Coarse-graining support: Support.Book.Ch05.Theorems.Section57.FiniteBasis #
Imported from the Apache-2.0 CoarseGraining development at commit
c7ddd76c08ade64fed1b8d2ca51be14dfee8deb4.
Finite-basis reduction for the quenched unit-vector maximum #
This file records the deterministic finite-dimensional reduction used in
Theorem t.homogenization.quenched: because the normalized block response is a
nonnegative quadratic form of the full-block vector, the maximum over unit
vectors is controlled by finitely many coordinate and pair probes.
The full-block matrix whose quadratic form is
J(Q,\overline A^{-1/2}e,\overline A^{1/2}e).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Finite coordinate and pair probes for the limiting-normalized J
quadratic on one cube.
Equations
- One or more equations did not get rendered due to their size.
Instances For
The normalized finite probe sum: coordinate probes are unchanged, while
plus/minus pair probes are scaled by 1/2 so their Euclidean square norm is at
most one.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Pointwise finite-basis control for a sampled coefficient field carrying an a.e.-ellipticity witness. This is the simultaneous version needed when the unit-vector supremum is packaged into a Chapter 2 scale response.