2.12: Inverse Trigonometric Functions
( \newcommand{\kernel}{\mathrm{null}\,}\)
One very useful application of implicit differentiation is to find the derivatives of inverse functions. We have already used this approach to find the derivative of the inverse of the exponential function — the logarithm.
We are now going to consider the problem of finding the derivatives of the inverses of trigonometric functions. Now is a very good time to go back and reread Section 0.6 on inverse functions — especially Definition 0.6.4. Most importantly, given a function f(x), its inverse function f−1(x) only exists, with domain D, when f(x) passes the “horizontal line test”, which says that for each Y in D the horizontal line y=Y intersects the graph y=f(x) exactly once. (That is, f(x) is a one-to-one function.)
Let us start by playing with the sine function and determine how to restrict the domain of sinx so that its inverse function exists.
Let y=f(x)=sin(x). We would like to find the inverse function which takes y and returns to us a unique x-value so that sin(x)=y.
- For each real number Y, the number of x-values that obey sin(x)=Y, is exactly the number of times the horizontal straight line y=Y intersects the graph of sin(x).
- When −1≤Y≤1, the horizontal line intersects the graph infinitely many times. This is illustrated in the figure above by the line y=0.3.
- On the other hand, when Y<−1 or Y>1, the line y=Y never intersects the graph of sin(x). This is illustrated in the figure above by the line y=−1.2.
This is exactly the horizontal line test and it shows that the sine function is not one-to-one.
Now consider the function
y=sin(x)with domain −π2≤x≤π2
This function has the same formula but the domain has been restricted so that, as we'll now show, the horizontal line test is satisfied.
As we saw above when |Y|>1 no x obeys sin(x)=Y and, for each −1≤Y≤1, the line y=Y (illustrated in the figure above with y=0.3) crosses the curve y=sin(x) infinitely many times, so that there are infinitely many x's that obey f(x)=sinx=Y. However exactly one of those crossings (the dot in the figure) has −π2≤x≤π2.
That is, for each −1≤Y≤1, there is exactly one x, call it X, that obeys both
sinX=Yand−π2≤X≤π2
That unique value, X, is typically denoted arcsin(Y). That is
sin(arcsin(Y))=Yand−π2≤arcsin(Y)≤π2
Renaming Y→x, the inverse function arcsin(x) is defined for all −1≤x≤1 and is determined by the equation
sin(arcsin(x))=xand−π2≤arcsin(x)≤π2.
Note that many texts will use sin−1(x) to denote arcsine, however we will use arcsin(x) since we feel that it is clearer 1; the reader should recognise both.
Since
sinπ2=1sinπ6=12
and −π2≤π6,π2≤π2, we have
arcsin1=π2arcsin12=π6
Even though
sin(2π)=0
it is not true that arcsin0=2π, and it is not true that arcsin(sin(2π))=2π, because 2π is not between −π2 and π2. More generally
arcsin(sin(x))= the unique angle θ between −π2 and π2 obeying sinθ=sinx=xif and only if −π2≤x≤π2
So, for example, arcsin(sin(11π16)) cannot be 11π16 because 11π16 is bigger than π2. So how do we find the correct answer? Start by sketching the graph of sin(x).
It looks like the graph of sinx is symmetric about x=π2. The mathematical way to say that “the graph of sinx is symmetric about x=π2” is “sin(π2−θ)=sin(π2+θ)” for all θ. That is indeed true 2.
Now 11π16=π2+3π16 so
sin(11π16)=sin(π2+3π16)=sin(π2−3π16)=sin(5π16)
and, since 5π16 is indeed between −π2 and π2,
arcsin(sin(11π16))=5π16(and not 11π16).
Derivatives of Inverse Trig Functions
Now that we have explored the arcsine function we are ready to find its derivative. Lets call
arcsin(x)=θ(x),
so that the derivative we are seeking is dθdx. The above equation is (after taking sine of both sides) equivalent to
sin(θ)=x
Now differentiate this using implicit differentiation (we just have to remember that θ varies with x and use the chain rule carefully):
cos(θ)⋅dθdx=1dθdx=1cos(θ)substitute θ=arcsinxddxarcsinx=1cos(arcsinx)
This doesn't look too bad, but it's not really very satisfying because the right hand side is expressed in terms of arcsin(x) and we do not have an explicit formula for arcsin(x).
However even without an explicit formula for arcsin(x), it is a simple matter to get an explicit formula for cos(arcsin(x)), which is all we need. Just draw a right–angled triangle with one angle being arcsin(x). This is done in the figure below 3.
Since sin(θ)=x (see 2.12.2), we have made the side opposite the angle θ of length x and the hypotenuse of length 1. Then, by Pythagoras, the side adjacent to θ has length √1−x2 and so
cos(arcsin(x))=cos(θ)=√1−x2
which in turn gives us the answer we need:
ddxarcsin(x)=1√1−x2
The definitions for arccos, arctan and arccot are developed in the same way. Here are the graphs that are used.
The definitions for the remaining two inverse trigonometric functions may also be developed in the same way 4 In fact, there are two different widely used definitions of arcsecx. Under our definition, below, θ=arcsecx takes values in 0≤θ≤π. Some people, perfectly legitimately, define θ=arcsecx to take values in the union of 0≤θ<π2 and π≤θ<3π2. Our definition is sometimes called the “trigonometry friendly” definition. The definition itself has the advantage of simplicity. The other definition is sometimes called the “calculus friendly” definition. It eliminates some absolute values and hence simplifies some computations. Similarly there are two different widely used definitions of arccscx. 5 One could also define arccot(x)=arctan(1/x) with arccot(0)=π2. We have chosen not to do so, because the definition we have chosen is both continuous and standard.. But it's a little easier to use
cscx=1sinxsecx=1cosx
arcsinx is defined for |x|≤1. It is the unique number obeying
sin(arcsin(x))=xand−π2≤arcsin(x)≤π2
arccosx is defined for |x|≤1. It is the unique number obeying
cos(arccos(x))=xand0≤arccos(x)≤πarctanx is defined for all x∈R. It is the unique number obeying
tan(arctan(x))=xand−π2<arctan(x)<π2arccscx=arcsin1x is defined for |x|≥1. It is the unique number obeying
csc(arccsc(x))=xand−π2≤arccsc(x)≤π2\boldsymbol{\ \ \ \ \ \ \ \}Because csc(0) is undefined, arccsc(x) never takes the value 0.
arcsecx=arccos1x is defined for |x|≥1. It is the unique number obeying
sec(arcsec(x))=xand0≤arcsec(x)≤π\boldsymbol{\ \ \ \ \ \ \ \}Because sec(π/2) is undefined, arcsec(x) never takes the value π/2.
arccotx is defined for all x∈R. It is the unique number obeying
cot(arccot(x))=xand0<arccot(x)<πTo find the derivative of arccos we can follow the same steps:
- Write arccos(x)=θ(x) so that cosθ=x and the desired derivative is dθdx.
- Differentiate implicitly, remembering that θ is a function of x:
−sinθdθdx=1dθdx=−1sinθddxarccosx=−1sin(arccosx).
- To simplify this expression, again draw the relevant triangle
from which we see
sin(arccosx)=sinθ=√1−x2.
- Thus
ddxarccosx=−1√1−x2.
Very similar steps give the derivative of arctanx:
- Start with θ=arctanx, so tanθ=x.
- Differentiate implicitly:
sec2θdθdx=1dθdx=1sec2θ=cos2θddxarctanx=cos2(arctanx).
- To simplify this expression, we draw the relevant triangle
from which we see
cos2(arctanx)=cos2θ=11+x2
- Thus
ddxarctanx=11+x2.
An almost identical computation gives the derivative of arccotx:
- Start with θ=arccotx, so cotθ=x.
- Differentiate implicitly:
−csc2θdθdx=1ddxarccotx=dθdx=−1csc2θ=−sin2θ=−11+x2
from the triangle
To find the derivative of arccsc we can use its definition and the chain rule.
θ=arccscxtake cosecant of both sidescscθ=xbut cscθ=1sinθ, so flip both sidessinθ=1xnow take arcsine of both sidesθ=arcsin(1x)
Now just differentiate, carefully using the chain rule :
dθdx=ddxarcsin(1x)=1√1−x−2⋅−1x2
To simplify further we will factor x−2 out of the square root. We need to be a little careful doing that. Take another look at examples 1.5.6 and 1.5.7 and the discussion between them before proceeding.
=1√x−2(x2−1)⋅−1x2=1|x−1|⋅√x2−1⋅−1x2note that x2⋅|x−1|=|x|.=−1|x|√x2−1In the same way we can find the derivative of the remaining inverse trig function. We just use its definition, a derivative we already know and the chain rule.
ddxarcsec(x)=ddxarccos(1x)=−1√1−1x2⋅(−1x2)=1|x|√x2−1
By way of summary, we have
The derivatives of the inverse trigonometric functions are
ddxarcsin(x)=1√1−x2ddxarccsc(x)=−1|x|√x2−1ddxarccos(x)=−1√1−x2ddxarcsec(x)=1|x|√x2−1ddxarctan(x)=11+x2ddxarccot(x)=−11+x2
Exercises
Stage 1
Give the domains of each of the following functions.
(a) f(x)=arcsin(cosx)(b) g(x)=arccsc(cosx)(c) h(x)=sin(arccosx)
A particle starts moving at time t=10, and it bobs up and down, so that its height at time t≥10 is given by cost. True or false: the particle has height 1 at time t=arccos(1).
The curve y=f(x) is shown below, for some function f. Restrict f to the largest possible interval containing 0 over which it is one--to--one, and sketch the curve y=f−1(x).
Let a be some constant. Where does the curve y=ax+cosx have a horizontal tangent line?
Define a function f(x)=arcsinx+arccscx. What is the domain of f(x)? Where is f(x) differentiable?
Stage 2
Differentiate f(x)=arcsin(x3). What is the domain of f(x)?
Differentiate f(t)=arccostt2−1. What is the domain of f(t)?
Differentiate f(x)=arcsec(−x2−2). What is the domain of f(x)?
Differentiate f(x)=1aarctan(xa), where a is a nonzero constant. What is the domain of f(x)?
Differentiate f(x)=xarcsinx+√1−x2. What is the domain of f(x)?
For which values of x is the tangent line to y=arctan(x2) horizontal?
Evaluate ddx{arcsinx+arccosx}.
Find the derivative of y=arcsin(1x).
Find the derivative of y=arctan(1x).
Calculate and simplify the derivative of (1+x2)arctanx.
Show that ddx{sin(arctan(x))}=(x2+1)−3/2.
Show that ddx{cot(arcsin(x))}=−1x2√1−x2.
Determine all points on the curve y=arcsinx where the tangent line is parallel to the line y=2x+9.
For which values of x does the function f(x)=arctan(cscx) have a horizontal tangent line?
Stage 3
Let f(x)=x+cosx, and let g(y)=f−1(y) be the inverse function. Determine g′(y).
f(x)=2x−sin(x) is one--to--one. Find (f−1)′(π−1).
f(x)=ex+x is one--to--one. Find (f−1)′(e+1).
Differentiate f(x)=[sinx+2]arcsecx. What is the domain of this function?
Suppose you can't remember whether the derivative of arcsine is 1√1−x2 or 1√x2−1. Describe how the domain of arcsine suggests that one of these is wrong.
Evaluate lim
Suppose f(2x+1)=\dfrac{5x-9}{3x+7}\text{.} Evaluate f^{-1}(7)\text{.}
Suppose f^{-1}(4x-1)=\dfrac{2x+3}{x+1}\text{.} Evaluate f(0)\text{.}
Suppose a curve is defined implicitly by
\arcsin(x+2y)=x^2+y^2 \nonumber
Solve for y' in terms of x and y\text{.}