Mathematics Department

Dr. Hassan Ibrahim

Dr. Mohamed Salman
Dr. Rami Younes
Spring 2022
MATH-203
Assignment 4
Solution

Exercise 1:
(a) Consider the set 𝑺 = { 𝒙 = (𝒙𝟏 , 𝒙𝟐 )|𝒙𝟏 = 𝒙𝟐 𝐨𝐫 𝒙𝟏 = −𝒙𝟐 }. Show that 𝑺 is not a subspace
of ℝ𝟐 .
Solution:
Notice that (1,1) and (1, −1) are two vectors in 𝑆.
However, (1,1) + (1, −1) = (2,0).
since 𝑥1 ≠ 𝑥2 (2 ≠ 0) and 𝑥1 ≠ − 𝑥2 (−2 ≠ 0), it follows that (2,0) does not belong to 𝑆.
𝑆 is not closed under addition⟹ 𝑆 is not a subspace of ℝ2 .

(b) Find the standard matrix of the linear transformation T: ℝ𝟑 → ℝ𝟐 defined by

𝑻(𝒙, 𝒚, 𝒛) = (𝒙 − 𝒚, 𝒚 − 𝒛).
Solution:
𝑇(1,0,0) = (1,0)
𝑇(0,1,0) = (−1,1).
𝑇(0,0,1) = (0, −1).
1 −1 0
The standard matrix 𝐴 = [𝑇 (𝐞𝟏 ) 𝑇(𝐞𝟐 ) 𝑇(𝐞𝟑 )] = [ ].
0 1 −1

𝟑 𝟑 𝟑
𝟑 𝟓 𝟏
(c) Consider the matrix 𝑨 = [ ]
−𝟐 𝟒 −𝟖
−𝟐 −𝟒 𝟎

(i) Is 𝐮 = (1, 0, 1) in null (A)? Justify your answer.

(ii) Find the nullity of 𝑨.
(iii) Compute a basis B for the column space of 𝑨 then extend B to a basis ℝ𝟒 .

Page 1 of 4
Solution:
3 3 3 6
1
3 5 1 4
(i) 𝐴𝐮 = [ ] [0] = [ ] ≠ 𝟎 ⟹ 𝐮 𝑖𝑠 not 𝑖𝑛 null(𝐴)
−2 4 −8 −10
1
−2 −4 0 −2

(ii) Find the reduced echelon form of A:

−3𝑅1 +𝑅2 →𝑅2
3 3 3 1 1 1 2𝑅1 +𝑅3 →𝑅3 1 1 1
3 5 1 1/3𝑅1 →𝑅1 3 5 1 2𝑅1 +𝑅4 →𝑅4 0 2 −2
[ ]⇒ [ ]⇒ [ ]
−2 4 −8 −2 4 −8 0 6 −6
−2 −4 0 −2 −4 0 0 −2 2
1/2𝑅2 →𝑅2 −1𝑅2 +𝑅3 →𝑅3
1 1 1 1/6𝑅3→𝑅3 1 1 1 𝑅2 +𝑅4 →𝑅4 1 0 2
0 2 −2 1/2𝑅4 →𝑅4 0 1 −1 −1𝑅2 +𝑅1→𝑅1 0 1 −1
[ ]⇒ [ ]⇒ [ ]
0 6 −6 0 1 −1 0 0 0
0 −2 2 0 −1 1 0 0 0

nullity(𝐴) = nb. of columns − rank(𝐴) = 3 − 2 = 1.

(iii)
3 3 3 1 0 2
3 5 1 0 1 −1
𝐴=[ ]~[ ]
−2 4 −8 0 0 0
−2 −4 0 0 0 0

By the column space structure theorem, a basis of col(𝐴) is given by the columns of 𝐴 that correspond to
the columns of the leading 1’s in an echelon form of 𝐴. Then

𝐮𝟏 = (3,3, −2, −2), 𝐮𝟐 = (3,5,4, −4) form a basis of col(A)

The standard basis of ℝ4 are

𝐞𝟏 = (1,0,0,0), 𝐞𝟐 = (0,1,0,0, ), 𝐞𝟑 = (0,0,1,0), 𝐞𝟒 = (0,0,0,1)

To extend B to a basis of ℝ4 , let 𝐶 be a matrix whose rows are the vectors 𝐮𝟏 , 𝐮𝟐 , 𝐞𝟑 , 𝐞𝟒

3 3 −2 −2 1 1 −2/3 −2/3
3 5 4 −4 𝑟𝑜𝑤 𝑒𝑐ℎ𝑒𝑙𝑜𝑛 𝑓𝑜𝑟𝑚 0 1 3 −1
𝐶=[ ]→ [ ]
0 0 1 0 0 0 1 0
0 0 0 1 0 0 0 1

From the echelon form of 𝐶, it is clear that rank(C) = 4. Thus, the vectors 𝐮𝟏 , 𝐮𝟐 , 𝐞𝟑 , 𝐞𝟒 are linearly
independent and form a basis of ℝ4 .

Page 2 of 4
(d) Show that if 𝐮 and 𝐯 are nonzero eigenvectors of A that correspond to distinct eigenvalues
𝝀 𝐚𝐧𝐝 µ, then 𝐮 and 𝐯 must be linearly independent.
Solution:
Let 𝐮 and 𝐯 be non-zero eigenvectors that correspond to eigenvalues 𝜆 and µ respectively, with λ ≠ µ.
Assume that 𝐮 and 𝐯 are linearly dependent. Then for some 𝑘: 𝐮 = 𝑘𝐯. It follows that:
𝐮 = 𝑘𝐯 ⟹ 𝐀𝐮 = 𝑘𝐀𝐯 ⟹ λ𝐮 = 𝑘µ𝐯
(then using 𝐮 = k𝐯) ⟹ λ𝐮 = µ𝐮 ⟹ ( λ − µ)𝐮 = 𝟎
λ ≠ µ ⟹ 𝐮 = 𝟎, 𝐜𝐨𝐧𝐭𝐫𝐚𝐝𝐢𝐜𝐭𝐢𝐨𝐧.

Exercise 2:

𝟒 −𝟏
(a) Use Cayley-Hamilton theorem in order to find 𝑨𝟔 if 𝑨 = [ ]
𝟎 𝟐
𝟏 −𝟐 𝟏
(b) Consider the matrix 𝑨 = [𝟎 𝟎 𝟏]
𝟎 𝟎 −𝟑
(i) Is 𝑨 nonsingular? Justify your answer.
(ii) Prove that 𝑨 is diagonalizable.
(iii) Compute a matrix 𝑷 such that 𝑷−𝟏 𝑨𝑷 = 𝒅𝒊𝒂𝒈(−𝟑, 𝟏, 𝟎).

Solution
4 −1
(a) Use Cayley-Hamilton theorem in order to find 𝐴6 if 𝐴 = [ ]
0 2
4−𝜆 −1
The characteristic equation: det(𝐴 − 𝜆𝐼) = [ ] = 𝜆2 − 6𝜆 + 8 = 0
0 2−𝜆
Replace 𝜆 by 𝐴 ⟹ 𝐴2 − 6𝐴 + 8𝐼 = 0 ⟹ 𝐴2 = 6𝐴 − 8𝐼
𝐴3 = 6𝐴2 − 8𝐴 = 6(6𝐴 − 8𝐼) − 8𝐴 = 28𝐴 − 48𝐼,
𝐴4 = 28𝐴2 − 48𝐴 = 28(6𝐴 − 8𝐼) − 48𝐴 = 120𝐴 − 224𝐼,
𝐴5 = 120𝐴2 − 224𝐴 = 120(6𝐴 − 8𝐼) − 224𝐴 = 496𝐴 − 960𝐼,
𝐴6 = 496𝐴2 − 960𝐴 = 496(6𝐴 − 8𝐼) − 960𝐴 = 2016𝐴 − 3968𝐼,
4 −1 1 0 4096 −2016
𝐴6 = 2016 [ ] − 3968 [ ]=[ ].
0 2 0 1 0 64

1 −2 1
(b) Consider the matrix 𝐴 = [0 0 1]
0 0 −3
(i) Is 𝐴 nonsingular? Justify your answer.
No, because of det(𝐴) = (1) (0) (−3) = 0 ⟹ the matrix 𝐴 has no inverse 𝐴−1.

Page 3 of 4
(ii) Prove that 𝐴 is diagonalizable.
The characteristic equation:
1−𝜆 −2 1
det(𝐴 − 𝜆𝐼) = [ 0 −𝜆 1 ] = (1 − 𝜆) (−𝜆) (−3 − 𝜆) = 0
0 0 −3 − 𝜆
⟹ 𝜆 = 1, 0 𝑎𝑛𝑑 − 3
Since the eigenvalues are distinct ⟹ the matrix 𝐴 is diagonalizable.
(iii) Compute a matrix 𝑃 such that 𝑃−1 𝐴𝑃 = 𝑑𝑖𝑎𝑔(−3, 1, 0)
A non-zero eigenvector 𝑋1 corresponding to the eigenvalue 𝜆 = 1:
0 −2 1 𝑥1 0
𝑥
[0 −1 1 ] [ 2 ] = [0] ⟹ −4𝑥3 = 0, −𝑥2 + 𝑥3 = 0 𝑎𝑛𝑑 − 2𝑥2 + 𝑥3 = 0 ⟹ 𝑥2 , 𝑥3 = 0
0 0 −4 𝑥3 0
𝑥1 𝑡 1 1
𝑥
𝐿𝑒𝑡 𝑥1 = 𝑡 ⟹ [ 2 ] = [0] = 𝑡 [0] ⟹ 𝑋1 = [0]
𝑥3 0 0 0
A non-zero eigenvector 𝑋2 corresponding to the eigenvalue 𝜆 = 0:
1 −2 1 𝑥1 0
[0 0 1 ] [ 𝑥2 ] = [ 0] ⟹ −3𝑥3 = 0, 𝑥3 = 0 𝑎𝑛𝑑 𝑥1 − 2𝑥2 + 𝑥3 = 0 ⟹ 𝑥3 = 0
0 0 −3 3 𝑥 0
𝑥1 2𝑡 2 2
𝑥
𝐿𝑒𝑡 𝑥2 = 𝑡 ⟹ 𝑥1 = 2𝑡 ⟹ [ 2 ] = [ 𝑡 ] = 𝑡 [1] ⟹ 𝑋2 = [1]
𝑥3 0 0 0
A non-zero eigenvector 𝑋3 corresponding to the eigenvalue 𝜆 = −3:
4 −2 1 𝑥1 0
[0 3 1] [𝑥2 ] = [0] ⟹ 3𝑥2 + 𝑥3 = 0 𝑎𝑛𝑑 4𝑥1 − 2𝑥2 + 𝑥3 = 0
0 0 0 𝑥3 0
𝑥1 5 5 5
5 𝑡
𝐿𝑒𝑡 𝑥2 = 𝑡 ⟹ 𝑥3 = −3𝑡, 𝑥1 = 𝑡 ⟹ [𝑥2 ] = [ 4 ] = 𝑡 [ 4 ] ⟹ 𝑋3 = [ 4 ]
4 𝑥3 𝑡 1 1
−3𝑡 −3 −3
−1
0 0
3
5/4 1 2 1
−1
𝑃 = [𝑋3 𝑋1 𝑋2 ] = [ 1 0 1] , 𝑃 = 1 −2 −
−3 0 0 4
1
[0 1 3 ]
−1
0 0
3 5
1 1 −2 1 1 2 −3 0 0
−1
⟹ 𝑃 𝐴𝑃 = 1 −2 − [0 0 1] [ 4 ] = [ 0 1 0].
4 0 0 −3 1 0 1
0 0 0
1 −3 0 0
[0 1 3 ]

Page 4 of 4