[S] "Skewed" Multivariate Random Generator

T.E.Diaz (tediaz@gwis2.circ.gwu.edu)
Fri, 8 May 1998 02:20:46 +0000


Hi All,

I want to be able to generate multivariate observations (to represent
test scores) from a distribution that is not symmetric. I will also
be simulating from the multivariate normal distribution using
rmvnorm()- another separate condition. So I was thinking that a
meaningful approach to generate skewed multivariate observations
would be to start with an output from rmvnorm() and then carry out a
transformation in order to introduce a specified skewness. Does
anybody know of a procedure along this line?

Thanks in advance for any assistance - pointer to an S code or a
reference to an analytical solution.

Ted Diaz
George Washington University
tediaz@gwu.edu
-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news