Skip to content

References

References

The reference shelf: where the library sits against the rest of the Python ecosystem, the papers its methods and systems come from, and the complete public API generated from the source docstrings. Everything here is meant to be looked up, not read front to back — start from Start or Analysis for the prose-first tour.

Benchmarks

Benchmarks →

A head-to-head comparison against the established Python dynamical-systems libraries on a shared set of classic tasks — short and long ODE integration, integration accuracy, the Lyapunov family, correlation dimension, the bifurcation sweep, basins, fixed points, the Poincaré section, and a spread of from-a-signal complexity measures. It records both speed (every task) and, where there is a ground truth, precision (the estimate and its deviation from the literature value). The short version: integration is the clear strength — the Rust engine runs roughly two orders of magnitude faster than scipy.integrate.solve_ivp while returning the whole dense trajectory in one call — and the analysis toolkit is accurate wherever a literature value exists.

Bibliography

Bibliography →

The original literature, in two parts. The methods half cites the papers behind the analysis toolkit — the Lyapunov, dimension, entropy, recurrence, surrogate and basin routines each trace back to a specific source, never to another library. The systems half is the catalogue's provenance: every built-in system carries the paper that defined it (and, where one exists, its DOI), pulled straight from the registry so the page cannot drift from the code.

API reference

API reference →

The complete public API, one page per area, rendered from the docstrings in the source:

  • Top level — everything importable straight from tsdynamics.
  • Base classes — ContinuousSystem, DelaySystem, DiscreteMap, StochasticSystem, Trajectory, and the System protocol.
  • Derived systems — PoincareMap, StroboscopicMap, TangentSystem, EnsembleSystem, ProjectedSystem, WrappedSystem.
  • Analysis — the full quantifier toolkit.
  • Transforms — spectra, preprocessing, and feature extraction.
  • Data & state-space — Trajectory, regions, samplers, set distances.
  • Solvers — the by-name method registry and auto-stiffness.
  • Registry — programmatic discovery of the built-in catalogue.
  • Utilities — the shared output-grid helper.

The visualization layer has its own dedicated section, and each built-in system is documented on its own generated page under Systems.