KPSS interpolated p-value
by Henrique Andrade
Dear Gretl Community,
I'm a KPSS test newbie and I would like to request your help. Please take a
look at the following small script:
<script>
open australia.gdt
loop foreach i 0 1 2 3 4 --quiet
kpss $i PAU
endloop
</script>
Looking at the interpolated p-values I've found results that are greater
than 1. I would like to know if this is possible.
Best regards,
--
Henrique C. de Andrade
Doutorando em Economia Aplicada
Universidade Federal do Rio Grande do Sul
www.ufrgs.br/ppge
14 years, 5 months
gretl crashes
by Artur T.
Hi,
Using the newest cvs under Ubuntu 10.04, I try to estimate a restricted
VECM (imposed restrictions on alpha and beta). But when I try to
estimate it using the gui gretl crashes. Using a script this does not
happen.
The output is:
----------------
Got a cross-alpha restriction
gretl_cholesky_decomp_solve: rcond = 9.02684e-07 (info = 0)
gretl_cholesky_decomp_solve: rcond = 5.34278e-10 (info = 0)
gretl_invert_symmetric_matrix:
dpotrf failed with info = 4 (n = 10)
input matrix (10 x 10)
0.015089 -1.7891e-06 6.9771e+08 3.7403e+08 -2.9628e+10
6.8545e+09 1.0860e+09 2.5572e+09 5.1781e+09 -4.0054e+09
-1.7891e-06 3.9619e-08 -4.1770e+06 -2.2392e+06 1.7738e+08
-4.1036e+07 -6.5016e+06 -1.5309e+07 -3.1000e+07 2.3980e+07
6.9771e+08 -4.1770e+06 3.4236e+22 1.8353e+22 -1.4538e+24
3.3634e+23 5.3288e+22 1.2548e+23 2.5408e+23 -1.9654e+23
3.7403e+08 -2.2392e+06 1.8353e+22 9.8389e+21 -7.7937e+23
1.8031e+23 2.8567e+22 6.7267e+22 1.3621e+23 -1.0536e+23
-2.9628e+10 1.7738e+08 -1.4538e+24 -7.7937e+23 6.1736e+25
-1.4283e+25 -2.2629e+24 -5.3284e+24 -1.0790e+25 8.3461e+24
6.8545e+09 -4.1036e+07 3.3634e+23 1.8031e+23 -1.4283e+25
3.3043e+24 5.2352e+23 1.2327e+24 2.4962e+24 -1.9309e+24
1.0860e+09 -6.5016e+06 5.3288e+22 2.8567e+22 -2.2629e+24
5.2352e+23 8.2944e+22 1.9531e+23 3.9548e+23 -3.0592e+23
2.5572e+09 -1.5309e+07 1.2548e+23 6.7267e+22 -5.3284e+24
1.2327e+24 1.9531e+23 4.5990e+23 9.3125e+23 -7.2035e+23
5.1781e+09 -3.1000e+07 2.5408e+23 1.3621e+23 -1.0790e+25
2.4962e+24 3.9548e+23 9.3125e+23 1.8857e+24 -1.4587e+24
-4.0054e+09 2.3980e+07 -1.9654e+23 -1.0536e+23 8.3461e+24
-1.9309e+24 -3.0592e+23 -7.2035e+23 -1.4587e+24 1.1283e+24
Segmentation fault
----------------------------
Cheers,
Artur
14 years, 5 months
First time to plot a series after installation of gretl
by yinung＠Gmail
Dear all,
Recently I found that if I plot a series right after a new installation of
gretl for windows (typically upgrading to a new release), it takes a pretty
long time to generate a graph. Some of my students might think that gretl is
"dead" at that moment but in fact it is not.
Does anyone have a clue about this?
Thanks
Yi-Nung Yang
14 years, 5 months
list command in a loop environment
by Artur T.
Hi gretl community,
I just tried to run a script which conducts an OLS estimation and ADF
test based on a list of series. I am pretty sure that a while ago this
worked fine. But now I receive the following error running a simple
script as the one you find below:
"? loop foreach i level
Data types not conformable for operation"
A similar script is the following:
--------
list level = y yp r dp cf q q2 i debt_costs dulc poil debt_networth
debt_equity
loop foreach i level
ols $i const
endloop
-----------------
Any ideas what the problem could be?
Cheers,
Artur
14 years, 5 months
Simple way to construct the index of a series??
by Artur T.
Hi gretl community,
is there a simple way to generate a index series of a variable, e.g.
Lapeyres (year x =100) or Paasche??
At the moment I use a spreadsheet to construct it easily but this
requires the ex- and import of the series of interest every time...
Thanks in advance!
artur
14 years, 5 months
panel unit root tests in CVS
by Allin Cottrell
The topic of testing for unit roots in panel data came up here not
so long ago.
In CVS and the gretl snapshots for Windows and OS X there's now
code in place for doing such testing. It's not in the GUI yet (and
neither is it documented), but is implemented as an extension to
the existing "adf" command. If anyone's interested in testing, I'd
appreciate any comments.
Here's the story. If you issue the "adf" command with panel data
you get (a) a brief account of the test for each panel unit plus
(b) a joint test for al units, if possible. (The per-unit test
output should probably be suppressed if there's an excessive
number of units in the panel, but right now you always get it.)
A couple of restrictions. First, with the regular adf test for
time-series data you can supply a list containing more than one
series and get multiple tests, but with panel data we do only one
series per command; if the input list contains more than one
series we ignore all but the first one. Second, the regular adf
command defaults to performing 3 variants of the test -- with
constant, with linear trend, and with quadratic trend -- though
you can change that via the command options. With panel data we
only do one variant at a time, by default the test with constant.
Some details: The joint tests have the null hypothesis that the
selected series is non-stationary for all panel units; the
alternative is that the series is stationary for at least one
unit. The test statistics are taken from Im, Pesaran and Shin
(Journal of Econometrics, 2003) and Choi (Journal of International
Money and Finance, 2001).
The IPS test involves computing the average value of the
Dickey-Fuller t-statistic (t_bar). If the lag order for the test
is non-zero, t_bar is referred to the distribution that IPS call
W_tbar, which is asymptotically N(0,1). If the lag order is zero
but the time-series length differs across the units, we use the
IPS Z_tbar statistic, which again is N(0,1) asymptotically. If the
order is zero and T is the same for all units we report exact
critical values as tabulated in the IPS article.
(Note that the IPS test cannot be done if the --gls option to the
adf command is given. There are also constraints on the minimal
T for each unit.)
The other test reported (Choi) is based on the p-values from the
per-unit Dickey-Fuller tests. Three variants are given: the
inverse chi-square, inverse normal and logit tests. These differ
in how the p-values are "aggregated" and the distribution that the
resulting statistic follows. For the inverse chi-square test the
null is rejected if the chi-square value is big enough; for the
others it is rejected if the test statistic takes on a large
enough negative value (which is also the case for all variants of
the IPS test).
I've tested some of this against the ipshin.ado extension that's
available for stata. We're pretty close in the cases I've tried
but I'm not sure that ipshin gets everything right so I'm not
necessarily trying to replicate exactly it reports. If anyone can
test against other IPS or Choi test implementations that would be
of interest.
Allin Cottrell
14 years, 5 months
"gig"?
by Olle Olsson
Hi,
I just downloaded version 1.9.1 for Windows and in the Variables>Time Series
menu in the GUI, there is a strange item entitled "gig". What's this?
/Olle Olsson, Sweden
14 years, 5 months
panel unit root tests
by Francisco López Herrera
The point is that when I open a panel data structured data set the adf test is
nos allowable, so my question is: how can I conduct panel data unit root
test?...With my best regards...flh(Parameswara das)
________________________________
De: "gretl-users-request(a)lists.wfu.edu" <gretl-users-request(a)lists.wfu.edu>
Para: gretl-users(a)lists.wfu.edu
Enviado: lunes, 12 de julio, 2010 17:44:32
Asunto: Gretl-users Digest, Vol 42, Issue 17
Send Gretl-users mailing list submissions to
gretl-users(a)lists.wfu.edu
To subscribe or unsubscribe via the World Wide Web, visit
http://lists.wfu.edu/mailman/listinfo/gretl-users
or, via email, send a message with subject or body 'help' to
gretl-users-request(a)lists.wfu.edu
You can reach the person managing the list at
gretl-users-owner(a)lists.wfu.edu
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Gretl-users digest..."
Today's Topics:
1. Re: Gretl-users Digest, Vol 42, Issue 16 (Francisco L?pez Herrera)
----------------------------------------------------------------------
Message: 1
Date: Mon, 12 Jul 2010 15:44:30 -0700 (PDT)
From: Francisco L?pez Herrera <francisco_lopez_herrera(a)yahoo.com.mx>
Subject: Re: [Gretl-users] Gretl-users Digest, Vol 42, Issue 16
To: gretl-users(a)lists.wfu.edu
Message-ID: <177798.99016.qm(a)web52901.mail.re2.yahoo.com>
Content-Type: text/plain; charset="iso-8859-1"
Excuse me, but how can I issue the adf test with panel data?...flh(Parmeswara
das)
________________________________
De: "gretl-users-request(a)lists.wfu.edu" <gretl-users-request(a)lists.wfu.edu>
Para: gretl-users(a)lists.wfu.edu
Enviado: lunes, 12 de julio, 2010 16:47:07
Asunto: Gretl-users Digest, Vol 42, Issue 16
Send Gretl-users mailing list submissions to
gretl-users(a)lists.wfu.edu
To subscribe or unsubscribe via the World Wide Web, visit
http://lists.wfu.edu/mailman/listinfo/gretl-users
or, via email, send a message with subject or body 'help' to
gretl-users-request(a)lists.wfu.edu
You can reach the person managing the list at
gretl-users-owner(a)lists.wfu.edu
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Gretl-users digest..."
Today's Topics:
1. panel unit root tests in CVS (Allin Cottrell)
2. Re: panel unit root tests in CVS (Anutechia Asongu)
3. Re: panel unit root tests in CVS (Summers, Peter)
4. Re: panel unit root tests in CVS (Allin Cottrell)
5. Re: panel unit root tests in CVS (Anutechia Asongu)
----------------------------------------------------------------------
Message: 1
Date: Mon, 12 Jul 2010 14:13:43 -0400 (EDT)
From: Allin Cottrell <cottrell(a)wfu.edu>
Subject: [Gretl-users] panel unit root tests in CVS
To: Gretl users <gretl-users(a)lists.wfu.edu>
Message-ID: <Pine.A41.4.58.1007121339540.2785386(a)f1n11.sp2net.wfu.edu>
Content-Type: TEXT/PLAIN; charset=US-ASCII
The topic of testing for unit roots in panel data came up here not
so long ago.
In CVS and the gretl snapshots for Windows and OS X there's now
code in place for doing such testing. It's not in the GUI yet (and
neither is it documented), but is implemented as an extension to
the existing "adf" command. If anyone's interested in testing, I'd
appreciate any comments.
Here's the story. If you issue the "adf" command with panel data
you get (a) a brief account of the test for each panel unit plus
(b) a joint test for al units, if possible. (The per-unit test
output should probably be suppressed if there's an excessive
number of units in the panel, but right now you always get it.)
A couple of restrictions. First, with the regular adf test for
time-series data you can supply a list containing more than one
series and get multiple tests, but with panel data we do only one
series per command; if the input list contains more than one
series we ignore all but the first one. Second, the regular adf
command defaults to performing 3 variants of the test -- with
constant, with linear trend, and with quadratic trend -- though
you can change that via the command options. With panel data we
only do one variant at a time, by default the test with constant.
Some details: The joint tests have the null hypothesis that the
selected series is non-stationary for all panel units; the
alternative is that the series is stationary for at least one
unit. The test statistics are taken from Im, Pesaran and Shin
(Journal of Econometrics, 2003) and Choi (Journal of International
Money and Finance, 2001).
The IPS test involves computing the average value of the
Dickey-Fuller t-statistic (t_bar). If the lag order for the test
is non-zero, t_bar is referred to the distribution that IPS call
W_tbar, which is asymptotically N(0,1). If the lag order is zero
but the time-series length differs across the units, we use the
IPS Z_tbar statistic, which again is N(0,1) asymptotically. If the
order is zero and T is the same for all units we report exact
critical values as tabulated in the IPS article.
(Note that the IPS test cannot be done if the --gls option to the
adf command is given. There are also constraints on the minimal
T for each unit.)
The other test reported (Choi) is based on the p-values from the
per-unit Dickey-Fuller tests. Three variants are given: the
inverse chi-square, inverse normal and logit tests. These differ
in how the p-values are "aggregated" and the distribution that the
resulting statistic follows. For the inverse chi-square test the
null is rejected if the chi-square value is big enough; for the
others it is rejected if the test statistic takes on a large
enough negative value (which is also the case for all variants of
the IPS test).
I've tested some of this against the ipshin.ado extension that's
available for stata. We're pretty close in the cases I've tried
but I'm not sure that ipshin gets everything right so I'm not
necessarily trying to replicate exactly it reports. If anyone can
test against other IPS or Choi test implementations that would be
of interest.
Allin Cottrell
------------------------------
Message: 2
Date: Mon, 12 Jul 2010 11:28:55 -0700 (PDT)
From: Anutechia Asongu <simplice_peace(a)yahoo.com>
Subject: Re: [Gretl-users] panel unit root tests in CVS
To: Gretl list <gretl-users(a)lists.wfu.edu>
Message-ID: <578737.56898.qm(a)web110307.mail.gq1.yahoo.com>
Content-Type: text/plain; charset="iso-8859-1"
Hi Cottrell,
?????????????? I must? appreciate your concern for this issue. Infact, I brought
up the issue of panel unit test some while ago here. I have two major concerns;
-Do I need to download the software again in a? bid to get this additional
properties of Panel unit incorporated?
-What about Panel Cointegration and Panel VEC/VAR?. Any idea on what the headway
could be?
???????????? Thanks in advance
--- On Mon, 7/12/10, Allin Cottrell <cottrell(a)wfu.edu> wrote:
From: Allin Cottrell <cottrell(a)wfu.edu>
Subject: [Gretl-users] panel unit root tests in CVS
To: "Gretl users" <gretl-users(a)lists.wfu.edu>
Date: Monday, July 12, 2010, 11:13 AM
The topic of testing for unit roots in panel data came up here not
so long ago.
In CVS and the gretl snapshots for Windows and OS X there's now
code in place for doing such testing. It's not in the GUI yet (and
neither is it documented), but is implemented as an extension to
the existing "adf" command. If anyone's interested in testing, I'd
appreciate any comments.
Here's the story. If you issue the "adf" command with panel data
you get (a) a brief account of the test for each panel unit plus
(b) a joint test for al units, if possible. (The per-unit test
output should probably be suppressed if there's an excessive
number of units in the panel, but right now you always get it.)
A couple of restrictions. First, with the regular adf test for
time-series data you can supply a list containing more than one
series and get multiple tests, but with panel data we do only one
series per command; if the input list contains more than one
series we ignore all but the first one. Second, the regular adf
command defaults to performing 3 variants of the test -- with
constant, with linear trend, and with quadratic trend -- though
you can change that via the command options. With panel data we
only do one variant at a time, by default the test with constant.
Some details: The joint tests have the null hypothesis that the
selected series is non-stationary for all panel units; the
alternative is that the series is stationary for at least one
unit. The test statistics are taken from Im, Pesaran and Shin
(Journal of Econometrics, 2003) and Choi (Journal of International
Money and Finance, 2001).
The IPS test involves computing the average value of the
Dickey-Fuller t-statistic (t_bar). If the lag order for the test
is non-zero, t_bar is referred to the distribution that IPS call
W_tbar, which is asymptotically N(0,1). If the lag order is zero
but the time-series length differs across the units, we use the
IPS Z_tbar statistic, which again is N(0,1) asymptotically. If the
order is zero and T is the same for all units we report exact
critical values as tabulated in the IPS article.
(Note that the IPS test cannot be done if the --gls option to the
adf command is given. There are also constraints on the minimal
T for each unit.)
The other test reported (Choi) is based on the p-values from the
per-unit Dickey-Fuller tests. Three variants are given: the
inverse chi-square, inverse normal and logit tests. These differ
in how the p-values are "aggregated" and the distribution that the
resulting statistic follows. For the inverse chi-square test the
null is rejected if the chi-square value is big enough; for the
others it is rejected if the test statistic takes on a large
enough negative value (which is also the case for all variants of
the IPS test).
I've tested some of this against the ipshin.ado extension that's
available for stata. We're pretty close in the cases I've tried
but I'm not sure that ipshin gets everything right so I'm not
necessarily trying to replicate exactly it reports. If anyone can
test against other IPS or Choi test implementations that would be
of interest.
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.wfu.edu/pipermail/gretl-users/attachments/20100712/dcecb262/...
------------------------------
Message: 3
Date: Mon, 12 Jul 2010 14:30:53 -0500
From: "Summers, Peter" <peter.summers(a)ttu.edu>
Subject: Re: [Gretl-users] panel unit root tests in CVS
To: Gretl list <gretl-users(a)lists.wfu.edu>
Message-ID:
<8885DA157278484489B1EBC8B8C043E3FDC4B234DA(a)COTTUS.ttu.edu>
Content-Type: text/plain; charset="us-ascii"
Allin,
Many thanks for implementing this! Once again your development work is highly
correlated with my current projects ;-).
I've successfully run the new command & everything seems sensible (I haven't
tried other implementations so nothing to compare to). One question/suggestion
though: To use the new feature, I had to create a new panel data set from an
existing time series one. This was no problem given the example on p. 25 of the
guide, but I was wondering if it would be possible to have something like a
"--panel" option to use with a list of series. Something like
list varlist = gdp cons inv
adf 0 varlist --panel
So you'd get the IPS & Choi tests at the end. Would that be feasible and/or have
non-negligible demand?
PS
-----Original Message-----
From: gretl-users-bounces(a)lists.wfu.edu
[mailto:gretl-users-bounces@lists.wfu.edu] On Behalf Of Allin Cottrell
Sent: Monday, July 12, 2010 1:14 PM
To: Gretl users
Subject: [Gretl-users] panel unit root tests in CVS
The topic of testing for unit roots in panel data came up here not
so long ago.
In CVS and the gretl snapshots for Windows and OS X there's now
code in place for doing such testing. It's not in the GUI yet (and
neither is it documented), but is implemented as an extension to
the existing "adf" command. If anyone's interested in testing, I'd
appreciate any comments.
Here's the story. If you issue the "adf" command with panel data
you get (a) a brief account of the test for each panel unit plus
(b) a joint test for al units, if possible. (The per-unit test
output should probably be suppressed if there's an excessive
number of units in the panel, but right now you always get it.)
A couple of restrictions. First, with the regular adf test for
time-series data you can supply a list containing more than one
series and get multiple tests, but with panel data we do only one
series per command; if the input list contains more than one
series we ignore all but the first one. Second, the regular adf
command defaults to performing 3 variants of the test -- with
constant, with linear trend, and with quadratic trend -- though
you can change that via the command options. With panel data we
only do one variant at a time, by default the test with constant.
Some details: The joint tests have the null hypothesis that the
selected series is non-stationary for all panel units; the
alternative is that the series is stationary for at least one
unit. The test statistics are taken from Im, Pesaran and Shin
(Journal of Econometrics, 2003) and Choi (Journal of International
Money and Finance, 2001).
The IPS test involves computing the average value of the
Dickey-Fuller t-statistic (t_bar). If the lag order for the test
is non-zero, t_bar is referred to the distribution that IPS call
W_tbar, which is asymptotically N(0,1). If the lag order is zero
but the time-series length differs across the units, we use the
IPS Z_tbar statistic, which again is N(0,1) asymptotically. If the
order is zero and T is the same for all units we report exact
critical values as tabulated in the IPS article.
(Note that the IPS test cannot be done if the --gls option to the
adf command is given. There are also constraints on the minimal
T for each unit.)
The other test reported (Choi) is based on the p-values from the
per-unit Dickey-Fuller tests. Three variants are given: the
inverse chi-square, inverse normal and logit tests. These differ
in how the p-values are "aggregated" and the distribution that the
resulting statistic follows. For the inverse chi-square test the
null is rejected if the chi-square value is big enough; for the
others it is rejected if the test statistic takes on a large
enough negative value (which is also the case for all variants of
the IPS test).
I've tested some of this against the ipshin.ado extension that's
available for stata. We're pretty close in the cases I've tried
but I'm not sure that ipshin gets everything right so I'm not
necessarily trying to replicate exactly it reports. If anyone can
test against other IPS or Choi test implementations that would be
of interest.
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
------------------------------
Message: 4
Date: Mon, 12 Jul 2010 17:36:46 -0400 (EDT)
From: Allin Cottrell <cottrell(a)wfu.edu>
Subject: Re: [Gretl-users] panel unit root tests in CVS
To: Gretl list <gretl-users(a)lists.wfu.edu>
Message-ID: <Pine.A41.4.58.1007121732460.1589636(a)f1n11.sp2net.wfu.edu>
Content-Type: TEXT/PLAIN; charset=UTF-8
On Mon, 12 Jul 2010, Anutechia Asongu wrote:
> I must appreciate your concern for this issue. Infact, I brought
> up the issue of panel unit test some while ago here. I have two
> major concerns;
> -Do I need to download the software again in a? bid to get this
> additional properties of Panel unit incorporated?
If you want to use newly added features then, yes, you will have
to download the latest snapshot.
> -What about Panel Cointegration and Panel VEC/VAR?. Any idea on
> what the headway could be?
Our priority on panel data right now is fuller support for dynamic
panel models (i.e. adding Blundell-Bond). I'm not sure when
support for cointegration analysis might follow.
Allin Cottrell
------------------------------
Message: 5
Date: Mon, 12 Jul 2010 14:47:04 -0700 (PDT)
From: Anutechia Asongu <simplice_peace(a)yahoo.com>
Subject: Re: [Gretl-users] panel unit root tests in CVS
To: Gretl list <gretl-users(a)lists.wfu.edu>
Message-ID: <157560.9284.qm(a)web110306.mail.gq1.yahoo.com>
Content-Type: text/plain; charset="iso-8859-1"
Many Thanks Allin, I'll do that right away.
WE BECOME THE CHANGE WE WANT
?????????????? www.pleho.org
--- On Mon, 7/12/10, Allin Cottrell <cottrell(a)wfu.edu> wrote:
From: Allin Cottrell <cottrell(a)wfu.edu>
Subject: Re: [Gretl-users] panel unit root tests in CVS
To: "Gretl list" <gretl-users(a)lists.wfu.edu>
Date: Monday, July 12, 2010, 2:36 PM
On Mon, 12 Jul 2010, Anutechia Asongu wrote:
> I must appreciate your concern for this issue. Infact, I brought
> up the issue of panel unit test some while ago here. I have two
> major concerns;
> -Do I need to download the software again in a? bid to get this
> additional properties of Panel unit incorporated?
If you want to use newly added features then, yes, you will have
to download the latest snapshot.
> -What about Panel Cointegration and Panel VEC/VAR?. Any idea on
> what the headway could be?
Our priority on panel data right now is fuller support for dynamic
panel models (i.e. adding Blundell-Bond). I'm not sure when
support for cointegration analysis might follow.
Allin Cottrell
-----Inline Attachment Follows-----
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.wfu.edu/pipermail/gretl-users/attachments/20100712/cd29e6e9/...
------------------------------
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
End of Gretl-users Digest, Vol 42, Issue 16
*******************************************
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.wfu.edu/pipermail/gretl-users/attachments/20100712/0bc646a6/...
------------------------------
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
End of Gretl-users Digest, Vol 42, Issue 17
*******************************************
14 years, 5 months
Re: [Gretl-users] Gretl-users Digest, Vol 42, Issue 16
by Francisco López Herrera
Excuse me, but how can I issue the adf test with panel data?...flh(Parmeswara
das)
________________________________
De: "gretl-users-request(a)lists.wfu.edu" <gretl-users-request(a)lists.wfu.edu>
Para: gretl-users(a)lists.wfu.edu
Enviado: lunes, 12 de julio, 2010 16:47:07
Asunto: Gretl-users Digest, Vol 42, Issue 16
Send Gretl-users mailing list submissions to
gretl-users(a)lists.wfu.edu
To subscribe or unsubscribe via the World Wide Web, visit
http://lists.wfu.edu/mailman/listinfo/gretl-users
or, via email, send a message with subject or body 'help' to
gretl-users-request(a)lists.wfu.edu
You can reach the person managing the list at
gretl-users-owner(a)lists.wfu.edu
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Gretl-users digest..."
Today's Topics:
1. panel unit root tests in CVS (Allin Cottrell)
2. Re: panel unit root tests in CVS (Anutechia Asongu)
3. Re: panel unit root tests in CVS (Summers, Peter)
4. Re: panel unit root tests in CVS (Allin Cottrell)
5. Re: panel unit root tests in CVS (Anutechia Asongu)
----------------------------------------------------------------------
Message: 1
Date: Mon, 12 Jul 2010 14:13:43 -0400 (EDT)
From: Allin Cottrell <cottrell(a)wfu.edu>
Subject: [Gretl-users] panel unit root tests in CVS
To: Gretl users <gretl-users(a)lists.wfu.edu>
Message-ID: <Pine.A41.4.58.1007121339540.2785386(a)f1n11.sp2net.wfu.edu>
Content-Type: TEXT/PLAIN; charset=US-ASCII
The topic of testing for unit roots in panel data came up here not
so long ago.
In CVS and the gretl snapshots for Windows and OS X there's now
code in place for doing such testing. It's not in the GUI yet (and
neither is it documented), but is implemented as an extension to
the existing "adf" command. If anyone's interested in testing, I'd
appreciate any comments.
Here's the story. If you issue the "adf" command with panel data
you get (a) a brief account of the test for each panel unit plus
(b) a joint test for al units, if possible. (The per-unit test
output should probably be suppressed if there's an excessive
number of units in the panel, but right now you always get it.)
A couple of restrictions. First, with the regular adf test for
time-series data you can supply a list containing more than one
series and get multiple tests, but with panel data we do only one
series per command; if the input list contains more than one
series we ignore all but the first one. Second, the regular adf
command defaults to performing 3 variants of the test -- with
constant, with linear trend, and with quadratic trend -- though
you can change that via the command options. With panel data we
only do one variant at a time, by default the test with constant.
Some details: The joint tests have the null hypothesis that the
selected series is non-stationary for all panel units; the
alternative is that the series is stationary for at least one
unit. The test statistics are taken from Im, Pesaran and Shin
(Journal of Econometrics, 2003) and Choi (Journal of International
Money and Finance, 2001).
The IPS test involves computing the average value of the
Dickey-Fuller t-statistic (t_bar). If the lag order for the test
is non-zero, t_bar is referred to the distribution that IPS call
W_tbar, which is asymptotically N(0,1). If the lag order is zero
but the time-series length differs across the units, we use the
IPS Z_tbar statistic, which again is N(0,1) asymptotically. If the
order is zero and T is the same for all units we report exact
critical values as tabulated in the IPS article.
(Note that the IPS test cannot be done if the --gls option to the
adf command is given. There are also constraints on the minimal
T for each unit.)
The other test reported (Choi) is based on the p-values from the
per-unit Dickey-Fuller tests. Three variants are given: the
inverse chi-square, inverse normal and logit tests. These differ
in how the p-values are "aggregated" and the distribution that the
resulting statistic follows. For the inverse chi-square test the
null is rejected if the chi-square value is big enough; for the
others it is rejected if the test statistic takes on a large
enough negative value (which is also the case for all variants of
the IPS test).
I've tested some of this against the ipshin.ado extension that's
available for stata. We're pretty close in the cases I've tried
but I'm not sure that ipshin gets everything right so I'm not
necessarily trying to replicate exactly it reports. If anyone can
test against other IPS or Choi test implementations that would be
of interest.
Allin Cottrell
------------------------------
Message: 2
Date: Mon, 12 Jul 2010 11:28:55 -0700 (PDT)
From: Anutechia Asongu <simplice_peace(a)yahoo.com>
Subject: Re: [Gretl-users] panel unit root tests in CVS
To: Gretl list <gretl-users(a)lists.wfu.edu>
Message-ID: <578737.56898.qm(a)web110307.mail.gq1.yahoo.com>
Content-Type: text/plain; charset="iso-8859-1"
Hi Cottrell,
?????????????? I must? appreciate your concern for this issue. Infact, I brought
up the issue of panel unit test some while ago here. I have two major concerns;
-Do I need to download the software again in a? bid to get this additional
properties of Panel unit incorporated?
-What about Panel Cointegration and Panel VEC/VAR?. Any idea on what the headway
could be?
???????????? Thanks in advance
--- On Mon, 7/12/10, Allin Cottrell <cottrell(a)wfu.edu> wrote:
From: Allin Cottrell <cottrell(a)wfu.edu>
Subject: [Gretl-users] panel unit root tests in CVS
To: "Gretl users" <gretl-users(a)lists.wfu.edu>
Date: Monday, July 12, 2010, 11:13 AM
The topic of testing for unit roots in panel data came up here not
so long ago.
In CVS and the gretl snapshots for Windows and OS X there's now
code in place for doing such testing. It's not in the GUI yet (and
neither is it documented), but is implemented as an extension to
the existing "adf" command. If anyone's interested in testing, I'd
appreciate any comments.
Here's the story. If you issue the "adf" command with panel data
you get (a) a brief account of the test for each panel unit plus
(b) a joint test for al units, if possible. (The per-unit test
output should probably be suppressed if there's an excessive
number of units in the panel, but right now you always get it.)
A couple of restrictions. First, with the regular adf test for
time-series data you can supply a list containing more than one
series and get multiple tests, but with panel data we do only one
series per command; if the input list contains more than one
series we ignore all but the first one. Second, the regular adf
command defaults to performing 3 variants of the test -- with
constant, with linear trend, and with quadratic trend -- though
you can change that via the command options. With panel data we
only do one variant at a time, by default the test with constant.
Some details: The joint tests have the null hypothesis that the
selected series is non-stationary for all panel units; the
alternative is that the series is stationary for at least one
unit. The test statistics are taken from Im, Pesaran and Shin
(Journal of Econometrics, 2003) and Choi (Journal of International
Money and Finance, 2001).
The IPS test involves computing the average value of the
Dickey-Fuller t-statistic (t_bar). If the lag order for the test
is non-zero, t_bar is referred to the distribution that IPS call
W_tbar, which is asymptotically N(0,1). If the lag order is zero
but the time-series length differs across the units, we use the
IPS Z_tbar statistic, which again is N(0,1) asymptotically. If the
order is zero and T is the same for all units we report exact
critical values as tabulated in the IPS article.
(Note that the IPS test cannot be done if the --gls option to the
adf command is given. There are also constraints on the minimal
T for each unit.)
The other test reported (Choi) is based on the p-values from the
per-unit Dickey-Fuller tests. Three variants are given: the
inverse chi-square, inverse normal and logit tests. These differ
in how the p-values are "aggregated" and the distribution that the
resulting statistic follows. For the inverse chi-square test the
null is rejected if the chi-square value is big enough; for the
others it is rejected if the test statistic takes on a large
enough negative value (which is also the case for all variants of
the IPS test).
I've tested some of this against the ipshin.ado extension that's
available for stata. We're pretty close in the cases I've tried
but I'm not sure that ipshin gets everything right so I'm not
necessarily trying to replicate exactly it reports. If anyone can
test against other IPS or Choi test implementations that would be
of interest.
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.wfu.edu/pipermail/gretl-users/attachments/20100712/dcecb262/...
------------------------------
Message: 3
Date: Mon, 12 Jul 2010 14:30:53 -0500
From: "Summers, Peter" <peter.summers(a)ttu.edu>
Subject: Re: [Gretl-users] panel unit root tests in CVS
To: Gretl list <gretl-users(a)lists.wfu.edu>
Message-ID:
<8885DA157278484489B1EBC8B8C043E3FDC4B234DA(a)COTTUS.ttu.edu>
Content-Type: text/plain; charset="us-ascii"
Allin,
Many thanks for implementing this! Once again your development work is highly
correlated with my current projects ;-).
I've successfully run the new command & everything seems sensible (I haven't
tried other implementations so nothing to compare to). One question/suggestion
though: To use the new feature, I had to create a new panel data set from an
existing time series one. This was no problem given the example on p. 25 of the
guide, but I was wondering if it would be possible to have something like a
"--panel" option to use with a list of series. Something like
list varlist = gdp cons inv
adf 0 varlist --panel
So you'd get the IPS & Choi tests at the end. Would that be feasible and/or have
non-negligible demand?
PS
-----Original Message-----
From: gretl-users-bounces(a)lists.wfu.edu
[mailto:gretl-users-bounces@lists.wfu.edu] On Behalf Of Allin Cottrell
Sent: Monday, July 12, 2010 1:14 PM
To: Gretl users
Subject: [Gretl-users] panel unit root tests in CVS
The topic of testing for unit roots in panel data came up here not
so long ago.
In CVS and the gretl snapshots for Windows and OS X there's now
code in place for doing such testing. It's not in the GUI yet (and
neither is it documented), but is implemented as an extension to
the existing "adf" command. If anyone's interested in testing, I'd
appreciate any comments.
Here's the story. If you issue the "adf" command with panel data
you get (a) a brief account of the test for each panel unit plus
(b) a joint test for al units, if possible. (The per-unit test
output should probably be suppressed if there's an excessive
number of units in the panel, but right now you always get it.)
A couple of restrictions. First, with the regular adf test for
time-series data you can supply a list containing more than one
series and get multiple tests, but with panel data we do only one
series per command; if the input list contains more than one
series we ignore all but the first one. Second, the regular adf
command defaults to performing 3 variants of the test -- with
constant, with linear trend, and with quadratic trend -- though
you can change that via the command options. With panel data we
only do one variant at a time, by default the test with constant.
Some details: The joint tests have the null hypothesis that the
selected series is non-stationary for all panel units; the
alternative is that the series is stationary for at least one
unit. The test statistics are taken from Im, Pesaran and Shin
(Journal of Econometrics, 2003) and Choi (Journal of International
Money and Finance, 2001).
The IPS test involves computing the average value of the
Dickey-Fuller t-statistic (t_bar). If the lag order for the test
is non-zero, t_bar is referred to the distribution that IPS call
W_tbar, which is asymptotically N(0,1). If the lag order is zero
but the time-series length differs across the units, we use the
IPS Z_tbar statistic, which again is N(0,1) asymptotically. If the
order is zero and T is the same for all units we report exact
critical values as tabulated in the IPS article.
(Note that the IPS test cannot be done if the --gls option to the
adf command is given. There are also constraints on the minimal
T for each unit.)
The other test reported (Choi) is based on the p-values from the
per-unit Dickey-Fuller tests. Three variants are given: the
inverse chi-square, inverse normal and logit tests. These differ
in how the p-values are "aggregated" and the distribution that the
resulting statistic follows. For the inverse chi-square test the
null is rejected if the chi-square value is big enough; for the
others it is rejected if the test statistic takes on a large
enough negative value (which is also the case for all variants of
the IPS test).
I've tested some of this against the ipshin.ado extension that's
available for stata. We're pretty close in the cases I've tried
but I'm not sure that ipshin gets everything right so I'm not
necessarily trying to replicate exactly it reports. If anyone can
test against other IPS or Choi test implementations that would be
of interest.
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
------------------------------
Message: 4
Date: Mon, 12 Jul 2010 17:36:46 -0400 (EDT)
From: Allin Cottrell <cottrell(a)wfu.edu>
Subject: Re: [Gretl-users] panel unit root tests in CVS
To: Gretl list <gretl-users(a)lists.wfu.edu>
Message-ID: <Pine.A41.4.58.1007121732460.1589636(a)f1n11.sp2net.wfu.edu>
Content-Type: TEXT/PLAIN; charset=UTF-8
On Mon, 12 Jul 2010, Anutechia Asongu wrote:
> I must appreciate your concern for this issue. Infact, I brought
> up the issue of panel unit test some while ago here. I have two
> major concerns;
> -Do I need to download the software again in a? bid to get this
> additional properties of Panel unit incorporated?
If you want to use newly added features then, yes, you will have
to download the latest snapshot.
> -What about Panel Cointegration and Panel VEC/VAR?. Any idea on
> what the headway could be?
Our priority on panel data right now is fuller support for dynamic
panel models (i.e. adding Blundell-Bond). I'm not sure when
support for cointegration analysis might follow.
Allin Cottrell
------------------------------
Message: 5
Date: Mon, 12 Jul 2010 14:47:04 -0700 (PDT)
From: Anutechia Asongu <simplice_peace(a)yahoo.com>
Subject: Re: [Gretl-users] panel unit root tests in CVS
To: Gretl list <gretl-users(a)lists.wfu.edu>
Message-ID: <157560.9284.qm(a)web110306.mail.gq1.yahoo.com>
Content-Type: text/plain; charset="iso-8859-1"
Many Thanks Allin, I'll do that right away.
WE BECOME THE CHANGE WE WANT
?????????????? www.pleho.org
--- On Mon, 7/12/10, Allin Cottrell <cottrell(a)wfu.edu> wrote:
From: Allin Cottrell <cottrell(a)wfu.edu>
Subject: Re: [Gretl-users] panel unit root tests in CVS
To: "Gretl list" <gretl-users(a)lists.wfu.edu>
Date: Monday, July 12, 2010, 2:36 PM
On Mon, 12 Jul 2010, Anutechia Asongu wrote:
> I must appreciate your concern for this issue. Infact, I brought
> up the issue of panel unit test some while ago here. I have two
> major concerns;
> -Do I need to download the software again in a? bid to get this
> additional properties of Panel unit incorporated?
If you want to use newly added features then, yes, you will have
to download the latest snapshot.
> -What about Panel Cointegration and Panel VEC/VAR?. Any idea on
> what the headway could be?
Our priority on panel data right now is fuller support for dynamic
panel models (i.e. adding Blundell-Bond). I'm not sure when
support for cointegration analysis might follow.
Allin Cottrell
-----Inline Attachment Follows-----
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.wfu.edu/pipermail/gretl-users/attachments/20100712/cd29e6e9/...
------------------------------
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
End of Gretl-users Digest, Vol 42, Issue 16
*******************************************
14 years, 5 months
gretlpre.tex problem
by Robert Levinson
Hi all -
I'm having trouble getting TeX to work properly with Gretl 1.9.1 under Mac OS X 10.6.4.
Suppose for instance that I want to see the results of an OLS regression as TeX output (the problem surfaces no matter what procedure I run). After running the regression I choose the LaTeX drop-down menu and select View -> Tabular. I get an error box that says it "Failed to process TeX file". On closing that error a log becomes visible, saying at the beginning --
-----------
This is pdfTeX, Version 3.1415926-1.40.10 (TeX Live 2009) (format=pdflatex 2010.7.8) 8 JUL 2010 14:04
entering extended mode
%&-line parsing enabled.
**\batchmode \input window
(./window.tex
! LaTeX Error: Missing \begin{document}.
See the LaTeX manual or LaTeX Companion for explanation.
Type H <return> for immediate help.
...
l.1 \begin{center}
You're in trouble here. Try typing <return> to proceed.
(more follows)
-----------
The problem seems to be that Gretl isn't using the gretlpre.tex file that I have sitting in my working directory. I tried the sample gretlpre code in the current Gretl manual. I also tried deleting the file (the manual says that if there's no gretlpre.tex file the preamble defaults to the standard preamble). The problem remains. When instead I choose to save the TeX output (rather than viewing it) the regression results are indeed saved as a TeX file. but again lacks the preamble that's in gretlpre.tex.
Any help appreciated.
Many thanks.
- Bob
14 years, 5 months
