ggplot - extract the values associated with breaks in a continuous colour scale - r

Question

I would like to extract the breaks and the colour values associated with a ggplot continuous colour scale. There are multiple answers to finding the colour associated with each date point (like this), which can also be used to get discrete scale values, but I haven't seen an approach for a continuous colour scale. I don't want to force the scales, just retrieve the values that ggplot generates.
example:
library(ggplot)
df <- data.frame(x = 1:10, y = 1:10, col = 11:20)
ggplot(df) +
geom_point(aes(x = x, y = y, colour = col))
I would like to get a data frame showing breaks (12.5, 15, 17.5, 20) and the colour values associated with them.
Many thanks!

Answer 1

There are two ways of doing this, once with building the plot and once without building the plot.
If we build the plot;
library(ggplot2)
df <- data.frame(x = 1:10, y = 1:10, col = 11:20)
ggplot(df) +
geom_point(aes(x = x, y = y, colour = col))
We can extract the scale and use it to retrieve the relevant information.
# Using build plot
build <- ggplot_build(last_plot())
scale <- build$plot$scales$get_scales("colour")
breaks <- scale$get_breaks()
colours <- scale$map(breaks)
data.frame(breaks = breaks, colours = colours)
#> breaks colours
#> 1 NA grey50
#> 2 12.5 #1D3F5E
#> 3 15.0 #2F638E
#> 4 17.5 #4289C1
#> 5 20.0 #56B1F7
Alternatively, we can skip building the plot and use the scales themselves directly, provided we 'train' the scales by showing it the limits of the data.
scale <- scale_colour_continuous()
scale$train(range(df$col))
breaks <- scale$get_breaks()
colours <- scale$map(breaks)
data.frame(breaks = breaks, colours = colours)
As you can see, the default breaks algorithm produces an out-of-bounds break. If you want to use the information later on, it might be good to filter those out.