Eigen values and eigen vectors
Characteristic roots or Eigen values:
The roots of characteristic equation
|𝜆𝐼 − 𝐴| = 0 are called characteristic roots or
characteristic values or eigen values or latent
roots or proper values of the matrix A.
Characteristic vectors or Eigen vectors:
A non zero vector X is called a Characteristic
vectors or Eigen vectors of a square matrix A,
there exists a eigen vector 𝜆 such that 𝐴𝑋 = 𝜆𝑋
Properties of eigen vectors
1) The eigen vectors X of a square matrix is
not unique
2) No two distinct eigen value corresponds to
a eigen vector X of a sqare matrix A.
3) Different eigen vectors correspond to the
same eigen value 𝜆 of a sqare matrix A.
Example 1: Find the eigen values and eigen
2 0 0
vectors of the matrix 𝐴 = [−1 1 0]
1 −4 0
Solution: The characteristic equation is
|𝜆𝐼 − 𝐴| = 0
𝜆−2 0 0
=> | 1 𝜆 − 1 0| = 0
−1 4 𝜆
=> (𝜆 − 2)(𝜆 − 1)( 𝜆) = 0
∴ 𝜆 = 0, 1, 2 (eigen values)
𝑥1
Now by definition 𝑋 = [ 𝑥2 ] is an eigen vectors
𝑥3
of A corresponding to the eigenvalue of 𝜆 if and
only if X is a non-trivial solution of
[𝜆𝐼 − 𝐴]𝑋 = 0
𝜆−2 0 0 𝑥1 0
=> [ 1 𝜆 − 1 0] [𝑥2 ] = [0] ------(1)
−1 4 𝜆 𝑥3 0
Case 1: Eigen vector corresponding to eigen
value 𝜆 =0, put 𝜆 = 0 in (1), we get
−2 0 0 𝑥1 0
[ 1 −1 0] [𝑥2 ] = [0] =>
−1 4 0 𝑥3 0
Forming the linear system, we have
2𝑥1 = 0 − − − − − − − −(2)
{ 𝑥1 − 𝑥2 = 0 − − − − − − − (3)
−𝑥1 + 4𝑥2 = 0 − − − − − −(4)
Solving these we get 𝑥1 = 0 and 𝑥2 = 0
Hence 𝑥3 is a free variable, say 𝑥3 = 𝑎.
Therefore, the eigen vector of A corresponding to
the eigenvalue of 𝜆 = 0 are non-zero vector of
0
the form X1 = [0]
𝑎
0
In particular let, a = 1 then X1= [0] is an eigen
1
vector corresponding to the eigenvalue 𝜆 = 0.
Case 2: Eigen vector corresponding to eigen
value 𝜆 = 1, put 𝜆 = 1 in (1), we get
−1 0 0 𝑥1 0
[ 1 1 0] [𝑥2 ] = [0] =>
−1 4 1 𝑥3 0
Performing i) R13, ii) R32(1), iii) R1(-1),
1 −4 −1 𝑥1 0
[1 0 0 ] [𝑥2 ] = [0] =>
0 0 0 𝑥3 0
Now, the rank of the coefficient matrix A and that of the
augmented matrix [𝐴𝐻 ] are the same. i,e. 𝜌(𝐴) = 2 =
𝜌(𝐴𝐻 ) . So the given system is consistent.
Since the system has 3 unknown variables whereas rank is
2. So the system has infinitely many solution. The general
solution contains n - r = 3 - 2 = 1 arbitrary constant.
Solving these we get 𝑥1 = 0 and 4𝑥2 + 𝑥3 = 0
Hence 𝑥3 is a free variable, say 𝑥3 = 𝑎.
𝑥3 𝑎
𝑥2 = − =−
4 4
0 0
Thus X2 = [ 0𝑎]= [ 0 ] is an eigen vector
− −4
4
corresponding to the eigenvalue of 𝜆 = 1.
Case 3: Eigen vector corresponding to eigen
value 𝜆 = 2, put 𝜆 = 2 in (1), we get
0 0 0 𝑥1 0
[ 1 1 0] [𝑥2 ] = [0] =>
−1 4 2 𝑥3 0
Performing i) R13, ii) R21(1), iii) R1(-1), R2(− 41)
1 −4 −2 𝑥1 0
2
[0 1 ] [𝑥2 ] = [0] =>
5
𝑥3 0
0 0 0
Obviously, the rank of the coefficient matrix A and that of
the augmented matrix [𝐴𝐻 ] are the same. i,e. 𝜌(𝐴) =
2 = 𝜌(𝐴𝐻 ) . So the given system is consistent.
Since the system has 3 unknown variables whereas rank is
2. So, the system has infinitely many solutions. The
general solution contains n - r = 3 - 2 = 1 arbitrary
constant.
Hence 𝑥3 is a free variable, say 𝑥3 = 𝑎.
2𝑥3 2𝑎
𝑥2 = − =−
5 5
2𝑎
𝑥1 = 4𝑥2 + 2𝑥3 =
5
2𝑎
5 2
Thus X3 = [ 2𝑎] = [−2] is an eigen vector
−
5 5
𝑎
corresponding to the eigenvalue of 𝜆 = 2.
