I've experienced the following issue  trying to augment a time-series dataset within a function. Even though adding some observations initially works, after restricting the sample and calling "smpl full" again the additional observations get lost. This happens with latest git on ubuntu: <hansl> open denmark.gdt -q function void foo (series s)     eval $t2        # T=55     dataset addobs 10     eval $t2        # T=65     smpl 1974:2 1974:3     smpl full     eval $t2        # T=55 end function foo(LRM) </hansl> Is this is really intended? If yo, why?

Am 27.04.2019 um 13:03 schrieb Artur Tarassow: > I still face this issue. Any idea or comment on this? > Am 24.04.19 um 19:32 schrieb Artur Tarassow: >>      eval $t2        # T=65 >>      smpl 1974:2 1974:3 >>      smpl full >>      eval $t2        # T=55 >> end function >> Is this is really intended? If yo, why? Looks like a bug to me. But have you tried anything else than $t2 to verify the length?

On Sat, 27 Apr 2019, Sven Schreiber wrote: > Am 27.04.2019 um 13:03 schrieb Artur Tarassow: > >> I still face this issue. Any idea or comment on this? > >> Am 24.04.19 um 19:32 schrieb Artur Tarassow: > >>>      eval $t2        # T=65 >>>      smpl 1974:2 1974:3 >>>      smpl full >>>      eval $t2        # T=55 >>> end function > >>> Is this is really intended? If yo, why? > > Looks like a bug to me. But have you tried anything else than $t2 to > verify the length? Hmm, extending the dataset within a function is a "special concession", which doesn't survive after the function returns. The amendment to the C code that permits this is marked "experimental, 2015-07-28". However, if we're allowing this at all then I guess $t2 should adjust accordingly and "smpl full" should work as one might expect. (Though the usual effect of "smpl full" inside a function is to return to whatever sample range was in force at caller level.) That's now done it git.