Abstract
We consider the problem of computing a modest number of the smallest eigenvalues along with orthogonal bases for the corresponding eigenspaces of a symmetric positive definite operatorA defined on a finite dimensional real Hilbert spaceV. In our applications, the dimension ofV is large and the cost of invertingA is prohibitive. In this paper, we shall develop an effective parallelizable technique for computing these eigenvalues and eigenvectors utilizing subspace iteration and preconditioning forA. Estimates will be provided which show that the preconditioned method converges linearly when used with a uniform preconditioner under the assumption that the approximating subspace is close enough to the span of desired eigenvectors.
Similar content being viewed by others
References
O. Axelsson, A generalized conjugate gradient, least squares method, Numer. Math. 51 (1987) 209–228.
G. Birkhoff and A. Schoenstadt (eds.),Elliptic Problem Solvers II (Academic Press, New York, 1984).
P. E. Bjørstad and O. B. Wildlund, Solving elliptic problems on regions partitioned into substructures, in:Elliptic Problem Solvers II, eds. G. Birkhoff and A. Schoenstadt (Academic Press, New York, 1984) pp. 245–256.
J. H. Bramble,Multigrid Methods, Pitman Research Notes in Mathematics Series (Longman Sci. Tech., London). Copublished with Wiley, New York, 1993.
J. H. Bramble and J. E. Pasciak, New convergence estimates for multigrid algorithms, Math. Comp. 49 (1987) 311–329.
J. H. Bramble and J. E. Pasciak, New estimates for multigrid algorithms including the V-cycle, Math. Comp. 60 (1993) 447–471.
J. H. Bramble, J. E. Pasciak and A. H. Schatz, An iterative method for elliptic problems on regions partitioned into substructures, Math. Comp. 46 (1986) 361–369.
J. H. Bramble, J. E. Pasciak and A. H. Schatz, The construction of preconditioners for elliptic problems by substructuring I, Math. Comp. 47 (1986) 103–134.
J. H. Bramble, J. E. Pasciak and A. H. Schatz, The construction of preconditioners for elliptic problems by substructuring II, Math. Comp. 49 (1987) 1–16.
J. H. Bramble, J. E. Pasciak and A. H. Schatz, The construction of preconditioners for elliptic problems by substructuring III, Math. Comp. 51 (1988) 415–430.
J. H. Bramble, J. E. Pasciak and A. H. Schatz, The construction of preconditioners for elliptic problems by substructuring IV, Math. Comp. 53 (1989) 1–24.
J. H. Bramble, J. E. Pasciak, J. Wang and J. Xu, Convergence estimates for product iterative methods with applications to domain decomposition, Math. Comp. 57 (1991) 1–21.
J. H. Bramble, J. E. Pasciak and J. Xu, Parallel multilevel preconditioners, Math. Comp. 55 (1990) 1–22.
Z. Cao, Generalized Rayleigh quotient matrix and bloc algorithm for solving large sparse symmetric generalized eigenvalue problems, Numerical Math. J. Chinese Univ. 5 (1983) 342–348.
T. F. Chan, R. Glowinski, J. Periaux and O. B. Widlund (eds.),Domain Decomposition Methods (SIAM, Philadelphia, PA, 1989).
T. F. Chan, R. Glowinski, J. Periaux and O. B. Widlund (eds.),3rd Int. Symp. on Domain Decomposition Methods for Partial Differential Equations (SIAM, Philadelphia, PA, 1990).
N. Chetty, M. Weinert, T. S. Rahman and J. W. Davenport, Vacancies and impurities in aluminum and magnesium, Phys. Rev. B (1995) 6313–6326.
J. Davidson, The iterative calculation of a few of the lowest eigenvalues and corresponding eigenvectors of large real symmetric matrices, J. Comput. Phys. 17 (1975) 87–94.
J. Davidson, Matrix eigenvector methods, in:Methods in Computational Molecular Physics (Reidel, Boston, 1983) pp. 95–113.
M. Dryja and O. Widlund, An additive variant of the Schwarz alternating method for the case of many subregions, Technical Report 339, Courant Institute of Mathematical Sciences (1987).
E. G. D'yakonov,Optimization in Solving Elliptic Problems (CRC, Boca Raton, 1995).
E. G. D'yakonov and A. V. Knyazev, Group iterative method for finding lower-order eigenvalues, Moscow Univ. Comput. Math. Cybern. 2 (1982) 32–40.
E. G. D'yakonov and A. V. Knyazev, On an iterative method for finding lower eigenvalues, Russian J. Numer. Anal. Math. Modelling 7 (1992) 473–486.
E. G. D'yakonov and M. Yu. Orekhov, Minimization of the computational labor in determining the first eigenvalues of differential operators, Math. Notes 27 (1980) 382–391.
E. G. D'yakonov, Iteration methods in eigenvalue problems, Math. Notes 34 (1983) 945–953.
R. Glowinski, G. H. Golub, G. A. Meurant and J. Periaux (eds.),1st Int. Symp. on Domain Decomposition Methods for Partial Differential Equations (SIAM, Philadelphia, PA, 1988) pp. 144–172.
R. Glowinski, Y. A. Kuznetzov, G. Meurant, J. Periaux and O. B. Widlund (eds.),4th Int. Symp. on Domain Decomposition Methods for Partial Differential Equations (SIAM, Philadelphia, PA, 1991) pp. 263–289.
S. K. Godunov, V. V. Ogneva and G. P. Prokopov, On the convergence of the modified steepest descent method in application to eigenvalue problems, Trans. Amer. Math. Soc. 2 (1976) 105.
V. P. Il'in,Numerical Methods of Solving Electrophysical Problems (Nauka, Moscow, 1985) (in Russian).
T. Kato,Perturbation Theory for Linear Operators (Springer, New York, 1976).
A. V. Knyazev, Convergence rate estimates for iterative methods for a mesh symmetric eigenvalue problem, Russian J. Numer. Anal. Math. Modelling 2 (1987) 371–396.
A. V. Knyazev,Computation of Eigenvalues and Eigenvectors for Mesh Problems: The Algorithms and Error Estimates (Dept. Numer. Math., USSR Acad. Sci., Moscow, 1986) (in Russian).
A. V. Knyazev, A preconditioned conjugate gradient method for eigenvalue problems and its implementation in a subspace, in:Eigenwertaufgaben in Natur- und Intgenieurwissenschaften und ihre Numerische Behandlung, Oberwolfach (1990), Int. Ser. Numer. Math., Vol. 96 (Birkhäuser, Basel, 1991) pp. 143–154.
A. V. Knyazev, New estimates for Ritz vectors, CIMS NYU 677 (New York, 1994). Also Math. Comp., to appear.
A. V. Knyazev and A. L. Skorokhodov, The preconditioned gradient-type iterative methods in a subspace for partial generalized symmetric eigenvalue problem, SIAM J. Numer. Anal. 31 (1994) 1226.
N. Kosugi, Modification of the Liu-Davidson method for obtaining one or simultaneously several eigensolutions of a large real symmetric; the preconditioned gradient-type iterative methods in a subspace for partial generalized symmetric eigenvalue problem, J. Comput. Phys. 55 (1984) 426–436.
D. E. Longsine and S. F. McCormick, Simultaneous Raileigh-quotient minimization methods forAx=λBx, Linear Algebra Appl. 34 (1980) 195–234.
S. F. McCormick and T. Noe, Simultaneous iteration for the matrix eigenvalue problem, Linear Algebra Appl. 16 (1977) 43–56.
R. B. Morgan, Davidson's method and preconditioning for generalized eigenvalue problems, J. Comput. Phys. 89 (1990) 241–245.
R. B. Morgan and D. S. Scott, Generalizations of Davidson's method for computing eigenvalues of sparse symmetric matrices, SIAM J. Sci. Statist. Comput. 7 (1986) 817–825.
R. B. Morgan and D. S. Scott, Preconditioning the Lanczos algorithm for sparse symmetric eigen-value problems, SIAM J. Sci. Comput. 14 (1993) 585–593.
C. W. Murray, S. C. Racine and E. R. Davidson, Improved algorithms for the lowest few eigenvalues and associated eigenvectors of large matrices, J. Comput. Phys. 103 (1992) 382–389.
B. N. Parlett,The Symmetric Eigenvalue Problem (Prentice-Hall, Englewood Cliffs, NJ, 1980).
W. V. Petryshyn, On the eigenvalue problemTu −λSu=0 with unbounded and non-symmetric operatorsT andS, Philos. Trans. R. Soc. Math. Phys. Sci. 262 (1968) 413–458.
A. Ruhe, SOR-methods for the eigenvalue problem with large sparse matrices, Math. Comp. 28 (1974) 695–710.
A. Ruhe, Iterative eigenvalue algorithm based on convergent splittings, J. Comput. Phys. 19 (1975) 110–120.
Y. Saad,Numerical Methods for Large Eigenvalue Problems (Halsted Press, New York, 1992).
M. Sadkane, Block-Arnoldi and Davidson methods for unsymmetric large eigenvalue problems, Numer. Math. 64 (1993) 195–211.
B. A. Samokish, The steepest descent method for an eigenvalue problem with semi-bounded operators, Izv. Vyssh. Uchebn. Zaved. Mat. 5 (1958) 105–114 (in Russian).
D. S. Scott, Solving sparse symmetric generalized eigenvalue problems without factorization, SIAM J. Numer. Anal. 18 (1981) 102–110.
A. Stathopoulos, Y. Saad and C. F. Fischer, Robust preconditioning of large, sparse, symmetric eigenvalue problems, J. Comput. Appl. Math. (to appear). Also report TR-93-093 of AHPCRC, University of Minneapolis (1993).
P. Vassilevski, Hybrid V-cycle algebraic multilevel preconditioners, Preprint, Bulgarian Academy Sciences, Sofia, Bulgaria (1987).
Author information
Authors and Affiliations
Additional information
Communicated by D.N. Arnold
This manuscript has been authored under contract number DE-AC02-76CH00016 with the U.S. Department of Energy. Accordingly, the U.S. Government retains a non-exclusive, royalty-free license to publish or reproduce the published form of this contribution, or allow others to do so, for U.S. Government purposes. This work was also supported in part under the National Science Foundation Grants No. DMS-9007185, DMS-9501507 and NSF-CCR-9204255, and by the U.S. Army Research Office through the Mathematical Sciences Institute, Cornell University.
Rights and permissions
About this article
Cite this article
Bramble, J.H., Pasciak, J.E. & Knyazev, A.V. A subspace preconditioning algorithm for eigenvector/eigenvalue computation. Adv Comput Math 6, 159–189 (1996). https://doi.org/10.1007/BF02127702
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF02127702