Could someone help me to find How I can calculate the probabilities that dynare give us when we use the model_comparison command?
I’ve also noticed that Dynare gives me always the comparison with Laplace’s ikelihood, even when I use the option modifiedharmonicmean. Is it some kind of bug or (most probabily) some mistake that I made?
Thanks!
Fernando
Dear Fernando, I don’t understand your first question.There is nothing to compute here. Dynare reports the Posterior Odds ratios and the posterior probabilty of each model (last column). Regarding your second question, did you run a metropolis for all the model to be compared ? Otherwise Dynare uses the Laplace approximation of the marginal posterior density.
Best, Stéphane.