# RE: basic questions...

From: mats karlsson <mats.karlsson>
Date: Thu, 31 Mar 2011 18:10:37 +0200

Dear Matt and all,

In the code you gave, DEL is a protection when F=0 as it may be in a =
lag-time model, not only for IWRES but also for the model as such. You =
can’t have a proportional error model with a prediction of zero. =
Add an absolute error component or switch from a lag to a transit =
compartment delay model are two possibilities. In general, I find IWRES =
useful for diagnosing the residual error model by plotting abs(IWRES) =
versus IPRED when I have reasonable amount of data per subject. An =
appropriate error model should result in a lack of trend. Others may =
have other ways of assuring that the residual error model is =
appropriate. Better alternatives are always welcome.

Best regards,

Mats

Mats Karlsson, PhD

Professor of Pharmacometrics

Dept of Pharmaceutical Biosciences

Uppsala University

Sweden

Postal address: Box 591, 751 24 Uppsala, Sweden

Phone +46 18 4714105

Fax + 46 18 4714003

From: owner-nmusers
On Behalf Of Fidler,Matt,FORT WORTH,R&D
Sent: Thursday, March 31, 2011 5:30 PM
To: Nick Holford; nmusers
Subject: RE: [NMusers] basic questions...

Paola,

I agree with Nick that there may be little use for the DEL and W =
variables.

The standard coding may be something like:

\$ERROR

IPRED = F

IRES = DV-IPRED

W = THETA(#)*F

DEL = 0

IF(W.EQ.0) DEL = 1

IWRES = IRES/(W+DEL)

Y = F+EPS(1)*W

The W represents the “weight” of the error model. In =
the above example this is proportional error.

The DEL is used to keep the unlikely divide by zero error from appearing =
when IRES/W is calculated. When W=0, IRES=RES.

Like Nick, I’m don’t find too much use in this. If W =
= 0, I think the IWRES should be missing instead of calculated as RES.

Matt.

From: owner-nmusers
On Behalf Of Nick Holford
Sent: Thursday, March 31, 2011 9:03 AM
To: nmusers
Subject: Re: [NMusers] basic questions...

Paola,

On 31/03/2011 3:02 p.m., di Gion, Paola wrote:

Dear NONMEM users,

as beginners in NONMEM we have the following questions:

- What is the meaning of "DEL" and "W" (in the subroutine \$ERROR)?

They are sort of like a virus that is spread by copying and pasting from =
old control streams passed around among beginners.

These are not standard NM-TRAN variables. Some people like to use them =
to create some diagnostic variables like IWRES. I find no use for them.

- What is the meaning of "TOL" and when should it be used?

TOL is an option for the differential equation solving subroutines =
It specifies the tolerance for the numerical solution of the =
differential equations. A reasonable starting value is TOL=6 for NM7 =
(with NSIG=2 and SIGL=6).

- Assumed we need to use \$DES, is this possible only with NONMEM version =
7 or can we use it also in NONMEM 6?

You can use \$DES with any version of NONMEM. It usually runs faster with =
NM6 than NM7. However, in some cases ADVAN13 in NONMEM7 will run faster =
than ADVAN6,8.9. Its hard to predict which ADVAN is best so just try =
different ones and use the one that runs fastest.

Thanks a lot and kind regards from rainy Cologne

Its raining in Cape Town too!

Paola & Lisa

--
Nick Holford, Professor Clinical Pharmacology
Dept Pharmacology & Clinical Pharmacology
University of Auckland,85 Park Rd,Private Bag 92019,Auckland,New Zealand
tel:+64(9)923-6730 fax:+64(9)373-7090 mobile:+64(21)46 23 53
email: n.holford
http://www.fmhs.auckland.ac.nz/sms/pharmacology/holford

_____

This e-mail (including any attachments) is confidential and may be =
legally privileged. If you are not an intended recipient or an =
authorized representative of an intended recipient, you are prohibited =
from using, copying or distributing the information in this e-mail or =
its attachments. If you have received this e-mail in error, please =
notify the sender immediately by return e-mail and delete all copies of =
this message and any attachments.

Thank you.

Received on Thu Mar 31 2011 - 12:10:37 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.