T S R Matrices

From:"Ludden, Thomas" Subject:[NMusers] T S R Matrices Date:Tue, March 12, 2002 9:36 am According to Guide VIII, one should be able to obtain the R matrix by including the option PRINT=R on the $COV record. Tom Ludden

From:"Stephen Duffull" Subject:[NMusers] RE: T S R matrices Date:Mon, March 11, 2002 7:49 pm Hi Sorry to continue this thread further... Could someone please clarify why the R matrix may not be available when selecting MATRIX=R in the $COV statement? I'm not sure why the method would fail to provide the matrix? I have often requested this matrix and NONMEM returns the answer ... R MATRIX ALGORITHMICALLY SINGULAR AND ALGORITHMICALLY NON-POSITIVE DEFINITE COVARIANCE STEP ABORTED The confusing thing for me is that it doesn't matter that the R matrix is singular since I'm not asking NONMEM to find its inverse - I'm just interested in seeing the matrix itself. Is there some way in which I can get NONMEM to provide the matrix even if it is singular. Regards Steve ***************************************** Stephen Duffull School of Pharmacy University of Queensland Brisbane 4072 Australia Tel +61 7 3365 8808 Fax +61 7 3365 1688 http://www.uq.edu.au/pharmacy/duffull.htm

From:"Gastonguay, Marc" Subject:RE: [NMusers] RE: T S R matrices Date:Tue, March 12, 2002 7:00 am Steve, I think your results make sense, although I see how this can be confusing. When you ask for MATRIX=R your are actually asking NONMEM to use the inverse R matrix. From NONMEM Help Topic $COVARIANCE: "MATRIX=R requests that the inverse R matrix be used. MATRIX=S requests that the inverse S matrix be used." Regards, Marc Gastonguay

From:"Stephen Duffull" Subject:RE: [NMusers] T S R Matrices Date:Tue, March 12, 2002 5:20 pm Thanks for your patience. PRINT=R gave the printout that I needed. Regards Steve ***************************************** Stephen Duffull School of Pharmacy University of Queensland Brisbane 4072 Australia Tel +61 7 3365 8808 Fax +61 7 3365 1688 http://www.uq.edu.au/pharmacy/duffull.htm