October 2026 - Gretl-users - gretlml.univpm.it

grel 2026d released

by Allin Cottrell

See https://gretl.sourceforge.net/ This is primarily a bug-fix release. Shortly after putting out gretl 2026c we noticed a nasty newly-introduced bug -- a crash triggered as follows: estimate a VAR via the gretl GUI, pull down the Analysis menu in the VAR output window, select Forecast, then select a specific series. Here's the change log: 2026-10-29 version 2026d - Fix bug: potential crash when forecasting based on a VAR estimated in the GUI - Fix bug: the "must-assign" attribute of packaged functions was not being registered in the function-package GUI - Fix bug: potential crash in the deseas() function on getting an error from X-13ARIMA-SEATS - Fix bug: missing data in the updated AWM macro dataset - L-BFGS-B: the code (originally translated from Fortran via f2c) has been rewritten in idiomatic C - L-BFGS-B: more robust handling of a criterion that is undefined at some points: back off on encountering a non-finite criterion or gradient, rather than terminating - L-BFGS-B: never evaluate the criterion outside of the specified bounds (one-sided numerical derivatives at a bound; guard against rounding error in the line search) - fix bug: BFGSmin, BFGSmax, BFGScmin, BFGScmax, NRmin and NRmax could return a criterion value that did not correspond exactly to the final parameter vector - Delete some unused variables flagged by clang - markdown formatter: handle bulleted lists not preceded by a blank line - macOS: add installer packages for gretl_edit - PairPlot addon: upgrade to PDF documentation -- Allin Cottrell Professor Emeritus of Economics Wake Forest University, NC

2 weeks, 1 day

1
0
0 / 0

event study example

by Theodoros Panagiotidis

Dear All, Does anybody have an event study code and an accompanying dataset that one could use in teaching? regards Theo -- Theodore Panagiotidis Department of Economics University of Macedonia Thessaloniki, Greece ResearchGate <https://www.researchgate.net/profile/Theodore_Panagiotidis> Econpapers <https://econpapers.repec.org/RAS/ppa1363.htm> Linkedin <https://www.linkedin.com/in/theodore-panagiotidis-1371b3169/>

2 weeks, 5 days

3
5
0 / 0

PREVIOUS ISSUE WITH PANEL DATA

by Brian Revell

Apologies checked my data and the dep variable had somehow been corrupted. No problem after replacing it with uncorrupted version). *Brian Revell*

1 month

1
0
0 / 0

pANEL DATA ISSUES WITH NEW VERSION OF GRETL

by Brian Revell

I have tried to run both a fixed effects and dynamic panel data model with 5 panel members and 20 observations per panel and 3 independent variables The fixed effects says *missing data* and stops (it was never a problem before) and the dynamic panel estimation specification say *out of memory *(never happened before). Suggestions please *Brian Revell*

1 month

1
0
0 / 0

gretl 2026c released

by Allin Cottrell

As usual, see https://gretl.sourceforge.net 2026-10-12 version 2026c - new addons kmeans and PairPlot, from A. Tarassow's function packages of the same names, with various enhancements and speedups - new function kmeans(): algorithm ASA136 for the k-means problem - new function permute(): calculate all permutations of a vector of up to 8 elements - "fcast" command: introduce the --exponentiate and --integrate options for forecasting levels from a model in logarithms and/or first differences - matrix code: drop the experimental hand-crafted OpenMP paths for multiplication, X'X and element-wise add/subtract, relying on the BLAS and native single-threaded code; lower the default blas_mnk_min threshold to 80000 accordingly - extra addon: new function convex_hull() - the "omp_mnk_min" set-variable is now an obsolete no-op - GUI: promote the "Packages" entry to top level - GUI: forecast from single-equation model, support write and copy in CSV format - GUI: fix for call to function package with no arguments - GUI: fix for the options shown in connection with making a forecast from a model - use of "to" in a range-of-lags expression: support this in a wider range of contexts - recession bars in time-series plots: extend support of this feature to daily data - enhance functionality of ui-maker for function packages - fix bug: GUI plot editor breaks factorized plots - fix bug: crash in GUI data editor on Windows - fix bug: correlation matrix heatmap broken if the graph-name of an included series contains '%' - fix bug: crash in adf test on panel data when the --difference option is given - fix bug: potentially incorrect output from xtab when given string-valued series as input - fix bug: potential crash or memory leak with the "smpl" command in certain cases of compounding restrictions - fix bug: possible crash in the "bds" command under certain conditions - gretl_edit: support loading of function definitions and emacs-style "jump to declaration" - MS Windows: update doc regarding use of MS-MPI -- Allin Cottrell Professor Emeritus of Economics Wake Forest University, NC

1 month

1
0
0 / 0

Re: yahoo_get error

by Theodoros Panagiotidis

It works smoothly! Thanks! On Sun, 11 Oct 2026 at 15:24, RICCARDO LUCCHETTI < r.lucchetti(a)staff.univpm.it> wrote: > The new version of the paxkage (3.10) is now available for download. The > yahoo_price() function should work correctly in weekly datasets too. > > Testing would be appreciated! 🙂 > ------------------------------ > *From:* Riccardo (Jack) Lucchetti <p002264(a)staff.univpm.it> > *Sent:* Friday, 09 October 2026 01:32:40 > *To:* gretl-users(a)gretlml.univpm.it <gretl-users(a)gretlml.univpm.it> > *Subject:* [Gretl-users] Re: yahoo_get error > > On 08/10/2026 22:07, Cottrell, Allin wrote: > > On Thu, Oct 8, 2026 at 3:18 PM Theodoros Panagiotidis <tpanag(a)uom.edu.gr> > wrote: > >> It is weekly > > OK, so it's a bug in yahoo_get. Either the conditionality for > > referencing 'amth' should be ($pd == 4 || $pd ==12) -- not that the > > data are not daily -- or else, if the weekly frequency is not > > supported, that should be made explicit. > > I'm working on a revised version of the package that should handle weekly > datasets correctly. > > ------------------------------------------------------- > 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)gretlml.univpm.it > To unsubscribe send an email to gretl-users-leave(a)gretlml.univpm.it > Website: > https://gretlml.univpm.it/postorius/lists/gretl-users.gretlml.univpm.it/ > _______________________________________________ > Gretl-users mailing list -- gretl-users(a)gretlml.univpm.it > To unsubscribe send an email to gretl-users-leave(a)gretlml.univpm.it > Website: > https://gretlml.univpm.it/postorius/lists/gretl-users.gretlml.univpm.it/ > -- Theodore Panagiotidis University of Macedonia Thessaloniki, Greece & Research Associate HO, LSE <https://www.lse.ac.uk/Hellenic-Observatory> Linkedin <https://www.linkedin.com/in/theodore-panagiotidis-1371b3169/> GooglePage <https://sites.google.com/view/prof-theodore-panagiotidis/bio?authuser=0>

1 month

1
0
0 / 0

yahoo_get error

by Theodoros Panagiotidis

? Tesla = yahoo_price("TSLA", 0, 0) TSLA imported ok > desc_s ~= ", " ~ amth The symbol 'amth' is undefined Data error *** error in function yahoo_price, line 35 > desc_s ~= ", " ~ amth -- Theodore Panagiotidis University of Macedonia Thessaloniki, Greece & Research Associate HO, LSE <https://www.lse.ac.uk/Hellenic-Observatory> Linkedin <https://www.linkedin.com/in/theodore-panagiotidis-1371b3169/> GooglePage <https://sites.google.com/view/prof-theodore-panagiotidis/bio?authuser=0> <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_camp...> Virus-free.www.avast.com <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_camp...> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

1 month

4
7
0 / 0

How to get L2 and CWC variables quickly

by Ramki S

When working with panel data, I need to create the cwc (centered within cluster) variables quickly. Although there is "add time mean.." will do exaclty this job in creating L2 variable, I have to do one more step to create the cwc variable. Can we have some function to do the both? I am using the terminology used in multi-level models. series Q_output_L2 = pmean(Q_output) series Q_output_cwc = Q_output - Q_output_L2 series PF_fuelprice_L2 = pmean(PF_fuelprice) series PF_fuelprice_cwc = PF_fuelprice - PF_fuelprice_L2 series LF_loadfactor_L2 = pmean(LF_loadfactor) series LF_loadfactor_cwc = LF_loadfactor - LF_loadfactor_L2

1 month

3
2
0 / 0

Error using winters (wes) from fcModels

by ramakrishnasalagrama＠gmail.com

? List04 = wes(Cement, 0, null, 1, 0.4, 0.03, 0.3, 2, 0) Index value -1 is out of bounds *** error within loop in function wes > genr f[j+1]=(F[j] + m*T[j])*S[j+m-$pd] I am getting the above error while trying to estimate the holt-winters from fcModels. Any help? https://www.dropbox.com/scl/fi/nklcilb87qdtc2dm1klr6/aus_production_cemen...

1 month

4
11
0 / 0

Package updates (September 2026)

by Riccardo (Jack) Lucchetti

Dear all, this message is to inform the community about the activity in our function package repository. During the month of September 2026, one package was updated and a new one was introduced. The new package is "nic_gig", by Mary Mertzanidou and Theodore Panagiotidis. This package is used jointly with the "gig" addon (univariate GARCH models), and lets you plot the "News Impact Curve" for a GARCH model and a GJR model. The updated package is "BreitungCandelonTest" (for frequency-wise Granger causality tests), by Sven Schreiber, which is mostly a bugfix. As usual, download and enjoy! ------------------------------------------------------- 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 -------------------------------------------------------

1 month, 1 week

1
0
0 / 0

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Gretl-users October 2026