RE: covariance step
Hi Varsha,
AS Nick had mentioned in his reply, if you have NONMEM VI Level 2.0 you can add
the following options to $EST to help get around the warning and implement the
covariance step to get standard errors for the parameters:
$EST METHOD=CONDITIONAL INTER MAXEVAL=9999 PRINT=5
NOTHETABOUNDTEST NOOMEGABOUNDTEST NOSIGMABOUNDTEST
Since you did hit a boundary condition (the final parameter estimate was at
least two decimal places off from the initial estimate you provided) and you
are using NONMEM VI Level 1.0 you could also simply update your initial
estimates to be the final values of the previous run and then rerun the
existing model and that might work as well.
Best wishes,
Scott Van Wart
Quoted reply history
-----Original Message-----
From: [email protected] [mailto:[email protected]] On
Behalf Of Nick Holford
Sent: Friday, July 17, 2009 4:27 AM
To: nmusers
Subject: Re: [NMusers] covariance step
Varsha,
The message you show below is not an error. It is just a warning. Errors
and warnings are very different. An error will cause NONMEM to stop and
usually give you some clue to the problem. A warning does not mean there
is a problem -- it is just a clue that there may be a problem and its up
to the user to do some thinking. There may be nothing for you to worry
about.
As others have pointed out the message you show often turns up in
NONMEM VI because the OMEGA estimate is close to 0. Many people found
this message irritating when NONMEM VI first came out so in later
versions you can tell NONMEM (via NM-TRAN) abbreviated code) not to
issue this kind of warning.
In the end it is up to you to make a decision about whether the model is
reasonable or not. NONMEM has no idea. Indeed it has been discussed many
times on this list that there is typically no difference between models
which NONMEM reports as successful versus those that NONMEM terminates
because of rounding errors. You have to learn to take responsibility for
modelling decisions and not pay too much attention to the messages that
NONMEM issues.
Nick
Varsha Mehta wrote:
> Dear NONMEM users:
>
> I have a data set for 30 infants where I am running a 2 COMPT PK model. I got
> the following error message for the covariance step. I need some help with
> trying to fix this error. Any suggestions? should I increase the upper
> boundries of initial estimates?
>
>
>
> MINIMIZATION SUCCESSFUL
> NO. OF FUNCTION EVALUATIONS USED: 289
> NO. OF SIG. DIGITS IN FINAL EST.: 3.0
> 0PARAMETER ESTIMATE IS NEAR ITS BOUNDARY
> THIS MUST BE ADDRESSED BEFORE THE COVARIANCE STEP CAN BE IMPLEMENTED
>
> any help is appreciated.
> Thanks
>
>
> Varsha Mehta, MS(CRDSA), Pharm.D., FCCP
> Clinical Associate Professor
> Pharmacy, Pediatrics and Communicable Diseases
> Clinical Pharmacist Neonatal Critical Care
> University of Michigan
> (O) 734-936-8985
> (F) 734-936-6946
> [email protected]
>
> **********************************************************
> Electronic Mail is not secure, may not be read every day, and should not be
> used for urgent or sensitive issues
>
--
Nick Holford, Professor Clinical Pharmacology
Dept Pharmacology & Clinical Pharmacology
University of Auckland, 85 Park Rd, Private Bag 92019, Auckland, New Zealand
[email protected] tel:+64(9)923-6730 fax:+64(9)373-7090
mobile: +33 64 271-6369 (Apr 6-Jul 20 2009)
http://www.fmhs.auckland.ac.nz/sms/pharmacology/holford