Discrete & Computational Geometry

, Volume 22, Issue 1, pp 105–117 | Cite as

Properties of Random Triangulations and Trees

  • L. Devroye
  • P. Flajolet
  • F. Hurtado
  • M. Noy
  • W. Steiger

Abstract.

Let T n denote the set of triangulations of a convex polygon K with n sides. We study functions that measure very natural ``geometric'' features of a triangulation τ∈ T n , for example, Δ n (τ) which counts the maximal number of diagonals in τ incident to a single vertex of K . It is familiar that T n is bijectively equivalent to B n , the set of rooted binary trees with n-2 internal nodes, and also to P n , the set of nonnegative lattice paths that start at 0 , make 2n-4 steps X i of size \(\pm\) 1, and end at X 1 + . . . +X 2n-4 =0 . Δ n and the other functions translate into interesting properties of trees in B n , and paths in P n , that seem not to have been studied before. We treat these functions as random variables under the uniform probability on T n and can describe their behavior quite precisely. A main result is that Δ n is very close to logn (all logs are base 2 ). Finally we describe efficient algorithms to generate triangulations in T n uniformly, and in certain interesting subsets.

Keywords

Interesting Property Efficient Algorithm Binary Tree Internal Node Convex Polygon 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Copyright information

© Springer-Verlag New York Inc. 1999

Authors and Affiliations

  • L. Devroye
    • 1
  • P. Flajolet
    • 2
  • F. Hurtado
    • 3
  • M. Noy
    • 3
  • W. Steiger
    • 4
  1. 1.School of Computer Science, McGill University, Montreal, Quebec, Canada H3A 2K6 luc@kriek.cs.mcgill.ca CA
  2. 2.INRIA, Rocquencourt, France Philippe.Flajolet@inria.fr FR
  3. 3.Universitat Politecnica de Catalunya, Spain hurtado@ma2.upc.es noy@ma2.upc.es ES
  4. 4.Department of Computer Science, Rutgers University, New Brunswick, NJ 08903, USA steiger@cs.rutgers.eduUS

Personalised recommendations