On Thu, May 2, 2024 at 7:37 AM Riccardo (Jack) Lucchetti
<p002...@staff.univpm.it> wrote:
>
> On 02/05/2024 13:16, Sven Schreiber wrote:
> > Am 02.05.2024 um 12:53 schrieb Riccardo (Jack) Lucchetti:
> >> On 02/05/2024 11:57, Sven Schreiber wrote:
> >>
> >>> Well, I thought so, too. However, even inside a function there seems
> >>> to be a problem. Test case: [...]
> >>
> >> I managed to track down the problem somehow. It looks as if operator
> >> precedence might be the issue:
> > Well, that's what Rigoberto said at the beginning, didn't he ? :-)
>
> I've come to think that this is indeed the case: I had misunderstood the
> actual meaning of Rigoberto's message.

Yes. The parser is getting confused by the <list>.<series_id>(<lag>)
formulation and is in effect parenthesizing

L.$i / L.$i(-1) - 1 # should be implicitly (L.$i / L.$i(-1)) - 1

as if it were

L.$i / (L.$i(-1) - 1)

In general, of course, '/' has higher precedence than '-' but it looks
like the right paren in "(-1)" is being taken as closing the division
expression.

Allin
_______________________________________________
Gretl-users mailing list -- gretl-users@gretlml.univpm.it
To unsubscribe send an email to gretl-users-le...@gretlml.univpm.it
Website: 
https://gretlml.univpm.it/postorius/lists/gretl-users.gretlml.univpm.it/

Reply via email to