Cleanlab Files

This release is non-breaking when upgrading from v2.7.0, mostly focused on documentation and testing improvements. The most notable update is a:

• New identifier column issue manager – detects sequential numerical columns that might influence your model. This feature is available as a preview and requires additional setup to use with Datalab.

Other Updates: - 📖 Docs & Readme: Improved clarity. - 🛠 Test suite: More stability and consistency.

What's Changed

• Added issue manager for detecting identifier columns by @MaxJoas in https://github.com/cleanlab/cleanlab/pull/1120
• Revised non-IID section in Datalab tutorial to show overall dataset score first before per-example insights by @gordon-lim in https://github.com/cleanlab/cleanlab/pull/1221
• Fixed numpy2 compatibility by @GaetanLepage in https://github.com/cleanlab/cleanlab/pull/1224
• Updated CI to macos-13 instead of 12 by @jwmueller in https://github.com/cleanlab/cleanlab/pull/1219
• Test improvements by @gordon-lim in https://github.com/cleanlab/cleanlab/pull/1218; @jwmueller in https://github.com/cleanlab/cleanlab/pull/1220
• General updates to docs and docs build system by @misteroh in https://github.com/cleanlab/cleanlab/pull/1212, https://github.com/cleanlab/cleanlab/pull/1213, https://github.com/cleanlab/cleanlab/pull/1216; @maxbuchan in https://github.com/cleanlab/cleanlab/pull/1226, https://github.com/cleanlab/cleanlab/pull/1227; @elisno in https://github.com/cleanlab/cleanlab/pull/1208, https://github.com/cleanlab/cleanlab/pull/1231

New Contributors

• @MaxJoas made their first contribution in https://github.com/cleanlab/cleanlab/pull/1120
• @misteroh made their first contribution in https://github.com/cleanlab/cleanlab/pull/1212
• @GaetanLepage made their first contribution in https://github.com/cleanlab/cleanlab/pull/1224
• @maxbuchan made their first contribution in https://github.com/cleanlab/cleanlab/pull/1226

Full Changelog: https://github.com/cleanlab/cleanlab/compare/v2.7.0...v2.7.1