Hi to all,
Folks I'm a little rusty...
What's now being meant by the 'roots' of an ARIMA model, please?
 
Best wishes
Richard Hudson
 
Dr RJF Hudson Qld Australia
rjfhud@powerup.com.au
----- Original Message -----
From: Sven Schreiber
To: gretl-users@lists.wfu.edu
Sent: Thursday, February 02, 2012 5:31 AM
Subject: Re: [Gretl-users] Retrieving the roots of an ARIMA

On 01/31/2012 09:13 PM, Allin Cottrell wrote:
> On Tue, 31 Jan 2012, Henrique Andrade wrote:
>
>> Is there any way to retrieve the roots of an ARIMA model? Is there an
>> accessor that allows this kind of operation?
>
> They are attached to the model in question under the name
> "roots". While there is no "$roots" accessor as such, it's now
> possible to retrieve them (in CVS and snapshots) via the
> special form $["roots"]. This form will work for any named
> data attached to the last model in the form of a scalar,
> series or matrix. Example:
>
> open data9-7
> arma 1 2 ; QNC
> matrix r = $["roots"]
> r
>
> This form treats the last model as if it were a "bundle", from
> which any suitable named data can be extracted.
>

Ho ho -- hold it :-)

Seriously, isn't this new syntax being introduced a little hastily? So
far everything that could be accessed got its $-accessor. This $[] thing
seems to be all new. I'm not saying it's bad, but it seems to exist in
parallel to the old-style accessor. After all, all existing accessors
could now be re-defined, for example $coeff -> $["coeff"]. Maybe that's
the way to go (old-style accessors could be kept as syntactic sugar for
backward-compatibility).

cheers,
sven

_______________________________________________
Gretl-users mailing list
Gretl-users@lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users