On Fri, 15 Oct 2021, Sven Schreiber wrote:
Am 14.10.2021 um 19:53 schrieb Sven Schreiber:
> Am 14.10.2021 um 18:14 schrieb Allin Cottrell:
>>
>> That's now upgraded in git: given a time-series dataset, recognized
>> lag series are extended by "dataset addobs". That's not yet the
case
>> for panel data.
>>
> Wow, that was quick! I expected it to be more involved and complex to
> change that handling...
>
> Will test next week at the latest.
>
I already could test just now (on Linux), and it seems to work nicely! I
also tested with a dynamic forecast stemming from an autoregressive
element in addition to the lagged exogenous, and also by modifying an
existing model and having gretl do the resetting of the sample, without
problems.
Thanks for closing this gap, Allin.
Thanks for testing. It turned that getting "dataset addobs" to
realize that it could extend lag series was quite simple. The
trickier thing -- now done, I think -- was to get the GUI mechanism
to put up the right message regarding adding extra observations.
There are two relevant messages. One assumes that adding
observations will extend the possible forecast range, asks how many
you want to add, and does it for you if you ask for more than 0. The
other is non-committal, just saying that adding observations is a
possibility and indicating the menu item to do that. I think the
choice between these should now be more fitting to the context.
Allin