From: Sven Schreiber <svetosch@gmx.net>
To: Gretl list <gretl-users@lists.wfu.edu>
Sent: Tue, December 8, 2009 11:25:31 AM
Subject: Re: [Gretl-users] using lags
Raz Lev schrieb:
> hi again,
>
> I am working on a script for a gmm estimation.
>
> some of my variables are used in lags in some of the equations of the
> model I am trying to estimate.
> When I try running the script I receive an error saying that missing
> values were encountered. I have checked (using ok) that the input series
> have no missing values in them.
>
> I therefore suspect this might be happening when trying to use the lag
> for the first observation in the series.
> Could this be the reason or does gretl 'know' how to handle this (skip
> the first observation maybe?)? If it could be the cause, any suggestions
> on handling it would be much appreciated.
I think you're right, so you would have to reduce the sample
accordingly. To verify that the assumption is correct, I suggest to do
this by hand. Later you could add something
like
smpl --no-missing inputvar(-3)
or maybe
smpl +3 0
#(according to the help, this may not work, but I think it would be nice
if it did)
(and revert back to the old sample with 'smpl full')
See the built-in command help for smpl.
please report back if and what worked
cheers,
sven
_______________________________________________
Gretl-users mailing list
Gretl-users@lists.wfu.eduhttp://lists.wfu.edu/mailman/listinfo/gretl-users