Regression Using Group of Lagged Series

[kszynkar]

Hi,

I have a series called "wip" and a group called "x". Is it possible to do a regression using the set of series in x as independent variables and wip as the dependent variable?

Thanks

[EViews Glenn]

If I'm understanding the question correctly

equation eq1.ls whip c x

should do the trick. If you are doing this from the dialog, just enter the dependent WHIP, followed by the regressors (I've included the constant C).

[kszynkar]

Ah, indeed it does. I'm actually trying to do:

equation wt_eq_s.ls @pc(wip) @pc(x)

which doesn't work. Without the "@pc" function it works.

[EViews Gareth]

Unfortunately you can not use functions of groups. Only functions of original series.

You can write a simple program that will make a group containing the functions of the series:

Code: Select all

group pcx for !i=1 to x.@count pcx.add @pc(x(!i)) next

Then use pcx as your regressor group.

[kszynkar]

that works. Thanks Eviews Gareth