Third party projects and code snippets

This page references projects and code snippets (gists) which are compatible with the scikit-learn API conventions.

Projects

Different models and algorithms

• lightning Large-scale linear classification and regression in Python/Cython
• libOPF Optimal path forest classifier
• pyIPCA Incremental Principal Component Analysis
• sklearn_pandas bridge for scikit-learn pipelines and pandas data frame with dedicated transformers.
• py-earth Multivariate adaptive regression splines
• [HMMLearn] (https://github.com/hmmlearn/hmmlearn) Hidden Markov Models
• sklearn-compiledtrees generate a C++ implementation of the predict function for decision trees (and ensembles) trained by sklearn. Useful for latency-sensitive production environments.
• glm-sklearn scikit-learn compatible wrapper around the GLM module in statsmodel.

Application-specific projects

• [nilearn] (http://nilearn.github.io/) NeuroImaging with scikit-learn

Gists

• Multi-Layer-Perceptron neural network classifier trained by SGD
• Non-Negative Garotte
• Kernel SGD
• Fuzzy K-means and K-medians
• Fast svmlight / libsvm file loader
• Kernel k-means
• Non-negative Least-Squares
• Sparse Filtering Unsupervised feature learning based on sparse-filtering
• Non-negative Matrix Factorization for I-divergence

Other

Code snippets that do not follow the fit / predict / transform API.

• Adaptive Lasso (no class provided)
• Generating data with non-parametric Gaussian mixture models. Useful if you need "random" data that should have non-trivial structure.