Lower bounds for the minimum eigenvalue of Hadamard product of an M-matrix and its inverse

For the Hadamard product A ° A⁻¹ of an M-matrix A and its inverse A⁻¹, we give new lower bounds for the minimum eigenvalue of A ° A⁻¹. These bounds are strong enough to prove the conjecture of Fiedler and Markham [An inequality for the Hadamard product of an M-matrix and inverse M-matrix, Linear Algebra Appl. 101 (1988) 1-8].     

A trace inequality for M-matrices and the symmetrizability of a real matrix by a positive diagonal matrix

The question of whether a real matrix is symmetrizable via multiplication by a diagonal matrix with positive diagonal entries is reduced to the corresponding question for M-matrices and related to Hadamard products. In the process, for a nonsingular M-matrix A, it is shown that tr(A^-1A^T) ? n, with equality if and only if A is symmetric, and that the minimum eigenvalue of A^-1 ° A is ? 1 with equality in the irreducible case if and only if A is positive diagonally symmetrizable.     

Nonnegative matrices with power invariant zero patterns

If A is a nonsingular M-matrix, the elements of the sequence {A^?k} all have the same zero pattern. Using the Drazin inverse, we show that a similar zero pattern invariance property holds for a class of matrices which is larger than the generalized M-matrices.     

M-matrices leading to semiconvergent splittings

An M-matrix as defined by Ostrowski is a matrix that can be split into A = sI ? B, s > 0, B ? 0 with s ? ρ(B), the spectral radius of B. M-matrices with the property that the powers of T = (1/s)B converge for some s are studied and are characterized here in terms of the nonnegativity of the group generalized inverse of A on the range space of A, extending the well-known property that A^{? 1} ? 0 whenever A is nonsingular.     

Classes of general H-matrices

Let M(A) denote the comparison matrix of a square H-matrix A, that is, M(A) is an M-matrix. H-matrices such that their comparison matrices are nonsingular are well studied in the literature. In this paper, we study characterizations of H-matrices with either singular or nonsingular comparison matrices. The spectral radius of the Jacobi matrix of M(A) and the generalized diagonal dominance property are used in the characterizations. Finally, a classification of the set of general H-matrices is obtained.     

On simultaneous diagonalization of one Hermitian and one symmetric form

It is remarked that if A, B ? M_n(C), A = A^t, and B? = B^t, B positive definite, there exists a nonsingular matrix U such that (1) ū^tBU = I and (2) U^tAU is a diagonal matrix with nonnegative entries. Some related actions of the real orthogonal group and equations involving the unitary group are studied.     

LU decomposition of M-matrices by elimination without pivoting

It is shown that if A or ?A is a singular M-matrix satisfying the generalized diagonal dominance condition y^TA?0 for some vector y? 0, then A can be factored into A = LU by a certain elimination algorithm, where L is a lower triangular M-matrix with unit diagonal and U is an upper triangular M-matrix. The existence of LU decomposition of symmetric permutations of A and for irreducible M-matrices and symmetric M-matrices follow as colollaries. This work is motivated by applications to the solution of homogeneous systems of linear equations Ax = 0, where A or ?A is an M-matrix. These applications arise, e.g., in the analysis of Markov chains, input-output economic models, and compartmental systems. A converse of the theorem metioned above can be established by considering the reduced normal form of A.     

An index classification of M-matrices

An M-matrix as defined by Ostrowski [5] is a matrix that can be split into A = sI ? B, where s > 0, B ? 0, with s ? r(B), the spectral radius of B. Following Plemmons [6], we develop a classification of all M-matrices. We consider v, the index of zero for A, i.e., the smallest nonnegative integer n such that the null spaces of Aⁿ and Aⁿ⁺¹ coincide. We characterize this index in terms of convergence properties of powers of s^?1B. We develop additional characterizations in terms of nonnegativity of the Drazin inverse of A on the range of A^v, extending (as conjectured by Poole and Boullion [7]) the well-known property that A^?1?0 whenever A is nonsingular.     

On the convergence of the symmetric successive overrelaxation method

The symmetric successive overrelaxation (SSOR) iterative method is applied to the solution of the system of linear equations $\text{A}\text{x}\text{=}\text{b}$, where A is an n×n nonsingular matrix. We give bounds for the spectral radius of the SSOR iterative matrix when A is an Hermitian positive definite matrix, and when A is a nonsingular M-matrix. Then, we discuss the convergence of the SSOR iterative method associated with a new splitting of the matrix A which extends the results of Varga and Buoni [1].     

Theorems on M-splittings of a singular M-matrix which depend on graph structure

Let $\text{A=M?Nε}\text{R}{}^{\mathrm{n\; n}}$ be a splitting. We investigate the spectral properties of the iteration matrix M^-1N by considering the relationships of the graphs of A, M, N, and M^-1N. We call a splitting an M-splitting if M is a nonsingular M-matrix and N?0. For an M-splitting of an irreducible Z-matrix A we prove that the circuit index of M^-1N is the greatest common divisor of certain sets of integers associated with the circuits of A. For M-splittings of a reducible singular M-matrix we show that the spectral radius of the iteration matrix is 1 and that its multiplicity and index are independent of the splitting. These results hold under somewhat weaker assumptions.     

On the inverse mean first passage matrix problem and the inverse M-matrix problem

The inverse mean first passage time problem is given a positive matrix M∈R^n,n, then when does there exist an n-state discrete-time homogeneous ergodic Markov chain C, whose mean first passage matrix is M? The inverse M-matrix problem is given a nonnegative matrix A, then when is A an inverse of an M-matrix. The main thrust of this paper is to show that the existence of a solution to one of the problems can be characterized by the existence of a solution to the other. In so doing we extend earlier results of Tetali and Fiedler.     

Backward error analysis for linear systems associated with inverses ofH-matrices

In this paper, bounds on the growth factors resulting from Gaussian elimination applied to inverses ofH-matrices are developed and investigated. These bounds are then used in the error analysis for solving linear systemsAx =b whose coefficient matricesA are of this type. For each such system our results show that the Gaussian elimination without pivoting can proceed safely provided that the elements of the inverse of a certainM-matrix (associated with the coefficient matrixA) are not excessively large. We exhibit a particularly satisfactory situation for the special case whenA itself is an inverse of anM-matrix. Part of the first section of this paper is devoted to a discussion on some results of de Boor and Pinkus for the stability of triangular factorizations of systemsAx =b, whereA is a nonsingular totally nonnegative matrix, and to the explanation of why the analysis of de Boor and Pinkus is not applicable to the case when the coefficient matrixA is an inverse of anM-matrix.Research supported in part by NSF Grant MCS-8102114.Research supported in part by the U.S. Army Research Office under contract No. DAAG-29-81-K-0132 and in part by NSF Grant MCS-8219500.     

A new inversion-free method for a rational matrix equation

Motivated by the classical Newton-Schulz method for finding the inverse of a nonsingular matrix, we develop a new inversion-free method for obtaining the minimal Hermitian positive definite solution of the matrix rational equation X+A^∗X^-1A=I, where I is the identity matrix and A is a given nonsingular matrix. We present convergence results and discuss stability properties when the method starts from the available matrix AA^∗. We also present numerical results to compare our proposal with some previously developed inversion-free techniques for solving the same rational matrix equation.     

The inverse M-matrix problem

One aspect of the inverse M-matrix problem can be posed as follows. Given a positive n × n matrix A=(a_ij) which has been scaled to have unit diagonal elements and off-diagonal elements which satisfy 0 < y ? a_ij ? x < 1, what additional element conditions will guarantee that the inverse of A exists and is an M-matrix? That is, if A^?1=B=(b_ij), then b_ii> 0 and b_ij ? 0 for i≠_j. If n=2 or x=y no further conditions are needed, but if n ? 3 and y < x, then the following is a tight sufficient condition. Define an interpolation parameter s via x²=sy+(1?s)y²; then B is an M-matrix if s^?1 ? n?2. Moreover, if all off-diagonal elements of A have the value y except for a_ij=a_jj=x when i=n?1, n and 1 ? _j ? n?2, then the condition on both necessary and sufficient for B to be an M-matrix.     

M-matrix characterizations.I—nonsingular M-matrices

The purpose of this survey is to classify systematically a widely ranging list of characterizations of nonsingular M-matrices from the economics and mathematics literatures. These characterizations are grouped together in terms of their relations to the properties of (1) positivity of principal minors, (2) inverse-positivity and splittings, (3) stability and (4) semipositivity and diagonal dominance. A list of forty equivalent conditions is given for a square matrix A with nonpositive off-diagonal entries to be a nonsingular M-matrix. These conditions are grouped into classes in order to identify those that are equivalent for arbitrary real matrices A. In addition, other remarks relating nonsingular M-matrices to certain complex matrices are made, and the recent literature on these general topics is surveyed.     

Congruence of Hermitian matrices by Hermitian matrices

Two Hermitian matrices A,B∈M_n(C) are said to be Hermitian-congruent if there exists a nonsingular Hermitian matrix C∈M_n(C) such that B=CAC. In this paper, we give necessary and sufficient conditions for two nonsingular simultaneously unitarily diagonalizable Hermitian matrices A and B to be Hermitian-congruent. Moreover, when A and B are Hermitian-congruent, we describe the possible inertias of the Hermitian matrices C that carry the congruence. We also give necessary and sufficient conditions for any 2-by-2 nonsingular Hermitian matrices to be Hermitian-congruent. In both of the studied cases, we show that if A and B are real and Hermitian-congruent, then they are congruent by a real symmetric matrix. Finally we note that if A and B are 2-by-2 nonsingular real symmetric matrices having the same sign pattern, then there is always a real symmetric matrix C satisfying B=CAC. Moreover, if both matrices are positive, then C can be picked with arbitrary inertia.     

On theLU factorization ofM-matrices

Summary In this paper, we give in Theorem 1 a characterization, based on graph theory, of when anM-matrixA admits anLU factorization intoM-matrices, whereL is a nonsingular lower triangularM-matrix andU is an upper triangularM-matrix. This result generalizes earlier factorization results of Fiedler and Pták (1962) and Kuo (1977). As a consequence of Theorem 1, we show in Theorem 3 that the conditionx ^T A0 ^T for somex>0, for anM-matrixA, is both necessary and sufficient forPAP ^T to admit such anLU factorization for everyn×n permutation matrixP. This latter result extends recent work of Funderlic and Plemmons (1981). Finally, Theorem 1 is extended in Theorem 5 to give a characterization, similarly based on graph theory, of when anM-matrixA admits anLU factorization intoM-matrices.Dedicated to Professor Ky Fan on his sixty-seventh birthday, September 19, 1981.Research supported in part by the Air Force Office of Scientific Research, and by the Department of Energy     

A generalization of N-matrices

Ky Fan defines an N-matrix to be a matrix of the form A = tI ? B, B ? 0, λ < t < ?(B), where ?(B) is the spectral radius of B and λ is the maximum of the spectral radii of all principal submatrices of B of order n ? 1. In this paper, we define the closure (N₀-matrices) of N-matrices by letting λ ? t. It is shown that if A ∈ Z and A^-1 < 0, then A ∈ N. Certain inequalities of N-matrices are shown to hold for N₀-matrices, and a method for constructing an N-matrix from an M-matrix is given.     

Involutory functions and Moore-Penrose inverses of matrices in an arbitrary field

Let F be a field, and M be the set of all matrices over F. A function ? from M into M, which we write ?(A) = A^s for A∈M, is involutory if (1) (AB)^s = B^sA^s for all A, B in M whenever the product AB is defined, and (2) (A^s)^s = A for all A∈M. If ? is an involutory function on M, then A^s is n×m if A is m×n; furthermore, Rank A = Rank A^s, the restriction of ? to F is an involutory automorphism of F, and (aA + bB)^s = a^sA^s + b^sB^s for all m×n matrices A and B and all scalars a and b. For an A∈M, an Ã∈M is called a Moore-Penrose inverse of A relative to ? if (i) AÃA = A, ÃAÃ = Ã and (ii) (AÃ)^s = AÃ, (ÃA)^s = ÃA. A necessary and sufficient condition for A to have a Moore-Penrose inverse relative to ? is that Rank A = Rank AA^s = Rank A^sA. Furthermore, if an involutory function ? preserves circulant matrices, then the Moore-Penrose inverse of any circulant matrix relative to ? is also circulant, if it exists.