dynamicslab/pysindy

Health v2 · maintenance only

Flags: no_license

Adoption not part of the score

1893 stars · 384 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

PySINDy is a Python library for discovering governing equations of dynamical systems from measurement data using the Sparse Identification of Nonlinear Dynamical systems (SINDy) method and related sparse regression techniques. The resulting models are interpretable symbolic equations that can be used for prediction, control, and theoretical analysis.

Use cases

• discover differential equations from time-series data
• fit sparse regression models to identify nonlinear dynamics
• infer governing equations of a physical system from measurements
• build interpretable models of dynamical systems for prediction and control
• perform system identification on experimental or simulation data
• learn symbolic equations of motion from trajectory data

When to choose

• you have time-series measurements and want interpretable symbolic equations rather than a black-box model
• you need to identify sparse nonlinear dynamics with techniques like SR3, MIOSR, or Bayesian regression
• you want a well-documented, actively maintained scientific Python package with tutorials

When to avoid

• you need black-box forecasting accuracy rather than interpretable equations
• your data is noisy, sparse, or lacks the sampling quality SINDy methods require
• you need a general-purpose deep learning time-series model instead of equation discovery

Facets

library · maturity active

machine-learning data-science simulation math machine-learning data-science simulation python cross-platform system-identification sparse-regression dynamical-systems model-discovery nonlinear-dynamics interpretable-models scientific-computing algorithms

Member repositories

For agents

markdown · JSON · MCP: product_card(name="dynamicslab/pysindy")