bayesianroc 0.0.9

Bayesian ROC analysis toolkit

bayesianroc

bayesianroc is a Python package written by Franz Mayr and André Carrington, for Bayesian analysis of ROC plots (including Binary Chance), over the whole plot or in a region of interest.

Please read 'Bayesian ROC Tookit Documentation.docx' from the Github page for details.

Features

• Classes:

  • BayesianROC: a subclass of DeepROC (from the Deep ROC Toolkit). It computes measures related to the Chance and Bayesian iso performance baselines and produces associated plots.

• Example of Classification and Analysis

  • Test_Binary_Chance.py: creates a BayesianROC object and performs classification and analysis. Questions are asked as input: you may hit enter to accept defaults, except it is recommended that you change the costs to see the effect of Binary Chance.

Installation

The package can be installed from PyPi:

pip install bayesianroc