Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.
Updated
Statement
Setting (pp. 1--3). For a convex polygon split by an antipodal cut into chains and , the distance matrix is the matrix with entries , and its skeleton (entries equal to kept, all others set to ) is the - cut matrix of . A real matrix has
- the diagonal property if its entries are positive and it has no submatrix with ;
- the obtuse angle property if its entries are positive and it has no acute angle submatrix, where a real matrix, , is an acute angle matrix when there are , , , with and .
A real matrix with both properties is distance-like; by Propositions 2 and 3 (p. 4) every distance matrix is distance-like. A - matrix is pattern feasible (after Fishburn and Reeds) if it avoids nine listed small matrices and every staircase matrix , (p. 2).
For integers , a real matrix is a cycle with an intersection-free edge if there are positive integers and at most , and and at most , with and for and for each , indices taken modulo (p. 3). The staircase matrix , , and every real matrix whose skeleton is the pattern feasible matrix with rows , , , are such cycles.
Theorem 2 (p. 3, quoted). "No cycle with an intersection-free edge is a distance-like matrix."
Consequence (p. 3). No distance-like matrix has skeleton , so the pattern feasible matrix is not a - cut matrix; this answers in the negative Fishburn and Reeds's question whether every pattern feasible matrix is a - cut matrix.
Proof pointer
Section 2.3 (p. 5). The proof shows that such a cycle fails the obtuse angle property: from the cycle's index sequences it picks four rows and four columns, the choice depending on the relative position of two extremal indices, whose intersection is an acute angle submatrix.
Read depth
Claims checked: the definitions and Theorem 2 were read clause by clause on the page images of arXiv:1009.2216v3, and the proof in Section 2.3 was followed in outline. Nothing here is independently reviewed.
Dependencies
None in the corpus. Proposition 2, cited by the paper from Pach and Tardos, and Proposition 3, proved on p. 4, give the link to convex polygons.
Source. A. Aggarwal, On unit distances in a convex polygon, Discrete Math. 338 (2015), no. 3, 88--92, doi:10.1016/j.disc.2014.10.009; the edition read is named on the source card.
Bears on
- Problem 96: With Propositions 2 and 3, Theorem 2 shows that no distance matrix of an antipodal cut of a convex polygon is a cycle with an intersection-free edge; it gives no bound on by itself.