No Arabic abstract
An empty simplex is a lattice simplex with only its vertices as lattice points. Their classification in dimension three was completed by White in 1964. In dimension four, the same task was started in 1988 by Mori, Morrison, and Morrison, with their motivation coming from the close relationship between empty simplices and terminal quotient singularities. They conjectured a classification of empty simplices of prime volume, modulo finitely many exceptions. Their conjecture was proved by Sankaran (1990) with a simplified proof by Bober (2009). The same classification was claimed by Barile et al. in 2011 for simplices of non-prime volume, but this statement was proved wrong by Blanco et al. (2016+). In this article we complete the classification of $4$-dimensional empty simplices. In doing so we correct and complete the classification claimed by Barile et al., and we also compute all the finitely many exceptions, by first proving an upper bound for their volume. The whole classification has: - One $3$-parameter family, consisting of simplices of width equal to one. - Two $2$-parameter families (the one in Mori et al., plus a second new one). - Forty-six $1$-parameter families (the 29 in Mori et al., plus 17 new ones). - $2461$ individual simplices not belonging to the above families, with volumes ranging between 29 and 419. We characterize the infinite families of empty simplices in terms of lower dimensional point configurations that they project to, with techniques that can be applied to higher dimensions and larger classes of lattice polytopes.
A lattice $d$-simplex is the convex hull of $d+1$ affinely independent integer points in ${mathbb R}^d$. It is called empty if it contains no lattice point apart of its $d+1$ vertices. The classification of empty $3$-simplices is known since 1964 (White), based on the fact that they all have width one. But for dimension $4$ no complete classification is known. Haase and Ziegler (2000) enumerated all empty $4$-simplices up to determinant 1000 and based on their results conjectured that after determinant $179$ all empty $4$-simplices have width one or two. We prove this conjecture as follows: - We show that no empty $4$-simplex of width three or more can have determinant greater than 5058, by combining the recent classification of hollow 3-polytopes (Averkov, Krumpelmann and Weltge, 2017) with general methods from the geometry of numbers. - We continue the computations of Haase and Ziegler up to determinant 7600, and find that no new $4$-simplices of width larger than two arise. In particular, we give the whole list of empty $4$-simplices of width larger than two, which is as computed by Haase and Ziegler: There is a single empty $4$-simplex of width four (of determinant 101), and 178 empty $4$-simplices of width three, with determinants ranging from 41 to 179.
Let $X={x_1,ldots,x_n} subset mathbb R^d$ be an $n$-element point set in general position. For a $k$-element subset ${x_{i_1},ldots,x_{i_k}} subset X$ let the degree ${rm deg}_k(x_{i_1},ldots,x_{i_k})$ be the number of empty simplices ${x_{i_1},ldots,x_{i_{d+1}}} subset X$ containing no other point of $X$. The $k$-degree of the set $X$, denoted ${rm deg}_k(X)$, is defined as the maximum degree over all $k$-element subset of $X$. We show that if $X$ is a random point set consisting of $n$ independently and uniformly chosen points from a compact set $K$ then ${rm deg}_d(X)=Theta(n)$, improving results previously obtained by Barany, Marckert and Reitzner [Many empty triangles have a common edge, Discrete Comput. Geom., 2013] and Temesvari [Moments of the maximal number of empty simplices of a random point set, Discrete Comput. Geom., 2018] and giving the correct order of magnitude with a significantly simpler proof. Furthermore, we investigate ${rm deg}_k(X)$. In the case $k=1$ we prove that ${rm deg}_1(X)=Theta(n^{d-1})$.
In this paper we completely characterize lattice ideals that are complete intersections or equivalently complete intersections finitely generated semigroups of $bz^noplus T$ with no invertible elements, where $T$ is a finite abelian group. We also characterize the lattice ideals that are set-theoretic complete intersections on binomials.
We describe an algorithm for testing the completeness of caps in PG(r; q), q even. It allowed us to check that the 95256-cap in PG(12; 4) recently found by Fu el al. (see [14]) is complete.
We show that, for every set of $n$ points in the $d$-dimensional unit cube, there is an empty axis-parallel box of volume at least $Omega(d/n)$ as $ntoinfty$ and $d$ is fixed. In the opposite direction, we give a construction without an empty axis-parallel box of volume $O(d^2log d/n)$. These improve on the previous best bounds of $Omega(log d/n)$ and $O(2^{7d}/n)$ respectively.