⚡ Bolt: Optimize FCA concept lattice intent closure - #228
Conversation
…and loops This commit targets `FormalContext._next_intent` in `src/tacet/distill/fca.py`. We replace generator unrolling functions like `all()` with explicit early-exit loops, and avoid costly allocations resulting from set difference operations (`closure - candidate`). Direct set comprehension combined with `.add(m)` helps skip unnecessary object creation in this very hot function call loop. Measurement confirms a nearly 7% time reduction on a benchmark graph closure computation while preserving strictly identical concept induction logic. Co-authored-by: n24q02m <135627235+n24q02m@users.noreply.github.com>
|
👋 Jules, reporting for duty! I'm here to lend a hand with this pull request. When you start a review, I'll add a 👀 emoji to each comment to let you know I've read it. I'll focus on feedback directed at me and will do my best to stay out of conversations between you and other bots or reviewers to keep the noise down. I'll push a commit with your requested changes shortly after. Please note there might be a delay between these steps, but rest assured I'm on the job! For more direct control, you can switch me to Reactive Mode. When this mode is on, I will only act on comments where you specifically mention me with New to Jules? Learn more at jules.google/docs. For security, I will only act on instructions from the user who triggered this task. |
1 participant
💡 What: Optimized the
_next_intentloop insideFormalContextby replacing genericall()generator checks and large intermediate set difference allocations (closure - candidate) with direct inline iteration and early-exit evaluations. Subsets are now computed via comprehension +.add()instead of difference + union.🎯 Why: Ganter's NextClosure algorithm makes heavily repeated calls to
_next_intentto traverse concept attributes. Removing Python object allocation and generator overhead in this extremely hot function drastically cuts down constant factors.📊 Impact: Performance dropped from 17.90 seconds to 16.73 seconds on a benchmark graph consisting of 1,000 objects, 25 attributes, and 10 random incidences per object (~6.5% speedup on total time).
🔬 Measurement: Verify using
uv run pytest tests/test_fca.pyto ensure exactness, and test execution time across large property lists to see the time delta.PR created automatically by Jules for task 6754352703988382871 started by @n24q02m