Efficiently and easily integrating differential equations with JiTCODE, JiTCDDE, and JiTCSDE

TL;DR

This paper introduces Python modules JiTCODE, JiTCDDE, and JiTCSDE for efficient, symbolic, just-in-time compiled numerical integration of various differential equations, especially suited for large systems and Lyapunov exponent estimation.

Contribution

The paper presents a new family of Python modules that simplify and accelerate the integration of ordinary, delay, and stochastic differential equations with symbolic input and JIT compilation.

Findings

Modules efficiently handle large systems of differential equations.

They enable automated estimation of Lyapunov exponents.

Performance analysis shows high efficiency and usability.

Abstract

We present a family of Python modules for the numerical integration of ordinary, delay, or stochastic differential equations. The key features are that the user enters the derivative symbolically and it is just-in-time-compiled, allowing the user to efficiently integrate differential equations from a higher-level interpreted language. The presented modules are particularly suited for large systems of differential equations such as used to describe dynamics on complex networks. Through the selected method of input, the presented modules also allow to almost completely automatize the process of estimating regular as well as transversal Lyapunov exponents for ordinary and delay differential equations. We conceptually discuss the modules' design, analyze their performance, and demonstrate their capabilities by application to timely problems.