Introduction

The project aims to integrate cppyy, an automatic Python-C++ bindings generator, with BioDynaMo. This will enable a hybrid modeling environment where users can call existing C++ agents from Python and also create new agents in Python that inherit directly from C++ base classes. This combines the performance of C++ with the accessibility, flexibility and libraries of Python Mentor: Vassil Vassilev

Overview

While BioDynaMo’s performance-critical core is written in C++, many researchers, students, and data scientists prefer Python for prototyping, analysis, and education. This project integrates cppyy, an automatic run-time Python-C++ bindings generator, with BioDynaMo to enable hybrid agent development. Unlike static binding tools , cppyy requires no manual binding code, generates bindings on-the-fly, and crucially supports cross-inheritance—allowing Python classes to inherit from C++ base classes. The main goal of the project is to demonstrate that BioDynaMo can have simulation agents in Python that seamlessly inherit from and extend existing C++ agent implementations, combining C++ performance with Python flexibility.

Goals

The goal is to explore the possibility of cppyy integration with BioDynaMo by creating a Python class that inherits from a C++ BioDynaMo agent class, overrides virtual methods, and integrates with the simulation’s agent management system, resolving the main technical challenges around cross-language polymorphism and type handling. Below are the steps along the way.

  1. Demonstrate a pre-existing, compiled C++ BioDynaMo agent that can be loaded and its methods can be called from a Python script.
  2. Create a new agent class written entirely in Python that inherits from a C++ BioDynaMo agent.
  3. Test the newly made class in a BioDynaMo simulation.
  4. Polish and document everything.