Module Reference

This page is a source-oriented map rather than an autodoc page. The package imports optional waveform libraries at module import time, so importing every module during a documentation build can fail on machines that only have the core documentation dependencies installed.

Command Entry Point

JenpyROQ.__main__ handles:

  • parsing --config-file;

  • creating output directories;

  • setting the JenpyROQ logger;

  • selecting serial, multiprocessing or MPI execution;

  • setting random seeds;

  • looping over linear and quadratic run types;

  • saving metadata and launching diagnostics.

Configuration Parser

JenpyROQ.initialise handles:

  • the help text shown by python -m JenpyROQ --help;

  • default configuration dictionaries;

  • type conversion from INI strings;

  • list parsing for enrichment-cycle options;

  • activation and skipping of sampled parameters;

  • training-range and test-value parsing;

  • git-state recording.

ROQ Core

JenpyROQ.jenpyroq.JenpyROQ owns the core state:

  • active waveform approximant;

  • parameter index maps;

  • frequency grid;

  • basis tolerances and training-cycle controls;

  • parameter transforms from mc,q or m,q to m1,m2;

  • waveform generation through the selected wrapper;

  • pre-selection and enrichment loops;

  • empirical interpolation nodes and basis interpolants.

Waveform Wrappers

JenpyROQ.waveform_wrappers defines the wrapper registry WfWrapper and the non-LAL approximant names. The registered wrapper class depends on optional imports:

  • LALWf for LALSimulation approximants;

  • WfTEOBResumS for teobresums-giotto;

  • WfMLGW for mlgw-bns-standalone;

  • WfBajes for bajes and NRPMw-family approximants.

Linear Algebra

JenpyROQ.linear_algebra exposes:

Function

Purpose

scalar_product(vec1, vec2, df, weights=1.)

Computes 4 * df * real(vdot(vec1, vec2)) after optional weighting.

normalise_vector(vec, df)

Divides a vector by the square root of its scalar product.

projection(u, v)

Computes the projection helper used during basis construction.

gram_schmidt(basis, vec, df)

Iteratively subtracts projections onto existing basis vectors and normalises the residual.

Post Processing

JenpyROQ.post_processing produces:

  • pre-selection residual plots;

  • empirical interpolation error plots;

  • outlier-count plots;

  • waveform comparison plots;

  • representation-error plots;

  • random validation tests;

  • basis-parameter histograms;

  • empirical-frequency histograms.