Hi,

It would be much appreciated if someone could explain how to draw/derive the credible set of a variable given that the model is estimated. Thank you.

What do you mean? Dynare allows you to estimate HPDIs.

Thanks, Professor.

My problem might not be clear. I wanted to draw credible bands of a variable (time series) derived from my model. But I could not locate the code to do it with reference to the manual.

Please define the concept

Do you the output from the `smoother`

-option of `estimation`

?

Yes,

I have estimated the parameters (Bayesian) and the model.

estimated_params;

ALPH, beta_pdf,0.35,0.01; so on

estimation(datafile=data_model1_bayesian_estimation, graph_format=pdf, mh_jscale=0.6, mode_compute=6, mode_check, prior_trunc=0, mh_replic=100000,

mh_nblocks=4, raftery_lewis_diagnostics, consider_all_endogenous, conf_sig=0.95);

I obtained the results from the oo_.SmoothedVariables.

Result looks like below,

Year Result

1971 0.335605029511308

1972 0.284030009014477

1973 0.356588867355969

1974 0.345517915172426

1975 0.428387096972830

Now, I want to know how to get the credible bands of this variable.

Thank you.

You did not specify the `smoother`

-option, which should give you `oo_.SmoothedVariables.HPDinf`

and `oo_.SmoothedVariables.HPDsup`

.

Dear Professor, Thanks a lot. I got it done!