Bruno Postle 33d79be3fe Cold-start bootstrap: diverse initial population for blank-slate search ... When the seed is an undivided bare plot (init.dom), auto-generate pop_size random topologies before the memetic loop starts, each evaluated at child_budget. This crosses the zero-feasibility region that single-seed chaining cannot escape — the programme-house cold start was stalling at 18 fails after 2000 evals vs urb-evolve's 6. Auto-detection via seed_root.divided preserves the existing single-seed path for warm starts from existing designs; all previous tests pass unchanged. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>		2026-06-13 23:29:12 +01:00
..
test_dom_corpus.py	Phase 3 gate (homemaker-py-uxz): native fitness 35/35 corpus parity; retire oracle from search	2026-06-13 21:44:42 +01:00
test_driver.py	Cold-start bootstrap: diverse initial population for blank-slate search	2026-06-13 23:29:12 +01:00
test_genome.py	Topology genome: base tree + per-storey deltas + type assignment	2026-06-12 13:52:32 +01:00
test_graph.py	Native fitness: adjacency graph + merge_divided (homemaker-py-3y7)	2026-06-12 21:57:45 +01:00
test_innerloop.py	Fix cma convergence test: assert basin convergence, not final-digit polish	2026-06-12 09:46:13 +01:00
test_operators.py	Fix benchmark cell arg order and mutate_swap on undivided trees	2026-06-12 23:26:22 +01:00
test_oracle.py	Geometry inner loop: batched full-objective ratio optimiser (CMA-ES)	2026-06-12 09:42:24 +01:00