🔢 mathematics

Complexity of the Zero Set of a Matrix Schubert Ideal

This paper investigates the complexity of torus-fixed affine subvarieties within matrix Schubert varieties, demonstrating that for a fixed dimension $n$ , the possible complexity values range from 0 to $(n-1)(n-3)$ with the sole exception of 1.

Original authors: Laura Escobar, Cesar Meza

Published 2026-05-27

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Laura Escobar, Cesar Meza

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you have a giant grid of numbers, like a spreadsheet, but instead of just numbers, you are looking at all the possible ways you can fill this grid while following specific rules. Mathematicians call these rules "rank conditions," which basically mean certain smaller blocks of numbers inside your big grid can't get too "complicated" or "full."

This paper is about exploring the shapes formed by these grids, which the authors call Matrix Schubert varieties. Think of these shapes as complex, multi-dimensional landscapes.

The Main Characters: The Grid and the Torus

The Grid ( $X_w$ ): Imagine a specific type of spreadsheet defined by a permutation (a specific reordering of numbers like 1, 2, 3 becoming 3, 1, 2). This spreadsheet has rules about how many independent rows or columns its sub-blocks can have.
The Torus ( $T$ ): Now, imagine a magical set of tools that can stretch or shrink the rows and columns of your spreadsheet independently, but in a very specific, balanced way. In math, this is called a "torus action." It's like having a remote control that can zoom in on specific parts of your grid without breaking the rules.
The "Useless" Space ( $C^k$ ): Sometimes, your grid has a huge section that is completely free and empty. You can fill this section with anything, and it doesn't change the core rules. The authors realized that to understand the true shape, they needed to cut this empty, free-floating section out. What remains is a smaller, tighter shape called $Y_w$ .

The Concept of "Complexity"

The paper introduces a concept called complexity. Think of this as a measure of how "wild" or "unpredictable" the shape is when you play with those magical stretching tools.

Complexity 0 (The Tame Ones): If the complexity is 0, the shape is very orderly. It's like a perfectly symmetrical crystal or a simple geometric solid. Mathematicians call these "toric varieties," and they are easy to map out using simple shapes like triangles and squares (polytopes).
High Complexity (The Wild Ones): If the complexity is high, the shape is chaotic. The stretching tools can twist it in so many different directions that it becomes very hard to describe with simple maps.

The authors wanted to answer a simple question: For a grid of a fixed size (say, $n \times n$ ), what are all the possible "wildness levels" (complexity numbers) we can find?

The Big Discovery

The authors found a very specific pattern in the answers:

The Maximum Wildness: For a grid of size $n$ , there is a specific "most chaotic" shape possible. The authors calculated exactly how chaotic it can get: the maximum complexity is $(n-1)(n-3)$ .
- Analogy: If your grid is $5 \times 5$ , the most chaotic shape you can make has a complexity of 8. If it's $10 \times 10$ , the max is 63.
- They also found the exact rule (permutation) that creates this most chaotic shape. It's a very specific, slightly messy reordering of numbers.
The Missing Number (The "1" Problem): The authors discovered that you can create shapes with complexity 0, 2, 3, 4, and so on, all the way up to the maximum.
- But there is one gap: You cannot create a shape with complexity 1.
- Analogy: Imagine a staircase where you can step on the ground (0), then skip a step and land on 2, then 3, 4, etc. You can never land on step 1. It's a mathematical "gap" that simply doesn't exist for these shapes.
Filling the Gap: They proved that for any size grid (as long as it's big enough, $n \ge 4$ ), you can find a shape for every complexity number you want, except for that missing 1.

How They Did It (The "Lego" Method)

To prove they could hit every number, they used a clever construction method:

They started with the "most chaotic" shape (the maximum complexity).
They showed that by swapping out a small corner of the rules for a simpler set of rules (like swapping a complex Lego block for a simpler one), they could reduce the complexity by exactly the right amount.
By doing this repeatedly, they could "dial down" the complexity from the maximum all the way down to 0, hitting every integer in between.

Summary

In short, this paper maps out the "chaos levels" of a specific type of mathematical shape. They found that:

There is a hard limit to how chaotic these shapes can get.
You can achieve almost every level of chaos below that limit.
The only level you can never achieve is 1. It's a unique, forbidden number in this mathematical universe.

This helps mathematicians understand the structure of these shapes better, knowing exactly which "wildness levels" are possible and which are impossible.

Technical Summary: Complexity of the Zero Set of a Matrix Schubert Ideal

Problem Statement
The paper investigates the combinatorial complexity of specific subvarieties within the framework of $T$ -varieties. A $T$ -variety is a normal variety equipped with an action of an algebraic torus $T$ . The complexity of such a variety is defined as the difference between its dimension and the dimension of a maximal $T$ -orbit (or equivalently, $\dim(X) - \dim(\sigma)$ , where $\sigma$ is the weight cone). Lower complexity indicates that the variety is more amenable to combinatorial methods; for instance, complexity 0 corresponds to toric varieties.

The authors focus on matrix Schubert varieties $X_w$ , introduced by Fulton, which are affine varieties of $n \times n$ matrices defined by rank conditions on submatrices determined by a permutation $w \in S_n$ . While $X_w$ carries a natural $T \times T$ action (where $T \simeq (\mathbb{C}^*)^n$ ), its complexity is often trivialized by a large affine factor $\mathbb{C}^k$ . To address this, the authors study the "essential" part of the variety, denoted $Y_w$ , defined by the decomposition $X_w \cong Y_w \times \mathbb{C}^k$ where $k$ is maximal. The defining ideal of $Y_w$ coincides with that of $X_w$ .

The central problem is to characterize the set of nonnegative integers that can be realized as the complexity of $Y_w$ for a fixed $n \geq 4$ . Previous work established that complexity 1 is impossible and that the set of achievable complexities excludes only 1 when $n$ varies over multiples of 4. This paper restricts the scope to a fixed symmetric group $S_n$ to determine the precise range of achievable complexities.

Methodology
The authors employ a combination of algebraic geometry, combinatorics, and graph theory:

Geometric Decomposition: They define $Y_w$ as the projection of $X_w$ onto the linear subspace spanned by entries not in the "northwest diagram" $NW(w)$. The dimension of $Y_w$ is given by $|L'(w)|$ , where $L'(w)$ is a specific skew diagram derived from the Rothe diagram $D(w)$ of the permutation.
Torus Action and Weight Cones: The complexity $d_w$ of $Y_w$ under the $T \times T$ action is calculated using the formula:
$d_w = \dim(Y_w) - \dim(\sigma_w)$
where $\sigma_w$ is the weight cone associated with the action.
Graph-Theoretic Translation: The weight cone $\sigma_w$ is identified as the edge cone of a directed bipartite graph $G_w$ . The vertices of $G_w$ correspond to the nonempty rows and columns of the diagram $L(w)$ , and edges correspond to the entries in $L(w)$ .
Complexity Formula: Using the dimension formula for edge cones, the complexity is expressed as:
$d_w = |L'(w)| - |V(G_w)| + |C(G_w)|$
where $|V(G_w)|$ is the number of vertices and $|C(G_w)|$ is the number of connected components. This is further refined using the cyclomatic number $\nu(G_w)$ of the underlying undirected graph:
$d_w = \nu(G_w) + |\text{dom}(w)| - |D(w)|$
where $\text{dom}(w)$ is the "dominant piece" of the Rothe diagram.
Constructive Proofs: To establish the range of achievable complexities, the authors utilize lemmas regarding the modification of Rothe diagrams. Specifically, they show how replacing a region of a permutation's diagram with the diagram of a smaller permutation affects the complexity, allowing them to construct permutations with specific complexity values.

Key Contributions and Results

Maximum Complexity: The paper determines the maximum possible complexity for $Y_w$ with $w \in S_n$ (for $n \geq 4$ ).
- Theorem 4.2: The maximum complexity is $(n-1)(n-3)$ .
- Uniqueness: This maximum is uniquely achieved by the permutation $s_{n-1} = [1, \dots, n-2, n, n-1]$ . The associated graph $G_w$ for this permutation is isomorphic to the complete bipartite graph $K_{n-1, n-1}$ .
Achievable Complexity Values: The paper characterizes the set of all integers $d$ that can be realized as the complexity of some $Y_w$ for a fixed $n$ .
- Theorem 4.8: For a fixed $n \geq 4$ , the set of achievable complexities is $\{0, 2, 3, \dots, (n-1)(n-3)\}$ .
- Exclusion of 1: The paper confirms that complexity 1 is never achievable for any $w \in S_n$ (relying on prior results by Escobar–Mészáros and Donten-Bury–Escobar–Portakal).
- Gap Filling: The authors demonstrate that every integer between 0 and the maximum (excluding 1) is achievable. They handle the specific case of complexity 4 for $n=5$ (which was not covered by their general interval construction) by explicitly exhibiting the permutation $12534$.

Significance
The paper provides a complete classification of the complexity spectrum for matrix Schubert varieties (specifically their essential parts $Y_w$ ) within a fixed symmetric group $S_n$ . By establishing that the complexity can take any integer value from 0 up to the theoretical maximum $(n-1)(n-3)$ , with the sole exception of 1, the authors clarify the combinatorial landscape of these varieties. This result refines previous knowledge which was limited to cases where $n$ varied over multiples of 4. The work connects the geometric properties of determinantal varieties to the structural properties of bipartite graphs (specifically cyclomatic numbers and edge cones), offering a precise combinatorial tool for analyzing the $T$ -variety structure of matrix Schubert varieties.

The Main Characters: The Grid and the Torus

The Concept of "Complexity"

The Big Discovery

How They Did It (The "Lego" Method)

Summary

More like this