9 Gauss Elimination

Gauss Elimination
ES 204
Numerical Methods in Engineering
How to solve small number of
equations (n3) by graphical method?
Doable for 2 linear equations by plotting them on

Cartesian coordinates with one axis to x1 and the other to x2
For 3 equations, each equation is a plane in 3D

coordinate system
Point where 3 planes intersect is the solution
Beyond 3 equations, graphical methods break down;

little practical value
How to indicate singular and ill-
conditioned systems graphically?
2 equations are coincident hard to identify exact

2 equations as parallel lines
infinite number of solutions intersection point
no solution; singular
singular ill-conditioned
sensitive to roundoff error
How to solve small number of
equations (n3) by Cramers rule?
Assume we have [A]{x} = {b}
Each unknown is a fraction of 2 determinants with denominator D and numerator obtained
from D by replacing the column of coefficients by constants b1, b2, . . . , bn
Example:
Determinant of 2x2:
b11 a12 a13 a11 b12 a13 a11 a12 b13
b21 a22 a23 a21 b22 a23 a21 a22 b23
b a32 a33 a b32 a33 a a32 b33
Cramers rule for 3x3 A matrix: x1 31 x2 31 x3 31
a11 a12 a13 b11 a12 a13 b11 a12 a13
a21 a22 a23 b21 a22 a23 b21 a22 a23
a31 a32 a33 b31 a32 a33 b31 a32 a33
Example: Cramers rule
Use Cramers rule to solve
How to use Octaves det function?
For more than three equations, Cramers rule becomes impractical

Number of equations increases, determinants are time consuming to evaluate by hand/PC
Singular systems have D = 0

Ill-conditioned systems have D near 0 or very small
How to solve small number of equations
(n3) by elimination of unknowns?
Noncomputer solution technique; combines equations ; popular method
Eliminate x1
Follow directly from Cramers rule
Method extremely tedious to be

implemented manually
What is Naive Gauss Elimination?
Multiply by a21/a11
Subtract to eliminate x1
Solve a general set of n equations

Modified 2nd equation
Repeat to eliminate x1 from 2nd to n equations (e.g. how to eliminate x1 in 3rd equation?)
All equations from 2nd to n will be modified due to elimination of x1
pivot equation; a11 is the pivot element
1st step: eliminate x1 from 2nd to n equation (done!)

2nd step: eliminate x2 from 3rd to n equation
Now we use modified 2nd equation as pivot

equation, and a22 as pivot element
New set after x1 elimination
Multiply by a32/a22, then subtract to:
Eliminates x2 in 3rd equation

Eliminate x2 from 3rd to n equation using 2nd
equation as pivot
Continue using the remaining pivot equations

Final step is to use the (n 1)th equation to
eliminate the xn1 term from the nth equation

Prime symbol indicates number
of modification
FORWARD ELIMINATION Phase (upper triangular matrix)
Your set of equations after forward

elimination phase
Value of x from 1 to n-1 in terms of

coefficients and constants
BACK SUBSTITUTION Phase
Example: Naive Gauss Elimination
0.1
3x1 0.1x2 0.2 x3 7.85
3
0.3
3x1 0.1x2 0.2 x3 7.85
3
Solve x by Gauss elimination
0.190000
7.00333x2 0.293333x3 19.5617
7 .00333
New set after eliminating x1

in 2nd and 3rd equations
New set after eliminating x2 in 3rd equation

(upper triangular form)
Example: Naive Gauss Elimination
3
b a12 j x j
1
b21 a123 x3
2
j 3
x2 1

a 22 a122
3
b a10j x j

0
b10 a120 x2 a130 x3

1
j 2
x1 0

a 11 a110
exact solution
Verification of numerical solution

by substitution to original set
How to implement Naive Gauss
elimination in Octave?
coefficient matrix A and right-hand-side vector b
are combined as augmented matrix
forward elimination - 2 nested loops
outer loop moves down the matrix from one pivot row to the next
inner loop moves below the pivot row to each of the

subsequent rows where elimination is to take place
What is pivoting?
naive during elimination and back substitution phases, it is
possible that a division by zero can occur
pivot element is close to zero
if magnitude of pivot element is small compared to other
elements, round-off errors can be introduced
partial pivoting before normalization, determine the coefficient with the largest absolute
value in the column below the pivot element
Rows can then be switched so that the largest element is the pivot element
complete pivoting - If columns as well as rows are searched for the largest element and then
switched
Example: Partial Pivoting
Use Gauss elimination to solve
Exact solution is x1 = 1/3 and x2 = 2/3
Multiplying first equation by 1/(0.0003)

Eliminate x1 from the second equation
Substitute back into the first equation
Result is very sensitive to the number of

significant figures carried in the computation
Example: Partial Pivoting
If the equations are solved in reverse order
Forward elimination and substitution:
Results less sensitive to the number of

significant figures due to partial pivoting
How to implement Gauss elimination
with pivoting in Octave?
Same code with GaussNaive
except with the addition of the
pivoting code
determine the largest available

coefficient in the column below the pivot
element using the max function
y is the largest element in the

vector x, and i is the index
corresponding to that element
How to evaluate determinant with
Gauss elimination?
Assume you are done with forward elimination on a 3x3 system
In general for a triangular form matrix:
After forward elimination:
With pivoting:
p represents the number of times that rows are pivoted
Determinant has value in assessing system condition (ill-conditioned; singular)

Example: Solution of a Tridiagonal
System
Solve the following tridiagonal system:
Transform the matrix to upper triangular form

Multiply first equation by factor e2/f1 and
subtract result from second equation
Creates a zero in place of e2 ;transforms other
coefficients to new values
Notation
g2 is unmodified because the element above it in

the first row is zero
Perform a similar calculation for the

third and fourth rows:
Example: Solution of a Tridiagonal
System
upper triangular form
Back substitution to generate the final solution

How to implement solution of
tridiagonal system in Octave?
algorithm does not include partial pivoting
Final notes:
Computational effort for Gauss elimination is
proportional to n3
Tridiagonal systems proportional to n

9 Gauss Elimination

Uploaded by

Copyright:

Available Formats

9 Gauss Elimination

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

9 Gauss Elimination

Uploaded by

Copyright:

Available Formats

How can small systems of linear equations be solved graphically or analytically?

How can small systems of linear equations be solved graphically or analytically?

What techniques are used to solve larger systems of equations numerically?

What techniques are used to solve larger systems of equations numerically?

Gauss Elimination

Doable for 2 linear equations by plotting them on

For 3 equations, each equation is a plane in 3D

Beyond 3 equations, graphical methods break down;

2 equations are coincident hard to identify exact

For more than three equations, Cramers rule becomes impractical

Singular systems have D = 0

Follow directly from Cramers rule

Method extremely tedious to be

Solve a general set of n equations

pivot equation; a11 is the pivot element

1st step: eliminate x1 from 2nd to n equation (done!)

Now we use modified 2nd equation as pivot

Eliminates x2 in 3rd equation

Continue using the remaining pivot equations

New set after x2 elimination

Your set of equations after forward

Value of x from 1 to n-1 in terms of

New set after eliminating x1

New set after eliminating x2 in 3rd equation

b10 a120 x2 a130 x3

Verification of numerical solution

forward elimination - 2 nested loops

inner loop moves below the pivot row to each of the

Exact solution is x1 = 1/3 and x2 = 2/3

Multiplying first equation by 1/(0.0003)

Substitute back into the first equation

Result is very sensitive to the number of

Forward elimination and substitution:

Results less sensitive to the number of

determine the largest available

y is the largest element in the

In general for a triangular form matrix:

After forward elimination:

p represents the number of times that rows are pivoted

Determinant has value in assessing system condition (ill-conditioned; singular)

Transform the matrix to upper triangular form

g2 is unmodified because the element above it in

Perform a similar calculation for the

upper triangular form

Back substitution to generate the final solution

algorithm does not include partial pivoting

You might also like