NONMEM Users Network Archive

Hosted by Cognigen

Re: simulation question

From: Ethan Wu <ethan.wu75>
Date: Fri, 6 Feb 2009 05:25:55 -0800 (PST)

Dear Martin and Elodie and others,   Thanks for pointing out this nice =
function of PsN--  I think I will give it a try this time.    =

tin.bergstrand
oglemail.com>; Ethan Wu <ethan.wu75

ation question Dear all,   PsN is useful tool for this purpos=
e but I think that the  functionality that should be used is called “St=
ochastic Simulation and Estimation (SSE)”. The Bootstrap functionality is=
 primarily used for obtaining parameter uncertainties.   A detailed use=
rguide for the SSE functionality is presented at the PsN webpage (see link =
below).   http://psn.sourceforge.net/PDF_docs/sse_userguide.pdf  =

lp (psn-general
  From:owner-nmusers
m] On Behalf Of Sebastian Ueckert Sent: Friday, February 06, 2009 10:51=

ation question   Dear Ethan, the simplest solution to solve your pr=
oblem would be to use the bootstrap command of PSN (http://psn.sourceforge.=
net/). With PSN installed you would simply do: bootstrap final_model.m=
od -samples=200 PSN would take care of unsuccessful runs and provide=
 a nice summary of the individual estimates. Best regards Sebastian=


udy designs by:  -- run 200 simulations with the final model (develope=
 from original dataset)  -- fit the final model to the 200 simulated da=
taset To achieve above, I used $SIM SUBPROB=200 option  however, no=
nmem would  completely stop after running into estimation problem a=
t one specific simulation/estimation cycles, for some designs it stop eve=
n before 10th iterations. Is there anyway nonmem could continue go on? =
Received on Fri Feb 06 2009 - 08:25:55 EST

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.