NONMEM Users Network Archive

Hosted by Cognigen

please unsubscibe

From: Fran Fernandez Campos <mran82>
Date: Wed, 1 Jun 2011 08:10:19 +0000

Thank you

Subject: [NMusers] please unsubscibe
Date: Tue, 31 May 2011 16:19:45 -0400
From: Rob

Thank you,


From: owner-nmusers
 Behalf Of Eleveld, DJ
Sent: Tuesday, May 31, 2011 2:48 PM
To: ???; nmusers
Subject: RE: [NMusers] question about shinkage
Hi Li,

Well, do you have rich data and a small number of subjects?

How much shrinkage exactly? A very small negative number might just be due =
to (hopefully) non-important numerical issues. It could also be due to ear=
ly termination of the estimation, not doing enough iterations, problems=
 with rounding errors, etc.

The use of shrinkage to diagnose model problems isnt powerful enough to try=
 to solve a problem without knowing anything about the model or the data. S=
o, it depends on your problem. What you usually encounter is that high s=
hrinkage means that a dataset is not informative enough to estimate a param=
ater in the individuals. I would interpret negative shrinkage as meaning th=
at something went wrong with the estimation. In that case you cant trust t=
he resulting estimations (or shrinkage for that matter) to be meaningful an=

You might want to look a constructing likelihood profiles for you model est=
imations as well. I find they work nicely in conjuction with considering s=

Douglas Eleveld

-----Original Message-----
From: owner-nmusers
Sent: Tue 5/31/2011 4:33 PM
To: nmusers
Subject: [NMusers] question about shinkage

Hi dear all,

I have a question about shinkage. I read an article about shinkage (Radojka
M. Savic and Mats O. Karlsson Importance of Shrinkage in Empirical Bayes
Estimates for Diagnostics: Problems and Solutions 2009) and try to use
shinkage to diagnose my model. An ETA-shinkage is negative in my result.
According to the article, negative shinkage may occur in the situation wh=
a parameter variance is fixed to a lower value than the true value or in
rich data from a small number of subjects. I wonder that if the parameter
variance is fixed, shinkage is 100% in my result. And if it is the data
problem, why is the shinkage of this kind of data negative? Besides, I
wonder that whether the negative shinkage indicate the model
misspecification? How important is shinkage to diagnose a model? Is it more
used to evaluate the relationship between the covariate and parameters or t=
choose a model?

Li Mengyao

De inhoud van dit bericht is vertrouwelijk en alleen bestemd voor de geadre=
sseerde(n). Anderen dan de geadresseerde(n) mogen geen gebruik maken van di=
t bericht, het niet openbaar maken of op enige wijze verspreiden of verme=
nigvuldigen. Het UMCG kan niet aansprakelijk gesteld worden voor een incomp=
lete aankomst of vertraging van dit verzonden bericht.

The contents of this message are confidential and only intended for the eye=
s of the addressee(s). Others than the addressee(s) are not allowed to use =
this message, to make it public or to distribute or multiply this message=
 in any way. The UMCG cannot be held responsible for incomplete reception o=
r delay of this transferred message. =

Received on Wed Jun 01 2011 - 04:10:19 EDT

The NONMEM Users Network is maintained by ICON plc. Requests to subscribe to the network should be sent to:

Once subscribed, you may contribute to the discussion by emailing: