Python module for two-sample statistical tests


License
GPL-3.0
Install
pip install twosampyle==0.1.3

Documentation

twosampyle

Build Status

twosampyle is a minimalist Python module for testing two sample (A/B) tests. It was created with the following design objectives:

  1. Ease of use: twosampyle has a uniform API and all statistical tests share the same methods and attributes. Think of something along the lines of sci-kit learn's API, but for statistical tests instead of machine learning.

  2. Minimalism : twosampyle only includes functionality that you'll need, and (in its current form) only incudes the following methods:

Getting Started: Testing with twosampyle

twosampyle currently includes the following statistical tests:

  • Nonparametric

      - `PermTest()`: Implementation of the classic permutation test.
      - `PermutedChiSquare()`: Nonparametric implementation of the chisquare test for independence.
    
  • Parametric

      - `ttest`:
      - `Chisquare`:
    
  • Bayesian

      - `BayesianAB`: Bayesian A/B Test
    

As state previously, each test contains the following methods