[R-sig-eco] Bayesian analysis with MCMC

[Bob O'Hara]

On 01/18/2014 04:39 AM, 明 wrote:
> Dear all
>       I have a model with four paramters, I want to estimate the parameter uncertainty, so Bayesian analysis with MCMC method is applied.
>       But every sigle mcmc chain seems give quite different parameter marginal distributions.
>       In order to get the true parameter marginal distributions, I do like this:
>       (1) I take 100 MCMC chain, and each chain has 10000 iterations.
>       (2) caculate the different parameter marginal distributions according to the frequence of paramter in step 1 sampling.
>       The result seems reasonable. but is it right?
>       Looking forward for your reply, Thanks in advance
>    
> Han Ming
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-ecology mailing list
> R-sig-ecology at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
Have you checked for convergence? If the chains are giving different 
marginal distributions, this suggests they haven't converged, so they 
are not sampling from the correct posterior distribution (or, at least, 
some chains aren't).

Bob

-- 
Bob O'Hara

Biodiversity and Climate Research Centre
Senckenberganlage 25
D-60325 Frankfurt am Main,
Germany

Tel: +49 69 7542 1863
Mobile: +49 1515 888 5440
WWW:   http://www.bik-f.de/root/index.php?page_id=219
Blog: http://blogs.nature.com/boboh
Journal of Negative Results - EEB: www.jnr-eeb.org