1  The Magic of {ggplot2}

1.1 data.frames and How to Use Them

Let’s review the basics of data.frames.

A data.frame is a table-like format which has the following properties:

  • Columns can each have a different type (numeric, character, boolean, factor)
  • Columns are called “variables”
  • Rows correspond to a single observation (ideally)
  • Can be subset or filtered based on criteria

Here’s an example of a table:

Individual variables within a data.frame can be accessed with the $ operator (such as gap1992$pop). We won’t use this very often, as the tidyverse lets us access the variables without it, as you’ll see.

1.1.1 Exercise

  • Run colnames() on the gap1992 data to see what’s in each column.
  • Then see how many rows there are in the dataset using nrow().
Solution 
##run colnames here on gap1992
colnames(gap1992)               
##run nrow() on gap1992
nrow(gap1992)

1.2 Thinking about aesthetics

Now that we’ve learned a little about the data.frame, we can get to the fun part: making graphs.

The first thing we are going to is think about how we represent variables in a plot.

How do we visually represent a variable in our dataset? Take a look at this graph. What variable is mapped to y, and what is mapped to x, and what is mapped to color?




1.3 Mapping variables to produce geometric plots

A statistical graphic consists of:

  • A mapping of variables in data to
  • aes()thetic attributes of
  • geom_etric objects.

In code, this is translated as:

Let’s take the above example code apart. A ggplot2 call always starts with the ggplot() function. In this function, we need two things:

  1. data - in this case, gap1992.
  2. mapping - An aesthetic mapping, using the aes() function.

In order to map our variables to aesthetic properties, we will need to use aes(), which is our aes()thetic mapping function. In our example, we map x to log(gdpPercap) and y to log(pop).

Finally, we can superimpose our geometry on the plot using geom_point().

1.3.1 Exercise

Based on the graph below, map the appropriate variables to the x, and y aesthetics. Run your plot. Remember, you can try plots out in the console before you submit your answer.

Hint: Look at the graph. If you need the variable names, they are listed below.

Solution 
ggplot(data=gap1992, 
    mapping = aes(
      x = log(gdpPercap), 
      y = lifeExp 
      )) + 
geom_point()

1.4 More about aes()

For geom_point(), there are lots of other aesthetics. The important thing to know is that aesthetics are properties of the geom. If you need to know the aesthetics that you can map to a geom, you can always use help() (such as help(geom_point)).

For a list of aesthetics for geom_point(), look here: https://ggplot2.tidyverse.org/reference/geom_point.html#aesthetics and look at all the aesthetic mappings. For example:

1.5 Aesthetics

geom_point() understands the following aesthetics (required aesthetics are in bold):

  • x
  • y
  • alpha
  • colour
  • fill
  • group
  • shape
  • size
  • stroke

1.6 Quick Check

Which of the following is not a mappable aesthetic to geom_point()?




1.7 Points versus lines

The great thing about ggplot2 is that it’s easy to swap geometric representations. Instead of x-y points, we can plot the data as a line graph by swapping geom_line() for geom_point().

1.7.1 Exercise

First run the code to see the plot with points. Change the geom_point() in the following graph to geom_line(). What happened?

How did the visual presentation of the data change?

Tip 
ggplot(gap1992, aes(x = log(gdpPercap), y = lifeExp, color=continent)) +
  geom_line()

1.8 Geoms are layers on a ggplot

We are not restricted to a single geom on a graph! You can think of geoms as layers on a graph. Thus, we can use the + symbol to add geoms to our base ggplot() statement.

1.8.1 Exercise

Add both geom_line() and geom_point() to the following ggplot. Are the results what you expected?

Solution 
ggplot(gap1992, aes(x = log(gdpPercap), y = lifeExp, color=continent)) +
## add code here
  geom_line() + geom_point()

1.9 Quick Check about ggplot2

What does the + in a ggplot statement do?

For example:




1.10 Final Challenge: Recreate this Gapminder Plot

Your final challenge is to completely recreate this graph using the gap1992 data.

  • If you need to remember variable names, you can always call head(gap1992) or colnames(gap1992) in the console.
  • Recreate the above graphic by mapping the right variables to the right aesthetic elements. Remember, you can try plots out in the console before you submit your answer.

1.10.1 Exercise

Tip 
ggplot(gap1992, aes(x = log(gdpPercap), 
    y = lifeExp, 
    color = continent,
    size = pop
    )) + ggtitle("Gapminder for 1992") + 
    geom_point()

1.11 What you learned in this chapter

  • Basic ggplot2 syntax.
  • Plotting x-y data using ggplot2 using both geom_point() and geom_bar().
  • Mapping variables in a dataset to visual properties using aes()
  • geoms correspond to layers in a graph.
  • That ggplot2 can make some pretty cool graphs
  • That you can do this!

More Resources