How to assign a maximum value shown on a colour ramp in R

Question

This is probably a simple thing to do in R, but I want to set a maximum value shown on a predefined colour ramp. I've been asked to use cmocean colour palettes and have a numeric variable with values ranging from 0 to ~10. However, I want the darkest colour to apply to values => 5. For example, any value of 5 and above will be displayed as the same/ "maximum" colour available. I don't want to remove the data above 5 so subsetting isn't an option. I hope this makes sense, thank you!

score 1 · Answer 1 · answered Dec 22 '20 at 15:13

I guess something like this? You cut the values according to a defined breaks, and in your case, 5 to 10 is 1 break and you assign one color to it:

library(cmocean)
x = sort(runif(100,0,10))
pal = cmocean("balance")(11)
lbl = cut(x,breaks = c(seq(0,5,length.out=10),10),right=FALSE)
names(pal) = unique(lbl)

Plot it, using the code from this post for a colorbar:

layout(matrix(c(1,2),ncol=2),widths=c(5,1))

plot(values,rep(1,length(values)),col=pal[lbl],pch=20)

my.colors = pal
z=matrix(1:10,nrow=1)
y=seq(0,5,len=10)
image(1,y,z,col=pal,axes=FALSE,xlab="",ylab="")
axis(2)

score 0 · Accepted Answer · answered Dec 22 '20 at 09:56

There is at least one way to do this, see if this approach works for your use-case:

library(tidyverse)
library(palmerpenguins)
#install.packages("cmocean")
library(cmocean)
library(cowplot)
library(ggbeeswarm)

plot_1 <- penguins %>% 
  na.omit() %>%
  ggplot(aes(x = species, y = flipper_length_mm, color = body_mass_g)) +
  geom_quasirandom()  +
  scale_color_cmocean(name = "deep")

plot_2 <- penguins %>% 
  na.omit() %>%
  ggplot(aes(x = species,
             y = flipper_length_mm,
             color = body_mass_g)) +
  geom_quasirandom()  +
  scale_color_cmocean(name = "deep",
                      limits = c(2500, 4000),
                      oob = scales::squish,
                      labels = c("2500", "3000", "3500", "4000+"))
cowplot::plot_grid(plot_1, plot_2, labels = "AUTO")

Explanation: Panel "A" (plot_1) doesn't have scale limits, Panel "B" (plot_2) has scale limits. The oob option ("Out Of Bounds) is also needed in addition to the scale limits in Panel "B" to 'keep' the darkest colour, otherwise the colours above 4000 are grey. I also altered the labels on the legend (labels option) to change "4000" to "4000+", but up to you whether you think that's useful or not.

Thanks for the response, this worked for me! But I just want to confirm that it isn't getting rid of anything above 4000 (for your example), it's just assigning the darkest colour to those values? I only ask because my plots before and after using your method look very similar. — BKeys, Jan 14 '21 at 06:00
It isn't getting rid of anything, it's just labelling points >4000 as the darkest shade of the colour scale. If you did want to get rid of points >4000, you can use a 'filter', e.g. `penguins %>% filter(flipper_length <= 4000) %>% ggplot(...` — jared_mamrot, Jan 14 '21 at 08:33
If this answer is satisfactory, please [accept it](https://stackoverflow.com/help/someone-answers) — jared_mamrot, Jan 14 '21 at 08:34

How to assign a maximum value shown on a colour ramp in R

2 Answers2