A Unified Python Interface to a Variety of Software for Solving Ordinary Differential Equations

Abstract

Nowadays there is a jungle of practical software and numerical methods to solve the ordinary differential equations (ODE) related problems. This thesis aims to introduce a Python interface that collects a variety of such software and methods, and make use of them through an unified interface. It should be in a simple and minimal sense for novice users of ODE. Newcomers to ODE can easily apply numerical methods to solve simple ODE problems, and trivially switch between these solvers. At the same time, for professional users, it can be 'extended' to be an advanced interface, which offers the expected functionality of advanced methods and great flexibility. The most important contribution of my work is the Python code, which is composed by 4 files: ODE.py, ODEPACK.py, ODEPACK_demo.py, RungeKutta.py. Readers can get a better understanding with a rough look at the original code. This interface has been integrated into the popular Python module 'scitools' as a new component. As an opensource package, the complete code can be downloaded directly from the website of 'scitools' for free. Up till now, many different ODE methods have been integrated into this interface. As an growing interface, we expect further more ODE methods to be integrated in the future.