Solutions for Math 225 Assignment #4 1

(1) Let T : R3 → R2 be the linear transformation given by

T (x, y, z) = (x + y + z, x − z).
Find the matrix [T ]B1 ,B2 representing T for
(a) B1 and B2 the standard bases of R3 and R2

Solution. Since
T (e1 ) = (1, 1) = 1 · e1 + 1 · e2
T (e2 ) = (1, 0) = 1 · e1 + 0 · e2
T (e3 ) = (1, −1) = 1 · e1 + (−1) · e2
we obtain
 
1 1 1
[T ]B1 ,B2 = .
1 0 −1

(b) B1 = {(1, 1, 0), (1, 0, 1), (0, 1, 1)} and B2 = {(3, 4), (4, 5)}.

Solution. For every v = (x, y) ∈ R2 ,

3 4 x0
          
x 0 3 0 4 x
=x +y ⇔ =
y 4 5 y 4 5 y0
 0  −1  
x 3 4 x
⇔ 0 =
y 4 5 y
 0   
x −5 4 x
⇔ 0 =
y 4 −3 y
Then
 
1     −1     
2 3 4 3 4 2 3 4 −6
T 1 = = =
1 4 5 4 5 1 4 5 5
0
 
1     −1     
2 3 4 3 4 2 3 4 −10
T 0 = = =
0 4 5 4 5 0 4 5 8
1
 
0     −1     
2 3 4 3 4 2 3 4 −14
T 1 = = =
−1 4 5 4 5 −1 4 5 11
1
1
http://www.math.ualberta.ca/˜xichen/math22514w/hw4sol.pdf
1
2

Therefore,
 
−6 −10 −14
[T ]B1 ,B2 =
5 8 11


(2) Let T : R3 → R3 be the linear transformation satisfying

T (e1 ) = e2 , T (e2 ) = e3
and K(T ) = Span{(1, 1, 1)}, where K(T ) is the kernel of T .
(a) Find the matrix [T ]B,B representing T under the standard
basis B of R3 .
(b) Find the range R(T ) of T .

Solution. Since (1, 1, 1) ∈ K(T ), T (1, 1, 1) = 0 and hence

T (e3 ) = T ((1, 1, 1) − e1 − e2 )
= T (1, 1, 1) − T (e1 ) − T (e2 ) = −e2 − e3 .
Therefore,
T (e1 ) = 0 · e1 + 1 · e2 + 0 · e3
T (e2 ) = 0 · e1 + 0 · e2 + 1 · e3
T (e3 ) = 0 · e1 + (−1) · e2 + (−1) · e3
and  
0 0 0
[T ]B,B = 1 0 −1 .
0 1 −1
And
R(T ) = Span{T (e1 ), T (e2 ), T (1, 1, 1)} = Span{e2 , e3 }
= {(x, y, z) : x = 0}.

(3) Let V = {f (x) ∈ R[x] : deg f < 3} be the vector space of
real polynomials in x of degree < 3 and let T1 : V → V and
T2 : V → V be two linear transformations given by
T1 (f (x)) = xf 0 (x) and T2 (f (x)) = f (1 − x).
(a) Find T1 + T2 and its matrix representation [T1 + T2 ]B,B .
(b) Find T1 ◦ T2 and its matrix representation [T1 ◦ T2 ]B,B .
(c) Find T2 ◦ T1 and its matrix representation [T2 ◦ T1 ]B,B .
 3

(d) Verify that

[T1 + T2 ]B,B = [T1 ]B,B + [T2 ]B,B
[T1 ◦ T2 ]B,B = [T1 ]B,B [T2 ]B,B
[T2 ◦ T1 ]B,B = [T2 ]B,B [T1 ]B,B
Here B is the basis {1, x, x2 }.

Solution. Since
(T1 + T2 )(f (x)) = T1 (f (x)) + T2 (f (x)) = xf 0 (x) + f (1 − x)
(T1 + T2 )(1) = x(1)0 + 1 = 1 · 1 + 0 · x + 0 · x2
(T1 + T2 )(x) = x(x)0 + (1 − x) = 1 · 1 + 0 · x + 0 · x2
(T1 + T2 )(x2 ) = x(x2 )0 + (1 − x)2 = 1 · 1 − 2 · x + 3 · x2
we obtain
 
1 1 1
[T1 + T2 ]B,B = 0 0 −2
0 0 3
Since
(T1 ◦ T2 )(f (x)) = T1 (T2 (f (x)) = T1 (f (1 − x)) = −xf 0 (1 − x)
(T1 ◦ T2 )(1) = −x · 0 = 0 · 1 + 0 · x + 0 · x2
(T1 ◦ T2 )(x) = −x · 1 = 0 · 1 − 1 · x + 0 · x2
(T1 ◦ T2 )(x2 ) = −x(2(1 − x)) = 0 · 1 − 2 · x + 2 · x2
we obtain
 
0 0 0
[T1 ◦ T2 ]B,B = 0 −1 −2
0 0 2
Since
(T2 ◦ T1 )(f (x)) = T2 (T1 (f (x)) = T2 (xf 0 (x)) = (1 − x)f 0 (1 − x)
(T2 ◦ T1 )(1) = (1 − x) · 0 = 0 · 1 + 0 · x + 0 · x2
(T2 ◦ T1 )(x) = (1 − x) · 1 = 1 · 1 − 1 · x + 0 · x2
(T2 ◦ T1 )(x2 ) = (1 − x)(2(1 − x)) = 2 · 1 − 4 · x + 2 · x2
we obtain
 
0 1 2
[T2 ◦ T1 ]B,B = 0 −1 −4
0 0 2
4

Since
T1 (1) = x(1)0 = 0 · 1 + 0 · x + 0 · x2
T1 (x) = x(x)0 = 0 · 1 + 1 · x + 0 · x2
T1 (x2 ) = x(x2 )0 = 0 · 1 + 0 · x + 2 · x2
we obtain  
0 0 0
[T1 ]B,B = 0 1 0 .
0 0 2
Since
T2 (1) = 1 = 1 · 1 + 0 · x + 0 · x2
T2 (x) = 1 − x = 1 · 1 − 1 · x + 0 · x2
T2 (x2 ) = (1 − x)2 = 1 · 1 − 2 · x + 1 · x2
we obtain  
1 1 1
[T2 ]B,B = 0 −1 −2 .
0 0 1
Therefore,
   
0 0 0 1 1 1
[T1 ]B,B + [T2 ]B,B = 0 1 0 + 0 −1 −2
0 0 2 0 0 1
 
1 1 1
= 0 0 −2 = [T1 + T2 ]B,B
0 0 3
  
0 0 0 1 1 1
[T1 ]B,B [T2 ]B,B = 0 1 0 0 −1 −2
0 0 2 0 0 1
 
0 0 0
= 0 −1 −2 = [T1 ◦ T2 ]B,B

0 0 2
and
  
1 1 1 0 0 0
[T2 ]B,B [T1 ]B,B = 0 −1 −2 0 1 0
0 0 1 0 0 2
 
0 1 2
= 0 −1 −4 = [T2 ◦ T1 ]B,B
0 0 2
 5


(4) Which of the following statements are true and which are false?
Justify your answer.
(a) Let T : V → W and S : U → V be two linear transfor-
mations. Then the kernel K(S) of S is contained in the
kernel K(T ◦ S) of T ◦ S.

Proof. True since

u ∈ K(S) ⇒ S(u) = 0 ⇒ T (S(u)) = 0 ⇒ u ∈ K(T ◦ S).

(b) Let T : V → W and S : U → V be two linear transforma-
tions. Then the range R(T ) of T is contained in the range
R(T ◦ S) of T ◦ S.

Proof. False. The correct statement is

R(T ◦ S) ⊂ R(T ).
Choose S = 0 and T 6= 0. Then R(T ◦ S) = {0} and
R(T ) 6= {0}. 
(c) Let T1 : V → W and T2 : V → W be two linear transfor-
mations. Then R(T1 + T2 ) = R(T1 ) + R(T2 ).
Proof. False. Choose T1 = T 6= 0 and T2 = −T . Then
T1 + T2 = 0 and hence R(T1 + T2 ) = {0}. But
R(T1 ) = R(T2 ) = R(T ) ⇒ R(T1 ) + R(T2 ) = R(T ) 6= {0}.

(d) If T : V → V is a linear transformation satisfying R(T ) =
R(T 2 ), then R(T 2013 ) = R(T 2014 ), where
T n = |T ◦ T {z
◦ ... ◦ T} .
n times

Proof. True since

R(T ) = R(T 2 ) ⇒ T (V ) = T 2 (V ) ⇒ T n−1 (T (V )) = T n−1 (T 2 (V ))
⇒ T n (V ) = T n+1 (V ) ⇒ R(T n ) = R(T n+1 )
for all n ≥ 1. 
6

(5) Find a linear transformation T : R[x] → R[x] such that

K(T ) = {0} and R(T ) 6= R[x].

Solution. Let T (f (x)) = xf (x). First of all, T is a linear trans-

formation since
T (f (x) + cg(x)) = x(f (x) + cg(x)) = xf (x) + cxg(x)
= T (f (x)) + xT (g(x))
for all f (x), g(x) ∈ R[x] and c ∈ R.
Second,
K(T ) = {f (x) : xf (x) ≡ 0} = {0}.
Finally, 1 6∈ R(T ) since 1 6≡ xf (x) for any f (x) ∈ R[x].
Therefore, R(T ) 6= R[x]. 
(6) Find the kernels and ranges of the following linear transforma-
tions:
(a) T : R2 → R3 given by T (x, y) = (x, y, x + y);

Solution. The kernel of T is

K(T ) = {(x, y) : T (x, y) = 0}
= {(x, y) : x = y = x + y = 0)} = {(0, 0)}.
The range of T is
R(T ) = Span{T (e1 ), T (e2 )}
= Span{(1, 0, 1), (0, 1, 1)} = {(s, t, s + t) : s, t ∈ R}
= {(x, y, z) : x + y − z = 0}.

(b) T : Mn×n (R) → Mn×n (R) given by T (A) = A + AT ;
Solution. The kernel of T is
K(T ) = {A : T (A) = 0} = {A : A = −AT }
i.e., the subspace of skew-symmetric matrices.
We claim that the range of T is the subspace
W = {A : A = AT }
of symmetric matrices.
Since (A + AT )T = AT + A, T (A) is symmetric for all A.
Therefore, R(T ) ⊂ W .
 7

On the other hand, for every A ∈ W , A = AT and hence

1 1 1
T ( A) = (A + AT ) = (2A) = A.
2 2 2
Therefore, W ⊂ R(T ). In conclusion, R(T ) = W . 

(c) T : R[x] → R2 given by T (f (x)) = (f (1), f 0 (2));

Solution. The kernel of T is

K(T ) = {f (x) : T (f (x)) = 0} = {f (x) : f (1) = f 0 (2) = 0}.
Every polynomial f (x) can be written as
f (x) = (x − 1)(x − 2)2 p(x) + r(x)
where p(x), r(x) ∈ R[x]| with deg r(x) < 3.
Note that every polynomial g(x) = (x − 1)(x − 2)2 p(x)
satisfies
g(1) = g 0 (2) = 0
and hence g(x) ∈ K(T ). Therefore, f (x) ∈ K(T ) if and
only if r(x) = a + bx + cx2 ∈ K(T ), i.e.,
a + b + c = b + 4c = 0 ⇔ r(x) = c(3 − 4x + x2 ).
Therefore,
K(T ) = {(x − 1)(x − 2)2 p(x) + c(x2 − 4x + 3) :
p(x) ∈ R[x], c ∈ R}
= Span{x2 − 4x + 3, (x − 1)(x − 2)2 , (x − 1)2 (x − 2)x,
..., (x − 1)(x − 2)2 xn , ...}
The range of T is
R(T ) = Span{T (1), T (x), T (x2 ), ...}.
Since
Span{T (1), T (x)} = Span{(1, 0), (1, 1)} = R2
we conclude that R(T ) = R2 . 

(d) T : R[x] → R[x] given by T (f (x)) = xf 0 (x) + f (x + 1).

8

Solution. Let f (x) = a0 + a1 x + ... + an xn be a polynomial

of degree n ≥ 0 (an 6= 0). Then
T (f (x)) = xf 0 (x) + f (x + 1)
= x(a1 + ... + nan xn−1 )
+ (a0 + a1 (x + 1) + ... + an (x + 1)n )
= (n + 1)an xn + g(x)
where deg g(x) ≤ n − 1. Therefore, deg T (f (x)) = n and
hence T (f (x)) = 0 if and only if f (x) ≡ 0. Consequently,
K(T ) = {0}.
The range R(T ) is
R(T ) = Span{T (1), T (x), T (x2 ), ..., T (xn ), ...}
= Span{1, 2x + 1, 2x2 + (x+)2 , ..., nxn + (x + 1)n , ...}
Since deg(nxn + (x + 1)n ) = n,
{nxn + (x + 1)n : n = 0, 1, 2, ...}
is a basis of R[x]. Therefore, R(T ) = R[x]. 

(7) Let Mm×n (R) be the vector space of m × n real matrices and
T : Mm×n (R) → Mm×n (R) be the linear transformation given
by
T (A) = P AQ
for some m × m matrix P and n × n matrix Q. Show that
K(T ) = {0} if and only if both P and Q are nonsingular.

Proof. Suppose that both P and Q are nonsingular.

If A ∈ K(T ), then T (A) = P AQ = 0. Since P and Q are
invertible,
P AQ = 0 ⇒ P −1 (P AQ)Q−1 = 0 ⇒ A = 0
and hence K(T ) = {0}.
Suppose that K(T ) = {0}.
If P is singular, P x = 0 for some m × 1 column vector x 6= 0;
then P AQ = 0 for
 
A = x x ... x
and hence A 6= 0 ∈ K(T ). Contradiction. Therefore, P is
nonsingular.
 9

If Q is singular, QT x = 0 for some n×1 column vector x 6= 0;

then P AQ = 0 for
 T
A = x x ... x
and hence A 6= 0 ∈ K(T ). Contradiction. Therefore, Q is
nonsingular. 
(8) Let T : V → W and S : U → V be two linear transformations.
Show that T ◦ S = 0 (i.e. T (S(u)) = 0 for all u ∈ U ) if and
only if R(S) ⊂ K(T ).

Proof. T (S(u)) = 0 for all u ∈ U ⇔ S(u) ∈ K(T ) for all u ∈ U

⇔ R(S) ⊂ K(T ).