Re: [Gretl-devel] OLS Forecasting
On Sun, 8 Feb 2009, Henrique wrote:
I had tested this new gretl feature using one of my databases. At
first, it worked very fine. But after this initial test I replicate
the estimation (and forecasting) exercise using EViews and I realized
that the results differ.
In gretl I'd selected the option "static forecast" (and the same in
EViews).
After this, I changed the EViews option to "dynamic forecast" and the
results were equal to the gretl "static forecast" predicted values.
Is something wrong?
Yes, the dynamic/static thing is not right yet for auto-integrated
forecasts; I'll work on this some more.
Allin.