0% found this document useful (0 votes)
110 views8 pages

Mata22 Problem Set

This document contains solutions to mathematical problems from a course at the University of Toronto Scarborough. The problems involve matrix algebra concepts such as matrix multiplication, transposes, symmetric matrices, and linear transformations. Key steps are shown for solving problems involving properties of matrices, linear combinations, and matrix equations.

Uploaded by

jbob15267
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
0% found this document useful (0 votes)
110 views8 pages

Mata22 Problem Set

This document contains solutions to mathematical problems from a course at the University of Toronto Scarborough. The problems involve matrix algebra concepts such as matrix multiplication, transposes, symmetric matrices, and linear transformations. Key steps are shown for solving problems involving properties of matrices, linear combinations, and matrix equations.

Uploaded by

jbob15267
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/ 8

University of Toronto Scarborough

Department of Computer & Mathematical Sciences

MAT A23H 2014/2015

Solutions #3

1. (Fraleigh & Beauregard, pages 46 – 48)


 
  2 −1  
4 1 −2  14 −2
(16) B C = 0 6 = .
5 −1 3 1 −5
  −3 2  
2 −1   3 3 −7
4 1 −2
CB = 0 6  =  30 −6 18 .
5 −1 3
−3 2 −2 −5 12
   2   
2 0 0 2 0 0 4 0 0
(17) A =  0 −1 0 . (a) A2 =  0 (−1)2 0  =  0 1 0 .
0 0 1 0 0 12 0 0 1
 7   
2 0 0 128 0 0
7 7
(b) A =  0 (−1) 0 =  0 −1 0 .
7
0 0 1 0 0 1
      
0 0 −1 0 0 −1 0 0 −1 −2 0 0
(18) A =  0 2 0 . (a) A2 =  0 2 0  0 2 0 = 0 4 0 
2 0 0 2 0 0 2 0 0 0 0 −2
   
(−2)3 0 0 0 0 −1
7 2 3 3
(b) A = (A ) A =  0 4 0   0 2 0 =
3
  0 0 (−2)  2 0 0
−8 0 0 0 0 −1 0 0 8
 0 64 0   0 2 0 =
  0 128 0 .
0 0 −8 2 0 0 −16 0 0
 
  4   
(19) x y = −2 3 −1  −1  = −14 = −14
  3  
4   −8 12 −4
y x =  −1  −2 3 −1 =  2 −3 1 .
3 −6 9 −3
(21) (a) True; (b) False; (c) False; (d) True; (e) False; (f) True; (g) True; (h) True;
(i) False; (j) True.
(23) We know that if v, w are column vectors with n components we can write the dot
product v·w as v T w in matrix notation. (Since v·w = w·v we could even use
w·v.) For the A, b and c of this question we have (A b)·(A c) = (A b)T (A c) =
(bT AT ) (A c).
MATA23H Solutions # 3 page 2

(24) Using the hint, c A = ((c A)T )T = (AT cT )T . We know that AT cT is a linear
combination of the column vectors of AT where the coefficient of the j th column
vector is cj . Since the transpose of a linear combination of column vectors is a
linear combination of row vectors with the same coefficients, we see that c A =
(AT cT )T is the linear combination of the row vectors of A, where the coefficient
of the ith row vector is ci .
     
(25) Let A = aij , B = bij ∈ Mn,k (R).  Now A + B  = a ij + b ij ∈ Mn,k (R). Since
addition in R commutes, aij + bij = bij + aij = B + A. Since B + A is also
n × k, we have A + B = B + A.
 
(27) Let A = aij ∈ Mn,k (R) and consider the the ij th entry of (r + s) A, (r + s) aij .
Now (r + s) aij = r aij + s aij , which is the ij th entry in r A + s A. Since both
(r + s) A and r A + s A ∈ Mn,k (R) we have (r + s) A = r A + s A.
(29) Let A ∈ Mn,k (R) and let B, C ∈ Mk,p (R). Now the ij th entry of A(B + C)
is (ith row of A) · (j th column of B + C) =(ith row of A) · (j th column of B) +
(ith row of A) · (j th column of C) (using the 2nd property of the dot product). This
is just the ij th entry in A B + A C. Since A(B + C) and A B + A C ∈ Mn,p (R),
they are equal.
  T
  T T
 T
(30) If
 A = a ij ∈ M n,k (R), then A = a ji ∈ M k,n (R), so (A ) = aji =
T T
aij ∈ Mn,k (R). Hence (A ) = A.
     
(31) Let A = aij , B = bij  ∈ M n,k
   (R). Now A + B = a ij + b ij ∈ Mn,k (R), so
(A + B)T = aji + bji = aji + bji = AT + B T ∈ Mk,n (R)
     
(32) Let A = aiℓ ∈ Mn,k (R), B = bℓj ∈ Mk,p (R) so A B = cij ∈ Mn,p (R). Mow
Xk X k
T
 
(A B) = cji , where cji = aiℓ bℓj = bℓj aiℓ , but this is just the (j th column
ℓ=1 ℓ=1
of B)·(ith row of A) = (j th row of B T )·(ith column of AT ), which is the jith entry
of B T AT . Since both (A B)T and B T AT ∈ Mp,n (R) we have (A B)T = B T AT .
(35) If B ∈ Mm,n (R) and B = AT then
(a) A ∈ Mn,m (R).
(b) A AT ∈ Mn,n (R).
(c) AT A ∈ Mm,m (R).
(36) Since v, w are column vectors in Rn we can regard v as an n × 1 matrix and
wT as an 1 × n matrix. Hence v wT is n × n. Now (v wT )T = (wT )T v T = w v T
and (w v T )T = (v T )T wT = v wT , we have that v wT and w v T are transposes of
each other.
   
1 1
= hji = HnT . Since Hn = HnT , Hn
   
(37) Now Hn = hij = =
i+j−1 j+i−1
is a symmetric matrix.
(38) Since A is a square matrix, AT is also square and of the same size so A + AT is
defined. Now (A + AT )T = AT + (AT )T = AT + A = A + AT ; i.e., A + AT is
symmetric.
MATA23H Solutions # 3 page 3

(39) Let A ∈ Mn,k (R), then AT ∈ Mk,n (R) and A AT is defined. Now (A AT )T =
(AT )T AT = A AT ; i.e., A AT is symmetric.
(40) (a) (A2 )T = (A A)T = AT AT = (AT )2 and (A3 )T = (A A2 )T = (A2 )T AT =
(AT )2 AT = (AT )3 (using #32).
(b) The generalization is An )T = (AT )n , n ≥ 1. To prove, we will use two steps.
1. base step (n = 1): where we have AT = AT .
2. inductive step (assume true for n = k and show true for n = k + 1): Now
inductive
(Ak+1 )T = (A Ak )T = (Ak )T AT = (AT )k AT = (AT )k+1 .
hypothesis
Hence it is true for all n
 
(41) (a) Let A = aij ∈ Mm,n (R). Now the j th component of A ej can be regarded as
the (ith row of A)·(ej ) = ai1 , ai2 , · · · , aij , · · · , ain ·ej = aij , j 
= 1, 2, ·· · , n,
a1j
 a2j 
since only the j th component of ej is non-zero. Hence A ej =  .. , the
 
 . 
am j
th
j column vector of A.
(b) (i) If A x = 0 for all x we have A ej = 0 for j = 1, 2, · · · , n. Now part (a)
=⇒ that the j th column of A is 0 for all j = 1, 2, · · · , n. Hence A = O,
the zero matrix.
(ii) (A x = B x, for each x) if and only if ((A − B) x = 0, for each x) if and
only if (A − B = O, by part (b)(i)) if and only if (A = B).
(42) (A + B)2 = (A + B)(A + B) = A2 + B A + A B + B 2 = A2 + 2 A B + B 2 if and
only if A B = B A. We already know that A B is generally not equal to B A, so
we seek acounterexample:
      
1 0 0 1 0 1 1 0
Let A = , B= so A B = and B A = . Now
1 0  0 0 0 1    00 
2 2 1 2 2 1 0 0 2 0 0
(A + B) = and A + 2 A B + B = + + =
   1 1  1 0 0 2 0 0
1 2 2 1
6= .
1 2 1 1
1 1
(44) We will try writing A as A = (A + AT ) + (A − AT ). From #38 we know that
2 2
 T
1 1 1 1
(A+AT ) is symmetric. Now (A − AT ) = (AT −(AT )T ) = (AT −A) =
2 2 2 2
1 T
− (A − A ), so this matrix is skew-symmetric. Hence A is written as the sum
2
of a symmetric matrix and a skew-symmetric matrix.
If A = B + C is a 2nd sum, with B symmetric and C skew-symmetric, we have
AT = (B + C)T = B T + C T = B − C. Adding the two expressions, we have
1
A + AT = 2B =⇒ B = (A + AT ) and subtracting we have A − AT = 2 C =⇒
2
MATA23H Solutions # 3 page 4

1
C = (A − AT ). Hence the decomposition is unique.
2
       
2 0 0 1 0 1 2 0 2 1 0 1 2 0 0
(46)  0 r 0   0 1 0  =  0 r 0  and  0 1 0   0 r 0  =
 0 0 2  1 0 1 2 0 2 1 0 1 0 0 2
2 0 2
 0 r 0 . Since both products are equal for all r, we say they commute
2 0 2
for all r ∈ R.
   
2. Let A = aij , B = bij ∈ Mn (R).
 
(a) Since A and B are square matrices, so is A + B = aij + bij . Now tr(A + B) =
(a11 + b11 ) + (a22 + b22 ) + · · · + (ann + bnn ) = (a11 + a22 + · · · + ann ) + (b11 +
b22 + · · · + bnn = tr(A) + tr(B).
 
(b) If λ ∈ R, λ A = λ aij and tr(λ A) = λ a11 + λ a22 + · · · + λ ann = λ (a11 +
a22 + · · · + ann ) = λ tr(A).
 T  
(c) Since AT = aij = aji . tr(AT ) = a11 + a22 + · · · + ann = tr(A).
 Xn 
(d) Since A and B are n × n matrices, A B = cij = aiℓ bℓj and B A =
ℓ=1
 n
X 
dij = bik akj are also n × n matrices. Hence tr(A B) = c11 + c22 + · · · + cnn =
k=1
(a11 b11 + a12 b21 + · · · + a1n bn1 ) + (a21 b12 + a22 b22 + · · · + a2n bn2 ) + · · · + (an1 b1n +
an2 b2n + · · · + ann bnn . Now tr(B A) = d11 + d22 + · · · + dnn = (b11 a11 + b12 a21 + · · · +
b1n an1 ) + (b21 a12 + b22 a22 + · · · + b2n an2 ) + · · · + (bn1 a1n + bn2 a2n + · · · + bnn ann ) =
(a11 b11 + a21 b12 + · · · an1 b1n ) + (a12 b21 + a22 b22 + · · · + an2 b2n ) + · · · (a1n bn1 + a2n bn2 +
collect all f irst terms
· · · ann bnn ) = (a11 b11 + a12 b21 + · · · + a1n bn1 ) + (a21 b12 + a22 b22 +
second terms etc
· · · + a2n bn2 ) + · · · + (an1 b1n + an2 b2n + · · · + ann bnn = tr(A B).

3. (Fraleigh & Beauregard, pages 68 – 71)


     
2 4 −2 R1 → 21 R1 1 2 −1 R2 →R2 −4 R1 1 2 −1
(2) (a)  4 8 3   4 8 3  R3 →R3 +R1  0 0 7 
−1 −3 0 ∼ −1 −3 0 ∼ 0 −1 −1
 
1 2 −1
R2 ↔R3
 0 −1 −1 .

0 0 7
(b) Since
 thereis a pivot in each column we conclude that this matrix reduces to
1 0 0
 0 1 0 .
0 0 1
(It is also correct to continue to row reduce.)
MATA23H Solutions # 3 page 5

     
0 2 −1 3 1 1 −3 3 1 1 −3 3
R2 →R2 +R1
 −1 1 2 0  R1 ↔R3  −1 1
 2 0  R4 →R4 −R1  0 2 −1 3 
 
(3) (a) 
 1 1 −3 3  ∼  0 2 −1 3   0 2 −1 3 

1 5 5 9 1 5 5 9 0 4 8 6
   
1 1 −3 3 1 1 −3 3
R3 →R3 −R2
 0 2 −1 3  R3 ↔R4  0 2 −1 3 
R4 →R4 −2 R2    .
 0 0 0 0  ∼  0 0 10 0 

0 0 10 0 0 0 0 0
   
1
1 1 −3 3 1
1 1 0 3
R2 → 2 R2
 0 1 − 1 3  R2 →R2 + 2 R3  0 1 0 3  R1 →R1 −R2
(b) R3 → 10 R3 
1
 0 0
2 2  R →R +3 R
1 1 3
 2 
1 0   0 0 1 0  ∼
∼ ∼
0 0 0 0 0 0 0 0
3 
1 0 0 2

 0 1 0 3 
 0 0 1 0 .
 2 

0 0 0 0
   
0 0 1 2 −1 −4 2 4 −1 3 2 −1
R1 ↔R3 R2 ↔R3
(6) (a)  0 0 0 1 −1 3   0 0 0 1 −1 3 
∼ ∼
 2 4 −1 3 2 −1  0 0 1 2 −1 −4
2 4 −1 3 2 −1
 0 0 1 2 −1 −4 .
0 0 0 1 −1 3
1
1 2 − 2 23 1 − 12 1 2 − 12 0 5
   
1
R2 →R2 −2 R3
2
−5
R1 → 2 R1
(b)  0 0 1 2 −1 −4  R1 →R1 − 32 R3  0 0 1 0 1 −2 

0 0 0 1 −1 3  ∼ 0 0 0 1 −1 3
1 2 0 0 3 −6
R1 →R1 + 21 R2
 0 0 1 0 1 −2 .

0 0 0 1 −1 3
 
1 0 2 0 1
x1 + 2 x3 = 1
(9)  0 1 1 3 −2  gives equations . Since
x2 + x3 + 3 x4 = −2
0 0 0 0 0
the third and fourth columns are pivot free, we put x4 = t and  x3 = s. Then 
1 − 2s
 −2 − s − 3t 
x2 = −2 − s − 3t and x1 = 1 − 2s giving the general solution x =  
 s 
t
with s, t ∈ R. Putting x3 = 3 and x4 = −2, we have the particular solution,
−5
 1 
x=  3 .

−2
MATA23H Solutions # 3 page 6

 
1 0
0 0 0
x1 = 0
 0 1
0 0 2 
  x2 = 2
(11)  0 0
1 0 −5   gives equations . The unique solution is
 x 3 = −5
 0 0
0 1 2 
x4 = 2
0 0
0 0 0
 
0
 2 
x=  −5 .

2
 
1 −1 2 0 3 1
 0 0 0 1 4 2 
(12) 
 0
. Since the third row has all zeros on the left of the
0 0 0 0 −1 
0 0 0 0 0 0
partition and −1 6= 0 on the right of the partition, the system is inconsistent and
there is no solution.
4 x1 − 3 x2 = 10
(14) Writing the given equations, , as an augmented ma-
  8 x1 −  x2 = 10 
4 −3 10 R2 →R2 −2 R1 4 −3 10
trix we have . The new equations
8 −1 10 ∼ 0 5 −10
4 x1 − 3 x2 = 10
are . The second gives x2 = −2 so the first becomes
5 x2 = −10  
1
4 x1 −3(−2) = 10 =⇒ 4 x1 = 4 =⇒ x1 = 1. The unique solution is x = .
−2
2x + y − 3z = 0
(16) The linear system, 6x + 3y − 8z = 0 , has augmented matrix
  2 x − y
 + 5 z = −4   
2 1 −3 0 R2 →R2 −3 R1 2 1 −3 0 2 1 −3 0
R2 ↔R3
 6 3 −8 0  R3 →R3 −R1  0 0 1 0   0 −2 8 −4 .

2 −1 5 −4 ∼ 0 −2 8 −4 0 0 1 0
2x + y − 3z = 0
We now have equations − 2 y + 8 z = −4 . The third gives z = 0.
z = 0
Then the second becomes −2y + 8(0) = −4 =⇒ y = 2 and  the first becomes
−1
2x + 2 − 3(0) = 0 =⇒ x = −1. The unique solution is x =  2 .
0
x1 − 3 x2 + x3 = 2
(18) The linear system, , has augmented matrix
  3 x 1 − 8 x2 + 2 x3 = 5
 
1 −3 1 2 R2 →R2 −3 R1 1 −3 1 2
. We now have equations
3 −8 2 5 ∼ 0 1 −1 −1
x1 − 3 x2 + x3 = 2
. Since the third column is pivot free, we put
x2 − x3 = −1
x3 = t. Then x2 = −1 + t and the first equation becomes x1 − 3(−1 + t) = 2 =⇒
MATA23H Solutions # 3 page 7

 
−1 + 2 t
x1 = −1 + 2 t. The general solution is x =  −1 + t  , t ∈ R.
t
x1 + 4 x2 − 2 x3 = 4
(19) The linear system, 2 x1 + 7 x2 − x3 = −2 , has augmented matrix
  2 x1 + 9x2 − 7 x3 = 1

1 4 −2 4 R2 →R2 −2 R1 1 4 −2 4
R3 →R3 +R2
 2 7 −1 −2  R3 →R3 −2 R1  0 −1 3 −10 

 2 9 −7 1  ∼ 0 1 −3 −7
1 4 −2 4
 0 −1 3 −10  which indicates an inconsistent system. There is no so-
0 0 0 −17
lution.  
2 x1 + 8 x2 = 16 2 8 16
(22) The linear system, , has augmented matrix
 5 x1 − 4 x2 =  −8   5 −4 −8
R1 → 12 R1 1 4 8 R2 →R2 −5 R1 1 4 8 R2 → −1
24
R 2 1 4 8
∼ 5 −4 −8  ∼ 0 −24 −48   ∼ 0 1 2
R1 →R1 −4 R2 1 0 0 0
. The unique solution is x = .
∼ 0 1 2 2
x1 − 2 x3 + x4 = 6
(23) The linear system, 2 x1 − x2 + x3 − 3 x4 = 0 , has augmented
 9 x1 − 3 x2 − x3  7 x4 = 4 
1 0 −2 1 6 R2 →R2 −2 R1 1 0 −2 1 6
R2 →−R2
matrix  2 −1 1 −3 0  R3 →R3 −9 R1  0 −1 5 −5 −12 

 9 −3 −1 −7 4 ∼  0 −3 17 −16 −50
1 0 −2 1 6 1 0 −2 1 6
R3 →R3 +3 R2 R3 → 21 R3
 0 1 −5 5 12   0 1 −5 5 12 
∼ ∼
0 −3 17 −16 −50 0 0 2 −1 −14
1 0 0 0 −8
   
1 0 −2 1 6 R2 →R2 +5 R3
5
 0 1 −5 5 12  R1 →R1 +2 R3  0 1 0 2
−23 . Since the fourth
1
0 0 1 − 2 −7 ∼ 1
0 0 1 − 2 −7
column is pivot free, we put x 4 = t giving the general solution
−8
 
 −23 − 52 t 
x=  −7 + 1 t  , t ∈ R.

2
t
(26) b is in the span of v 1 , v 2 , v 3 , if there are scalars  x1, x2 , x3 such that v =
  x 1
x1 v 1 + x2 v 2 + x3 v 3 ; i.e., if the system v 1 v 2 v 3  x2  = b has a solution. We
 x3 
1 −4 1 8 R2 →R2 −3 R1
write this system as an augmented matrix:  3 −12 5 26  R3 →R3 −2 R1
2 −9 −1 14 ∼
MATA23H Solutions # 3 page 8

   
1 −4 1 8 1 −4 1 8
R2 ↔R3
 0 0 2 2   0 −1 −3 −2 . Since this system is consis-

0 −1 −3 −2 0 0 2 2
tent, b is in the span of v 1 , v 2 , v 3 .
(29) (a) False; (b) False; (c) True; (d) True; (e) False; (f) True; (g) True; (h) True;
(i) False; (j) True.

You might also like