A Question for the Setting of Priors

[Butree]

Hi, guys,

I'm currently working as a research assistant in a project, being asked to estimate a model using traditional Bayesian approach. When I try to write the prior for the autoregressive compoment of the cost-push shock, I meet some difficulties- I was asked to set 0 as the prior mean(or some number close enough to 0), and in most cases the priors for autoregressive components would be set to beta distribution. but when I set it up as beta_pdf,0,0.05, it shouldn't work because of the properties of beta distribution. So I wanna ask whether there's some alternative distribution , allowing for 0 as an option?

Thanks and best of luck.

[jpfeifer]

Either use a modified beta from -1 to 1 as documented in the manual or use a uniform distribution.