Page 2 of 2

Re: COMPUTABLE GENERAL EQUILIBRIUM

Posted: Wed Mar 30, 2016 12:36 pm
by EViews Gareth
It appears as though the forecast series is identical to the actual series, other than those observations for which the actual data are not observed.

If the forecasts are identical to the actual data, all evaluation statistics will be equal to zero.

Re: COMPUTABLE GENERAL EQUILIBRIUM

Posted: Thu Mar 31, 2016 1:52 am
by piorember1990
So what didn't I do correctly?
I created the workfile, defined the date, opened the series, went to proc and selected automatic ARIMA forecasting. I defined both the estimation sample and the forecast length. I moved to options and selected forecast comparison graph.
After which the forecast series was created. I opened it and went to view and selected forecast evaluation.
Please where exactly did I get it wrong sir.
Thank you.

Re: COMPUTABLE GENERAL EQUILIBRIUM

Posted: Thu Mar 31, 2016 3:41 am
by EViews Gareth
It is hard to say what you did wrong without knowing exactly what you're trying to do.

You need to think about what exactly your goal is - and over which observations. Forecast evaluation can only occur over observation for which you have actual data and forecast data.