Professional Documents
Culture Documents
September 2014
Chapter 1
Matrices and Linear Systems
1.1
1.1.1
Exercises
Ex0. Which of the following systems are linear
x
+
x
2x
=
0
x
+
2
x2 2x3
1
2
3
1
4x
3x1 4x2 + x3
3x3 = 2
2
3
x1
x1 + 2x2 + 4x3
7x2 + 4x3 = 5
=3
x + 3xy 4y
=2
11x + y
= 3
=9
=3
Ex1. Rewrite the following linear systems into the matrix forms
x y + 3z
=1
2x
+
x
=
1
1
3
4
x
+ 2z = 2
4x
=
2
2 x3 3x4
2x + 2y z = 3
x 1
+ 4x3
=0
3x + y 3z = 1
2y
+ 3t
=0
+ 2z t = 2
1 2 1
4 13 5
where A =
0 1 2 B = 2 5 2
1 1 6
1 3 1
3
c) BD and DB
11 2
C= 1 3
2 3
d) BA and AB
1 2 0
D=
1 1 3
1 0 2
a)
h
i 3
0 2 1 1
1
b)
2 h
i
2 1 2
1
c)
2 1
i
1 2 2
2
1
b) AC
where A =
Ex5. Compute
1 0 1
2 1 2
c) CA
B=
d) DB
2 5 2
1 3 1
the product
201
1 99
0 0
2
100
3
1 0
55
1 5
0 1
e) DC
0
C=
1
2
2 1
201
100
55
99
D=
1 2 0
1 0 0
0 1 0
3
0 0 1
5
Give an observation about the results obtained when we multiply a matrix (of a suitable size)
by matrix
1 0 0
I3 =
0
1
0
0 0 1
on the left and on the
0 0 1
0 1 0
1 0 0
right.
201 1 99
22 503 3
5
1 95
0 0 1
1 0 0
0 0 1
1 0 0
0 1 0
0 1 0
201
100
55
99
201
100
55
99
Give observations about rows of the result matrix gotten when we multiply a matrix (having
3 rows) respectively by matrices
0 0 1
P31 =
0 1 0
1 0 0
0 1 0
0 0 1
; P21 = 1 0 0 ; P312 = 1 0 0
0 0 1
0 1 0
on the left. Similarly, find a matrix (of size 3 3) which is multiplied on the left of every
matrix (having 3 rows) to make this matrix exchange row 2 and row 3 and keep row 1 in the
result
92
matrix.
79
50
0 1 0 ;
3
9
1 0 0
0 0 1
92 50
8 1
79
0 1 0
1 0 0 ;
3
9
0 0 1
92 50
8 1
79
0 0 1
1 0 0
3
9
0 1 0
Give observations about columns of the result matrix gotten by multiplying a matrix (having
3 columns) respectively by matrices
0 1 0
0 0 1
1 0 0 ;
;
P
=
P31 =
0
1
0
21
0 0 1
1 0 0
0 0 1
P312 =
1
0
0
0 1 0
on the right. Find a matrix (of size 3 3) which is multiplied on the right of every matrix
A (having 3 columns) to get the result matrix whose columns in order from left to right are
column 3, column 1 and then column 2 of A.
Ex6. Compute the following products
1 0 0
a1 b1 c1 d1
k 1 0 a1 b2 c2 d2
0 0 1
a3 b 3 c 3 d 3
1 0 0 0
a1 b 1 c 1
0 1 0 0 a b c
1
2
2
0 0 1 0 a3 b 3 c 3
k 0 0 1
X Y Z
Now compare the result matrix with the matrix on the right of each product above to give
some observations. Find a 3 3 matrix A such that when we multiply A by any matrix B
(having 3 rows), the result matrix keeps row 1 and row 2 of B and its row 3 is sum of row 3
of B and 3 times row 2 of B.
Ex7. Calculate the matrix multiplications then compare two result matrices.
1 0 0
1 0 0
1 0 0
1 0 0
C=
D=
k 1 0 0 1 0
0 1 0 k 1 0
0 0 1
h 0 1
h 0 1
0 0 1
Now multiply matrix C by any matrix B (having 3 rows), then observe the rows of result
matrix in concerning with rows of B. For example,
a b
B=
X Y
x y
take
a b c 1
a
b
c
X Y Z 1
X
Y
Z
P
=
x y z 1 x s + 2X y u + 2Y z v + 2Z
s u v w
s
u
v
2 3 0 3
0 2 1 2
= 1 2 2 0
Q
1
2
2
0
1 3 1 1
1 3
1 1
Ex8. Determine the products below
2
1 0 0
k 1 0 ,
0 0 1
Ex9 . Show
then
1
0 0
0
nk
=
0
1
0
1
n ka na 1
where n := 1 + 2 + 3 + . . . + (n 1) =
Show that for every n 2
1 0 0
n
0 0
1 0 0
k 1 0
,
.
.
.
,
1 0
0 1
0 0 1
3
0 0
1 0 0
1 0 0
1
k 1 0 =
1 0
3k
= 2k 1 0 ,
a 1
ka 2a 1
0 a 1
(1 + 2)ka
1 0
k 1
0 a
n(n 1)
2
k 1 0 =
nk
1
0
b a 1
n ka + nb na 1
3a 1
1
2x y
=0
x + 2y
= 15
(1)
x y
=1
(2)
x y
=3
(3)
2x + y
=3
(5)
(4)
6x + 3y
=9
(6)
Use determinant of order 2 to determine if a system has some solutions or not. If it has a
unique solution, find it.
Draw the lines whose the equations are (1) and (2) respectively in xy-plane. Then find the
insection of two these lines. Compare the coordinates of the intersect point with the solution
we found before.
Draw the lines having their equations (3) and (4) respectively. Observe about their intersection.
Draw the lines having their equations (5) and (6) respectively. Observe about their intersection.
Ex11. Give the equations of three lines in xy plane
ax + by = k
cx + dy = h
mx + ny = l
Decribe the relative position of these lines when the system with these three equations
has no solution
has a unique solution
has more than one solution. In this case, determine how many solution this system could
have.
Ex12. Use the substitution method to solve the linear systems
2x y + z = 0
(1)
x
+ 2z = 1
(a)
2x + y z
x + 2y z = 3
(2)
2x y + z
=3
(b)
2y + z
4x y
3x y + 3z = 2
4x
=4
(3)
(c)
3z
(i)
(ii)
= 1
(iii)
For each system, daw the pictures of all of its equations in the Cartesian coordinate system.
Determine their insection if it exists.
Ex13. Find the column 2 and row 3 of the result matrix of the product
2
4
3
1 1 0 0
2
2
3
0 2 3 1
1 2 2
1 2 0 3
3
0 1
Ex14. Suppose linear system Ax = b has at least one solution, where A = (aij )1im is a
1jn
m n matrix. Show that the system can not have exactly two solutions. That means that if
this systems is consistent, it has either only one solution or infinitively many solutions.
1.1.2
Gaussian Elimination
Ex1. Which of the following matrices are echelon. For each matrix, if it is not echelon,
use Elementary Row Operations to change
1 1 0 0
B=
A=
0
2
0
0
0 0 0 3
0
0 1 0 2 1
3
0 2 0 0 0
C=
0
0 0 0 0 0
0
0 0 2 0 0
2 1
0 1
1
2x1
+ x3
x1 + 2x2 + 3x3
x1 + x2 2x3
=4
=0
=1
x1 + 2x2 + 3x3 x4
2x2 + 2x3 + x4
=4
= 5
=5
x1 x2
x1 + 2x2
3x1 + x2
2x 3x
1
2
Ex3. Use Gaussian Elimination to find b such that the following systems have
only one solution.
more than one solution, find 3 particular solutions of the systems in this cases.
no solution.
x1 2x2 + x3
2x1 + x2 + 3x3
x1 + x2 + bx3
=3
=4
=0
x1
x3 + 2x4
2x1 x2
+ 4x4
=0
=2
=b
=1
=3
= 1
=0
x + 2y
=3
x + y
= 3
x+y+z
x + y
=2
= 2
y 4z
3x y + 2z
2x + y + 2z
x + 2y + 3z
2x
+z
x1 + 2x2
x4
2x2 2x3 + x4 + x5
3x1 + x2
+ x4 + 2x5
= 9
= 2
=3
=7
=0
x
+ 2z
2x y
= 11
=5
=8
=8
z + 3t
2x + y
=8
=4
+ 4t = 4
Ex5. Solve two systems following at the same time by applying Gauss-Jordan elimination on
the extra augmented matrix with two extra columns are the right hand side column vectors
of two systems.
x + y 2z
x + 2y + 3z
2x + 3y + z
=3
=5
=4
Notation. Here 2 systems have the same left hand side of all their equations. Right hand
side of the first system is the column 1 of the right hand side matrix. And that of the second
system is the column 2 of this matrix.
Hint: Use Gauss-Jordan elimination on the extra augmented matrix
1 2 3
Aext =
1 2
2 3
5 7
4 7
3
1
These system can be rewritten in form of the matrix equation (in which matrix is considered
as unknowns) as the following
1 2
1 2
2 3
5 7
X
=
3
1
4 7
10
x1 x2
X=
y1 y2
z1 z2
Ex6. Solve the matrix equation by using Gauss-Jordan elimination (see Ex5 )
1.2
1.2.1
1 2
2
X =
0 1
1 7 12
Exercises
Ex1. Calculate
A+B
AB
A2B
AD +BD
2
2 1
where A =
1 7
2
(A+B)D
B=
3 2
3 1
(A+B)2
11
A2 +AB +BA+B 2
2 2
D=
1
5
3 0
11
b. Compute
a 0 0
x b 0
0 y c
Then find a, b, c, x, y such
is I3 (the found matrix is not identity matrix)
that the product
1 1 2
. Compare the matrices 2A and A+A, 3A and A+A+A.
Ex5. Give matrix A =
3 1 3
Try to prove that kA = |A + A +
{z. . . + A} for all m n matrix A and k is a natural number.
k times
AB
EC
DC
where A =
2
C=
1 0 0
0 1 0
2
;
6 5
a c
D=
c b
d a
A E
I3 D
E=
4 111
2 13
A E
B
AB + EC
I3 D
C
BI3 + DC
13 3 4
B=
6
2
I3
1 2
0 1
12
Ex8. a. Calculate and compare two products AT A and (AT A)T where
A=
1 3
3
2
1
b. Prove that AT A = (AT A)T and AAT = (AAT )T for every matrix A of size m n.
Note. A n n matrix B satisfying the equality B = B T is called a symmetric matrix. In this
exercise, matrices AT A and AAT are symmetric.
2
Ex9. Check that the following matrix A is
guaranteed
A + A 2I2 = O2 ,
3 1
.
where O2 is the 2 2 zero matrix and A =
2 0
1.2.2
Rank of Matrix
2 3 0 1
A=
3
1
2
1
4 6 0 2
3 1 1
2 3
2
B=
0
2
3
1 2
2
3 3
2
3 1 2
2 1 0 1
C=
4 2 5 1
1 2 2 2
3 1 1
4 3 0
D=
1 0 0
7 4 1
1 1 1
1
1
1 1 1
E=
1
1
1
1
1 1
1
1
1 3
F =
5
1 2 2
a)A =
2
3 1
2 3 1
b)A =
3 1 1
1 1 1
3 2 2 1
c)A =
1 0
3 2 1
13
Ex3. a. Compare the rank and the number min{m, n} where m, n are the number of rows
and columns respectively of the matrices in Ex1.
b. In general, where matrix A has size of m n, prove that rank(A) min{m, n}.
Ex4. a. Find the rank of matrices A, B, A + B, AC where
2 2 1
3
A=
2
3 1 1
0 2 1 2
2 1
0 2
B=
4 5 3 1
1
3
1
2
3 5
C=
1 2
3 1
2 1 2
a)
0 3
x1
x2 = 2
2
1 a 1
x3
1
1 1
b)
2
x1
x2 =
2
2
2 4 a + 12
x3
a a 13
0
14
c)
1 a1 1
0
1
0
a2
0
0
1
0 a1
d)
0
3
2
2
1
2
3
0
a
2 a2 + 2a + 5 3
3
8
y
3
z = 3
s
2
7
t
3
x
y =
a+8
z
2x
+
y
=
5
2x + y
x + y
x + y + z
=2
3x y
2x + 3y
= 9
3x y + 2z
2x + 3y
=7
x1 + x2 + 3x3
3x1 + 4x2
+ 2x4 + 3x5
=8
= 1
=9
=5
=2
= 9
=7
2x1 + 3x2
2x4
x1 2x2 + x3 + x4
=1
=1
Express the solution in form of column vectors where the column vector whose components
are dependent variables is represented by a linear combination of constant column vectors.
For each system, check in case the system is consistent that:
(rank of matrix of coefficients) + (number of free variables)=(number of variables)
1.3
1.3.1
15
number.
Write down the formula of expanding determinant (of oder n) along the column j.
b.Use properties of determinant to show that all the determinants following are zero
2 1
0 0
2 1 3
2 1 3
321 1 3
0 1 3
2 1 3
21 1 3
a
1 b 1 3 b
1 c + 1 2 c
2 d 2 4 d
1 a 1
0 1 3
2 1 3
4 3 9
2 1
2
3
1
1 1
2 2
0
1
0
2 3
1
0
1
3 2 0
11 21 99 2
1
Ex2. Use the method introduced in Practice to find the determinant of 3 3 matrices
A=
2
1
3
3
3 1
B=
2
0
0
3
C=
6
3
1
0 7 2
Ex3. Compute the determinant of matrices by using Elementary Row Operations to change
matrices into the triangular form.
2 1 2
2 3
1
3 2
1
3
3
0
2 3
0
5
1
2 3
2 1
3
3 2 1
5 3 1 0
2 1 4
2 3 0
1
0 0
Ex4. Expand the determinants along a appropriate row or column then compute them
2 3
0 2 1
2 2 2
2 2 1
0 0 1
0 2
2 1
11 0 1
23 0 0
0
0
0
0
2 3
3
0
0
1
0
5
6 2 3 9
1 3 1
2
4
2
16
b. Try to prove
A O
= |A||C|
B C
numbers below
3 0 0
4 0 0
and
2 1 1
1 2 2
4
0 0 5 4
3 2 1 1
2 1 2 4
2 3 2 1
2 4 2 3
0 0 3 2
0 0 4 3
2 3 2 4
2 4 1 1
3 2 0 0
2 1 0 0
0
2 3
2 4
1 1
.
2 2
0 3
O B
C A
and
and
and
= |C||B|
3 2 3 4
.
2 1 5 4
2 3
2 4
3 2
.
4 3
3 2 2 4
.
2 1 1 1
A B
O C
= |A||C|
A B
C O
= |C||B|
17
Proof. Suppose that A = (aij )1i,jn . Consider the first determinant, the others prove
similarly.
Step 1. Expand the first
A O
det(D) =
B C
where D1j (1 j n) is the determinant of order 2n 1 given from the determinant det(D)
by deleting row 1 and column j.
A1j O(n1)n
where A1j , B 0
However, for each j = 1, 2, ..., n, D1j is also in form D1j =
0
C
B
are submatrices of A, B correspondingly, especially A1j is size of (n 1) (n 1).
Step 2. Expand each determinant D1j (of order now going down to 2n 1) along the first
row of A1j . For instance, with the first one, D11 , of such determinants, we have
0
0
0
D11 = a22 D22
+ (1)a23 D23
+ . . . + (1)2+n a2n D2n
0
A2k O(n2)n
0
(k = 2, ..., n) and A0 is a (n 2) (n 2) submatrix of A11 .
where D2k =
2k
00
C
B
Continue this process until Step n 1, then Step n,
Step n 1. Each expansion of the determinants (of order n + 2) in this step is in form
anr O1n
anl O1n
a(n1)q
+ (1)a(n1)p
C
C
Bn1
Bn1
Step n. Each expansion of the determinants (of order n + 1) in this step is in form
anr O1n
= ans |C|
C
Bn1
Finally, combine n steps, we get
X
det(D) =
"
=
#
X
where each ordered ntuple (j1 , j2 , ..., jn ) is a permutation of {1, 2, ..., n} and (j1 , j2 , ..., jn ) =
0 or 1, depending on (j1 , j2 , ..., jn ).
18
Recommand. Expand the determinant following and all its subdeterminants along their first
row to make clear about a permutation and the function .
a11 a12 a13 a14
a
21 a22 a23 a24
a31 a32 a33 a34
a41 a42 a43 a44
Note. Since this proof, we can imply that the formula
A O
= |A||C|
B C
also works in case A is a m m matrix, C is a p p matrix. That means that A and C must
not have the same size.
c. The formula below for the determinant of a block matrix is incorrect
A B
= |AD| |CB|
C D
where A, B, C, D
1 2 3
2 1 4
1 3 1
3 2 0
1
1 2
1 7
1 3
3 4
det
and det
2 1
0 5
3 2
4 1
7
5
2 3
3 4
2 3 1
2
3 4 2
0
2 1 0
4
2 1
0
1 3
2 1
1
and
and
2 3 2 1
3 4 0 2
2 3 1 2 1 3
3 4 2 0 2 1
2 1 0 4 1 3
19
Proof. Suppose that A = (aij )1i,jn , B = (bij )1i,jn . Consider the determinant following
(of order 2n)
A O
In B
By Ex7, this determinant is |A||B|. Morever, using Elementary Row Operations to transform
the determinant by the way following
Step 1. Multiply row n + i of the determinant by a1i then add this product into row 1 for
each i = 1, 2, ..., n. At the end of this step, row 1 of the determinant becames
h
i
0 0 ... 0 ...
{z
}
|
{z
}
|
n components n components
where the last n components is just the row vector
a11 B1 + a12 B2 + . . . + a1n Bn = (row 1 of A) times B = row 1 of AB
Here, Bj (j = 1, 2, ..., n) is the row j of matrix B.
Step 2. Multiply row n + i of the determinant by a2i then add this product into row 2 for
each i = 1, 2, ..., n. At the end of this step, row 2 of the determinant becames
h
i
0 0 ... 0 ...
|
{z
}
|
{z
}
n components n components
where the last n components is just the row vector
a21 B1 + a22 B2 + . . . + a2n Bn = (row 2 of A) times B = row 2 of AB
Continue this process up to Step n. Finally, we get the new determinant
O AB
In B
whose value is the same as before. However, by Ex7, we get the value of the new determinant
equal to | In |.|AB| = |AB|. So,
A O O AB
= |AB|
|A||B| =
=
In B In B
20
Row Operations
1
1 4
16
1 94 942
1 184 1842
1.3.2
1
a5
a25
a35
a45
a1 a2 a3 a4
a21 a22 a23 a24
a31 a32 a33 a34
a41 a42 a43 a44
Operations
a1
a21
a31
a41
1
2 3 4
5
4 9 16 25
8 27 64 125
1
Ex1. Solve the system using the determinant if it has only one solution
3x y
x 3y
x 3y = 2
= 2
(3)
(2)
(1)
9x 3y
2x 6y = 4
2x + y = 3
x 3y + z
(i) 2x + y 3z
2y z
(a)
=2
=3
=0
2x + y + z
(ii)
x + y 3z
x 2y + 10z
2x1
+ x3 2x4
x1 + x2 3x3
3x1 2x2
5x2
+ 3x4
+ 2x4
=4
= 3
= 1
=5
(b)
=2
=1
= 1
=1
=0
x 3y
(iii) 3x + 2y 2z
2x + 5y 2z
5x1
+ x3 2x4
x1 + x2 3x3
3x1 2x2
5x2
+ 3x4
+ 2x4
=3
=4
= 3
=4
= 3
= 1
=5
21
Ex2. Use Cramers rule to show that all homogenous linear systems Ax = On1 , where A
is a square matrix of nonzero determinant, have only one solution which is the zero solution
x = On1 .
Ex3. Consider the case where Cramers rule fails to solve the homogenous linear system Ax =
On1 , where A is n n matrix. That means that det(A) = 0. For example the homogenous
system
2x1 + x2 + 3x3
3x1 3x2 + x3
3x1
+ 10x3
=0
=0
=0
=0
4x2 + x3 + 6x4
x + x + x + x
1
2
3
4
=0
=0
=0
Show that two these systems have infinitively many solutions. Show that in general case, such
systems have infinitively many solutions.
Ex4. Give of linear system Ax = b, where A is a square matrix of size n. Observe about
the relationship between the number of solutions and all determinants det(A), det(Aj ) (j =
1, 2, ..., n) when n = 2. Then give a prediction of that relationship in general case, as n 3.
Using Gaussian Elimination to solve the systems (i) (iii) and (a) (b) in Ex1 which do not
have a unique solution to check your prediction.
Ex5. Determine such that the systems below have a unique solution.
x ( 1)y + 2z + t
2y
+
2z
=
4
3x +
y + 2z
3x
+
y
z
=
3
2x
+z
2x 3y + z = 1
3y + 2z t
=2
=3
=2
=5
Ex6. Show that the equation of the line through the distinct points (a, b) and (c, d) can be
written
x y 1
a b 1
c d 1
=0
(a1 , b1 ) are collinear if and only if
=0
22
Ex8. Show that the equation of the plane passing through three noncollinear points (a1 , b1 , c1 ),
(a2 , b2 , c2 ) and (a3 , b3 , c3 ) can be represented as
1.3.3
a1 b 1 c 1
a2 b 2 c 2
a3 b 3 c 3
1
1
=0
1
1
Invertible Matrix
Ex1. Find the matrix inverse (if it exists) of the matrices below by using two different
methods
1 2 0
A=
3 0 1
2 1 2
0 1 3
B=
0 0 1
2 1 3
a b
D=
c d
1 3
3 0 1 0
C=
0
1
3
3
1 1 2 2
1 1
E=
1
1
3 2
1
1
1
this equality.
b. Find A1 where A =
0 1
0
(We can check this equality)
3
2
1
and we know that A + A 5A 2I3 = O.
2
Ex3. Solve the linear systems below by finding the inverse of the matrix of coefficients for
each system.
2x 3y
=1
3x + y
=2
x 3y + z
2x + y 3z
2y z
=2
=3
=0
23
(AB)1 = B 1 A1
1
0
0
1 1
0
B=
0 1 1
0
0 1