Sufficient Conditions for a Graph kG Admitting all [1,k]-factors

Jie Wu1
1School of Economic and management, Jiangsu University of Science and Technology, Zhenjiang, Jiangsu 212100, China.

Abstract

Let g,f:V(G){0,1,2,3,} be two functions satisfying g(x)f(x) for every xV(G). A (g,f)-factor of G is
defined as a spanning subgraph F of G such that g(x)dF(x)f(x) for every xV(G). An (f,f)-factor is simply called
an f-factor. Let φ be a nonnegative integer-valued function defined on V(G). Set
Dg,feven={φ:g(x)φ(x)f(x) for every xV(G) and xV(G)φ(x) is even}.
If for each φDg,feven, G admits a φ-factor, then we say that G admits all (g,f)-factors. All (g,f)-factors
are said to be all [1,k]-factors if g(x)1 and f(x)k for any xV(G). In this paper, we verify that for a connected multigraph
G satisfying NG(X)=V(G) or |NG(X)|>(1+1k+1)|X|1 for every XV(G), kG admits all [1,k]-factors, where
k2 is an integer and kG denotes the graph derived from G by replacing every edge of G with k parallel edges.

Keywords: Graph, Spanning subgraph, φ-factor, [1,k]-factors

1. Introduction

The graphs discussed here are multigraphs, which may admit multiple edges but do not admit loops. A graph is called a simple graph if it admits neither multiple edges nor loops. For convenience, we simply call a multigraph a graph when we show notations and definitions. Let G be a graph. We use V(G) and E(G) to denote the vertex set and the edge set of G, respectively. For xV(G), we write dG(x) for the degree of x in G, and NG(x) for the set of the vertices adjacent to x in G. We write NG(X)=xXNG(x) for any XV(G). We denote by I(G) the set of isolated vertices of G, and by ωk(G) the number of components of G with order at least k. Specially, we write i(G)=|I(G)| and ω(G)=ω1(G). Let kG denote the graph derived from G by replacing every edge of G with k parallel edges.

Let g,f:V(G){0,1,2,3,} be two functions satisfying g(x)f(x) for every xV(G). A (g,f)-factor of G is defined as a spanning subgraph F of G such that g(x)dF(x)f(x) for every xV(G). An (f,f)-factor is simply called an f-factor. If g(x)1 and f(x)k for any xV(G), then a (g,f)-factor is called a [1,k]-factor, where k1 is a fixed integer. A [1,1]-factor is simply called a 1-factor.

Let φ be a nonnegative integer-valued function defined on V(G). In the following, we write Dg,feven={φ:g(x)φ(x)f(x) for every xV(G)and xV(G)φ(x) is even}. If for each φDg,feven, G admits a φ-factor, then we say that G admits all (g,f)-factors. All (g,f)-factors are said to be all [1,k]-factors if g(x)1 and f(x)k for any xV(G).

Lots of authors studied factors [1-16] and all factors [17-20] of graphs. The neighborhood conditions for graphs having factors were derived by many authors [21-27]. Lu, Kano and Yu [18] characterized a graph G such that kG admits all [1,k]-factors. Lu, Kano and Yu’s results generalized Tutte’s 1-factor theorem [28].

Theorem 1.([18]). Let k2 be an integer, and let G be a connected multigraph. Then kG admits all [1,k]-factors if and only if ki(GX)+ωk+1(GX)|X|+1 for any XV(G).

Using Theorem 1, we verify some results related to all [1,k]-factors in graphs. Our main results will be shown in Section 2.

2. Next section

In this section, we discuss the relationship between neighborhood and all [1,k]-factors, and verify two results related to all [1,k]-factors.

Theorem 2.Let k be an integer with k2, and let G be a connected multigraph. If G satisfies NG(X)=V(G) or |NG(X)|>(1+1k+1)|X|1 for every XV(G), then kG admits all [1,k]-factors.

 Assume that G satisfies the hypothesis of Theorem 2, but kG does not admit all [1,k]-factors. Then by Theorem 1, we derive (1)ki(GX)+ωk+1(GX)|X|+2 for some subset XV(G). The following proof will be divided into two cases by the value of i(GX).

Case 1. i(GX)=0.

According to (1), we get ωk+1(GX)|X|+2, which implies that GX has at least |X|+2 components with order at least k+1. We denote by Y the set of vertices of any |X|+1 components of GX with order at least k+1. It is clear that NG(Y)V(G). Thus, we derive (2)|NG(Y)|>(1+1k+1)|Y|1.

On the other hand, we easily see that |NG(Y)||X|+|Y|. Combining this with (2), we have |X|+|Y||NG(Y)|>(1+1k+1)|Y|1, namely, (3)|X|>1k+1|Y|1.

Note that |Y|(k+1)(|X|+1). Then using (3), we admit |X|>1k+1|Y|1(|X|+1)1=|X|, this is a contradiction.

Case 2. i(GX)>0.

Obviously, NG(V(G)X)V(G). Thus, we have (4)|NG(V(G)X)|>(1+1k+1)|V(G)X|1=(1+1k+1)(|V(G)||X|)1. Using (4) and |NG(V(G)X)||V(G)|i(GX), we obtain |V(G)|i(GX)|NG(V(G)X)|>(1+1k+1)(|V(G)||X|)1, that is, (5)|V(G)|+(k+1)i(GX)<(k+2)|X|+k+1.

On the other hand, we easily see that (6)|V(G)|i(GX)+|X|+(k+1)(|X|+2)=i(GX)+(k+2)|X|+2(k+1).

It follows from (5) and (6) that (k+2)|X|+k+1>|V(G)|+(k+1)i(GX)(GX)+(k+2)|X|+2(k+1)+(k+1)i(GX)=(k+2)i(GX)+(k+2)|X|+2(k+1) which implies (k+2)i(GX)+k+1<0, which is a contradiction. Theorem 2 is proved. ◻

Theorem 3. For any positive integer k with k2, there exist infinitely many graphs G that satisfy NG(Y)=V(G) or |NG(Y)|(1+1k+1)|Y|1 for every YV(G), but kG does not admit all [1,k]-factors.

 Let r1 be an integer. We construct a graph G=Kr((r+2)Kk+1), where Kr denotes the complete graph of order r, Kk+1 denotes the complete graph of order k+1 and means “join”.

Next, we demonstrate that NG(Y)=V(G) or |NG(Y)|(1+1k+1)|Y|1 for every YV(G). We shall consider two cases by the value of |Y|.

Case 1. |Y|=1.

In this case, we obviously have that |NG(Y)|k+r>1k+1=(1+1k+1)|Y|1.

Case 2. |Y|2.

Case 2.1. YV(Kr).

It is clear that NG(Y)=V(G).

Case 2.2. YV(Kr)=.

Let Kk+1(i), 1ir+2, denote the disjoint copies of Kk+1 in GV(Kr). Write b1=#{i:|YV(Kk+1(i))|=1}, b2=#{i:|YV(Kk+1(i))|=2},, and bk+1=#{i:|YV(Kk+1(i))|=k+1}. Thus, we derive |Y|=b1+2b2++(k+1)bk+1 and |NG(Y)|=r+kb1+(k+1)(b2++bk+1). If |Y|(k+1)(r+1+(k1)b1+(k1)b2+(k2)b3++bk), then we admit |NG(Y)|=r+|Y|+(k1)b1+(k1)b2+(k2)b3++bk(1+1k+1)|Y|1.

In the following, we may assume that |Y|>(k+1)(r+1+(k1)b1+(k1)b2+(k2)b3++bk). Note that YV(Kr)=. Thus, we have |Y||V(G)||V(Kr)|=(r+2)(k+1). Hence, we gain (k+1)(r+1+(k1)b1+(k1)b2+(k2)b3++bk)<|Y|(r+2)(k+1), namely, (k1)b1+(k1)b2+(k2)b3++bk<1, which implies b1=b2==bk=0, and so |Y|=(k+1)bk+1. If bk+1=r+2, then we derive Y=V((r+2)Kk+1), and so NG(Y)=V(G). Next, we assume that bk+1r+1. Clearly, |NG(Y)|=r+(k+1)bk+1=(k+2)bk+11+r+1bk+1(k+2)bk+11=(1+1k+1)|Y|1.

Consequently, we verify that NG(Y)=V(G) or |NG(Y)|(1+1k+1)|Y|1 for every YV(G).

Set X=V(Kr). Then i(GX)=0 and ωk+1(GX)=r+2. Thus, we deduce ki(GX)+ωk+1(GX)=r+2=|X|+2>|X|+1. Therefore, kG does not admit all [1,k]-factors by Theorem 1. Theorem 3 is verified. ◻

References:

  1. K. Huang, K. Lih, A note on m-near-factor-critical graphs, European Journal of Combinatorics 80(2019)273–276.

  2. K. Ota, T. Tokuda, A degree condition for the existence of regular factors in K1,n-free graphs, Journal of Graph Theory 22(1996)59–64.

  3. Y. Egawa, K. Kotani, 4-factors in 2-connected star-free graphs, Discrete Mathematics 309(2009)6265–6270.

  4. N. Haghparast, M. Kano, S. Maezawa, K. Ozeki, Connected odd factors of graphs, Australasian Journal of Combinatorics 73(1)(2019):200–206.

  5. S. Zhou, T. Zhang, Z. Xu, Subgraphs with orthogonal factorizations in graphs, Discrete Applied Mathematics 286(2020)29–34.

  6. S. Zhou, Y. Zhang, Z. Sun, The Aα-spectral radius for path-factors in graphs, Discrete Mathematics 347(5)(2024)113940.

  7. S. Zhou, Z. Sun, H. Liu, D-index and Q-index for spanning trees with leaf degree at most k in graphs, Discrete Mathematics 347(5)(2024)113927.

  8. S. Zhou, Some results about component factors in graphs, RAIRO–Operations Research 53(3)(2019)723–730.

  9. S. Zhou, Z. Sun, F. Yang, A result on P3-factor uniform graphs, Proceedings of the Romanian Academy, Series A: Mathematics, Physics, Technical Sciences, Information Science 23(1)(2022)3–8.

  10. S. Zhou, Remarks on path factors in graphs, RAIRO-Operations Research 54(6)(2020)1827–1834.

  11. S. Zhou, Y. Zhang, Sufficient conditions for fractional [a,b]-deleted graphs, Indian Journal of Pure and Applied Mathematics, DOI: 10.1007/s13226-024-00564-w

  12. G. Liu, L. Zhang, Tougness and the existence of fractional k-factors of graphs, Discrete Mathematics 308(2008)1741–1748.

  13. S. Zhou, Y. Xu, Z. Sun, Degree conditions for fractional (a,b,k)-critical covered graphs, Information Processing Letters 152(2019)105838.

  14. S. Zhou, Q. Pan, L. Xu, Isolated toughness for fractional (2,b,k)-critical covered graphs, Proceedings of the Romanian Academy, Series A: Mathematics, Physics, Technical Sciences, Information Science 24(1)(2023)11–18.

  15. S. Wang, W. Zhang, Independence number, minimum degree and path-factors in graphs, Proceedings of the Romanian Academy, Series A: Mathematics, Physics, Technical Sciences, Information Science 23(3)(2022)229–234.

  16. X. Lv, A degree condition for fractional (g,f,n)-critical covered graphs, AIMS Mathematics 5(2)(2020)872–878.

  17. T. Niessen, A characterization of graphs having all (g,f)-factors, Journal of Combinatorial Theory, Series B 72(1998)152–156.

  18. H. Lu, M. Kano, Q. Yu, Characterizations of graphs G having all [1,k]-factors in kG, Discrete Mathematics 342(2019)111580.

  19. Y. Yuan, R. Hao, Toughness condition for the existence of all fractional (a,b,k)-critical graphs, Discrete Mathematics 342(2019)2308–2314.

  20. S. Zhou, T. Zhang, Some existence theorems on all fractional (g,f)-factors with prescribed properties, Acta Mathematicae Applicatae Sinica, English Series 34(2)(2018)344–350.

  21. M. Kano, A sufficient condition for a graph to have [a,b]-factors, Graphs and Combinatorics 6(1990)245–251.

  22. H. Matsuda, A neighborhood condition for graphs to have [a,b]-factors, Discrete Mathematics 224(2000)289–292.

  23. S. Zhou, Z. Sun, Some existence theorems on path factors with given properties in graphs, Acta Mathematica Sinica, English Series 36(8)(2020)917–928.

  24. S. Zhou, Z. Sun, Binding number conditions for P2-factor and P3-factor uniform graphs, Discrete Mathematics 343(3)(2020)111715.

  25. W. Gao, W. Wang, A tight neighborhood union condition on fractional (g,f,n,m)-critical deleted graphs, Colloquium Mathematicum 149(2017)291–298.

  26. Y. Yuan, R. Hao, Neighborhood union conditions for fractional [a,b]-covered graphs, Bulletin of the Malaysian Mathematical Sciences Society 43(2020)157–167.

  27. A. Anderson, Sufficient conditions for matchings, Proceedings of the Edinburgh Mathematical Society 18(1973)129–136.

  28. W. Tutte, The factorization of linear graphs, Journal of the London Mathematical Society 22(1947)107–111.