discretizing viridis ggplot color scale

Question

I have an ordered factor variable that I would like to plot using ggplot2. Is there any way I can use scale_color_viridis(), a continuous color scale, with this ordered factor without casting the factor to numeric? The straightforward

iris$Sepal.Width <- ordered(iris$Sepal.Width)

ggplot(iris, aes(Sepal.Length, Petal.Length, color=Sepal.Width)) + 
  geom_point() + 
  scale_color_continuous()

doesn't work.

What's wrong with casting it to numeric? Can you please delete the same question with different title. — mtoto, Mar 10 '17 at 21:42

score 13 · Accepted Answer · answered Mar 10 '17 at 21:46

13

Viridis has a discrete = TRUE option.

iris$Sepal.Width <- ordered(iris$Sepal.Width)

ggplot(iris, aes(Sepal.Length, Petal.Length, color=Sepal.Width)) + 
geom_point() + 
viridis::scale_color_viridis(discrete = TRUE)

answered Mar 10 '17 at 21:46

detroyejr

1,094
1
9
14

score 3 · Answer 2 · answered Apr 11 '18 at 15:32

3

Last version of {ggplot2} (dev: 2.2.1.9000) now has a viridis scale included.
You can use scale_colour_viridis_d() for discrete values or scale_fill_viridis_c() for continuous values.

In your case :

iris$Sepal.Width <- ordered(iris$Sepal.Width)

ggplot(iris, aes(Sepal.Length, Petal.Length, color=Sepal.Width)) + 
  geom_point() + 
  scale_colour_viridis_d()

answered Apr 11 '18 at 15:32

Sébastien Rochette

6,536
2
22
43

This is now in ggplot2 stable release 3.0.0 – Droplet Jul 27 '18 at 19:04

discretizing viridis ggplot color scale

2 Answers2