4:29 p.m.
Hi,
I just answered a gretl-related question about the variance inflation
factors on Stackexchange
(https://stats.stackexchange.com/questions/298612/why-is-variance-inflatio...).
When looking at the 'vif' command basically for the first time ever, my
first reaction was "why is this a command"? It would seem more natural
to call it with "modtest --vif" or something like that. (Although I
admit it wouldn't provide any p-values.)
Related, shouldn't there be a scripting way to access the output of vif
without doing the calculations manually?
cheers,
sven
2:09 a.m.
On Sat, 2 Sep 2017, Sven Schreiber wrote:
Hi,
I just answered a gretl-related question about the variance inflation factors
on Stackexchange
(https://stats.stackexchange.com/questions/298612/why-is-variance-inflatio...).
When looking at the 'vif' command basically for the first time ever, my first
reaction was "why is this a command"? It would seem more natural to call it
with "modtest --vif" or something like that. (Although I admit it wouldn't
provide any p-values.)
Yes, it's not a "test" in the strict sense.
Related, shouldn't there be a scripting way to access the output
of vif
without doing the calculations manually?
No doubt there should. But I'm now in the thick of a big teaching
semester and I won't be able to do anything about it soon.
Registering a feature request might be a good idea.
Allin
10:58 a.m.
Am 06.09.2017 um 04:09 schrieb Allin Cottrell:
On Sat, 2 Sep 2017, Sven Schreiber wrote:
> Related, shouldn't there be a scripting way to access the
output of
> vif without doing the calculations manually?
No doubt there should. But I'm now in the thick of a big teaching
semester and I won't be able to do anything about it soon. Registering a
feature request might be a good idea.
Right, I will do that. But first let me consolidate a few things here;
in an off-list email from last your we had this discussion about the
fact that with 'vif' you also get a Belsley-Kuh-Welch printout, and that
there's Lee's bkw package:
"
... the bkw.gfn package now of course provides a solution, but why
have that code duplication? So would it be possible to add a hansl
interface to that piece of built-in code?
Yes, that's something we should do. Right now Jack and I are working on
the Nelder-Mead optimizer, but I'll put it on my "to do" list.
"
Also let me remind everybody that there's a recent native cnumber()
function which is somewhat related.
Maybe some of these interfaces could be consolidated (with extra options
or stuff like that).
cheers,
sven
9:35 a.m.
Am 06.09.2017 um 12:58 schrieb Sven Schreiber:
Am 06.09.2017 um 04:09 schrieb Allin Cottrell:
> semester and I won't be able to do anything about it soon.
Registering
> a feature request might be a good idea.
This is now request #101,
https://sourceforge.net/p/gretl/feature-requests/101/
thanks,
sven
2663
days inactive
2673
days old
3 comments
2 participants
participants (2)
-
Allin Cottrell -
Sven Schreiber