You are on page 1of 23

Linear Algebra and Differential Equations

Lecture 15
More about Eigenvalues and Eigenvectors

July 12, 2016

Linear Algebra and Differential EquationsLecture 15 More aboutJuly


Eigenvalues
12, 2016 and Eigenvect
1 / 23

Eigenvalues and Eigenvectors for a 3 3 Matrix


Find the eigenvalues and eigenvectors of the matrix

1 1 1
A= 0 2 1
1 1 3
The characteristic equation is

1
1
1

2
1
0 = 0
1
1 3

2
1
= (1 )
1 3
= (1 )(2 )(3 )











1

1
+ 1


2 1

Linear Algebra and Differential EquationsLecture 15 More aboutJuly


Eigenvalues
12, 2016 and Eigenvect
2 / 23

The eigenvectors are non trivial solutions of


(A I)v = 0
For = 1 this becomes


0 1 1
a1
0
0 1 1 a2 = 0
1 1 2
a3
0

This implies

a2 + a3 = 0
a1 a2 + 2a3 = 0

3
v1 = 1
1

Linear Algebra and Differential EquationsLecture 15 More aboutJuly


Eigenvalues
12, 2016 and Eigenvect
3 / 23

For = 2 this becomes


1 1 1
b1
0
0
0 1 b2 = 0
1 1 1
b3
0
This implies

b3 = 0
b1 b2 = 0

1
v2 = 1
0

For = 3 this becomes

2 1 1
c1
0
0 1 1 c2 = 0
c3
0
1 1 0
This implies

c2 = c3
c1 = c2

1
v2 = 1
1

Linear Algebra and Differential EquationsLecture 15 More aboutJuly


Eigenvalues
12, 2016 and Eigenvect
4 / 23

So if we form the change of basis matrix using the eigenvectors as columns

3 1 1
P= 1 1 1
1 0 1
then we can diagonalize the original matrix as

1 0 0
P1 AP = 0 2 0
0 0 3

Linear Algebra and Differential EquationsLecture 15 More aboutJuly


Eigenvalues
12, 2016 and Eigenvect
5 / 23

Repeated Eigenvalues
Consider the matrix

1 4 2
A= 0 2 0
1 5 4

The characteristic equation is



1
4
2

0
2

0

1
5
4




=0

For = 3 the eigenvector is given by


2 4 2
a1
0
0 1 0 a2 = 0
1
5
1
a3
0

( 3)( 2)2 = 0

1
v3 = 0
1

Linear Algebra and Differential EquationsLecture 15 More aboutJuly


Eigenvalues
12, 2016 and Eigenvect
6 / 23

For = 2 the eigenvector is

1 4 2
0 0 0
1 5 2

given by

a1
0
a2 = 0
a3
0

2
v2 = 0
1

We are still short a vector for the change of basis matrix. The solution is
to introduce a generalised eigenvector or power eigenvector as a solution
to the equation
(A 2I)2 w = 0
One obvious solution is v2 , another solution is given by
(A 2I)w = v2

Linear Algebra and Differential EquationsLecture 15 More aboutJuly


Eigenvalues
12, 2016 and Eigenvect
7 / 23

1 4 2
b1
2
0 0 0 b2 = 0
1 5 2
b3
1
This gives the equations
b1 + 4b2 2b3 = 2

b1 + 5b1 + 2b3 = 1

Which leads to a possible w and change of basis matrix P

0
1
2
0
0
1/9
w = 1/9
P= 0
7/9
1 1 7/9
P doesnt quite diagonalize the original matrix

3 0 0
P1 AP = 0 2 1
0 0 2

Linear Algebra and Differential EquationsLecture 15 More aboutJuly


Eigenvalues
12, 2016 and Eigenvect
8 / 23

Another example
Consider the matrix

2
0 0
A = 2 0 2
3 3 5

The characteristic equation is





2

0 = (2 )
= (2 )(2 5 + 6) = ( 2)2 ( 3)
3 5
For = 3 the eigenvector is given by


1 0 0
a1
0
2 3 2 a2 = 0
3 3 2
a3
0

0
v3 = 2
3

Linear Algebra and Differential EquationsLecture 15 More aboutJuly


Eigenvalues
12, 2016 and Eigenvect
9 / 23

For = 2 the eigenvector is given by


0
0 0
v1
0
2 2 2 v2 = 0
3 3 3
v3
0
This is only one equation
v1 + v2 v3 = 0
So there are two free variables and we can pick
vectors to span the subspace

1
v2 = 0
v2? =
1

two linearly independent

0
1
1

Linear Algebra and Differential EquationsLecture 15 More about


July
Eigenvalues
12, 2016 and 10
Eigenvect
/ 23

The change of basis matrix and its inverse are

1 0 0
1
0
0
3 2
P= 0 1 2
P1 = 2
1 1 3
1 1 1
and

2 0 0
P1 AP = 0 2 0
0 0 3

so repeated eigenvalues dont always lead to problems with diagonalization.

Linear Algebra and Differential EquationsLecture 15 More about


July
Eigenvalues
12, 2016 and 11
Eigenvect
/ 23

When is a matrix diagonalizable?


An n n matrix is diagonalizable if and only if its has n linearly
independent eigenvectors.
Suppose our matrix is A and it has n linearly independent basis vectors
v1 , v2 , , vn . Construct the change of basis matrix P as a row of these
P = [v1

v2

vn ]

Now
AP = [Av1

Av2

Avn ] = [1 v1

2 v2

n vn ]

because the vs are eigenvectors of A.


Lets define a new matrix, D, a diagonal matrix with the eigenvalues along
the main diagonal.

Linear Algebra and Differential EquationsLecture 15 More about


July
Eigenvalues
12, 2016 and 12
Eigenvect
/ 23

D=

1 0
0 2
..
.. . .
.
.
.
0 0

0
0
..
.

But
PD = [1 v1

2 v2

n vn ]

and so
PD = AP
or
D = P1 AP
and our original matrix has been diagonalized.

Linear Algebra and Differential EquationsLecture 15 More about


July
Eigenvalues
12, 2016 and 13
Eigenvect
/ 23

Distinct eigenvalues Linear independence


Let the eigenvectors of a matrix A be v1 , v2 , , vn . Clearly v1 is linearly
independent by itself. Now if
1 v1 + 2 v2 = 0
then
(A 1 I)(1 v1 + 2 v2 ) = 2 (2 1 )v2 = 0
and 2 = 0. Suppose the first k vectors are linearly independent and
suppose
1 v1 + 2 v2 + + k vk + k+1 vk+1 = 0
0 = (A k+1 I)(1 v1 + 2 v2 + + k vk + k+1 vk+1 )
= 1 (1 k+1 )v1 + 2 (2 k+1 )v2 + + k (k k+1 )vk

Linear Algebra and Differential EquationsLecture 15 More about


July
Eigenvalues
12, 2016 and 14
Eigenvect
/ 23

But the eigenvalues are distinct and the vectors v1 , v2 , , vk are linearly
independent
1 = 2 = = k = 0

k+1 = 0

The cases of k = 1 and k = 2 are true and so by induction all the vectors
have to be linearly independent.
The space spanned by the eigenvectors is called the eigenspace so another
way of stating the previous theorem about diagonalization is that a matrix
is diagonalizable if the eigenspace of the matrix is the whole space.

Linear Algebra and Differential EquationsLecture 15 More about


July
Eigenvalues
12, 2016 and 15
Eigenvect
/ 23

Functions of matrices
If a matrix A is diagonalizable then if can be expressed in the form
A = PDP1
for a change of basis matrix P and a diagonal
easy to compute
m
1
0
0 m
2

Dm = .
..
..
..
.
.
0
0

matrix D. Powers of D are


0
0
..
.
m
n

To calculate Am
Am = (PDP1 )m = (PDP1 )(PDP1 ) (PDP1 ) = PDm P1

Linear Algebra and Differential EquationsLecture 15 More about


July
Eigenvalues
12, 2016 and 16
Eigenvect
/ 23

So for any function that can be expressed


function of a diagonalizable matrix by

f (1 )
0
0
f (2 )

f (A) = P .
..
..
.
0

as a power series we find the

..
.

0
0
..
.

f (n )

1
P

For example
A2
+
eA = I + A +
2!
e 1 0
0 e 2

= P .
..
..
..
.
.
0
0

0
0
..
.
e n

1
P

Linear Algebra and Differential EquationsLecture 15 More about


July
Eigenvalues
12, 2016 and 17
Eigenvect
/ 23

An Example
Find B1/2 if


B=

The eigenvalues are given by



6
5

3
4

6 5
3 4



=0

= 1, 9

For = 1


5 5
3 3



a1
a2



b1
b2

0
0

0
0

v1 =

1
1

For = 9


3 5
3 5

v9 =

5
3

Linear Algebra and Differential EquationsLecture 15 More about


July
Eigenvalues
12, 2016 and 18
Eigenvect
/ 23

The diagonal matrix, the change of basis matrix and its inverse are






1 3 5
1 0
1 5
1
D=
P=
P =
0 9
1 3
8 1 1
Since
D = P1 BP

B = PDP1

then
1/2 1
B1/2 = PD
P 

 

1 5
1 0 1 3 5
=
1 3
0 3 8 1 1


1 9 5
=
4 3 7

Linear Algebra and Differential EquationsLecture 15 More about


July
Eigenvalues
12, 2016 and 19
Eigenvect
/ 23

What if the matrix isnt diagonalizable?


By using power (generalized) eigenvectors we can
reduce to the form

1 1 0
0 1 1

P1 AP = 0 0 1
..
..
.. . .
.
.
.
.
0 0 0

almost do it. We can


0
0
0
..
.

with the eigenvalues on the diagonal and ones on the diagonal above the
main diagonal in the submatrix containing the repeated root(s). Let us
write it in the form
A = P(D + N)P1

Linear Algebra and Differential EquationsLecture 15 More about


July
Eigenvalues
12, 2016 and 20
Eigenvect
/ 23

Where D is the diagonal matrix

0 1 0
0 0 1

N= 0 0 0
.. .. ..
. . .

and N has the form

0
0

0
N2 = 0
.
..
. ..
0 0 0 0

Am = P(D + N)m P1




m
m
m
m1
= P(D +
D
N+
Dm2 N2 + )P1
1
2
= P(Dm + mDm1 N)P1
m
1 mm1
0

1
m1
m
0
1
m1

m
0
0

= P
1
.
.
.
..
..
..
..
.
0
0
0

0
0
0
..
.
m
n

1
P

Linear Algebra and Differential EquationsLecture 15 More about


July
Eigenvalues
12, 2016 and 21
Eigenvect
/ 23

An example
Find A50 if


A=

3
1
4 1

The characteristic equation is


2 2 = 1 = 0
The eigenvector is given by


  
v1
0
2
1
=
4 2
v2
0

=1

v1 =

1
2

To get another vector for our change of basis matrix we need to solve


 

 
2
1
v1
1
1
?
=
v1 =
4 2
v2
2
1

Linear Algebra and Differential EquationsLecture 15 More about


July
Eigenvalues
12, 2016 and 22
Eigenvect
/ 23

then


P=

A50 =

1
1
2 1


=

1 1
2
1

50 

1
1
2 1



1 1
0 1

1
1
2 1



101
50
200 99

1 50
0 1


=
=



1 1
2
1
1 1
2
1




Linear Algebra and Differential EquationsLecture 15 More about


July
Eigenvalues
12, 2016 and 23
Eigenvect
/ 23

You might also like