system estimation on large cross-section problem under win XP
by Franck Nadaud

Hi listers, greetings from Paris !
Well, I checked on my system and The loop runs ok with the ols command but not
with the system commands.
however, will the loop runs, gretl tells me there is a near singularity in my
data.
I try to check the script by calling a simpler system, with the strict minimal
input : the prices and income. I will then attempt to increase complexity to
check the behavior of GRETL.
Note also that I spotted some problems with the data, serious enough to cause
to regress on singular matrices (duplicated or zero columns) for some of the
loop index values. I fixed that but the problem persists.
I check to see if it is the system command that pose problems. I tell you back
when i have anything new
thanx & regards
cheers
Franck
--
Franck Nadaud
CIRED
UMR 8568 CNRS - EHESS, ENPC, ENGREF, CIRAD
45 bis avenue de la Belle Gabrielle
94736 Nogent-sur-Marne Cedex
TEL: 33-1-43-94-73-94
FAX: 33-1-43-94-73-70
MOB: 06-07-39-92-75
France
13 years, 12 months

tsls tests
by Summers, Peter

Dear all,
I've got a couple questions regarding tsls and the automatic specification tests gretl does.
First, in a model with k (potentially) endogenous regressors, gretl seems to use k-1 degrees of freedom for the Hausman test rather than k. My impression is that gretl uses the "variable addition" form (eg, Greene p. 323), is that right? If so, shouldn't the the df be k?
The specific model I'm using is from Verbeek's "schooling" data set. Log wage is the dependent variable, with "black", "south76" and "smsa76" as exogenous regressors; education, experience and its square ("ed76", "exp76", "exp762") are instrumented regressors. The instruments are age, its square, and a "near college" dummy (plus the 3 exogenous x's). If I "align" the regressor and instrument lists like this:
Model 21: TSLS, using observations 1-3010
Dependent variable: lwage76
Instrumented: ed76 exp76 exp762
Instruments: const black smsa76 south76 nearc4a age76 sq_age76
then I don't get the weak instruments test. But if I switch the order to this:
Instruments: const black smsa76 south76 age76 sq_age76 nearc4a
I get the test, but with all 0's for the critical values:
Weak instrument test -
Cragg-Donald minimum eigenvalue = 1.53611
Critical values for TSLS bias relative to OLS:
bias 5% 10% 20% 30%
value 0.00 0.00 0.00 0.00
Similar things happen with over-identified models (although in that case I get non-zero critical values). I don't know as much about these tests as I should, but does the ordering of the instruments really matter?
Thanks in advance for any help!
PS
===============================
Dr. Peter Summers
Assistant Professor
Department of Economics
Texas Tech University
===============================
13 years, 12 months

Eviews model to GRETL system
by Erhan Azrai

Hello
I am having trouble to convert a piece eviews code below to gretl
system. The code is used to solve an OLG model of 8 simultaneous
equations. Can it even be done? Thank you.
model mv20den
'equations for the model = 8 simultaneous equations
'Equation 1 : gamma_EA ~human capital accumulation
mv20den.append gamma_EA = (A_9 * (z_tilde ^ (-nu_3))) - 1
'Equation 2 : gamma_M ~growth accumulation
mv20den.append gamma_M = A_12 * (z_tilde ^ (1 -phi_2) ) * chi_tilde * phi_tilde
'Equation 3 : z2 ~
mv20den.append z2 = A_12 * (z_tilde ^ (1 -phi_2) ) * chi_tilde * phi_tilde
.
.
.
'Solve the model
mv20den.solveopt(o=g)
mv20den.solveopt(z=n)
solve mv20den
Best,
Erhan
13 years, 12 months

Panel corrected standard errors
by Kristyna Pokorna

Hi,
I'd need to use Prais-Winsten model (Panel Corrected Standard Errors, PSCE) as in my panel data set there are only few countries included compared to the length of the time series. Therefore the model preset in gretl (model described by Arellano and Bond) is not appropriate. Please, is it possible to invoke the PSCE estimator in gretl? If it is possible, please could you give some instructions what should be done?
Thanks in advance for help!
Best regards,
Kristyna
13 years, 12 months

Re: [Gretl-users] Gretl-users Digest, Vol 39, Issue 22
by Stefan Rehm

> Message: 4
> Date: Tue, 20 Apr 2010 09:22:45 +0200 (CEST)
> From: "Riccardo (Jack) Lucchetti" <r.lucchetti(a)univpm.it>
> Subject: Re: [Gretl-users] Postgresql ODBC connection
> To: Gretl list <gretl-users(a)lists.wfu.edu>
> Message-ID: <alpine.DEB.2.00.1004200918580.19309(a)ec-4.econ.univpm.it>
> Content-Type: text/plain; charset="iso-8859-15"
>
> On Mon, 19 Apr 2010, Stefan Rehm wrote:
>
> > Hi,
> > i want to extract data from an existing PostgreSQL database on
> > localhost. I tested the ODBC connection with OpenOffice and it works
> > nice after entering the database user and password.
> > But i'm still confused to extract the data with Gretl. I connected via
> > the gretl console to the database named eix with
> >
> > open dsn=eix user=databaseuser password=password --odbc
> >
> > and there are no output from the gretl console. how can i access the
> > data?
>
> You're halfway through. The "open" command establishes the connection.
> The actual data retrieval is a job for the "data" command, which stores
> the output of an SQL query. Example:
>
> <script>
> qry = "select foo from bar"
> data myseries query=qry --odbc
> </script>
Hmm, no chance. There is no message after the open command. So i'm not
shure if everything work fine. To retrieve the data i use this workflow:
open dsn='Postgresql eix' user=postgres password=xxx --odbc
[no message in gretl console, then i tried:]
data x query="SELECT id FROM person" --odbc
Can somebody help me out?
Best Regards
Stefan
13 years, 12 months

system estimation on large cross-section problem under win XP
by Franck Nadaud

Dear gretl listers, greetings from Paris !
I am currently doing some system estimation with a large cross section under
win XP on GRETL 1.8.7 (GUI). My problem is the following:
I have a big data base (N = 49.808 obs.) declared in cross section format.
Actually, it is a pooling of (small) parts of five households budget and
expanditures surveys for years 1979, 1984, 1989, 1995 and 2001.
I coded a survey number variable (s = 1..5) for each year (79, 84, .. 01)
declared it as a discrete variable.
My objective is to estimate a demand system for each year from a script were I
call the data for each s = 1..5 with commands like :
smpl SURVNUM=1 --restrict --replace
system name=AIDS
equation WLOG_W ... variables
equation WCAR_W ... variables
end system
restrict AIDS
b[1,3]-b[2,2]=0
end restrict
estimate AIDS method=ols --iterate
however when i run this small segment, gretl crashes and i dont see why.
I began by writing a loop on the variable SURVNUM, like the following:
loop i = 1..5
smpl SURVNUM=$i --restrict --replace
... commands
end loop
smpl full
but when i saw it does not run properly I decided to try the slow way (ie: to
sample manually), but even then it does not work.
Another remark: in my program, after the system estimation, there is a lot of
matrix / series computations but I am not sure if GRETL can handle matrices
indexing under smpl commands. If not I will go the dirty way : split may base
into 5 smaller bases and estimate the scripts opening each of them in
sequence. But before commiting such a sin, I would like to verify if the case
is desperate or not !
cheers folks & regards
Franck
--
Franck Nadaud
CIRED
UMR 8568 CNRS - EHESS, ENPC, ENGREF, CIRAD
45 bis avenue de la Belle Gabrielle
94736 Nogent-sur-Marne Cedex
TEL: 33-1-43-94-73-94
FAX: 33-1-43-94-73-70
MOB: 06-07-39-92-75
France
13 years, 12 months

off topic - scholarship
by Kehl Dániel

Dear list-members,
sorry for this off-topic e-mail, I hope it is not a big problem for any
of you.
I am a PhD student and also teaching statistics at the University of Pécs.
I am close to win a scholarship of 1 year to a US university.
All I need is a receiver professor and/or department in the US, who is
ready to do some paperwork (invitation letter, things like that).
In the US I want to take part at as many stat courses I can and of
course to take part in the research activities if possible.
The topics I am interrested in (but not an expert yet) are: categorical
data analysis, GLM, nonparametrics, multivariate analysis.
My only contact to any university people in the US is this mailing list,
thats why I wrote this short e-mail.
Please if anyone could help or give me tips in this topic feel free to
write.
Sorry for disturbing again (hope you won't drop me from this useful list):
Daniel,
Hungary
13 years, 12 months

Postgresql ODBC connection
by Stefan Rehm

Hi,
i want to extract data from an existing PostgreSQL database on
localhost. I tested the ODBC connection with OpenOffice and it works
nice after entering the database user and password.
But i'm still confused to extract the data with Gretl. I connected via
the gretl console to the database named eix with
open dsn=eix user=databaseuser password=password --odbc
and there are no output from the gretl console. how can i access the
data?
Best Regards
Stefan
13 years, 12 months

Corrgm command from Console
by Matteo Chinazzi

Hello,
quick question, how do I get the Gnuplot graph of ACF and PACF using the
corrgm command from console?
Because I am able to get only the ASCII plot, while I would rather need
to plot (and save) the gnuplot graph of corrgm from console (either
within gretl_x11 or within gretlcli environments..).
I have tried to look at the source code (describe.c file), but I am
still not sure on how to proceed..
Thanks for your help!
Matteo
13 years, 12 months

Re: [Gretl-users] two simple time series questions
by robert pisani

Charles --
Actually, I've received quite a lot of helpful response from another
group, some of which are leading to the solution I sought. If you're
interested, I can send a more mathematical description with graphs,
etc., of the various time series in question, but it appeared to me
that this was not the place for such. Please let me know. And
thanks for your suggestions.
Robert
BTW, it's Kolmogorov.
At 04:22 AM 4/19/2010, you wrote:
>Since nobody else has suggested anything I suggest a few points. (
>The lack of response is not altogether surprising since you don't
>define your objective with much scientific clarity.)
>
>One issue is whether you want to test some properties of the series
>or whether you want to test for similarity in the sense of correlation.
>If the first, then one would have thought you would want to measure
>the autocorrelations and partial autocorrelations and then compare
>them with the mean in some arbitrary way.
>If the second then you might consider cumulating the observations
>and using something like the Kolmorgorov-type tests to see whether
>the series are similar.
>
>Alternatively and very crudely you might take first differences and
>calculate the Baumol efficiency criterion which = mean - k *
>standard deviation (k being arbittrarily chosen to reflect a
>weighting or trade-off between the "risingness" reflected in the
>mean and the unsmoothness reflected in the standard deviation. This
>is an old measure suggested to measure investment performance but it
>might get you started even though it ignores the statistical
>properties of the time series.
>
>Charles Ward
>
>
>(1)
>
>On 16 April 2010 17:09, robert pisani
><<mailto:r.pisani@mac.com>r.pisani(a)mac.com> wrote:
>
>Two simple time series questions:
>
>1- I want a measure of a time series that describes how closely it
>resembles a given collection of time series. Thus, the time series
>collection may be contained in column 2 to n of an Excel file, with
>each column containing one of the collection, with the time series to
>be measured in column 1.
>
>2- Define a measure of a time series that measures how close it is to
>a smoothly rising (but not necessarily monotonically rising)
>series. Thus using such a measure, one could rank a collection of
>time series so that those most smoothly rising and fastest rising
>rank highest/lowest using the measure. Thus the time series to be
>measured may reside in column 1 of an Excel spread sheet, with the
>second column being the numbers 1, 2, 3, . . . etc.
>
>The proximity of the given series to the collection or to "smooth
>rising" should be measured in some rms sense. I realize that these
>requirements are not completely defined. I'm looking for ideas.
>
>
>_______________________________________________
>Gretl-users mailing list
><mailto:Gretl-users@lists.wfu.edu>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
13 years, 12 months