TOPICS
Search

Nonintersecting Diagonal Problem


The nonintersecting diagonal problem asks for the maximum number d(n) of pairwise disjoint unit-square diagonals in an n×n array of squares, where n is a positive integer. Selected diagonals may neither cross nor share an endpoint. For even n, d(n)=n(n+1)/2 (Boyland et al. 2017). Song (2026) proved that for odd n,

 d(n)=(n(n+1))/2+|_(n+1)/6_|,

where |_x_| is the floor function. The values for n=1, 2, ... are 1, 3, 6, 10, 16, 21, 29, 36, 46, 55, ... (OEIS A264041).

Song (2026) reports AI assistance with the proof and Lean formalization. The formal reduction assumes the Fisher-Fitzpatrick grid-cover theorem.


Explore with Wolfram|Alpha

References

Boyland, P.; Pintér, G.; Laukó, I.; Roth, I.; Schoenfield, J. E.; and Wasielewski, S. "On the Maximum Number of Non-Intersecting Diagonals in an Array." J. Integer Seq. 20, Article 17.2.4, 2017. https://cs.uwaterloo.ca/journals/JIS/VOL20/Pinter/pinter3.html.Fisher, D. C. and Fitzpatrick, S. L. "The Isometric Path Number of a Graph." J. Combin. Math. Combin. Comput. 38, 97-110, 2001.Sloane, N. J. A. Sequence A264041 in "The On-Line Encyclopedia of Integer Sequences."Song, L. "Nonintersecting Unit Diagonals in Odd Square Arrays." 16 Oct 2026. https://arxiv.org/abs/2610.33042.

Cite this as:

Weisstein, Eric W. "Nonintersecting Diagonal Problem." From MathWorld--A Wolfram Resource. https://mathworld.wolfram.com/NonintersectingDiagonalProblem.html

Subject classifications