Unit - 2
Rate of convergence of the above methods
Convergence-
A sequence of iterates is said to converge with order if
For some constant c > 0.
Note-
- If p = 1, the sequence is said to be linearly convergent to .
- If p = 2, the sequence is said to be quadratically convergent to .
- If p = 3, the sequence is said to be cubically convergent to . And so on.
Order and Rate of Convergence
Let be the given equation. Let α be its exact root.
Then
Then the positive number p is said to be order of convergence if,
If p=1 , then convergence is linear and if p=2 , then it is quadratic.
a) Bisection Mehtod: The percentage error defined by
The rate of convergence for bisection method is
i.e. it converges linearly.
b) Regula-Falsi Method: The rate of convergence is
Here p=1.618 and is linear.
c) Newton Raphson Method: The rate of convergence is
Here p=2 so the convergence is quadratic
d) Secant Method: The rate of convergence is
Order of convergence is non integer and is greater than one , so is super linear.
Note: Order of Convergence
1) Bisection Method : p=1 (linear)
2) Regula-Falsi Method: 1 < p < 2 (superlinear).
3) Newton Raphson Method: p = 2 (quadratic).
4) Secant Method: p=1.618 (superlinear).
Gaussian Elimination Method
In this method we eliminate successively the unknown so that the equation (1) remain with the single unknown and reduce to upper triangular system. At last with help of back substitution we calculate the values of the remaining unknowns.
Consider a system of n linear equation in n unknown
… …… ….. … (1)
… …… …. …
To convert the above system into upper triangular matrix we eliminate from the second, third, fourth …., n equations above by multiplying the first equation by added them to the corresponding equations second, third, fourth,…., n equation. We get
… …… ….. …
… …… …. …
Repeating the above method for we get finally the upper triangular form.
Upper Triangular form of above
……………………….. (i)
Thus .
Then we calculate the values of .
Note: In (i) the coefficient is the pivot element and the equation is called the pivot equation. If then the above method fails and if it is close to zero the round off error may occur.
If or very small compared to other coefficient of the equation, then we find the largest available coefficient in the column given below the pivot equation and then interchange the two rows to obtain new pivot variable this is known as partial pivoting.
Example1 Apply Gauss Elimination method to solve the equations:
Given Check Sum (sum of coefficient and constant)
-1 …. (I)
-16 …. (ii)
5 …. (iii)
(I)We eliminate x from (ii) and (iii)
Apply eq(ii)-eq(i) and eq(iii)-3eq(i) we get
-1 ….(i)
-15 ….(iv)
8 ….(v)
(II) We eliminate y from eq(v)
Apply
-1 ….(i)
-15 ….(iv)
73 ….(vi)
(III) Back Substitution we get
From (vi) we get
From (iv) we get
From (i) we get
Hence the solution of the given equation is
Example2 Solve the equation by Gauss Elimination Method:
Given
Rewrite the given equation as
… (i)
….(ii)
….(iii)
…(iv)
(I) We eliminate x from (ii),(iii) and (iv) we get
Apply eq(ii) + 6eq(i), eq(iii) -3eq(i), eq(iv)-5eq(i) we get
…(i)
….(v)
….(vi)
…(vii)
(II) We eliminate y from (vi) and (vii) we get
Apply 3.8 eq(vi)-3.1eq(v) and 3.8eq(vii)+5.5eq(v) we get
…(i)
….(v)
…(viii)
…(ix)
(III) We eliminate z from eq (ix) we get
Apply 9.3eq (ix) + 8.3eq (viii), we get
… (i)
….(v)
…(viii)
350.74u=350.74
Or u = 1
(IV) Back Substitution
From eq(viii)
Form eq(v), we get
From eq(i),
Hence the solution of the given equation is x=5, y=4, z=-7 and u=1.
Example3 Apply Gauss Elimination Method to solve the following system of equation:
Given … (i)
… (ii)
… (iii)
(I) We eliminate x from (ii) and (iii)
Apply we get
… (i)
… (iv)
… (v)
(II) We eliminate y from (v)
Apply we get
… (i)
… (vi)
… (vii)
(III) Back substitution
From (vii)
From (vi)
From (i)
Hence the solution of the equation is
Example4: Solve the system by Gauss Elimination method using partial pivoting
Given exact solution is
Given equations are
Using partial pivoting we rewrite the given equations as
(1)
(2)
Using Gauss elimination method
Multiplying (1) by (-0.0003120/0.5000) + (2) we get
Or
Substituting value of y in equation (1) we get
Hence
Example5: Solve the system of linear equations
Using partial pivoting by Gauss elimination method we rewrite the given equations as
(1)
(2)
(3)
Apply and
(1)
(4)
(5)
Apply )
(1)
(4)
Or .
Putting value of z in we get .
Putting values of y and z in we get .
Hence the solution of the equation is .
Gauss-Jordan method-
In this method we reduce the given system of equations AX = b to a diagonal system of equations Ix = d, here I is the identity matrix.
The reduced system gives the solution vector X.
Example: By using Gauss-Jordan method, solve the system of equations-
Sol:
Here the augmented matrix will be-
By performing elementary row transformation, and eliminations, we get-
Now making the pivots as 1, ((
We obtain,
Hence the solution of the system is-
Matrix Inversion using Gauss Jordan method-
We know that X will be the inverse of a matrix A if
Where I is an identity matrix of order same as A.
We write, using elementary operation we covert the matrix A in to a upper triangular matrix. Then compare this matrix with each corresponding column and using back substitution.
The determinant of this coefficient matrix is the product of the diagonal elements of the upper triangular matrix.
If this |A|=0 then inverse does not exist.
Example1
Solve the system of equations
Here A =
The augmented system is
Apply
Apply
Here
Which is non zero so the inverse exists.
Comparing the diagonal matrix with corresponding columns.
By back Substitution we get
Form first part of above
Similarly we solve second and third part and obtain a matrix whose columns are
Which is the required inverse
Example2: Find the inverse and determinant of-
Let [A:I]=
Apply
Apply
=
Apply
Here therefore exist.
On comparing the coefficient matrix with each column of RHS matrix.
From first term back substitution
We get
From second and third we get
Hence
Jacobi’s Iteration method:
Let us consider the system of simultaneous linear equation
The coefficients of the diagonal elements are larger than the all other coefficients and are non zero. Rewrite the above equation we get
Take the initial approximation we get the values of the first approximation of .
By the successive iteration we will get the desired the result.
Example: Use Jacobi’s method to solve the system of equations:
Since
So, we express the unknown with large coefficient in terms of other coefficients.
Let the initial approximation be
2.35606
0.91666
1.932936
0.831912
3.016873
1.969654
3.010217
1.986010
1.988631
0.915055
1.986532
0.911609
1.985792
0.911547
1.98576
0.911698
Since the approximation in ninth and tenth iteration is same up to three decimal places, hence the solution of the given equations is
Example: Solve by Jacobi’s Method, the equations
Given equation can be rewrite in the form
… (i)
..(ii)
..(iii)
Let the initial approximation be
Putting these values on the right of the equation (i), (ii) and (iii) and so we get
Putting these values on the right of the equation (i), (ii) and (iii) and so we get
Putting these values on the right of the equation (i), (ii) and (iii) and so we get
0.90025
Putting these values on the right of the equation (i), (ii) and (iii) and so we get
Putting these values on the right of the equation (i), (ii) and (iii) and so we get
Hence solution approximately is
Example: Use Jacobi’s method to solve the system of the equations
Rewrite the given equations
Let the initial approximation be
1.2
1.3
0.9
1.03
0.9946
0.9934
1.0015
Hence the solution of the above equation correct to two decimal places is
Gauss Seidel method:
This is the modification of the Jacobi’s Iteration. As above in Jacobi’s Iteration, we take first approximation as and put in the right hand side of the first equation of (2) and let the result be . Now we put right hand side of second equation of (2) and suppose the result is now put in the RHS of third equation of (2) and suppose the result be the above method is repeated till the values of all the unknown are found up to desired accuracy.
Example: Use Gauss –Seidel Iteration method to solve the system of equations
Since
So, we express the unknown of larger coefficient in terms of the unknowns with smaller coefficients.
Rewrite the above system of equations
Let the initial approximation be
3.14814
2.43217
2.42571
2.4260
Hence the solution correct to three decimal places is
Example: Solve the following system of equations
By Gauss-Seidel method.
Rewrite the given system of equations as
Let the initial approximation be
Thus the required solution is
Example: Solve the following equations by Gauss-Seidel Method
Rewrite the above system of equations
Let the initial approximation be
Hence the required solution is
Solving Eigen value problems (Power method)-
Consider the eigen value problem
The Eigen values of a matrix A are given by the roots of the characteristic equation
If the matrix A is of order n, then expanding the determinant, we obtain the characteristic equation as
For the given matrix we write the characteristic equation, by expanding we find the roots . These are called the Eigen values.
Now suppose be the solution of the system of homogeneous equations, corresponding to the Eigen value
These vectors are called the Eigen vectors.
To find the approximate values of all the Eigen values and Eigen vectors, iteration method or power method is used.
Power method is used when only the largest and/or the smallest Eigen values of a matrix are desired.
Procedure for Power method-
Step-1: First we choose an arbitrary real vector , basically is chosen as-
Step-2: Compute , , , , ………… Put
Step-3: Compute , ,
Step-4: The largest Eigen value is
The error in can be find as-
The Eigen vector corresponding to is
How do we determine the smallest Eigen value?
If is the Eigen value of A, then the reciprocal is the Eigen value of , then the reciprocal of the largest Eigen value of will be the smallest Eigen value of A.
Example: Find the largest Eigen value and the corresponding Eigen vector of the matrix
Also find the error in the value of the largest Eigen value.
Sol.
Let us choose the initial vector
Then
and
Now put , then-
Hence the largest Eigen value is-
And the corresponding Eigen vector is-
The error can be calculated as-
References:
- E. Kreyszig, “Advanced Engineering Mathematics”, John Wiley & Sons, 2006.
- P. G. Hoel, S. C. Port and C. J. Stone, “Introduction to Probability Theory”, Universal Book Stall, 2003.
- S. Ross, “A First Course in Probability”, Pearson Education India, 2002.
- W. Feller, “An Introduction to Probability Theory and Its Applications”, Vol. 1, Wiley, 1968.
- N.P. Bali and M. Goyal, “A Text Book of Engineering Mathematics”, Laxmi Publications, 2010.
- B.S. Grewal, “Higher Engineering Mathematics”, Khanna Publishers, 2000.