Concepts of Modern Mathematics I (Math 202)

Virginia Intermont College

Jason Riedy
Fall semester, 2008
2
These pages are available as PDF, either as one growing PDF document for the
entirety or as individual documents for each sessions notes.
If you have diculties viewing these or have particular accessibility needs, please
mail me at [email protected].
Contents
I Introduction 11
1 Syllabus 13
1.1 Concepts of Modern Mathematics I . . . . . . . . . . . . . . . . . 13
1.2 Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.3 Instructor: Jason Riedy . . . . . . . . . . . . . . . . . . . . . . . 13
1.4 Text . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.5 Grading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.6 On homework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.7 Submitting homework . . . . . . . . . . . . . . . . . . . . . . . . 14
2 Syllabus schedule 15
II Notes for chapters 1 and 2 17
3 Notes for 18 August 19
3.1 Syllabus and class mechanics . . . . . . . . . . . . . . . . . . . . 19
3.2 Introductions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.3 First homework . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.4 Problem solving . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.4.1 Categories . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.4.2 Polyas stages . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.4.3 Understand the problem . . . . . . . . . . . . . . . . . . . 20
3.4.4 Devise a plan . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.4.5 Carry out the plan . . . . . . . . . . . . . . . . . . . . . . 21
3.4.6 Looking back . . . . . . . . . . . . . . . . . . . . . . . . . 21
4 Notes for 20 August 23
4.1 Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.1.1 Problem solving . . . . . . . . . . . . . . . . . . . . . . . 23
4.1.2 Categories . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.2 Todays goal: Problem solving principles . . . . . . . . . . . . . . 23
4.3 Polyas principles . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.3.1 Understand the problem . . . . . . . . . . . . . . . . . . . 24
3
4 CONTENTS
4.3.2 Divise a plan . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.3.3 Carry out the plan . . . . . . . . . . . . . . . . . . . . . . 24
4.3.4 Examine your solution . . . . . . . . . . . . . . . . . . . . 25
4.4 Two closely related tactics, guessing and making a list . . . . . . 25
4.4.1 Guessing . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.4.2 Example 1.3 from the text . . . . . . . . . . . . . . . . . . 25
4.4.3 Tabling / Making a list . . . . . . . . . . . . . . . . . . . 26
4.4.4 Example 1.4 from the text: a counting problem. . . . . 26
4.4.5 Dierent example to show bisection . . . . . . . . . . . . 28
4.5 Next time: More problem solving ideas. . . . . . . . . . . . . . . 28
4.6 Homework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
5 Notes for 22 August 31
5.1 Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.1.1 Polyas problem solving principles . . . . . . . . . . . . . 31
5.1.2 Tactic: Guessing . . . . . . . . . . . . . . . . . . . . . . . 31
5.1.3 Tactic: Tabling / Making an orderly list . . . . . . . . . . 32
5.1.4 Example of creating a list . . . . . . . . . . . . . . . . . . 32
5.1.5 Tactic not from the text: Orderly creation of a partial list 34
5.2 New tactic: Drawing a diagram . . . . . . . . . . . . . . . . . . . 35
5.2.1 Understanding the problem . . . . . . . . . . . . . . . . . 35
5.2.2 Devise a plan . . . . . . . . . . . . . . . . . . . . . . . . . 36
5.2.3 Carry out the plan . . . . . . . . . . . . . . . . . . . . . . 36
5.2.4 Looking back . . . . . . . . . . . . . . . . . . . . . . . . . 36
5.3 Homework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
6 Solutions for rst weeks assignments 39
6.1 Problem Set 1.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
6.1.1 Problem 10 . . . . . . . . . . . . . . . . . . . . . . . . . . 39
6.1.2 Problem 13 . . . . . . . . . . . . . . . . . . . . . . . . . . 41
6.2 Example like 1.3 with no solution . . . . . . . . . . . . . . . . . . 42
6.3 Problem Set 1.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
6.3.1 Problem 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
6.3.2 Problem 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
6.3.3 Problem 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
6.4 Consider solving Example 1.3 with a table . . . . . . . . . . . . . 46
6.5 More in Problem Set 1.2 . . . . . . . . . . . . . . . . . . . . . . . 46
6.5.1 Problem 6: whoops, not assigned . . . . . . . . . . . . . . 46
6.5.2 Problem 13 . . . . . . . . . . . . . . . . . . . . . . . . . . 47
6.5.3 Problem 18 . . . . . . . . . . . . . . . . . . . . . . . . . . 48
6.5.4 Problem 20 . . . . . . . . . . . . . . . . . . . . . . . . . . 48
7 Notes for 25 August 51
7.1 Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
7.2 Draw a diagram, follow dependencies . . . . . . . . . . . . . . . . 51
7.2.1 Understanding the problem . . . . . . . . . . . . . . . . . 51
CONTENTS 5
7.2.2 Devise a plan . . . . . . . . . . . . . . . . . . . . . . . . . 52
7.2.3 Carry out the plan . . . . . . . . . . . . . . . . . . . . . . 52
7.2.4 Looking back . . . . . . . . . . . . . . . . . . . . . . . . . 53
7.3 Look for a pattern . . . . . . . . . . . . . . . . . . . . . . . . . . 54
7.3.1 Example: What is the last digit of 7
100
. . . . . . . . . . 54
7.4 Patterns and representative special cases . . . . . . . . . . . . . . 55
7.4.1 Sums over Pascals triangle . . . . . . . . . . . . . . . . . 56
7.5 Homework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
8 Notes for 27 August 59
8.1 Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
8.2 Ruling out possibilities . . . . . . . . . . . . . . . . . . . . . . . . 59
8.2.1 Logic puzzles . . . . . . . . . . . . . . . . . . . . . . . . . 60
8.3 The pigeonhole principle . . . . . . . . . . . . . . . . . . . . . . . 60
8.4 Mathematical reasoning . . . . . . . . . . . . . . . . . . . . . . . 61
8.5 Next time: structures and kinds of proofs . . . . . . . . . . . . . 63
8.6 Homework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
9 29 August: Review of previous notes 65
10 Solutions for second weeks assignments 67
10.1 Patterns: The 87th digit past the decimal in 1/7? . . . . . . . . . 67
10.2 Patterns: Units digit of 3
100
. . . . . . . . . . . . . . . . . . . . 67
10.3 Problem set 1.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
10.3.1 Arithmetic progressions: Problem 6 . . . . . . . . . . . . 68
10.3.2 Problem 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
10.3.3 Problem 11 . . . . . . . . . . . . . . . . . . . . . . . . . . 69
10.3.4 Problem 20 . . . . . . . . . . . . . . . . . . . . . . . . . . 70
10.4 Problem set 1.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
10.4.1 Reducing possibilities: Problem 7 . . . . . . . . . . . . . . 71
10.4.2 Logic puzzle: Problem 9 . . . . . . . . . . . . . . . . . . . 72
10.4.3 Pigeonholes: Problem 12 . . . . . . . . . . . . . . . . . . . 72
10.4.4 Pigeonholes: Problem 13 . . . . . . . . . . . . . . . . . . . 72
10.4.5 Pigeonholes: Problem 14 . . . . . . . . . . . . . . . . . . . 72
10.5 Inductive or deductive? . . . . . . . . . . . . . . . . . . . . . . . 73
11 Notes for 1 September 75
11.1 Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
11.2 Proof . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
11.3 Direct proof . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
11.4 Proof by contrapositives . . . . . . . . . . . . . . . . . . . . . . . 77
11.5 Homework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
12 Notes for 3 September 79
12.1 Proof review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
12.2 Inductive proof . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
6 CONTENTS
12.3 Starting with set theory . . . . . . . . . . . . . . . . . . . . . . . 81
12.4 Language of set theory . . . . . . . . . . . . . . . . . . . . . . . . 81
12.5 Basic denitions . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
12.6 Translating sets into (and from) English . . . . . . . . . . . . . . 82
12.7 Next time: Relations between and operations on sets . . . . . . . 83
12.8 Homework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
13 Notes for 8 September 85
13.1 Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
13.2 Relations and Venn diagrams . . . . . . . . . . . . . . . . . . . . 85
13.3 Translating relations into (and from) English . . . . . . . . . . . 86
13.4 Consequences of the set relation denitions . . . . . . . . . . . . 87
13.5 Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
13.6 Homework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
14 Solutions for third weeks assignments 89
14.1 Induction: Sum of rst n integers . . . . . . . . . . . . . . . . . . 89
14.2 Problem set 2.1 (p83) . . . . . . . . . . . . . . . . . . . . . . . . 90
14.2.1 Problem 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
14.2.2 Problem 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
14.2.3 Problem 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
14.2.4 Problem 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
14.2.5 Problem 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
14.2.6 Problem 27 . . . . . . . . . . . . . . . . . . . . . . . . . . 91
15 Notes for 10 September 93
15.1 Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
15.1.1 Denitions . . . . . . . . . . . . . . . . . . . . . . . . . . 93
15.1.2 Relations . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
15.1.3 Operations . . . . . . . . . . . . . . . . . . . . . . . . . . 94
15.2 From sets to whole numbers . . . . . . . . . . . . . . . . . . . . . 94
15.3 Homework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
16 Notes for 12 September 97
16.1 Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
16.2 Addition of whole numbers . . . . . . . . . . . . . . . . . . . . . 98
16.3 Subtraction of whole numbers . . . . . . . . . . . . . . . . . . . . 99
16.4 Multiplication of whole numbers . . . . . . . . . . . . . . . . . . 100
16.5 Monday: Division and exponentials . . . . . . . . . . . . . . . . . 101
16.6 Homework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
17 Solutions for fourth weeks assignments 103
17.1 Problem set 2.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
17.1.1 Problem 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
17.1.2 Problem 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
17.1.3 Problem 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
CONTENTS 7
17.1.4 Problem 13 . . . . . . . . . . . . . . . . . . . . . . . . . . 104
17.1.5 Problem 21 . . . . . . . . . . . . . . . . . . . . . . . . . . 104
17.1.6 Problem 23 . . . . . . . . . . . . . . . . . . . . . . . . . . 104
17.1.7 Why answering problem 32 would be a bad idea. . . . . . 105
17.2 Problem set 2.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
17.2.1 Problem 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
17.2.2 Problem 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
17.2.3 Problem 11 . . . . . . . . . . . . . . . . . . . . . . . . . . 106
17.2.4 Problem 24 . . . . . . . . . . . . . . . . . . . . . . . . . . 106
17.3 Write 2 + 3 using disjoint sets. . . . . . . . . . . . . . . . . . . . 106
17.4 Illustrate 2 + 3 using Peano arithmetic. . . . . . . . . . . . . . . 106
17.5 Problem set 2.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
17.5.1 Problem 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
17.5.2 Problem 10 . . . . . . . . . . . . . . . . . . . . . . . . . . 107
17.5.3 Problem 26 . . . . . . . . . . . . . . . . . . . . . . . . . . 108
17.6 Illustrate 2 3 using Peano arithmetic. You do not need to expand
addition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
17.7 Illustrate (1 2) 3 = 1 (2 3) using a volume of size six. . . . . . 108
18 Notes for the fth week: review 109
18.1 Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
18.2 Problem solving . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
18.2.1 Understand the problem . . . . . . . . . . . . . . . . . . . 110
18.2.2 Devise a plan . . . . . . . . . . . . . . . . . . . . . . . . . 110
18.2.3 Carry out the plan . . . . . . . . . . . . . . . . . . . . . . 111
18.2.4 Look back at your solution . . . . . . . . . . . . . . . . . 111
18.3 Set theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
18.3.1 Denitions and mappings . . . . . . . . . . . . . . . . . . 111
18.3.2 Cardinality and one-to-one correspondence . . . . . . . . 112
18.4 Operations and whole numbers . . . . . . . . . . . . . . . . . . . 113
19 First exam and solutions 115
III Notes for chapters 3, 4, and 5 117
20 Notes for the sixth week: digits, bases, and operations 119
20.1 Positional Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . 119
20.2 Converting Between Bases . . . . . . . . . . . . . . . . . . . . . . 121
20.2.1 Converting to Decimal . . . . . . . . . . . . . . . . . . . . 122
20.2.2 Converting from Decimal . . . . . . . . . . . . . . . . . . 123
20.3 Operating on Numbers . . . . . . . . . . . . . . . . . . . . . . . . 123
20.3.1 Multiplication . . . . . . . . . . . . . . . . . . . . . . . . . 124
20.3.2 Addition . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
20.3.3 Subtraction . . . . . . . . . . . . . . . . . . . . . . . . . . 126
20.3.4 Division and Square Root: Later . . . . . . . . . . . . . . 127
8 CONTENTS
20.4 Homework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
21 Solutions for sixth weeks assignments 129
21.1 Problem set 3.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
21.2 Problem set 3.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
21.3 Problem set 3.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
21.4 Problem set 3.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
22 Notes for the seventh week: primes, factorization, and modular
arithmetic 135
22.1 Divisibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
22.2 Primes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
22.3 Factorization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
22.4 Modular Arithmetic . . . . . . . . . . . . . . . . . . . . . . . . . 139
22.5 Divisibility Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
22.6 Homework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
23 Solutions for seventh weeks assignments 143
23.1 Problem set 4.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
23.2 Two diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
23.3 Problem set 4.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
23.4 A familiar incomplete integer . . . . . . . . . . . . . . . . . . . . 146
24 Notes for the eighth week: GCD, LCM, and ax +by = c 147
24.1 Modular arithmetic . . . . . . . . . . . . . . . . . . . . . . . . . . 147
24.2 Divisibility rules . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
24.3 Greatest common divisor . . . . . . . . . . . . . . . . . . . . . . 149
24.4 Least common multiple . . . . . . . . . . . . . . . . . . . . . . . 150
24.5 Euclidean GCD algorithm . . . . . . . . . . . . . . . . . . . . . . 151
24.6 Linear Diophantine equations . . . . . . . . . . . . . . . . . . . . 152
24.7 Homework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
25 Solutions for eighth weeks assignments 157
25.1 Problem set 4.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
25.2 Computing GCDs . . . . . . . . . . . . . . . . . . . . . . . . . . 158
25.3 Computing LCMs . . . . . . . . . . . . . . . . . . . . . . . . . . 159
25.4 Linear Diophantine equations . . . . . . . . . . . . . . . . . . . . 159
26 Notes for the ninth week: ax +by = c, fractions 161
26.1 Linear Diophantine equations . . . . . . . . . . . . . . . . . . . . 161
26.1.1 In general. . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
26.1.2 The other example . . . . . . . . . . . . . . . . . . . . . . 164
26.2 Into real numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
26.2.1 Operator precedence . . . . . . . . . . . . . . . . . . . . . 165
26.3 Rational numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
26.4 Review of rational arithmetic . . . . . . . . . . . . . . . . . . . . 167
CONTENTS 9
26.4.1 Multiplication and division . . . . . . . . . . . . . . . . . 167
26.4.2 Addition and subtraction . . . . . . . . . . . . . . . . . . 169
26.4.3 Comparing fractions . . . . . . . . . . . . . . . . . . . . . 170
26.5 Complex fractions . . . . . . . . . . . . . . . . . . . . . . . . . . 171
26.6 Homework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
27 Solutions for ninth weeks assignments 175
27.1 Diophantine equations . . . . . . . . . . . . . . . . . . . . . . . . 175
27.2 Problem set 6.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
27.3 Problem set 6.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
27.4 Problem set 6.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
28 Notes for the tenth week: Irrationals and decimals 179
28.1 Real numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
28.2 Exponents and roots . . . . . . . . . . . . . . . . . . . . . . . . . 180
28.2.1 Positive exponents . . . . . . . . . . . . . . . . . . . . . . 180
28.2.2 Zero exponent . . . . . . . . . . . . . . . . . . . . . . . . 181
28.2.3 Negative exponents . . . . . . . . . . . . . . . . . . . . . . 182
28.2.4 Rational exponents and roots . . . . . . . . . . . . . . . . 183
28.2.5 Irrational numbers . . . . . . . . . . . . . . . . . . . . . . 184
28.3 Decimal expansions and percentages . . . . . . . . . . . . . . . . 185
28.3.1 Representing rationals with decimals . . . . . . . . . . . . 186
28.3.2 The repeating decimal expansion may not be unique! . . . 189
28.3.3 Rationals have terminating or repeating expansions . . . 189
28.3.4 Therefore, irrationals have non-repeating expansions. . . . 190
28.3.5 Percentages as rationals and decimals . . . . . . . . . . . 191
28.4 Fixed and oating-point arithmetic . . . . . . . . . . . . . . . . . 191
28.4.1 Rounding rules . . . . . . . . . . . . . . . . . . . . . . . . 192
28.4.2 Floating-point representation . . . . . . . . . . . . . . . . 193
28.4.3 Binary fractional parts . . . . . . . . . . . . . . . . . . . . 194
28.5 Homework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
29 Second exam and solutions 199
30 Third exam, due 1 December 201
31 Third exam solutions 203
32 Final exam 205
IV Resources 207
33 Math Lab 209
34 On-line 211
34.1 Educational Standards . . . . . . . . . . . . . . . . . . . . . . . . 211
10 CONTENTS
34.2 General mathematics education resources . . . . . . . . . . . . . 211
34.3 Useful software and applications . . . . . . . . . . . . . . . . . . 212
Part I
Introduction
11
Chapter 1
Syllabus
1.1 Concepts of Modern Mathematics I
Initial home page: http://jriedy.users.sonic.net/VI/math202-f08/
Meets MWF 9.00am-9.50am in room 210 on the second oor of the J. F. Hicks
Memorial Library.
The original syllabus is available, and notes will be posted as available.
Homework problems are posted in each sessions notes.
1.2 Goals
Gain practice in mathematical reasoning and problem solving.
Review material relevant for elementary education.
Fit mathematics into its historical and practical contexts.
1.3 Instructor: Jason Riedy
email: Jason Riedy <[email protected]>
instant messages (sometimes): [email protected]
oce hours: MW 1.30pm-2.30pm in the Math Lab (see Section 33 below)
or by appointment. Or you may nd me many afternoons at Java Js in
Bristol or Zazzyz in Abingdon.
13
14 CHAPTER 1. SYLLABUS
1.4 Text
Long, Calvin T. and DeTemple, Duane W. Mathematical Reasoning for Elemen-
tary Teachers, fourth edition. Addison Wesley, 2005. ISBN 0-321-28696-0
1.5 Grading
Standard 10-point scale, 3 points on either side for -/+ grades.
The homework is 20%, three mid-term exams are 20% each, and the nal counts
for 40%. This adds to 120%; the nal counts as two 20% scores, and the lowest
20% score is dropped.
1.6 On homework
Some problems will be given in every class. The weeks problems will be collected
on the following Monday.
Mathematics is a social endeavour. Groups are encouraged, but everyone must
turn in their own work. At some point, you will be asked to present a homework
problem, its solution, and your reasoning to the class.
Also, there may be solutions available for problems. But try tackling the problem
yourself (or with your group) rst. Practice is important.
Write out sentences and not sequences of expressions. Explain your approaches.
This class is as much about the reasoning process as the results.
1.7 Submitting homework
Groups are ne, turn in your own work. Homework is due in or before class on
Mondays.
Note that you may email homework. However, I dont use Microsoft
TM
products
(e.g. Word), and software packages are notoriously nicky about translating
mathematics.
If youre typing it (which I advise just for practice in whatever tools you use),
you likely want to turn in a printout. If you do want to email your submission,
please produce a PDF or PostScript document.
Chapter 2
Syllabus schedule
Chapters 1 and 2 Scheduled for 18 August through 17 September. Roughly
2.5 weeks on chapter 1, then 1.5 weeks on chapter 2.
First exam Scheduled for 19 September.
Chapters 3, 4, and 5 Scheduled for 22 September through 29 October.
Second exam Scheduled for 31 October.
Chapters 6 and 7 Scheduled for 3 November through 26 November.
Third exam Scheduled for 28 November.
Review Scheduled for 1 and 3 December.
Final exam Ocial time: Saturday, 6 December from 3.30pm until 5.30pm.
15
16 CHAPTER 2. SYLLABUS SCHEDULE
Part II
Notes for chapters 1 and 2
17
Chapter 3
Notes for 18 August
Notes also available as PDF.
3.1 Syllabus and class mechanics
The original syllabus is available.
3.2 Introductions
Backgrounds
Goals? (i.e. which grade levels?)
Remember: Im a mathematician, not an elementary educator. help keep me on
track for what you need.
Reminder that elementary is not a judgement, its a description. when a
mathematician or scientist uses elementary, they mean to understand or solve
a problem almost from rst principles.
(e.g. Elementary functions like sine and cosine cannot be built simply from other
functions.)
3.3 First homework
After checking the table of contents, for which topics in chaps 2-7 would you like
the most review?
19
20 CHAPTER 3. NOTES FOR 18 AUGUST
Via email, please. Id like to have them while planning the next few classes.
3.4 Problem solving
Goal is to build up mathematical common sense
3.4.1 Categories
Problems to nd
Problems to prove
Emphasis for early years is the rst, problems to nd.
3.4.2 Polyas stages
Understand the problem
Divise a plan
We will explore a taxonomy of plans
Carry out the plan
Examine the solution
Continuing example from Plya; uses the geometry of right triangles. A bit past
elementary level, but it illustrates the right aspects.
3.4.3 Understand the problem
What is known, the data?
What is unknown?
What is the relation or condition linking the data with the unknown?
initial guess: is it possible to satisfy the condition?
Find the diagonal of a box (rectangular parallelpiped, like the classroom) of
which the length, width, and height are known.
draw the gure
(dialog not including additional prodding, etc)
Me What is the unknown?
class length of diagonal
3.4. PROBLEM SOLVING 21
Me What is the data?
class length, width, and height
Me Help me write it. What notation should be use?
class x for the unknown, a, b, c for the sides.
Me So what is the relationship between x, a, b, and c in words? Dont be fancy,
etc.
class x is the diagonal of the box with side lengths a, b, and c.
Me Does the problem sound reasonable? Can the condition be satised at all?
class yes, a, b, and c completely determine the box, and so also the diagonal
3.4.4 Devise a plan
Do you know of a related problem?
(tea kettle joke about reducing to an already solved problem)
Look at the unknown. Do you know of a similar problem with the same type of
unknown? Its a length.
If remembering length of a hypotenuse, good!
Otherwise: can you think of a similar but simpler problem?
Keep close to the actual problem by checking that all relevant data are used,
and that the entire condition is used.
(most explicit hint to use: Can you nd a triangle?)
3.4.5 Carry out the plan
Add additional notation for the new length, d.
Apply Pythagorus.
Smaller triangle: a
2
+b
2
= d
2
.
Apply to larger triangle: d
2
+c
2
= x
2
.
So: a
2
+b
2
+c
2
= x
2
.
3.4.6 Looking back
Can you check or verify the result? Test it?
Did you use all the data?
22 CHAPTER 3. NOTES FOR 18 AUGUST
Can you reach the same conclusion dierently?
Does this reduce to the simpler case? Do the dimensions match?
Can you think of similar problems to which this may apply? (e.g. distance from
the center to a corner)
Chapter 4
Notes for 20 August
Notes also available as PDF.
4.1 Review
4.1.1 Problem solving
Goal is to build up mathematical common sense
4.1.2 Categories
Problems to nd
Problems to prove
4.2 Todays goal: Problem solving principles
Review the principles
Two tactics: Guessing and tabling.
4.3 Polyas principles
These are principles and not a recipe. Use these to form a problem-solving
plan. (problem solving as a problem. . . )
23
24 CHAPTER 4. NOTES FOR 20 AUGUST
Understand the problem
Divise a plan
We will explore a taxonomy of plans
Carry out the plan
Examine the solution
4.3.1 Understand the problem
Often helps to rephrase a few ways.
In English (or whatever is appropriate)
With mathematical notation
Determine what may be relevant.
Sketch the problem graphically, with numbers, with physical items. . .
Whatever works for you on this problem.
Decide if the problem may have a reasonable solution.
4.3.2 Divise a plan
Taxonomy of plans in the text.
Will explore shortly.
The last be ingenious. . . Dont take it seriously. Far more work is
accomplished by being systematic.
The ingenious part comes with practice, building up connections in your
mind.
Sure Im lucky. And the more I practice, the luckier I get. Gary
Player, golfer
One goal: Turn a strategy into a list of questions to ask students.
When I think about the strategy, I consider what I want to write about it.
A major goal of mathematics is communicating ideas well.
4.3.3 Carry out the plan
Details depend on the plan. . .
May work, may not work.
4.4. TWOCLOSELYRELATEDTACTICS, GUESSINGANDMAKINGALIST25
Dead-ends are common when approaching new styles of problems.
Moving through the plan requires attention to detail.
Mathematicians and scientists grow to loath +/- signs.
Eventually, you learn which details can be xed later.
4.3.4 Examine your solution
Very, very important.
Check your results somehow, possibly by varying the problem a little.
Try to generalize a little.
Interpret your results. Often provides a check in itself, or leads to an
alternate route.
Common in mathematics: First publication of a result is long and
hairy. Interested people being interpreting it, and a short or more
direct proof is found.
Erdos and the book proof.
4.4 Two closely related tactics, guessing and mak-
ing a list
Both are forms of searching
Guessing: heuristic
Tabling: methodical
4.4.1 Guessing
Begins as a hunch
A few examples later becomes a guess
More examples (tting into a mathematical dialog) becomes a conjecture.
4.4.2 Example 1.3 from the text
Note: Phrases like pretty clear and obvious are dangerous.
First problem to solve: What is the problem?
26 CHAPTER 4. NOTES FOR 20 AUGUST
Want to extend the pattern. What is the pattern?
Rephrase the problem. Variables?
Not for elementary students, but it helps to stay ahead of them. . .
Before guessing look for relationships. A little larger here a little smaller
there.
Now start guessing and checking.
Use the relationships, and look for more. This helps target your
guesses.
That is the search. . .
Now have two examples. Look back at what weve done, and look ahead
to what we can do.
Look for other relationships.
Construct another problem with a solution
4.4.3 Tabling / Making a list
Search through the space of answers methodically.
Ensures you wont miss anything.
Blind table construction can be long. . .
4.4.4 Example 1.4 from the text: a counting problem.
When you read it, note the follow,
The technique is to be methodical in constructing the table.
Take care to chose one method and follow it.
If you remember truth tables from logic, same idea.
Example 1.4, page 15 (?):
Make an orderly list
How many dierent total scores could you make if you hit the dart-
board shown with three darts?
(Three nested circles. Scores 10 for the innermost, 5 for the middle,
and 1 for the outer ring.)
Understand the problem
4.4. TWOCLOSELYRELATEDTACTICS, GUESSINGANDMAKINGALIST27
Three darts hit the dartboard and each scores a 1, 5, or 10. The
total score is the sum of the scores for the three darts. There could
be three 1s, two 1s and a 5, one 5 and two 10s, and so on. The fact
that we are told to nd the total score when throwing three darts is
just a way of asking what sums can be made using three numbers,
each of which is either 1, 5, or 10.
Devise a plan
If we just write down sums hit or miss, we will almost surely overlook
some of the possibilities. Using an orderly scheme instead, we can
make sure that we obtain all possible scores. Lets make such a list.
We rst list the score if we have three 1s. then two 1s and one 5,
then two 1s and no 5s, and so on. In this way, we can be sure that
no score is missed.
Carry out the plan
Number of 1s Number of 5s Number of 10s Total Score
3 0 0 3
2 1 0 7
2 0 1 12
1 2 0 11
1 1 1 16
1 0 2 21
0 3 0 15
0 2 1 20
0 1 2 25
0 0 3 30
The possible total scores are listed.
Look back
Here the key to the solution was in being very systematic. We were
careful rst to obtain all possible scored with three 1s, then two 1s,
then no 1s. With two 1s there could be either a 5 or a 10 as shown.
For one 1 the only possibilities are two 5s and no 10s, one 5 and one
10, or no 5s and two 10s. Constructing the table in this orderly way
makes it clear that we have not missed any possibilities.
My additional notes:
Note the similarity with numbers. Consider listing all numbers that t the
following two properties:
1. Use only the digits 0, 1, 2, and 3.
2. Have at most three digits.
Then drop those numbers whose digits do not add to three.
28 CHAPTER 4. NOTES FOR 20 AUGUST
Thats another way to construct a table like this: Consider a larger table where
it is easier to be systematic, then remove numbers that do not t the problem.
4.4.5 Dierent example to show bisection
Modied problem 12.a from the texts chapter review exercises.
Geometric progression Sequence of numbers dened by a starting number
and a constant. The second number is generated by multiplying by the
constant, the third by multiplying again.
Consider the sequence where 3 is the starting number and two is the constant.
First: 3 = 3
Second: 6 = 3 2
Third: 12 = 6 2 = 3 2
2
and so forth.
Which term in the sequence is 768?
Solve by a list? Could be long.
What is 2
10
? What is 3 2
10
?
Know third and 11th term. Third is smaller, 11th is larger.
Which term to try next?
Half-way is the 7th term: 2
6
3 = 192
Now what region? (> 7th, < 11th)
9th: 768. DONE
Calculated three terms (11th, 7th, 9th) rather than ve.
4.5 Next time: More problem solving ideas.
4.6 Homework
Groups are ne, turn in your own work. Homework is due in or before class on
Mondays.
Write out (briey) your approach to each problem.
Problem set 1.1:
Problems 10, 13 (Hint: nd a way to make a smaller table)
4.6. HOMEWORK 29
Construct an example like 1.3 where there is no solution. Explain what
lead you to its construction. (Hint:
Problem set 1.2:
Problems 5, 7, 8
Consider solving Example 1.3 with a table starting at a=1, b=1, c=1. How
long would the table be if you step through the choices? How many entries
would you check if you bisect the choices? (Hints:
1. Dont make the lists, but rather count the steps in the method for
making the list.
2. Or just go ahead and use a program or spreadsheet to build the lists.)
Note that you may email homework. However, I dont use Microsoft
TM
products
(e.g. Word), and software packages are notoriously nicky about translating
mathematics.
If youre typing it (which I advise just for practice in whatever tools you use),
you likely want to turn in a printout. If you do want to email your submission,
please produce a PDF or PostScript document.
30 CHAPTER 4. NOTES FOR 20 AUGUST
Chapter 5
Notes for 22 August
Notes also available as PDF.
5.1 Review
5.1.1 Polyas problem solving principles
Principles are not a recipe, but they are a way to cook.
Understand the problem
Rephrase it, play with examples, sketch it, determine your goals
Divise a plan
Covered two search-based plans, guessing and tabling
Carry out the plan
Requires attention to detail
Examine the solution
Check your result, generalize, try dierent problems
5.1.2 Tactic: Guessing
Good for nding a solution satisfying a relationship or emulating an
example.
While rephrasing the problem, look for relationships to guide the guesses.
If stuck, try related or simpler examples.
31
32 CHAPTER 5. NOTES FOR 22 AUGUST
Looking back at the problem can help you target guesses better on similar
future problems.
Examples:
Number games
Getting started on just about any problem. Guessing helps nd
examples while understanding the problem.
5.1.3 Tactic: Tabling / Making an orderly list
Good for listing problems and some counting problems.
Find the number of ways to make change given certain coins.
Can be good for satisfying relationships when guesses run dry.
Search through the space of answers methodically.
Ensures you wont miss anything.
Takes attention to detail in the plan of making the table.
The simpler the plan, the fewer mistakes in execution.
Examples:
Logic tables
Counting ways to make a particular sum
(Amusingly, this is a computationally dicult problem. There
is no known method that is fundamentally better than listing the
possibilities.)
5.1.4 Example of creating a list
Problem 11 in set 1.2:
When Anita made a purchase she gave the clerk a dollar and received
21 cents in change. Complete this table to show what Anitas change
could have been.
# dimes # nickles # pennies
2 0 1
Why does the list start with one penny?
Only way to make any change ending in 1.
Can you use that to simplify the problem?
5.1. REVIEW 33
Use 20 cents rather than 21, then add 1 to the last column.
Removing extraneous details can help simplify a table-making method.
What now? Decide on a systematic method for lling in the table.
Methods for making methods:
Find a rule each row of the table must satisfy.
A conserved quantity is good, here the total amount.
Find rules relating the columns.
How many nickles per dime? Pennies per. . . ?
Decide on a transformation style from one row to the next.
Try to push from one side to the other.
Will need to back-track sometimes.
Look for groups to separate the table, possibly nested.
How many dimes can be used? Each number determines a group.
Within each dime group, how many nickels?
Denes a recursive function, may recall from Math 131.
Sometimes careful construction leads to a direct solution.
Not this time, at least not without some number theory. (See
linear Diophantine equations and how to solve them.)
Extra idea: Extend the table by one column to have a total. Check along
the way.
The goal of self-checks is not only to nd errors earlier but also to
help locate the error.
Designing good checks is an art in itself, but can help with the table
design.
An aphorism from computer programming:
Everyone knows that debugging is twice as hard as writing a program
in the rst place. So if youre as clever as you can be when you write
it, how will you ever debug it? Brian Kernighan (C, Unix, awk,
. . . )
Ideal end results:
One dime becomes two nickels, one nickel becomes ve pennies
Each dime quantity denes a group. Each nickel quantity denes a (fairly
trivial) group.
34 CHAPTER 5. NOTES FOR 22 AUGUST
Within a group, push change across the row.
No number means a zero.
# dimes # nickles # pennies total cents = 20
2 2*10 = 20
1 2 10+2*5 = 20
1 1 5 10+5+5*1 = 20
1 10 10+10*1 = 20
4 4*5 = 20
3 5 3*5+5*1 = 20
2 10 2*5+10*1 = 20
1 15 1*5+15*1 = 20
20 20*1 = 1
Remember: The real answer has one extra penny, so add 1 to the penny
column.
5.1.5 Tactic not from the text: Orderly creation of a par-
tial list
One example is bisection.
Works when there is a natural order in the result values.
Example last time was nding a term in an increasing sequence.
Pick a place to check the result value.
Consider an increasing sequence like the example in the last session:
If its too small, pick a second larger place. Lets call it A.
If still too small, repeat. Replace the old staring point A.
If too large, pick a second smaller place.
If still too large. . . Same idea.
Once you nd a bracketing point, B, consider a point half-way between
A and B.
If the half-way point is too small, replace A and repeat.
If its too large, replace B and repeat.
5.2. NEW TACTIC: DRAWING A DIAGRAM 35
5.2 New tactic: Drawing a diagram
Talk about a coincidence, although clearly the example has no relationship to
actual car models. Example 1.5 from the text, but done a little dierently.
Example 1.5: Draw a diagram
In a stock car race the rist ve nishers in some order were a Ford,
a Pontiac, a Chevy, a Buick, and a Dodge.
1. The Ford nished seven seconds before the Chevy.
2. The Pontiac nished six seconds after the Buick.
3. The Dodge nished eight seconds after the Buick.
4. The Chevy nished two seconds before the Pontiac.
In what order did the cars nish the race?
5.2.1 Understanding the problem
What information do we have?
Make of ve dierent nishers.
Only one dimension is involved: time.
Four statements relating their nishing times.
Four relative time relationships.
The Buick is mentioned most often, but every make is mentioned at least
once and no relationships are obviously repeated.
Is this enough information?
Five cars, four relationships, one dimension.
Consider: (car)-(car)-(car)-(car)-(car). Four relationships.
Could be exactly enough information.
Try rephrasing the problem.
Place ve points (F, P, C, B, and D) on a line such that
1. F is seven units to the right of C,
2. P is six units to the left of B,
3. D is eight units to the left of B, and
4. C is two units to the right of P.
36 CHAPTER 5. NOTES FOR 22 AUGUST
5.2.2 Devise a plan
All distances are relative. We need a starting point.
Pick one. B volunteers by appearing twice.
Apply all relationships with B on the right.
Then we will have two new cars placed. Use all relationships with those
on the right.
Repeat until nished or stuck.
If stuck with this method, is there any solution?
No. All times are relative. There would be at least two groups of cars
with no relationships between the groups.
5.2.3 Carry out the plan
Now we get to draw. Sorry, but Im just using tables.
Using relationship 2 and 3,
P(6) -(5) -(4) -(3) -(2) -(1) B
D(8) -(7) -(6) -(5) -(4) -(3) -(2) -(1) B
Simplifying the presentation:
D - P - - - - - B
Now we have D and P available for resolving the rules.
Only one appears on the right of a rule, P in rule 4.
Place C by rule 4:
D - P -(1) C(2) - - - B
Now C is available, so place F by rule 1:
D - P - C -(1) -(2) -(3) B(4) -(5) -(6) F(7)
Or without the counts:
D - P - C - - - B - - F
So the nal nishing order is F, then B, then C, then P, and then D.
5.2.4 Looking back
How can we check this result?
Did we place all the cars? Yes.
5.3. HOMEWORK 37
Did we use all the statements? Yes, so there can be no inconsistency.
What helped with executing the plan?
The way the statements were written let us build a concrete plan.
We could follow the chain of cars.
Variations on the problem:
Would three statements be enough?
No. One car would be left out of the relationships. Counting to
determine if a problem is soluble will return.
Would more statements guarantee a solution?
No. Not if the statements are inconsistent.
No. Some cars may not be related to others.
If the statements were consistent and not repeated, how many would
we need to guarantee a solution exists?
The total number of consistent relationships possible is a counting
problem in itself. The result is every way to choose two cars from
ve, 5 choose 2 = 5!/(2!(5-2)!) = 10, where 5! is 5 factorial
or 5*4*3*2*1.
The minimum number of relationships is 4 (from above), the
maximum is 10.
Theres a kind of order here from a threshold. If a solution is
guaranteed by K consistent relationships, it also is guaranteed
by K + 1.
You could search for counter-examples from 5 to 9, or you could
use the order here to bisect.
5.3 Homework
Groups are ne, turn in your own work. Homework is due in or before class on
Mondays.
Write out (briey) your approach to each problem.
Problem set 1.2:
Problem 13, for making a list.
Problem 18, for drawing a diagram.
38 CHAPTER 5. NOTES FOR 22 AUGUST
Problem 20, somewhat combining all of the above. Hints: Try tech-
niques on a smaller problem. Find a numerical property about the
number of edges exposed. And consider two limiting cases to govern
how many possibilities exist.
Note that you may email homework. However, I dont use Microsoft
TM
products
(e.g. Word), and software packages are notoriously nicky about translating
mathematics.
If youre typing it (which I advise just for practice in whatever tools you use),
you likely want to turn in a printout. If you do want to email your submission,
please produce a PDF or PostScript document.
Chapter 6
Solutions for rst weeks
assignments
Also available as PDF.
Note: These are my approaches to these problems. There are many ways to
tackle each.
6.1 Problem Set 1.1
6.1.1 Problem 10
Part a
Following the construction rules for one pass forward:
3 4 ?
7 7
14
What remains is to ll in the upper-right corner. The rule is that the 7 below
the ? is the sum of 4 and the missing number ?. So the nal result is
3 4 3
7 7
14
39
40 CHAPTER 6. SOLUTIONS FOR FIRST WEEKS ASSIGNMENTS
Part b
In this case, no side-by-side numbers are provided. One must start by working
on the rule from the apex backward. In the rst step, the row above 16 must
add to 16, so the missing number is 7. Subtracting 7 from the left and right
sides produces the two corner numbers. The nal gure is
0 7 2
7 9
16
Part c
Here there is not enough information to progress in either direction. A solution
comes from guessing, building a table, or nding relationships. A table would
include numbers from 0 to 9 for each of the four open spots.
For nding one solution, note the problem is symmetric. So the numbers along
the left side are equal to the numbers along the right. We can use this information
to produce a smaller table.
However, symmetry provides sucient information to work up the gure. The
two numbers above the 10 are the same and add to 10, so each is ve. Then the
corners must be four, and one solution is
4 1 4
5 5
10
Part d
Both parts a and b were formed by applying the rules directly. There was no
room for changing any of the values, so these are the only solutions.
For part c, the solution above reduced the problem until there was a unique
solution. But the original problem did not require that the numbers down each
side stay the same, so we are free to try other patterns. Given the rule for adding
adjacent entries in one row to form the next, we know we can shift numbers so
long as their sums are the same.
Pushing a one from the left to the right produces a few more examples from the
initial solution on the left:
4 1 4
5 5
10
3 1 5
4 6
10
2 1 6
3 7
10
6.1. PROBLEM SET 1.1 41
6.1.2 Problem 13
One way to create a table for this problem would be to list all sequences of
the numbers 2, 3, 4, 5, and 6. The consider mapping the sequences to the
diagrams row-wise; the rst three numbers ll the three circles in rst row, the
next number lls the next circle down, etc.
Filling in the table rst by rotating the initial sequence,
2 3 4 5 6 No, 2 + 3 + 4 = 3 + 5 + 6
6 2 3 4 5 YES, 6 + 2 + 3 = 2 + 4 + 5
5 6 2 3 4 YES, 5 + 6 + 2 = 6 + 3 + 4
This order happens to answer both part a and b almost immediately. In the
worst case, however, there are 5! = 120 rows in this table.
So two examples satisfying the relationship are
6 2 3
4
5
5 6 2
3
4
Answering part c by an exhaustive table requires all 120 rows because the answer
is no. This part also indicates how to reduce the table.
The top row of three and the middle column of three add to the same quantity.
They also share one number where they intersect. That number does not matter
at all. And the order of the remaining numbers within the row or column also
does not matter.
So each row of a full table need provide only which number is left out and then
two pairs. And note that after picking the number and the rst pair, only a
single pair remains. So chosing to drop 2 and then picking the pair (3, 5) leaves
5 and 6 which forms only one pair. There are ve numbers to drop out and six
ways to pick a pair from the remaining four numbers, so a complete table was
only 30 lines and not 120.
To check the last part, however, we know 3 will be the number left out. This
leaves only six lines to check, far fewer than 120.
left out rst pair second pair
3 2, 4 5, 6 6 = 11
2, 5 4, 6 7 = 10
2, 6 4, 5 8 = 9
4, 5 2, 6 9 = 8
4, 6 2, 5 10 = 7
5, 6 2, 4 11 = 6
Examining the sums in the table, note that one always is odd and one always is
even. Dropping 3 from the initial sequence leaves 2, 4, 5, and 6. Of those, three
42 CHAPTER 6. SOLUTIONS FOR FIRST WEEKS ASSIGNMENTS
are even and only one is odd. The sum of two even numbers is even, and the
sum of an even and an odd number is odd. So there is no way to choose two
pairs from 2, 4, 5, 6 such that both add to an even or to an odd number. Hence
the sums cannot be equal.
Sometimes taking care in reducing a table is more work, but it can make plain
relationships you can use in future problems to avoid making any table at all.
6.2 Example like 1.3 with no solution
I apologize for chopping o the hint. It suered from a cut and paste error and
should have pointed you to the even/odd idea.
Guessing and creating small tables of examples are perfectly good approaches.
Another is to use a similar even/odd relationship.
The sum of the smaller circles is half the sum of the larger circles. So if the
larger circles add to an odd value, there is no way to ll the smaller circles with
integers adding to half that odd value.
Fill the outer circles with odd number of odd numbers. The resulting sum is
odd.
Without the hint, you may simply ll the larger circles with 1, 1, and 1. Then you
need a zero somewhere in the smaller circles, and I disallowed zeros or negative
numbers during class. I did not realize that the text allowed non-integers and
negative numbers in the problem.
6.3 Problem Set 1.2
6.3.1 Problem 5
Generally, it is far easier to see the rule if the guesses are placed in order,
answering a portion of part c. Assume that each of the childrens rules are
determined by a relatively simple formula of the chosen number.
Part a
Make a table to see if the guesses could satisfy the rule:
Chosen: 0 2 4 5 8
Returned: -3 7 17 22 37
First guess: -3 7 17 22 37
Second guess: -3 7 17 22 37
6.3. PROBLEM SET 1.2 43
So the guesses may have found the rule. If the guesses really are the same, that
is more evidence. However, the problem does not provide rules for the initial
rules. A students rule could be return a random number. If the rules are
required to be linear, that is of the form ax +b where the chosen number is x,
then the ve choices above are more than the two required to determine the line.
(Two points determine a line, and each guess places a point.)
The two guesses appear the same. Consider the guesses algebraically, so the
rst guess is 5i 3 and the second is 5(i 1) + 2 where i is the chosen number.
Expanding the latter, 5i 5 + 2 = 5i 3, so the rules really are the same.
Part b
The table of provided data:
Chosen: 0 1 3 7 9
Returned: 1 2 10 50 82
The following questions could help guide your guesses,
Is the sequence formed by a line like ax +b? Such a curve grows slowly.
The next step up is a quadratic curve like a parabola. Does the data t
some simple formula ax
2
+bx +c?
The next step is a cubic with x
3
. Do the numbers increase very quickly?
Starting at zero helps; you see the constant term must be one. Subtracting one
from the other returned numbers,
Chosen: 0 1 3 7 9
Returned 1: 0 1 9 49 81
So a reasonable guess is x
2
+ 1.
Another route to this result:
The curve grows a little more quickly than we would expect from a line. Also, if
this were a line, the rst two points determine its structure as x + 1, but this
does not t the third point, 3 + 1 = 4 = 10.
A quadratic is determined by three points. From the rst, (0, 1), we see that
c = 1. Then the second (1, 2) and third (3, 10) points set up the equations
a + b = 1 and 9a + 3b = 9. We already know this cannot be a line, so a = 0.
Assuming a and b are integers, that leaves a = 1 and b = 0. So the rule may be
x
2
+ 1. That ts the other data.
44 CHAPTER 6. SOLUTIONS FOR FIRST WEEKS ASSIGNMENTS
Part c
The table of provided data:
Chosen: 0 1 2 3 4
Returned: 7 10 13 16 19
This grows slowly, so its reasonable to guess a linear relationship. The zero
term gives a constant 7, and the rst term suggests a slope of 3 so 3 1 +7 = 10.
Continuing with the other points conrms that this is a reasonable guess.
Choosing the rst few numbers starting from zero denitely helps make the
relationship clear. However, it can be useful to pick a few consecutive numbers
further away to see how quickly the function grows. In this case, we would
expect choosing 10 and 11 to result in two numbers as close together as the
results from 0 and 1 if the problem were linear. If the student had returned two
numbers much further apart, we could avoid trying to t the data to a line.
6.3.2 Problem 7
There are many straight-forward ways to build a table here.
One starts by placing all four coins in one denomination, then moves a coin to
the right:
N D Q total
4 0 0 20
3 1 0 25
3 0 1 40
2 2 0 30
2 1 1 45
2 0 2 60
1 3 0 35
1 2 1 50
1 1 2 65
1 0 3 80
0 4 0 40
0 3 1 55
0 2 2 70
0 1 3 85
0 0 4 100
There are 15 lines, two of which are shared. So there are 14 dierent quantities
possible.
Another method would write out denominations rst and then try to form
them with some combination of coins. The smallest possible amount is 20 for
using four nickles, and the largest possible is 100 for four quarters. There are
6.3. PROBLEM SET 1.2 45
(10020)/5 = 16 possible table entries. But 75 and 90 cannot be formed, leaving
14 dierent possible quantities.
6.3.3 Problem 8
This is a classical problem used in discrete mathematics and introductory com-
puter science classes, although often starting with a dollar to make tabling less
practical.
There are 49 ways.
There are at most two quarters, at most ve dimes, at most 10 nickles, and at
most 50 pennies per line of a table with the following heading:
Pennies Nickles Dimes Quarters total
To generate the table, start with two quarters and then shift amounts over as in
other problems. You used the conserved quantity, the total amount, to guide
your next choice.
Another method for solving this problem is to set up recurrence relationships
and build a slightly dierent table.
Consider making change for an amount N. And consider four dierent ways for
making such change:
A
N
with only pennies,
B
N
with nickles and pennies,
C
N
with dimes, nickles, and pennies, and
D
N
with quarters, dimes, nickes, and pennies.
Say we start at N and the full collection of possible coins. Then either the
change contains a quarter or it does not. If it does contain with a quarter, then
we change the remaining N 25 in the same way, possibly with more quarters.
If not, then we change N no quarters. So
D
N
= C
N
+D
N25
.
Similarly,
C
N
= B
N
+C
N10
, and
B
N
= A
N
+B
N5
.
We can begin constructing a table of values by N starting from the extreme
case N = 0. There is only one way of making no change at all, so A
0
= B
0
=
C
0
= D
0
= 1. There also is only one way of making change with pennies, so
A
N
= 1 for all N. And the relations above show we need only rows where N is
a multiple of ve and provide formulas for every entry.
46 CHAPTER 6. SOLUTIONS FOR FIRST WEEKS ASSIGNMENTS
N A
N
B
N
C
N
D
N
0 1 1 1 1
5 1 2 2 2
10 1 3 4 4
15 1 4 6 6
20 1 5 9 9
25 1 6 12 13
30 1 7 16 18
35 1 8 20 24
40 1 9 25 31
45 1 10 30 39
50 1 11 36 49
6.4 Consider solving Example 1.3 with a table
Note that each of a, b, and c can be at most one less than the smallest number
adjacent to them (I ruled out zeros in class). So there are at most 10 possibilities
for a, 14 for b, and 10 for c. Listing all would require 10 10 14 = 1400 entries.
But also one variable is completely determined by the other two, so we would
need a list of 10 14 = 140 possibilities for a and b (or c and b), or 10 10 = 100
for a and c.
Or we could note that each choice of a immediately determines a possible choice
of c. Then we simply verify that some b satises the sums. This leaves a total
table of 10 lines.
Assume there is a reasonable order for bisection. A worst case sequence of entries
we could check would be 1, 10, 5, 7, 8, 9; so 6 entries total. The actual sequence
here would be 1, 10, 5, 7, 6; so 5 entries.
6.5 More in Problem Set 1.2
6.5.1 Problem 6: whoops, not assigned
Whoops. I solved this but never assigned it. Keeping the solution
here for examples.
There are many approaches. I provide a few overly clever ones below.
The rst two parts only require following the rules to ll the triangles.
Part a only requires following the rules to ll the triangle:
8
7 1
11 4 5
6.5. MORE IN PROBLEM SET 1.2 47
Part b is similar, but requires lling in 8 = 19 11:
19
8 11
9 1 12
The next two fall to the same tabling technique outlined in the solution above
with some minor modications.
For part c, the outer sum is odd, so there is no integer solution. The text
apparently allows non-integer solutions. Then the following triangle suces and
can be found by doubling all the entries, solving, and halving all the entries
again:
7
2.5 4.5
11 8.5 13
In part d, the negative numbers can be dealt with by adding a suciently large
number to the vertices of the triangle, solving the problem, and then subtracting
half that number from the intermediate nodes.
The rst number we add should be even to maintain the even/odd balance
(even+even is even, even+odd is odd, so adding an even doesnt change the
number of evens and odds). And adding a number larger than the largest given
number will shift all the intermediates above zero. So we chose the smallest even
number larger than 11 and hence add 12 across the outer nodes. Solving:
10
3 7
19 16 23
Subtracting half 12 (or 6) from the intermediate notes gives a solution to the
original problem:
-2
-3 1
7 10 11
6.5.2 Problem 13
The area is the product of the length and width. Knowing the lengths are whole
numbers, we can write a list of all integers that divide into 120 cleanly. We need
only include half the entries; the length and width are interchangable.
The perimeter is twice the sum of the length and width. Including the perimeter
in the table solves the second part of the problem.
48 CHAPTER 6. SOLUTIONS FOR FIRST WEEKS ASSIGNMENTS
length width area perimeter
1 120 120 242
2 60 120 124
3 40 120 86
4 30 120 68
5 24 120 58
6 20 120 52
8 15 120 46
10 12 120 44
So the shape with the smallest perimeter is the nearly square 10 12. This is
true in general: The shape with the least perimeter or surface area enclosing the
greatest volume is the closest to a circle or sphere. Here, a square is as close to
a circle as you can get with four sides, and 10 12 is as close as you can get to
a square.
6.5.3 Problem 18
Drawing diagrams for smaller problems reveals the sequence in the following
table:
Sections on each side Number of posts
1 4
2 8
3 12
4 16
The relationship above is that there are four times as many posts as sections, so
the nal result is 40 posts for 10 sections.
6.5.4 Problem 20
Each square has four edges, and joining two squares along one edge removes one
from both but two from their sum. Hence all possible perimeters all are even.
What is the largest number of exposed edges?
Simple counting arguments give an overestimate. Given nine squares, there can
be at most 36 edges exposed, or four for each square. But each square must be
joined to at least one other. Again, joining reduces the number of exposed edges
by one per square. Each square must be joined, so each contributes at most
three edges to the total. So there can be at most 27 edges exposed, or 26 taking
into account the previous paragraph.
But they squares are all connected. The form hiding the fewest edges occurs
when placing all squares in a line. Shifting around squares from this shape only
hides more edges, so we know the largest perimeter is 20.
6.5. MORE IN PROBLEM SET 1.2 49
And what is the smallest number? There must be at least four edges exposed,
as all the squares are linked and there must be at least four larger edges of at
least one edge each.
In general, the most circle-like gure has the least exposed surface. In this
case, the shape would be square, exposing 12 edges. Shifting any block around a
square exposes more edges, so the square is the smallest with a perimeter of 12.
The remaining problem is to take the largest (or smallest) shape and shift blocks
around to form all the intermediates. Forming each of 14, 16, and 18 is fairly
simple.
50 CHAPTER 6. SOLUTIONS FOR FIRST WEEKS ASSIGNMENTS
Chapter 7
Notes for 25 August
Notes also available as PDF.
7.1 Review
7.2 Draw a diagram, follow dependencies
Talk about a coincidence, although clearly the example has no relationship to
actual car models. Example 1.5 from the text, but done a little dierently.
Example 1.5: Draw a diagram
In a stock car race the rist ve nishers in some order were a Ford,
a Pontiac, a Chevy, a Buick, and a Dodge.
1. The Ford nished seven seconds before the Chevy.
2. The Pontiac nished six seconds after the Buick.
3. The Dodge nished eight seconds after the Buick.
4. The Chevy nished two seconds before the Pontiac.
In what order did the cars nish the race?
7.2.1 Understanding the problem
What information do we have?
Make of ve dierent nishers.
Only one dimension is involved: time.
51
52 CHAPTER 7. NOTES FOR 25 AUGUST
Four statements relating their nishing times.
Four relative time relationships.
The Buick is mentioned most often, but every make is mentioned at least
once and no relationships are obviously repeated.
Is this enough information?
Five cars, four relationships, one dimension.
Consider: (car)-(car)-(car)-(car)-(car). Four relationships.
Could be exactly enough information.
Try rephrasing the problem.
Place ve points (F, P, C, B, and D) on a line such that
1. F is seven units to the right of C,
2. P is six units to the left of B,
3. D is eight units to the left of B, and
4. C is two units to the right of P.
7.2.2 Devise a plan
All distances are relative. We need a starting point.
Pick one. B volunteers by appearing twice.
Apply all relationships with B on the right.
Then we will have two new cars placed. Use all relationships with those
on the right.
Repeat until nished or stuck.
If stuck with this method, is there any solution?
No. All times are relative. There would be at least two groups of cars
with no relationships between the groups.
7.2.3 Carry out the plan
Now we get to draw. Sorry, but Im just using tables.
Using relationship 2 and 3,
P(6) -(5) -(4) -(3) -(2) -(1) B
D(8) -(7) -(6) -(5) -(4) -(3) -(2) -(1) B
7.2. DRAW A DIAGRAM, FOLLOW DEPENDENCIES 53
Simplifying the presentation:
D - P - - - - - B
Now we have D and P available for resolving the rules.
Only one appears on the right of a rule, P in rule 4.
Place C by rule 4:
D - P -(1) C(2) - - - B
Now C is available, so place F by rule 1:
D - P - C -(1) -(2) -(3) B(4) -(5) -(6) F(7)
Or without the counts:
D - P - C - - - B - - F
So the nal nishing order is F, then B, then C, then P, and then D.
7.2.4 Looking back
How can we check this result?
Did we place all the cars? Yes.
Did we use all the statements? Yes, so there can be no inconsistency.
What helped with executing the plan?
The way the statements were written let us build a concrete plan.
We could follow the chain of cars.
Following dependencies often is called working backwards.
Variations on the problem:
Would three statements be enough?
No. One car would be left out of the relationships. Counting to
determine if a problem is soluble will return.
Would more statements guarantee a solution?
No. Not if the statements are inconsistent.
No. Some cars may not be related to others.
If the statements were consistent and not repeated, how many would
we need to guarantee a solution exists?
The total number of consistent relationships possible is a counting
problem in itself. The result is every way to choose two cars from
54 CHAPTER 7. NOTES FOR 25 AUGUST
ve, 5 choose 2 = 5!/(2!(5-2)!) = 10, where 5! is 5 factorial
or 5*4*3*2*1.
The minimum number of relationships is 4 (from above), the
maximum is 10.
Theres a kind of order here from a threshold. If a solution is
guaranteed by K consistent relationships, it also is guaranteed
by K + 1.
You could search for counter-examples from 5 to 9, or you could
use the order here to bisect.
7.3 Look for a pattern
Incredibly important at many levels.
Mathematics generally is about characterizing patterns.
Immediately useful for
extending from provided data, and
continuing relationships found when considering problems.
7.3.1 Example: What is the last digit of 7
100
The initial problem is straight-forward; there appears to be little more to
understand. One useful relationship is that there are at most 9 possible nial
digits. (Zero is not possible.)
With so few possible digits, a good initial plan is forming a table:
Number Last digit
7
1
7
7
2
9
7
3
3
7
4
1
7
5
7
7
6
9
.
.
.
.
.
.
We certainly dont want to extend this to 7
100
. However, note that the last digit
of 7
4
is 1. Then 7 1 = 7 begins the pattern anew.
To check, we could guess that the last digit of 7
8
is 1. Continuing the table
conrms the guess.
7.4. PATTERNS AND REPRESENTATIVE SPECIAL CASES 55
So 7
i
has the last digit 1 for all i that are multiples of four. And thus the last
digit of 7
100
is 1.
7.4 Patterns and representative special cases
A special case is quite literally a case set aside as special.
A representative special case is a special case that accurately represents
the general case.
Consider teaching a property over all cases by illustrating with a specic
case.
The specic case must not depend on which case is used. . .
Will use texts example: sums of rows of Pascals triangle
56 CHAPTER 7. NOTES FOR 25 AUGUST
7.4.1 Sums over Pascals triangle
A wonderful source of examples and relationships
Related to polynomials, probability distributions, and even fractals.
Written in rather boring table form, each entry is the sum of the entry directly
above and above to the left:
#
0 1
1 1 1
2 1 2 1
3 1 3 3 1
4 1 = 0 + 1 4 = 1 + 3 6 = 3 + 3 4 = 3 + 1 1 = 1 +0
Problem: What is the sum of the 20
th
row? The 200
th
?
Understanding the problem:
How is the triangle formed?
Each entry is the sum of the two above.
Can we continue this pattern? Yes.
Do we want to write 200 rows? No.
What else can we do?
Plan:
Form the sums.
Look for a pattern.
Reason about all cases given a representative case.
The result:
# sum
0 1 1
1 1 1 2
2 1 2 1 4
3 1 3 3 1 8
4 1 4 6 4 1 16
A guess:
Each rows sum is twice the previous rows sum.
Starting from the zeroth row as 1, the n
th
row is 2
n
.
7.5. HOMEWORK 57
The next few rows check.
Consider a specic case, forming the fourth row:
# sum
3 1 3 3 1 8
4 1 = 0 + 1 4 = 1 + 3 6 = 3 + 3 4 = 3 + 1 1 = 1 +0 16
What do we know about the sum of each row?
Its the sum of each entry.
In turn, each entry is the sum of two entries from the previous row.
So the sum must be the sum of sums of pairs from the previous row,
or twice the previous rows sum.
By looking at a special case but applying only general reasoning, we have proven
that each rows sum is twice the previous rows sum.
We used no properties specic to the third or fourth rows.
We could have chosen (without loss of generality) any row and applied
the same reasoning.
Thus the reason applies to all rows.
And the nal answers:
2
20
= 1 048 576. A megabyte (MiB in correct units, not MB) is 2
20
bytes.
2
200
has 61 digits, none of which are particularly elucidating. For this style
of problem, leaving 2
200
unevaluated is much better. But, for completeness,
the answer is in the notes.
1 606 938 044 258 990 275 541 962 092 341 162 602 522 202 993 782 792 835 301 376
7.5 Homework
Groups are ne, turn in your own work. Homework is due in or before class on
Mondays.
Write out (briey) your approach to each problem.
Patterns: What is the 87
th
digit past the decimal point in the expansion
of 1/7?
Patterns and related patterns: Using the result of 7
100
above, what is the
last digit of 3
100
?
From problem set 1.3:
58 CHAPTER 7. NOTES FOR 25 AUGUST
Arithmetic progressions: problem 6 (see the text, esp. around exam-
ples 1.8 and 1.9)
Problem 9, and feel free to criticize the use of likely or probably
here as well.
Problem 11.
Problem 20.
Note that you may email homework. However, I dont use Microsoft
TM
products
(e.g. Word), and software packages are notoriously nicky about translating
mathematics.
If youre typing it (which I advise just for practice in whatever tools you use),
you likely want to turn in a printout. If you do want to email your submission,
please produce a PDF or PostScript document.
Chapter 8
Notes for 27 August
Notes also available as PDF.
8.1 Review
A special case is quite literally a case set aside as special.
A representative special case is a special case that accurately represents
the general case. with a specic case.
The specic case must not depend on which case is used. . .
We proved that the sum of the i
th
row is 2
i1
.
Will return to proofs next time. Today, two tactics / principles and
discussion of general mathematical reasoning.
8.2 Ruling out possibilities
Weve already seen this in building tables, guessing, etc.
If the result must be even, dont list odd numbers.
In the diagram where we lled in circles along the sides of a triangle
to equal integers in the vertices, we never needed to guess larger than
the smallest integer.
Generally, use whatever rules apply to reduce your problem.
In an extreme case, only one result may be left standing.
59
60 CHAPTER 8. NOTES FOR 27 AUGUST
8.2.1 Logic puzzles
Classical problem structure. Provide lists of items along with properties
linking the lists. Then derive the only possible matching between the lists.
Zillions of logic puzzles available on-line or in puzzle magazines.
Each of Bill, John, Fred, and Jim are married to one of Judy, Gretchen,
Margie, and Loretta.
1. Judys husbands name does not begin with J.
2. Margies husbands name has the same letter twice.
3. The name of Lorettas husband has three letters.
General strategy: Form a table to track possibilities.
Fill in what is known from each rule (noted (1), (2), (3) below).
Cross out possiblities that cannot occur (similar notation).
What possibilities remains? Judy and Fred
That leaves one spot left, John and Gretchen
Judy Gretchen Margie Loretta
Bill (2) (2) Yes(2) (2)
John No(1) Yes (2) (3)
Fred Yes (2) (3)
Jim No(1) (3) (2) Yes(3)
For an example that does not work, see problem 8 in set 1.4 (page 49). Whoever
wrote that has no idea what is in a decent banana split or double-dip cone.
8.3 The pigeonhole principle
If there are more pigeons than pigeonholes, then at least one hole
holds more than one pigeon.
Thought to have been presented rst by Dirichlet in 1834 as the shelf principle.
Useful for proving or demonstrating a fact through counting.
Assume there are 14 black socks and 6 white socks in the drawer. Without
looking, how many socks must you retrieve to have two of the same color?
How might you consider solving this? Write out all possibilities for
each number of socks you retrieve. Yuck.
Here there are two holes, black and white.
You need draw three socks to guarantee two of the same color!
8.4. MATHEMATICAL REASONING 61
Frighteningly powerful principle.
Ignoring baldness, there are at least two people in the tri-cities area
with the same number of hairs on their head.
There are about 150 thousand hairs in a typical head of hair.
According to the 2000 census, there are 480 091 people in the
tri-cities area.
If we assume no one has over 480 000 hairs, and each hair count
is a category / pigeonhole, then there must be two people with
the same number of hairs.
A lossless le compression system must expand some les.
Compressed les are pigeonholes.
There are fewer possible les of smaller size.
Hence if all les can compress to smaller sizes, there will be two
les that compress to the same le.
The pigeonhole principle and its variations are an indispensable tool of mathe-
matics!
There is a wonderful description and exploration of dierent phrasings from
Edgar Dijkstra:
http://www.cs.utexas.edu/users/EWD/transcriptions/EWD09xx/EWD980.html
Even more examples through the references at
http://www.maa.org/editorial/knot/pigeonhole.html
8.4 Mathematical reasoning
Two key forms of reasoning in mathematics:
Inductive Making an educated guess from prior observations.
Deductive If premises are satised, conclusion follows.
Premises also are known as hypotheses, suppositions, or other similar terms.
Typically,
problems to nd use inductive reasoning, and
problems to prove apply deductive reasoning.
62 CHAPTER 8. NOTES FOR 27 AUGUST
But nding a proof is in many ways inductive.
Problem solving so far has been inductive. Take example problems and their
solutions. Emulate the solutions on similar problems.
History from the western view:
Old example: Egyptian papyri (1900bc-1800bc)
Consisted of arithmetic tables followed by a list of worked problems.
Solve new problems by imitating previous ones.
No pure symbols; only count or calculate with real items. (Hiero-
glyphics hurt.)
Continued through to Greek times
Geometry replaced explicit counting.
No variables but rather geometric gures.
Every concept had to be illustrated geometrically.
Pythagoras constructed proportions from lengths.
However, deductive reasoning began in earnest.
Around 600bc, Greek mathematicians began discussing and prov-
ing theorems.
Euclids Elements, 300bc, developed systematic and rigorous
proofs.
Proofs complicated by the restriction to geometric gures.
Algebra, the beginning of fully abstract proofs:
500bc for Babylonians!
200ad for Greeks (Diophantus of Alexandria)
Spread widely from Persians, Muhammad ibn M us a al-khw arizm in
820ad.
(transliteration of his books title gave algebra, his name gives
algorithm)
So modern proof only became feasible 1200 years ago.
Wasnt widely adopted until the nineteenth century, 200 years ago.
So if proof seems dicult, remember that humanity took a long, long time
to develop the idea.
The concept of proof still is evolving!
Historical summary from Prof. Steven Krantz at
8.5. NEXT TIME: STRUCTURES AND KINDS OF PROOFS 63
http://www.math.wustl.edu/ sk/eolss.pdf
Remember to take great care with the premises in both forms of reasoning!
Inductive reasoning generalizes from examples.
If the examples are not appropriate, the result will be incorrect.
Say there is a line of women waiting for a rest room. You assume its the
ladys room. But if there is only one rest room. . .
8.5 Next time: structures and kinds of proofs
8.6 Homework
Groups are ne, turn in your own work. Homework is due in or before class on
Mondays.
Write out (briey) your approach to each problem.
In problem set 1.4 (p48):
For reducing the number of possibilites: problem 7
Logic puzzle: problem 9
Pigeonholes: Problems 12 (ignoring leap years), 13, 14
Note that 12 wants the number that guarantees two people have
the same birthday.
Which of these statements demonstrate inductive reasoning, and which
demonstrate deductive reasoning? Justify.
It has rained for the past week. It will rain tomorrow.
All men are mortal. Socrates is a man. Therefore, Socrates is mortal.
Satellite-based network access does not function through heavy rain.
It is raining heavily. I cannot upload the notes right now.
The next number after 3, 8, 13, 18, and 23 is 28.
Note that you may email homework. However, I dont use Microsoft
TM
products
(e.g. Word), and software packages are notoriously nicky about translating
mathematics.
If youre typing it (which I advise just for practice in whatever tools you use),
you likely want to turn in a printout. If you do want to email your submission,
please produce a PDF or PostScript document.
64 CHAPTER 8. NOTES FOR 27 AUGUST
Chapter 9
29 August: Review of
previous notes
See Chapter 8.
65
66 CHAPTER 9. 29 AUGUST: REVIEW OF PREVIOUS NOTES
Chapter 10
Solutions for second weeks
assignments
Also available as PDF.
Note: These are my approaches to these problems. There are many ways to
tackle each.
10.1 Patterns: The 87th digit past the decimal
in 1/7?
What are the digits of 1/7? A computer or calculator computes something
like 0.142857142857143. We start to see a pattern: 142857 is repeated. The
chunk has length six, so digit 6i + 1 will be 1, 6i + 2 will be 4, and so on. Digit
87 = 6 14 + 3, so we expect the 87
th
digit will be 2.
We could prove this in a few ways if desired. One would be to carry out long
division for seven steps. On the seventh, we would see the same problem as the
initial problem, namely 1.0/7. The same problem with the same data must have
the same solution, and thus the pattern must repeat.
10.2 Patterns: Units digit of 3
100
The quick solution is to note that 3 appears in the orbit of 7
100
directly before
1, because 7 3 = 21. So in a sense which we will formalize later, 3 is the inverse
of 7 with respect to the units digit. We need merely to count backwards in the
table generated by powers of 7.
67
68 CHAPTER 10. SOLUTIONS FOR SECOND WEEKS ASSIGNMENTS
The more direct approach is to note that powers of 3 give a nal digit sequence
of 3, 9, 7, and 1 repeating after the 1. The sequence works in chunks of four,
and we see that the entry for 100 = 4 25 + 4 is 1.
10.3 Problem set 1.3
10.3.1 Arithmetic progressions: Problem 6
Part a. The number of increments by seven is (86 2)/7 = 12. To check, form
2 + 12 7 and verify it is 86.
Part b. One method is to re-apply Gausss technique (using inductive reasoning).
There are 13 entries in the sequence, 2 and then 12 increments by 7. Pairing
these up, we have 6 added terms of the form 2 + 86 = 9 + 79 = 72 + 16 =
65 +23 = 58 +30 = 51 +37 = 88 and one remaining term of 44. So the result is
6 88 + 44 = 572.
Note that the paired terms added to 88, while the lone term was 88/2 = 44.
Playing with the form above 6 88 + (1/2)88 = 88 (12 + 1)/2, recalling the
n(n + 1)/2 form for the sum from 1 to n.
Another method is to work with the summation form directly to nd
12

i=0
(2 + 7i) = 13 2 + 7
12

i=0
i
= 26 + 7
12(12 + 1)
2
= 572.
Part c. There is a slight trick to this. Often it is more obvious to start with the
zero
th
term as I have above. Then the form of the i
th
term is 2 + 7i, showing
the initial term directly. However, many people prefer counting from one. To
convert the form, substitute n 1 for i and see that 2 + 7(n 1) = 5 + 7n is
the form of the n
th
form.
Part d. One method of computing the sum again is to emulate Gausss trick
with the observation from Part a. We expect the sum to be the number of terms,
n, times the rst plus the last, 2 +(5 +7n) = 3 +7n, divided by two. So the
sum is n(7n 3)/2.
Another method is to work with summation form,
n

i=0
(2 + 7i) = (n + 1) 2 + 7
n

i=0
i
= 2(n + 1) + 7
n(n + 1)
2
.
10.3. PROBLEM SET 1.3 69
With sucient algebra, this does simplify to the other form.
10.3.2 Problem 9
Part a. Personally, Id point out that in their limited imagination, the writers
likely intend 16 to be the next number.
Part b.
n 2
n
n
2
n + 2 n
3
5n
2
+ 10n 4
1 2 2 2
2 4 4 4
3 8 8 8
4 16 14 20
Part c. There are many, many functions that could generate 2, 4, and 8. One
needs more information to decide on which function.
And the wording of the rst problem uses likely. To me, likely infers some
probability, but there is none.
Better wording would ask for the reason why the student believes the sequence
extends to their guesses.
10.3.3 Problem 11
Following the hint, consider rearranging B B R R to B R B R. That requires
one swap in the middle. The number of swaps cannot be fewer, so this is the
minimum number.
Now rearrange B B B R R R to B R B R B R. Two swaps move the left-most
R into position, B R B B R R. The rst two letters are correct, and the rest
match the previous problem. No fewer swaps could have moved an R into the
second-left-most position; there must be at least one swap for each position
moved. So we know the fewest swaps beyond the rst case is two, for a minimum
total of three.
At this point, its reasonable to guess that the number of swaps is either 2
#B1
1
or

#B1
i=1
i. Either ts the data so far. However, at each stage we moved one R
into place and then reduced to a known subproblem. This adds moves, so the
summation seems the most likely.
To verify that

#B1
i=1
i = 4 (4 + 1)/2 = 10 swaps suces, use the following:
70 CHAPTER 10. SOLUTIONS FOR SECOND WEEKS ASSIGNMENTS
Start with: B B B B B R R R R R
4 swaps later: B R B B B B R R R R
3 swaps later: B R B R B B B R R R
2 swaps later: B R B R B R B B R R
1 swap later: B R B R B R B R B R
Total number of swaps: 10
To support that this is the minimum, we rely on inductive reasoning. At each
stage, we add the minimum number of necessary swaps to the subproblem, which
in turn is solved minimally. To formalize this, we could apply mathematical
induction.
10.3.4 Problem 20
Part a. The sums are 12, 24, 48, and 64. One pattern is that the sum is four
times the repeated quantity on the anti-diagonal. Another is to note that the
repeated quantity is one larger than the upper-left corner and one smaller than
the lower-right corner.
So adding along opposite anti-diagonals is like adding the same number entries
from the main anti-diagonal. So given a block of the form
k 1 k
k k + 1
we can add the two far anti-diagonals to get (k 1) + (k + 1) = 2k. Along with
the main diagonal, the total sum is (k 1) + (k + 1) +k +k = 4k.
Part b. The sums are 54, 108, and 144.
Part c. The blocks follow a similar pattern as the previous part:
k 2 k 1 k
k 1 k k + 1
k k + 1 k + 2
Again, add along opposite anti-diagonals to match +2 and 2, +1 and 1.
So the sum must be 9k.
Part d. In each case, the sum is the number of squares times the element
repeated along the main anti-diagonal. If the square is side-length n and the
main anti-diagonal holds k, the sum is n
2
k.
10.4. PROBLEM SET 1.4 71
10.4 Problem set 1.4
10.4.1 Reducing possibilities: Problem 7
Part a. Writing out the conditions that are easy to express gives
x = 2k + 1 for some k,
x > 1,
x < 100,
x > 20,
x < 5 7 = 35, and
x = 5i for some i.
Combining the rst and last, we know x is an odd multiple of ve, or x = 5(2j +1)
for some j. The inequalities reduce to 20 < x < 35. The only possibility in this
range is x = 25, and the digits here add to seven.
Part b. Again, translating the conditions and rewriting them to be positive
rather than negated (so not even becomes odd) gives
x = 2k + 1 for some k,
x = 11i for some i,
x > 20,
x = 3j for some j, and
x < 79.
An odd multiple of three and nine can be expressed by composing the require-
ments, so x = 3 11 (2p + 1) = 33(2p + 1) for some integer p. The bounds here
are not redundant, so 20 < x < 79. The only multiples of 33 in this region are
33 and 66. Both have digits whose sums are even (divisible by two). So there
are two possible results.
Part c. Translating this is not quite as useful, giving
x = 2k for some k,
x = 3j
1
+i
1
for some j and i = 0,
x = 4j
1
+i
1
for some j and i = 0,
x 81, and
x 64.
One observation helps: An even number not divisible by four has only one factor
of 2. So the rst and third requirements become
x = 2 (2p + 1) for some p.
72 CHAPTER 10. SOLUTIONS FOR SECOND WEEKS ASSIGNMENTS
So we are looking for even numbers in 64 x 80 with only one factor of two
and that are not multiples of three. This reduces to looking for an odd z = x/2
in 32 x 40 that is not a multiple of three.
The odd non-multiples of three in this range are 35 and 37. So z is either 35 or
37, and x is either 70 or 74.
10.4.2 Logic puzzle: Problem 9
The problem assumes neither bigamy nor marriage to immediate relations is
allowed, which is reasonable in the US. The true (T) and false (F) values are
subscripted by a number if they are a direct implication of that rule. The forth
rule was useless.
Kitty Sarah Josie Anne
David F
3
F
5
T F
Will F
3
T
5
F
1
F
5
Floyd F
3
F
5
F
2
T
Gus T
3
F
3
F
3
F
3
10.4.3 Pigeonholes: Problem 12
Part a. To ll 365 days and guarantee one is repeated, you need 366 people.
As an amazing aside, though, with only 23 people the chance of two sharing a
birthday is over 50%!
Part b. The result here is twice 365 plus one, or 731.
10.4.4 Pigeonholes: Problem 13
There are only ten units digits possible, zero through nine. If one digit is repeated,
the dierence between the two numbers with the same units digit must have
units digit zero and be divisible by ten. With eleven numbers, we must have
two with the same units digit.
10.4.5 Pigeonholes: Problem 14
Part a. Note there are six ways single-digit numbers (or the units digit) can
add to ten, 1 + 9, 2 + 8, 3 + 7, 4 + 6, and 5 + 5. If we use those as pigeonholes
for the units digit, then we know one of those holes must be lled twice with
seven numbers. Now within the holes we have another set of holes with one hole
for each digit. If both holes are lled, then the sum is divisible by ten. If only
one hole is lled, then the numbers share a units digit and their dierence is
divisible by ten.
10.5. INDUCTIVE OR DEDUCTIVE? 73
Part b. With six numbers, we can ll each hole once. Any consecutive sequence
of six numbers suce, e.g. 1, 2, 3, 4, 5, 6.
10.5 Inductive or deductive?
It has rained for the past week. It will rain tomorrow.
This is inductive because it is based solely on repeated observations.
All men are mortal. Socrates is a man. Therefore, Socrates is mortal.
Deductive reasoning begins with premises and draws a conclusion as in
this example.
Satellite-based network access does not function through heavy rain. It is
raining heavily. I cannot upload the notes right now.
Again, this example sets up data and rules from which it draws a conclusion.
This is deductive reasoning.
The next number after 3, 8, 13, 18, and 23 is 28.
This could fall either way, depending on how you decide on the next number.
If you assume this is an arithmetic sequence and calculate the next term,
this could be considered deductive reasoning. Recognizing an arithmetic
sequence without further information is inductive in itself. If you just look
and see a pattern, the process is completely inductive.
Often the initial phase of problem solving, understanding the problem, is
an exercise in reasoning inductively.
74 CHAPTER 10. SOLUTIONS FOR SECOND WEEKS ASSIGNMENTS
Chapter 11
Notes for 1 September
Notes also available as PDF.
11.1 Review
Two key forms of reasoning in mathematics:
Inductive Making a guess from prior observations.
Deductive If premises are satised, conclusion follows.
Typically,
problems to nd use inductive reasoning, and
problems to prove apply deductive reasoning.
But nding a proof is in many ways inductive.
Problem solving so far has been inductive. Take example problems and their
solutions. Emulate the solutions on similar problems.
Remember to take great care with the premises in both forms of reasoning!
Inductive reasoning generalizes from examples.
If the examples are not appropriate, the result will be incorrect.
Mathematics is representative reasoning.
We model the real world and represent pieces of it with symbols (numbers,
letters, digits, etc.).
Then we reason about how the symbols interact.
Only takes a few rules to build a massive system.
75
76 CHAPTER 11. NOTES FOR 1 SEPTEMBER
Amazingly, the system often mimics some aspect of the real world.
11.2 Proof
1. You start with a guess, possibly after seeing a pattern.
2. A few tests, and the guess becomes a conjecture.
3. Once the conjecture is proven, you have a theorem.
A lemma is a theorem leading to other theorems. Typically a technical
result leading to the main theorem.
A corollary is a subsequent theorem, a simple result after a primary
theorem.
Other items:
Denitions are dicult to dene.
Axioms are fundamental denitions. The classic example:
Two points dene a line.
11.3 Direct proof
Example 1.16 from the text:
Theorem: If n is a non-negative integer, then n
2
is either a multiple of 4 or
one larger than a multiple of 4.
(A whole number is a non-negative integer.)
What are we trying to show? Rephrasing with mathematical symbols, we
want to prove that for every non-negative integer n there is some integer k where
n
2
= 4k or n
2
= 4k + 1.
To prove this, it doesnt matter what k is for a given n or which of the two terms
apply. For simpler forms like these, however, its often useful to construct the
result.
Initial exploration.
What else? Build a little table.
11.4. PROOF BY CONTRAPOSITIVES 77
n n
2
form
0 0 4 0
1 1 4 0 + 1
2 4 4 1
3 9 4 2 + 1
4 16 4 4
5 25 4 6 + 1
Recognize a pattern: Evens have one form, odds another. This breaks the
problem into two cases.
If n is even, then n = 2i for some integer i. Then n
2
= 4i and k = i.
If n is odd, then n = 2i+1 for some integer i. Then n
2
= (2i+1)
2
= 4i
2
+4i+1 =
4(i
2
+i) + 1 and k = i
2
+i.
To summarize, we have proven the following:
If n is a non-negative integer, then n
2
is either 4i or 4(i
2
+i) + 1 for
some integer i.
11.4 Proof by contrapositives
Sometimes called indirect reasoning, often called proof by contradiction. Purists
hate that phrase.
Logically, if p then q is equivalent to if not q then not p. This is the contrapositive.
Thus if you prove that the negation of your conclusion q implies the negation of
your hypothesis p, you have proven that your hypothesis p implies the conclusion
q. Clear? No? This is why the method often is called proof by contradiction.
That route is often easier to understand.
For a demonstration, remember the denition of a prime number. A prime is an
integer p = 1 that only can be divided cleanly by 1 and p itself.
Theorem: There are innitely many prime numbers.
Proof: Suppose there is some largest prime P. Then form the (very large)
integer Q = 2 3 5 7 . . . P +1, one larger than the product of all primes. Now
Q is not divisible by any prime, so Q must itself be prime. But Q is larger than
P, contradicting the assumption that P is the largest prime.
What have we actually proven? If there is a largest prime P, then there is a
larger prime Q. Thus, there can be no largest prime.
(The symbol is one traditional ending for a proof. It is used instead of the
letters QED, for quod erat demonstrandum. Literally it means that which was
to be demonstrated.)
78 CHAPTER 11. NOTES FOR 1 SEPTEMBER
The texts Example 1.17 is another very nice example of this style.
11.5 Homework
Groups are ne, turn in your own work. Homework is due in or before class on
Mondays.
Write out (briey) your approach to each problem.
Prove the formula for the sum of the rst n positive integers by mathe-
matical induction. That is, prove 1 + 2 + 3 + +n = n(n + 1)/2. The
base case here is n = 1. Then show that the n
th
term transforms into the
(n + 1)
th
term by adding n + 1.
Note that you may email homework. However, I dont use Microsoft
TM
products
(e.g. Word), and software packages are notoriously nicky about translating
mathematics.
If youre typing it (which I advise just for practice in whatever tools you use),
you likely want to turn in a printout. If you do want to email your submission,
please produce a PDF or PostScript document.
Chapter 12
Notes for 3 September
Notes also available as PDF.
12.1 Proof review
1. You start with a guess, possibly after seeing a pattern.
2. A few tests, and the guess becomes a conjecture.
3. Once the conjecture is proven, you have a theorem.
A lemma is a theorem leading to other theorems. Typically a technical
result leading to the main theorem.
A corollary is a subsequent theorem, a simple result after a primary
theorem.
Other items:
Denitions are dicult to dene.
Axioms are fundamental denitions. The classic example:
Two points dene a line.
A direct proof takes the given hypotheses and applies rules directly to produce
the conclusion. We used a direct proof to show that n
2
is either a multiple of
four (can be written as 4k for some k) or one larger than a multiple of four
(4k + 1) when n is a non-negative integer.
An indirect proof, often called proof by contradiction or proof by contrapositive,
proves that the opposite of the conclusion implies the opposite of the hypothesis.
Thus if the hypothesis is true, the conclusion must be true. We used proof by
contradiction to show that there are innitely many primes. We actually proved
79
80 CHAPTER 12. NOTES FOR 3 SEPTEMBER
that if there is a largest prime, then there is a larger number not divisible by
any known prime and thus prime itself. The contradiction, constructing a prime
larger than the largest prime, proved that there is no largest prime. Along with
the known existence of primes, the fact there is no largest prime implies there
must be innitely many primes.
12.2 Inductive proof
Remember Pascals triangle? We proved that the sum of the n
th
row is 2
n
if we
start counting rows from zero.
Written in rather boring table form, each entry is the sum of the entry directly
above and above to the left:
#
0 1
1 1 1
2 1 2 1
3 1 3 3 1
4 1 = 0 + 1 4 = 1 + 3 6 = 3 + 3 4 = 3 + 1 1 = 1 +0
We will use our result as an example of mathematical induction. This is a proof
technique that formalizes inductive reasoning and using a representative case.
1. Start with verifying a base case.
2. Then assume the n
th
case.
3. Use that to prove the (n + 1)
st
case.
We want to prove the following:
Theorem: The sum of the n
th
line of Pascals triangle is 2
n
.
Proof:
Here, the base case is the 0
th
line:
# sum
0 1 1 = 2
0
Then let line n be the sequence
# sum
n Pn,1 Pn,2 Pn,3 Pn,n+1 2
n
Now we need to prove that line (n + 1) holds true. In our notation, item j
th
on
line (n + 1) is formed be P
n+1,j
= P
n,j1
+P
n,j
, setting the entries outside the
table (P
n,0
and P
n,n+2
) to zero.
12.3. STARTING WITH SET THEORY 81
Then we need to show that

n+2
j=1
P
n+1,j
= 2
n+1
.
n+2

j=1
P
n+1,j
=
n+2

j=1
(P
n,j1
+P
n,j
)
=
n+2

j=1
P
n,j1
+
n+2

j=1
P
n,j
=
_
_
P
n,0
+
n+1

j=1
P
n,j
_
_
+
_
_
P
n,n+2
+
n+1

j=1
P
n,j
_
_
=
n+1

j=1
P
n,j
+
n+1

j=1
P
n,j
= 2
n
+ 2
n
= 2 2
n
= 2
n+1
.
12.3 Starting with set theory
Our coverage of set theory mostly will cover denitions. We will use some
concepts to work with problem solving and to illustrate proofs. Then we will
use sets to build up whole numbers (0, 1, 2, . . .) and basic arithmetic on those
numbers.
12.4 Language of set theory
We will cover just enough set theory to use later.
Cardinalities are important for probability. We dont have time to cover
probability suciently well, so we will not explore the sizes of sets deeply.
This is known as nave set theory. We do not dene absolutely everything,
nor do we push set theorys logical limits. Much.
Goals:
Impart some of the language necessary for later chapters.
Practice reasoning in a formal setting.
One key aspect is what to do in extreme cases like empty sets.
Set up straight-forward examples for logic.
82 CHAPTER 12. NOTES FOR 3 SEPTEMBER
12.5 Basic denitions
To start, we require unambiguous denitions of terms and items. When a term
or item is unambiguously dened, it is called well-dened.
set An unordered collection of unique elements.
Curly braces: {A, B, C} is a set of three elements, A, B, and C.
Order does not matter: {cat, dog} is the same set as {dog, cat}.
Repeated elements do not matter: {1, 1, 1} is the same set as {1}.
Can be implicit: {x| x is an integer, x > 0, x < 3} is the same set as
{1, 2}.
Read the implicit form as the set of elements x such that x is an
integer, x > 0, and x < 3. Or the set of elements x where . . .
Other symbols that sometimes stand for such that: :, (reversed
)
Implicit (or set-builder) form can include formula or other bits left of
the bar. {3x| x is a positive integer} is the set {3, 6, 9, . . .}.
element Any item in a set, even other sets. (Also entry, member, item, etc.)
This is not ambiguous. If something is in a set, it is an item of that
set. It doesnt matter if the item is a number or a grape.
{A, {B, C}} is a set of two elements, A and {B, C}.
None of the following are the same: {A, {B, C}}, {A, B, C}, {{A, B}, C}.
empty set Or null set. Denoted by rather than {}.
This is a set on its own.
{} is the set of the empty set, which is not empty.
Think of sets as bags. An empty bag still is a bag, and if a bag
contains an empty bag, the outer bag is not empty.
Implicit denitions can hide empty sets.
For example, the set {x| x is an odd integer divisible by 2} is .
singleton A set with only one element.
{1} and {} both are singletons (or sometimes singleton sets).
12.6 Translating sets into (and from) English
From English:
12.7. NEXT TIME: RELATIONS BETWEENANDOPERATIONS ONSETS83
The days of the week:
{Monday, Tuesday, Wednesday, Thursday, Friday, Saturday, Sunday}
Of course, were using a representation of the days and not the days
themselves. That is how we reason about things; we model them and
represent them by symbols.
The days when homework is due:
{25
th
of August, 1
st
of September, . . .}
We could list them all.
{ every Monday after the 18
th
of August 2008 until after the 1
st
of
December }
Or: {x| x is a Monday, x is after the 18
th
of August, and x is on or
before the 1
st
of December }
To English:
{2, 3, 4}:
The set containing two, three, and four.
{x| x is an integer and x > 0}:
The positive integers, also called the counting numbers or the natural
numbers.
Often written as J
+
. The integers often are written as J (because the
I form can be dicult to read), rationals as Q (for quotients), the
reals as R.
{2x 1 | x J
+
}
The set whose members have the form 2x 1 where x is a positive
integer.
Cannot list all the entries; this is an innite set.
Here, the odd integers.
12.7 Next time: Relations between and opera-
tions on sets
12.8 Homework
Groups are ne, turn in your own work. Homework is due in or before class on
Mondays.
84 CHAPTER 12. NOTES FOR 3 SEPTEMBER
Write out (briey) your approach to each problem.
Prove the formula for the sum of the rst n positive integers by mathe-
matical induction. That is, prove 1 + 2 + 3 + +n = n(n + 1)/2. The
base case here is n = 1. Then show that the n
th
term transforms into the
(n + 1)
th
term by adding n + 1.
Problem set 2.1 (p83):
Problems 1, 2, 4, 5, 6, 27
Note that you may email homework. However, I dont use Microsoft
TM
products
(e.g. Word), and software packages are notoriously nicky about translating
mathematics.
If youre typing it (which I advise just for practice in whatever tools you use),
you likely want to turn in a printout. If you do want to email your submission,
please produce a PDF or PostScript document.
Chapter 13
Notes for 8 September
Notes also available as PDF.
13.1 Review
set An unordered collection of unique elements.
Curly braces: {A, B, C} is a set of three elements, A, B, and C.
Can be implicit or in set builder notation: {x| x is an integer, x >
0, x < 3} is the same set as {1, 2}.
Order does not matter, repeated elements do not matter.
element Any item in a set, even other sets. (Also entry, member, item, etc.)
empty set Or null set. Denoted by rather than {}.
This is a set on its own.
{} is the set of the empty set, which is not empty.
singleton A set with only one element.
13.2 Relations and Venn diagrams
(Someday I will include Venn diagrams for these in the notes.)
element of The expression x A states that x is an element of A. If x / A,
then x is not an element of A.
4 {2, 4, 6}, and 4 / {x| x is an odd integer }.
85
86 CHAPTER 13. NOTES FOR 8 SEPTEMBER
There is no x such that x , so {x| x } is a long way of writing
.
subset If all entries of set A also are in set B, A is a subset of B.
superset The reverse of subset. If all entries of set B also are in set A, then A
is a superset of B.
proper subset If all entries of set A also are in set B, but some entries of B
are not in A, then A is a proper subset of B.
{2, 3} is a proper subset of {1, 2, 3, 4}.
equality Set A equals set B when A is a subset of B and B is a subset of A.
Order does not matter. {1, 2, 3} = {3, 2, 1}.
The symbols for these relations are subject to a little disagreement.
Many basic textbooks write the subset relation as , so A B when A
is a subset of B. The same textbooks reserve for the proper subset.
Supersets are .
This keeps a supercial similarity to the numerical relations and <. In
the former the compared quantities may be equal, while in the latter they
must be dierent.
Most mathematicians now use for any subset. If a property requires
a proper subset, it often is worth noting specically. And the only
non-proper subset of a set is the set itself.
Extra relations are given for emphasis, e.g. or for proper subsets and
or to emphasize the possibility of equality.
Often a proper subset is written out: A B and A = B.
Ill never remember to stick with the textbooks notation. My
use of is for subsets and not proper subsets.
13.3 Translating relations into (and from) En-
glish
From English:
The train has a caboose.
Its reasonable to think of a train as a set of cars (they can be
reordered).
The cars are the members.
Hence, caboose train
13.4. CONSEQUENCES OF THE SET RELATION DEFINITIONS 87
The VI volleyball team consists of VI students.
VI volleyball team VI students
There are no pink elephants.
pink elephants =
To English:
x todays homework set.
x is a problem in todays homework set.
Todays homework this weeks homework.
Todays homework is a subset of this weeks homework.
13.4 Consequences of the set relation denitions
Every set is a subset of itself. Expected.
If A = B, then every member of A is a member of B, and every
member of B is a member of A. This is what we expect from equality, but
we did not dene set equality this way. Follow the rules:
A = B imples A B and B A.
Because A B, every member of A is a member of B.
Because B A, every member of B is a member of A.
The empty set is a subset of all sets. Unexpected! This is a case of
carrying the formal logic to its only consistent end.
For some set A, A if every member of is in A.
But has no members.
Thus all of s members also are in A.
This is called a vacuous truth.
The alternatives would not be consistent, but proving that requires more ma-
chinery that we need.
13.5 Operations
union The union of two sets A and B, denoted by A B, is the set consisting
of all elements from A and B.
A B = {x| x A or x B}.
88 CHAPTER 13. NOTES FOR 8 SEPTEMBER
Remember repeated elements do not matter: {1, 2}{2, 3} = {1, 2, 3}.
intersection The intersection of two sets A and B, denoted A B, is the set
consisting of all elements that are in both A and B.
A B = {x| x A and x B}.
{1, 2} {2, 3} = {2}.
{1, 2} {3, 4} = {} = .
set dierence The set dierence of two sets A and B, written A \ B, is the
set of entries of A that are not entries of B.
A\ B = {x| x A and x B}.
Sometimes written as AB, but that often becomes confusing.
If A and B share no entries, they are called disjoint. One surprising consequence
is that every set A has a subset disjoint to the set A itself.
No sets (not even ) can share elements with because has no elements.
So all sets are disjoint with .
The empty set is a subset of all sets.
So all sets are disjoint with at least one of their subsets!
Can any other subset be disjoint with its superset? No.
13.6 Homework
Groups are ne, turn in your own work. Homework is due in or before class on
Mondays.
Write out (briey) your approach to each problem.
Problem set 2.1 (p83):
Problems 7, 8, 10, 20, 24
Note that you may email homework. However, I dont use Microsoft
TM
products
(e.g. Word), and software packages are notoriously nicky about translating
mathematics.
If youre typing it (which I advise just for practice in whatever tools you use),
you likely want to turn in a printout. If you do want to email your submission,
please produce a PDF or PostScript document.
Chapter 14
Solutions for third weeks
assignments
Also available as PDF.
Note: These are my approaches to these problems. There are many ways to
tackle each.
14.1 Induction: Sum of rst n integers
Prove the formula for the sum of the rst n positive integers by
mathematical induction. That is, prove 1+2+3+ +n = n(n+1)/2.
The base case here is n = 1. Then show that the n
th
term transforms
into the (n + 1)
st
term by adding n + 1.
Theorem: Let n be a positive integer. Then the sum of the rst n positive
integers is
n

i=1
i =
n(n + 1)
2
.
Proof: We proceed by induction. First note that
1

i=1
i = 1 =
1(1 + 1)
2
.
Now assume that
n

i=1
i =
n(n + 1)
2
.
89
90 CHAPTER 14. SOLUTIONS FOR THIRD WEEKS ASSIGNMENTS
We must show that
n+1

i=1
i =
(n + 1)(n + 2)
2
.
To proceed, we pull the n + 1 term out of the sum to see that
n+1

i=1
i = n + 1 +
n

i=1
i = (n + 1) +
n(n + 1)
2
.
Then we factor out n + 1 and simplify to obtain
(n + 1) +
n(n + 1)
2
= (n + 1)
_
1 +
n
2
_
= (n + 1)
2 +n
2
=
(n + 1)(n + 2)
2
,
proving the result.
14.2 Problem set 2.1 (p83)
14.2.1 Problem 1
Wikipedia can be very useful.
Part a. {California, Oregon, Idaho, Utah, Arizona}. (Note that we could include
Nevada; it certainly shares a border with itself.)
Part b. {Mississippi, Missouri, Maine, Montana, Maryland, Massachusetts, Michigan, Minnesota}
Part c. {Arizona}
(Thanks to Chris Fields for pointing out the states I missed.)
14.2.2 Problem 2
Part a. {a, c, e, h, i, l, m, n, o, s, t, y}, possibly including capital L if you
treat that as a separate letter.
Part b. {a, e, m, t}.
14.2.3 Problem 4
Many answers are reasonable.
14.2. PROBLEM SET 2.1 (P83) 91
Part a. {x|x J, 10 < x < 15}, or perhaps {10 +i|i J, 0 < i < 5}.
Part b. {2i|i J, 3 i 8}.
Part c. {4 + 4i|i J, 0 i < 5}.
Part d. {i
2
+ 2i + 2|i J, 0 i < 4} or
{1 +i|i = the sum of the rst j positive odd integers where j 4}
14.2.4 Problem 5
Many answers are reasonable.
Part a. {2i|i J, i > 6}.
Part b. {(2i + 1)
2
|i J, i 12}.
Part c. {3i|i J
+
}
14.2.5 Problem 6
Part a. No, there are many cities with each name.
Part b. Yes, the cities are specic. There is only one Idaho, Montana, or Texas
in the world. There are two Georgias, but only one has a town called Duluth.
(The other Georgia does not use the same alphabet, but I am pretty sure no
town there transliterates to Duluth.)
Part c. No, there is no universal measure of smart.
Part d. Yes, this is specic. However, this likely is empty. How many people
have a gpa of exactly 3.5? If you have an odd number of credits, that is not
possible. And if you have an even number, you would have to have exactly as
many As as Bs.
14.2.6 Problem 27
These are just my rambling responses.
A set of china is a collection of dishware. However, there are many repeated
elements that are identical (until chipped), so this is not entirely a mathematical
set. If you label each piece, however, then each element has its own identity and
you do have a set.
In a family, pretty much everyone has their own identity and is not repeated,
even if they are twins. So families could be considered sets. If you rank the
family by age or otherwise, the ranked family is ordered and thus not a set.
An aggregate in my mind is a summary and not a listing, but Im poisoned by
programming. So to me, an aggregate is a function you apply to a set to nd a
value. For example, the mean or the median is an aggregate. But Im sure there
are other uses.
92 CHAPTER 14. SOLUTIONS FOR THIRD WEEKS ASSIGNMENTS
A class in a class room denitely is a set. Each member is unique and identiable.
And by the time there is a nal ranking (grade), the class is over. Mathematically,
however, class means something beyond set. The collection of all sets is a class
and not a set; no set can contain itself as an element.
In my wifes collection of antique sewing machines, each is identiable and
appears unique. Thus, a set.
Chapter 15
Notes for 10 September
Notes also available as PDF.
15.1 Review
15.1.1 Denitions
set An unordered collection of unique elements.
element Any item in a set, even other sets. (Also entry, member, item, etc.)
empty set Or null set. Denoted by rather than {}.
singleton A set with only one element.
15.1.2 Relations
element of The expression x A states that x is an element of A. If x / A,
then x is not an element of A.
subset A B if all entries of set A also are in set B.
superset The reverse of subset. B A when A B.
proper subset If all entries of set A also are in set B, but some entries of B
are not in A, then A is a proper subset of B.
equality A = B when A B and B A.
93
94 CHAPTER 15. NOTES FOR 10 SEPTEMBER
15.1.3 Operations
union A B = {x| x A or x B}.
intersection A B = {x| x A and x B}.
set dierence A\ B = {x| x A and x B}.
complement and universe The complement of a set with respect to a
given universal set U is the set A
c
= U \ A = {x| x U and x / A}.
Sometimes written A or A

. Take care with the universal set; there is no

universal universal set.
15.2 From sets to whole numbers
Now well show how to construct whole numbers from sets.
nominal A number used for indentication, like your student ID.
ordinal A number used for ordering, as in rst, second, etc.
cardinal A number that counts the number of entries in some set.
We are moving into counting elements of sets. This can extend into probability
and other topics, but for now we are interested just in counting and numbers.
Two more denitions are useful:
one-to-one correspondence Two sets A and B are in one-to-one correspon-
dence if there is a pairing of elements (a, b) such that each a A and b B
belongs to exactly one pair. The pairing is also called a bijection.
equivalent sets Two sets A and B are equivalent if they are in one-to-one
correspondence. We write A B for equivalence and A B for non-
equivalence.
So what do equivalent sets share? Consider establishing a one-to-one correspon-
dence between two sets A and B. If the sets do not have the same number of
elements, then at least one element will not be paired. Remember indirect proof,
or proof by the contrapositive? We have just proven that all equivalent sets have
the same number of elements.
The number of elements is a rather important property and gains its own
notation:
cardinality Writen |A|, or rarely n(A), the cardinality of A is the number of
elements in A. So |{a, b, 7}| = 3.
We now have a link from the cardinality of sets to cardinal integers 0. These
are the whole numbers. Moreover, now we can construct the whole numbers.
Starting from the unique empty set, || = 0.
15.3. HOMEWORK 95
Dene an operation S(x) = x {x}. This is the successor function. Then
S() = {} = {}, a set with one element. Continue applying the successor
to see:
|| 0
S() |{}| 1 {0}
S(S()) |{, {}}| 2 {0, 1}
S(S(S())) | {, {}, {, {}}} | 3 {0, 1, 2}
By applying the successor function n times to the empty set, we obtain a
set of cardinality n. Starting from nothing (literally, ), we obtain the whole
numbers! This is a construction from Giuseppe Peano, a 19th century Italian
mathematician. After the association between and 0, we identify the number
k with the set of all numbers preceding k.
The text has examples of building numbers with dierent physical gizmos. Its
worth thinking about the 2-D block idea, particularly in regards to factors and
factorization. Here, you could imagine S as adding one more unit onto a line.
Starting from a point, , applying S grows the numbers one at a time.
We also can impose an ordering on these numbers. Given two sets A and B with
|A| = a and |B| = b, we say that a < b if A is equivalent to some proper subset
of B, or A C where C B. We say a b if A is equivalent to some subset of
B.
Now we also can see the dierence between nite sets and the concept of innite
sets. A nite set is equivalent to some number of successor applications to the
empty set. You can count nite numbers.
An innite set, however, is equivalent to no number of successor applications. If
it were, we could add one more to have a larger set. There are dierent kinds of
innities, but we wont worry about that right now. And note that you can have
a one-to-one correspondence between innite sets. For example, {x| x J
+
} has
a one-to-one correspondence with {
1
x
| x J
+
}. Each integer in the rst set is
matched with exactly one in the second set.
15.3 Homework
Groups are ne, turn in your own work. Homework is due in or before class on
Mondays.
Write out (briey) your approach to each problem.
Problem set 2.2 (p97):
Problems 1, 2, 6, 13, 21, 23
Why would answering problem 32 be a bad idea?
96 CHAPTER 15. NOTES FOR 10 SEPTEMBER
Note that you may email homework. However, I dont use Microsoft
TM
products
(e.g. Word), and software packages are notoriously nicky about translating
mathematics.
If youre typing it (which I advise just for practice in whatever tools you use),
you likely want to turn in a printout. If you do want to email your submission,
please produce a PDF or PostScript document.
Chapter 16
Notes for 12 September
Notes also available as PDF.
16.1 Review
nominal A number used for indentication, like your student ID.
ordinal A number used for ordering, as in rst, second, etc.
cardinal A number that counts the number of entries in some set.
one-to-one correspondence Two sets A and B are in one-to-one correspon-
dence if there is a pairing of elements (a, b) such that each a A and b B
belongs to exactly one pair. The pairing is also called a bijection.
equivalent sets Two sets A and B are equivalent if they are in one-to-one
correspondence. We write A B for equivalence and A B for non-
equivalence.
cardinality Writen |A|, or rarely n(A), the cardinality of A is the number of
elements in A. So |{a, b, 7}| = 3.
We now have a link from the cardinality of sets to cardinal integers 0. These
are the whole numbers. Moreover, now we can construct the whole numbers.
Starting from the unique empty set, || = 0.
Dene an operation S(x) = x {x}. This is the successor function. Then
S() = {} = {}, a set with one element. Continue applying the successor
to see:
97
98 CHAPTER 16. NOTES FOR 12 SEPTEMBER
|| 0
S() |{}| 1 {0}
S(S()) |{, {}}| 2 {0, 1}
S(S(S())) | {, {}, {, {}}} | 3 {0, 1, 2}
After the association between and 0, we identify the number k with the set of
all numbers preceding k. Thus we follow Guiseppe Peanos construction of the
whole numbers from nothing but sets.
nite set Equivalent to some number of successor applications S() to the empty
set.
innite set Cannot be equivalent to any number of successor applications, as
we could apply once more to obtain a larger set.
16.2 Addition of whole numbers
So how do we dene addition? The text has one method that assumes A and B
are disjoint. That is, A B = . Then |A B| = |A| +|B|, so we could dene
addition by the cardinality of a union of disjoint sets.
Its worth a brief detour to discuss what happens to the cardinality when the
sets are not disjoint. This is a tool used frequently in probability.
|{a, b} {c, d}| = |{a, b, c, d}| = 4, where |{a, b}| = |{c, d}| = 2. If they share
an element, though, the unions cardinality still is four. |{a, b, c} {c, d}| =
|{a, b, c, d}| = 4. But now |{a, b, c}| = 3.
The rule for counting elements for a union of two sets is |AB| = |A|+|B||AB|.
The rst two terms count shared elements twice, and the last term removes one
of those. For three terms, |A B C| = |A| +|B| +|C| |A B| |A C|
|B C| +|A B C|, where you add back in elements you removed too many
times.
But for general addition, we start by dening a + 0 = a using A = A as a
guide. Then we dene addition for other numbers recursively by
a + 0 = a, and
a +S(b) = S(a +b).
So a + 1 = a +S(0) = S(a + 0) = S(a), the successor of a. If we think of S(b)
as b + 1, then the second rule is a + (b + 1) = (a +b) + 1.
16.3. SUBTRACTION OF WHOLE NUMBERS 99
A small example working through the recursive denition:
1 + 2 = 1 +S(1)
= S(1 + 1)
= S(1 +S(0))
= S(S(1 + 0))
= S(S(1))
= S(2)
= 3.
Obviously, we dont want to work through many of these.
So we dene addition as decomposing one number into its sequence of successors
and then applying those successors. For the line model, we extend one line by
the length of the other.
Some properties of addition on whole numbers:
closure If a and b are two whole numbers, a + b is a whole number. Whole
numbers are closed over addition.
commutative a +b = b +a
associative a + (b +c) = (a +b) +c
identity a + 0 = a
Each of these is straight-forward to prove using the texts model of unions of
disjoint sets. Working with Peano arithmetic is trickier but possible.
Terminology: The terms being added are called addends or summands.
16.3 Subtraction of whole numbers
Subtraction is the rst property we will dene implicitly. The dierence of a
and b, written a b, is dened to be the whole number c where a = b + c. If
no such c exists, then subtraction is not dened. There is no integer c 0 such
that 1 = 5 +c, so 1 5 is not dened.
Thus subtraction is not closed over the whole numbers. It also is not com-
mutative and not associative. However, because a + 0 = a, we know that
a 0 = a. Thus subtraction does have an identity on one side. However, 0 a
is not dened unless a = 0. There is no c such that 0 = a +c unless a = c = 0.
Terminology: In a b, a is the minuend and b is the subtrahend.
100 CHAPTER 16. NOTES FOR 12 SEPTEMBER
16.4 Multiplication of whole numbers
We are accustomed to thinking of multiplication as repeated addition. For
example, 4a = a +a +a +a. Or we can rearrange the repeated addition as in
4a = a + 3a = a + (a + 2a) = a + (a + (a +a)). This arrangement provides the
recursive denition of multiplication:
a 0 = 0, and
a S(b) = a + (a b)
Again, thinking of S(b) as b + 1, a (b + 1) = a + (a b). We pull apart one side
into a sequence of ones and repeatedly add the other side.
We will jump ahead and assume multiplication is commutative. So write the
example of 4a as a 4:
a 4 = a S(3)
= a + (a S(2))
= a + (a +a S(1))
= a + (a + (a +a S(1)))
= a + (a + (a + (a +a 0))
= a +a +a +a + 0.
And with this recursive denition,
a 1 = a S(0)
= a + (a 0)
= a + 0 = a,
and one is the multiplicative identity.
Dening multiplication in terms of repeated addition lets multiplication inherit
the closure property. The product a b exists and is a whole number for every
pair of whole numbers a and b. And we could push through and show that
multiplication is commutative and associative. Another less formal model
makes these properties more clear, however.
A quick note on terminology: In 15 = 5 3, 15 is the product, and ve and three
are factors. The term factor only applies to integers. Two other terms applied
to the operands (5 and 3) in the context of repeated addition are multiplicand
and multiplier. If you consider 15 = 53 = 5+5+5, then ve is the multiplicand
and three is the multiplier. These terms still are used when building multiplier
circuits. One term, the multiplicand, is shifted and added according to the
multiplier. We will return to this later.
But for the less formal models, consider multiplication of two numbers as tracing
out areas and volumes.
16.5. MONDAY: DIVISION AND EXPONENTIALS 101
The drawing can have its axes ipped around without changing the area. Thus,
ab = ba and multiplication is commutative.
For the associative property, consider a volume traced out by three numbers.
Putting parenthesis in dierent places is the same as counting in dierent
directions, but the nal count always is the same. Thus multiplication is
associative.
There is one additional property that combines addition and multiplication: the
distributive property. Here a(b +c) = ab +ac. Drawing a box and separating it
into two pieces demonstrates the distributive property well. Thinking of repeated
addition here also is simple. We have either b + c copies of a, or b copies of a
added to c copies.
16.5 Monday: Division and exponentials
I need to cover these still. They provide examples of undened behavior; its
worth exploring that a bit.
To be followed by a little review.
16.6 Homework
Groups are ne, turn in your own work. Homework is due in or before class on
Mondays.
Write out (briey) your approach to each problem.
Problem set 2.3 (p111):
2, 5, 11, 24
Write 2 + 3 using disjoint sets.
Illustrate 2 + 3 using Peano arithmetic.
Problem set 2.4 (p130):
5, 10
26
Illustrate 2 3 using Peano arithmetic. You do not need to expand addition.
Illustrate (1 2) 3 = 1 (2 3) using a volume of size six.
Note that you may email homework. However, I dont use Microsoft
TM
products
(e.g. Word), and software packages are notoriously nicky about translating
mathematics.
102 CHAPTER 16. NOTES FOR 12 SEPTEMBER
If youre typing it (which I advise just for practice in whatever tools you use),
you likely want to turn in a printout. If you do want to email your submission,
please produce a PDF or PostScript document.
Chapter 17
Solutions for fourth weeks
assignments
Also available as PDF.
Note: These are my approaches to these problems. There are many ways to
tackle each.
17.1 Problem set 2.2
17.1.1 Problem 1
(a) June 13
th
: nominal, rst: ordinal
(b) eleventh: ordinal, second: ordinal, 6-iron: nominal, 160 yards: cardinal
(c) 7 pin: nominal, sixth frame: ordinal, 9: cardinal
17.1.2 Problem 2
(a) There are ve letters in the phrase, {P,A,N,M,B}, so we can associate 1
P, 2 A, etc. The two sets are equivalent.
(b) The second set has one more element than the rst, so the sets are not
equivalent.
(c) We can associate o t, n w, and e o, where the left quantities come
from {o, n, e} and the right from {t, w, o}. The two sets are equivalent.
(d) {0} has in element, and has none, so the sets are not equivalent.
103
104 CHAPTER 17. SOLUTIONS FOR FOURTH WEEKS ASSIGNMENTS
17.1.3 Problem 6
(a) We can use the function f(w) = w +1 to construct a bijection between W
and N.
(b) The function f(d) = d +1 creates a one-to-one mapping between D and E.
(c) The function f(n) = 10
n
creates a one-to-one mapping between N and
{10, 100, . . .}.
17.1.4 Problem 13
The drawing will have
|B C| |A B C| = 5 in the unlabeled portion of B C,
|B| |A B| |B C| + |A B C| = 28 in the remaining unlabeled
portion of B,
|A C| |A B C| = 8 in the unlabeled portion of A C,
|A| |A B| |A C| + |A B C| = 15 in the remaining unlabeled
portion of A,
|C| |B C| |A C| + |A B C| = 10 in the remaining unlabeled
portion of C, and
|U| (10 + 7 + 5 + 28 + 8 + 15 + 10) = 17 in U but outside A, B, and C.
17.1.5 Problem 21
(a) The function f(a) = a is the one-to-one correspondence from A to itself.
(b) If A B, there is a function f(a) = b where each b B appears for exactly
one a A and the function is dened for every b B and a A. The
inverse function g(b) = a where b = f(a) shows B A.
(c) If A B and B C, then there are one-to-one functions f(a) = b
and g(b) = c for each mapping. The function h(a) = g(f(a)) then is a
one-to-one mapping between A and C.
17.1.6 Problem 23
For (a) and (b), the table is Pascals triangle, which we already have in the
notes. For (c), we want the entry P
6,3
in our earlier notation. There are 20 such
ways.
17.2. PROBLEM SET 2.3 105
17.1.7 Why answering problem 32 would be a bad idea.
Wow. Never, ever give someone a list of all your identifying numbers. Identity
theft is a serious problem. While quite often all these numbers are available for
a little work, at least make a criminal work for them.
17.2 Problem set 2.3
17.2.1 Problem 2
Substituting into |AB| = |A| +|B| |AB|, we see that 10 = 5 +8 |AB|,
or that |A B| = 3.
17.2.2 Problem 5
Im just going to show these as lines. Illustrating them on a number line is
equivalent. A better diagram for the last would wrap portions of the result in
parentheses to show how the line extended.
(a) 3 + 5 = +
=
= 8
(b) 5 + 3 =
+
=
= 8
(c) 4 + 2 =
+
=
= 6
(d) 0 + 6 =
+
=
= 6
(e) 3 + (5 + 7) =
+ ( + )
= +
=
= 15
(f ) (3 + 5) + 7 =
( + ) +
106 CHAPTER 17. SOLUTIONS FOR FOURTH WEEKS ASSIGNMENTS
= +
=
= 15
17.2.3 Problem 11
Again, Im just going to show these as lines. Illustrating them on a number
line is equivalent.
(a) 7 3 =

=
= 4
(b) 7 4 =

=
= 3
(c) 7 7 =

= = 0
(d) 7 0 =
-
=
= 7
17.2.4 Problem 24
Draw with two shadings and show that the intersection is shaded twice.
17.3 Write 2 + 3 using disjoint sets.
If we let 2 {a, b} and 3 {c, d, e}, then 2+3 {a, b}{c, d, e} = {a, b, c, d, e}.
17.4 Illustrate 2 + 3 using Peano arithmetic.
We dened addition with
a + 0 = a, and
a +S(b) = S(a +b).
17.5. PROBLEM SET 2.4 107
Here, 3 = S(2), so
2 + 3 = 2 +S(2)
= S(2 + 2)
= S(2 +S(1))
= S(S(2 + 1))
= S(S(2 +S(0)))
= S(S(S(2 + 0)))
= S(S(S(2)))
= 5.
17.5 Problem set 2.4
17.5.1 Problem 5
(a) Any set that contains only 1 and some other number is closed because 1
is the multiplicative identity.
(b) Any set that contains only 1 and some other number is closed because 1
is the multiplicative identity.
(c) 2 4 = 8 / {0, 2, 4}, so this is not closed.
(d) The product of even numbers always is even, so this is closed.
(e) The product of odd numbers cannot be even, so they must be odd and
this set is closed.
(f ) 2
2
2
3
= 2
5
/ {1, 2, 2
2
, 2
3
}, so this set is not closed.
(g) The product of powers of two is a power of two, so this set is closed.
(h) Similarly, the product of powers of seven is a power of seven so this set
is closed. Both of these are closed because {0, 1, 2, . . .} is closed under
addition; a
i
a
j
= a
i+j
, moving the property up to the superscript.
17.5.2 Problem 10
Each product is equal to the appropriate shaded area, and the sum is equal to
the entire rectangle. The letters correspond directly to the positions.
I had completely forgotten about the FOIL mnemonic. After long enough, its
just a part of what you do.
108 CHAPTER 17. SOLUTIONS FOR FOURTH WEEKS ASSIGNMENTS
17.5.3 Problem 26
The operation is closed because no new shapes are introduced in the operators
table of results.
The operation is commutative because the operation table is symmetric across
the diagonal axis. Thus a b = b a.
Because x = x, the symbol is s identity.
After the identity, there are only two symbols remaining. Thus the commu-
tative property combined with the identity means that this operator must be
associative.
17.6 Illustrate 2 3 using Peano arithmetic. You
do not need to expand addition.
We dened multiplication with
a 0 = 0, and
a S(b) = a + (a b).
So
2 3 = 2 S(2)
= 2 + (2 2)
= 2 + (2 S(1))
= 2 + (2 + (2 1))
= 2 + (2 + (2 S(0)))
= 2 + (2 + (2 + 2 0))
= 2 + 2 + 2 = 6.
17.7 Illustrate (1 2) 3 = 1 (2 3) using a volume
of size six.
Draw two 1 2 3 boxes made of 1 1 1 boxes. To illustrate (1 2) 3, show
that the 1 2 portion is stacked 3 times. To illustrate 1 (2 3), show that the
2 3 portion is stacked across 1 time.
Chapter 18
Notes for the fth week:
review
Notes also available as PDF.
18.1 Review
Structure of the upcoming test:
Eight questions. Chose six and solve them.
Thus expect about seven minutes per question.
Remember to read and answer the entire question.
Closed book, etc. Calculators are ne but not necessary.
Bring scratch paper and paper for writing up your results. Separately.
Answers and explanations need to be indicated clearly.
No questions are intended to be trick questions.
Will cover the following topics:
problem solving,
set theory,
operations on whole numbers.
Remember that solutions for the homework problems are available on-line:
http://jriedy.users.sonic.net/VI/math202-f08/.
109
110 CHAPTER 18. NOTES FOR THE FIFTH WEEK: REVIEW
18.2 Problem solving
Polyas principles:
1. Understand the problem.
2. Devise a plan.
3. Carry out the plan.
4. Look back at your solution.
This is not a simple 1-2-3-4 recipe. Understanding the problem may include
playing with little plans, or trying to carry out a plan may lead you back to
trying to understand the problem.
18.2.1 Understand the problem
Read the entire problem.
Read the whole problem.
Read all of the problem.
One comment about the homeworks: Most people answer only part of any
given problem.
Determine what you have and what you want.
To indicate an answer clearly to someone else (like me), you need to know
what the answer is.
Consider rephrasing the problem, either in English or symbolically.
Rephrasing the problem may help you remember solution methods.
Try some examples.
This is close to devising a plan. Sometimes you may stumble upon an
answer.
Look for relationships between the data.
Examples may help nd relationships. The relationships may help you
decide on a plan. Mathematics is about relationships between dierent
entities; symbolic mathematics helps abstract away the entities themselves.
18.2.2 Devise a plan
Sometimes plans are trivial, or so simple it seems pointless to make them
specic. But write it out anyways. Often the act of putting a plan into words
helps nd aws in the plan.
18.3. SET THEORY 111
Try to devise a plan that you can check along the way. The earlier you detect a
problem, the easier you can deal with it.
Some plans weve considered:
Guessing and checking.
Try a few combinations of the data. See what falls out. This is good for
nding relationships and understanding the problem.
Searching using a list.
If you know the answer lies in some range, you can search that range
systematically by building a list.
Finding patterns.
When trying examples, keep an eye open for patterns. Sometimes the
patterns lead directly to a solution, and sometimes they help to break a
problem into smaller pieces.
Following dependencies / working backwards.
Be sure to understand what results depend on which data. Look for
dependencies in the problem. Sometimes pushing the data you have
through all the dependencies will break the problem into simpler sub-
problems.
18.2.3 Carry out the plan
Attention to detail is critial here.
When building a list, be sure to carry out a well-dened procedure. Or when
looking for patterns, be systematic in the examples you try. Dont jump around
randomly.
18.2.4 Look back at your solution
Can you check your result? Sometimes trying to check reveals new relationships
that could lead to a better solution.
Think about how your solution could help with other problems.
18.3 Set theory
18.3.1 Denitions and mappings
set An unordered collection of unique elements.
112 CHAPTER 18. NOTES FOR THE FIFTH WEEK: REVIEW
You can write a set by listing its entries, {1, 2, 3, 4}, or through set builder
notation, {x| x is a positive integer, x < 5}.
empty set The unique set with no elements: {} or . Be sure to know the
relations between the empty set and other sets, and also how the empty
set behaves in operations.
element of You write x A to state that x is an element of A. The symbol is
not an E but is almost a Greek . Think of a pitchfork.
Note that 1 {1, 2} and {1} {{1}, {2}}, but {1} {1, 2}.
subset Given two sets A and B, A B if every element of A is also an element
of B.
One implication is that A for all sets A. This statement is vacuously
true.
Here {1} {1, 2} and {1} {{1}, {2}}.
superset Given two sets A and B, B A if every element of A is also an
element of B.
proper subset or superset A subset or superset relation is proper if it im-
plies the sets are not equal.
Venn diagram A blobby diagram useful for illustrating operations and relations
between two or three sets.
union A B = {x| x A or x B}. The union contains all elements of both
sets.
intersection A B = {x| x A and x B}. The intersection contains only
those elements that exist in both sets.
set dierence A \ B = {x| x A and x / B}. The set dierence contains
elements of the rst set that are not in the second set. It cannot contain
any elements of the second set.
You can write the result of multiple operations in set-builder notation,
(A B) C = {x| x A and x B} C
= {x| (x A and x B) or x C}.
18.3.2 Cardinality and one-to-one correspondence
The cardinality of a set is a count of its elements. So |{a, b, c}| = 3. An innite
set like the set of all positive integers has no simple cardinality.
To sets are equivalent, or A B, if there is a one-to-one correspondence
between the two sets. With such a correspondence, each element of one set is
matched with exactly one element of the other set.
18.4. OPERATIONS AND WHOLE NUMBERS 113
For nite sets, the two sets must be the same size. If they werent, there would
be at least one unmatched element.
For nite sets, establishing a one-to-one correspondence is straight-forward. You
just need to list which entries correspond. For {1, 2, 3} and {a, b, c}, such a list
might be
1 c, 2 a, 3 b.
Or you could provide a formula. For A = {1, 2, 3} and B = {2, 4, 6}, a mapping
could be a A f(a) B where
f(a) = 2a.
For innite sets, a formula is the most straight-forward way. Consider the
extending the above sets to be innite. For sets A = {1, 2, 3, . . .} and B =
{2, 4, 6, . . .}. the same mapping as before shows they are equivalent. For the sets
A = {1, 2, 3, . . .} and B = {2, 3, 4, . . .}, a mapping function would be f(a) = a+1.
18.4 Operations and whole numbers
closure For any operation, the result of the operation always lies in the set.
For whole numbers, x +y is closed but x y is not.
commutative For numbers, x+y = y+x and xy = yx. For sets, AB = BA
and A B = B A.
associative For numbers, x + (y + z) = (x + y) + z and x(yz) = (xy)z. For
sets, A (B C) = (A B) C) and A (B C) = (A B) C.
distributive For numbers, the only version is the distribution of multiplication
over addition or x(y +z) = xy +xz. For sets, both unions and intersections
distribute. So A (B C) = (A B) (A C) and A (B C) =
(A B) (A C).
There are multiple ways to illustrate addition and multiplication.
For addition, one illustration is piles of rocks or symbols. This is somewhat like
set unions, except we assume every element of each set is unique. Equivalently,
we assume all sets are disjoint. Then |AB| = || = 0, so |AB| = |A| +|B|
|A B| |A| +|B|.
Another illustration is the number line. Each number is represented by a length:
0
1
2
Then addition appends lines:
1+2 + 3
114 CHAPTER 18. NOTES FOR THE FIFTH WEEK: REVIEW
This makes the commutative and associative properties fairly obvious; the line
has the same nal length regardless.
This is essentially the same as Peano arithmetic covered earlier, except addition
in Peano arithmetic moves a bar from one number to the other until the sum is
reached.
Multiplication often is best illustrated by areas or volumes.
Chapter 19
First exam and solutions
Available as PDF.
115
116 CHAPTER 19. FIRST EXAM AND SOLUTIONS
Part III
Notes for chapters 3, 4, and
5
117
Chapter 20
Notes for the sixth week:
digits, bases, and operations
Notes also available as PDF.
What we will cover from Chapter 4:
Numbers and digits in dierent bases, with historical context
Arithmetic, digit by digit
And additionally, Ill give a brief summary of computer arithmetic.
20.1 Positional Numbers
A number is a concept and not just a sequence of symbols. We will be discussing
ways to express numbers.
Before our current form of expressing cardinal numbers:
Piles of rocks dont work well for merchants.
Marks on sticks, then marks on papyrus.
Marking numbers is costly. A large number becomes a large number of marks.
Many marks lead to many errors. Merchants dont like errors. So people started
using symbols rather than plain marks.
An intermediate form, grouping:
Egyptian: Dierent symbols for dierent levels of numbers: units, tens,
hundreds. Grouping within the levels.
119
120CHAPTER 20. NOTES FOR THE SIXTHWEEK: DIGITS, BASES, ANDOPERATIONS
Roman: Symbols for groups, with addition and subtraction of symbols for
smaller groups.
Greek (and Hebrew and Arabic): Similar, but using all their letters for
many groups.
Early Chinese: Denote the number of marks in the group with a number
itself. . .
Getting better, but each system still has complex rules. The main problems
are with skipping groups. We now use zero to denote an empty position, but
these systems used varying amounts of space. Obviously, this could lead to trade
disagreements. Once zeros were adopted, many of these systems persisted in
trade for centuries.
Now into forms of positional notation, shorter and more direct:
Babylonian:
Two marks, tens and units.
Now the marks are placed by the number of 60s.
Suers from complicated rules about zeros.
(Using 60s persists for keeping time...)
Mayan:
Again, two kinds of marks for ves and units.
Two positional types: by powers of 20, and by powers of 20 except
for one power of 18.
(Note that 18 20 = 360, which is much closer to a year.)
Essentially equivalent to what we use, but subtraction in Mayan is
much easier to see.
(many other cultures adopted similar systems (e.g. Chinese rods)
Current: Hindu-Arabic numeral system
The characters dier between cultures, but the idea is the same. The characters
often are similar as well. Originated in the region of India and was carried west
through trade. No one knows when zero was added to the digits. The earliest
rm evidence is in Arab court records regarding a visitor from India and a
description of zero from around 776 AD. The rst inscription found with a zero
is from 876 AD in India. However, the Hindu-Arabic system was not adopted
outside mathematics even in these cultures. Merchants kept to a system similar
to the Greek and Hebrew systems using letters for numbers.
Leonardo Fibonacci brought the numerals to Europe in the 13
th
century (after
1200 AD) by translating an Arabic text to Latin. By 15
th
century, the nu-
20.2. CONVERTING BETWEEN BASES 121
meral system was in wide use in Europe. During the 19
th
century, this system
supplanted the rod systems in Asia.
The nal value of the number is based on the positions of the digits:
1234 = 1 10
3
+ 2 10
2
+ 3 10
1
+ 4 10
0
.
We call ten the base. Then numbers becomes polynomials in terms of the base
b,
1234 = b
3
+ 2 b
2
+ 3 b
1
+ 4.
Here b = 10.
So we moved from marks, where 1000 would require 1000 marks, to groups,
where 1000 may be a single mark but 999 may require dozens of marks. Then
we moved to positional schemes where the number of symbols depends on the
logarithm of the value; 1000 = 10
3
requires 4 = 3 + 1 symbols.
After looking at other bases, we will look into operations (multiplication, addition,
etc.) using the base representations.
20.2 Converting Between Bases
Only three bases currently are in wide use: base 10 (decimal), base 2 (binary),
and base 16 (hexadecimal). Occasionally base 8 (octal) is used, but that is
increasingly rare. Other conversions are useful for practice and for seeing some
structure in numbers. The structure will be useful for computing.
Before conversions, we need the digits to use. In base b, numbers are expressed
using digits from 0 to b 1. When b is past 10, we need to go beyond decimal
numerals:
Value: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Digit: 0 1 2 3 4 5 6 7 8 9 A B C D E F
Upper- and lower-case are common.
So in hexadecimal, DECAFBAD is a perfectly good number, as is DEADBEEF. If
there is a question of what base is being used, the base is denoted by a subscript.
So 10
10
is a decimal ten and 10
2
is in binary.
To nd values we recognize more easily, we convert to decimal. Then we will
convert from decimal.
122CHAPTER 20. NOTES FOR THE SIXTHWEEK: DIGITS, BASES, ANDOPERATIONS
20.2.1 Converting to Decimal
Converting to decimal using decimal arithmetic is straight-forward. Remember
the expansion of 1234 with base b = 10,
1234 = 1 10
3
+ 2 10
2
+ 3 10
1
+ 4 10
0
= b
3
+ 2 b
2
+ 3 b
1
+ 4.
Each digit of DEAD has a value, and these values become the coecients. Then
we expand the polynomial with b = 16. In a straight-forwart way,
DEAD = D 16
3
+E 16
2
+A 16
1
+D
= 13 16
3
+ 14 16
2
+ 10 16 + 13
= 13 4096 + 14 256 + 10 16 + 13
= 57005.
We an use Horners rule to expand the polynomial in a method that often is
faster,
DEAD = ((13 16 + 14) 16 + 10) 16 + 13
= (222 16 + 10) 16 + 13
= 3562 16 + 13
= 57005.
Lets try a binary example. Convert 1101
2
to decimal:
1101
2
= (((1 2 + 1) 2 + 0) 2 + 1
= (3 2 + 0) 2 + 1
= 6 2 + 1
= 13.
Remember the rows of a truth table for two variables? Here,
11
2
= 2 + 1 = 3,
10
2
= 2 + 0 = 2,
01
2
= 0 + 1 = 1, and
00
2
= 0 + 0 = 0.
20.3. OPERATING ON NUMBERS 123
20.2.2 Converting from Decimal
To convert to binary from decimal, consider the previous example:
13 = 8 +5
= 8 + 4 +1
= 1 2
3
+ 1 2
2
+ 0 2
1
+ 1 2
0
= 1101
2
.
At each step, we nd the largest power of two less than the remaining number.
Another example for binary:
293 = 256 +37
= 256 + 32 +5
= 256 + 32 + 4 + 1
= 1 2
8
+ 1 2
5
+ 1 2
2
+ 1
= 100100101
2
.
And in hexadecimal,
293 = 256 +37
= 1 256 + 2 16 + 5
= 125
16
.
You can see why some people start remembering powers of two.
If you have no idea where to start converting, remember the relations b
log
b
x
= x
and log
b
x = log x/ log b. Rounding log
b
x up to the larger whole number gives
you the number of base b digits in x.
The text has another version using remainders. We will return to that in the
next chapter. And conversions to and from binary will be useful when we discuss
how computers manipulate numbers.
20.3 Operating on Numbers
Once we split a number into digits (decimal or binary), operations can be a bit
easier.
We will cover multiplication, addition, and subtraction both
to gain familiarity with positional notation, and
to compute results more quickly and mentally.
124CHAPTER 20. NOTES FOR THE SIXTHWEEK: DIGITS, BASES, ANDOPERATIONS
Properties of positional notation will help when we explore number theory.
We will use two properties frequently:
Both multiplication and addition commute (a + b = b + a) and re-
associate (a +b) +c = a + (b +c).
Multiplication distributes over addition, so a(b +c) = ab +ac.
Multiplying powers of a common base adds exponents, so b
a
b
c
= b
a+c
.
20.3.1 Multiplication
Consider multiplication. I once had to learn multiplication tables for 10, 11, and
12, but these are completely pointless.
Any decimal number multiplied by 10 is simply shifted over by one digit,
123 10 = (1 10
2
+ 2 10
1
+ 3 10
0
) 10
= 1 10
3
+ 2 10
2
+ 3 10
1
= 1230.
Multiplying by 11 = 1 10 + 1 is best accomplished by adding the other number
to itself shifted,
123 11 = 123 (10 + 1) = 1230 + 123 = 1353.
And for 12 = 1 10 + 2, you double the number,
123 12 = 123 (10 + 2) = 1230 + 246 = 1476.
Multiplying longer numbers quickly follows the same pattern of shifting and
adding. We can expand 123 123 = 123 (1 10
2
+ 2 10 + 3) to
123
123
369
2460
12300
15129
Another method expands the product of numbers as a product of polynomials,
working one term at a time. This is essentially the same but not in tabular form:
123 123 = (1 10
2
+ 2 10 + 3) (1 10
2
+ 2 10 + 3)
= (1 10
2
+ 2 10 + 3) (1 10
2
+ 2 10) + (1 10
2
+ 2 10 + 3) 3
= (1 10
2
+ 2 10 + 3) (1 10
2
+ 2 10) + (3 10
2
+ 6 10 + 9)
= . . .
20.3. OPERATING ON NUMBERS 125
This form splits the sums apart as well; we will cover that next.
Bear in mind that short-term memory is limited to seven to eight pieces of
information. Structure mental arithmetic to keep as few pieces in ight as
possible. One method is to break multiplication into stages. In long form, you can
group the additions. For example, expanding 123123 = 123(110
2
)+(12323) =
123 (1 10
2
) + (123 2 10 + 123 3),
123
123
369
2460
2829
12300
15129
Assuming a small number uses only one slot in your short-term memory, need
track only where you are in the multiplier, the current sum, the current product,
and the next sum. That leaves three to four pieces of information to use while
adding.
One handy trick for 15% tips: divide by ten, divide that amount by two, and
add the pieces. We can use positional notation to demonstrate how that works,
x 15% = (x 15)/100
= ((x (10 + 5))/100
= ((x 10) + (x (10/2)))/100
= x/10 + (x/10)/2
20.3.2 Addition
Digit-by-digit addition uses the commutative and associative properties:
123 + 456 = (1 10
2
+ 2 10 + 3) + (4 10
2
+ 5 10 + 6)
= (1 + 4) 10
2
+ (2 + 5) 10 + (3 + 6)
= 579.
Naturally, when a digit threatens to roll over ten, it carries to the next digit.
Expanding the positional notation,
123 + 987 = (1 10
2
+ 2 10 + 3) + (9 10
2
+ 8 10 + 7)
= (1 + 9) 10
2
+ (2 + 8) 10 + (3 + 7)
= 10 10
2
+ 10 10 + 10.
126CHAPTER 20. NOTES FOR THE SIXTHWEEK: DIGITS, BASES, ANDOPERATIONS
Because the coecients are greater than b 1 = 9, we expand those coecients.
Commuting and reassociating,
123 + 987 = 10 10
2
+ 10 10 + 10
= (1 10 + 0) 10
2
+ (1 10 + 0) 10 + (1 10 + 0)
= 1 10
3
+ 1 10
2
+ 1 10 + 0
= 1110.
However, when working quickly, or when the addition will be used in another
operation, you do not need to expand the carries immediately. This is called a re-
dundant representation because numbers now have multiple representations.
You can represent 13 as 1 10 + 3 or simply as 13.
If you work that way mentally, you need to keep the intermediate results in
memory. So during multiplying, you only need to work out the carries every
three to four digits. . .
20.3.3 Subtraction
In systems with signed numbers, we know that subtracting a number is the same
as adding its negation: a b = a +(b). So we expect the digit-by-digit method
to work with each digit subtracted, and it does. Because a = 1 a, we can
distribute the sign over the digits:
456 123 = (4 10
2
+ 5 10 + 6) (1 10
2
+ 2 10 + 3)
= (4 10
2
+ 5 10 + 6) + ((1 10
2
+ 2 10 + 3))
= (4 10
2
+ 5 10 + 6) + (1 10
2
+2 10 +3)
= (4 1) 10
2
+ (5 2) 10 + (6 3)
= 333.
As with carrying, borrowing occurs when a digit goes negative:
30 11 = (3 10
1
+ 0) (1 10
1
+ 1)
= (3 1) 10
1
+ (0 1)
= 2 10
1
+1
= 1 10
1
+ (10 1)
= 1 10
1
+ 9
= 19.
Again, you can use a redundant intermediate representation of 2 10
1
1 if
youre continuing to other operations. And if all the digits are negative, you
20.4. HOMEWORK 127
can factor out 1,
123 456 = (1 10
2
+ 2 10 + 3) (4 10
2
+ 5 10 + 6)
= (1 4) 10
2
+ (2 5) 10 + (3 6)
= (3) 10
2
+ (3) 10 + (3)
= (3 10
2
+ 3 10 + 3)
= 333.
20.3.4 Division and Square Root: Later
We will cover these later with number theory.
20.4 Homework
Problem Set 3.1 (p154)
8, 16
32, 34, 35
Problem Set 3.2 (p161)
3, 4, 5, 6, 8, 11, 20
Problem Set 3.3 (p177)
10, 12, 20 (the zeros are indicators; nd the least base where the sum
is correct)
Problem Set 3.4 (p188)
5
17 (this is Napiers method; he made physical rods to accelerate the
process)
19, 33
128CHAPTER 20. NOTES FOR THE SIXTHWEEK: DIGITS, BASES, ANDOPERATIONS
Chapter 21
Solutions for sixth weeks
assignments
Also available as PDF.
Note: These are my approaches to these problems. There are many ways to
tackle each.
21.1 Problem set 3.1
Problem 8 Assuming the accounting style and not the calender style, the rst
number to add is 5 20
2
+ 6 20 + 13 and the second is 11 20 + 8, but we
dont need these forms. We can add Mayan digit by Mayan digit instead.
The bottom digit is three bars and six dots, which simplies to four bars
and one dot, or one dot and one dot carried up to the next digit. The
next digit is three bars and three dots, where one of those dots is the carry.
There are no carries here, so the top digit is just a bar. The nal result:
One dot in the bottom digit, three bars and three bars in the
next digit up, then a single bar in the top-most digit.
Problem 16 24 872 = 2 10
4
+4 10
3
+8 10
2
+7 10
1
+2 10
0
, 3 071 =
3 10
3
+0 10
2
+7 10
1
+1 10
0
Problem 32 500 +60 + 9 = 569
Problem 34 300 +80 + 5 = 385
Problem 35 2 10 + 18 = 2 10 + 10 + 8 = (2 + 1) 10 + 8 = 3 10 + 8 = 38.
129
130 CHAPTER 21. SOLUTIONS FOR SIXTH WEEKS ASSIGNMENTS
21.2 Problem set 3.2
Problem 3 The last digit, the one corresponding to 6
0
= 1, is constant down
the table. The rst digit, the one corresponding to 6
1
= 6, is constant
across the table. Along each diagonal, both digits increase by one at each
step. There are other patterns, but these are some of the most obvious.
Problem 4 The next two rows:
60 61 62 63 64 65
70 71 72 73 74 75
Problem 5 413
5
= 4 5
2
+ 1 5 + 3 = 108
2004
5
= 2 5
3
+ 4 = 254
10
5
= 1 5 = 5
100
5
= 1 5
2
= 25
1000
5
= 1 5
3
= 125
2134
5
= 2 5
3
+ 1 5
2
+ 3 5 + 4 = 294
Problem 6 413
6
= 4 6
2
+ 1 6 + 3 = 153
2004
6
= 2 6
3
+ 4 = 436
10
6
= 1 6 = 6
100
6
= 1 6
2
= 36
1000
6
= 1 6
3
= 216
2134
6
= 2 6
3
+ 1 6
2
+ 3 6 + 4 = 490
Problem 8 362 = 2422
5
27 = 102
5
5 = 10
5
25 = 100
5
Problem 11 2
0
= 1
2
1
= 2
2
2
= 4
2
3
= 8
2
4
= 16
2
5
= 32
2
6
= 64
2
7
= 128
2
8
= 256
2
9
= 512
2
10
= 1024
21.2. PROBLEM SET 3.2 131
1101
2
= 13
111
2
= 7
1000
2
= 8
10101
2
= 21
24 = 11000
2
18 = 10010
2
2 = 10
2
8 = 1000
2
0 = 0
2
= 00000
2
1 = 1
2
= 00001
2
2 = 10
2
= 00010
2
3 = 11
2
= 00011
2
4 = 100
2
= 00100
2
5 = 101
2
= 00101
2
6 = 110
2
= 00110
2
7 = 111
2
= 00111
2
8 = 1000
2
= 01000
2
9 = 1001
2
= 01001
2
10 = 1010
2
= 01010
2
11 = 1011
2
= 01011
2
12 = 1100
2
= 01100
2
13 = 1101
2
= 01101
2
14 = 1110
2
= 01110
2
15 = 1111
2
= 01111
2
16 = 10000
2
= 10000
2
17 = 10001
2
= 10001
2
18 = 10010
2
= 10010
2
19 = 10011
2
= 10011
2
20 = 10100
2
= 10100
2
21 = 10101
2
= 10101
2
22 = 10110
2
= 10110
2
23 = 10111
2
= 10111
2
24 = 11000
2
= 11000
2
25 = 11001
2
= 11001
2
26 = 11010
2
= 11010
2
27 = 11011
2
= 11011
2
28 = 11100
2
= 11100
2
29 = 11101
2
= 11101
2
30 = 11110
2
= 11110
2
31 = 11111
2
= 11111
2
One of the more important patterns to see is how the digits repeat once
132 CHAPTER 21. SOLUTIONS FOR SIXTH WEEKS ASSIGNMENTS
padded to the left by zeros. The units (or right-most) digit alternates
0, 1, 0, 1, . . .. The next alternates in pairs, 0, 0, 1, 1, 0, 0, 1, 1, . . .. The
next in groups of four, the next in groups of eight, and so forth.
In one sentence: Theyre the same thing.
Problem 20 Base two digits take two values, so a three-digit numeral may
take one of 2 2 2 = 8 values.
The problem is the same, so there are eight possible subsets. This is
how we described the power-set operation, and P(A) = 2
|A|
for any
set A.
The same argument or reference provides 2
4
= 16 subsets.
The general result is 2
n
.
21.3 Problem set 3.3
Problem 10 In order, the steps are associativity, associativity, commu-
tativity, associativity, associativity, and distributivity of addition
over multiplication.
Problem 12 + 0 1 2 3
0 0 1 2 3
1 1 2 3 0
2 2 3 0 1
3 3 0 1 2
Problem 20 Here, 1 + 4 = 0 in the units digit, so the base must be 5.
2 + 4 = 0 two places over, and 3 + 1 = 4 without a carry before it, so
the base must be 6.
There are no carries in the sum at all, so the base is indeterminant.
The base can be any integer at least 7 because the largest digit
that appears is 6.
3+1 = 0, but the before it 4+4 = 4 must carry and itself be afected
by a carry. The base is 5.
Swap the subtraction around, so 236 + 254 = 523. If 6 + 4 = 3 plus a
carry, the base is 7.
247 + 254 = 523. With 7 + 4 = 3 plus a carry, the base is 8.
28E+254 = 523. Now E+4 = 3 plus a carry, so the base is E+1 = F
or 15.
21.4. PROBLEM SET 3.4 133
21.4 Problem set 3.4
Problem 5 In order of blanks, distributivity of addition over multiplica-
tion, commutativity of multiplication, associativity of addition,
and distributivity of addition over multiplication.
Problem 17 3 7 4
0 0/6 1/4 0/8 2
8 0/3 0/7 0/4 1
0 1/5 3/5 2/0 5
4 1 0
So 374 215 = 80410.
The lattice algorithm works digit by digit. This is the same as the
method in class but using diagonals rather than writing in columns.
I would hope so, but it hasnt so far. Working with a set of Napiers
bones could help. Those are physical rods you use to build these
columns.
Problem 19 The Egyptian algorithm essentially converts the multiplier
to binary. Then it doubles the multiplicand at each step and adds
the intermediate product to the result if the corresponding bit of the
multiplier is 1.
I have never heard of the duplation algorithm (perhaps duplication?).
But one way to write out the method is to convert 24 to binary,
24 = 11000
2
. Now start with an accumulator of 0:
bit of 17 doubling of 71 accumulator
0
0 71 0
0 142 0
0 284 0
1 568 568
1 1136 1704
So 71 24 = 1704.
Problem 33 Problem Actual answer Calculators answer
8 3 24 34
9 5 45 55
4 2 8 18
8 4 32 42
a b x 10 x
9 6 64 10 = 54 64
134 CHAPTER 21. SOLUTIONS FOR SIXTH WEEKS ASSIGNMENTS
When the decimal logic was done outside a chip (the early 80s), this type of
failure could happen from a simple short-circuit. Now its highly doubtful;
a failure like this would require massive trauma to the chip, leading to
general failure.
Chapter 22
Notes for the seventh week:
primes, factorization, and
modular arithmetic
Notes also available as PDF.
This week, we will cover the following topics from Chapter 4 and Chapter 5:
divisibility and prime numbers,
factorization into primes,
modular arithmetic, and
nding divisibility rules.
Im pulling modular arithmetic (clock arithmetic) from Chapter 5 because it
explains divisibility rules. Well return to the clock form in the future.
Once upon a time, number theory was both decried and revered as being pure
mathematics with no practical applications. That is no longer remotely true.
There are oblique applications in error correction (e.g. how CDs still play when
scratched), but one overwhelming, direct application is in encryption.
So I also will discuss at some point
Eulers totient function ((n)) and the RSA encryption algorithm.
The RSA algorithm is at the core of the secure socket layer (SSL) protocol used
to secure web access (the https prex, colored locks, etc.).
135
136CHAPTER 22. NOTES FOR THE SEVENTHWEEK: PRIMES, FACTORIZATION, ANDMODULAR ARITHMETIC
22.1 Divisibility
When dening operations on integers, we skipped division. As with subtraction,
the integers are not closed over division; 1/2 is not an integer. So we dene
division implicitly.
For any integers a and b, we can write
b = q a +r,
where q is an integer called the quotient, and r < |a| is a non-negative integer
called is the remainder, residue, or residual. We will see that requiring
0 r < |a| is very important and makes division well-dened.
Then a divides b, or a | b, when r = 0. Alternately, b is a multiple of a and a
is a divisor of b. If we cannot write b = q a + r with r = 0, then a does not
divide b, or a b. When a | b, then we dene division as b/a = q.
For example,
14 = 2 7 + 0, so 7 | 14 and 14/7 = 2, and
20 = 2 7 + 6, so 7 20 and 20/7 is not an integer.
In the latter case, though, 20/7 = 2 + 6/7, which rounds down to 2.
Some other examples showing extreme and negative cases,
6 = 3 2 + 0, so 2 | 6 and 6/2 = 3,
6 = 3 2 + 0, so 2 | 6 and 6/ 2 = 3,
6 = 3 2 + 0, so 2 | 6 and 6/ 2 = 3,
7 = 4 2 + 1, so 2 7,
7 = 4 2 + 1, so 2 7,
7 = 3 2 + 1, so 2 7 (note: not 4 2 1),
5 = 0 10 + 5, so 10 5, and
0 = 0 13 + 0, so 13 | 0 and 0/13 = 0.
What about when a = 0? Then b = q 0 +b is true for any quotient q. Without
further restrictions on q, division by zero is not be well-dened. In calculus and
some applications, there are times when you ll in the hole left by a division by
zero by some obvious completion.
But is the form b = qa +r well-dened when a = 0?
Theorem: The expansion b = qa + r with 0 r < |a| is unique for a = 0, so
division is well-dened.
Proof. We begin by assuming there are two ways of expanding b = qa +r. Then
we show that the forms must be identical.
22.2. PRIMES 137
Let there be two distinct ways of writing b = qa +r with a = 0,
b = q
1
a +r
1
, and
b = q
2
a +r
2
,
with 0 r
1
< |a| and 0 r
2
< |a|.
If r
1
= r
2
, then b r
1
= b r
2
. From the equations above b r
1
= q
1
a and
b r
2
= q
2
a, so q
1
a = q
2
a or (q
1
q
2
)a = 0. Because a = 0, q
1
= q
2
and the
forms are identicall.
For r
1
= r
2
, we know one of them is larger. Without loss of generality, assume
r
1
< r
2
. Then there is some positive integer k such that increases r
1
to match
r
2
, or r
2
= r
1
+k. Note that k r
2
.
Substituting for r
2
, we see that b = q
2
a +r
1
+k, or equivalently b k = q
2
a +r
1
.
Now we can subtract this equation from b = q
1
a +r
1
to obtain
k = (q
1
q
2
)a = z a + 0
for some quotient z.
So a | k, but k r
2
< |a|. The only way we can satisfy this is if q
1
q
2
= 0 and
q
1
= q
2
. Thus also k = 0 and r
1
= r
2
. So we cannot have to dierent ways to
write b = q a +r, and our form of division is well-dened.
Some useful properties of divisibility:
If d | a and d | b, then d | ra + sb for all integers r and s. A quick proof:
a = q
a
d and b = q
b
d, so ra+sb = rq
a
d+sq
b
d = (rq
a
+sq
b
)d, then d|ra+sb.
If a | b and b | c, then a | c. Quick proof: b = q
a
a, c = q
b
b, so c = q
b
(q
a
a) =
(q
b
q
a
)a.
If a | bc and a b, then a | c.
22.2 Primes
Divisibility gives a numbers a multiplicative structure thats dierent than the
digit-wise structure we previously examined.
To build the structure, we start from numbers which cannot be decomposed. An
integer p > 1 is called a prime number if its only divisors are 1 and p itself. We
will explain why 1 is not considered prime when we discuss factorization. All
other numbers are composite and must have some prime divisor.
138CHAPTER 22. NOTES FOR THE SEVENTHWEEK: PRIMES, FACTORIZATION, ANDMODULAR ARITHMETIC
Consider possible divisors of 11,
11 = 11 1 + 0 so 1 | 11,
11 = 5 2 + 1 so 2 11, and
11 = 3 3 + 2 so 3 11.
We can stop at 3. Because multiplication is commutative, any divisors come in
pairs. The smaller of the pair must be

11 3.3; thats the point where any

pairs a b are repeated as b a.
So the only divisor less than

11 is 1, and 11 is prime.
How many primes are there?
Theorem: There are innitely many primes.
Proof. Assume there are only k primes p
1
, p
2
, . . . , p
k
and all other numbers are
composite. Then let n = p
1
p
2
p
k
+ 1, one larger than the product of all
primes.
Consider dividing n by some prime, say p
k
. Then we can write n = (p
1
p
2

p
k1
)p
i
+ 1. Given the form is unique and r = 1, p
k
does not divide n. We
could have chosen any of the primes, so p
i
n for all i = 1, . . . , k. Thus no prime
divides n.
For a number n to be composite, it must have some factor or divisor other than
1 and n. If that factor is not prime, then the factor has another factor, and so
forth until you reach some prime. Because of transitivity of division (a | b and
b | c imply a | c), the prime must divide n. Here, though, no primes divide n, so
n cannot be composite and must be prime itself.
So assuming there are k primes leads to a contradiction because we can construct
one more. Thus there are either no primes or innitely many. We demonstrated
that 11 is prime, so there must be innitely many primes.
There are mountains of unanswered questions about prime numbers. Consider
the pairs of primes (3, 5), (5, 7), (11, 13), and (17, 19). Each are separated by
two. Are there innitely many such pairs? No one knows. Similarly, there are
Mersenne primes of the form 2
n
1. No one knows how many Mersenne
primes exist.
22.3 Factorization
A factorization of a number is a decomposition into factors. So 24 = 8 3 is a
factorization of 24, as is 24 = 4 2 3. A prime factorization is a factorization
22.4. MODULAR ARITHMETIC 139
into primes. Here 24 = 2 2 2 3 is a prime factorization of 24. We use exponents
to make this easier to write, and 24 = 2
3
3.
You can be systematic about prime factorization and discover the primes through
the sieve of Eratosthenes. Consider nding a prime factorization of 110.
We start just by writing possible factors. Technically we need integers only

1100 33.6, but we only ll enough here to demonstrate the point.

2 3 4 5 6 7 8 9 10
11 12 13 14 15 16 17 18 19 20
The rst possible factor is 2, and 2 | 1100. We divide by two until the result is
not divisible by 2. This gives 1100 = 2
2
275. Then we cross out all multiples of
2; these cannot divide 275. Because 275 = 91 3 + 2, 3 275. But we also know
no multiples of 3 divide 275, so we cross out all remaining multiples of 3.
The next not crossed out is 5, which divides 275. Now 1100 = 2
2
5
2
11. We
showed 11 is prime before, but lets continue this method. Cross out all multiples
of ve. The next number to try is 7, which does not divide 11. But we can cross
out all multiples of 7. The next free number is 11, which we have again shown
to be prime.
In our (short) list, it happens that only primes remain. We have sieved out all
the non-primes. Actually, once we removed all multiples of primes

20, only
primes remained.
Moving from one prime to the next is a systematic method both for nding
prime numbers and for nding a prime factorization.
Factorizations provide a useful mechanism for working
We will not prove the following, but is often called the fundamental theorem
of arithmetic:
Theorem: Every integer greater than one has a unique prime factorization.
22.4 Modular Arithmetic
Factorization itself will prove useful later. Now we will explore modular arithmetic
and nd some quick rules for determining when d | a for some d.
Modular arithmetic is arithmetic on remainders.
Consider expressions of 7 and 4 in terms of multiples of 3 plus remainders:
7 = 2 3 +1 and 4 = 1 3 +1. Now 11 = 7 +4 = (2 3 +1) +(1 3) +1 = 3 3 +2.
Note that the sum of the remainders was < 3 and was the new remainder itself.
If the remainder is 3, we can just pull a three out of it: 11 + 8 = (3 3 +
2) + (2 3 + 2) = 5 3 + 4. To convert this into the correct form, note that
140CHAPTER 22. NOTES FOR THE SEVENTHWEEK: PRIMES, FACTORIZATION, ANDMODULAR ARITHMETIC
4 = 1 3 + 1, and 19 = 5 3 + (1 3 + 1) = 6 3 + 1. We need consider only the
sum of remainders to compute the results remainder.
The remainder of the sum just wraps around. Think about time. If you add a
few hours and cross 12, the result just wraps around. So 1:00 is the same as
13:00 or 25:00.
We dont identify 1:00 as just one time but a member of a set of all times that are
one hour after a multiple of 12. Similarly, we can identify numbers as elements
of sets where all members have the same remainder relative to a given divisor.
The congruence class of r modulo a is {x| q : x = qa +r}. If a number b is
in the congruence class of r modulo a, we write b r (mod a).
The canonical member of a congruence class is its least positive member. Just as
we dont naturally consider 25:00 as 1:00, we tend to identify congruence classes
by the least r. So while 13 87 (mod 2) is correct (both 13 and 87 are odd),
we prefer 13 1 (mod 2).
We dene addition and multiplication on entire congruence classes. For the
operation to be dened, the modulus of each class must be the same. Then were
adding numbers of the form b
1
= q
1
a +r
1
for b
1
r
1
(mod a) and b
2
= q
2
a +r
2
for b
2
r
2
(mod a). As in our example above, the remainders add. Here
b
1
+b
2
= q
1
a +r
1
+q
2
a +r
2
= (q
1
+q
2
)a + (r
1
+r
2
) r
1
+r
2
(mod a).
Identifying congruence classes by their least positive element, we can write a
table showing all additions modulo 4:
+ (mod 4) 0 1 2 3
0 0 1 2 3
1 1 2 3 0
2 2 3 0 1
3 3 0 1 2
Addition of congruence classes maintains the additive identity that we expect,
b + 0 b (mod a).
Note that every class has an additive inverse, a class where b + (b) 0
(mod a). Remember that we forced the residual to be positive when we dened
division. Then we can see that the inverse of 1 modulo 4 is 1 = 1 4 + 3 3
(mod 4).
Another way to see this is that the canonical representation of b is the least
number which increases b to be equal to the modulus a. So the inverse of 1 is 3
because 1 + 3 = 4 0 (mod 4).
We also dene multiplication on congruence classes.
22.5. DIVISIBILITY RULES 141
If b
1
= q
1
a +r
1
and b
2
= q
2
a +r
2
, then
b
1
b
2
= (q
1
a +r
1
) (q
2
a +r
2
)
= q
1
q
2
a
2
+q
1
r
2
a +q
2
r
1
a +r
1
r
2
= (q
1
q
2
a +q
1
r
2
+q
2
r
1
)a +r
1
r
2
r
1
r
2
(mod a).
So we need only multiply remainders.
Identifying congruence classes by their least positive element, we can write a
table showing all multiplications modulo 4:
(mod 4) 0 1 2 3
0 0 0 0 0
1 0 1 2 3
2 0 2 0 2
3 0 3 2 1
Again, there is a multiplicative identity, b 1 b (mod a).
Unlike plain integer division, some congruence classes have an inverse. The
only integer that has an integer inverse is 1. But modulo 4, both 1 and 3 have
multiplicative inverses. Here 3 3 = 9 1 (mod 3).
22.5 Divisibility Rules
Using modular arithmetic and positional notation, we can derive some quick
divisibility tests.
First, consider divisibility by powers of 2 and 5. The factorization of 10 = 2 5,
and so 10
k
= 2
k
5
k
. So 2
k
| 10
k
and 5
k
| 10
k
, or 10
k
0 (mod 2)
k
and 10
k
0
(mod 5)
k
.
Now remember how to expand positional notation. We know that 1234 =
1 10
3
+ 2 10
2
+ 3 10
1
+ 4. So 1 10
3
+ 2 10
2
+ 3 10
1
+ 4 0 + 0 + 0 + 4
(mod 2) 0 (mod 2). Divisibility by 2 depends only on the nal digit. Similarly,
1234 0 + 0 + 0 + 4 (mod 5), and divisibility by 5 depends only on the nal
digit.
For 2
2
= 4 and 5
2
= 25, all but the last two digits are equivalent to zero. And
for 2
3
= 8 and 5
3
= 125, all but the last three digits are equivalent to zero. So
one divisibility rule:
When testing for divisibility by 2
k
or 5
k
, we need only consider the
last k digits.
Now consider divisibility by 3 or 9. We know that 10 1 (mod 3) and 10 1
(mod 9). Using modular arithmetic, 123 = 1 10
2
+ 2 10 + 3 1 + 2 + 3
142CHAPTER 22. NOTES FOR THE SEVENTHWEEK: PRIMES, FACTORIZATION, ANDMODULAR ARITHMETIC
(mod 3) 6 (mod 3) 0 (mod 3). Hence 3 | 123 because the sum of its digits
is divisible by 3.
Similarly, 10 1 (mod 3). So 123 1+2+3 (mod 9) 6 (mod 9), and 9 123.
If the sum of the digits is greater than 9, simply add those digits.
Test for divisibility by 3 or 9 by adding the numbers digits and
checking that sum. If that sum is greater than 9, add the digits again.
Repeat until the result is obvious.
Other primes are not so straight-forward. Divisibility by 7 is a pain; there is an
example method in the texts problems for Section 5.1.
The rule for 11 is worth exploring. Because 10 < 11, the canonical member of
its congruence class is just 10. But there is another member of interest, 10 1
(mod 1)1. So you can alternate signs on alternate digits from the right. So
123456 1 + 2 3 + 4 5 + 6 3 (mod 1)1, and 11 123456.
For divisibility by 6, 12, 18, or other composite numbers, factor the divisor and
test for divisibility by each factor. To test for divisibility by 72 = 2
3
3
2
= 8 9,
test for divisibility by 8 and by 9.
22.6 Homework
Problem Set 4.1 (p242):
Problem 2, but dont repeat the drawings.
4, 7, 9, 10, 13, 14, 23, 24
Also draw diagrams showing that 8 18 and 3 11.
Problem Set 4.2 (p252):
1, 2, 8, 14, 15
Take a familiar incomplete integer, 679 . Using the expression of 679
as N = 10
4
x
4
+x
0
+6790, use 8 | N to nd x
0
? Given that, use 9 | N to
nd x
4
. Now if 72 turkeys cost $ 679 , what is the total?
Chapter 23
Solutions for seventh weeks
assignments
Also available as PDF.
Note: These are my approaches to these problems. There are many ways to
tackle each.
23.1 Problem set 4.1
Problem 2 The (non-repeated) factorizations are 1 35 and 5 7. Drawing those
as boxes is straight-forward.
Problem 4 Factors of 18 1 2 3 6 9 18
Quotient 18 9 6 3 2 1
Problem 7 48 = 1 48 = 2 24 = 3 16 = 4 6, so the factors are 1, 2, 3, 4,
6, 16, 24, and 48.
54 = 1 54 = 2 27 = 3 18 = 6 9, so the factors are 1, 2, 4, 6, 9 18,
27, 54.
The largest common factor then is 6.
Problem 9 48 = 2
4
3
1
108 = 2
2
3
3
2250 = 2
2
3
2
5
3
24 750 = 2
1
3
2
5
3
11
1
143
144CHAPTER 23. SOLUTIONS FOR SEVENTH WEEKS ASSIGNMENTS
Problem 10 Yes, because all primes are shared and no powers of the
primes exceed those of a.
No, because 3
2
has a larger exponent than the 3
1
in a.
a/b = (2
3
3
1
7
2
)/(2
2
3
1
) = 2
32
3
11
7
20
= 2
1
7
2
.
There are (2 + 1) (1 + 1) (2 + 1) = 18 factors of a.
We can make a list by running up the exponents:
2
0
3
0
7
0
2
1
3
0
7
0
2
2
3
0
7
0
2
3
3
0
7
0
2
0
3
1
7
0
2
1
3
1
7
0
2
2
3
1
7
0
2
3
3
1
7
0
2
0
3
0
7
1
2
1
3
0
7
1
2
2
3
0
7
1
2
3
3
0
7
1
2
0
3
1
7
1
2
1
3
1
7
1
2
2
3
1
7
1
2
3
3
1
7
1
2
0
3
0
7
2
2
1
3
0
7
2
2
2
3
0
7
2
2
3
3
0
7
2
2
0
3
1
7
2
2
1
3
1
7
2
2
2
3
1
7
2
2
3
3
1
7
2
Problem 13 No, it is only true that at least one prime factor cannot exceed

n. Consider 2 47 = 94, where both 2 and 47 are prime. We have

94 < 10 but 47 > 10 >

94. But 2 <

94.
Problem 14 No. If n is not prime, some of its factors may split between b and
c. For example, 2 3 = 6 | 18 = 2 9, but 6 2 and 6 9. The factors of
n = 6, 2 and 3, are split between a = 2 and b = 9.
Problem 23 Here it is true. Consider the prime factorizations of b and c. If
p | bc, then p must appear in one or both of those prime factorizations,
and thus it must divide at least one of b and c.
Problem 24 Again, use the prime factorization of n. Because p and q are
primes, they must appear in that factorization. Then pq | n because both
23.2. TWO DIAGRAMS 145
appear, so you can commute products around to group (pq) and the rest
of the factorization.
23.2 Two diagrams
8 18: 18 = 2 8+2, so you can draw and count: +
3 11: 11 = 3 3 + 2: +
23.3 Problem set 4.2
Problem 1 1554 is even, so divisible by 2, does not end in 5 or 0, so
is not divisible by 5, and has digits that add to 0 (mod 3), so is
divisible by 3.
1999 is not even, so is not divisible by 2, does not end in 5 or 0,
so is not divisible by 5, and has digits that add to 1 (mod 3), so
is not divisible by 3.
805 is not divisible by 2, is divisible by 5, and is not divisible
by 3.
2450 is divisible by 2, is divisible by 5, and is not divisible by
3.
Problem 2 2 and 3
2 and 5
3 and 5
2, 3, and 5
Problem 8 The missing digit must be divisible by 2. Also, the sum of the
digits must be congruent to 0 modulo 3. The non-blank digits already add
to 0 (mod 3), so the missing digit must be a multiple of 3. There is only
one even multiple of 3 less than 10, so the missing digit must be 6.
Problem 14 Expanding the positional notation and simplifying, abc, abc =
a(10
5
+10
2
)+b(10
4
+10
1
)+c(10
3
+10
0
) = (a10
2
+b10
1
+c)(10
3
+10
0
) =
abc 1001. Now 1001 = 7 11 13, so abc, abc is divisible by each of those.
Problem 15 ab ba = a 10 + b (b 10 + a) = (a b) 10 + (b a).
Because 10 1 (mod 9), this becomes a b + b a 0 (mod 9).
The result always is a multiple of 9. Equivalently, we can rearrange
146CHAPTER 23. SOLUTIONS FOR SEVENTH WEEKS ASSIGNMENTS
(ab)10+(ba) = (ab)10+1(ab) = (ab)(101) = 9(ab),
giving also which multiple of 9.
Here the dierence is (ac) 10
2
+0+(ca) because the middle digit
always cancels. Again, the result always is a multiple of 9 and we can
rearrange (a c) 10
2
+ 0 + (c a) = (a c) 10
2
+1 (a c) =
(a c) (10
2
1) = 99(a c) to see the result is a multiple of 99.
23.4 A familiar incomplete integer
Take a familiar incomplete integer, 679 . Using the expression of
679 as N = 10
4
x
4
+x
0
+6790, use 8 | N to nd x
0
. Given that,
use 9 | N to nd x
4
. Now if 72 turkeys cost $ 679 , what is the
total?
If 8 | N then 8 divides the last three digits, so 8 | 790 +x
0
. Thus 790 +x
0
0
(mod 8). Because 790 6 (mod 8), we know that x
0
2 (mod 8). The only
decimal digit satisfying x
0
2 (mod 8) is x
0
= 2.
Now we have N = 10
4
x
4
+ 6792. For 9 | N, the sum of the digits must be zero
modulo 9. Thus x
4
+ 6 + 7 + 9 + 2 0 (mod 9), or x
4
+ 6 0 (mod 9). Thus
x
4
= 3, and N = 36792.
(I forgot the decimal place in the problem, so these are very expensive turkeys.)
So if 72 turkeys cost $36792, each turkey costs $511. If I had remembered the
decimal place correctly, the turkeys cost $5.11 each.
Chapter 24
Notes for the eighth week:
GCD, LCM, and ax + by = c
Notes also available as PDF.
What we covered last week:
divisibility and prime numbers,
factorization into primes,
modular arithmetic,
nding divisibility rules,
This weeks topics:
review modular arithmetic and nding divisibility rules,
greatest common divisors and least common factors,
Euclids algorithm for greatest common divisors, and
solving linear Diophantine equations.
These all are useful when you deal with integral numbers of things
24.1 Modular arithmetic
Remember the divisibility form for b with respect to dividing by a = 0,
b = q a +r, with 0 r < |a|.
This form is unique for a given a and b.
147
148CHAPTER 24. NOTES FOR THE EIGHTHWEEK: GCD, LCM, ANDAX+BY = C
Consider a = 5. There are only ve possible values of r, zero through four.
Because the form is unique, we can place every b into one of r congruence
classes. Each congruence class is a set. For a = 5, we have the following classes:
{. . ., -10, -5, 0, 5, 10, . . .} = {5k + 0 | k J}
{. . ., -9, -4, 1, 6, 11, . . .} = {5k + 1 | k J}
{. . ., -8, -3, 2, 7, 12, . . .} = {5k + 2 | k J}
{. . ., -7, -2, 3, 8, 13, . . .} = {5k + 3 | k J}
{. . ., -6, -1, 4, 9, 14, . . .} = {5k + 4 | k J}
We say that two numbers are in the same congruence class for a given a by
b c (mod a).
Or b is equivalent to c modulo a. A collection of one entry from each set is called
a complete residue system. We typically select the least positive numbers,
those in bold above.
We dene arithmetic on congruence classes by arithmetic on the remainders. The
remainders wrap around every multiple of the modulus. For example, addition
modulo 4 and modulo 5 are dened as follows:
+ (mod 4) 0 1 2 3
0 0 1 2 3
1 1 2 3 0
2 2 3 0 1
3 3 0 1 2
+ (mod 5) 0 1 2 3 4
0 0 1 2 3 4
1 1 2 3 4 0
2 2 3 4 0 1
3 3 4 0 1 2
4 4 0 1 2 3
This works as you expect. Addition is commutative and associative. There is
an additive identity, because b + 0 0 (mod a). Unlike the positive integers,
there also is an additive inverse for every residue because b + (a b) 0
(mod a).
Multiplication likewise is commutative and associative, and there is a mul-
tiplicative identity, 1. The unusual aspect appears with the multiplicative
inverse. Some residues have inverses, and some dont:
(mod 4) 0 1 2 3
0 0 0 0 0
1 0 1 2 3
2 0 2 0 2
3 0 3 2 1
(mod 5) 0 1 2 3 4
0 0 0 0 0 0
1 0 1 2 3 4
2 0 2 4 1 3
3 0 3 1 4 2
4 0 4 3 2 1
The dierence here is that 5 is prime while 4 is composite. Any factor of the
modulus will not have a multiplicative inverse.
24.2. DIVISIBILITY RULES 149
24.2 Divisibility rules
One common application of modular arithmetic (besides telling time) is in
testing whether one integer divides another. We use modular arithmetic and
positional notation. Both help us break the larger problem, testing divisibility
of a potentially large number, into the smaller problems of breaking apart the
number and evaluating expressions in modular arithmetic.
If a | b (a divides b), then b 0 (mod a). So we can test for divisibility by
expanding b in positional notation and evaluating the operations modulo a.
When the divisor is small, a straight-forward evaluation is simplest. Because
10 1 (mod 3), we can test for divisibility by 3 by adding the numbers digits
modulo 3. For example,
1234 10
3
+ 2 10
2
+ 3 10 + 4 (mod 3)
1
3
+ 2 1
2
+ 3 1 + 4 (mod 3)
1 + 2 + 3 + 4 1 + 2 + 0 + 1 1 (mod 3).
Hence 3 1234. The same trick applies to 9 because 10 1 (mod 9).
When the divisor is closer to a power of 10, using a negative element of the
congruence class may be useful. For 11, remember that 10 and 1 are in the
same congruence class because 10 = 0 11+10 and 1 = 1 11+10. So 10 1
(mod 11) and we can expand the powers of ten,
1234 10
3
+ 2 10
2
+ 3 10 + 4 (mod 11)
(1)
3
+ 2 (1)
2
+ 3 (1) + 4 (mod 11)
1 + 2 +3 + 4 (mod 11) 2 (mod 11).
Hence 11 1234. Here, the trick form is that you start from the units digit
and then alternate subtracting and adding digits.
For more complicated examples, we can factor the divisor. To test if a number
is divisible by 72, factor 72 = 2
3
3
2
= 8 9. Then test if the number is divisible
by 8 and by 9.
If a | b and c | b, then it may be true that ac | b. This is certainly true of a and
b are powers of dierent primes. The key point is that a and b share no common
divisors. Note that 72 = 6 12, 6 | 24, and 12 | 24, but obviously 72 24 because
24 < 72.
24.3 Greatest common divisor
So nding common divisors is useful for testing divisibility. The greatest common
divisor of numerator and denominator reduces a fraction into its simplest form.
150CHAPTER 24. NOTES FOR THE EIGHTHWEEK: GCD, LCM, ANDAX+BY = C
In general, common divisors help break problems apart.
Written (a, b) or gcd(a, b), the greatest common divisor of a and b is
the largest integer d 1 that divides both a and b.
Well discuss a total of two methods for nding the greatest common divisor.
The rst uses the prime factorization, and the second uses the divisibility form in
the Euclidean algorithm. Later well extend the Euclidean algorithm to provide
integer solutions x and y to equations ax +by = c.
The prime factorization method factors both a and b. Consider a = 1400 =
2
3
5
2
7 and b = 1350 = 2 3
3
5
2
.
Lining up the factorizations and remembering that x
0
= 1, we have
a = 1400 = 2
3
3
0
5
2
7
1
, and
b = 1350 = 2
1
3
3
5
2
7
0
.
Now chose the least exponent for each factor. Then
d = 2
1
3
0
5
2
7
0
= 50
is the greatest common divisor. For more than two integers, factor all the
integers and nd the least exponents across the corresponding factors in all of
the factorizations.
For an example use, reduce a fraction a/b = 1350/1400 to its simplest form. To
do so, divide the top and bottom by d = 50. Then a/b = 1350/1400 = 27/28.
Now we can state the requirement about divisibility given some factors:
If two relatively prime integers a and b both divide c, then ab divides
c.
Some other properties of the gcd:
Because the gcd is positive, (a, b) = (|a|, |b|).
(a, b) = (b, a)
If the gcd of two numbers is 1, or (a, b) = 1, then a and b are called
relatively prime.
24.4 Least common multiple
Before the other method for nding the gcd, we consider one related quantity.
The least common multiple, often written lcm(a, b), is the least
number L a and L b such that a | L and b | L.
There are clear, every day uses. Think of increasing a recipe when you can only
buy whole bags of some ingredient. You need to nd the least common multiple
24.5. EUCLIDEAN GCD ALGORITHM 151
of the recipes requirement and the bags quantity. Or when you need to nd
the next day two dierent schedules intersect.
Again, you can work from the prime factorizations
a = 1400 = 2
3
3
0
5
2
7
1
, and
b = 1350 = 2
1
3
3
5
2
7
0
.
Now the least common multiple is the product of the larger exponents,
lcm(a, b) = 2
3
3
3
5
2
7
1
= 37 800.
And for more than two integers, take the maximum across all the exponents of
corresponding factors.
Another relation for two integers a and b is that
lcm(a, b) =
ab
d
.
So given a = 1350, b = 1400, and d = 50,
lcm(1350, 1400) =
1350 1400
50
=
1 890 000
50
= 37 800.
This does not hold directly for more than two integers.
24.5 Euclidean GCD algorithm
Another method for computing the gcd of two integers a and b is due to Euclid.
This often is called the rst algorithm expressed as an abstract sequence of steps.
We start with the division form of b in terms of a = 0,
b = qa +r with 0 r < a.
Because (a, b) = (|a|, |b|), we can assume both a and b are non-negative. And
because (a, b) = (b, a), we can assume b a.
Let d = (a, b). Last week we showed that if d|a and d|b, then d|ra +sb for any
integers r and s. Then because d|a and d|b, we have d|b qa or d|r. So we have
that d = (b, a) also divides r. Note that any number that divides a and r also
divides b, so d = (a, r).
Continuing, we can express a in terms of r as
a = q

r +r

with 0 r

< r.
Now d|r

and d = (r, r

). Note that r

< r < a, so the problem keeps getting

smaller! Eventually, some remainder will be zero. Then the previous remainder
is the greatest common divisor.
152CHAPTER 24. NOTES FOR THE EIGHTHWEEK: GCD, LCM, ANDAX+BY = C
1. Find q
0
and r
0
in b = q
0
a +r
0
with 0 r
0
< a.
2. If r
0
= 0, then (a, b) = a.
3. Let r
1
= a to make the loop easier to express.
4. Then for i = 1, . . .
(a) Find q
i
and r
i
in r
i2
= q
i
r
i1
+r
i
with 0 r
i
< r
i1
.
(b) If r
i
= 0, then (a, b) = r
i1
and quit.
(c) Otherwise continue to the next i.
Consider calculating (53, 77). Following the steps, we have
77 = 1 53 + 24,
53 = 2 24 + 5,
24 = 4 5 + 4,
5 = 1 4 + 1, and
4 = 4 1 + 0.
And thus (53, 77) = 1.
For another example, take (128, 308). Then
308 = 2 128 + 52,
128 = 2 52 + 24,
52 = 2 24 + 4, and
24 = 6 4 + 0.
So (128, 308) = 4.
24.6 Linear Diophantine equations
Later in the semester, we will examine linear equations ax + by = c over real
numbers. But many every-day applications require integer solutions. We can
use the Euclidean algorithm to nd one integer solution to ax +by = c or prove
there are none. Then we can use the computed gcd to walk along the line to all
integer solutions.
Say we need to solve ax+by = c for integers a, b, and c to nd integer solutions
x and y.
Let d = (a, b). Then, as before, d | ax +by for all integers x and y. So d | c for
any solutions to exist. If d c, then there are no integer solutions. If a and b
are relatively prime, then (a, b) = 1 and solutions exist for any integer c.
24.6. LINEAR DIOPHANTINE EQUATIONS 153
Consider solving ax + by = d. Because d | c, we can multiply solutions to
ax +by = d by c/d to obtain solutions of ax +by = c. To solve ax +by = d we
work backwards after using the Euclidean algorithm to compute d = (a, b).
Say the algorithm required k steps, so d = r
k1
. Working backward one step,
d = r
k1
= r
k3
q
k1
r
k2
= r
3
q
k1
(r
k4
q
k2
r
k3
)
= (1 +q
k1
q
k2
)r
3
q
k1
r
k4
.
So d = r
k1
= i r
k3
+j r
k4
where i and j are integers. Continuing, the gcd
d can be expressed as an integer combination of each pair of remainders.
Returning to the example of (77, 53),
1 = 5 1 4,
= 5 1 (24 5 5) = 5 5 1 24
= 5 (53 2 24) 1 24 = 5 53 11 24
= 5 53 11 (77 1 53) = 16 53 11 77.
To solve 53x +77y = 22, we start with 53 16 +77 (1) = 1. Multiplying by 22,
53 (16 22) + 77 (1 22) = 22,
and x = 352, y = 22 is one solution.
But if there is one solution, there are innitely many! Remember that d = (a, b),
so a/d and b/d are integers. Given one solution x = x
0
and y = y
0
, try
substituting x = x
0
+t (b/d) and y = y
0
t (a/d) for any integer t. Then
a(x
0
+t (b/d)) +b(x
0
t (a/d)) = ax
0
+bx
0
+t (ab/d)) +t (ba/d))
= ax
0
+bx
0
= c.
Actually, all integer solutions to ax +by = c are of the form
x = x
0
+t (b/d), and y = y
0
t (a/d),
where t is any integer, d = (a, b), and x
0
and y
0
are a solution pair.
Another example, consider solving 12x+25y = 331. First we apply the Euclidian
algorithm to compute (12, 25) = 1:
25 = 2 12 + 1, and
12 = 12 1 + 0.
Substituting back,
12 (2) + 25 1 = 1, and12 (662) + 25 331 = 331.
154CHAPTER 24. NOTES FOR THE EIGHTHWEEK: GCD, LCM, ANDAX+BY = C
So we can generate any solution to 12x + 25y = 331 with the equations
x = 662 + 25t and y = 331 12t.
Using these, we can nd a smaller solution. Try making x non-negative with
662 + 25t 0,
25t 662, thus
t > 26.
Substituting t = 27,
x = 13, and y = 7.
Interestingly enough, this must be the only non-negative solution. A larger t will
force y negative, and a smaller t forces x negative. But the solution for t = 26 is
still small,
x = 12, and y = 19.
24.7 Homework
Practice is absolutely critical in this class.
Groups are ne, turn in your own work. Homework is due in or before class on
Mondays.
Problem set 4.3, p264:
6, 7, 13 (using any method, not the specied one), 15, 18, 25
Compute the following using both the prime factorization method and
the Euclidean algorithm:
(720, 241)
(64, 336)
(15, 75)
Compute the least common multiples:
lcm(64, 336)
lcm(11, 17)
lcm(121, 187)
lcm(2025, 648)
Find two integer solutions to each of the following, or state why no solutions
exist:
64x + 336y = 32
24.7. HOMEWORK 155
33x 27y = 11
31x 27y = 11
Note that you may email homework. However, I dont use Microsoft
TM
products
(e.g. Word), and software packages are notoriously nicky about translating
mathematics.
If youre typing it (which I advise just for practice in whatever tools you use),
you likely want to turn in a printout. If you do want to email your submission,
please produce a PDF or PostScript document.
156CHAPTER 24. NOTES FOR THE EIGHTHWEEK: GCD, LCM, ANDAX+BY = C
Chapter 25
Solutions for eighth weeks
assignments
Also available as PDF.
Note: These are my approaches to these problems. There are many ways to
tackle each.
25.1 Problem set 4.3
Problem 6 Every integer divides zero, so (0, n) = n appears correct for n > 0.
Indeed, 0 = 0 n and n = 1 n, so n divides both. The only point for
discussion is in the denition of the greatest common divisor and divisibility.
The text denes it in an obtuse way that disallows (0, n) = n.
Most mathematicians would allow n | 0 and (0, n) = n. Indeed, n | 0
for all n is stated on the rst page of one of the most respected number
theory textbooks
1
. So, alas, the correct response is to nd out what
standardized test uses which denition. This is yet another reason why
standardized tests in higher mathematics are useless.
Problem 7 Again, by the texts denitions, this cannot be true. By most
common mathematical denitions, lcm(0, n) = 0. This is another example
where Im interested in seeing how you state and defend your position.
Problem 13 One method for nding both is to nd the prime factoriza-
tions 18 = 23
2
, 24 = 2
3
3, and 12 = 2
2
3. Then (18, 24, 12) = 23 = 6
and lcm(18, 24, 12) = 2
3
3
2
= 72.
1
Hardy, G., & Wright, E. (1979). An Introduction to the Theory of Numbers. Oxford:
Clarendon Press.
157
158 CHAPTER 25. SOLUTIONS FOR EIGHTH WEEKS ASSIGNMENTS
Here 8 = 2
3
, 20 = 2
2
5, and 14 = 2 7. So (8, 20, 14) = 2 and
lcm(8, 20, 14) = 2
3
5 7 = 280.
Verify: 2 280 = 560, 8 20 14 = 2240. So the two quantites are not
equal. The product of the gcd and lcm only equals the product of the
numbers when there are only two numbers involved or when no pairs
of numbers any factors.
Problem 15 (24, 18) = 6, and (6, 12) = 6. Also, lcm(24, 18) = 72 and
lcm(72, 12) = 72. Both give the same result, as they must.
(8, 20) = 4, (4, 14) = 2. lcm(8, 20) = 40, lcm(40, 14) = 280. Again,
these agree.
The gcd and lcm can be computed pairwise. That is, you can compute
each over the rst pair of numbers, then compute over that result
and the next number, etc.
Problem 18 Find the largest jump such that you can reach all the numbers
when starting from zero.
Take each number as a jump size. Find the rst number where all
the jumps land simultaneously.
Problem 25 This is looking for the least common multiple of 45 and 96. Think
of it like the number line problem above. You mark a notch on each
gear. Starting with that notch on zero, you roll each gear to generate
the jumps. You are looking for the point where they coincide. So
lcm(45, 96) = lcm(3 15, 3 32) = 3 15 32 = 1440 teeth will go past. In
terms of the smaller hear, this is 1440/45 = 32 revolutions.
25.2 Computing GCDs
Compute the following using both the prime factorization method
and the Euclidean algorithm:
(720, 241)
(64, 336)
(15, 75)
Prime factorizations:
241 is prime. So (720, 241) = 1.
64 = 2
6
, 336 = 2
4
3 7. (64, 336) = 2
4
= 16.
(15, 75) = (15, 75). 15 = 3 5, 75 = 3 5
2
, so (15, 75) = 15.
Euclidean algorithm:
25.3. COMPUTING LCMS 159

720 = 2 241 + 238,

241 = 1 238 + 3,
238 = 79 3 +1
3 = 3 1 + 0.
So (720, 241) = 1.

336 = 5 64 +16,
64 = 4 16 + 0.
So (336, 64) = 16.
(15, 75) = (15, 75):
75 = 5 15 + 0.
So (15, 75) = 15.
25.3 Computing LCMs
Compute the least common multiples:
lcm(64, 336)
lcm(11, 17)
lcm(121, 187)
lcm(2025, 648)
lcm(64, 336) = 64 336/(336, 64) = 1344
Both are prime, so lcm(11, 17) = 11 17 = 187
lcm(121, 187) = lcm(11
2
, 11 17) = 11
2
17 = 2057
lcm(2025, 648) = lcm(3
3
5
2
, 2
3
3
4
) = 2
3
3
4
5
2
= 16200
25.4 Linear Diophantine equations
Find two integer solutions to each of the following, or state why no
solutions exist:
64x + 336y = 32
160 CHAPTER 25. SOLUTIONS FOR EIGHTH WEEKS ASSIGNMENTS
33x 27y = 11
31x 27y = 11
From a previous problem, we have that 336 = 64 5 + 16. Thus 336 1 +
64 5 = 16 and 336 2 + 64 10 = 32. So one solution is x
0
= 10 and
y
0
= 2. The general solution is x = x
0
+t 336/(336, 64) = 10 +21t and
y = y
0
t 64/(336, 64) = 2 4t for any integer t. Another solution then
is x(1) = 10 +1 21 = 11 and y(1) = 2 4 1 = 2.
Here, (33, 27) = (3 11, 3
3
) = 3. Now 3 11, so there are no solutions.
Now 31 is prime, so (31, 27) = 1 | 11 and there are solutions. Running
through the Euclidean algorithm we see that
31 = 27 1 + 4,
27 = 4 6 + 3, and
4 = 3 1 + 1.
Starting from the bottom and substituting for the previous remainder,
4 + 3 (1) = 1,
4 + (27 + 4 (6)) 1 = 27 (1) + 4 7 = 1,
27 (1) + (31 + 27 (1)) 7 = 31 7 + 27 (8) = 1.
We nd that 31 7 +27 (8) = 1, so 31x27y = 11 has an initial solution
of x
0
= 7 11 = 77 and y
0
= 1 8 11 = 88.
The general solutions have the form
x = x
0
+t
27
(31, 27)
= 77 27t, and
y = y
0
t
31
(31, 27)
= 88 31t,
Another solution is given by x(1) = 77 27 1 = 50 and y(1) = 88 31 1 = 57.
Chapter 26
Notes for the ninth week:
ax + by = c, fractions
Notes also available as PDF.
26.1 Linear Diophantine equations
In a few weeks, we will examine linear equations ax +by = c over real numbers.
But many every-day applications require integer solutions. We can use the
Euclidean algorithm to nd one integer solution to ax +by = c or prove there
are none. Then we can use the computed gcd to walk along the line to all integer
solutions.
Some solvable problems:
A 98 pound box contains 5 pound bags of sugar and 12 pound sacks
of oranges. How many of each are in the box?
Or:
Say you need a digital image in a 4 : 3 aspect ratio (x : y) that
includes a 50 pixel border along each side. What sizes are possible
for the inner image?
Consider the latter problem. Rephrasing algebraically,
4
3
=
x + 100
y + 100
, or
3x 4y = 100.
We start by solving
3x 4y = 1
161
162CHAPTER 26. NOTES FOR THE NINTHWEEK: AX+BY = C, FRACTIONS
and then multiplying the base solutions by 100. This case has one easy solution,
x = 1 and y = 1, with 3 1 4 1 = 3 + 4 = 1. Another solution is
x = 3 and y = 2.
In fact, there are inntely many solutions to 3x 4y = 1 given by
x = 1 + 4t, and
y = 1 + 3t
for any integer t. You can substitute these expressions into 3x 4y to verify the
result. Scaling the right-hand side by 100, solutions to 3x 4y = 100 are given
by
x = 100 + 4t, and
y = 100 + 3t.
For x > 0 and y > 0, we need t > 33. So the rst positive solutions are given by
t = 34, 35, . . . and are
(x, y) {. . . , (36, 2), (40, 5), (44, 8), (48, 11), (52, 14), (56, 17), (60, 20), . . .}.
26.1.1 In general. . .
Say we need to solve ax+by = c for integers a, b, and c to nd integer solutions
x and y. In general, equations over integers are called Diophantine equations
after Diophantus of Alexandria (approx. 200AD-290AD). He was the rst known
to study these equations using algebra. The form ax +b = c describes linear
Diophantine equations.
Let d = (a, b). Then, as before, d | ax +by for all integers x and y. So d | c for
any solutions to exist. If d c, then there are no integer solutions. If a and b
are relatively prime, then (a, b) = 1 and solutions exist for any integer c.
Consider solving ax + by = d. Because d | c, we can multiply solutions to
ax +by = d by c/d to obtain solutions of ax +by = c. To solve ax +by = d we
work backwards after using the Euclidean algorithm to compute d = (a, b).
Say the algorithm required k steps, so d = r
k1
. Working backward one step,
d = r
k1
= r
k3
q
k1
r
k2
= r
3
q
k1
(r
k4
q
k2
r
k3
)
= (1 +q
k1
q
k2
)r
3
q
k1
r
k4
.
So d = r
k1
= i r
k3
+j r
k4
where i and j are integers. Continuing, the gcd
d can be expressed as an integer combination of each pair of remainders.
26.1. LINEAR DIOPHANTINE EQUATIONS 163
Returning to the example of (77, 53), we found
77 = 1 53 + 24,
53 = 2 24 + 5,
24 = 4 5 + 4,
5 = 1 4 + 1, and
4 = 4 1 + 0.
Working from the second to the last,
1 = 5 1 4,
= 5 1 (24 5 5) = 5 5 1 24
= 5 (53 2 24) 1 24 = 5 53 11 24
= 5 53 11 (77 1 53) = 16 53 11 77.
To solve 53x +77y = 22, we start with 53 16 +77 (1) = 1. Multiplying by 22,
53 (16 22) + 77 (1 22) = 22,
and x = 352, y = 22 is one solution.
But if there is one solution, there are innitely many! Remember that d = (a, b),
so a/d and b/d are integers. Given one solution x = x
0
and y = y
0
, try
substituting x = x
0
+t (b/d) and y = y
0
t (a/d) for any integer t. Then
a(x
0
+t (b/d)) +b(x
0
t (a/d)) = ax
0
+bx
0
+t (ab/d)) +t (ba/d))
= ax
0
+bx
0
= c.
Actually, all integer solutions to ax +by = c are of the form
x = x
0
+t (b/d), and y = y
0
t (a/d),
where t is any integer, d = (a, b), and x
0
and y
0
are a solution pair.
Another example, consider solving 12x+25y = 331. First we apply the Euclidian
algorithm to compute (12, 25) = 1:
25 = 2 12 + 1, and
12 = 12 1 + 0.
Substituting back,
12 (2) + 25 1 = 1, and
12 (662) + 25 331 = 331.
164CHAPTER 26. NOTES FOR THE NINTHWEEK: AX+BY = C, FRACTIONS
So we can generate any solution to 12x + 25y = 331 with the equations
x = 662 + 25t and y = 331 12t.
Using these, we can nd a smaller solution. Try making x non-negative with
662 + 25t 0,
25t 662, thus
t > 26.
Substituting t = 27,
x = 13, and y = 7.
Interestingly enough, this must be the only non-negative solution. A larger t will
force y negative, and a smaller t forces x negative. But the solution for t = 26 is
still small,
x = 12, and y = 19.
26.1.2 The other example
Our other posed problem:
A 98 pound box contains 5 pound bags of sugar and 12 pound sacks
of oranges. How many of each are in the box?
So we need to solve 5x + 12y = 98, and start with 5x + 12y = 1.
Computing (12, 5),
12 = 2 5 + 2,
5 = 2 2 + 1, and
2 = 2 1 + 0.
So (12, 5) = 1. Because 1 | 98, there are innitely many integer solutions. We
need to nd the non-negative solutions from those.
For a base solution,
1 = 5 2 2
= 5 2 (12 2 5)
= 5 (5) + 12 (2).
So x
0
= 5 and y
0
= 2 solve 5x + 12y = 1. Multiplying by 98,
x
0
= 490 and y
0
= 196
solve 5x + 12y = 98.
26.2. INTO REAL NUMBERS 165
To nd all solutions,
x = 490 + 12t, and
y = 196 5t.
To nd non-negative solutions, rst consider how to make y positive. Here
t = 40 makes y = 4. Trying x, x = 10. So one solution is
x
+
= 10 and y
+
= 5.
With t = 39, y is negative. And with t = 41, x is negative. So this is the
only possible solution for the actual problem.
26.2 Into real numbers
Weve used real numbers without much thought. For the next week and a half,
well ll in a few details.
Rational numbers
Arithmetic and comparisons
Decimal expansion (and other bases)
Percentages
Irrational numbers
Show that non-rational real numbers exist
Square roots, cube roots, and other radicals
Computing
Floating-point arithmetic (arithmetic with restricted rationals)
This week will cover topics in rational numbers and hence fractions. For some
people, this will be old hat. For others, this is a continuing thorn in their sides.
This presentation will be a bit dierent than the texts more typical structure. I
hope that this dierence may help some who struggle with rationals and fractions
by providing reasons for the rules.
26.2.1 Operator precedence
A quick aside on the order in which operations can be applied. Working with
fractions stress operator precedence.
Operations generally dont pass through straight lines, whether horizontal for
fractions or vertical for absolute value.
166CHAPTER 26. NOTES FOR THE NINTHWEEK: AX+BY = C, FRACTIONS
Parentheses force an order. Work from the inner outwards.
The general order of precedence between operations:
1. exponents, then
2. multiplication and division (which really are the same thing), then
3. addition, subtraction, and negation (again, these are the same thing).
Within a class, operations proceed from left to right.
Go through a parenthetical clause and compute every exponent, then every
multiplication from left to right, and then every addition from left to right.
When in doubt, use parentheses when you write expressions.
26.3 Rational numbers
Rational numbers are ratios of integers. In a fraction
n
d
, n is the numerator
and d is the denominator. The rational numbers form a set,
Q = {
a
b
| a J, b J, b = 0},
where J is the set of all integers. Let R be the set of all real numbers, then
J Q R. The integer on top, a, is the numerator; the integer on the bottom,
b, is the denominator.
Note that this is a very formal construction. We just plop one integer atop
another and call it a number. Amazing that it works.
Fractions represent ratios and proportions. When you state that 1 in 10 people
are attractive to mosquitos
1
, thats a rational number. We wont reach probabil-
ity, where we learn to interpret these ratios correctly, but we will cover basic
manipulations of rational numbers.
Two fun points:
There are only as many rational numbers as there are non-negative integers
(and hence integers)! Both sets are innite, but you can construct a
mapping from each non-negative integer to and from a corresponding
fraction.
Between any two real numbers of any sort, there is a rational number. The
size of the separation does not matter! There always exists a rational num-
ber arbitrarily close to a given real. Consider taking a decimal/calculator
expansion and chopping it o once its close enough.
1
http://www.webmd.com/a-to-z-guides/features/are-you-mosquito-magnet
26.4. REVIEW OF RATIONAL ARITHMETIC 167
26.4 Review of rational arithmetic
Rational arithmetic is based on integer arithmetic. The following properties will
be inherited by multiplication and addition for rationals q, r, and s:
closure q +r Q, qr Q
commutative q +r = r +q, qr = rq,
associative q + (r +s) = (q +r) +s, q(rs) = (qr)s, and
distributive q(r +s) = qr +qs.
One homework question is to take the operation denitions below and verify
some of these properties.
The following are somewhat formal denitions to show how to construct rationals
along strict rules.
26.4.1 Multiplication and division
We start with multiplication and division. Let a, b, k, x, y J, so all the variables
are integers. We will extend these variables to run over the rational numbers
shortly.
The denition of multiplying fractions:
a
b

x
y
=
ax
by
.
As an example
3
7

5
2
=
15
14
.
The denition of a relationship between division and fractions:
a/b = a
1
b
=
a
b
so 8/2 = a
1
2
=
8
2
.
We need to be a little careful here. Integer division was dened only when b | a,
so this expression formally only holds when b | a. We relax this restriction later
to allow the variables to run over rational numbers.
An important consequence is that
a =
a
1
for all a.
This leads to very useful technique, expressing 1 as a fraction:
1 = k/k =
k
k
168CHAPTER 26. NOTES FOR THE NINTHWEEK: AX+BY = C, FRACTIONS
for any k = 0. Remember that in the divisibility form k = 1 k + 0, so k | k and
k/k = 1.
Next we show what the text calls the fundamental property of rational numbers,
which is not terribly fundamental. First we show that 1 is the multiplicative
identity for rationals by using the fact that 1 is the multiplicative identity
for integers,
a
b
1 =
a
b

1
1
=
a 1
b 1
=
a
b
.
Using this fact, we show that
a
b
=
ak
bk
for any k = 0,
a
b
=
a
b
1 =
a
b

k
k
=
ak
bk
.
Now we introduce proper fractions. A proper fraction is a rational
a
b
where
the numerator a and denominator b are relatively prime. That is gcd(a, b) = 1
and they share no common factors. Every fraction is equal to some proper
fraction. Given gcd(a, b) = d, we can factor out the common divisor,
a
b
=
a

d
b

d
=
a

d
d
=
a

.
So for 15 and 35, (15, 35) = 5 and
15
35
=
3 5
7 5
=
3
7
.
A fraction that is not proper is improper. An improper fraction is a redundant
representation, and keeping fractions improper sometimes helps speed operations.
Every rational number with a non-zero numerator has a multiplicative inverse.
This uses only the expression for 1 and the relationship between integer division
and fractions. Using that a/a = 1 and commutativity of integer multiplication,
1 = (ab)/(ab) =
ab
ab
=
ab
ba
=
a
b

b
a
.
So if a = 0, then
b
a
is the multiplicative inverse of
a
b
. As an example,
3
5

5
3
=
15
15
= 1.
With a multiplicative inverse, we can dene division of rationals analogously
to the fractional form of division of integers,
a
b
/
x
y
=
a
b

y
x
=
ay
bx
.
For example,
3
5
/
5
7
=
3
5

7
5
=
21
25
.
26.4. REVIEW OF RATIONAL ARITHMETIC 169
26.4.2 Addition and subtraction
When adding rational numbers, you must ensure both ratios have the same
denominators. This is the same as ensuring measurements are all in the same
units; both numerators need measured by the same denominator.
The denition for adding fractions:
a
b
+
x
y
=
ay
by
+
bx
by
=
ay +bx
by
.
Later we will use the least common multiple of b and y to work with a smaller
initial denominator. So
1
2
+
1
3
=
3
6
+
2
6
=
5
6
.
Rational numbers have additive identities:
a
b
+
0
b
=
a + 0
b
=
a
b
.
We prefer there to be only one additive identity. We can use the fundamental
property above to prove that all additive identities are equal to
0
1
:
0
b
=
0 b
1 b
=
0
1

b
b
=
0
1
1 =
0
1
.
Given that 1 | 0,
0
1
= 0/1 = 0. So zero is the additive identity for rationals as
well as integers.
Like integers, rationals have additive inverses:
a
b
+
a
b
=
a +a
b
=
0
b
= 0.
Given the additive inverse exists, we can dene the negation of a rational as

a
b
=
a
b
,
and then we dene subtraction in terms of addition as
a
b

x
y
=
a
b
+
x
y
=
ay bx
xy
.
So
1
2

1
3
=
3
6
+
2
6
=
1
6
.
170CHAPTER 26. NOTES FOR THE NINTHWEEK: AX+BY = C, FRACTIONS
26.4.3 Comparing fractions
We start with some high-level denitions and nd the common product rule for
comparing fractions.
First, a quick review of integer ordering. We say an integer is negative if it has
a negative sign, e.g. -1. An integer is positive if it is neither zero nor negative,
or equivalently if the integer is also a counting number. We start an ordering of
the integers by saying that a positive i > 0, a negative i < 0, and 0 = 0.
Then we can compare two integers i and j by their dierence. There are three
cases:
If i j > 0, then i > j.
If i j < 0, then i < j.
Finally, if i j = 0, then i = j.
This phrasing may help with the common confusion regarding comparisons and
multiplication by negative numbers.
Consider two integers 3 < 5. That 3 < 5 implies 3 5 < 0 (and we know it is -2).
Now multiply both sides here by -1. If 3 5 < 0, that implies 3 5 is negative,
and in turn 1 (3 5) = 5 3 = 2 is positive. Thus multiplying both sides
by -1 (and hence any negative number) requires reversing the comparison. Here
1(35) = 53 > 0. But by the distributive property, 1(35) = (3)(5)
as well, so (3) (5) > 0 and 3 > 5.
Returning to rationals, the integers are a subset, so an order on the rationals
should respect the same ordering on the integer subset.
A positive fraction is equal to some fraction where both the numerator and
denominator are positive integers. So both the numerator and denominator must
have the same sign,
3
5
, or
3
5
=
3
5

1
1
=
3
5
.
A negative fraction is equal to some fraction where the numerator is negative
and the denominator is positive. Here the signs must be opposite,
3
5
, or
3
5
=
3
5

1
1
=
3
5
.
As we saw with the additive identity, a zero fraction is equal to the integer
zero and has a zero numerator. The sign of zero does not matter in rational
arithmetic (although it may in a computers oating-point arithmetic).
Given two rationals r and q, r is strictly less than q, r < q, if q r is positive.
Thus
q
n
q
d

r
n
r
d
=
q
n
r
d
q
d
r
n
q
d
r
d
> 0.
26.5. COMPLEX FRACTIONS 171
We can always move negative signs into the numerator, so we assume that q
d
and r
d
are positive. But remember to convert the fraction into having
a positive denominator! Then the above relation
q
n
r
d
q
d
r
n
> 0 or, equivalently, q
n
r
d
> q
d
r
n
.
So
q
n
q
d
>
r
n
r
d
when q
n
r
d
> q
d
r
n
.
By symmetry, then we can compare two rational numbers by comparing appro-
priate products.
If q
n
r
d
> q
d
r
n
, then
qn
q
d
>
rn
r
d
.
If q
n
r
d
< q
d
r
n
, then
qn
q
d
<
rn
r
d
.
If q
n
r
d
= q
d
r
n
, then
qn
q
d
=
rn
r
d
.
Consider comparing
1
2
and
1
3
,
1 3 > 1 2
1
2
>
1
3
.
And for the negations
1
2
and
1
3
,
1 3 < 1 2
1
2
<
1
3
.
As an example of why you need to force the denominator to be positive, consider
1
2
and
1
3
.
1 3 < 1 2
1
2
<
1
3
.
This is because we are in essence multiplying both sides by the product of their
denominators. That product is negative, so we would have to ip the sign. Its
just as easy to remember to make the denominator positive.
26.5 Complex fractions
So far, the numerator and denominator have been integers. We can loosen
the denition slightly and allow complex fractions where the numerator and
denominator are rational numbers. We extend the division denition to map
complex fractions into fractions with integral numerators and denominators,
a
b
x
y
=
a
b
/
x
y
=
a
b

y
x
=
ay
bx
.
We could use this denition to show that all the arithmetic operations work as
expected on complex fractions.
172CHAPTER 26. NOTES FOR THE NINTHWEEK: AX+BY = C, FRACTIONS
Working with complex fractions sometimes allows adding fractions without using
a massive denominator.
Let L be the least common multiple of b and y. Then b | L and y | L, so L/b
and L/y are integers. We can manipulate the addition denition slightly by
introducing L,
a
b
+
x
y
=
a
b
1
+
x
y
1
=
a
1
b
1
+
x
1
y
1
=
L
L

_
a
1
b
1
+
x
1
y
1
_
=
a
L
b
L
+
x
L
y
L
=
a(L/b)
L
+
x(L/y)
L
=
a(L/b) +x(L/y)
L
.
With 75 = lcm(15, 25),
7
15
+
8
25
=
7 3 + 8 5
75
=
61
75
.
Quite often there is less work in reducing the result into proper form if you use
the least common multiple as the denominator.
26.6 Homework
Practice is absolutely critical in this class.
Groups are ne, turn in your own work. Homework is due in or before class on
Mondays.
Repeated, because I didnt cover the material last week: Find
two integer solutions to each of the following, or state why no solutions
exist:
64x + 336y = 32
33x 27y = 11
31x 27y = 11
Problem set 6.1 (p354)
2, 6, 11, 12
32
26.6. HOMEWORK 173
Problem set 6.2 (p375)
6, 18, 13, 25
Problem set 6.3 (p389)
2
8, 9
12
Note that you may email homework. However, I dont use Microsoft
TM
products
(e.g. Word), and software packages are notoriously nicky about translating
mathematics.
If youre typing it (which I advise just for practice in whatever tools you use),
you likely want to turn in a printout. If you do want to email your submission,
please produce a PDF or PostScript document.
174CHAPTER 26. NOTES FOR THE NINTHWEEK: AX+BY = C, FRACTIONS
Chapter 27
Solutions for ninth weeks
assignments
Also available as PDF.
Note: These are my approaches to these problems. There are many ways to
tackle each.
27.1 Diophantine equations
See the previous weeks solutions.
27.2 Problem set 6.1
Problem 2 Im not going to draw this, but it should be fairly straight-forward.
This was more an exercise in something handy for elementary classes.
Problem 6
20
60
=
1
3

30
60
=
1
2

5
7

25
100
=
1
4

25
100
=
1
4

3
12
=
1
4

2
3
175
176 CHAPTER 27. SOLUTIONS FOR NINTH WEEKS ASSIGNMENTS

3
4
(A quart is made of quarters... There are four cups in a quart.
Cooking is a source of bizarre but traditional units.)
Problem 11
4
5
=
64
65
=
24
30

6
9
=
32
33
=
2
3

7
25
=
207
2025
=
140
500

18
3
=
36
31
=
6
1
=
16
11
=
6
1
Problem 12
18
42
=
63
67
=
3
7
.
Here (18, 49) = 1 and (5, 14) = 1, so both are in lowest common terms.
That form is unique, and these fractions dier, so these fractions
cannot be equal. Following the texts method, you want to compare
218
249
=
36
98
and
75
714
=
35
98
, because lcm(49, 14) = 98.

9
25
=
209
2025
=
180
500
=
140
500
.

24
144
=
2
12
=
1
6
,
32
96
=
1
3
=
2
6
. These are not equal.
Problem 32 Yes, this is a general property. But you need to provide some
examples.
27.3 Problem set 6.2
Problem 6
5
7

10
5
= 2

20
24
=
5
6

56
65

76
60
=
19
15

100
200
=
1
2

31
100

321
450
=
107
150
Problem 18 1

1
4
1
Problem 13 3
5
2
=
15
2

2
3

3
2
= 1

3
4
2 =
3
2
27.4. PROBLEM SET 6.3 177
Problem 25 For each, you proceed by solving the row, column, or diagonal
that has only one open spot. Repeating suces to ll the squares.

1
2
1
12
5
12
1
4
1
3
5
12
1
4
7
12
1
6

8
15
1
5
4
15
1
15
1
3
3
5
2
5
7
15
2
15
27.4 Problem set 6.3
Problem 2 Reassociate to add the rst two terms with the common
denominator of 5 rst.
Commute the terms in the parenthesis and reassociate to add terms
with common denominator of 4 rst.
Use the distributive property to pull out the
2
3
, then add the eighths.
Commute and reassociate to multiply
3
4

4
3
= 1 rst.
Problem 8
2
3

4
7
+
2
3

3
7
=
2
3
(
4
7
+
3
7
) =
2
3
1 =
2
3

4
5

2
3

3
10

2
3
= (
4
5

3
10
)
2
3
=
1
2

2
3
=
1
3

4
7

3
2

4
7

6
4
=
4
7
(
3
2

6
4
) =
4
7
0 = 0
(
4
7

2
5
)/
2
7
=
42
75

7
2
=
4
5
Problem 9 adding fractions with a common denominator
multiplying fractions
distributing multiplication over addition
adding fractions with a common denominator
multiplying fractions
Problem 12 First subtraction of 32, then multiplication by the inverse of
9
5
.

Celsius -40 -20 0 10 20 45 100

Fahrenheit -40 -13 32 50 68 104 212
As seen in the table above, both agree at 40. If F = C, then
F =
5
9
(F 32),
9
5
F = F 32, 32 =
14
5
F, F =
80
7
11.
178 CHAPTER 27. SOLUTIONS FOR NINTH WEEKS ASSIGNMENTS
Chapter 28
Notes for the tenth week:
Irrationals and decimals
Notes also available as PDF.
exponents, roots, and irrationals
decimals and percentages
oating-point arithmetic
exponents, roots, and irrationals
exponents, rules, etc.
extending to negative exponents: rationals
extending to rational exponents leads to roots
roots to/from exponents
28.1 Real numbers
We wont dene the real numbers. That requires more time than we can allow
here. We will simply use the reals, denoted R, as more than the rationals.
This was the state of aairs until around 1872 when Richard Dedekind nally
discovered a way to construct real numbers formally.
So the reals t into our system of sets on the very top,
natural numbers whole numbers J Q R.
Look up the term Dedekind cut for more on actually dening real numbers.
179
180CHAPTER 28. NOTES FOR THE TENTHWEEK: IRRATIONALS ANDDECIMALS
28.2 Exponents and roots
We will cover:
denition for positive integer exponents,
rules,
zero exponents,
negative exponents, and
rational exponents and roots.
28.2.1 Positive exponents
Weve already used positive exponents when discussing the digit representation
of numbers:
10 = 10 = 10
1
100 = 10 10 = 10
2
1000 = 10 10 10 = 10
3
10000 = 10 10 10 10 = 10
4
.
.
.
.
.
.
.
.
.
In general, for any number (integer, rational, or real), the number raised to an
integer exponent is dened as:
a
1
= a,
a
2
= a a,
a
3
= a a a,
.
.
.
a
k
=
k
..
a a a . . . a .
For example,
2
3
= 8, and
_
2
3
_
2
=
4
9
.
28.2. EXPONENTS AND ROOTS 181
Negative numbers have signs that bounce around:
(5)
1
= 5,
(5)
2
= 25,
(5)
3
= 125, and
(5)
4
= 625.
With the symbolic denition, we can show other properties of exponentiation:
(ab)
3
= (ab) (ab) (ab)
= (a a a) (b b b) (by commutativity and associativity)
= a
3
b
3
.
In general,
(ab)
k
= a
k
b
k
.
For example,
1000 = 10
3
= (2 5)
3
= 2
3
5
3
= 8 125.
Or when multiplying numbers raised to powers, we have that exponents add as
in
a
k
a
m
=
k
..
a . . . a
m
..
a . . . a
=
k+m
..
a a
= a
k+m
.
For example,
10
2
10
3
= 100 1000 = 100000 = 10
5
.
And numbers raised to powers multiple times multiply exponents as in
(a
k
)
m
=
m
..
a
k
a
k
a
k
. . . a
k
= a
km
.
For example,
100
2
= (10
2
)
2
= 10
4
= 10000.
28.2.2 Zero exponent
Consider the following relationship between integer exponents and division:
a
3
= a
4
/a,
a
2
= a
3
/a, and
a
1
= a
2
/a.
182CHAPTER 28. NOTES FOR THE TENTHWEEK: IRRATIONALS ANDDECIMALS
Reasoning inductively, we suspect that
a
0
= a
1
/a = 1.
Using the rule above for adding exponents along with the additive identity
property that k + 0 = k, we can deduce that
a
k
= a
k+0
= a
k
a
0
.
So for any a = 0,
a
0
= 1 when a = 0.
Why cant we dene this for a = 0? 0
k
= 0 for any integer k > 0. So
0 = 0
k
= 0
k
0
0
does not help to dene 0
0
; were left with 0 = 0 0
0
. Because
0 x = 0 for any x, 0
0
can be anything.
Examples:
5
0
= 1
(73)
0
= 1
0
0
is undened. . .
28.2.3 Negative exponents
Continuing inductively for a = 0,
a
0
= 1,
a
1
= a
0
/a =
1
a
, and
a
2
= a
1
/a =
1
a

1
a
=
1
a
2
.
Again, we can use the fact that exponents add to derive this deductively:
1 = a
0
= a
k+k
= a
k
a
k
,
and so a
k
is the multiplicative inverse of a
k
, and we previously showed that to
be
1
a
k
. We have shown that
a
k
=
1
a
k
for all a = 0.
For example:
2
2
=
1
2
2
=
1
4
28.2. EXPONENTS AND ROOTS 183
is the inverse of
2
2
= 4.
Also,
_
2
3
_
1
=
1
2
3
=
3
2
is the multiplicative inverse of
2
3
.
And
_
2
3
_
2
=
1
(
2
3
)
2
=
1
4
9
=
9
4
is the multiplicative inverse of
_
2
3
_
2
=
2
3
.
28.2.4 Rational exponents and roots
So weve played with division and exponents. Consider now reasoning inductively
using the multiplication rule for exponents:
a
4
= (a
2
)
2
,
a
2
= (a
1
)
2
, and so
a
1
= (a
1
2
)
2
.
We call a
1
2
the square root of a and write

a.
But

a is only dened some of the time. Over integers, there clearly is no
integer b such that b
2
= 2, so

2 is not dened over the integers and fractional

exponents are not closed over integers.
Also, the product of two negative numbers is positive, and the product of two
positive numbers is positive, so there is no real number whose square is negative.
Hence for real a,

a is undened for a < 0.

184CHAPTER 28. NOTES FOR THE TENTHWEEK: IRRATIONALS ANDDECIMALS
Remember that (b)
2
= (1)
2
b
2
= b
2
, so the square root may be either positive
or negative!
2
2
= 4,
(2)
2
= 4, hence

4 = 2.
In most circumstances,

a means the positive root, often called the principal
square root. When you hit a square-root key or apply a square root in a
spreadsheet, you get the principal square root.
Other rationals provide other roots:
a
1
= (a
1
3
)
3
is the cube root,
3

a = a
1
3
.
Here, though, (a)
3
= (1)
3
a
3
= (a
3
), and there is no worry about the sign
of the cube root.
Using (a
k
)
m
= a
km
, we also have
a
2
3
=
3

a
2
= (
3

a)
2
The exponential operator can be dened on more than just the rationals, but we
wont go there. However, remember that I mentioned the rationals are dense in
the reals. There is a rational number close
28.2.5 Irrational numbers
There are more reals than rationals. This is a very non-obvious statement. To
justify it, we will
prove that

2 is not rational, and
generalize that proof to other roots.
Remember the table to show that there are as many integers as rationals? You
cannot construct one for the reals. I might show that someday. Its shockingly
simple but still a mind-bender. But for now, a few simple examples suce to
make the point.
Theorem: The number

2 is not rational.
Proof. Suppose

2 were a rational number. Then

2 =
a
b
28.3. DECIMAL EXPANSIONS AND PERCENTAGES 185
for some integers a and b. We will show that any such a and b, two must divide
both and so (a, b) 2. Previously, we explained that any fraction can be reduced
to have (a, b) = 1. Proving that (a, b) 2 shows that we cannot write

2 as a
fraction.
Now if

2 =
a
b
, then 2 =
a
2
b
2
and 2b
2
= a
2
. Because 2 | 2b
2
, we also know that
2 | a
2
. In turn, 2 | a
2
and 2 being prime imply that 2 | a and thus a = 2q for
some integer q.
With a = 2q, a
2
= 4q
2
. And with a
2
= 2b
2
, 2b
2
= 4q
2
or b
2
= 2q
2
. Now 2 | b as
well as 2 | a, so (a, b) 2.
Theorem: Suppose x and n are positive integers and that
n

x is rational. Then
n

x is an integer.
Proof. Because
n

a is rational and positive, there are positive integers a and b

such that
n

x =
a
b
.
We can assume further that the fraction is in lowest terms, so (a, b) = 1. Now
we show that b = 1.
As in the previous proof,
n

x =
a
b
implies that x b
n
= a
n
.
If b 1, there is a prime p that divides b. And as before, p | b implies
p | a, contradicting the assumption that (a, b) = 1. Thus b = 1 and
n

x is an
integer.
With decimal expansions, we will see that rational numbers have repeating
expansions. Irrational numbers have decimal expansions that never repeat.
There are some fascinating properties of the expansions
Irrational numbers come in two kinds, algebraic and transcendental. We
wont go into the dierence in detail, but numbers like

2 are algebraic, and
numbers like and e are transcendental.
28.3 Decimal expansions and percentages
Remember positional notation:
1 234 = 1 10
3
+ 2 10
2
+ 3 10
1
+ 4 10
0
.
Given negative exponents, we can expand to the right of 10
0
. General English
notation uses a decimal point to separate the integer portion of the number
from the rest.
186CHAPTER 28. NOTES FOR THE TENTHWEEK: IRRATIONALS ANDDECIMALS
So with the same notation,
1 234.567 = 1 10
3
+ 2 10
2
+ 3 10
1
+ 4 10
0
+5 10
1
+ 6 10
2
+ 7 10
3
.
Operations work in exactly the same digit-by-digit manner as before. When any
position goes over 9, a factor of 10 carries into the next higher power of 10. If
any digit becomes negative, a factor of 10 is borrowed frrom the next higher
power of 10.
Other languages use a comma to separate the integer from the rest and also use
a period to mark o powers of three on the other side, for example
1, 234.567 = 1.234, 567.
You may see this if you play with locales in various software packages. Obvi-
ously, this can lead to massive confusion among travellers. (A price of 1.234 is
not less than 2 but rather greater than 1000.)
Typical international mathematical and science publications use a period to
separate the integer and use a space to break groups of three:
1, 234.567 = 1 234.567.
28.3.1 Representing rationals with decimals
What is the part to the right of the decimal point? It often is called the
fractional part of the number, giving away that it is a representation of a
fraction.
Here we consider the decimal representation of rational numbers
1
a
for dierent
integers a. We will see that the expansions fall into two categories:
1. some terminate after a few digits, leaving the rest zero; and
2. some repeat a trailing section of digits.
For rational numbers, these are the only two possibilities.
We can nd the decimal expansions by long division.
Two simple examples that terminate:
0. 5
2 1. 0
1. 0
0. 2
5 1. 0
1. 0
Note that 2 | 10 and 5 | 10, so both expansions terminate immediately with
1
2
= .5 and
1
5
= .2.
28.3. DECIMAL EXPANSIONS AND PERCENTAGES 187
Actually, all fractions with a denominator consisting of powers of 2 and ve have
terminating expansions. For example,
1
2
2
=
1
4
= 0.25,
1
5
3
=
1
125
= 0.008, and
1
2 5
2
=
1
50
= 0.02.
What if the denominator a in
1
a
does not divide 10, or a 10? Then the expansion
does not terminate, but it does repeat. If the denominator has no factors of 2
or 5, it repeats immediately.
Examples of repeating decimal expansions:
0. 3 3 . . .
3 1. 0 0 0
. 9
. 1 0
9
1 0
0. 1 4 2 8 5 7 1 . . .
7 1. 0 0 0 0 0 0 0 0
. 7
3 0
2 8
2 0
1 4
6 0
5 6
4 0
3 5
5 0
4 9
1 0
7
3
We write these with a bar over the repeating portion, as in
1
3
= 0.3, and
1
7
= 0.142857.
We say that 0.3 has a period of 1 and 0.142857 has a period of 6.
We could write 0.2 = 0.20, but generally we say that this terminates once we
reach the repeting zeros.
If the denominator a contains factors of 2 or 5, the repeating portion occurs a
number of places after the decimal. For example, consider
1
6
=
1
23
and
1
45
=
1
59
:
188CHAPTER 28. NOTES FOR THE TENTHWEEK: IRRATIONALS ANDDECIMALS
0. 1 6 6 . . .
6 1. 0 0 0 0
. 6
4 0
3 6
4 0
0. 0 2 2 . . .
45 1. 0 0 0 0
. 9 0
1 0 0
9 0
1 0
So the decimal representations are
1
6
= 0.16, and
1
45
= 0.02.
The hard way to determine the period of a repeating fraction
Note that for all non-negative integer k,
10
k
0 (mod 2),
10
k
0 (mod 5), and
10
k
1 (mod 3).
These tell us that the expansions have periods of 0, 0, and 1.
For seven,
10
0
1 (mod 7),
10
1
3 (mod 7),
10
2
2 (mod 7),
10
3
6 (mod 7),
10
4
4 (mod 7),
10
5
5 (mod 7), and
10
6
1 (mod 7),
so the period is of length 7.
For 45,
10
0
1 (mod 45),
10
1
10 (mod 45), and
10
2
10 (mod 45).
28.3. DECIMAL EXPANSIONS AND PERCENTAGES 189
This is a little more complicated, but the pattern shows that there is one initial
digit before hitting a repeating pattern, exactly like the expansion
1
45
= 0.02.
In each case, we are looking for the order of 10 modulo the denominator. Finding
an integer with a large order modulo another integer is a building block in RSA
encryption used in SSL (the https prex in URLs).
28.3.2 The repeating decimal expansion may not be unique!
One common stumbling block for people is that the repeating decimal expansion
is not unique.
Let
n = 0.9 = 0.99999.
Then multiplying n by 10 shifts the decimal over one but does not alter the
pattern, so
10n = 9.9 = 9.99999.
Given
10n = 9.9, and
n = 0.9,
we can subtract n from the former.
9n = 9.9 0.9 = 9.
With 9n = 9, we know n = 1. Thus 1 = 0.9!
This is a consequence of sums over innite sequences, a very interesting and
useful topic for another course. But this technique is useful for proving that
rationals have repeating expansions.
28.3.3 Rationals have terminating or repeating expansions
Theorem: A decimal expansion that repeats (or terminates) represents a
rational number.
Proof. Let n be the number represented by a repeating decimal expansion.
Without loss of generality, assume that n > 0 and that the integer portion is
zero. Now let that expansion have d initial digits and then a period of length
p. Here we let a terminating decimal be represented by trailing 0 digits with a
period of 1.
190CHAPTER 28. NOTES FOR THE TENTHWEEK: IRRATIONALS ANDDECIMALS
For example, let d = 4 and p = 5. Then n looks like
n = 0 . d
1
d
2
d
3
d
4
p
1
p
2
p
3
p
4
p
5
.
Then 10
d
n leaves the repeating portion to the right of the decimal. Following
our example d = 4 and p = 5,
10
4
n = d
1
d
2
d
3
d
4
. p
1
p
2
p
3
p
4
p
5
.
Because it repeats, 10
d+p
n has the same pattern to the right of the decimal. In
our running example,
10
4+5
n = d
1
d
2
d
3
d
4
p
1
p
2
p
3
p
4
p
5
. p
1
p
2
p
3
p
4
p
5
.
So 10
d+p
n 10
d
n has zeros to the right of the decimal and is an integer k. In
our example,
k = 10
4+5
n 10
4
n = d
1
d
2
d
3
d
4
p
1
p
2
p
3
p
4
p
5
d
1
d
2
d
3
d
4
.
We assumed n > 0, so the dierence above is a positive integer. The fractional
parts cancel out.
Now n =
k
10
d+p
10
d
is one integer over another and thus is rational.
Theorem: All rational numbers have repeating or terminating decimal expan-
sions.
Proof. This is a very dierent style of proof, using what we have called the
pidgeonhole principle. Without loss of generality, assume the rational number
of interest is of the form
1
d
for some positive integer d.
At each step in long division, there are only d possible remainers. If some
remainder is 0, the expansion terminates.
If no remainder is 0, then there are only d1 possible remainders that appear. If
the expansion is taken to length d, some remainder must appear twice. Because
of the long division procedure, equal remainders leave equal sub-problems, and
thus the expansion repeats.
28.3.4 Therefore, irrationals have non-repeating expansions.
So we know that any repeating or terminating decimal expansion represents a
rational, and that all rationals have terminating or repeating decimal expansions.
Thus, we have the following:
Corollary: A number is rational if and only if it has a repeating decimal
expansion.
28.4. FIXED AND FLOATING-POINT ARITHMETIC 191
So if there is no repeating portion, the number is irrational. One example,
0.101001000100001 ,
has an increasing number of zero digits between each one digit. This number is
irrational.
Its beyond our scope to prove that is irrational, but it is. Thus the digits of
do not repeat.
28.3.5 Percentages as rationals and decimals
Percentage comes from per centile, or part per 100. So a direct numerical
equivalent to 85% is
85% =
85
100
= .85.
We can expand fractions to include decimals in the numerator and denominator.
The decimals are just rationals in another form, and we already explored complex
fractions with rational numerators and denominators.
So we can express decimal percentages,
85.75% =
85.75
100
= .8575.
Everything else just works. To convert a fraction into a percentage, there are
two routes. One is to convert the denominator into 100:
1
2
=
50
100
= 50%.
Another is to produce the decimal expansion and then multiply that by 100:
1
7
= 0.142857 = 14.2857142857%.
Converting a percentage into a proper fraction required dropping the percentage
into the numerator and then manipulating it appropriately:
85.75% =
85.75
100
=
8575
100
100
=
8575
10000
=
343
400
.
28.4 Fixed and oating-point arithmetic
So far we have considered innite expansions, ones that are not limited to a set
number of digits. Computers (and calculators) cannot store innite expansions
that do not repeat, and those that do require more overhead than they are worth.
192CHAPTER 28. NOTES FOR THE TENTHWEEK: IRRATIONALS ANDDECIMALS
Instead, computers round innite results to have at most a xed number
of signicant digits. Operations on these limited representations incur some
round-o error, leading to a tension between computing speed and the precision
of computed results. One important fact to bear in mind is that precision
does not imply accuracy. The following is a very precise but completely
in-accurate statement:
The moon is made of Camembert cheese.
First well cover dierent rounding rules from the perspective of xed-point
arithmetic, or arithmetic using a set number of digits to the right of the decimal
plce. Then well explain oating-point arithmetic where the decimal point
oats through a xed number of signicant digits.
We will not cover the errors in oating-point operations, but we will cover the
errors that come from typical binary representation of decimal data.
The points you need to take away from this are the following:
Using a limited number of digits (or bits) to represent real numbers leads
to some inherent, representationall error.
Representing every-day decimal quantities in binary also incurs some
representational error.
Despite the doom-like points above, oating-point arithmetic often provides
results that are accurate enough. We wont be able to cover why this is, but the
high-level reasons include:
using far more digits of precision than initially appear necessary, and
carrying intermediate results to even higher precision.
28.4.1 Rounding rules
Generally, computer arithmetic can be modelled as computing the exact result
and then rounding that exact result into an economical representation.
truncation or rounding to zero With this rounding method, digits beyond
the stored digits are simply dropped.
rounding half-way away from zero This is the texts method of round half
up. A number is rounded to the nearest representable number. In the
half-way case, where the digits beyond the number of digits stored are
5000 , the number is rounded upwards.
rounding half-way to even This is the preferred method for rounding in
general. A number is rounded to the nearest representable number. In
the half-way case, where the digits beyond the number of digits stored are
5000 , the number is rounded so the nal stored digit is even.
28.4. FIXED AND FLOATING-POINT ARITHMETIC 193
There are more rounding methods, but these suce for our discussion. Rounding
rules are hugely important in banking and nance, and there are quite a few
versions required by dierent regulations and laws.
Examples of each rounding method above, rounding to two places after the
decimal point:
initial number truncate round half up round to nearest even
1
3
= 0.3 0.33 0.33 0.33
1
7
= 0.142857 0.14 0.14 0.14
0.444 0.44 0.44 0.44
0.445 0.44 0.45 0.44
0.4451 0.44 0.45 0.45
0.446 0.44 0.45 0.45
0.455 0.44 0.46 0.46
Rounding error is the absolute dierence between the exact number and the
rounded, stored representation. In the table above, the rounding error in
representing
1
3
is |
1
3
0.33| = |
1
3

33
100
| = |
100
300

99
300
| =
1
300
= 0.003. Note that
here the rounding error is 1% of the exact result. That error is large because we
use only two digits.
Note that you cannot round in stages. Consider round-to-nearest-even applied
to 0.99455 and rounding to two places after the point:
Incorrect Correct
0.99455 0.99455
0.9946
0.995
1.00 0.99
28.4.2 Floating-point representation
Consider repeatedly dividing by 10 in xed-point arithmetic that carries two
digits beyond the decimal:
1 10 = 0.10,
0.1 10 = 0.01,
0.01 10 = 0.00.
So ( ( 1 10 ) 10 ) 10 evaluates to 0! This phenomenon is called underow,
where a number grows too small to be represented. A similar phenomenon,
overow, occurs when a number becomes too large to be represented. Computer
arithmetics dier on how they handle over- and underow, but generally overow
produces an symbol and underow produces 0.
Floating-point arithmetic compensates for this by carrying a xed number of
signicant digits rather than a xed number of fractional digits. The position
194CHAPTER 28. NOTES FOR THE TENTHWEEK: IRRATIONALS ANDDECIMALS
of the decimal place is carried in an explicit, integer exponent. This allows
oating-point numbers to store a wider range and actually makes analysis of the
round-o error easier.
In oating-point arithmetic,
1 10 = 1.000 10
0
,
0.1 10 = 1.000 10
1
,
0.01 101.000 10
2
,
.
.
.
This continues until we run out of representable range for the integer exponents.
We leave the details of oating-point underow for another day (if youre unlucky).
28.4.3 Binary fractional parts
Just as integers can be converted to other bases, fractional parts can be converted
as well.
Each position to the right of the point (no longer the decimal point) corresponds
to a power of the base. For binary, the typical computer representation,
1
2
= 2
1
= 0.1
2
= 0.5,
1
4
= 2
2
= 0.01
2
= 0.25,
1
8
= 2
3
= 0.001
2
= 0.125.
So a binary fractional part can be expanded with powers of two:
0.1101
2
=
1
2
1
+
1
2
2
+
0
2
3
+
1
2
4
= 0.8125.
To nd a binary expansion, we need to carry out long division in base 2. I wont
ask you to do that.
The important part to recognize is that nite decimal expansions may have
innite, repeating binary expansions! Remember that in decimal, 2 | 10 and
5 | 10, so negative powers of 2 and 5 have terminating decimal expansions. In
binary, only 2 | 2, so only powers of 2 have terminating binary expansions.
Numbers you expect to be exact are not. Consider 0.1. Its binary expansion is
0.1 = 0.00011
2
.
28.5. HOMEWORK 195
A ve-bit xed-point representation would use
0.1 0.00011
2
.
The error in representing this with a ve-digit xed-point representation is
0.00625, or over 6%.
In a ve-bit oating-point representation,
0.1 1.1001
2
2
4
.
The error here is less than 0.0024, or less than 0.24%. You can see what
oating-point gains here.
Ultimately, though, in a limited binary fractional representation, adding ten
dimes does not equal one dollar! This is why often programs slanted towards
nance (e.g. spreadsheets) use a form of decimal arithmetic. On current common
hardware, decimal arithmetic is implemented in software rather than hardware
and is orders of magnitude slower than binary arithmetic.
28.5 Homework
Practice is absolutely critical in this class.
Groups are ne, turn in your own work. Homework is due in or before class on
Mondays.
Problem set 7.1 (p421):
1, 2, 3, 4
14
25
28
Problem set 7.2 (p433):
4
5 (I wont get a chance to cover this, but scientic notation is a good
exercise in positional notation and rounding.)
9
Problem set 7.4 (p457);
2, 3, 4
18
On rounding and oating point arithmetic:
196CHAPTER 28. NOTES FOR THE TENTHWEEK: IRRATIONALS ANDDECIMALS
Round each of the following to the nearest tenth (one place after the
decimal) using round to nearest even, round to zero (trunca-
tion), and round half-up:
86.548
86.554
86.55
Compute the following quantities with a computer or a calculator.
Write what type of computer/calculator you used and the
software package if its a computer. Compute it as shown. Do
not simplify the expression before computing it, and do not re-enter
the intermediate results into the calculator or computer program. Also
compute the expressions that do not include 10
16
by hand exactly.
There should be a dierence between the exact result and the displayed
result in some of these cases. Remember to work from the innermost
parentheses outward.
(
10 times
..
0.1 + 0.1 + 0.1 + 0.1 + 0.1 + 0.1 + 0.1 + 0.1 + 0.1 + 0.1) 1
( (
10 times
..
0.1 + 0.1 + 0.1 + 0.1 + 0.1 + 0.1 + 0.1 + 0.1 + 0.1 + 0.1) 1)
10
16
, where 10
16
often is entered as 1e16. If the result overows
(signals an error) on various calculators, replace 10
16
by 10
8
in
this and later portions.
( ( (2 3) 1) 3) + 1
( ( ( (2 3) 1) 3) + 1) 10
16
( ( (6 7) 1) 7) + 1
( ( ( (6 7) 1) 7) + 1) 10
16
The object of this rst part is to demonstrate round-o error. The
rst to problems, adding 0.1 repeatedly, may see no error if the device
calculates in decimal. The latter four parts should see some error
regardless of the base used.
Now copy down the number displayed by the rst calculation in each
of the following. Re-enter it as x in the second calculation.
1 3, then 1 3 x where x is the number displayed.
If you have a calculator or program with , , then x where
x is the number displayed.
The object here is to see that the number displayed often is not the
number the computer or calculator has stored.
28.5. HOMEWORK 197
Note that you may email homework. However, I dont use Microsoft
TM
products
(e.g. Word), and software packages are notoriously nicky about translating
mathematics.
If youre typing it (which I advise just for practice in whatever tools you use),
you likely want to turn in a printout. If you do want to email your submission,
please produce a PDF or PostScript document.
198CHAPTER 28. NOTES FOR THE TENTHWEEK: IRRATIONALS ANDDECIMALS
Chapter 29
Second exam and solutions
Available as PDF.
199
200 CHAPTER 29. SECOND EXAM AND SOLUTIONS
Chapter 30
Third exam, due 1
December
Available as PDF. Remember, this is due on 1 December, 2008.
201
202 CHAPTER 30. THIRD EXAM, DUE 1 DECEMBER
Chapter 31
Third exam solutions
Available as PDF.
203
204 CHAPTER 31. THIRD EXAM SOLUTIONS
Chapter 32
Final exam
Available as PDF. As a warning, there are typos.
205
206 CHAPTER 32. FINAL EXAM
Part IV
Resources
207
Chapter 33
Math Lab
See the Math Lab Information Sheet for details.
Room 209 of the J. F. Hicks Memorial Library. Tutoring and additional material.
Run by Prof. Charlotte Ingram.
209
210 CHAPTER 33. MATH LAB
Chapter 34
On-line
As with all things, question the provenance of on-line resources before relying
upon them. This list is not comprehensive and does not provide endorsements;
this list is just a starting point.
34.1 Educational Standards
Virginia DOE Standards of Learning resources for mathematics
National Council of Teachers of Mathematics standards
34.2 General mathematics education resources
Encyclopedia:
Planet Math
Wolfram Mathworld
Texts:
Wikibooks
George Cains list of online mathematics textbooks
Alex Stef(?)s list of texts
Education and Coursework:
The Math Forum
Internet School Library Media Center Mathematics Resources K-12
211
212 CHAPTER 34. ON-LINE
Math Archives K-12 Teaching Materials
Mathematics Association of Americas professional development resources
Elementary?: Maths is Fun (British, obviously)
Late high school through graduate: Webcast@Berkeley, MIT Open Course-
Ware
Everyday items: SIAMs Math Matters! publications
others. . .
34.3 Useful software and applications
This list is for future reference. Each item has a somewhat steep learning curve
that is outside our scope. These may not be immediately useful for this course,
but they can be useful for playing with ideas quickly.
Exploratory and programming environments:
Linear algebra: Octave
Statistics: R
Geometry: Geomview
Algebra: Maxima, YACAS, others. . .
Spreadsheet: OpenOce, SIAG, others. . . Note: spreadsheets often are
made notorious for their poor quality arithmetic.