[Roadmap] Processes and Random Walks

[M-Lampert]

Overview

In PathPy3, we had a module processes that implemented random walks and other processes like epidemic spreading. There have been some small efforts to migrate the module to PathpyG, however, all of the migrated code is still optimised for Pathpy3 and does not make use of torch-operations and the GPU as desired in PathpyG. The goal of this issue is to redesign this module from the ground up with pytorch and the GPU in mind. Therefore, the already implemented parts (see module processes) will not be part of the main branch and, therefore, the first beta version of PathpyG, since we expect many breaking API changes.

Tasks

• Design new abstract-process class that enables the implementation of network processes on simple and higher-order networks
• Implement a random walk process
• Add visualisation capabilities to the processes
• (Optional) Implement other processes like epidemic spreading.