Start here

Complete, runnable experiments. Each walks through a real file under examples/ line by line — open the file in your editor and read alongside, since the code blocks are pulled from it and cannot drift.

• EMG classification - examples/synthetic/emg_classification.py. Two-class Rest/Fist with CatBoost on RMS+MAV features. The best first read: it establishes the App + Stream + Pipeline pattern every other page assumes.
• EMG regression with VHI - examples/synthetic/emg_regression.py. Continuous 5-DOF target with a CatBoost multi-output regressor, recorded via VHI's gRPC control plane and pushed back via the LSL data plane.
• Examples directory - one-paragraph summary of each runnable example under examples/synthetic/, with the run command and what's tweakable.

Once these make sense, the rest of the guides are task recipes, grouped by job: driving a device, recording, models and features, extending the framework.