Hi Ahreum,

Error messages in WinBugs are terrible to interpret. I believe that this message is indicating that the program is unable to perform the calculations of the model due to illogical or extreme values selected by the sampler.

The first thing I would do is calculate values wherever you use a function. For example calculate the logit transformation by hand instead of using the logit function. I have fixed this exact error by making this change once.

Secondly, I would carefully go over your priors to make sure that they don't allow parameter values that could cause an error in calculation. For example,

d~dnorm(0,1)

c

Dear Dan Gwinn

Thank you for your answer. It solves my problem too.