feature_engine
Feature engineering and selection open-source Python library compatible with sklearn.
What is feature_engine?
Feature engineering and selection open-source Python library compatible with sklearn.
Overview of feature_engine
What it is
Feature-engine is an open-source Python library for feature engineering and selection, compatible with Scikit-learn. It provides multiple transformers that implement fit() and transform() methods to learn parameters from data and apply transformations. The library is licensed under BSD-3-Clause.
Key features
- Missing data imputation: includes MeanMedianImputer, CategoricalImputer, RandomSampleImputer, and others.
- Categorical encoding: provides OneHotEncoder, OrdinalEncoder, MeanEncoder, WoEEncoder, RareLabelEncoder, etc.
- Discretisation: methods such as EqualFrequencyDiscretiser, EqualWidthDiscretiser, DecisionTreeDiscretiser.
- Outlier handling: Winsoriser, ArbitraryOutlierCapper, OutlierTrimmer for capping or removing outliers.
- Variable transformation: LogTransformer, BoxCoxTransformer, YeoJohnsonTransformer, and more.
- Feature selection: includes DropCorrelatedFeatures, RecursiveFeatureElimination, MRMR, and others.
Use cases
Feature-engine is used to engineer and select features for machine learning models. Its transformers can be integrated into Scikit-learn pipelines for tasks such as handling missing data, encoding categorical variables, discretising continuous variables, capping outliers, transforming variables, and selecting relevant features. It also includes transformers for datetime, text, and time series features.
Getting started
Install via pip: pip install feature_engine Or via conda: conda install -c conda-forge feature_engine Or clone the repository: git clone https://github.com/feature-engine/feature_engine.git
License and pricing
Feature-engine is open source under the BSD-3-Clause license.
Based on the project's public README and repository metadata.