Abstract
A graph G is called super edge-magic if there exists a bijection , where , such that is a constant for every edge . Such a case, f is called a super edge magic labeling of G. A bipartite graph G with partite sets A and B is called consecutively super edge-magic if there exists a super edge-magic labeling f with the property that and . The super edge-magic deficiency of a graph G, denoted by , is either the minimum nonnegative integer n such that is super edge-magic or +∞ if there exists no such n. The consecutively super edge-magic deficiency of a bipartite graph G, denoted by , is either the minimum nonnegative integer n such that is consecutively super edge-magic or +∞ if there exists no such n. In this paper, we study the super edge-magic deficiency of some graphs. We investigate the (consecutively) super edge-magic deficiency of forests with two components. We also investigate the super edge-magic deficiency of a 2-regular graph and join product of with an isolated vertex.
Keywords: (Consecutively) super edge-magic graph, (Consecutively) super edge-magic deficiency, Forest with two components, 2-regular graph, Join product graph
(Consecutively) super edge-magic graph; (Consecutively) super edge-magic deficiency; Forest with two components; 2-regular graph; Join product graph
1. Introduction
Let G be a finite and simple graph having vertex set and edge set , where and . A labeling of G is a bijection . Weight of an edge xy under the labeling f is . The labeling f is called an edge-magic labeling if for any . In such a case, G is called an edge-magic graph and the constant k is called the magic constant of the labeling f. An edge-magic labeling f of G with additional property that is called a super edge-magic (SEM) labeling. Thus, a SEM graph is a graph that admits a SEM labeling. The concept of an edge-magic labeling was introduced by Kotzig and Rosa in 1970 [1]. Meanwhile, the terminology of a SEM labeling was introduced by Enomoto et al. in 1998 [2]. In 2001, Muntaner-Batle [3] introduced the concept of special SEM labeling of a bipartite graph. A special SEM labeling of a bipartite graph G with partite sets A and B is a SEM labeling f of G with the property that and . In 2007, Oshima [4] called such a labeling consecutively SEM labeling.
The following lemma proved by Figueroa-Centeno et al. [5] provides sufficient and necessary conditions for a SEM graph.
Lemma 1.1
[5] A graph G is SEM if and only if there exists a bijection such that the set of all edge-sums consists of q consecutive integers. In this case, f extends to be a SEM labeling of G with magic constant .
The next lemma proved by Enomoto et al. [2] gives sufficient condition for non-existence of a SEM labeling of a graph.
Lemma 1.2
[2] If G is a SEM graph then .
Moreover, Kotzig and Rosa [1] also proved that for every graph G there exists a nonnegative integer n such that is an edge-magic graph. This fact leads them to introduce the concept of edge-magic deficiency of a graph. The edge-magic deficiency of a graph G, , is defined as the minimum nonnegative integer n such that is an edge-magic graph. This concept motivated Figueroa-Centeno et al. [6] to introduce the concept of super edge-magic deficiency of a graph. The super edge-magic deficiency (SEMD) of a graph G, , is defined as either the minimum nonnegative n such that is a SEM graph or +∞ if there exists no such n. In 2016, Ichishima et al. [7] give the notion of consecutively super edge-magic deficiency of a bipartite graph. The consecutively super edge-magic deficiency (consecutively SEMD) of a bipartite graph G, , is defined to be either the smallest nonnegative integer n with the property that is consecutively SEM or +∞ if there exists no such n.
Many researchers have investigated the SEMD of some classes of graphs. The complete results on this subject can be seen in a dynamic survey of graph labeling by Gallian [8]. In this paper, we study the (consecutively) SEMD of forests with two components, where its components are non isomorphic combs, isomorphic a subdivision of and , and union of a comb and a subdivision of or . Moreover, we find SEMD of 2-regular graph . By applying Ichishima's et al. [9] and Cichacz's et al. [10] methods to this result, we prove that some 2-regular graphs with a large order have zero SEMD, and then we also obtain the SEMD of union of cycles and paths. In addition, we study the SEMD of graph for any integer and .
2. The SEMD of forests with two components
Figueroa-Centeno et al. [6] proved that the SEMD of all forests is finite. Next, the same authors [11] investigated the SEMD of some classes of forests with two components, such as , and . Based on these results, they gave the following conjecture.
Conjecture 2.1
[11] If F is a forest with two components then .
Inspired by Conjecture 2.1, many researchers have investigated the SEMD of some forests with two components. Baig et al. [12] found the SEMD of union of combs and stars. Javed et al. [13] studied the SEMD of forests with two components consisting of combs, generalized combs, and stars. In particular, they found the SEMD of two isomorphic combs. Imran and Mukhtar [14] showed that forests with two components consisting of stars and subdivision of stars have zero SEMD. Krisnawati et al. [15] investigated SEMD of two non isomorphic of a subdivision of and .
In this section, we study SEMD of disjoint union of two non isomorphic combs. We also continue Krisnawati's et al. [15] work to find SEMD of disjoint union of two isomorphic of a subdivision of and . We prove that these graphs have zero SEMD. Some of these graphs also have zero consecutively SEMD. Additionally, we investigate the (consecutively) SEMD of union of a comb and a subdivision of or .
A comb, with , is a tree consisting of path , whose vertices are , together with edges . We investigate zero (consecutively) SEMD of two non isomorphic combs in the following theorem.
Theorem 2.2
Let . For any and , .
Proof
Let be a graph with vertex set and edge set as follows.
Thus, is a graph of order and size .
For any and , define a labeling . Without loss of generality, we assume . We consider the proof based on the value of n.
Case 1. n is odd.
Case 2. n is even.
It is not hard to verify that the set of all edge-sums generated by f is . By Lemma 1.1, f extends to a SEM labeling of with magic constant . Hence, for any and , .
Next, let A and B be partite sets of .
For odd n,
For even n,
Since and , so is a consecutively SEM graph. Thus, for any and , . □
As an illustration of the proof of Theorem 2.2, see the Fig. 1.
Figure 1.
The (consecutively) SEM labeling of Cb3 ∪ Cb6 and Cb4 ∪ Cb6.
A subdivision of a star , denoted by , where , , is a graph obtained by inserting vertices to each edge of the star . Now, we study (consecutively) SEMD of disjoint union of two isomorphic of subdivision of and in the following theorems.
Theorem 2.3
Let . For any even , and for any odd , .
Proof
Let be a graph with
Thus, is a graph of order and size .
Let be an even integer. Define a labeling as follows.
The set of all edge-sums generated by f is . By Lemma 1.1, f extends to a SEM labeling of with magic constant . Hence, for any even , .
Moreover, let A and B be partite sets of , where
It is easy to see that and . So, is a consecutively SEM graph. Thus, for any even .
Now, to show for any odd , let us consider the graph , where and . Define a labeling as follows.
To label the rest of vertices, we consider the following cases.
- (i)
For and , set
- (ii)
For and , set
The set of all edge-sums generated by g is . By Lemma 1.1, g extends to a SEM labeling of with magic constant . Hence, for any odd , . □
Theorem 2.4
Let . For any , .
Proof
The graph is a graph of order and size . Let
Define a labeling as follows.
To label the vertices of the second component of , we consider two following cases.
Case 1. n is even.
For , set and
For , set and
For even , label as follows.
- (i)
For ,
- (ii)
For ,
Case 2. n is odd.
Set and
For all cases, the set of all edge-sums generated by the labeling f is . By Lemma 1.1, f extends to a SEM labeling of with magic constant . Therefore, for any , . □
To clarify the proof of Theorem 2.3, Theorem 2.4, see Fig. 2.
Figure 2.
The (consecutively) SEM labeling of 2T(6,5,5) and 2T(7,6,6,7).
Next, we give the upper bound of (consecutively) SEMD of and .
Theorem 2.5
Let . For any , .
Proof
Let be a graph having vertex and edge sets as follows.
For any , let be a graph with
Thus, has vertices and edges.
For , label and with and , respectively. These vertex labelings can be extended to a SEM labeling of with magic constant .
For , define a labeling as follows.
Case 1. n is even.
, , for , and
Case 2. n is odd.
, for , and
For all cases, it can be verified that the set of all edge-sums generated by the labeling f is . By Lemma 1.1, f extends to a SEM labeling of with magic constant . Hence, for any .
Furthermore, let A and B be partite sets of where
Since for any , and , then is a consecutively SEM graph. Thus, . □
Theorem 2.6
Let . For any ,
Proof
Define the vertex and edge sets of as follows.
For any even , let be a graph with
Next, define a labeling as follows.
For , set and
For , set and
For even , label as follows.
- (i)
For ,
- (ii)
For ,
It can be checked that the set of all edge-sums is . Hence by Lemma 1.1, f extends to a SEM labeling of with magic constant . Therefore, for any even .
Now, for any odd , let be a graph with
Next, define a labeling as follows.
Under the labeling g, the set of all edge-sums is . Hence by Lemma 1.1, g extends to a SEM labeling of with magic constant . Therefore, for any odd n. □
Fig. 3 shows an illustration of the proof of Theorem 2.5, Theorem 2.6.
Figure 3.
The (consecutively) SEM labeling of Cb7 ∪ T(8,7,7)∪3K1 and the SEM labeling of Cb6 ∪ T(7,6,6,7)∪5K1.
3. The SEMD of and related graphs
Enomoto et al. [2] stated that a cycle is SEM if and only if n is odd. Figueroa-Centeno et al. [6] investigated the SEMD of the cycle . They also proved the following result.
Theorem 3.1
[6] If G is a graph of size such that every vertex in has even degree then .
In 2005, Figueroa-Centeno et al. [11] investigated the SEMD of , , , and conjectured that
In 2009, Holden et al. [16] showed that for any integer , for any integer , and for any integer have a strong vertex-magic labeling, which is equivalent to a SEM labeling. Based on their results, they proposed the following conjecture.
Conjecture 3.2
[16] A 2-regular graph of odd order is SEM if and only if it is not one of , , or .
Motivated by Conjecture 3.2, Figueroa-Centeno et al. [17] showed that some 2-regular graphs with two components are SEM. They proved that is SEM if and only if is even, is SEM if and only if is odd, is SEM if and only if is even, and is SEM for any even and odd . Ichishima and Oshima [18] determined the SEMD of for even m and n, and for arbitrary n when and 7. In this section, we investigate the SEMD of .
Theorem 3.3
The SEMD of a 2-regular graph is given by
and .
Proof
First, let be a graph with vertex and edge sets
Next, we show that for . Let where . Define a labeling as follows.
To label , for , we consider the following cases.
Case 1. is even.
Case 2. is odd.
It can be checked that for all cases, the set of all edge-sums is . By Lemma 1.1, f extends to a SEM labeling of with magic constant . Hence, for .
Next, we prove that for . Let , . For , label , and with , and , respectively. For , define a labeling as follows.
It is not hard to verify that the set of all edge-sums is . Hence by Lemma 1.1, f extends to a SEM labeling of with magic constant . Therefore, for .
Now, let for , . Define a labeling as follows.
where z is the vertex of . For the other vertices, we consider the following cases.
Case 1. For any odd , set
Case 2. For any even , set
To label the rest of vertices, we consider two subcases.
- •
Subcase 2.1. For , set
- •
Subcase 2.2. For , set
It can be checked that under the labeling f, the set of all edge-sums is . By Lemma 1.1, f extends to a SEM labeling of G with magic constant . Thus, for .
To show that , label with , respectively. Finally, for is a direct consequence of Theorem 3.1. □
In 2013, Ichishima et al. [9] introduced the concept of a pseudo SEM graph. A graph G with isolated vertices is called pseudo SEM if there exists a bijection such that the set consists of consecutive integers. Such a function is called a pseudo SEM labeling. They also proved the following results.
Theorem 3.4
[9] Let m be an odd integer. If is any pseudo SEM 2-regular graph then is a SEM 2-regular graph, where and are the greatest common divisor and the least common multiple of integers m and , respectively.
Corollary 3.5
[9] Let m be an odd integer such that and , . If is any pseudo super edge-magic graph then is a SEM graph.
By applying Theorem 3.4 and Corollary 3.5 to a part of Theorem 3.3, we have the following Lemma.
Lemma 3.6
For any odd m and , the following SEMD of some graphs are hold.
Proof
It is easy to verify that the labeling f in the proof of Theorem 3.3 is a pseudo SEM labeling of for and . By applying Theorem 3.4 and Corollary 3.5 to these 2-regular pseudo SEM graphs, we obtain (i) and (iii). By removing edges of the SEM 2-regular graph with the smallest edge-sums , where , we get (ii). In a similar way, we obtain the result (iv) by removing 12 edges of the SEM 2-regular graph with the smallest edge-sums , where . □
As an illustration of the proof of Lemma 3.6 (i) and (ii), see Fig. 4.
Figure 4.
(a) The pseudo SEM labeling of 2C3 ∪ C10 ∪ K1; (b) The SEM labeling of 2C9 ∪ C30 ∪ C3, which is obtained by applying Corollary 3.5 to Fig. 4(a) for m = 3; (c) The SEM labeling of 16P3 ∪ C3, which is obtained by removing 16 edges with the smallest edge-sums {27,28,⋯,42} of the graph in Fig. 4(b).
Cichacz et al. [10] in 2017 introduced a new method to expand the classes of known (strong) vertex-magic labeling of 2-regular graphs as the following theorem.
Theorem 3.7
[10] For , let be an integer and . If G is (strong) vertex-magic then is (strong) vertex-magic for every integer r.
Ngurah [19] discovered that the method introduced in the proof of Theorem 3.7 is also valid for pseudo SEM graphs. By applying this fact to our results, we have the following lemma.
Lemma 3.8
For any odd m and , the following SEMD of some graphs are satisfied.
Proof
The result (i) and (iv) are obtained by applying Theorem 3.7 to the fact that for and are pseudo SEM, respectively. We gain (ii) and (iii) from (i) by removing edges with the edge-sums and , where , respectively. Moreover, we obtain the result (v), (vi), (vii), and (viii) from (iv) by deleting edges with the edge-sums , , , and , where , respectively. □
Fig. 5 shows the construction given in the proof of Lemma 3.8 (i), (ii), and (iii).
Figure 5.
(a) The SEM labeling of 2C9 ∪ C30 ∪ C3, which is obtained by applying Theorem 3.7 to Fig. 4(a) for r = 1; (b) The SEM labeling of 3P3 ∪ C9 ∪ C30 ∪ C3, which is obtained by removing 3 edges with the smallest edge-sums {27,28,29} in Fig. 5(a); (c) The SEM labeling of 3P3 ∪ C9 ∪ 3P10 ∪ C3, which is obtained by removing 3 edges with the smallest edge-sums {30,31,32} in Fig. 5(b).
4. The SEMD of join product of union of a star and a path with an isolated vertex
To present our results, we need the concept of dual labeling. A dual labeling of a SEM labeling f is defined as for all and for all . Baskoro et al. [20] proved that the dual of a SEM labeling is also a SEM labeling.
Ngurah and Simanjuntak [21] studied the SEMD of join products of a path, a star, and a cycle, respectively, with isolated vertices. Generally, they showed that the join product of a SEM graph with isolated vertices has finite SEMD. In [22], the same authors investigated the SEMD of join product of a graph G which has certain properties with an isolated vertex. They gave a necessary condition for to have zero SEMD as the following lemma.
Lemma 4.1
[22] Let G be a graph with no cycle and minimum degree one. If then G is a tree or a forest.
They showed that Lemma 4.1 is attainable. In particular, they proved that the join product of some forests with an isolated vertex has zero SEMD, such as if and only if and if and only if . In this section, we study the SEMD of join product graph for any integer and .
For any integer and , let be a graph having vertex and edge sets:
Thus, the graph has vertices and edges.
Now, we give necessary and sufficient conditions of to have zero SEMD for .
Theorem 4.2
if and only if ; and for , if and only if .
Proof
Firstly, we show that for , . For , label with ; ; and , respectively. These vertex labelings can be extended to a SEM labeling of for .
Next, we prove that for every . Suppose that for every . Then by Lemma 1.1, there exists a bijection such that is a set of consecutive integers. Since has vertices and edges, there are two possibilities of S, namely or and they are dual to each other. Thus, we can consider . Based on the degree of all vertices of , the sum of all elements in S contains times of label of z, times of label of c, three times of label of and two times of label of the remaining vertices. Hence,
or
Furthermore, to get edge-sums 3, 4, and 5 in S, then the vertices of labels 1, 2 and 3 are adjacent to each other or the vertex of label 1 is adjacent to the vertices of labels 2, 3, and 4. By using the above equation and this fact, we have two following cases.
Case 1. There is a triangle with the vertices of labels 1, 2 and 3. Since every triangle in share a common vertex z, hence and . We can check that the equation has no solution for . If then and its solution is and . However, the solution does not lead to a SEM labeling of for any .
Case 2. The vertex of label 1 is adjacent to the vertices of label 2, 3, and 4. Since the vertex of label 1 must have minimum degree 3, the possibilities of its vertex are z, c or .
- (i)
If then , where . It is easy to verify that this equation has no solution.
- (ii)
If then . It can be checked that this equation has no solution if . If then . However, this solution does not lead to a SEM labeling of for any .
- (iii)
If then . Since has degree three, which is adjacent to z, thus . If then . If then either or . If then , thus . However, this solution cannot be extended to a SEM labeling of .
Hence, for every . Therefore, if and only if .
Now, we show that for , if and only if . For and , label with and , respectively. Moreover, for and , label with and , respectively. It can be checked that these labelings extend to a SEM labeling of and for . Conversely, suppose that for . Then for , we have
and for ,
By a similar argument as in the proof of , we can show that and for every . Hence, for , if and only if . □
Since , , and do not SEM for almost all of n, thus we try to find its SEMD. Our result is as follows.
Theorem 4.3
For any integer , the SEMD of is given by:
- (i)
if and ; and ; and and .
- (ii)
if and , where r is any positive integer.
Proof
- (i)
As a consequence of Theorem 4.2, we obtain that for , for , and for . Hence, the remaining case is to show that for ; for ; and for . Thus, let us consider the graph and suppose w is the vertex of .It is not hard to check that these labelings extend to a SEM labeling of for ; for ; and for .
- –
For and , label with and , respectively.- –
For and , label with ; ; ; and , respectively.- –
For and , label with ; ; and , respectively.- (ii)
Let , where and . Define a labeling as follows.
where , . For ,
The remaining labels are used to label isolated vertices of H. It can be verified that the set of all edge-sums is a consecutive integer , where . By Lemma 1.1, f extends to a SEM labeling of H with magic constant . Therefore, for any and , , . □
5. Conclusion
In this paper, we study the (consecutively) SEMD of some graphs. We find the exact value or upper bound of the (consecutively) SEMD of forests with two components. We also find the exact value of the SEMD of a 2-regular graph for almost all of n. By using this and previous known results, we obtain the exact value or upper bound of the SEMD of some 2-regular graphs and union of cycles and paths. Moreover, we provide the necessary and sufficient conditions of to gain zero SEMD and the upper bound of SEMD of this graph for the remaining cases.
Declarations
Author contribution statement
V.H. Krisnawati: Conceived and designed the analysis; Analyzed and interpreted the data; Wrote the paper.
A.A.G. Ngurah: Analyzed and interpreted the data; Contributed analysis tools or data; Wrote the paper.
N. Hidayat, A.R. Alghofari: Conceived and designed the analysis; Contributed analysis tools or data.
Funding statement
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of interests statement
The authors declare no conflict of interest.
Additional information
No additional information is available for this paper.
Acknowledgements
The authors would like to thank to the reviewers for their valuable comments and suggestions.
Contributor Information
Vira Hari Krisnawati, Email: virahari@ub.ac.id.
Anak Agung Gede Ngurah, Email: aag.ngurah@unmer.ac.id.
References
- 1.Kotzig A., Rosa A. Magic valuation of finite graphs. Can. Math. Bull. 1970;13:451–461. [Google Scholar]
- 2.Enomoto H., Llado A., Nakamigawa T., Ringel G. Super edge-magic graphs. SUT J. Math. 1998;34:105–109. [Google Scholar]
- 3.Muntaner-Batle F.A. Special super edge-magic labelings of bipartite graphs. J. Comb. Math. Comb. Comput. 2001;39:107–120. [Google Scholar]
- 4.Oshima A. Consecutively super edge-magic tree with diameter 4. Rev. Bull. Calcutta Math. Soc. 2007;15:87–90. [Google Scholar]
- 5.Figueroa-Centeno R., Ichishima R., Muntaner-Batle F.A. The place of super edge-magic labelings among other classes of labelings. Discrete Math. 2001;231:153–168. [Google Scholar]
- 6.Figueroa-Centeno R., Ichishima R., Muntaner-Batle F.A. On the super edge-magic deficiency of graphs. Electron. Notes Discrete Math. 2002;11:299–314. [Google Scholar]
- 7.Ichishima R., Muntaner-Batle F.A., Oshima A. The consecutively super edge-magic deficiency of graphs and related concepts. Electron. J. Graph Theory Appl. 2020;8:71–92. [Google Scholar]
- 8.Gallian J.A. A dynamic survey of graph labeling. Electron. J. Comb. 2019;DS6 [Google Scholar]
- 9.Ichishima R., Muntaner-Batle F.A., Oshima A. Enlarging the classes of super edge-magic 2-regular graphs. AKCE Int. J. Graphs Comb. 2013;340:129–146. [Google Scholar]
- 10.Cichacz S., Froncek D., Singgih I. Vertex magic total labelings of 2-regular graphs. Discrete Math. 2017;340:3117–3124. [Google Scholar]
- 11.Figueroa-Centeno R., Ichishima R., Muntaner-Batle F.A. Some new results on the super edge-magic deficiency of graphs. J. Comb. Math. Comb. Comput. 2005;55:17–31. [Google Scholar]
- 12.Baig A.Q., Ahmad A., Baskoro E.T., Simanjuntak R. On the super edge-magic deficiency of forests. Util. Math. 2011;86:147–159. [Google Scholar]
- 13.Javed S., Riasat A., Kanwal S. On super edge-magicness and deficiencies of forests. Util. Math. 2015;98:149–169. [Google Scholar]
- 14.Imran M., Mukhtar A. On super edge-magic total labeling of forests consisting of stars and subdivided stars. Int. J. Math. Soft Comput. 2017;7:1–14. [Google Scholar]
- 15.Krisnawati V.H., Ngurah A.A.G., Hidayat N., Alghofari A.R. On the super edge-magic deficiency of forests. AIP Conf. Proc. 2019;2192 doi: 10.1016/j.heliyon.2020.e05561. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Holden J., McQuillan D., McQuillan J.M. A conjecture on strong magic labelings of 2-regular graphs. Discrete Math. 2009;309:4130–4136. [Google Scholar]
- 17.Figueroa-Centeno R., Ichishima R., Muntaner-Batle F.A., Oshima A. A magical approach to some labeling conjectures. Discuss. Math., Graph Theory. 2011;31:79–113. [Google Scholar]
- 18.Ichishima R., Oshima A. On the super edge-magic deficiency of 2-regular graphs with two components. Ars Comb. 2016;129:437–447. [Google Scholar]
- 19.Ngurah A.A.G. On (super) edge-magic deficiency of some classes of graphs. Electron. J. Graph Theory Appl. 2019 submitted for publication. [Google Scholar]
- 20.Baskoro E.T., Sudarsana I.W., Cholily Y.M. How to construct new super edge-magic graphs from some old ones. J. Indones. Math. Soc. 2005;11:155–162. [Google Scholar]
- 21.Ngurah A.A.G., Simanjuntak R. Super edge-magic deficiency of join product graphs. Util. Math. 2017;105:279–289. [Google Scholar]
- 22.Ngurah A.A.G., Simanjuntak R. On the super edge-magic deficiency of join product and chain graphs. Electron. J. Graph Theory Appl. 2019;7:157–167. [Google Scholar]





