REPETITIONS OF PAK-STANLEY LABELS IN G-SHI ARRANGEMENTS CARA BENNETT LUCY MARTINEZ AVA MOCK GORDON ROJAS KIRBY AND ROBIN TRUAX Abstract. Given a simple graph G one can dene a hyperplane arrangement called the G-Shi

2025-04-29 0 0 624.39KB 34 页 10玖币
侵权投诉
REPETITIONS OF PAK-STANLEY LABELS IN G-SHI ARRANGEMENTS
CARA BENNETT, LUCY MARTINEZ, AVA MOCK, GORDON ROJAS KIRBY, AND ROBIN TRUAX
Abstract.
Given a simple graph
G
, one can define a hyperplane arrangement called the
G
-Shi
arrangement. The Pak-Stanley algorithm labels the regions of this arrangement with
G
-parking
functions. When
G
is a complete graph we recover the full Shi arrangement, and the Pak-Stanley
labels give a bijection with ordinary parking functions. However, for proper subgraphs
GKn
,
while the Pak-Stanley labels still include every
G
-parking function, some appear more than once.
These repetitions of Pak-Stanley labels are a topic of interest in the study of
G
-Shi arrangements and
G
-parking functions. Furthermore,
G
-parking functions are connected to many other combinatorial
objects (for example, superstable configurations in chip-firing). In studying these repetitions, we
can draw on existing results about these objects such as Dhar’s Burning Algorithm. Conversely, our
results have implications for the study of these objects as well.
The key insight of our work is the introduction of a combinatorial model called the Three Rows
Game. Analyzing the histories of this game and the ways in which they can induce the same
outcomes allows us to characterize the multiplicities of the Pak-Stanley labels. Using this model, we
develop a classification theorem for the multiplicities of the Pak-Stanley labels of the regions in the
Pn
-Shi arrangement, where
Pn
is the path graph on
n
vertices. Then, we generalize the Three Rows
Game into the
T
-Three Rows Game. This allows us to study the multiplicities of the Pak-Stanley
labels of the regions in
T
-Shi arrangements, where
T
is any tree. Finally, we discuss the possibilities
and difficulties in applying our method to arbitrary graphs. In particular, we analyze multiplicities
in the case when
G
is a cycle graph, and prove a uniqueness result for maximal
G
-parking functions
for all graphs using the Three Rows Game.
1. Introduction
Ahyperplane in
Rn
is an affine subspace of dimension
n
1. A hyperplane arrangement is a collection
of finitely many hyperplanes. Given a hyperplane arrangement
A
, its complement
Rn\SH∈A H
splits into connected components called regions. The focus of this paper is on the combinatorial
properties of the regions of the
G
-Shi arrangement and their labels by
G
-parking functions, as
defined by Duval, Klivans, and Martin [2]. The
G
-Shi arrangement
S
(
G
) is defined for any graph
G= (V, E) by
S(G) = {xixj= 0,1| {i, j} ∈ Ewith i<j}.
When
G
=
Kn
, the
G
-Shi arrangement is the Shi arrangement, which has (
n
+ 1)
n1
regions. See
[3] for a survey of the Shi arrangement.
Since the regions of the Shi arrangement are equinumerous with parking functions of length
n
, a
natural problem is to find a bijection between regions of the Shi arrangement and parking functions
of length
n
. Pak and Stanley (1996) and Athanasiadis and Linusson (1999) gave two such bijections
[6],[1]. The Pak-Stanley algorithm, defined later, labels the regions of the
G
-Shi arrangement with
G
-parking functions in such a way that every
G
-parking function appears. When
G
is complete,
each
G
-parking function appears exactly once as a Pak-Stanley label on a region of the
G
-Shi
arrangement. If
G
is not complete, some
G
-parking functions appear more than once as Pak-Stanley
labels in the
G
-Shi arrangement. Consider the examples of
G
=
K3
and
G
=
P3
in Figures 1a and
1b respectively.
Date: October 26, 2022.
1
arXiv:2210.13613v1 [math.CO] 24 Oct 2022
(0,0,0)
(0,1,0) (0,0,1)
(1,0,0)
(1,1,0)
(2,0,0) (1,0,1)
(0,0,2)
(0,1,1)(0,2,0)
(2,1,0)
(2,0,1)
(1,0,2)
(0,1,2)
(0,2,1)
(1,2,0)
(a) Labelling the K3-Shi arrangement.
(0,1,1) (0,0,1) (1,0,1)
(0,1,0) (0,0,0) (1,0,0)
(0,2,0) (0,1,0) (1,1,0)
(b) Labelling the P3-Shi arrangement.
Figure 1. Example Pak-Stanley Labels for G-Shi Arrangements
Figures 1a and 1b depict
G
-Shi arrangements. Notice that on the left, each label appears once
(has multiplicity 1), but on the right, the label (0
,
1
,
0) appears twice (has multiplicity 2). In this
paper, we work towards a characterization of the “multiplicities” of the Pak-Stanley labels in
G
-Shi
arrangements by introducing a combinatorial model called the Three Rows Game defined in Section
4.2
In Section 3, we introduce the Shi adjacency digraph to describe the adjacencies of regions in the
G
-Shi arrangement and their relationship to the Pak-Stanley labels of these regions. Section 4
introduces the Three Rows game, a combinatorial model for analyzing multiplicities of Pak-Stanley
labels in
Pn
-Shi arrangements for path graphs
Pn
. In particular, we prove the following result to
characterize multiplicities of the Pak-Stanley labels in G-Shi arrangements for path graphs.
Theorem
(Path Multiplicity Theorem)
.
Suppose
p
= (
p1, . . . , pn
) is a Pak-Stanley label of a region
of
S
(
Pn
). A run
r
of length
k
in
p
is a section of
p
of the form (0
,
1
,...,
1
,
0) with
k
1s. If the
length of a run
r
is denoted
`
(
r
), then the multiplicity of the label
p
in in the
Pn
-Shi arrangement is
µ(p) = Y
runs rin p
(`(r) + 1).
Besides analyzing path graphs, we also study multiplicities of Pak-Stanley labels in
G
-Shi arrange-
ments for all trees in Section 5, for example characterizing multiplicities for star graphs with a similar
theorem. Then in Section 6, we discuss how the Three Rows Game can be played on all graphs,
not just on trees, and the added complexity of such an extension. In particular, we completely
characterize multiplicities of the Pak-Stanley labels in
G
-Shi arrangements for cycle graphs with an
analogous Cycle Multiplicity Theorem. Finally, we prove the following fact about multiplicities of
the maximal labels of regions of G-Shi arrangements.
Corollary. Let Gbe a graph and pa maximal G-parking function. Then phas multiplicity 1.
2
2. Background
We begin with a discussion of some prerequisite topics: the
G
-Shi arrangement, the Pak-Stanley
algorithm, chip-firing, and superstable configurations. An important note is that for the remainder
of the paper, all of our graphs are assumed to be connected, finite, and simple.
2.1.
The G-Shi Arrangement and the Pak-Stanley Algorithm.
First, we formally define the
G-Shi arrangement.
Definition 2.1
(Shi Arrangement)
.
The Shi arrangement
Sn
is the hyperplane arrangement in
Rn
with hyperplanes xixj= 0 and xixj= 1 for each i, j ∈ {0,1,2, . . . , n 1}with i<j.
Definition 2.2
(
G
-Shi Arrangement)
.
Given a graph
G
= (
V, E
) with
V
=
{
0
, . . . , n
1
}
, the
G
-Shi arrangement
S
(
G
) is the hyperplane arrangement in
Rn
with hyperplanes
xixj
= 0 and
xixj
= 1 for each
{i, j} ∈ E
with
i < j
[4]. In this case,
G
is called the defining graph of the
arrangement.
It is important to note that the
G
-Shi arrangement depends on the labelling 0
, . . . , n
1 of the
vertices of
G
. Indeed, consider the isomorphic graphs
K4\ {
0
,
1
}
and
K4\ {
0
,
2
}
. The former has a
G
-Shi arrangement with 84 regions, whereas the latter has a
G
-Shi arrangement with 85 regions.
Over the course of this paper, we will demonstrate that in certain special cases the multiplicities do
not depend on the labelling of the vertices.
Figure 1a illustrates a projection of the Shi arrangement onto the hyperplane
x0
+
x1
+
x2
= 0,
allowing us to visualize it in 2 dimensions without losing any regions or adjacency relations. This
projection trick is also used in Figure 1b.
Next, we discuss
G
-parking functions and the Pak-Stanley algorithm (which surjectively assigns
G-parking functions to the regions of the G-Shi arrangement).
Definition 2.3
(Outdegree)
.
Given a graph
G
= (
V, E
), a subset
S
of
V
, and a vertex
v
, the
outdegree
outdegS
(
v
)of
v
with respect to
S
is the number of edges from
v
to vertices outside
S
. In
particular, outdeg(v) is the outdegree of v(which equals deg(v) when Gis undirected).
Definition 2.4
(
G
-Parking Function)
.
Let
G
be an undirected graph on vertices
V
=
{
0
,
1
, . . . , n
1
, q}
. In this case,
q
is called the sink. A
G
-parking function is an
n
-tuple (
a0, a1, . . . , an1
) such
that for any non-empty subset S⊆ {0, . . . , n 1}, there exists vSsuch that av<outdegS(v).
Definition 2.5
(
G
)
.
Let
G
= (
V, E
) be a graph. Then,
G
is the graph obtained from
G
by
adding a sink vertex qand an edge between qand each vV.
Next, we will discuss the Pak-Stanley algorithm [6], whose behavior is the central topic of our paper.
Definition 2.6
(Pak-Stanley Algorithm)
.
The Pak-Stanley algorithm maps
G
-parking functions
to the regions of the
G
-Shi arrangement [6]. It assigns each region
R
an
n
-tuple
λ
(
R
) of nonnegative
integers, called its Pak-Stanley label as follows:
(1)
The region in which
x0> x1>· · · > xn1
and
x0xn1<
1 is called the base region and
denoted R0. Define λ(R0) = (0,0,...,0).
(2) Suppose λ(R) has been defined, and that R0is a region such that
(a) Rand R0share a boundary facet, which is part of a hyperplane HS(G).
3
(b) Rlies in the same half-space of Has R0.
In this case, we define λ(R0) = (λ(R) + eiif His given by xixj= 0 with i < j,
λ(R) + ejif His given by xixj= 1 with i < j.
Here, eidenotes the ith standard basis vector of Rn.
The set of Pak-Stanley labels for the regions of the
G
-Shi arrangement are called the Pak-Stanley
labels for G; these are the same as the G-parking functions [4].
The examples of
G
=
K3
and
G
=
P3
are displayed in Figures 1a and 1b. Notice every (
K3
)
-parking
function appears in Figure 1a, and every (
P3
)
-parking function appears in Figure 1b. Furthermore,
each (
K3
)
-parking function appears exactly once in the
K3
-Shi arrangement. As it turns out, both
of these are general patterns that are proved in [4]. Specifically, we have the following:
Theorem 2.7
([4], Corollary 2.8)
.
Every
G
-parking function occurs as a label in the Pak-Stanley
algorithm on the G-Shi arrangement.
It is well known that there are (
n
+ 1)
n1
regions in the Shi arrangement
Sn
. Similarly, there
are (
n
+ 1)
n1
parking functions (and therefore (
n
+ 1)
n1
(
Kn
)
-parking functions). Thus, every
(
Kn
)
-parking function occurs as a label precisely once in the Pak-Stanley algorithm on the
Kn
-Shi
arrangement.
2.2.
Chip-Firing, Superstable Configurations, and Dhar’s Burning Algorithm.
We begin
this section by recounting some of the definitions in Chapter 2 of The Mathematics of Chip-firing
by Klivans [5]. For more detail, we refer the reader to this book.
In this section, G= (V, E) is a graph on n+ 1 vertices with sink vertex q.
Definition 2.8 (Chip Configuration).Achip configuration for Gis a non-negative integer vector
c= (c0, c1, . . . , cn1)Zn
0
where each coordinate of the vector corresponds to the number of “chips” at a particular vertex.
That is, the
i
th coordinate
ci
represents the number of chips at the vertex
vi
. Notice that we do not
consider the number of chips on the sink vertex q.
Definition 2.9
(Firing)
.
Given a chip configuration
c
for
G
, a non-sink vertex
v
fires by sending
one chip to each of its neighbors (possibly including the sink). That is, the configuration
c
is
replaced by c0= (c0
0, . . . , c0
n1)Znwhere
c0
i=
cideg i i =v
ci+ 1 {i, v} ∈ E
ciotherwise.
This is a legal fire if c0is a chip configuration.
Definition 2.10 (Stable).A chip configuration cis stable if there are no legal fires.
Now let us transition to discussing superstable configurations.
4
Definition 2.11
(Graph Laplacian)
.
Let
G
= (
V, E
) be a graph on
n
vertices
v0, . . . , vn1
. The
graph Laplacian ∆(G) is the n×nmatrix given by
ij =
deg(vi)i=j
1i6=jand {vi, vj} ∈ E
0 otherwise.
Definition 2.12
(Reduced Laplacian)
.
Let ∆ be the Laplacian of a graph
G
with sink
q
. Then,
the reduced graph Laplacian of
G
with respect to
q
, denoted ∆
q
(
G
) or ∆
q
, is the matrix obtained
from ∆ by deleting the row and column corresponding to q.
Definition 2.13
(Cluster-Fire)
.
Let
G
be a graph on
n
+ 1 vertices with sink
q
. Consider a chip
configuration
c
= (
c0, . . . , cn1
). Let
S
be a subset of the vertices of
G
. A cluster-fire at
S
replaces
c
with a new configuration
c0
given by sending a chip to each neighbor of
v
for each
vS
. Formally,
c0=cq(G)χS
where
χSRn
is the characteristic vector of
S
; that is,
χS
has
i
th coordinate equal to 1 if
iS
and 0 if i6∈ S. Such a cluster-fire is called legal if c0is still a chip configuration.
Definition 2.14
(Superstable Configuration)
.
Let a graph
G
be given with a chip configuration
c
.
The configuration cis called superstable if there are no legal cluster-fires from c.
Example 2.15.
The left-hand configuration is not superstable, as one can fire at the three vertices
whose chip labels are bolded. However, the right-hand side is superstable.
Figure 2. Non-Superstable and Superstable Configurations.
Now, the main reason why superstable configurations are central to our paper is the following result:
Theorem 2.16
([5], Theorem 3.6.3)
.
If
G
is a graph with sink vertex
q
, then the
G
-parking
functions of Gare precisely the set of superstable configurations of G.
Proof. By definition; see Definitions 2.4 and 2.14.
We conclude with a discussion of some important results about superstable configurations. First,
we will discuss Dhar’s Burning Algorithm.
Definition 2.17
(Dhar’s Burning Algorithm)
.
Let
G
be a graph with sink
q
, and
c
be a chip
configuration on
G
. Envision the chips as firefighters protecting the vertices that they are on. Then,
light the sink qon fire, and repeat the following process:
(1) If a vertex vis on fire, the fire spreads along all the edges of vtowards v’s neighbors.
5
摘要:

REPETITIONSOFPAK-STANLEYLABELSING-SHIARRANGEMENTSCARABENNETT,LUCYMARTINEZ,AVAMOCK,GORDONROJASKIRBY,ANDROBINTRUAXAbstract.GivenasimplegraphG,onecande neahyperplanearrangementcalledtheG-Shiarrangement.ThePak-StanleyalgorithmlabelstheregionsofthisarrangementwithG-parkingfunctions.WhenGisacompletegraph...

展开>> 收起<<
REPETITIONS OF PAK-STANLEY LABELS IN G-SHI ARRANGEMENTS CARA BENNETT LUCY MARTINEZ AVA MOCK GORDON ROJAS KIRBY AND ROBIN TRUAX Abstract. Given a simple graph G one can dene a hyperplane arrangement called the G-Shi.pdf

共34页,预览5页

还剩页未读, 继续阅读

声明:本站为文档C2C交易模式,即用户上传的文档直接被用户下载,本站只是中间服务平台,本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私,请立即通知玖贝云文库,我们立即给予删除!
分类:图书资源 价格:10玖币 属性:34 页 大小:624.39KB 格式:PDF 时间:2025-04-29

开通VIP享超值会员特权

  • 多端同步记录
  • 高速下载文档
  • 免费文档工具
  • 分享文档赚钱
  • 每日登录抽奖
  • 优质衍生服务
/ 34
客服
关注