diff --git a/docs/ci-verification.md b/docs/ci-verification.md index 63ba3e9aa8..52b62d0481 100644 --- a/docs/ci-verification.md +++ b/docs/ci-verification.md @@ -1,8 +1,25 @@ # CI verification -## Current selected-family replay: #6030 +## Current selected-family replay: #7000 with #7001 The [selected record](../certificates/selected-result.json) binds conditional +κ = 0.00363005817 to the [application review](research/pr-intake-evidence/pr7000-application/PROOF-REVIEW.md), +[validation receipts](research/pr-intake-evidence/pr7000-application/VALIDATION.json), +and [explicit source closure](../certificates/frontier7000-source.json). +`make frontier7000-verify` follows that exact-source selected replay. The review +reuses eleven authenticated preceding executions and records seven fresh serial +application, scalar and complete native-catalogue executions. The earlier +monolithic timeout is retained separately and is not completed coverage. +Consult the receipts for exact commands, source heads and limits. + +The five inherited framework assumptions remain. No complete uniform word, +literal multiplier execution, independent human proof or formal-kernel proof +is asserted. Routine research submissions do not need to rerun this +selected-family replay. The sections below preserve historical checkpoints. + +## Historical selected-family replay: #6030 + +The [historical checkpoint](../certificates/checkpoints/frontier6030.json) binds conditional κ = 0.003125804 to the [proof review](research/pr-intake-evidence/pr6030-audit/PROOF-REVIEW.md), [validation receipts](research/pr-intake-evidence/pr6030-audit/VALIDATION.json), and [explicit source closure](../certificates/frontier6030-source.json). diff --git a/research/historical-selected-replay-gap/README.md b/research/historical-selected-replay-gap/README.md index 55d8cc3b49..36d9a873ca 100644 --- a/research/historical-selected-replay-gap/README.md +++ b/research/historical-selected-replay-gap/README.md @@ -69,7 +69,11 @@ python3 -B scripts/research_package.py --package research/historical-selected-re The direct command prints `PASS_HISTORICAL_SELECTED_REPLAY_GAP`. A registered replay prints `PASS_REGISTERED_RESEARCH_PACKAGE`. Both require Git history -containing the two inspected commits. +containing the two inspected commits and the source and manifest objects at +`2485805a6d47f36ea83ddde6327d453ac82d44dc`. The verifier uses a temporary +sparse checkout of that commit, checks the declared source digests there, and +uses its planner and manifest inventory. The recorded rejections and selection +therefore remain reproducible after later changes to the live Makefile or CI guide. ## Open next questions diff --git a/research/historical-selected-replay-gap/SOURCES.json b/research/historical-selected-replay-gap/SOURCES.json index c1010eec95..62bb8b3950 100644 --- a/research/historical-selected-replay-gap/SOURCES.json +++ b/research/historical-selected-replay-gap/SOURCES.json @@ -4,13 +4,13 @@ "files": [ { "path": "Makefile", - "usage": "input", + "usage": "provenance", "git_blob": "0d683a06b9e7ee162496108ce9a7c5d2ee94fccf", "sha256": "4f2d4f2dcd15139ec82658a975f85ed300897385f85799218539f4c8d60567da" }, { "path": "certificates/partial-swap-network.json", - "usage": "input", + "usage": "provenance", "git_blob": "eccafa830c16170c36faf93045ddf1f638f7bcd4", "sha256": "4c0b9d37e1034b2076e30a70f8e514416298c462d9ee98bc87fe6556b14df786" }, @@ -22,7 +22,7 @@ }, { "path": "docs/ci-verification.md", - "usage": "input", + "usage": "provenance", "git_blob": "63ba3e9aa8bdc5e942b56e34244b92b16f854085", "sha256": "c7d3027e9ad8bcbff68f07aa9576b30fb2f3c76c44d876d40c7fdbf4a3e2228a" }, @@ -34,163 +34,163 @@ }, { "path": "research/clifford-state-performance/SOURCE.json", - "usage": "input", + "usage": "provenance", "git_blob": "6f98a0061a5104dfab82c9e806f819b41811f684", "sha256": "396d0f08df24952248140d5c65c51b8c902ef1c580892ff7ea182890530f67ee" }, { "path": "research/community-batch1047-audit/CONTENT-COVERAGE.json", - "usage": "input", + "usage": "provenance", "git_blob": "5e0f92af463ec12c5e1dcc3effd44081a88af647", "sha256": "07a9daeaf2bda308a92e9a4684f13d5fc766c7772d74801fceb7f4599fdab226" }, { "path": "research/community-batch1099-audit/CONTENT-COVERAGE.json", - "usage": "input", + "usage": "provenance", "git_blob": "9e13266bf85e627f1b7dc225e9530ea69fd2c7f0", "sha256": "421e1211fb1897c45621dc6801c276e53f2e6f335583c9c9049e535f87c0a3b1" }, { "path": "research/community-batch932-audit/CONTENT-COVERAGE.json", - "usage": "input", + "usage": "provenance", "git_blob": "cc4d03eb59b9ab3f952735f7e88ff18c2d15d7af", "sha256": "184629cac91a5cfaca50330a88596008b824221433867dcf8117539c913c284a" }, { "path": "research/community-batch970-audit/CONTENT-COVERAGE.json", - "usage": "input", + "usage": "provenance", "git_blob": "8b9f4111badf7de5d97684799560b55110e9da79", "sha256": "23188ed920e38a54006cb883e181585620821c63bf7e8b5e05d32719417cc37e" }, { "path": "research/community-frontier741-audit/CONTENT-COVERAGE.json", - "usage": "input", + "usage": "provenance", "git_blob": "a5300d4e7a5381a3f443fc54c768ca6f0f236d68", "sha256": "3c820fb30e1b4d9bcfc6a7ab524375e5b2abd66b940af2c509146104f0bcd79c" }, { "path": "research/community-frontier890-audit/CONTENT-COVERAGE.json", - "usage": "input", + "usage": "provenance", "git_blob": "d7ff6d777b01d69b6ce84caf25d3b7e4a4d2f211", "sha256": "6b185a9902652f893f3ebd8ada92d2f11b6de6533a9a4d11718a8a4496101128" }, { "path": "research/community-hyperplane599-audit/SOURCE.json", - "usage": "input", + "usage": "provenance", "git_blob": "546f37d326d10a99910ae8f908e71ef768524d98", "sha256": "65099d764bfcc39bda5e4d9966783611b094c3522dbdd1435cd498d657a8ef8e" }, { "path": "research/community-k492-audit/SOURCE.json", - "usage": "input", + "usage": "provenance", "git_blob": "e538481e4cb16ed936d3195f2dfb5b5a9ca0a76b", "sha256": "2c0f4fa5bc150d05658af2e7c54c4d2c337dd52b7efd3d69530ae73bcbdf452a" }, { "path": "research/community-reconciliation92-audit/MANIFEST.json", - "usage": "input", + "usage": "provenance", "git_blob": "2a2e93685173d5f13cbb5abe740f5d90352e0228", "sha256": "5f501f3c118131e7d069f30ad77186eb2ad6961f9680a0ef105f4514ed41808a" }, { "path": "research/partial-swap-digest/baseline.py", - "usage": "input", + "usage": "provenance", "git_blob": "a2f0190704db83bd8ef48e46e723c49b925804fe", "sha256": "81e9af92cdf38243a540cdcff2062778e449467848d444da37a9553c12ca4c34" }, { "path": "research/r420-two-sided-basis/SOURCES.json", - "usage": "input", + "usage": "provenance", "git_blob": "5d718e11e62f4bcadecf1e02d6b61f87dbe190b0", "sha256": "d9810b0d139c5a1211fd2c1f0d142187cc1fca30026e1e7bc4f9ecfacf1596d1" }, { "path": "research/skip-clones/SOURCE.json", - "usage": "input", + "usage": "provenance", "git_blob": "8d1bb3b9c8189e35e5d8bfdfa632d7473e8be0d5", "sha256": "0b68e63e35a97ce6b0cdb6e291b88ebaa8f2dd4afb70b2498f10ee89428258dc" }, { "path": "research/split-skip/SOURCE.json", - "usage": "input", + "usage": "provenance", "git_blob": "2ab610b6ea2edd2865fda793ef2168b0c76afb4f", "sha256": "7c2b4646a3ecfef6f881da5c43e4bd0fa9a4d1b00548b8ab458d897e16735134" }, { "path": "scripts/historical-ci-inputs/Makefile", - "usage": "input", + "usage": "provenance", "git_blob": "aff68409021c9d401a6f8013b9dfd7dbcc30f2a8", "sha256": "ea8830dd79da493e2d75baf753d733ff7efbcf2b4794afe41fae0ac8f3055646" }, { "path": "scripts/historical-ci-inputs/ledger-batch932-run_isolated_tests.py", - "usage": "input", + "usage": "provenance", "git_blob": "65db599ab0b1dea0312c4ca048e0bd0f44c3736f", "sha256": "96ea543e6c707424afe247e77bd86c82152f90fc76a0773c4232145da5cdf8d1" }, { "path": "scripts/historical-ci-inputs/ledger-batch932-test_isolated_test_runner.py", - "usage": "input", + "usage": "provenance", "git_blob": "34820a8f24678f918f4d0e7a35380d8013fc3495", "sha256": "628391b6abfbc8674ceec1751867a4392f2cc0b5b66886243e139af86cf355ac" }, { "path": "scripts/historical-ci-inputs/ledger-batch970-verify_selected_result.py", - "usage": "input", + "usage": "provenance", "git_blob": "c755b76e0d44ba4917aa94440ebe54daefce5b2a", "sha256": "d7fd929726b326aeeaf0a3996475adec88eee9aac09ae0bebeeaa0bc1fc8f199" }, { "path": "scripts/historical-ci-inputs/ledger-frontier741-ci-verification.md", - "usage": "input", + "usage": "provenance", "git_blob": "7f3c012504187917a728b9fe87043f0fc8cd006a", "sha256": "a854022e0e1d5d4db4d60e586be3ea56005013ef14a1fc06d02bc06e8529c4b0" }, { "path": "scripts/historical-ci-inputs/partial-swap-context.json", - "usage": "input", + "usage": "provenance", "git_blob": "ad65114886bf34a4fd27e79f3be502389e689faa", "sha256": "833a9642f33aeab2ffea17ec9c6b82e0a953f90192b2c82e10bef21538dccba9" }, { "path": "scripts/historical-ci-inputs/partial-swap-network.json", - "usage": "input", + "usage": "provenance", "git_blob": "52075f2463f60af5e16571347f24274fce73a1ea", "sha256": "f3c4be597b2f2136da4ae2cc80c7454f63c90aa01f84e5b8131cee4dd55de815" }, { "path": "scripts/partial_swap/shared.py", - "usage": "input", + "usage": "provenance", "git_blob": "44a0e454c70f9c4b98e258b8443612e6a119904b", "sha256": "5a0646a50b9f3e8ea55168620a78e8853b0ab810517cefc8cfb8696aedee278a" }, { "path": "scripts/replay_historical_check.py", - "usage": "input", + "usage": "provenance", "git_blob": "1e92fff2782970ef71378f594f2350d245a4db26", "sha256": "5467619c78183d3adc82b0cb55af28e5c1a01fca2b0588677816244b53e2bafd" }, { "path": "scripts/research_package.py", - "usage": "input", + "usage": "provenance", "git_blob": "091b4c8f58bac9a9580d76025b887dc5c873d05f", "sha256": "e1256016e7f51827d04fa2179255989f639ae1d97afaa9c16265f85a0c6abd70" }, { "path": "scripts/run_isolated_tests.py", - "usage": "input", + "usage": "provenance", "git_blob": "a96ab35b315ce492e7fbbfd6c5ae211aa066b541", "sha256": "49eb56dd0fc0d8d8ffdb209970a7adf2176223580646f575d5c22e59543b381c" }, { "path": "scripts/verify_selected_result.py", - "usage": "input", + "usage": "provenance", "git_blob": "6882a4700e410dec569b51ae2eaf1a19cc019195", "sha256": "d6b305eac5b9bdb8172c469ad30a1655a60fa0aceac8b1bf610e1b73f51babd8" }, { "path": "tests/test_isolated_test_runner.py", - "usage": "input", + "usage": "provenance", "git_blob": "343b24abf6719b81805eba5a84c74e57634c8288", "sha256": "3a5470d0b139828b90e19bf9583ddeafcd69f8e34f2824664eb3a2958a52e34d" } diff --git a/research/historical-selected-replay-gap/VERIFY.json b/research/historical-selected-replay-gap/VERIFY.json index 98553e72ed..ab7efa3471 100644 --- a/research/historical-selected-replay-gap/VERIFY.json +++ b/research/historical-selected-replay-gap/VERIFY.json @@ -4,36 +4,7 @@ "python": "3.12", "requirements": [], "inputs": [ - "Makefile", - "certificates/partial-swap-network.json", - "docs/ci-verification.md", - "research/clifford-state-performance/SOURCE.json", - "research/community-batch1047-audit/CONTENT-COVERAGE.json", - "research/community-batch1099-audit/CONTENT-COVERAGE.json", - "research/community-batch932-audit/CONTENT-COVERAGE.json", - "research/community-batch970-audit/CONTENT-COVERAGE.json", - "research/community-frontier741-audit/CONTENT-COVERAGE.json", - "research/community-frontier890-audit/CONTENT-COVERAGE.json", - "research/community-hyperplane599-audit/SOURCE.json", - "research/community-k492-audit/SOURCE.json", - "research/community-reconciliation92-audit/MANIFEST.json", - "research/partial-swap-digest/baseline.py", - "research/r420-two-sided-basis/SOURCES.json", - "research/skip-clones/SOURCE.json", - "research/split-skip/SOURCE.json", - "scripts/historical-ci-inputs/Makefile", - "scripts/historical-ci-inputs/ledger-batch932-run_isolated_tests.py", - "scripts/historical-ci-inputs/ledger-batch932-test_isolated_test_runner.py", - "scripts/historical-ci-inputs/ledger-batch970-verify_selected_result.py", - "scripts/historical-ci-inputs/ledger-frontier741-ci-verification.md", - "scripts/historical-ci-inputs/partial-swap-context.json", - "scripts/historical-ci-inputs/partial-swap-network.json", - "scripts/partial_swap/shared.py", - "scripts/replay_historical_check.py", - "scripts/research_package.py", - "scripts/run_isolated_tests.py", - "scripts/verify_selected_result.py", - "tests/test_isolated_test_runner.py" + "scripts/research_package.py" ], "commands": [ { diff --git a/research/historical-selected-replay-gap/verify.py b/research/historical-selected-replay-gap/verify.py index 83b995941a..9d266c5a4e 100644 --- a/research/historical-selected-replay-gap/verify.py +++ b/research/historical-selected-replay-gap/verify.py @@ -6,8 +6,10 @@ """ import hashlib import importlib.util +import json import subprocess import sys +import tempfile from pathlib import Path ROOT = Path(__file__).resolve().parents[2] @@ -61,31 +63,47 @@ def require(ok, message): raise SystemExit(message) -def load_checker(): - path = ROOT / "scripts/replay_historical_check.py" - spec = importlib.util.spec_from_file_location("historical_selected_replay_gap_checker", path) +def load_module(path): + spec = importlib.util.spec_from_file_location(path.stem, path) module = importlib.util.module_from_spec(spec) spec.loader.exec_module(module) return module +def checkout_snapshot(snapshot, sources, runner, env): + paths = [runner.canonical_relative_path(row["path"]) for row in sources["files"]] + git = ["git", "-c", "maintenance.auto=false", "-c", "gc.auto=0", + "-c", "core.autocrlf=false"] + subprocess.run([*git, "clone", "--quiet", "--shared", "--no-tags", "--no-checkout", + "--", str(ROOT), str(snapshot)], cwd=ROOT, env=env, check=True) + subprocess.run([*git, "update-ref", "--no-deref", "HEAD", INSPECTED], + cwd=snapshot, env=env, check=True) + # The pinned planner reads omitted manifests from the sparse HEAD/index. + patterns = "".join(f"/{path.as_posix()}\n" for path in paths) + subprocess.run([*git, "sparse-checkout", "set", "--no-cone", "--stdin"], + cwd=snapshot, env=env, input=patterns.encode("utf-8"), check=True) + subprocess.run([*git, "read-tree", "-mu", "HEAD"], cwd=snapshot, env=env, check=True) + for path, row in zip(paths, sources["files"], strict=True): + require(hashlib.sha256((snapshot / path).read_bytes()).hexdigest() == row["sha256"], + "historical checkout source digest changed: " + path.as_posix()) + + def entries(checker): for name, config in checker.CONFIG.items(): for entry in [config, *config.get("additional_inputs", [])]: yield name, entry -def reject(checker, entry): +def reject(root, checker, entry): try: - checker.authenticated_input(ROOT, entry) + checker.authenticated_input(root, entry) except ValueError as exc: return str(exc) raise SystemExit("historical input was admitted: " + entry["path"]) -def main(): - require(not sys.flags.optimize, "assertions must stay enabled") - checker = load_checker() +def check_snapshot(root, env): + checker = load_module(root / "scripts/replay_historical_check.py") selected = [entry for name, entry in entries(checker) if name == "tooling-ledger" and entry["path"] == SELECTED_PATH] require(len(selected) == 1, "selected-result checker overlay is not unique") @@ -94,7 +112,7 @@ def main(): require(entry["accepted_live_sha256"] == [SELECTED_ACCEPTED], "accepted selected-checker successors changed") frozen = (checker.INPUTS / entry["input"]).read_bytes() - live = (ROOT / SELECTED_PATH).read_bytes() + live = (root / SELECTED_PATH).read_bytes() require(hashlib.sha256(frozen).hexdigest() == SELECTED_PIN, "frozen selected checker bytes changed") require(hashlib.sha256(live).hexdigest() == SELECTED_LIVE, "live selected checker digest changed") require(SELECTED_LIVE not in (SELECTED_PIN, SELECTED_ACCEPTED), "live checker was already admitted") @@ -102,7 +120,7 @@ def main(): # live successor, not a broken comparison. require(SELECTED_PIN in (SELECTED_PIN, *entry["accepted_live_sha256"]), "pin membership control failed") - require(reject(checker, entry) == REJECTIONS[SELECTED_PATH][1], + require(reject(root, checker, entry) == REJECTIONS[SELECTED_PATH][1], "selected checker rejection text changed") counts = {path: 0 for path in REJECTIONS} @@ -110,7 +128,7 @@ def main(): for _name, item in entries(checker): if item["path"] not in REJECTIONS: continue - message = reject(checker, item) + message = reject(root, checker, item) counts[item["path"]] += 1 messages.setdefault(item["path"], message) for path, (count, message) in REJECTIONS.items(): @@ -118,20 +136,31 @@ def main(): require(messages[path] == message, "rejection text changed for " + path) require(sum(counts.values()) == 19, "rejection total changed") ci = [item for _name, item in entries(checker) if item["path"] == CI_PATH] - require(len(ci) == 1 and hashlib.sha256((ROOT / CI_PATH).read_bytes()).hexdigest() == CI_LIVE, + require(len(ci) == 1 and hashlib.sha256((root / CI_PATH).read_bytes()).hexdigest() == CI_LIVE, "selected CI note digest changed") diff = subprocess.check_output( - ["git", "diff", "--name-only", PARENT, INSPECTED], cwd=ROOT, text=True) + ["git", "diff", "--name-only", PARENT, INSPECTED], cwd=root, env=env, text=True) changed = sorted(name for name in diff.splitlines() if name) require(changed == CHANGED, "inspected commit diff changed") - sys.path.insert(0, str(ROOT / "scripts")) - from research_package import plan_details - details = plan_details(ROOT, set(changed)) + planner = load_module(root / "scripts/research_package.py") + details = planner.plan_details(root, set(changed)) require([row["package"] for row in details] == PLAN_PACKAGES, "helper plan selection changed") matched = {name for row in details for names in row["causes"].values() for name in names} require(sorted(set(changed) - matched) == UNMATCHED, "unmatched plan paths changed") - require(plan_details(ROOT, set()) == [], "empty change set selected a package") + require(planner.plan_details(root, set()) == [], "empty change set selected a package") + + +def main(): + require(not sys.flags.optimize, "assertions must stay enabled") + sources = json.loads(Path(__file__).with_name("SOURCES.json").read_bytes()) + require(sources["base_commit"] == INSPECTED, "historical checkout base changed") + runner = load_module(ROOT / "scripts/research_package.py") + env = runner.git_source_env() + with tempfile.TemporaryDirectory(prefix="historical-selected-replay-") as temporary: + snapshot = Path(temporary) / "snapshot" + checkout_snapshot(snapshot, sources, runner, env) + check_snapshot(snapshot, env) print("PASS_HISTORICAL_SELECTED_REPLAY_GAP") print("selected_checker_live=" + SELECTED_LIVE) print("rejections=19 distinct_paths=7 plan_packages=9") diff --git a/research/primal-dual-completion-rank/PROOF.md b/research/primal-dual-completion-rank/PROOF.md new file mode 100644 index 0000000000..ef1586eac7 --- /dev/null +++ b/research/primal-dual-completion-rank/PROOF.md @@ -0,0 +1,121 @@ +# Minimum rank of an invertible correction with paired support constraints + +Let $V$ be a finite-dimensional vector space over a field $K$, let +$S\le V$, and let $f:S\to V$ be linear and injective. Let +$\pi,g:V\to Q$ be surjective linear maps to the same vector space, with +$\pi f=g|_S$. We seek $P\in\mathrm{GL}(V)$ with $P|_S=f$ and +$\pi P=g$. + +For the inverse-transpose interpretation, suppose covectors $a_i$ must +be sent to $b_i$ by $P^{-\mathsf T}$, where both lists are independent. +Use $g(v)=(a_i(v))_i$ and $\pi(v)=(b_i(v))_i$. Then +$P^{-\mathsf T}a_i=b_i$ exactly when $b_i(Pv)=a_i(v)$ for all $v$. +The compatibility hypothesis is equality of the prescribed pairings on +$S$. If the source and target covector lists have dependencies, a common +independent sublist suffices when all corresponding relations agree. + +Put + +~~~text +H = ker(g - pi) +A = S ∩ H +B = f(S) ∩ H +S0 = {s in S : f(s) = s}. +~~~ + +Then the minimum of $\operatorname{rank}(P-I)$ is + +$$ +\operatorname{rank}(g-\pi)+\max(\dim A,\dim B)-\dim S_0. +$$ + +Equivalently, writing $d(s)=f(s)-s$, the minimum is + +$$ +\operatorname{rank}(g-\pi)+\operatorname{rank}d +-\min(\operatorname{rank}(\pi d),\operatorname{rank}(g d)). +$$ + +Indeed $(g-\pi)s=\pi d(s)$ and $(g-\pi)f(s)=g d(s)$ for every +$s\in S$, by compatibility. Thus the ranks of $\pi d$ and $g d$ +give the codimensions of $A$ in $S$ and $B$ in $f(S)$, +respectively. Also $S_0=\ker d$ lies in both $A$ and $B$. + +## Lower bound + +For any admissible $P$, set $F=\ker(P-I)$. Then $F\le H$, since +$gv=\pi Pv=\pi v$ on $F$, and $F\cap A=S_0$. +If $f(s)\in F$, then $P^2s=Ps$. Injectivity of $P$ gives +$Ps=s$ and $f(s)=s\in S_0$. Hence $F\cap B=S_0$ as well. +Applying the dimension formula to $F+A$ and $F+B$ inside $H$ gives + +$$ +\dim F\le\dim H-\max(\dim A,\dim B)+\dim S_0. +$$ + +Taking codimensions in $V$ gives the bound. + +## Attainment + +Choose complements $A'$ and $B'$ of $A\cap B$ in $A$ and $B$. +Take bases $a_1,\ldots,a_u$ of $A'$ and $b_1,\ldots,b_v$ of +$B'$. Let $C$ be a complement of $A+B$ in $H$. +With $p=\min(u,v)$, take + +$$ +F=S_0\oplus\operatorname{span}(a_i+b_i:1\le i\le p)\oplus C. +$$ + +This sum is direct. Inside $A+B=(A\cap B)\oplus A'\oplus B'$, +the $A'$ and $B'$ projections of the middle summand are both +injective. Therefore $F\cap A=F\cap B=S_0$, and +$\dim F=\dim S_0+p+\dim H-\dim(A+B)$ reaches the bound. + +On $W=S+F$, define $L(s+x)=f(s)+x$ for $s\in S,x\in F$. +The equality $S\cap F=S_0$ makes this well defined. The equality +$f(S)\cap F=S_0$ makes it injective: if $f(s)=-x\in S_0$, +then $f$ fixes $-x$ and injectivity of $f$ gives $s=-x$. +The map $L$ is onto $W'=f(S)+F$ and satisfies $\pi L=g$ on $W$. + +To extend $L$, map a basis of $W\cap\ker g$ to its image, a +basis of $W'\cap\ker\pi$. Extend both to bases of the full kernels +and pair the added vectors arbitrarily. Next choose a basis of +$g(W)$ and take lifts in $W$. Pair them with their $L$-images in +$W'$. Complete that quotient basis to a basis of $Q$ and pair any +remaining $g$-lifts with $\pi$-lifts of the same basis vectors. +These pairs specify $P$ between two bases of $V$. They preserve $L$ +on $W$ and satisfy $\pi P=g$. In particular $P$ fixes $F$, +so the lower bound is attained. The construction uses basis extensions +and linear solves over $K$, including $K=\mathbb F_2$. + +## What invertibility costs + +For arbitrary endomorphisms $T$ with the same equations $T|_S=f$ +and $\pi T=g$, the minimum rank of $T-I$ is + +$$ +\operatorname{rank}(g-\pi)+\operatorname{rank}d-\operatorname{rank}(\pi d). +$$ + +In this case every fixed space lies in $H$ and meets $A$ in $S_0$. +For attainment take $S_0$ plus a complement of $A$ in $H$ as +the fixed subspace and define $L$ as before. On a complement of +$S+F$ extend $L$ by choosing any $\pi$-lift of $g$ on each +basis vector. This proves the relaxed formula. The increase imposed +by invertibility is exactly +$\max(0,\operatorname{rank}(\pi d)-\operatorname{rank}(g d))$. + +For a smallest-dimensional example with a strict increase, take +$V=K^2$, $S=Ke_1$, $f(e_1)=e_2$, $\pi=(0,1)$ and +$g=(1,1)$. The two constraint equations force the shape + +$$ +P=\begin{pmatrix}0&a\\1&1\end{pmatrix}. +$$ + +At $a=0$ the matrix is singular with correction rank one. +If $a\ne0$, both $\det P$ and $\det(P-I)$ equal $-a$. +The invertible minimum is two over every field. In dimensions at most +one, no strict increase is possible. A nonzero $d$ in dimension one +requires $S=V$, where the injective $f$ already fixes the whole +completion. If $d=0$, the two formulas coincide. diff --git a/research/primal-dual-completion-rank/README.md b/research/primal-dual-completion-rank/README.md new file mode 100644 index 0000000000..ed36c726af --- /dev/null +++ b/research/primal-dual-completion-rank/README.md @@ -0,0 +1,104 @@ +# Exact correction rank for paired primal and dual support maps + +This contribution was produced by Nick Petalas. Apache-2.0. + +**Claim type.** Scoped algebraic proof with exact finite corroboration. + +Let an injective map $f:S\to V$ prescribe the primal action. Let +surjections $g,\pi:V\to Q$ prescribe $P^{-\mathsf T}$ on dual support +through the equation $\pi P=g$. Under the compatibility condition +$\pi f=g|_S$, the minimum of $\operatorname{rank}(P-I)$ among +invertible completions $P$ is + +$$ +\operatorname{rank}(g-\pi)+\operatorname{rank}d +-\min(\operatorname{rank}(\pi d),\operatorname{rank}(g d)), +\qquad d(s)=f(s)-s. +$$ + +[PROOF.md](PROOF.md) constructs a largest possible fixed subspace and +extends the paired constraints by basis completion. The proof works over +every field. Allowing singular completions can lower the minimum. A +two-dimensional example has relaxed minimum one and invertible minimum +two over every field. + +The support-completion route in GQ08 uses low-rank corrections to limit +the size of the auxiliary inverse in G17. The formula gives the smallest +possible rank for that step before factoring or pricing the coefficients. +It also detects when a low-rank completion of the primal and row equations +is too optimistic for a reversible word. [Atlas, #5803](https://github.com/CrocSwap/integer-mult-bounds/pull/5803). + +## Evidence and reproduction + +The verifier enumerates every binary matrix action in dimensions one +through three. For every primal subspace and dual row space it groups +the matrices by their literal primal and quotient actions, then compares +the minima with and without invertibility to the proof's formula. For +each dual row space it uses one basis. A different basis only changes +both quotient maps by the same invertible change of coordinates. + +| Dimension | All matrices | Invertible matrices | Compatible contracts | Strict rank increases | +|---|---:|---:|---:|---:| +| 1 | 2 | 1 | 4 | 0 | +| 2 | 16 | 6 | 118 | 6 | +| 3 | 512 | 168 | 21,232 | 2,016 | + +The verifier also checks the explicit two-dimensional gap and rejects a +mutated minimum that omits the invertibility charge. Its JSON output +records the counts and hashes of the six package files. The verification +commands are read-only. The census uses only the standard library. Its +matrix size is bounded by three, and live collections use managed Python +allocation within that finite domain. + +From the repository root, run these with CPython 3.12. A different +installation can substitute its interpreter path. + +~~~sh +/Users/nick/git/integer-mult-bounds/.venv/bin/python3.12 -I -W error -B research/primal-dual-completion-rank/verify.py +/Users/nick/git/integer-mult-bounds/.venv/bin/python3.12 -I -O -W error -B research/primal-dual-completion-rank/verify.py +/Users/nick/git/integer-mult-bounds/.venv/bin/python3.12 -B scripts/research_package.py --package research/primal-dual-completion-rank +/Users/nick/git/integer-mult-bounds/.venv/bin/python3.12 -B scripts/check_research_commit.py --package research/primal-dual-completion-rank --check-source-history +~~~ + +The direct marker is `PASS_PRIMAL_DUAL_COMPLETION_RANK`. The two +repository commands report `PASS_REGISTERED_RESEARCH_PACKAGE` and +`PASS_RESEARCH_COMMIT_BYTES`, respectively. + +## Scope + +This is the algebraic part of GQ08. The construction works with compatible +maps over fields. Dyadic-unit realizability, coefficient growth, a retained +caller, signed native realization, owner restoration, tape and precision +costs, and the complete signed moment remain open. The selected $\kappa$ +and its five framework assumptions are outside this scope. +Independent human and formal proof-kernel review are unclaimed. +Predecessor scientific replays, Lean, Docker and the selected cold replay +were unrun. + +## Open next questions + +- Does a named retained cut admit an optimal correction with paid scalar + coefficients? First extract its two exact support maps and compute the + field minimum. The resulting completion needs a literal word and a + complete restoration and caller invoice. +- Which paired contracts over $\mathbb Z[1/2]$ attain the field + minimum by a dyadic-unit completion? First test the optimal fixed + subspace for a complementary saturated submodule and check its + reductions at odd primes. A field complement need not give an + admissible scalar word. + +## Credit + +David Ganz / Dugongue posed GQ08 and supplied G16's extension +parameterization and G17's low-rank inverse formula. Ali Romman's +dyadic helper-basis work is a related implementation precedent for +one normalized row/column pair. The ordinary block rank-completion +problem is classical. The contribution here is the fixed-space bound +and attaining construction under simultaneous invertibility and paired +support constraints. [#5803](https://github.com/CrocSwap/integer-mult-bounds/pull/5803), +[#5611](https://github.com/CrocSwap/integer-mult-bounds/pull/5611), +[Epperly, Govindarajan and Chandrasekaran, §2](https://arxiv.org/abs/2106.11267). + +[Advance claim](https://github.com/CrocSwap/integer-mult-bounds/discussions/1166#discussioncomment-19693317). +The atlas and dyadic precedent are pinned as historical provenance in +[SOURCES.json](SOURCES.json). diff --git a/research/primal-dual-completion-rank/SOURCES.json b/research/primal-dual-completion-rank/SOURCES.json new file mode 100644 index 0000000000..80b2c850d4 --- /dev/null +++ b/research/primal-dual-completion-rank/SOURCES.json @@ -0,0 +1,30 @@ +{ + "schema_version": 1, + "base_commit": "e54523517d0f03e58352e647c403a5af3e525a57", + "files": [ + { + "path": "certificates/selected-result.json", + "usage": "provenance", + "git_blob": "75d77b328ea02ded535e7f23aee5b1e55d0bb1a8", + "sha256": "154c3d30455d778fd3343b6b4851f0fa4665f0f7625b0adb3675e2bafbbf6bf1" + }, + { + "path": "docs/research/agent-task-catalog.json", + "usage": "provenance", + "git_blob": "3404fbdd5e7529a0674eef26114b5071e4586cec", + "sha256": "ead6cdc9d0af6d58e8a49744d8ec8c7c15e13d2ba93c276c434bcf9b51b93a81" + }, + { + "path": "research/frontier-mathematical-atlas/volume-geometry.md", + "usage": "provenance", + "git_blob": "4f094d7163f989c88ed57d0351438aa2d8319d43", + "sha256": "42973fe7eb01181adcb604878f6402b02539724fc2620761464e1fbd3bc1c1e4" + }, + { + "path": "research/dyadic-helper-basis-completion/README.md", + "usage": "provenance", + "git_blob": "595d586859ca22e488a675bb1b78985fa8157d36", + "sha256": "6eec44155e30cf8b100cfd2e6378d0dd60da1cdf496e3b88251a40018fe9bcb3" + } + ] +} diff --git a/research/primal-dual-completion-rank/TASK.json b/research/primal-dual-completion-rank/TASK.json new file mode 100644 index 0000000000..73079101cf --- /dev/null +++ b/research/primal-dual-completion-rank/TASK.json @@ -0,0 +1,83 @@ +{ + "schema_version": 1, + "task_id": "complex-rank-shape-decoder", + "priority_revision": 8, + "base_commit": "e54523517d0f03e58352e647c403a5af3e525a57", + "original_card": { + "id": "complex-rank-shape-decoder", + "question": "Can a legal decoder or coframe change improve the full signed moment, including changed helper targets?", + "starting_points": [ + "docs/research/current-priorities.md", + "docs/research/research-family-status.md", + "certificates/selected-result.json", + "docs/research/pr-intake-evidence/pr5599-completion/PROOF-REVIEW.md", + "docs/research/recursive-invoice-preflight.md" + ], + "methods": [ + "Construct and independently check one source-bound instance of the stated first experiment.", + "Seek an exact counterexample or target-specific complete-budget obstruction to that same proposed change." + ], + "acceptance_gate": "A source-bound arbitrary-input identity and legal producer/consumer boundary with the complete paid and productive target invoice, including restoration, ownership, native alignment and extra charges. Price under the actual current model; historical ceiling.py assumptions must be established before use. A favorable optimistic relaxation keeps a route open but is not a completed supplier. A scoped local result remains useful; expansion requires the family-level budget gate.", + "first_step": { + "command": [ + "python3", + "-B", + "scripts/research_selected_baseline.py" + ], + "expected": "PASS_RESEARCH_BASELINE_BINDINGS_ONLY", + "inspect": "Read revision 8, the selected review and live exact-source handoffs. Then follow this card's bounded first_experiment.", + "scope": "Source-bound orientation only, not a supplier replay or theorem proof." + }, + "first_experiment": "Price one complete changed itinerary with both paid and productive histograms. Check #5693/#5698 for overlap and their exact hypotheses before another coframe search. Use the signed invoice preflight where its boundary contract agrees; an inconclusive concave comparison is not a power crossing. Stop if an optimistic complete budget fails at the chosen target or native delivery remains unspecified.", + "bottleneck_link": "Judge the actual current supplier contract and complete signed target/paid bill; frozen-profile ceilings require an applicability proof.", + "allowed_variations": [ + "One exact source head, helper/cut region or quantified proposition and changed hypothesis; check live handoffs first. Different seeds are not different scopes." + ], + "scope_change_triggers": [ + "Changing source head, helper/cut domain, operation alphabet or caller boundary without recording the new scope.", + "Treating an archived proposal or local numerical pass as complete supplier admission." + ], + "overlap_terms": [ + "coframe", + "signed moment", + "5693", + "5698" + ], + "existing_checks": [ + { + "scope": "Selected PR5599: complete helper-subspace construction and independent caller replay; five framework assumptions remain.", + "evidence": "docs/research/pr-intake-evidence/pr5599-completion/PROOF-REVIEW.md", + "remaining": "The next experiment must identify uncovered source-specific work. Successor archive intake does not transfer this admission.", + "status": "reviewed", + "source_commit": "09dd0965711ca7eeb169057a420abab1c9052137", + "selected_record_sha256": "2a0220e866417774606e0f991e41b9c1d26b1f43b97857238cf478dbc7d330b5" + } + ] + }, + "scope_status": "changed", + "actual_scope": "Prove the exact minimum rank of P-I over invertible field-linear completions with prescribed primal action on a subspace and prescribed inverse-transpose action on dual support.", + "scope_change_reason": "GQ08 in David Ganz / Dugongue PR5803 seeks a cheap installed primal and dual support completion. This scoped algebraic step determines the optimum correction rank before a caller or a paid word is chosen.", + "bottleneck_link": "The correction rank fixes the minimal dimension in the atlas G17 low-rank inverse formula. Source installation and its complete signed paid and productive invoice are separate.", + "budget_minutes": 60, + "stop_rule": "Ship the first exact rank formula with a construction and a complete small binary corroboration.", + "outcome": "completed", + "evidence": "PROOF.md gives the universal construction and a strict two-dimensional separation. The verifier enumerates all binary matrices through dimension three and checks 21,354 normalized contracts, including 2,022 strict gaps.", + "author": "Nick Petalas", + "ai_assistance": "Independent human and formal proof-kernel review are unclaimed.", + "open_questions": [ + { + "question": "Can a named retained cut use a rank-optimal correction in a paid scalar word?", + "why_it_matters": "The algebraic optimum can become a bound on the size of the auxiliary inverse in a source-bound construction.", + "first_step": "Extract the exact primal and dual support maps and compute their minimum over the coefficient field.", + "blocked_by": "A concrete compatible map pair, a literal word and its complete restored caller invoice." + }, + { + "question": "Which compatible dyadic contracts attain the field minimum in GL over Z[1/2]?", + "why_it_matters": "An optimal field factor may use forbidden odd denominators or a nonunit determinant.", + "first_step": "Test whether an optimal fixed subspace admits a complementary saturated dyadic submodule, including at odd-prime reductions.", + "blocked_by": "A criterion for compatible basis extension over the dyadic module." + } + ], + "question_source": "https://github.com/CrocSwap/integer-mult-bounds/pull/5803", + "claim_url": "https://github.com/CrocSwap/integer-mult-bounds/discussions/1166#discussioncomment-19693317" +} diff --git a/research/primal-dual-completion-rank/VERIFY.json b/research/primal-dual-completion-rank/VERIFY.json new file mode 100644 index 0000000000..84d103ed67 --- /dev/null +++ b/research/primal-dual-completion-rank/VERIFY.json @@ -0,0 +1,34 @@ +{ + "schema_version": 1, + "status": "ready", + "python": "3.12", + "requirements": [], + "inputs": [], + "commands": [ + { + "argv": [ + "{python}", + "-I", + "-W", + "error", + "-B", + "{package}/verify.py" + ], + "timeout_seconds": 120, + "checks": "Complete binary rank census through dimension three and explicit invertibility-gap control." + }, + { + "argv": [ + "{python}", + "-I", + "-O", + "-W", + "error", + "-B", + "{package}/verify.py" + ], + "timeout_seconds": 120, + "checks": "Complete binary rank census through dimension three and explicit invertibility-gap control." + } + ] +} diff --git a/research/primal-dual-completion-rank/verify.py b/research/primal-dual-completion-rank/verify.py new file mode 100644 index 0000000000..d63ff33c3b --- /dev/null +++ b/research/primal-dual-completion-rank/verify.py @@ -0,0 +1,171 @@ +#!/usr/bin/env python3 +"""Corroborate the paired-support formulas by complete small binary censuses.""" + +from functools import reduce +from hashlib import sha256 +from itertools import combinations, product +import json +from operator import xor +from pathlib import Path +import platform + + +def require(condition, message): + if not condition: + raise ValueError(message) + + +def span(vectors): + points = {0} + for vector in vectors: + points |= {point ^ vector for point in points} + return frozenset(points) + + +def subspaces(dimension): + nonzero = range(1, 1 << dimension) + return sorted({ + tuple(sorted(span(vectors))) + for size in range(dimension + 1) + for vectors in combinations(nonzero, size) + }) + + +def quotient_table(covectors, order): + basis = [] + for row in covectors: + if row not in span(basis): + basis.append(row) + return tuple( + sum(((v & row).bit_count() % 2) << i for i, row in enumerate(basis)) + for v in range(order) + ) + + +def dimension(points): + count = len(points) + require(count > 0 and count & (count - 1) == 0, + f"invalid subspace cardinality: {count}") + return count.bit_length() - 1 + + +def matrix_actions(size): + """Enumerate the literal action table of every binary square matrix.""" + order = 1 << size + actions = [] + for columns in product(range(order), repeat=size): + table = tuple( + reduce(xor, (c for i, c in enumerate(columns) if x & (1 << i)), 0) + for x in range(order) + ) + invertible = len(set(table)) == order + rank = dimension({x ^ value for x, value in enumerate(table)}) + actions.append((table, invertible, rank)) + return actions + + +def completion_minima(actions, source, pi): + """Group enumerated matrices by the exact equations each one satisfies.""" + ordinary = {} + invertible = {} + quotient_size = len(set(pi)) + for table, unit, rank in actions: + target = tuple(table[x] for x in source) + if len(set(target)) != len(source): + continue + g = tuple(pi[image] for image in table) + if len(set(g)) != quotient_size: + continue + contract = (target, g) + ordinary[contract] = min(rank, ordinary.get(contract, rank)) + if unit: + invertible[contract] = min(rank, invertible.get(contract, rank)) + require(ordinary.keys() == invertible.keys(), + "some compatible contract has no invertible completion") + return ordinary, invertible + + +def predicted_minima(source, target, pi, g): + ambient = dimension(range(len(pi))) + fixed = {x for x, y in zip(source, target, strict=True) if x == y} + equalizer = {x for x in range(len(pi)) if g[x] == pi[x]} + base = ambient - dimension(equalizer) - dimension(fixed) + source_inside = dimension(equalizer.intersection(source)) + target_inside = dimension(equalizer.intersection(target)) + return base + source_inside, base + max(source_inside, target_inside) + + +def check_minima(source, target, pi, g, observed): + expected = predicted_minima(source, target, pi, g) + require(observed == expected, + f"incorrect minima for {source}, {target}, {pi}, {g}: " + f"enumerated {observed}, formula {expected}") + + +def census(size): + actions = matrix_actions(size) + spaces = subspaces(size) + quotients = [quotient_table(space, 1 << size) for space in spaces] + gap_counts = [0] * (size + 1) + for source, pi in product(spaces, quotients): + ordinary, invertible = completion_minima(actions, source, pi) + for (target, g), free_rank in ordinary.items(): + unit_rank = invertible[target, g] + check_minima(source, target, pi, g, (free_rank, unit_rank)) + gap_counts[unit_rank - free_rank] += 1 + return { + "dimension": size, + "matrices": len(actions), + "invertible_matrices": sum(unit for _, unit, _ in actions), + "primal_subspaces": len(spaces), + "dual_subspaces": len(quotients), + "contracts": sum(gap_counts), + "counts_by_rank_increase": gap_counts, + } + + +def check_gap_fixture(): + source, target = (0, 1), (0, 2) + pi, g = (0, 0, 1, 1), (0, 1, 1, 0) + ordinary, invertible = completion_minima(matrix_actions(2), source, pi) + observed = ordinary[target, g], invertible[target, g] + require(observed == (1, 2), f"unexpected two-dimensional fixture: {observed}") + check_minima(source, target, pi, g, observed) + try: + check_minima(source, target, pi, g, (1, 1)) + except ValueError as error: + require("enumerated (1, 1), formula (1, 2)" in str(error), + "wrong reason for rejecting a missing invertibility charge") + else: + raise ValueError("accepted a missing invertibility charge") + return {"relaxed": observed[0], "invertible": observed[1]} + + +def package_hashes(): + directory = Path(__file__).resolve().parent + names = ("PROOF.md", "README.md", "SOURCES.json", "TASK.json", + "VERIFY.json", "verify.py") + hashes = {} + for name in names: + with (directory / name).open("rb") as stream: + raw = stream.read(65_537) + require(0 < len(raw) <= 65_536, f"unexpected package file size: {name}") + hashes[name] = sha256(raw).hexdigest() + return hashes + + +def main(): + fixture = check_gap_fixture() + reports = [census(size) for size in range(1, 4)] + print(json.dumps({ + "status": "PASS_PRIMAL_DUAL_COMPLETION_RANK", + "python": platform.python_version(), + "censuses": reports, + "sharp_fixture": fixture, + "rejected_missing_invertibility_charge": 1, + "sha256": package_hashes(), + }, sort_keys=True)) + + +if __name__ == "__main__": + main()