On nearly Gorenstein affine semigroups

An affine semigroup is a finitely generated subsemigroup of $(\mathbb Z_{\ge 0}^d, +)$, and a numerical semigroup is an affine semigroup with $d = 1$. A growing body of recent work examines shifted families of numerical semigroups, that is, families of numerical semigroups of the form $M_n = \langle n + r_1, \ldots, n + r_k \rangle$ for fixed $r_1, \ldots, r_k$, with one semigroup for each value of the shift parameter $n$. It has been shown that within any shifted family of numerical semigroups, the size of any minimal presentation is bounded (in fact, this size is eventually periodic in $n$). In this paper, we consider shifted families of affine semigroups, and demonstrate that some, but not all, shifted families of 4-generated affine semigroups have arbitrarily large minimal presentations.

Affine semigroups—discrete analogues of convex polyhedral cones—mark the cross-roads of algebraic geometry, commutative algebra and integer programming. They constitute the combinatorial background for the theory of toric varieties, which is their main link to algebraic geometry. Initiated by the work of Demazure [17] and Kempf, Knudsen, Mumford and Saint-Donat [33] in the early 70s, toric geometry is still a very active area of research. However, the last decade has clearly witnessed the extensive study of affine semigroups from the other two perspectives. No doubt, this is due to the tremendously increased computational power in algebraic geometry, implemented through the theory of Grobner bases, and, of course, to the modern computers. In this article we overview those aspects of this development that have been relevant for our own research, and pose several open problems. Answers to these problems would contribute substantially to the theory. The paper treats two main topics: (1) affine semigroups and several covering properties for them and (2) algebraic properties for the corresponding rings (Koszul, Cohen-Macaulay, different “sizes” of the defining binomial ideals). We emphasize the special case when the initial data are encoded into lattice polytopes. The related objects—polytopal semigroups and algebras— provide a link with the classical theme of triangulations into unimodular simplices. We have also included an algorithm for checking the semigroup covering property in the most general setting (Section 4). Our counterexample to certain covering conjectures (Section 3) was found by the application of a small part of this algorithm. The general algorithm could be used for a deeper study of affine semigroups.

We present a techinique which constructs an affine semigroup Se,d (resp. a monomial variety) from a given affine semigroup S (resp. monomial variety). The two affine semigroups (resp. monomial varieties) have a number of algebraic properties in common. Using this technique we are able to unify and extend a number of results scattered in the literature.

We present several new algorithms for computing factorization invariant values over affine semigroups. In particular, we give (i) the first known algorithm to compute the delta set of any affine semigroup, (ii) an improved method of computing the tame degree of an affine semigroup, and (iii) a dynamic algorithm to compute catenary degrees of affine semigroup elements. Our algorithms rely on theoretical results from combinatorial commutative algebra involving Gröbner bases, Hilbert bases, and other standard techniques. Implementation in the computer algebra system GAP is discussed.

Wilf Conjecture on numerical semigroups is a question posed by Wilf in 1978 and is an inequality connecting the Frobenius number, embedding dimension and the genus of the semigroup. The conjecture is still open in general. We prove that this Wilf inequality is preserved under gluing of numerical semigroups. If the numerical semigroups minimally generated by [Formula: see text] and [Formula: see text] satisfy the Wilf inequality, then so does their gluing which is minimally generated by [Formula: see text]. We discuss the extended Wilf’s Conjecture in higher dimensions for certain affine semigroups and prove an analogous result.

We provide algorithms for performing computations in generalized numerical semigroups, that is, submonoids of [Formula: see text] with finite complement in [Formula: see text]. These semigroups are affine semigroups, which in particular implies that they are finitely generated. For a given finite set of elements in [Formula: see text] we show how to deduce if the monoid spanned by this set is a generalized numerical semigroup and, if so, we calculate its set of gaps. Also, given a finite set of elements in [Formula: see text] we can determine if it is the set of gaps of a generalized numerical semigroup and, if so, compute the minimal generators of this monoid. We provide a new algorithm to compute the set of all generalized numerical semigroups with a prescribed genus (the cardinality of their sets of gaps). Its implementation allowed us to compute (for various dimensions) the number of numerical semigroups of higher genus than has previously been computed.

Let $\CaC\subset \Q^p$ be a rational cone. An affine semigroup $S\subset \CaC$ is a $\CaC$-semigroup whenever $(\CaC\setminus S)\cap \N^p$ has only a finite number of elements. In this work, we study the tree of $\CaC$-semigroups, give a method to generate it and study their subsemigroups with minimal embedding dimension. We extend Wilf's conjecture for numerical semigroups to $\CaC$-semigroups and give some families of $\CaC$-semigroups fulfilling the extended conjecture. We also check that other conjectures on numerical semigroups seem to be also satisfied by $\CaC$-semigroups.

LetS be a uniform semigroup (this includes all topological groups and affine semigroups). We show that a certain space of uniformly continuous functions onS has a left invariant mean iffS has the fixed point property for uniformly continuous affine actions ofS on compact convex sets. This is closely related to but independent of the results of T. Mitchell in [13] and A. Lau in [10]. Interesting examples and consequences are given for the special cases of topological groups and affine convolution semigroups of probability measures on a locally compact semigroup or group.

The aim of this thesis is to study discrete structures associated with group- based models. Group-based models are statistical models on phylogenetic trees that can be parametrized by polynomial maps. The algebraic varieties given by these polynomial maps are toric. By the well-known correspondence between toric varieties and polyhedral fans, we can associate discrete structures, such as lattice polytopes and affine semigroups, with group-based models. We follow three main lines in this dissertation. In Chapters 2 and 3, we study the Hilbert polynomials of group-based models. This is motivated by a result of Buczyńska and Wiśniewski stating that the Hilbert polynomial of the Jukes- Cantor binary model on a trivalent tree does not depend on the shape of the tree. In Chapter 2, we give a simple combinatorial proof to this statement, and in Chapter 3, we show that the analogous statement does not hold for the Kimura 3-parameter model. In Chapters 4 and 5, we study the phylogenetic semigroups on graphs that generalize the Jukes-Cantor binary model on trees. In Chapter 4, we study the maximal degrees of the minimal generators of these semigroups. In Chapter 5, we investigate the minimal generators of the phylogenetic semigroups on graphs with a few holes, extending the work of Buczyńska. Finally, in Chapter 6, we establish a connection between Berenstein-Zelevinsky triangles from representation theory and group-based models. This is motivated by the recent work of Sturmfels, Xu, and Manon related to conformal block algebras.

Affine semigroups and second duals

Dualizing complexes of seminormal affine semigroup rings and toric face rings

Let C⊆Np be an integer cone. A C-semigroup S⊆C is an affine semigroup such that the set C\\S is finite. Such C-semigroups are central to our study. We develop new algorithms for computing C-semigroups with specified invariants, including genus, Frobenius element, and their combinations, among other invariants. To achieve this, we introduce a new class of C-semigroups, termed B-semigroups. By fixing the degree lexicographic order, we also research the embedding dimension for both ordinary and mult-embedded N2-semigroups. These results are applied to test some generalizations of Wilf’s conjecture.

We give a formula for a Frobenius vector of a Gorenstein simplicial affine semigroup $S$, and when the semigroup is Cohen-Macaulay we give an algorithm computing the set of minimal Frobenius vectors of $S$ for a special class of semigroups.

We study the topological rigidity of affine semigroups of algebraic actions. In the first part of the paper we given an algebraic condition for rigidity that unifies previous rigidity results and we settle an old question of Walters. In the second part we study a generalized notion of the rigidity of hyperbolic actions.

Let A⊆B be cancellative abelian semigroups, and let R be an integral domain. We show that the semigroup ring R[B] can be decomposed, as an R[A]-module, into a direct sum of R[A]-submodules of the quotient ring of R[A]. In the case of a finite extension of positive affine semigroup rings, we obtain an algorithm computing the decomposition. When R[A] is a polynomial ring over a field, we explain how to compute many ring-theoretic properties of R[B] in terms of this decomposition. In particular, we obtain a fast algorithm to compute the Castelnuovo–Mumford regularity of homogeneous semigroup rings. As an application we confirm the Eisenbud–Goto conjecture in a range of new cases. Our algorithms are implemented in the Macaulay2 package MonomialAlgebras.