```Dear All,

I am wondering if WinBUGS is the one that I can use to do the very basic
Bayesian updating? I mean something like "if the prior distribution of a
parameter is gamma, and new data has poisson distribution, then the
posterior distribution of the parameter is another gamma". However,in all
of the examples of WinBUGS, people are trying to model the relationship
between a parameter theta(which is a parameter in the distribution of a
variable y) and other variables  x. In my case, I just want to update the
parameter without complex modelling. In the following model I have passed
the "check model" but when I load the data, it shows "expected variable
name".
Context: here y is a variable called "trip distance per person per
day".Suppose I have old data yi, from which Crystal Ball gives a good
fitting distribution of beta(alpha=5.06,beta=18.59).Based on it,I assume
the prior distributions of alpha is normal(5.06,0.001) and
beta~normal(18.59,0.001). Now I have some new data yi(new),which I suppose
comes from a population of another beta distribution (i.e.,the function to
build likelihood).  I want to know what the updated distributions of alpha
and beta are. Can I use the means of updated alpha and beta to get the
updated distribution of variable y.

Your help to me is highly appreciated.

Regards,
Yongping Zhang

model
{
for (i in 1:N) {
y[i] ~ dbeta(alpha, beta)
}
alpha ~ dnorm(5.06,0.001)
beta ~ dnorm(18.59,0.001)
}

list
(y=c(1.00,1.50,8.50,3.50,8.67,6.00,4.00,4.00,4.00,7.00,8.00,4.25,8.33),
N=13)#Data

list (alpha=5, beta=18) # intials

-------------------------------------------------------------------
This list is for discussion of modelling issues and the BUGS software.