Foundations — Applications — approximation, evaluating limits

This page assumes nothing. Every symbol the parent Applications note uses is unpacked here, one at a time, each resting on the one before. If a symbol below still feels shaky, that is exactly the symbol to slow down on.

1. A power $x^n$ — repeated multiplication

The picture. Think of $x^n$ as area/volume growing: $x^1$ is a line of length $x$, $x^2$ a square, $x^3$ a cube. For a small $x$ (say $x=0.1$), each higher power is a shrinking sliver: $0.1,0.01,0.001,\dots$ — every step is ten times tinier.

Why the topic needs it. A Taylor series is built entirely out of powers of $(x-a)$. The whole trick — "far-away terms are negligible" — only works because for small inputs, higher powers vanish fast. Look at the figure: by $x^4$ the bar is invisible.

2. Factorial $n!$ — the shrinking denominator

The picture. Factorials explode upward — but in a Taylor series they sit underneath as denominators, so they crush each term downward. Powers grow, factorials grow faster, so the ratio $\frac{x^n}{n!}$ collapses toward zero.

Why the topic needs it. Every coefficient in the four standard series carries a $\frac{1}{n!}$. That is why Taylor series converge so beautifully: the factorial denominator eventually beats any fixed power in the numerator. It is the engine behind "the leftover error is tiny."

3. The summation sign $\sum$ — a compact "add up" instruction

The picture. Imagine a conveyor belt. $n$ is the item counter; for each $n$ you drop the term $a_n$ into a growing pile, and $\sum$ is the total in the pile. When the top is $\infty$, the belt never stops — that is an infinite series.

Why the topic needs it. The parent's master formula $f(x)={\sum }_{n=0}^{\infty }\frac{f^{(n)}(a)}{n!}(x-a{)}^n$ is unreadable until $\sum$ is a friend. It's shorthand for "add up infinitely many polynomial pieces."

4. Function and its value at a point — $f(x)$, $f(a)$

The picture. $f(a)$ is the height of the curve directly above the point $a$ on the horizontal axis. That height is the very first ingredient of the polynomial twin: before matching slope or curve, we match height.

Why the topic needs it. Taylor series are built near a chosen point $a$ (the "centre"). Everything is measured by how far you drift, $(x-a)$, from that home base. When $a=0$ the series gets the special name Maclaurin.

5. The derivative $f^{\prime }(x)$ and higher derivatives $f^{(n)}(x)$

The picture. Three readings of the curve at the point $a$:

• $f(a)$ = height (where you are),
• $f^{\prime }(a)$ = tilt (which way and how steeply you're heading),
• $f^{\prime \prime }(a)$ = bend (how sharply the road curves).

Why the topic needs it. The coefficient $c_n=\frac{f^{(n)}(a)}{n!}$ is literally built from these derivatives. Each derivative supplies one more piece of information — height, then tilt, then bend, then finer wiggles — so the polynomial matches the true function ever more closely. That is the entire logic of why Taylor's formula is the "best" polynomial fit.

6. Approximately equal $\approx$, absolute value $|\cdot |$, less-than $<$, $\le$

The picture. $|x-a|$ is the distance between the input $x$ and the centre $a$ on the number line — no negatives allowed for a distance. In an error bound like $|R_N(x)|\le \frac{M}{(N+1)!}|x-a{|}^{N+1}$, the left side reads "the size of the leftover error" and the right side is "our guaranteed ceiling on it."

Why the topic needs it. Approximation is worthless without honesty about how wrong we might be. $\le$ turns a guess into a guarantee; $|\cdot |$ makes "how far off" a clean, sign-free quantity.

7. The remainder $R_N$ and the bound $M$

The picture. $P_N$ is the visible polynomial curve; $R_N$ is the tiny vertical gap between it and the true curve. As $N$ grows, that gap shrinks. $M$ is a "worst-case guess" that lets us wall the gap in without knowing it exactly.

Why the topic needs it. This is the difference between "roughly $0.0998$" and "$0.0998334$, correct to 6 decimals — provably." The remainder machinery is the trust in the whole method. (Explored fully in Lagrange Remainder & Error Estimation.)

8. The indeterminate form $\frac{0}{0}$

The picture. Two runners both sprinting toward zero. Whoever shrinks slower dominates the ratio. Series expansion is our slow-motion camera: it reveals the leading power of each runner, and the ratio of those leading powers is the answer.

Why the topic needs it. Half the parent note is about killing these limits by substituting series and cancelling the lowest power. Without recognising $\frac{0}{0}$ as "a race, not a number," the trick has no motivation. (L'Hôpital's Rule is the alternative decider.)

Prerequisite map

Read it top-down: raw symbols (powers, factorials, derivatives) fuse into the coefficient, the summation sign strings coefficients into the series, and the series then branches into the topic's two jobs — approximating (guarded by the remainder) and killing limits (triggered by $\frac{0}{0}$).

Equipment checklist

Test yourself — reveal only after answering.

Connections

• Parent: Applications
• Taylor & Maclaurin Series
• Lagrange Remainder & Error Estimation
• Radius & Interval of Convergence
• Alternating Series Test
• L'Hôpital's Rule
• Power Series — Operations (add, multiply, differentiate)