#661: compatible g-lift vs exact-lab ΔZ after the #633/#653 repairs - #663
Draft
LightChainr wants to merge 1 commit into
Draft
#661: compatible g-lift vs exact-lab ΔZ after the #633/#653 repairs#663LightChainr wants to merge 1 commit into
LightChainr wants to merge 1 commit into
Conversation
…over of #633). Lift the frozen 9-decile consensus g (725 not in the freeze) into the affine chart of the exact-lab quantile via DZ_Q[g](u) = (g(u)-g(a))/W - (Q(u)-Q(a))(g(b)-g(a))/W^2 with shared anchors 0.2/0.8, and compare to the repaired-census site tangent DeltaZ = Z_{L=4} - Z_{L=3} (PR #653, bond dual_fail = 0) with unoriented |cosine|. Primary unoriented angle 24.87 deg (|cos| = 0.9073); Q = site L=4 gives 24.72 deg, repaired bond gives 25.81 deg. Chart-exact: the lift direction is invariant under Q -> cQ + d, so recovering Q on the deciles from the census Z is exact, not interpolation. Equal-weighting g is a labelled sensitivity (delta < 0.3 deg). No N725 rerun; N725 g-vs-DeltaZ remains undetermined (no second production size). Tiny-torus/exact-lab DeltaZ cannot prove an asymptotic mismatch in either direction. Brings in the #655 lineage script and n725-zflow-corrected.json unchanged from analysis/633-n725-spin0, and the repaired census-exact.json from repair/632-bond-lab-and-wrapping-l3l4.
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Closes #661 (leftover of #633; not a new production, not #622).
What this does
One comparison, one notes file: lifts the frozen 9-decile consensus g into the affine chart of the exact-lab quantile with the issue's formula
shared anchors 0.2 / 0.8 (both on g's grid; no interpolation of g), and compares unoriented (|cosine|) to ΔZ = Z_{L=4} − Z_{L=3} recomputed from the repaired census (PR #653, bond
dual_fail = 0; PR #628's broken bond Z is asserted against, not copied).Result (spin0 primary, Q = site L=3)
Chart-exactness: Q on the deciles is recovered from the census Z exactly up to affine, and
DZ_{cQ+d}[g] = DZ_Q[g]/c, so the angle does not depend on the recovery — no anchor is missing, nothing left undetermined on the lab side.Scope, stated not hidden
n725-zflow-corrected.jsonis g-lift vs Z_725 (same-size parallelism), not the g-vs-ΔZ test.Files
notes/issue-633-glift-exact-lab-20260908.md— deliverable noteresults/probe-invariant-shape/issue661-glift-exact-lab.json— angles/vectorsscripts/probe_invariant_shape/issue661_glift_exact_lab.py+tests/test_issue661_glift_exact_lab.py(6 tests, passing)results/probe-invariant-shape/census-exact.json— repaired Repair #632 bond laboratory; wrapping-type census L=3,4 (#640) #653 version overlaid (bond dual_fail 118133 → 0)results/probe-invariant-shape/n725-zflow-corrected.json,scripts/threshold_quantile_lineage.py— Restore spin-0 N725 Q/Z from the committed histogram (#633) #655 lineage, unchanged fromanalysis/633-n725-spin0Reproduce:
python3 scripts/probe_invariant_shape/issue661_glift_exact_lab.py python3 -m unittest discover -s tests -p "test_issue661*"