[Gretl-users] Re: Dynamic Panel Model
Hi Sven
I was trying to simplify the query by omitting the i index for panel
members as it was not germane to the querywhich was about the form the
forecast equation takes.
I have figured it out and manually chrcked the gretl fitted values to
confirm what has been done.. It would appear that the data are reintegrated
to non differenced form with appropriate collection of terms and estimated
coefficients .
If Gretl in general produced an estimation equation summary format (as per
OLS) for time series models, it would be a useful additional output.
Thanks
Brian
On Sun, 16 Nov 2025, 18:18 Sven Schreiber, <sven.schreiber(a)fu-berlin.de>
wrote:
Am 16.11.2025 um 11:19 schrieb Brian Revell:
Just to clarify my understanding of the generation of fitted values for
levels of Yt
do the estimated coefficients transpose directly into the specified
equation
Yt=a1 Yt-1+ b1Xt + d +ut (which begs the question of recovery of d )
or if the GUI difference estimator is the default estimator , are the
fitted values derived through expansion of the differenced form of
the equation with terms collected such that Yt=(1+a1)Yt-1.....etc.?..
Hi, it's not totally clear to me what the question is. The subject says
dynamic panel, but the equation does not look like a panel model. For
example, no unit/group index i. So please be more specific on what you have
actually estimated, maybe also with some gretl output.
thanks
sven
_______________________________________________
Gretl-users mailing list -- gretl-users(a)gretlml.univpm.it
To unsubscribe send an email to gretl-users-leave(a)gretlml.univpm.it
Website:
https://gretlml.univpm.it/postorius/lists/gretl-users.gretlml.univpm.it/
Attachments:
- attachment.htm (text/html — 3.5 KB)