[Gretl-devel] Re: 'omit const' doesn't work
On Mon, 30 Sep 2019, Sven Schreiber wrote:
Am 30.09.2019 um 03:21 schrieb Allin Cottrell:
> OK, but these are just use cases for determining the statistical
> significance of a single (unique) regressor. I agree that task is not
> that uncommon, but it's easily accomplished without use of "omit".
>
> If you take a look at the help for omit it's pretty clear (I think) that
> it's not intended for that purpose: "It calculates a Wald test for the
> joint significance of the variables in varlist, which should be a subset
> of the independent variables in the model last estimated." The subset
> can have just one member, but it's expected to be a proper subset. And
> "By default the restricted model is estimated and it replaces the
> original as the 'current model'".
To be honest I don't find that totally clear (I thought subset was meant
in contrast to superset), but it's no big deal, and I admit that I
haven't checked back on the doc for 'omit'.
However, I guess at least the error message could be adjusted, it
currently says "not available for this estimator".
Yes, that's not an apt message.
But anyway, second thoughts on my part: since "omit" is the most
concise command for checking the significance of a regressor or set
of regressors, we might as well support the limiting case where all
regressors are omitted. (Previously "omit" would fail, not only for
a sole regressor, but also where the entire set of k > 1 regressors
is omitted.)
So that's now in git, with a slight revision of the help for "omit"
to match.
Allin