Changelog
Source:.github/py_changelog.md
All notable changes to the Python shaprpy library will be documented in this file. Changes to the underlying shapr R package is documented in shapr CHANGELOG
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
[0.4.0] - 2025-10-20
Added
-
NEW:
Shaprclass for exploration and analysis of explanation results- Object-oriented interface for working with Shapley value explanations
- Methods for extracting results:
get_results(),get_explanation_dict(),get_r_object() - Summary and printing functionality:
summary(),print() - Integration with SHAP library via
to_shap()method for plotting
- Enhanced example scripts demonstrating new class-based functionality
- Improved R package import handling with better path management
Changed
-
BREAKING:
explain()function now returns aShaprobject instead of a plain dictionary - Restructured internal module organization with private
_explain.pyand_rutils.pymodules - Updated all example scripts, tests and README to demonstrate new class-based API
- Improved lazy loading for R package dependencies
[0.3.0] - 2025-08-27
Added
- Publish
shaprpyon PyPI - Restructured the
shaprpypackage for improved modularity and maintainability - Test suite with snapshot testing for output validation using syrupy
- GitHub Actions CI/CD pipeline for automated testing across Python versions
- Input validation tests for error handling
Changed
- BREAKING: Minimum Python version requirement increased from 3.8+ to 3.10+
- Updated package classifiers to reflect Python 3.10+ requirement