Research article

Proof of a conjecture on the ϵ-spectral radius of trees

  • Received: 06 October 2022 Revised: 13 November 2022 Accepted: 16 November 2022 Published: 05 December 2022
  • MSC : 05C50

  • The ϵ-spectral radius of a connected graph is the largest eigenvalue of its eccentricity matrix. In this paper, we identify the unique n-vertex tree with diameter 4 and matching number 5 that minimizes the ϵ-spectral radius, and thus resolve a conjecture proposed in [W. Wei, S. Li, L. Zhang, Characterizing the extremal graphs with respect to the eccentricity spectral radius, and beyond, Discrete Math. 345 (2022) 112686].

    Citation: Jianping Li, Leshi Qiu, Jianbin Zhang. Proof of a conjecture on the ϵ-spectral radius of trees[J]. AIMS Mathematics, 2023, 8(2): 4363-4371. doi: 10.3934/math.2023217

    Related Papers:

    [1] Huan Xu, Tao Yu, Fawaz E. Alsaadi, Madini Obad Alassafi, Guidong Yu, Jinde Cao . Some spectral sufficient conditions for a graph being pancyclic. AIMS Mathematics, 2020, 5(6): 5389-5401. doi: 10.3934/math.2020346
    [2] Wafaa Fakieh, Zakeiah Alkhamisi, Hanaa Alashwali . On the Aα-spectra of graphs and the relation between Aα- and Aα-spectra. AIMS Mathematics, 2024, 9(2): 4587-4603. doi: 10.3934/math.2024221
    [3] Maria Adam, Dimitra Aggeli, Aikaterini Aretaki . Some new bounds on the spectral radius of nonnegative matrices. AIMS Mathematics, 2020, 5(1): 701-716. doi: 10.3934/math.2020047
    [4] He Song, Longmin Wang, Kainan Xiang, Qingpei Zang . The continuity of biased random walk's spectral radius on free product graphs. AIMS Mathematics, 2024, 9(7): 19529-19545. doi: 10.3934/math.2024952
    [5] Jianwei Du, Xiaoling Sun . On symmetric division deg index of trees with given parameters. AIMS Mathematics, 2021, 6(6): 6528-6541. doi: 10.3934/math.2021384
    [6] Ximing Fang . The simplified modulus-based matrix splitting iteration method for the nonlinear complementarity problem. AIMS Mathematics, 2024, 9(4): 8594-8609. doi: 10.3934/math.2024416
    [7] Yufei Huang, Hechao Liu . Bounds of modified Sombor index, spectral radius and energy. AIMS Mathematics, 2021, 6(10): 11263-11274. doi: 10.3934/math.2021653
    [8] Xu Li, Rui-Feng Li . Shift-splitting iteration methods for a class of large sparse linear matrix equations. AIMS Mathematics, 2021, 6(4): 4105-4118. doi: 10.3934/math.2021243
    [9] Qin Zhong, Ling Li . Notes on the generalized Perron complements involving inverse N0-matrices. AIMS Mathematics, 2024, 9(8): 22130-22145. doi: 10.3934/math.20241076
    [10] Ali Raza, Mobeen Munir, Tasawar Abbas, Sayed M Eldin, Ilyas Khan . Spectrum of prism graph and relation with network related quantities. AIMS Mathematics, 2023, 8(2): 2634-2647. doi: 10.3934/math.2023137
  • The ϵ-spectral radius of a connected graph is the largest eigenvalue of its eccentricity matrix. In this paper, we identify the unique n-vertex tree with diameter 4 and matching number 5 that minimizes the ϵ-spectral radius, and thus resolve a conjecture proposed in [W. Wei, S. Li, L. Zhang, Characterizing the extremal graphs with respect to the eccentricity spectral radius, and beyond, Discrete Math. 345 (2022) 112686].



    Let G be a simple connected graph with vertex set V(G)={v1,v2,,vn}. The distance between vertices vi and vj in G, denoted by dG(vi,vj), is the length of the shortest path between vi and vj. The distance matrix of G, denoted by D(G), is the n×n matrix whose (i,j)-entry is equal to dG(vi,vj). Let A(G) be the adjacency matrix of G. The adjacency matrix and the distance matrix of a graph are well-studied matrix classes in the field of spectral graph theory. For more details about the study of classes of matrices associated with graphs, we refer to [2,3,4].

    For uV(G), eG(u) or e(u) denotes the eccentricity of u in G, which is equal to the largest distance from u to other vertices in G. The diameter of a graph G, denoted by diam(G), is defined to be the maximum of the eccentricity of all the vertices of G. A vertex v is said to be an eccentric vertex of the vertex u if dG(u,v)=e(u). A vertex uV(G) is said to be a diametrical vertex of G if e(u)=diam(G). If each vertex of G is a diametrical vertex and has a unique eccentric vertex, then G is called a diametrical graph.

    A matching in G is a set of edges without common vertices. The maximum matching is a matching with the maximum size in G. The matching number is the size of a maximum matching in G.

    The eccentricity matrix of the graph G, denoted by ϵ(G), is defined as [14]

    ϵ(G)u,v={dG(u,v),if dG(u,v)=min{eG(u),eG(v)},0,otherwise.

    It is also known as the DMAX-matrix defined by Randić in [10] as a tool for chemical graph theory. The eigenvalues of the eccentricity matrix of a graph G are called the eccentricity eigenvalues, or ϵ-eigenvalues, of G. Since ϵ(G) is symmetric, the ϵ-eigenvalues are all real, which are denoted by ξ1(G)ξ2(G)ξn(G). As usual, ξ1(G) is called the ϵ-spectral radius of G, denoted also by ρϵ(G).

    Recently, the ϵ-spectral radius has received much attention. Wang et al. [13] determined sharp lower and upper bounds for the ϵ-spectral radius of graphs and identified the corresponding extremal graphs. Wei et al. [16] determined the n-vertex trees with minimum ϵ-spectral radius. Furthermore, in [16], the authors identified all trees with given order and diameter having the minimum ϵ-spectral radius. He and Lu [5] identified the n-vertex trees with fixed odd diameter having the maximum ϵ-spectral radius. Wei, Li and Zhang [17] characterized the n-vertex trees having the second minimum ϵ-spectral radius and identified the n-vertex trees with small matching number having the minimum ϵ-spectral radii. For more advances on the ϵ-eigenvalues, one may be referred to [7,8,9,12,15].

    Let ˆTa,bn,5 be the n-vertex tree obtained from P5=v1v2v3v4v5v6 by attaching a pendent edges to v3, b pendent edges and a path with length 2 to v4, where a,b1 and a+b=n8, see Figure 1.

    Figure 1.  ˆTa,bn,5.

    Let Tn,d be the set of trees with order n and diameter d. Wei, Li and Zhang [17] proved in their Theorem 4.6 that, among the n-vertex trees with matching number 5 in d4Tn,d, the tree ˆT1,n9n,5 has the minimum ϵ-spectral radius if 10n16, and the trees ˆTγ,n8γn,5 or ˆTγ,n8γn,5 have the minimum ϵ-spectral radius if n17, where γ=1144(48n461206n17). They commented that "since it is abnormal for the eccentricity matrices of trees of diameter 4", their Theorem 4.6 does not solve the problem for d=4. So they proposed the following conjecture.

    Conjecture 1.1. Among the n-vertex trees with matching number 5 with n10, the tree ˆT1,n9n,5 has the minimum ϵ-spectral radius if 10n16, and the trees ˆTγ,n8γn,5 or ˆTγ,n8γn,5 have the minimum ϵ-spectral radius if n17, where γ=1144(48n461206n17).

    In this article, to avoid the earlier difficulty, we propose new graph transformations, characterize the certain structures of the extremal trees and determine the unique tree in Tn,4 with matching number 5 having the minimum ϵ-spectral radius, so give an affirmative answer to the conjecture.

    In this section, we introduce some preliminary results which will be used to in our proofs. The following Lemma is a well-known result in the theory of nonnegative matrices, see [4,6].

    Lemma 2.1. [4,6] Let A and B be two nonnegative irreducible matrices with same order. If Ai,jBi,j for each i,j, then ρ(A)ρ(B) with equality if and only if A=B, where ρ(A) and ρ(B) denote the spectral radius of A and B, respectively.

    Mahato et al. [8] obtained a lower bound for the ϵ-spectral radius of a graph with given diameter, which will be used in the proof of Lemma 3.1 and Lemma 3.2.

    Lemma 2.2. If G is a connected graph with diameter d2, then ξ1(G)d with equality if and only if G is a diametrical graph.

    Lemma 2.3. [14] The eccentricity matrix of a tree with order n2 is irreducible.

    Note that for a tree T with at least two vertices, ϵ(T) is an irreducible nonnegative matrix by Lemma 2.3. Thus by the Perron-Frobenius theorem, ρϵ(T) is simple, and there is a positive unit eigenvector of ϵ(T), which is called Perron vector of ϵ(T), corresponding to ρϵ(T).

    Lemma 2.4. [17] ρϵ(ˆTa,bn,5) is the largest root of f(λ)=0, where f(λ)=λ432aλ216bλ2141λ2+512ab+1312a+800b+2050.

    Lemma 2.5. [17] Among d4Tn,d with matching number 5, ˆT1,n9n,5 is the tree with minimum ϵ-spectral radius for 10n16, and ˆTγ,n8γn,5 or ˆTγ,n8γn,5 are trees with the minimum ϵ-spectral radius for n17, where γ=1144(48n461206n17).

    Let Sn,(a0,a1,,a) be the tree obtained by adding an edge between the center v0 of a star Sa0+1 and the center vi of the star Sai+1 for each i=1,2,,, see Figure 2. Any tree with diameter 4 is of the form Sn,(a0,a1,,a) with 2 and ai1 for each i=1,2,,.

    Figure 2.  Tree Sn,(a0,a1,a2,,a).

    Lemma 3.1. Let TSn,(a0,a1,,a) with 2 and 1a1a2a, where n=l+1+k=0ak. If there exists k such that 1k1 and ak2, then

    ρϵ(Sn,(a0,a1,,ak1,,a+1))<ρϵ(T).

    Proof. Let Ui=V(Sai+1){vi} for 0i. We partition V(T) into U0,{v0},{v1},...,{v},U1,...,U, then ϵ(T) can be written as

    (0a0,a00a0,10a0,10a0,10a0,13Ja0,a13Ja0,a23Ja0,al01,a000002J1,a12J1,a22J1,al01,a0000001,a13J1,a23J1,al01,a000003J1,a101,a23J1,al01,a000003J1,a13J1,a201,al3Ja1,a02Ja1,10a1,13Ja1,13Ja1,10a1,a14Ja1,a24Ja1,al3Ja2,a02Ja2,13Ja2,10a2,13Ja2,14Ja2,a10a2,a24Ja2,al3Jal,a02Jal,13Jal,13Jal,10al,14Jal,a14Jal,a20al,al).

    Let x be a Perron eigenvector corresponding to ρ:=ρϵ(T), whose coordinate with respect to vertex v is xv.

    Since ρxu0=3j=1ujUjxuj for any u0U0, we have xu0=xu0 for any u0,u0U0. Similarly, we have xuj=xuj for any xuj,xujUj with j=1,2,...,. Thus

    ρxu0=3(a1xu1+a2xu2++axu);ρxv0=2(a1xu1+a2xu2++axu);ρxv1=3(0+a2xu2++axu);ρxv2=3(a1xu1+0++axu);ρxv=3(a1xu1++a1xu1+0);ρxu1=2xv0+3(a0xu0+0+xv2++xv)+4(0+a2xu2++axu);ρxu2=2xv0+3(a0xu0+xv1+0++xv)+4(a1xu1+0++axu);ρxu=2xv0+3(a0xu0+xv1++xv1+0)+4(a1xu1++a1xu1+0).

    By eliminating xu0, xv0,xv from the above system, we obtain

    {ρ2xu1=(9a0+95)a1xu1+(9a0+914+4ρ)a2xu2++(9a0+914+4ρ)axu;ρ2xu2=(9a0+914+4ρ)a1xu1+(9a0+95)a2xu2++(9a0+914+4ρ)axu;ρ2xu=(9a0+914+4ρ)a1xu1+(9a0+914+4ρ)a2xu2++(9a0+95)axu.

    Let c=9a0+914. Since xui>0 for all 1i, ρ is the largest root of fak,a(λ)=0, where

    fak,a(λ)=|λ2(c+9)a1(c+4λ)a2(c+4λ)a(c+4λ)a1λ2(c+9)a2(c+4λ)a(c+4λ)a1(c+4λ)a2λ2(c+9)a|=|λ2(c+9)a1(c+4λ)a2(c+4λ)a(λ2+4λa19a1)λ2+4λa29a20(λ2+4λa19a1)0λ2+4λa9a|.=Πi=1(λ2+4aiλ9ai)|λ2(c+9)a1λ2+4a1λ9a1(c+4λ)a2λ2+4a2λ9a2(c+4λ)aλ2+4aλ9a110101|=(1i=1ai(c+4λ)λ2+4aiλ9ai)Πi=1(λ2+4aiλ9ai).

    By Lemma 2.2, ρdiam(T)=4. Then ρ2+4aiρ9ai>0 for all 1i. Hence, ρ is the largest root of gak,a(λ)=0, where

    gak,a(λ)=1i=1ai(c+4λ)λ2+4aiλ9ai.

    Since for λ>0,

    gak,a(λ)=i=14aiλ2+2aicλ+36a2i+4a2ic(λ2+4aiλ9ai)2>0,

    we have gak,a(λ) is monotonically increasing for λ>0. Thus it is sufficient to prove gak1,a+1(λ)>gak,a(λ) for λρ.

    By a direct calculation, one has

    Lemma 3.2. Let Tb=Sn,4(n8b,1,1,1,b) with 1bn9. Then ρϵ(Tb)6+36n191 with equality if and only if b=1.

    Proof. By the proof of Lemma 3.1, ρϵ(Tb) is the largest root of

    139(n8b)+9414+4λλ2+4λ9b(9(n8b)+9414+4λ)λ2+4bλ9b=0,

    i.e., fb(λ)=0, where

    fb(λ)=λ48λ3+(9b2+(209n)b27n+141)λ2+(144b2+(872144n)b)λ+(324(nb)1719)b.

    Note that

    f1(λ)=λ48λ3+(17036n)λ2+(1016144n)λ+324n2043=(λ2+4λ9)(λ212λ+22736n).

    By a direct calculation, ρϵ(T1)=6+36n191.

    Suppose now that 2bn9. Then it is sufficient to prove f1(λ)>fb(λ) for λρϵ(Tb). By a direct calculation, one has

    f1(λ)fb(λ)=(9b2+9nb20b9n+29)λ2+(144b2+144nb872b144n+1016)λ+324b2+1719b324nb+324n2043=(b1)((9n9b29)λ2+(144n144b1016)λ+9(36b36n+227))=(b1)g(λ),

    where g(λ)=(9n9b29)λ2+(144n144b1016)λ+9(36b36n+227) with 2bn9. By Lemma 2.2, we get ρϵ(Tb)diam(T)=4. Since 9n9b29=9(nb)2952>0 and 144n144b10162(9n9b29)=144(nb)10162(9n9b29)<0, we have g(λ) is montoncially increasing for λ>0. Then g(λ)g(ρϵ(Tb))g(4)=396(nb8)+683>0. Thus, f1(λ)>fb(λ) for λρϵ(Tb). We note that a similar treatment has been used in studying the Estrada indices, see [1,11].

    Lemma 3.3. Let T1=Sn,4(n9,1,1,1,1) and T1=Sn,5(0,1,1,1,1,n10). For n11, we have ρϵ(T1)>ρϵ(T1).

    Proof. Let NT1(v0){v1,v2,v3,v4}={w0,w1,,wn10}, NT1(vi){v0}={ui} for i=1,2,3,4. Then

    T1T1n10i=1v0wi+n10i=1w0wi.

    Let V1={u1,u2,u3,u4}, V2={v1,v2,v3,v4} and V3={w1,w2,,wn10}. As we pass from T1 to T1, we have

    ϵ(T1)u,vϵ(T1)u,v={2if u{v0},vV3,1if uV1,vV3,3if uV2,vV3,0otherwise.

    Thus, ϵ(T1)>ϵ(T1). Since T1 and T1 are n-vertex trees, by Lemma 2.3, ϵ(T1) and ϵ(T1) are irreducible. By Lemma 2.1, ρϵ(T1)>ρϵ(T1).

    Theorem 3.1. Let T be an n-vertex tree with diameter 4 and matching number 5. Then ρϵ(T)6+36n191 with equality if and only if TSn,4(n9,1,1,1,1).

    Proof. Let T be the tree having minimum ϵ-spectral radius among n-vertex trees with diameter 4 and matching number 5. Then TSn,(a0,a1,,a). Since the matching number of T is 5, =4 for a01 and =5 for a0=0. Thus, assume that TSn,4(a0,a1,a2,a3,a4) with a01 and 1a1a2a3a4 or TSn,5(0,b1,b2,b3,b4,b5) with 1b1b2b3b4b5. In the former case, we have by Lemma 3.1 that a1=a2=a3=1, and by Lemma 3.2 that a4=1, so a0=n9, that is, TSn,4(n9,1,1,1,1). In the latter case, we have by Lemma 3.1 that b1=b2=b3=b4=1, so TSn,5(0,1,1,1,1,n10). By Lemma 3.3, TSn,4(n9,1,1,1,1).

    Theorem 3.2. Among the n-vertex trees with matching number 5 with n10, ˆT1,n9n,5 is the tree with minimum ϵ-spectral radius for 10n16, and ˆTγ,n8γn,5 or ˆTγ,n8γn,5 are trees with the minimum ϵ-spectral radius for n17, where γ=1144(48n461206n17).

    Proof. Let T be the tree minimizing the ϵ-spectral radius in Tn,d with matching number 5. If d=4, then TSn,4(n9,1,1,1,1) by Theorem 3.1. If d5, then TˆTa,bn,5 by Lemma 2.5. Let ρϵ be the spectral radius of ˆTa,bn,5. Then by Lemma 2.4, ρϵ is the largest root of f(λ)=0, where

    f(λ)=λ432aλ216bλ2141λ2+512ab+1312a+800b+2050.

    Note that a+b=n8. Thus

    f(λ)=λ4(16n+16a+13)λ2+512a(n7a)+800n4350.

    Since n10, then 512a(n7a)+800n4350>0. Thus

    ρϵ(ˆTa,bn,5)=16n+16a+13+(16n+16a+13)24(512a(n7a)+800n435)2<16n+16a+1316n+16(n8)+13=32n115<36n155+1236n191=6+36n191=ρϵ(Sn,4(n9,1,1,1,1)).

    Combining with Lemma 2.5, the result follows.

    In this contribution, we characterize the unique tree among all trees with diameter 4 and matching number 5 that minimizes the ϵ-spectral radius. This confirms a conjecture in [17]. By combining the results from [17], the trees with minimum ϵ-spectral radius among all trees with matching number r5 have been characterized completely. For further study, one may try to determine the trees with the minimum ϵ-spectral radius among all trees with matching number r6 or even among trees with given fraction matching number.

    The authors would like to thank anonymous referees for helpful comments and suggestions which improved the original version of the paper. This work was supported by Natural Science Foundation of Guangdong Province (No.2021A1515010028).

    The authors declare that they have no competing interests.



    [1] A. Alhevaz, M. Baghipur, Y. Shang, Merging the spectral theories of distance Estrada and distance signless Laplacian Estrada indices of graphs, Mathematics, 7 (2019), 995. https://doi.org/10.3390/math7100995 doi: 10.3390/math7100995
    [2] M. Aouchiche, P. Hansen, Distance spectra of graphs: a survey, Linear Algebra Appl., 458 (2014), 301–386. https://doi.org/10.1016/j.laa.2014.06.010 doi: 10.1016/j.laa.2014.06.010
    [3] R. B. Bapat, Graphs and matrices, 2 Eds., London: Springer, 2014. https://doi.org/10.1007/978-1-4471-6569-9
    [4] A. E. Brouwer, W. H. Haemers, Spectra of graphs, New York: Springer, 2012. https://doi.org/10.1007/978-1-4614-1939-6
    [5] X. He, L. Lu, On the largest and least eigenvalues of eccentricity matrix of trees, Discrete Math., 345 (2022), 112662. https://doi.org/10.1016/j.disc.2021.112662 doi: 10.1016/j.disc.2021.112662
    [6] R. A. Horn, C. R. Johnson, Matrix analysis, 2 Eds., Cambridge: Cambridge University Press, 2013. https://doi.org/10.1017/CBO9781139020411
    [7] I. Mahato, R. Gurusamy, M. R. Kannan, S. Arockiaraj, Spectra of eccentricity matrices of graphs, Discrete Appl. Math., 285 (2020), 252–260. https://doi.org/10.1016/j.dam.2020.05.029 doi: 10.1016/j.dam.2020.05.029
    [8] I. Mahato, R. Gurusamy, M. R. Kannan, S. Arockiaraj, On the spectral radius and the energy of eccentricity matrix of a graph, Linear Multilinear Algebra, in press. https://doi.org/10.1080/03081087.2021.2015274
    [9] A. K. Patel, L. Selvaganesh, S. K. Pandey, Energy and inertia of the eccentricity matrix of coales-cence of graphs, Discrete Math., 344 (2021), 112591. https://doi.org/10.1016/j.disc.2021.112591 doi: 10.1016/j.disc.2021.112591
    [10] M. Randić, DMAX-matrix of dominant distances in a graph, MATCH Commun. Math. Comput. Chem., 70 (2013), 221–238.
    [11] Y. Shang, Bounds of distance Estrada index of graphs, Ars Comb., 128 (2016), 287–294.
    [12] J. Wang, X. Lei, W. Wei, X. Luo, S. Li, On the eccentricity matrix of graphs and its applications to the boiling point of hydrocarbons, Chemometr. Intell. Lab. Sys., 207 (2020), 104173. https://doi.org/10.1016/j.chemolab.2020.104173 doi: 10.1016/j.chemolab.2020.104173
    [13] J. Wang, M. Lu, L. Lu, F. Belardo, Spectral properties of the eccentricity matrix of graphs, Discrete Appl. Math., 279 (2020), 168–177. https://doi.org/10.1016/j.dam.2019.10.015 doi: 10.1016/j.dam.2019.10.015
    [14] J. Wang, M. Lu, F. Belardo, M. Randić, The anti-adjacency matrix of a graph: eccentricity matrix, Discrete Appl. Math., 251 (2018), 299–309. https://doi.org/10.1016/j.dam.2018.05.062 doi: 10.1016/j.dam.2018.05.062
    [15] J. Wang, L. Lu, M. Randić, G. Z. Li, Graph energy based on the eccentricity matrix, Discrete Math., 342 (2019), 2636–2646. https://doi.org/10.1016/j.disc.2019.05.033 doi: 10.1016/j.disc.2019.05.033
    [16] W. Wei, X. He, S. Li, Solutions for two conjectures on the eigenvalues of the eccentricity matrix, and beyond, Discrete Math., 343 (2020), 111925. https://doi.org/10.1016/j.disc.2020.111925 doi: 10.1016/j.disc.2020.111925
    [17] W. Wei, S. Li, L. Zhang, Characterizing the extremal graphs with respect to the eccentricity spectral radius, and beyond, Discrete Math., 345 (2022), 112686. https://doi.org/10.1016/j.disc.2021.112686 doi: 10.1016/j.disc.2021.112686
  • This article has been cited by:

    1. Jianping Li, Leshi Qiu, Jianbin Zhang, On the least eccentricity eigenvalue of graphs, 2023, 336, 0166218X, 47, 10.1016/j.dam.2023.03.029
    2. Leshi Qiu, Jianping Li, Jianbin Zhang, On the eccentricity energy and eccentricity spectral radius of graphs with odd diameter, 2023, 57, 0399-0559, 3141, 10.1051/ro/2023168
    3. Lu Huang, Aimei Yu, Rong-Xia Hao, The ɛ-spectral radius of trees with perfect matchings, 2025, 363, 0166218X, 110, 10.1016/j.dam.2024.11.028
  • Reader Comments
  • © 2023 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(1393) PDF downloads(52) Cited by(3)

Figures and Tables

Figures(2)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog