Explicit estimates for Artin \(L\)-functions: Duke's short-sum theorem and Dedekind zeta residues (Q2162826)

The main result of the paper is the following one, namely Theorem 1, to be quoted as Theorem 1: Let \(L(s,\chi)\) be an entire Artin \(L\)-function that satisfy the Generalized Riemann Hypothesis, in which \(\chi\) has degree \(d\) and conductor \(N\). Then: \[ \left\vert\log L(1,\chi)- \sum \frac{\chi(p)}{p}\right\vert\le 13,53\,d; \] here the summation has only been extended over those primes \(p\) satisfying \(o\le \sqrt{\log N}\). Theorem 1 is an explicit version of an older theorem due to \textit{W. Duke} [Compos. Math. 136, No. 1, 103--115 (2003; Zbl 1013.11072)]: \[ \log L(1,\chi)= \sum_{p\le\sqrt{\log N}} \frac{\chi(p)}{p}+ O(1). \] Theorem 1 is an explicit version of Duke's theorem [loc. cit.]; here ``explicit'' means that no implicit constants are left being unspecified. The proof of Theorem 1 is carried out in detail and fills the greater part of the paper. At some places the computer language Mathematica and Python have been use used but only very seldom. The way of reasoning is very clearly done and worthwhile to read. The author's methods are also of use in attacking related number theoretical problems.