Chapter II

Linear
equations and
inequalities

1-1
Many problems in economics can be modelled as a system of linear equations or a system of

linear inequalities. In this chapter, we consider some basic properties of such systems and

discuss general solution procedures.

1-2
Section I
SYSTEMS OF LINEAR
EQUATIONS

1-3
Preliminaries

At several points in previous chapters we have

been confronted with systems of linear
equations. The following example of determining
feasible production programmes also leads to a
system of linear equations.

1-4
Example
•

1-5
Thus, we get the following system of three linear equations with four variables:

1-6
1-7
1-8
Definition

1-9
Matrix representation

The system (8.1) can also be written in matrix representation

Ax = b,

1-10
Vector representation

1-11
Basic notions

1-12
 Existence and uniqueness of a
solution

We investigate in which cases a system has a

solution. To this end, we introduce the notion of
the rank of a matrix.

1-13
r(A) = p ≤ min{m, n}

1-14
Example 1

1-15
Example 2

1-16
1-17
Augmented matrix

1-18
Determining consistency

1-19
Determining consistency

1-20
Unique solution/ Infinite solutions

1-21
 Elementary transformations;
solution procedures
Solution procedures for systems of linear
equations transform the given system into a
‘system with easier structure’. The following
theorem characterizes some
transformations of a given system of linear
equations such that the set of solutions
does not change.

1-22
Elementary / Equivalent
transformations

1-23
Solution procedures

Gauss Jordan elimination

Gaussian elimination
Matrix inversion

1-24
 Gauss Jordan Elimination
The method systematically transforms an augmented matrix into a

reduced form. The system corresponding to a reduced augmented matrix

is called a reduced system. As we shall see, reduced systems are easy to

solve.

1-25
The following matrices are in reduced form

1-26
1-27
Solution

1-28
1-29
1-30
1-31
1-32
1-33
1-34
1-35
1-36
1-37
1-38
1-39
1-40
Gaussian elimination

The Gaussian elimination method is known as the row reduction for

solving linear equations systems. It consists of a sequence of

operations performed on the corresponding matrix of coefficients.

To perform row reduction on a matrix, we have to complete a

sequence of elementary row operations to transform the matrix till

we get 0s (i.e., zeros) on the lower left-hand corner of the matrix as

much as possible. That means the obtained matrix should be an

The obtained matrix will be in row echelon form. The matrix is said
upper triangular matrix.
to be in reduced row-echelon form when all of the leading

coefficients equal 1, and every column containing a leading 1-41

Gaussian Elimination Gauss Jordan Elimination

1-42
Example

1-43
Solution

Reduce the augmented matrix of the system to a triangular form

1-44
1-45
Example 2

1-46
1-47
Matrix inversion

1-48
Matrix inversion using pivoting

Steps to find the inverse of a matrix using Gauss-Jordan method

1-49
Example

1-50
Example
Consider the matrix

1-51
1-52
we obtain

1-53
Section 2
SYSTEMS OF LINEAR
INEQUALITIES

1-54
1-55
1-56
Properties of feasible solutions

1-57
1-58
when considering a system of m inequalities with two

non-negative variables, the feasible region is described by the intersection of m half-planes

1-59
THEOREM 8.10 The feasible region M of system (8.13) is either empty or a convex set with at most a finite

number of extreme points.

1-60
Graphical solution

1-61
Example

1-62
Graphical solution

1-63
Example
Two goods G1 and G2 are produced by
means of two raw materials R1 and
R2 with the capacities of 50 and 80 units,
respectively. To produce 1 unit of G1, 1
unit of R1 and 1 unit of R2 are required. To
produce 1 unit of G2, 1 unit of R1 and 2
units of R2 are required. The price of G1 is
3 EUR per unit, the price of G2 is 2 EUR
per unit and at least 60 EUR worth of
goods need to be sold.
1-64
Let xi be the number of produced units of Gi, i ∈ {1, 2}. A feasible production programme has to satisfy the

following constraints:

This is a system of linear inequalities with only two variables, which can be easily solved graphically.

The convex set of feasible solutions has five extreme points described by the vectors xi (or points Pi), i = 1,

2, . . . , 5:

1-65
1-66
1-67
THEOREM 8.13 Any extreme point of the feasible region M of system (8.13) corresponds to at least one basic

feasible solution, and conversely, any basic feasible solution corresponds

exactly to one extreme point.

1-68
A solution procedure
According to Theorem 8.11 we have to generate all extreme points in order to describe the

feasible region of a system of linear inequalities. Using Theorems 8.13 and 8.14, respectively,

this can be done by generating all basic feasible solutions of the given system.

In this section, we restrict ourselves to the case when system (8.13) is given in the special form

Ax ≤ b, x ≥ 0, with b ≥ 0,

where A is an m×n matrix and we assume that r(A) = m

1-69
Inequalities are transformed into equations by introducing a slack variable in each constraint,

i.e. for the ith constraint

where ui ≥ 0 is a so-called slack variable (i ∈ {1, 2, . . . ,m}).

as an initial basic feasible solution, i.e. the variables x1,

x2, . . . , xn are the non-basic variables and the slack variables

u1, u2, . . . , um are the basic variables

1-70
 Example
Consider the system of linear inequalities presented in Example 8.10. Introducing slack variables u1, u2 and

u3, we obtain the initial tableau in rows 1 to 3 below. Now, the goal is to generate all basic feasible solutions

of the given system of linear equations

1-71
1-72
1-73
Therefore, in the above example, there are six basic feasible solutions and four basic infeasible solutions. From

rows 1 to 18, we get the following basic feasible solutions. (The basic

variables are printed in bold face.)

1-74
Deleting now the introduced slack variables u1, u2, u3, we get the corresponding extreme

points P1 with the coordinates (0, 0), P2 with the coordinates (4, 0), P3 with the coordinates

(17, 13) and P4 with the coordinates (10, 20). The fifth and sixth basic feasible solutions

correspond to extreme point P1 again. (In each of them, exactly one basic variable has value

zero.)

1-75
It can be seen that our computations have started from point P1, then we have moved to the adjacent extreme

point P2, then to the adjacent extreme point P3 and finally to P4.

1-76