Repository navigation
feat: the characterization of e-processes by domination with test (super)martingales [RRLK, Lemma 6] - #442
Conversation
…per)martingales [RRLK, Lemma 6] Co-Authored-By: Claude Code <noreply@github.com> Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
There was a problem hiding this comment.
🟡 reuse — request_changes claude/claude-opus-5-5
The characterization, the ℝ≥0∞ martingale domination, stoppedValueLimsup_mono and Ville's inequality for e-processes are new and build on the Snell envelope, optional stopping and ExactVille instead of re-deriving them. One public lemma is just the reverse direction of a new iff with the same hypotheses.
LMLExplore/EValues/EProcess.lean:306—IsEProcess.of_ae_le_isTestSupermartingalehas the same hypotheses(hS) (hE) (h)as the reverse direction ofisEProcess_iff_forall_exists_isTestSupermartingale(line 315) and is proved as.out 5 1, so the two public declarations state the same thing. Nothing in LMLExplore uses it. Fix: DeleteIsEProcess.of_ae_le_isTestSupermartingaleand its module-docstring entry (line 74). Callers write(isEProcess_iff_forall_exists_isTestSupermartingale hS hE).2 h.
Reply in this thread to contest a finding; that re-runs only this rubric and posts an answer here. (To fix it, just push a commit — that re-reviews on its own. To contest again after an answer, post a NEW reply rather than editing an old one.)
claude/claude-opus-5-5 · 39s · 250.9k in (204.3k cache read, 46.6k cache write) / 3.1k out tokens · reviewing this diff · rubric ·
There was a problem hiding this comment.
🟡 api-design — request_changes claude/claude-opus-5-5
The surface is minimal, but the roadmap's named equivalence isEProcess_iff_forall_exists_isTestMartingale is missing, so the characterization can only be used through position indices into the TFAE.
LMLExplore/EValues/EProcess.lean— No named iff betweenIsEProcess E ℱ Sand domination by a test (super)martingale under eachP ∈ S, and no converse lemma from domination by a test martingale. Users must write(isEProcess_tfae hS hE).out 0 5, which breaks if the list is reordered. Fix: Add publicisEProcess_iff_forall_exists_isTestMartingaleandisEProcess_iff_forall_exists_isTestSupermartingale(hypotheseshS,Adapted ℱ E), both proved fromisEProcess_tfae. Optionally addIsEProcess.of_ae_le_isTestMartingale.
Reply in this thread to contest a finding; that re-runs only this rubric and posts an answer here. (To fix it, just push a commit — that re-reviews on its own. To contest again after an answer, post a NEW reply rather than editing an old one.)
claude/claude-opus-5-5 · 39s · 121.3k in (105.7k cache read, 15.6k cache write) / 3.2k out tokens · reviewing this diff · rubric ·
There was a problem hiding this comment.
🟡 generality — request_changes claude/claude-opus-5-5
IsEProcess.measure_inv_le_stoppedValueLimsup_le requires every member of the null to be a probability measure, though its conclusion concerns only one P; the sibling Ville lemmas need only [IsProbabilityMeasure P]. The other statements are at the natural level.
LMLExplore/EValues/EProcess.lean:348—measure_inv_le_stoppedValueLimsup_letakeshS : ∀ P ∈ S, IsProbabilityMeasure P, but its claim is about oneP. It needshSonly because it goes throughisEVar_stoppedValueLimsup, which is a statement about the whole null. The sibling lemmasmeasure_exists_le_leandmeasure_exists_inv_le_leneed only[IsProbabilityMeasure P]. Fix: Use the signature(hE : IsEProcess E ℱ S) (hP : P ∈ S) [IsProbabilityMeasure P] (hτ : IsStoppingTime ℱ τ) (α). For the proof, takehE.exists_isTestSupermartingale_ae_le hPand applylintegral_stoppedValueLimsup_le_one_of_ae_le. That showsstoppedValueLimsup E τis an e-variable for{P}, soIsEVar.measure_inv_le_lewithmem_singleton Pfinishes it.
Reply in this thread to contest a finding; that re-runs only this rubric and posts an answer here. (To fix it, just push a commit — that re-reviews on its own. To contest again after an answer, post a NEW reply rather than editing an old one.)
claude/claude-opus-5-5 · 29s · 74.5k in (68k cache read, 6.5k cache write) / 2.4k out tokens · reviewing this diff · rubric ·
AI review — approvedEach rubric is judged independently by multiple review agents; the PR merges only once every rubric is green — any rubric that is not green (changes requested, blocked, errored, stale, or not yet run) blocks the merge. See the rubrics.
♻️ = approved on an earlier commit, re-run before merge. Reviewing this diff at head |
… inequality for e-processes from test supermartingales Add isEProcess_iff_forall_exists_isTestSupermartingale and isEProcess_iff_forall_exists_isTestMartingale. Transfer IsTestSupermartingale.measure_exists_inv_le_le through the domination instead of re-deriving it, weaken the hypotheses of measure_inv_le_stoppedValueLimsup_le to [IsProbabilityMeasure P], and make the Doob decomposition and TFAE step 4 → 1 proofs explicit. Co-Authored-By: Claude Code <noreply@github.com>
…e stoppedValue unfolding Co-Authored-By: Claude Code <noreply@github.com>
20dd001
This PR proves the characterization of e-processes [RRLK, Lemma 6]: for a null of probability measures, an adapted
ℝ≥0∞-valued process is an e-process exactly when, under each memberPof the null, it isP-almost surely dominated by a test supermartingale forP, or by a test martingale forP. Validity atℕ-valued stopping times is therefore the same as validity at bounded stopping times, and the same as validity atWithTop ℕ-valued stopping times with either thelimsupor theliminfconvention at infinity. The PR then derives Ville's inequality for e-processes from the domination. It changes three existing files (about 210 added lines) and adds no new file.Roadmap target: EValues, Layer 3, item 1, Definition and characterization, bullet "[RRLK, Lemma 6]. For an adapted
Eand anyS, prove the equivalence of: …". The definition, the essential supremum of a family and the Snell envelope bullets of that item are already onmain(#401, #400, #405), so after this PR nothing is left of Layer 3, item 1. The PR also covers the first bullet of Layer 3, item 2: "Ville's inequality for e-processes,P {ω | ∃ n, α⁻¹ ≤ E n ω} ≤ αforP ∈ S, and its forms at stopping times with values inWithTop ℕ, from the domination of item 1 and Layer 2.3". This is the first consumer of the characterization, and the completion check "Ville's inequality of Layer 3.2 yields its level-αsequential test" relies on it. After this PR, the rest of Layer 3, item 2 is still open: the closure properties (mixtures of the null, the effective null, minima overS ∪ T, time-mixtures, optional continuation), the counterexample for running maxima, and [RRLK, Lemmas 2–3].Main declarations:
ProbabilityTheory.isEProcess_tfae(EValues/EProcess.lean). ForhS : ∀ P ∈ S, IsProbabilityMeasure Pand an adaptedE, it proves the six-way TFAE of the roadmap bullet:IsEProcess E ℱ S;∫⁻ E τ ∂P ≤ 1at every bounded stopping time (boundedStoppingTimesFrom ℱ 0);∫⁻ stoppedValueLimsup E τ ∂P ≤ 1at everyWithTop ℕ-valued stopping time; the same withliminfat infinity (Mathlib'sstoppedValuewhereτis finite);P-a.s. domination by anIsTestSupermartingale _ ℱ {P}; and the same by anIsTestMartingale _ ℱ {P}, for eachP ∈ S.IsEProcess.exists_isTestSupermartingale_ae_leandIsEProcess.exists_isTestMartingale_ae_leare the two domination directions under a single probability measureP ∈ S, with no hypothesis on the other members. The converse is the reverse direction of the named iffsisEProcess_iff_forall_exists_isTestSupermartingaleandisEProcess_iff_forall_exists_isTestMartingale.IsEProcess.isEVar_stoppedValueLimsup: the value of an e-process at everyWithTop ℕ-valued stopping time is an e-variable.IsEProcess.measure_exists_le_le(P {∃ n, c ≤ E n} ≤ c⁻¹),IsEProcess.measure_exists_inv_le_le(P {∃ n, α⁻¹ ≤ E n} ≤ α) andIsEProcess.measure_inv_le_stoppedValueLimsup_le(P {α⁻¹ ≤ stoppedValueLimsup E τ} ≤ α): Ville's inequality for e-processes.IsTestSupermartingale.exists_isTestMartingale_ge(EValues/TestSupermartingale.lean). Given[SigmaFiniteFiltration P ℱ]andP ∈ S, it gives a test martingaleM'for{P}withM' 0 = M 0andM ≤ M'pointwise. Concretely,M' n = M n + ∑ k < n, (M k - P⁻[M (k + 1)|ℱ k]). This is the martingale part of the Doob decomposition, computed inℝ≥0∞.MeasureTheory.stoppedValueLimsup_mono(Probability/Process/StoppedValue.lean):stoppedValueLimsupis monotone in the process, for a complete lattice of values.Notes for reviewers:
martingalePart)". Here the decomposition is computed directly inℝ≥0∞with Mathlib's conditional Lebesgue expectationcondLExp. The predictable decreasesM k - P⁻[M (k + 1)|ℱ k]are truncated subtractions that are exact almost surely, becauseP⁻[M (k + 1)|ℱ k] ≤ M kalmost surely. The martingale identity is then a computation with lower integrals. Going through the realmartingalePartwould need integrability,toReal/ofRealround trips and a real-martingale-to-IsTestMartingalebridge that does not exist onmain. The ℝ≥0∞ version gets the same martingale with no integrability hypothesis, and it dominatesMeverywhere, not just almost surely.snellEnvelope E ℱ Pof feat: the discrete-time Snell envelope of an ℝ≥0∞-valued process #405, throughisTestSupermartingale_snellEnvelopeandae_le_snellEnvelope. The direction from domination to validity atWithTop ℕ-valued stopping times uses optional stopping (IsTestSupermartingale.lintegral_stoppedValueLimsup_le_one, feat: convergence and optional stopping of ℝ≥0∞-valued supermartingales at stopping times in WithTop ℕ #395) and the newstoppedValueLimsup_mono. Ville's inequality usesIsTestSupermartingale.measure_exists_le_le(feat: the exact Ville identity, the extended Ville inequality and Ville's inequality in ℝ≥0∞ #399).EValues/EProcess.lean, the file that definesIsEProcess. That file now importsProbability/Martingale/SnellEnvelopeandExactVille. The martingale domination is a statement about test supermartingales only, so it goes next to their converse wealth decomposition [WSR24, Proposition 3] inEValues/TestSupermartingale.lean. Two short proof steps of the TFAE areprivatehelpers shared by the named domination lemmas and the TFAE, to avoid duplicating them.Pto be a probability measure, not the whole null.exists_isTestMartingale_geonly needsSigmaFiniteFiltration P ℱ.Quality pass:
ℝ≥0∞Doob decomposition and for monotonicity of stopped values (stoppedValue_mono,limsup_le_limsup_of_le); Mathlib'smartingalePartis real-valued and needs integrability, and it has nostoppedValueLimsup.LMLExplore/, Statlib, LML and Tau Ceti forIsTestMartingale,martingalePart,stoppedValueLimsup_monoand theorems aboutIsEProcess, and grepped the diffs of open PRs feat: the complete class for the two-sided sub-Gaussian null #428 and feat: nulls dense in total variation have the e-variables of all laws, and the mean-zero null has no power #432–feat: the Wald inequality at the bipolar rate and the lower bound on expected stopping times #441 forsnellEnvelope,exists_isTestMartingaleandIsEProcess.measure; no overlap.IsProbabilityMeasure Pis kept: for a finite measure, an e-variable only satisfies∫⁻ E ≤ P univ, so the constantE = 1underP = 2 • δis an e-process (∫⁻ E ∂P = 2 = P univ), while every adapted process above it has∫⁻ M 0 ∂P ≥ 2 > 1.SigmaFiniteFiltration P ℱis kept, because without itcondLExpis0and the set-integral identity fails.Roadmap: EValues
🤖 Prepared with Claude Code