RE: Help: Non-positive semi-definite message

From: "KOWALSKI, KENNETH G. [R&D/1825]" <kenneth.g.kowalski@pharmacia.com> Subject: RE: Help: Non-positive semi-definite message Date: Tue, 21 Aug 2001 17:27:50 -0500 Nick, >I was thinking of a value of around 0.3 which I have seen for CL and V. This >seems more reasonable than 0 or 1. I disagree. If you arbitrarily fix the correlation to 0.3 when NONMEM wants to estimate it to be 1.0 I think you will still run into simulations of individual data that can be very unrealistic more so than restricting the correlation to 1.0. To me the real proof of what is more reasonable is to do a simulation (posterior predictive check). I'll bet you a six-pack that 1.0 will be more reasonable than 0.3. Ken