Documentation

LeanPool.Superdiffusion.Algsuperdiff.Section4.Provider.ExcessDecay.BoundaryCoveringSlot

BoundaryCoveringSlot #

Ported from scottnarmstrong/Superdiffusion at 7f20476385d4af418f110303085ad0740c9eba46. Module imports, support namespace and local option directives adapted for Lean Pool.

The boundary lane's covering cube, read at the slot n+3 #

CoveringSlotObstruction measured exactly this: on the boundary branch the covering cube c + □_{n+2}, c = wellPlacedCentre x m (n+2), leaves the n+2 good-event slot's cube, and always sits inside the n+3 one. The frozen general clause puts its slot at n+3 precisely so that this positive half is usable. This module cashes it in, at the level the boundary Caccioppoli step consumes.

Three steps, each proved elsewhere and composed here:

  1. the half-open containment — CoveringSlotObstruction's open-cube containment c + □_{n+2} ⊆ z + □_{n+3} re-derived in the half-open realization the off-grid transports require (the margin is ample: the centre moves by less than 6·3^n and the half-open child reaches 4.5·3^n, against the parent's 13.5·3^n);
  2. the off-grid transport at depth 1 — OffGridErrorFluxCorrected's root transport at K = □_{n+3}, P = □_{n+2}, so the depth factor is a single 3^{u·1}, i.e. 3^{s/8} ≤ 3^{1/8} < 1.15;
  3. the good event at the n+3 slot — ReindexSlot's (n+3) caps, read at L ≥ n+3 (the frozen binder n + 3 ≤ m ≤ L).

No antisymmetric-shift conversion is used: nothing below converts Λ or λ by the antisymmetric-shift invariance. The whole chain runs at the (n+3) family parentRebasedFamily M L (n+3) c z ω — the covering cube's own frame at the parent's constant — which is the family InteriorRebase builds and the one the equation transports to for free.

The q = 1 ingredients are supplied here, at the covering cube: CaccioppoliInteriorPrefactor's constant-1 comparisons Λ_{u,1} ≤ Λ_{u/2,2} and λ_{u,1}^{-1} ≤ λ_{u/2,2}^{-1} turn the q = 2 ratio cap into the q = 1 ingredients Λ_{u,1}, λ_{u,1}^{-1} and Θ that caccioppoliWithRHSPrefactor reads.

What this does not do #

It does not prove the boundary clause. The boundary lane still needs its analytic half (the trace-measure Poincaré and the Stampacchia step), the boundary energy assembly at this family, and the L²/window transport onto the general clause's window. Nothing below is an instance, or a fraction, of any source node.

References #

1. The half-open containment of the covering cube in the n+3 parent #

2. The off-grid transport at depth one #

The covering cube's coarse-graining error against the general clause's representative.

At every index t strictly above the good-event index s/8 and at most 1/2, the covering cube's q = 2 error at the (n+3) re-based family is bounded by the frozen clause's own fluxCorrectedErrorRepresentative M L (n+3) ⟨s/8⟩, with the depth-one factor 3^{s/8} (at most 3^{1/8} < 1.15).

3. The covering cube's ellipticity data, capped on the good event #