Wiki
Wiki

Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.

Updated

External and contextual interfaces for Yip's Theorem 1.3


Dependencies of the representation proof

The proof of Yip's Theorem 1.3 invokes no external research theorem. Its elementary analytic inputs are written at their application points:

  • for t≠0t\ne0, the equation x−it=eiθx^{-it}=e^{i\theta} has arbitrarily large positive real solutions, obtained by solving a linear congruence for log⁡x\log x;
  • for g(y)=y−(1+it)g(y)=y^{-(1+it)}, the identity g(n)−g(x)=∫xng′(y) dyg(n)-g(x)=\int_x^n g'(y)\,dy and ∣g′(y)∣=∣1+it∣y−2|g'(y)|=|1+it|y^{-2} give the finite-block estimate; and
  • finite harmonic mass gives absolute convergence of ∑n−(1+it)\sum n^{-(1+it)}, so block endpoints and increasing enumerations have the same sum.

The first two items are steps of the proof of Lemma 2.1 (p. 2), sketched on [[analysis/yip_2025_problem_erdos_ingham/lemma_2_1|the Lemma 2.1 page]]; the print obtains the second from the mean value theorem, and the integral form above gives the same bound. The last closes the proof of Theorem 1.3 (p. 3), sketched on the [[analysis/yip_2025_problem_erdos_ingham/theorem_1_3|Theorem 1.3 page]], and also lets the set be listed in increasing order for Problem 967. No recursive external proof reconstruction is needed for the selected theorem.

Contextual Erdős--Ingham equivalence

Yip's Theorem 1.2 on printed p. 1 is a restatement of Erdős and Ingham's Theorem 4. In Yip's integer setting, let 1<a1<a2<⋯1<a_1<a_2<\cdots be finite or infinite and suppose ∑kak−1<∞\sum_k a_k^{-1}<\infty. Yip states the equivalence of:

  1. 1+∑kak−1−it≠01+\sum_k a_k^{-1-it}\ne0 for every real tt;
  2. for every nondecreasing f:R≥0→R≥0f:\mathbb R_{\ge0}\to\mathbb R_{\ge0} which vanishes on [0,1)[0,1),
f(x)+∑kf(x/ak)∼(1+∑kak−1)x(x→∞)f(x)+\sum_k f(x/a_k) \sim\left(1+\sum_k a_k^{-1}\right)x \quad(x\to\infty)

implies f(x)∼xf(x)\sim x.

The published 1964 theorem has the broader initial setup of a finite or infinite nondecreasing real sequence 1<a1≤a2≤⋯1<a_1\le a_2\le\cdots with finite reciprocal sum. Its function class I\mathcal I consists of nonnegative, nondecreasing, locally bounded real functions which vanish below 11. The exact historical statement and page locators are filed at [[analysis/erdos_1964_arithmetical_tauberian_theorems/_index|the Erdős--Ingham source home]].

This equivalence supplies historical motivation and a consequence of a zero, but it is not used to construct the set in Theorem 1.3 or to transfer that set to Problem 967. Its external proof is therefore outside the selected direct proof chain and receives no proof credit here.