The magnitude of metric spaces (Q374018)

The author introduces a notion of size for metric spaces; here \(\infty\) is permitted as distance in a metric space. The author shows that metric spaces have magnitude. For a finite metric space the author introduces the following definitions: The similarity matrix \(\zeta_A \in \mathbb R^{A\times A} \) is defined by \(\zeta_A(a,b)=e^{-d(a,b)} \, (a,b \in A)\). A weighting on \(A\) is a function \(w:A\rightarrow \mathbb R\) such that \(\sum_b \zeta_A(a,b) w(b) =1\) for all \(a\in A\). The space \(A\) has magnitude if it admits at least one weighting; its magnitude is then \(| A| =\sum_a w(a)\) for any weighting \(w\), and is independent of the weighting chosen. A metric space is positive definite if every finite subspace is positive definite. The magnitude of a compact positive definite space \(A\) is \(| A| =\sup\{| B| :B\) is a finite subspace of \(A \} \in [0,\infty]\). The article has a wider scope, the author defines the magnitude of a matrix, gives some background on enriched categories and introduces the concept of magnitude for enriched categories.