Remark 4 [03QK]
Original official author HTML, exact retained edition. Historical TeX conversion verdicts remain unchanged. Cited-edition alignment and mathematical self-containment are not assessed.
Complete original source context · Original author HTML
Remark 4
1) Since the matrix defined by the metric is positive, the function is convex. In particular, there is locally well-defined Legendre transform of . This fact will be used later, when we will discuss the duality of Monge-Ampère manifolds.
2) It seems plausible that in the case when all are simply-connected, and for , the metric space is a homological sphere of dimension . In all examples it is in fact homeomorphic to .