ANALYSIS

Analysis can briefly be described as the study of infinite processes, particularly as found
in calculus and related subjects. We start with an introduction to the first principles of
the subject, namely the axioms of the real number system. Thereafter we study the deep
concepts of sequences, continuity, integration, and series. All of these concepts involve infinite
processes of one kind or another.

ORGANIZATION

1. Textbook
Haggarty, Rod. Fundamentals of Mathematical Analysis (Second edition), Prentice
Hall/Pearson Education, 1993.
2. Lecturers
(a) Dr Mokhwetha Mabula (Module coordinator)
Botany Building 2-5; Tel: +27 12 420-6818; Email: mokhwetha.mabula@up.ac.za
(b) Dr Marten Wortel
Botany Building 2-11; Tel: +27 12 420-2525; Email: marten.wortel@up.ac.za
3. Lectures
See the 2022 timetable or view on clickUP.
4. Tutorial classes
These are compulsory. Unless announced otherwise, you should prepare the exercises
of the study units that were covered in class during the week preceding the tutorial.
You will register yourself to one of the tutorial groups some time during the first
or second week of the semester. You may not attend any other tutorial group without
the permission of the coordinator or the lecturer/tutor for your assigned group. Tuto-
rial classes provides more opportunity for you to understand your material and to ask
questions where you do not understand. Note that Tutors and Lecturers will give
you guidance on how to solve problems not to solve problems for you.
5. Consulting hours
Hours of consultation with lecturers will be displayed on their office doors (and on
clickUP). Students may consult lecturers only during the consulting hours as indicated,
or by appointment. This policy also holds before tests and examinations. In other
words, lecturers are only available during their normal consulting hours on the day
before a test. This policy aims at encouraging students to plan their work and to work
continuously.
6. Announcements
Announcements may be made in class and/or on clickUP.

1
 ASSESSMENT AND RELATED MATTERS

The examination and test instructions in the yearbook must be followed meticulously.

1. Material for tests and the exam

Material for the semester tests and the exam will be posted on clickUP. Announcements
may also be made in class.
In the tests and exam, there will be a mixture of questions designed to test
(a) your knowledge of basic ideas, definitions, and theory
(b) your ability to prove specified results, and
(c) the depth of your understanding of the module material by proving new results
and/or solving new problems.
Due to the nature of the subject, it is policy that no partial marks are given for defini-
tions that are mathematically wrong, even if 90% of the correct words and symbols are
given in the answer.

2. Absence from tests You must notify the lecturer within three days of the class or
semester test if you were absent due to illness or some other valid reason. You must
present convincing proof of the reason for your absence, for example medical certificate.
In the case of the exam, the relevant faculty office should be informed of the absence.

There will be no ”sick tests” for class tests and semester tests. If you miss a class test
for a valid reason, then that test will excluded when your total mark is calculated. if you
miss semester test ( and you have acceptable excuse) then you may write the special
test at a time that will be announced. It will cover the work of both semester tests.
Note that this will only replace one semester test. Hence if you miss both semester
tests, your special test will only replace one of them. The other one you will get zero
for it.
3. Absence from semester tests and special test
Absence from a semester test, with a valid reason, must be reported to the module coor-
dinator within three working days. Supporting documentation (for example a medical
certificate) must be presented. Failure to adhere to these rules may result in a mark of
zero being awarded for the test that was not written.
Students who do not write one of the semester tests, and have a valid excuse for this,
may write the special test. The dates for this will be announced on click-up.
The scope of the special test is the combined scopes of both semester tests. The mark of
the special test will be substituted for the mark of the semester test that was not written.
The mark of the special test may be substituted for that of at most one semester test
that was not written (with a valid reason).
• Valid original sick notes are accepted if issued by a medical doctor registered at
the Health Professions Council of South Africa (HPCSA). The only other type
of sick note that is accepted are those issued by an Advanced Practice Nurse (a
registered nurse with a postgraduate qualification) as determined by the South
African Nursing Council who has a BHCF practice number, provided that the
diagnosis falls only within their specific field of specialisation.
• An affidavit will only be accepted if supported by substantiating documentation,
e.g. case report or criminal charge with case number obtained from a police station,
valid medical certificate for injuries, a death certificate for a funeral, etc. Please
note that submission of fraudulent sick notes and affidavits is a criminal offense,
which will lead to disciplinary action and may result in dismissal.

2
 4. Semester test dates
The semester test dates will be available on the UP Website and will also be announced
in class and/or on ClickUP.

5. Tutorial classes and class tests

By and large the material for the weekly tutorial will be exercises as listed in this study
guide, on the class work of the preceding week. However, some additional problems
maybe given. Class tests may combine work from more than one tutorial.
Class tests date will be announced on click-up.

6. Marked tests and assignments

File away safely all your assignments, class tests and semester tests after receiving them.
They are the only proof that you have if, for example, you claim that there is an error
in our record of your marks.

7. Memo or Solutions
Only partial memo/solutions for assignments, class tests, tutorials and semester tests
will be provided for selected questions not for all questions.

8. Semester mark
Your semester mark will be computed as follows:
Semester test 1 : 35% Semester test 2: 35%
Class Tests: and Assignments: 30%
Remark: There is a total of four class tests and one assignment for the semester and all
will be marked and counted for your semester mark.

9. Final examination
A minimum semester mark of 40% is required for admission to the exam. The date will
be announced on ClickUP.
Exam timetable clashes or absence from the exam should be reported to your faculty
administration, not to your lecturer.

10. Final mark

Your final mark will be computed as follows:
Semester mark: 40% Examination mark: 60%
To pass the module, a final mark of at least 50% is required together with and a sub-
minimum of 40% for the examination.

11. Supplementary examination

A supplementary examination is granted under the following conditions
(a) a final mark of between 40% and 49% has been obtained and either the examination
or semester mark is at least 50%; or
(b) a pass mark has been obtained, but the required sub-minimum in the examination
has not been obtained.
If a student who qualifies to write the supplementary exam does not write the supple-
mentary exam for any reason whatsoever, or fails the module after having written the
supplementary exam, then that student fails the module.

CONTACT TIME AND STUDY HOURS

3
This module carries a weight of 12 credits, indicating that, on average, a student should
spend some 120 hours to master the required skills (including time for preparation for tests
and the examination). This means that, on average, you should devote some 8.5 hours of
study time per week to this module. The scheduled contact time is approximately 3.5 hours
per week, which means that another 5 hours per week of own study time should be
devoted to the module.
It is not possible in a course of this nature for the lecturer to give detailed proofs or expla-
nation of every result. You will have to learn some of the material on your own. Of course,
the lecturers are available for consultation in case of any difficulty.

DISCIPLINARY ISSUES

The policy of the Department is to refer all incidents in which there is a suspicion of dis-
honesty or other irregularity to the Disciplinary Committee of the University. In the recent
past, some students have had to leave the University on account of their dishonest behaviour
in tests in the Department of Mathematics and Applied Mathematics.

4
 STUDY THEMES

THEME 0: FOUNDATIONAL ISSUES

Unit 0.1: Implication and proof 1 lectures

Source: Haggarty, Section 1.1
Learning objectives:
On completion of this study unit you should be able to

1. state what it means for two statements (statement forms, to be more precise) to be
logically equivalent (self-study and will be asked in the assignments class and
semester tests, and in the exam).
2. write down the truth table that defines the composite statement P =⇒ Q (self-study
will be asked in the assignments class and semester tests, and in the exam).
3. define both the converse and the contra-positive of the composite statement P =⇒ Q,
and know (and prove) which of these are logically equivalent (self-study and will be
asked in the assignments class and semester tests, and in the exam).
4. prove statements such as in Example 3: (i) directly, (ii) indirectly by proving the contra-
positive, and (iii) with a proof by contradiction.

Problems: Haggarty, Exercises 1.1 (p. 24) no. 4–5

Problems 1 (p. 37) no. 2(a,b), 3(a,b)
√
Note: For no. 3(a): change the question to “... is less than or equal to n.” Can you
show that without the “or equal to” that the statement is false?

6
 THEME 1: THE REAL NUMBERS

Note: At this stage of the course it is essential to learn to give complete proofs. This means
that every step should use valid logic, combinining previously proved conclusion(s) and/or
axiom(s) to form a new conclusion. (See Haggarty, Section 1.1 for valid logic.)
In later lecture units, when we are more familiar with the axioms of the real number system,
we may skip small steps. However, even in such cases, we may skip steps only when it is
easy, for anyone familiar with the axioms and their consequences, to fill in the gaps.

Unit 1.1: The arithmetic and order axioms of R 1 12 lectures

Source: Haggarty, Section 2.1 (pp. 41–44) and Section 2.2 (pp. 50–57)
Learning objectives:
On completion of this study unit you should be able to

1. explain the purpose of the arithmetic and order axioms of R.

2. write down at least three of the arithmetic axioms and two of the order axioms of R.
3. write down at least two arithmetic and one order property of R that follow from the
axioms.
4. use the order properties of R to solve inequalities.
5. define the minimum (or least element) and the maximum (or greatest element) of a
set S.
6. define the absolute value of a real number.
7. state and prove the triangle inequality.

Remark: Pages 45–49 of Section 2.1 are excluded.

Problems: Haggarty, Exercises 2.2 (p. 58) no. 1–4

7
Unit 1.2: Completeness Axiom 2 12 lectures
Source: Haggarty, Section 2.2.3 and Section 2.3
Learning objectives:
On completion of this study unit you should be able to

1. define the concepts of upper bound, lower bound, least upper bound (supremum), and
greatest lower bound (infimum) of a set.
2. state and prove alternative formulations for supremum and infimum.
3. determine whether a given set is bounded from below or bounded from above.
4. find the supremum and the infimum of a given set (if they exist).
5. state the completeness axiom of R.
6. prove the Archimedean postulate: The set N of natural numbers does not have an upper
bound.
7. prove that Q is dense in R.

Problems: Haggarty, Exercises 2.2 (p. 58) no. 5

Exercises 2.3 (p. 64) no. 2(a–c), 4
Problems 2 (pp. 65–66) no. 4(a), 6, 7(d), 8

Additional problems:

1. Let A and B be any non-empty bounded subsets of R. Prove the following statements:
(a) If c is any non-negative real number and C = {ca : a ∈ A} then sup(C) = c sup(A).
(b) If C = {x + y : x ∈ A, y ∈ B} then sup(C) = sup(A) + sup(B).
(c) If C = {−a : a ∈ A} then sup(C) = − inf(A).
2. If A ⊆ B and A and B are bounded above, what can one say about sup(A) and sup(B)?
3. If a non-empty subset S of R has an upper bound M ∈ S, prove that sup(S) = M .
4. Say in your own words what the connection is between the maximum of a set S and
the supremum of S.
5. Let S be a non-empty bounded subset of R. Prove that sup(S) = ` if and only if
(a) ` is an upper bound of S, and
(b) for every real number ε > 0 there exists an element x ∈ S such that x > ` − ε.

8
 THEME 2: SEQUENCES

Unit 2.1: Sequences and limits 2 12 lectures

Source: Haggarty, Section 3.1–3.3.4, Appendix to Theme 2, and class notes.
Learning objectives:
On completion of this study unit you should be able to
1. write down the definition of a sequence as a function from N to R.
2. define the concepts of convergence and divergence of a sequence of real numbers.
3. prove that a sequence can have at most one limit.
4. prove that a convergent sequence is bounded.
5. prove the convergence of specific sequences.
6. know and apply the rules 3.1.2, the sandwich rule 3.1.3, and the composite rule 3.1.4.
7. state the limits of certain basic sequences and use it to find and prove the limits of other
sequences.
8. know what a null sequence is.
9. be able to prove and apply the power rule.
10. know the basic null sequences of Theorem 3.2.2.
11. know what it means for a sequence to tend to infinity.
12. know and apply the reciprocal rule 3.3.4 (p. 87).

Problems: Haggarty, Exercises 3.1 (pp. 78–79) no. 1(b), 2, 3(a), 3(c), 4(a)
Exercises 3.2 (pp. 83–84) no. 1, 2(a), 3, 4(b), 5
Exercises 3.3 (pp. 89–90) no. 1(b), 2
Problems 3 (pp. 98–100) no. 1(a), 2, 3(a,b,e), 4(a), 5(a–c)

Unit 2.2: Monotone sequences 1 12 lectures

Source: Haggarty, Section 3.4.1, Appendix to Theme 2, and class notes.
Learning objectives:
On completion of this study unit you should be able to

1. write down the definitions of an increasing sequence, a decreasing sequence, and a

monotone sequence.
2. state and prove the monotone convergence theorem.
3. use the monotone convergence theorem to prove the convergence of specific sequences.
4. define the concepts: countable and uncountable sets.
5. give a demonstration that the set Q of rational numbers is countable by using a diagonal
argument.
6. use a repeated interval division argument to prove that the interval [0, 1] is uncountable.
7. explain how the above-mentioned result implies that R is uncountable.

Problems: Haggarty, Exercises 3.4 (pp. 97–98) no. 1(a–d), 2(a–d)

Problems 3 (pp. 98–100) no. 6(a–e), 8

9
Unit 2.3: Subsequences 2 lectures
Source: Haggarty, Section 3.3.5 and 3.4.2
Learning objectives:
On completion of this study unit you should be able to

1. write down the definition of a subsequence.

2. prove that (a) every subsequence of a convergent sequence of real numbers is convergent;
and (b) if a sequence diverges to infinity, then every subsequence of it also diverges to
infinity.
3. use subsequences to prove the divergence of specific sequences.
4. state and prove the Bolzano-Weierstrass Theorem.
5. Cauchy sequences and the convergence of Cauchy sequences

Problems: Haggarty, Exercises 3.3 (pp. 89–90) no. 3(a–c), 4

Problems 3 (pp. 98–100) no. 7, 9

10
 THEME 3: CONTINUOUS FUNCTIONS

Note: Special attention should be paid to the ε − δ definition, or characterization, of conti-

nuity. Students neglecting this will have difficulties with all the units of this theme, because
this definition is used so many times in proofs and arguments.

Unit 3.1: Limits and Continuity 2 lectures (partially revision)

Source: Haggarty, Section 5.1 and 5.2
Learning objectives:
On completion of this study unit you should

1. be able to write down the definition of a limit of a function f (x) as x tends to infinity,
and the definition of a limit of a function f (x) as x tends to a point a ∈ R.
2. be able to write down the ε − δ definitions of the left-hand and right-hand limits.
3. know and be able to use the limit laws 5.1.3 to 5.1.6.
4. formulate the definition of continuity of a function at a point c.
5. formulate the ε − δ definition of continuity of a function at a point c.
6. be able use the ε − δ definition of continuity to prove that certain basic functions are
continuous, and that certain other functions are not continuous.
7. be able to write down both the limit-formulation and the ε − δ formulation of left-
continuity and right-continuity of a function f (x) at a point c.
8. know and be able to prove that a function can have at most one limit at a point.
9. know and be able to prove that if f is continuous at a, and (xn ) converges to a, then
(f (xn )) converges to f (a).

Problems: Exercises 5.1 (p. 144) no. 2, 3, 6

Exercises 5.2 (pp. 152–153) no. 2, 3, 4(b)
Problems 5 (pp. 163–164) no. 6, 7

11
Unit 3.2: Boundedness of Continuous Functions 2 lectures
Source: Haggarty, Section 5.3.1
Learning objectives:
On completion of this study unit you should

1. know and be able to prove that a continuous function on a closed and bounded interval
is bounded from above and from below.
2. know and be able to prove that a continuous function on a closed and bounded interval
attains its minimum and its maximum values on the interval.

Problems:

(1) Consider an interval I = [a, b] and a function f which is continuous on I. Suppose that
f is unbounded on I.
Let c denote the midpoint of the interval [a, b]. Since f is unbounded on [a, b], f is
unbounded on at least one of the intervals [a, c] or [c, b]. Let I1 = [a1 , b1 ] be that half
of [a, b] on which f is unbounded. If f is unbounded on both halves, let [a1 , b1 ] denote
the left half.
Continue this process repeatedly, so that for each n ∈ N, In+1 = [an+1 , bn+1 ] is that
half of the interval In = [an , bn ] on which f is unbounded. If f is unbounded on both
halves of the interval In , then let In+1 be the left half.
(a) Find an expression for the length of each interval In .
(b) Explain why the set
A = {an : n ∈ N}.
has a supremum and that d = sup A belongs to the interval [a, b].
(c) Show that the sequence (an ) is increasing.
(d) Prove that there are numbers K and L, and a number δ > 0 such that, for every
x ∈ [a, b]
|x − d| < δ ⇒ K < f (x) < L.
(e) Show that there is a number n ∈ N such that In ⊂ (d − δ, d + δ).
(f) Explain why we have now arrived at a contradiction.
(2) Consider an interval I = [a, b] and a function f that is continuous on I.
Assume that f is not bounded on I.
(a) Use the fact that f is not bounded to prove that there is a sequence (xn ) in [a, b]
for which |f (xn )| > n for each n ≥ 1.
(b) Now show that our assumption that f is not bounded on I leads to a contradiction,
concluding that f is in fact bounded on I.

12
Unit 3.3: The Intermediate Value Theorem 2 12 lectures
Source: Haggarty, Sections 5.2.7 and 5.3
Learning objectives:
On completion of this study unit you should

1. know and be able to prove the intermediate value theorem.

2. be able to use the intermediate value theorem.
3. know and be able to prove that a continuous function maps intervals onto intervals.
4. know that if a continuous function on an interval is invertible then the inverse is also
continuous (p. 151).

Problems: Haggarty, Exercises 5.3 (p. 162) no. 1, 2, 3

Problems 5 (pp. 163–164) no. 8, 9(d), 10

Additional problems:

1. Let I = [a, b] be an interval, and suppose that f is continuous on I such that f (a) <
γ < f (b). Assume that f (x) 6= γ for all x ∈ I.
Let c1 be the midpoint of the interval I. If f (c1 ) < γ, let I1 be the right half of I. If
f (c1 ) > γ, let I1 be the left half of I.
Continuous this process repeatedly, so that cn denotes the midpoint of the interval In .
If f (cn ) < γ, let In+1 be the right half of In , and if f (cn ) > γ, we let In+1 be the left
half of In .
b−a
Let In = [an , bn ]. Note that the length of In is .
2n
(a) Explain why the sequence (bn ) is decreasing.
(b) Explain why c = inf{bn : n ∈ N} exists, and satisfies a ≤ c ≤ b.
(c) Explain why f (c) cannot be strictly greater than γ.
(d) Explain why f (c) cannot be strictly less than γ. Thus conclude that f (c) = γ.
(f) Explain why c 6= a and c 6= b.
2. Consider the function f (x) = x3 − 2.
(a) Explain, using a suitable theorem, why the equation f (x) = 0 has at least one
solution in the interval [0, 2].
(b) The method of successive bisection, used in problem (3), is used to construct the
following algorithm for finding a numerical approximation of the solution of the
equation f (x) = 0.
Let I1 = [0, 2].
For n ≥ 1, let cn denote the midpoint of the interval In .
If f (cn ) > 0, let In+1 denote the left half of In .
If f (cn ) < 0, let In+1 denote the right half of In .
If f (cn ) = 0, the algorithm stops.
We think of the points c1 , c2 , ..., cn , ... so obtained as a sequence of approxima-
tions for the solution c of the equation f (x) = 0 . Indeed, note that the solu-
tion c belongs to each of the intervals In . Since the length of the interval In is
(2 − 0)/2n−1 , it follows that |c − cn | < 2−0 1
2n = 2n−1 . This expression is called the
error of approximation. Clearly this error tends to 0 as n tends to infinity.
√
(i) Use this algorithm to find an approximation for 3 2, with an error less than 10−2 .

13
 THEME 4: INTEGRATION

Unit 4.1: Definition of the integral 1 12 lectures

Source: Haggarty, Section 7.1 (pp. 211–221)
Learning objectives:
After completion of this unit you should

1. know what upper and lower sums are, and be able to compute their values in specific
cases.
2. be able to write down the definition of the Riemann integral.
3. be able to prove Theorem 7.1.4 and the corollary to Theorem 7.1.4 (given in the remarks
below).

Remarks:

1. Note that only bounded functions are considered.

2. Since area has not been defined, the discussion related to area is merely to motivate the
definition of the integral.
3. Corollary to Theorem 7.1.4: If P1 and P2 are partitions of [a, b] such that P1 ⊂ P2 ,
then L(P1 ) ≤ L(P2 ) and U (P1 ) ≥ U (P2 ).

Problems: Haggarty, Exercises 7.1 (pp. 234–235) no. 2, 3

Additional problems:

1. In the proof of Theorem 7.1.4 we read: “In a similar fashion it can be shown that
L(P ) ≤ L(P 0 ).” Prove it.
2. The proof of Theorem 7.1.4 ends with: “Thus L ≤ U.” Provide the reason.
3. Write a complete proof of the corollary to Theorem 7.1.4.

Unit 4.2: Existence of the integral 2 12 lectures

Source: Haggarty, Section 7.1 (pp. 221–226)
Learning objectives:
After completion of this unit you should be able to

1. state Riemann’s condition.

2. prove Theorems 7.1.5 and 7.1.6.
3. prove Theorems 7.1.8 and 7.1.9 (Uniform Continuity Theorems)

Self-study

4. prove the small span theorem (see Appendix to Theme 4).

5. prove Theorem 7.1.9 using the small span theorem.

Problems:

1. Prove Theorem 7.1.6 for a decreasing function.

2. Prove Theorem 7.1.9 using the small span theorem.

14
*3. Prove Theorem 7.1.9 (using the small span theorem) for the following two cases where
the function is not continuous everywhere (but is assumed to be bounded):
(a) The function f is not continuous at a, but is continuous everywhere else.
(b) The function f is not continuous at c ∈ (a, b), but is continuous everywhere else.

Unit 4.3: Riemann sums 1 12 lectures

Source: Appendix to theme 4
Learning objectives:
After completion of this unit you should be able to

1. write down the definition of a Riemann sum.

2. prove that a sequence of Riemann sums converge under certain conditions to the Rie-
mann integral.

Unit 4.4: Properties of the integral 1 lecture

Source: Haggarty, Section 7.1 (pp. 226–229)
Learning objectives:
After completion of this unit you should be able to state and prove the properties of the
Riemann integral.
Remark: In the case of Property (2) note that the integral on the left exists if both integrals
on the right exist and both integrals on the right exist if the integral on the left exists.
Problems:

1. Complete the proof of Theorem 7.1.10(2) of Haggarty.

2. Prove Theorem 7.1.10(1).
3. Prove Theorem 7.1.10(3).
4. Prove Theorem 7.1.10(4).
5. Assume that an increasing function is integrable and use the properties of the integral,
to prove that a decreasing function is integrable. Hint: If the function f is decreasing
and g(x) = f (a) − f (x), then the function g is increasing.

15
 THEME 5: SERIES

Unit 5.1: Definition of infinite series 2 lectures

Source: Haggarty, Section 4.1
Learning objectives:
On completion of this study unit you should be able to

1. write down the definition of an nth partial sum of a sequence, the definition of a con-
vergent series and the definition of a divergent series.
2. show the convergence or divergence of certain series such as in Examples 1–3 and the
exercises.
3. know the definition of a geometric series and prove Result 4.1.1.
4. prove the vanishing condition 4.1.2 and apply it to show that a sequence is divergent.
5. show with an example that there are divergent series whose terms nevertheless tend to
zero.
6. state and apply the sum and scalar product rules for series.
7. state for which values of p the p-series converges and diverges.
8. prove that the harmonic series diverges.

Problems: Haggarty, Exercises 4.1 (pp. 107–108) no. 1–4

Unit 5.2: Series Tests 1 12 lectures

Source: Haggarty, Section 4.2 (excluding Section 4.2.9)
Remark: By definition a series ∞
P
r=1 ar is a positive-term series if each ar > 0.
Learning objectives:
On completion of this study unit you should be able to

1. state and prove the first comparison test.

2. state the second comparison test.
3. state D’Alembert’s ratio test.
4. state the alternating series test.
5. state the integral test, and apply it as in Example 5.
6. state the definition of a rearrangement of a series, and the definition of an absolutely
convergent series.
7. state and prove the theorem informally known as “absolute convergence implies conver-
gence.”
8. state the theorem that refers to arbitrary values that can be obtained by rearrangements
of a conditional convergent series.
9. state the rearrangement rule.
10. apply all the above-mentioned results and tests as in the examples and exercises.

Problems: Haggarty, Exercises 4.2 (p. 120) no. 1–6

Additional problem: Prove that the partial sums of a positive series form an increasing
sequence.

16
 APPENDICES
These appendices can be seen as a supplement to the textbook. The study guide refers to
these appendices.

Glossary

Set Subset Intersection Union

Complement Domain Range Image
Injective function Surjective function Bijective function Composition
Inverse Transitive property Algebraic axiom Order axiom
Lower bound Upper bound Supremum Infimum
Continuous Left-continuous Right-continuous Convergent sequence
Divergent sequence Power rule Null sequence Countable set
Uncountable set Subsequence Bounded function Upper sum
Lower sum Riemann integrable Small span theorem Partial sum
Series Vanishing condition Geometric Series Comparison test
Ratio test Conditionally convergent Rearrangement test

APPENDIX to THEME 2

Unit 2.1. Define a sequence in R to be a function

X : N −→ R : n 7→ X(n).
We will follow the tradition of denoting the value of X at n (that is, X(n)) by xn , and
so, denote the function X by the notation X = (xn ) or X = (xn : n ∈ N). We use the
parentheses to indicate that the ordering induce by that in N is crucial. We will, therefore,
notationally distinguish between the sequence X = (xn : n ∈ N) and the set {xn : n ∈ N} of
values of this sequence, the latter set being called the range of the sequence.

Unit 2.2. Countable sets

1. The empty set ∅ is said to have 0 elements.

2. If n ∈ N, a set S is said to have n elements if there exists a bijection from the set
Nn := {1, 2, . . . , n} onto S.
3. A set is said to be finite if it is either empty or it has n elements for some n ∈ N.
4. A set S is said to be infinite if it is not finite.
5. A set S is said to be denumerable (or countably infinite) if there exists a bijection of N
onto S.
6. A set S is said to be countable if it is either finite or denumerable.

Theorem. The set Q of rational numbers is countable.

Informal Proof. We will enumerate Q by the diagonal procedure ...

Uncountability of sets
A set S is said to be uncountable if it is not countable.
Theorem. The interval [0, 1] is not countable.
Theorem. The set R of real numbers is not countable.

17
 APPENDIX to THEME 4

As in the textbook we assume our function f is defined on the interval [a, b]. The definitions
of Mi and mi are also as in the textbook.
The small span theorem for continuous functions
Definition: The span of f over the partition P = {x0 , x1 , · · · , xn }, is defined by

span(P ) = max{Mi − mi : i = 1, 2, · · · , n}.

Exercise: Suppose that c is a point such that a < c < b, P1 is a partition of the interval
[a, c], and P2 a partition of the interval [c, b]. Prove that if span(P1 ) < ε and span(P2 ) < ε,
then there exists a partition P of [a, b] such that span(P ) < ε.
Exercise: Prove that if δ is a number such that span(P ) > δ for every partition P of the
interval [a, b], then either span(P ) > δ for each partion P of [a, c], or span(P ) > δ for each
partition P of [c, b].
Definition: A function f has the small span property on the interval [a, b] if for every ε > 0,
there exists a partition P for which span(P ) < ε.
Note: It can be shown that a function has the small span property if and only if it is
“uniformly continuous” (Haggarty, Section 7.1.8, p. 224) on [a, b].
Theorem. (Small span theorem) Every function f that is continuous on [a, b] has the small
span propery.
Proof. Suppose the theorem is false. Then, there exists a real number ε∗ > 0 such that
span(P ) > ε∗ for each partition P of [a, b].
Let c1 = (a + b)/2. Now either
1. span(P ) > ε∗ for each partition P of [a, c], or
2. span(P ) > ε∗ for each partition P of [c, b].
In the first case, let a1 = a and b1 = c. In the second case, let a1 = c and b1 = b. Note that
b1 − a1 = 21 (b − a).
Consider the statement that we have numbers a1 ≤ a2 ≤ · · · ≤ an and b1 ≥ b2 ≥ · · · ≥ bn ,
such that bn − an = 2−n (b − a) and span(P ) > ε∗ for each partition P of [an , bn ].
We have already seen that the statement is true for n = 1.
Suppose that the statement is true for n = m.
Then span(P ) > ε∗ for each partition P of [am , bm ]. Let cm+1 = (am + bm )/2. Then either
span(P ) > ε∗ for each partition P of [am , cm+1 ], or span(P ) > ε∗ for each partition P
of [cm+1 , bm ]. In the first case, let am+1 = am and bm+1 = cm+1 . In the second case, let
am+1 = cm+1 and bm+1 = bm . Note that bm+1 − am+1 = 12 (bm − am ) = 2−(m+1) (b − a). Also
clearly am+1 ≥ am and bm+1 ≤ bm in both cases.
Thus the statement is true for n = m + 1. This concludes our induction.
Now {an } is not only increasing but also bounded above, hence limn→∞ an exists. Since {bn }
is decreasing and bounded below, limn→∞ bn exists.
So, limn→∞ (bn − an ) exists. From bn − an = 2−n (b − a) we have

lim an = lim bn .
n→∞ n→∞

Denote this limit by p. Then (provide reason), there exists a δ > 0 such that
ε∗ ε∗
f (p) − < f (x) < f (p) +
2 2

18
for each x ∈ (p − δ, p + δ).
Thus |f (x) − f (y)| < ε∗ for all x, y in (p − δ, p + δ).
Now by our induction statement, there exists a natural number N such that aN and bN are
in (p − δ, p + δ) and span(P ) > ε∗ for every partition P of [aN , bN ]. So |f (x) − f (y)| > ε∗ for
some x, y in [aN , bN ].
This contradiction shows that our assumption, that f does not have the small span property,
is false. 
Riemann sums
Consider a function f which is bounded on the closed interval [a, b]. Divide the interval into
subintervals using a finite set of points P = {x0 , x1 , . . . , xn }, with a = x0 < x1 < · · · <
xn = b. Recall that P is called a partition of [a, b]. It partitions [a, b] into n subintervals
[x0 , x1 ], [x1 , x2 ], . . . , [xn−1 , xn ].
Choose, for each k, a sample point x∗k ∈ [xk−1 , xk ] and let Q = {x∗1 , x∗2 , . . . , x∗n }. Also, let
hk = xk − xk−1 for each k. The sum
n
X
S(P, Q) = f (x∗k )hk
k=1

is called a Riemann sum.

We denote max{hk : k = 1, 2, . . . , n} by h.
We will not prove the following useful result.
Theorem. Suppose the function f is integrable on [a, b]. Then, for any ε > 0 there exists a
Rb
δ > 0 such that | a f (x)dx − S(P, Q)| < ε for all partitions P satisfying h < δ, and for all
corresponding sample point sets Q.
Exercises

1. Prove the above-mentioned theorem in the case in which f is continuous on [a, b]. Hint:
Use Riemann’s condition.
2. Use the above-mentioned theorem to prove that if f is integrable on [a, b], then any se-
quence of Riemann sums S(Pn , Qn ), with partition Pn and sample points Qn , converges
to the Riemann integral provided that lim(hn ) = 0.

19
 APPENDIX to THEME 5
Pn k
Theorem. Suppose that the series fn (x) =def k=0 ak x satisfies the condition that
sup |f (x) − fn (x)| converges to zero as n → ∞. Then
a≤x≤b
Z b ∞ Z
X b
f (x)dx = ak xk dx.
a k=0 a

Proof. Let ε > 0 be given. Then by assumption there is an integer N so that if n > N then
ε
sup |f (x) − fn (x)| < . For such n we have
a≤x≤b b−a
Z b n Z
X b Z b
n
f (x)dx − an x dx = f (x)dx − fn (x)
a k=0 a a
Z b
≤ |f (x) − fn (x)|dx
a
Z b
ε
≤ dx
a b−a
= ε.

Since ε > 0 is arbitrary, we are done.

R1 2 2
Example: As an application we evaluate 0 ex dx. Let f (x) = ex .
x2k
Note that ex = ∞
2 P
k=0 k! .
x2k P∞
Letting fn (x) denote the nth partial sum of this series, we get that f (x)−fn (x) =
k=n+1 k! .
P∞ 1
P∞ 1 in x, we find that |f (x) −fn (x)| ≤ k=n+1
Using the fact that this expression is increasing
1
k!
P∞that 10 ≤ x ≤ 1. Since k=0 k! is a convergent series (converging to e ), we
for all x such
have that k=n+1 k! converges to zero as n → ∞. So sup0≤x≤1 |f (x) − fn (x)| converges to
zero as n → ∞ and the theorem can be applied.
By the theorem we have
1 ∞ Z 1
x2k
Z X
f (x)dx = dx
0 k!
k=0 0
∞
X 1
= .
(2k + 1)(k!)
k=0

Remark: The value of the integral is 1.4626517459 (rounded to 10 decimal places), and it
is easy to check that for example with n = 10 we have nk=0 (2k+1)(k!)
1
P
= 1.4626517447

THEME 0: FOUNDATIONAL ISSUES

Unit 0.1: Implication and proof

1 21 lectures
Source: Haggarty Section 1.1
Learning objectives:
On completion of this study unit you should be able to

20
 1. state what it means for two statements (statement forms, to be more precise) to be
logically equivalent.
2. write down the truth table that defines the composite statement P =⇒ Q.
3. define both the converse and the contra-positive of the composite statement P =⇒ Q,
and know (and prove) which of these are logically equivalent.
4. prove statements such as in Example 3 firstly directly, secondly indirectly by proving
the contra-positive, and thirdly with a proof by contradiction.

Problems: Haggarty, p.24 Exercises 1.1 numbers 4-5; p. 37 Problems 2(a)(b), 3(a)(b)

THEME 1: THE REAL NUMBERS

Note: At this stage of the course it is essential to learn to give complete proofs. This means
that every step should use valid logic, combinining previously proved conclusion(s) and/or
axiom(s) to form a new conclusion. (See Haggarty 1.1 for valid logic.)
In later lecture units, when we are more familiar with the axioms of the real number system,
we may skip small steps. However, even in such cases, we may skip steps only when it is
easy, for anyone familiar with the axioms and their consequences, to fill in the gaps.

Unit 1.1: The arithmetic and order axioms of R

1 12 lectures
Source: Haggarty Section 2.1 (pages 41-44) and Section 2.2 (pages 50-57)
Learning objectives:
On completion of this study unit you should be able to

1. explain the purpose of the arithmetic and order axioms of R.

2. write down at least three of the arithmetic axioms and two of the order axioms of R.
3. write down at least two arithmetic and one order property of R that follow from the
axioms.
4. use the order properties of R to solve inequalities.
5. define the minimum (or least element) and the maximum (or greatest element) of a set
S.
6. define the absolute value of a real number.
7. state and prove the Triangle Inequality.

Remark: Pages 45-49 of section 2.1 are excluded.

Problems: Haggarty Exercises 2.2 numbers 1-4

Unit 1.2: Completeness Axiom

2 12 lectures
Source: Haggarty Section 2.2.3 and Section 2.3
Learning objectives:
On completion of this study unit you should be able to

1. define the concepts of upper bound, lower bound, least upper bound (supremum) and
greatest lower bound (infimum) of a set.
2. state and prove alternative formulations for supremum and infimum.

21
 3. determine whether a given set is bounded from below or bounded from above.
4. find the supremum and the infimum of a given set (if they exist).
5. state the Completeness Axiom of R.
6. prove the Archimedean Postulate: The set N of natural numbers does not have an upper
bound.
7. prove that Q is dense in R.

Problems:
Exercise 2.2 number 5;
Exercise 2.3 numbers 2(a), 2(b), 2(c), 4;
Problems 4(a), 6, 7(d), 8.

Additional problems:

1. Let A and B be any non-empty bounded subsets of R. Prove the following statements:
(a) If c is any non-negative real number and C = {ca : a ∈ A} then sup(C) = c sup(A).
(b) If C = {x + y : x ∈ A, y ∈ B} then sup(C) = sup(A) + sup(B).
(c) If C = {−a : a ∈ A} then sup(C) = − inf(A).
2. If A ⊆ B and A and B are bounded above, what can one say about sup(A) and sup(B)?
3. If a non-empty subset S of R has an upper bound M ∈ S, prove that sup(S) = M .
4. Say in your own words what the connection is between the maximum of a set S and
the supremum of S.
5. Let S be a non-empty bounded subset of R. Prove that sup(S) = ` if and only if
(a) ` is an upper bound of S, and
(b) for every real number  > 0 there exists an element x ∈ S such that x > ` − .

THEME 2: SEQUENCES

Unit 2.1: Sequences and limits

2 21 lectures
Source: Haggarty Section 3.1 - 3.3 (up to 3.3.4) Appendix to Theme 2, and class notes
Learning objectives:
On completion of this study unit you should be able to

1. write down the definition of a sequence as a function from N to R.

2. define the concepts of convergence and divergence of a sequence of real numbers.
3. prove that a sequence can have at most one limit.
4. prove that a convergent sequence is bounded.
5. prove the convergence of specific sequences.
6. know and apply the Rules 3.1.2, the Sandwich Rule 3.1.3 and the Composite Rule 3.1.4
7. state the limits of certain basic sequences and use it to find and prove the limits of other
sequences.
8. know what a null sequence is

22
 9. be able to prove and apply the Power rule
10. know the Basic null sequences of Theorem 3.2.2
11. know what it means for a sequence to tend to infinity
12. know and apply the Reciprocal rule 3.3.4 (p.87)

Problems: Exercises 3.1: 1(b), 2, 3 (a), (c), 4 (a); 3.2: 1, 2 (a), 3, 4 (b), 5; 3.3: 1 (b), 2.
Problems 1 (a), 2, 3 (a), (b), (e) 4 (a), 5 (a), (b), (c)

Unit 2.2: Monotone sequences

2 21 lectures
Source: Haggarty Section 3.4.1
Learning objectives:
On completion of this study unit you should be able to

1. write down the definitions of an increasing sequence, a decreasing sequence and a mono-
tone sequence.
2. state and prove the Monotone Convergence Theorem.
3. use the Monotone Convergence Theorem to prove the convergence of specific sequences.
4. define the concepts: countable and uncountable sets.
5. give a demonstration that the set Q of rational numbers is countable by using a diagonal
argument.
6. use a repeated interval division argument to prove that the interval [0, 1] is uncountable
7. explain how the above-mentioned result implies that R is uncountable

Problems: Exercises 3.4: 1 (a), (b), (c), (d), 2 (a), (b),(c),(d)

Problems 6 (a),(b), (c), (d), (e), 8

Unit 2.3: Subsequences

1 lecture
Source: Haggarty Section 3.3.5 and 3.4.2
Learning objectives:
On completion of this study unit you should be able to

1. write down the definition of a subsequence.

2. prove that every subsequence of a convergent/ divergent sequence of real numbers is
convergent /divergent.
3. use subsequences to prove the divergence of specific sequences.
4. state and prove the Bolzano-Weierstrass Theorem.

Problems: Exercises 3.3: 3 (a), (b), (c), 4

Problems 7, 9

23
THEME 3: CONTINUOUS FUNCTIONS
Unit 3.1: Limits and Continuity
Note: Special attention should be paid to the  − δ definition, or characterization, of
continuity. Students neglecting this will have difficulties with all the units of this theme,
because this definition is used so many times in proofs and arguments.
2 lectures (Partially revision.)
Source: Haggarty Section 5.1 and 5.2
Learning objectives:
On completion of this study unit you should

1. be able to write down the definition of a limit of a function f (x) as x tends to infinity,
and the definition of a limit of a function f (x) as x tends to a point a ∈ R
2. be able to write down the  − δ definitions of the left-hand and right-hand limits
3. know and be able to use the Limit Laws 5.1.3 to 5.1.6
4. formulate the definition of continuity of a function at a point c
5. formulate the  − δ definition of continuity of a function at a point c
6. be able use the  − δ definition of continuity to prove that certain basic functions are
continuous, and that certain other functions are not continuous
7. be able to write down both the limit-formulation and the  − δ formulation of left-
continuity and right-continuity of a function f (x) at a point c
8. know and be able to prove that a function can have at most one limit at a point
9. know and be able to prove that if f is continuous at a, and (xn ) converges to a, then
(f (xn )) converges to f (a)

Problems:

(1) Page 144 nr. 2, 3, 6

(2) Page 152 nr. 2, 3, 4(b)
(3) Page 163 nr. 6, 7

Unit 3.2: Boundedness of Continuous Functions

2 lectures
Source: Haggarty Section 5.3.1
Learning objectives:
On completion of this study unit you should

1. know and be able to prove that a continuous function on a closed and bounded interval
is bounded from above and from below.
2. know and be able to prove that a continuous function on a closed and bounded interval
attains its minimum and its maximum values on the interval.

Problems:

(1) Consider an interval I = [a, b] and a function f which is continuous on I. Suppose that
f is unbounded on I.
Let c denote the midpoint of the interval [a, b]. Since f is unbounded on [a, b], f is
unbounded on at least one of the intervals [a, c] or [c, b]. Let I1 = [a1 , b1 ] be that half

24
 of [a, b] on which f is unbounded. If f is unbounded on both halves, let [a1 , b1 ] denote
the left half.
Continue this process repeatedly, so that for each n ∈ N, In+1 = [an+1 , bn+1 ] is that
half of the interval In = [an , bn ] on which f is unbounded. If f is unbounded on both
halves of the interval In , then let In+1 be the left half.
(a) Find an expression for the length of each interval In .
(b) Explain why the set
A = {an : n ∈ N}.
has a supremum and that d = sup A belongs to the interval [a, b].
(c) Show that the sequence (an ) is increasing.
(d) Prove that there are numbers K and L, and a number δ > 0 such that, for every
x ∈ [a, b]
|x − d| < δ ⇒ K < f (x) < L. (1)
(e) Show that there is a number n ∈ N such that In ⊂ (d − δ, d + δ).
(f) Explain why we have now arrived at a contradiction.
(2) Consider an interval I = [a, b] and a function f that is continuous on I.
Assume that f is not bounded on I.
(a) Use the fact that f is not bounded to prove that there is a sequence (xn ) in [a, b]
for which |f (xn )| > n for each n ≥ 1.
(b) Now show that our assumption that f is not bounded on I leads to a contradiction,
concluding that f is in fact bounded on I.

Unit 3.3: The Intermediate Value Theorem

2 21 lectures
Source: Haggarty Sections 5.2 and 5.3
Learning objectives:
On completion of this study unit you should

1. know and be able to prove the Intermediate Value Theorem.

2. be able to use the Intermediate Value Theorem.
3. know and be able to prove that a continuous function maps intervals onto intervals.
4. know that if a continuous function is invertible then the inverse is also continuous (p.
151).

Problems:
(1) Page 162 nr. 1, 2, 3.
(2) Page 163 nr. 8, 9(d), 10.
(3) Let I = [a, b] be an interval, and suppose that f is continuous on I such that f (a) <
γ < f (b). Assume that f (x) 6= γ for all x ∈ I.
Let c1 be the midpoint of the interval I. If f (c1 ) < γ, let I1 be the right half of I. If
f (c1 ) > γ, let I1 be the left half of I.
Continuous this process repeatedly, so that cn denotes the midpoint of the interval In .
If f (cn ) < γ, let In+1 be the right half of In , and if f (cn ) > γ, we let In+1 be the left
half of In .
Let In = [an , bn ]. Note that the length of In is
b−a
.
2n

25
 (a) Explain why the sequence (bn ) is decreasing.
(b) Explain why c = inf{bn : n ∈ N} exists, and satisfies a ≤ c ≤ b.
(c) Explain why f (c) cannot be strictly greater than γ.
(d) Explain why f (c) cannot be strictly less than γ. Thus conclude that f (c) = γ.
(f) Explain why c 6= a and c 6= b.
(4) Consider the function f (x) = x3 − 2.
(a) Explain, using a suitable theorem, why the equation f (x) = 0 has at least one
solution in the interval [0, 2].
(b) The method of successive bisection, used in problem (3), is used to construct the
following algorithm for finding a numerical approximation of the solution of the
equation f (x) = 0.
Let I1 = [0, 2].
For n ≥ 1, let cn denote the midpoint of the interval In .
If f (cn ) > 0, let In+1 denote the left half of In .
If f (cn ) < 0, let In+1 denote the right half of In .
If f (cn ) = 0, the algorithm stops.
We think of the points c1 , c2 , ..., cn , ... so obtained as a sequence of approximations
for the solution c of the equation f (x) = 0 . Indeed, note that the solution c belongs
to each of the intervals In . Since the length of the interval In is (2 − 0)/2n−1 , it
follows that
2−0 1
|c − cn | < n
= n−1 .
2 2
This expression is called the error of approximation. Clearly this error tends to 0
as n tends to infinity.
√
(i) Use this algorithm to find an approximation for 3 2, with an error less than
10−2 .

THEME 4: INTEGRATION
Lecture Unit 4.1 Definition of the integral
1 21 lectures
Source Haggarty Section 7.1 p 211-221
After completion of this unit you should

1. know what upper and lower sums are, and be able to compute their values in specific
cases;
2. be able to write down the definition of the Riemann integral;
3. be able to prove Theorem 7.1.4 and the corollary to Theorem 7.1.4 of the remarks below.

Remarks

1. Note that only bounded functions are considered.

2. Since area has not been defined, the discussion related to area is merely to motivate the
definition of the integral.
3. Corollary to Theorem 7.1.4 If P1 and P2 are partitions of [a, b] such that P1 ⊂ P2 ,
then L(P1 ) ≤ L(P2 ) and U (P1 ) ≥ U (P2 ).

Exercises

26
 1. Exercises 7.1 in Haggarty: Numbers 2 and 3.
2. In the proof of Theorem 7.1.4 we read: “In a similar fashion it can be shown that
L(P ) ≤ L(P 0 ).” Prove it.
3. The proof of Theorem 7.1.4 ends with: “Thus L ≤ U.” Provide the reason.
4. Write a complete proof of the Corollary to Theorem 7.1.4.

Lecture Unit 4.2 Existence of the integral

2 21 lectures
Source Haggarty Section 7.1 p 221-226
After completion of this unit you should be able to

1. state Riemann’s condition;

2. prove Theorems 7.1.5 and 7.1.6
3. prove the small span theorem (see Appendix to Theme 4)
4. prove Theorem 7.1.9 using the small span theorem.

Exercises

1. Prove Theorem 7.1.6 for a decreasing function.

2. Prove Theorem 7.1.9 using the small span theorem.
*3. Prove Theorem 7.1.9 (using the small span theorem) for the following two cases where
the function is not continuous everywhere:
(a) The function f is not continuous at a, but continuous everywhere else.
(b) The function f is not continuous at c ∈ (a, b), but continuous everywhere else.

Lecture Unit 4.3 Riemann sums

1 21 Lectures
Source Appendix to theme 4
After completion of this unit you should be able to

1. write down the definition of a Riemann sum

2. prove that a sequence of Riemann sums converge under certain conditions to the Rie-
mann integral

Lecture Unit 4.4 Properties of the integral

1 lecture
Source Haggarty Section 7.1 p 226-229
After completion of this unit you should be able to state and prove the properties of the
Riemann integral.
Remarks

1. In the case of Property (2) note that the integral on the left exists if both integrals on
the right exist and both integrals on the right exist if the integral on the left exists.

Exercises

1. Complete the proof of Theorem 7.1.10(2) of Haggarty.

2. Prove Theorem 7.1.10(1).

27
 3. Prove Theorem 7.1.10(3).
4. Prove Theorem 7.1.10(4).
5. Assume that an increasing function is integrable and use the properties of the integral,
to prove that a decreasing function is integrable. Hint: If the function f is decreasing
and g(x) = f (a) − f (x), then the function g is increasing.

THEME 5: SERIES

Unit 5.1: Definition of infinite series

2 lectures
Source: Haggarty Section 4.1
Learning objectives:
On completion of this study unit you should be able to

1. write down the definition of an nth partial sum of a sequence, the definition of a con-
vergent series and the definition of a divergent series.
2. show the convergence or divergence of certain series such as in Examples 1-3 and the
exercises.
3. know the definition of a geometric series and prove Result 4.1.1
4. prove the vanishing condition 4.1.2 and apply it to show that a sequence is divergent.
5. show with an example that there are divergent series whose terms nevertheless tend to
zero.
6. state and apply the Sum - and Scalar product rules for series.
7. state for which values of p the p-series converges and diverges.
8. prove that the harmonic series diverges.

Problems:Exercises 1-4

Unit 5.2: Series Tests

1 12 lectures
Source: Haggarty Section 4.2, excluding the Cauchy product of series (4.2.9)
Remark: By definition a series ∞
P
r=1 ar is a positive-term series if each ar > 0.
Learning objectives:
On completion of this study unit you should be able to

1. state and prove the First comparison test.

2. state the Second comparison test.
3. state D’Alembert’s ratio test.
4. state the Alternating series test.
5. state the Integral test, and apply it as in Example 5.
6. state the definition of a rearrangement of a series, and the definition of an absolutely
convergent series.
7. state and prove the theorem informally known as “absolute convergence implies conver-
gence.”

28
 8. state the theorem that refers to arbitrary values that can be obtained by rearrangements
of a conditional convergent series.
9. state the rearrangement rule.
10. apply all the above-mentioned results and tests as in the examples and exercises.

Problems:Exercises 1-6
Additional problem: Prove that the partial sums of a positive series form an increasing
sequence.

APPENDIX to THEME 2
(These appendices can be seen as a supplement to the textbook. The study guide refers to
these appendices.)

Unit 2.1.1. Define a sequence in R to be a function

X : N −→ R : n 7→ X(n).

We’ll follow the tradition of denoting the value of X at n (that is, X(n)) by xn , and so, denote
the function X by the notation X = (xn ) or X = (xn : n ∈ N). We use the parentheses
to indicate that the ordering induce by that in N is crucial. We’ll, therefore, notationally
distinguish between the sequence X = (xn : n ∈ N) and the set {xn : n ∈ N} of values of this
sequence, the latter set being called the range of the sequence.

Unit 2.1.2 Countable sets

1. The empty set ∅ is said to have 0 elements.

2. If n ∈ N, a set S is said to have n elements if there exists a bijection from the set
Nn := {1, 2, . . . , n} onto S.
3. A set is said to be finite if it is either empty or it has n elements for some n ∈ N.
4. A set S is said to be infinite if it is not finite.
5. A set S is said to be denumerable (or countably infinite) if there exists a bijection of N
onto S.
6. A set S is said to be countable if it is either finite or denumerable.

Theorem The set Q of rational numbers is countable.

Informal Proof. We’ll enumerate Q by the diagonal procedure ...

Unit 2.1.3 Uncountability of sets

A set S is said to be uncountable if it is not countable.

Theorem The interval [0, 1] is not countable. (We will do another proof in class.)

Proof. The proof proceeds by contradiction and uses the fact that every real number x ∈ [0, 1]

29
has a decimal representation x = 0.b1 b2 b3 . . ., where bi = 0, 1 . . . , 9. Suppose that there is an
enumeration x1 , x2 , x3 , . . . of all numbers in [0, 1], which we display as:

x1 = 0.b11 b12 b13 . . . b1n , . . .

x2 = 0.b21 b22 b23 . . . b2n . . .
x3 = 0.b31 b32 b33 . . . b3n . . .
... ...
xn = 0.bn1 bbn2 bn3 . . . bnn . . . ,

Now define a real number y := 0.y1 y2 y3 . . . yn . . . by setting y1 := 2 if b11 ≥ 5 and y1 := 7 if

b11 ≤ 4. In general, let 
2 if bnn ≥ 5,
yn :=
7 if bnn ≤ 4.

Then y ∈ [0, 1]. Note that the numbber y is not equal to any of the numbers with two
decimal representations, since yn 6= 0, 9 for all n ∈ N. Further, since y and xn differ in the
nthe decimalplace, then y 6= xn for any n. Therefore, y is not included in the enumeration
of [0, 1], and so, the hypothesis is contradicted.

Unit 2.2 Monotone sequences

Theorem The set R of real numbers is not countable.

Proof. (Will be discussed in class)

APPENDIX to THEME 3

Theorem Let f : A → B be a continuous bijection, with A and B closed intervals. If

f is differentiable at a ∈ A and f 0 (a) 6= 0, then f −1 is differentiable at b = f (a) ∈ B and
(f −1 )0 (b) = f 01(a) .

Proof Let A = [p, q], and suppose that a is not an endpoint of A. Consider the function

f (a + σ) − f (a)


 if σ 6= 0
E(σ) = σ

 0
f (a) if σ = 0

with domain [p − a, q − a]. Since f is differentiable at a, the function E is continuous at 0.

Let

σ(h) = f −1 (b + h) − f −1 (b),

where h ∈ [inf B − b, sup B − b]. Since f is a continuous bijection on A, f − is continuous on

B. Therefore

lim σ(h) = f −1 (b) − f −1 (b) = 0 = σ(0). (2)

h→0

30
 Thus σ(h) is continuous at h = 0. For h 6= 0 we have
f −1 (b + h) − f −1 (b) f −1 (b + h) − f −1 (b)
=
h (b + h) − b

σ(h)
=
f (f −1 (b + h)) − f (f −1 (b))

σ(h)
= (3)
f (f −1 (b) + σ(h)) − f (f −1 (b))

σ(h)
=
f (a + σ(h)) − f (a)

1
=
E(σ(h))
Since E is continuous at σ = 0, and σ is continuous at h = 0, it follows from (2) and the
Composite Rule 5.2.6 that

lim E(σ(h)) = E(σ(0)) = E(0) = f 0 (a).

h→0

Since f 0 (a) 6= 0, it now follows from (3) and the Quotient Rule 5.1.3 that
f −1 (b + h) − f −1 (b) 1 1
lim = lim = 0 .
h→0 h h→0 E(σ(h)) f (a)

Thus f −1 is differentiable at b and (f −1 )0 (b) = 1

f 0 (a) .

APPENDIX to THEME 4
As in the textbook we assume our function f is defined on the interval [a, b]. The definitions
of Mi and mi are also as in the textbook.
The small span theorem for continuous functions
Definition: The span of f over the partition P = {x0 , x1 , · · · , xn }, is defined by

span(P ) = max{Mi − mi : i = 1, 2, · · · , n}.

Exercise: Suppose that c is a point such that a < c < b, P1 is a partition of the interval
[a, c], and P2 a partition of the interval [c, b]. Prove that if span(P1 ) <  and span(P2 ) < ,
then there exists a partition P of [a, b] such that span(P ) < .
Exercise: Prove that if δ is a number such that span(P ) > δ for every partition P of the
interval [a, b], then either span(P ) > δ for each partion P of [a, c], or span(P ) > δ for each
partition P of [c, b].
Definition: A function f has the small span property on the interval [a, b] if for every  > 0,
there exists a partition P for which span(P ) < .
Note: It can be shown that a function has the small span property if and only if it is
“uniformly continuous” (Def 7.1.8, p.224) on [a, b].
Theorem (Small span theorem) Every function f that is continuous on [a, b] has the small
span propery.
Proof Suppose the theorem is false. Then, there exists a real number ∗ > 0 such that
span(P ) > ∗ for each partition P of [a, b].
Let c1 = (a + b)/2. Now either

31
 1. span(P ) > ∗ for each partition P of [a, c], or
2. span(P ) > ∗ for each partition P of [c, b].
In the first case, let a1 = a and b1 = c. In the second case, let a1 = c and b1 = b. Note that
b1 − a1 = 12 (b − a).
Consider the statement that we have numbers a1 ≤ a2 ≤ · · · ≤ an and b1 ≥ b2 ≥ · · · ≥ bn ,
such that bn − an = 2−n (b − a) and span(P ) > ∗ for each partition P of [an , bn ].
We have already seen that the statement is true for n = 1.
Suppose that the statement is true for n = m.
Then span(P ) > ∗ for each partition P of [am , bm ]. Let cm+1 = (am + bm )/2. Then either
span(P ) > ∗ for each partition P of [am , cm+1 ], or span(P ) > ∗ for each partition P
of [cm+1 , bm ]. In the first case, let am+1 = am and bm+1 = cm+1 . In the second case, let
am+1 = cm+1 and bm+1 = bm . Note that bm+1 − am+1 = 12 (bm − am ) = 2−(m+1) (b − a). Also
clearly am+1 ≥ am and bm+1 ≤ bm in both cases.
Thus the statement is true for n = m + 1. This concludes our induction.
Now {an } is not only increasing but also bounded above, hence limn→∞ an exists. Since {bn }
is decreasing and bounded below, limn→∞ bn exists.
So, limn→∞ (bn − an ) exists. From bn − an = 2−n (b − a) we have
lim an = lim bn .
n→∞ n→∞

Denote this limit by p. Then (provide reason), there exists a δ > 0 such that
∗ ∗
f (p) − < f (x) < f (p) +
2 2
for each x ∈ (p − δ, p + δ).
Thus |f (x) − f (y)| < ∗ for all x, y in (p − δ, p + δ).
Now by our induction statement, there exists a natural number N such that aN and bN are
in (p − δ, p + δ) and span(P ) > ∗ for every partition P of [aN , bN ]. So |f (x) − f (y)| > ∗ for
some x, y in [aN , bN ].
This contradiction shows that our assumption, that f does not have the small span property,
is false. 
Riemann sums
Consider a function f which is bounded on the closed interval [a, b]. Divide the interval into
subintervals using a finite set of points P = {x0 , x1 , . . . , xn }, with a = x0 < x1 < · · · <
xn = b. Recall that P is called a partition of [a, b]. It partitions [a, b] into n subintervals
[x0 , x1 ], [x1 , x2 ], . . . , [xn−1 , xn ].
Choose, for each k, a sample point x∗k ∈ [xk−1 , xk ] and let Q = {x∗1 , x∗2 , . . . , x∗n }. Also, let
hk = xk − xk−1 for each k. The sum
n
X
S(P, Q) = f (x∗k )hk
k=1
is called a Riemann sum.
We denote max{hk : k = 1, 2, . . . , n} by h.
We will not prove the following useful result.
Theorem Suppose the function f is integrable on [a, b]. Then, for any  > 0 there exists
Rb
a δ > 0 such that | a f − S(P, Q)| <  for all partitions P satisfying h < δ, and for all
corresponding sample point sets Q.
Exercises

32
 1. Prove the above-mentioned theorem in the case in which f is continuous on [a, b]. Hint:
Use Riemann’s condition.
2. Use the above-mentioned theorem to prove that if f is integrable on [a, b], then any se-
quence of Riemann sums S(Pn , Qn ), with partition Pn and sample points Qn , converges
to the Riemann integral provided that lim(hn ) = 0.

APPENDIX to THEME 5
Pn k
Theorem Suppose that the series fn (x) =def k=0 ak x satisfies the condition that
sup |f (x) − fn (x)| converges to zero as n → ∞. Then
a≤x≤b
Z b ∞ Z
X b
f (x)dx = an xn dx.
a k=0 a

Proof Let  > 0 be given. Then by assumption there is an integer N so that if n > N then

sup |f (x) − fn (x)| < . For such n we have
a≤x≤b b−a
Z b n Z
X b Z b
n
| f (x)dx − an x dx| =| f (x)dx − fn (x)|
a k=0 a a
Z b
≤ |f (x) − fn (x)|dx
a
Z b

≤ dx
a b−a
=.

Since  > 0 is arbitrary, we are done.

R1 2 2
Example: As an application we evaluate 0 ex dx. Let f (x) = ex .
x2k
Note that ex = ∞
2 P
k=0 k! .
x2k P∞
Letting fn (x) denote the nth partial sum of this series, we get that f (x)−fn (x) =
k=n+1 k! .
P∞ 1
P∞ 1 in x, we find that |f (x) −fn (x)| ≤ k=n+1
Using the fact that this expression is increasing
1
k!
P that 10 ≤ x ≤ 1. Since k=0 k! is a convergent series (converging to e ), we
for all x such
have that ∞ k=n+1 k! converges to zero as n → ∞. So sup0≤x≤1 |f (x) − fn (x)| converges to
zero as n → ∞ and the theorem can be applied.
By the theorem we have
1 ∞ Z 1
x2k
Z X
f (x)dx = dx
0 k!
k=0 0
∞
X 1
= .
(2k + 1)(k!)
k=0

Remark: The value of the integral is 1.4626517459 (rounded to 10 decimal places), and it
is easy to check that for example with n = 10 we have nk=0 (2k+1)(k!)
1
P
= 1.4626517447

33