Sie sind auf Seite 1von 5

MATH 413 Homework 2

David Chung
September 25, 2014
1. Is it possible that vectors v1 , v2 , v3 are linearly dependent, but the vectors
w1 = v1 + v2 , w2 = v2 + v3 , and w3 = v3 + v1 are linearly independent?
Let
v1 = v2 + v3
v2 = v2
v3 = v3

(1)
(2)
(3)

w1 = v1 + v2
= v2 + v3 + v2
= ( + 1)v2 + v3
w2 = v2 + v3
w3 = v3 + v1
= v3 + v2 + v3
= v2 + ( + 1)v3

(4)
(5)
(6)
(7)
(8)
(9)
(10)

with , being scalars. Then

w1 , w2 , w3 are all linear combinations of v2 , v3 . Thus w1 , w2 , w3 are linearly


independent.
P
2.Let u span(S). Then u = ni=1 i ui for ui S1 , i R. But, if ui S1 ,
then ui S2 . Thus the right hand side is also in span(S2 ). This implies that
u span(S2 ). To prove the second part, start with v V . Since S1 S2 ,
from the first part it follows that span(S1 ) span(S2 ). This means that
V span(S2 ). Hence span(S2 ) = V .
1

1 0 0
3a. 0 1 0
0 0 0

1 0 0
3b. 0 1 0
0 0 1

3
1

0
cos(30 ) sin(30 ) 0
2
2
3
3c. sin(30 ) cos(30 ) 0 = 12
0
2
0
0
1
0
0 1




1 0
0 0
4. Let A =
and B =
0 0
1 1


AB =

1 0
0 0



0 0
1 1

10+01
=
00+01


0 0
=
0 0


0 0
1
BA =
1 1
0

01+00
=
11+10


0 0
=
0 0



a b
s
5a. Suppose A =
and x =
c d
t

Ax =


(11)

10+01
00+01

(12)
(13)

0
0


(14)

00+00
10+10


(15)
(16)

where x 6= ~0 and Ax = ~0.

as+bt
cs+dt

= ~0


(17)
(18)

So a = bts , b = ast , c = dts , and d = cst . If we let = st , and = st ,


then we get a = b, b = a, c = d, and d = c. However, only works if
s 6= 0 and if t 6= 0, but not both at the same time, since x 6= ~0. If s 6= 0,
then b = d = 0 and if t 6= 0, then a = c = 0.


 
a c
u
T
So we let A =
and y =
where y 6= ~0.
b d
v


A y=

=

=

a c
b d



au + cv
bu + dv

u
v



(19)
(20)

bu + dv
bu + dv


(21)

Now we can see that the first row is just a scalar multiple of the second row.
So now we can try and solve for either u or v. Lets try solving for u.
bu + dv = 0
bu = dv
dv
u=
b

(22)
(23)
(24)

So if we let v = 0, then b = 0 and u 6= 0. Also if we let v be anything other


than 0, lets say v = 1, then we have u = db , so now we have . . .


bu + dv
bu + dv

b
b

d
b
d
b

+ d(1)
=
+ d(1)


d + d
=
d + d
 
0
=
0
= ~0


(25)
(26)
(27)
(28)

So without loss of generality, this works the same for the rest of the
cases and all of the cases.
3

5b. We want to show that Ax = ~0 AT Ax = ~0. The leftward direction


can be broken down into two cases. The first case is where AT is invertible,
while the other one is while AT is not invertible.
Ax = ~0 AT Ax = ~0 Case 1 AT is invertible. So assume Ax 6= ~0, we let
B = Ax.
AT B = ~0
A

(29)
T ~

(A B) = A

0
(AT AT )B = AT ~0
B = ~0

(30)
(31)
(32)

Ax = ~0 AT Ax = ~0 Case 2 AT is not invertible.


Then ad = bc (determinants), so . . .
AT Ax = 0

(a2 + c2 )x1 + (ab + cd)x2 = 0


(ab + cd)x1 + (b2 + d2 )x2 = 0
(ab + cd)
x2
(a2 + c2 )
(ab + cd)
x2 = 2
x1
(b + d2 )
x1 =

(33)
(34)
(35)
(36)
(37)
(38)

Plug in (37) into ax1 + bx2 = 0.


a(ab + cd)
x2 + bx2
a2 + c 2
(a2 b + cad)
x2 + bx2

a2 + c 2
(a2 b + c2 b)
2
x2 + bx2
a + c2
b(a2 + c2 )
x2 + bx2
2
a + c2
b(a2 + c2 )
2
x2 + bx2
a + c2
bx2 + bx2

=0

(39)

=0

(40)

=0

(41)

=0

(42)

=0

(43)

=0

(44)

=0

(45)

=0

(46)

=0

(47)

=0

(48)

=0

(49)

Now we can plug in (38) into cx1 + dx2 = 0.


d(ab + cd)
x1
b2 + d2
(adb + cd2 )
x1
cx1 + 2
b + d2
(b2 c + cd2 )
cx1 + 2
x1
b + d2
c(b2 + d2 )
cx1 + 2
x1
b + d2
cx1 + cx1
cx1 +

Ax = ~0 AT Ax = ~0

Ax = ~0

(50)

A (Ax) = A (~0)
AT (~0) = AT (~0)
T

~0 = ~0
Thus, Ax = ~0 AT Ax = ~0.
5

(51)
(52)
(53)

Das könnte Ihnen auch gefallen