softpy: A User-Friendly Python Library for Soft Computing

Soft computing refers to a computational paradigm aimed at providing approximate and heuristic solutions to hard problems, often adopting strategies inspired by nature or human reasoning: evolutionary computation and fuzzy systems are prime examples of this approach. In this article we present softpy, a Python library for soft computing designed to support the development of soft computing-based applications in both practical contexts as well as in educational use. Compared to other existing libraries and frameworks, softpy provides a common implementation of both evolutionary computing and fuzzy systems functionalities under an integrated API compatible with the Python data science ecosystem. After describing the design philosophy of the library, we provide a self-contained documentation of its contents and functionalities, and then provide a simple example aimed at illustrating its use in the development of an hybrid soft computing application. Finally, we describe potential directions for improvement and future extensions of the library.