DEPARTMENT OF MATHEMATICS, IIT GUWAHATI

MA102: Mathematics II Mid Semester Exam May 07, 2022

PART I

Roll No.: Name:

Tutorial Group: T Stud. Sign: Inv. Sign:

1. Write the final answer in the spaces provided for answers. Each correct answer gets a score of 1. 5

QUESTION ANSWER
  
 a b 
(i) If S =   ∈ M2 (R) : a = b then dim(S) is equal to 3
 c d 

(ii) If p(λ) = (λ−1)(λ−2)(λ+5)2 is the characteristic polynomial of a matrix A and 1

A is NOT diagonalizable then the geometric multiplicity of −5, as an eigenvalue of
A, is equal to

(iii) If the third column of RREF (A) of a square matrix A is equal to [1, −3, 0, 0]t c[−1, 3, 1, 0]T , c 6= 0
and rank(A) = 3 then a non-trivial solution to Ax = 0 is equal to

(iv) If A is a 4 × 4 matrix such that rank(A) = 1 then rank(adj(A)) is equal to 0

(v) If T : R4 [x] → R5 [x] is a linear transformation such that T (1 + 2x) = x and 3x2 − x
T (1 + 3x) = x2 then T (2 + 7x) is equal to

2. For each of the following questions, ONE or MORE THAN ONE option(s) is/are correct. Write only
(all) the correct option(s) in the space provided for answer, for full score. There is no negative or partial marks.
Each correct answer gets a score of 2. For example, if options A and B are the only correct choices for a given
question, write A, B for full score. 10

(i) Let A be a 4 × 4 matrix such that the system Ax = [1, −1, 1, 1]t is consistent but the system Ax = [1, 1, 1, 1]t
is not consistent. Then which of the following is/are correct?

(A) Ax = [1, −1, 1, 1]t has infinitely many solutions

(B) Ax = [1, −1, 1, 1]t has a unique solution

(C) Ax = [1, −1, 1, 1]t has exactly two solutions

(D) Ax = [1, 0, 1, 1]t has no solution

ANSWER: A, D

P.T.O
(ii) Let A and B be non-zero real matrices of order 2 × 2 and 2 × 4, respectively such that AB = 0. Then which
of the following is/are correct?

(A) columns of A are linearly dependent

(B) columns of B are linearly dependent

(C) col(B) = (row(A))⊥

(D) col(B) 6= (row(A))⊥

ANSWER: A, B, C

(iii) Let T : R2 → R3 be a linear transformation such that T [1, 0]t = [1, 2, 0]t and T [0, 1]t = [2, 3, 1]t . Then which
of the following is/are correct?

(A) T is onto

(B) T is one-one

(C) if {v1 , v2 } ⊂ R2 is linearly independent then {T (v1 ), T (v2 )} is linearly independent

(D) if {v1 , v2 } ⊂ R2 is such that {T (v1 ), T (v2 )} is linearly independent then {v1 , v2 } is linearly independent

ANSWER: B, C, D

(iv) Let A be a real symmetric matrix of order 4 such that 1, −1, 2 are eigenvalues of A, and the third column of
A is the zero column. Then which of the following is/are correct?

(A) 0 is an eigenvalue of A

(B) [0, 0, 1, 0]t is an eigenvector of A

(C) [0, 0, 1, 1]t is an eigenvector of A

(D) [0, 0, 2, 0]t is an eigenvector of A

ANSWER: A, B, D

(v) Let 1, −1, 3 be the eigenvalues of a 3 × 3 real matrix A. Then which of the following is/are correct?

(A) A2 is invertible

(B) algebraic multiplicity of 1 as an eigenvalue of A2 is 2

(C) geometric multiplicity of 1 as an eigenvalue of A2 is 2

(D) geometric multiplicity of 1 as an eigenvalue of A2 is 1

ANSWER: A, B, C
 DEPARTMENT OF MATHEMATICS, IIT GUWAHATI

MA102: Mathematics II Mid Semester Exam Maximum Marks: 40

Date: May 07, 2022 Time: 2 pm - 4 pm
Instructions

• This exam has questions in two parts (Part I & Part II) given in two separate sheets.

• Write answers of Part I (given separately) in the question paper itself and return to the invigilator within
35 minutes (at 2:35 PM) of the exam.

• Write answers of Part II in the given answer script.

• This exam has questions for a total of 40 marks. Part I has 15 marks and Part II has 25 marks.

• Any page used for rough work must have a heading Rough Work. That page will NOT be evaluated.

• No question in this exam requires any clarification from the instructors/invigilators. Even if a question has an
error or incomplete data, you are advised to write answer according to your understanding or write reasons
for why it is not possible to solve partially or completely that question by citing errors/insufficient data.

• Necessary justifications must be given for all your arguments.

PART II

1. If U = {A ∈ M2 (R) : A is upper triangular} and V = {A ∈ M2 (R) : A is symmetric} then find a basis of

U + V . Is U + V a direct sum? 4
       
a b a ba b − c 0 c
Solution: Let 



 ∈ M2 (R)
 . Then 



=
 
 
+
 


 .
c d c d d c
    
0
 
0


       
 
 1 0
 0  0
1  0
0 0

  
   
Hence M2 (R) = U + V , and a basis of U + V is .
  
 ,  ,  ,  
        
 
 0

0
 
0 0
 
1 0
 

 0 1 



Aliter:
     
 
 1  0
0  0
1 0

  
   
Note that dim(U ) = 3, and a basis of U is 


, 
 
, 
 


 .
 
 0

0
 
0 0
 

 0 1 


     
 
1 0
 0 1
 0 0

  

   
Also, dim(V ) = 3, and a basis of V is 


, 
 
, 
 


 .
 
 0

0
 
1 0
 

 0 1 



   
 
1  0
0 0

 


  
Since U ∩ V is the set of 2 × 2 real diagonal matrices, dim(U ∩ V ) = 2, and a basis of U ∩ V is 


, 
 


 .
 
 0

0
 

 0 1 


Hence dim(U + V ) = dim(U ) + dim(V ) − dim(U ∩ V ) = 3 + 3 − 2 = 4. Since U + V


is a subspace
  
of M2(R) and

 
 1  0
0  0
1  0
0 0

 
    
dim(M2 (R)) = dim(U +V ), hence M2 (R) = U +V . Therefore a basis of U +V is  , 
 
, 
 
, 
 


 .
 
 0

0
 
0 0
 
1 0
 

 0 1 


Since U ∩ V 6= {0}, hence U + V is not a direct sum.

   
1 0 0 3
2. Find elementary matrices E1 , E2 , E3 such that E3 E2 E1 A =  , where A =  . 2
0 1 1 1
      
0 3 1 1 0 1
1.  R2 ↔ R1  , the corresponding elementary matrix is E1 =  .
1 1 −−−−−−→ 0 3 1 0
     
1 1 1 1 1 0
2.  R2 → 1 R2   the corresponding elementary matrix is E2 =  .
3
0 3 −
− −−−− −→ 0 1 1
0 3
     
1 1 1 0 1 −1
3.  R1 → R1 − R2   the corresponding elementary matrix is E3 =  .
0 1 −−−−−−−−−−→ 0 1 0 1

All possible permutations of the above elementary row operations and the corresponding elementary matrices
E1 , E2 , E3 are correct answers.

3. If A, B are real square matrices of the same order, then show that col(AB) ⊆ col(A) and row(AB) ⊆ row(B).
If rank(AB) = rank(B) then check whether null(AB) = null(B). 4

Solution: Let y ∈ col(AB). Then y = ABz for some z ∈ Rn , and so y = A(Bz) for Bz ∈ Rn . This implies
that y ∈ col(A). Hence col(AB) ⊆ col(A).

Let y ∈ row(AB). Then yt = zt AB for some z ∈ Rn , and so yt = (zt A)B for zt A ∈ Rn . This implies that
y ∈ row(B). Hence row(AB) ⊆ row(B).
Aliter
Let y ∈ row(AB) then y ∈ col(AB)t , that is y ∈ col(B t At ). Hence by the previous result y ∈ col(B t ) which
implies y ∈ row(B). Hence row(AB) ⊆ row(B).
If rank(AB) = rank(B) then by the Rank-Nullity Theorem nullity(AB) = nullity(B), since AB and B have
the same number of columns.

Further, let x ∈ null(B) so that Bx = 0. Then Bx = 0 ⇒ ABx = 0 ⇒ x ∈ null(AB).

Hence null(B) ⊆ null(AB).

Now nullity(AB) = nullity(B) and null(B) ⊆ null(AB) together gives null(B) = null(AB).

4. Let T : R2 [x] → R3 be such that T (a + bx + cx2 ) = [2a − b, a + b − 3c, c − a]t for all a + bx + cx2 ∈ R2 [x].
Find rank(T ) and nullity(T ). Also, find [T ]C←B with respect to the ordered bases B = {1, 1 + x, 1 + x2 } and
C = {e2 , 2e1 , e3 }. 5

Solution: Since range(T ) is spanned by {T (1), T (x), T (x2 )}, hence {T (1), T (x), T (x2 )} contains a basis of
range(T ).

Note that T (1) = [2, 1, −1]t , T (x) = [−1, 1, 0]t and T (x2 ) = [0, −3, 1]t . Clearly {T (1), T (x)} is linearly
independent, and hence rank(T ) ≥ 2.

Also, T (x) = (− 12 )T (1) + (− 12 )T (x2 ), which implies x + 1

2 + 12 x2 ∈ ker(T ), and hence nullity(T ) ≥ 1.

Now from rank(T ) + nullity(T ) = 3, rank(T ) ≥ 2 and nullity(T ) ≥ 1, we get rank(T ) = 2 and nullity(T ) = 1.

Aliter:

We have

T (a + bx + cx2 ) = 0 ⇔[2a − b, a + b − 3c, c − a]T = 0

 
2 −1 0
⇔[a, b, c]t is a solution of the system Ax = 0, where A = 
 
 1 1 .
−3

−1 0 1
  
1 0 −1
  t 2
Since RREF (A) = 
0 1 −2, we have null(A) = span{[1, 2, 1] }. Hence ker(T ) = span{1 + 2x + x }.


0 0 0

Therefore nullity(T ) = 1, and by Rank-Nullity Theorem, rank(T ) = 3 − nullity(T ) = 2.

Since T (1) = [2, 1, −1]t = e2 + 2e1 − e3 , T (1 + x) = [1, 2, −1]t = 2e2 + 12 (2e1 ) − e3 and

T (1 + x2 ) = [2, −2, 0]t = −2e2 + 2e1 , we have

 
1 2 −2
 
[T ]C←B = 
 1
1
2 .
1

−1 −1 0

 
1 1 1
 
−1 −1 5. Find all the distinct eigenvalues of A and a basis for each of the corresponding
5. Let A =  

0 0 3
eigenspaces. Check whether the matrix A is diagonalizable. 5
 
1−λ 1 1
 
Solution: We have det(A − λI) = det −1 − λ 5  = λ2 (3 − λ).
 −1 
0 0 3−λ
Hence the distinct eigenvalues of A are 0 and 3.
 
1 1 0
  t
Since RREF (A − 0.I) = RREF (A) = 
0 0 1, we have null(A) = span{[1, −1, 0] }.


0 0 0
 
−2 1 1
 
Also note that A − 3I =  −1 −4 5 , and so [1, 1, 1]t ∈ null(A − 3I).
 
0 0 0
 
−2 1 1
 
 0 −9 9, we have rank(A − 3I) = 2, and hence nullity(A − 3I) = 1. This implies
Since REF (A − 3I) =  

0 0 0
null(A − 3I) = span{[1, 1, 1]t }.
 
1 0 −1
 
Or Alternately, since RREF (A − 3I) =  0 1 −1 , we have null(A − 3I) = span{[1, 1, 1]t }.
 
0 0 0

Thus a basis of the eigenspace E0 is {[1, −1, 0]t }, and a basis of the eigenspace E3 is {[1, 1, 1]t }.

Since geometric multiplicity of 0 is 1, which is strictly less than the algebraic multiplicity of 0, hence A is not
diagonalizable.

Or since there does not exist 3 linearly independent eigenvectors of A, hence A is not diagonalizable.

6. If W = {[x, y, z]t ∈ R3 : 2x − y + 3z = 0, y + 3z = 0} then find a basis of W ⊥ . If possible, find a non-trivial

subspace U of R3 such that u.w = 0 for all u ∈ U and for all w ∈ W , but U 6= W ⊥ . Also, find projW (v) and
perpW (v), where v = [−4, −1, 4]t . 5

Solution:
  
2 −1 3
Note that W = null .
0 1 3
   
2 −1 3 1 0 3
Now RREF of   is equal to  .
0 1 3 0 1 3

Hence W = span{[−3, −3, 1]t }, and so W ⊥ = {[x, y, z]t ∈ R3 : −3x − 3y + z = 0} = span{[−1, 1, 0]t , [1, 0, 3]t }.

So {[−1, 1, 0]t , [1, 0, 3]t } is a basis of W ⊥ .

If we take U = span{[−1, 1, 0]t }, then U satisfies the given condition but U is a proper subset of W ⊥ , that is,
U 6= W ⊥ .

Since [−4, −1, 4]t = [−3, −3, 1]t + 2[−1, 1, 0]t + [1, 0, 3]t , we have

projW (v) = [−3, −3, 1]t and

perpW (v) = +2[−1, 1, 0]t + [1, 0, 3]t = [−1, 2, 3]t .

——— ——— END ——— ———