[R-sig-ME] About lmer function

[Douglas Bates]

On Thu, Apr 1, 2010 at 9:19 AM, Douglas Bates <bates at stat.wisc.edu> wrote:
> I suggest sending queries like this to the
> R-SIG-Mixed-Models at R-project.org mailing list, which I am taking the
> liberty of cc:'ing on this reply.
>
> On Thu, Apr 1, 2010 at 2:32 AM,  <Edgar.Torres at ifremer.fr> wrote:
>> Hello professor Bates,
>>
>> I have had problems with the function lmer in the package lme4, I'm trying
>> to run a GLMM with a poisson distribution. But I receive the warning message
>>
>> Warning message: In mer_finalize(ans) : iteration limit reached without
>> convergence (9)
>>
>> I tried to put more iteration with control=list(maxIter=500), but actually
>> the program justo run 300 times, as the default.
>>
>> the code that I wrote is this:
>>
>> GLMM1 <- lmer(OPTOT ~ fYEAR + (fYEAR | BOAT), data=DATA1, family=poisson,
>> na.action=na.omit, control=list(maxIter=500,msVerbose=T), verbose=T)
>
> Ben Bolker contributed a patch for this that Martin Maechler
> incorporated in a new release of lme4.  The release has been uploaded
> to CRAN but not yet installed (I think the CRAN maintainers are taking
> a well-deserved vacation).

By the way, before trying to increase the maximum number of iterations
I would first use verbose = TRUE in the current call to see where the
iterations are going.

>
>>
>> I am not really sure that the way that I wrote the control statement is the
>> right one.
>>
>> Thanks a lot
>>
>> Edgar TORRES
>> Phd student in Ecosystems
>> Centre de Recherche Halieutique  (IRD)
>> Avenue Jean Monnet BP 171
>> 34203 Sète cedex France
>>
>>
>