The last idea of the course: let both coordinates depend on a third variable, so a curve becomes the path of a moving point. Covers sketching from a parametrization by reading each coordinate's behaviour separately, eliminating the parameter by substitution or by a trigonometric identity, the standard recipes for parametrizing graphs, segments and ellipses, and the two adjustments that reverse an orientation or delay a start. Closes with the cycloid, a curve that has no usable equation in x and y at all.
Subject: Trigonometry · 68 slides · symbolic lesson
Open the interactive version of this deck
Title
Trigonometry · Section 11.10
A curve as the path of a moving point
Objectives
Every curve in this course so far has been pinned down by one equation relating x and y. That leaves out the curves that are the graph of no function either way, and it leaves out the whole idea of travelling along a curve. Both gaps close when each coordinate gets its own function of a new variable.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, pp. 1048-1059 — the whole of the final section
This is the last section of the course, and it is the one that changes what a curve is allowed to be. Until now a curve has been the graph of an equation: you hand the equation a point and it tells you yes or no. That is a good description of a shape, and a poor description of a journey.
Here you give each coordinate its own function of a new variable, usually called t and usually imagined as time. The curve becomes the set of places a moving point visits, and it comes with a direction of travel built in. Two things fall out at once. Curves that fail both line tests, which no single function could ever produce, become easy. And the direction, which no equation in the two coordinates can express, becomes part of the answer.
The lesson has five working parts: what a parametrization is, how to sketch one, how to eliminate the parameter and what that costs, how to find a parametrization for a described curve, and how to adjust one whose direction or starting point is wrong. It ends with the cycloid, the curve that makes the case for the whole idea.
Warm-up
Discussion prompt
In Section 10.2 you modelled an object moving around a circle of radius r at a constant angular frequency. Write down its two coordinates as functions of time, without looking back, and say what the sign of the angular frequency controls.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1048 — the circular motion equations from p. 732 recalled
Write your answer down before revealing it. If the two equations came back to you quickly, you already know more about this section than the title suggests.
Circular motion was the first parametrization you met, back in Section 10.2. The radius fixes the size of the circle, the angular frequency fixes the speed, and the sign of the angular frequency fixes the direction. What you were not told at the time is that the pair of equations is an object with a name.
Notice which features of the motion live in which part of the description. The set of points visited is just a circle of a given radius; it does not care about the angular frequency at all. Everything the angular frequency controls, speed and direction, is invisible in the phrase a circle of radius r. Hold on to that split, because the rest of the lesson runs on it.
Section
Part 1
Concept
Imagine a bug crawling across a table top. It starts somewhere, wanders, loops back across its own trail, and heads off. The trail it leaves is a perfectly good curve, and no equation of the form y equals something in x can produce it.
Figure (svg): A looping path traced by a moving point: it starts high on the left, swings down to the left, loops back and crosses itself at the point (0, 1), then heads off to the upper right. A dashed vertical line at x = 2 cuts the path three times; a dashed horizontal line at y = minus 1 cuts it twice.
The dashed vertical line meets the trail three times, so the trail fails the Vertical Line Test. The dashed horizontal line meets it twice, so it fails the Horizontal Line Test as well. It is the graph of no function of x and of no function of y.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1048 — the bug crawling across a table top
Look at the two dashed lines before anything else. The red one is a vertical line test and it meets the trail three times, so there is no function of the first coordinate that produces this picture. The purple one is a horizontal line test and it meets the trail twice, so turning the picture on its side does not help either.
That is a real obstacle, not a technicality. Every graphing tool you have used so far takes one coordinate in and gives the other out, and this curve simply cannot be fed to one.
Now find the opening. At any single instant the bug is at exactly one place. So its position along the table, left to right, is a perfectly good function of time, and so is its position front to back. The trail crosses itself because the bug passes the point marked Q twice, at two different instants, which contradicts nothing at all. The failure of the line tests is a failure of the picture to be a graph; it is not a failure of the motion to be describable.
Concept
Here is the opening the picture leaves. The bug can be in only one place at any one instant, so its position is a function of time even though it is not a function of x. Record each coordinate separately.
\[ \begin{cases} x = f(t) \\ y = g(t) \end{cases} \]
parameter — The independent variable, traditionally t, that both coordinate functions are functions of. It is often thought of as time, but it need not be.
parametrization — A system of two such equations together with an interval for the parameter. It gives a curve as the set of points it produces, and an orientation as the order in which they arrive.
Neither f nor g has to be one to one. That is precisely how the trail can cross itself: the bug reaches the crossing point at two different instants.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1048 — the definition and footnote 2
Read the two equations as one object. They are a system, so they are always used together at the same value of the parameter: put in a value, get out two numbers, plot one point. Reading only one of the two equations tells you nothing about where the point is.
The word parameter matters. The new variable is not a third coordinate and it does not appear anywhere on the picture. It is a label attached to each point, saying when that point is visited, and when you draw the curve the labels are thrown away.
The sentence about one-to-one is the one to remember from this slide. If the trail crosses itself, then two different labels sit on the same point, so both coordinate functions repeat an output. Neither of them can be one to one. That is not a defect; it is exactly the freedom that makes this description stronger than an equation.
Notation
Annotate
On: \( \begin{cases} x = f(t) \\ y = g(t) \end{cases} \quad a \le t \le b \)
Work through the four annotations in order, and pay most attention to the last one. Students lose more marks on the interval than on anything else in this section.
The interval is not a footnote to the answer. Change it and you change which points are on the curve, where the journey starts, where it stops, and sometimes how many times each point is visited. A pair of equations without an interval is an incomplete answer, in the same way that an integral without limits is a different object from one with them.
The second annotation is worth a moment too. Because each coordinate is an ordinary function of one variable, every tool you already have applies to it: domain, range, increasing and decreasing, turning points, end behaviour. The next part of the lesson is nothing but the systematic use of those tools, one coordinate at a time.
Intuition
An equation relating x and y answers one question: is this point on the curve, yes or no. A parametrization answers a much richer question: where is the point at each moment. The second answer contains the first, and a good deal more besides.
That extra information is what makes the impossible possible. Asking a single function of x to produce two heights above one x is a contradiction. Asking a moving point to pass over the same x twice, at two different moments, is an ordinary Tuesday.
It is also the reason physics and computer graphics almost never use equations in x and y for a path. What a simulation or an animation actually needs is the position at each instant, which is exactly what the pair of functions gives.
It can feel like a cheat to solve a problem by adding a variable. Here is why it is not. An equation relating the two coordinates carries one bit of information per point: in, or out. A parametrization carries an address for every instant, and from those addresses you can recover the membership test, but not the other way round.
The second paragraph is the heart of it. Asking a single function to give two heights above the same input is a contradiction in terms. Asking a moving point to pass over the same input twice, at two different moments, is completely ordinary. Nothing was cheated; the question was asked in a form that has an answer.
If you go on to physics, graphics or engineering, this is the form you will actually meet. A simulation does not ask whether a point lies on a trajectory; it asks where the object is at this instant, and the parametrization answers directly. The equation in the two coordinates, when it exists at all, is usually the less useful description.
Concept
orientation — The direction of travel along the curve, that is, the direction of increasing parameter. It belongs to the parametrization, never to the curve.
Figure (svg): Two copies of the same sideways parabola. On the left, y = 2t minus 1, and the arrows run upward from (1, minus 5) to (6, 5). On the right, y = 1 minus 2t, and the arrows run downward from (1, 5) to (6, minus 5).
The curve itself is only a set of points, and a set of points is not going anywhere. Change the second equation and the same set of points is travelled the other way. Arrows on a sketch are how the orientation is recorded.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1048 — the curve is devoid of orientation
Compare the two panels carefully. Every single point on the left panel is also on the right panel, and vice versa. As sets they are identical, and any equation satisfied by one is satisfied by the other.
The arrows are the only difference, and they came from changing the second equation. That tells you where orientation lives: not in the curve, in the parametrization. A curve is a set of points, and a set does not go anywhere.
This is why a sketch of a parametric curve is incomplete without arrows on it. If a problem gives you a parametrization and you hand back a shape, you have thrown away half of what you were given. Put at least two arrows on every parametric sketch you draw from here on, and put them on while you still have the parameter in front of you.
Discrimination
Take the parametrization on the left of the last figure: the first coordinate is t squared minus three and the second is two t minus one, for t at least minus two.
Sort into buckets
Sort each statement by whether it is a fact about the set of points, or a fact the parametrization added.
Sort all six before checking. The test for each statement is simple: imagine re-parametrizing the same set of points some other way. If the statement could then become false, it belongs to the schedule.
The shape, the vertex and the line-test verdict all survive, because they are determined by which points are present and nothing else. Re-parametrize however you like and a rightward-opening parabola stays a rightward-opening parabola with the same vertex.
Direction, starting point, and the fact that a particular point is reached at a particular value of the parameter all fail that test. Reverse the parametrization and the curve is travelled downward and starts somewhere else, with no point of the plane moving at all. If you ever find yourself unsure which side a statement falls on, ask whether an equation in the two coordinates could express it. If it could not, it belongs to the schedule.
Concept
\[ x = r\cos(\omega t), \qquad y = r\sin(\omega t) \]
Figure (svg): A circle of radius 2960 about the origin, with dots at t equal to 0, 6, 12 and 18 hours at the right, top, left and bottom, and arrows running counter-clockwise.
The book's own example puts Lakeland Community College on a circle of radius 2960 miles, turning once as the earth turns. One revolution is completed as the hour runs from zero up to twenty-four.
\[ x = 2960\cos\!\left(\tfrac{\pi}{12}t\right), \qquad y = 2960\sin\!\left(\tfrac{\pi}{12}t\right) \]
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, pp. 1048-1049 — the circular motion equations and Example 10.2.7
Nothing on this slide is new except the vocabulary, and that is the point. You have been writing parametrizations since Chapter 10 without the word.
The example in the book is nicely concrete: a college on the surface of the earth, sitting at a distance of two thousand nine hundred and sixty miles from the axis, carried once round as the earth turns. The parameter is the hour, and the coefficient inside the cosine and the sine is chosen so that twenty-four hours is exactly one revolution.
Check that coefficient for yourself rather than taking it on trust. You want the angle to advance by one full turn as the hour advances by twenty-four, so the coefficient is a full turn divided by twenty-four, which is pi over twelve. Getting that conversion right, from a period you are given to the number that goes inside the trigonometric function, is a skill from Chapter 10 that this section keeps using.
Worked example
Sketch the curve, and mark the direction of travel.
\[ \begin{cases} x = t^2 - 3 \\ y = 2t - 1 \end{cases} \quad t \ge -2 \]
Choose friendly values of t, starting at the left end
Why: The interval starts at minus two, so start there and step up by one.
\[ t = -2,\, -1,\, 0,\, 1,\, 2,\, 3 \]
Evaluate both coordinates at each t
Why: Each value of t gives one point, computed from the two formulas independently.
| t | x = t² − 3 | y = 2t − 1 | point |
|---|---|---|---|
| −2 | 1 | −5 | (1, −5) |
| −1 | −2 | −3 | (−2, −3) |
| 0 | −3 | −1 | (−3, −1) |
| 1 | −2 | 1 | (−2, 1) |
| 2 | 1 | 3 | (1, 3) |
| 3 | 6 | 5 | (6, 5) |
Plot the points in the order of the table
Why: The order is the new information; join them smoothly and draw arrows the way t increases.
\[ (1,-5) \to (-2,-3) \to (-3,-1) \to (-2,1) \to (1,3) \to (6,5) \]
Figure (svg): A parabola opening to the right with vertex (minus 3, minus 1), traced from (1, minus 5) at t equals minus 2, left and up through (minus 3, minus 1) at t equals 0, then right and up to (6, 5) at t equals 3, with arrows pointing the way t increases.
Check the turning point
Why: The first coordinate is smallest when t is zero, so the leftmost point of the curve is at t equal to zero, and the second coordinate there is minus one. That matches the vertex the sketch shows, and the squared t in the first equation is why the path doubles back at all.
\[ x(0) = -3, \quad y(0) = -1 \;\Longrightarrow\; \text{leftmost point } (-3,-1) \]
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1049 — Example 11.10.1
This is the plain method, and it is worth doing once before the systematic one arrives. Choose values of the parameter, compute both coordinates at each, plot, join, and arrow.
Notice how the table is built. The two coordinates are computed completely independently of each other; nothing in the first column feeds the second. What ties them together is only that they were computed at the same value of the parameter, which is why they belong on the same row and produce a single point.
The check step is the habit to copy. The first coordinate is a square, so it is smallest when the parameter is zero, and that instantly tells you where the leftmost point of the curve is without plotting anything more. Whenever one coordinate has an obvious maximum or minimum, evaluate there: it hands you the extreme point of the whole curve for almost no work.
One warning about the arrows. They do not run left to right across the page here, they run up it, and the curve doubles back on itself. If you had been given only the six points with no order, you would probably have joined them in a different sequence.
Prediction
A parametric curve passes through one and the same point of the plane at two different values of the parameter.
Predict first
What does a crossing like that force to be true about the two coordinate functions?
Correct: Neither of them can be one to one
Why: Two different parameter values give the same first coordinate and also the same second coordinate. A one to one function never repeats an output, so neither coordinate function can be one to one. Nothing is undefined and nothing is illegitimate: a crossing is exactly what parametrizations are for.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1048 — footnote 2
Commit to one option first. The wording is deliberately tempting: three of the four options sound like something has gone wrong.
Nothing has. A crossing means two different values of the parameter give the same point of the plane, so both coordinate functions return the same output twice. A function that repeats an output is not one to one, so neither of them is.
That is a fact about the functions, not a fault in them. Being one to one is a property that matters when you want to invert a function, and you do not want to invert these. In fact the failure is what makes the whole method more powerful than a graph: a graph can never cross itself, and a path can.
Section
Part 2
Concept
Plotting points works, but it leaves you guessing between them. There is a systematic method, and it is the same two-plane idea you used for polar graphs: draw the easy graphs first, then read them off.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1050 — the method used throughout Example 11.10.2
Plotting points works but it leaves the spaces between the points to guesswork, and on an unbounded interval it never finishes. The systematic method fixes both problems, and it should feel familiar: it is the same two-plane approach you used to graph polar equations, with one auxiliary graph replaced by two.
Read the six steps as a division of labour. The graph of the first coordinate against the parameter is in charge of left and right. The graph of the second coordinate against the parameter is in charge of up and down. Neither one knows anything about the other; you are the one who reads them at the same moment.
The splitting step is what keeps it manageable. Cut the interval at every turning point of either graph, and on each piece each coordinate has one fixed direction, so the motion on that piece is one of four things: up and right, up and left, down and right, down and left. Sketching then becomes bookkeeping rather than guesswork.
Picture it
Figure (svg): Three panels. Left: x equals t squared minus 3 against t, a parabola with its minimum at t equals 0. Middle: y equals 2t minus 1 against t, a rising straight line. Right: the resulting curve in the plane, a parabola on its side traced upward.
The two panels on the left are ordinary function graphs, of the kind you have drawn since the first chapter. Neither of them is the curve. The curve is what you get by reading them at the same value of the parameter and plotting the pair.
Take a moment with this figure before moving on, because the rest of the part depends on reading it fluently.
The left panel says the first coordinate falls until the parameter reaches zero and rises afterwards, so the point moves left, then right. The middle panel says the second coordinate rises throughout, so the point moves up the whole time. Put those together: up and left, then up and right, with the turn happening at the same moment the left panel turns.
The right panel is the result, and notice that it is a different kind of picture from the other two. The other two have the parameter on the horizontal axis; the right one has no parameter anywhere. That is why the arrows have to be drawn: they are the last trace of the information the first two panels carried.
Worked example
\[ \begin{cases} x = t^3 \\ y = 2t^2 \end{cases} \quad -1 \le t \le 1 \]
Find the range of each coordinate
Why: Cubing sends minus one to minus one and one to one; doubling a square runs from zero up to two.
\[ -1 \le x \le 1, \qquad 0 \le y \le 2 \]
Describe each coordinate's behaviour
Why: The cube is increasing over the whole interval; the square falls to zero at t equal to zero and then rises.
\[ x \uparrow \text{ on } [-1,1]; \quad y \downarrow \text{ on } [-1,0], \; y \uparrow \text{ on } [0,1] \]
Evaluate at the ends and at the turning value
Why: Those three values of t are the ones the behaviour changes at.
\[ t=-1: (-1,2), \quad t=0: (0,0), \quad t=1: (1,2) \]
Transfer to the plane
Why: Right and down to the origin, then right and up: the point arrives at the origin and leaves in a different direction.
\[ (-1,2) \searrow (0,0) \nearrow (1,2) \]
Eliminate the parameter to name the shape
Why: Solve the first equation for t and substitute into the second.
\[ t = \sqrt[3]{x} \;\Longrightarrow\; y = 2\left(\sqrt[3]{x}\right)^{2} = 2x^{2/3} \]
Figure (svg): Three panels. Left: x equals t cubed against t, increasing throughout. Middle: y equals 2t squared against t, falling to zero at t equals 0 then rising. Right: the curve in the plane, two branches meeting in a sharp point at the origin, traced from (minus 1, 2) down to the origin and back up to (1, 2).
Check the sharp point
Why: At t equal to zero both coordinates are momentarily standing still, which is why the path has a corner rather than a smooth turn. Substituting x equal to one eighth into the eliminated equation gives one half, and the parametrization at t equal to one half gives the same point.
\[ x = \tfrac18: \; 2x^{2/3} = 2\cdot\tfrac14 = \tfrac12; \quad t=\tfrac12: \; \left(\tfrac18, \tfrac12\right) \]
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1050 — Example 11.10.2 part 1
Follow the order of the steps, because it is the order the method prescribes: ranges first, then behaviour, then a few points, then transfer, then eliminate.
The ranges box the curve in. The first coordinate never leaves the interval from minus one to one and the second never leaves the interval from zero to two, so the whole picture lives in a small rectangle. Knowing that before you plot anything stops you from drawing a curve that wanders off.
The interesting feature is the corner. Both coordinate graphs are smooth, so it is genuinely surprising that the path is not. The reason is in the middle panel: at the moment the parameter is zero, the second coordinate is momentarily not changing, and the first coordinate, being a cube, is momentarily not changing either. The point slows to a standstill, and a point that stops is free to leave in a completely different direction.
The eliminated equation confirms it. The curve is two times the cube root of the first coordinate, squared, a shape you met in Section 5.3, and it has a cusp at the origin. Worth noticing: the algebra alone would have given you the shape but not the fact that the point pauses there.
Step zero
\[ \begin{cases} x = 2e^{-t} \\ y = e^{-2t} \end{cases} \quad t \ge 0 \]
Discussion prompt
The interval here is unbounded, so a table of values can never finish. Name the two things you should work out before plotting anything, and say why each one is needed.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1051 — Example 11.10.2 part 2, the analysis of end behaviour
Write your answer before revealing. The instinct on seeing two exponentials is to reach for a calculator, and this slide exists to delay that by about a minute.
The interval is unbounded, so no table can ever be complete, and the two facts that replace it are the ranges and the end behaviour. The exponential of a negative quantity is strictly between zero and one and equals one when the exponent is zero, which pins both coordinates into half-open intervals. That is the box.
The end behaviour says where the curve is heading. Both coordinates shrink toward zero as the parameter grows, so the curve approaches the origin. The word approaches is doing real work: it gets arbitrarily close and never arrives, which is why the origin will be drawn as an open circle. Get those two facts first and the friendly values only have to fill in the middle.
Worked example
\[ \begin{cases} x = 2e^{-t} \\ y = e^{-2t} \end{cases} \quad t \ge 0 \]
Pick friendly values of t
Why: The parameter sits in an exponent, so the friendly values are natural logarithms.
\[ t = 0, \; \ln 2 \approx 0.693, \; \ln 3 \approx 1.099 \]
Evaluate at each one
Why: Use the fact that e to the minus natural log of a is one over a.
\[ (2,1), \quad \left(1, \tfrac14\right), \quad \left(\tfrac23, \tfrac19\right) \]
Read the direction from the two graphs
Why: Both coordinates are decreasing for every t at least zero, so the point moves left and down throughout.
\[ x \downarrow, \; y \downarrow \;\Longrightarrow\; \text{left and down, forever} \]
Eliminate the parameter
Why: Solve the first equation for the exponential and substitute into the second, rewritten as a square.
\[ e^{-t} = \frac{x}{2} \;\Longrightarrow\; y = \left(e^{-t}\right)^{2} = \frac{x^{2}}{4} \]
Figure (svg): The parabola y equals x squared over 4 drawn dashed, with the piece from (2, 1) down to the origin drawn solid and arrowed toward the origin, an open circle at the origin, and dots at (2, 1), (1, one quarter) and (two thirds, one ninth).
Check the endpoint and the open end
Why: At t equal to zero the formula gives the point two comma one, and one squared over four is one quarter at x equal to one, matching the second friendly point. The origin satisfies the eliminated equation but no value of the parameter produces it, so it is drawn open.
\[ \tfrac{2^2}{4} = 1 \;\checkmark, \quad \tfrac{1^2}{4} = \tfrac14 \;\checkmark, \quad 2e^{-t} \ne 0 \text{ for any } t \]
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1051 — Example 11.10.2 part 2
The choice of friendly values is the small idea worth stealing here. With the parameter sitting in an exponent, whole numbers are not friendly at all; natural logarithms are, because the exponential undoes them and leaves a fraction.
The direction needs no case analysis this time. Both coordinates are decreasing for every value of the parameter, so the motion is left and down throughout, with no turning points anywhere. That makes this the simplest curve in the example set, and it is a good one to check your method against.
The elimination uses a small trick worth keeping. Rather than solving for the parameter, solve for the whole exponential, then notice that the second coordinate is the square of it. Spotting that one expression is a power of another is often quicker than inverting anything.
Finally, mind the open circle. The origin satisfies the eliminated equation, so if you hand in the equation alone you have claimed a point that no value of the parameter produces. That is the first hint of the gap this part of the lesson is really about.
Trap
Having found the eliminated equation, the shape gets named and the ends get guessed:
\[ y = \frac{x^2}{4} \]
so the curve runs from the origin up to the right, forever.
Wrong at both ends.
The parameter interval fixes the ends, and nothing else does. Here the first coordinate is at most two and never zero, so the curve stops at the point two comma one and never reaches the origin.
\[ 0 < x \le 2 \]
Always convert the interval on the parameter into an interval on x or y, and state it beside the eliminated equation.
This mistake happens at the very end of an otherwise correct solution, which is what makes it expensive. The algebra is right, the shape is right, and then the ends get filled in from a memory of what parabolas look like rather than from the interval that was given.
The discipline that prevents it: after eliminating, immediately translate the interval on the parameter into an interval on one of the coordinates, and write that interval down next to the equation before you do anything else. Here the exponential is at most one and never zero, so the first coordinate runs over the half-open interval from zero to two, and that single line rules out both wrong ends at once.
Tweak it
Parameter explorer
The curve plots the first coordinate of an ellipse parametrization against the parameter, not the ellipse itself. Slide the semi-axis and read off the largest and smallest values the first coordinate takes, and the parameter values where they happen.
\[ x(t) = 1 + {a}\cos t \]
Read the axes before you slide anything. The horizontal axis here is the parameter, not the first coordinate, and the curve is not the ellipse. What is plotted is one coordinate of the ellipse parametrization against the parameter, which is exactly the left-hand panel of the three-panel figures you have been drawing.
Now slide the semi-axis. The curve stretches vertically and nothing else changes: the peaks stay at the parameter values zero and one full turn, and the trough stays at the half turn. The highest and lowest values are the centre plus and minus the semi-axis.
Connect that back to the ellipse. The horizontal extent of the ellipse is precisely the range of this curve, and the parameter values where this curve peaks and troughs are precisely when the moving point is at the right and left ends of the ellipse. Everything the first panel of a three-panel sketch tells you is visible here.
Fill the middle
Still with the exponential pair from Example 11.10.2 part two, at the friendly value where the parameter is the natural logarithm of two.
Fill in the blanks
x = 2e^1 = \frac14, \qquad y = e^___ = ___
Why: e to the minus natural log of two is one half, so twice it is one. For the second coordinate, minus two times the natural log of two is the natural log of one quarter, so the exponential of it is one quarter. The point is one comma one quarter, which is the middle dot in the figure.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1051 — the friendly values of Example 11.10.2 part 2
Fill in both blanks before checking, and do it without a calculator. The whole point of a friendly value is that the arithmetic comes out clean.
For the first coordinate, the exponential of minus the natural logarithm of two is one half, and twice a half is one. For the second, minus twice the natural logarithm of two is the natural logarithm of one quarter, so the exponential of it is one quarter. You could also get there faster by using the relationship you already found: the second coordinate is the first squared over four, and one squared over four is a quarter.
That second route is worth noticing. Once you have eliminated the parameter, the eliminated equation becomes a free check on every point you compute, and it usually takes less work than the original formula.
Section
Part 3
Concept
The pair of equations is a system, so the technique from Section 8.7 applies: solve one of them for the parameter and substitute the result into the other. What comes out is a single equation in x and y, which you can usually recognise.
\[ y = g(t) \;\Longrightarrow\; t = g^{-1}(y) \;\Longrightarrow\; x = f\!\left(g^{-1}(y)\right) \]
Solve whichever of the two is easier to invert. A linear coordinate is the obvious choice; a squared one is not, because taking a root loses a sign.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, pp. 1049-1050 — substitution and elimination as in Section 8.7
This is the same technique as solving a system by substitution from Section 8.7, and it needs no new ideas. Solve whichever equation is easier to invert, put the result into the other, and simplify.
The word easier is doing the work. A linear coordinate inverts perfectly: there is exactly one value of the parameter giving each output, so nothing is lost. A squared coordinate does not: taking a square root forces a choice of sign, and choosing wrongly silently deletes half the curve.
So the rule of thumb is to look at both equations first and invert the tamer one. In the example coming up, one coordinate is a square and the other is linear, and inverting the square instead of the line would turn a clean two-line solution into a case analysis.
Worked example
Return to the first example and turn the picture into an equation.
\[ \begin{cases} x = t^2 - 3 \\ y = 2t - 1 \end{cases} \quad t \ge -2 \]
Solve the easier equation for t
Why: The second one is linear in t, so it inverts with no loss.
\[ y = 2t - 1 \;\Longrightarrow\; t = \frac{y+1}{2} \]
Substitute into the other equation
Why: Now every t is gone.
\[ x = \left(\frac{y+1}{2}\right)^{2} - 3 \]
Rearrange into a recognisable form
Why: Add three, then multiply both sides by four.
\[ (y+1)^{2} = 4(x+3) \]
Name the curve
Why: From Section 7.3 this is a parabola with a horizontal axis.
\[ \text{vertex } (-3,-1), \text{ opening right} \]
Restrict, to match the parameter interval
Why: The parameter starts at minus two, so the second coordinate starts at minus five and only grows.
\[ t \ge -2 \;\Longrightarrow\; y = 2t-1 \ge -5 \]
Figure (svg): The parabola (y plus 1) squared equals 4(x plus 3) drawn dashed for all y, with the part where y is at least minus 5 drawn solid and arrowed upward, and a marked cut at (1, minus 5).
Check the restriction at the cut
Why: Put y equal to minus five into the equation: four times x plus three is sixteen, so x is one, and the curve does start at the point one comma minus five. The equation with the restriction describes the same set of points as the parametrization, and the one thing that cannot be recovered is the direction of travel.
\[ (-5+1)^2 = 16 = 4(x+3) \;\Longrightarrow\; x = 1 \]
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1049 — eliminating the parameter in Example 11.10.1
You already have the sketch for this curve, so watch what the equation adds and what it takes away.
What it adds is a name. The rearranged form is instantly recognisable from Section 7.3 as a parabola with a horizontal axis, and it hands you the vertex directly. Your sketch suggested a parabola; the equation proves it and locates the vertex exactly.
What it takes away is visible in the figure. The dashed part of the parabola satisfies the equation but is not on the curve, because no value of the parameter at least minus two produces it. That is what the restriction on the second coordinate is for, and translating the interval on the parameter into that restriction is a required step, not an optional polish.
And then there is the part no restriction can fix. Look at the figure again: the equation plus the restriction gives you exactly the right solid arc, and nothing whatsoever about which end it starts from. Whenever a problem asks for the direction too, keep the parametrization beside your answer.
Notation
Annotate
On: \( (y+1)^{2} = 4(x+3), \quad y \ge -5 \)
Step through the four annotations and treat them as a checklist for every elimination you ever do.
The first two are about completeness. The equation names the shape, and the restriction cuts it down to the part that is actually travelled. A solution that gives the equation but no restriction is an incomplete answer, and it is the single most common way to lose a mark in this section.
The third annotation is about honesty. There are no arrows in an equation and there is no way to put any in. If you need the direction recorded, record it in words alongside.
The fourth is a warning about what is coming. Sometimes the restriction cannot do its job at all, because the parametrization visits part of the curve more than once and a restriction can only remove points, never count them. Two slides from now you meet exactly that case, so read the equation there with suspicion already in place.
Intuition
It is worth being clear about why this is not just an inconvenience of the method. An equation in x and y is a membership test: you feed it a point and it answers yes or no. Nothing in a yes-or-no test can encode an order.
Orientation is an ordering of the points, and infinitely many different orderings produce the same set. So the moment you throw away the parameter you throw away the ordering, and no amount of extra restriction on x or y can put it back, because restrictions only remove points.
Practical consequence: if a problem asks for direction as well as shape, do the elimination last, and keep the parametrization alongside your final answer.
It is worth understanding this as a matter of principle rather than as a quirk of the method, because that is what stops you from hunting for a trick that recovers the arrows.
Here is the argument in one line. An equation in the two coordinates is a membership test, and a membership test returns yes or no for each point independently. An order is not a property of individual points; it is a relation between them. So there is nowhere in the equation for the order to live.
Restrictions do not change that. A restriction is just another membership test, so adding restrictions only removes points; it never introduces an order among the ones that remain.
The practical consequence is a rule about the order in which you work. Do the sketching and the arrows first, while the parameter is still in front of you, and eliminate last. If you eliminate first and then try to reconstruct the direction from the equation, you will be trying to recover information that no longer exists.
Concept
When the two coordinates are built from sines and cosines, solving for the parameter drags you into inverse trigonometric functions and their restricted ranges. There is a much better move: solve for the trigonometric functions themselves, then relate them with an identity.
\[ \cos t = \frac{x-h}{a}, \quad \sin t = \frac{y-k}{b} \;\Longrightarrow\; \left(\frac{x-h}{a}\right)^{2} + \left(\frac{y-k}{b}\right)^{2} = 1 \]
A reciprocal identity does the same job when a cosecant, secant or cotangent appears, and a Pythagorean identity does it when a tangent and a secant appear together.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1050 — the strategy for trigonometric parametrizations
You could solve a cosine equation for the parameter using an inverse cosine, and you should not. The inverse cosine has a restricted range, so it gives you back only part of the interval, and the algebra that follows is unpleasant.
The better move keeps the parameter inside the trigonometric function. Isolate the cosine in one equation and the sine in the other, and then use the identity that connects them. The parameter vanishes in one step, and nothing was restricted along the way.
Learn to spot which identity the parametrization is built for. A cosine and a sine call for the Pythagorean Identity in its basic form. A cosecant, secant or cotangent beside its partner calls for a reciprocal identity. A tangent and a secant together call for the Pythagorean Identity in the form that relates those two. The choice is dictated by which functions appear, so read the pair before you start.
Worked example
\[ \begin{cases} x = \sin t \\ y = \csc t \end{cases} \quad 0 < t < \pi \]
Find the range of each coordinate
Why: On this open interval the sine is positive and at most one, so its reciprocal is at least one.
\[ 0 < x \le 1, \qquad y \ge 1 \]
Evaluate at three friendly values
Why: The two outer values are supplementary, so their sines agree.
\[ t=\tfrac{\pi}{6}: \left(\tfrac12, 2\right), \quad t=\tfrac{\pi}{2}: (1,1), \quad t=\tfrac{5\pi}{6}: \left(\tfrac12, 2\right) \]
Examine the two open ends
Why: As the parameter approaches either end the sine approaches zero from above and its reciprocal grows without bound.
\[ t\to 0^{+} \text{ or } t\to\pi^{-}: \quad x\to 0^{+}, \; y\to\infty \]
Eliminate the parameter with a reciprocal identity
Why: The cosecant is one over the sine, and the sine is x.
\[ y = \csc t = \frac{1}{\sin t} = \frac{1}{x} \]
Describe the motion on each half
Why: Up to the quarter turn the sine increases and the cosecant decreases; after it, both reverse.
\[ \left(0,\tfrac{\pi}{2}\right): \text{right and down}; \quad \left(\tfrac{\pi}{2},\pi\right): \text{left and up} \]
Figure (svg): Three panels. Left: x equals sine t rising to 1 at t equals pi over 2 and falling back, and y equals cosecant t falling to 1 and rising again, both over the interval from 0 to pi. Middle: for t up to pi over 2 the point runs down the curve y equals 1 over x toward the point 1 comma 1. Right: for t beyond pi over 2 the point runs back up the very same arc.
Check what no restriction can fix
Why: The two dots at the sixth and the five-sixths of a half turn are the same point, so the arc from the vertical asymptote to one comma one is travelled twice. Restricting x to the interval from zero to one gives the right set of points but records one trip, never two.
\[ \sin\tfrac{\pi}{6} = \sin\tfrac{5\pi}{6} = \tfrac12 \;\Longrightarrow\; \text{the arc is covered twice} \]
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, pp. 1051-1052 — Example 11.10.2 part 3 and footnote 3
This is the most important example in the section, and the reason is in the last step rather than the algebra.
The algebra is genuinely easy. A cosecant is one over a sine, and the sine is the first coordinate, so the second coordinate is one over the first. The curve lies on a hyperbola you have known since early algebra.
The behaviour is where it gets interesting. Look at the friendly values: a sixth of a half turn and five sixths of a half turn give the same point, because supplementary angles have the same sine. So the point runs out along the arc, turns around at the quarter turn, and comes back over the same ground. The right-hand panel shows two arrows on one curve pointing opposite ways, and both belong there.
Now ask what equation could record that. Restricting the first coordinate to the interval from zero to one gives exactly the right set of points, so the set is not the problem. The problem is that the arc is covered twice, and a set does not count multiplicity. This is the case the book promises in a footnote, and it is the clearest possible demonstration that a parametrization is strictly more information than an equation with restrictions.
Worked example
\[ \begin{cases} x = 1 + 3\cos t \\ y = 2\sin t \end{cases} \quad 0 \le t \le \tfrac{3\pi}{2} \]
Find the range of each coordinate
Why: The cosine and the sine each swing between minus one and one over this interval.
\[ -2 \le x \le 4, \qquad -2 \le y \le 2 \]
Evaluate at the quarter turns
Why: Those four values of t are where each coordinate turns around.
\[ (4,0), \; (1,2), \; (-2,0), \; (1,-2) \]
Track the motion on each quarter
Why: One coordinate changes direction at each quarter turn.
\[ \nwarrow \text{ then } \swarrow \text{ then } \searrow, \text{ stopping at } (1,-2) \]
Solve for the two trigonometric functions
Why: Isolate the cosine in the first equation and the sine in the second.
\[ \cos t = \frac{x-1}{3}, \qquad \sin t = \frac{y}{2} \]
Apply the Pythagorean Identity
Why: Square both and add; the left side is one.
\[ \left(\frac{x-1}{3}\right)^{2} + \left(\frac{y}{2}\right)^{2} = 1 \]
Name the curve
Why: From Section 7.4 this is an ellipse centred at one comma zero.
\[ \frac{(x-1)^{2}}{9} + \frac{y^{2}}{4} = 1 \]
Figure (svg): An ellipse centred at (1, 0) with horizontal semi-axis 3 and vertical semi-axis 2. The arc from (4, 0) counter-clockwise through (1, 2) and (minus 2, 0) to (1, minus 2) is solid and arrowed; the remaining quarter from (1, minus 2) back to (4, 0) is dashed.
Check the extent against the interval
Why: Three quarter turns are travelled, counter-clockwise, and the fourth is not. Putting the three-quarter-turn value into both formulas gives one comma minus two, which is where the solid arc stops.
\[ t=\tfrac{3\pi}{2}: \; x = 1+3(0) = 1, \; y = 2(-1) = -2 \]
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1052 — Example 11.10.2 part 4
The elimination here is the standard trigonometric one, and it is worth learning as a fixed routine: isolate the cosine, isolate the sine, square both, add, set equal to one.
Watch what happens to the constants as you do it. The one added to the first coordinate becomes the one subtracted inside the squared numerator, and the three multiplying the cosine becomes the three in the denominator, which is then squared to nine. Every number in the parametrization has a job in the final equation, which is a good way to check you have not dropped one.
The interval is what makes this example more than a routine. Three quarter turns are travelled and the fourth is not, so the curve is three quarters of an ellipse. The equation cannot say that, and neither can a restriction on one coordinate alone, since the missing quarter shares its range of first coordinates with a quarter that is travelled.
So the honest final answer has three parts: the equation of the ellipse, the fact that only three quarters of it is traced, and the direction. Write all three.
Error analysis
Annotate
On: \( x=\sin t,\; y=\csc t \;\Longrightarrow\; y=\frac1x, \; 0 < x \le 1 \)
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1052 — the point of footnote 3, made good
Read the line and try to locate the problem before revealing the annotations. It is not in the algebra, which is entirely correct.
The error is in what is being claimed. The restricted equation does describe the same set of points as the parametrization, and if the question had asked for the set, this would be a complete answer. The implied claim is stronger: that the restricted equation describes the same curve, meaning the same journey.
It does not, because the arc is travelled twice and an equation lists points without counting visits. This is the one situation in the section where no amount of extra restriction rescues the equation, and recognising it is the skill being tested.
The repair is not more algebra. It is a sentence: give the equation, give the restriction, and then say in words that the arc is covered out and back as the parameter runs across the interval. On an examination that sentence is worth as much as the algebra above it.
Faded example
An ellipse is parametrized by one plus three cosines for the first coordinate and two sines for the second. Fill in the two denominators that make the Pythagorean Identity appear.
Fill in the blanks
\left(\frac32}\right)^___ + \left(\frac______}\right)^___ = 1
Why: The denominators are exactly the numbers multiplying the cosine and the sine, because those are what you divide by to isolate each trigonometric function. Three goes under the shifted first coordinate and two under the second, and squaring and adding turns the left side into the Pythagorean Identity.
Fill both blanks before checking. If you hesitate, recompute rather than guess, because the two plausible wrong answers here are instructive.
The denominators are the numbers that multiplied the cosine and the sine, because isolating a trigonometric function means dividing by its coefficient. Three goes under the shifted first coordinate and two under the second.
The tempting error is to write nine and four instead, remembering the final form of an ellipse equation where the denominators are squares. Both are right, in different places: three and two go inside the brackets that are then squared, and nine and four appear once you carry out the squaring. Write the version with the brackets first, every time, and the squares take care of themselves.
Trap
From the three-quarter ellipse:
\[ \frac{(x-1)^{2}}{9} + \frac{y^{2}}{4} = 1 \]
so the curve is an ellipse.
Wrong, twice over.
It is three quarters of an ellipse, travelled counter-clockwise from the point four comma zero. Eliminating gives you the curve the points lie on, which may be strictly bigger than the curve itself.
\[ 0 \le t \le \tfrac{3\pi}{2} \]
Finish every elimination with two extra lines: the restriction that trims the equation down, and a sentence saying where the journey starts and which way it goes.
This is the same trap as the earlier one about endpoints, now dressed in trigonometric clothing, and it catches people who got the hard part right.
The eliminated equation tells you the curve the points lie on. That curve may be strictly larger than the curve you were given, and here it is: a whole ellipse rather than three quarters of one.
Build a habit of finishing with two extra lines rather than one. The first line is the restriction, derived from the interval on the parameter. The second is a sentence naming the starting point and the direction. With those two lines your answer contains everything the parametrization contained; without them it contains strictly less.
Section
Part 4
Concept
Figure (svg): Four small panels. First: the graph of a function with dots at t equal to minus 1, 0, 1 and 2, traced left to right. Second: a sideways parabola with dots at y equal to minus 1, 0, 1, traced bottom to top. Third: a line segment with dots at t equal to 0, one half and 1. Fourth: an ellipse with a dot at t equal to 0 on the right and arrows running counter-clockwise.
| to parametrize | let | with |
|---|---|---|
| y = f(x), x over an interval | x = t, y = f(t) | t over that same interval |
| x = g(y), y over an interval | x = g(t), y = t | t over that same interval |
| the segment from (x₀, y₀) to (x₁, y₁) | x = x₀ + (x₁ − x₀)t, y = y₀ + (y₁ − y₀)t | 0 ≤ t ≤ 1 |
| (x − h)²/a² + (y − k)²/b² = 1 | x = h + a cos t, y = k + b sin t | 0 ≤ t < 2π, counter-clockwise |
The last two are worth reading as sentences: a segment is the starting point plus the displacement times t, and an ellipse is its centre plus a cosine and a sine scaled to the two semi-axes.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1053 — Parametrizations of Common Curves
Up to now you have been given a parametrization and asked what it draws. From here the question runs the other way: you are given a curve and asked for a parametrization. These four recipes handle most of what you will be asked for.
The first two barely deserve the name. If one coordinate is already written as a function of the other, promote the input to be the parameter and the recipe writes itself. The only thing to be careful about is which coordinate is the input, since the second recipe is for curves solved for the first coordinate in terms of the second.
The last two are worth reading as English sentences rather than formulas. A segment is a starting point plus a displacement times the parameter. An ellipse is its centre plus an offset whose two parts are the semi-axes times a cosine and a sine.
One caution about the last row. The stated interval gives one full turn counter-clockwise starting from the right-hand end of the horizontal axis. If your problem wants a different direction or a different start, the recipe is still the right beginning; Part 5 shows you how to adjust it.
Notation
Annotate
On: \( x = x_0 + (x_1-x_0)t, \quad y = y_0 + (y_1-y_0)t \)
This formula looks worse than it is, and the annotations are there to make it readable rather than memorable.
Read it as starting point plus displacement times the parameter. At parameter zero you are at the starting point, because the displacement term vanishes. At parameter one you have travelled the whole displacement, so you are at the terminal point. Everything in between is a fraction of the way.
That is also why the interval from zero to one is not negotiable. Stop earlier and you fall short of the terminal point; run longer and you shoot past it along the same line.
A useful consequence to notice: because both coordinates are multiplied by the same parameter, the motion is along a straight line at a steady rate, so the parameter value one half lands exactly at the midpoint, regardless of how long the segment is. Every directed segment you parametrize this way runs on the same clock, which is precisely what makes joining two of them a matter of delaying the second.
Intuition
The first two recipes are almost too simple to be called recipes: promote one coordinate to be the parameter and the other follows. They work precisely because a graph already describes one coordinate as a function of the other.
The last two are both about splitting a description into a fixed part and a moving part. A segment is a fixed starting point plus a fixed direction scaled by t. An ellipse is a fixed centre plus an offset whose two components are a scaled cosine and a scaled sine.
And the ellipse recipe is not something to memorise. It is the Pythagorean Identity read backwards: you want the squares of two quantities to add to one, so make one of them a cosine and the other a sine.
Understanding where the recipes come from is worth more than memorising them, because you will meet curves the table does not cover.
The first two come from the definition of a graph. A graph already expresses one coordinate as a function of the other, so naming the input as the parameter costs nothing and changes nothing.
The last two share a shape: a fixed part plus a moving part. For a segment the fixed part is the starting point and the moving part is a fixed direction scaled by the parameter. For an ellipse the fixed part is the centre and the moving part is an offset that goes round and round.
And the ellipse recipe is the Pythagorean Identity read backwards. You want two squared quantities that always add to one, and you already know two functions that do exactly that. If you ever forget the recipe, write down what you want the identity to give you and the parametrization falls out in one line.
Worked example
Two curves given as graphs. The second has no formula you can write down, which turns out not to matter.
\[ \text{(1) } y = x^{2}, \; -3 \le x \le 2 \qquad \text{(2) } y = f^{-1}(x), \; f(x) = x^{5}+2x+1 \]
Apply the first recipe to part one
Why: The curve is already in the form y equals a function of x, so promote x to be the parameter.
\[ x = t, \quad y = t^{2}, \quad -3 \le t \le 2 \]
Interchange the variables in part two
Why: The graph of the inverse is the graph of x equals f of y.
\[ y = f^{-1}(x) \iff x = y^{5} + 2y + 1 \]
Apply the second recipe rather than solving
Why: There is no elementary formula for y here, but none is needed: name the second coordinate t.
\[ y = t, \quad x = t^{5} + 2t + 1 \]
Fix the interval
Why: An odd-degree polynomial has all real numbers as its range, so the second coordinate of the inverse runs over all real numbers.
\[ -\infty < t < \infty \]
Figure (svg): The steep curve y equals x to the fifth plus 2x plus 1 and its mirror image in the dashed line y equals x, which is the graph of the inverse function, with matching dots at (1, 4) and (4, 1).
Check both answers
Why: In part one, x equals t makes y equal t squared the same as y equal x squared, over the same interval. In part two, t equal to one gives the point four comma one, and f of one is four, so the inverse really does send four to one.
\[ f(1) = 1+2+1 = 4 \;\Longrightarrow\; f^{-1}(4) = 1 \;\checkmark \]
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1053 — Example 11.10.3 parts 1 and 2
Part one is a formality, included so that part two has something to lean on.
Part two is the striking one. You are asked to parametrize the graph of an inverse function whose formula you cannot write down, and the answer is a two-line calculation. The graph of the inverse is the graph of the original with the roles of the coordinates exchanged, so the equation to parametrize is the original polynomial with the coordinates swapped.
Now the second recipe applies: the equation gives the first coordinate as a function of the second, so name the second coordinate as the parameter. There was never any need to solve a fifth-degree equation, and it is just as well, since no formula for that solution exists in general.
This is a good example of the theme of the whole section. The obstacle was that the curve is not the graph of a nice function of the first coordinate. The escape was to stop insisting that one coordinate be a function of the other, and let both be functions of something else.
Worked example
Find a parametrization for the segment that starts at the point two comma minus three and ends at the point one comma five.
Find the displacement in the first coordinate
Why: End minus start.
\[ x_1 - x_0 = 1 - 2 = -1 \]
Find the displacement in the second coordinate
Why: End minus start again, minding the two minus signs.
\[ y_1 - y_0 = 5 - (-3) = 8 \]
Write starting point plus displacement times t
Why: Each coordinate separately.
\[ x = 2 + (-1)t = 2 - t, \qquad y = -3 + 8t \]
State the interval
Why: The recipe uses the unit interval, so that t equal to one lands on the terminal point.
\[ 0 \le t \le 1 \]
Eliminate the parameter to identify the line
Why: Solve the first for t and substitute into the second.
\[ t = 2 - x \;\Longrightarrow\; y = -3 + 8(2-x) = -8x + 13 \]
Figure (svg): The line y equals minus 8x plus 13 drawn dashed, with the segment from (2, minus 3) to (1, 5) drawn solid and arrowed upward, and dots at t equal to 0, one quarter, one half, three quarters and 1.
Check both endpoints
Why: At t equal to zero the first coordinate is two and the line gives minus sixteen plus thirteen, which is minus three. At t equal to one the first coordinate is one and the line gives minus eight plus thirteen, which is five. Both endpoints are right and the direction runs from the first to the second.
\[ x=2: \; y=-3 \;\checkmark \qquad x=1: \; y=5 \;\checkmark \]
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1054 — Example 11.10.3 part 3
The word directed in the question is not decoration. It is telling you that a bare equation of a line will not answer it: you must say which end you start from.
Work the displacements first, and be careful with the signs. From two to one is minus one, and from minus three to five is plus eight; the second one catches people because subtracting a negative feels like an extra step. Once you have those two numbers the parametrization writes itself.
The check is worth copying because it verifies both ends, not just one. Eliminating the parameter gives an ordinary line, and a line through the right two points is not enough on its own, since a line is infinite in both directions. What finishes the check is confirming that parameter zero lands on the initial point and parameter one on the terminal point.
Notice the midpoint in the figure. Because the parameter multiplies both displacements equally, the halfway value of the parameter lands at the halfway point of the segment, which is a quick sanity check you can do in your head.
Worked example
\[ x^{2} + 2x + y^{2} - 4y = 4 \]
Complete the square in each variable
Why: Add one and four to both sides.
\[ (x+1)^{2} + (y-2)^{2} = 4 + 1 + 4 = 9 \]
Put it in the recipe's form
Why: Divide through by nine, so both semi-axes are three.
\[ \frac{(x+1)^{2}}{9} + \frac{(y-2)^{2}}{9} = 1 \]
Read off the two trigonometric functions
Why: Match each squared fraction against the Pythagorean Identity.
\[ \cos t = \frac{x+1}{3}, \qquad \sin t = \frac{y-2}{3} \]
Rearrange into the parametrization
Why: Multiply by three and move the constants across.
\[ x = -1 + 3\cos t, \qquad y = 2 + 3\sin t \]
State the interval
Why: One full revolution, counter-clockwise.
\[ 0 \le t < 2\pi \]
Figure (svg): A circle of radius 3 centred at (minus 1, 2), with dots at (2, 2), (minus 1, 5), (minus 4, 2) and (minus 1, minus 1) for t equal to 0, pi over 2, pi and 3 pi over 2, arrows running counter-clockwise, and the centre marked.
Verify by substituting straight back
Why: Expanding the whole left side collapses to nine times the Pythagorean Identity minus five, so the equation holds at every value of the parameter, not merely at a few.
\[ 9\cos^{2}t + 9\sin^{2}t - 5 = 9(1) - 5 = 4 \;\checkmark \]
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1054 — Example 11.10.3 part 4
The equation is a circle in disguise, and the first job is completing the square in both variables, a technique from Chapter 7 that this section assumes.
Once the equation is in centre-and-radius form, dividing by the radius squared puts it in the shape the recipe wants, with both denominators equal because a circle is an ellipse with equal semi-axes. Then match each squared fraction against a squared trigonometric function and rearrange.
The verification here is better than checking a few points, so study it. Substituting the two expressions into the original equation and expanding, every term involving the parameter either cancels or collapses into nine times the Pythagorean Identity. So the equation holds for every value of the parameter at once, not merely at the four values you happened to test.
Get into the habit of asking whether your check is a spot check or a proof. Spot checks catch arithmetic slips; a substitution like this one catches structural errors too.
Worked example
\[ \frac{x^{2}}{4} + \frac{y^{2}}{9} = 1, \quad \text{left half only} \]
Read off the two semi-axes
Why: The denominators are the squares of the semi-axes.
\[ a = 2, \qquad b = 3 \]
Write the recipe with centre at the origin
Why: No shift is needed here.
\[ x = 2\cos t, \qquad y = 3\sin t \]
Decide which parameter values give the left half
Why: The first coordinate is negative exactly for Quadrant II and Quadrant III angles.
\[ \tfrac{\pi}{2} \le t \le \tfrac{3\pi}{2} \]
Substitute back into the equation
Why: Each fraction cancels to a squared trigonometric function.
\[ \frac{4\cos^{2}t}{4} + \frac{9\sin^{2}t}{9} = \cos^{2}t + \sin^{2}t = 1 \]
Figure (svg): An ellipse with horizontal semi-axis 2 and vertical semi-axis 3. The left half, from (0, 3) at t equals pi over 2 down through (minus 2, 0) to (0, minus 3) at t equals 3 pi over 2, is solid and arrowed; the right half is dashed.
Check the three landmark values
Why: The quarter turn gives zero comma three, the half turn gives minus two comma zero, and the three-quarter turn gives zero comma minus three. The arc runs down the left side, exactly the half that was asked for.
\[ \left(0,3\right) \to (-2,0) \to (0,-3) \]
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1055 — Example 11.10.3 part 5
The parametrization itself takes one line, since the equation is already in the recipe's form with the centre at the origin. All the work is in the interval.
Think about which angles put the moving point on the left. The first coordinate is twice a cosine, so it is negative exactly when the cosine is negative, which happens for Quadrant II and Quadrant III angles. That is the interval from a quarter turn to three quarters of a turn.
Two things to watch. First, the semi-axes are not where you might expect: the larger one belongs to the vertical direction here, because the larger denominator sits under the second coordinate. Second, the endpoints of the interval are included, since the top and bottom of the ellipse are on the boundary of the left half and the problem wants a connected arc.
The check at the end costs three substitutions and confirms both the shape and the direction: the point starts at the top, swings out to the left, and finishes at the bottom.
Picture it
Figure (svg): Four copies of the unit circle. The first is traced counter-clockwise from the point one comma zero; the second twice as fast; the third clockwise from the same point; the fourth clockwise starting from the bottom. In each, a large dot marks where the point is at t equal to 0 and a smaller one where it is at t equal to pi over 4.
The recipes give one answer each, but a curve has infinitely many parametrizations. All four panels trace the unit circle; they differ in where the clock starts, which way it runs, and how fast.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1048 — footnote 1: note the indefinite article
Four panels, four parametrizations, one curve. Look at the large dot in each panel to see where the parameter zero puts you, and the small dot to see where a quarter of a half turn has taken you.
The second panel is the one people find hardest to see. It traces exactly the same points as the first, starting in the same place and going the same way, but it arrives everywhere twice as early, and over the stated interval it goes round twice. Speed is a genuine difference between parametrizations even when direction and starting point agree.
This is why the book is careful to say a parametrization rather than the parametrization. There are infinitely many, and the recipes in the table simply give you a convenient one. Whenever a problem specifies a direction or a starting point, you should expect to have to adjust whatever the recipe hands you.
Sorting
Sort into buckets
Sort each parametrization by whether it traces the whole unit circle exactly once as the parameter runs over the stated interval.
Sort all six, and use one test on each: how much total angle is swept as the parameter crosses the interval.
Three of them sweep exactly one full turn. They differ in direction and in starting point, and neither of those affects the answer to the question asked, which is only about covering the circle once.
The two that fail are instructive and they fail in opposite ways. One sweeps only half a turn, so it draws a semicircle and misses half the points entirely. The other sweeps two full turns, so it draws the whole circle but visits every point twice. Whenever a coefficient appears in front of the parameter inside a trigonometric function, check the total angle swept before anything else.
Matching
Match the pairs
Why: The segment uses starting point plus displacement times t, with displacements minus five and seven. The second curve is already solved for x in terms of y, so the second coordinate becomes the parameter and the bounds come from the two given points. The circle has both semi-axes equal to five. The last ellipse must first be divided by nine, which gives semi-axes three and one about the centre one comma zero.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, pp. 1059-1060 — Exercises 25, 29, 31 and 36
Each of these is one of the four recipes with one small twist, and the twist is what the exercise is really testing.
For the segment, compute both displacements and mind the signs. For the second curve, notice that it is solved for the first coordinate in terms of the second, so the second coordinate becomes the parameter and the bounds come from the second coordinates of the two named points, not the first.
The two conic rows need a preliminary step each. The circle has to be recognised as having both semi-axes equal to five. The ellipse is not yet in the recipe's form and must be divided by nine first, which turns the nine multiplying the squared second coordinate into a one underneath it. Skipping that division is the standard error here, and it produces semi-axes of three and three instead of three and one.
Comparison
Comparison matrix
| curve to parametrize | first coordinate | second coordinate | interval |
|---|---|---|---|
| y = 4 − x² from (−2, 0) to (2, 0) | x = t | y = 4 − t² | −2 ≤ t ≤ 2 |
| the segment from (0, 0) to (3, 4) | x = 3t | y = 4t | 0 ≤ t ≤ 1 |
| the circle x² + y² − 6y = 0, counter-clockwise | x = 3 cos t | y = 3 + 3 sin t | 0 ≤ t < 2π |
| the right half of x²/4 + y²/9 = 1 | x = 2 cos t | y = 3 sin t | −π/2 ≤ t ≤ π/2 |
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, pp. 1059-1060 — Exercises 27 and 33; the other two rows in the same style
Fill in each blank before checking, and say out loud which recipe you are using for each row.
The first row is the graph recipe, so the second coordinate is the same formula with the parameter in place of the first coordinate. The second row is the segment recipe, whose interval is always from zero to one. The third row needs completing the square first, which puts the centre at zero comma three with radius three.
The last row is the one to think about. The right half of the ellipse is the Quadrant I and Quadrant IV angles, and the natural way to write that interval is from minus a quarter turn to a quarter turn rather than in two pieces. Writing it as the union of an interval near a full turn and an interval near zero also describes the right set but makes a mess of the journey, since the parameter would have to jump.
Section
Part 5
Concept
reversing the orientation — Replace every occurrence of the parameter by its negative, in the equations AND in the inequalities that bound it. The same points are traced the other way.
\[ t \mapsto -t: \quad x = f(-t), \; y = g(-t), \; a \le -t \le b \]
shift of parameter — Replace every occurrence of the parameter by the parameter minus a constant, again in the equations and in the bounds. The whole journey is delayed by that constant.
\[ t \mapsto t - c: \quad x = f(t-c), \; y = g(t-c), \; a \le t-c \le b \]
Both are mechanical, and both go wrong in exactly the same way: someone substitutes into the two equations and forgets the inequality. The bounds are part of the parametrization.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1055 — Adjusting Parametric Equations
Both adjustments are substitutions, and both are mechanical. The only skill is doing them completely.
Reversing replaces the parameter by its negative. The same points come out, in the opposite order. Delaying replaces the parameter by the parameter minus a constant, which pushes the whole journey later by that constant without changing the route or the direction at all.
The word everywhere in both definitions is the one that matters. The parameter appears in the two equations and it also appears in the inequality that bounds it, and the inequality is just as much part of the parametrization as the equations are. Substitute into the equations only and you get a set of equations that trace a completely different piece of the curve.
One more piece of advice on order. Reverse first, then delay. Reversing moves the starting point, so if you delay first you will have to redo the calculation of where to start.
Worked example
Find a parametrization of the curve that starts at the point two comma four and follows the parabola to end at the point minus one comma one, with the journey starting at parameter zero.
Parametrize the parabola with the first recipe
Why: It is already in the form y equals a function of x, and the two endpoints fix the bounds.
\[ x = t, \; y = t^{2}, \quad -1 \le t \le 2 \]
Notice the direction is wrong
Why: That version starts at the point minus one comma one, which is the required finish.
\[ \text{starts } (-1,1) \to \text{ends } (2,4) \]
Reverse: replace t by its negative everywhere
Why: Including inside the inequality.
\[ x = -t, \; y = (-t)^{2}, \quad -1 \le -t \le 2 \]
Simplify
Why: The square kills the sign, and multiplying the inequality by minus one flips it.
\[ x = -t, \; y = t^{2}, \quad -2 \le t \le 1 \]
Delay: replace t by t minus two everywhere
Why: The journey currently begins two units too soon.
\[ x = -(t-2), \; y = (t-2)^{2}, \quad -2 \le t-2 \le 1 \]
Simplify again
Why: Expand the square and add two throughout the inequality.
\[ x = 2-t, \; y = t^{2}-4t+4, \quad 0 \le t \le 3 \]
Figure (svg): Two copies of the parabola y equals x squared between x equals minus 1 and x equals 2. On the left the arrows run right and up from (minus 1, 1) to (2, 4) with t from minus 1 to 2. On the right the arrows run left and down from (2, 4) to (minus 1, 1) with t from 0 to 3.
Check the two endpoints and the shape
Why: At parameter zero the point is two comma four and at parameter three it is minus one comma one, as required. And the second coordinate is the square of the first, so the point never leaves the parabola.
\[ y = t^{2}-4t+4 = (2-t)^{2} = x^{2} \;\checkmark \]
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1055 — Example 11.10.4 part 1
Read the question carefully before looking at the work. Two things are specified beyond the shape: which end the journey starts from, and what value the parameter has at the start. Each of those needs one adjustment.
The recipe gives the right arc travelled the wrong way, so reversing comes first. Watch the inequality as the negative goes in: multiplying an inequality by minus one swaps the two ends and flips both signs, which is why the interval comes out running from minus two to one.
Now the direction is right and the parameter starts at minus two rather than zero, so the journey starts two units too soon and a delay of two fixes it. Substitute the parameter minus two everywhere, including the inequality, and add two throughout to get the interval from zero to three.
The check is doubly worthwhile here. Confirming the two endpoints catches an error in the bounds, and confirming that the second coordinate really is the square of the first catches an error in the expansion. Both are easy slips in a calculation with this many substitutions.
Trap
Reversing the parametrization of the parabola:
\[ x = -t, \quad y = t^{2} \]
\[ -1 \le t \le 2 \]
Wrong. The inequality was left alone.
The substitution has to reach every occurrence of the parameter, and the bounds are occurrences. Replacing it inside the inequality and then solving gives a different interval.
\[ -1 \le -t \le 2 \iff -2 \le t \le 1 \]
Left unchanged, the bounds would trace the arc from the point one comma one to the point minus two comma four, which is not the curve that was asked for at all.
This is the single most common error in the whole section, and it is entirely mechanical, which means it is entirely preventable.
Say the rule to yourself as you work: every occurrence, including the ones inside the inequality. A useful physical habit is to underline each occurrence of the parameter in your written line, bounds included, before you substitute, and then check that every underline has been dealt with.
Look at what the error costs. Left unchanged, the interval traces an arc of the same parabola between two completely different points, so the answer is not slightly wrong; it is a different curve. There is no partial credit hiding in a mistake like that, and there is no way to notice it without checking the endpoints, which is why the final check step is never optional.
Worked example
Parametrize the path that goes from the origin along a straight line to the point three comma four, then along a straight line to the point five comma zero.
Parametrize the first segment
Why: Starting point the origin, displacement three and four.
\[ x = 3t, \; y = 4t, \quad 0 \le t \le 1 \]
Parametrize the second segment on its own clock
Why: Starting point three comma four, displacement two and minus four.
\[ x = 3+2t, \; y = 4-4t, \quad 0 \le t \le 1 \]
Notice the clocks collide
Why: Both pieces want the interval from zero to one, so as written they happen at the same time.
\[ \text{first ends at } t=1, \text{ second begins at } t=0 \]
Delay the second piece by one unit
Why: Replace the parameter by the parameter minus one, in the equations and in the bounds.
\[ x = 3+2(t-1), \; y = 4-4(t-1), \quad 0 \le t-1 \le 1 \]
Simplify
Why: Expand, and add one throughout the inequality.
\[ x = 1+2t, \; y = 8-4t, \quad 1 \le t \le 2 \]
Write the whole path as one piecewise parametrization
Why: One clock now runs from zero to two.
\[ f(t)=\begin{cases}3t & 0\le t\le 1\\ 1+2t & 1\le t\le 2\end{cases} \quad g(t)=\begin{cases}4t & 0\le t\le 1\\ 8-4t & 1\le t\le 2\end{cases} \]
Figure (svg): A two-part path: a segment from the origin up to (3, 4), then a segment down to (5, 0). Dots mark t equal to 0 at the origin, t equal to 1 at the corner and t equal to 2 at the end, with arrows along both pieces.
Check the handover
Why: At parameter one the first piece gives three comma four and the second piece gives one plus two comma eight minus four, which is also three comma four. The two definitions agree there, so the path is unbroken, and parameter two gives five comma zero.
\[ f(1): \; 3 = 1+2 \;\checkmark, \qquad g(1): \; 4 = 8-4 \;\checkmark \]
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1056 — Example 11.10.4 part 2
This example is where the delay earns its keep. Each piece of the path is a segment, so each is easy on its own, and the problem is that both recipes hand you the same interval from zero to one.
Two pieces both happening between zero and one is not a path; it is two paths at once. Delaying the second piece by one unit slides it to the interval from one to two, so it begins exactly where and when the first piece stops.
The final answer looks more imposing than it is: two piecewise formulas, one for each coordinate, sharing a single clock that runs from zero to two. Read it as two ordinary parametrizations laid end to end.
The check on the handover is the important one. At the joining value both branches must give the same point, or the path has a jump in it. They do here, which is the guarantee that the two pieces are genuinely joined rather than merely near each other. Any time you build a piecewise parametrization, check that value from both sides.
Worked example
Parametrize the unit circle, oriented clockwise, with parameter zero corresponding to the point zero comma minus one.
Start from the standard parametrization
Why: Counter-clockwise, beginning at the point one comma zero.
\[ x = \cos t, \; y = \sin t, \quad 0 \le t < 2\pi \]
Reverse the orientation
Why: Replace the parameter by its negative everywhere, then use the Even and Odd Identities.
\[ x = \cos(-t) = \cos t, \; y = \sin(-t) = -\sin t, \quad -2\pi < t \le 0 \]
Find when the wanted starting point is reached
Why: It is the bottom of the circle, three quarters of the way round clockwise.
\[ x=0,\; y=-1 \iff t = -\tfrac{3\pi}{2} \]
Slide the interval so it begins there
Why: Any interval of length one full turn covers the circle, so keep the equations and start the parameter at that value.
\[ x = \cos t, \; y = -\sin t, \quad -\tfrac{3\pi}{2} \le t < \tfrac{\pi}{2} \]
Delay by three quarters of a turn
Why: Replace the parameter by the parameter minus that amount, in the equations and in the bounds.
\[ x = \cos\!\left(t-\tfrac{3\pi}{2}\right), \; y = -\sin\!\left(t-\tfrac{3\pi}{2}\right), \quad 0 \le t < 2\pi \]
Simplify with the Sum and Difference Formulas
Why: The cosine of the difference is minus the sine, and minus the sine of the difference is minus the cosine.
\[ x = -\sin t, \qquad y = -\cos t, \quad 0 \le t < 2\pi \]
Figure (svg): The unit circle with dots at (0, minus 1), (minus 1, 0), (0, 1) and (1, 0) for t equal to 0, pi over 2, pi and 3 pi over 2, and arrows running clockwise from the bottom of the circle.
Check the start and the first quarter turn
Why: At parameter zero the point is zero comma minus one, the bottom of the circle. At a quarter turn it is minus one comma zero, the left of the circle, so the motion really is clockwise.
\[ t=0: (0,-1) \;\checkmark \qquad t=\tfrac{\pi}{2}: (-1,0) \;\checkmark \]
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1056 — Example 11.10.4 part 3
This is the hardest calculation in the section and also the most useful, because the cycloid derivation runs on its answer. Take it slowly.
The strategy is the one from the previous slide: fix the direction first, then the starting point. Reversing turns the counter-clockwise circle into a clockwise one, and the Even and Odd Identities tidy the result into a cosine and minus a sine.
The starting point is still wrong: at parameter zero you are at the right of the circle, and the bottom is reached three quarters of a turn earlier in the clockwise direction. Because any interval of one full turn covers the whole circle, you are free to simply slide the interval to begin there, and then delay by that amount so the parameter begins at zero.
The last simplification uses the Sum and Difference Formulas. Do not take it on faith; expand both expressions and watch the cosine of three quarters of a turn vanish and the sine of it turn into minus one. The final parametrization, minus a sine and minus a cosine, is elegant enough to be worth remembering: it is the clockwise circle starting at the bottom.
Concept
Everything so far had an equation in x and y waiting behind it. Now comes the curve that justifies the whole apparatus. Roll a wheel along a straight road and mark one point on its rim; the path that point traces is called a cycloid.
cycloid — The curve traced by a fixed point on the rim of a circle as the circle rolls without slipping along a straight line.
There is no formula giving the height of the traced point as a function of how far along the road you are, and no tidy equation relating the two coordinates. The parametrization, on the other hand, comes out in a few lines.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1056 — the cycloid introduced
Everything so far has had an equation waiting behind it, which might leave you thinking parametrizations are a convenience rather than a necessity. The cycloid settles that.
The definition is physical and needs no algebra: mark a point on the rim of a wheel and roll the wheel along a straight road. The mark traces a curve. You have seen it without noticing, on a bicycle valve at night or a reflector on a moving wheel.
There is no useful equation relating the two coordinates. You can write an implicit relation with an inverse cosine in it, but it is unreadable and useless, and it cannot be solved for either coordinate. Meanwhile the parametrization takes about five lines to derive and states everything cleanly. That asymmetry is the argument for the whole section.
Picture it
Figure (svg): A wheel of radius r rolling right along the x-axis, drawn at the start, the half turn and the full turn, with a spoke to the marked rim point in each. The marked point starts on the ground, rises to height 2r at the half turn, and returns to the ground one circumference later; small dots on the path mark the quarter and three-quarter turns.
Watch the marked point across one turn. It sits on the road, climbs the back of the wheel, reaches twice the radius at the top when the wheel has turned half way, and comes back down to the road exactly one circumference further along.
Trace the marked point across the five stages with your eye before reading anything else. It starts on the road, rises up the back of the wheel, reaches its highest point at twice the radius when the wheel has turned half way, and comes back down to the road exactly one circumference along.
Two features are worth noticing now, because the equations will explain them later. The point touches the road only at isolated instants, and at those instants the curve comes to a sharp point rather than a smooth minimum. That is because the contact point of a rolling wheel is momentarily at rest.
Also compare the marked point with the centre. The centre glides along a horizontal line at a perfectly steady rate. The rim point does not: it is stationary at the bottom and moving fastest at the top, which is why the arch is not a circle or a parabola or anything else you already know.
Worked example
Let the wheel have radius r and let the angle be the amount of clockwise rotation it has undergone. Find the coordinates of the marked rim point.
Start from the previous example, renaming the parameter
Why: A rim point starts at the bottom and the wheel turns clockwise, which is exactly the parametrization just derived.
\[ x = -\sin\theta, \qquad y = -\cos\theta \]
Scale from the unit circle up to radius r
Why: Stretching by r in both directions multiplies both coordinates by r.
\[ x = -r\sin\theta, \qquad y = -r\cos\theta \]
Locate the centre of the wheel
Why: It stays r above the road, and rolling without slipping means the distance travelled equals the arc length turned through.
\[ v t = r\theta \;\Longrightarrow\; \text{centre } = (r\theta,\, r) \]
Shift the rotating offset to sit on the moving centre
Why: Add the centre's coordinates to the offset.
\[ x = -r\sin\theta + r\theta, \qquad y = -r\cos\theta + r \]
Factor out r
Why: This is the standard form.
\[ x = r(\theta - \sin\theta), \qquad y = r(1 - \cos\theta), \quad \theta \ge 0 \]
Figure (svg): A wheel of radius r whose centre sits at the point r theta comma r after rolling through an angle theta. The marked rim point P is reached from the centre by going r sine theta to the left and r cosine theta down; the arc already traced runs from the origin to P.
Check three positions
Why: At angle zero both coordinates are zero, so the point starts on the road. At a half turn the sine is zero and the cosine is minus one, giving the highest point at twice the radius. At a full turn the point is back on the road, one circumference along.
\[ \theta=0: (0,0); \; \theta=\pi: (\pi r, 2r); \; \theta=2\pi: (2\pi r, 0) \]
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, pp. 1056-1057 — the derivation of the cycloid
This derivation is five short steps and each one is something you have already done, so do not be intimidated by the result.
The key decision is the first one. A rim point of a rolling wheel starts at the bottom and goes round clockwise as the wheel rolls to the right, and you derived exactly that parametrization in the previous example. Everything after is adjustment.
Scaling to a wheel of radius r multiplies both coordinates, because stretching the unit circle by that factor in both directions is what turns it into a circle of that radius. Then the centre has to be located: it stays exactly the radius above the road, and rolling without slipping means the distance rolled equals the arc turned through, which is the radius times the angle.
The last move is the one to be careful about conceptually. The scaled circle equations give the offset from the centre to the rim point; the centre's coordinates give where that offset is based. Adding them is what places the rotating offset on the moving centre, and factoring out the radius gives the standard form.
Check it at three angles, as the last step does. Zero, a half turn and a full turn give the bottom, the top and the bottom again, which is exactly what the figure showed.
Worked example
Find the parametric equations of the cycloid traced by a circle of radius three, and give sensible window settings for a graphing utility.
Put the radius into the general equations
Why: Returning to t as the name of the parameter.
\[ x = 3(t - \sin t), \qquad y = 3(1 - \cos t), \quad t \ge 0 \]
Choose the parameter range for one arch
Why: One full revolution of the wheel makes exactly one arch.
\[ 0 \le t < 2\pi \]
Find the horizontal extent
Why: The centre is at three t, so over one revolution it travels three times a full turn.
\[ 0 \le x \le 6\pi \approx 18.85 \]
Find the vertical extent
Why: The cosine runs from one down to minus one, so the second factor runs from zero to two.
\[ 0 \le y \le 6 \]
Extend to three arches
Why: Tripling the parameter range triples the horizontal range and leaves the height alone.
\[ 0 \le t < 6\pi \;\Longrightarrow\; 0 \le x \le 18\pi \approx 56.55 \]
Figure (svg): Two panels. Above: one arch of the cycloid for a circle of radius 3, from the origin up to height 6 at x equal to 3 pi and back to the ground at 6 pi, with dots at the quarter turns. Below: the same curve continued to three arches, reaching x equal to 18 pi.
Check the quarter-turn points and the aspect ratio
Why: At a quarter turn the point is at about one point seven one comma three, and at three quarters at about seventeen point one four comma three: the arch is not symmetric in time even though it is symmetric in shape. Keeping the vertical range at zero to six while the horizontal range is zero to eighteen pi squashes the picture, which is why the book warns about the window.
\[ t=\tfrac{\pi}{2}: (3-3, 3) \to \left(\tfrac{3\pi}{2}-3, 3\right) \approx (1.71, 3) \]
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, pp. 1057-1058 — Example 11.10.5 and the window discussion
With the general equations in hand this is a substitution, and the interesting part is choosing a window for the picture.
The horizontal extent comes from the centre, not from the rim point. The centre sits at three times the parameter, so over one revolution it travels three times a full turn, which is six pi, a little under nineteen. The vertical extent runs from the road to twice the radius, so from zero to six.
The book's warning about the window is worth taking seriously, and the figure shows why. If you set the vertical range to zero to six while the horizontal range is zero to six pi, the picture is stretched vertically by a factor of about three and the arch looks far too pointed. A cycloid only looks like a cycloid when both axes use the same scale.
One last observation from the check. The quarter-turn point sits at about one point seven one along and the three-quarter-turn point at about seventeen point one four. Those are not symmetric about the middle of the arch in terms of how much time each half takes to traverse at a steady horizontal rate, and that asymmetry in speed is exactly what the derivation predicted: slow near the road, fast near the top.
Edge cases
Parameter explorer
The curve plots the height of the rim point against the parameter, for a wheel of the radius you choose. Slide the radius. How high does the point get, and how far apart in the parameter are two visits to the road?
\[ y(t) = {r}(1 - \cos t) \]
Once again, read the axes first. The horizontal axis is the parameter and the curve is the height of the rim point, not the cycloid itself. What you are looking at is the second panel of a three-panel sketch of the cycloid.
Slide the radius and two things stay put while one changes. The zeros stay exactly one full turn apart, whatever the radius, because the point touches the road once per revolution and the revolution is measured by the parameter. The peaks stay at the half turns. What changes is the height, which is always twice the radius.
Notice the shape of each hump. It is not a parabola or a semicircle; it is a raised cosine, flat at the bottom and rounded at the top. That flatness at the bottom is the reason the actual cycloid has a sharp corner where it meets the road: the height leaves zero very slowly while the horizontal position is also barely moving.
Estimation
A wheel of radius three rolls through one full turn. Its centre moves in a straight line for a distance of six pi, about eighteen point eight five. The marked rim point wanders up and over, so its journey is longer.
Predict first
Guess the length of the path the rim point traces over one arch, as a multiple of the radius.
Correct: About 8 times the radius
Why: The arch is exactly eight times the radius long, which for radius three is twenty-four. Measuring the drawn curve numerically gives twenty-four to five decimal places. The straight run of the centre is only six pi, about eighteen point eight five, so the rim point travels about a quarter further than the wheel rolls. Proving the number needs the arc length formula from calculus, but a sketch is enough to rule out the two extreme options.
\[ \text{one arch} = 8r = 24 \quad \text{versus} \quad 2\pi r = 6\pi \approx 18.85 \]
Guess before checking, and make the guess from the picture rather than from a formula. The centre travels about nineteen units in a straight line, and the rim point clearly wanders further than that, so anything at or below six times the radius is out.
The answer is exactly eight times the radius, which for a wheel of radius three makes twenty-four. The figure's curve was measured numerically and comes out at twenty-four to five decimal places, so the clean number is not a coincidence or an approximation.
Two things make that number memorable. It is a whole-number multiple of the radius, with no pi in it at all, even though the width of the arch is full of pi. And it means the rim point travels about a quarter further than the wheel rolls, which is a fact you can feel on a bicycle: a reflector on the rim covers noticeably more ground than the bike does.
Proving it needs the arc length formula from calculus, which is one of the first places parametric equations show up again. Consider that a preview.
Real world
Exercise 42 closes the section by asking for parametric equations that model a passenger's position on the London Eye. Here is the same calculation with round numbers.
Discussion prompt
A wheel of radius sixty metres has its lowest point five metres above the ground and turns once every thirty minutes. You board at the bottom and the wheel carries you first toward the negative x direction. Write your position as a pair of functions of the number of minutes since boarding.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1060 — Exercise 42
Write your two functions before revealing, and treat this as a test of whether the adjustments have stuck.
There are three pieces of information to translate. The boarding point at the bottom and the direction of turn tell you that the clockwise-from-the-bottom parametrization is the right starting point. The radius tells you what to multiply both coordinates by. The height of the lowest point tells you where the centre is, which is what gets added to the second coordinate.
The period is the fourth piece and it is the one people fumble. You want the angle to advance by a full turn as the minutes advance by thirty, so the coefficient inside the trigonometric functions is a full turn divided by thirty, which simplifies to pi over fifteen. This is the same conversion as the college on the rotating earth, and it will be the same one every time.
Check at three moments as the answer does. Boarding puts you five metres up, a quarter of the ride puts you level with the centre, and half the ride puts you at the top at one hundred and twenty-five metres. If any of those comes out wrong, the error is almost always in the coefficient or in the constant added to the second coordinate.
Section
Part 6
Pattern
Figure (svg): A flow diagram with two entry points. Going one way: given the equations, read each coordinate against the parameter, transfer to the plane, mark the orientation, then eliminate the parameter to name the shape and add back any restriction. Going the other way: given a curve, choose the matching recipe, then reverse or delay the parameter to fix the direction and the start.
The diagram has two rows because the section asks two opposite questions, and the order of operations is different for each.
Going from equations to a picture, the rule is do not eliminate first. Sketch from the two coordinate graphs, put the arrows on while the parameter is still in front of you, and only then eliminate to name the shape. If you eliminate first you will have thrown away the direction before you recorded it.
Going from a curve to equations, pick the recipe first and then ask two questions of what it gives you: which way does it go, and where is the point when the parameter is zero. Fix the direction with a reversal and the start with a delay, in that order.
The line at the bottom of the diagram is the one that will save you the most marks. Every time you change the parameter, change the bounds. It is worth writing it at the top of your page before an examination.
Ranking
Put in order
Order the steps that turn the standard counter-clockwise circle into a clockwise one starting at a chosen point, with the parameter beginning at zero.
Why: Direction comes before starting point, because reversing moves the starting point around and you would have to find it again. Once the direction is right, locate the wanted start, slide the interval to begin there, and only then delay so the parameter begins at zero. The final check costs one line and catches a dropped sign immediately.
Drag the six steps into order before checking. Only two of the orderings are genuinely tempting, and the difference between them is worth thinking through.
Direction must be fixed before the starting point. Reversing a parametrization moves everything around, so a starting point located before the reversal will be wrong afterwards and you will have to find it a second time.
The two middle steps are also easy to merge in your head, and keeping them separate is what makes the calculation reliable. First find the parameter value that reaches the wanted point, then slide the interval to begin there, and only then perform the delay that moves that value to zero.
The final check is not padding. It costs two substitutions and it catches the dropped sign that this calculation invites.
Counterexample
Discussion prompt
Someone claims that finding a parametrization means finding the parametrization. Give three genuinely different parametrizations of the segment from the origin to the point one comma two, and say what each pair of them disagrees about.
Write down your three before revealing, and try to make them differ in three different ways rather than three variations on the same idea.
Reversing gives one easy answer, and it disagrees with the original about direction. Substituting a square for the parameter gives another, and it agrees about direction and endpoints but disagrees about speed: the point crawls at the start and hurries at the end. Rescaling the parameter gives a third, which agrees about everything except the reading on the clock.
The general lesson is worth stating plainly. Given any parametrization, substituting any increasing function of a new parameter produces another parametrization of the same curve with the same direction, and any decreasing one reverses it. So there are genuinely infinitely many, and the question is never which is correct but which is convenient for the problem in front of you.
Check
Check your understanding
Two parametrizations produce exactly the same set of points. Which property could still differ between them?
Answer: B
Why: Everything that depends only on which points are present is shared: the line-test verdict, the corners, the equation the points satisfy. The order of visits is not determined by the set of points, so it can differ, and it is exactly what the arrows on a sketch record.
The question is about what is determined by the set of points and what is not, which is the distinction this lesson keeps returning to.
Three of the options are properties of the set: whether it passes a line test, where its corners are, and which equations its points satisfy. Two parametrizations producing the same set of points must agree about all three, because all three are computed from the points alone.
The order of visits is not computed from the points; it is extra information carried by the parametrization. That is why it is the only thing on the list that can differ, and it is exactly what the arrows on your sketches record.
Check
Check your understanding
Eliminate the parameter from x = 2 + 4 cos t, y = −1 + 5 sin t. Which equation results?
Answer: A
Why: Isolate each trigonometric function: the cosine is x minus two over four and the sine is y plus one over five. Squaring and adding gives one, so each squared numerator sits over the SQUARE of the number that multiplied it, giving sixteen and twenty-five.
Run the standard routine rather than trying to spot the answer. Isolate the cosine, isolate the sine, square both, add, set equal to one.
The two things that go wrong are the signs of the shifts and the denominators. Isolating the cosine from a coordinate written as two plus four cosines gives the quantity two subtracted on top, not added, and the whole fraction is squared so the denominator becomes sixteen rather than four.
The last option is a useful check on your understanding rather than a serious contender. The two multipliers are different, so the curve is stretched by different amounts in the two directions and cannot be a circle. If you ever produce an equation whose two denominators are equal from a parametrization whose two multipliers are not, you have made an algebra error.
Check
Check your understanding
The parametrization x = t, y = t³ for 1 ≤ t ≤ 4 is reversed by replacing t with −t. What is the new interval on t?
Answer: B
Why: The substitution reaches the inequality too: one is at most minus t, which is at most four. Multiplying through by minus one reverses both inequality signs and gives minus four at most t at most minus one. The new parametrization is x equals minus t and y equals minus t cubed on that interval.
This is the trap slide made into a question, deliberately, because repetition is the only thing that fixes this error.
The substitution goes into the inequality as well as into the equations. Replacing the parameter by its negative inside the inequality and then multiplying through by minus one flips both signs and swaps the two ends, which is how the interval from one to four becomes the interval from minus four to minus one.
The wrong answers are the three ways people get there. Leaving the bounds alone is the classic slip. Negating only one end forgets that multiplying an inequality by a negative number affects both sides. And shifting the interval to start at zero is a different adjustment altogether, one that was not asked for here.
Explain it to yourself
Discussion prompt
In two or three sentences, explain why a curve's orientation cannot be recovered from an equation in x and y, and say what that means for the order in which you should work a problem that asks for both the shape and the direction.
Write your sentences out before revealing. Being able to say this clearly is a good test of whether the central idea of the lesson has landed.
The argument has two halves. First, an equation in the two coordinates is a membership test, answering yes or no for each point separately, so there is nowhere in it for an ordering of the points to live. Second, restrictions do not help, because a restriction is another membership test and can only remove points, never order them.
The practical half of the answer matters just as much. Since the information is destroyed by eliminating and cannot be recovered afterwards, the order of work is forced: sketch and arrow first, eliminate last, and keep the parametrization next to the final equation. If you find yourself trying to deduce a direction from an equation, stop, because you are trying to recover something that is no longer there.
Exit ticket
\[ \begin{cases} x = 2\cos t \\ y = \sec t \end{cases} \quad 0 \le t < \tfrac{\pi}{2} \]
Discussion prompt
Eliminate the parameter, state the restriction that the interval forces, and say in one sentence which way the point travels.
Stitz & Zeager, College Trigonometry, Ch. 11 Applications of Trigonometry — §11.10 Parametric Equations §11.10, p. 1059 — Exercise 13
This one problem uses most of the lesson: a reciprocal identity, a range, a restriction and a direction.
Start with the identity. A secant is one over a cosine, and the cosine is half the first coordinate, so the second coordinate is two over the first. That is a branch of a familiar hyperbola.
Then the restriction. Over the given interval the cosine runs from one down toward zero without reaching it, so the first coordinate runs from two down toward zero without reaching it, and the second coordinate is at least one. Both restrictions come from the interval, not from the equation.
Finally the direction, which is the part the equation cannot give you. As the parameter increases the cosine decreases, so the point moves left and upward, climbing the branch and hugging the vertical axis without ever touching it. If your answer gave the equation and stopped, you have given the shape and left out the journey, which is the one mistake this lesson has been built to prevent.
Recap
| what you are given | what to do | what you get |
|---|---|---|
| a pair of equations and an interval | graph each coordinate against the parameter, then transfer | the curve with its arrows |
| a pair with algebraic coordinates | solve one for the parameter and substitute | an equation in x and y, plus a restriction |
| a pair with trigonometric coordinates | solve for the trigonometric functions and use an identity | an equation in x and y, plus a restriction |
| a graph, a segment, a circle or an ellipse | the four recipes on the Part 4 table | one parametrization out of infinitely many |
| a parametrization with the wrong direction or start | replace t by minus t, or by t minus c, bounds included | the same curve, correctly scheduled |
\[ x = r(t - \sin t), \qquad y = r(1 - \cos t) \]
The cycloid is the argument for the whole idea: a curve with no usable equation in x and y at all, written down in one line as a path. That closes the course, and it opens calculus, where the same two functions give velocity, arc length and area.
OpenStax Algebra and Trigonometry 2e, §10.6 Parametric Equations §10.6 — the same material in OpenStax
OpenStax Algebra and Trigonometry 2e, §10.7 Parametric Equations: Graphs §10.7 — graphing parametric equations, with further applications
Five rows and one idea. A parametrization gives each coordinate as a function of a third variable, and that buys you two things an equation in the two coordinates cannot supply: curves that are the graph of no function either way, and a direction of travel.
The three techniques are sketching, eliminating and parametrizing. Sketch by graphing each coordinate against the parameter and transferring; eliminate by substitution when the coordinates are algebraic and by an identity when they are trigonometric, and always add the restriction that the interval forces; parametrize from the four recipes, and then adjust with a reversal or a delay, changing the bounds every time.
The cycloid is the closing argument. It is a completely ordinary physical curve with no usable equation in the two coordinates, and one line of parametric equations describes it exactly.
This is also the last section of the course, and it points straight at the next one. In calculus the same two functions are differentiated to give velocity, integrated to give arc length and area, and the cycloid is one of the first curves you will meet again there.
Want this taught 1-on-1? Alexander tutors Trigonometry — $55/session, free consultation.