Remote Access Mathematics of Computation
Green Open Access

Mathematics of Computation

ISSN 1088-6842(online) ISSN 0025-5718(print)



Improved formulas for complete and partial summation of certain series.

Authors: Herbert E. Salzer and Genevieve M. Kimbro
Journal: Math. Comp. 15 (1961), 23-39
MSC: Primary 65.00
MathSciNet review: 0121972
Full-text PDF Free Access

Abstract | References | Similar Articles | Additional Information

Abstract: In two previous articles one of the authors gave formulas, with numerous examples, for summing a series either to infinity (complete) or up to a certain number n of terms (partial) by considering the sum of the first j terms $ {S_j}$, or some suitable modification $ {S'_j}$, closely related to $ {S_j}$, as a polynomial in 1/j. Either $ {S_\infty }$ or $ {S_n}$ was found by m-point Lagrangian extrapolation from $ {{S}_{{{j}_{0}}}}$, $ {{S}_{{{j}_{0}}-1}}$, $ \cdot \cdot \cdot $, $ {{S}_{{{j}_{0}}-m+1}}$ to 1/j = 0 or 1/j = 1/n respectively. This present paper furnishes more accurate m-point formulas for sums (or sequences) $ {S_j}$ which behave as even functions of 1/j. Tables of Lagrangian extrapolation coefficients in the variable $ 1/{j^2}$ are given for: complete summation, m = 2(1)7, $ {j_0}$ = 10, exactly and 20D, m = 11, $ {j_0}$ = 20, 30D; partial summation, m = 7, $ {j_0}$ = 10, n = 11(1)25(5)100, 200, 500, 1000, exactly. Applications are made to calculating $ \pi $ or the semi-perimeters of many-sided regular polygons, Euler's constant,

$\displaystyle 1 + \sum\limits_{r = 1}^j {\left\{ {\tfrac{{ - 1}}{{{{(4r - 1)}^2... ... \tfrac{1}{{{3^2}}} + \tfrac{1}{{{5^2}}} - \cdots } \,{\text{for}}\,j = \infty $

(Catalan's constant), calculation of a definite integral as the limit of a suitably chosen sequence, determining later zeros of $ {J_v}(x)$ from earlier zeros for suitable v, etc. A useful device in many cases involving sums of odd functions, is to replace $ {S_j}$ by a trapezoidal-type $ {S'_j}$ which is seen, from the Euler-Maclaurin formula, to be formally a series in $ 1/{j^2}$. In almost every example, comparison with the earlier (1/j)-extrapolation showed these present formulas, for 7 points, to improve results by anywhere from around 4 to 9 places.

References [Enhancements On Off] (What's this?)

  • [1] H. E. Salzer, ``A simple method for summing certain slowly convergent series,'' J. Math. Phys., v. 33, 1955, p. 356-359. MR 0068315 (16:863d)
  • [2] H. E. Salzer, ``Formulas for the partial summation of series,'' MTAC, v. 10, 1956, p. 149-156. MR 0081528 (18:416a)
  • [3] L. F. Richardson, ``The approximate arithmetical solution by finite differences of physical problems involving differential equations, with an application to the stresses in a masonry dam,'' Philos. Trans. Roy. Soc. London, Ser. A, v. 210, 1910, p. 307-357.
  • [4] L. F. Richardson & J. A. Gaunt, ``The deferred approach to the limit,'' Philos. Trans. Roy. Soc. London, Ser. A, v. 226, 1927, p. 299-361.
  • [5] G. Blanch, ``On the numerical solution of parabolic partial differential equations,'' J. Res. Nat. Bur. Standards, v. 50, 1953, p. 343-356. MR 0059078 (15:474b)
  • [6] H. C. Bolton & H. I. Scoins, ``Eigenvalues of differential equations by finite-difference methods,'' Proc. Cambridge Philos. Soc., v. 52, pt. 2, 1956, p. 215-229. MR 0079344 (18:72d)
  • [7] M. G. Salvadori, ``Extrapolation formulas in linear difference operators,'' Proc. First U. S. Nat. Congress of Applied Mech., Edwards Bros., 1952, p. 15-18. MR 0060911 (15:747a)
  • [8] Herrmann, ``Bestimmung der trigonometrischen Functionen aus den Winkeln und der Winkel aus den Functionen, bis zu einer beliebigen Grenze der Genauigkeit,'' Kaiserliche Akademie der Wissenschaften, Wien, Mathematisch-naturwissenschaftliche Classe, Sitzungsberichte, 1, pt. IV, 1848, p. 176-177. Reprinted in Table of Sines and Cosines to Fifteen Decimal Places at Hundredths of a Degree, Nat. Bur. Standards. Appl. Math. Ser. No. 5, 1949, p. 92-93. MR 0030289 (10:740d)
  • [9] E. Whittaker & G. Robinson, The Calculus of Observations, 4th edition, Blackie and Son, London, 1954, p. 135.
  • [10] H. T. Davis, Tables of the Higher Mathematical Functions, v. II, Principia Press, Bloomington, Indiana, 1935, p. 282, 284-285, 304. Davis cites the earlier work of J. W. L. Glaisher in relation to Catalan's constant in Mess. of Math., v. 6, 1876, p. 71-76, Proc. London Math. Soc., v. 8, 1877, p. 200-201, Mess. of Math., v. 42, 1913, p. 35-58.
  • [11] G. N. Watson, Theory of Bessel Functions, 2nd edition, Cambridge University Press, 1952, p. 506.
  • [12] NBS, Tables of Spherical Bessel Functions, v. II, New York, Columbia University Press, 1947, p. 318.
  • [13] W. G. Bickley & J. C. P. Miller, ``Notes on the evaluation of zeros and turning values of Bessel functions, II, The McMahon Series,'' Phil. Mag., s. 7, v. 36, 1945, p. 125. MR 0012903 (7:82d)

Similar Articles

Retrieve articles in Mathematics of Computation with MSC: 65.00

Retrieve articles in all journals with MSC: 65.00

Additional Information

Article copyright: © Copyright 1961 American Mathematical Society

American Mathematical Society