Degree sequence condition for Hamiltonicity in tough graphs
Abstract
Generalizing both Dirac’s condition and Ore’s condition for Hamilton cycles, Chvátal in 1972 established a degree sequence condition for the existence of a Hamilton cycle in a graph. Hoàng in 1995 generalized Chvátal’s degree sequence condition for 1-tough graphs and conjectured a -tough analogue for any positive integer . Hoàng in the same paper verified his conjecture for and recently Hoàng and Robin verified the conjecture for . In this paper, we confirm the conjecture for all .
Keywords. Degree sequence; Hamiltonian cycle; Toughness
1 Introduction
Graphs considered in this paper are simple, undirected, and finite. Let be a graph. Denote by and the vertex set and edge set of , respectively. The degree of a vertex in is denoted by . If and are non-adjacent in , then is obtained from by adding the edge . We write if two vertices and are adjacent in ; and write otherwise. For , denote by and the subgraph of induced on and , respectively. For , we write for . For two integer , we let .
Let be an integer. The non-decreasing sequence is a degree sequence of graph if the vertices of can be labeled as such that for all . In 1972, Chvátal [3] proved the following well known result.
Theorem 1.
Let be a graph with degree sequence , where is an integer. If for all , implies , then is Hamiltonian.
Hoàng [5, Conjecture 1] in 1995 conjectured a toughness analogue for the theorem above. We let denote the number of components of . For a real number , we say is t-tough if for all such that . The largest for which is -tough is called the toughness of and is denoted . If is complete, is defined to be . Chvátal [4] defined this concept in 1973 as a measure of a graph’s “resilience” under the removal of vertices. Hoàng’s conjecture can now be stated as follows.
Conjecture 2.
Let and be integers, and be a graph with degree sequence . Suppose that satisfies the following predicate :
Then, if is -tough, is Hamiltonian.
Hoàng in the same paper [5, Theorem 3] proved that the Conjecture holds for . Since every hamiltonian graph must necessarily be 1-tough, the statement for generalizes Theorem 1. Recently, Hoàng and Robin [6] proved that the Conjecture is true for . In this paper, we confirm Conjecture 2 for all .
Theorem 3.
Let be an integer and be a -tough graph on vertices with degree sequence . If for all it holds that implies , then is Hamiltonian.
A graph is pancyclic if contains cycles of any length from to . As a consequence of Theorem 3 and a result of Hoàng [5, Theorem 7] that if a -tough graph satisfies and is Hamiltonian, then is pancyclic or bipartite, we also obtain the following result.
Corollary 4.
Let be an integer and be a -tough graph on vertices with degree sequence . If for all it holds that implies , then is pancyclic or bipartite.
2 Proof of Theorem 3
We will need the following result by Bauer et al. [1] and our closure lemma for -tough graphs with .
Theorem 5.
Let be any real number and be a -tough graph on vertices . If , then is Hamiltonian.
Theorem 6 (Toughness Closure Lemma).
Let be an integer, be a -tough graph on vertices, and let distinct be non-adjacent with . Then is Hamiltonian if and only if is Hamiltonian.
The following toughness closure concept was given by Hoàng and Robin [6]. Let be an integer, and be a -tough graph on vertices. Then the -closure of is formed by repeatedly adding edges joining vertices and such that and are non-adjacent in the current graph and their degree sum is at least in the current graph, until no such pair remains. By the same argument as showing that the Hamiltonian closure of a graph is well defined (e.g., see [2, Lemma 4.4.2]), the -closure of is well defined. Thus by Theorem 6, we will consider the -closure of instead when we prove Theorem 3. We mostly adopted the ideas used by Hoàng and Robin in [6].
Proof of Theorem 3.
As satisfies the property implies that any supergraph of obtained from by adding missing edges also satisfies the property , by Theorem 6, it suffices to work with the -closure of . For the sake of notation, we just assume that itself is its -closure. We may assume that is not Hamiltonian. Thus is not complete and so by being 4-tough.
Let be all the vertices of such that for all . Thus, we have that implies . By Theorem 1, if for all , then is Hamiltonian. So, we assume that there exists some positive integer such that . Then as , we have . Choose to be minimum with the property that . Then for all . Since , we must have .
Let . We say that is complete to if for all and such that , we have . If for all and such that , we call a universal clique of . Clearly, vertices in a universal clique have degree in . We will show that has a universal clique of size larger than . In particular, this gives that . By Theorem 5, this proves that is Hamiltonian, a contradiction to the assumption that is not Hamiltonian. Let
Claim 2.1.
For all positive integer , is a clique complete to .
Proof of Claim 2.1.
If for some and for some , then . Thus, . This in turn implies that is a clique in , since . ∎
Claim 2.2.
Let be any positive integer. If for every , it holds that implies , then is a universal clique in .
Proof of Claim 2.2.
Assume there exists a positive integer that satisfies the hypothesis, but is not a universal clique. Choose to be maximum such that there exists for some such that . By Claim 2.1, . Thus by the assumption of this claim. By the maximality of , we have for all . So, , which gives . But, this implies , a contradiction. ∎
Let be a universal clique in of maximum size.
Claim 2.3.
We have .
Proof of Claim 2.3.
Suppose that . As is a universal clique in , we have for all . If , then , which contradicts . Thus . Note that for any as every vertex of has degree . Let . As for all , each has at most neighbor from in , and so we have
if , | ||||
if . |
Since , we have . However, we get , contradicting the toughness of . Thus, Claim 2.3 must hold. ∎
Claim 2.4.
For all positive integer such that , we have .
Proof of Claim 2.4.
Suppose such that . By the hypothesis, . That is, there are at least vertices of degree at least , indicating . ∎
Claim 2.5.
We have for all integer with .
Proof of Claim 2.5.
Assume there exists such that and . Choose such an to be minimum. It suffices to show that is a universal clique: by Claims 2.3 and 2.4, we have . Rearranging gives , a contradiction. Thus we show that is a universal clique in the following. By Claim 2.1, is a clique complete to . Therefore, to apply Claim 2.2, we show that every vertex for belongs to the set or satisfies but .
We first show that for all . Consider for now that . If , then . By the minimality of , we get . Thus . If , then . In either case, we have shown . For any , we have . Now for , suppose . By the minimality of , we have if . We have if . By the minimality of , we have for all . This completes the proof. ∎
Claim 2.6.
We have .
Proof of Claim 2.6.
We suppose to the contrary that . Let . Then . By Claim 2.5, we have . If , then all vertices from are contained in a universal clique of and so we have . This gives , a contradiction of the assumption that . Thus there exists such that . We choose such an to be maximum. Since , we have , which gives . Then by Claim 2.5 and the argument in the second paragraph in the proof of Claim 2.5, we have . By the maximality of , we have for all and so . This gives , which contradicts that . ∎
Claim 2.7.
We have .
Proof of Claim 2.7.
Assume . Then, as , we have . By Claim 2.2 and the choice of , we know that is a universal clique. Therefore, by Claims 2.4 and 2.6, we get . Observe that for , we have
This gives . Thus , a contradiction. ∎
As , Theorem 5 implies that is Hamiltonian, a contradiction to our assumption that is not Hamiltonian. This completes the proof.
3 Proof of Theorem 6
Denote by an orientation of a cycle . We assume that the orientation is clockwise throughout the rest of this paper. For , denotes the path from to along . Similarly, denotes the path between and which travels opposite to the orientation. We use to denote the immediate successor of on and to denote the immediate predecessor of on . If , then and . We use similar notation for a path when it is given an orientation. Theorem 7 is needed in the proof of Theorem 6, and we prove Theorem 7 in the last section.
Theorem 7.
Let be rational and be a -tough graph on vertices. Suppose that is not Hamiltonian, but there exists such that has a Hamilton cycle . Then, for any distinct , we have that .
Theorem 6 (Toughness Closure Lemma).
Let be an integer, be a -tough graph on vertices, and let distinct be non-adjacent with . Then is Hamiltonian if and only if is Hamiltonian.
Proof.
It is clear that is Hamiltonian implies that is Hamiltonian. For the converse, we suppose that is Hamiltonian but is not. Then again, this implies that is not complete and so .
As is Hamiltonian, has a Hamilton path connecting and . Let be such a path, where and . We will orient to be from to , and write for two vertices and such that is at least as close to along as is. Our goal is to find a cutset of with size less than and so arriving a contradiction to the toughness of . For this purpose, based on the assumption that is not Hamiltonian, we look at how the neighbors of and are arranged along this path , and their adjacency relations.
The first two assertions below follow directly from the assumption that is not Hamiltonian, and the last two are corollaries of the first two.
Claim 3.1.
Let distinct and suppose and . Then the following holds.
-
(1)
If , then and .
-
(2)
If , then and .
-
(3)
If and additionally , then for any with .
-
(4)
If and additionally , then for any with .
Since and and do not have two common neighbors that are consecutive on by Claim 3.1(1) above, each of and is expected to have many neighbors that are consecutive on . Thus we define neighbor intervals for and , respectively, as set of consecutive vertices on that are all adjacent to or . For , and with and such that , we call a -interval and denote it by if but .
Given and , by Claim 3.1(1), we know that the two intervals can have at most one vertex in common. In case that they do have a common vertex, then it must be the case that . In this case, we let and call it a joint-interval. Finally, for with , we define interval-gaps to be sets of consecutive vertices on that are all not adjacent to either or . A parallel-gap is such that and that , or , or but . A crossing-gap is such that and that and . By the range of and in the above definition, we see that each of and is not contained in any interval-gaps.
Let be the set of -intervals that are not joint-intervals, be the set of -intervals that are not joint-intervals, and be the set of joint-intervals. Let
Claim 3.2.
Each crossing-gap contains at least two vertices and there are at least distinct crossing-gaps when .
Proof of Claim 3.2.
For the first part, suppose for some is a crossing-gap with a single vertex. Then gives a Hamilton cycle of . We have , and with respect to the cycle , we have and . However, , contradicting Theorem 7. For the second part, assume that . Let the common neighbors of and be with . Thus for each is a set of vertices such that and . By the first part of this claim and Claim 3.1(1), we know that each of for contains at least two vertices that are adjacent to neither nor . By finding a minimal sub-path of such that the predecessor of its left end is a neighbor of , the successor of its right end is a neighbor of , we can find two distinct vertices with the following properties: , , , and . Then is a crossing-gap. Since and are disjoint for distinct , we can find distinct crossing-gaps. ∎
Let be the total number of distinct parallel-gaps and be the total number of distinct crossing-gaps. We let the set of parallel-gaps be , and let . We also let the set of crossing-gaps be , and let .
Claim 3.3.
We have .
Proof of Claim 3.3.
By the definition, the three sets are pairwise disjoint. Thus . Also, by our definition, we have and so . Since , and and are contained in an -interval, -interval, or joint-interval, it follows that there are exactly interval-gaps. By Claim 3.2, . As and are not contained in any interval-gaps, we get
As and , we get . Therefore,
as desired. ∎
Claim 3.4.
For any , if is a crossing-gap of size 2, then for any such that .
Proof of Claim 3.4.
We will show that has less than neighbors in , to arrive a contradiction to being -tough.
By Claim 3.1(1)-(2), we know that for any with on , we have ; and for with on , we have . Thus vertices from and are non-neighbors of . Let
if , | ||||
if . |
Then is a Hamilton cycle of . The predecessors and successors of vertices below are all taken with respect to . As is not Hamiltonian, both and are independent sets in . When , since and , it then follows that for any . As a consequence, we get . When , since and , it then follows that for any . As a consequence, we get .
The arguments above indicate that for every distinct vertex , there is a unique non-neighbor of that is corresponding to . Thus has at least non-neighbors on . Then by Claim 3.3 that , we get
a contradiction. ∎
We now construct a cutset of such that . To do so, we define the following sets:
Let
if is a -interval, | ||||
otherwise. |
We prove the following claims regarding what vertices are in and the size of .
Claim 3.5.
Let for some . Then , or , or is contained in a parallel-gap of size one such that , or is contained in a crossing-gap of size two, or is the right endvertex of a crossing-gap of size three.
Proof of Claim 3.5.
By the definition of , we know that either is a neighbor of or , or is contained in a parallel-gap of size one, or a crossing-gap of size two or three. If , then by the definition of , we have . If , then by the definition of , we have . If is contained in a parallel-gap of size one, then by the definition of , we know that . As is a parallel-gap, implies . If is contained in crossing-gap of size three, then is the right endvertex of a crossing-gap of size three by the definition of . ∎
Claim 3.6.
We have .
Proof of Claim 3.6.
For each crossing-gap of size , we let if , if , and if . Note that by the definition of , only one vertex was deleted from the -interval containing . Now by the definition of and Claim 3.3, we have
where the last inequality follows as when , and by the definition of and the fact that for all from Claim 3.2. ∎
Claim 3.7.
We have .
Proof of Claim 3.7.
For the sake of contradiction, suppose is connected. Let and . Then, there must exists a path in connecting a vertex of and a vertex and is internally-disjoint with . Suppose that for some and . By Claim 3.5, we know that , or , or when the -interval containing has size at least two, and that . By Claim 3.1(1) and (4), we know that . Thus contains at least three vertices. As is internally-disjoint with , are from interval-gaps of .
As again, , or , or when the -interval containing has size at least two. Since , Claim 3.1(4) implies that . Thus is not the right endvertex of any crossing-gap. By Claim 3.4, is not the left endvertex of any crossing-gap of size two. Thus by Claim 3.5, is a parallel-gap of size one such that . Now with in the place of , the same arguments as above imply that , if exists, is a parallel-gap of size one such that . Similarly, for any , if exists, we deduce that is a parallel-gap of size one such that . As and , we get a contradiction to Claim 3.1(4). ∎
4 Proof of Theorem 7
Theorem 7.
Let be rational and be a -tough graph on vertices. Suppose that is not Hamiltonian, but there exists such that has a Hamilton cycle . Then, for any distinct , we have that .
Proof.
Suppose to the contrary that there are distinct for which . As is not hamiltonian, is not a complete graph. Thus .
For and , let and . For , we let and . We will construct a cutset of such that . For this purpose, we define the following sets:
In the following, we prove some properties of these sets.
Claim 4.1.
We have .
Proof of Claim 4.1.
Suppose to the contrary that there exists . If , then is a Hamilton cycle of . If , then is a Hamilton cycle of . ∎
If there are with , then is a Hamilton cycle in . Thus we have the following claim.
Claim 4.2.
The set is an independent set in .
Claim 4.3.
We have and .
Proof of Claim 4.3.
Clearly . Observe that and . By Claim 4.1, we have ; and by Claim 4.2, we have . Thus,
(1) | |||||
which gives . For the second part, it follows from (1) by noting that . ∎
We will take a subset of with size at least and show that deleting less than vertices from produces at least components, and thus contradicts the assumption that is 4-tough. We let
Claim 4.4.
We have .
Proof of Claim 4.4.
As , it suffices to show . Let . Then we have by the definition of . Also, we have because and is an independent set by Claim 4.2. Furthermore, , as otherwise that contradicts being independent in . Thus . Consider next that . Then we have by the definition of . Also, we have and by the same argument as above. Thus . Therefore we have
as desired. ∎
Claim 4.5.
The set is an independent set in .
Proof of Claim 4.5.
Since is an independent set by Claim 4.2, for any , since and , it follows that ; and for any , since and , it follows that . Thus it not adjacent to any vertex from . Next, let distinct such that . Consider first that . By symmetry, we assume that is in between and along . Then is a Hamilton cycle of . Next consider . By symmetry, we assume that is in between and along . Then is a Hamilton cycle of . Finally, consider and . Then is a Hamilton cycle in . Therefore, is an independent set in . ∎
We show that all except at most vertices of correspond to a vertex from . For this purpose, we introduce three new sets as follows.
We can think of the definition of above as a mapping from to vertices in . For , we say that a vertex generates if when , and if when .
A chord of is an edge with and . Two chords and of that do not share any endvertices form a crossing if the four vertices appear along in the order or . We say that form a crossing with if there exist distinct vertices and such that such that and are crossing chords of .
Claim 4.6.
For and , there exist no such that , , and and form a crossing.
Proof of Claim 4.6.
We proceed by contradiction. Assume that , and are as described in the claim. The definitions of and are symmetric up to reversing the direction of and exchanging the roles of and . Thus we assume that and consider two cases regarding or below. In each case, we construct a Hamilton cycle of , thereby achieving a contradiction to the assumption that is not Hamiltonian.
Consider first that . We let a Hamilton cycle of be defined as follows according to the location of the vertex on :
if (in this case ). See Figure 1(a). | ||||
if (in this case ). | ||||
if (in this case ). |
Consider then that . We let a Hamilton cycle of be defined as follows according to the location of the vertex on :
if (in this case ). See Figure 1(b). | ||||
if (in this case ). |


Lastly, let . In this case, we have . Let be the vertex that generates . Then is constructed according to the location of on :
if . See Figure 2. | ||||
if . | ||||
if . | ||||
if . |

∎
We will show next that . In this way, after deleting of at most vertices, we will get at least components and so get a contradiction to the toughness of .
Claim 4.7.
We have .
Proof of Claim 4.7.
By the definition of , we know that . Now to prove Claim 4.7, it suffices to show that because if this is true then we get
We show that , which would get us the desired upper bound by the first part of Claim 4.3.
The proof requires several cases. In most cases, we show that for each distinct element of in the given case, there is a distinct element of . Let and such that generates . Recall that and . Since the definitions of and are symmetric up to reversing the direction of and exchanging the roles of and , we prove the case only.
Consider first that . We may assume as otherwise . Now we must have since otherwise and form a crossing if and and form a crossing if , contradicting Claim 4.6. Therefore and so . Thus in the following cases, we assume . Recall that we assumed .
Suppose first that . Then implies . Since , we must have . This implies that . We next claim that , as otherwise and form a crossing. Thus .
Suppose then that . Then implies . As , we get . Also, . Otherwise, is a Hamilton cycle in . Thus . In particular, in this case, . For otherwise, suppose , then is a Hamilton cycle in . Thus .
Lastly, consider . Then implies . As , we must have , which gives . By Claim 4.6, . Lastly, , as otherwise is a Hamilton cycle in . Thus . Since , it follows that .
The three sets , , and are disjoint, we have when , and we have when . Thus the argument above implies that distinct vertices from correspond to distinct vertices from . Therefore , as desired. ∎
References
- [1] D. Bauer, H. J. Broersma, J. van den Heuvel, and H. J. Veldman. Long cycles in graphs with prescribed toughness and minimum degree. Discrete mathematics, 141(1-3):1–10, 1995.
- [2] J. A. Bondy and U. S. R. Murty. Graph theory with applications. American Elsevier Publishing Co., Inc., New York, 1976.
- [3] V. Chvátal. On hamilton’s ideals. Journal of Combinatorial Theory, Series B, 12(2):163–168, 1972.
- [4] V. Chvátal. Tough graphs and hamiltonian circuits. Discrete Mathematics, 5(3):215–228, 1973.
- [5] C. T. Hoàng. Hamiltonian degree conditions for tough graphs. Discrete mathematics, 142(1-3):121–139, 1995.
- [6] C. T. Hoàng and C. Robin. A closure lemma for tough graphs and hamiltonian degree conditions. Discrete Mathematics, 347(4):113872, 2024.