On the centrosymmetric permutations in a class

TL;DR

This paper explores the relationship between the growth rate of permutation classes and their centrosymmetric elements, proposing conjectures and proving results for sum closed classes and geometric grid classes.

Contribution

It introduces conjectures about growth rates of centrosymmetric permutations and proves them for sum closed classes with growth rate at most approximately 2.30522.

Findings

The growth rate of a permutation class may be strictly greater than that of its centrosymmetric elements.

Conjecture: equality of growth rates holds for sum closed classes.

Proved the conjecture for classes with growth rate ≤ 2.30522.

Abstract

A permutation is centrosymmetric if it is fixed by a half-turn rotation of its diagram. Initially motivated by a question by Alexander Woo, we investigate the question of whether the growth rate of a permutation class equals the growth rate of its even-size centrosymmetric elements. We present various examples where the latter growth rate is strictly less, but we conjecture that the reverse inequality cannot occur. We conjecture that equality holds if the class is sum closed, and we prove this conjecture in the special case where the growth rate is at most $\xi \approx 2.30522$, using results from Pantone and Vatter on growth rates less than $\xi$. We prove one direction of inequality for sum closed classes and for some geometric grid classes. We end with preliminary findings on new kinds of growth-rate thresholds that are a little bit larger than $\xi$.