Math 1200

Calculus I

Fall 2022
Instructor: Keira Gunn

Lecture Notes

1
Table of Contents 2. Rules for Differentiation 15

2.0 Learning Outcomes 16

2.1 Derivatives of Polynomials and Exponentials 17
Notation
1. Limits, Continuity, and Rates of Change 4
Constant Rule
1.0 Learning Outcomes 5 Power Rule
1.1 Limits 6 Sum of Functions
Exponentials
Limits
2.2 Product Rule and Quotient Rule 18
Limit laws Product Rule
1.2 Evaluating Limits 7 Quotient Rule
When 𝑓(𝑎) Exists 2.3 Derivatives of Trigonometric Functions 19
Infinite Limits Trigonometric Limits
Evaluating Limits Algebraically Trigonometric Functions
2.4 Chain Rule 20
Squeeze Theorem
Composite Functions
1.3 Continuity 8 Chain Rule
Continuity 2.5 Implicit Differentiation 21
Piecewise Functions 2.6 Inverse Trigonometric Functions and Logarithms 21
Intermediate Value Theorem Inverse Trigonometric Functions
1.4 Limits at Infinity 9 Logarithms
𝑥 Approaching Infinity Logarithmic Differentiation
Asymptotes 2.7 Related Rates 22
Related Rates
1.5 Rates of Change 11
Strategies for Solving Related Rates Problems
Average Rate of Change
Tangent Lines
Instantaneous Rates of Change
1.6 The Derivative 13
The Limit Definition of Derivative
Differentiability
Position, Velocity, and Acceleration
The Derivative as a Function

2
3. Applications of Derivatives 24 4. Integration 38

3.0 Learning Outcomes 25 4.0 Learning Outcomes 39

3.1 L’Hopital’s Rule 26 4.1 Antiderivatives 40
Indeterminate forms Indefinite Integral
L’Hopital’s Rule Antiderivatives
Applying L’Hopital’s Rule
Initial Value Problems
3.2 Local Extrema 27
Intervals of Increase and Decrease 4.2 Areas Underneath Curves 41
Critical Numbers The Velocity Problem
Local Extrema The Definite Integral
First Derivative Test for Local Extrema Riemmann Sums
3.3 Concavity 29 Computing Areas
The Second Derivative 4.3 The Fundamental Theorem of Calculus 44
Concavity of a Graph
FTC I
Inflection Points
The Second Derivative Test for Local Extrema FTC II
3.4 Graphing 30 Net Change Theorem
3.5 Absolute Extrema 31 4.4 Substitution 46
Absolute Maximum and Minimum The Indefinite Integral
Closed and Bounded Intervals The Definite Integral
Single Critical Point Theorem 4.5 Areas Between Curves 47
3.6 Optimization 32 4.6 Average Value of a Function 48
Strategies for Optimization Problems
Average Value Formula
3.7 Linear Approximation 33
Differentials Mean Value Formula for Integrals
Linear Approximation
3.8 Newton’s Method 34
Newton’s Method
Where Newton’s Method Fails
3.9 Mean Value Theorem 35
Rolle’s Theorem
Upper Bound for the Number of Zeroes
Mean Value Theorem

3
 Section 1

Limits, Continuity, and Rates

of Change

4
Learning Outcomes

By the end of this section, you will be able to:

 Determine the limit of a function at 𝑥 = 𝑎 by looking at the graph of the function.

 Find the limit of a function at 𝑥 = 𝑎 by analyzing the values of the function for values near 𝑎.
 Compute the limit of a function at 𝑥 = 𝑎 algebraically using the limit laws.
 Compute a limit using the Squeeze Theorem
 Find the discontinuities of a function from its graph.
 Find the discontinuities of a function from its equation using limits.
 Find the discontinuities of a piecewise function.
 Use the Intermediate Value Theorem to approximate the location of a root of a function.
 Use the Intermediate Value Theorem to find a lower bound on the number of roots of a function.
 Compute the limit of a function to infinity algebraically.
 Determine the equations of any asymptotes of a function.
 Compute the average rate of change of a function over a specified interval.
 Compute the instantaneous rate of change of a function at a given point.
 Determine the equation of a tangent line to a function at a given point.
 Compute the derivative of a function at a given point.
 Interpret the derivative of a function as the rate of change.
 Determine if a function is differentiable at a point given its graph.
 Find velocity given a position function.
 Find acceleration given a velocity function.
 Compute the derivative of a function.
 Determine the location of any horizontal tangents to a function.

5
1.1 Limits lim 𝑓(𝑥) = 𝐿
𝑥→𝑎
We define the left-hand side of 𝑎to be all values that are “less than 𝑎” and the In other words, lim 𝑓(𝑥) exists only if both the left and right limits exist and
𝑥→𝑎
right-hand side of 𝑎 is all values that are “greater than 𝑎”.
are equal.
The use of left and right refers to the position of these values on a number
line where the left-hand side is the negative numbers and the right-hand side
is the positive numbers. If lim+ 𝑓(𝑥) = lim− 𝑓(𝑥) = 𝐿 , then lim 𝑓(𝑥) = 𝐿.
𝑥→𝑎 𝑥→𝑎 𝑥→𝑎

The Limit 2) Determine the limits at 𝑥 = −2, 0, 2, 4 for the function in 1).
As 𝑥 gets closer and closer to (but never equal to) 𝑎 from the left-hand side,
if 𝑓(𝑥) gets closer and closer to (in fact, as close as possible, and maybe 𝑥 2 − 𝑥 −2
even equal to) some value 𝐿 (a finite value or an infinite one) then we say 3) Evaluate 𝑓(𝑥) = at 𝑥 = −2, −1.1, −1.01, −1.001 to determine
𝑥+1
that 𝑳 is the left-hand limit of 𝒇(𝒙) as 𝒙 approaches 𝒂. We denote this by the left-hand limit of 𝑓(𝑥)as 𝑥 approaches -1, and evaluate at 𝑥 = 0,
−0.9, −0.99, −0.999 to determine the right-hand limit.
lim 𝑓(𝑥) = 𝐿 𝑥 2 − 𝑥 −2
𝑥 →𝑎 − Use these to evaluate lim if it exists.
𝑥→−1 𝑥+1
.
Similarly, as 𝑥 gets approachesa from the right-hand side, if 𝑓(𝑥) approaches 4) Use a table of values to determine the left and right-hand limits of 𝑓(𝑥) =
some value 𝐿 then we say that 𝑳 is the right-hand limit of 𝒇(𝒙) as 𝒙 1⁄ as 𝑥 approaches 0. Determine lim 1⁄ if it exists.
𝑥 𝑥 𝑥→0
approaches 𝒂. We denote this by:
Limit Laws
lim+ 𝑓(𝑥) = 𝐿 Let 𝑐 be a real number, and suppose the limits to 𝑎 of 𝑓(𝑥) and 𝑔(𝑥) both
𝑥 →𝑎
exist, then:
1. lim 𝑐 𝑓(𝑥) = 𝑐 lim 𝑓(𝑥)
The only difference between the two notations is the + or – above the 𝑎. 𝑥→𝑎 𝑥→𝑎

1) For the graph below, determine the left and right hand limits of 𝑓(𝑥) as 𝑥 2. lim [𝑓(𝑥) ± 𝑔(𝑥)] = lim 𝑓(𝑥) ± lim 𝑔(𝑥)
𝑥→𝑎 𝑥→𝑎 𝑥→𝑎
approaches −2, 0, 2, 4.
3. lim [𝑓(𝑥) ∙ 𝑔(𝑥)] = (lim 𝑓(𝑥)) ∙ (lim 𝑔(𝑥))
𝑥→𝑎 𝑥→𝑎 𝑥→𝑎

𝑓(𝑥) lim 𝑓(𝑥)

𝑥→𝑎
4. lim = if lim 𝑔(𝑥) ≠ 0
𝑥→𝑎 𝑔(𝑥) lim 𝑔(𝑥)
𝑥→𝑎
𝑥→𝑎

𝑛
5. lim [𝑓(𝑥)]𝑛 = [lim 𝑓(𝑥)]
𝑥→𝑎 𝑥→𝑎

These laws can be proved by using what is called the precise definition of the
limit. The precise definition gives us a calculable method to finding limits that
is more well-defined than our current definition of “it goes towards some
If 𝑓(𝑥) approaches 𝐿 as 𝑥 approaches 𝑎 (from either side) then we say 𝑳 is
number”. This is a topic that is not covered in this course.
the limit of 𝒇(𝒙) as 𝒙 approaches 𝒂. We denote this by:

6
5) Suppose that lim 𝑓(𝑥) = 5 and lim 𝑔(𝑥) = −2 Evaluating Limits Algebraically
𝑥→3 𝑥→3 𝟎
For a limit, we have to re-express the function in a way where we can
a) find lim[2𝑓(𝑥) + 4𝑔(𝑥)] b) find lim[𝑓(𝑥)2 ∙ 𝑔(𝑥)] 𝟎
𝑥→3 𝑥→3 cancel factors in the hopes of eliminating the zero problems.
[𝑔(𝑥)]2 Rational functions will usually require simplification and/or factoring, and
c) find lim
𝑥→3 𝑓(𝑥)− 2 radical functions will usually require rationalization.

1.2 Evaluating Limits 3) Evaluate the following limits:

2𝑥 2 + 3𝑥−2 𝑥 2 − 3𝑥+2
a) lim b) lim 𝑥 2 − 2𝑥+1
When 𝒇(𝒂) Exists 𝑥→−2 𝑥2− 4 𝑥→1
For polynomials, radicals, rationals, exponentials, logarithms, trig functions (𝑥+ℎ)2 − 𝑥 2 √𝑥+7−3
c) lim d) lim
and other functions that we use on a regular basis, we have the property: ℎ→0 ℎ 𝑥→2 𝑥− 2
2𝑥2 + 𝑥
1 2
lim 𝑓(𝑥) = 𝑓(𝑎), when 𝑓(𝑎) exists. e) lim ( − ) f) lim 𝑒 𝑥
𝑥→𝑎
𝑥→1 𝑥−1 𝑥2− 1 𝑥→0

This means that, for most functions, if we can plug our 𝑎 into 𝑓(𝑥) then we do
so to evaluate the limit. Squeeze Theorem
The Squeeze Theorem states that:
We have this property because of the continuity of these functions. We will If 𝑔(𝑥) is continuous and𝑓(𝑥) ≤ 𝑔(𝑥) ≤ ℎ(𝑥) near 𝑎 and
discuss continuity more in the next section. lim 𝑓(𝑥) = lim ℎ(𝑥) = 𝐿, then lim 𝑔(𝑥) = 𝐿
𝑥→𝑎 𝑥→𝑎 𝑥→𝑎

1) Evaluate the following limits: An explanation of the theorem can be seen in

a) lim(𝑥 2 + 3𝑥 − 1) b) lim 𝑒 𝑥−1 the illustration below. We can see that
𝑥→2 𝑥→1
because 𝑔(𝑥) is “trapped” between 𝑓(𝑥)and
Infinite Limits ℎ(𝑥), and 𝑓(𝑥) and ℎ(𝑥) both limit to 𝐿, 𝑔(𝑥)
When 𝑓(𝑎) does not exist, we have to do further analysis. We will typically is forced towards 𝐿 as well.
0 non-zero
get one of two cases: or .
0 0

non-zero
For a limit, both the left and right hand sides will go to infinity, since
𝟎
the top is nearing a finite value and the bottom is getting smaller and smaller. 4) Suppose that 2𝑥 − 1 ≤ 𝑓(𝑥) ≤ 𝑥 2 for all 𝑥.
In these cases, we simply have to determine if the left-hand limit and right- Show that lim 𝑓(𝑥) = 1
hand limit are going to positive or negative infinity. (They may not necessarily 𝑥→1

go to the same infinity)

The Squeeze Theorem is a useful tool for solving limits of complex functions
that can be bounded for simpler ones. For example, we will often use the
2) Evaluate the following limits:
fact that −1 ≤ sin 𝑥 ≤ 1 and −1 ≤ cos 𝑥 ≤ 1 to find a bound for a function
a) lim 𝑥⁄𝑥 − 1 b) lim 1⁄𝑥 2 involving trigonometric expressions.
𝑥→1 𝑥→0

1
5) Evaluate lim𝑥→0+ 𝑥 sin ( ).
𝑥

7
1.3 Continuity Piece-Wise Functions
Continuity only really becomes interesting when we look at piecewise
functions.
Continuity For a function defined as follows:
We say that a function is continuous at 𝑎 if the limit of the function exists at 𝑥2 𝑥 < −1
a and if the limit is equal to 𝑓(𝑎) 𝑓(𝑥) = { 𝑥 + 1 −1 ≤ 𝑥 < 1
𝐥𝐢𝐦 𝒇(𝒙) = 𝒇(𝒂) 2𝑥 − 1 𝑥≥1
𝒙→𝒂

In the previous section, we defined many limits by their continuity properties, We can sketch the function by drawing the three “sub-functions” and only
and now we are defining continuity using the limit. These are circular keeping the appropriate regions. The illustration on the right (next page)
definitions and we would require the precise definition of the limit to fix this would be the graph of 𝑓(𝑥).
problem.

1) Determine if 𝑓(𝑥) is continuous at 𝑥 = −2, 0, 2, 4. If 𝑓 is not continuous,

explain which condition of continuity is not met.

We can see that, aside from discontinuities that occur within each interval
from the function itself, we can also get discontinuities at the endpoints of
A function is continuous on an interval 𝐼 if it is continuous for all 𝑎 in 𝐼. each interval.
polynomials, rationals, exponentials, logarithms, trig functions, inverse trig
functions,absolute values are all continuous everywhere they exist. To find the discontinuities of a piecewise function we can either sketch the
In other words, the function is continuous on the entire domain. function, or look for discontinuities amongst the discontinuities of each
individual function (as long as they occur over their respective interval) or
2) Find where the following functions are continuous: endpoints where the limit does not exist or is not finite.
𝑥+2
a) 𝑓(𝑥) = 6𝑥 5 − 4𝑥 2 + 2 b) 𝑦 = 𝑥2− 1
3) Find all discontinuities for the function
c) 𝑦 = log(4𝑥 + 3) d) 𝑦 = 𝑒 𝑥−1 0 𝑥 < −1
𝑓(𝑥) = { 1/𝑥 −1 ≤ 𝑥 < 1
For radical functions (for example, a square root function) we have to be 2−𝑥 𝑥≥1
careful when using the domain to determine continuity. For example, √𝑥 is
4) Find the values of 𝑚 and 𝑏 so that 𝑓(𝑥) is continuous.
continuous on (0, ∞), whereas the domain is [0, ∞).
𝑥2 𝑥 < −1
𝑓(𝑥) = {𝑚𝑥 + 𝑏 −1 ≤ 𝑥 ≤ 1
ln 𝑥 𝑥>1

8
Intermediate Value Theorem bounds to find the precise number of solutions for an equation even if the
Suppose that for some continuous solutions might not be solvable.
function 𝑓 we have that 𝑓(−1) = −1
and 𝑓(1) = 1. When we graph these
two points, it is clear that the function 1.4 Limits at Infinity
must pass through the 𝑥-axis since it
is continuous (if it did not pass through
𝒙 Approaching Infinity
the 𝑥-axis, it would have to jump over
Instead of looking what happens to 𝑓(𝑥) as 𝑥 approaches a finite value 𝑎, we
it; causing a discontinuity at that
point). In other words, we know that can also look at what happens to the function as 𝑥 gets infinitely larger, in
there is a value 𝑥 = 𝑐 in the interval either the positive or negative direction. We denote these as:
lim 𝑓(𝑥) or lim 𝑓(𝑥)
(−1, 1) with 𝑓(𝑐) = 0. 𝑥→∞ 𝑥→−∞

In general, this gives us the Intermediate Value Theorem: 1) Evaluate 𝑓(𝑥) = 1⁄𝑥 at 𝑥 = 10, 100, 1000, 10000, and at
If 𝑓(𝑥) is a continuous function with 𝑓(𝑎) < 𝑁 < 𝑓(𝑏) 𝑥 = −10, −100, −1000 and −10000 and make a conclusion regarding
(or 𝑓(𝑏) < 𝑁 < 𝑓(𝑎)) then there is 𝑐 in (𝑎, 𝑏) with 𝑓(𝑐) = 𝑁. the values of lim 1⁄𝑥 and lim 1⁄𝑥 .
In other words, to get from (𝑎, 𝑓(𝑎)) to (𝑏, 𝑓(𝑏)) a continuous function must 𝑥→∞ 𝑥→−∞

pass through all of the values between 𝑓(𝑎) and 𝑓(𝑏).

When solving limits at infinity, the algebraic steps are typically very different
0
In the previous example, we had 𝑁 = 0, 𝑎 = −1, and 𝑏 = 1. from conventional limits. This is because we rarely get a “ ” occurring (or
0
Notice that the Intermediate Value Theorem says nothing about the location zeroes at all) so our strategy is not to cancel out zeroes. Instead, we
∞
of 𝑥 = 𝑐; just that it exists. In fact, many possible 𝑐-values could exist in the typically get limits of the form . We can deal with these by forcing the
∞
interval (𝑎, 𝑏).
expression “1⁄𝑥 ” to occur, since we know that
5) Show that 𝑓(𝑥) = 2 for some 𝑐 in [1, 2] for the function. lim 1⁄𝑥 = 0.
𝑥→∞
𝑓(𝑥) = 𝑥 + sin 𝑥.
2) Evaluate the limit: lim (𝑥 2 − 3𝑥 + 1).
6) Suppose that 𝑓(𝑥) is a continuous function with 𝑓(3) = 2 and 𝑓(𝑥) = 4 𝑥→∞
has no solution. Is it possible that 𝑓(1) = 7?
Using limit laws, it is an immediate result from this that lim 𝑐⁄𝑥 𝑛 = 0
𝑥→±∞
We can use the Intermediate Value Theorem to help us determine how many for any constant 𝑐and for any positive number𝑛 and so we can use this as
solutions there are to an equation by “bringing all terms to one side” and well.
letting the equation be “𝑓(𝑥) = 0” for the corresponding 𝑓(𝑥). From here, we
can get a lower bound on the number of solutions by finding 𝑥 values that When solving limits at infinity involving rational functions, we can factor the
give positive or negative outputs (since, in order to go from positives to highest occurring power of 𝑥 from the top and the highest occurring power of
negatives or vice versa, the function must pass through 0). 𝑥 from the bottom and then use limit laws to cancel the resulting “ 1⁄𝑥 ” terms
to simplify the expression.
7) Show that the equation ln 𝑥 + 2 = 𝑥 has at least 2 solutions.

Later in the semester, we will find a method to find an upper bound on the
number of solutions to an equation. We will be able to combine these
9
3) Evaluate the following limits: We indicate a horizontal asymptote with a dotted horizontal line. Notice that
as 𝑥 moves further and further to the right or left (and so goes to positive or
3𝑥−1 2𝑥 2 + 𝑥−3
a) lim b) lim negative infinity), the value of 𝑦 tends towards some finite number.
𝑥→∞ 𝑥 2 − 2𝑥+1 𝑥→−∞ 4𝑥−1
5) Determine the equations of the vertical and horizontal asymptotes for the
3𝑥 2 + 2𝑥−1 √𝑥 2 + 1
c) lim d) lim following function:
𝑥→∞ 4−5𝑥 2 𝑥→−∞ 2𝑥

e) lim (ln 𝑥 − ln(𝑥 + 1))

𝑥→∞

We can also use the squeeze theorem to evaluate infinite limits when we can
create upper and lower bounds for the limited function.

4) Use the squeeze theorem to evaluate the following:

2
1+𝑒 −𝑥
lim
𝑥→∞ 𝑥

Asymptotes So, we get a vertical asymptote at 𝑥 = 𝑎 where lim± 𝑓(𝑥) = ±∞.

𝑥→𝑎
Asymptotes describe the “extreme” behaviour of a function. They are
typically straight lines that the function gets “closer and closer to”. Most commonly, these occur in rational functions where the denominator is
Graphically, it will appear as though the curve of the function moves towards equal to zero (and the numerator is not equal to zero); however they can also
the asymptote and eventually becomes almost the same line. occur in a logarithm where the “inside” is zero or the tangent function where
𝜋
the “inside” is + 𝑘𝜋.
A Vertical Asymptote will appear similar to the following sketches: 2

6) Determine the equations of the vertical asymptotes for the following

functions:
𝑥 2 − 3𝑥+2
a) 𝑓(𝑥) = b) 𝑓(𝑥) = ln 𝑥 + 𝑥
𝑥 2 − 5𝑥+4
𝑥𝑒 −𝑥 𝜋
c) 𝑦 = d) 𝑓(𝑥) = tan (3𝑥 + )
2𝑥+1 4

We indicate a vertical asymptote with a dotted vertical line. Notice that as

Also, we get a horizontal asymptote at
the 𝑥-values move closer to the dotted line, the 𝑦-values increase towards
𝑦 = lim 𝑓(𝑥) or 𝑦 = lim 𝑓(𝑥)
positive or negative infinity. 𝑥→∞ 𝑥→−∞
if these limits both exist and are finite.
A Horizontal Asymptote
will appear similar to To find a horizontal asymptote, we take the limits to positive and negative
those in the following infinity and determine if they are finite.
sketch:
7) Determine the equations of the horizontal asymptotes for the following
functions:

10
 𝑥 2 − 3𝑥+2 Tangent Lines
a) 𝑓(𝑥) = b) 𝑓(𝑥) = ln 𝑥 + 𝑥
𝑥 2 − 5𝑥+4
The tangent line at a point 𝑥 = 𝑎 is the line that touches the curve at
𝑥𝑒 −𝑥 √4𝑥 2 + 1
c) 𝑦 = d) 𝑔(𝑥) = 𝑥 = 𝑎 and “has the same direction”.
2𝑥+1 𝑥+2

To draw an asymptote, we draw the dotted line where indicated, and force
the curve to “move towards the line” in an appropriate way.

Note that a vertical asymptote cannot be crossed by a curve, but a horizontal

asymptote can be crossed infinitely many times.

8) Sketch a graph with a horizontal asymptote at 𝑦 = 1 and vertical

asymptotes at 𝑥 = 2 and 𝑥 = 5

1.5 Rates of Change

For values of 𝑥 near 𝑎, we can see that the tangent line looks very much like
Average Rate of Change the curve.
We can find the average rate of change of a function 𝑓(𝑥) over an interval
[𝑎, 𝑏] by treating the function as though it were linear and using the slope 3) For the following function, sketch the tangent lines at 𝑥 = −2, 0, 3.
between 𝑎and 𝑏.
∆𝑓 𝑓(𝑏)− 𝑓(𝑎)
We get average rate of change = =
∆𝑥 𝑏−𝑎

1) After 4 hours of work, Jamal has earned $63; however, after 9 hours of
work he has earned $138. What is Jamal’s average rate of pay from
between 4 and 9 hours?

2) A car drives 174 km between 2 pm and 5 pm. What is its average speed
over this time?

If we draw a straight line

between two points 𝑥 = 𝑎 and We lose this property when the function is not continuous at 𝑥 = 𝑎, or if the
𝑥 = 𝑏 on the graph of a function, function suddenly “changes direction” at 𝑥 = 𝑎.
we call this line the secant line.
In the following two examples, the tangent lines do not exist:
In other words, the average rate
of change is simply the slope of
the secant line.

11
 For a linear function, the instantaneous rate of change is always just the
slope of the line. Since the tangent of a function is a straight line that “looks
like” the function, its instantaneous rate of change is the slope of the tangent!

5) Use a tangent line to approximate the

instantaneous rate of change for the following
function at 𝑥 = 2.

In the example on the left, the function has a discontinuity at 𝑥 = 𝑎. While

the line drawn does have the same direction as the curve, it does not meet
the curve at point 𝑥 = 𝑎.
In the example on the right, if we try to draw a line that has the direction on Drawing tangents by hand is not a very practical way to determine
the left side of the point, it does not have the same direction as on the right instantaneous rates of change. Instead, consider the following graph of a
side. We get the same result if we try to draw a tangent using the right side function, its tangent at 𝑥 = 𝑎, and several secant lines between 𝑎 and various
of the point. different 𝑥 values.

The tangent line to 𝒇(𝒙) at 𝒙 = 𝒂 exists only if 𝒇 is

continuous at 𝒙 = 𝒂 and if the graph of 𝒚 = 𝒇(𝒙) is
smooth at 𝒙 = 𝒂.

4) For the following function, find the values of 𝑥 where there the tangent
line does not exist:

Notice that as the 𝑥 values get closer to 𝑎, the secant line starts to look more
and more like the tangent line. In fact, we could say that:

lim (secant through 𝑥 and 𝑎) = (tangent at 𝑥 = 𝑎)

𝑥→𝑎
But, more importantly, we have that the slope of the secants through 𝑥 and a
Instantaneous Rates of Change will approach the slope tangent, or:
The instantaneous rate of change of a function 𝑓(𝑥) is the rate of change
of 𝑓 at any given instant. For example, a car might have an average of lim (slope of secant through 𝑥 and 𝑎) = (slope of tangent at 𝑥 = 𝑎)
𝑥→𝑎
speed of 80 km/h over a 3 hour time span, but at any given instant within that The slope of the secant can actually be quantified using our slope formula.
span it could be traveling at 50 km/h, 100 km/h, 20 km/h, etc. Recall, the slope of the secant through 𝑎 and 𝑥 can be found by
𝑓(𝑥)− 𝑓(𝑎)
.
𝑥−𝑎

12
In other words, we have: Differentiability
𝑓(𝑥)− 𝑓(𝑎)
𝒇(𝒙)− 𝒇(𝒂) We say that 𝑓(𝑥)is differentiable at 𝑥 = 𝑎 if lim𝑥→𝑎 . exists and is
Slope of tangent at 𝒂: 𝐥𝐢𝐦𝒙→𝒂 . 𝑥−𝑎
𝒙−𝒂
finite. In other words, 𝑓 is differentiable at 𝑥 = 𝑎 if 𝑓 ′ (𝑎) exists.
Or:
𝒇(𝒙)− 𝒇(𝒂) Graphically, 𝑓(𝑥) is differentiable at 𝑥 = 𝑎 if the following three criteria are
Instantaneous rate of change at 𝒂: 𝐥𝐢𝐦𝒙→𝒂 . met:
𝒙−𝒂
 𝑓(𝑥)is continuous at 𝑥 = 𝑎
 𝑓(𝑥)is smooth at 𝑥 = 𝑎
6) Compute the instantaneous rates of change for
1  The tangent at 𝑥 = 𝑎 is not vertical
a) 𝑓(𝑥) = at 𝑥 = −2.
𝑥
b) 𝑓(𝑥) = √2𝑥 + 3 at 𝑥 = 3. 3) For which values of 𝑥 is 𝑓(𝑥) not differentiable?

7) Suppose that the total profit in hundreds of dollars from selling 𝑥 items is
given by 𝑃(𝑥) = 3𝑥 2 − 5𝑥 + 9. Find the instantaneous rate of change of
profit with respect to the number of items produced when 𝑥 = 2. (This
number is called the marginal profit at 𝑥 = 2).

1.6 The Derivative

Limit Definition of Derivative

Another name for the slope of the tangent or the instantaneous rate of
4) Find the values of 𝑎 and 𝑏 so that 𝑓(𝑥) is differentiable at 𝑥 = 0.
change at 𝑥 = 𝑎 is the derivative of 𝒇(𝒙)at 𝒙 = 𝒂, denoted 𝑓 ′ (𝑎) 𝑎𝑥 + 𝑏 𝑥≤0
(said as “𝑓prime of 𝑎”). 𝑓(𝑥) = { 2
𝑥 +𝑥+1 𝑥 >0
Limit Definition of Derivative (at a point)
𝒇(𝒙) − 𝒇(𝒂) Position, Velocity, and Acceleration
𝒇′ (𝒂) = 𝐥𝐢𝐦𝒙→𝒂 .
𝒙−𝒂 If we are given a position function 𝑝(𝑡) (where 𝑡 is time) then we can find
′ (𝑎)
instantaneous velocity at 𝑡 = 𝑐 by finding 𝑝′(𝑐) (since velocity is the rate of
If the limit is infinite, we say that 𝑓 does not exist. change of position over time).
Similarly, since acceleration is the rate of change of velocity over time, we
1) 𝑓(𝑥) = 2𝑥 2 + 3. Compute 𝑓′(3). can find instantaneous acceleration at 𝑡 = 𝑐 by finding 𝑣 ′ (𝑐) (where 𝑣(𝑡) is
the velocity function)
2) Find the equation of the tangent line to 𝑓(𝑥) = 𝑥 2 + 𝑥 at 𝑥 = 1.
5) The position of a particle in metres is given by 𝑝(𝑡) = 3𝑡 2 − 𝑡 + 5 (for 𝑡 in
seconds). Find the velocity of the particle at 2 seconds.

13
 1
The Derivative as a Function b) 𝑓(𝑥) =
𝑥 2 +1
As long as 𝑓(𝑥) is differentiable at 𝑥 = 𝑎, we are able to compute the
c) 𝑓(𝑥) = √𝑔(𝑥) + 2 where𝑔(𝑥) is differentiable
derivative at 𝑥 = 𝑎 by using the limit definition of derivative given earlier.
If we are given a function 𝑓(𝑥) and asked to compute the derivative for
multiples values of 𝑥, computing multiple limits is not practical. Instead, we 7) For which values of 𝑥 is there a horizontal tangent to the graph of
can find an expression for a function that represents the derivative of 𝑓(𝑥) at 𝑓(𝑥) = 𝑥 3 − 4𝑥 + 3 ?
any given 𝑥 value. Naturally, we will denote this function by 𝑓 ′ (𝑥).
8) If a position function is given by 𝑝(𝑡) = 3𝑡 2 − 𝑡 + 5 (in metres, 𝑡 in
Consider the following graph: seconds), then find 𝑣(𝑡) and 𝑎(𝑡) (the velocity and acceleration functions,
respectively).
Use 𝑣(𝑡) and 𝑎(𝑡) to determine the velocity and acceleration of the
particle at 2 seconds.

Instead of taking a limit of secants to a fixed point 𝑎, we will take them to an

arbitrary point 𝑥. Also, instead of using random “second points” that move
towards 𝑥, we will look at second points that are a distance of ℎ away from 𝑥
(so at 𝑥 + ℎ). As these points move towards 𝑥, the value of ℎ will go to zero.
We will be constructing secant lines through the points(𝑥, 𝑓(𝑥)) and
(𝑥 + ℎ, 𝑓(𝑥 + ℎ)).
The slopes of these secant lines can be found by:
𝑓(𝑥 + ℎ) − 𝑓(𝑥) 𝑓(𝑥 + ℎ) − 𝑓(𝑥)
=
(𝑥 + ℎ) − 𝑥 ℎ

And so, the derivative at 𝑥 can be found by the formula:

Limit Defintion of Derivative (as a function)
𝒇(𝒙+𝒉)− 𝒇(𝒙)
𝒇′ (𝒙) = 𝐥𝐢𝐦
𝒉→𝟎 𝒉

6) For the following functions, use the limit definition of derivative to find
𝑓′(𝑥).
a) 𝑓(𝑥) = 𝑥 2 − 4𝑥 + 3

14
 Section 2

Rules for Differentiation

1
Learning Outcomes

By the end of this section, you will be able to:

 Compute the derivative of power functions
 Compute the derivative of polynomial functions
 Compute the derivative of exponential functions
 Compute limits involving trigonometric functions
 Compute the derivative of trigonometric functions
 Compute the derivative of products of functions using the product rule
 Compute the derivative of quotients offunctions using the quotient rule
 Compute the derivative of composite functions using the chain rule
 Compute the derivative of logarithmic functions
 Compute the derivative of inverse trigonometric functions
 Use logarithms to simplify differentiate
 Use implicit differentiation with respect to time to set up a an equation involving rates of change
 Find a rate of change of a quantity given the rate of change of a related quantity.

16
 = lim (𝑛𝑥 𝑛−1 + ℎ𝑃(𝑥, ℎ)) = 𝑛𝑥 𝑛−1 + 0 ⋅ 𝑃(𝑥, ℎ)
2.1 Polynomials and Powers ℎ→0
(by evaluating the limit)
= 𝑛𝑥 𝑛−1 ■
Notation
The derivative of a function, can be written in any of the following ways:
𝑑𝑦 𝑑 Constant Multiples of a Function
𝑓 ′ (𝑥) = = 𝑓(𝑥) = 𝐷𝑥 𝑓(𝑥) If 𝑔′ (𝑥) exists and 𝑓(𝑥) = 𝑘𝑔(𝑥) for 𝑘 in 𝐑, then 𝑓 ′ (𝑥) = 𝑘𝑔′ (𝑥) or:
𝑑𝑥 𝑑𝑥 𝑑 𝑑
We can denote the second derivative (the derivative of the derivative) with: 𝑘 ∙ 𝑔(𝑥) = 𝑘 𝑔(𝑥)
𝑑2 𝑦 𝑑2 𝑑𝑥 𝑑𝑥
𝑓 ′′ (𝑥) = 2 = 2 𝑓(𝑥) = 𝐷𝑥𝑥 𝑓(𝑥)
𝑑𝑥 𝑑𝑥
pf: In the limit definition of derivative for 𝑘𝑔(𝑥), we can factor the 𝑘 from the
Constant Rule limit to be left with 𝑘times the limit definition of derivative for 𝑔(𝑥).
■
If 𝑓(𝑥) = 𝑘, where 𝑘 is any real number, then 𝑓 ′ (𝑥) = 0, or:
𝑑
𝑘=0 3) Compute the derivatives of the following functions:
𝑑𝑥 5
a) 𝑦 = 12𝑥 3 b) 𝑓(𝑥) = 2
𝑥
pf: if 𝑓(𝑥) = 𝑘, then the graph of the function is a horizontal line, where the
slope is always zero. ■ Addition of Functions
If 𝑔′ (𝑥)and ℎ′ (𝑥)both exist and 𝑓(𝑥) = 𝑔(𝑥) + ℎ(𝑥)then
1) 𝑓(𝑥) = 8. Find 𝑓 ′ (𝑥). 𝑓 ′ (𝑥) = 𝑔′ (𝑥) + ℎ′ (𝑥)or:
𝑑 𝑑 𝑑
Power Rule [𝑔(𝑥) + ℎ(𝑥)] = 𝑔(𝑥) + ℎ(𝑥)
𝑑
𝑑𝑥 𝑑𝑥 𝑑𝑥
If 𝑓(𝑥) = 𝑥 𝑛 (𝑛 ≠ 0), then 𝑓 ′ (𝑥) = 𝑛𝑥 𝑛−1 , or 𝑥 𝑛 = 𝑛𝑥 𝑛−1 .
𝑑𝑥
𝑑 𝑛 4) Compute the derivatives of the following functions:
𝑛−1
𝑥 = 𝑛𝑥 a) 𝑦 = 4𝑥 3 − 2𝑥 + 3 b) 𝑔(𝑥) = 𝑎2 −
12
𝑑𝑥 𝑥
𝑥+1
c) 𝑓(𝑥) =
2) Compute the derivatives for the following functions. √𝑥
1
a) 𝑓(𝑥) = 𝑥 8 b) 𝑓(𝑥) = c) 𝑓(𝑥) = √𝑥
𝑥 pf: In the limit definition of derivative for 𝑔(𝑥) + ℎ(𝑥) we can separate into
two limits; one with the 𝑔 terms and one with the 𝑓 terms. This will result in
pf: For any power 𝑛, expanding the expression (𝑥 + ℎ)𝑛 gives: the limit definition of derivatives for 𝑔(𝑥)and ℎ(𝑥)separately. ■
(𝑥 + ℎ)𝑛 = 𝑥 𝑛 + 𝑛𝑥 𝑛−1 ℎ + ℎ2 𝑃(𝑥, ℎ) (where 𝑃(𝑥, ℎ) is a polynomial in 𝑥
and ℎ) **** NOTE: we can find the derivative of a sum of functions by finding the
So, subtracting 𝑥 𝑛 gives: sum of the derivatives, however we cannot extend the same principle to
(𝑥 + ℎ)𝑛 − 𝑥 𝑛 = 𝑛𝑥 𝑛−1 ℎ + ℎ2 𝑃(𝑥, ℎ) products. That is, [𝑓(𝑥)𝑔(𝑥)]′ ≠ 𝑓 ′ (𝑥)𝑔′ (𝑥). We will learn how to deal with
We now have the limit definition of derivative for 𝑓(𝑥) = 𝑥 𝑛 : these in the next section.
𝑓(𝑥 + ℎ) − 𝑓(𝑥) (𝑥 + ℎ)𝑛 − 𝑥 𝑛 5) Find all values of 𝑥 where the tangent is horizontal for the function
𝑓 ′ (𝑥) = lim = lim
ℎ→0 ℎ ℎ→0 ℎ 𝑓(𝑥) = 𝑥 3 + 4𝑥 2 + 4
𝑛𝑥 𝑛−1 ℎ + ℎ2 𝑃(𝑥, ℎ) ℎ(𝑛𝑥 𝑛−1 + ℎ𝑃(𝑥, ℎ))
= lim = lim
ℎ→0 ℎ ℎ→0 ℎ 6) If 𝑔′ (2) = 3, ℎ′ (2) = 5 and 𝑓(𝑥) = 3𝑔(𝑥) − 2ℎ(𝑥), then compute 𝑓 ′ (2).
(by factoring the ℎ’s.)

17
7) Find the parabola with equation 𝑦 = 𝑎𝑥 2 + 𝑏𝑥 whose tangent line at (1, 1) 𝑑 𝑑 𝑑
[𝑔(𝑥)ℎ(𝑥)] = [ 𝑔(𝑥)] ℎ(𝑥) + [ ℎ(𝑥)] 𝑔(𝑥)
is𝑦 = 3𝑥 − 2. 𝑑𝑥 𝑑𝑥 𝑑𝑥

1) Compute the derivative of 𝑓(𝑥) = (𝑥 2 + 1)(2𝑥 2 − 3) in two ways:

Exponential Functions a) By first using the product rule.
To compute the derivative of an exponential function, we will first need to b) By expanding and simplifying and using the power rule
𝑒 ℎ −1
establish the limit: lim =1 pf: Let 𝑓(𝑥) = 𝑔(𝑥)𝑝(𝑥)(I chose 𝑝 so that there is no confusion with the ℎ in
ℎ→0 ℎ
the limit). So, we have:
𝑒 ℎ −1 𝑓(𝑥 + ℎ) − 𝑓(𝑥) 𝑔(𝑥 + ℎ)𝑝(𝑥 + ℎ) − 𝑔(𝑥)𝑝(𝑥)
8) Use a table of values to show that lim = 1. 𝑓 ′ (𝑥) = lim = lim
ℎ→0 ℎ ℎ→0 ℎ ℎ→0 ℎ
now, we add 0 = 𝑔(𝑥)𝑝(𝑥 + ℎ) − 𝑔(𝑥)𝑝(𝑥 + ℎ) to the top:
Using the definition of derivative, as a result we will get that if 𝑔(𝑥 + ℎ)𝑝(𝑥 + ℎ) + [𝑔(𝑥)𝑝(𝑥 + ℎ) − 𝑔(𝑥)𝑝(𝑥 + ℎ)] − 𝑔(𝑥)𝑝(𝑥)
= lim
𝑓(𝑥) = 𝑒 𝑥 , then 𝑓 ′ (𝑥) = 𝑒 𝑥 , or ℎ→0 ℎ
𝑑 𝑥 rearranging gives:
𝑒 = 𝑒𝑥 [𝑔(𝑥 + ℎ)𝑝(𝑥 + ℎ) − 𝑔(𝑥)𝑝(𝑥 + ℎ)] + [𝑔(𝑥)𝑝(𝑥 + ℎ) − 𝑔(𝑥)𝑝(𝑥)]
𝑑𝑥 = lim
In other words, the derivative of 𝑒 𝑥 is itself! ℎ→0 ℎ
factor 𝑝(𝑥 + ℎ) from the left and 𝑔(𝑥) from the right:
9) Differentiate the following: 𝑝(𝑥 + ℎ)[𝑔(𝑥 + ℎ) − 𝑔(𝑥)] + 𝑔(𝑥)[𝑝(𝑥 + ℎ) − 𝑝(𝑥)]
= lim
a) 𝑦 = 4𝑒 𝑥 ℎ→0 ℎ
b) 𝑓(𝑥) = 6𝑥 2 − 12𝑒 𝑥 limit laws allow us to rearrange to:
𝑔(𝑥 + ℎ) − 𝑔(𝑥) 𝑝(𝑥 + ℎ) − 𝑝(𝑥)
= lim [𝑝(𝑥 + ℎ) ] + lim [𝑔(𝑥) ]
10) 𝑓(𝑥) = 𝑒 𝑥 + 𝑥. Find the equation of the tangent at 𝑥 = 0. ℎ→0 ℎ ℎ→0 ℎ
finally, evaluating the limit and recognizing the limit definition of derivative for
pf: The definition of derivative gives: 𝑔 and 𝑝 gives us:
𝑑 𝑥 𝑒 𝑥+ℎ − 𝑒 𝑥 = 𝑝(𝑥 + 0)𝑔′ (𝑥) + 𝑔(𝑥)𝑝′ (𝑥) = 𝑔′ (𝑥)𝑝(𝑥) + 𝑝′ (𝑥)𝑔(𝑥) ∎
𝑒 = lim
𝑑𝑥 ℎ→0 ℎ
using exponent laws we get: 2) Differentiate the following functions using the product rule:
𝑒 𝑥𝑒ℎ − 𝑒 𝑥 𝑒 𝑥 (𝑒 ℎ − 1) 3
a) 𝑓(𝑥) = √𝑥 (2𝑥 + 1)
= lim = lim
ℎ→0 ℎ ℎ→0 ℎ b) 𝑦 = (𝑥 5 + 𝑥 3 + 𝑥)(𝑥 4 + 1)
Limit laws allow us to pull the 𝑒 𝑥 out of the limit to get: c) 𝑔(𝑥) = 𝑥 2 𝑒 𝑥
𝑒ℎ − 1
= 𝑒 𝑥 lim = 𝑒 𝑥 (1) = 𝑒 𝑥 ∎ 3) The force required to move an object can be calculated by 𝐹 = 𝑚 ⋅ 𝑎
ℎ→0 ℎ
where 𝐹 is force in Newtons, 𝑚 is mass in kg and 𝑎 is acceleration in
m/s2. A 5 kg snowball is gaining mass at a rate of 50 g/s as it rolls down
2.2 The Product Rule and Quotient Rules a hill. The snowball is accelerating down the hill at a rate of 4 m/s 2, but
the acceleration is decreasing by
Product Rule 1 m/s2 every 10 seconds. Find the rate of change of the net force that is
If 𝑔′ (𝑥) and ℎ′ (𝑥) exist, and 𝑓(𝑥) = 𝑔(𝑥)ℎ(𝑥) then causing the snowball to roll down the hill.
𝑓 ′ (𝑥) = 𝑔′ (𝑥)ℎ(𝑥) + ℎ′ (𝑥)𝑔(𝑥), or:

18
Quotient Rule As well as the sine addition formula
If 𝑔 ′ (𝑥)
and ℎ ′ (𝑥)
exist, ℎ(𝑥) ≠ 0 and 𝑓(𝑥) =
𝑔(𝑥)
then sin(𝐴 + 𝐵) = sin 𝐴 cos 𝐵 + sin 𝐵 cos 𝐴.
ℎ(𝑥)
𝑔′ (𝑥)ℎ(𝑥)−ℎ′ (𝑥)𝑔(𝑥) sin 𝑥
𝑓 ′ (𝑥) = [ℎ(𝑥)]2
, or: 1) Use a table of values to estimate the value of lim .
𝑥→0 𝑥
𝑑 𝑑
𝑑 𝑔(𝑥) [ 𝑔(𝑥)] ℎ(𝑥) − [ ℎ(𝑥)] 𝑔(𝑥)
= 𝑑𝑥 𝑑𝑥
2) Evaluate the following using limit laws (for c you will need the fact that
𝑑𝑥 ℎ(𝑥) [ℎ(𝑥)]2
cos 2 𝜃 + sin2 𝜃 = 1).
sin(2𝑥) sin(5𝑥) cos(𝑥)−1
a) lim b) lim c) lim
𝑥→0 𝑥 𝑥→0 sin(3𝑥) 𝑥→0 𝑥
The proof for quotient rule is similar to that of the chain rule.
Trigonometric Functions
2−𝑥 2
4) Compute the derivative of 𝑓(𝑥) = in two ways: If 𝑓(𝑥) = sin 𝑥 then 𝑓 ′ (𝑥) = cos 𝑥, or
𝑥
𝑑
a) By first using the quotient rule. sin 𝑥 = cos 𝑥
b) By separating the quotient into two then using power rule. 𝑑𝑥

5) Differentiate the following functions using the quotient rule: pf: Let 𝑓(𝑥) = sin 𝑥. Starting with the definition of derivative we have:
2𝑥+1 𝑥𝑒 𝑥
sin(𝑥 + ℎ) − sin(𝑥)
√𝑥 𝑓 ′ (𝑥) = lim
a) 𝑓(𝑥) = b) 𝑓(𝑥) = c) 𝑔(𝑥) = ℎ→0 ℎ
3𝑥−2 𝑥 2 +1 𝑥 2 +1
Using the addition formula for sine this gives:
6) Suppose that 𝑓(3) = −1, 𝑓 ′ (3) = 2, 𝑔(3) = 4, and 𝑔′ (3) = 1. Compute sin 𝑥 cos ℎ + cos 𝑥 sin ℎ − sin 𝑥
= lim
the following: ℎ→0 ℎ
a) ℎ′ (3) if ℎ(𝑥) = 𝑓(𝑥)𝑔(𝑥) Regrouping and applying limit laws gives:
𝑓(𝑥) sin 𝑥 (cos ℎ − 1) cos 𝑥 sin ℎ
b) ℎ′ (3)if ℎ(𝑥) = = lim + lim
𝑔(𝑥) ℎ→0 ℎ ℎ→0 ℎ
cos ℎ − 1 sin ℎ
= sin 𝑥 lim + cos 𝑥 lim
7) Determine the equation of the tangent at 𝑥 = 1 for the function ℎ→0 ℎ ℎ→0 ℎ
𝑥2 + 1 Using the previous limit results we have:
𝑓(𝑥) = = sin 𝑥 (0) + cos 𝑥 (1) = cos 𝑥 ∎
𝑥−2

8) Find all 𝑥-values where there is a horizontal tangent to the function: Using a similar proof (and the cosine addition formula) we also have that if
𝑒𝑥 𝑓(𝑥) = cos 𝑥 then 𝑓 ′ (𝑥) = − sin 𝑥, or
𝑓(𝑥) = 𝑑
𝑥−1 cos 𝑥 = − sin 𝑥
𝑑𝑥
3) Differentiate the following:
a) 𝑦 = 10 sin 𝑥 b) 𝑓(𝑥) = 𝑒 𝑥 cos 𝑥 c) 𝑔(𝑥) = tan 𝑥
2.3 Derivatives of Trigonometric Functions
4) A 10 foot ladder rests against a vertical wall. Let 𝜃be the angle between
Trigonometric Limits the top of the ladder and the wall and let 𝑥 be the distance from the
To find the derivatives of trigonometric functions, we will need to evaluate bottom of the ladder to the wall. If the bottom of the ladder slides away
𝜋
two limits: from the wall, how fast does 𝑥 change with respect to𝜃 when 𝜃 =
sin 𝑥 cos(𝑥)−1 3
lim and lim
𝑥→0 𝑥 𝑥→0 𝑥

19
5) Find the equation of the tangent line at 𝑥 = 𝜋 for the function 1) Split the composite functions into two simple functions.
1
𝑓(𝑥) = 𝑥 cos 𝑥. a) 𝑓(𝑥) = (3𝑥 + 1)2 b) 𝑦 = 4 c) 𝑔(𝑥) = 𝑒 cos 𝑥
(√𝑥+3)

We can use the quotient rule to show that if 𝑓(𝑥) = tan 𝑥, then
1
𝑓 ′ (𝑥) = 2 , or Chain Rule
cos 𝑥 If 𝑔(𝑥) and ℎ(𝑥) are differentiable functions and 𝑓(𝑥) = 𝑔(ℎ(𝑥))then
𝑑 1
tan 𝑥 = = sec 2 𝑥 𝑓 ′ (𝑥) = 𝑔′ (ℎ)ℎ′ (𝑥), or:
𝑑𝑥 cos 2 𝑥
𝑑 𝑑 𝑑 𝑑𝑔 𝑑ℎ
6) Find the location of all horizontal tangents to the curve 𝑔(ℎ(𝑥)) = 𝑔(ℎ) ℎ(𝑥) = ⋅
𝑑𝑥 𝑑ℎ 𝑑𝑥 𝑑ℎ 𝑑𝑥
𝑦 = 2𝑥 − tan 𝑥 for 𝑥 in [0,2𝜋).
That is, “the derivative of the outside function times the derivative of the
Similarly, we can differentiate the secondary trig functions to get: inside function”.
𝑑
sec 𝑥 = tan 𝑥 sec 𝑥 2) Differentiate the functions:
𝑑𝑥 1
𝑑 a) 𝑓(𝑥) = (3𝑥 + 1)2 b) 𝑦 = 4
csc 𝑥 = − cot 𝑥 csc 𝑥 (√𝑥+3)
𝑑𝑥 c) 𝑔(𝑥) = 𝑒 cos 𝑥
d) 𝑦 = sin 4 (𝑥)
𝑑
cot 𝑥 = − csc 2 𝑥
𝑑𝑥 pf: Let 𝑓(𝑥) = 𝑔(𝑝(𝑥)). By the definition of derivative, we have:
𝑔(𝑝(𝑥 + ℎ)) − 𝑔(𝑝(𝑥))
7) Differentiate the following: 𝑓 ′ (𝑥) = lim
tan 𝑥+1 ℎ→0 ℎ
a) 𝑦 = 𝑒 𝑥 csc 𝑥 b) 𝑓(𝑥) =
cot 𝑥+1
𝑝(𝑥+ℎ)−𝑝(𝑥)
We can multiply by 1 = to get:
𝑝(𝑥+ℎ)−𝑝(𝑥)
𝑔(𝑝(𝑥 + ℎ)) − 𝑔(𝑝(𝑥)) 𝑝(𝑥 + ℎ) − 𝑝(𝑥)
2.4 Chain Rule 𝑓 ′ (𝑥) = lim ∙
ℎ→0 ℎ 𝑝(𝑥 + ℎ) − 𝑝(𝑥)
𝑔(𝑝(𝑥+ℎ))−𝑔(𝑝(𝑥)) 𝑝(𝑥+ℎ)−𝑝(𝑥)
Composite Functions = lim ∙
ℎ→0 𝑝(𝑥+ℎ)−𝑝(𝑥) ℎ
𝑔(𝑝(𝑥+ℎ))−𝑔(𝑝(𝑥)) 𝑝(𝑥+ℎ)−𝑝(𝑥)
Recall, a composite function 𝑓(𝑥) = 𝑔(ℎ(𝑥)) is a function where 𝑓(𝑥) is = lim ∙ lim
ℎ→0 𝑝(𝑥+ℎ)−𝑝(𝑥) ℎ→0 ℎ
formed by “plugging ℎ(𝑥) into 𝑔(𝑥)”.
Since 𝑝(𝑥) is the “input” for the 𝑔(𝑝) function, the limit definition of derivative
For example, if ℎ(𝑥) = 𝑥 2 + 1 and 𝑔(𝑥) = √𝑥 we get gives us:
𝑔(ℎ(𝑥)) = 𝑔(𝑥 2 + 1) = √𝑥 2 + 1 = 𝑔′(𝑝) 𝑝′(𝑥) ∎

The Chain Rule will give us a tool for differentiating composite functions, but 3) 𝑓(3) = 1, 𝑔(2) = 3, 𝑓 ′ (3) = 4, and 𝑔′ (2) = −2. If ℎ(𝑥) = 𝑓(𝑔(𝑥)), then
first we will have to be able to identify such functions and express them in evaluate ℎ′ (2).
terms of their two simpler functions, the “inside function” and the “outside
function”. 4) Differentiate the following functions:
2
𝑒𝑥
If 𝑓(𝑥) = √𝑥 2 2
+ 1, then we could have ℎ(𝑥) = 𝑥 + 1 as the “inside function” a) 𝑓(𝑥) = b) 𝑦 = √3𝑥 + 1 cos(𝑒 𝑥 )
𝑥 2 +1
and √ℎ as the “outside function”. Usually, the inside function is enclosed in 𝑥
brackets or something similar. 5) Use the fact that 𝑎 𝑥 = 𝑒 ln 𝑎 = 𝑒 𝑥 ln 𝑎 to differentiate 𝑦 = 𝑎 𝑥 .

20
6) The motion of a spring that is subject to a frictional force or a damping In practice, when we differentiate these “functions of 𝑦” we can just
force (such as a shock absorber in a car) is often modeled by the product 𝑑𝑦
differentiate as if 𝑦 was our variable, then affix a “ ” to the end of the
𝑑𝑥
of an exponential function and a sine or cosine function. Suppose the
expression.
equation of motion of a point on such a spring is
𝑠(𝑡) = 2𝑒 −1.5𝑡 sin 2𝜋𝑡 𝑑𝑦
where 𝑠 is measured in centimetres and 𝑡 in seconds. Find the velocity 3) Find for the following relations:
𝑑𝑥
after 𝑡 seconds. a) 𝑥 2 + 𝑦 3 = 2
2
b) 𝑒 𝑥 + sin(𝑦 2 + 1) = √ 𝑦
c) 𝑥 2 𝑦 4 = 𝑒 2𝑦
2.5 Implicit Differentiation
𝑑𝑦
When we solve for , it will usually
𝑑𝑥
Implicit Differentiation be in terms of both 𝑥 and 𝑦. The
Up to this point, we have been concerned with differentiating functions. value (for a particular 𝑥 and 𝑦 which
For example, if 𝑓(𝑥) = 3𝑥 2 − 2𝑥 + 1, we would have 𝑓 ′ (𝑥) = 6𝑥 − 2. are on the original curve) will
Since 𝑓 ′ (𝑥) is the derivative of 𝑓(𝑥) and (6𝑥 − 2) is the derivative of represent the slope of the tangent
(3𝑥 2 − 2𝑥 + 1), we can see that we have simply “applied differentiation” to at the point (𝑥, 𝑦) (which must be on
both sides of the equation. This is consistent with our fundamental rule of the original curve
algebra (do the same thing to both sides of an equation).
4) Find the equation of all tangent lines to the curve at 𝑥 = 1:
Suppose the same expression were written as 𝑓(𝑥) + 2𝑥 = 3𝑥 2 + 1 instead. 𝑦 2 𝑥 + 𝑥 3 = 1 + 𝑦𝑥 2
We can differentiate both sides of the equation to get:
𝑑 𝑑
[𝑓(𝑥) + 2𝑥] = [3𝑥 2 + 1] evaluating the derivatives gives: 𝑑2𝑦
𝑑𝑥 𝑑𝑥 5) 𝑥 3 + 𝑦 3 = 1. Find (the second derivative of 𝑦).
𝑑𝑥 2
𝑓 ′ (𝑥) + 2 = 6𝑥 solving for 𝑓 ‘(𝑥) we have
𝑓 ′ (𝑥) = 6𝑥 − 2 6) Find the locations of any horizontal and vertical tangents to the curve:
We get the same thing! 𝑥 2 + 2𝑥𝑦 + 4𝑦 2 = 48
𝑑𝑦
1) 𝑦 + 2𝑥 = 𝑒 𝑥 + 1. Find
𝑑𝑥

We can use this reasoning to differentiate relationships between 𝑥 and 𝑦 that

2.6 Inverse Trigonometric Functions and Logarithms
𝑑
cannot be expressed as functions by differentiating (applying to) both
𝑑𝑥 Inverse Trigonometric Functions
sides of an equation. Consider 𝑦 = arcsin 𝑥. Equivalently, we can write this function as 𝑥 = sin 𝑦.
𝑑𝑦
For the purposes of differentiation, we can think of 𝑦 as being a function of 𝑥, From here, we can implicity differentiate the equation to solve for ⁄𝑑𝑥 .
so 𝑦 = 𝑦(𝑥).
𝑑𝑦 1) Use implicit differentiation to differentiate arcsin 𝑥, arccos 𝑥, and arctan 𝑥.
So, 𝑦 ′ (𝑥) = (it’s the derivative of 𝑦 with respect to 𝑥)
𝑑𝑥
𝑑 𝑑 𝑑𝑦
Also, 𝑔(𝑦) = 𝑔(𝑦(𝑥)) = 𝑔′ (𝑦)𝑦 ′ (𝑥) = 𝑔′ (𝑦) ( )
𝑑𝑥 𝑑𝑥 𝑑𝑥

2) Differentiate 𝑦 3 with respect to 𝑥.

21
As a result, we will get the following results: We can now use logarithm laws to simplify the right hand side to get:
𝑑 1 ln 𝑓(𝑥) = 5 ln(2𝑥 + 1) + 3 ln(3𝑥 2 − 1) + ln 𝑒 𝑥+2
arcsin 𝑥 = = 5 ln(2𝑥 + 1) + 3 ln(3𝑥 2 − 1) + (𝑥 + 2)
𝑑𝑥 √1 − 𝑥 2
𝑑 −1 From here, implicit differentiation is fairly simple.
arccos 𝑥 =
𝑑𝑥 √1 − 𝑥 2
𝑑 1 8) Differentiate the following:
arctan 𝑥 = (4𝑥+1)3 (2𝑥−3)7
𝑑𝑥 1 + 𝑥2 a) 𝑓(𝑥) = (2𝑥 + 1)5 (3𝑥 2 − 1)3 𝑒 𝑥+2 b) 𝑔(𝑥) = √
𝑥 2 +1

2) Differentiate the following:

a) 𝑓(𝑥) = 𝑥 arccos 𝑥 b) 𝑦 = arctan(𝑒 𝑥 ) 9) Use logarithmic differentiation to differentiate the following:
2
a) 𝑓(𝑥) = 𝑥 𝑥 b) 𝑔(𝑥) = (cos 𝑥) 𝑥 .
Logarithms
3) Differentiate 𝑦 = ln 𝑥 (𝑥 > 0) and 𝑦 = ln(−𝑥) (𝑥 < 0).
1 2.7 Related Rates
If 𝑓(𝑥) = ln |𝑥|, then 𝑓 ′ (𝑥) = or
𝑥
𝑑 1 Related Rates
ln |𝑥| =
𝑑𝑥 𝑥 Related Rates problems are applications involving rates of change of a value
The proof of this is a combination of the results from question 9. over time.
Intuitively, when we speak of a rate of change we are generally referring to
In practice, we will typically focus on positive valued 𝑥and ignore the absolute the change over time unless otherwise specified.
value symbol in the logarithm.
1) The population on a small island can be found by the function
4) Differentiate the following:
a) 𝑓(𝑥) = 6 ln 𝑥 b) 𝑔(𝑥) = 𝑒 𝑥 ln 𝑥 c) 𝑦 =
𝑥 15,000𝑒 0.02𝑡
ln 𝑥 𝑃(𝑡) =
5) Find the locations of all horizontal tangents to the curve 5𝑒 0.02𝑡 − 2
𝑦 = 𝑥 ln|𝑥|. (where 𝑡 is the number of years after 2012).
Find how quickly the population will be decreasing in 2025.
6) Differentiate the following:
3 2
a) 𝑓(𝑥) = ln(𝑥 2 + 𝑥 + 1) b) 𝑓(𝑥) = ln( √𝑥 ) c) 𝑦 = 𝑒 𝑥 ln(𝑥 2 + 1) If two quantities are related to each other by some equation, and one
quantity is changing then this will affect the other quantity. If we know how
7) Differentiate 𝑦 = log 𝑎 𝑥 quickly one quantity changes then we can find how quickly the other quantity
changes using calculus.
Logarithmic Differentiation
We can use logarithms to simplify differentiation that would require heavy If 𝑥 and 𝑦 are our quantities with 𝑦 = 𝑓(𝑥), and we know that 𝑥 is changing at
𝑑𝑥
use of product rule or quotient rule. a rate of 2 units/second then we can express this as (the rate of change of
𝑑𝑡
𝑑𝑦
𝑥 over time). We can solve for the rate of change of 𝑦 over time ( ) by
For example, consider 𝑓(𝑥) = (2𝑥 + 1)5 (3𝑥 2 − 1)3 𝑒 𝑥+2 𝑑𝑡
To differentiate 𝑓(𝑥), we would have to apply a triple product rule, as well as differentiating the equation with respect to 𝑡 (we differentiate with respect to t
multiple chain rules. Instead, we can take the logarithm of both sides to get: because the “𝑑𝑡” on the bottom of the derivative expression).
ln 𝑓(𝑥) = ln[(2𝑥 + 1)5 (3𝑥 2 − 1)3 𝑒 𝑥+2 ]
22
When we “differentiate with respect to 𝑡”, we do this using implicit You may have to make substitutions for some variables using existing
differentiation. So, relationships. In particular, if triangles show up in your sketch then look
𝒅 𝒅𝒚 𝒅 𝒅𝒙 out for similar triangles and pythagorean theorem
[𝒈(𝒚)] = 𝒈′ (𝒚) [𝒉(𝒙)] = 𝒉′ (𝒙) .
𝒅𝒕 𝒅𝒕 𝒅𝒕 𝒅𝒕
5. Differentiate the relationship with respect to 𝑡.
𝜋
2) 𝑦 = cos(3𝑥 + 𝜋). When 𝑥 = , 𝑥 is increasing by 2 units per second. 6. Plug in all values except for the unknown rate of change.
2
Find how quickly 𝑦 is changing at this time.
Here, you can finally plug in the known values. You may have to use the
Using this, we can also solve related rates problems for non-function original equation or other relationships to solve for some of the
relationships between 𝑥 and 𝑦. We simply differentiate implicitly with respect unknowns
to 𝑡! 7. Solve for the rate of change you are looking for.

3) 𝑦 2 − 2𝑥𝑒 𝑦 + 𝑥 2 𝑦 = 4. At the point (−2, 0), 𝑥 is decreasing by 3 units per 5) A spherical snowball with radius 8 cm is melting such that its radius is
minute. Find the rate of change of 𝑦. shrinking by 4 mm/min. At what rate is the volume of the snowball
decreasing?
4) If two resistors with resistance 𝑅1 and 𝑅2 are connected in parallel, then 6) A 20 meter ladder is rested against a wall. The ladder begins to slip
the total resistance 𝑅, measured in ohms (Ω), is given by such that when the ladder reaches 16 metres high, the base is slipping
away at a rate of 10 meters per minute.
1 1 1 a) At what rate is the height of the ladder decreasing?
= +
𝑅 𝑅1 𝑅2 b) At what rate is the angle between the ladder and the ground
decreasing?
If 𝑅1 and 𝑅2 are increasing at rates of 0.3 Ω/s and 0.2 Ω/s, respectively, 7) An ice cream cone (whose shape makes a perfect cone) has radius 2 cm
how fast is 𝑅 changing when 𝑅1 = 80 Ω and and height 8 cm. Ice cream is leaking from the bottom of the cone at a
𝑅2 = 100 Ω? rate of 3 cm3 each minute. When the depth of the remaining ice cream in
the cone is 2 cm, at what rate is the depth changing?
Strategy for Solving Related Rates Problems
1. If possible, sketch the scenario using variables for the quantities whose 8) Car A is 120 km North of point P and headed North. Car B is currently
rates have been mentioned in the question . 50 km East of point P and driving towards P. If Car A is driving at 50
It is recommended to not put known values in the place of any of km/h and car B is driving at 60 km/h, then what is the rate of change of
these variables. Since these values can change, they should stay as the distance between them?
variables; otherwise, when we differentiate they will be treated as if
9) A runner sprints around a circular track of radius 100 m at a constant
they are constants. In general, the only values you should include in speed of 7 m/s. The runner’s friend is standing at a distance 200 m from
your sketch are constants. the center of the track. How fast is the distance between the friends
2. Identify/label the rate you are looking for. changing when the distance between them is 200 m?

3. Identify/label all rates that you are given.

4. Determine a relationship that uses only the variables from the last few
steps.
Again, do not plug in known values unless they are constants.

23
 Section 3

Applications of
Derivatives

24
Learning Outcomes

By the end of this section, you will be able to:

0 ∞
 Use L’Hopital’s Rule to solve limits of the form: “ ”, “ ”, “∞ · 0”, “∞0 ”, “0∞ ”, “1∞ ”
0 ∞
 Determine where 𝑓 ′ > 0 or 𝑓 ′ < 0 given the graph of a function.
 Find the critical numbers of a function.
 Find the intervals of increase and decrease for a function.
 Find the local extrema of a function by analyzing the first derivative.
 Determine where 𝑓 ′′ > 0 or 𝑓 ′′ < 0 given the graph of a function.
 Find the intervals of concavity for a function.
 Find the inflection points of a function.
 Find the local extrema of a function by using the second derivative test on critical points.
 Sketch the graph of a function using information regarding domain, intercepts, asymptotes, intervals of increase and decrease, and intervals of concavity.
 Find the absolute minimum or maximum of a function from its graph
 Find the absolute minimum or maximum of a continuous function over a closed and bounded interval
 Find the absolute minimum or maximum of a continuous function when there is a single critical point on an interval.
 Solve for the maximum or minimum value of a quantity in an application.
 Use differentials to approximate the change of 𝑦 that occurs as a result of changing 𝑥.
 Estimate the value of a function at a point using a linear approximation at a nearby point
 Use Newton’s Method to approximate the zeroes of a function.
 Use Rolle’s Theorem to show the existence of a critical number over an interval.
 Find an upper bound on the number of zeroes to a function.
 Use the Mean Value Theorem to show the existence of a point 𝑥 = 𝑐 in an interval (𝑎, 𝑏) such that 𝑓 ′ (𝑐) is the secant slope between 𝑥 = 𝑎 and 𝑥 = 𝑏.

25
3.1 L’Hopital’s Rule 2) Evaluate the limit lim
𝑥 2 −9
by:
𝑥→3 𝑥 2 −2𝑥−3
a) using L’hopital’s rule
Indeterminate Forms b) factoring and cancelling
When evaluating the limits lim 𝑓(𝑥), recall that the first step is usually to see
𝑥→𝑎
if 𝑓(𝑎) exists, and if it does then this is the value of the limit. 𝑒𝑥− 𝑥 − 1
3) Evaluate the limit lim
In some cases, 𝑓(𝑎) does not exist, but we are able to immediately 𝑥→0 𝑥2
1
determine that the limit is infinite (for example, if the form is “ ”). 4
0 4) Show that L’hopital’s rule doesn’t work for the limit lim+ and explain
In other cases, we have to do algebraic manipulations in order to cancel out 𝑥→2 𝑥−2
the factors that are causing 𝑓(𝑥) to not exist at 𝑥 = 𝑎. (for example, if the why the result is consistent with the rule.
0 0
form is “ ”. The form “ ” is called an indeterminate form because we cannot
0 0 We can also apply l’hopital’s rule to evaluate infinite limits.
immediately determine if the limit will be finite, infinite, or exist at all.
3𝑥 2 −2𝑥+1
If “plugging in 𝑎” (𝑎 can be infinite, but the idea of “plugging it in” makes less 5) Evaluate the limit lim by:
𝑥→∞ 2𝑥 2 −3𝑥+4
0 ∞
sense) gives any of the following forms: “ ”, “ ”, “∞ · 0”, “∞0 ”, “00 ”, “1∞ ”, then a) using L’hopital’s rule
0 ∞
b) dividing the top and bottom by 𝑥 2 and simplifying.
we say that the form is indeterminate.
ln 𝑥
L’Hopital’s Rule 6) Evaluate the limit lim .
𝑥→∞ 𝑥
𝑒 𝑥 −1 1–1 0
Consider the limit lim . If we plug in 𝑥 = 0 we get = ; however we will
𝑥→0 𝑥 0 0
Applying L’Hopital’s Rule
not be able to factor the top in order to cancel the 𝑥 on the bottom. Instead,
In order to use L’Hopital’s Rule, the function in the limit must be of the form
we can evaluate this limit using a technique called L’hopital’s rule. 𝑓(𝑥)
If lim 𝑓(𝑥) = lim 𝑔(𝑥) = 0 or lim 𝑓(𝑥) = lim 𝑔(𝑥) = ±∞ , .
𝑥→𝑎 𝑥→𝑎 𝑥→𝑎 𝑥→𝑎 𝑔(𝑥)
𝑓(𝑥) 𝑓 ′(𝑥)
then lim = lim
𝑥→𝑎 𝑔(𝑥) 𝑥→𝑎 𝑔′ (𝑥) “0 · ∞”
Suppose we have lim 𝑓(𝑥)𝑔(𝑥) where 𝑓(𝑎) = 0 and 𝑔(𝑎) → ∞; we can
𝑒 𝑥 −1 𝑥→𝑎
1) Evaluate the limit lim using L’hopital’s rule. rewrite this limit as either lim
𝑓(𝑥)
(to give “0/0”) or as lim
𝑔(𝑥)
(to give
𝑥→0 𝑥
𝑥→𝑎 1/𝑔(𝑥) 𝑥→𝑎 1/𝑓(𝑥)
“∞/∞”).
Pf: We will prove the case where 𝑓(𝑎) = 𝑔(𝑎) = 0 and 𝑓 and 𝑔 are both
continuous. (proving cases other than this are quite tedious and subtle work 7) Evaluate the following limits:
with the limits)
a) lim+ 𝑥 ln 𝑥 b) lim+ 𝑡 2 𝑒 1/𝑡
We have 𝑥→0 𝑡→0
𝑓(𝑥)−𝑓(𝑎)
𝑓 ′ (𝑥) [ ] 𝑓(𝑥) − 𝑓(𝑎)
lim ′ 𝑥−𝑎
= lim 𝑔(𝑥)−𝑔(𝑎) = lim “∞0 ”, “00 ”, “1∞ ”
𝑥→𝑎 𝑔 (𝑥) 𝑥→𝑎 𝑥→𝑎 𝑔(𝑥) − 𝑔(𝑎) These are indeterminate since, when we “plug in 𝑎”, the power rule that
[ ]
𝑥−𝑎
But since 𝑓(𝑎) = 𝑔(𝑎) = 0, this gives should get applied for the base contradicts the rule that exponent rule that
𝑓 ′ (𝑥) 𝑓(𝑥) − 0 𝑓(𝑥) should get applied for the exponent. For example, with “00 ”, 0 to the power
lim ′ = lim = lim ∎ of anything should always be zero, but anything to the power of 0 should be
𝑥→𝑎 𝑔 (𝑥) 𝑥→0 𝑔(𝑥) − 0 𝑥→0 𝑔(𝑥)
one (except for 0, of course). We can also deal with these by rewriting the
limits and applying l’Hopital’s.

26
 1) Find the intervals of increase and decrease for the sketch of
Suppose we have lim 𝑓(𝑥) 𝑔(𝑥) where we have one of the forms above. We 𝑦 = 𝑓(𝑥) below
𝑥→𝑎
can solve this by first solving the logarithm of this limit. This will allow us to
“pull the 𝑔(𝑥) down”.
lim ln(𝑓(𝑥) 𝑔(𝑥) ) = lim 𝑔(𝑥) ln 𝑓(𝑥)
𝑥→𝑎 𝑥→𝑎
Notice that this turns “∞0 ” into “0 · ∞”, “00 ” into “0 · ∞”, and “1∞ ” into “∞ · 0”.
We can now solve each of these limits by using the previous technique
(writing as the product as a fraction, then applying l’hopital’s). Suppose our
result is 𝐿.
To find the original limit, we simply find 𝑒 𝐿 , since limit laws and the properties
of logarithms and exponentials give:
lim ln(𝑓(𝑥)𝑔(𝑥) ) ln( lim 𝑓(𝑥)𝑔(𝑥) )
𝑒 𝐿 = 𝑒 𝑥→𝑎 =𝑒 𝑥→𝑎 = lim 𝑓(𝑥) 𝑔(𝑥) Critical Numbers
𝑥→𝑎
A critical number for 𝑓(𝑥) is a number in the domain of 𝑓 where 𝑓 ′ (𝑥) = 0 or
8) Evaluate the following limits: 𝑓 ′ (𝑥) does not exist. The derivative of a continuous function is continuous
a) lim 𝑥 1/𝑥 b) lim 𝑥 𝑥
2 where defined, and so it can only change signs either where 𝑓 ′ (𝑥) = 0 or
𝑥→∞ 𝑥→0+ 𝑓 ′ (𝑥) does not exist; that is why these numbers are so crucial.
2 3𝑥
c) lim (1 + ) .
𝑥→∞ 𝑥
In order to find where a function is increasing or decreasing, we can first find
the critical numbers by finding where 𝑓 ′ (𝑥) = 0 or 𝑓 ′ (𝑥) does not exist. We
***Be sure to note the difference between the use of l’Hopital’s Rule can then use these numbers to break the entire real number line into
and the quotient rule for differentiating. intervals, on which the sign will be entirely positive or entirely negative.

3.2 Local Extrema

2) For the sketch of 𝑦 = 𝑓(𝑥) in question 1), find the critical numbers.
Intervals of Increase and Decrease
3) Find the critical numbers for the following functions:
We say that a function 𝑓(𝑥) is increasing on an interval if for any 𝑥1 and 𝑥2
a) 𝑔(𝑥) = 3𝑥 2 + 2𝑥 − 1
in that interval with 𝑥2 > 𝑥1 we have 𝑓(𝑥2 ) > 𝑓(𝑥1 ) (in other words, the 2𝑥
function is “going up” from left to right). The function is decreasing if 𝑓(𝑥1 ) > b) 𝑓(𝑥) = (𝑥−1)2
𝑓(𝑥2 ) (or the function is “going down” from left to right).
For example, if we find critical numbers at 𝑥 = −2, 1, 5 we would have the
If 𝑓 ′ (𝑥) > 0 for all 𝑥 in the interval, then 𝑓 is increasing on the interval. (Since
following number line:
slopes are positive)
If 𝑓 ′ (𝑥) < 0 for all 𝑥 in the interval, then 𝑓 is decreasing on the interval. (Since
slopes are negative)
If 𝑓 ′ (𝑥) = 0 for all 𝑥 in the interval, then 𝑓 is constant on the interval. (Since We can see that this breaks R into the intervals (−∞, −2), (−2, 1), (1, 5), and
the slope is neither positive nor negative) (5, ∞).

To determine the sign of the interval, we just need to plug in a “test value”
that occurs in that interval.

27
For the previous intervals we could use the test numbers:
−3 in (−∞, −2); 0 in (−2, 1);
3 in (1, 5); 10 in (5, ∞).

We then determine if 𝑓 ′ (𝑥) is positive or negative for these numbers and it

will tell us the sign of 𝑓 ′ over the entire interval. ***Note: We define “nearby” to mean all of the points in any open interval
containing 𝑥 = 𝑐. In the third example, we see that the point is not the
Suppose we get the following greatest of all points, but just the ones nearby.

6) Find all local extrema for the sketch of

𝑦 = 𝑓(𝑥) below

The intervals on which 𝑓 ′ (𝑥) is positive are intervals of increase and the
intervals on which 𝑓 ′ (𝑥) is negative are intervals of decrease.

According to the example, we would have (−∞, −2) and (5, ∞) as intervals of
decrease and (−2, 1) and (1, 5) as intervals of increase. Notice that the
endpoints are not included in the intervals, since, at that instant, the function
is constant (even in the case of 𝑥 = 1, where the function is increasing on
both sides). First Derivative Test for Local Extrema
If 𝑓(𝑥) is continuous on any open interval (𝑎, 𝑏) then local extrema can only
4) Find the intervals of increase and decrease for the following functions: occur at 𝑥 = 𝑐 if 𝑐 is a critical point.
a) 𝑔(𝑥) = 3𝑥 2 + 2𝑥 − 1 If 𝑓 is decreasing on (𝑎, 𝑐) and increasing on (𝑐, 𝑏) then 𝑥 = 𝑐 is a local
2𝑥
b) 𝑓(𝑥) = (𝑥−1)2 minimum. ⋱𝑓(𝑐) ⋰
If 𝑓 is increasing on (𝑎, 𝑐) and decreasing on (𝑐, 𝑏) then 𝑥 = 𝑐 is a local
5) Population density near a heavily polluting factory can be found by the maximum.⋰ 𝑓(𝑐) ⋱
function 𝑃(𝑥) where 𝑃 is the number of people per square kilometer (in
thousands)and 𝑥 is the distance from the factory in kilometers. When finding the y-value of the local extrema, be sure to plug 𝑐 back into the
𝑃(𝑥) = (5 − 2𝑥)𝑒 𝑥 − 4, 0 < 𝑥 < 2. original function and not the derivative (plugging it into the derivative will
At what distances are the populations increasing as you move away from always return “0” or “DNE”).
the factory?
7) Find local extrema for the following functions:
Local Extrema a) 𝑔(𝑥) = 3𝑥 2 + 2𝑥 − 1
2𝑥
A local (or relative) maximum is a point (𝑐, 𝑓(𝑐)) such that 𝑓(𝑐) is greater b) 𝑓(𝑥) = (𝑥−1)2
than all of the “nearby” points.
Similarly, a local (or relative) minimum is a point (𝑐, 𝑓(𝑐)) such that 𝑓(𝑐) is 8) Find the coordinates of the local extrema for the following functions:
less than all of the “nearby” points. a) 𝑦 = 𝑥 3 − 𝑥 2 − 𝑥 − 6
2
b) 𝑔(𝑥) = 𝑥𝑒 −2𝑥
2
c) 𝑓(𝑥) = ln |𝑥 − 1|.
d) ℎ(𝑥) = sec 𝑥 + tan 𝑥 0 ≤ 𝑥 ≤ 2𝜋
28
3.3 Concavity If we look closer at the
graph of a concave up
Concavity of a Graph function and its slopes, we
The concavity of a graph is a term used to describe the curvature. see that as we move
We say that a function is concave down if it bends in the downwards from left to right, the
direction, as pictured below: value of the slopes
becomes more and
more positive. That is, the
slope (or the first
derivative) is increasing. If the first derivative is increasing, then the second
derivative (the derivative of the first derivative) is positive.
and we say that a function is concave up if it bends in the upwards direction,
as pictured below: A function is concave up on an
interval if 𝑓 ′′ (𝑥) > 0 for all 𝑥 in the
interval.
A function is concave down on an
interval if 𝑓 ′′ (𝑥) < 0 for all 𝑥 in the
Notice that a function is not necessarily increasing or decreasing if it is
interval.
concave up or down.
Finding intervals of concavity of a function is very similar to finding the
Another way of looking at concavity is that if the curve occurs below the
intervals of increase/decrease.
tangent, then it is concave down and if it occurs above the tangent then it is
The only way for the sign of the second derivative to change is if
concave up.
𝑓 ′′ (𝑥) = 0 or if 𝑓 ′′ (𝑥) does not exist, so we first find these points.
Then, we break 𝐑 into open intervals using these numbers and determine the
sign of 𝑓 ′′ (𝑥) on each interval by using a test point within the interval.

2) Find intervals of concavity for the function𝑓(𝑥) = ln |𝑥 2 − 4|.

Inflection Points
1) Find the intervals of concavity for the function sketched below: An inflection point is a point on the curve 𝑦 = 𝑓(𝑥) where the concavity
changes from positive to negative or vice versa. It is the second derivative
equivalent of a local extrema.

3) Find the inflection points for the function sketched below:

29
4) Find the inflection points for the following functions: C vertical asymptotes (demoninator = 0 for rational, inside = 0 for
a) 𝑓(𝑥) = 3𝑥 4 + 10𝑥 3 − 12𝑥 2 + 4𝑥 − 5 logarithm)
2
b) 𝑦 = 𝑥𝑒 −𝑥 horizontal asymptotes (limit of function to ±∞)
D intervals of inc/decrease and coordinates of any local extrema
Second Derivative Test for Local Extrema
If 𝑓 ′ (𝑐) = 0 and 𝑓 ′′ (𝑐) > 0, then 𝑓(𝑐) is a local minimum. E intervals of concavity and the coordinates of any inflection points
If 𝑓 ′ (𝑐) = 0 and 𝑓 ′′ (𝑐) < 0, then 𝑓(𝑐) is a local maximum.
If 𝑓 ′ (𝑐) = 0 and 𝑓 ′′ (𝑐) = 0 or does not exist, then no conclusion can be
made. Once this information is gathered, we can sketch the graph to match this
information. A good strategy is to follow this order:
1. Plot ALL found points: local extrema, inflection points, etc.
2. Sketch all asymptotes using a dotted line.
3. Connect the points using a simple smooth curve, being sure to bring the
curve “close” to asymptotes and not to connect points where the function
is not defined (ie, never cross a vertical asymptote).
5) Determine all local mins/maxs using the second derivative test if 4. For regions that are not “between points” or guided by asymptotes, use
possible; if the second derivative test fails, use the first derivative test to your intervals of increase/decrease and concavity to extend the curve.
analyze the points. 5. Check that your graph matches the data that you’ve collected. You may
4 7
a) 𝑓(𝑥) = 𝑥 4 − 8𝑥 2 − 5 b) 𝑦 = 14𝑥 3 − 𝑥 3 have to make small adjustments to perfect the sketch (particularly with
concavity).
** note: if 𝑓 ′′ (𝑐) = 0, it is not necessarily true that the point is neither a
minimum or maximum. Consider the functions 𝑓(𝑥) = 𝑥 3 and
𝑓(𝑥) = 𝑥 4 . 1) Sketch the function 𝑓(𝑥) = 𝑥 3 − 6𝑥 2 − 15𝑥 + 14.
In both cases we get 𝑓 ′ (0) = 0, and 𝑓 ′′ (0) = 0, but in 𝑥 4 we have a minimum
𝑥
of 0 at 𝑥 = 0 however in 𝑥 3 we have neither a maximum or minimum at 𝑥 = 2) Sketch the function 𝑓(𝑥) = (2
– 𝑥)2
.
0. When we get 𝑓 ′′ (𝑐) = 0 we should use the first derivative test to analyze
the point. 3) Sketch the function 𝑓(𝑥) = ln |𝑥 2 − 1|.

4) Sketch the function 𝑓(𝑥) = 2𝑥 − cos 𝑥 over the interval [0, 2𝜋].
3.4 Graphing
𝑒𝑥
5) Sketch the graph of the function 𝑓(𝑥) =
We can combine the information from this section, as well as information 1+2𝑒 𝑥

regarding domain, intercepts, and asymptotes to sketch a reasonably

accurate sketch of any graph.

When we graph a function, we will make note of the following information:

A domain of the function
B any key points
30
6) Sketch a graph of 𝑦 = 𝑓(𝑥) where 𝑓(𝑥) has all of the following properties: 2) Find the absolute maximum or minimum (if they exist) for the following
functions by sketching them on the specified intervals
 𝑓 is continuous and differentiable for all 𝑥 except 𝑥 = 3. 1
a) 𝑓(𝑥) = 3(𝑥 – 2)2 + 5 b) 𝑓(𝑥) = on [1, ∞)
𝑥
 𝑓 has intercepts at (0,0) and (2,0). c) 𝑓(𝑥) = −|𝑥 + 2| on 𝐑
 lim 𝑓(𝑥) = −1; lim 𝑓(𝑥) = 2.
𝑥→−∞ 𝑥→∞ Closed and Bounded Intervals
 𝑓 has a vertical asymptote at 𝑥 = 3. If 𝑓(𝑥) is continuous on a closed and bounded interval [𝑎, 𝑏] then the
absolute extrema both exist and they will occur either at a critical number or
 𝑓 is increasing on (−∞, 1); 𝑓 is decreasing on (1, 3) and (3, ∞); at one of the endpoints 𝑎 or 𝑏.
𝑓(1) = 1.
 𝑓 is concave up on (−∞, 0) and (3, ∞); 𝑓 is concave down on (0,3).

3.5 Absolute Extrema

abs min at 𝑥 = 𝑎 abs min at 𝑥 = 𝑐 abs min at 𝑥 = 𝑐2
Absolute Maximum and Minimum abs max at 𝑥 = 𝑐 abs max at 𝑥 = 𝑎 abs max at 𝑥 = 𝑐1
The absolute maximum of a function on an interval is the greatest value
that the function attains on that interval; if the absolute maximum occurs at In practice, to find the absolute extrema of 𝑓(𝑥) on a closed and bounded
𝑥 = 𝑐 then 𝑓(𝑐) ≥ 𝑓(𝑥) for all 𝑥 in the interval. interval we need to:
1. Find the critical numbers
Similarly, the absolute minimum of a function is the least value attained; if 2. Plug all of the critical numbers and the endpoints back into 𝑓(𝑥)
the absolute minimum occurs at 𝑥 = 𝑐 then 𝑓(𝑐) ≤ 𝑓(𝑥) for all 𝑥 in the 3. Choose the least value to be our absolute minimum and the greatest
interval. value to be our absolute maximum.

This works because the extrema must occur within this set of numbers, and
so we only need to check those to find our extrema.

3) Find the absolute extrema for the following functions on the specified
closed and bounded intervals.
abs min at 𝑥 = 𝑎 abs max at 𝑥 = 𝑏 abs max at 𝑥 = 𝑏 a) 𝑓(𝑥) = 2𝑥 3 − 9𝑥 2 − 24𝑥 + 10 on [0, 5]
no abs max no abs min no abs min b) 𝑓(𝑥) = ln(𝑥 2 + 1) on [−1, 1]
2
c) 𝑓(𝑥) = 𝑥 3 on [−2, 1]
1) Find the absolute maximum and absolute minimum given the graphs of
the following functions (if they exist)
Single Critical Point Theorem
If 𝑓(𝑥) is continuous on any interval, and the only critical point occurs at 𝑥 =
𝑐, then 𝑥 = 𝑐 gives an absolute maximum if it is a local maximum and 𝑥 = 𝑐
gives an absolute minimum if it is a local minimum

a) b)
31
 Strategy for Solving Optimization Problems
1. Identify what is being optimized.
The question will discuss many quantities, but for one in particular you
will be asked to find the greatest or least value of it.
This quantity will usually follow the “least/most” words from above.

abs min at 𝑥 = 𝑎 abs max at 𝑥 = 𝑏 abs min at 𝑥 = 1 2. Determine an expression for the quantity being optimized that is relevant
to the information given.
If the only critical point is at 𝑥 = 𝑐 and this is a local minimum, then there For example, if you’re optimizing the area of a circle, 𝐴 = 𝐿𝑊 isn’t a
cannot possible be a lesser value for 𝑓(𝑥): if there were, then there would useful expression for area.
have to be another critical point somewhere in order for the function to move It may be helpful to draw a diagram to see relationships better.
towards this lesser value!
3. If it is not already done, express the quantity to be optimized as a
4) Find the critical point for the following functions. Determine if the critical function of one variable.
point gives the absolute maximum or minimum This may require you to find relationships between the other variables to
a) 𝑓(𝑥) = 𝑥 4 + 𝑥 2 make substitutions.
b) 𝑔(𝑥) = √𝑥 2 + 4
2𝑥−1
c) ℎ(𝑥) = 2 on (0, ∞) 4. Find the critical numbers of the function.
𝑥
Where 𝑓 ′ (𝑥) = 0 or where 𝑓 ′ (𝑥) does not exist.
There will almost always be only one critical number within the domain
3.6 Optimization of the function.
An optimization problem is an application problem where you are asked to 5. Verify that the critical number you found is in fact the max/min as
solve for an absolute extrema. requested.
You have to make sure the critical number is actually the value that is
1) Between 0 and 20 seconds, a particle’s displacement in metres is given being solved for. Typically, you only know that
by 𝑝(𝑡) = −𝑡 3 + 9𝑡 2 + 120𝑡 − 400. 𝑓 ′ (𝑥) = 0, but this does not say it’s a max or min.
Determine when the particle is furthest from position 0. If the domain is closed and bounded, you can use a closed and
bounded interval test
The difficult part of an optimization problem is setting up the function to be Otherwise, if there is one critical number you can use a first or second
optimized. Once this is done, we can use the previous calculus methods to derivative test to determine if it’s a local min/max and therefore absolute
find the maximum or minimum values as requested. min/max

An optimization problem can always be identified by its use of words 6. Answer the question.
suggesting the “most” or “least” of something. These words usually end in “- Be sure that the answer you give is an appropriate response to the
est” such as fastest, largest, shortest, or have most, least, minimum, question. Sometimes you will be asked to give the 𝑥 value that gives a
maximum in the expression. max/min, and sometimes you will be asked for the max/min (so the
value of 𝑓(𝑥)).

2) Two non-negative numbers have a sum of 10. Find the two numbers if
the sum of their squares is maximized.

32
 They refer to infinitesimally small changes in 𝑥 and 𝑦. When 𝛥𝑥 is small, we
3) A 10m fence is to be built between two perpendicular walls to form a tend to use 𝑑𝑥 = 𝛥𝑥. Subsequently, 𝛥𝑦 will be small as well so we can use
triangular pen (the walls will form the two other sides of the right angled 𝑑𝑦 = 𝛥𝑦.
triangle). What angle should the fence make with the wall in order to
maximize the area in the pen? When 𝑦 = 𝑓(𝑥), this gives us the equation for finding the differential 𝑑𝑦:
𝑑𝑦 = 𝑓 ′ (𝑥) 𝑑𝑥
4) An open top box with square base is to be constructed to hold 300 cm 3.
The sides of the box are to be made of a material that costs $0.20/cm 2 2) Find 𝑑𝑦 if 𝑦 = ln(3𝑥 + √𝑥)
and the bottom of the box will cost $0.12/cm 2. What are the dimensions
of the least expensive box that can be made?
5) Find the point on the line 𝑦 = 2𝑥 that is closest to the point (4, 6) 𝑥–1
3) Find 𝑑𝑦 if 𝑦 = if 𝑥 = 5 and 𝛥𝑥 = 0.2
3𝑥 + 1
6) A hydro wire is to be built between a power plant and an island that is
located 6 km down-shore, and 2 km from the coast. We can use differentials to approximate the change that takes place in a
It costs $1000/km to lay the wire on land and $1500/km to lay the wire function of 𝑥 given a small change in 𝑥.
over water.
What is the least expensive way to lay the wire? 4) A 30cm x 30cm photo is placed in a frame with a 2 mm border.
Approximate the area of the border using differentials.
3.7 Linear Approximations
Linear Approximations
If 𝑦 is a function of 𝑥, and the derivative of 𝑦 exists then we have Going back to the equation 𝛥𝑦 ≈ 𝑓 ′ (𝑥)𝛥𝑥 for 𝑦 = 𝑓(𝑥).
𝑑𝑦 Δ𝑦 We have 𝛥𝑥 = 𝑥2 − 𝑥1 and 𝛥𝑦 = 𝑦2 − 𝑦1 = 𝑓(𝑥2 ) − 𝑓(𝑥1 ), so
= lim
𝑑𝑥 Δ𝑥→0 Δ𝑥
𝑓(𝑥2 ) − 𝑓(𝑥1 ) ≈ 𝑓 ′ (𝑥)(𝑥2 − 𝑥1 )
So, for small 𝛥𝑥, we have
𝑑𝑦 Δ𝑦 𝑑𝑦 𝑓(𝑥2 ) ≈ 𝑓(𝑥1 ) + 𝑓 ′ (𝑥)(𝑥2 − 𝑥1 )
≈ or Δ𝑦 ≈ Δ𝑥
𝑑𝑥 Δ𝑥 𝑑𝑥
If we let 𝑥1 = 𝑎 (a fixed, known value) and 𝑥2 = 𝑥 (a variable) then we have
our linearization: 𝑓(𝑥) ≈ 𝑓(𝑎) + 𝑓 ′ (𝑎)(𝑥 − 𝑎)
If 𝑦 = 𝑓(𝑥), this gives us
𝑳(𝒙) = 𝒇(𝒂) + 𝒇′ (𝒂)(𝒙 − 𝒂)
𝚫𝒚 ≈ 𝒇′ (𝒙)𝚫𝒙
This is our linear approximation formula.
Notice: this is just the tangent line for 𝑓(𝑥) at 𝑥 = 𝑎.
1) 𝑦 = 𝑒 𝑥 .
If 𝑥 is increased from 0 to 0.1, approximate the change in 𝑦.

Differentials
We call 𝒅𝒙 and 𝒅𝒚 the differentials of 𝑥 and 𝑦 (respectively).

33
This formula is very convenient if we are trying to approximate the value of
𝑓(𝑥) where 𝑥 is “near” some value 𝑎 that is “easy” to evaluate in the function. Notice that as long as the behaviour of the curve of 𝑦 = 𝑓(𝑥) is not too
complicated, 𝑥2 should be closer to the zero of the function.
5) 𝑓(𝑥) = 𝑥 10 . Linearly approximate the value of 𝑓(1.1) and 𝑓(0.99) using
𝑎 = 1. We now perform the same actions on 𝑥2 .
That is, we find the linearization of 𝑓(𝑥) at 𝑥2
6) Suppose 𝑓(2) = 3 and 𝑓 ′ (2) = −2. and then 𝑥3 will be the zero of that
a) Use a linear approximation to estimate 𝑓(1.9). linearization.
b) If 𝑓 ′′ (2) = 1, will your estimate from a) be above or below the actual
value of 𝑓(1.9)? In most cases, this procedure will give us a
sequence of 𝑥𝑖 ’s that will get closer and
4 closer to the zero of the function.
7) Linearly approximate
2.1

We can use this technique to get a fairly

8) Linearly approximate √62 accurate approximation for the zeros of a function. However, it is not
practically to continually keep recalculating the linearizations, then solving for
their zeroes.
3.8 Newton’s Method
Instead, consider the linearization formula:𝐿(𝑥) = 𝑓(𝑎) + 𝑓 ′ (𝑎)(𝑥 − 𝑎).
Newton’s Method Our first linearization is for 𝑎 = 𝑥1 , so we have:
Not all equations can be solved algebraically. In particular, it is not always 𝐿(𝑥) = 𝑓(𝑥1 ) + 𝑓 ′ (𝑥1 )(𝑥 − 𝑥1 )
possible to find a solution to the equation 𝑓(𝑥) = 0 for some functions. Now, we solve for the zero of 𝐿(𝑥).
0 = 𝑓(𝑥1 ) + 𝑓 ′ (𝑥1 )(𝑥 − 𝑥1 )
Newton’s Method is a technique for approximating the zeroes of a function
by a series of linear approximations. −𝑓′(𝑥1 )(𝑥 − 𝑥1 ) = 𝑓(𝑥1 )
𝑓(𝑥1 )
1) Let 𝑓(𝑥) = 𝑥 3 + 2𝑥 − 1. 𝑥 − 𝑥1 =
−𝑓 ′ (𝑥1 )
a) Find 𝐿1 (𝑥), the linear approximation of 𝑓(𝑥) at 𝑎 = 0.
𝑓(𝑥1 )
b) Solve 𝐿1 (𝑥) = 0. 𝑥 = 𝑥1 −
𝑓 ′(𝑥1 )
c) Evaluate 𝑓(𝑥) at the zero of 𝐿1 (𝑥).
𝑓(𝑥1 )
Newton’s Method is as follows: This zero then becomes our 𝑥2 . So, we have 𝑥2 = 𝑥1 − .
𝑓 ′(𝑥1 )

We start by choosing an initial value,

Since we find 𝑥3 , 𝑥4 , etc. all the same way, we get the iterative formula for
which we can call 𝑥1 .
our sequence:
𝑓(𝑥𝑛 )
Next, we construct the linearization of 𝑥𝑛+1 = 𝑥𝑛 − ′
𝑓(𝑥) at 𝑥1 . This is denoted by 𝐿 in the 𝑓 (𝑥𝑛 )
illustration. This is how we will quickly compute the sequence of terms that will usually
limit towards the zero of our function.
As long as 𝐿 is not horizontal, it will
intersect with the 𝑥-axis, and thus have a zero. This zero will be 𝑥2 .
34
2) Use Newton’s method to approximate a zero of 𝑓(𝑥) = 𝑥 3 + 2𝑥 − 1. 1
1) Verify that 𝑓(𝑥) = √𝑥 − 𝑥 is continuous and differentiable on [0, 9] and
3
Begin with 𝑥1 = 0.
that 𝑓(0) = 𝑓(9). Find 𝑐 in (0,9] such that 𝑓 ′ (𝑐) = 0.
3) Approximate the solution to the equation 𝑒 𝑥 + 𝑥 = 2
2) Let 𝑓(𝑥) = 1 – 𝑥 2/3 . Show that 𝑓(−1) = 𝑓(1) but there is no number 𝑐
4) To four decimal places, find all solutions to the equation 𝑥 = 2 sin 𝑥
in (−1, 1) such that 𝑓 ′ (𝑐) = 0. Why does this not contradict Rolle’s
3 Theorem?
5) Use Newton’s method to approximate √15 to four decimal places.
Pf: Suppose that 𝑓(𝑥) is differentiable on [𝑎, 𝑏] and 𝑓(𝑎) = 𝑓(𝑏). There are
Where Newton’s Method Fails two cases. Either 𝑓(𝑥) = 𝑓(𝑎) (and so 𝑓(𝑥) = 𝑓(𝑏)) for all 𝑥 in (𝑎, 𝑏) OR
It was mentioned that when the linearization is horizontal, then it will have no there is some 𝑥 in (𝑎, 𝑏) such that 𝑓(𝑥) ≠ 𝑓(𝑎).
zero. Notice that in the formula for Newton’s Method, this will give 𝑓 ′ (𝑥𝑛 ) =
0, and so the value of 𝑥𝑛+1 will not exist. If 𝑓(𝑥) = 𝑓(𝑎) for all 𝑥 in (𝑎, 𝑏) then 𝑓(𝑥) is constant on [𝑎, 𝑏] (this is the first
illustration above). If 𝑓(𝑥) is constant, then 𝑓 ′ (𝑥) = 0 for all 𝑥 in (𝑎, 𝑏) and so
In fact, if 𝑓 ′ (𝑥𝑛 ) is small, there will be a large (and often unwanted) jump we have our conclusion.
between 𝑥𝑛 and 𝑥𝑛+1 .
If 𝑓(𝑥) ≠ 𝑓(𝑎) for some 𝑥 in (𝑎, 𝑏), then either 𝑓(𝑥) < 𝑓(𝑎) or 𝑓(𝑥) > 𝑓(𝑎). If
As well, if the curve is complex between 𝑥1 and our zero, it is much less likely 𝑓(𝑥) < 𝑓(𝑎), then 𝑓(𝑥) < 𝑓(𝑏) and so the function “goes down” from 𝑓(𝑎)
that the approximations will approach the zero. and then must “go back up” to 𝑓(𝑏). (this is the case in the second
illustration, notice we get a local minimum). So, 𝑓 ′ changes from positive to
In most cases, if it doesn’t appear as though Newton’s method is working for negative and so must go through zero at least once (by the Intermediate
some 𝑥1 , then we can try another. Value Theorem). We can let 𝑐 be any of these places where 𝑓 ′ = 0.
3
Similiary, if 𝑓(𝑥) > 𝑓(𝑎) we will get a local maximum and therefore some 𝑐
6) Use Newton’s method to solve √𝑥 = 0 using 𝑥1 = 𝑎 for any 𝑎 ≠ 0. where 𝑓′ = 0.

Notice that Rolle’s Theorem only states there is at least one solution
3.9 Mean Value Theorem to 𝑓 ′ (𝑥) = 0. It does not indicate how many solutions there are. (Consider
the third illustration above).
Rolle’s Theorem
If 𝑓(𝑥) is continuous and differentiable on the interval [𝑎, 𝑏] and 𝑓(𝑎) = Upper Bound for the Number of Zeroes
𝑓(𝑏), then there is some c in (𝑎, 𝑏) such that 𝑓 ′ (𝑐) = 0. We used the Intermediate Value Theorem to give us a lower bound on the
number of zeroes. We can use Rolle’s Theorem to give us an upper bound
The illustrations below should explain the theorem. Essentially, if 𝑓 attains on this number.
the same height at two points, then there must be a horizontal tangent If 𝑓(𝑥) is continuous and differentiable and has 𝑛 critical
between them. numbers then 𝑓(𝑥) has at most 𝑛 + 1 zeroes.

35
Notice in this illustration that there are 7 zeroes to the function, and 6
locations with 𝑓 ′ (𝑥) = 0.

Pf: Suppose that 𝑓(𝑥) has 𝑛 critical numbers. We will show that it is
impossible to have at least 𝑛 + 2 zeroes.

Suppose we did have 𝑛 + 2 zeroes. This gives us 𝑎1 , 𝑎2 , …, 𝑎𝑛+2

such that 𝑓(𝑎1 ) = 𝑓(𝑎2 ) = … = 𝑓(𝑎𝑛+2 ) = 0. We can order so that 𝑎1 < 𝑎2 <
… < 𝑎𝑛+2 .

According to Rolle’s Theorem, we should be able find a critical number in 7) Verify that 𝑓(𝑥) = 3𝑥 2 + 2𝑥 + 5 is continuous and differentiable on the
each of the intervals (𝑎1 , 𝑎2 ), (𝑎2 , 𝑎3 ), …, (𝑎𝑛+1 , 𝑎𝑛+2 ). There are 𝑛 + 1 interval [−1, 1] and find c that verifies the conclusion of the Mean Value
distinct intervals and so this will give 𝑛 + 1 distinct critical numbers. But, we Theorem.
had said there were only 𝑛 critical numbers and so this is impossible.
Pf: We prove the Mean Value Theorem by basically “slanting” our function
Therefore there can only be at most 𝑛 + 1 zeroes to the function.
and applying Rolle’s Theorem.
3) Show that 𝑓(𝑥) = 𝑥 3 + 𝑥 2 − 5𝑥 + 1 has at most 3 real zeroes.
Suppose 𝑓(𝑥) is continuous and differentiable on [𝑎, 𝑏].
𝑥 Let 𝑆 be the secant line between 𝑥 = 𝑎 and 𝑥 = 𝑏. So 𝑆 has slope
4) Show that the equation 𝑒 + 𝑥 = 2 has at most one solution. 𝑓(𝑏)−𝑓(𝑎)
and goes through the point (𝑎, 𝑓(𝑎)) and thus 𝑆 has equation
𝑏−𝑎
If there are 𝑛 critical numbers to a function, it does not necessarily mean that (𝑆 − 𝑓(𝑎)) =
𝑓(𝑏)−𝑓(𝑎)
(𝑥 − 𝑎), or:
there are 𝑛 + 1 zeroes. It just means there are at most 𝑛 + 1 zeroes. (ie, 𝑏−𝑎
𝑓(𝑏) − 𝑓(𝑎)
consider 𝑓(𝑥) = 𝑥 2 + 1). 𝑆 = 𝑓(𝑎) + (𝑥 − 𝑎)
𝑏−𝑎
We can use Rolle’s Theorem in combination with the Intermediate Value
We can construct a new function 𝑔(𝑥) by subtracting 𝑆 from 𝑓(𝑥).
theorem to find the exact number of zeroes to a function. 𝑓(𝑏)−𝑓(𝑎)
𝑔(𝑥) = 𝑓(𝑥) − 𝑆 = 𝑓(𝑥) − 𝑓(𝑎) − (𝑥 − 𝑎)
𝑏−𝑎
5) Show that 𝑥 3 + 𝑥 2 − 5𝑥 + 1 = 0 has exactly 3 real roots.
𝑓(𝑏)−𝑓(𝑎)
We will get that 𝑔(𝑎) = 𝑓(𝑎) − 𝑓(𝑎) − (𝑎 − 𝑎) = 0.
6) Show that the equation 𝑒 𝑥 + 𝑥 = 2 has exactly one solution. 𝑏−𝑎
We will also get that
𝑓(𝑏) − 𝑓(𝑎)
Mean Value Theorem 𝑔(𝑏) = 𝑓(𝑏) − 𝑓(𝑎) − (𝑏 − 𝑎)
Rolle’s Theorem is a particular case of the Mean Value Theorem. 𝑏−𝑎
𝑔(𝑏) = 𝑓(𝑏) − 𝑓(𝑎) − (𝑓(𝑏) − 𝑓(𝑎)) = 0
If 𝑓(𝑥) is continuous and differentiable on an interval [𝑎, 𝑏]
then there is some 𝑐 in (𝑎, 𝑏) such that: And so 𝑔(𝑎) = 𝑔(𝑏) = 0. From this, we can apply Rolle’s Theorem to
𝑓(𝑏) − 𝑓(𝑎) show that there is some 𝑐 in (𝑎, 𝑏) with 𝑔′ (𝑐) = 0.
𝑓′(𝑐) =
𝑏−𝑎

That is, there is some point in the interval whose tangent has the same slope
as the secant between 𝑎 and 𝑏.

36
 𝑓(𝑏)−𝑓(𝑎)
Now, 𝑔′ (𝑐) = 𝑓 ′ (𝑐) − 𝑆′(𝑐). Since 𝑆 is linear with slope , we have
𝑏−𝑎
′ (𝑐) 𝑓(𝑏)−𝑓(𝑎)
𝑆 = . So, we have:
𝑏−𝑎
𝑓(𝑏) − 𝑓(𝑎)
0 = 𝑔′ (𝑐) = 𝑓 ′ (𝑐) − .
𝑏−𝑎
𝑓(𝑏)−𝑓(𝑎)
And therefore 𝑓 ′ (𝑐) = . ∎
𝑏−𝑎

8) Does there exist a continuous and differentiable function 𝑓 such that

𝑓(0) = −1, 𝑓(2) = 4, and 𝑓 ′ (𝑥) ≤ 2 for all 𝑥?

9) If 𝑓(𝑥) is a continuous and differentiable function with 𝑓(1) = 10 and

𝑓 ′ (𝑥) ≥ 2 for 1 ≤ 𝑥 ≤ 4, how small can 𝑓(4) possibly be?

10) Use the Mean Value Theorem to prove the inequality

|sin 𝑏 − sin 𝑎| ≤ |𝑏 − 𝑎| for all 𝑎 and 𝑏.

37
 Section 4:

Integration

38
Learning Outcomes

By the end of this section, you will be able to:

 Evaluate an antiderivative for a simple function.
 Evaluate a specific function given its derivative and an initial value of the function.
 Estimate the area underneath a curve using approximating rectangles with right or left endpoints or midpoints.
 Express the area underneath a curve as a limit of Riemann sums.
 Evaluate a definite integral by interpreting it as areas underneath a curve.
 Use the Fundamental Theorem of Calculus (II) and anti-derivatives to evaluate a definite integral.
 Use the Fundamental Theorem of Calculus (I) to differentiate an integral function.
 Use substitution to simplify a definite or indefinite integral.
 Compute the area enclosed by two curves.
 Compute the average value of a function over an interval
 Use and explain the Mean Value Theorem of Integrals

39
4.1 Antiderivatives 3) Evaluate the following antiderivatives:
a) ∫ sec 2 𝑥 𝑑𝑥
b) ∫ 𝑒 𝑥 𝑑𝑥
The Indefinite Integral 𝑑𝑥
Suppose that 𝑓(𝑥) and 𝐹(𝑥) are functions such that 𝐹′(𝑥) = 𝑓(𝑥). c) ∫
𝑥
We say that 𝑓(𝑥) is the derivative of 𝐹(𝑥), and we can also say that 𝐹(𝑥) is d) ∫ √𝑥 𝑑𝑥
an antiderivative of 𝑓(𝑥).
Simple Antiderivatives
1) Let 𝐹(𝑥) = ln | sec 𝑥 + tan 𝑥 | and 𝑓(𝑥) = sec 𝑥. For the following, we have 𝐹′(𝑥) = 𝑓(𝑥)
Show that 𝐹(𝑥) is an antiderivative of 𝑓(𝑥).
𝑓(𝑥) = 𝑎 ⇒ 𝐹(𝑥) = 𝑎𝑥 + 𝐶
We can also write: 𝐹(𝑥) = ∫ 𝑓(𝑥) 𝑑𝑥 𝑥 𝑛+1
𝑓(𝑥) = 𝑥 𝑛 (𝑛 ≠ −1) ⇒ 𝐹(𝑥) = + 𝐶
𝑛+1
In this notation, we call 𝐹(𝑥) the indefinite integral of 𝑓(𝑥). The "𝑑𝑥" is like 1
𝑓(𝑥) = ⇒ 𝐹(𝑥) = ln |𝑥| + 𝐶
a multiplicative piece that we will learn more about later. 𝑥

𝑓(𝑥) = 𝑒 𝑥 ⇒ 𝐹(𝑥) = 𝑒 𝑥 + 𝐶
2 𝑥
2) Find an antiderivative for the function 𝑓(𝑥) = 3𝑥 + 𝑒 .
𝑓(𝑥) = cos 𝑥 ⇒ 𝐹(𝑥) = sin 𝑥 + 𝐶

A function 𝑓(𝑥) has an infinite number of antiderivatives. If 𝐹(𝑥) is an 𝑓(𝑥) = sin 𝑥 ⇒ 𝐹(𝑥) = − cos 𝑥 + 𝐶
antiderivative of f(x), then 𝐹(𝑥) + 𝐶 will also be for any real C, since (𝐹(𝑥) + 𝑓(𝑥) = sec 2 𝑥 ⇒ 𝐹(𝑥) = tan 𝑥 + 𝐶
𝐶)′ = (𝐹(𝑥))′ + (𝐶)′ = 𝐹′(𝑥) + 0 = 𝑓(𝑥). 𝑓(𝑥) = sec 𝑥 tan 𝑥 ⇒ 𝐹(𝑥) = sec 𝑥 + 𝐶
1
𝑓(𝑥) = ⇒ 𝐹(𝑥) = arctan 𝑥 + 𝐶
Graphically, we can think of 𝐹(𝑥) as a function whose "slope function" is 1+𝑥 2
1
𝑓(𝑥). We can see that an upwards or downwards shift of 𝐹(𝑥) will not affect 𝑓(𝑥) = ⇒ 𝐹(𝑥) = arcsin 𝑥 + 𝐶
√1−𝑥 2
the values of the slope, and thus 𝑓(𝑥).
Similar to the derivative, working with addition and subtraction of functions,
and scalar multiples is easy. That is,
∫[𝑎𝑓(𝑥) + 𝑏𝑔(𝑥)] 𝑑𝑥 = 𝑎 ∫ 𝑓(𝑥) 𝑑𝑥 + 𝑏 ∫ 𝑔(𝑥) 𝑑𝑥

4) Evaluate the following:

10 2 4
When evaluating antiderivatives, we will usually find the "simplest" ∫ (3𝑥 5 − 1+𝑥 2
+ 4𝑒 𝑥 + −
𝑥 𝑥3
+ 2 sec 𝑥 tan 𝑥 ) 𝑑𝑥
antiderivative (one without a constant value), then add a "+ 𝐶" to express the
set of all antiderivatives.
Initial Value Problems
In many applications, it will be useful for us to determine which of the
For example, ∫ sin 𝑥 𝑑𝑥 = − cos 𝑥 + 𝐶.
antiderivatives is the correct one. In order to solve these problems, we must

40
be given the value of the function at a certain point. This set of coordinates Another way of looking at this, is that the displacement can be found by
is called an initial value. computing the area formed by the rectangle underneath the curve.

Suppose we are given that 𝑔′ (𝑥) = 6𝑥 + 1 and 𝑔(0) = 2. Next, suppose we are given the
The antiderivative of 𝑔′ (𝑥) = 6𝑥 + 1 is 𝑔(𝑥) = 3𝑥 2 + 𝑥 + 𝐶. graph of a piecewise constant
Since 𝑔(0) = 2, we can plug in 𝑥 = 0 to get velocity function. Since the velocity
𝑔(0) = 3(0)2 + (0) + 𝐶 = 𝐶 and therefore 𝐶 = 2. is constant over intervals, we can
So, 𝑔(𝑥) = 3𝑥 2 + 𝑥 + 2. calculate the displacement over
each interval by 𝑑 = 𝑣𝑡 and add
5) Solve for 𝑔(𝑥) these displacements together.
𝑔′ (𝑥) = sin 𝑥 + 𝑥; 𝑔(0) = 3
Here, we get
6) A car is traveling at 16 m/s. The brakes are applied such that the car 𝑑 = (1)(3) + (2)(2) + (1)(1) + (3)(1) + (2)(1) = 13 m.
decelerates at a constant 4 m/s2. What is the stopping distance of the Again, this can be viewed at as the area underneath the curve of the
car? function.

Finally, suppose we are given the

4.2 The Area underneath a Curve
graph of any velocity function.
There is no simple way to
The Velocity Problem calculate the displacement over
this time interval; however, we can
Suppose we are given the graph of approximate it by finding
a constant velocity function, as to estimating the average velocities
the right. Since the velocity is over intervals.
constant, we can calculate
displacement over the time interval Here, we will break down the 6
by: 𝑑 = 𝑣𝑡. seconds into three 2 second
intervals and approximate the
Here we get 𝑑 = (2)(6) = 12 m. average velocities as about 1.7 m/s,
2.2 m/s, and 2.8 m/s (from left to
right)

Using this approximation, we get

𝑑 ≈ (1.7)(2) + (2.2)(2) + (2.8)(2) = 13.4 m.

41
We can get an even better
approximation by using smaller time
intervals. We can try six intervals of
1 second to give us the
approximation. As we can see,
there is less "room for error" as the
smaller intervals allow us to get a
better idea of what the average
velocity should be.

We have 𝑑 ≈ 1.6 + 1.6 + 2.1 + 2.4 + 2.7 + 2.9 = 13.3 m

When we take smaller and smaller rectangles, the displacement

approximations get closer and closer to the actual value, and the sum of the
area of the rectangles gets closer and closer to the area underneath the
1) Approximate the displacement over the first second for the object whose curve.
velocity graph is given below:
The Definite Integral
As we have seen, there are reasons for us to want to determine the area
underneath a curve (ie, to find the displacement given a velocity curve).
In calculus, we denote the area underneath the curve of 𝑓(𝑥) over the
interval [𝑎, 𝑏] by:
𝒃
∫𝒂 𝒇(𝒙)𝒅𝒙 ,
which we also call the definite integral of 𝒇(𝒙) from 𝒂 to 𝒃.

The function 𝑓(𝑥) is called the integrand and 𝑎 and 𝑏 are the bounds of
integration.
Notice that the only differences in notation from the indefinite integral are the
bounds of integration. We will see that in computation, the definite integral
and the indefinite integral are closely related.
We can also see that our method of approximating using average velocities
is the same as approximating the area underneath the curve using 2) Compute the following integrals by interpreting the curves geometrically
"approximating rectangles" 4 1
a) ∫1 |𝑥 − 2|𝑑𝑥 b) ∫−1 √1 − 𝑥 2 𝑑𝑥

42
 3) Approximate the area underneath the curve 𝑓(𝑥) = 𝑒 𝑥 on [0, 2] using 4
subintervals and any choice of 𝑥𝑖∗ .
Riemann Sums
If we are asked to approximate the area It is often more convenient to choose the xi*'s according to some pattern.
underneath the curve of 𝑓(𝑥) over [𝑎, 𝑏], Some common examples are choosing right endpoints, left endpoints, or
we can do so using a Riemann Sum midpoints.
Approximation. If [𝑎, 𝑏] is the interval, then the n sub-intervals are:
[𝑥0 , 𝑥1 ], [𝑥1 , 𝑥2 ], …, [𝑥𝑖−1 , 𝑥𝑖 ], …, [𝑥𝑛−1 , 𝑥𝑛 ] where
Riemann Sum Approximations are essentially how we approximated the 𝑥0 = 𝑎, 𝑥1 = 𝑎 + 𝛥𝑥, 𝑥2 = 𝑎 + 2𝛥𝑥, …, 𝑥𝑖 = 𝑎 + 𝑖𝛥𝑥,…, 𝑥𝑛 = 𝑎 +
displacement for the velocity curve earlier; we use appropriate approximating 𝑛𝛥𝑥 = 𝑏.
rectangles.
From this, we can see that:
There will be 𝑛 approximating The left endpoints are 𝑥0 , 𝑥1 , …, 𝑥𝑛−1
rectangles with equal width The right endpoints are 𝑥1 , 𝑥2 , …, 𝑥𝑛 and
𝛥𝑥. The midpoints are ½(𝑥0 + 𝑥1 ), ½ (𝑥1 + 𝑥2 ), …, ½ (𝑥𝑛−1 + 𝑥𝑛 )
Since the interval has length
(𝑏 – 𝑎), we divide this into 𝑛
equal subintervals of length
𝑏–𝑎
𝛥𝑥 = .
𝑛
The height of each rectangle left endpoints right endpoints midpoints
will be determined by the
value of 𝑓(𝑥) for some 𝑥 4) Approximate the area underneath the curve 𝑓(𝑥) = 𝑒 𝑥 on [0, 2] using 𝑛 =
inside that interval. If we are discussing the 𝑖 th interval, we usually denote 4 and:
this representative 𝑥 value as 𝑥𝑖∗ a) left endpoints
b) right endpoints
How we choose these 𝑥𝑖∗ 's is not important, as long as they are in the c) midpoints
specified interval.
5) Approximate the area underneath the curve 𝑓(𝑥) = 𝑒 𝑥 on [0, 2] using
The width of each rectangle is 𝛥𝑥 and the height will be 𝑓(𝑥𝑖∗ ). So, we can right endpoints and:
find the area of the rectangle in the 𝑖 th interval by: a) 𝑛 = 8
𝐴rectangle = length × width = 𝑓(𝑥𝐼∗ ) 𝛥𝑥 b) 𝑛 = 16
So, the area for all of [𝑎, 𝑏] can be approximated by: ** note, the actual value for this area is 𝑒 2 − 1 ≈ 6.389
𝑏 𝑛

∫ 𝑓(𝑥)𝑑𝑥 ≈ 𝑓(𝑥1∗ )Δ𝑥 + 𝑓(𝑥2∗ )Δ𝑥 + ⋯+ 𝑓(𝑥𝑛∗ )Δ𝑥 = ∑ 𝑓(𝑥𝑖∗ )Δ𝑥

𝑎 𝑖=1

43
Computing Area pf: The first property is a result of the calculation of 𝛥𝑥. If we are going from
As we have discussed, we can get better and better approximations by using 𝑏 to 𝑎 instead of 𝑎 to 𝑏 this will cause 𝛥𝑥 to be negative.
smaller rectangles. In other words, the approximation will approach the The second property is clear from interpretation of the definite integral as
actual area if we compute the sum of the areas as the width of the rectangles an area underneath a curve (as long as 𝑎 < 𝑏 < 𝑐). In fact, using the first
(𝛥𝑥) goes to zero. This is the limit of Riemann Sums as 𝑛 goes to infinity! property we also have that it works for any arrangement of 𝑎, 𝑏, and 𝑐.

We have: We get the same distributive/additive properties as with the indefinite integral
𝒃 𝒏 by simply applying the properties over the summation. That is,
∫ 𝒇(𝒙)𝒅𝒙 = 𝐥𝐢𝐦 ∑ 𝒇(𝒙∗𝒊 )𝚫𝒙 𝑏 𝑏 𝑏

𝒂 𝒏→∞ ∫ [𝑐 ∙ 𝑓(𝑥) + 𝑑 ∙ 𝑔(𝑥)] 𝑑𝑥 = 𝑐 ∫ 𝑓(𝑥)𝑑𝑥 + 𝑑 ∫ 𝑔(𝑥)𝑑𝑥

𝒊=𝟏 𝑎 𝑎 𝑎

Recall from calculus I, “𝑑𝑥” refers to "infinitesimally small change in 𝑥". Here, Another nice property is that it does not matter which variable we integrate
the 𝑑𝑥 in the integral is a natural replacement for the 𝛥𝑥 in the summation along. That is,
(which goes to zero, as 𝑛 goes to infinity). 𝑏 𝑏
∫ 𝑓(𝑥)𝑑𝑥 = ∫ 𝑓(𝑡)𝑑𝑡
𝑎 𝑎
6) Express the following integrals as a limit of Riemann Sums using right
endpoints 4.3 The Fundamental Theorem of Calculus
𝜋 ln 4
a) ∫0 sin2 𝑥 𝑑𝑥 b) ∫ln 2 𝑥𝑒 𝑥 𝑑𝑥
Fundamental Theorem of Calculus Part I
1 4
7) Express the following limit as a definite integral lim ∑𝑛𝑖=1 4𝑖 ∙
𝑛→∞ 1+
𝑛
𝑛 If 𝑓 is continuous on [𝑎, 𝑏], then the function
𝑥
𝑔(𝑥) = ∫ 𝑓(𝑡)𝑑𝑡 𝑎≤𝑥≤𝑏
8) Evaluate the area underneath the curve 𝑓(𝑥) = 1 + 2𝑥 over [1, 4] using 𝑎
right-endpoints by setting up the area as a limit as a Riemann sum and is continuous on [𝑎, 𝑏] and differentiable on (𝑎, 𝑏) and 𝑔′(𝑥) = 𝑓(𝑥).
evaluating the limit.
1) Differentiate the following functions:
This also gives us a few convenient properties of integrals. 𝑥 5
𝑏 𝑎 a) 𝑓(𝑥) = ∫2 sin(𝑡 2 )𝑑𝑡 b) 𝑓(𝑥) = ∫𝑥 ln(3𝑡 + 1) 𝑑𝑡
∫ 𝑓(𝑥)𝑑𝑥 = − ∫ 𝑓(𝑥)𝑑𝑥 𝑥2 2𝑥
𝑎 𝑏 c) 𝑦 = ∫2 arcsin 𝑡 𝑑𝑡 d) 𝑔(𝑥) = ∫𝑥 𝑒 4𝑡 𝑑𝑡
𝑏 𝑐 𝑐 𝑥+ℎ 𝑥
𝑔(𝑥+ℎ)− 𝑔(𝑥) ∫𝑎 𝑓(𝑡)𝑑𝑡 − ∫𝑎 𝑓(𝑡)𝑑𝑡
∫ 𝑓(𝑥)𝑑𝑥 + ∫ 𝑓(𝑥)𝑑𝑥 = ∫ 𝑓(𝑥)𝑑𝑥 pf: 𝑔′ (𝑥) = lim = lim
ℎ→0 ℎ ℎ→0 ℎ
𝑎 𝑏 𝑎
by properties of definite integrals, we have
𝑥+ℎ
5 5 1 ∫𝑥 𝑓(𝑡) 𝑑𝑡
9) If ∫1 𝑓(𝑥)𝑑𝑥 = 3 and ∫3 𝑓(𝑥)𝑑𝑥 = 1 then evaluate ∫3 𝑓(𝑥)𝑑𝑥 . = lim
ℎ→0 ℎ

44
 𝑥
This is the area underneath the curve pf: Let 𝑔(𝑥) = ∫𝑎 𝑓(𝑡)𝑑𝑡 .
between 𝑥 + ℎ and 𝑥. As ℎ goes So, we have 𝑔′(𝑥) = 𝑓(𝑥) from FTC I, and so 𝑔(𝑥) is an antiderivative of
zero, this is "approximately a 𝑓(𝑥). Thus, 𝐹(𝑥) = 𝑔(𝑥) + 𝐶 for some 𝐶.
rectangle" with width ℎ and height
somewhere between 𝑓(𝑥 + ℎ) and We have
𝑓(𝑥). Since 𝑥 + ℎ is close to 𝑥, we 𝑏 𝑏
can use the value 𝑥 to obtain our 𝑔(𝑏) = ∫ 𝑓(𝑡)𝑑𝑡 = ∫ 𝑓(𝑥)𝑑𝑥
𝑎 𝑎
height estimate of 𝑓(𝑥) to get: 𝑏
area under 𝑓 between 𝑥 and 𝑥 + ℎ So, 𝐹(𝑏) = ∫𝑎 𝑓(𝑥)𝑑𝑥 + 𝐶
= lim
ℎ→0 h
𝑎
𝑓(𝑥) ⋅ ℎ Also, we have 𝑔(𝑎) = ∫𝑎 𝑓(𝑡)𝑑𝑡 = 0
= lim = 𝑓(𝑥)
ℎ→0 ℎ (since there would be "0" area between 𝑎 and itself)

Since this derivative exists, we have that 𝑔 is continuous and So 𝐹(𝑎) = 0 + 𝐶 = 𝐶

differentiable. ■ Thus,

The second part of the Fundamental Theorem of Calculus finally ties the 𝑏
Thus, 𝐹(𝑏) − 𝐹(𝑎) = (∫𝑎 𝑓(𝑥)𝑑𝑥 + 𝐶) − (𝐶)
antiderivative and the definite integral together.
𝑏
= ∫𝑎 𝑓(𝑥)𝑑𝑥 ■
Fundamental Theorem of Calculus Part II
If 𝑓 is continuous on [𝑎, 𝑏] and 𝐹 is any antiderivative of 𝑓, then
𝒃 The Net Change Theorem
∫ 𝒇(𝒙)𝒅𝒙 = 𝑭(𝒃) − 𝑭(𝒂)
𝒂 Notice that all examples for definite integrals have used positive valued
functions only.
To simplify computations, we will typically write 𝐹(𝑥)|𝑏𝑎 instead of The Net Change Theorem states that the fundamental theorem of calculus
𝐹(𝑏) − 𝐹(𝑎). still holds true even if the function attains negative values on the interval.

2) Evaluate the following: When interpreting a definite integral as "area under a curve", we are actually
2 4 calculating the area between the curve and the 𝑥-axis, measured from the 𝑥-
a) ∫0 𝑒 𝑥 𝑑𝑥 b) ∫1 (1 + 2𝑥)𝑑𝑥
axis upwards. When we have negatively valued functions, a definite integral
𝜋/2 1/√2 1
c) ∫𝜋/3 cos 𝑥 𝑑𝑥 d) ∫1/2 √ 2 𝑑𝑣 will consider the "area between the curve and the 𝑥-axis" to be negative,
1−𝑣
since it is measured in the downwards direction.
3) Water flows into a tank at a rate of (1 + 𝑡 2 )−1 Litres per minute (𝑡 is time
in minutes) . Assuming the tank has 5 L to start, how much water is in
the tank after the first half hour?

45
To compute the definite integral of a
function that is both positively and
negatively valued, we subtract the Over the first three seconds, velocity is positive and so the object travels in
"negative area" from the "positive the positive direction (by a total of 3 metres). Over the final three seconds,
area" to get the net area. velocity is negative, and so the object travels in the negative (or reverse)
direction by 4 metres.

4) Interpret the integral as areas to evaluate

5
∫ (|3 − 𝑥| − 1)𝑑𝑥
0
The result is a net displacement of 3 – 4 = −1 metres.
This is consistent with the Riemann Sum definition,
𝑛 If we were to measure distance, we treat all numbers as "positive" (in other
𝑏
∫ 𝑓(𝑥)𝑑𝑥 = lim ∑ 𝑓(𝑥𝑖∗ )Δ𝑥 words, all areas are positively valued).
𝑛→∞
𝑎 𝑖=1 We would have a distance of 3 metres in the first three seconds and 4
metres in the final three seconds and thus a total distance of 7 metres.
where negatively valued regions of the function will cause the integral to
decrease. 6) A particle has velocity defined by the function
𝑣(𝑡) = 𝑡 2 – 4𝑡 + 3 measured in m/s and 𝑡 is time in seconds.
5) Evaluate the following integrals: a) Calculate the displacement between 0 and 5 seconds.
3 b) Calculate the distance travelled between 0 and 5 seconds.
a) ∫0 (1 − 2𝑥)𝑑𝑥
4
b) ∫2 (𝑥 3 − 𝑒 𝑥 )𝑑𝑥
4.4 Substitution
To highlight why this definition of "negative areas" makes sense, it is easiest
to look at a velocity vs time graph and use it to measure displacement and The Indefinite Integral
distance. Recall the Chain Rule from differential calculus:
A reminder that displacement does not consider the path taken, only the (𝐹(𝑔(𝑥))) ′ = 𝐹 ′ (𝑔(𝑥))𝑔′ (𝑥)
resulting position whereas distance measures every step along the way.
So, this gives us: ∫ 𝐹 ′ (𝑔(𝑥))𝑔′ (𝑥) 𝑑𝑥 = 𝐹(𝑔(𝑥)) + 𝐶.
If we let 𝑔(𝑥) = 𝑢, we can write this as
∫ 𝐹 ′ (𝑔(𝑥))𝑔′ (𝑥) 𝑑𝑥 = 𝐹(𝑢) + 𝐶
But, 𝐹(𝑢) = ∫ 𝐹 ′ (𝑢)𝑑𝑢, So:
∫ 𝐹 ′ (𝑔(𝑥))𝑔′ (𝑥)𝑑𝑥 = ∫ 𝐹 ′ (𝑢)𝑑𝑢

46
If we let 𝐹 be an antiderivative of 𝑓, we have: and the right side is:
𝑔(𝑏)
∫ 𝒇(𝒈(𝒙))𝒈′ (𝒙) 𝒅𝒙 = ∫ 𝒇(𝒖) 𝒅𝒖 ∫
𝑔(𝑏)
𝑓(𝑢)𝑑𝑢 = 𝐹(𝑢)|𝑔(𝑎) = 𝐹(𝑔(𝑏)) − 𝐹(𝑔(𝑎))
𝑔(𝑎)

In practice, we will often want to use this technique when integrating and so both sides are equal. ■
composite functions. We let 𝑢 = 𝑔(𝑥) ("the inside function")
And, using differentials we have 𝑑𝑢 = 𝑔′ (𝑥) 𝑑𝑥. In practice, we need to change the bounds of integration in order to use
substitution on a definite integral.
Consider ∫ 𝑥 2 sin(𝑥 3 + 1) 𝑑𝑥.
sin(𝑥 3 + 1) is a composite function with (𝑥 3 + 1) on "the inside", so we let 4
For example, Suppose we have ∫0 √2𝑥 + 1 𝑑𝑥 .
𝑢 = 𝑥 3 + 1.
We let 𝑢 = 2𝑥 + 1, so 𝑑𝑢 = 2 𝑑𝑥.
Thus, 𝑑𝑢 = 3𝑥 2 𝑑𝑥, or 𝑥 2 𝑑𝑥 = 𝑑𝑢/3.
We also need to write the first bound as 𝑢(0) = 2(0) + 1 = 1 and the
We can first rearrange the integral to get:
second bound as 𝑢(4) = 2(4) + 1 = 9. This gives:
= ∫ sin(𝑥 3 + 1) (𝑥 2 𝑑𝑥). 91
Now we can substitute 𝑥 3 + 1 = 𝑢 and 𝑥 2 𝑑𝑥 = 𝑑𝑢/3 to get: = ∫1 2 √𝑢 𝑑𝑢 which we can now integrate to get:
1
= ∫ sin(𝑢) 𝑑𝑢/3 =
3
∫ sin 𝑢 𝑑𝑢. 1 9 1 2 3 9 1 26
= ∫ 𝑢1/2 𝑑𝑢 = [ 𝑢2 ] = (27 − 1) =
This is easy to integrate! 2 1 2 3 1 3 3
1
= − cos 𝑢 + 𝐶
3
We now replace u with the original substitution, 𝑢 = 𝑥 3 + 1 2) Evaluate the following:
1 2 2 𝑒 1/𝑥 2
= − cos(𝑥 3 + 1) + 𝐶
3
a) ∫0 (𝑥 − 1)25 𝑑𝑥 b) ∫1 2 𝑑𝑥 c) ∫1 𝑥 √𝑥 − 1 𝑑𝑥
𝑥
1) Integrate the following:
a) ∫ 𝑒 3𝑥 𝑑𝑥 b) ∫ √2𝑥 + 1 𝑑𝑥 4.5 Areas Between Curves
2+4𝑥
c) ∫ sin2 𝜃 cos 𝜃 𝑑𝜃 d) ∫
𝑥+𝑥 2
𝑑𝑥
Earlier, we discussed the definite integral as being the area (with +/- value)
The Definite Integral underneath a curve.
If 𝑔′ is continuous on [𝑎, 𝑏] and 𝑓 is continuous on the range of 𝑔(𝑥), and We can compute the area between two curves using a similar idea.
𝑢 = 𝑔(𝑥), then:
𝒃 𝒈(𝒃) If we wish to compute the area between 𝑓(𝑥) and 𝑔(𝑥) over the interval [𝑎, 𝑏]
∫ 𝒇(𝒈(𝒙))𝒈′ (𝒙)𝒅𝒙 = ∫ 𝒇(𝒖)𝒅𝒖 where 𝑓(𝑥) > 𝑔(𝑥) then we can simply subtract the area of 𝑔(𝑥) over [𝑎, 𝑏]
𝒂 𝒈(𝒂)
from the area of 𝑓(𝑥), as below:

pf: the left side is:

𝑏
𝑏
∫ 𝑓(𝑔(𝑥))𝑔′ (𝑥)𝑑𝑥 = 𝐹(𝑔(𝑥))|𝑎 = 𝐹(𝑔(𝑏)) − 𝐹(𝑔(𝑎))
𝑎

47
So, if 𝑓(𝑥) > 𝑔(𝑥) on [𝑎, 𝑏] then the area between 𝑓(𝑥) and 𝑔(𝑥) can be In these cases, we find the points a and b where the curves intersect and
found by: then compute the area between the curves over the interval [𝑎, 𝑏].

𝑏 𝑏 𝑏
3) Compute the area enclosed by the curves 𝑦 = √𝑥 and 𝑦 = 𝑥 2
∫ 𝑓(𝑥)𝑑𝑥 − ∫ 𝑔(𝑥)𝑑𝑥 = ∫ [𝑓(𝑥) − 𝑔(𝑥)] 𝑑𝑥
𝑎 𝑎 𝑎
If the curves intersect in more than one place, then we must split the integral
When asked to integrate the difference between 𝑓 and 𝑔, it will be up to us to up into intervals such that 𝑓(𝑥) > 𝑔(𝑥) over entire intervals or 𝑔(𝑥) > 𝑓(𝑥)
determine which of 𝑔 or 𝑓 is greater than the other on the interval. We can over entire intervals.
usually determine this by a quick sketch of the curves.
If we choose the wrong function to be "on top" then we will get a negative 4) Compute the area enclosed by the curves 𝑦 = 𝑥 3 – 𝑥 and
answer. In this case, we can just switch our choice of 𝑓 and 𝑔 and we will 𝑦 = 3𝑥 – 3𝑥 3 .
get the correct answer!
In some cases, it will be easier to compute the area by integrating along 𝑦
instead of 𝑥. If the curves are easier to express as 𝑥(𝑦) then this may be an
1) Compute the area between the curves 𝑦 = √3𝑥 + 1 and 𝑦 = 𝑒 −𝑥 on the
option.
interval [1, 5]

5) Find the area enclosed by the line 𝑦 = 𝑥 – 1 and the parabola

When one of the functions is negatively valued, the formula still works.
𝑦 2 = 2𝑥 + 6
Consider the illustration below:

4.6 Average Value of a Function

Average Value Formula

If we have 𝑛 distinct 𝑦-values, then we would find the average of these 𝑦-
If we are asked for the area between curves, we do not consider any of the values by the formula:
space to be negatively valued, so the formula above works. 𝑦1 + 𝑦2 + … + 𝑦𝑛
𝑦𝑎𝑣𝑒 =
𝑛
2) Compute the area between the curves 𝑦 = 𝑥 2 − 6𝑥 + 5 and
𝑦 = (𝑥 + 3)−1 over the interval [2, 4] To compute the average value of a continuous measure (defined by a
In many cases, we will be asked to find continuous function 𝑓(𝑥)) over an interval [𝑎, 𝑏], we can approximate the
the area enclosed by two curves. In value using a similar formula.
this case, the curves will meet in at
least two places in such a way that the We will divide the interval into n sub-intervals of equal length
curves create a border for some 𝛥𝑥 = (𝑏– 𝑎)/𝑛. We can approximate the contribution over each interval by
enclosed area. choosing a test point, 𝑥𝑖∗ in each interval.

48
 𝑓(𝑥1∗ )+ 𝑓(𝑥2∗ )+ …+𝑓(𝑥𝑛
∗)
So, we get : 𝑓𝑎𝑣𝑒 ≈
𝑛
pf: Suppose that 𝑓_ave = 𝐴. There are two cases: either 𝑓(𝑥) is a constant
.
function or 𝑓(𝑥) is not.
To get better and better approximations, we would take the limit as 𝑛 goes to
infinity. This gives:
If 𝑓(𝑥) is constant, then we must
𝑓(𝑥1∗ ) + 𝑓(𝑥2∗ ) + ⋯ + 𝑓(𝑥𝑛∗ )
𝑓𝑎𝑣𝑒 = lim have 𝑓(𝑥) ≡ 𝐷 for some constant
𝑛→∞ 𝑛
𝐷. This will give us that the area
𝑏−𝑎 1 Δ𝑥 underneath 𝑓 from 𝑎 to 𝑏 is a
Now, recall, 𝛥𝑥 = and so = . This gives:
𝑛 𝑛 𝑏−𝑎 rectangle with area (𝑏 – 𝑎) ⋅ 𝐷
𝑓(𝑥1∗ ) + 𝑓(𝑥2∗ ) +
⋯ + 𝑓(𝑥𝑛∗ ) 1 1
𝑓𝑎𝑣𝑒 = lim ⋅ Δ𝑥 (see the illustration) and so 𝑓𝑎𝑣𝑒 = AREA = [(𝑏 − 𝑎) ⋅ 𝐷] = 𝐷.
𝑛→∞ (𝑏 − 𝑎) 𝑏−𝑎 𝑏−𝑎

1 Since we have that 𝑓ave = 𝐴, this means that 𝐷 = 𝐴, and that in fact all 𝑥’s
= lim [𝑓(𝑥1∗ ) + 𝑓(𝑥2∗ ) + ⋯ + 𝑓(𝑥𝑛∗ )] ⋅ Δ𝑥 in [𝑎, 𝑏] are equal to 𝐴, and so any of them could be our choice of 𝑐.
𝑏 − 𝑎 𝑛→∞

The limit is the definition of the definite integral! So we have: If 𝑓(𝑥) is not constant, then there
1 𝑏 are values less than (or greater
𝑓𝑎𝑣𝑒 = ∫ 𝑓(𝑥)𝑑𝑥 than) 𝐴. If all of the values are
𝑏−𝑎 𝑎
less than 𝐴, then 𝑓(𝑥) < 𝐴 and so
the area underneath 𝑓 from 𝑎 to 𝑏
1) Compute the average of the following functions over the specified
is less than (𝑏 – 𝑎) ⋅ 𝐴 (see the illustration), and so the average is less
intervals:
ln 𝑥 2
than 𝐴, a contradiction. Thus, if there are values less than 𝐴, there must
a) 𝑓(𝑥) = [1, 𝑒] b) 𝑓(𝑡) = 𝑡𝑒 −𝑡 [0, 5] also be values above 𝐴. So, by the intermediate Value Theorem, there
𝑥
must also be a value at 𝐴 and so this will be our 𝑐.
2) Show that the average rate of change of 𝑓(𝑥) can be found using Similarly, if there are values greater than 𝐴, there must also be values
𝑓(𝑏)−𝑓(𝑎)
. less than 𝐴.
𝑏−𝑎
■
3
Mean Value Theorem for Integrals 4) If 𝑓 is continuous and ∫1 𝑓(𝑥)𝑑𝑥 = 8, show that 𝑓(𝑥) = 4 at least once on
If 𝑓 is continuous on [𝑎, 𝑏], then there exists a number 𝑐 in the interval [1,3].
[𝑎, 𝑏] such that 𝑓(𝑐) = 𝑓ave

3) 𝑓(𝑥) = 𝑥 2 + 3𝑥 + 1. Find the average over the interval [0, 4] and find a
value 𝑐 in [0, 4] such that 𝑓(𝑐) = 𝑓𝑎𝑣𝑒 .

49