A note on upper bounds on the cp-rank (Q1039748): Difference between revisions

A matrix \(A\) is called completely positive if it is a Hermitian square of a nonnegative matrix, i.e., if \(A=B^tB\) for some (entrywise) nonnegative matrix \(B\). The minimal number of columns in such \(B\) is called the cp-rank of \(A\). In the paper under review the author gives new short proofs of theorems of \textit{J.~Hannah} and \textit{T.~Laffey} [Linear Algebra Appl. 55, 1--9 (1983; Zbl 0519.15007)] and \textit{F.~Barioli} and \textit{A.~Berman} [Linear Algebra Appl. 363, 17--33 (2003; Zbl 1042.15012)] giving an upper bound on the cp-rank of a matrix in terms of its rank and the number of zeros in a full rank principal submatrix.