for fill, color, linetype, and shape.

2. Using scale_..._discrete(guide = "none")

This approach is specific to discrete scales. It works by setting the guide for a particular scale to "none". For instance, to remove the legend associated with the fill aesthetic:

ggplot(data, aes(x = mpg, y = disp, group = gear)) +
  geom_point(aes(color = vs)) +
  geom_point(aes(shape = factor(cyl))) +
  geom_line(aes(linetype = factor(gear))) +
  geom_smooth(aes(fill = factor(gear), color = gear)) + 
  theme_bw() +
  scale_fill_discrete(guide = "none")

3. show.legend = FALSE within geom_*() layers

You can suppress the legend for a specific layer by adding show.legend = FALSE to the geom_*() function. This is useful when you want to remove the legend for a particular geom without affecting others.

ggplot(data, aes(x = mpg, y = disp, group = gear)) +
  geom_point(aes(color = vs), show.legend = FALSE) +
  geom_point(aes(shape = factor(cyl)), show.legend = FALSE) +
  geom_line(aes(linetype = factor(gear)), show.legend = FALSE) +
  geom_smooth(aes(fill = factor(gear), color = gear), show.legend = FALSE) + 
  theme_bw()

4. Selective Legend Control with guide_legend()

For more refined control, especially when you want to selectively hide parts of a legend, you can use guide_legend() within guides(). For example, to remove the color from the fill legend while keeping the fill legend itself visible:

ggplot(data, aes(x = mpg, y = disp, group = gear)) +
  geom_point(aes(color = vs)) +
  geom_point(aes(shape = factor(cyl))) +
  geom_line(aes(linetype = factor(gear))) +
  geom_smooth(aes(fill = factor(gear), color = gear)) + 
  theme_bw() +
  guides(fill = guide_legend(override.aes = list(color = NA)))

Best Practices

Clarity: Strive for clarity in your visualizations. If a legend is unnecessary or obscures the data, remove it.
Consistency: Maintain consistency in your legend styles and positioning across different plots.
Selective Control: Use show.legend = FALSE and guide_legend() for fine-grained control when you need to customize legends for specific layers or aesthetics.