vectorized table operations in r data.table

Question

Based on given averages and standard deviations, I'm trying to fill columns with various sample runs. What I have right now is:

A = data.table(ave = c(20, 3), std = c(.1, 1))
A[, paste0("scenario", c(1:3)) := rnorm(2, ave, std)]

Which results in the table:

ave  std scenario1 scenario2 scenario3
 20  0.1   20.2377   20.2377   20.2377
  3  1.0    2.6497    2.6497    2.6497

So clearly, its running rnorm on one vector, then repeating that same vector for each column. How do I make it recalculate each column? I've tried things like

matrix(rnorm(6, ave, std), 2, 3)

or using transpose(lapply(...)) but neither seems to work.

Answer 1

您可以为每列调用rnorm ：

A[, paste0("scenario", c(1:3)) := lapply(1:3, function(x) rnorm(2, ave, std))]

Answer 2

我们可以使用replicate

A[, paste0("scenario",c(1:3)) := replicate(3, rnorm(2, ave, std), simplify = FALSE)]