5.
Diagonalization
The goal here is to develop a useful factorization A = PDP 1 , when A is n n. We can use this to
compute A k quickly for large k.
The matrix D is a diagonal matrix (i.e. entries off the main diagonal are all zeros).
D k is trivial to compute as the following example illustrates.
EXAMPLE: Let D =
5 0
0 4
. Compute D 2 and D 3 . In general, what is D k , where k is a
positive integer?
Solution:
D2 =
D3 = D2D =
5 0
5 0
0 4
0 4
52
5 0
0 42
0 4
0
0
and in general,
5k
Dk =
EXAMPLE: Let A =
P=
1 1
1 2
D=
6 1
2
5 0
0 4
0 4k
. Find a formula for A k given that A = PDP 1 where
and P 1 =
Solution:
A 2 = PDP 1 PDP 1 = PDP 1 P DP 1 = PDDP 1 = PD 2 P 1
Again,
A 3 = A 2 A = PD 2 P 1 PDP 1 = PD 2 P 1 P DP 1 = PD 3 P 1
In general,
A k = PD k P 1 =
1 1
5k
1 2
0 4k
2 5k 4k
5 k + 4 k
2 5 k 2 4 k 5 k + 2 4 k
A square matrix A is said to be diagonalizable if A is similar to a diagonal matrix, i.e. if
A = PDP 1 where P is invertible and D is a diagonal matrix.
When is A diagonalizable?
A.)
(The answer lies in examining the eigenvalues and eigenvectors of
Note that
6 1
6 1
=5
=4
5 4
and
3
Altogether
6 1
1 1
1 2
6 1
1 1
1 2
5 8
Equivalently,
1 1
___
1 2
___
or
6 1
2
1 1
5 0
1 1
1 2
0 4
1 2
In general,
v1 v2 vn
v1 v2 vn
and if
v1 v2 vn
is invertible, A equals
v1 v2 vn
THEOREM 5
v1 v2 vn
The Diagonalization Theorem
An n n matrix A is diagonalizable if and only if A has n linearly independent eigenvectors.
In fact, A = PDP 1 , with D a diagonal matrix, if and only if the columns of P are n linearly
independent eigenvectors of A. In this case, the diagonal entries of D are eigenvalues of
A that correspond, respectively, to the eigenvectors in P.
EXAMPLE: Diagonalize the following matrix, if possible.
A=
0 0
2 1
1 0 1
Step 1. Find the eigenvalues of A.
detA I = det
= 2 2 1 = 0.
Eigenvalues of A: = 1 and = 2.
Step 2. Find three linearly independent eigenvectors of A.
By solving A I x = 0, for each value of , we obtain the following:
0
Basis for = 1:
v1 =
1
1
1
0
Basis for = 2:
v2 =
v3 =
Step 3: Construct P from the vectors in step 2.
0 0 1
P=
1 1
1 0
Step 4: Construct D from the corresponding eigenvalues.
1 0 0
D=
0 2 0
0 0 2
Step 5: Check your work by verifying that AP = PD
AP =
0 0 1
0 0
2 1
1 1
1 0 1
1 0
0 0 1
PD =
0
=
0 2
1 2
1 0 0
0 2
1 1
0 2 0
1 0
0 0 2
1 2
EXAMPLE: Diagonalize the following matrix, if possible.
2 4 6
A=
0 2 2
0 0 4
Since this matrix is triangular, the eigenvalues are = 2 and = 4. By solving A I x = 0 for
each eigenvalue, we would find the following:
1
Basis for = 2 :
v1 =
0
0
5
Basis for = 4 :
v2 =
1
1
Every eigenvector of A is a multiple of v 1 or v 2 which means there are not three linearly
independent eigenvectors of A and by Theorem 5, A is not diagonalizable.
2 0 0
EXAMPLE: Why is A =
2 6 0
diagonalizable?
3 2 1
Solution: Since A has three eigenvalues ( 1 = ____, 2 = ____, 3 = ____) and since
eigenvectors corresponding to distinct eigenvalues are linearly independent, A has three linearly
independent eigenvectors and it is therefore diagonalizable.
THEOREM 6
An n n matrix with n distinct eigenvalues is diagonalizable.
EXAMPLE: Diagonalize the following matrix, if possible.
A=
0 0
0 0
24 12 2 0
0
0 2
Solution: Eigenvalues: 2 and 2 (each with multiplicity 2).
Solving A I x = 0 yields the following eigenspace basis sets.
1
Basis for = 2 :
v1 =
v2 =
6
0
v3 =
3
0
0
Basis for = 2 :
0
1
0
v4 =
0
0
1
v 1 , v 2 , v 3 , v 4 is linearly independent P = v 1 v 2 v 3 v 4 is invertible
where
2
1 0 0 0
P=
0 1 0 0
6 3 1 0
0 0 0 1
THEOREM 7
and
D=
A = PDP 1 ,
0 0 0
0 2 0 0
0
0 2 0
0 0 2
Let A be an n n matrix whose distinct eigenvalues are 1 , , p .
a. For 1 k p, the dimension of the eigenspace for k is less than or equal to the
multiplicity of the eigenvalue k .
b. The matrix A is diagonalizable if and only if the sum of the dimensions of the distinct
eigenspaces equals n, and this happens if and only if the dimension of the eigenspace for
each k equals the multiplicity of k .
c. If A is diagonalizable and k is a basis for the eigenspace corresponding to k for each k,
then the total collection of vectors in the sets 1 , , p forms an eigenvector basis for R n .
