Triangular-row conjecture for optimal n-mean configurations

About 11 years old · traced to

Let nn be a positive integer, and set

N=⌊2n⌋,N=\lfloor\sqrt{2n}\rfloor,

and

J=N−∣n−N(N+1)2∣.J=N-\left|n-\frac{N(N+1)}{2}\right|.

An optimal configuration is a set of nn points minimizing the quantization error for the uniform distribution on an equilateral triangle. Arrange its points in NN rows, with the jjth row indexed by jj.

Triangular-row conjecture. For most nn, there is an optimal configuration with NN rows. The jjth row has jj points for j≤Jj\leq J. If n>N(N+1)/2n>N(N+1)/2, the rows with j>Jj>J each have one extra point, so the jjth row has j+1j+1 points; if n<N(N+1)/2n<N(N+1)/2, they each have one fewer point, so the jjth row has j−1j-1 points.

The conjecture describes the row structure suggested by the numerical search and by the observed proximity of optimal configurations to triangular lattices. It remains open in the supplied source.

References

Primary source

Carl P. Dettmann and Mrinal Kanti Roychowdhury, “Quantization for uniform distributions on equilateral triangles”, arXiv:1508.00498 (2017).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.