Coordinate Geometry — Lines, Distances, Midpoints

Coordinate geometry translates geometric relationships into algebra using the Cartesian plane. The distance between two points is

(x_{2} - x_{1})^{2} + (y_{2} - y_{1})^{2}

, their midpoint is the average of their coordinates, and straight lines are described by linear equations whose slope encodes direction. Applies wherever position, alignment, and separation must be computed exactly — navigation, computer graphics, physics, engineering. Founded by René Descartes and Pierre de Fermat in the 1630s.

Coordinate geometry (also called analytic geometry) studies geometric objects — points, lines, segments — by assigning numerical coordinates to points in a plane and expressing geometric facts as algebraic equations. A point is an ordered pair $(x, y)$ giving its horizontal and vertical displacement from a fixed origin. This matters because it converts visual reasoning into computation: questions like "are these three points collinear?" or "what is the shortest distance between this point and this line?" become solvable by formula rather than construction. The surprising fact is that the entire apparatus rests on a single tool — the Pythagorean theorem — repurposed as a distance formula, so that all of plane geometry becomes a branch of algebra.

The Core Insight

Once you fix a coordinate system, every geometric relationship becomes an algebraic equation, and every algebraic equation describes a geometric shape. Distance is just the hypotenuse of a right triangle built from coordinate differences; a line's "steepness" is a single number (the slope); and parallelism, perpendicularity, and collinearity all reduce to simple arithmetic conditions on slopes and coordinates.

Formal Statement

The Cartesian Plane and a Point

The Cartesian plane is the set of all ordered pairs of real numbers $(x, y)$ . Two perpendicular number lines — the $x$ -axis (horizontal) and $y$ -axis (vertical) — meet at the origin $O = (0, 0)$ . For a point $P = (x_{1}, y_{1})$ :

$x_{1}$ is the abscissa (signed horizontal distance from the $y$ -axis),
$y_{1}$ is the ordinate (signed vertical distance from the $x$ -axis).

The plane is divided into four quadrants, numbered I–IV counterclockwise from the upper right.

Distance Between Two Points

For points $P_{1} = (x_{1}, y_{1})$ and $P_{2} = (x_{2}, y_{2})$ , the distance $d$ between them is

d = (x_{2} - x_{1})^{2} + (y_{2} - y_{1})^{2}

⟳form a right triangle with legs

∣ x_{2} - x_{1} ∣

and

∣ y_{2} - y_{1} ∣

; apply the Pythagorean theorem to the hypotenuse

\overline{P_{1} P_{2}}

Here $d \geq 0$ always, and $d = 0$ iff $P_{1} = P_{2}$ . The order of the points does not matter because the differences are squared.

Midpoint of a Segment

The midpoint $M$ of the segment joining $P_{1} = (x_{1}, y_{1})$ and $P_{2} = (x_{2}, y_{2})$ is

M = (\frac{x _{1} + x _{2}}{2}, \frac{y _{1} + y _{2}}{2})

⟳the midpoint's coordinates are the arithmetic means of the endpoints' coordinates; equivalently,

M = P_{1} + \frac{1}{2} (P_{2} - P_{1})

Slope of a Line

The slope (gradient) $m$ of the line through $P_{1} = (x_{1}, y_{1})$ and $P_{2} = (x_{2}, y_{2})$ , with $x_{1} \neq = x_{2}$ , is

m = \frac{y _{2} - y _{1}}{x _{2} - x _{1}} = \frac{Δ y}{Δ x}

⟳slope is "rise over run" — the ratio of vertical change to horizontal change between any two points on the line.

A vertical line ( $x_{1} = x_{2}$ ) has undefined slope. A horizontal line has slope $0$ .

Equations of a Line

A non-vertical line may be written in several equivalent forms:

y - y_{1} = m (x - x_{1}) (point-slope form)

⟳rearrange the slope definition

m = \frac{y - y _{1}}{x - x _{1}}

for a general point

(x, y)

on the line.

y = m x + b (slope-intercept form)

⟳set

x = 0

in [4] to identify

b

as the

y

-intercept; expand and collect terms.

A x + B y + C = 0 (general form, A, B not both 0)

⟳clear fractions and move all terms to one side; this form also handles vertical lines where

B = 0

Parallel and Perpendicular Lines

Let two non-vertical lines have slopes $m_{1}$ and $m_{2}$ .

They are parallel iff $m_{1} = m_{2}$ (and they are distinct lines).
They are perpendicular iff

m_{1} \cdot m_{2} = - 1 ⟺ m_{2} = - \frac{1}{m _{1}}

⟳rotating a direction vector

(1, m_{1})

9 0^{\circ}

gives

(- m_{1}, 1)

, whose slope is

- 1/ m_{1}

; the dot product of perpendicular direction vectors is zero.

Distance from a Point to a Line

The perpendicular distance from a point $P_{0} = (x_{0}, y_{0})$ to the line $A x + B y + C = 0$ is

d = \frac{∣ A x _{0} + B y _{0} + C ∣}{A ^{2} + B ^{2}}

⟳project the vector from any point on the line to

P_{0}

onto the unit normal

\frac{( A , B )}{A ^{2} + B ^{2}}

; the absolute value of that projection is the distance.

Why It Holds

The Distance Formula from the Pythagorean Theorem

Take points $P_{1} = (x_{1}, y_{1})$ and $P_{2} = (x_{2}, y_{2})$ .
Construct a third point $Q = (x_{2}, y_{1})$ — directly below or above $P_{2}$ and level with $P_{1}$ . This makes $△ P_{1} Q P_{2}$ a right triangle with the right angle at $Q$ .
The horizontal leg $\overline{P_{1} Q}$ has length $∣ x_{2} - x_{1} ∣$ (same $y$ , differing $x$ ).
The vertical leg $\overline{Q P_{2}}$ has length $∣ y_{2} - y_{1} ∣$ (same $x$ , differing $y$ ).
By the Pythagorean theorem, the hypotenuse satisfies $d^{2} = ∣ x_{2} - x_{1} ∣^{2} + ∣ y_{2} - y_{1} ∣^{2}$ .
Since $∣ a ∣^{2} = a^{2}$ for any real $a$ , this is $d^{2} = (x_{2} - x_{1})^{2} + (y_{2} - y_{1})^{2}$ .
Taking the non-negative square root gives equation [1]. $■$

The Midpoint Formula

Let $M = (\overset{x}{ˉ}, \overset{y}{ˉ})$ be the midpoint of $\overline{P_{1} P_{2}}$ .
Because $M$ bisects the segment, the horizontal displacement from $P_{1}$ to $M$ equals that from $M$ to $P_{2}$ : $\overset{x}{ˉ} - x_{1} = x_{2} - \overset{x}{ˉ}$ .
Solving: $2 \overset{x}{ˉ} = x_{1} + x_{2}$ , so $\overset{x}{ˉ} = \frac{x _{1} + x _{2}}{2}$ .
The identical argument on vertical displacements gives $\overset{y}{ˉ} = \frac{y _{1} + y _{2}}{2}$ .
Hence $M = (\frac{x _{1} + x _{2}}{2}, \frac{y _{1} + y _{2}}{2})$ , which is equation [2]. $■$

René Descartes | 1596–1650

The French philosopher and mathematician René Descartes published La Géométrie in 1637 as an appendix to his Discours de la méthode. There he showed that geometric curves could be described by algebraic equations in two variables, unifying two fields the Greeks had kept apart. The modern "Cartesian" coordinate system bears his Latinized name (Cartesius). Working independently and slightly earlier, Pierre de Fermat developed equivalent ideas in an unpublished manuscript, Ad locos planos et solidos isagoge. Descartes' notation — using letters near the end of the alphabet ( $x, y, z$ ) for unknowns — survives essentially unchanged today.

Building Intuition

Think of the plane as a city grid. Each point is an address: walk $x$ blocks east and $y$ blocks north from the central plaza (the origin).

The distance formula answers "how far apart are two addresses as the crow flies?" — not along the streets, but in a straight diagonal, which is why a right triangle and Pythagoras appear.
The midpoint is the meeting place exactly halfway between two friends: average where each one starts.
The slope is the grade of a hill: rise 3 metres over every 4 metres you walk forward and the slope is $3/4$ . A bigger slope means a steeper climb; a negative slope means you're descending.
Perpendicular slopes multiplying to $- 1$ feels strange until you picture turning a corner: if a road climbs steeply, the road crossing it at a right angle must descend gently in the opposite sense — steep and shallow trade places, and the sign flips.

A segment from $A(1,2)$ to $B(7,6)$ with its midpoint $M$, slope right triangle, and a perpendicular line through $M$.

Worked Examples

Distance, Midpoint, and Slope of a Segment

Given:

A = (1, 2)

and

B = (7, 6)

(coordinates in metres). Find: the length of

\overline{A B}

, its midpoint, and its slope. Step 1 — Distance, using [1]:

d = (7 - 1)^{2} + (6 - 2)^{2} = 6^{2} + 4^{2} = 36 + 16 = 52 = 213 .

Numerically

213 \approx 7.21

m. Step 2 — Midpoint, using [2]:

M = (\frac{1 + 7}{2}, \frac{2 + 6}{2}) = (4, 4) .

Step 3 — Slope, using [3]:

m = \frac{6 - 2}{7 - 1} = \frac{4}{6} = \frac{2}{3} .

d = 213 \approx 7.21 m, M = (4, 4), m = \frac{2}{3}

Equation of a Line and Distance from a Point

Given: the line

ℓ

passing through

A = (1, 2)

with slope

m = \frac{2}{3}

, and the external point

P_{0} = (6, 1)

. Find: the equation of

ℓ

in general form, and the perpendicular distance from

P_{0}

ℓ

. Step 1 — Point–slope form, using [4]:

y - 2 = \frac{2}{3} (x - 1) .

Step 2 — Convert to general form [6]. Multiply by 3:

3 y - 6 = 2 (x - 1) = 2 x - 2 ⟹ 2 x - 3 y + 4 = 0.

A = 2, B = - 3, C = 4

. Step 3 — Distance from $P_{0} = (6, 1)$ , using [8]:

d = \frac{∣2 ( 6 ) - 3 ( 1 ) + 4∣}{2 ^{2} + ( - 3 ) ^{2}} = \frac{∣12 - 3 + 4∣}{13} = \frac{13}{13} = 13 .

Numerically

13 \approx 3.61

units.

ℓ : 2 x - 3 y + 4 = 0, d = 13 \approx 3.61 units

Common Misconceptions

Sign and order errors are the most common mistakes.

In the distance formula, the differences are squared, so order does not matter — but students sometimes forget to square, or take $∣ x_{2} - x_{1} ∣ \cdot ∣ y_{2} - y_{1} ∣$ instead of the sum of squares.
The midpoint uses addition then division by 2, not subtraction. $(\frac{x _{1} - x _{2}}{2}, \dots)$ is wrong.
For slope, subtract the coordinates in the same order in numerator and denominator: $\frac{y _{2} - y _{1}}{x _{2} - x _{1}}$ , never $\frac{y _{2} - y _{1}}{x _{1} - x _{2}}$ .
A vertical line has undefined slope, not "infinite" or "zero" slope. Treat it separately as $x = constant$ .
In equation [8], the numerator must use $A x_{0} + B y_{0} + C$ with the line written as $= 0$ . If you start from $y = m x + b$ , rewrite it first.

Claim (false): "If two lines have slopes whose product is negative, they are perpendicular."

Consider $ℓ_{1}$ with slope $m_{1} = 2$ and $ℓ_{2}$ with slope $m_{2} = - 1$ . Their product is $2 \cdot (- 1) = - 2$ , which is negative. But for perpendicularity we need the product to equal exactly $- 1$ , by equation [7]. Here $- 2 \neq = - 1$ , so the lines meet at an angle that is not $9 0^{\circ}$ . (The actual angle $θ$ satisfies $tan θ = \frac{m _{1} - m _{2}}{1 + m _{1} m _{2}} = \frac{3}{- 1} = 3$ , giving $θ \approx 71. 6^{\circ}$ .) A negative product alone is not enough — the precise value $- 1$ is required.

Connections

Coordinate geometry is the bridge linking pure synthetic geometry to algebra and analysis.

It rests on the Pythagorean Theorem, which supplies the distance formula directly.
The slope condition for perpendicularity is the two-dimensional case of the Dot Product of vectors, where perpendicular vectors have zero dot product.
Lines extend naturally to Conic Sections — circles, parabolas, ellipses, hyperbolas — each a quadratic equation in $x$ and $y$ , with the circle $(x - h)^{2} + (y - k)^{2} = r^{2}$ being the distance formula set equal to a constant.
Slope is the discrete forerunner of the Derivative: the derivative is the limiting slope of a curve at a point.
The whole framework generalises to Vectors in the Plane and to three dimensions, where a third coordinate $z$ and an extended distance formula appear.

Real-World Applications

GPS and navigation: receivers compute distances between coordinate positions to triangulate location; route planners use midpoints and straight-line distances as estimates.
Computer graphics & game engines: every sprite, polygon vertex, and camera ray is a coordinate; collision detection uses point-to-line and point-to-point distance formulas constantly.
Surveying and civil engineering (CAD): land plots, road gradients (literally "slope" as percent grade), and property boundaries are stored as coordinates and analysed with these formulas.
Robotics & path planning: a robot computes the perpendicular distance from its position to a planned line of travel to stay on course.
Physics — projectile and kinematics: trajectories are plotted as $(x, y)$ pairs; displacement is a distance-formula computation between start and end points.
Machine learning: the Euclidean distance [1] underlies $k$ -nearest-neighbours classification and clustering algorithms; the point-to-line distance [8] appears in support vector machines (the "margin").
Air-traffic control & maritime routing: closest-approach calculations between moving objects reduce to minimising a distance function between parametrised lines.

Historical Development

c. 300 BCE — Euclid's Elements systematises synthetic geometry without coordinates; ratios and constructions dominate.
c. 200 BCE — Apollonius of Perga studies conic sections using a precursor to coordinates ("diameters" and "ordinates") but no algebra.
14th c. — Nicole Oresme graphs quantities against time, an early use of coordinate-like representation of varying magnitudes.
1629–1636 — Pierre de Fermat writes Ad locos planos et solidos isagoge, deriving equations of lines and conics, anticipating analytic geometry (published posthumously, 1679).
1637 — René Descartes publishes La Géométrie, formally fusing algebra and geometry; the coordinate method spreads across Europe.
1684–1690s — Newton and Leibniz develop calculus, treating slope as the instantaneous rate of change, building directly on coordinate geometry.
1748 — Leonhard Euler's Introductio in analysin infinitorum standardises the modern treatment of functions and curves in the coordinate plane.
19th c. — Generalisation to $n$ -dimensional and non-Euclidean spaces (Riemann, Cayley) shows coordinate geometry is one instance of a far broader framework.
20th–21st c. — Digital computers make coordinate geometry the operational language of graphics, CAD, GIS, and scientific computing.

Subtleties & Limits

The formulas above assume a flat Euclidean plane with the standard ("Pythagorean") notion of distance. Several deep questions and limitations arise when these assumptions are relaxed:

On the surface of a sphere (e.g. the Earth), the straight-line distance formula gives the chord through the planet, not the travel distance along the surface; the correct measure is the great-circle (geodesic) distance, which uses spherical trigonometry, not [1].
Replacing the Euclidean metric with the taxicab metric $d = ∣ x_{2} - x_{1} ∣ + ∣ y_{2} - y_{1} ∣$ produces a self-consistent geometry where "circles" are diamonds and midpoints are no longer unique — illustrating that distance is a choice of axioms, not an absolute.
Determining the geometry of physical space is ultimately an empirical and unresolved matter: general relativity describes spacetime as curved, so the flat coordinate geometry of high school is an excellent local approximation, not the final word on real-world distance.
Even within the plane, generalising "the line minimising total distance to many points" leads to optimisation problems (the Fermat point, geometric median) that have no closed-form solution in general.