Graph info

Should I use this graph?


This graph requires:

✅ a categorical variable

Description

“In general, pie charts work well when the goal is to emphasize simple fractions, such as one-half, one-third, or one-quarter.”

“They also work well when we have very small datasets.” - Claus O. Wilke, Fundamentals of Data Visualization (2019)

Pie-charts are ideal for comparing the proportions of categorical variable values, and we can build pie-charts using the ggpubr package.

Getting set up

PACKAGES:

Install packages.

Code 
install.packages("ggpubr")
library(ggpubr)
install.packages("ggplot2movies")
library(ggplot2movies) 
library(ggplot2)

DATA:

Remove the missing values and "NC-17" from mpaa and summarise the count and percent.

Code 
movie_pie <- ggplot2movies::movies |>
  filter(mpaa != "" & mpaa != "NC-17") |> 
  group_by(mpaa) |> 
  summarise(cnt = n()) |> 
  mutate(
    perc = round(cnt / sum(cnt), 3),
    mpaa = factor(mpaa, 
          levels = c("PG", "PG-13", "R")))
glimpse(movie_pie)
Rows: 3
Columns: 3
$ mpaa <fct> PG, PG-13, R
$ cnt  <int> 528, 1003, 3377
$ perc <dbl> 0.108, 0.204, 0.688

The grammar

CODE:

Create labels with labs()

Initialize the graph with ggplot() and provide data

Assign "perc" to x

Assign labs to label

Assign "in" to lab.pos

Assign "white" to lab.font and color

Assign "mpaa" to fill

Remove legend with theme(legend.position = "none")

Code 
labs <- paste0(movie_pie$mpaa, " (", 
               (100*movie_pie$perc), "%)")
labs_pie <- labs(
  title = "Percent MPAA ratings for IMDB movies",
  x = "Percent MPAA rating")

ggp2_pie <- ggpubr::ggpie(movie_pie, 
          x = "perc", label = labs, 
          lab.pos = "in", lab.font = "#ffffff",
          fill = "mpaa", color = "#ffffff") + 
    theme(legend.position = "none")  
    
ggp2_pie + 
  labs_pie

GRAPH: