Testing suite

This page describes the current MORS test suite and how to run it. The suite covers the spectral synthesis pipeline in full, the Spectrum class, and regression tests for the Spada and Baraffe stellar evolution tracks with two ODE solvers. The physical model (physicalmodel.py), rotational evolution solver (rotevo.py), and parameter handling are not yet covered by dedicated tests.

Running the tests

Prerequisites

Install MORS with the development dependencies:

pip install -e .[develop]

Download the stellar evolution data (required for Spada and Baraffe track tests):

export FWL_DATA=/path/to/your/data
mors download all

Run the full suite

coverage run -m pytest

Run a specific file

pytest tests/test_spada_RB.py
pytest tests/test_spectrum.py

Run with verbose output

pytest -v

View coverage report

coverage run -m pytest
coverage report

Current test files

test_spada_FE.py: Spada tracks, Forward Euler solver

Runs four parametrised integration tests using the Forward Euler ODE solver. Each test creates a mors.Star at a given mass and initial rotation rate and checks three quantities at a given age against regression values from a known-good run:

Tolerance: rtol=1e-6. The test calls mors.DownloadEvolutionTracks('Spada') to ensure data are present.

test_spada_RB.py: Spada tracks, Rosenbrock solver

Identical structure to test_spada_FE.py but uses the default RosenbrockFixed solver. The expected values differ from the Forward Euler test because the two solvers produce slightly different numerical results at the same tolerance settings.

test_baraffe.py: Baraffe tracks

Runs two parametrised integration tests using Baraffe et al. (2015) tracks. Each test creates a mors.BaraffeTrack at a given mass and checks three quantities at a given age against regression values:

Tolerance: rtol=1e-5. Note that Baraffe tracks use time in years, not Myr.

test_spectrum.py: Spectrum class and helpers

Unit tests for mors.spectrum. No stellar evolution data required. Covers:

• WhichBand: wavelength-to-band lookup, including overlap regions
• ScaleToSurf / ScaleTo1AU: round-trip scaling consistency
• PlanckFunction_surf: monotonically increasing with temperature, positive values
• Spectrum.LoadDirectly: ascending wavelength ordering, NaN/zero flux sanitisation, binwidth length
• Spectrum.CalcBandFluxes: band integration with constant flux (analytically verifiable)
• Spectrum.ExtendShortwave: extension length, wavelength bounds, constant flux value
• Spectrum.ExtendPlanck: extension length, wavelength bounds, positive flux
• TSV round-trip: write and reload with tolerance matching the %1.4e format

test_synthesis.py: Spectral synthesis

Unit tests for mors.synthesis. Uses monkeypatch to replace Value, Percentile, Lxuv, Lbol, and PlanckFunction_surf with lightweight fakes, so no stellar evolution data are required. Covers:

• GetProperties: flux budget consistency: \(F = L / (4\pi \, \mathrm{AU}^2)\), UV remainder definition, Planck band integral
• CalcBandScales: scale factor \(Q_k = F_k^\mathrm{hist} / F_k^\mathrm{modern}\) for each band
• CalcScaledSpectrumFromProps: correct band scale factor applied per wavelength, including overlap regions
• FitModernProperties: correct initial guess shape, correct unpacking of the minimize result for both fixed-age and free-age cases

CI matrix

Tests run automatically on every push and pull request to main via GitHub Actions:

Coverage is reported in the GitHub Actions summary. A coverage badge is generated from the ubuntu-latest + Python 3.12 run on main.