1:42 p.m.
On Fri, October 13, 2006 15:22, Allin Cottrell wrote:
> I can see the point, although personally I use (or would like to
> use) gretl differently.
I can see a case for having the behaviour that you expected (i.e.
estimating a model via the gui updates the "last model" as seen
from the console). That wouldn't be difficult to do.
Hmm. I don't know. I'm not able to say precisely why, but I'm not comfortable
with the idea that the result of a script may be influenced by something
outside the script itself. Sven, could you give us a little more detail as to
what your plan is?
(cc:ing this to the devel list, because IMO this discussion really belongs there)
Riccardo (Jack) Lucchetti
Dipartimento di Economia
Facoltà di Economia "G. Fuà"
Ancona
1:57 p.m.
New subject: [Gretl-users] accessing $vcv etc.
On Fri, 13 Oct 2006, Riccardo (Jack) Lucchetti wrote:
On Fri, October 13, 2006 15:22, Allin Cottrell wrote:
>> I can see the point, although personally I use (or would like to
>> use) gretl differently.
>
> I can see a case for having the behaviour that you expected (i.e.
> estimating a model via the gui updates the "last model" as seen
> from the console). That wouldn't be difficult to do.
Hmm. I don't know. I'm not able to say precisely why, but I'm
not comfortable with the idea that the result of a script may be
influenced by something outside the script itself.
I agree on that. But I think Sven was talking about something a
bit different from executing a script, namely interactive work at
the console. I believe the question is: should interactive work in
the gui program (by point and click or via console commands) be
treated as a single "thread", or should there be a distinct
"console thread", with its own notion of "last model"?
Personally, I'm not sure about this, but I have tripped over the
issue myself once or twice. E.g. estimate a model in the gui,
open the console, type "genr uh = $uhat", and get an error message
because there's no $uhat available...
Allin.
2:05 p.m.
New subject: [Gretl-users] accessing $vcv etc.
On Fri, October 13, 2006 15:57, Allin Cottrell wrote:
I agree on that. But I think Sven was talking about something a
bit different from executing a script, namely interactive work at
the console. I believe the question is: should interactive work in
the gui program (by point and click or via console commands) be
treated as a single "thread", or should there be a distinct
"console thread", with its own notion of "last model"?
Personally, I'm not sure about this, but I have tripped over the
issue myself once or twice. E.g. estimate a model in the gui,
open the console, type "genr uh = $uhat", and get an error message
because there's no $uhat available...
As far as I'm concerned, I'd rather have the present behaviour than
? ols y x
# do some random stuff in the gui
? genr u = $uhat
# get the wrong thing
My 2 cents go to the "2 threads" options.
Riccardo (Jack) Lucchetti
Dipartimento di Economia
Facoltà di Economia "G. Fuà"
Ancona
2:09 p.m.
New subject: [Gretl-users] accessing $vcv etc.
On Fri, 13 Oct 2006, Riccardo (Jack) Lucchetti wrote:
As far as I'm concerned, I'd rather have the present
behaviour than
? ols y x
# do some random stuff in the gui
? genr u = $uhat
# get the wrong thing
My 2 cents go to the "2 threads" options.
So let's go with the status quo for now. I'll try to remember to
document this behaviour!
Allin.
3:59 p.m.
New subject: [Gretl-users] accessing $vcv etc.
El Viernes, 13 de Octubre de 2006 16:09, Allin Cottrell escribió:
On Fri, 13 Oct 2006, Riccardo (Jack) Lucchetti wrote:
> As far as I'm concerned, I'd rather have the present behaviour than
>
> ? ols y x
> # do some random stuff in the gui
> ? genr u = $uhat
> # get the wrong thing
>
> My 2 cents go to the "2 threads" options.
So let's go with the status quo for now. I'll try to remember to
document this behaviour!
Allin.
Is there a way to recover the coefficients of the ols regression run trough
the GUI?
I tried, in the model window "/Save/Define a new variable" and writing
beta1=$coeff(const) this did not work.
I tried also in a gretl console "genr beta1=$coeff(const)", but we already
knew this is in a different thread and it cannot work, as it was.
--
Ignacio Díaz-Emparanza
Dpto. de Economía Aplicada III (Econometría y Estadística)
UPV-EHU
5:02 p.m.
New subject: [Gretl-users] accessing $vcv etc.
On Mon, 16 Oct 2006, Ignacio Díaz-Emparanza wrote:
Is there a way to recover the coefficients of the ols regression run
trough
the GUI?
I tried, in the model window "/Save/Define a new variable" and writing
beta1=$coeff(const) this did not work.
That should work, and does in current CVS. I think it may have
got broken briefly as a side-effect of some other changes.
Allin.
7:20 p.m.
New subject: [Gretl-users] accessing $vcv etc.
[devel-list only now; just keep me cc-ed please]
Riccardo (Jack) Lucchetti schrieb:
On Fri, October 13, 2006 15:22, Allin Cottrell wrote:
>> I can see the point, although personally I use (or would like to
>> use) gretl differently.
> I can see a case for having the behaviour that you expected (i.e.
> estimating a model via the gui updates the "last model" as seen
> from the console). That wouldn't be difficult to do.
>
Hmm. I don't know. I'm not able to say precisely why, but I'm not
comfortable
with the idea that the result of a script may be influenced by something
outside the script itself. Sven, could you give us a little more detail as to
what your plan is?
First of all, I don't have a plan, and I am not asking for a change. You
should first check with other users what's the preferred way.
To answer your question: Suppose some things are best done
interactively. (I guess you agree, otherwise there wouldn't be much need
for a gui-gretl.) Then there are other things where automatization by
scripts is good, e.g. because gretl doesn't have the corresponding
built-in function. So I would like to freely move back and forth between
manipulating stuff via menus and doing something on the CLI. That's
currently not possible because of the mentioned separation.
That's all, really.
-sven
6401
days inactive
6404
days old
6 comments
4 participants
participants (4)
-
Allin Cottrell -
Ignacio Díaz-Emparanza -
Riccardo (Jack) Lucchetti -
Sven Schreiber