ENGG 1120 A&D S TUDENT ’ S NAME :

L INEAR A LGEBRA FOR E NGINEERS S TUDENT ’ S I D :

M IDTERM E XAMINATION
16 March 2023, 19:00–20:30 HKT

Question Points Score

Matrix Calculations 25
Solving Systems of Linear Equations 25
Determinants 25
Rank of a Matrix 25
Total: 100

Exam Duration: 90 minutes.

This is a closed-book, closed-notes examination.

Your solutions have to be written by hand.

One A4 double-sided cheat-sheet is allowed.

You are allowed to use a calculator from the university-approved list of calculators.

Even if you are unable to prove the answer of one part,

you may use that answer in other parts.
Always justify your answers.

Please write your name and student ID at the top of the first page of your solutions.

The question paper, scratch sheets and the answer book must be returned after the
exam.

PAGE 1 OF 9
ENGG 1120 A&D M IDTERM E XAMINATION
L INEAR A LGEBRA FOR E NGINEERS 16 M ARCH 2023, 19:00–20:30 HKT

1. (a) [5 points] Consider the matrix

 
0 0 0 0 0 1
0 0 0 0 2 0
 
0 0 0 3 0 0
M=
0
.
 0 4 1 0 0
0 5 0 0 2 0
6 0 0 0 0 3
Compute (adj(M))−1 .
Solution: First, det(M) = (−1)(2)(−3)(4)(−5)(6) = −720. By using the relation M adj(M) =
det(M)I, we have
 
0 0 0 0 0 1
0 0 0 0 2 0
 
−1 1 −1 0 0 0 3 0 0
(adj(M)) = M= .
det(M) 0 0
720  4 1 0 0
0 5 0 0 2 0
6 0 0 0 0 3

(b) [5 points] Let X be an m × n real matrix and Y be an n × m real matrix, where m, n are positive
integers. Suppose X is row equivalent to YT , and Y is row equivalent to the n × m zero matrix.
Find the set of all possible X.
Solution: There exists a sequence of m×m elementary matrices E1 , E2 , . . . , Ek such that E1 E2 . . . Ek X =
YT . On the other hand, there exists a sequence of n × n matrices E′1 , E′2 , . . . , E′k′ such that
E′1 E′2 . . . E′k′ Y = 0n×m . As all elementary matrices are invertible, we have Y = 0n×m , and
thus X = 0m×n . This solution in unique, so the set of all possible X is {0m×n }.

(c) [5 points] For any positive integer k, let Ak be a k × k matrix defined as

 
0 0 0 ··· 0 0 0
1 0 0 · · · 0 0 0
 
0 1 0 · · · 0 0 0
 
Ak =  ... ... ... . . . ... ... ...  .
 
 
0 0 0 · · · 0 0 0
 
0 0 0 · · · 1 0 0
0 0 0 ··· 0 1 0
In other words, the (i, j)-th entry of Ak is 1 if i = j + 1 for j ∈ {1, 2, . . . , k − 1}. Otherwise, the entry
is 0. Define the 0-th power of a square matrix to be the identity matrix. Find the minimal α + β in
terms of m and n such that (Am )α B(An )β = 0m×n for any arbitrary m × n real matrix B, where
α, β are non-negative integers and m, n are positive integers.
Solution: The result of multiplying Am on the left side of any m × n matrix C can be obtained
by shifting down all entries in C by one row, and filling in the first row with 0’s. Similarly, the
result of multiplying An on the right side of any m × n matrix C can be obtained by shifting left
all entries in C by one column, and filling in the last column with 0’s. Therefore, the worst case is
a m × n matrix B where its (1, n)-th entry is non-zero. We need to shift either m rows downwards
or n columns leftwards to make it a zero matrix. Therefore, the minimal α + β is min{m, n}.

(d) [5 points] Show that zT CT Cz ≥ 0 for any m × n real matrix C and any n × 1 real vector z, where
m, n are positive integers.

PAGE 2 OF 9
ENGG 1120 A&D S TUDENT ’ S NAME :
L INEAR A LGEBRA FOR E NGINEERS S TUDENT ’ S I D :

T T T
Pm that2 Cz is an m × 1 real matrix. Write Cz = (a1 , a2 , . . . , am ) . Then, z C Cz =
Solution: Note
T
(Cz) Cz = i=1 ai ≥ 0.
√ √
1+ 5 1− 5
(e) [5 points] Define ϕ = 2 , ψ= 2 , and
 0  
1 1 1 0
= .
1 0 0 1

Given that      
1 1 1 ϕ ψ ϕ 0 1 −ψ
=√ .
1 0 5 1 1 0 ψ −1 ϕ
Show that n   " √1 n+1 #
ϕ − √1 ψ n+1

1 1 1 5 5
= √1 ϕn − √1 ψ n
1 0 0 5 5

for any non-negative integer n.

 
ϕ ψ
Hint: You may find the inverse of first.
1 1
Solution: Note that  −1  
ϕ ψ 1 1 −ψ
=√ .
1 1 5 −1 ϕ
Therefore,
 n     n   
1 1 1 1 ϕ ψ ϕ 0 1 −ψ 1
=√
1 0 0 5 1 1 0 ψn −1 ϕ 0
!
√1 ϕn+1 − √1 ψ n+1
 n+1
ψ n+1
 
1 ϕ 1 5 5
=√ n n = √1 ϕn √1 ψ n
.
5 ϕ ψ −1 5
− 5

PAGE 3 OF 9
ENGG 1120 A&D M IDTERM E XAMINATION
L INEAR A LGEBRA FOR E NGINEERS 16 M ARCH 2023, 19:00–20:30 HKT

2. Consider the following system of equations, where a is a real parameter.

x1 + (a + 1)x2 +a2 x3 = a3
(1 − a)x1 + (1 − 2a)x2 = a3 .
x1 + (a + 1)x2 +ax3 = a2

(a) [5 points] Find the solution of the given system for a = −1.
Solution:
x1 + x3 = −1
2x1 + 3x2 = −1
x1 − x3 = 1
From the first and the third equation, it is found that x1 = 0 and x3 = −1. Hence, we have
x2 = − 31 , and the solution of the system is
 
1
x= 0, − , −1
3

(b) [5 points] Find the value(s) of a such that the given system has unique solution.
Solution: The augmented matrix is equivalent to

a + 1 a2 a3
 
1
 1−a 1 − 2a 0 a3 
1 a+1 a a2
a2
 
1 a+1 a
∼  1 − a 1 − 2a 0 a3 
0 0 a − a a − a2
2 3

a2
 
1 a+1 a
∼  0 a2 − 2a a2 − a 2a3 − a2 
0 0 a2 − a a3 − a2
a2
 
1 a+1 a
∼  0 a(a − 2) 0 a3 
2
0 0 a(a − 1) a (a − 1)

When a ̸= {0, 1, 2}, the coefficient matrix and the augmented matrix are of rank 3 so the solution
of the system is unique.

(c) [5 points] Find the value(s) of a such that the given system has no solution.
Solution: By the augmented matrix in last part, when a = 2, the cofficient matrix is of rank 2 and
the augmented matrix is of rank 3. Therefore the system has no solution.

(d) [10 points] Find the value(s) of a such that the given system has infinitely many solutions. Solve
the system for the found value(s) of a.
Solution: When a = 0, both the coefficient matrix and the augmented matrix are of rank 1 , we
have infinitely many solutions.
When a = 0, the linear system is reduced and equivalent to

x1 + x2 = 0.

PAGE 4 OF 9
ENGG 1120 A&D S TUDENT ’ S NAME :
L INEAR A LGEBRA FOR E NGINEERS S TUDENT ’ S I D :

Assigning the parameters x1 = s and x3 = t, the set of solutions is given by

{(s, −s, t) | s, t ∈ R}
When a = 1, both the coefficient matrix and the augmented matrix are of rank 2, so we also have
infinitely many solutions.
In this case, the linear system is equivalent to

x1 + 2x2 + x3 = 1
− x2 = 1
thus x2 = −1 and x1 + x3 = 3.
Assigning the parameter x1 = s, the set of solutions is given by

{(s, −1, 3 − s) | s ∈ R}
In conclusion, the linear system has infinitely many solutions for a = 0 and a = 1.

PAGE 5 OF 9
ENGG 1120 A&D M IDTERM E XAMINATION
L INEAR A LGEBRA FOR E NGINEERS 16 M ARCH 2023, 19:00–20:30 HKT

3. Let An be the n × n matrix with aii = 1 (for i = 1, . . . , n), ai,i+1 = ai+1,i = 1 (for i = 1, . . . , n − 1), and
all the other entries are 0, i.e., An has the following forms:
 
1 1 0 ··· ··· 0
 .. .. .. .. 
1
 . . . .

.. .. .. .. .. 
  " #
. . . . .

0 1 1 h i
An =   for n ≥ 3, A2 = , A1 = 1 .
. .
 .. . . . . . . . . . . . 0 1 1

 
. .. .. .. 
. . . . 1
. 
0 ··· ··· 0 1 1

(a) [10 points] Prove that, for n ≥ 3,

det(An ) = det(An−1 ) − det(An−2 ).

Hence, or otherwise, find det(A9 ) and det(A10 ).

Solution: By the cofactor expansion along the first row, then

1 1 0 ··· ··· ··· 0

 
0 1 1 0 ··· ··· 0
 .. .. .. .. .. 
 
. 1 . . . .
 .. .. 
 
.. .. .. ..
det(An ) = (1) det(An−1 ) − (1) det 
. 0 . . . . . 
. . .. .. .. ..
 .. ..

 . . . . 0 
. . .. .. ..
 .. ..

. . . 1
0 0 ··· ··· 0 1 1 (n−1)×(n−1)

By the cofactor expansion down the first column, then

det(An ) = det(An−1 ) − det(An−2 )

Since
1 1
det(A1 ) = det[1] = 1, det(A2 ) = = 1(1) − 1(1) = 0,
1 1

then

det(A3 ) = 0 − 1 = −1, det(A4 ) = −1 − 0 = −1, det(A5 ) = −1 − (−1) = 0,

det(A6 ) = 0 − (−1) = 1, det(A7 ) = 1 − 0 = 1 det(A8 ) = 1 − 1 = 0,
det(A9 ) = 0 − 1 = −1, det(A10 ) = −1 − 0 = −1

(b) [3 points] Using the results in part (a), or otherwise, solve the homogeneous linear system A9 x = 0.
Solution: Since det(A9 ) ̸= 0, the homogeneous linear system has only the trivial solution, i.e.
x = 0.
(c) [7 points] Consider the nonhomogeneous linear system A10 x = b, where b1 = 1 and bi = 0 (for
i = 2, . . . , 10). Using the results in part (a), or otherwise, find x1 by determinants (Cramer’s Rule).

PAGE 6 OF 9
ENGG 1120 A&D S TUDENT ’ S NAME :
L INEAR A LGEBRA FOR E NGINEERS S TUDENT ’ S I D :

Solution: Let

D = det(A10 ) = −1

Let
1 1 0 ··· ··· ··· 0
 
0 1 1 0 ··· ··· 0
 .. .. .. .. .. 
 
. 1 . . . .
 .. .. 
 
.. .. .. ..
. 0
D1 = det  . . . . .
. . .. .. .. ..
 .. ..

 . . . . 0 
. . .. .. ..
 .. ..

. . . 1
0 0 ··· ··· 0 1 1 10×10

By the cofactor expansion down the first column, then

D1 = (1) det(A9 ) = −1

Thus
D1 −1
x1 = = =1
D −1

(d) [5 points] Using the results in part (a), or otherwise, find det(2A−1
10 ).

Solution:
210
det(2A−1
10 ) = = −1024
det(A10 )

PAGE 7 OF 9
ENGG 1120 A&D M IDTERM E XAMINATION
L INEAR A LGEBRA FOR E NGINEERS 16 M ARCH 2023, 19:00–20:30 HKT

4. Let  
1 4 2
A = −λ λ 3 
1 0 −2λ
and B = AT A, where λ is an unspecified real parameter.
(a) [10 points] Find all values of λ such that rank(A) = 2.
Solution:
rank(A) = 2 implies det(A) = 0.

1 4 2
4 2 1 4
det(A) = −λ λ 3 = − 2λ = (12 − 2λ) − 2λ(λ + 4λ)
λ 3 −λ λ
1 0 −2λ
= −10λ2 − 2λ + 12 = −2(λ − 1)(5λ + 6) = 0.

Thus, λ must be equal to 1 or −6/5.

• When λ = 1,
     
1 4 2 1 4 2 1 4 2
A =  −1 1 3 ∼ 0 5 5 ∼ 0 1 1 
1 0 −2 0 1 1 0 0 0

The number of nonzero rows of the echelon form of A is 2, and hence rank(A) = 2.

• When λ = −6/5,
     
1 4 2 1 4 2 1 4 2
A =  6/5 −6/5 3 ∼ 6 −6 15  ∼  0 10 −1 
1 0 12/5 5 0 12 0 0 0

We have rank(A) = 2.

Thus, when λ = 1 or −6/5, rank(A) = 2.

(b) [3 points] Let λ = 1, show that the matrix B is not invertible. Justify your answer.
Solution:
det(B) = det(AT A) = det(AT ) det(A) = det(A) det(A) = (det(A))2 .
When λ = 1, det(A) = 0, in this case, det(B) = 0, and B is not invertible.
(c) [12 points] Let λ = 0, determine the rank of matrix B, and find B−1 if B is invertible.
Solution:
det(A) = 0 if and only if λ = 1 or −6/5, thus when λ = 0, det(B) = (det(A))2 ̸= 0.
rank(B) = 3, and B is invertible.
    
1 0 1 1 4 2 2 4 2
B =  4 0 0   0 0 3  =  4 16 8 
2 3 0 1 0 0 2 8 13

PAGE 8 OF 9
ENGG 1120 A&D S TUDENT ’ S NAME :
L INEAR A LGEBRA FOR E NGINEERS S TUDENT ’ S I D :

   
2 4 2 1 0 0 1 0 0 1 −1/4 0
(B | I) =  4 16 8 0 1 0  ∼ 0 1 0 −1/4 11/72 −1/18  .
2 8 13 0 0 1 0 0 1 0 −1/18 1/9
 
1 −1/4 0
Hence, B−1 =  −1/4 11/72 −1/18 
0 −1/18 1/9

************************************ End of exam paper ************************************

PAGE 9 OF 9