Re: Help: Non-positive semi-definite message

From: Nick Holford <n.holford@auckland.ac.nz> Subject: Re: Help: Non-positive semi-definite message Date: Wed, 22 Aug 2001 07:16:41 +1200 Ken, "KOWALSKI, KENNETH G. [R&D/1825]" wrote: > > In the setting when NONMEM wants to estimate the correlation for an > off-diagonal element of omega to 1 what value do you propose fixing the > covariance value to? That is *exactly* what I am proposing. The analogous situation would be fixing KA to a reasonable value when you dont have enough data in the absorption phase to estimate it (Janet Wade did some simulations on this in JPB). If you pick a reasonable value then the other estimates will be reasonable. If you fix the covariance at various values (and > estimate everything else) I believe you will find that the minimum OFV will > correspond to the estimate of the covariance which leads to the correlation > being set to 1.0. Thus, any fixed value that results in the correlation > being constrained to less than 1.0 should have a higher OFV and hence will > be less parsimonious. I would accept that fixing the covariance would constrain the OFV but sometimes you should not believe the OFV as the only criterion of a reasonable fit. >I don't know how one arbitrarily sets the correlation > to be something between 0 and 1 when the data/design do not provide > sufficient information to estimate it different from 1. If you set it > arbitrarily close to 1 like 0.95 or 0.99 so as not to increase the OFV too > much then you have a "near perfect correlation" and your concern would still > exist at least approximately. 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. -- Nick Holford, Divn Pharmacology & Clinical Pharmacology University of Auckland, 85 Park Rd, Private Bag 92019, Auckland, New Zealand email:n.holford@auckland.ac.nz tel:+64(9)373-7599x6730 fax:373-7556 http://www.phm.auckland.ac.nz/Staff/NHolford/nholford.htm