CHAPTER 4

Inverse Functions

U nder the right circumstances, a function f will have a so-called inverse,

a function f °1 that “undoes” the effect of f . Whereas f sends an
input x to the number f (x), the function f °1 sends the number f (x) back
to x. We describe f °1 intuitively below before giving an exact definition.
First consider a function f . From its graph, y
y = f (x)
we can describe the rule f as follows. For
an input x, to find the corresponding output
f (x)
value f (x), move vertically from point x on the
x-axis until reaching the graph of f . Then f
move horizontally to reach the output num-
ber f (x) on the y-axis. x
x

The inverse function f °1 reverses this by y

y = f (x)
sending any number f (x) back to x. Thus
the inputs for f °1 are the outputs f (x) of f .
Think of f °1 as a rule that works this way: f (x)
Given an input value f (x), move horizontally
f °1
from this point on the y-axis to the graph of f .
Then go vertically to the reach the output x x
x
on the x-axis.
But this process doesn’t work for just any f . y
y = f (x)
Take the f on the right. Starting at f (x) on
the y-axis, we can move horizontally to two f ( x)
points on the graph, then down to two values
of x. How can we say which x is the right
output for the input f (x)? The problem is
due to the horizontal line through f (x) that x
x x
touches the graph of f twice. ? ?

The above discussion means that a function f will have an inverse

provided that no horizontal line meets its graph at more than one point. A
function that meets this requirement is called one-to-one.
60 Inverse Functions

4.1 One-to-one Functions and Their Inverses

Now we will formalize the previous page’s discussion. In order to define the
inverse of a function, we first require that it be one-to-one.

Definition 4.1 A function f (x) is one-to-one if no horizontal line meets

its graph at more than one point.
p
For example, the function f (x) = x + 1 But the function f (x) = x2 is not
is one-to-one because no horizontal one-to-one because there exist
line intersects its graph more than horizontal lines intersecting the
once. (Each horizontal line touches graph at more than one point.
the graph at one point or none at all.)
y
y y = x2
p
y= x+1

x x
The term “one-to-one” comes from the fact that any one output number f (x)
corresponds to exactly one (and not more than one) input number x.
With one-to-one functions defined, we can now define the inverse of
such a function. On the previous page we said the inverse of f should be a
° ¢
function f °1 sending f (x) back to x, which means f °1 f (x) = x.

Definition 4.2 Any one-to-one function f with domain X and range Y

has an inverse, which is a function f °1 with domain Y and range X ,
° ¢ ° ¢
for which f °1 f (x) = x for any x in X , and f f °1 (y) = y for any y in Y .

As an illustration, the diagram below shows the graph of a one-to-one

function f with domain X and range Y .
The inverse f °1 sends any b in Y to the y
number x in X for which f (x) = b. Thus 8 y = f (x)
<
√ ! Y b
the number x :
f °1 (b) = .
for which f (x) = b
x
° ¢ ° ¢ x = f °1 (b)
Thus f f °1 (b) = b. This and f °1 f (x) = x | {z }
mean that f and f °1 “undo” one another. X
Above we have b = f (x) and f °1 (b) = x. In general, if x and y are two
numbers for which y = f (x), then f °1 (y) = x. In other words, the equations
y = f (x) and f °1 (y) = x express the same relationship between x and y.
One-to-one Functions and Their Inverses 61

On the previous page we encountered the equation

√ !
the number x
f °1 (b) = . (4.1)
for which f (x) = b

This is a rule for f °1 . It says that when you are dealing with a number b
and are trying to find f °1 (b), ask yourself what number x you’d have to plug
into f to get b. Then f °1 (b) equals that number x. This kind of backwards
thinking can often lead to an easy answer.
Example 4.1 Let f (x) = 2 + x + 2 x . This function increases as x increases,
so it is one-to-one and thus has an inverse. Find f °1 (3), f °1 (5) and f °1 (8).
√ !
the number x
Let’s start with f (3). Equation (4.1) says f (3) =
°1 °1
.
for which f (x) = 3
Try plugging a few values into f . You will soon hit upon f (0) = 2 + 0 + 20 = 3.
Thus f °1 (3) = 0. Also f (1) = 2 + 1 + 21 = 5, so f °1 (5) = 1. We are on a roll:
f (2) = 2 + 2 + 22 = 8, so f °1 (8) = 2.
But computing, say, f °1 (7) is not so easy because there is no obvious x for
which f (x) = 7. This is not to say that f °1 (7) doesn’t exist (it does). It’s just
not something we can calculate mentally.

To close, we remark that the function f °1 is pronounced “ f inverse.”

Please note that f °1 is the symbol for the function defined by Definition 4.2.
It is not the reciprocal of f . That is, f °1 (x) 6= f (1x) . If we ever wanted to
° ¢°1
express the reciprocal of f (x) we would write f (x) = f (1x) .
And finally, a question. Does the function f °1 have an inverse? The
° ¢
equation f f °1 (y) = y means that f undoes the effect of f °1 , so f is the
° ¢°1
inverse of f °1 . In other words f °1 = f.

Exercises for Section 4.1

1. Suppose f ( x) = 2 x . Find f °1 (8), f °1 (4), f °1 (2), f °1 (1), and f °1 (0.5).

2. Suppose f ( x) = 10 x . Find f °1 (1000), f °1 (100), f °1 (10), f °1 (1), and f °1 (0.1).
p
3. Suppose f ( x) = x + x3 . Find f °1 (2), f °1 (10), f °1 (°2), f °1 (0), and f °1 ( 3 3 + 3).
4. Suppose f ( x) = 3 x + x3 . Find f °1 (4), f °1 (17), f °1 (1), f °1 (54) and f °1 (°2/3).
5. Suppose f ( x) = x + sin( x). Find f °1 (0), f °1 (º), f °1 (º/2 + 1) and f °1 (2º).
6. Suppose f ( x) = x + cos( x). Find f °1 (1), f °1 (º/2), f °1 (º ° 1) and f °1 (2º + 1).
62 Inverse Functions

4.2 Graphing the Inverse of a Function

We now explore how the graph of f °1 is related to the graph of f . As starting
point consider the one-to-one function f graphed below.

4
y = f (x)
3
2
1
y = f (x)
°4 °3 °2 °1 0 1 2 3 4
°1
°2
°3

Let’s draw the graph of f °1 . The range of f is the interval [°2, 3.5], so
this is the domain of f °1 . To graph f °1 we will pick some values in this
interval, plug them into f °1 , make a table, and sketch the graph.
We learned how to find f °1 (b) in Section 4.1. For example, f °1 (2) is the
number x for which f (x) = 2. The graph of f shows f (1) = 2, so f °1 (2) = 1.
Similarly f (4) = 3.5, so f °1 (3.5) = 4. Continuing, we get

f °1 (°2) = °4, f °1 (°1) = °3, f °1 (0) = °2, f °1 (1) = °1, f °1 (3) = 3.

Next we tally these values in a table, plot the points and connect them.
For comparison we show the graph of f °1 (dashed) with that of f (solid).

y = f °1 (x)
4
x f °1 ( x) y = f (x)
3
°2 °4
2
°1 °3
1
0 °2
1 °1 °4 °3 °2 °1 0 1 2 3 4
2 1 °1

3 3 °2

3.5 4 °3
°4

There is a striking relationship between the graphs of f and f °1 . The

graph of f °1 is the graph of f reflected across the dotted line. This line is
the graph of the equation y = x. (Think of it as the graph of the equation
y = 1 · x + 0, so it is a line with slope 1 and y-intercept 0.)
It turns out that the graph of f °1 is always the graph of f reflected
across the line y = x. Before justifying this we look at one more example.
Graphing the Inverse of a Function 63

p
Example 4.2 Consider the function f (x) = x, with domain [0, 1) and
range [0, 1). Then its inverse also has domain and range [0, 1). Moreover,
√ ! √ !
the number x for the number x for
f °1 (b) = = p = b2 .
which f (x) = b which x = b
p
Thus the inverse of f (x) = x is the function f °1 (x) = x2 .
p
The graphs of y = x and its inverse y = x2 are sketched below. Notice
that because the domain of f °1 is [0, 1), the graph of f °1 (x) = x2 is plotted
only on this domain. (The inverse f °1 has no negative inputs because
p
f (x) = x has no negative outputs.) Just as in the previous example, the
graph of f °1 is the graph of f reflected across the line y = x.

y
y = f °1 (x) = x2
4

3
p
2 y = f (x) = x
1
x
0 1 2 3 4

To see why the graph of f °1 is always the graph of f reflected across

the line y = x, recall that the graph of y = f (x) consists of all points (x, f (x))
where x is in the domain of f . Because the inverse f °1 sends each number
f (x) to x, its graph consists of the points ( f (x), x). This amounts to saying
that for any point (a, b) on the graph of f there is a corresponding point (b, a)
on the graph of f °1 .
y
y=x
(a, b)
b

a (b, a)
x
a b

Now think about how points (a, b) and (b, a) are related. The picture above
shows that any point (a, b) on the graph of f reflects across the line y = x to
(b, a), which is on the graph of f °1 . We have established the following fact.

Fact: The graph of f °1 is the graph of f reflected across the line y = x.

64 Inverse Functions

Exercises for Section 4.2

1. Below is the graph of f ( x) = 2 x . 1

2. Below is the graph of f ( x) = .
2x
Find f °1 (4), f °1 (2), f °1 (1), f °1 (1/2), Find f °1 (4), f °1 (2), f °1 (1), f °1 (1/2),
f °1 (1/4), f °1 (1/8), and f °1 (1/16). f °1 (1/4), f °1 (1/8), and f °1 (1/16).
Sketch the graph of f °1 . Sketch the graph of f °1 .
4 4
3 3
2 2
1 1

°4 °3 °2 °1 0 1 2 3 4 °4 °3 °2 °1 0 1 2 3 4
°1 °1
°2 °2
°3 °3

3. A function f is graphed below. 4. A function f is graphed below.

Find f °1
(°4), f °1
(°3), f °1
(°2), Find f °1 (°3), f °1 (°2), f °1 (°1),
f °1 (0), f °1 (2), f °1 (3) and f °1 (4). f °1 (0), f °1 (2), and f °1 (4).
Sketch the graph of f °1 . Sketch the graph of f °1 .
4 4
3 3
2 2
1 1

°4 °3 °2 °1 0 1 2 3 4 °4 °3 °2 °1 0 1 2 3 4
°1 °1
°2 °2
°3 °3

5. Here is a table for some values of a one-to-one function f . Use it to make a table
for f °1 . Sketch the graphs of f and f °1 .

x °4 °3 °2 °1 0 1 2 3 4
f ( x) 5 4 3 2 1 0.75 0. 5 0.25 0

6. Here is a table for some values of a one-to-one function f . Use it to make a table
for f °1 . Sketch the graphs of f and f °1 .

x °4 °3 °2 °1 0 1 2 3 4
f ( x) °4 °3.5 °3 °2.5 °1 1 3 3.5 5
Finding Inverses 65

4.3 Finding Inverses

This section reviews a technique for finding the function f °1 (x) when f (x) is
given as an algebraic expression.
As a point of departure we note one of many uses for inverses: they
can be used to solve certain equations. For example, suppose we have an
equation of form
y = f (x)

and we want to solve it for x in terms of y. That is, we want to isolate the x on
one side of the equation, and have on the other side an expression involving
the variable y. If f happens to have an inverse, then we can take f °1 of
° ¢
both sides of the above equation to get f °1 (y) = f °1 f (x) . As Definition 4.2
° ¢
assures us that f °1 f (x) = x, this becomes

f °1 (y) = x,

and we have now solved for x in terms of y.

In summary, solving y = f (x) for x yields x = f °1 (y).
Now suppose we know the function f (x) but do not know what f °1 (x) is.
The above discussion tells us that we can find f °1 (x) by algebraically solving
the equation y = f (x) for x, and we will get x = f °1 (y). Of course we will
probably want to use x as the independent variable and y as a dependent
variable, so we can interchange them to get y = f °1 (x), and the inverse is at
hand. Here is a summary of our technique.

How to compute the inverse of f (x)

1. Write y = f (x)
2. Interchange x and y to get x = f (y).
3. Solve x = f (y) for y to get y = f °1 (x).

Example 4.3 Find the inverse of the function f (x) = x3 + 1.

Let us carry out the steps of our new procedure.
1. Write y = f (x), which in this case is y = x3 + 1.
2. Interchange x and y to get x = y3 + 1.
3. Next we solve x = y3 + 1 for y.
x ° 1 = y3
p
x ° 1 = y.
3
p p
Therefore y = 3
x ° 1, and we conclude f °1 (x) =
x ° 1.
3

p
The inverse has now been computed and it is f °1 (x) = 3 x ° 1.
66 Inverse Functions

Let’s round out Example 4.3 by comparing the graphs of f (x) and f °1 (x).
This is a good opportunity to use the graph-shifting techniques of Section 2.4.
the graph of f (x) = x3 + 1 is the graph of y = x3 shifted up one unit, and the
p p
graph of f °1 (x) = 3 x ° 1 is the graph of y = 3 x shifted right one unit. The
shifted graphs are sketched below. As expected, one is the reflection of the
other across the line y = x.

y y = f (x) = x3 + 1

p
y = f °1 (x) =
3
x°1
x

If you are ever in doubt that the inverse you have computed is correct,
° ¢
there is a way to check it. Definition 4.2 says both equations f °1 f (x) = x
° °1 ¢
and f f (x) = x must hold. Verifying one (or both) of these assures you that
your work is correct. In the example just done, we started with f (x) = x3 + 1
p
and obtained f °1 (x) = 3 x ° 1. Note that
° ¢ p
f °1 f (x) = 3
f (x) ° 1
p
3
= x3 + 1 ° 1
p
3
= x3
= x.
° ¢
The fact that f °1 f (x) = x indicates that our work was correct. The inverse
f °1 literally “undoes” the effect of f , sending any number f (x) back to x.
This section has used only the letter f for a function. Of course other
letters can be used. A one-to-one function g will have an inverse g°1 , etc.
x
Example 4.4 Is the function g(x) = one-to-one? If so, find its inverse.
x°1
To answer the first question, let’s sketch the graph of g and see whether
any horizontal line crosses its graph more than once. To draw the graph we
can manipulate g slightly and apply shifting. Notice that

x 1 + (x ° 1) 1 x°1 1
g(x) = = = + = + 1.
x°1 x°1 x°1 x°1 x°1
Finding Inverses 67

1 1
As g(x) = + 1, we see that its graph is the graph of y = shifted
x°1 x
right one unit and up one unit. This is sketched below. As no horizontal
line crosses the graph more than once, g is one-to-one, so it has an inverse.
y

y = g(x)
x

Now we carry out our procedure for computing g°1 from g. The first step
is to write the equation y = g(x), which in this case is
x
y= .
x°1

Next we interchange x and y to get

y
x= .
y°1

Now we must solve this for y. We want to isolate y, so as a fist step we get
the y out of the denominator by multiplying both sides by y ° 1.
y
x(y ° 1) = (y ° 1)
y°1
xy° x = y

Now we need to collect all occurrences of y on one side. Doing this, we get

xy° y = x.

To isolate y we factor it out on the left and divide both sides by x ° 1.

y(x ° 1) = x
y(x ° 1) x
=
x°1 x°1
x
y =
x°1
x
Now that the equation has been solved for y we find that g°1 (x) = .
x°1
x x
In summary, the inverse of g(x) = is the function g°1 (x) = .
x°1 x°1
 68 Inverse Functions

Interestingly, the inverse of g is g itself, that is, g is its own inverse. This
is just a coincidence – most functions are not equal to their inverses. But
perhaps we should not have been surprised that this g is its own inverse:
Looking at the above graph of g, we see that it is symmetric with respect
to the line y = x, that is, reflecting it across the line does not yield a new
graph. Thus the graphs of g and g°1 are identical, so they are the same
function.

Our technique of finding the inverse of f by solving x = f (y) for y has its
limitations because the equation may be difficult or impossible to solve.
Consider the function f (x) = x + sin(x) from Exercise 5 on page 61. By
“reverse engineering” we can compute f °1 (x) for certain convenient values
of x. For instance, if asked about f °1 (º/2 + 1) we would (after some thought)
note that f (º/2) = º/2 + sin(º/2) = º/2 + 1 and therefore f °1 (º/2 + 1) = º/2.
But actually finding a formula for f °1 (x) is problematic. It involves
solving the equation x = y + sin(y) for y. This is a very difficult problem
because our standard equation solving techniques cannot isolate the y.

Exercises for Section 4.3

Each of the following functions is one-to-one. Use the methods of this section
to find their inverses.

1. f (x) = (x ° 3)3 ° 1 1
2. f (x) = °
x
p
3. g(x) = ° 5 x + 2 4. f (x) = 2 ° x3
2° x 2 ° 3x
5. f (x) = 6. f (x) =
x+5 x+5
7. f (x) = x3 + 3x2 + 3x + 1 8. f (x) = x3 + 3x2 + 3x
(Hint: factor first.) (Hint: compare to Exercise 7.)

9. f (x) = 2 ° x2 on domain [0, 1) 2

10. h(x) = p
3
x
1 1
11. f (µ ) = 12. f (x) = 1 °
µ+3 x
p 1
13. g(x) = 3 ° 5 3 4x ° 3 14. f (w) =
w3 + 3
Restricting the Domain 69

4.4 Restricting the Domain

If a function is not one-to-one, then it has no inverse. This is the case for
f (x) = x2 graphed below, left. A horizontal line crosses its graph twice, so f
is not one-to-one and thus has no inverse.

y = f (x) = x2 y = f (x) = x2 y = f (x) = x2

p
y = f °1 (x) = x

x x x

Occasionally we will be in a situation where it is still desirable to find

an “inverse” of such a function. This task is not as hopeless as it may seem.
One approach is to restrict the domain of f to make it one-to-one.
For the function f (x) = x2 , we could declare the domain to be the interval
[0, 1), rather than its natural domain (°1, 1). This middle graph above
shows y = f (x) with this domain. Now f is one-to-one, and indeed it has
p
an inverse f °1 (x) = x. The right-hand graph shows f with this restricted
p
domain, along with its inverse f °1 (x) = x.
Modifying the domain of a function to make it one-to-one is called re-
stricting the domain. We will rarely have to use this technique, so it is not
necessary to work any exercises involving it. It is used only in Chapter 6
where we develop the useful idea of inverse trig functions.
A function like sin(x) is clearly not-one-to one because many horizontal
lines cross it infinitely many times. But if we restrict its domain to the
interval [° º2 , º2 ], then the resulting function (graphed bold below) is one-to-
one and has an inverse.
y
y = sin(x)

x
° º2 º
2

Chapter 6 will develop this idea. We will restrict the domains of the
trigonometric functions to make them one-to-one. We will then be able to
define such functions as sin°1 (x), tan°1 (x), etc.
70 Inverse Functions

4.5 Exercise Solutions for Chapter 4

Exercises for Section 4.1

1. Suppose f ( x) = 2 x . Find f °1 (8), f °1 (4), f °1 (2), f °1 (1), and f °1 (0.5).

f (3) = 8, so f °1 (8) = 3 f (2) = 4, so f °1 (4) = 2 f (1) = 2, so f °1 (2) = 1

f (°1) = 2°1 = 1
2 = 0.5, so f °1 (0.5) = °1

p
3. Suppose f ( x) = x + x3 . Find f °1 (2), f °1 (10), f °1 (°2), f °1 (0), and f °1 ( 3 3 + 3).

f (1) = 2, so f °1 (2) = 1 f (2) = 10, so f °1 (10) = 2 f (°1) = °2, so f °1 (°2) = °1

p p p
3 3
p p p
f ( 3) = 3 + 3 = 3 + 3, so f °1 ( 3 + 3) = 3
3 3 3 3 3

5. Suppose f ( x) = x + sin( x). Find f °1 (0), f °1 (º), f °1 (º/2 + 1) and f °1 (2º).

f (0) = 0, so f °1 (0) = 0 f (º) = º, so f °1 (º) = º
f (º/2) = º/2 + 1, so f °1 (º/2 + 1) = º/2 f (2º) = 2º, so f °1 (2º) = 2º

Exercises for Section 4.2

1. Below is the graph of f ( x) = 2 x . 2. A function f is graphed below.
f (2) = 4, so f °1
(4) = 2 f (°4) = °4, so f °1 (°4) = °4
f (1) = 2, so f °1 (2) = 1 f (°2) = °3, so f °1 (°3) = °2
f (0) = 1, so f °1 (1) = 0 f (°1) = °2, so f °1 (°2) = °1
f (°1) = 1/2, so f °1 (1/2) = °1, f (0) = °0, so f °1 (0) = 0
f (°2) = 1/4, so f °1 (1/4) = °2 f (1) = 2, so f °1 (2) = 1
f (°3) = 1/8, so f °1 (1/8) = °3 f (2) = 3, so f °1 (3) = 2
f (°4) = 1/16, so f °1 (1/16) = °4. f (4) = 4, so f °1 (4) = 4

4 4
3 3
2 2
1 1 y = f °1 ( x)

°4 °3 °2 °1 0 1 2 3 4 °4 °3 °2 °1 0 1 2 3 4
°1 °1
°2 y = f °1 ( x) °2
°3 °3
Exercise Solutions for Chapter 4 71

3. Here is a table for some values of a one-to-one function f . Use it to make a table
for f °1 . Sketch the graphs of f and f °1 .
5
4
x °4 °3 °2 °1 0 1 2 3 4 3
f ( x) 5 4 3 2 1 0.75 0. 5 0.25 0 y = f (x) 2
1
x 5 4 3 2 1 0.75 0.5 0.25 0 °4 °3 °2 °1 0 1 2 3 4 5
°1
f ( x) °4 °3 °2 °1 0 1 2 3 4 °1
°2
°1 (x)
°3 y = f

Exercises for Section 4.3

p
1. Find the inverse: f ( x) = ( x ° 3)3 ° 1 3. Find the inverse: g( x) = ° 5 x + 2
p
( x ° 3)3 ° 1
5
y = y = ° x+2
p
x = ( y ° 3)3 ° 1 x = ° 5 y+2
3 ≥ p ¥5
x+1 = ( y ° 3) x5 = ° 5 y+2
p
3
x+1 = y°3
p
3
x5 = °( y + 2)
y = x+1+3
p y = °2 ° x5
°1 3
f ( x) = x+1+3
g°1 ( x) = °2 ° x5
2° x 7. Find the inverse: f ( x) = x3 + 3 x2 + 3 x + 1
5. Find the inverse: f ( x) =
x+5
2° x y = x3 + 3 x2 + 3 x + 1
y =
x+5 x = y3 + 3 y2 + 3 y + 1
2° y
x =
y+5 x = ( y + 1)3
x( y + 5) = 2° y p q
3
3
x = ( y + 1)3
x y + 5x = 2° y
p
xy+ y = 2 ° 5x 3
x = y+1
y( x + 1) = 2 ° 5x p
3
y = x°1
2 ° 5x
y = p
x+1 f °1 ( x) = 3
x°1
°1 2 ° 5x
f ( x) =
x+1
9. Find the inverse: f ( x) = 2 ° x2

y = 2 ° x2
x = 2 ° y2
y2 = 2° x
p
y = 2° x
°1
p
f ( x) = 2° x
 72 Inverse Functions

1 p
11. Find the inverse of f (µ ) = 13. Find the inverse: g( x) = 3 ° 5 3 4 x ° 3
µ+3
p3
1 y = 3 ° 5 4x ° 3
y = p
µ+3 x = 3 ° 5 3 4y ° 3
p
1 3° x = 5 3 4y ° 3
µ = 3° x p
y+3 = 3
4y ° 3
5
µ ( y + 3) = 1 µ ∂
3° x 3
= 4y ° 3
µ y + 3µ = 1 5
µ ∂3
3° x
µy = 1 ° 3µ +3 = 4y
5
1 ° 3µ µ ∂
y = 1 3° x 3 3
µ y = +
4 5 4
1 ° 3µ µ ∂3
1 3° x 3
f °1 (µ ) = g°1 ( x) = +
µ 4 5 4