Implementation of WENDy in Python, featuring automated critical radius selection for optimal parameters recovery.
WENDy is a forward-solver-free algorithm for estimating parameters in differential equations using the weak form. See https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10002818/ for details about WENDy. This repository contains code accompanying the paper:
Weak Form Scientific Machine Learning: Test Function Construction for System Identification by A. Tran and D. M. Bortz. PDF on arXiv: https://arxiv.org/pdf/2507.03206
Copyright 2025, All Rights Reserved. This codebase was developed by April Tran and adapted from the MATLAB implementation of WENDy, originally written by Dan Messenger.