Compute as code

Author reusable checklist definitions, evaluate compute expressions, and publish derived datasets with nominal.experimental.compute_as_code. This module is experimental and requires pip install 'nominal[compute]'.

For the checklist workflow, see Authoring checklists as code. The propulsion test-stand example creates and revises thermal, pressure, and controller checks, evaluates a run, and inspects generated events.

API Reference

Module: nominal.experimental.compute_as_code.

Classes

ChannelBinding	A channel resolved relative to the run or asset where a checklist executes.
CheckDefinition	A reusable failure condition written in the frontend compute language.
ChecklistBuilder	Collect source checks and publish complete checklist revisions.

Functions

commit_derived_definition	Replace a derived dataset's definition by creating a new commit.
compute_series	Compute a derived series by applying a nominal.compute expression to a set of channels.
create_derived_dataset	Create a derived dataset defined by a nominal.compute graph.
get_derived_definition	Fetch a dataset's derived definition (its compute spec plus the commit that produced it).