Abstract
We present an inherently parallel method using a Taylor polynomial as an estimator for several numerical examples of strictly diagonally dominant and banded triangular Toeplitz systems. We demonstrate the effectiveness of the method using multiple precisions for each of four test cases, and compare the results versus classical Jacobi and Newton methods. We also show that the residual norms can be driven down to any arbitrary precision with less hardware requirements than existing methods. Finally, we use truncated Taylor series as initial estimates for the Newton, and in each case, we obtain equal or better accuracy than the Newton method alone, while using less vector computations.
Similar content being viewed by others
Notes
An FFT is not required. The result of this operation is a column of 1's.
An FFT is not required. The result of this operation is a column of 1's.
An FFT is not required. The result of this operation is a column of 1's.
An FFT is not required. The result of this operation is a column of 1's.
An FFT is not required. The result of this operation is a column of 1's.
References
Advanpix L Multiprecision computing toolbox for matlab, yokohama, japan, 2008–2020
Belhaj S, Dridi M (2014) A note on computing the inverse of a triangular toeplitz matrix. Appl Math Comput 236:512–523
Bini D (1984) Parallel solution of certain Toeplitz linear systems. SIAM J Comput 13(2):268–276
Bini D, Pan V (1986) Polynomial division and its computational complexity. J Complex 2(3):179–203. https://doi.org/10.1016/0885-064X(86)90001-4
Bini D, Pan V (1993) Improved parallel polynomial division. SIAM J Comput 22(3):617–626
Bini D, Pan V (1993) Parallel computations with toeplitz-like and hankel-like matrices. In: Proceedings of the 1993 International Symposium on Symbolic and Algebraic Computation, pp 193–200. ACM (1993)
Bini D, Pan VY (1994) Polynomial and matrix computations. Volume 1: Fundamental algorithms. Progress in Theoretical Computer Science. Birkhäuser Verlag, Boston-Basel-Berlin-Stuttgart
Bini DA, Codevico G, Van Barel M (2003) Solving toeplitz least squares problems by means of newton’s iteration. Numerical Algorithms 33(1–4):93–103
Bini DA, Meini B (1999) Effective methods for solving banded toeplitz systems. SIAM J Matrix Anal Appl 20(3):700–719
Bini DA, Meini B (2001) Approximate displacement rank and applications. Contemp Math 281:215–232
Bini DA, Meini B (2009) The cyclic reduction algorithm: from poisson equation to stochastic processes and beyond. Numer Algorithms 51(1):23–60
Böttcher A, Halwass M (2013) Wiener-hopf and spectral factorization of real polynomials by newton’s method. Linear Algebra Appl 438(12):4760–4805
Chow E, Anzt H, Scott J, Dongarra J (2018) Using jacobi iterations and blocking for solving sparse triangular systems in incomplete factorization preconditioning. J Parallel Distrib Comput 119:219–230
Commenges D, Monsion M (1984) Fast inversion of triangular toeplitz matrices. IEEE Trans Autom Control 29:250–251
Cooley JW, Tukey JW (1965) An algorithm for the machine calculation of complex fourier series. Math Comput 19(90):297–301
Grcar J, Sameh A (1981) On certain parallel toeplitz linear system solvers. SIAM J Sci Stat Comput 2(2):238–256
van der Hoeven J (2010) Newton’s method and fft trading. J Symb Comput 45(8):857–878
Huang J, Huang TZ, Belhaj S (2013) Scaling bini’s algorithm for fast inversion of triangular toeplitz matrices. J Comput Anal Appl 15(5):858–867
Isaacson E, Keller HB (1966) Analysis of numerical methods. Wiley, United States
Kapralos (2021) Estimator. Internet Website, Last accessed 10/21/2021. Https://github.com/BDDTT/Estimator
Kapralos M, Wolinetz A, Murphy BJ (2016) Parallel solution of diagonally dominant banded triangular toeplitz systems using taylor polynomials. In: 18th IEEE International Conference on High Performance Computing and Communications; 14th IEEE International Conference on Smart City; 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016, Sydney, Australia, December 12-14, 2016, pp 601–607
Larriba-Pey JL, Navarro JJ, Jorba A, Roig O (1996) Review of general and toeplitz vector bidiagonal solvers. Parallel Comput 22(8):1091–1126
Lin FR, Ching WK, Ng MK (2004) Fast inversion of triangular toeplitz matrices. Theoret Comput Sci 315(2):511–523
Lv XG, Huang TZ, Le J (2008) A note on computing the inverse and the determinant of a pentadiagonal toeplitz matrix. Appl Math Comput 206(1):327–331
Malyshev A, Sadkane M (2014) Fast solution of unsymmetric banded toeplitz systems by means of spectral factorizations and woodbury’s formula. Numer Linear Algebra Appl 21(1):13–23
McNally JM, Garey L, Shaw R (2008) A communication-less parallel algorithm for tridiagonal toeplitz systems. J Comput Appl Math 212(2):260–271
Murphy BJ (2011) Acceleration of the inversion of triangular toeplitz matrices and polynomial division. Comput Algebra Sci Comput: Lect Notes Comput Sci 6885:321–332
Pan VY, Branham S, Rosholt RE, Zheng AL (1999) Newton’s iteration for structured matrices. In: Fast reliable algorithms for matrices with structure, pp. 189–210. SIAM
Pan VY, Rami Y, Wang X (2002) Structured matrices and newton’s iteration: unified approach. Linear Algebra Appl 343:233–265
Schönhage A (2000) Variations on computing reciprocals of power series. Inf Process Lett 74:41–46
Sieveking M (1972) An algorithm for division of power series. Computing 10(1–2):153–156
Stpiczyński P (2015) Fast solver for toeplitz bidiagonal systems of linear equations. Ann Univ Mariae Curie-Sklodowska, sectio AI-Inform 1(1):1–7
Sun XH, Moitra S (1996) A fast parallel tridiagonal algorithm for a class of CFD applications. NASA Langley Technical Report Server
Trench WF (1964) An algorithm for the inversion of finite toeplitz matrices. J Soc Ind Appl Math 12(3):515–522
Trench WF (1974) Inversion of toeplitz band matrices. Math Comput 28(128):1089–1095
Trench WF (1985) Explicit inversion formulas for toeplitz band matrices. SIAM J Algebr Discrete Methods 6(4):546–554
Trench WF (1987) A note on solving nearly triangular toeplitz systems. Linear Algebra Appl 93:57–65
Wang X, Huang Y (2012) A fast algorithm for inversion of real lower triangular toeplitz matrices. J Comput Anal Appl 14(1)
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Kapralos, M. Taylor polynomials as an estimator for certain toeplitz matrices. J Supercomput 78, 13245–13275 (2022). https://doi.org/10.1007/s11227-022-04373-y
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11227-022-04373-y