Newton Polytopes and Analytic Spread
Benjamin Drabkin, Benjamin Oltsik · 2025
Read Newton Polytopes and Analytic Spread by Benjamin Drabkin, Benjamin Oltsik (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-00671.txt.
Introduction
Benjamin Drabkin and Benjamin Oltsik use a geometric tool to study a size measure for monomial ideals. In algebra, an ideal is a set of polynomials closed under certain operations. The Newton polytope is a shape built from the exponents in those polynomials. Analytic spread measures how many variables a minimal reduction of the ideal needs. Their 2025 paper ties the two together, bounding the spread by the faces and walls of the polytope.
The bound lets them classify the simplest monomial ideals, called basic ones. They then compute the spread exactly for several families of ideals. Worked families show the method is usable, not just theoretical. The authors offer the shape-first approach as a shortcut around heavier algebra.
Catalog record
| CREATOR | Benjamin Drabkin, Benjamin Oltsik |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 45 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-00671
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Newton Polytopes and Analytic Spread Benjamin Drabkin, Benjamin Oltsik Abstract . Using the Newton polytope and polyhedron, we study analytic spread and ideal reductions of monomial ideals. We determine a bound for analytic spread based on halfspaces and hyperplanes of the Newton polytope, and we classify basic monomial ideals. We then apply this method to calculate the analytic spread for a few families of monomial ideals. 2020 Mathematics Subject Classification Primary: 13A02, 13E05, 13H15 1. Introduction Let K be an infinite field, and let I\subseteq R=K[x_{1},\dots,x_{n}] . The analytic spread, \ell(I)=\dim R[It]\otimes_{R}K , is an invariant which captures information about the generators of the powers of I . Notably, \ell(I) gives the minimal number of generators in a reduction of I . In [ 1 ] , Bivià-Ausina proves that the analytic spread of a monomial ideal can be determined from its Newton polyhedron, \operatorname{NP}(I) . In particular, the analytic spread is one higher than the maximum dimension of a compact face of \operatorname{NP}(I) . This is a very powerful tool for computing analytic spread, as it allows one to use convex geometric methods to solve what would otherwise be algebraic questions. When given a sketch of NP(I), one can immediately identify the analytic spread. However, when NP(I) is implicitly given by halfspaces and hyperplanes, it is often complicated to determine which faces of the Newton polyhedron are compact. In this paper, we show that the analytic spread of any monomial ideal can be characterized in terms of its Newton polytope, \operatorname{np}(I) , and consider applications of this characterization. In Section 2, we give an overview of necessary background information. In Section 3, we introduce and prove our characterization of the analytic spread of monomial ideals. Theorem 1 (Theorem 3.4 ) . Let I be a monomial ideal, and let \operatorname{np}(I) have hyperplanes defined by \begin{array}[]{c}\mathbf{w}_{1}\cdot\mathbf{u}=b_{1}\\ \mathbf{w}_{2}\cdot\mathbf{u}=b_{2}\\ \vdots\\ \mathbf{w}_{s}\cdot\mathbf{u}=b_{s}\end{array} and halfspaces defined by \begin{array}[]{c}\mathbf{h}_{1}\cdot\mathbf{u}\leq c_{1}\\ \mathbf{h}_{2}\cdot\mathbf{u}\leq c_{2}\\ \vdots\\ \mathbf{h}_{t}\cdot\mathbf{u}\leq c_{t}\end{array}. Suppose that there exist \alpha_{1},\dots,\alpha_{s}\in\mathbb{R} and \beta_{1},\dots,\beta_{t}\in\mathbb{R}_{\geq 0} such that \mathbf{W}:=\sum_{i=1}^{s}\alpha_{i}\mathbf{w}_{i}+\sum_{i=1}^{t}\beta_{j}\mathbf{h}_{i} has all negative entries. Then \ell(I)\leq n+1-(s+k) where k is the minimal number of non-zero \beta_{j} required to achieve \mathbf{W} with all negative entries. In many cases, this provides an easier method of computing analytic spread than determining the compact faces of the Newton polyhedron. In Section 4, we use this result to give a characterization of basic monomial ideals. Theorem 2 (Theorem 4.5 ) . Let I be a monomial ideal in K[x_{1},\ldots,x_{n}] , and assume \operatorname{np}(I) lies on the s hyperplanes, \{\mathbf{w}_{i}\cdot\mathbf{u}=b_{i}:i=1,\ldots,s\} , where \mathbf{w}_{i}=(a_{i,1},\ldots,a_{i,n}) , presented so that the s\times n matrix of \mathbf{w}_{i} ’s is in reduced row echelon form. Then I is basic if and only if all of the following conditions hold: (1) \mu(I)\leq n , (2) s=n-\mu(I)+1 , (3) for every j\in\{1,\ldots,n\} , there exists i_{j} such that a_{i_{j},j}>0 . In Section 5, we demonstrate applications of Theorem 1 . We show that the inequality of 1 is an equality for all Newton polyhedra in three dimensions, and we compute the analytic spread of two families of monomial ideals. In particular, for monomial ideals with with r disjointly generated associated primes, we show that \ell(I)=n-r+1 . For monomial ideals that are intersections of two prime powers, we show that \ell(I)=n-1 . 2. Background In this paper, all rings are commutative and unital. Definition 2.1 . Let R be a ring, and let J\subseteq I be ideals. We call J a reduction of I if, for some n\in{\mathbb{Z}}_{>0} , JI^{n}=I^{n+1} . Example 2.2 . Let R=K[x,y] , J=(x^{m},y^{m}) , and I=(x,y)^{m} for any m . It is clear that J\subseteq I is a reduction since JI=I^{2} . A reduction of an ideal shares many traits with the original. For example, an ideal and its reduction have the same radical, minimal primes, and height [ 6 ] . Importantly, an ideal J\subseteq I is a reduction if and only if \overline{J}=\overline{I} . Definition 2.3 . Let J be a reduction of I , then J is a minimal reduction of I if it contains no proper reductions of I . If I has no reductions other than itself, then I is called basic . A main goal of this paper is to classify precisely when certain families of monomial ideals are basic. To do this, we need the following: Definition 2.4 . Let I\subseteq R be an ideal. The Rees algebra of I is \mathcal{R}(I)=\bigoplus_{n=1}^{\infty}I^{n}t^{n}. If R is local (or graded) with maximal ideal (or unique maximal homogeneous ideal) \mathfrak{m} , the analytic spread is \ell(I)=\dim\left(\mathcal{R}(I)\otimes_{R}\left.\raise 1.0pt\hbox{$R$}\right/\hskip-2.0pt\raise-3.0pt\hbox{$\mathfrak{m}$}\right)=\dim\left(\bigoplus_{n=1}^{\infty}\frac{I^{n}}{\mathfrak{m}I^{n}}t^{n}\right), where \dim(M) is the Krull dimension of M . An important fact that we will use later in the paper is that \ell(I)\leq\dim R . While not immediately apparent, there is a critical relation between analytic spread and reductions of ideals: Theorem 2.5 (Corollary 8.3.6, Proposition 8.3.7 [ 6 ] ) . Let R be a local (or graded) ring, and let I be a (homogeneous) ideal. Then, if the residue field of R is infinite and J\subseteq I is a reduction, then J is a minimal reduction of I if and only if \mu(J)=\ell(I) . As an immediate consequence, Corollary 2.6 . Let R be a local (or graded) ring with infinite residue field, and let I be a (homogeneous) ideal. Then I is basic if and only if \mu(I)=\ell(I) . We now introduce a few tools to help work with monomial ideals. Definition 2.7 . Let R=K[x_{1},\ldots,x_{n}] , and let I be a monomial ideal. The Newton polytope of I is \operatorname{np}(I):=\operatorname{conv}\{\mathbf{b}\in{\mathbb{Z}}_{\geq 0}^{n}:{\mathbf{x}}^{\mathbf{b}}\in G(I)\}, where G(I) is the minimal generating set of monomials for I , \mathbf{b}=(b_{1},\ldots,b_{n}) , and \mathbf{x}^{\mathbf{b}}=x_{1}^{b_{1}}\cdots x_{n}^{b_{n}} . The vector \mathbf{b} is called the exponent vector . Define the Newton polyhedron of I as \operatorname{NP}(I):=\operatorname{conv}\{\mathbf{b}\in{\mathbb{Z}}_{\geq 0}^{n}:{\mathbf{x}}^{\mathbf{b}}\in I\}. Equivalently, \operatorname{NP}(I)=\operatorname{np}(I)+{\mathbb{R}}_{\geq 0}^{n}, where the above addition is the Minkowski sum. An important fact about \operatorname{NP}(I) is that a monomial \mathbf{x}^{\mathbf{b}}\in\overline{I} if and only if \mathbf{b}\in\operatorname{NP}(I) . Thus, a monomial reduction of an ideal will have the same Newton polyhedron, but not necessarily the same Newton polytope. As with any convex polyhedron, the Newton polytope and Newton polyhedron can be expressed as an intersection of a set of hyperplanes and halfspaces. In particular, a minimal set of the hyperplanes on which the polytope lies is unique up to linear combinations. Also, halfspaces can be chosen so that the intersection between the polytope and the supporting hyperplane of the halfspace is a facet of the polytope. We call such a set of halfspaces, facet-defining . A remarkable result from Bivià-Ausina [ 1 ] relates analytic spread with the Newton polyhedron: Theorem 2.8 (Bivià-Ausina [ 1 ] ) . Let I be a monomial ideal. Then \ell(I) is equal to one more than the largest dimension of a compact face of \operatorname{NP}(I) . Example 2.9 . Let R=K[x,y,z] , and let I_{1}=(xy,xz,yz) and I_{2}=(xy,y^{4}z^{4},z^{4}x^{4}) . Using the notation that (u,v,w) is the exponent vector corresponding with x^{u}y^{v}z^{w} , the bounding halfspaces for the Newton polyhedra are as follows: \operatorname{NP}(I_{1})=\begin{cases}u+v\geq 1\\ u+w\geq 1\\ v+w\geq 1\\ u+v+w\geq 2\\ u,v,w\geq 0,\end{cases}\quad\text{and}\quad\operatorname{NP}(I_{2})=\begin{cases}u+3v\geq 4\\ 3u+v\geq 4\\ 4u+w\geq 4\\ 4v+w\geq 4\\ u,v,w\geq 0\end{cases} Figures 1 and 2 are depictions of \operatorname{NP}(I_{1}) and \operatorname{NP}(I_{2}) respectively, as well as the corresponding Newton polytopes shaded.