pysindy

Why we included this project

PySINDy turns noisy time-series measurements into readable differential equations. It implements the SINDy method, which searches for a compact, interpretable model of how a system's state evolves instead of treating the data as a black box. That makes it useful for anyone working with physical systems, whether you are studying fluid flows, designing controllers, or building simulations. The package includes a library of candidate terms, several sparse regression optimizers, and tools for handling derivatives and noisy data, so you can go from raw measurements to a validated model without leaving Python. Teams already doing scientific computing in notebooks will find it slots in easily.