The Setting
The question concerns functionals along an infinite branch of a finite level tree: can they detect a common vector while tending pointwise to zero on the closed span of differences of a dense sequence?
Let be a real Banach space. For a subset of and a selfmap , a subset of is invariant for when .✦ Minimal invariant set data consist of a nonempty, norm-closed, bounded, convex, norm-separable, weakly compact subset of and a fixed-point-free nonexpansive selfmap of , such that every nonempty norm-closed convex invariant subset of equals .✦ All norms are the original ambient norm.
The closed difference span of and its whole closed unit ball , viewed as a subset of , are (1) Weak compactness of a subset of means compactness of its image in the weakly topologized ambient space.✦ Write for the copy of with the induced weak topology.✦
Normalized minimal invariant data are minimal invariant set data with , a norm-dense sequence in , metrizable , and weakly compact from (1).✦ Write for the class of these data.
Trees and Switching Data
An ordered rooted tree has finite levels with and , and nondecreasing surjective parent maps .✦ Its root is .✦ For , the ancestor map is the composition of the intervening parent maps, with the identity.✦ A coherent branch is a choice such that whenever .✦ The positive level nodes form the set .✦
For functions on a set , a prefix detector for is a point satisfying (2)✦ Finite prefix witnesses in a subset of mean that, at every positive depth, some node has a detector in as in (2).
A weighted ordered tree is an ordered rooted tree with a positive weight for each child , satisfying for every parent .✦ A tree of nonexpansive maps on a convex set , indexed by an ordered rooted tree , is a family consisting of a nonexpansive selfmap for each node.✦ For an assignment , the weighted child average is the map from to given by .✦ For an assignment from to , the cutoff evaluation is the point of obtained at the root by repeated weighted child averaging and application of the node maps strictly above the cutoff.✦ At depth zero, .
For an anchor and , the anchored averaging map is the selfmap of given by .✦ For an anchor and , the weighted resolvent is the unique fixed point in of the anchored averaging map .✦ For positive integers , the discrete resolvent is the selfmap of defined by .✦ The terminal resolvent sequence at an anchor is the sequence in defined by for .✦ The generated resolvent family is the least family of selfmaps of containing the identity and all , and closed under composition and convex mixing.✦ For selfmaps of , a convex mixture is the selfmap of given by this linear combination, where and .✦
Candidate ordinals start at zero and count discarded candidates. For an ordinal , its candidate budget is below. An injective labeling assigns retained budgets to positive level nodes. The cumulative budget minimum and dyadic tolerance are defined by (3) For and , the ideal switched cutoff is an assignment on . The ancestor displacement sum at uses the same point at every proper ancestor: (4)
Switching tree data for consist of a weighted ordered tree with such an evaluation, an anchor , the labeling , and functionals of norm at most one at the positive level nodes. The maps are generated, and is the identity. There are prediction points and an ultrafilter on refining the cofinite filter. With the budgets from (3) and cutoffs from (4), their conditions are (5) These conditions hold at every parent and child ; the weak convergence in (5) concerns the evaluated root outputs in the weakly topologized ambient space. Write for the class of switching tree data for .
The Claim
Theorem 1. Let be an ordered rooted tree, let be a nonempty subset of a set , and let . Suppose there are finite prefix witnesses in . There are a coherent branch and points such that (6)✦
Let be a real normed space, weakly compact, , and . Suppose every finite initial family of halfspaces has a common point in . There is such that (7)✦
Let have norm at most one. For and , assume and . Then (8)✦
Let be a summable real family and injective. Then (9)✦
For any subset with a norm-dense sequence in , (10)✦
Let be a real normed space, , and with . Assume for every . For every , (11)✦
Let and be an ordered rooted tree. At every node let , and . At positive levels assume and for every . Suppose the real family is summable. There is a positive depth such that, for each and , some satisfies (12)✦
For every , (13)✦
For every real Banach space , (14)✦
What Is Checked
The branch assertion (6) is stated over arbitrary real-valued detector functions on finite ordered levels, whereas (7) uses continuous linear functionals on a weakly compact subset of a real normed space. The detector estimate (8) uses the ambient norm, and (9) concerns a summable real family indexed by an arbitrary set. Assertions (10) and (11) use norm-closed real linear spans. Assertions (12), (13) and (14) are stated over normalized minimal invariant data, including weak compactness of the whole ball .
For , the anchored averaging map is a strict contraction with contraction bound in the ambient norm.✦ At every positive integer , the resolvent weight satisfies .