Determinants of partition matrices (Q1912271): Difference between revisions

Given two partitions, \(\mu\) and \(\lambda\), of the integer \(n\), the number of refinements of \(\mu\) into \(\lambda\) is defined to be the number of ways of taking a set partition with cardinalities given by \(\mu\) and refining it into a set partition with cardinalities given by \(\lambda\). Given a partition \(\lambda\) of \(n\), we define \(\lambda+ 1^n\) to be the partition of \(2n\) obtained by adding parts of size 0 as necessary so that there are exactly \(n\) parts, and then adding 1 to each part. We define \(\lambda\cup 1^n\) to be the partition of \(2n\) which has all of the parts in \(\lambda\) plus additional \(n\) parts of size 1. Finally, given \(n\), we let \(\lambda_1,\dots, \lambda_{p(n)}\) be the partitions of \(n\) listed in reverse lexicographic order, and we define \(M_n\) to be the \(p(n)\times p(n)\) matrix for which the entry in row \(i\), column \(j\) is the number of refinements of \(\lambda_j+ 1^n\) into \(\lambda_i\cup 1^n\). The author proves that for all positive integers \(n\), the determinant of \(M_n\) is 1.