Advanced Calculus: Lecture Notes For MA 440/540 & 441/541 2015/16

ADVANCED CALCULUS
Lecture notes for

MA 440/540 & 441/541
2015/16
Rudi Weikard
logHxL
x
1 2 3 4 5
-1
-2
Based on lecture notes by G. Stolz and G. Weinstein

Version of September 3, 2016
1
Contents
First things first 1

0.1. The goals 1
0.2. The rules 1
0.3. Hints 2
0.4. The language of mathematics 3
Chapter 1. The real numbers 5

1.1. Field axioms 5
1.2. Order axioms 6
1.3. The induction principle 8
1.4. Counting and infinity 10
1.5. The least upper bound axiom 11
Chapter 2. Sequences and series 15

2.1. Sequences 15
2.2. Sums and the Σ-notation 18
2.3. Series 18
Chapter 3. A zoo of functions 23
Chapter 4. Continuity 27
4.1. Limits of functions 27
4.2. Continuous functions 28
4.3. The intermediate value theorem and some of its consequences 29
4.4. Uniform convergence and continuity 29
Chapter 5. Differentiation 33
5.1. Derivatives 33
5.2. The mean value theorem and Taylor’s theorem 35
5.3. Uniform convergence and differentiation 36
Chapter 6. Integration 39
6.1. Existence and uniqueness of integrals 39
6.2. Properties of integrals 41
6.3. The fundamental theorem of calculus 42
6.4. Integration of piecewise continuous functions 42
6.5. Uniform convergence and integration 43
Chapter 7. Special topics 45

7.1. Generalized limits 45
7.2. Trigonometric functions and their inverses 46
i
ii CONTENTS
7.3. Analytic geometry 48

7.4. Sets of measure zero and some consequences 49
Appendix A. Some set theory and logic 51
A.1. Elements of logic 51
A.2. Basics of set theory 52
A.3. Functions 53
A.4. The recursion theorem 54
Index 55
First things first
0.1. The goals

Our goal in this class is threefold:
(1) to obtain a body of knowledge in Advanced Calculus, the basis of the analysis of
real-valued functions of one real variable;
(2) to learn how to communicate ideas and facts in both a written and an oral form;
(3) and, perhaps most importantly, to become acquainted with — indeed, to master
— the process of creating mathematics.
In conducting this class we shall try to model a mathematical community in which
both collaboration and competition are prevalent. This community is — no, you are —
on the verge of discovering the foundations for a number of rules and recipes which have
been successfully in use for some time. In the process you will recreate a body of knowledge
almost as if you were the first to discover it. However, as we have only nine months to do
this rather than a century or two, there will be some help available to you, most prominently
in the form of these notes which will delineate broadly a path in which discovery will (or
could) proceed.
In this course it is allowed and, in fact, required to criticize the person on the board for
flaws or incomplete arguments (you are a scientific community). Criticism has to be leveled
in a professional manner, in particular, it has to be free from any personal insults. At the
same time you have to learn to accept criticism without taking it personally. By learning
to stand up for your ideas (or to accept that you made a mistake) you may get something
out of this course which is of value not only in mathematics.
0.2. The rules

The following rules, based on intellectual and academic honesty, will be in force.
(1) Everybody will have the opportunity to present proofs of theorems. You will have
the proof written out on paper and present it with the help of a document camera.
(2) The audience (including the instructor) may challenge a statement made in the
course of the proof at any point.
(3) If the presenter is able to defend the challenged statement, he or she proceeds; if
not, the presenter must sit down earning no points for this problem and losing the
right to present again that day. The challenger may present his or her solution or
elect to receive a challenge reward (see rules (10) and (12)).
(4) A proof of a theorem will be considered correct if no one has objections (or further
objections). Its written version will then be “published” by uploading it to Canvas
(it should have a title and the list of authors). The presenter and, if applicable,
his or her collaborators (see rule (9)) will earn a total of 10 points at this time.
1
2 FIRST THINGS FIRST
(5) During class the instructor has the final decision on determining whether an ar-
gument may stand or not. His verdict may still be challenged after a proof is
“published” (see rule (6)).
(6) If someone other than an author discovers a flaw in a “published” proof, he or she
will get the opportunity to explain the mistake and present a correct proof for a
total of 20 points.
(7) While presenting proofs you may only refer to those axioms and theorems in the
notes which occur before the one you are working on, to published proofs of such
theorems, to the definitions, and to the appendix.
(8) You must give credit where credit is due, i.e., during your presentation you must
declare the points at which you had help and by whom.
(9) It is also possible to report joint work. In such a case 4 points will be earned for
the presentation while the other 6 are evenly distributed among the collaborators.
(10) The successor of a presenter will be chosen as the student with the smallest number
of points among the volunteers taking into account the modification by rules (3),
(11) and (12). A die is rolled, if necessary.
(11) You may volunteer for a particular problem by an e-mail to me. This (in the order
received) establishes priority among volunteers with the same number of points.
(12) For a student who has earned a challenge reward 20 points will be subtracted from
his or her current score for the purpose of determining a successor. At the time
such a student is selected to prove a theorem the challenge reward expires.
(13) Class attendance and participation is required. Absences from class are recorded
in Canvas. After 10 absences from class 10 points will be subtracted from your
class score and the count of absences is set again to zero.
(14) There will be no partial credit except as described above to share credit.
0.3. Hints
• As there is no partial credit, be well prepared to answer whatever questions may
arise.
• There is no need to take notes, final proofs will be “published”. Think along
instead.
• Try to earn points early.
• The index at the end of these notes may lead you to some of the necessary defini-
tions (let me know of any omissions).
• Often a proof will be fairly simple once one realizes that a previous result or a
previous method of proof can be used. Thus constant participation in class, even
when it’s other people’s turn to present, is highly advised! You should go over the
proofs again at home and see if they are still clear to you, possibly rewriting them
with more details.
• Mathematical reasoning takes time. You may expect some frustration — without
it there would never be a sense of accomplishment. Plan to spend a lot of time
thinking about a problem before writing down a final solution.
• Exercises are for private study. Points are only earned for proofs of theorems.
• More difficult problems are marked with a ∗.
• See me if you need help!
0.4. THE LANGUAGE OF MATHEMATICS 3
0.4. The language of mathematics

Mathematics is a highly formalized subject. In many ways learning it is similar to learn-
ing a foreign language. This language rests on two (tightly interwoven) pillars: logic and set
theory. In this course we assume that you have a basic familiarity (possibly unconsciously)
with both. Nevertheless Appendix A collects some fundamental material about set theory
which you may (and will have to) use in your proofs. If you like you may also find some
information in my algebra notes which are on my website1. There, at the ends of Sections
1.2 and 1.3, are references to books on these subjects.
Statements (true or false) are the bread and butter of mathematics. The more important
kind of statements are the following:
• Axioms: In a given mathematical theory some statements are taken for granted.
Such statements are called axioms. The axioms, in fact, characterize the theory.
Changing an axiom means to consider a different theory. Euclidean geometry, for
instance, relies on five axioms. One of them is the axiom of parallels. After trying
for centuries to infer the parallel axiom from the other axioms mathematicians of
the nineteenth century developed non-Euclidean geometries in which the parallel
axiom is replaced by something else.
• Definitions: A mathematical definition specifies the meaning of a word or phrase
leaving no ambiguity. It may be considered an abbreviation. For instance, the
statement “A prime number is a natural number larger than one such that if it
divides a product of two natural numbers it divides one of the factors.” defines
the word prime number.
• Theorems: A theorem is a true statement of a mathematical theory requiring
proof. It is usually of the form “p implies q”. For example the theorem “If n is
even, then n2 is divisible by 4.” is of this form. Sometimes, when a statement
hinges only on the axioms, the theorem could simply be something like “2 is a
prime number.”.
Many authors also use the words proposition, lemma, and corollary. Logically these are
theorems and we will not use these words (but you should still look them up in a dictionary
if you ever want to talk to other people).
1http://www.math.uab.edu/~rudi/teaching/algebra.pdf
CHAPTER 1
The real numbers
We will introduce the real numbers R by a series of axioms, namely the field axioms, the
order axioms, and the least upper bound axiom. This means we will be sure of the precise
properties of the real numbers. Of course, you probably already have some intuition as to
what real numbers are, and these axioms are not meant to substitute for that intuition.
However, when writing your proofs, you should make sure that all your statements follow
from these axioms or their consequences proved previously. You will be amazed about the
rich world being created — by you — from these axioms.
1.1. Field axioms

Definition 1. A binary operation on a set A is a function from A × A to A. It is
customary to express a binary operation as a ? b (or with other symbols in place of ?). An
element e ∈ A is called an identity if e ? a = a ? e = a for all a ∈ A. An element b ∈ A is
called an inverse of a ∈ A if a ? b = b ? a = e, assuming e is an identity.
Axiom 1 (Field axioms). For each pair x, y ∈ R, there is a unique element denoted
x + y ∈ R, called the sum of x and y, such that the following axioms of addition are
satisfied:
(A1): (x + y) + z = x + (y + z) for all x, y, z ∈ R (associative law of addition).
(A2): x + y = y + x for all x, y ∈ R (commutative law of addition).
(A3): There exists an additive identity 0 ∈ R (in other words: x + 0 = x for all
x ∈ R).
(A4): Each x ∈ R has an additive inverse, i.e., an inverse with respect to addition.
Furthermore, for each pair x, y ∈ R, there is a unique element denoted by x · y ∈ R (or
simply xy), called the product of x and y, such that the following axioms of multiplication
are satisfied:
(M1): (xy)z = x(yz) for all x, y, z ∈ R (associative law of multiplication).
(M2): xy = yx for all x, y ∈ R (commutative law of multiplication).
(M3): There exists a multiplicative identity 1 ∈ R \ {0} (in other words: 1 · x = x
for all x ∈ R).
(M4): Each x ∈ R \ {0} has a multiplicative inverse, i.e., an inverse with respect to
multiplication.
Finally, multiplication and addition satisfy the distributive law:
(D): (x + y)z = xz + yz for all x, y, z ∈ R.
In formulating the distributive law we have made use of the convention to let multipli-
cation take precedence over addition, i.e., x + yz is short for x + (yz).
Any set which satisfies all the above axioms is called a field. Thus R, but also the set
of rational numbers with which you are familiar, are fields.
5
6 1. THE REAL NUMBERS
Exercise 1. Show that one may define binary operations in the set {0, 1} which turn
it into a field. (This field is called Z2 .)
Theorem 1. The additive identity in R is unique.
Exercise 2. Is −0 an additive identity?
Theorem 2. Every real number x has a unique additive inverse.
The unique additive inverse of x ∈ R is called the negative of x and is denoted by −x.
For simplicity we will usually write x − y in place of x + (−y).
Theorem 3. x + y = x + z if and only if y = z, assuming that x, y, z ∈ R.
Theorem 4. Suppose x, y ∈ R. Then the following two statements hold.
(1) If x + y = x then y = 0.
(2) If x + y = 0 then y = −x.
Theorem 5. −(−x) = x for all x ∈ R.
Theorem 6. There is a unique multiplicative identity in R and every non-zero real
number x has a unique multiplicative inverse.
The unique multiplicative inverse of x ∈ R \ {0} is called the reciprocal of x and is
denoted by x−1 . We will also use the notation x1 and 1/x in place of x−1 and we will usually
write xy or x/y for xy −1 .
Theorem 7. If x, y, z ∈ R and x 6= 0, then the following statements hold.
(1) xy = xz if and only if y = z.
(2) If xy = x then y = 1.
(3) If xy = 1 then y = x−1 .
(4) (x−1 )−1 = x.
Theorem 8. For every x ∈ R we have 0x = 0.
Theorem 9. 0 does not have a reciprocal and neither is it the reciprocal of any number.
Theorem 10. If x and y are non-zero real numbers, then xy 6= 0.
Theorem 11. Let x, y ∈ R. Then
(1) (−1)x = −x,
(2) (−x)y = −(xy) = x(−y), and
(3) (−x)(−y) = xy.
Theorem 12. Let a, b ∈ R and x, y ∈ R \ {0}. Then a/x + b/y = (ay + bx)/(xy).
1.2. Order axioms

We will now state the next set of axioms for the real numbers. We use the following
notation: If A is a subset of R then
−A = {x ∈ R : −x ∈ A}.
Exercise 3. Convince yourself that a ∈ −A if and only if −a ∈ A.
Theorem 13. Suppose A is a subset of R. Then −(−A) = A.
Axiom 2 (Order axioms). The set of real numbers R is an ordered field. This means
that (in addition to being a field) R has the following property: There is a set P ⊂ R such
that
1.2. ORDER AXIOMS 7
(O1): −P ∩ P = ∅,
(O2): −P ∪ {0} ∪ P = R, and
(O3): If a, b ∈ P then a + b ∈ P and ab ∈ P.
We mention in passing that the rational numbers also form an ordered field.
Theorem 14. 0 is neither in P nor in −P.
Exercise 4. Show that the field Z2 in Exercise 1 can not be ordered.
Definition 2. The elements in P are called positive and those in −P are called negative.
The non-negative numbers are those in (−P)c = P ∪ {0} while the non-positive numbers are
those in Pc = −P ∪ {0}.
Definition 3. Let a and b be real numbers. We say that a < b or, equivalently, b > a
if b − a ∈ P. We say that a ≤ b or, equivalently, b ≥ a if b − a ∈ P ∪ {0}.
Theorem 15. 1 > 0.
Theorem 16. Suppose a, b ∈ R. Then a ≤ b if and only if a < b or a = b.
Theorem 17. Let x, y ∈ R. Then either x ≤ y or y ≤ x (or both).
Theorem 18. Let x, y ∈ R. If x ≤ y and y ≤ x then x = y.
Theorem 19. For any two real numbers x and y exactly one of the following three
statements is true: x < y, x = y, or x > y.
Theorem 20. Let x, y, z ∈ R. If x < y and y < z, then x < z.
Theorem 21. Let x, y, z ∈ R. Then x + y < x + z if and only if y < z.
Theorem 22. Let x, y, z ∈ R.
(1) If 0 < x and y < z, then xy < xz.
(2) If x < 0 and y < z, then xz < xy.
Theorem 23. If x, y ∈ R and 0 < x < y, then 0 < y −1 < x−1 .
These consequences of the order axioms show that one can represent the real numbers
on the familiar number line. The relationship x < y is to be interpreted as “x lies to the
left of y”. It is often helpful to think along these lines when trying to devise proofs. Here
2 = 1 + 1 and 3 = 2 + 1.
-2 -1 - 12 0 1 1
3 2
1 2 3
Exercise 5. Identify the theorems which guarantee the ordering indicated in this sketch
of the number line.
Theorem 24. If x, y ∈ R and x < y, then there exists a number z ∈ R such that
x < z < y.
This shows that any field satisfying the order axioms will have many elements.
Definition 4. Let a, b be real numbers. Each of the following types of subsets of R
is called a finite interval : (a, b) = {x ∈ R : a < x < b}, [a, b) = {x ∈ R : a ≤ x < b},
(a, b] = {x ∈ R : a < x ≤ b}, and [a, b] = {x ∈ R : a ≤ x ≤ b}. The following sets
are called infinite intervals: (−∞, a) = {x ∈ R : x < a}, (−∞, a] = {x ∈ R : x ≤ a},

(a, ∞) = {x ∈ R : a < x}, [a, ∞) = {x ∈ R : a ≤ x}, and (−∞, ∞) = R. The intervals
(a, b), (−∞, a), and (a, ∞) are called open intervals while [a, b], (−∞, a], and [a, ∞) are
called closed intervals. The interval (−∞, ∞) is considered both an open and a closed
interval.
Note that the empty set is also considered to be an interval, in fact both an open and
a closed interval since ∅ = (a, b) = [a, b] if a > b.
1.3. The induction principle

Definition 5. A subset S of R is called inductive, if 1 ∈ S and if x + 1 ∈ S whenever
x ∈ S.
Theorem 25. The sets R, P, and [1, ∞) are all inductive.
Definition 6. The intersection of all inductive sets is called the set of natural numbers
and is denoted by N. The set Z = −N∪{0}∪N is called the set of integers or whole numbers.
The set Q = {m/n : m ∈ Z, n ∈ N} is called the set of rational numbers.
Exercise 6. If the sets A and B are inductive, then so is A ∩ B.
Theorem 26. N is inductive. In particular, 1 is a natural number.
Theorem 27. If n is a natural number, then n ≥ 1.
Theorem 28 (The induction principle). If M ⊂ N is inductive, then M = N.
The induction principle gives rise to an important method of proof, the so called induc-
tion proofs of which we will see many: M = N, if M has the three properties (i) M ⊂ N,
(ii) 1 ∈ M , and (iii) ∀n ∈ M : n + 1 ∈ M .
Definition 7. If n ∈ N we call n + 1 the successor of n and n the predecessor of n + 1.
Theorem 29. 1 is the only natural number without a predecessor.
Theorem 30. (n, n + 1) ∩ N = ∅ whenever n ∈ N.
Induction may also be used to define concepts. We give some important examples (note
that these depend on the recursion theorem, see Appendix A).
Definition 8. Let n be a natural
Pn number and g a function from N to R. The sum of
the first n terms of g, denoted by k=1 g(k), is defined inductively by
1
X n+1
X n
X
g(k) = g(1) and g(k) = g(n + 1) + g(k).
k=1 k=1 k=1
Qn
Also, the product of the first n terms of g, denoted by k=1 g(k), is defined inductively by
1
Y n+1
Y n
Y
g(k) = g(1) and g(k) = g(n + 1) g(k).
k=1 k=1 k=1
Definition 9 (Factorial). For natural numbers n one defines the number n! (pro-
nounced ‘n factorial’) by
Yn
n! = k.
k=1
In particular, 1! = 1, 2! = 2, and 3! = 2 · 3 = 6. One also defines 0! = 1 (get used to it).
1.3. THE INDUCTION PRINCIPLE 9
Qn
Definition 10 (Powers). If x ∈ R and n ∈ N, then the number k=1 x is called the
n-th power of x. It is denoted by xn . We also define x0 = 1 and, for x 6= 0, x−n = (x−1 )n .
In particular 00 = 1 (get used to that, too).
Theorem 31. For any n ∈ N
n
X n
X
cak = c ak
k=1 k=1
when c and the ak are real numbers.

Theorem 32. For any n ∈ N and real numbers ak and bk
n
X n
X n
X
(ak + bk ) = ak + bk .
k=1 k=1 k=1
Theorem 33. If the ak are real numbers and n ∈ N, then

n
X
an − a0 = (ak − ak−1 ).
k=1
A sum like the one on the right of the previous equation is called a telescoping sum.
Theorem 34. If m and n are natural numbers, then so are n + m and nm.
Theorem 35. The sum of the first n natural numbers is n(n + 1)/2, i.e.,
n
X 1
k= n(n + 1)
2
k=1
for all n ∈ N.
Theorem 36. Let a be real number different from 1 and n a natural number. Then
n
X 1 − an+1
a0 + ak = .
1−a
k=1
Theorem 37 (Bernoulli’s1 inequality). If a > −1 and n ∈ N, then (1 + a)n ≥ 1 + na.

Theorem 38. For all n ∈ N it is true that
n
X n(n + 1)(2n + 1)
k2 = .
6
k=1
Theorem 39. If a 6= 0 and m, n ∈ Z, then am an = am+n and (am )n = amn . Moreover,

if a, n ∈ N, then an ∈ N.
Next, we present the well ordering principle. It is actually equivalent to the induction
principle and sometimes comes handy in proofs.
Theorem 40 (The well-ordering principle).* Let S ⊂ N and S 6= ∅. Then S contains a
smallest element (denoted by min S).
Theorem 41. Suppose 0 < q ∈ Q. Then there exist m, n ∈ N such that m/n = q and,
if k/` = q for k, ` ∈ N, then m ≤ k and n ≤ `.
1Jacob Bernoulli (1654 – 1705)

In the situation of the previous theorem m/n is called the representation of q in lowest
terms. A negative rational number q also has a representation in lowest terms, namely
−m/n, if m/n is the representation in lowest terms of −q > 0.
Theorem 42. There is no rational number r for which r2 = 2.
Is there are real number whose square is 2? It will turn out that introducing the real
numbers allows to handle this deficiency. In fact, that’s why real numbers were invented.
1.4. Counting and infinity

This section deals with counting and the notion of infinity. For a given n ∈ N let
Zn = {k ∈ N : k ≤ n}.
Definition 11. A set X is called finite if it is empty or if there exists an n ∈ N and a

surjective function ϕ : Zn → X. Otherwise the set is called infinite.
If X is a non-empty finite set, then min{k ∈ N : ∃ surjective function ϕ : Zk → X} is
called the number of elements2 in X and is denoted by #X.
Exercise 7. Find out what Z1 , Z2 , and Z3 actually are. Then find all functions
ϕ : Z1 → Z1 and all functions ψ : Z2 → Z2 . Which of these are injective and which are
surjective? Also, find all injective functions τ : Z3 → Z3 .
Exercise 8. Think about how counting is related to surjections, injections, and bijec-
tions.
Theorem 43. #Zn ≤ n.
Theorem 44. #Zn = n.
Theorem 45. Suppose X is a non-empty finite set. If there is a bijective function from
X to Zn , then #X = n. Conversely, if #X = n and ϕ : Zn → X is surjective, then it is, in
fact, bijective.
Theorem 46. N is infinite.
Theorem 47. Suppose X and Y are sets and X ⊂ Y . Then the following two state-
ments are true:
(1) If X is infinite, then Y is infinite.
(2) If Y is finite, then X is finite.
Definition 12. A set X is called countable if it is empty or if there exists a surjective

function φ : N → X. Otherwise it is called uncountable. A set which is countable but not
finite is called countably infinite.
Theorem 48. Every finite set is countable and N is countably infinite.
Theorem 49.* Z, N × N, and Q are countably infinite.
2This number exists by Theorem 40.

1.5. THE LEAST UPPER BOUND AXIOM 11
1.5. The least upper bound axiom

Note that the set of rational numbers, just like the set of real numbers, is an ordered field
(identify the set P which shows this). So far, for all we know, these sets could be identical.
Only the least upper bound axiom, which is our final axiom, distinguishes between the two.
Definition 13. Let S be a subset of R. The number b is called an upper bound of S if
x ≤ b for all x ∈ S. It is called a lower bound of S if x ≥ b for all x ∈ S. If S has an upper
(or lower) bound it is called bounded above (or bounded below). S is called bounded if it has
both an upper and a lower bound.
Exercise 9. Every element of P is an upper bound of −P. What is the least upper
bound of −P?
Definition 14. Let S be a subset of R, which is bounded above, and U the set of all
its upper bounds. A number c is called a least upper bound of S or a supremum of S if it
is an upper bound of S and a lower bound of U . Similarly, if S is bounded below and L is
the set of all lower bounds of S, then a number c which is a lower bound of S and an upper
bound of L is called a greatest lower bound of S or an infimum of S.
Theorem 50. Suppose a, b ∈ R and a < b. Then b is a least upper bound and a is a
greatest lower bound of (a, b).
Theorem 51. If a least upper or a greatest lower bound of S exists, then it is unique.
We are now ready to state our final axiom of the real numbers. This means that
all desired results on the real numbers have to be proven by using nothing more than
Axioms 1 — 3 (and results which are known to follow from these axioms).
Axiom 3 (Least Upper Bound Axiom). Any non-empty subset of R which is bounded
above has a least upper bound.
Both Q and R satisfy Axioms 1 and 2 (they are both ordered fields). It is Axiom 3
which makes all the difference (a tremendous difference as we will see) between the rational
and the real numbers.
Theorem 52. Let S be a subset of R. The set S is bounded above if and only if −S is
bounded below and the number b is an upper bound of S if and only if −b is a lower bound
of −S.
Theorem 53. If S is a non-empty subset of R which is bounded below, then S has a
greatest lower bound.
Definition 15. If S is a subset of R which has a least upper bound we will denote this
uniquely defined number by sup S. If sup S is an element of S, it is called a maximum of S
and is denoted by max S. Similarly, the greatest lower bound of a set S (when it exists) is
denoted by inf S. If inf S is an element of S, it is called a minimum of S and is denoted by
min S.
Theorem 54. If either of the numbers sup S and inf(−S) exist, then so does the other.
Moreover, in this case, sup S = − inf(−S).
Theorem 55. Let S be a bounded non-empty subset of R. Then inf S ≤ sup S.
Theorem 56. Suppose that ∅ = 6 T ⊂ S ⊂ R.
(1) If S is bounded above, then T is bounded above and sup T ≤ sup S.
(2) If S is bounded below, then T is bounded below and inf T ≥ inf S.

Theorem 57 (Archimedian3 property). Let a, b ∈ R and a > 0. Then there exists a
natural number n such that na > b.
The following innocent looking theorem will be very important later on when we will
look at limits.
Theorem 58. Let x be a real number such that 0 ≤ x < 1/n for every natural number
n. Then x = 0.
Theorem 59. If x and y are real numbers and x < y, then there exists q ∈ Q such that
x < q < y.
Because of this last theorem one says that the rational numbers are dense in R but
recall from Theorem 42 that despite of this there are still holes on the number line. This is
why the real numbers are sorely needed.
Our next goal is to prove the existence of square roots of positive real numbers. We
will then use this to show that the rational numbers do not satisfy the least upper bound
axiom.
Theorem 60. Suppose x, y ≥ 0 are real numbers. Then x < y if and only if x2 < y 2 .
Theorem 61. Let y > 0 be a real number and E = {z ∈ R : z > 0, z 2 < y}. Then E is
not empty and bounded above.
Theorem 62.* Let y > 0 be a real number. Then there exists a real number x > 0 such
that x2 = y.
Definition 16. Let x and y be real numbers. x is called a square root of y if x2 = y.
Theorem 63. A real number y < 0 has no square root. The number y = 0 has exactly
one square root, namely 0. A real number y > 0 has exactly two square roots; these are
negatives of each other.
√
Definition
√ 17. If x > 0, then we denote by x the unique positive square root of x.
Also 0 = 0.
We will usually call this number the square root of y ≥ 0, even if we actually should be
more precise and call it the non-negative square root.
√ √ √
Theorem 64. If a ≥ 0 and b ≥ 0, then ab = a b.
Theorem 65. The rational numbers Q do not satisfy the Least Upper Bound Axiom.
More precisely, if A ⊂ Q is bounded above and U is the set of all rational upper bounds of
A, then U may not have a least element.
As the final subject of this chapter we will discuss absolute values.
Definition 18. The absolute value of x ∈ R is defined by
(
x if x ≥ 0,
|x| =
−x if x < 0.
√
Theorem 66. Let x, y be any real numbers. Then x2 = |x| and |xy| = |x| |y|.
3Archimedes of Syracuse (ca. 287 BC – 212 BC)
1.5. THE LEAST UPPER BOUND AXIOM 13
Theorem 67. Suppose a and x are real numbers and c is a positive real number. Then
−|x| ≤ ±x ≤ |x|. Moreover, the interval (a − c, a + c) equals the set {x ∈ R : |x − a| < c}.
Theorem 68 (Triangle inequality). If x, y ∈ R, then
|x + y| ≤ |x| + |y|
and
|x + y| ≥ |x| − |y|.
Definition 19. Let X be a set. A function d : X × X → [0, ∞) is called a metric or
distance function if it has the following properties:
(1) d(x, y) = 0 if and only if x = y.
(2) d(x, y) = d(y, x) for all x, y ∈ X.
(3) d(x, y) ≤ d(x, z) + d(z, y) for all x, y, z ∈ X.
(X, d) is then called a metric space. The number d(x, y) is called the distance between x
and y.
Property (3) is also called triangle inequality.
Theorem 69. The function R × R → [0, ∞) : (x, y) 7→ |x − y| is a metric.
Note that, by Theorem 67, the interval (a − c, a + c) is the set of all real numbers whose
distance from a is less than c.
CHAPTER 2
Sequences and series
2.1. Sequences
Definition 20. A sequence of real numbers is a function which maps N to R.
For a sequence x : N → R we will usually write n 7→ xn , where xn is traditionally used
for the value x(n) of the function x at n.
Exercise 10. For the sequence x : n 7→ 1/n and y : n 7→ 1/n2 consider where the
points xn and yn fall on the number line.
We have the impression that these sequences approach (whatever that might mean)
zero. How could we make such a notion precise?
Exercise 11. Show that |1/n − 0| < 1/100 for all n > 100 and |1/n − 0| < 1/1000 for
all n > 1000.
Exercise 12. Intuitively, what number L do the entries of the sequence n 7→ xn =
(n2 + (−1)n )/n2 approach?
Exercise 13. Consider the sequence from the previous exercise and suppose ε = 1/100.
Find an N ∈ R such that |xn − L| < ε for all n > N . What about ε = 1/500? What will
happen for even smaller ε? Note that ε describes the error we allow.
Definition 21. Let x : n 7→ xn be a sequence. We say that x converges to the real
number L if for every positive real number ε there is an N ∈ R such that |xn − L| < ε
whenever n > N ; or, more concisely,
∀ε > 0 : ∃N ∈ R : ∀n > N : |xn − L| < ε.
We say that the sequence x converges or is convergent if there is a real number L such
that x converges to L. If a sequence is not convergent, we say that it diverges or is divergent.
Theorem 70. If c ∈ R, then x : n 7→ xn = c converges to c.
Theorem 71. The sequence n 7→ 1/n converges to 0.
Theorem 72. If the sequence x converges to L1 and also to L2 , then L1 = L2 .
Definition 22. If a sequence x converges to L, then L is called the limit of x and we
write
lim xn = L.
n→∞
Definition 23. We say that a sequence is bounded above (or below) if its range is
bounded above (or below). A sequence is called bounded if it is bounded both above and
below.
Theorem 73. A finite set of real numbers is bounded.
15
16 2. SEQUENCES AND SERIES
Theorem 74. If a sequence converges, then it is bounded.

Theorem 75. The sequence n 7→ n diverges.
Theorem 76. The sequence n 7→ (−1)n diverges.
Theorem 77. If limn→∞ xn = L, then limn→∞ |xn | = |L|.
Theorem 78. If limn→∞ xn = L and limn→∞ yn = M , then limn→∞ (xn +yn ) = L+M .
Theorem 79. If c ∈ R and limn→∞ xn = L, then limn→∞ cxn = cL.
Theorem 80. If limn→∞ xn = L and limn→∞ yn = M , then limn→∞ xn yn = LM .
Theorem 81. If limn→∞ xn = 0 and the sequence y is bounded, then limn→∞ xn yn = 0.
Theorem 82. If xn 6= 0 for all n ∈ N and if limn→∞ xn = L 6= 0, then limn→∞ 1/xn =
1/L.
Theorem 83. For every k ∈ N it holds that limn→∞ n−k = 0.
Theorem 84. limn→∞ (n2 − n)/(3n2 + 1) = 1/3.
Theorem 85. If the sequences x and y have limits and if xn ≤ yn for all n ∈ N, then
limn→∞ xn ≤ limn→∞ yn .
Theorem 86. If the sequences x and y both have limit L and if xn ≤ zn ≤ yn for all
n ∈ N, then limn→∞ zn = L.
Theorem 87. If a > 0 is a real number, then limn→∞ (1 + a)−n = 0. Moreover, if
c ∈ (−1, 1), then limn→∞ cn = 0.
Definition 24. A sequence x of real numbers is called non-decreasing (or non-increas-
ing) if xn ≤ xn+1 (or xn ≥ xn+1 ) for all n ∈ N. It is called strictly increasing (or strictly
decreasing) if the inequalities are strict.
Warning: Different people use somewhat different notation here: increasing may be
used both for non-decreasing or for strictly increasing. One is on the safe side if one uses
non-decreasing and strictly increasing as we do here.
Theorem 88. If the sequence x is non-decreasing and bounded above, then it converges
to sup{xn : n ∈ N}. Similarly, if x is non-increasing and bounded below, then it converges
to inf{xn : n ∈ N}.
Theorem 89. Suppose the sequence x is bounded and define sequences x̂ and x̌ by
setting x̂n = sup{xk : k ≥ n} and x̌n = inf{xk : k ≥ n}. Then both x̂ and x̌ are convergent.
Definition 25. Suppose the sequence x is bounded. Then the limit of the sequence
n 7→ sup{xk : k ≥ n} is called the limit superior of x and is denoted by lim supn→∞ xn , i.e.,
lim sup xn = lim sup{xk : k ≥ n}.
n→∞ n→∞
Similarly, the limit of the sequence n 7→ inf{xk : k ≥ n} is called the limit inferior of x and
is denoted by lim inf n→∞ xn , i.e.,
lim inf xn = lim inf{xk : k ≥ n}.
n→∞ n→∞
Theorem 90. Let x be the sequence defined by xn = (−1)n + 1/n. Then we have
lim supn→∞ xn = 1 and lim inf n→∞ xn = −1.
2.1. SEQUENCES 17
Theorem 91. Suppose x is a bounded sequence. Then lim inf n→∞ xn ≤ lim supn→∞ xn .
Theorem 92. Let x be a bounded sequence. Then lim supn→∞ −xn = − lim inf n→∞ xn .
Theorem 93. Let x and y be bounded sequences such that xn ≤ yn for all n ∈ N.
Then lim supn→∞ xn ≤ lim supn→∞ yn and lim inf n→∞ xn ≤ lim inf n→∞ yn .
Theorem 94. Let x and y be bounded sequences. Then
lim sup(xn + yn ) ≤ lim sup xn + lim sup yn
n→∞ n→∞ n→∞
and
lim inf (xn + yn ) ≥ lim inf xn + lim inf yn .
n→∞ n→∞ n→∞
Theorem 95. The inequalities in Theorem 94 may be strict.

Theorem 96. Suppose the sequence x is bounded and ε > 0. Then there is an N ∈ R
such that −ε + lim inf n→∞ xn ≤ xn ≤ ε + lim supn→∞ xn whenever n > N .
Definition 26. If x is a sequence of real numbers and k is a strictly increasing sequence
of natural numbers, then the sequence n 7→ xkn is called a subsequence of x.
For instance, the sequences n 7→ 1/(2n) and n 7→ 1/(n+1) are subsequences of n 7→ 1/n.
Theorem 97. If a sequence of real numbers converges to L ∈ R then so does every one
of its subsequences.
Theorem 98.* If the sequence x is bounded, then there exists a subsequence of x which
converges to lim supn→∞ xn and one which converges to lim inf n→∞ xn . In particular, every
bounded sequence has convergent subsequences.
The last statement in the previous theorem is called the Bolzano1-Weierstrass2 theorem.
Theorem 99. Let x be a bounded sequence. Then lim supn→∞ xn = lim inf n→∞ xn if
and only if x converges. In this case it holds that
lim inf xn = lim xn = lim sup xn .
n→∞ n→∞ n→∞
Definition 27. A sequence x is called a Cauchy3 sequence or Cauchy for short if

∀ε > 0 : ∃N ∈ R : ∀n, m > N : |xn − xm | < ε.
Theorem 100. A convergent sequence is Cauchy.
Theorem 101. Any Cauchy sequence is bounded.
Theorem 102. If x is a Cauchy sequence and x0 a convergent subsequence of x, then
x is convergent and x and x0 have the same limit.
Theorem 103. A sequence of real numbers converges if and only if it is a Cauchy
sequence.
The fact that every Cauchy sequences converges does not hold in Q. It is a consequence
of the least upper bound axiom. In fact, one can show that the least upper bound axiom
and the requirement that every Cauchy sequence converges are equivalent. To refer to this
essential property easily one says that R is complete.
1Bernhard Bolzano (1781 – 1848)
2Karl Weierstrass (1815 – 1897)
3Augustin-Louis Cauchy (1789 – 1857)
2.2. Sums and the Σ-notation

In Definition 8 we introduced a notation for the sum of finitely many terms using the
Σ-symbol. It will convenient to extend this notation to a more general setting.
Definition 28. If m and n are any two integers and f a function from Z to R, we
define (
n
X 0 if m > n,
f (k) = Pn+1−m
k=m k=1 f (m − 1 + k) if m ≤ n.
Note that, if m ≤ n, only the values f (j) with m ≤ j ≤ n are needed in the above
definition.
The following theorems are useful when dealing with finite sums. In each of them f is
a real-valued function with an appropriate domain.
Pn Pn+p
Theorem 104. Suppose m, n, p ∈ Z. Then k=m f (k) = k=m+p f (k − p).
Pn Pn
Theorem 105. Suppose m, n ∈ Z. Then k=m f (k) = k=m f (m + n − k).
Theorem 106. Suppose k, m, n ∈ Z and k ≤ m ≤ n. Then
n
X m
X n
X
f (j) = f (j) + f (j).
j=k j=k j=m+1
Pm Pn Pn Pm
Theorem 107. If m, n ∈ N ∪ {0}, then k=0 `=0 f (`, k) = `=0 k=0 f (`, k).
Theorem 108. Any permutation is a finite composition of transpositions.
Theorems 105 and 107 are special cases of the following one.
Theorem 109 Pn commutative law). Suppose n ∈ N and π : Zn → Zn is
Pn(The generalized
bijective. Then k=1 f (k) = k=1 f (π(k)). Moreover, P if A is a finite setPwith n elements
n n
and φ and ψ are bijective functions from Zn to A, then k=1 f (φ(k)) = k=1 f (ψ(k)).
This theorem allows for the following definition.
P Definition Pn29. If A is a finite set and f a function from A to R, then we define

a∈A f (a) = k=1 f (φ(k)) where n = #A and φ : Zn → A is bijective.
2.3. Series
Definition 30. If x : N → R is a sequence of real numbers the sequence
n
X
s : N → R : n 7→ sn = xk
k=1
P∞
is a called the sequence of partial sums of x or a series. We will denote s by k=1 xk .
Thus a series is actually a sequence. Conversely, in view of Theorem 33, a sequence may
also be cast as a series:
Pgiven a sequence s defining x1 = s1 and xn = sn − sn−1 Pshows that
∞ ∞
s is a series, viz., s = k=1 xk . We emphasize that (in this context) the symbol k=1 xk is
not to be thought of as a real number. It is only a concise notation for a sequence of partial
sums. Indeed, if this sequence ofPpartial sums does not converge, it is not sensible to assign
∞
any real number to the symbol k=1 xk .
2.3. SERIES 19
Definition 31. Suppose x is a sequence ofPreal numbers. If the sequence s of partial

∞
sums of x converges to L ∈ R, we say the series k=1 xk converges. We then write (abusing
notation slightly)
X∞
xk = lim sn = L.
n→∞
k=1
Of course, a series (i.e., a sequence of partial sums) may also

P∞ diverge.
We will also use the following notation: If N ∈ Z we define k=N xk to be the sequence
Pn+N −1 P∞
of partial sums n 7→ k=N xk . If this is convergent we denote its limit also by k=N xk .
P∞
Theorem 110. If −1 < a < 1, then the geometric series n=0 an converges to 1/(1−a).
P∞ 1
Theorem 111. n=1 n(n+1) = 1.
P∞ P∞
Theorem 112. If c ∈ R and the series n=1 xn converges, then the series n=1 cxn
converges and
X∞ X∞
cxn = c xn .
n=1 n=1
P∞ P∞ P∞
Theorem 113. If n=1 xn converges and n=1 yn converges, then n=1 (xn + yn )
converges and
∞
X ∞
X ∞
X
(xn + yn ) = xn + yn .
n=1 n=1 n=1
P∞
Theorem 114. If the series n=1 xn converges, then limn→∞ xn = 0.
P∞
Theorem 115. The series n=1 (−1)n diverges.
P∞
Theorem 116 (Cauchy Criterion). The series n=1 xn converges if and only if the
following criterion is satisfied: For every ε > 0 there is an N ∈ R such that, for all m ∈ N
and k ∈ N with m > k > N ,
m
X
xn < ε.

n=k+1
P∞
Theorem 117. The harmonic series n=1 1/n diverges.
This example shows that there is no converse to Theorem 114, i.e., the fact that the
sequence converges to 0 does not guarantee the convergence of the associated sequence of
partial sums.
P∞
Theorem 118. The alternating harmonic series n=1 (−1)n+1 /n is convergent.
The following inequality is a generalized version of the triangle inequality.
Theorem 119. Let xn be a real number for every n ∈ N. Then
k k
X X
x ≤ |xn |

n

n=1 n=1
for any k ∈ N.
Theorem 120 (Comparison Test). Let x, y, and z be sequences P∞ and assume there is
an N ∈ NP such that |xn | ≤ yn ≤ zn for all n ≥ N . If the series n=1 yn converges, then
∞
the series n=1 xn converges, too, and
∞ ∞
X X
x ≤ yn .

n

n=N n=N
P∞ P∞
If the series n=1 yn diverges, then the series n=1 zn diverges, too.
P∞
Theorem 121. If 2 ≤ k ∈ N, then n=1 n−k converges.
Theorem 122 (Raabe’s4 Test). Suppose that there is an N ∈ N and a β > 1 such that

an+1 β
an ≤ 1 − n + 1

P∞
whenever n ≥ N . Then the series n=1 an converges.
Theorem 123 (Ratio Test). Suppose that xn 6= 0 for all n ∈ N and that

xn+1
lim sup
< 1.
n→∞ xn
P∞
Then the series n=1 xn converges.
P∞
Theorem 124. Let a be a fixed real number. Then the series n=0 an /n! converges.
Definition 32. We set
∞
X
exp(a) = an /n!
n=0
and
∞
X
exp(1) = 1/n! = e.
n=0
Definition 33. For integers n and k with n ≥ k ≥ 0 let

n n!
= .
k k!(n − k)!
These numbers are called binomial coefficients.
Theorem 125. For integers n and k with 1 ≤ k ≤ n it holds that

n+1 n n
= + .
k k k−1
Theorem 126 (Binomial Theorem). Let a, b ∈ R and n a non-negative integer. Then
n
n
X n n−k k
(a + b) = a b .
k
k=0
n
Theorem 127. lim supn→∞ (1 + 1/n) ≤ e.
Theorem 128. lim inf n→∞ (1 + 1/n)n ≥ e.
Combining Theorem 127 and Theorem 128 we get
e = lim (1 + 1/n)n .
n→∞
4Joseph Ludwig Raabe (1801 – 1859)

2.3. SERIES 21
Theorem 129. Let x ≥ 0 and n ∈ N. Then there exists a unique y ≥ 0 such that
y n = x.
Definition 34. The unique number
√ y given by Theorem 129 is called the (non-negative)
n-th root of x ≥ 0. We write y = n x.
√
Inspired by Theorem 39 the notation y = n x = x1/n is frequently used.
√
Theorem 130. If p > 0, then limn→∞ n p = 1.
√
Theorem 131. limn→∞ n n = 1.
p
P∞(Root Test). Let x be a sequence such that lim supn→∞ |xn | < 1.
n
Theorem 132
Then the series n=1 xn converges.
The following inequality shows that the root test is more precise than the ratio test.
The advantage of the ratio test is that it is usually more easily applicable.
Theorem 133. Suppose a is a bounded sequence of positive numbers. Then
√ an+1
lim sup n an ≤ lim sup
n→∞ n→∞ an
if the right hand side exists.
P∞ P∞
Definition 35. ThePseries n=1 xn is said to converge absolutely if the series n=1 |xn |
∞
converges. If the series n=1 xn converges, but does not converge absolutely, then we say
that it converges conditionally.
P∞
Theorem 134. If the series n=1 xn converges absolutely, then it converges and

X∞ X ∞
xn ≤ |xn |.

n=1 n=1
Definition 36. Let π : N → N be a bijective sequence. Let Px∞be a given sequence and
define a sequence
P∞ y by yn = x πn . Then we say that the series n=1 yn is a rearrangement
of the series n=1 xn .
P∞ P∞
Theorem 135.* P∞Suppose thatP∞ n=1 xn converges absolutely and let n=1 yn be a
rearrangement of n=1 xn . Then n=1 yn converges absolutely, and
∞
X ∞
X
yn = xn .
n=1 n=1
P∞
Let n=1 xn be a convergent series which does not converges absolutely. Then there is
a rearrangement of the series such that the sequence of its partial sums is not bounded above
or below. Moreover, given any real number s, there is a rearrangement which converges to
s. We will not try to prove this statement.
CHAPTER 3
A zoo of functions
Definition 37. Let X be a set and f, g functions from X to R. We define their sum,
difference, and product respectively by (f ± g)(x) = f (x) ± g(x) and (f g)(x) = f (x)g(x)
for all x ∈ X. If g(x) 6= 0 for all x ∈ X we also define the quotient of f and g by
(f /g)(x) = f (x)/g(x).
The composition of functions (g ◦ f )(x) = g(f (x)) is defined in the appendix.
Definition 38. Let S ⊂ R. A function f : S → R is called non-decreasing (or non-
increasing) if f (x) ≤ f (y) (or f (x) ≥ f (y)) whenever x, y ∈ S and x ≤ y. It is called
strictly increasing (or strictly decreasing) if the inequalities are strict. A function is called
monotone if it is non-increasing or non-decreasing. It is called strictly monotone if it is
strictly increasing or strictly decreasing.
Definition 39. If n is a non-negative integer and a0 , a1 , . . ., an are real numbers, then
the function p : R → R defined by
n
X
p(x) = ak xk
k=0
is called a polynomial function or a polynomial for short. The integer n is called the degree
of p if an 6= 0. The zero function is also a polynomial but no degree is assigned to it.
Theorem 136. If p and q are polynomials of degree n and k, respectively, then p + q
and pq are also polynomials. The degree of p + q is the larger of the numbers n and k unless
n = k in which case the degree is at most n. The degree of pq equals n + k.
Theorem 137. A polynomial p of degree n has at most n zeros, i.e., there are at most
n points x such that p(x) = 0.
Definition 40. Let p and q be polynomials and assume q 6= 0, i.e., q is not the zero
polynomial. Let S = {x ∈ R : q(x) 6= 0}. Then, the function r : S → R given by
p(x)
r(x) =
q(x)
is called a rational function.
Every polynomial is a rational function (choose q = 1).
Definition 41. Let J be an interval in R. A function f : J → R is called an algebraic
function if there is a natural number n and polynomials p0 , p1 , ..., pn , not all zero, such
that
p0 (x) + p1 (x)f (x) + · · · + pn (x)f (x)n = 0
for all x ∈ J. A function which is not algebraic is called transcendental.
23
24 3. A ZOO OF FUNCTIONS
Every rational function is algebraic (choose n = 1). Other

√ examples of algebraic func-
√
tions are f (x) = n x for n ∈ N and J = [0, ∞) and f (x) = 1 − x2 for J = [−1, 1].
Definition 42. The function exp : R → R defined by
∞
X xk
exp(x) =
k!
k=0
is called the exponential function.

The exponential function is transcendental but we will not prove this fact.
Theorem 138. Let En (x) denote the partial sums of the series defining the exponential
function, i.e.,
n
X xk
En (x) = .
k!
k=0
If |x|, |y| ≤ a and if a ≥ 1, then
(2a2 )n
|En (x)En (y) − En (x + y)| ≤ .
n!
Theorem 139. Suppose x, y ∈ R. Then
exp(x + y) = exp(x) exp(y).
Theorem 140. exp(x) > 1 for all x > 0 and exp(x) > 0 for all x ∈ R.
Theorem 141. exp(x) = 1 if and only if x = 0 and exp(x) = exp(y) if and only if
x = y. The exponential function is strictly increasing.
Theorem 142. The range of the exponential function is (0, ∞).
Since exp : R → (0, ∞) is bijective it has an inverse function.
Definition 43. The function log : (0, ∞) → R defined by
log(x) = y if and only if x = exp(y)
(i.e., the inverse function of exp) is called the logarithmic function.
Theorem 143. The logarithmic function is a bijection from (0, ∞) to R. log(1) = 0 and
log(x) is positive for x > 1 and negative for 0 < x < 1. Moreover, log(xy) = log(x) + log(y)
if x, y > 0. Finally, if x > 0, then log(1/x) = − log(x).
Theorem 144. If 0 < x ∈ R, p, q ∈ Z and q 6= 0, then xp/q = exp(p log(x)/q).
Definition 44. Let x be a positive real number and r any real number. Then we define
the r-th power of x by xr = exp(r log(x)).
Since exp(1) = e and hence log(e) = 1 we have, in particular, that exp(r) = er whenever
r ∈ R.
Theorem 145. Suppose x > 0 and r, s ∈ R. Then log(xr ) = r log(x), xr xs = xr+s , and
(x ) = xrs .
r s
P∞ x2k+1 P∞ x2k
Theorem 146. Suppose x > 0. Then the series k=0 (2k+1)! and k=0 (2k)! are con-
vergent.
3. A ZOO OF FUNCTIONS 25
Definition 45. The function sin : R → R defined by

∞
X x2k+1
sin(x) = (−1)k
(2k + 1)!
k=0
is called the sine function.
The function cos : R → R defined by
∞
X x2k
cos(x) = (−1)k
(2k)!
k=0
is called the cosine function.
The function tan = sin / cos (which is not defined on all of R) is called the tangent
function.
The functions sin, cos, and tan are called trigonometric functions. They are not one-to-
one but restrictions to certain intervals are. On these one can define their inverse functions.
Before we can study these issues more deeply we need better tools, which will be provided
in the next chapters.
CHAPTER 4
Continuity
4.1. Limits of functions

Definition 46. Let S ⊂ R and a ∈ R. We say that a is a limit point of S if there exists
a sequence x such that xn ∈ S \ {a} for every n ∈ N and limn→∞ xn = a. A point a ∈ S
which is not a limit point of S is called an isolated point of S.
Exercise 14. Find the limit points and the isolated points of {1/n : n ∈ N} and those
of the interval (0, 1).
Theorem 147. Let S be a subset of R. The point a ∈ S is an isolated point of S if and
only if there exists a positive δ such that (a − δ, a + δ) ∩ S = {a}.
Definition 47. Let S be a subset of R, f : S → R a function, and a ∈ R a limit point of
S. We say that f converges to the real number b as x tends to a if for every positive ε there
is a positive δ such that for all x ∈ S for which 0 < |x − a| < δ we have that |f (x) − b| < ε
or, more formally, if
∀ε > 0 : ∃δ > 0 : ∀x ∈ S : 0 < |x − a| < δ ⇒ |f (x) − b| < ε.
It is important to note that the point a does not necessarily have to be an element of
S.
Theorem 148. Suppose f : S → R converges to both b1 and b2 as x tends to a. Then
b1 = b2 .
Definition 48. If f : S → R converges to b as x tends to a, we call b the limit of f at
a and we write limx→a f (x) = b.
Thus limits, when they exist, must be unique. We emphasize, however, that the limit
of f at a does not have to be equal to f (a) even if a ∈ S. Definition 47 requires that a can
be approached by elements of S other than a. We do not define a limit of a function at an
isolated point of its domain.
Exercise 15. Does the function x 7→ (x + 2)/x have a limit at a = 4? Find an interval
(c, d) such that (x + 2)/x is closer to that limit than ε = 0.001 for all x ∈ (c, d). What would
δ be?
Theorem 149.* Let S, f , and a be as in Definition 47. Then, f has a limit at a if and
only if the sequence n 7→ f (un ) is convergent whenever u : N → S \ {a} is a sequence with
limit a. In this situation all limits in question are equal to each other.
Theorem 150. Let S ⊂ R, a ∈ R a limit point of S, and f : S → R and g : S → R
functions which have limits at a. Then f ± g have limits at a. In fact,
lim (f ± g)(x) = lim f (x) ± lim g(x).
x→a x→a x→a
27
28 4. CONTINUITY
Theorem 151. Let S ⊂ R, a ∈ R a limit point of S, and f : S → R and g : S → R

functions which have limits at a. Then f g has a limit at a. In fact,
lim (f g)(x) = lim f (x) lim g(x).
x→a x→a x→a
Theorem 152. Let S, a, f , and g be as in Theorem 151. Suppose that g(x) 6= 0 for all
x ∈ S and that limx→a g(x) 6= 0. Then f /g has a limit at a. In fact,
lim (f /g)(x) = lim f (x)/ lim g(x).
x→a x→a x→a
4.2. Continuous functions

Definition 49. Let S ⊂ R and a ∈ S. A function f : S → R is called continuous at a
if and only if for every ε > 0 there exists a δ > 0 such that for all x ∈ S with |x − a| < δ it
holds that |f (x) − f (a)| < ε. Concisely,
∀ε > 0 : ∃δ > 0 : ∀x ∈ S : |x − a| < δ ⇒ |f (x) − f (a)| < ε.
We say that f : S → R is continuous on S (or simply continuous) if f is continuous at
every point of S.
Theorem 153. Suppose S ⊂ R, f : S → R, and a ∈ S is an isolated point of S. Then
f is continuous at a.
Theorem 154. Suppose S ⊂ R, f : S → R, and a ∈ S is not an isolated point of S.
Then f is continuous at a if and only if limx→a f (x) exists and equals f (a).
Theorem 155. Let c a fixed real number.
(1) The constant function f : R → R : x 7→ c is continuous.
(2) The identity function f : R → R : x 7→ x is continuous.
(3) The absolute value function f : R → R : x 7→ |x| is continuous.
Theorem 156. The Heaviside function f : R → R defined by
(
0 for x < 0,
f (x) =
1 for x ≥ 0,
is not continuous at the point a = 0 but continuous everywhere else.
Theorem 157. The function f : R → R defined by
(
x for x ∈ Q,
f (x) =
0 for x ∈ R \ Q
is continuous at a = 0 but not continuous at any other point in R.
Theorem 158. Let S ⊂ R, a ∈ S, and f and g be functions defined on S which are
continuous at a. Then the following statements are true.
(1) f ± g is continuous at a.
(2) f g is continuous at a.
(3) If g(a) 6= 0, then f /g is continuous at a.
Theorem 159. All polynomials are continuous on R.
Theorem 160. Let p and q be polynomials. Then the rational function r = p/q is
continuous on S = {x ∈ R : q(x) 6= 0}, its domain.
4.4. UNIFORM CONVERGENCE AND CONTINUITY 29
Theorem 161. Let S1 and S2 be subsets of R and consider the functions f : S1 → S2

and g : S2 → R. If f is continuous at a ∈ S1 and g is continuous at f (a), then g ◦ f is
continuous at a.
Theorem 162. Let S ⊂ R and f : S → R a continuous function. Then |f | : S → R
defined by |f |(x) = |f (x)| is a continuous function.
√ p
Theorem 163. The function x 7→ x is continuous on [0, ∞) and the function x 7→ |x|
is continuous on R.
√
Theorem 164. The function x 7→ 1 − x2 is continuous on [−1, 1].
Theorem 165. Suppose I ⊂ R is an interval and f : I → R is either strictly increasing
or strictly decreasing. If f (I) is also an interval, then f is continuous.
Theorem 166. If f : S → R is strictly increasing, then f is one-to-one and f −1
is strictly increasing. The statements where “strictly increasing” is replaced by “strictly
decreasing” also holds.
Theorem 167. The exponential function and the logarithmic function are continuous
on their respective domains.
Theorem 168. Suppose r ∈ R. Then the power function x 7→ xr is continuous on
(0, ∞).
4.3. The intermediate value theorem and some of its consequences

The following theorem is very useful in determining the range of a function. We have
used the central idea of its proof in showing the existence of roots (Theorems 62 and 129)
and in establishing the range of the exponential function (Theorem 142).
Theorem 169 (Intermediate Value Theorem).* Let a, b ∈ R be such that a < b. Let
f : [a, b] → R be a continuous function, and let A ∈ R be such that f (a) ≤ A ≤ f (b) or
f (b) ≤ A ≤ f (a). Then there exists x ∈ [a, b] such that f (x) = A.
Theorem 170. Let a, b ∈ R with a < b and f : [a, b] → R a continuous function. Then
f ([a, b]) is bounded.
there are points x1 and x2 in [a, b] such that f (x1 ) = sup f ([a, b]) and f (x2 ) = inf f ([a, b]).
Recall that a supremum is called a maximum and an infimum a minimum, if it is an
element of the set in question.
f ([a, b]) is a closed and bounded interval.
4.4. Uniform convergence and continuity

Definition 50. Let S ⊂ R and, for each n ∈ N, let fn be a function from S to R. The
map n 7→ fn is called a sequence of functions. We say that n 7→ fn converges pointwise to a
function f : S → R if for each point x ∈ S the numerical sequence n 7→ fn (x) converges to
f (x).
Theorem 173. Define the functions fn : [0, 1] → R by fn (x) = xn for n ∈ N. The
sequence n 7→ fn converges pointwise to the function which is 0 on [0, 1) but 1 at 1.
Note that each fn is continuous but that the limit function is not.
30 4. CONTINUITY
Definition 51. Let S ⊂ R and n 7→ fn a sequence of functions on S. We say that

n 7→ fn converges uniformly to a function f : S → R if for every ε > 0 there exists N ∈ R
such that for all n > N and all x ∈ S we have that |fn (x) − f (x)| < ε.
To reiterate, consider the statements
(1) ∀x ∈ S : ∀ε > 0 : ∃N ∈ R : ∀n > N : |fn (x) − f (x)| < ε.
(2) ∀ε > 0 : ∃N ∈ R : ∀x ∈ S : ∀n > N : |fn (x) − f (x)| < ε.
A sequence n 7→ fn converges to f pointwise if (1) holds and uniformly if (2) holds.
Exercise 16. Let fn be the sequence defined in Theorem 173. Given ε > 0 find N so
that |fn (x)| < ε for all n > N where x is, in turn, each of the values 1/10, 9/10, and 99/100.
Theorem 174. Uniform convergence implies pointwise convergence but not vice versa.
Theorem 175. Let S ⊂ R and n 7→ fn a sequence of continuous functions on S which
converges uniformly to a function f : S → R. Then f is continuous on S.
Theorem 176 (Cauchy criterion for uniform convergence). Let S ⊂ R and fn : S → R
for n ∈ N. The sequence of functions n 7→ fn converges uniformly on S if and only if
for all ε > 0 there exists N ∈ R such that for all n, m > N and all x ∈ S it holds that
|fn (x) − fm (x)| < ε.
A series of functions is defined as the sequence of partial sums of functions with a
common domain. Thus the definitions of pointwise and uniform convergence extend also to
series of functions. For instance, the exponential function is a series of polynomials (powers
to be more precise).
Theorem 177 (Weierstrass M -test). Let S ⊂ R and suppose n 7→ gn is a sequence of
real-valued functions defined on S. Assume that there
P∞ are non-negative numbers Mn such
that |gn (x)| ≤ Mn for all x ∈ S and that the series n=1 Mn converges. Then the following
statements hold:
P∞
(1) The series Pn=1 gn (x) converges absolutely for every x ∈ S.
∞
(2) The series n=1 gn converges uniformly in S.
P∞Definition 52. Let x0 ∈ R and an ∈ R for all non-negative integers n. Then the series
n
n=0 an (x − x 0 ) is called a power series.
P∞
P∞Theorem 178. Suppose that x, t ∈ R and |x| < |t|. If tn converges, then
n=0 anP
n
P ∞ n ∞ n
n=0 an x converges absolutely. If, however, n=0 an x diverges, then n=0 an t diverges
also.
P∞
Definition 53. Let the series S = n=0 an xn and the set
∞
X
A = {x ≥ 0 : |an |xn converges}
n=0
be given. If A is bounded the number sup A is called the radius of convergence of S. If A

is unbounded, we say that S has infinite radius of convergence.
The following theorem explain the choice of the name “radius of convergence”.
Theorem 179. Let S and A be as in Definition 53. If A is bounded, then S converges
absolutely when |x| < sup A and diverges when |x| > sup A. If A is unbounded, then S
converges absolutely for all x ∈ R.
4.4. UNIFORM CONVERGENCE AND CONTINUITY 31
P∞
Theorem 180. Suppose the power series n=0 an xn has radius of convergence R > 0
and that 0 < ε < R. Then the power series converges uniformly on S = {x ∈ R : |x| ≤
R − ε}. If it has infinite radius of convergence and r > 0, then it converges uniformly on
S = {x ∈ R : |x| ≤ r}.
Theorem 181. The power series defining the exponential function or the sine or cosine
functions have infinite radius of convergence. Each of these functions is continuous on R.
CHAPTER 5
Differentiation
Throughout this chapter S denotes a set without isolated points. The most important
case is when S is an interval (finite or infinite; open, closed, or half-closed).
5.1. Derivatives
Definition 54. Let f : S → R be a function, and a ∈ S. We say that f is differentiable
at a if
f (x) − f (a)
lim
x→a x−a
exists. The limit is called the derivative of f at a and is commonly denoted by f 0 (a). If
f : S → R is differentiable at every point a ∈ S, we say that f is differentiable on S.
The expression
f (x) − f (a)
,
x−a
defined for x ∈ S \ {a} is called a difference quotient.
The difference quotient has an obvious geometric interpretation as the slope of a secant
passing through two points on the graph of f . The notation f 0 for the derivative of f was
introduced by Newton1. We will not use Leibniz’s2 notation involving differentials (df /dx).
Theorem 182. Let f : S → R be differentiable at a ∈ S. Then f is continuous at a.
Theorem 183. The constant function and the identity function are differentiable on
R. Their derivatives at any point are 0 and 1, respectively.
Theorem 184. The function h : R → [0, ∞) : x 7→ |x| is differentiable on R \ {0} but
not at zero.
√
Theorem 185. The function f : [0, ∞) → [0, ∞) : x 7→ x is differentiable on (0, ∞)
but not at zero.
Theorem 186. Let f : S → R and g : S → R be functions which are differentiable at
a ∈ S and let c ∈ R be a constant. Then
(1) cf is differentiable at a and (cf )0 (a) = cf 0 (a) and
(2) f ± g are differentiable at a, and (f ± g)0 (a) = f 0 (a) ± g 0 (a).
Theorem 187 (Product or Leibniz rule for derivatives). Let f : S → R and g : S → R
be functions which are differentiable at a ∈ S. Then, f g is differentiable at a and
(f g)0 (a) = f 0 (a)g(a) + f (a)g 0 (a).
1Isaac Newton (1643 – 1727)

2Gottfried Wilhelm Leibniz (1646 – 1716)
33
34 5. DIFFERENTIATION
Theorem 188 (Quotient rule for derivatives). Let f : S → R and g : S → R be

functions which are differentiable at a ∈ S. Also assume that g(a) 6= 0. Then, f /g is
differentiable at a and
g(a)f 0 (a) − f (a)g 0 (a)
(f /g)0 (a) = .
[g(a)]2
Theorem 189. All polynomials are differentiable on R. In particular, if n ∈ N and
f (x) = xn , then
f 0 (x) = nxn−1 .
Theorem 190. Any rational function is differentiable on its domain. In particular, if

n ∈ N and g(x) = x−n , then
g 0 (x) = −nx−n−1 for x ∈ R \ {0}.
Note that the rule for the differentiation of the power function with an integer power
works for all n ∈ Z.
Theorem 191. Suppose a ∈ S, f : S → R, and m ∈ R. Then f is differentiable at a

with derivative m if and only if there exists a function h : S → R which is continuous at a
with h(a) = 0 such that
f (x) = f (a) + m(x − a) + (x − a)h(x).
Since, in a vicinity of a, the term (x − a)h(x) is generally much smaller than f (a) +
m(x − a), the theorem says that f (x) may be well approximated by f (a) + m(x − a) if and
only if it is differentiable at a. The function x 7→ f (a) + m(x − a) is called the tangent line
of f at the point (a, f (a)).
Theorem 192 (Chain Rule). Suppose S1 and S2 are intervals, f : S1 → R, g : S2 → R

and f (S1 ) ⊂ S2 . If f is differentiable at a and g is differentiable at f (a), then g ◦ f is
differentiable at a, and
(g ◦ f )0 (a) = g 0 (f (a))f 0 (a).
Let f : S → R be a differentiable function on S. Then x 7→ f 0 (x) defines another

function f 0 : S → R.
Definition 55 (Higher order derivatives). Let f : S → R be a differentiable function

on S. If f 0 is differentiable at some point a ∈ S, then we denote its derivative at a by f 00 (a),
and we call it the second order derivative of f at a. The function f is then called twice
differentiable at a.
Higher order derivatives may now be defined recursively. Assuming, for some n ∈ N,
that f has derivatives up to order n on S (denoted by f (k) , k = 1, ..., n) and that the n-th
order derivative is differentiable at a, one defines the (n + 1)-st order derivative f (n+1) (a)
as the derivative of f (n) at a. It is also customary to use f (0) instead of f .
Theorem 193. Polynomials have derivatives of all orders on all of R. In fact, if p is a

polynomial of degree d, then p(n) (x) = 0 for all x ∈ R and all integers n > d.
Theorem 194. The exponential functions has derivatives of all orders on all of R, in
fact exp(n) (x) = exp(x) for all n ∈ N and all x ∈ R.
5.2. THE MEAN VALUE THEOREM AND TAYLOR’S THEOREM 35
5.2. The mean value theorem and Taylor’s theorem

In this section a and b always denote two real numbers such that a < b.
Definition 56 (Local extrema of a function). Let f : S → R be a function. We say
that f has a local minimum (or a local maximum) at x0 ∈ S, if there exists a number δ > 0
such that f (x) ≥ f (x0 ) (or f (x) ≤ f (x0 )) for every x ∈ (x0 − δ, x0 + δ) ∩ S. The point x0
is called a local extremum if it is either a local minimum or a local maximum.
Theorem 195. Let f : (a, b) → R be differentiable at x0 ∈ (a, b). If f has a local
maximum or a local minimum at x0 , then f 0 (x0 ) = 0.
Theorem 196 (Rolle’s3 Theorem). Let f : [a, b] → R be continuous on [a, b] and
differentiable on (a, b). Suppose that f (a) = f (b). Then, there exists c ∈ (a, b) such that
f 0 (c) = 0.
Theorem 197 (The mean value theorem). Let f, g : [a, b] → R be continuous on [a, b]
and differentiable on (a, b). Then there exists c ∈ (a, b) such that
(f (b) − f (a))g 0 (c) = (g(b) − g(a))f 0 (c).
In particular, there is a c ∈ (a, b) such that
f (b) − f (a) = f 0 (c)(b − a).
Theorem 198. Let f : [a, b] → R be continuous on [a, b] and differentiable on (a, b).
Suppose that f 0 (x) = 0 for all x ∈ (a, b). Then there exists C ∈ R such that f (x) = C for
all x ∈ [a, b].
Theorem 199. Let f : [a, b] → R be continuous on [a, b] and differentiable on (a, b).
Then the following statements hold:
(1) If f 0 (x) ≥ 0 for all x ∈ (a, b), then f is non-decreasing on [a, b].
(2) If f 0 (x) > 0 for all x ∈ (a, b), then f is strictly increasing on [a, b].
Analogous results hold, of course, if f 0 (x) ≤ 0 or f 0 (x) < 0 for all x ∈ (a, b).
Theorem 200. Let f : (a, b) → R be differentiable and assume that f 0 (x) > 0 (or
0
f (x) < 0) for all x ∈ (a, b). Then the inverse function g of f is differentiable on its domain
and g 0 = 1/(f 0 ◦ g).
Theorem 201. The logarithmic function is differentiable on (0, ∞) and log0 (x) = 1/x.
Theorem 202. Suppose r ∈ R. Then the power function pr : x 7→ xr is differentiable
on (0, ∞) and p0r (x) = rxr−1 .
Theorem 203 (Taylor’s4 theorem).* Let f : (a, b) → R have n + 1 derivatives on (a, b)
and let x and x0 be two distinct points in (a, b). Then there exists c between x and x0 such
that
n
X f (k) (x0 ) f (n+1) (c)
f (x) = (x − x0 )k + (x − x0 )n+1 .
k! (n + 1)!
k=0
Theorem 204. Let f : (a, b) → R be a function which together with its derivative is
differentiable on (a, b) and whose second derivative is continuous on (a, b). Let x0 ∈ (a, b)
and suppose f 0 (x0 ) = 0. Then the following two statement are true.
(1) If f 00 (x0 ) < 0, then f has a local maximum at x0 .
(2) If f 00 (x0 ) > 0, then f has a local minimum at x0 .
3Michel Rolle (1652 – 1719)
4Brook Taylor (1685 – 1731)
36 5. DIFFERENTIATION
5.3. Uniform convergence and differentiation

Theorem 205. Define the functions fn : [0, 1] → R by fn (x) = xn /n for n ∈ N. The
sequence n 7→ fn converges uniformly to the function f = 0, but the sequence n 7→ fn0 does
not converge to f 0 = 0.
Note that the operations of taking a limit and taking a derivative do not necessarily
commute.
Theorem 206. Suppose that the functions fn : [a, b] → R are differentiable, that the
sequence of functions n 7→ fn0 converges uniformly on [a, b] and that, for some x0 ∈ [a, b],
the numerical sequence n 7→ fn (x0 ) is also convergent. Then n 7→ fn converges uniformly
on [a, b].
Theorem 207. Suppose functions fn with the properties of Theorem 206 are given.
For fixed x ∈ [a, b] and n ∈ N define Hn : [a, b] → R by
(
(fn (t) − fn (x))/(t − x) if t 6= x
Hn (t) =
fn0 (x) if t = x.
Then each Hn is continuous and the sequence n 7→ Hn converges uniformly on [a, b].
Theorem 208. Suppose that the functions fn : [a, b] → R are differentiable, that the
sequence of functions n 7→ fn0 converges uniformly on [a, b] and that, for some x0 ∈ [a, b],
the numerical sequence n 7→ fn (x0 ) is also convergent. Then n 7→ fn converges uniformly
on [a, b] to a differentiable function on [a, b] and, for each x ∈ [a, b],
( lim fn )0 (x) = lim fn0 (x).
n→∞ n→∞
Theorem 209.PSuppose that the functions gk : [a, b] → R are differentiable, that the
∞ 0
series of functions P k=0 gk converges uniformly on [a, b] and that, for some x0 ∈ [a, b],
∞ P∞
the numerical series k=0 gk (x0 ) also converges. Then k=0 gk converges uniformly to a
differentiable function on [a, b] and, for each x ∈ [a, b],
∞
!0 ∞
X X
gk (x) = gk0 (x).
k=0 k=0
P∞ n
P∞
Theorem 210. The power series n=0 cn x and n=1 ncn xn−1 have the same radius
of convergence.
Theorem P∞ 211. Let a > 0. Suppose the function f : (−a, a) → R is defined by the
power series n=0 cn xn whose radius of convergence is at least a. Then f has derivatives
of all orders on (−a, a).
Theorem 212. sin and cos have derivatives of all orders on R. In fact, sin0 = cos and
0
cos = − sin.
Definition 57. Suppose f : (a, b) → R has derivatives of all orders at x0 ∈ (a, b).
Consider the power series
∞
X f (n) (x0 )
Tf (x) = (x − x0 )n .
n=0
n!
If there is a positive r such that f (x) = Tf (x) for all x ∈ (x0 − r, x0 + r) then f is called
analytic at x0 and Tf is called the Taylor series of f at x0 .
5.3. UNIFORM CONVERGENCE AND DIFFERENTIATION 37
Analytic functions are the realm of Complex Analysis, perhaps the most beautiful sub-
ject in Mathematics.
Theorem 213. The function x 7→ 1/(1 − x) defined on (−∞, 1) is analytic at zero.
CHAPTER 6
Integration
Throughout this chapter a and b denote two real numbers such that a < b.
6.1. Existence and uniqueness of integrals

Definition 58. A partition P of [a, b] is a finite subset of [a, b] which contains both a
and b. If the number of elements in P is n + 1 we will label them so that
a = x0 < x1 < . . . < xn−1 < xn = b.
A partition P 0 is called a refinement of P if P ⊂ P 0 . P ∪ Q is called the common refinement
of the partitions P and Q.
Definition 59. If P is a partition of [a, b] with n + 1 elements and f : [a, b] → R is a
bounded function we define the lower Riemann sum L(f, P ) and the upper Riemann sum
U (f, P ) by
Xn
L(f, P ) = mj (xj − xj−1 )
j=1
and
n
X
U (f, P ) = Mj (xj − xj−1 )
j=1
where mj = inf{f (x) : xj−1 ≤ x ≤ xj } and Mj = sup{f (x) : xj−1 ≤ x ≤ xj }.

Integration was put first on firm footing by Cauchy in 1823. The above definition is
due to Darboux1. Riemann2 sums are actually such where mj or Mj is replaced by f (tj )
for some tj ∈ [xj−1 , xj ].
Theorem 214. If P is a partition of [a, b] and f : [a, b] → R a bounded function, then
L(f, P ) ≤ U (f, P ).
Theorem 215. If P is a partition of [a, b], P 0 a refinement of P , and f : [a, b] → R a
bounded function, then L(f, P ) ≤ L(f, P 0 ) ≤ U (f, P 0 ) ≤ U (f, P ).
Theorem 216. If P and Q are partitions of [a, b] and f : [a, b] → R a bounded function,
then L(f, P ) ≤ U (f, Q).
Theorem 217. Suppose f : [a, b] → R is a bounded function. Then the set {L(f, P ) :
P is a partition of [a, b]} is bounded above and thus has a supremum. Likewise, the set
{U (f, P ) : P is a partition of [a, b]} is bounded below and has an infimum.
1Jean-Gaston Darboux (1842 – 1917)

2Bernhard Riemann (1826 – 1866)
39
40 6. INTEGRATION
Definition 60. Suppose f : [a, b] → R is a bounded function. The number

I− (f, a, b) = sup{L(f, P ) : P is a partition of [a, b]}
is called the lower Riemann integral of f over [a, b] and the number
I+ (f, a, b) = inf{U (f, P ) : P is a partition of [a, b]}
is called the upper Riemann integral of f over [a, b].
Theorem 218. Suppose f : [a, b] → R is a bounded function. Then I− (f, a, b) ≤
I+ (f, a, b).
Definition 61. Suppose f : [a, b] → R is a bounded function. If I− (f, a, b) = I+ (f, a, b)
we say that f is Riemann integrable (or just integrable) over [a, b]. It is customary to denote
the common value of I− (f, a, b) and I+ (f, a, b) by
Z b
f (x)dx.
a
R
Leibniz thought of the integral sign as an elongated s, the initial letter of sum. He
thought of the integral as a sum of infinitely many “infinitesimally” narrow rectangles of
Rb
width dx and height f (x). Today we might as well use the notation a f but Leibniz’s
notation still has the following two advantages: (1) In physics, where quantities have units,
it becomes clear what the unit of the integral is, e.g., if x denotes time and f velocity
the integral will be a distance. (2) For the integrand xr the use of dx indicates that one
integrates x 7→ xr rather than r 7→ xr .
Theorem 219. Suppose f : [a, b] → R is a bounded function. Then f is Riemann
integrable if and only if, for every positive ε, there is a partition P such that
U (f, P ) − L(f, P ) < ε.
Theorem 220. Let c ∈ [0, 1] and suppose f : [0, 1] → R is defined by
(
0 for 0 ≤ x ≤ c,
f (x) =
1 for c < x ≤ 1.
R1
Then f is Riemann integrable and 0 f = 1 − c.
Theorem 221 (Dirichlet’s function). The function f : [0, 1] → R defined by
(
1 for x ∈ [0, 1] ∩ Q,
f (x) =
0 for x ∈ [0, 1] ∩ (R \ Q)
is not Riemann integrable.
Exercise 17. Compute I± for the function given in Theorem 221.
Theorem 222. The identity function on [a, b] is Riemann integrable.
Rb
Exercise 18. Compute a x dx.
Theorem 223. Monotone functions on finite closed intervals are Riemann integrable.
Definition 62. Let S be a subset of R. Then f : S → R is called uniformly continuous
on S if for every ε > 0 there is a δ > 0 such that for all x, y ∈ S for which |x − y| < δ it is
true that |f (x) − f (y)| < ε.
6.2. PROPERTIES OF INTEGRALS 41
Theorem 224. Uniformly continuous functions on finite closed intervals are Riemann
integrable.
As a matter of fact, the qualifier “uniformly” is not necessary in the previous theorem,
because any function which is continuous on a finite closed interval is uniformly continuous
there. However, to prove this fact one needs the Heine-Borel theorem, which we have
not available. Instead, we will try for a different proof for the integrability of continuous
functions in Theorem 236 following an idea advocated by C. Bennewitz3.
The concept of the Riemann integral has proved to be too weak and too cumbersome for
many applications in differential equations, probability, and elsewhere. It is too weak since
many functions one might want to integrate are not Riemann integrable, e.g., Dirichlet’s
function. It is too cumbersome since it is sometimes difficult to show when limit processes
can be interchanged with integration (despite what we show in Section 6.5). To remedy
this Lebesgue4 introduced a more powerful integral called the Lebesgue integral. This is the
topic of more advanced courses in Real Analysis.
6.2. Properties of integrals

Theorem 225. Suppose f is Riemann integrable over [a, b] and [r, s] ⊂ [a, b]. Then f
is Riemann integrable over [r, s].
Theorem 226. Suppose f is bounded on [a, b] and c ∈ (a, b). Then
I± (f, a, c) + I± (f, c, b) = I± (f, a, b).
Rs
This theorem suggest we define I± (f, r, s) and r f also for the case where r ≥ s.
RDefinition 63. Suppose Rf : [a, b] →

R rR is a bounded function and r, s ∈ [a, b]. We
s s
set r f = 0 for r = s and r f = − s f for r > s. Similarly, I± (f, r, r) = 0, and
I± (f, r, s) = −I± (f, s, r).
Theorem 227. If f and g are Riemann integrable over [a, b], then so is f + g and
Rb Rb Rb
a
(f + g) = a f + a g.
Theorem 228. If f is Riemann integrable over [a, b] and α ∈ R, then αf is Riemann
Rb Rb
integrable over [a, b] and a αf = α a f .
The properties in Theorems 227 and 228 combine to establish linearity of the Riemann
Rb Rb Rb
integral: a (αf + βg) = α a f + β a g.
Theorem 229 (Monotonicity). Suppose f, g : [a, b] → R are Riemann integrable. If
Rb Rb
f (x) ≤ g(x) for all x ∈ [a, b], then a f ≤ a g.
Definition 64. Given a real-valued function f we define its positive and negative parts
f + and f − by f + (x) = max{f (x), 0} and f − (x) = − min{f (x), 0}.
Note that both f + and f − are non-negative functions.
Theorem 230. Suppose f is a real-valued function. Then f − = (−f )+ , f = f + − f − ,
and |f | = f + + f − .
Theorem 231. If f is Riemann integrable over [a, b], then so are its positive and negative
parts.
3Private communication and unpublished lecture notes
4Henri Lebesgue (1875 – 1941)
42 6. INTEGRATION
Theorem 232 (Triangle Inequality). Suppose f : [a, b] → R is Riemann integrable on

[a, b]. Then |f | is Riemann integrable over [a, b] and
Z
b Z b
f ≤ |f |.

a a
6.3. The fundamental theorem of calculus

Theorem 233. Suppose f is bounded on [a, b]. Then the functions F± : [a, b] → R :
x 7→ I± (f, a, x) are continuous on [a, b].
Theorem 234. Suppose f is bounded on [a, b]. Then the functions F± : [a, b] → R : x 7→
I± (f, a, x) are differentiable at any point c where f is continuous in which case F±0 (c) = f (c).
Definition 65. A function F : [a, b] → R is called a primitive or an antiderivative of

f : [a, b] → R if it is continuous and if F 0 = f on (a, b). The function F : [a, b] → R is called
a piecewise primitive or a piecewise antiderivative, if it is continuous and if there is a finite
subset D of [a, b] such that F 0 = f on [a, b] \ D.
Theorem 235. Suppose f : [a, b] → R is Riemann integrable over [a, b] and F : [a, b] →
R is a piecewise antiderivative of f . Then
Z b
f = F (b) − F (a).
a
The combination of Theorems 234 (for the case where F± coincide) and 235 is known as
the Fundamental Theorem of Calculus. Its essence is that the operations of integration and
differentiation are in some sense inverses of each other. But note that Theorem 234 makes
no statement as to the size of the set where F 0 does not exist while Theorem 235 requires
that set to be finite. This imbalance is a shortcoming of the Riemann integral and is only
overcome by the more powerful Lebesgue integral.
6.4. Integration of piecewise continuous functions

Definition 66. A function f : [a, b] → R is called piecewise continuous if there exists
a finite subset D of [a, b] such that f is continuous on [a, b] \ D.
Theorem 236. Bounded, piecewise continuous functions on finite closed intervals have
piecewise antiderivatives and are Riemann integrable.
Theorem 237 (Integration by parts). Suppose f, g : [a, b] → R are bounded and piece-
wise continuous on [a, b] and let F and G, respectively, be piecewise antiderivatives. Then
Z b Z b
Fg + f G = F (b)G(b) − F (a)G(a).
a a
Theorem 238 (Substitution rule). Suppose f : [a, b] → R is continuous on [a, b] and

that g : [a, b] → R is differentiable with a continuous derivative g 0 . Then
Z g(b) Z b
f= (f ◦ g)g 0 .
g(a) a
6.5. UNIFORM CONVERGENCE AND INTEGRATION 43
6.5. Uniform convergence and integration

Theorem 239. For n ∈ N define the functions fn : [0, 1] → R by
(
2n2 x if 0 ≤ x ≤ 1/n
fn (x) =
0 if 1/n < x ≤ 1.
R1
The sequence n 7→ fn converges pointwise to the function f = 0, but the sequence n 7→ 0
fn
converges to 1.
The operations of integration and taking limits may not always be interchanged.
Theorem 240. Suppose that n 7→ fn : [a, b] → R is a sequence of functions which
are Riemann integrable over [a, b] and that this sequence converges uniformly to a function
f : [a, b] → R. Then f is Riemann integrable and
Z b Z b
f = lim fn .
a n→∞ a
Theorem 241. Suppose that n 7→ gn : [a, b]P → R is a sequence of functions which are
∞
Riemann integrable over [a, b] and that the series n=0 gn converges uniformly to a function
g : [a, b] → R. Then g is Riemann integrable and
Z b ∞ Z b
X
g= gn .
a n=0 a
Theorem 242. For all x for which |x| < 1 we have

∞
X (−x)n
log(1 + x) = − .
n=1
n
In particular, the function x 7→ log(1 + x) defined on (−1, ∞) is analytic at zero.
CHAPTER 7
Special topics
7.1. Generalized limits

Definition 67. Let a be a real number. If r > 0, the set (a−r, a+r) = {x : |x−a| < r}
is called a neighborhood of a. The sets (c, ∞), where c ∈ R, are called neighborhoods of ∞
while the sets (−∞, c) are called neighborhoods of −∞. The set of all neighborhoods of
a ∈ R ∪ {−∞, ∞} is denoted by N (a).
Definition 68. We say that ∞ is a limit point of S ⊂ R if S is not bounded above.
Similarly, if S is not bounded below, we say that −∞ is a limit point of S.
Theorem 243. The point a ∈ R ∪ {−∞, ∞} is a limit point of S ⊂ R if and only if for
every neighborhood V of a we have V ∩ S \ {a} =
6 ∅.
Theorem 244. Suppose S ⊂ R, f : S → R, a ∈ R is a limit point of S, and b ∈ R.
Then the following statement is true: If
∀U ∈ N (b) : ∃V ∈ N (a) : ∀x ∈ S : x ∈ V \ {a} ⇒ f (x) ∈ U,
then f converges to b as x tends to a.
The same conclusion holds when S = N and a = ∞ (in which case f is a sequence).
Note that in place of ∀x ∈ S : x ∈ V \ {a} ⇒ f (x) ∈ U we might as well have written
∀x ∈ V ∩ S \ {a} : f (x) ∈ U . Because of the previous theorem we may now extend the
definition of convergence to include cases where a and/or b are in {−∞, ∞}.
Definition 69. Let S be a subset of R, f : S → R a function, and a ∈ R ∪ {−∞, ∞}
a limit point of S. We say that f converges to b ∈ R ∪ {−∞, ∞} as x tends to a if
∀U ∈ N (b) : ∃V ∈ N (a) : ∀x ∈ V ∩ S \ {a} : f (x) ∈ U.
Theorem 245. Suppose f : S → R converges to both b1 and b2 as x tends to a. Then
b1 = b2 .
Definition 70. If f : S → R converges to b as x tends to a, we call b the limit of f at
a and we write limx→a f (x) = b.
Theorem 246. limx→∞ e−x = 0.
Theorem 247. The function f : R → [0, ∞) defined by
( 2
e−1/x if x 6= 0
f (x) =
0 if x = 0
is continuous on all of R.
Theorem 248. Let (a, b) be an open interval (allowing for a = −∞ and b = ∞) and
let f, g : (a, b) → R be differentiable functions. Assume g 0 (x) 6= 0 for all x ∈ (a, b) and
45
46 7. SPECIAL TOPICS
limx→a f 0 (x)/g 0 (x) = L for some real number L. Then g is one-to-one on (a, b) and, for
every ε > 0, there is a point c ∈ (a, b) such that for all x, y ∈ (a, c) we have g(x) 6= 0 and
f (x) − f (y)
L−ε< <L+ε
g(x) − g(y)
as long as x 6= y.
Theorem 249 (L’Hôpital’s1 rule I). Let (a, b) be an open interval (allowing for a = −∞
and b = ∞) and let f, g : (a, b) → R be differentiable functions. Assume g 0 (x) 6= 0
for all x ∈ (a, b), limx→a f (x) = limx→a g(x) = 0, and limx→a f 0 (x)/g 0 (x) = L where
L ∈ R ∪ {−∞, ∞}. Then
f (x)
lim = L.
x→a g(x)
Theorem 250 (L’Hôpital’s rule II). Let (a, b) be an open interval (allowing for a = −∞
and b = ∞) and let f, g : (a, b) → R be differentiable functions. Assume g 0 (x) 6= 0
and g(x) > 0 for all x ∈ (a, b), limx→a g(x) = ∞, and limx→a f 0 (x)/g 0 (x) = L where
L ∈ R ∪ {−∞, ∞}. Then
f (x)
lim = L.
x→a g(x)
There are analogous theorems for the case where x tends to b.

The next two theorems compare the growth of positive powers with that of the loga-
rithmic and exponential functions.
Theorem 251. Suppose r is a positive real number. Then limx→∞ log(x)/xr = 0 and
limx→0 xr log(x) = 0.
Theorem 252. Suppose r is a real number. Then limx→∞ xr /ex = 0.
Theorem 253. The function f defined in Theorem 247 has derivatives of all orders on
all of R. In particular, f (n) (0) = 0 for every non-negative integer n.
Theorem 254. The function f defined in Theorem 247 has a Taylor series Tf at x0 = 0
with infinite radius of convergence. Tf (x) = f (x) holds for x = 0 but nowhere else.
This shows that analyticity of a function is an even stronger property than having
derivatives of any order.
7.2. Trigonometric functions and their inverses

Theorem 255 (Pythagoras2). (sin x)2 + (cos x)2 = 1 for all x ∈ R.
It follows immediately from this that | sin x| ≤ 1 and | cos x| ≤ 1.
Theorem 256 (Addition theorems). Let x, y ∈ R. Then
sin(x + y) = sin(x) cos(y) + cos(x) sin(y)
and
cos(x + y) = cos(x) cos(y) − sin(x) sin(y).
Theorem 257. cos has a unique zero in (0, 2).
1Guillaume François Antoine, Marquis de l’Hôpital (1661 – 1704)

2Pythagoras of Samos (ca. 570 – ca. 495 BC)
7.2. TRIGONOMETRIC FUNCTIONS AND THEIR INVERSES 47
Definition 71. The unique zero of cos in (0, 2) is denoted by π/2, i.e., π is twice that
zero.
Theorem 258. π > 0, cos(π/2) = 0, and sin(π/2) = 1.
Theorem 259. sin(x + π/2) = cos(x), sin(x + π) = − sin(x), and sin(x + 2π) = sin(x)
for all x ∈ R.
Theorem 260. sin(x) = 0 if and only if x = mπ with m ∈ Z.
Theorem 261. Suppose sin(x + a) = sin(x) for all x ∈ R. Then a is an integer multiple
of 2π.
Theorem 262. sin is strictly increasing on [−π/2, π/2]. Also, cos is strictly decreasing
on [0, π].
Definition 72. A function f : (a, b) → R is called convex on (a, b), if
f ((1 − t)x + ty) ≤ (1 − t)f (x) + tf (y)
for all x, y ∈ (a, b) and all t ∈ [0, 1]. It is called concave on (a, b), if −f is convex on (a, b).
Note that the graph of the function t 7→ (1 − t)f (x) + tf (y) describes a straight line
segment joining f (x) and f (y) in the coordinate plane. Convexity means that the graph of
a function lies below any such segment.
Theorem 263. f is convex on (a, b) if and only if
f (v) − f (u) f (w) − f (v)
≤
v−u w−v
whenever a < u < v < w < b.
Theorem 264. If f : (a, b) → R is twice differentiable and f 00 ≥ 0 on (a, b), then f is
convex on (a, b).
Theorem 265. The sine function is positive in (0, π) and negative in (π, 2π). It is
strictly increasing in (0, π/2) and (3π/2, 2π) and strictly decreasing in (π/2, 3π/2). Finally,
it is concave in (0, π) and convex in (π, 2π).
This theorem allows to draw a sketch of the sine function on [0, 2π] and by periodicity
on all of R. The graph of the cosine function is simply obtained by shifting the graph of the
sine function.
Theorem 266. tan is defined on R\{(2m+1)π/2 : m ∈ Z} and has range R. Moreover,
tan(x + π) = tan(x) for all x in its domain.
Theorem 267. The tangent function is negative and concave in (−π/2, 0) and positive
and convex in (0, π/2). It is strictly increasing in the entire interval (−π/2, π/2).
Definition 73 (Inverse trigonometric functions). arcsin, arccos, and arctan are the
inverse functions of sin |[−π/2,π/2] , cos |[0,π] , and tan |(−π/2,π/2) , respectively.
Theorem 268. arcsin : [−1, 1] → [−π/2, π/2]√is strictly increasing and continuous. It
is differentiable on (−1, 1). In fact arcsin0 (x) = 1/ 1 − x2 .
Theorem 269. arctan : R → (−π/2, π/2) is strictly increasing and differentiable. In
fact arctan0 (x) = 1/(1 + x2 ).
P∞
Theorem 270. If |x| < 1, then arctan(x) = n=0 (−1)n x2n+1 /(2n + 1).
Theorem 271. π/4 = arctan(1) = arctan(1/2) + arctan(1/3).
505 6115
Theorem 272. 162 ≤π≤ 1944 .
48 7. SPECIAL TOPICS
7.3. Analytic geometry

Theorem 273. The function
p
d : R2 × R2 → [0, ∞) : ((x1 , y1 ), (x2 , y2 )) 7→ (x1 − x2 )2 + (y1 − y2 )2
is a metric on R2 .
Definition 74. The function d defined in Theorem 273 is called the Euclidean metric
or Euclidean distance in R2 .
Definition 75. The set of all points in R2 which have a given distance r > 0 from a
given point p0 = (x0 , y0 ) is called a circle of radius r about p0 . The set of all points whose
distance from p0 is less than r is called the (open) disk of radius r about p0 . The circle of
radius 1 about the point (0, 0) is called the unit circle. The unit disk is the set of all points
whose distance from (0, 0) is less than 1.
Theorem 274. The unit circle is the set
{(x, y) ∈ R2 : x2 + y 2 = 1} = {(cos(t), sin(t)) : t ∈ [0, 2π]}.
Definition 76. A plane curve is an ordered pair (x, y) of continuous functions defined
on some interval [a, b].
Definition 77. The curve p2 : [a2 , b2 ] → R2 is called equivalent to the curve p1 :
[a1 , b1 ] → R2 if there exists a strictly increasing, continuous, and surjective function ϕ :
[a1 , b1 ] → [a2 , b2 ] such that p2 ◦ ϕ = p1 .
Theorem 275. Equivalence of curves is an equivalence relation, i.e., it has the following
three properties:
(1) Any curve is equivalent to itself (reflexivity).
(2) If the curve p2 is equivalent to the curve p1 , then p1 is equivalent to p2 (symmetry).
(3) If p2 is equivalent to p1 and p3 is equivalent to p2 , then p3 is equivalent to p1
(transitivity).
Definition 78. The curve (x, y) : [a, b] → R2 is called continuously differentiable if the
derivatives of x and y (the components of the curve) exist and are p continuous. The pair
(x0 , y 0 ) is then called the velocity of the curve and the function t 7→ x0 (t)2 + y 0 (t)2 is called
the speed of the curve.
Exercise 19. Note that, in view of Theorem 274 or a proper generalization of it, a
circle may be viewed as a continuously differentiable curve.
Definition 79. The length of the continuously differentiable curve (x, y) : [a, b] → R2
is defined as Z bp
x02 + y 02 .
a
Note that for constant speed v the length of a curve is v(b − a) (speed times time).
Theorem 276. Suppose two curves are continuously differentiable, equivalent to each
other, and that the function ϕ establishing the equivalence has a continuous derivative.
Then the curves have the same length.
The importance of this observation is that it shows length to be a property of (certain)
geometric objects in the plane rather than of functions with values in the plane.
Theorem 277. The length of the unit circle (its circumference) is 2π.
7.4. SETS OF MEASURE ZERO AND SOME CONSEQUENCES 49
Definition 80. Let f, F : [a, b] → R be two continuous functions with the property
f (x) ≤ F (x). Suppose S = {(x, y) ∈ R2 : a ≤ x ≤ b and f (x) ≤ y ≤ F (x)}. Then the area
of S is given by
Z b
A= (F − f ).
a
Exercise 20. Suppose f (x) = 0 and F (x) = h. What kind of set is S and what is its
area?
Theorem 278. The area of the unit disk is π.
7.4. Sets of measure zero and some consequences

Definition 81. A subset A of R is said to have measure
S∞ zero if for every
P∞ε > 0 there exist
bounded open intervals (an , bn ), n ∈ N, such that A ⊂ n=1 (an , bn ) and n=1 (bn −an ) < ε.
Theorem 279. The set of rational numbers has measure zero.
Theorem 280.* A countable union of sets of measure zero has itself measure zero.
Definition 82. For any function f : [a, b] → R and any n ∈ N let
D(f, n) = {x ∈ [a, b] : ∀δ > 0 : ∃y ∈ [a, b] : |y − x| < δ and |f (y) − f (x)| > 1/n}.
S∞
Theorem 281. If f : [a, b] → R, then n=1 D(f, n) is the set of all points for which f
fails to be continuous.
Theorem 282. Suppose f is a bounded real-valued function on [a, b], (c, d) ⊂ [a, b], and
D(f, n) ∩ (c, d) 6= ∅. Then sup{f (x) : c ≤ x ≤ d} − inf{f (x) : c ≤ x ≤ d} > 1/n.
Theorem 283. If f is Riemann integrable on [a, b], then D(f, n) is a set of measure
zero.
Theorem 284. If f is Riemann integrable on [a, b] the set of all points where f is not
continuous is a set of measure zero.
This theorem shows that, in some sense, the class of Riemann integrable functions is
quite small, emphasizing again the need for a different notion of integral. The converse of
this theorem is also true but the proof of this requires a little more preparation so we will
skip it.
Theorem 285.* Suppose that there are intervals (an , bn ), n ∈ N, such that [0, 1] ⊂
S∞ SN
n=1 (an , bn ). Then there is an N ∈ N such that [0, 1] ⊂ n=1 (an , bn ).
Theorem 286. The interval [0, 1] does not have measure zero.
Theorem 287. The set of real numbers is uncountable.
APPENDIX A
Some set theory and logic
In this section we briefly recall some notation and a few facts from set theory and logic1.
Some of what is said below is intentionally vague lest we should write a book on the matter.
A.1. Elements of logic

A statement is a sentence for which it can (in principle) be determined whether it is
true or false (“true” and “false” are called truth values). Statements may be connected to
form new statements. If p and q are statements we may form the statement “p and q”,
the statement “p or q”, and the statement “not p” (negation). Very important is also the
connective “implies”: “p implies q” is an abbreviation for “q or not p” (this entails no logical
implication). Propositional logic deals with determining the truth values of such compound
statements given the status of their components.
Matters become more interesting when the internal structure of statements is also con-
sidered. The simplest statements with internal structure are those where something (the
predicate) is said about something else (the subject). For instance, the sentences “London
is a city.” and “Rome is the capital of Georgia.” are such statements. Next one introduces
variables, i.e., symbols meant to represent something unspecified. A formula is a sentence
containing a variable, which becomes a statement after specifying the variable. For instance,
in the sentence “x2 + 16 = 25.” x represents a number but we do not specify which one. At
this point logic becomes interwoven with set theory since, in general, we need to make clear
what kind of thing a variable is.
There are two more important ways of turning formulas into statements. If F (x) is
a formula involving the variable x we write ∀x : F (x) to mean that F (x) becomes a true
statement whenever x is replaced by any specific value. We write ∃x : F (x) to mean that
there is some specific value, say v, such that F (v) is a true statement. The negation of the
statement “∀x : F (x)” is the statement “∃x : not F (x)” and the negation of “∃x : F (x)’ is
“∀x : not F (x)”.
Both ∀x : F (x) and ∃x : F (x) are statements despite the occurrence of a variable. Such
variables are called bound . For example, the statement “ Every state has a capital.” would
be formalized as ∀x : ∃y : y is the capital of x. The variables are there only for convenience;
notice that the plain English version does not even make explicit use of them. A variable
which is not bound is called free. A sentence containing a P free variable is a formula not a
n
statement. To give another example consider the formula k=1 k 2 = n2 + n/2. Here k is
aP bound variable but n is a free variable. One may want to determine the truth value of
n
“ k=1 k 2 = n2 + n/2” for a specific n; in contrast, it makes no sense to specify k.
1For more details you may check my lecture notes on Algebra at http://www.math.uab.edu/
~rudi/teaching/algebra.pdf and the books referenced there.
51
52 A. SOME SET THEORY AND LOGIC
A.2. Basics of set theory

We will not formally define the terms set and element but we think of a set (or a
collection or a family) as a single entity which collects a variety of other entities, the elements
(or members) of the set. A little more precisely, we assume that a set determines its elements
and vice versa. In particular, two sets A and B are equal if and only if they contain precisely
the same elements. We will use the following phrases: “x is an element (or a member) of
(the set or collection) A”, “x belongs to A”, or “A contains x” and this relationship is
denoted by x ∈ A. Otherwise, if x does not belong to A, we write x 6∈ A. The set with no
elements is called the empty set and is denoted by ∅.
There are essentially two ways to specify a set. Firstly, one can list all the elements
of a set, e.g., {2, 3, 5} is the set containing the numbers 2, 3, and 5. Secondly, a set might
collect elements which all share a certain property or certain properties. For instance,
{x : x is a city} denotes the set of all cities.
If A and B are sets and every element of B is an element of A, we say that B is a subset
of A or that A includes B, and we write B ⊂ A or A ⊃ B. So, to check if B ⊂ A one picks
an arbitrary element x ∈ B and shows that it is in A. The meaning of ‘arbitrary’ here is
simply that the only fact we use about x is that it is in B. In this way we have checked the
inclusion simultaneously for all x ∈ B. A subset B of A is called a proper subset of A if it
is different from A.
Let F be a non-empty collection of sets. Then we define the intersection of all sets in
F by
\
A = {x : x ∈ A for every A ∈ F} = {x : (∀A ∈ F : x ∈ A)}.
A∈F
If F = {A, B} we denote their intersection simply by A ∩ B. If A ∩ B = ∅ we say that A

and B are disjoint. Similarly, the union of all sets in F is defined by
[
A = {x : x ∈ A for some A ∈ F} = {x : (∃A ∈ F : x ∈ A)}.
A∈F
The union of two sets A and B is denoted by A ∪ B.
Theorem A.1. Let A, B and C be sets. Then the following statements hold:
(1) Commutative laws:
A∪B =B∪A and A ∩ B = B ∩ A.
(2) Associative laws:
(A ∪ B) ∪ C = A ∪ (B ∪ C) and (A ∩ B) ∩ C = A ∩ (B ∩ C).
(3) Distributive laws:
\ \
A∪( X) = (A ∪ X) and
X∈F X∈F
[ [
A∩( X) = (A ∩ X).
X∈F X∈F
The set X \ A which consists of all the elements of X which do not belong to A, i.e.,
x ∈ X \ A if and only if x ∈ X and x 6∈ A. If A is a subset of X, then X \ A is called the
complement of A in X. If it is clear what X is, then the complement is also denoted Ac .
A.3. FUNCTIONS 53
Theorem A.2 (De Morgan’s laws). Let E be a set and suppose that F is a non-empty
collection of subsets of E. Then De Morgan’s laws hold:
!c
[ \
X = X c,
X∈F X∈F
!c
\ [
X = X c.
X∈F X∈F
These formulas can be expressed concisely, if not precisely, as follows: The complement
of a union is the intersection of the complements and the complement of an intersection is
the union of the complements.
A.3. Functions
Let X and Y be sets. An ordered pair is an object of the form (x, y) where x ∈ X and
y ∈ Y . The set of all such ordered pairs (x, y) is called the Cartesian product of X and Y ,
and is written X × Y . A subset f of X × Y is a function, if for each x ∈ X there is a unique
y ∈ Y such that (x, y) ∈ f . As the concept of function is extremely important we repeat
the preceding statement in a more formal way: f ⊂ X × Y is a function if and only if
(i) ∀x ∈ X : ∃y ∈ Y such that (x, y) ∈ f and
(ii) if (x, y1 ) ∈ f and (x, y2 ) ∈ f , then y1 = y2 .
The more conventional notation for functions is y = f (x) instead of (x, y) ∈ f . If f ⊂ X × Y
is a function, we write f : X → Y or, to define it precisely, f : X → Y : x 7→ y = f (x), e.g.,
f : R → [0, 1] : x 7→ 1/(1 + x2 ). We call X the domain of f , Y the codomain or target of f ,
and say that f maps X to Y . The element f (x) ∈ Y is called the image of x under f . The
set of those y ∈ Y for which there is an x ∈ X such that y = f (x) is called the range or
image of f , and will be written as ran(f ). Let f : X → Y . If A ⊂ X, we define the function
f |A : A → Y , called the restriction of f to A, by setting (f |A )(x) = f (x) for every x ∈ A.
We define the image of A under f as the set
f (A) = {y ∈ Y : ∃x ∈ A : f (x) = y} = {f (x) : x ∈ A}.
In other words f (A) = ran(f |A ).
Theorem A.3. Let f : X → Y and A, B ⊂ X. Then f (A ∪ B) = f (A) ∪ f (B) and
f (A ∩ B) ⊂ f (A) ∩ f (B).
Equality does not always hold in the second case. Can you think of an example where
it does not?
If A ⊂ Y , we define the pre-image of A under f , denoted by f −1 (A) ⊂ X, by:
f −1 (A) = {x ∈ X : f (x) ∈ A}.
Theorem A.4. Let f : X → Y and A, B ⊂ Y . Then f −1 (A ∪ B) = f −1 (A) ∪ f −1 (B)
and f −1 (A ∩ B) = f −1 (A) ∩ f −1 (B).
A function f : X → Y is onto (or surjective) if ran(f ) = Y . A function f : X → Y is
one-to-one (or injective) if f (x1 ) = f (x2 ) implies that x1 = x2 . If f is both one-to-one and
onto, we say that f is bijective. A bijective function from a finite set X to itself is called a
permutation of X. A permutation π is called a transposition if π(x) = x for all but at most
two elements x of the domain of π.
Suppose now that f : X → Y is an injective function. Then there exists a unique
function g : f (X) → X such that g(f (x)) = x for all x ∈ X, and f (g(y)) = y for all
54 A. SOME SET THEORY AND LOGIC
y ∈ f (X). The function g is denoted f −1 , and is called the inverse function of f . Note that
notation is abused here in the sense that f −1 is used to describe two different things. Only
the context tells which of the two meanings is used. In any case, caution is advised. These
two different meanings are exemplified in
f −1 ({y}) = {f −1 (y)}
which holds for all y ∈ Y if f : X → Y is bijective.
Given a function f : X → Y and a function g : Y → Z their composition g ◦ f is the
function from X to Z defined by
(g ◦ f )(x) = g(f (x)).
Theorem A.5. The inverse of an injective function is injective. The composition of
injective functions is injective and the composition of surjective functions is surjective.
A.4. The recursion theorem

The following theorem states that it is ok to make recursive definitions. It is not strictly
a part of set theory but requires also a proper definition of the set of natural numbers.
Theorem A.6 (The Recursion Theorem). Let X be a non-empty set, f a function from
X to X, and x1 an element of X. Then there is one and only one function u : N → X such
that u(1) = x1 and u(n + 1) = f (u(n)) for every n ∈ N.
Index
arccos, 47 uniform, 30
arcsin, 47 convex, 47
arctan, 47 countable, 10
countably infinite, 10
analytic, 36 curve
antiderivative, 42 plane, 48
piecewise, 42
area, 49 de Morgan’s laws, 53
associative law of addition, 5 degree, 23
associative law of multiplication, 5 dense, 12
derivative, 33
bijective, 53 higher order, 34
binary operation, 5 second order, 34
binomial coefficients, 20 differentiable, 33
Bolzano-Weierstrass theorem, 17 twice, 34
bound variable, 51 disjoint, 52
bounded, 11 disk, 48
sequence, 15 unit, 48
bounded above, 11 distance, 13
bounded below, 11 distance function, 13
divergence
Cartesian product, 53 of a sequence, 15
Cauchy sequence, 17 of a series, 19
circle, 48 domain, 53
unit, 48
codomain, 53 element, 52
commutative law of addition, 5 empty set, 52
commutative law of multiplication, 5 Euclidean metric, 48
complement, 52 extremum
complete, 17 local, 35
composition, 54
concave, 47 factorial, 8
continuity formula, 51
uniform, 40 free variable, 51
continuous function, 53
at a, 28 algebraic, 23
on S, 28 cosine, 25
convergence exponential, 24
absolute, 21 logarithmic, 24
conditional, 21 polynomial, 23
of a function, 27 rational, 23
of a sequence, 15 sine, 25
of a series, 19 tangent, 25
pointwise, 29 functions
55
56 INDEX
trigonometric, 25 sequence, 16
Fundamental Theorem of Calculus, 42 non-increasing
function, 23
greatest lower bound, 11 sequence, 16
non-negative, 7
identity, 5 non-positive, 7
additive, 5 number
multiplicative, 5 natural, 8
image rational, 8
of a function, 53 whole, 8
of a point under a function, 53
of a set under a function, 53 one-to-one, 53
inclusion, 52 onto, 53
inductive, 8 ordered pair, 53
infimum, 11
injective, 53 part
integer, 8 negative, 41
intersection, 52 positive, 41
interval partial sums, 18
closed, 8 partition, 39
finite, 7 permutation, 53
infinite, 8 polynomial, 23
open, 8 positive, 7
inverse, 5 power, 24
additive, 5 integer, 9
multiplicative, 5 power series, 30
inverse function, 54 pre-image, 53
isolated point, 27 predecessor, 8
predicate, 51
least upper bound, 11 primitive, 42
length piecewise, 42
of a curve, 48 product, 5
limit, 15
of a function, 27 radius, 48
limit inferior, 16 radius of convergence, 30
limit point, 27 range, 53
limit superior, 16 rational number, 8
linearity, 41 rearrangement, 21
lower bound, 11 reciprocal, 6
lower Riemann integral, 40 refinement, 39
lower Riemann sum, 39 common, 39
lowest terms, 10 restriction, 53
Riemann integrable, 40
maximum, 11
local, 35 sequence, 15, 29
measure zero, 49 sequence of partial sums, 18
metric, 13 series, 18
minimum, 11 alternating harmonic, 19
local, 35 geometric, 19
monotone, 23 harmonic, 19
strictly, 23 set, 52
speed, 48
natural numbers, 8 statement, 51
negation, 51 strictly decreasing
negative number, 7 function, 23
negative of a number, 6 sequence, 16
neighborhood, 45 strictly increasing
non-decreasing function, 23
function, 23 sequence, 16
INDEX 57
subject, 51
subsequence, 17
subset, 52
proper, 52
successor, 8
sum, 5
supremum, 11
surjective, 53
target, 53
Taylor series, 36
telescoping sum, 9
transcendental, 23
transposition, 53
triangle inequality
for a metric, 13
for the absolute value, 13
truth value, 51
uncountable, 10
union, 52
unit circle, 48
unit disk, 48
upper bound, 11
upper Riemann integral, 40
upper Riemann sum, 39
variable, 51
velocity, 48
whole numbers, 8

Advanced Calculus: Lecture Notes For MA 440/540 & 441/541 2015/16

Uploaded by

Copyright:

Available Formats

Advanced Calculus: Lecture Notes For MA 440/540 & 441/541 2015/16

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Advanced Calculus: Lecture Notes For MA 440/540 & 441/541 2015/16

Uploaded by

Copyright:

Available Formats

ADVANCED CALCULUS

Lecture notes for

Based on lecture notes by G. Stolz and G. Weinstein

First things first 1

Chapter 1. The real numbers 5

Chapter 2. Sequences and series 15

Chapter 3. A zoo of functions 23

Chapter 7. Special topics 45

7.3. Analytic geometry 48

0.1. The goals

0.2. The rules

0.4. The language of mathematics

The real numbers

1.1. Field axioms

1.2. Order axioms

are called infinite intervals: (−∞, a) = {x ∈ R : x < a}, (−∞, a] = {x ∈ R : x ≤ a},

1.3. The induction principle

when c and the ak are real numbers.

Theorem 33. If the ak are real numbers and n ∈ N, then

Theorem 37 (Bernoulli’s1 inequality). If a > −1 and n ∈ N, then (1 + a)n ≥ 1 + na.

Theorem 39. If a 6= 0 and m, n ∈ Z, then am an = am+n and (am )n = amn . Moreover,

1Jacob Bernoulli (1654 – 1705)

Theorem 42. There is no rational number r for which r2 = 2.

1.4. Counting and infinity

Definition 11. A set X is called finite if it is empty or if there exists an n ∈ N and a

Theorem 43. #Zn ≤ n.

Theorem 44. #Zn = n.

Theorem 46. N is infinite.

Definition 12. A set X is called countable if it is empty or if there exists a surjective

Theorem 48. Every finite set is countable and N is countably infinite.

Theorem 49.* Z, N × N, and Q are countably infinite.

2This number exists by Theorem 40.

1.5. The least upper bound axiom

(2) If S is bounded below, then T is bounded below and inf T ≥ inf S.

Sequences and series

Theorem 74. If a sequence converges, then it is bounded.

Theorem 95. The inequalities in Theorem 94 may be strict.

Definition 27. A sequence x is called a Cauchy3 sequence or Cauchy for short if

2.2. Sums and the Σ-notation

P Definition Pn29. If A is a finite set and f a function from A to R, then we define

Definition 31. Suppose x is a sequence ofPreal numbers. If the sequence s of partial

Of course, a series (i.e., a sequence of partial sums) may also

Definition 33. For integers n and k with n ≥ k ≥ 0 let

4Joseph Ludwig Raabe (1801 – 1859)

Every rational function is algebraic (choose n = 1). Other

is called the exponential function.

Definition 45. The function sin : R → R defined by

4.1. Limits of functions

Theorem 151. Let S ⊂ R, a ∈ R a limit point of S, and f : S → R and g : S → R

4.2. Continuous functions

Theorem 161. Let S1 and S2 be subsets of R and consider the functions f : S1 → S2

4.3. The intermediate value theorem and some of its consequences

4.4. Uniform convergence and continuity

Definition 51. Let S ⊂ R and n 7→ fn a sequence of functions on S. We say that

be given. If A is bounded the number sup A is called the radius of convergence of S. If A

1Isaac Newton (1643 – 1727)

Theorem 188 (Quotient rule for derivatives). Let f : S → R and g : S → R be

Theorem 190. Any rational function is differentiable on its domain. In particular, if

g 0 (x) = −nx−n−1 for x ∈ R \ {0}.