I want to plot a grouped density plot using ggplot. Usually, you would use aes(fill = variable_to_group_by) to obtain several densities, each colored/filled by variable_to_group_by.
However, I would like the fill of each density distribution to take the color of a continuous variable (which is unique within each group) instead of the discrete factor that I am using for grouping.
So, I would like to have a distribution per cut as below, but the ‘fill’ color of the distributions should rather be by the continuous variable mean_carat(made below).
diamonds %>%
group_by(cut) %>%
mutate(mean_carat = mean(carat)) %>%
ungroup() %>%
ggplot(aes(x = price, fill = cut, color = mean_carat)) +
geom_density(alpha = 0.3)
Thanks in advance!
>Solution :
Using the group aes you could group by cut but fill by mean_carat:
library(dplyr)
library(ggplot2)
diamonds %>%
group_by(cut) %>%
mutate(mean_carat = mean(carat)) %>%
ungroup() %>%
ggplot(aes(x = price, group = cut, fill = mean_carat)) +
geom_density(alpha = 0.3)
