Foundations of
Data Science

Spring 2026 | Data 2 (399)
Jeffrey M. Girard | Lecture 13c

Roadmap

• Apply complete themes to rapidly change the look of plots

• Customize individual theme elements for fine-grained control

• Combine multiple plots into complex layouts using patchwork

Themes

Themes

• Themes let you control non-data elements
  • Font size and color
  • Ticks and gridlines
  • Legend position

• Complete themes change many elements
  • theme_*()

• Individual elements can also be changed
  • theme(...)
  • element_*()

Setup for Themes

library(tidyverse)
# Create a base plot to add themes to
p <- 
  ggplot(mpg, aes(x = displ, y = hwy)) +
  geom_jitter(
    aes(color = drv, shape = drv),
    size = 3, alpha = 0.8
  ) + 
  labs(
    x = "Engine displacement (L)",
    y = "Highway MPG",
    color = "Drive",
    shape = "Drive",
    title = "Fuel Efficiency"
  )

p

Classic Theme

# theme_classic() removes gridlines and gives a clean axis
p + theme_classic(base_size = 24)

Black and White Theme

# theme_bw() adds a border and keeps major gridlines
p + theme_bw(base_size = 24)

Minimal Theme

# theme_minimal() removes the background box for a modern look
p + theme_minimal(base_size = 24)

Legend Position

# Move the legend to the bottom, top, left, or "none"
p + theme(legend.position = "bottom")

Legend Coordinates

# Move the legend inside the plot using relative coordinates (0 to 1)
p + theme(legend.position = c(0.85, 0.8))

Text Elements

# Use element_text() to change font properties
p + theme(plot.title = element_text(face = "bold", color = "red"))

Line Elements

# Use element_line() to modify gridlines and axes
p + theme(panel.grid.major = element_line(linetype = "dashed", color = "red"))

Blanking

# Use element_blank() to remove specific features entirely
p + theme(panel.grid.minor = element_blank(), axis.ticks.x = element_blank())

Reusable Themes

# Store your custom theme in an object
# Then apply it to multiple plots
my_theme <- 
  theme_minimal(base_size = 30) +
  theme(
    legend.position = "bottom",
    plot.title = 
      element_text(face = "bold"),
    panel.grid.minor = element_blank()
  )

p + my_theme

Layouts

Layouts

• We may want to combine multiple plots into one
  • e.g., for communication and publishing
  • e.g., to combine related “subfigures”
  • e.g., to create complex figure layouts

• To do so, we can use the {patchwork} package
  • This redefines the +, |, / operators
  • We use these to combine ggplot objects
  • We can control layout and annotations

Setup for Patchwork

# Create example plots to work with
p1 <- ggplot(mpg, aes(x = displ)) + 
  geom_histogram(bins = 10, fill = "steelblue", color = "white") + 
  labs(x = "Engine Displacement", y = NULL)

p2 <- ggplot(mpg, aes(x = hwy)) + 
  geom_histogram(bins = 10, fill = "seagreen", color = "white") + 
  labs(x = "Highway MPG", y = NULL)

p3 <- ggplot(mpg, aes(x = displ, y = hwy)) + 
  geom_point(alpha = 0.5) +
  labs(x = "Engine Displacement", y = "Highway MPG")

Combining Plots: Basic Addition

# The + operator attempts to lay out the plots logically
library(patchwork)

p1 + p2

Controlling the Layout

# Use plot_layout() to explicitly state columns or rows
p1 + p2 + p3 + plot_layout(ncol = 3)

Adding Spacers

# Use plot_spacer() to leave empty areas in the grid
p1 + plot_spacer() + p2 + plot_layout(ncol = 3)

Side-by-Side

# The | operator explicitly places plots side-by-side
p1 | p2 | p3

Above-and-Below

# The / operator explicitly places plots above and below
p1 / p3

Mixing and Matching

# Parentheses determine nesting and grouping
(p1 | p2) / p3

Relative Sizes

# Pass a vector to widths or heights to adjust proportions
(p1 | p2) / p3 + plot_layout(heights = c(1, 2))

Overall Annotations

# plot_annotation() adds a master title, subtitle, or caption
(p1 | p2) / p3 + 
  plot_annotation(
    title = "Vehicle Engine Size and Fuel Economy",
    caption = "Data from fueleconomy.gov"
  )

Automated Tagging

# tag_levels automatically labels subfigures (e.g., A, B, C or 1, 2, 3)
(p1 | p2) / p3 + 
  plot_annotation(tag_levels = "A")