LawSynthGitHub

Tutorials

Longer, narrative walkthroughs that chain many LawSynth features into one story. Each is runnable end to end and deterministic/offline. For short, single-task recipes, see the cookbook.

  1. From a messy sensor CSV to a trusted, shareable model

— profile → prep → discover → explain → validate → backtest → ensemble → report → export → register.

  1. A predator–prey ecology walkthrough

— generate data, discover with the ecology recipe, read the laws, run what-ifs and scenario boards, and export.

  1. Monitoring a running system

— discover a model of "normal", score fresh batches for drift, and decide.

Prerequisites: a built lawsynth CLI and the lawsynth Python SDK (see installation); optionally the lawsynth-notebook package for interactive views. Verify your setup with lawsynth doctor.