[Gretl-users] Re: About the PMG package
Am 24.03.2021 um 08:35 schrieb Riccardo (Jack) Lucchetti:
On Wed, 24 Mar 2021, Reynaldo Senra wrote:
> For me It is always uncomfortable to write about this because I
am not
> skillful enough to program and I really acknowledge and appreciate the
> efforts all developers do to make available new packages for all Gretl
> community.
No problem. In fact, bug reports and feature requests such as yours help
us develop the package and make it better. Thank you for that.
Absolutely.
> The issue is that the new version of PMG package apparently
doesn't
> report the short run estimates. I really believe that by adding this,
> the package will be just perfect. On the other hand, the lack of this
> feature is not desirable.
I'm afraid don't quite understand what you mean. Printing out the
individual coefficients for all unit is IMO undesirable. It would be a
lot of output; in most cases, unnecessary. Note that you can access the
full set of coefficients for each unit via the "fullcoeffs" element of
the output bundle. Perhaps we could add an auxiliary function for
pretty-printing them. Sven?
I suspect that maybe it wasn't clear enough that in the latest version
you can obtain those coefficients, so let's see what Reynaldo says about
that.
The question is, what's the use case? In my view it might be to compute
impulse responses (dynamic multipliers) for which you need all those
coefficients, but you don't really need to have them printed.
cheers
sven