The height of Mallows trees

Abstract

Random binary search trees are obtained by recursively inserting the elements σ(1),σ(2),…,σ(n) of a uniformly random permutation σ of [n]={1,…,n} into a binary search tree data structure. Devroye (J. Assoc. Comput. Mach. 33 (1986) 489–498) proved that the height of such trees is asymptotically of order c∗logn, where c∗=4.311… is the unique solution of clog((2e)/c)=1 with c≥2. In this paper, we study the structure of binary search trees Tn,q built from Mallows permutations. A Mallows(q) permutation is a random permutation of [n]={1,…,n} whose probability is proportional to qInv(σ), where Inv(σ)=|{i<j:σ(i)>σ(j)}|. This model generalizes random binary search trees, since Mallows(q) permutations with q=1 are uniformly distributed. The laws of Tn,q and Tn,q−1 are related by a simple symmetry (switching the roles of the left and right children), so it suffices to restrict our attention to q≤1. We show that, for q∈[0,1], the height of Tn,q is asymptotically (1+o(1))(c∗logn+n(1−q)) in probability. This yields three regimes of behaviour for the height of Tn,q, depending on whether n(1−q)/logn tends to zero, tends to infinity or remains bounded away from zero and infinity. In particular, when n(1−q)/logn tends to zero, the height of Tn,q is asymptotically of order c∗logn, like it is for random binary search trees. Finally, when n(1−q)/logn tends to infinity, we prove stronger tail bounds and distributional limit theorems for the height of Tn,q.