Numerical Analysis: Bridging Theoretical Calculus and Numerical Computation

Numerical analysis serves as the rigorous bridge between the infinite precision of theoretical calculus and the finite, discrete constraints of computer hardware. This slide establishes the foundational definitions of limits, continuity, and differentiability to show that while calculus provides the "exact" analytical destination, numerical computation provides the "approximate" path there, constrained by the tolerances ($\varepsilon$) and intervals ($\delta$) defined in classical real analysis.

1. The Foundation: Limits and Sequential Approximation

We move from the theoretical abstraction of limits to the computational reality that a processor cannot approach zero; it can only approach a machine epsilon.

Definition 1.1: The Limit

A function $f$ defined on a set $X$ has the limit $L$ at $x_0$, written $\lim_{x \to x_0} f(x) = L$, if, given any real number $\varepsilon > 0$, there exists a $\delta > 0$ such that $|f(x) - L| < \varepsilon$, whenever $x \in X$ and $0 < |x - x_0| < \delta$.

Definition 1.3: Sequence Convergence

A sequence $\{x_n\}_{n=1}^{\infty}$ has the limit $x$ if, for any $\epsilon > 0$, there exists a positive integer $N(\epsilon)$ such that $|x_n - x| < \epsilon$ whenever $n > N(\epsilon)$. This justifies our iterative algorithms.

2. Continuity and Differentiability: Safety Requirements

In numerical software, Continuity (Definition 1.2) and Differentiability (Definition 1.5) are not just academic properties; they are "safety requirements" for numerical stability. Theorem 1.6 proves that if a function is differentiable at $x_0$, it is continuous at $x_0$, ensuring that small measurement errors do not result in catastrophic output jumps.

🎯 Real-World Case: The Ideal Gas Law

Consider $PV = nRT$. In theoretical calculus, we assume variables are exact. In numerical computation, we acknowledge $P$ and $V$ are limits of measured sequences.

$T = \frac{PV}{nR} = \frac{(1.00)(0.100)}{(0.00420)(0.08206)} = 290.15 \text{ K} = 17^\circ\text{C}$

QUESTION 1

In calculus, we learn that $e = \lim_{h \to 0} (1 + h)^{1/h}$. What is the approximation of $e$ when $h = 0.04$?

2.6658

2.7183

2.7048

2.5000

QUESTION 2

Which theorem establishes that differentiability is a stronger condition than continuity?

Theorem 1.6

Theorem 1.4

Definition 1.1

Rolle's Theorem

QUESTION 3

What happens to the relative error when subtracting two nearly equal numbers in machine arithmetic?

It remains constant

It can increase dramatically due to significant digit cancellation

It always decreases

It becomes zero

QUESTION 4

Using four-digit rounding arithmetic, what is the result of $0.54617 - 0.54601$?

0.0002

0.00016

0.0001

0.00022

QUESTION 5

According to Theorem 1.4, $f$ is continuous at $x_0$ if and only if for any sequence $\{x_n\}$ converging to $x_0$:

The sequence $f(x_n)$ converges to $f(x_0)$

The sequence $f(x_n)$ is strictly increasing

The sequence $x_n$ is constant

The derivative $f'(x_n)$ is zero