Visualizing WNBA ages

library(tidyverse)
library(patchwork)

# This is from exercise 7
wnba <- read_csv("https://github.com/andrewheiss/dataviz-projects_07-relationships/raw/refs/heads/main/data/wnba.csv")

Distributions

We can plot the distribution of ages with dots—each dot is a player:

plot_dots <- ggplot(wnba, aes(x = age)) +
  geom_dotplot(
    binwidth = 1,
    dotsize = 0.4,
    fill = "darkorange"
  )
plot_dots

↑ that means that there’s 1 19-year-old player, 2 20-year-old players, 15 25-year-old players, and so on. The dots show the general shape of the ages of WNBA players.

Instead of using dots, we can use a histogram. Here, the bin width is 1, so that the bars show the count of players between 19-20, 20-21, 21-22, and so on. It matches the dotplot:

plot_hist <- ggplot(wnba, aes(x = age)) +
  geom_histogram(
    binwidth = 1,
    boundary = 0,
    color = "white",
    fill = "darkred"
  )
plot_hist

A density plot is basically a histogram, but smoothed out

plot_density <- ggplot(wnba, aes(x = age)) +
  geom_density(fill = "dodgerblue", color = NA)
plot_density

It shows that most players are around 25 years old and still matches the general shape of the points and the histogram.

Here are all three combined:

plot_dots / plot_hist / plot_density

or in one plot:

ggplot(wnba, aes(x = age)) +
  geom_histogram(
    aes(y = after_stat(count)),
    binwidth = 1,
    color = "white",
    fill = "darkred"
  ) +
  geom_dotplot(
    aes(y = after_stat(count)),
    binwidth = 1,
    dotsize = 0.5,
    fill = "darkorange"
  ) +
  geom_density(
    aes(y = after_stat(density) * nrow(wnba) * 1),
    color = "dodgerblue",
    linewidth = 2
  )

Violin plots

Violin plots are density plots that are mirrored.

Like, here’s a regular density plot:

ggplot(wnba, aes(x = age)) +
  geom_density(fill = "dodgerblue", color = NA)

And here’s a violin plot, which is the same thing, but with a second density flipped down at the bottom (I added a horizontal line to show where it’s reflected):

ggplot(wnba, aes(x = age, y = 0)) + 
  geom_violin(fill = "dodgerblue", color = NA) + 
  geom_hline(yintercept = 0)

Half distributions

The different half distributions from {gghalves} let you show different types of graphs simultaneously.

Like here’s a half violin (or just a density) with stacked dots:

library(gghalves)

ggplot(wnba, aes(x = 0, y = age)) +
  geom_half_violin(side = "r", fill = "dodgerblue", color = NA) +
  geom_half_dotplot(
    stackdir = "down",
    binwidth = 1,
    dotsize = 0.3,
    fill = "darkorange"
  ) +
  coord_flip()

Or a histogram with stacked dots:

ggplot(wnba, aes(x = 0, y = age)) +
  geom_histogram(
    aes(y = age),
    inherit.aes = FALSE,
    binwidth = 1,
    color = "white",
    fill = "darkred"
  ) +
  geom_half_dotplot(
    stackdir = "down",
    binwidth = 1,
    dotsize = 0.3,
    fill = "darkorange"
  ) +
  coord_flip(xlim = c(-35, 25))

Or a half violin with jittered points:

ggplot(wnba, aes(x = 0, y = age)) +
  geom_half_violin(side = "r", fill = "dodgerblue", color = NA) +
  geom_half_point(side = "l", position = position_jitter(height = 0, width = 0, seed = 1234)) +
  coord_flip()