RE: When should a long run be aborted?
From: "Sam Liao" sliao@pharmaxresearch.com
Subject: RE: [NMusers] When should a long run be aborted?
Date: Mon, November 22, 2004 2:09 pm
Hi Paul:
I would like to suggest you to try log-transform of your DV and F by FO
method instead of FOCE method. You can check out some discussion of this
on the usersnet.
I had experience fitting one parent, 3 metabolites in 8 DE recently, it took
more than 24 hrs to converge with log-transformed data using FO method. So,
if you try my suggestion, I expected it converged in few hours. A faster
CPU will certainly be helpful to you too.
Best regards,
Sam Liao
PharMax Research