perf: reduce unit test suite runtime by removing redundant visualization and reducing grid sizes

[Jammy2211]

Overview

Profile and optimize the PyAutoGalaxy unit test suite to reduce total runtime by 25-35%. The source code is stable, so this only touches test code and test configuration. Follows the same approach proven on PyAutoFit (61s→41s).

Plan

• Remove redundant visualize_before_fit() calls from aggregator test fixtures — these duplicate work already done by search.fit() and no aggregator test asserts on visualization outputs
• Disable FITS/CSV file output in test visualization config — real disk I/O that no test depends on
• Reduce interferometer simulate-and-fit grid sizes from 51×51 to 31×31 where assertions are grid-size-independent

Detailed implementation plan

Affected Repositories

• PyAutoGalaxy (primary)

Work Classification

Library

Branch Survey

Repository	Current Branch	Dirty?
./PyAutoGalaxy	main	minor (CLAUDE.md, docs, scrap.py)

Suggested branch: feature/unit-test-profiling

Implementation Steps

1. Remove visualize_before_fit() line from 4 files:

   • test_autogalaxy/aggregator/conftest.py:51
   • test_autogalaxy/aggregator/ellipse/conftest.py:51
   • test_autogalaxy/aggregator/imaging/test_aggregator_fit_imaging.py:33
   • test_autogalaxy/aggregator/ellipse/test_aggregator_fit_ellipse.py:33

2. Disable FITS/CSV flags in test_autogalaxy/config/visualize.yaml:

   • fits_fit, fits_galaxy_images, fits_model_galaxy_images → false
   • fits_dirty_images → false
   • csv_reconstruction → false

3. Reduce grid sizes in test_autogalaxy/interferometer/test_simulate_and_fit_interferometer.py:

   • test__perfect_fit: 51×51→31×31 grid, 7→3 UV wavelengths, 7×7→5×5 mesh
   • test__linear_light_profiles_agree: 51×51→31×31 grid

4. Run full test suite to verify all 852 tests pass

Key Files

• test_autogalaxy/aggregator/conftest.py — redundant viz call
• test_autogalaxy/aggregator/ellipse/conftest.py — redundant viz call
• test_autogalaxy/aggregator/imaging/test_aggregator_fit_imaging.py — redundant viz call
• test_autogalaxy/aggregator/ellipse/test_aggregator_fit_ellipse.py — redundant viz call
• test_autogalaxy/config/visualize.yaml — FITS/CSV output flags
• test_autogalaxy/interferometer/test_simulate_and_fit_interferometer.py — grid sizes

Original Prompt

Click to expand starting prompt

Profile and speed up unit tests in @PyAutoGalaxy.

Goal

Reduce the test suite runtime by identifying and optimizing slow tests. The source code is stable, so safe changes to test parameters are acceptable — but prompt the user before changing any numerical assertion values.

Approach (proven on PyAutoFit)

1. Profile: Run pytest <test_dir> --durations=0 -q to get wall-clock times for all tests
2. Identify: Find the top 30+ slowest tests and read their source to understand bottlenecks
3. Categorize: Group by cause — large data arrays, excessive iterations/max_steps, expensive fixtures, real sampler initialization (Dynesty)
4. Optimize (in order of impact):
   • Reduce n_obs, data array sizes, range(N) loop counts where tests use more data than needed for their assertion tolerances
   • Reduce max_steps, n_iters, maxcall values in EP/sampler tests that validate plumbing not convergence
   • Reduce number_of_steps in grid search fixtures to shrink the job count (e.g. 10->5 gives 25 jobs instead of 100)
   • Adjust query/assertion values if data range changes (e.g. if range(100)->range(50), update assertions that referenced value 50 to value 25)
   • Leave tests with inherent sampler initialization overhead (DynestyStatic with maxcall<=5) as-is
5. Verify: Run full suite, confirm all tests pass, compare before/after total time

What NOT to change

• Don't change numerical reference values in assertions without user approval
• Don't change fixture scoping to module/session — this conflicts with function-scoped set_config_path autouse fixtures
• Don't mock real sampler tests — they test actual sampling behavior
• Don't change test_messages.py::test_normal_simplex — the slowness is numerical quadrature integration

Known pattern: FactorValue hash collision

If you encounter a KeyError: <GaussianPrior id=0> in AbstractJacobian.grad(), this is a known bug where FactorValue (id=0) hash-collides with a GaussianPrior that gets id=0. This was fixed in PyAutoFit PR #1182 — check that the fix is present in the installed autofit version.

Target

25-35% reduction in total test runtime. On PyAutoFit this took 61s -> 41s.

[Jammy2211]

Smoke Test Results — 2026-04-06

Workspace	Passed	Failed	Total
autofit_workspace	7	0	7
autogalaxy_workspace	3	4	7
autolens_workspace	4	5	9
autolens_workspace_test	5	0	5

All failures are pre-existing — KeyError in parameter_lists_for_paths when PYAUTOFIT_TEST_MODE=2 encounters MGE/Basis models with profile_list. Not related to this PR's test-only changes.

[Jammy2211]

Shipped — 2026-04-06

PR

• perf: reduce unit test runtime from 82s to 28s #322

Summary

Reduced PyAutoGalaxy unit test runtime from 82s to ~28s (66% reduction). Moved 17 aggregator tests to the new autogalaxy_workspace_test repo as standalone integration scripts matching the autolens_workspace_test/scripts/aggregator/ pattern. Added plt.tight_layout and plt.colorbar monkeypatches to the plot_patch fixture. Reduced interferometer grid sizes, UV wavelengths, and MGE test gaussian counts.

Session Notes

• Aggregator tests can't be optimized in place — the I/O (database creation, FITS writes, visualization) is inherent to what they test. Moving them to a workspace_test repo was the right strategy.
• The plt.tight_layout and plt.colorbar monkeypatches together saved ~12s by skipping expensive matplotlib font metric and colorbar rendering that was wasted since savefig is already no-oped.
• The autogalaxy_workspace_test repo mirrors the autolens_workspace_test pattern: standalone scripts in scripts/aggregator/ with local config, no pytest.

[Jammy2211]

Workspace Test Repo Created — 2026-04-06

New repo: https://github.com/PyAutoLabs/autogalaxy_workspace_test

Contains 4 standalone aggregator integration scripts (17 tests) migrated from test_autogalaxy/aggregator/:

• scripts/aggregator/galaxies.py — GalaxiesAgg
• scripts/aggregator/fit_imaging.py — FitImagingAgg + ImagingAgg
• scripts/aggregator/fit_interferometer.py — FitInterferometerAgg + InterferometerAgg
• scripts/aggregator/ellipse.py — EllipsesAgg + FitEllipseAgg + MultipolesAgg

All tests pass. Matches the autolens_workspace_test/scripts/aggregator/ pattern.