whitebeam

whitebeam is a python module for decision trees without the use of sklearn and SciPy

docs
test
package

The project was started in 2020 by Kieran Molloy as a coursework component for SCC461 at Lancaster University

Installation

Dependencies

whitebeam requires:

• Python (>= 3.6)
• NumPy (>= 1.14)
• Cython (>= 0.29.21)
• joblib (>= 1.0.0)

User Installation

If you already have a working directory, the easiest installation is via pip

pip install whitebeam

However it is advisable to install within a virtual environment

virtualenv venv
venv/bin/activate
pip install whitebeam

There will soon be more detailed installation instructions in the docs.

Changelog

See the changelog for a history of notable changes to whitebeam

Development

Development is planned to be ongoing however this may change due to university requirements.

Source Installation

Installing from source requires python-dev which can be installed using

sudo apt-get install python3-dev

then cloning this repo, installing requirements and building which is handled by poetry

git clone https://github.com/K-Molloy/whitebeam
pip install -r requirements.txt
python setup.py build_ext --inplace

again, it is advisable to use a virtual environment.

Testing

After installation, the test suite can be launched from outside the source directory ( although pytest >= 5.01 must be installed)

pytest whitebeam