4:34 p.m.
good evening
i have a problem with gretl_var.
first with the function ar_model i can specify the lags
FOR EXAMPLE: Yt= Yt-3 + Y t-2 + Yt-6 +CONST but with gretl I CAN'T i
just enter the lag order .
HOW can i specify the lags .
my project work personalized lags please answer me .
thanks for all
6:54 p.m.
Aymen,
If your lags differ in each equation, it's not really a VAR. However you could specify
the equations separately and estimate them as a SUR model. Have a look at the
"system" command.
I hope this helps,
PS
________________________________________
From: gretl-users-bounces(a)lists.wfu.edu [gretl-users-bounces(a)lists.wfu.edu] on behalf of
aymen kaabi [kaabiayman(a)gmail.com]
Sent: Sunday, April 28, 2013 12:34 PM
To: Gretl list
Subject: [Gretl-users] vector autoregressive
good evening
i have a problem with gretl_var.
first with the function ar_model i can specify the lags
FOR EXAMPLE: Yt= Yt-3 + Y t-2 + Yt-6 +CONST but with gretl I CAN'T i just enter
the lag order .
HOW can i specify the lags .
my project work personalized lags please answer me .
thanks for all
7:14 p.m.
I think Aymen means to have the same lags in each equation, but with a
"gappy" structure, for example include the first and third lags, but not
the second (everywhere). Then the VAR-OLS-estimation would still apply.
Vaguely I think there was a discussion about this some time back, but I
don't remember the result. Of course it's still possible to specifiy
this manually as a system and do OLS, but it's not very convenient.
sorry,
sven
Am 28.04.2013 20:54, schrieb Summers, Peter:
Aymen,
If your lags differ in each equation, it's not really a VAR. However
you could specify the equations separately and estimate them as a SUR
model. Have a look at the "system" command.
I hope this helps,
PS ________________________________________ From:
gretl-users-bounces(a)lists.wfu.edu [gretl-users-bounces(a)lists.wfu.edu]
on behalf of aymen kaabi [kaabiayman(a)gmail.com] Sent: Sunday, April
28, 2013 12:34 PM To: Gretl list Subject: [Gretl-users] vector
autoregressive
good evening
i have a problem with gretl_var.
first with the function ar_model i can specify the lags
FOR EXAMPLE: Yt= Yt-3 + Y t-2 + Yt-6 +CONST but with gretl I CAN'T
i just enter the lag order .
HOW can i specify the lags .
my project work personalized lags please answer me .
thanks for all
_______________________________________________ Gretl-users mailing
list Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
5:16 p.m.
thanks sven .
that's exactly what i want .
who understand what i want please describe me how can i do this.
2013/4/28 Sven Schreiber <svetosch(a)gmx.net>
I think Aymen means to have the same lags in each equation, but with
a
"gappy" structure, for example include the first and third lags, but not
the second (everywhere). Then the VAR-OLS-estimation would still apply.
Vaguely I think there was a discussion about this some time back, but I
don't remember the result. Of course it's still possible to specifiy
this manually as a system and do OLS, but it's not very convenient.
sorry,
sven
Am 28.04.2013 20:54, schrieb Summers, Peter:
> Aymen,
>
> If your lags differ in each equation, it's not really a VAR. However
> you could specify the equations separately and estimate them as a SUR
> model. Have a look at the "system" command.
>
> I hope this helps,
>
> PS ________________________________________ From:
> gretl-users-bounces(a)lists.wfu.edu [gretl-users-bounces(a)lists.wfu.edu]
> on behalf of aymen kaabi [kaabiayman(a)gmail.com] Sent: Sunday, April
> 28, 2013 12:34 PM To: Gretl list Subject: [Gretl-users] vector
> autoregressive
>
> good evening
>
>
> i have a problem with gretl_var.
>
> first with the function ar_model i can specify the lags
>
> FOR EXAMPLE: Yt= Yt-3 + Y t-2 + Yt-6 +CONST but with gretl I CAN'T
> i just enter the lag order .
>
> HOW can i specify the lags .
>
> my project work personalized lags please answer me .
>
>
> thanks for all
>
> _______________________________________________ Gretl-users mailing
> list Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
>
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
6:53 p.m.
Here is my take. Simple VARs are estimated by least squares applied to
several equations at once. Since each equation contains the same
variables, there is no gain from system estimation. So, just estimate them
one by one. Use least squares e.g.,
For lags 2, 3 and 6 with an intercept:
ols Y const Y(-2, -3, -6)
Cheers,
Lee
On Mon, Apr 29, 2013 at 12:16 PM, aymen kaabi <kaabiayman(a)gmail.com> wrote:
thanks sven .
that's exactly what i want .
who understand what i want please describe me how can i do this.
2013/4/28 Sven Schreiber <svetosch(a)gmx.net>
> I think Aymen means to have the same lags in each equation, but with a
> "gappy" structure, for example include the first and third lags, but not
> the second (everywhere). Then the VAR-OLS-estimation would still apply.
>
> Vaguely I think there was a discussion about this some time back, but I
> don't remember the result. Of course it's still possible to specifiy
> this manually as a system and do OLS, but it's not very convenient.
>
> sorry,
> sven
>
>
> Am 28.04.2013 20:54, schrieb Summers, Peter:
> > Aymen,
> >
> > If your lags differ in each equation, it's not really a VAR. However
> > you could specify the equations separately and estimate them as a SUR
> > model. Have a look at the "system" command.
> >
> > I hope this helps,
> >
> > PS ________________________________________ From:
> > gretl-users-bounces(a)lists.wfu.edu [gretl-users-bounces(a)lists.wfu.edu]
> > on behalf of aymen kaabi [kaabiayman(a)gmail.com] Sent: Sunday, April
> > 28, 2013 12:34 PM To: Gretl list Subject: [Gretl-users] vector
> > autoregressive
> >
> > good evening
> >
> >
> > i have a problem with gretl_var.
> >
> > first with the function ar_model i can specify the lags
> >
> > FOR EXAMPLE: Yt= Yt-3 + Y t-2 + Yt-6 +CONST but with gretl I CAN'T
> > i just enter the lag order .
> >
> > HOW can i specify the lags .
> >
> > my project work personalized lags please answer me .
> >
> >
> > thanks for all
> >
> > _______________________________________________ Gretl-users mailing
> > list Gretl-users(a)lists.wfu.edu
> > http://lists.wfu.edu/mailman/listinfo/gretl-users
> >
>
> _______________________________________________
> Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
>
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
--
Lee Adkins
Professor of Economics
lee.adkins(a)okstate.edu
learneconometrics.com
7:46 p.m.
Ok, so I remembered some things correctly (see below), but indeed had
forgotten the most important part: It's already there! Just push the
"lags" button in the VAR specification dialog and enter your individual
lag structure (comma-separated I think).
See http://lists.wfu.edu/pipermail/gretl-devel/2009-February/001682.html
for the original discussion, already four years ago... time passes
quickly...
It remains to be said that the CLI version is still undocumented at
least in the online help.
hth,
sven
Am 29.04.2013 20:53, schrieb Lee Adkins:
Here is my take. Simple VARs are estimated by least squares applied
to
several equations at once. Since each equation contains the same
variables, there is no gain from system estimation. So, just estimate
them one by one. Use least squares e.g.,
For lags 2, 3 and 6 with an intercept:
ols Y const Y(-2, -3, -6)
Cheers,
Lee
On Mon, Apr 29, 2013 at 12:16 PM, aymen kaabi <kaabiayman(a)gmail.com
<mailto:kaabiayman@gmail.com>> wrote:
thanks sven .
that's exactly what i want .
who understand what i want please describe me how can i do this.
2013/4/28 Sven Schreiber <svetosch(a)gmx.net <mailto:svetosch@gmx.net>>
I think Aymen means to have the same lags in each equation, but
with a
"gappy" structure, for example include the first and third lags,
but not
the second (everywhere). Then the VAR-OLS-estimation would still
apply.
Vaguely I think there was a discussion about this some time
back, but I
don't remember the result. Of course it's still possible to specifiy
this manually as a system and do OLS, but it's not very convenient.
sorry,
sven
8:31 p.m.
thanks Sven Schreiber
i talk about the library but you talk about the program.
i search solution to perform var "gappy" structure, for example include
the first and third lags
with functions in library . if you an idea please help me.
2013/4/29 Sven Schreiber <svetosch(a)gmx.net>
Ok, so I remembered some things correctly (see below), but indeed
had
forgotten the most important part: It's already there! Just push the
"lags" button in the VAR specification dialog and enter your individual
lag structure (comma-separated I think).
See http://lists.wfu.edu/pipermail/gretl-devel/2009-February/001682.html
for the original discussion, already four years ago... time passes
quickly...
It remains to be said that the CLI version is still undocumented at
least in the online help.
hth,
sven
Am 29.04.2013 20:53, schrieb Lee Adkins:
> Here is my take. Simple VARs are estimated by least squares applied to
> several equations at once. Since each equation contains the same
> variables, there is no gain from system estimation. So, just estimate
> them one by one. Use least squares e.g.,
>
> For lags 2, 3 and 6 with an intercept:
>
> ols Y const Y(-2, -3, -6)
>
> Cheers,
> Lee
>
>
> On Mon, Apr 29, 2013 at 12:16 PM, aymen kaabi <kaabiayman(a)gmail.com
> <mailto:kaabiayman@gmail.com>> wrote:
>
> thanks sven .
>
> that's exactly what i want .
>
> who understand what i want please describe me how can i do this.
>
>
>
> 2013/4/28 Sven Schreiber <svetosch(a)gmx.net <mailto:svetosch@gmx.net
>>
>
> I think Aymen means to have the same lags in each equation, but
> with a
> "gappy" structure, for example include the first and third lags,
> but not
> the second (everywhere). Then the VAR-OLS-estimation would still
> apply.
>
> Vaguely I think there was a discussion about this some time
> back, but I
> don't remember the result. Of course it's still possible to
specifiy
> this manually as a system and do OLS, but it's not very
convenient.
>
> sorry,
> sven
>
>
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
9:07 p.m.
Am 29.04.2013 22:31, schrieb aymen kaabi:
thanks Sven Schreiber
i talk about the library but you talk about the program.
i search solution to perform var "gappy" structure, for example include
the first and third lags
with functions in library . if you an idea please help me.
Hm, ok, I see. Well if I were you, I would first think about maybe
calling the gretl console program to do what you want instead of digging
inside the library, but that's just my ignorance.
Apart from that, you may want to look at the file lib/src/var.c in the
source, I'm pasting a comment part from it that looks relevant
(including line numbers from the online cvs browser tool) -- especially
the OPT_S thing. Directly above it is a 'maybe_get_lags_list' routine
which belongs in there as well, it seems.
Good luck,
sven
(Clarification for other users reading this: You typically do not (want
to) deal with these things, don't get frightened!)
/**
2467 * gretl_VAR:
2468 * @order: lag order for the VAR
2469 * @list: specification for the first model in the set.
2470 * @dset: dataset struct.
2471 * @opt: if includes %OPT_R, use robust VCV;
2472 * if includes %OPT_H, use HAC VCV;
2473 * if includes %OPT_I, print impulse responses;
2474 * if includes %OPT_F, print forecast variance decompositions;
2475 * if includes %OPT_D, add seasonal dummies;
2476 * if includes %OPT_N, do not include a constant.
2477 * if includes %OPT_Q, do not show individual regressions.
2478 * if includes %OPT_T, include a linear trend.
2479 * if includes %OPT_L, test for optimal lag length (only).
2480 * if includes %OPT_S, look for a specific set of lags
2481 * @prn: gretl printing struct.
2482 * @errp: location to receive error code.
2483 *
2484 * Estimate a vector auto-regression (VAR), print and save
2485 * the results.
2486 *
2487 * Returns: pointer to VAR struct, which may be %NULL on error.
2488 */
9:14 p.m.
thanks a lot Sven Schreiber i didn't find the word to thank you .
2013/4/29 Sven Schreiber <svetosch(a)gmx.net>
Am 29.04.2013 22:31, schrieb aymen kaabi:
> thanks Sven Schreiber
>
>
> i talk about the library but you talk about the program.
>
> i search solution to perform var "gappy" structure, for example include
> the first and third lags
>
> with functions in library . if you an idea please help me.
>
>
Hm, ok, I see. Well if I were you, I would first think about maybe
calling the gretl console program to do what you want instead of digging
inside the library, but that's just my ignorance.
Apart from that, you may want to look at the file lib/src/var.c in the
source, I'm pasting a comment part from it that looks relevant
(including line numbers from the online cvs browser tool) -- especially
the OPT_S thing. Directly above it is a 'maybe_get_lags_list' routine
which belongs in there as well, it seems.
Good luck,
sven
(Clarification for other users reading this: You typically do not (want
to) deal with these things, don't get frightened!)
/**
2467 * gretl_VAR:
2468 * @order: lag order for the VAR
2469 * @list: specification for the first model in the set.
2470 * @dset: dataset struct.
2471 * @opt: if includes %OPT_R, use robust VCV;
2472 * if includes %OPT_H, use HAC VCV;
2473 * if includes %OPT_I, print impulse responses;
2474 * if includes %OPT_F, print forecast variance decompositions;
2475 * if includes %OPT_D, add seasonal dummies;
2476 * if includes %OPT_N, do not include a constant.
2477 * if includes %OPT_Q, do not show individual regressions.
2478 * if includes %OPT_T, include a linear trend.
2479 * if includes %OPT_L, test for optimal lag length (only).
2480 * if includes %OPT_S, look for a specific set of lags
2481 * @prn: gretl printing struct.
2482 * @errp: location to receive error code.
2483 *
2484 * Estimate a vector auto-regression (VAR), print and save
2485 * the results.
2486 *
2487 * Returns: pointer to VAR struct, which may be %NULL on error.
2488 */
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
2:38 p.m.
On Mon, 29 Apr 2013, Sven Schreiber wrote:
Ok, so I remembered some things correctly (see below), but indeed
had
forgotten the most important part: It's already there! Just push the
"lags" button in the VAR specification dialog and enter your individual
lag structure (comma-separated I think).
See http://lists.wfu.edu/pipermail/gretl-devel/2009-February/001682.html
for the original discussion, already four years ago... time passes
quickly...
It remains to be said that the CLI version is still undocumented at
least in the online help.
Thanks for chasing this down, Sven.
The CLI counterpart to the special use of the "Lags" button in
the GUI is appending an option string of the form
--lags=<lags>
e.g.
--lags=1,2,4 # or --lags="1,2,4"
The reason this has never been documented is that it was a
quick-and-dirty hack. It's not nice to have a lag order as the
first argument to the "var" command plus an option to specify
lags at the end. The right way to do this would be to allow
use of the name of a matrix in place of the scalar "order"
argument, as in
matrix p = {1,2,4}
var p Ylist
I guess, though, that implementing that functionality has not
seemed like a high priority to date.
Anyway, as things stand, if one wants to use the libgretl API
to estimate a VAR with "gappy" lags one has to emulate the
setting of the "--lags=..." option, as in
<C>
GRETL_VAR *v;
int err = 0;
set_optval_string(VAR, OPT_S, "1,2,4");
v = gretl_VAR(4, Ylist, dataset, OPT_S, prn, &err);
</C>
Note that this is not guaranteed to be stable API. If we get
around to allowing a matrix argument for the lag order the
clunky OPT_S business may disappear.
Allin Cottrell
5:18 p.m.
On Tue, 30 Apr 2013, Allin Cottrell wrote:
On Mon, 29 Apr 2013, Sven Schreiber wrote:
> Ok, so I remembered some things correctly (see below), but indeed had
> forgotten the most important part: It's already there! Just push the
> "lags" button in the VAR specification dialog and enter your individual
> lag structure (comma-separated I think).
>
> See http://lists.wfu.edu/pipermail/gretl-devel/2009-February/001682.html
> for the original discussion, already four years ago... time passes
> quickly...
>
> It remains to be said that the CLI version is still undocumented at
> least in the online help.
Thanks for chasing this down, Sven.
The CLI counterpart to the special use of the "Lags" button in the GUI is
appending an option string of the form
--lags=<lags>
e.g.
--lags=1,2,4 # or --lags="1,2,4"
The reason this has never been documented is that it was a quick-and-dirty
hack. It's not nice to have a lag order as the first argument to the "var"
command plus an option to specify lags at the end. The right way to do this
would be to allow use of the name of a matrix in place of the scalar "order"
argument, as in
matrix p = {1,2,4}
var p Ylist
I guess, though, that implementing that functionality has not seemed like a
high priority to date.
OK, it's now implemented in CVS, and the undocumented
"--lags=" option has been removed. Formulations of the
following sort will now work:
var 4 Ylist # still OK, of course
matrix p = {1,2,4}
var p Ylist # give a named vector
var {1,2,4} Ylist # give an inline vector
I'll document this shortly.
In regard to the libgretl API, a second parameter to gretl_VAR
has been inserted, namely a list of lags. The argument can be
given as NULL for regular VAR behavior. Examples of use:
<C>
/* regular VAR specification */
v = gretl_VAR(4, NULL, dataset, OPT_NONE, prn, &err);
/* gappy specification */
int laglist[4] = {3, 1, 2, 4};
v = gretl_VAR(4, laglist, dataset, OPT_NONE, prn, &err);
</C>
"laglist" is in standard gretl list form, with the first
element indicating the number of elements that follow. So the
second VAR will include the three lags 1, 2 and 4.
Allin Cottrell
7:47 p.m.
thanks lee
but only Sven Schreiber who understand me . i just want find a function do
what Sven Schreiber explain exactly i know how can i do ar with different
lags like Yt= Yt-1+ Yt-6 + Yt-3 not lag order but with var i didn't find
solution.
2013/4/29 Lee Adkins <lee.adkins(a)okstate.edu>
Here is my take. Simple VARs are estimated by least squares applied
to
several equations at once. Since each equation contains the same
variables, there is no gain from system estimation. So, just estimate them
one by one. Use least squares e.g.,
For lags 2, 3 and 6 with an intercept:
ols Y const Y(-2, -3, -6)
Cheers,
Lee
On Mon, Apr 29, 2013 at 12:16 PM, aymen kaabi <kaabiayman(a)gmail.com>wrote:
> thanks sven .
>
> that's exactly what i want .
>
> who understand what i want please describe me how can i do this.
>
>
>
> 2013/4/28 Sven Schreiber <svetosch(a)gmx.net>
>
>> I think Aymen means to have the same lags in each equation, but with a
>> "gappy" structure, for example include the first and third lags, but
not
>> the second (everywhere). Then the VAR-OLS-estimation would still apply.
>>
>> Vaguely I think there was a discussion about this some time back, but I
>> don't remember the result. Of course it's still possible to specifiy
>> this manually as a system and do OLS, but it's not very convenient.
>>
>> sorry,
>> sven
>>
>>
>> Am 28.04.2013 20:54, schrieb Summers, Peter:
>> > Aymen,
>> >
>> > If your lags differ in each equation, it's not really a VAR. However
>> > you could specify the equations separately and estimate them as a SUR
>> > model. Have a look at the "system" command.
>> >
>> > I hope this helps,
>> >
>> > PS ________________________________________ From:
>> > gretl-users-bounces(a)lists.wfu.edu [gretl-users-bounces(a)lists.wfu.edu]
>> > on behalf of aymen kaabi [kaabiayman(a)gmail.com] Sent: Sunday, April
>> > 28, 2013 12:34 PM To: Gretl list Subject: [Gretl-users] vector
>> > autoregressive
>> >
>> > good evening
>> >
>> >
>> > i have a problem with gretl_var.
>> >
>> > first with the function ar_model i can specify the lags
>> >
>> > FOR EXAMPLE: Yt= Yt-3 + Y t-2 + Yt-6 +CONST but with gretl I CAN'T
>> > i just enter the lag order .
>> >
>> > HOW can i specify the lags .
>> >
>> > my project work personalized lags please answer me .
>> >
>> >
>> > thanks for all
>> >
>> > _______________________________________________ Gretl-users mailing
>> > list Gretl-users(a)lists.wfu.edu
>> > http://lists.wfu.edu/mailman/listinfo/gretl-users
>> >
>>
>> _______________________________________________
>> Gretl-users mailing list
>> Gretl-users(a)lists.wfu.edu
>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>
>
>
> _______________________________________________
> Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
>
--
Lee Adkins
Professor of Economics
lee.adkins(a)okstate.edu
learneconometrics.com
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
8:49 a.m.
If I understood well, this is possible in JMulTi under "Subset
restrictions".Manny
Date: Mon, 29 Apr 2013 18:16:58 +0100
From: kaabiayman(a)gmail.com
To: gretl-users(a)lists.wfu.edu
Subject: Re: [Gretl-users] vector autoregressive
thanks sven .
that's exactly what i want .
who understand what i want please describe me how can i do this.
2013/4/28 Sven Schreiber <svetosch(a)gmx.net
I think Aymen means to have the same lags in each equation, but with a
"gappy" structure, for example include the first and third lags, but not
the second (everywhere). Then the VAR-OLS-estimation would still apply.
Vaguely I think there was a discussion about this some time back, but I
don't remember the result. Of course it's still possible to specifiy
this manually as a system and do OLS, but it's not very convenient.
sorry,
sven
Am 28.04.2013 20:54, schrieb Summers, Peter:
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
Aymen,
If your lags differ in each equation, it's not really a VAR.
However
you could specify the equations separately and estimate them as a SUR
model. Have a look at the "system" command.
I hope this helps,
PS ________________________________________ From:
gretl-users-bounces(a)lists.wfu.edu [gretl-users-bounces(a)lists.wfu.edu]
on behalf of aymen kaabi [kaabiayman(a)gmail.com] Sent: Sunday, April
28, 2013 12:34 PM To: Gretl list Subject: [Gretl-users] vector
autoregressive
good evening
i have a problem with gretl_var.
first with the function ar_model i can specify the lags
FOR EXAMPLE: Yt= Yt-3 + Y t-2 + Yt-6 +CONST but with gretl I
CAN'T
i just enter the lag order .
HOW can i specify the lags .
my project work personalized lags please answer me .
thanks for all
_______________________________________________ Gretl-users mailing
list Gretl-users(a)lists.wfu.edu
8:53 a.m.
Ok! Thanx.Manny
Date: Wed, 1 May 2013 10:52:01 +0200
From: svetosch(a)gmx.net
To: gretl-users(a)lists.wfu.edu
Subject: Re: [Gretl-users] vector autoregressive
Am 01.05.2013 10:49, schrieb Manny B:
> If I understood well, this is possible in JMulTi under "Subset
> restrictions".
> Manny
>
No, that's something else, which is what Peter meant (not a "real" VAR
anymore, so OLS no longer efficient).
-sven
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
4032
days inactive
4035
days old
14 comments
6 participants
participants (6)
-
Allin Cottrell -
aymen kaabi -
Lee Adkins -
Manny B -
Summers, Peter -
Sven Schreiber