The Zetafast algorithm for computing zeta functions
Abstract
We express the Riemann zeta function of argument with imaginary part in terms of three absolutely convergent series. The resulting simple algorithm allows to compute, to arbitrary precision, and its derivatives using at most summands for any , with explicit error bounds. It can be regarded as a quantitative version of the approximate functional equation. The numerical implementation is straightforward. The approach works for any type of zeta function with a similar functional equation such as Dirichlet -functions, or the Davenport-Heilbronn type zeta functions.
Cite
@article{arxiv.1703.01414,
title = {The Zetafast algorithm for computing zeta functions},
author = {Kurt Fischer},
journal= {arXiv preprint arXiv:1703.01414},
year = {2017}
}
Comments
v1: 13 pages, 1 figure; v2: references updated, new material added, v3: estimates sharpened and simplified, v4: estimate for N sharpened and acknowledgment added, v5: estimates sharpened and misprints corrected, v6: corrected inaccuracy in proof for positive integer arguments and typos, v7: 2 misprinted plus signs in (1.1) and (1.12) corrected