Overlaying Pie Charts in ggplot2

Question

I am making a pie chart to go along with a series of plots all made in ggplot2. The data I'm using have two categories broken in to a total of three subcategories. Basically, the data look like this:

    Category Category_Value Super_Category
    <fctr>   <dbl>          <dbl>
 1  A        0.03733874     1
 2  B        0.66732754     0
 3  C        0.29533372     1

Here is the basic pie chart I have at the subcategory level:

And here is what I'd like to have (or something similar):

I had never made a pie chart in ggplot2 before, so here is my basic code to generate the top plot:

pie.chart <- ggplot(pie.data, aes(x = "", y = Category_Value, fill = Category, width = 1)) +
  geom_bar(width = 1, stat = "identity") +
  coord_polar("y", start = 0) +
  scale_fill_manual(values = c("#4DAF4A", "#377EB8", "#E41A1C")) +
  theme_bw() + 
  theme(
    axis.title.x = element_blank(),
    axis.title.y = element_blank(),
    panel.border = element_blank(),
    panel.grid   = element_blank(),
    axis.ticks   = element_blank()
  )

Is this something that's doable? I messed around with making another plot grouped at the major category level and overlaying them without success.

Answer 1

You could use annotate to get an approximation of your picture.

Firstly I've used your small subset of data

pie.data <- data.frame(
  Category = c("A", "B", "C"),
  Category_Value = c(0.03733874, 0.66732754, 0.29533372),
  Super_Category = c(1,0,1))

Then I've appplied your code

pie.chart <- ggplot(pie.data, aes(x = "", y = Category_Value, fill = Category, width = 1)) +
  geom_bar(width = 1, stat = "identity") +
  coord_polar("y", start = 0) +
  scale_fill_manual(values = c("#4DAF4A", "#377EB8", "#E41A1C")) +
  theme_bw() + 
  theme(
    axis.title.x = element_blank(),
    axis.title.y = element_blank(),
    panel.border = element_blank(),
    panel.grid   = element_blank(),
    axis.ticks   = element_blank()
  )

And Finnaly I drew line using

pie.chart + annotate("rect", xmin = 1.5, xmax = 1.9, ymin = 0,2, ymax = 0.30, alpha=0,colour = "black")

And the output:

Note that because you have more data that my sample you will have to play with the settings values of ymin = 0,2, ymax = 0.30 in annotate in order your line will cover values that you want.