bivariate probit and the cluster option
by artur.bala.tn
Dear all,
In a bivariate probit estimation I passed the "cluster" option refering, by mistake, to a non-existing variable. Obviously, gretls doesn't take into account the cluster command as long as the cluster variable doesn't exist. But the tricky stuff is that the output estimation changed (z and p-value) with respect to a previous non-cluster estimation with same independent variables.
For replication, try to run the following script from gretl's practice files adding a "fake" cluster variable.
Best,
Artur
open greene25_1.gdt
# regressors for first equation
list x1 = const age avgexp
# regressors for second equation
list x2 = const age income ownrent selfempl
# no-clustered estimation
biprobit anydrg cardhldr x1 ; x2
# faked clustered estimation
biprobit anydrg cardhldr x1 ; x2 --cluster=noclustervariable
10 years
Error message for tramolin
by Tim Nall
Following instructions on this list, I went to files-->Function files -->
On local machine. This popped up a set of (unchecked) check boxes. I
checked all of them, then closed that popup. I went to the menus, found
nothing. Went back to "On local machine" and double-clicked the "tramolin"
option. Aha, the desired function starts. i selected "BOTH" as the
(existing) series name and "BOTH_tra" as the "save to" name. Clicked OK,
got error message (below). Went back to the tramolin popup and clicked
"Help"; the only thing that suggests possible problems is that you must
have TRAMO installed. I do in fact have TRAMO installed.
[Error message below]
The shell command is not activated.
*** error in function tramolin, line 56
> ! del /Q *.t && cd C:\Documents and Settings\teacher\Application
Data\gretl\tramo && "C:\Program Files\tramo\tramo.exe" -i sername -k
sername.ser && cd C:\Documents and Settings\teacher\Application
Data\gretl\tramo\graph\series && echo "ser_adj" > sername.txt
Thanks TMN
10 years
File export options
by oleg_komashko＠ukr.net
Menu text "Export to GNU R" should be changed to "Export to text", extension from ".R" to ".txt". They are text files, not .RData fises. ".R" extension is for R scripts. Everything works correctly but labels are misleading. One should guess to import file.R in R as a text data file.
10 years
a small suggestion (actually two)
by Tim Nall
I was happy when this list told me about Files --> Function files -->
On server. It contains several things I'd been reading about, but
couldn't find on gretl. I have two small suggestions that are purely
for convenience:
1) I have installed 4 files, but can only find 1 of them on the
menus... And I installed tramolin then found TRAMO downloadable
elsewhere (?)... Not to kick a dead horse, but in the table that is
displayed, a new column that shows menu path where the downloaded file
can be found would be very nice . [At this moment I am looking for the
best way to filter data.. on this list i have already posted about
several problems that zeroes are causing me.]
2) probably too late for this suggestion: Me personally, I would
*never* have looked under the "Files" menu for these. In fact, I
didn't. I did however look under "Tools", which to me would be the
logical place. I wouldn't think: "Mmmm, no ridge regression, should I
look for it under Files?" i would think to look under Tools. But the
first suggestion may be more doable, since the second is a larger
change. Thanks TMN
On Sun, Apr 27, 2014 at 11:24 PM, Riccardo (Jack) Lucchetti
<r.lucchetti(a)univpm.it> wrote:
> On Sun, 27 Apr 2014, Andreï | Андрей Викторович wrote:
>
>> Dear Riccardo,
>>
>> The filter function is quite useful. There is a slight misprint in the PDF
>> Command Reference: the second sum starts from {i=i}; of course, it should
>> say {i=1}.
>
>
> Of course. It's now fixed in CVS, thanks.
>
> -------------------------------------------------------
> Riccardo (Jack) Lucchetti
> Dipartimento di Scienze Economiche e Sociali (DiSES)
>
> Università Politecnica delle Marche
> (formerly known as Università di Ancona)
>
> r.lucchetti(a)univpm.it
> http://www2.econ.univpm.it/servizi/hpp/lucchetti
> -------------------------------------------------------
> _______________________________________________
> Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
--
Best regards,
Timothy M. Nall
Assistant Professor
National Quemoy University
Kinmen, Taiwan
10 years
Question: rho and DW stat don't appear
by Tim Nall
May I ask, what does it mean if you take the log difference of both the
dependent and independent variables in an OLS, and then rho and DW stat
don't even appear in the output? The White's test value is quite small...
if i swap dependent/independent, the rho and DW stat reappear... Thanks TMN
10 years
quickly run through ARIMA models, varying parameters, acquiring AIC, BIC, HQC?
by Tim Nall
All,
Please forgive my simple questions. For ARIMA modelling, using the ACF
and PACF to determine the pdq parameters requires subjective judgment
based on experience. Answers are seldom clear-cut. As the subject
header says, can a somewhat more objective path be found in quickly
running through ARIMA models, varying parameters, acquiring AIC, BIC,
HQC, and comparing? I don't think the time-series-->VAR lag selection
option gets at precisely the same thing. Thank you TMN
10 years
Problems with regression model forecast into future with 2 or 3 observations (T+1, T+2, T+3)
by Ondřej Dvouletý
Dear All,
I would like to ask you for help with Gretl software. I am writing you with
an problem which I face during making forecast / prediction. I estimated
econometric model with one depended variable and four explanatory
variables. Method of estimating was OLS or Cochrane - Orcutt (both with
same result). When I try to make a forecast into the future (*and Yes, I
added new observations into the new data set*) I am unable to forecast into
the future.
*Gretl notices:* There are no observations available for forecasting out of
sample. If you wish, you can add observations (Data menu, Edit data), or
you can shorten the sample range over which the model is estimated (Sample
menu).
*The problem is that I have vacant places for forecast but the Gretl does
not make it. When I put into the empty places any values (without sense for
example 1, 2...) The forecast is made. *
Do you have any procedure how to fix the problem? I tried this several
times and there was no way how to do that. How can I make forecast without
entering nonsense values just when I add new empty observations and I want
to know forecasted values.
Thank you very much for your reply.
*Ondrej Dvoulety*
Mobil.: +420 728 431 027
e-mail: ondradvoulety(a)gmail.com
Skype: ondradvoulety
[image: ICON] <http://cz.linkedin.com/in/ondrejdvoulety/>
Studeneves 98
273 79 TURANY U SLANEHO
10 years
gretl needs updating export of data?
by oleg_komashko＠ukr.net
This is how Gretl exports to csv Create text file (tab between columns) year gdp 2000 8010672 2001 8070670.31 2002 8198842.23 2003 8421918.34 2004 8776525.02 2005 9089701.34 2006 9413052.38 2007 9709513.84 2008 9674648.114 2009 9333377.657 2010 9615405.926 2011 9782251.723 2012 9974404.718 import in gretl run logs gdp genr gdp10=gdp/10 genr gdp100=gdp/100 genr gdp1000=gdp/1000 genr gdp10000=gdp/10000
export to CSV (top lines)
obs gdp l_gdp gdp10 gdp100 gdp1000 gdp10000 2000 8010672.000000000000 15.8962852107 801067.199999999953 80106.720000000001 8010.672000000000 801.0672000 2001 8070670.309999999590 15.9037470988 807067.030999999959 80706.703099999999 8070.670310000000 807.0670310 (heap of junk in last dig.) (func. 12 dig. +(.)) (no junk digits)
For example, my problem with R3.1 was caused not by large numbers, but by large number of junk digits in gdp May be it would be better to set 16 symbols for a number (15+1) and set "e" format for numbers greater than some N, say 10^6?
10 years
Re: [Gretl-users] armax-auto option occasionally generates this error
by Tim Nall
This one likes to crash (attached as text file). I wonder, is armax_auto
aware of the sample range I selected via GUI? Not that it matters to me,
but it may be important.
TMN
On Sat, Apr 26, 2014 at 8:15 PM, <oleg_komashko(a)ukr.net> wrote:
> Can you supply the data set? As attachment, may be.
>
> --- Оригінальне повідомлення ---
> Від кого: "Tim Nall" <tnall.ling(a)gmail.com>
> Дата: 26 квітня 2014, 13:19:16
>
> Thanks to all for the replies. Will try to find TRAMO. meanwhile, the
> armax-auto option occasionally generates this error:
>
> ? armax(4, 4, BOTH, null, 1, 1, 0, 1, 0)
> BFGS: initial value of objective function is not finite
> *** error in function armax, line 82
> > loop q = 0..maxq
>
> On Sat, Apr 26, 2014 at 4:46 PM, Riccardo (Jack) Lucchetti
> <r.lucchetti(a)univpm.it> wrote:
> > On Sat, 26 Apr 2014, oleg_komashko(a)ukr.net wrote:
> >
> >> You can use armax_auto (see time series menu in Gretl GUI). The function
> >> has a drawback: the cycle crushes on encountering a unit root in MA part of
> >> the model Also you can use auto.arima{forecast} in R My experience tells it
> >> would be better to determine the order of integration (use unit root tests)
> >> beforehand and use, say "d=1" in auto.arima (type ?auto.arima in R console)
> >
> >
> > That's a function package, you may need to install it first: go to
> > Files>Function Files>On server.
> >
> > -------------------------------------------------------
> > Riccardo (Jack) Lucchetti
> > Dipartimento di Scienze Economiche e Sociali (DiSES)
> >
> > Università Politecnica delle Marche
> > (formerly known as Università di Ancona)
> >
> > r.lucchetti(a)univpm.it
> > http://www2.econ.univpm.it/servizi/hpp/lucchetti
> > -------------------------------------------------------
> > _______________________________________________
> > Gretl-users mailing list
> > Gretl-users(a)lists.wfu.edu
> > http://lists.wfu.edu/mailman/listinfo/gretl-users
>
>
>
> --
> Best regards,
> Timothy M. Nall
> Assistant Professor
> National Quemoy University
> Kinmen, Taiwan
>
> _______________________________________________
> Gretl-users mailing listGretl-users@lists.wfu.eduhttp://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
>
--
Best regards,
Timothy M. Nall
Assistant Professor
National Quemoy University
Kinmen, Taiwan
10 years
Drastic discrepancy in performance
by Andreï | Андрей Викторович
Dear developers and users,
I noticed that an attempt to address the cycle variable heavily depends on
the use of dollar sign. Consider the following file:
<hansl> #Code for DFtaumu.inp
# Dickey---Fuller's tau_mu distribution
set stopwatch
nulldata 10000
scalar T = 100
scalar tranches=10
scalar iters=10000
matrix y = zeros(T,1)
matrix DFtaumu = zeros(iters,1)
matrix RES
matrix VARB
loop j=1..tranches --quiet
loop i=1..iters --quiet
y = cum(mnormal(T, 1))+5*j # <--- That is the one!
DFtaumu[i] = (mols(y[3:], ones(T-2,1)~y[2:T-1], &RES,
&VARB)[2]-1)/VARB[2,2]^0.5
endloop
sprintf fn "%02d", $j
mwrite(DFtaumu, "DFTM-(a)fn.mat")
printf "Tranche %d of %d: %f s elapsed\n", $j, tranches, $stopwatch
endloop
</hansl>
I am studying the influence of the order of magnitude of the constant on
the DF \tau_\mu distribution. The numbers are relatively small for
demonstrative purposes. When I run it through gretlcli or gretlcli-mpi, the
average running time is 0.2 s per loop of *j* (one tranche).
gretlcli-mpi DFtaumu.inp
However, it seems more correct to use the *$j* instead of just *j* in the
loop, so I modified the 14th line by adding a $:
y = cum(mnormal(T, 1))+5*$j
It seems more correct, yet the performance fell abruptly, the average
running time increased by 50%. When the sample size is increased to
*nulldata 100000* and *scalar iters=100000*, the average execution time for
one tranche increases from 2.01 (without $ sign) to 2.85 seconds (with $
sign).
Why does such a slight change cause such tremendous fall in performance
with seemingly no effect on the properties of the output (I compared the
distributions of DFtaumu and found no difference). Why? What kind of
further optimisation and improvement should be undertaken in order to avoid
such pitfalls?
Thank you very much in advance!
--
Yours sincerely, | С уважением,
Andreï V. Kostyrka. | Андрей Викторович Костырка.
http://kostyrka.ru, http://kostyrka.ru/blog
10 years
