paper

On the Lipschitz Constant of the RSK Correspondence

arXiv:1012.1819

Abstract

We view the RSK correspondence as associating to each permutation a Young diagram , i.e. a partition of . Suppose now that is left-multiplied by transpositions, what is the largest number of cells in that can change as a result? It is natural refer to this question as the search for the Lipschitz constant of the RSK correspondence. We show upper bounds on this Lipschitz constant as a function of . For , we give a construction of permutations that achieve this bound exactly. For larger we construct permutations which come close to matching the upper bound that we prove.

Updated presentation based on comments made by reviewers. Accepted for publication to JCTA