Test 2: Problem 1 Points Scores
Test 2: Problem 1 Points Scores
Test 2: Problem 1 Points Scores
Points 6 12 10 10 12 10 50+10
Scores
There are 5 problems worth 50 points total and a bonus problem worth up to 10 points.
Show all work. Always indicate carefully what you are doing in each step; otherwise it may not be possible
to give you appropriate partial credit.
1. [6 points] Let W1 and W2 be linear subspaces of a vector space V such that W1 + W2 = V and
W1 ∩ W2 = {0}. Prove that for each vector α ∈ V there are unique vectors α1 ∈ W1 and α2 ∈ W2
such that α = α1 + α2 .
Solution: Since W1 + W2 = V , every vector α ∈ V can be represented as
α = α1 + α2
We must then show that this representation is unique. Suppose that there are two other vectors β1 ∈ W1
and β2 ∈ W2 such that
α = β1 + β2 .
Then
α − α = 0 = (α1 + α2 ) − (β1 + β2 )
so
α1 − β1 = β2 − α2 .
Call the above vector γ. Then since α1 and β1 are both in W1 , their difference γ must also be in W1 .
Similarly, γ must be in W2 , and so
γ ∈ W1 ∩ W2 .
Therefore γ = 0, that is, α1 = β1 and α2 = β2 .
2. [12 points] Consider the vectors in R4 defined by
(a) [8 points] What is the dimension of the subspace W of R4 spanned by the three given
vectors? Find a basis for W and extend it to a basis B of R4 .
Solution:
α1 −1 0 1 2 −1 0 1 2 −1 0 1 2 −1 0 1 2
α2 = 3 4 −2 5 → 0 4 1 11 → 0 4 1 11 → 0 4 1 11
α3 1 4 0 9 1 4 0 9 0 4 1 11 0 0 0 0
We can use {α1 , α2 } as a basis for W (or the first two rows of the above matrix). To extend this to a
basis for R4 , we can add any two vectors which are linearly independent to α1 , α2 , and the other vector.
For example,
B = {α1 , α2 , (0, 0, 1, 0), (0, 0, 0, 1)}
will do nicely.
(b) [4 points] Use a basis B of R4 as in (a) to characterize all linear transformations T : R4 → R4
that have the same null space W . What can you say about the rank of such a T ? What is therefore
the precise condition on the values of T on B?
Solution: If T is a linear transformation whose null space is W , then T (α1 ) = 0 and T (α2 ) = 0.
If β is any other vector not in W , then T (β) 6= 0 (or else β would be in the null space of T ). Since
What are the coordinates of the vector (a, b, c) in the ordered basis B?
Solution:
a 1 1 1
b = x1 0 + x2 1 + x3 0 ,
c −1 1 0
a = x1 + x2 + x3 , b = x2 , c = −x1 + x2
x1 = b − c, x2 = b, x3 = a − 2b + c
Answer: (b − c, b, a − 2b + c)
4. [10 points] Let V be the vector space over R of all real polynomial functions p of degree at
most 2. For any fixed a ∈ R consider the shift operator T : V → V with (T p)(x) = p(x + a).
Explain why T is linear and find the range and null space of T . Is T an isomorphism? Write down
the matrix of T with respect to the ordered basis B = {1, x, x2 }.
Solution: Why T is linear: [T (p1 + cp2 )](x) = (p1 + cp2 )(x + a) = p1 (x + a) + cp2 (x + a) =
(T p1 )(x) + (T p2 )(x)
Null space: If T p = 0, then p(x + a) = 0 for all x. Hence, p(x) = 0 for all x. So the null space is trivial.
Then T is an isomorphism, because its null space is trivial and its range is V
(a) [3 points] What is the matrix of T in the standard ordered basis for R2 ?
Solution: T (1, 2) = (−2, 1) = x1 (1, 2) + x2 (1, −1) = (x1 + x2 , 2x1 − x2 ), −1 = 3x1 , 5 = −3x2
µ ¶ µ 1 2
¶
x1 y1 −3
TB = = 3
x2 y2 − 35 1
3
(d) [3 points] Prove that for every real number c the operator (T − cI) is invertible.
Solution: Notice that T 2 = −I. Then (T − cI)(T + cI) = T 2 − c2 I = −(1 + c2 )I. For every
1
real c the number 1 + c2 is positive, thus nonzero. Hence, − 1+c 2 (T + cI) is the inverse of T − cI.
Bonus Problem [up to 10 points] Let T, U ∈ L(V, V ) be linear operators on the finite dimensional
vector space V . Prove that the rank of the composition U T is less than or equal to the minimum
of the ranks of T and U .
Solution: (thanks to Junmeng Chen)
Note that the range of U T is a subspace of the range of U , because T (V ) ⊂ V implies that U (T (V )) ⊂
U (V ). Hence the rank of U T is less than or equal to the rank of U .
On the other hand, the null space of U T contains the null space of T , because T (α) = 0 implies
that U (T (α)) = 0. Hence the nullity of U T is greater than or equal to the nullity of T . Since
rank(T ) = dim(V ) − null(T ) and rank(U T ) = dim(V ) − null(U T ), it follows that the rank of U T
is less than or equal to the rank of T .