ggplot2

ggplot2 is a visualization package included in the tidyverse. It follows the Grammar of Graphics (GoG), which involves building graphs from the following components:

Image from The Grammar of Graphics by Leland Wilkinson

Check out the ggplot cheatsheet.

Elements in ggplot2 funtions

The vocabulary of ggplot can be difficult to parse at first. Here are the essential components:

• Aesthetics: Visual properties of the objects in your plot, e.g. axis of data, size, shape, color, pattern, fill of variables, alpha

• Geoms: Geometric objects representing data, such as lines, bars, points

• Facets: Subgrouping of data

• Statistics: Additional functions like regression lines

• Scales: Legends and labels

• Coordinate System: Cartesian, polar, etc.

• Themes: Background visuals

Install/load tidyverse

The very first time you want to use a package you first need to install it.

R

# if you have never downloaded tidyverse uncomment the line below and run to install it
install.packages('tidyverse')
# load tidyverse
library(tidyverse)

We will do a similar step with our penguin data that we will be visualizing.

R

install.packages("palmerpenguins")
library(palmerpenguins)

We use the View() function to look at the data frame and check that we have tidy data: each variable is a column and each observation is a row.

R

View(penguins)

Building layers

Let’s start by laying the foundations of our plot with the ggplot() function. We add in our data, letting us create a blank plot.

R

ggplot(data=penguins)

Output

Now we need to add aesthetics and geometric objects. Aesthetics are what you plot (x, y, size, color, fill, shape), and geoms are how you plot aesthetics (point, line, bar, boxplot). We specify aesthetics in aes().

Here, we set up axes to show the relationship between the variables bill_length_mm and bill_depth_mm. The data will still not be visualized, but the axes show the range of the data.

R

ggplot(data=penguins,aes(x=bill_length_mm,y=bill_depth_mm))

Output

Now we can decide what kind of plot to make. Let’s start with a simple scatter plot. We need to add the geom (geometry), which here is geom_point().

R

ggplot(data=penguins,aes(x=bill_length_mm,y=bill_depth_mm))+
       geom_point()

Output

We can now see our data! However, it is difficult to see any pattern at the moment.

Let’s group together data from each species. We can do this by adding color=species to the aes(), which gives each species its own color. This will also create a legend.

R

ggplot(data=penguins,aes(x=bill_length_mm,y=bill_depth_mm, color=species))+
         geom_point()

Output

In addition to color, you also add other aesthetics: fill, shape, linewidth, and alpha (transparency).

R

ggplot(data = penguins, aes(x = bill_length_mm, y = bill_depth_mm, shape = species)) +
   geom_point()

Output

If we specify a color outside of aesthetics, such as within geom_point(), every data point will be that color. We pick the specific color in quotes.

R

ggplot(penguins, aes(x = bill_length_mm, y = bill_depth_mm)) +
   geom_point(color = "red")

Output

Let’s try making another type of plot. Here, we make a boxplot of bill depth by species with geom_boxplot().

R

ggplot(data = penguins, aes(x = species, y = bill_depth_mm)) +
   geom_boxplot()

Output

We can make a histogram of bill depth with geom_histogram.

R

ggplot(data = penguins, aes(x = bill_depth_mm)) +
   geom_histogram()

Output

Like with our scatter plot, we can separate out species with color, here specified with fill.

R

ggplot(data = penguins, aes(x=bill_depth_mm, fill=species)) +
   geom_histogram(binwidth = 0.25)

Output

Facets

Another way to separate out groups is with facets. Facets are essentially panels showing each group individually. We specify the facets as their own layer in facet_wrap().

R

ggplot(data = penguins, aes(x = bill_depth_mm)) +
   geom_histogram(binwidth = 0.25) +
   facet_wrap(~ species)

Output

Customizing our plot

ggplot has many options for customizing plots. We will go into the very basics of those options here.

We will start by saving a simple colored box plot to a variable named myplot.

R

myplot<- ggplot(data = penguins, aes(x = species, y = bill_depth_mm, color = species)) +
   geom_boxplot()
myplot

Output

Once the plot is saved as a variable, we can add axes labels with xlab() and ylab().

R

myplot+
   xlab("Species")+
   ylab("Bill Depth")

Output

We can also change the title of the legend. Depending on various factors, such as how you are distinguishing groups, there are different functions for this. For this specific case, we use the function scale_color_discrete().

R

myplot+
   xlab("Species")+
   ylab("Bill Depth")+
   scale_color_discrete(name="Species of Penguin")

Output

Themes

The default theme in ggplot has a light gray background with a faint grid. There are many other themes you can use in ggplot, such as theme_minimal.

R

myplot+
   xlab("Species")+
   ylab("Bill Depth")+
   scale_color_discrete(name="Species of Penguin")+
      theme_minimal()

Output

This is one of many pre-built themes available. It is also possible to make a custom theme.

If you would like to play with other pre-built themes, try the ggthemes package!

R

install.packages('ggthemes')
library(ggthemes)
+theme_tufte()
+theme_fivethirtyeight()
+theme_economist()
+theme_wsj()
+theme_solarized()

Saving plots

Finally, to save a plot, you can use the ggsave() function, specifying the desired file name.

R

penguins_plot<-myplot+
   xlab("Species")+
   ylab("Bill Depth")+
   scale_color_discrete(name="Species of Penguin")+
      theme_minimal()

ggsave("penguins_plot.pdf", penguins_plot, device="pdf")

Output

## Saving 7 x 5 in image

Challenge

Try to recreate the following plot from the penguins data set:

Solution

R

ggplot(data = penguins, aes(x = flipper_length_mm,y = body_mass_g)) +
   geom_point(aes(color = species,
                  shape = species),
                  size = 3,
                  alpha = 0.8) +
scale_color_manual(values = c("darkorange","purple","cyan4")) +
labs(title = "Penguin size, Palmer Station LTER",
      subtitle = "Flipper length and body mass for Adelie, Chinstrap and Gentoo Penguins",
      x = "Flipper length (mm)",
      y = "Body mass (g)",
      color = "Penguin species",
      shape = "Penguin species") +
theme(legend.position = c(0.2, 0.7),
      plot.title.position = "plot",
      plot.caption = element_text(hjust = 0, face= "italic"),
      plot.caption.position = "plot")