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 201–300 of 810
Localization in Matrix Computations: Theory and Applications
Benzi, M. (2016) · Lecture notes in mathematics · doi:10.1007/978-3-319-49887-4_4 · Cited by 45
“The bounds are a consequence of fundamental error bounds for Krylov subspace approximations to the matrix exponential due to Hochbruch and Lubich [109].”
BackgroundResultInfluentialA fourth-order commutator-free exponential integrator for nonautonomous differential equations
Thalhammer, M. (2006) · SIAM Journal on Numerical Analysis · doi:10.1137/05063042 · Cited by 33
Garcia, F., Bonaventura, L., Net, M., & Sánchez, J. (2014) · Journal of Computational Physics · doi:10.1016/j.jcp.2014.01.033 · Cited by 34
“In this section, we introduce the two different classes of methods for integrating stiff ordinary differential equations that have been considered in this paper, namely the IMEX backward differentiation formulae (IMEX-BDF) [7,8,21] and the exponential time differencing schemes (ETD) [15,18,22,23].”
MethodUnconditionally stable integration of Maxwell’s equations
Verwer, J. G., & Botchev, M. A. (2009) · Linear Algebra and its Applications · doi:10.1016/j.laa.2008.12.036 · Cited by 33
On the performance of exponential integrators for problems in magnetohydrodynamics
Einkemmer, L. D., Tokman, M., & Loffeld, J. J. (2016) · Journal of Computational Physics · doi:10.1016/j.jcp.2016.11.027 · Cited by 35
An investigation of GPU-based stiff chemical kinetics integration methods
Curtis, N., Niemeyer, K. E., & Sung, C. (2017) · Combustion and Flame · doi:10.1016/j.combustflame.2017.02.005 · Cited by 35
“Furthermore, Hochbruck and Lubich [35, 36] demonstrated that the action of the matrix exponential on a vector obtained using Krylov subspace approximation converges faster than corresponding Krylov methods for the solution of linear equations.”
MethodAn efficient second-order linear scheme for the phase field model of corrosive dissolution
Gao, H., Ju, L., Duddu, R., & Li, H. (2019) · Journal of Computational and Applied Mathematics · doi:10.1016/j.cam.2019.112472 · Cited by 36
Haut, T. S., Babb, T., Martinsson, P., & Wingate, B. (2015) · IMA Journal of Numerical Analysis · doi:10.1093/imanum/drv021 · Cited by 32
“[15] and [11]).”
Chebyshev matrix product state approach for time evolution
Halimeh, J. C., Kolley, F., & McCulloch, I. P. (2015) · Physical Review B · doi:10.1103/physrevb.92.115130 · Cited by 36
Resonantly enhanced tunneling and transport of ultracold atoms on tilted optical lattices
Rubbo, C. P., Manmana, S. R., Peden, B. M., Holland, M., & Rey, A. M. (2011) · Physical Review A · doi:10.1103/physreva.84.033638 · Cited by 33
Exponential integrators for time–fractional partial differential equations
Garrappa, R. (2013) · The European Physical Journal Special Topics · doi:10.1140/epjst/e2013-01973-1 · Cited by 32
“In the last years great efforts have been devoted to device efficient methods for evaluating matrix functions [19], with special emphasis on exponential and related functions [3,4,20,26,33,35,39].”
Recursion relations for the extended Krylov subspace method
Jagels, C., & Reichel, L. (2010) · Linear Algebra and its Applications · doi:10.1016/j.laa.2010.08.042 · Cited by 32
“Functions of interest in applications include f(t) := exp(t), f(t) := √ t, f(t) := ln(t); see [1, 2, 4, 5, 7, 8, 9, 11, 12, 13, 14, 18, 19, 20, 27] for applications and discussions on numerical methods for the evaluation of (1.”
BackgroundParallel solution in time of ODEs: some achievements and perspectives
Amodio, P., & Luigi Brugnano, A. (2008) · Applied Numerical Mathematics · doi:10.1016/j.apnum.2008.03.024 · Cited by 32
“The approach that we shall consider is similar to what suggested, for example, in [23,18], even though different approaches can be also considered (see, e.”
BackgroundResultTime Integration and Steady-State Continuation for 2d Lubrication Equations
Beltrame, P., & Thiele, U. (2010) · SIAM Journal on Applied Dynamical Systems · doi:10.1137/080718619 · Cited by 32
“We show that the necessary dimensionK is about one hundred, while in [31, 73, 38, 98], K ≈ 10 is sufficient.”
BackgroundA second-order Magnus-type integrator for nonautonomous parabolic problems
González, C., Ostermann, A., & Thalhammer, M. (2005) · Journal of Computational and Applied Mathematics · doi:10.1016/j.cam.2005.04.036 · Cited by 32
“From a telescopic identity, it follows nm = n−1∑ j=m+1 nj+1(e hAj − ehAm)e(tj−tm)Am + n∑ j=m+1 e(tn+1−tj+1)Am(ehAj − ehAm)e(tj−tm)Am .”
Matching exponential-based and resolvent-based centrality measures
Aprahamian, M., Higham, D. J., & Higham, N. J. (2015) · Journal of Complex Networks · doi:10.1093/comnet/cnv016 · Cited by 39
“There is a wide range of alternative algorithms based on Krylov subspace projection techniques [2], [26], [42].”
MethodExponential Fourier Collocation Methods for First-Order Differential Equations
Wu, X., & Wang, B. (2018) · Springer eBooks · doi:10.1007/978-981-10-9004-2_3 · Cited by 31
A massively parallel exponential integrator for advection-diffusion models
Martínez, Á., Bergamaschi, L., Caliari, M., & Vianello, M. (2009) · Journal of Computational and Applied Mathematics · doi:10.1016/j.cam.2009.01.024 · Cited by 35
Jiménez-Galán, Á., Argenti, L., & Martín, F. (2013) · New Journal of Physics · doi:10.1088/1367-2630/15/11/113009 · Cited by 35
“For moderately intense fields, this operation is efficiently accomplished with a Krylov scheme [76] which generally”
MethodA block Krylov subspace time‐exact solution method for linear ordinary differential equation systems
Botchev, M. A. (2013) · Numerical Linear Algebra with Applications · doi:10.1002/nla.1865 · Cited by 33
“, [10–20]).”
A physically based approach to the accurate simulation of stiff fibers and stiff fiber meshes
Michels, D. L., Mueller, J., & Sobottka, G. (2015) · Computers & Graphics · doi:10.1016/j.cag.2015.10.001 · Cited by 31
“Efficient determination of Krylov dimension: To determine a conservative value of m such as the Krylov-based evaluations of matrix functions remain below a given error threshold, we follow [22]: the absolute error”
BackgroundTime-Domain Simulations of the Nonlinear Maxwell Equations Using Operator-Exponential Methods
Pototschnig, M., Niegemann, J., Tkeshelashvili, L., & Busch, K. (2009) · IEEE Transactions on Antennas and Propagation · doi:10.1109/tap.2008.2011181 · Cited by 31
“In order to guarantee that the entire Lawson4 integrator (22) is fourth order in time, we evaluate the matrix exponentials through Krylov-subspace techniques [7], [12]–[14] with a Krylov-subspace dimension of 6.”
MethodError Analysis of Krylov Methods In a Nutshell
Hochbruck, M., & Lubich, C. (1998) · SIAM Journal on Scientific Computing · doi:10.1137/s1064827595290450 · Cited by 31
“This is shown in [10] using Faber polynomials like in [5, 15].”
MethodSelf-citationTambue, A., & Ngnotchouye, J. M. T. (2016) · Applied Numerical Mathematics · doi:10.1016/j.apnum.2016.04.013 · Cited by 30
Gharibnejad, H., Schneider, B. I., Leadingham, M., & Schmale, H. J. (2019) · Computer Physics Communications · doi:10.1016/j.cpc.2019.05.019 · Cited by 30
On the Matrix Mittag–Leffler Function: Theoretical Properties and Numerical Computation
Popolizio, M. (2019) · Mathematics · doi:10.3390/math7121140 · Cited by 30
“The convergence of Krylov subspace methods can be quite slow when the spectrum of A is large; this phenomenon was primarily studied for the exponential [44] and successively for the ML function [12].”
BackgroundGao, H., Ju, L., Li, X., & Duddu, R. (2019) · Journal of Computational Physics · doi:10.1016/j.jcp.2019.109191 · Cited by 33
Linear estimation of continuous-discrete linear state space models with multiplicative noise
Jiménez, J. C., & Ozaki, T. (2002) · Systems & Control Letters · doi:10.1016/s0167-6911(02)00150-0 · Cited by 30
“, those based on rational PadN e approximations [5], the Schur decomposition [5] or Krylov sub-space methods [6] (for a recent review see [19]).”
MethodRunge–Kutta integrators yield optimal regularization schemes
Rieder, A. (2005) · Inverse Problems · doi:10.1088/0266-5611/21/2/003 · Cited by 30
“Exponential integrators are Runge–Kutta-like schemes which solve linear ODEs exactly, see, for instance, Hochbruck, Lubich and Selhofer [7].”
MethodNumerical time propagation of quantum systems in radiation fields
Alvermann, A., Fehske, H., & Littlewood, P. B. (2012) · New Journal of Physics · doi:10.1088/1367-2630/14/10/105008 · Cited by 36
Jiang, T., & Zhang, Y. (2016) · Journal of Computational Physics · doi:10.1016/j.jcp.2016.01.021 · Cited by 30
“By the analysis theorems provided in [10, 16], an error estimation of the Krylov subspace approximation (26) for this case is”
MethodEfficient design of exponential-Krylov integrators for large scale computing
Tokman, M., & Loffeld, J. J. (2010) · Procedia Computer Science · doi:10.1016/j.procs.2010.04.026 · Cited by 30
A variational splitting integrator for quantum molecular dynamics
Lubich, C. (2003) · Applied Numerical Mathematics · doi:10.1016/j.apnum.2003.09.001 · Cited by 30
Self-citationApproximating Semidefinite Packing Programs
Iyengar, G., Phillips, D. J., & Stein, C. (2011) · SIAM Journal on Optimization · doi:10.1137/090762671 · Cited by 29
“Currently, the best known techniques for efficiently computing the matrix exponential rely on using the Lanczos method to computing the basis of the Krylov subspaces associated with the matrix Γ [10, 13] or (I + θΓ)−1 [31] for an appropriately chosen θ.”
Rainwater, G., & Tokman, M. (2014) · Journal of Computational Physics · doi:10.1016/j.jcp.2014.03.012 · Cited by 29
Circuit simulation via matrix exponential method for stiffness handling and parallel processing
Weng, S.-H., Chen, Q., Wong, N., & Cheng, C. (2012) · doi:10.1145/2429384.2429469 · Cited by 29
“Previous work [19], [20] showed that MEXP with adaptive step control leads to a several”
BackgroundInexact Inverse Subspace Iteration with Preconditioning Applied to Non-Hermitian Eigenvalue Problems
Robbé, M., Sadkane, M., & Spence, A. (2009) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/060673795 · Cited by 32
“6) should be replaced by the larger one N = 3 l(∂E) 2π 2 (see [12]).”
BackgroundGeiger, S., Lord, G. J., & Tambue, A. (2012) · Computational Geosciences · doi:10.1007/s10596-011-9273-z · Cited by 35
“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 [13, 28, 32].”
MethodInfluentialLow-Rank Updates of Matrix Functions
Beckermann, B., Kreßner, D., & Schweitzer, M. (2018) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/17m1140108 · Cited by 28
Weak exponential schemes for stochastic differential equations with additive noise
Mora, C. M. (2005) · IMA Journal of Numerical Analysis · doi:10.1093/imanum/dri001 · Cited by 28
Shift-invert Lanczos method for the symmetric positive semidefinite Toeplitz matrix exponential
Pang, H., & Sun, H. (2010) · Numerical Linear Algebra with Applications · doi:10.1002/nla.747 · Cited by 28
“Shift-invert technique However, it is shown in [11, 16] that when ‖ Tn‖2 becomes larger, more iterations in the Lanczos process may be needed to achieve a given accuracy for the approximation to the vector e− Tn r .”
BackgroundAn exponential time-integrator scheme for steady and unsteady inviscid flows
Li, S., Luo, L., Wang, Z. J., & Ju, L. (2018) · Journal of Computational Physics · doi:10.1016/j.jcp.2018.03.020 · Cited by 28
Mott polaritons in cavity-coupled quantum materials
Kiffner, M., Coulthard, J., Schlawin, F., Ardavan, A., & Jaksch, D. (2019) · New Journal of Physics · doi:10.1088/1367-2630/ab31c7 · Cited by 28
“We calculate the optical conductivity with the full system Hamiltonian using Krylov subspace methods [49] for a half-filled electronic system”
MethodApplication of operator splitting to the Maxwell equations including a source term
Botchev, M. A., Faragó, I., & Horváth, R. (2008) · Applied Numerical Mathematics · doi:10.1016/j.apnum.2008.03.031 · Cited by 28
ROSENBROCK-KRYLOV METHODS FOR LARGE SYSTEMS OF DIFFERENTIAL EQUATIONSPreprint
Tranquilli, P., Adrian, S., & Sandu, A. (n.d.) · Cited by 28
Geometric numerical integration of the assignment flow
Zeilmann, A., Savarino, F., Petra, S., & Schnörr, C. (2020) · Inverse Problems · doi:10.1088/1361-6420/ab2772 · Cited by 31
Floquet oscillations in periodically driven Dirac systems
Junk, V., Reck, P., Gorini, C., & Richter, K. (2020) · Physical review. B./Physical review. B · doi:10.1103/physrevb.101.134302 · Cited by 27
Crouseilles, N., Einkemmer, L. D., & Massot, J. (2020) · Journal of Computational Physics · doi:10.1016/j.jcp.2020.109688 · Cited by 27
Just SLaQ When You Approximate: Accurate Spectral Distances for Web-Scale Graphs
Tsitsulin, A., Munkhoeva, M., & Perozzi, B. (2020) · International World Wide Web Conference (WWW) · doi:10.1145/3366423.3380026 · Cited by 33
“reduce the variance of the estimate, we apply the 2-term polynomial variance reduction technique. Although matrix-vector product approximation bounds for the matrix exponential have been well studied [21], analogous error estimates for the von Neumann entropy remain an open problem in numerical linear algebra. We summ…”
MethodElectron Dynamics with the Time-Dependent Density Matrix Renormalization Group
Baiardi, A. (2021) · Journal of Chemical Theory and Computation · doi:10.1021/acs.jctc.0c01048 · Cited by 33
Nearly reducible finite Markov chains: Theory and algorithms
Sharpe, D., & Wales, D. J. (2021) · The Journal of Chemical Physics · doi:10.1063/5.0060978 · Cited by 29
Mei, L., Huang, L., & Wu, X. (2022) · SIAM Journal on Scientific Computing · doi:10.1137/21m1412475 · Cited by 27
The generalized Kadanoff-Baym ansatz. Computing nonlinear response properties of finite systems
Balzer, K., Hermanns, S., & Bönitz, M. (2013) · Journal of Physics Conference Series · doi:10.1088/1742-6596/427/1/012006 · Cited by 27
Global extended Krylov subspace methods for large-scale differential Sylvester matrix equations
Sadek, E. M., Bentbib, A. H., Sadek, L., & Alaoui, H. T. (2019) · Journal of Applied Mathematics and Computing · doi:10.1007/s12190-019-01278-7 · Cited by 26
Chen, H., & Sun, H. (2021) · Journal of Scientific Computing · doi:10.1007/s10915-021-01431-0 · Cited by 26
Filtered Conjugate Residual‐type Algorithms with Applications
Saad, Y. (2006) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/060648945 · Cited by 26
“, [25, 23, 5, 16, 15]), though the attention was primarily focussed on the case when f is analytic (e.”
BackgroundExplicit schemes for time propagating many-body wave functions
Frapiccini, A. L., Hamido, A., Schröter, S., Pyke, D., Mota-Furtado, F., O’Mahony, P. F., Madroñero, J., Eiglsperger, J., & Piraux, B. (2014) · Physical Review A · doi:10.1103/physreva.89.023418 · Cited by 26
Sotero, R. C., Trujillo‐Barreto, N. J., Jiménez, J. C., Carbonell, F., & Rodríguez‐Rojas, R. (2008) · Journal of Computational Neuroscience · doi:10.1007/s10827-008-0109-3 · Cited by 29
“This was achieved in the present paper by using Krylov subspace methods (Hochbruck and Lubich 1997).”
MethodSuman, S. K., & Kumar, A. (2022) · Archives of Computational Methods in Engineering · doi:10.1007/s11831-021-09690-8 · Cited by 25
Optimal control of high-harmonic generation by intense few-cycle pulses
Solanpää, J., Budagosky, J. A., Shvetsov-Shilovski, N. I., Castro, A., Rubio, Á., & Räsänen, E. (2014) · Physical Review A · doi:10.1103/physreva.90.053402 · Cited by 25
Göckler, T., & Grimm, V. (2014) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/140964655 · Cited by 25
“[20]), but for only slightly larger, mildly stiff matrices, the method deteriorates and for stiff matrices an efficient approximation is not possible at all.”
BackgroundMatrices, moments, and rational quadrature
Lagomasino, G. L., Reichel, L., & Wunderlich, L. (2008) · Linear Algebra and its Applications · doi:10.1016/j.laa.2008.04.047 · Cited by 25
“Bounds for the approximation error introduced by replacing a non-rational function f by a rational one are discussed in, e.g., [3, 10, 22].”
BackgroundArray-representation integration factor method for high-dimensional systems
Wang, D., Zhang, L., & Nie, Q. H. (2013) · Journal of Computational Physics · doi:10.1016/j.jcp.2013.11.002 · Cited by 25
“Even in a cluster where computing a 32 3 × 32 3 matrix exponential is possible, the CPU time needed will be about 2 hours, and computation of a 64 3 × 64 3 matrix exponential takes more than a day On the other hand, AcIIF runs normally with good accuracy, showing clear advantages in handling larger…”
Ju, L., Liu, X., & Leng, W. (2014) · Discrete and Continuous Dynamical Systems - B · doi:10.3934/dcdsb.2014.19.1667 · Cited by 25
“The flexibility of compact representation allows either direct calculation of the exponentials of matrices or using Krylov subspace [4, 10, 22] for non-constant diffusion coefficients to compute their exponential matrix-vector multiplications for saving further in storages and cost.”
MethodA literature survey of matrix methods for data science
Stoll, M. (2020) · GAMM-Mitteilungen · doi:10.1002/gamm.202000013 · Cited by 30
“Often the main work goes into solving linear systems with (zI − A) and once direct solvers become infeasible approaches based on Krylov subspace methods have shown very good performance for approximating the evaluation of f(A)b [139, 159, 85, 119, 159].”
MethodRate of convergence of local linearization schemes for initial-value problems
Jiménez, J. C., & Carbonell, F. (2005) · Applied Mathematics and Computation · doi:10.1016/j.amc.2005.01.118 · Cited by 27
“Examples are the approximations based on the Schur decomposition [29], the rational Pade approximations [30] and the Krylov subspace methods [30].”
MethodJiménez, J. C., & de la Cruz, H. (2011) · BIT Numerical Mathematics · doi:10.1007/s10543-011-0360-2 · Cited by 28
Analysis of Krylov subspace approximation to large-scale differential Riccati equations
Koskela, A., & Mena, H. (2020) · ETNA - Electronic Transactions on Numerical Analysis · doi:10.1553/etna_vol52s431 · Cited by 26
“4.1, similarly to the bounds given in [17], exhibits a hump before it starts to decrease in case ktAk > 1. Improved bounds for special cases of the matrix A are possible by using, e.g., results of [23]. 4.2. Refined error bounds for the Lyapunov equation. Although Theorem 4.1 shows the superlinear convergence speed…”
InfluentialExcitations in two-component Bose gases
Kleine, A., Kollath, C., McCulloch, I. P., Giamarchi, T., & Schollwöck, U. (2008) · New Journal of Physics · doi:10.1088/1367-2630/10/4/045025 · Cited by 29
XDiag: Exact diagonalization for quantum many-body systems
Wietek, A., Staszewski, L., Ulaga, M., Ebert, P. L., Karlsson, H., Sarkar, S., Shackleton, H., Sinha, A., & Soares, R. D. (2026) · SciPost Physics Codebases · doi:10.21468/scipostphyscodeb.70 · Cited by 32
“[ 53 ] .”
Wang, Y., Shi, T., & Chen, C.-C. (2021) · Physical Review X · doi:10.1103/physrevx.11.041028 · Cited by 29
“The error of evaluating the vector propagation is well controlled by ϵ n ≤ 12 e − ð ρδ t Þ 2 = 16 n ð e ρδ t= 4 n Þ n given the Krylov dimension n ≥ ρδ t= 2 and spectral radius ρ ¼ j E max − E min j [119].”
Auxiliary Hamiltonian representation of the nonequilibrium Dyson equation
Balzer, K., & Eckstein, M. (2014) · Physical Review B · doi:10.1103/physrevb.89.035148 · Cited by 24
Error Estimates for Polynomial Krylov Approximations to Matrix Functions
Diele, F., Moret, I., & Ragni, S. (2008) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/070688924 · Cited by 24
The restarted shift-and-invert Krylov method for matrix functions
Moret, I., & Popolizio, M. (2012) · Numerical Linear Algebra with Applications · doi:10.1002/nla.1862 · Cited by 24
“/, convex, we recall a general result of [9] that says m 1,k.”
BackgroundSemi-Lagrangian Exponential Integration with Application to the Rotating Shallow Water Equations
Peixoto, P. S., & Schreiber, M. (2019) · SIAM Journal on Scientific Computing · doi:10.1137/18m1206497 · Cited by 24
“e exponentials exist (see [29]), two approaches are currently most commonly researched in this context: Krylov subspace solvers and rational approximations. Krylov solvers, such as those presented in [28], are used in [15] and [25] for the matrix exponentiation of a dynamic linearization of the shallow water system. F…”
BackgroundShifted GMRES for oscillatory integrals
Olver, S. (2009) · Numerische Mathematik · doi:10.1007/s00211-009-0264-0 · Cited by 28
Fast inexact subspace iteration for generalized eigenvalue problems with spectral transformation
Xue, F., & Elman, H. C. (2010) · Linear Algebra and its Applications · doi:10.1016/j.laa.2010.06.021 · Cited by 24
“For details about Ca and Cb, see [28], [19] and [12].”
BackgroundCovariance Matrix Estimation via Network Structure
Lan, W., Fang, Z., Wang, H., & Tsai, C. (2017) · Journal of Business and Economic Statistics · doi:10.1080/07350015.2016.1173558 · Cited by 27
“Moreover, model (2.3) can be viewed as a matrix polynomial with order d (see Dennis, Traub, and Weber 1976 and Hochbruck and Lubich 1997).”
InfluentialKrylov precise time‐step integration method
Fung, T. C., & Chen, Z. (2006) · International Journal for Numerical Methods in Engineering · doi:10.1002/nme.1737 · Cited by 24
“As mentioned in References [22, 23], the main idea of the Krylov subspace method is to approximate the matrix functions (A · t){v} by ([A] · t){v} ≈ [Vm] · ([Hm] · t){e1} (23) where {e1} = [1, 0, . . . , 0]T.”
MethodThe generalized Kadanoff-Baym ansatz. Computing nonlinear response properties of finite systemsPreprint
Balzer, K., Hermanns, S., & Bönitz, M. (2016) · Cited by 24
Krylov-subspace approach for the efficient control of quantum many-body dynamics
Larocca, M., & Wisniacki, D. A. (2021) · Physical Review A · doi:10.1103/physreva.103.023107 · Cited by 26
Nan, C., & Song, H. (2022) · Journal of Computational Physics · doi:10.1016/j.jcp.2022.111028 · Cited by 23
Enhancement of Krylov Subspace Spectral Methods by Block Lanczos Iteration
Lambers, J. V. (2008) · AIP conference proceedings · doi:10.1063/1.2990930 · Cited by 23
“Unless grid refinement is used during the Lanczos process, such error affects the recursion coefficients, which in turn affect the quadrature nodes and weights, and therefore the accuracy of the temporal approximation of each component.”
MethodGöckler, T., & Grimm, V. (2013) · SIAM Journal on Numerical Analysis · doi:10.1137/12089226x · Cited by 23
“The main motivation to examine rational Krylov subspace methods is the favorable property that the convergence of these methods does not deteriorate for matrices with large norm in contrast to the standard Krylov subspace method (e.g. [16], [25]).”
Analyzing the positivity preservation of numerical methods for the Liouville-von Neumann equation
Riesch, M., & Jirauschek, C. (2019) · Journal of Computational Physics · doi:10.1016/j.jcp.2019.04.006 · Cited by 28
“The algorithm bases on Krylov subspace methods [28,29] like the approaches presented in [30,31].”
MethodError Bounds for the Krylov Subspace Methods for Computations of Matrix Exponentials
Wang, H., & Ye, Q. (2017) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/16m1063733 · Cited by 27
“We observe that when τ is relatively small, our new a priori bound is comparable to Hochbruch and Lubich’s bound, but as τ increases, our bound improves significantly.”
BackgroundMethodResultInfluentialApproximate dynamics leading to more optimal control: Efficient exact derivatives
Jensen, J. H. M., Møller, F., Sørensen, J. J., & Sherson, J. F. (2021) · Physical Review A · doi:10.1103/physreva.103.062612 · Cited by 28
“— Another common way of approximating the time evolution for extended applicability is through the use of Krylov-Lanczos subspace methods [61, 74, 75] where exponential operator applications are performed without explicit construction.”
BackgroundA polynomial method based on Fejèr points for the computation of functions of unsymmetric matrices
Novati, P. (2002) · Applied Numerical Mathematics · doi:10.1016/s0168-9274(02)00139-3 · Cited by 25
“Concerning the rate of convergence, KPMs generally show a very satisfactory behavior (see [13]), but they also present the typical disadvantages of the projective schemes, that is, the construction of the projection subspaces.”
BackgroundMethodGavnholt, J., Rubio, Á., Olsen, T., Thygesen, K. S., & Schiøtz, J. (2009) · Physical Review B · doi:10.1103/physrevb.79.195405 · Cited by 23
Borisov, A. G., & Shabanov, S. V. (2005) · Journal of Computational Physics · doi:10.1016/j.jcp.2005.03.034 · Cited by 23
Zimmerling, J., Wei, L., Urbach, H. P., & Remis, R. (2016) · Journal of Computational Physics · doi:10.1016/j.jcp.2016.03.057 · Cited by 23
“Finally, we mention that it is possible to give a general error bound for polynomial Krylov methods (see, for example, [15]), but such a bound is usually not tight and leads to a gross overestimation of the required number of iterations.”
BackgroundThe ReLPM Exponential Integrator for FE Discretizations of Advection-Diffusion Equations
Bergamaschi, L., Caliari, M., & Vianello, M. (2004) · Lecture notes in computer science · doi:10.1007/978-3-540-25944-2_57 · Cited by 25
Error Bounds for the Lanczos Methods for Approximating Matrix Exponentials
Ye, Q. (2013) · SIAM Journal on Numerical Analysis · doi:10.1137/11085935x · Cited by 22
“Several more refined bounds have been obtained in [9, 10, 15] but they all suggest similar dependence of convergence on τ‖A‖.”
Using the Restricted-denominator Rational Arnoldi Method for Exponential Integrators
Novati, P. (2011) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/100814202 · Cited by 22
“This technique has been used, for instance, in [22] and [13].”
BackgroundMethodFast exponential time integration scheme for option pricing with jumps
Lee, S. T., Liu, X., & Sun, H. (2010) · Numerical Linear Algebra with Applications · doi:10.1002/nla.749 · Cited by 22
“Unfortunately, past analysis [10] showed that the iterative number m could get quite large when ‖An‖2 increases.”
BackgroundMethodA Krylov‐subspace based solver for the linear and nonlinear Maxwell equations
Busch, K., Niegemann, J., Pototschnig, M., & Tkeshelashvili, L. (2007) · physica status solidi (b) · doi:10.1002/pssb.200743290 · Cited by 25
Time-accurate and highly-stable explicit peer methods for stiff differential problems
Conte, D., Pagano, G., & Paternoster, B. (2023) · Communications in Nonlinear Science and Numerical Simulation · doi:10.1016/j.cnsns.2023.107136 · Cited by 21
Many-Body Majorana Braiding without an Exponential Hilbert Space
Mascot, E., Hodge, T., Crawford, D., Bedow, J., Morr, D. K., & Rachel, S. (2023) · Physical Review Letters · doi:10.1103/physrevlett.131.176601 · Cited by 31
Exponential integrators in time-dependent density-functional calculations
Kidd, D., Covington, C., & Varga, K. (2017) · Physical review. E · doi:10.1103/physreve.96.063307 · Cited by 21
de la Cruz, H., Biscay, R., Jiménez, J. C., & Carbonell, F. (2012) · Mathematical and Computer Modelling · doi:10.1016/j.mcm.2012.08.011 · Cited by 24
“This have become feasible due to advances in the computation of matrix exponentials (see, e.g., [26], [56], [17], [12], [25]) and multiple integrals involving matrix exponentials (see, e.g., [9], [59]).”
Background
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.