Try to write a circle as .
Go ahead. Pick a formula. You'll get stuck almost immediately, because for most values of a circle has two values, one on top, one on the bottom. A single function can only ever spit out one per . A circle just doesn't fit that mold.
So how does a compass draw one? How does a GPS unit trace your walking route, loops and all? How does a video game move a ball along a curved arc?
They all cheat the same way. They stop asking "what's in terms of " and start asking "where is the point at time ." That's the whole idea behind a parametric equation, and once you see it, a lot of curves that used to feel impossible become almost easy.
What does a parametric equation actually mean?
Picture a bug crawling across a table, dragging an inked leg behind it. Every second, you jot down two numbers: how far east it's moved, and how far north.
That's it. That's the whole idea. You're not writing one rule that connects east and north directly. You're writing two separate rules, one for east, one for north, and both of them are driven by the same clock.
We call that clock (for time, though it doesn't have to be time at all, just a shared input). We write the two rules as and . Watch what happens when the bug walks steadily east while bobbing up and down like it's riding a wave:
One clock, two rules. x(t) crawls east at a steady pace while y(t) bobs up and down, watch how the two strips below combine into the path above.
Scrub or hit play. Notice the two strips underneath, they're each showing one rule in isolation. climbs in a straight line because the bug moves east at a constant rate. waves up and down because the bug's north-south position oscillates on its own schedule. Neither strip knows about the other. Combine them at every instant and you get the curved path up top.
That's a parametric equation: two independent functions of the same input, traced out together.
Why can this describe shapes a normal function can't?
Here's the payoff. A regular function has to pass the vertical line test, one per , no exceptions. That rules out circles, spirals, and loops before you even start.
A parametric curve has no such rule. and are both free to wander however tells them to, including doubling back on themselves. That's exactly how you get a circle.
The classic example, the one you'll see everywhere once you know it, is this pair:
Drag the point around the circle. x and y each swing between -1 and 1 on their own, but x squared plus y squared never budges from 1.
Drag the point around. and each swing back and forth between and , completely independently, and yet the point never strays from the circle. Watch the readout in the corner: stays locked at no matter where you drag. That's not a coincidence, it's the Pythagorean identity, , showing up as a picture instead of a proof.
One equation, and it draws a shape that "y = f(x)" simply can't touch.
Does changing the parameter change the shape?
Now for the interesting part. What if we speed up the clock?
Try replacing with in the same circle equations, , . Does the picture change?
Same circle, same path, different clocks. Slide k and watch the orange dot lap the teal one, the shape never changes, only the timing does.
Slide up and down. The path stays exactly the same circle. What changes is how fast the point travels around it. At , the orange dot laps the circle twice as often as the teal one. At , it crawls at half speed. Check the lap counters, the ratio always matches exactly.
That's the part a lot of people miss: a parametric equation doesn't just encode a shape, it encodes a journey along that shape, including speed and direction. Two totally different parametrizations can trace the identical curve while disagreeing completely about timing. The shape is only half the story.
How do we get back to a familiar equation?
So parametric equations are more powerful than . But sometimes you don't need that power, you just want the plain old equation back. How do you undo the parametrization?
You eliminate the parameter. Solve one equation for , then substitute it into the other. Take the simplest case, and . Since already equals , just swap for in the second equation:
...and there's your familiar parabola. For the circle, you don't even solve for , you just square both equations and add them, and the cancels out via that same identity from before:
Toggle the example. The dashed line is drawn straight from the Cartesian formula, the solid line is drawn by sweeping t, they sit exactly on top of each other.
Substitute t = x directly into y = t², since x = t.
Toggle between the two examples. The dashed curve is drawn straight from the plain Cartesian formula, no involved at all. The solid curve is drawn by sweeping through its whole range. They land on the exact same pixels, because eliminating the parameter doesn't change the shape, it just changes which variable is doing the describing.
The short version
A parametric equation describes a curve using two separate functions, and , both driven by the same input instead of one variable depending directly on the other. That extra freedom is exactly what lets it draw circles, spirals, and loops that a plain equation can never represent.
The parameter carries more information than the shape alone, it also encodes speed and direction, which is why the same curve can come from wildly different-looking equations. And whenever you do want the plain equation back, eliminating the parameter, solving for in one equation and substituting into the other, gets you there without changing the picture at all.
Next time you see a circle, a spiral, or anything that loops back on itself, you'll know why it needed two functions and a clock instead of one.
All visualizations are interactive React components running entirely in your browser. Every curve is recomputed live from its parametric formula as you drag, scrub, or slide, nothing is a precomputed image. No libraries beyond React.