Next Article in Journal
A Note on Shape Vector Fields on Hypersurfaces
Previous Article in Journal
Relationship between Body Composition Asymmetry and Specific Performance in Taekwondo Athletes: A Cross-Sectional Study
Previous Article in Special Issue
Controller Design and Stability Analysis for a Class of Leader-Type Stochastic Nonlinear Systems
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Tricyclic Graph with Minimum Randić Index

1
School of Computer Science and Technology, Hengyang Normal University, Hengyang 421002, China
2
School of Information and Communication, Shenzhen Institute of Information Technology, Shenzhen 518172, China
*
Author to whom correspondence should be addressed.
These authors contributed equally to this work.
Submission received: 12 September 2023 / Revised: 12 October 2023 / Accepted: 18 October 2023 / Published: 20 November 2023
(This article belongs to the Special Issue Advances in Graph Theory and Symmetry/Asymmetry)

Abstract

:
The Randić index of a graph G is the sum of ( d G ( u ) d G ( v ) ) 1 2 over all edges u v of G, where d G ( u ) denotes the degree of vertex u in G. In this paper, we investigate a few graph transformations that decrease the Randić index of a graph. By applying those transformations, we determine the minimum Randić index on tricyclic graphs and characterize the corresponding extremal graphs.

1. Introduction

In this paper, we are concerned with undirected simple connected graphs, unless otherwise specified. Let G = ( V , E ) be such a graph with n vertices and m edges, which is addressed as an ( n , m ) -graph in what follows. A graph is cyclic if it contains at least one cycle, otherwise it is acyclic. More specifically, an ( n , n + k ) -graph is a tree, unicyclic, bicyclic, tricyclic, or tetracyclic, if k = 1 , 0 , 1 , 2 , 3 , respectively. Denote N G ( v ) as the neighbors of vertex v in G, and d G ( v ) = | N G ( v ) | the degree of v. As usual, let Δ ( G ) = max { d G ( v ) | v V } and δ ( G ) = min { d G ( v ) | v V } . A pendant vertex (or leaf) is a vertex of degree one. The star S n is a tree with n 1 pendant vertices, and the path P n is a tree with two pendant vertices. For vertex v, we call N G p ( v ) = { u N G ( v ) | d G ( u ) = 1 } pendant neighbors of v, and N G n ( v ) = { u N G ( v ) | d G ( u ) > 1 } non-pendant neighbors. Furthermore, let d G p ( v ) = | N G p ( v ) | and d G n ( v ) = | N G n ( v ) | . If E ˜ is an edge set, then G E ˜ denotes the graph formed from G by deleting edges in E ˜ , while G + E ˜ means the graph from G by adding edges in E ˜ . If graphs G and H are isomorphic, we can write it as G H .
The Randić index (or R index for short) of G is defined as
R ( G ) = u v E 1 d G ( u ) d G ( v ) .
This structural descriptor was proposed as a branching index [1] by Milan Randić in 1975. Since then, the mathematical properties of the R index have been studied extensively. For a comprehensive survey, see [2,3,4,5]. Remarkably, connections between the R index and other graph descriptors have been discovered, such as the graph energy [6], the normalized Laplacian matrix [7], and the information content of the underlying molecular graph [8]. Furthermore, the general Randić index [9,10,11] has attracted much attention in recent years.
From the view of extremal graph theory, Bollobás and Erdös [12] first proved that S n is the unique graph with the minimum R index for all n-vertex graphs and n-vertex trees. Trees with the second to the fourth minimum R indices have been determined by Zhao and Li in [13]. Caporossi et al. [14] and P. Yu [15] showed that P n attains the maximum R index in n-vertex trees. In [14], trees and unicyclic graphs with the first and second maximum R indices and bicyclic graphs with the maximum R index are also considered. The unique unicyclic graph with the minimum R index has been determined by Gao and Lu in [16]. Du and Zhou [17] investigated more minimum and maximum R indices of trees, unicyclic, and bicyclic graphs. For instance, the second to the fifth maximum and the second minimum R indices of bicyclic graphs. Furthermore, the tricyclic and tetracyclic graphs with the maximum and the second maximum R indices have been determined in [18,19], while the counterparts with the minimum Randić indices are still unidentified for now.
In this paper, we investigate a few graph transformations that decrease the R index of graphs. With the aid of these transformations, we derive a tricyclic graph with the minimum Randić index as follows.
Theorem 1.
If G is a tricyclic graph of order n 4 , then R ( G ) n 4 + 3 n 1 + 1 . And, the equality holds if and only if G T R n , where T R n is obtained by attaching n 4 pendant vertices to one vertex of a complete graph K 4 shown as Figure 1.

2. Preliminaries

The following lemmas are required mainly for characterizing transformations in the next section.
Lemma 1.
Suppose f ( x ) is twice differentiable, and f ( x ) > 0 . Let S = f ( a ) f ( b ) f ( c ) + f ( d ) with a + d = b + c , a b d and a c d . Then S 0 with equality holding if and only if a = b or a = c .
Proof. 
If a = b , then d = c from a + d = b + c , thereby S = 0 . Similarly S = 0 if a = c . Without loss of generality, we may assume that a < b c < d . Thus, we obtain S = f ( ξ 1 ) ( a b ) + f ( ξ 2 ) ( d c ) = ( b a ) ( f ( ξ 2 ) f ( ξ 1 ) ) = ( b a ) ( ξ 2 ξ 1 ) f ( η ) > 0 , where a < ξ 1 < b c < ξ 2 < d and ξ 1 < η < ξ 2 . Therefore, the lemma holds clearly. □
Lemma 2.
If x 2 , y 2 , then f ( x , y ) = y 2 + 2 x y y 1 + 1 x y + x 1 > 0 .
Proof. 
Let g ( x , y ) = f ( x , y ) y 2 + 2 x y + y 1 + 1 x y + x 1 y ( x + y 1 ) = ( x 3 + 2 x ) y 2 + ( 4 x 4 x + 7 + 3 x 10 x ) y + 4 x 4 x + 8 x 8 x with x 2 and y 2 . Note that it suffices to show that g ( x , y ) > 0 .
Observe that ( x 3 + 2 x ) = 1 1 x 3 2 > 0 , so x 3 + 2 x 2 3 + 2 2 > 0 . Hence, g ( x , y ) y = 2 y ( x 3 + 2 x ) + ( 4 x 4 x + 7 + 3 x 10 x ) 4 ( x 3 + 2 x ) + ( 4 x 4 x + 7 + 3 x 10 x ) = 4 x + 3 x 2 x 5 4 2 + 3 2 2 2 5 > 0 since ( 4 x + 3 x 2 x 5 ) = 2 x + 1 x 3 2 3 x 2 > 0 . As a consequence, g ( x , y ) g ( x , 2 ) = 2 + 2 x 4 x 2 + 2 2 4 2 > 0 as ( 2 + 2 x 4 x ) = 2 x 3 2 2 x 2 > 0 . Hence, the lemma holds easily. □
Lemma 3.
If x 3 ,   y 3 and k 2 5 , and let
f ( x , y ) = x 3 x + y 3 y + 1 x y x + y 5 x + y 1 + k x k x + y 1 .
Then, f ( x , y ) > 0 and f ( y , x ) > 0 .
Proof. 
Since f ( x , y ) x 3 x + y 3 y + 1 x y x + y 5 x + y 1 + 2 5 1 x 1 x + y 1 = h ( x , y ) , then we only have to show h ( x , y ) > 0 . Consider the gradient
h ( x , y ) y = 1 2 1 y 1 x + y 1 + 3 1 x 2 y 3 2 4 2 5 5 2 x + y 1 3 2 x 1 4 η 3 2 + 3 1 3 2 y 3 2 4 2 5 5 2 x + y 1 3 2 ( 3 1 2 + 3 1 3 ) ( 4 2 5 ) 2 x + y 1 3 2 = 2 5 1 3 2 x + y 1 3 2 > 0 ,
where y < η < x + y 1 , hence h ( x , y ) h ( x , 3 ) .
Let g ( x ) = x ( x + 2 ) x 3 + 1 3 + 2 5 x + x 2 + 2 5 x + 2 h ( x , 3 ) = 2 3 3 x 2 + 4 15 + 12 5 100 10 3 15 x
+ 304 + 8 15 72 5 60 3 15 . Since the axis of symmetry of g ( x ) is 4 15 + 12 5 100 10 3 15 / 2 3 3 / 2 2.16493 < 3 , hence g ( x ) g ( 3 ) = 12 5 / 5 + 4 / 15 + 4 15 / 3 0.06408 > 0 , implying h ( x , 3 ) > 0 .
Analogously, it can be shown that f ( y , x ) > 0 holds. Therefore, the lemma follows immediately. □
Lemma 4.
If x 3 , y 4 , then
f ( x , y ) = x 3 x + 1 x y x 2 x + y 1 + 1 2 1 x 1 x + y 1 0 .
Proof. 
Consider the gradient f ( x , y ) y = x 2 + 2 2 2 ( x + y 1 ) 3 2 1 2 x y 3 2 . Since
f ( x , y ) y x 2 + 2 2 2 ( x + y 1 ) 3 2 + 1 2 x y 3 2 4 x + y 1 3 x = x x 2 + 2 2 2 x + y 1 3 y 3 x x 2 + 2 2 2 x + 3 3 4 3 = 63 ( x 3 ) + ( 60 2 76 ) 64 x 2 + ( 252 64 2 2 ) x + 9 ( x 3 ) 64 > 0 ,
where the last inequality holds by x 3 , and consequently f ( x , y ) y > 0 .
Therefore, it is sufficient to show f ( x , 4 ) 0 . Let q ( x ) = f ( x , 4 ) x ( x + 3 ) ( x 3 + 1 2 + 2 2 x + x 2 + 2 2 x + 3 ) = 2 x 2 + 10 2 51 4 x + 81 30 2 4 . Since the axis of symmetry of q ( x ) is ( 10 2 51 4 ) / 4 2.3 < 3 , then q ( x ) q ( 3 ) = 0 . Note that q ( x ) and f ( x , 4 ) have the same sign, so the proof is complete. □
Lemma 5.
Let x 1 , x 2 , y be integers with y 3 , and let f ( x 1 , x 2 , y ) = x 1 + 1 2 + k x 1 + x 2 + y 2 + 1 2 y + 1 y ( x 1 + x 2 ) x 1 + y 2 + 2 + k x 1 + x 2 + y 2 1 2 , then f ( x 1 , x 2 , y ) > 0 , if either of the following is satisfied: (1) x 1 1 ,   2 x 2 10 ,   k = 0 ; (2) x 1 = 0 ,   2 x 2 6 ,   k = 1 3 .
Proof. 
(1) Let h ( t ) = t 2 + 2 x 2 2 t + x 2 3 2 with t 1 . Note that h = t 2 2 x 2 + 3 2 2 2 t + x 2 5 2 < 0 . Then, we have f ( x 1 , x 2 , y ) x 1 = x 1 2 + 2 x 2 + ( 1 2 1 y ) 2 x 1 + x 2 3 2 ( x 1 + y 2 ) 2 + 2 x 2 2 ( x 1 + y 2 ) + x 2 3 2 > h ( x 1 ) h ( x 1 + y 2 ) > 0 , hence it suffices to prove f ( 1 , x 2 , y ) > 0 .
Let A = y 2 + 1 2 y + 1 y ( 1 + x 2 ) + 1 + 1 2 1 + x 2 1 2 , and B = y 1 + 2 x 2 + y 1 . First note that 1 + 1 2 1 + x 2 1 2 1 + 1 2 1 + 10 1 2 > 0 , which means A > 0 . Let g ( x 2 , y ) = f ( 1 , x 2 , y ) ( A + B ) y ( x 2 + 1 ) ( x 2 + y 1 ) = ( A 2 B 2 ) y ( x 2 + 1 ) ( x 2 + y 1 ) = a 1 y 5 2 + a 2 y 2 + a 3 y 3 2 + a 4 y + a 5 y 1 2 + a 6 . It is easy to see f ( 1 , x 2 , y ) > 0 if g ( x 2 , y ) > 0 .
Now, let b 1 = a 1 , and b i = 3 b i 1 + a i for i = 2 , 3 , 4 , 5 , 6 . With assistance of computer, one can get all values of b i for 2 x 2 10 as shown in Table 1, and conclude that b i > 0 for each i.
We find that g ( x 2 , y ) = b 1 y 5 2 + a 2 y 2 + a 3 y 3 2 + a 4 y + a 5 y 1 2 + a 6 3 b 1 y 2 + a 2 y 2 + a 3 y 3 2 + a 4 y + a 5 y 1 2 + a 6 = b 2 y 2 + a 3 y 3 2 + a 4 y + a 5 y 1 2 + a 6 . By repeating this process, we arrive at g ( x 2 , y ) b 6 > 0 . Therefore, f ( x 1 , x 2 , y ) > 0 holds.
(2) Since the result in this case can be proved by a similar argument as (1), so we omit the details and only give the values of b i in Table 2.
The lemma therefore follows easily. □
Lemma 6.
If y z 2 , then f ( y , z ) = ( 1 y z 1 y + 1 z 1 ) ( 1 z + 1 y 1 z 1 1 y + 1 ) > 0 .
Proof. 
Observe that f ( y , z ) = ( 1 1 z ) ( 1 1 y ) ( 1 1 z 1 ) ( 1 1 y + 1 ) . It is obvious that f ( y , 2 ) = 1 1 2 1 1 y > 0 .
Now, we consider the case y z > 2 . Let h ( y , z ) = ln ( 1 1 z ) ( 1 1 y ) ln ( 1 1 z 1 ) ( 1 1 y + 1 ) with y z > 2 . Note that f ( y , z ) has the same sign with h ( y , z ) since ln ( x ) is increasing for x > 0 . Let g ( t ) = ln ( 1 1 t ) for t > 1 , and g ( t ) = 3 t 2 4 t 2 t 1 2 > 0 . Hence, we obtain h ( y , z ) = g ( z 1 ) g ( z ) g ( y ) + g ( y + 1 ) > 0 by Lemma 1. Therefore, the lemma holds easily. □
Lemma 7.
Let f ( y , z ) = y + k y + z + k z + k y z with y z 2 , then f ( y , z ) > f ( y + 1 , z 1 ) if it meets one of the following conditions: (1) k 0 and k = 0 ; (2) k 1 , k = 1 .
Proof. 
(1) Let g ( t ) = t + k t with t 1 , and note that g ( t ) = t 3 k 4 t 5 2 > 0 . Then we have f ( y , z ) f ( y + 1 , z 1 ) = g ( z 1 ) g ( z ) g ( y ) + g ( y + 1 ) > 0 by Lemma 1.
(2) Let h ( t ) = t + k + 1 t with t 1 , and note that h ( t ) = t 3 k 3 4 t 5 2 > 0 . By Lemma 6, we have f ( y , z ) f ( y + 1 , z 1 ) z + k z + y + k y z 1 + k z 1 y + 1 + k y + 1 + 1 z + 1 y 1 z 1 1 y + 1 = h ( z 1 ) h ( z ) h ( y ) + h ( y + 1 ) > 0 , where the last inequality follows from Lemma 1. □
Lemma 8.
If x 4 , y 4 , then f ( x , y ) = y 3 + 3 3 + 2 2 + 1 x y + x 4 + 3 3 + 2 x x + y 7 + 2 3 3 + 2 x + y 3 ( 1 3 + 1 6 ) > 0 .
Proof. 
Since
f ( x , y ) x = 4 2 3 3 1 y 2 x 3 2 4 2 2 3 3 2 ( x + y 3 ) 3 2 + ( 1 2 x 1 2 x + y 3 ) > 4 2 3 3 1 4 2 x 3 2 4 2 2 3 3 2 ( x + y 3 ) 3 2 > 3 3 1 2 2 ( x + y 3 ) 3 2 > 0 ,
hence f ( x , y ) f ( 4 , y ) . And, moreover,
f ( 4 , y ) y = 5 2 3 3 2 2 2 y 3 2 4 2 3 3 2 2 y + 1 3 2 + ( 1 2 y 1 2 y + 1 ) = 5 2 3 3 2 2 2 y 3 2 + 1 4 η 3 2 4 2 3 3 2 2 y + 1 3 2 > 1 3 + 1 2 1 2 y + 1 3 2 > 0 ,
where y < η < y + 1 . Therefore, f ( x , y ) f ( 4 , 4 ) 0.05036 > 0 . □

3. Transformations Decreasing Randić Index

To find tricyclic graphs with a small Randić index, we provide some transformations that decrease the Randić index of graphs. It is worth noting that all transformations defined here preserve the number of vertices and edges of a graph. For simplicity, we will not repeat this property in the sequel.
Definition 1
(Transformation I). Suppose G is a graph with two adjacent vertices u and v such that N G ( v ) N G ( u ) = . Let graph G = G { v w | w N G ( v ) u } + { u w | w N G ( v ) u } , and we write the transformation as G = Γ ( G , u , v ) .
Theorem 2.
Suppose G is a graph with two adjacent vertices u and v such that N G ( v ) N G ( u ) = . Let G = Γ ( G , u , v ) . Then R ( G ) > R ( G ) if G meets one of the following conditions:
(1) 
v has only one non-pendant neighbor u and d G ( v ) 2 ;
(2) 
v has two non-pendant neighbors u and w with d G ( u ) d G ( w ) ;
(3) 
v has three non-pendant neighbors u , v 1 , v 2 and u has three non-pendant neighbors v , u 1 , u 2 such that 1 d G ( u 1 ) + 1 d G ( u 2 ) + 1 d G ( v 2 ) + 1 d G ( v 2 ) 2 5 ;
(4) 
d G ( v ) 4 and u has three non-pendant neighbors v , u 1 , u 2 with 1 d G ( u 1 ) + 1 d G ( u 2 ) 1 2 .
Proof. 
Let E ˜ be the edge set of G that are not incident with u or v, and R ˜ = p q E ˜ 1 d G ( p ) d G ( q ) . And, let x = d G ( u ) , y = d G ( v ) , Δ = R ( G ) R ( G ) . Then,
R ( G ) = R ˜ + z N G ( u ) v 1 x d G ( z ) + z N G ( v ) u 1 y d G ( z ) + 1 x y R ( G ) = R ˜ + z N G ( u ) v 1 d G ( z ) + z N G ( v ) u 1 d G ( z ) + 1 1 x + y 1 .
(1) Note that x , y 2 and v has y 1 pendant neighbors. Then,
Δ = y 1 y + 1 x y y x + y 1 + z N G ( u ) v 1 d G ( z ) ( 1 x 1 x + y 1 ) y 1 y + 1 x y y x + y 1 = y 2 + 2 x y y 1 + 1 x x + y 1 + 1 1 x 1 y 1 x + y 1 > 0 ,
where the last inequality holds by Lemma 2 and 1 y > 1 x + y 1 .
(2) Notice that x 2 , y 2 , d G ( w ) x , and v has y 2 pendant neighbors in this case. Then,
Δ y 2 y + 1 x y y 1 y + x 1 + 1 d G ( w ) 1 y 1 y + x 1 + z N G ( u ) v 1 d G ( z ) ( 1 x 1 x + y 1 ) y 2 y + 1 x y y 1 y + x 1 + 1 x 1 y 1 y + x 1 = y 2 + 2 x y y 1 + 1 x y + x 1 > 0 ,
where the last inequality holds by Lemma 2.
(3) Let k 1 = 1 d G ( u 1 ) + 1 d G ( u 2 ) , k 2 = 1 d G ( v 1 ) + 1 d G ( v 2 ) , and p ( x , y ) = x 3 x + y 3 y + 1 x y x + y 5 x + y 1 . Note that x , y 3 and k 1 + k 2 2 5 from the condition. Then,
Δ = p ( x , y ) + k 1 x k 1 x + y 1 + k 2 y k 2 x + y 1 = k 1 k 1 + k 2 p ( x , y ) + k 1 + k 2 x k 1 + k 2 x + y 1 + k 2 k 1 + k 2 p ( x , y ) + k 1 + k 2 y k 1 + k 2 x + y 1 > 0 ,
where the last inequality holds by Lemma 3.
(4) Obviously, x 3 , y 4 and N G ( v ) u . Then
Δ = x 3 x + 1 x y x 2 x + y 1 + z N G ( v ) u 1 d G ( z ) 1 y 1 x + y 1 + 1 d G ( u 1 ) + 1 d G ( u 2 ) 1 x 1 x + y 1 > x 3 x + 1 x y x 2 x + y 1 + 1 2 1 x 1 x + y 1 0 ,
where the last inequality holds by Lemma 4. □
Definition 2
(Transformation II). Suppose G is a graph with two vertices u and v. Let G arise from G by moving k pendant neighbors from u to v, and the transformation can be written as G = Θ ( G , u , v , k ) . When all pendant neighbors of u are moved to v, this case will be written as G = Θ ( G , u , v , ) .
Theorem 3.
Suppose G is a graph, and there is a cycle C = v 0 v 1 v 2 v 0 in G with d G p ( v 1 ) 1 and d G n ( v 1 ) = 2 . Let G = Θ ( G , v 1 , v 0 , ) . Then, R ( G ) > R ( G ) if either of the following is satisfied:
(1) 
d G p ( v 0 ) 1 , 2 d G n ( v 0 ) 10 ;
(2) 
2 d G n ( v 0 ) 6 and there is a vertex u N G ( v 0 ) { v 1 , v 2 } with d G ( u ) 3 .
Proof. 
Let x 1 = d G p ( v 0 ) , x 2 = d G n ( v 0 ) and y = d G ( v 1 ) . Let f ( x ) = 1 x , then f ( x ) = 3 4 x 5 2 > 0 . And, note that 2 + ( x 1 + x 2 + y 2 ) = y + ( x 1 + x 2 ) , 2 x 1 + x 2 ( x 1 + x 2 + y 2 ) and 2 y ( x 1 + x 2 + y 2 ) , then we get 1 2 1 x 1 + x 2 1 y + 1 x 1 + x 2 + y 2 0 by Lemma 1. Then,
R ( G ) R ( G ) = x 1 x 1 + x 2 + y 2 y + 1 y x 1 + x 2 x 1 + y 2 + 1 2 x 1 + x 2 + y 2 + z N G n ( v 0 ) { v 1 , v 2 } 1 d G ( z ) 1 x 1 + x 2 1 x 1 + x 2 + y 2 1 d G ( v 2 ) 1 2 1 x 1 + x 2 1 y + 1 x 1 + x 2 + y 2 x 1 x 1 + x 2 + y 2 y + 1 y x 1 + x 2 x 1 + y 2 + 1 2 x 1 + x 2 + y 2 + z N G n ( v 0 ) { v 1 , v 2 } 1 d G ( z ) 1 x 1 + x 2 1 x 1 + x 2 + y 2 1 2 1 2 1 x 1 + x 2 1 y + 1 x 1 + x 2 + y 2 ,
where the last inequality follows by d G ( v 2 ) 2 .
(1) Note that x 1 1 , 2 x 2 10 and y 3 . Then,
R ( G ) R ( G ) x 1 x 1 + x 2 + y 2 y + 1 y x 1 + x 2 x 1 + y 2 + 1 2 x 1 + x 2 + y 2 1 2 1 2 1 x 1 + x 2 1 y + 1 x 1 + x 2 + y 2 = x 1 + 1 2 x 1 + x 2 + y 2 + 1 2 y + 1 y x 1 + x 2 x 1 + y 2 + 2 x 1 + x 2 + y 2 1 2 > 0 ,
where the last inequality holds by (1) of Lemma 5.
(2) Obviously, the assertion holds if d G p ( v 0 ) 1 by (1). Hence, we only need to consider the case d G p ( v 0 ) = 0 . Note that x 1 = 0 ,   2 x 2 6 ,   y 3 ,   d G ( u ) 3 , then,
R ( G ) R ( G ) y 2 y + 1 y x 2 y 2 + 1 2 x 2 + y 2 + 1 d G ( u ) 1 x 2 1 x 2 + y 2 1 2 1 2 1 x 2 1 y + 1 x 2 + y 2 1 2 + 1 3 x 2 + y 2 + 1 2 y + 1 y x 2 y 2 + 2 + 1 3 x 2 + y 2 1 2 > 0 ,
where the last inequality holds by (2) of Lemma 5. □
Theorem 4.
Suppose G is a graph with two vertices u and v such that d G ( u ) d G ( v ) 2 and d G p ( u ) d G p ( v ) 1 . Let G = Θ ( G , v , u , 1 ) , then R ( G ) > R ( G ) if d G n ( v ) = d G n ( u ) and z N G n ( u ) v 1 d G ( z ) = z N G n ( v ) u 1 d G ( z ) .
Proof. 
Let E ˜ be the edge set of G that are not incident with u or v, and x = d G ( u ) 2 , y = d G ( v ) 2 . And, observe that | d G n ( u ) | + z N G n ( u ) v 1 d G ( z ) = | d G n ( v ) | + z N G n ( v ) u 1 d G ( z ) , denoted by k.
If u and v are not adjacent, then k = z N G n ( u ) ( 1 d G ( z ) 1 ) 0 , so
R ( G ) = p q E ˜ 1 d G ( p ) d G ( q ) + z N G n ( u ) 1 x d G ( z ) + x | N G n ( u ) | x + z N G n ( v ) 1 y d G ( z ) + y | N G n ( v ) | y = p q E ˜ 1 d G ( p ) d G ( q ) + x + k x + y + k y R ( G ) = p q E ˜ 1 d G ( p ) d G ( q ) + x + 1 + k x + 1 + y 1 + k y 1 .
Therefore, we get R ( G ) > R ( G ) by (1) of Lemma 7.
If u and v are adjacent, then k = 1 + z N G n ( u ) v ( 1 d G ( z ) 1 ) 1 , so
R ( G ) = p q E ˜ 1 d G ( p ) d G ( q ) + z N G n ( u ) v 1 x d G ( z ) + z N G n ( v ) u 1 y d G ( z ) + x | N G n ( u ) | x + y | N G n ( v ) | y + 1 x y = p q E ˜ 1 d G ( p ) d G ( q ) + x + k x + y + k y + 1 x y R ( G ) = p q E ˜ 1 d G ( p ) d G ( q ) + x + 1 + k x + 1 + y 1 + k y 1 + 1 ( x + 1 ) ( y 1 ) ,
hence we have R ( G ) > R ( G ) by (2) of Lemma 7. Thus, the proof is complete. □
Lemma 9.
Suppose G is a graph with two vertices u and v such that N G n ( u ) v = N G n ( v ) u and d G p ( u ) > 0 , d G p ( v ) > 0 . Let G = Θ ( G , v , u , ) , then R ( G ) > R ( G ) .
Proof. 
Observe that if we exchange pendant neighbors of v and u, R ( G ) does not change. Hence, the assertion holds easily from Theorem 4. □

4. Main Results

4.1. Undeletable Subgraph and Classification of Tricyclic Graphs

We need the following important definition to start our analysis.
Definition 3
Suppose G is a cyclic graph, then the undeletable subgraph ϕ ( G ) of G is defined as a maximum subgraph without a pendant vertex, i.e., the subgraph arising from G by deleting all pendant vertices recursively.
Obviously, ϕ ( G ) is connected and δ ( ϕ ( G ) ) 2 . Moreover, an undeletable subgraph of a graph is unique. And, it is easy to verify that the undeletable subgraph of a unicyclic graph is a cycle.
With the definition and Theorem 2, we are able to prove the following crucial lemma:
Lemma 10.
Suppose G is a cyclic ( n , m ) -graph with an undeletable subgraph ϕ ( G ) . Then, there exists a ( n , m ) -graph G such that R ( G ) > R ( G ) if there is a vertex w V ( G ) V ( ϕ ( G ) ) with d G ( w ) > 1 .
Proof. 
Let G ^ = G E ( ϕ ( G ) ) by deleting all edges of ϕ ( G ) . By the definition of ϕ ( G ) , G ^ contains no cycle, i.e., G ^ is a forest.
Let T be the tree of G ^ containing w. We claim that T contains exactly one vertex of V ( ϕ ( G ) ) . First, assume that T contains no vertex of V ( ϕ ( G ) ) , then T is not connected with vertices of V ( ϕ ( G ) ) in G, thereby G is disconnected, which is a contradiction. Now, assume that T contains at least two vertices of V ( ϕ ( G ) ) , and denote two of which by x and y, then there is a unique path in T that connects them containing a vertex z V ( ϕ ( G ) ) since E ( T ) E ( ϕ ( G ) ) = . And, there exists a path in ϕ ( G ) that connects u and v since ϕ ( G ) is connected. Therefore, vertex z lies on a cycle of G, implying that it belongs to V ( ϕ ( G ) ) , which contradicts the fact that z V ( ϕ ( G ) ) . So, T contains exactly one vertex of V ( ϕ ( G ) ) , say v 0 .
Let P = v 0 v 1 v h be the longest path from v 0 to all other vertices in T. Note that h 2 because a path from v 0 to a pendant vertex containing w is of length at least two. Hence, d G ( v h 1 ) 2 , d G ( v h 2 ) 2 , and v h 2 is the only non-pendant neighbor of v h 1 . Then, by (1) of Theorem 2, there is an ( n , m ) -graph G = Γ ( G , v h 2 , v h 1 ) such that R ( G ) > R ( G ) . □
For a graph G with an undeletable subgraph ϕ ( G ) , if d G ( w ) = 1 for each w V ( G ) V ( ϕ ( G ) ) , i.e., each v V ( ϕ ( G ) ) only has pendant neighbors in V ( G ) V ( ϕ ( G ) ) , it is said to be a pendant-maximized graph.
Lemma 11.
Suppose G is a pendant-maximized tricyclic ( n , n + 2 ) -graph with undeletable subgraph ϕ ( G ) . If there is a vertex v V ( ϕ ( G ) ) with exactly two non-adjacent neighbors in ϕ ( G ) , then there is an ( n , n + 2 ) -graph G such that R ( G ) > R ( G ) ; otherwise, ϕ ( G ) must be one of the 15 graphs as shown in Figure 2, Figure 3 and Figure 4 up to isomorphism.
Proof. 
(1) We first prove the “if” part. Without loss of generality, let the neighbors of v in ϕ ( G ) be u and w with d G ( u ) d G ( w ) . Observe that d G n ( v ) = 2 because G is pendant-maximized and v has two neighbors in ϕ ( G ) . Hence, N G ( v ) N G ( u ) = since u and w are non-adjacent. Therefore, there is an ( n , n + 2 ) -graph G = Γ ( G , u , v ) such that R ( G ) > R ( G ) by (2) of Theorem 2.
(2) Now, the “otherwise” part. By the definition of undeletable subgraph, ϕ ( G ) is a tricyclic graph, that is, | E ( ϕ ( G ) ) | = | V ( ϕ ( G ) ) | + 2 and δ ( ϕ ( G ) ) 2 . In the remaining argument, all degree and neighbors are constrained in ϕ ( G ) .
We claim that 4 | V ( ϕ ( G ) ) | 10 , where the lower bound is obvious by checking graphs of order one to four. We first show there are at most six vertices of degree two in ϕ ( G ) . Notice that each vertex v of degree two must lie on a cycle of length three because the neighbors of v must be adjacent. Moreover, each cycle of length three contains at most two vertices of degree two, otherwise the cycle is disconnected with other parts of ϕ ( G ) . Since there are at most three edge-disjoint cycles in a tricyclic graph, thereby at most three edge-disjoint cycles of length three. Hence, by Handshaking Lemma, we have 2 × 6 + 3 × ( | V ( ϕ ( G ) ) | 6 ) ( | V ( ϕ ( G ) ) | + 2 ) × 2 , implying | V ( ϕ ( G ) ) | 10 .
Therefore, we can list all possible graph structures for ϕ ( G ) with 4 | V ( ϕ ( G ) ) | 10 , which are exactly those shown in Figure 2, Figure 3 and Figure 4. Thus the proof is complete. □
Let TR i j ( n ) be the set of n-vertex tricyclic graphs obtained from U i j by attaching n | V ( U i j ) | pendant vertices to U i j . It is evident that graphs belonging to TR i j ( n ) are pendant-maximized. On the other hand, if a pendant-maximized tricyclic graph G with ϕ ( G ) U i j , then G TR i j ( n ) .

4.2. Relations between TR i j ( n )

Suppose A and B are two graph sets, and if for any graph G A , there is a graph G B such that R ( G ) > R ( G ) , then this relation is written as R ( A ) > R ( B ) or R ( B ) < R ( A ) . Our remaining task is to figure out the above described relations between all TR i j ( n ) .
Lemma 12.
R ( TR 7 4 ( n ) ) > R ( TR 7 3 ( n ) ) > R ( TR 7 2 ( n ) ) > R ( TR 7 1 ( n ) ) .
Proof. 
We prove the results in the order of left to right.
(1) Suppose G is a graph in TR 7 4 ( n ) with undeletable subgraph labelled as U 7 4 in Figure 4. It may be assumed that d G p ( v 4 ) = 0 ; otherwise, by Lemma 9, pendant neighbors of v 4 can be moved to v 3 without increasing R ( G ) since N G n ( v 3 ) v 4 = N G n ( v 4 ) v 3 . Moreover, we may assume d G p ( v 6 ) = 0 by similarly reasoning. Then clearly, d G ( v 4 ) = d G ( v 6 ) = 2 .
Consider first when d G p ( v 2 ) > 0 . If d G p ( v 3 ) > 0 , by (1) of Theorem 3, moving pendant neighbors of v 3 to v 2 reduces R ( G ) . So we may assume d G p ( v 3 ) = 0 . Since d G n ( v 2 ) = d G n ( v 1 ) = 3 , and 1 v 3 + 1 v 4 + 1 v 7 + 1 v 8 > 2 2 > 2 5 , therefore graph G = Γ ( G , v 1 , v 2 ) satisfies R ( G ) > R ( G ) appealing to (3) of Theorem 2.
Now we turn to the case of d G p ( v 2 ) = 0 , that is, d G ( v 2 ) = 3 . Note that d G n ( v 7 ) = d G n ( v 1 ) = 3 , and 1 v 2 + 1 v 8 + 1 v 5 + 1 v 6 > 1 2 + 1 3 > 2 5 , thus graph G = Γ ( G , v 1 , v 7 ) satisfies R ( G ) > R ( G ) again by (3) of Theorem 2. It is not difficult to check that G and G both belong to TR 7 3 ( n ) . Thus we have R ( TR 7 4 ( n ) ) > R ( TR 7 3 ( n ) ) .
(2) Let G TR 7 3 ( n ) be a graph with undeletable subgraph as U 7 3 in Figure 4. As before, we may assume d G p ( v 4 ) = 0 . Let G = Γ ( G , v 2 , v 1 ) , and obviously G TR 7 2 ( n ) . Since d G ( v 1 ) 4 , d G n ( v 2 ) = 3 and 1 v 3 + 1 v 4 1 2 , then we get R ( G ) > R ( G ) by (4) of Theorem 2. Thus R ( TR 7 3 ( n ) ) > R ( TR 7 2 ( n ) ) holds.
(3) Using similar arguments as (2), there is a graph G = Γ ( G , v 2 , v 1 ) TR 7 1 ( n ) such that R ( G ) > R ( G ) . Hence R ( TR 7 3 ( n ) ) > R ( TR 7 2 ( n ) ) . □
Lemma 13.
R ( TR 6 3 ( n ) ) > R ( TR 6 2 ( n ) ) > R ( TR 6 1 ( n ) ) .
Proof. 
We prove the relations from left to right.
(1) Let G TR 6 3 ( n ) be a graph with ϕ ( G ) as U 6 3 in Figure 4. As before, we assume that d G p ( v 6 ) = 0 .
Let S = 1 v 2 + 1 v 3 + 1 v 5 + 1 v 6 . If d G p ( v 2 ) = 0 , then S > 1 2 + 1 3 > 2 5 . And if d G p ( v 2 ) > 0 , we may assume d G p ( v 3 ) = 0 ; otherwise R ( G ) can be reduced by moving pendant neighbors of v 3 to v 2 according to (1) of Theorem 3. Then, we have S > 1 2 + 1 2 > 2 5 . Moreover, d G n ( v 1 ) = d G n ( v 4 ) = 3 , hence there is a graph G = Γ ( G , v 1 , v 4 ) with R ( G ) > R ( G ) from (3) of Theorem 2. And, it is evident that G TR 6 2 ( n ) , thus we obtain R ( TR 6 3 ( n ) ) > R ( TR 6 2 ( n ) ) .
(2) Suppose G is a graph in TR 6 2 ( n ) with an undeletable subgraph as U 6 2 in Figure 4. Using analogous arguments as (1), we can show that 1 v 1 + 1 v 3 + 1 v 4 + 1 v 5 > 1 2 + 1 4 > 2 5 . Additionally, note that d G n ( v 2 ) = d G n ( v 6 ) = 3 . By (3) of Theorem 2, there is a graph G = Γ ( G , v 2 , v 6 ) TR 6 1 ( n ) such that R ( G ) > R ( G ) . So, it follows R ( TR 6 2 ( n ) ) > R ( TR 6 1 ( n ) ) easily. □
Lemma 14.
R ( TR 4 2 ( n ) ) > R ( TR 4 1 ( n ) ) , R ( TR 7 1 ( n ) ) > R ( TR 4 1 ( n ) ) , R ( TR 6 1 ( n ) ) > R ( TR 4 1 ( n ) ) .
Proof. 
We prove the three relations in the order of left to right.
(1) Let G TR 4 2 ( n ) with the undeletable subgraph as U 4 2 in Figure 3. As before, we assume d G p ( v 6 ) = 0 , i.e., d G ( v 6 ) = 2 . And, notice that d G ( v 4 ) 4 , d G n ( v 7 ) = 3 . Then, by (4) of Theorem 2, graph G = Γ ( G , v 4 , v 7 ) TR 4 1 ( n ) satisfies R ( G ) > R ( G ) . Thus, R ( TR 4 2 ( n ) ) > R ( TR 4 1 ( n ) ) holds clearly.
(2) Let G TR 7 1 ( n ) with an undeletable subgraph as U 7 1 in Figure 4. As before, we assume d G p ( v 3 ) = d G p ( v 5 ) = 0 . Note that d G n ( v 1 ) = 6 and v 5 N G ( v 1 ) { v 2 , v 3 } . If d G p ( v 2 ) > 0 , then by (2) of Theorem 3, R ( G ) can be reduced by moving pendant neighbors of v 2 to v 1 . Similarly, it holds for v 4 . So we may assume d G p ( v 2 ) = d G p ( v 4 ) = 0 .
Let d i = d G ( v i ) , and we have d 2 = d 3 = d 4 = d 5 = 2 and d 1 6 . Let G = G v 2 v 3 + v 2 v 5 , then R ( G ) R ( G ) = ( 1 d 1 d 3 + 1 d 1 d 5 + 1 d 2 d 3 + 1 d 4 d 5 ) ( 1 d 1 + 1 d 1 ( d 5 + 1 ) + 1 d 2 ( d 5 + 1 ) + 1 d 4 ( d 5 + 1 ) ) = 1 d 1 ( 2 1 1 3 ) + 1 2 6 1 6 ( 2 1 1 3 ) + 1 2 6 0.1169 > 0 . It is easy to check that G TR 4 1 . Thus, R ( TR 7 1 ( n ) ) > R ( TR 4 1 ( n ) ) follows.
(3) Let G TR 6 1 ( n ) with the undeletable subgraph as U 6 1 in Figure 4. As before, we assume d G p ( v 5 ) = d G p ( v 7 ) = 0 . Moreover, we may assume d G p ( v 3 ) = 0 . Otherwise, note that d G n ( v 1 ) = 4 and v 5 N G ( v 1 ) { v 2 , v 3 } with d G ( v 5 ) = 2 , then appealing to (2) of Theorem 3, moving pendant neighbors of v 3 to v 1 will decrease R ( G ) .
Now, notice that d G n ( v 2 ) = 4 , d G ( v 3 ) = 2 and v 3 N G ( v 2 ) { v 6 , v 7 } , we can decrease R ( G ) by moving pendant neighbors of v 6 to v 2 if d G p ( v 6 ) > 0 . Hence, we only have to consider the case d G ( v 6 ) = d G ( v 7 ) = 2 .
Let d i = d G ( v i ) , and notice that d 1 4 , d 2 4 . Let G = G v 6 v 7 + v 6 v 1 , then R ( G ) R ( G ) > ( 1 d 2 d 7 + 1 d 6 d 7 ) ( 1 d 2 + 1 d 6 ( d 1 + 1 ) ) = 1 d 2 ( 1 2 1 ) + 1 2 1 2 ( d 1 + 1 ) 1 4 ( 1 2 1 ) + 1 2 1 10 > 0 . It is easy to see that G TR 4 1 ( n ) , thus R ( TR 6 1 ( n ) ) > R ( TR 4 1 ( n ) ) holds. □
Lemma 15.
R ( TR 5 2 ( n ) ) > R ( TR 5 1 ( n ) ) > R ( TR 3 1 ( n ) ) , R ( TR 4 1 ( n ) ) > R ( TR 3 1 ( n ) ) , R ( TR 3 2 ( n ) ) > R ( TR 3 1 ( n ) ) .
Proof. 
We prove the assertions from left to right.
(1) Let G TR 5 2 ( n ) with the undeletable subgraph as U 5 2 in Figure 3. As before, we assume d G p ( v 6 ) = 0 . Observe that N G n ( v 1 ) v 2 = N G n ( v 2 ) v 1 = { v 3 , v 4 } . By Lemma 9, we can move pendant neighbors of v 2 to v 1 and do not increase R ( G ) if d G p ( v 2 ) > 0 . Therefore, we may assume that d G p ( v 2 ) = 0 . Notice that d G n ( v 4 ) = d G n ( v 7 ) = 3 and 1 v 1 + 1 v 2 + 1 v 5 + 1 v 6 > 1 2 + 1 3 > 2 5 , hence there is G = Γ ( G , v 4 , v 7 ) TR 5 1 ( n ) such that R ( G ) > R ( G ) from (3) of Theorem 2. Thus, we obtain R ( TR 5 2 ( n ) ) > R ( TR 5 1 ( n ) ) .
(2) Let G TR 5 1 ( n ) with the undeletable subgraph as U 5 1 in Figure 3. By analogous arguments as (1), we may assume that d G p ( v 6 ) = d G p ( v 2 ) = 0 , that is, d G ( v 6 ) = 2 , d G ( v 2 ) = 3 . And, note that d G n ( v 4 ) = 4 and v 2 N G ( v 4 ) { v 5 , v 6 } . Then, according to (2) of Theorem 3, moving pendant neighbors of v 5 to v 4 reduces R ( G ) if d G p ( v 5 ) > 0 . So, we may assume that d G p ( v 5 ) = 0 .
Let G = G v 5 v 6 + v 5 v 1 , and let d i = d G ( v i ) . Note that d 1 3 , d 4 4 , d G ( v 5 ) = d G ( v 6 ) = 2 . Then, R ( G ) R ( G ) > ( 1 d 4 d 6 + 1 d 5 d 6 ) ( 1 d 4 + 1 ( d 1 + 1 ) d 5 ) = 1 d 4 ( 1 2 1 ) + 1 4 1 2 ( d 1 + 1 ) 1 4 ( 1 2 1 ) + 1 4 1 2 ( 3 + 1 ) = 0 . It is easy to verify that G TR 3 1 ( n ) . Therefore, R ( TR 5 1 ( n ) ) > R ( TR 3 1 ( n ) ) holds.
(3) Let G TR 4 1 ( n ) with the undeletable subgraph as U 4 1 in Figure 3. Using similar arguments as (2), we may assume that d G p ( v 2 ) = d G p ( v 5 ) = d G p ( v 6 ) = 0 , i.e., d G ( v 2 ) = d G ( v 5 ) = d G ( v 6 ) = 2 .
Let G = G v 3 v 2 + v 3 v 6 , and let d i = d G ( v i ) . Note that d 4 5 . Then, R ( G ) R ( G ) > ( 1 d 2 d 4 + 1 d 4 d 6 + 1 d 2 d 3 + 1 d 5 d 6 ) ( 1 d 4 + 1 d 4 ( d 6 + 1 ) + 1 d 3 ( d 6 + 1 ) + 1 d 5 ( d 6 + 1 ) ) = 1 d 4 ( 2 1 1 3 ) + 1 d 3 ( 1 2 1 3 ) + 1 4 1 6 1 5 ( 2 1 1 3 ) + 1 4 1 6 0.01879 > 0 . It is clear that G TR 3 1 ( n ) . Therefore, we obtain R ( TR 4 1 ( n ) ) > R ( TR 3 1 ( n ) ) as desired.
(4) Let G TR 3 2 ( n ) with the undeletable subgraph as U 3 2 in Figure 2. Let S = 1 v 2 + 1 v 3 . If d G p ( v 2 ) = 0 , i.e., d G ( v 2 ) = 3 , then clearly S 1 3 . Otherwise, note that d G n ( v 2 ) = 3 , d G p ( v 2 ) > 0 , according to (1) of Theorem 3, moving pendant neighbors of v 3 to v 2 decreases R ( G ) if d G p ( v 2 ) > 0 . So, S 1 2 in this case. As a consequence, we have S 1 3 by the above argument. Similarly, 1 v 5 + 1 v 6 1 3 , implying that 1 v 2 + 1 v 3 + 1 v 5 + 1 v 6 > 2 5 . And, notice that d G n ( v 1 ) = d G n ( v 4 ) = 3 and N G ( v 1 ) N G ( v 4 ) = . Appealing to (3) of Theorem 2, graph G = Γ ( G , v 1 , v 4 ) satisfies R ( G ) > R ( G ) with G TR 3 1 ( n ) . Therefore, we have R ( TR 3 2 ( n ) ) > R ( TR 3 1 ( n ) ) . □
Before proceeding with more relations, let us define some essential functions and graph classes. Let
F 1 ( n ) = n 4 + 3 n 1 + 1 , F 2 ( n ) = n 9 2 + 3 2 2 n 1 + 3 2 4 , F 3 ( n ) = n 5 + 2 3 3 + 2 n 1 + 1 + 6 3 .
For i = 1 , 2 , 3 , let T R i * ( n ) be n-vertex graphs in TR i 1 ( n ) with a vertex of degree n 1 . It is worth noting that all pendant vertices of T R i * ( n ) are adjacent to a single vertex. Further, it can be verified easily that T R 1 * ( n ) T R n .
Lemma 16.
If G TR 1 1 ( n ) , then R ( G ) F 1 ( n ) with equality if and only if G T R 1 * ( n ) .
Proof. 
If G T R 1 * ( n ) , then clearly R ( G ) = F 1 ( n ) . If G T R 1 * ( n ) , at least two vertices of ϕ ( G ) have pendant neighbors. Suppose the undeletable subgraph ϕ ( G ) is labelled as U 1 1 in Figure 2. Without loss of generality, we may assume that d G p ( v 1 ) d G p ( v 2 ) d G p ( v 3 ) d G p ( v 4 ) . Note that N G n ( v 1 ) v 2 = N G n ( v 2 ) v 1 = { v 3 , v 4 } . By Lemma 9, moving pendant neighbors of v 2 to v 1 will decrease R ( G ) if d G p ( v 2 ) > 0 . Similarly, this holds for v 3 and v 4 . So, we can conclude that R ( G ) > F 1 ( n ) if G T R 1 * ( n ) . So, the proof is complete. □
Lemma 17.
If G TR 2 1 ( n ) , then R ( G ) F 2 ( n ) with equality if and only if G T R 2 * ( n ) .
Proof. 
Suppose the undeletable subgraph ϕ ( G ) is labelled as U 2 1 in Figure 2. If G T R 2 * ( n ) , i.e., one of v 1 , v 2 is adjacent to all pendant neighbors, then obviously R ( G ) = F 2 ( n ) . Then, let us consider the case G T R 2 * ( n ) .
Case 1.
d G p ( v 1 ) > 0 , d G p ( v 2 ) > 0 , d G p ( v 3 ) = d G p ( v 4 ) = d G p ( v 5 ) = 0 .
It is easy to see that N G n ( v 1 ) v 2 = N G n ( v 2 ) v 1 = { v 3 , v 4 , v 5 } . By Lemma 9, R ( G ) can be reduced by moving pendant neighbors of v 2 to v 1 , implying R ( G ) > F 2 ( n ) .
Case 2.
one of v 3 , v 4 , v 5 has pendant neighbors.
We may assume that d G p ( v 3 ) > 0 , d G p ( v 4 ) = d G p ( v 5 ) = 0 . Notice that d G n ( v 1 ) = 4 , d G ( v 4 ) = 2 , and v 4 N G ( v 1 ) { v 2 , v 3 } . By (2) of Theorem 3, we can move pendant neighbors of v 3 to v 1 to reduce R ( G ) . Thus, we have R ( G ) > F 2 ( n ) .
Case 3.
at least two of v 3 , v 4 , v 5 have pendant neighbors.
Suppose d G p ( v 3 ) > 0 , d G p ( v 4 ) > 0 without loss of generality. Note that N G n ( v 3 ) v 4 = N G n ( v 4 ) v 3 = { v 1 , v 2 } . Then, again appealing to Lemma 9, pendant neighbors of v 4 can be moved to v 3 with R ( G ) decreased. Then, we arrive at Case 2, thus R ( G ) > F 2 ( n ) .
Therefore, it completes the proof. □
Lemma 18.
If G TR 3 1 ( n ) , then R ( G ) F 3 ( n ) with equality if and only if G T R 3 * ( n ) .
Proof. 
Suppose the undeletable subgraph ϕ ( G ) is labelled as U 3 1 in Figure 2. If G T R 3 * ( n ) , i.e., v 1 is adjacent to all pendant vertices, then obviously R ( G ) = F 3 ( n ) . So, we suppose that G T R 3 * ( n ) .
Case 1.
d G p ( v 2 ) = d G p ( v 3 ) = 0 .
Consider first d G p ( v 5 ) > 0 . And, observe that d G n ( v 1 ) = 4 , d G ( v 2 ) = 2 and v 2 N G ( v 1 ) { v 3 , v 5 } . Then, by (2) of Theorem 3, we can move pendant neighbors of v 5 to v 1 and get R ( G ) smaller. Likewise, this holds for d G p ( v 4 ) > 0 . Therefore, we obtain R ( G ) > F 3 ( n ) .
Case 2.
one of d G p ( v 2 ) > 0 , d G p ( v 3 ) > 0 holds. Without loss of generality, suppose d G p ( v 2 ) > 0 , d G p ( v 3 ) = 0 .
Subcase 2.1.  d G p ( v 4 ) = d G p ( v 5 ) = 0 . Let G be obtained from G by moving pendant neighbors of v 2 to v 1 . Observe that d G ( v 4 ) = d G ( v 5 ) = 2 , d G ( v 3 ) = 3 . Let x = d G ( v 1 ) 4 and let y = d G ( v 2 ) = d G p ( v 2 ) + d G p ( v 2 ) 4 . So, R ( G ) R ( G ) = ( 6 6 + y 3 + 3 3 + 2 2 + 1 x y + x 4 + 3 3 + 2 x ) ( 6 3 + 1 3 + x + y 7 + 2 3 3 + 2 x + y 3 ) = y 3 + 3 3 + 2 2 + 1 x y + x 4 + 3 3 + 2 x x + y 7 + 2 3 3 + 2 x + y 3 ( 1 3 + 1 6 ) > 0 , where the last inequality holds by Lemma 8. Thus, we obtain R ( G ) > F 3 ( n ) .
Subcase 2.2.  d G p ( v 4 ) > 0 , d G p ( v 5 ) = 0 . Observe that d G n ( v 4 ) = 2 , d G n ( v 2 ) = 3 , moving pendant neighbors of v 4 to v 2 will reduce R ( G ) from (1) of Theorem 3. Then, we arrive at Subcase 2.1.
Subcase 2.3.  d G p ( v 4 ) = 0 , d G p ( v 5 ) > 0 . By analogous argument as Case 1, we can move pendant neighbors of v 5 to v 1 , so we get to Subcase 2.1.
Subcase 2.4.  d G p ( v 4 ) > 0 , d G p ( v 5 ) > 0 . Similarly, as Subcase 2.2, pendant neighbors of v 4 can be moved to v 2 and R ( G ) will decrease. Then, we arrive at Subcase 2.3.
According to the four subcases, we obtain R ( G ) > F 3 ( n ) in this case.
Case 3.
d G p ( v 2 ) > 0 , d G p ( v 3 ) > 0 .
Subcase 3.1.  d G p ( v 4 ) = d G p ( v 5 ) = 0 . Note that d G n ( v 2 ) = d G n ( v 3 ) = 3 and u N G n ( v 2 ) v 3 1 d G ( u ) = u N G n ( v 3 ) v 2 1 d G ( u ) = 1 d G ( v 4 ) + 1 d G ( v 1 ) . By Theorem 9, R ( G ) can be reduced by moving pendant neighbors of v 3 to v 2 . Then, we get the Subcase 2.1.
Subcase 3.2.  d G p ( v 4 ) + d G p ( v 5 ) > 0 . Notice that d G n ( v 2 ) = 3 , d G n ( v 4 ) = 2 . By (1) of Theorem 3, pendant neighbors of v 4 can be moved to v 2 with R ( G ) decreased if d G p ( v 4 ) > 0 . Analogously, this holds for v 5 . Thus, we arrive at Subcase 3.1.
Now, we can conclude that R ( G ) > F 3 ( n ) if vertices other than v 1 of ϕ ( G ) have pendant neighbors. Thus, the proof is complete. □
Lemma 19.
R ( TR 3 1 ( n ) ) > R ( TR 2 1 ( n ) ) > R ( TR 1 1 ( n ) ) .
Proof. 
Suppose A , B are two graph sets with min G A R ( G ) < min G B R ( G ) . Let G A A be a graph satisfying R ( G A ) = min G A R ( G ) . Then, for any graph G B B , we have R ( G B ) min G B R ( G ) > R ( G A ) , that is, R ( A ) < R ( B ) .
So, by Lemmas 16–18, it suffices to show that F 3 ( n ) > F 2 ( n ) > F 1 ( n ) . Observe that n 5 for G TR 3 1 ( n ) or G TR 2 1 ( n ) . Hence, F 3 ( n ) F 2 ( n ) = 1 + 6 3 3 2 4 + 2 3 1 2 1 2 n 1 1 + 6 3 3 2 4 + 2 3 1 2 1 2 4 0.06297 > 0 . And, F 2 ( n ) F 1 ( n ) = 3 2 4 1 + 3 2 1 2 3 n 1 3 2 4 1 + 3 2 1 2 3 4 0.005295 > 0 . Therefore, the assertion holds clearly. □
We draw all the relations mentioned here in Figure 5, in which A B represents R ( A ) < R ( B ) .

4.3. The Proof of Theorem 1

Now we are ready to prove our main result.
Proof of Theorem 1.
First, note that F 1 ( n ) = n 4 + 3 n 1 + 1 and T R 1 * ( n ) T R n , so it is equivalent to show that R ( G ) F 1 ( n ) with equality if and only if G T R 1 * ( n ) . Let F = TR i j ( n ) be the union of all TR i j ( n ) .
If G T R 1 * ( n ) , it is clear that R ( G ) = F 1 ( n ) .
If G TR 1 1 ( n ) { T R 1 * ( n ) } , we have R ( G ) > F 1 ( n ) by Lemma 16.
If G F TR 1 1 ( n ) , by Lemmas 12–15, 19 together, we obtain R ( G ) > R ( T R 1 * ( n ) ) = F 1 ( n ) .
If G F and G is pendant-maximized, by Lemma 11, we can find a graph G F such that R ( G ) > R ( G ) F 1 ( n ) .
If G F and G is not pendant-maximized, by Lemmas 10 and 11, we will again find a graph G F such that R ( G ) > R ( G ) F 1 ( n ) .
Therefore, the theorem holds clearly. □

5. Conclusions

In our current work, we investigate three kinds of graph transformations that decrease the Randić index of graphs, which may be valuable for studying relations between the Randić index and the structure of graphs. For instance, Theorem 4 implies that the pendant neighbors of two vertices of a graph connect to its Randić index predictably. By applying these transformations systematically, the minimum Randić index of tricyclic graphs is determined with the corresponding extremal graphs. In fact, the minimum Randić index of trees, unicyclic, and bicyclic graphs could be obtained by the analogous method without much effort.

Author Contributions

Conceptualization, L.T. and Z.S.; methodology, L.T. and Z.S.; software, L.T.; validation, L.T., Z.S. and M.L.; formal analysis, Z.S.; investigation, M.L.; writing—original draft preparation, L.T.; writing—review and editing, Z.S.; visualization, Z.S.; supervision, M.L.; project administration, L.T.; funding acquisition, L.T. and Z.S. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by Hunan Provincial Natural Science Foundation of China (grant number 2019JJ40005), the Science and Technology Plan Project of Hunan Province (grant number 2016TP1020), the 14th Five-Year Plan Key Disciplines and Application-oriented Special Disciplines of Hunan Province (grant number Xiangjiaotong[2022]351), the Open Fund Project of Hunan Provincial Key Laboratory of Intelligent Information Processing and Application for Hengyang Normal University (grant number 2022HSKFJJ012), and the Shenzhen higher education stability support program (grant number 20220820085638002).

Data Availability Statement

Data are contained within the article.

Acknowledgments

The authors are grateful to the anonymous referee for valuable comments which helped us to improve the manuscript.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Randić, M. On characterization of molecular branching. J. Am. Chem. Soc. 1975, 97, 6609–6615. [Google Scholar] [CrossRef]
  2. Gutman, I.; Furtula, B. Recent Results in the Theory of Randić Index; Univerzitet Kragujevac: Kragujevac, Serbia, 2008. [Google Scholar]
  3. Li, X.; Gutman, I. Mathematical Aspects of Randić-Type Molecular Structure Descriptors; Univerzitet Kragujevac: Kragujevac, Serbia, 2006. [Google Scholar]
  4. Li, X.; Shi, Y. A survey on the Randić index. MATCH Commun. Math. Comput. Chem. 2008, 59, 127–156. [Google Scholar]
  5. Randić, M. The connectivity index 25 years after. J. Mol. Graph. Model. 2001, 20, 19–35. [Google Scholar] [CrossRef] [PubMed]
  6. Arizmendi, G.; Arizmendi, O. Energy of a graph and Randić index. Lin. Algebra Appl. 2021, 609, 332–338. [Google Scholar] [CrossRef]
  7. Das, K.C.; Sun, S.; Gutman, I. Normalized Laplacian eigenvalues and Randić energy of graphs. MATCH Commun. Math. Comput. Chem. 2017, 77, 45–59. [Google Scholar]
  8. Gutman, I.; Furtula, B.; Katanic, V. Randić index and information. AKCE Int. J. Graphs Comb. 2018, 15, 307–312. [Google Scholar] [CrossRef]
  9. Hu, Y.; Li, X.; Yuan, Y. Trees with minimum general Randić index. MATCH Commun. Math. Comput. Chem. 2004, 52, 119–128. [Google Scholar]
  10. Palacios, J.L. Applications of Radon’s inequalities to generalized topological descriptors. Contrib. Math. 2023, 8, 30–37. [Google Scholar] [CrossRef]
  11. Vetrik, T.; Balachandran, S. General Randić index of unicyclic graphs with given number of pendant vertices. Discret. Math. Lett. 2022, 8, 83–88. [Google Scholar]
  12. Bollobás, B.; Erdös, P. Graphs of extremal weights. Ars Combin. 1998, 50, 225–233. [Google Scholar] [CrossRef]
  13. Li, X.; Zhao, H. Trees with small Randić connectivity indices. MATCH Commun. Math. Comput. Chem. 2004, 51, 167–178. [Google Scholar]
  14. Caporossi, G.; Gutman, I.; Hansen, P.; Pavlović, L. Graphs with maximum connectivity index. Comput. Biol. Chem. 2003, 27, 85–90. [Google Scholar] [CrossRef]
  15. Yu, P. An upper bound on the Randić index of trees. J. Math. Study 1998, 31, 225–230. (In Chinese) [Google Scholar]
  16. Gao, J.; Lu, M. On the Randić index of unicyclic graphs. MATCH Commun. Math. Comput. Chem. 2005, 53, 377–384. [Google Scholar]
  17. Du, Z.; Zhou, B. On Randić indices of trees, unicyclic graphs, and bicyclic graphs. Int. J. Quantum Chem. 2011, 111, 2760–2770. [Google Scholar] [CrossRef]
  18. Dehghan-Zadeh, T.; Ashrafi, A.R.; Habibi, N. Maximum and Second Maximum of Randić Index in the Class of Tricyclic Graphs. MATCH Commun. Math. Comput. Chem. 2015, 74, 137–144. [Google Scholar]
  19. Dehghan-Zadeh, T.; Ashrafi, A.R.; Habibi, N. Tetracyclic graphs with extremal values of Randić index. Boll. Unione Mat. Ital. 2015, 8, 9–16. [Google Scholar] [CrossRef]
Figure 1. The structure of T R n .
Figure 1. The structure of T R n .
Symmetry 15 02086 g001
Figure 2. Graphs containing no edge-disjoint cycle.
Figure 2. Graphs containing no edge-disjoint cycle.
Symmetry 15 02086 g002
Figure 3. Graphs containing one edge-disjoint cycle.
Figure 3. Graphs containing one edge-disjoint cycle.
Symmetry 15 02086 g003
Figure 4. Graphs containing three edge-disjoint cycles.
Figure 4. Graphs containing three edge-disjoint cycles.
Symmetry 15 02086 g004
Figure 5. Relations between all TR i j ( n ) .
Figure 5. Relations between all TR i j ( n ) .
Symmetry 15 02086 g005
Table 1. Values of b i calculated by computer.
Table 1. Values of b i calculated by computer.
x 2 b 1 b 2 b 3 b 4 b 5 b 6
22.9141.9754.2504.7976.22412.32
32.8283.7429.8967.2828.12819.11
42.6347.31118.3410.1510.9029.61
52.36312.7429.4313.1014.3343.61
62.03320.0643.0515.9218.2860.95
71.65729.3059.1318.4422.6181.51
81.24340.4777.6020.5427.23105.2
90.796753.5898.4022.1132.07131.9
100.323768.64121.523.0637.06161.5
Table 2. Values of b i calculated by computer.
Table 2. Values of b i calculated by computer.
x 2 b 1 b 2 b 3 b 4 b 5 b 6
21.6330.85240.51971.5862.7484.759
31.4491.4432.9122.4613.2257.122
41.1383.7767.9143.6974.59913.00
50.74437.94415.364.8986.59622.15
60.292514.0025.155.8129.03234.43
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Tang, L.; Song, Z.; Lin, M. Tricyclic Graph with Minimum Randić Index. Symmetry 2023, 15, 2086. https://0-doi-org.brum.beds.ac.uk/10.3390/sym15112086

AMA Style

Tang L, Song Z, Lin M. Tricyclic Graph with Minimum Randić Index. Symmetry. 2023; 15(11):2086. https://0-doi-org.brum.beds.ac.uk/10.3390/sym15112086

Chicago/Turabian Style

Tang, Liangwen, Zhumei Song, and Mugang Lin. 2023. "Tricyclic Graph with Minimum Randić Index" Symmetry 15, no. 11: 2086. https://0-doi-org.brum.beds.ac.uk/10.3390/sym15112086

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop