Random generation of Q-convex sets

  • Authors:
  • Sara Brunetti;Alain Daurat

  • Affiliations:
  • Dipartimento di Scienze Matematiche e Informatiche, Università di Siena, Pian dei Mantellini, Siena, Italy;LSIIT UMR CNRS-ULP, Pôle API, Boulevard Sébastien Brant, Illkirch-Graffenstaden, France

  • Venue:
  • Theoretical Computer Science
  • Year:
  • 2005

Quantified Score

Hi-index 5.23

Visualization

Abstract

The problem of randomly generating Q-convex sets is considered. We present two generators. The first one uses the Q-convex hull of a set of random points in order to generate a Q-convex set included in the square [0, n)2. This generator is very simple, but is not uniform and its performance is quadratic in n. The second one exploits a coding of the salient points, which generalizes the coding of a border of polyominoes. It is uniform, and is based on the method by rejection. Experimentally, this algorithm works in linear time in the length of the word coding the salient points. Besides, concerning the enumeration problem, we determine an asymptotic formula for the number of Q-convex sets according to the size of the word coding the salient points in a special case, and in general only an experimental estimation.