[FIX] image.clean_img() kwargs integration

[MIZwally]

Fixed the bug in issue #4048; reformats the kwargs in clean_kwargs so they can be properly passed into the signal.clean function. Additionally, upon the use of a sample_mask kwarg, the size of the nifti is updated to reflect the new number of volumes after cleaning.

• Closes [BUG] clean_img with sample_mask passed to signal.clean fails because output dimensions are not the same as the input #4048

[github-actions]

👋 @MIZwally 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.

[bthirion]

Can you add a test to enforce the behavior ?

[Remi-Gau]

@MIZwally

You can use the code from @ymzayek here: #4048 (comment)

to add a test in this file:
https://github.com/nilearn/nilearn/blob/main/nilearn/image/tests/test_image.py

See our contributing doc regarding tests: https://nilearn.github.io/stable/development.html#tests

[Remi-Gau]

You should be able to fix code formatting issues with pre-commit:

pip install pre-commit
pre-commit install
pre-commit run -a

[codecov]

Codecov Report

Merging #4051 (f8ac76e) into main (2a0eb8f) will increase coverage by 0.08%.
Report is 4 commits behind head on main.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #4051      +/-   ##
==========================================
+ Coverage   91.50%   91.59%   +0.08%     
==========================================
  Files         143      143              
  Lines       16077    16079       +2     
  Branches     3341     3340       -1     
==========================================
+ Hits        14712    14728      +16     
+ Misses        820      804      -16     
- Partials      545      547       +2     

Flag	Coverage Δ
macos-latest_3.10	91.51% <100.00%> (+0.01%)	⬆️
macos-latest_3.11	91.51% <100.00%> (?)
macos-latest_3.12	91.51% <100.00%> (+0.01%)	⬆️
macos-latest_3.8	?
macos-latest_3.9	91.47% <100.00%> (+0.01%)	⬆️
ubuntu-latest_3.10	91.51% <100.00%> (+0.01%)	⬆️
ubuntu-latest_3.11	91.51% <100.00%> (+0.01%)	⬆️
ubuntu-latest_3.12	91.51% <100.00%> (+0.01%)	⬆️
ubuntu-latest_3.8	?
ubuntu-latest_3.9	91.47% <100.00%> (?)
windows-latest_3.10	91.46% <100.00%> (+0.02%)	⬆️
windows-latest_3.11	91.46% <100.00%> (+0.02%)	⬆️
windows-latest_3.12	91.46% <100.00%> (+0.02%)	⬆️
windows-latest_3.8	?
windows-latest_3.9	91.43% <100.00%> (+0.02%)	⬆️

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

Files	Coverage Δ
nilearn/image/image.py	95.71% <100.00%> (+0.03%)	⬆️

... and 13 files with indirect coverage changes

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

[bthirion]

Thx. You need to add the corresponding test.

[Remi-Gau]

make sure to update the changelog and add your name in the citation.cff file: see instructions here

https://nilearn.github.io/stable/development.html#changelog

[Remi-Gau]

I can see that you made a commit to update the changelog but it does not appear on the PR.

See all the changes listed in this PR.

Also make sure to run black and flake8 on your code: #4051 (comment)

And make sure to add a test: #4051 (comment)

[Remi-Gau]

OK it seems the test is failing because the data fetcher is failing.

@ymzayek I will invetigate but in case you know how your code snippet worked back then and not now let me know.
#4048 (comment)

[ymzayek]

@Remi-Gau sorry haven't had time to review this yet. Just in general I wouldn't recommend that snippet as it is for a test. That snippet was just a quick way to reproduce the error in a python environment and not for pytest. We should always use minimal/generated data for our tests. We don't want to rely on fetching data for tests as this could cause unrelated error due to e.g. network issues

[Remi-Gau]

yeah I was getting to the same conclusion after 5 minutes I am already in "mocking hell" just to get the test data set up.

Will see if we can generate some dummy fmriprep data + confounds with our _utils.

@MIZwally bare with us, while we get this sorted.

[ymzayek]

@MIZwally you can add a test like:

#import at the top
from nilearn._utils.data_gen import _basic_confounds

#replace current test with
def test_clean_img_sample_mask(img_4d_rand_eye):
    """Check sample mask can be passed as a kwarg and used correctly."""
    length = img_4d_rand_eye.shape[3]
    confounds = _basic_confounds(length)
    # exclude last time point
    sample_mask =  np.arange(length - 1)

    img = image.clean_img(
        img_4d_rand_eye,
        detrend=True,
        standardize="zscore_sample",
        confounds=confounds,
        standardize_confounds="zscore_sample",
        **{"clean__sample_mask": sample_mask},
    )
    # original shape is (10, 10, 10, 10)
    assert img.shape == (10, 10, 10, 9)

@Remi-Gau I can refactor following this PR to either make _basic_confounds "public" or make a fixture.

[bthirion]

LGTM, thx !

[ymzayek]

@Remi-Gau did you want to have a final look at this since you've requested changes?

[Remi-Gau]

Good with me.
thanks @MIZwally