Page 1 of 1

Point Forecasting

PostPosted: Tue Oct 11, 2016 5:41 pm
by MelleMa
Hello,

could anybody answer my question, please?

My estimation line is as folows:

estimation(datafile=dataset_variate,mh_replic = 50000,mode_file=version1_mode, mode_compute=6,mh_nblocks= 2,mh_drop=0.2,mh_jscale=0.35,order=1,mode_check,irf=0, forecast=8, graph_format=none)

I supposed to have 8 forecasts for each variables (for T+1, T+2, ..., T+8 where T is the last observation of my dataset), but I have 9 forecasts for each variable.

Why is it so?

Thanks a lot for your attention to my question!
Sorry if the answer exists somewhere in the forum, I was looking for but I found nothing.

Re: Point Forecasting

PostPosted: Tue Oct 11, 2016 7:17 pm
by jpfeifer
Please try the unstable version. In 4.4.3 the Bayesian forecasts stored the initial values as well. See http://www.dynare.org/DynareWiki/KnownBugs