Abstract
This article focuses on the design and development of a new robust and efficient general-purpose incomplete Cholesky factorization package HSL_MI28, which is available within the HSL mathematical software library. It implements a limited memory approach that exploits ideas from the positive semidefinite Tismenetsky-Kaporin modification scheme and, through the incorporation of intermediate memory, is a generalization of the widely used ICFS algorithm of Lin and Moré. Both the density of the incomplete factor and the amount of memory used in its computation are under the user's control. The performance of HSL_MI28 is demonstrated using extensive numerical experiments involving a large set of test problems arising from a wide range of real-world applications. The numerical experiments are used to isolate the effects of scaling, ordering, and dropping strategies so as to assess their usefulness in the development of robust algebraic incomplete factorization preconditioners and to select default settings for HSL_MI28. They also illustrate the significant advantage of employing a modest amount of intermediate memory. Furthermore, the results demonstrate that, with limited memory, high-quality yet sparse general-purpose preconditioners are obtained. Comparisons are made with ICFS, with a level-based incomplete factorization code and, finally, with a state-of-the-art direct solver.
- Patrick R. Amestoy, Timothy A. Davis, and Iain S. Duff. 2004. Algorithm 837: AMD, an approximate minimum degree ordering algorithm. ACM Trans. Math. Softw. 30, 3, 381--388. Google ScholarDigital Library
- Owe Axelsson, Igor Kaporin, Igor Konshin, Andrey Kucherov, Maya Neytcheva, Ben Polman, and Alex Yeremin. 2000. Comparison of algebraic solution methods on a set of benchmark problems in linear elasticity. Final Report of the STW project NNS.4683. Department of Mathematics, University of Nijmegen.Google Scholar
- Owe Axelsson and Niels Munksgaard. 1983. Analysis of incomplete factorizations with fixed storage allocation. In Preconditioning Methods: Analysis and Applications. Topics in Computational Mathematics, Vol. 1, Gordon & Breach, New York, 219--241.Google Scholar
- Lourenco Beirão da Veiga, Vitaliy Gyrya, Konstantin Lipnikov, and Gianmarco Manzini. 2009. Mimetic finite difference method for the Stokes problem on polygonal meshes. J. Comput. Phys. 228, 19, 7215--7232. Google ScholarDigital Library
- Michele Benzi. 2002. Preconditioning techniques for large linear systems: a survey. J. Comput. Phys. 182, 2, 418--477. Google ScholarDigital Library
- Michele Benzi, John C. Haws, and Miroslav Tůma. 2000. Preconditioning highly indefinite and nonsymmetric matrices. SIAM J. Sci. Comput. 22, 4, 1333--1353. Google ScholarDigital Library
- Michele Benzi and M. Tůma. 2003. A robust incomplete factorization preconditioner for positive definite matrices. Numer. Linear Algebra Appl. 10, 5--6, 385--400.Google ScholarCross Ref
- Elisabeth H. Cuthill and James McKee. 1969. Reducing the bandwidth of sparse symmetric matrices. In Proceedings of the 24th National Conference of the ACM. ACM Press, 157--172. Google ScholarDigital Library
- Timothy A. Davis and Yifan Hu. 2011. The University of Florida Sparse Matrix Collection. ACM Trans. Math. Softw. 38, Article 1. Google ScholarDigital Library
- Elizabeth D. Dolan and Jorge J. Moré. 2002. Benchmarking optimization software with performance profiles. Math. Prog. 91, 2, 201--213.Google ScholarCross Ref
- Iain. S. Duff and Jacko Koster. 2001. On algorithms for permuting large entries to the diagonal of a sparse matrix. SIAM J. Matrix Anal. Appl. 22, 4, 973--996. Google ScholarDigital Library
- Iain S. Duff and Stéphane Pralet. 2005. Strategies for scaling and pivoting for sparse symmetric indefinite problems. SIAM J. Matrix Anal. Appl. 27, 2, 313--340. Google ScholarDigital Library
- Roland W. Freund and Nöel M. Nachtigal. 1990. An implementation of the look-ahead Lanczos algorithm for non-Hermitian matrices, part II. Tech. Rep., TR 90--46. RIACS, NASA Ames Research Center.Google Scholar
- Thomas George, Anshul Gupta, and Vivek Sarin. 2009, revised 2011. An empirical analysis of the performance of preconditioners for SPD systems. Res. Rep. RC 24737, IBM.Google Scholar
- Thomas George, Anshul Gupta, and Vivek Sarin. 2012. An empirical analysis of the performance of preconditioners for SPD systems. ACM Trans. Math. Softw. 38, Article 24. Google ScholarDigital Library
- Nicholas I. M. Gould, Yifan F. Hu, and Jennifer A. Scott. 2007. A numerical evaluation of sparse direct symmetric solvers for the solution of large sparse, symmetric linear systems of equations. ACM Trans. Math. Softw. 33, Article 10. Google ScholarDigital Library
- Nicholas I. M. Gould and Jennifer A. Scott. 2004. A numerical evaluation of HSL packages for the direct solution of large sparse, symmetric linear systems of equations. ACM Trans. Math. Softw. 30, 3, 300--325. Google ScholarDigital Library
- Anshul Gupta and Thomas George. 2010. Adaptive techniques for improving the performance of incomplete factorization preconditioning. SIAM J. Sci. Comput. 32, 1, 84--110. Google ScholarDigital Library
- Jonathan D. Hogg and Jennifer A. Scott. 2008. The effects of scalings on the performance of a sparse symmetric indefinite solver. Tech. Rep., RAL-TR-2008-007, Rutherford Appleton Laboratory.Google Scholar
- Jonathan D. Hogg and Jennifer A. Scott. 2011. HSL_MA97: A bit-compatible multifrontal code for sparse symmetric systems. Tech. Rep., RAL-TR-2011-024, Rutherford Appleton Laboratory.Google Scholar
- Jonathan D. Hogg and Jennifer A. Scott. 2014. Pivoting strategies for tough sparse indefinite systems. ACM Trans. Math. Softw. 40, 1, Article 4. DOI: http://dx.doi.org/10.1145/2513109.2513113 Google ScholarDigital Library
- HSL. 2013. A collection of Fortran codes for large-scale scientific computation. http://www.hsl.rl.ac.uk.Google Scholar
- David Hysom and Alex Pothen. 2001. A scalable parallel algorithm for incomplete factor preconditioning. SIAM J. Sci. Comput. 22, 6, 2194--2215. Google ScholarDigital Library
- Alan Jennings and George M. Malik. 1977. Partial elimination. J. Inst. Math. Appl. 20, 3, 307--316.Google ScholarCross Ref
- Alan Jennings and George M. Malik. 1978. The solution of sparse linear equations by the conjugate gradient method. Int. J. Numer. Methods Engrg. 12, 1, 141--158.Google ScholarCross Ref
- Mark T. Jones and Paul E. Plassmann. 1995a. Algorithm 740: Fortran subroutines to compute improved incomplete Cholesky factorizations. ACM Trans. Math. Softw. 21, 1, 18--19. Google ScholarDigital Library
- Mark T. Jones and Paul E. Plassmann. 1995b. An improved incomplete Cholesky factorization. ACM Trans. Math. Softw. 21, 1, 5--17. Google ScholarDigital Library
- Igor E. Kaporin. 1998. High quality preconditioning of a general symmetric positive definite matrix based on its UTU+UTR+RTU decomposition. Numer. Linear Algebra Appl. 5, 6, 483--509.Google ScholarCross Ref
- George Karypis and Vipin Kumar. October 1997. METIS: A software package for partitioning unstructured graphs, partitioning meshes, and computing fill-reducing orderings of sparse matrices (Version 3.0). Tech. Rep., University of Minnesota, Department of Computer Science and Army HPC Research Center.Google Scholar
- Chih-Jen Lin and Jorge J. Moré. 1999. Incomplete Cholesky factorizations with limited memory. SIAM J. Sci. Comput. 21, 1, 24--45. Google ScholarDigital Library
- Konstantin Lipnikov, Mikhail Shashkov, Daniil Svyatskiy, and Yuri Vassilevski. 2007. Monotone finite volume schemes for diffusion equations on unstructured triangular and shape-regular polygonal meshes. J. Comput. Phys. 227, 1, 492--512. Google ScholarDigital Library
- Konstantin Lipnikov, Daniil Svyatskiy, and Yuri Vassilevski. 2009. Interpolation-free monotone finite volume method for diffusion equations on polygonal meshes. J. Comput. Phys. 228, 3, 703--716. Google ScholarDigital Library
- Thomas A. Manteuffel. 1980. An incomplete factorization technique for positive definite linear systems. Math. Comp. 34, 150, 473--497.Google Scholar
- John K. Reid and Jennifer A. Scott. 1999. Ordering symmetric sparse matrices for small profile and wavefront. Int. J. Numer. Methods Eng. 45, 1737--1755.Google ScholarCross Ref
- Daniel Ruiz. 2001. A scaling algorithm to equilibrate both rows and columns norms in matrices. Tech. Rep., RAL-TR-2001-034, Rutherford Appleton Laboratory, Chilton, Oxfordshire, England.Google Scholar
- Daniel Ruiz and Bora Uçar. 2011. A symmetry preserving algorithm for matrix scaling. Tech. Rep., INRIA RR-7552, INRIA, Grenoble, France.Google Scholar
- Yousef Saad. 1994. ILUT: A dual threshold incomplete LU factorization. Numer. Linear Algebra Appl. 1, 4, 387--402.Google ScholarCross Ref
- Yousef Saad. 1996. Iterative Methods for Sparse Linear Systems. PWS Publishing Co., Boston. Google ScholarDigital Library
- Jennifer A. Scott and Miroslav Tůma. 2011. The importance of structure in incomplete factorization preconditioners. BIT 51, 2, 385--404.Google ScholarCross Ref
- Jennifer A. Scott and Miroslav Tůma. 2014. On positive semidefinite modification schemes for incomplete Cholesky factorization. SIAM J. Sci. Comput. 36, 2, A634--A667.Google ScholarCross Ref
- Scott W. Sloan. 1986. An algorithm for profile and wavefront reduction of sparse matrices. Int. J. Numer. Methods Eng. 23, 239--251.Google ScholarCross Ref
- Scott W. Sloan. 1989. A Fortran program for profile and wavefront reduction. Int. J. Numer. Methods Eng. 28, 2651--2679.Google ScholarCross Ref
- Made Suarjana and Kincho H. Law. 1995. A robust incomplete factorization based on value and space constraints. Int. J. Numer. Methods Eng. 38, 1703--1719.Google ScholarCross Ref
- Miron Tismenetsky. 1991. A new preconditioning technique for solving large sparse linear systems. Linear Algebra Appl. 154--156, 331--353.Google Scholar
- Ichitaro Yamazaki, Zhaojun Bai, Wenbin Chen, and Richard Scalettar. 2009. A high-quality preconditioning technique for multi-length-scale symmetric positive definite linear systems. Numer. Math. Theory, Methods and Appl. 2, 4, 469--484.Google ScholarCross Ref
Index Terms
- HSL_MI28: An Efficient and Robust Limited-Memory Incomplete Cholesky Factorization Code
Recommendations
Symmetric Permutations for I-matrices to Delay and Avoid Small Pivots During Factorization
In this article, we present several new permutations for I-matrices making these more suitable for incomplete LU-factorization preconditioners used in solving linear systems by iterative methods. A general matrix can be transformed by row permutation as ...
MIQR: A Multilevel Incomplete QR Preconditioner for Large Sparse Least-Squares Problems
This paper describes a multilevel incomplete QR factorization for solving large sparse least-squares problems. The algorithm builds the factorization by exploiting structural orthogonality in general sparse matrices. At any given step, the algorithm ...
Low-rank updates of balanced incomplete factorization preconditioners
Let Ax = b be a large and sparse system of linear equations where A is a nonsingular matrix. An approximate solution is frequently obtained by applying preconditioned iterations. Consider the matrix B = A + PQT where P,Qźźn k$P, Q \in \mathbb {R}^{n \...
Comments