A density matrix approach to the convergence of the self-consistent field iteration

Parikshit Upadhyaya; Elias Jarlebring; Emanuel H. Rubensson

doi:10.3934/naco.2020018

Article Contents

2021, Volume 11, Issue 1: 99-115. Doi: 10.3934/naco.2020018

This issue Previous Article Improving whale optimization algorithm for feature selection with a time-varying transfer function Next Article A PID control method based on optimal control strategy

A density matrix approach to the convergence of the self-consistent field iteration

1.
Lindstedtsvägen 25, Department of Mathematics, SeRC - Swedish e-Science research center, Royal Institute of Technology, SE-11428 Stockholm, Sweden
2.
Division of Scientific Computing, Department of Information Technology, , Uppsala University, Box 337, SE-75105 Uppsala, Sweden

^* Corresponding author
^* Corresponding author

Received: October 31, 2018

Revised: November 30, 2019

Early access: March 2020

Published: March 2021

Abstract Full Text(HTML) Figure(7) Related Papers Cited by

Abstract

In this paper, we present a local convergence analysis of the self-consistent field (SCF) iteration using the density matrix as the state of a fixed-point iteration. Conditions for local convergence are formulated in terms of the spectral radius of the Jacobian of a fixed-point map. The relationship between convergence and certain properties of the problem is explored by deriving upper bounds expressed in terms of higher gaps. This gives more information regarding how the gaps between eigenvalues of the problem affect the convergence, and hence these bounds are more insightful on the convergence behaviour than standard convergence results. We also provide a detailed analysis to describe the difference between the bounds and the exact convergence factor for an illustrative example. Finally we present numerical examples and compare the exact value of the convergence factor with the observed behaviour of SCF, along with our new bounds and the characterization using the higher gaps. We provide heuristic convergence factor estimates in situations where the bounds fail to well capture the convergence.

Keywords:

Mathematics Subject Classification: Primary: 65F15, 65H17.

Citation:

Full Text(HTML)

Figure 1. Schematic illustration of elements of $ \Omega_3 $ as indices of $ R $ for the real-valued problem in subsection 4.1 with $ n = 7, p = 3, \alpha = 10.0 $

Download: Full-size image PowerPoint slide

Figure 2. Convergence factor and bounds for the illustrative example

Download: Full-size image PowerPoint slide

Figure 3. Norm of $ \mathcal{L}(x_1x_2^H) $ and $ \mathcal{L}(x_2x_3^H) $

Download: Full-size image PowerPoint slide

Figure 4. Variance of $ \delta_1 $ and $ \delta_2 $ with $ \epsilon $

Download: Full-size image PowerPoint slide

Figure 5. Complex-valued problem for $ n = 30 $ ((a),(b) and (d)), $ p = 15, \alpha = 40.0 $ ((a),(b) and (c))

Download: Full-size image PowerPoint slide

Figure 6. Real-valued problem for $ n = 60 $, $ \alpha = 5.0 $, $ p = 25 $

Download: Full-size image PowerPoint slide

Figure 7. Water molecule problem with $ n = 13, p = 5 $

Download: Full-size image PowerPoint slide

Related Papers

Cited by

References

[1]	Z. Bai, D. Lu and B. Vandereycken, Robust Rayleigh quotient minimization and nonlinear eigenvalue problems, , SIAM J. Sci. Comput., 40 (2018), A3495–A3522. doi: 10.1137/18M1167681.
[2]	D. R. Bowler and T. Mizayaki, $\mathcal{O}$($n$) methods in electronic structure calculations, , Rep. Prog. Phys., 75 (2012), 036503, http://iopscience.iop.org/article/10.1088/0034-4885/75/3/036503. doi: 10.1088/0034-4885/75/3/036503.
[3]	Y. Cai, L.-H. Zhang, Z. Bai and R.-C. Li, On an eigenvector-dependent nonlinear eigenvalue problem, SIAM J. Matrix Anal. Appl., 39 (2018), 1360-1382. doi: 10.1137/17M115935X.
[4]	E. Cancés and C. L. Bris, On the convergence of SCF algorithms for the Hartree-Fock equations, M2AN, Math. Model. Numer. Anal., 34 (2000), 749-774. doi: 10.1051/m2an:2000102.
[5]	T. Helgaker, P. Jorgensen and J. Olsen, Molecular Electronic-Structure Theory, John Wiley and Sons, 2000. doi: 10.1002/9781119019572.
[6]	H. V. Henderson and S. R. Searle, Vec and vech operators for matrices, with some uses in jacobians and multivariate statistics, Can J. Stat., 7 (1979), 65-81. doi: 10.2307/3315017.
[7]	N. Higham, Functions of Matrices, Society for Industrial and Applied Mathematics, 2008. doi: 10.1137/1.9780898717778.
[8]	T. Kato, Perturbation Theory for Linear Operators, 2nd edition, Springer-Verlag, 1995. doi: 10.1007/978-3-642-66282-9.
[9]	A. Levitt, Convergence of gradient-based algorithms for the Hartree-Fock equations, ESAIM: Math. Model. Numer. Anal., 46 (2012), 1321-1336. doi: 10.1051/m2an/2012008.
[10]	X. Liu, X. Wang, Z. Wen and Y. Yuan, On the convergence of the self-consistent field iteration in Kohn-Sham density functional theory, SIAM J. Matrix Anal. Appl., 35 (2014), 546-558. doi: 10.1137/130911032.
[11]	X. Liu, Z. Wen, X. Wang, M. Ulbrich and Y. Yuan, On the analysis of the discretized Kohn-Sham density functional theory, SIAM J. Numer. Anal., 53 (2015), 1758-1785. doi: 10.1137/140957962.
[12]	C. E. McCulloch, Symmetric matrix derivatives with applications, J. Amer. Stat. Assoc., 77 (1982), 679-682. doi: 10.2307/2287736.
[13]	A. Messiah, Quantum Mechanics, Dover Publications, 1999.
[14]	T. T. Ngo, M. Bellalij and Y. Saad, The trace ratio optimization problem for dimensionality reduction, SIAM J. Matrix Anal. Appl., 31 (2010), 2950-2971. doi: 10.1137/090776603.
[15]	P. Pulay, Convergence acceleration of iterative sequences. The case of scf iteration, Chem. Phys. Lett., 73 (1979), 393-398. doi: 10.1016/0009-2614(80)80396-4.
[16]	T. Rohwedder and R. Schneider, An analysis for the DIIS acceleration method used in quantum chemistry calculations, J. Math. Chem., 49 (2011), 1889-1914. doi: 10.1007/s10910-011-9863-y.
[17]	E. Rudberg, Quantum Chemistry for Large Scale Systems, PhD thesis, Royal Institute of Technology, 2007, Available at http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-4561.
[18]	E. Rudberg, E. H. Rubensson and P. Saƚek, Kohn-Sham density functional theory electronic structure calculations with linearly scaling computational time and memory usage, J. Chem. Theory Comput., 7 (2011), 340-350. doi: 10.1021/ct100611z.
[19]	E. Rudberg, E. H. Rubensson, P. Saƚek and A. Kruchinina, Ergo: An open-source program for linear-scaling electronic structure calculations, SoftwareX, 7 (2018), 107-111. doi: 10.1016/j.softx.2018.03.005.
[20]	Y. Saad, J. T. Chelikowsky and S. M. Shontz, Numerical methods for electronic structure calculations of materials, SIAM Rev., 52 (2010), 3-54. doi: 10.1137/060651653.
[21]	R. E. Stanton, Intrinsic convergence in closed-shell SCF calculations. A general criterion, J. Chem. Phys., 75 (1981), 5416-5422. doi: 10.1063/1.441942.
[22]	A. Szabo and N. S. Ostlund, Modern Quantum Chemistry: Introduction to Advanced Electronic Structure Theory, Dover Publications, 1996.
[23]	C. Yang, W. Gao and J. C. Meza, On the convergence of the self-consistent field iteration for a class of nonlinear eigenvalue problems, SIAM J. Matrix Anal. Appl., 30 (2009), 1773-1788. doi: 10.1137/080716293.
[24]	L.-H. Zhang, L.-Z. Liao and M. K. Ng, Fast algorithms for the generalized Foley Sammon discriminant analysis, SIAM J. Matrix Anal. Appl., 31 (2010), 1584-1605. doi: 10.1137/080720863.
[25]	L.-H. Zhang, W. H. Yang and L.-Z. Liao, A note on the trace quotient problem, Opt. Lett., 8 (2014), 1637-1645. doi: 10.1007/s11590-013-0680-z.