diff --git a/.beads/issues.jsonl b/.beads/issues.jsonl index c7f95df..3d4ee5a 100644 --- a/.beads/issues.jsonl +++ b/.beads/issues.jsonl @@ -22,7 +22,7 @@ {"id":"homemaker-py-1p0","title":"Geometry inner loop: full-objective equal-offset ratio optimiser","description":"DESIGN.md §5.1, §7 Phase 1. Productionise experiments/optimize_fullfitness.py into homemaker: optimise(topology, x0=None) -\u003e (geometry, fitness). DOF = equal-offset division ratios of free branches (solver.free_branches, lowest-storey cut ownership), clipped to [eps, 1-eps]. Objective = full oracle fitness (never a proxy — §4.2 falsified). Must support warm-start x0 (§5.6) and a population/batch evaluation mode so each iteration scores via one batched oracle call (§4.6).","acceptance_criteria":"Reproduces or exceeds §4.5 gains (x1.24–x1.67, no new failures) on 2f45907, candidate-002, c964435; works as a library call on any corpus .dom","status":"closed","priority":1,"issue_type":"feature","assignee":"Bruno Postle","owner":"bruno@postle.net","created_at":"2026-06-11T23:36:58Z","created_by":"Bruno Postle","updated_at":"2026-06-12T08:46:31Z","started_at":"2026-06-12T00:14:19Z","closed_at":"2026-06-12T08:46:31Z","close_reason":"innerloop.optimise() lands: batched CMA-ES sigma ladder (0.05/0.15, IPOP popsize doubling, deterministic seeding) over equal-offset free-branch ratios vs full oracle fitness; warm-start x0 supported. Acceptance vs unprojected originals: x1.65/x1.66/x1.58 against bars x1.24/x1.67/x1.59, no new failures, 46 oracle calls vs NM's 200. Two near-bar results accepted as reproduced-within-noise (1% tol) — draw spread brackets the single-NM-draw bars; approved by Bruno 2026-06-12. Gotchas: equal-offset projection of legacy unequal cuts loses fitness/adds failures (midpoint projection used); pycma seed=0 means clock-seeded.","dependencies":[{"issue_id":"homemaker-py-1p0","depends_on_id":"homemaker-py-av5","type":"blocks","created_at":"2026-06-12T00:39:33Z","created_by":"Bruno Postle","metadata":"{}"}],"dependency_count":1,"dependent_count":3,"comment_count":0} {"id":"homemaker-py-8cs","title":"Experiment: warm-vs-cold start of inner loop (Lamarckian inheritance)","description":"DESIGN.md §5.6, §4.6. Warm-starting a child topology's inner loop from the parent's optimised ratios is the main lever for cutting per-topology cost (~3 min/topology cold). Apply single topology mutations to optimised corpus designs, re-optimise warm (surviving cuts keep values, new cuts get heuristic defaults) vs cold, compare oracle-call counts to convergence at equal final fitness.","acceptance_criteria":"Speedup factor measured across \u003e=10 mutated topologies; decision recorded (expect order-of-magnitude; if \u003c2x, revisit §4.6 Phase-2 scoping)","notes":"Experiment script committed (experiments/warm_vs_cold.py, 1cc86c8) and machinery validated oracle-free; one mutated child scored through the oracle OK. Waiting on homemaker-py-gp2 reference run to finish, then execute under URB_NO_OCCLUSION=1 (3 parents x 400 evals + 12 children x 2 x 200 evals, ~1.5-2 h oracle time). Default budgets: parent 400, child 200; target = evals to 95% of best final.","status":"closed","priority":1,"issue_type":"task","owner":"bruno@postle.net","created_at":"2026-06-11T23:36:58Z","created_by":"Bruno Postle","updated_at":"2026-06-12T11:44:45Z","closed_at":"2026-06-12T11:44:45Z","close_reason":"Measured (URB_NO_OCCLUSION=1, parent budget 400, child 200, 12 single mutations across 3 designs): cold start reached 95% of warm final in 0/12 cases within budget — speedup unbounded at practical budgets; warm finals beat cold finals x1.2-x4 in 12/12; 6/12 warm starts were within 95% at 1 eval (near-neutral mutations). Decision: Lamarckian warm-starting is MANDATORY in the memetic driver (homemaker-py-b39), not an optimisation; cold starts produce strictly worse geometry at equal budget. Note: 2 undivides were exactly fitness-neutral (same-type merge == Merge_Divided equivalence) — locality datum for homemaker-py-nyb.","dependencies":[{"issue_id":"homemaker-py-8cs","depends_on_id":"homemaker-py-1p0","type":"blocks","created_at":"2026-06-12T00:39:34Z","created_by":"Bruno Postle","metadata":"{}"}],"dependency_count":1,"dependent_count":0,"comment_count":0} {"id":"homemaker-py-av5","title":"Batched oracle: score many .dom files per invocation","description":"oracle.py currently scores one .dom per urb-fitness.pl call (~1.65 s/dom). DESIGN.md §4.6: batching amortises Perl startup to ~0.99 s/dom and is required so population/batch optimisers can score a whole generation in one oracle call. Extend oracle.py with a batch API: write N .dom files, one perl invocation, parse N .score/.fails pairs. Keep the single-file path for compatibility.","acceptance_criteria":"Batch of 35 corpus files scores in one perl invocation; per-file results identical to single-file calls; measured s/dom reported","status":"closed","priority":1,"issue_type":"task","assignee":"Bruno Postle","owner":"bruno@postle.net","created_at":"2026-06-11T23:36:56Z","created_by":"Bruno Postle","updated_at":"2026-06-12T00:14:06Z","started_at":"2026-06-11T23:50:40Z","closed_at":"2026-06-12T00:14:06Z","close_reason":"score_batch() lands in oracle.py; 35-file corpus parity verified single-vs-batch (1e-12 rel fitness, exact fail sets); 0.98 s/dom batched vs 1.27 single, x1.30","dependency_count":0,"dependent_count":1,"comment_count":0} -{"id":"homemaker-py-qi6","title":"Circulation placement to clear not-connected / access fails","description":"The 94g finish-time collapse cannot touch the \"not-connected\" (level N not\nconnected) and related access/inaccessible fails: they are properties of the\nCIRCULATION skeleton (c/o/s cells), which the collapse deliberately never\nrelabels (they form the structure the room assignment is layered onto). On the\nharbor-house best layout 2 of the 15 fails are not-connected (levels 0 and 1);\nthese are out of scope for any label optimisation.\n\nGOAL: a search/repair step that places or reshapes circulation so every usable\nspace is reachable and each storey's circulation graph is connected. Candidate\nmechanisms: (a) a mutation/operator that inserts a circulation cell to bridge a\ndisconnected component (graph.py already computes connected components +\nconnected_circulation); (b) a finish-time repair that re-types a boundary cell to\ncirculation where it reconnects the graph at least net-fail cost; (c) bias the\nouter search toward connected topologies via the graded signal.\n\nInteracts with 94g: circulation placement changes which cells are skeleton vs\nassignable, so it should run BEFORE the label collapse (collapse then optimises\nlabels over the improved skeleton). Also interacts with the public-access pin\n(94g) — better circulation placement can supply invariant inside public access,\nremoving the need to pin a room provider.\n\nMeasure on the 6 evolved layouts from the 94g sweep (not-connected + access +\ninaccessible fail counts). Related: 94g, homemaker-py-2g5.","status":"open","priority":2,"issue_type":"feature","owner":"bruno@postle.net","created_at":"2026-07-18T10:12:27Z","created_by":"Bruno Postle","updated_at":"2026-07-18T10:12:27Z","dependencies":[{"issue_id":"homemaker-py-qi6","depends_on_id":"homemaker-py-94g","type":"discovered-from","created_at":"2026-07-18T11:12:27Z","created_by":"Bruno Postle","metadata":"{}"}],"dependency_count":0,"dependent_count":0,"comment_count":0} +{"id":"homemaker-py-qi6","title":"Circulation placement to clear not-connected / access fails","description":"The 94g finish-time collapse cannot touch the \"not-connected\" (level N not\nconnected) and related access/inaccessible fails: they are properties of the\nCIRCULATION skeleton (c/o/s cells), which the collapse deliberately never\nrelabels (they form the structure the room assignment is layered onto). On the\nharbor-house best layout 2 of the 15 fails are not-connected (levels 0 and 1);\nthese are out of scope for any label optimisation.\n\nGOAL: a search/repair step that places or reshapes circulation so every usable\nspace is reachable and each storey's circulation graph is connected. Candidate\nmechanisms: (a) a mutation/operator that inserts a circulation cell to bridge a\ndisconnected component (graph.py already computes connected components +\nconnected_circulation); (b) a finish-time repair that re-types a boundary cell to\ncirculation where it reconnects the graph at least net-fail cost; (c) bias the\nouter search toward connected topologies via the graded signal.\n\nInteracts with 94g: circulation placement changes which cells are skeleton vs\nassignable, so it should run BEFORE the label collapse (collapse then optimises\nlabels over the improved skeleton). Also interacts with the public-access pin\n(94g) — better circulation placement can supply invariant inside public access,\nremoving the need to pin a room provider.\n\nMeasure on the 6 evolved layouts from the 94g sweep (not-connected + access +\ninaccessible fail counts). Related: 94g, homemaker-py-2g5.","notes":"LANDED (2026-07-18) mechanism (c) — graded circulation-connectivity signal (DESIGN §18). graph.circulation_connectivity(G) = largest-circ-component fraction [0,1]; summed over storeys it rides the score_with_grade proximity channel, gated by conf flag conn_grade (replaces the §11.4 leaf-grade on that channel). Secondary comparator key (-n_fails, grade, fitness) only — scalar fitness and fail count byte-identical (verified). Wired conn_grade through driver _overrides_for/_fitness_for/_evaluate/search (enabling it implies the grade key); evolve --conn-grade (HOMEMAKER_CONN_GRADE, default OFF). 9 new tests (tests/test_conn_grade.py), 276 pass; 60-eval CLI smoke confirms plumbing. PENDING: full-budget A/B to confirm the gradient actually pulls runs toward connected circulation and clears 'not connected' fails; if graded key alone insufficient, follow-on is an insert/relocate-circulation mutation operator (mechanism a) which now has a gradient to climb. Keeping in_progress until the A/B verdict.","status":"in_progress","priority":2,"issue_type":"feature","assignee":"Bruno Postle","owner":"bruno@postle.net","created_at":"2026-07-18T10:12:27Z","created_by":"Bruno Postle","updated_at":"2026-07-18T17:06:53Z","started_at":"2026-07-18T12:17:08Z","dependencies":[{"issue_id":"homemaker-py-qi6","depends_on_id":"homemaker-py-94g","type":"discovered-from","created_at":"2026-07-18T11:12:27Z","created_by":"Bruno Postle","metadata":"{}"}],"dependency_count":0,"dependent_count":0,"comment_count":0} {"id":"homemaker-py-7fm","title":"Geometry/topology search for shape-intrinsic fails (long-thin useless cells)","description":"The finish-time collapse (94g) proved that a large share of the harbor-house best\nlayout's residual fails are GEOMETRY-INTRINSIC, not label slack: long-thin cells\nthat are useless whatever room usage is assigned. Their width / proportion /\ncrinkliness fails cannot be cleared by any relabelling (94g searches labels only,\nnever geometry) — confirmed by collapse_global clearing only ~2-3 of 15 fails on\nthe best layout, the rest shape- or building-level bound.\n\nGOAL: a search/repair operator that reshapes such cells so the space becomes\nusable — e.g. re-solving a subtree's division ratios, merging a sliver into a\nneighbour, or a targeted division-ratio mutation biased by the offending factor\n(narrowest-width, aspect, crinkliness). Unlike the collapse this MUST move\ngeometry (division ratios / tree shape), and must be evaluated for net fail-count\neffect (a reshape that fixes width may add size elsewhere — same shuffle risk the\n94g threshold objective addressed for labels).\n\nSCOPE: width/proportion/crinkliness fails on inside room cells whose geometry no\nroom type can satisfy. Explicitly NOT relabelling (that is 94g, done). Candidate\nmechanisms: (a) inner-loop solve already optimises ratios — check why it leaves\ndegenerate cells (local optimum? target-dim conflict?); (b) a finish-time\n\"deslim\" operator + re-solve; (c) an operators.py mutation weighted toward high-\naspect leaves. Measure on the same 6 evolved layouts used for the 94g sweep.\n\nSee bd memory collapse-global-94g-and-any-label-usage-optimisation for the\nlabel-vs-geometry boundary. Related: 94g (label collapse), homemaker-py-2g5\n(occlusion/daylight rebuild, which feeds crinkliness).","status":"open","priority":2,"issue_type":"feature","owner":"bruno@postle.net","created_at":"2026-07-18T10:11:43Z","created_by":"Bruno Postle","updated_at":"2026-07-18T10:11:43Z","dependencies":[{"issue_id":"homemaker-py-7fm","depends_on_id":"homemaker-py-94g","type":"discovered-from","created_at":"2026-07-18T11:11:42Z","created_by":"Bruno Postle","metadata":"{}"}],"dependency_count":0,"dependent_count":0,"comment_count":0} {"id":"homemaker-py-94g","title":"Global cell↔room collapse: generalise 9o5 matching to all spaces (WFC-style)","description":"Generalise the 9o5 superposition/collapse from interchange-classes to a GLOBAL cell↔room assignment: evolution searches unlabelled floorplans (tree shape + circulation/room/outside), a collapse function optimally LABELS each candidate. Mechanically an assignment problem — N cells (computed area/width/proportion/level/graph-pos) ↔ M required rooms (target dims + level + adjacency) — minimising total fail-cost; reuse Fitness._best_assignment (Hungarian/brute) but over the whole leaf set instead of one class. WFC framing = the constructive algorithm: each cell a distribution over types, PROPAGATE hard constraints exactly (level, requires_below service stacks, must-have adjacency) to prune, OBSERVE lowest-entropy cell, collapse to best-fitting room weighted by fit-quality, backtrack on contradiction.\n\nMOTIVATION (harbor-house evolved-3M-nols-3.dom, best layout, 15 fails): ~11 of 15 are LABEL-RELATIVE — 4 size (0/lrrll,0/rllll,1/lrlr,1/lrrll), 3 width (0/rlrrlr,0/rrllrr,0/rrrr), 2 proportion (0/rlrlr,1/rlrlr), 2 wrong-level (me1 L1-\u003e0, r L0-\u003e1). A cell fails size/width/proportion because the room ASSIGNED to it wants dims the cell lacks; relabel to a room it fits and the fail vanishes. Wrong-level is a HARD constraint a level-respecting collapse honours for free. Only 4 are shape-intrinsic and out of scope: crinkliness x2 (0/llll,0/rlllr) + not-connected x2 (level 0/1). Realistic target: 15 -\u003e ~4-6.\n\nWHY IT MAY SUCCEED WHERE 9o5 WAS FALSIFIED (xi7): 9o5 went negative because (1) auto-derived classes were semantically wrong (harbor-house 8-code chain, see homemaker-py-b3v) and (2) collapse perturbed feasibility (ON ADDS fails, 38v33/48v43). A GLOBAL, hard-constraint-respecting collapse sidesteps both: no fragile similarity classes; never violates level/adjacency/stack so it cannot ADD feasibility fails — only improve or match.\n\nRISK: search-landscape flattening. fitness = max-over-labellings makes the objective flatter/noisier (many topologies collapse to similar best scores), removing the gradient evolution climbs — the likely cause of 9o5's negative verdict, AMPLIFIED at full scope. Mitigations to A/B: (a) collapse only at FINISH (search on committed types, one relabel pass at end — cheap, strictly cannot worsen final score); (b) warm-start collapse from evolved labels as a local polish.\n\nRECOMMENDED FIRST STEP (cheapest, ~1 day, strictly cannot worsen final layout): FINISH-TIME global collapse — after a normal run, one optimal cell-\u003eroom assignment over the full leaf set with hard constraints enforced, then re-score. Measures empirically how many of the 11 label-relative fails are real slack vs already-optimal. If it clears a meaningful chunk, justifies the in-search WFC collapse + the landscape-flattening A/B. Does NOT fix crinkliness/connectivity (need geometry + circulation-placement work, file separately).\n\nFiles: fitness.py (_best_assignment, collapse_superposition), programme.py (constraints), driver.py (finish hook), graph.py (adjacency for propagation).","notes":"WIRED + PUBLIC-ACCESS TERM DONE.\n1. Public-access pin (preserve_public_access=True, default): when the building's\n ONLY street access is an l/k ROOM neighbour of a public outside leaf (no\n circulation fallback — the existential building check the per-leaf objective\n can't see), that room leaf is PINNED (kept + its demand slot decremented) so\n the collapse can't drop \"no outside public access\". On the best layout this\n turns 15-\u003e13 into 15-\u003e12 (the lone regression removed, zero new fails). Sweep\n total 172-\u003e171, still monotone across all 6.\n2. Keep-better wrapper Fitness.collapse_finish(root, **kw) -\u003e (tree, base, coll,\n applied): scores on throwaway copies (score_with_fails merges in place),\n returns collapsed only if fails don't increase. Safety belt (config already\n monotone here, not proven so in general).\n3. Wiring: driver.collapse_best(result, programme_dir, ...) updates result.best\n (lineage +collapse, canonical re-score). evolve.py runs it after the sharing\n polish behind --collapse/--no-collapse (default ON). Standalone CLI\n homemaker-collapse file.dom (pyproject entry) writes \u003cstem\u003e.collapsed.dom;\n flags --adjacency/--public-access/--objective/--keep-better. Verified the\n written dom independently re-scores 15-\u003e12.\n4. Tests: tests/test_collapse_global.py (6) — demand-set relabel, level hard\n constraint, c/o/s exclusion, no-op safety, keep-better/unmerged. 267 pass.\n\nSTILL OPEN (separate work, not label slick): geometry-intrinsic fails — long-thin\nuseless cells (width/proportion/crinkliness) and not-connected — need geometry/\ntopology + circulation-placement search, NOT the collapse (see bd memory\ncollapse-global-94g...). In-search WFC collapse A/B also still open (xi7 risk).","status":"closed","priority":2,"issue_type":"feature","assignee":"Bruno Postle","owner":"bruno@postle.net","created_at":"2026-07-17T16:17:12Z","created_by":"Bruno Postle","updated_at":"2026-07-18T10:13:43Z","started_at":"2026-07-17T21:07:05Z","closed_at":"2026-07-18T10:13:43Z","close_reason":"FINISH-TIME global cell→room collapse DELIVERED (commits d52cce6, da18ef7, 880a214):\nFitness.collapse_global (c/o/s partition, level hard constraint, adjacency\nrelaxation, threshold objective, public-access pin) + collapse_finish keep-better\nwrapper + driver.collapse_best + evolve --collapse hook + homemaker-collapse CLI +\n6 tests (267 pass). Best harbor-house layout 15→12 fails; monotone across 6 evolved\nlayouts (sweep 195→171). Label search only — cannot fix geometry/building-intrinsic\nfails.\n\nRemaining scope spun out: homemaker-py-7fm (shape-intrinsic reshape), homemaker-py-qi6\n(circulation placement / not-connected), homemaker-py-qpk (in-search WFC collapse\nA/B, xi7 landscape-flattening risk). Closing 94g as the finish-time thrust is done.","dependencies":[{"issue_id":"homemaker-py-94g","depends_on_id":"homemaker-py-b3v","type":"related","created_at":"2026-07-17T17:23:20Z","created_by":"Bruno Postle","metadata":"{}"},{"issue_id":"homemaker-py-94g","depends_on_id":"homemaker-py-xi7","type":"related","created_at":"2026-07-17T17:23:06Z","created_by":"Bruno Postle","metadata":"{}"}],"dependency_count":0,"dependent_count":0,"comment_count":0} {"id":"homemaker-py-3l6","title":"Leaf-sharing default makes internal fitness diverge from canonical homemaker-fitness score","description":"Default is --leaf-sharing (on). Leaf sharing is a fitness-evaluation knob (fitness.py:415 quality_size, plus edge cap and count check): a shared leaf of code X is credited as satisfying k programme entries, with its size Gaussian re-centred on k*target. The evolve internal objective therefore rewards genomes that under-materialise the programme. When the winning .dom is re-scored by the canonical homemaker-fitness (leaf_sharing off), those un-materialised copies become 'missing required space ... (critical)' fails.\n\nObserved on examples/harbor-house (init.dom, budget 3M, workers 2):\n - leaf sharing ON (default): internal best 1.03e-05, but canonical score 6.73e-29 with 90 fails (15 critical missing-room).\n - --no-leaf-sharing (warm-started to full budget): internal and canonical agree at 4.19e-06, 15 fails, 0 critical -- ~9500x better than the prior best 3m.dom (4.41e-10).\n\nSo the default silently optimises an objective the canonical scorer does not credit, and writes a catastrophically worse .dom than its reported internal fitness implies.\n\nOptions to consider:\n 1. Make --no-leaf-sharing the default (strict per-leaf baseline agrees with canonical scorer).\n 2. Before writing output, re-score best-so-far with leaf_sharing off and warn (or refuse) if it regresses vs internal fitness.\n 3. Materialise/unfold shared leaves into k distinct rooms when writing the .dom, so the output satisfies the per-room programme.\n 4. Keep sharing as an early-phase relaxation only and anneal leaf_share_factor down to 0 before finishing (see related annealing investigation).","notes":"FIXED (option 3+2 combined, auto-finish): leaf-sharing runs now unfold+polish+rescore before write so output is honest under the canonical scorer.\n\nImplementation:\n- driver.polish_finish(result, programme_dir, polish_budget, ...): deep-copies best, operators.unfold_shared_leaves() to materialise the count deficit, then warm-starts a leaf_sharing=False search (bootstrap=False) from the unfolded genome. polish_budget\u003c=0 -\u003e single rescore eval only (used on interrupt). Stitches evals/topologies/sigs/restarts/history onto the sharing run; history tagged share:/polish: since the two objectives are not comparable. Returned best.fitness is canonical (leaf_sharing off =\u003e internal==canonical).\n- evolve.py: new --polish-budget flag (env HOMEMAKER_POLISH_BUDGET, default -1=auto=budget//2, 0=unfold+rescore only). main() calls polish_finish when --leaf-sharing on; interrupt forces polish_budget=0 for a fast honest output.\n\nVerified end-to-end (harbor-house, budget 3000 + polish 1500): reported polish fitness 4.79788e-27 EXACTLY matches canonical homemaker-fitness, 0 critical fails (was: internal 1.03e-05 vs canonical 6.73e-29 w/ 15 critical). Tiny budget so absolute quality low but honesty restored. Tests: driver.polish_finish x3 (test_driver.py), full suite 254 pass.\n\nDefault kept --leaf-sharing on per decision (sharing's topology-search speed retained; output made honest by the finish). Schedule B in-run annealing remains kpu.","status":"closed","priority":2,"issue_type":"bug","assignee":"Bruno Postle","owner":"bruno@postle.net","created_at":"2026-07-05T16:24:34Z","created_by":"Bruno Postle","updated_at":"2026-07-15T08:00:26Z","started_at":"2026-07-15T06:56:29Z","closed_at":"2026-07-15T08:00:26Z","close_reason":"Closed","dependency_count":0,"dependent_count":1,"comment_count":0} @@ -80,26 +80,26 @@ {"id":"homemaker-py-erc.6","title":"Experiment: inner-loop slack-expansion objective term","description":"Inner-loop counterpart to plot-fill construction. If Diagnostic B shows the inner loop has room to expand leaves into slack but no objective gradient to do so (the scalar rewards hitting target area but not exceeding it where slack exists), add a term/incentive so the ratio optimiser pushes leaf boundaries out to consume neighbouring slack and satisfy size, rather than parking at target.\n\nCONDITIONAL on Diagnostic B: build this only if B localizes the gap to the inner loop (room to expand, no gradient); if B shows construction targets too-small dims, prefer the plot-fill construction sibling. Must preserve the §5.4 inner-loop cliff / §4.9 lexicographic protection — the term sits where it cannot displace the fail-count ordering. A/B vs §12.2 baseline, seeds 0/1/2, 20000 evals, staged, default-OFF. Record DESIGN.md §13.6.","notes":"DEPRIORITISED by Diagnostic B (§13.2). B shows the inner loop CANNOT repair undersize: the slack is depth-driven maldistribution baked into the frozen topology, and the equal-offset ratio DOF cannot shrink a 14x leaf to feed a starved one without trading into shape fails (0.5^n cliff). Wrong DOF and wrong direction — the blocker is slicing POSITION, not a missing expansion reward. Fix belongs upstream in construction/topology (erc.4 re-scoped, erc.3). Keep as a low-priority follow-up only if a depth-balanced construction still leaves a residual size gradient the inner loop could pick up.","status":"closed","priority":4,"issue_type":"feature","owner":"bruno@postle.net","created_at":"2026-06-22T23:16:24Z","created_by":"Bruno Postle","updated_at":"2026-06-28T13:22:22Z","closed_at":"2026-06-28T13:22:22Z","close_reason":"wont-fix (DESIGN §13.7): Diag B (§13.2) showed the inner loop cannot repair undersize (wrong DOF — slicing position, frozen-topology ratios). Superseded by depth-balanced construction (erc.4). Condition unmet.","dependencies":[{"issue_id":"homemaker-py-erc.6","depends_on_id":"homemaker-py-erc","type":"parent-child","created_at":"2026-06-23T00:16:23Z","created_by":"Bruno Postle","metadata":"{}"},{"issue_id":"homemaker-py-erc.6","depends_on_id":"homemaker-py-erc.2","type":"blocks","created_at":"2026-06-23T00:16:47Z","created_by":"Bruno Postle","metadata":"{}"}],"dependency_count":1,"dependent_count":0,"comment_count":0} {"id":"homemaker-py-erc.5","title":"Experiment: compactness-aware cuts (minimize leaf perimeter/area)","description":"Attacks the #1 factor, crinkliness (346) — a per-leaf perimeter/area property DISTINCT from proportion (aspect ratio). Proportion-aware seeding (leu.2) sizes splits but does not bias toward balanced, square-ish subdivision. Add a KD-tree-style 'keep both children compact' cut rule (prefer the cut orientation/position that minimises summed child perimeter/area) in construction.\n\nCONDITIONAL on Diagnostic A: if A shows per-leaf shape-fail is FLAT across densities (floor intrinsic to slicing density), better cuts at the same leaf count will not pay → this should be closed wont-fix in favour of leaf-sharing. Only build if A shows shape-fail RISES with density. A/B vs §12.2 baseline, seeds 0/1/2, 20000 evals, staged, default-OFF. Record DESIGN.md §13.5.","notes":"DEPRIORITISED by erc.1 verdict (§13.1): per-leaf shape-fail flat vs slicing density and cuts already squarest (_size_divisions_from_targets picks squarest rotation) yet still ~1.8 fails/leaf =\u003e little compactness headroom at fixed leaf count. Floor is intrinsic to leaf COUNT, not cut quality. Revisit only if leaf-sharing (erc.3) underdelivers.","status":"closed","priority":4,"issue_type":"feature","owner":"bruno@postle.net","created_at":"2026-06-22T23:16:21Z","created_by":"Bruno Postle","updated_at":"2026-06-28T13:22:17Z","closed_at":"2026-06-28T13:22:17Z","close_reason":"wont-fix (DESIGN §13.7): Diag A (§13.1) showed the floor is intrinsic to leaf COUNT not cut quality; revisit condition was 'only if leaf-sharing underdelivers' but leaf-sharing OVER-delivered (−32…−39%, §13.3). Condition unmet.","dependencies":[{"issue_id":"homemaker-py-erc.5","depends_on_id":"homemaker-py-erc","type":"parent-child","created_at":"2026-06-23T00:16:21Z","created_by":"Bruno Postle","metadata":"{}"},{"issue_id":"homemaker-py-erc.5","depends_on_id":"homemaker-py-erc.1","type":"blocks","created_at":"2026-06-23T00:16:43Z","created_by":"Bruno Postle","metadata":"{}"}],"dependency_count":1,"dependent_count":0,"comment_count":0} {"id":"homemaker-py-2g5","title":"Rebuild occlusion/daylight/sun subsystem in Python (post-Phase-5, after optimisation fully native)","description":"DESIGN.md §6 port scope — a whole subsystem, not a term. quality_daylight (Leaf.pm:281-296) needs Urb::Misc::Sun + Urb::Field::Occlusion (+CIESky); quality_uncrinkliness also takes the occlusion object. Indoor spaces return 1 for daylight; cost is outdoor spaces + crinkliness. Port Sun_horizontal (262980-minute normalisation) and the occlusion wall set from Dom-\u003eWalls.","acceptance_criteria":"Daylight and crinkliness factors match Perl (float tolerance) across the corpus, including multi-storey cases","notes":"Re-scoped 2026-06-12: occlusion disabled in the Urb oracle instead of ported (see homemaker-py-gp2). Native fitness ships with simple crinkliness (illumination factor = 1, in homemaker-py-gnw). This issue is now the eventual Python occlusion rebuild, only after optimisation works entirely in Python. Restores outdoor-daylight and shaded-wall selection pressure.\nReframed 2026-06-17: orthogonal to epic homemaker-py-c4c. This is fitness FIDELITY (restoring daylight + shaded-wall selection pressure to match Perl), not search CAPABILITY — it changes what 'good' means, not the search's ability to find good. It will NOT improve final designs in the sense currently sought. Stays P4, deferred until the topology-search-quality epic lands and optimisation is fully native.","status":"open","priority":4,"issue_type":"feature","owner":"bruno@postle.net","created_at":"2026-06-11T23:38:25Z","created_by":"Bruno Postle","updated_at":"2026-06-17T19:14:48Z","dependency_count":0,"dependent_count":0,"comment_count":0} -{"_type":"memory","key":"homemaker-py-pythonpath-set-pythonpath-home-bruno-src","value":"homemaker-layout PYTHONPATH: package installed as 'homemaker-layout' via pip install -e . so 'import homemaker_layout' works from anywhere without PYTHONPATH. For running tests use 'python -m pytest' from project root /home/bruno/src/homemaker-layout (pyproject.toml adds src/ automatically). Never try pip show homemaker — that's the old homemaker-addon conflict."} -{"_type":"memory","key":"user-preference-bruno-this-is-a-fedora-system","value":"User preference (Bruno): this is a Fedora system — NEVER install Python packages via pip without asking first; always ask whether to install the rpm via dnf (e.g. python3-cma) before considering pip. Applies to any dependency additions."} -{"_type":"memory","key":"cli-tool-style-prefer-python-m-homemaker-module","value":"CLI tool style: prefer python -m homemaker.module --parameters pattern, installable via pip install -e . with pyproject.toml entry_points. Not standalone bin/ scripts."} -{"_type":"memory","key":"experiment-harness-gotcha-the-leaf-sharing-relaxed-objective","value":"Experiment harness gotcha: the leaf-sharing RELAXED objective (§13.3) is injected ONLY by monkeypatching fitness.load_config in the parent process (run_staged_search.py / probe scripts). This is parent-process-only and does NOT propagate into ProcessPoolExecutor workers (n_workers\u003e1), which re-import fitness fresh and score under the STRICT on-disk patterns.config -\u003e r.n_fails MISMATCH (worker strict vs parent relaxed re-score). ALL §13.x floor runs were therefore SERIAL. Any future PARALLEL leaf-sharing experiment will silently mis-score until leaf_sharing lives on disk/CLI (tracked: homemaker-py-x3b). The parallel driver itself is correct; both paths score via load_config(programme_dir)."} +{"_type":"memory","key":"deceptive-valleys-in-topology-search-when-every-single","value":"Deceptive valleys in topology search: when every single-step mutation from a target state passes through a high-fail intermediary (e.g. level_fix displaces a room into 5+ new fails), a compound operator that atomically applies two coordinated changes can escape. Design compound operators to land on the low-fail state directly, bypassing the deceptive gradient. Programme-house example: level_compound_fix atomically moves the level-constrained room AND re-inserts the displaced room adjacent to C in one step (operators.py, 2026-06-14)."} +{"_type":"memory","key":"9o5-multi-use-leaves-is-path-a-superposition","value":"9o5 multi-use leaves is path (a) — superposition as SEARCH RELAXATION that COLLAPSES to specific usage at the end, NOT path (b) loose-fit/no-collapse. Bruno's intent: codes with SIMILAR leaf requirements form an interchangeable equivalence class; during evolution the solver doesn't commit which leaf serves which specific usage (smoother landscape, no fighting over exact leaf usage); at the end the layout is CONDENSED to specific usages by brute-forcing the in-class assignment (3 interchangeable usages over 3 leaves = 3! = 6 combinations to check, pick best). 'Derive automatically' compatibility = requirement-similarity grouping. This reverses the issue's stated 'path b preferred' note."} {"_type":"memory","key":"experiment-seeding-pitfall-run-search-scaled-py-s","value":"Experiment seeding pitfall: run_search_scaled.py's default PH_SEED (c964…dom) is a FINISHED programme-house design — passing it warm-starts and floors at ~3 fails, NOT a blank-slate topology search. For blank-slate runs comparable to §11.5/§11.6 baselines, seed from examples/programme-house/init.dom (a bare undivided plot; driver bootstrap auto-triggers only on bare plots). Bit the 6zy sweep — first pass used c964 and falsely showed 3-fail floor across the whole grid."} -{"_type":"memory","key":"never-use-corpus-filenames-candidate-001-dom-candidate","value":"Never use corpus filenames (candidate-001.dom, candidate-002.dom, generated.dom, init.dom, etc.) as --output targets when running experiments. These are test fixtures. Always write experimental outputs to scratch/ or a timestamped path. Lesson from 2026-06-14: warm-start runs overwrote candidate-001/002.dom and broke graph tests."} -{"_type":"memory","key":"urb-oracle-nondeterminism-urb-fitness-pl-output-varies","value":"Urb oracle nondeterminism: urb-fitness.pl output varies run-to-run from Perl hash-order randomisation — .fails line ORDER shuffles (compare sorted, use oracle.Score.fail_lines) and the score float can flip by ~1 ULP (compare with math.isclose rel_tol=1e-12, never ==). Not a batching artifact; affects single runs too. Matters for the Phase 3 native-fitness parity gate (homemaker-py-uxz)."} -{"_type":"memory","key":"proportion-aware-constructive-seeding-leu-2-12-2","value":"Proportion-aware constructive seeding (leu.2/§12.2): sizing seed cuts from target AREAS only regresses (thin slivers wreck aspect); you must ALSO pick each cut's rotation for child squareness. It is a convergence ACCELERATOR via a deeper local optimum around the constructed topology: wins where that topology is roughly right and budget is scarce (harbor -13%, maple -10% at 20k evals) but DELAYS small programmes where the seed must be restructured by undivide (programme-house regresses at fixed budget, yet reaches the floor given budget - speed, not asymptote). Default-on. Also: n_storeys must honour storey_minimum, not just level: keys (programme-house storey_minimum:2, all rooms level:0 - was seeded 1 storey short; cq1)."} +{"_type":"memory","key":"homemaker-py-3l6-fix-leaf-sharing-evolve-runs","value":"homemaker-py-3l6 fix: leaf-sharing evolve runs now auto-finish before write via driver.polish_finish — unfold_shared_leaves() then a warm-started leaf_sharing=False polish search (--polish-budget, default budget//2). Makes the written .dom honest under canonical homemaker-fitness (internal==canonical when leaf_sharing off). Interrupt path forces polish_budget=0 (unfold+rescore only). This is yaa's unfold-then-polish, made automatic; Schedule B annealing is still kpu."} {"_type":"memory","key":"run-to-run-reproducibility-in-homemaker-layout-serial","value":"Run-to-run reproducibility in homemaker-layout: serial search (workers=1) is byte-for-byte deterministic; parallel (workers\u003e1) is now deterministic too AFTER fixing driver._run_batch to admit futures in submission order (was as_completed/completion order, bug xcy). Reproducibility holds only for a FIXED worker count — serial vs parallel differ because children-per-iteration is 1 vs n_workers (different batch granularity), which is expected, not a bug. The constructive seeder was NEVER nondeterministic: _assign_adjacency_aware has unique idx tiebreaks; comparing topologies with Python builtin hash() of the signature STRING is invalid (PYTHONHASHSEED salts str hashing per process) — use a stable hash (sha1) or genome.signature equality."} {"_type":"memory","key":"urb-fitness-bug-found-fixed-2026-06-12","value":"Urb fitness bug found+fixed 2026-06-12 (patch in /home/bruno/src/urb, uncommitted): ProgrammeDriven.pm ratio_o/ratio_type grepped case-insensitively over the ratios hash and took the FIRST key — nondeterministic (x4.5 score swings) for designs with mixed-case type classes (both 'c' circulation and 'C' covered). Fixed to SUM the class (matches Is_Circulation//Is_Outside semantics); 35/35 corpus scores unchanged. CRITICAL for homemaker-py-3y7/gnw: the native port must implement class-SUM ratios. Building.pm has the same unpatched pattern (site-driven path, not used by our oracle). Also: the memetic search reward-hacked this bug before the fix — search results predating it are noise artifacts."} -{"_type":"memory","key":"deceptive-valleys-in-topology-search-when-every-single","value":"Deceptive valleys in topology search: when every single-step mutation from a target state passes through a high-fail intermediary (e.g. level_fix displaces a room into 5+ new fails), a compound operator that atomically applies two coordinated changes can escape. Design compound operators to land on the low-fail state directly, bypassing the deceptive gradient. Programme-house example: level_compound_fix atomically moves the level-constrained room AND re-inserts the displaced room adjacent to C in one step (operators.py, 2026-06-14)."} -{"_type":"memory","key":"ld2-13-6-interior-o-seed-diagnostic-all","value":"ld2/§13.6 interior-O seed diagnostic: ALL crinkliness fails in the constructed bal+share seed are UNDER-exposed (crink\u003c0.62, landlocked rooms with no facade + no uncovered-O neighbour) — zero over-exposed sliver fails. So the erc crinkliness residual is genuine under-daylighting, validating the interior light-well premise. Default outside_divisor=6 was too sparse (null: harbor 147-\u003e142, crinkliness even rose). odiv=3 is the seed-optimal joint setting: harbor seed fails 147-\u003e129 (-18), maple 219-\u003e206 (-14), landlocked fails drop, at cost of more leaves (harbor +4, maple +8). Because it ADDS leaves it carries the §13.4 wash-out risk; A/B to convergence pending."} -{"_type":"memory","key":"strategy-decision-2026-06-12-bruno-occlusion-daylight","value":"Strategy decision 2026-06-12 (Bruno): occlusion/daylight is ORTHOGONAL to building a scalable optimiser. Disable it in Urb (env flag, homemaker-py-gp2) rather than port it; native fitness uses simple crinkliness (illumination factor = 1); rebuild occlusion in Python only after optimisation is fully native (homemaker-py-2g5, now P4). Consequence: all scores change when the flag flips — re-baseline corpus/.score, DESIGN \\$4.5 gains, gate bars at one clean boundary AFTER homemaker-py-1p0 closes; Phase-2 urb-evolve benchmark must run with the same flag."} -{"_type":"memory","key":"9o5-multi-use-leaves-is-path-a-superposition","value":"9o5 multi-use leaves is path (a) — superposition as SEARCH RELAXATION that COLLAPSES to specific usage at the end, NOT path (b) loose-fit/no-collapse. Bruno's intent: codes with SIMILAR leaf requirements form an interchangeable equivalence class; during evolution the solver doesn't commit which leaf serves which specific usage (smoother landscape, no fighting over exact leaf usage); at the end the layout is CONDENSED to specific usages by brute-forcing the in-class assignment (3 interchangeable usages over 3 leaves = 3! = 6 combinations to check, pick best). 'Derive automatically' compatibility = requirement-similarity grouping. This reverses the issue's stated 'path b preferred' note."} {"_type":"memory","key":"island-model-psk-14-is-a-null-priming","value":"Island model (psk, §14) is a NULL: priming a population from N converged independent elites + crossover-heavy migration does not beat best-of-N at equal total budget (maple island 124 vs control 116). The child_probe instrument shows WHY: area-matched crossover across independently-converged elites almost never synthesizes (1-3 of ~64 children beat the better parent, max drop 2-5) because the slicing encoding is non-canonical (9gp), so splices are disruptive not combinatorial. Search-machinery null #3 after graded-objective and niching/restarts; residual stays geometry/shape-bound."} -{"_type":"memory","key":"programme-house-optimisation-result-2026-06-14-15","value":"Programme-house optimisation result (2026-06-14/15): best achievable is 1 fail (l1 wrong level, score ~0.005). 0 fails is geometrically impossible: l1 (min 27m²) must occupy ll (~23m²) at level 0, which eliminates the t3-adj-C provider; dividing ll into lll(l1)+llr(C) gives llr proportion ~6:1 (fails). Python memetic optimizer achieves 1 fail in 50k evals vs Perl optimiser's 2-3 fails. Winning topology: TWO C nodes at level 0 — ll(C) for t3-adj-C via geometric contact, rl(C) for staircase via tree-sibling adjacency to rrr(O). Best .dom: scratch/from-warmstart-fixed.dom and scratch/from-compound3-fixed.dom."} {"_type":"memory","key":"warm-x0-initialization-bug-pattern-when-a-topology","value":"warm_x0 initialization bug pattern: when a topology operator explicitly sets division ratios on a newly-created node (e.g. compound_fix sets node.division=[0.25,0.25] for t3), parent.ratios has no entry for that node (it was a leaf). warm_x0 defaults it to 0.5, corrupting the inner loop's starting point and making the operator invisible to lex comparison. Fix: only propagate child ratios for nodes where the parent node was NOT already divided; stale hidden nodes revealed by structural mutations (swap flipping b.below) must NOT contribute their pre-writeback values. See driver.py lines 259-267 (fixed 2026-06-14)."} +{"_type":"memory","key":"homemaker-py-pythonpath-set-pythonpath-home-bruno-src","value":"homemaker-layout PYTHONPATH: package installed as 'homemaker-layout' via pip install -e . so 'import homemaker_layout' works from anywhere without PYTHONPATH. For running tests use 'python -m pytest' from project root /home/bruno/src/homemaker-layout (pyproject.toml adds src/ automatically). Never try pip show homemaker — that's the old homemaker-addon conflict."} +{"_type":"memory","key":"ld2-13-6-interior-o-seed-diagnostic-all","value":"ld2/§13.6 interior-O seed diagnostic: ALL crinkliness fails in the constructed bal+share seed are UNDER-exposed (crink\u003c0.62, landlocked rooms with no facade + no uncovered-O neighbour) — zero over-exposed sliver fails. So the erc crinkliness residual is genuine under-daylighting, validating the interior light-well premise. Default outside_divisor=6 was too sparse (null: harbor 147-\u003e142, crinkliness even rose). odiv=3 is the seed-optimal joint setting: harbor seed fails 147-\u003e129 (-18), maple 219-\u003e206 (-14), landlocked fails drop, at cost of more leaves (harbor +4, maple +8). Because it ADDS leaves it carries the §13.4 wash-out risk; A/B to convergence pending."} +{"_type":"memory","key":"programme-house-optimisation-result-2026-06-14-15","value":"Programme-house optimisation result (2026-06-14/15): best achievable is 1 fail (l1 wrong level, score ~0.005). 0 fails is geometrically impossible: l1 (min 27m²) must occupy ll (~23m²) at level 0, which eliminates the t3-adj-C provider; dividing ll into lll(l1)+llr(C) gives llr proportion ~6:1 (fails). Python memetic optimizer achieves 1 fail in 50k evals vs Perl optimiser's 2-3 fails. Winning topology: TWO C nodes at level 0 — ll(C) for t3-adj-C via geometric contact, rl(C) for staircase via tree-sibling adjacency to rrr(O). Best .dom: scratch/from-warmstart-fixed.dom and scratch/from-compound3-fixed.dom."} +{"_type":"memory","key":"user-preference-bruno-this-is-a-fedora-system","value":"User preference (Bruno): this is a Fedora system — NEVER install Python packages via pip without asking first; always ask whether to install the rpm via dnf (e.g. python3-cma) before considering pip. Applies to any dependency additions."} {"_type":"memory","key":"adjacency-in-binary-slicing-tree-is-structural-not","value":"Adjacency in binary slicing tree is structural, not geometric: the inner-loop NM cannot fix topological adjacency failures. Two paths exist: (1) tree-sibling adjacency — a node is adjacent to its sibling in the tree; (2) cross-zone geometric adjacency — leaves from different subtrees that happen to share a boundary. Staircase/adjacency fails require a topology mutation that changes which nodes are siblings or which zones touch. This was proved empirically on programme-house: staircase fail from rot=0 layout could not be fixed by NM but was fixed by level_retype creating a two-C topology (2026-06-14/15)."} -{"_type":"memory","key":"correction-to-urb-fitness-bug-memory-bruno-2026","value":"CORRECTION to urb-fitness-bug memory (Bruno, 2026-06-12): 'C' is NOT a 'covered' type — Is_Covered is a geometric predicate (indoor space above). Urb's generic types are canonically UPPERCASE: C=circulation, O=outside, S=sahn (get_space_types qw/C O S/; corpus is 100% uppercase, never 'c'/'o' leaves). The mixed-case designs that fired the latent ratio_type first-match bug were created by homemaker's own operator type pool emitting lowercase 'c'/'o' — fixed: driver/operators now emit uppercase generics only, and class checks use t[0].lower() in 'cos'. The Urb class-sum patch stays as defensive hardening (zero impact on canonical designs). Native port (3y7/gnw): treat type classes case-insensitively, generics canonically uppercase."} -{"_type":"memory","key":"homemaker-py-3l6-fix-leaf-sharing-evolve-runs","value":"homemaker-py-3l6 fix: leaf-sharing evolve runs now auto-finish before write via driver.polish_finish — unfold_shared_leaves() then a warm-started leaf_sharing=False polish search (--polish-budget, default budget//2). Makes the written .dom honest under canonical homemaker-fitness (internal==canonical when leaf_sharing off). Interrupt path forces polish_budget=0 (unfold+rescore only). This is yaa's unfold-then-polish, made automatic; Schedule B annealing is still kpu."} +{"_type":"memory","key":"strategy-decision-2026-06-12-bruno-occlusion-daylight","value":"Strategy decision 2026-06-12 (Bruno): occlusion/daylight is ORTHOGONAL to building a scalable optimiser. Disable it in Urb (env flag, homemaker-py-gp2) rather than port it; native fitness uses simple crinkliness (illumination factor = 1); rebuild occlusion in Python only after optimisation is fully native (homemaker-py-2g5, now P4). Consequence: all scores change when the flag flips — re-baseline corpus/.score, DESIGN \\$4.5 gains, gate bars at one clean boundary AFTER homemaker-py-1p0 closes; Phase-2 urb-evolve benchmark must run with the same flag."} +{"_type":"memory","key":"urb-oracle-nondeterminism-urb-fitness-pl-output-varies","value":"Urb oracle nondeterminism: urb-fitness.pl output varies run-to-run from Perl hash-order randomisation — .fails line ORDER shuffles (compare sorted, use oracle.Score.fail_lines) and the score float can flip by ~1 ULP (compare with math.isclose rel_tol=1e-12, never ==). Not a batching artifact; affects single runs too. Matters for the Phase 3 native-fitness parity gate (homemaker-py-uxz)."} +{"_type":"memory","key":"cli-tool-style-prefer-python-m-homemaker-module","value":"CLI tool style: prefer python -m homemaker.module --parameters pattern, installable via pip install -e . with pyproject.toml entry_points. Not standalone bin/ scripts."} {"_type":"memory","key":"multi-storey-staircase-consistency-when-dividing-or-retyping","value":"Multi-storey staircase consistency: when dividing or retyping a circulation (C) leaf at one level, the same structural change should be propagated to the matching leaf on ALL other storeys so the stair core path is maintained. The optimizer cannot fix staircase disruptions through trial-and-error geometry alone — it requires a synchronized multi-level operator that applies the same topology change to every storey simultaneously."} +{"_type":"memory","key":"proportion-aware-constructive-seeding-leu-2-12-2","value":"Proportion-aware constructive seeding (leu.2/§12.2): sizing seed cuts from target AREAS only regresses (thin slivers wreck aspect); you must ALSO pick each cut's rotation for child squareness. It is a convergence ACCELERATOR via a deeper local optimum around the constructed topology: wins where that topology is roughly right and budget is scarce (harbor -13%, maple -10% at 20k evals) but DELAYS small programmes where the seed must be restructured by undivide (programme-house regresses at fixed budget, yet reaches the floor given budget - speed, not asymptote). Default-on. Also: n_storeys must honour storey_minimum, not just level: keys (programme-house storey_minimum:2, all rooms level:0 - was seeded 1 storey short; cq1)."} {"_type":"memory","key":"unfold-strategy-for-shared-leaves-homemaker-py-8iv","value":"Unfold strategy for shared leaves (homemaker-py-8iv, resolved 2026-07-16): use the BALANCED GRID (operators._grow_balanced/_size_subtree_equal), NOT circulation-aware slicing. Slicing a shared leaf perpendicular to its access edge so every child touches the corridor was implemented + A/B-tested and LOST decisively (150k-eval warm-start polish from evolved-3M: slice 41 fails/3.5e-14 vs grid 25 fails/2.4e-09, grid ahead at every milestone). Reason: k rooms all touching one wall are intrinsically thin slices; that geometric debt (proportion/long/width) is unfixable without topology change, while the grid's squarer children let local search re-route access cheaply via level_retype/place_missing/level_fix. Lesson: at the sharing-\u003eno-sharing transition, prioritise squarer children and leave access to local search; do not reintroduce slicing in Schedule B (kpu)."} +{"_type":"memory","key":"experiment-harness-gotcha-the-leaf-sharing-relaxed-objective","value":"Experiment harness gotcha: the leaf-sharing RELAXED objective (§13.3) is injected ONLY by monkeypatching fitness.load_config in the parent process (run_staged_search.py / probe scripts). This is parent-process-only and does NOT propagate into ProcessPoolExecutor workers (n_workers\u003e1), which re-import fitness fresh and score under the STRICT on-disk patterns.config -\u003e r.n_fails MISMATCH (worker strict vs parent relaxed re-score). ALL §13.x floor runs were therefore SERIAL. Any future PARALLEL leaf-sharing experiment will silently mis-score until leaf_sharing lives on disk/CLI (tracked: homemaker-py-x3b). The parallel driver itself is correct; both paths score via load_config(programme_dir)."} +{"_type":"memory","key":"never-use-corpus-filenames-candidate-001-dom-candidate","value":"Never use corpus filenames (candidate-001.dom, candidate-002.dom, generated.dom, init.dom, etc.) as --output targets when running experiments. These are test fixtures. Always write experimental outputs to scratch/ or a timestamped path. Lesson from 2026-06-14: warm-start runs overwrote candidate-001/002.dom and broke graph tests."} {"_type":"memory","key":"collapse-global-94g-and-any-label-usage-optimisation","value":"collapse_global (94g) and any label/usage optimisation CANNOT fix geometry-intrinsic fails. The harbor-house 15-fail best layout contains long-thin cells that are useless whatever room usage is assigned — their width/proportion/crinkliness fails are shape-bound, not label slack. Two consequences: (1) do not over-claim collapse gains — only ~2-3 of that layout's fails are reclaimable relabel slack, the rest are geometry- or building-level bound; (2) the threshold objective must not be tuned to 'pass' a degenerate cell via a permissive room type — a metric-pass on a physically useless space is gaming, not a fix. Real remedies for these are geometry/topology search (cell shape) and circulation placement, filed separately, not the collapse."} +{"_type":"memory","key":"correction-to-urb-fitness-bug-memory-bruno-2026","value":"CORRECTION to urb-fitness-bug memory (Bruno, 2026-06-12): 'C' is NOT a 'covered' type — Is_Covered is a geometric predicate (indoor space above). Urb's generic types are canonically UPPERCASE: C=circulation, O=outside, S=sahn (get_space_types qw/C O S/; corpus is 100% uppercase, never 'c'/'o' leaves). The mixed-case designs that fired the latent ratio_type first-match bug were created by homemaker's own operator type pool emitting lowercase 'c'/'o' — fixed: driver/operators now emit uppercase generics only, and class checks use t[0].lower() in 'cos'. The Urb class-sum patch stays as defensive hardening (zero impact on canonical designs). Native port (3y7/gnw): treat type classes case-insensitively, generics canonically uppercase."} diff --git a/DESIGN.md b/DESIGN.md index f342c20..c93aad4 100644 --- a/DESIGN.md +++ b/DESIGN.md @@ -2414,3 +2414,54 @@ the collapse *inside* search per-eval (rather than finish-time) is `homemaker-py gated on the 9o5 landscape-flattening risk (§13 / `homemaker-py-xi7`) and its own A/B. Tests: `tests/test_collapse_global.py` ×6 (demand-set relabel, level hard constraint, c/o/s exclusion, no-op safety, keep-better/unmerged); 267 pass. + +## 18. Graded circulation-connectivity signal (`homemaker-py-qi6`) — in progress + +**Motivation — the binary fail is flat.** After the §17 collapse, the residual fails on the +harbor-house set are dominated by `level N not connected` (2 of the best layout's 12; also on +5 of the 6 sweep layouts). That fail comes from `connected_circulation` (`graph.py`): remove +every non-circulation vertex from a storey's adjacency graph and require the remaining +circulation cells (`C` stairs plus the `cr`/`st` room-codes that collide with the c/s prefix) +to form ONE connected component. On the evolved layouts they instead fragment into **4–7 +components per storey**. + +**Why finish-time repair fails (measured, negative).** The obvious §17-style companion — a +finish-time pass that re-types boundary cells to circulation to bridge the components, kept +only if the fail count does not rise — was prototyped (Steiner-MST bridge set per disconnected +storey, keep-better guard) and measured on the 6 layouts: **195 → 560 fails (+365)**. The +`not connected` fail is *binary* (one fail per storey regardless of fragmentation), but each +storey needs 3–7 bridge cells, and every needed-room→circulation conversion triggers a +missing-room fail cascade (2–5 fails) that dwarfs the single connectivity fail it clears. +Keep-better reverts every one → no-op. **Conclusion: connectivity cannot be bought at finish +time when every cell is a needed room; it must come from the outer search allocating connected +circulation topology.** But the binary fail gives the search *zero gradient* — a 7-component +storey scores identically (both in fail count and in the `0.5^n` scalar) to a 2-component one — +so the search cannot tell it is making progress. + +**Mechanism — a graded proximity on the same channel §11.4 built.** `graph.circulation_connectivity(G)` +returns the fraction of circulation cells in the largest connected circulation component ∈ +[0,1] (1.0 = a single connected spine, lower = more fragmented, 0.0 = no circulation), measured +on the same circ subgraph the fail uses so the two agree at the connected endpoint. Summed over +storeys it is the graded proximity scalar `Fitness.score_with_grade` already carries for the +outer comparator, gated by the `conn_grade` conf flag: when on it *replaces* the §11.4 leaf +quality-proximity on that channel (a distinct, better-motivated use — §11.4 was rejected because +within a fail-tier the `0.5^n` scalar is NOT flat there and grade merely displaced a working +signal; connectivity is the opposite case, genuinely flat under the binary fail). Like §11.4 it +leaves the scalar fitness and fail count **byte-identical** (verified) — it is only the secondary +key `(-n_fails, grade, fitness)` (driver `use_lex and use_grade`), strictly beneath fail-count so +the §6 missing-space hierarchy and the §5.4 inner-loop cliff are untouched. Among equally-failing +neighbours the search now prefers the one whose circulation is closer to one component, restoring +the gradient toward connected topologies. + +**Wiring.** `conn_grade` threads through `_overrides_for`/`_fitness_for`/`_evaluate` and the +`search` signature; enabling it implies the grade key. `evolve.py` exposes `--conn-grade` +(env `HOMEMAKER_CONN_GRADE`, default OFF); the grade is read off the optimised tree, one extra +native eval per child. + +**Status / next.** Signal, fitness wiring, CLI, and 9 tests landed (`tests/test_conn_grade.py`: +pure-graph fraction contract, non-circ cells ignored, monotone under (dis)connection, and the +score/fail-count-invariance of the flag). The A/B — does the gradient actually pull evolve runs +toward connected circulation and clear `not connected` fails — needs full-budget runs and is +pending (short 60-eval smoke run confirms the plumbing only). If the graded key alone is +insufficient, the follow-on is an insert/relocate-circulation mutation operator (mechanism (a), +still `homemaker-py-qi6`) that now has a gradient to climb. 276 tests pass. diff --git a/src/homemaker_layout/driver.py b/src/homemaker_layout/driver.py index 2d16137..d81866f 100644 --- a/src/homemaker_layout/driver.py +++ b/src/homemaker_layout/driver.py @@ -40,11 +40,14 @@ _CHILD_INNER_KW: dict = {} def _overrides_for(leaf_sharing: bool, superpose: bool, - max_share: int | None = None) -> dict | None: + max_share: int | None = None, + conn_grade: bool = False) -> dict | None: """Run-level conf overrides for the native evaluator (None when all off). ``max_share`` (homemaker-py-kpu) overrides the evaluator's ``leaf_share_max`` grain cap for the in-run annealing ramp; ``None`` leaves the config default. + ``conn_grade`` (homemaker-py-qi6) turns the graded proximity scalar into the + circulation-connectivity signal (§18). """ ov: dict = {} if leaf_sharing: @@ -53,13 +56,16 @@ def _overrides_for(leaf_sharing: bool, superpose: bool, ov["superpose"] = True if max_share is not None: ov["leaf_share_max"] = int(max_share) + if conn_grade: + ov["conn_grade"] = True return ov or None @functools.lru_cache(maxsize=None) def _fitness_for(programme_dir: str, leaf_sharing: bool = False, superpose: bool = False, - max_share: int | None = None) -> "fitness.Fitness": + max_share: int | None = None, + conn_grade: bool = False) -> "fitness.Fitness": """Cached Fitness evaluator per (programme dir, leaf_sharing) (config load is the cost). @@ -70,7 +76,7 @@ def _fitness_for(programme_dir: str, leaf_sharing: bool = False, inner loop instead of reading the on-disk (sharing-free) patterns.config. Cached per process — workers fork their own copy. """ - overrides = _overrides_for(leaf_sharing, superpose, max_share) + overrides = _overrides_for(leaf_sharing, superpose, max_share, conn_grade) conf, cost = fitness.load_config(programme_dir, overrides=overrides) return fitness.Fitness(conf, cost) @@ -146,7 +152,8 @@ def _evaluate(root: dom.Node, programme_dir, urb_root, x0, budget, inner_kw, best_n_fails: int | None = None, leaf_sharing: bool = False, superpose: bool = False, - max_share: int | None = None) -> tuple[Individual, int]: + max_share: int | None = None, + conn_grade: bool = False) -> tuple[Individual, int]: # §12.3 shape-feasibility pre-filter (homemaker-py-9gp.1): if even the best # achievable (proportion-aware) geometry of this topology already has at least # as many shape fails as the incumbent's TOTAL fails — and exceeds the tunable @@ -154,11 +161,12 @@ def _evaluate(root: dom.Node, programme_dir, urb_root, x0, budget, inner_kw, # eval instead of spending the full inner-loop budget. The best_n_fails guard # makes the proxy safe: a topology whose shape-fail floor is still below the # incumbent is never discarded. Pruned individuals are tagged and never admitted. - overrides = _overrides_for(leaf_sharing, superpose, max_share) + overrides = _overrides_for(leaf_sharing, superpose, max_share, conn_grade) if (feasibility_max_shape_fails is not None and best_n_fails is not None): pred = operators.predicted_shape_fails( root, _reqs_for(str(programme_dir)), - _fitness_for(str(programme_dir), leaf_sharing, superpose, max_share)) + _fitness_for(str(programme_dir), leaf_sharing, superpose, max_share, + conn_grade)) if pred > feasibility_max_shape_fails and pred >= best_n_fails: ind = Individual(root=root, fitness=0.0, n_fails=pred, ratios={}, lineage=f"pruned/{lineage}", grade=0.0, @@ -173,7 +181,8 @@ def _evaluate(root: dom.Node, programme_dir, urb_root, x0, budget, inner_kw, grade = 0.0 if want_grade: _, _, grade = _fitness_for( - str(programme_dir), leaf_sharing, superpose, max_share).score_with_grade( + str(programme_dir), leaf_sharing, superpose, max_share, + conn_grade).score_with_grade( copy.deepcopy(root)) ind = Individual(root=root, fitness=r.fitness, n_fails=r.n_fails, ratios=innerloop.ratio_map(root), lineage=lineage, @@ -209,6 +218,7 @@ def search( base_p: float = 1.0, child_probe=None, use_grade: bool = False, + conn_grade: bool = False, tournament_k: int = 2, niche_by_signature: bool = False, restart_patience: int | None = None, @@ -300,6 +310,9 @@ def search( # Kept default-off for reproducibility. Strictly beneath -n_fails ⇒ the # missing-space hierarchy (§6) is preserved and the inner-loop cliff (§5.4) # is untouched. + # homemaker-py-qi6 §18: the connectivity signal rides the same grade channel, + # so enabling it enables the grade secondary key. + use_grade = use_grade or conn_grade if use_lex and use_grade: _key = lambda ind: (-ind.n_fails, ind.grade, _rank_fitness(ind)) elif use_lex: @@ -402,7 +415,7 @@ def search( best_nf = result.best.n_fails if result.best is not None else None full = [ (root, programme_dir, urb_root, x0, budget_, kw_, lin, use_grade, - mx, best_nf, leaf_sharing, superpose, max_share) + mx, best_nf, leaf_sharing, superpose, max_share, conn_grade) for root, x0, budget_, kw_, lin in tasks ] if _pool is not None: @@ -483,7 +496,8 @@ def search( want_grade=use_grade, leaf_sharing=leaf_sharing, superpose=superpose, - max_share=max_share) + max_share=max_share, + conn_grade=conn_grade) n_evals += used admit(seed_ind, pop) diff --git a/src/homemaker_layout/evolve.py b/src/homemaker_layout/evolve.py index e6111c6..2242c04 100644 --- a/src/homemaker_layout/evolve.py +++ b/src/homemaker_layout/evolve.py @@ -95,6 +95,16 @@ def _parse_args(argv=None) -> argparse.Namespace: "requirements) form equivalence classes and each candidate " "collapses every superposed leaf to its best in-class usage " "before scoring (default: off)") + p.add_argument("--conn-grade", dest="conn_grade", + action=argparse.BooleanOptionalAction, + default=_env_bool("HOMEMAKER_CONN_GRADE", False), + help="homemaker-py-qi6 (§18): graded circulation-connectivity " + "signal. Adds a secondary comparator key (beneath fail " + "count, above fitness) = per-level largest-circ-component " + "fraction, giving the search a gradient toward connected " + "circulation that the binary 'not connected' fail lacks. " + "Does not change the scalar fitness or fail count " + "(default: off)") p.add_argument("--anneal-grain", type=str, default=os.environ.get("HOMEMAKER_ANNEAL_GRAIN"), metavar="LADDER", @@ -161,6 +171,7 @@ def main(argv=None) -> int: print(f"leaf sharing : {args.leaf_sharing} (factor={args.leaf_share_factor})", file=sys.stderr) print(f"superpose : {args.superpose}", file=sys.stderr) + print(f"conn grade : {args.conn_grade}", file=sys.stderr) print(f"output : {out or 'stdout'}", file=sys.stderr, flush=True) anneal_ladder = None @@ -209,6 +220,7 @@ def main(argv=None) -> int: leaf_sharing=args.leaf_sharing, leaf_share_factor=args.leaf_share_factor, superpose=args.superpose, + conn_grade=args.conn_grade, log=lambda m: print(m, file=sys.stderr, flush=True), ) _finish_sharing = args.leaf_sharing diff --git a/src/homemaker_layout/fitness.py b/src/homemaker_layout/fitness.py index d97788a..ea68428 100644 --- a/src/homemaker_layout/fitness.py +++ b/src/homemaker_layout/fitness.py @@ -214,6 +214,13 @@ class Fitness: # leaf to its best in-class usage before scoring, so search optimises the # condensed objective directly and the relaxation gap is removed. self._superpose = bool(self.conf("superpose")) + # homemaker-py-qi6 graded circulation-connectivity signal (DESIGN.md §18): + # default OFF. When on, the graded proximity scalar (want_grade) is the + # per-level largest-circ-component fraction instead of the §11.4 leaf + # quality-proximity — a secondary comparator key giving the outer search + # a gradient the binary "level N not connected" fail lacks. Leaves the + # scalar fitness and fail count untouched, exactly like §11.4. + self._conn_grade = bool(self.conf("conn_grade")) from .programme import CLASS_CAP as _CLASS_CAP self._class_cap = int(self.conf("superpose_class_cap") or _CLASS_CAP) self._interchange_classes: list | None = None # lazily derived @@ -1453,10 +1460,18 @@ class Fitness: ) cost += se.cost value += se.value - if want_grade: # §11.4 outer-comparator signal only; off by default + if want_grade and not self._conn_grade: # §11.4 signal; off by default for le in se.leaves: grade += _leaf_grade(le.factors) + # §18 (homemaker-py-qi6): repurpose the grade channel for the graded + # circulation-connectivity signal — sum of per-level largest-circ-component + # fractions, higher when circulation is closer to a single connected spine. + # Secondary comparator key only; score and fail count are untouched. + if want_grade and self._conn_grade: + for gc in graph_circ: + grade += graph_mod.circulation_connectivity(gc) + building_factor = self.evaluate_building(root, tracking) value *= building_factor diff --git a/src/homemaker_layout/graph.py b/src/homemaker_layout/graph.py index 158f926..9226cf6 100644 --- a/src/homemaker_layout/graph.py +++ b/src/homemaker_layout/graph.py @@ -188,6 +188,32 @@ def _connected_outside_inplace(G: nx.Graph) -> None: G.remove_nodes_from(to_remove) +def circulation_connectivity(G: nx.Graph) -> float: + """Fraction of circulation cells in the largest connected circulation + component — a continuous [0,1] proximity to a single connected circulation + spine (1.0 = fully connected, lower = more fragmented, 0.0 = no circulation). + + Companion graded signal for the binary ``level N not connected`` fail + (``connected_circulation``, homemaker-py-qi6). That fail fires identically + whether a level's circulation is split into 2 components or 7, so it is FLAT + across fragmentation and gives the outer search no gradient to climb toward + connectivity. This proxy restores the gradient: among equally-failing + layouts, the one whose circulation is closer to a single component scores + higher. Measured on the same circ subgraph the fail uses (all non-circulation + vertices removed), so the two agree at the connected endpoint (proxy == 1.0 + iff ``connected_circulation`` is True on a non-empty circ set). + """ + gc = G.copy() + gc.remove_nodes_from( + [v for v in list(gc.nodes()) if not dom.is_circulation(v)] + ) + n = gc.number_of_nodes() + if n == 0: + return 0.0 + largest = max((len(c) for c in nx.connected_components(gc)), default=0) + return largest / n + + def connected_circulation(G: nx.Graph) -> bool: """True iff circulation nodes are non-empty and connected; mirrors ``Urb::Dom::Connected_Circulation`` (Storey.pm:106). diff --git a/tests/test_conn_grade.py b/tests/test_conn_grade.py new file mode 100644 index 0000000..fe2fede --- /dev/null +++ b/tests/test_conn_grade.py @@ -0,0 +1,113 @@ +"""Tests for the graded circulation-connectivity signal (homemaker-py-qi6, §18). + +Covers: + - graph.circulation_connectivity: largest-circ-component fraction, non-circ + cells ignored, empty → 0.0, monotone under (dis)connection. + - Fitness conn_grade wiring: repurposes the graded proximity scalar, leaves the + scalar fitness and fail count byte-identical (secondary comparator key only). +""" + +import copy +from pathlib import Path + +import networkx as nx +import pytest + +from homemaker_layout import dom as dom_mod +from homemaker_layout.dom import Node +from homemaker_layout.graph import circulation_connectivity +from homemaker_layout.fitness import Fitness, load_config + +HARBOR = Path(__file__).parent.parent / "examples" / "harbor-house" + + +# --------------------------------------------------------------------------- # +# circulation_connectivity — pure graph contract +# --------------------------------------------------------------------------- # + +def _circ(*ids): + # bare, unlinked nodes: level_of == 0 → is_usable → is_circulation for c/s + return [Node(type=t) for t in ids] + + +def test_fully_connected_is_one(): + a, b, c = _circ("C", "C", "S") + G = nx.Graph([(a, b), (b, c)]) + assert circulation_connectivity(G) == 1.0 + + +def test_two_components_is_half(): + a, b, c, d = _circ("C", "C", "C", "C") + G = nx.Graph([(a, b), (c, d)]) # two disjoint pairs of 4 circ cells + assert circulation_connectivity(G) == 0.5 + + +def test_non_circulation_cells_ignored(): + # largest circ component is {a,b} of 3 circ cells → 2/3; the room cells r/s + # bridging them do NOT count as circulation, so the split stands. + a, b, lone = _circ("C", "C", "C") + r1, r2 = Node(type="b1"), Node(type="k1") + G = nx.Graph([(a, b), (a, r1), (r1, r2), (r2, lone)]) + assert circulation_connectivity(G) == pytest.approx(2 / 3) + + +def test_no_circulation_is_zero(): + r1, r2 = Node(type="b1"), Node(type="k1") + G = nx.Graph([(r1, r2)]) + assert circulation_connectivity(G) == 0.0 + assert circulation_connectivity(nx.Graph()) == 0.0 + + +def test_connecting_a_component_raises_the_grade(): + a, b, c, d = _circ("C", "C", "C", "C") + split = nx.Graph([(a, b), (c, d)]) # 0.5 + joined = nx.Graph([(a, b), (b, c), (c, d)]) # 1.0 + assert circulation_connectivity(joined) > circulation_connectivity(split) + + +# --------------------------------------------------------------------------- # +# Fitness conn_grade wiring — must not perturb score or fail count +# --------------------------------------------------------------------------- # + +@pytest.mark.skipif(not HARBOR.is_dir(), reason="harbor-house example absent") +@pytest.mark.parametrize("name", ["evolved-3M-nols-3.dom", "evolved-3M.dom"]) +def test_conn_grade_leaves_score_and_fails_untouched(name): + conf, cost = load_config(HARBOR) + conf_cg, _ = load_config(HARBOR, overrides={"conn_grade": True}) + fit, fit_cg = Fitness(conf, cost), Fitness(conf_cg, cost) + + root = dom_mod.load(str(HARBOR / name)) + s_base, f_base = fit.score_with_fails(copy.deepcopy(root)) + s_cg, f_cg, grade = fit_cg.score_with_grade(copy.deepcopy(root)) + + assert s_cg == pytest.approx(s_base) + assert f_cg == f_base + # grade is the sum of per-level fractions ∈ [0, n_levels]; harbor layouts are + # partially disconnected, so it is strictly positive and below the level count. + n_levels = len(dom_mod.levels(dom_mod.load(str(HARBOR / name)))) + assert 0.0 < grade <= n_levels + + +@pytest.mark.skipif(not HARBOR.is_dir(), reason="harbor-house example absent") +def test_more_connected_layout_scores_higher_grade(): + conf_cg, cost = load_config(HARBOR, overrides={"conn_grade": True}) + fit = Fitness(conf_cg, cost) + + def grade_of(name): + _, _, g = fit.score_with_grade(dom_mod.load(str(HARBOR / name))) + return g + + # evolved-3M has one fully-connected storey; nols-3 is fragmented on both. + assert grade_of("evolved-3M.dom") > grade_of("evolved-3M-nols-3.dom") + + +@pytest.mark.skipif(not HARBOR.is_dir(), reason="harbor-house example absent") +def test_conn_grade_off_uses_leaf_grade_not_connectivity(): + # With the flag off, the grade is the §11.4 leaf quality-proximity, which is a + # different (smaller, here) scalar — the two channels must not collide. + conf, cost = load_config(HARBOR) + conf_cg, _ = load_config(HARBOR, overrides={"conn_grade": True}) + root = dom_mod.load(str(HARBOR / "evolved-3M-nols-3.dom")) + _, _, g_leaf = Fitness(conf, cost).score_with_grade(copy.deepcopy(root)) + _, _, g_conn = Fitness(conf_cg, cost).score_with_grade(copy.deepcopy(root)) + assert g_leaf != g_conn