Chapter 2

Solution of
Non-Linear
Equations
Introduction
Bisection Method
Regula-Falsi Method
Method of iteration
Newton - Raphson Method
Secant Method
Muller’s Method
Graeffe’s Root Squaring
Method
Secant
Method
We choose x0, x1 close to the root
‘a’ of f (x) =0 such that f (x0) f(x1)
As a next approximation x2 is
obtained as the point of
intersection of y = 0 and the chord
passing through the points
(x0, f(x0 )), (x1 f(x1 )).
f ( x1 )  f ( x0 )
y  f ( x0 )  ( x  x0 ),
x1  x0
Putting y = 0 we get
x0 f ( x1 )  x1 f ( x0 )
x2  x 
f ( x1 )  f ( x0 )
Generally

xn 1 f ( xn )  xn f ( xn 1 )
xn 1  , n  12,3
f ( xn )  f ( xn 1 )
This sequence converges to the
root ‘b’ of f (x) = 0 i.e. f( b ) = 0.
 Convergence of


Secant Method
Here the sequence is generated by the rule
xn 1 f ( xn )  xn f ( xn 1 )
xn 1 
f ( xn )  f ( xn 1 )
starting with x0 and x1 as
{x0 , x1 }
It will converge to ‘a’ ,that is f ( a ) = 0
The Secant method
converges faster
than linear and
slower than
Newton’s quadratic.
Example
Do three iterations of secant
method to find the root of

f ( x)  x  3x  1  0,
3

taking

x0  1, x1  0.5
n  1,
f ( x0 )  f (1)  1,
f ( x1 )  f (0.5)   0.375
x0 f ( x1 )  x1 f ( x0 )
x2 
f ( x1 )  f ( x0 )
(1)(0.375)  (0.5)(1)
  0.2
0.375  (1)
n  2,
f ( x2 )  f (0.2)  0.408
x1 f ( x2 )  x2 f ( x1 )
x3 
f ( x2 )  f ( x1 )
(0.5)(0.408)  0.2(0.375)
  0.3563
0.408  (0.375)
n  3,
f ( x3 )  f (0.3563)  0.02367
x2 f ( x3 )  x3 f ( x2 )
x4 
f ( x3 )  f ( x2 )
(0.2) f (0.3563)  0.3563 f (0.2)

f (0.3563)  f (0.2)
(0.2)(0.02367)  (0.3563)(0.408)
  0.3477
0.02367  0.408
x5  0.3473, f ( x5 )  0.0000096
and x5  x4  0.0004.
Though X5 is not the true
root, yet this is good
approximation to the root
and convergence is faster
than bisection.
Comparison:
In secant method, we do not
check whether the root lies in
between two successive
approximates Xn-1, and Xn.
This checking was imposed after
each iteration, in Regula –Falsi
method.
Muller’s
Method
In Muller’s method, f (x) = 0 is
approximated by a second
degree polynomial; that is by a
quadratic equation that fits
through three points in the
vicinity of a root. The roots of
this quadratic equation are
then approximated to the roots
of the equation f (x) = 0.
This method is iterative in
nature and does not
require the evaluation of
derivatives as in Newton-
Raphson method. This
method can also be used to
determine both real and
complex roots of f (x) = 0.
Suppose, xi  2 , xi 1 , xi
be any three distinct
approximations to a root
of f (x) = 0.
f ( xi 2 )  fi 2 , f ( xi 1 )  fi 1
f ( xi )  fi .
Noting that any three distinct
points in the (x, y)-plane
uniquely, determine a
polynomial of second degree.
A general polynomial of
second degree is given by

f ( x)  ax  bx  c
2
Suppose, it passes through the points

( xi 2 , fi 2 ), ( xi 1 , fi 1 ), ( xi , fi )

then the following equations will

be satisfied
ax 2
i 2  bxi  2  c  f i  2
ax  bxi 1  c  f i 1
2
i 1

ax  bxi  c  f i
i
2
Fig: Quadratic polynomial.
Eliminating a, b, c, we obtain

2
x x 1 f
2
xi  2 xi  2 1 fi 2
2
0
xi 1 xi 1 1 f i 1
2
xi xi 1 fi
which can be written as
( x  xi 1 )( x  xi )
f  fi 2
( xi  2  xi 1 )
( x  xi  2 )( x  xi )
 fi 1
( xi 1  xi  2 )( xi 1  xi )
( x  xi  2 )( x  xi 1 )
 fi
( xi  xi  2 )( xi  xi 1 )
That was a second degree
polynomial.
Now, introducing the notation
h  x  xi ,
hi  xi  xi 1 ,
hi 1  xi 1  xi 2
The above equation can be written as
The above equation can be written as
( h  hi )h
f  fi 2
 hi 1 (  hi 1  hi )
( h  hi  hi 1 )h
 f i 1
( hi 1 )(  hi )
(h  hi  hi 1 )(h  hi )
 fi
( hi  hi 1 )hi
We further define
h x  xi
 
hi xi  xi 1
hi
i 
hi 1
 i  1  i
With these substitution we
get a simplified Equation as
1
f  [ (  1) f i  2
2

i i

 (  1   )i i fi 1
i
1

(  1   )i fi ]
i
1
Or

f   ( fi  2   fi 1i i  fi i )
2
i
2
i
1

 [ f i  2   f 
i
2
i 1 i
2

 fi (i   i )] i
1
 fi
To compute  ,set f = 0, we obtain
i ( fi  2i  fi 1 i  fi )  gi    i fi  0
2

where
gi  f i  2 i  f i 1 i  f i (i   i )
2 2

A direct solution will lead to loss of

accuracy and therefore to obtain
max accuracy we rewrite as:
f i i gi
  i ( f i  2 i  f i 1 i  fi )  0
 2

so that,

1 gi  [ gi2  4 fi i i ( fi  2 i  fi 1 i  f i )]1/ 2


or  2 fi  i

2 f i i

gi  [ gi  4 f i i i ( f i  2 i  f i 1 i  f i )]
2 1/ 2

Here, the positive sign must be so

chosen that the denominator becomes
largest in magnitude.
we can get a better
approximation to the root, by
using

xi 1  xi  hi 
Example
Find the root of the
equation x – x–1 = 0
3

using Muller’s method.