Spanning subset sums for finite Abelian groups (Q5931447)

Given a finite commutative group \(G\), the author is studying the number \(w(G)\) defined as the maximal \(k\) with the property that there is a collection \(s_1,\dots,s_k\) of distinct elements of \(G\) and a \(g\in G\) which is not representable in the form \(\sum\varepsilon_is_i\), \(\varepsilon_i=0\) or 1. The quantity \(e(G)\) is defined similary, the only difference being that the representation \(\varepsilon_i\equiv 0\) of the zero element is excluded. The paper describes these quantities for almost all groups; the only groups where the exact answer is not known are cyclic groups of order \(pq\), with primes \(p,q\) satisfying \(p+[2\sqrt{p-2}]+1<q<2p\). For these groups \(w(G)\) is either \(p+q-2\) or \(p+q-1\) and \(e(G)\) is either \(p+q-3\) or \(p+q-2\). The study is heavily based on previous work done mainly by Diderrich, Gao, Hamidoune, Mann.