Diagonalizablity of a substochastic matrix

102 Views Asked by At

Let D+E be a stochastic matrix i. e., D+E be the matrix with each row sum 1, where D and E are non-zero nonnegetive, irreducible matrices, I-D being non-singular. Can it be proved that D+kE is diagonalizable for 0< |k| <1? I tried different pair of matrices (D, E), but each time it is diagonalizable, but I can't prove it generally. Please help me. Thanks in advance.