R help - Mar 2011 - a question on R optimization functions

Dear All,

I use nlminb or optim for maximizing likelihood functions. Sometimes,
the parameter values happen to be NA, then the program will hang there
and iterate forever without stopping. No error message will be
produced. So I can not use error catch method such as "try". Are there
any suggestions how I can circumvent this problem? Maybe I can time a
function, and if the time exceeds a threshold, it will be stopped. I
am not sure if this is feasible in R. As a note, I am running
simulations with thousands of replicates, so there needs to be a
systematic way of doing this. Are there any suggestions on how to do
this? Thank you very much!

Regards,
Dajiang

Dajiang J. Liu <dajiang.liu <at> gmail.com> writes:
> I use nlminb or optim for maximizing likelihood functions. Sometimes,
> the parameter values happen to be NA, then the program will hang there
> and iterate forever without stopping. No error message will be
> produced. So I can not use error catch method such as "try". Are
there
> any suggestions how I can circumvent this problem? Maybe I can time a
> function, and if the time exceeds a threshold, it will be stopped. I
> am not sure if this is feasible in R. As a note, I am running
> simulations with thousands of replicates, so there needs to be a
> systematic way of doing this. Are there any suggestions on how to do
> this? Thank you very much!
   It would be very surprising (I suppose not impossible, but
*very* surprising) if this behavior were generated by R's optimization
functions; it would be much more likely for it to be an infinite
loop/hang somewhere inside your objective function.

  See ?setTimeLimit (which will fail if the problem is inside
C code that you have written ...)

  Ben Bolker

It seems more likely that the return value from your function is NA or NaN or
Inf. This might then result in an NA parameter value being calculated for the
next step. This is possible, for example,  because the line search extends
outside the feasible region. You can re-write your function to check for that
case and return a large negative value (or positive if minimizing), or stop() if
that is more appropriate.

Paul
> -----Original Message-----
> From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-
> project.org] On Behalf Of Dajiang J. Liu
> Sent: March 25, 2011 3:59 AM
> To: r-help at r-project.org
> Subject: [R] a question on R optimization functions
> 
> Dear All,
> 
> I use nlminb or optim for maximizing likelihood functions. Sometimes,
> the parameter values happen to be NA, then the program will hang there
> and iterate forever without stopping. No error message will be
> produced. So I can not use error catch method such as "try". Are
there
> any suggestions how I can circumvent this problem? Maybe I can time a
> function, and if the time exceeds a threshold, it will be stopped. I
> am not sure if this is feasible in R. As a note, I am running
> simulations with thousands of replicates, so there needs to be a
> systematic way of doing this. Are there any suggestions on how to do
> this? Thank you very much!
> 
> Regards,
> Dajiang
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-
> guide.html
> and provide commented, minimal, self-contained, reproducible code.
===================================================================================
La version fran?aise suit le texte anglais.

------------------------------------------------------------------------------------

This email may contain privileged and/or confidential information, and the Bank
of
Canada does not waive any related rights. Any distribution, use, or copying of
this
email or the information it contains by other than the intended recipient is
unauthorized. If you received this email in error please delete it immediately
from
your system and notify the sender promptly by email that you have done so. 

------------------------------------------------------------------------------------

Le pr?sent courriel peut contenir de l'information privil?gi?e ou
confidentielle.
La Banque du Canada ne renonce pas aux droits qui s'y rapportent. Toute
diffusion,
utilisation ou copie de ce courriel ou des renseignements qu'il contient par
une
personne autre que le ou les destinataires d?sign?s est interdite. Si vous
recevez
ce courriel par erreur, veuillez le supprimer imm?diatement et envoyer sans
d?lai ?
l'exp?diteur un message ?lectronique pour l'aviser que vous avez ?limin?
de votre
ordinateur toute copie du courriel re?u.