MAINT: Deselect logistic regression cases that fail from small numeric differences

[david-cortes-intel]

Description

Deselects a few conformance tests that started failing with deviances in the order of 1e-6 after this PR: #2752

---

Checklist:

Completeness and readability

• Git commit message contains an appropriate signed-off-by string (see CONTRIBUTING.md for details).
• I have resolved any merge conflicts that might occur with the base branch.

Testing

• All CI jobs are green or I have provided justification why they aren't.

[david-cortes-intel]

/intelci: run

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

Flag	Coverage Δ
azure	80.47% <ø> (ø)
github	82.07% <ø> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.
see 3 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[ethanglaser]

This looks good but it appears neither deselection is picked up. For top one I am guessing it is because we already have this exact same deselection https://github.com/uxlfoundation/scikit-learn-intelex/blob/main/deselected_tests.yaml#L298, so maybe it makes sense to remove the original line and version from this one. For second line I think the deselection logic is dicey with parenthesis and possibly quotes. Best to confirm it works locally.

[david-cortes-intel]

Fixed.

[david-cortes-intel]

/intelci: run

[david-cortes-intel]

@ethanglaser @icfaust Looks like it's still executing this deselected test for sklearn==1.0.2:

test_estimators[StackingClassifier(estimators=[('est1',LogisticRegression(C=0.1)),('est2',LogisticRegression(C=1))])-check_sample_weights_invariance(kind=ones)]

Any ideas about what's happening here?

[ethanglaser]

For second line I think the deselection logic is dicey with parenthesis and possibly quotes. Best to confirm it works locally. I did take a brief look at this late last week and was unable to figure out the best way to deselect this specific subset of tests but I can take a closer look today.

[ethanglaser]

Update: issue is with the whitespace between "<" and "1.1"

Also I suggest fully writing out the test name, as there are 50+ tests that get deselected by the current string

[ethanglaser]

/intelci: run

[david-cortes-intel]

Update: issue is with the whitespace between "<" and "1.1"

Also I suggest fully writing out the test name, as there are 50+ tests that get deselected by the current string

There's dozens of failures from parametrizations of this test depending on the machine. This should cover all the cases where in theory these small differences could show, even if they aren't failing right now.