Results from Metropolis-Hastings

1

Hi there!

I got some results that I am attaching, as you can see in rhe MCMC results table 1, the HPD’s are NaN values. Why? I got the following message from Matlab:

Log data density is -Inf.
posterior_moments: There are not enough draws computes to compute HPD Intervals. Skipping their computation.
posterior_moments: There are not enough draws computes to compute deciles. Skipping their computation.

I also do not understand the bounds of the prior information table as well as why the Stdev from the MCMC table and the posterior maximization differs.

Thanks!

I provide the fiiles if anyone wproblem.zip (593.2 KB) ant to replicate the problem.

2

You set mh_replic=1. With only 1 draw, you cannot meaningfully generate a distribution.