Functions preserving matrix groups and iterations for the matrix square root

被引:39
作者
Higham, NJ [1 ]
Mackey, DS
Mackey, N
Tisseur, F
机构
[1] Univ Manchester, Sch Math, Manchester M13 9PL, Lancs, England
[2] Western Michigan Univ, Dept Math, Kalamazoo, MI 49008 USA
关键词
automorphism group; bilinear form; sesquilinear form; scalar product; adjoint; Frechet derivative; stability analysis; perplectic matrix; pseudo-orthogonal matrix; Lorentz matrix; generalized polar decomposition; matrix sign function; matrix pth root; matrix square root; structure preservation; matrix iteration; Newton iteration;
D O I
10.1137/S0895479804442218
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
For which functions f does A is an element of G double right arrow f(A). G when G is the matrix automorphism group associated with a bilinear or sesquilinear form? For example, if A is symplectic when is f( A) symplectic? We show that group structure is preserved precisely when f( A(-1)) = f(A)(-1) for bilinear forms and when f(A(-*)) = f(A)(-*) for sesquilinear forms. Meromorphic functions that satisfy each of these conditions are characterized. Related to structure preservation is the condition f((A) over bar) = <(f(A))over bar>, and analytic functions and rational functions satisfying this condition are also characterized. These results enable us to characterize all meromorphic functions that map every G into itself as the ratio of a polynomial and its "reversal," up to a monomial factor and conjugation. The principal square root is an important example of a function that preserves every automorphism group G. By exploiting the matrix sign function, a new family of coupled iterations for the matrix square root is derived. Some of these iterations preserve every G; all of them are shown, via a novel Frechet derivative-based analysis, to be numerically stable. A rewritten form of Newton's method for the square root of A. G is also derived. Unlike the original method, this new form has good numerical stability properties, and we argue that it is the iterative method of choice for computing A(1/2) when A. G. Our tools include a formula for the sign of a certain block 2 x 2 matrix, the generalized polar decomposition along with a wide class of iterations for computing it, and a connection between the generalized polar decomposition of I + A and the square root of A is an element of G.
引用
收藏
页码:849 / 877
页数:29
相关论文
共 29 条
[1]  
BINI DA, 2004, 454 MANCH CTR COMP M
[2]  
Bolshakov Y, 1997, LINEAR ALGEBRA APPL, V261, P91
[3]   Extension of isometries in finite-dimensional indefinite scalar product spaces and polar decompositions [J].
Bolshakov, Y ;
vanderMee, CVM ;
Ran, ACM ;
Reichstein, B ;
Rodman, L .
SIAM JOURNAL ON MATRIX ANALYSIS AND APPLICATIONS, 1997, 18 (03) :752-774
[5]   Computing the square root and logarithm of a real P-orthogonal matrix [J].
Cardoso, JR ;
Kenney, CS ;
Leite, FS .
APPLIED NUMERICAL MATHEMATICS, 2003, 46 (02) :173-196
[6]  
Cartan H., 1971, DIFFERENTIAL CALCULU
[7]   Approximating the logarithm of a matrix to specified accuracy [J].
Cheng, SH ;
Higham, NJ ;
Kenney, CS ;
Laub, AJ .
SIAM JOURNAL ON MATRIX ANALYSIS AND APPLICATIONS, 2001, 22 (04) :1112-1125
[8]  
Greene RE., 1997, Function theory of one complex variable
[9]   THE MATRIX SIGN DECOMPOSITION AND ITS RELATION TO THE POLAR DECOMPOSITION [J].
HIGHAM, NJ .
LINEAR ALGEBRA AND ITS APPLICATIONS, 1994, 212 :3-20
[10]   NEWTON METHOD FOR THE MATRIX SQUARE ROOT [J].
HIGHAM, NJ .
MATHEMATICS OF COMPUTATION, 1986, 46 (174) :537-549