Colouring wrld_simpl countries with a variable

Question

This seemed at first rather easy but turned out to be impossible to solve. I want to plot the wrld_simpl map but coloring the countries with a color gradient that depends on a variable. The countries with no values for the variable should not be filled but still show their borders.

Here is the dataframe with the variable averCL, used to color the polygons

d=read.csv(https://www.dropbox.com/scl/fi/kkvstxwjn6xrj3ezjok6t/example.csv?rlkey=cfq8sosg8ypa329l5yy1gbqyc&dl=0)

first I merge it with wrld_simpl

ws=merge(wrld_simpl, d, by = "NAME", all.x = TRUE)

Then, I create a color palette for the variation in averCL and give white color to the countries with no values of the variable

ws=merge(wrld_simpl, d, by = "NAME", all.x = TRUE)
ws$col=as.factor(ws$averCL)
levels(ws$col)=colorRampPalette(c("lightpink", "darkred"))(length(levels(ws$col)))
ws$col[is.na(ws$col)]<-"white"

Finally, I thought the countries should be plotted with that color gradient, but no...other colors appear (green..yellow..) instead. Why?

 plot(ws, col =  ws$col)

Answer 1

Its using the integer value behind the factor for the colouring, and then getting the colour from the current palette.

Convert the factor to character if you want to use a factor column's levels as colouring values:

plot(ws, col=as.character(ws$col))

Also I recommend you convert as much of your workflow to sf and associated spatial data objects, as well as using tmap to make thematic maps.