Rev | Line | |
---|
[37] | 1 | function x = randnorm(n, m, S, V) |
---|
| 2 | % RANDNORM Sample from multivariate normal. |
---|
| 3 | % RANDNORM(n,m) returns a matrix of n columns where each column is a sample |
---|
| 4 | % from a multivariate normal with mean m (a column vector) and unit variance. |
---|
| 5 | % RANDNORM(n,m,S) specifies the standard deviation, or more generally an |
---|
| 6 | % upper triangular Cholesky factor of the covariance matrix. |
---|
| 7 | % This is the most efficient option. |
---|
| 8 | % RANDNORM(n,m,[],V) specifies the covariance matrix. |
---|
| 9 | % |
---|
| 10 | % Example: |
---|
| 11 | % x = randnorm(5, zeros(3,1), [], eye(3)); |
---|
| 12 | |
---|
| 13 | if nargin == 1 |
---|
| 14 | x = randn(1,n); |
---|
| 15 | return; |
---|
| 16 | end |
---|
| 17 | [d,nm] = size(m); |
---|
| 18 | x = randn(d, n); |
---|
| 19 | if nargin > 2 |
---|
| 20 | if nargin == 4 |
---|
| 21 | if d == 1 |
---|
| 22 | S = sqrt(V); |
---|
| 23 | else |
---|
| 24 | S = chol(V); |
---|
| 25 | end |
---|
| 26 | end |
---|
| 27 | if d == 1 |
---|
| 28 | x = S .* x; |
---|
| 29 | else |
---|
| 30 | x = S' * x; |
---|
| 31 | end |
---|
| 32 | end |
---|
| 33 | if nm == 1 |
---|
| 34 | x = x + repmat(m, 1, n); |
---|
| 35 | else |
---|
| 36 | x = x + m; |
---|
| 37 | end |
---|
Note: See
TracBrowser
for help on using the repository browser.