weather-p1: CT-F16 kills the steering rescue — and the 0.684 plateau was the instrument, not the data - #929
Merged
Conversation
…el scoring) Committing the probe with its four bars stated, before any fetch, so the read cannot be tuned to the result. CT-F16a >= 0.70 sign consistency vs the 500/600/700 hPa steering flow; CT-F16b the PAIRED test (sd of the signed offset must drop >= 10%); CT-F16c anti-vacuity (permuted + 90-deg-rotated references must BOTH stay under 0.70, reported first and voiding the rest if they fail); CT-F16d the level sweep, descriptive. Scope stated up front: this is a RE-SCORING of CT-F14's own 19 storms with ONLY the motion reference changed -- the MECHANISTIC test, explicitly NOT a verdict. A fresh-sample verdict is CT-F17, named and not run. Co-Authored-By: Claude <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CcpLeEC3XK8Eye53GKBVvi
… control scored the headline Bars were committed before the run (05f0900). Both primary bars FAIL, in the direction that cuts against this arc's own hypothesis. CT-F16a sign consistency vs 500/600/700 hPa steering flow: 0.579 (11/19), p=0.324, against a 0.70 bar -- WORSE than the surface displacement's 0.684 on the identical storms. CT-F16b paired residual: sd 68.29 -> 87.71 deg, 28.4% WIDER, where a >=10% tightening was predicted. Opposite direction. CT-F16d level sweep improves MONOTONICALLY toward the surface (400 hPa 0.579 -> 850 hPa 0.684; sd 89.5 -> 77.0), best at 850 hPa, OUTSIDE the 400-650 hPa band the height ladder predicted. No mid-tropospheric optimum exists on this sample. Report SS9.2 called steering level "the single most promising fix". It is now measured and it is not a fix; that row is superseded in place. THE CONTROL IS THE LARGER FINDING. F16c scored two deliberately WRONG references through the same pipeline. The 90-deg-rotated steering reference returned 13/19 = 0.684, p=0.0835 -- numerically identical to CT-F14's headline, the figure this arc has carried as "suggestive". At n=19 the ladder is 11->0.324, 13->0.0835, 14->0.0318. CT-F14 was never one storm short of significance; it was one storm short of distinguishability from an answer built to be wrong. Rule banked: an anti-vacuity control measures the RESOLVING POWER of the instrument, not only the test it is attached to. SECOND, INDEPENDENT FINDING: the sign test conflates a systematic rotation with a correct prediction. Weak steering (<10 m/s, n=6): sign 0.833, median |err| 103 deg. Strong steering (>=10 m/s, n=13): sign 0.462, median |err| 55 deg. corr(speed,|err|) = -0.407. Magnitude accuracy improves with steering strength as physics expects while sign consistency moves the opposite way -- because a one-sided sign test on a distribution not centred at zero reports which SIDE the bias falls on. This arc has used it as the primary instrument since SS4. NOT falsified: the height ladder decomposed the FIELD per level about its own centre; CT-F16 keeps the SURFACE dipole and swaps the FLOW reference. Different quantities. The ladder stands; its operational reading is what died. The structural claim (SS9.1) is untouched. Co-Authored-By: Claude <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CcpLeEC3XK8Eye53GKBVvi
…esultant resolves the same rows at p=0.005
Operator: "die irrationale Aufsummierung hilft, dass der Dipol nicht auf 0.68
kollabiert." Measured on the SAME 19 stored storms, no fetch, post-hoc and
explicitly NOT a verdict:
referent sign<0 R_bar mu Rayleigh p
surface (F14) 0.684 0.516 -30.2deg 0.0050
steering (F16) 0.579 0.343 -40.5deg 0.107
CONTROL rot+90 0.684 0.343 -130.5deg 0.107
CONTROL permuted 0.421 0.142 -- 0.689
Three findings:
1. The 0.684 plateau was a property of the STATISTIC, not the data. The sign
test collapses each error vector to one bit; 19 bits saturate below the
14/19 floor. The vector Aufsummierung (circular resultant) resolves the
identical rows at p=0.0050 -- concentration and offset become two numbers
instead of eating each other. The arc's ~-40deg systematic offset is now
ESTIMATED (-30.2 +/- 36.5 deg) instead of penalized.
2. The wrong referent is VISIBLE. The rotated control that scored an
indistinguishable 0.684 under the sign test shows the same R_bar with mu
shifted 100.3 deg -- separated far beyond both CIs. Instrument hierarchy:
real (0.516) > structured-but-wrong (0.343, wrong mu) > permuted (0.142,
below the uniform floor 0.203).
3. Every prior sign-consistency number in the arc (2/2, 6/10, 8/10, 13/19)
was read through an instrument that cannot estimate the offset it
penalizes and cannot distinguish a rotated referent at these n. Bounded in
BOTH directions: the sign test neither established the claim nor could it
have.
Faltung reading folded into SS5.13: the resultant is the first circular
Fourier coefficient (Faltung with e^{i theta}); the W6 two-component fit is a
DEconvolution (component mix conv apparatus noise, +/-3-7 deg from CT-F4);
on Z_256 the circular Faltung is substrate-native (DistanceLut::circular's
domain, FFT-able).
Not promoted: same sample, post-hoc. CT-W6 = the pre-registered circular-
statistics use; CT-F17 = the fresh-sample verdict.
Co-Authored-By: Claude <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CcpLeEC3XK8Eye53GKBVvi
|
Warning Review limit reached
Next review available in: 58 minutes You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository. How can I continue?After more reviews become available, a review can be triggered using the To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews. How do review limits work?CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability. For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window. Please refer docs for additional details. Review details⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (6)
ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (6)
📝 WalkthroughWalkthroughCT-F16 adds a paired steering-flow re-scoring probe for 19 CT-F14 storms, stores pressure-level results, and adds circular-resultant analysis with rotated and permuted controls. Reports document failed steering-level criteria and preserve unaffected height-ladder findings. ChangesCT-F16 analysis
Estimated code review effort: 4 (Complex) | ~45 minutes Sequence Diagram(s)sequenceDiagram
participant Analyst
participant Probe as comet_tail_f16.py
participant ERA5 as ERA5 WeatherBench2
participant Artifact as comet_tail_f16.json
participant Report as COMET_TAIL_REPORT.md
Analyst->>Probe: run paired CT-F16 rescoring
Probe->>ERA5: retrieve storm pressure and wind fields
ERA5-->>Probe: return meteorological measurements
Probe->>Artifact: write errors, level sweep, and verdicts
Analyst->>Report: record steering and circular-result findings
Possibly related PRs
Poem
✨ Finishing Touches📝 Generate docstrings
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Bugbot couldn't run - usage limit reachedBugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit. A user or team admin can review and increase usage limits in the Cursor dashboard. (requestId: serverGenReqId_c3085e81-95fe-4383-85b4-5f4e4a338338) |
Owner
Author
|
@coderabbitai review Requesting a review while this stays a draft. This PR carries statistical machinery and pre-registered verdicts — the material where review has been most productive on this arc (the Specifically worth adversarial eyes:
Generated by Claude Code |
|
|
AdaWorldAPI
marked this pull request as ready for review
August 12, 2026 10:48
AdaWorldAPI
pushed a commit
that referenced
this pull request
Aug 12, 2026
Post-merge board hygiene per the Mandatory Board-Hygiene Rule. #929 was a mixed PR (2 probes + report sections + epiphany), so its entry records the non-hygiene half: * CT-F16: the steering rescue failed both pre-registered bars; the level sweep is monotone toward the SURFACE (best 850 hPa). The ladder survives as a measurement; its operational reading is dead. * F16c's rotated control scored 0.684 = CT-F14's headline -- the sign test could not distinguish the real referent from a wrong one at n=19. Rule banked: a control's score is the floor any rate-headline must clear. * The circular resultant resolves the same 19 rows at p=0.0050, estimates the offset (-30.2 deg) instead of penalizing it, and separates the rotated control by 100.3 deg. Post-hoc, instrument demo, not a promotion -- but sign-consistency numbers are retired as verdict-grade instruments across the arc. All 13 figures in the entry verified against the committed JSONs BEFORE landing (13/13 exact -- the #927 self-verification, now standard practice). Append-only audit: zero removed lines across both board files (pure prepend). THIS PR IS PURE HYGIENE -- no type, plan, deliverable, epiphany or code. Per the termination clause it generates no further obligations; the chain stops here. Co-Authored-By: Claude <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CcpLeEC3XK8Eye53GKBVvi
AdaWorldAPI
pushed a commit
that referenced
this pull request
Aug 12, 2026
…h, fixed in all three homes Codex P2 on #930: "identical R-bar with mu shifted 100.3 deg" fused two different comparisons. Identical R-bar belongs to steering<->rotated (0.343 = 0.343, mu shifted exactly 90 -- rotation preserves concentration by construction); the 100.3 deg separation belongs to surface<->rotated, where R-bar is NOT identical (0.516 vs 0.343). The composite was false in all three homes: LATEST_STATE, the #929 arc entry, report SS5.13. Stated correctly the finding is STRONGER -- the control separates from surface in BOTH channels. Unmerged board entries composed in place (append-only rule's unmerged-PR allowance); the merged report gets a dated correction note. Lesson banked: the 13/13 figure verification checked every NUMBER and still missed this -- a relation between two individually correct numbers can be false. Verify comparative claims AS claims: every "identical/same/larger" must name both operands, and the check must evaluate the relation. Also codex P2 #2: the LATEST_STATE shipped-PR table had stalled at #780. Added #926-#929 rows plus an explicit gap-note row for #781-#925 (carried by PR_ARC_INVENTORY) -- honest gap, not silent reconstruction of ~150 rows. Co-Authored-By: Claude <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CcpLeEC3XK8Eye53GKBVvi
AdaWorldAPI
pushed a commit
that referenced
this pull request
Aug 12, 2026
…ng, and the append-only audit could not fail Operator: "930 has comments / check also previous 5 / if you want go back another 5." All 68 review comments on #920-#930 enumerated and checked against the TREE, not the merge. Depth is non-uniform and said so: #922-#930 finding-by-finding; #920 (27) and #921 (5) spot-checked on P1/governance only (both clean), ~30 older findings there left explicitly UNVERIFIED. Clean: #922/#924/#925/#929 zero comments; all 27 of #926's findings fixed in the tree (equal-budget grid_pts + regenerated E2 JSON, seam-wrapping subgrid_min, find_center->None, F7d 35->40, CT_F12 NO-VERDICT, persisted storm metadata, np.roll longitude, __file__-relative write, net-decay E6, the 93-97%->90.9-94.3% headline); #923's plan-status P2 resolved. Three open, all frozen in append-only ledgers, all corrected in NEW entries: 1. "+92.76 Pa moves R2 in the 5th decimal" is refuted by the report's own carve table 15 lines above it: carve A's +92.76 Pa moved R2 0.9212->0.9129 (0.0083, THIRD decimal); carve D's +1.59 Pa moved it 2.4e-06 (SIXTH). The var() BUG was blind at every magnitude; the STATISTIC is near-blind only in the single-digit-Pa regime -- which is exactly where "lossless" was claimed. 2. "10 probe scripts with committed JSON" undercounts: 15 .py added, 11 with a committed .json. 3. "+13/-0, +10/-0, +0/-0": measured +13/-0, +17/-0, and EPIPHANIES absent from the net diff -- it was +1/-1 at 0f9e6bc (the in-place edit), zeroed by the revert. The +0/-0 was a revert artefact, not evidence of purity. And the audit METHOD is replaced: zero-deletions proves ADDITIVE, not PREPEND. The suffix check (new.endswith(old)) proves it. Both halves measured before banking: a constructed end-append scores zero-del=True/suffix=False; a true prepend True/True. It also fires on THIS PR (LATEST_STATE suffix=False at +15/-0) -- correctly, since #930 inserts table rows mid-file and composes an unmerged entry in place. A False is a demand for justification, not a verdict. Same defect shape as #930's relation error, twice in three days: two individually-true numbers asserted of one pair when each belongs to a different one. A figure-by-figure check cannot catch it; the relation must be evaluated as a claim. Already standing in report SS10.1 and the worker preamble of weather-w-probes-v1 (both verified present, not assumed). Co-Authored-By: Claude <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CcpLeEC3XK8Eye53GKBVvi
AdaWorldAPI
pushed a commit
that referenced
this pull request
Aug 12, 2026
#930 merged as MIXED (started as #929 hygiene, grew into the #920-#930 open-review sweep + report SS10 + weather-w-probes-v1 worker briefs, per its own twice-corrected title/body). Arc entry + LATEST_STATE shipped-PR row added. The #928/#929 shipped-table rows are left untouched (superseded by the sweep entry's correction table, not edited in place -- append-only). Suffix-checked before commit: PR_ARC_INVENTORY suffix=True (pure append); LATEST_STATE suffix=False (a table-row mid-file insert, the justified case the sweep entry itself named). Co-Authored-By: Claude <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CcpLeEC3XK8Eye53GKBVvi
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Two probes, both pre-registered before their runs, both landing against the arc's own hypotheses
CT-F16 — the leading moderator, measured and failed (
05f09005→5f066f67)§9.2 called the steering level "the single most promising fix" for the directional claim. Paired re-scoring of CT-F14's own 19 storms with only the motion reference changed (surface 6h displacement → 500/600/700 hPa disk-mean flow; same centres, byte-identical
err_deg):The anti-vacuity control is the larger finding: a deliberately 90°-rotated reference scored 13/19 = 0.684, p=0.0835 — numerically identical to CT-F14's headline. At n=19 the sign-test ladder is 11→0.324, 13→0.0835, 14→0.0318: CT-F14 was never one storm short of significance; it was one storm short of distinguishability from an answer built to be wrong. Second, independent finding: stratified by steering strength, magnitude accuracy improves (median |err| 103°→55°) while sign consistency drops (0.833→0.462) — a one-sided sign test on a distribution not centred at zero reports which side the bias falls on, not whether the prediction holds.
Not falsified: the height ladder (different quantity — field-per-level vs surface dipole) and the entire structural claim (§9.1).
§5.13 — the circular resultant resolves the same rows the sign test couldn't (
bd2f93d6)Operator: "die irrationale Aufsummierung hilft, dass der Dipol nicht auf 0.68 kollabiert." Measured on the same 19 stored storms, no fetch, post-hoc and explicitly not a verdict:
The Faltung reading is folded into §5.13: the resultant is the first circular Fourier coefficient; the planned CT-W6 two-component fit (geo/neighbor + bow-wave) is a deconvolution; on
Z_256circular Faltung is substrate-native (DistanceLut::circular()'s own domain).Contents
comet_tail_f16.py/.json— pre-registration committed before the runcomet_tail_resultant_instrument.py/.json— the instrument comparison, deterministic bootstrapCOMET_TAIL_REPORT.md§5.12 + §5.13, §9.2 steering row superseded in placeEPIPHANIES.md—E-THE-CONTROL-SCORED-THE-HEADLINE-1Queued, named, not run: CT-W6 (two-component vector deconvolution of the dipole: neighbor far-field + bow-wave, global coefficients across 19×2 observations, circular metrics, F16c-standard controls), CT-W2s/W5/W7 (sunflower collision nodes, spiral-ADI, Gegendruck), CT-F17 (fresh-sample verdict).
🤖 Generated with Claude Code
https://claude.ai/code/session_01CcpLeEC3XK8Eye53GKBVvi
Generated by Claude Code
Summary by CodeRabbit
Documentation
Data & Analysis