The Art of R Programming

Simulation Programming in R

One of the most common uses of R is simulation. Let’s see what kinds of tools R has available for this application.

Built-In Random Variate Generators

As mentioned, R has functions to generate variates from a number of different distributions. For example, rbinom() generates binomial or Bernoulli random variates.^[3]

Let’s say we want to find the probability of getting at least four heads out of five tosses of a coin (easy to find analytically, but a handy example). Here’s how we can do this:

> x <- rbinom(100000,5,0.5)
> mean(x >= 4)
[1] 0.18829

First, we generate 100,000 variates from a binomial distribution with five trials and a success probability of 0.5. We then determine which of them has a value 4 or 5, resulting ...