CHAPTER I

Numerical Solution of Equations of a Single Variable

 This chapter focuses on numerical solution of equations of a single variable, which
appear in the general form:

( )

Graphically, a solution (or root) of f(x) = 0 refers to the point of intersection of f(x)
and the x-axis. Therefore, depending on the nature of the graph of f(x) in relation to the
x-axis, the equation above may have a unique solution, multiple solutions, or no
solution.

A root of an equation can sometimes be determined analytically resulting in an exact

solution in closed form:

 For instance, the equation e3x − 2 = 0 can be solved analytically to obtain a

unique equation .
 In most situations, however, this is not possible and the root(s) must be found
numerically.

Numerical methods for solving an equation are divided into two main categories:

1.1 Bracketing methods:

1.1.1 Bisection Method:

It is assumed that f(x) is continuous on an interval [a, b] and has a root there so that
f(a) and f(b) have opposite signs, hence f(a)f(b) < 0.

Fig.1. Approximate solution of 2 − x + cos x = 0.

The procedure goes as follows:

 Locate the midpoint of [a, b], that is, . If f(a) and f(c1) have opposite
signs, the interval [a, c1] contains the root and will be retained for further
analysis; that is, the left end is retained while the right end is adjusted. If f(b) and
 f(c1) have opposite signs, we continue with [c1, b]; that is, the right end is kept
while the left end is adjusted.
 The process is repeated until the length of the most recent interval [a, b] satisfies
the desired accuracy.
 The initial interval [a, b] has length (b – a). Beyond that, the first generated
interval has length ( ) , the next interval ( ), and so on. Thus, the
nth interval constructed in this manner has length (b − a)/2n− 1, and because it
brackets the root, the absolute error associated with the nth iteration satisfies:
| | (b>a)
 it can be shown that the number N of iterations needed to meet the tolerance
condition satisfies:
( ) ( )

Fig.2. Method of bracketing methods.

 Example1: Use bisection method to find the root of f(x) = x3 − 10x2 + 5 = 0 that
lies in the interval (0.6, 0.8).

 Solution: f(x) is continuous in the interval [0.6, 0.8] and f(0.6)xf(0.8)=-1.4350<0.

a b f(a) f( ) f(b)

0.6 0.8 0.7 1.6160 0.4430 -0.8880

0.7 0.8 0.75 0.4430 -0.2031 -0.8880
0.7 0.75 0.725 0.4430 0.1248 -0.2031
0.725 0.75 0.7375 0.1248 -0.0379 -0.2031
0.725 0.7375 0.7312 0.1248 0.0438 -0.0379
0.7312 0.7375 0.7344 0.0444 0.0033 -0.0379
0.7344 0.7375 0.7360 0.0027 -0.0176 -0.0379
0.7344 0.7360 0.7352 0.0027 -0.0078 -0.0183
 0.7344 0.7352 0.7348 0.0027 -0.0026 -0.0078
0.7344 0.7348 0.7346 0.0027 4.5858e- -0.0026
005
0.7346 0.7348 0.7347 4.5858e- -0.0013 -0.0026
005
 Example2: Use bisection method to find the root of f(x) = x3 − 10x2 + 5 = 0 that
lies in the interval (0.6, 0.8) with MATLAB.
 Solution:
clc;clear all;close all
a=0.6;
b=0.8;
f=@(x)(x^3-10*x^2+5);
for i=1:100
c=(a+b)/2;
if f(a)*f(c)<0
b=c;
else
a=c;
end
if abs(f(a))<1.0e-4
break
end
end
fprintf(' The root is: %6.4f\n ',a)
fprintf('The number of iterations is: %d\n',i)

1.2 Open methods:

1.2.1 Newton–Raphson Method:

Newton’s method is the most commonly used open method to solve f(x) = 0, where f’
is continuous.
Consider the graph of f(x) in Fig.3.
 Start with an initial point x1 and locate the point (x1, f(x1)) on the curve.
 Draw the tangent line to the curve at that point, and let its x-intercept be x2. The
equation of this tangent line is:
( ) ( )( )
 Therefore, its x-intercept is found by setting y = 0 and solving for x:
( )
( )

 Once x2 is available, locate the point (x2, f(x2)), draw the tangent line to the curve
at that point, and let x3 be its x-intercept, which is found as:
( )
( )

 Continue this process until the sequence {xn} converges to the intended root r.
 Fig.3. Geometry of Newton’s method.

 Example3: Use the Newton’s method to find a root of the polynomial equation
f(x). For the initial point use x1=0.7:

f(x) = x3 − 10x2 +5.

 Solution:
The derivative of the function is : ( ) , so that:
x f(x) f’(x) ( )
( )

x1 0.7 0.4430 -12.5300 0.7354

x2 0.7354 -0.0104 -13.0856 0.7346
x3 0.7346 4.5858e-005 -13.0731 0.7346

 Example4: Use the Newton’s method to find a root of the polynomial equation
f(x) with MATLAB. For the initial point use x1=0.7:

f(x) = x3 − 10x2 +5.

 Solution:

clc;clear all;close all

x1=0.7;
f=@(x)(x^3-10*x^2+5);
df=@(x)(3*x^2-20*x);
x=x1;
for i=1:100
y=x;
x=y-(f(x)/df(x));
if x-y<1.0e-4
break
end
end
fprintf(' The root is: %6.4f\n ',x)
fprintf('The number of iterations is: %d\n',i)
1.3 MATLAB Built-In Functions:
 The fzero function finds the roots of f(x) = 0 for a real function f(x).

>> f=@(x)(x^3-10*x^2+5);
>> fzero(f,[0.6,0.8])
 roots(C) computes the roots of the polynomial whose coefficients are the
elements of the vector C.
>> C=[1,-10,0,5];
>> x=roots(C)
 solve seeks zeros of the expressions or solutions of the equations.
>> x=solve('x^3-10*x^2+5=0');
>> double(x)
Bibliography
[1] Ramin S Esfandiari. Numerical methods for engineers and scientists using
MATLAB
®. Crc Press, 2017.
[2] Kiusalaas Jaan. Numerical methods in engineering with matlab, 2009.
[3] Tao Pang. An introduction to computational physics, 1999.
[4] Chris Woodford and Chris Phillips. Numerical methods with worked examples:
Matlab edition. Springer Science & Business Media, 2011.