Dear Bugs Users,
In conclusion, if one has some questionable results on his posterior
covariance terms (specially if there is
not much data supply), I would advise the following procedure :
If teta follows a Wishart distribution,
1. Fit the model taking every teta[i] ~ dnorm(0,tau[i]), tau[i] ~
dgamma(0.001,0.001).
2. Fit teta ~ dmnorm(mteta[],T[,]) with T ~ dwish(R,p+2) where the R
diagonal terms are the estimations of the 1/tau[i]
at the first step, 0 elsewhere.
As David Spiegelhalter says, "this is a little bit cheating as you are
using the data to construct
the prior, but as your df is minimal it should not be a problem".
You may compare the results with the ones obtain taking a Wishart prior
with p df.
Many thanks to those who wrote to me with suggestions and comments.
Etienne
Etienne Le Bihan
ACTA Informatique
149, rue de Bercy
75595 Paris Cedex 12
Tel : 01.40.04.50.25
mailto:[log in to unmask]
-------------------------------------------------------------------
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
|