Linear Algebra Take Home Test 4

7/16/2009
Solutions
1. Let V and W be nite dimensional vector spaces and suppose T: V W is linear. Prove that
T is injective if and only if dim(V ) = rank(T).
Proof: () Suppose T is injective, then ker T = {0}, so Null(T) = Dim(ker T) = 0 so by the
rank + nullity theorem
dim(V ) = Rank(T) + Null(T) = Rank(T).
() Suppose Rank(T) = dim(V ) then by rank + nullity theorem Null(T) = 0 so ker T = {0}
and this says T is injective.
2. Let V and W be nite dimensional vector spaces and suppose T: V W is linear. Prove that
T cannot be surjective if dim(V ) < dim(W)
Proof: Suppose dim(V ) < dim(W) and assume for a contradiction that T is surjective. Then
by denition T(V ) = W. This says
dim(W) = dim(T(V )) = Rank(T)
By the rank + nullity theorem we now know
dim(V ) = dim(W) + Null(T).
But by assumption, dim(V ) < dim(W) and so
dim(W) > dim(W) + Null(T)
or in otherwords, Null(T) < 0. But the Null(T) = dim(ker T) and dimension is the number of
elements in a set. So Null(T) < 0 doesnt make sense, its a contradiction. Therefore T cannot
be surjective.
3. Let V and W be nite dimensional vector spaces and suppose T: V W is linear. Prove that
T cannot be injective if dim(V ) > dim(W).
Proof: Suppose dim(V ) > dim(W) and assume for a contradiction that T is injective. Then
we know ker T = {0} so Null(T) = 0 so by the rank + nullity theorem
dim(V ) = Rank(T) = dim(T(V )).
But we know that dim(V ) > dim(W) so this says
dim(W) < dim(T(V )).
However, T(V ) is a subspace of W and so we know dim(T(V )) < dim(W) and this is a
contradiction. Therefore, T cannot be injective.
4. Let T: P
1
(R) P
1
(R) be the linear transformation dened by
T(p(x)) = p

(x).
1
Let B
1
= {1, x} and B
2
= {1 + x, 1 x} be two bases for P
1
(R). Use the fact that

1 1
1 1

1
=

1
2
1
2
1
2

1
2

to nd [T]
B
2
B
2
Solution: Since T(1) = 0 = 0 1 + 0 x and T(x) = 1 = 1 1 + 0 x we know
[T]
B
1
B
1
=

0 1
0 0

Now note Id(1 + x) = 1 + x = 1 1 + 1 x and Id(1 x) = 1 1 1 x. Therefore

[Id]
B
1
B
2
= Q =

1 1
1 1

So by proposition 13 [T]
B
2
B
2
= Q
1
[T]
B
1
B
1
Q, that is,
[T]
B
2
B
2
=

1
2
1
2
1
2

1
2

0 1
0 0

1 1
1 1

1
2

1
2
1
2

1
2

5. Let V = P
1
(R), and, for p(x) V , dene f
1
, f
2
V

by
f
1
(p(x)) =

1
0
p(t)dt and f
2
p(x) =

2
0
p(t)dt.
Prove that {f
1
, f
2
} is a basis for V

, and nd a basis for V for which it is the dual basis.
Solution: Let B = {p
1
(x), p
2
(x)} = {a
1
+ b
1
x, a
2
+ b
2
x}. Now if we want {f
1
, f
2
} to behave
like a dual basis, then
1 = f
1
(p
1
(x)) =

1
0
a
1
+ b
1
tdt = a
1
+
1
2
b
1
0 = f
2
(p
1
(x)) =

2
0
a
1
+ b
1
tdt = 2a
1
+ 2b
1
0 = f
1
(p
2
(x)) =

1
0
a
2
+ b
2
tdt = a
2
+
1
2
b
2
1 = f
2
(p
2
(x)) =

2
0
a
2
+ b
2
tdt = 2a
2
+ 2b
2
.
This tells us a
1
= 2, b
1
= 2, a
2
= 1/2, and b
2
= 1 so B = {2 2x, 1/2 + x}. Now lets
check that B really is a basis. Since its got 2 elements, all we need to show is that its linearly
independent. Suppose there are c
1
and c
2
such that
c
1
(2 2x) + c
2
(1/2 + x) = 0
then
(2c
1
1/2c
2
) 1 + (2c
1
+ c
2
) x = 0.
We already know {1, x} is a basis for P
1
(R) so
2c
1
1/2c
2
= 0
2c
1
+ c
2
= 0
Equivalently, c
1
= c
2
= 0 and therefore, B is a basis and were done.
2