Affine invariant points (Q891077)

Let \(n \in \mathbb N\). A subset \(K\) of \(\mathbb R^n\) is called a convex body if \(K\) is a compact convex set with nonempty interior. The space of convex bodies in \(\mathbb R^n\) is denoted by \(\mathcal K_n\). The space \(\mathcal K_n\) is endowed with the Hausdorff metric. A map \(p :\mathcal K_n \to\mathbb R^n\) is called an affine invariant point if \(p\) is continuous and \(p(T(K)) = T(p(K))\) holds for every affine map on \(\mathbb R^n\) and every \(K \in \mathcal K_n\). It is shown that, for every \(n \geq 2\), the set of all affine invariant points is an infinite-dimensional affine space (Theorem 1). For \(K \in \mathcal K_n\), one can consider the set \(\mathfrak P_n(K)\) of all \(p(K)\) such that \(p : \mathcal K_n \to \mathbb R^n\) is an affine invariant point. If \(P_n(K)\) is \((n-1)\)-dimensional, then it can be described as the set of all \(x \in \mathbb R^n\) satisfying \(T x = x\) for all affine transformations \(T\) that map \(K\) onto \(K\) (Theorem 2). The set of all \(K \in \mathcal K_n\), for which \(\mathfrak P_n(K)=\mathbb R^n\) is fulfilled, is open and dense in \(\mathcal K_n\) (Theorem 3).