gretl 2026a released
by Allin Cottrell
As usual, see https://gretl.sourceforge.net/
The biggest change is our update from the Mersenne Twister (SFMT) to
xoshiro256+ for pseudo-random number generation.
2026-01-30 version 2026a
- mat2list(): fix potential crashing bug; enhance by using
the t1 and t2 values on the matrix argument, if present
- binary logit/probit: use QR decomposition in tricky cases
where the Hessian may otherwise end up not p.d.
- "difftest" command: support the --split-by option for the
rank sum test
- "print" command: add --head (-h) and --tail (-t) options
as convenient abbreviations for the --range option
- "gnuplot" command, --y2axis option: accept a named list
as parameter, or comma-separated series
- "open" command: support the --all-cols option for xlsx
and gnumeric importation
- "data" command: fix possible problem with long db.nomics
series identifiers
- "xcorrgm" command: save a $result matrix and add a --quiet
option
- "quantreg" command: support $coeff and $stderr accessors
for the multiple-tau case
- "gnuplot" and "plot" commands with --band option: support
giving the center and width of the band as scalars if they
are not varying
- fix bug: possible crash from kdsmooth() when there's no
disturbance in the observation equation and exact diffuse
initialization is not employed
- fix bug in regls() with cross validation using MPI: it
could happen that too many MPI processes were employed,
leading to broken results
- fix bug in GUI deletion dialog: destroying the dialog
could be taken as indicating "OK"
- updates and fixes for documentation
- macos arm64 build: ensure that __divdc3 is defined
- 64-bit Windows build: ensure that libreadline is supplied
- random number generator: replace Mersenne Twister with
xoshiro256+
--
Allin Cottrell
Professor Emeritus of Economics
Wake Forest University, NC
1 month
new adalasso version
by Sven Schreiber
Dear all,
an update to the "adalasso" contributed function package by myself has
been released recently, version 0.2. It implements the adaptive Lasso.
There were two reasons: First there was a bug where it was actually
impossible to choose plain OLS for the first weighting stage. (The
default there is a ridge estimator.) Secondly, there turned out to be a
little gretl bug in conjunction with the parallel computing layer MPI
(fixed in forthcoming gretl 2026a) which could be triggered by the
package. This could lead to wrong results ('nan' values for some
standard errors), now there is a workaround in place in the package, for
older gretl versions up to the current 2025c.
As a very small new feature, you can now directly get the full
coefficient vector for the "best" model (optimal tuning parameter),
including those elements that were set to zero. (Usually you just get
the subset of the non-zero coefficients, or the full matrix for all models.)
Don't hesitate to report any bugs that you come across (in this or other
packages)!
As usual, the new package version is available for download from within
the gretl program.
cheers
sven
1 month, 1 week
'obs' function parameter type
by Sven Schreiber
Hi,
at the beginning of ch. 14 of the user guide there's the following
footnote: "An additional parameter type is available for GUI use, namely
obs; this is equivalent to int except for the way it is represented in
the graphical interface for calling a function."
Even though I have authored various function packages, I wasn't aware of
this. Can anybody point me to a package that uses this facility?
thanks
sven
1 month, 2 weeks
Macroeconomic model using Gretl
by Carlos Andrade
Dear Gretl users,
How to build a macroeconomic model using Gretl?
Atenciosamente,
Carlos Antonio Soares de Andrade
Economista e Consultor de Empresas
Telefone: (83) 9 98117113
1 month, 3 weeks
gretl in a tablet?
by Cristián Arturo Ducoing Ruiz
Dear Gretl users,
I have a student without access to a laptop and she wants to install gretl
in a tablet (Android, Lenovo TAB K11)
Thanks!
--
Atte.
Cristián Arturo Ducoing Ruiz
1 month, 3 weeks
Package Updates (November/December 2025)
by Riccardo (Jack) Lucchetti
Dear all,
this message is to inform the community about the activity in our
function package repository: during the months of November and December
2025, 4 packages were updated to a new version and a new one was
released. (I'm putting two months together because last month I just
forgot, hehehe)
The updates were:
"basis_functions", by myself (splines, fractional polynomials, Fourier
series etc)
"CSDpanel.zip, by Sven Schreiber and Jörg Breitung (test and estimators
for panels with cross-section dependence)
"season_plot", by Artur Tarassow (Buys Ballot plots)
"hausman_taylor", by Allin Cottrell (Hausman-Taylor panel model)
And some stats on 2025: we had over 24,000 downloads for our function
packages. The top ten:
downloads package name description
747 lpmfx logit/probit marginal effects
661 getYahoo To download financial data from Yahoo Finance
616 yahoo_get Downloader of daily financial data from Yahoo
514 auto_arima Return best SARIMA(X) model according to information
criteria value
512 heatmap Heatmap, contour and 3D plots
507 FEP Forecast Evaluation Package
498 ADMBP ARDL Dynamic Multiplier Bootstrap Package
448 SB Stationary bootstrap resampling of time series
427 PPtest Phillips-Perron unit root test
404 adalasso adaptive Lasso
Kudos to the contributors!
-------------------------------------------------------
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, 3 weeks
