Diagonalization Matrix
Diagonalization Matrix
Diagonalization
Diagonalization of Matrix
Mongi BLEL
Table of contents
2 Diagonalization
Definition
If A ∈ Mn (R) and λ ∈ R.
λ is called an eigenvalue of the matrix A if there is X ∈ Rn \ {0}
such that
AX = λX .
The corresponding nonzero X are called eigenvectors of the matrix
A.
Example
1 1 1
If A is the matrix A = , then the vector X = is an
−3 5 1
eigenvector for A because AX = 2X . The corresponding eigenvalue
is λ = 2.
Remark Note that if AX = λ and c is any real number, then
A(cX ) = cAX = c(λX ) = λ(cX ). Then, if X is an eigenvector
of A, then so is cX for any nonzero number c.
Definition
If A ∈ Mn (R), the polynomial
qA (λ) = |A − λI |
Example
1 3
Find all of the eigenvalues and eigenvectors of A = .
2 2
1−λ 3
Compute the characteristic polynomial qA (λ) = =
2 2−λ
3
(λ + 1)(λ − 4). The roots of qA (λ) are −1 and 4. X1 = is an
−2
1
eigenvector for A with respect to the eigenvalue −1 and X2 =
1
is an eigenvector for A with respect to the eigenvalue 4.
Example
Find
the eigenvalues
of the following
matrix
1 −1 0 5 4 2 −1 4 −2
A = −1 2 −1, A = 4 5 2, A = −3 4 0 .
0 −1 1 2 2 2 −3 1 3
Definition
If A is a matrix with characteristic polynomial p(),
If A is a matrix with characteristic polynomial qA (λ), the
multiplicity of a root λ of qA is called the algebraic multiplicity of
the eigenvalue λ.
1 0 0
Example Let A = −1 1 −1. The characteristic function of
1 0 2
the matrix A is
λ−1 0 0
qA (λ) = 1 λ−1 1 = (λ − 1)2 (2 − λ).
−1 0 λ−2
The eigenvalue λ = 1 has algebraic multiplicity 2, while λ = 2 has
algebraic multiplicity 1.
Definition
Let A ∈ Mn (R) and λ an eigenvalue of the matrix A. The set
Eλ = {X ∈ Rn ; AX = λX }
is called the eigenspace of A associated to the eigenvalue λ.
Remark
If λ is an eigenvalue of the matrix A ∈ Mn (R), then
Eλ = {X ∈ Rn ; AX = λX } is vector sub-space of Rn . Its
dimension is called the the geometric multiplicity of λ.
Theorem
The geometric multiplicity of an eigenvalue is less than or equal to
its algebraic multiplicity.
Definition
A matrix that has an eigenvalue whose geometric multiplicity is
less than its algebraic multiplicity is called defective.
Theorem
If A ∈ Mn (R) and X1 , . . . , Xm are eigenvectors for different
eigenvalues λ1 , . . . , λm , then X1 , . . . , Xm are linearly independent.
Proof
The proof is by induction. The result is true for m = 1.
Assume the result true for m and let X1 , . . . , Xm+1 be eigenvectors
for different eigenvalues λ1 , . . . , λm+1 .
Definition
A matrix A ∈ Mn (R) is called diagonalizable if there exists an
invertible matrix P ∈ Mn (R) such that the matrix P −1 AP is
diagonal .
Remark
If X1 , . . . , Xn are the columns of the matrix P, then the columns of
the matrix AP are: AX1 , . . . , AXn .
Moreover if
λ1 0 . . . . . . 0
0 λ2 0 . . . ..
.
D = ... ..
. ..
. ..
0 .
.. ..
.
. . . . ..
. 0
0 ... ... 0 λn
then the columns of the matrix PD are: λ1 X1 , . . . , λn Xn .
Then P −1 AP = D ⇐⇒ PD = AP and the columns of the matrix
P form a basis of Rn of eigenvectors of the matrix A.
Theorem
A matrix A ∈ Mn (R) is diagonalizable if and only if it has n
eigenvectors linearly independent. These vectors form a basis of
the vector space Rn .
Examples
Prove that the following matrices are diagonalizable and find an
invertible matrix P ∈ Mn (R) such that the matrix P −1 AP is
diagonal 15
and find A .
1 −1 0 5 4 2 −1 4 −2
A = −1 2 −1, A = 4 5 2, A = −3 4 0 .
0 −1 1 2 2 2 −3 1 3
Theorem
If A ∈ Mn (R) and the characteristic function
Remark
For example, if a matrix A ∈ Mn (R) and has n different
eigenvalues, then A is diagonalizable.
Example
5 4
Consider the matrix A = . The characteristic polynomial
−4 −3
of the matrix A is
5−λ 4
qA (λ) = = (1 − λ)2 .
−4 −3 − λ
Example
−10 −6
Consider the matrix A = . The characteristic polyno-
18 11
mial of the matrix A is
−10 − λ −6
qA (λ) = = (λ − 2)(1 + λ).
18 11 − λ
Then the matrix is diagonalizable.
E−1 = h(−2, 3)i and E2 = h(1,
−2)i.
−1 0
The diagonal matrix is D =
0 2
−2 1
and the matrix P is P = .
3 −2
Example
5 0 4
Consider the matrix A = 2 1 5 . The characteristic poly-
−4 0 −3
nomial of the matrix A is
5−λ 0 4
qA (λ) = 2 1−λ 5 = (1 − λ)3 .
−4 0 −3 − λ
Example
1 0 0
Consider the matrix A = −1 1 −1. The characteristic poly-
1 0 2
nomial of the matrix A is
1−λ 0 0
qA (λ) = −1 1 − λ −1 = (1 − λ)2 (2 − λ).
1 0 2−λ
E1 = h(0, 1, 0), (1, 0, −1)i and E2 = h(0, 1, −1)i.
Then the matrix is diagonalizable.
1 0 0
the diagonal matrix is D = 0 1 0
0 0 2
0 1 0
and P = 1 0 1 .
0 −1 −1
Mongi BLEL Diagonalization of Matrix
Eigenvalue and Eigenvector
Diagonalization
Example
5 −3 0 9
0 3 1 −2
Consider the matrix A = 0
.
0 2 0
0 0 0 2
The characteristic polynomial of the matrix A is
5 − λ −3 0 9
0 3−λ 1 −2
qA (λ) = = (5 − λ)(3 − λ)(2 − λ)2 .
0 0 2−λ 0
0 0 0 2−λ
0 0 0 2
1 3 1 −1
0 2 1 2
and P = 0 0 −1 0 .
0 0 0 1
Example
2 2 −1
Consider the matrix A = 1 3 −1.
−1 −2 2
The characteristic polynomial of the matrix A is
2−λ 2 −1
qA (λ) = 1 3 − λ −1 = −(λ − 1)2 (λ − 5).
−1 −2 2 − λ
Example
7 4 16
Consider the matrix A = 2 5 8 .
−2 −2 −5
The characteristic polynomial of the matrix A is
7−λ 4 16
qA (λ) = 2 5−λ 8 = −(λ − 3)2 (λ − 1).
−2 −2 −5 − λ
E3 = h(1, −1, 0), (4, 0, −1)i, E1 = h(2, 1, −1)i.
Then the matrix A is diagonalizable.
1 0 0
The diagonal matrix is D = 0 3 0 and the matrix P is P =
0 0 3
2 1 4
1 −1 0
−1 0 −1
Example
2 −1 0 12
0 1 0 12
Consider the matrix A = −1 1 1 −1. The characteristic
1 −1 1 3
polynomial of the matrix A is
1
2 − λ −1 0 2
1
0 1−λ 0
qA (λ) = 2 = (1 − λ)(2 − λ)3 .
−1 1 1 − λ −1
1 −1 1 3−λ