A NEW PERTURBATION BOUND FOR THE LDU FACTORIZATION OF DIAGONALLY DOMINANT MATRICES

被引:14
作者
Dailey, Megan [1 ]
Dopico, Froilan M. [2 ,3 ]
Ye, Qiang [4 ]
机构
[1] Indiana Univ Kokomo, Kokomo, IN 46904 USA
[2] Univ Carlos III Madrid, CSIC UAM UCM UC3M, Inst Ciencias Matemat, Leganes 28911, Spain
[3] Univ Carlos III Madrid, Dept Matemat, Leganes 28911, Spain
[4] Univ Kentucky, Dept Math, Lexington, KY 40506 USA
基金
美国国家科学基金会;
关键词
accurate computations; column diagonal dominance pivoting; diagonally dominant matrices; diagonally dominant parts; LDU factorization; rank-revealing decomposition; relative perturbation theory; BACKWARD ERROR; EIGENVALUE; ACCURATE;
D O I
10.1137/13093858X
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
This work introduces a new perturbation bound for the L factor of the LDU factorization of (row) diagonally dominant matrices computed via the column diagonal dominance pivoting strategy. This strategy yields L and U factors which are always well-conditioned and, so, the LDU factorization is guaranteed to be a rank-revealing decomposition. The new bound together with those for the D and U factors in [F. M. Dopico and P. Koev, Numer. Math., 119 (2011), pp. 337-371] establish that if diagonally dominant matrices are parameterized via their diagonally dominant parts and off-diagonal entries, then tiny relative componentwise perturbations of these parameters produce tiny relative normwise variations of L and U and tiny relative entrywise variations of D when column diagonal dominance pivoting is used. These results will allow us to prove in a follow-up work that such perturbations also lead to strong perturbation bounds for many other problems involving diagonally dominant matrices.
引用
收藏
页码:904 / 930
页数:27
相关论文
共 40 条
[1]  
Alfa AS, 2002, NUMER MATH, V90, P401, DOI 10.1007/S002110100289
[2]  
[Anonymous], 1959, The Theory of Matrices
[3]   PERTURBATION BOUNDS FOR THE LDLH AND LU DECOMPOSITIONS [J].
BARRLUND, A .
BIT, 1991, 31 (02) :358-363
[4]   Linear perturbation theory for structured matrix pencils arising in control theory [J].
Bora, S ;
Mehrmann, V .
SIAM JOURNAL ON MATRIX ANALYSIS AND APPLICATIONS, 2006, 28 (01) :148-169
[5]   STRUCTURED EIGENVALUE CONDITION NUMBER AND BACKWARD ERROR OF A CLASS OF POLYNOMIAL EIGENVALUE PROBLEMS [J].
Bora, Shreemayee .
SIAM JOURNAL ON MATRIX ANALYSIS AND APPLICATIONS, 2009, 31 (03) :900-917
[6]   ACCURATE SOLUTION OF STRUCTURED LEAST SQUARES PROBLEMS VIA RANK-REVEALING DECOMPOSITIONS [J].
Castro-Gonzalez, Nieves ;
Ceballos, Johan ;
Dopico, Froilan M. ;
Molera, Juan M. .
SIAM JOURNAL ON MATRIX ANALYSIS AND APPLICATIONS, 2013, 34 (03) :1112-1128
[7]   RIGOROUS PERTURBATION BOUNDS OF SOME MATRIX FACTORIZATIONS [J].
Chang, X. -W. ;
Stehle, D. .
SIAM JOURNAL ON MATRIX ANALYSIS AND APPLICATIONS, 2010, 31 (05) :2841-2859
[8]   MULTIPLICATIVE PERTURBATION ANALYSIS FOR QR FACTORIZATIONS [J].
Chang, Xiao-Wen ;
Li, Ren-Cang .
NUMERICAL ALGEBRA CONTROL AND OPTIMIZATION, 2011, 1 (02) :301-316
[9]   Some features of Gaussian elimination with rook pivoting [J].
Chang, XW .
BIT NUMERICAL MATHEMATICS, 2002, 42 (01) :66-83
[10]   On the sensitivity of the LU factorization [J].
Chang, XW ;
Paige, CC .
BIT NUMERICAL MATHEMATICS, 1998, 38 (03) :486-501