From: "=?big5?B?YmVhdXR5ZW15?=" <beautyemy@sinamail.com>
Subject: [NMusers] error message from $covariance
Date: Wed, 14 Nov 2001 18:41:10 +0800

Dear NONMEM users:
After running $covarience, I've got error message as following:

MINIMIZATION SUCCESSFUL
NO. OF FUNCTION EVALUATIONS USED: 1163
NO. OF SIG. DIGITS IN FINAL EST.: 3.2
R MATRIX ALGORITHMICALLY SINGULAR
AND ALGORITHMICALLY NON-POSITIVE-SEMIDEFINITE
COVARIANCE STEP ABORTED

Could anyone give me any advise how to refine my model??
Thanks a lot!!

==================================================================
·s®öºô¡D§K¶O 50MB 1q¤l«H¿c http://www.sinamail.com

 

 

*****

 

 

From: bvatul@ufl.edu
Subject: Re: [NMusers] error message from $covariance
Date: Wed, 14 Nov 2001 12:42:53 -0500

Hello

Please check the estimates of etas. Either you might
simplify your model by reducing the etas or you may fix
the eta to zero. You might be observing some etas which
are very low, something like E-10 or so.

Hope this helps

Atul
Post-doctoral fellow
University of Florida