Unit 5, lesson 61

Compositions Involving Inverse Trig Functions

How to evaluate a trig function of an inverse trig value exactly with a reference triangle, why undoing in the other order can fail, and how principal ranges fix every sign.

  • Theory part 121
  • Practice part 122
  • Inverse Trigonometric Functions

A ramp climbs 3 meters for every 4 meters of horizontal run, so its angle is θ=arctan⁡34\theta = \arctan \frac{3}{4}. You never need that angle in degrees to know its sine. The right triangle with legs 3 and 4 has hypotenuse 5, so sin⁡θ=35\sin \theta = \frac{3}{5} exactly. Expressions that put a trigonometric function and an inverse trigonometric function together come up often in calculus, and the main difficulty is the restricted range each inverse function returns.

Key terms

Principal ranges

Definition Each inverse trigonometric function returns exactly one angle, taken from a fixed set of angles called its principal range. In radians,

FunctionDomainPrincipal range
arcsin⁡x\arcsin x−1≤x≤1-1 \le x \le 1[−π2,π2]\left[-\frac{\pi}{2}, \frac{\pi}{2}\right]
arccos⁡x\arccos x−1≤x≤1-1 \le x \le 1[0,π][0, \pi]
arctan⁡x\arctan xall real xx(−π2,π2)\left(-\frac{\pi}{2}, \frac{\pi}{2}\right)
arcsec⁡x\operatorname{arcsec} x∣x∣≥1\lvert x \rvert \ge 1[0,π2)∪(π2,π]\left[0, \frac{\pi}{2}\right) \cup \left(\frac{\pi}{2}, \pi\right]
arccsc⁡x\operatorname{arccsc} x∣x∣≥1\lvert x \rvert \ge 1[−π2,0)∪(0,π2]\left[-\frac{\pi}{2}, 0\right) \cup \left(0, \frac{\pi}{2}\right]

So θ=arcsin⁡x\theta = \arcsin x means both sin⁡θ=x\sin \theta = x and −π2≤θ≤π2-\frac{\pi}{2} \le \theta \le \frac{\pi}{2}, and the same pattern holds for the others. The ranges for arcsec and arccsc are the convention used in this course, and with them, for every xx with ∣x∣≥1\lvert x \rvert \ge 1,

arcsec⁡x=arccos⁡1x,arccsc⁡x=arcsin⁡1x.\begin{gathered} \operatorname{arcsec} x = \arccos \frac{1}{x}, \\[0.8em] \operatorname{arccsc} x = \arcsin \frac{1}{x}. \end{gathered}

Some books choose other ranges for these two.

In plain words An equation like sin⁡θ=12\sin \theta = \frac{1}{2} has infinitely many solutions, and the inverse function picks exactly one of them. The principal range is the rule that says which one, and it tells you the quadrant of the answer before you compute anything.

Example The angles with sine −12-\frac{1}{2} include −π6-\frac{\pi}{6}, 7π6\frac{7\pi}{6} and 11π6\frac{11\pi}{6}. Only −π6-\frac{\pi}{6} lies in [−π2,π2]\left[-\frac{\pi}{2}, \frac{\pi}{2}\right], so arcsin⁡(−12)=−π6\arcsin\left(-\frac{1}{2}\right) = -\frac{\pi}{6}. For cosine the answer looks different. The range [0,π][0, \pi] contains no negative angles, so arccos⁡(−12)=2π3\arccos\left(-\frac{1}{2}\right) = \frac{2\pi}{3}, an angle in the second quadrant.

Trig function after its inverse

Definition A composition of a trigonometric function with an inverse trigonometric function applies one after the other. When the inverse acts first,

sin⁡(arcsin⁡x)=xandcos⁡(arccos⁡x)=xfor every x in [−1,1],\begin{gathered} \sin(\arcsin x) = x \\[0.4em] \text{and} \quad \cos(\arccos x) = x \\[0.4em] \text{for every } x \text{ in } [-1, 1], \end{gathered}

and

tan⁡(arctan⁡x)=xfor every real x.\begin{gathered} \tan(\arctan x) = x \\[0.4em] \text{for every real } x. \end{gathered}

For ∣x∣>1\lvert x \rvert > 1 the expressions sin⁡(arcsin⁡x)\sin(\arcsin x) and cos⁡(arccos⁡x)\cos(\arccos x) are undefined, because xx is not in the domain of the inner function.

In plain words The inverse function finds an angle with the given sine, cosine or tangent, and the outer function reads that value back. This order always returns the input, as long as the inner function accepts it.

Example Reading back gives cos⁡(arccos⁡(−23))=−23\cos\left(\arccos\left(-\frac{2}{3}\right)\right) = -\frac{2}{3} and tan⁡(arctan⁡50)=50\tan(\arctan 50) = 50. On the other hand sin⁡(arcsin⁡2)\sin(\arcsin 2) has no value at all, since no real angle has sine 22.

Inverse after the trig function

Definition The equations

arcsin⁡(sin⁡x)=x,arccos⁡(cos⁡x)=x,arctan⁡(tan⁡x)=x\begin{gathered} \arcsin(\sin x) = x, \\[0.4em] \arccos(\cos x) = x, \\[0.4em] \arctan(\tan x) = x \end{gathered}

hold exactly when xx lies in the principal range of the inverse function, that is, for −π2≤x≤π2-\frac{\pi}{2} \le x \le \frac{\pi}{2}, for 0≤x≤π0 \le x \le \pi and for −π2<x<π2-\frac{\pi}{2} < x < \frac{\pi}{2} respectively. For any other xx at which the expression is defined, its value is the one angle in the principal range that has the same sine, cosine or tangent as xx.

In plain words Taking the sine forgets which of many angles you started with. The inverse function can only return the angle in its own range, so you get xx back only when xx was already there.

Example The angle 5π6\frac{5\pi}{6} lies outside [−π2,π2]\left[-\frac{\pi}{2}, \frac{\pi}{2}\right], and its sine is 12\frac{1}{2}. So

arcsin⁡(sin⁡5π6)=arcsin⁡12=π6.\arcsin\left(\sin \frac{5\pi}{6}\right) = \arcsin \frac{1}{2} = \frac{\pi}{6}.

In the same way tan⁡3π4=−1\tan \frac{3\pi}{4} = -1, which gives arctan⁡(tan⁡3π4)=−π4\arctan\left(\tan \frac{3\pi}{4}\right) = -\frac{\pi}{4}.

Exact evaluation

Definition Exact evaluation gives a value as an exact expression, such as a fraction, a radical or a rational multiple of π\pi, rather than as a rounded decimal. To evaluate a composition such as cos⁡(arcsin⁡x)\cos(\arcsin x) exactly, name the inner angle θ=arcsin⁡x\theta = \arcsin x and write down the two facts it carries,

sin⁡θ=xand−π2≤θ≤π2.\sin \theta = x \quad \text{and} \quad -\frac{\pi}{2} \le \theta \le \frac{\pi}{2}.

Then compute the outer function of θ\theta from those facts, by recognizing a special angle, by a Pythagorean identity or by a reference triangle.

In plain words Give the inner angle a name. The inverse function tells you one trigonometric value of that angle and the interval it lives in, and that is enough to find any other value.

Example Let θ=arccos⁡(−22)\theta = \arccos\left(-\frac{\sqrt{2}}{2}\right). Then cos⁡θ=−22\cos \theta = -\frac{\sqrt{2}}{2} with 0≤θ≤π0 \le \theta \le \pi, so θ=3π4\theta = \frac{3\pi}{4} and

sin⁡(arccos⁡(−22))=sin⁡3π4=22.\begin{aligned} \sin\left(\arccos\left(-\frac{\sqrt{2}}{2}\right)\right) &= \sin \frac{3\pi}{4} \\[0.8em] &= \frac{\sqrt{2}}{2}. \end{aligned}

No special angle is needed for cos⁡(arcsin⁡13)\cos\left(\arcsin \frac{1}{3}\right). With θ=arcsin⁡13\theta = \arcsin \frac{1}{3} the identity gives cos⁡2θ=1−19=89\cos^2 \theta = 1 - \frac{1}{9} = \frac{8}{9}, and cosine is never negative on [−π2,π2]\left[-\frac{\pi}{2}, \frac{\pi}{2}\right], so

cos⁡(arcsin⁡13)=89=223.\cos\left(\arcsin \frac{1}{3}\right) = \sqrt{\frac{8}{9}} = \frac{2\sqrt{2}}{3}.

Reference triangle

Definition When the inverse function returns an angle θ\theta with 0<θ<π20 < \theta < \frac{\pi}{2}, which happens for arcsin⁡x\arcsin x and arccos⁡x\arccos x with 0<x<10 < x < 1 and for arctan⁡x\arctan x with x>0x > 0, the angle θ\theta is an acute angle of a right triangle. A reference triangle for θ\theta is a right triangle whose sides match the known ratio, for example opposite side aa and hypotenuse cc when sin⁡θ=ac\sin \theta = \frac{a}{c}. The Pythagorean theorem gives the third side, and every other trigonometric value of θ\theta is a ratio of two sides.

In plain words Turn the input into a triangle. A sine of 27\frac{2}{7} means opposite side 2 and hypotenuse 7, and once the third side is known you can read off anything.

Example Let θ=arcsin⁡27\theta = \arcsin \frac{2}{7}. The input is positive, so θ\theta is acute. Draw a right triangle with opposite side 22 and hypotenuse 77. The adjacent side is 49−4=35\sqrt{49 - 4} = 3\sqrt{5}, and so

tan⁡(arcsin⁡27)=235=2515.\tan\left(\arcsin \frac{2}{7}\right) = \frac{2}{3\sqrt{5}} = \frac{2\sqrt{5}}{15}.

Signs from the quadrant

Definition The principal range places the angle θ\theta in a quadrant decided by the sign of the input.

AngleInput between 00 and 11Input between −1-1 and 00
θ=arcsin⁡x\theta = \arcsin xQuadrant IQuadrant IV, −π2<θ<0-\frac{\pi}{2} < \theta < 0
θ=arccos⁡x\theta = \arccos xQuadrant IQuadrant II, π2<θ<π\frac{\pi}{2} < \theta < \pi

Arctan follows the arcsin pattern for inputs of any size. Every positive input gives an angle in Quadrant I and every negative input an angle in Quadrant IV. The inputs 00 and ±1\pm 1 of arcsin and arccos, and the input 00 of arctan, give angles on an axis, such as arcsin⁡(−1)=−π2\arcsin(-1) = -\frac{\pi}{2} and arccos⁡(−1)=π\arccos(-1) = \pi. For a negative input, build the reference triangle from the absolute value of the input, find the size of the answer from it, and then give the answer the sign its function has in that quadrant.

In plain words The triangle gives the size of the answer and the quadrant gives the sign. Arcsin and arctan of a negative number point into Quadrant IV, while arccos of a negative number points into Quadrant II. The one exception is the input −1-1 for arcsin and arccos, which lands on an axis.

Example Let θ=arccos⁡(−35)\theta = \arccos\left(-\frac{3}{5}\right), an angle in Quadrant II. The reference triangle has adjacent side 3, hypotenuse 5 and opposite side 4. In Quadrant II sine is positive and tangent is negative, so

sin⁡θ=45,tan⁡(arccos⁡(−35))=−43.\begin{gathered} \sin \theta = \frac{4}{5}, \\[0.8em] \tan\left(\arccos\left(-\frac{3}{5}\right)\right) = -\frac{4}{3}. \end{gathered}

Compare φ=arcsin⁡(−35)\varphi = \arcsin\left(-\frac{3}{5}\right), which lies in Quadrant IV. There cosine is positive, so cos⁡φ=45\cos \varphi = \frac{4}{5}.

Algebraic expressions in x

Definition A composition of a trigonometric function with an inverse trigonometric function of a variable can be written without trigonometry. For every xx in [−1,1][-1, 1],

cos⁡(arcsin⁡x)=1−x2,sin⁡(arccos⁡x)=1−x2,\begin{gathered} \cos(\arcsin x) = \sqrt{1 - x^2}, \\[0.4em] \sin(\arccos x) = \sqrt{1 - x^2}, \end{gathered}

for every xx in (−1,1)(-1, 1),

tan⁡(arcsin⁡x)=x1−x2,\tan(\arcsin x) = \frac{x}{\sqrt{1 - x^2}},

and for every real xx,

sin⁡(arctan⁡x)=x1+x2,cos⁡(arctan⁡x)=11+x2.\begin{gathered} \sin(\arctan x) = \frac{x}{\sqrt{1 + x^2}}, \\[0.8em] \cos(\arctan x) = \frac{1}{\sqrt{1 + x^2}}. \end{gathered}

In plain words These are the reference triangle answers with a letter in place of a number. The principal ranges make the same formulas correct for negative xx too, so no separate cases are needed.

Example Take θ=arctan⁡x\theta = \arctan x. For x>0x > 0 draw a triangle with opposite side xx and adjacent side 11, so the hypotenuse is 1+x2\sqrt{1 + x^2} and sin⁡θ=x1+x2\sin \theta = \frac{x}{\sqrt{1 + x^2}}. For x<0x < 0 the angle is in Quadrant IV, where sine is negative, and the formula is negative too. At x=−1x = -1 it gives

sin⁡(arctan⁡(−1))=−12=−22,\sin(\arctan(-1)) = \frac{-1}{\sqrt{2}} = -\frac{\sqrt{2}}{2},

which matches sin⁡(−π4)\sin\left(-\frac{\pi}{4}\right).

Absolute values

Definition For every real number uu,

u2=∣u∣.\sqrt{u^2} = \lvert u \rvert.

So the Pythagorean identities give, wherever both sides are defined,

1−sin⁡2θ=∣cos⁡θ∣,1+tan⁡2θ=∣sec⁡θ∣,sec⁡2θ−1=∣tan⁡θ∣.\begin{gathered} \sqrt{1 - \sin^2 \theta} = \lvert \cos \theta \rvert, \\[0.4em] \sqrt{1 + \tan^2 \theta} = \lvert \sec \theta \rvert, \\[0.4em] \sqrt{\sec^2 \theta - 1} = \lvert \tan \theta \rvert. \end{gathered}

The absolute value bars may be dropped only when the range of θ\theta guarantees that the quantity inside them is not negative.

In plain words A square root never comes out negative, so taking the root of a square loses the sign. You get the sign back from the principal range, not from the algebra.

Example Let θ=arcsec⁡x\theta = \operatorname{arcsec} x with ∣x∣≥1\lvert x \rvert \ge 1. Then sec⁡θ=x\sec \theta = x and tan⁡2θ=x2−1\tan^2 \theta = x^2 - 1, so ∣tan⁡θ∣=x2−1\lvert \tan \theta \rvert = \sqrt{x^2 - 1}. Tangent is not negative on [0,π2)\left[0, \frac{\pi}{2}\right) and not positive on (π2,π]\left(\frac{\pi}{2}, \pi\right], so

tan⁡(arcsec⁡x)={x2−1if x≥1,−x2−1if x≤−1.\begin{aligned} &\tan(\operatorname{arcsec} x) \\[0.4em] &\quad = \begin{cases} \sqrt{x^2 - 1} & \text{if } x \ge 1, \\ -\sqrt{x^2 - 1} & \text{if } x \le -1. \end{cases} \end{aligned}

At x=−2x = -2 the angle is arcsec⁡(−2)=2π3\operatorname{arcsec}(-2) = \frac{2\pi}{3}, and indeed tan⁡2π3=−3\tan \frac{2\pi}{3} = -\sqrt{3}.

Branch-dependent simplification

Definition A simplification is branch-dependent when the simplified form changes from one interval of the variable to another, because a principal range or an absolute value is involved. For example, for every integer kk,

arctan⁡(tan⁡x)=x−kπfor −π2+kπ<x<π2+kπ,\begin{gathered} \arctan(\tan x) = x - k\pi \\[0.4em] \text{for } -\frac{\pi}{2} + k\pi < x < \frac{\pi}{2} + k\pi, \end{gathered}

and

arcsin⁡(sin⁡x)={xif −π2≤x≤π2,π−xif π2≤x≤3π2,\begin{aligned} &\arcsin(\sin x) \\[0.4em] &\quad = \begin{cases} x & \text{if } -\frac{\pi}{2} \le x \le \frac{\pi}{2}, \\ \pi - x & \text{if } \frac{\pi}{2} \le x \le \frac{3\pi}{2}, \end{cases} \end{aligned}

with the same pattern repeating every 2π2\pi.

In plain words One formula does not fit every input. Split the line into pieces and, on each piece, find the expression that has the right sine, cosine or tangent and also lands in the principal range.

Example For π2≤x≤3π2\frac{\pi}{2} \le x \le \frac{3\pi}{2} the number π−x\pi - x lies in [−π2,π2]\left[-\frac{\pi}{2}, \frac{\pi}{2}\right] and sin⁡(π−x)=sin⁡x\sin(\pi - x) = \sin x, which is why arcsin⁡(sin⁡x)=π−x\arcsin(\sin x) = \pi - x there. Test it at x=4π3x = \frac{4\pi}{3}. The sine is −32-\frac{\sqrt{3}}{2}, so

arcsin⁡(sin⁡4π3)=arcsin⁡(−32)=−π3=π−4π3.\begin{aligned} \arcsin\left(\sin \frac{4\pi}{3}\right) &= \arcsin\left(-\frac{\sqrt{3}}{2}\right) \\[0.8em] &= -\frac{\pi}{3} = \pi - \frac{4\pi}{3}. \end{aligned}

Check yourself

  1. Find the exact value of arccos⁡(cos⁡7π6)\arccos\left(\cos \frac{7\pi}{6}\right).
Show the answer

The cosine of 7π6\frac{7\pi}{6} is −32-\frac{\sqrt{3}}{2}. The angle in [0,π][0, \pi] with this cosine is 5π6\frac{5\pi}{6}, so

arccos⁡(cos⁡7π6)=5π6.\arccos\left(\cos \frac{7\pi}{6}\right) = \frac{5\pi}{6}.
  1. Find sec⁡(arctan⁡(−512))\sec\left(\arctan\left(-\frac{5}{12}\right)\right) and sin⁡(arctan⁡(−512))\sin\left(\arctan\left(-\frac{5}{12}\right)\right).
Show the answer

The angle θ=arctan⁡(−512)\theta = \arctan\left(-\frac{5}{12}\right) lies in Quadrant IV. Its reference triangle has opposite side 5, adjacent side 12 and hypotenuse 13. In Quadrant IV cosine and secant are positive and sine is negative, so

sec⁡θ=1312,sin⁡θ=−513.\sec \theta = \frac{13}{12}, \qquad \sin \theta = -\frac{5}{13}.
  1. Write tan⁡(arccos⁡x)\tan(\arccos x) as an algebraic expression in xx, and say for which xx it holds.
Show the answer

Let θ=arccos⁡x\theta = \arccos x, so cos⁡θ=x\cos \theta = x and 0≤θ≤π0 \le \theta \le \pi. Sine is not negative on [0,π][0, \pi], so sin⁡θ=1−x2\sin \theta = \sqrt{1 - x^2}, and

tan⁡(arccos⁡x)=1−x2x,−1≤x≤1,  x≠0.\begin{gathered} \tan(\arccos x) = \frac{\sqrt{1 - x^2}}{x}, \\[0.8em] -1 \le x \le 1, \; x \neq 0. \end{gathered}

At x=0x = 0 the angle is π2\frac{\pi}{2}, where tangent is undefined. For −1<x<0-1 < x < 0 the formula is negative, as it should be in Quadrant II.

How this lesson connects

Where this leads in Calculus