Refactor unit tests: granularity, descriptive names, parametrize by Jammy2211 · Pull Request #405 · PyAutoLabs/PyAutoLens

Jammy2211 · 2026-03-25T19:53:48Z

Summary

Split omnibus tests into focused single-scenario tests in test_fit_imaging.py, test_tracer.py, and all point-source fit files — each test now covers exactly one scenario so a failure immediately identifies the broken case
Descriptive test names now encode the input condition and expected outcome (e.g. test__figure_of_merit__pixelization_inversion__expected_log_evidence instead of test__fit_figure_of_merit)
pytest.mark.parametrize applied to multi-class/multi-case tests: FitPointDataset position fitting classes, plane_index_via_redshift_from, extract_plane_index_of_profile, upper plane index with light profile
New plot tests covering functions added in the recent rewrite of fit_imaging_plots.py and tracer_plots.py that had zero coverage: subplot_fit_x1_plane, subplot_fit_log10_x1_plane, subplot_tracer_from_fit, subplot_fit_combined, subplot_fit_combined_log10, save_tracer_fits, save_source_plane_images_fits
All numerical assertions preserved exactly unchanged

Files changed

File	Change
`test_autolens/imaging/test_fit_imaging.py`	Split `test__fit_figure_of_merit` (10+ scenarios) + sub_2 variants into individual tests; split dict and subtracted image tests
`test_autolens/lens/test_tracer.py`	Split `test__has`, plane/galaxy sorting tests, convergence/potential/deflection tests; parametrize redshift lookups and extract_plane_index
`test_autolens/imaging/plot/test_fit_imaging_plots.py`	Add 7 new tests for uncovered plot functions; split `test__subplot_of_planes`
`test_autolens/lens/plot/test_tracer_plots.py`	Add 2 new tests for `save_tracer_fits` / `save_source_plane_images_fits` using `tmp_path`
`test_autolens/point/fit/test_fit_dataset.py`	Split by matching/non-matching name; parametrize position fitting classes
`test_autolens/point/fit/test_fluxes.py`	Rename tests to be self-describing
`test_autolens/point/fit/positions/image/test_pair*.py`	Rename + split multi-scenario tests
`test_autolens/point/fit/positions/image/test_abstract.py`	Rename for clarity
`test_autolens/point/fit/positions/source/test_separations.py`	Split 2 inline scenarios into 3 focused tests

Test plan

python -m pytest test_autolens/imaging/test_fit_imaging.py -v
python -m pytest test_autolens/lens/test_tracer.py -v
python -m pytest test_autolens/imaging/plot/ -v
python -m pytest test_autolens/lens/plot/ -v
python -m pytest test_autolens/point/fit/ -v

🤖 Generated with Claude Code

- Split omnibus tests into focused single-scenario tests throughout test_fit_imaging.py, test_tracer.py, and point fit files so failures pinpoint the exact broken scenario rather than a broad test function - Renamed all tests to encode the input condition and expected outcome (e.g. test__figure_of_merit__pixelization_inversion__expected_log_evidence) - Parametrized multi-class/multi-redshift cases with pytest.mark.parametrize: FitPointDataset position fitting classes, plane_index_via_redshift_from, extract_plane_index_of_profile, and upper_plane_index_with_light_profile - Added coverage for new plotting functions lacking any tests: subplot_fit_x1_plane, subplot_fit_log10_x1_plane, subplot_tracer_from_fit, subplot_fit_combined, subplot_fit_combined_log10, save_tracer_fits, save_source_plane_images_fits - All numerical assertions preserved exactly unchanged Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Consistent with test_fluxes.py rename pattern: test names now encode the tracer type used (mock vs real isothermal). Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Copilot

Pull request overview

This PR refactors the AutoLens test suite to make failures more actionable by splitting multi-scenario “omnibus” tests into single-purpose tests, improving test naming, and using pytest.mark.parametrize where appropriate. It also adds new plot and FITS-output tests to cover recently introduced plotting / output helpers that previously had no coverage.

Changes:

Split large multi-scenario tests into focused single-scenario tests and renamed tests to encode conditions + expected outcomes.
Added parametrized tests for multi-class / multi-case behavior (e.g., point position fitting classes, redshift/plane-index helpers).
Added new plot and FITS-writing tests for tracer / fit imaging plotting utilities.

Reviewed changes

Copilot reviewed 12 out of 12 changed files in this pull request and generated 5 comments.

Show a summary per file

File	Description
`test_autolens/point/fit/test_time_delays.py`	Renames / clarifies time-delay fit tests (mock tracer vs real tracer).
`test_autolens/point/fit/test_fluxes.py`	Renames / clarifies flux fit tests (mock tracer vs real tracer).
`test_autolens/point/fit/test_fit_dataset.py`	Splits matching vs non-matching point name tests; parametrizes position-fit class expectations; factors common fixtures.
`test_autolens/point/fit/positions/source/test_separations.py`	Splits combined scenarios into separate tests for mass/no-mass and multi-plane behavior.
`test_autolens/point/fit/positions/image/test_pair_repeat.py`	Splits repeat-allocation behavior into separate focused tests with full quantity assertions.
`test_autolens/point/fit/positions/image/test_pair_all.py`	Renames tests to be scenario-specific and keeps edge-case coverage (inf / duplicates).
`test_autolens/point/fit/positions/image/test_pair.py`	Renames the image-pair residuals test to be explicit about behavior.
`test_autolens/point/fit/positions/image/test_abstract.py`	Refocuses on multi-plane scaling behavior for the abstract fit class.
`test_autolens/lens/test_tracer.py`	Splits and parametrizes tracer behavior tests for clearer single-failure diagnosis.
`test_autolens/lens/plot/test_tracer_plots.py`	Adds tests for FITS-writing helpers (`save_tracer_fits`, `save_source_plane_images_fits`) using `tmp_path`.
`test_autolens/imaging/test_fit_imaging.py`	Major split of `FitImaging` tests into granular, scenario-specific tests while preserving numeric assertions.
`test_autolens/imaging/plot/test_fit_imaging_plots.py`	Adds coverage for new plotting helpers and splits plane-subplot tests for clearer intent.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-03-25T20:04:56Z

test_autolens/point/fit/positions/image/test_pair.py

+import numpy as np
+import pytest
+


numpy and pytest are imported but not used in this test file. Consider removing them to keep the test minimal and reduce noise.

Suggested change

import numpy as np

import pytest

Copilot · 2026-03-25T20:04:57Z

test_autolens/point/fit/test_time_delays.py

-    # Replace 2.5 with expected model time delay from your tracer
-    assert fit.model_time_delays.in_list[1] == pytest.approx(-573.994580905, 1.0e-4)
-    assert fit.log_likelihood == pytest.approx(-22600.81488747, 1.0e-4)
+import numpy as np


numpy is imported but not used in this test file. Consider removing the unused import to keep the test focused.

Suggested change

import numpy as np

Copilot · 2026-03-25T20:04:57Z

test_autolens/point/fit/positions/image/test_abstract.py

+    assert fit_0.model_data[0, :] == pytest.approx(
+        scaling_factor * fit_1.model_data.array[0, :], 1.0e-1


These two assertions are identical, so the second one is redundant and may be a copy/paste mistake (it looks like you intended to assert a different row / index). Consider removing the duplicate or changing the second assertion to cover the other coordinate/component you meant to validate.

Suggested change

assert fit_0.model_data[0, :] == pytest.approx(

scaling_factor * fit_1.model_data.array[0, :], 1.0e-1

assert fit_0.model_data[1, :] == pytest.approx(

scaling_factor * fit_1.model_data.array[1, :], 1.0e-1

Copilot · 2026-03-25T20:04:57Z

test_autolens/point/fit/positions/image/test_abstract.py

-    assert fit_0.model_data[0, :] == pytest.approx(
-        scaling_factor * fit_1.model_data.array[0, :], 1.0e-1
-    )
+import numpy as np


numpy is imported but no longer used in this test file. Removing unused imports keeps the tests easier to maintain and avoids confusion about expected dependencies.

Suggested change

import numpy as np

Copilot · 2026-03-25T20:04:58Z

test_autolens/imaging/test_fit_imaging.py

+    masked_imaging_covariance_7x7, image_7x7, psf_3x3, noise_map_7x7, mask_2d_7x7
+):
+    dataset = al.Imaging(
+        data=image_7x7,
+        psf=psf_3x3,
+        noise_map=noise_map_7x7,
+        over_sample_size_lp=2,
+    )
+    masked_imaging_7x7 = dataset.apply_mask(mask=mask_2d_7x7)
+


This test builds a new Imaging dataset and masked_imaging_7x7, but then fits using masked_imaging_covariance_7x7 instead, so the dataset created here is unused. Either remove this unused setup, or use the newly created masked dataset so the test actually exercises the intended sub-2 oversampling path.

Suggested change

masked_imaging_covariance_7x7, image_7x7, psf_3x3, noise_map_7x7, mask_2d_7x7

):

dataset = al.Imaging(

data=image_7x7,

psf=psf_3x3,

noise_map=noise_map_7x7,

over_sample_size_lp=2,

)

masked_imaging_7x7 = dataset.apply_mask(mask=mask_2d_7x7)

masked_imaging_covariance_7x7,

):

…rtion - test_abstract.py: remove unused `import numpy as np`; fix duplicate assertion so second check covers row [1, :] not [0, :] again - test_fit_imaging.py: remove unused fixtures and dead setup block from covariance test (only masked_imaging_covariance_7x7 is needed) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

…index The second assert was a copy-paste of the first (both checked row [0, :]). Changing it to row [1, :] introduced a genuinely failing assertion since the scaling relationship does not hold for the second position. Simply remove the duplicate instead. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Jammy2211 and others added 2 commits March 25, 2026 19:53

Rename test_time_delays.py tests for clarity

d5cbe62

Consistent with test_fluxes.py rename pattern: test names now encode the tracer type used (mock vs real isothermal). Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Jammy2211 requested a review from Copilot March 25, 2026 19:59

Copilot started reviewing on behalf of Jammy2211 March 25, 2026 19:59 View session

Copilot AI reviewed Mar 25, 2026

View reviewed changes

Jammy2211 and others added 2 commits March 25, 2026 20:12

Jammy2211 merged commit e1d4678 into main Mar 25, 2026
8 checks passed

Jammy2211 deleted the claude/refactor-unit-tests branch April 2, 2026 11:44

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Refactor unit tests: granularity, descriptive names, parametrize#405

Refactor unit tests: granularity, descriptive names, parametrize#405
Jammy2211 merged 4 commits intomainfrom
claude/refactor-unit-tests

Jammy2211 commented Mar 25, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Mar 25, 2026

Uh oh!

Copilot AI Mar 25, 2026

Uh oh!

Copilot AI Mar 25, 2026

Uh oh!

Copilot AI Mar 25, 2026

Uh oh!

Copilot AI Mar 25, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

		assert fit_0.model_data[0, :] == pytest.approx(
		scaling_factor * fit_1.model_data.array[0, :], 1.0e-1

Conversation

Jammy2211 commented Mar 25, 2026

Summary

Files changed

Test plan

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants