To broaden its reach, VACUUMMS now features a C++ API with Python bindings via Pybind11, packaged as a Conda distribution. This enables researchers to start with a simulated sample (e.g., a single LAMMPS frame), execute a few lines of Python code, and generate both vivid free volume visualizations and statistical void space profiles. Outputs can be seamlessly integrated into Jupyter notebooks, enhancing reproducibility and collaboration. For instance, a user can process a molecular dynamics snapshot to produce images and metrics in under ten lines of code—streamlining workflows for materials scientists.
This presentation will highlight VACUUMMS’s key features—variational pathfinding, visualization, and statistical analysis—while detailing its development journey. We encapsulate C++ core functionality, bridge it to Python with Pybind11, manage builds with CMake, and distribute via Conda, ensuring accessibility across platforms. Applications to amorphous material design and the software’s role in advancing computational materials science will be discussed, with examples drawn from LAMMPS-based simulations.