Roots of Equations

Secant Method
Norhayati Rosli

SLIDE | 1
1
 Instructor

ASSOC. PROF. DR. NORHAYATI ROSLI

Centre for Mathematical Sciences (CSM)
Universiti Malaysia Pahang (UMP)
Expertise: Numerical Methods, Mathematical Modelling
(Deterministic & Stochastic), Probability Theory,
Risk Based Inspection (RBI)
Tel: +609-5492276
Email: norhayati@ump.edu.my
https://psm.ump.edu.my/norhayati

SLIDE | 2
 Roots of Equation
• Introduction
• Graphical Method
• Incremental Search
• Bracketing Method
✓Bisection Method
✓False Position Method
• Open Method
✓Newton Raphson Method
✓Secant Method
Prepared by Norhayati Rosli

SLIDE | 3
 Learning Outcomes

By the end of this topic, students should be able to:

✓ Determine the roots of the equation using secant method.

Prepared by Norhayati Rosli

SLIDE | 4
 Introduction
In many cases, the derivative of a function is very difficult to find or even is
not differentiable.
Alternative approach is by using secant method.
The slope in Newton’s Raphson method is substituted with backward finite
divided difference.

Secant Method
Introduction

f ( xi −1 ) − f ( xi )
f '( xi ) =
Raphson
Newton

xi −1 − xi

The secant method formula is:

 f ( xi )( xi −1 − xi ) 
xi +1 = xi −  
 f ( xi −1 ) − f ( xi )  Figure 7: Graphical Illustration of
Secant Method
Prepared by Norhayati Rosli

SLIDE | 5
 Secant Method
For the continuous function, 𝑓 𝑥 = 0:
Step 1: Choose initial values, 𝑥−1 and 𝑥0 . Find 𝑓(𝑥−1 ) and 𝑓(𝑥0 ).
Step 2: Compute the next estimate, 𝑥𝑖+1 by using secant method
formula
 f ( xi )( xi −1 − xi ) 
xi +1 = xi −  

Secant Method
Introduction

 f ( xi −1 ) − f ( xi ) 
Raphson
Newton

Step 3: Calculate the approximate percent relative error, 𝜀𝑎

xi +1 − xi
a = 100%
xi +1
Step 4: Compare 𝜀𝑠 with 𝜀𝑎 . If 𝜀𝑎 < 𝜀𝑠 , the computation is stopped.
Otherwise, repeat Step 2.
Prepared by Norhayati Rosli

SLIDE | 6
 Secant Method
Example 9
2 3
Determine one of the real root(s) of 𝑓 𝑥 = −12 − 21𝑥 + 18𝑥 − 2.4𝑥 by using
secant method with initial guesses of 𝑥−1 = 1.0 and 𝑥0 = 1.3. Perform the computation
until 𝜀𝑎 < 5%.
Solution

Secant Method
Introduction

First iteration,𝑥−1 = 1.0 and 𝑥0 = 1.3

Raphson
Newton

𝑓 1.0 =-17.4, 𝑓 1.3 = −14.528

 f ( x0 )( x−1 − x0 ) 
x1 = x0 −  
 f ( x−1 ) − f ( x0 ) 
 −14.1528(1 − 1.3) 
= 1.3 −   = 2.6075
 − 17.4 + 14.1528 
2.6075 − 1.3
a =  100% = 50.14%   s
2.6075
Prepared by Norhayati Rosli

SLIDE | 7
 Secant Method
Solution (Cont.)
Continue the second iteration and the results are summarised as follows:

No. of 𝑖 𝒙𝒊−𝟏 𝒙𝒊 𝒇(𝒙𝒊−𝟏 ) 𝒇(𝒙𝒊 ) 𝒙𝒊+𝟏 𝜺𝒂 (%)

Secant Method
Iteration
Introduction

1 0 1 1.3 -17.4 -14.1527 2.6075 50.14

Raphson
Newton

2 1 1.3 2.6075 -14.1528 13.0780 1.9796 31.72

3 2 2.6075 1.9796 13.0780 -1.6519 2.0500 3.44

Therefore, after three iterations the approximated root of 𝑓(𝑥) is 𝑥3 = 2.0500 with
𝜀𝑎 = 3.44%.

Prepared by Norhayati Rosli

SLIDE | 8
 Conclusion

Bracketing Method Open Method

Need two initial guesses Can involve one or more initial
guesses

Secant Method
The root is located within an interval Not necessarily bracket the root.
Introduction

prescribed by a lower and an

Raphson
Newton

upper bound.
Always work but converge slowly Do not always work (can diverge)
but when they do they usually
converge much more quickly.

Prepared by Norhayati Rosli

SLIDE | 9
 References

✓ Norhayati Rosli, Nadirah Mohd Nasir, Mohd Zuki Salleh, Rozieana Khairuddin,
Nurfatihah Mohamad Hanafi, Noraziah Adzhar. Numerical Methods, Second
Edition, UMP, 2017.
✓ Chapra, C. S. & Canale, R. P. Numerical Methods for Engineers, Sixth Edition,
McGraw–Hill, 2010.

SLIDE | 10
Thank You

SLIDE | 11