credit-scoring 0.2.0

This library help researcher have quick result with woe, iv, result reporting with credit scoring problem

Project Description

Credit-Score

Credit Score is a Python module for credit scoring problems, help developers/ researchers applied woe-iv for easier. More, It provides some importance reports about credit scoring contains: GINI, KS, LIFT, GAIN, IRL, LR, ... help users save time to choose the best model.

Currently, credit-scoring package handles only binary target

Installation

This package requires numpy, pandas, sklearn and plotly. We can use this package for Python 3, and pandas must be >= 1.0

To install the package, run this command:

python setup.py install

or

pip install credit_scoring

If you want install the development of this package, you can visit: https://github.com/minhtcuet/creditscoring

Usage

This version, we focus on reporting first. You can have quick GINI, KS, LIFT, .... with 2 input parameters contains: y_target(binary type) and y_predict(probability)

Example

GINI Calculated

from credit_scoring.AUC import GINI

gini = GINI(y_predict, y_label) print(gini)_