The Banach space of quasinorms on a finite-dimensional space (Q2238556): Difference between revisions

This paper is a clever reflection about the meaning of ``interpolation'' (real, complex, whatever) in Banach spaces. The original approach to the topic is already present in the title: The space of quasinorms defined on \(\mathbb{K}^n\) is a Banach space. Surely the reader would expect something like: the space of norms defined on \(\mathbb{K}^n\) is a Banach space, but that is not how things are. This is one of the deep points in the paper: the realization that the study of Banach spaces requires the study of quasinorms. The first part of the paper contains the proof of the following nice result: the space of quasinorms (up to dilations) on \(\mathbb{K}^n\) can be endowed with a natural norm and the resulting Banach space is isometric to the space of continuous functions (up to constant functions) on the corresponding projective space. The second part of the paper contains a bunch of interesting digressions connecting their approach with the Banach-Mazur compactum, namely, the compact space of norms on \(\mathbb{K}^n\) when this space is endowed with the Banach-Mazur distance.