Data visualization: Design choices

Warning: Missing `trust` will be set to FALSE by default for RDS in 2.0.0.

Recall: common components of ggplot2

• Common components of ggplot2 include:
  • ggplot(): the main function to create a plot
  • aes(): the function to specify the mapping of variables to axes
  • geom_*() or stat_*(): functions to add layers to the plot
  • labs(): the function to change labels of axes and title
  • theme(): the function to adjust non-data plot elements such as text size,

 

ggplot(
  data = my_data, 
  mapping = aes(x = col1, y = col2)
  ) +
  
  geom_*() +          
  
  labs() +       
  
  theme()                                  

In this lesson, we will focus on labels, themes, and facets.

Basic labelling using labs(): titles and axes

• We can assign many different labels to a plot using the labs() function

We have the following options for labelling:

ggplot() +     
  geom_*() +          
  labs(
    x = "Label for x-axis",
    y = "Label for y-axis",
    color = "Label for color legend",
    fill = "Label for fill legend",
    title = "Title for the whole plot",
    subtitle = "Appears smaller, under the title", 
    caption = "Appears at the bottom of the plot, good for context" 
  ) +       
  theme()                                  

1

Typically all you need for a plot with two variables.

2

Only needed if you have a legend for color or fill.

Example 1: New labels for our plot of age vs. self-reported health

• The following example adds all possible labels to the plot
• We do NOT need to add every possible label!!

ggplot(
  hrs_00, 
  aes(x = age_yr, y = srh)
  ) +
  geom_boxplot() +
  geom_jitter(alpha = 0.2) +
  labs(
    x = "Age (years)",
    y = "Self-reported health",
    title = "Self-reported health by age",
    subtitle = "Data from the Health and Retirement Study (HRS)",
    caption = "Source: HRS, 2022 wave"
  )

Example 2: Labels for color

• Note that the HRS dataset came with really nice labels embedded in the .rda file

ggplot(
  hrs_00,
  aes(x = age_yr, y = srh, color = srh)
  ) +
  geom_jitter()

ggplot(
  hrs_00,
  aes(x = age_yr, y = srh, color = srh)
  ) +
  geom_jitter() +
  labs(
    x = "Age (years)",
    color = "Self-reported \nhealth"
  )

1

Take away “at Interview” for x-axis

2

Added \n to condense the horizontal space that the legend title takes up.

Example 3: Same plot without our nice, pre-labelled variables

• If we use the .xlsx file, we lose the labels and the SRH level order.

hrs_xlsx <- import(here("data", "hrs_data.xlsx")) |> drop_na()

ggplot(
  hrs_xlsx,
  aes(x = age_yr, y = srh, color = srh)
  ) +
  geom_jitter()

ggplot(
  hrs_xlsx,
  aes(x = age_yr, y = srh, color = srh)
  ) +
  geom_jitter() +
  labs(
    x = "Age (years)",
    y = "Self-reported health", 
    color = "Self-reported \nhealth"
  )

2

Added \n to condense the horizontal space that the legend title takes up.

Facets

• Facets allow us to create multiple plots based on one or more discrete variables
  • Discrete meaning categorical or whole number values

 

• We create multiple plots based on subsets of the data
  • If each subset is determined by one variable, we use facet_wrap()
  • If each subset is determined by two variables, we use facet_grid()

Example 4: facet_wrap()

• We use srh as the facet variable to create a histogram of age for each level of self-reported health

ggplot(
  hrs_00,
  aes(x = age_yr)
  ) +
  geom_histogram() +
  facet_wrap(~ srh)

1

The ~ indicates that we are faceting by a single variable, srh.

`stat_bin()` using `bins = 30`. Pick better value `binwidth`.

Example 5: facet_grid()

• We use srh and diab as the facet variables to create a histogram of age for each level of self-reported health and diabetes status

ggplot(
  hrs_00,
  aes(x = age_yr)
  ) +
  geom_histogram() +
  facet_grid(diab ~ srh)

1

The ~ indicates that we are faceting by two variables, diab and srh. The first variable is used for the rows and the second variable is used for the columns.

`stat_bin()` using `bins = 30`. Pick better value `binwidth`.

Complete themes

• We can change the entire appearance of our plot with themes
• There are several set themes in ggplot2: check out the page on complete themes for more!

 

ggplot(hrs_00) +
  geom_histogram(aes(x = age_yr)) + 
  theme_grey()

1

This is the default theme!

`stat_bin()` using `bins = 30`. Pick better value `binwidth`.

ggplot(hrs_00) +
  geom_histogram(aes(x = age_yr)) +
  theme_bw()

`stat_bin()` using `bins = 30`. Pick better value `binwidth`.

ggplot(hrs_00) +
  geom_histogram(aes(x = age_yr)) +
  theme_classic()

`stat_bin()` using `bins = 30`. Pick better value `binwidth`.

ggplot(hrs_00) +
  geom_histogram(aes(x = age_yr)) +
  theme_minimal()

`stat_bin()` using `bins = 30`. Pick better value `binwidth`.

Modifying themes

• We can change various components of a theme using the theme() function

 

• We can change the look of the title, axes, facets, and legend
  • The input options for theme() is VAST

 

• We’re only going to touch the surface here!

Example 6: Modifying the title and axes

ggplot(
  hrs_00, 
  aes(x = age_yr, y = srh) 
  ) +
  geom_boxplot() + 
  geom_jitter(alpha = 0.2) + 
  labs(
    x = "Age (years)", 
    y = "Self-reported health",
    title = "Self-reported health by age",
  ) +
  theme(
    plot.title = element_text(size = 30, face = "bold", color = "blue"),
    axis.title.x = element_text(size = 20, face = "italic", color = "red"),
    axis.title.y = element_text(size = 20, color = "red"),
    axis.text.y = element_text(size = 16, angle = 45)
  )

1

Changes the title size, makes it bold, and changes the color to blue

2

Changes the x-axis title size, makes it italic, and changes the color to red

3

Changes the y-axis title size and color to red

4

Changes the y-axis text size and rotates it 45 degrees

Example 7: Modifying the facets

ggplot(
  hrs_00,
  aes(x = age_yr)
  ) +
  geom_histogram() +
  facet_grid(diab ~ srh) +
  theme(
    strip.text.x = element_text(size = 16, color = "blue"),
    strip.text.y = element_text(size = 16, color = "red")
  )

1

Changes the facet column label size and color to blue

2

Changes the facet row label size and color to red

`stat_bin()` using `bins = 30`. Pick better value `binwidth`.

Example 8: Modifying the legend

ggplot(
  hrs_00,
  aes(x = age_yr, y = srh, color = srh)
  ) +
  geom_jitter() +
  labs(
    x = "Age (years)", 
    color = "Self-reported health"
  ) +
  theme(
    legend.position = "bottom"
  )

1

Moving the legend to the bottom of the plot instead of the right (default)

Saving plots

• We can save a plot as an image for later use: way better quality than taking a screenshot of the plot
• We need to assign the plot to an R object, and then we can save the R object as an image or pdf
• We use ggsave() to save an R object (that’s a plot)

Let’s say I saved my plot from Example 8 as plot_age_srh:

plot_age_srh <- ggplot(
  hrs_00,
  aes(x = age_yr, y = srh, color = srh)
  ) +
  geom_jitter() +
  labs(
    x = "Age (years)", 
    color = "Self-reported health"
  ) +
  theme(
    legend.position = "bottom"
  )

1

I assign the output (aka my plot) to plot_age_srh

ggsave(
  filename = here(
    "images", 
    "plot_age_srh.png"
    ),
  plot = plot_age_srh,
  width = 5,
  height = 3,
  units = "in",
  dpi = 200)

1

We need to tell R where we want the image to end up in our files

2

We give R the name of the plot

3

We need to adjust the height and width to make the plot readable in the image

Wrap-up

• We went over additional layers for ggplot2

 

• We can use labels to make plots more readable to viewers
• We can use facets to wrap plots by subsets of data
• We can use themes to change the overall appearance of a plot
• We can use ggsave() to save plots as images

 

• Check out more on ggplot2: it has a lot of capabilities!

Resources

• ggplot2 Cheatsheet in html form
• ggplot2 Cheatsheet in pdf form
• List of complete themes
• Modify components of a theme
• Page on ggsave()
• Online textbook for ggplot2: https://ggplot2-book.org/
• Another online resource for data visualization with ggplot2: https://socviz.co/index.html#preface
• Advanced Data Viz Class at Duke

 

• If you found these lessons on Data Visualization fun, I highly suggest you look into Dr. Bedrick’s Data Visualization course in the Spring!