Abstract
Let P ∈ C[z] be a polynomial which has a cluster of k zeros near v 0 ∈ C. The coefficients of the factor F of P corresponding to this root cluster can be computed from the coefficients of P by Newton iteration in C k, applied to the mapping F ↦ P mod F.
This paper presents and analyzes a numerical algorithm based on this idea. The algorithm uses only polynomial multiplication and division. In particular, there is no need to solve linear systems.
The analysis results in an explicit quantitative condition for the coeficcients of P which guarantees the algorithm to converge when started with P 0 = (z — v 0)k. Efficient test procedures for this condition are described. Finally, another starting value condition in terms of root sizes is proved and illustrated with some examples.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Bini, D. and Pan, V.Polynomial and Matrix Computations, Volume 1: Fundamental Algorithms. Birkhäuser, Boston, 1994.
Friedman, J. On the convergence of Newton’s method.J. of Complexity 5(1989), 12 – 33.
Gourdon, X. Algorithmique du théorème fondamental de l’algèbre. Rapport de Recherche 1852, INRIA, February 1993.
Gourdon, X., November 1995. Personal communication.
Grau, A. A. The simultaneous improvement of a complete set of approximate factors of a polynomial.SIAM J. Numer. Analysis 8(1971), 425 – 438.
Henrici, P.Applied and Computational Complex Analysis, vol.1. Wiley & Sons, New York, 1974.
Jenkins, M. A. and Traub, J. F. A three-stage variable-shift iteration for polynomial zeros and its relation to generalized rayleigh iteration.Numer. Math.14 (1970), 252 – 263.
Kantorovich, L. and Akilov, G.Functional Analysis in Normed Spaces. MacMillan, New York, 1964.
Kim, M.-H. On approximate zeros and rootfinding algorithms for a complex polynomial.Math. Comp.51 (1988), 707 – 719.
Kirrinnis, P.Zur Berechnung von Partialbruchzerlegungen und Kurvenintegralen rationaler Funktionen. Dissertation, Univ. Bonn, Institut für Informatik, January 1993.
Kirrinnis, P. Partial fraction decomposition in C(z) and simultaneous Newton iteration for factorization in C[z]. Preprint, Univ. Bonn. Submitted for publication, September 1995.
Lehmer, D. H. A machine method for solving polynomial equations.J. Assoc. Comput. Mach. 8(1961), 151 – 162.
Malajovich, G. and Zubelli, J. P. A fast and stable algorithm for splitting polynomials. Preprint, Dept. de Matematica Aplicada, Univ. Federal do Rio de Janeiro, January 1996.
Neff, C. A. and Reif, J. H. An 0(n1+e logb) algorithm for the complex root problem. InProc. 35th Ann. IEEE Symp. on Foundations of Computer Science(1994), pp. 540 – 547.
Pan, V. Y. Sequential and parallel complexity of approximate evaluation of polynomial zeros.Comput. Math. Appl. 14(1987), 591 – 622.
Pan, V. Y. New techniques for approximating complex polynomial zeros. InProc. 5th Ann. ACM-SIAM Symp. on Discrete Algorithms(1994), pp. 260 – 270.
Pan, V. Y. Optimal (up to polylog factors) sequential and parallel algorithms for approximating complex polynomial zeros. InProc. 27th Ann. ACM Symp. on the Theory of Computing (STOC’95)(1995), pp. 741 – 750.
Renegar, J. On the cost of approximating all roots of a complex polynomial.Math. Programming 32(1985), 319 – 336.
Renegar, J. On the worst-case complexity of approximating zeros of polynomials.J. of Complexity 3(1987), 90 – 113.
Schänhage, A. Asymptotically fast algorithms for the numerical multiplication and division of polynomials with complex coefficients. InComputer Algebra EU-ROCAM ’82 (Marseille 1982) (1982), J. Calmet, Ed., vol. 144 ofSpringer Lect. Notes in Computer Science, Springer, pp. 3–15.
Schönhage, A. The fundamental theorem of algebra in terms of computational complexity. Tech. rep., Univ. Tübingen, 1982.
Schönhage, A. Equation solving in terms of computational complexity. InProc. Int. Congress of Mathematicians, Berkeley, CA(1986), pp. 131 – 153.
Schönhage, A. Ordinary root finding. Preprint, Univ. Bonn, January 1996.
Schönhage, A., Grotefeld, A. F. W., and Vetter, E.Fast Algorithms: A Multitape Turing Machine Implementation. B.I. Wissenschafts verlag, Mannheim, 1994.
Shub, M. and Smale, S. Computational complexity: On the geometry of polynomials and a theory of cost: Part I.Ann. scient. Ec. Norm. Sup., 4e série 18(1985), 107 – 142.
Shub, M. and smale, S. Computational complexity: On the geometry of polynomials and a theory of cost: II.SIAM J. Comput. 15(1986), 145 – 161.
.Smale, S. The fundamental theorem of algebra and complexity theory. Bull. (New Series)Amer. Math. Soc.4(1981), 1 – 36.
Smale, S. Newton’s method estimates from data at one point. InThe Merging of Disciplines: New Directions in Pure, Applied, and Computational Mathematics(1985), R. E. Ewing, K. I. Gross, and C. F. Martin, Eds., Springer, pp. 185 – 196.
Smale, S. On the efficiency of algorithms of analysis.Bull. (New Series) Amer. Math. Soc. 13(1985), 87 – 121.
Smale, S. Algorithms for solving equations. InProc. Int. Congress of Mathematicians, Berkeley, CA(1986), pp. 172 – 195.
Weyl, H. Randbemerkungen zu Hauptproblemen der Mathematik. II. Fundamentalsatz der Algebra und Grundlagen der Mathematik.Math. Zeitschr. 20(1924), 142 – 150.
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1997 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Kirrinnis, P. (1997). Newton Iteration Towards a Cluster of Polynomial Zeros. In: Cucker, F., Shub, M. (eds) Foundations of Computational Mathematics. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-60539-0_15
Download citation
DOI: https://doi.org/10.1007/978-3-642-60539-0_15
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-61647-4
Online ISBN: 978-3-642-60539-0
eBook Packages: Springer Book Archive