Skip to content

jkleint/categorical-encoding

 
 

Categorical Encoding Methods

Travis Status Coveralls Status CircleCI Status DOI

A set of scikit-learn-style transformers for encoding categorical variables into numeric by means of different techniques.

Important Links

Documentation: http://contrib.scikit-learn.org/categorical-encoding/

Encoding Methods

  • Backward Difference Contrast [2][3]
  • BaseN [6]
  • Binary [5]
  • Hashing [1]
  • Helmert Contrast [2][3]
  • LeaveOneOut [4]
  • Ordinal [2][3]
  • One-Hot [2][3]
  • Polynomial Contrast [2][3]
  • Sum Contrast [2][3]
  • Target Encoding [7]
  • Weight of Evidence [8]

Usage

The package by itself comes with a single module and an estimator. Before installing the module you will need numpy, statsmodels, and scipy.

To install the module execute:

$ python setup.py install

or

pip install category_encoders

or

conda install -c conda-forge category_encoders

To use:

import category_encoders as ce

encoder = ce.BackwardDifferenceEncoder(cols=[...])
encoder = ce.BaseNEncoder(cols=[...])
encoder = ce.BinaryEncoder(cols=[...])
encoder = ce.HashingEncoder(cols=[...])
encoder = ce.HelmertEncoder(cols=[...])
encoder = ce.LeaveOneOutEncoder(cols=[...])
encoder = ce.OneHotEncoder(cols=[...])
encoder = ce.OrdinalEncoder(cols=[...])
encoder = ce.PolynomialEncoder(cols=[...])
encoder = ce.SumEncoder(cols=[...])
encoder = ce.TargetEncoder(cols=[...])
encoder = ce.WOEEncoder(cols=[...])

All of these are fully compatible sklearn transformers, so they can be used in pipelines or in your existing scripts. If the cols parameter isn't passed, every non-numeric column will be encoded. Please see the docs for transformer-specific configuration options.

Examples

from category_encoders import *
import pandas as pd
from sklearn.datasets import load_boston

# prepare some data
bunch = load_boston()
y = bunch.target
X = pd.DataFrame(bunch.data, columns=bunch.feature_names)

# use binary encoding to encode two categorical features
enc = BinaryEncoder(cols=['CHAS', 'RAD']).fit(X, y)

# transform the dataset
numeric_dataset = enc.transform(X)

In the examples directory, there is an example script used to benchmark different encoding techniques on various datasets.

The datasets used in the examples are car, mushroom, and splice datasets from the UCI dataset repository, found here:

datasets

Contributing

Category encoders is under active development, if you'd like to be involved, we'd love to have you. Check out the CONTRIBUTING.md file or open an issue on the github project to get started.

License

BSD 3-Clause

References:

  1. Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML.
  2. Contrast Coding Systems for categorical variables. UCLA: Statistical Consulting Group. from https://stats.idre.ucla.edu/r/library/r-library-contrast-coding-systems-for-categorical-variables/.
  3. Gregory Carey (2003). Coding Categorical Variables. from http://psych.colorado.edu/~carey/Courses/PSYC5741/handouts/Coding%20Categorical%20Variables%202006-03-03.pdf
  4. Strategies to encode categorical variables with many categories. from https://www.kaggle.com/c/caterpillar-tube-pricing/discussion/15748#143154.
  5. Beyond One-Hot: an exploration of categorical variables. from http://www.willmcginnis.com/2015/11/29/beyond-one-hot-an-exploration-of-categorical-variables/
  6. BaseN Encoding and Grid Search in categorical variables. from http://www.willmcginnis.com/2016/12/18/basen-encoding-grid-search-category_encoders/
  7. A Preprocessing Scheme for High-Cardinality Categorical Attributes in Classification and Prediction Problems. from https://kaggle2.blob.core.windows.net/forum-message-attachments/225952/7441/high%20cardinality%20categoricals.pdf
  8. Weight of Evidence (WOE) and Information Value Explained. from https://www.listendata.com/2015/03/weight-of-evidence-woe-and-information.html

About

A library of sklearn compatible categorical variable encoders

Resources

License

Code of conduct

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 95.8%
  • Shell 2.3%
  • TeX 1.9%