Anti-magic labeling of graphs
Yu-Chang Liang
April 28, 2017
Liang ()
report
April 28, 2017
1 / 66
Outline
1
Introduction
2
Tool
3
Known results
4
Our research
5
Further work
6
References
Liang ()
report
April 28, 2017
2 / 66
History Definition Magic square is an arrangement of numbers in a square grid, where the numbers in each row, and in each column, and the numbers in the forward and backward main diagonals, all add up to the same number.
Liang ()
report
April 28, 2017
3 / 66
Labeling
A graph labeling is a mapping f satisfied: f : A −→ B. S where A may be V (G), E(G), F (G), V (G) E(G), etc, and B be the set of numbers( the element of a group). Goal: The structure of φf (x) which is the function depend with f . P e.g., φf (v) = e∈E(v) f (e), or φf (uv) = f (u) − f (v)
Liang ()
report
April 28, 2017
4 / 66
History Definition (Sedl´aˇcek 1963) A graph G is called magic if G has a labeling of the edges with distinct integers such that for each vertex v the sum of the labels of all edges incident with v is the same for all v.
Liang ()
report
April 28, 2017
5 / 66
History
Definition (Stewart 1966, 1967) A graph G is called super-magic if G has a labeling of the edges with distinct consecutive integers such that for each vertex v the sum of the labels of all edges incident with v is the same for all v.
Definition (Stewart 1966, 1967) A graph G is called semi-magic if G has a labeling of the edges with integers such that for each vertex v the sum of the labels of all edges incident with v is the same for all v.
Liang ()
report
April 28, 2017
6 / 66
History
Definition An S-weighting of a graph G is a map w : E(G) −→ S. k-weighting is an S-weighting with S = {1, . . . , k}.
Definition
P Let φw (v) = e∈E(v) w(e). A weighting w is proper if φw (v) is a proper coloring of G.
Liang ()
report
April 28, 2017
7 / 66
History
Conjecture ((1,2,3)-Conjecture, 2004) Every graph without isolated edges has a proper 3-weighting.
Liang ()
report
April 28, 2017
8 / 66
History
Definition An total S-weighting of a graph G is a map w : E(G) k-weighting is an S-weighting with S = {1, . . . , k}.
S
V (G) −→ S.
Definition
P Let φw (v) = w(v) + e∈E(v) w(e). A weighting w is proper if φw (v) is a proper coloring of G.
Liang ()
report
April 28, 2017
9 / 66
History
Conjecture (1-2-Conjecture, 2007) Every graph has a proper total 2-weighting.
Liang ()
report
April 28, 2017
10 / 66
History Definition (Rosa, A 1967) A graph G is called graceful if f is an injection from the vertices of G to the set {0, 1, . . . E(G)} such that, when each edge xy is assigned the label | f (x) − f (y) |, the resulting edge labels are distinct.
Liang ()
report
April 28, 2017
11 / 66
History
Conjecture (1967) All trees are graceful.
Liang ()
report
April 28, 2017
12 / 66
Definition Definition (Hartsfild and Ringel 1990) An anti-magic labeling of G is a bijection from the set of edges to the set {1, 2, . . . , |E(G)|}, such that all the vertex-sums are pairwise distinct, where the vertex-sum s(v) of a vertex v is the sum of labels of all edges incident with v.
Liang ()
report
April 28, 2017
13 / 66
Tool
1 : Algorithm. (Strategy) 2 : Combinatorial Nullstellensatz. 3 : Probability method.
Liang ()
report
April 28, 2017
14 / 66
Strategy
1 : Consider the structure of G. 2 : Design the function φf (v). 3 : characterize the graph G and f .
Liang ()
report
April 28, 2017
15 / 66
Strategy of paths
Liang ()
report
April 28, 2017
16 / 66
Strategy of paths
Liang ()
report
April 28, 2017
17 / 66
Strategy of cycles
Liang ()
report
April 28, 2017
18 / 66
Strategy of cycles
Liang ()
report
April 28, 2017
19 / 66
Combinatorial Nullstellensatz
Let f be the labeling of G, and f (ei ) = xi , then g(x1 , x2 , . . . , xm ) =
Y
(xi − xj )
1≤i
Y
(φf (vi ) − φf (vj )).
1≤i
If there are x1 , x2 , . . . , xm , such that g(x1 , x2 , . . . , xm ) 6= 0, then G is anti-magic.
Liang ()
report
April 28, 2017
20 / 66
Combinatorial Nullstellensatz
Let η be the index function. g(x1 , x2 , . . . , xm ) =
X η
cη
Y
η(i)
xi
.
i
If cη 6= 0, and |Si | ≥ η(i) + 1, then there are xi ∈ Si , such that g(x1 , x2 , . . . , xm ) 6= 0
Liang ()
report
April 28, 2017
21 / 66
Combinatorial Nullstellensatz
Liang ()
report
April 28, 2017
22 / 66
Combinatorial Nullstellensatz
g(x1 , x2 , . . . , x5 ) = (x1 + x2 − x2 − x3 )(x2 + x3 − x3 − x4 ) (x3 + x4 − x4 − x5 )(x4 + x5 − x5 − x1 )(x5 + x1 − x1 − x2 ) (x1 + x2 − x3 − x4 )(x2 + x3 − x4 − x5 )(x3 + x4 − x5 − x1 ) Y (x4 + x5 − x1 − x2 )(x5 + x1 − x2 − x3 ) (xi − xj ). 1≤i
The coefficient of x41 x42 x43 x44 x45 is −980
Liang ()
report
April 28, 2017
23 / 66
Conjecture
Conjecture (Hartsfild and Ringel 1990) All trees except K2 have an anti-magic labeling.
Liang ()
report
April 28, 2017
24 / 66
Conjecture
Liang ()
report
April 28, 2017
25 / 66
Conjecture
Conjecture (Hartsfild and Ringel 1990) All connected graphs except K2 have an anti-magic labeling.
Liang ()
report
April 28, 2017
26 / 66
Conjecture
Liang ()
report
April 28, 2017
27 / 66
Trees
Theorem (Chawathe and Krishna 2002) Every complete m-ary tree is anti-magic.
Theorem (Kaplan, Lev and Roditty 2009) Every tree with at most one vertex of degree 2 is anti-magic. (The proof has an error)
Liang ()
report
April 28, 2017
28 / 66
Regular graphs
Theorem (Hartsfild and Ringel 1990) Every 2-regular graph is anti-magic.
Theorem (Cranston 2009) Every k-regular bipartite graph is anti-magic.
Theorem (T-M Wang and G-H Zhang 2013) Regular Graph with Particular Factors is anti-magic.
Liang ()
report
April 28, 2017
29 / 66
Cartesian product graphs
Theorem (T-M Wang 2005) Let G be the Cartesian product of cycles Cn1 Cn2 · · · Cnk , then G is anti-magic.
Theorem (T-M Wang 2005) If R is an r-regular anti-magic graph with r > 1 then RCn is anti-magic.
Liang ()
report
April 28, 2017
30 / 66
(ω, k)-antimagic
Theorem (Wong and Zhu 2011) If G has a universal vertex, then G is weighted-2-antimagic. If G has a prime number of vertices and has a Hamiltonian path, then G is weighted-1-antimagic.
Theorem (Wong and Zhu 2011) If G has a prime number of vertices and has a Hamiltonian path, then G is weighted-1-antimagic.
Liang ()
report
April 28, 2017
31 / 66
(ω, k)-antimagic
Theorem (Hefetz 2005) Let G be a graph with q edges and at most one isolated vertex and no isolated edges is(ω, 2q − 4)-antimagic.
Theorem (Hefetz 2005) Let G be a graph with p > 2 vertices that admits a 1-factor is (p − 2)-antimagic.
Theorem (Hefetz 2005) Let G be a graph with p vertices and maximum degree n − k, where k ≥ 3 is any function of p is (3k − 7)-antimagic
Liang ()
report
April 28, 2017
32 / 66
Cartesian product graphs
Theorem (Zhang and Sun 2009) If R is an anti-magic regular graph then for any graph G, GR is anti-magic.
Theorem (Zhang and Sun 2009) Let G be the Cartesian product of many paths Pn1 Pn2 · · · Pnt , then G is anti-magic, provided that at least one of the paths has length at last 3.
Liang ()
report
April 28, 2017
33 / 66
Tree
Theorem (Chawathe and Krishna 2002) Every complete m-ary tree is anti-magic.
Theorem (Kaplan, Lev and Roditty 2009) Every tree with at most one vertex of degree 2 is anti-magic. (The proof has an error)
Liang ()
report
April 28, 2017
34 / 66
General graphs Theorem (Yilma 2013) If G is a graph with maximum degree ∆ = |V (G)| − 3, and |V (G)| ≥ 9 then G is anti-magic.
Theorem (Hefetz, Saluz and Tran 2005, 2010) If G is a graph with pk vertices which admits a Cp -factor, then G is anti-magic.
Theorem (Alon, Kaplan, Lev, Roditty and Yuster 2004) If G is a graph with minimum degree δ(G) ≥ C log |V (G)|, then G is anti-magic.
Theorem (Eccles 2014) The graphs of large linear size are anti-magic. Liang ()
report
April 28, 2017
35 / 66
Survey
[Joseph A. Gallian 1997 2016] A Dynamic Survey of Graph Labeling.
Liang ()
report
April 28, 2017
36 / 66
Results
Theorem (Liang, Wong and Zhu 2014) If T 6= K2 is a tree with at most one vertex of degree 2, then T is anti-magic.
Liang ()
report
April 28, 2017
37 / 66
Consider the structure of tree T .
Liang ()
report
April 28, 2017
38 / 66
Design the function φf (v).
Each vertex has at most one in-edge. Goal: the labeling of in-edge are distinct implies φf (v) are distinct.
Liang ()
report
April 28, 2017
39 / 66
characterize the graph T and f . The sum of labeling of the out-edge are the same(based on the group).
Liang ()
report
April 28, 2017
40 / 66
Trees Definition Suppose G is a graph and e = (u, v) is an edge of G. A subdivision of e is the operation of replacing e = (u, v) with a path (u, we , v), where we is a new vertex.
Liang ()
report
April 28, 2017
41 / 66
Trees
Definition Suppose T = (V, E) is a tree and A ⊆ E. Denote by T ∗ (A) the tree obtained by subdividing each edge in A. Let T ∗ = T ∗ (E) be the tree which obtained by subdivide each edge of T .
Liang ()
report
April 28, 2017
42 / 66
Trees Theorem (Liang, Wong and Zhu 2014) If T is a tree with V2 (T ) = ∅ and T ∗ is obtained from T by subdividing every edge, then T ∗ is anti-magic.
Liang ()
report
April 28, 2017
43 / 66
Trees Theorem (Liang, Wong and Zhu 2014) If T is a tree in which every vertex v ∈ V (T ) has an even number of sons, then for any A ⊆ E(T ), T ∗ (A) is obtained from T by subdividing every edge in A is anti-magic.
Liang ()
report
April 28, 2017
44 / 66
Trees Definition Let Vi (T ) beSthe set of vertices of T of degree i, and Veven (T ) = i>0 V2i (T ). If Veven (T ) induces a path, then we assume Veven (T ) = (v1 , v2 , . . . , vk−1 ), and P (v0 , v1 , . . . , vk ) is also a path.
Liang ()
report
April 28, 2017
45 / 66
Trees Theorem (Liang, Wong and Zhu 2014) Assume T is a tree, V2 (T ) induces a path Pk−1 and each component of T \ V2 (T ) is arbitrary complete d-ary tree but not a star, then T is anti-magic.
Liang ()
report
April 28, 2017
46 / 66
Trees Theorem (Liang, Wong and Zhu 2014) Assume T is a tree, Veven (T ) induces a path and |Veven (T )| is odd, then T is anti-magic.
Liang ()
report
April 28, 2017
47 / 66
Trees Theorem (Liang, Wong and Zhu 2014) Assume T is a tree, Veven (T ) induced a path, and |Veven (T )| is even. Let 2s = |Veven (T )|. If deg(vs ) 6= deg(vk ) + 1 or deg(vs+1 ) 6= deg(v0 ) + 1, then T is anti-magic.
Liang ()
report
April 28, 2017
48 / 66
Cartesian product graphs
Theorem (Liang and Zhu 2013) Assume G is a k-regular graph with k ≥ 2. For any connected graph H with |E(H)| ≥ |V (H)| − 1 ≥ 1, HG is anti-magic.
Liang ()
report
April 28, 2017
49 / 66
Cartesian product graphs
Theorem (Liang and Zhu 2013) Assume H is a graph with |E(H)| ≥ |V (H)| − 1. If each component of H has a vertex of odd degree, then the prism of H is anti-magic.
Theorem (Liang and Zhu 2013) If H is a graph with m ≥ 2n − 2, where m = |E(G)|, and n = |V (G)|, then the prism of H is anti-magic.
Theorem (Liang and Zhu 2013) If G is a graph with ∆(G) = |V (G)| − 1, GP2 is anti-magic.
Liang ()
report
April 28, 2017
50 / 66
Cartesian product graphs Theorem (Liang and Zhu 2013) Assume G is obtained from a regular graph by adding a universal vertex u, then for any positive integer k, Gk is anti-magic.
Liang ()
report
April 28, 2017
51 / 66
Cartesian product graphs
Theorem (Liang and Zhu 2013) If each of G1 and G2 is obtained from a (d − 1)-regular graph by adding a universal vertex, then G1 G2 is anti-magic.
Theorem (Liang and Zhu 2013) If G1 , G2 are double star graphs, then G1 G2 is anti-magic.
Liang ()
report
April 28, 2017
52 / 66
Regular graphs
Theorem (Liang and Zhu 2014, Cranston, Liang and Zhu 2015, Chang, Liang, Pan and Zhu 2015 ) Let k > 1 be a positive integer, and G is a k-regular graph, then G is anti-magic.
Liang ()
report
April 28, 2017
53 / 66
idea
Definition : Given v ∗ is a vertex of G, let Li = {u : dG (u, v ∗ ) = i}. Goal 1 : For any u, v ∈ Li , s(v) 6= s(u). Goal 2 : For any x ∈ Li−1 , y ∈ Li , s(x) > s(y). Idea : For each vertex v ∈ Li , N (v) ⊆ Li−1
Liang ()
report
S
Li
S
Li+1 .
April 28, 2017
54 / 66
Partition
Partition : Fix the vertex v ∗ of G, Let L0 S = {v ∗ } and for i ≥ 1, ∗ Li = {u : dG (u, v ) = i}. Assume V (G) = pi=0 Li .
Liang ()
report
April 28, 2017
55 / 66
Partition
Liang ()
report
April 28, 2017
56 / 66
Labeling Let Ei = E(G[Li ]) ∪ E(G[Li−1 , Li ]). We label the edges in Ep , Ep−1 , . . . , E1 recursively in this order.
Liang ()
report
April 28, 2017
57 / 66
Labeling The partial-sum ω(v) of a vertex v is the sum of labels of all edges incident with v except σ(v). Ordering {ω(v)|v ∈ Lj } by {ωi |1 ≤ j ≤ |Lj |} such that ωi ≤ ωi+1 , for all i.
So for any u, v ∈ Li , s(v) 6= s(u). Liang ()
report
April 28, 2017
58 / 66
Proof of theorem
We label the open trail in Gσ [Li , Li−1 ] by [s, l] as follow:
Liang ()
report
April 28, 2017
59 / 66
Further work
1 : Tree. Cartesian product graph. Halin graph. Cartesian power graph. 2 : (ω, k)- anti-magic. 3 : New tool for general graph. 4 : Hyper graph version. 5 : Directed graph version.
Liang ()
report
April 28, 2017
60 / 66
References N. Alon, G. Kaplan, A. Lev, Y. Roditty, and R. Yuster, Dense graphs are antimagic, J. Graph Theory, 47 (2004), 297-309. A. Rosa, On certain valuations of the vertices of a graph, Theory of Graphs (Internat. Symposium, Rome, July 1966), Gordon and Breach, N. Y. and Dunod Paris (1967) 349-355. Y. Cheng, A new class of antimaig Cartesian product graphs, Discrete Math., 308 (2008), 6441-6448. Y. Cheng, Lattice grids and prisms are antimagic, Theoret. Comput. Sci., 374 (2007), 66-73. P. D. Chawathe, and V. Krishna, Antimagic labelings of complete m-ary trees, Number theory and discrete mathematics (Chandigarh, 2000), 77-80, Trends Math., Birkh¨ auser, Basel, 2002. D. W. Cranston, Regular bipartite graphs are antimagic, J. Graph Theory, 60 (2009), 173-182. Liang ()
report
April 28, 2017
61 / 66
References
D. W. Cranston, Y. Liang and X. Zhu, Regular Graphs of Odd Degree Are Antimagic, J Graph Theory, 80 (2015), 28-33. T. Eccles, Graphs of large linear size are antimagic, manuscript, 2014, arXiv:1409.3659 J. A. Gallian, A Dynamic survey on Graph Labeling, The Electronic Journal of Combinatorics, 15 (2008). S. W. Golomb, How to number a graph, Graph theory and computing, 23-37. Academic Press, New York, 1972. P. Hall , On representatives of subsets, J.Lond. Mat. Sc., 10 (1935), 26-30.
Liang ()
report
April 28, 2017
62 / 66
References N. Hartsfield, and G. Ringel, Pearls in Graph Theory, Academic Press, INC., Boston, 1990, pp. 108-109, Revised version 1994. D. Hefetz, Anti-magic graphs via the Combinatorial Nullstellensatz, J. Graph Theory, 50 (2005), 263-272 D. Hefetz, H.T.T. Tran, and A. Saluz, An application of the Combinatorial Nullstellensatz to a graph labelling problem, J. Graph Theory, 65 (2010), 70-82. G. Kaplan, A. Lev, and Y. Roditty, On zero-sum partitions and anti-magic trees, Discrete Math., 309 (2009), 2010-2014. Y. Liang, T. Wong, and X. Zhu, Anti-magic labeling of trees, Discrete Math. 331 (2014), 9-14. Y. Liang, and X. Zhu, Anti-magic labeling of cubic graphs, J. Graph Theory, 75 (2014), 31-36. Liang ()
report
April 28, 2017
63 / 66
References Y. Liang, and X. Zhu, Anti-magic labelling of Cartesian product of graphs, Theoret. Comput. Sci., 477 (2013), 1-5. G. Ringel, Problem 25, in Theory of Graphs and its Applications, Proc. Symposium Smolenice 1963, Prague (1964), 162. A. Rosa, On certain valuations of the vertices of a graph, 1967 Theory of Graphs (Internat. Sympos., Rome, 1966) pp. 349-355 Gordon and Breach, New York; Dunod, Paris. J. Sedl´aˇcek, Problem 27, in Theory of Graphs and its Applications, Proc. Symposium Smolenice, June, (1963) 163-167. T. M. Wang, Toroidal grids are antimagic, computing and combinatorics, Lecture Notes in Comput. Sci., 3595, Springer, Berlin(2005), 671-679. Liang ()
report
April 28, 2017
64 / 66
References
T. M. Wang, and C. C. Hsiao, On antimagic labeling for graph products, Discrete Math., 308 (2008), 3624-3633. D. B. West, Introduction to Graph Theory, Second Edition, Prentice Hall, Inc., Upper Saddle River, NJ, 1996. T. Wong, and X. Zhu, Antimagic labelling of vertex weighted graphs , J. Graph Theory, 70 (2012), 348-359. Z. B. Yilma, Antimagic properties of graphs with large maximum degree , J. Graph Th., 72 (2013), no. 4, 367-373. Y. Zhang, and X. Sun, The antimagicness of the Cartesian product of graphs, Theoret. Comput. Sci., 410 (2009), 727-735.
Liang ()
report
April 28, 2017
65 / 66
Thanks for your attention!
Liang ()
report
April 28, 2017
66 / 66