Categorical Encoding#

Feature-engine’s categorical encoders replace the categories of the variable with estimated or arbitrary numbers.

Summary of Feature-engine’s encoders characteristics

Transformer	Regression	Classification	Multi-class	Description
`OneHotEncoder()`	√	√	√	Adds dummy variables to represent each category
`OrdinalEncoder()`	√	√		√ Replaces categories with an integer
`CountFreuencyEncoder()`	√	√	√	Replaces categories with their count or frequency
`MeanEncoder()`	√	√	x	Replaces categories with the targe mean value
`WoEEncoder()`	x	√	x	Replaces categories with the weight of the evidence
`DecisionTreeEncoder()`	√	√		√ Replaces categories with the predictions of a decision tree
`RareLabelEncoder()`	√	√		√ Groups infrequent categories into a single one

Feature-engine’s categorical encoders encode only variables of type categorical or object by default. From version 1.1.0, you have the option to set the parameter ignore_format to True to make the transformers also accept numerical variables as input.

Other categorical encoding libraries#

For additional categorical encoding transformations, visit the open-source package Category encoders.