NONMEM Users Network Archive

Hosted by Cognigen

NM7 Question on METHOD=IMP

From: Vu, Thuy <thuy>
Date: Thu, 29 Oct 2009 16:12:21 -0700

Dear NM7 users,

According to the examples in the users guide, the IMP step followed the SAE=
M step with EONLY=1 to evaluate the objective function. In all of my mode=
ls (for the same data set), I have these two $EST steps:
$EST METHOD=SAEM INTER FILE=saem.txt NBURN=3000 NITER=2000 PRINT==
5
CTYPE=3 GRD=DDDDDDDDDSSDDD SEED=20091013 ISAMPLE=2
$EST METHOD=IMP EONLY=1 FILE=imp.txt NITER=300 ISAMPLE=3000 PRINT=
=1

In my base model, the OBJ trend in the IMP step is as follows (truncated he=
re):
 iteration 296 OBJ= -1796.0722480934633
 iteration 297 OBJ= -1802.1845380634943
 iteration 298 OBJ= -1795.4946269639131
 iteration 299 OBJ= -1795.8286754523187
 Elapsed estimation time in seconds: 230513
 iteration 300 OBJ= -1791.4880075805286

When I add IOV on one of the parameters (to the base model), I got the foll=
owing trend, which showed the OBJ became worse:
 iteration 296 OBJ= 52215.178158178285
 iteration 297 OBJ= 51758.149614882743
 iteration 298 OBJ= 51850.531967877905
 iteration 299 OBJ= 53377.558672094958
 Elapsed estimation time in seconds: 314605
 iteration 300 OBJ= 53030.267320528052

When I compared the GOF plots and parameter estimates for these two models,=
 they looked pretty much the same. I could not find any obvious differences=
 between them. How would I interpret/compare the OBJ for these 2 models? Wh=
at could be the reasons for such difference in the OBJ for the 2 models?

Thank you in advance for your help,
-Thuy




Received on Thu Oct 29 2009 - 19:12:21 EDT

The NONMEM Users Network is maintained by ICON plc. Requests to subscribe to the network should be sent to: nmusers-request@iconplc.com.

Once subscribed, you may contribute to the discussion by emailing: nmusers@globomaxnm.com.