Simframe: A Python Framework for Scientific Simulations

Abstract

Simframe is a Python framework to facilitate scientific simulations. The scope of the software is to provide a framework which can hold data fields, which can be used to integrate differential equations, and which can read and write data files. Conceptually, upon initialization Simframe is an empty frame that can be filled with Fields containing data. Fields are derived from numpy.ndarrays (Harris et al., 2020), but with extended functionality. The user can then specify differential equations to those data fields and can set up an integrator which is integrating those fields according the given differential equations. Therefore, Simframe can only work with data, that can be stored in NumPy arrays. Data fields that should not be integrated themselves, but are still required for the model, can have an update function assigned to them, according to which they will be updated once per integration step. Simframe contains a number of integration schemes of different orders, both for explicit and implicit integration. Furthermore, Simframe includes methods to read and write output files. Due to its modular structure, Simframe can be extended at will, for example, by implementing new integration schemes and/or user-defined output formats.