I frequently use kernel density plots to illustrate distributions. These are easy and fast to create in R like so:
set.seed(1)
draws <- rnorm(100)^2
dens
With the polygon()
function, see its help page and I believe we had similar questions here too.
You need to find the index of the quantile values to get the actual (x,y)
pairs.
Edit: Here you go:
x1 <- min(which(dens$x >= q75))
x2 <- max(which(dens$x < q95))
with(dens, polygon(x=c(x[c(x1,x1:x2,x2)]), y= c(0, y[x1:x2], 0), col="gray"))
Output (added by JDL)