Erdős-Ko-Rado for perfect matchings

From MaRDI portal
Publication:2400976

DOI10.1016/J.EJC.2017.05.005zbMATH Open1369.05173arXiv1409.2057OpenAlexW2730819323MaRDI QIDQ2400976FDOQ2400976


Authors: Nathan Lindzey Edit this on Wikidata


Publication date: 31 August 2017

Published in: European Journal of Combinatorics (Search for Journal in Brave)

Abstract: A perfect matching of a complete graph K2n is a 1-regular subgraph that contains all the vertices. Two perfect matchings intersect if they share an edge. It is known that if mathcalF is family of intersecting perfect matchings of K2n, then |mathcalF|leq(2(n1)1)!! and if equality holds, then mathcalF=mathcalFij where mathcalFij is the family of all perfect matchings of K2n that contain some fixed edge ij. We give a short algebraic proof of this result, resolving a question of Godsil and Meagher. Along the way, we show that if a family mathcalF is non-Hamiltonian, that is, mcupmotcongC2n for any m,minmathcalF, then |mathcalF|leq(2(n1)1)!! and this bound is met with equality if and only if mathcalF=mathcalFij. Our results make ample use of a somewhat understudied symmetric commutative association scheme arising from the Gelfand pair (S2n,S2wrSn). We give an exposition of a few new interesting objects that live in this scheme as they pertain to our results.


Full work available at URL: https://arxiv.org/abs/1409.2057




Recommendations




Cites Work


Cited In (15)

Uses Software





This page was built for publication: Erdős-Ko-Rado for perfect matchings

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2400976)