English

The hyperbolic geometry of random transpositions

Probability 2016-08-16 v3 Combinatorics

Abstract

Turn the set of permutations of nn objects into a graph GnG_n by connecting two permutations that differ by one transposition, and let σt\sigma_t be the simple random walk on this graph. In a previous paper, Berestycki and Durrett [In Discrete Random Walks (2005) 17--26] showed that the limiting behavior of the distance from the identity at time cn/2cn/2 has a phase transition at c=1c=1. Here we investigate some consequences of this result for the geometry of GnG_n. Our first result can be interpreted as a breakdown for the Gromov hyperbolicity of the graph as seen by the random walk, which occurs at a critical radius equal to n/4n/4. Let TT be a triangle formed by the origin and two points sampled independently from the hitting distribution on the sphere of radius anan for a constant 0<a<10<a<1. Then when a<1/4a<1/4, if the geodesics are suitably chosen, with high probability TT is δ\delta-thin for some δ>0\delta>0, whereas it is always O(n)-thick when a>1/4a>1/4. We also show that the hitting distribution of the sphere of radius anan is asymptotically singular with respect to the uniform distribution. Finally, we prove that the critical behavior of this Gromov-like hyperbolicity constant persists if the two endpoints are sampled from the uniform measure on the sphere of radius anan. However, in this case, the critical radius is a=1log2a=1-\log2.

Keywords

Cite

@article{arxiv.math/0411011,
  title  = {The hyperbolic geometry of random transpositions},
  author = {Nathanaël Berestycki},
  journal= {arXiv preprint arXiv:math/0411011},
  year   = {2016}
}

Comments

Published at http://dx.doi.org/10.1214/009117906000000043 in the Annals of Probability (http://www.imstat.org/aop/) by the Institute of Mathematical Statistics (http://www.imstat.org)

R2 v1 2026-07-22T17:11:47.366Z