Table of Contents
Fetching ...

Additive growth amongst images of linearly independent analytic functions

Samuel Mansfield

TL;DR

This work develops a framework to prove additive growth for images and iterated sum-difference sets under real-analytic, linearly independent derivative families. By introducing 1- and $k$-independence of analytic functions and leveraging squeezing, equidistribution, and discrete-derivative analysis (with Hurwitz’s theorem to control zeros), the authors derive a recursive bound $|2^{n-1}f(A)-(2^{n-1}-1)f(A)| \gg |A|^{\phi(n)}$ with $\phi(1)=1$ and $\phi(n)=1+1/(1+1/\phi(n-1))$, approaching the golden ratio $\phi=(1+\sqrt{5})/2$. They extend these results to $|2^n f(A-A)-(2^n-1)f(A-A)|$ under a flat-discrete-derivative independence condition, showing growth for polynomials of degree $m\ge n+1$ and for $f=\arctan(e^x)$ with $n=3$, improving bounds on the additive growth of angles in Cartesian products. The approach unifies convex-function growth with a broader class of analytic functions and yields new growth bounds for both set-analytic and geometric (angle) configurations, with clear implications for additive combinatorics and geometric incidence problems.

Abstract

Let $\mathcal{F}$ be a set of $n$ real analytic functions with linearly independent derivatives restricted to a compact interval $I$. We show that for any finite set $A \subset I$, there is a function $f \in \mathcal{F}$ that satisfies $$|2^{n-1}f(A)-(2^{n-1}-1)f(A)|\gg_{\mathcal{F},I} |A|^{φ(n)},$$ where $φ:\mathbb{N} \to \mathbb{R}$ satisfies the recursive formula $$φ(1)=1, \quad φ(n)=1+\frac{1}{1+\frac{1}{φ(n-1)}} \quad \text{for } n\geq 2.$$ The above result allows us to prove the bound $$|2^nf(A-A)-(2^n-1)f(A-A)| \gg_{f,n,I} |A|^{1+φ(n)}$$ where $f$ is an analytic function for which any $n$ distinct non-trivial discrete derivatives of $f'$ are linearly independent. This condition is satisfied, for instance, by any polynomial function of degree $m \geq n+1$. We also check this condition for the function $\arctan(e^x)$ with $n=3$, allowing us to improve upon a recent bound on the additive growth of the set of angles in a Cartesian product due to Roche-Newton.

Additive growth amongst images of linearly independent analytic functions

TL;DR

This work develops a framework to prove additive growth for images and iterated sum-difference sets under real-analytic, linearly independent derivative families. By introducing 1- and -independence of analytic functions and leveraging squeezing, equidistribution, and discrete-derivative analysis (with Hurwitz’s theorem to control zeros), the authors derive a recursive bound with and , approaching the golden ratio . They extend these results to under a flat-discrete-derivative independence condition, showing growth for polynomials of degree and for with , improving bounds on the additive growth of angles in Cartesian products. The approach unifies convex-function growth with a broader class of analytic functions and yields new growth bounds for both set-analytic and geometric (angle) configurations, with clear implications for additive combinatorics and geometric incidence problems.

Abstract

Let be a set of real analytic functions with linearly independent derivatives restricted to a compact interval . We show that for any finite set , there is a function that satisfies where satisfies the recursive formula The above result allows us to prove the bound where is an analytic function for which any distinct non-trivial discrete derivatives of are linearly independent. This condition is satisfied, for instance, by any polynomial function of degree . We also check this condition for the function with , allowing us to improve upon a recent bound on the additive growth of the set of angles in a Cartesian product due to Roche-Newton.

Paper Structure

This paper contains 13 sections, 17 theorems, 102 equations, 1 figure.

Key Result

Theorem 1.1

Let $A \subset \mathbb{R}$ be a finite convex set, then

Figures (1)

  • Figure 1: Point sets $P$ determining $\Omega(|P|)$ distinct angles

Theorems & Definitions (35)

  • Theorem 1.1: Ruzsa-Shakan-Solymosi-Szemerédi shakanetc
  • Definition 1.2
  • Theorem 1.3: Hanson---Roche-Newton---Rudnev MishaBrandonOlly
  • Definition 1.4
  • Theorem 1.5: Bradshaw peter
  • Definition 1.6
  • Example 1.7
  • Theorem 1.8
  • Theorem 1.9
  • Theorem 1.10
  • ...and 25 more