Hi, everyone,

recently I am doing a project, and with the binary time series data, I plan to fit the dynamic model with Winbugs,

y(t) ~ dbern(p(t))
logit(p(t))=epsilon(t)+gamma(t)
epsilon(t)=epsilon(t-1)+omega(t),omega(t) ~ Normal(0,tau1)
gamma(t)=-gamma(t-1)-...-gamma(t-12)+delta(t),delta(t) ~ Normal(0,tau2)

this is a very standard generalized dynamic model, but there are some convergence problem when I fit it in winbugs.
Does anyone fit such a model in winbugs before? if does, would you like to give me the sample code for study?
Many thanks.

Shawn