SKLL 4.0.0

💥 Breaking changes 💥

• scikit-learn has been updated to v1.3.0. This could mean that the same SKLL experiments when run with SKLL 3.2.0 could yield different results.

💡 New features 💡

• Add BaggingClassifier and BaggingRegressor support by @desilinguist in #742
• Add support for HistGradientBoostingClassifier and HistGradientBoostingRegressor by @desilinguist in #743
• Include model fit times in learning curves by @desilinguist in #745
• Add neg_root_mean_squared_error metric and objective for regressors by @desilinguist in #741
• Add support for Python 3.11 by @desilinguist in #749

🛠 Bugfixes & Improvements 🛠

• Apply code formatting and other minor changes. by @desilinguist in #724
• Use pathlib.Path where possible. by @desilinguist in #725
• Migrate to new codecov uploader. by @desilinguist in #728
• Add type hints to skll.config module by @desilinguist in #729
• Add type hints to skll.data module & improve types in skll.config by @desilinguist in #730
• Bug fix in feature set split method by @tamarl08 in #731
• Add type hints to skll.experiments module by @desilinguist in #732
• Add type hints to skll.learner module + other refactoring by @desilinguist in #734
• Add type hints inskll.utils module and in all other remaining files. by @desilinguist in #736
• Improve docstrings and create linkable type hints (Part 1) by @desilinguist in #737
• Improve docstrings and type hints (Part 2). by @desilinguist in #738
• Improve docstrings and type hints (Part 3) by @desilinguist in #739
• Improve docstrings & type hints (Part 4) by @desilinguist in #740
• Migrate tests to nose2 instead of nose by @desilinguist in #747
• Stop using sklearn's private _scorer API for custom metrics in SKLL. by @desilinguist in #751
• Fix a few typos, etc. in the documentation by @mulhod in #712

🙏🏽 Code reviewers 🙏🏽

In no particular order: @dblandan, @mulhod, @Frost45, @tamarl08, @damien2012eng

New Contributors

• @tamarl08 made their first contribution in #731

Full Changelog: v3.2.0...v4.0.0