Refined Wilf-equivalences by Comtet statistics (Q2055135)

The study of permutation classes defined by pattern avoidance was started by \textit{D. E. Knuth} [The art of computer programming. Vol. 1: Fundamental algorithms. 3rd ed. Reading, MA: Addison-Wesley (1997; Zbl 0895.68055)] and grew since into a separate field of mathematics. A key notion in this field is Wilf-equivalence: two sets of patterns are Wilf-equivalent if the same number of permutations of \(n\) elements avoids either of them, for any \(n\). This paper launches the systematic study of the Wilf-equivalence relations refined by the statistics \(\mathrm{comp}(\pi)\) and \(\mathrm{iar}(\pi)\) where \(\mathrm{comp}(\pi)\) is the number of components of the permutation \(\pi\) and \(\mathrm{iar}(\pi)\) is the number of its ascending runs. The term Comtet statistic is used for any statistic that is equidistributed with \(\mathrm{comp}\) as this permutation statistic was first studied by Comtet. The paper contains bijective proofs of the symmetry of the joint distribution \((\mathrm{comp}, \mathrm{iar})\) over several Catalan and Schröder classes, preserving the values of the left-to-right maxima. It provides a complete classification of \(\mathrm{comp}\)- and \(\mathrm{iar}\)-Wilf-equivalences for length \(3\) patterns and pairs of length \(3\) patterns. For these pattern avoiding classes and for separable permutations, the authors refine the \((\mathrm{comp}, \mathrm{iar})\) distribution with the descent statistic and compute the generating functions. Furthermore, they refine Wang's descent-double descent-Wilf equivalence between separable permutations and \((2413, 4213)\)-avoiding permutations by the Comtet statistic \(\mathrm{iar}\).