Re: [Gretl-devel] ARMA developments
On Sat, 28 Apr 2018, Sven Schreiber wrote:
Am 27.04.2018 um 23:43 schrieb Allin Cottrell:
> I'm pleased to announce that our code for exact ML estimation of ARMA
> models is now (in git and snapshots) MUCH faster than it used to be (like,
> two orders of magnitude faster in many cases).
> Thanks go to Oleh for pushing us in this direction!
Speed is always cool, of course. But a real-world application example would
be nice where there's a user-noticeable difference. I mean I just tried an
artificial dataset with 9999 obs and current arma felt instantaneous already.
Actually, the new setup has been in git for at least a couple of weeks.
Are people using bootstraps on ARMA specifications or stuff like
that? And/or
simulate some option prices based on an ARMA DGP for example? That would make
a nice comparison with other software.
The best way I can think for making the speed improvement useful would be
adding some kind of "auto" option to the arima command to perform
automatic selection of the arma orders.
-------------------------------------------------------
Riccardo (Jack) Lucchetti
Dipartimento di Scienze Economiche e Sociali (DiSES)
Università Politecnica delle Marche
(formerly known as Università di Ancona)
r.lucchetti(a)univpm.it
http://www2.econ.univpm.it/servizi/hpp/lucchetti
-------------------------------------------------------