arXiv subjects
Publications and source records attributed to Maria Chudnovsky.
Fix g>1. Every graph of large enough tree-width contains a g x g grid as a minor; but here we prove that every four-edge-connected graph of large enough tree-width contains a g x g grid as an immersion (and hence contains any fixed graph with maximum degree at most four as an immersion). This result has a number of applications.
Let $B$ and $R$ be two simple graphs with vertex set $V$, and let $G(B,R)$ be the simple graph with vertex set $V$, in which two vertices are adjacent if they are adjacent in at least one of $B$ and $R$. For $X \subseteq V$, we denote by $B|X$ the subgraph of $B$ induced by $X$; let $R|X$ and $G(B,R)|X$ be defined similarly. We say that the pair $(B,R)$ is {\em additive} if for every $X \subseteq V$, the sum of the clique numbers of $B|X$ and $R|X$ is at least the clique number of $G(B,R)|X$. In this paper we give a necessary and sufficient characterization of additive pairs of graphs. This is a numerical variant of a structural question studied in \cite{ABC}.
Let $\mathcal{G}$ be the class of all graphs with no induced four-edge path or four-edge antipath. Hayward and Nastos \cite{MS} conjectured that every prime graph in $\mathcal{G}$ not isomorphic to the cycle of length five is either a split graph or contains a certain useful arrangement of simplicial and antisimplicial vertices. In this paper we give a counterexample to their conjecture, and prove a slightly weaker version. Additionally, applying a result of the first author and Seymour \cite{grow} we give a short proof of Fouquet's result \cite{C5} on the structure of the subclass of bull-free graphs contained in $\mathcal{G}$.
A class $\mathcal{G}$ of graphs is said to be {\em $χ$-bounded} if there is a function $f:\mathbb{N} \rightarrow \mathbb{R}$ such that for all $G \in \mathcal{G}$ and all induced subgraphs $H$ of $G$, $χ(H) \leq f(ω(H))$. In this paper, we show that if $\mathcal{G}$ is a $χ$-bounded class, then so is the closure of $\mathcal{G}$ under any one of the following three operations: substitution, gluing along a clique, and gluing along a bounded number of vertices. Furthermore, if $\mathcal{G}$ is $χ$-bounded by a polynomial (respectively: exponential) function, then the closure of $\mathcal{G}$ under substitution is also $χ$-bounded by some polynomial (respectively: exponential) function. In addition, we show that if $\mathcal{G}$ is a $χ$-bounded class, then the closure of $\mathcal{G}$ under the operations of gluing along a clique and gluing along a bounded number of vertices together is also $χ$-bounded, as is the closure of $\mathcal{G}$ under the operations of substitution and gluing along a clique together.
For a graph $G$ and a set of graphs $\mathcal{H}$, we say that $G$ is {\em $\mathcal{H}$-free} if no induced subgraph of $G$ is isomorphic to a member of $\mathcal{H}$. Given an integer $P>0$, a graph $G$, and a set of graphs $\mathcal{F}$, we say that $G$ {\em admits an $(\mathcal{F},P)$-partition} if the vertex set of $G$ can be partitioned into $P$ subsets $X_1,..., X_P$, so that for every $i \in \{1,..., P\}$, either $|X_i|=1$, or the subgraph of $G$ induced by $X_i$ is $\{F\}$-free for some $F \in \mathcal{F}$. Our first result is the following. For every pair $(H,J)$ of graphs such that $H$ is the disjoint union of two graphs $H_1$ and $H_2$, and the complement $J^c$ of $J$ is the disjoint union of two graphs $J_1^c$ and $J_2^c$, there exists an integer $P>0$ such that every $\{H,J\}$-free graph has an $(\{H_1,H_2,J_1,J_2\},P)$-partition. Using a similar idea we also give a short proof of one of the results of \cite{heroes}. Our final result is a construction showing that if $\{H,J\}$ are graphs each with at least one edge, then for every pair of integers $r,k$ there exists a graph $G$ such that every $r$-vertex induced subgraph of $G$ is $\{H,J\}$-split, but $G$ does not admits an $(\{H,J\},k)$-partition.
A conjecture due to the fourth author states that every $d$-regular planar multigraph can be $d$-edge-coloured, provided that for every odd set $X$ of vertices, there are at least $d$ edges between $X$ and its complement. For $d = 3$ this is the four-colour theorem, and the conjecture has been proved for all $d\le 8$, by various authors. In particular, two of us proved it when $d=7$; and then three of us proved it when $d=8$. The methods used for the latter give a proof in the $d=7$ case that is simpler than the original, and we present it here.
It was conjectured by the third author in about 1973 that every $d$-regular planar graph (possibly with parallel edges) can be $d$-edge-coloured, provided that for every odd set $X$ of vertices, there are at least $d$ edges between $X$ and its complement. For $d = 3$ this is the four-colour theorem, and the conjecture has been proved for all $d\le 7$, by various authors. Here we prove it for $d = 8$.
Chudnovsky and Seymour's structure theorem for claw-free graphs has led to a multitude of recent results that exploit two structural operations: {\em compositions of strips} and {\em thickenings}. In this paper we consider the latter, proving that every claw-free graph has a unique optimal {\em antithickening}, where our definition of {\em optimal} is chosen carefully to respect the structural foundation of the graph. Furthermore, we give an algorithm to find the optimal antithickening in $O(m^2)$ time. For the sake of both completeness and ease of proof, we prove stronger results in the more general setting of trigraphs.
A special case of a combinatorial theorem of De Bruijn and Erdos asserts that every noncollinear set of n points in the plane determines at least n distinct lines. Chen and Chvatal suggested a possible generalization of this assertion in metric spaces with appropriately defined lines. We prove this generalization in all metric spaces induced by connected chordal graphs.
One of the De Bruijn - Erdos theorems deals with finite hypergraphs where every two vertices belong to precisely one hyperedge. It asserts that, except in the perverse case where a single hyperedge equals the whole vertex set, the number of hyperedges is at least the number of vertices and the two numbers are equal if and only if the hypergraph belongs to one of simply described families, near-pencils and finite projective planes. Chen and Chvatal proposed to define the line uv in a 3-uniform hypergraph as the set of vertices that consists of u, v, and all w such that {u,v,w} is a hyperedge. With this definition, the De Bruijn - Erdos theorem is easily seen to be equivalent to the following statement: If no four vertices in a 3-uniform hypergraph carry two or three hyperedges, then, except in the perverse case where one of the lines equals the whole vertex set, the number of lines is at least the number of vertices and the two numbers are equal if and only if the hypergraph belongs to one of two simply described families. Our main result eneralizes this statement by allowing any four vertices to carry three hyperedges (but keeping two forbidden): the conclusion remains the same except that a third simply described family, complements of Steiner triple systems, appears in the extremal case.
Reed's $ω$, $Δ$, $χ$ conjecture proposes that every graph satisfies $χ\leq \lceil\frac 12(Δ+1+ω)\rceil$; it is known to hold for all claw-free graphs. In this paper we consider a local strengthening of this conjecture. We prove the local strengthening for line graphs, then note that previous results immediately tell us that the local strengthening holds for all quasi-line graphs. Our proofs lead to polytime algorithms for constructing colourings that achieve our bounds: $O(n^2)$ for line graphs and $O(n^3m^2)$ for quasi-line graphs. For line graphs, this is faster than the best known algorithm for constructing a colouring that achieves the bound of Reed's original conjecture.
Let G be a digraph (without parallel edges) such that every directed cycle has length at least four; let $β(G)$ denote the size of the smallest subset X in E(G) such that $G\X$ has no directed cycles, and let $γ(G)$ be the number of unordered pairs {u,v} of vertices such that u,v are nonadjacent in G. It is easy to see that if $γ(G) = 0$ then $β(G) = 0$; what can we say about $β(G)$ if $γ(G)$ is bounded? We prove that in general $β(G)$ is at most $γ(G)$. We conjecture that in fact $β(G)$ is at most $γ(G)/2$ (this would be best possible if true), and prove this conjecture in two special cases: 1. when V(G) is the union of two cliques, 2. when the vertices of G can be arranged in a circle such that if distinct u,v,w are in clockwise order and uw is a (directed) edge, then so are both uv and vw.
A graph G is perfect if for every induced subgraph H, the chromatic number of H equals the size of the largest complete subgraph of H, and G is Berge if no induced subgraph of G is an odd cycle of length at least 5 or the complement of one. The "strong perfect graph conjecture" (Berge, 1961) asserts that a graph is perfect if and only if it is Berge. A stronger conjecture was made recently by Conforti, Cornuejols and Vuskovic -- that every Berge graph either falls into one of a few basic classes, or it has a kind of separation that cannot occur in a minimal imperfect graph. In this paper we prove both these conjectures.