BACE Econometric Package Has Been Released
Researchers Marcin Błażejowski and Jacek Kwiatkowski launched a new Bayesian averaging tool for the gretl platform.
Updated on Sept. 25, 2026 in Economics — General
Economists Marcin Błażejowski and Jacek Kwiatkowski have released the BACE software package for the gretl environment. The tool facilitates Bayesian averaging of classical estimates to improve statistical modeling.
Why it matters
The package addresses critical issues in statistical modeling by managing model uncertainty and improving computational speed. It provides researchers with an integrated way to combine classical and Bayesian estimation techniques.
The BACE package supports a diverse range of models, including logit, probit, ordered logit, ordered probit, multinomial logistic, Poisson, Tobit, interval, and linear regression. The software utilizes Bayesian information criterion variants.
The players
Marcin Błażejowski
He is an economist and co-developer of the BACE software package.
Jacek Kwiatkowski
He is an economist and co-developer of the BACE software package.
The details
Developed for the gretl platform, the package implements a model-building strategy that uses Bayesian techniques alongside classical estimation. It additionally provides users with specific measures of jointness to interpret data relationships.
Timeline
September 25, 2026: The BACE package was officially published.
The Big Picture
The introduction of BACE follows the ongoing development of the gretl econometric analysis environment as a standard for open-source statistical software. It extends the library of available tools for researchers using the platform for advanced statistical analysis.
Researchers and students currently using gretl can now incorporate Bayesian averaging into their existing workflows without needing to switch to proprietary software. This update simplifies the transition from simple linear regression to more complex, multi-model probabilistic analysis.
The takeaway
The BACE package demonstrates how specialized tools can lower the barrier to entry for complex Bayesian analysis in economics. Users should ensure their datasets are compatible with the specific regression models supported by the gretl host environment before implementation.
Further reading
For more information on the current state of professional research software, visit Economics — General.
More information
Read the full details on the BACE package journal article for technical specifications.
Source note: This article includes information reported by Jstatsoft.
