This page is a toolbox. We open every drawer, name every tool in plain words, draw the picture it stands for, and say why the parent topic the path-dependence note cannot work without it. Read top to bottom — each item is built only from the ones above it.
Picture: the burnt-orange dot sits at (x,y). Walk x units along the horizontal axis, then y units up. One pair = one location.
Why the topic needs it: the whole subject asks what happens to a machine's output as the input point(x,y) slides toward a special target point (a,b). Without a way to name points, we can't even talk about "sliding toward."
Picture: think of f as the height of a landscape above the flat sheet. Every ground position (x,y) has a terrain height f(x,y) floating above it.
Why the topic needs it: the "limit" question is "as I walk toward the target on the ground, does the terrain height settle to one number?" No two-input function, no question.
This is the single most important symbol in the parent note, so we build it slowly.
How the formula is born — the right triangle. Go from (a,b) to (x,y) in two steps: first move horizontally a gap of (x−a), then vertically a gap of (y−b). These two moves are the two legs of a right triangle; the straight arrow between the points is its hypotenuse.
Pythagoras says (leg)² + (leg)² = (hypotenuse)². So
distance2=(x−a)2+(y−b)2⟹distance=(x−a)2+(y−b)2.
We square each gap so that a negative gap (moving left or down) counts the same as a positive one — length can't be negative.
We take the square root at the end to undo the squaring and return to honest length units.
Why the topic needs it: the definition of a limit says the input must be within some small radius of the target. This formula is that "within distance." (We name that radius in item 7.)
Why the topic needs it:∣f(x,y)−L∣ measures how far the outputf(x,y) is from the candidate answer L, regardless of whether it overshoots or undershoots. It is the output-side twin of the input-side distance in item 4.
Why the topic needs it: paths can only ever disprove a limit. To prove one exists you must win the ε–δ game on the whole disk. These two symbols are the referee's rulebook.
The line through(a,b) with slope m is y=b+m(x−a). Different m picks a different straight road into the target.
Why the topic needs it: the cheapest paths to test are straight lines. Sweeping m over all values tests infinitely many directions at once — and if the answer still depends on m, the limit is already dead (Worked Example 1 in the parent note).
Picture: the deep-teal arrow has length r and swings through angle θ. The tip's shadow on the horizontal axis is rcosθ (that's x); its height is rsinθ (that's y).
Test yourself — you should be able to answer each before reading the parent note.
What do the two numbers in (x,y) tell you?
How far right/left (x) and how far up/down (y) — one exact spot on the plane.
What does the arrow in (x,y)→(a,b) mean, and why never "="?
The wandering point creeps toward the fixed target without landing on it, because f may be undefined right at the target.
Where does (x−a)2+(y−b)2 come from?
Pythagoras on the right triangle whose legs are the horizontal gap (x−a) and vertical gap (y−b); it's the length of the straight arrow between the points.
Why square the gaps and then take a root?
Squaring kills the sign so left/down count like right/up; the root returns to honest length units.
What is L in a limit statement?
The candidate landing value — the single number we hope the output settles to, tested via ∣f−L∣.
What is an open disk 0<(x−a)2+(y−b)2<ρ?
All points nearer than ρ to the target, excluding the rim and the centre itself.
Why is a disk (not an interval) the source of all the trouble?
A disk has infinitely many approach directions, so a single value L must work no matter how you sneak in.
Which of ε,δ controls the output, which the input?
ε = output tolerance around L; δ = input radius around (a,b).
What is the line through (a,b) with slope m?
y=b+m(x−a); sweeping m tests all straight directions at once.
Translate x=rcosθ,y=rsinθ: what do r and θ measure?
r = distance from origin (how close), θ = direction (from which way).
What does leftover θ as r→0 signal?
Direction-dependence, i.e. the limit does not exist.
What tool proves a limit exists, and why can't paths do it?
A shrinking bound ∣f−L∣≤g(r)→0 via the Squeeze Theorem; paths test only some routes, a bound covers the whole disk.
What does the degree of a term count, and why care?
Sum of the powers of x and y; equal degrees cancel size and leave direction (danger), higher top degree forces 0.