The RATS Software Forum

Dear Tom:
I'm trying to get the best lags for my AR(p), and I'm wondering can I get that by using Varlagselect or there is some special program for AR model's lag. Thank you

I'm trying to forecasting my AR(p) and here is my code, qurterly GDP = qgdp, assuming the lag is 3 and I'm doing on-step forecast :

linreg qgdp
# constant qgdp{1 to 3}
uforecast(STATIC) fqgdp 2000:1 2011:4
@uforeerrors qgdp fqgdp

Am I right forecasting the gdp?

Snow wrote:Dear Tom:
I'm trying to get the best lags for my AR(p), and I'm wondering can I get that by using Varlagselect or there is some special program for AR model's lag. Thank you

You can, though there's also an @ARAUTOLAGS procedure that's specifically designed for autoregressions. Note, however, that @ARAUTOLAGS is designed for stationary processes, while @VARLAGSELECT allows non-stationary models.

Snow wrote:I'm trying to forecasting my AR(p) and here is my code, qurterly GDP = qgdp, assuming the lag is 3 and I'm doing on-step forecast :

linreg qgdp
# constant qgdp{1 to 3}
uforecast(STATIC) fqgdp 2000:1 2011:4
@uforeerrors qgdp fqgdp

Am I right forecasting the gdp?

That's correct. One-step ahead forecasts for QGDP.