.. _related_projects: ===================================== Related Projects ===================================== Projects implementing the scikit-learn estimator API are encouraged to use the `scikit-learn-contrib template `_ which facilitates best practices for testing and documenting estimators. The `scikit-learn-contrib GitHub organization `_ also accepts high-quality contributions of repositories conforming to this template. Below is a list of sister-projects, extensions and domain specific packages. Interoperability and framework enhancements ------------------------------------------- These tools adapt scikit-learn for use with other technologies or otherwise enhance the functionality of scikit-learn's estimators. **Data formats** - `sklearn_pandas `_ bridge for scikit-learn pipelines and pandas data frame with dedicated transformers. - `sklearn_xarray `_ provides compatibility of scikit-learn estimators with xarray data structures. **Auto-ML** - `auto-sklearn `_ An automated machine learning toolkit and a drop-in replacement for a scikit-learn estimator - `autoviml `_ Automatically Build Multiple Machine Learning Models with a Single Line of Code. Designed as a faster way to use scikit-learn models without having to preprocess data. - `TPOT `_ An automated machine learning toolkit that optimizes a series of scikit-learn operators to design a machine learning pipeline, including data and feature preprocessors as well as the estimators. Works as a drop-in replacement for a scikit-learn estimator. - `Featuretools `_ A framework to perform automated feature engineering. It can be used for transforming temporal and relational datasets into feature matrices for machine learning. - `Neuraxle `_ A library for building neat pipelines, providing the right abstractions to both ease research, development, and deployment of machine learning applications. Compatible with deep learning frameworks and scikit-learn API, it can stream minibatches, use data checkpoints, build funky pipelines, and serialize models with custom per-step savers. - `EvalML `_ EvalML is an AutoML library which builds, optimizes, and evaluates machine learning pipelines using domain-specific objective functions. It incorporates multiple modeling libraries under one API, and the objects that EvalML creates use an sklearn-compatible API. **Experimentation and model registry frameworks** - `MLFlow `_ MLflow is an open source platform to manage the ML lifecycle, including experimentation, reproducibility, deployment, and a central model registry. - `Neptune `_ Metadata store for MLOps, built for teams that run a lot of experiments. It gives you a single place to log, store, display, organize, compare, and query all your model building metadata. - `Sacred `_ Tool to help you configure, organize, log and reproduce experiments - `Scikit-Learn Laboratory `_ A command-line wrapper around scikit-learn that makes it easy to run machine learning experiments with multiple learners and large feature sets. **Model inspection and visualization** - `dtreeviz `_ A python library for decision tree visualization and model interpretation. - `eli5 `_ A library for debugging/inspecting machine learning models and explaining their predictions. - `sklearn-evaluation `_ Machine learning model evaluation made easy: plots, tables, HTML reports, experiment tracking and Jupyter notebook analysis. Visual analysis, model selection, evaluation and diagnostics. - `yellowbrick `_ A suite of custom matplotlib visualizers for scikit-learn estimators to support visual feature analysis, model selection, evaluation, and diagnostics. **Model selection** - `scikit-optimize `_ A library to minimize (very) expensive and noisy black-box functions. It implements several methods for sequential model-based optimization, and includes a replacement for ``GridSearchCV`` or ``RandomizedSearchCV`` to do cross-validated parameter search using any of these strategies. - `sklearn-deap `_ Use evolutionary algorithms instead of gridsearch in scikit-learn. **Model export for production** - `sklearn-onnx `_ Serialization of many Scikit-learn pipelines to `ONNX `_ for interchange and prediction. - `skops.io `__ A persistence model more secure than pickle, which can be used instead of pickle in most common cases. - `sklearn2pmml `_ Serialization of a wide variety of scikit-learn estimators and transformers into PMML with the help of `JPMML-SkLearn `_ library. - `sklearn-porter `_ Transpile trained scikit-learn models to C, Java, Javascript and others. - `m2cgen `_ A lightweight library which allows to transpile trained machine learning models including many scikit-learn estimators into a native code of C, Java, Go, R, PHP, Dart, Haskell, Rust and many other programming languages. - `treelite `_ Compiles tree-based ensemble models into C code for minimizing prediction latency. - `micromlgen `_ MicroML brings Machine Learning algorithms to microcontrollers. Supports several scikit-learn classifiers by transpiling them to C code. - `emlearn `_ Implements scikit-learn estimators in C99 for embedded devices and microcontrollers. Supports several classifier, regression and outlier detection models. **Model throughput** - `Intel(R) Extension for scikit-learn `_ Mostly on high end Intel(R) hardware, accelerates some scikit-learn models for both training and inference under certain circumstances. This project is maintained by Intel(R) and scikit-learn's maintainers are not involved in the development of this project. Also note that in some cases using the tools and estimators under ``scikit-learn-intelex`` would give different results than ``scikit-learn`` itself. If you encounter issues while using this project, make sure you report potential issues in their respective repositories. Other estimators and tasks -------------------------- Not everything belongs or is mature enough for the central scikit-learn project. The following are projects providing interfaces similar to scikit-learn for additional learning algorithms, infrastructures and tasks. **Time series and forecasting** - `Darts `_ Darts is a Python library for user-friendly forecasting and anomaly detection on time series. It contains a variety of models, from classics such as ARIMA to deep neural networks. The forecasting models can all be used in the same way, using fit() and predict() functions, similar to scikit-learn. - `sktime `_ A scikit-learn compatible toolbox for machine learning with time series including time series classification/regression and (supervised/panel) forecasting. - `skforecast `_ A python library that eases using scikit-learn regressors as multi-step forecasters. It also works with any regressor compatible with the scikit-learn API. - `tslearn `_ A machine learning library for time series that offers tools for pre-processing and feature extraction as well as dedicated models for clustering, classification and regression. **Gradient (tree) boosting** Note scikit-learn own modern gradient boosting estimators :class:`~sklearn.ensemble.HistGradientBoostingClassifier` and :class:`~sklearn.ensemble.HistGradientBoostingRegressor`. - `XGBoost `_ XGBoost is an optimized distributed gradient boosting library designed to be highly efficient, flexible and portable. - `LightGBM `_ LightGBM is a gradient boosting framework that uses tree based learning algorithms. It is designed to be distributed and efficient. **Structured learning** - `HMMLearn `_ Implementation of hidden markov models that was previously part of scikit-learn. - `PyStruct `_ General conditional random fields and structured prediction. - `pomegranate `_ Probabilistic modelling for Python, with an emphasis on hidden Markov models. - `sklearn-crfsuite `_ Linear-chain conditional random fields (`CRFsuite `_ wrapper with sklearn-like API). **Deep neural networks etc.** - `skorch `_ A scikit-learn compatible neural network library that wraps PyTorch. - `scikeras `_ provides a wrapper around Keras to interface it with scikit-learn. SciKeras is the successor of `tf.keras.wrappers.scikit_learn`. **Federated Learning** - `Flower `_ A friendly federated learning framework with a unified approach that can federate any workload, any ML framework, and any programming language. **Privacy Preserving Machine Learning** - `Concrete ML `_ A privacy preserving ML framework built on top of `Concrete `_, with bindings to traditional ML frameworks, thanks to fully homomorphic encryption. APIs of so-called Concrete ML built-in models are very close to scikit-learn APIs. **Broad scope** - `mlxtend `_ Includes a number of additional estimators as well as model visualization utilities. - `scikit-lego `_ A number of scikit-learn compatible custom transformers, models and metrics, focusing on solving practical industry tasks. **Other regression and classification** - `ML-Ensemble `_ Generalized ensemble learning (stacking, blending, subsemble, deep ensembles, etc.). - `lightning `_ Fast state-of-the-art linear model solvers (SDCA, AdaGrad, SVRG, SAG, etc...). - `py-earth `_ Multivariate adaptive regression splines - `gplearn `_ Genetic Programming for symbolic regression tasks. - `scikit-multilearn `_ Multi-label classification with focus on label space manipulation. - `seglearn `_ Time series and sequence learning using sliding window segmentation. - `fastFM `_ Fast factorization machine implementation compatible with scikit-learn **Decomposition and clustering** - `lda `_: Fast implementation of latent Dirichlet allocation in Cython which uses `Gibbs sampling `_ to sample from the true posterior distribution. (scikit-learn's :class:`~sklearn.decomposition.LatentDirichletAllocation` implementation uses `variational inference `_ to sample from a tractable approximation of a topic model's posterior distribution.) - `kmodes `_ k-modes clustering algorithm for categorical data, and several of its variations. - `hdbscan `_ HDBSCAN and Robust Single Linkage clustering algorithms for robust variable density clustering. As of scikit-learn version 1.3.0, there is :class:`~sklearn.cluster.HDBSCAN`. - `spherecluster `_ Spherical K-means and mixture of von Mises Fisher clustering routines for data on the unit hypersphere. **Pre-processing** - `categorical-encoding `_ A library of sklearn compatible categorical variable encoders. As of scikit-learn version 1.3.0, there is :class:`~sklearn.preprocessing.TargetEncoder`. - `imbalanced-learn `_ Various methods to under- and over-sample datasets. - `Feature-engine `_ A library of sklearn compatible transformers for missing data imputation, categorical encoding, variable transformation, discretization, outlier handling and more. Feature-engine allows the application of preprocessing steps to selected groups of variables and it is fully compatible with the Scikit-learn Pipeline. **Topological Data Analysis** - `giotto-tda `_ A library for `Topological Data Analysis `_ aiming to provide a scikit-learn compatible API. It offers tools to transform data inputs (point clouds, graphs, time series, images) into forms suitable for computations of topological summaries, and components dedicated to extracting sets of scalar features of topological origin, which can be used alongside other feature extraction methods in scikit-learn. Statistical learning with Python -------------------------------- Other packages useful for data analysis and machine learning. - `Pandas `_ Tools for working with heterogeneous and columnar data, relational queries, time series and basic statistics. - `statsmodels `_ Estimating and analysing statistical models. More focused on statistical tests and less on prediction than scikit-learn. - `PyMC `_ Bayesian statistical models and fitting algorithms. - `Seaborn `_ Visualization library based on matplotlib. It provides a high-level interface for drawing attractive statistical graphics. - `scikit-survival `_ A library implementing models to learn from censored time-to-event data (also called survival analysis). Models are fully compatible with scikit-learn. Recommendation Engine packages ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - `implicit `_, Library for implicit feedback datasets. - `lightfm `_ A Python/Cython implementation of a hybrid recommender system. - `OpenRec `_ TensorFlow-based neural-network inspired recommendation algorithms. - `Surprise Lib `_ Library for explicit feedback datasets. Domain specific packages ~~~~~~~~~~~~~~~~~~~~~~~~ - `scikit-network `_ Machine learning on graphs. - `scikit-image `_ Image processing and computer vision in python. - `Natural language toolkit (nltk) `_ Natural language processing and some machine learning. - `gensim `_ A library for topic modelling, document indexing and similarity retrieval - `NiLearn `_ Machine learning for neuro-imaging. - `AstroML `_ Machine learning for astronomy. Translations of scikit-learn documentation ------------------------------------------ Translation's purpose is to ease reading and understanding in languages other than English. Its aim is to help people who do not understand English or have doubts about its interpretation. Additionally, some people prefer to read documentation in their native language, but please bear in mind that the only official documentation is the English one [#f1]_. Those translation efforts are community initiatives and we have no control on them. If you want to contribute or report an issue with the translation, please contact the authors of the translation. Some available translations are linked here to improve their dissemination and promote community efforts. - `Chinese translation `_ (`source `__) - `Persian translation `_ (`source `__) - `Spanish translation `_ (`source `__) - `Korean translation `_ (`source `__) .. rubric:: Footnotes .. [#f1] following `linux documentation Disclaimer `__