More examples

Two tutorials run on every build of this site, and every number on them is the output of the code above it: Validate NumPy arrays and Custom checks and read-only arrays.

Two further files fit declarations to the shape of a real consumer. They ship with the package, and this site does not execute them:

File	What it is
forecast_scale.py	declarations fitted to the method shape of a real forecasting scaler, on NumPy
locos_capture.py	declarations fitted to the operand relationships of a real attribution run, on PyTorch

locos_capture.py composes storage with older Pydantic validators. It is not the native storage path: for that, see An owned computed result.

Run one from a clone of the repository:

uv run python packages/tenspec/examples/forecast_scale.py