Math 101 Course Pack

Calculus II

Notes By William Thompson

Contents

1 Integration Techniques and Applications 6

1.1 Brief Review of Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.2 (Section 8.1) Some Integration Techniques and Tricks . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.2.1 Completing the Square . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.2.2 Using Trigonometric Identities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.2.3 Long Division . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.2.4 Rewriting (a matter of perception) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.2.5 Dumb Luck . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.2.6 Mixing Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.2.7 Organizing Definite Integral Substitution . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.3 (Section 7.1) Logarithms and Exponentials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
1.3.1 Integrals Involving the Natural Logarithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
1.3.2 Integrals Involving the Exponential Function . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.3.3 Other Logarithmic and Exponential Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . 26
1.4 (Section 7.2) Separable Differential Equations and Modeling . . . . . . . . . . . . . . . . . . . . . . 28
1.4.1 Separable Differential Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
1.4.2 Unlimited Population Growth Model (Malthus Model) . . . . . . . . . . . . . . . . . . . . . 29
1.4.3 Radioactive Decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
1.4.4 Heat Transfer: Newton’s Law of Cooling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
1.5 (Section 8.2) Integration By Parts (IBP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
1.5.1 Establishing Integration by Parts and LIPET . . . . . . . . . . . . . . . . . . . . . . . . . . 33
1.5.2 Dominant Logarithms and Inverse Functions in LIPET . . . . . . . . . . . . . . . . . . . . 35
1.5.3 Dominant Polynomials in LIPET . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
1.5.4 Dominant Exponential or Trigonometric Functions in LIPET . . . . . . . . . . . . . . . . . 37
1.5.5 Substitutions Leading to IBP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
1.5.6 Integrals of Lonely Logarithms and Inverse Functions (Ninja’s) . . . . . . . . . . . . . . . . 39
1.5.7 The Tabular Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
1.6 (Section 8.3) Trigonometric Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
1.6.1 Products of Sines and Cosines of Different Composition . . . . . . . . . . . . . . . . . . . . 42
1.6.2 Products of Sines and Cosines, Same Input, Raised to a Power . . . . . . . . . . . . . . . . 44
1.6.3 Eliminating Roots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
1.6.4 Lonely Powers of Tangent or Secant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
1.6.5 Powers of Tangent and Secant with Same Input . . . . . . . . . . . . . . . . . . . . . . . . . 52
1.7 (Section 8.4) Trigonometric Substitution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
1.7.1 Forms of Trigonometric Subsitution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
1.7.2 Sine Substitutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
1.7.3 Tangent Substitutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

1
 1.7.4 Secant Substitutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
1.7.5 More Obscure u Terms in Trigonometric Substitution . . . . . . . . . . . . . . . . . . . . . 59
1.8 (Section 8.5) Partial Fractions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
1.8.1 Linear Terms Present . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
1.8.2 Irreducible Quadratic Terms Present . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
1.8.3 Special Case Where All Terms Are Linear . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
1.8.4 Example of Computation an Integral by Partial Fractions . . . . . . . . . . . . . . . . . . . 65
1.9 (Section 8.8) Improper Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
1.9.1 Defining Improper Integrals and Convergence . . . . . . . . . . . . . . . . . . . . . . . . . . 67
1.9.2 p-Integrals and the Integral Comparison Tests . . . . . . . . . . . . . . . . . . . . . . . . . . 69
1.9.3 Vertical Asymptotes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
1.9.4 Intuition is Lost in the Land of Infinity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
1.10 (Section 6.1) Volumes Using Cross Sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
1.10.1 Defining Volumes by Cross Sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
1.10.2 Solids of Revolution and their Volume by Washers . . . . . . . . . . . . . . . . . . . . . . . 78
1.11 (Section 6.2) The Shell Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
1.12 (Section 6.3) Arc-Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
1.12.1 Formulating the Arc-Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
1.12.2 Computation of Arc-Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
1.12.3 The Surface Area of a Solid of Revolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

2 Series 91
2.1 (Section 10.1) Introduction to Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
2.1.1 Ways to Describe a Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
2.1.2 Convergence and Divergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
2.1.3 Useful Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
2.1.4 Monotonic Sequences and the Monotonic Convergence Theorem . . . . . . . . . . . . . . . 99
2.2 (Section 10.2) Infinite Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
2.2.1 Defining an Infinite Sum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
2.2.2 Geometric Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
2.2.3 Telescoping Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
2.2.4 The n-th Term Divergence Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
2.2.5 Properties of Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
2.3 (Section 10.3) Integral Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
2.3.1 Constructing the Integral Test and Examples . . . . . . . . . . . . . . . . . . . . . . . . . . 110
2.3.2 Error Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
2.4 (Section 10.4) Comparison Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
2.4.1 The Direct Comparison Test for Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
2.4.2 The Limit Comparison Test for Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
2.5 (Section 10.5) Absolute Convergence, Ratio and Root Tests . . . . . . . . . . . . . . . . . . . . . . 119
2.5.1 The Absolute Convergence Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
2.5.2 The Ratio Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
2.5.3 The Root Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
2.6 (Section 10.6) Alternating Series Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
2.6.1 Conditional Convergence and the Alternating Series Test . . . . . . . . . . . . . . . . . . . 126
2.6.2 Error Estimation of Series that Converge by the AST . . . . . . . . . . . . . . . . . . . . . 129
2.7 (Section 10.7) Power Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
2.7.1 Power Series Functions and Their Domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

2
 2.7.2 Operations of Power Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
2.8 (Section 10.8) Taylor and MacLaurin Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
2.8.1 The Unreasonable Assumption of Being a Taylor Series . . . . . . . . . . . . . . . . . . . . 138
2.8.2 Examples of Constructing a Taylor Series Directly . . . . . . . . . . . . . . . . . . . . . . . 139
2.8.3 Taylor Polynomials of Order N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
2.9 (Section 10.9) Convergence of Taylor Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
2.9.1 Taylor’s Theorem and Approximating Functions . . . . . . . . . . . . . . . . . . . . . . . . 142
2.9.2 Taylor Series with Bounded Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
2.9.3 Important Analytic Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
2.9.4 Approximations Using Taylor Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
2.10 (Section 10.10) Binomial Series and Applications of Taylor Series . . . . . . . . . . . . . . . . . . . 148
2.10.1 Binomial Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
2.10.2 Applications of Taylor Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150

3 Coordinate Systems 153

3.1 (Appendix A7) Complex Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
3.1.1 Introducing Complex Numbers and Algebraic Operations . . . . . . . . . . . . . . . . . . . 154
3.1.2 Argand Diagrams and Polar Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
3.1.3 Euler’s Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
3.1.4 Roots of Complex Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
3.2 (Section 11.1) Parametric Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
3.2.1 Defining Parametric Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
3.2.2 Graphing a Parametric Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
3.2.3 Standard Representation of a Circle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
3.2.4 Transformations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
3.3 (Section 11.2) Calculus of a Parameterized Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
3.3.1 The Slope of a Parametric Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
3.3.2 Higher Order Derivatives and Vertical/Horizontal Tangency . . . . . . . . . . . . . . . . . . 166
3.3.3 Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
3.3.4 Arc-Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
3.4 (Section 11.3) Polar Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
3.4.1 Revisiting Polar Coordinates and Non-Uniqueness . . . . . . . . . . . . . . . . . . . . . . . 172
3.4.2 Polar to Cartesian (Points), Cartesian to Polar (Equations) . . . . . . . . . . . . . . . . . . 173
3.4.3 Cartesian to Polar (Points), Polar to Cartesian (Equations) . . . . . . . . . . . . . . . . . . 175
3.4.4 Graphing Sinusoidal Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
3.5 (Section 11.4) Graphing Polar Curves in the Cartesian Plane . . . . . . . . . . . . . . . . . . . . . 178
3.5.1 Plotting Curves in the xy-Plane Based on Their rθ-Graph . . . . . . . . . . . . . . . . . . . 178
3.5.2 Graphing Polar Curves Using GeoGebra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
3.5.3 Calculus in Polar Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
3.6 (Section 11.5) Area Trapped by Polar Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
3.6.1 Formulating the Wedged Area Trapped by Polar Curves . . . . . . . . . . . . . . . . . . . . 186
3.6.2 Examples of Computing Area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
3.6.3 Splits in Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189

3
Table of Trigonometric Identities

Basic Identities

• cos2 (θ) + sin2 (θ) = 1

• tan2 (θ) + 1 = sec2 (θ)

• sin(2θ) = 2 sin(θ) cos(θ)

• cos(2θ) = cos2 (θ) − sin2 (θ)

Half Angle Identities

1 − cos(2θ)
• sin2 (θ) =
2
1 + cos(2θ)
• cos2 (θ) =
2
Ptolemy’s Identities

• sin(A + B) = sin(A) cos(B) + cos(A) sin(B)

• cos(A + B) = cos(A) cos(B) − sin(A) sin(B)

tan(A) + tan(B)
• tan(A + B) =
1 − tan(A) tan(B)
Product to Sum
1 1
• sin(A) sin(B) = cos(A − B) − cos(A + B)
2 2
1 1
• cos(A) cos(B) = cos(A − B) + cos(A + B)
2 2
1 1
• sin(A) cos(B) = sin(A − B) + sin(A + B)
2 2
Even and Odd Properties

• sin(−θ) = − sin(θ)

• cos(−θ) = cos(θ)

• tan(−θ) = − tan(θ)

Reflections
π 
• sin − θ = cos(θ) and sin (π − θ) = sin(θ)
2
π 
• cos − θ = sin(θ) and cos (π − θ) = − cos(θ)
2
π 
• tan − θ = cot(θ) and tan (π − θ) = − tan(θ)
2

4
Table of Antiderivatives
Z
f (x) f (x)dx

1
xn xn+1 + C if n 6= 1
n+1

sin(x) − cos(x) + C

cos(x) sin(x) + C

tan(x) ln | sec(x)| + C

sec(x) ln | sec(x) + tan(x)| + C

cot(x) ln | sin(x)| + C

csc(x) − ln | csc(x) + cot(x)| + C

sec2 (x) tan(x) + C

csc2 (x) − cot(x) + C

sec(x) tan(x) sec(x) + C

csc(x) cot(x) − csc(x) + C

ex ex + C

1
ln |x| + C
x
1
√ arcsin(x) + C
1 − x2
1
arctan(x) + C
1 + x2
1
√ arcsec|x| + C
x x2 − 1
ax
ax +C
ln(a)

5
Chapter 1

Integration Techniques and Applications

6
1.1 Brief Review of Integration
The first month focuses heavily on picking up where Calculus I left off...integration.

There are two types of integration in Calculus I,

b n  
b−a b−a
Z X
• Definite Integration: f (x)dx = lim f a+k
a n→∞ n n
k=1
Z
• Indefinite Integration: f (x)dx = F (x) + C where F 0 (x) = f (x).

Indefinite Integration

This is when we try to find the antiderivative of a function. Usually we do this to apply it to definite integra-
tion. That is,
Z
If F 0 (x) = f (x) then f (x)dx = F (x) + C

We call F (x) an antiderivative of f (x) and F (x) + C the most generalized antiderivative of f (x).

Note: Do not forget the constant of integration!

Definite Integration

This is defined by the use of Riemann sums,

Z b n
X
f (x)dx = lim f (a + k∆x)∆x
a n→∞
k=1

b−a
where ∆x = . This geometrically represents the area between the graphs of y = f (x) and y = 0 over the
n
region a ≤ x ≤ b.

Another, and in my personal opinion, a better way to think of this integral is that it adds up the area of a
b−a
bunch of rectangles with area f (x)dx which we represent as f (a + k∆x)∆x for ∆x = of infinitesimal size.
n
The reason why to view it as a summation of terms f (x)dx rather than area is because in the future you might
replace these terms with other quantities that shouldn’t be interpreted geometrically.

7
 Example: Consider a metal rod where we measure along only it’s length from the bottom x = 0 to x = L if
it is L units in length.

Let ρ(x0 ) represent the density of the bar at the point x = x0 . As dx is thought of as an infinitesimal of ∆x, a
volume portion of the bar, then ρ(x)dx is the product of density and volume in a small region of the bar yielding
the mass in that region. To get the total mass of the bar we sum up everything to obtain
Z L
Mass = ρ(x)dx
0
Note: This example illustrates that like the derivative, there are two interpretations of integration. One physical
and one geometric. Do not pick favourites on how you like to interpret integration, you need the flexibility of both
to solve problems.

By the second part of the Fundamental Theorem of Calculus there is a relationship between antiderivatives
and definite integration. This is given by
Z b
f (x)dx = F (b) − F (a)
a
where F (x) is an antiderivative of f (x) on [a, b].

Example: Find the area between y = x and y = x2 − 2x.

8
 Z π/6
Example: Compute (2 sec(θ) + sec(θ) tan(θ))dθ
0

There are basic rules to integration:

Z Z Z
• (f (x) + g(x))dx = f (x)dx + g(x)dx
Z Z
• kf (x)dx = k f (x)dx

Integration (indefinite) undoes differentiation. Undoing something is hard! Unlike differentiation, we are very
limited in ways to find an antiderivative. I like to compare differentiation and antidifferentiation to a plate.
Differentiation would be smashing the plate and antidifferentiation would be gluing it back together.

9
 Substitution (undoing the chain rule)

This dealt with integrals of the form

Z b
f (g(x))g 0 (x)dx
a

By letting u = g(x) we have du = g 0 (x)dx. Moreover, when x = a then u = g(a) and when x = b then u = g(b).
So this becomes
Z b Z g(b)
0
f (g(x))g (x)dx = f (u)du
a g(a)

This type of substitution letting u = g(x) is called a push-forward substitution. There is another type of
substitution called a pull-back substitution that we cover when we discuss trigonometric substitution.
Z 1
2
Example: Compute xe−x dx
0

10
 There are two special types of substitution that we do which don’t seem as obvious to fit into the form above.
Consider the following integral
Z b
f (x + c)dx
a
By letting u = x + c we obtain du = dx. The bounds (if present) change as u = a + c and u = b + c when
x = a and x = b respectively. This gives us
Z b Z b+c
f (x + c)dx = f (u)du
a a+c

I like to call these shifts.

Z
x
Example: Compute dx
(x + 1)2020

11
 Now consider the integral below,
Z b
f (kx)dx
a
where k 6= 0. We may use the substitution u = kx to obtain du = kdx. The bounds change as well to integrate
from u = ka to u = kb. Now we have a problem with the above, we don’t have a kdx term in the integral. We get
around this issue as follows:
Z b Z b Z kb
1 1
f (kx)dx = f (kx)(kdx) = f (u)du
a k a k ka
where we just use the fact that 1 = k/k to introduce the k term we need. I like to call these scalings.
Z π/15
Example: Compute sin(5x)dx.
0

12
1.2 (Section 8.1) Some Integration Techniques and Tricks
Most integral aren’t immediately clear on what the antiderivative is. Often algebraic tricks are required or some
clever use of identities.

1.2.1 Completing the Square

Very useful! This technique is use to turn two terms where a variable appears into a single term. For terms like
x2 + ax it turns them into shifts, which you can sub for. A reminder of how to complete the square:

Z
dx
Example: Compute √ .
(x − 2) x2 − 4x + 3

13
1.2.2 Using Trigonometric Identities
We will learn more about using trig identities as the course develops. Currently a very helpful identity to know if
cos2 (x) + sin2 (x) = 1. The example below is an example similar to the idea of the previous example, to turn two
terms into a single term that is easier to control.
Z π/2 p
Example: Compute 1 − cos(θ) dθ.
0

14
1.2.3 Long Division
This is for rational expressions where the degree of the numerator is higher than or equal to the degree of the
polynomial on the denominator. A reminder that a rational function is of the form:

P (x)
Q(x)
Where P (x) and Q(x) are polynomials. Fome examples are as follows:

To evaluate these integrals we use long division to put it into a more favorable form. This doesn’t always work,
but we will learn a another technique to pass such a roadblock when we study partial fractions in this chapter.
Z 3
x +x
Example: Compute dx.
x−1

15
(Continued...)

16
1.2.4 Rewriting (a matter of perception)
Look to rewrite an expression. Sometimes a good technique is to look for common factors. This is a little tricky
at first. When we say common factors, we don’t only mean common factors of a polynomial.
Z
Example: Compute (27e9x + e12x )1/3 dx.

17
1.2.5 Dumb Luck
I’m sorry to say but integrals, as you have probably seen at this point, are not as straight-forward to compute as
derivatives. We know really only one technique so far to compute integrals, substitution.

Sometimes we can make a substitution that changes the integral into something we can evaluate without a
good reason why it would have worked initially. Math is often preached as a logical step from one to the next, but
it honestly isn’t always that case. Some things are solved by trial and error. This isn’t a fault in math, this is a
fault in the nature of the universe and it’s reality.

Note: Saying an answer just came to you without a logical reason when there IS a logical reason does not
justify the result. It just comes off as a lack of work.
Z √
x
Example: Compute dx.
1 + x3

18
 Some ‘dumb luck’ methods are a bit more clear on where to start. Sometimes it’s because we want to reduce
the complexity of the number of terms. Below we do an example of using a substitution to ‘eliminate’ a constant
to avoid expansion.
Z
Example: Compute 27e−3x (1 + 3e−x )4 dx.

19
1.2.6 Mixing Techniques
It is possible you may have to mix techniques to complete a problem!

x−6
Z
Example: Compute √ dx.
8x − x2

20
1.2.7 Organizing Definite Integral Substitution
One last note is when evaluating integrals is to be as organized as possible. Usually integrals require several steps
and if another person, whether it be a colleague or a professor, sees work splattered randomly across the page or is
broken into several obscure segments...they will just give up on trying to follow your reasoning. Best way to learn
not to do this is to observe how people professionally present information and to mimic it into your technique.
Aesthetic and presentation is crucial. You also want to ease the amount of work you have to do by avoiding several
side calculations that don’t remain consistent with the flow of your solution.

One such issue I commonly see students do is to not substitute for the bounds in definite integrals when solving
something by substitution. Many extra steps are taken and it loses focus of the reader. While this is minor at
this stage, it does make a difference when we study trigonometric substitution. It cuts down on time drastically
when we substitute for the bounds in this section.
Z 4 Z −1/2
Example: Given that f (t) is a continuous function on R such that f (t)dt = 3 compute f (2t + 5)dt.
1 −2

21
1.3 (Section 7.1) Logarithms and Exponentials
The textbook goes through extensive measures to demonstrate how to reconstruct the exponential and all properties
of the logarithm if we initially define the natural logarithm by
Z x
1
ln(x) = dt, x>0
1 t
the process is long and worth a read. What we will do is give examples of computing logarithmic and expo-
nential type integrals.

1.3.1 Integrals Involving the Natural Logarithm

Z x
1
It is defined as ln(x) = dt or as the inverse function of f (x) = ex . Since it is the inverse of f (x) = ex then
1 t
eln(x) = x and ln(ex ) = x
The logarithm is amazing in that it handles exponents and multiplication operations extremely well.

Properties

• ln(ab) = ln(a) + ln(b)

• ln(a/b) = ln(a) − ln(b)

• ln(ab ) = b ln(a)

Note

The notation ln ab is ambiguous. For example, ln(x2 ) = ln(x · x) but (ln(x))2 = (ln(x)) · (ln(x)). In general
notice that ln(ab ) 6= (ln(a))b so you must always specify where to place the parenthesis unless an author
explicitly states what the notation means.
Z
dx
Since ln |x| = + C it is commonly found in dealing with integrals of the form
x
Z b 0
f (x)
dx
a f (x)

since u = f (x) gives du = f 0 (x)dx and thus

f (b)
b
f 0 (x) f (b)
Z Z
du 
dx = = ln |u|
a f (x) f (a) u f (a)
Z 0
3dx
Example: Compute
−1 3x − 2

22
 Z
ln(ln(x))
Example: Compute dx.
x ln(x)

Z
1
Example: Compute dx.
arctan(4x)(1 + 16x2 )

23
1.3.2 Integrals Involving the Exponential Function
The counterpart to the logarithm. It is denoted either f (x) = ex or f (x) = exp(x) and can be defined in several
different ways. It was first defined by the following limit
 x n
ex = lim 1 +
n→∞ n
Z x
dt
or you may define it as the inverse of ln(x) = . We will learn later that another common definition some
1 t
use is
n
x
X xk
e = lim
n→∞ k!
k=0

which will be discussed in detail in chapter 10.

It has the amazing property that

Z
d x
[e ] = ex and ex dx = ex + C
dx
Because it is what some might call a “fixed function” in calculus, it appears very commonly when solving most
problems.

Properties

• ea+b = ea eb
ea
• = ea−b
eb
1
• e−a =
ea
• (ea )b = eab

Note
These are just regular exponent laws. It is because, surprisingly, the way the exponential is defined as above
(using limits) you would not expect it to be a function with these properties! It turns out that exp(x) shares
all the properties that exponential laws give, thus we use an alternate notation exp(x) = ex to make things
easier for us. We represent the exponential function as a number to the variable because nothing changes
if we do this.

24
 Z √ln(π)
2 2
Example: Compute 2xex cos(ex )dx.
0

25
1.3.3 Other Logarithmic and Exponential Integrals
We have the logarithms and exponentials of other bases

f (x) = ax and g(x) = loga (x)

They are inverses and thus related by

aloga (x) x and loga (ax ) = x

with the integration and differentiation rules

ax
Z
d x
[a ] = (ln(a))ax ax dx = +C
dx ln(a)
Z
d 1
[loga (x)] = loga (x)dx = LEARN THIS LATER!
dx x ln(a)

Z e
2 ln(10) log10 (x)
Example: Compute dx.
1 x

26
 Z π/4  tan(t)
1
Example: Compute sec2 (t)dt.
0 3

27
1.4 (Section 7.2) Separable Differential Equations and Modeling
1.4.1 Separable Differential Equations
Definition
A differential equation is an equation involving an unknown function and its derivatives.

Definition
A separable differential equation (SDE) is one of the form y 0 = f (y)g(x).

Example: Which of the following are SDE’s?

dy dy dy dy
= ex+y = sin(x + y) = x2 log2 (x) ln(y) = xy + x
dx dx dx dx

Procedure for Solving an SDE

1. Provided g(y) 6= 0, divide both sides by g(y) to obtain

1 dy
= f (x)
g(y) dx

2. Integrate both sides with respect to x to obtain

Z Z
1 dy
dx = f (x)dx
g(y) dx

dy
3. Use the substitution rule for integrals dy = dx to obtain
dx
Z Z
1
dy = f (x)dx
g(y)

4. Integrate the previous expression

Note that solutions to SDE’s are usually implicit and can seldomly be solved for explicitly.

Example: Solve explicitly for y(x) if xy 0 = 1 + y 2 .

28
1.4.2 Unlimited Population Growth Model (Malthus Model)
Malthus Model
A population growth with no limitations imposed is modeled by


 dy = ky (Differential Equation)
dt
y(0) = y (Initial Condition)
0

where k is a constant and t is time. It has the unique solution

y = y0 ekt

Proof: Following the procedure of solving an SDE we have

1 dy
⇒ =k
y dt
Z Z
1 dy
⇒ dt = kdt
y dt
Z Z
1
⇒ dy = kdt
y

⇒ ln |y| = kt + C
By taking the exponential of both sides we obtain

eln |y| = ekt+C ⇒ |y| = eC ekt

As an additional assumption, we are using this to model the growth of a positive quantity of something (e.g.
population). As a result, we may assume that y(t) > 0 and so the above simplifies to

y(t) = eC ekt
Now we invoke the condition y0 = y(0) = eC e0 = eC (1) = eC and so the solution is given by

y(t) = y0 ekt


29
 Example: The biomass of a yeast culture in an experiment is 29g. After 30 min the mass is 37g. Assuming
the equation for unlimited population growth models this, how long will it take for the mass to double from it’s
initial population size?

30
1.4.3 Radioactive Decay
Radioactive Decay Model

A substance undergoing radioactive decay is modeled by


 dy = −ky (Differential Equation)
dt
y(0) = y (Initial Condition)
0

where k > 0. This has the unique solution

y(t) = y0 e−kt
which is obtained from the previous model by replacing k with −k. The constant k is usually obtained by
knowing the half-life of a substance. It satisfies the equation

ln(2)
Half-Life =
k

Example: The half-life of carbon-14 is 5730 years. Find the age of a sample in which 10% of the radioactive
substance has decayed.

31
1.4.4 Heat Transfer: Newton’s Law of Cooling
Newton’s Law of Cooling

If H(t) is the temperature of an object at time t and Hs is the constant surrounding temperature the
model is given by


 dH = −k(H − H ) (Differential Equation)
s
dt
H(0) = H (Initial Condition)
0

where k is a constant. You can obtain the unique solution by acknowledging the fact that because Hs is a
constant then

dH d
= [H − Hs ]
dt dt
d
thus the above differential equation becomes [H − Hs ] = −k(H − Hs ). By the previous model with
dt
y = H − Hs we obtain the solution H − Hs = (H0 − Hs )e−kt . Solving for H gives us

H(t) = Hs + (H0 − Hs )e−kt

Example: In Breaking Bad, Walter and Jesse cook “soup”. Accidentally the soup is heated to 98o C by Jesse
screwing up again. After Walter yells at Jesse and tells him ‘I am the one who cooks’ he tries to cool the soup
by immersing it in a container surrounded by 18o C water. After 5 minutes the temperature of the soup is 38o C.
The batch will be ruined if it doesn’t reach 20o C within 10 minutes. Is this batch spoiled?

32
1.5 (Section 8.2) Integration By Parts (IBP)
1.5.1 Establishing Integration by Parts and LIPET
Integration by parts is an integration rule that deals with undoing the product rule. Let u(x) and v(x) be
differentiable functions. Then by the product rule,

d dv du
[uv] = u +v
dx dx dx
Integrating both sides with respect to x over [a, b] yields
Z b Z b Z b
d dv du
[uv] dx = u dx + v dx
a dx a dx a dx
b Z b Z b
 dv du
⇒ uv  =
 u dx + v dx
a a dx a dx
Then rewrite this as
Z b
b Z b
dv  du
u dx = uv  − v dx
a dx a a dx
This is also commonly represented as the following.
Integration by Parts

Let u(x) and v(x) be differentiable functions on an interval (a, b) and continuous on [a, b]. Then...
Z b
b Z b

udv = uv  − vdu
a a a

The idea is that you manipulate the product rule to obtain a way to express an integral product in the form
of another integral product. The hope is that this rearrangement simplifies the integration. Notice that when you
dv
start with u on the left hand side you differentiate it in the right hand side. When you start with on the left
dx
hand side you find the anti-derivative v on the right hand side. In short, you often make one term in the product
simpler while you make the other term more complicated. The hope is that the simplification of one term outdoes
complicating the other to make the integral doable.
Z
Example: Compute x cos(x)dx using u = x and dv = cos(x)dx.

33
 Be careful to make good choices! Selecting which term to differentiate and which one to integrate matters!
Let’s revisit the previous example where we makeZthe other choice of assigning u and v.
Example: Compute one iteration of IBP for x cos(x)dx using u = cos(x) and dv = xdx.

You can see the previous example did not work out well after a single iteration. The purpose is to simplify the
expression! Not make it even more complicated! Thankfully there is an acronym called LIPET which helps us
determine which term to set as our u-term in some cases.
LIPET
Z
Consider an integral of the form f (x)g(x)dx where f (x) and g(x) are either a Logarithmic function,
Inverse trigonometric function, Polynomial function, Exponential function, or Trigonometric function.
LIPET is an acronym that tells you that you set your u term as the first function that is present on the
list:

• L = Logarithmic function

• I = Inverse trigonometric function

• P = Polynomial function

• E = Exponential function

• T = Trigonometric function

Example: The above LIPET system does not hold for the following integrals! Can you explain why? Ability
to find an antiderivative is NOT the reason!

1 − t2 t
Z p Z Z
x 1 − x2 dx ln(y)ey tan(y)dy e dt
1 + t2

34
1.5.2 Dominant Logarithms and Inverse Functions in LIPET
Logarithmic and inverse functions are dramatically more complicated that their derivatives. For this reason, they
often become a good first choice as what to set your u-term to be when performing integration by parts! The
derivative of these expressions are usually some rational type function.
Z e
Example: Compute x3 ln(x)dx.
1

35
1.5.3 Dominant Polynomials in LIPET
When we encounter polynomials as the first thing on the list in LIPET, it’s easy to see how the expression reduce
to a simpler form at each step! Specifically, these are integrals that are the product of a polynomial with either
an exponential or trigonometric function.
Z
Example: Compute (x2 + 2x) cos(2x)dx.

Z
Example: Compute x sec2 (x)dx.

36
1.5.4 Dominant Exponential or Trigonometric Functions in LIPET
These are integrals where both terms in the product cycle back to a multiple of their original form after differen-
tiation and integration. For example look at the chain of cos(x) after differentiating multiple times.

d d
[cos(x)] = − sin(x) → [− sin(x)] = − cos(x)
dx dx
and so after differentiating twice we are back to a multiple of cos(x). Similarly differentiating e5x once gives
5e5x , which is a multiple of itself. For these you use IBP by selecting one term in the product to keep being dif-
ferentiated and the other term to keep being integrated until we wind up with a multiple of our original integral.
Then you solve the equation for the integral.
Z
Example: Compute e3x sin(2x)dx.

37
1.5.5 Substitutions Leading to IBP
This is the integration by parts form of “Dumb Luck”. You make a substitution that happens to lead to IBP.
Z π2 √
Example: Compute sin( x)dx.
0

38
1.5.6 Integrals of Lonely Logarithms and Inverse Functions (Ninja’s)
Products might be hidden in plain sight when you encounter integrals. Such is the case for lonely integrands like
Z Z Z
arctan(x)dx log3 (x)dx (ln(x))2 dx

where a product that may be exploited by IBP is actually hidden. If f (x) is a logarithmic or inverse function
then you may rewrite the integral as
Z Z
f (x)dx = f (x) · 1dx

to which you set u = f (x) and dv = 1dx. This doesn’t necessarily fall under LIPET, since we don’t have
“constants” on that list. I would just remember it as a special exception.
Z
Example: Compute ln(x)dx.

Z
Example: Compute arccos(x)dx.

39
1.5.7 The Tabular Method
Sometimes you may have to use IBP several times. Fortunately enough there is a pattern you can use in a table
to obtain the result. Take for instance the following,

Z Z
000 00
uv dx = uv − u0 v 00 dx (1.1)
 Z  Z
= uv − u v − u v dx = uv − u v + u00 v 0 dx
00 0 0 00 0 00 0 0
(1.2)
Z
= uv 00 − u0 v 0 + u00 v − u000 vdx (1.3)

Notice how one term keeps being applied derivatives to while the other term keeps being integrated. The
product is taken between the two and the sign alternates. We can record this in a table quite conveniently.
Z
Example: Use the tabular method on the previous example e3x sin(2x)dx.

40
 Z
Example: Use the tabular method to compute (x4 + 3x2 )3x dx.

Note
Under LIPET, the tabular method works best on polynomial, exponential, or trigonometric dominant
integrals. It does not work well on logarithmic or inverse trigonometric dominant integrals.

41
1.6 (Section 8.3) Trigonometric Integrals
We know how to integrate some basic trigonometric functions.
Z
• sin(x)dx = − cos(x) + C
Z
• csc2 (x)dx = − cot(x) + C
Z
• tan(x)dx = ln | sec(x)| + C

and so forth. But what about more complicated terms like

Z Z
sin2020 (x) cos3 (x)dx and tan2 (x) sec2 (x)dx?

The textbook Thomas’ Calculus really doesn’t give you justice on the techniques to solve these problems. It
avoids categorizing techniques to not add to the density of material, but it’s better to know it than attack these
problems half blind!

1.6.1 Products of Sines and Cosines of Different Composition

This is to deal with products of sines and cosines, not raised to a power, where the inputs of both differ. For
example integrals of the form
Z
cos(5x) sin(2x)dx

To deal with these, due to the fact that both terms in the product, do cycle, you can use “repetitive in all
terms” IBP. However, a quicker trick is to use the product to sum identities.

Procedure for Integrals Consisting of a Products of Sine and/or Cosine with Different Inputs

1. Use one of the following three appropriate product to sum formulas:

1
• cos(mx) cos(nx) = (cos((n + m)x) + cos((n − m)x))
2
1
• sin(mx) sin(nx) = (cos((n − m)x) − cos((n + m)x))
2
1
• sin(nx) cos(mx) = (sin((n + m)x) + sin((n − m)x))
2
2. Integrate the remaining basic integral.

These are the so called “product to sum” formulas. For example,

1 1
(cos((3 − 2)x) − cos((3 + 2)x)) = (cos(x) − cos(5x))
sin(3x) sin(2x) =
2 2
which turns the product into something easily, and quickly, integrable.

42
 Z
Example: Compute cos(2x) sin(4x)dx.

43
1.6.2 Products of Sines and Cosines, Same Input, Raised to a Power
These are integrals of the form
Z b
sinm (x) cosn (x)dx
a
where the input might be scaled or shifted. These divide into two cases:

Z b
Procedure for Computing sinm (x) cosn (x)dx
a

• m and/or n is odd:

1. Take off a single term from the odd power and collect it with your dx term to form a du.
2. Now that the remaining terms are raised to an even power, convert them to the other trigono-
metric function using sin2 (x) + cos2 (x) = 1.
3. Complete the u-substitution and integrate with the du term formed in the first step.

• m and n are even:

1. Convert all terms using the identities

1 − cos(2x) 1 + cos(2x)
sin2 (x) = and cos2 (x) =
2 2
2. If all your terms are not basic integrals, expand everything.
3. If you encounter all even terms repeat the first step on those integrand terms. Else, your integrand
terms are either of the case where you have a basic integral or m and/or n are odd. Use the
appropriate procedure on those.
Z
Example: Compute sin3 (x) cos3 (x)dx.

44
 Z
Example: Compute cos7 (x)dx.

45
 Z
Example: Compute sin2 (x) cos2 (x)dx.

46
1.6.3 Eliminating Roots
p
Procedure for Computing Integrals Containing 1 ± cos(mx) where m is Even

1. Use the double angle identity 1 + cos(2x) = 2 cos2 (x) or 1 − cos(2x) = 2 sin2 (x) to eliminate the square
root.

2. Use appropriate integration techniques afterwards.

Z π/4 p
Example: Compute 1 − cos(4x)dx.
0

47
 p
Procedure for Computing Integrals Containing 1 ± sin(nx)
p
1 ∓ sin(nx)
1. Multiply the integrand by the conjugate p
1 ∓ sin(nx)

2. Expand the terms and use the identity 1 − sin2 (nx) = cos2 (nx) to eliminate the square root.

3. Use appropriate integration techniques afterwards.

Z π/4 p
Example: Compute 1 − sin(2x)dx.
0

48
1.6.4 Lonely Powers of Tangent or Secant
This is for integrals of the form
Z Z
m
tan (x)dx or secn (x)dx

Unlike the case with powers of sine and cosine, they require their own categorization.
Z b Z b
m
Procedure for Computing tan (x)dx or secn (x)dx
a a

• n ≥ 3:

1. Pull of a sec2 (x) term and perform IBP with dv = sec2 (x)dx. Your integral will reduce as the
following reduction formula:

secn−2 (x) tan(x) n − 2

Z Z
n
sec (x)dx = + secn−2 (x)dx
n−1 n−1
2. Complete the integration if you are at a basic integral of sec2 (x) or sec(x). Otherwise, repeat
the first step on the appropriate integral.

• m ≥ 3 is odd:

sec(x)
1. Pull off a single tangent term to and multiply by to form
sec(x)
Z Z
m sec(x) tan(x)
tan (x)dx = tanm−1 (x) dx
sec(x)
2. Construct du = sec(x) tan(x)dx and convert the remaining tangent terms to secants using
tan2 (x) = sec2 (x) − 1.
3. Complete your u-substitution using u = sec(x) and integrate.

• m ≥ 2 is even:

1. Convert all your tangent terms into secants using the identity tan2 (x) = sec2 (x) − 1.
2. If the result is a basic integral, integrate it. Otherwise, expand the result to construct a sum of
powers of secants.
3. Integrate the basic integral terms of sec2 (x), sec(x) and constants, then use the procedure of
integrating higher powers of secants mentioned above.
Z
Example: Compute tan2 (x)dx.

49
 Z
Example: Compute sec4 (x)dx.

50
 Z
Example: Compute sec3 (x)dx.

51
1.6.5 Powers of Tangent and Secant with Same Input
This is for integrals of the form
Z
tanm (x) secn (x)dx

where the inputs may be shifted or scaled. The strategies for power products of cosecant and cotangent are
identical. In this, it is assumed that m, n ≥ 1 (not a lonely integrand).

Z b
Procedure for Computing a Non-Lonely tanm (x) secn (x)dx
a

• m is odd:

1. Pull off a single tan(x) and sec(x) term and construct du = sec(x) tan(x)dx.
2. Convert the remaining tangent terms using the identity tan2 (x) = sec2 (x) − 1.
3. Complete the u-substitution with u = sec(x) and integrate.

• n is even:

1. Pull of a sec2 (x) term and construct du = sec2 (x)dx.

2. Convert the remaining secant terms using the identity sec2 (x) = tan2 (x) + 1.
3. Complete the u-substitution with u = tan(x) and integrate.

• m is even and n is odd:

1. Convert all the tangent terms to secants using the identity tan2 (x) = sec2 (x) − 1.
2. Expand your integrand to get a sum of powers of secants.
3. Use the appropriate integration techniques for integrating lonely powers of secants.
Z
Example: Compute sec(x) tan3 (x)dx.

52
(Continued...)

Z
At Home Exercise: Use the appropriate procedure to compute tan2 (x) sec3 (x)dx

53
1.7 (Section 8.4) Trigonometric Substitution
1.7.1 Forms of Trigonometric Subsitution
These are used to deal with integrals containing the terms

a2 − u2 or a2 + u2 or u2 − a2
where a is a constant and u is the variable of integration. This is where we use a substitution method called a
pullback substitution. The two substitution methods pullback and push-forward are as follows:

• Pullback: Substitutions of the form x = g(u).

• Push-forward: Substitutions of the form u = h(x).

The above terms resemble trigonometric identities

• a2 − u2 resembles 1 − sin2 (θ) = cos2 (θ)

• a2 + u2 resembles 1 + tan2 (θ) = sec2 (θ)

• u2 − a2 resembles sec2 (θ) − 1 = tan2 (θ)

Example: Convert the following integral

x4
Z
dx
9 + 4x2
to a trigonometric integral using the substitution 2x = 3 tan(θ). Do not complete the integration.

54
Procedure for Computing Integrals Containing u2 ± a2 and a2 − u2

The procedure for all the following is essentially the same. Use the substitution u = a ×
(Appropriate Trig Function), use techniques of trigonometric integrals to complete the integration, then
convert back.

• Containing a2 − u2 :

1. Let u = a sin(θ) and compute du = a cos(θ)dθ.

2. Complete the substitution and use an appropriate integration technique to compute the resulting
trigonometric integral.
3. Convert all terms back. If you encounter a lonely θ use the identity θ = arcsin(u/a). If you
encounter a double angle, reduce to a single angle using a double angle identity. If you encounter
u
a non-sine function, represent the identity sin(θ) = as a triangle and solve for the values of
a
other trigonometric functions.

• Containing a2 + u2 :

1. Let u = a tan(θ) and compute du = a sec2 (θ)dθ.

2. Complete the substitution and use an appropriate integration technique to compute the resulting
trigonometric integral.
3. Convert all terms back. If you encounter a lonely θ use the identity θ = arctan(u/a). If you
encounter a double angle, reduce to a single angle using a double angle identity. If you encounter
u
a non-sine function, represent the identity tan(θ) = as a triangle and solve for the values of
a
other trigonometric functions.

• Containing u2 − a2 :

1. Let u = a sec(θ) and compute du = a sec(θ) tan(θ)dθ.

2. Complete the substitution and use an appropriate integration technique to compute the resulting
trigonometric integral.
3. Convert all terms back. If you encounter a lonely θ use the identity θ = arcsec(u/a). If you
encounter a double angle, reduce to a single angle using a double angle identity. If you encounter
u
a non-sine function, represent the identity sec(θ) = as a triangle and solve for the values of
a
other trigonometric functions.

55
1.7.2 Sine Substitutions
π π
In these integrals we always have a domain restriction of − ≤ θ ≤ .
2 2
Z
x 3
Example: Compute √ dx.
5 − 4x2

56
1.7.3 Tangent Substitutions
π π
In this we always have a domain restriction of − < θ < as well.
2 2
Z
1
Example: Compute dx.
(4x2 + 9)2

57
1.7.4 Secant Substitutions
For terms of the form u2 − a2 you must have a restriction of |u| > a for secant substitution to make any sense. If
π π
u > a then 0 < θ < and if u < −a then < θ < π.
2 2
Z √ 2
x − 25
Example: Compute dx.
x

58
1.7.5 More Obscure u Terms in Trigonometric Substitution
√
ln(3 3)
et
Z
Example: Compute √ dt.
ln(3) e2t + 9

59
 √ √
Z
Example: Compute x 1 − xdx.

60
1.8 (Section 8.5) Partial Fractions
This is, quite bleakly stated, as the process of ‘unsimplifying’ a fraction to integrate it.
3x + 11 4 1
Example: Since = − then
x2 − x − 6 x−3 x+2
Z Z Z
3x + 11 4 1
2
dx = dx − dx = 4 ln |x − 3| − ln |x + 2| + C
x −x−6 x−3 x+2

P (x)
For f (x) = , a rational function with deg(P (x)) < deg(Q(x)) where P (x) and Q(x) are polynomials you
Q(x)
start by factoring Q(x). We want to express, like above, this rational in the form

p1 (x) p2 (x) pn (x)

f (x) = + + · · · + sn
q1s1 (x) q2s2 (x) qn (x)

where deg(pi (x)) =deg(qi (x)) − 1 (this allows us to represent it in the form of a logarithmic integral). This
form is called the partial fraction decomposition of f (x). The forms we suggest in the decomposition depend
on how Q(x) factors.

1.8.1 Linear Terms Present

Suppose Q(x) is factored, the following terms present means we suggest the following form.

Term in Q(x) Suggested Term(s) in Decomposition to Add

A
ax + b
ax + b

A1 A2 Ak
(ax + b)k + + ··· +
ax + b (ax + b)2 (ax + b)k

3x + 4
Example: Suggest a form for the partial fraction decomposition of f (x) = .
(2x + 1)2 (x − 1)

61
1.8.2 Irreducible Quadratic Terms Present
Definition

A quadratic of the form ax2 + bx + c is called irreducible if b2 − 4ac < 0. That is, it cannot be factored
into linear factors over the real numbers.

Suppose Q(x) is factored, the following terms present means we suggest the following form.

Irreducible term in Q(x) Suggested Term(s) in Decomposition to Add

Ax + B
ax2 + bx + c
ax2 + bx + c

A1 x + B1 A2 x + B 2 Ak x + Bk
(ax2 + bx + c)k 2
+ 2 2
+ ··· +
ax + bx + c (ax + bx + c) (ax2 + bx + c)k

3x2 + 4x
Example: Suggest a form for the partial fraction decomposition of f (x) = .
(x2 + 1)2 (x − 1)2 (x − 3)

62
 x+1
Example: Construct the partial fraction decomposition f (x) = .
x4 + x2

63
1.8.3 Special Case Where All Terms Are Linear
This is the BEST case scenario. You can solve for all the constants in the easiest manner possible. It’s best to
see this “trick” by example.

3x2 + 1
Example: Compute the partial fraction decomposition of f (x) = .
(x − 3)(x − 4)(x − 2)

64
1.8.4 Example of Computation an Integral by Partial Fractions
x2 − x + 2
Z
Example: Compute dx.
x3 − 1

65
(Continued...)

Note
If the degree of the denominator is higher than the numerator, you must use long division first to re-express
the integrand! Note also that some of the integrals which may be solved by partial fractions may also be
solved by trigonometric substitution.

66
1.9 (Section 8.8) Improper Integrals
1.9.1 Defining Improper Integrals and Convergence
This section truthfully makes more sense upon the introduction of sequences, but we make do for the present
time. These are integrals over a region of space that, in a sense, are infinite in size. You will find that some re-
sults might be counter intuitive at first, although the results are entirely based upon the way that things are defined.

Consider the function y = e−x/2 . As this function is entirely positive it makes sense to talke about the area
Z b
between it and the x-axis over the region [0, b] for some value b. This defines the area function A(b) = e−x/2 dx.
0

For each value of b the area is finite. Indeed, we may compute

Z b
b
−x/2 −x/2  −b/2

A(b) = e dx = −2e  = −2(e − 1)
0 0
For each increased value of b we extend the area. It is possible that the area settles down to a finite value even
though the region where we take it over is infinite. That is, by taking the limit as b → ∞ we obtain

lim A(b) = lim −2(e−b/2 − 1) = −2(0 − 1) = 2

b→∞ b→∞

This is an illustration of an improper integral. There are two types of improper integrals, one where the region
of integration is infinite and the other where there is a vertical asymptote in the region of integration.

67
 Definition
An integral with infinite limits (i.e. integral over an infinite region) is called a Type I Integral. It is
defined as:

Z ∞ Z b 
1. f (x)dx = lim f (x)dx
a b→∞ a
Z b Z b 
2. f (x)dx = lim f (x)dx
−∞ a→−∞ a
Z ∞ Z c Z ∞
3. f (x)dx = f (x)dx + f (x)dx, where c is any real number and the two integrals are defined
−∞ −∞ x
as in the previous two points.

All these definitions assume that f (x) is continuous over the region of integration.

Often we care little about the value of the integral. We often care about the behaviour of the integral. There
are two options, either an integral results in concrete finite value or it does not.
Z ∞
1
Example: Determine whether the following integral dx is a finite value or not.
1 x

Definition
If an integral results in a finite value we say the integral converges. If an integral is not convergent we say
it is divergent. The present tense verbal forms of these are that the integral converges or it diverges.

68
1.9.2 p-Integrals and the Integral Comparison Tests
Often we determine convergence or divergence of an integral by comparing it (in some manner) to a simpler
and more well known integral that we understand the behaviour of. The simplest integrals we understand the
convergent/divergent behaviour of are the p-Integrals.
Definition
Z ∞
dx
A Type I p-integral is an integral of the form .
1 xp

Theorem
Consider the Type I p-integral. If p > 1 the p-Integral converges. If p ≤ 1 the p-Integral diverges.

Proof: If p = 1 then by the previous example we saw that the integral diverges. If p > 1 then by the power
rule we obtain the following
Z ∞ ∞  
dx 1 1−p 
 1 1−p
= x  = lim b −1
1 xp 1−p 1 1 − p b→∞
We can see that if p > 1 then the exponent of b1−p is negative and thus
1
p < 1 ⇒ lim =0
b→∞ bp−1

On the other hand, if p < 1 then the exponent of b1−p is positive and thus

p > 1 ⇒ lim b1−p = ∞

b→∞

thus establishing the above theorem as desired .

Note
Z b
dx
Many people casually call every integral of the form p
a p-integral. There is no problem with this
a x Z ∞
dx
usually because the context is often clear. Usually, an integral of the form where a > 0 is colloquially
a xp
Z b
dx
called a Type I p-integral. On the other hand, integrals of the form p
where b > 0 is colloquially called
0 x
a Type II p-integral.

69
Below is the theorem we use to compare unknown integrals of Type I Improper to those that are known.

Direct Comparison Test

Let f and g be continuous functions on [a, ∞) with 0 ≤ f (x) ≤ g(x) for all x ≥ a. Then...
Z ∞ Z ∞
1. f (x)dx converges if g(x)dx converges.
a a
Z ∞ Z ∞
2. g(x)dx diverges if f (x)dx diverges.
a a
Z ∞
dx
Example: Determine whether or not the integral converges or diverges.
1 1 + x4

Z ∞
dx
Example: Determine whether or not the integral √ converges or diverges.
6 6 x2 − 25

70
We have one more test for Type I Integrals.

Limit Comparison Test

Let f and g be continuous and positive functions on [a, ∞) with

f (x)
lim =L
x→∞ g(x)
Z ∞ Z ∞
where L is positive and finite, i.e. 0 < L < ∞. Then f (x)dx and g(x)dx converge together or
a a
diverge together.

∞
1 − e−x
Z
Example: Determine whether or not the integral dx converges or diverges.
1 x

71
1.9.3 Vertical Asymptotes
You can only invoke the Fundamental Theorem of Calculus if the function is continuous over the region of inte-
gration. Thus the Fundamental Theorem of Calculus does not apply to the integral

Z 1
dx
−1 x2

1

Finding an antiderivative and computing F (x) will give you the incorrect answer. This is because there is
−1
a vertical asymptote disrupting the continuity over this region of integration. We define such functions as below.

Definition
If the integrand is discontinuous (and has a vertical asymptote) in the region of integration we call it a
Type II Integral. If f (x) is continuous over (a, c) ∪ (c, b) (i.e. only has a vertical asymptote at x = c)
then we define
Z b Z m  Z b 
f (x)dx = lim f (x)dx + lim f (x)dx
a m→c− a x→c+ n

Z 1
1
Example: Determine whether or not the integral dx converges or diverges.
−1 x2

72
 Definition
Z 1
dx
A Type II p-integral is an integral of the form .
0 xp

Theorem
Consider the Type II p-integral. If p ≥ 1 the p-integral diverges. If p < 1 the p-integral converges.

Note
This is almost opposite of the result for Type I p-integrals.

The Type II improper integrals has their own comparison theorem. We will omit the limit comparison version
and just mention the direct comparison one to keep things simple.

Direct Comparison Test for Type II Integrals

Let f and g be continuous functions on I = (a, c) ∪ (c, b) with 0 ≤ f (x) ≤ g(x) on I. Then...
Z b Z b
1. f (x)dx converges if g(x)dx converges.
a a
Z b Z b
2. g(x)dx diverges if f (x)dx diverges.
a a

73
1.9.4 Intuition is Lost in the Land of Infinity
Gabriel’s horn is a vuvuzela like surface obtained by rotating the curve f (x) = 1/x around the x-axis where x ≥ 1.

∞ ∞
Z Z r
1 1 1
It has a volume given by V = π dx and surface area SA = 2π 1+ dx.
1 x2 1 x x4
Example: Show that the surface area of Gabrielle’s Horn is infinite, yet the volume is finite.

74
1.10 (Section 6.1) Volumes Using Cross Sections
1.10.1 Defining Volumes by Cross Sections
We approach computing the volume of a solid by use of cross-sections. Cross-sections are region formed by
intersecting a solid with a plane.

The idea is to find the area of each cross section and then add them up, through means of integration, to get
the volume. Thus in this chapter you should not think of integration as area under a curve but rather the more
physical interpretation of summing a quantity over a region.

Definition
Z b
The volume of a solid with cross sectional area A(x) from x = a to x = b is V = A(x)dx
a

Procedure

1. ? SKETCH THE SOLID AND A TYPICAL CROSS-SECTION ?

2. Find a formula for A(x)

3. Find the limits of integration

Z b
4. Compute V = A(x)dx
a

Cavalieri’s Principle

If two solids have the same height and cross-sectional area at every point along that height, then they have
the same volume.

75
Example: Find the volume of a pyramid of height h whose base is square with sides of length L.

76
 Example: Find the volume of a cheese wedge cute from a circular cylinder of radius r if the angle between
the top and bottom is π/6 radians.

77
1.10.2 Solids of Revolution and their Volume by Washers
A solid of revolution is a solid obtained by rotating a curve around a line parallel (or equal) to an axis.

One may notice that the cross-sections are “so-called” washers (flat donut shapes).

To find the area of a washer we use the fact that it’s area is the difference of two circles.

78
 p
Example: Continue with the example above. Consider the region bounded by x = y + 4, x = 1, y = 0 and
y = 5 and form the solid of revolution by revolving the region about the y-axis. Construct the cross-sectional area
A(y) at each moment 1 ≤ y ≤ 5 and use this to construct the volume of the solid.

79
 The Washer Method for Solids of Revolution
Let R be a region in the xy-plane that is revolved about an axis. If it is revolved around a line of the form
y = C (including the x-axis) then your volume is given by
Z b
V =π (R2 − r2 )dx
a

where A(x) = R(x)2 − r(x)2 represents the cross sectional area as the difference of the bigger and smaller
radii over the region [a, b]. If it is revolved around a line of the form x = C (including the y-axis) then your
volume is given by
Z d
V =π (R2 − r2 )dy
c
2 2
where A(y) = R(y) − r(y) represents the cross sectional area as the difference of the bigger and smaller
radii over the region [c, d].

In either case, if r = 0 we call this The Disk Method instead.

Note
The picture is what tells you how to set up your R and r. There is no formula for constructing it and
is situation specific depending on your picture. Don’t try asking for one because if anyone has a specific
formula for you that covers “all cases”, I can construct a counterexample where it fails.

Example: Find the volume of the solid of revolution formed by the region bounded by y = x2 − 2x and y = x
and revolving it about the line y = 4.

80
(Continued...)

81
1.11 (Section 6.2) The Shell Method
This section covers an alternate method to compute the volume of a solid of revolution. This is the method by
looking at a solid more like a matryoshka (Russian) doll. For these, the idea is to find the surface area of a “shell”
and add them up. If you want to think of the washer method as deconstructing a cake into its pancake like layers,
then the shell method is like deconstructing an onion into its spherical like layers (as opposed to rings).

In this formulation of the solid, we see that the three-dimensional object is layered like a Russian doll where
each layer is a tube-like cylinder.

If we imagine that the radius of a cylinder is r and the height is h, then the surface are of this object is A = 2πrh.

82
 Once again you integrate to add them up, obtaining the volume.

The Shell Method for Solids of Revolution

Let R be a region in the xy-plane that is revolved about an axis. If it is revolved around a line of the form
y = C (including the x-axis) then your volume is given by
Z d
V = 2π rhdy
c
where A(y) = 2πr(y)h(y) represents the surface area of a typical shell where the domain of the radius is
[c, d]. If it is revolved around a line of the form x = C (including the y-axis) then your volume is given by
Z b
V = 2π rhdx
a
where A(x) = 2πr(x)h(x) represents the surface area of a typical shell where the domain of the radius is
[a, b].

Note
The variable you integrate will be opposite of that for washers! This becomes very evident when drawing a
picture!

Example: Consider the region bounded by the curve y = sin(x) and the x-axis over [0, π]. Find the volume
of the solid of revolution obtained by revolving this region about the y-axis.

83
 (Continued...)

Example: Determine the volume of the solid obtained by rotating the region bounded by x = (y − 2)2 and
y = x about the line y = −1.

84
(Continued...)

85
1.12 (Section 6.3) Arc-Length
1.12.1 Formulating the Arc-Length
The purpose of this section is to both construct a good definition for the length of a curve and give some examples.
We start with a curve y = f (x) and consider approximating the length by computing the lengths of the segments
through a series of consecutive points P1 , ..., Pn as in the image below.

If we let |Pi−1 Pi | represent the length of the line segment connecting point Pi−1 (xi−1 , yi−1 ) to point Pi (xi , yi )
then we form an approximation given by
n
X
Length ≈ |Pi−1 Pi |
i=1
b−a
For the sake of argument, lets assume the points are evenly spaced so that ∆x = xi − xi−1 = . Then, we
n
obtain the approximation
s
n p n  2
X X f (x i ) − f (xi−1 )
Length ≈ (xi − xi−1 )2 + (f (xi ) − f (xi−1 ))2 = 1+ ∆x
xi − xi−1
i=1 i=1
whereby we use the Mean Value Theorem (MVT) to say that the secant slopes can be represented in terms
of the derivative for an intermediary value in the interval (xi , xi−1 ). That is, (f (xi ) − f (xi−1 ))/(xi − xi−1 ) = f 0 (x∗i )
for some point x∗i in (xi , xi−1 ). Thus we obtain
n p
X
Length ≈ 1 + (f 0 (xi ))2 ∆x
i=1
Now, as we take the limit as n → ∞ our approximation gets better.

86
But as we take this limit we have, by definition, a formulation of the definite integral.
n q Z bp
DEF
X
Length = lim 0 ∗ 2
1 + (f (xi )) ∆x = 1 + f 0 (x)2 dx
n→∞ a
i=1

and similar formulations may be made for curves expressed in the form of x = g(y).
Definition
Let y = f (x) be a function that is continuous on [a, b] and differentiable on (a, b). The arc-length of the
graph of its representing curve over [a, b] is
Z bp
L= 1 + f 0 (x)2 dx
a
Similarly, let x = g(y) be a function that is continuous on [c, d] and differentiable on (c, d). The arc-length
of the graph of its representing curve over [c, d] is
Z dp
L= 1 + g 0 (y)2 dy
x
Z
In either case, we commonly abbreviate an Arc-Length integral as ds.

87
1.12.2 Computation of Arc-Length
x2 ln(x)
Example: Find the length of y = − over [1, 3].
2 4

3x 2/3
 
2
Example: Show that the arc-length of the function y = + 1 over 0 ≤ x ≤ 32/3 is much more
2 3
difficult to compute in terms of x instead of in terms of y. Compute the length.

88
Example: Determine the length of y = ln(sec(x)) over the interval [0, π/4]

89
1.12.3 The Surface Area of a Solid of Revolution
Definition
Consider the solid of revolution obtained by revolving the function y = f (x) over the interval [a, b] about
the x-axis. We define the surface area as
Z b Z b p
SA = 2πf (x)ds = 2πf (x) 1 + f 0 (x)2 dx
a a
Similarly, if we revolved a function x = g(y) over the interval [c, d] about the y-axis then we define the
surface area of the resulting solid of revolution to be
Z d Z d p
SA = 2πg(y)ds = 2πg(y) 1 + g 0 (y)2 dy
c c
p
Example: Determine the surface area of the solid obtained by rotating y = 9 − x2 in [−2, 2] about the
x-axis.

90
Chapter 2

Series

91
2.1 (Section 10.1) Introduction to Sequences
Definition
A sequence is an ordered list a1 , a2 , a3 , ... and may be abbreviated as {ai } or {ai }∞
i=1

These sequences may literally be any list of objects. In mathematics we always imply it to be infinite if no
context is given although people use the term sequence colloquially to describe lists of finite objects.

Example: You can have a sequence of animals: {Dogs, Birds, Quoakkas, Lizards, Picasso’s Cats, ...}

Example: You can have a sequence of dates: {June 3rd, July 12th, August 20th, June 3rd, September 10th,...}

Hence some sequences can repeat elements in them but their order is important.

2.1.1 Ways to Describe a Sequence

There are, in essence, three common ways to describe a sequence.
Explicit Description of a Sequence

These are sequences of the form an = f (n) for some function where n is a sequence of integers starting from
some initial value n ≥ k.

n(n + 1)
Example: Triangular numbers are given by the sequence Tn = where n ≥ 1. We compute
2
1(2) 2(3) 3(4) 4(5)
T1 = =1 T2 = =3 T3 = =6 T4 = = 10
2 2 2 2

Example: Arithmetic Sequences are given by an = a + (n − 1)d for n ≥ 1 where a is some initial value and
d is a called the common difference. Here you start at some value a and add the number d to obtain the next
number in the sequence. For example if an = 10 + 5(n − 1) for n ≥ 1 the first few terms of the sequence are...

a1 = 10 + 5(0) = 10 a2 = 10 + 5(1) = 15 a3 = 10 + 5(2) = 20 a4 = 10 + 5(3) = 25

so you start at a1 = 10 and add 5 each time to get the next number.

92
 Recursive Definition of a Sequence

These are sequences of the form an = f (a1 , a2 , ..., an−1 ) for n ≥ k where initial values a1 , a2 , ..., ak−1 are
given. You generate the new values from the previous values.

Example: The most famous sequence is undeniably the Fibonacci sequence and is given by fn = fn−1 + fn−2
for n ≥ 2 and f0 = 1, f1 = 1. Give the next few numbers in the Fibonacci sequence.

Example: The logistic recursive sequence is given by xn+1 = rxn (1 − xn ) for n ≥ 0 where x0 is some specified
3 1
initial value in [0, 1]. For example if k = and x0 = then
2 2
  
3 1 1 3
x1 = 1− =
2 2 2 8
  
3 3 3 45
x2 = 1− =
2 8 8 128
  
3 45 45 11205
x3 = 1− =
2 128 128 32768
and so forth. This sequence is best calculated using a computer. It exhibits a very interesting behaviour as
you vary the parameter r. I recommend reading about the relation of the Logistic Map and Chaos Theory. Below
is an excellent video provided by Veritasium that described the Logistic Map:

==Link to Veritasium Video==

93
The Description of a Sequence by a Mathematical Anarchist

This is where you specify a sequence by random rules described in English (Well, English for this class).

Example: The look and say sequence is the sequence as follows

1, 11, 21, 1211, 111221, 312211, ...

Determine the rule for generating elements of this sequence and generate the next two numbers of the sequence.

94
2.1.2 Convergence and Divergence
Sequences (like we’ve seen with improper integrals) either approach a single finite value or they do not. That is,
they either converge or diverge.

n2 + 1
Example: Consider the sequence an = where n ≥ 1. Determine the limit as n → ∞. Does it converge
3n2 + 3
or diverge?

cos(n)
Example: All previous limit laws hold. Squeeze, l’Hôpital, etc. For example consider an = where
n
n ≥ 1. Determine the limit as n → ∞. Does it converge or diverge?

95
  nπ 
Example: Consider the sequence an = cos where n ≥ 0. Determine the limit as n → ∞. Does it
2
converge or diverge?

√ √
Example: Given that the recursive sequence an+1 = 2 + an , for n ≥ 1 with a1 = 2 converges, determine
it’s limit.

96
 Example: Provided a sequence converges you can determine the limit of a recursive equation by setting up
an equation you may solve for. For example consider the sequence obtained by successive ratio’s of the Fibonacci
sequence Rn = fn+1 /fn where n ≥ 0.

97
2.1.3 Useful Results
Theorem
If f (x) is continuous and {an } is a sequence in the domain of f then lim f (an ) = f ( lim an ).
n→∞ n→∞

n
Example: Consider f (x) = ln(x) and an = where n ≥ 1. Determine lim f (an ).
n+1 n→∞

Theorem
These are useful limits to memorize. As n → ∞,

• ln(n)/n → 0

• n1/n → 1

• (ln(n))1/n → 1

• a1/n → 1 if a > 0

• an → 0 if |a| < 1
 a n
• 1+ → ea
n
• an /n! → 0
 n
n
Example: Consider an = for n ≥ 1. Determine the limit as n → ∞.
n+1

98
2.1.4 Monotonic Sequences and the Monotonic Convergence Theorem
Definition
A sequence {an } is...

• non-increasing if an+1 ≤ an for all n.

• non-decreasing if an+1 ≥ an for all n.

A sequence that is either non-increasing or non-decreasing is also called monotonic.

Definition
A sequence {an } is bounded...

• below by M if an ≥ M for all n.

• above by M if an ≤ M for all n.

• by M if |an | ≤ M for all n.

1
Example: Consider an = where n ≥ 1. This is decreasing and bounded below by M = −2. Explain why.
n

Monotonic Convergence Theorem

If {an } is either bounded below and non-increasing or bounded above and non-decreasing then the sequence
converges.

an−1
Example: Argue that the sequence an = where n ≥ 1 converges without solving for it explicitly.
n

99
2.2 (Section 10.2) Infinite Series
2.2.1 Defining an Infinite Sum
Intuitive Description: An infinite series (or commonly just called a series) is the sum of an infinite sequence of
numbers. That is, if {ak }∞
k=1 is a sequence then a series whose terms are this sequence is

∞
X
ak = a1 + a2 + a3 + ....
k=1

The problem with this definition is that it doesn’t give context of what it means to take an infinite sum. We
understand limits and can work from there. It should illustrate the fact that as we taking more values to add, we
are approaching some value. Thus we need to somehow incorporate this into the definition.

(Actual) Definition

Given a sequence {ak } as above we define the sequence of partial sums

n
X
Sn = ak = a1 + a2 + · · · + an
k=1

We say that the series

∞
X
ak
k=1

converges to L if lim Sn = L. If the Sequence {Sn } diverges we say the series diverges.
n→∞

Now, most of the time you will never be able to find this limit L. It is often impossible and just calculated
using very abstract techniques or approximated by computers. Mostly we care about convergence or divergence.
This section focuses on two series you can actually find the value of while the rest of this chapter focuses on just
determining convergence or divergence without ever being able to find this value.

Example: Consider the series

∞
X
(−1)k
k=1
n
X
We have that the partial sums are Sn = (−1)k . We compute a few terms to find that
k=1
S1 = (−1)1 = −1
S2 = (−1)1 + (−1)2 = −1 + 1 = 0
S3 = (−1)1 + (−1)2 + (−1)3 = −1 + 1 − 1 = −1
S4 = (−1)1 + (−1)2 + (−1)3 + (−1)4 = −1 + 1 − 1 + 1 = 0
At this point you can probably tell that {Sn }∞
n=1 = {−1, 0, −1, 0, −1, 0, −1, ...}. We see that for this sequence
lim Sn does not exist. It oscillates between two values. Therefore the above series diverges.
n→∞

100
2.2.2 Geometric Series
Definition
A geometric series is a series of the form
∞
X
ark−1 = a + ar + ar2 + ar3 + · · ·
k=1

Theorem
If |r| < 1 then
∞
X a
ark−1 =
1−r
k=1

else if |r| ≥ 1 the series diverges.

n
X
Proof: Consider the partial sum Sn = ark = a + ar + ar2 + · · · + arn−1 . We may derive an explicit formula
k=1
for the partial sums of this particular series as follows. Scale partial sum by r to obtain

rSn = ar + ar2 + ar3 + · · · + arn

Then form the difference Sn − rSn to obtain

(1 − r)Sn = Sn − rSn = (a + ar + ar2 + · · · + arn−1 ) − (ar + ar2 + ar3 + · · · + arn ) = a − arn

due to many of the terms canceling out. Now, solve for the partial sum to obtain

a(1 − rn )
Sn =
1−r
We observe that this sequence converges provided that |r| < 1 and diverges otherwise. In particular, if |r| < 1
then lim rn = 0 and thus
n→∞
∞
X a(1 − 0) a
ark−1 = lim Sn = =
n→∞ 1−r 1−r
k=1


101
 ∞  
X e k
Example: Determine if converges or diverges. If it converges find it’s sum.
π
k=0

∞  
X 1
Example: Determine if ln converges or diverges. If it converges find it’s sum.
3k
k=1

102
 1 1 1 1
Example: Determine if the series 1 − + − + − · · · converges or diverges. If it converges find it’s sum.
2 4 8 16

Example: Express the number 1.414414414414... as the ratio of two integers.

103
2.2.3 Telescoping Series
Definition
A telescoping series is a series of the form
∞
X
(ak−m − ak−l )
k=1

You can find the sum (or determine convergence/divergence) of these series by investigating and finding a
closed formula for the partial sums.
∞
X 40k
Example: Determine whether the series converges or diverges. If it converges find it’s
(2k − 1)(2k + 1)
k=1
sum.

104
 ∞ r !
X n+1
Example: Determine if the series ln converges or diverges. If it converges find it’s sum.
n
n=1

105
2.2.4 The n-th Term Divergence Test
∞
X
If you are given the fact that a series an converges what can you say about the convergence of the sequence
n=1
in the sum, {an }? Notice that we may single out elements of the sequence using partial sums by the following

Sn − Sn−1 = (a1 + a2 + · · · + an−1 + an ) − (a1 + a2 + · · · + an−1 ) = an

∞
X
Then if we are told an converges, that implies that lim Sn = L for some number L. Thus we obtain
n→∞
n=1
lim Sn = L and lim Sn−1 = L
n→∞ n→∞

and thus

lim an = lim (Sn − Sn−1 ) = L − L = 0

n→∞ n→∞

This gives us a result but nothing that is useful.

(Mostly Useless But Still Should Know This) Theorem

∞
X
If an converges then lim an = 0.
n→∞
n=1

We can reword this theorem to become useful but it requires a little bit of logic.

Definition and Result

Given a statement “If p then q” the contrapositive of this statement is “If NOT q then NOT p”. The
contrapositive of a statement has the same logical implication.

Example: Consider the statement “If it is a crow then it is black” (where we are talking about the common
crow before any biology people start talking about blue jays). So if you see a crow you can say it is black. This
also implies that the following statement is also true, “If it is not black then it is not a crow”.

With this we can form the useful theorem.

n-th Term Divergence Test

∞
X
If it is NOT TRUE that lim an = 0 then it is NOT TRUE that an converges. That is, if lim an 6= 0
n→∞ n→∞
n=1
∞
X
then an diverges. The statement lim an 6= 0 is still satisfied if an diverges.
n→∞
n=1

106
 ∞
X  nπ 
Example: Consider the series cos .
2
n=1

∞ 
X π −n
Example: Consider the series 1− .
n
n=1

107
 X 4k 2 − k 4 1
Example: A friend has determined that for the series that it diverges because lim ak = − =6 0.
10 + 2k 4 k→∞ 2
k=1
However, they wrote down
∞
X 4k 2 − k 4 1
=−
10 + 2k 4 2
k=1

Explain why the student is incorrect.

∞
X 1 1
Example: Consider the series . We see that lim = 0. Is there anything you can say about the value
n n→∞ n
n=1
convergence or divergence of the series? Can you conclude anything about the value of the series?

Definition
∞
X 1
The series is called the Harmonic Series.
n
n=1

108
2.2.5 Properties of Series
Theorem
X X
Let an and bn be two convergent series and let k be a constant. Then...
X X X
1. (an + bn ) = an + bn
X X X
2. (an − bn ) = an − bn
X X
3. kan = k an

and consequently all the above expressions are convergent series.

There are also corollaries that occur as a consequence of the above theorem.

Theorem

1. Every non-zero constant multiple of a divergent series is also divergent.

X X X X
2. If an converges and bn diverges, then both (an + bn ) and (an − bn ) diverge.

∞  k !
X 2 k
Example: Explain why the series − diverges.
3 k+1
k=1

109
2.3 (Section 10.3) Integral Test
2.3.1 Constructing the Integral Test and Examples
Formulating the Integral Test

Suppose that f is a continuous, positive and decreasing function such that f (n) = an on the interval [1, ∞).
Then...
Z ∞ ∞
X Z ∞
f (x)dx < an < a1 + f (x)dx
1 n=1 1

∞
X
To demonstrate this result, we consider rewriting the series as an = a1 + a2 + a3 + · · · as
n=1
∞
X
an = a1 · 1 + a2 · 1 + a3 · 1 + · · ·
n=1

which we will interpret each term ai · 1 to geometrically represent the area of a rectangle of height H = ai and
width W = 1. Now, we may fit these rectangles to the curve of y = f (x) as follows to obtain

which holds graphically since f (1) = a1 , f (2) = a2 , ... etc. Based on this, we see that the sum of the rectangles
forms an over approximation and we obtain the lower bound of
Z ∞ ∞
X
f (x)dx < an
1 n=1

For the upper bound, we take the above picture and move all the rectangles one unit over to the left. This
still is allowed since by the symmetry of the rectangle we still have f (1) = a1 , f (2) = a2 , ... etc.

110
 However, we are taking the area for n ≥ 1, and thus we obtain
∞
X Z ∞
an < f (x)dx
n=2 1

to which we add a1 to both sides of the inequality to obtain

∞
X Z ∞
an < a1 + f (x)dx
n=1 1

as desired. From this, we can determine the behaviour of certain series by a representing integral and vice
versa (essentially a comparison test).

The Integral Test

Let f (x) be continuous, positive, and decreasing on [k, ∞) and that f (n) = an then
Z ∞ ∞
X
f (x)dx and an
k n=k

either both converge or both diverge.

Note
Z ∞ ∞
dx X 1
Since and both converge or diverge together the nature of convergence for p-integrals is the
1 xp np
n=1
same as the p-series form.

111
Note
A COMMON mistake that students make is that they assume that the integral is equal to the series! THIS
IS NOT THE CASE! They just have the same convergent behaviour! That is,
Z ∞ ∞
X
f (x)dx 6= f (n)
k n=k

In this course, you should assume that unless you have a Geometric Series, Telescoping Series, or (a yet
to be mentioned) Taylor Series then all hope of finding an exact value for a series is lost. Very advanced
techniques are required.
∞
X 1
Example: Determine the nature of convergence for .
n2 +1
n=1

∞
X
Example: Determine the nature of convergence for ne−n .
n=2

112
 ∞
X 1
Example: Determine the nature of convergence for .
n ln(n)
n=2

Note
We will be discussing other series tests so that we have a wide range of tools at our disposal for determining
the behaviour of series. A BIG indication that you might want to start with an integral test is if a logarithm
is present.
∞  k Z ∞ x
X 2 2
At Home Exercise: Show that 6= dx.
3 1 3
k=1

113
2.3.2 Error Estimation
We only know how to find the sum of geometric and telescoping series, but we can approximate others by partial
sums
N
X ∞
X
an ≈ an
n=k n=k
∞
X N
X
for very very large N . Let S = an and SN = an . We can form the error in the estimation (summing
n=k n=k
up to the term n = N ),

RN = S − SN
Suppose that S converges under the integral test. If f (n) = an under the required conditions earlier then one
can show the error is bounded by
Z ∞ Z ∞
f (x)dx ≤ RN ≤ f (x)dx
N +1 N
by rearranging the inequality formed at the beginning of this section.
∞
X 1
Example: Estimate by bounding it with N = 10 terms.
n2
n=1

114
2.4 (Section 10.4) Comparison Tests
2.4.1 The Direct Comparison Test for Series
Direct Comparison Test
X X
Let an and bn be series with 0 ≤ an ≤ bn for all n. Then
X X
• If bn converges then an converges
X X
• If an diverges then bn diverges

The logic behind this is identical to that for improper integrals.

∞
X 1
Example: Determine whether or not the series converges or diverges.
n3 +n
n=1

∞
X 2n + 3 n
Example: Determine whether or not the series converges or diverges.
3n + 4 n
n=1

115
2.4.2 The Limit Comparison Test for Series
Limit Comparison Test
an
Suppose an > 0 and bn > 0 and let lim = L. Then provided
n→∞ bn
X X
• 0 < L < ∞, then an and bn both converge or diverge
X X
• L = 0, then if bn converges then an converges
X X
• L = ∞, then if bn diverges then an diverges

Think about these results intuitively!

∞ r
X n5 + 3n
Example: Determine whether or not the series converges or diverges.
2n2 + 4
n=3

116
 ∞
X 1 + n ln(n)
Example: Determine whether or not the series converges or diverges.
n2 + 5
n=2

117
 ∞
X ln(n)
Example: Determine whether or not the series √ n converges or diverges.
n=1
ne

118
2.5 (Section 10.5) Absolute Convergence, Ratio and Root Tests
2.5.1 The Absolute Convergence Test
Definition
An alternating series is a series of the form
∞
X
(−1)n an
n=k

where an ≥ 0 for all n ≥ k.

∞
X 4n+1
Example: Determine if the alternating series (−1)n+1 converges or diverges.
5n
n=0

We can handle them for geometric series. Other tests will be required for other series.

Absolute Convergence Test

X X
If |an | converges then an converges.

Proof: Notice that |an | is either an or −an by definition of the absolute value (depending on the sign of an ).
Thus we can say

0 ≤ an + |an | ≤ |an | + |an | = 2|an |

X X
Since we are assuming |an | is convergence then 2|an | is also convergent. As an + |an | and 2|an | are
X X
non-negative them by a comparison test (an + |an |) is convergent as 2|an | is. Then we may write
X X X
an = (an + |an |) − |an |
X
and as an is the difference of two convergent series, it is also convergent. 

119
Note
This is currently our only test mentioned that handles series with negative terms in determining convergence.
The divergence test handles series with negative terms but only determines divergence. Be careful on how
to interpret this theorem. There is NO COMPARISON happening here. The comparison tests previously
mentioned only apply to series Xwhose terms are non-negative. Comparison does not apply to series that
have negative terms. Also, if |an | diverges there is nothing you can conclude!

∞
X (−1)n+2
Example: Determine the nature of convergence of .
n2
n=1

∞
X sin(n)
Example: Determine the nature of convergence of .
n3
n=1

120
 ∞
3 n
X  
n
Example: Determine the nature of convergence of (−1) 1 + .
n
n=1

Definition
X X
If |an | is convergent we say an is absolutely convergent.

Thus all but the last example mentioned are series that are absolutely convergent. Absolute convergence is
a level of convergence that is VERY strong. There are many things you can do with series that is absolutely
convergent when manipulating them... while for series that are convergent (but not absolutely convergent) you
are very restricted.

121
2.5.2 The Ratio Test
This is everybody’s favourite test.

The Ratio Test

X
Let an be any series and suppose
 
 an+1 
lim  =L
n→∞  an 

Then if...

• L < 1 the series converges absolutely.

• L > 1 the series diverges.

• L = 1 then the test is inconclusive. You must apply a different test as this one does not work.

Note
This test is useful for FACTORIALS (especially!!!), polynomials, and simple exponents.
∞
X 3n (n + 1)
Example: Determine the nature of convergence of .
n!
n=0

122
 ∞
X n!n!
Example: Determine the nature of convergence of .
(2n)!
n=1

123
2.5.3 The Root Test
The Root Test
X
Let an be any series and suppose

lim |an |1/n = L

n→∞

then if...

• L < 1 then the series converges absolutely.

• L > 1 then the series diverges.

• L = 1 then the test is inconclusive. You must use a different test as this one does not work.

Note
This is useful for bad exponents
∞  
X
n 1
Example: Determine the nature of convergence of sin √ .
n=1
n

124
 ∞ 2
X 2n
Example: Determine the nature of convergence of .
n2
n=3

∞
X n!
Example: Determine the nature of convergence of .
(−n)n
n=1

125
2.6 (Section 10.6) Alternating Series Test
2.6.1 Conditional Convergence and the Alternating Series Test
Definition
X X X
If a series bn converges but |bn | diverges then series bn is called conditionally convergent.

Note
Conditionally convergent series DO converge, but they have a “lower level” of convergence. The level of
convergence is rather slow and weak. When working with conditionally convergent series you are rather
limited to what you can do with them algebraically.

The Alternating Series Test

X
The series (−1)n+1 an converges if all the following are satisfied:

• All an ≥ 0.

• All terms in {an } are eventually all non-increasing.

• lim an = 0.
n→∞

Proof: Let’s examine even partial sums. Note that as we assume an+1 ≤ an (non-increasing) then an − an+1 ≥
0. We have

S2 = a1 − a2

S4 = a1 − a2 + a3 − a4 = S2 + a3 − a4 ≥ S2 + 0 = S2

S6 = S4 + a5 − a6 ≥ S4 + 0 = S4
So {S2n } is increasing. Also S2n ≤ a1 since

S2n = a1 − a2 + a3 − a4 + · · · a2n−1 − a2n = a1 − (a2 − a3 ) − (a4 − a5 ) − · · · − (a2n−2 − a2n−1 ) − a2n ≤ a1

as an − an+1 ≥ 0. Thus by monotonic convergence {S2n } converges (i.e. S2n → L). If N is odd then for SN
we have lim S2m+1 = lim (S2m + a2m+1 ) = L + 0 = L. 
m→∞ m→∞

Note
X
To show a series is absolutely convergent you need to show |an | converges and THAT’S IT. To show a
X
series is conditionally convergent you need to show an converges under the alternating series test AND
X
show that |an | diverges by some other test.

126
 ∞
X (−1)n+1
Example: Determine the nature of convergence of the alternating Harmonic series . Is it condi-
n
n=1
tionally or absolutely convergent?

∞
X (−1)n n2
Example: Determine the nature of convergence of .
n2 + 5
n=1

127
 ∞ √
X (−1)n−3 n
Example: Determine the nature of convergence of . Is it conditionally or absolutely conver-
n+4
n=0
gent?

128
2.6.2 Error Estimation of Series that Converge by the AST
X
The sum (−1)n+1 an = S always lies between two successive partial sums SN and SN +1 . Furthermore the error
(remainder) is bounded by

RN = |S − SN | ≤ an+1
∞
X (−1)n
Example: Find the error in approximating using 15 terms.
n2
n=1

129
2.7 (Section 10.7) Power Series
2.7.1 Power Series Functions and Their Domain
Sometimes we can define a function by a series. When the terms in the sum are simple polynomials we call it a
“power series”.

Definition
A power series about x = a (called centered at a) is a function of the form
∞
X
f (x) = cn (x − a)n = ck (x − a)k + ck+1 (x − a)k+1 + · · ·
n=k

where all cn are constant.

1
Example: Find a power series form of f (x) = and find the domain of the series form.
1−x

N
X
Think of it like this, the partial sums SN (x) = xn−1 = 1+x+x2 +· · ·+xN are polynomials that approximate
n=1
1
f (x) = . The approximation gets better as N → ∞.
1−x

130
This is an example I like as it illustrates that a region of convergence describes the domain of such a function.

Example: For what values of x does

∞
X xn
f (x) = (−1)n+1
n
n=1

converge? (i.e. what is the domain of f (x)?)

131
 So the ratio test (or less commonly the root test) gives a fence between convergence and divergence, then you
check to see if the fence itself is dangerous (divergent) or not (convergent).

Theorem
A series will either converge absolutely at a point x = a, in an interval |x − a| < R or everywhere. The
region of convergence is called the interval of convergence (check boundary). The number R is called
the radius of convergence.
∞
X xn
Example: Determine the interval and radius of convergence of f (x) = .
n!
n=0

132
 ∞ 
1 n
X 
Example: Determine the interval and radius of convergence of f (x) = 1+ (x − 2)n and be sure to
n
n=1
check the endpoints.

133
2.7.2 Operations of Power Series
Addition and Scalar Multiplication
X X
If an xn and bn xn converge for |x| < R then
∞ ∞ ∞
! !
X X X
n n
an x + bn x = (an + bn )xn
n=0 n=0 n=0

and
∞
X ∞
X
kan xn = k an xn
n=0 n=0

converge for |x| < R where k is a real number.

Multiplication
X X
If an xn and bn xn converge for |x| < R then
∞
! ∞ ! ∞ n
!
X X X X
n n
an x bn x = ak bn−k xn
n=0 n=0 n=0 k=0

converges for |x| < R.

Substitution
X X
If an xn converges for |x| < R and g(x) is continuous then an (g(x))n converges for all values x such
that |g(x)| < R.

1
Example: Consider f (x) = . Find an expression for this function in power series form and determine
4 + x2
the interval of convergence.

134
 Differentiation
∞
X
If f (x) = cn (x − a)n converges for |x − a| < R then
n=0
∞
X
0
f (x) = ncn (x − a)n−1
n=1
∞
X
f 00 (x) = n(n − 1)cn (x − a)n−2
n=2

..
.

∞
X
f (k) (x) = n(n − 1) · · · (n − (k + 1))cn (x − a)n−k
n=k

all converge over |x − a| < R.

∞
1 1 X
Example: Consider the power series representation of the function f (x) = given by = xn
1−x 1−x
n=0
which converges over |x| < 1. Use this to construct power series representations of its derivatives.

135
 Integration
∞
X
If f (x) = cn (x − a)n converges over |x − a| < R then
n=0
∞
!
(x − a)n+1
Z X
f (x)dx = cn +C
n+1
n=0

converges over |x − a| < R.

∞
1 X
Example: Consider = (−1)n xn which converges over |x| < 1. Use this to construct a power series
1+x
n=0
representation of f (x) = ln(1 + x) and determine the interval of convergence.

136
2.8 (Section 10.8) Taylor and MacLaurin Series
A Taylor series (and Maclaurin) is a polynomial series that (potentially) approximates a known function f (x).
∞
1 X
Let’s assume we are GIVEN that a known function f (x) has a power series representation (e.g. = xn
1−x
n=0
for |x| < 1) and consider it’s representation
∞
X
f (x) = an (x − a)n = a0 + a1 (x − a) + a2 (x − a)2 + a3 (x − a)3 + · · ·
n=0

Note
We are ASSUMING a KNOWN function f (x) has a power series form. This is a really massive assumption!!

Then we shall attempt to solve for all terms in the sequence {an }. We see that

f (a) = a0 + a1 (0) + a2 (0)2 + · · · = a0

f 0 (x) = a1 + 2a2 (x − a) + 3a3 (x − a)2 + · · · ⇒ f 0 (a) = a1 + 2a2 (0) + 3a3 (0)2 + · · · = a1

f 00 (x) = 2a2 + 3 · 2a3 (x − a) + · · · → f 00 (a) = 2a2 + 3 · 2a3 (0) + · · · = 2a2

f 000 (x) = 3 · 2a3 + · · · ⇒ f 000 (a) = 3 · 2a3 + 4 · 3 · 2(0) + · · · = 3 · 2a3

..
.

f (k) (x) = k · (k − 1) · · · (2)(1)ak + (k + 1)(k)(k − 1) · · · (2)ak+1 (x − a) + · · · ⇒ f (k) (a) = k · (k − 1) · · · (2)(1)ak = k!ak

f (k) (a)
and thus ak = . Thus if we are TOLD a specific function has a power series representation and we want
k!
to compute it, we can use this to determine the representation of it in power series form without clever algebraic
tricks or relating it to a geometric series.

Definition
Let f be a function with derivatives of all orders about an interval containing x = a. Then the Taylor
series about x = a of f (x) is the series
∞
X f (n) (a)
P (x) = (x − a)n
n!
n=0

If a = 0 it is called a Maclaurin series (just to give him some credit)

137
2.8.1 The Unreasonable Assumption of Being a Taylor Series
There is still one question, for each value of x within an appropriate interval is it true that P (x) = f (x)? It
required a massive assumption, that the function f (x) COULD BE expressed as a power series. Knowing whether
or not you can do this requires tools we don’t have. You can always compute the Taylor series of any function but
you might not get equality of P (x) and f (x). Let’s have a brief chat about how functions are classified in calculus.

Definition

Let I be an open interval and denote C 0 (I) as the collection of functions that are continuous on I (called
class zero). We also further denote C k (I) as the collection of functions whose derivatives all the way up
to order k are continuous on I (called class k) and denote C ∞ (I) as the collection of function who have
derivatives of ALL orders and are continuous on I (called smooth functions). Lastly, we also represent
functions who have a power series representation on I as A(I) (called analytic functions).

Note
To illustrate this, f (x) = |x| is a class zero function on R (the collection of all real numbers) as it is
∞
1 X
continuous but it’s derivative is not continuous. The function f (x) = = xn is an analytic function
1−x
n=0
on (−1, 1) as it has a power series representation on this interval.

Here’s how the structuring in calculus works...

C 0 (I) ⊃ C 1 (I) ⊃ C 2 (I) ⊃ · · · C ∞ (I) ⊃ A(I)

where A ⊃ B means the collection of all objects in B can be found in A. By this chain, analytic is the UL-
TIMATE form of differentiability. You can have functions with derivatives of ALL orders and are continuous but
still don’t have a power series representation! Making your way down the chain is harder at each step. Functions
further down the chain are much smoother and nicer than functions near the start of the chain.

Example Show the function

(
x2 x>0
f (x) =
−x2 x≤0

is strictly a class one function on R (i.e. lies in the collection C 1 (R) but not C 2 (R)).

138
2.8.2 Examples of Constructing a Taylor Series Directly
Example: Compute the Taylor series of f (x) = sin(x) about x = 0.

139
Example: Find the Taylor series of f (x) = ex about x = 0.

140
2.8.3 Taylor Polynomials of Order N
Definition
The Taylor polynomial of order N to f (x) at x = a is
N
X f (n) (a) f 00 (a) f (N ) (a)
PN (x) = (x − a)n = f (a) + f 0 (a)(x − a) + (x − a)2 + · · · + (x − a)N
n! 2! N!
n=0

Definition
If P (x) is the Taylor polynomial of f (x) we call f (x) the generating function of P (x).

Two questions remain in this theory of wanting to replace f (x) with P (x) (because polynomials are nicer to
work with):

• When can we expect f (x) = P (x) for each x? That is, when can we expect that lim PN (x) = f (x)?
N →∞

• If the previous condition is satisfied (yay!) how accurate are the finite order approximations PN (x)? (Because
we might want to understand the function using a computer and computers don’t do infinity so using some
smaller terms to approximate values might be good enough).

Example: We can’t always expect that P (x) = f (x). With great difficulty one can show that for
(
0 x=0
f (x) = − 12
e x x 6= 0

That derivatives of all orders exist and are continuous with the further fact that f (n) (0) = 0. Thus for a Taylor
series centered at x = 0 we have P (x) = 0. However, clearly f (x) is not identically zero! Thus they are not equal.
This function is an example of a function that is C ∞ (R) but not A(R)!

141
2.9 (Section 10.9) Convergence of Taylor Series
2.9.1 Taylor’s Theorem and Approximating Functions
Alright, it’s time to answer the big question. When does the Taylor series P (x) equal the generating function
f (x)? When does PN (x) → f (x) as N → ∞?

Do the polynomials approach f (x) as N → ∞? By the mean value theorem we have the following theorem to
help us out.

Taylor’s Theorem

If f is of class C N on an interval containing a and b then there exists a number c between a and b such that
∞
!
f (N +1) (c) X f (n) (a) f (N +1) (c)
f (b) = PN (b) + (b − a)N +1 = (b − a)n + (b − a)N +1
(N + 1)! n! (N + 1)!
n=0

This theorem extends (in a complicated way) so that in an interval I where f is of class C ∞ then for all x in
I containing a,

f (x) = PN (x) + RN (x)

f (N +1) (x)
for all N where RN (x) = (x − a)N +1 for some c between x and a.
(N + 1)!

142
 Alright, this might seem complicated. We’re working VERY HARD to approximate the function f (x) with
the Taylor polynomials. Let’s break down what’s happening, we have said that if the function is smooth (nice
enough) then we can represent it as the sum of two functions

f (x) = PN (x) + RN (x)

Now here’s the part we have to work on. Since PN (x) → P (x) as N → ∞ then we need to only show that this
other weird function RN (x) satisfies RN (x) → 0 as N → ∞. This will allow us to conclude that

f (x) = PN (x) + RN (x) ⇒ lim f (x) = lim (PN (x) + RN (x)) ⇒ f (x) = P (x) + 0 = P (x)
N →∞ N →∞

Often one can estimate RN (x) without ever knowing c.

Example: Show the function f (x) = ex is equal to it’s Taylor series.

143
2.9.2 Taylor Series with Bounded Derivatives
Theorem

If there is a constant M such that |f (N +1) (t)| ≤ M for all t between x and a then RN (x) satisfies

|x − a|N +1
|RN (x)| ≤ M
(N + 1)!
If this holds for every N and the Taylor conditions are satisfied then P (x) = f (x).

Example: Show that f (x) = sin(x) is equal to it’s Taylor series.

144
2.9.3 Important Analytic Functions
The following is a short list of some analytic functions and their Taylor series form.

Important Taylor Series Expression

∞
X xn
• ex = in I = R
n!
n=0
∞
X (−1)n x2n
• cos(x) = in I = R
(2n)!
n=0
∞
X (−1)n x2n+1
• sin(x) = in I = R
(2n + 1)!
n=0
∞
1 X
• = xn in I = (−1, 1)
1−x
n=0
∞
X (−1)n xn+1
• ln(1 + x) = in I = (−1, 1)
n+1
n=0

Note
Ahem, for future calculus courses (and this one) you should have these MEMORIZED. No if’s and’s or
but’s. Many expressions for other functions are obtained using these.

Example: Use the above to indirectly construct a Taylor Series for the integral function
Z x
2 2
F (x) = √ e−t dt
π 0
.

Note
The above integral function is the famous “Error Function”.

145
2.9.4 Approximations Using Taylor Series
Example: Find P4 (x) of f (x) = ex cos(x).

146
In an alternating series you can see how well PN (x) approximates f (x).

x3
Example: For what values of x will P3 (x) = x − approximate sin(x) with an error no bigger than 3 × 10−4 ?
3!

147
2.10 (Section 10.10) Binomial Series and Applications of Taylor Series
In this section we discuss a bit of practicality of using and developing Taylor series. Before we do so, we discuss
Binomial series.

2.10.1 Binomial Series

These are the series representation of (1 + x)m . Before you get all “just expand it and it becomes a polynomial.
Polynomials are their own Taylor series” we consider the case where m is not a positive integer as well.

We defin the binomial coefficient

 
m m(m − 1)(m − 2) · · · (m − k + 1)
=
k k!
for k ≥ 1 where k is an integer and we set
   
m m m(m − 1)
=m =
1 2 2!
We compute for f (x) = (1 + x)m at x = 0,

f (x) = (1 + x)m ⇒ f (0) = 1

f 0 (x) = m(1 + x)m−1 ⇒ f 0 (0) = m

f 00 (x) = m(m − 1)(1 + x)m−2 ⇒ f 00 (0) = m(m − 1)

and so forth. Hence
f (k) (0) = m(m − 1)(m − 2) · · · (m − k + 1)
and thus the series is
∞ ∞  
X m(m − 1)(m − 2) · · · (m − k + 1) k
X m
P (x) = 1 + x =1+ xk
k! k
k=1 k=1

One can show that by the ratio test and inspecting the endpoints it converges (absolutely) for |x| < 1. Showing
P (x) = (1 + x)m is a little tricky. The book shows a trick and guides you in one of the exercises of this section.

148
 √
Example: Form a series representation of 5.

149
2.10.2 Applications of Taylor Series
One should not underestimate the “power” of power series. Their complexity is worth the effort. They are the
foundation of much theory.
∞
X xn x2 x3 x4
Complex Variables: We know that ex = =1+x+ + + + ···
n! 2! 3! 4!
n=0
Let x = iθ then
∞
iθ
X (iθ)n θ2 θ3 θ4 θ5
e = = 1 + iθ − −i + + i + ···
n! 2! 3! 4! 5!
n=0

since i0 = 1, i1 = i, i2 = −1, i3 = −i, i4 = 1, ...and the pattern repeats. Thus

θ2 θ4 θ6 θ3 θ5 θ7
   
iθ
e = 1− + − + ··· + i θ − + − + · · · = cos(θ) + i sin(θ)
2! 4! 6! 3! 5! 7!
Ta-da!

Probability Theory: One of (if not the most) important distribution is the bell curve!

Many natural occurrences follow a bell curve distribution. Ideally, in a class grades you want them to follow a
bell (normal) distribution. The probability of something occuring is given by the area between two points under
the curve. In quantum physics, for example, the probability of a particle existing in a location can follow a bell
curve.

150
 Example: In the ground state of the harmonic oscillator, there is a non-zero probability of finding a particle
outside the classically allowed region. The probability of finding the particle outside of the region [−a, a] is given
by

mω a
r Z  mω 
F (a) = 1 − exp − x2 dx
π~ −a ~
r
~
Determine the probability of finding the particle outside the region where a = .
mω

151
 Approximating Values: One can show using
∞ ∞
1 1 X X
2
= 2
= (−x2 )n = (−1)n x2n
1+x 1 − (−x )
n=0 n=0

is convergent for |x| < 1. Upon integrating and solving to obtain C = 0,

∞
X (−1)n x2n+1
arctan(x) =
2n + 1
n=0

holds for −1 < x ≤ 1 (after checking the endpoints). Thus we may compute
∞
X (−1)n
π
= arctan(1) =
4 2n + 1
n=0
∞
X (−1)n 4 4 4 4
⇒π=4 = 4 − + − + − ···
2n + 1 3 5 7 9
n=0

however it does converge very slowly. You’ll need several terms to for a good approximation (the reason it’s
slow is due to conditional convergence).

Evaluating Limits: One can use Taylor series to compute various limits.

ex − (1 + x)
Example: Compute lim .
x→0 x2

152
Chapter 3

Coordinate Systems

153
3.1 (Appendix A7) Complex Numbers
3.1.1 Introducing Complex Numbers and Algebraic Operations
Everything we’ve dealt with is over the collection of measurable quantities, the real numbers R. The complex
numbers C are an extension of the reals R such that we allow solutions to the equation

x2 + 1 = 0
√
The “positive” solution x = + −1 = i gives rise to “imaginary” parts of numbers. The space C consists of all
numbers of the form

a + bi
where a and b are real.

Example: The equation x2 + 4x + 5 = 0 has solutions

Addition and multiplication are defined as expected.

Example: Compute (3 + 4i) − (−1 + 2i) and (2 + i)(1 + 3i).

Definition
The conjugate of z = a + bi is denoted and defined as z̄ = a − bi.

Definition
The norm/modulus of z = a + bi is defined as
√ p
|z| = z · z̄ = a2 + b2

154
Definition
We define division for z = a + bi, w = c + di as
z z w̄ z w̄
= =
w ww̄ |w|2
Essentially you compute it by rationalizing the denominator.

2 + 3i
Example: Compute .
1 − 7i

155
3.1.2 Argand Diagrams and Polar Coordinates
There are two components to a complex number we need to measure instead of just the one component that real
numbers have. Thus real numbers are represented by a single quantity while complex numbers are represented as
a pair of numbers (thus a point). We call this representation an Argand diagram.

We can also represent them by “polar coordinates”. The point makes an angle with the x-axis and a radius
out from the origin.

From the diagram a = r cos(θ) and b = r sin(θ) where r = |z|.

Definition
Consider the complex number z = a + ib and let θ be the angle between the segment connecting point
representation of z with the origin and the positive x-axis. Provided that r represents the length of this
segment, the Polar Representation of the point is

z = r(cos(θ) + i sin(θ))

156
Example: Express 1 − i in polar coordinates.

157
3.1.3 Euler’s Formula
Euler’s Formula

eiθ = cos(θ) + i sin(θ)

This allows us to write the polar form as

a + bi = r(cos(θ) + i sin(θ)) = reiθ

This form allows us to exploit exponent properties.

Products: Taking the product gives (r1 eiθ1 )(r2 eiθ2 ) = r1 r2 ei(θ1 +θ2 )

r1 eiθ1 r1
Quotients: Taking the quotient gives = ei(θ1 −θ2 )
r2 eiθ2 r2
Powers: Taking the power gives (reiθ )n = rn eiθn .

That last point is the most interesting because it gives us the following theorem.

De Moivre’s Theorem

(cos(θ) + i sin(θ))n = cos(nθ) + i sin(nθ)

Proof: Combine eiθ = cos(θ) + i sin(θ) with (eiθ )n = einθ . 

You can use this to generate identities for trigonometric functions!

Example: Use De Moivre’s Theorem to generate identities for cos(2θ) and sin(2θ).

158
3.1.4 Roots of Complex Numbers
Every polynomial of degree n always has n-roots over C (called the Fundamental Theorem of Algebra). We can
use the exponential form to ‘easily’ solve

zn = C
for z where C is a constant (possibly a complex number C = a + bi).
Procedure for Finding All Roots of a Complex Number

Start with an equation of the form z n = C.

1. Express C in exponential form as C = reiθ . Then

z n = reiθ

2. Use the periodicity of the complex exponential (sine and cosine). Since sine and cosine are periodic
with period 2π then

z n = r exp(i(θ + 2πk))

for k any integer.

3. Take the n-th root of both sides,

  
1/n θ + 2πk
z=r exp i
n

for k = 0, 1, 2, 3, ..., n − 1.

The reason it goes from k being all integers to just the first n values is because afterwards it repeats. So
we can obtain the n roots from the first n values of k starting from zero.

Example: Find the all the fourth roots of −16.

159
3.2 (Section 11.1) Parametric Equations
3.2.1 Defining Parametric Curves
A curve C given by f (x, y) = 0 gives us a graph but it falls short on a few aspects.
Definition
If C is a curve and (x, y) is any point on C then provided there exists functions such that
(
x = f (t)
y = g(t)
we call it a parametric representation of C.

Example: Consider y 2 = x. Then y = t and x = t2 is a parametric representation of this curve. All points
can be mapped out by choosing values of t.

Why is this useful? For modeling the position of a particle by time.

Example: Suppose that the position of a particle follows the trajectory of the curve y 2 = x2 + 1 given by the
equations y = sec(πt) and x = tan(πt) where −π/2 < t < π/2. Determine the position of the particle after 1.25
seconds. Graph and Determine the orientation of the particle flow along the curve.

160
3.2.2 Graphing a Parametric Curve
Technique #1: Plotting Points

This is the worst way and I don’t really condone it.

Example: Attempt to graph x = 2 cos(t), y = sin(t) for 0 ≤ t ≤ 2π by plotting a few points.

If you do it this way, you might as well use GeoGebra.

Graphing Using GeoGebra

Using GeoGebra Classic 6, you can graph parametric curves given by x = f (t) and y = g(t) from t = a to
t = b using the command

Curve(f(t),g(t),t,a,b)

161
Technique #2: Finding the Curve C

This is usually done by solving one equation for t and plugging it in the other.

Example: Consider the curve given by x = t2 + t, y = 2t − 1.

162
3.2.3 Standard Representation of a Circle
Consider the circle x2 + y 2 = R2 .

x y
We note cos(t) = , sin(t) = . So a parametric representation is
R R
(
x = R cos(t)
y = R sin(t)
for 0 ≤ t < 2π. The orientation is counter-clockwise as that represents the direction of the increasing angle
(which is variable the circle is parametrized by).

3.2.4 Transformations
Reverse Orientation Translation by a Point

163
3.3 (Section 11.2) Calculus of a Parameterized Curve
3.3.1 The Slope of a Parametric Curve
Let C be a curve given parametrically by x = f (t), y = g(t). We still want to discuss the slope of the tangent line
dy
.
dx
By the chain rule

dy dy dx dy dy/dt
= ⇒ =
dt dx dt dx dx/dt

Notation
People commonly use the following,

dy dy
y0 = ẏ =
dx dt
to denote differentiation with respect to space and time respectively. Thus the above may be stated as
ẏ
y0 =
ẋ

Theorem
ẏ ẋ
Provided ẋ 6= 0 we have y 0 = . Provided ẏ 6= 0 we have x0 = .
ẋ ẏ

2
Example: (Witch of Agnesi) Consider the curve given by x = 2t, y = .
1 + t2

164
Example: Find the tangent line to x = t5 − 4t3 , y = t2 at (0, 4).

165
3.3.2 Higher Order Derivatives and Vertical/Horizontal Tangency
Horizontal: ẏ = 0, (provided ẋ 6= 0) Vertical: ẋ = 0, (provided ẏ 6= 0)

We can also talk about concavity. Since

d
d dt [y]
[y] =
dx ẋ
then by the same procedure,
h i
d dy
d2 y
 
d dy dt dx
= =
dx2 dx dx ẋ

Theorem
Let x(t) and y(t) be parametric equations of a curve. Then
d 0
dt [y ]
y 00 =
ẋ

Example: Consider x = 1 − t2 , y = t7 + t5 .

166
3.3.3 Integration
Definition
Z b
The signed area between y(x) and the x-axis over [a, b] is ydx. The signed area between x(y) and the
Z d a

y-axis over [c, d] is xdy.

c

Example: Compute the area trapped between the curve x = cos3 (t), y = sin3 (t) and the x-axis over 0 ≤ t ≤ 2π.

167
(Continued...)

168
Example: Find the area enclosed by the y-axis and the curve x = t − t2 , y = 1 + e−t .

169
3.3.4 Arc-Length
We learned that the length of y(x) over [a, b] is
Z bp
L= 1 + y 02 dx
a
ẏ
Let’s examine the integrand. Since y 0 = then
ẋ
s  2 p
p ẏ ẋ2 + ẏ 2
1 + y 02 dx = 1 + dx = dx
ẋ ẋ
1
However dx = t0 dx = dt. So the length of x(t), y(t) over t = c to t = d is the following.
ẋ
Theorem
The arc length of a parametrized curve x(t), y(t) over c ≤ t ≤ d is
Z dp
L= ẋ2 + ẏ 2 dt
c

Example: Compute the length of a circle of radius R using the standard parametrization.

170
Example: Compute the length of the curve x = ln(sec(t) + tan(t)) − sin(t), y = cos(t) over 0 ≤ t ≤ π/3.

171
3.4 (Section 11.3) Polar Coordinates
3.4.1 Revisiting Polar Coordinates and Non-Uniqueness
We already talked a bit about polar form with complex numbers. The idea is that we measure points in the
xy-plane using an angle θ and radius r.

Consider the point (r, θ) = (2, π/6).

The expression for a point is not necessarily unique! For example, technically (2, π/6) = (2, π/6 + 2π).

In fact, (2, π/6) = (2, π/6 + 2πn) for any integer n. Polar coordinates also allows negative distance (unless
specified otherwise)!

172
 For example consider the angle θ = π/6 + π = 7π/6.

If we take the radius r = −2 then we measure the distance backwards, so (2, π/6) = (−2, 7π/6) = (−2, 7π/6 +
2πn) for any integer n as well.

3.4.2 Polar to Cartesian (Points), Cartesian to Polar (Equations)

Expressing Cartesian Values in Terms of Polar Values

Consider a point (x, y) in Cartesian that makes an angle θ with the positive x axis and is a distance r from
the origin. Then...
(
x = r cos(θ)
y = r sin(θ)

We can use this to convert equations to polar form. This is very useful for equation with the term x2 + y 2 .
Given a specified r and θ this allows you to find the Cartesian equivalent of a point in polar coordinates.

Example: Find the Cartesian representation of the point (r, θ) = (2, π/6).

173
Example: Find the polar curve equation of x4 + y 4 + 2x2 y 2 + 2x3 + 2xy 2 − y 2 = 0.

174
3.4.3 Cartesian to Polar (Points), Polar to Cartesian (Equations)
Expressing Polar Values in Terms of Cartesian Values

Solving the equations x = r cos(θ) and y = r sin(θ) for r and θ gives

( 2
r = x2 + y 2
y
tan(θ) =
x

Where you solve for them explicitly depending on the “Branch” (this means what region the angle in the
Cartesian plane: 0 ≤ θ < 2π, 2π ≤ θ < 4π, etc... and whether you use a positive or negative radius).

Example: Convert (x, y) = (−4, −4) to polar coordinates.

Example: Convert the equation r = −8 cos(θ) to Cartesian.

175
3.4.4 Graphing Sinusoidal Curves
We will be talking about graphing polar curves from their equations in the next section. To do so, we will need
to review how to graph sinusoidal curves.

Review:

r = A sin(Bθ + C) + D or r = A cos(Bθ + C) + D

• The amplitude is |A| (height of the curve is 2|A|).

2π
• The period is (length of one wave).
|B|
• The phase shift is −C/B (horizontal shift).

• The vertical shift is D (vertical translation).

Example: Graph the curve r = 2 sin(4θ − 2) + 3 in the rθ-plane.

176
Example: Graph the curve r = 1 + 2 sin(2θ) in the rθ-plane.

177
3.5 (Section 11.4) Graphing Polar Curves in the Cartesian Plane
3.5.1 Plotting Curves in the xy-Plane Based on Their rθ-Graph
When graphing functions of the form r = f (θ) in the Cartesian plane, the best way is to follow along by graphing
it initially in the rθ-plane and then the xy-plane. This is best seen by example.

Example: Graph r = 1 − cos(θ) in the Cartesian plane.

178
Example: Graph r = 1 + sin(θ) in the Cartesian plane.

179
Example Graph r = cos(2θ) in the Cartesian plane.

180
Example: Graph r = 1 + 2 sin(θ) in the Cartesian plane.

181
Example: Graph r2 = 4 cos(2θ) in the Cartesian plane.

182
3.5.2 Graphing Polar Curves Using GeoGebra
Using Geogebra Classic 6, there is a way to plot polar curves of the form r = f (θ) with ease. Below is the command
that you use to obtain this.
GeoGebra Polar Graphing

To graph the curve r = f (θ) over the region a ≤ θ ≤ b use the command

Curve[(f(t);t),t,a,b]

Note that the semicolon is important!

Example: Use GeoGebra Classic 6 to graph the polar curve r(θ) = sin(θ) + sin3 (5θ/2).

183
3.5.3 Calculus in Polar Coordinates
We can use r = f (θ) to describe a curve parametrically (in terms of a parameter θ) as
(
x = r cos(θ) = f (θ) cos(θ)
y = r sin(θ) = f (θ) sin(θ)
ẏ dy/dθ
From this use y 0 = = to obtain the following formula for the slope.
ẋ dx/dθ
Theorem
Let r = f (θ) be a differentiable curve in polar coordinates. Then the Cartesian slope of the curve at (r, θ)
is given by the following.

f 0 (θ) sin(θ) + f (θ) cos(θ)


dy 
=
dx (r,θ) f 0 (θ) cos(θ) − f (θ) sin(θ)

You can also apply the same logic to finding the arc-length by creating a parametrization but it always simplifies
to the following formula for a curve r = f (θ) and is given by the following.
Theorem
Let r = f (θ) be a differentiable curve in polar coordinates. Then the Arc-Length of the curve over α ≤ θ ≤ β
is given by
s  2
Z β
2
dr
L= r + dθ
α dθ

Example: Compute the slope of the tangent line to r = 1 + 2 cos(θ) at θ = π/4.

184
Example: Find the length of the curve r = 1 − cos(θ) for 0 ≤ θ ≤ 2π.

185
3.6 (Section 11.5) Area Trapped by Polar Curves
3.6.1 Formulating the Wedged Area Trapped by Polar Curves
Consider a curve r = f (θ) for α ≤ θ ≤ β.

To find the area trapped above we take wedges instead of rectangles for simplicity.

β−α
We subdivide the angles into equal widths ∆θn = each with a radius of rk = f (θk ) for some θk in each
n
subregion. The formula for the area of each sector is
1 1
Ak = rk2 ∆θn = f (θk )2 ∆θn
2 2
Then we take n → ∞ and add them up to obtain
Z β
1
A= f (θ)2 dθ
2 α

186
3.6.2 Examples of Computing Area
Example: Compute the area trapped inside of r = 1 + cos(θ).

187
Example: Find the area trapped inside r = 1 and r = 1 − cos(θ).

188
3.6.3 Splits in Integration
When integrating between two objects splits are a common occurence

The behaviour changes at θ = c.

Example: Find the area common to the curves r = 9 cos(θ) and r = 9 sin(θ).

189
.

190