CSC431

Computational Science and Numerical

Method

Dr. O. J. Oyelade

1 2/10/2014
 Solving linear systems
Consider the system of ݊ linear equations and ݊
unknowns, given by:

ܽ ‫ݔ‬
ܽ11 ‫ݔ‬1 + ܽ
ܽ12 ‫ݔ‬
‫ݔ‬2 + · · · + ܽ
ܽ1݊
݊ ‫ܾ = ݊ݔ‬
ܾ1 ,
ܽ21 ‫ݔ‬1 + ܽ22 ‫ݔ‬2 + · · · + ܽ2݊ ‫ܾ = ݊ݔ‬2 ,
.
.
.
ܽ݊1 ‫ݔ‬1 + ܽ݊2 ‫ݔ‬2 + · · · + ܽ݊݊ ‫ = ݊ݔ‬bn .

2 2/10/2014
 We can write this system as the matrix
equation
Ax = b, with:

3 2/10/2014
 Gaussian Elimination
The standard numerical algorithm to solve a system of
linear equations is called Gaussian Elimination. We can
illustrate this algorithm by example.

Consider the system of equations:

‫ ݔ‬+ 2‫ݔ‬2
−3‫ݔ‬1 ‫ ݔ‬− ‫ݔ‬3
‫ = ݔ‬−1,
‫ ݔ‬− 6‫ݔ‬2
6‫ݔ‬1 ‫ ݔ‬+ 7‫ݔ‬3
‫ = ݔ‬−7,
3‫ݔ‬1 − 4‫ݔ‬2 + 4‫ݔ‬3 = −6.

4 2/10/2014
 To perform Gaussian elimination, we form an
Augmented Matrix by combining the matrix A with
the column vector b:

Row reduction is performed on the matrix as follows:

(1) multiply any row by a constant
(2) add multiple of one row to another row
(3) interchange the order of any rows
The goal is to convert the original matrix into an upper-
triangular matrix.
5 2/10/2014
We start with the first row of the matrix and work our way down
as follows:
1. We multiply the first row by 2
2. add it to the second row
3. and add the first row to the third row

6 2/10/2014
We then go to the second row. We multiply this row by −1
and add it to the third row:

The resulting equations can be determined from the matrix

and are given by:
‫ݔ‬1 + 2‫ݔ‬
−3‫ݔ‬ ‫ݔ‬2 − ‫ݔ‬3 = −1
‫ݔ‬2 + 5‫ݔ‬
−2‫ݔ‬ ‫ݔ‬3 = −9
‫ݔ‬3 = 2.
−2‫ݔ‬

7 2/10/2014
These equations can be solved by backward substitution,
starting from the last equation and working backwards. We
have
−2‫ݔ‬‫ݔ‬3 = 2 → ‫ݔ‬3 = −1
−2‫ݔ‬‫ݔ‬2 = −9 − 5‫ݔ‬‫ݔ‬3 = −4 → ‫ݔ‬2 = 2,
−3‫ݔ‬‫ݔ‬1 = −1 − 2‫ݔ‬‫ݔ‬2 + ‫ݔ‬3 = −6 → ‫ݔ‬1 = 2.

Therefore, we have:

8 2/10/2014
Extension of method of elimination to large sets of
equations by developing a systematic scheme
or algorithm to eliminate unknowns
and to back substitute.
• As in the case of the solution of two equations, the
technique for n equations consists of two phases:
– Forward elimination of unknowns
– Back substitution

9 2/10/2014
10 2/10/2014
Forward Elimination
A set of n equations and n unknowns
a11 x1 + a12 x2 + a13 x3 + ... + a1n xn = b1
a21 x1 + a22 x2 + a23 x3 + ... + a2 n xn = b2
. .
. .
. .
an1 x1 + an 2 x2 + an 3 x3 + ... + ann xn = bn

(n-1) steps of forward elimination

 Forward Elimination
Step 1
For Equation 2, divide Equation 1 by a11 and multiply by a21 .

 a21 
 a ( a11 x1 + a12 x2 + a13 x3 + ... + a1n xn = b1 )
 11 
a21 a21 a21
a21 x1 + a12 x2 + ... + a1n xn = b1
a11 a11 a11
 Forward Elimination
Subtract the result from Equation 2.
a21 x1 + a22 x2 + a23 x3 + ... + a2 n xn = b2
a21 a21 a21
− a21 x1 + a12 x2 + ... + a1n xn = b1
a11 a11 a11
_________________________________________________

 a21   a21  a21

 a22 −  
a12  x2 + ... +  a2 n − 
a1n  xn = b2 − b1
 a11   a11  a11

or a x + ... + a x = b
'
22 2
'
2n n
'
2
Forward Elimination
Repeat this procedure for the remaining equations to
reduce the set of equations as
a11 x1 + a12 x2 + a13 x3 + ... + a1n xn = b1
'
a22 x2 + a23
'
x3 + ... + a2' n xn = b2'
'
a32 x2 + a33
'
x3 + ... + a3' n xn = b3'
. . .
. . .
. . .

an' 2 x2 + an' 3 x3 + ... + ann

'
xn = bn'

End of Step 1
 Forward Elimination
Step 2
Repeat the same procedure for the 3rd term of Equation 3.

a11 x1 + a12 x2 + a13 x3 + ... + a1n xn = b1

'
a22 x2 + a23
'
x3 + ... + a2' n xn = b2'
"
a33 x3 + ... + a3" n xn = b3"
. .
. .
. .

an" 3 x3 + ... + ann

"
xn = bn"
End of Step 2
 Forward Elimination
At the end of (n-1) Forward Elimination steps, the system of equations will look like

a11 x1 + a12 x 2 + a13 x3 + ... + a1n x n = b1

'
a22 x2 + a23
'
x3 + ... + a2' n xn = b2'
"
a33 x3 + ... + a3" n xn = b3"
. .
. .
. .

( n −1) (n −1 )
ann xn = bn

End of Step (n-1)

Matrix Form at End of Forward
Elimination

a11 a12 a13 ⋯ a1n   x1   b1 

0 a'
a ' '
⋯ a 2 n x2     b2' 
 22 23    
0 0 a "
33 ⋯ a"3n   x3  =  b3" 
    
 ⋮ ⋮ ⋮ ⋯ ⋮  ⋮   ⋮ 
 0   xn  bn(n-1 ) 
(n −1 )
0 0 0 ann
Back Substitution Starting Eqns
a11 x1 + a12 x 2 + a13 x3 + ... + a1n x n = b1
'
a22 x2 + a23
'
x3 + ... + a2' n xn = b2'
"
a33 x3 + ... + an" xn = b3"
. .
. .
. .

( n −1) (n −1 )
ann xn = bn
 Back Substitution
Start with the last equation because it has only one unknown

( n −1)
b
xn = n
( n −1)
a nn
 Back Substitution
( n −1)
b
xn = n
( n −1)
a nn

bi(i −1) − ai(,ii−+11) xi +1 − ai(,ii−+12) xi + 2 − ... − ai(,in−1) xn

xi = (i −1) for i = n − 1,...,1
aii
n
(i −1) (i −1)
bi − ∑ aij x j
j =i +1
xi = (i −1) for i = n − 1,...,1
a ii
Gaussian elimination
Gaussian elimination
23 2/10/2014
Another example
The upward velocity of a rocket is given at three
different times

Table 1 Velocity vs. time data.

Time, t (s ) Velocity, v (m/s )

5 106.8
8 177.2
12 279.2

The velocity data is approximated by a polynomial as:

v(t ) = a1t 2 + a 2 t + a 3 , 5 ≤ t ≤ 12.

Find the velocity at t=6 seconds .
Example 1 Cont.
Assume

v (t ) = a 1 t 2 + a 2 t + a 3 , 5 ≤ t ≤ 12.

Results in a matrix template of the form:

 t 12 t1 1  a1   v1 
 2  a  = v 
t2 t2 1  2  2
 t 32 t3 1   a 3   v 3 


Using data from Table 1, the matrix becomes:

 25 5 1  a 1  106 . 8 
 64 8 1  a  = 177 . 2 
  2  
144 12 1  a 3   279 . 2 
Example 1 Cont.

 25 5 1  a 1   106 .8   25 5 1 ⋮ 106 .8 
 64 8 1  a  = 177 .2  ⇒  64 8 1 ⋮ 177 .2 
   2    
144 12 1  a 3   279 .2  144 12 1 ⋮ 279 .2 

1. Forward Elimination
2. Back Substitution
Forward Elimination
Number of Steps of Forward Elimination

Number of steps of forward elimination is

(n−1)=(3−1)=2
Forward Elimination: Step 1
 25 5 1 ⋮ 106.8  Divide Equation 1 by 25 and
 64 8 1 ⋮ 177.2 
  multiply it by 64, . 64
= 2.56
144 12 1 ⋮ 279.2 25
[25 5 1 ⋮ 106.8]× 2.56 = [64 12.8 2.56 ⋮ 273.408]

. [64 8 1 ⋮ 177.2]
− [64 273.408]
Subtract the result from Equation 2
12.8 2.56 ⋮
[0 − 4.8 − 1.56 ⋮ − 96.208]

 25 5 1 ⋮ 106.8 
Substitute new equation for Equation 2  0 − 4.8 − 1.56 ⋮ − 96.208
 
144 12 1 ⋮ 279.2 
Forward Elimination: Step 1 (cont.)
 25 5 1 ⋮ 106.8  Divide Equation 1 by 25 and
 0 − 4.8 − 1.56 ⋮ − 96.208
  multiply it by 144, . 144
144 12 1 ⋮ 279.2  = 5.76
25

[25 5 1 ⋮ 106.8]× 5.76 = [144 28.8 5.76 ⋮ 615.168]

.
Subtract the result from Equation 3 [144 12 1 ⋮ 279.2]
− [144 28.8 5.76 ⋮ 615.168]
[0 − 16.8 − 4.76 ⋮ − 335.968]

Substitute new equation for Equation 3 25 5 1 ⋮ 106.8 

 0 − 4.8 − 1.56 ⋮ − 96.208 
 
 0 − 16.8 − 4.76 ⋮ − 335.968
 Forward Elimination: Step 2
25 5 1 ⋮ 106.8  Divide Equation 2 by −4.8
 0 − 4.8 − 1.56 ⋮ − 96.208  and multiply it by −16.8,
 
 0 − 16.8 − 4.76 ⋮ − 335.968 − 16.8 .
= 3 .5
− 4.8
[0 − 4.8 − 1.56 ⋮ − 96.208]× 3.5 = [0 − 16.8 − 5.46 ⋮ − 336.728]

Subtract the result from Equation 3 [0 − 16.8 − 4.76 ⋮ 335.968]

− [0 − 16.8 − 5.46 ⋮ − 336.728]
[0 0 0.7 ⋮ 0.76]

Substitute new equation for Equation 3

25 5 1 ⋮ 106.8 
 0 − 4.8 − 1.56 ⋮ − 96.208
 
 0 0 0.7 ⋮ 0.76 
Back Substitution
 Back Substitution
25 5 1 ⋮ 106.8  25 5 1   a1   106.8 
 0 − 4.8 − 1.56 ⋮ − 96.2 ⇒  0 − 4.8 − 1.56 a  = − 96.208
     2  
 0 0 0.7 ⋮ 0.7   0 0 0.7   a3   0.76 

Solving for a3
0.7 a3 = 0.76
0.76
a3 =
0.7
a3 = 1.08571
Back Substitution (cont.)
25 5 1   a1   106.8 
 0 − 4.8 − 1.56 a  = − 96.208
   2  
 0 0 0.7   a3   0.76 

Solving for a2
− 4.8a2 − 1.56a3 = −96.208
− 96.208 + 1.56a3
a2 =
− 4.8
− 96.208 + 1.56 ×1.08571
a2 =
− 4.8
a2 = 19.6905
Back Substitution (cont.)
25 5 1   a1   106.8 
 0 − 4.8 − 1.56 a  = − 96.2
   2  
 0 0 0.7   a3   0.76 

Solving for a1
25a1 + 5a2 + a3 = 106.8
106.8 − 5a2 − a3
a1 =
25
106.8 − 5 ×19.6905 − 1.08571
=
25
= 0.290472
Naïve Gaussian Elimination Solution
 25 5 1  a1   106 .8 
 64 8 1   a  = 177 .2 
   2  
144 12 1  a 3   279 .2 

 a1  0.290472 
a  =  19.6905 
 2  
 a3   1.08571 
Example 1 Cont.
Solution  a1  0.290472
a  =  19.6905 
The solution vector is
 2  
 a3   1.08571 

The polynomial that passes through the three data points is then:

v (t ) = a1t 2 + a2t + a3
= 0.290472 t 2 + 19.6905t + 1.08571, 5 ≤ t ≤ 12

v 6 = 0.290472 6 + 19.6905(6 ) + 1.08571

( ) ( )2

= 129.686 m/s.
Naïve Gauss Elimination
Pitfalls
Pitfall#1. Division by zero
10 x2 − 7 x3 = 3
6 x1 + 2 x2 + 3x3 = 11
5 x1 − x2 + 5 x3 = 9

0 10 − 7  x1   3 
6 2 3   x2  = 11
    
5 − 1 5   x3   9 
Is division by zero an issue here?
12 x1 + 10 x2 − 7 x3 = 15
6 x1 + 5 x2 + 3x3 = 14
5 x1 − x2 + 5 x3 = 9

12 10 − 7  x1  15
6 5 3   x2  = 14
    
 5 − 1 5   x3   9 
 Is division by zero an issue here? YES
12 x1 + 10 x2 − 7 x3 = 15
6 x1 + 5 x2 + 3x3 = 14
24 x1 − x2 + 5 x3 = 28

12 10 − 7  x1  15 12 10 − 7  x1   15 

6 5 3   x2  = 14  0
     0 6.5  x2  = 6.5
    
24 − 1 5   x3  28 12 − 21 19   x3  − 2

Division by zero is a possibility at any step of forward

elimination
Pitfall#2. Large Round-off Errors
 20 15 10  x1   45 
− 3 − 2.249 7   x  = 1.751
   2  
 5 1 3   x3   9 
Exact Solution

 x1  1
 x  = 1
 2  
 x3  1
Pitfall#2. Large Round-off Errors
 20 15 10  x1   45 
− 3 − 2.249 7   x  = 1.751
   2  
 5 1 3   x3   9 
Solve it on a computer using 6 significant digits with chopping

 x1   0.9625 
 x  =  1.05 
 2  
 x3  0.999995
 Pitfall#2. Large Round-off Errors
 20 15 10  x1   45 
− 3 − 2.249 7   x  = 1.751
   2  
 5 1 3   x3   9 

Solve it on a computer using 5 significant digits with chopping

 x1   0.625 
 x  =  1.5 
 2  
 x3  0.99995
Is there a way to reduce the round off error?
 Avoiding Pitfalls
Increase the number of significant digits
• Decreases round-off error
• Does not avoid division by zero
 Avoiding Pitfalls
Gaussian Elimination with Partial Pivoting
• Avoids division by zero
• Reduces round off error
Gauss Elimination with
Partial Pivoting
Pitfalls of Naïve Gauss Elimination

Possible division by zero

Large round-off errors
 Avoiding Pitfalls
Increase the number of significant digits
• Decreases round-off error
• Does not avoid division by zero
 Avoiding Pitfalls
Gaussian Elimination with Partial Pivoting
• Avoids division by zero
• Reduces round off error
 What is Different About Partial
Pivoting?

At the beginning of the kth step of forward elimination, find the maximum of

akk , ak +1,k ,................, ank

If the maximum of the values is a pk
in the p th row, k ≤ p ≤ n, then switch rows p and k.
Matrix Form at Beginning of 2nd Step of
Forward Elimination

a11 a12 a13 ⋯ a1n   x1   b1 

 0 a' a '
⋯ ' 
a 2 n x2   ' 
b2
 22 23    
 0 a32
'
a '
33 ⋯ a3' n   x3  =  b3' 
    
 ⋮ ⋮ ⋮ ⋯ ⋮  ⋮   ⋮ 
 0 a'n 2 an' 3 an' 4 ann'
  xn  bn' 
Example (2nd step of FE)
6 14 5.1 3.7 6   x1   5 
0 − 7 6 1 
2 x2    − 6
    
0 4 12 1 11  x3  =  8 
    
0 9 23 6 8   x4   9 
0 − 17 12 11 43  x5   3 
Which two rows would you switch?
Example (2nd step of FE)

6 14 5.1 3.7 6   x1   5 
0 − 17 12 11 43  x   3 
  2   
0 4 12 1 11  x3  =  8 
    
0 9 23 6 8   x4   9 
0 − 7 6 1 2   x5  − 6

Switched Rows
 Gaussian Elimination
with Partial Pivoting
A method to solve simultaneous linear equations
of the form [A][X]=[C]

Two steps
1. Forward Elimination
2. Back Substitution
Forward Elimination
Same as naïve Gauss elimination method except that we switch
rows before each of the (n-1) steps of forward elimination.
Example: Matrix Form at Beginning of 2nd
Step of Forward Elimination

a11 a12 a13 ⋯ a1n   x1   b1 

 0 a' a'23 ⋯ a'2 n   x2  b2' 
 22    
 0 a32' '
a33 ⋯ a3n   x3  =  b3 
' '

    
 ⋮ ⋮ ⋮ ⋯ ⋮  ⋮   ⋮ 
 0 a'n 2 ' '
an 3 an 4 ann   xn  bn 
'
    '
Matrix Form at End of Forward
Elimination

a11 a12 a13 ⋯ a1n   x1   b1 

 0 a' a'23 ⋯ a'2 n   x2   b2' 
 22    
0 0 "
a33 ⋯ a3n   x3  =  b3 
" "

    
 ⋮ ⋮ ⋮ ⋯ ⋮  ⋮   ⋮ 
 0 0 ann   xn  bn 
   
(n −1 ) (n-1 )
0 0
Back Substitution Starting Eqns
a11 x1 + a12 x 2 + a13 x3 + ... + a1n x n = b1
'
a22 x2 + a23
'
x3 + ... + a2' n xn = b2'
"
a33 x3 + ... + an" xn = b3"
. .
. .
. .

( n −1) (n −1 )
ann xn = bn
Back Substitution
( n −1)
b
xn = n
( n −1)
a nn

n
(i −1) (i −1)
bi − ∑ aij x j
j =i +1
xi = (i −1) for i = n − 1,...,1
a ii
Example 2
Solve the following set of equations by
Gaussian elimination with partial pivoting

 25 5 1  a 1   106 .8 
 64 8 1   a  = 177 .2 
   2  
144 12 1  a 3   279 .2 
 Example 2 Cont.

 25 5 1  a 1   106 .8   25 5 1 ⋮ 106 . 8 
 64 1  a  = 177 .2  ⇒ 
 8  2    64 8 1 ⋮ 177 . 2 
144 12 1  a 3   279 .2  144 12 1 ⋮ 279 . 2 

1. Forward Elimination
2. Back Substitution
Forward Elimination
Number of Steps of Forward Elimination

Number of steps of forward elimination is (n−1)=(3−1)=2

Forward Elimination: Step 1

Examine absolute values of first column, first row
and below.
25 , 64 , 144
• Largest absolute value is 144 and exists in row 3.
• Switch row 1 and row 3.

 25 5 1 ⋮ 106.8  144 12 1 ⋮ 279.2

 64 8 1 ⋮ 177.2  ⇒  64 8 1 ⋮ 177.2 
   
144 12 1 ⋮ 279.2  25 5 1 ⋮ 106.8 
Forward Elimination: Step 1 (cont.)
144 12 1 ⋮ 279.2 Divide Equation 1 by 144 and
 64 8 1 ⋮ 177.2  multiply it by 64, . 64
  = 0.4444
 25 5 1 ⋮ 106.8  144

[144 12 1 ⋮ 279.2]× 0.4444= [63.99 5.333 0.4444 ⋮ 124.1]

.
Subtract the result from Equation 2 [64 8 1 ⋮ 177.2]
− [63.99 5.333 0.4444 ⋮ 124.1]
[0 2.667 0.5556 ⋮ 53.10]
Substitute new equation for Equation 2 144 12 1 ⋮ 279.2
 0 2.667 0.5556 ⋮ 53.10 
 
 25 5 1 ⋮ 106.8 
Forward Elimination: Step 1 (cont.)
144 12 1 ⋮ 279.2 Divide Equation 1 by 144 and
 0 2.667 0.5556 ⋮ 53.10  multiply it by 25, .25
  = 0.1736
 25 5 1 ⋮ 106.8  144

[144 12 1 ⋮ 279.2 ]× 0.1736 = [25.00 2.083 0.1736 ⋮ 48.47 ]

.
Subtract the result from Equation 3 [25 5 1 ⋮ 106.8]
− [25 2.083 0.1736 ⋮ 48.47]
[0 2.917 0.8264 ⋮ 58.33]

Substitute new equation for Equation 3 144 12 1 ⋮ 279.2

 0 2.667 0.5556 ⋮ 53.10
 
 0 2.917 0.8264 ⋮ 58.33
 Forward Elimination: Step 2

Examine absolute values of second column, second row
and below.
2.667 , 2.917
• Largest absolute value is 2.917 and exists in row 3.
• Switch row 2 and row 3.

144 12 1 ⋮ 279.2 144 12 1 ⋮ 279.2

 0 2.667 0.5556 ⋮ 53.10 ⇒  0 2.917 0.8264 ⋮ 58.33
   
 0 2.917 0.8264 ⋮ 58.33  0 2.667 0.5556 ⋮ 53.10
Forward Elimination: Step 2 (cont.)
Divide Equation 2 by 2.917 and
144 12 1 ⋮ 279.2 multiply it by 2.667,
 0 2.917 0.8264 ⋮ 58.33
  2.667
 0 2.667 0.5556 ⋮ 53.10  = 0.9143 .
2.917
[0 2.917 0.8264 ⋮ 58.33]× 0.9143 = [0 2.667 0.7556 ⋮ 53.33]

.
[0 2.667 0.5556 ⋮ 53.10]
− [0 53.33]
Subtract the result from Equation 3
2.667 0.7556 ⋮
[0 0 − 0.2 ⋮ − 0.23]

144 12 1 ⋮ 279.2 
Substitute new equation for Equation 3  0 2.917 0.8264 ⋮ 58.33 
 
 0 0 − 0.2 ⋮ − 0.23
Back Substitution
 Back Substitution
144 12 1 ⋮ 279.2  144 12 1   a1   279.2 
 0 2.917 0.8264 ⋮ 58.33  ⇒  0 2.917 0.8264 a  =  58.33 
     2  
 0 0 − 0.2 ⋮ − 0.23  0 0 − 0.2   a3  − 0.23

Solving for a3
− 0.2a3 = −0.23
− 0.23
a3 =
− 0.2
= 1.15
 Back Substitution (cont.)
144 12 1   a1   279.2 
 0 2.917 0.8264 a  =  58.33 
   2  
 0 0 − 0.2   a3   − 0.23

Solving for a2
2.917a2 + 0.8264a3 = 58.33
58.33 − 0.8264a3
a2 =
2.917
58.33 − 0.8264 ×1.15
=
2.917
= 19.67
 Back Substitution (cont.)
144 12 1   a1   279.2 
 0 2.917 0.8264 a  =  58.33 
   2  
 0 0 − 0.2   a3   − 0.23

Solving for a1
144a1 + 12a2 + a3 = 279.2
279.2 − 12a2 − a3
a1 =
144
279.2 − 12 ×19.67 − 1.15
=
144
= 0.2917
Gaussian Elimination with Partial Pivoting
Solution

 25 5 1  a 1   106 .8 
 64 8 1  a  = 177 .2 
  2  
144 12 1  a 3   279 .2 

 a1  0.2917 
a  =  19.67 
 2  
 a 3   1.15 
Partial Pivoting: Example
Consider the system of equations
10 x1 − 7 x2 = 7
− 3 x1 + 2.099 x2 + 6 x3 = 3.901
5 x1 − x2 + 5 x3 = 6
In matrix form
 10 − 7 0  x1   7 
− 3 2.099 6  x  3.901
   2 =  
 5 − 1 5  x3   6 

Solve using Gaussian Elimination with Partial Pivoting using five

significant digits with chopping
 Partial Pivoting: Example
Forward Elimination: Step 1
Examining the values of the first column
|10|, |-3|, and |5| or 10, 3, and 5
The largest absolute value is 10, which means, to
follow the rules of Partial Pivoting, we switch
row1 with row1.

Performing Forward Elimination

 10 − 7 0  x1   7  10 −7 0  x1   7 
− 3 2.099 6  x  = 3.901

 5
 2   
− 1 5  x3   6 
⇒  0 − 0.001 6  x  = 6.001

 0 2.5
 2   
5  x3   2.5 
 Partial Pivoting: Example
Forward Elimination: Step 2
Examining the values of the first column
|-0.001| and |2.5| or 0.0001 and 2.5
The largest absolute value is 2.5, so row 2 is
switched with row 3

Performing the row swap

10 −7 0  x1   7  10 −7 0  x1   7 
 0 − 0.001 6  x  = 6.001

 0 2.5
 2   
5  x3   2.5 
⇒ 0
 2 . 5 5   x  =  2. 5 
 2   
 0 − 0.001 6  x3  6.001
 Partial Pivoting: Example
Forward Elimination: Step 2

Performing the Forward Elimination results in:

10 − 7 0   x1   7 
 0 2.5 5   x  =  2.5 
  2   
 0 0 6.002  x3  6.002
 Partial Pivoting: Example
Back Substitution
Solving the equations through back substitution

10 − 7 0   x1   7  6.002
x3 = =1
 0 2.5 5   x  =  2.5  6.002
  2   
 0 0 6.002  x3  6.002 2.5 − 5 x3
x2 = = −1
2.5

7 + 7 x 2 − 0 x3
x1 = =0
10
 Partial Pivoting: Example
Compare the calculated and exact solution
The fact that they are equal is coincidence, but it
does illustrate the advantage of Partial Pivoting

 x1   0   x1   0 
[X ] calculated =  x2  = − 1 [X ] exact =  x 2  = − 1
 x3   1   x3   1 
81 2/10/2014
82 2/10/2014
Determinant of a Square Matrix
Using Naïve Gauss Elimination
Example
 Theorem of Determinants
If a multiple of one row of [A]nxn is added or subtracted to
another row of [A]nxn to result in [B]nxn then det(A)=det(B)
 Theorem of Determinants
The determinant of an upper triangular matrix [A]nxn is given by

det (A ) = a11 × a 22 × ... × a ii × ... × a nn

n
= ∏a
i =1
ii
Forward Elimination of a
Square Matrix

Using forward elimination to transform [A]nxn to an upper triangular matrix,

[U]nxn.

[A ]n × n → [U ] n × n

det ( A ) = det (U )
Example

Using naïve Gaussian elimination find the

determinant of the following square matrix.

 25 5 1
 64 8 1

144 12 1
Forward Elimination
 Forward Elimination: Step 1
 25 5 1 Divide Equation 1 by 25 and
 64 8 1
  multiply it by 64, . 64
= 2.56
144 12 1 25
[25 5 1]× 2.56 = [64 12.8 2.56]
. [64 8 1]
Subtract the result from Equation 2
− [64 12.8 2.56]
[0 − 4.8 − 1.56]

Substitute new equation for Equation 2  25 5 1 

 0 − 4.8 − 1.56
 
144 12 1 
Forward Elimination: Step 1 (cont.)
25 5 1  Divide Equation 1 by 25 and
 0 − 4.8 − 1.56
  multiply it by 144, . 144
= 5.76
144 12 1   25
[25 5 1]× 5.76 = [144 28.8 5.76]
.
[144 12 1]
Subtract the result from Equation 3
− [144 28.8 5.76]
[0 − 16.8 − 4.76]

Substitute new equation for Equation 3 25 5 1 

 0 − 4.8 − 1.56 
 
 0 − 16.8 − 4.76
 Forward Elimination: Step 2
Divide Equation 2 by −4.8
25 5 1  and multiply it by −16.8,
 0 − 4.8 − 1.56 
  − 16.8 .
 0 − 16.8 − 4.76 = 3 .5
− 4.8
([0 − 4.8 − 1.56]) × 3.5 = [0 − 16.8 − 5.46]
.
Subtract the result from Equation 3
[0 − 16.8 − 4.76]
− [0 − 16.8 − 5.46]
[0 0 0.7]
Substitute new equation for Equation 3 25 5 1 
 0 − 4.8 − 1.56
 
 0 0 0.7 
 Finding the Determinant
After forward elimination

 25 5 1 25 5 1 
 64 8 1 →  0 − 4.8 − 1.56
   
144 12 1  0 0 0.7 
.

det (A ) = u11 × u 22 × u 33
= 25 × (− 4 . 8 )× 0 . 7
= − 84 . 00
Summary
-Forward Elimination
-Back Substitution
-Pitfalls
-Improvements
-Partial Pivoting
-Determinant of a Matrix