Characterizations of graph classes by forbidden configurations

Similar documents
Vertex coloring, chromatic number

Vertex coloring, chromatic number

Colouring graphs with no odd holes

Graphs and Discrete Structures

The Structure of Bull-Free Perfect Graphs

MATH 350 GRAPH THEORY & COMBINATORICS. Contents

How many colors are needed to color a map?

The Six Color Theorem

The following is a summary, hand-waving certain things which actually should be proven.

Matching Algorithms. Proof. If a bipartite graph has a perfect matching, then it is easy to see that the right hand side is a necessary condition.

Mock Exam. Juanjo Rué Discrete Mathematics II, Winter Deadline: 14th January 2014 (Tuesday) by 10:00, at the end of the lecture.

Math 443/543 Graph Theory Notes 11: Graph minors and Kuratowski s Theorem

Treewidth and graph minors

Small Survey on Perfect Graphs

Vertex Cover is Fixed-Parameter Tractable

Graphs with no 7-wheel subdivision Rebecca Robinson 1

Math 777 Graph Theory, Spring, 2006 Lecture Note 1 Planar graphs Week 1 Weak 2

Immersion in four-edge-connected graphs

Jordan Curves. A curve is a subset of IR 2 of the form

Coloring graphs without fan vertex-minors and graphs without cycle pivot-minors

HW Graph Theory Name (andrewid) - X. 1: Draw K 7 on a torus with no edge crossings.

A taste of perfect graphs (continued)

Problem Set 3. MATH 776, Fall 2009, Mohr. November 30, 2009

K 4 C 5. Figure 4.5: Some well known family of graphs

Lecture Notes on Graph Theory

Ma/CS 6b Class 11: Kuratowski and Coloring

Planarity: dual graphs

1 Matching in Non-Bipartite Graphs

Kyle Gettig. Mentor Benjamin Iriarte Fourth Annual MIT PRIMES Conference May 17, 2014

Fast Skew Partition Recognition

Computing Linkless and Flat Embeddings of Graphs in R 3

The clique number of a random graph in (,1 2) Let ( ) # -subgraphs in = 2 =: ( ) We will be interested in s.t. ( )~1. To gain some intuition note ( )

Math 575 Exam 3. (t). What is the chromatic number of G?

Maximum number of edges in claw-free graphs whose maximum degree and matching number are bounded

November 14, Planar Graphs. William T. Trotter.

The clique number of a random graph in (,1 2) Let ( ) # -subgraphs in = 2 =: ( ) 2 ( ) ( )

arxiv: v3 [cs.ds] 26 Sep 2013

Diskrete Mathematik und Optimierung

Math 776 Graph Theory Lecture Note 1 Basic concepts

Abstract. A graph G is perfect if for every induced subgraph H of G, the chromatic number of H is equal to the size of the largest clique of H.

The k-in-a-path problem for claw-free graphs

arxiv: v1 [cs.dm] 13 Apr 2012

γ(ɛ) (a, b) (a, d) (d, a) (a, b) (c, d) (d, d) (e, e) (e, a) (e, e) (a) Draw a picture of G.

Excluding induced subgraphs

Planar graphs. Chapter 8

Jordan Curves. A curve is a subset of IR 2 of the form

2. Lecture notes on non-bipartite matching

Vertex 3-colorability of claw-free graphs

On the Relationships between Zero Forcing Numbers and Certain Graph Coverings

Claw-Free Graphs With Strongly Perfect Complements. Fractional and Integral Version.

Math 3012 Applied Combinatorics Lecture 12

Excluding induced subgraphs

MAT 145: PROBLEM SET 6

Diskrete Mathematik und Optimierung

Theorem 3.1 (Berge) A matching M in G is maximum if and only if there is no M- augmenting path.

WUCT121. Discrete Mathematics. Graphs

Math 778S Spectral Graph Theory Handout #2: Basic graph theory

Discrete Wiskunde II. Lecture 6: Planar Graphs

arxiv: v1 [math.co] 4 Apr 2011

arxiv: v4 [math.co] 4 Apr 2011

A generalization of Mader s theorem

Some new results on circle graphs. Guillermo Durán 1

Constructions of k-critical P 5 -free graphs

Adjacent: Two distinct vertices u, v are adjacent if there is an edge with ends u, v. In this case we let uv denote such an edge.

Ma/CS 6b Class 4: Matchings in General Graphs

Advanced Combinatorial Optimization September 17, Lecture 3. Sketch some results regarding ear-decompositions and factor-critical graphs.

Discharging and reducible configurations

INTRODUCTION TO GRAPH THEORY. 1. Definitions

On Cyclically Orientable Graphs

Partial Characterizations of Circular-Arc Graphs

Planar Graphs. 1 Graphs and maps. 1.1 Planarity and duality

Math 776 Homework. Austin Mohr. June 15, Theorem 1.1. The following assertions are equivalent for a graph T : (i) T is a tree;

Diskrete Mathematik und Optimierung

Rigidity, connectivity and graph decompositions

Lecture 3: Recap. Administrivia. Graph theory: Historical Motivation. COMP9020 Lecture 4 Session 2, 2017 Graphs and Trees

Basics of Graph Theory

A graph is finite if its vertex set and edge set are finite. We call a graph with just one vertex trivial and all other graphs nontrivial.

Graph Theory S 1 I 2 I 1 S 2 I 1 I 2

How many colors are needed to color a map?

Discrete mathematics II. - Graphs

Lecture 6: Graph Properties

Discrete mathematics

arxiv: v1 [math.co] 28 Sep 2010

Block Duplicate Graphs and a Hierarchy of Chordal Graphs

Class Six: Coloring Planar Graphs

The strong chromatic number of a graph

Perfect Matchings in Claw-free Cubic Graphs

A structure theorem for graphs with no cycle with a unique chord and its consequences

A Vizing-like theorem for union vertex-distinguishing edge coloring

Assignment 1 Introduction to Graph Theory CO342

Critical and Anticritical Edges with respect to Perfectness

Matching Theory. Figure 1: Is this graph bipartite?

if for every induced subgraph H of G the chromatic number of H is equal to the largest size of a clique in H. The triangulated graphs constitute a wid

On Possible Counterexamples to Negami s Planar Cover Conjecture

On a perfect problem

HOMEWORK 4 SOLUTIONS. Solution: The Petersen graph contains a cycle of odd length as a subgraph. Hence,

Edge colorings. Definition The minimum k such that G has a proper k-edge-coloring is called the edge chromatic number of G and is denoted

Bipartite Roots of Graphs

[Ramalingam, 4(12): December 2017] ISSN DOI /zenodo Impact Factor

Lecture 2 : Counting X-trees

Transcription:

Characterizations of graph classes by forbidden configurations Zdeněk Dvořák September 14, 2015 We consider graph classes that can be described by excluding some fixed configurations. Let us give some examples. Theorem 1 (Kuratowski). A graph G is planar if and only if it contains neither K 5 nor K 3,3 as a topological minor (i.e., does not contain a subgraph isomorphic to a subdivision of K 5 or K 3,3 ). Theorem 2. A graph G is bipartite if and only if it does not contain any odd cycle as a subgraph. Let us give a few definitions. We consider the following partial orders on graphs: induced subgraph H G subgraph H G topological minor H t G (a subdivision of H is a subgraph of G) minor H m G (H can be obtained from a subgraph of G by contracting edges) Note that H G H G H t G H m G. Let be any of these orders. We say that a class G of graphs is -closed (subgraph-closed, minor-closed,... ) if for all graphs H and G, if H G and G G, then H G. Examples: The class of all planar graphs is minor-closed (and thus it also is topological-minor-closed, subgraph-closed and induced-subgraph-closed). 1

The class of all bipartite graphs is subgraph-closed, but not topologicalminor-closed. The class of all graphs whose connected components are cliques is induced-subgraph-closed, but not subgraph-closed. For a set F of graphs, let Forb (F) = {G : ( F F)F G} denote the class of graphs that do not contain any element of F in the sense defined by. Examples: Forb t (K 5, K 3,3 ) = planar graphs. Forb (C 3, C 5, C 7,...) = bipartite graphs. Forb (K 1,2 ) = graphs whose connected components are cliques. A graph F is a -obstruction for a graph class G if F G, but F G for all F F. Let Obst (G) be the set of all -obstructions for G. Examples: K 5 is a topological-minor-obstruction for planar graphs, since K 5 is not planar, but all proper topological minors of K 5 are planar. K 6 is not a topological-minor-obstruction for planar graphs, since K 5 < t K 6 and K 5 is not planar. We say that a partial order on graphs is locally finite if {H : H G} is finite for every graph G. Lemma 3. Let G be a class of graphs, and let be a locally finite order. The following claims are equivalent. (a) G is -closed (b) G = Forb (F) for some set F (c) G = Forb (Obst (G)) Proof. (a) (c) First, suppose that G G. Since G is -closed, every H G satisfies H G, and thus H is not a -obstruction for G. Hence, G Forb (Obst (G)). Consider now any graph G G. The set S = {H G : H G} is finite, and thus it contains a -minimal element F. Then F G, 2

but F G for every F F, i.e., F is a -obstruction for G. Hence G Forb (Obst (G)). Therefore, G = Forb (Obst (G)). (c) (b) Trivial, let F = Obst (G). (b) (a) Consider any G G. Since G = Forb (F), we have F G for every F F. Consequently, if H G, then also F H. Therefore, H Forb (F) = G. Since this holds for every G G and every H G, the class G is -closed. 1 Subgraph-closed classes Let P n denote a path with n vertices, and let tk 2 denote the matching of size t. Simple examples: Forb (C 3, C 4, C 5,...) = forests Forb (C 3, C 5, C 7,...) = bipartite graphs Forb (P 2 ) = isolated vertices Forb (P 3 ) = isolated vertices and edges Forb (K 1,n ) = maximum degree at most n 1 Forb (2K 2 ) = isolated vertices, or a star plus isolated vertices, or a triangle plus isolated vertices. Forb (tk 2 ) is the class of graphs with maximum matching of size at most t 1, which can be described explicitly using Tutte s theorem. The following approximate description is often more useful. A set X V (G) is a vertex cover if every edge of G is incident with a vertex of X, i.e., G X has no edges. Theorem 4. Every graph in Forb (tk 2 ) has a vertex cover of size at most 2(t 1). Conversely, every graph with vertex cover of size at most t belongs to Forb ((t + 1)K 2 ). 3

Proof. Suppose that tk 2 G. Let M G be a maximum matching, E(M) t 1. Then G V (M) has no edges, i.e., V (M) is a vertex cover for G of size at most 2(t 1). Conversely, if X is a vertex-cover of G of size at most t, then every edge of a matching in G intersects X, and thus such a matching has at most X edges. Consequently, (t + 1)K 2 G. We can also obtain a similar approximate characterization for Forb (P n ). The tree-depth of a graph G is the smallest d 1 for that there exists a rooted forest T of depth at most d with vertex set V (G), such that every edge of G joins a vertex with its ancestor or descendant in T. Examples: Graphs of tree-depth 1 consist of isolated vertices. Graphs of tree-depth at most 2 consist of stars. The path P 2 n 1 has tree-depth n, the path P 2 n has tree-depth n + 1. Theorem 5. If P n G, then G has tree-depth at most n 1. Conversely, if G has tree-depth at most n, then P 2 n G. Proof. Suppose that P n G. We can assume that G is connected, as otherwise we consider each component separately. Run depth-first search from any vertex of G, and let T be the resulting tree. Then T G, hence P n T, and thus T has depth at most n 1. Observe also that every edge of G joins a vertex with its ancestor or descendant in T. Conversely, if P 2 n G, then G has at least as large tree-depth as P 2 n, which is n + 1. 2 Induced-subgraph-closed classes For a graph H, let H denote the complement of H, that is the graph with the same vertex set in that two distinct vertices are adjacent if and only if they are not adjacent in H. Let us mention a famous recent result. A graph G is perfect if ω(h) = χ(h) for every H G. Perfect graphs are interesting, since we can determine their chromatic number as well as the size of maximum clique in polynomial 4

time. A hole is a cycle of length at least 4. An anti-hole is a complement of a hole. The following characterization of perfect graphs was proposed by Berge in 1961, and finally proved by Chudnovsky, Robertson, Seymour, and Thomas in 2002. Theorem 6. A graph is perfect if and only if it contains neither odd hole nor an odd anti-hole as an induced subgraph. That is, Forb (C 5, C 7, C 7, C 9, C 9,...) = perfect graphs. Another well-known result concerns line-graphs. A graph G is a linegraph of H if V (G) = E(H), and two vertices of G are adjacent if and only if the corresponding edges of H are incident with the same vertex. Theorem 7. A graph G is a line-graph of some graph if and only if it does not contain any of the following graphs as an induced subgraph: Further examples: Forb (C 3, C 4, C 5,...) = forests Forb (C 3, C 5, C 7,...) = bipartite graphs Forb (P 2 ) = isolated vertices Forb (P 3 ) = all components are cliques Because C 4 is equal to 2K 2, Forb (2K 2 ) contains exactly the complements of graphs in Forb (C 4 ), and in particular complements of all graphs without cycles of length at most 4. The exact description of Forb (2K 2 ) is not known. See homework exercises for some partial results. The description of Forb (K 1,3 ) (claw-free graphs) is known, but it is extremely complicated. 5

3 Exercises 1. ( ) Let G be a -closed class of graphs, where is a locally finite order. Show that Obst (G) F for every set F such that G = Forb (F). 2. ( ) Describe the graphs in Forb (P 4 ). 3. ( ) Prove that P 2 n has tree-depth n + 1. 4. ( ) Prove that Forb (C 3, C 5, C 7,...) = bipartite. 5. ( ) Describe the graphs in Forb (2K 2, C 3, C 5, C 7,...), that is bipartite graphs without induced matching of size 2. 6