Spans and convex combinations of boundary-valued continuous functions

Abstract

For an $(n\ge 2)$-dimensional real Banach space $E$ with unit ball $E_{\le 1}$ and a topological space $X$ arbitrary elements in $C(X,E_{\le 1})$ are always expressible as linear combinations of at most three functions valued in the unit sphere $\partial E_{\le 1}$. On the other hand, for normal $X$, $C(X,E_{\le 1})$ can only be the convex hull of $C(X,\partial E_{\le 1})$ if the covering dimension of $X$ is strictly smaller than $\dim E$. A variant of this remark is the characterization of normal $X$ with $\dim X<\dim E$ as precisely those for which $C(X,E_{\le 1})$ is the convex hull of nowhere-vanishing continuous $X\to E_{\le 1}$ or, equivalently, that of continuous functions $X\to E_{[r,1]}$, $r\in (0,1)$ valued in arbitrarily thin spherical shells. This extends a number of results due to Peck, Cantwell, Bogachev, Mena-Jurado, Navarro-Pascual and Jim\'enez-Vargas and others revolving around the realizability of the unit ball of $C(X,E)$ as a convex hull of its extreme points for strictly convex and/or complex $E$.