Process reference

vaft.process is the computation layer of VAFT. Almost every function in it takes plain NumPy arrays and scalars and returns arrays, tuples or dataclasses; it does not read or write ODS. The ODS-aware layer lives in vaft.omas (mainly vaft.omas.process_wrapper), which pulls geometry and signals out of an ODS, calls into vaft.process, and writes the results back.

Rule of thumb: vaft.process is the math, vaft.omas.compute_* is the API you usually call.

Every public processing function is documented under one contract, and these pages are generated from those docstrings by python -m vaft.process.catalog. Where a formula page answers what is this quantity and when is it valid, a process page answers how is this input turned into this output: the parameters and returns with their units, the processing steps in order, the defaults that change the result and what kind of value each is, the conventions assumed, the machine scope, the limitations, and the provenance of the method. Nothing here is written by hand, so the site and vaft.process.describe("<name>") always agree.

import vaft.process as P

print(P.describe("repair_clipped_interval"))     # one routine, rendered
P.search("vest.yaml")                             # every routine whose text mentions it
P.list_processes(category="signal_processing")   # imports only that submodule

The discovery layer is loaded on first use only; import vaft.process.signal_processing never touches it.

Categories

The layer is being brought under the contract one submodule at a time (#252). A category gets a reference page when every function in it conforms; until then its count is shown and its page is absent.

Machine-independent or VEST policy

Every function’s Applicability section opens with one of two sentences, and each page badges it. Machine-independent means the caller supplies every machine-specific number – an acquisition rail, a chord length, a processing window – and the routine infers nothing about the hardware; the VEST values are read from vest.yaml by vaft.machine_mapping, never here. VEST-specific means the routine embeds a VEST constant, a VEST column name or a VEST acquisition-era rule, and its page says which and for what shot range.

Where a default came from

A default that changes the result is classified on its page as one of: physical constant, literature value, diagnostic calibration, empirical estimate, validated-workflow default, machine-specific setting, acquisition-era policy, legacy compatibility value, or numerical convenience. A legacy compatibility value is one that arrived with a ported routine and whose derivation is not recorded; the page says so rather than inventing one.

Refreshing this snapshot

From a checkout of the develop branch, run:

python -m vaft.process.catalog --output docs/_data/process_catalog.yml

The snapshot records the SHA-256 of every vaft/process/*.py source file, the private _equilibrium_parametric.py included; documentation validation compares them when VAFT_REGISTRY_SOURCE points to the corresponding source checkout. The same text is available offline as vaft.process.describe("<name>"), vaft.process.search("<text>") and vaft.process.list_processes(category="<category>").

What these objects mean scientifically – which concept a plot draws, which diagnostic measures it, which Data Dictionary path represents it – is generated in the scientific ontology explorer.