RE: Behavioural/Result differences for NONMEM version 5.
Dear Chris,
You are switching both NONMEM versions and compiler versions and operating
system versions. I'd be amazed if you found identical results on all your
models. Only the very simple or very well defined ones will not move around,
but for any real-life model, numerical outcome will depend on all these factors
plus additional influences like the operating system itself (i.e. Linux,
Windows, OSX), running 32 or 64 bit versions of OS and/or software, the whole
multitude of compiler switches, running NONMEM with or without MPI, and finally
hardware differences like Intel or AMD processors.
To get back to your question: I'd be surprised if the Redhat version would
affect your results but changing the Intel compiler version almost certainly
will (unless you've been particularly fortunate with your choice of compiler
switches) and so will changing NONMEM version.
Good luck,
Rik Schoemaker, PhD
Life Science Services - Exprimo
Senior Consultant
Phone:
+31 20 441 6410
E-mail:
[email protected] <mailto:[email protected]%20>
Web:
http://www.sgs.com/exprimo
SGS Exprimo NV
Generaal De Wittelaan 19A Bus 5
B-2800 Mechelen, Belgium
Switchboard: +32 15 27 32 45
This e-mail is confidential. It is also privileged or otherwise protected by
work product immunity or other legal rules. The information is intended to be
for use of the individual or entity named above. If you are not the intended
recipient, please be aware that any disclosure, copying, distribution or use of
the contents of this information is prohibited. You should therefore delete
this message from your computer system. If you have received the message in
error, please notify us by reply e-mail. The integrity and security of this
message cannot be guaranteed on the Internet. Thank you for your co-operation.
Quoted reply history
From: [email protected] [mailto:[email protected]] On
Behalf Of Chris Franklin
Sent: 05 August 2013 17:16
To: [email protected]
Subject: [NMusers] Behavioural/Result differences for NONMEM version 5.
Hi,
I know we all predominantly use NONMEM versions after NONMEM version 5. But we
retain old versions to reproduce or work on historic activities.
Anyway the question/issue:
We are upgrading our infrastructure from RedHat 4 to RedHat 5 and Intel
compiler from 10 to 12.
We now have models that worked historically but we get anomalous behaviour
varying from differences in theta and eta estimates in the 3rd significant
digit up to runs that hang during the convergence steps.
Has anyone else experienced these?
Note: we have been informed of two compiler warnings during the installation
phase:
ifort: command line remark #10010: option '-Vaxlib' is deprecated and will be
removed in a future release. See '-help deprecated'
ifort: command line remark #10279: option '-xW' is deprecated and will be
removed in a future release. See '-help deprecated'
Your feedback/advice on this is most welcome.
Regards,
Chris Franklin
________________________________
This e-mail was sent by GlaxoSmithKline Services Unlimited
(registered in England and Wales No. 1047315), which is a
member of the GlaxoSmithKline group of companies. The
registered address of GlaxoSmithKline Services Unlimited
is 980 Great West Road, Brentford, Middlesex TW8 9GS.