Mh_jscale

hi
the same mh_jscale seems to yield different acceptance ratios in dynare version 3 .065 and version 4. wonder where the adjustment has taken place.

thanks

reuben

Hi Reuben,

Is your problem systematic? The effective acceptance ratio depends on the sequence of proposals. Two different sequences of proposals may lead the mcmc in very different regions of the posterior distribution… In this case the acceptance ratios will be very different. If you think that the problem is systematic, can you post an example?

Best,
Stéphane.