Psymple: A Python package for complex systems modelling

The modelling of complex systems, which are characterised by high degrees of interdependence amongst components, is continually evolving in response to developing computing and visualisation power.Examples such as ecological, economic, and social systems are classically modelled using statistical or correlative techniques, which have natural shortcomings when used for predictive modelling outside of the range of their parameterisation data.Predictive modelling is improved by deterministic models that capture the evolution of a system, or by hybrid models composed of both deterministic and correlative parts.These models often feature significantly increased size and complexity, and there has been a growing requirement for so-called next generation modelling platforms that are built to a specification requiring modular and data-first implementations that drive clear, adaptable, reusable and accessible modelling practices.The psymple platform is designed to facilitate the development of hybrid complex system models and modelling frameworks.It allows users to link together arbitrary combinations of modular differential equation systems and functional components to build categorical diagrams representing a complex system.A compilation process automatically generates simulatable system equations from these diagrams using the symbolic mathematics package sympy.Ultimately, this allows users to focus on the components and interactions of models, rather than their complex equation structure.