CHAPTER ONE
INTRODUCTION TO NUMERICAL ANALYSIS
1.1 Overview
Numerical analysis is the branch of mathematics concerned with the design and analysis of
algorithms
for approximating solutions to continuous mathematical problems. This chapter introduces the
foundations
of numerical methods, their necessity, limitations, and examples illustrating their application.
1.2 Why Numerical Methods?
Many mathematical problems do not possess closed-form analytical solutions. Even when they do,
analytical solutions may be too complex for practical use. Numerical methods provide a way to
compute
approximate solutions with controllable accuracy.
Examples include:
- Solving nonlinear equations such as f(x) = cos(x) - x.
- Solving systems of linear equations arising from engineering applications.
- Approximating integrals where classical calculus techniques fail.
1.3 Sources of Error
Errors in numerical computation arise from:
1. **Modeling error** – difference between physical problem and mathematical model.
2. **Truncation error** – error from approximating infinite processes by finite ones.
3. **Round-off error** – due to finite precision in computers.
Example:
Approximating e^x with finite series:
e^x = 1 + x + x²/2! + ... + x■/n! + R■.
1.4 Types of Numerical Methods
Numerical methods are broadly classified as:
- Direct methods (e.g., Gaussian elimination)
- Iterative methods (e.g., Jacobi, Newton-Raphson)
Example (Newton-Raphson):
x_{n+1} = x_n - f(x_n)/f'(x_n).
Applied to f(x) = x² - 2:
Starting with x■ = 1.5, iterations converge to √2.
1.5 Numerical Stability and Convergence
A numerical method should be stable (small errors do not grow uncontrollably) and convergent
(iterations approach true solution).
Example of unstable process:
Subtracting nearly equal numbers:
x = (1 - cos(0.001)) / (0.001) causes large round-off error.
1.6 Floating Point Representation
Real numbers are approximated using floating-point arithmetic:
x = ± (1.d■d■d■...) × 10^e.
Example:
0.000345 = 3.45 × 10■■.
Errors arise from normalization and rounding.
1.7 Solving Nonlinear Equations
Common methods:
1. Bisection method – guaranteed convergence.
2. Newton-Raphson method – fast but requires derivative.
3. Secant method – derivative-free Newton method.
Example (Bisection):
Solve f(x) = x³ - x - 1 = 0 in [1, 2].
Midpoints converge to approximately 1.3247.
1.8 Interpolation
Interpolation constructs a function that matches known data points.
Example:
Given points (0,1), (1,3), (2,2):
Linear interpolation between (1,3) and (2,2):
y = 3 + (2 - 3)(x - 1) = 3 - (x - 1).
Polynomial interpolation:
Lagrange formula:
L(x) = Σ y_i L_i(x).
1.9 Numerical Differentiation
Derivative approximations like:
f'(x) ≈ (f(x+h) - f(x)) / h.
Example:
If f(x) = e^x, h = 0.01:
f'(1) ≈ (e^1.01 - e^1)/0.01.
1.10 Numerical Integration
Approximating integrals using:
- Trapezoidal rule
- Simpson’s rule
Example (Trapezoidal Rule):
∫■¹ x² dx ≈ (h/2)[f(0) + 2f(0.5) + f(1)]
= 1/6 ≈ 0.1667.
1.11 Initial Value Problems (ODEs)
Euler’s method:
y_{n+1} = y_n + h f(x_n, y_n).
Example:
dy/dx = x + y, y(0) = 1.
With h = 0.1:
y■ = 1 + 0.1(0 + 1) = 1.1.
1.12 Conclusion
This chapter has provided an extensive introduction to the concepts, errors, methods, and
examples
that form the foundation of numerical analysis. Subsequent chapters build on these principles with
advanced algorithms and real-world applications.