Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Sharp diameter bound on one segment between collinear roots

Proved
ErdosProblems.Erdos1041.PaperCompleteR21.sharp_collinear_root_diameter_monic

by willcook · Sep 24, 2026 · Mathlib c5ea003 (Lean v4.30.0)

erdos-1041-collinearpolynomial-rootssharp-inequality

Let f∈C[X]f\in\mathbb C[X]f∈C[X] be monic of degree n≥2n\ge2n≥2. Suppose all its root occurrences lie on an affine line b+dRb+d\mathbb Rb+dR, where b,d∈Cb,d\in\mathbb Cb,d∈C and ∣d∣=1|d|=1∣d∣=1. List those occurrences, with multiplicity, as b+dyib+d y_ib+dyi​ for real yiy_iyi​. For any greatest pairwise distance DDD among them, there are distinct indices j,kj,kj,k with yj≤yky_j\le y_kyj​≤yk​ and no listed coordinate strictly between them such that every zzz on their closed segment satisfies

∣f(z)∣≤(D/2)n2n−1cos⁡n(π/(2n)).|f(z)|\le\frac{(D/2)^n}{2^{n-1}\cos^n(\pi/(2n))}.∣f(z)∣≤2n−1cosn(π/(2n))(D/2)n​.

The selected indices can denote the same coordinate when roots repeat. The bound applies to one selected segment, not every pair; the separate equality and best-constant theorems establish sharpness. Will Cook authored the source paper and directed its research infrastructure (AI agents did most research and drafting; he did not independently verify every claim); Erdős–Herzog–Piranian had previously proved a contained collinear root segment under their stated hypotheses. Separately, ani supplied the degree-seven Hausdorff counterexample to the broader #1041 benchmark; that counterexample is distinct from this sharp-collinear proof.

Where to inspect the proof. This page shows the theorem statement with a by sorry stub; the accepted proof is stored separately under View graph → Solutions & Sketches. That route may require sign-in. A proved consequence imports this theorem in the same Lean environment.

Preamble
import Definitions.Def_ErdosProblems_Erdos1041_SharpCollinearChebyshev
import Definitions.Def_ErdosProblems_Erdos1041_PaperCompleteR21_CollinearDiameterWhole
import Mathlib
import Mathlib.Algebra.Polynomial.Degree.IsMonicOfDegree
import Mathlib.Algebra.Polynomial.Roots
import Mathlib.Analysis.SpecialFunctions.Trigonometric.Chebyshev.RootsExtrema
import Mathlib.Data.Real.Basic
import Mathlib.RingTheory.Polynomial.ScaleRoots
import Mathlib.Tactic.Choose
import Mathlib.Tactic.FieldSimp
import Mathlib.Tactic.Linarith
import Mathlib.Tactic.NormNum
import Mathlib.Tactic.Positivity
import Mathlib.Tactic.Ring
import Mathlib.Topology.Algebra.Polynomial
import Mathlib.Topology.Order.IntermediateValue

/-!
# Erdős 1041: the sharp Chebyshev bound for collinear roots, with sharpness

Paper-form restatement of

* `res:sharp-collinear-root-diameter`
  (`paper/1041/erdos-1041-lemniscate-newton-flow.tex`, line 1155),
* `thm:sharp-collinear-diameter`
  (`paper/reasoning-parts/erdos1041/core.tex`, line 1715),
* `cor:collinear-erdos-1041`
  (`paper/reasoning-parts/erdos1041/core.tex`, line 1769).

A monic complex polynomial of degree `n` whose zero occurrences are collinear is
written here in its factored form `∏ k, (X - C (base + dir * y k))` with
`‖dir‖ = 1` and `y : Fin n → ℝ`: that is exactly "monic of degree `n` with all
zero occurrences on one line", the line being `base + dir * ℝ`.  The diameter
`D` of the zero occurrences is carried as the hypothesis that `D` is the
greatest pairwise distance between zero occurrences.

The existing tree modules `SharpCollinearAlternation` and
`SharpCollinearChebyshev` supply the alternation/Chebyshev kernel: for a monic
real `p` of degree `m + 2` vanishing at `±1` and alternating in sign at `m + 1`
ordered interior points, one of those points has `|p| ≤ comparisonBound (m+2)`.
This file supplies everything the paper's proof needs around that kernel:

* the rigid normalisation (translation, rotation, scaling by `R = D/2`) and the
  transport identity `‖f (base + dir * s)‖ = R ^ n * |q ((s - mid)/R)|`;
* the existence of a maximiser of `|q|` in each root gap, and the fact that it
  is interior;
* the sign alternation of those gap maxima, proved directly from the product
  form of `q`;
* the upgrade from the kernel's pointwise conclusion to a bound on the WHOLE
  selected segment, which is where "`c i` is the gap maximum" is used;
* the repeated-zero (constant path) case;
* the sharpness clauses: the scaled Chebyshev configuration, its extreme zeros
  at distance `D`, a point of every adjacent gap at which the bound is attained
  with equality, and hence that no smaller constant works in any degree.
-/

set_option autoImplicit false
set_option maxHeartbeats 1600000

noncomputable section


open Polynomial Finset Set
open ErdosProblems.Erdos1041.SharpCollinearChebyshev

/-! ### The constant `C_n = 1 / (2^(n-1) cos^n (π/(2n)))` -/









/-! ### A monic product of linear factors -/



/-! ### Sign alternation between consecutive root gaps -/



/-! ### Gap maxima -/



/-! ### The normalised core: a whole gap below `C_n` -/



/-! ### Transport between the root line and `ℝ` -/









/-! ### The sharp collinear root-diameter theorem -/



/-! ### The Erdős case: a short curve inside the unit lemniscate -/



/-! ### Sharpness: the scaled Chebyshev root configuration -/







































/-! ### "Best possible in every degree" -/







/-! ### From an abstract monic polynomial with collinear zeros -/

open ErdosProblems.Erdos1041.PaperCompleteR21

Formal statement
theorem ErdosProblems.Erdos1041.PaperCompleteR21.sharp_collinear_root_diameter_monic {n : ℕ} (hn : 2 ≤ n) (f : ℂ[X])
    (hf : f.IsMonicOfDegree n) (base dir : ℂ) (hdir : ‖dir‖ = 1)
    (hcol : ∀ z ∈ f.roots, ∃ t : ℝ, z = base + dir * (t : ℂ)) :
    ∃ y : Fin n → ℝ, f = (∏ k, (X - C (base + dir * (y k : ℂ)))) ∧
      ∀ D : ℝ, IsGreatest {d : ℝ | ∃ j k : Fin n,
          d = dist (base + dir * (y j : ℂ)) (base + dir * (y k : ℂ))} D →
        ∃ j k : Fin n, j ≠ k ∧ y j ≤ y k ∧
          (∀ l : Fin n, y l ≤ y j ∨ y k ≤ y l) ∧
          dist (base + dir * (y j : ℂ)) (base + dir * (y k : ℂ)) ≤ D ∧
          ∀ z ∈ segment ℝ (base + dir * (y j : ℂ)) (base + dir * (y k : ℂ)),
            ‖f.eval z‖
              ≤ 1 / (2 ^ (n - 1) * Real.cos (Real.pi / (2 * (n : ℝ))) ^ n)
                * (D / 2) ^ n := by sorry
Source
Lean source: https://github.com/wcook04/plectis-erdos-lean/blob/c93c2e4dd86a2e317e0cb650ea244fee1afd59c2/ErdosProblems/Erdos1041/PaperCompleteR21/CollinearDiameterWhole.lean#L965-L981; paper theorem and historical comparison: https://github.com/wcook04/plectis-erdos/blob/6917e15ec4abc2623512254da93221e446eeb707/paper/1041/erdos-1041-lemniscate-newton-flow.tex#L1193-L1246

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me