I get upset when software dies and refuses to give me an answer. I'd
much rather have a routine give me a wrong answer -- with an error
message -- than just an error message. Maybe refuse to print standard
errors when the hessian is singular, but at least give me a progress
report
May I interject a comment?
When data is generated from a specified model with reasonable
parameter
values, it should be possible to fit such a model successful,
or is this
me being stupid?
Let me take a turn at being stupid. Why should this be true? That is, why
should it be possible
I agree that the model is not fitting the Lesaffre data well, but my point was
to show that glmmADMB is numerically stable. Numerical
stability is obviously a nice property, but becomes particularly important
when one wants to do parametric bootstrappin, which I think is needed
for these kinds of
Of course it is generally possible to generate datasets for a perfectly
well-defined model that are hard to fit, but in this particular case I
feel it should be possible. In my observations, glmm.admb is far more
numerically stable fitting GLMM's than other software I've seen. Further
, I
Douglas Bates wrote:
The Laplace method in lmer and the default method in glmm.admb,
which according to the documentation is the Laplace approximation,
produce essentially the same model fit. One difference is the
reported value of the log-likelihood, which we should cross-check, and
another
On 12/19/05, Hans Julius Skaug [EMAIL PROTECTED] wrote:
Douglas Bates wrote:
The Laplace method in lmer and the default method in glmm.admb,
which according to the documentation is the Laplace approximation,
produce essentially the same model fit. One difference is the
reported value of
On 12/15/05, Roel de Jong [EMAIL PROTECTED] wrote:
Dear R-users,
because lme(r) glmmpql, which are based on Penalized Quasi Likelihood,
are not very robust with Bernoulli responses,
The current version of lmer takes method = PQL (the default) or
Laplace or AGQ although AGQ is not available
Dear R-users,
because lme(r) glmmpql, which are based on Penalized Quasi Likelihood,
are not very robust with Bernoulli responses, I wanted to test glmmADMB.
I run the following simulation study:
500 samples are drawn with the model specification:
y =
Dear R-users,
Half a year ago we put out the R package glmmADMB for fitting
overdispersed count data.
http://otter-rsch.com/admbre/examples/glmmadmb/glmmADMB.html
Several people who used this package have requested
additional features. We now have a new version ready.
The major new feature is