bruno/homemaker-layout
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

homemaker-py-2g7.5: full acceptance-criteria A/B (harbor+maple, 3 seeds, 20k budget)

Browse source 
CP-SAT assignment does not clearly beat greedy end-to-end: worse on
harbor-house hard fails, roughly a wash on maple-court, and the reassign
operator fired in only 1 of 18 runs. Keeps assign_solver="greedy" and
enable_reassign=False as defaults; closes the bead's remaining acceptance
criterion. Raw results in experiments/results/.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01LSwQwpEaHFBkeVSDDWd75S
This commit is contained in:
Bruno Postle 2026-08-05 00:01:00 +01:00
parent cf634ae949
commit ade7f9a131
2 changed files with 40 additions and 21 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

42
.beads/issues.jsonl
View file

File diff suppressed because one or more lines are too long

19
experiments/results/ab_cpsat_assign_20k_harbor_maple.log Normal file
View file

@ -0,0 +1,19 @@
=== harbor-house 2026-08-04T10:52:41+01:00 ===
seed 0: | greedy hard=8 soft=36 fit=5.386e-15 1862.0s | cpsat hard=15 soft=22 fit=1.388e-13 1869.1s | reassign hard=11 soft=36 fit=5.057e-16 2037.1s reassign_fired=0
seed 1: | greedy hard=13 soft=31 fit=1.422e-16 1487.8s | cpsat hard=13 soft=35 fit=1.881e-16 1774.5s | reassign hard=5 soft=39 fit=3.977e-16 1816.8s reassign_fired=0
seed 2: | greedy hard=7 soft=39 fit=2.434e-17 1784.6s | cpsat hard=11 soft=39 fit=6.704e-18 1709.9s | reassign hard=9 soft=31 fit=9.487e-15 1673.1s reassign_fired=0
budget=20000 n_seeds=3 programme=harbor-house
greedy : mean hard=9.333 soft=35.333 fitness=1.85089e-15 wall=1711.5s
cpsat : mean hard=13.000 soft=32.000 fitness=4.63301e-14 wall=1784.5s
reassign : mean hard=8.333 soft=35.333 fitness=3.46334e-15 wall=1842.3s mean_reassign_fired=0.0
=== maple-court 2026-08-04T15:19:37+01:00 ===
seed 0: | greedy hard=30 soft=72 fit=6.825e-34 2735.1s | cpsat hard=23 soft=57 fit=5.002e-26 3384.4s | reassign hard=25 soft=79 fit=4.567e-34 3221.9s reassign_fired=0
seed 1: | greedy hard=13 soft=76 fit=3.645e-30 2981.7s | cpsat hard=25 soft=73 fit=1.047e-32 3229.9s | reassign hard=49 soft=71 fit=1.48e-39 2921.1s reassign_fired=1
seed 2: | greedy hard=25 soft=69 fit=6.827e-31 2882.0s | cpsat hard=19 soft=75 fit=8.893e-32 3404.5s | reassign hard=32 soft=83 fit=1.06e-37 3071.1s reassign_fired=0
budget=20000 n_seeds=3 programme=maple-court
greedy : mean hard=22.667 soft=72.333 fitness=1.44285e-30 wall=2866.3s
cpsat : mean hard=22.333 soft=68.333 fitness=1.66721e-26 wall=3339.6s
reassign : mean hard=35.333 soft=77.667 fitness=1.5226e-34 wall=3071.4s mean_reassign_fired=0.3
=== ALL DONE 2026-08-04T23:03:32+01:00 ===