The graphs of cosine and sine, and the four numbers that describe every transformation of them. Covers periodicity as the property that makes one cycle sufficient, the fundamental cycle and its five quarter marks, the quarter-mark method for graphing any sinusoid without tracking transformations one at a time, the identification of amplitude, period, phase shift and vertical shift from a formula and from a graph, and the recognition that a cosine plus a sine at the same frequency is itself a single sinusoid.
Subject: Trigonometry · 68 slides · symbolic lesson
Open the interactive version of this deck
Title
Trigonometry · Section 10.5.1
One cycle, four numbers, and every wave you will ever need to draw
Objectives
Until now cosine and sine have been numbers attached to angles. Here they become curves, and every stretched, squashed, shifted version of those curves turns out to be described by exactly four numbers.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, pp. 790-800 — learning goals for subsection 10.5.1
So far cosine and sine have been machines that turn an angle into a number. You have built them from the unit circle, worked out their values at the special angles, and pushed them through a page of identities. In this lesson they finally become pictures, and the pictures turn out to be easier to work with than the formulas ever were.
The payoff is a kind of compression. Any curve you can make from the basic cosine or sine wave, by stretching it, squashing it, flipping it or sliding it, is completely described by four numbers. Once you can extract those four numbers from a formula, from a graph, or from a physical description, you can draw the curve, and you can recreate the formula from the drawing.
Work through the five outcomes in order. The first two are mechanical and you should be able to do them without thinking by the end. The last three are where the real understanding sits, and they are also where the exam questions live.
Warm-up
Discussion prompt
Without looking anything up, write the cosine of zero, of a quarter turn, of a half turn, of three quarters of a turn and of a full turn, in that order. Then do the same for sine. What shape are you describing if you read those numbers as heights?
Write your ten numbers down before reading on. If you have to hesitate on any of them, go back to the unit circle for two minutes, because everything in this lesson is built on them and nothing here will stick without them.
Now look at the shape you have described. The cosine list starts high, falls to the bottom, and climbs back to where it started. The sine list starts in the middle, rises to the top, comes back through the middle, drops to the bottom, and returns. Those are the two curves you are about to draw, at their five most important points.
There is a second thing worth noticing. Both lists use the same three values, one and zero and minus one, in a different order. That is the first hint that the two graphs are really one graph, seen from two starting positions, which is exactly what the middle of this part will prove.
Section
Part 1
Concept
Figure (svg): The unit circle with a ray at two thirds of pi meeting the circle at a marked point; a solid arc shows the angle and a dashed arc shows a whole extra turn returning to the same point, with a dashed vertical dropping from the point to the horizontal axis.
Cosine and sine were defined as the coordinates of the point where an angle in standard position meets the unit circle. Two angles that differ by a whole number of full turns are coterminal, so they meet the circle at the very same point.
\[ \cos(t + 2\pi k) = \cos(t), \qquad \sin(t + 2\pi k) = \sin(t), \quad k \in \mathbb{Z} \]
So the two functions cannot hold any new information beyond one turn. Everything they will ever do, they do in a single lap of the circle.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 790 — before Definition 10.3
Look at the circle. The solid arc is the angle you started with, and the dashed arc is a whole extra lap on top of it. Both arcs end at the same place, because a full lap brings you back to where you were. The point on the circle has not moved, so neither of its coordinates has changed.
That is the entire content of the equation on the slide, and it is worth seeing as geometry rather than as algebra. Cosine and sine read off the coordinates of a point on a circle, and a circle has no memory of how many times you went round it.
Keep hold of the consequence. These functions cannot contain any information beyond one lap. If you know what cosine does on one interval of length one full turn, you know what it does everywhere, forever, in both directions. That is what makes graphing them possible at all, because otherwise you would face a curve of infinite length with no shortcut.
Concept
periodic function — A function f is periodic if there is a real number c with f(t + c) = f(t) for every t in the domain of f.
the period — The smallest positive number p for which f(t + p) = f(t) for every t in the domain, when such a smallest number exists.
Two separate ideas hide in that definition. Being periodic only asks that some shift leaves the function alone. The period is the smallest such shift, and the book is careful to say when such a smallest one exists.
A constant function is the warning label. Every positive number leaves it unchanged, so it is periodic; but there is no smallest positive number, so it has no period at all.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 790 — Definition 10.3
Read the two boxes as two separate claims, because they are. The first says a shift exists that does nothing. The second picks out the smallest such shift and gives it a name. It is entirely possible to have the first without the second, and the book makes a point of saying so.
The constant function is the example that keeps you honest. Shift it by any amount you like and nothing happens, so it satisfies the first box many times over. But ask for the smallest positive shift that works and there is no answer, because for every positive number you name there is a smaller positive number that also works. So the constant function is periodic and has no period.
Most of the time this distinction will not bother you, because the functions you meet have a genuine smallest shift. It matters here because it forces you to read definitions as written. The word smallest in that definition is doing real work, and a definition that says smallest is making a promise that something might fail to keep.
Concept
Being unchanged by a shift of two pi is easy. Showing that nothing smaller works takes one line, and it is a nice piece of reasoning.
Suppose some positive p works
Why: Assume the shift leaves cosine alone at every input.
\[ \cos(t + p) = \cos(t) \quad \text{for all } t \]
\[ \text{Put } t = 0: \quad \cos(p) = \cos(0) = 1 \]
\[ \cos(p) = 1 \iff p = 2\pi k \quad \text{for an integer } k \]
The smallest positive multiple of two pi is two pi itself, so no shorter shift can work. Sine is handled the same way, or by the cofunction identity.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 790 — the argument after Definition 10.3
Half of this claim is easy and you already have it: a shift of one full turn does nothing, because of the circle picture on the last slide. The hard half is showing that no shorter shift works, and the trick is a common one in mathematics. Instead of testing every possible shift, assume you have one that works and squeeze information out of that assumption.
The squeeze is a single substitution. If the shift works for every input, it works for the input zero in particular. Feeding in zero turns a statement about all inputs into a statement about one number, and that number turns out to have cosine equal to one.
Now you use what you know about where cosine equals one, which is only at the whole multiples of a full turn. So the mystery shift was a multiple of a full turn all along, and the smallest positive one of those is a full turn itself. Notice how little this argument needed. It would work just as well for any function whose value at zero is attained nowhere else inside a turn.
Notation
Annotate
On: \( f(t + p) = f(t) \quad \text{for every } t \)
Click through the four annotations in order. The first two are about reading the equation correctly: the shift is applied to the input, and the equality is exact, not approximate. Students who skip this step often end up thinking a function is periodic because its graph looks roughly the same in two places.
The third annotation is the one that catches people out. A shift that works at a handful of inputs is worth nothing. The definition asks for every input in the domain, which is a very strong demand, and it is what makes periodicity such a useful property when a function actually has it.
The last annotation reminds you that a periodic function has infinitely many shifts that work, not one. Every whole multiple of the period also works. The period is the smallest, and it is the one worth quoting, because every other working shift is a multiple of it.
Sorting
Sort into buckets
Sort each function by whether some positive shift leaves it completely unchanged.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 790 — the constant-function remark
Sort all six before revealing. For each one, ask yourself a concrete question: is there a distance I could slide this graph horizontally so that it lands exactly on top of itself?
Two of them clearly cannot. A parabola and an exponential both keep increasing on the right, so any horizontal slide leaves the new copy above or below the old one, never on it. Nothing about their shape repeats.
The distance to the nearest whole number is the interesting one. Its graph is a row of identical triangles, a sawtooth, and it repeats every single unit. It is periodic without being smooth and without being wavy, which is a useful antidote to the idea that periodic means wave-shaped. The constant function goes in the periodic bucket, but remember the previous slide: it is periodic with no period at all.
Concept
| property | cosine | sine |
|---|---|---|
| domain | all real numbers | all real numbers |
| range | from −1 to 1 | from −1 to 1 |
| continuous and smooth | yes: no gaps, corners or asymptotes | yes: no gaps, corners or asymptotes |
| even or odd | even: cos(−x) = cos(x) | odd: sin(−x) = −sin(x) |
| period | 2π | 2π |
Four of the five rows are identical. The one genuine difference, evenness against oddness, turns out to be geometric rather than deep: the two graphs are the same curve, slid sideways.
Continuous and smooth is what licenses the very last step of every graph you will draw, which is joining the plotted points with a flowing curve rather than with straight segments.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, pp. 790-791 — Theorem 10.22
This table is less a theorem than a summary of everything you already knew, written in one place. Read down the two columns and notice how much is shared. Same inputs allowed, same outputs possible, same smoothness, same period.
The single genuine difference is in the fourth row. Cosine is even, meaning it does not notice the sign of its input; sine is odd, meaning it flips sign when its input does. Later in the lesson you will use the odd property of sine as a tool, to force a negative coefficient out of an argument, so it is worth having it firmly in mind.
The continuity and smoothness row is the one that licenses your pencil. When you have plotted five or nine points and you join them up, you are allowed to draw a flowing curve with no corners and no jumps precisely because this theorem says there are none. Without that guarantee, joining points would be a guess.
Worked example
Build the graph from the table of common values, assuming nothing about its shape.
\[ \text{Plot } y = \cos(x) \text{ for } 0 \le x \le 2\pi \]
Evaluate at the four quarter marks and the endpoint
Why: These are the quadrantal angles, where the unit-circle point sits on an axis.
\[ \cos: \; 1, \; 0, \; -1, \; 0, \; 1 \quad \text{at } x = 0, \tfrac{\pi}{2}, \pi, \tfrac{3\pi}{2}, 2\pi \]
Fill in the four eighth marks
Why: At every odd multiple of a quarter pi the point sits on a diagonal, so both coordinates have size root two over two.
\[ \cos\left(\tfrac{\pi}{4}\right) = \tfrac{\sqrt2}{2}, \quad \cos\left(\tfrac{3\pi}{4}\right) = -\tfrac{\sqrt2}{2}, \quad \cos\left(\tfrac{5\pi}{4}\right) = -\tfrac{\sqrt2}{2}, \quad \cos\left(\tfrac{7\pi}{4}\right) = \tfrac{\sqrt2}{2} \]
Plot the nine points and join them smoothly
Why: Theorem 10.22 forbids corners, so the join is a flowing curve.
\[ (0,1), \left(\tfrac{\pi}{4}, \tfrac{\sqrt2}{2}\right), \left(\tfrac{\pi}{2}, 0\right), \; \ldots, \; (2\pi, 1) \]
Figure (svg): The graph of y equals cosine of x from zero to two pi, with nine plotted points at the eighth marks: it starts at height one, crosses zero at pi over two, reaches minus one at pi, crosses zero again at three pi over two and returns to one at two pi.
Check that the two ends agree
Why: The height at zero and the height at two pi are both 1, as they must be, because those inputs are coterminal. That agreement is exactly what lets the cycle tile the line with no jump.
\[ \cos(0) = \cos(2\pi) = 1 \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 791 — the fundamental cycle of cosine
This is the graph being built rather than recalled, and it is worth doing once slowly. You are not drawing a shape you remember; you are evaluating a function at nine inputs and seeing what shape appears.
Take the eighth marks seriously. With only the five quarter marks you would have a maximum, a minimum and two zeros, and you would have to guess how the curve gets between them. The four extra points at the odd quarter-pi angles pin that down: they show the curve is already most of the way down by the time it is halfway to the axis, which is what gives the wave its rounded top rather than a triangular one.
The check at the end is small but it is the whole reason this is called a cycle. The height at the start and the height at the end are equal, so when you paste a copy of this piece onto the end of it, the two pieces meet without a step. If those two values disagreed, the tiling would fail and the function would not be periodic.
Picture it
Figure (svg): The cosine curve drawn across four periods from minus four pi to four pi, all cycles identical, with the single cycle from zero to two pi drawn much thicker than the rest.
Copy the highlighted cycle and paste it end to end, forever, in both directions. That is not a picture of the graph; it is the graph, because periodicity guarantees each copy is exact.
This is why the rest of the lesson never draws more than one cycle. Drawing one cycle is a complete answer to the question of what the function looks like.
The thick stretch in the middle is the piece you just built. Everything else on the picture is that piece, copied and pasted, over and over, to the left and to the right.
This is not an approximation or a convenient way of drawing. Periodicity says the copies are exact. So a picture of one cycle is not a partial answer to what does this function look like; it is a complete one, provided you say which cycle you have drawn.
From here on, every graphing question in this lesson asks for one cycle only. When you are asked to graph one cycle of some transformed sinusoid, you are being asked for a finite amount of work that nevertheless determines an infinite curve. That is a good deal, and it comes entirely from the definition on the second slide of this part.
Worked example
The same nine inputs, now feeding the other coordinate of the unit-circle point.
\[ \text{Plot } y = \sin(x) \text{ for } 0 \le x \le 2\pi \]
Evaluate at the quarter marks
Why: Sine is the second coordinate, so it starts on the axis rather than at the top.
\[ \sin: \; 0, \; 1, \; 0, \; -1, \; 0 \quad \text{at } x = 0, \tfrac{\pi}{2}, \pi, \tfrac{3\pi}{2}, 2\pi \]
Fill in the eighth marks
Why: Positive on the first half, negative on the second.
\[ \sin\left(\tfrac{\pi}{4}\right) = \sin\left(\tfrac{3\pi}{4}\right) = \tfrac{\sqrt2}{2}, \qquad \sin\left(\tfrac{5\pi}{4}\right) = \sin\left(\tfrac{7\pi}{4}\right) = -\tfrac{\sqrt2}{2} \]
Figure (svg): The graph of y equals sine of x from zero to two pi, with nine plotted points: it starts at zero, rises to one at pi over two, returns to zero at pi, falls to minus one at three pi over two and comes back to zero at two pi.
Check against the Pythagorean identity
Why: At a quarter pi both coordinates are root two over two, and the squares add to one, so the point really is on the unit circle.
\[ \left(\tfrac{\sqrt2}{2}\right)^2 + \left(\tfrac{\sqrt2}{2}\right)^2 = \tfrac12 + \tfrac12 = 1 \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 792 — the fundamental cycle of sine
Exactly the same procedure, one coordinate over. Where cosine reads the horizontal coordinate of the unit-circle point, sine reads the vertical one, so the values you get are the same numbers arriving in a different order.
Notice where sine starts. At the input zero, the point on the circle is at the far right, whose vertical coordinate is zero. So the sine graph begins on the axis and rises, while the cosine graph begins at its crest. That difference in starting position is the only difference between the two pictures, as the next slide shows.
The check uses the Pythagorean identity as a sanity test on the eighth-mark values. If your two coordinates at a quarter of a turn really are the coordinates of a point on the unit circle, their squares must add to one. They do, which confirms that root two over two is the right number rather than, say, one half.
Concept
The resemblance between the two pictures is not a coincidence. Two identities from the last section turn one into the other.
\[ \sin(x) = \cos\left(\frac{\pi}{2} - x\right) \quad \text{(cofunction)} \]
\[ \cos\left(\frac{\pi}{2} - x\right) = \cos\left(-\left(x - \frac{\pi}{2}\right)\right) = \cos\left(x - \frac{\pi}{2}\right) \quad \text{(cosine is even)} \]
Figure (svg): The cosine and sine curves on one set of axes from minus a half pi to five halves pi, the sine curve identical to the cosine curve moved a quarter period to the right, with a horizontal arrow at height one running from zero across to pi over two.
Subtracting a quarter pi from the input shifts a graph a quarter pi to the right. So the sine curve is the cosine curve moved right by a quarter of a period, and any statement about one is a statement about the other.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 792 — the cofunction argument
This is the single most useful structural fact about the two graphs, and the derivation is only two lines. The cofunction identity rewrites sine as a cosine with a reflected input; the even property of cosine then removes the reflection, leaving a plain subtraction inside.
Subtracting a constant from the input of any function slides its graph to the right by that constant. This is not special to trigonometry; it is the same rule you used on parabolas. So the sine curve is the cosine curve pushed right by a quarter of a turn, which is exactly a quarter of a period.
Look at the picture and verify it with your eyes. Put a finger on any crest of the cosine curve and slide right by a quarter period; you land on a crest of the sine curve. This is why the choice between writing a sinusoid as a cosine or as a sine is a matter of taste rather than of substance, and why the last two parts of this lesson will keep producing both answers.
Worked example
Periodicity is a calculation tool, not just a description of a picture.
\[ \text{Find } \cos\left(\frac{27\pi}{4}\right) \]
Write one period with the same denominator
Why: Two pi is eight quarters of pi.
\[ 2\pi = \frac{8\pi}{4} \]
Subtract whole periods until the angle lands inside one turn
Why: Three periods is twenty-four quarters of pi.
\[ \frac{27\pi}{4} - 3\left(\frac{8\pi}{4}\right) = \frac{27\pi - 24\pi}{4} = \frac{3\pi}{4} \]
Evaluate the reduced angle
Why: Three quarters of pi lies in the second quadrant with reference angle a quarter pi, and cosine is negative there.
\[ \cos\left(\frac{3\pi}{4}\right) = -\frac{\sqrt2}{2} \]
Check the removal was legal
Why: Three whole periods were taken away, and periodicity says each removal leaves the value untouched; a decimal check agrees, since both sides are about minus 0.7071.
\[ \cos\left(\frac{27\pi}{4}\right) = \cos\left(\frac{3\pi}{4}\right) = -\frac{\sqrt2}{2} \approx -0.7071 \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 790 — periodicity of cosine
You have met angle reduction before, but it is worth seeing again in the language of periods rather than of coterminal angles, because it is the same operation viewed from the graph instead of from the circle.
The arithmetic is easier if you first write one period with the same denominator as the angle. Once one period is eight quarters of pi, subtracting periods is subtracting eights from twenty-seven, and you can see immediately that three of them come out, leaving three quarters of pi.
The check is the part to copy into your own work. Each subtraction was legal only because the function is periodic, so say so. And a quick decimal comparison, evaluating both the original and the reduced angle on a calculator, catches an arithmetic slip in five seconds. Reduction errors are almost always off by one period, and a decimal check finds them at once.
Prediction
\[ \cos(0.3) \approx 0.9553. \quad \text{What is } \cos(0.3 + 100\pi)? \]
Predict first
One hundred pi is added to the input. What happens to the value?
Correct: About 0.9553
Why: One hundred pi is fifty whole periods, because each period is two pi. Adding a whole number of periods returns you to the same point of the unit circle, so the value is completely unchanged. The tempting wrong answer flips the sign, which is what an odd number of half turns would do, not an even number of whole turns.
Commit to one option before revealing. The number 0.3 is deliberately not a special angle: the point of the question is that you do not need to know anything about it.
One hundred pi is fifty full turns. Fifty full turns of the circle put you back exactly where you started, so the point has not moved and the value has not changed. You do not need to evaluate anything; you only need to recognise that the added amount is a whole number of periods.
If you were tempted by the sign-flipped option, you were probably thinking of half turns. Adding one half turn does flip the sign of cosine. But one hundred pi is fifty whole turns, not an odd number of half turns, so no flip occurs. When a question adds a large multiple of pi, always ask first whether the multiple is even or odd.
Trap
A reasonable-sounding line, straight after Definition 10.3:
Since a constant function repeats, it is periodic, so it has a period. Its period must be the smallest positive shift that works.
Wrong. There is no smallest one.
Every positive number leaves a constant function unchanged, and the positive numbers have no smallest member. The function is periodic, and it has no period. Definition 10.3 says the period exists only when that smallest number exists, and this is precisely why.
\[ f(t) = 5 \]
\[ f(t + c) = f(t), \; \text{every } c > 0 \]
The wrong line is not careless; it is a reasonable inference from the words, and that is what makes it a good trap. Periodic sounds as though it should imply having a period, and for almost every function you will meet, it does.
The failure is very specific. It needs a function so repetitive that arbitrarily small shifts work, and the constant function is the only one you will meet in this course that manages it. The set of positive shifts that work has no smallest member, and a definition that asks for a smallest member gets no answer.
The wider lesson is about how to read a definition. When a definition says smallest, or largest, or unique, treat that word as a claim that needs checking rather than as decoration. The book chose to include the constant function as an exercise precisely so that you would meet this once, early, in a harmless place.
Section
Part 2
Concept
the argument of a trigonometric function — The whole expression inside the cosine or the sine, and nothing else.
\[ f(x) = 1 - 5\cos(2x - \pi) \]
The argument of the function f is x, because x is what you feed to f. But the argument of the cosine is the expression below, and it is that expression the rest of this part works with.
\[ 2x - \pi \]
Keeping the two straight matters, because every step of the method to come operates on the inside expression and never on the whole formula.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 793 — the argument of a trigonometric function
This slide looks like vocabulary and is really the pivot of the whole lesson. Every technique from here on operates on the inside expression, so if you cannot spot it reliably, nothing else will work.
The word argument is used in two senses on this slide and you should be able to keep them apart. The argument of the function itself is whatever you feed to the function, which here is the variable. The argument of the cosine is the expression handed to the cosine, which is only part of the formula. When a question says the argument, look at the context to see which one is meant, and in this section it is almost always the second.
A practical tip: draw a loop around the inside expression in pencil before you do anything else. The factor in front, the number added at the end, and the minus sign at the front of the whole thing are all outside that loop and play no part in the next step.
Concept
Figure (svg): One cycle of the cosine curve with dashed vertical lines at zero, pi over two, pi, three pi over two and two pi, and the five points on the curve labelled maximum, zero, minimum, zero, maximum.
On the basic cosine curve the five inputs below are exactly where something worth plotting happens: the two ends of the cycle, the low point, and the two crossings of the axis. Between them the curve just travels.
\[ 0, \quad \frac{\pi}{2}, \quad \pi, \quad \frac{3\pi}{2}, \quad 2\pi \]
Now the key move. For a transformed sinusoid, the same five values are still where the interesting things happen, but they are values of the argument, not of x. So set the argument equal to each of them in turn and solve for x. Five one-line solves give the five x-coordinates directly, with no need to apply the transformations in order.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 793 — the quarter marks
Look at the picture first. On the basic cosine cycle, those five dashed lines catch everything you would want to plot: the crest at each end, the trough in the middle, and the two places the curve crosses the axis. Between them nothing surprising happens, so five points and a smooth join reproduce the cycle.
Now the idea that makes the rest of the lesson easy. Suppose the cosine is being fed some expression in x rather than x itself. The cosine still does not care where the number came from; it behaves the same way it always does. So the crests and troughs of the new curve happen exactly when the expression being fed in takes those same five values.
That turns a graphing problem into five linear equations. Set the inside expression equal to each of the five landmark numbers, solve each for x, and you have the five inputs at which the interesting things happen. You never have to decide whether to stretch before shifting, which is the classic way transformation problems go wrong.
Worked example
Graph one cycle and state the period.
\[ f(x) = 3\cos\left(\frac{\pi x - \pi}{2}\right) + 1 \]
Name the argument
Why: Everything inside the cosine, including the division by two.
\[ \frac{\pi x - \pi}{2} \]
Set the argument equal to zero and solve
Why: This locates the start of the cycle.
\[ \frac{\pi x - \pi}{2} = 0 \;\Longrightarrow\; \pi x = \pi \;\Longrightarrow\; x = 1 \]
Repeat for the other four quarter marks
Why: Each is the same two-step solve: multiply by two, add pi, divide by pi.
\[ \frac{\pi x - \pi}{2} = \tfrac{\pi}{2},\; \pi,\; \tfrac{3\pi}{2},\; 2\pi \;\Longrightarrow\; x = 2,\; 3,\; 4,\; 5 \]
| argument equals | equation | x |
|---|---|---|
| 0 | (πx − π)/2 = 0 | 1 |
| π/2 | (πx − π)/2 = π/2 | 2 |
| π | (πx − π)/2 = π | 3 |
| 3π/2 | (πx − π)/2 = 3π/2 | 4 |
| 2π | (πx − π)/2 = 2π | 5 |
Feed each x back into the whole formula
Why: At those five inputs the cosine takes the values 1, 0, minus 1, 0, 1, and then the 3 and the plus 1 act.
\[ y = 3(1)+1,\; 3(0)+1,\; 3(-1)+1,\; 3(0)+1,\; 3(1)+1 = 4,\; 1,\; -2,\; 1,\; 4 \]
Figure (svg): One cycle of y equals three cosine of the quantity pi x minus pi over two, plus one, drawn from x equals one to x equals five, with the five points one comma four, two comma one, three comma minus two, four comma one and five comma four marked and a dashed horizontal line at height one.
Check the period two ways
Why: The cycle was graphed from 1 to 5, a width of 4; and in standard form the coefficient of x inside the cosine is a half pi, so two pi divided by a half pi is also 4.
\[ 5 - 1 = 4, \qquad \frac{2\pi}{\pi/2} = 4 \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 793 — Example 10.5.1, part 1
Follow the first solve carefully and the other four are copies of it. Setting the inside expression equal to zero gives a linear equation, and clearing the denominator and the factor of pi leaves the answer 1. Notice that the pi cancels completely; the five inputs come out as plain whole numbers.
The table is worth copying as a layout. Writing the five equations in a column, with their solutions beside them, keeps the bookkeeping straight and makes it obvious if one of the five is out of step. The solutions should always come out evenly spaced, because the quarter marks are evenly spaced and the solving is linear. Here they are 1, 2, 3, 4, 5, and if one of them had been 3.5 you would know at once that you had slipped.
The substitution step is where the outside numbers finally act. The cosine of the five landmark values is one, zero, minus one, zero, one; multiplying by 3 and adding 1 turns those into the five heights. Doing it in that order is much quicker than substituting the raw inputs into the whole formula.
The check gives the period twice, from the graph and from the formula, and the two agree. Getting the period from the width of your own drawing is free, so there is no excuse for not doing it.
Fill the middle
The same first and last solve, on a different argument. Give each answer as a multiple of pi.
Fill in the blanks
2x - \frac\pi/67\pi/6 = 0 \;\Longrightarrow\; x = ___ \qquad\quad 2x - \frac______ = 2\pi \;\Longrightarrow\; x = ___
Why: Add a third of pi to both sides, then halve. The first equation gives two x equal to a third of pi, so x is a sixth of pi. The second gives two x equal to two pi plus a third of pi, which is seven thirds of pi, so x is seven sixths of pi. The gap between the two answers is pi, which is the period, and that is a free check.
Fill both blanks before checking. Only the first and the last quarter mark are asked for, because those two are the ones that give you the period straight away.
The solving is the same two moves each time: add a third of pi to both sides, then divide by two. The only place people slip is in the second equation, where two pi and a third of pi have to be combined into a single fraction before halving. Two pi is six thirds of pi, so the total is seven thirds of pi, and half of that is seven sixths.
Then check yourself the way the last slide did. The distance between your two answers should be the period, which here is two pi divided by two, that is pi. Seven sixths of pi minus one sixth of pi is indeed pi, so both answers are consistent.
Worked example
Graph one cycle and state the period. The coefficient of x inside the sine is negative, which changes nothing about the method.
\[ g(x) = \frac12\sin(\pi - 2x) + \frac32 \]
Name the argument and set it to zero
Why: Solving for x now involves dividing by minus two.
\[ \pi - 2x = 0 \;\Longrightarrow\; x = \frac{\pi}{2} \]
Solve for the other four quarter marks
Why: Because x is multiplied by a negative number, the solutions come out in decreasing order.
\[ \pi - 2x = \tfrac{\pi}{2},\; \pi,\; \tfrac{3\pi}{2},\; 2\pi \;\Longrightarrow\; x = \tfrac{\pi}{4},\; 0,\; -\tfrac{\pi}{4},\; -\tfrac{\pi}{2} \]
Feed each back into g
Why: The sine values at the quarter marks are 0, 1, 0, minus 1, 0, and the half and the three halves act on them.
\[ y = \tfrac32,\; 2,\; \tfrac32,\; 1,\; \tfrac32 \]
Figure (svg): One cycle of y equals one half sine of the quantity pi minus two x, plus three halves, drawn from minus a half pi to a half pi, with five marked points and a dashed horizontal line at height three halves.
Check the period from the endpoints
Why: One cycle was graphed from minus a half pi to a half pi, and that interval has length pi; the same answer comes from two pi over the size of the coefficient of x, which is two.
\[ \frac{\pi}{2} - \left(-\frac{\pi}{2}\right) = \pi = \frac{2\pi}{2} \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, pp. 793-794 — Example 10.5.1, part 2
The new feature here is the negative coefficient of x inside the sine. Many students freeze at this point and start hunting for a rule about reflections. There is no need: the method does not care about the sign, because solving a linear equation does not care about the sign.
What does change is the order of the answers. Dividing by a negative number reverses the sense of the list, so the five inputs come out decreasing rather than increasing. That is not an error and it does not need fixing. Plot the five points wherever they land, and the cycle appears running from the leftmost to the rightmost.
Read the period off the two extreme inputs rather than off the first and last in your list. Here the smallest is minus a half pi and the largest is a half pi, so the cycle is pi wide. The formula gives the same thing, using the size of the coefficient of x and ignoring its sign.
Error analysis
Annotate
On: \( 3\cos\left(\frac{\pi x - \pi}{2}\right) + 1 = 0, \tfrac{\pi}{2}, \pi, \tfrac{3\pi}{2}, 2\pi \)
Read the line before opening the annotations. It has the right ingredients in the wrong places, which is what makes it worth studying rather than laughing at.
The five landmark numbers are things the cosine is fed. They are inputs, not outputs. Setting the entire function equal to them asks where the graph reaches those heights, which is a genuine question but a completely different one, and not one that locates crests and troughs.
There is a nice practical way to catch this in your own work. If the equations you have written need an inverse cosine to solve, you have set the wrong side equal to the quarter marks. The right version always gives you linear equations, and linear equations are why this method is quick.
Worked example
Graph one cycle, and state the period, amplitude, phase shift and vertical shift.
\[ y = \sin(2x - \pi) \]
Set the argument to each quarter mark
Why: Add pi, then halve.
\[ 2x - \pi = 0, \tfrac{\pi}{2}, \pi, \tfrac{3\pi}{2}, 2\pi \;\Longrightarrow\; x = \tfrac{\pi}{2}, \tfrac{3\pi}{4}, \pi, \tfrac{5\pi}{4}, \tfrac{3\pi}{2} \]
Read the five heights
Why: There is no stretch and no vertical shift, so the heights are the bare sine values.
\[ y = 0, \; 1, \; 0, \; -1, \; 0 \]
Figure (svg): One cycle of y equals sine of the quantity two x minus pi, drawn from a half pi to three halves pi, starting on the axis, rising to one at three quarters pi, back to zero at pi, down to minus one at five quarters pi and back to zero.
State the four numbers
Why: Amplitude one, no vertical shift, and the cycle begins at a half pi rather than at zero.
\[ \text{period } \pi, \quad \text{amplitude } 1, \quad \text{phase shift } \tfrac{\pi}{2}, \quad \text{vertical shift } 0 \]
Check the phase shift against the graph
Why: The graph crosses upward through the axis at a half pi, exactly where the plain sine curve crosses upward at zero, so the whole picture has moved right by a half pi.
\[ \sin\left(2\left(x - \tfrac{\pi}{2}\right)\right) = \sin(2x - \pi) \;\checkmark \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 809 — Exercise 6
This is the first one to try covering the words and following only the maths. Five solves, five substitutions, a picture, four numbers. If you can follow that chain without the prose, you have the method.
The substitution step is unusually easy here because there is nothing outside the sine: no stretch, no vertical shift. So the five heights are just the sine values at the quarter marks, and the wave runs between minus one and one exactly as the basic sine curve does.
Pay attention to the check. The claim that the phase shift is a half pi is tested against the drawing: the upward crossing of the axis, which happens at zero for the plain sine curve, has moved to a half pi. Rewriting the argument with the two factored out makes the same point algebraically. Two descriptions of the same fact, agreeing, is what a check should look like.
Ranking
Put in order
Put a complete graphing run into the order you would actually write it.
Why: The argument has to be named before anything can be set equal to a quarter mark, and the heights cannot be found before the inputs are. Reading the period off the two end inputs comes last, and it doubles as a free check on the arithmetic.
Put them in order before checking. The point of the exercise is not that the order is hard to guess but that writing it out once makes you notice which steps depend on which.
Two dependencies are real. You cannot set anything equal to a quarter mark until you have decided what the argument is, and you cannot find the heights until you have the inputs. Everything else follows from those two.
The last step is the one most often skipped, and it is nearly free. Reading the period off the distance between your first and last input takes two seconds and catches almost every arithmetic slip you could have made in the solving stage. Make it a habit now, before the formulas get harder.
Section
Part 3
Concept
sinusoid — Any curve obtained from the basic cosine or sine graph by stretching, squashing, reflecting or shifting it. Every one of them is described by four numbers.
Figure (svg): A wave two units tall above and below a dashed horizontal baseline at height one, with a vertical double-headed measurement from the baseline up to a crest labelled amplitude, and a horizontal measurement from one crest to the next labelled period.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, pp. 794-795 — the definition of a sinusoid
The definition is deliberately loose: a sinusoid is anything you can get from a basic cosine or sine wave by the standard transformations. What matters is the claim that follows it, which is that four numbers pin one down completely.
Study the picture and make sure you can point at each of the three visible quantities. The amplitude is measured from the middle of the wave up to the crest, not from the axis and not from the trough. The baseline is where that middle sits. The period is the distance from any point to the next point where the wave is doing the same thing.
The fourth quantity, the phase shift, cannot be drawn on a picture of a wave alone, because it says how far this wave sits from where an unshifted one would sit. You need a reference to see it, which is why the picture shows only three. Keep that asymmetry in mind: three of the four are properties of the curve you can measure with a ruler, and the fourth is a comparison.
Concept
Write any sinusoid in the standard form below, with the coefficient of x inside taken positive, and the four quantities can simply be read off.
\[ C(x) = A\cos(\omega x + \varphi) + B \qquad S(x) = A\sin(\omega x + \varphi) + B, \quad \omega > 0 \]
| quantity | formula | in words |
|---|---|---|
| period | 2π/ω | one full repeat |
| amplitude | |A| | crest height above the baseline |
| phase shift | −φ/ω | positive means right, negative means left |
| vertical shift | B | the baseline |
The positive number omega is called the angular frequency: it counts how many cycles the wave completes over an interval of length two pi. The quantity phi on its own is called the phase, but it is the combination below that moves the picture.
\[ \text{phase shift} = -\frac{\varphi}{\omega} \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 795 — Theorem 10.23
This theorem is the engine of the rest of the lesson, and it is worth learning as a table rather than as a sentence. Two of the four quantities are read straight off, and two require a small calculation with the coefficient of x.
The condition that the coefficient of x must be positive is not decoration. If you apply the formulas to a negative coefficient you will get a negative period, which is meaningless, and the phase shift will come out with the wrong sign. There is a slide later in this part on how to force it positive, and you should use it every time.
The name angular frequency is worth knowing because it is what physicists and engineers call this number, and it has a concrete meaning: it counts how many cycles fit into an interval of length one full turn. Doubling it halves the period. If you ever forget whether to multiply or divide, think about that sentence rather than about the formula.
Notation
Annotate
On: \( C(x) = A\cos(\omega x + \varphi) + B \)
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 795 — Theorem 10.23
Take the annotations in pairs. The first and last, the stretch factor and the vertical shift, act outside the cosine and are read off directly. The middle two act inside, and inside is where the complications live.
The third annotation is the one worth writing on your hand. The added constant sits next to the coefficient of x, not next to x, so the shift it causes has been scaled by that coefficient. This is the single most common error in this section, and the next few slides attack it from three directions.
There is also a quiet warning in the first annotation. The amplitude is the size of the stretch factor, so it can never be negative. If a question asks for the amplitude and you write a negative number, you have answered a different question. The sign of the stretch factor tells you whether the wave has been flipped, which is real information, but it is not part of the amplitude.
Tweak it
Parameter explorer
This is the curve of A times cosine of x. Drag A upward and downward, and watch what happens to the wave as A passes through zero into negative values.
\[ y = A\cos(x), \quad A = {A} \]
Drag the control slowly from its starting value up to the top of its range, and watch what stays still. The places where the curve crosses its middle do not move, and the spacing of the crests does not change. Only the height changes.
Now pull the control down through zero. At zero the wave vanishes into a flat line, which makes sense: multiplying every height by zero flattens everything. Keep going and the curve reappears upside down, with troughs where the crests used to be.
That is the whole reason the amplitude is defined as the size of the stretch factor rather than the factor itself. Two settings of equal size and opposite sign give waves of the same height; one is simply the other turned over. So the amplitude answers how tall, and the sign answers which way up, and those are separate questions.
Tweak it
Parameter explorer
This is the curve of the cosine of omega times x, over a fixed stretch of the horizontal axis. Increase omega and count how many complete waves fit into the window.
\[ y = \cos(\omega x), \quad \omega = {w} \]
The window you are looking at has a fixed width, so as you raise the control the waves have to get narrower to fit. Count them. At the starting setting there are two complete waves; push the control to twice that value and there are four.
That counting is the meaning of the name angular frequency. The number tells you how many cycles are completed over an interval of one full turn, and the period, which is how long one cycle takes, is therefore one full turn divided by that number. Bigger frequency, shorter period: they pull in opposite directions.
Notice also what does not change. The crests reach the same height no matter where the control sits. Squeezing a wave horizontally does nothing to it vertically, which is why the amplitude and the period can be discussed completely independently of each other.
Tweak it
Parameter explorer
This is the curve of the cosine of x minus c. Increase c from zero and watch the whole wave travel. What is the smallest positive c at which the picture becomes indistinguishable from the one at c equal to zero?
\[ y = \cos(x - c), \quad c = {c} \]
Start at zero and increase the control slowly. The whole wave travels to the right as one rigid object, keeping its height and the spacing of its crests exactly as they were.
Stop at a few interesting places on the way. Around one and a half you are looking at the sine curve, which is the fact you proved earlier in this lesson. Around three you are looking at the cosine curve turned upside down, because a shift of half a turn moves every crest onto a trough.
Keep going to the top of the range. At about six and a quarter the picture becomes indistinguishable from where you started, because you have shifted by a whole period and a whole period of shifting does nothing. This is worth remembering when your answer to a fitting question differs from the book by exactly one period: both are right.
Tweak it
Parameter explorer
This is the curve of the cosine of x plus B. Move B up and down, and say which features of the wave change and which are untouched.
\[ y = \cos(x) + B, \quad B = {B} \]
Move the control up and down and watch the whole wave rise and fall as one piece. Nothing about the wave itself changes, which is the point: this is the one parameter that touches none of the other three.
What does change is where the wave sits relative to the horizontal axis. Start with the control at zero and the wave spends half its time below the axis. Raise the control past one and the wave never dips below the axis at all, because its lowest point has been lifted above zero.
That last observation is how these models get used in practice. A depth of water, a temperature, or a height above the ground cannot be negative, so a real-world sinusoid nearly always has a baseline chosen to keep the whole wave above zero. The baseline is usually the most physically meaningful of the four numbers: it is the average value of whatever is oscillating.
Concept
Reading the shift straight off the added constant is the single most common slip in this section. Factoring out the coefficient of x shows what the shift really is.
\[ A\cos(\omega x + \varphi) + B = A\cos\left(\omega\left(x + \frac{\varphi}{\omega}\right)\right) + B \]
Written that way it is an ordinary horizontal translation of the wave whose argument is omega times x. Adding a quantity to x moves a graph to the left by that quantity, so the picture slides by minus phi over omega.
Figure (svg): Two waves of the same height and the same period, one a faint reference and one shifted a small distance to the left, with a horizontal arrow between their crests marking a shift of pi over six to the left.
In the picture the added constant is a third of pi and the coefficient of x is two, so the wave moves left by a sixth of pi, half as far as the constant alone would suggest.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 795 — the remark on phase
The formula for the phase shift looks arbitrary until you see this one line of algebra, and then it never looks arbitrary again. Factor the coefficient of x out of the whole argument, and what is left inside the bracket is x plus something.
Once the inside reads x plus something, you are back on ground you know from earlier algebra courses: adding to the input moves the graph left by that amount. And the something is the added constant divided by the coefficient of x, which is exactly the formula.
Look at the picture and check the size of the effect. The added constant is a third of pi, but the wave has only moved a sixth of pi, half as far, because the coefficient of x is two. Squeezing a wave to half its width also halves every horizontal distance inside it, including the shift. If you can reconstruct that sentence, you will never need to memorise the formula.
Worked example
Check the theorem against a graph already drawn by hand.
\[ f(x) = 3\cos\left(\frac{\pi x - \pi}{2}\right) + 1 \]
Split the argument into standard form
Why: Divide each term of the numerator by two, so the coefficient of x is visible.
\[ \frac{\pi x - \pi}{2} = \frac{\pi}{2}x + \left(-\frac{\pi}{2}\right) \]
Read off the four letters
Why: Compare with A cosine of omega x plus phi, plus B.
\[ A = 3, \quad \omega = \frac{\pi}{2}, \quad \varphi = -\frac{\pi}{2}, \quad B = 1 \]
Compute the period and the amplitude
Why: Two pi divided by omega, and the size of A.
\[ \text{period} = \frac{2\pi}{\pi/2} = 4, \qquad \text{amplitude} = |3| = 3 \]
Compute the phase shift and the vertical shift
Why: Minus phi over omega, and B itself.
\[ -\frac{\varphi}{\omega} = -\frac{-\pi/2}{\pi/2} = 1, \qquad B = 1 \]
Check every number against the hand-drawn cycle
Why: The graph ran from 1 to 5, a width of 4; its crest was at height 4 and its trough at minus 2, so it reaches 3 above and below the line at height 1; and the cycle started at 1 instead of at 0.
\[ \tfrac12(4 - (-2)) = 3 \;\checkmark \qquad \tfrac12(4 + (-2)) = 1 \;\checkmark \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, pp. 795-796 — Theorem 10.23 applied to f
This is the theorem being tested against a graph you have already drawn by hand, which is the best way to start trusting it. Nothing here is new information; it is the same curve described a second way.
The first step is the one that needs care. The argument was written as a single fraction, which hides the coefficient of x, so you have to split the fraction into two terms before anything can be read off. Once it is split the four letters are visible, and note that the added constant is negative, since it was being subtracted.
Work through the check line by line. The width of your drawn cycle gives the period. Half the vertical distance between crest and trough gives the amplitude. The average of crest and trough gives the baseline. And the cycle started at 1 rather than at 0, which is the phase shift. Four numbers from the formula, four numbers from the picture, and all four agree. Those last two calculations, half the difference and half the sum, are the tools the next part is built on.
Worked example
The theorem requires a positive coefficient of x. This formula does not have one, so the odd property of sine is used to fix that before anything is read off.
\[ g(x) = \frac12\sin(\pi - 2x) + \frac32 \]
Factor minus one out of the argument
Why: The inside becomes minus the quantity two x minus pi.
\[ \frac12\sin\left(-(2x - \pi)\right) + \frac32 \]
Pull the minus sign out using the odd identity
Why: Sine of minus an angle is minus the sine of that angle.
\[ = -\frac12\sin(2x - \pi) + \frac32 = -\frac12\sin\left(2x + (-\pi)\right) + \frac32 \]
Read off the four letters
Why: Now the coefficient of x inside is a positive two.
\[ A = -\frac12, \quad \omega = 2, \quad \varphi = -\pi, \quad B = \frac32 \]
Compute the four quantities
Why: The amplitude is the size of A, so the minus sign is discarded there but not elsewhere.
\[ \text{period} = \frac{2\pi}{2} = \pi, \quad \text{amplitude} = \frac12, \quad -\frac{\varphi}{\omega} = \frac{\pi}{2}, \quad B = \frac32 \]
Check against the graph, including the apparent mismatch
Why: The hand-drawn cycle ran from minus a half pi to a half pi, so the theorem's phase shift of a half pi is where that cycle ENDS, not where it starts. A fresh cycle does begin there: the rewriting introduced a reflection, and the reflected wave starts one cycle later. Both descriptions give the same curve.
\[ g\left(\tfrac{\pi}{2}\right) = -\tfrac12\sin(0) + \tfrac32 = \tfrac32 \;\checkmark \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 796 — Theorem 10.23 applied to g
The theorem cannot be used as it stands, because the coefficient of x inside the sine is negative. Fixing that is a two-line job and you should do it automatically from now on, before reading off anything at all.
The tool is the odd property of sine from Theorem 10.22. Factor a minus sign out of the whole argument, then let the odd property carry that minus sign out in front of the sine. The coefficient of x is now positive, the added constant has changed sign, and there is a new minus sign attached to the stretch factor.
The final check is the subtle part of this slide, and the book flags it too. The theorem reports a phase shift of a half pi, but the cycle you drew by hand ended there instead of starting there. Both are correct descriptions of the same curve. The rewriting introduced a flip, and the flipped wave begins its cycle exactly one cycle later than the one you drew. Whenever the theorem and your drawing seem to disagree about where a cycle starts, look for a reflection or a whole period.
Trap
A very common line when the coefficient of x is not one:
\[ y = 5\cos(2x + \tfrac{\pi}{3}) \]
\[ \text{shift} = \tfrac{\pi}{3} \text{ left?} \]
Wrong whenever the coefficient of x is not one.
Factor the coefficient out before reading anything. The constant inside is divided by it, so the true shift is a sixth of pi to the left, half of what the raw constant suggests.
\[ 5\cos\left(2\left(x + \tfrac{\pi}{6}\right)\right) \]
\[ -\frac{\varphi}{\omega} = -\frac{\pi/3}{2} = -\frac{\pi}{6} \]
The safe habit is never to quote a shift from a formula you have not yet written in standard form.
This is the most reliable way to lose marks in this section, and it happens because the quantity you want and the quantity you can see are different numbers that look interchangeable.
Two things go wrong at once in the wrong version. The constant has not been divided by the coefficient of x, so the distance is wrong; and the sign has not been negated, so the direction can be wrong too. Notice that when the coefficient of x happens to be one, the distance is accidentally right, which is why the error survives so long: it works on all the easy examples.
The defence is a rule about process rather than about formulas. Never quote a phase shift from a formula you have not yet written in standard form with a positive coefficient of x. If you always factor first, there is nothing left to get wrong.
Matching
Match the pairs
Why: Two of the four act on their own: the size of A is the amplitude and B is the baseline. The other two act through a quotient, because omega multiplies x before phi is added, so both the period and the shift are found by dividing by omega.
Match all four before checking. Two of them are straightforward and two need a second thought, and it is worth noticing which are which.
The stretch factor and the vertical shift act on their own: one sets how tall, the other sets how high. Neither is divided by anything, because neither lives inside the cosine.
The other two act through a quotient, and the reason is the same for both. Inside the cosine, x is multiplied by the frequency before anything else happens, so both the horizontal length of a cycle and the horizontal position of a cycle get scaled by that factor. Dividing by it undoes the scaling. If you can say that sentence in your own words, both formulas stop being things to memorise.
Worked example
State the period, amplitude, phase shift and vertical shift.
\[ y = -\frac32\cos\left(2x + \frac{\pi}{3}\right) - \frac12 \]
Read off the four letters
Why: The formula is already in standard form and the coefficient of x is positive.
\[ A = -\frac32, \quad \omega = 2, \quad \varphi = \frac{\pi}{3}, \quad B = -\frac12 \]
Amplitude and period
Why: The size of A, and two pi over omega.
\[ \text{amplitude} = \left|-\tfrac32\right| = \frac32, \qquad \text{period} = \frac{2\pi}{2} = \pi \]
Phase shift and vertical shift
Why: Minus phi over omega is negative, which means a shift to the left.
\[ -\frac{\varphi}{\omega} = -\frac{\pi/3}{2} = -\frac{\pi}{6}, \qquad B = -\frac12 \]
Say what the negative A does
Why: The wave is flipped, so the cycle begins at a minimum rather than at a maximum.
\[ \text{maximum} = -\tfrac12 + \tfrac32 = 1, \qquad \text{minimum} = -\tfrac12 - \tfrac32 = -2 \]
Check at the start of the cycle
Why: The cycle begins at minus a sixth of pi; substituting gives minus 2, the minimum, confirming both the shift and the flip.
\[ y\left(-\tfrac{\pi}{6}\right) = -\tfrac32\cos(0) - \tfrac12 = -2 \;\checkmark \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 809 — Exercise 11
This one has a negative stretch factor, a negative baseline, and a phase shift that comes out negative, so every sign in the problem is working against you. Take it slowly and write each letter down separately before computing anything.
The amplitude is three halves, not minus three halves, because the amplitude is a size. The minus sign is not lost, though: it tells you the wave is flipped, so the cycle begins at a minimum rather than a maximum, and the step about maximum and minimum values is where that information is cashed in.
The check is the one worth copying. Evaluate the function at the start of the cycle and see whether you get what your description predicts. Here the description says the cycle begins at a minimum of minus 2, and substituting gives exactly minus 2. That single evaluation tests the phase shift, the flip and the vertical shift all at once.
Comparison
Comparison matrix
| formula | amplitude | period | phase shift | vertical shift |
|---|---|---|---|---|
| y = 3 sin(x) | 3 | 2π | 0 | 0 |
| y = sin(3x) | 1 | 2π/3 | 0 | 0 |
| y = cos(x − π/2) | 1 | 2π | π/2 right | 0 |
| y = cos(3x − 2π) + 4 | 1 | 2π/3 | 2π/3 right | 4 |
| y = −(1/3) cos(x/2 + π/3) | 1/3 | 4π | 2π/3 left | 0 |
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 809 — Exercises 1, 2, 4, 7 and 8
Fill in every blank before revealing. The rows get harder going down, and the last row combines all four difficulties at once.
Two rows deserve a second look. In the fourth row the constant inside is a whole two pi, which looks alarming until you divide by the coefficient of x and find a shift of two thirds of pi to the right; and the added 4 outside is the baseline, unaffected by anything happening inside.
The last row has a fractional coefficient of x, which makes the period longer rather than shorter, since dividing by a number smaller than one makes things bigger. It also has a negative stretch factor, so remember that the amplitude is a third and not minus a third. If you got the direction of the shift wrong, check the sign convention again: a positive constant inside gives a shift to the left.
Prediction
\[ y = 4\sin(-2\pi x + \pi) \]
Predict first
What is the period of this sinusoid?
Correct: 1
Why: The coefficient of x inside is negative, so first use the odd property to write the function as minus four times the sine of the quantity two pi x minus pi. Now omega is two pi, and the period is two pi divided by two pi, which is 1. Whenever a pi appears as a factor of x, the pi cancels out of the period and the answer is a plain number, which is exactly what you want for a wave measured in seconds or metres rather than in radians.
\[ 4\sin(-2\pi x + \pi) = -4\sin(2\pi x - \pi), \quad \omega = 2\pi, \quad \text{period} = \frac{2\pi}{2\pi} = 1 \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 809 — Exercise 12
Commit to an answer before revealing. The negative coefficient of x is the first hurdle and the factor of pi is the second, and each one on its own is enough to derail a quick guess.
Deal with the sign first, exactly as you did two slides ago: factor the minus out of the argument and let the odd property of sine carry it to the front. Now the coefficient of x is positive and the theorem applies.
Then notice what happens to the pi. The coefficient of x carries a factor of pi, and the period formula divides a two pi by it, so the pi cancels and the period is a plain 1. This is how sinusoids that model real quantities are almost always written: you want a period of 12 seconds or 24 hours, not of some multiple of pi, so the pi is built into the coefficient of x precisely so that it will cancel.
Section
Part 4
Concept
Given a picture instead of a formula, each of the four quantities can be recovered by measuring, and then the letters follow by rearranging Theorem 10.23.
| measure this on the graph | and it gives |
|---|---|
| the width of one cycle, call it p | ω = 2π/p |
| half the distance from the lowest point to the highest | the amplitude, so A = that number (or its negative) |
| the average of the highest and lowest values | the vertical shift B |
| the x-value where a cycle begins | the phase shift, so φ = −ω times the phase shift |
\[ A = \frac{\text{max} - \text{min}}{2}, \qquad B = \frac{\text{max} + \text{min}}{2} \]
Those last two formulas are worth memorising in the form given. Half the spread is the amplitude; the midpoint of the spread is the baseline.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, pp. 796-797 — Example 10.5.2
Everything in this part is the last part read from right to left. You measure four things on a picture and then rearrange the four formulas to recover the letters in the standard form.
The two formulas on the slide are the ones to commit to memory, and they come as a pair for a reason. Half the difference of the highest and lowest values is the amplitude, and half their sum is the baseline. Computing both together, from the same two numbers, makes it almost impossible to confuse them.
The phase shift is the only one of the four that needs a decision rather than a measurement, because you have to choose which cycle you are describing and whether you are fitting a cosine or a sine. A cosine cycle starts at a crest; a sine cycle starts on the baseline heading upward. Deciding that first makes the rest mechanical.
Picture it
Figure (svg): One cycle of a sinusoid drawn from x equals minus one to x equals five, starting at a maximum height of five halves, dropping to a minimum of minus three halves at x equals two and returning to five halves at x equals five, with dashed lines marking the maximum, the minimum and the midline.
The top line and the bottom line are half of the answer already. The gap between them is four, so half the gap is two and the amplitude is two. The midpoint between them is a half, so the baseline is a half.
The horizontal data is on the picture too. The cycle runs from minus one to five, so it is six units wide, and it starts at a crest, which is where a cosine cycle starts.
Look at the three dashed horizontal lines. Two of them touch the curve, at its highest and lowest points, and the third sits exactly halfway between them. That third line is the baseline, and drawing it in is the first thing you should do on any fitting problem.
Once it is drawn, two of the four numbers are already yours. The distance from the baseline up to the top line is the amplitude. The height of the baseline is the vertical shift. Both come from the same two readings.
The horizontal information is on the picture too, and it is worth reading before you start calculating. The cycle drawn here runs from minus one to five, so it is six units wide. And it begins at a crest, which is what a cosine cycle does, so fitting a cosine to it will be the easier of the two jobs.
Worked example
Find a cosine function whose graph is the cycle shown on the previous slide.
\[ C(x) = A\cos(\omega x + \varphi) + B \]
Find omega from the width of the cycle
Why: One cycle runs from minus one to five.
\[ p = 5 - (-1) = 6 = \frac{2\pi}{\omega} \;\Longrightarrow\; \omega = \frac{\pi}{3} \]
Find phi from the phase shift
Why: A cosine cycle starts at a crest, and this crest is at minus one, so the phase shift is minus one.
\[ -\frac{\varphi}{\omega} = -1 \;\Longrightarrow\; \varphi = \omega = \frac{\pi}{3} \]
Find the amplitude from half the spread
Why: The range runs from minus three halves up to five halves.
\[ A = \frac12\left(\frac52 - \left(-\frac32\right)\right) = \frac12(4) = 2 \]
Find the vertical shift from the average of the extremes
Why: Add the two ends of the range and halve.
\[ B = \frac12\left(\frac52 + \left(-\frac32\right)\right) = \frac12(1) = \frac12 \]
Assemble the formula
Why: Substitute the four numbers into the standard cosine form.
\[ C(x) = 2\cos\left(\frac{\pi}{3}x + \frac{\pi}{3}\right) + \frac12 \]
Check at two of the marked points
Why: At x equal to minus one the argument is zero, giving the crest; at x equal to 2 the argument is pi, giving the trough.
\[ C(-1) = 2(1) + \tfrac12 = \tfrac52 \;\checkmark \qquad C(2) = 2(-1) + \tfrac12 = -\tfrac32 \;\checkmark \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, pp. 796-797 — Example 10.5.2, part 1
Work in the order the steps give, because it is the order in which the information becomes available. The width of the cycle gives the frequency; the start of the cycle then gives the constant inside; and the vertical readings give the remaining two.
The second step is the one people get backwards. The phase shift is minus one, because the cycle starts one unit to the left of the origin, and the constant inside is minus the frequency times that shift. Two negatives make the constant positive here, so do the arithmetic rather than guessing the sign.
The check evaluates the finished formula at two of the marked points. Choose points where the argument comes out as a landmark value, as here, and no calculator is needed: at the first point the argument is zero and the cosine is one, at the second it is half a turn and the cosine is minus one. If the two agree with the graph, your four numbers are almost certainly all correct, because a wrong one would show up at one or other of those points.
Fill the middle
A different sinusoid, described only by its highest and lowest values: the maximum is 11 and the minimum is minus 1.
Fill in the blanks
A = \frac65 = ___, \qquad B = \frac______ = ___
Why: Half the difference of the extremes is the amplitude: twelve halved is 6. Half their sum is the baseline: ten halved is 5. So the wave runs 6 units above and 6 units below the line at height 5, which does indeed reach 11 at the top and minus 1 at the bottom.
Fill both blanks before checking, and do them together rather than one at a time. That is the habit the last few slides have been building.
Half the difference of the extremes is the amplitude and half their sum is the baseline. Here the extremes are 11 and minus 1, so the difference is 12 and the sum is 10, giving 6 and 5.
Test your answers by rebuilding the picture from them. A wave of amplitude 6 sitting on a baseline of 5 reaches up to 11 and down to minus 1, which is what you were given. Doing this reverse check takes a few seconds and catches the classic slip of reading the amplitude as the maximum, which is the subject of a slide two on from here.
Worked example
Most of the work is already done: the period, the amplitude and the vertical shift do not care whether the formula uses cosine or sine. Only the phase shift changes.
\[ S(x) = A\sin(\omega x + \varphi) + \frac12, \quad \omega = \frac{\pi}{3}, \quad A = 2 \]
Ask where a SINE cycle begins
Why: Not at a crest: a sine cycle starts on the baseline and heads upward.
\[ \text{start where } y = \tfrac12 \text{ and the curve is rising} \]
Figure (svg): The same sinusoid continued to the right past x equals nine, with the stretch beginning at x equals seven halves drawn thicker to show a cycle that starts on the midline going upward.
Read that starting point from the extended graph
Why: Continuing the curve to the right, the next upward crossing of the baseline is at seven halves.
\[ \text{phase shift} = \frac72 \]
Turn the phase shift into phi
Why: Multiply by minus omega.
\[ -\frac{\varphi}{\omega} = \frac72 \;\Longrightarrow\; \varphi = -\frac72\cdot\frac{\pi}{3} = -\frac{7\pi}{6} \]
Assemble the formula
Why: Same A, same omega, same B, new phi.
\[ S(x) = 2\sin\left(\frac{\pi}{3}x - \frac{7\pi}{6}\right) + \frac12 \]
Check that the two formulas agree at the marked points
Why: At x equal to minus one both give five halves, and at x equal to 2 both give minus three halves, so the sine formula and the cosine formula describe the very same curve.
\[ S(-1) = 2\sin\left(-\frac{3\pi}{2}\right) + \tfrac12 = \tfrac52 \;\checkmark \qquad S(2) = 2\sin\left(-\frac{\pi}{2}\right) + \tfrac12 = -\tfrac32 \;\checkmark \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 797 — Example 10.5.2, part 2
Three of the four numbers carry over unchanged, and it is worth being clear about why. The height of the wave, the length of a cycle and the position of its middle are properties of the curve itself. They do not know or care which function you are going to write it with.
Only the starting point changes, and that is because cosine and sine start their cycles in different places. A cosine cycle starts at a crest; a sine cycle starts on the baseline going upward. Look at the extended picture and find the first such point to the right: it is at seven halves, where the curve crosses its middle on the way up.
Turning that shift into the constant inside is a single multiplication, and the sign matters: the constant is minus the frequency times the shift. The final check is the one that really settles the matter, because it evaluates both the cosine answer and the sine answer at the same two points and gets the same values. Two very different-looking formulas, one curve.
Concept
Neither answer above is the answer. A cycle of the same curve begins at infinitely many places, and allowing a negative amplitude doubles the choices again.
\[ S(x) = 2\sin\left(\frac{\pi}{3}x - \frac{7\pi}{6}\right) + \frac12 = -2\sin\left(\frac{\pi}{3}x - \frac{\pi}{6}\right) + \frac12 \]
The second version starts its cycle at one half instead of at seven halves, and takes the amplitude factor negative so that the flipped sine curve still runs downward from there. Both are correct, and identities convert either into the other.
So when you are asked to fit a sinusoid, expect your answer to differ from the back of the book. Test it by evaluating both at a couple of the marked points, which settles the question in ten seconds.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 797 — the remark after Example 10.5.2
This is the slide to remember when your answer does not match the back of the book. A fitting problem has infinitely many correct answers, and there is no way to tell from the question which one is wanted.
There are two independent sources of freedom. You can start your cycle at any of the infinitely many points where a cycle of the right kind begins, each one a whole period from the last. And you can take the stretch factor negative, which flips the wave and shifts the starting points by half a period.
So do not try to guess the book's choice. Produce any correct formula, then test it: evaluate it at two or three of the marked points on the graph and check the values match. That test takes under a minute and it is proof, which guessing never is.
Error analysis
Annotate
On: \( \text{max} = \tfrac52, \; \text{min} = -\tfrac32 \;\Longrightarrow\; A = \tfrac52, \; B = 0 \)
Read the line before opening the annotations, and see whether you can say in one sentence what assumption has been smuggled in.
The assumption is that the wave is centred on the horizontal axis. When that is true the highest point really is the amplitude, which is why this error survives every textbook example where the baseline happens to be zero. The moment a wave is lifted or lowered, it fails.
The detection method in the third annotation is worth adopting generally. Take your proposed amplitude and baseline, work out the highest and lowest values they predict, and compare with the picture. Here the prediction is a wave running from minus five halves to five halves, and the picture runs from minus three halves to five halves. They disagree, so something is wrong, and you know it before you have written a formula.
Real world
Figure (svg): Left: a circle of radius three with its centre four units above the ground, a marked car on the rim and a dashed horizontal showing its height. Right: the height of that car plotted against time over two full turns, a smooth wave between one and seven with its middle at four.
Discussion prompt
An observation wheel has radius 3 metres and its centre sits 4 metres above the ground. It turns once every 12 seconds, and a car starts at the lowest point. Write the car's height as a function of time, then say what each of the four parameters means in the physical situation.
Write your own formula before revealing. Start by deciding which of the four numbers the problem has handed you directly, because three of them are stated almost in words.
The radius is the amplitude, because the car rises that far above the axle and falls that far below it. The height of the axle is the baseline, because that is the middle of the motion. The time for one turn is the period, so the coefficient of x follows by dividing a full turn by twelve. Only the starting position takes a decision, and starting at the bottom is what forces the minus sign in front of the amplitude: an unflipped cosine would start at the top.
Look at the two panels together. The left one shows a point going round a circle; the right one shows its height plotted against time. The height coordinate of a point moving round a circle at a steady rate is exactly a sinusoid in time, and that single sentence is why these curves model tides, alternating current, daylight hours, and the vibration of a guitar string.
Section
Part 5
Concept
The sum identity turns the standard form into a combination of a plain cosine and a plain sine at the same frequency.
\[ C(x) = A\cos(\omega x + \varphi) + B = A\cos(\omega x)\cos(\varphi) - A\sin(\omega x)\sin(\varphi) + B \]
\[ S(x) = A\sin(\omega x + \varphi) + B = A\sin(\omega x)\cos(\varphi) + A\cos(\omega x)\sin(\varphi) + B \]
Read those from right to left and they say something surprising. Any combination of a cosine and a sine at one common frequency, plus a constant, is already a sinusoid; it has just been written with its amplitude and phase scrambled into two coefficients.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, pp. 797-798 — the expanded forms
These two lines are just the sum identities from the previous section applied to the standard forms, so there is nothing new to prove. What is new is the direction in which you are going to read them.
Read from left to right and they say a sinusoid can be pulled apart into a cosine piece and a sine piece at the same frequency. Read from right to left and they say something far more useful: if you ever meet a cosine piece and a sine piece at the same frequency added together, you are looking at a single sinusoid that has been written in disguise.
Notice that the two coefficients in the expansion both contain the amplitude and the phase, tangled together. Untangling them is what the next few slides do, and it turns out to need only the Pythagorean identity and a little care with signs.
Worked example
Show this is a sinusoid by putting it in the cosine standard form.
\[ f(x) = \cos(2x) - \sqrt3\,\sin(2x) \]
Match it against the expanded cosine form
Why: The frequency is visible and there is no constant term.
\[ \cos(2x) - \sqrt3\,\sin(2x) = A\cos(2x)\cos\varphi - A\sin(2x)\sin\varphi, \quad \omega = 2, \; B = 0 \]
Equate the two coefficients
Why: The coefficient of the cosine on each side, then the coefficient of the sine.
\[ A\cos\varphi = 1, \qquad A\sin\varphi = \sqrt3 \]
Eliminate phi with the Pythagorean identity
Why: Square both equations and add; the squares of cosine and sine add to one.
\[ A^2 = A^2\cos^2\varphi + A^2\sin^2\varphi = 1^2 + (\sqrt3)^2 = 4 \]
Choose A and solve for phi
Why: Take A positive; then the two equations ask for an angle with cosine a half and sine root three over two.
\[ A = 2: \quad \cos\varphi = \frac12, \; \sin\varphi = \frac{\sqrt3}{2} \;\Longrightarrow\; \varphi = \frac{\pi}{3} \]
Write the answer
Why: Substitute A, omega and phi into the standard cosine form.
\[ f(x) = 2\cos\left(2x + \frac{\pi}{3}\right) \]
Figure (svg): Three curves on one set of axes: a faint wave of height one, a second faint wave of height about one point seven, and their sum drawn thickly, a clean wave of height two with the same period as both.
Check by expanding the answer
Why: Run the sum identity forwards and the original formula must reappear.
\[ 2\left[\cos(2x)\tfrac12 - \sin(2x)\tfrac{\sqrt3}{2}\right] = \cos(2x) - \sqrt3\,\sin(2x) \;\checkmark \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 798 — Example 10.5.3, part 1
The strategy is one you have used before in algebra: two expressions are equal for every value of the variable, so the coefficients of matching pieces must be equal. Here the matching pieces are the cosine term and the sine term.
That gives two equations in two unknowns, but they are not linear and you cannot just substitute. The way through is to square both and add, because the squares of the cosine and the sine of the same angle always add to one. The unknown angle vanishes and the amplitude drops out.
With the amplitude known, each original equation becomes a statement about the cosine and the sine of the phase. You need both to identify the angle, because either one alone leaves two candidates. Here cosine is a half and sine is root three over two, and the only angle in one turn with both is a third of pi.
The check runs the sum identity forwards and lands back on the original expression. That is a real verification, not a restatement, and the picture supports it: the thick curve, the sum of the two faint ones, is a clean wave of height two with the same period as its parts.
Concept
The two coefficient equations look like a dead end, because they involve both unknowns. Multiplying the Pythagorean identity by the square of the amplitude is what unlocks them.
\[ \cos^2\varphi + \sin^2\varphi = 1 \;\Longrightarrow\; A^2\cos^2\varphi + A^2\sin^2\varphi = A^2 \]
The left side is the sum of the squares of the two known coefficients, so the amplitude comes out immediately, with no reference to phi at all.
\[ a\cos(\omega x) + b\sin(\omega x) \quad\text{has amplitude}\quad \sqrt{a^2 + b^2} \]
Geometrically the pair of coefficients is a point in the plane, and the amplitude is simply its distance from the origin. The phase is the angle that point makes, which is why finding it needs the signs of both coefficients and not just their sizes.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 810 — Exercise 36
This slide is short but it is the conceptual heart of the last part of the lesson. The two coefficient equations look unsolvable because the two unknowns are multiplied together in both of them.
Multiplying the Pythagorean identity through by the square of the amplitude is the move that breaks the deadlock. It produces an equation whose left side is exactly the sum of the squares of the two known coefficients, and whose right side is the square of the amplitude alone. The phase has been eliminated without ever being found.
The geometric reading is worth carrying away. Treat the two coefficients as the coordinates of a point in the plane. Then the amplitude is the distance from the origin to that point, and the phase is the angle that point makes in standard position. That is why you need the signs of both coefficients to pin the phase down, and why you can predict which quadrant the phase lies in before doing any work at all.
Worked example
The same function, now matched against the expanded sine form. Watch which coefficient goes with which equation.
\[ \cos(2x) - \sqrt3\,\sin(2x) = A\sin(2x)\cos\varphi + A\cos(2x)\sin\varphi \]
Equate the coefficients, carefully
Why: In the sine form it is the sine term that carries cosine phi, and the cosine term that carries sine phi.
\[ A\cos\varphi = -\sqrt3, \qquad A\sin\varphi = 1 \]
Find A the same way
Why: Square and add.
\[ A^2 = (-\sqrt3)^2 + 1^2 = 4 \;\Longrightarrow\; A = 2 \]
Find phi from the signs
Why: Cosine negative and sine positive puts the angle in the second quadrant, with reference angle a sixth of pi.
\[ \cos\varphi = -\frac{\sqrt3}{2}, \; \sin\varphi = \frac12 \;\Longrightarrow\; \varphi = \frac{5\pi}{6} \]
Write the answer
Why: Same amplitude, same frequency, different phase.
\[ f(x) = 2\sin\left(2x + \frac{5\pi}{6}\right) \]
Check by expanding, and against the cosine version
Why: Expanding reproduces the original, and the two phases differ by exactly a quarter turn, which is the general relation of Exercise 35.
\[ 2\left[\sin(2x)\left(-\tfrac{\sqrt3}{2}\right) + \cos(2x)\tfrac12\right] = \cos(2x) - \sqrt3\,\sin(2x) \;\checkmark \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 799 — Example 10.5.3, part 2
The same function, the same method, but the bookkeeping changes and that is the whole difficulty of this slide. In the sine expansion it is the sine term that carries the cosine of the phase, and the cosine term that carries the sine of the phase.
Write the expanded form out and line the terms up before equating anything. If you match them by position rather than by name you will get the two equations the wrong way round, the phase will come out in the wrong quadrant, and the check at the end will fail.
The amplitude comes out the same as before, which makes sense: it is the height of the curve, and the curve has not changed. Only the phase differs, and it differs by exactly a quarter turn from the cosine version. That relationship is general, and it is Exercise 35 in the book: the same curve written as a sine and as a cosine has phases a quarter turn apart.
Sorting
Sort into buckets
Sort each formula by whether it can be written as a single stretched, shifted cosine.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 799 — the remark on matching arguments
Sort all six before revealing. The test to apply is whether the expression can be written as a single stretched, shifted cosine, not whether it is periodic. Every one of these is periodic.
Three of them are the standard case: a cosine and a sine at one common frequency, possibly with a constant added, so the coefficient-matching method applies directly. Two of them mix different frequencies and are out.
The product of a sine and a cosine is the interesting one, and it is worth pausing on. It looks nothing like the standard shape, but the double-angle identity from the last section turns it into half the sine of twice the angle, which is a plain sinusoid. So the honest test is not whether an expression already looks like the standard form, but whether identities can bring it there.
Trap
The method seems to run just as well here:
\[ \cos(2x) - \sqrt3\,\sin(3x) \]
\[ A^2 = 1 + 3 = 4 \]
\[ \Longrightarrow\; 2\cos\left(2x + \tfrac{\pi}{3}\right)? \]
Wrong. The two frequencies are different.
The coefficient matching only works when both terms carry the same expression inside. Here one is at double the frequency and the other at triple, so there is no common argument to factor out, and no single wave can describe the result.
Figure (svg): A repeating but lumpy curve with unequal humps, drawn across two of its periods, quite unlike a smooth single wave.
The curve is still periodic, with period two pi, but the humps within one period are different sizes. A sinusoid never does that.
The wrong version is tempting because the arithmetic all works. Squaring and adding the coefficients gives a perfectly good number, and the angle exists. Nothing in the calculation complains.
What has been skipped is the step where the two terms are supposed to share a common inside expression that can be factored out. Here one term runs at twice the speed and the other at three times, so there is no common expression, and the coefficient-matching argument never gets started.
Look at the picture to see what the sum actually does. It repeats, so it is periodic, but within one repeat the humps are different heights and different widths. A sinusoid is a single shape repeated exactly; this is two shapes repeated exactly. The book points out in a footnote that the curve does look wave-like, which is precisely why you have to check the arguments match rather than trusting the picture.
Worked example
Write this as a single cosine and as a single sine, taking the phase between zero and two pi.
\[ f(x) = \sqrt2\,\sin(x) + \sqrt2\,\cos(x) + 1 \]
Match against the cosine form
Why: Omega is 1 and the constant term is B, so B is 1.
\[ A\cos\varphi = \sqrt2, \qquad -A\sin\varphi = \sqrt2 \;\Longrightarrow\; A\sin\varphi = -\sqrt2 \]
Find the amplitude
Why: Square and add the two right-hand sides.
\[ A^2 = (\sqrt2)^2 + (-\sqrt2)^2 = 4 \;\Longrightarrow\; A = 2 \]
Find the cosine phase
Why: Cosine positive and sine negative is the fourth quadrant.
\[ \cos\varphi = \frac{\sqrt2}{2}, \; \sin\varphi = -\frac{\sqrt2}{2} \;\Longrightarrow\; \varphi = \frac{7\pi}{4} \]
Match against the sine form
Why: Now the sine term carries cosine phi and the cosine term carries sine phi; both are positive, so the angle is in the first quadrant.
\[ A\cos\varphi = \sqrt2, \; A\sin\varphi = \sqrt2 \;\Longrightarrow\; \varphi = \frac{\pi}{4} \]
Write both answers
Why: Same amplitude, same baseline, two phases.
\[ C(x) = 2\cos\left(x + \frac{7\pi}{4}\right) + 1, \qquad S(x) = 2\sin\left(x + \frac{\pi}{4}\right) + 1 \]
Check both at zero
Why: The original gives root two plus one; so does each answer, since the cosine of seven quarters pi and the sine of a quarter pi are both root two over two.
\[ \sqrt2 + 1 \approx 2.4142 = 2\left(\tfrac{\sqrt2}{2}\right) + 1 \;\checkmark \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, p. 809 — Exercise 25
This one adds a constant on the end, which is the baseline, and it is the easiest of all four numbers: it comes across unchanged and plays no part in the coefficient matching.
Watch the sign in the first step. In the cosine expansion the sine term appears with a minus sign in front, so a positive sine coefficient in the original forces a negative value for the amplitude times the sine of the phase. Getting that sign right is what puts the phase in the fourth quadrant rather than the first.
The question asks for the phase between zero and a full turn, which is why the cosine answer is seven quarters of pi rather than minus a quarter of pi. Those are the same angle, and if you wrote the negative version you are not wrong, only outside the requested range. Add a full turn and you are there.
Finally, notice the relationship between the two answers: the sine phase is a quarter of pi and the cosine phase is seven quarters, and those differ by a half pi, which is the quarter-turn rule again.
Worked example
Write this in both standard forms, with the phase between zero and two pi.
\[ f(x) = 2\sqrt3\,\cos(x) - 2\sin(x) \]
Match against the cosine form
Why: Compare the coefficient of cosine x, then of sine x.
\[ A\cos\varphi = 2\sqrt3, \qquad -A\sin\varphi = -2 \;\Longrightarrow\; A\sin\varphi = 2 \]
Find the amplitude
Why: Square and add.
\[ A^2 = (2\sqrt3)^2 + 2^2 = 12 + 4 = 16 \;\Longrightarrow\; A = 4 \]
Find the cosine phase
Why: Both coordinates positive is the first quadrant.
\[ \cos\varphi = \frac{\sqrt3}{2}, \; \sin\varphi = \frac12 \;\Longrightarrow\; \varphi = \frac{\pi}{6}, \quad C(x) = 4\cos\left(x + \frac{\pi}{6}\right) \]
Match against the sine form
Why: The roles of the two coefficients swap, so cosine phi is now negative while sine phi is positive: the second quadrant.
\[ A\cos\varphi = -2, \; A\sin\varphi = 2\sqrt3 \;\Longrightarrow\; \varphi = \frac{2\pi}{3}, \quad S(x) = 4\sin\left(x + \frac{2\pi}{3}\right) \]
Check the quarter-turn relation and a value
Why: The two phases differ by exactly a half pi, as Exercise 35 predicts, and both forms give two root three at zero, matching the original.
\[ \frac{2\pi}{3} - \frac{\pi}{2} = \frac{\pi}{6} \;\checkmark \qquad f(0) = 2\sqrt3 \approx 3.4641 = 4\cos\left(\tfrac{\pi}{6}\right) \;\checkmark \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.5 Graphs of the Trigonometric Functions §10.5, pp. 809-810 — Exercises 29 and 35
By now the procedure should feel routine, so use this one to build speed. Match, square and add, then read the quadrant off the two signs.
The quadrant reasoning is the part worth slowing down for. In the cosine form both the cosine and the sine of the phase come out positive, which puts the phase in the first quadrant. In the sine form the roles of the two coefficients swap, the cosine of the phase turns negative, and the phase moves to the second quadrant. Nothing about the curve has changed; only the description has.
The check does two things at once. It confirms that the two phases differ by a quarter turn, which is the general relation from Exercise 35, and it evaluates both forms at a single convenient input and compares with the original. If a fitting answer passes both of those tests, you can stop worrying about it.
Section
Part 6
Pattern
Figure (svg): A flow diagram: start from a formula or from a graph. A formula is put into standard form and the four parameters are read off; a sum of a cosine and a sine at one frequency is first combined; a graph gives the amplitude from half the spread and the baseline from the average. All routes meet at the quarter marks, which give the picture.
This diagram is the whole lesson on one slide, and it is arranged by what you are given rather than by what you have to do, which is how exam questions actually arrive.
The three entry points on the left are the three kinds of question. A formula, a graph, or a sum of two waves. Each has its own short preparation step, and all three funnel into the same middle box: a formula in standard form with four known numbers.
From there the route is fixed. Read the four quantities, then set the argument equal to the five quarter marks and solve if a drawing is wanted. The one step people skip is forcing the coefficient of x positive, and skipping it corrupts the phase shift every time. Put it in your written procedure so that it cannot be forgotten under pressure.
Check
Check your understanding
For one cycle of y = 2 sin(4x + π), the quarter-mark equations give which first and last values of x?
Answer: A
Why: Set the argument 4x + π equal to 0: x = −π/4. Set it equal to 2π: x = π/4. The cycle is therefore π/2 wide, which agrees with the period 2π divided by ω = 4.
This question is only about the first and last of the five solves, because those two are what determine where the cycle sits and how wide it is.
Set the inside expression equal to zero and then to a full turn, and solve each. The two answers are a quarter of pi apart on either side of the origin, which makes the cycle a half pi wide. That agrees with the period found from the formula, and the agreement is the check you should always run.
The most tempting wrong answer ignores the added constant and gives the cycle of an unshifted wave. If you picked that one, go back and notice that the constant inside is what moves the cycle, and that here it moves it a quarter of pi to the left.
Check
Check your understanding
What is the phase shift of y = 5 cos(2x + π/3) − 1?
Answer: B
Why: The shift is −φ/ω = −(π/3)/2 = −π/6, and a negative shift means a move to the left. Factoring the 2 out of the argument shows the same thing: 2(x + π/6).
Two things have to be right here: the size of the shift and its direction. Three of the four options get exactly one of them right, which is why guessing will not help.
Divide the constant inside by the coefficient of x, and then negate. The constant is a third of pi, the coefficient is two, so the size is a sixth of pi; the constant is positive, so the shift is negative, which means to the left.
If the formula is not sticking, factor instead. Pulling the two out of the argument turns the inside into x plus a sixth of pi, and adding to the input always moves a graph to the left. The factoring version is slower but it is impossible to get backwards, so use it whenever you are unsure.
Check
Check your understanding
What is the amplitude of f(x) = 3 sin(x) + 3√3 cos(x)?
Answer: A
Why: The amplitude of a cos(ωx) + b sin(ωx) is the square root of a² + b². Here that is the square root of 27 + 9, which is the square root of 36, namely 6.
The instinct to add the two coefficients is strong and it is wrong, so it is worth understanding why rather than just remembering not to.
Adding amplitudes would be correct if the two waves reached their crests at the same moment, because then the peaks would stack. A sine and a cosine at the same frequency never do: one is at its crest exactly where the other is crossing its middle. So their sum is taller than either one but shorter than the two added together.
The right answer comes from squaring and adding, which is the distance formula in disguise. Treat the two coefficients as the legs of a right triangle and the amplitude is the hypotenuse. Here the legs are 3 and three root three, the squares are 9 and 27, and the hypotenuse is 6.
Explain it to yourself
Discussion prompt
In two or three sentences, explain why setting the expression inside the cosine equal to zero, a half pi, pi, three halves pi and two pi, and then solving for x, produces the five points you need. Say what would go wrong if you set the whole formula equal to those five values instead.
Write two or three sentences before revealing. Being able to say why a method works is the difference between a technique you can apply to a familiar question and one you can adapt to an unfamiliar one.
The key observation is that the cosine has no idea what has been done to x before the result reaches it. It sees one number and responds to that number. So its landmark behaviour, crest, zero, trough, zero, crest, is tied to the value of the argument, not to the value of x.
The second half of the question is the more useful one. If you can say precisely what goes wrong when you set the whole formula equal to the quarter marks, you will never make that mistake, because you will recognise the symptoms: equations that are not linear, and equations with no solution at all.
Exit ticket
One cycle of a sinusoid begins at a maximum of 9 when x is 2, drops to a minimum of minus 3, and returns to 9 when x is 10.
Discussion prompt
Find a cosine formula for this curve, and check it at both ends of the cycle and at the trough.
Do this one on paper before revealing anything. It uses all four parts of the lesson and nothing else, and it should take about three minutes.
Start with the two vertical readings, because half the difference and half the sum give you the amplitude and the baseline immediately. Then use the width of the cycle for the coefficient of x, and the position of the crest for the phase shift. Only then assemble the formula.
Check at three places rather than one: both ends of the cycle and the trough halfway along. At each of them the argument should come out as a landmark value, so you can evaluate without a calculator. If all three agree with the description, your formula is right, whatever it looks like compared with anyone else's.
Recap
| given | what to do | what you get |
|---|---|---|
| a formula | put it in standard form with ω positive | amplitude |A|, period 2π/ω, shift −φ/ω, baseline B |
| a graph | half the spread, average of the extremes, width of a cycle | A, B, ω, and the shift from where a cycle starts |
| a cosine plus a sine, same ω | match coefficients, square and add | amplitude √(a² + b²), phase from the signs |
| any of the above | set the argument to the five quarter marks and solve | the five points, and the drawing |
\[ C(x) = A\cos(\omega x + \varphi) + B \]
\[ \text{period } \frac{2\pi}{\omega}, \quad \text{amplitude } |A|, \quad \text{shift } -\frac{\varphi}{\omega}, \quad \text{baseline } B \]
Next, the same graphing question for the secant and the cosecant, where the reciprocal of a value that passes through zero forces vertical asymptotes into the picture.
OpenStax Algebra and Trigonometry 2e, §8.1 Graphs of the Sine and Cosine Functions §8.1 — the same material, with an emphasis on modelling
Four numbers describe every sinusoid there is, and this table is a map of how to get them from whatever you happen to be given. A formula needs standard form first. A graph needs half the spread and the average of the extremes. A cosine plus a sine at one frequency needs the coefficients squared and added.
Two habits are worth carrying out of this lesson. Always force the coefficient of x positive before quoting anything, and always divide the constant inside by that coefficient before quoting a shift. Those two steps between them prevent the great majority of the errors this section generates.
Next comes the same graphing question for the secant and the cosecant. Those are reciprocals of cosine and sine, so wherever cosine or sine passes through zero, its reciprocal blows up, and the graphs acquire vertical asymptotes. The four parameters survive in a modified form, but the amplitude does not, because a curve that runs off to infinity has no height to measure.
Want this taught 1-on-1? Alexander tutors Trigonometry — $55/session, free consultation.