bean plot

R code of the ggplot2 example

library(ggplot2)
library(dplyr)
library(palmerpenguins)

birds <- penguins |>
  filter(!is.na(bill_depth_mm)) |>
  mutate(position = as.integer(species))

means <- birds |>
  group_by(species, position) |>
  summarise(depth = mean(bill_depth_mm), .groups = "drop")

## The bean: a mirrored density; inside, a short line per penguin and a long
## one for the mean.
p <- ggplot(birds, aes(x = species, y = bill_depth_mm, fill = species)) +
  geom_violin(colour = NA, alpha = 0.25) +
  geom_segment(aes(x = position - 0.08, xend = position + 0.08, yend = bill_depth_mm, colour = species),
               linewidth = 0.3, alpha = 0.6) +
  geom_segment(data = means, aes(x = position - 0.3, xend = position + 0.3, y = depth, yend = depth),
               colour = "black", linewidth = 0.7, inherit.aes = FALSE) +
  scale_y_continuous(labels = lex_number, breaks = lex_break_breaks, expand = lex_break_expand) +
  labs(title = "Gentoo penguins have the shallowest bills",
       subtitle = "Short lines: one penguin each; long black line: the mean",
       x = NULL, y = "Bill depth (mm)",
       caption = "Data: palmerpenguins (Horst, Hill & Gorman 2020)") +
  theme_lex_viz(grid = "y", arrows = "y", break_not_zero = "y") +
  theme(legend.position = "none")