Distance \(k\)-sectors exist (Q991184)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Distance \(k\)-sectors exist |
scientific article |
Statements
Distance \(k\)-sectors exist (English)
0 references
2 September 2010
0 references
Since the seminal paper of \textit{T. Asano, J. Matoušek} and \textit{T. Tokuyama} [Adv. Math. 212, No.~1, 338--360 (2003; Zbl 1185.68768)] a lot of recent work has been done. The paper under review is devoted to the notion of distance \(k\)-sectors: A distance \(k\)-sector of two disjoint, nonempty, closed sets \(P\) and \(Q\) in Euclidean spaces, where \(k \geqslant 2\) is an integer, is a \((k - 1)\)-tuple \((C_{1},C_{2},\dots ,C_{k - 1})\) such that \(C_i\) is the bisector of \(C_{i - 1}\) and \(C_{i+1}\) for every \(i=1,2,\dots ,k - 1\), where \(C_{0}=P\) and \(C_k=Q\). A new notion of \(k\)-gradation for \(P\) and \(Q\) is introduced and its existence (even in an arbitrary metric space) is obtained using the Knaster-Tarski fixed point theorem. The existence of \(k\)-sectors is achieved then as a consequence. The use of a fixed point theorem to prove the existence of the distance trisector curve between two points was already conjectured in the cited paper. A final part of the paper is devoted to discuss the computational methods that can be used to draw \(k\)-sectors (approximation by polygons, pixelization, etc.\dots). The analysis of time complexity is left as a future research problem. Uniqueness remains as an open problem. It is only conjectured in the Euclidean space. An example for the \(\ell_1\) metric in the plane is shown where different \(3\)-sectors are obtained for the same pair of sets \(P,Q\). The plane with the \(\ell_1\) metric is a geodesic metric space because for every two distinct points \(x,y\in {\mathbb R}^2\) there is a metric segment connecting them, i.e., an isometry mapping \(\gamma:[a,b]\to {\mathbb R}^2\) of an interval \([a,b]\subset{\mathbb R}^2\) with \(\gamma(a) = x\) and \(\gamma(y) = b\). Nevertheless there is a difference with other geodesic metric space. With the \(\ell_1\) metric, the set of points between \(x\) and \(y\), i.e., \(\{z\in{\mathbb R}^2| d_1(x,z) + d_1(z,y) = d_1(x,y)\), is not just the image of a metric segment, but, in general, a two dimensional subset. This behavior could be responsible of the non uniqueness of \(k\)-sectors for the \(\ell_1\) metric in the plane.
0 references
distance \(k\)-sectors
0 references
Knaster-Tarski fixed point theorem
0 references
metric space
0 references
distance trisector curve
0 references
Euclidean space
0 references
geodesic metric space
0 references