Formulas for calculating the dimensions of the sums and the intersections of a family of linear subspaces with applications (Q2316760)

Let \(\mathcal{M}_{1}\) and \(\mathcal{M}_{2}\) be subspaces of a finite dimensional vector space. It is known that then \[ \dim (\mathcal{M}_{1}+\mathcal{M}_{2})=\dim \mathcal{M}_{1}+\dim \mathcal{M}_{2}-\dim (\mathcal{M}_{1}\cap \mathcal{M}_{2}). \tag{1} \] In this paper, the author first revisits some formulas for calculating the dimensions of the intersection of a family \(\mathcal{M}_{1},\mathcal{M}_{2},\dots,\mathcal{M}_{k}\) of vector subspaces of \(\mathbb{C}^{m}\). Then a formula that extends (1) and links the dimensions of the sum and the intersection of a family \(\mathcal{M}_{1},\mathcal{M}_{2},\dots,\mathcal{M}_{k}\) of vector subspaces of \(\mathbb{C}^{m}\) is obtained. Let \(\mathbb{C}^{m\times n}\) denote the set of all \(m\times n\) complex matrices and let \(r(A)\) and \(\mathcal{R}(A)\) stand for the rank and the column space of \(A\in \mathbb{C}^{m\times n}\), respectively. We say that \(A^{-}\in \mathbb{C}^{n\times m}\) is an inner generalized inverse of \(A\in \mathbb{C}^{m\times n}\) if \(AA^{-}A=A\). Note that every matrix \(A\in \mathbb{C}^{m\times n}\) has an inner generalized inverse however it is not necessarily unique. In the second part of the paper, the author presents some applications of the dimensional formulas. In the example that follows, the author determines the maximum and the minimum dimensions of the intersection of column spaces of matrix products involving inner generalized inverses. Let \(A_{i}\in \mathbb{C}^{m_{i}\times n}\), \(i=1,2,\ldots ,k\). Then \[ \begin{aligned} &\max_{A_{1}^{-},\ldots ,A_{k}^{-}}\dim \left[ \mathcal{R}(A_{1}^{-}A_{1})\cap \mathcal{R}(A_{2}^{-}A_{2})\cap \ldots \cap \mathcal{R}(A_{k}^{-}A_{k})\right] \\ &=\min \left\{ r(A_{1}),r(A_{2}),\ldots ,r(A_{k})\right\} \end{aligned} \] and \[ \begin{aligned} &\min_{A_{1}^{-},\ldots ,A_{k}^{-}}\dim \left[ \mathcal{R}(A_{1}^{-}A_{1})\cap \mathcal{R}(A_{2}^{-}A_{2})\cap \ldots \cap \mathcal{R}(A_{k}^{-}A_{k})\right] \\ &=\max \left\{ 0,r(A_{1})+r(A_{2})+\ldots +r(A_{k})-(k-1)n\right\} . \end{aligned} \]