You are on page 1of 5

4.

4 Newton-Raphson method

Let, [a0 , b0 ] be an initial interval containing the only root  of the given equation f ( x)  0 and let f (x) be

continuously differentiable sufficient number of times in [a0 , b0 ] .

Let x0  [a0 , b0 ] be an initial approximation to  . We set initial approximation of  as x0  a0 or b0 . Let

x1  x0  h be the exact root closer to x0 so that f ( x1 )  f ( x0  h)  0 , where h is sufficiently small.

Then, expanding f ( x0  h) by Taylor’s series about x0 , we obtain

h2
f ( x0  h)  f ( x0 )  hf ( x0 )  f ( x0 )  ...  0 .
2!

Now, neglecting the terms containing h 2 and higher powers of h , we have

f ( x0 )  hf ( x0 )  0 .

This implies

f ( x0 )
h .
f ( x0 )

Therefore, we take the first approximation to  as

x1  x0  h
f ( x0 )
 x0 
f ( x0 )

Continuing similarly, we get successive approximations (2 nd approximation, 3rd approximation and so on) to  as

f ( x1 )
x2  x1  ,
f ( x1 )

f ( x2 )
x3  x2  ,
f ( x2 )

and so on.

1
In general, the (n+1)th approximation to  is given by

f ( xn )
xn 1  xn  , n  0,1,2... (4.4.1)
f ( xn )

This is the iteration scheme of Newton-Raphson method.

4.4.1 Condition of convergence

The iteration scheme shows that, Newton-Raphson method is only a particular case of fixed point iteration method.

In this case, the equation f ( x)  0 can to be rewritten as

f ( x)
x   ( x)  x 
f ( x)

  f ( x)2  f ( x) f ( x)  f ( x) f ( x)


Therefore,  ( x)  1   
  f ( x)2   f ( x)2

Thus, the condition of convergence of Newton-Raphson method is

 ( x)  1, for all x  [a0 , b0 ]

f ( x) f ( x)
This implies 1, for all x  [a0 , b0 ]
 f ( x)2
2
Hence, the condition of convergence of Newton-Raphson method is f ( x) f ( x)  f ( x) , for all x  [a0 , b0 ] .

4.4.2 Order of convergence for Newton-Raphson method

We know that, error at n-th iteration  n    xn , i.e.   xn   n

Therefore,

 n2
0  f ( )  f ( xn   n )  f ( xn )   n f ( xn )  f (n ) , applying Taylor series expansion
2

2
where min{ , xn }  n  max{  , xn } .

f ( xn )
Now, iteration formula is xn 1  xn 
f ( xn )

Therefore,

f ( xn )
  xn 1    xn 
f ( xn )

  n 2 f ( n ) 
This implies,  n 1   n   n  
 2 f ( xn ) 

Hence,

 n 2 f ( n )
 n1   (4.4.2.1)
2 f ( xn )

where min{ , xn }  n  max{  , xn } .

This is the error equation.

If the iteration converges, then xn ,  n   .

Therefore,

 n 1 1 f ( )
lim 
n   n2 2 f ( )

which shows that the order of convergence of the Newton-Raphson method is 2 and the asymptotic error constant is

1 f ( )
.
2 f ( )

3
Example 8:

Find the real root of the equation 10 x  x  4  0 correct to six significant digits using Newton-Raphson method.

Solution:

Let f ( x)  10x  x  4  0 .

We first apply method of tabulation in order to find the location of rough value of the root.

Table 13: location of the root

x f (x)

0 -3

1 7

We note that f(0)<0 and f(1)>0. Thus the given equation changes its sign within the interval [0, 1]. Therefore, there

exists at least one real root of the equation within [0, 1].

Now, f ( x)  10x ln 10  1

We choose initial approximation x0  0 .

The successive iterations generated by the eq. (4.4.1) have been presented in Table 14.

Table 14: Table for finding real root

n f ( xn ) f  x n 1 
xn 1  xn 
f ( xn )

0 0.90837932 5.00641

4
1 0.65355360 1.15709

2 0.55178472 0.11453

3 0.53934062 0.00144869

4 0.53917915 2.39265E-7

5 0.53917912 6.18949E-15

Hence, the required real root of the given equation is 0.5391791 correct to six significant digits.

You might also like