Differentiation Rules 3

3.4 The Chain Rule

 The Chain Rule
Suppose you are asked to differentiate the function

The differentiation formulas you learned in the previous sections of this chapter
do not enable you to calculate F(x).

Observe that F is a composite function. In fact, if we let

y = f(u) = and let u = g(x) = x2 + 1, then we can write
y = F(x) = f(g(x)), that is, F = f  g.

We know how to differentiate both f and g, so it would be useful to have a rule

that tells us how to find the derivative of F = f  g in terms of the derivatives of f
and g.
3
 The Chain Rule
It turns out that the derivative of the composite function f  g is the product of the
derivatives of f and g. This fact is one of the most important of the differentiation
rules and is called the Chain Rule.

It seems plausible if we interpret derivatives as rates of change. Regard du/dx as

the rate of change of u with respect to x, dy/du as the rate of change of y with
respect to u, and dy/dx as the rate of change of y with respect to x.
If u changes twice as fast as x and y changes three times as fast as u, then it
seems reasonable that y changes six times as fast as x, and so we expect that

4
The Chain Rule

5
 The Chain Rule
The Chain Rule can be written either in the prime notation

(f  g)(x) = f(g(x))  g(x)

or, if y = f(u) and u = g(x), in Leibniz notation:

Equation 3 is easy to remember because if dy/du and du/dx were quotients, then
we could cancel du.

Remember, however, that du has not been defined and du/dx should not be
thought of as an actual quotient.
6
 Example 1 – Using the Chain Rule
Find F'(x) if F(x) = .

Solution 1: prime notation

(Using Equation 2): We have expressed F as
F(x) = (f  g)(x) = f(g(x)) where f(u) = and g(x) = x2 + 1.

Since
and g(x) = 2x

we have F(x) = f(g(x))  g(x)

7
 Example 1 – Solution cont’d

Solution 2: Leibniz notation

(Using Equation 3): If we let u = x2 + 1 and y = , then

8
 The Chain Rule
When using Formula 3 we should bear in mind that dy/dx refers to the derivative
of y when y is considered as a function of x (called the derivative of y with
respect to x), whereas dy/du refers to the derivative of y when considered as a
function of u (the derivative of y with respect to u).
For instance, in Example 1, y can be considered as a function of x
(y = ) and also as a function of u (y = ).

Note that

whereas

9
 The Chain Rule
In general, if y = sin u, where u is a differentiable function of x, then, by the Chain
Rule,

Thus

In a similar fashion, all of the formulas for differentiating trigonometric functions

can be combined with the Chain Rule.

10
 The Chain Rule
Let’s make explicit the special case of the Chain Rule where the outer function f
is a power function.
If y = [g(x)]n, then we can write y = f(u) = un where u = g(x).

By using the Chain Rule and then the Power Rule, we get

11
 Example 3 – Using the Chain Rule with the Power Rule

Differentiate y = (x3 – 1)100.

Solution:
Taking u = g(x) = x3 – 1 and n = 100 in (4), we have

= (x3 – 1)100

= 100(x3 – 1)99 (x3 – 1)

= 100(x3 – 1)99  3x2

= 300x2(x3 – 1)99

12
 The Chain Rule
We can use the Chain Rule to differentiate an exponential function with any base
a > 0. Recall that a = eln a. So
ax = (eln a)x = e(ln a)x
and the Chain Rule gives
(ax) = (e(ln a)x) = e(ln a)x (ln a)x

= e(ln a)x  ln a = ax ln a

because ln a is a constant. So we have the formula

13
 The Chain Rule
In particular, if a = 2, we get

(2x) = 2x ln 2

We had given the estimate

(2x)  (0.69)2x

This is consistent with the exact formula (6) because ln 2  0.693147.

The reason for the name “Chain Rule” becomes clear when we make a longer
chain by adding another link.

14
The Chain Rule
Suppose that y = f(u), u = g(x), and x = h(t), where f, g, and h
are differentiable functions.

Then, to compute the derivative of y with respect to t, we use

the Chain Rule twice:

15
Tangents to Parametric Curves

16
 Tangents to Parametric Curves
We have discussed curves defined by parametric equations

x = f(t) y = g(t)

The Chain Rule helps us find tangent lines to such curves. Suppose f and g are
differentiable functions and we want to find the tangent line at a point on the
curve where y is also a differentiable function of x.

Then the Chain Rule gives

17
 Tangents to Parametric Curves
If dx/dt  0, we can solve for dy/dx:

Equation 7 (which you can remember by thinking of canceling the dt’s) enables
us to find the slope dy/dx of the tangent to a parametric curve without having to
eliminate the parameter t. We see from (7) that the curve has a horizontal
tangent when dy/dt = 0 (provided that dx/dt  0) and it has a vertical tangent
when dx/dt = 0 (provided that dy/dt  0).

18
 Example 10
Find an equation of the tangent line to the parametric curve

x = 2 sin 2t y = 2 sin t

at the point ( 1). Where does this curve have horizontal or vertical tangents?

Solution:
At the point with parameter value t, the slope is

19
 Example 10 – Solution cont’d

The point ( 1) corresponds to the parameter value t = /6, so the slope of the
tangent at that point is

20
 Example 10 – Solution cont’d

An equation of the tangent line is therefore

or

Figure 2 shows the curve and its tangent line.

Figure 2

21
 Example 10 – Solution cont’d

The tangent line is horizontal when dy/dx = 0, which occurs when cos t = 0 (and
cos 2t  0), that is, when t = /2 or 3/2.
(Note that the entire curve is given by 0  t  2.)

Thus the curve has horizontal tangents at the points (0, 2) and (0, –2), which we
could have guessed from Figure 2.

The tangent is vertical when dx/dt = 4 cos 2t = 0

(and cos t  0), that is, when t = /4, 3/4, 5/4, or 7/4.

The corresponding four points on the curve are (2, ).

If we look again at Figure 2, we see that our answer appears
to be reasonable.

22
3.5 Implicit Differentiation
 Implicit Differentiation
The functions that we have met so far can be described by expressing one
variable explicitly in terms of another variable—for example,
y= or y = x sin x
or, in general, y = f(x).

Some functions, however, are defined implicitly by a relation between x and y

such as
x2 + y2 = 25
or
x3 + y3 = 6xy

24
 Implicit Differentiation
In some cases it is possible to solve such an equation for y as an explicit function
(or several functions) of x.
For instance, if we solve Equation 1 for y, we get
y= , so two of the functions determined by the implicit Equation 1 are
f(x) = and g(x) = .
The graphs of f and g are the upper and lower semicircles of the circle
x2 + y2 = 25. (See Figure 1.)

Figure 1 25
 Implicit Differentiation
It’s not easy to solve Equation 2 for y explicitly as a function of x by hand. (A
computer algebra system has no trouble, but the expressions it obtains are very
complicated.)

Nonetheless, (2) is the equation of a curve called the folium of Descartes

shown in Figure 2 and it implicitly defines y as several functions of x.

The folium of Descartes

Figure 2 26
 Implicit Differentiation
The graphs of three such functions are shown in Figure 3.

Graphs of three functions defined by the folium of Descartes

Figure 3

When we say that f is a function defined implicitly by Equation 2, we mean that

the equation

x3 + [f(x)]3 = 6x f(x)

is true for all values of x in the domain of f.

27
 Implicit Differentiation
Fortunately, we don’t need to solve an equation for y in terms of x in order to find
the derivative of y.
Instead we can use the method of implicit differentiation.

This consists of differentiating both sides of the equation with respect to x and
then solving the resulting equation for y'.

In the examples and exercises of this section it is always assumed that the given
equation determines y implicitly as a differentiable function of x so that the
method of implicit differentiation can be applied.

28
 Example 1 – Finding a Tangent Line Implicitly
(a) If x2 + y2 = 25, find .

(b) Find an equation of the tangent to the circle x2 + y2 = 25 at the point (3, 4).

Solution 1:
(a) Differentiate both sides of the equation x2 + y2 = 25:

29
 Example 1 – Solution cont’d

Remembering that y is a function of x and using the Chain Rule, we have

Thus

Now we solve this equation for dy/dx:

30
 Example 1 – Solution cont’d

(b) At the point (3, 4) we have x = 3 and y = 4, so

An equation of the tangent to the circle at (3, 4) is therefore

y–4= (x – 3) or 3x + 4y = 25

Solution 2:
(b) Solving the equation x2 + y2 = 25, we get y = .

The point (3, 4) lies on the upper semicircle y =

and so we consider the function f(x) = .

31
 Example 1 – Solution cont’d

Differentiating f using the Chain Rule, we have

So

and, as in Solution 1, an equation of the tangent is 3x + 4y = 25.

32
3.6 Inverse Trigonometric Functions and Their Derivatives
 Inverse Trigonometric Functions and Their Derivatives

You can see from Figure 1 that the sine function y = sin x is not one-to-one
(use the Horizontal Line Test).

Figure 1

But the function f(x) = sin x, –/2  x  /2, is one-to-one (see Figure 2).

Figure 2
34
Inverse Trigonometric Functions and Their Derivatives

The inverse function of this restricted sine function f exists

and is denoted by sin–1 or arcsin. It is called the inverse sine
function or the arcsine function.

Since the definition of an inverse function says that

f –1(x) = y f(y) = x

we have

Thus, if –1  x  1, sin–1x is the number between –/2 and /2

whose sine is x.
35
Example 1
Evaluate (a) sin–1 and (b) tan(arcsin ).

Solution:
(a) We have

because sin(/6) = and /6 lies between –/2 and /2.

36
Example 1 – Solution cont’d

(b) Let  = arcsin , so sin  = .

Then we can draw a right triangle with angle  as in Figure

3 and deduce from the Pythagorean Theorem that the third
side has length .

Figure 3

This enables us to read from the triangle that

37
 Inverse Trigonometric Functions and Their Derivatives

The cancellation equations for inverse functions become, in this case,

38
Inverse Trigonometric Functions and Their Derivatives

The inverse sine function, sin–1, has domain [–1, 1] and range [–/2, /2],
and its graph, shown in Figure 4, is obtained from that of the restricted
sine function (Figure 2)
by reflection about the line y = x.

We know that the sine function f is continuous, so the inverse sine

function is also continuous.

Figure 2 Figure 4 y = sin–1x = arcsin x

39
Inverse Trigonometric Functions and Their Derivatives

We can use implicit differentiation to find the derivative of

the inverse sine function, assuming that it is differentiable.
(The differentiability is certainly plausible from its graph in
Figure 4.)

Let y = sin–1x. Then sin y = x and –/2  y  /2.

Differentiating sin y = x implicitly with respect to x, we obtain

and

40
Inverse Trigonometric Functions and Their Derivatives

Now cos y  0 since –/2  y  /2, so

Therefore

41
Inverse Trigonometric Functions and Their Derivatives

The inverse cosine function is handled similarly. The restricted

cosine function f(x) = cos x, 0  x  , is
one-to-one (see Figure 6) and so it has an inverse function
denoted by cos–1 or arccos.

Figure 6 y = cos x, 0  x  
42
Inverse Trigonometric Functions and Their Derivatives

The inverse cosine function, cos–1, has domain [–1, 1] and

range [0, ] and is a continuous
function whose graph is
shown in Figure 7.

Figure 7 y = cos–1x = arccos x

Its derivative is given by

43
 Inverse Trigonometric Functions and Their Derivatives

The tangent function can be made one-to-one by restricting it to the interval (–/2,
/2).
Thus the inverse tangent function
is defined as the inverse of the
function f(x) = tan x, –/2  x  /2,
as shown in Figure 8.

Figure 8

It is denoted by tan–1 or arctan.

44
Inverse Trigonometric Functions and Their Derivatives

The inverse tangent function, tan–1 = arctan, has domain

and range (–/2, /2).

Its graph is shown in Figure 10.

Figure 10 y = tan–1x = arctan x

45
 Inverse Trigonometric Functions and Their Derivatives

We know that

and

and so the lines x = /2 are vertical asymptotes of the graph of tan.

Since the graph of tan–1 is obtained by reflecting the graph of the restricted
tangent function about the line y = x, it follows that the lines y = /2 and y = –/2
are horizontal asymptotes of the graph of tan–1.

46
Inverse Trigonometric Functions and Their Derivatives

This fact is expressed by the following limits:

The formula for the derivative of the arctangent function is

derived in a way that is similar to the method we used for
arcsine.

47
Inverse Trigonometric Functions and Their Derivatives

If y = tan–1x, then tan y = x. Differentiating this latter equation

implicitly with respect to x, we have

and so

48
3.7 Derivatives of Logarithmic Functions
Derivatives of Logarithmic Functions
In this section we use implicit differentiation to find the
derivatives of the logarithmic functions y = loga x and, in
particular, the natural logarithmic function y = ln x.

50
Derivatives of Logarithmic Functions
If we put a = e in Formula 1, then the factor ln a on the right
side becomes ln e = 1 and we get the formula for the
derivative of the natural logarithmic function loge x = ln x:

By comparing Formulas 1 and 2, we see one of the main

reasons that natural logarithms (logarithms with base e) are
used in calculus: The differentiation formula is simplest when
a = e because ln e = 1.
51
Example 1
Differentiate y = ln (x3 + 1).
Solution:
To use the Chain Rule, we let u = x3 + 1.
Then y = ln u, so

52
Derivatives of Logarithmic Functions
In general, if we combine Formula 2 with the Chain Rule as
in Example 1, we get

or

53
Example 6
Find f(x) if f(x) = ln |x|.
Solution:
Since

it follows that

Thus f(x) = 1/x for all x  0.

54
Derivatives of Logarithmic Functions
The result of Example 6 is worth remembering:

55
Logarithmic Differentiation

56
Logarithmic Differentiation
The calculation of derivatives of complicated functions involving
products, quotients, or powers can often be simplified by taking
logarithms.

The method used in the following example is called logarithmic

differentiation.

57
Example 7 – Logarithmic Differentiation

Differentiate

Solution:
We take logarithms of both sides of the equation and use
the Laws of Logarithms to simplify:

ln y = ln x + ln (x2 + 1) – 5 ln (3x + 2)

Differentiating implicitly with respect to x gives

58
Example 7 – Solution cont’d

Solving for dy/dx, we get

Because we have an explicit expression for y, we can

substitute and write

59
Logarithmic Differentiation

If f(x) < 0 for some values of x, then ln f(x) is not defined, but we can write
|y| = |f(x)| and use Equation 4. We illustrate this procedure by proving the
general version of the Power Rule.

60
Logarithmic Differentiation
In general, there are four cases for exponents and bases:

1. (ab) = 0 (a and b are constants)

2. [f(x)]b = b[f(x)]b –1f(x)

3. [ag(x)] = ag(x)(ln a)g(x)

4. To find (d/dx)[f(x)]g(x), logarithmic differentiation can be

used.

61
The Number e as a Limit

62
The Number e as a Limit
We have shown that if f(x) = ln x, then f(x) = 1/x. Thus
f(1) = 1. We now use this fact to express the number e as
a limit.

From the definition of a derivative as a limit, we have

63
 The Number e as a Limit
Because f(1) = 1, we have

Then, by the continuity of the exponential function, we have

64
The Number e as a Limit
Formula 5 is illustrated by the graph of the function
y = (1 + x)1/x in Figure 4 and a table of values for small
values of x. This illustrates the fact that, correct to seven
decimal places,
e  2.7182818

Figure 4

65
The Number e as a Limit
If we put n = 1/x in Formula 5, then n → as x → 0+ and
so an alternative expression for e is

66