01-30-2015 07:39 AM
I have a question about proc model. When I want to simulate correlated observations using copulae I use the proc model like this (this is a dummy code):
proc model out=sim(where=(_REP_ > 0)) data=histdata sdata=covmat;
y = 0;
errormodel y ~ Uniform(0,1);
z = 0;
errormodel z ~ Uniform(0,1);
solve y z / random=500 seed=12345 copula=(normal );
The covariance matrix that i load into the procedure is not positive semi-definite, so i get the following NOTE in the log:
NOTE: The covariance matrix of the residuals is not positive semidefinite. An approximation of this matrix will be used instead.
My question is that is there any way (using ODS or sg like that) to get the final covariance matrix that SAS created and that was used for the simulation?