William Link writes:
> Suppose A = A[1], A[2], ..., A[k] is a factor, and one wants to fix
> certain levels (such as A[1] = 0, for identifiability), while letting
> other values be stochastic.
>
> How can one do that in BUGS?
[...]
How about creating a hierarchical model letting the A[k]'s come from a
common distribtution, say a normal:
A[k] ~ dnorm(mu.A,tau.A)
and then assign non-informative (diffuse) normal and gamma priors to
mu.A and tau.A? This may be a more sensible model (?) and would also
improve the estimates of each A[k]...
Jarle
--
Jarle Tufto, [log in to unmask]
ICAPB, University of Edinburgh, www.ed.ac.uk/~jarlet
Kings' Buildings, West Mains Road, Phone: +44-131-650-5508
Edinburgh EH9 3JT, Scotland, UK Fax: +44-131-667-3210
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|