LINEAR ALGEBRA AND ANALYTICAL GEOMETRY

CHAPTER TWO
SYSTEMS OF LINEAR
EQUATIONS
2.1 INTRODUCTION
The theory of matrices has been successfully applied in diverse fields and areas such as mathematics,
engineering, social and biological sciences, economics etc. We shall apply this theory to a system of m linear
equations in n unknowns or variables. A finite set of linear equations in n unknowns x1, x2, …, xn is called a
system of linear equations or a linear system.
A general system of m linear equations in n unknowns x1, x2, …, xn can be written as
a11 x1 + a12 x2 + . . . + a1n xn = b1
a21 x1 + a22 x2 + . . . + a2n xn = b2
………………………………. (1)
……………………………….
am1 x1 + am2 x2 + . . . + amn xn = bm
where ai j and bj for i  1, 2,  , m; j  1,2, n are constants. These numbers ai j are called the coefficients of
the system. The system (1) is known as “Non-homogeneous System of Linear Equations”.
If each bi in (1) is zero, the system becomes:
a11 x1 + a12 x2 + . . . + a1n xn = 0
a21 x1 + a22 x2 + . . . + a2n xn = 0
………………………………. (2)
……………………………….
am1 x1 + am2 x2 + . . . + amn xn = 0
The system of equations (2) is called the “Homogeneous System of Linear Equations”. A sequence of n
numbers s1, s2, s3, …sn, for which (1) is satisfied when we substitute x1 = s1, x2 = s2, …,xn = sn, is called a
solution of (1). The set of all such solutions is called solution set or the general solution of (1).
A system of equations that has no solution is said to be inconsistent. On the other hand, if there exists at least
one solution of the system, the system is called consistent system.
The system (2) always possesses one solution, namely the zero n-tupled (0, 0, 0, … , 0) solution called the zero
or trivial solution. Any other solution, if it exists, is called a non-zero or non –trivial solution.
NOTE: The following examples will help the readers to understand the concept of consistent and
inconsistent systems. The method is based on school algebra and does not need matrix concepts.
2.1.1 Geometric Interpretation and Existence of Solution
From analytical geometry, we know that an equation ax  by  c represents a straight line. Now consider two
equations in two unknowns x and y:
a11 x + a12 y = b1
a21 x + a22 y = b2
Each equation represents a straight line in the plane R2. There are three possible cases:
(a) No solution if the lines are parallel.
(b) Precisely one solution if lines intersect.
(c) Infinitely many solutions if lines coincide. For instance,
x+y=1 x+y=1 x+y=1
x+y=0 x -y=0 2x +2 y = 2

1 1 1

Case (a) No Solution exists Case (b) Precisely one solution exists Case (c) Infinitely many solutions exist.

Page 1 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
Mathematically speaking:
a11 a12 b1 a11 a
(a) No solution exists if   (b) Precisely one solution if  12
a21 a22 b2 a21 a22
a11 a12 b1
(c) Infinitely many solutions if  
a21 a22 b2
EXAMPLE 01: Solve the system of linear equations:
(a) 3x  2 y  12 (i )
7x  4 y  2 (ii ) .
Multiply equation (i) by 2 and add the two equations, we get
6 x  4 y  24
7x  4 y  2
________________
13 x  26  x  2 

Substituting x  2 into (i), we get 6  2 y  12  y  3  Thus, the system is “consistent” and has a unique
solution (x, y) = (2, 3).
3x  2 y  4  iii 
(b) .
6x  4 y  9 (iv)
Multiply equation (iii) by 2 and subtract it from equation (iv), we get

6x+4y=8
6x+4y=9
- - -
0 = -1 (This is not possible)
This shows that the given system is “inconsistent” and hence possesses no solution.
x y  2 (v )
(c)
2x  2 y  4 (vi)
If we multiply equation (v) by 2, we shall get an equation that is identical to equation (vi). This means that
equations (v) and (vi) are equivalent or identical. If we substitute x = 1 and y =1, both equations are satisfied.
Therefore, (x, y) = (1, 1) forms a solution of the above system. You will notice that there are many other pairs of
solutions, e.g.;
x  2 and y  0, x  3 and y  1, x  4 and y  2 and so on.
This shows that this system of equations is “consistent and has infinitely many solutions”.
System of linear equations helps us to solve practical problems. The following example shows one of the
applications. We shall discuss many application problems in the coming sections.
EXAMPLE 02: If 20 pounds of rice and 10 pounds of potatoes cost $16.20 and 30 pounds of rice and 12
pounds of potatoes cost $23.04, how much will 10 pounds of rice and 50 pounds of potatoes cost?
Solution: Let the cost of 1 pound of rice be $x and the cost of 1 pound of potatoes be $y. Then according to the
question,
20 x  10 y  16.20 i
30 x  12 y  23.04  ii 
Multiply (i) by 6 and (ii) by 4, we have
120 x  60 y  97.2  iii 
120 x  48 y  92.16  iv 
Subtracting (iv) from (iii), we have
12 y  5.04  y  0.42 
Substituting this value of y into (i), we obtain
20 x  10 0.42   16.20  x  0.6 

Thus, 10 pounds of rice cost $10  0.6  $6, and 50 pounds of potatoes cost $50  0.42  $21 

Page 2 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY

2.1.2 Matrix Notation of System of Linear Equations

Using the matrix notation, the linear system (1) may be written as:
Ax=b (3)
where A, x and b are matrices given by:
 a11 a12  a1n   x1   b1 
a a22  a2n  x  b 
A   21 , x  2  and b   2  
         
     
am1 am2  amn   xn  bm 
The matrix A is called the system matrix, the right side constants bi form a column vector b and the unknowns
xi form the column vector x.
 a11 a12  a1n b1 
a  a2n b2 
The matrix Ab  
21 a22
      
 
am1 am2  amn bmn 
is called the augmented matrix of the system (1). We see that Ab is obtained by augmenting the matrix A by
column vector b. The matrix Ab determines the system (1) completely, because it contains all the given
numbers appearing in (1). The unknowns xi are immaterial because instead of xi if we use some other variable,
the solution of the system will not change.
2.2 ROW EQUIVALENT AUGMENTED MATRICES
Let A b be the augmented matrix for the system of equations (1) and Cd be its row equivalent matrix, obtained by
the following three operations:
(i) Any two equations of (1) are interchanged. This is shown by writing Ri j.
(ii) Any equation of (1) is multiplied by a non-zero constant. This is shown by writing c Ri.
(iii) A constant multiple of an equation (1) is added to another equation. This is shown by writing
Ri + c Rj.
The three operations listed above are called elementary or admissible operations for the system (1). The
resulting system of equations will be equivalent to the previous one, and therefore have the same solution. Thus
by transforming the augmented matrix Ab into row equivalent augmented matrix Cd, we can easily solve the
given system of equations. The two methods namely:
(i) GAUS’S Elimination Method and
(ii) GAUS’S-JORDAN Methods
are used to solve the system (1). These methods are known as “Direct Methods” or “Analytic Methods.”
2.2.1 Gauss Elimination Method
One of the several methods employed to solve a system of m linear equations in n variables, is known as
Gaussian elimination method named after its inventor, the famous German mathematician Carl Friedrich Gauss
(1777-1855). Let us consider the linear system (1) of section 2.1
In matrix form it can be expressed as A x = b, where A is the matrix of coefficients of order m × n, x and b are
column matrices/vectors of order n × 1 and m × 1 respectively. The following four steps can be applied to solve
a system using Gauss’s elimination method.
Step1. Change the system of linear equations to the form A x = b.
Step2. Form the augmented coefficient matrix Ab by including the elements of b as an extra column in the
matrix A.
Step3. Convert the augmented matrix into echelon form by using elementary row operations.
Step4. Find x by detaching the fourth column back to its original position on the right hand side of the matrix
equation A x = b. This procedure is known as “Backward Substitution”.

Page 3 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY

NOTE: (i) A system A x = b is called over-determined if it has more equations than unknowns i.e. m > n,
determined if the number of equations is equal to the number of unknowns, i.e. m = n and
underdetermined if A x = b has fewer equations than unknowns i.e. m < n.

(ii) When an augmented matrix is written in the row echelon form, the variables corresponding to the
first non-zero elements (or leading elements) in each row are called leading or pivotal variables. All
other variables are called free or non-leading variables.
(iv) If m = n and |A| ≠ 0, then system (1) can be solved by
(a) Cramer’s rule discussed in Chapter 3.
(b) Using the inverse matrix method x = A-1 b.

EXAMPLE 01: Use Gauss’s elimination method to solve the system of linear equations (m = n)
x1 + 5 x2 + 2 x 3 = 9
x1 + x2 + 7 x 3 = 6
- 3 x2 + 4 x 3 = -2
Solution:
Step1. We change the system of linear equations in matrix form Ax  b 

1 5 2   x1   9 
A  1 1 7  , x   x  and b   6  
0  3 4   2  2
   
 x 3 
Step2. We form the augmented coefficient matrix Ab by including the constants, an extra column in the
matrix.
1 5 2 9 
Ab  1 1 7 6 
0  3 4  2 
 
Step3. We convert this augmented coefficient matrix into an echelon form using elementary row operations.
1 5 2 9  1 5 2 9
Ab  1 1 7 6  R 2    1 R1  0  4 5  3  R3    1 R2
0  3 4  2  0  3 4  2 
 
1 5 2 9  1 5 2 9
 0 4 5  3 R2  4 R3  0 0 1 1 R23
0 1 1 1  0 1 1 1
1 5 2 9
 
0 1 1 1


0 0 1 1

Step4. Find x by detaching the fourth column back to its original position on the right hand side of the matrix
equation Ax  b 
1 5 2   x1  9
0 1  1  x   1 
0 0 1   2  1
   x3   
x1 +5 x2 + 2 x 3 = 9 (i)
x2 - x3=1 (ii)
x 3 = -1 (iii)
From (iii), we get x3  1 . Substituting x3  1 into (ii), we get x2  1  1  x2  2  Now put
x3  1 and x 2  2 in (i), we get x  3  1

Thus, the required solution of the given system of equations is: x1 = 3, x2 = 2, x3 = 1.

EXAMPLE 02: A yarn merchant sells brands A, B, C of yarn each of which is a blend of Pakistani,
Egyptian and American cotton in the ratio 1:2:1, 2:1:1 and 2:0:2 respectively. If one kilogram of A, B, C
costs 40, 50 and 60 rupees respectively, find the cost of a kilogram of cotton of each country.
Page 4 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY

Solution: Let the cost of each kilogram of Pakistani, Egyptian and American cotton be x1 , x 2 and x 3 rupees
respectively. Then the above-mentioned problem is formulated into the following mathematical model
involving equations:
1 2 1
x1  x 2  x3  40
4 4 4
2 1 1
x1  x 2  x3  50 ,
4 4 4
2 2
x1  x3  60
4 4
which is equivalent to the following system of equations:
x1  2 x 2  x3  160
2 x1  x 2  x3  200 (multiply each equation by 4).
2 x1  2 x3  240
Step1. We change the system of linear equations in matrix form Ax  b 

1 2 1   x1  160 
A  2 1 1 , x   x  , and b   200 
 
2 0 2  2  240
   x 3   

Step2. We form the augmented coefficient matrix Ab by including the constants, an extra column in the
matrix.
1 2 1 160 
Ab   2 1 1 200 


 2 0 2 240 
Step3. We convert this augmented coefficient matrix into an echelon form using elementary row operations.
1 2 1 160 
A  2 1 1 200 R    2  R , R    2  R
b 2 1 3 1
2 0 2 240
1 2 1 160  1 2 1 160
 1
 
0 3 1  120
   1 R2 ,    R3  0 3 1 120 R23
  4
0 4 0  80 
 0 1 0 20 
1 2 1 160 1 2 1 160
 0 1 0 20  R3    3 R2  
0 1 0 20 

0 3 1 120 
0 0 1 60 

Step4. Find x by detaching the fourth column back to its original position on the right hand side of the matrix
equation Ax  b 
1 2 1  x1  160
0 1 0  x 2    20  

0 0 1  x3   60 
Expanding from top, we have
x1  2 x2  x3  160  i
0  x1  x2  0  x3  20  ii 
0  x1  0  x2  x3  60  iii 
Solving equations (i) through (iii) using backward substitution, we get
x1 = 60, x2 = 20 and x3 = 60. Thus, the cost of each kilogram of Pakistani, Egyptian and American cotton is
Rs.60/, Rs.20/ and Rs.60/ respectively.

Page 5 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
2.2.3 Gauss-Jordan Method
Gauss elimination process is based on:
(i) changing the augmented matrix Ab in triangular/echelon form and
(ii) making the backward substitution.

The Gauss’s method is based upon transferring the augmented matrix into an “Echelon” form where backward
substitution takes more operations (addition and multiplications). In Gauss-Jordan’s method, one reduces the
augmented matrix into “Reduced Echelon” form where one does not need the process of backward substitution.
This saves a lot of manual or computing time. Let us see how this method works.
Consider the system (1) of section 2.1. Following four steps can be applied to solve a system of m linear
equations in n variables using Gauss-Jordan method.
Step1. Change the system of linear equations to the form Ax  b 
Step2. Form the augmented coefficient matrix Ab by including the constants, an extra column in the matrix.
Step3. Convert the augmented matrix into “Reduced Echelon” form by using elementary row operations.
Step4. Find x by detaching the fourth column back to its original position on the
right hand side of the matrix equation Ax  b 
EXAMPLE 03: Use Gauss-Jordan method to solve the system of linear equations (m = n)
x 1  5x 2  2x 3  9
x 1  x 2  7x 3  6
 3x 2  4x 3  2
Solution:
Step1. We change the system of linear equations in matrix form Ax  b 

1 5 2   x1   9 

A 1 1 7 , x   x  and b   6  

0  3 4   2  2
   x 3   

Step2. We form the augmented coefficient matrix Ab by including the constants, an extra column in the
matrix.
1 5 2 9 

Ab  1 1 7 6  .
0  3 4  2
Step3. We convert this augmented coefficient matrix into reduced echelon form using elementary row
operations.
1 5 2 9  1 5 2 9
Ab  1 1 7 6  R 2    1 R1  0  4 5  3  R3    1 R2
0  3 4  2  0  3 4  2 
 
1 5 2 9  1 5 2 9
 0 4 5  3 R 2  4 R 3  0 0 1 1  R 23
0 1
 0 1

 1 1   1 1
1 5 2 9 1 0 7 4
 0 1 1 1  R1    5  R 2  0 1  1 1  R1    7  R3 , R 2  R3
   
0 0 1 1 0 0 1 1
1 0 0  3
 0 1 0 2 
0 
 0 1 1 

Step4. Find x by detaching the fourth column back to its original position on the right hand side of the matrix
equation Ax  b 
1 0 0  x1   3
0 1 0  x 2    2  

0 0 1  x3   1 

Page 6 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
Expanding from top, we have
x1  0  x2  0  x3  3 i
0  x1  x2  0  x3  2  ii 
0  x1  0  x2  x3  1  iii 
From (i), (ii) and (iii) we have the required solution of the given system of linear equations
x1 = -3, x2 = 2, x3 = 1.
EXAMPLE 04: Use Gauss-Jordan method to solve the system of linear equations
x1  x 2  x 3  a
x 1   1  a  x 2  x 3  2a
x 1  x 2   1  a  x 3  3a
Solution:
Step1. We change the system of linear equations in matrix form Ax  b 
1 1 1   x1  a 

A  1 1 a 1 , x   x 2  , and
 b   2a  
1 1 1  a     3a 
   x 3   

Step2. We form the augmented coefficient matrix Ab by including the constants, an extra column in the
matrix.
1 1 1 a 

Ab  1 1  a 1 2a  
1 1  a 3a 

 1
Step3. We convert this augmented coefficient matrix into reduced echelon form using elementary row
operations.
1 1 1 a
Ab  1 1  a 1 2a  R 2    1 R1 , R3    1 R1
1 1 1  a 3a 
 
1 1 1 a 1 1 1 a
1 1
 0 a 0 a   R1,   R3  0 1 0 1  R1    1 R2
 a a
0 0 a 2a  0 0 1 2 
1 0 1 a  1 1 0 0 a  3
 0 1 0 1  R1    1 R3  
0 1 0 1 
0 0 1 2  
0 0 1 2 
Step4. Find x by detaching the fourth column back to its original position on the right hand side of the matrix
equation Ax  b 
1 0 0  x1  a  3
0 1 0   x    1  
0 0 1   2   2 
   x 3   
Expanding from top, we have
x1  0  x2  0  x3  a  3 i
0  x1  x2  0  x3  1  ii 
0  x1  0  x2  x3  2  iii 
From (i), (ii) and (iii) we get the required solution of the given system of linear equations:
x1 = a - 3, x2 = 1, x3 = 2.
2.2.4 Applications of Gauss-Jordan Method
In this section we shall produce different problems from different engineering disciplines, viz; traffic
engineering, electric networking and many diverse areas of applied mathematics.

Page 7 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
EXAMPLE 05: An animal feed is to be made from corn, soybeans, and cotton seed. Determine how many
units of each ingredient are needed to make a feed that supplies 1800 units of fiber, 2800 units of fat, and
2200 units of protein, given that one unit of each ingredient provides the numbers of units shown in the
table below. The table states, for example, that a unit of corn provides 10 units of fiber, 30 units of fat,
and 20 units of protein.

Corn Soybean Cotton seed

Units of Fiber 10 20 30
Units of Fat 30 20 40
Units of Protein 20 40 25

Solution: Let x represent the required number units of corn, y the number of units of soybeans, and z the umber
of units of cottonseed. Since the total amount of fiber is to be 1800,
10 x  20 y  30 z  1800 
The feed must supply 2800 units of fat, so
30 x  20 y  40 z  2800 
Finally, since 2200 units of protein are required,
20 x  40 y  25 z  2200 
Then we must solve this system of equations:
10 x  20 y  30 z  1800 

30 x  20 y  40 z  2800 
20 x  40 y  25 z  2 200


1
Step1. We change the system of linear equations in matrix form Ax  b 
10 20 30  x 1800 
A  30 20 40  , x   y  , and b   2800  
20 40 25 z 2200 
     
Step2. We form the augmented coefficient matrix Ab by including the constants, an extra column in the
matrix.
10 20 30 1800 
Ab  30 20 40 2800  
 20 40 25 2200 
 
Step3. We convert this augmented coefficient matrix into reduced echelon form using elementary row
operations.
10 20 30 1800 
Ab  30
1 1 1
20 40 2800  R ,   R ,   R
 20  10 1  10  2  5  3
40 25 2200 

1 2 3 180 
 3 2 4 280 R    3 R , R    4  R
  2 1 3 1
4 8 5 440
1 2 3 180  1 2 3 180
 0 4 5
 1  1
 260    R ,    R  0 1 1.25 65  R    2  R
0 7

 280  4 2  7 3 0  1 2
 0  0 1 40 
1 0 0.5 50 
 0 1 1.25 65  R    0.5  R , R    1.25  R
0  1 3 2 3
 0 1 40 
1 0 0 30 
 0 1 0 15  
 
0 0 1 40 

From the above matrix it is obvious that the feed should contain 30 units of corn, 15 units of soybeans, and 40
units of cottonseed.

Page 8 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY

EXAMPLE 06: The housing department of the Punjab Government plans to undertake four housing
projects and lists material requirements for each house in each of the project as follows:
Project1 Project2 Project3 Project4
Paint  in 100 gallons  1 2 1 1.5
Wood  in 10,000 cu.ft.  3 4 2.5 2.5
Bricks  in millions  1 2 1.5 1
Labour  in 1000 hours  10 10 9 8
If the supplier delivers 6800 gallons of paint, 1,420,000 cubic feet of wood, 67 million bricks and 4,48,000
hours of labor, find the number of houses built for each project.
Solution: Let x1 , x 2 , x 3 and x 4 be the numbers of houses built under Project-1, Project-2, Project-3 and
Project-4 respectively. Then according to the conditions of the problem:
1  x1  2  x 2  1  x3  1.5  x 4  68
3  x1  4  x 2  2.5  x3  2.5  x 4  142
1  x1  2  x 2  1.5  x3  1  x 4  64
10  x1  10  x 2  9  x3  8  x 4  448
The augmented matrix is:
1 2 1 1 .5 68 
 3 4 2.5 2.5 142 
Ab   
 1 2 1 .5 1 64


10 10 9 8 448

We convert this augmented coefficient matrix into reduced echelon form using elementary row operations:
 1 2 1 1.5 68 
 3 4 2.5 2.5 142 
Ab    R 2    3 R1 , R3    1 R1 , R 4    10  R1
 1 2 1.5 1 64 
10 10 9 8 448
1 2 1 1.5 68 
0 2  0.5 2  62   1
 R  R2 ,    R2 , 2 R3 ,   1 R4
0 0 0.5  0.5 4  1  2
 
0  10 1 7  232
1 0 0.5  0.5 6 
0 1 0.25 1 31 
 R    10 R2
0 0 1 1 8 4
 
0 10 1 7 232
1 0 0.5  0.5 6 
0 1 0.25 1 31 
 R    0.5 R3 , R2    0.25 R3 , R4  1.5 R3
0 0 1 1 8  1
 
0 0  1.5 3  78
1 0 0 0 10 
0
1 0 1.25 33  1 
   R
0 0 1 1  8  4.5  4
 
0 0 0  4.5  90
1 0 0 0 10  1 0 0 0 10 
0
0 1 0 1.25 33  1 0 0 8 
 R    1.25 R4, R3  R4 
0 0 1 1  8 2 0 0 1 0 12 
   
0 0 0 1 20  0 0 0 1 20

Solving, we obtain: x1  10, x2  8, x3  12, x4  20 

Page 9 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
This means that with the available material, 10 houses can be built for Project-1, 8 houses for Project-2, 12
houses for Project-3 and 20 houses for Project-4.
EXAMPLE 07: A soap manufacturer decides to spend 600,000 rupees on radio, magazine and TV
advertising. If he spends as much on TV advertising as on magazines and radio together, and the amount
spent on magazines and TV combined equals five times that spent on radio, what is the amount to be
spent on each type of advertising?
Solution: Let x, y, z be the amounts in rupees spent on radio, magazines and TV advertising respectively. Then,
according to the given conditions, we have
x  y  z  600,000 z  x y y  z  5x
OR
x  y  z  600,000
x yz 0
5x  y  z  0
The augmented matrix is:
1 1 1 600,000 
Ab  1 1  1 0  R    1 R , R    5  R
5  1  1  2 1 3 1
 0 
1 1 1 600,000  1 1 1 600,000 
 1  1
 0 0 2  600,000    R2 ,    R3  
0 0 1 300,000 
 R23
 2  6
0 6 6  300,000  
0 1 1 1500,000

1 1 1 600,000 1 0 0 100,000 
 0 1 1 500,000 R1    1 R2 , R2    1 R3  
0 1 0 200,000
0 0 1 300,000 
0 0 1 300,000 

Thus, x = 100, 000, y = 200, 000, z = 300, 000. That is Rs.100, 000 is spent on radio, Rs.200, 000 is spent on
magazines and Rs.300, 000 is spent on TV advertising.

EXAMPLE 08: Find the three numbers whose sum is 34 where the sum of the first and second is 7 and
the sum of the second and third is 22.
Solution: Let x, y and z be the required three numbers. Then according to the question, we have
x  y  z  34
x y 7
y  z  22
To solve the above system of equations, we use the Gauss-Jordan method and have
1 1 1 34  1 1 1 34 
Ab  1 1 0 7  R 2    1 R1  0 0 1  27    1 R2
0 1 1 22  
  0 1 1 22 

1 1 1 34  1 0 0 12 
 0 0 1 27  R 23  0 1 1 22 R 2    1 R3
0  0 
 1 1 22   0 1 27 
1 0 0 12 
 0 1 0  5 
 
0 0 1 27 
Thus, x  12, y  5, z  27 

EXAMPLE 09: Three brands of fertilizer are available that provide nitrogen, phosphoric acid, and
soluble potash to the soil. One bag of each brand provides the following units of each nutrient.
Brand Nitrogen Phosphoric Acid Potash
A 1 3 2
B 2 1 0
C 3 2 1
Page 10 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY

For ideal growth, the soil in a certain country needs 18 units of nitrogen, 23 units of phosphoric acid, and
13 units of potash per acre. How many bags of each brand of fertilizer should be used per acre for ideal
growth?
Solution: Let x, y and z be the number of bags of each brand of fertilizer s should be used per acre for ideal
growth respectively. According to the question,
x  2 y  3z  18
3x  y  2 z  23
2x  z  13
To solve the above system of equations, we use the Gauss-Jordan method and have
1 2 3 18 
Ab   3 1 2 23 R    3 R , R    2  R
2 0 1 13  2 1 3 1
 
1 2 3 18  1 2 3 18 
 0  5  7  31   1 R ,   1 R  0 5 7 31 R    1 R
  2 3   2 3
0  4  5  23 0 4 5 23
1 2 3 18  1 0  1 2 
 0 1 2 8  R    2  R , R    4  R  0 1 2
 1
8    R
0 4 5 23 1 2 3 2 0 0  3  9   9  3
   
1 0  1 2  1 0 0 5 
 0 1 2 8  R  R , R    2  R  0 1 0 2 
0 0 1 3  1 3 2 3 0 0 1 3 
   
 x  5, y  2, z  3 
This shows that 5 bags of brand A, 2 bags of brand B, and 3 bags of brand C should be used per acre for ideal
growth.
2.2.5 Solution by Matrix Inverse Method
In this section we present the matrix inverse method to solve the system of linear equations A x = b, where A is
square matrix and |A| ≠ 0.
As mentioned earlier, this method is based on the fact that if A x = b and if A is invertible matrix, , then
x = A-1 b. For example, consider the system of equations:

2x + y – z = 1, 2y + z = 2, 5x + 2y – 3z = 1
In matrix form it may be written as:

2 1  1  x  1 
0
 2 1   y   2  A x = b  x = A-1 b
5 2  3  z  1 
1
x 2 1  1 1   8 1  3  1  3
 y   0 1   2    5 2  2   1
    2    1 [See Example 01 part (v) section 1.9]
 z  5 2  3 1   10 1  4 1   4 

2 1  1  8 1  3
 1   5 2 
NOTE Inverse of matrix A = 0 2  is  1 

5 2  3
 
 10 1  4

Thus x = 3, y = -1 and z = 4 is the solution of the given system.
2.3 HOMOGENEOUS SYSTEM OF LINEAR EQUATIONS
In the previous section we studied the system of linear equations A x = b. Such a system of equations is known
as “Non-Homogeneous System of Linear Equations”. In this section we shall consider and study the
“Homogeneous System of Linear Equations” where b = O.
Consider a system of m homogeneous equations in n unknowns x1 , x 2 ,  , x n ;

Page 11 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
a11 x1  a12 x 2    a1n x n  0 

a 21 x1  a 22 x 2    a 2 n x n  0 

 
 

a m1 x1  a m 2 x 2    a mn x n  0
In matrix form the above system of equations may be expressed as:

Ax=O (1)
where,

 a11 a12  a1n   x1  0 

a a22 
 a2 n  x  0 
A   21 , x   2  and O    
        
     
am1 am 2  amn   xn  0 
Obviously x1  0, x 2  0,  , x n  0, form a solution of (1). This is known as trivial solution. All other
solutions (if exist) are called non-trivial solutions of (1).
EXAMPLE 01: Solve the following system of homogeneous equations (m = n)
x y z 0
4x  5y  2z  0
2x  3y 0
Solution: We change the above system of equations to the matrix form Ax  O .
1 1 1   x   0 
Or  4 5 2   y   0  
 2 3 0   z  0 
    
1 1 1 
A  4 5 2 R    4  R , R    2  R
b 2 1 3 1
2 3 0
1 1 1  1 1 1 
 0 1 - 2 R    1 R  0 1 - 2 
0 1 - 2  3 2 0 0 
   0 
This last matrix implies that
1 1 1   x  0 
 0 1  2   y   0  
    
0 0 0   z  0

x yz 0 i
This gives
y  2z  0  ii 
From equation (ii) we get y  2 z  Putting y  2 z in (i) we get x  2 z  z  0  x  3 z 

Let z = k, the solutions are x  3k , y  2k and z  k  Here k is known as a “free parameter” or “an
arbitrary constant”; hence it can hold any value. Thus for different values of k, there are infinite solutions of the
above Homogeneous System of Equations. For example,
If k = 1, then x  3, y  2 and z  1 
If k = 2, then x  6, y  4 and z  2  etc.
EXAMPLE 02: Solve the following linear homogeneous equations (m < n)
2w  3x  y  z  0
4w  6x  2y  2z  0
 6w  12x  3y  4z  0
Page 12 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
Solution: Above equations can be written in the matrix form Ax  O , that is
 w  0 
 2 3  1  1
   
 4  6  2 2   x   0  
  6 12  y
 4    
0
 3
 z  0 
To solve the above system of equations, we use the Gauss-Jordan method and have
2 3  1  1
A  4 6 2 2  R    2  R , R  3R
b   2 1 3 1
  6 12 3  4

2 3 - 1 - 1 2 3 - 1 - 1
 0 - 12 0
1 1
4    R ,   R  0 -3 0 1 R  R
0  4 2  7  3 0
0 - 7  
 3 2
 21  3 0 - 1
 w  0 
2 3 1  1 2 3 1  1    
0 x 0
 0 3 0 1    3 0 1       
   y  0 
0 0 0 0  0 0 0 0     
 z  0 
2w  3x  y  z  0 i
This gives
 3x  z  0  ii 
From  ii  , we get x  z 3 . Putting this into  i  , we get
z
2 w  3   y  z  0  2 w  z  y  z  0  2 w  y  0  y  2 w 
3
Thus, x  z 3 and y = 2w. Take z = k1 and w = k2, we get x = k1/3 and y = 2k2 (It may be noted that there are
two parameters k1 and k2).Hence, x  k1 / 3, y  2k 2 , z  k1 and w  k 2 constitute the general solution
where k1 and k 2 are arbitrary constants. We can obtain an infinite number of solutions by giving different
values to k1 and k 2 .
2.4 CONSISTENCY CRITERIAN
Notice that the system of linear equations
x  2y  4
3x  2 y  2
has a unique solution . The system of equations
x  2y  4
3 x  6 y  12
has an infinite number of solutions, and the system of linear equations
x  2y  4
3x  6 y  5

has no solution. [Students are advised to solve the above three systems by using School Algebra or otherwise
and convince themselves].
We shall discuss in detail that under what conditions the system of linear equations (either non-homogeneous or
homogeneous) will possess the solution.
1. Consistency Criteria for the Non-Homogeneous System of Linear Equations
(a) When the system of linear equations A x = b is a square system and
i. The Rank of A = Rank of Ab = n (where n is number of unknowns or equations) then the system
has unique solution.
ii. The Rank of A = Rank Ab < n then the system has infinite solutions.
iii. The Rank A ≠ Rank Ab the system is said to be inconsistent and hence has no solution.
(b) When the system of linear equations A x = b is rectangular system and
i. The Rank of A = Rank of Ab, then the system has an infinite number of solutions.
ii. The Rank A ≠ Rank Ab, then the system is said to be inconsistent and hence has no solution.
Page 13 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
2. Consistency Criteria for Homogenous System of Linear Equations
(a) When the homogeneous system A x = O is a square system and
i. The Rank of matrix A= n, the system has only a trivial/zero solution.
i. The Rank A < n then the system has infinite many solutions.
(b) When the homogeneous system A x = O is a rectangular system and
i. The Rank of A = n, where n is the number of unknowns or variables, then the system has an infinite
solutions.
The following flowchart shows the systematic “Consistency Criterion” for the system of linear equations.

System of Equation

Non-homogeneous A x = b Homogeneous Ax=0

m=n m≠n m=n m≠n

Infinite solutions No solution if Zero solution if Infinitely many

if Rank(A) = Rank(A) ≠ Rank(A) = n solutions if
Rank(Ab)) Rank(Ab) Rank(A) < n

Infinitely many
solutions
ifRank(A) = n

Unique solution Infinite solutions No solution if

If Rank(A) = if Rank (A) < Rank(A) ≠
Rank((Ab) Rank(Ab) Rank(Ab)

EXAMPLE 01: Show that the system of equations 2x  6y  11, 6x  20y  6z  3, 6y  18z  1 is not
consistent (m = n).
Solution: The augmented matrix for the above system of equations is
2 6 0  11

Ab  6 20 6 3 

0 6  18 1 
Now we reduce the matrix into echelon form using elementary row operations:
2 6 0  11 2 6 0  11
 1
Ab  6 20  6  3  R    3 R  0 2  6 30     R 2
0 6  18  1  2 1  0 6  18 
 1 
2
  
2 6 0  11 2 6 0  11 
 0 1 3 15  R3    6  R 2  0 1 3 15  
0  18 1 
 0  91

 6  0 0

From the last matrix we can easily observe that Rank A  2 and Rank Ab  3 
Page 14 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
Rank A  Rank Ab , therefore the given equations are “inconsistent”. (Since 0x + 0 y + 0 z = -91, that
is, 0 = - 91 is not possible.)

EXAMPLE 02: Discuss the consistency of the following system of equations

2x  3y  4z  11, x  5y  7z  15, 3x  11y  13z  25.
If found consistent, solve it (m = n).

Solution: The augmented for the above system of equations is:

2 3 4 11 
Ab  1 5 7 15  
3 11 13 25
 
Now we reduce the matrix into echelon form using elementary row operations:
2 3 4 11 
Ab  1 5 7 15  R12
3 11 13 25
 
1 5 7 15 
 2 3 4 11  R 2    2  R1 , R3    3 R1
3 11 13 25
 
1 5 7 15 
 1
 0  7  10  19    1 R 2 ,    R3
0  4  8  20  4
 
1 5 7 15 1 5 7 15
 0 7 10 19 R 23  0 1 2 5  R    7 R2
0 1 2 5  0 7 10 19 3
   
 1 5 7 15   1 5 7 15 
 1
 0 1 2 5    R3  0 1 2 5
0 0  4  16 4  0 0 1 4 
   
From the last matrix we can easily observe that Rank A  Rank Ab  3  Therefore the given system is
consistent and has a unique solution. Also, we have

x  5 y  7 z  15 i
y  2z  5  ii 
z4  iii 
Solving, we have
x = 2, y = -3, z = 4
as a solution of the given system.
EXAMPLE 03: Test for consistency and solve the following system of linear equations
5x  3y  7z  4, 3x  26y  2z  9, 7x  2y  10z  5 (m = n)
Solution: The augmented for the above system of equations is:
5 3 7 4

Ab  3 26 9
2 

 7 2 10 5 

Now we reduce the matrix into echelon form using elementary row operations:
5 3 7 4 5 3 7 4 
Ab   3 26 2 9  R3    2  R 2  3 26 2 9  R13
7 2 10 5 1  50 6  13
   
1  50 6  13
 3 26 2 9  R 2    3 R1 , R 3    5  R1
 
5 3 7 4 

Page 15 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
1  50 6  13
 1   1 
 0 176  16 48   R ,  R
0 253  23 69  16  2  23  3
 
1  50 6  13 1  50 6  13
 0 11 1 3  R3    1 R2  0 11 1 3 
   
0 11  1 3  0 0 0 0 
From the last matrix we can easily observe that
Rank A  Rank Ab  2  3  No. of unknowns.
Thus the system is “consistent having infinite solutions”. From the last matrix we have
x  50 y  6 z  13  i
11 y  z  3  ii 
From  ii  , we have y   3  z  / 11 . Substituting it into  i  , we have
50 50 150
x  3  z   6 z  13 è x
z  6 z  13
11 11 11
  50  66  150 16 7 1
x
11
z 
11
 13 è x  z è x   7  16 z  
  11 11 11
Let z = k then, x = (7 – 16 k)/11, y = (3 + k)/11.

EXAMPLE 04: Examine the following system for a non-trivial solution (m < n)
x 1  x 2  2x 3  x 4  0
3x 1  2x 2  x4  0
4x 1  x 2  2x 3  2x 4  0
1 -1 2 1
 1
Solution: The matrix of the coefficients is: A  3 2 0 
4
 1 2 2

Now we reduce the matrix into echelon form using elementary row operations:
1 -1 2 1
A  3 2 0 1  R2    3 R1, R3    4  R1
4 1 2 2

1 1 2 1  1 1 2 1 
1
 0 5 6  2 R3    1 R2  0 5 6  2   R2
    5
0 5 6  2 0 0 0 0  
1 1 2 1  1 0 4 5 3 5 
 0 1 6 5  2 5 R1  R2  0 1 6 5  2 5 
0  0 
 0 0 0   0 0 0 

The rank of A is 2 < 4 (the number of variables), therefore, the system of equations is consistent having infinite
number of “non-trivial solutions”.
The first two rows of the last matrix give the following relations:

4 3
x1  x3  x4  0 i
5 5
6 2 From  i  and  ii  , we have
x 2  x3  x 4  0  ii 
5 5
 4 3  6 2
x1   x3  x 4  and x 2  x3  x 4 , where x3 and x4 arte arbitrary.
5 5  5 5
Let x3  k1 and x 4  k 2 , then

Page 16 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY

4 3  6 2 Hence, we have
x1   k1  k 2  and x2  k1  k 2 
5 5  5 5
4 3  6 2
x1   k1  k 2 , x 2  k1  k 2 , x3  k1 and x 4  k 2 
 5 5  5 5
We can get an infinite number of solutions by giving different values to k1 and k 2 
EXAMPLE 05: Show that the system
2x1  x2  3x3  a
3x1  x 2  5x 3  b
 5x1  5x2  21x3  c
is inconsistent if c  2a  3b 
Solution: The augmented matrix is:
 2 1 3 a
Ab   3 1 5 b 
 5  5 
 21 c
Now we apply the elementary row operations:
 2  1 3 a  2 1 3 a 
Ab   3 1  5 b  R    1 R  1 2 8 b  a  R12
  5  5 21 c  2 1  5
 5 21 c 

 
1 2 8 b  a
 2 1 3 a  R2    2  R1 , R3  5 R1
 5 
 5 21 c 
1 2 8 ba  1 2 8 ba 
 0 5 19 3a  2b  R3  R2  0 5 19 3a  2b  
0  0 
 5  19 c  5b  5a   0 0 c  3b  2 a 

The given system of equations will be inconsistent, if c  3b  2 a  0  c  2a  3b 

EXAMPLE 06: Find the values of k such that the system of equations
x  ky  3z  0, 4x  3y  kz  0, 2x  y  2z  0
has a non – trivial solution (m = n).
Solution: The matrix of coefficients is
1 k 3
A  4 3 k


2 1 2

Now we reduce the matrix into echelon form using elementary row operations:
1 k 3  1 k 3 
A   4 3 k  R 2    4  R1 , R 3    2  R1  0 3  4k k  12 R2    2  R3
2 1 2  0 1  2k  4 
 
1 k 3  1 k 3 
 0 1 k  4 R3    1  2k  R2  0 1 k  4  .
0 1  2k  4  0 0 2k 2  9k 

For non – trivial solution  

2k 2  9k  0  k 2k  9  0  k  0 or k 
9

2
EXAMPLE 07: Find the value of k such that the following system of equations has non – trivial solutions.
Also find the solution.

2x  3y  2z  0, 3x  y  3z  0, 7x  ky  z  0
Solution: The matrix of coefficients for the above system is:

Page 17 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
2 3 - 2
A  3 - 1 3 

7 k - 1
Now we reduce the matrix into echelon form using elementary row operations:
2 3 - 2
A  3 - 1 3  R 2    1 R1
7 k - 1 
 
2 3  2
 1  4 5  R12
 
 7 k  1
1  4 5 
 2 3  2 R    2  R1 , R 3    7  R1
7 k  1  2
 
 1  4 5 
1
 0 11  12   R
   11  2
 0 k  28  36 
1  4 5 
 12 
 0 1   R 3     k  28  R 2
11
0 k  28  36 
 
1 4 5 
 0 1  11 / 12 
0 
 36  12 k  28  / 11
 0

12
For non – trivial solution we must have:  36   k  28  0 è 396  12k  308  0
11
è 12k  60  0 è k  5
Putting this value of k in the above equations, we get:
x  4 y  5z  0 i
y  12 / 11z  0  ii 
From  ii  we have y  12 / 11 z
 12  48 7
Substituting it into  i  , we have x  4   z  5z  0  x z  5z  0  x   z
 11  11 11
Let z = a, then x = -7 a /11, y = 12 a/11 which is the required solution of the given system.

EXAMPLE 08: Determine for what values of λ and μ the following system of equations
x  y  z  6, x  2y  3z  10, x  2y  λz  μ
has (i) no solution (ii) a unique solution (iii) infinite number of solutions.

Solution: The augmented matrix for the above system of equations is:
1 1 1 6 
Ab  1 2 3 10  .
1 2   

Now we reduce the matrix into echelon form using elementary row operations:
1 1 1 6 
Ab  1 2 3 10  R2    1 R1 , R3    1 R1
1 2   
 
1 1 1 6  1 1 1 6 
 0 1 2 4  R    2 R  0 1 2 4 
  3 2  
 0 1  1   6 0 0  3   10 
Page 18 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
(i) There is no solution if Rank A  Rank Ab
i.e.   3  0    3 and   10  0    10 
(ii) There is a unique solution if Rank A  Rank Ab  3
i.e.   3  0    3 and  may have any value.
(iii) There are infinite solutions if Rank A  Rank Ab  3
i.e.   3  0    3 and   10  0    10 
EXAMPLE 09: For what value of λ the following system of linear equations
 5  λ x 1  4x 2  2x 3  0
4x 1   5  λ  x 2  2x 3  0
2x 1  2x 2   2  λ   0
have a non-trivial solutions. Find these solutions.
Solution: The matrix of coefficients is
5   4 2 
A 4 5 2  
 2 2 2   
We reduce this matrix to the echelon form by applying elementary row operations:
5   4 2   2 2 2  
1
A 4 5 2  R13  4 5 2    R1
 2   
2  
2
 2 2   5   4

 1 1 1   2
 4 5 2  R 2    4  R1 , R3     5    R1
5   
 4 2 
 
1 1 1  2 
 0 1    2  2 
  6  7  
2
0   1 
 2 
 
1 1 1  2 
 0 1   2 1    
1   1
R2 , 

 R3 ;  assuming 1    0 
   1    6      1    1  
0  1    
 2 
   
1 1 1   2  1 0 3 2 
 0 1  2  R1    1 R 2 , R 3  R 2  A  0 1 2 
  6      10 
0  1 2 
  0 0 
 2 
For the non-trivial solution, rank of A < 3 (the number of unknowns), so we must have
  10
 0    10  It implies that
2
 10 
x1   3   x3  0 or x1  2 x3  0 or x1  2 x3 and x 2  2 x3  0 or x 2  2 x3 ,
 2
where x 3 is arbitrary.
If 1    0 or   1 , then above system becomes:
1 1 1 2
0 0 1  1 
 0   x1  x 2  2 x3  0  x1   x 2  2 x3 , where x 2 and x 3 are both arbitrary.
 0   
0 0 
EXAMPLE 10: For what value of λ the system of linear equations

Page 19 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
 3  λ x1  x 2  x 3  0
x1   1  λ x 2  x 3  0
x1  x 2   1  λ x 3  0
has a non-trivial solutions. Find these solutions.
Solution: The matrix of the coefficients is
 3    1 1 
A   1  1    1 

 1 1 1    
We reduce this matrix to the echelon form by applying elementary row operations:
 3    1 1 
A  1  1    1  R12
 1    
 1 1
 1  1    1 
  3    1 1  R 2     3    R1 , R3    1 R1
 1    

 1 1

1  1    1 

 0 2  4  
2
  2    1 R3
 
0   

1  1 1  1  1 1 
1
 0 2  4     2   R3 ;   0  0 2  4     2  R23
2 2
0 
 
  
0 
  1 1 

 
1  1 1 
1  R       1 R 2 , R3    2  4  2  R 2
 
 0 1
0 2  4    2   2  1  
 
1 0 2 
 0 1 1 .
0 0  4  5  2 
 
For non-trivial solutions rank of A < 3 (the number of unknowns), so we must have
 4  5   2  0 or  2  5  4  0    4  0 or   1  0    1, 4 
Now when   1, then from the last matrix, we have
x1  x3  0  x1   x 3 and x 2  x3  0  x 2   x 3 
When   4,
x1  2 x3  0  x1  2 x 3 and x 2  x3  0  x 2   x 3 
Hence, for   1, x1  x2   x3 : for   4, x1  2 x3 , x2   x3 
EXAMPLE 11: A zoo charges $6 for adults, $3 for students, and $0.50 for children. One morning 79
people enter and pay a total of $207. Determine the possible numbers of adults, students, and children.
Solution: Let A denote the number of adults, S be the number of students and C the number of children. Then
according to the question,
A  S  C  79 (Number of people)
6 A  3S  0.5C  207 (Total Amount)
To solve the above system of equations, we use the Gauss-Jordan method and have
1 1 1 79 
Ab 
6 3 0.5 207   2  R 2
1 1 1 79 
 R    12  R1

12 6 1 414
 2

Page 20 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
1 1 1 79  1 

   R
 0  6  11  534 6  2
1 1 1 79 
 R    1 R 2

0 1 11 6 89   1
1 0  5 6  10
 

0 1 11 6 89 

From the last matrix, we have
5 5 11 11
A C  10  A  C  10 and S  C  89  S  C  89 
6 6 6 6
Possible values for C are C  12,18, 24,30,36,42, and 48  Because for other values of C, the result will
be either a fraction or a negative number.
Hence the possible number of adults, students and children are shown in the following table:

Children 12 18 24 30 36 42 48
Adults 0 5 10 15 20 25 30
Students 67 56 45 34 23 12 1

EXAMPLE 12: Three species of bacteria co-exist in a test tube and feed on three foods F1 , F 2 , F 3 .
Suppose that a bacterium of the i th species consumes on the average an amount a ij of the j th food
per day where
a 11  1, a 12  1, a 13  1,
a 21  1, a 22  2, a 23  3,
a 31  1, a 32  3, a 33  5 
Further suppose that there are 15000, 30000, 45000 units of food of types F1 , F2 , F3 respectively.
Assuming that all food is consumed, what are the populations of the three species that can co-exist in this
environment?
Solution: Let the population of the three species of bacteria be x1 , x 2 , x 3  Since the i th species
1 1 1  x1  15000 
 3  x 2   30000   That is,
consumes a ij units of food of type we have 1 2
1 3 5  x3  45000
x1  x2  x3  15000 i
x1  2 x2  3x3  30000  ii 
x1  3x2  5 x3  45000  iii 
1 1 1
 3
The matrix of the co-efficients is: A  1 2 

1 3 5

The augmented matrix is:
1 1 1 15000 
Ab  1 2 3 30000 R 2    1 R1 , R3    1 R1
1 3 5 45000
1 1 1 15000 
 0 1 2 15000  R3    2  R2
0 2 4 30000
1 1 1 15000
 0 1 2 15000 R1    1 R2
0 0 0 0 
1 0 1 0 
 0 1 2 15000 .
0 0 0 0 
 x1  x3  0  x1  x3 and x 2  2 x3  15000  x 2  15000  2 x3 , where x3 is arbitrary.
Page 21 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
From  i , we see that the total populations of the three species that can co-exist is 15000.
If we take x3  3000, then x1  3000 and x 2  9000 
Hence, the populations of the three species that can co-exist in this environment are 3000, 9000, and 3000.
2.5 NETWORK FLOW PROBLEMS
Many problems in every day life involve flow of traffic along a network of roads, or of water through pipes, or
of current along conductors. In such networks there are points (called junctions or nodes) where a new flow
either enters or leaves the system. The basic idea to be exploited in such systems is that flow is conserved i.e.
the flow into the system is equal to the total flow out of the system. This means that at each junction (or point of
intersection) the total flow into must equal the total flow out. This condition leads to a linear system of equations
for the unknown variables. For example, Fig. 1 shows a flow of 40 into a node and a total flow of x1  x2  5
out of the node. Since we assume that the total flow into a node is equal to the total flow out, it follows that the
flows x1 and x2 must satisfy the linear equation 40  x1  x2  5, or equivalently, x1  x2  35 

In this section we will discuss network problems associated x1

with traffic flow and electrical currents in conductors.
NOTE: The network problems we discuss in this section are 40 x2
kept very simple so the computational details do not obscure
the ideas.
x3
EXAMPLE 01: In a certain part of the city two sets of one way streets intersect
as shown in the following figure:

610 310
450 x1

A D 640
x2 x4
520
B x3 C 600

480 390

The average hourly volume of traffic entering and leaving this part during such hours is given in the
figure. Determine the amount of traffic between each of the four junctions.
Solution: We denote the number of automobiles between junctions by x1 , x2 , x3 , and x4  Using the fact
that the number of automobiles entering a given junction must equal that leaving the same, we obtain the
following equations for the unknowns x1 , x2 , x3 , x4 
x1  45 0  x2  610 

x2  52 0  x3  4 80 

x3  3 90  x4  6 00 
x4  64 0  x  3 10 
1 

These equations can be written in the standard form as

x1  x2  160
x2  x3  40
x3  x4  210
x1  x4  330
The augmented matrix for the above system of equations can be transformed into the reduced row echelon form
as follows:
1  1 0 0 160  1  1 0 0 160 
0 1  1 0  40  0 1  1 0  40
Ab=   R 4    1 R1 
 1 210
 R1  R2 , R4    1 R2
0 0 1  1 210  0 0 1

1 0 0  1 330   0 1 0  1 170  

Page 22 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
1 0 1 0 120  1 0 0  1 330 
0 1 1 0  40  0 1 0  1 170 

0 0 1 1 210
 R1  R3 , R 2  R3 , R 4    1 R3  
0 0 1  1 210

   
0 0 1 1 
210  0
 0 0 0 
0 

From the last matrix, the equivalent system of equations is:

x  x  330, x  x  170, x  x  210 
1 4 2 4 3 4

Here is a free variable. There are an infinite number of possible solutions. There is no unique solution. If the
amount of traffic between a pair of junctions were known, the problem could be exactly solved.
EXAMPLE 02: A part of Lahore’s road network for vehicular traffic is as shown by arrows in the
following diagram:
(a) Write the equations indicating the traffic flow given in the diagram.
(b) Show that the traffic flow along AB, CD can be expressed in terms of the traffic flow along AD.
(c) If the stretch AD or the stretch AB is closed, then show that the solution to the problem is unique.
Solution:
(a) Suppose that x1 , x 2 and x 3 denote the number of vehicles going along AB, CD and AD respectively:
The required diagram is shown below. The symbol indicates “One way traffic”
360 x1 250
Equating the incoming traffic to the outgoing
traffic at each intersection, we have the following
mathematical model:
x3
x1 + x3 = 360 (i)
x2 + x3 = 390 (ii)
x1 + 250 = x2 + 220  x1 – x2 = - 30 (iii) 390 x2 220
(b) Taking the augmented matrix and reducing it to the echelon form, we have
1 0 1 360  1 0 1 360  1 0 1 360
Ab  0 1 1 390  R3    1 R1  0 1 1 390  R  R2  0 1 1 390 .
1  1 0  30 0  1  1  390 3 0 
     0 0 0 

 x1  x3  360  x1  360  x3 and x2  x3  390  x2  390  x3 

Thus, x1 and x 2 can be expressed in terms of x3 , where x 3 is an arbitrary integer such that
0  x 3  360 
(c) If the stretch AD is closed, then x 3  0 and x1  360, x 2  390  Or if the stretch AB is closed,
then x1  0 and x 2  30, x3  360 
EXAMPLE 03: The flow of traffic (in vehicles per hour) through a network of streets is shown in the
adjacent figure.
Solve the system for x i , i  1,2,3,4  Also show the traffic flow when: 200 x
2 100
(a) x  0 A D
4
(b) x  100  x
4 1
x
4
Solution: 100 C
B x
3 100
Using the fact that flow into the junction is equal to flow out, we obtain
x2  200  x1, x1  100  x3 , x3  200  x4 , x4  100  x2
These equations can be written in the standard form as
x1  x 2  200
x1  x3  100
Page 23 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
x 3  x 4  200
x2  x 4  100
The augmented matrix for the above system of equations can be transformed into the reduced row echelon form
as follows:

1  1 0 0 200  1  1 0 0 200 
1 0  1 0  100 0 1  1 0  300
Ab=   R 2    1 R1 
1
 R1  R2 , R4    1 R2
0 0 1  1 200  0 0 1 200

0 1 0  1  100 0 1 0 
 1  100 

1 0 1 0  100  1 0 0 1 100 
0 1 1 0  300 0 1 0 1  100 

1
 R1  R 3 , R 2  R 3 , R 4    1 R3  
1

0 0 1 200
 0 0 1 200

0 0 1 1 
200  0
 0 0 0 0 

From the last matrix, the equivalent system of equations is:
x1  x4  100, x2  x4  100, x3  x4  200 

Here x4 is a free variable. With x4  k , we obtain

x1  100  k , x2  100  k , x3  200  k 
The solution is not unique. This means that there are a large number of possible flows of traffic.
(a) When x4  0, that is; no flow along the street CD, we obtain the unique solution
x1  100, x 2  100, x3  200 
(b) When x  100, we have the unique solution
4
x1  200, x2  0, x3  300 

NOTE: We normally want the flows in a network to be nonnegative. For instance, in the above example
part (b) x 2  100 indicates that traffic was flowing from A to D rather than in the prescribed direction
from D to A.
EXAMPLE 04: (a) Set up a system of equations that represents traffic flow for the network shown in the following
figure. (The numbers give the average flows into and out of the network at peak traffic hours.)
(b) Solve the system of equations. What is the traffic flow if x 6  300 and x7  1300 vehicles per hour?
400

x1 x2

800 A B C 600

x5 x4 x3

600 1600
F x6 E x7 D
400 400

Solution: (a) Since the flow into a node is equal to the flow out, we obtain the following system of equations:
800  x1  x5  Node A , x1  x4  x2  400  Node B 
x2  x3  600  Node C  , x3  1600  x7  400  Node D 
x7  x4  x6  Node E  , x5  x6  400  600  Node F  
These equations can be written in the standard form as
x1  x5  800  Node A , x1  x2  x4  400  Node B  ,

Page 24 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
x2  x3  600  Node C  , x3  x7  1200  Node D  ,
x4  x6  x7  0  Node E  , x5  x6  1000  Node F  

The augmented matrix for the above system of equations can be transformed into reduced row echelon form as
follows:
1 0 0 0 1 0 0 800 
1  1 0 1 0 0 0 400 

0 1  1 0 0 0 0 600 
Ab=   R    1 R1
0 0 1 0 0 0  1  1200 2
0 0 0 1 0 1 1 0 
 
0 0 0 0 1 1 0 1000 
1 0 0 0 1 0 0 800 
0  1 0 1  1 0 0  400 
0 1  1 0 0 0 0 600

     1 R
0  1  1200 2
0 0 1 0 0
0 0 0 1 0 1 1 0 
0 0 
 0 0 1 1 0 1000 
1 0 0 0 1 0 0 800 
0 1 0  1 1 0 0 400 
0 1  1 0 0 0 0 600

  R    1 R
0 0 1 0 0 0  1  1200 3 2
0 0 0 1 0 1 1 0 
0 0 0 1000 
 0 1 1 0 
1 0 0 0 1 0 0 800 
0 1 0  1 1 0 0 400 
0 0  1 1  1 0 0 200

     1 R
0  1  1200 3
0 0 1 0 0
0 0 0 1 0 1 1 0 
0 0 0 
 0 1 1 0 1000 
1 0 0 0 1 0 0 800 
0 1 0  1 1 0 0 400 
0 0 1  1 1 0 0  200 
   R    1 R
0 0 1 0 0 0  1  1200 4 3
0 0 0 1 0 1  1 0 
0 0 0 0 1 1 0 1000 
 
1 0 0 0 1 0 0 800 
0 1 0 1 1 0 0 400 

0 0 1 1 1 0 0  200 
  R  R4 , R3  R4 , R5    1 R4
0 0 0 1 1 0 1  1000 2
0 0 0 1 0 1 1 0 
 
0 0 0 0 1 1 0 1000 

Page 25 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
1 0 0 0 1 0 0 800 
0 1 0 0 0 0 1  600 
 
0 0 1 0 0 0 1  1200
  R    1 R5 , R4  R5 , R6    1 R5
0 0 0 1 1 0 1  1000 1
 
0 0 0 0 1 1 0 1000 
0 0 0 0 1 1 0 1000 

1 0 0 0 0 1 0  200 
0 1 0 0 0 0 1  600 
 
0 0 1 0 0 0 1  1200
  
0 0 0 1 0 1 1 0 
 
0 0 0 0 1 1 0 1000 

0 0 0 0 0 0 0 0 

From the last matrix, the equivalent system of equations is:
x1  x 6  200  x1  x 6  200  x 2  x 7  600  x 2  x 7  600 
x 3  x 7  1200  x 3  x 7  1200  x 4  x 6  x 7  0  x 4  x 7  x 6 
x 5  x 6  1000  x 5  1000  x 6 

(b) If x6  300 and x7  1300, then (vehicles per hour)

x1  100, x  700, x  100, x  1000, x  700 
2 3 4 5

2.6 ELECTRICAL NETWORKS

An electrical system may be in the form of a circuit or a network. The essential components of such a system
are a resistor, a capacitor (or a condenser), an inductor and a battery or generator as a source of voltage or
electromotive force (e.m.f.).
A resistor uses electrical energy and converts it into heat. An electric light bulb, electric tube and electric heater
are examples of resistors. A capacitor stores electrical energy and an inductor opposes a change in electrical
current. The electromotive force (supplied by a battery or generator) is the cause of flow of current.

2.6.1 Symbols, Graphical Representation and Units

C R
(a) (b)
Battery

(e)

L V

(c) (d)
Figure 2.6
In a circuit
(i) a capacitor is denoted by C, its capacitance is measured in farads (F) and its graphical representation is
shown in figure 2.6 (a).
(ii) A resistor is denoted by R, its resistance is measured in ohms    and its graphical representation is
displayed in figure 2.6 (b).
(iii) An inductor is denoted by L, its conductance is measured in henries (H) and its graphical representation is
displayed in figure 2.6 (c).
(iv) The electromotive force is measured in volts (V). The e.m.f. of a battery is fixed and may be denoted by E.
The e.m.f. of a generator varies periodically and may be denoted by E 0 sin t , where  is the angular
frequency. Graphical representations of a battery and generator are displayed in figure 2.6 (d, e).
2.6.2 Electrical Quantities
The common electrical quantities are charge (quantity of electricity) denoted by q; current i.e. the amount of
charge flowing per second through the cross section of a wire, denoted by I  dq dt ; where the time t is
measured in seconds. The charge is measured in coulombs and the current in amperes.
Page 26 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
The flow of current in a circuit is governed by Ohm’s law and Kirchoff’s laws which can be stated as follows:
Ohm’s Law: The voltage drop across a resistor is the product of the current and the resistance.
Kirchoff’s First Law (Junction Rule): The sum of the currents flowing into a node is equal to the sum of the
currents flowing out.
Kirchoff’s Second Law (Current Rule): The algebraic sum of the voltage drops around a closed circuit is
equal to the total voltage in the circuit.
NOTE: When applying the current rule, select a direction (clockwise or counterclockwise) around the
closed circuit and then consider all voltages and currents positive when in this direction and negative
when in the opposite direction. This is why the term algebraic sum is used in the current rule.
EXAMPLE 01: Find currents in each branch of the circuit shown in the following figure:

F E
6 I 1

20V
I 2

A B
4 10V

2 I3
C D
Solution: Using Kirchoff’s first law (junction rule) at nodes A and B, we have
I  I  I  I  I  I  0
2 3 1 1 2 3
i
and by Kirchoff’s second law (current rule) for the circuits ABCD and ABEF, we have
4 I  2 I  10
2 3

 ii 
4 I 2  6 I 1  20  iii 

Equations 1 ,  2, and  3 can be written in the standard form as

I 1
 I 2
 I 3
 0
 4I 2
 2I 3
 10
6I 1
 4I 2
 20
The corresponding augmented matrix can be reduced as follows:
1  1  1 0  1 1 1 0
   1 1
A  0  4
b
2 10  R    6  R  0 3 1
4 2 10    R ,   R
2 3
 2  2
6 4 0 20 0 10 6 20
1  1  1 0  1 1 1 0 

 0 2  1  5 R    2  R  1  5
3 2
0 2 R 23

0 5 3 10  
0 1 5 20 

1  1  1 0 

 0 1 5 20  R  R , R    2  R 1 2 3 2

0 2  1  5
1 0 4 20 
 1 
 0 1 5 20   R
 11 
3

0 0  11  45
1 0 4 20  1 0 0 40 11 
 0 1 5 20  R    4  R , R    5 R
1 3 2 3

0 1 0  5 11

0 0 1 45 11 
0 0 1 45 11 


Page 27 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
Hence, from the last matrix, we have
40 5 45
I 
, I  , I  
11 11 11
1 2 3

EXAMPLE 02: Determine the currents I , I , and I 1 2 3 for the electrical network shown in the
following figure:
20 5V

I1 A
I2 10
B C
I3
10 10V

D
Solution: Applying Kirchoff’s second law to the loops BDCB and BCAB, we obtain the equations:
 10 I 2  10 I 3  10  BDCB 
20 I1  10 I 2  5  BDCB 
Applying Kirchoff’s first law to either of the nodes B or C, we have I 1  I 2  I 3  Thus the above equations
can be written in the standard form as:
I1  I2  I3  0
 10 I 2  10 I 3  10
20 I1  10 I 2  5
The corresponding augmented matrix can be reduced as follows:
1 1 1 0 
 1  1
Ab   0  10 10 10   R2 ,   R3
 10  5
20 10 0 5 
1 1 1 0 
 0 1  1  1 R3    4  R1
4 2 0 1 

1 1 1 0
 0 1 1  1 R1  R2 , R3    6  R2
0 6 4 1 
1 0 2  1
 1 
 0 1 1  1  R3
0  10 
0 10 7 
1 0 2  1 1 0 0 0 .4 
 0 1 1  1 R1  2 R3 , R2  R3 
0 1 0  0.3
0 0 1 0.7  
0 0 1 0 .7 

Therefore, the currents are
I1  0.4, I 2  0.3, I 3  0. 7 
Since I2 is negative, the current flow is from C to B rather than from B to C, as tentatively assigned in the given
figure.

Page 28 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
EXAMPLE 03: Find currents in the circuit shown in the following figure:

I1 5

5V
10
I2 B
A

5
10V
I3

D
Solution: Kirchoff’s first law applied at junction A gives
I 2  I1  I 3   I1  I 2  I 3  0
Similarly, Kirchoff’s second law applied to circuits ABC and ABD gives
10 I 2  5 I1  5 and 10 I 2  5 I 3  10 
Thus the above equations can be written in the standard form as:
I1  I2  I3  0
5 I1  10 I 2  5
10 I 2  5I 3  10

The corresponding augmented matrix can be reduced as follows:

1 1 1 0 1 1 1 0
1 1
Ab  5 10 0 5   R2 ,   R3  1 2 0 1  R2    1 R1
0 5 5
10 5 10 0 2 1 2
1 1 1 0
 0 3 1 1  R2    1 R3
0 2 1 2
1 1 1 0
 0 1 2  1 R1  R2 , R3    2  R2
0 2 1 2 
1 0 1  1
1
 0 1 2  1  R3
5
0 0 5 4 
1 0 1  1
 0 1 2  1  R1  R3 , R2  2 R3
0 0 1 4 5

1 0 0  1 5

0 1 0 3 5 

0 0 1 4 5 

Page 29 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
1 3 4
Therefore, the required currents are: I2   , I 2  , I3  
5 5 5

EXERCISE 2
01. Solve the following systems of linear equations:
x  y  z  3 x  2y  3z  1
 i x  2y  3z  4  ii  3x  y  z  2
x  4y  9z  6 4x  2y  z  3
x  y  2z  3
 iii  x  2y  3z  5
3x  4y  5z   13
2x  x  2x  x  6
x  2x  x  1 1 2 3 4
1 2 3 3x  6x  6x  12 x  36
 iv  3x  2x  2x  2  iv  1 2 3 4
1 2 3 4x  3x  3x  3x  1
7x  2x  3x  5 1 2 3 4
1 2 3 2x  2x  x  x  10
1 2 3 4
3x  2x  2x  0
1 3 4
 iv  x  7x  4x  9x  0
1 2 3 4
7x  7x  5x  0
1 2 4
02. Test the consistency of the following equations and solve them if possible:
3x  3y  2z  1 t  w  x  y  z  1
x  2y  4 2t  w  3x  4z  2
 i  ii 
10 y  3z  2 3t  2w  2x  y  z  1
2x  3y  z  5 t  x  y  z  0
3w  6x  y  z  0 4x  5y  z  2
 iii  w  2x  5y  3z  0  iv  3x  y  2z  9
2w  4x  3y  z  3 x  4y  z  5

03. Find the value of k for which the following system of equations is consistent:
3x  2y  2z  3
x  ky  3z  0
4x  y  2z  7

04. Find the value of  for which the system of equations

x  y  4z  1
x  2y  2z  1
x  y  z  1
will have a unique solution.
05. For what values of  and  , the following system of equations
2x  3y  5z  9
7x  3y  2z  8 will have
2x  3y  z  
(i) unique solution (ii) no solution.

06. Determine the values of a and b for which the system

3  2 1   x  b
5  8 9   y   3
2 1 a   z   1
    
(i) unique solution (ii) has no solution and, (iii) has infinitely many solutions.
Page 30 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY

07. Choose  that makes the following system of linear equations consistent and find the general solution of
the system for that  .
x  y  z  t  2
2y  4z  2t  3
x  2y  z  2t  
08. Show that the system of equations
3x  4y  5z  a
4x  5y  6z  b
5x  6y  7z  c
doesn’t have a solution unless a  c  2b  Solve the equations when a  b  c  1 
09. Find the condition on  for which the system of equations
3x  y  4z  3
x  2y  3z  2
6x  5y  z  3
has a unique solution. Find the solution for   5 
10. Find the value of k such that the system of equations
2x  3y  2z  0
3x  y  3z  0
7x  ky  z  0
has non – trivial solutions. Find the solutions.
11. Find the values of k, such that the system of equations
4x 1
 9x 2
 x
3
 0
kx 1
 3x 2
 kx 3
 0
x 1
 4x 2
 2x 3
 0
has non – trivial solutions. Hence, find the solution of the system.
12. Find values of  for which the following system of equations has a non – trivial solution.
3 x1  x2  x 3
 0
2 x1  4 x2  x 3
 0
8 x1  4 x2  6x 3
 0
13. Find the value of k for which the following system of equations

4x  2y  5z  0
x  ky  2z  0
2x  y  z  0
has a non – trivial solution.
14. Determine the non – trivial solution (if any) of the following homogeneous system of linear equations:

x1  2x2  3x3  0 x1  x2  x3  0
i 2x1  5x2  2x3  0  ii  2x1  3x2  x3  0
3x1  x2  4x3  0 x1  4x2  2x3  0

Page 31 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY

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

 iii  3x1  7x2  2x3  4x4  0  iv  2x1  4x2  x3  0
4x1  3x2  5x3  2x4  0 3x1  2x2  2x3  0
15. For the following network, (a) set up the system of equations that describes traffic flow; (b) determine the
flows x1 , x 2 and x3 if x 4  100; and (c) determine the maximum and minimum values for x1 if all the
flows are constrained to be nonnegative.
800 400

400 A x1 B 600

x4 x2
400 C D

200 x3 200 200

16. The flow of traffic (number of vehicles per hour) through a network of streets is shown in the following
figure:

400 x1 600

x2 x3 x4

300 x5 100

17. For the electrical networks (i) – (iii), determine the currents in the various branches:

4 2V
I1
3 I2
(i)
I3
4 4V

I1 3V

R1  4
I2
(ii)
R 32

I3

R3  1 4V

Page 32 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY

I1 5V

R1  1
I2
(iii)
R2  2
I3

R3  1 8V

ANSWERS TO EXERCISE 2
20 3 1
01.  i  x  2, y  1, z  0   ii  x  ,y ,z    iii x  1, y  0, z  2 
29 29 29
3 k 1  5k
 iv  x 
1
,x 
2
, x  k   v  x  2, x  1, x  1, x  3 
3 1 2 3 4
4 8
2  1  
 vi  x    k  k , x    29k  14k , x  k , x  k 
1 3 1 2 2 21  1 2 3 1 4 2

02.  i  x  2, y  1, z  4   ii  t  3k  k  1, w  3k  1, x  k  2 k  1, y  k , z  k 

1 2 1 1 2 1 2
 iii  x  k , y  1, z  2, w  2 k  1   iv  x  2, y  1, z  1 
1 7
03. k   04.   
4 10
05.  i  5  ii    5,   9 
1 1
06.  i  a  3,  ii  a  3, b  ,  iii  a  3, b  
3 3
7 1 3
07.   , x   3k , y   2k  k , z  k , t  k 
2 2 2 2 2 1 2 1
08. x  k  1, y  2k  1, z  k  09
5k 4 13k 9
  5, x    , y  , z k
7 7 7 7
7a 12a
10. k  5, x   , y , z  a 11. k  1, x1  2 , x 2   , x3   
11 11
1
12.   1  13. k  
2
2 3
14  i  x1  0, x 2  0, x3  0  ii  x1  k , x2  k , x3  k , where k is an arbitrary value assigned to
5 5
the
free variable x 3   iii  The system must have a non – zero solution because there are more unknowns
than equations.  iv  x1  0, x2  0, x3  0 
x1  x4  1200
x1  x2  1000
15.  a   b x1  1100, x2  100, x3  500;  c 
x3  x4  600
x2  x3  400
The minimum
value is x1  600 and the maximum value is x1  1000 
16 (a) x1   k1  k 2  700, x2   k1  k 2  300, x4   k 2  100  (b)
x1  600, x2  200, x4  0 

Page 33 of 34
 LINEAR ALGEBRA AND ANALYTICAL GEOMETRY
(c) x1  500, x2  100, x4  0 
17.  i  I1  0.05, I 2  0.6, I 3  0.55   ii  I1  0, I 2  1, I 3  1 
 iii  I1  1, I 2  2, I 3  1 

Page 34 of 34