Numerical Methods
Presented by: Dr. Soubhagya Baddi
Course Code: MC &NM
Course: 1BMATE201
Institution: DSCE
Date: 26-03-2026
Introduction
Numerical methods are mathematical techniques used to approximate
solutions to problems that cannot be solved analytically. They are essential in
engineering, physics, computer science, and applied mathematics, especially
when dealing with complex equations, large datasets, or real-world systems.
Importance of Numerical Methods:
Real-world problems are rarely exact: Many physical systems (fluid
dynamics, climate models, structural engineering) require approximations.
Computational efficiency: Algorithms allow solving problems too large for
symbolic methods.
Error control: Numerical methods quantify and minimize approximation
errors.
What Are Numerical Methods?
• Numerical methods are techniques for approximating solutions to
mathematical problems that are difficult or impossible to solve
analytically, such as integrals, differential equations, or nonlinear
equations. These methods relay on iterative calculations, where an initial
guess is refined through repeated steps until the solution is sufficiently
accurate.
• They are essential in fields like engineering, physics, finance, and data
science, where exact solutions are often impractical.
1. What is a Root of an Equation?
• A root (or solution) of an equation 𝑓 𝑥 = 0 is a value of 𝑥 that
satisfies the equation.
• Types of equations:
• Algebraic equations: polynomials
Example: 𝑥 3 − 4𝑥 + 1 = 0
• Transcendental equations: involve functions like 𝑒 𝑥 , sin 𝑥 , log 𝑥
Example: 𝑒 𝑥 − 𝑥 − 2 = 0
Many such equations cannot be solved exactly, so we use numerical
methods to find approximate roots.
Why Numerical Methods?
• Exact solutions may not exist or are difficult to find.
• Useful for engineering, physics, and real-world problems.
• Provide approximate answers with desired accuracy.
• We try to find a value of 𝑥such that: 𝑓 𝑥 ≈ 0
• Most methods start with:
• An initial guess or
• An interval where the root lies
Classification of Numerical Methods
(A) Bracketing Methods
• These methods trap the root in an interval.
1. Bisection Method
Based on Intermediate Value Theorem
• If 𝑓 𝑎 and 𝑓 𝑏 have opposite signs, root lies in 𝑎 𝑏
• 𝑓 𝑎 ⋅𝑓 𝑏 <0
• Idea:
• Divide interval into two halves repeatedly
• Select subinterval where sign changes
• Advantages:
• Always converges (guaranteed)
• Disadvantages:
• Slow
Classification of Numerical Methods
• 2. Regula-Falsi (False Position Method)
• Similar to bisection but uses a straight line approximation
• Faster than bisection in many cases
Classification of Numerical Methods
(B) Open Methods
• These methods do not require bracketing.
• 1. Newton-Raphson Method
• Uses tangent to approximate root.
𝑓 𝑥𝑛
• 𝑥𝑛+1 = 𝑥𝑛 − 𝑓′ 𝑥𝑛
• Idea:
• Start with initial guess
• Draw tangent and find next approximation
• Advantages:
• Very fast (quadratic convergence)
• Disadvantages:
• Requires derivative
• May diverge if guess is poor
Classification of Numerical Methods
• 2. Secant Method
• Similar to Newton but does not use derivative
• Uses two previous points
• Formula:
𝑓 𝑥𝑛 𝑥𝑛 −𝑥𝑛−1
• 𝑥𝑛+1 = 𝑥𝑛 −
𝑓 𝑥𝑛 −𝑓 𝑥𝑛−1
• Advantages:
• Faster than bisection
• No derivative needed
Classification of Numerical Methods
• 3. Fixed Point Iteration Method
• Rewrites equation as:
•𝑥=𝑔 𝑥
• Then iterates:
• 𝑥𝑛+1 = 𝑔 𝑥𝑛
• Condition for convergence:
• ∣ 𝑔′ 𝑥 ∣< 1
Graphical Interpretation
• Root is where graph of 𝑓 𝑥 intersects x-axis.
• Numerical methods approximate this intersection.
• geometric interpretation of the main numerical methods used to find
roots of equations 𝑓 𝑥 = 0 in terms of the graph of 𝑦 = 𝑓 𝑥 .
What does “finding a root” mean
geometrically?
• A root is simply the point where the curve cuts the x-axis.
•𝑦=𝑓 𝑥
• So, all numerical methods are different ways of approximating the x-
intercept of this curve.
Bisection Method (Interval Halving)
• Geometric Idea:
• Start with two points 𝑎and 𝑏such that:
• 𝑓 𝑎 and 𝑓 𝑏 have opposite signs
• This means the curve must cross the x-axis between them.
• 𝑓 𝑎 ⋅𝑓 𝑏 <0
• Visualization:
• You have a curve crossing the x-axis somewhere between 𝑎and 𝑏
• Take the midpoint 𝑚 = 𝑎 + 𝑏 /2
• Check which half contains the crossing
• Repeat
• Geometrically:
You are zooming into the x-axis intersection step by step.
Regula-Falsi Method (Chord Method)
• Geometric Idea:
• Instead of midpoint, draw a straight line (chord) between:
• 𝑎 𝑓 𝑎 and 𝑏 𝑓 𝑏
• Visualization:
• The chord intersects the x-axis at a point
• That intersection is taken as the next approximation
• Geometrically:
You replace the curve locally with a secant line and use its intercept.
• Interpretation:
• Better approximation than bisection because it considers slope of the
curve
Example 1:
• Using the Regula–Falsi method, find the root of the equation
c𝑜𝑠𝑥−𝑥=0 , that lies between 0 & 1, carry out three iterations.
Solution:
Step 1: Define the function
• 𝑓 𝑥 = cos 𝑥 − 𝑥
• At 𝑥 = 0: 𝑓 0 = 1 − 0 = 1 > 0
• At 𝑥 = 1: 𝑓 1 = cos 1 − 1 ≈ 0.5403 − 1 = −0.4597 < 0
• So, the root lies between 0 and 1.
Example 1:
• Step 2: Regula–Falsi formula
• If 𝑓 𝑎 and 𝑓 𝑏 have opposite signs, the next approximation is:
𝑎𝑓 𝑏 −𝑏𝑓 𝑎
•𝑥=
𝑓 𝑏 −𝑓 𝑎
• Step 3: Iterations
• Iteration 1
• 𝑎 = 0, 𝑏 = 1
0⋅𝑓 1 −1⋅𝑓 0 −1 −1
• 𝑥1 = = = ≈ 0.685
𝑓 1 −𝑓 0 −0.4597−1 −1.4597
• 𝑓 𝑥1 = cos 0.685 − 0.685 ≈ 0.774 − 0.685 = 0.089 > 0
• So, new interval: 0.685 1 .
Example 1:
• Iteration 2
• 𝑎 = 0.685, 𝑏 = 1
0.685⋅𝑓 1 −1⋅𝑓 0.685 0.685 −0.4597 −1 0.089
• 𝑥2 = =
𝑓 1 −𝑓 0.685 −0.4597−0.089
−0.315−0.089 −0.404
• 𝑥2 = ≈ ≈ 0.737
−0.5487 −0.5487
• 𝑓 𝑥2 = cos 0.737 − 0.737 ≈ 0.741 − 0.737 = 0.004 > 0
• New interval: 0.737 1 .
Example 1:
• Iteration 3
• 𝑎 = 0.737, 𝑏 = 1
0.737⋅𝑓 1 −1⋅𝑓 0.737 0.737 −0.4597 −1 0.004
• 𝑥3 = =
𝑓 1 −𝑓 0.737 −0.4597−0.004
−0.339−0.004 −0.343
• 𝑥3 = ≈ ≈ 0.739
−0.4637 −0.4637
• 𝑓 𝑥3 = cos 0.739 − 0.739 ≈ 0.739 − 0.739 ≈ 0.0001
Example 1:
• After 3 iterations, the root is approximately:
• 𝑥 ≈ 0.739
• This matches the known solution of cos 𝑥 = 𝑥 ,which is about
0.739085.
Graphical Representation
Interpretation
• Iteration 1: The secant line between 𝑓 0 and 𝑓 1 intersects at 𝑥1
≈ 0.685.
• Iteration 2: New interval 0.685 1 gives 𝑥2 ≈ 0.737
• Iteration 3: Interval 0.737 1 gives 𝑥3 ≈ 0.739.
• Convergence: The method stabilizes at the root 𝑥 ≈ 0.739, where
𝑓 𝑥 ≈ 0.
Example 2:
Using the Regula–Falsi method, find the root of the equation
𝑥 3 − 4𝑥 − 9 = 0. Correct to four decimal places.
• Solution: We solve
𝑓 𝑥 = 𝑥 3 − 4𝑥 − 9 = 0
• Step 1: Choose initial interval
Check values:
𝑓 2 = 8 − 8 − 9 = −9, 𝑓 3 = 27 − 12 − 9 = 6
Since signs are opposite, the root lies in 2 3 .
Example 2:
• Step 2: Regula–Falsi formula
𝑎𝑓 𝑏 −𝑏𝑓 𝑎
• 𝑥=
𝑓 𝑏 −𝑓 𝑎
• Iterations
• Iteration 1:
2 6 − 3 −9 39
𝑥1 = = = 2.6000
6 − −9 15
𝑓 2.6 = −1.824 negative
New interval: 2.6 3
• Iteration 2:
2.6 6 − 3 −1.824
𝑥2 = = 2.6930
6 − −1.824
𝑓 2.693 ≈ −0.242
Example 2:
• New interval: 2.693 3
• Iteration 3:
• 𝑥3 = 2.7050
𝑓 2.705 ≈ −0.010
• Iteration 4:
• 𝑥4 = 2.7055
• 𝑓 2.7055 ≈ −0.002
• Iteration 5 (refinement):
• Final answer: 𝑥 ≈ 2.7056
Graphical Representation
Example 3:
• Using the Regula–Falsi method, find the root of the equation 𝑥− 3𝑒−𝑥=0, carry out three iterations.
• Solution: We solve the equation
𝑓 𝑥 = 𝑥 − 3𝑒 −𝑥 = 0
• Step 1: Choose initial interval
• Check values:
• 𝑓 0 = 0 − 3 = −3 negative
3
• 𝑓 1 = 1 − 3𝑒 −1 = 1 − 𝑒 ≈ −0.103 negative
• 𝑓 2 = 2 − 3𝑒 −2 ≈ 2 − 0.406 = 1.594 positive
• So, the root lies in 1 2 .
• Step 2: Regula–Falsi Formula
𝑎𝑓 𝑏 −𝑏𝑓 𝑎
• 𝑥= 𝑓 𝑏 −𝑓 𝑎
Example 3:
• Iteration 1
• Take 𝑎 = 1, 𝑏 = 2
• 𝑓 1 = −0.103, 𝑓 2 = 1.594
1 1.594 −2 −0.103 1.594+0.206
• 𝑥1 = = = 1.0606
1.594− −0.103 1.697
• 𝑓 𝑥1 = 1.0606 − 3𝑒 −1.0606 ≈ 0.0216 positive
• New interval: 1 1.0606
• Iteration 2
• 𝑎 = 1, 𝑏 = 1.0606
1 0.0216 −1.0606 −0.103 0.0216+0.1092
• 𝑥2 = = = 1.0496
0.0216− −0.103 0.1246
• 𝑓 𝑥2 ≈ −0.0010 negative
• New interval: 1.0496 1.0606
Example 3:
• Iteration 3
• 𝑎 = 1.0496, 𝑏 = 1.0606
1.0496 0.0216 −1.0606 −0.0010 0.0227+0.0011
• 𝑥3 = = = 1.0507
0.0216− −0.0010 0.0226
• 𝑓 𝑥3 ≈ 0.0001 positive
• Final Answer (after 3 iterations):
• 𝑥 ≈ 1.0507
Graphical Representation
Example 4:
Find the approximate value of the real root of the equation 𝑥^3− 3𝑥+4=0
using the method of Regula-Falsi. (Carry out three iterations).
• Solution: We solve
𝑓 𝑥 = 𝑥 3 − 3𝑥 + 4 = 0
• Step 1: Choose initial interval
• 𝑓 −3 = −27 + 9 + 4 = −14 negative
• 𝑓 −2 = −8 + 6 + 4 = 2 positive
• So, the root lies in − 3 − 2 .
• Step 2: Regula–Falsi Formula
𝑎𝑓 𝑏 −𝑏𝑓 𝑎
•𝑥=
𝑓 𝑏 −𝑓 𝑎
Example 4:
• Iteration 1
• 𝑎 = −3, 𝑏 = −2
• 𝑓 𝑎 = −14, 𝑓 𝑏 = 2
−3 2 − −2 −14 −6−28
• 𝑥1 = = = −2.125
2− −14 16
3
• 𝑓 𝑥1 = ൫−2.125) − 3 −2.125 + 4 = −9.595 + 6.375 + 4
= 0.780 positive
• New interval: − 3 − 2.125
Example 4:
• Iteration 2
• 𝑎 = −3, 𝑏 = −2.125
• 𝑓 𝑎 = −14, 𝑓 𝑏 = 0.780
−3 0.780 − −2.125 −14 −2.34−29.75
• 𝑥2 = = = −2.171
0.780− −14 14.78
• 𝑓 𝑥2 ≈ 0.131 positive
• New interval: − 3 − 2.171
Example 4:
• Iteration 3
• 𝑎 = −3, 𝑏 = −2.171
• 𝑓 𝑎 = −14, 𝑓 𝑏 = 0.131
−3 0.131 − −2.171 −14 −0.393−30.394
• 𝑥3 = = = −2.178
0.131− −14 14.131
• 𝑓 𝑥3 ≈ 0.020 positive
• Final Answer (after 3 iterations):
• 𝑥 ≈ −2.178
Graphical Representation
Newton–Raphson Method (Tangent Method)
• Geometric Idea:
• At a point 𝑥𝑛 ,draw the tangent to the curve.
𝑓 𝑥𝑛
• 𝑥𝑛+1 = 𝑥𝑛 −
𝑓′ 𝑥𝑛
• Visualization:
• Start at a point on the curve 𝑥𝑛 𝑓 𝑥𝑛
• Draw the tangent line there
• Extend it to meet the x-axis
• That intersection gives the next point 𝑥𝑛+1
• Geometrically:
You are replacing the curve by its tangent (best linear approximation).
• Interpretation:
• Very fast because tangent closely follows the curve near the root
Example 1:
Using the Newton-Raphson method, find a real root of 𝑥 𝑠𝑖𝑛𝑥+ 𝑐𝑜𝑠𝑥=0
nearer to x = π, ( Here x is in radians). Carry out three iterations.
Solution: We are given the equation
𝑥 sin 𝑥 + cos 𝑥 = 0
and we need to find a real root near 𝑥 = 𝜋using the Newton–Raphson
method.
• Step 1: Define the function
• Let 𝑓 𝑥 = 𝑥 sin 𝑥 + cos 𝑥
• Using the product rule, the derivative is: 𝑓 ′ 𝑥 = 𝑥 cos 𝑥
Example 1:
Step 2: Newton–Raphson formula
The iteration formula is:
𝑓 𝑥𝑛
𝑥𝑛+1 = 𝑥𝑛 − ′
𝑓 𝑥𝑛
Step 3: Initial approximation
Since the root is near 𝑥 = 𝜋, 𝑥0 = 𝜋 ≈ 3.1416
Step 4: Iterations
Iteration 1:
𝑓 𝑥0 = 𝜋 ⋅ sin 𝜋 + cos 𝜋 = 0 − 1 = −1
𝑓 ′ 𝑥0 = 𝜋 ⋅ cos 𝜋 = 𝜋 −1 = −3.1416
−1
𝑥1 = 3.1416 − = 3.1416 − 0.3183 = 2.8233
−3.1416
Example 1:
Iteration 2:
𝑓 2.8233 ≈ 2.8233 ⋅ sin 2.8233 + cos 2.8233 ≈ 2.8233 0.312 + −0.950
≈ 0.882 − 0.950 = −0.068
𝑓 ′ 2.8233 = 2.8233 ⋅ cos 2.8233 ≈ 2.8233 −0.950 = −2.682
−0.068
𝑥2 = 2.8233 − = 2.8233 − 0.0254 = 2.7979
−2.682
Iteration 3:
𝑓 2.7979 ≈ 2.7979 ⋅ sin 2.7979 + cos 2.7979 ≈ 2.7979 0.334 + −0.943
≈ 0.934 − 0.943 = −0.009
𝑓 ′ 2.7979 ≈ 2.7979 −0.943 = −2.638
−0.009
𝑥3 = 2.7979 − = 2.7979 − 0.0034 = 2.7945
−2.638
𝑥 ≈ 2.7945
Example 2:
Using the Newton-Raphson method, find a real root of the equation
𝑒 𝑥 − 4x = 0, correct to four decimal places.
Solution: We need to solve
𝑒 𝑥 − 4𝑥 = 0
using the Newton–Raphson method, correct to four decimal places.
Define the function 𝑓 𝑥 = 𝑒 𝑥 − 4𝑥
𝑓′ 𝑥 = 𝑒𝑥 − 4
Newton–Raphson Formula
𝑓 𝑥𝑛
𝑥𝑛+1 = 𝑥𝑛 − ′
𝑓 𝑥𝑛
Example 2:
• Step 2: Choose Initial Approximation
Check values:
𝑓 0 =1>0
𝑓 1 = 𝑒 − 4 ≈ −1.282 < 0
So, the root lies between 0and 1.
Take:
𝑥0 = 0.5
• Step 3: Iterations
Iteration 1:
𝑓 0.5 = 𝑒 0.5 − 2 ≈ 1.6487 − 2 = −0.3513
𝑓 ′ 0.5 = 1.6487 − 4 = −2.3513
−0.3513
𝑥1 = 0.5 − = 0.5 − 0.1495 = 0.3505
−2.3513
Example 2:
• Iteration 2:
𝑓 0.3505 ≈ 𝑒 0.3505 − 1.402 ≈ 1.4197 − 1.402 = 0.0177
𝑓 ′ 0.3505 ≈ 1.4197 − 4 = −2.5803
0.0177
𝑥2 = 0.3505 − = 0.3505 + 0.0069 = 0.3574
−2.5803
• Iteration 3:
𝑓 0.3574 ≈ 𝑒 0.3574 − 1.4296 ≈ 1.4303 − 1.4296 = 0.0007
𝑓 ′ 0.3574 ≈ 1.4303 − 4 = −2.5697
0.0007
𝑥3 = 0.3574 − = 0.3574 + 0.0003 = 0.3577
−2.5697
• Iteration 4 (for accuracy):
𝑓 0.3577 ≈ 0.0000 very small
• 𝑥 ≈ 0.3577
Example 3:
Find the approximate value of the real root of the equation 𝑥 3 − 3 = 0 by the
Newton-Raphson method by taking x = 2. Carry out three iterations.
Solution: We need to solve
𝑥3 − 3 = 0
using the Newton–Raphson method, starting with 𝑥0 = 2 ,and perform three
iterations.
Step 1: Define the function
𝑓 𝑥 = 𝑥3 − 3
𝑓 ′ 𝑥 = 3𝑥 2
Newton–Raphson Formula
𝑓 𝑥𝑛
𝑥𝑛+1 = 𝑥𝑛 − ′
𝑓 𝑥𝑛
Example 3:
Step 2: Iterations
Iteration 1:
𝑥0 = 2
𝑓 2 = 8 − 3 = 5, 𝑓 ′ 2 = 12
5
𝑥1 = 2 − = 2 − 0.4167 = 1.5833
12
Iteration 2:
𝑓 1.5833 ≈ ൫1.5833)3 − 3 ≈ 3.97 − 3 = 0.97
𝑓 ′ 1.5833 ≈ 3൫1.5833)2 ≈ 7.52
0.97
𝑥2 = 1.5833 − ≈ 1.5833 − 0.1290 = 1.4543
7.52
Iteration 3:
𝑓 1.4543 ≈ ൫1.4543)3 − 3 ≈ 3.07 − 3 = 0.07
𝑓 ′ 1.4543 ≈ 3൫1.4543)2 ≈ 6.34
0.07
𝑥3 = 1.4543 − ≈ 1.4543 − 0.0110 = 1.4433
6.34
𝑥 ≈ 1.4433
Secant Method
• Geometric Idea:
• Similar to Newton method but instead of tangent:
• Use a secant line through two points
• Visualization:
• Take two points on curve: 𝑥𝑛−1 𝑓 𝑥𝑛−1 , 𝑥𝑛 𝑓 𝑥𝑛
• Draw a line joining them
• Find where it cuts the x-axis
• Geometrically:
You approximate the curve using a line through two previous points.
• Interpretation:
• Avoids derivative but still captures slope behavior
Fixed Point Iteration
• Geometric Idea:
• Rewrite equation as:
• 𝑥=𝑔 𝑥
• Visualization:
• Plot:
• 𝑦=𝑔 𝑥
• 𝑦 = 𝑥(straight line at 45°)
• Their intersection gives the root
• Iteration process:
• Start with 𝑥0
• Move vertically to curve 𝑦 = 𝑔 𝑥
• Then horizontally to line 𝑦 = 𝑥
• Repeat
• This creates a zig-zag path converging to intersection.
• Interpretation:
• Convergence depends on slope of 𝑔 𝑥
Summary (Geometric View)
Method Geometric Tool Used Idea
Bisection Interval shrinking Zoom into root
Regula-Falsi Chord (straight line) Linear interpolation
Newton-Raphson Tangent Best local approximation
Secant Secant line Approximate slope
Fixed Point Curve intersection Iterative mapping
Comparison of Methods
Method Type Speed Reliability
Bisection Bracketing Slow Very high
Regula-Falsi Bracketing Medium High
Newton-Raphson Open Very fast Medium
Secant Open Fast Medium
Applications
• Solving engineering equations
• Electrical circuits
• Fluid mechanics
• Finding eigenvalues
• Optimization problems
Conclusion
• Numerical methods give approximate roots
• Choice depends on:
• Speed required
• Availability of derivative
• Nature of function
Bisection = safest, Newton = fastest
Computational Tools
• Modern numerical methods are implemented using programming
languages and platforms such as Python, MATLAB, R, and Julia. These
tools allow visualization, automation, and efficient computation of
complex problems, making numerical methods accessible even for
large-scale systems.