Plot multiple Pie chart in different size and position in r-CodePudding

I would like to make a multiple pie plot.

df <- read.table(text="Position Depth Group Value
A   50  G1  30
A   50  G2  60
A   50  G3  10
A   200 G1  10
A   200 G2  10
A   200 G3  30
B   50  G1  24
B   50  G2  48
B   50  G3  8
B   100 G1  8
B   100 G2  8
B   100 G3  24
C   100 G1  36
C   100 G2  72
C   100 G3  12
C   600 G1  12
C   600 G2  12
C   600 G3  36",
header=TRUE)

Position: the different sampling site;
Depth: the sampling water depth;
Group: three different animals;
Value: the count of animals.

I'd like to make the figure like:
the size of pie:total count of Group(G1 G2 G3);
pie composition:Group(G1%, G2%, G3%)

CodePudding user response：

One option to achieve your desired result would be via the scatterpie::geom_scatterpie which however requires some data wrangling. First you have to convert your categorical Position column to a numeric. Second we have to transform the numeric Position to the same scale as Depth. Third, we have to rename the Value column to value (geom_scatterpie demands that in case of working with data in long format). Finally I have chosen the radius such that the pie area reflects the sum of values. Depending on your output format you probably have to do some rescaling of the radius, e.g. multiplied by 3.

library(dplyr)
library(scatterpie)
library(ggplot2)

df <- df |>
  mutate(x = as.numeric(factor(Position)), x = scales::rescale(x, to = range(Depth))) |>
  rename(value = Value) |>
  group_by(Position, Depth) |>
  mutate(r = 2 * sqrt(sum(value) / 2 / pi))

ggplot()  
  geom_scatterpie(aes(x = x, y = Depth, fill = Group, r = 3 * r), data = df, cols = "Group", long_format = TRUE)  
  scale_x_continuous(breaks = unique(df$x), labels = unique(df$Position))  
  scale_y_reverse(breaks = unique(df$Depth))  
  coord_fixed()

DATA

df <- structure(list(Position = c(
  "A", "A", "A", "A", "A", "A", "B",
  "B", "B", "B", "B", "B", "C", "C", "C", "C", "C", "C"
), Depth = c(
  50L,
  50L, 50L, 200L, 200L, 200L, 50L, 50L, 50L, 100L, 100L, 100L,
  100L, 100L, 100L, 600L, 600L, 600L
), Group = c(
  "G1", "G2", "G3",
  "G1", "G2", "G3", "G1", "G2", "G3", "G1", "G2", "G3", "G1", "G2",
  "G3", "G1", "G2", "G3"
), Value = c(
  30L, 60L, 10L, 10L, 10L, 30L,
  24L, 48L, 8L, 8L, 8L, 24L, 36L, 72L, 12L, 12L, 12L, 36L
)), class = "data.frame", row.names = c(
  NA,
  -18L
))