1

A Research Project
On The Topic

“ GaussElimination Method For Solving

System Of Linear Equation”

As a part of the course

For the Degree of
Bachelor of Science
2023-24

Under the Supervision of : Submitted By:

Dr. Sharon Moses Kriti
B.Sc. Mathematics

Department of Mathematics
St. John’s College, Agra
2

St. John’s College, Agra

Dr. Sharon Moses Department

of Mathematics

Certificate of Supervisor

This is certified that Kriti D/o of Mr.

Bhupendra singh who is submitting the
Research Project for the partial fulfilment
of the Degree of 'Bachelor of
Science' on the topic “Gauss
Elimination Method for solving
the System of Linear Equations”
conducted extensions and investigations
in the Department of Mathematics under
my supervision and guidance.
The research project embodies the work
of Kriti and to be best of my knowledge, it
has not been submitted to any other
3

institution or university for award of any

Degree or Diploma.

Date: Signature :

Acknowledgement

All the investigations and study regarding this

work have been done by me under the
appreciable and encouraging guidance of my
respected competent supervisor Dr. Sharon
Moses, Sen. Lecturer, Department of
mathematics , St. John’s College , Agra.
I would like to express my sincere gratitude to
all those who have contributed to the
completion of this research project .
First and foremost, I am deeply thankful to my
supervisor, Dr. Sharon Moses , for his
invaluable guidance, support, and
encouragement throughout this research
endeavour. His expertise, mentorship, and
constructive feedback have been instrumental
in shaping the direction and quality of this
work.
4

I would also like to extend my appreciation to

my fellow research scholars who provided
valuable insights, assistance, and
encouragement during various stages of this
project.
Furthermore, I am grateful to Department of
Mathematics, St. John’s College, Agra for their
support and resources, which facilitated the
execution of this research project.
I am also indebted to my parents for their
unwavering support and encouragement
throughout this journey. Their patience,
positivity, and belief in my abilities have been
a constant source of inspiration and
motivation.
Thank you to all who have contributed to this
project in any capacity. I am truly grateful for
the opportunity to undertake this research
endeavour.

Kriti
B.Sc. Mathematics
6th semester
5

Preface

It is with great pleasure and a sense of

acknowledgment that I present this
research project titled “ Gauss
Elimination Method: Theory,
Applications, and Advancements.”
This project represents a comprehensive
investigation into one of the foundational
6

techniques in numerical linear algebra,

the Gauss elimination method.
This project represents the diligent study
and analysis about working principle of
gauss elimination method , its working
criteria, applications in various problems,
reflection on the group of equations in
which this method gives low accuracy or
fails to solve the equations and its
alternative and modifications in working
rule.
This research project aims to detailed
study of method from how it evolve to
various improvements made in it.
I hope that the work will reflect analysis
of equations and area of applications of
method that I have done in this regard.
I trust that the content will contribute
meaningfully to the body of subject .

Contents
7

 Introduction
1. Linear Equations & Linear
Systems
2. Elementary Operations
3. Echelon Form
 Working rule : Consistent & Non
consistent systems
 Applications
 Limitations
 History of evolution of Method
 Conclusion
 References
8

Abstract
In this project our focus will be on the most basic
method for solving linear algebraic Systems, known
as Gaussian Elimination in honor of one of the all-
time mathematical Greats — the early nineteenth
century German mathematician Carl Friedrich
Gauss. As the father of linear algebra, his name will
occur repeatedly throughout this text. Gaussian
Elimination is quite elementary, but remains one of
the most important algorithms in applied (as well
as theoretical) mathematics. Our initial focus will
be on the most important class of systems: those
involving the same number of equations as
unknowns. Although we will eventually develop
techniques for handling completely general linear
systems.

Linear equations :
A linear equation in the variables x1, x2 , x3 , ......, xn
is an equation of the form
a1x1 + a1x2 + a3x3 + ............. +anxn = b
Where a1, a2, a3, ......., an and b are fixed numbers.
These fixed numbers are called parameters.
Typical linear equations are
x+2y = 4 and 7x+3y-z = 0
The equations are called linear because their
graphs are straight lines ( when it is equation in
9

two variables) and planes ( when it is an equation

in three variables).
The key feature of a linear equations is that each
term of the equation is either a constant term or a
term of order one (that is, a constant coefficient
times one of the variables).
Linear systems :
A linear system in the variables x1 , x2 , x3 , ........., xn
is a collection of one or more linear equations in
these variables. The general form of a linear

with m equations in n variables.

system

a11x11 + a12x12 +............. + a1nx1n = b1

a21x21 + a22x22 +..............+ a2nx2n = b2
. . .

.
. . .

.
. . .

am1xm1 + am2xm2 + .......... + amnxmn = bm

.

Where a11 , a12 ,......., amn and b1 , b2 ,......., bm are fixed

equations and n variables is called an m×n linear

numbers called parameters. A linear system with m

system.
3x+4y+7z =2 ;
x+2y+3z =5 ;
Example :

2x+y+z =6 ;
consists a 3×3
linear system.
Coefficient matrix & Augmented matrix :
10

A linear system can be completely described by a

rectangular table of numbers, called a matrix. The
coefficient matrix of a linear system is the matrix

variables, and is often denoted A. The augmented

consisting of all coefficients in front of the

matrix of the linear system is a matrix that

contains the entries in the coefficient matrix,
extended with the constants on the right side of

augmented matrix will be (m×n+1) for a linear

each equation, and it is often denoted Â. Order of

system of m equations in n variables.

In the above example,
3
3 4 7; 2
4 7

Coefficient matrix, A = 1 2 3 ; Augmented

matrix, Â = 1 2 3 ; 5
2 1 1
2 1 1;6

 Elementary Row Operations :

The following operations on the rows of matrix are
called Elementary Row Operations :
 To interchange two rows of matrix.
 To change a row by adding to it a multiple of
another row.
 To multiply each element in a row by the same
non-zero number.
The rows of the augmented matrix corresponds to
the equations of the linear system, and to perform
11

row operations on the matrix corresponds to

performing operations on the equations of the

row operations preserve the solutions of the linear

system. But the POINT TO BE NOTED that these

system.
Our aim is to use elementary row operations to
obtain a new and simpler form of the linear system.
We say that two linear systems are row equivalent
if you can get from one to the other using
elementary row operations.
What is the notation of elementary row operations?
Let’s take an example,
 ax +by + cz=f
bx +cy + az=g
cx + ay+ bz=h
Consist a system of linear
equations.

Augmented matrix, Â =
ab c; f
bc a; g
c ab;h

Now we will replace the second row with the sum

of the second row and (−1) times the first row.
With symbols, the elementary row operation

R2 → R2 + (-1) R1
represented as

Now, Â =
ab c; f
b−a c−b a−c ; g−f
c ab;h

 Echelon Forms :
12

An echelon form is a term used in the military to

describe the arrangement of rows (of troops or
ships etc.) in which each successive row extends
further than
the row in front of it . A 'leading entry' is the first
non-zero element in a row.
DEFINITION : A matrix is in the echelon form (or
row echelon form ) if it has following properties :
1.All non-zero rows are above of any rows of all
zeros.
2.Each leading entry of a row is in a column to
the right of the leading entry of the row above
it.
3.All entries in a column below a leading entry
are zeros.
For example:
∆ • • • • • • •
0 ∆ • • • • • •
0 0 0 ∆ • • • •
0 0 0 0 0 ∆ • •
0 0 0 0 0 0 ∆ •
0 0 0 0 0 0 0 ∆
In above matrix, the ∆s are leading entries (non-
zero) and •s can be any value.
DEFINITION : A matrix is in reduced echelon
form (or reduced row echelon form) if it is in
echelon form, and furthermore:
1.The leading entry in each nonzero row is 1.
13

2.Each leading 1 is the only nonzero entry in its

column.
For example :
1 0 • 0 0 0 • 0
0 1 • 0 0 0 • 0
0 0 0 1 0 0 • 0
0 0 0 0 1 0 • 0
0 0 0 0 0 1 • 0
0 0 0 0 0 0 0 0
Where •s can be any value.
NOTE : Each matrix is equivalent to one and only
one reduced echelon matrix. So , the leading
entries of any echelon form of a given matrix are in

 Pivot position : A pivot position in a matrix A

the same positions.

echelon form of A .
is the position of a leading 1 in the reduced

Gauss-Elimination
 WORKING RULE : Gaussian elimination is
an efficient method for solving any linear
system using systematic elimination of
variables. It can be described the following way
:
Stage-1 : Write down the augmented matrix
of the linear system.
Stage-2 : Use elementary row operations to
reduced the matrix to an echelon form.
14

Stage-3 : Solve the linear system of the

echelon form using back substitution.
Let’s the process with an example.

Example : A system of linear equations:

x−2 y + z=5 ;
2 x−5 y+ 4 z=−3;

Solution : The augmented matrix for this system

x−4 y +6 z=10 ;

is 3×4 matrix.
1−2 1 ; 5
2−5 4 ;−3
1−4 6 ; 10

Now we will reduce this matrix into echelon

form by using elementary row operations.
R2 → R2 – 2R1
R 3 → R3 – R1
i.e.
1−2 1 ; 5
0−12 ;−13
0−25 ; 5

The circled entry is the first nonzero entry in the

first row and all the entries below this are 0. Such a
circled entry is called a pivot.
R3 → R3 – 2R2
1−2 1 ; 5
0−12 ;−13
0 0 1 ; 31

The above reduced matrix represent the linear

system :
x−2 y + z=5
− y +2 z=−13
15

z=31

These equation can be solved successively by

Backward substitution.
z=31 ;
y=13+ 2 z=75 ;
x=5+2 y−z =124.

 We can continue Gaussian elimination to

simplify the augmented matrix further. This is
called the Gauss-Jordan Process. Here, we
ensure that all the pivots are equal to 1 and
moreover all the other entries in the column
containing the pivot are 0. In other words, we
have 0’s not only below but also above the
pivot. Matrix,
1−2 1 ; 5
0−12 ;−13
0 0 1 ; 31
Elementary operations ,
R1 → R1 – R3
R2 → 2R3 – R2

This gives,
1 0 0 ; 124
0 1 0 ; 75
0 0 1 ; 31

This simple augmented matrix quickly gives the

desired solution .
x=124 ; y =75 ; z=31;

This is an example of linear system with unique

solution. Now, we will see some examples of linear
16

system with infinitely many solutions and linear

system with no solutions .
 Linear system with infinitely many
solutions : First we will see an example then

Example :
after define the terms.
A system of linear equations:
x−2 y + z−u+ v=5 ;
2 x−5 y+ 4 z +u−v=−3 ;

Solution :
x−4 y +6 z +2u−v=10 ;

The augmented matrix for this

system is
1−2 1−1 1 ; 5
2−5 4 1−1 ;−3

Elementary row operations,

1−4 6 2−1 ; 10

R2 →R2 - 2R1 1−2 1−1 1 ; 5

R3 →R3 - R1 0−12 3−3;−13
→ 0−25 3−2 ;5

R3 →R3 - 2R2 1−2 1−1 1 ; 5

→ 0−12 3−3;−13
0 0 1−3 4 ; 31

R2 → (-R2 )
1−2 1−1 1 ; 5

→ 0 1−2−3 3; 13
0 0 1−3 4 ; 31

R1 → R1 + 2R2
→ 1 0−3−7 7 ; 31
0 1−2−3 3; 13
0 0 1−3 4 ; 31

R2 → R2 + 2R3 1 0 0−16 19 ; 124

17

R1 → R1 + 3R3 0 1 0−911; 75
→ 0 0 1−3 4 ; 31

By backward substitution,
x−16 u+19 v=124 ;
x=124 +16 u−19 v
y−9u+ 11v=75 ;
y=75+ 9 u−11 v
z−3 u+ 4 v =31;
z=31+3 u−4 v
Here we say that u and v are independent variables and
x, y, z are dependent variables.

This graph represent the solution space for above

example of linear system. The solution contains a plane
and all points that lies in the plane (i.e. infinite points) are
the solutions of linear system .
 Linear system with no solution : Let’s take

Example :
an example
A system of linear
equations:
x−5 y +4 z=3 ;
x−5 y +3 z=6 ;
18

Solution : The augmented matrix for this

2 x−10 y +13 z =5 ;

system is
1−5 4 ; 3
1−5 3 ; 6
2−10 13 ; 5

Elementary row operations,

R2 → R 2 - R 1 1−5 4 ; 3
R3 → R3 - 2R1 0 0−1; 3
→ 0 0 5 ;−1
R3 → R3 + 5R2
→ 1−5 4 ; 3
0 0−1; 3
0 0 0 ; 14

By backward substitution,
→ The bottom row corresponds to the
equation
0. z=14

→ Hence the system has no solutions.

Now, we will talk about the observations from the
above examples.

variables . Thus the augmented matrix  will be of

Consider a linear system of m equations in n

order (m×n+1) and reduced echelon form Û

Û=
19

The entries denoted by ∗ and the ci ’s are real

numbers; they may or may not be zero. The Pi ’s
denote the pivots; they are non-zero. Note that

of U and that any column of Û has at most one

there is exactly one pivot in each of the first r rows

pivot. Hence r ≤ m and r ≤ n.

than the number of equations) and cr+k 6 ≠ 0

1.If r < m (the number of non-zero rows is less

for some k ≥ 1, then the (r + k)th row

0 = cr+k and so the system has no solutions

corresponds to the self-contradictory equation

2.If (i) r = m or (ii) r < m and cr+k = 0 for all k ≥ 1,

(inconsistent system) .

then there exists a solution of the system

(consistent system) .

Û contains a pivot, then xj is called a basic

Basic and Free variable : If the jth column of

variable; otherwise xj is called a free variable

(independent variable.)

APPLICATIONS : The first historically known

application of the Gaussian elimination is for
solving a system of linear equations. In the
further development, this method can also be
used to find the rank and base of a matrix,
solve the inverse of an invertible square
matrix, calculate the determinant of a square
matrix. As for Gauss-Jordan reduction, it has
similar functions as Gaussian elimination,
20

except for calculating the determinant which is

not possible in Gauss-Jordan reduction. Both
methods rely on the elementary row operations
to reduce a matrix into an echelon form (in
Gaussian elimination) or a reduced echelon
form (in Gauss-Jordan reduction).
Let’s discuss about few applications of this
method.
 Finding determinant of a square matrix :
In this regard, Gauss has given few principles
for elementary row operations which are
describing different changes in the determinant
of some numerical. In order to do the
row operations we have to follow these rules
through the elimination process.
1.swapping of two different rows will result into
the value of the determinant multiplying by -1.
2.Multiplication of some scalar quantities to a
particular row of the determinant will result
into the determinant value multiplied by the
same scalar quantity.
3.Addition of two rows or multiplying scalar with
the row addition do not change the
determinant value.
If the Gaussian elimination has been applied to a
square matrix A to derive its row echelon matrix as
considered B, then the determinant of the provided
square matrix A will be the quotient by d with all
the products of the diagonal elements of B where d
is considered as the scalar product by which the
21

determinant was multiplied through the use of the

above three principles of Gaussian elimination.
Πdiag ( B )
det ( A )=
d

 To recognize the inverse of different

matrix : Various elimination techniques of
Gaussian and the related Gauss-Jordan
elimination would be used for deriving different
inverse matrices if the inverse of the
respective matrices exists. If a matrix is a
square matrix, then the row reduction could be
used for determining its respective inverse
matrix. Consider a matrix A of order n x n. For
finding its inverse, it should be observed that
the identity matrix of order n x n is an
augmented matrix that is to the right of the
matrix A. Augmented matrix will be like [A | I]
which is of the order of n x 2n. Now by using
the method of elementary row operations
formula, the reduced row echelon form could
be formed in the order of a n x 2n matrix. The
matrix A would be an invertible matrix if and
only if its left block of the matrix will be formed
as the identity matrix for the same respective
order. If the identity matrix could be formed
then the right block of the matrix would be the
final inverse matrix of A that would be termed
as A-1 . It has been concluded that if there
cannot be formed the left block as the identity
22

matrix then it should be considered as the not

For example : Matrix A has been considered as

invertible matrix.

following,

A =
2−1 0
−1 2−1
0−12
In order to find its respective inverse matrix A-1, the

considered and it would be the order of a 3×6 matrix.

following form of an augmented matrix can be

(A|I) =
2−1 0 ; 1 0 0
−1 2−1 ; 0 10
0−12 ; 0 0 1
Through applying the elimination method of Gauss or
Gauss-Jordan, the row reduced echelon form could be
obtained from the given augmented matrix as
following.
311
100;

(I|B) =
424
1 1
010; 1
2 2
113
001;
424

As it has been seen that the left block of the

augmented matrix has been obtained as an identity
matrix “I”, these elementary products could be
denoted by B. Thus, it has been showed that BA=I and
it implies that B=A-1 .

 To compute the basis and ranks of a matrix :

The algorithm for different Gaussian elimination
would also be applied in any m x n matrix to find its
respective bases and ranks. Through this way, a 6×9
23

matrix would be taken for evaluating its respective

row echelon expression like the following.

Where, the star marked position could be any arbitrary

entries and the a, b, c, d, e is considered as the nonzero
entries. The rank of the matrix could also be known from
this form. As there are 5 different nonzero rows in the
matrix T, 5 is considered as the rank of the matrix. It has
been also observed that the columns numbered 1, 3, 4, 7,
9 have elements a, b, c, d, e respectively. These will be
considered as the basis of the matrix A where the vector
space of the matrix A is spanned over these columns.

 LIMITATIONS : Gaussian elimination has been the

work horse of solving linear equations for a long
time. It is a good method for solving not too large
systems of equations.
There are some pitfalls, most notably the problem of loss
of accuracy due to small pivots. This usually can be
overcome by always choosing the largest element in the
column/row/ rest matrix as pivot.
There is another problem that rears its ugly head: fill-in.
Most practical problems result in very sparse matrices,
sometimes structured, sometimes unstructured. If you
use Gauss elimination that sparsity is partially or totally
lost and consequently calculation times soar. Today with
very large 3D problems with numbers of unknowns into
the millions Gauss elimination is no longer workable and
this has renewed interest in very powerful iterative
24

methods GMRES and Krylov space methods like BiCGStab

and IDRs.
The Gaussian elimination method may produce
inaccurate results when the terms in the augmented
matrix are rounded off.
For example , assume the following equation

0.1445 x +0.2546 y +0.3541 z=1

When you convert the system of equations into matrix

form, you might want to round off the co-efficients to say
2 significant digits (0.1445 would be rounded off to 0.14) .
And you need to carry out multiple operations on the
matrix to reduce it to it’s row echelon form, during the
course of which there is magnification of the error
percentage.
As a result of this, the final results obtained after back
substitution may be very much different from the
expected results. This is one of the biggest drawbacks of
gauss elimination method.

Background
History

Carl Friedrich Gauss

lived from 1777 to
1855, in Germany. He
is often called “the
greatest
mathematician since
antiquity.”
25

When Gauss was around 17 years old, he developed a

method for working with inconsistent linear systems,
called the method of least squares. A few years later (at
the advanced age of 24) he turned his attention to a
particular problem in astronomy. In 1801 the Sicilian
astronomer Piazzi discovered a (dwarf) planet, which he
named Ceres, in honor of the patron goddess of Sicily.
Piazzi took measurements
of Ceres’ position for 40 nights, but then lost track of it
when it passed behind the sun. Piazzi had only tracked
Ceres through about 3 degrees of sky. Gauss however
then succeeded in calculating the orbit of Ceres, even
though the task seemed hopeless on the basis of so few
observations. His computations were so accurate that the
astronomer Olbers located Ceres again later the same
year.
In the course of his computations Gauss had to solve
systems of 17 linear equations. Since Gauss at first
refused to reveal the methods that led to this amazing
accomplishment, some even accused him of sorcery.
Eight years later, in 1809, Gauss revealed his methods of
orbit computation in his book ‘Theoria Motus Corporum
Coelestium.’
Although Gauss invented this method (which Jordan then
popularized), it was a reinvention. As linear systems were
being solved by a similar method in China 2,000 years
earlier.

Conclusion
Gauss elimination method is an efficient method to solve
the system of linear equations. We have discussed above
about the step wise method of solving equations and
application of Gauss elimination and challenges while
26

using it. Although Computing technology has brought a

significant impact in resolving a particular computational
problem, e.g., in solving a system of linear equations, a
collection of equations with the same set of variables
which can be drawn as straight lines in the Cartesian
graph. In that case, the stability and efficiency of a
method is a primary concern, especially in a large
computation for there will be more time needed to solve
it Than to solve the smaller ones . Therefore, deciding
which method to use is an important thing. In this case,
this method one of the most widely used method.

References

1. Eivind Eriksen 'Linear Systems and Gaussian

Elimination' September 2, 2011 BI Norwegian
Business School.
2. Gaussian Elimination By Mark Crovella.
3. Prof. Neeta Nataraja IIT Bombay.
Journal Ilmu Komputer Indonesia (JIK) , Volume 7 , No
: 2, November 2022 'Conceptual and Practical
Review of Gaussian elimination and Gauss-Jordan
Reduction'.
4. International journal of Multidisciplinary educational
Research (Peer Reviewed and Refereed Journal:
VOLUME:13, ISSUE:1(1), JANUARY: 2024)’
Applications of Gaussian elimination approach in
different process of finding solution of a linear
equations system' Scopus Review ID:
A2B96D3ACF3FEA2A Online Copy of Article
Publication Available :http://www.ijmer.in
5. Article of Jos van Kan, Former assistant professor
Numerical analysis at Delft University of Technology
https://www.quora.com/profile/Jos-van-Kan .
27

6. Image, “Carl Friedrich Gauss” by Gottlieb Biermann

A. Wittmann (photo) – Gauß-Gesellschaft Göttingen
e.V. (Foto: A. Wittmann).. Licensed under Public
Domain via Wikimedia Commons.
••Thank you ••