Probability, Random Variables, and Random Processes: Theory and Signal Processing Applications
by John J. Shynk
4.7 GENERATING DEPENDENT RANDOM VARIABLES
One can easily generate a multivariate pdf
for N independent marginal pdfs as the product
, and this can be done even if the marginal pdfs are from different distribution families as illustrated in previous examples. Unfortunately, there are not many widely used multivariate distributions for dependent random variables, as in the case of the bivariate Gaussian and Student's t distributions where the correlation coefficient ρ appears explicitly. Dependency can be introduced for other random variables by implementing the following procedure, which we describe for the bivariate case (N = 2). This method does not yield a closed-form expression for the desired joint pdf, but instead is a means of generating samples for a computer simulation:
- Let
be a random vector with a bivariate standard Gaussian pdf and a prespecified value for the correlation coefficient
. Note that for the standard bivariate pdf, the mean is zero and
. This matrix was defined ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access