On Krylov Subspace Approximations to the Matrix Exponential Operator
Open accessThe DOI matches its Crossref record; 29 of the 30 references it cites were found in a registry or an index, and 1 are not checked yet.
Abstract
Krylov subspace methods for approximating the action of matrix exponentials are analyzed in this paper. We derive error bounds via a functional calculus of Arnoldi and Lanczos methods that reduces the study of Krylov subspace approximations of functions of matrices to that of linear systems of equations. As a side result, we obtain error bounds for Galerkin-type Krylov methods for linear equations, namely, the biconjugate gradient method and the full orthogonalization method. For Krylov approximations to matrix exponentials, we show superlinear error decay from relatively small iteration numbers onwards, depending on the geometry of the numerical range, the spectrum, or the pseudospectrum. The convergence to exp$(\tau A)v$ is faster than that of corresponding Krylov methods for the solution of linear equations $(I-\tau A)x=v$, which usually arise in the numerical solution of stiff ordinary differential equations (ODEs). We therefore propose a new class of time integration methods for large systems of nonlinear differential equations which use Krylov approximations to the exponential function of the Jacobian instead of solving linear or nonlinear systems of equations in every time step.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Analogous to the eigenvalue problem in DMRG, both Eqs. ( 7) and ( 8) can be integrated using a Lanczos scheme [28] in order to keep the computational cost at O(D 3 ).”
From Unifying time evolution and optimization with matrix product states (2016) · cites this paper as [28]
Citations
Cites
Solving Ordinary Differential Equations II
Hairer, E., & Wanner, G. (1996) · Springer Series in Computational Mathematics · doi:10.1007/978-3-642-05221-7 · Cited by 4,342
Solving Ordinary Differential Equations II
Hairer, E., & Wanner, G. (1991) · Springer Series in Computational Mathematics · doi:10.1007/978-3-662-09947-6 · Cited by 1,456
Theory of Functions of a Complex Variable
Markushevich, A. I., Silverman, R. A., & Gillis, J. (1966) · Physics Today · doi:10.1063/1.3048359 · Cited by 1,211
Unitary quantum time evolution by iterative Lanczos reduction
Park, T. J., & Light, J. C. (1986) · The Journal of Chemical Physics · doi:10.1063/1.451548 · Cited by 888
Uniform Numerical Methods for Problems with Initial and Boundary Layers.
Berger, A. E., Doolan, E. P., Miller, J. J. H., & Schilders, W. H. A. (1982) · Mathematics of Computation · doi:10.2307/2007353 · Cited by 754
Analysis of Some Krylov Subspace Approximations to the Matrix Exponential Operator
Saad, Y. (1992) · SIAM Journal on Numerical Analysis · doi:10.1137/0729014 · Cited by 670
Numerical Ranges of Operators on Normed Spaces and of Elements of Normed Algebras
Bonsall, F. F., & Duncan, J. (1971) · Cambridge University Press eBooks · doi:10.1017/cbo9781107359895 · Cited by 614
Propagation Methods for Quantum Molecular Dynamics
Kosloff, R. (1994) · Annual Review of Physical Chemistry · doi:10.1146/annurev.pc.45.100194.001045 · Cited by 602
Krylov subspace methods for solving large unsymmetric linear systems
Saad, Y. (1981) · Mathematics of Computation · doi:10.1090/s0025-5718-1981-0616364-6 · Cited by 449
Generalized Runge-Kutta Processes for Stable Systems with Large Lipschitz Constants
Lawson, J. D. (1967) · SIAM Journal on Numerical Analysis · doi:10.1137/0704033 · Cited by 328
Research works printed but not linked: 5
- A. Friedli, Verallgemeinerte Runge‐Kutta Verfahren zur Lösung steifer Differentialgleichungen , in Lecture Notes in Mathematics 631, Springer‐Verlag, Berlin, 1978, pp. 35–50. · doi:10.1007/bfb0067462
- Hochbruck M., 1997, SIAM J. Sci. Comput., 18
- Knizhnerman L. A., 1991, Comput. Math. and Math. Phys., 31, 1
- Knizhnerman L. A., 1992, Comput. Math. and Math. Phys., 32, 1199
- D. R. Taylor, Analysis of the look‐ahead Lanczos algorithm , Ph.D. thesis, University of California at Berkeley, Berkeley, CA, 1982.
Cited by
Items 101–200 of 810
Out-of-time-order correlations and the fine structure of eigenstate thermalization
Brenes, M., Pappalardi, S., Mitchison, M. T., Goold, J., & Silva, A. (2021) · Physical review. E · doi:10.1103/physreve.104.034120 · Cited by 71
Data-sparse approximation to the operator-valued functions of elliptic operator
Gavrilyuk, I. P., Hackbusch, W., & Khoromskij, B. N. (2003) · Mathematics of Computation · doi:10.1090/s0025-5718-03-01590-4 · Cited by 61
“In the matrix case, the most commonly used algorithms for approximating the action of the matrix exponential are based on Krylov subspace methods (see [20, 17]).”
MethodData-sparse approximation to a class of operator-valued functions
Gavrilyuk, I. P., Hackbusch, W., & Khoromskij, B. N. (2004) · Mathematics of Computation · doi:10.1090/s0025-5718-04-01703-x · Cited by 60
Grimm, V. (2005) · Numerische Mathematik · doi:10.1007/s00211-005-0583-8 · Cited by 59
“[9]).”
On the Convergence of Krylov Subspace Methods for Matrix Mittag–Leffler Functions
Moret, I., & Novati, P. (2011) · SIAM Journal on Numerical Analysis · doi:10.1137/080738374 · Cited by 59
“Since we are dealing with any generalized ML function, following the lines of [17], we use formula (3.”
BackgroundMethodInfluentialA multi-resolution approach to heat kernels on discrete surfaces
Vaxman, A., Ben‐Chen, M., & Gotsman, C. (2010) · ACM Transactions on Graphics · doi:10.1145/1778765.1778858 · Cited by 59
Symplectic and multisymplectic numerical methods for Maxwell’s equations
Sun, Y., & Tse, P. S. P. (2010) · Journal of Computational Physics · doi:10.1016/j.jcp.2010.12.006 · Cited by 59
“There are other more efficient approaches to solve this linear system with large L, for example Krylov subspace methods [15,4].”
MethodExtreme-Scale Phase Field Simulations of Coarsening Dynamics on the Sunway TaihuLight Supercomputer
Zhang, J., Zhou, C., Wang, Y., Ju, L., Du, Q., Chi, X., Xu, D., Chen, D., Liu, Y., & Liu, Z. (2016) · International Conference for High Performance Computing, Networking, Storage and Analysis (SC) · doi:10.1109/sc.2016.3 · Cited by 67
“Computing matrix exponentials has been a problem of much interests to the scientific computing community, see for example [39], [40], [41] and [42], [43].”
BackgroundConvergence analysis of the extended Krylov subspace method for the Lyapunov equation
Knizhnerman, L., & Simoncini, V. (2011) · Numerische Mathematik · doi:10.1007/s00211-011-0366-3 · Cited by 59
“We consider an external mapping from the family used for instance in [14], [28],”
MethodBian, L., & Gebraeel, N. (2014) · Naval Research Logistics (NRL) · doi:10.1002/nav.21583 · Cited by 58
Real-time energy dynamics in spin- 12 Heisenberg chains
Langer, S., Heyl, M., McCulloch, I. P., & Heidrich‐Meisner, F. (2011) · Physical Review B · doi:10.1103/physrevb.84.205115 · Cited by 64
Charge and spin dynamics driven by ultrashort extreme broadband pulses: A theory perspective
Moskalenko, A. S., Zhu, Z.-G., & Berakdar, J. (2017) · Physics Reports · doi:10.1016/j.physrep.2016.12.005 · Cited by 60
“based on Chebyshev polynomials, or the iterative Lanczos method generating an orthogonal basis in the corresponding Krylov subspace [164] are commonly used to approximate the action of the exponential operator on the state [156,160,163].”
MethodShift-Invert Arnoldi Approximation to the Toeplitz Matrix Exponential
Lee, S. T., Pang, H., & Sun, H. (2010) · SIAM Journal on Scientific Computing · doi:10.1137/090758064 · Cited by 56
“In fact, the Krylov subspace methods recently have become an efficient means to approximate the matrix exponential [4] [8] [9] [10] [11] [12] [17] [18] [22] [23] [24] [26] [28] [31], especially when the matrix is very large and sparse.”
BackgroundMethodWANG, L. (2004) · Journal of the Mechanics and Physics of Solids · doi:10.1016/j.jmps.2004.05.001 · Cited by 56
On the use of matrix functions for fractional partial differential equations
Garrappa, R., & Popolizio, M. (2010) · Mathematics and Computers in Simulation · doi:10.1016/j.matcom.2010.10.009 · Cited by 55
Stopping Criteria for Rational Matrix Functions of Hermitian and Symmetric Matrices
Frommer, A., & Simoncini, V. (2008) · SIAM Journal on Scientific Computing · doi:10.1137/070684598 · Cited by 64
“, [21] for superlinear convergence results for exp(x)), except possibly for the first phase, where almost complete stagnation may occur.”
BackgroundDeflated Restarting for Matrix Functions
Eiermann, M., Ernst, O. G., & Güttel, S. (2011) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/090774665 · Cited by 54
Althaus, S. L. (1996) · Political Communication · doi:10.1080/10584609.1996.9963092 · Cited by 54
On Restart and Error Estimation for Krylov Approximation of $w=f(A)v$
Tal‐Ezer, H. (2007) · SIAM Journal on Scientific Computing · doi:10.1137/040617868 · Cited by 56
Exponential time integration and Chebychev discretisation schemes for fast pricing of options
Tangman, D. Y., Gopaul, A., & Bhuruth, M. (2007) · Applied Numerical Mathematics · doi:10.1016/j.apnum.2007.07.005 · Cited by 53
Decay Bounds for Functions of Hermitian Matrices with Banded or Kronecker Structure
Benzi, M., & Simoncini, V. (2015) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/151006159 · Cited by 59
Multirate infinitesimal step methods for atmospheric flow simulation
Wensch, J., Knoth, O., & Galant, A. (2009) · BIT Numerical Mathematics · doi:10.1007/s10543-009-0222-3 · Cited by 52
“Lubich and Hochbruck [ 11 ] use a Krylov-space approximation to the exponential function.”
MethodDynamics and decoherence in the central spin model in the low-field limit
Stanek, D., Raas, C., & Uhrig, G. S. (2013) · Physical Review B · doi:10.1103/physrevb.88.155305 · Cited by 66
A Krylov projection method for systems of ODEs
Celledoni, E., & Moret, I. (1997) · Applied Numerical Mathematics · doi:10.1016/s0168-9274(97)00033-0 · Cited by 51
Effective Approximation for the Semiclassical Schrödinger Equation
Bader, P., Iserles, A., Kropielnicka, K., & Singh, P. (2014) · Foundations of Computational Mathematics · doi:10.1007/s10208-013-9182-8 · Cited by 51
“In this case, as noted by [10], the error does not decrease substantially till m ≥ ρ, decreasing rapidly thereafter.”
BackgroundSolution of 3D time-domain electromagnetic problems using optimal subspace projection
Zaslavsky, M., Druskin, V. L., & Knizhnerman, L. (2011) · Geophysics · doi:10.1190/geo2011-0088.1 · Cited by 51
“That is why (where applicable) the SLDM is far superior to the explicit FDTD ( Druskin and Knizhnerman, 1988, 1994; Gallopulos and Saad, 1992; Hochbruck and Lubich, 1997).”
BackgroundLäuchli, A. M., & Werner, P. (2009) · Physical Review B · doi:10.1103/physrevb.80.235117 · Cited by 51
Chen, S., & Zhang, Y. (2011) · Journal of Computational Physics · doi:10.1016/j.jcp.2011.01.010 · Cited by 51
“As discovered in [31], the convergence to ev is substantially faster than that of corresponding Krylov methods for the solution of linear system (I DtA)x = v, which arises in the implicit treatment of diffusion or other high order derivative terms, at least unless a good preconditioner is available.”
BackgroundDoublon dynamics in the extended Fermi-Hubbard model
Hofmann, F., & Potthoff, M. (2012) · Physical Review B · doi:10.1103/physrevb.85.205127 · Cited by 62
An interpolatory approximation of the matrix exponential based on Faber polynomials
Moret, I., & Novati, P. (2001) · Journal of Computational and Applied Mathematics · doi:10.1016/s0377-0427(00)00261-2 · Cited by 51
On the use of exponential time integration methods in atmospheric models
Clancy, C., & Pudykiewicz, J. A. (2013) · Tellus A Dynamic Meteorology and Oceanography · doi:10.3402/tellusa.v65i0.20898 · Cited by 50
“compared with the rational functions arising from the implicit approach (Hochbruck and Lubich, 1997; Loffeld and Tokman, 2013).”
BackgroundInfluentialKnizhnerman, L., Druskin, V. L., & Zaslavsky, M. (2009) · SIAM Journal on Numerical Analysis · doi:10.1137/080715159 · Cited by 50
“, [27, 26, 35, 9, 18]; however, the basic idea first appeared in the classical work of Hestenes and Stiefel [17].”
BackgroundGoto, S., & Danshita, I. (2019) · Physical review. B./Physical review. B · doi:10.1103/physrevb.99.054307 · Cited by 56
Efficient time integration for discontinuous Galerkin approximations of linear wave equations
Hochbruck, M., Pažur, T., Schulz, A., Thawinan, E., & Wieners, C. (2014) · ZAMM ‐ Journal of Applied Mathematics and Mechanics / Zeitschrift für Angewandte Mathematik und Mechanik · doi:10.1002/zamm.201300306 · Cited by 60
“We discuss the relation between standard explicit or implicit Runge-Kutta methods and polynomial or rational Krylov subspace methods for approximating the matrix exponential function directly [11, 14, 16, 21, 32].”
BackgroundMethodSelf-citationPadé approximation for the exponential of a block triangular matrix
Dieci, L., & Papini, A. (2000) · Linear Algebra and its Applications · doi:10.1016/s0024-3795(00)00042-2 · Cited by 56
Jiménez, J. C. (2002) · Applied Mathematics Letters · doi:10.1016/s0893-9659(02)00041-1 · Cited by 54
Gaudreault, S., & Pudykiewicz, J. A. (2016) · Journal of Computational Physics · doi:10.1016/j.jcp.2016.07.012 · Cited by 53
A restarted Lanczos approximation to functions of a symmetric matrix
Ilić, M., Turner, I., & Simpson, D. P. (2009) · IMA Journal of Numerical Analysis · doi:10.1093/imanum/drp003 · Cited by 55
“…a scalar, analytic function f(t) : D ⊂ C → C is the Lanczos approximation f (A) b ≈ ‖b‖Vmf (Tm) e1, b = ‖b‖Vme1 (1.1) where AVm = VmTm + βmvm+1e T m (1.2) is the Lanczos decomposition and the columns of Vm form an orthonormal basis for the Krylov subspace Km(A, b) = span { b, Ab, . . . , Am−1b } .”
BackgroundInfluentialExponential integrators for stiff elastodynamic problems
Michels, D. L., Sobottka, G., & Weber⋆, A. (2014) · ACM Transactions on Graphics · doi:10.1145/2508462 · Cited by 53
“In this context Krylov subspace projection methods have been studied in detail by Saad [1992], Hochbruck and Lubich [1997], and Sidje [1998].”
MethodToro, E. F., & Clarke, J. F. (1998) · Fluid mechanics and its applications · doi:10.1007/978-94-015-9137-9 · Cited by 49
B-SERIES METHODS CANNOT BE VOLUME-PRESERVING
Iserles, A., Quispel, G. R. W., & Tse, P. S. P. (2007) · BIT Numerical Mathematics · doi:10.1007/s10543-006-0114-8 · Cited by 48
“However, if the source-free ODE is of the form ˙ y = Ay + b, where A ∈ sl(n) and y, b ∈ R n , then it is known that exponential integrators such as the explicit ERK methods of Hochbruck, Lubich and Selhofer [9] [10] produce exact solutions for linear ODEs and hence such numerical methods are volume-preserving for line…”
MethodProducing coherent excitations in pumped Mott antiferromagnetic insulators
Wang, Y., Claassen, M., Moritz, B., & Devereaux, T. (2017) · Physical review. B./Physical review. B · doi:10.1103/physrevb.96.235142 · Cited by 48
Light-Enhanced Spin Fluctuations and d -Wave Superconductivity at a Phase Boundary
Wang, Y., Chen, C.-C., Moritz, B., & Devereaux, T. (2018) · Physical Review Letters · doi:10.1103/physrevlett.120.246402 · Cited by 48
Photoexcitation of a Light-Harvesting Supramolecular Triad: A Time-Dependent DFT Study
Spallanzani, N., Rozzi, C. A., Varsano, D., Baruah, T., Pederson, M. R., Manghi, F., & Rubio, Á. (2009) · The Journal of Physical Chemistry B · doi:10.1021/jp900820q · Cited by 48
Convergence of Restarted Krylov Subspace Methods for Stieltjes Functions of Matrices
Frommer, A., Güttel, S., & Schweitzer, M. (2014) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/140973463 · Cited by 47
“f(A)b, the action of a function of a matrix A ∈ Cn×n on a vector b ∈ C by m iterations of Arnoldi’s method [8, 22,26,31], one computes an Arnoldi decomposition”
MethodGeneralized Polar Decompositions for the Approximation of the Matrix Exponential
Zanna, A., & Munthe–Kaas, H. (2002) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/s0895479800377551 · Cited by 47
“See [8] for accurate bounds on the number m of iterations of Krylov methods.”
MethodHamiltonian-based impurity solver for nonequilibrium dynamical mean-field theory
Gramsch, C., Balzer, K., Eckstein, M., & Kollar, M. (2013) · Physical Review B · doi:10.1103/physrevb.88.235106 · Cited by 49
Further development of efficient and accurate time integration schemes for meteorological models
Luân, V. T., Pudykiewicz, J. A., & Reynolds, D. R. (2018) · Journal of Computational Physics · doi:10.1016/j.jcp.2018.10.018 · Cited by 45
“In 1997, Hochbruck and Lubich [39] analyzed Krylov subspace methods for efficiently computing the action of a matrix exponential (with large norm) on some vector.”
Coherent matter waves emerging from Mott-insulators
Rodríguez, K., Manmana, S. R., Rigol, M., Noack, R. M., & Muramatsu, A. (2006) · New Journal of Physics · doi:10.1088/1367-2630/8/8/169 · Cited by 52
Parallel exponential Rosenbrock methods
Luân, V. T., & Ostermann, A. (2016) · Computers & Mathematics with Applications · doi:10.1016/j.camwa.2016.01.020 · Cited by 44
Reachability Analysis of Large Linear Systems With Uncertain Inputs in the Krylov Subspace
Althoff, M. (2019) · IEEE Transactions on Automatic Control · doi:10.1109/tac.2019.2906432 · Cited by 50
“Several possibilities have been developed for approximating e v in the Krylov subspace [37], [55], [56].”
BackgroundDunnett, A., & Chin, A. W. (2021) · Physical review. B./Physical review. B · doi:10.1103/physrevb.104.214302 · Cited by 51
Maset, S., & Zennaro, M. (2008) · Mathematics of Computation · doi:10.1090/s0025-5718-08-02171-6 · Cited by 44
A Filon-type asymptotic approach to solving highly oscillatory second-order initial value problems
Wang, B., Liu, K., & Wu, X. (2013) · Journal of Computational Physics · doi:10.1016/j.jcp.2013.03.009 · Cited by 43
Numerical analysis of parabolic problems with dynamic boundary conditions
Kovács, B., & Lubich, C. (2016) · IMA Journal of Numerical Analysis · doi:10.1093/imanum/drw015 · Cited by 43
“…the entire function j ( z ) = ( e z − 1 ) / z . methods The matrix-vector product j ( t A ) vvv can be approximated efficiently by Krylov subspace ; see Hochbruck & Lubich (1997); Saad (1992). b Since the largest eigenvalues of the symmetric positive semi-definite matrix b A are of size O ( h − 2 )…”
Self-citationExponential time integration using Krylov subspaces
Schulze, J., Schmid, P. J., & Sesterhenn, J. L. (2008) · International Journal for Numerical Methods in Fluids · doi:10.1002/fld.1902 · Cited by 50
“The Krylov subspace approximation to ( A)v converges toward the exact solution as fast as e Av converges to its exact equivalent, see [10].”
BackgroundBesse, C., Dujardin, G., & Lacroix–Violet, I. (2017) · SIAM Journal on Numerical Analysis · doi:10.1137/15m1029047 · Cited by 43
“Moreover the numerical computation of the linear group generated by i∆ is much simpler in the case of the torus [36].”
Zhang, Y. A., Hua, W., Bennett, K., & Mukamel, S. (2014) · Topics in current chemistry · doi:10.1007/128_2014_618 · Cited by 45
Time-Domain Analysis of Large-Scale Circuits by Matrix Exponential Method With Adaptive Control
Weng, S.-H., Chen, Q., & Cheng, C. (2012) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2012.2189396 · Cited by 42
“Fortunately, MEXP only needs the product of e A v, which could be approximated efficiently using Krylov subspace projection [22], [23].”
MethodThe Shape of Data: Intrinsic Distance for Data DistributionsPreprint
Tsitsulin, A., Munkhoeva, M., Mottin, D., Karras, P., Bronstein, A. M., Oseledets, I. V., & Müller, E. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1905.11141 · Cited by 53
“To obtain the error bounds, we use the Theorem 2 in [28], which we recite below.”
MethodComputing $f(A)b$ via Least Squares Polynomial Approximations
Chen, J., Anitescu, M., & Saad, Y. (2011) · SIAM Journal on Scientific Computing · doi:10.1137/090778250 · Cited by 45
“The case f(A) = exp(A) was extensively studied (see, e.g., [28, 13, 17 , 18, 3]).”
Blanes, S., Casas, F., & Thalhammer, M. (2017) · Computer Physics Communications · doi:10.1016/j.cpc.2017.07.016 · Cited by 54
“…approximates the action of a matrix-exponential on a vector, instead of computing the matrix-exponential itself, see [1,26,35,38,41] and references therein; in this respect, procedures that involve only matrix-vector products such as Chebyshev or Krylov methods are usually preferred, see [25,31].”
BackgroundCollective neutrino oscillations with tensor networks using a time-dependent variational principle
Cervia, M. J., Siwach, P., Patwardhan, A. V., Balantekin, A. B., Coppersmith, S. N., & Johnson, C. W. (2022) · Physical review. D/Physical review. D. · doi:10.1103/physrevd.105.123025 · Cited by 46
Numerical approximation of the product of the square root of a matrix with a vector
Allen, E. J., Baglama, J., & Boyd, S. (2000) · Linear Algebra and its Applications · doi:10.1016/s0024-3795(00)00068-9 · Cited by 41
“The only other numerical method available for estimating A~c employs a Krylov procedure for approximating functions [17].”
MethodAn exponential integrator for advection-dominated reactive transport in heterogeneous porous media
Tambue, A., Lord, G. J., & Geiger, S. (2010) · Journal of Computational Physics · doi:10.1016/j.jcp.2010.01.037 · Cited by 45
“The Krylov subspace technique is based on the idea of projecting the operator on a “small” Krylov subspace of the matrix via the Arnoldi process [23, 24].”
MethodThe extended Krylov subspace method and orthogonal Laurent polynomials
Jagels, C., & Reichel, L. (2009) · Linear Algebra and its Applications · doi:10.1016/j.laa.2009.03.006 · Cited by 40
“, [4, 10, 14, 21] for discussions on this approach.”
Dynamic properties of the local linearization method for initial-value problems
Jiménez, J. C., Biscay, R., Mora, C. M., & Rodriguez, L. M. (2002) · Applied Mathematics and Computation · doi:10.1016/s0096-3003(00)00100-4 · Cited by 41
Improved Variants of the Hutch++ Algorithm for Trace Estimation
Persson, D., Cortinovis, A., & Kreßner, D. (2022) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/21m1447623 · Cited by 66
Relativistic ionization dynamics for a hydrogen atom exposed to superintense XUV laser pulses
Lindblom, T. K., Selstø, S., & Lindroth, E. (2017) · Physical Review A · doi:10.1103/physreva.95.043403 · Cited by 43
3D transient electromagnetic modeling using a shift-and-invert Krylov subspace method
Zhou, J., Liu, W., Li, X., & Qi, Z. (2018) · Journal of Geophysics and Engineering · doi:10.1088/1742-2140/aab1d7 · Cited by 39
“It has been proven that Krylov subspace methods always yield super-linear convergence rates to approximate the exponential function (Hochbruck and Lubich 1997).”
BackgroundOn the Method of Neumann Series for Highly Oscillatory Equations
Iserles, A. (2004) · BIT Numerical Mathematics · doi:10.1023/b:bitn.0000046810.25353.95 · Cited by 39
“In principle, the exponentials can be approximated by rational approximations [3], Krylov subspace methods [ 8 ] or special methods that are designed to respect Lie-group structure [5, 16].”
MethodTHE COMPUTATION OF FUNCTIONS OF MATRICES BY TRUNCATED FABER SERIES
Moret, I., & Novati, P. (2001) · Numerical Functional Analysis and Optimization · doi:10.1081/nfa-100105314 · Cited by 39
“For various cases of interest, like those considered for instance in [17, 25, 26], the previous general bounds can be specialized, taking into account the particular structure of .”
BackgroundIlić, M., Turner, I., & Anh, V. (2008) · International Journal of Stochastic Analysis · doi:10.1155/2008/104525 · Cited by 38
Assessing the saturation of Krylov complexity as a measure of chaos
Español, B. L., & Wisniacki, D. A. (2023) · Physical review. E · doi:10.1103/physreve.107.024217 · Cited by 38
“Krylov methods were first introduced in the mathematical literature as an efficient way to perform matrix exponentiation [20, 21].”
MethodApproximation of functions of large matrices with Kronecker structure
Michele Benzi, A., & Simoncini, V. (2016) · Numerische Mathematik · doi:10.1007/s00211-016-0799-9 · Cited by 38
“5 (see [26]), hence we further split I1 as”
InfluentialGuidotti, P., Kim, Y., & Lambers, J. V. (2013) · SIAM Journal on Imaging Sciences · doi:10.1137/120882895 · Cited by 38
“KSS methods compute a Jacobi matrix corresponding to each Fourier coefficient, in contrast to traditional Krylov subspace methods (see, for example, [13, 14, 23]) that normally use only a single Krylov subspace generated by the initial data or the solution from the previous time step.”
MethodJiménez, J. C., & Ozaki, T. (2003) · International Journal of Control · doi:10.1080/0020717031000138214 · Cited by 45
“…conditional moments x̂ t=tk and P t=tk is reduced to use a con- venient algorithm to compute matrix exponentials, e.g. those based on rational Padé approximations (Golub and Van Loan 1996), the Schur decomposition (Golub and Van Loan 1996) or Krylov subspace methods (Hochbruck and Lubich 1997).”
MethodExponential Fourier Collocation Methods for Solving First-Order Differential Equations
Wang, B., Wu, X., Meng, F., & Fang, Y. (2017) · Journal of Computational Mathematics · doi:10.4208/jcm.1611-m2016-0596 · Cited by 37
“Moveover, many different approaches to evaluating this action in an efficient way have been proposed in the literature, see, e.g. [1,2,4,22,23,27,35,36].”
BackgroundConvergence Analysis of Exponential Time Differencing Schemes for the Cahn-Hilliard Equation
Li, X. L. X., Ju, L., & Meng, X. (2019) · Communications in Computational Physics · doi:10.4208/cicp.2019.js60.12 · Cited by 37
Resolvent Krylov subspace approximation to operator functions
Grimm, V. (2011) · BIT Numerical Mathematics · doi:10.1007/s10543-011-0367-8 · Cited by 37
A higher order local linearization method for solving ordinary differential equations
la Cruz, H. D., Biscay, R., Carbonell, F., Ozaki, T., & Jiménez, J. C. (2006) · Applied Mathematics and Computation · doi:10.1016/j.amc.2006.06.096 · Cited by 41
“, [3], [4], [21], [22], [23], [24], [25]).”
BackgroundMethodInfluentialTambue, A., Berre, I., & Nordbotten, J. M. (2012) · Advances in Water Resources · doi:10.1016/j.advwatres.2012.12.004 · Cited by 41
“Computing matrix exponentials of stiff matrices is a notorious problem in numerical analysis [33], but new developments for both Léja points and Krylov subspace techniques [3, 22, 40, 41, 45] have led to efficient numerical approaches; see e.”
Symmetry-protected tensor networks
Hubig, C. (2017) · Elektronische Hochschulschriften der LMU München (Ludwig-Maximilians-Universität München) · doi:10.5282/edoc.21348 · Cited by 80
Incorporating Animal Movement Into Distance Sampling
Glennie, R., Buckland, S. T., Langrock, R., Gerrodette, T., Ballance, L. T., Chivers, S. J., & Scott, M. D. (2020) · Journal of the American Statistical Association · doi:10.1080/01621459.2020.1764362 · Cited by 36
“The sparse structure of G can be exploited to accelerate computations using the Krylov subspace approximation (Hochbruck and Lubich 1997) and, when movement is isotropic and G block-Toeplitz, the two-dimensional discrete Fourier transform (Lee 1986).”
MethodComputation of the Exponential of Large Sparse Skew-Symmetric Matrices
Buono, N. D., Lopez, L., & Peluso, R. I. (2005) · SIAM Journal on Scientific Computing · doi:10.1137/030600758 · Cited by 36
“Polynomial approximants, instead, are based on the approximation of exp(τA)q1 on some Krylov subspace of dimension m < n, but these techniques are computationally competitive only when m is small compared with n [11, 14, 25].”
BackgroundMethodRational approximation to trigonometric operators
Grimm, V., & Hochbruck, M. (2008) · BIT Numerical Mathematics · doi:10.1007/s10543-008-0185-9 · Cited by 36
“For the approximation of y(�) = f(�2A)v by Krylov methods, it is known from [ 18 ], that the approximation in the Krylov space”
MethodSelf-citationLanger, S., Schuetz, M. J. A., McCulloch, I. P., Schollwöck, U., & Heidrich‐Meisner, F. (2012) · Physical Review A · doi:10.1103/physreva.85.043618 · Cited by 35
Fast Computation of the Matrix Exponential for a Toeplitz Matrix
Kreßner, D., & Luce, R. (2018) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/16m1083633 · Cited by 35
“For a matrix T of large norm, the Krylov subspace method can be expected to converge slowly [15].”
MethodConvergence Analysis of Projection Methods for the Numerical Solution of Large Lyapunov Equations
Simoncini, V., & Druskin, V. L. (2009) · SIAM Journal on Numerical Analysis · doi:10.1137/070699378 · Cited by 35
“The following mapping is a modified version of the external mapping used for instance in [15], ψ(τ) = γ1 − γ2(1 − 1 τ )τ, τ = σe, (4.”
Computing a matrix function for exponential integrators
Lu, Y. Y. (2003) · Journal of Computational and Applied Mathematics · doi:10.1016/j.cam.2003.08.006 · Cited by 35
“In this case, the Krylov subspace method [4, 8, 3] can be used.”
MethodBisetti, F. (2011) · Combustion Theory and Modelling · doi:10.1080/13647830.2011.631032 · Cited by 38
“Krylov subspace methods [19,20] allow one to calculate approximate and computationally economical matrix functions by projecting the original matrix onto a subspace of (much) lower dimensionality.”
MethodEfficient approximation of the exponential operator for discrete 2D advection–diffusion problems
Bergamaschi, L., Caliari, M., & Vianello, M. (2002) · Numerical Linear Algebra with Applications · doi:10.1002/nla.288 · Cited by 41
“In several papers [5, 8, 9, 11, 12, 13] explicit time integration schemes have been proposed for the solution of (2), which rest on the”
BackgroundKrylov-space approach to the equilibrium and nonequilibrium single-particle Green’s function
Balzer, M., Gdaniec, N., & Potthoff, M. (2011) · Journal of Physics Condensed Matter · doi:10.1088/0953-8984/24/3/035603 · Cited by 34
Preserving geometric properties of the exponential matrix by block Krylov subspace methods
Lopez, L., & Simoncini, V. (2006) · BIT Numerical Mathematics · doi:10.1007/s10543-006-0096-6 · Cited by 34
“represent the de-facto alternative to far more costly procedures for approximating exp(A)v [37]; this has motivated the study of the convergence properties of Krylov subspace based approaches; see, e.g., [19, 28 , 35, 42].”
BackgroundGauger, C. M., Leinen, P., & Yserentant, H. (2000) · SIAM Journal on Numerical Analysis · doi:10.1137/s0036142999352564 · Cited by 34
“23) converges very fast for entire functions φ, much faster than the conjugate gradient method for the solution of a linear system with the coefficient matrix I− τA; see [4].”
MethodInfluentialBarthel, T., Kasztelan, C., McCulloch, I. P., & Schollwöck, U. (2009) · Physical Review A · doi:10.1103/physreva.79.053627 · Cited by 34
Krylov subspace exponential time domain solution of Maxwell’s equations in photonic crystal modeling
Botchev, M. A. (2015) · Journal of Computational and Applied Mathematics · doi:10.1016/j.cam.2015.04.022 · Cited by 37
“These methods include Krylov subspace methods [86, 25, 52, 32, 72, 26, 43, 44, 27], the Chebyshev polynomials, scaling and squaring with Padé or Taylor approximations and other methods [82, 23, 75, 17,”
Rosenbrock--Krylov Methods for Large Systems of Differential Equations
Tranquilli, P., & Sandu, A. (2014) · SIAM Journal on Scientific Computing · doi:10.1137/130923336 · Cited by 33
“A related family of methods are exponential integrators [15], which use matrix exponentials of the Jacobian as part of the solution process and evaluate matrix exponential times vector products via Krylov space methods [18, 6, 36, 16, 17].”
MethodART: Adaptive residual-time restarting for Krylov subspace matrix exponential evaluations
Botchev, M. A., & Knizhnerman, L. (2019) · Journal of Computational and Applied Mathematics · doi:10.1016/j.cam.2019.06.027 · Cited by 38
Tamascelli, D., Rosenbach, R., & Plenio, M. B. (2015) · Physical Review E · doi:10.1103/physreve.91.063306 · Cited by 45
Metadata
- Authors
- Hochbruck, M., & Lubich, C.
- Affiliations
- University of Tübingen (DE)
- Venue
- SIAM Journal on Numerical Analysis
- Published
- 1997-10
- Volume
- 34
- Issue
- 5
- Pages
- 1911-1925
- ISSN
- 0036-1429 (ISSN-L)
- 1095-7170
- OpenAlex
- W1984347942
- Cited by
- 795
- References
- 30
- Citations by year
- 2026: 27 · 2025: 23 · 2024: 19 · 2023: 23 · 2022: 23 · 2021: 37 · 2020: 42 · 2019: 41 · 2018: 46 · 2017: 41 · 2016: 42 · 2015: 32 · 2014: 44 · 2013: 44 · 2012: 30
- FWCI
- 14.18
- Type
- journal article
- Language
- en
- Open access
- green
- doi.org
- Licence
- other-oa
- Locations
- SIAM Journal on Numerical Analysis (publishedVersion)
- Repository KITopen (Karlsruhe Institute of Technology) (publishedVersion)
- KITopen (other-oa)
- Topics
- Physical Sciences › Computer Science › Computational Theory and Mathematics › Matrix Theory and Algorithms
- Physical Sciences › Mathematics › Numerical Analysis › Numerical Solvers for Ordinary Differential Equations
- Physical Sciences › Engineering › Computational Mechanics › Advanced Numerical Methods in Computational Mathematics
- Keywords
- matrix exponential, Krylov subspace methods, Arnoldi method, Lanczos method, functional calculus, error bounds, biconjugate gradient method, numerical range, pseudospectra, superlinear convergence, stiff ordinary differential equations, time integration methods, nonlinear differential equations, Jacobian matrix
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.