Abstract
Equally-weighted formulas for numerical differentiation at a fixed pointx=a, which may be chosen to be 0 without loss in generality, are derived for (1)\(f^{(m)} (0) = k\left\{ {\sum\limits_{i = n + 1}^{2n} f (x_i ) - \sum\limits_{i = 1}^n f (x_i )} \right\} + R_{2n} \) whereR 2n =0 whenf(x) is any (2n)th degree polynomial. Equation (1) is equivalent to (2)\(k\left\{ {\sum\limits_{i = n + 1}^{2n} {x_i^r } - \sum\limits_{i = 1}^n {x_i^r } } \right\} = \delta _m^r m!,r = 1,2, \ldots ,2n\),r=1,2,..., 2n. By choosingf(x)=1/(z−x),x i fori=1,..., n andx i fori=n+1,..., 2n are shown to be roots ofg n (z) andh n (z) respectively, satisfying (3)\(e^{ - (m - 1){! \mathord{\left/ {\vphantom {! k}} \right. \kern-\nulldelimiterspace} k}z^m } g_n (z) = h_n (z)\left( {1 + \frac{{c_1 }}{{z^{2n + 1} }} + \frac{{c_2 }}{{z^{2n + 2} }} + \cdots } \right)\). It is convenient to normalize withk=(m−1)!. LetP s (z) denotez s · numerator of the (s+1)th diagonal member of the Padé table fore x, frx=1/z, that numerator being a constant factor times the general Laguerre polynomialL −2s−1 s (x), and letP s (X i )=0, i=1, ...,s. Then for anym, solutions to (1) are had, for2n=2ms, forx i , i=1, ...,ms, andx i , i=ms+1,..., 2ms, equal to all them th rootsX 1/m i and (−X i )1/m respectively, and they give {(2s+1)m−1}th degree accuracy. For2sm≦2n≦(2s+1)m−1, these (2sm)-point solutions are proven to be the only ones giving (2n)th degree accuracy. Thex i 's in (1) always include complex values, except whenm=1, 2n=2. For2sm<2n≦(2s+1)m−1,g n (z) andh n (z) are (n−sm)-parameter families of polynomials whose roots include those ofg ms (z) andh ms (z) respectively, and whose remainingn−ms roots are the same forg n (z) andh n (z). Form>1, and either 2n<2m or(2s+1)m−1<2n<(2s+2)m, it is proven that there are no non-trivial solutions to (1), real or complex. Form=1(1)6, tables ofx i are given to 15D, fori=1(1)2n, where 2n=2ms ands=1(1) [12/m], so that they are sufficient for attaining at least 24th degree accuracy in (1).
Similar content being viewed by others
References
Salzer, H. E.: Optimal Points for Numerical Differentiation. Numer. Math.2, 214–227 (1960).
Milne-Thomson, L. M.: The Calculus of Finite Differences, London: Macmillan 1933
Salzer, H. E.: Equally Weighted Quadrature Formulas over Semi-Infinite and Infinite Intervals. J. of Math. and Phys.34, 54–63 (1955).
Perron, O.: Die Lehre von den Kettenbrüchen, 2nd ed., Leipzig: B. G. Teubner 1929, (repr. by Chelsea, New York, 1950); 3rd ed., vol. II. Stuttgart: 1957.
Bateman Manuscript Project, Higher Transcendental Functions, vols. I and II. McGraw-Hill, New York: 1953.
Author information
Authors and Affiliations
Additional information
General Dynamics/Astronautics. A Division of General Dynamics Corporation.
Rights and permissions
About this article
Cite this article
Salzer, H.E. Equally-weighted formulas for numerical differentiation. Numer. Math. 4, 381–392 (1962). https://doi.org/10.1007/BF01386336
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF01386336