Hardcode hexadecimal string values as color in ggplot

Question

I have a data frame with "x" and "y" columns as numeric values, and a third column "cluster" as a hexidecimal string, an example seen below:

library(ggplot2)
library(scales)
colList = c(scales::hue_pal()(3),"#520090")
dat = data.frame(x=runif(100,0,1),y=runif(100,0,1),cluster=sample(1:4, 100, replace=T))
dat$cluster = factor(dat$cluster)
levels(dat$cluster) = c(colList)
head(dat)

I am trying to create a scatterplot with "x" and "y" columns mapped to the x and y axis, and with those points colored according to the hexadecimal value stored in the "cluster" column. I have tried the following:

ggplot(dat,aes(x,y)) + 
  geom_point(aes(colour = cluster), alpha=0.5)

However, this simply assigns the default first four values stored in scales::hue_pal()(4), and I have changed the last one to a dark purple color with hexadecimal value #520090. I also am trying to change the default hexadecimal values from appearing as the text in the legend. I tried unsuccessfully to hardcode in "Cluster 1", "Cluster 2", ..., "Cluster 4" as the legend text:

ggplot(dat,aes(x,y)) + 
  geom_point(aes(colour = cluster), alpha=0.5) + 
  theme(legend.text = element_text("Cluster 1", "Cluster 2", "Cluster 3", "Cluster 4"))

Any advice is much appreciated!

Answer 1

In order to color the dots based on the cluster identity, the cluster name (i.e., your hex values) needd to be mapped to a set of aesthetic values.

Since you want to have the hex values from the cluster column to represent actual colors, you can use the scale_color_manual function and give the levels of the cluster column as the values parameter. To changes the labels, simply set the desired labels value.

ggplot(dat, aes(x,y)) + geom_point(aes(colour = cluster), alpha=0.5) +
scale_color_manual(values = levels(dat$cluster), 
                   labels = c("Cluster1","Cluster2","Cluster3", "Cluster4"))