The Setting
Continuous linear functionals attached to a tree impose simultaneous conditions on a single vector along a branch. Leaf switching supplies the finite differences to which these functionals are applied.
Let be a real Banach space, let denote its weakly topologized copy, and let be the canonical identification. All norm and nonexpansiveness conditions use the original ambient norm and its induced dual norm. For , its weak carrier has the induced topology; weak compactness of means compactness of in .
Minimal invariant data consist of a nonempty, norm-closed, bounded, convex, norm-separable, weakly compact subset and a fixed-point-free nonexpansive map , with minimal among its nonempty norm-closed convex invariant subsets.✦ A subset is invariant relative to when .✦ The closed difference span and its whole closed unit ball , viewed as subsets of , are given by (1), with norm closure and the original ambient norm: (1) Normalized minimal invariant data consist of minimal invariant data with , a sequence , written , with norm-dense range in , metrizability of , and weak compactness of .✦
For an anchor and , the anchored averaging map is defined in (2). The weighted resolvent is the unique fixed point of the anchored averaging map for an anchor and in the parameter range of (2). The discrete resolvent and terminal resolvent sequence , for , have the indices in (2): (2) The generated maps form the least family of selfmaps of containing the identity and the , and closed under composition and binary convex combinations.✦ For selfmaps , their convex combination with coefficients and is the map .✦
Tree Evaluations
An ordered rooted tree is a sequence of finite levels with and , together with nondecreasing surjective parent maps .✦ Its root is .✦ The ancestor of , for , is obtained by successive parent maps, with .✦ Positive-level nodes retain their depth; for a natural number , the finite nonroot truncation is .
A weighted ordered tree is an ordered rooted tree equipped with a weight for each child , with the weights over the children of every parent summing to one.✦ The root path weight of a node of a weighted ordered tree is the product of the edge weights from to , with .✦ A tree evaluation on a weighted ordered tree with values in a convex set is a family of nonexpansive selfmaps , one for each node.✦ For and , the weighted child average of at is the vector in given by (3): (3)✦ The cutoff evaluation of repeatedly takes these averages and applies the node maps strictly above level ; it applies no map at the cutoff itself, and .✦
For and , the terminal descendants, their first and last members, and their inclusive lower and exclusive upper endpoints are defined by (4): (4) Given and , the switched leaf assignment is a map from to that retains below the integer cutoff and uses elsewhere.✦ For switched leaf assignments on , the evaluated switching root is a point of , the normalized leaf switch vector for a leaf is a vector in , and the roots and are the points of before and after switching the descendants of , all defined in (5): (5) For and , the ideal switched cutoff is the assignment in (6): (6)✦ The ancestor displacement sum , for and , is the sum of the displacements of under the proper ancestor maps, as in (7): (7)✦
Switching Tree Data
A candidate ordinal is zero-based and counts discarded candidates as well as retained ones. For a candidate ordinal and an injective ordinal assignment on the positive-level nodes, the dyadic budget , the retained budgets , the cumulative budget minimum , and the dyadic tolerance are the positive real numbers defined in (8): (8)
Switching tree data over normalized minimal invariant data consist of a weighted ordered tree with an evaluation on by generated maps, with the identity, an anchor , and an injective ordinal assignment . They include functionals at positive-level nodes, an ultrafilter on extending the cofinite filter, and prediction limits , subject to (9). Here , with , and the convergence is in after evaluation of the root outputs: (9) The zero root detector extension is the family of functionals on all nodes with and the given functionals at positive-level nodes. For a family on a common set and , simultaneous path detection of is the relation defined in (10): (10)✦
The Claim
Theorem 1. Let be switching tree data over normalized minimal invariant data . For every , (11)✦
For a general weighted tree evaluation on a convex subset , let be a real linear subspace containing , and suppose for all . Fix , and , continuous linear functionals with at every node, and real numbers for . Assume the block detection bounds (12): (12) Then a normalized leaf switch vector satisfies (13)✦
In the paper, the result reads: for the tree of Proposition 4.1 and every , there are a leaf and such that for every nonroot ancestor of , including .
What Is Checked
Assertion (11) is stated over normalized minimal invariant data and switching tree data; (13) is stated over weighted tree evaluations with the subspace and diameter assumptions above and the bounds (12).
The descendant sets in (4) are nonempty at every later level.✦
For , the anchored averaging map in (2) is a strict contraction with factor ✦, and the reciprocal weights satisfy for every positive integer .