fix(genai,#18733): DALL-E ex.4 corriger unites HSV Pillow 0-255 + circularite teinte

[jsboige]

Grain: MED/notebook-python -- lane myia-po-2027:CoursIA -- prev: MED/notebook-python #18754

Summary

Corrige K04 de l'audit Astra 2026-10-01 (parent #18731, audit-reassessment CONFIRMED par ai-01, rejoue avec Pillow).

DALL-E 3 Foundation (MyIA.AI.Notebooks/GenAI/Image/01-Foundation/01-1-OpenAI-DALL-E-3.ipynb), exercice 4 (Indice 1, cellule 30) :

• Ancien : intervalles en degrés (0-60°, 180-240°) — faux pour Pillow qui encode H sur 0-255.
• Ancien : moyenne arithmétique simple — échoue sur la circularité de la teinte (témoin audit : deux rouges (255,0,6) et (255,6,0) → moyenne 127 = cyan au lieu de rouge).

Correction :

• Explicite les unités 0-255 (bleu pur = 170, pas 240).
• Propose de masquer les pixels peu saturés (un gris a H = 0).
• Propose proportion de pixels par secteur (chaud H < 30 ou H > 230 ; froid 100 < H < 170) ou moyenne circulaire avec concentration.

Markdown-only, pas de re-exec (C.3 — aucune cellule code modifiée).

Acceptation

• Un rouge qui traverse la coupure reste rouge (moyenne circulaire proposée)
• Un bleu pur est classé froid (H = 170 explicite)
• Un gris n'est pas déclaré chaud (masquage S faible proposé)
• Mélange sans direction dominante signalé (concentration circulaire proposée)

See #18733

🤖 Generated with Claude Code

[github-actions]

Golden-Set Execution (H.7 P3)

✅ 8/8 notebooks passed (certified reproducible)

Notebook	Status	Time
2.1-Workflow-ML.ipynb	✅ SUCCESS	3.5s
2.2-Descente-de-gradient.ipynb	✅ SUCCESS	3.9s
2.3-Regression-lineaire-logistique.ipynb	✅ SUCCESS	4.2s
2.4-Arbres-Forets-Ensembles.ipynb	✅ SUCCESS	4.4s
Search-01-StateSpace.ipynb	✅ SUCCESS	3.0s
SL-1-LogicalLearning.ipynb	✅ SUCCESS	2.1s
rl_4_multi_armed_bandits.ipynb	✅ SUCCESS	15.9s
GameTheory-04c-NashExistence-Python.ipynb	✅ SUCCESS	2.3s

Pinned lockfile: scripts/notebook_tools/golden_set.lock.txt (H.7 P3, axe A #4208)

[github-actions]

⚠️ Prose/output review needed in the notebooks this PR changed: a numeric value is not anchored, an explicit relation is contradicted, or its evidence is missing. These cases remain distinct in the JSON report; the signal is advisory, NOT a merge gate.

Scope = notebooks CHANGED in this PR, not the whole corpus. Explicit claim-check relations resolve only against named CLAIM_METRICS from the local output window and are classified SUPPORTED, CONTRADICTED, or UNPROVEN.
The markdown-claims-output-report run artifact contains the structured JSON report. See python scripts/check_markdown_claims_output.py --help for re-running locally.
Detector rationale: c.290 / c.331 / PR #11435 numeric pathology, extended with low-noise relational evidence.

[github-actions]

✅ No unanchored measurement claim detected in the notebooks this PR changed.

Scope = notebooks CHANGED in this PR, not the whole corpus. The stale-claim-report run artifact holds the structured JSON.
Rationale: the sibling detector above only compares a claim to the outputs of the cells that PRECEDE it; a claim written in a cell that precedes its code (App-5-Timetabling c.2/c.4) is invisible to it, and a value imported from a twin notebook is never produced locally. See python scripts/check_stale_claims.py --help.

[github-actions]

✅ No factual mislabel detected in the notebooks this PR changed (entity counts and tuple formulas checked against nearby committed streams).

Scope = notebooks CHANGED in this PR, not the whole corpus. The factual-mislabel-report run artifact holds the structured JSON.
Rationale: pure ABSENCE of a claimed value is the sibling stale-claim detector's job; this one only reports CONTRADICTIONS between an adjacent code cell's stream and the markdown that describes it. See python scripts/check_factual_mislabel.py --help.

[github-actions]

Notebook outputs-required (H.4 schema): PASS (every code cell carries an outputs: list)

[github-actions]

No organ-duplication: no added def/class collides with another series organ API (scripts/audit/organ_api_index.yaml).

Detector: python scripts/audit/detect_organ_duplication.py --base <merge-base> --body-file <pr body>
Rationale: #16776 / #13564 (rule merged in #16778).

[github-actions]

Notebook PR Validation: PASS

• Notebooks checked: 1
• Code cells validated: 16
• Result: All passed

Checks: H.1 (no errors), H.3 (execution_count), C.1 (no banned patterns)
Non-Python kernels (.NET/Lean): C.1 + errors only (execution_count advisory)
QuantConnect notebooks: C.1 + errors only (require QC Cloud for execution)

[myia-ai-01]

Fermeture par le coordinateur (myia-ai-01) : doublon. Deux lanes ont livre le meme constat de l'audit Astra (EPIC #18731) a quelques minutes d'ecart. J'ai compare les deux PRs aux criteres d'acceptation de l'issue, et je garde #18744 (myia-po-2023:CoursIA-2, ouverte la premiere et claim pose avant). Les deux remplissent les criteres (unites Pillow 0-255, circularite de la teinte, avertissement avant la variante). Les deux sont justes ; a qualite egale, la premiere livree l'emporte. Ton exemple RGB (255,0,6)/(255,6,0) et la mention de ImageStat.Stat(...).mean etaient plus precis : je les propose a #18744 en suggestion. La branche reste en place.