R barplot - keep same colours after sorting - r

Question

I want to plot percentages for 3 variables (a,b,c) one after the others. So I have a matrix (%) for a set of activities for variable a, b and c.
dta = structure(c(0.0073, 0.1467, 0.0111, 0.0294, 0.0451, 0.0031, 0.1823,
0.0452, 0.2212, 0.1123, 7e-04, 0.1138, 0.0723, 0.1649, 0.0634),
.Dim = c(5L, 3L),
.Dimnames = list(c("c Work", "e Travel/Commute",
"f Cooking", "g Housework", "h Odd jobs"),
c("a", "b", "c")))
However, I would like to plot each variables sorted and but keeping the same colours for the set of activities.
So this is the colours of the activities.
library(RColorBrewer)
rc = c(brewer.pal(n = 5, name = 'Set2'))
kol = list()
kol$act <- c("c Work", "e Travel/Commute", "f Cooking", "g Housework", "h Odd jobs" )
kol$colours <- rc
kol = as.data.frame(kol)
act colours
1 c Work #66C2A5
2 e Travel/Commute #FC8D62
3 f Cooking #8DA0CB
4 g Housework #E78AC3
5 h Odd jobs #A6D854
So here are my barplots
par(mfrow = c(2,2))
barplot(dta[,1], horiz = T, las = 2, col = kol$colours)
barplot(dta[,2], horiz = T, las = 2, col = kol$colours)
barplot(dta[,3], horiz = T, las = 2, col = kol$colours)
So I want is to sort by keep the same colours for the activities
par(mfrow = c(2,2))
barplot(sort(dta[,1]), horiz = T, las = 2)
barplot(sort(dta[,2]), horiz = T, las = 2)
barplot(sort(dta[,3]), horiz = T, las = 2)
How can I make it "match" ?

Answer 1

You can use the function match to match the names of the "entities" and the desired colours, for example, for the first column:
kol$colours[match(names(sort(dta[,1])), kol$act)]
so, to obtain your barplot, just do:
par(mfrow = c(2,2), mar=c(5, 8, 4, 1)) # also modifying the margins to make the names fit in
for (i in 1:3) {
barplot(sort(dta[,i]), horiz = T, las = 2, col=kol$colours[match(names(sort(dta[, i])), kol$act)])
}