如何使用 R 从给定分布中模拟 m 个大小为 n 的随机样本？

Question

I know how to generate a random sample of size n from a standard statistical distribution, say exponential. But if I want to generate m such random samples of size n (ie m vectors of dimension n) how can I do it?

Answer 1

To create a n by m matrix containing m samples of size n you can use:

x <- replicate(m, rnorm(n, ...))

Obviously substituting rnorm with other distributions if desired. If you then want to store these in separate individual vectors then you can use

v <- x[ , i] 

This puts the i th column of x into v , which corresponds to the i th sample. It may be easier/quicker to just use a simple for loop altogether though:

for(i in 1:m){
  name <- paste("V", i, sep = "")
  assign(name, rnorm(n, ...))
}

This generates a random sample at each iteration, and for stage i , names the sample Vi . By the end of it you'll have m random samples named V1 , V2 , ..., Vm .