7

After making a figure in ggplot and using a viridis palette, I would like to get a list of Hex color codes for all colors which were randomly selected for that particular plot. Is that possible?

Here is the data:

df = data.frame(Cell = c(rep("13a",5), rep("1b",5), rep("5b",5)),
                condition = rep(c("a","b","c","d","e"), 3),
                variable = c(58,55,36,29,53, 57,53,54,52,52, 45,49,48,46,45))

And the plot:

library(ggplot2)
library(viridis)

windows(width=4, height=4 )

ggplot(df, aes(x = condition, y = variable, group = Cell, color = Cell)) +
geom_point(aes(color = Cell))+
geom_line()+
scale_color_viridis(discrete=TRUE)
Dave Gruenewald
  • 5,329
  • 1
  • 23
  • 35
user7395965
  • 397
  • 5
  • 17
  • 3
    `scales::viridis_pal()(number_of_categories)`. You've got 3 colors, so `scales::viridis_pal()(3)`. There's no randomness. – Gregor Thomas Jan 08 '20 at 18:10

1 Answers1

12

You can manually recreate the hexes used with a scales::viridis_pal()(n) call (proposed by @Gregor):

scales::viridis_pal()(length(unique(df$Cell)))

[1] "#440154FF" "#21908CFF" "#FDE725FF"

However, you can also access the underlying data of any ggplot object using ggplot_build:

Let's first save your plot as gg:

gg <- ggplot(df, aes(x = condition, y = variable, group = Cell, color = Cell)) +
  geom_point(aes(color = Cell))+
  geom_line()+
  scale_color_viridis(discrete=TRUE)

Now to access the underlying components:

ggplot_build(gg)

Since we really are only interested in the data:

ggplot_build(gg)$data

Which gives us:

[[1]]
      colour x  y group PANEL shape size fill alpha stroke
1  #440154FF 1 58     1     1    19  1.5   NA    NA    0.5
2  #440154FF 2 55     1     1    19  1.5   NA    NA    0.5
3  #440154FF 3 36     1     1    19  1.5   NA    NA    0.5
4  #440154FF 4 29     1     1    19  1.5   NA    NA    0.5
5  #440154FF 5 53     1     1    19  1.5   NA    NA    0.5
6  #21908CFF 1 57     2     1    19  1.5   NA    NA    0.5
7  #21908CFF 2 53     2     1    19  1.5   NA    NA    0.5
8  #21908CFF 3 54     2     1    19  1.5   NA    NA    0.5
9  #21908CFF 4 52     2     1    19  1.5   NA    NA    0.5
10 #21908CFF 5 52     2     1    19  1.5   NA    NA    0.5
11 #FDE725FF 1 45     3     1    19  1.5   NA    NA    0.5
12 #FDE725FF 2 49     3     1    19  1.5   NA    NA    0.5
13 #FDE725FF 3 48     3     1    19  1.5   NA    NA    0.5
14 #FDE725FF 4 46     3     1    19  1.5   NA    NA    0.5
15 #FDE725FF 5 45     3     1    19  1.5   NA    NA    0.5

[[2]]
      colour x  y group PANEL size linetype alpha
1  #440154FF 1 58     1     1  0.5        1    NA
2  #440154FF 2 55     1     1  0.5        1    NA
3  #440154FF 3 36     1     1  0.5        1    NA
4  #440154FF 4 29     1     1  0.5        1    NA
5  #440154FF 5 53     1     1  0.5        1    NA
6  #21908CFF 1 57     2     1  0.5        1    NA
7  #21908CFF 2 53     2     1  0.5        1    NA
8  #21908CFF 3 54     2     1  0.5        1    NA
9  #21908CFF 4 52     2     1  0.5        1    NA
10 #21908CFF 5 52     2     1  0.5        1    NA
11 #FDE725FF 1 45     3     1  0.5        1    NA
12 #FDE725FF 2 49     3     1  0.5        1    NA
13 #FDE725FF 3 48     3     1  0.5        1    NA
14 #FDE725FF 4 46     3     1  0.5        1    NA
15 #FDE725FF 5 45     3     1  0.5        1    NA
Dave Gruenewald
  • 5,329
  • 1
  • 23
  • 35