Brandenburg Technical University Cottbus

Department 1, Institute of Mathematics

Chair for Numerical Mathematics an Scientific Computing
Prof. Dr. G. Bader, Dr. A. Pawell

Problem Session to the Course: Mathematics I

Environmental and Resource Management WS 2002/03
Solutions to Sheet No. 13 (Deadline: January, 27/28 2002)

Homework
H 13.1: Eigenvalues of A: λ1 = −1, λ2/3 = 1
Eigenvectors:
λ1 = −1:
 
2 0 0
1
 0 1 1  ~x = 0, ~x = (0, 1, −1)T , ~c1 = √ (0, 1, −1)T
0 1 1 2

λ2/3 = 1:

1
~c2 = √ (0, 1, 1)T , ~c3 = (1, 0, 0)T
2

0 0 1
 

C= √1 √1 0 .
2 2
− √12 √1
2
0

q(~x) = ~xT A~x = x21 + 2x2 x3 .

B = C, C T AC = diag(1, −1, −1)

q(C~y ) = (C~y )T A(C~y ) = ~y T C T AC~y = ~y T diag(1, −1, −1)~y = y12 − y22 − y32 .
 
1−λ 1 1
H 13.2: det(A − λE) = det  0 1−λ 5 =
0 −1 −1 − λ

(1 − λ)[(1 − λ)(−1 − λ) + 5] = (1 − λ)[4 + λ2 ] = 0

 ⇒ λ1 = 1, λ2/3 = ±2i

Eigenvectors:
 
1−λ 1 1
 0 1−λ 5  ~x = ~0
0 −1 −1 − λ

λ1 = 1:
 
0 1 1
 0 0 5 , ~x = t(1, 0, 0)T
0 −1 −2

λ1 = 2i:
 
1 − 2i 1 1
 0 1 − 2i 5  ~x = ~0
0 −1 −1 − 2i
 
1 − 2i 1 1  T
2i
 0 −1 −1 − 2i  ~x = ~0, ~x = , −1 − 2i, 1
1 − 2i
0 0 0

λ1 = −2i:
 
1 + 2i 1 1
 0 1 + 2i 5  ~x = ~0
0 −1 −1 + 2i
 
1 + 2i 1 1  T
~ −2i
 0 −1 −1 + 2i  ~x = 0, ~x = , −1 + 2i, 1
1 + 2i
0 0 0
 
1−λ 1
H 13.3: det(A − λE) = det = (1 − λ)(3 − λ) − 1 = λ2 − 4λ + 2.
1 3−λ
Eigenvalues:
√
λ1/2 = 2 ± 2
√
Eigenvectors: λ = 2 + 2
√
√
 
−1 − 2 1
√1 ~x = ~0, ~x = p √ (−1 + 2, 1)T
1 1− 2 4−2 2
√
λ=2− 2
√
√
 
−1 + 2 1
√1 ~x = ~0, ~x = p √ (−1 − 2, 1)T
1 1+ 2 4+2 2

2
  √ √ 
−1+
√ 2 −1− 2
√
√ √
B 4−2 2 4+2 2 
√ 1 √ √ 1
√
4−2 2 4+2 2

 √ 
2+ 2 0√
B −1 = B T , B −1 AB =
0 2− 2

Additional Problems
 
5−λ 2 0
P 13.1: det(A − λE) = det  2 5−λ 0 =
0 0 3−λ

(3 − λ) (5 − λ)2 − 4 = λ2 − 10λ + 21 = 0

⇒ λ1 = 7, λ2,3 = 3.
   
−2 2 0 1
1  
 2 −2 0  ~x = ~0, ~x1 = √ 1
0 0 −4 2 0
     
2 2 0 1 0
1 
 2 2 0  ~x = ~0, ~x2 = √ −1  , ~x3 =  0 
0 0 0 2 0 1
 
1 1 0
1 
C=√ 1 −1 √0 
2 0 0 2
   
1 1 0 5 2 0 1 1 0
1
C T AC =  1 −1 √0   2 5 0   1 −1 √0  =
2
0 0 2 0 0 3 0 0 2
     
14 0 0 7 0 0 λ1 0 0
1
0 6 0 = 0 3 0  =  0 λ2 0 
2
0 0 6 0 0 6 0 0 λ3

b)

q(~x) = ~xT A~x = 5x21 + 4x1 x2 + 5x22 + x23

 
7 0 0
q(~y ) := q(C~y ) = q(~x) = ~y T  0 3 0  ~y = 7y12 + 3y22 + 3y32
0 0 6

The transformation ~x = C~y is an orthogonal transformation of the coor-

dinate system. det C = 1 ⇒ the transformation is a rotation of the
coordinate system.
     
1 1 0 0 1 0 1 −1 0
1  1
C=√ 1 −1 √0  =  1 0 0  √  1 1 √0  =
2 0 0 2 0 0 1 2 0 0 2

3
 π π
 


0 1 0 cos 4
− sin 4
0
 1 π π
0 0   sin 4 cos 4 0 
0 0 1 0 0 1

Reflection about x2 = x1 and a rotation by π4 around the origin.


T 0, i 6= j
P 13.2: C = (~c1 , . . . , ~cn ), ~ci ~cj = , A~ci = λi~ci .
1, i = j

C T AC = C T (A~c1 , . . . , A~cn ) = C T (λ1~c1 , . . . , λn~cn ) =

λ1~cT1 ~c1 λn~cT1 ~cn

 
...
 ...  = diag(λ1 , . . . , λn ).
λ1~cTn ~c1 ... λn~cTn ~cn

P 13.3: The quadratic form q is given by

 
1 −1
q(~x) = x21 − 2x1 x2 + 2x22 = ~xT A~x = ~xT ~x
−1 2

Eigenvalues of A

(1 − λ)(2 − λ) − 1 = 1 − 3λ + λ2 = 0
√
3 5
λ1/2 = ±
2 2

Eigenvectors:

√ √ !T
3 5 1 1 5
λ= + ⇒ ~c1 = q √ − ,1
2 2 10 5 2 2
4 − 2

√ √ !T
3 5 1 1 5
λ= − ⇒ ~c1 = q √ + ,1
2 2 10 5 2 2
4 + 2

C = (~c1 , ~c2 )
√ √ !
3 5 3 5
C T AC = diag + , −
2 2 2 2

√ ! √ !
T T T 3 5 3 5
q(C~y ) = (C~y ) AC~y = ~y C AC~y = + y12 + − y22
2 2 2 2

Problems can be downloaded from the internet site:

http://www.math.tu-cottbus.de/∼pawell/education/erm/erm.html