Wiki
Wiki

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

Updated


Statement

Notation (printed p. 254): for n∈N∪{∞}n\in\mathbb N\cup\{\infty\}, J(K,n)J(K,n) is the set of points P∈∂KP\in\partial K such that there is a circle centred at PP that meets ∂K\partial K in at least nn points, and ∣X∣|X| is the 1-dimensional Hausdorff measure (perimeter) of X⊂R2X\subset\mathbb R^2.

Theorem 1.3 (printed p. 254). "Let ε>0\varepsilon>0, then there is a convex body KεK_\varepsilon such that

∣J(Kε,∞)∣∣∂Kε∣>1−ε."\frac{|J(K_\varepsilon,\infty)|}{|\partial K_\varepsilon|}>1-\varepsilon."

The paper adds (p. 254) that if K0K_0 is a segment or an acute triangle, KεK_\varepsilon can be built so that Kε→K0K_\varepsilon\to K_0 in the Hausdorff metric as ε→0\varepsilon\to0. It suggests that part of the difficulty of finding a bound on N(K)N(K) uniform in KK may come from this theorem and Theorem 1.4.

Source. I. Bárány and E. Roldán-Pensado, A question from a famous paper of Erdős, Discrete Comput. Geom. 50 (2013), 253--261, doi:10.1007/s00454-013-9507-z; the definitions and Theorem 1.3 on printed p. 254. The edition read is identified on the source card.

Read depth. Claims checked: the definitions and the theorem were read clause by clause on the page image. The proof (pp. 259--260) was read for structure only, and nothing here is independently reviewed.

Proof pointer

§ 3, pp. 259--260. Near a triangle A1A2A3A_1A_2A_3, choose BiB_i close to AiA_i so that A1B1A2B2A3B3A_1B_1A_2B_2A_3B_3 is a convex hexagon with acute angles ∠AiBiAi+1\angle A_iB_iA_{i+1}, and apply Lemma 3.1 (p. 258) with N=∞N=\infty on each AiBiAi+1Bi+1A_iB_iA_{i+1}B_{i+1}; near a segment [A,B][A,B] the same is done with a convex quadrilateral ACBDACBD with acute angles at CC and DD. Not checked here.

Dependencies

Lemma 3.1 (p. 258) of the paper.

Bears on

  • Problem 982: the theorem concerns the circles centred at points of a convex curve in Erdős's 1946 convex-curve statement, the strongest of the conjectures his paper poses on p. 248; it says nothing about distinct distances from a vertex of a convex polygon and leaves the problem's statement undecided.