200237 Mathematics for Engineers 1

WEEK 5

MATRIX ALGEBRA Continued

Things I MUST do:

 Read through these notes before my lecture.

 Read Chapter 5: Pages 387-395.

LEARNING OUTCOMES
At the end of this section students should be able to:

 Use Gaussian elimination to solve a set of linear equations.

 Use row operations to find the inverse of a matrix.

 Find the eigenvalues and corresponding eigenvectors for a matrix.

Matrix Algebra Continued 1

 MATRIX ALGEBRA Continued
Reduction Methods

Matrix and determinant methods for solving simultaneous equations are generally inefficient.
Another approach, known as reduction or elimination methods, is more efficient, and can handle
those cases where the previous matrix and determinant methods fail.

When do matrix and determinant methods fail?

Well, matrix and determinant methods for solving simultaneous equations rely on det A not
equalling zero. If det A = 0, then other methods (that is, reduction methods) must be used to
solve the system of linear equations.

Gaussian Reduction

Reduction methods are a systemisation of the method of elimination. Consider for example the
following system of linear equations:

x  2y  z  1 (1)
2x  3y  z  3 (2)
3x  y  2 z  2 (3)

To solve using the elimination method, we can begin by multiplying equation (1) by 2 and
subtracting equation (2) to eliminate x. Similar steps are then taken to eliminate either y or z,
leaving one equation in one unknown to solve. Using substitution, we can then find the values of
the other variable(s).

These calculations can be carried out by working with the array of coefficients and constants,
known as an augmented matrix. That is, we consider
1 2 1 1
 
2 3 1 3
3 1 2 2 

The procedure is to use row operations on the array to reduce it to one in which all the elements
below the main diagonal are zero. The eliminations are carried out using row operations
according to the following rules for what is allowed:

1. Multiplication or division of a row by a constant.

2. Interchange of two rows.
3. Addition/subtraction of one row multiplied by a constant to/from another row.

The sequence of eliminations can now be written out as follows: begin with the array

1 2 1 1
 
2 3 1 3
3 1 2 2 


Matrix Algebra Continued 2

Now we perform row operations on this array as follows:

1 2 1 1
 
0 1 1 1 R2 : R2  2 R1
0 7 5 1  R3 : R3  3R1

1 2 1 1
 
0 1 1 1
0 0 2  8  R3 : R3  7 R2


This array can now be used to find z, then y, and finally x. That is:

From Row 3: 2 z   8  z   4 .

From Row 2: y  z  1  y  4  1  y   3 .

From Row 1: x  2 y  z  1  x  6  4  1  x  2  1  x   1

In summary, the method is to use row operations on the array of coefficients so as to reduce the
array to one in which all elements below the main diagonal are zero. This method is referred to
as Gaussian elimination.

Basic Method

For a 3  3 system:

Step 1 : If possible, get a “1” in the top left-hand corner.

1 . . .
 
. . . .
. . . . 


Step 2 : Get zeros as the first term in Rows 2 and 3, using Row 1.

1 . . .
 
0 . . .
0 . . . 


Step 3 : Get a zero as the second element of Row 3, using Rows 2 and 3.

1 . . .
 
0 . . .
0 0 . . 


Step 4 : Use Row 3 to find z, then Row 2 to find y, and Row 1 to find x.

Matrix Algebra Continued 3

Examples

1. Solve, using Gaussian elimination, the following system of linear equations:

2 x  3 y  4 z   19
6x  4 y  2z  8
x  5 y  4 z  23
Solution

Rewrite the three equations in a different order:

x  5 y  4 z  23
6x  4 y  2z  8
2 x  3 y  4 z   19

When we now write the first array, a “1” appears as the first element in Row 1; that is, we get

1 5 4 23 
 
6 4 2 8
2 3 4 19 

We now proceed to “get zeros” as the first elements in Rows 2 and 3:

NOTE: The first element in Row 2 is 6 times the first element in Row 1, and the first element in
Row 3 is twice the first element in Row 1; so the appropriate row operations will be:

1 5 4 23 
 
 0  26  26 130  R2 : R2  6 R1
 0 13  4  65  R3 : R3  2 R1
 
To get a zero as the second element in Row 3, two further row operations are needed.

1 5 4 23 
 
0 1 1 5 R2 : R2   26 
 0 13  4  65 

NOTE: The second element in Row 3 is – 13 times the second element in Row 2, so:

1 5 4 23 
 
0 1 1 5
0 0 9 0  R3 : R3  13R2


From Row 3, we have 9 z  0 , so z  0 . From Row 2, we have y  z  5 , so y  5 . Finally

from Row 1, we have x  5 y  4 z  23 , so x  25  23 , so x   2 .

Hence the solution is x   2 , y  5 and z  0 .

Matrix Algebra Continued 4

2. Solve, using Gaussian elimination, the following system of linear equations:

4 x  2 y  3z   1
3x  2 y  2 z  5
x  3y  2z  1
Solution

Rewrite the three equations in a different order

x  3y  2z  1
3x  2 y  2 z  5
4 x  2 y  3z   1

When we now write down the first table, a “1” appears as the first element in Row 1; that is,

1 3 2 1
 
3 2 2 5
4 2 3 1 


We now proceed to “get zeros” as the first elements in Rows 2 and 3:

1 3 2 1
 
0 7 4 2 R2  R2  3R1
0 14 11  5  R3  R3  4 R1


Now to get a zero as the second element in Row 3, only one further row operation is needed:

1 3 2 1
 
0 7 4 2
0 0 3  9  R3  R3  2 R2


From Row 3, 3 z   9 , so z  3 .

From Row 2, 7 y  4z  2
7y  12  2
7y  14
y 2

From Row 1, x  3 y  2 z  1
x  6  6 1
x 1

Hence the solution is x  1 , y  2 and z  3 .

Matrix Algebra Continued 5

Non-Unique Solutions

A set of equations where the determinant of the coefficient matrix is equal to zero will have a
non-unique solution. This means that when we solve the set of equations, we will either have an
infinite number of solutions or no solution at all.

Infinite Solutions – A Redundant System of Linear Equations

Examples

1. Find all solutions of the following system of linear equations:

x y z  4
2x  3y  z  3
3x  5 y  3z  2
Solution

 1 1 1 1 1 1
 
Here A   2  3 1  , and det A  2  3 1  1 9  5   1  6  3  1 10  9   0 .
 3 5 3  3 5 3
 

Hence we know that there is a non-unique solution for this set of equations. In fact this set of
equations reduces to (each student should check that they can do this):

1 1 1 4
 
0 1  3 5 
0 0 0 0 


From R3 : 0 x  0 y  0 z  0  0  0 .

This is a true statement, but not useful in determining a solution! This means that the set of
equations are redundant, meaning that there is no unique solution for x, y and z. In fact a
redundancy tells us that there are an infinite number of solutions that can be generated that will
satisfy the given equations.

To solve, let one of the unknowns be a parameter (for example, let z  a ), and express the
others in terms of this.

So, letting z  a , then we find that…

From Row 2,  y  3 z   5
 y  3a   5
y  3a  5  y  5  3a

From Row 1, x  y  z  4
x   5  3a   a  4
x  5  3a  a  4  x  5  4a  4  x  9  4a

Matrix Algebra Continued 6

Hence the general solution is given by x  9  4a , y  5  3a , z  a .

A particular solution is found by substituting any value for a. For example, we could let a  0 :
then x  9 , y  5 , z  0 .

Or let a   1 : then x  13 , y  8 , z   1 , etc.

2. Solve the system of linear equations below.

x  2 y  3z  6
2x  y  4z  2
4 x  3 y  2 z  14
Solution
1 2 3 6
 
2 1 4 2
4 3 2 14 

Now we perform row operations on the augmented matrix as usual:

1 2 3 6
 
0  5 10 10  R2  R2  2 R1
0  5 10 10  R3  R3  4 R1


1 2 3 6
 
 0  5 10 10 
0 0 0 0  R3  R3  R2


Let z  a ; then from Row 2, we find that

 5 y  10 z   10
 5 y  10a   10
y  2a  2  y  2  2 a
But now from Row 1:
x  2 y  3z  6
x  2  2  2a   3a  6
x  4  4a  3a  6  x  4  a  6  x  2  a

Hence the infinite number of solutions can be generated by letting a equal any number in

x  2a 

y  2  2a 
z  a 


Matrix Algebra Continued 7

Gaussian reduction can be used on systems of any size.

Example

Show that the system

x1  2 x2  3x3  x4  0
x1  3x2  x3  x4  2
2 x1  x2  x3  3x4  3
4 x1  4 x2  x3  5 x4  5

has infinitely many solutions, and find all solutions.

Solution

The initial matrix is as follows.

 1 2 3 1 0
 
 1 3 1 1 2
 2 1 1 3 3
 
 4 4 1 5 5 

Now we do the usual row operations.

1 2 3 1 0
 
0 1 4 0 2 R2  R2  R1
0 5 7 1 3 R3  R3  2 R1
 
0 4 11 1 5  R4  R4  4 R1

 1 2 3 1 0
 
 0 1 4 0 2
 0 0 27 1 13  R3  R3  5 R2
 
 0 0 27 1 13  R4  R4  4 R2

 1 2 3 1 0
 
 0 1 4 0 2
 0 0 27 1 13 
 
 0 0 0 0 0  R4  R4  R3

Let x4  a . Then from Row 3:

27 x3  x4  13
27 x3  a  13
13  a
x3 
27

Matrix Algebra Continued 8

From Row 2:
x2  4 x3  2
 13  a 
x2  4    2
 27 
 13  a  2  4a
x2  2  4   
 27  27
Finally from Row 1:
x1  2 x2  3 x3  x4  0
 2  4a   13  a 
x1  2    3 a  0
 27   27 
35 38a
x1    0
27 27
35  38a
x1 
27
Hence the general solution is
35  38a
x1 
27
2  4a
x2 
27
13  a
x3 
27
x4  a

No Solution – An Inconsistent System of Linear Equations

Examples

1. Find all solutions, if they exist, of the following set of equations:

x y  z  4
2x  3y  z  3
x  2 y  2z   2
Solution

1 1 1
Here det A  2 3 1  1 6  2   1  4  1  1  4  3   0 : non-unique solution! The
1 2 2
augmented matrix is
 1 1 1 4
 
 2 3 1 3 .
 1 2 2 2 
 

Matrix Algebra Continued 9

This reduces to
1 1 1 4
 
0 1  3 5 .
0 0 0  1 


The last row gives 0x + 0y + 0z = – 1, i.e. 0 = – 1, which is impossible. This means the
equations are inconsistent, and there is no solution.

2. Show that the following system of equations has no solution.

x  2 y  3z   1
3x  y  2 z  7
5x  3 y  4 z  2
Answer

First we write down the augmented matrix, then we row reduce it as always.

The augmented matrix and relevant row operations are written below.

1 2 3 1 
 
3 1 2 7
5 3 4 2 


1 2 3 1 
 
 0 7 11 10  R2  R2  3R1
 0 7 11 7  R3  R3  5 R1


1 2  3 1 
 
 0 7 11 10 
0 0 0  3  R3  R3  R2


Row 3 says 0  x  0  y  0  z   3 , i.e. 0 = – 3. This is impossible; therefore this system

of linear equations is inconsistent and has no solution.

Unlike the determinant and matrix methods of solving systems of equations, the reduction
methods can be used when the number of unknowns and equations are not equal. By using row
reductions to produce zeroes below the diagonal, we can find the solutions of the set of equations
as before. When there are more equations than unknowns, the equations are usually inconsistent
and so there will be no solution. When there are more unknowns than equations the equations
will usually be redundant, and there will be an infinite number of solutions.

Matrix Algebra Continued 10

Example

1. Solve the following set of equations:

x  2y  z  4
2x  y  z  3
 x  3y  2
x  y  2z  5
Solution

The augmented matrix and relevant row operations are below.

 1 2 1 4
 
 2 1 1 3
 1 3 0 2
 
 1 1 2 5 

 1 2 1 4 
 
 0 5 1 5  R2  R2  2 R1
 0 5 1 6  R3  R3  R1
 
 0 1 1 1  R4  R4  4 R1

 1 2 1 4 
 
 0 0 0 1 
R2  R2  R3
 0 5 1 6 
 
 0 1 1 1 

But now R2  0 x  0 y  0 z  1 , an inconsistency; therefore there is no solution.

2. Solve the system of equations below.

x  y  2 z  2t   1
2x  y  z  t   1
3x  y  z  2t   2
Solution

NOTE: This system of equations has four unknowns and three equations. There will be an
infinite number of solutions for this system, even though it turns out that the reduction does not
lead to a redundancy. Here is the augmented matrix:

 1 1 2 2 1 
 
2 1 1 1 1  .
3 1 1  2  2 


Matrix Algebra Continued 11

The row operations and resulting tables are below.

 1 1 2 2 1 
 
 0 3 5 3 1 R2  R2  2 R1
 0 4 7 4 1  R3  R3  3R1


 1 1 2 2 1 
 
 0 3 5 3 1
 0 1 2 1 0  R3  R3  R2


 1 1 2 2 1 
 
 0 1 2 1 0
 0 3 5 3 1  R2  R3


 1 1 2 2 1 
 
 0 1 2 1 0
0 0 1 0 1  R3  R3  3R2


From Row 3, z  1 .

Now from Row 2, y  2 z  t  0 . Let t  a ; then Row 2 becomes y  2 z  a  0 . As z  1 ,

this becomes y  2  a  0 , and so y  2  a .

From Row 1:
x  y  2 z  2t   1
x   2  a   2  2a   1
x  2  a  2  2a   1
x  a 1

Hence the general solution is given by

x  a 1 
y  2  a 

z  1 
t  a 

Homogeneous Linear Systems

Definition: A system of linear equations is said to be homogeneous if all the constants on the
right-hand side are zero.

Matrix Algebra Continued 12

For example
x y  0
2x  y  0

is a homogeneous set of equations, whereas the following is not:

x  2y  z  t  0
2x  y  z  t  1
 x  y  3z  t  0
2x  3y  2z  t  0

A homogeneous system of equations always has one solution, namely the solution with all
unknowns equal to zero (known as the trivial solution). Thus a homogeneous system of
equations cannot be inconsistent. When solving homogeneous linear systems, there are only two
possibilities:

1. The system has only one solution, the trivial solution; or

2. The system has an infinite number of solutions in addition to the trivial solution.

Test for Non-Zero Solution for Homogeneous Equations

It can be shown that for n homogeneous linear equations to have a non-trivial solution, we must
have det A  A  0 .

For example, consider the following:

ax1  by1  cz1  0

ax2  by2  cz2  0
ax3  by3  cz3  0

These equations will have a non-trivial solution provided that

a1 b1 c1
a2 b2 c2  0 .
a3 b3 c3

Example

Solve the following system of linear equations using Gaussian reduction:

3x  y  3z  0
x  4 y  2z  0
3x  10 y  12 z  0

Matrix Algebra Continued 13

Solution

First, check to see if this homogeneous system of equations will have a non-trivial solution; that
is, find the value of det A.

3 1 3
4 2 1 2 1 4
det A  1 4 2  3 1  3
 10 12 3 12 3 10
3 10 12
 3    48  20   1    12  6   3    10  12 
 0

As det A = 0, the equations will have a non-trivial solution. Therefore, we now know that an
infinite set of solutions exists, and so using reduction we should get a redundancy.

Now we can begin the row operations. Remember to rearrange the equations to get a “1” as the
first element in row 1; that is,
x  4 y  2z  0
3x  y  3z  0
3x  10 y  12 z  0

This can be written as

1 4 2 0
 
3 1 3 0
 3 10 12 0 


Now, using row operations:

1 4 2 0
 
 0 11  9 0 R2  R2  3R1
 0  22 18 0  R3  R3  3R1


1 4 2 0
 
 0 11  9 0
0 0 0 0  R3  R3  2 R2


As previously, we know that this system will have an infinite number of solutions, as the
reduction has led to a redundancy.

Letting z = a, from Row 2 we find that

11y  9a  0
9a
y  
11

Matrix Algebra Continued 14

Then from Row 1,
x  4 y  2z  0
36a
x  2a  0
11
14a
x 
11

Hence the infinite number of solutions are of the form

14a
x
11
9a
y
11
z a

Inversion of Matrices by Reduction

We can use reduction methods to find matrix inverses. Essentially the method is as follows:

 a11 a12 a13 

 
Let A   a21 a22 a23  , for example. Then starting with the array
a a33 
 31 a32
 a11 a12 a13 1 0 0
 
 a21 a22 a23 0 1 0
a a32 a33 0 0 1 
 31

and using row reduction methods, we finally end up with

1 0 0 b11 b12 b13 

 
0 1 0 b21 b22 b23 
0 0 1 b31 b32 b33 


 
where the b 's are the elements in the inverse A 1 of matrix A.

That is,
 b11 b12 b13 

1 
A   b21 b22 b23  .
b b33 
 31 b32

Matrix Algebra Continued 15

Example

 2 1 
Find the inverse of A   .
5 3
Solution

To find A1 , we begin as follows: look at the array

 2 1 1 0
 
5 3 0 1

Do row operations to get a 1 in the top left-hand corner:

 1 1 
1  0
R1 
R1
 2 2 
5 3 0 1 
2


Now we reduce the matrix on the left to the 2  2 identity matrix:

 1 1 
1  0
 2 2  R2  R2  5R1
0 11

5
1 

 2 2 

 1 
 1 1 2
0 
 2  2
0 5 2  R2  R2
 1   11
 11 11 

1 0 3 1 
  1
 11 11 
R1  R1  R2
 
5 2  2
 0 1 
 11 11 

 3 1 
 11 11 
Hence A1   .
5 2 
 
 11 11 

Check: AA1  A1 A  I ? The student should verify this!!

Matrix Algebra Continued 16

 Eigenvalues and Eigenvectors
In many technical problems, equations of the form AX   X occur frequently. Here A is a
square matrix,  is a scalar, and X is a non-zero column matrix or vector that looks like:

 x1 
 
 x2  .
 
 
 xn 

The values of  are called the eigenvalues of the matrix A, and the corresponding solutions (X)
are called the eigenvectors of A.

Example
a b  x1 
In the 2  2 case, suppose A    and X    . Then the equation becomes:
c d   x2 

 a b   x1   x1 
      
 c d  x2   x2 
or
ax1  bx2   x1
cx1  dx2   x2
In other words,
 a    x1  bx2  0
cx1   d    x2  0
As a matrix equation, this is
a b   x1  0
     
 c d     x2  0
or
 A  I  X  0.

This is then a set of homogeneous equations, which has only the trivial solution of X  0 , or an
infinite number of solutions including the trivial solution.

Procedure for Finding Eigenvalues and Eigenvectors

The matrix equation AX   X , which can be rewritten in the form  A   I  X  0 , represents

a homogeneous system of n equations in n unknowns. Now we know we have the trivial solution,
namely X = 0; but we are looking for nontrivial solutions for X. The system will have a nontrivial
solution if and only if the coefficient matrix  A   I  is singular; that is, if and only if

det  A   I   0 .

Matrix Algebra Continued 17

So to find the eigenvalues  , we solve the characteristic equation det  A   I   0 ; and to find
the corresponding eigenvectors X, we set up an augmented matrix and use reduction to solve the
set of homogeneous equations  A   I  X  0 .

A trivial solution to  A   I  X  0 would be X = 0.

Remember, X will have an infinite number of solutions. Can you say why?

Examples
1 5
1. Find the eigenvalues and corresponding eigenvectors for the matrix A   .
2 4
Solution

To find the eigenvalues , solve the characteristic equation for  ; so solve A  I  0 , i.e.

1  5
 0.
2 4

Evaluating this determinant, we obtain

1    4     10  0
4  5   2  10  0
 2  5  6  0
   6    1  0
Hence   6 or    1 .

To find the eigenvectors, solve  A   I  X  0 for X, using each value of  in turn.

Letting   6 and solving  A   I  X  0 , the augmented matrix will be:

1 6 5 0
 
 2 46 0

We know that there will be an infinite number of solutions; so using row reductions as before,
we should obtain a redundancy from

 5 5 0
 .
 2 2 0

You check that this reduces to

0 0 0
 .
1 1 0

Matrix Algebra Continued 18

So from Row 2, we get:
x1  x2  0  x1  x2 .

 x1   x1   1
Hence X        x1   is the general form of the eigenvector corresponding to the
 x2   x1   1
eigenvalue   6 .

The eigenvector is then often given as just the vector without the constant multiple out the front.
 1
In other words, we often say the eigenvector corresponding to the eigenvalue of 6 is X    .
 1
Now, letting    1 :  A   I  X  0 becomes

11 5 0 2 5 0
    .
 2 4 1 0 2 5 0

This obviously reduces to

2 5 0
 .
0 0 0

5
Thus 2 x1  5 x2  0  x1   x2 ; and the corresponding eigenvector is
2
  5 x2   5
 x1  
X   2   x   2 ,
 x2   x   
2

 2   1 

 5 
or just X   2  , or even X    5  (multiplying throughout by 2).
 1   
   2

2. Find the eigenvalues and corresponding eigenvectors for the matrix

6 2 5 
 
A 1 1 1 .
 8 2 7 

Solution

To find the eigenvalues , solve A  I  0 , i.e.

6   2 5
1 1  1  0.
8 2 7

Matrix Algebra Continued 19

Then the working looks as follows:

  6    1    7     2  2 7    8  5  2  8 1     0

  6    7  8   2  2   2     1  5  10  8   0
  6     2  8  9   2  2  50  40  0
 6 2  48  54   3  8 2  9  38  52  0
  3  2 2    2  0
 3  2 2    2  0

Now substitute values for  (remember they must be factors of the constant term 2); for example,
try   1 . Here 1  2 1  1  2  0 . So   1 is a solution.
3 2

Then try   2 . Here  2   2  2   2  2  0 . So   2 is a solution.

3 2

Finally try    1 . Here  1  2  1  1  2  0 . So    1 is a solution.

3 2

Hence the eigenvalues are   1 ,   2 and    1 .

Now to find the corresponding eigenvectors, solve  A   I  X  0 for the three (3) values of .

Letting   1 ,  A   I  X  0 becomes

 6  1 2  5   x1   0 
    
 1 11 1   x2    0 
 8 2 7  1   x3   0 


 7 2  5   x1   0 
    
 1 0 1   x2    0 
 8 2 6   x3   0 


Now do row operations on the augmented matrix below to solve this system of equations.
Remember that you should always end up with a zero row!

 7 2 5 0 0 2 2 0 R1  R1  7 R2
   
 1 0 1 0  1 0 1 0
 8 2 6 0   0 2 2 0  R3  R3  8 R2
 

0 2 2 0
 
1 0 1 0
0 0 0 0  R3  R3  R1


Matrix Algebra Continued 20

From Row 2, we obtain x1  x3  0 , so x1   x3 .

From Row 1, we obtain 2 x2  2 x3  0 , so x2   x3 .

 x1    x3   1 
     
So X   x2     x3   x3   1  is the general form of the eigenvector corresponding to the
x   x   1
 3   3  
eigenvalue   1 .

We now repeat this process with   2 and    1 . Letting   2 ,  A   I  X  0 becomes

 6  2 2 5   x1   0 
    
 1 1 2 1   x2    0 
 8 2 7  2   x3   0 


 8 2  5   x1   0 
    
 1 1 1   x2    0 
 8 2 5   x3   0 


Now do row operations on the augmented matrix below to solve this system of equations.

 8 2 5 0
 
 1 1 1 0
 8 2 5 0 


 0 6 3 0 R1  R1  8 R2
 
 1 1 1 0
 0 6 3 0  R3  R3  8 R2


 0 6 3 0
 
 1 1 1 0
 0 0  R3  R3  R1
 0 0

From Row 1, we obtain  6 x2  3x3  0 ; so x3  2 x2 .

From Row 2, we obtain x1  x2  x3  0 ; whence x1  x2  2 x2  0 , so x1   x2 .

 x1    x2   1 
     
So X   x2    x2   x2  1  is the general form of the eigenvector corresponding to the
 x   2x   2
 3  2  
eigenvalue   2 .

Matrix Algebra Continued 21

Finally, we repeat this process with    1 . Letting    1 ,  A   I  X  0 becomes

 6  1 2  5   x1   0 
    
 1 11 1   x2    0 
 8 2 7  1   x3   0 


 5 2  5   x1   0 
    
 1 2 1   x2    0 
 8 2 8   x3   0 


Now do row operations on the augmented matrix below to solve this system of equations.

 5 2 5 0
 
 1 2 1 0
 8 2 8 0 


 0 12 0 0 R1  R1  5 R2
 
 1 2 1 0
 0  18 0  R3  R3  8 R2
 0

 0 12 0 0
 
 1 2 1 0
 0 0  R3  R3  1.5 R1
 0 0

From Row 1, we obtain 12 x2  0 , so x2  0 .

From Row 2, we obtain x1  2 x2  x3  0 ; but x2  0 , so x1  x3  0 , and x1   x3 .

 x1    x3   1 
     
So X   x2    0   x3  0  is the general form of the eigenvector corresponding to the
x   x   1
 3   3  
eigenvalue    1 .

Matrix Algebra Continued 22

Tutorial Questions

 1 3 
1. Find the inverse of A    using reduction methods.
 2 4
2. Solve the following simultaneous equations by reduction.
x  2y  z  5
x  3y  2
(a) (b) 2x  y  z  0
2x  y   3
 x  y  3z  8

x  y  4z  1 x1  2 x2  3 x3  2
(c)  2x  7 y  6z  2 (d) 4 x1  5 x2  6 x3  8
x  9 y  8z  3 7 x1  8 x2  9 x3  14

3. Find the eigenvalues and corresponding eigenvectors for the following matrices.

 0 3 0  1 2 7 
 1  3     
(a)   (b)  1 2 0 (c)  0 1  3 
 2 4  1 1  0 2 
 1  0

 3 2 1
 
4. Show that  1 4 1  has an eigenvalue of   2 , and find the corresponding
 4 2 
 0
eigenvector.

 3 2 1
 
5. Given that   3 is an eigenvalue of  1 4 1  , find a corresponding eigenvector.
 4 2 
 0

ANSWERS

1  4 3 
Q1.   
10   2 1 

Q2. (a) x  1 , y  1 (b) x  4, y  3, z  5

(c) Inconsistent; that is, no solution (d) x1  2  a , x2   2a , x3  a

 x1    3a   3 
Q3. (a) Eigenvalue   1 , eigenvector       a 
 x2   2a   2
 x1    x2   1 
Eigenvalue   2 , eigenvector       x2  
 x2   x2   1

Matrix Algebra Continued 23

 0 0
   
Q3. (b) Eigenvalue   1 , eigenvector  0   x3  0 
x  1
 3  
 3x2   3
   
Eigenvalue    1 , eigenvector  x2   x2  1 
  2x   2 
 2   
  x2   1 
   
Eigenvalue   3 , eigenvector  x2   x2  1 
 0  0
   
 x1  1
   
(c) Eigenvalue   1 , eigenvector  0   x1  0 
 0  0
   
 x2  1
   
Eigenvalue    1 , eigenvector  x2   x2  1 
 0  0
   
 9 x3   9
   
Eigenvalue   2 , eigenvector   x3   x3   1 
 x   1
 3  

 0   0  x1   1
       
Q4.  x2   x2  1  Q5.  5 x1   x1  5 
  2x   2    4x   4 
 2     1   

Extra Questions

TEXT: Page 370 5.5.3 Exercises, Question 74

Page 399 5.7.3 Exercises, Questions 96 and 97.

More Questions 6
Do you understand this topic?
1. Can we use reduction to solve every set of simultaneous equations?
2. If det A = 0, what type of solution(s) can you expect to a set of simultaneous equations?
3. What is: (i) a redundancy (ii) an inconsistency?
4. What is a homogeneous set of equations? How do you tell if that set of homogeneous
equations has a nontrivial set of solutions?
5. How do you find an eigenvalue and an eigenvector?
6. Any other questions from the lecture notes?

Matrix Algebra Continued 24

Questions

1. For the following set of equations:

x  y  2z  0
2 x  4 y  5z  0
6 x  13z  0
(a) Answer the questions below without using reduction techniques.
(i) What is one solution for this set of equations?
(ii) Explain why this set of equations has an infinite number of solutions.
(b) Verify by reduction that this set of equations has an infinite set of solutions.
(c) Find all solutions for this set of equations.

2. Three machines together produce 650 parts per hour. Twice the production of the second
machine is 10 parts per hour more than the sum of the production of the other two
machines. If the first machine operates for 3 hours and the other two for 2 hours, 1550 parts
are produced. Let x, y and z be the number of parts produced by the first, second and third
machines respectively. Write down a system of equations relating x, y and z, and solve this
system to determine how many parts are produced by each machine.

3. Use reduction to solve the system of equations

a1 x  b1 y  c1
a2 x  b2 y  c2
Show that the result obtained is the same as that obtained using determinants.

 4 6 6
4. (i) Determine the eigenvalues of the matrix A   0 1

0 .
 1 1 
 1

(ii) Find ONE corresponding eigenvector for the matrix given in (i).

Answers
 13a 6 
 
1. (i) x=y=z=0 (ii) det A = 0, so an infinite no. of solutions (c)  a 6 
 
 a 
x  y  z  650
c1b2  c2b1 c a  c1a2
2. 2 y  x  z  10  x  250, y  220, z  180 3. x  , y 2 1
a1b2  a2b1 a1b2  a2b1
3x  2 y  2 z  1550

3 2  6
4.    2, X   0  ;    1, X   0  ;   1, X   6 
   
1 1  1 
     

Matrix Algebra Continued 25