unite-toolbox 0.2.1

A toolbox for practical applications of information theory.

UNITE Toolbox

Unified diagnostic evaluation of scientific models based on information theory

The UNITE Toolbox is a Python library for incorporating Information Theory into data analysis and modeling workflows. The toolbox collects different methods of estimating information-theoretic quantities in one easy-to-use Python package. Currently, UNITE includes functions to calculate entropy $H(X)$, Kullback-Leibler divergence $D_{KL}(p||q)$, and mutual information $I(X; Y)$, using three methods:

• Kernel density-based estimation (KDE)
• Binning using histograms
• k-nearest neighbor-based estimation (k-NN)

Installation

Although the code is still highly experimental and in very active development, a release version is available on PyPI and can be installed using pip.

pip install unite_toolbox

Or uv.

uv add unite_toolbox

Check the pyproject.toml for requirements.

How-to

In the documentation please find tutorials on the general usage of the toolbox and some applications.