ggplot2 theming

library(tidyverse)

── Attaching packages ─────────────────────────────────────── tidyverse 1.3.2 ──
✔ tibble  3.1.8      ✔ dplyr   1.0.10
✔ tidyr   1.2.1      ✔ stringr 1.5.0 
✔ readr   2.1.3      ✔ forcats 0.5.2 
✔ purrr   0.3.5      
── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
✖ dplyr::filter() masks stats::filter()
✖ dplyr::lag()    masks stats::lag()

library(palmerpenguins)

theme_set(theme_gray())

Scales formatters

library(scales)

Attaching package: 'scales'

The following object is masked from 'package:purrr':

    discard

The following object is masked from 'package:readr':

    col_factor

# according to wine spectator 
# https://top100.winespectator.com/lists/
wine_ranking <- 
  tibble(
    rank = c(1, 2, 3, 4),
    price = c(80, 130, 78, 65)
  )

wine_baseplot <- 
  ggplot(wine_ranking, 
         aes(
           x = rank,
           y = price
         )
        )+
    geom_col()
wine_baseplot

wine_baseplot+
  scale_x_continuous(
    labels = label_ordinal()
  )+
  scale_y_continuous(
    labels = label_dollar()
  )

wine_baseplot+
  scale_x_continuous(
    labels = label_ordinal(
      rules = ordinal_french()
    )
  ) +
  scale_y_continuous(
    labels = label_dollar(
      prefix = "",
      suffix = "\u20ac"
    )
  )

Figure size, caption, alt-text

We can configure the size of figures and other aspects of their appearance in a quarto document by setting certain quarto chunk options.

```{r}
#| label: fig-wine-price
#| fig-width: 4
#| fig-height: 1
#| out-width: 75%
#| fig-align: center
#| fig-cap: "The relationship between wine ranking and price"
#| fig-cap-location: margin
#| fig-alt: "A bar graph representing the relationship between the top four wine's ranks and their price. The most expensive bottle is ranked second."
wine_baseplot
```

Figure 1: The relationship between wine ranking and price

Pre-set themes

penguin_baseplot <- 
  ggplot(
    data = drop_na(penguins), 
    aes(
      x = bill_length_mm, 
      y = bill_depth_mm, 
      color = species
    )
  )+
  geom_point()+
  facet_wrap(
    ~island,
    labeller = label_both
  )+
  labs(title = "Penguins")

penguin_baseplot

Default themes

theme_bw()

penguin_baseplot +
  theme_bw()

theme_minimal()

penguin_baseplot+
  theme_minimal()

theme_void()

penguin_baseplot + 
  theme_void()

theme_dark()

penguin_baseplot +
  theme_dark()

theme_light()

penguin_baseplot +
  theme_light()

theme_classic()

penguin_baseplot +
  theme_classic()

Theme extensions

ggthemes

library(ggthemes)

A limited set of examples

theme_tufte()

penguin_baseplot +
  theme_tufte()

theme_fivethirtyeight()

penguin_baseplot +
  theme_fivethirtyeight()+
  scale_color_fivethirtyeight()

theme_excel()

penguin_baseplot +
  theme_excel()+
  scale_color_excel()

theme_economist()

penguin_baseplot +
  theme_economist()+
  scale_color_economist()

theme_wsj()

penguin_baseplot +
  theme_wsj()+
  scale_color_wsj()

theme_stata()

penguin_baseplot +
  theme_stata() +
  scale_color_stata()

Dark mode

{ggdark} has a number of dark mode themes based on the default ggplot2 themes, as well as ggdark::dark_mode() to turn any given theme “darkmode”

library(ggdark)

dark_theme_grey()

penguin_baseplot +
  dark_theme_grey()

dark_mode(theme_stata())

penguin_baseplot +
  dark_mode(theme_stata())

Taking control of themes

Each plot theme component is made up of one of the following:

• rectangles, ggplot2::element_rect()

• lines, ggplot2::element_line()

• text, ggplot2::element_text()

For example, to change the polarizing grey background to a different color, we need to

• identify what component of the plot we want to change (the panel)

• identify what aspect of of that component we want to change (panel.background)

• identify what element we’re modifying (element_rect)

penguin_baseplot +
  theme(
    panel.background = element_rect(fill = "steelblue")
  )+
  labs(caption = "this is not a better color solution")

Often it makes sense to start with an existing theme, and then start tweaking it.

penguin_baseplot +
  theme_minimal(
    # change font size
    base_size = 14
  ) +
  theme(
    # drop the "minor" breaks
    panel.grid.minor = element_blank(),
    # slender grey dashed major breaks
    panel.grid.major = element_line(
      color = "grey80", 
      linewidth = 0.25,
      linetype = "dashed"
    ),
    # dashed border around each facet
    panel.border = element_rect(
      fill = NA,
      color = "grey50",
      linetype = "dashed"
    ),
    # move the legend to the top
    legend.position = "top",
    # force each plotting area to be square
    aspect.ratio = 1
  ) -> sticker_plot

sticker_plot

Fonts

To change the font of a plot, you need to use the family argument in element_text().

Getting fonts to show up right in an R plot is an ongoing struggle, but the recent showtext package improves things a bit. Let’s experiment by adding two fonts from google fonts.

library(showtext)
font_add_google(name = "Lobster", family = "Lobster")
font_add_google(name = "Ubuntu", family = "Ubuntu")

sticker_plot +
  theme(
   text = element_text(family = "Ubuntu"),
   plot.title = element_text(family = "Lobster", size = 30),
   strip.text = element_text(face = "bold", color = "grey20")
  )