Partition identities involving gaps and weights. II (Q1265257): Difference between revisions

[Part I, Trans. Am. Math. Soc. 349, 5001-5019 (1997; Zbl 0893.11042).] If \(\pi\) is a partition, let \(\sigma(\pi)\) denote the sum of the parts of \(\pi\). If \(T\) is a set of partitions, let \(p_T(n)\) denote the number of partitions of \(n\) that belong to \(T\). Let \(S\subset T\). The author, extending some prior results, obtains several identities of the form \[ p_T(n) = \sum \bigl\{w_{S,T} (\pi): \pi\in S,\;\sigma(\pi) =n\bigr\}, \] where the weights \(w_{S,T}(n)\) are positive integers. Among the cases considered are: (1) \(T=\) set of all partitions, \(S=\) set of all partitions into parts \(\not \equiv 2\pmod 4\); (2) \(T\) as in (1), \(S=\) set of all partitions such that odd parts do not repeat; (3) \(T\) as in (1), \(S=\) set of all partitions such that no part occurs 3 or more times; (4) \(T=R_1=\) set of all Rogers-Ramanujan partitions of the first kind, \(S=\) set of all partitions such that the parts differ by at least 4; (5) \(T=R_2=\) set of all Rogers-Ramanujan partitions of the second kind, \(S\) as in (4); (6) \(T=\) set of all partitions into distinct parts, \(S=R_1\).