A system of linear equations related to the transportation problem with application to probability theory (Q1104859)

A system of \(m+n\) linear equations \[ (1)\quad \sum^{n}_{j=1}p_{ij}x_{ij}=u_ i,\quad 1\leq i\leq m,\quad \sum^{m}_{i=1}p_{ij}x_{ij}=v_ j,\quad 1\leq j\leq n, \] is considered. (For \(p_{ij}\equiv 1\), this system becomes the system of constraints of a transportation problem.) The rank of the coefficient matrix C of this system and a solvability condition for the system are found. In this connection, the relation between the reducibility of the \(m\times n\) matrix \(P=(p_{ij})\) and the rank of C is discussed. The main result is Theorem 2 saying that if P is reducible then the rank of C is \(m+n-1\). The next theorem then says that the system (1) is consistent if \(\sum^{m}_{i=1}u_ i=\sum^{n}_{j=1}v_ j\) and P is irreducible. In the last section, an application to conditional expectation in a finite discrete probability space is given.