There are several channels to connect with scikit-learn developers for assistance, feedback, or contributions.
Note: Communications on all channels should respect our Code of Conduct.
- Main Mailing List: Join the primary discussion platform for scikit-learn at scikit-learn Mailing List.
- Commit Updates: Stay informed about repository updates and test failures on the scikit-learn-commits list.
If you have questions, this is our general workflow.
- Stack Overflow: Some scikit-learn developers support users using the [scikit-learn] tag.
- General Machine Learning Queries: For broader machine learning discussions, visit Stack Exchange.
When posting questions:
- Please use a descriptive question in the title field (e.g. no "Please help with scikit-learn!" as this is not a question)
- Provide detailed context, expected results, and actual observations.
- Include code and data snippets (preferably minimalistic scripts, up to ~20 lines).
- Describe your data and preprocessing steps, including sample size, feature types (categorical or numerical), and the target for supervised learning tasks (classification type or regression).
Note: Avoid asking user questions on the bug tracker to keep the focus on development.
- GitHub Discussions Usage questions such as methodological
- Stack Overflow Programming/user questions with [scikit-learn] tag
- GitHub Bug Tracker Bug reports - Please do not ask usage questions on the issue tracker.
- Discord Server Current pull requests - Post any specific PR-related questions on your PR, and you can share a link to your PR on this server.
Encountered a bug? Report it on our issue tracker
Include in your report:
- Steps or scripts to reproduce the bug.
- Expected and observed outcomes.
- Python or gdb tracebacks, if applicable.
- The ideal bug report contains a :ref:`short reproducible code snippet <minimal_reproducer>`, this way anyone can try to reproduce the bug easily.
- If your snippet is longer than around 50 lines, please link to a gist or a github repo.
Tip: Gists are Git repositories; you can push data files to them using Git.
The following companies (listed in alphabetical order) offer support services related to scikit-learn and have a proven track record of employing long-term maintainers of scikit-learn and related open source projects:
scikit-learn has presence on various social media platforms to share updates with the community. The platforms are not monitored for user questions.
Note: The scikit-learn Gitter room is no longer an active community. For live discussions and support, please refer to the other channels mentioned in this document.
This documentation is for |release|. Documentation for other versions can be found here, including zip archives which can be downloaded for offline access.
We no longer provide a PDF version of the documentation, but you can still generate it locally by following the :ref:`building documentation instructions <building_documentation>`. The most recent version with a PDF documentation is quite old, 0.23.2 (released in August 2020), but the PDF is available here.