Some results about inverse-positive matrices (Q656668)

The notion of a sub-direct sum of two matrices was introduced by \textit{S. M. Fallat} and \textit{C. R. Johnson} [Linear Algebra Appl. 288, No. 1-3, 149--173 (1999; Zbl 0973.15013)] to study positivity classes of matrices, especially the class of \(M\)-matrices. In the paper under review the authors study the problem of inverse positivity of a sub-direct sum, given that each of the summands is inverse positive. (Recall that a square matrix \(A\) is inverse positive, if \(A^{-1}\) exists and \(A^{-1}\) is entrywise nonnegative and that invertible \(M\)-matrices form a subclass of inverse positive matrices). The authors also present sufficient conditions on the structure of an inverse positive matrix ensuring that it could be written as a sub-direct sum of two inverse positive matrices.