The Finite Dimensional Approximation Property and the AR-Property in Needle Point Spaces (Q4399232)

A subset \(X\) of a linear metric space \(Y\) is said to have the finite-dimensional approximation property (FDAP) if there exists an \(F\)-norm \(\|\cdot\|\) on \(Y\), a sequence \(\{r_n\}\) of maps from \(X\) into finite-dimensional subsets \(X_n\) of \(X\) and a summable sequence \(\{s_n\}\) of positive numbers such that \(\liminf_{n\to\infty} s^{-1}_n(1+ \dim X_n)\|x-r_n(x)\|= 0\) for every \(x\in X\). A non-zero point \(a\in Y\) is called a needle point if for every \(\varepsilon>0\) there exists a finite subset \(A(a,\varepsilon)\) of \(Y\setminus\{0\}\) such that \(\|b\|\leq\varepsilon\) for every \(b\in A(a,\varepsilon)\), for every \(b\in \text{conv}(A(a,\varepsilon)\cup \{0\})\) there exists an \(\alpha\in[0,1]\) such that \(\|b-\alpha a\|<\varepsilon\) and \(a\in \text{conv}(A(a,\varepsilon)\cup \{0\})\). The space \(Y\) is said to be a needle point space if \(Y\) is a separable complete linear metric space in which every non-zero point is a needle point. The author proves that every needle point space \(Y\) contains a dense linear subspace \(E\) with the following properties: (i) \(E\) contains a non-empty compact convex set with no extreme points; (ii) \(E\) has the FDAP implying that every non-empty convex subset of \(E\) is an AR.