BestCatEncode: Detects categorical variables in a given dataset and encodes...

In akunuriYoshitha/CatEncode:

Description

Takes in a dataset, detects categorical variables and encodes the data using multiple encoding techniques:

Usage

BestCatEncode(data, dv)

Arguments

data	Any dataset with atleast one categorical field that is to be encoded
dv	Dependent variable in the given dataset

Details

Contrast Encoding

• Dummy Coding

• Simple Coding

• Deviation Coding

• Polynomial Coding

• Reverse Helmert Coding

• Helmert Coding

• Forward Difference Coding

• Backward Difference Coding

Generic Encoders

• Label/Integer Encoding

• One-Hot Encoding

• Binary Encoding

• Frequency Encoding

• Hashing Encoding

• Probability Ratio Encoding

Bayesian Encoders

• Mean Target Encoding

• Leave One Out Encoding

• Weight Of Evidence Encoding

• James Stein Encoding

• M-Estimator Encoding

Each of these 19 datasets are passed to a logistic model to choose the best possible dataset.

Value

Returns a list with an encoded dataset, performance metrics and a fit file to fit the test data.

akunuriYoshitha/CatEncode documentation built on July 16, 2021, 4:16 p.m.