Adaptive meshes for strongly-positive metric controls #
Moise Chapter 6 first subdivides the source until every simplex is small compared with the prescribed strongly-positive tolerance. This file supplies that step for the locally finite adaptive triangulation. Strong positivity gives a uniform lower bound on a compact neighborhood of each point; continuity then shrinks that neighborhood until its image has small diameter. Subordination to the resulting open cover converts the existing setwise polygonal graph replacement into a pointwise controlled approximation.
A neighborhood on which phi has a fixed positive lower bound and the image of f has
diameter at most one quarter of that bound.
Instances For
Strong positivity and local compactness produce controlled neighborhoods for every point.
A chosen controlled neighborhood at every point.
Equations
Instances For
The part of a mesh tolerance reserved for staying inside an open target region. When the
region is all of the plane, the constant bound 1 is used instead.
Equations
- LeanEval.Topology.ClassificationOfSurfaces.Moise.regionMeshControl V f x = if hcomp : Vᶜ.Nonempty then min 1 (LeanEval.Topology.ClassificationOfSurfaces.Moise.frontierDistance V (f x) / 4) else 1
Instances For
The target-region mesh control is strongly positive along any continuous map landing in that region.
A minimum of two strongly-positive controls is strongly positive.
The ambient open cover obtained from controlled neighborhoods in the open subspace U.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Every selected cover set retains its quantitative scale and image-diameter bounds.
The adaptive locally finite triangle complex subordinate to the quantitative cover.
Equations
- One or more equations did not get rendered due to their size.
Instances For
The controlled adaptive complex with its tolerance automatically reduced near the frontier of an open target region.
Equations
- K.regionControlledAdaptiveComplex U hU V hV f hf hmem phi hphi = K.controlledAdaptiveComplex U hU f hf (LeanEval.Topology.ClassificationOfSurfaces.Moise.regionSafeControl V f phi) ⋯
Instances For
Each adaptive face inherits one quantitative scale from the cover member containing it.
Subordination makes every edge image small enough for the canonical simultaneous polygonal replacement to satisfy the original strongly-positive pointwise control.
The canonical replacement graph on the controlled adaptive mesh is a pointwise
phi-approximation of the original realization.
Every edge belonging to a controlled adaptive face has image diameter bounded by that face's selected scale.
The boundary error on one adaptive face is bounded by a positive scale which is at most both the prescribed pointwise tolerance and one.
The adaptive mesh discharges the exact face-boundary estimate consumed by the cellwise Schoenflies extension.
If the original controlled realization lies in the model half-plane, so does its entire simultaneous polygonal replacement graph.
The adaptive face-boundary estimate for the frontier-reduced tolerance.
The tolerance consumed by the filling is uniformly bounded and frontier-relative.
The frontier-reduced adaptive fillings lie in the open target region.
The frontier-reduced adaptive fillings form a locally finite family in the open target.
The region-controlled adaptive complex carries distinct vertex triples on distinct faces:
the hfaces entry condition of exists_polygonalReplacement holds unconditionally.
Complete locally finite cellwise replacement once the remaining side-separation condition is supplied. Region containment and local finiteness are now consequences, not hypotheses.
Complete locally finite cellwise replacement with no residual side hypothesis. The
separation entry of exists_polygonalReplacement is discharged by the facewise comparison
map after shrinking the arc approximation controls; the shrunken realization keeps the map,
the region, and the frontier-reduced tolerance, so all adaptive control theorems reapply.