The maximal and minimal ranks of \(A - BXC\) with applications (Q1419680): Difference between revisions

Maximal and minimal ranks of a linear matrix expression \(A - BXC\) over an arbitrary field \(\mathbb{F}\), where \(A \in {\mathbb F}^{m \times n}\), \(B \in {\mathbb F}^{m \times k}\) and \(C \in {\mathbb F}^{l \times n}\), are investigated with regard to a variant matrix \(X \in {\mathbb F}^{k \times l}\). The obtained results on the extremal ranks of the given linear matrix expression can be used for finding extremal ranks of other linear and nonlinear matrix expressions. The rank invariance and the range invariance of the expression \(A - BXC\) with respect to its variant matrix \(X\) is discussed. On the basis of generalized inverses of matrices the general solution to the rank equation \(\text{rank}(A- BXC)+ \text{rank}(BXC)= \text{rank}(A)\) is presented and the minimal rank of \(A - BXC\) subject to this equation is obtained.