Dear List,
I have a linear mixed model of the form
Y[i,j] ~ dnorm(mu[i,j],tau[i])
mu[i,j] <- alpha[i] + beta[i]*t[j]
where Y[i,j] is the observation in individual i at time t[j]
I model alpha[i] as a fixed effect with a 'vague' normal prior...
alpha[i] ~ dnorm(0,0.001)
I model the growth rates and the log variance as normally distributed
random effects...
At the moment I'm using the following
beta[i] ~ dnorm(mu.beta,tau.beta)
mu.beta ~ dnorm(0,0.001)
tau.beta ~ dgamma(0.001,0.001)
tau[i] <- 1/(exp(log.var[i]))
log.var[i] ~ dnorm(mu.log.var,tau.log.var)
mu.log.var ~ dnorm(0,0.001)
tau.log.var ~ dgamma(0.001,0.001)
Generally, this model seems to work fine. However, I've noticed that for
some datasets, the variance in beta seems too small, and that this may be a
consequence of the gamma prior for tau.beta. I'd like to implement
alternative priors that are more reference-like in nature, such as the
uniform shrinkage prior. I'd be grateful on information on the appropriate
syntax in BUGS.
TIA
Simon
Simon D.W. Frost, M.A., D.Phil.
Department of Pathology
University of California, San Diego
UCSD Treatment Center
150 W. Washington St., Suite 100
San Diego, CA 92103
USA
Tel: +1 619 543 8080 x275
Fax: +1 619 298 0177
Email: [log in to unmask]
WWW: http://www.simonfrost.com
-------------------------------------------------------------------
To mail the BUGS list, mail to [log in to unmask]
You can search old messages at www.jiscmail.ac.uk/lists/bugs.html
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
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]
|