1.

2 Gaussian Elimination

1
 Echelon Forms
• This matrix which have following properties is in reduced row-echelon
form :
1. If a row does not consist entirely of zeros, then the first nonzero
number in the row is a 1. We call this a leading 1.
2. If there are any rows that consist entirely of zeros, then they are
grouped together at the bottom of the matrix.
3. In any two successive rows that do not consist entirely of zeros, the
leader 1 in the lower row occurs farther to the right than the
leader 1 in the higher row.
4. Each column that contains a leader 1 has zeros everywhere else.
• A matrix that has the first three properties is said to be in row-echelon
form .
• A matrix in reduced row-echelon form is of necessity in row-echelon
form, but not conversely.

2
 Example-1: Row-Echelon &
Reduced Row-Echelon form
• reduced row-echelon form:
1 2 0
0 1
1 0 0 4  1 0 0  0 0 1 3 0 0
0 1 0 7 ,0 1  0

0 , 0 0 0 ,  
0 0  0 0
0 0 11  0 1  0 0 0 0
0
0
 row-echelon
form:
1 4  3 7  1 1 0 0 1 2 6 0
0 26, 0 01, 0 0
1 0 1 1
 0 5  0  1 
0 10 00 0 0 0
3
Example-2: Row-Echelon and Reduced
Row-Echelon form
• All matrices of the following types are in row-echelon form
( any real numbers substituted for the *’s. ) :
1 * * * * * * *
0 *
1 * * * 1 * * * 1 * *  0 0 1 * * * * *
    1  0 *
* * , 0 1 *, 0 *  0 0 0 1 * * * * *
0 1 * * , 0
* * 0   0 0 0  0 0 0 0 1 * * * 
 0 1 * 0  0 *
0 0 1 0 0 0  0 0 0 0 0 0 0 1
0 0 0 0 0 *

0
All 0matrices
0 1 0 of the following types are in reduced
row-
echelon form ( any real numbers 1 *substituted
0 0 0 * *for
0 the
 0 *
*’s.
1 ) :
0 0 0 1 0 0 * 1 0 *   0 0 1 0 0 * * 0 
     0 *
* 0 , 0 * , 0 *  0 0 0 1 0 * * 0 *
0 1 0   1 0   1 , 0

* 0 0 * 0 0 0  0 0 0 0 1 * * 0
0    0 *
1  0 0 0
0 0 1 0 1 0 0 0  0 0 0 0 0 0 0 1
0 * 4
0 0 0 0 0
 Example-3: Solutions of Linear Systems (a)

Suppose that the augmented matrix for a system

of linear equations have been reduced by row
operations to the given reduced row-echelon
form. Solve the system.
1 0 0 5
(a) 0 1 0
 0 4 

2 (a)
Solution
x 
0 1
the corresponding
system 5
of equations is :
y
5
 Example-3: Solutions of Four Linear Systems
(b1)

1 00 4 1
(b) 0 1 
0 62 
0 2 
0 1 3

Solution (b)
free
1. The x1  4x4  - variables
corresponding x2 1  2x4 
system of equations 6
is : x3  3x4  2
leading
variables

6
Example-3: Solutions of Four Linear Systems
(b2)

x1  - 1 - 2. We see that the free variable

4x4 can be assigned an arbitrary value,
say t, which then determines
x 23  26 -- 2 x44
values of the leading variables.
3x

3. There are infinitely

x1  1  4t,
many solutions, and the
general solution is x 2  6  2t,
given by the formulas x 3  2  3t,
x4  t

7
Example-3: Solutions of Four Linear Systems
(c1)

1 6 0 0 4  2
 1 
(c) 0 0 1 0 3 

0 
0 0 1 5 02  

0 0 0 0 0
Solution
(c)
1. The 4th row of zeros x1  6x2  4x5  - 2
leads to the equation
places no restrictions on x3  3x5  1
the solutions. Thus, we x4  5x5  2
can omit this equation.

8
 Example-3: Solutions of Four Linear Systems
(c2)
Solution (c)
x1  - 2 - 6x2 -
2. Solving for the leading
variables in terms of the 4x5
free variables: x3  1- 3x5
x4  2 - 5x5
3. The free variable can
be assigned an arbitrary x1  - 2 - 6 s - 4t
value , there are ,x  s
2
infinitely many solutions,
x3  1 -
and the general solution
3t
x4  2 -
is given by the formulas.
5t,
x4  t 9
Example-3: Solutions of Four Linear Systems(d)

1 0 0
 0 
0 0 1 

(d) 0 1 2
Solution (d):
0
the last equation in the corresponding
system of equation is
0 0x1  0x2  0x3 
1
Since this equation cannot be satisfied,
there is
no solution to the system.
10
 Elimination Methods (1/7)

• We shall give a step-by-step elimination

procedure that can be used to reduce any matrix
to reduced row-echelon form.

0 0 2 0 7 12
2 28
4 10 6 12
 2

4 5 6 5
1
11
 Elimination Methods (2/7)
• Step1. Locate the leftmost column that does not consist
entirely of zeros.
0 0 2 0 7 12
4 10 6 12 
2
28
Leftmost nonzero column
2 4 5 6 5
• 1
Step2.
 Interchange the top row with another row, to bring a
nonzero entry to top of the column found in Step1.

2 4 10 6 12 28 The 1th and 2th rows in

 0 0 2 0 7 12
the preceding matrix
 were interchanged.
2 4 5 6 5
1

12
 Elimination Methods (3/7)

• Step3. If the entry that is now at the top of the column found in
Step1 is a, multiply the first row by 1/a in order to introduce a
leading 1.
1 2 5 3 6 14
 0 0 2 0 7 12
The 1st row of the

2 4 5 6 5 preceding matrix was
1 multiplied by 1/2.
• Step4. Add suitable multiples of the top row to the rows below
so that all entires below the leading 1 become zeros.
1 2 5 3 6 14 
-2 times the 1st row of the
0 0 2 0 7 12 
preceding matrix was

added to the 3rd row.
0 0 5 0 17 
29
13
 Elimination Methods (4/7)

• Step5. Now cover the top row in the matrix and begin again
with Step1 applied to the submatrix that remains.
Continue in this way until the entire matrix is in
row-echelon form.


1
0 0 22 05 37 6 12  14 
Leftmost nonzero
 29
0 0 5 0 17 column in the
submatrix

1 2 5 3 6 14  The 1st row in the

0 0 1 0 2 6

7 submatrix was multiplied
by -1/2 to introduce a
0 0 5 0 17  leading 1.
29
14
 Elimination Methods (5/7)
• Step5 (cont.)
-5 times the 1st row of the
6 14 submatrix was added to the
1 2 5
  2nd row of the submatrix to
3 72
0 0 1 1
 6 introduce a zero below the
0 2  leading 1.
1
The top row in the submatrix was
01 02 05 3 6
14 
covered, and we returned again
00 0 1 0  72  6 Step1.
0 0 0 0 12 1 
Leftmost nonzero
column in the new
1 2 5 3 6 14  submatrix
0 0 1  072  6
The first (and only) row in
the new submetrix was
0 0 0 0 1 2 multiplied by 2 to
introduce a leading 1.
 The entire matrix is now in row-echelon
form. 15
 Elimination Methods (6/7)
• Step6. Beginning with las nonzero row and working upward, add
suitable multiples of each row to the rows above to introduce zeros
above the leading 1’s.
2 5 3 6 7/2 times the 3rd row of
1
the preceding matrix was
0 1
14 added to the 2nd row.

 0 1 0 0 
0  -6 times the 3rd row was
0 0 0 0 1 12 added to the 1st row.


2 0 3 0
10 02  5 13 00 120 5 times the 2nd row was
added to the 1st row.
0 00
01 00 10

 The last matrix is in reduced row-echelon
2 
form.
0 0 0 0 1 16
 Elimination Methods (7/7)

• Step1~Step5: the above procedure produces a row-

echelon form and is called Gaussian elimination.
• Step1~Step6: the above procedure produces a reduced
row-echelon form and is called Gauss-Jordan
elimination.
• Every matrix has a unique reduced row-echelon form
but a row-echelon form of a given matrix is not unique.

17