Coordinate geometry

A plane, an origin, and two numbers per point

A single point on a flat surface has no name of its own until two reference directions are agreed on and the point's distance along each one gets measured. Coordinate geometry is what happens once that agreement is in place: every point becomes a pair of numbers, and geometric questions — where something is, which way it sits relative to something else — turn into algebra you can compute rather than a drawing you have to eyeball.

The pair-of-numbers plane carries a name that reflects its history: it is often called the Cartesian plane, after Rene Descartes, who is credited with the underlying idea in his 1637 work La Geometrie. What Descartes actually wrote pinned a location down with two unknown line segments rather than the ready-made grid taught today, but trading a picture for two numbers is the same move this concept is built on.

A linear function's rule, f(x) = mx + b, already generates points one at a time from an input; coordinate geometry is what happens once those points sit on an actual plane together — plotting them, measuring between them, and reasoning about the lines and shapes several of them form at once. Two coordinates, read in a fixed order, mean exactly one point and no other.

That two-numbers-per-location trick does not stop at a flat plane — every later idea that pins something down with a list of numbers, whether a position in physical space or a location in some other kind of space entirely, is doing the same trick with a longer list.

The x-axis, the y-axis, and four quadrants

Two number lines, crossed at a right angle, are all it takes to build the plane. The horizontal one is the x-axis, the vertical one is the y-axis, and the single point where they cross — where both read zero — is the origin. Every other point on the plane gets located the same way: how far along the x-axis, then how far along the y-axis, written as an ordered pair (x, y).

The order matters. (x, y) and (y, x) are not the same point unless the two numbers happen to be equal, like a street address built from two independent parts, read in a fixed order, that together name one exact place and no other: two numbers, read in a fixed order, that together specify one location and no other.

The two axes also cut the plane into four regions called quadrants, numbered I through IV counterclockwise starting from the upper right. A point's quadrant is fixed entirely by the sign of its coordinates: both positive lands in quadrant I, a negative x with a positive y lands in quadrant II, both negative in quadrant III, and a positive x with a negative y in quadrant IV. Sit exactly on an axis and the point is not in any quadrant at all — its coordinates are just (x, 0) or (0, y).

Finding the point exactly between two points

Two points on the plane, (x1, y1) and (x2, y2), pin down a segment between them, and that segment has an exact midpoint with no ruler required. Average the two x-coordinates, average the two y-coordinates, and the pair that results, ((x1 + x2) / 2, (y1 + y2) / 2), is that midpoint.

The averaging is not a coincidence dressed up as a formula. It works like two people walking toward each other at the same steady pace and meeting exactly halfway between their starting points, because moving halfway along each coordinate separately lands on exactly the same spot as moving halfway along the segment itself.

The formula treats horizontal and vertical distance completely independently: the x-coordinates get averaged with no reference to the y-coordinates at all, and vice versa. That independence is what makes the midpoint of (2, 10) and (8, 2) a plain (5, 6) — average 2 and 8 to get 5, average 10 and 2 to get 6 — regardless of how far apart the two points sit or which direction the segment between them runs.

Testing whether points line up

Linear functions already hand you the tool for comparing two points' steepness: the slope between them. Coordinate geometry puts that same tool to a new job — deciding whether three points, taken together, sit on one straight line at all.

Three points, (x1, y1), (x2, y2), and (x3, y3), are collinear exactly when the slope from the first point to the second matches the slope from the second point to the third. Match, and there is no bend between them — one straight line passes through all three. Differ, even slightly, and the middle point sits off to one side, and no single line reaches all three.

That comparison works like a taut string stretched from the first point to the last: if the middle point doesn't nudge the string off its path, all three sit on one straight line: the middle point either falls exactly on the straight path between the outer two, or it does not, and a matching pair of slopes is what confirms which case you are looking at.

One case is worth flagging separately: a vertical run between two of the points makes the slope formula divide by zero, which is undefined rather than equal to any number. Three points that share an x-coordinate are collinear by inspection — they sit on a single vertical line — and need no slope arithmetic at all to confirm it.

Checking whether a point lies on a line

A line's equation is buildable two ways already familiar from linear functions: point-slope form from a single point and a slope, or slope-intercept form once the intercept is known. Coordinate geometry puts that finished equation to a different job — checking whether some other point actually belongs to the line.

The check is substitution. Take a candidate point's x and y, plug them into the line's equation in place of the variables, and see whether the two sides come out equal. They do, and the point lies on the line; they don't, and the point misses it, no matter how close it looks on a rough sketch.

Take the line y = 2x - 1 and a candidate point (3, 5): substituting gives 2(3) - 1 = 5, which matches the point's y, so (3, 5) sits on the line. The point (3, 4) fails the same test — 2(3) - 1 = 5, not 4 — and misses the line entirely, even though it looks close on a sketch.

This substitution test is the same move that later confirms a graphed crossing point is real rather than an estimate, and the same move that checks whether a shape's corners actually sit where its sides say they should.

When two lines never meet, and when they meet at a right angle

Two lines on the same plane relate to each other in one of a few honest ways, and slope alone — already familiar as a single line's steepness — decides two of the most useful ones: whether the lines are parallel, and whether they are perpendicular.

Parallel lines share the exact same slope and never cross, no matter how far the plane extends in either direction; two lines with slopes of 3 stay the same vertical distance apart forever, whatever their intercepts happen to be. Different intercepts keep them from being the same line drawn twice, but equal slope is the entire test.

Perpendicular lines meet at a right angle, and their slopes carry a stricter relationship: each is the negative reciprocal of the other, so their product is always -1. A line with slope 2 meets a perpendicular line with slope -1/2 at a square corner; flip either sign, or invert only one of them, and the angle stops being square.

A horizontal line, with slope 0, is perpendicular only to a vertical line, whose slope is undefined rather than any number — the negative-reciprocal rule breaks down exactly there, and a vertical-horizontal pair gets checked by sight instead of arithmetic.

Where two lines cross

Two lines that are not parallel cross exactly once, and that single crossing point is the one (x, y) pair that lies on both lines at the same time. Reading it off a graph is a matter of finding where the two drawings overlap and reading the coordinates at that spot.

A crossing read off a graph is only a candidate until it is checked: substitute the point's coordinates into both lines' equations and confirm both come out true, the same test that already confirms whether any single point sits on any single line. A candidate that satisfies one equation but not the other was misread off the picture, not an actual crossing.

Two distinct lines with equal slope run parallel and share no crossing at all. Two lines with equal slope and equal intercept are not two lines but one, drawn twice, and every point on it counts as shared rather than none.

A crossing point is the plane's way of answering a question with two conditions at once — wherever a rule governing one line meets a rule governing another, the crossing is the single spot that honors both.

Describing shapes by their coordinates

A shape drawn on the plane is nothing more than its vertices, listed as coordinate pairs, plus the understanding that straight sides connect them in order. A triangle needs three vertices, a rectangle four, and the shape is completely pinned down the moment the coordinates are — no separate drawing required to know what it looks like.

Comparing two sides' slopes reuses the same tool that decides whether two whole lines are parallel: a rectangle's opposite sides carry equal slopes, and testing that equality directly from the vertices confirms the shape actually is a rectangle rather than a slightly skewed quadrilateral that only looks like one.

The midpoint formula answers a different, equally practical question about the same shape: where a side's exact center sits, useful for anything that needs to balance on it or split it evenly, computed straight from that side's two endpoint coordinates with no extra measuring involved.

The same habit — pinning a location down with a short list of numbers, then reasoning about it with arithmetic instead of a ruler — is exactly what a machine does when it tracks its own position on a map or represents an idea as a point in some other kind of space; a plane and two coordinates are simply the smallest version of that idea.