OPFython: A Python implementation for Optimum-Path Forest

Gustavo Henrique de Rosa, João P. Papa

Abstract: OPFython is an open-sourced Python package that implements Optimum-Path Forest algorithms using object-oriented programming and a straightforward structure. It provides an alternative implementation to the standard LibOPF package, which heavily depends on the C language and occasionally hinders fast prototyping. Additionally, OPFython provides documented code, unitary tests, and examples that assist users in learning how to work with the package. Such features are well-suited for researchers and developers interested in exploring alternative state-of-the-art machine learning algorithms.