scikit-learn/scikit-learn

sklearn.cluster

Clustering algorithms including K-Means (with Cython Lloyd/Elkan/minibatch kernels), DBSCAN, affinity propagation, spectral, hierarchical, and OPTICS.

sklearn.decomposition

Matrix decomposition methods such as PCA, NMF, FastICA, dictionary learning, LDA, and kernel PCA.

sklearn.datasets

Loading and fetching sample datasets (including remote downloads from OpenML and other hosts) plus synthetic data generators and SVMLight format I/O.

sklearn.callback

Callback framework for monitoring estimator fitting, including progress bars, scoring monitors, and callback transport/context plumbing.

sklearn._loss

Loss functions and link functions for gradient boosting and linear models, with a Cython core generated from a tempita template (_loss.pyx.tp).

sklearn.compose

ColumnTransformer for per-column preprocessing pipelines and transformed-target regressors.

sklearn.covariance

Covariance estimation including empirical, shrunk, robust (MCD), graphical lasso, and elliptic envelope estimators.

Global configuration & base API (sklearn/_config.py, sklearn/base.py, sklearn/pipeline.py)

Thread-local global config (set_config/config_context), BaseEstimator machinery, and Pipeline composition of estimators.