ggplot2 Essentials

Estimated time: 30 minutes

Overview

Questions

• How do I create plots with ggplot2?
• How do I save the plots I created?

Objectives

• Use ggplot2 to generate plots.
• Know the three components to a basic ggplot2: data, aesthetic mappings, geometric objects.
• Manipulate the aesthetic mappings of a plot.
• Save a plot created with ggplot() to disk.

Plotting is a key component of exploratory data analysis and a powerful way to identify patterns, trends, and relationships between variables in your dataset.

In this lesson, we will use the ggplot2 package, a widely used system for creating clear and flexible data visualisations in R.

ggplot2 is built on the grammar of graphics, which is the idea that any plot can be built from the same set of components: a data set, mapping aesthetics, and graphical layers:

• Data are the data that you, the user, provide.

• Aesthetic Mappings are what connect the data to the graphics. They tell ggplot() how to use your data to affect how the graph looks, such as changing what is plotted on the X or Y axis, or the size or colour of different data points.

• Geometric objects (geoms) are the parts of the plot that we can see, such as points, lines, or bars. Each geom creates a different type of plot (e.g. scatter plots, histograms, bar charts). In ggplot2, plots are built up in layers, with each geom added as new layer.

Data

---

In this lesson we will use the mpg dataset, which is included with the ggplot2 package. This dataset contains fuel economy data for a range of cars. Fuel efficiency is recorded in miles per gallon (mpg), which is commonly used in the United States (in Australia, we more often use litres per 100 km). Higher values indicate better fuel efficiency.

Because mpg is built into ggplot2, we can use it directly without needing to load an external file.

Setup

We are going to be using functions from the ggplot2 package to create visualizations of data. Functions are predefined bits of code that automate more complicated actions. R itself has many built-in functions, but we can access many more by loading other packages of functions and data into R.

If you don’t have a blank, untitled script open yet, go ahead and open one with Shift+Cmd+N (Mac) or Shift+Ctrl+N (Windows). Then save the file to your scripts/ folder, and title it workshop_code.R.

Earlier, you had to install the ggplot2 package by running install.packages("tidyverse"). That installed the package onto your computer so that R can access it. In order to use it in our current session, we have to load the package using the library() function.

R

library(ggplot2)

We will use the mpg dataset that is prepackaged with ggplot.

In R, the str() function provides a summary of the dataset’s internal structure. It tells you the total number of observations (rows) and variables (columns), the data type of each column (e.g., numeric, character, or factor), and displays the first few values of each variable.

R

str(mpg)

OUTPUT

tibble [234 × 11] (S3: tbl_df/tbl/data.frame)
 $ manufacturer: chr [1:234] "audi" "audi" "audi" "audi" ...
 $ model       : chr [1:234] "a4" "a4" "a4" "a4" ...
 $ displ       : num [1:234] 1.8 1.8 2 2 2.8 2.8 3.1 1.8 1.8 2 ...
 $ year        : int [1:234] 1999 1999 2008 2008 1999 1999 2008 1999 1999 2008 ...
 $ cyl         : int [1:234] 4 4 4 4 6 6 6 4 4 4 ...
 $ trans       : chr [1:234] "auto(l5)" "manual(m5)" "manual(m6)" "auto(av)" ...
 $ drv         : chr [1:234] "f" "f" "f" "f" ...
 $ cty         : int [1:234] 18 21 20 21 16 18 18 18 16 20 ...
 $ hwy         : int [1:234] 29 29 31 30 26 26 27 26 25 28 ...
 $ fl          : chr [1:234] "p" "p" "p" "p" ...
 $ class       : chr [1:234] "compact" "compact" "compact" "compact" ...

Data dictionary

The main variables we will use in this lesson are:

• displ: engine size (litres)
• hwy: highway fuel efficiency (miles per gallon)
• cty: city fuel efficiency (miles per gallon)
• class: type of car (e.g. SUV, compact)
• drv: drive type (front-wheel, rear-wheel, 4-wheel)

Challenge

Know your data

Before creating any plots, it’s useful to briefly explore the dataset to understand what variables are available and what they represent.

Take a minute to explore the dataset:

• What types of variables are included (numeric, categorical)?
• Which variable might you use for the x-axis if you wanted to explore engine size?
• Which variable could you use for the y-axis to represent fuel efficiency?
• Is there a categorical variable you could use to group points?

Show me the solution

• numeric: displ, year, cyl, cty, hwy
• categorical: manufacturer, model, trans, drv, fl, class
• engine size: displacement
• fuel efficiency: mpg
• any of the categorical variables can be used to group points

Plotting with ggplot2

---

Let’s start off building an example using the mgp data.

The most basic function is ggplot(), which lets R know that we’re creating a new plot. Any of the arguments we give the ggplot() function are the global options for the plot - they apply to all layers on the plot.

R

ggplot(data = mpg)

Here we called ggplot() and told it what data we want to show on our plot. This is not enough information to actually draw anything. However, it does create a blank plot that helps demonstrate how the components are put together. We’re essentially providing the base layer for other elements to be added on to.

Next we’re going to add in the mapping aesthetics using the aes() function. aes() tells ggplot() how variables in the data map to aesthetic properties of the plot, such as which columns of the data should be used for the x and y locations.

R

ggplot(data = mpg, 
       mapping = aes(x = displ, y = hwy))

Here we told ggplot() we want to plot the “displ” column of the data frame on the x-axis, and the “hwy” column on the y-axis. Notice that we didn’t need to explicitly pass aes these columns (e.g. x = mpg[, "disp"]). This is because ggplot() is designed to look in the data for that column!

Notice again that we still don’t have a plot. The third and final component needed to make a plot is a geom function to tell ggplot() how to visually represent the data. For example, if we want to use points to represent the data in our plot, we use the geom_point() function.

The + is used to add layers to a plot.

R

ggplot(data = mpg, 
       mapping = aes(x = displ, y = hwy)) +
  geom_point()

There we have it! A scatter plot of points that represents the relationship between x (‘displacement’) and y (‘highway fuel efficiency’) in our dataset.

These three components — data, aesthetic mappings, and geoms — are all you need to create a basic plot in ggplot2.

In the rest of this lesson, we will build on this foundation to create different types of visualisations and explore how to adapt them for different questions.

Challenge

Aesthetic Mappings

Modify the example so that the plot shows how city fuel efficiency relates to engine displacement.

• What can we say about the relationship?
• How does it compare to the relationship with highway efficiency?
• (BONUS) How could we modify this plot to make the difference with highway more obvious?

Show me the solution

• City fuel efficiency (cty) and engine displacement (displ) have a strong negative correlation.
  • As engine displacement (size in liters) increases, city fuel economy (miles per gallon) noticeably decreases, meaning larger engines consume significantly more fuel.
• Highway miles per gallon (hwy) are generally higher than city miles.
  • Because of this, the hwy is higher on the y-axis and stretches across a slightly wider range.
• We could put both sets of data on the same plot.

R

ggplot(data = mpg, 
       mapping = aes(x = displ, y = cty)) + 
  geom_point()

Saving the plot

---

The ggsave() function allows you to save a plot created with ggplot2 quickly and easily with just a filename:

R

ggsave(filename = "fig/my_ggplot.png")

OUTPUT

Saving 7 x 7 in image

Notice that we did not provide any additional arguments to ggsave().

• If we omit the plot argument, it will automatically save the last plot you created with ggplot.
• If we omit the device argument, it will use the file extension to determine the device.

In this episode, we introduced the grammar of graphics and the core components of ggplot2: data, aesthetic mappings, and geometric layers. Using these components, we created our first plot with geom_point() and explored how variables can be mapped to visual properties such as colour.

In the next episodes, we will build on these basics to explore our dataset in more detail. We will use a selection of commonly used geoms to examine relationships between variables, the distribution of individual variables, differences between groups, and the composition of our data.

Key Points

• ggplot2 can quickly create simple plots for exploratory data analysis.
• ggplot2 plots are built around the grammer of graphics where the key components of a plot can be extended with the additional geom layers.