“…Then the sum ࣷ is called the direct sum of subspaces U and W. In addition, let ρ(A) and λ max (A) denote spectral radius and the maximal eigenvalue of a square matrix A, respectively. In many problems of practical importance [2,3,10,13,18,19,21], one is concerned with an approximation of a solution of restricted linear equations Ax = b, (1.1) where, A ࢠ C m × n , b ∈ C m and x ࢠ T. By using the restricted generalized inverse of the matrix A relative to the subspace T, the expression by Ben-Israel and Greville [12] for the general solution of consistent linear system (1.1) is given as x = P T (AP T ) (1) …”