Scribd
Upload
110 views5 pages

Midterm1 Solution

The document contains the solution to 6 problems in linear algebra. Problem 1 shows that a set V is not a vector space over R with the given operations. Problem 2 proves that the union of two subspaces W1 and W2 of a vector space V is a subspace if and only if one is contained in the other. Problem 3 proves that the span of the union of sets S1 and S2 is equal to the sum of their individual spans.

Uploaded by

Karthik Kumar K
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
110 views5 pages

Midterm1 Solution

The document contains the solution to 6 problems in linear algebra. Problem 1 shows that a set V is not a vector space over R with the given operations. Problem 2 proves that the union of two subspaces W1 and W2 of a vector space V is a subspace if and only if one is contained in the other. Problem 3 proves that the span of the union of sets S1 and S2 is equal to the sum of their individual spans.

Uploaded by

Karthik Kumar K
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

Linear Algebra

Solution to Midterm 1

Problem 1: Let V = {(a1 , a2 ) : a1 , a2 R}. Define addition of elements of V coordinatewise, and for (a1 , a2 ) in V and c R, define
n(0,0)
if c=0
c(a1 , a2 ) =
(ca1 , ac2 ) if c6=0.
Is V a vector space over R with these operations? Justify your answer. (5 points)
Solution. No! If c, d R, c + d 6= 0, c 6= 0, d 6= 0, then
(c + d)(a1 , a2 ) = ((c + d)a1 ,

a2
)
c+d

usually is not equal to


c(a1 , a2 ) + d(a1 , a2 ) = (ca1 + da1 ,

a1 a2
+ ).
c
d

(VS8) does not hold.

Problem 2: Let W1 and W2 be subspaces of a vector space V . Prove that W1 W2 is a


subspace of V if and only if W1 W2 or W2 W1 . (9 points)
Proof. () that W1 W2 or W2 W1 , then W1 W2 = W1 or W2 .
Since W1 and W2 are subspaces V ,
we have W1 W2 is also a subspace of V
() Suppose that W1 W2 is a subspace of V .
Also suppose that W1 6 W2 and W2 6 W1 , then there exist u, v V such that u
W1 \W2 , v W2 \W1 .
u, v W1 W2 u + v W1 W2 .
If u + v W1 , then (u) + (u + v) W1 v W1
If u + v W2 , then (u + v) + (v) W2 u W2
Hence W1 W2 or W2 W1 .

Problem 3: Show that if S1 and S2 are arbitrary subsets of a vector space V , then
span(S1 S2 ) = span(S1 ) + span(S2 ). (9 points)
1

P
Pn
Proof. Let u span(S1 S2 ), then u = m
i=1 ai vi +
j=1 bj wj , for some scalars ai , i =
1, , m, bj , j = 1, , n, where vi , i = 1, , m, are in S1 and wj , j = 1, , n, are in
P
Pn
S2 . Since m
a
v
is
in
span(S
)
and
i
i
1
i=1
j=1 bj wi is in span(S2 ), we have u span(S1 ) +
span(S2 ). Hence span(S1 S2 ) span(S1 ) + span(S2 ).
Now let v = x + y span(S1 ) + span(S2 ), where x span(S1 ) and y span(S2 ). We
Pm
can write x =
i=1 ai vi , for some scalars ai , i = 1, , m and vi S1 , i = 1, , m
Pn
and y = j=1 bj wj , for some scalars bj , j = 1, , n and wi S2 , j = 1, , n. Then
Pm
Pn
we can see that v = x + y =
a
v
+
i
i
i=1
j=1 bj wj is in span(S1 S2 ), since vi , i =
1, , m, wj , j = 1, , n are in S1 S2 . Hence span(S1 ) + span(S2 ) span(S1 S2 ).
Therefore span(S1 S2 ) = span(S1 ) + span(S2 ).

Problem 4: Prove that a set S is linear dependent if and only if S = {0} or there exist
distinct vectors v, u1 , u2 , , un in S such that v is a linear combination of u1 , u2 , , un .
(9 points)

Proof. () If S is linearly dependent and S 6= {0}, then there exist distinct vectors
u0 , u1 , , un S such that
a0 u0 + a1 u1 + + an un = 0
with at least one of the scalars a0 , a1 , , an is not zero, say a0 6= 0.
Then we have






a2
an
a1
u1 +
u2 + +
un .
u0 =
a0
a0
a0
Hence v = u0 is a linear combination of u1 , u2 , , un .
() If S = {0}, then its clear that S is linearly dependent.
Assume that there exist distinct vectors v, u1 , u2 , , un S such that v is a linear
combination of u1 , u2 , , un , say
v = a1 u 1 + a2 u 2 + an u n ,
for some scalars a1 , a2 , , an .
Then we have
0 = (1)v + a1 u1 + a2 u2 + an un .
Hence S is linearly dependent.

Problem 5: Prove that if W1 is any subspace of a finite-dimensional vector space V ,


then there exists a subspace W2 of V such that V = W1 W2 . (9 points)
Proof. Let = {u1 , , un } be a basis for W1 . Since W1 is a subspace of V . By Replacement Theorem, we can extend to a basis for V , say = {u1 , , un , un+1 , , um }. Let
W2 = span({un+1 , , um }).
Claim that V = W1 W2 .
1. V = W1 + W2 .
If v V , then
v=

m
X
i=1

ai ui =

n
X
i=1

ai ui +

m
X

ai ui W1 + W2 ,

for some scalars ai , i = 1, , m.

i=n+1

This implies that V W1 + W2 . But by the definition of W1 + W2 , we also know that


W1 + W2 V . Hence V = W1 + W2 .
2. W1 W2 = {0}.
P
P
Let u W1 W2 . Then u = ni=1 bi ui = m
i=n+1 ci ui , for some scalars b1 , , bn , cn+1 , , cm .
Then we have
n
m
X
X
bi ui +
(ci )ui = 0.
i=1

i=n+1

But is linearly independent, since is a basis. Hence b1 = = bn = cn+1 = = cm =


0. This implies that u = 0. That is W1 W2 = {0}. By 1 and 2, we have V = W1 W2 .
We are done!

Problem 6: Prove that if W1 and W2 are finite-dimensional subspaces of a vector space
V , then the subspace W1 + W2 is finite-dimensional, and dim(W1 + W2 ) = dim(W1 ) +
dim(W2 ) dim(W1 W2 ). (9 points)
(Hint: Start with a basis {u1 , u2 , , uk } for W1 W2 and extend this set to a basis
{u1 , u2 , , uk , v1 , v2 , , vm } for W1 and to a basis {u1 , u2 , , uk , w1 , w2 , , wp } for
W2 .)
Proof. dim(W1 W2 ) dim(V )
W1 W2 has a finite basis = {u1 , u2 , , uk }.
We can extend to a basis 1 = {u1 , u2 , , uk , v1 , v2 , , vm } for W1 and to a basis
2 = {u1 , u2 , , uk , k1 , k2 , , kp } for W2 .
Let = {u1 , u2 , , uk , v1 , v2 , , vm , w1 , w2 , , wp }.
We claim that is a basis for W1 + W2 .

To prove the claim, we need to check that


1. is linearly independent.
Let a1 u1 + + ak uk + b1 v1 + + bm vm + c1 w1 + + cp wp = 0, for some scalars
a1 , , ak , b1 , , bm , c1 , , cp .
Then (b1 )v1 + + (bm )vm = a1 u1 + + ak uk + c1 w1 + + cp wp W1 W2 .
Since is a basis for W1 W2 , we have (b1 )v1 + + (bm )vm = d1 u1 + + dk uk for
some scalars d1 , , dk .
d1 u1 + + dk uk + b1 v1 + + bm vm = 0
d1 = = dk = b1 = = bm = 0, since 1 is a basis for W1 .
a1 u1 + + ak uk + c1 w1 + + cp wp = 0
a1 = = ak = c1 = = cp = 0, since 2 is a basis for W2 .
Hence is linearly independent.
2. W1 + W2 = span().

Let u = v + w W1 + W2 , where v W1 and w W2 , be any vector in W1 + W2 .


Since 1 is a basis for W1 and 2 is a basis for W2 , we can find some scalars x1 , , xk , y1 , , ym , z1 , , zk ,
such that
u = (x1 u1 + + xk uk + y1 v1 + ym vm ) + (z1 u1 + + zk uk + t1 w1 + + tp wp )

= (x1 + z1 )u1 + + (xk + zk )uk + y1 v1 + ym vm + t1 w1 + + tp wp
That is, W1 + W2 span().
It is easy to see that span() W1 + W2 .
Hence W1 + W2 = span().
Therefore, is a basis for W1 + W2 .
Finally, we have
dim(W1 + W2 ) = k + m + p = (k + m) + (k + p) k = dim(W1 ) + dim(W2 ) dim(W1 W2 ).

Bonus Problem 7: Prove that the intersection of three 6-dimensional subspaces in R8
is not the zero vector space {0}. (5 points)
Proof. Let U, V and W be three 6-dimensional subspaces of R8 .
Then U + V is a subspace of R8 and dim(U + V ) 8.

By Problem 6, we have
dim(U V ) = dim(U ) + dim(V ) dim(U + V ) 6 + 6 8 = 4.
Similarly, we have dim(U W ) 4.
Since U V U and U W U , U V and U W are subspaces of U .
This implies that (U V ) + (U W ) is a subspace of U .
Hence, dim ((U V ) + (U W )) dim(U ) = 6.
By Problem 6 again, we have
dim(U V W ) = dim ((U V ) (U W ))
= dim(U V ) + dim(U W ) dim ((U V ) + (U W ))
4 + 4 6 = 2.
Therefore, the intersection of three 6-dimensional subspaces in R8 has at least dimension
two and can not be the zero vector space {0}.


You might also like