Mathematics 3410 - Homework 2

Only selected problems will be graded.

In this homework set, I will use the following fact:

Fact. The span of a set of vectors is a subspace.
Let V be a real vector space and v1 , . . . , vk ∈ V . Then

U = span(v1 , . . . , vk ) = {c1 v1 + · · · + ck vk | c1 , . . . , ck ∈ R}

is a subspace of V .
Proof. (i) Note that ~0 lies in U . This is since

~0 = 0 · v1 + · · · + 0vk .

(ii) Let u, w ∈ U where

u = c1 v1 + · · · + ck vk and w = d1 v1 + · · · + dk vk .

Thus

u + w = (c1 v1 + · · · + ck vk ) + (d1 v1 + · · · + dk vk ) = (c1 + d1 )bv1 + · · · + (ck + dk )vk ∈ U.

(iii) Let c ∈ R and u ∈ U . Then

cu = c(c1 v1 + · · · + ck vk ) = (cc1 )v1 + · · · + (cck )vk ∈ U.

It follows from the subspace test that U is a subspace of V .

1. In each case determine if U is a subspace of V . Justify your answer in each case.

(i) U = {x ∈ R2 | x1 + x2 = 1}, V = R2 .
Solution. This is not a subspace of R2 since ~0 = (0, 0) does not lie on this
line as 0 + 0 6= 1.
 
a
(ii) U = {x ∈ R3 | x =  b  for some a, b ∈ R}, V = R3 .
a+b
Solution. Since x = (a, b, a + b) = a(1, 0, 1) + b(0, 1, 1) for a, b ∈ R this is
a plane passing through the origin and therefore is a subspace. In fact, it
follows that
U = Span((1, 0, 1), (0, 1, 1))
which is a subspace as shown above.
(iii) U = {x ∈ R3 | x1 + 2x2 < 0}, V = R3 .
Solution. Note that ~0 = (0, 0, 0) is not in this set since 0 + 2 · 0 = 0. Since
~0 ∈
/ U , this is not a subspace of R3 .
(iv) U = {x ∈ R3 | x21 + x22 + x23 = 0}, V = R3 .
Solution. Observe that the only vector lying in this set is ~0 = (0, 0, 0). (This
is since if a vector x = (x1 , x2 , x3 ) has one non-zero entry then x21 +x22 +x23 > 0
and is not in the set. Therefore this set equals {~0} and is a subspace.
(v) U = {x ∈ R3 | x21 + x22 + x23 = −1}, V = R3 .
Solution. Observe that U = φ, the empty set. This is since if x = (x1 , x2 , x3 ) ∈
R3 , then
x21 + x22 + x23 ≥ 0 6= −1.
Since (0, 0, 0) is not in this set this is not a subspace.

1
      
3 2 1
(vi) U = {x ∈ R3 | x =  0  + s  1  + t  2  for some s, t ∈ R}, V = R3 .
1 1 1
Solution. Observe that (3, 0, 1) = 2(2, 1, 1) + (−1)(1, 2, 1). Therefore if
x ∈ U , then
         
3 2 1 2 1
x = 0 + s 1 + t 2 = (s + 2) 1 + (t − 1) 2 
        
1 1 1 1 1

for some s, t ∈ R. It follows that U is a subset of Span((2, 1, 1), (1, 2, 1)). On

the other hand, if y ∈ Span((2, 1, 1), (1, 2, 1)) then
         
2 1 3 2 1
y = a 1 + b 2 = 0 + (a − 2) 1 + (b + 1) 2  ∈ U
        
1 1 1 1 1

for some a, b ∈ R. Thus Span((2, 1, 1), (1, 2, 1)) is a subset of U . It follows

that
U = Span((2, 1, 1), (1, 2, 1))
and is thus a subspace of R3 by the above fact.
2. Let A be an n × n matrix. Verify that

V = {x ∈ Rn | Ax = 3x}

is a subspace of Rn .
Solution. We shall check this is a subspace of Rn .
(i) We have that ~0 ∈ V since A~0 = 0 = 3~0.
(ii) Let x ∈ V and c ∈ R. We have that

Ax = 3x.

Note that
A(cx) = c(Ax) = c(3x) = 3(cx)
and thus cx ∈ V .
(iii) Let x, y ∈ V . This means that

Ax = 3x and Ay = 3y.

Therefore
A(x + y) = Ax + Ay = 3x + 3y = 3(x + y)
and it follows that V is a subspace of Rn .
3. (a) Let U and V be subspaces of Rn . Define the intersection of U and V to be

U ∩ V = {x ∈ Rn : x ∈ U and x ∈ V }.

Show that U ∩ V is a subspace of Rn .

(b) Is U ∪ V = {x ∈ Rn : x ∈ U or x ∈ V } a subspace of Rn ? Give proof or give
counterexample.
Solution.
(i) Since U and V are subspaces, we have ~0 ∈ U and ~0 ∈ V and thus ~0 ∈ U ∩ V .
(ii) If c ∈ R and x ∈ U ∩ V we have cx ∈ U (since U is a subspace) and cx ∈ V
(since V is a subspace). It follows that cx ∈ U ∩ V .

2
 (iii) If x and y ∈ U ∩ V . We have x + y ∈ U (since x and y ∈ U and U
is a subspace), and x + y ∈ V (since x and y ∈ V and V is a subspace), so
x+y ∈U ∩V.
It follows that U ∩ V is a subspace of Rn .
(b) Is U ∪ V = {x ∈ Rn : x ∈ U or x ∈ V } a subspace of Rn ? In general, this is
not necessarily a subspace. For example, let

U = {(x1 , 0) ∈ R2 | x1 ∈ R},
V = {(0, x2 ) ∈ R2 | xx ∈ R}.

Note that
U ∪ V = {(x1 , x2 ) | x1 x2 = 0}.
(This is the cross containing the x1 -axis and the x2 -axis.) It was explained in
class why this is not a subspace of R2 . It is not closed under vector addition. We
have (1, 0), (0, 1) ∈ U ∪ V but their sum (1, 0) + (0, 1) = (1, 1) ∈
/ U ∪V.
4. For each of the following vectors b ∈ R4 , decide whether b is a linear combination
of
v1 = (1, 0, 1, −2), v2 = (0, −1, 0, 1), v3 = (1, −2, 1, 0).
If b is a linear combination, then find c1 , c2 , c3 ∈ R such that b = c1 v1 + c2 v2 +
c3 v3 . If not, then prove it is not possible.
(a) b = (1, 1, 1, 1),
(b) b = (1, −1, 1, −1).
Solution. b is a linear combination of v1 , v2 , and v3 if and only if there exist real
numbers c1 , c2 , and c3 such that

c1 v1 + c2 v2 + c3 v3 = b.

As explained in class, this is the same as solving the matrix equation Ax = b

where A is the 4 × 3 matrix whose columns are v1 , v2 , and v3 and x = (c1 , c2 , c3 )
(We really mean a column vector here.). Thus it suffices to reduce the augmented
matrices    
1 0 1 1 1 0 1 1
 0 −1 −2 1 
 and  0 −1 −2 −1 
 

 1 0 1 1   1 0 1 1 
−2 1 0 1 −2 1 0 −1
and determine if each system is consistent or not.
(a) Let’s consider the first augmented matrix. Applying the row operations R3 →
R3 − R1 and R4 → R4 + 2R1 we find its equivalent to
 
1 0 1 1
 0 −1 −2 1 
 .
 0 0 0 0 
0 1 2 3

Then the row operation R4 → R4 + R2 yields

 
1 0 1 1
 0 −1 −2 1 
 .
 0 0 0 0 
0 0 0 4

Since the last row reads 0 = 4, it follows that this system is inconsistent. Therefore
b = (1, 1, 1, 1) is not a linear combination of v1 , v2 , and v3 .

3
 (b) We now deal with the second augmented matrix. Applying the row operations
R3 → R3 − R1 and R4 → R4 + 2R1 we find its equivalent to
 
1 0 1 1
 0 −1 −2 1 
 .
 0 0 0 0 
0 1 2 1

Then the row operation R4 → R4 + R2 yields

 
1 0 1 1
 0 −1 −2 1 
 
 0 0 0 0 
0 0 0 0

and multipliying R2 by (−1) we obtain

 
1 0 1 1
 0 1
 2 −1 

 0 0 0 0 
0 0 0 0

The pivot variables are c1 and c2 and the non-pivot variable is c3 . The equations
we have are c1 + c3 = 1 and c2 + 2c3 = 1 or

c1 = −c3 + 1 and c2 = −2c3 + 1.

Therefore        
c1 −c3 + 1 −1 1
x =  c2  =  −2c3 + 1  = c3  −2  +  1 
c3 c3 1 0
Choosing c3 = 0 produces a linear combination c1 = 1, c2 = 1, and c3 = 0 or

(1, −1, 1, −1) = 1 · (1, 0, 1, −2) + 1 · (0, −1, 0, 1) + 0(1, −2, 1, 0).

5. Let k ∈ N and v1 , . . . , vk ∈ Rn . Prove that

Span(v1 , . . . , vk ) = Span(v1 , . . . , vk , v) if and only if v ∈ Span(v1 , . . . , vk ).

Solution. Let’s first prove that =⇒ . Span(v1 , . . . , vk ) = Span(v1 , . . . , vk , v).

Therefore
v ∈ Span(v1 , . . . , vk , v) = Span(v1 , . . . , vk ).
Now we prove the opposite direction. If v ∈ Span(v1 , . . . , vk ), then there exist
constants c1 , . . . , ck such that

v = c1 v1 + · · · + ck vk

and it suffices to prove that

k
X
Span(v1 , . . . , vk ) = Span(v1 , . . . , vk , cj vj ).
j=1
Pk
We set A = Span(v1 , . . . , vk ) and B = Span(v1 , . . . , vk , j=1 cj vj ). Let x ∈ A.
Then ∃ a1 , . . . , ak ∈ R such that
k
X
x = a1 v1 + · · · + ak vk = a1 v1 + · · · + ak vk + 0( cj vj ).
j=1

4
 It follows that x ∈ B and hence A ⊂ B. On the other hand, let x ∈ B. Then we
have ∃ b1 , . . . , bk+1 ∈ R such that
k
X
x = b1 v1 + · · · + bk vk + bk+1 ( cj vj )
j=1

= (b1 + bk+1 c1 )v1 + · · · + (bk + bk+1 ck )vk .

Thus x ∈ A and hence B ⊂ A. This establishes A = B.

6. Let f (t) = at2 + bt + c ∈ P2 where a, b, c ∈ R. Show that f (t) is a linear
combination of the polynomials p1 = (t − 1)2 , p2 = t − 1, and p3 = 1. This shows
that
P2 = span{p1 , p2 , p3 }.
Solution. Let at2 + bt + c = x(t − 1)2 + y(t − 1) + z. Expanding out we see that

at2 + bt + c = x(t2 − 2t + 1) + y(t − 1) + z

= xt2 + (−2x + y)t + (x − y + z).

Equating coefficients we obtain

a=x
b = −2x + y
c = x − y + z.

It follows that x = a, y = b + 2x = b + 2a = 2a + b, and z = c − x + y =

c − a + (b + 2a) = c + b + a = a + b + c. This system has unique solution x = a,
y = 2a + b, z = a + b + c for any a, b, and c. So any polynomial f (t) = at2 + bt + c
can be written as a linear combination of p1 , p2 , and p3 .