• 검색 결과가 없습니다.

8.3 Symmetric, Skew-Symmetric, and Orthogonal Matrices

N/A
N/A
Protected

Academic year: 2022

Share "8.3 Symmetric, Skew-Symmetric, and Orthogonal Matrices"

Copied!
13
0
0

로드 중.... (전체 텍스트 보기)

전체 글

(1)

중앙대학교 건설환경플랜트공학과 교수

김 진 홍

- 5주차 강의 내용 -

(2)

8.3 Symmetric, Skew-Symmetric, and Orthogonal Matrices

Definitions

A real square matrix A = [ajk] is called

symmetric if transposition leaves it unchanged,

(1) AT = A, thus

skew-symmetric if transposition gives the negatives of A,

(2) AT = -A, thus

orthogonal if transposition gives the inverse of A,

(3) AT = A-1,

jk

,

kj

a

a

jk

,

kj

a

a  

Ex. 1) Symmetric, Skew-Symmetric, and Orthogonal Matrices The matrices below are symmetric, skew-symmetric and orthogonal,

respectively.

3 2 3 2 3

1 3

1 3 2 3 2

3 2 3 1 3 2

, 0 20 12

20 0

9

12 9

0 , 4 2 5

2 0 1

5 1 3

Chap. 8 Linear Algebra : Matrix Eigenvalue Problems

(3)

Ex. 2) Symmetric, Skew-Symmetric, and Orthogonal Matrices

3 2 3 1 3

2 3

1 3 2 3

2 3

2 3 2 4

1

, 0 12 8

12 0 4

8 4 0 , 8 1 4

1 8 4

4 4 7

Chap. 8 Linear Algebra : Matrix Eigenvalue Problems

Any real square matrix A may be written as the sum of a symmetric matrix R and a skew-symmetric matrix S, where

(4) ( )

2 ) 1

2 (

1 T

A A

A A S

R T

Ex. 3) Illustration of Formula (4)













0 0 . 6 5 . 1

0 . 6 0 5 . 1

5 . 1 5 . 1 0

0 . 3 0 . 2 5 . 3

0 . 2 0 . 3 5 . 3

5 . 3 5 . 3 0 . 9

3 4 5

8 3 2

2 5 9

S R A

(4)

Theorem 1

Eigenvalues of Symmetric and Skew-Symmetric Matrices

(a) The eigenvalues of a symmetric matrix are real.

(b) The eigenvalues of a skew-symmetric matrix are pure imaginary or zero.

A=

 

3 1

1

3 det(A-λI)= (3 ) 1 0, 4, 2

3 1

1

3 2

 

  

 Ex) 

☞ The eigenvalues of a symmetric matrix A are real.



 

1 0 1

0 det(B-λI) =    i

  

 1 0,

1

1 2

Ex) B=

☞ The eigenvalues of a skew-symmetric matrix A are pure imaginary.

Chap. 8 Linear Algebra : Matrix Eigenvalue Problems

C= det(C-λI)=

Ex)

8 1 4

1 8 4

4 4 7

0 ) 9 )(

9 ( 8

1 4

1 8

4

4 4

7

2

λ λ λ λ

λ

9 , 9

λ

☞ The eigenvalues of a symmetric matrix A are real.

(5)

Theorem 4

Determinant of an Orthogonal Matrix

The determinant of an orthogonal matrix are the value +1 or -1.

Theorem 5

Eigenvalues of an Orthogonal Matrix

The eigenvalues of an orthogonal matrix

A

are real or complex conjugates in pairs and have absolute value

1

.













3 2 3 2 3 1

3 1 3 2 3 2

3 2 3 1 3

2 det(C-λI)

1 0

3 2 3

2

2

3

   

   

Ex) C =

, 0 3 2 2

3

3

2

  

   

6 11 , 5

1 i



1 ) 6 / 11 ( ) 6 / 5

( 22

☞ The eigenvalues of an orthogonal matrix C are real or complex conjugates in pairs and have absolute value 1.

Chap. 8 Linear Algebra : Matrix Eigenvalue Problems

(6)

8.4 Similarity, Diagonalization

Definition

Similar Matrices, Similarity Transformation

An n x n matrix  is called similar to n x n matrix A if (4)

for some (nonsingular) n x n matrix P. This information, which gives  from A, is called a similarity transformation.

AP P

A  

1

Theorem 3

Eigenvalues and Eigenvectors of Similar Matrices

If

Â

is similar to

A

, then

Â

has the same eigenvalues as

A

. Furthermore, if

x

is an eigenvector of

A

, then

y = P-1x

is an eigenvector of

Â

corres- ponding to the same eigenvalue.

Ex. 4) Eigenvalues and Vectors of Similar Matrices

 

 

 

1 4

3

A 6

and

 

 

4 1

3 P 1

Chap. 8 Linear Algebra : Matrix Eigenvalue Problems

(7)



 





 



 

 

 

 

2 0

0 3 4 1

3 1 1 4

3 6 1 1

3 A 4



 

 

1 4

3

A 6 det(A-λI) = 5 6 12 0, 3,2

1 4

3

6 2

 

   

 From 



 

 

2 0

0 ˆ 3

A det( -λI) = ( 3)( 2) 0, 3,2

2 0

0

3     

   

From

☞ Similar Matrices have the same eigenvalues.

When

 





 

 

 4

, 3 3 4

; 3 0 4

3 2 4

;

2 A I x1 x2 X1

1 , , 0

0

; 0 0 0

0 2 1

ˆ 1 1

 





 



I x Y

A 1 1 1

1 0 4 3 1 1

3

4 Y

X

P

When

 





 

 

 1

, 1

; 4 0 4

3 3 3

;

3 A I x1 x2 X2

0 , , 1

0

; 1 0 0

0 3 0

ˆ 1 2

I y Y

A 2 2

1

0 1 1 1 1 1

3

4 Y

X

P

☞ y = P-1x is an eigenvector of Â

☞ y = P-1x is an eigenvector of  eigenvector of Â

eigenvector of Â

eigenvector of A

eigenvector of A

Chap. 8 Linear Algebra : Matrix Eigenvalue Problems

(8)

Theorem 4

Diagonalization of a Matrix

If an n x n matrix

A

has a basis of eigenvectors, then

(5) D = X-1AX

is diagonal, with the eigenvalues of

A

as the entries on the main diagonal.

Here

X

is the matrix with these eigenvectors as column vectors.

Ex. 5) Diagonalization



 

 

10 6

9

A 5 Characteristic equation ;

,

11

with λ (A xI) 0. Its eigenvector is

1 2

1 X 3

Chap. 8 Linear Algebra : Matrix Eigenvalue Problems

det(A-λI) = 5 4 0, 1,4

10 6

9

5 2

 

λ λ λ

λ λ

0 9 6 1 2

x x

2 3 X1

,

14

with λ (A4I)x0. 9x19x2 0 Its eigenvector is

1 1 X2

3 2

1

1 1

X

4 0

0 1 1 2

1 3 10 6

9 5 3 2

1

1AX 1 X D

(9)













4 3 1

1 1 3

2 1 1

3 . 9 8 . 1 7 . 17

5 . 5 0 . 1 5 . 11

7 . 3 2 . 0 3 . 7

2 . 0 2 . 0 8 . 0

7 . 0 2 . 0 3 . 1

3 . 0 2 . 0 7 . 0

1AX X D













0 0 0

0 4 0

0 0 3

0 12 3

0 4 9

0 4 3

2 . 0 2 . 0 8 . 0

7 . 0 2 . 0 3 . 1

3 . 0 2 . 0 7 . 0

Chap. 8 Linear Algebra : Matrix Eigenvalue Problems

Ex. 5) Diagonalization

 

 

5 . 5 0 . 1 7 . 17

5 . 5 0 . 1 5 . 11

7 . 3 2 . 0 3 . 7 A

Characteristic equation ;

 

3

 

2

 12   0 , 0 ,

4 ,

3

2 3

1

     

,

1

 3

with

 (

A

 

1I

)

x

 0 .

Its eigenvector is

1 3 1

T

,

2

  4

with

 (

A

 

2I

)

x

 0 .

Eigenvector is

1 1 3

T

,

3

 0

with

(

A

 

3I

)

x

 0 .

Eigenvector is

2 1 4

T 



4 3 1

1 1 3

2 1 1 X

(10)

Chap. 8 Linear Algebra : Matrix Eigenvalue Problems

Ex. 6) Diagonalization





1 2 1

0 1 6

1 2 1 A

, 0 ) 3 )(

4 ( , 0 12 )

det(AλI λ3λ2λλ λλ  3 ,

4 ,

0 2 3

1     

From Ex. 3) 8.1

13 6 1 X1



1 2 1 X2

2 3 2 X3

2 1 13

3 2 6

2 1 1 X

21 / 2 7 / 1 21 / 9

28 / 3 7 / 2 28 / 9

12 / 1 0 12 / 1 X 1

21 / 2 7 / 1 21 / 9

28 / 3 7 / 2 28 / 9

12 / 1 0 12 / 1

1AX X D

1 2 1

0 1 6

1 2 1

2 1 13

3 2 6

2 1 1

3 0 0

0 4 0

0 0 0

8.5 Complex Matrices and Forms Notations

is obtained from by replacing each entry ( ; real) with its complex conjugate . Also, is the transpose of , hence the conjugate transpose of A.

] [ajk

AA[ajk] ,

i

ajk  AT[akj]

i ajk 

A

(11)

Definition

Hermitian, Skew-Hermitian, and Unitary Matrices A square matrix A = is called

Hermitian if , that is ,

skew-Hermitian if , that is ,

unitary if

] [ a

jk

A

A Takj

ajk A

A T  

a

kj

  a

jk

1

 A A T

Ex. 2) Hermitian, Skew-Hermitian, and Unitary Matrices





 

 

 



 

 

i i

i C i

i B i

i A i

2 3 1 2 1

2 3 1 2

1 2 ,

2 , 3

7 3

1

3 1 4

are Hermitian, skew-Hermitian, and unitary matrices.

Chap. 8 Linear Algebra : Matrix Eigenvalue Problems

Ex. 1) Notations

If ,

5 2 6

1 4

3 

 

 

i i

A i then ,

5 2 6

1 4

3

i i

A i and

 

 

i i

AT i

5 2 1

6 4 3

(12)

- Theorem 1 Eigenvalues

(a) The eigenvalues of a Hermitian matrix are real.

* symmetric matrix

(b) The eigenvalues of a skew-Hermitian matrix are pure imaginary or zero.

(c) The eigenvalues of a unitary matrix have absolute value

1

.

* orthogonal matrix Ex. 3) Illustration of Theorem 1

From the matrices in Example 2,

Matrix Characteristic Equation Eigenvalues

Hermitian 9, 2

Skew-Hermitian Unitary

0 18

2 11

 

0 8

2

 2   

i

0

2

   1 

i

i i, 2 4 

i

i 2

3 1 2 , 1 2 3 1 2

1   

* skew-symmetric matrix

Chap. 8 Linear Algebra : Matrix Eigenvalue Problems

(13)

감사합니다

참조

관련 문서

This direct approach of the Jacobi method, which we call the direct Jacobi method, treats the matrix ˜ H as a real symmetric method, and it updates the half of its elements

가산성 백색 가우시 안잡음(additive white Gaussian noise, AWGN) 채널에서 모의실험 결과, 제안된 3차원 직교 주파수분할다중화는 기 존 시스템에 비하여 훨씬 향상된

고속 Center Weighted Hadamard 행렬이 고속알고리즘의 Flow Chart와 밭담 관계는 다음 그림과 같다..

In this paper, we investigate the symmetric arctic rank and characterize the strong linear preservers of sets of completely positive matrices defined by symmetric arctic rank

More precisely, the conjugate gradient least squares (CGLS) method has been extended to construct the offered algorithm which obtains the symmetric arrowhead solution group of

Using the covering technique, Malniˇ c and Potoˇ cnik [32] classified the vertex-transitive elementary abelian coverings of the Petersen graph when the fibre-preserving group

[6] Simply connected and complete 3-dimensional Sasakian space forms M 3 (c) of constant holomorphic sectional curvature c are isomorphic to one of the following unimodular Lie

The current study aims to show the significance of connectivity analysis on the SPD space and introduce inference algorithms on the SPD manifold, such as regression