Abstract
For a high dimensional problem, a randomized Gram-Schmidt (RGS) algorithm is beneficial in computational costs as well as numerical stability. We apply this dimension reduction technique by random sketching to Krylov subspace methods, e.g. to the generalized minimal residual method (GMRES). We propose a flexible variant of GMRES with the randomized Gram-Schmidt–based Arnoldi iteration to produce a set of basis vectors of the Krylov subspace. Even though the Krylov basis is no longer \(l_2\) orthonormal, its random projection onto the low dimensional space achieves \(l_2\) orthogonality. As a result, the numerical stability is observed which turns out to be independent of the dimension of the problem even in extreme scale problems. On the other hand, as the harmonic Ritz values are commonly used in GMRES with deflated restarting to improve convergence, we consider another deflation strategy, for instance disregarding the singular vectors associated with the smallest singular values. We thus introduce a new algorithm of the randomized flexible GMRES with singular value decomposition (SVD)–based deflated restarting. At the end, we carry out numerical experiments in the context of compressible turbulent flow simulations. Our proposed approach exhibits a quite competitive numerical behaviour to existing methods while reducing computational costs.
Similar content being viewed by others
Data Availability
The authors confirm that the data supporting the findings of this study are available within the article.
Code Availability
All code and scripts to reproduce plots can be found at Jang’s GitHub (https://github.com/Yongseok7717/RandomizedGMRES).
References
Saad, Y., Schultz, M.H.: GMRES: a generalized minimal residual algorithm for solving nonsymmetric linear systems. SIAM J. Sci. Stat. Comput. 7(3), 856–869 (1986)
Paige, C.C., Saunders, M.A.: Solution of sparse indefinite systems of linear equations. SIAM J. Numer. Anal. 12(4), 617–629 (1975)
Nachtigal, N.M., Reddy, S.C., Trefethen, L.N.: How fast are nonsymmetric matrix iterations? SIAM J. Matrix Anal. Appl. 13(3), 778–795 (1992)
Embree, M.: How descriptive are GMRES convergence bounds? arXiv:2209.01231 (2022)
Morgan, R.B.: GMRES with deflated restarting. SIAM J. Sci. Comput. 24(1), 20–37 (2002)
Giraud, L., Gratton, S., Pinel, X., Vasseur, X.: Flexible GMRES with deflated restarting. SIAM J. Sci. Comput. 32(4), 1858–1878 (2010)
Röllin, S., Fichtner, W.: Improving the accuracy of GMRES with deflated restarting. SIAM J. Sci. Comput. 30(1), 232–245 (2008)
Simoncini, V.: On the convergence of restarted Krylov subspace methods. SIAM J. Matrix Anal. Appl. 22(2), 430–452 (2000)
Parks, M.L., De Sturler, E., Mackey, G., Johnson, D.D., Maiti, S.: Recycling Krylov subspaces for sequences of linear systems. SIAM J. Sci. Comput. 28(5), 1651–1674 (2006)
Daas, H.A., Grigori, L., Hénon, P., Ricoux, P.: Recycling Krylov subspaces and truncating deflation subspaces for solving sequence of linear systems. ACM Trans. Math. Softw. (TOMS) 47(2), 1–30 (2021)
Carvalho, L.M., Gratton, S., Lago, R., Vasseur, X.: A flexible generalized conjugate residual method with inner orthogonalization and deflated restarting. SIAM J. Matrix Anal. Appl. 32(4), 1212–1235 (2011)
Balabanov, O., Grigori, L.: Randomized Gram-Schmidt process with application to GMRES. SIAM J. Sci. Comput. 44(3), 1450–1474 (2022)
Balabanov, O., Grigori, L.: Randomized block Gram-Schmidt process for solution of linear systems and eigenvalue problems. arXiv:2111.14641 (2021)
Nakatsukasa, Y., Tropp, J.A.: Fast & accurate randomized algorithms for linear systems and eigenvalue problems. arXiv:2111.00113 (2021)
Bjorck, A.: Numerical Methods for Least Squares Problems. SIAM, Philadelphia, PA (1996)
Achlioptas, D.: Database-friendly random projections: Johnson-Lindenstrauss with binary coins. J. Comput. Syst. Sci. 66(4), 671–687 (2003)
Halko, N., Martinsson, P.-G., Tropp, J.A.: Finding structure with randomness: probabilistic algorithms for constructing approximate matrix decompositions. SIAM Rev. 53(2), 217–288 (2011)
Woodruff, D.P.: Sketching as a tool for numerical linear algebra. Found. Trends Theor. Comput. Sci. 10(1–2), 1–157 (2014)
Balabanov, O., Nouy, A.: Randomized linear algebra for model reduction. Part I: Galerkin methods and error estimation. Adv. Comput. Math. 45(5), 2969–3019 (2019)
Saad, Y.: A flexible inner-outer preconditioned GMRES algorithm. SIAM J. Sci. Comput. 14(2), 461–469 (1993)
Paige, C.C., Parlett, B.N., Vorst, H.A.: Approximate solutions and eigenvalue bounds from Krylov subspaces. Numer. Linear Algebra Appl. 2(2), 115–133 (1995)
Gaul, A., Gutknecht, M.H., Liesen, J., Nabben, R.: A framework for deflated and augmented Krylov subspace methods. SIAM J. Matrix Anal. Appl. 34(2), 495–518 (2013)
Giraud, L., Langou, J., Rozloznik, M.: The loss of orthogonality in the Gram-Schmidt orthogonalization process. Comput. Math. Appl. 50(7), 1069–1075 (2005)
Abdelmalek, N.N.: Round off error analysis for Gram-Schmidt method and solution of linear least squares problems. BIT Numer. Math. 11(4), 345–367 (1971)
Giraud, L., Langou, J., Rozložník, M., Eshof, Jvd: Rounding error analysis of the classical Gram-Schmidt orthogonalization process. Numerische Mathematik 101(1), 87–100 (2005)
Björck, Å.: Solving linear least squares problems by Gram-Schmidt orthogonalization. BIT Numer. Math. 7(1), 1–21 (1967)
Van Loan, C.F., Golub, G.: Matrix computations (Johns Hopkins studies in mathematical sciences). Matrix Computations (1996)
Tropp, J.A.: Improved analysis of the subsampled randomized Hadamard transform. Adv. Adapt. Data Anal. 3(01n02), 115–126 (2011)
Wang, S.: A practical guide to randomized matrix computations with MATLAB implementations. arXiv:1505.07570 (2015)
Jacobs, E.N., Ward, K.E., Pinkerton, R.M.: The Characteristics of 78 Related Airfoil Section from Tests in the Variable-Density Wind Tunnel. Technical report, NASA (1933)
Allmaras, S.R., Johnson, F.T.: Modifications and clarifications for the implementation of the Spalart-Allmaras turbulence model. In: Seventh International Conference on Computational Fluid Dynamics (ICCFD7), vol. 1902, pp. 1–11 (2012). XBig Island, HI
Crivellini, A., Bassi, F.: An implicit matrix-free discontinuous Galerkin solver for viscous and turbulent aerodynamic simulations. Comput. Fluids 50(1), 81–93 (2011)
Cambier, L., Heib, S., Plot, S.: The ONERA elsA CFD software: input from research and feedback from industry. Mech. Ind., (2013)
Baker, A.H., Jessup, E.R., Kolev, T.V.: A simple strategy for varying the restart parameter in GMRES(m). J. Comput. Appl. Math. 230(2), 751–761 (2009)
Arts, T., De Rouvroit, M.L.: Aero-thermal performance of a two dimensional highly loaded transonic turbine nozzle guide vane: a test case for inviscid and viscous flow computations. In: Turbo Expo: Power for Land, Sea, and Air, vol. 79047, pp. 001–01106 (1990). American Society of Mechanical Engineers
Arts, T., De Rouvroit, M.L., Rutherford, A.: Aero-thermal Investigation of a Highly Loaded Transonic Linear Turbine Guide Vane Cascade. von Karman Institute for Fluid Dynamics, Technical report (1990)
Funding
This work is supported by DGAC (Direction Générale de l’Aviation Civile) in the frame of the SONICE project.
Author information
Authors and Affiliations
Contributions
Y. Jang: writing — original draft, review and editing, investigation, methodology, visualization, software. L. Grigori: conceptualization, formal analysis, supervision. E. Martin: review and editing. C. Content: data curation, project administration.
Corresponding author
Ethics declarations
Conflict of interest
The authors declare no competing interests.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Jang, Y., Grigori, L., Martin, E. et al. Randomized flexible GMRES with deflated restarting. Numer Algor (2024). https://doi.org/10.1007/s11075-024-01801-3
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s11075-024-01801-3
Keywords
- Randomized Gram-Schmidt orthogonalization algorithm
- Random sketching
- Krylov subspace methods
- Implicit restarting with deflation