White's test for heteroskedasticity (GUI)
by Henrique
Dear Gretl community,
When I try to perform a "White's test for heteroskedasticity" on OLS
models with only one term (x = b0 + b1y) Gretl shows me only the "White's
test for heteroskedasticity (squared only)". I believe that this could be a
little bit confusing and I suggest to introduce an alert when we try to
perform an "White's test for heteroskedasticity" on regressions (OLS) that
have only one explanatory variable.
I'd have attached a session file that shows this problem.
Best,
Henrique C. de Andrade
Doutorando em Economia Aplicada
Universidade Federal do Rio Grande do Sul
www.ufrgs.br/ppge
16 years
fracdiff and fraclag
by Klein, Christoph
Hi,
I have created another patch. This one implements a more intelligent calculation of the fracdiff function in the case of autoreg evaluations. In this case it justs calculates the currently evaluated observation. Additionally it adds a fraclag command to gretl. The fractional lag operator is defined as 1-(1-L)^d. Is it ok to "overload" the fracdiff command for this purpose?
Christoph
16 years
autoreg and bessel
by Klein, Christoph
Hi,
another patch for support of autoreg formulas in the new bessel function.
Christoph
16 years
bessel functions
by Klein, Christoph
Hi,
I wrote a mail yesterday, but somehow it got lost?! I'm working on the bessel code right now. The attached patch implements a bessel function which can be called like this:
bessel(<type>, <order>, <value>) example bessel("K",0.5,1).
the function works for J,I,K,Y bessel functions. For K bessel there was no support for fractional orders in cephes, thats why there is additonal code for these cases. the fractional order for k and negative orders for I work reasonably well in the range lambda<10 and value<20. The attached test script checks against the implementation of R.
It's already quite late here. If you have questions or sugestions I will be glad to answer them tomorrow in the morning!
Christoph
16 years
Sv: Gretl-users Digest, Vol 26, Issue 5
by Lars Pålsson-Syll
Re: cross-tabulation without original raw data
Hi,
Is there any possibility of performing a cross-tabulation in gretl when
you only have
aggregated frequencies (I'm thinking of something analogous to Stata's
"tabi" command )?
Beste regards,
Lars P Syll
Lars Pålsson Syll
Professor of Civics
Malmö University
>>> <gretl-users-request(a)lists.wfu.edu> 09-03-28 19:36 >>>
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. Runs test (Ana Jes?s L?pez)
2. Re: Runs test (Ignacio Diaz-Emparanza)
3. RE: Recursive resuduals.... (Mariusz Doszy?)
4. figure, --loess-fit and loop (artur tarassow)
5. Re: figure, --loess-fit and loop (Allin Cottrell)
6. Re: figure, --loess-fit and loop (artur tarassow)
7. getting familiar with the scripts interface (Raz Lev)
8. Re: getting familiar with the scripts interface (Allin Cottrell)
----------------------------------------------------------------------
Message: 1
Date: Tue, 17 Mar 2009 12:06:18 +0100
From: Ana Jes?s L?pez <anaj(a)uniovi.es>
Subject: [Gretl-users] Runs test
To: "'Gretl list'" <gretl-users(a)lists.wfu.edu>
Message-ID: <002101c9a6f0$660f2440$322d6cc0$@es>
Content-Type: text/plain; charset=utf-8
Hello
I am having problems while running the runs test in levels, since the
number of runs (R) in the variable 'c_cemento'is always equal to 1 and
therefore the p value is always very low thus rejecting the null
hypothesis of randomness.
This problem does not appear when the test is run over the
first-difference series.
Thanks and best regards
Ana J.L├│pez
anaj(a)uniovi.es
------------------------------
Message: 2
Date: Tue, 17 Mar 2009 13:35:48 +0100
From: "Ignacio Diaz-Emparanza" <ignacio.diaz-emparanza(a)ehu.es>
Subject: Re: [Gretl-users] Runs test
To: Gretl list <gretl-users(a)lists.wfu.edu>
Message-ID: <200903171335.48879.ignacio.diaz-emparanza(a)ehu.es>
Content-Type: text/plain; charset="utf-8"
El Tuesday 17 March 2009 12:06:18 Ana Jes├║s L├│pez escribi├│:
> Hello
> I am having problems while running the runs test in levels, since the
> number of runs (R) in the variable 'c_cemento'is always equal to 1 and
> therefore the p value is always very low thus rejecting the null
hypothesis
> of randomness. This problem does not appear when the test is run over
the
> first-difference series. Thanks and best regards
This is normal if c_cemento is always positive. The runs test is based
on the
number of positive runs and negative runs, and tests randomness around
the
zero value. If all numbers are positive, there is only one run.
(See for example http://en.wikipedia.org/wiki/Wald-Wolfowitz_runs_test )
--
Ignacio Diaz-Emparanza
DEPARTAMENTO DE ECONOMÍA APLICADA III (ECONOMETRÍA Y ESTADÍSTICA)
UPV/EHU
Avda. Lehendakari Aguirre, 83 | 48015 BILBAO
T.: +34 946013732 | F.: +34 946013754
www.et.bs.ehu.es
------------------------------
Message: 3
Date: Tue, 17 Mar 2009 16:33:19 +0100
From: Mariusz Doszy? <madosz(a)wp.pl>
Subject: RE: [Gretl-users] Recursive resuduals....
To: "'Gretl list'" <gretl-users(a)lists.wfu.edu>
Message-ID: <000701c9a715$bbd28e60$3377ab20$@pl>
Content-Type: text/plain; charset="utf-8"
Hello...
I've installed current snapshot and now it works fine... But thanks...
Best wishes,
Mariusz
Poland
-----Original Message-----
From: gretl-users-bounces(a)lists.wfu.edu
[mailto:gretl-users-bounces@lists.wfu.edu] On Behalf Of Ignacio
Diaz-Emparanza
Sent: Tuesday, March 17, 2009 11:41 AM
To: Gretl list
Subject: Re: [Gretl-users] Recursive resuduals....
El Monday 16 Ma> Hello.I'm trying to write a script computing recursive residuals.
> You're probably familiar with the concept how to obtain such a
> residual: after estimating model for (n+1) observations (with dummy
> variable equals 1 for observation number (n+1)) t-value for dummy
> variable is multiplied by standard error of residuals. Unfortunately
something is wrong.
I see no problem. The output I obtained is in the attachment.
------------------------------
Message: 4
Date: Sat, 28 Mar 2009 12:30:21 +0100
From: artur tarassow <artur.tarassow(a)googlemail.com>
Subject: [Gretl-users] figure, --loess-fit and loop
To: gretl-users(a)lists.wfu.edu
Message-ID:
<e58e62710903280430y31778b2dne2568a9eb51261ad(a)mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
Hello,
when I try to generate a xy-plot with the "--loess-fit" option using
the loop command
foreach i 6 10 100
gnuplot hph_$i hpwa_lab_$i --loess-fit --output=hp_h_wa_lab_$i.eps
endloop
the compiled files will be corrupted. For the other "fit"-options it
works well.
Is there a reason for this?
Best,
Artur
------------------------------
Message: 5
Date: Sat, 28 Mar 2009 10:10:47 -0400 (EDT)
From: Allin Cottrell <cottrell(a)wfu.edu>
Subject: Re: [Gretl-users] figure, --loess-fit and loop
To: Gretl list <gretl-users(a)lists.wfu.edu>
Message-ID: <Pine.A41.4.58.0903281007260.250364(a)f1n11.sp2net.wfu.edu>
Content-Type: TEXT/PLAIN; charset=US-ASCII
On Sat, 28 Mar 2009, artur tarassow wrote:
> when I try to generate a xy-plot with the "--loess-fit" option using
> the loop command
>
> foreach i 6 10 100
> gnuplot hph_$i hpwa_lab_$i --loess-fit --output=hp_h_wa_lab_$i.eps
> endloop
>
> the compiled files will be corrupted. For the other
> "fit"-options it works well.
>
> Is there a reason for this?
Yes ;-) The file were not actually corrupted, they were just
plain text gnuplot command files, not EPS. The trouble is that
when you select the loess plot option a special graphing function
is called, and that function was not getting the message that a
specific output format was wanted.
This is now fixed in CVS; I'll put up a fixed Windows snapshot
shortly.
Allin.
------------------------------
Message: 6
Date: Sat, 28 Mar 2009 15:33:54 +0100
From: artur tarassow <artur.tarassow(a)googlemail.com>
Subject: Re: [Gretl-users] figure, --loess-fit and loop
To: Gretl list <gretl-users(a)lists.wfu.edu>
Message-ID:
<e58e62710903280733x201d7f05ua05cd24d71707ed3(a)mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
Excellent work Allin,
thanks very much for that! ;)
Best,
Artur
2009/3/28 Allin Cottrell <cottrell(a)wfu.edu>:
> On Sat, 28 Mar 2009, artur tarassow wrote:
>
>> when I try to generate a xy-plot with the "--loess-fit" option using
>> the loop command
>>
>> foreach i 6 10 100
>> gnuplot hph_$i hpwa_lab_$i --loess-fit --output=hp_h_wa_lab_$i.eps
>> endloop
>>
>> the compiled files will be corrupted. For the other
>> "fit"-options it works well.
>>
>> Is there a reason for this?
>
> Yes ;-) áThe file were not actually corrupted, they were just
> plain text gnuplot command files, not EPS. áThe trouble is that
> when you select the loess plot option a special graphing function
> is called, and that function was not getting the message that a
> specific output format was wanted.
>
> This is now fixed in CVS; I'll put up a fixed Windows snapshot
> shortly.
>
> Allin.
> _______________________________________________
> Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
>
------------------------------
Message: 7
Date: Sat, 28 Mar 2009 09:13:55 -0700 (PDT)
From: Raz Lev <raz_lev(a)yahoo.com>
Subject: [Gretl-users] getting familiar with the scripts interface
To: gretl-users(a)lists.wfu.edu
Message-ID: <792976.52655.qm(a)web34601.mail.mud.yahoo.com>
Content-Type: text/plain; charset="us-ascii"
hi
I am trying to learn how to use the scripting interface - is there any
trainDoes anyone have any tips about getting started? I am trying to do basic
things like cleaning up a sample, getting descriptive stats and running
some regressions and even though I can do all of this through the GUI I
would like to get familiar with the console.
thanks.
"I often see problems in the world and wonder why someone doesn't fix
them. Then I realize that I am someone, and that makes me feel bad
because it is all my fault."
http://dilbert.com/blog/entry/fixing_the_world/
16 years
[OT] Gnuplot Book: Galley Proofs Received!
by Philipp K. Janert
After a long delay, due to technical difficulties during
final layout, today I received the first set of galley
proofs for review!
This means that the actual printed book should definitely
be available for shipping within the next few weeks.
For more details on the book, see:
www.manning.com/janert
Many thanks to everyone for your patience!
Best,
Ph.
16 years
A time series question
by Henrique
Dear Gretl community,
I have three variables (time series) named x, y and z. I need to know
how "x" changes the impact of "y" on "z". Any suggestions?
Best,
Henrique C. de Andrade
Doutorando em Economia Aplicada
Universidade Federal do Rio Grande do Sul
www.ufrgs.br/ppge
16 years
Sv: Gretl-users Digest, Vol 29, Issue 4
by Lars Pålsson-Syll
Re: xtab --matrix
Thanks a lot Allin for adding this handy function!
Best regards,
Lars Palsson Syll
Lars Pålsson Syll
Professor of Civics
Malmö University
>>> <gretl-users-request(a)lists.wfu.edu> 09-06-15 16:51 >>>
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: Sv: Gretl-users Digest, Vol 26, Issue 5 (Allin Cottrell)
2. Heteroskedasticty.. (bulent.miran(a)ege.edu.tr)
3. Re: Sv: Gretl-users Digest, Vol 26, Issue 5 (Allin Cottrell)
4. gretl dropping variables (Allin Cottrell)
5. Re: autoreg formula and bessel function
(Riccardo (Jack) Lucchetti)
6. Re: gretl dropping variables (Riccardo (Jack) Lucchetti)
7. Re: gretl dropping variables (Allin Cottrell)
8. A time series question (Henrique)
----------------------------------------------------------------------
Message: 1
Date: Sun, 14 Jun 2009 13:03:08 -0400 (EDT)
From: Allin Cottrell <cottrell(a)wfu.edu>
Subject: Re: [Gretl-users] Sv: Gretl-users Digest, Vol 26, Issue 5
To: Gretl list <gretl-users(a)lists.wfu.edu>
Message-ID: <Pine.A41.4.58.0906141301320.696684(a)f1n11.sp2net.wfu.edu>
Content-Type: TEXT/PLAIN; charset=US-ASCII
On Sun, 14 Jun 2009 bulent.miran(a)ege.edu.tr wrote:
> I checked out the older versions and saw it begins with
heteroscedasticity
> corrected not WLS.
Ah, that's just a matter of different labeling. The
"hereroskedasticity corrected" model is in fact just WLS, with the
weights based on the estimated error variance.
Allin Cottrell
------------------------------
Message: 2
Date: Sun, 14 Jun 2009 20:13:43 +0300 (EEST)
From: bulent.miran(a)ege.edu.tr
Subject: [Gretl-users] Heteroskedasticty..
To: "Gretl list" <gretl-users(a)lists.wfu.edu>
Message-ID: <4562.88.245.228.68.1244999623.squirrel(a)posta.ege.edu.tr>
Content-Type: text/plain; charset="iso-8859-9"
Yes.. I also guessed so. However for beginners it may be a confusion. I
wonder if it is possible to label it with a little more explanation for
preventing confusions?
B Miran
EGE ▄N▌VERS▌TES▌
-----------------------------------------------------------------------
Bu elektronik posta ve onunla iletilen bⁿtⁿn dosyalar sadece
g÷ndericisi tarafindan almasi amaclanan yetkili gercek ya da tⁿzel
kisinin kullanimi icindir. Eger s÷z konusu yetkili alici
degilseniz bu elektronik postanin icerigini aciklamaniz,
kopyalamaniz, y÷nlendirmeniz ve kullanmanizkesinlikle yasaktir
ve bu elektronik postayi derhal silmeniz gerekmektedir.
EGE ▄N▌VERS▌TES▌ bu mesajin icerdigi bilgilerin do≡rulu≡u veya
eksiksiz oldugu konusunda herhangi bir garanti vermemektedir. Bu
nedenle bu bilgilerin ne sekilde olursa olsun iceriginden,
iletilmesinden,
alinmasindan ve saklanmasindan sorumlu degildir.
Bu mesajdaki g÷rⁿsler yalnizca g÷nderen kisiye aittir ve
EGE ▄N▌VERS▌TES▌'nin g÷rⁿslerini yansitmayabilir
-----------------------------------------------------------------------
This e-mail and any attachments may contain confidential and
privileged information. If you are not the intended recipient,
please notify the sender immediately by return e-mail, delete this
e-mail and destroy any copies. Any dissemination or use of this
information by a person other than the intended recipient is
unauthorized and may be illegal.
EGE UNIVERSITY makes no warranty as to the accuracy or completeness
of any information contained in this message and hereby excludes any
liability of any kind for the information contained therein or for
the information transmission, reception, storage or use of
such in any way whatsoever. The opinions expressed in this
message belong to sender alone and may not necessarily reflect the
opinions of EGE UNIVERSITY.
16 years
gretl dropping variables
by Allin Cottrell
A further response to Nicolai Striewe's question in
http://lists.wfu.edu/pipermail/gretl-users/2009-May/003346.html :
There are three circumstances in which gretl deliberately drops
regressors from a specification given by the user:
1. (all models) the variable is all zeros
2. (all models) the variable is exactly collinear with
other included regressors
3. (panel data, fixed effects estimator) the variable
is time-invariant
If a regressor is dropped when none of these conditions is
satisfied, that's a bug (unless I'm forgetting any other
legitimate conditions, though I don't think so).
So far as I know, we don't have any evidence of a bug of this sort
to date.
Allin Cottrell
16 years
Limited Number of Independent Variables - Random Effects - Panel Data
by Nicolai C. Striewe
Dear All,
I have to add to my recent request that the independent variables that I
want to add to the existing 54 independent variables are unit-invariant
(only time varying) macroeconomic variables. Only if they are unit-invariant
one of the time-dummies is omitted. It actually does not omit the time dummy
when I add a time-varying and unit-varying variable. My question is why is a
time dimension/time dummy omitted as one unit-invariant time-varying
variable is added?
Many thanks and best regards.
Nicolai
___________________________________
Nicolai Striewe
Research Assistant
Real Estate Management Institute
Aareal Professorship Real Estate Banking
Department of Finance, Accounting and Real Estate
European Business School (EBS)
Greater Frankfurt
GERMANY
-----Ursprüngliche Nachricht-----
Von: Nicolai C. Striewe [mailto:Striewe.ebs@rem-institute.org]
Gesendet: Mittwoch, 13. Mai 2009 14:55
An: 'gretl-users(a)lists.wfu.edu'
Betreff: Limited Number of Independent Variables - Random Effects - Panel
Data
Dear all.
May I pose the following problem with the random-effects panel data model
that I created:
The model consists of 54 independent variables (1 constant, 8 focus
variables, 5 control variables, and 40 time dummies). When I extend the
model by one focus variable it cuts off one time dummy. It seems to be a
restriction of gretl, that if more than 54 variables are specified in the
model, gretl leaves out all independent variables exceeding this limit
without a notification of omitting some variables.
Does gretl restrict the number of independent variables to 54? If yes, is
there a way around that restriction?
Many thanks in advance for your comments and thoughts.
Best regards,
Nicolai Striewe
___________________________________
Nicolai Striewe
Research Assistant
Real Estate Management Institute
Aareal Professorship Real Estate Banking
Department of Finance, Accounting and Real Estate
European Business School (EBS)
Greater Frankfurt
GERMANY
16 years
