| Library: | times |
| See also: |
| Quantlet: | normalcorr | |
| Description: | generates correlated pseudo random normal variates using the Cholesky factorization. |
| Usage: | x = normalcorr(n,C) | |
| Input: | ||
| n | scalar, time, n represents the number of observations to be generated in each series | |
| C | p x q matrix, correlation matrix, e.g. obtained with corr() | |
| Output: | ||
| x | ( n, sqrt(size(C)) )-dimensional array, containing the simulated trajectories | |
library("multi")
library("times")
randomize(100)
C = reshape(#(1,0.5,0.5,1),#(2,2))
x = normalcorr(200,C) ; simulation of correlated normal variates
corr(x)
Contents of r [1,] 1 0.52418 [2,] 0.52418 1 Note: The output vectors (x) approximately have a correlation structure defined by C.