On the set of points represented by harmonic subseries

Basic Information

• Paper ID: 2405.07681
• Title: On the set of points represented by harmonic subseries
• Author: Vjekoslav Kovač (University of Zagreb)
• Classification: math.NT (Number Theory), math.CA (Classical Analysis), math.CO (Combinatorics)
• Publication Date: May 2024 (arXiv v3: September 12, 2024)
• Paper Link: https://arxiv.org/abs/2405.07681

Abstract

This paper constructively resolves an open problem posed by Erdős and Graham in their 1980 monograph on combinatorial number theory through the design of a "convergence game" (Alice versus Bob). The author proves that the three-dimensional point set represented by harmonic subseries $\left\{\left(\sum_{n\in A}\frac{1}{n}, \sum_{n\in A}\frac{1}{n+1}, \sum_{n\in A}\frac{1}{n+2}\right): A \subset \mathbb{N}, \sum_{n\in A}\frac{1}{n}<\infty\right\}$ possesses non-empty interior. This generalizes an unpublished two-dimensional result of Erdős and Straus.

Research Background and Motivation

Problem Origins

1. Erdős's Series of Unit Fraction Problems: Paul Erdős posed numerous problems concerning the representation of numbers as finite or infinite sums of distinct unit fractions, which have driven the development of new techniques in number theory and combinatorics.
2. Two-Dimensional Erdős-Straus Result: Erdős and Straus (unpublished) proved that for all strictly increasing sequences of positive integers $(a_k)$ satisfying $\sum_k 1/a_k < \infty$, the point set $\left\{\left(x, y\right): x=\sum_k\frac{1}{a_k}, y=\sum_k\frac{1}{1+a_k}\right\}$ contains a non-empty open set.
3. Three-Dimensional Generalization Problem: Erdős and Graham posed in their 1980 monograph: does the three-dimensional (or higher-dimensional) case also hold? That is, considering $\left(x, y, z\right) = \left(\sum_k\frac{1}{a_k}, \sum_k\frac{1}{1+a_k}, \sum_k\frac{1}{2+a_k}\right)$

Significance of the Problem

• Theoretical Significance: This is a fundamental problem in harmonic series theory, involving the topological properties of "achievement sets"
• Higher-Dimensional Challenge: Compared to the two-dimensional case, the three-dimensional problem requires more refined arithmetic identities and control strategies
• Constructive Proof: This paper provides explicit construction, even computing specific open balls

Limitations of Existing Methods

• Achievement set theory has primarily focused on one-dimensional cases or special instances
• Research on topological properties of higher-dimensional vector-valued series is limited
• Lack of arithmetic tools necessary for addressing three-dimensional problems

Core Contributions

1. Resolution of a 40+ Year Open Problem: Constructively proves the affirmative answer to the Erdős-Graham three-dimensional problem (Theorem 1)
2. Innovative Game-Theoretic Method: Introduces a "convergence game" framework, transforming the problem into a strategic game between Alice and Bob
3. Key Arithmetic Lemma: Discovers and proves a fundamental arithmetic identity (Lemma 2), reducing the problem to perturbation series through linear transformation
4. Explicit Construction: Not only proves existence but also computes a specific open ball: radius $10^{-24}$, centered near approximately $(2.588\times 10^{-6}, 2.588\times 10^{-6}, 2.588\times 10^{-6})$
5. Elementary Methods: Uses minimal number-theoretic tools, relying primarily on clever arithmetic identities and convergence analysis

Detailed Methodology

Task Definition

Input: Target point $q = (q_1, q_2, q_3) \in \mathbb{R}^3$ located in a specific rectangular region
Output: Infinite set $A \subset \mathbb{N}$ such that $\left(\sum_{n\in A}\frac{1}{n}, \sum_{n\in A}\frac{1}{n+1}, \sum_{n\in A}\frac{1}{n+2}\right) = q$ and $\sum_{n\in A}\frac{1}{n}<\infty$

Core Strategic Architecture

Step One: Linear Transformation Reduction

Through Lemma 2, using the matrix $M = \begin{pmatrix} 1 & 0 & 0 \\ 3 & -4 & 1 \\ 1 & -2 & 1 \end{pmatrix}$ the original problem is transformed into a perturbation series problem. The key identity is: $M\begin{pmatrix} 1/(an) \\ 1/(an+1) \\ 1/(an+2) \end{pmatrix} = \begin{pmatrix} 1/(an) + O(1/n^4) \\ 2/(a^2n^2) + O(1/n^4) \\ 2/(a^3n^3) + O(1/n^4) \end{pmatrix}$

Step Two: Arithmetic Identity Construction

Discovers special finite sets $S_1, S_2, S_3, T_1, T_2, T_3 \subset \mathbb{N}$ such that by adding terms from $S_j$ and removing terms from $T_j$, one can "move" in the $j$-th coordinate direction: $\left(\sum_{a\in S_j} - \sum_{a\in T_j}\right) M\begin{pmatrix} 1/(an) \\ 1/(an+1) \\ 1/(an+2) \end{pmatrix} = \frac{c_j}{n^j}e_j + O\left(\frac{1}{n^4}\right)$

Specific construction:

• $S_1 = \{45, 72, 144, 160, 432, 480\}$, $T_1 = \{48, 60, 120, 720, 1440, 4320\}$
• $S_2 = 11\cdot\{16, 20, 240\}$, $T_2 = 11\cdot\{15, 24, 120\}$
• $S_3 = 7\cdot\{10, 30, 60\}$, $T_3 = 7\cdot\{12, 15\}$

These identities are based on Pythagorean identities and clever combinations of unit fractions.

Step Three: Sparsification Technique

To avoid repeated indices, the form $n = a(k^2m+1)$ is used, where $m = 2310 = 2\cdot 3\cdot 5\cdot 7\cdot 11$ is the product of all relevant prime factors, and $k \geq K$ ($K=14$).

Game Rules

The Game Between Alice and Bob:

• Initial Point: $p = \sum_{l=K}^{\infty}\sum_{j=1}^3 \sum_{a\in T_j} M\begin{pmatrix} 1/(a(l^2m+1)) \\ \cdots \end{pmatrix}$
• Round Structure: Proceeds with $k = K, K+1, \ldots$, each round corresponding to index $n = a(k^2m+1)$
• Alice's Action: For each coordinate $j$, decides whether to "activate" movement in that direction ($\epsilon_{k,j} \in \{0,1\}$)
• Bob's Interference: Adds at most $C/(k^2m+1)^4$ perturbation to each coordinate
• Alice's Decision Rule: $\epsilon_{k,j} = 1 \iff x_{k,j} + \frac{3c_j}{(k^2m+1)^j} \leq q_j$

Technical Innovations

1. Non-Greedy Strategy: Unlike classical greedy algorithms, Alice maintains a "safety distance" $3c_j/(k^2m+1)^j$, avoiding overshoot
2. Tail Control: Through precise estimation $\sum_{l=k}^{\infty}\frac{C}{(l^2m+1)^4} < \frac{c_j}{(k^2m+1)^j}$ ensuring future perturbations remain controllable
3. Bidirectional Approximation: Proves two key properties (Claims 1 and 2):
   • $\epsilon_{k,j}=1$ occurs infinitely often (ensuring not falling below target)
   • $\epsilon_{k,j}=0$ occurs infinitely often (ensuring not exceeding target)
4. Cauchy Sequence Argument: From $|x_{k+1}-x_k| = O(1/k^2)$ derives convergence

Experimental Setup

Nature of "Experiments"

This is a pure theoretical mathematics paper; "experiments" manifest as:

1. Computability of constructive proofs
2. Explicit parameter calculations
3. Numerical verification of specific open sets

Computational Tools

• Software: Mathematica 13.0.0
• Applications:
  • Verification of arithmetic identities
  • Computation of optimal constant $C = 8.7649\times 10^{-8}$
  • Determination of $K=14$ satisfying inequalities (4.2) and (4.3)
  • Calculation of initial point $p$ and target region

Parameter Settings

• Constant $C$: Obtained through optimization as $C = 8833/100776960000$
• Starting Index $K$: Determined through numerical verification and integral estimation as $K=14$
• Coefficients: $c_1 = 1/180$, $c_2 = 1/348480$, $c_3 = 1/1029000$

Experimental Results

Main Result (Theorem 1)

Theorem: The set $\left\{\left(\sum_{n\in A}\frac{1}{n}, \sum_{n\in A}\frac{1}{n+1}, \sum_{n\in A}\frac{1}{n+2}\right): A \subset \mathbb{N}, \sum_{n\in A}\frac{1}{n}<\infty\right\} \subseteq \mathbb{R}^3$ possesses non-empty interior.

Explicit Open Ball (Section 5)

Through constructive proof, the following is computed:

• Center Point: $\begin{pmatrix} 2.58842922\ldots \\ 2.58842919\ldots \\ 2.58842916\ldots \end{pmatrix} \times 10^{-6}$
• Radius: $10^{-24}$

This is obtained through the following steps:

1. Computation of rectangular region $Q$ (formula 4.4)
2. Selection of maximum inscribed ball in $Q$
3. Transformation to ellipsoid via $M^{-1}$
4. Estimation of minimum axis length of ellipsoid

Verification of Key Inequalities

For $k \geq 14$ and $j=1,2,3$, the following are verified: $\sum_{l=k}^{\infty}\frac{3C}{(l^2m+1)^4} < \frac{c_j}{(k^2m+1)^j}$$\sum_{l=k+1}^{\infty}\frac{c_j}{(l^2m+1)^j} > \frac{4c_j}{(k^2m+1)^j}$

Proof Structure Verification

Through proof by contradiction in Claims 1 and 2, the following are established:

• Sequence $(x_k)$ is a Cauchy sequence
• The limit equals exactly the target point $q$
• For each point in the target region, there exists a corresponding set $A$

Related Work

Unit Fraction Representation Problems

1. Kakeya (1914): Earliest research on topological properties of achievement sets
2. Guthrie-Nymann-Sáenz (1988-2000): Completely resolved one-dimensional case, discovering four topological types
3. Graham (1964): Research on series with smooth replaceable terms

Achievement Set Theory

1. Bartoszewicz et al. (2013-2018): Research on achievement sets in the plane, including geometric series and conditionally convergent series
2. Morán (1989, 1994): Research on fractal properties and dimensions of achievement sets
3. Laltanpuia-Singh (2008): Research from vector measure perspective

Uniqueness of This Work

• First Three-Dimensional Result: Resolves the Erdős-Graham problem posed in 1980
• Elementary Methods: Does not rely on advanced fractal theory or measure theory
• Constructive: Provides explicit algorithms and parameters, rather than mere existence proofs
• Game-Theoretic Perspective: Innovatively introduces adversarial game framework

Conclusions and Discussion

Main Conclusions

1. Affirmative Answer: The Erdős-Graham three-dimensional problem is answered affirmatively
2. Generalizability: The method can in principle be extended to higher dimensions, but requires more complex arithmetic identities
3. Computability: Not only proves existence but also computes specific parameters

Limitations

1. Very Small Open Ball: Radius of only $10^{-24}$ indicates that interior points, while existing, are "sparse"
2. Higher-Dimensional Generalization: The paper does not address four or higher dimensions; construction of arithmetic identities becomes increasingly difficult
3. Optimality Unknown: Unclear whether larger interior open sets can be found
4. Specific Form: Only addresses the case $(1/n, 1/(n+1), 1/(n+2))$; other shift forms are not discussed

Future Directions

1. Higher-Dimensional Generalization: Seeking arithmetic identities for four or higher dimensions
2. Optimal Boundaries: Research on the precise size of interior open sets
3. General Shifts: Consider more general forms such as $(1/n, 1/(n+d_1), 1/(n+d_2))$
4. Algorithm Optimization: Improve Alice's strategy to obtain larger open sets
5. Fractal Dimension: Research the Hausdorff dimension of the entire point set

In-Depth Evaluation

Strengths

1. Problem Importance: Resolves a classical open problem of 40+ years, possessing significant theoretical value
2. Method Innovation:
   • Novel and intuitive game-theoretic perspective
   • Excellent pedagogical design of "warm-up game" (Section 2)
   • Highly skillful discovery of arithmetic identities
3. Proof Completeness:
   • Constructive proof with each step verifiable
   • Gradual transition from simple to complex games
   • Rigorous bidirectional approximation argument (Claims 1 and 2)
4. Computability:
   • All constants explicitly provided
   • Mathematica verification enhances credibility
   • Section 5 provides specific open ball
5. Writing Quality:
   • Clear structure, rigorous logic
   • "Warm-up" in Section 2 greatly enhances readability
   • Systematic notation

Weaknesses

1. Quantitative Aspects of Results:
   • Open ball radius of $10^{-24}$ is extremely small, with limited practical significance
   • No discussion of whether this approaches optimality
2. Method Limitations:
   • Construction of arithmetic identities relies on computer search, lacking systematic theory
   • Difficulties in higher-dimensional generalization insufficiently discussed
   • Sparsification parameter $m=2310$ appears to be obtained through trial, lacking theoretical guidance
3. Technical Details:
   • Proof of Lemma 2 primarily relies on verification, lacking deeper insight
   • Why these specific sets $S_j, T_j$? Is there a systematic construction method?
4. Insufficient Generalization Discussion:
   • No attempt at four-dimensional case
   • Do other types of series (such as $1/(n+a), 1/(n+b), 1/(n+c)$) apply?
5. Limited Connection to Achievement Set Theory:
   • While related literature is mentioned, deeper discussion of the relationship between this paper's methods and existing theory is lacking
   • Can this result be derived from a more general framework?

Impact

1. Theoretical Contribution:
   • Resolution of classical problem will be widely cited
   • Game-theoretic method may inspire research on other problems
2. Methodological Contribution:
   • "Convergence game" framework possesses generality
   • Arithmetic reduction techniques may apply to other unit fraction problems
3. Practical Value:
   • Pure theoretical result with no direct applications
   • Techniques may inspire algorithm design
4. Reproducibility:
   • Fully reproducible with all parameters explicitly provided
   • Mathematica code can reproduce computations

Applicable Scenarios

1. Number Theory Research: Researchers studying unit fraction representation problems
2. Combinatorics: Extensions of achievement set theory
3. Real Analysis: Fine analysis of series convergence
4. Teaching: Section 2's "game" design suitable for advanced mathematics instruction

Key References

1. Erdős & Graham (1980): Old and new problems and results in combinatorial number theory - Posed original problem
2. Guthrie & Nymann (1988): Completely resolved one-dimensional achievement set problem
3. Graham (1964): Research on series with smooth replaceable terms
4. Bartoszewicz et al. (2015-2018): Modern research on two-dimensional achievement sets
5. Kakeya (1914): Foundational work in achievement set theory

Overall Assessment

This is an excellent pure mathematics paper that resolves a long-standing open problem using elementary yet highly skillful methods. The paper's greatest strengths are:

1. Innovative game-theoretic framework transforming complex convergence problems into intuitive strategic games
2. Clever arithmetic identities achieving crucial dimensional reduction
3. Constructive proof not only proving existence but computing specific parameters

Main weaknesses lie in the quantitative aspects of results (extremely small open ball) and difficulties in higher-dimensional generalization. Nevertheless, this represents significant progress in the field and will have lasting impact. The writing is clear, with Section 2's "warm-up game" design exemplary in making complex proofs accessible.

Recommendation Index: ⭐⭐⭐⭐⭐ (5/5)
Difficulty Level: Advanced undergraduate/graduate level (requires background in real analysis and elementary number theory)