Release Notes¶
Note
This project uses Semantic Versioning for version numbering:
The version number is in the format of
MAJOR.MINOR.PATCH. An increment in theMAJORversion indicates incompatible API changes, an increment in theMINORversion indicates added functionality in a backward-compatible manner, and an increment in thePATCHversion indicates backward-compatible bug fixes.Major version zero (0.y.z) is for beta releases. Anything may change at any time. The public API should not be considered stable till version 1.y.z.
Version 0.6.0 (Beta) - Forthcoming¶
Added
Add COPT benchmarks (#15).
Changed
Remove
PyYAML>=6.0.0dependency by replacingconfig.yamlwithconfig.py(#9).
Version 0.5.1 (Beta) - 2026-04-29¶
Added
Add a GitHub Actions workflow for automatic releases on PyPI.
Add a
CITATION.cfffile for citation metadata.[docs] Add
.readthedocs.yamlfor Read the Docs integration.[docs] Add sitemap using
sphinx_sitemap.
Changed
Make
matplotliban optional dependency.Update the dependencies for benchmarking.
[docs] Update the logos in the documentation.
Fixed
Fix a bug caused by inconsistent placeholder solver names.
Version 0.5.0 (Beta) - 2026-04-17¶
We are excited to announce the first public beta release of BendersLib, a Python library dedicated to Benders decomposition! To get started, please see the Quickstart section in our documentation. We welcome any feedback. Please report any issues or suggestions on our GitHub Issues page. We also welcome contributions to the library.
Added
Benders Decomposition Variants: Classical Benders Decomposition, Combinatorial Benders Decomposition, Generalized Benders Decomposition, L-shaped Method (linear and convex recourse), Integer L-shaped Method, and Logic-based Benders Decomposition.
Extensibility: Annotated Benders Decomposition, Custom Cut Generation, Custom Subproblem Solvers, and Callbacks.
Enhancement Options: Branch-and-check Method, Parallel Subproblem Solving, Multi-cut Generation, Cut Normalization, and IIS-based Feasibility Cut Generation (
use_iis_cut).Solver Agnostic: Built-in interfaces for popular solvers.