on unicyclic reflexive graphs. the spectrum of a simple graph (non-oriented, without loops and...

21
ON UNICYCLIC REFLEXIVE GRAPHS

Upload: blaise-lang

Post on 01-Jan-2016

221 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

ON UNICYCLIC REFLEXIVE GRAPHS

Page 2: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

• The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix, along with the usual assumption

• The Interlacing theorem: Let be the eigenvalues of a

graph G and eigenvalues of its induced subgraph H. Then the inequalities

, , hold.• Reflexive graphs are graphs having

1 2( , ,..., )n 0,1

1 2 ... .n

1 2 ... n

1 2 ... k

n k i i i 1,2,...,i k

2 2.

Page 3: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

• Graph G is a maximal reflexive graph inside a given class of graphs C if G is reflexive and any extension G+v that belongs to C has

• Theorem (Smith): For a simple graph G (resp. )

if and only if each component of G is an induced subgraph (resp. proper induced subgraph) of one of the graphs of Fig. 1, all of which have

2 2.

1 2G 1 2G

1 2.G

Page 4: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

Figure 1.Connected graphs that have their largest eigenvalue (the index) equal to 2 are known as Smith graphs

1

1

3

2

n n -1

C n W n

1 2 n

Page 5: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

• Theorem (Schwenk): Given a graph G, let ( ) denote the

set of all cycles containing a vertex and an edge of G, respectively. Then

(i)

(ii)

where Adj(v) denotes the set of neighbors of v, while G – V(C) is the graph obtained from G by removing the vertices belonging to the cycle C.

C v C uvv

uv

2 ,G G v G v u G V Cu Adj v C C v

P P P P

2 ,G G uv G v u G V CC C uv

P P P P

Page 6: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

• Corollary 1. Let G be a graph obtained by joining a vertex of a

graph to a vertex of a graph by an edge. Let ( ) be the subgraph of ( ) obtained by deleting the vertex ( ) from (resp. ). Then

• Corollary 2. Let G be a graph with a pendant edge , being of degree 1. Then where ( ) is the graph obtained from G (resp. ) by deleting the vertex (resp. )

1v

1G 2v 2G

1G 2G

1G 2G1v 2v 1G 2G

1 2 1 2' ' .G G G G GP P P P P

1 2v v 1v

1 2

,G G GP P P

1G 2G

1v 2v1G

Page 7: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

Theorem RS

Let G be a graph with a cut vertex u.i. If at least two components of G-u are supergraphs

of Smith graphs, and if at least one of them is a proper supergraph, then

ii. If at least two components of G-u are Smith graphs, and the rest are subgraphs of Smith graphs, then

iii. If at most one component of G-u is a Smith graph, and the rest are proper subgraphs of Smith graphs, then

2 2.G

2 2.G

2 2.G

Page 8: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

Maximum number of loaded vertices of the cycle

in unicyclicreflexive graph

Page 9: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

Theorem 1.

The cycle of unicyclic reflexive graph of length greater than 8 cannot have more than 7 loaded vertices.

Theorem 2.

The cycle of unicyclic reflexive graph of length greater than 10 cannot have more than 6 loaded vertices.

Page 10: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

The length of the cycle with six loaded vertices

Theorem 3. Maximal length of the cycle of unicyclic reflexive graph with 6

loaded vertices is l = 12.

Page 11: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,
Page 12: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

The length of the cycle with five loaded vertices

Theorem 4.

1. Maximal length of the cycle of unicyclic reflexive graph with 5 loaded vertices, if these vertices are not consecutive, is l = 14.

2. Maximal length of the cycle of unicyclic reflexive graph with 5 consecutive loaded vertices is l = 16.

Page 13: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

The length of the cycle with four loaded vertices

Theorem 5.

1. Maximal length of the cycle of unicyclic reflexive graph with 4 loaded vertices, if there are no consecutive loaded vertices on the cycle, is l = 16.

2. Maximal length of the cycle of unicyclic reflexive graph with 4 loaded vertices, if there are are two (but not three, and not four) consecutive loaded vertices on the cycle is l = 21.

3. Maximal length of the cycle of unicyclic reflexive graph with 4 loaded vertices, if there are are three (but not four) consecutive loaded vertices on the cycle is l = 38.

4. The length of the cycle of unicyclic reflexive graph with 4 consecutive loaded vertices, has no upper bound.

Page 14: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

Let m, p, n, q be the lengths of the paths (a; b), (c; d), (a; d), (b; c), respectivelyThe length of the cycle of graph G is l = m + p + q + n

PG(2) = mpqn-4mpn-4mnq-4pqm- 4pqn+12mn+12mq+12pq+12np+16nq+16mp-32m - 32n - 32p - 32q

Page 15: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

m = 1: PG(2) = -3pqn + 8pn + 12nq + 8pq - 20n - 20q - 16p - 32 n

m n q PG(2) λ2≤2 λ2>2 l1. 1 2 2 4p-64 p≤16 p≥17 212. 1 2 3 6p-60 p≤10 p≥11 163. 1 2 4 8p-56 p≤7 p≥8 144. 1 2 5 10p-52 p≤5 p≥6 135. 1 2 6 12p-48 p≤4 p≥5 136. 1 2 7 14p-44 p≤3 p≥4 137. 1 2 8 16p-40 p≤2 p≥3 138. 1 2 9 18p-36 p≤2 p≥3 149. 1 2 10 20p-32 p≤1 p≥2 14

10. 1 2 11 22p-28 p≤1 p≥2 1511. 1 2 12 24p-24 p≤1 p≥2 1612. 1 3 3 5p-44 p≤8 p≥9 1513. 1 3 4 4p-28 p≤7 p≥8 1514. 1 3 5 3p-12 p≤4 p≥5 1315. 1 3 6 2p+4 / p≥1 /16. 1 3 7 p+20 / p≥1 /17. 1 3 8 36 / p≥1 /

Page 16: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

m = n = 1: PG(2) = 5pq - 8p - 8q – 52

p≥35: PG(2)= 167q - 332 > 0

p≥11 and q≥3: PG(2)= 47q - 140 > 0

p PG(2) λ2≤2 λ2>2 l

1. 2 2q-68 q≤34 q≥35 38

2. 3 7q-76 q≤10 q≥11 15

3. 4 12q-84 q≤7 q≥8 13

4. 5 17q-92 q≤5 q≥6 12

5. 6 22q-100 q≤4 q≥5 12

6. 7 27q-108 q≤4 q≥5 13

7. 8 32q-116 q≤3 q≥4 13

8. 9 37q-124 q≤3 q≥4 14

9. 10 42q-132 q≤3 q≥4 15

Page 17: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

m=n=p=1: PG(2) = -3q - 60 < 0

Page 18: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

The length of the cycle with three loaded vertices

Theorem 6. Let G be the unicyclic reflexive graph with exactly three

loaded vertices of the cycle, and let m, n and k be the lengths of the paths between its loaded vertices, p= min (m,n,k).

1. If p≥3 then the maximal length of the cycle is 18. 2. If p=2: 2.1. m=n=2, the length of the cycle is not bounded. 2.2. m=2, n≥3, k≥3, maximal length of the cycle is 23. 3. If p=1: 3.1. m=n=1, or m=1, n=2, the length of the cycle is not bounded. 3.2 m=1, n≥3, k≥3, maximal length of the cycle is 40.

Page 19: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

Let m, n, k be the lengths of the paths (a; b), (b; c), (c; a), respectively

The length of the cycle of graph G is l = m + n + k

p = min (m,n,k)

PG(2) = -mnk + 4mn + 4mk + 4nk - 12m - 12n - 12k

Page 20: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

p=1: p=3:

p=2: p=4:

m n k l1 3 k≤36 7≤l≤401 4 k≤11 8≤l≤161 5 k≤7 9≤l≤131 6 k≤6 10≤l≤13

m n k l2 3 k≤18 8≤l≤232 4 k≤10 9≤l≤162 5 k≤7 10≤l≤142 6 k≤6 11≤l≤14

m n k l

3 3 k≤12 9≤l≤18

3 4 k≤9 10≤l≤16

3 5 k≤7 11≤l≤15

3 6 k≤6 12≤l≤15

m n k l

4 4 k≤8 12≤l≤16

4 5 k≤7 13≤l≤16

4 6 k≤6 14≤l≤16

Page 21: ON UNICYCLIC REFLEXIVE GRAPHS. The spectrum of a simple graph (non-oriented, without loops and multiple edges) is the spectrum of its adjacency matrix,

p=5:1. m=n=k=5, l=152. m=n=5, k=6, l=163. m=5, n=k=6, l=17

p=6:m=n=k=6, l=18