[R] Generate normally distributed correlation matrix

Richard Wang richcmwang at gmail.com
Thu Apr 28 15:37:02 CEST 2011


I would like to ask a statistics questions.  One way to generate a correlation matrix with normally distributed entries is by generating a matrix A with uniform or normal rows.  Then normalize each row to get B.  Then B times transpose B is the desired matrix.  However, the standard deviation of the entries is not large enough for my purpose.  Is there a method to generate a normally distributed correlation matrix with a controlled (larger) standard deviation?

Thanks for any help.


More information about the R-help mailing list