在 R 上绘制 3D 饼图时出现错误消息

Question

I currently have trouble plotting a 3D pie chart whereas it had worked well with a very similar dataset.

Here is my dataset:

structure(list(type_de_sejour = structure(1:4, levels = c("Ambulatoires", 
"Externes", "Fictifs", "Hospitalisé"), class = "factor"), nb_sejours_2021 = c(20794, 
365, 0, 7866)), row.names = c(NA, -4L), class = "data.frame")

And here is my code:

lab <- paste0(round(pie1_PGS$nb_sejours_2021/sum(pie1_PGS$nb_sejours_2021) * 100, 2), 
"%")


pie3D(pie1_PGS$nb_sejours_2021, radius = 0.75,
      height = 0.1,
      theta = 0.7,
      border = "white",
      col = rainbow(length(lab)),
      shade = 0.5,
      labels = lab,
      labelcol = "red",
      labelcex = 0.75,
      explode = 0.2,
      main = "Répartition des séjours 2021 par type")

I get the following message of error:

"Error in seq.default(start, end, by = angleinc) : 
  (to - from) / by incorrect"

Also, I would like to plot a legend to indicate what the colours mean (they take the values of variable type_de_sejour). I have seen several posts here, but I can't seem to manage to do it on this dataset, so I would welcome any help regarding this issue too. Here is the code I added:

legend(0.5, 1.5, c("Ambulatoires","Externes", "Hospitalisé", 
"Séances"), cex = 0.3,
       fill = rainbow(length(lab)))

I think the problem is that the legend is too big as regards the plot...

Answer 1

Add legend

You can use the function legend like this:

pie1_PGS <- structure(list(type_de_sejour = structure(1:4, levels = c("Ambulatoires", 
                                                                      "Externes", "Fictifs", "Hospitalisé"), class = "factor"), nb_sejours_2021 = c(20794, 
                                                                                                                                                    365, 0, 7866)), row.names = c(NA, -4L), class = "data.frame")

pie1_PGS <- pie1_PGS[!(pie1_PGS$nb_sejours_2021 == 0),]

lab <- paste0(round(pie1_PGS$nb_sejours_2021/sum(pie1_PGS$nb_sejours_2021) * 100, 2), 
              "%")

library(plotrix)
pie3D(pie1_PGS$nb_sejours_2021, 
      radius = 0.75,
      height = 0.1,
      theta = 0.7,
      border = "white",
      col = rainbow(length(lab)),
      shade = 0.5,
      labels = lab,
      labelcol = "red",
      labelcex = 0.75,
      explode = 0.2,
      main = "Répartition des séjours 2021 par type") 

legend(0.1, 0.9, pie1_PGS$type_de_sejour, cex = 0.7, fill = rainbow(length(lab)))

^{Created on 2022-08-11 by the reprex package (v2.0.1)}

You should remove the rows with 0 value because you can't show them in a pie chart. You can use the following code:

pie1_PGS <- structure(list(type_de_sejour = structure(1:4, levels = c("Ambulatoires", 
                                                                      "Externes", "Fictifs", "Hospitalisé"), class = "factor"), nb_sejours_2021 = c(20794, 
                                                                                                                                                    365, 0, 7866)), row.names = c(NA, -4L), class = "data.frame")

pie1_PGS <- pie1_PGS[!(pie1_PGS$nb_sejours_2021 == 0),]

lab <- paste0(round(pie1_PGS$nb_sejours_2021/sum(pie1_PGS$nb_sejours_2021) * 100, 2), 
              "%")

library(plotrix)
pie3D(pie1_PGS$nb_sejours_2021, 
      radius = 0.75,
      height = 0.1,
      theta = 0.7,
      border = "white",
      col = rainbow(length(lab)),
      shade = 0.5,
      labels = lab,
      labelcol = "red",
      labelcex = 0.75,
      explode = 0.2,
      main = "Répartition des séjours 2021 par type") 

^{Created on 2022-08-10 by the reprex package (v2.0.1)}