Bourbaki

Proposition 7

Tag 07XU, permanent. Cite it as [Bourbaki, Tag 07XU].

Topological Vector Spaces, chapter V, § 1. Prehilbertian spaces and hilbertian spaces

Let E be a Hausdorff prehilbertian space and let Ψ\Psi be a non-empty directed increasing set of non-empty, convex, complete subsets of E. Put A=⋃H∈ΨHA = \bigcup_{H \in \Psi} H and suppose that the closure N of A is complete. Then N is convex and we have

pN(x)=lim⁡H∈ΨpH(x) for all x∈E. p_N(x) = \lim_{H \in \Psi} p_H(x) \text{ for all } x \in E.

It is clear that A is convex, hence its closure N is convex (II, p. 13). With the notations of prop. 6, d(x,N)=inf⁡H∈Ψd(x,H)d(x, N) = \inf_{H \in \Psi} d(x, H), and consequently d(x,N)d(x, N) is the limit of d(x,H)d(x, H) with respect to the section filter of Ψ\Psi. Since pH(x)∈Hp_H(x) \in H and

lim⁡H∈Ψ∥x−pH(x)∥=lim⁡H∈Ψd(x,H)=d(x,N), \lim_{H \in \Psi} \|x - p_H(x)\| = \lim_{H \in \Psi} d(x, H) = d(x, N),
it follows from cor. 1 of V, p. 11 that pH(x)p_H(x) tends to the projection pN(x)p_N(x) of xx onto N with respect to the section filter of Ψ\Psi.

Bourbaki puts the proof directly after the statement, so what is above is the statement and its proof together.

Cited by

Nothing.

Cites

Transcribed from the printed volume. Not translated.