Bruno Postle
ac5913151a
Cold-start re-baseline after the DESIGN.md 38.10/38.11 objective change. Single worker (avoids homemaker-py-b8g), scored by the shipped scorer from the programme directory. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01MJ84Feep79Hhm3E4zZJmnB
60 lines
3.3 KiB
Text
60 lines
3.3 KiB
Text
WARNING: 418 m2 per storey needs ~86 m of daylit wall; the plot's non-private perimeter gives 53 m. Roughly 49 m2 of courtyard closes the gap. (DESIGN.md §38.3)
|
|
seed : /home/bruno/src/homemaker-layout/examples/harbor-house/init.dom
|
|
programme : harbor-house
|
|
budget : 500000
|
|
pop : 16
|
|
child_budget : 80
|
|
workers : 1
|
|
rng seed : 0
|
|
leaf sharing : True (factor=3)
|
|
superpose : False
|
|
multi_use : False
|
|
conn grade : False
|
|
use tiers : False
|
|
bridge circulation : False
|
|
ruin recreate : False
|
|
collapse in-search : True
|
|
shapecurve warmstart : False
|
|
shapecurve prune : False
|
|
output : /home/bruno/src/homemaker-layout/examples/harbor-house/coldstart-500000-s0.dom
|
|
[ 80 evals] best 4.30115e-34 (fails 101) via construct/0
|
|
[ 240 evals] best 6.34775e-22 (fails 62) via construct/2
|
|
[ 1760 evals] best 1.0381e-21 (fails 61) via swap 0/lrr
|
|
[ 1920 evals] best 1.07712e-21 (fails 61) via rotate 0/lrll
|
|
[ 2720 evals] best 1.67039e-20 (fails 58) via level_retype 1/lrll<->0/rllr
|
|
[ 2960 evals] best 1.75641e-20 (fails 58) via retype 1/lrrl->ws1
|
|
[ 5120 evals] best 1.79841e-20 (fails 58) via core_undivide noop
|
|
[ 5280 evals] best 3.39494e-20 (fails 57) via crossover rllll<->rllr
|
|
[ 5440 evals] best 3.5999e-20 (fails 57) via core_divide noop
|
|
[ 6720 evals] best 6.90053e-20 (fails 56) via rotate 0/ll
|
|
[ 7840 evals] best 7.05248e-20 (fails 56) via core_divide noop
|
|
[ 8080 evals] best 7.16548e-20 (fails 56) via level_retype 1/rlrl<->0/rllr
|
|
[ 9600 evals] best 1.35685e-18 (fails 52) via retype 0/lrrl->da1
|
|
[ 10000 evals] best 1.41061e-18 (fails 52) via core_divide noop
|
|
[ 15200 evals] best 1.43527e-18 (fails 52) via core_undivide noop
|
|
[ 18400 evals] best 4.70767e-18 (fails 50) via swap 0/rrr
|
|
[ 18560 evals] best 4.78877e-18 (fails 50) via crossover rlrrl<->rlllr
|
|
[ 19200 evals] best 4.96874e-18 (fails 50) via rotate 0/rrl
|
|
[ 22560 evals] best 5.00177e-18 (fails 50) via core_undivide noop
|
|
[ 23360 evals] best 5.02163e-18 (fails 50) via retype 1/rlll->of
|
|
[ 24880 evals] best 5.02826e-18 (fails 50) via core_divide noop
|
|
[ 27200 evals] best 5.02858e-18 (fails 50) via core_undivide noop
|
|
[ 29600 evals] best 8.13881e-18 (fails 50) via level_retype 1/rlrl<->0/llrr
|
|
[ 30960 evals] best 8.2257e-18 (fails 50) via core_undivide noop
|
|
[ 37440 evals] best 8.31661e-18 (fails 50) via core_divide noop
|
|
[ 39200 evals] best 8.34909e-18 (fails 50) via crossover rlr<->rlr
|
|
[ 40240 evals] best 8.43219e-18 (fails 50) via core_undivide noop
|
|
[ 46640 evals] best 8.44241e-18 (fails 50) via core_undivide noop
|
|
[ 50800 evals] best 9.76987e-18 (fails 50) via crossover rllll<->llrr
|
|
[ 51920 evals] best 1.00136e-17 (fails 49) via rotate 0/rll
|
|
[ 52000 evals] best 1.00136e-17 (fails 49) via core_divide noop
|
|
[ 52400 evals] best 1.01413e-17 (fails 49) via swap 0/llr
|
|
[ 61440 evals] best 4.26469e-17 (fails 48) via rotate 0/rlrl
|
|
[ 62800 evals] best 4.28559e-17 (fails 48) via core_divide noop
|
|
[ 63840 evals] best 4.2856e-17 (fails 48) via retype 0/lrlll->da1
|
|
[ 66560 evals] best 4.28816e-17 (fails 48) via core_undivide noop
|
|
[ 67440 evals] best 4.2888e-17 (fails 48) via rotate 0/rlll
|
|
[ 67680 evals] best 1.73994e-16 (fails 46) via level_retype 0/rllll<->1/rrr
|
|
[ 70160 evals] best 1.73995e-16 (fails 46) via core_undivide noop
|
|
[ 72240 evals] best 1.73995e-16 (fails 46) via crossover lrrl<->rrll
|
|
[ 75680 evals] best 1.73995e-16 (fails 46) via core_undivide noop
|