crucio 0.1.8

Crucio is a python sci-kit learn inspired package for class imbalance. It use some classic methods for class balancing taking as parameters a data frame and the target column.

Crucio is a python sci-kit learn inspired package for class imbalance. It use some classic methods for class balancing taking as parameters a data frame and the target column.

This version of crucio has the next methods of feature selection:

1. ADASYN

2. ICOTE (Immune Centroids Oversampling)

3. MTDF (Mega-Trend Difussion Function)

4. MWMOTE (Majority Weighted Minority Oversampling Technique)

5. SMOTE (Synthetic Minority Oversampling Technique)

6. SMOTENC (Synthetic Minority Over-sampling Technique for Nominal and Continuous)

7. SMOTETOMEK (Synthetic Minority Oversampling Technique + Tomek links for undersampling)

8. SMOTEENN (Synthetic Minority Oversampling Technique + ENN for undersampling)

9. SCUT (SMOTE and Clustered Undersampling Technique)

10. SLS (Safe-Level-Synthetic Minority Over-Sampling TEchnique)

11. TKRKNN (Top-K ReverseKNN)

All these methods takes the pandas Data Frame and y column to balance on.

How to use crucio

To use balancer from crucio you should just import the balancer from crucio in the following framework:

`` from crucio import SMOTE ``

class names are written above in parantheses.

Next create a object of this algorithm (I will use ADASYN method as an example).

` method = ADASYN() `

To balance the dataset on the target column use the ‘balance’ function, using as parameters the pandas Data Frame and the column that you want to balance.

` new_dataframe = method.balance(df, 'target') `

Returned value is a new data frame with the target column balanced.

With love from Sigmoid.

We are open for feedback. Please send your impression to [e-mail]