Re: [Gretl-users] ARIMA model with exogenous variable
On Wed, 1 Jul 2009 Didier.Monselesan(a)csiro.au wrote:
I have estimated the follwing models
arima 2 1 0; GMSL_CW const GISS_GSST
arima 2 0 0; d_GMSL_CW const GISS_GSST
As expected the estimated model parameters are identical in both
cases... However, I could not explain how GRETL computes the
fitted values and residuals for the levels (not the differences)
in model 1). I was expecting the fitted values of model 1) to be
the cumulated sums of fitted values of model 2)...
I'm attaching a script giving a version of the demonstration in
the Gretl User's Guide (Example 20.1 on ARIMA forecasting). I have
modified the example to resemble more closely the model you're
estimating, and have added a comparison of fitted values (as
accessed via $yhat) alongside the comparison of results from
the "fcast" command.
Hopefully this should show you what's going on.
Allin Cottrell
Attachments:
- armaxfit.inp (text/plain — 917 bytes)