H^k norm of an iterated family #
Evans, Partial Differential Equations (2nd ed.), §5.2.2 (p. 259) defines
‖u‖_{W^{k,p}(U)} = (Σ_{|α| ≤ k} ∫_U |D^α u|^p dx)^{1/p}
for multi-indices α. A family HasIteratedWeakDerivOn V k u indexes derivatives by lists of
directions, one list per ordered sequence of differentiations, so a multi-index of order j
corresponds to as many lists as it has orderings. iteratedNorm is the same sum taken over
lists:
iteratedNorm H = (Σ_{j ≤ k} Σ_{α ∈ (Fin d)^j} ‖D_α u‖²_{L²(V)})^{1/2}.
Weak derivatives commute, so each multi-index contributes its L² norm squared with the
multiplicity of its orderings, between 1 and j!; the two norms are equivalent with constants
depending on d and k alone. Weak derivatives on an open set are unique, so the value does
not depend on the family chosen.
Main declarations #
iteratedNorm: the norm.norm_D_le_iteratedNorm: every entry of the family is bounded by it.iteratedNorm_le: a uniform bound on the entries bounds it, with a constant indandk.
H^k(V) norm of an iterated family, summed over lists of directions of length at
most k.
Equations
Instances For
Every entry of the family up to order k is bounded by the norm.
The norm is a uniform bound on the family.
The number of lists of directions of length at most k, as a real number.
Equations
- EllipticPdes.Regularity.listCount d k = ∑ m ∈ Finset.range (k + 1), ↑d ^ m
Instances For
Norm bound from a uniform bound on the family, with the factor √N for N the number
of lists of length at most k.