[R] glmmPQL: Na/NaN/Inf in foreign function call
spencer.graves at pdf.com
Thu Jan 19 04:40:05 CET 2006
I have not seen a reply to this post, and unfortunately, I'm not
smart enough to help you. If you would still like help from this
listserve, please provide a terse, reproducible example that someone can
in seconds copy from your email into R and presumably get the same error
message you report. If you have that, then anyone with interest in
glmmPQL can try to help you. Without that, your email must reach
someone who has gotten that specific error message from glmmPQL with
sufficient regularity that the error message has become solidly recorded
in their brains. Moreover, this person must also have the time and
interest to reply. Don't gamble on a long shot. Play the odds. >
PLEASE do read the posting guide! "www.R-project.org/posting-guide.html".
David Reitter wrote:
> I'm using glmmPQL, and I still have a few problems with it.
> In addition to the issue reported earlier, I'm getting the following
> error and I was wondering if there's something I can do about it.
> Error in logLik.reStruct(object, conLin) : Na/NaN/Inf in foreign
> function call (arg 3)
> ... Warnings:
> 1: Singular precistion matrix in level -1, block 4
> 4: ""
> The interaction terms are
> primed ~ log(dist) * role
> random = ~ dist | target.utt / prime.utt
> The family is binomial (logit).
> I ensured that log(dist) is always [0; 1]. Role is a factor (binary).
> target.utt and prime.utt are categories as well.
> This is version 2.1.1 - is the latest version more reliable?
> Thanks for any help you can give me.
> R-help at stat.math.ethz.ch mailing list
> PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
More information about the R-help