How do you generate multivariate Gaussian random numbers in R?

Question

How do we generate data points following a Gaussian (normal) distribution in R?

Suppose I want to generate points in 2d (or higher dimensional) space that follow a Gaussian distribution. How do I do this using R?

Answer 1

Gaussian distributions are for one dimensional random variables. You can generate them using rnorm.

rnorm(100, mean = 3, sd = 2)

For the higher dimensional case you want a multivariate normal distribution instead. Try mvrnorm in the MASS package, or rmvnorm in the mvtnorm package.

library(mvtnorm)
rmvnorm(100, mean = c(3, 5), sigma = matrix(c(1, 0.5, 0.5, 2), nrow = 2))

Further reading: ?Distributions and the CRAN Task View on distributions.

Answer 2

One dimensional: ?rnorm. More dimensions: install and load package mvtnorm and use rmvnorm().