How to assign specfic colours to specifc categorical variables in R?

问题

I'm a complete beginner to R and have this question. I'm using the following code to generate a colour list and then create a massive scatterplot matrix. I want to assign specific colours to the first column of my matrix(categorical with 4 categories). Running this code works fine but how do I verify that the colours that I intend to specify for each of the categorical variables is correct?

Basically I want to achieve green for 'control', orange for 'low', brown for 'medium' and black for 'high'.

col.list<-c("green","orange","brown","black")

palette(col.list)

pairs(Indices[,4:17], col=Indices[,1])

Thank you for any help!

回答1:

The way you're doing it is correct. If you want to check that indeed the colours correspond to your group, you can, for example do it that way (here with a reproducible example):

set.seed(1)
a <- data.frame(Group=factor(sample(c("control","low","medium","high"),20,TRUE),
                             levels= c("control","low","medium","high")),
                x=rnorm(20),y=rnorm(20))
col.list <- c("green","orange","brown","black")
palette(col.list)
pairs(a[,2:3], col=a[,1])

What col=a[,1] does is actually palette()[a[,1]] (which works IF the content of the column is a factor or an integer), so let's see:

palette()[a[,1]]
[1] "orange" "orange" "brown"  "black"  "green"  "black"  "black"  "brown"  "brown"  "green"  "green"  "green"  "brown"  "orange"
[15] "black"  "orange" "brown"  "black"  "orange" "black" 

table(a[,1], palette()[a[,1]])
         black brown green orange
  control     0     0     4      0
  low         0     0     0      5
  medium      0     5     0      0
  high        6     0     0      0

The only thing you really have to worry about is that the content of Indices[,1] is a factor whose levels are ordered in the same order as the corresponding color list.

来源：https://stackoverflow.com/questions/19874240/how-to-assign-specfic-colours-to-specifc-categorical-variables-in-r

标签

plot

scatter