[MAINT] Use rng fixture for setting up `np.random.RandomState()` - part 1 #4052

ymzayek · 2023-10-13T14:33:32Z

Part of use rng fixture #3926 (to avoid big diff)

Changes proposed in this pull request:

Use rng fixture in modules _utils, connectome, datasets, decoding, decomposition, glm

github-actions · 2023-10-13T14:40:15Z

👋 @ymzayek Thanks for creating a PR!

Until this PR is ready for review, you can include the [WIP] tag in its title, or leave it as a github draft.

Please make sure it is compliant with our contributing guidelines. In particular, be sure it checks the boxes listed below.

PR has an interpretable title.
PR links to Github issue with mention Closes #XXXX (see our documentation on PR structure)
Code is PEP8-compliant (see our documentation on coding style)
Changelog or what's new entry in doc/changes/latest.rst (see our documentation on PR structure)

For new features:

There is at least one unit test per new function / class (see our documentation on testing)
The new feature is demoed in at least one relevant example.

For bug fixes:

There is at least one test that would fail under the original bug conditions.

We will review it as quick as possible, feel free to ping us with questions if needed.

nilearn/connectome/tests/test_connectivity_matrices.py

nilearn/decoding/tests/test_decoder.py

nilearn/glm/tests/test_regression.py

Co-authored-by: Remi Gau <remi_gau@hotmail.com>

codecov · 2023-10-13T16:03:19Z

Codecov Report

Merging #4052 (4af8880) into main (96755a4) will not change coverage.
Report is 3 commits behind head on main.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #4052   +/-   ##
=======================================
  Coverage   91.60%   91.60%           
=======================================
  Files         143      143           
  Lines       16072    16072           
  Branches     3339     3339           
=======================================
  Hits        14722    14722           
  Misses        804      804           
  Partials      546      546

Flag	Coverage Δ
macos-latest_3.10	`91.49% <ø> (ø)`
macos-latest_3.11	`91.49% <ø> (ø)`
macos-latest_3.12	`91.49% <ø> (ø)`
macos-latest_3.8	`91.45% <ø> (ø)`
macos-latest_3.9	`91.46% <ø> (ø)`
ubuntu-latest_3.10	`91.49% <ø> (ø)`
ubuntu-latest_3.11	`91.49% <ø> (ø)`
ubuntu-latest_3.12	`91.49% <ø> (ø)`
ubuntu-latest_3.8	`91.45% <ø> (ø)`
ubuntu-latest_3.9	`91.46% <ø> (ø)`
windows-latest_3.10	`?`
windows-latest_3.11	`91.43% <ø> (ø)`
windows-latest_3.12	`?`
windows-latest_3.8	`91.39% <ø> (ø)`
windows-latest_3.9	`91.40% <ø> (ø)`

Flags with carried forward coverage won't be shown. Click here to find out more.

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

bthirion

LGTM, thx !

bthirion · 2023-10-14T07:33:40Z

nilearn/decoding/tests/test_searchlight.py

@@ -5,14 +5,14 @@
 from nibabel import Nifti1Image
 from sklearn.model_selection import KFold

+from nilearn.conftest import _rng
 from nilearn.decoding import searchlight


 def _make_searchlight_test_data(frames):


Why not make rng an argument of the test in that case ?

AFAICT because rng is a fixture and fixtures can only be passed as arguments to a test function (or another fixture) but not to a utility function (pretty sure it is more complicated than that but this is the super short version)

bthirion · 2023-10-14T07:37:01Z

nilearn/glm/tests/test_regression.py

    # If design matrix contains an intercept, the
    # mean of the residuals should be 0 (short of
    # some numerical rounding errors)
-    Xintercept[:, 0] = 1
-    model = OLSModel(design=Xintercept)
+    X[:, 0] = 1


Isn't X modified ?

yeah but it gets regenerated every time the fixture is invoked so eventhough it is modified in this test, the next test will get a "fresh new X".

Start using rng fixture

3abdb0e