Fix overlapping interval union

[samuelbray32]

Description

• Fixes Incorrect interval union for overlapping intervals #1519
  • Correction to logic of Interval.union when overlapping intervals exist
  • Updates test values
• Enforce properly ordered Intervals
  • each interval in the list should have a [start, stop] value with start <= stop

Checklist:

• NA If this PR should be accompanied by a release, I have updated the CITATION.cff
• NA If this PR edits table definitions, I have included an alter snippet for release notes.
• NA If this PR makes changes to position, I ran the relevant tests locally.
• NA If this PR makes user-facing changes, I have added/edited docs/notebooks to reflect the changes
• I have updated the CHANGELOG.md with PR number and description.

[Copilot]

Pull request overview

This PR addresses incorrect behavior in Interval.union when interval lists contain overlapping intervals (fixing #1519) and introduces stricter validation intended to ensure interval lists are well-formed.

Changes:

• Fix Interval.union start/stop detection logic for overlapping intervals and add internal consistency checks.
• Add _extract_only + _validate_intervals to validate extracted interval-like inputs.
• Update interval helper tests to cover the overlapping-union case and adjust existing expected values.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 3 comments.

File	Description
src/spyglass/common/common_interval.py	Updates union algorithm and introduces interval extraction + validation helpers.
tests/common/test_interval_helpers.py	Updates/extends test cases for Interval.union behavior.

Comments suppressed due to low confidence (1)

tests/common/test_interval_helpers.py:379

• New interval validation behavior is introduced via _validate_intervals, but there are no tests asserting that invalid intervals (e.g. [start > stop], including the single-interval case) raise a ValueError, or that unsorted-but-valid inputs are handled as intended. Adding a focused test (or two) here would lock in the new contract and prevent regressions.

@pytest.mark.parametrize(
    "one, two, expected_result",
    [
        (
            np.array([[0, 3]]),
            np.array([[2, 4]]),
            np.array([[0, 4]]),
        ),
        (
            np.array([[0, 1]]),
            np.array([[2, 4]]),
            np.array([[0, 1], [2, 4]]),
        ),
        (
            np.array([[1, 3], [9, 10]]),
            np.array([[0, 2]]),
            np.array([[0, 3], [9, 10]]),
        ),
        (
            np.array([[0, 1]]),
            np.array([[2, 1e11]]),
            np.array([[0, 1], [2, 1e11]]),
        ),
    ],
)
def test_interval_list_union(interval_obj, one, two, expected_result):
    ret = interval_obj(one).union(two).times
    assert np.array_equal(ret, expected_result), "Problem with Interval.union"

---

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

[Copilot]

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 1 comment.

---

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