Fixed #1151 Forward query_idx from aamp_match to core.mass_absolute

[cagataykenter]

Fixes #1151.

aamp_match accepts query_idx and forwards it to core._find_matches, but never forwards it to core.mass_absolute. The normalized twin, motifs.match, forwards it to both core.mass and core._find_matches.

core._find_matches seeds candidate_idx = query_idx and then breaks when D[candidate_idx] > atol + max_distance, so it depends on D[query_idx] == 0 — which is what core.mass_absolute's query_idx parameter exists to establish.

query_idx was added to core.mass and core.mass_absolute in #860; #856 wired it into core.mass from motifs.match the next day, but aamp_motifs.py was not updated. core.mass_absolute's query_idx branches have been unreachable ever since — which is why they carry # pragma: no cover. They are now covered, so the pragmas are removed.

Changes

• stumpy/aamp_motifs.py — forward query_idx to core.mass_absolute.
• stumpy/core.py — drop the two # pragma: no cover markers on mass_absolute's now-reachable query_idx branches.
• tests/test_aamp_motifs.py — add test_aamp_match_query_idx. There was previously no query_idx test in either tests/test_motifs.py or tests/test_aamp_motifs.py.

Behaviour

For a correct self-join this is a no-op: _mass_absolute uses scipy.spatial.distance.cdist, which returns exactly 0.0 for the identical row, so D[query_idx] was already the minimum and the self-match was already returned first.

What it restores is the behaviour when query_idx does not describe Q — the case mass_absolute's warning exists to catch:

	match	aamp_match before	aamp_match after
rows returned	(1, 2)	(0,)	(1, 2)
self-match first	yes	no	yes
warns	yes	no	yes

Verification

• Correct usage: output byte-identical to main (np.array_equal → True), single- and multi-dimensional.
• aamp_motifs (the internal caller, which always passes query_idx=candidate_idx with the real subsequence): identical motif_distances and motif_indices across 30 seeds.
• tests/test_aamp_motifs.py, tests/test_motifs.py, tests/test_precision.py: all pass under NUMBA_DISABLE_JIT=1.
• The new test alone covers core.py:1375, :1376, :1410, :1411 with no partial branches, so coverage report --fail-under=100 still passes with the pragmas gone.
• black --check, isort --profile black --check-only, flake8: clean.

[gitnotebooks]

Review these changes at https://app.gitnotebooks.com/stumpy-dev/stumpy/pull/1152

[seanlaw]

@cagataykenter Thank you for the thorough explanation and for discovering this issue. It is greatly appreciated! Please allow me some time to review

[seanlaw]

@cagataykenter Thank you again for this excellent contribution! I am curious how you discovered this issue and what you use STUMPY for?