English

Representations of analytic functions as infinite products and their application to numerical computations

Number Theory 2012-04-05 v3

Abstract

Let DD be an open disk of radius 1\le 1 in C\mathbb C, and let (ϵn)(\epsilon_n) be a sequence of ±1\pm 1. We prove that for every analytic function f:DCf: D \to \mathbb C without zeros in DD, there exists a unique sequence (αn)(\alpha_n) of complex numbers such that f(z)=f(0)n=1(1+ϵnzn)αnf(z) = f(0)\prod_{n=1}^{\infty} (1+\epsilon_nz^n)^{\alpha_n} for every zDz \in D. From this representation we obtain a numerical method for calculating products of the form pprimef(1/p)\prod_{p \text{prime}} f(1/p) provided f(0)=1f(0)=1 and f(0)=0f'(0) = 0; our method generalizes a well known method of Pieter Moree. We illustrate this method on a constant of Ramanujan π1/2pprimep2pln(p/(p1))\pi^{-1/2}\prod_{p \text{prime}} \sqrt{p^2-p}\ln(p/(p-1)). From the properties of the exponents αn\alpha_n, we obtain a proof of the following congruences, which have been the subject of several recent publications motivated by some questions of Arnold: for every n×nn \times n integral matrix AA, every prime number pp, and every positive integer kk we have trApktrApk1(modpk)\text{tr} A^{p^k} \equiv \text{tr} A^{p^{k-1}} (\text{mod}\,{p^k}).

Keywords

Cite

@article{arxiv.1202.1335,
  title  = {Representations of analytic functions as infinite products and their application to numerical computations},
  author = {Marcin Mazur and Bogdan V. Petrenko},
  journal= {arXiv preprint arXiv:1202.1335},
  year   = {2012}
}

Comments

Several editorial changes have been made