NONMEM Users Network Archive

Hosted by Cognigen

Re: compiler option issue

From: Nick Holford <n.holford>
Date: Sat, 03 Feb 2007 11:53:54 +1300

Maud,

NONMEM is itself unreliable at determining its termination status. Differences in termination status are not necessarily due to compiler options. When random instances of data are studied the parameter estimates obtained by NONMEM are essentially indistiguishable irrespective of whether termination is successful or fails due to rounding errors.

The Compaq VF 6.6 compiler is now obsolete. It has been replaced by the Intel VF compiler. The compiler will makes its own choices for generating code depending on the processor it finds. This can be explicitly controlled by the user to create code for a specific target processor. If you want a qualified NONMEM installation on each computer you may wish to create different versions depending on the target processor.

It is important to use the same compiler version and compiler options for each NONMEM run as was used to compile the main NONMEM system.

I would recommend that you use NMQual for installation and qualification of NONMEM on each machine.
http://metruminstitute.org/downloads/index.shtml
If you use NMQual to install NONMEM on each target machine then you can be assured that the processor and compiler option issues will be taken care of. If you know that the processor is the same on the target machine then you can use a network copy of an NMQual generated NONMEM for NONMEM runs or copy the files to each target machine.

Nick

> Destree Maud wrote:
>
> Dear All,
> I work as application analyst and have to provide support to NONMEM users.
> I have recently discovered NONMEM and have a few questions about compiler and optimization options.
> I use NONMEM v on a HP xw6400 workstation with Compaq Visual Fortran 6.6.
> I was first suprised to see that for a particular run I had a rounding error terminaison with /optimize:1 /fpe:0
> and that the job terminated well with /optimize:4 /fpe:0. I would have expected the reverse behaviour.
>
> My second question is how important is it the recompile the whole NONMEM code when changing PC?
> Is it possible to recompile NONMEM on a laptop (also with Compaq Visual Fortran 6.6) then copy sources to the workstation knowing that with PdxPOP or nmfe5.bat only a part of code is recompiled on the machine used for the run?
>
> Is it important to have the same optimisation options for both parts of code?
>
> Could you provide some advice about these issues?
> Thank you
> Best regards
>
> Maud
>
> Maud Destrée
>
> IT Global Drug Development
> UCB Pharma S.A.
>
> ---------------------------------------------------------
> Legal Notice: This electronic mail and its attachments are intended solely for the person(s) to whom they are addressed and contain information which is confidential or otherwise protected from disclosure, except for the purpose for which they are intended. Dissemination, distribution, or reproduction by anyone other than the intended recipients is prohibited and may be illegal. If you are not an intended recipient, please immediately inform the sender and return the electronic mail and its attachments and destroy any copies which may be in your possession. UCB screens electronic mails for viruses but does not warrant that this electronic mail is free of any viruses. UCB accepts no liability for any damage caused by any virus transmitted by this electronic mail.
>
> ---------------------------------------------------------

--
Nick Holford, Dept Pharmacology & Clinical Pharmacology
University of Auckland, 85 Park Rd, Private Bag 92019, Auckland, New Zealand
email:n.holford
http://www.health.auckland.ac.nz/pharmacology/staff/nholford/
Received on Fri Feb 02 2007 - 17:53:54 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.