[This article was first published on R – Statistical Odds & Ends, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

To plot multiple figures on a single canvas in base R, we can change the graphical parameter mfrow. For instance, the code below tells R that subsequent figures will by drawn in a 2-by-3 array:

par(mfrow = c(2, 3))


If we then run this next block of code, we will get the image below:

set.seed(10)
n <- 50; p <- 3
x <- matrix(runif(n * p), ncol = p)
for (j in 1:p) {
plot(x[, j], x[, j]^2 + 0.1 * rnorm(n),
xlab = paste0("x", j), ylab = paste0("f(x", j, ")^2 + noise"))
plot(x[, j], x[, j]^3 + 0.1 * rnorm(n),
xlab = paste0("x", j), ylab = paste0("f(x", j, ")^3 + noise"))
}


Notice how the plots are filled in by row? That is, the first plot goes in the top-left corner, the next plot goes to its right, and so on. What if we want the plots to be filled in by column instead? For instance, in our example above each feature is associated with two panels. It would make more sense for the first column to filled with plots related to x1, and so on.

There is an easy fix for that: instead of specifying the mfrow parameter, specify the mfcol parameter. See the results below:

par(mfcol = c(2, 3))
set.seed(10)
n <- 50; p <- 3
x <- matrix(runif(n * p), ncol = p)
for (j in 1:p) {
plot(x[, j], x[, j]^2 + 0.1 * rnorm(n),
xlab = paste0("x", j), ylab = paste0("f(x", j, ")^2 + noise"))
plot(x[, j], x[, j]^3 + 0.1 * rnorm(n),
xlab = paste0("x", j), ylab = paste0("f(x", j, ")^3 + noise"))
}


References: