Hi George,
this bug is fixed in unstable. Please, tell me if it works for you.
Best
Michel
On 10/12/2011 09:50 PM, George wrote:
Hi Sebastien
running Dynare 4.2.2 with conditional_forecast(periods =40,parameter_set = posterior_mean,controlled_varexo = (em)) I got error
??? Error using ==> eval Reference to non-existent field 'posterior_mean'.
Error in ==> get_posterior_parameters at 46 xparam(m) = eval(['oo_.posterior_' type '.shocks_std.' name1]);
Error in ==> imcforecast at 57 xparam = get_posterior_parameters('mean');
I.e. when checking oo_ I found only oo_.posterior_mode ans = parameters: [1x1 struct] shocks_std: [1x1 struct]
and no .posterior_mean despite it being specified in the model file
This posterior_mean model specification used to work with 4.2.1 but it looks like recent changes made posterior_mode the only option (I have not tried other options but it works with posterior_mode specified in mod file also) .
- what is role of constrained_vars in cond. forecasting, how is it
chosen or set and why is it set constrained_vars_ = 1; in the <model>.m file?
Best regards George
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev