pirl-unc / hitlist / 25054171068

Build # 25054171068

Build Type

push

github

Committed by web-flow

Commit Message

v1.29.2: Add --version flag + vectorize sample-match loops (#173) (#174)

Two changes that landed together since both are small:

1. **`hitlist --version`**: top-level CLI now exposes the package
   version. Was reported as missing by a user on v1.29.1.

2. **Vectorize three per-row Python loops** in
   `generate_observations_table()` sample-match resolution (issue #173).
   Replace list comprehensions over the 4.4M-row index with
   `MultiIndex.isin` / `.reindex()` so the membership checks happen in
   pandas C code instead of Python.

Profile, warm cache, on a built observations index (4.4M rows):
- before: 34.8s wall, 6.9M `_isna` calls (3.6s in `isna`)
- after:  24.9s wall, `_isna` no longer in top-20 (~10s saved, 28%)

Output is identical: same row counts, same `sample_match_type`
distribution, same `sample_mhc` fill behavior. The dtype fix ( cast
lookup keys to float to match `obs["_pmid_int"]`'s float64 dtype) is
called out in code comments so future readers don't reintroduce a
silent dtype-mismatch miss.

Coverage Stats

3175 of 4647 relevant lines covered (68.32%)

0.68 hits per line

Source File

Press 'n' to go to next uncovered line, 'b' for previous

Source Not Available