ezmsg
Messaging and Multiprocessing.
ezmsg is a pure-python implementation of a directed acyclic graph (DAG) pub/sub messaging pattern based on LabGraph which is optimized and intended for use in constructing real-time neural signal processing implementations. ezmsg implements much of the LabGraph API (with a few notable differences), and owes a lot of its design to the LabGraph developers/project. Afterall, imitation is the sincerest form of flattery.
The ezmsg library eases the creation of modular systems whose elements can be switched out easily. Consider the use case of building a processing pipeline where you will be experimenting with adding and removing steps until you find an optimal workflow. ezmsg allows you to easily separate each step into a discrete entity and piece together a workflow from those entities. Check out the Examples to see how this works.
Why use ezmsg over a comparable tool? ezmsg is extremely fast and uses Python’s (currently) new multiprocessing.shared_memory module to facilitate efficient message passing without C++ or any compilation/build tooling. It is easy to install and contains less boilerplate than similar frameworks. It also provides a framework for building processing pipelines while keeping best practices in mind.
Check out this Google Colab for an example ezmsg notebook to experiment with.