Dear Allstat users,
I would like some advices, if possible, about the use of convergence diagnostic methods for Gibbs Sampling.
Suppose I have got four available methods:
1 Geweke
2 Gelman&Rubin
3 Raftery & Lewis
4 Heidelberger & Welch
Without using all of them, can I simply choose one of them? Is one of them particularly recommended? Which is the criteria of selecting the most appropriate method?
any comment or help will be appreciated
Stefano
|