MTL101::Linear Algebra and Differential Equations

Tutorial 3

Department of Mathematics
Indian Institute of Technology Delhi

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 1 / 107

Question 1

Question 1
Find a basis of the row space of the following matrices:
   t
1 2 3 1 0 2 5
 2 3 4 ,  0 3 1 1  .
3 4 5 3 1 0 1

Recall:
Let A ∈ Mm×n (F), Ri . Denote the i-th row (1 ≤ i ≤ m) by
Ri ∈ M1×n (F). Then Span(R1 , R2 , . . . Rm ) is called row space of A.
Dimension of row space of A = row rank of A.
If a matrix is row echelon matrix, then non zero rows form a basis of
the row space.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 2 / 107

Question 1(a)

Solution:
   
1 2 3 1 2 3
R3 →R3 −3R1
 2 3 4 − −−−−−−− →  0 −1 −2 
R2 →R2 −2R1
3 4 5 0 −2 −4
   
1 2 3 1 2 3
R →R −2R2 R2 →−R2
−−3−−−3−−−→  0 −1 −2  − −−−−→  0 1 2 
0 0 0 0 0 0
 
1 0 −1
R →R −2R2
−−1−−−1−−−→  0 1 2 .
0 0 0
Hence, the required basis of row space is {(1, 0, −1), (0, 1, 2)}.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 3 / 107

Question 1(b)

Solution:
   
 1 0 3
t 1 0 3
1 0 2 5  0 3 1  R4 →R4 −5R1  0 3 1 
 2 1 0 −
 0 3 1 1  =  −−−−−−→  
R3 →R3 −2R1  0 1 −6 
3 1 0 1
 5 1 1  0 1 −14
1 0 3 1 0 3
8
R2 →R2 −3R3  0 0 19  R4 →R4 + 19 R2  0 0 19 
−− −−−−−→  − − − −−−− −→  
R4 →R4 −R3  0 1 −6   0 1 −6 
0 0 −8 0 0 0
   
1 0 3 1 0 3
R2 ↔R3  0 1 −6 
 R3 → 19 R3  0 1 −6 
1  
−− −−→   0 0 19  −−−−−−→  0 0 1 .
0 0 0 0 0 0
Hence, the required basis is {(1, 0, 3), (0, 1, −6), (0, 0, 1)}.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 4 / 107

Question 2

Question 2
For i ∈ {1, 2, . . . , n}, define pi : Fn → F by pi (x1 , x2 , . . . , xn ) = xi (the i-th
projection).
(a) Show that it is a linear transformation.
(b) If T : Fn → F is a linear transformation then it is an F -linear
combination of the projections, that is, T = a1 p1 + a2 p2 · · · + an pn
for a1 , . . . an ∈ F.
(c) Further, show that S : Fm → Fn is a linear transformation if and only
if for each i ∈ {1, 2, . . . , n}, the composition pi ◦ S : Fm → F is a
linear transformation.
(d) If S : Fm → Fn is a linear transformation then S(x1 , x2 , . . . xm )
= (y1 , y2 , . . . yn ) where yi = ai1 x1 + ai2 x2 + · · · + aim xm for aij ∈ F
with (1 ≤ i ≤ n, 1 ≤ j ≤ m).

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 5 / 107

Question 2(a)

Question 2(a)
For i ∈ {1, 2, . . . , n}, define pi : Fn → F by pi (x1 , x2 , . . . , xn ) = xi (the i-th
projection).
(a) Show that it is a linear transformation.

Recall:
Suppose V and W are vector spaces over the same field F. A map
T : V → W is called a linear transformation if

T (au + bv ) = aT (u) + bT (v )

for any a, b ∈ F and any u, v ∈ V .

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 6 / 107

Question 2(a)

Solution:
Let u = (x1 , x2 , . . . , xn ), v = (y1 , y2 , . . . , yn ) ∈ Fn and a, b ∈ F, then

pi (au + bv ) = pi (a(x1 , x2 , . . . xn ) + b(y1 , y2 , . . . yn ))

= pi (ax1 + by1 , ax2 + by2 , . . . , axn + byn )
= axi + byi
= api (x1 , x2 , . . . xn ) + bpi (y1 , y2 , . . . yn )
= api (u) + bpi (v ).

Hence, pi is a linear transformation for i ∈ {1, 2, . . . , n}.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 7 / 107

Question 2(b)

Question 2(b)
For i ∈ {1, 2, . . . , n}, define pi : Fn → F by pi (x1 , x2 . . . , xn ) = xi (the i-th
projection).
(b) If T : Fn → F is a linear transformation then it is an F -linear
combination of the projections, that is, T = a1 p1 + a2 p2 · · · + an pn for
a1 , . . . an ∈ F.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 8 / 107

Question 2(b)
Solution:
Let u = (x1 , x2 , . . . , xn ) ∈ Fn .
For the standard basis {e1 , e2 , . . . , en } of Fn , we have
u = (x1 , x2 , . . . , xn ) = (x1 e1 + x2 e2 + · · · + xn en ).
Since, T is a linear transformation,

T (x1 , x2 , . . . , xn ) = T (x1 e1 + x2 e2 + · · · + xn en )
= x1 T (e1 ) + x2 T (e2 ) + · · · + xn T (en ) ∈ F,

This implies, T (e1 ), T (e2 ), . . . , T (en ) ∈ F.

For ai ∈ F (1 ≤ i ≤ n), suppose T (ei ) = ai , then

T (x1 , x2 , . . . , xn ) = x1 a1 + x2 a2 + · · · + xn an
= a1 p1 (x1 , x2 , . . . , xn ) + · · · + an pn (x1 , x2 , . . . , xn )
= (a1 p1 + a2 p2 + · · · + an pn )(x1 , x2 , . . . , xn ).

Hence, T = a1 p1 + a2 p2 · · · + an pn for a1 , . . . an ∈ F.
(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 9 / 107
Question 2(c)

Question 2(c)
For i ∈ {1, 2, . . . , n}, define pi : Fn → F by pi (x1 , x2 . . . , xn ) = xi (the i-th
projection).
(c) Further, show that S : Fm → Fn is a linear transformation if and only if
for each i ∈ {1, 2, . . . n}, the composition pi ◦ S : Fm → F is a linear
transformation.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 10 / 107

Question 2(c)

Solutions:
Let S is a linear transformation.
Let u, v ∈ Fm and a, b ∈ F. Then,

(pi ◦ S)(au + bv ) = pi (S(au + bv ))

= pi (aS(u) + bS(v ))
= api (S(u)) + bpi (S(v ))
= a(pi ◦ S)(u) + b(pi ◦ S)(u)

Hence, the composition pi ◦ S is a linear transformation for

i ∈ {1, 2, . . . , n}.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 11 / 107

Question 2(c) contd...

Converse Part
Let the composition pi ◦ S is a linear transformation for each
i ∈ {1, 2, . . . n}.
For v ∈ Rn , v = (p1 (v ), p2 (v ), . . . , pn (v )).
Let a, b ∈ F and u, w ∈ Fm , Then

S(au + bv ) = (p1 (S(au + bv )), p2 (S(au + bv )), . . . , pn (S(au + bv )))

= (p1 ◦ S(au + bv ), p2 ◦ S(au + bv ), . . . , pn ◦ S(au + bv ))
= (a(p1 ◦ S))(u), a(p2 ◦ S))(u), . . . , a(pn ◦ S))(u))
+ ((b(p1 ◦ S))(v ), b(p2 ◦ S))(v ), . . . , b(pn ◦ S))(v )))
= aS(u) + bS(v )

Hence, S is a linear transformation.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 12 / 107

Question 2(d)

Question 2(d)
For i ∈ {1, 2, . . . , n}, define pi : Fn → F by pi (x1 , x2 . . . , xn ) = xi (the i-th
projection).
(d) If S : Fm → Fn is a linear transformation then S(x1 , x2 , . . . xm )
= (y1 , y2 , . . . yn ) where yi = ai1 x1 + ai2 x2 + · · · + aim xm for aij ∈ F with
(1 ≤ i ≤ n, 1 ≤ j ≤ m).

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 13 / 107

Question 2(d)
Solution:
Let u = (x1 , x2 , . . . , xm ) ∈ Fm .
For the standard basis {e1 , e2 , . . . , em } of Fm , we know
u = (x1 , x2 , . . . , xm ) = (x1 e1 + x2 e2 + · · · + xm em ).
Since, S is a linear transformation,
S(x1 , x2 , . . . , xm ) = S(x1 e1 + x2 e2 + · · · + xm em )
= x1 S(e1 ) + x2 S(e2 ) + · · · + xm S(em )
This implies, S(e1 ), . . . , S(em ) ∈ Fn .
Let
S(e1 ) = (a11 , . . . , an1 ),
S(e2 ) = (a12 , . . . , an2 ),
···
S(em ) = (a1m , . . . , anm )

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 14 / 107

Question 2(d) contd...

Hence,

S(x1 , x2 , . . . , xm ) = x1 S(e1 ) + x2 S(e2 ) + · · · + xm S(em )

= x1 (a11 , . . . , an1 ) + · · · + xm (a1m , . . . , anm )
= (a11 x1 + a12 x2 + · · · + a1m xm ,
a21 x1 + a22 x2 + · · · + a2m xm ,
. . . , an1 x1 + an2 x2 + · · · + anm xm ).

Hence, S(x1 , x2 , . . . , xm ) = (y1 , y2 , . . . yn ), where,

yi = ai1 x1 + ai2 x2 + · · · + aim xm for aij ∈ F.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 15 / 107

Question 3

Question 3
Find the rank and nullity of the following linear transformations. Also
write a basis of the range space in each case.
(a) T : F3 → F3 defined by T (x, y , z) = (x + y + z, x − y + z, x + z).
(b) Assume that 0 ≤ m ≤ n. T : Fn → Fm defined by
T (x1 , x2 , . . . , xn ) = (x1 , x2 , . . . , xm ).

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 16 / 107

Question 3(a)

Question 3(a)
(a) T : F3 → F3 defined by T (x, y , z) = (x + y + z, x − y + z, x + z).

Recall: Suppose T : V → W is a linear transformation. Then

The set ker (T ) := {v ∈ V : T (v ) = 0} is called the null space of T.
The set T (V ) := {T (v ) : v ∈ V } is called the range space of T .
The dimension of ker (T ) is called the nullity of T and the dimension
of T (V ) is called the rank of T .
Rank-Nullity theorem: rank(T)+nullity(T)=dim(V).

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 17 / 107

Question 3(a)

Solution:
The null space of T is defined as
{(x, y , z) : x + y + z = 0, x − y + z = 0, x + z = 0}, which is the
solution space of certain homogeneous system of linear equations,
The null space can be expressed as the solution space of Ax = b, i.e.,

1 1 1 ! x ! 0 !
1 −1 1 y = 0 .
1 0 1 z 0

On solving the above system for (x, y , z), we get,

x + y + z = 0, y = 0, x + z = 0

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 18 / 107

Question 3(a) contd...

Hence, null space is given by {(−a, 0, a), a ∈ R} and (1, 0, −1) is a

basis of Ker(T).
Hence, nullity(T ) = 1.
By Rank-nullity theorem,

Rank(T ) = dim(R3 ) − Nullity (T )

= 3 − 1 = 2.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 19 / 107

Question 3(a) contd...

Let {e1 , e2 , e3 } be the standard basis of R3 . Then, T (R3 ) is

generated by {T (e1 ), T (e2 ), T (e3 )} given by,

T (1, 0, 0) = (1, 1, 1),

T (0, 1, 0) = (1, −1, 0),
T (0, 0, 1) = (1, 1, 1).

1 1 1 !
Consider matrix A = 1 −1 0 which is obtained by arranging
1 1 1
the images of standard basis row-wise.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 20 / 107

Question 3(a) contd...

Hence, the row space of A generates range space of T i.e., basis of

row space of A is a basis of range space.
1 1 1 ! 1 1 1 ! R →R + 1 R
R3 →R3 −R1 1 1 2
1 −1 0 −−−−−−−→ 0 −2 −1 −−−−−−−2−→
R2 →R2 −R1
1 1 1 0 0 0
1 0 1/2 !
R2 → −1 R
1 0 1/2 !
2 2
0 −2 −1 −−−−− −→ 0 1 1/2
0 0 0 0 0 0
Hence, the basis of range space of T is {(1, 0, 21 ), (0, 1, 12 )}.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 21 / 107

Question 3(b)

Question 3(b)
Find the rank and nullity of the following linear transformations. Also
write a basis of the range space in each case.
(b) Assume that 0 ≤ m ≤ n. T : Fn → Fm defined by
T (x1 , x2 , . . . , xn ) = (x1 , x2 , . . . , xm ).

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 22 / 107

Question 3(b)

Solution:
Let {e1 , e2 , . . . , en } be standard basis of Fn and {e10 , e20 , . . . , em
0 } be
m
standard basis of F where m ≤ n.
Case(i). Let m < n.
Then, by definition of T, we have

T (e1 ) = e10 ,
T (e2 ) = e20 ,
···
0
T (em ) = em ,
T (em+1 ) = · · · = T (en ) = 0.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 23 / 107

Question 3(b) contd...

The matrix representation of T is given by,

 
1 0 0 ··· 0 0 ··· 0
0 1
 0 ··· 0 0 ··· 0

A = 0 0
 1 ··· 0 0 ··· 0

 .. .. .. .. .. 
. . . . . 0 ··· 0
0 0 0 ··· 1 0 ··· 0 m×n

Rank(T)=Rank(A)= Number of non-zero rows in row-echelon form.

Hence, Rank(T)=m and using Rank-Nullity theorem nullity(T)=n-m.
Also, {e10 , e20 , . . . , em
0
} is the basis for range space.
Case(ii). Let m = n.
Then, matrix representation of T is given by identity matrix.
Hence, rank(T)=n, nullity(T)=0 and {e10 , e20 , . . . , en0 } is the basis for
range space.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 24 / 107

Question 4

Question 4
Write the matrix representations of the linear operators with respect to the
ordered basis B.
(a) T : R2 → R2 where T (x, y ) = (x, y ), B = {(1, 1), (1, −1)}
(b) D : Pn+1 → Pn+1 such that
D(a0 + a1 x + · · · + an x n ) = a1 + 2a2 x!+ · · · + nan x n−1 , B n
! = {1, x, . . . , x }.
x y x +w z
(c)T : M2 (F) → M2 (F), T = , B=
z w z +w x
 ! ! ! !
1 0 0 1 0 0 0 0
, , ,
0 0 0 0 1 0 0 1

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 25 / 107

Question 4(a)

Question 4(a)
Write the matrix representations of the linear operators with respect to the
ordered basis B.
(a) T : R2 → R2 where T (x, y ) = (x, y ), B = {(1, 1), (1, −1)}

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 26 / 107

Question 4(a)

Solution:
Let B = {(1, 1), (1, −1)} is the ordered basis of R2 and
T (x, y ) = (x, y ).
As,

T (1, 1) = (1, 1) = 1(1, 1) + 0(1, −1)

T (1, −1) = (1, −1) = 0(1, 1) + 1(1, −1)

Hence, the matrix representation of T is given by,

!
1 0
[T ]B =
0 1

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 27 / 107

Question 4(b)

Question 4(b)
Write the matrix representations of the linear operators with respect to the
ordered basis B.
(b) D : Pn → Pn such that D(a0 + a1 x + · · · + an x n )
= a1 + 2a2 x + · · · + nan x n−1 , B = {1, x, . . . , x n }.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 28 / 107

Question 4(b)

Solution:
Let B = {1, x, . . . , x n } is the ordered basis of R2 and
D(a0 + a1 x + · · · + an x n ) = a1 + 2a2 x + · · · + nan x n−1 .
As,

D(1) = 0 = 0.1 + 0.x + · · · + 0.x n

D(x) = 1 = 1.1 + 0.x + · · · + 0.x n
D(x 2 ) = 2x = 0.1 + 2.x + · · · + 0.x n

..
.

D(x n ) = nx n−1 = 0.1 + 0.x + · · · + n.x n−1 + 0.x n .

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 29 / 107

Question 4(b)

Hence, the matrix representation of T is given by,

 
0 1 0 ··· 0
0 0 2
 · · · 0 
[D]B =  ... ... ... .. .. 

 . . 
0 0 0 · · · n
0 0 0 · · · 0 n+1×n+1

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 30 / 107

Question 4(c)

Question 4(c)
Write the matrix representations of the linear operators with respect to the
ordered basis B. ! !
x y x +w z
(c)T : M2 (F) → M2 (F), T = ,
z w z +w x
 ! ! ! !
1 0 0 1 0 0 0 0
B= , , ,
0 0 0 0 1 0 0 1

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 31 / 107

Question 4(c)

Solution:
( ! ! ! !)
1 0 0 1 0 0 0 0
Let B = , , , is the ordered
0 0 0 0 1 0 0 1
! !
x y x +w z
basis and T = .
z w z +w x

! !
1 0 1 0
T =
0 0 0 1
! ! ! !
1 0 0 1 0 0 0 0
=1 +0 +0 +1
0 0 0 0 1 0 0 1

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 32 / 107

Question 4(c)

! !
0 1 0 0
T =
0 0 0 0
! ! ! !
1 0 0 1 0 0 0 0
=0 +0 +0 +0
0 0 0 0 1 0 0 1

! !
0 0 0 1
T =
1 0 1 0
! ! ! !
1 0 0 1 0 0 0 0
=0 +1 +1 +0
0 0 0 0 1 0 0 1

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 33 / 107

Question 4(c)

! !
0 0 1 0
T =
0 1 1 0
! ! ! !
1 0 0 1 0 0 0 0
=1 +0 +1 +0
0 0 0 0 1 0 0 1

 
1 0 0 1
0 0 1 0
Hence, the matrix representation of T is given by, [T ]B = 
0

0 1 1
1 0 0 0

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 34 / 107

Question 5

Question 5
Suppose dimV = dimW < ∞ and T : V → W is a linear transformation.
Show that the following statements are equivalent
(a) T is an isomorphism.
(b) T is injective (i.e., one to one).
(c) kerT = 0.
(d) T is surjective (i.e., onto).

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 35 / 107

Question 5

Recall:
Let T : V → W is a linear transformation. Then
T is injective if its null space is the zero space.
T is surjective if range space=W.
If U is a subspace of V such that dim(U)=dim(V), then U=V.
A linear transformation T is said to be an isomorphism if it is one-one
and onto.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 36 / 107

Question 5

Solution:
(a) =⇒ (b), obvious.
(b) =⇒ (c)
Suppose T is injective.
This implies T (x) = T (y ) =⇒ x = y for x, y ∈ V
If possible, suppose KerT 6= 0, then ∃ a non-zero vector
x ∈ KerT such that T (x) = 0 = T (0) =⇒ x = 0, A
CONTRADICTION.
Hence, KerT = 0.
(c) =⇒ (d)
Suppose KerT = 0, hence, nullity(T)=0.
Since dim(V)=dim(W), by Rank-nullity theorem, we get
rank(T)=dim(V)=dim(W).
Since, range(T) is a subspace of W and dim(range(T))=dim(W),
This implies range space of T is W itself.
Hence, T is surjective.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 37 / 107

Question 5

Solution:
(d) =⇒ (a)
Let T : V → W is a surjective linear transformation and
dim(V)=dim(W).
Hence, rank(T)=dim(W).
By Rank-nullity theorem, we have nullity(T)=0, i.e. ker(T)={0}
Thus for any x,y ∈ V such that T(x)=T(y), we have
T(x) - T(y) = 0
i.e., T(x-y) = 0,
=⇒ x-y ∈ ker(T)={0}
=⇒ x − y = 0, and hence x=y.
This implies, T is one to one.
Hence, T is a one-one, onto linear transformation. Hence, T is an
isomorphism.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 38 / 107

Question 6

Question 6
Suppose m > n. Justify the following statements:
(a) There is no one to one (injective) R-linear transformation from Rm to
Rn .
(b) There is no onto (surjective) R-linear tranformation from Rn to Rm .

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 39 / 107

Question 6(a)

Question 6(a)
Suppose m > n. Justify the following statements:
(a) There is no one to one (injective) R-linear transformation from Rm to Rn .

Solution:
If possible, suppose T : Rm → Rn is a one-one linear transformation and
m > n.
Hence, nullity(T)=0.
By Rank-nullity theorem, we have

Rank(T ) + Nullity (T ) = dim(Rm )

Rank(T ) + 0 = m > n

Since, Range(T) is a subspace of Rn . Hence, rank(T)≤ n.

Hence, the assumption was wrong, there is no one to one R-linear
transformation from Rm to Rn .
(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 40 / 107
Question 6(b)

Question 6(b)
Suppose m > n. Justify the following statements:
(b) There is no onto (surjective) R-linear tranformation from Rn to Rm .

Solution:
If possible, suppose T : Rn → Rm is a onto linear transformation and m > n.
Hence, rank(T)=m.
By Rank-nullity theorem, we have

Rank(T ) + Nullity (T ) = dim(Rn )

m + nullity (T ) = n < m
nullity (T ) = n − m < 0

Since, null-space(T) is a subspace of Rn . Hence, 0 ≤nullity(T)≤ n.

Hence, the assumption was wrong, there is no onto R-linear transformation
from Rn to Rm .
(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 41 / 107
Question 7

Question 7
Find the eigenvalues, eigenvectors and dimension of eigen-spaces of the
following operators.
(a) T : R2 → R2 with T (x, y ) = (x + y , x),
(b) T : R2 → R2 with T (x, y ) = (y , x),
(c) T : R2 → R2 with T (x, y ) = (y , −x)
(d) T : C2 (C) → C2 (C) with T (x, y ) = (y , −x).
(e) T : Cn → Cn with T (x1 , x2 , . . . , xn ) = (xn , x1 , . . . , xn−1 ).
(f) T : C2 → C2 with T (z1 , z2 ) = (z1 − 2z2 , z1 + 2z2 ).

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 42 / 107

Question 7

Recall:
Suppose T : V → V is a linear operator on a vector space V .
A scalar λ is said to be an eigenvalue of T if there is a nonzero vector
v ∈ V such that T (v ) = λv and v is called an eigenvector of T
associated to the eigenvalue λ.
Collection of all eigen-vectors v corresponding to λ, along with the 0
vector, is called the eigen space of λ.
Eigen-values of T are the root of the characteristic polynomial
|λI − A|, where A is the matrix representation of T with respect to
the standard basis.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 43 / 107

Question 7(a)

Question 7(a)
Find the eigenvalues, eigenvectors and dimension of eigen-spaces of the
following operator. T : R2 → R2 with T (x; y ) = (x + y , x),

Solution:
Consider the standard ordered basis B = {(1, 0), (0, 1)} of R2 .
!
1 1
The matrix representation of T with respect of B is A = .
1 0
The characteristic polynomial of A is given by det(λI − A) = 0,

λ − 1 −1
= 0 =⇒ λ2 − λ − 1 = 0
−1 λ − 0
√
1± 5
=⇒ λ =
2

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 44 / 107

Question 7(a) contd...

√
1± 5
Hence, λ = 2 are the eigen-values of T.
If X is eigen vector corresponding to eigen value λ, then

AX = λX
(A − λI )X = 0
! ! !
1−λ 1 x 0
=⇒ = .
1 −λ y 0
√
1+ 5
For λ1 = 2 , the eigen-vector X1 is the non-zero solution of
√ ! ! !
1− 5
2 1 √
x 0
= .
1 − 1+2 5 y 0

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 45 / 107

Question 7(a) contd...
√ !
1+ 5
On solving for x, y , we get X1 = .
2
Hence, eigen space Wλ1 = {aX1 ; a ∈ R}. Hence, dimension of eigen
space for λ1 is 1.
√
1− 5
For λ2 = 2 , the eigen-vector X2 is the solution of
√ ! ! !
1+ 5
2 1 √
x 0
= .
1 − 1−2 5 y 0

√ !
1− 5
On solving for x, y , we get X2 = .
2
Hence, eigen space Wλ2 = {bX2 ; b ∈ R}. Hence, dimension of eigen
space for λ2 is 1.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 46 / 107

Question 7(b)
Question 7(b)
Find the eigenvalues, eigenvectors and dimension of eigen-spaces of the following
operators. (b) T : R2 → R2 with T (x; y ) = (y , x),

Solution:
Consider the standard basis B = {(1, 0), (0, 1)} for R2 .
!
0 1
The matrix representation of T with respect to B is A = .
1 0

Solving det(λI − A) = 0, we get the characteristic polynomial λ2 − 1 = 0.

Hence, λ = ±1 are the eigen values of T.
If X is eigen vector corresponding to eigen value λ, then
(A − λI )X = 0
! ! !
−λ 1 x 0
=⇒ = .
1 −λ y 0

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 47 / 107

Question 7(b) contd...

For λ = 1, the eigen-vector X1 is the solution of

! ! !
−1 1 x 0
= .
1 −1 y 0

−x + y = 0, x − y = 0
=⇒ x − y = 0 =⇒ x = y .
!
1
Hence, X1 = is eigen vector corresponding to eigen value λ = 1.
1

Hence, eigen space Wλ1 = {aX1 ; a ∈ R}. Hence, dimension of eigen space
for λ1 is 1.
!
1
Similarly, the eigen-vector corresponding to λ2 = −1 is X2 = .
−1

Hence, eigen space Wλ2 = {bX2 ; b ∈ R}. Hence, dimension of eigen space
for λ2 is 1.
(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 48 / 107
Question 7(c)

Question 7(c)
Find the eigenvalues, eigenvectors and dimension of eigen-spaces of the
following operators.
(c) T : R2 → R2 with T (x; y ) = (y , −x)

Solution:
The matrix representation
! of T with respect of standard basis B is
0 1
A= .
−1 0
Hence, the characteristic polynomial is λ2 + 1 = 0.
Above polynomial does not have real root. Hence, the linear
transformation does not have real eigen values.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 49 / 107

Question 7(d)

Question 7(d)
Find the eigenvalues, eigenvectors and dimension of eigen-spaces of the
following operators. (d) T : C2 (C) → C2 (C) with T (x; y ) = (y , −x).

Solution:
Consider the standard basis B = (1, 0), (0, 1) of C2 (C).
Then, matrix representation
! of T with respect to standard basis is
0 1
A= .
−1 0
The characteristic polynomial of T is λ2 + 1 = 0.
Hence, the eigen values of T are λ = ±i.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 50 / 107

Question 7(d) contd...
For λ = i, the eigen-vector X1 is given by
(A − iI )X1 = 0
! ! !
−i 1 x 0
=
−1 −i y 0
=⇒ −ix + y = 0, −x − iy = 0
=⇒ ix = y
!
1
Hence eigen vector X1 = .
i
Hence, eigen space Wλ1 = {aX1 ; a ∈ C}. Hence, dimension of eigen space
for λ1 is 1.
!
1
Similarly, the eigen-vector corresponding to λ2 = −i is X2 = .
−i
Hence, eigen space Wλ2 = {bX2 ; b ∈ C}. Hence, dimension of eigen space
for λ2 is 1.
(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 51 / 107
Question 7(e)

Question 7(e)
T : Cn → Cn with T (x1 , x2 , . . . , xn ) = (xn , x1 , . . . , xn−1 ).

Solution:
The standard basis of Cn (C) is given by
B = {(1, 0, . . . , 0), (0, 1, . . . , 0), . . . , (0, 0, . . . , 1)}.
 of T with respect to B is
Then,matrix representation
0 0 ··· 0 1
1 0 · · · 0 0 
 
A = 0 1 · · · 0 0.
 
 .. .. . .
 . . · · · .. .. 
0 0 ··· 1 0
(This type of matrix is known as companion matrix)

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 52 / 107

Question 7(e)

The characteristic polynomial is given by

−λ 0 · · · 0 1
1 −λ · · · 0 0
det(T − λI ) = 0 1 ··· 0 0 =0
.. .. .. ..
. . ··· . .
0 0 ··· 1 −λ
λn − 1 = 0.
Hence, the eigen values of T are the n−th root of unity.
Note that, λn = 1 for any eigenvalue λ of the above matrix.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 53 / 107

Question 7(e)

Let X = (x1 , x2 , . . . , xn )t be an eigen-vector for the general eigenvalue

λ.
Then X 6= 0, i.e., atleast one of xi is non-zero. For simplicity, we are
assuming x1 6= 0 (as each λ is non-zero, we can divide by λ in case of
need and can go ahead with the similar proof)
and TX = λX . On comparing both sides we get

xn = λx1 , x1 = λx2 , x2 = λx3 , . . . , xn−1 = λxn .

Writing each variable in terms of x1 to get

xn = λx1 , xn−1 = λ2 x1 , xn−2 = λ3 x1 , . . . ,

x2 = xn−(n−2) = λn−2+1 x1 = λn−1 x1 , x1 = λn x1 = 1x1 = x1 .

Thus, X = x1 (λn , λn−1 , . . . , λ2 , λ)t is an eigenvector corresponding to eigen
value λ. (as x1 6= 0, we can divide by x1 to get a vector free from x1 )

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 54 / 107

Question 7(f)

Question 7(f)
Find the eigenvalues, eigenvectors and dimension of eigen-spaces of the
following operators. (f) T : C2 → C2 with T (z1 , z2 ) = (z1 − 2z2 , z1 + 2z2 ).

Solution:
! of T with respect to B is
The matrix representation
1 −2
=⇒ [T ] = .
1 2
Hence, the eigen values of T are given by
1 − λ −2
det(λI − A) = =0
1 2−λ
λ2 − 3λ + 4 = 0
√
3±i 7
=⇒ λ =
2
.
(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 55 / 107
Question 7(f) contd...

√
3+i 7
For λ1 = 2 , the eigen-vector X1 is given by,

(A − λ1 I )X1 = 0
√ ! ! !
− 1+i2 7 −2√
x 0
=
1 1−i 7 y 0
2
√
1+i 7
=⇒ − x − 2y = 0.
2
!
−4√
Hence, eigen-vector X1 = .
1+i 7

Hence, eigen space Wλ1 = {aX1 ; a ∈ C}. Hence, dimension of eigen space
for λ1 is 1.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 56 / 107

Question 7(f)

√
3−i 7
Similarly, the eigen-vector corresponding to λ2 = 2 is
√ !
1+i 7
X2 = .
−2
Hence, eigen space Wλ2 = {bX2 ; b ∈ C}. Hence, dimension of eigen
space for λ2 is 1.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 57 / 107

Question 8

Question 8
Find a basis B such that [T ]B is a diagonal matrix in case T is
diagonalizable. Find P such that [T ]B = P[T ]S P −1 where S is the
standard basis in each case.
(a) T : C2 → C2 defined by T (x, y ) = (y , −x).
(b) T : C3 → C3 defined by
T (x, y , z) = (5x − 6y − 6z, −x + 4y + 2z, 3x − 6y − 4z).
(c) T : C2 → C2 defined by T (x, y ) = (xcosθ + ysinθ, −xsinθ + ycosθ).

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 58 / 107

Question 8

Recall:
A linear operator T : V → V is diagonalizable (over F)
iff V has a basis B with respect to which the matrix of T is diagonal.
if and only if dimV is equal to the sum of the dimensions of the eigen
spaces of T .
if all eigen values are distinct.

Let T is diagonalizable.
Let B is the collection of distinct eigen-vector corresponding to
different eigen values.
Then, [T ]B = P[T ]S P −1 where S is the standard basis and P −1 is
the change of basis matrix from S to B.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 59 / 107

Question 8(a)

Question 8(a)
T : C2 → C2 defined by T (x, y ) = (y , −x).

Solution:
The matrix representation
! of T with respect to standard basis S is,
0 1
[T ]S = .
−1 0
The characteristic polynomial of [T ]S is, λ2 + 1 = 0, it has distinct
roots λ = ±i.
Hence, T is diagonalizable.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 60 / 107

Question 8(a) contd...

Define, B = {(1, i), (1, −i)} the collection of distinct eigen-vectors.

Then matrix representation of T with respect to B is,

T (1, i) = (i, −1) = a(1, i) + b(1, −i)

T (1, −i) = (−i, −1) = c(1, i) + d(1, −i)

!
i 0
=⇒ [T ]B = .
0 −i

The matrix P −1 is the change of basis matrix, given by

(1, i) = 1.(1, 0) + i.(0, 1)

(1, −i) = 1.(1, 0) − i(0, 1)
!
1 1
Hence, P −1 = .
i −i

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 61 / 107

Question 8(a) contd...

Verification:
!t ! !
1 1
0 1 1 1
P[T ]S P −1 = 2
−i
2
i
2 2 −1 0 i −i
!
i 0
= = [T ]B
0 −i

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 62 / 107

Question 8(b)

Question 8(b)
T : C3 → C3 defined by T (x, y , z) = (5x − 6y − 6z, −x + 4y + 2z, 3x − 6y − 4z).

Solution:
The matrix representation of T with respect to standard basis S is,
5 −6 −6
!
[T ]S = −1 4 2 .
3 −6 −4
The characteristic polynomial of [T ]S is, (λ − 2)2 (λ − 1) = 0.
The eigen vector corresponding to λ = 1 is given by,

AX = X
5 −6 −6 x x
! ! !
−1 4 2 y = y
3 −6 −4 z z

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 63 / 107

Question 8(b) contd...

Solving the above system for x, y , x,

4 −6 −6 ! 0 6 2 ! R →R + −1 R
R3 →R3 +3R2 3 3 3 1
−1 3 2 −
−−−−−−− → −1 3 2 −−−−−−−− −→
R1 →R1 +4R2
3 −6 −5 0 3 1

0 6 2 ! R →1R 0 3 1 ! −1 3 2 !
1 1 R1 →R2
−1 3 2 −−−−2−→ −1 3 2 −− −−→ 0 3 1
0 0 0 0 0 0 0 0 0
We get, x − 3y − 2z = 0 3y + z = 0
Hence, The eigen space correponding to λ = 1 is spanned by
h(3, −1, 3)i.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 64 / 107

Question 8(b) contd...

The eigen vector corresponding to λ = 2 is given by,

(A − 2I )X = 0

3 −6 −6 ! x ! 0 !
−1 2 2 y = 0
3 −6 −6 z 0
Solving the above system for x, y , x, we get −x + 2y + 2z = 0.
The eigen space correponding to λ = 2 is spanned by
h(2, 1, 0), (2, 0, 1)i.
Hence, dim(eigen-spaces)= 3 =dim(V). So, V is diagonalizable.
Define B = {(3, −1, 3), (2, 1, 0), (2, 0, 1)}. Then, [T ]B is diagonal
3 2 2 !−1
matrix and change of basis matrix P −1 = −1 1 0 .
3 0 1
(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 65 / 107
Question 8(c)

Question 8(c)
T : C2 → C2 defined by T (x, y ) = (xcosθ + ysinθ, −xsinθ + ycosθ).

Solution:
The matrix representation!of T with respect to standard basis S is,
cos θ sin θ
[T ]S = .
− sin θ cos θ
The characteristic polynomial of [T ]S is,

λ2 − 2λ cos θ + 1) = 0
=⇒ λ = cos θ ± i sin θ

Since, both the eigen values are distinct. Hence, T is diagonalizable.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 66 / 107

Question 8(c) contd...

The eigen vector corresponding to λ = cos θ + i sin θ is given by,

−i sin θ sin θ
det(A − λI ) = =0
− sin θ −i sin θ

=⇒ −i sin θx + sin θy = 0
Hence, the eigen space is spanned by h(1, i)i.
The eigen vector corresponding to λ = cos θ − i sin θ is given by,

i sin θ sin θ
det(A − λI ) = =0
− sin θ i sin θ

=⇒ ix + y = 0
Hence, the eigen space is spanned by h(1, −i)i.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 67 / 107

Question 8(c) contd...

Define B = {(1, i), (1, −i)}. Then, [T ]B is diagonal matrix and

!−1
1 1
change of basis matrix P −1 = .
i −i

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 68 / 107

Question 9
Characteristic polynomial of a matrix is satisfied by the matrix(Cayley
Hamilton). Use it to find(invertibility and) the inverse of the following
operators.
(a) (x, y , z) → (x + y + z, x + z, −x + y ).
(b) (x, y , z) → (x, x + 2y , x + 2y + 3z).

Recall:
If p(x) is the Characteristic polynomial of a matrix A then p(0) =
det(A)
A square matrix A is invertible iff det(A) 6= 0

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 69 / 107

Question 9(a)
Solution:
Let T denotes the given linear operator.
Then, the matrix representation of T with respect to the standard ordered
basis is,
 
1 1 1
A= 1 0 1
−1 1 0

Let p(X ) = det(XI − A) is the characteristic polynomial of A.

Here
p(X ) = (X − 1)2 (X + 1) = X 3 − X 2 − X + 1
Since p(0) = 1 6= 0, hence, A is invertible.
Thus, by Cayley-Hamilton theorem,

A3 − A2 − A + I = O =⇒ A2 − A − I + A−1 = O

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 70 / 107

Question 9(a) contd...

Hence

A−1 = I + A − A2
     
1 0 0 1 1 1 1 2 2
= 0 1 0  +  1 0 1  − 0 2 1 
0 0 1 −1 1 0 0 −1 0
 
1 −1 −1
=  1 −1 0 
−1 2 1

which is matrix representation of inverse of given operator with

respect to the standard ordered basis.
Hence, T −1 (x, y , z) = (x − y − z, x − y , −x + 2y + z).

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 71 / 107

Question 9(b)

Question 9(b)
(b) (x, y , z) → (x, x + 2y , x + 2y + 3z).

Solution:
Let T denotes the given linear operator.
Then, the matrix representation of T with respect to the standard
ordered basis is,
 
1 0 0
A = 1 2 0
1 2 3
.
Let p(X ) = det(XI − A) is the characteristic polynomial of A.
Here, p(X ) = (X − 1)(X − 2)(X − 3) = X 3 − 6X 2 + 11X − 6, since,
p(0) = −6 6= 0, hence, A is invertible.
Thus, by Cayley-Hamilton theorem, A3 − 6A2 + 11A − 6I = 0.
(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 72 / 107
Question 9(b) contd...

Hence,
1
A−1 = (A2 − 6A + 11I )
6      
1 0 0 6 0 0 11 0 0
1 
= 3 4 0 − 6 12 0  −  0 11 0 
6
6 10 9 6 12 18 0 0 11
 
1 0 0
= − 12 1
2 0
0 − 3 131

which is matrix representation of inverse of given operator with

respect to standard ordered basis.
Hence, T −1 (x, y , z) = (x, y −x z−y
2 , 3 ).

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 73 / 107

Question 10
Which of the following is an inner product.
(a) h(x1 , y1 ), (x2 , y2 )i = x1 x2 + y1 y2 + 3 on R2 over R.
(b) h(x1 , y1 ), (x2 , y2 )i = x1 x2 − y1 y2 on R2 over R.
(c) h(x1 , y1 ), (x2 , y2 )i = y1 (x1 + 2x2 ) + y2 (2x1 + 5x2 ) on R2 over R.
(d) h(x1 , y1 ), (x2 , y2 )i = x1 x2 + y1 y2 on C2 over C.
(e) h(x1 , y1 ), (x2 , y2 )i = x1 x¯2 − y1 y¯2 on C2 over C.
(f) If A, B ∈ Mn (C ) define hA, Bi = Trace(AB̄).
(g) Suppose C [0, 1] is the space of continuous complex valued functions
R1
on the interval [0, 1] and for f , g ∈ C [0, 1], hf , g i := 0 f (t)g (t) dt.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 74 / 107

Question 10

Recall
Let V be a vector space over F (where F = R or C). A map V × V → F
denoted by (u, v ) → hu, v i is called an inner product on V if the following
properties hold:
(a) hu, ui ∈ F and ≥ 0 for each u ∈ V ;
(b) hu, ui = 0 if and only if u = 0;
(c) hau + bv , w i = ahu, w i + bhv , w i;
(d) hu, v i = hv , ui (the complex conjugate).
A vector space together with an inner product is called an inner product
space.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 75 / 107

Question 10

Question 10(a)
Which of the following is an inner product.
(a) h(x1 , y1 ), (x2 , y2 )i = x1 x2 + y1 y2 + 3 on R2 over R.

Solution:
For u = (0, 0), hu, ui = 3 6= 0.
Hence, it is not an inner product.

Question 10(b)
h(x1 , y1 ), (x2 , y2 )i = x1 x2 − y1 y2 on R2 over R.

Solution:
For u = (1, −1), hu, ui = h(1, −1), (1, −1)i = 1.1 − (−1)(−1) = 0.
Hence, it is not an inner product.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 76 / 107

Question 10

Question 10(c)
h(x1 , y1 ), (x2 , y2 )i = y1 (x1 + 2x2 ) + y2 (2x1 + 5x2 ) on R2 over R.

Solution:
For u = (1, −1),
hu, ui = h(1, −1), (1, −1)i = (−1).(3) + (−1)(7) = −10  0.
Hence, it is not an inner product.

Question 10(d)
h(x1 , y1 ), (x2 , y2 )i = x1 x2 + y1 y2 on C2 over C.

Solution:
For u = (i, 0), hu, ui = h(i, 0), (i, 0)i = i.i + 0 = −1  0.
Hence, it is not an inner product.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 77 / 107

Question 10

Question 10(e)
h(x1 , y1 ), (x2 , y2 )i = x1 x¯2 − y1 y¯2 on C2 over C.

Solution:
For u = (i, i), hu, ui = h(i, i), (i, i)i = i.(−i) − i(−i) = 0.
Hence, it is not an inner product.

Question 10(f)
If A, B ∈ Mn (C ) define hA, Bi = Trace(AB̄).

Solution:  
1 −1
For A =
1 −1
 
0 0
hA, Ai=Trace(AĀ)=Trace = 0.
0 0
Hence, it is not an inner product.
(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 78 / 107
Question 10(g)

Question 10(g)
Suppose C [0, 1] is the space of continuous complex valued functions on
R1
the interval [0, 1] and for f , g ∈ C [0, 1], hf , g i := 0 f (t)g (t) dt.

Solution:
Let f (t), g (t) ∈ C [0, 1], then,
Z 1
hg (t), f (t)i : = g (t)f (t) dt
0
Z 1
= g (t)f (t) dt
0
Z 1
= g (t)f (t) dt
0
Z 1
= f (t)g (t) dt = hf (t), g (t)i.
0

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 79 / 107

Question 10(g) contd...

Z 1 Z 1
hf (t), f (t)i = f (t)f (t) dt = |f (t)|2 dt ≥ 0
0 0
2
and hf (t), f (t)i = 0 iff |f (t)| = 0 iff f (t) = 0

For a, b ∈ C and f (t), g (t), h(t) ∈ C [0, 1]

Z 1
haf (t) + bg (t), h(t)i = [af (t) + bg (t)]h(t) dt
0
Z 1 Z 1
=a f (t)h(t) dt + b g (t)h(t) dt
0 0
= ahf (t), h(t)i + bhg (t), h(t)i

Hence, V is an inner product space.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 80 / 107

Question 11

a b
Suppose A = ∈ M2 (R) is such that a > 0 and
b d
det(A) = ad − b 2 > 0. Show that hX , Y i = X t AY is an inner product on
R2 .

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 81 / 107

Question 11
Solution:
Let X , Y ∈ R2 , then,
hY , X i = Y t AX = (Y t AX )t as Y t AX ∈ R
= X t At Y = X t AY = hX , Y i.
 
x
Let X = 1 6= 0
x2
hX , X i = X t AX = ax1 2 + 2bx1 x2 + dx2 2
2 
ad − b 2
 
b
= a x1 + x2 + x2 2 > 0
a a
For m, n ∈ R
hmX + nY , Z i = (mX + nY )t AZ
= m(X t AZ ) + n(Y t AZ )
= mhX , Z i + nhY , Z i
Hence, it is an inner product space.
(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 82 / 107
Question 12
p
Suppose V is an inner product space. Define kv k = hv , v i. Show the
following statements.
(a)kv k = 0 if and only if v = 0.
(b) For a ∈ F , kav k = |a|kv k.
(c) ku + v k ≤ kuk + kv k.
(d)|kv k − kw k| ≤ kv − w k.
(e) hu, v i = 0 then ku + v k2 = kuk2 + kv k2 .

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 83 / 107

Question 12
Solution: (a) If v = 0 then clearly kv k = 0. If kv k = 0 implies that
hv , v i = 0 which further implies that v = 0.
(b) kav k2 = hav , av i = ahv , av i = aāhv , v i = |a|2 kv k2

=⇒ kav k2 = |a|2 kv k2

Taking square root both sides, kav k = |a|kv k.

(c)ku + v k2 = hu + v , u + v i = hu, u + v i + hv , u + v i
= hu, ui + hu, v i + hv , ui + hv , v i
= kuk2 + hu, v i + hu, v i + kv k2
= kuk2 + 2Rehu, v i + kv k2
≤ kuk2 + 2|hu, v i| + kv k2
≤ kuk2 + 2kukkv k + kv k2 . (By Cauchy Schwarz Inequality)

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 84 / 107

Question 12 contd...

=⇒ ku + v k2 ≤ (kuk + kv k)2
Taking square root, ku + v k ≤ kuk + kv k
(d) Consider kv k = kv − w + w k ≤ kv − w k + kw k
=⇒ kv k − kw k = kv − w k (1)
Similarly, kw k = kw − v + v k = k(−1)(v − w ) + v k ≤ kv − w k + kv k
=⇒ kw k − kv k = kv − w k (2)
From(1) and (2), |kv k − kw k| ≤ kv − w k.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 85 / 107

Question 12 Contd...

(e) Given that hu, v i = 0 implies that hv , ui = 0. Now,

ku + v k2 = hu + v , u + v i
= hu, u + v i + hv , u + v i
= hu, ui + hu, v i + hv , ui + hv , v i
= kuk2 + kv k2

=⇒ ku + v k2 = kuk2 + kv k2 .

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 86 / 107

Question 13
Use standard inner product on R2 over R to prove the following
statement: ”A parallelogram is a rhombus if and only if its diagonals are
perpendicular to each other.”

Recall:
The vectors u, v ∈ R2 are said to be perpendicular (orthogonal) if and only
if hu, v i = 0.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 87 / 107

Question 13

Solution:
Let u, v are adjacent sides of parallelogram, then, u + v and u − v
represent diagonals of parallelogram.
Consider hu + v , u − v i = hu, ui − hu, v i + hv , ui − hv , v i.
Since, the filed is R, we have hu, v i = hv , ui. Hence,

hu + v , u − v i = kuk2 − kv k2 = 0 if and only if kuk = kv k.

Hence, A parallelogram is a rhombus if and only if its diagonals are

perpendicular to each other.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 88 / 107

Question 14
Find with respect to the standard inner product of R3 , an orthonormal
basis containing (1, 1, 1).

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 89 / 107

Question 14
Solution:
Consider a basis B = {(1, 1, 1), (1, 0, 0), (0, 1, 0)} of R3 containing (1, 1, 1).
Denote u1 = (1, 1, 1), u2 = (1, 0, 0), u3 = (0, 1, 0).
u1
Now, using Gram Schmidt process, we have, v1 = ku1 k

2
ku1 k = hu1 , u1 i = 1.1 + 1.1 + 1.1 = 3
 
Hence, v1 = √1 , √1 , √1 .
3 3 3

Let

w2 = u2 − hu2 , v1 iv1
   
1 1 1 1 1 1
= (1, 0, 0) − h(1, 0, 0), √ ,√ ,√ i √ ,√ ,√
3 3 3 3 3 3
 
2 1 1
= ,− ,− .
3 3 3

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 90 / 107

Question 14 contd...

s      
2 2 2 1 1 1 1
kw2 k = . + − − + − −
3 3 3 3 3 3
√
2
=⇒ kw2 k = √
3
√ 
w2
Hence, v2 = kw2 k = √23 , − √16 , − √16 .
Let

w3 = u3 − hu3 , v1 iv1 − hu3 , v2 iv2

   
1 1 1 1 1 1
= (0, 1, 0) − h(0, 1, 0), √ , √ , √ i √ , √ , √
3 3 3 3 3 3
√ ! √ !
2 1 1 2 1 1
− h(0, 1, 0), √ , − √ , − √ i √ , − √ , − √ .
3 6 6 3 6 6

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 91 / 107

Question 14 contd...

     
1 1 1 1 1 1 1 1
= (0, 1, 0) − , , + ,− ,− = 0, , − .
3 3 3 3 6 6 2 2
     
2 1 1 1 1 1
kw3 k = 0.0 + + − − =
2 2 2 2 8
1
=⇒ kw3 k = √ .
2 2
 
w3 1 1
Hence, v3 = kw3 k = 0, 2 , − 2 .
√ √

Hence the required orthonormal basis is

(  √ !  )
1 1 1 2 1 1 1 1
√ ,√ ,√ , √ , −√ , −√ , 0, √ , − √
3 3 3 3 6 6 2 2

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 92 / 107

Question 15
Find an orthonormal basis of P3 = {f (x) ∈ R[x] : deg f (x) < 3} with
R1
respect to the inner product defined by hf , g i := 0 f (t)g (t) dt.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 93 / 107

Question 15
Solution:
Let B = {1, x, x 2 } is the basis of Pn .
Denote u1 = 1, u2 = x, u3 = x 2 .
Using Gram Schmidt process, we have
u1
v1 =
ku1 k
2 R1
ku1 k = hu1 , u1 i = 0
1.1 dt = 1,
Hence, v1 = 1.
Let

w2 = u2 − hu2 , v1 iv1
= x − hx, 1i1
Z 1
1
=x− t.1 dt = x − .
0 2

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 94 / 107

Question 15 contd...

Z 1
2 1 1
kw2 k = (t − )(t − ) dt
0 2 2
Z 1
1 1
= (t 2 + − t) dt = .
0 4 12
w2
√
Hence, v2 = kw2 k = 12(x − 21 ).
Let
w3 = w3 − hw3 , v1 iv1 − hw3 , v2 iv2
√ 1 √ 1
= x 2 − hx 2 , 1i1 − hx 2 , 12(x − )i 12(x − )
2 2
Z 1 Z 1
1 1
= x2 − t 2 dt − 12(x − ) t 2 (t − ) dt
0 2 0 2
1
= x2 − x +
6
(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 95 / 107
Question 15 contd...

Z 1
1 1
kw3 k2 =(t 2 − t + )(t 2 − t + ) dt
0 6 6
Z 1
4 1 1 1
= (t 4 − 2t 3 + t 2 − t + ) dt = .
0 3 3 36 180
w3
√
Hence, v3 = kw 3k
= 180(x 2 − x + 16 ).
Hence, the required orthonormal basis is
√ 1 √
 
1
1, 12(x − ), 180(x 2 − x + )
2 6

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 96 / 107

Question 16
Suppose W is a subspace of the finite dimensional inner product space.
Define W ⊥ := {v ∈ V : hw , v i = 0 for all w ∈ W }. Show the following
statements.
(a) W ⊥ is a subspace of V.
(b) W ∩ W ⊥ = 0.
(c) V = W ⊕ W ⊥ .
(d) (W ⊥ )⊥ = W .

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 97 / 107

Question 16(a)

Question 16(a)
W ⊥ is a subspace of V.

Solution:
Since, h0, w i = 0 ∀ w ∈ W . Hence, 0 ∈ W ⊥ =⇒ W ⊥ 6= φ.
Let v1 , v2 ∈ W ⊥ and a, b ∈ F .
Then, for w ∈ W ,

hav1 + bv2 , w i = ahv1 , w i + bhv2 , w i = a.0 + b.0 = 0.

This implies, av1 + bv2 ∈ W ⊥ .

Hence, W ⊥ is a subspace of V .

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 98 / 107

Question 16(b)

Question 16(b)
W ∩ W ⊥ = 0.
Solution:
Let u ∈ W ∩ W ⊥ .
This implies u ∈ W and hence, hu, v i = 0 ∀ v ∈ W ⊥ .
Also, u ∈ W ⊥ and hence, hu, w i = 0 ∀ w ∈ W .
i.e., hu, w i = 0 ∀ w ∈ WUW ⊥ .
Thus, hu, ui = 0, and hence u = 0.
Also note that 0 ∈ W ∩ W ⊥ . Hence, W ∩ W ⊥ = {0}.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 99 / 107

Question 16(c)

Question 16(c)
V = W ⊕ W ⊥.
Solution:
Let V is a finite dimensional vector space and W is subspace of V , hence,
W is finite dimensional.
Let dim(W ) = m.
Let B = {u1 , u2 , ..um } be an orthonormal basis of W , such that
(
0 i 6= j
hui , uj i =
1 i =j

Let v ∈ V and consider

n
X
w =v− hv , ui iui , (3)
i=1

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 100 / 107
Question 16(c) contd...

then, for k = 1, 2, ..., n

n
X
hw , uk i = hv − hv , ui iui , uk i
i=1
n
X
= hv , uk i − hv , ui ihui , uk i
i=1
= hv , uk i − hv , uk i = 0

Hence, w is orthogonal to each basis vector of basis of W, this implies,

w ∈ W ⊥.
Now, (3) can be written as
n
X
v =− hv , ui iui + w
i=1

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 101 / 107
Question 16(c) contd...

This implies, v ∈ W + W ⊥ . Hence, V ⊆ W + W ⊥ , but

W + W ⊥ ⊆ V , hence, V = W + W ⊥ .
On the other hand, W ∩ W ⊥ = {0}.
Hence, V = W ⊕ W ⊥ .

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 102 / 107
Question 16(d)

Question 16(d)
(W ⊥ )⊥ = W .

Solution:
Let w ∈ W , then hw , v i = 0 ∀ v ∈ W ⊥ .
Hence, w ∈ (W ⊥ )⊥ , =⇒ W ⊆ (W ⊥ )⊥ .
Since, V = W ⊕ W ⊥ =⇒ dimV = dimW + dimW ⊥
Replacing W ⊥ instead of W in above equation, we get

V = W ⊥ ⊕ (W ⊥ )⊥ =⇒ dimV = dimW ⊥ + dim(W ⊥ )⊥ .

Hence, dimW = dim(W ⊥ )⊥ =⇒ W = (W ⊥ )⊥ .

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 103 / 107
Question 17
Suppose W = {(x, y ) ∈ R2 : x + y = 0}. Find the shortest distance of
(a, b) ∈ R2 from W with respect to (i) the standard inner product, (ii) the
inner product defined by h(x1 , y1 ), (x2 , y2 )i = 2x1 x2 + y1 y2 .

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 104 / 107
Question 17

Recall:
Suppose V is an inner product space and W is a proper subspace of
V. Given v ∈ V , a vector w0 ∈ W is said to be a best approximation
of v if for each w ∈ W we have

||v − w0 || ≤ ||v − w ||.

Let W be a finite dimensional subspace of V. Suppose

{w1 , w2 , . . . , wn } be an orthogonal basis of W . Then the best
approximation of v ∈ V is given by
n
X hv , wi i
w0 = wi .
||wi ||2
i=1

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 105 / 107
Question 17(i)

Question 17(i)
Suppose W = {(x, y ) ∈ R2 : x + y = 0}. Find the shortest distance of
(a, b) ∈ R2 from W with respect to (i) the standard inner product.

Solution:
Consider an orthogonal basis B = {1, −1} of W .
Let v = (a, b) ∈ R2 .
Then, the best approximation of v is,
n
X hv , wi i h(a, b), (1, −1)i
w0 = 2 wi = 2 (1, −1)
i=1 kwi k k(1, −1)k
a.1 + b.(−1) a−b b−a
= (1, −1) = ( , ).
2 2 2
Hence, the shortest distance between (a, b) and W is,
(a+b)
kv − w0 k = k(a, b) − ( a−b b−a a+b a+b
2 , 2 )k = k( 2 , 2 )k =
√ .
2

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 106 / 107
Question 17(ii)

Question 17(ii)
The inner product defined by h(x1 , y1 ), (x2 , y2 )i = 2x1 x2 + y1 y2 .

Solution:
Consider an orthogonal basis B = {1, −1} of W .
Let v = (a, b) ∈ R2 .
Then, the best approximation of v is,
n
X hv , wi i h(a, b), (1, −1)i
w0 = 2 wi = 2 (1, −1)
i=1 kwi k k(1, −1)k
2a.1 + b.(−1) 2a − b b − 2a
= (1, −1) = ( , ).
2.1.1 + (−1)(−1) 3 3

Hence, the shortest distance between (a, b) and W is, q

√
6(a+b)2 2(a+b)
kv − w0 k = k(a, b) − ( 2a−b b−2a a+b 2b+2a
3 , 3 )k = k( 3 , 3 )k = 9 = √
3
.

(Maths Dept., IIT Delhi) MTL101, Tutorial-3 Semester-II, 2020-21 107 / 107