Am 29.12.2020 um 09:29 schrieb Artur Bala:
Ok, this is not such a great topic but I'm going to submit it anyway...
In an estimation output, gretl uses the expression "using
observations...", for example:
and here "using observations..." loses its meaning given
that the actual
observations used in the estimation are not in a row anymore.
I think it's a valid point. If the obs are non-continguous, I guess the
hyphen should be dropped (giving the false impression of a range) and
simply the number of obs given. I'm less sure whether more information
on the precise subsampling should be given in the header, it might
become too large.