Restricted partition pairs

From MaRDI portal





The author finds an explicit representation of the generating function for pairs of partitions: \((M \geq p_ N \geq p_{N-1}\geq \cdots \geq p_ 1\geq0\); \(M \geq q_ N \geq q_{N-1} \geq \cdots \geq q_ 1\geq 0)\), \(\sum p_ i+q_ i=n\). \(M\) and \(N\) are fixed bounds, with the additional restriction that for fixed parameters \(a\), \(b\), \(\alpha\), and \(\beta\): \[ p_ i-q_{i+1-a} \geq 1-\alpha, \qquad q_ i-p_{i+1-b} \geq 1- \beta. \] The reason for studying such pairs of partitions is that they encompass and generalize partitions for which restrictions have been placed on certain hook differences (the difference between the length of the \(i\)th row and the \(i\)th column, for example). It is useful to think of the \(p_ i\)s as forming the rows to the right of the Durfee square in the Ferrers graph and the \(q_ i\)s as forming the columns below the Durfee square. The restrictions in the displayed equations then become hook differences. One of the corollaries of Burge's representation of the generating function is the observation that it is symmetric in \(a\) and \(\alpha\) and in \(b\) and \(\beta\). A more important corollary is a doubly bounded form of the Rogers-Ramanujan identities that provides a unifying generalization of several classical but previously unrelated results that each imply the Rogers-Ramanujan identities.




Cited in
(29)








This page was built for publication: Restricted partition pairs

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1801786)