Contributing

Since researchers are the intended audience of this library, we value correctness and readability over complex performance optimizations and broad functionality. Although note that we reuse scikit-learn’s built-in functions whenever we can, and that has somewhat different priorities.

We welcome all types of contributions: correcting bugs in code, typos in documentation or implementation of new algorithms. Our inclusion criteria for new algorithm is pretty relaxed- any algorithm that has been published in a peer-reviewed journal/conference is eligible. Please view this guideline from scikit-learn before you open a pull request.