arXiv++ Combinatorics

A celebrated theorem of Stiebitz asserts that any graph with minimum degree at least $s+t+1$ can be partitioned into two parts which induce two subgraphs with minimum degree at least $s$ and $t$, respectively. This resolved a conjecture of Thomassen. In this paper, we prove that for $s,t\geq 2$, if a graph $G$ contains no cycle of length four and has minimum degree at least $s+t-1$, then $G$ can be partitioned into two parts which induce two subgraphs with minimum degree at least $s$ and $t$, respectively. This improves the result of Diwan, who proved the same statement for graphs of girth at least five. Our proof also works for the case of variable functions, in which the bounds are sharp as showing by some polarity graphs. As a corollary, it follows that any graph containing no cycle of length four with minimum degree at least $k+1$ contains $k$ vertex-disjoint cycles.

Bollobás and Scott [5] conjectured that every graph $G$ has a balanced bipartite spanning subgraph $H$ such that for each $v\in V(G)$, $d_H(v)\ge (d_G(v)-1)/2$. In this paper, we show that every graphic sequence has a realization for which this Bollobás-Scott conjecture holds, confirming a conjecture of Hartke and Seacrest [10]. On the other hand, we give an infinite family of counterexamples to this Bollobás-Scott conjecture, which indicates that $\lfloor (d_G(v)-1)/2\rfloor$ (rather than $(d_G(v)-1)/2$) is probably the correct lower bound. We also study bipartitions $V_1, V_2$ of graphs with a fixed number of edges. We provide a (best possible) upper bound on $e(V_1)^λ+e(V_2)^λ$ for any real $λ\geq 1$ (the case $λ=2$ is a question of Scott [13]) and answer a question of Scott [13] on $\max\{e(V_1),e(V_2)\}$.

Let $K^{(r)}_{s_1,s_2,\cdots,s_r}$ be the complete $r$-partite $r$-uniform hypergraph and $ex(n,K^{(r)}_{s_1,s_2,\cdots,s_r})$ be the maximum number of edges in any $n$-vertex $K^{(r)}_{s_1,s_2,\cdots,s_r}$-free $r$-uniform hypergraph. It is well-known in the graph case that $ex(n,K_{s,t})=Θ(n^{2-1/s})$ when $t$ is sufficiently larger than $s$. In this note, we generalize the above to hypergraphs by showing that if $s_r$ is sufficiently larger than $s_1,s_2,\cdots,s_{r-1}$ then $$ex(n, K^{(r)}_{s_1,s_2,\cdots,s_r})=Θ\left(n^{r-\frac{1}{s_1s_2\cdots s_{r-1}}}\right).$$ This follows from a more general Turán type result we establish in hypergraphs, which also improves and generalizes some recent results of Alon and Shikhelman. The lower bounds of our results are obtained by the powerful random algebraic method of Bukh. Another new, perhaps unsurprising insight which we provide here is that one can also use the random algebraic method to construct non-degenerate (hyper-)graphs for various Turán type problems. The asymptotics for $ex(n, K^{(r)}_{s_1,s_2,\cdots,s_r})$ is also proved by Verstraëte independently with a different approach.

Let $k$ and $\ell$ be positive integers. A cycle with two blocks $c(k,\ell)$ is an oriented cycle which consists of two internally (vertex) disjoint directed paths of lengths at least $k$ and $\ell$, respectively, from a vertex to another one. A problem of Addario-Berry, Havet and Thomassé (2007) asked if, given positive integers $k$ and $\ell$ such that $k+\ell\ge 4$, any strongly connected digraph $D$ containing no $c(k,\ell)$ has chromatic number at most $k+\ell-1$. In this paper, we show that such digraph $D$ has chromatic number at most $O((k+\ell)^2)$, improving the previous upper bound $O((k+\ell)^4)$ obtained by Cohen, Havet, Lochet and Nisse (2016). In fact, we are able to find a digraph which shows that the answer to the above problem is no. We also show that if in addition $D$ is Hamiltonian, then its underlying simple graph is $(k+\ell-1)$-degenerate and thus the chromatic number of $D$ is at most $k+\ell$, which is tight.

In this paper, we develop a method for studying cycle lengths in hypergraphs. Our method is built on earlier ones used in [21,22,18]. However, instead of utilizing the well-known lemma of Bondy and Simonovits [4] that most existing methods do, we develop a new and very simple lemma in its place. One useful feature of the new lemma is its adaptiveness for the hypergraph setting. Using this new method, we prove a conjecture of Verstraëte [37] that for $r\ge 3$, every $r$-uniform hypergraph with average degree $Ω(k^{r-1})$ contains Berge cycles of $k$ consecutive lengths. This is sharp up to the constant factor. As a key step and a result of independent interest, we prove that every $r$-uniform linear hypergraph with average degree at least $7r(k+1)$ contains Berge cycles of $k$ consecutive lengths. In both of these results, we have additional control on the lengths of the cycles, which therefore also gives us bounds on the Turán numbers of Berge cycles (for even and odd cycles simultaneously). In relation to our main results, we obtain further improvements on the Turán numbers of Berge cycles and the Zarankiewicz numbers of even cycles. We will also discuss some potential further applications of our method.

Let $H$ be a fixed graph. Denote $f(n,H)$ to be the maximum number of edges not contained in any monochromatic copy of $H$ in a 2-edge-coloring of the complete graph $K_n$, and $ex(n,H)$ to be the {\it Turán number} of $H$. An easy lower bound shows $f(n,H)\ge ex(n,H)$ for any $H$ and $n$. In \cite{KS2}, Keevash and Sudakov proved that if $H$ is an edge-color-critical graph or $C_4$, then $f(n,H)= ex(n,H)$ holds for large $n$, and they asked if this equality holds for any graph $H$ when $n$ is sufficiently large. In this paper, we provide an affirmative answer to this problem for an abundant infinite family of bipartite graphs $H$, including all even cycles and complete bipartite graphs $K_{s,t}$ for $t>s^2-3s+3$ or $(s,t)\in\{(3,3),(4,7)\}$. In addition, our proof shows that for all such $H$, the 2-edge-coloring $c$ of $K_n$ achieves the maximum number $f(n,H)$ if and only if one of the color classes in $c$ induces an extremal graph for $ex(n,H)$. We also obtain a multi-coloring generalization for bipartite graphs. Some related problems are discussed in the final section.

In this paper we determine the chromatic number of graphs with two odd cycle lengths. Let $G$ be a graph and $L(G)$ be the set of all odd cycle lengths of $G$. We prove that: (1) If $L(G)=\{3,3+2l\}$, where $l\geq 2$, then $χ(G)=\max\{3,ω(G)\}$; (2) If $L(G)=\{k,k+2l\}$, where $k\geq 5$ and $l\geq 1$, then $χ(G)=3$. These, together with the case $L(G)=\{3,5\}$ solved in \cite{W}, give a complete solution to the general problem addressed in \cite{W,CS,KRS}. Our results also improve a classical theorem of Gyárfás which asserts that $χ(G)\le 2|L(G)|+2$ for any graph $G$.

There has been extensive research on cycle lengths in graphs with large minimum degree. In this paper, we obtain several new and tight results in this area. Let $G$ be a graph with minimum degree at least $k+1$. We prove that if $G$ is bipartite, then there are $k$ cycles in $G$ whose lengths form an arithmetic progression with common difference two. For general graph $G$, we show that $G$ contains $\lfloor k/2\rfloor$ cycles with consecutive even lengths and $k-3$ cycles whose lengths form an arithmetic progression with common difference one or two. In addition, if $G$ is 2-connected and non-bipartite, then $G$ contains $\lfloor k/2\rfloor$ cycles with consecutive odd lengths. Thomassen (1983) made two conjectures on cycle lengths modulo a fixed integer $k$: (1) every graph with minimum degree at least $k+1$ contains cycles of all even lengths modulo $k$; (2) every 2-connected non-bipartite graph with minimum degree at least $k+1$ contains cycles of all lengths modulo $k$. These two conjectures, if true, are best possible. Our results confirm both conjectures when $k$ is even. And when $k$ is odd, we show that minimum degree at least $k+4$ suffices. This improves all previous results in this direction. Moreover, our results derive new upper bounds of the chromatic number in terms of the longest sequence of cycles with consecutive (even or odd) lengths.

The number of proper $q$-colorings of a graph $G$, denoted by $P_G(q)$, is an important graph parameter that plays fundamental role in graph theory, computational complexity theory and other related fields. We study an old problem of Linial and Wilf to find the graphs with $n$ vertices and $m$ edges which maximize this parameter. This problem has attracted much research interest in recent years, however little is known for general $m,n,q$. Using analytic and combinatorial methods, we characterize the asymptotic structure of extremal graphs for fixed edge density and $q$. Moreover, we disprove a conjecture of Lazebnik, which states that the Turán graph $T_s(n)$ has more $q$-colorings than any other graph with the same number of vertices and edges. Indeed, we show that there are infinite many counterexamples in the range $q = O({s^2}/{\log s})$. On the other hand, when $q$ is larger than some constant times ${s^2}/{\log s}$, we confirm that the Turán graph $T_s(n)$ asymptotically is the extremal graph achieving the maximum number of $q$-colorings. Furthermore, other (new and old) results on various instances of the Linial-Wilf problem are also established.

It is proved that there exists an absolute constant c > 0 such that for every natural number k, every non-bipartite 2-connected graph with average degree at least ck contains k cycles with consecutive odd lengths. This implies the existence of the absolute constant d > 0 that every non-bipartite 2-connected graph with minimum degree at least dk contains cycles of all lengths modulo k, thus providing an answer (in a strong form) to a question of Thomassen. Both results are sharp up to the constant factors.

Given a set of $t$ words of length $n$ over a $k$-letter alphabet, it is proved that there exists a common subsequence among two of them of length at least $\frac{n}{k}+cn^{1-1/(t-k-2)}$, for some $c>0$ depending on $k$ and $t$. This is sharp up to the value of $c$.

A graph is called diameter-$k$-critical if its diameter is $k$, and the removal of any edge strictly increases the diameter. In this paper, we prove several results related to a conjecture often attributed to Murty and Simon, regarding the maximum number of edges that any diameter-$k$-critical graph can have. In particular, we disprove a longstanding conjecture of Caccetta and Häggkvist (that in every diameter-2-critical graph, the average edge-degree is at most the number of vertices), which promised to completely solve the extremal problem for diameter-2-critical graphs. On the other hand, we prove that the same claim holds for all higher diameters, and is asymptotically tight, resolving the average edge-degree question in all cases except diameter-2. We also apply our techniques to prove several bounds for the original extremal question, including the correct asymptotic bound for diameter-$k$-critical graphs, and an upper bound of $(\frac{1}{6} + o(1))n^2$ for the number of edges in a diameter-3-critical graph.

Let $k$ and $r$ be two integers with $k \ge 2$ and $k\ge r \ge 1$. In this paper we show that (1) if a strongly connected digraph $D$ contains no directed cycle of length $1$ modulo $k$, then $D$ is $k$-colorable; and (2) if a digraph $D$ contains no directed cycle of length $r$ modulo $k$, then $D$ can be vertex-colored with $k$ colors so that each color class induces an acyclic subdigraph in $D$. The first result gives an affirmative answer to a question posed by Tuza in 1992, and the second implies the following strong form of a conjecture of Diwan, Kenkre and Vishwanathan: If an undirected graph $G$ contains no cycle of length $r$ modulo $k$, then $G$ is $k$-colorable if $r\ne 2$ and $(k+1)$-colorable otherwise. Our results also strengthen several classical theorems on graph coloring proved by Bondy, Erdős and Hajnal, Gallai and Roy, Gyárfás, etc.

Answering in a strong form a question posed by Bollobás and Scott, in this paper we determine the discrepancy between two random k-uniform hypergraphs, up to a constant factor depending solely on k.

Fifty years ago Erdős asked to determine the minimum number of $k$-cliques in a graph on $n$ vertices with independence number less than l. He conjectured that this minimum is achieved by the disjoint union of $l-1$ complete graphs of size $\frac{n}{l-1}$. This conjecture was disproved by Nikiforov who showed that the balanced blow-up of a 5-cycle has fewer 4-cliques than the union of 2 complete graphs of size $\frac{n}{2}$. In this paper we solve Erdős' problem for $(k,l)=(3,4)$ and $(k,l)=(4,3)$. Using stability arguments we also characterize the precise structure of extremal examples, confirming Erdős' conjecture for $(k,l)=(3,4)$ and showing that a blow-up of a 5-cycle gives the minimum for $(k,l)=(4,3)$.

A minimum feedback arc set of a directed graph $G$ is a smallest set of arcs whose removal makes $G$ acyclic. Its cardinality is denoted by $β(G)$. We show that an Eulerian digraph with $n$ vertices and $m$ arcs has $β(G) \ge m^2/2n^2+m/2n$, and this bound is optimal for infinitely many $m, n$. Using this result we prove that an Eulerian digraph contains a cycle of length at most $6n^2/m$, and has an Eulerian subgraph with minimum degree at least $m^2/24n^3$. Both estimates are tight up to a constant factor. Finally, motivated by a conjecture of Bollobás and Scott, we also show how to find long cycles in Eulerian digraphs.