Understanding the "cut" function in R

Question

I tried making the following histogram in R (randomly select 10% of all rows and color them red):

a = rnorm(100000,60000,1000)
b = a

c = data.frame(a,b)
color <- c("black", "red")     
color_1 <- sample(color, nrow(c), replace=TRUE, prob=c(0.9, 0.1))
c$color_1 = as.factor(color_1)


hist(c$a, col = c$color_1, 100000, main = "title")

legend("topleft", legend=c("group a", "group b"),
       col=c("red", "black"), lty = 1, cex=0.8)
title(
      sub = "some title")

Problem: But for some reason, the colors are not showing up:

I tried to see if other commands might get the colors to show up:

hist(c$a, col = color_1, 100000, main = "title")

Or trying to remove the color variable as a "factor":

a = rnorm(100000,60000,1000)
b = a

c = data.frame(a,b)
color <- c("black", "red")     
color_1 <- sample(color, nrow(c), replace=TRUE, prob=c(0.9, 0.1))
c$color_1 = color_1


hist(c$a, col = c$color_1, 100000, main = "title")

legend("topleft", legend=c("group a", "group b"),
       col=c("red", "black"), lty = 1, cex=0.8)
title(
      sub = "some title")

I also tried to follow the advice from this question here (Partially color histogram in R) :

h = hist(c$a, col = c$color_1, breaks = 100000, main = "title")

legend("topleft", legend=c("group a", "group b"),
       col=c("red", "black"), lty = 1, cex=0.8)
title(
      sub = "some title")



cuts <- cut(h$breaks, c(-Inf,Inf))
plot(h, col=cuts)

But this also did not work. I think this might be because I am not using the "cut" function correctly?

Can someone please show me how to fix this?

Thanks

Answer 1

Here is what I understand of the question:

1. Plot a vector's histogram;
2. 10% of the bars are randomly selected;
3. And have a different color.

First remake the example data set. Apparently, there is no need for a 2nd vector b. And the RNG seed is set, in order to make the results reproducible.

set.seed(2021)
a <- rnorm(100000, 60000, 1000)
c <- data.frame(a)
color <- c("black", "red")     
n_colors <- length(color)

Now get the histogram data but don't plot it. Then select as many color codes (at most n_colors) as counts. And plot the histogram.

h <- hist(c$a, breaks = "FD", plot = FALSE)
i_col <- sample(n_colors, length(h$counts), replace = TRUE, prob = c(0.6, 0.4))
plot(h, main = "title", col = color[i_col])

legend("topleft", legend=c("group a", "group b"),
       col=c("red", "black"), lty = 1, cex=0.8)
title(sub = "some title")