Hello, BUGS-users:
I'm trying to implement an MCMC to estimate model parameters in a
quasi-likelihood analysis. The BUGS model executes and provides believable
parameter estimates for various proper densities for the observations, so I
think the model itself is sound. I want to replace the surrogate density
for the observations with a quasi-likelihood. The model is estimating 13
parameters on only 42 observations.
I've tried the zeros trick used for specifying an arbitrary *density* and
this leads to unsatisfactory results (unreasonable parameter estimates) and
sometimes trap errors, although I can avoid the traps by using more
restrictive priors.
I suspect the trouble is because this quasi-likelihood does not correspond
to a legitimate likelihood function and that this analysis may not be
possible. However, I'm hoping that there is a way to do this, since
estimating equations exist for maxmium quasilikelihood problems (see, e.g.
McCullagh and Nelder Ch. 9).
I would be very appreciative to hear of any shared QL experiences with BUGS.
Thanks in advance.
Charles Annis, P.E.
[log in to unmask]
phone: 561-352-9699
eFAX: 503-217-5849
http://www.StatisticalEngineering.com
-------------------------------------------------------------------
This list is for discussion of modelling issues and the BUGS software.
For help with crashes and error messages, first mail [log in to unmask]
To mail the BUGS list, mail to [log in to unmask]
Before mailing, please check the archive at www.jiscmail.ac.uk/lists/bugs.html
Please do not mail attachments to the list.
To leave the BUGS list, send LEAVE BUGS to [log in to unmask]
If this fails, mail [log in to unmask], NOT the whole list
|