R help - Aug 2005 - glmmPQL error in logLik.reStruct

Dear R users,

I'm attempting to fit a GLM with random effects using the tweedie family
for the error structure.  I'm getting the error:

iteration 1 
Error in logLik.reStruct(object, conLin) : 
        NA/NaN/Inf in foreign function call (arg 3)

I'm running V2.1.0
I notice from searching the lists that the same error was reported in
May 2004 by Spencer Graves, but no-one was able to offer any
suggestions.  Running the sample code that he provided also still
produces the same error:

set.seed(8) 
N. <- 1000 
z <- rnorm(N.) 
pr.good <- exp(-1e-4*exp(2+2*z)) 
quantile(pr.good) 
DF. <- data.frame(yield=rbinom(N., N., pr.good)/N., 
       Offset=rep(-10, N.), nest=1:N.) 
fit <- glmmPQL(fixed=1-yield~offset(Offset), random=~1|nest, 
       family=binomial(link="cloglog"), 
       data=DF., weights=rep(N., N.)) 

A year on can anyone suggest what the problem might be?  Or in
particular, Spencer - did you come up with a solution?

Thanks in advance, 

Kechi Nzerem

Dear all,

As an update to my previous post, for anyone who is interested: Someone
has kindly told me that this error does not occur in the 2.2 pre-release
version.  I've run my particular model with my data and this problem is
solved in the 2.2 pre-release.

Cheers,

Kechi Nzerem
>-----Original Message-----
>From: Kechi Nzerem
>Sent: 03 August 2005 14:12
>To: 'r-help at lists.R-project.org'
>Subject: glmmPQL error in logLik.reStruct
>
>Dear R users,
>
>I'm attempting to fit a GLM with random effects using the tweedie
family
>for the error structure.  I'm getting the error:
>
>iteration 1
>Error in logLik.reStruct(object, conLin) :
>        NA/NaN/Inf in foreign function call (arg 3)
>
>I'm running V2.1.0
>I notice from searching the lists that the same error was reported in
May
>2004 by Spencer Graves, but no-one was able to offer any suggestions.
>Running the sample code that he provided also still produces the same
>error:
>
>set.seed(8)
>N. <- 1000
>z <- rnorm(N.)
>pr.good <- exp(-1e-4*exp(2+2*z))
>quantile(pr.good)
>DF. <- data.frame(yield=rbinom(N., N., pr.good)/N.,
>       Offset=rep(-10, N.), nest=1:N.)
>fit <- glmmPQL(fixed=1-yield~offset(Offset), random=~1|nest,
>       family=binomial(link="cloglog"),
>       data=DF., weights=rep(N., N.))
>
>A year on can anyone suggest what the problem might be?  Or in
particular,
>Spencer - did you come up with a solution?
>
>Thanks in advance,
>
>Kechi Nzerem