[Retracted] On the Extremal Trees for Some Bond Incident Degree Indices with a Fixed Number of Segments
Abstract
The Gourava indices and hyper-Gourava indices are introduced by Kulli in 2017. These graph invariants are related to the degree of vertices of a graph G. Let be the class of all n−vertex chemical trees with r segments. In this paper, we characterize the graphs with the maximum value of the above indices in the class of chemical trees. In addition to different degree sequences, sharp upper bounds on those indices of trees with a fixed number of segments are determined, and the corresponding extremal trees are characterized.
1. Introduction
The extremal Zagreb indices have been studied for certain classes of trees in last some decades. For more details, see [1, 5–15].
Here, θi,j(T) denotes the number of edges of adjacent vertices of degree i and j in graph T. For GO1, GO2, HGO1, and HGO2, the values of ψ(i, j) are (i + j) + ij, (i + j) × ij, [(i + j) + ij]2, and [(i + j) × ij]2, respectively.
Furthermore, we will use this system of equations to find the number of edges of different types in upcoming theorems.
Lemma 1. For any chemical tree with 3 ≤ r ≤ n − 1, the following result holds.
Proof. A sequence (d1, d2, …, dn) of n positive integers, where n ≥ 2, is a degree sequence of a tree of order n if and only if and . In addition to this, if we have a chemical tree T, the order of the tree n is equal to the sum of the number of vertices of degrees i, i ∈ {1,2,3,4}, i.e.,
By solving (19) and (20), we get the values of unknown V1 and V4 as
For V3 = 0, V1 = (2(r + 2)/3), V2 = n − r − 1, and V4 = ((r − 1)/3) with r ≡ 1(mod3).
For V3 = 1, V1 = (2r/3) + 1, V2 = n − r − 1, and V4 = (r/3) − 1 with r ≡ 0(mod3).
For V3 = 2, V1 = (2(r + 1)/3), V2 = n − r − 1, and V4 = ((r − 5)/3) with r ≡ 2(mod3). ■
2. Main Result
Recall be a class of all n−vertex chemical trees of degree sequence . Let be a tree for 3 ≤ r ≤ n − 2, which maximizes the Gourava indices and hyper-Gourava indices. For proving Theorem 1, we determine the structure of 1Tmax from the following lemmas for the above-mentioned indices.
Lemma 2. Let be a maximal tree. If a pendent vertex u is adjacent to a vertex v of , then 1Tmax does not have a vertex w of with adjacent to both non-pendent neighbors.
Proof. Suppose, on the contrary, that u ~ v such that u and v are pendent and branching vertices, respectively. There exists a vertex w ∈ 1Tmax of degree two, such that w ~ w1 and w ~ w2, where w1, w2 are non-pendent vertices, i.e., . Let T∗ = 1Tmax − {uv, ww1, ww2} + {uw, vw, w1, w2}, then . We have
For each case GO1( 1Tmax) < GO1(T∗), GO2( 1Tmax) < GO2(T∗), HGO1( 1Tmax) < HGO1(T∗), and HGO2( 1Tmax) < HGO2(T∗), which is a contradiction to the maximality of 1Tmax.
Lemma 3. If a vertex x ∈ V( 1Tmax) of maximum degree Δ, 3 ≤ Δ ≤ n − 2, is the only branching vertex in 1Tmax. Then, 1Tmax is a starlike tree.
Proof. In 1Tmax, we notice that a vertex has maximum degree greater than two. This implies 1Tmax ≠ Pn. Suppose, on the contrary, that a branching vertex y ∈ V( 1Tmax) different from x with . Let P : xu1u2 … uly be a path in 1Tmax also x ~ xi, 1 ≤ i ≤ dx − 1, y ~ ul and y ~ yj, 1 ≤ j ≤ dy − 1. Here, all xi and yj are pendent vertices. Now, we obtained another tree such that T∗ = 1Tmax − {y ~ yj|1 ≤ j ≤ dy − 1} + {x ~ yj|1 ≤ j ≤ dy − 1}. Then, we have two cases:
-
Case 1: if x ~ y, we have
(23) -
contradiction to the choice of 1Tmax.
-
Case 2: if x≁y, we have
Again contradiction to the choice of 1Tmax. For each case GO1( 1Tmax) < GO1(T∗), GO2( 1Tmax) < GO2(T∗), HGO1( 1Tmax) < HGO1(T∗), and HGO2( 1Tmax) < HGO2(T∗). So, 1Tmax contains only one branching vertex. Hence, 1Tmax is a starlike tree.
Theorem 1. Let , where 3 ≤ r ≤ n − 2, then for a degree sequence , we have
The equality holds if and only if 1Tmax has degree sequence .
Proof.
-
Case 1: if V2 < r and n < 2r + 1, then by Lemmas 2 and 3, we have θ1,2 = θ2,r = n − r − 1, θ2,2 = 0, and θ1,r = −n + 2r + 1. Now, solving (12)–(15), we get
(26) -
Case 2: if V2 ≥ r and n ≥ 2r + 1, then by Lemmas 2 and 3, we have θ1,r = 0, θ2,2 = n − 2r − 1, and θ1,2 = θ2,r = r. Now, solving (12)–(15), we get
For proving Theorem 2, Theorem 3, and Theorem 4, we determine the structures of , , and from the following lemmas.
Lemma 4. For 3 ≤ r ≤ n − 1, let be a tree with maximal Gourava indices and hyper-Gourava indices which contain maximum two vertices of degree three.
Proof. Suppose, on the contrary, that Tmax contains three vertices of degree three. Let P : u1u2u3 … ul be the longest path in Tmax. Assume that vertices ui, uj, and ul−1 are branching vertices of degree three. Suppose two pendent vertices v and w = ul are adjacent to ul−1. Let T∗ = Tmax − {ul−1v, ul−1w} + {uiv, ujw}. For ui, uj, ul−1 ∈ Tmax, we have four cases.
-
Case 1: if ui≁uj and uj≁ul−1.
(28) -
a contradiction to Tmax.
-
Case 2: if ui≁uj and uj ~ ul−1.
(29) -
a contradiction with Tmax.
-
Case 3: if ui ~ uj and uj≁ul−1.
(30) -
a contradiction with Tmax.
-
Case 4: if ui ~ uj ~ ul−1.
(31)
For each case GO1(Tmax) < GO1(T∗), GO2(Tmax) < GO2(T∗), HGO1(Tmax) < HGO1(T∗), and HGO2(Tmax) < HGO2(T∗), which is a contradiction with the maximality of Tmax. So, Tmax contains at most two vertices of degree three.
Lemma 5. Let be a tree which contains interval path of length one only where 3 ≤ r ≤ n − 1.
Proof. Suppose, on the contrary, that Tmax has an internal path of length greater than or equal to two. Let P : u1u2u3 … uk be an internal path of length greater than or equal to two in Tmax, where u1 and uk be the branching vertices and . Let w be a pendent vertex adjacent to some ui ∈ V(Tmax) other then uj, 1 < j < k. Let T∗ = Tmax − {uiw, u1u2, uk−1uk} + {u1uk, u2w, uk−1ui}, then and
Lemma 6. For 3 ≤ r ≤ n − 1, let be a maximal tree containing a pendent vertex w ∈ V(Tmax) adjacent to some vertex u ∈ V(Tmax) of , then Tmax contains a pendent path of length at most two.
Proof. Suppose, on the contrary, that Tmax has a pendent path of length greater than or equal to three. Let P : u1u2u3 … uk be a pendent path of length greater than or equal to three in Tmax and a pendent vertex w ∈ V(Tmax) adjacent to some u ∈ V(Tmax) of . Let T∗ = Tmax − {uw, u1u2, u2u3} + {u2w, u2u, u1u3}, then and
Lemma 7. For 3 ≤ r ≤ n − 1, if be a maximal tree contains a pendent vertex u ∈ V(Tmax) adjacent to a vertex v ∈ V(Tmax) of , then Tmax does not contain any vertex of degree two adjacent with any vertex of degree three.
Proof. Suppose, on the contrary, that a vertex s ∈ V(Tmax) of degree two is adjacent with a branching vertex t ∈ V(Tmax) of degree three and a pendent vertex u ∈ V(Tmax) is adjacent to a branching vertex v ∈ V(Tmax) of degree four. Let Ns(Tmax) = {q, t}. Let T∗ = Tmax − {qs, st, vu} + {qt, vs, su}, then and
Lemma 8. For 3 ≤ r ≤ n − 1, if be a maximal tree containing a vertex u ∈ V(Tmax) of du = 3, then vertex u is adjacent to at most one branching vertex in Tmax.
Proof. Suppose, on the contrary, that x, y, and z are three branching vertices in V(Tmax). Let P : u1u2u3 … uj−1ujuj+1 … ul be the longest path in Tmax containing x, y, and z. By Lemma 4, there are maximum two vertices of degree three in P. Let uj−1 = x, uj = y, and uj+1 = z. If there are at most two vertices of degree three including y in P, then suppose a vertex ui ∈ V(Tmax) of degree three for some i, 1 ≤ i ≤ j − 1, and a vertex uk ∈ V(Tmax) of degree four for some k, j + 1 ≤ k ≤ l, is adjacent only one branching vertex. Now, bearing in mind , and dx = 4. Let T∗ = Tmax − {xy, yz, ukuk+1} + {xz, uky, yuk+1}, then and
Corollary 1 (see [23].)If Δ = 4 of , then the graph induced by the vertices of degree four of Tmax is a tree where 3 ≤ r ≤ n − 1.
Theorem 2. Let , where 3 ≤ r ≤ n − 1, then for a degree sequence , we have
The equality holds if and only if 2Tmax has a degree sequence .
Proof. For with r ≥ 5, we have V4 ≥ 1. By Corollary 1, we get θ4,4 = V4 − 1 = ((r − 1)/3) − 1 = ((r − 4)/3).⇒θ4,4 = (r − 4/3).
-
Case 1: when V1 > V2, we have n < ((5r + 7)/3). By Lemmas 5 and 6, we have θ2,2 = 0. Also, from (12)–(15), we get
(37)(38)(39) -
From (37), (38), and (39), we get θ1,4 = ((5r − 3n + 7)/3), θ1,2 = θ2,4 = n − r − 1. Hence,
(40) -
Case 2: when V1 ≤ V2, we have n ≥ ((5r + 7)/3). By Lemmas 5 and 6, we have θ1,4 = 0. Also, from (12)–(15), we get
From (41), (42), and (43), we get θ4,2 = ((2r + 4)/3), θ2,2 = ((3n − 5r − 7)/3). Hence,
Theorem 3. Let , where 3 ≤ r ≤ n − 1, then for a degree sequence , we have
The equality holds if and only if 3Tmax has a degree sequence .
Proof. For with V4 ≥ 1 and r ≥ 6. By Corollary 1, we have θ4,4 = V4 − 1 = (r/3) − 2. ⇒θ4,4 = (r/3) − 2, θ3,3 = 0 and by Lemmas 5 and 8, we have θ3,4 = 1.
-
Case 1: when V2 < 2V4 + 2, we have n < ((5r + 3)/3)⇒θ2,2 = 0 and θ1,4 ≠ 0, then by Lemma 7, θ2,3 = 0. From (12)–(15), we get
(46)(47)(48)(49) -
From (46)–(48) and (49), we get θ1,4 = (5r/3) − n, θ1,2 = θ2,4 = n − r − 1. Hence,
(50) -
Case 2: when V2 = 2V4 + 2, we have n = ((5r + 3)/3). It follows that θ2,2 = 0 = θ1,4. From (12)–(15), we get
(51)(52)(53)(54) -
From (51)–(53) and (54), we get θ1,2 = n − r − 1, θ1,3 = (5r/3) − n + 2, θ2,3 = ((3n − 5r)/3), and θ2,4 = (2r/3) − 1. Hence,
(55) -
Case 3: when V2 > 2V4 + 2, we have n > ((5r + 3)/3). It further implies θ1,4 = 0 and θ2,2 ≠ 0; then, by Lemmas 5 and 6, we have θ1,3 = 0. From (12)–(15), we get
From (56)–(58) and (59), we get θ2,2 = (−5r/3) + n − 2 and θ2,4 = (2r/3) − 1. Hence,
Theorem 4. Let , where 3 ≤ r ≤ n − 1, then for a degree sequence , we have
The equality holds if and only if 4Tmax has a degree sequence .
Proof. For with V4 ≥ 1 and r ≥ 8. By Corollary 1, it follows θ4,4 = V4 − 1 = ((r − 8)/3). By Lemmas 5 and 8, we get θ3,3 = 0 and θ3,4 = 2.
-
Case 1: when V2 ≤ 2V4, we have n ≤ ((5r − 7)/3). Hence, θ2,2 = 0, ⇒θ1,4 ≠ 0, and by Lemma 7, θ2,3 = 0. From (12)–(15), we get
(62)(63)(64)(65) -
From (62)–(64) and (65), we get θ1,2 = θ2,4 = n − r − 1, θ1,4 = ((5r − 7)/3) − n. Hence,
(66) -
Case 2: when 2V4 + 1 ≤ V2 ≤ 2V4 + 3, we have ((5r − 4)/3) ≤ n ≤ ((5r + 2)/3) and θ1,4 = 0 = θ2,2. From (12)–(15), we get
(67)(68)(69)(70) -
From (67)–(69) and (70), we get θ1,2 = n − r − 1, θ1,3 = ((5r + 5)/3) − n, θ2,3 = ((−5r + 7)/3) + n, θ2,4 = ((2r − 10)/3). Hence,
(71) -
Case 3: when V2 > 2V4 + 3, we have n > ((5r + 2)/3) which implies θ1,4 = 0 and θ2,2 ≠ 0. Hence, by Lemmas 5 and 6, θ1,3 = 0. From (12)–(15), we get
(72)(73)(74)(75)
From (72)–(74) and (75), we get θ2,2 = ((3n − 5r − 5)/3), θ2,4 = ((2r − 10)/3). Hence,
3. Conclusions
Nowadays, many researchers are introduced to investigating properties of different molecular descriptors through topological indices. In this way, we have investigated upper bonds on Gourava indices and hyper-Gourava indices for some classes of n−vertex chemical trees with a fixed number of segments and vertices of degree two.
At this point, we left the lower bonds on Gourava and hyper-Gourava indices for the above classes of n−vertex chemical trees with a fixed number of segments as an open problem.
Conflicts of Interest
The authors declare that there are no conflicts of interest.
Open Research
Data Availability
The whole data are included within this article. However, the reader may contact the corresponding author for more details on the data.