pain AWOIHL 13S US SET THEORY ETE SEYMOUR LIPSCHUTZ The best student aid available for excelling on tests and improving grades: + 530 Solved Problems with completely worked-out solutions CMe uuu es Usable with all major textbooks + Provides comprehensive drill in problem solving StaySCHAUM’S OUTLINE OF THEORY AND PROBLEMS oF SET THEORY and Related Topies SEYMOUR LIPSCHUTZ, Ph.D. Professor of Mathematics Temple University SCHAUM’S OUTLINE SERIES McGRAW-HILL PUBLISHING COMPANY New York St.Louis SanFrancisco Auckland Bogoté Caracas Hamburg Lisbon London Madrid Mexico Milan Montreal NewDelhi OklahomaCity Paris SanJuan SéoPaulo Singapore Sydney Tokyo TorontoCopyright © 1904 by MeGraw-Hil, Ine. AN Rights Reserved, Printed in the United Stater of America, No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, ‘electronic, mechanteal, photocopying, recording, or otherwise, without the prior written permission of the publisher. ISBN 07-097086-6 19 20 21 22 23 24 25 26 27 28 29 30 SH SH9876543210Preface ‘The theory of sets lies at the foundations of mathematies. Concepts in set theory, such as functions and relations, appear explicitly or implicitly in every branch of mathematics. This text is an informal, non-axiomatic treatment of the theory of sets. ‘The material is divided into three Parts, since the logical development is thereby not disturbed while the usefulness as a text and reference book on any of several levels, is increased. Part I contains an introduction to the elementary operations of sets and a detailed discussion of the concept of a function and of a relation. Part II develops the theory of cardinal and ordinal numbers in the classical approach of Cantor. It also considers partially ordered sets and the axiom of choice and its equivalents, including Zorn’s lemma. Part III treats of those topics which are usually associated with elementary set theory. Of course, the particular presentation of certain topics is influenced by the author's preferences, For example, functions are introduced before relations and are not initially defined as sets of ordered pairs. Each chapter begins with clear statements of pertinent definitions, principles and theorems together with illustrative and other descriptive material. ‘This is fol- lowed by graded sets of solved and supplementary problems. ‘The solved problems serve to illustrate and amplify the theory, bring into sharp focus those fine points without which the student continually feels himself on unsafe ground, and provide the repetition of basic principles so vital to effective learning. Numerous proofs of theorems and derivations of basic results are included among the solved problems. ‘The supplementary problems serve as a complete review of the material of each chapter. Considerably more material has been included here than can be covered in most first courses. This has been done to make the book more flexible, to provide a more useful book of reference, and to stimulate further interest in the topics. ‘The following texts are suggested references. Those by Halmos and Kamke are especially recommended as auxiliary reading for Part II. Bourbaki, N., Theorie des Ensembles, Hermann, Paris, 1958 Halmos, P. R., Naive Set Theory, Van Nostrand, 1960 Hausdorff, F., Set Theory, Chelsea, 1957 Kamke, E., Theory of Sets, Dover, 1950 Kuratowski, C., Introduction to Set Theory and Topology, Addison-Wesley, 1962 Natanson, I. P., Theory of Functions of a Real Variable, Chap. 1, 2, 14, Ungar, 1955 T wish to take this opportunity to thank many of my friends and colleagues for invaluable suggestions and critical review of the manuscript. Particular thanks are extended to the staff of the Schaum Publishing Company for their excellent cooperation. SeYMouR Liescuurz Polytechnic Institute of Brooklyn January, 1964CONTENTS Part 1 Elementary Theory of Sets Pore Chapter SETS AND SUBSETS : 1 Sets. Notation. Finite and infinite sets. Equality of seta, Nall act. Subsets Proper subsets. Comparability. Sets of sets. Universal set. Power set Disjoint sets. Venn-Euler diagrams. Line diagrams, Axiomatic: development of set theory. Chapter BASIC SET OPERATIONS a wv Set operations. Union. Intersection. Difference. Complement. Operations on comparable sets. Chapter SETS OF NUMBERS 30 Sets of numbers. Real numbers. Integers. Rational numbers. Natural num- bors. Irrational numbers. Decimals and real numbers. Inequalities, Absolute value. Intervals. Properties of intervals. Infinite intervals. Bounded and ‘unbounded sets, Chapter FUNCTIONS 45 Definition. Mappings, operators, transformations, Range. One-one functions Onto functions. Identity function. Constant functions. Produet function, Associativity of products of functions, Inverse of a function. Inverse function. ‘Theorems on the inverse function. Chapter PRODUCT SETS AND GRAPHS OF FUNCTIONS. 66 Ordered pairs. Product set. Coordinate diagrams. Graph of = function. Graphs and coordinate diagrams. Functions as sets of ordered pairs. Product sets in general Chapter RELATIONS Spee ose eer = 81 Open sentences. Relations. Solution sets and graphs of relations. Relations as sets of ordered pairs. Inverse relations. Reflexive relations. Symmetric relations. Anti-symmetric relations, Transitive relations, Equivalence rela: tions. Domain and range of a relation. ‘Relations and functions. Chapter FURTHER THEORY OF SETS . 104 Algebra of sets. Principle of duality. Indexed sets. Generalized operations. Partitions. Equivalence relations and partitions. Chapter FURTHER THEORY OF FUNCTIONS, OPERATIONS 16 Functions and diagrams. Set functions. Real-valued functions. Algebra of real-valued functions. Rule of the maximum domain. Characteristic functions. Choice functions. Operations. Commutative operations, Associative opera- tions. Distributive operations. Identity elements. Inverse elements. Oper: tions and subsets.CONTENTS Part It Cardinals, Ordinals, and Transfnite Induction Page Chapter. 9 = CARDINAL NUMBERS 134 Equivalent sets. Denumerable seta. ‘The continuum. Cardinal numbers Cardinal arithmetic. Inequalities and cardinal numbers, Cantor's theorem. Schréeder-Bernstein theorem. Continuum hypothesis. Chapter 19 = PARTIALLY AND TOTALLY ORDERED SETS 150 Partially ordered sets. Totally ordered seta, Subscte of ordered seta, Totally ‘ordered subsets. First and last elements. Maximal and minimal elements, Upper and lower bounds. Similar sets. Order types. Chapter 1] | WELL-ORDERED SETS. ORDINAL NUMBERS 166 Well-ordered sets, ‘Transfinite induction. Limit elements, Initial segments. Similarity between well-ordered set and its subset. Comparison of well- ordered sets. Ordinal numbers. Inequalities and ordinal numbers. Ordinal ‘addition. Ordinal multiplication, Structure of ordinal numbers. Auxiliary ‘construction of ordinal numbers Chapter 12. AXIOM OF CHOICE. ZORN'S LEMMA. WELL-ORDERING THEOREM feeeceeese 119 Cartesian products and’ choice functions,’ Axiom’ of ‘choice. Zorn’s iemmin. Woell-ordering theorem. Cardinal and ordinal numbers. Alephs. Chapter 13 PARADOXES IN SET THEORY aoe 185, Introduction. Set ofall sets (Cantor's paradox), Russell's paradox. Set of all ordinal numbers (Burali-Forti paradox). Set of all cardinal numbers. Family of al sets equivalent to set, Family of all sets similar to a well-ordered set. Part HI Related Topics Chapter 14 | ALGEBRA OF PROPOSITIONS 187 Statements, Conjunction. Disjunction. Negation. Polynomials and Boolean polynomials. Propositions and truth tables. Tatolo- ses and contradictions. Logical equivalence. Algebra of propositions. Logical Jmpliestion. Logically true and logically equivalent statements. Chapter 15 QUANTIFIERS ee 208, Propositional functions and truth sets, Universal quantifier. Existential qua tifer. Negation of propositions which contain quantifiers. Counter-examples. Notation. Propositional functions containing more than one variable. Chapter 16 | BOOLEAN ALGEBRA 216 Definition, Duality in a Boolean algebra. Basie theorems. Order in a Boolean algebra. ‘Switching cireuit design. ‘Chapter 17 | LOGICAL REASONING ere . 225 Arguments, Axgumenta and Venn diagrams. Arguments and propositions ‘Arguments and quantifiers, Conditional statements and variations. INDEX 232Part I — Elementary Theory of Sets |Chapter 1 Sets and Subsets SETS A fundamental concept in all branches of mathematics is that of a set. Intuitively, a set is any well-defined list, collection, or class of objects. The objects in sets, as we shalll see from our examples, can be anything: numbers, people, letters, rivers, ete. These objects are called the elements or members of the set. Although we shall study sets as abstract entities, we now list ten particular examples of sets. Example 11: ‘The numbers 1, 3, 7, and 10, Example 1.2: The solutions of the equation 2*—22—2 Example 13: ‘The vowels of the alphabet: a, ¢,j,0 and u, Example 14: The people living on the earth, Example 15: The students Tom, Diek and Harry. Example 1.6: The students who are absent from school. Example 1.7: The countries England, France and Denmark. Example 18: The capital cities of Europe. ‘The numbers 2, 4,6,8, . ‘The rivers in the United States. Notice that the sets in the odd numbered examples are defined, that is, presented, by actually listing its members; and the sets in the even numbered examples are defined by stating properties, that is, rules, which decide whether or not a particular object is a member of the set. NOTATION Sets will usually be denoted by capital letters A, B, X, Y, ‘The elements in our sets will usually be represented by lower case letters 4, by By os If we define a particular set by actually listing its members, for example, let A consist of the numbers 1,3,7 and 10, then we write A = (1,8,7,10) that is, the elements are separated by commas and enclosed in brackets {). We call this ‘the tabular form of a set. But if we define a particular set by stating properties whi its elements must satisfy, for example, let B be the set of all even numbers, then we use a letter, usually z, to represent an arbitrary element and we write B = (z | ziseven) which reads “B is the set of numbers x such that x is even”. We call this the set-builder form of a set. Notice that the vertical line “|” is read “such that”. In order to illustrate the use of the above notation, we rewrite the sets in Examples 11-110, We denote the sets by A1, As, ...,Aw respectively. 12 SETS AND SUBSETS [owap. 1 Example 21: As = (1,3, 7,10). @ | st—a2-2 = 0). {a 6,10, 0). {2 | © is a person living on the earth}. (Tom, Dick, Harry). (e | x in a student and 2 is absent from school). (England, France, Denmark}. {| 2 is a capital city and 2 is in Europe). 2, 4, 6, By «3s (2 | x is a river and 2 is in the United States}. If an object x is a member of a set A, ie, A contains z as one of its elements, then we write | which can also be read “x belongs to A” or “z is in A”. If, on the other hand, an object is not a member of a set A, i.e. A does not contain x as one of its elements, then we write Example 210: zea It is common custom in mathematics to put a vertical line “|"" or indicate the opposite or negative meaning of the symbol. Example $1: Let A = {a,e,i,0,u). Then aed, DEA, eed, eA. Example Let B= (2| 2 iseven). Then 3¢B,6eB,11¢B, 1408. '/" through a symbol to FINITE AND INFINITE SETS Sets can be finite or infinite. Intuitively, a set is finite if it consists of a specific number of different elements, i.e, if in counting the different members of the set the counting process can come to an end. Otherwise a set is infinite. In a later chapter we give a precise definition of infinite and finite sets. Let M be the set of the days of the week. ‘Then Mf is finite, Let N = (2,4,6,8,-..}. Then N is infinite Let P = {2 | 2 is a river on the earth). Although it may be the number of rivers in the world, P is still a finite set. ult to count. EQUALITY OF SETS Set A is equal to set B if they both have the same members, i.e. if every element which belongs to A also belongs to B and if every element which belongs to B also belongs to A. We denote the equality of sets A and B by A=B Let A = (1,2,8,4) and B= (2,1,4,2). Then A=, that is, (1,2,8, 4) (8,1, 4,2), since each of the elements i, 2, 3 and 4 of A belongs to # and exch of the elements 3, 1,4 and 2 of B belongs to A. Note therefore that a set does not change if ite elements are rearranged Example 52: Let C = (5,6,5,7) and D = (7,5,7,6). Then C=D, that is, (5,6,5,7) = {7, 5, 7,8), since each clement of C belongs to D and each element of D belongs to'C. ‘Note that a set does not change if its elements are repeated, Also, the set (5, 6, 7) equals C and D. Example 53: Let B = {e | 24-82 =~2), F Example 5, (2,1) and G = (1,2,2,1). Then E =6CHAP. 1} SETS AND SUBSETS 3 NULL SET It is convenient to introduce the concept of the empty set, that is, a set which contains no elements. ‘This set is sometimes called the null set. We say that such a set is void or empty, and we denote it by the symbol 9. Example 61: Let A be the set of people in the world who are older than 200 years, According: ‘to known statisties A is the null set. Example 62: Let B = (| 2*= 4, 210 odd). Then B is the empty set, SUBSETS If every element in a set A is also a member of a set B, then A is called a subset of B. More specifically, A is a subset of B if xe A implies xe B. We denote this relationship by writing Ve) which can also be read “A is contained in B”. Example 71: The set C = (1, 8,5) is a subset of D = (5,4,3,2,1), since each number 1, 3 and 5 belonging to C also belongs to D. Example 72: The set B= {2,4,6} ia a subset of F = (6,2,4), since each number 2,4 and 6 belonging to E’also belongs to F. Note, in particular, that E =F, In a similar manner it ean be shown that every set is a subset of itecf. Example 73: Let G = (2 | x is even}, ie. G= (2,4,6,. tive power of 2}, ie. let F = (2, 4,8,16, ing. With the above definition of a subset, we are able to restate the definition of the equality of two sets: Definition 1.1:] Two sets A and B are equal, i.e. A=. Tf A is a subset of B, then we can also write BOA which reads “B is a superset of A” or “B contains A”. Furthermore, we write AGB or BDA Jy and let F = (e | 2 is a posi ‘Then FCG, ie. F is contained ;, if and only if ACB and BCA. if A is not a subset of B. In conclusion, we state: Remark 1.1: The null set 9 is considered to be a subset of every set. Remark 1.2: If A is not a subset of B, that is, if A¢'B, then there is at least one element in A that is not a member of B. PROPER SUBSET Since every set A is a subset of itself, we call B a proper subset of A if, first, B is a subset of A and, secondly, if B is not equal to A. More briefly, B is a proper subset of A if BCA and B+A is denoted by BCA and “B is a proper subset of A” is denoted by BCA We will continue to use the previous notation in which we do not distinguish between a subset and a proper subset. In some books “B is a subset of A”4 SETS AND SUBSETS [CHAP.1 COMPARABILITY ‘Two sets A and B are said to be comparable if ACB or BCA that is, if one of the sets is a subset of the other set. Moreover, two sets A and B are said to be not comparable if A¢B and BEA Note that if A is not comparable to B then there is an element in A which is not in B and, also, there is an element in B which is not in A. Example 81: Let set of B. Example 82: Let R = (a, 5) and S = (b,,d). Then R and S are not comparable, @eR and ayS and ce Sande R. , b) and B= (a,b,c). Then A is comparable to B, since A is a sub- THEOREM AND PROOF In mathematics, many statements can be proven to be true by the use of previous assumptions and definitions. In fact, the essence of mathematics consists of theorems and their proofs. We now prove our first ‘Theorem 11: If A is a subset of B and B is a subset of C then A is a subset of C, that ACB and BCC implies ACC Proof. (Notice that we must show that any element in A is also an element in C.) Let x be an element of A, that is, let reA. Since A is a subset of B, x also belongs to B, that is, xeB. But, by hypothesis, BCC; hence every element of B, which includes z, is a member of C. We have shown that xe A implies reC. Accordingly, by definition, ACC. SETS OF SETS It sometimes will happen that the objects of a set are sets themselves; for example, the set of all subsets of A. In order to avoid saying “set of sets”, it is common practice to say “family of sets” or “class of sets”. Under these circumstances, and in order to avoid confusion, we sometimes will let script letters A,B, denote families, or classes, of sets since capital letters already denote their elements. Example 9.1: In geometry we usually say “a family of lines” or “a family of curves” since lines land curves are themselves sets of points Example 92: The set {((2,8), (2}, (5,6)) ia a family of sets. Its members are the sets (2, 3), (2) and (5, 6). ‘Theoretically, it is possible that a set has some members which are sets themselves and some members which are not sets, although in any application of the theory of sets this case arises infrequently. Example 93: Let A = (2, (1,3), 4, (2,5)). Then A is not a family of sets; here some ele- ments of A are sets and some are not. UNIVERSAL SET In any application of the theory of sets, all the sets under investigation will likely be subsets of a fixed set. We call this set the universal set or universe of discourse. We denote this set by U.CHAP. 1) SETS AND SUBSETS 5 Example 101: In plane geometry, the universal set consists of all the points in the plane. Example 102: In human population studies, the universal set consists of all the people in the world, POWER SET ‘The family of all the subsets of any set S is called the power set of S. We denote the Power set of S by 2 Example 111: Let M = (a, 6}. Then 2 = ((a, 5}, (a), {0}, 0 Example 112: Let T= (4,17,8). Then BF = (7, (4,72, (4,8), (7, 8}, 4, (7, @, OY If a set S is finite, say S has n elements, then the power set of S can be shown to have 2" elements. This is one reason why the class of subsets of S is called the power set of S and is denoted by 2°. DISJOINT SETS If sets A and B have no elements in common, ic. if no element of A is in B and no element of B is in A, then we say that A and B are disjoint. Example 121: Let A = (1, 3,7, 8) and B = {2,4,7,9). Then A and B are not disjoint since Tis tn both sets, ie Te and Te B. Example 122: Let A be the positive numbers and let B be the negative numbers. ‘Then A and B are disjoint since no number is both positive and negative. Example 123: Let B= (2, y,2) and F = (r,0,0). Then H and F are disjoint, VENN-EULER DIAGRAMS A simple and instructive way of illustrating the relationships between sets is in the use of the so-called Venn-Euler diagrams or, simply, Venn diagrams. Here we represent a set by a simple plane area, usually bounded by a circle. Example 131: Suppose ACB and, say, A*B. ‘Then A and B can be described by either diagram: 2 G) Example 122: Suppose A and B are not comparable. ‘Then A and B can be represented by the diagram on the right if they are disjoint, or the diagram on the left if they are not disjoint. S O6 SETS AND SUBSETS [CHAP. 1 Example 183: Let A = (a,b, e,d) and B= {o,d,e,f). Then we illustrate these sets with a ‘Venn diagram of the form A LINE DIAGRAMS Another useful and instructive way of illustrating the relationships between sets is by the use of the so-called line diagrams. If ACB, then we write B on a higher level than A and connect them by a line: B 4 If ACB and BCC, we write — diagram of A, B and C is Example M1: Lat A = (a), B= (0) and C= (0, ). Then the a A 2 Brample 12: Let X= (2), ¥ = (0,4), 2 = (eae) and W gram of X,Y, 2Z and W is z No {2,4 ). Then the line dis- AXIOMATIC DEVELOPMENT OF SET THEORY In an axiomatic development of a branch of mathematics, one begins with: (1) undefined terms (2) undefined relations (8) axioms relating the undefined terms and undefined relations. ‘Then, one develops theorems based upon the axioms and definitions. Example 151: In an axiomatic development of Plane Euclidean Geometry: (1) “points” and “lines” are undefined terms, (2) “point on a line” or, equivalently, “line contai tion, Axiom 1; Two different points are on one and only one line. Axiom 2: Two different lines eannot contain more than one point in comCHAP. 1) SETS AND SUBSETS 7 In an axiomatic development of set theory: (1) “element” and “set” are undefined terms, (2) “element belongs to a set” is the undefined relation. (3) Two of the axioms are: Axiom of Extension: Two sets A and B are equal if and only if every element in A belongs to B and every element in B belongs to A. Axiom of Specification: Let P(x) be any statement and let A be any set. Then there exists a set B = (a | acA, P(a) is true) Here, P(2) is a sentence in one variable for which P(a) is true or false for any ae A. For example, P(2) could be the sentence “z*~ 4” or “‘x is a member of the United Nations”. There are other axioms which are not listed since the axioms concern concepts which are discussed later. Furthermore, as our treatment of set theory is mainly intuitive, especially Part I, we will refrain from any further discussion of the axiomatic develop- ment of set theory. Solved Problems NOTATION 1. Rewrite the following statements using set notatio (2) 2 does not belong to A. (4) F is not a subset of G. (2) Ris a superset of S. (5) H does not include D. (8) dis a member of B. Solution: ()2¢A ROS @deB FEE (6) HAD. 2 Let A= (x|2r=6} and let b=3. Does b= A? Solution: A Is a set which consists of the single element 3, that is, A= (3). ‘The number 3 belongs to it does not equal A. ‘There is a basic difference between an element x and the set (#). 3. Let M= (r,s,t}. In other words, M consists of the elements r, s and t. State whether each of the four statements is correct or incorrect. If a statement is incorrect, tell why. (@)reM ()rcM (e) (eM (@ (ICM Soluti (@) Correct. (©) Incorrect. The symbol © must connect two sets; it indicates that one ae i Accordingly, rcM is incorrect since r is a member of M, not m subset, (©) Incorrect. The symbol e should connect an object to a sat: it indicates that the object is a member of the set. Therefore (r}e Mf is incorrect since (r} is a subset of Mf, not a member of M. (@) Correct, subset of the other.8 SETS AND SUBSETS [CHAP. 1 4. State in words and then write in tabular form: Q) A = {@| = 4) @) B= («| #-2=5) (8) C = {a | x is positive, x is negative) (4) D = (a | isa letter in the word “correet”}. Solution: (1) Ie reads “A is the act of # auch that # squared equals four”. The only numbers which when ‘squared give four are 2 and —2; hence A = (2,~2). (2) Te reads “B is the set of such that # minus 2 equals 5", ‘The only solution is 7; hence = (7). (3) It reads “C is the set of = such that 2 is positive and z is negative”. There is no number which is both positive and negative: hence C is empty, that is, C= 0. (@) It reads “D is the set of «such that x is a letter in the word ‘correct’", ‘The indicated letters are core and t; thus D = {e,0,r,¢t). 5. Write these sets in a set-builder form: (1) Let A consist of the letters a,b,c, and e. (2) Let B= (2,4,6,8,...}. (8) Let C consist of the countries in the United Nations. (A) Let D = (8). (6) Let E be the Presidents Truman, Eisenhower and Kennedy. Solution: Notice first that set-description, ie. set-builder, form of a set need not be unique, It is only necessary that any description define the aame set. We give a few of the many possible answers to this problem. (1) A = (2 | 2 appears before f in the alphabet) = {2 | 2 is one of the frst five letters in the alphabet) (2) B = (2 | 2 is even and positive} (@) C = {| «isa country, « is in the United Nations) @) D = (e|2-2=1) = | m= 8) (8) EB = (2 | z was President after Franklin D. Roosevelt} FINITE AND INFINITE SETS 6. Which sets are finite? (1) The months of the year. (4) (2 | x is even) 2) (1,2,8, ...,99, 100). (©) (12,3, ...). (3) The people living on the earth, Solution: ‘The first three sets are finite, Although physically it might be impossible to count the number cof people on the earth, the set is still finite. The last two sets are infinite. If we ever try to count ‘the even numbers we would never come to the end. EQUALITY OF SETS 7. Which of these sets are equal: (r,,4), (867.8), (4867), (87502 Solution: ‘They are all equal to each other. Note that order and repetition do not change a set.CHAP. 1) SETS AND SUBSETS 9 8. Which of these sets are equal? (1) (# | # is a letter in the word “follow”}. (2) The letters which appear in the word “wolf”. (8) (2 | 2 is a letter in the word “flow”). (4) The letters f, 1, 0 and w. Solution: If sete are written in tabular form then it is easy to decide whether or not they are equal. After writing the four sets in tabular form, we see that they are all equal to the set {f,1,0,¥). NULL SET 9. Which word is different from the others, and why: (1) empty, (2) void, (8) zero, (4) null? Solution: ‘The first, second and fourth words refer to the set which contains no lements. The word zero refers to a specific number, Hence zero is different, 10. Which of the following sets are different: 9, (0}, (}? Solution: Each is different from the others. ‘The set (0) contains one clement, the number zero. The set contains no elements; itis the null set. ‘The set (Q) also contains one element, the null set; it is a set of sets 11. Which of these sets is the null set? (1) A = (| 2 isa letter before a in the alphabet). @) C= (| z4x). (2) B= (| 2*=9 and 2r=4). () D = (@|2+8=8). Solution: (1) Since @ is the first letter of the alphabet, set A contains no elements; hence A=. (2) There is no number which satisfies both equations 24=9 and 224; hence B is also the null set. (@) We assume that every object is iteelf, 20 C is empty. In fact, some books define the null set in this way, that is, O = @lews} (4) ‘The number zero satisfies the equation #-+8 = 8, s0 D consists of the clement zero, Accordingly D is not the empty set. SUBSETS 12, Let A = {2,y,2). How many subsets does A contain, and what are they? Solution: We list all the possible subsets of A. They are: (24,2), (v2), (2,21, (za), (2), (a), (2), and the null set . There are eight subsets of A. 13, Define the following sets of figures in the Euclidean plane: Q = (x| x isa quadrilateral} H = (2 | x is a rhombus} R = (2| 2 is a rectangle) S = (2| xis a square). Decide which sets are proper subsets of the others. Solution: Since a aquare has 4 right angles it is a rectangle, since it has 4 equal sides it is a rhombus, and since it has 4 sides it is a quadrilateral. Accordingly, SCQ, SCR, SCH, that is, S is a subset of the other three. Also, since there are examples of rectangles, rhombuses and quadrilaterals which are not aquares, S is a proper tubset of the other three. In a similar manner we see that R is a proper subset of Q, and H is a proper subset of @. There are no other relations among the sets.10 SETS AND SUBSETS [CHAP. 1 JA. Does every set have a proper subset? Solution: ‘The null set @ does not have a proper subset. Every other set does have as a proper subset, Some books do not call the null set a proper subset; in such ease, sets which contain only one element would not contain a proper subset. 15. Prove: If A is a subset of the null set Q, then A=9. Solution: ‘The null set 9 is a subset of every set; in particular @CA. By hypothesis, ACO. Hence by Definition 1, A= 9. 16. How does one prove that a set A is not a subset of a set B? Prove that A = (2,8,4,5} is not a subset of B = (x | xis even). Solution: It is necessary to show that there is at Teast one element in A which and 8¢B, we see that A is not a subset of B, that is, AGB. Notice that it whether or not there are other elements in A which are not in B. rot in B. Since 304 not necessary to know I. Let V=(d), W={ed), X= (a,b,c), Y=(a,b) and Z=(a,b,d). Determine whether each of the following statements is true or false. ( ¥cx @) Wz 6) VEY () ¥cx (@) X=W @ wey (4) Z9V (6) ZX (8) Yez (10) Wo¥ Solution: (1) Since each element in ¥ is a member of X, we conclude that YCX is true. (2) ‘The only element in V isd, and is also in Ws thus W is a superset of V and hence WV is false. (9) Since ae % and a¢ W, WeZ is true (8) Zin-a superset of V since the only element in V is & member of (6) Since deV and d¢¥, VY is troe (6) Since ceX and o¢ 7, then Z is not a superset of X, ie, ZX is true, (1) Via not a subset of X since de V and dy X; hence VCX is fale. (@) Each element in ¥ is a member of Z; henee YZ ia false (9) Since «eX and ey W, X=W is false. (40) Since ce W and e¢ ¥, W is not a subset of ¥ and hence WY is false. s hence ZV is true (3,tu,v,w}, B= (u,v,w,2,42), C= (8u%2}, D = (uv), B= (8,0) 3). Let X be an unknown set. Determine which ets 4,B,C,D,E or F can equal X if we are given the following information: (1) XCA and XCB (8) XGA and X¢C (®) X4B and XcC (4) XCB and X¢C Solution: (1) The only set which is a subset of both A and B is D. Notice that C, and F are not subsets of B since ee C,E,F and ey B. (2) Set X can equal C, H or F since these are subsets of C and, as was noted previously, they are not aubsets of B. (3) Only B is not a subset of either A or C. D and A are subsets of Thos X=B. (4) Both B and D are subsets of B, and are not subsets of C. All other seta violate at least one of the conditions. Hence X=B or X=D. ind C, B and F are subsets of C.CHAP. 1) SETS AND SUBSETS 11 19. Let A be a subset of B and let B be a subset of C, that is, let ACB and BCC. Suppose aecA,beB,ceC, and suppose d¢A,e¢B,f¢C. Which statements must be true? (acl, @)beA, Bced, U)deB, (Bega, ()f¥A Solution: (1) By Theorem 1.1, A is a subset of C. Then ac implies aeC, and the statement is always true, (2) Since the element be B need not be an element in A, the statement can he false. (8) The clement eC could be an element in A; hence ey A need not be true. (4) The element d, which is not in A, need not be in B; hence the statement might not be true, (8) Since e¢# and ACB, e¢A is always true. (6) Since /¥C and ACC, [4A ie always true. LINE DIAGRAMS 20. Construct a line diagram for the sets A = (a,b,c), B Solution: Since ADB, ADC and B and C are not comparable, we write . (a,b) and C= (a,¢). 21, Construct a line diagram of the sets X= (a,b,c), ¥=(a,b} and Z= (6). Solution: Here ZCY and YCX. So we write ; ! | | > 4 We do not write for the line from Z to redundant since ZCY and YCX already implies ZCX. 22. Construct a line diagram of the sets R= (r,8,t), S=(s) and T= {s,t,u}. Solution: Here SCR and SCT. Also, R and T are not comparable, Accordingly we write a 23. Let Q, R, H and S be the sets in Problem 18. Construct a line diagram for these sets. Solution: Bice QoP and Gath wera wie ee _. We now add tothe diagram. Since SCR and SCH, we complate the diagram as follows: @ a R 1H 22 SETS AND SUBSETS (CHAP. 1 24, Construct a line diagram for the sets V,W,X,¥ and Z in Problem 17. Solution: Since VCW and VcZ, we write ™ z oa Since ¥CZ, we add ¥ to the diagram: 7 z NON, Finaly, since ¥CX we complete the diagram as follows Dl 25. Let S be any set. Construct a line diagram for the sets @, S and the universal set U. Solution: ‘Since the null eet @ subset of every set, ie. OCS, we write s o Furthermore, since the universal set U is a superset of every set including S, we complete the diagram ‘a8 follows: { 9 MISCELLANEOUS PROBLEMS 26. Consider the following five statements: (1) ACB, (2) ADB, (8) A=B, (4) A and B are disjoint, (5) A and B are not comparable. Which statement best describes each @ CD Oo & © (@) The area of B is part of the area of A; hence ADB. (0) There are points in A which are not in B, and points in B which are not in A; hence A and B are not comparable. The sets are not disjoint since there are points which belong to both sets. (c) Here the sets are disjoint since no point lies in both sets. The sets are also not comparable. (@) The area of A is part of the area of B; hence ACB. Venn diagram? 27. Consider the following line diagram of sets A,B,C and D.CHAP. 1) SETS AND SUBSETS 18 31, 82. . Let A= Write a statement that relates each pair of sets in the diagram. ‘There should be six statements. Solution: We see first that CCB, DCB and BoA we conclude that CCA and DCA. Finally, connected by an increasing path. se these sets are connected by lines. By Theorem 1.1, 1¢ sets C and D are not comparable since they are not Construct possible Venn diagrams of sets A,B,C and D which have a line diagram in Problem 27. Solution: ‘We make two possible diagrams: B “(€& . ©@ ‘The main difference in the diagrams is that the sets C and D are disjoint in the second diagram, But both have the same line diagram. ). What is meant by the symbol ((2, 8)) ? Solution: ‘We have a set which contains one clement, the set consisting of the elements 2 and 3. Notice that, {2,3} belongs to ({2,8)); it is not a subset of ((2,3)). Also, we can say that ((2,8}} is a set of sets, 2, (4,5), 4). Which statements are incorrect and why? () 45)CA 2) (45)eA (8) (4,5) CA Solution: ‘The elements of A are 2, 4 and the set (4,5). ‘Therefore (2) is correct and (1) is an incorrect statement, (3) Is a correct statement since the set consisting of the single element, (4,5), is a subset of a. Let E = (2, (4,5), 4). Which statements are incorrect and why? () 5eE @) (eB (8) (5)CE Solution: Each statement is incorrect. ‘The elements of E are 2, 4 and the set (4,5); hence (1) and (2) are Incorrect. There are eight subsets of H and {5) is not one of them; go (3) is incorrect, Find the power set 2° of the set S = (8, (1,4}}. Solution: Notive first that $ contains two elements, 3 and the set (1,4). Therefore 2° contains 2 Glements: S itself, the null set, (2) and the set which contains (1,4) alone, ie, ((1,4)). More briefly, 2 = (5, (81,9), 9) Which of the following are undefined in an axiomatic development of set theory: (1) set, (2) subset of, (3) disjoint, (4) element, (5) equals, (6) belongs to, (7) superset of. Solution: ‘The only undefined concepts in set theory are set, element, and the relation “belongs to” (, (4) and (6),4 SETS AND SUBSETS (CHAP.1 34. Prove: Let A and B be non-empty, ie. Ax and B4Q. If A and B are disjoint, then A and B are not comparable. Solution: Since A and B are non-empty, there are elements aeA and beB. Furthermore, since A and B are disjoint, af B and bg A. Hence AGB and BGA, ic. A and B are not comparable. 85. Let A and B be not comparable. Must A and B be disjoint? Solution: No, ‘The sets in the following Venn diagram are not comparable; these sets are also not disjoint. Supplementary Problems NOTATION 36. Write in set notation: (2) Ria a superset of 7 (6) does not belong to A @) 2 in a member of ¥ (©) Bis included in F (8) M in not a subset of S (7) the empty set (4) the power set of W (8) B belongs to of ‘31, State in words: (Q) A = | 2 lives in Paris} (@) © = (@ | # is older than 21 years) (2) B = (2 | « speaks Danish) (4) D = (2 | # is a citizen of France). 38, Write in tabular form: ) P= @\at-2-2=0) @) Q = (@| # isa letter in the word “follow”? @ B= @|#=9,2-3=5) @) S = @ | 2ina vowel) ©) P= (| # isa digit in the number 2524). 39, Let E = (1,0). State whether each of the following statements is correct or incorrect. @) Wek @ eB |) WCH Mock WH OcE 40, In an axiomatic development of set theory, state which of these symbols represents an undefined relation: (1) ¢, (@)e, (8) 3.CHAP. 1) SETS AND SUBSETS 15 SUBSETS 41, Let B = (01,2). Find all the subsets of B. 42 Let F = (0,(1,2)). Find all the subsets of F. 8 Let © = & | t-oe+8 = 0) D = (| zis even). Complete the following statements by inserting C, 3 or “n OF) ABO) ALC (8) Boies, A) Asse D, (6) Bees D, (6) C...D. (not comparable) between each pair M, Let A = (1,2, ...,8,9), B = (2,4,6,8), C = (1,3,5,7,9}, D = (84,8), and E = (3,5). Which sets can equal X if we are given the following information? UW X and B are disjoint (8) XCA and X¢C @ XcD and X¢B (&) Xco and X44 45, State whether each of the following statements is correct or ‘incorrect. (1) Every subset of a finite set is finite. (2) Every subset of an infinite aet is infinite, MISCELLANEOUS PROBLEMS 46, Draw a line diagram for the sets A, 2,C and D of Problem 43. 41. Draw a line diagram for the sets A,B,C,D and E of Problem 44. State whether each of these statements is correct or incorrect: 1) 44,3) = @ 4,3) @) wc 2) 0,3,1,2,8,2) €(,2,3) (8) OC{{4)} @ Mea 49, State whether each of the following sets is finite or infinite: (1) The set of lines which are parallel to the z-axis, (2) The set of letters in the English alphabet, (8) The set of numbers which are multiples of 5. (4) The set of animals living on earth, (5) The set of numbers which are roots of the equation x" -+ 422" — 172" (6) The set of circles through the origin (0,0). 2a +19 50, State whether each of the following statements is correct or incorrect. Here Sis any set that is ae (1) Sez @ Sce @) (Ser (& (SC 51, Draw a line diagram for the seta in the following Venn diagram. C16 2», 0. a, 2 re a“ 4. a. SETS AND SUBSETS Answers to Supplementary Problems () ROT, @) weY, @) MES, (4) 2, (6) 2¥A, (6) BCF, (7) O, (8) Reot. (1) A is the set of 2 such that » lives in Paris, (2) Bis the set of x auch that » speaks Danish, (8) Cis the set of x such that z is older than 21 years, (A) D is the set of z such that 2 is a citizen of France. () P= 2-2), @ @= holw), @) R=, @ S= {aeiou) 6) T= 23,4). (A) incorrect, (2) incorrect, (8) correct, (4) correct, (5) incorrect ‘The symbol ¢ represents an undefined relation, There are eight subsets: B, (0,1), {0,2}, (1,2), {0}, (1), (2), 2. ‘There are four subsets: F, (0), ((1,2)1, 0. @) >, @) >, @) c, (A) ne, (6) c, @) c. (1) C2 @) DB @) A,B,D (4) None (1) correct, (2) incorrect ZN, S/ B (1) correct, (2) correct, (8) correct, (4) incorrect, (6) correct (41) infinite, (2) finite, (8) infinite, (4) finite, (6) finite, (6) infinite (1) correct, (2) incorrect, (8) incorrect, (4) correct (CHAP.1Chapter 2 Set Operations SET OPERATIONS In arithmetic we learn to add, subtract and multiply, that is, we assign to each pair of numbers z and y a number z+ ¥ called the sum of z and y, a number x—y called the difference of 2 and y, and a number zy called the product of x and y. These assignments are called the operations of addition, subtraction and multiplication of numbers. In this chapter we define the operations of union, intersection and difference of sets, that is, we will assign new sets to pairs of sets A and B. In a later chapter we will see that these set operations behave in a manner somewhat similar to the above operations on numbers. UNION ‘The union of sets A and B is the set of all elements which belong to A or to B or to both. We denote the union of A and B by AUB which is usually read “A union B’ Example 11: In the Venn diagram in Fig. 2-1, we have shaded AUB, ie. the area of A and the area of B. AUB is shaded Fig. 24 Example 12: Let $= (a,0,¢d} and T= (f,b,4,9). ‘Then SUT = (a,b,e,4,f,9) Example 13: Let P be the set of positive real numbers and let Q be the set of negative real numbers. Then PUQ, the union of P and Q, consists of all the real numbers except ‘The union of A and B may also be defined concisely by AUB = (@ | reA or reB) Remark 2.1: It follows directly from the definition of the union of two sets that AUB and BUA are the same set, AUB = BUA Remark 22: Both A and B are always subsets of AUB, that is, AC(AUB) and BC(AUB) W18 BASIC SET OPERATIONS (cHAP.2 In some books the union of A and B is denoted by A +B and is called the set-theoretic sum of A and B or, simply, A plus B. INTERSECTION The intersection of sets A and B is the set of elements which are common to A and B, that is, those elements which belong to A and which also belong to B. We denote the intersection of A and B by ANB which is read “A intersection B”. Example 21: In the Venn diagram in Fig. 22, we have shaded ANB, the area that is common ‘to both A and B. ‘ANB i shaded Fig.2-2 Example 22: Let S = (a,6,¢,d) and T= (/,b,d,9}. Then Sot = (a) Le, the multiples of 2; and let W Example 23: Let V = (2,4,6,. ‘multiples of 3. ‘Then Vow = (6,12,18, ...} ‘The intersection of A and B may also be defined concisely by ANB = (| wed, eB) Here, the comma has the same meaning as “and”, fe. the 18,6,9, Remark 23: It follows directly from the definition of the intersection of two sets that ANB = BNA Remark 24: Each of the sets A and B contains ANB as a subset, (ANB)CA and (ANB)CB Remark 2.5: If sets A and B have no elements in common, i.e. if A and B are disjoint, then the intersection of A and B is the null set, ie. ANB = 9. In some books, especially on probability, the intersection of A and B is denoted by AB and is called the set-theoretic product of A and B or, simply, A times B. DIFFERENCE ‘The difference of sets A and B is the set of elements which belong to A but which do not belong to B. We denote the difference of A and B by A-B which is read “A difference B” or, simply, “A minus B”.CHAP. 2] BASIC SET OPERATIONS 19 Example 81: In the Venn part of B. weram in Fig. 2-3, we have shaded A ~B, the area in A which is not AWB is shaded Fig. 23 (a,b, ed) and T = (f,b,d, 9) Then S-7 = (a0) Example 33: Let R be the set of real numbers and let @ be the set of rational numbers ‘Then R—Q consists of the irrational numbers, The difference of A and B may also be defined concisely by A-B = (z| zed, aeB) Remark 2.6: Set A contains A~B as a subset, ie., (A-B)cA Remark 2.7: The sets (A — B), AMB and (B— A) are mutually disjoint, that is, the inter- section of any two is the null set. The difference of A and B is sometimes denoted by A/B or A~B. Example 82: Let S COMPLEMENT The complement of a set A is the set of elements which do not belong to A, that is, the difference of the universal set U and A. We denote the complement of A by 7G Example 41: In the Venn diagram in outside of A. Here we the rectangle. 24, we shaded the complement of A, i.e. the arca jume that the universal set U consists of the area in is shaded Fig. Example 42: Let the universal set U be the English alphabet and let T= T= et wweh Example 43: Let F = (2,4,6,...), that odd numbers. "Here we (a, be). ‘Then the even numbers. Then B” = (1, 8,5, ...), the sume that the universal get is the natural numbers, eae ‘The complement of A may also be defined concisely by Al = (@|aeU,2¢A} or, simply, a {| ee A) We state some facts about sets which follow directly from the definition of the com- plement of a set.20 BASIC SET OPERATIONS (CHAP. 2 Remark 28: The union of any set A and its complement A’ is the universal set, ie., Ava’ = 0 Furthermore, set A and its complement A’ are disjoint, i And = 9 Remark 29: The complement of the universal set U is the null set, and vice versa, = U=9 md Or=U Remark 2.10: ‘The complement of the complement of a set A is the set A itself. More briefly a Our next remark shows how the difference of two sets can be defined in terms of the complement of a set and the intersection of two sets. More specifically, we have the following basic relationship: Remark 2.11: The difference of A and B is equal to the intersection of A and the com- plement of B, that is, A~B = AnB’ ‘The proof of Remark 2.11 follows directly from definitions: A-B (| reA,aeB) = (x | eed,zeBY) = ANB’ OPERATIONS ON COMPARABLE SETS ‘The operations of union, intersection, difference and complement have simple properties when the sets under investigation are comparable. ‘The following theorems can be proved. Theorem 2.1: Let A be a subset of B. ‘Then the intersection of A and B is precisely A, ela ACB implies ANB =A ‘Theorem 2.2: Let A be a subset of B. Then the union of A and B is precisely B, that is, ACB implies AUB = B ‘Theorem 23: Let A be a subset of B. Then B’ is a subset of 4’, that is, ACB implies B’cA’ We illustrate Theorem 2.8 by the Venn diagrams in Fig. 2-5 and 2-6. Notice how the area of B’ is included in the area of A’. @s BY ia shaded A’ is shaded Fig. 25 Fig.26 Theorem 24: Let A be a subset of B. Then the union of A and (B— A) is precisely B, that is, ACB implies AU(B-A) = BCHAP. 2] BASIC SET OPERATIONS 21 Solved Problems UNION 1. In the Venn diagrams below, shade A union B, that is, AUB: @ @ Solution: The union of A and B is the set of all elements which belong to A or to # or to both, We there- fore shade the area in A and B as follows @ r E E: A @ © © @ AUB is shaded 2 Let A = (1,2, A}, B= (2,4,6,8) and C = (8,4,5,6). Find (a) AUB, (0) AUC, (©) BUG, (d) BUB. To form the union of A and B we put all the elements from A together with all the elements from B. Accordingly, AUB = (1,2,3,4,6,8) Similarly, Notice that BUB is precisely B. 3. Let A,B and C be the sets in Problem 2. Find (1) (AUB)UC, (2) AU(BUC). Solutio (1) We first find (AUB) = (1,2,8,4,6,8). Then the union of (AUB) and C is (AUBUC = {1,2,8,4,6,8,5) (2) We first find (BUC) = (2,4,6,8,8,5). Then the union of A and (BUC) is AU(BUC) = {1,2,3,4,6,8,5) Notice that (AUB)UC = AU(BUC). 4, Let X= (Tom, Dick, Harry), Y = (Tom, Mare, Eric} and Z = (Mare, Eric, Edward). Find (a) XUY, (b) YUZ, (e) XUZ. Solution: ‘To find XUY we list the names of X with the names of ¥; thus XUY = (Tom, Dick, Harry, Mare, Brie) Similarly, YuZ = (Tom, Mare, Eric, Edward) XUZ = {Tom, Dick, Harry, Mare, Eric, Edward) 5. Let A and B be two sets which are not comparable. Construct the line diagram for the sets A, B and AUB. Solution: Notice first, by Remark 2.2, that A and B are both subsets of AUB, that is, AC(AUB) and BC(AUB) Accordingly, the line diagram of A,B and AUB is AuR22, BASIC SET OPERATIONS (CHAP. 2 6. Prove Remark 2.2; A and B are subsets of AUB. BUA we neod only show that A is a subset of AUB, that is, xeA implies Let # be a member of A. Then it follows that 2 is a member of A or B, ie. xe AUB. Thus AC(AUB), 7. Prove: A= AUA. Solution: By Definition 1.1, we must show that AC(AUA) and (AUA)CA. By Remark 22, AC(AUA). Now let #e(AUA). ‘Then, by the definition of union, zeA or 2eA; thus 2 belongs to A. Hence, (AUA)CA and by Definition 14, A = (AUA). UUA = U, where U is the w By Remark 22, UC(UUA). Since every set is @ subset of the universal set, (U'UA)CU and the ‘conclusion follows from Definition 1.1 9% Prove: QUA =A. Solution: By Remark 22, AC(AUQ). Now let 2¢(4UQ); then zeA or ze. By definition of the null set, 2¢Q; hence #cA. "We have shown that 7e(AUQ) implies xe, ie (AUQ)CA. By Definition 1.1, A= QUA. 10, Prove: AUB = 9 implies A=9 and B= 9. Solution: By Remark 22, AC(AUB), that, Hence by Definition 1.1, A=9. Ina ACQ. But @ is a subset of every set; in particular, OCA. imilar manner we can show that B= 0. INTERSECTION 11, In the Venn diagrams in Problem 1, shade the intersection of A and B, that is, ANB. Solution: ‘The intersection of A and B consists of the area that is common to both A and to B. To compute ANB, we first shade A with strokes slanting upward to the right (////) and we shade B with strokes ‘slanting downward to the right (\\\\), as follows ANB is shaded Notice that ANB is empty in (¢) where A and B are disjoint.CHAP. 2] BASIC SET OPERATIONS 28 12, Let A = (1,2,3,4), B= (2,4,6,8) and C = (©) BNC, (@) BOB. Soluti "To form the intersection of A and B, we list all the elements which are common to A and B; thus ANB = (24). Similarly, ANC = (84), BOC = (4,6) and BOB ~ {2,4,6,8). Notice that BOB is, in fact, B. 3,4,5,6). Find (a) ANB, (b) ANC, Let A, B and C be the sets in Problem 12. Find (a) (ANB)NC, (b) AN(BNC). Solution: (a) ANB = {2,4}. ‘Then the intersection of (2,4) with C is (ANB)NC = (4), (®) BNC = (4,6). The intersection of this set with A is {4}, that ia, An(BNG) = (4). Notice that (AnB)NC = An(Bn0). 1, Let A and B be two sets which are not comparable. Construct the line diagram of A, B and ANB. Soluti By Remark 2.4, ANB is a subset of both A and B, that is, (AMB)CA and (ANB)CB. Accordingly, we have the following line diagram = = 3 Ane 15, Prove Remark 2.4: (ANB) is a subset of A and of B. Solution: Let 2 be any element in ANB. By definition of intersection, x belongs to both A and B; in particular, ze. We have shown that 2eAnB implies 2eA, Le (ANB)CA. Similarly, (ANB)CB. 16. Prove: ANA =A. Solution: By Remark 2.4, (AQA)CA. Let z be any clement in A; then, obviously, z belongs to both A and A, i.e, x belongs to ANA. We have proven that zeA implies 2¢ (AA), ia AC(ANA). By Definition 1.4, Ana =A. 1, Prove: UNA = A, where U is the universal set. Solution: By Remark 2.4, (UA)CA. Let « be any element in A. Since UT is the universal set, x also belongs to U. Since zeA and 2eU, by definition of intersection, 2 e (UNA). We have shown that ce implies x © (UMA), Le. we have proven that AC(UNA). By Definition 11, (UnA) = A. 18, Prove: AND =. Solution: By Remark 2.4, (ANQ)CO. But the null set is a subset of every set; in particular, 9C(A NP). ‘Therefore AND = 0. DIFFERENCE 19, Let A = (1,2,8,4), B= (2,4,6,8) and C= (8,4,5,6). Find (a) (A—B), (b) (C—A), (©) (B—C), (a) (B— A), (¢) (B—B). Solution: (@) ‘The set A—B consists of the elements in A which are not in B. Since A = (1,2,8,4) and 2408, then A~B = (1,3). () The only elements in C which are not in A are 6 and 6; hence C—A @ B-C=28) @ B-A=(6,8) () B-B=9. 5,6.24 BASIC SET OPERATIONS [CHAP. 2 20. In the Venn diagrams exhibited in Problem 1, shade A minus B, that is, A—B. Solution: Tn each case the act A—B consists of the elements in A which are not in B, that is, the area in A which does not lie in B. © @ A~B is shaded () ® Notice, as in (c), that AB =A if A and B are disjoint. Notice also, if A is a subset of B. in (&), that A~B= 9 21. Suppose sets A and B are not comparable. Construct a line diagram for the sets A, B, (A~B), (B—A), @ and the universal set U. Solution: Notice first, by Remark 2.6, that (A~B)CA and (B~A)CB. Since @ is a subset of every clement and since, by Remark 2.7, (A~B) and (B—A) are not comparable, we can frst draw an —— : va Since AD(A~B) and BDB~A, we add A and B to the diagram as follows: Since U contains every set, we complete the diagram: u Ao > AWB BA oe {If we did not inclade U or @ to the diagram, then the line diagram would not be connected 22, Prove Remark 2.6: (A~B)CA. Solution: Let z be any element of the set AB. By definition of difference, eA and 2¢ B: in particular, a belongs to A. We have shown that 2¢(A—B) implies ze A; in other words, (A ~ B)CA, 3. Prove: (A~-B)NB = @. Solution: Let x belong to (A —B)NB. By definition of intersection, xe (A ~B) and eB. But by definition of difference, se and 2¢B. Since there is no element satisfying both eB and xy, then (A= BB = 9. COMPLEMENT 2A. Let U = (1,2,3,...,8,9), A= (1,2,3,4), B= (2,4,6,8) and C = (8,4,5,6). Find (a) A’, (0) BY, (c) (ANCY, (d) (AUBY, (6) (A’Y, (1) (B= CY Solution: (a) The set A’ consiste of the elements that are in U7 but not in A. Hence A’ = (5,6,7,8,9).CHAP. 2] BASIC SET OPERATIONS 25 (©) ‘The set which eonsists of the elements that are in U but not in B is B’ (©) (Ano = {3,4} and hence (ANC) = (1,2,5,6,7,8,9). (@ (AUB) = (1,2,8,4,6,8) and hence (AUB) = {6,7,9). (©) A! = (5,6,7,8,9) and hence (AY = (1,2,3,4), ie. (A WO) (B—C) = (2,8) and hence (B~C)' = (1,3,4,5,6,7,9}. 1,8, 5, 7, 9) 25, In the Venn diagram below, shade (a) B’, (b) (AUBY’, (c) (B~ AY’, (d) A’NB’. Solution: (a) Since B, the complement of B, consists of the elements which do not belong to B, shade the area outside of B. BY is shaded (®) First shade the area AUB; then (AUR) is the area outside of (AUB). an ) AUB is shaded (auBy (©) First shade B— A; then (B—A)’ is the area outside of BA. i! } BA is shaded (B= AY is shaded (@) First shade A’, the area outside of A, with strokes that slant upward to the right (////) and shade B’ with strokes that slant downward to the right (\\\\): then A'AB’ is the cross-hatched area. AY and BY are shaded A‘0B" i shaded Notice that the area of (AUB)' is the same as the area of A’B’.26 BASIC SET OPERATIONS (CHAP. 2 26, Prove De Morgan's Theorem: (AUB) = A’NB’, Solutic Let #¢ (AUB) then # does not belong to AUB, Therefore #9A and 2B, ie, ee A’ and 20’, land by the definition of intersection, 2 belongs to A’ B". We have shown that xe(AUBY' implies es (auBy c(a’nB) Now let yeA’OB"; then y belongs to A’ and y belongs to B’. Thus y¥A and y¢B and hence ¥¢(AUB), ie. ye(AUBY. We have shown that yc(A'NB") implies ye(AUBY, ie (A'nB)) c (AUBY Hence by Definition 1.1, (4’0B") = (AUBY'. MISCELLANEOUS PROBLEMS 27. Let _U = (a,b,c,d,¢}, A = (a,b,d} and B= (b,d,e}. Find (a) AUB, (b) BNA, (0) BY, ~A, (e) A'NB, (f) AUB’, (9) A’B’, (h) BY— A’, (i) (ANBY’, (j) (AUBY. (@) The union of A and B consists of the elements in A and the elements in B, ie. AUB = (a,b,d,e) (®) ‘The intersection of A and B consists of the elements which are common to A and B, ic. AnB = (bd) (©) The complement of B consists of the letters which are in U but not in B; thus B’ = (a,¢). (@ The set B~A consists of the elements in B which are not in A, ie. B-A = (¢), (©) A= (ee) and B= (6,d,0); then A’DB ) A= (a,b,d) and BY = (ae); then AUB! (@) A’ = (ee) and BY = (a,0); then A'NB’ (h) BY-A’ = (a). (From (6), ANB () From (a), AUB 28, In the Venn diagram below shade (1) AN(BUC), (2) (ANB)U(ANC), (8) AU(BNC), (4) (AUB)N(AUC). QO) LS Solutic ey, (1) First shade A with upward slanted strokes and shade BUC with downward slanted strokes; then A0(BUC) is the cross-hatched shaded An(BuC) (2) First shade ANB with upward slanted strokes and then shade ANC with downward slanted strokes; then (AMB)U(ANC) is the total area shaded, as shown below.CHAP. 2} BASIC SET OPERATIONS ‘AB and ANC are shaded Notice that An(BUC) (3) First shade A with upwat slanted strokes and shade BOC with downward slanted strokes; then AU(BN0) is the total area shaded. (ANB)UANG) is shaded (ANB)U(ANC). c A and BOC are shaded AU(BNC) is shaded (4) Firat shade AUB with upward slanted strokes and shade AUC with downward slanted strokes; then the cross-hatched area is (AUB)N(AUC). ‘AUB and AUG are shaded Notice that AU(BN0) (AUB)n(AUC) is shaded (AUB)N(AUO). 29. Prove: BA is a subset of A’. Solution: Let © belong to BA. Then zeB and 24 A; hence = is a member of A’. Since ze B—A implies eA’, B—A is subset of A’. 30. Prove: B-A’ = BNA. Solution: BoA’ = (e| eB, 2¢A) = (@ | zeB, zed) = BNA Supplementary Problems BL. Let the universal set U = (a,d,e,drefoh and let A = (a,b,cde), B= (a,0¢,9) and C= (hef,g). Find: (@ Ave () c-B wo A-B ( A-07 (9 (Aa-By @) BNA @ B (@) Buc Cina (10) (Anayy28 BASIC SET OPERATIONS [onap. 2 Prove If ANB = Q, then ACB’. In the Venn diagrams below, shade (1) VW, (2) W", (@) W-V, () WOW, 6) VW’, () V—W () Or @ o 34, Draw a Venn diagram for three non-empty sets A, B and C so that A, B and C will have the following properties: @ AcB, CcB, Anc BR @) Acc, A#c, BNC =9 @ ACB, CEB, Anc#o @ AcE, BCC, CHB, AnC 35. Determine: () Una ao ana ( Uea (9) Ana (®) Ava (4) Qua @u (8) A’UA (1) ona. 36, Complete the following statements by inserting C, 2 or ne (non comparable) between each pair of sets. Here A and B are arbitrary sets. () A.A 8) AN BHA (6) A AWB @ ALAnB () AL. AUB © AL B-A 31. ‘The formula A~B = ANB’ can define the difference of two sets using only the operations of {intersection and complement. Find a formula that ean define the union of two sets, AUB, again only using the operations of intersection and complement. . Prove: A—B is a subset of AUB. 2. Prove Theorem 2.1: ACB implies ANB = A, Prove: Let AnB =; then Bna’ = B. Prove Theorem 2.2: ACB implies AUB = B. Prove: A’—B' = B-A. Prove Theorem 23: ACB implies B’cA’. Prove: Let ANB = Q; then AUB’ = B’. $5. Prove: (AnBy' = A’UBY 48, Prove Theorem 2.4: ACB implies AU(B—A) = B. Answers to Supplementary Problems a.) U 8) A 6) ff (1) C= {b,e,f,9) (9) (6,4, f.0} (2) {a,0,¢} 4) af) (6) {,4, f.e,9) (8) {a,¢,d} a0) 0 Proof. Let #eA. Since A and B are disjoint, 27 B; hence 2 belongs to B’. We have shown that 20d implies 4B, ie. ACB 38. (@) a) @) o Ew) w VnW is shaded WV ie shaded VW" ie shaded @ 4 © v V'UW is shaded Vi—W" ia shadedCHAP. 2] BASIC SET OPERATIONS 29 @® a @ @: Or VinW ie shaded WV is shaded Vin" ie shaded 6) “ 6) UW is shaded “ ) ® ox @ o CL) &@ BMA GA BU HA HH HP (MU WU WA UHH 3% (2 2 B>2 MC Ene Ene 31. AUB = (A'nBYy,Chapter 3 Sets of Numbers SETS OF NUMBERS Although the theory of sets is very general, important sets which we meet in elementary mathematics are sets of numbers. Of particular importance, especially in analysis, is the set of real numbers which we denote by RE In fact, we assume in this chapter, unless otherwise stated, that the set of real numbers R* is our universal set. We first review some elementary properties of real numbers before applying our elementary principles of set theory to sets of numbers. The set of real numbers and its properties is called the real number system. REAL NUMBERS, R* One of the most important properties of the real numbers is that they can be repre- sented by points on a straight line. As in Fig. 3-1, we choose a point, called the origin, to represent 0 and another point, usually to the right, to represent 1. Then there is a natural way to pair off the points on the line and the real numbers, that is, each point will represent a unique real number and each real number will be represented by a unique point. We refer to this line as the real line. Accordingly, we can use the words point and number interchangeably. Fig. 8-1 Those numbers to the right of 0, ie. on the same side as 1, are called the positive ‘numbers, and those numbers to the left of 0 are called the negative numbers. ‘The num- ber 0 itself is neither positive nor negative. INTEGERS, Z ‘The integers are those real numbers 0, 1,2, 8, . ‘We denote the integers by Z; hence we can write Za... 0,1,2, ‘The integers are also referred to as the “whole” numbers, 30CHAP. 3] SETS OF NUMBERS aL One important property of the integers is that they are “closed” under the operations of addition, multiplication and subtraction; that is, the sum, product and difference of two integers is again an integer. Notice that the quotient of two integers, e.g. 8 and 7, need not be an integer; hence the integers are not closed under the operation of division. RATIONAL NUMBERS, @ The rational numbers are those real numbers which can be expressed as the ratio of two integers. We denote the set of rational numbers by Q. Accordingly, Q@ = (| 2 = plq where peZ, geZ) Notice that each integer is also a rational number since, for example, 5 =5/1; is a subset of Q. ‘The rational numbers are closed not only under the operations of addition, multiplica- tion and subtraction but also under the operation of division (except by 0). In other words, the sum, product, difference and quotient (except by 0) of two rational numbers is again a rational number. hence Z NATURAL NUMBERS, N’ The natural numbers are the positive integers. We denote the set of natural numbers by N; hence N = (2,8...) The natural numbers were the first number system developed and were used primarily, at one time, for counting. Notice the following relationships between the above number systems: ee ee eee ‘The natural numbers are closed only under the operations of addition and multiplica- tion. The difference and quotient of two natural numbers need not be a natural number. ‘The prime numbers are those natural numbers p, excluding 1, which are only divisible by 1 and p itself. We list the first few prime numbers: 2,8, 5,7, 11, 18, 17, 19, IRRATIONAL NUMBERS, Q ‘The irrational numbers are those real numbers which are not rational, that is, the set of irrational numbers is the complement of the set of rational numbers Q in the real numbers R*; hence Q’ denotes the irrational numbers. Examples of irrational numbers are V3, x, V2, ete. LINE DIAGRAM OF THE NUMBER SYSTEMS Fig. 3-2 below is a line diagram of the various sets of numbers which we have inves- tigated. (For completeness, the diagram includes the set of complex numbers, numbers of the form a+bi where a and b are real. Notice that the set of complex numbers is a superset of the set of real numbers.)32 SETS OF NUMBERS (CHAP. 8 Complex numbers Real numbers Rational numbers rational numbers Nexis mage feo) [L_Naturad nme] Finemaes —] Fig.3-2 DECIMALS AND REAL NUMBERS Every real number can be represented by a “nonterminating decimal”. ‘The decimal representation of a rational number p/q can be found by “dividing the denominator q into the numerator p". If the indicated division terminates, as for 3/8 = 375 we write 3/8 = .875000 or 3/8 = 374999, If the indicated division of q into p does not terminate, then it is known that a block of digits will continually be repeated; for example, 2/ll = 181818... We now state the basic fact connecting decimals and real numbers. The rational numbers correspond precisely to those decimals in which a block of digits is continually repeated, and the irrational numbers correspond to the other nonterminating decimals. INEQUALITIES ‘The concept of “order” is introduced in the real number system by the The real number a is leat than the real number b, written aa which reads “b is greater than a". Furthermore, we write a=b or b if a 8 Remark 3.1: Notice that the concept of order, i.e. the relation au, {x |2=2), = (e|2<8), {z|a=4), (2 |reR*) are called infinite intervals and are also denoted by A=(1,%), B=([2,=), C=(-=,3) D We plot these infinite intervals on the real line as follows: Boame 4 ee ° ee A is shaded E is shaded BOUNDED AND UNBOUNDED SETS Let A be a set of numbers; then A is called a bounded set if A is the subset of a finite interval. An equivalent definition of boundedness is [Definition 3.1:] Set A is bounded if there exists a positive number M such that jel=M for all eA. A set is called unbounded if it is not bounded. Notice, then, that A is a subset of the finite interval [—M, M]. Example 41: Let A = (1, 1/2, 1/3, ...). ‘Then A is bounded since A is certainly a subset of the closed interval [0,1] Example 42: Let A = (2,4,6,...). ‘Then A is an unbounded set. Example 43: Let A = (7, 950, ~478, 2822, 42). ‘Then A is bounded, Remark 33: If a set A is finite then it is necessarily bounded. If a set is infinite then it can be either bounded as in Example 4.1 or unbounded as in Example 4.2.36 SETS OF NUMBERS [cHaP. 8 Solved Problems SETS OF NUMBERS In the following problems, let R*,Q,Q/,Z,N and P denote respectively the real num- bers, rational numbers, irrational numbers, integers, natural numbers and prime numbers. 1. State whether each of the following is true or false. () -TeNn (6) -6e@ (a) Ben 2) y2eq () WeP (12) V94AeQ (8) 4eZ (8) $eZ (18) -222 (4) 9eP (9) V=Be@ (1d) ste R® (5) Be eQ (10) LeR® (15) V4 R* Soluti a LN contains only the postive integers; ~1 is negative. (2) True, YZ cannot be expressed as the ratio of two integers; hence YZ is not ra @ 2Z, the set of integers, contains all the “whole” numbers, positive and nega’ (4) False, 8 divides 9:20 9 is not prime. (9) False. > is not rational and neither is 3 ) The rational numbers include the integers. Also, ~6 = (~6/1) o 11 has no divisors except 11 and 1; hence 11 is prime. (8) False. is not an integer. V=5 is not a real number; hence, in particular, itis not an irrational number, (ao) 1 is a real number. (aa) ‘VB =2 which is a positive integer. (2) 9/4 = 3/2 which is a rational number. (3) Z consists of the positive and negative “whole” numbers. aa) + ia real and 90 is =*, (15) False, Yd = 2i is not real 2. Draw a line diagram of the sets of numbers R*, N and Q. Solutio Both N and Q are subsets of R#. However, N and Q’ are not comparable, According! a 3. To which of the sets R*,0,Q’,2,N and P does each of the following numbers belong? (1) -8/4, (2) 18, (8) V7. Solute (1) 8/4 @, the rational numbers, sinee it is the ratio of the two integers —3 and 4. Also, —8/4eR# since QCR*. (2) 13eP, since the only divisors of 18 are 18 and 1. 13 alao belongs to N,Z,@ and R#, since P is a subset of each of them. (3) V=7 is not a real number; hence it does not belong to any of the given sets. ‘the Tine 4. Let B= (2,4,6,...) and F = (1,3,5,.. of (1) addition, (2) multiplication? Solution: (1) The sum of two even numbers is even; hence E ia closed under the operation of addition, The sum of two odd numbers is not odd; hence F is not closed under the operation of addition (2) The product of two even numbers is even, and the product of two odd numbers is odd; hence both E and F are closed under the operation of multiplication. Are E and F closed under the operationsCHAP. 3] SETS OF NUMBERS 387 5. Consider the sets R*,Q,Q’,Z,N and P. Which of these are ot closed under the operations of (1) addition, (2) subtraction? Solution: (2) Q' and P, For example, —V2eQ and y2eQ’, but —v2+V2 = 0¢Q% ScP and SeP, but 345 = 8¢P. v2eQ' but y2—y2 = 0¢Q' SeN and 7eN, but 3—| 4eP. (2) QY,N and P. For exampl = -4eN; TeP and 8eP, but 73 INEQUALITIES AND ABSOLUTE VALUES 6. Write the following statements in notational form: (1) ais less than b. (4) is not less than b. (2) aisnot greater than or equal tod. (6) avis greater than or equal to b. (3) ais less than or equal to b. (6) a isnot greater than b, Solution: Recall that a vertical or slant line through a symbol designates the opposite meaning of the symbol. We write: ()a Ward, 7. Insert between the following pairs of numbers the correct symbol: <,> or =. (1) 8. (8) 8... (6) 8....9 (2) -4 @) 8.00.8 (6) ~e. af Solution We write <0 if b—a is positive, a> if b—a is negative, and @=8 if b—a=0. Then (1) 8>-%, B)—4>-8, (GHD, (4) BK 3, H=9, (6) —e< eld. & Prove: Ifa8or8w and w>1, or w<5 and 1w>1.38 SETS OF NUMBERS (CHAP. 3 LL. Give the precise definition of the absolute value function, that is, define Solution: By definition, 2) = { » if2=0 m2 ite <0, 12, Suppose |z|<0. Find x. Solution: Note that the absolute value of a number is always non-negative, that is, for every number 2, |2|=0. By hypothesis, |2|~0; hence |x) 13. Evaluate: (1) B-5| (6) |-8| + 8-1) 2) |-8+5) (7) [2-5|- 4-7] @) |-3-5] (8) 18 + |-1~4|-3-|-8) (4) |-2)\~|-6| (9) ||-2|~|-6)| 6) (8—7—|-6] (10) |—|-8]] Solution: (8-5) = |-3} @ [345 (3) |-8—5| = |-8] (@) [2] ~ [6] (8) [8=1|—-|-8) = [41 —|-5) (6) |-8|+|8—1) = |-8] + 2) () 2-8] -[4~7] (8) 184 |-1-4) = 18+5-s-8=7 ) |-21-|-6)) 10) |=[-8)] = +5) 14, Rewrite so that 2 is alone between the inequality signs: (1) 8<2-4<8 (8) -9<82<12 () 8<22-5<7 @) -1<2t8<2 () -6<-22<4 () -7<-2248<5 Solution: (1) By Ps, add 4 to each side of 8<2—4< 8 toget T< 2 < 12, (2) By Ps, add ~8 to each side of ~1<243<2 toget <2 <1, (3) By Ps, multiply each side of -9 < $2 < 12 by j to get -3<2 <4. (4) By Ps, multiply each side of —6 < ~2e < 4 by —} and then reverse the inequalities to get ced, (5) Add 5 to each side of 8< 22-5 <7 toget 8<2r<12, Then multiply by } to obtain 4=1 @) -2%-5, since -2>—9, (@) Both 2° 8 and 2°=8 are correct, since 2° @) 87, since 8<7. (6) Both = 9 and S=9 are correct, since 3'=8. (6) 3-11, since 9 > —11 INTERVALS 18. Rewrite the following intervals in set-builder form: (1) M = [-8,5), (2) S= (8,8), (8) T= [0,4], (4) W = (-7,-2]. Solution Recall that the parenthesis means that the endpoint does not belong to the interval; and the bracket means that the endpoint belongs to the interval. Accordingly, (2 | -B< 2 <5) | s<2<8) fe| O=%=4) 2} 19, Plot the intervals R = (-1,2], $= (2,2), T= (0,1) and W = [1,8] on the real line. Solution: ‘To plot R, firat circle its endpoints —1 and 2: ee a ar et ‘Since the endpoint 2 belongs to R, shade the circle around 2: oa ar pe vee : et40 SETS OF NUMBERS (CHAP. 3 Similarly, a ars 3 7 S is plotted po a nares ¥ t ¢ T is plotted => a ee | >. + : ¢ W is plotted 20. Let the real number a be less than b, i.e. a-1). Sketch the sets on the real line and then write the sets in interval notation. Soluti ‘The sets are all infinite intervals. Circle the end point and draw a ray on the side of the end point, where the set lies, as follows: Sera dr eats $ ‘ A to i 3 i ? t t zB In interval. nota (1, =). Note that the sets are defined by A = (—©,8), B= [2,=), C parenthesis is used on the side with the infinity symbol. D OPERATIONS ON INTERVALS 22, Let A=[-3,1) and B = [-1,2). (1) Sketch A and B on the same real line, (2) Using (1), sketch AUB, ANB and A—B on real lines. (8) Write AUB, ANB and A~B in interval notation. Soluti (2) On a real tine, shade A with strokes slanting upward to the right (////) and shade B with strokes slanting downward to the right (\) a cies a 2 _—— A and B are shadedCHAP. 3] SETS OF NUMBERS 41 (2) -AUB consists of points in either interval, that is, points which are shad 7 : eae AUB is shaded ACB consists only of those points which are in both A and B, cs tt ANB ia shaded A~B consists of the points in A which are not in B, ie, of the points which are shaded //// but rot eroas-hatched: ie aces ° r ? er (8) The above sketches indicate that AUB = (~8,2], ANB = [-1,1), and A-B = [-8,-1). 28. Under what conditions will the union of two disjoint intervals be an interval? Solution: First, the right endpoint of one of the intervals must be the left endpoint of the other. Secondly, ‘at the common endpoint one interval must be closed and the other interval open. For example, let ‘M=(-8,4) and N= (4,7). The right endpoint of M is the left endpoint of N; and M is open at 4 land N is closed at 4. MUN = [-8,7) and MN ~ 0. 24, In each case, sketch on a real line and write the resultant set in interval notation: (a) (@| #=-1) A {2|-82) Solution: Tn each case, sketch the left set with strokes //// and the set on the right with strokes \\. @ o © ‘The intersection is the null set since there are no lies in both interval ee coer re42 SETS OF NUMBERS (CHAP. 3 BOUNDED AND UNBOUNDED SETS 25, State whether each of the following sets is bounded or unbounded. @ (| 2<3) (@)_(& | @ is a positive power of 2) ©) (,3,5,7 ...) (©) (2,5,25,...) (2) (2,8-45,0,8%) (1, ~1, 1/2, -1/2, 1/8, -1/8, 1/4, ...) Solution: (a) This set is unbounded since there are negative numbers whose absolute values are arbitrarily large. In fact, this set is an infinite interval (~=,5), which eannot be contained in a finite interval () This set, the odd numbers, is unbounded. (6) Although the numbers in this set are very large, the sot is still bounded since it is finite, We ean choose the largest number as a bound, (@) The set of positive powers of 2, which are 2,4,8,16,..., is an unbounded set. (©) Since this set consists only of the two numbers 2 and 5, it is bounded. (A) Although there are an infinite number of numbers in this set, the set is stil bounded. It certainly is contained in the interval [—1, 1} 26. If two sets W and V are bounded, what can be said about the union and intersection of these sets? Solution: Both the union and the intersection of bounded sets are bounded. 21. If two sets R and S are unbounded, what can be said about the union and intersection of these sets? Solution: ‘The union of R and S must be unbounded, but the intersection of R and $ could be either bounded or unbounded. For example if R = (—=,3) and S ), then the intersection of these infinite intervals is the finite, and therefore bounded, interval |-2,3). But if R = (8,) and S = [-2,=), then the intersection is the infinite, and therefore unbounded, interval (3, =). Supplementary Problems SETS OF NUMBERS 28, State whether each of the following is true or false: (re@ ten () TeP () Bez @)1eP (At) Be @)3eZ (a) V1eQ 6) VOEN () VBeRe — (10) Veg (12) 2eQ 29, Draw a line diagram of the sets R¥, Z, Q’ and P. 30, Consider the sets R¥,Q,Q',Z,N and P. Which of these are not closed under the operations of (2) multiplication, (2) division (except by 0)? 31. Consider the sets fe | 2= 2 meN) = (2,4,8,16,...) fe | 2=3n, ne N) = (8,6,9,12,...) © = fe | = 9m me Z) = (1-4-6 -8,0,8,6, 0) Which of these are closed under the operations of (1) addition, (2) subtraction, (2) multiplication? 82, State whether each of the following is (a) always true, (6) sometimes true, (e) never true, Here, a0 and 60. @) eZ, beQ and a—beN. (8) aeP, beP and atbeP. @) aeP, beQ! and abeQ. (8) @eN, be@ and atbeQ. @) aeN, beZ and abe Z. (1) aeZ, beQ and a/beN. @) aeN, beQ and aide. (8) aeP, beZ and bard.CHAP. 3} SETS OF NUMBERS 43 INEQUALITIES AND ABSOLUTE VALUES 88, Write the following statements in notational form: (2) = fs not greater than y. (8) is not less than y. (2) The absolute value of is less than 4, (4) is greater than or equal to t. 34, Insert between the following pairs of numbers the correct symbol: <, > or =. Here, # denotes any real number. ) 5....-8 (ay 6) &...19 ted @) fa}. =8 (8) mess al3 ©) le). @) 2. 5. Rewrite the geometric relationships between the numbers using the inequality notation. (1) a Ties to the right of &. (2) lies to the left of y. () + lies between 5 and ~8. 36, Evaluate: a) 4 [Bl-1-5) @ [s8}- 2-1] @) 4-7) [2-3] + 1-8) ©® [I-81 I-91] (3) aed] (6) -2)+ 1-8) (® [26] 1 —9]) 31. Rewrite so that = is alone between the inequality signs: @) -2-1, T= fel e<-3), 45. Sketch each of the sets in Problem 44 on a real line, 4. Let A= (-4,2), B= (-1,0, C= (-#, 1]. Find, and write in interval notation, () AUB ( a-B Avc (A~c () BUC (ay B-c @) AnB () B-A @ Ane @) 0-4 (10) Boe. (2) CB BOUNDED AND UNBOUNDED SETS 47, Write each of the following sats in tabular form and state whether it is bounded or unbounded. B= (el e=(m,neN) G = (| e= Gy nen) F = (e/2=8,neN) H = (| 2eN, 2 < 2576) 48. Are the following statements (a) always true, (6) sometimes true, (c) never true, (1) If A is finite, A is bounded. (8) If A is a subset of [~28,79], A is finite, (2) If A is infinite, A is bounded. (4) If A is a subset of [-23, 79], A is unbounded,44 SETS OF NUMBERS (CHAP. 3 Answers to Supplementary Problems 2 ORO T OE ME 6107 TOF, OF, O07, OF 097 cs Re a L ‘30. (1) Q'andP = (2) Q', 2, N and P 3.) Bande) 0) Aunt 82, (1) Sometimes true, (5) Always true, (6) Sometimes true. (7) Sometimes true. @ (4) Never true, (6) Always trae, (8) Always true. 1 ) eey @) <4 Grey Were M> @> M= M< H< H< M< wH> (1) @>bor b Re ‘The image of —3 is 9; hence we can also write /(-3) = 9, or f:—8- 9. Let f assign to each country in the world its eapital city. Here, the domain of f is the set of countries in the world; the co-domain of f is the list of eapital eities in the world. The image of France is Paris, that is, (France) = Paris. Example 13: Let A = (a,b, ¢,d) and B= (a,b,c). Define a function f of A into B by the correspondence f(a) = b, fib) = ¢, fle) =e and fd) By this definition, the Image, for example, of b ise. Example 14: Let A = (-1, 1}. Let f assign to each rational number in R¥ the number 1, and to each irrational number in R# the number —1. ‘Then f:R# > A, and f can be defined concisely by 1 if x is rational fo = {4 EE Example 15: Let A= {a,b,¢,d) and B=(z,y,2}. Let f:A~ B be defined by the diagram: =< [=> Notice that the functions in Examples 1.1 and 1.4 are defined by specific formulas. But this need not always be the case, as is indicated by the other examples. The rules of correspondence which define functions can be diagrams as in Example 1.5, can be geographical as in Example 1.2, or, when the domain is finite, the correspondence can be listed for each element in the domain as in Example 1.3. 4546 FUNCTIONS [cHAP. 4 MAPPINGS, OPERATORS, TRANSFORMATIONS If A and B are sets in general, not necessarily sets of numbers, then a function f of A into B is frequently called a mapping of A into B; and the notation wes is then read “f maps A into B”. We can also denote a mapping, or function, f of A into a ASB or by the diagram © If the domain and co-domain of a function f are both the same set, say fava then f is frequently called an operator or transformation on A. As we will see later, operators are important special cases of functions. EQUAL FUNCTIONS If f and g are functions defined on the same domain D and if f(a) = 9(a) for every aeD, then the functions f and g are equal and we write feo Example 21: Let f(z) = 2 where 2 is a real number. Let o(<) = 2* where 2 is a complex num- ber. ‘Then the function f is not equal to g since they have different domains, Example 22: Let the function / be defined by the diagram Let a function g be defined by the formula g(2) = t {1,2}. Then f=g since they both have the same domain and since f and g assign the same image to each element in the domain. Example 23: Let {:R# > R# and g:R > R2, Suppose f is defined by fle) = z* and g by lv) = 9". Then f and g are equal functions, that is, /=g. Notice that 2 and y fare merely dummy variables in the formulas defining the functions. RANGE OF A FUNCTION Let f be a mapping of A into B, that is, let f:A~>B. Each element in B need not appear as the image of an element in A. We define the range of f to consist precisely of those elements in B which appear as the image of at least one element in A. We denote the range of f:A>B by WA) Notice that f(A) is a subset of B. Let the function f:R¢ + R# be defined by the formula f(z) range of f consists of the positive real numbers and zero, Example $2: Let f:A > B be the function in Example 1.8, Then /(A) = (6, ¢. Example 81: ‘Then the ONE-ONE FUNCTIONS Let f map A into B. Then f is called a one-one function if different elements in B are assigned to different elements in A, that is, if no two different elements in A have the same image. More briefly, f:A->B is one-one if f(a) = f(a’) implies a=a’ or, equiva- lently, aa’ implies f(a) * f(a’).CHAP. 4] FUNCTIONS 41 Example 41: Let the function f:R# > R# be defined by the formula f(z) = 2". Then f is not ‘2 one-one function since /(2) = f(—2) = 4, that is, since the image of two different real numbers, 2 and —2, is the same number, 4, Let the function f:R# > R# be defined by the formula f(z) =’. Then f is a ‘one-one mapping since the cubes of two different real numbers are themselves different. Example 43: ‘The function f which assigns to each country in the world its capital city is one-one rent countries have different capitals, that is, no ity is the capital of two different countries. Example 42: ONTO FUNCTIONS Let f be a function of A into B. Then the range f(A) of the function f is a subset of B, that is, f(A)CB. If f(A)=B, that is, if every member of B appears as the image of at least one element of A, then we say “f is a function of A onto B”, or “f maps A onto B”, or “f is an onto function”. Example 5.1: Let the function f:R# > R* be defined by the formula f(s) =. Then f is not ‘an onto function since the negative numbers do not appear in the range off, that is, zo negative number is the square of a real number Example 52: Let /:A~ B be the function in Example 1.8. Notice that (A) = (b,¢). Since B= (a,b), the range of f does not equal the eodomain, Le. fis not onto, Example 53: Let /:A > B be the function in Example 1.5. Notice that, WA) = (ey) = B that is, the range of f is equal to the co-domain B. Thus f m J is an onto mapping. 8 A onto B, i.e, IDENTITY FUNCTION Let A be any set. Let the function f:A~A be defined by the formula f(z) = 2, that is, let f assign to each element in A the element itself. Then f is called the identity function or the identity transformation on A. We denote this function by 1 or by 1s. CONSTANT FUNCTIONS A function f of A into B is called a constant function if the same element beB is assigned to every element in A. In other words, f:A~B is a constant function if the range of f consists of only one element. Example 61: Let the function f be defined by the diagram: [7 ‘Then f is not a constant function since the range of f consists of both 1 and 2. Example 62: Let the function f be defined by the diagram: PS ‘Then f is a constant function since 3 is assigned to every element in A, Example 63: Let f:R# > R® be defined by the formula f(2) = tion since 6 is assigned to every element. ‘Then f is a constant func-48 FUNCTIONS [cHAP.4 PRODUCT FUNCTION Let f be a function of A into B and let g be a function of B, the co-domain of f, into C. We illustrate the functions below. Oo Let ae A; then its image f(a) is in B which is the domain of g. Accordingly, we can find the image of f(a) under the mapping g, that is, we can find g(/(a)). Thus we have a rule which assigns to each element ae A a corresponding element g(f(a)) eC. In other words, we have a function of A into C. This new function is called the product function or composition function of f and g and it is denoted by (gef) or (af) More briefly, if f:A> B and g:B- C then we define a function (gf): > C by (ae fla) = 9(f(a)) Here = is used to mean equal by definition. We can now complete our diagram: “ ease We compute (9° f): A+ C by its definition: (fla) = ofa) = oy) = (web) = gb) = oe) = 7 (Ne) = of) = ou) = Notice that the function (gf) is equivalent to “following the arrows” from A to C in the diagrams of the functions / and g. Example 72: To each real number let f assign its square, let g assign the number plus 8, Le. let (2) (72 0W2) = fo) = 16) = 25, (eA = of@) = 4) = 7 Notice that the product functions (f° g) and (g © f) are not the same function. ‘We compute a general formula for these product functions: (Fo oe) = foe) = fle+8) = (+3) = hort (wo Me) = ofa) = ete let #2) = 2%, To each real number 243. Then Remark 4.1: Let f:A>B. Then lnof =f and fola=f that is, the product of any function and the identity function is the function itself.CHAP. 4 FUNCTIONS 49 ASSOCIATIVITY OF PRODUCTS OF FUNCTIONS Let f:A>B,g:BC and h:C~D. Then, as illustrated in Figure 4-1, we can form the produet function gof:A > C, and then the function ho(gof):A > D. Similarly, as illustrated in Figure 4-2, we can form the product function hog: B+ D and then the function (hog): + D. Fig Both he(gof) and (hog)ef are functions of A into D. A basic theorem on functions states that these functions are equal. Specifically, Theorem 41: Let f:A>B,g:B+C and h:C+D. Then (hegef = hoof) In view of Theorem 4.1, we can write hegof:A~>D without any parenthesis. INVERSE OF A FUNCTION Let f be a function of A into B, and let be B. Then the inverse of b, denoted by f(b) consists of those elements in A which are mapped onto }, that is, those elements in A which have b as their image. More briefly, if /:A +B then 1) = (| ced, fle) =d} Notice that /-*(b) is always a subset of A. We read f-' as “f inverse”. Example 81: Let the function /:A ~ B be defined by the diagram ! ‘Then J-"(2) = {6,¢}, since both b and ¢ have z aa their image point. Also, £-"(y) = {a}, as only’ is mapped into y. The inverse of z, f-*(2), is the null fet D, since no element of A is mapped into Example 82: Let f:R# > R®, the real numbers, be defined by the formula f(z) = 2". Then £4) = (2,-2}, since 4 is the image of both 2 and —2 and there is no other real number whose square is four. Notice that f ®, since there is no clement in R* whose square is —3.50 FUNCTIONS [oHaP. 4 Example 83: Let f be a function of the complex numbers into the complex numbers, where f is, defined by the formula f(z) = 2. Then f-*(-8) = (V3i,—V3i}, as the square of each of these numbers is ~3. Notice that the functions in Examples 82 and 83 are different although they are defined by the same formula, We now extend the definition of the inverse of a function. Let f:A+B and let D bbe a subset of B, that is, DCB. Then the inverse of D under the mapping f, denoted by #-1(D), consists of those elements in A which are mapped onto some element in D. More briefiy, 1D) = (@| ze, fe)eD) Example 81: Lat the function fA > be defined by the diagram a 2 ‘Then {-"({r,8}) = (y), since only y is mapped into r or ¢ Also f-*((r,t}) = tz,n.2) nee each element in A haa as its image r or t. Example 82: Let [:R* > R* be defined by f(s) = 2", and let 9] = (2 | 4=2=9) ‘Then ry) = Example 93: Let f:A > B be any function. Then /-'(B) = A, its image in B. If f(A) denotes the range of the function f, then yay = Further, if be B, then £406) = FY) Here /-! has two meanings, as the inverse of an element of B and of a subset of B. the inverse INVERSE FUNCTION, Let f be a function of A into B. In general, f-'(b) could consist of more than one element or might even be the empty set @. Now if’ f:A~B is a one-one function and an onto function, then for each beB the inverse f—'(b) will consist of a single element in A. We therefore have a rule that assigns to each beB a unique element f-*(b) in A. Accordingly, f~' is a function of B into A and we can write PBoa In this situation, when f:A + B is one-one and onto, we call f~ the inverse function of /. Example 101: Let the function f:A + B be defined by the diagram A f B a a, Notice that f is one-one and onto, Therefore J, the inverse fonction, exist We describe {-":B > A by the diagram B re ACHAP. 4] FUNCTIONS . BL Notice, further, that if we send the arrows in the opposite direction in the first diagram of f we essentially have the diagram of f-'. Example 102: Let the function f:A ~ B be defined by the diagram A , B a Since f(a) = y and fle) =v, the function f is not one-one, Therefore the inverse fanction f-* does not exist, “As f(y) = (a, e), we cannot assign both @ and c to ‘the element ye B. Let /:R# > Rf, the real numbers, be defined by (x) = 2%. Notice that / is one- fone and onto, Hence f-':R# ~ R* existe, In fact we have a formula which defines the inverse function, f-!(e) = Vz. Example 10. THEOREMS ON THE INVERSE FUNCTION Let a function f:A>B have an inverse function f-!:B > A. Then we see by the diagram that we can form the product function (f-1» f) which maps A into A, and we see by the diagram that we can form the product function (f -!) which maps B into B. We now atate the basic theorems on the inverse function: ‘Theorem 4.2: Let the function /:A->B be one-one and onto; ie. the inverse function #-':B> A exists, Then the product function (ftef:Ara is the identity function on A, and the product function (fef-):B>B is the identity function on B. ‘Theorem 4.3: Let f:A~B and g:B~A. Then g is the inverse function of f, ie. 9 =f", if the product function (ge f):A-> A is the identity function on A and (f°g):B~B is the identity function on B. Both conditions are necessary in Theorem 4.8 as we shall see from Bxample 11: Let A = (z,y) and let B= (a,6,¢). Define a function f:A + B by the dia- gram (a) below }— &) (@ ) Now define a function g:B > A by the diagram (b) above We compute (9 °f):A> A, (a Mle) = ole) = 9) = = — fw) = Lf) = oe) = v ‘Therefore the product function (g © f) is the identity function on A. But g is not the inverte function of f because the product function (f © g) is not the Wentity function on B, f not being an onto function62 FUNCTIONS [oHaP. 4 Solved Problems DEFINITION OF FUNCTION L “ State whether or not each of the diagrams defines a function of A = {a,b,c} into B= (2,4,2). 4 a @ ®@ ® Solution: (1) No. There is nothing assigned to the clement be A. (2) No. Two elements, # and z, are assigned to the clement eA. In a function, only one clement can be assigned to an element in the domain. (8) Yea, It is possible in a function for the same clement in the co-domain to be assigned to more than one element in the domain. Use a formula to redefine the following functions: (1) To each real number let fi assign its cube. (2) To each real number let f: assign the number 5. (8) To each positive number let fs assign its square, and to the other real numbers let fs assign the number 4. Solution: (1) The fonction f,, which is a mapping of R# into RE, can be defined by fs(e) = (2) Since fs assigns 5 to every number we can define fs by fala) = 6. @) Since there are two different rules which are used in defining /s, we define fs as follows: fle) = fiers Which of these statements is different from the others, and why? (1) f is a function of A into B. (8) f:x > f(z) (6) f is. a mapping of A into B. Q) f:A>B () ASB Solution: (3) i different from the others. We are not told what is the domain and the eo-domain in (8), whereas in all the others we are told that A is the domain and 2 is the co-domain, Let f(x) =2* define a function on the closed interval -2<2=<8. Find (1) H(4), (2) (8), (8) (E38). Solution: a) fd) = 4 = 16. (2) (2) has no meaning, ie. is undefined, since —3 is not in the domain of the funetion. (@) fit-8) = (¢- 3 = O—6t+9. But this formula is true only when ¢—9 is in the domain, i.e. when 2 = ¢—3 <8. In other words, ¢ must satisfy 1 = ¢ = 11, Let the function /:R* > R* be defined by _. [ 1 itz is rational He) = 1-1 ite (a) Express f in words. (b) Find /(3), f(z), /(2.1813. Solution: (@) The function f assigns the number 1 to each rational number and the number —1 to each irra- tional numberCHAP. 4) FUNCTIONS 5B 1, As 2.1813, 1. Since V2 ie (6) Since 4 is a rational number, f(j) = 1. Since x is an irrational number, f(r) is a repeating decimal, which represents a rational number, (2.1813...) irrational, (V2) = —1. 6. Let the function /:R* > R* be defined by Be—1 if z>8 Ne) = jat-2 if -2=2<8 Qe +8 if n<-2 Find (a) f(2), (0) (4), (¢) f(-1), (@) f(-8). Solution: (6) Since 2 belongs to the closed interval [-2,8], we use the formula f(s) = st—2. Hence 12) = 2-2 4-2ak () Since 4 belongs to (2, =), we use the formula f(z) = 82-1, ‘Thus f(d) = 8(4)—1 = 12-1 (©) Since —1 is in the interval [2,3], we use the formula f(x) = 2"—2. Computing, f(-1) = iF-2 1-2 5-1 (@) Since ~2 is less than 2, A-8) = 2-3) +8 = 843 Notice that there is only one function / defined even though there are three formulas which are used to define 7. The reader should not confuse formulas and functions, 2), we use the formula f(z) = 2¢+3, ‘Thus =8 belongs to ( 3. 7. Let A = (a,b,c) and B = (1,0). How many different functions are there from A into B, and what are they? Solution: We list all the functions of A into B by diagrams. In each function we assign either 1 or 0, but not both, to each clement in A ‘(Ce Notice that there are eight functions bh Z h h he ¢® i Te RANGE OF A FUNCTION & Let A= (1,2,3,4,5). Define a function f:A > A by the diagram a = ‘What is the range of the function /? Solution: ‘The range consists of all the image points, Since only the numbers 2, 5 and 5 appear as image points, the range of f is the aet (2, 8,5)54 a 10. nL. FUNCTIONS (CHAP. 4 Let W = (a,b,¢,d). Let a function f of W into W be defined by f(a) = fe) =a, f(d) Find the range of the function f: W > W. Solu ‘The range of f consists of those elements which appear as image points. Only a and ¢ appear fs the image points of elements in W. Accordingly, the range of f is (a,¢). Let V = {-2,-1,0,1,2}. Let the function g:V>R* be defined by the formula (2) = +1 Find the range of g. Solution: ‘We compute the image of each element of V. o-2) = (OFT = 4t1 = 5 ot) = (Ht = ati =e a0) = OF+1 = 041 =1 ot) = @'t1=14152 02) =@PF1=441=5 ‘Thus the range of g is the set of image points (5,2, 1, 2,5), i.e. the set (5,2, 1). Each of the following formulas defines a function from R* into R#. Find the range of each function, (1) fz) = 2, @) o(z) = sinz, (8) Mx) = a +1 Solution: () Every real number @ has a real cube root Ya: hence Wa) = Way = In other words, the range of f is the entire set of real numbers. (2) ‘The sine of any real number will lie in the closed interval {—1,1]. Also, all the numbers in this interval will be the aine of some real number. Consequently, the range of g is the interval (1,1) (@) If we add 1 to the aquare of each real number, we obtain the set of numbers which are greater than or equal to 1. In other words, the range of A is the infinite interval (1, »). EQUAL FUNCTIONS com Let the funetions fs, fafa, fe of R* into R* be defined by @ A@) = (©) fe) = # (0) fe(v) = y? — (@)_ fe assigns to each real number its square Which of these functions are equal? Solution: ‘They are all equal to each other. ‘The letters are merely dummy variables. Each function assigns ‘the same number to every real number. . Let the functions f,g and h be defined by (a) f(z) = 2* where 0=2=1 (0) g(y) = y? where 2=y=8 (©) h(z) = 2* where ze R* Which of these functions are equal? Solution: None of the functions are equal. Although the rules of correspondence are the same, the domains are different. Thus the functions are different.CHAP. 4] FUNCTIONS 55 ONE-ONE FUNCTIONS 14, Let A = {a,b,¢,d,¢}, and let B be the set of letters in the alphabet, Let the functions f, 9 and h of A into B be defined by: () fla) =r, f(b) =a, fle) =s, fad) =r, fle) (2) g(a) =a, o(b) = 6, g(e)=«, g(a) =", 9(6) (8) h(a) =z, h(b) =, le) =a, h(d) =, hie) State whether or not each of these functions is one-one. Solution: Note that in order for a function to be one-one, it must assign different image points to different elements in the domain, (1) fis not @ one-one fun« @) 9 is a one-one function. (8) his not a one-one function since Ka) = A. since f assigns r to both @ and d, ie. fla) = fd) 15. State whether or not each of the following functions is one-one. (1) To each person on the earth assign the number which corresponds to his age. (2) To each country in the world assign the number of people which live in the country. (8) To each book written by only one author assign the author. (4) To each country in the world which has a prime minister assign its prime minister. Solution: (1) Many people in the world have the same age: hence this function is not one-one. @) Although itis possible for two countries to have the same number of people, statistics show that, today this is not so; hence this function is one-one. (8) It is possible for two different books to have the same author; hence this function is not one-one. (#) No two different countries in the world have the same prime minister; thus this function is one-one. 16. Let A = [-1,1] = (2 | -1 1), B= [1,8] and C = [-3,-1]. Let the functions futA> Re, fa:B> R? and fo:C> R® be defined by the rule: ‘To each number assign its square. Which of the functions are one-one? Solution: ‘The function fi:A~>R# is not one-one since J) bers in the domain are assigned the same image. The function f::B+# is one-one since the squares of unequal positive numbers are themselves unequal. ‘Also, 4:0 R# is one-one since the squares of unequal negative-numbers are unequal. Notice, once again, that a formula itself does not define « function. In fact, we have just seen that the same formula gives rise to different functions which have different properties. JCP, that is, since two different: num: 17. Find the “largest” interval D on which the formula f(z) = 2* defines a one-one function. Solution: ‘As long as the interval D contains either positive or negative numbers, but not both, the function will be one-one. ‘Thus D ean be the Intervals [0,=) or (2,0). There ean be other infinite intervals on which f will be one-one, but they will be subsets of one of these two. 18 Can a constant function be one-one? Solution: If the domain of a function contains a single element, the function will be a constant function ‘and it will also be one-one.56 FUNCTIONS (CHAP. 4 19. On which sets A will the identity function 14:4 >A be one-one? Solution: ‘A can be any set. The identity function is always one-one 20. In Problem 7 we listed all possible functions of A = (a,b,c) into B of the functions are one-one? Solution: None of the functions is one-one. In each function, at least two elements have the same image. (1,0). Which ONTO FUNCTIONS 21 Let f:A>B. Find f(A), Solution: If f is an onto function then every element in the co-domain of f is in the range; hence f(A) =B. the range of f, if 7 an onto function. 22. Is the funetion f:A->A in Problem 8 onto? Solution: ‘The numbers 1 and 4 in the co-domain are not the images of any elements in the domain; hence f is not an onto function. In other words, (A) = (2,8,5) is a proper subset of A. 23, Let A =[-1,1]. Let funetions j, g and h of A into A be defined by: () f@)=2, 2) o@)=2, (8) A(z) = sinz Which function, if any, is onto? Solution: (0) No negative numbers appear in the range of /; hence f is not an onto fun (2) ‘The function g is onto, that is, 9(A) (2) The fonction h is nat onto, In fact, there is no number « in A such that sin 2 24, Can a constant funetion be an onto function? Solution: If the co-domain of a function f consists of a single element, then f is always a constant function and is onto. 25, On which sets A will the identity function 14:A-> A be onto? Solution: ‘The identity function is always onto; hence A can be any set. 26, In Problem 7 all possible functions of A = {a,b,c} into B= (1,0) are listed. Which of these, if any, are onto functions? Solutio ‘All the functions are onto except fi and fx. PRODUCT FUNCTIONS 27. Let the funetions f:A>B and g:B-C be defined by the diagram A f B ’ c etCHAP. 4] FUNCTIONS 87 (@) Find the product function (go f):A > C. (0) Find the ranges of f, g and gof. Solution: (@) We use the definition of the product function and compute: Na) = ola) = ow) = ¢ We NW) = fd) = oz) = 6 WN) = ae) = ow) = ¢ Notice that we arrive at the same answer if we “follow the arrows”: atyre boase aged wud (0) By the diagram, the range of f is {2,y}, and the range of g is (r,#,t). By (a), the range of °F is (a,t). Notice that the ranges of g and g°/ are different. 28, Let A = (1,2,3,4,5) and let the functions f:A+A and g:A~A be defined by: 40) = 8 12) = 5, 18) = 38, f4)=1, £5) =2 a1) = 4, 92) =1, 9(8)=1, 94) = 2, (5) Find the composition functions fg and g of. Solation: ‘We use the definition of the product function and compute: Goo) = fot) = 14) = 1 G2 0N2) = f@) = 7) = 3 ov) = fata) = fa) = 8 G20N8) = fad) = f2) = 6 (Fox) = fo) = 18) = 3 Also, (ea) = aay = of) = 1 (eA) = of@) = 96) = 8 (2A) = a) = 9) = 1 (@°N1) = ava) = 00) = 4 @ 116) = a6) = 92) = 1 Notice that the functions f and 9 °f are not equal. 29. Let the functions {:R* + R* and g:R* + R* be defined by fia) = 22+1, g(z) = at-2 Find formulas which define the product functions go f and fog. Solution: We first compute g°/:R# - R¢. Essentially we want to substitute the formula for f inside the formula for g. We use the definition of the product function as follows: (9° Ale) = ole) = gl@e+1) = @e+1—2 = det 4 4e—1 Perhaps the reader is more familiar with the process if the functions are defined as: v= fe) = 2t1, 2 = oy) = v-2 ‘Then y is eliminated from the two formul 2S 2 = Beata = bette ‘The reader should become familiar with the first method. It is necessary to realize that 2 ‘4 dummy variable, We now compute fog:R# > B®. Weave) = flo(e) = flet—2) = 22-2) 41 = Bt—358 FUNCTIONS (CHAP. 4 30. Let the functions f and g on the real numbers R* be defined by f(z) = a+ 22-8, — g(z) = 82-4 (1) Find formulas which define the product functions gof and fog. (2) Check the formulas by showing (gof)(2) = 9(f(2)) and (feg)(2) Solution: a) Went f(9(2)). U2) = (e+ 22-3) = Bet+22—3)— Ge oN2) = gle) = 182-4) = Be—4" + 2182~ 4) — @) Ne) = 92+ 6@)-13 = 2+12-19 = 1 U2) = g(2*+22)—3) = g() = 318) —4 (Fo9)@) = se" — 182) +5 = 36-3645 M92) = §(8(2)—4) = f2) = B+ 2@)-3 = 5 ae toe — 13 Qe — Ike +8 31 Prove: If is onto. Solution: Tet ¢ be any clement in C. Since g is onto, there exists an clement be B such that g(6)=e. Also, since f is onto, there existe an element ae such that f(e)=b. Now (9°f\(a) = g(f(a)) = o(d) ‘Thus for any ceC, we have shown there is at least one clement aeA such that (g°/)(a) ‘Therefore g°/ is an onto function. ~B is onto and g:B~C is onto, then the product function (gof):A + C 32, Prove Theorem 4.1: Let f:A> B, g:B~C and h:C> D; then (hoghef = ho(gef) Solution: ‘The two functions are equal if they assign the same image to each element in the domain, that is, if (h2 9) Mle) = (he (ge Ne) for every © A. Computing, (ho 9) 2 Ne) = (he fle) = Moe) (hog ofa) = Mig © Ale) = Moe) Hence (heaves = hewen INVERSE OF A FUNCTION 38. Let A = (1,2,8,4,5). Let the function f:A-A be defined by the diagram aa (1) #-*(2), (2) #08), (8) F(A), (4) F123, (5) 72,8, 4). lution: (1) #7") consists of those elements whose image is 2. Only 4 has the image 2; hence /-*(2) = (4) @) [-*(8) =O sinco 8 is not the image of any element in the domain, (8) F-4@) = (1,3,5} since f(1)=4, 1(8)=4, 7(6)=4 and since 4 is not the image of any other element, (4) J7*(,2) consiste of those elements whose image is either 1 or 2; hence f-"(1,2) = (2,4). (8) J-42,8,4) = (4,1,8,5) since each of these numbers, and no others, has 2, 3 or 4 as an image point.CHAP. 4) FUNCTIONS 69 ‘34, Let the function #:R*~R* be defined by f(x) = 2%. Find: @) £725), QB I*-9), BFUHLI), HMC #O), 6) F Solution: (1) f°*(25) = {5,5} since (6) = 25 and f(—5) = 25 and since the square of no other number is 25. @) #-"(-8) =O since there is no real number whose square is —9, 1 the equation z*=~9 has no real root. ((4,25)). ©) F-N(-A,A)) = (4,1) since [2] =1 implies [x"|<1, Le. if # belongs to [-1,1] then /( Delongs to [=1, 1]. (4) $-(#,0)) = {0} since 0 = Oe (-,0] and since no other number squared belongs to (~=, 0]. (6) f7*((4,25)) consists of those numbers whose squares belong to [4,25], ie. those numbers x such that Hence £4, 25) also fe | 2= <5 or 5 = 2-2) 35. Let f:A>B. Find f-1(f(A)), that is, find the inverse of the range of f. Solution: ‘Since the image of every element in A is in the range of f, ria) = A ‘under all circumstances, INVERSE FUNCTION 96. Let f:A-B, and let f have an inverse function f-!:B>A, State two properties of the function f. Solution: ‘The function f must be both one-one and onto. 37. Let W = (1,2,8,4,5), and let the functions /:W> W, g:W-> W and h:W> W be defined by the following diagrams: , 0 A | SS p< 7 | A yr BS Which of these functions, if any, has an inverse function? Solution: In order for a function to have an inverse, the function must be both one-one and onto. Only is one-one and onto; hence A, and only h, has an inverse function. 38, Let A = [-1,1]. Let the functions f,, fs, fs and fs of A into A be defined by: 2) file) = 2%, (2) fala) = 2%, (8) fo(a) = sinx, (4) fale) = sin dex State whether or not each of these functions has an inverse function. Solution: | (fo ater enone oot; hence J at ne Inver (2) fais one-one since xy implies 2% y*, Also, fr is onto. Hence fs has an inverse function, (8) fu is a one-one function but not onto; hence fs has no inverse, (4) fea n inverse function since it is one-one and onto.60 FUNCTIONS [CHAP. 4 39. Prove: Let f:A>B and g:B~C have inverse functions f-':B-> A and g™!:C > B. ‘Then the composition function g°f:A>C has an inverse function which is frog t:C> A, Solution: By Theorem 4.3, we must show that eoewon and oes By repeated use of Theorem 4.1, the associative law for composition of functions, we compute eo yewen 2W'eeN) = Stolen) of) eden = step ad Notice that we use the property that g-' 9 is the identity function, and that the product of 1, the identity function, and f is /. Similarly, weneureg) geeuteo = glee neg) 40, Let f:R*>R* be defined by f(z) = 2x-3. Note that f is one-one and onto, so f has an inverse function f-1:R* > R*. Find a formula that defines the inverse fune- tion f Solution: Lat y be the image of = under the function f. Then v= fle) = 22-3 Consequently, x will be the image of y unfier the inverse function f~', Le., e=r'w Solving for 2 in terms of y in the above equation, es tae Then FW) = w+ 92 ‘a formula defining the inverse function. Note y is merely dummy variable; hence Pa) = +92 also defines the inverse function. Moreover, the latter expression is preferable since z is customarily ‘used in defining functions. 41. Let f:R*>R* be defined by f(z) = 2*+5. Note f is one-one and onto, so f has an inverse function. Find a formula that defines f-'. Solution: Solve for # in terms of ys y= 2*-+5, y-5 =< and 2 = Vy—6. ‘Then the inverse function ie f-'(e) = Yz=8. 42, Let A = R*~{3) and B= R*—(1). Let the function f:A>B be defined by z-2 fe) = $33 ‘Then f is one-one and onto, Find a formula that defines f~ SolutiCHAP. 4) FUNCTIONS 61 MISCELLANEOUS PROBLEMS 43, Let the function f:R*>R* be defined by f(x) =2*-32+2. Find: (a) H(-8) (e) f(a) (@ f(2z—3) (mn) f(f(z +1) (b) f2)-f(-4)—) flv—2) (A) f@x—-3) + Ha+3) — (n) fle +h)— f(z) (©) fw) (9) fiz+h) (kt) flat Bx +2) (0) [f(z +h) ~fa)vr @) fe’) (h) fe +3) @ f@) Solution: ‘The function assigns to any clement the square of the clement minus $ times the element plus 2. (8 — 8-8) +2 = 94942 = 20 ey 3a) +2 M8) = (4 = 8-4) +2 = 80. Then 1) ~ f(-4) = 0-30 = ~20 (yt 8y) +2 = vay +2 Be) +2 = at aah +2 (et ale) 42 = at at se ya 2 8y—2) +2 = yh dye tt By te +2 ©) fe+h) = (e+hP~ Seth) +2 = att Bh EM Beh +D (8) fe+8) = (43)! Sle+8) +2 = W649) — Se 942 = at Sete (i) f@x—3) = (22—3)* ~ 3(2e—3) +2 = 4x*— 122+ 9-6r+9+2 = 4x*— 18x + 20 J) Using (h) and (i), we have ‘f@x — 8) + fle +3) = (Ax*— 182 + 20) + (a+ 32+2) = Bat — 152 + 22 () flet—S242) = (et 32 +2) — ale 8042) 42 = at Get + 100 Be © Mle) = fet 32+2) = 2462" + 102" — 32 (om) 92+ 0) = Mlle+D"~ Sle +1) +2)) = f(le4+ 2241 ~ 2-342) = fat—x) = (atx) — 82" 2) +2 = ot — Bet — 2s + B24 2 (0) By (o), leh) = #4 2ch+ M—Be— Bh 2, Hence Ale +h) — fx) = (a + 2h + hY— 8x — Bh +2) — (x*—3e+2) = Bch + hi — Bh (0) Using (n), we have (e+ h) — flel/h = Beh h—Shyih = Oe + h—3 44, Let the functions f:R*>R* and g:R*>R* be defined by f(x) = 27-8 and G2) = +5. Find (a) 4(5), (b) 9(—8), (c) a(f(2)), (a) F(9(8)), (e) 9(a—1), (f) Klg(a—1)), (9) 9(f(z)), (h) Kole +1), (i) 9(9(@)). Solution: (@) A) = 2-3 = 10-3 (Cares = 945 = 14 ‘H(2@)~3) = 94-3) = 94) = O45 = 6 A{S'+5)) = M(9+5) = f04) = 24) —2 = 26 (= +5. Mo+1+5 = t- 2046 wwe have Kola—)) = fle'—2a+6) = 2a 20+6)~3 = Yat da +9 (©) ole) = of@e—3) = (22a +5 = det 122 414 () Hole+4) = Mle+O + 5) = Nettoz+1+5) f+ 2246) = 2@+2246)—3 = Bet 4 deo © ole) = ole +8) = +O 45 = 24+ 1004 + 206 FUNCTIONS (CHAP. 4 Supplementary Problems DEFINITION OF A FUNCTION 45,” State whether or not each of the diagrams defines a function of {1,2,8} into (4,5,6). yt L<{ Lt ® @ ® 46, Redefine the following functions by use of formula: (1) To each real number let / assign its square plus 3. (2) To each real number let g assign the number plus the absolute value of the number. (3) To each real number greater than or equal to 3 let h assign the number cubed, and to each number less than $ let h assign the number 4. AT. Let the function f:R¢ + R# be defined by f(e) ) fle-2). i248. Find (1) (4), (2) (8), (8) fly—2), 48. Let the function g:R# + R# be defined by o(z) = {2 73* eS? fe Be it e=2 ) 916), (2) 910), (8) o(~2) 49, Let T = [-8,5) and let the function f:T+R# be defined by f(z) = 224 ©) 18), ( f0—2). 1 Find (@) 2), Bets ero 50, Let the function h:R# + RE be defined by (a) = 42*—|2| if xe (-2, 9]. and itecoo HB). Find (a) 18), (8) W(2), (€) M15), (@) AAG), Si, Let X= (2,3) and Y = {1,3,5}. How many different functions are there from X into Y? RANGE OF A FUNCTION 452, The following diagrams define functions f, g and h which map the set (1,2,8,4) into itself. rf) Go) & Find (1) the range of /, (2) the range of g, (8) the range of h. 58. Let W = (-1,0,2,5,11). Let the func range of f. [:W-R® be defined by fle) = z*—x—2 Find the 54, Consider the following six functions: fas(-@, 8) > Re fas[-1,4) > Re fe:[-5,9) > Re If each function is defined by the same formula, fa) = 2 i.e. if to each number + each function assigns 2, find the range of (1) fu, (2) fy @) fy (4) fu (8) fy ©) foCHAP. 4] FUNCTIONS 63 455. Consider the aix functions in Problem 64. If each function is defined by the formula fay = sie. If to each number # each function assigns 2%, find the range of (1) fy (2) fu (8) fa, (d) fu ©) fy ©) fo 56. Consider the functions in Problem 54. Suppose each function is defined by the formula fay = 2-8 Find the range of (1) fx, (2) fo, (8) fo, (4) Jo, (8) fa (@) fee 51, Consider the functions in Problem 64. Suppose each function is defined by the formula Ney = 2244 Find the range of (1) fs, (2) fay (8) fa (8) fy (8) fo, (8) fe 88. Suppose f:A + B. Which of the following is always true: () HAVE B, @) MAY=B, (8) MA)IB. ONE-ONE FUNCTIONS 58. Let f:X—Y, State whether or not each of the following properties defines a one-one function: 1) Ha) = 40) implies @=b (3) fla) # (0) implies a b 2) a=b implies fla) = 00) (4) ab implies fla) #10) 60, State whether or not each of the functions in Problem 64 is one-one. 61, State whether or not each of the functions in Problem 85 is one-one, 62, State whether or not each of the functions in Problem 52 is one-one, 68. Prove: If f:A-B is oneone and if 9:B->C is one-one, then the product function g ° f:A > C PRODUCT FUNCTIONS 6. is f:A~>B, g:B+A, h:C+B, F:B+C and G:C are pictured in the dia- State whether or not each of the following defines a product function and if it does, determine its domain and co-domain: Geeh, Qhef, WM Fes, MGs, Gorh, Fok, hoGg, Bho. 65. Consider the functions f, g and h in Problem 62. Find the product functions (1) fog, (2) he f, @) gogo 66. Let the functions /:R#—+R# and g:R#>R# be defined by fla) = +8241, gle) = 22-3 Find formulas which define the product functions (1) f°9, @) 9°f, @) 9e9, (d) fof. 61. Let the functions f:R# + RE and g:R* > RE be defined by Ha) = at 2h, ole) = a +1 Find (a) (A, (6) Foo, (@ (NI), (@) Ve oN).64 FUNCTIONS (CHAP. 4 INVERSE OF A FUNCTION 68. Let f:R* + R* be defined by f(e) = 241. Find (1) f*(6), (2) $-*(0), (8) F-*(10), (4) F>*(-5), () F120, 26), 6) 4((0,5), (D FL-B, 1, B) F-"(-B, 5). 69, Let g:R#+R* be defined by (2) sine, Find (2) o*(0), @) 97*Q), (8) 9°*(2), 4) oM((-1,1)). 10. Let f:A > B. Find f-"(B). ‘MISCELLANEOUS PROBLEMS TL, Let f:R#>R# be defined by f(z) = 82+4, Then / is one-one and onto. Give a formula that defines 12, Let A= R¥—(-1/2) and B= R—(1/2). Let f:A>B be defined by fa) = (e—aye+1) ‘Then f is one-one and onto. Find a formula that defines /~ 13, Let W = [0,=). Let the functions /:W->W, 9:W->W and h: WW be defined by MMe) = 2, ge) = att3, May = 242 ‘Which of these functions, if any, is onto? TA, Let the function f:R# > R# be defined by (2) =2t+2—2 Find @ 48) © fle-% —(@) Sw) () fe+®— fe) @ FG HB) © 8-12) 12) Heth) HU). are TB. Let f:A>B, 9:B A, andlet 9 °f=1,, the identity function on A. State whether each of the following is true or false. ao (8) 7 is a one-one function. (8) 9 is a one-one funetion. (2) f in an onto function. (4) g is an onto function. Answers to Supplementary Problems 45. (1) No, (2) Yes, (8) No we ite=3 48. () fe) = 248, @) oe) = 24 lel, Ale) = {i fees AT. (1) 8, (2) 24, (B) yh Sys tas — Ay + BE +3, (4) oh 82 +15 48. 2) 10, @ 2, G0 49. (@) 1, (®) Undefined since 6 is not in the domain, (¢) 2°—8t+1 when ~1=¢=7 50. (2) 6, (B) 28, (c) 19, (a) 45 51, Nine 52) 12,4, @) 4,2,3,4, (8) (1,3) 58. {0,—2, 18, 108) 5A. (1) [0,4], (2) [0,9], (8) [0,9], (4) (25, *), (6) (0,16), (6) [0,25]CHAP. 4] FUNCTIONS 6 55. 56, st. a. a 1m. n nn 1. mM. 8. (1) [8,8], 2) [0,27), (8) [27,0], (4) (#,—225}, (6) (1,64), (6) [-125,27) (1) [8-1], 2) [-8,0}, (8) [-6,~3}, (4) (#8), (5) [-4,), (6) [-8,0) (1) (0,81, (2) (4,20), (8) (2.4), (4) (6), (6) (2,12), (6) [-6,10) Ay () Yea, (2) No, (8) No, (4) Ye (1) No, (2) Yes, (8) Yes, (4) Yes, (5) No, (6) No ‘They are all one-one. Only g is one-one, We must show that (9°/\(a) = (g (2) implies a=b. Let (ofa) = (9 °f\t). Then, by definition af prot funciona) = (0 2 he) = 2/40) = aU). Since g ix enrane, fl) = 10), and since f is one-one, a=. Hence 9 °f is one-one. (1) g2f:A>A, (2) Not defined, (8) Fe f:A+C, (4) Not defined, (6) geh:C~A, (6) Feh:C>C, () hoGog:B>B, (8) hoG:A>B. °o rs () eo oy2) = Axor +1 (8) (92 o\2) = de -9 @ (ofa) = Be + 02-1 (4) Ge Nle) = a4 Ga + Lat + 152 + 1. (a) 10, (6) 15, (6) 5, (a) 624 @) 22) {3-8) —) fe | ~B= 2 = 3 or B= mo ao wo @) te | -2=2=2) @) | -2<2<2) () (es) -B8, 4,0, Bey.) = Ge | 2= me where me Z) (2) {e | 2 = G/2)+2en where neZ) (8) 9, (4) RF, the set of all real numbers PB) =A $9 @) = (e438 PG) = +220) Only f is onto. (a) 10 @-2 (0) 2th + RK @ @-3) wo ©) vty-2 (h) a+ Bat 20 — 9a wo (© #—32 QD Ptrhths2th-2 (Ba (1) False, (2) False, (8) True, (4) True, (5) FalseChapter 5 Product Sets and Graphs of Functions ORDERED PAIRS Intuitively, an ordered pair consists of two elements, say a and b, in which one of them, say a, is designated as the first element and the other as the second element. An ordered pair is denoted by (a,b) ‘Two ordered pairs (a,) and (¢,d) are equal if and only if a=e and b=d. Example 11: ‘The ordered pairs (2,3) and (2,2) are different, Example 1.2: The points in the Cartesian plane shown in Fig. 5-1 below represent ordered pairs cof real numbers. Example 13: The set {2,8} is not an ordered pair since the elements 2 and 3 are not distin- suished. Example 14: Ordered pairs can have the same first and second elements such as (1,1), (4,4) and (5,5). Although the notation (a,b) is also used to denote an open interval, the correct mean- ing will be clear from the context. Remark 5.1: An ordered pair (a,b) can be defined rigorously by (a,b) = { {a}, (a,b}} From this definition, the fundamental property of ordered pairs can be proven: (a,b) = (c,d) implies a=c and b=d PRODUCT SET Let A and B be two sets. The product set of A and B consists of all ordered pairs (a,b) where ae A and beB. It is denoted by AXB which reads “A cross B”. More concisely, AXB = {(a,b) | aed, beB) Let A= (1,2,8) and B= (a,b). ‘Then the product set AB = (ya), (1,0), 2a), (2,0), (a), (8,)) Let W = (at), Then wx Example 23: The Cartesian plane shown in itself, ie, R# x RE, The product set A x B is also called the Cartesian product of A and B. It is named after the mathematician Descartes who, in the seventeenth century, first investigated the set R* x R*. It is also for this reason that R* x R*, as pictured in Fig. 5-1, is called the Cartesian plane. Example 2 Example 22: (6,0), (0, (ba), (60) ig. 6.1 is the product set of the real numbers with 66CHAP. 5] PRODUCT SETS AND GRAPHS OF FUNCTIONS 67 Remark 5.2: If set A has » elements and set B has m elements then the product set Ax B has n times m elements, ie. nm elements. If either A or B is the null set then A xB is also the null set. Lastly, if either A or B is infinite and the other is not empty, then Ax B is infinite. Remark 53: The Cartesian product of two sets is not commutative; more specifically, AXBeBXA unless A= B or one of the factors is empty. COORDINATE DIAGRAMS ‘The reader is familiar with the Cartesian plane R* x R*, as shown in Fig. 5-1 below. Each point P represents an ordered pair (a,b) of real numbers. A vertical line through P meets the horizontal axis at a and a horizontal line through P meets the vertical axis at b, as in Fig. 5-1. B aH y . eae Fig.5-1 Fig.5-2 The Cartesian product of any two sets, if they do not contain too many elements, can be displayed on a coordinate diagram in a similar manner. For example, if A = (a,b, ¢,d} and B = (x,y,z), then the coordinate diagram of Ax B is as shown in Fig. 5-2 above. Here the elements of A are displayed on the horizontal axis and the elements of B are displayed on the vertical axis. Notice that the vertical lines through the elements of A and the horizontal lines through the elements of B meet in 12 points. These points repre- sent Ax B in the obvious way. The point P is the ordered pair (c, 1). GRAPH OF A FUNCTION Let f be a function of A into B, that is, let f:A>B. The graph f* of the function f consists of all ordered pairs in which aeA appears as a first element and its image appears as its second element. In other words, ft = (a,b) | aed, b=f(a)) Notice that /*, the graph of f:A > B, is a subset of A x B. Example 31: Let the function f:A > B be defined by the diagram A B es ‘Then fla) =2, 1(6) = 3, f(e) =2 and fld)=1. Hence the graph of f is F = (a2), (6,8), (6,2),68 PRODUCT SETS AND GRAPHS OF FUNCTIONS (CHAP. 5 Example 32: Let W = (1,2, 8,4}. Let the function /:W > R# be defined by fla) = 2438 ‘Then the graph of f is f= (0,4, 2,5), 8,8), 40) Example 83: Let N be the natural numbers 1,2,3, .... Let the function, ale) = a NN be defined by ‘Then the graph of g is 9° = (G1), 2,8), (82, (4,60, «9 PROPERTIES OF THE GRAPH OF A FUNCTION Let f:AB. We recall two properties of the function f. First, for each element acA there is assigned an element in B. Secondly, there is only one element in B which is assigned to each aeA. In view of these properties of f, the graph f* of f has the following two properties: Property 1: For each ae, there is an ordered pair (a,b) e /*. Property 2: Each ae A appears as the first element in only one ordered pair in f*, that is, (a, b)ef*, (a,c)ef* implies b= In the following examples, let A = (1,2,3,4) and B = (3, 4,5, 6). Example 41: The set of ordered pairs (4,5), 2,3), (4,6) cannot be the graph of a function of A into B since it violates Property 1. Specif- cally, $e. and there is no ordered pair in which 8 is a first clement. Example (2: The set of ordered pairs (11,5), 28), (8,6), 4,8), 2 40 ccannot he the graph of a function of A into B since it violates Property 2, that is, the element 2¢4 appears as the first element in two different ordered pairs (2,3) and (2,4), GRAPHS AND COORDINATE DIAGRAMS Let f* be the graph of a function f:A>B. As f* is a subset of Ax B, it can be dis- played, i.e. graphed, on the coordinate diagram of A x B. Example 51: Let f(z) = 2" define a function on the intervs ia diaplayed in Fig. 5-8 below in the usual we <4, Then the graph of f eed Fig. 53 Fig. 5-4 Fig. 55 Example 52: Let a function /:A > B be defined by the diagram shown in Fig. 5-4 above. Here f*, the graph of f, consists of the ordered pairs (a,2), (6,8), (1) and (d,2). ‘Then /* is displayed on the coordinate diagram of AX B as shown in Fig. 6-5 above.CHAP. 5] PRODUCT SETS AND GRAPHS OF FUNCTIONS 69 PROPERTIES OF GRAPHS OF FUNCTIONS ON COORDINATE DIAGRAMS Let f:A+B. Then f*, the graph of f, has the two properties listed previously: Property 1: For each ae A, there is an ordered pair (a,b) /*. Property 2: If (a,b)e/* and (a,c)e/*, then b=c. Therefore, if f* is displayed on the coordinate diagram of A x B, it has the following two properties: Property 1: Each vertical line will contain at least one point of /*. Property 2: Each vertical line will contain only one point of /*. Example 61: Let A b,¢) and B = {1,2,8). Consider the sets of points in the two coordinate diagrams of A x B below. al | 2 | abe rary ay ® In (1), the vertical line through } does not contain a point of the set; hence the set of points cannot be the graph of a function of A into B. In (2), the vertical line through a contains two points of the set; hence this set of points cannot be the graph of a function of A into B. Example 62: The circle 2*+y! = 9, pictured below, cannot be the graph of @ function since there are vertical lines which contain more than one point of the circle, sty = 9 is plotted FUNCTIONS AS SETS OF ORDERED PAIRS Let f* be a subset of Ax B, the Cartesian product of sets A and B; and let /* have the two properties discussed previously: Property 1: For each acA, there is an ordered pair (a,b) f*. Property 2: No two different ordered pairs in /* have the same first element. ‘Thus we have a rule that assigns to each element ac A the element beB that appears in the ordered pair (a,b) e f*. Property 1 guarantees that each element in A will have an image, and Property 2 guarantees that the image is unique. Accordingly, f* is a function of A into B. In view of the correspondence between functions f:A>B and subsets of Ax B with Property 1 and Property 2 above, we redefine a function by the Definition 5.1:] A function f of A into B is a subset of Ax B in which each ae A appears as the first element in one and only one ordered pair belonging to f.10 PRODUCT SETS AND GRAPHS OF FUNCTIONS [CHAP. 5 Although this definition of a function may seem artificial, it has the advantage that it does not use such undefined terms as “assigns”, “rule”, “correspondence”. Example 71: Let A = (a, 6,6) and B= (1,2,3). Furthermore, let 1 = (64,2), (6,0), 0,2) ‘Then f has Property 1 and Property 2. Hence / is a function of A into B, which is also illustrated in the following diagram: Example 72: Let V = (1,2,8} and W = (a,¢,4,0,u). Also, let fF = Ga, 2,4,8,9,2,0) ‘Then f is not a function of V into W since two different ordered pairs in f,(2,¢) and (2,1), have the same first element, If f is to be a function of V into W, then it cannot assign both ¢ and w to the element 2eV. Example 73: Let S = (1,2,3,4) and 7 = (1,3,5}. Let 1 = (2,2), 2,5), 4,3) ‘Then f is not a function of S into T since 3 does not appear as the frst element in any ordered pair belonging to f. ‘The geometrical implication of Definition 5.1 is stated in Remark 54: Let f be a set of points in the coordinate diagram of Ax B. If every vertical Tine contains one and only one point of f, then is a function of A into B. Let the function f:A->B be one-one and onto. Then the inverse func- tion f-* consists of those ordered pairs which when reversed, i.e. permuted, belong to f. More specifically, fr = (Ga) | @ben Remark 55: PRODUCT SETS IN GENERAL ‘The concept of a product set can be extended to more than two sets in a natural way. ‘The Cartesian product of sets A, B and C, denoted by AXBxC consists of all ordered triplets (a,b,c) where acA, beB and ecC. Analogously, the Cartesian product of sets As, A2,...,Ay, denoted by ALx AaxX 16+ X Ay consists of all ordered n-tuples (a1,2,...,ds) where ae As, ...,dy¢A,. Here an ordered n-tuple has the obvious intuitive meaning, that is, it consists of n elements, not necessarily distinct, in which one of them is designated as the first element, another as the second element, ete. Example 81: In three dimensional Euclidean geometry each point represents an ordered triplet, i.e ite z-component, its y-component and its z-component. Example 82: Let A = (a,0}, B= {1,2,9) and C= (2,y}. Then AXBXC = ((a,1,2), (a 1,0), (2,2), (a,2, v), (a8, 2), (4,3, 9), (1,2), Aw), (62,2), 2,0), (b, 8,2), (b, 3.)CHAP. 5) PRODUCT SETS AND GRAPHS OF FUNCTIONS a Solved Problems ORDERED PAIRS AND PRODUCT SETS L (John, Jim, Tom} and let V = (Betty, Mary). Find Wx V. W x V consists of all ordered pairs (a,6) where ac W and be V. Consequently, WXV = {(Gohn, Betty), (John, Mary), (Jim, Betty), im, Mary), (Tom, Betty), (Tom, Mary)} 2 Suppose the ordered pairs (x+y, 1) and (8, 2—y) are equal. Find x and y. Solution: If (e+y,1) = @,2—w) then, by the fundamental property of ordered pairs, ety =8 and ‘The solution of these simultaneous equations is B 3. Find the ordered pairs corresponding to the points I PJ P:,P2,Ps and P, which appear in the coordinate diagram 7 of AXB in Fig. 5-6. Here, A = {a,b,¢,d,e) and Ll B= (a, ¢,i,0,). Solution: ‘The vertical line through Ps crosses b on the A axis, and the P, horizontal line through P, crosses i on the B axis; thus P, corre- crEEG sponds to the ordered pair (6,1). Similarly, Ps = (oa), Ps = (2,4) and P. = (6,0). Fig. 56 4. Let A = (a,b), B= (2,8) and C= (3,4). Find (1) Ax(BUC), (2) (AX B)U(AXO), (8) AX(BNC), (4) (AX B)N(AXC) Solution: (1) First compute BUC = (2,3, 4). Then AX (BUG) = (a2), (0,8), (0,4), (0,2), (0,8), (0,4) (2) First find AX Band A x ¢: AXB = (4,2), (0,8), (b,2), (0,90 AXE = (4,8), (0,4), (2), (0,4 ‘Then compute the union of the two sets (AX BY(A XC) = (0,2), (0,8), (6,2), (0,3), (0,4, (0,49) Notice, from (1) and (2), that Ax(BUC) = (AXB)UIA XO) 8). ‘Then AX (BOG) = (0,3), (,9)} (4) In (2), AxB and AXC were computed. ‘The intersection of Ax B and AXC consiste of the ordered pairs which belong to both sets, ic, (AX B)AXC) = (a9), (6,3)) Notice, from (8) and (4), that, AX(BNQ = (AXBNAXO) (8) First compute BNC2 PRODUCT SETS AND GRAPHS OF FUNCTIONS 5. Sketch, by shading the appropriate area, the product set (2) 1=2<4) x (2) 22 on the coordinate diagram of R* x R¥. Solution: 3) (CHAP. 5 Draw light vertical lines through 1 and 4 on the horizontal axis, and horizontal lines through 2 and 8 on the vertical axis as shown in Fig, 5-7 below. ‘The rectangular area bordered by the four lines, together with three of its sides, represents the product set. Shade the diagram as shown in Fig, 5-8 below. Notice that the side of the rectangle which does not belong to the product set is shaded by a dotted line. Fig.5-1 6 Prov Solution: ACB and CCD implies (Ax C) c (Bx D). Let (2,9) be any clement in A XC; then 2cA and yeC. By hypothesis, A is a subset of B and Cis a subset of D; then zeB and yeD. Thus the ordered 7. Let A Solution: "A convenient method of finding AX B XC is through the socalled “tree diagram” shown bel AX BX C consists of the ordered triplets on the right of the “tree” B= (2,4) and C = (8,4,5). Find Ax Bx C. 2,8) (42,4) 1,2, 5) 14,3) (4,4) 45) (22,3) @2,4) 2,5) (2,43) waa) (24,5) (82,8) (32.4) 2,5) (54,3) G44) 4.5) ir (2,y) belongs to BXD. We have shown that (<,y)©AXC implies (z,y)eBXD; hence A XC is a subset of Bx D.CHAP. 5] PRODUCT SETS AND GRAPHS OF FUNCTIONS 3 & Prove: Ax (BNC) = (AxB)N(AXC). Solution: We first show AX (BAC) is a subset of (A xB) (A XC). Let (#¥) be any element of Ax (Bn); then ze A and ye BNC. By definition of intersection, y belongs to both B and to C. Since zeA and yeB, then (z,y)¢AXB. Also, since zeA and yeC, then (#,y)©AXC. Accord ingly, (z,y) belongs to the intersection of AX B and A XC. We have therefore proven A x (B.C) ¢ AX BOA KO, Next we want to show that (AX B) (Ax C) is a subset of A x (BC). Now let (s,1) be any clement of (A XB) (A x Ch; then (z,1) belongs to AB and (2,1) belongs to A xC. It now follows that 7A and weB, and zeA and weC. Since w belongs to doth B and C, then we BOC. We have zeA and weBXC; then (2,W) © Ax (BMC). We have just proven that (A XC) 0 (Ax 0) is'n subset of Ax (BNC). By Definition 1.1, the seta are equal. = (a,b), W = (1,2,3,4,5) and V = (8,5,7,9). Find (Sx Wn (SV). Solution: ‘The product set (S x W)N(S x V) can be found by first computing Sx W and S x V,_ and then computing the intersection of these sets, But, by Problem 8 (Sx W)nisxV) = Sx(WaV) ‘Then WOV = (8,5, and (Sx WAXY) = SX WAV) = {a,3),(a,5), (6,8), (0,5) GRAPHS OF FUNCTIONS 10, Let. W = (1,2,8,4}, and let the function f:W->R* be defined by the formula f(z) = 2%, Find the graph f* of the function f. Solution: First compute /(1) 1, (2) = 2 = 4, @) = 9 = 9, fd) = 4 = 16. The graph f* of f const of the ordered par jb. (2) where x41W- Accorinaly, f° = (1), (0, 80) (418)). IL Let _V = (a,b, c,d), and let the function g:V->V be defined by the diagram shown in Fig. 5-9 below. Find the graph g* of the function g, and sketch g* on the coordinate diagram of Vx V. ‘ LI SS et > A ne Solution: Note, by the diagram, g(a) (¢,0), (@,a)}. Plot the ordered above. 4, g(0) =e, ofc) = and g(d) =a. Hence g* = {(a,2), (bye) ‘of g* on the coordinate diagram of Vx V as shown in Fig, 5-10, 12, Let the function h:R*>R* be defined by h(x) = z+3. State whether or not each of the following ordered pairs belongs to the graph h* of the function h: (a) (2,6), (b) (8,11), (¢) (10,12), (a) (4,7), (e) (—6,-9), (A) (-1,2). Solution: (@) (2) = 248 = 5; 0 (2,6) does not belong to AY. (d) MA) = 443 = 7; 20 4, Tee hg) = 843 80 (8,11) does belong to AY. (2) M-8) = ~8 +8 = ~8; 0 (6-9) ¢ A. (©) (20) = 10+3 = 13; 90 (10,12) ¢ WY. OM) = AF8= 2 80-2) Er™ PRODUCT SETS AND GRAPHS OF FUNCTIONS (CHAP. 5 13, Let S = (a,¢,i,0,u). To each letter in S let the function g assign the letter which follows it in the alphabet. Find the graph g* of the function g. Solution: Firat find g(a) = 6, 9(e) ‘Thos p and 9(u) a, 6), (6.0 sid, (0, (2)? FUNCTIONS AS ORDERED PAIRS 14, Let V = (1,2,8,4). State whether or not each of the following sets of ordered pairs is a function of V into V. (1) fi = (2,3), (1,4), 2,1), 8,2), 4,4) (2) fe = ((8,1), (4,2), (1,1)) (8) fs = ((2,1), (8,4), (1,4), 2,1), (4,4)) (4) fe = ((2,8), (1,6), (4,2), (8, 4)} Solution: Note first, by Definition 5.1, that a subset fof VX is a function f:V—V if each eeV ‘appears as the first element in one and only one ordered pair in f. (1) Since two different ordered pairs (2,8)¢/: and (2,1)efs have the same first clement, fi is not a function of V into V. (2) The clement 2 V does not appear as the first clement in any ordered pair belonging to fs. Hence fu is not a function of V into V. (8) The set fr is a function of V into V. Although 2 appears as the first element in two ordered pairs, these two ordered pairs are equal (4) Although each clement in V appears as the first element in one and only one ordered pair belonging to fs, the set fs is not a function of V into V since f« is not a subset of VV. Specifically, (1,6) ef. but (1,)¢V XV. 15. Let W = (a,b,¢,d). State whether or not the set of points in each of the following coordinate diagrams of Wx W is a function from W into W. Het ore ttt ot “HE oft ttt “HEH EEE a oy @) co = Note first that a set of points on a coordinate diagram Is a function provided that every vertical Jine contains one and only one point of the set. (2) The vertical line throygh b contains two points of the set; hence the set is not a function from W into W. (2) Since each vertical line contains one and only one point of the set, this set is a function of W into W. ‘The fact that the horizontal line through ¢ contains three points does not violate: the properties of a funetion, (8) The vertical line through ¢ contains no point of the set; hence this set is not a function of W into W. (4) For the same reason as in (2), this set is a function of W into W.CHAP. 5] PRODUCT SETS AND GRAPHS OF FUNCTIONS © 16. 19, Let K = (1,2,8,4,5,6) and let $= (1,2,3,4). The | | 4 set g of points in the coordinate diagram of R x S in the figure at right is a function of R into S. 4 t (a) Find 9(2), 9(4), 0(6). (0) Find 9~* (2),0-"(8),0-*(4). #4 + (@) Find (| xe R, g(x) <8} 1 t Solution: {@) To find 9(2) we look at the point in g which lies on the Sarees ae vertical line through 2 the point i (2,4). Hence o(2 the second element in the ordered pair. ‘The vertical line through 4 contains the point (4,1) of g; hence g(4) = 1 ‘The vertical line through 6 contains the point (6,4); hence (6) = 4. (@) To find 9°*(2) we look at the points that lie on the horizontal line through 2. The points a1 (1,2) and (5,2). ‘Then g~*(2) consists of the first elements of these ordered pairs, Le g~' (2) {1,5}. Note that the ordered pairs (1,2) and (5,2) in g meun that g(1)=2 and p(5)~2 ‘The horizontal line through 3 contains only the point (8,3) of g: then g°*(8) = (3) ‘The horizontal line through 4 contains the points (2,4) and (6,4) of g: then g”'(4) (e) First note that g(l) = 2, g(2) = 4, 9(8) =3, 9(4) =1, 915) =2, o(6)=4. The set tx (2) <8} consists of those elements in R whose image is less than 3, ie. whose image i ‘The set is (1,4,5)- Geometrically, this set consists of the first elements of the points in g which lie below the horizontal line through 3. . Let h be a set of points in the coordinate diagram of Ex F which is a function of E into F. (a) If each horizontal line contains at most one point of h, what type of function is h? (0) If each horizontal line contains at least one point of h, what type of function is h? Solution: (a) If each horizontal line contains at most one point of A, then, for every ze F, h- (2) is empty, or consists of one clement in E. ‘Thus h is a one-one function. (6) If each horizontal line contains at least one point of f, then, for every ze F, h(x) is not empty. Hence A is an onto function, F = ((1,5), (8,1), (4,7), (-2,-8)} define a function of A into B? Solution: ‘The set f will define a function of A into B if f is a subset of AX B and if each clement in A appears as the first element in one and only one ordered pair in f. Accordingly, A must equal the set of first elements of f, ie. A = (1,3,4,~2): and B must contain the aet of second elements of f, he {5,1,7,-8) cB. Let W = [-4,4). State whether or not each of the following sets of points displayed on a coordinate diagram of Wx W is a function of W into W. ®76 PRODUCT SETS AND GRAPHS OP FUNCTIONS (CHAP. 5 Solution: 2 o Note first, by Remark 5.4, that set of points on a coordi vertical line contains one and only one point of the set. (1) Since vertical lines contain two points of the set, the set of points is not a function of W into W. (2) Since the vertical lines close to the vertical axis contain no point of the set, the set of points not a funetion of W into W. (8) ‘The vertical lines through the circle will contain two points of the set; hence the set of points in not a function of W into W. (4) ‘The set of points is a function of W into W since each vertical line contains one and only one point of the set, gram is a function If every 20, Let A= (a,0,¢,d). The set (2, ), (0 @), (6,4), (d,€)) is a one-one, onto function of A into A. Find the inverse function. Solution: ‘To find the inverse function, permute, inverse function ‘Then the . reverse, the elements in each ordered pair. {@,0), (4), (4,0), )) Supplementary Problems ORDERED PAIRS AND PRODUCT SETS 21, Suppose (y~2, 2¢+1) = (e~1,y +2). Find # and y. 22, Find the ordered pairs corresponding to the points P., Ps, Ps and Py which appear below in the coordinate diagram of (1,2,3,4) x (2,4, 6,8). il rea 28. Let W = (Mark, Erie, Paul} and let V Erie, David). Find (1) WV, (2) VXW, (8) VV. 2, ‘Sketch, by shading the appropriate ares, each of the following product sets on a coordinate diagram Reine = 3] x [1.8] @) [-3,) x (-#,2] 8} XB) 4) (8,2) x 8,8] 25, Let A = (2,3), B = (1,3,5) and C = (8,4). Construct the “tree diagram” of Ax B%C, as in Problem 1, and then find AB XC.CHAP. 5) PRODUCT SETS AND GRAPHS OF FUNCTIONS 7 1%, Let S = (a,,0), T= (6, then find Sx Tx W. d} and W = (a,d}. Construct the “tree diagram” of Sx Tx W and 21, Suppose sets V, W and Z have 3, 4 and § clements respectively. How many elements are there in () VX W XZ, (2) ZxVXW, B) WXZXVE 2%, Let A = BNC. Which, if any, of the following is true? () AXA = (BXBN(CXO) AXA = BONCXB) GRAPHS OF FUNCTIONS 2, Let M = (1,2,3,4,5) and let the function /:M > R* be defined by We) = t+ 2e-1 Find the graph of /. 30, Let W = (1,2,8,4) and let a function g:W > W be defined by the diagram (2) Find the graph of g. (2) Display the graph of g on the coordinate diagram of WX W. 31, Let the function h:R# > RE be defined by the formula Ma) = 22-1 State whether or not each of the following ordered pairs belongs to the graph of h: (2) (3,5), (8) (2,8), (0) (—4,—7), (A) BID, (€) (8,8), (7). 32, Let the function g ign to each name in the set {Betty, Martin, David, Alan, Rebecca) the number of different letters needed to spell the name, Find the graph of g. 83, Each of the following formulas defines a function of R# into R2, Plot the graph of each of these functions on a coordinate diagram of R# x R#, the Cartesian plane. ® H) © fe) ta =~ 2el FUNCTIONS AS ORDERED PAIRS BL Let W = (a,b,e,d). State whether or not each of the following seta of ordered pairs is a function of W into W. ) (0), (6,4), (@a), (ed), (a,d)) (8) (a,b), (B,), (eB), (2, 6)) 2) (Cd,4), (0), (0,8), (4,0)) 4) Kae), (0,0), (@,8), (ed) 85. Let V = (1,2,8,4). State whether or not the set of points in each of the following coor diagrams of VV ia s function of V into V. ae i ras eae Tee ay @) 6)8 PRODUCT SETS AND GRAPHS OF FUNCTIONS [omar. 5 38, Let A = (1,2,3,4,5). Let f be the set of points displayed in the first coordinate and let g be’the set of points displayed in the second diagram, 1, isialalalels ela] , +++ + ‘leletslala r iagram of AXA, Tepae ane aere es iinet fis displayed @ Is displayed ‘Thus f and g are functions of A into A. Find HW Or OH (1) product function fog (9) (2 | fle) = 4) (2) 95) (8) 9°") (6) product funetion g of (8) f-*((1,2)) (10) | o(@) > 2 31. Let the function f:A +B be displayed on a coordinate diagram of A x B. What geometrical property does f have if (1) / is one-one, (2) f is a constant function, (8) f is onto, (4) f has an inverse? 38, Let B= [-4,4). State whether or not each of the following sets of points displayed on a coordinate diagram of BB is a function of B into B. @ rc ‘MISCELLANEOUS PROBLEMS 39. Sketch, by shading the appropriate area, each of the following product sets on a coordinate diagram o @) of REX RE, () @|-8<2=2) x | -2-2) x | 229 @) | lel <2) x @ | 2>-2) Each of the following formulas defines a function of R# into R#. Plot each of these functions on coordinate diagram of R# xR. ext (2) fla) = 242\2| (8) fee) = ie eee © he) zit bi=2 2 ifz<-2 B-2ite>2 am te)CHAP. 5 PRODUCT SETS AND GRAPHS OF FUNCTIONS Answers to Supplementary Problems a e=2y 4), Pe = (2,8), Pa = 10), Pa = 0 2.0) Wx @) VxW = @) VV = ((Bric, Erie), (Bric, Day 4. a) ® 25. See Fig. 5-11 below. AXBXC 62). {(ark, Erie), (Mark, David), (Erie, Erie), (Erie, David), (Paul, Erie), (Paul, David). {(Brie, Mark), (David, Mark), (Erie, Erie), (David, Erie), (David, David)}. (David, Erie), (Erie, Paul), (David, Paul)}. @ a {(2 1,8), (2,44), (28,8), (28,4), (25,8), (25,4), 1,8), (1,4), (8,8,3), (83,4), (85,8), (8,5,4)) 28, See Fig. 6-12 below. SxTXW = =s 4 2 —<—! =} = 3 sj =i Fig. 51 4), (a,¢,4), (a,¢,d), (a,d,a), (a, d,d), esa), red), (da), (dad), ), (c,¢,d), (ed, a), (0, d, d)} 1——j . —=! is —s » —! —! — se) © v= Ke) @ v, =), first plot f(z.) = 0. ‘The curve f(z,y) = 0. will, in simple situations, partition the plane into various regions. The relation will consist of all the pointa in possibly one of more regions. Test one or more points in each region in order to determine whether or not all the points in the region belong to the relation. ‘The sketch of each of the above relations is as follows: a @ ety <0 Bay 5090 RELATIONS [cHar.6 ) “ ate DOMAINS, RANGES AND INVERSES 6. Consider the relation R = (1,5), (4,5), (1,4), (4,6), (8,7), (7,6)). Find (1) the domain of R, (2) the range of R, (3) the inverse of R. Solution: (1) The domain of R consists of toe set of frst clements in R; hence the domain of R is, 43,9) (2) The range of R consists of the set of second elements in R; hence the range of Ris 15,4,6,7) of the same pairs as are in ® but in the reverse order; hence R= (6,1), (A), 4,2), 64), 1,9, GT) (3) The inverse of # consis 7. Let T = {1,2,8,4,5}, and let a relation R in T be the set of points displayed in the following coordinate diagram of TXT iS Find (1) the domain of R, (2) the range of R, (8) the inverse of R. (4) Sketch R-* on a coordinate diagram of Tx T. Solution: (1) The element ze 7 is in the domain of if and only if the vertical line through z contains @ point of R. Thus the domain of R is the set (2,4,6), since the vertical line through each of these clements, and only these elements, contains points of R. @) The clement ze 7 is in the range of R if and only if the horizontal line through z contains a point of R. Thus the range of R is the set (1,2,4}, since the horizontal line through each of these lements, and only these elements, contains at least one point of R. (8) Since R 42,1), (4), (4.2), (4,4), (6,29) R (1,2), 4,2), A), (4,4), @5)) (4) Bis displayed on a coordinate diagram of Tx T as follows: Ris sketchedCHAP. 6] RELATIONS a1 Let R = ((z,u) | ze R*, yo Ré, 4z*+9y! = 36). The sketch of R on the coordinate diagram of R* x R# follows: Find (1) the domain of R, (2) the range of R, (3) R-*. Solution: (1) The domain of R is the interval [—8,3], since the vertical line through each of these numbers, and only these numbers, contains at least one point of R. (2) The range of R is the interval {-2,2], since the horizontal line through each of these elements, and only these elements, contains at least one point of R. (8) Ris found by erchanging = and y in the open sentence that defines ; hence RO = (ay) | ee R*, eR, 92t + Ay" = 96) 9. What relationships, if any, exist between the domain and range of a relation R and the domain and range of R-'? Solutio Since R~' consists of the same pairs as are in R except in the reverse order, each first element in R will be a second element in R-! and each second clement in R will be a first element in R*. Consequently, the domain of R is the range of R~* and the range of is the domain of 10, Let R be the relation in the natural numbers N = (1,2,8,...) defined by the open sentence “2r+y = 10”, that is, let R= ((z,v) | reN, yen, 2e+y = 10) Find (1) the domain of R, (2) the range of R, (8) R-'. Solution: Note first that the solution set of 22+ y = 10 R= (1,8), 2,6) 4, (4.2) even though there are an infinite number of elements in N. (1) ‘The domain of R, which consists of the frst clements of R, is (1, 2, 4). (2) ‘The range of R, which consista of the second elements of R, is (8, 6, 4,2). (@) Ris found by interchanging z and y in the open sentence that defines hence R= (ew | 2eN, yeN, et2y = 10F Also, as R-* consists of the same pairs as are in R except in the reverse order, R- can be defined by B+ = (8,1), 6,2), (42), COP REFLEXIVE RELATIONS 11, When is a relation R in a set A not reflexive? Solution: B is not reflexive if there is at least one element ae A such that (@,0) ¢ R. 12 Let W = (1,2,8,4} and R = (1,1), (1,3), (2,2), 8,1), (4,4). Is R reflexive? Solution: R iis not reflexive since 3eW and (3,8) ¢ R.92 RELATIONS [cHAP. 6 18. Let A be any set and let D be the “diagonal line” of Ax A, that is, D is the set of all (@,a) where ae. What relationship is there, if any, between any reflexive relation R in A and D? Solution: Every reflexive relation R in A must contain the “ set of B if B is reflexive. agonal line”. In other words, D is a sub- 14, Each of the following open sentences defines a relation R in the natural numbers N. State whether or not each is a reflexive relation. (1) “2 is less than or equal to y”. (8) “x+y = 10" (2) “x divides y”. (4) “x and y are relatively prime”. Solution: (1) Since for every Ge N, (aja)eR. Hence R is reflexive. (2) Since every number divides itself, the relation is reflexive, (8) Since 3+3¥ 10, 3 is not related to itself. Hence R is not reflexive, (A) The greatest common divisor of 6 and 5 is 5; thus (5,5) ¢R. Hence R is not reflexive, 15. Let H = (1,2,3}. Consider the following relations in E: Ri = (1,2), (8,2), 2,2), (2,3)) Re = {(1,2)) Re = (1,2), (2,8), (1,8)) Ry = EXE Rs = (1,1), (2,2), (2,8), (8,2), (8,8) State whether or not each of these relations is reflexive. Solution: If a relation in E is reflexive, then (1,1), (2,2) and (8,8) must belong to the relation. Therefore only Ry and Ry are reflexive, SYMMETRIC RELATIONS 16. When is a relation R in a set A not symmetric? Solution: not symmetric if there are elements ae A, be A such that (ad eR, (ba) eR Note a%}, otherwise (a,5) eR implies (b, a) « R. 17, Let -V = (1,2,8,4) and R = {(1,2), (8,4), (2,1), (8,8)). Is R symmetric? Solutio B ie not symmetric, since eV, 4eV, (3,4) eR and (4,8) ¢ R. 18. Is there any set A Solution: If A is the null set or if A contains only one element, then every relation in is symmetric. which every relation in A is symmetric? 19. Each of the following open sentences defines a relation R in the natural numbers N. State whether or not each relation is symmetric. (1) “x is less than or equal to y” (8) “x+y = 10” (2) “x divides y” (4) “2+2y = 10” Solution: (1) Since 3=5 but 543, (3,5) and (5,3)¢R. Thus R is not symmetric. (2) Since 2 divides 4 but 4 does not divide 2, (2,4) eR and (4,2)¢R. Hence R is not symmetric.CHAP. 6] RELATIONS 98 (8) If a+b =10 then b+a = 10; in other words, if (a,b) eR then (b,a)eR. Hence R is sym- metric. (4) Note that (2,4) eR but (4,2)¢R, ie. 2+2(4) = 10 but 442(2) # 10, Thus R is not sym- metric. 20. Let E = (1,2,3). Consider the following relations in E: Ri = (0,1), 2,1), 2,2), B,2),2,8)) Re = (1,1), 8,2), 2,3) Ra = ((1,1)) Rs = EXE Rs = {(1,2)} State whether or not each of these relations is symmetric. Solution: (1) Ri is not symmetric since (2,1) eR, but (1,2)¢R. —(4)_-Re is symmetric, (2) Re is symmetric. (8) Ra is symmetric. (8) Ry is not symmetric since (1,2) ¢ Rs but (21) ¢ Rs 21, Prove: Let R and R’ be symmetric relations in a set A; then ROR’ is a symmetric relation in A. Solution: ‘Note first that R and ’ are subsets of Ax A; hence ROR’ is also a subset of AXA and is, therefore, a relation in A. Let (a,b) belong to ROR’. Then (a,8)eR and (a,b) eR’. Since R and Ri are symmetric, (a) also belongs to R and (b,a) also belongs to R'; hence (b,a) © RNR’. We have shown that (a,) © RAR’ implies (b,a)e ROR’; hence ROR’ is symmetric. ANTI-SYMMETRIC RELATIONS 22, When is a relation R in a set A not anti-symmetric? Solution: B is not antisymmetric if there exists elements aeA,beA,a%b such that (a,b) cR and aeR. 23. Let W = {1,2,8,4} and R = {(1,2), (8,4), (2,2), (8,3), (2,1)}. Is R anti-symmetric? Solution: R ie not anti-symmeteic, since Le W, 2eW, 142, (1,2)eR and (21) eR, 24. Can a relation & in a set A be both symmetric and anti-symmetrie? Solution: ‘Any subset of the “diagonal line” of A x A, that is, any relation Rin A in which (a,b) eR implies a= 6 {is both symmetric and anti-symmetrie. Let E = (1,2,8}. Consider the following relations in B: Ri = (1,1), 2,1), 2,2), (8,2),(2,38)} Ra = ((1,1), (2,3), (8,2) Rr = ((L,1)) R= EXE Rs = ((1,2)) ‘State whether or not each of these relations is anti-symmetric. Solution: (1) Ry is not antisymmetric since (8,2)¢ Rs and (2,8) ¢ Ri. (2) Ry is antisymmetric. (8) Ris anti-symmetric. (A) Re is not anti-symmetric since (2,3) © R. and (@,2) ¢ Ru (8) Ry is not anti-symmetric for the same reasons as for R.94 RELATIONS [CHAP. 6 26, Let H = (1,2,3}. Give an example of a relation R in E such that R is neither sym- metric nor anti-symmetric. Solution: ‘The relation R = {(1,2), (2,1), (2,8)} is not symmetric since (2,8) eR but (8,2) ¢R. R is also not antisymmetric since (1,2)eR and (2,1)¢R. 21. Each of the following open sentences defines a relation R in the natural numbers N. State whether or not each of the relations is anti-symmetric. (1) “x is less than or equal toy” (8) “x +2y = 10" (2) “x is less than y” (4) “x divides y” Solution: (1) Since a= and 6=« implies a=, B is anti-symmetric. (2) Ifa}, then either ) is also divisible by 5, (a,d)eR implies (ba) eR Ris symmetric. Let (a,b) eR and (0,6) (a=) + (66) is R; then (a—b) and (b—e) are each divisible by 5. Hence (a—c) ie. (a6) belongs to R. Since (a,b) eR and (b,c)eR implies (a,c) eR Bis transitive, ‘Since R is reflexive, symmetric and transitive, R is, by definition, an equivalence relation. 98. Let R and R’ be relations in a set A. Prove each of the following two statements: (1) If R is symmetric and R’ is symmetric, then RUR’ is symmetric. (2) If Ris reflexive and R’ is any relation, then RUR’ is reflexive. Solution: (1) If (@,) eRURY, then (a,8) belongs to R or R’, which are symmetric, Hence (b, 2) also belongs to R or R’. Then (b,a)¢RUR’ and RUR’ is symmet (@) Bia reflexive if and only if R contains the “diagonal line” D of Ax A. But Dc Rand RC RUR’ implies D.CRUR'. Therefore RUR’ is reflexive. 39. Let R and R’ be relations in a set A. Show that each of the following statements is false by giving a counter example, that is, an example for which it is not true. (1) If Ris anti-symmetric and R’ is anti-symmetric, then RUR’ is anti-symmetric. (2) If R is transitive and R’ is transitive, then RUR’ is transitive. Solution: {(2,1)) are each anti-aymmetrie; but RUR’ = ((1,2),(2,1)) is not anti- {@2,8)) are cach transi but RUR’ = (1,2), (2,3) is not transitive, 40. Let R (@u) | eet, yeR sy Rr (zy) | teR*, yeR*, y= rt) Note that F and R’ are both relations in the real numbers. (1) Sketch the relation RAR’ on a coordinate diagram of R* x R*. (2) Find the domain of ROR’. (8) Find the range of ROR’. Solution: (1). Sketch R on a coordinate diagram of R# x R#, as in Problem 4, and shade R with strokes slanting ‘upward to the right (////); and on the same coordinate diagram sketch R’ with strokes slanting itched area is ROR’. Thus downward to the right (\\\\), as shown in Fig. 6-12. The cross ROR’ is displayed in Fig. 6-18, © eu Rand R’ are sketched Fig. 6-12 Fig. 61398 RELATIONS (CHAP. 6 (2) The domain of ROR’ is (~1,2), since a vertical ‘these points, will contain a point of RAR’. (3) The range of ROR is [0,4], since a horizontal line through each point in this interval, and only ‘these points, will contain at least one point of ROR’. ine through each point in this interval, and only Supplementary Problems BASIC PROPERTIES OF RELATIONS a, 4 46. a. Let R be the relation in A = (2,8,4,5) which is defined by the open sentence “x and y are relatively prime”, ie. “the only common divisor of 2 and y is 1”. (1) Find the solution set of 2, that is, write R as a sot of ordered pairs (2) Sketch Ron a coordinate diagram of AXA. Let R be the relation in B= (2,3, Write Ras a act of ordered pairs, 5,8) defined by the open sentence “| — yl is divisible by 3”. Let C = (1,2,8,4,5}, and let the relation R in C be the set of points displayed in the following coordinate diagram of Cx C. 5 4 14 aaeeateat (1) State whether each is true or false: (a) 1R4, (0) 2R5, (c) 3#1, (d) 5 #3. (2) Write each of the following subsets of C in tabular form: (@) { 8R2) ( (| @2eRP © fel aeR) — @) te | eRo Find (8) the domain of , (4) the range of R, (6) R~ Bach of the following open sentences defines a relation in the real numbers. Sketch each relation on coordinate diagram of R# x RE, Qym. Then mp Aw and therefore m# Mes At Since m was arbitrary, Mie, Ai = 9. 2%. (1) No, (2) No, (8) Yes, (4) Yes 21, There are five different partitions of V: {0,23 (0), 29%, (2.9), (18, 029% (O27. 28. Note that (a, 8)~(a,0), since a+b = b+o; hence R is reflexive. Suppose (a, »)~ (c,d). Then at+d~=b-+e, which implies that e+ =d+a, Thus (c,d)~ (0,0) and B is symmetric, Now suppose (a, 6)~ (c,d) and (c,d) ~(e,f). Then etd =d+e and ct+f=dte, Thus : (ata ser = te + (+e) and, subtracting ¢+d from both sides, a+/= +e. Thus (a, 2) ~ (6 Accordingly, R is an equivalence relation. and hence R is transitive,Chapter 8 Further Theory of Functions, Operations FUNCTIONS AND DIAGRAMS ‘As mentioned previously, the symbol ALB denotes a function of A into B. In a similar manner, the diagram A; A Cc p consists of letters A, B and C denoting sets, arrows f, g and h denoting functions f:A > B, g:B>C and h:A~ C, and the sequence of arrows {f,g} denoting the composite function gof:A~C. Each of the functions h: AC and gof:A~ C, that is, each arrow or se- ‘quence of arrows connecting A to C is called a path from A to C. Definition 81:] A diagram of functions is said to be commutative if for any sets X and Y in the diagram, any two paths from X to Y are equal. Example Li: commutative. ~~ b—z—>e Then ich =f, goi= 5 and gof = joh= goich. Example 12: ‘The functions :A + B and g:B + A are inverses if and only if the following diagrams are commutative: Here 1, and 1, are the Identity functions, RESTRICTIONS AND EXTENSIONS OF FUNCTIONS Let f be a function of A into C, ie. let f:A~C, and let B be a subset of A. Then f induces a function /’:B+C which is defined by #(b) = #00) for any beB. The function /’ is called the restriction of f to B and is denoted by AB Example 21: Let /:R# + R* be defined by f(@) = 2". Then FIN = {yD BM, BM) 4,16), «3 in the restrietion of f to N, the natural numbers, 116CHAP. 8] FURTHER THEORY OF FUNCTIONS, OPERATIONS uy Bxample 22: The set 9 = (2,8), (5,1), (8,7, 6,8), (9,5)) ie a function from (2, 5,3, 8,9} into N Then (2,8), (8,0, 0,5) 1 subset of 9, is the restriction of g to (2,3, 9), the set of first elements of the ordered pairs'in g. We can look at this situation from another point of view. Let f:A~>C and let B bea superset of A. Then a function F:B~ C is called an extension of f if, for every ae A, F(a) = fla) Let f be the function on the positive real numbers defined by fle) =x, that is, ‘the identity function. Then the absolute value function Example = [eter ll = 4-2 ite B by ‘Then (7) = (2,8). Let g:R¥ > R# be defined by g(2) = 2%, and let T = [8,4]. Then OT) = [168] = (e | 9< 2416) Now let of be the family of subsets of A, and let B be the family of subsets of B. It f:A>B, then f assigns to each set Trof a unique set f(7)eB. In other words, the function f: A> B induces a function f:e4>. Although each function is denoted by the same letter /, they are essentially two different functions. Notice that the domain of {icf > B consists of sets. Generally speaking, a function is called a set function if its domain consists of sets.118 FURTHER THEORY OF FUNCTIONS, OPERATIONS [cHAP. 8 REAL-VALUED FUNCTIONS A function f:A> R*, which maps a set into the real numbers, i.e, which assigns to each ae A a real number f(a)e R*, is called a real-valued function. ‘Those funetions which are usually studied in elementary mathematics, e.g, PEA) = dos + aah + oes + ya + ae t(z) = sinz, cos or tanz f(@) = loge or that is, polynomials, trigonometric functions and logarithmic and exponential functions, are special examples of real-valued functions. ALGEBRA OF REAL-VALUED FUNCTIONS Let 7, be the family of all real-valued funetions with the same domain D. Then many (aigebraic) operations are defined in J,. Specifically, let f:D~ R* and g:D~ RF, and let keR#. Then each of the following functions is defined as follows: (G+W:D> RE by (f+ H(t) = fle) +k (f):D+R* by (te) = If) eee sree ya) = Fe)" (f+9):D>R* by — (f= 9\(z) = Mz) + 9(z) (kf):D>R* by (ef\(z) = R(f(2)) (fg):D>R* by (foy(2) = fe)o(z) (flg):D>R* by (ae) = f(@)/a(z) (where g(t) #0) ‘Note that (fg):D-R is not the same as the composition function which was discussed previously. Example 4: Let D = {a,b}, and let f:D-> R# and g:D > R® be defined by: fa) =1, (0) =8 and g(a) =2, (6) = In other words, £ = Ca,0, 6,9 and g = ((a,2), Then Bf —20)la)_ = fla) 2y(a) = 311) —2(2) (8-208) = 3/(6) — 2916) = 818) 21-1) that is, 8f = 29 = (a, —2), (119) Also, since gi(2) = la(#)| and (9 +8)(2) = ole) +3, lol = €(,2,@0) and g +8 = {(a,5), (2) Example 42: Let f:R? + R# and g:R# + R# be defined by the formulas fa) = 22-1 and gfe) = 2 ‘Then formulas which define the functions (8/—29):R* > R¥ and (fg):R# > Re are found as follows: (f= 2o)(2) = 822-1) 22") = — 2x44 60-3. (ave) = (22 — 12") = 22 RULE OF THE MAXIMUM DOMAIN A formula of the form fa) = Ve, oz) =sinz, h(x) = VE does not, in itself, define a function unless there is given, explicitly or implicitly, a domain, ie. a set of numbers, on which the formula then defines a function. Hence the following expressions appear:CHAP. 8) FURTHER THEORY OF FUNCTIONS, OPERATIONS. 119 Let f(x) = x* be defined on [—2, 4}. Let g(2) = sinz be defined for 0=2~2r. However, if the domain of a function defined by a formula is the maximum set of real numbers for which the formula yields a real number, e.g., let f(z) = Wz for 240 then the domain is usually not stated explicitly. This convention is sometimes called the rule of the maximum domain. Example 6:1: Consider the following functions: fale) for 2=0 fe) for 242 fale) for 0 2 Or fiz) = tans for 24 w/t, weN ‘The domains of fx and fc need not have been explicitly stated since each consists of all those numbers for which the formula has meaning, that is, the functions could have been defined by writing fale) = We2) and fala) = tone Example 52: Consider the function f(c) = vI=7; its domain, unless otherwise stated, is [1,1], Here we implicitly assume that the co-domain is R¢. CHARACTERISTIC FUNCTIONS Let A be any subset of a universal set U. Then the real-valued function xaiU > (1,0) defined by we = [1 teed Xe 0 ify is called the characteristic function of A. Example 61: Let U = (a, ,e,d,¢} and A = {a,d,e). ‘Then the function of U into (1, 0) defined by the following diagram FF ZS —_ fa the characteristic fonction x, of A. Note further that any function f:U-> (1,0) defines a subset Ay = (@ | ee, fz)=1) of U and that the characteristic function x4, of Ay is the original function f. Thus there is a one-to-one correspondence between all subsets of U, ie. the power set of U, and the set of all functions of U into (1,0). CHOICE FUNCTIONS Let {Ai),,, be a family of non-empty subsets of B. Then a function FAder* B is called a choice function if, for every ic I, H(A) © As that is, if the image of each set is an element in the set.120 FURTHER THEORY OF FUNCTIONS, OPERATIONS [CHAP. 8 Example 71: Consider the following subsets, Ar = (2,9), As = (13,4, As = (2,5) of B= (1,2,8, 4,5) and consider the following functions of (4s, As, As) into B: rt) f o Note that f is not a choice function since f(As) = 2 does not belong to As, ie. f(s) ¢ Ax. Note farther that g is a choice function since g(A.)e As, g(As) © As and 9(As) © Ax Remark 81: Essentially a choice function, for a given family of sets, “chooses” an element from each set in the family. Whether or not a choice function exists for an arbitrary family of sets, is a question which lies at the foun- dation of the theory of sets. Chapter 11 will be devoted to this question. OPERATIONS ‘The reader is familiar with the operations of addition and multiplication of numbers, union and intersection of sets, and composition of functions. These operations are denoted atllone , arb=c, AUB=C, ANB=C, gof=h Im each situation, an element (c, C or h) is assigned to an original pair of elements. In other words, there is a function that assigns an element to each ordered pair of elements. We now introduce the precise Definition 8.2:] An operation « on a set A is a function of the Cartesian product Ax A into A, ie, aa a atb= Remark 82: The operation «:A x A~ A is sometimes referred to as a binary operation, and an n-ary operation is defined to be a function XAX 1X AOA (n times) We shall continue to use the word operation instead of binary operation. COMMUTATIVE OPERATIONS ‘The operation «:4.x A> A is called commutative if, for every abe A, a(a,b) = a(d, a) Example 81: Addition and multiplication of real numbers are commutat atb=b+a and ab = be Example 82: Let «:R# x R# > RE be the operation of subtraction defined by a: (2,4) #—y. ‘Then (5,1) = 4 and all, 5) = 4 Hence subtraction is not a commutative operation, re operations since Example 83: Union and intersection of sets are commutative operations, since AUB=BUA and ANB = BOACHAP. 8] FURTHER THEORY OF FUNCTIONS, OPERATIONS iat ASSOCIATIVE OPERATIONS The operation «:A x A~ A is called associative if, for every a,b,ce A, elo(a,b), 6) = fa, of )) In other words, if a(a,b) is written ab, then « is ascociative if (aebjec = ae(bec) Example 91: Addition and multiplication of real numbers are associative operations, since (atbte = atte) and — (ab)e = abe) Example 92: Let a:R# X R# + R# be the operation of division defined by a: (z,y) + a/v. ot ‘o(a(12,6), 2) = a(2,2) = 1 (12, a(6,2)) = a(12,8) = 4 Hence division is not an associative operation. Example 93: Union and intersection of sets are associative operations, since (AuB)uC = AU(BUQ) and (ANB) = An(Bnc) DISTRIBUTIVE OPERATIONS Consider the following two operations: AX Ard Bi:AXA>A ‘The operation « is said to distribute over the operation p if, for every a,b,ce A, a(a, (b,c) = lala, d), a(a,c)) In other words, if a(a,d) is written a+b, and (a,b) is written a4, then « distributes eee ae(bac) = (aeb)A(are) Example 101: The operation of multiplication of real numbers distributes over the operation of addition, since ery oe But the operation of addition of real numbers does not distribute over the operation of multiplication, since a+ (be) # (at Wate) Example 102: The operations of union and intersection of sets distribute over each other since AUB) = (AVB)N(AVO, AN(BUC) = (ANBL(ANG) IDENTITY ELEMENTS Let a:AXA >A be an operation written (a,b) = a+b. An element ec is called an identity element for the operation « if, for every element ac A, esa =ave=a Example IL1: Let a:R# x R# + R# be the operation of addition. Then 0 is an identity el ‘ment for addition since, for every real number ae R#, Ota =ae0=a, thai, O+a=e+0-0122 FURTHER THEORY OF FUNCTIONS, OPERATIONS (CHAP. 8 Example 112: Consider the operation of intersection of sets. Then U, the universal set, identity element, since for every set A (which is a subset of U), Uta = AtU= A, that Una = AnU =A Example 113: Consider the operation of multiplication of real numbers. Then the number 1 is an identity element since, for every real number a, lta = art = a, that is leas al=a ‘Theorem 8.1: If an operation a:4 x A> A_ has an identity element ee A, then it is the only identity element. ‘Thus we can speak of the identity element for an operation instead of an identity element. INVERSE ELEMENTS Let «:AxA >A be an operation written (a,b) = a+b, and let ecA be the identity element for «. Then the inverse of an element ac A, denoted by oo is an element in A with the following property: atea=arat=e Example 121: Consider the operation of addition of real numbers for which 0 is the identity ele- ment, Then, for any real number a, ite negative (a) is ite additive inverse since a=at-a=0, thats, (ata =at(-a =0 Example 122: Consider the operation of multiplication of rational numbers, for which 1 is the lentity clement. Then for any non-zero rational number p/a, where p and ¢ are integers, its reciprocal q/p is its multiplicative inverse, since (alpnola = (wfavarn) = 1 Example 123: Let a:N x NN be the operation of multiplication for which 1 is the identity clement; here N is the set of natural numbers. Then 2 has no multiplicative inverse, Since there is no element 2. withthe property epsae st In fact, no clement in N has a multiplicative inverse except 1 which has itself as an inverse OPERATIONS AND SUBSETS Consider an operation «:A Xx A> A and a subset B of A. Then B is said to be closed under the operation of « if, for every ,b’ eB, a(b, bi) eB that is, if (Bx B)CB Example 131: Consider the operation of addition of natural numbers. ‘Then the set of even num- bers is closed under the operation of addition since the sum of any two even numbers is always even. Moreover, the set of odd numbers is not closed under ‘the operation of addition since the sum of two odd numbers is not odd. i are closed under the operation of multi- Example 132: The four complex numbers 1,—1, plication.CHAP. 8] FURTHER THEORY OF FUNCTIONS, OPERATIONS 128 Solved Problems DIAGRAMS AND FUNCTIONS 1. In the adjoining diagram of functions, how many paths B are there from A to H and what are they? L | £ Solution: A "There are six pathe from A to B: as ct se ASBoE ADdCSE > i ASBSooE AdobDSE b ASB-SCLDSE ASDSE rof, soicf, tojeies, eh, tejoh, tog ‘As noted previously, the functions are written from right to left. 2 Suppose the adjoining diagram is commutative. Here 4—___&__v4 1a is the identity function on A. State all the informa- tion that is inferred from the diagram, 7 o Sol 8 First, since the diagram is commutative, g°f = 1. Furthermore, since g°/ is one-one, f must also be one-one; and since g°f is onto, g must also be onto, Tt need not be true that g = f-", since we do not know whether f°g = ly SET FUNCTIONS & Let (12,8), and let f:W> V be defined by the adjoining diagram. Find: (1) f({a,b,d)), @) flta.e)). Solution: a (0) Compute as follows: = Hla, b, dB) = Ge), 0), 1} = 3,8) = 12,3) | (2) Similarly, Na, ed) = Sa), fo = (2,2) = (2) Note that f({a,c}) = 2 is an incorrect statement since the image of the set (a,¢}, in this situa- tion, is a subset of V, Le, (2), and not an element in V. 4. Prove: Let f:A>B be one-one. Then the induced set function f:24>2" is also one-one. Here 2* and 2° are the power sets of A and B respectively. SNUCK and ¥ be two diferent subsets of A, iy em, Ver, X4¥ ‘Then there exists an clement z¢A with the property that zeX, 2¢¥ (or ze¥, 2¢X) Thus /(2)e/X) and, since f is oneone, fla) ¢{(¥) (or fla)ef(¥) and fla) ¢ f(X)). Hence J) #10) and, by definition, the induced set function ie also one-one, 5. Prove: Let f:A-B be onto. Then the induced set function /:24>2" is also an onto funetion. Solution: It ig necessary to show that each set in 2° is the image of at least one set in 2, Let Ye2" Since / is onto, $20) = e | te, flee ¥) is not empty, But ¥ ia the image of f-"(¥), i. ff-'(¥)) = ¥. Hence f:24+2* is onto,124 FURTHER THEORY OF FUNCTIONS, OPERATIONS [cHaP.8 REAL-VALUED FUNCTIONS 6 Let W = {a,b,c}, and let f and g be the following real-valued functions on W: f(a) = 1, fb) = -2, fle) g(a) = —2, g(b) = 0, g(e) Find each of the following functions: (1) f+ 2g, (2) fa — 2f. Solution: (2) Compute as follows: + 20\(0) = fle) +20(a) = 1-4 = ~% (+ 20N6) = 16) +290) = -2+0 = UF 2oNe) = fe) +20) = 342=5 Thus /+29 = (@,-B, (6-2), (5). (2) Similarly, Ug = 21) = fla) g(@) ~ 27(@) = (4y(-2) ~ 20) Ua ~ 2/0) = #00) o(b) ~ 2710) Ua 2Ne) = fe 96) ~ 2fe) Henee fg ~2f = {(a, ~4), (2,4), (e, ~8))- (-2)(0) = 2-2) (y(t) ~ 218) 7. Let f be the real-valued funetion with domain [-3,3] which is displayed on the fol- lowing coordinate diagram: . Plot and describe the graph of each of the following functions: (1) f+2, (2) [fl Solution: (1) Since, by definition, +22) = f(z) +2, each image value of the original function is increased by 2.’ Hence raise the entire graph of f two units in order to obtain the graph of f+2, as shown in Fig. 8:1 below. +2 in displayed U/l is displayed Fig. 6-1 Fig. 8-2 oe ae tin = wa = {49 fers Hence part of the graph of |/| is identical with that part of the graph of f which appears above ‘the zvaxis; and the remainder of the graph of |/| is the reflection, over the x-axis, of that portion of the graph of f which appears below the z-axis. See Fig. 8-2 above,CHAP. 8] FURTHER THEORY OF FUNCTIONS, OPERATIONS 125 8 Find the domain of each of the following real-valued functions: (1) file) = 1/2 where x>0 (8) fa(z) = log (2-1) = @) fal) = y38=2 (4) fale) = 2 where 0<2<4 lution: (1) ‘The domain is explicitly given as (2 | 2 > 0). @) No domain is explicitly given, so we apply the rule of the maximum domain. Since /s assumes real values only when 3—z™ 0, that is, when £8, the domain of fs is (z | z= 8). (3) No domain is explicitly given, so we again apply the rule of the maximum domain. Note that the logarithmic function is defined only on positive numbers. Hence fy has meaning only if 2-1>0, he. ifz>4, Thus the domain of fs is (2 | #> 1). (4) ‘The domain is explicitly given as (z | 0 9 ‘The real-valued function 0:4 > R* which is defined by Oa(z)=0 for every eA is called the zero function (on A). Prove that for any function ()f+0n=f and (2) 7-04 = 0% =R, Solution: (1) Since + 0,)(2) = Ha) +0,(2) = fle) +0 = fle) for every 2eA, f+0, = (2) Also, (f+0,)(2) = fl2)+04(2) = fle) +0 = 0 = O,(2) for every zeA. Hence f+0, = Ox [Note that the zero function has properties which are very similar to properties of the number 0. 10. Consider the real-valued function f = ((1,2), (2, -8), (8,—1)} (with domain (1, 2, 3)). Find (1) f+4, (2) |fl, @) fF Solution: (1) Since, by definition, (f+ 42) = f(z)+4, add 4 to each of the image values, ie. add 4 to the second element of each ordered pair belonging to f. Hence f+4 = G8, &D, G9) (2) Since f(z) = |f(2)|, substitute the absolute value of the second element for the second element Jn each ordered pair in f. Accordingly, Mi = 2, @9, @D) (8) Since f(z) = (fla), substitute the square of the second element for the second element in each ‘ordered pair in f. Consequently, F = (0,8, @9), GD) MISCELLANEOUS PROBLEMS ON FUNCTIONS 11. Prove: Let A and B be subsets of a universal set U. Then x,y = X4X9" vn ee ANB; thetfore eA and #2. Then Xanel®) = Ge4x,V@) = x42) xy(@) = QM) = 1 Let ye (ANB; thus ye A'UB'; and hence ved’ or veB' Then rang (W) = 0. Also, (xqxp)v) = x, xq(u) =O since x,(v) =O oF xg(v) = 0. Thus xyrp ANd x42, assign the same number to each element in U. Hence by definition, Kane = Xo 12. Consider the function: j(z)=2 where z=0. State whether or not each of the follow- ing functions is an extension of f. (1) gi(z) =a where o=-2 (8) 1:R*> RY (6) galz) =a where ee [-1,1) 2) 92(2) = [2| for all zeR* (4) galz) = (@+ 2/2 Solution: Note that « function /" is an extension of / if, first, the domain of {' is a superset of [0, =), the domain of f, and, secondly, if fe) == for every ze'[0,©).126 13. 4 FURTHER THEORY OF FUNCTIONS, OPERATIONS (omar. 8 (1). Since gs satisfies both of the above conditions, g: is an extension of f. 2 if ze [0,«) —piteco* ‘the absolute value function is an extension of f. (8) By definition of an identity fonction, (2) = # for every #e R#, Thus the identity function is an extension of f. = etapa = [tee © Since ge) = cet lene = {63 (8) ‘The domain of gs is not a superset of the domain of f; hence g. is not an extension of f. @) Since one) = [el = { 2 if ze(0,©) Oe ts os is an extension of f. As sets, what is the relationship between a function f:A>B and the restriction of f to a subset A’ of A? Solution: The striction of f to A’, ie. f|A’, ie a subset of f. For #eA! implies 2eA and, hence, (fla) © fA’ implies, f(@)) © f Consider the subsets A: = (1,2,3), Az = (1,5}, Ay = (2,4,5} and Ay = (8,4) of B= (1,2,3,4,5). State whether or not each of the following functions of (Ai, As, As, Aa} into B is a choice function. Q ft (2) fe @) fr @) fe Solution: (1) Since f(A,) =2 is not a member of As, fs Is not a choice function. (2) Note that f,(A,) belongs to Ai, for each i; hence fs is» choice function. (8) Also, fs(A.) © Ay for each é; hence fe is a choice function. (4) Note that (As) =1 does not belong to As; thus fis not a choice funetion. (An 2), (Aa, 2), (As, 8), (Ae, 4)} {(As, 1), (Aa, 1), (As, 4), (Ae, 4)} {(As, 2), (As, 1), (As, 4), (Ae 3)} (As, 8), (As, 5), (As, 1), (As, 3)} OPERATIONS 5. 16. Let a:NxN > N be the operation of least common multiple (Lem), i.e, e(a,b) = a+b = Lem. of aandb (1) Is.« commutative? (2) Is « associative? (8) Find the identity element of «. (4) Which elements in N, if any, have inverses and what are they? Solution: (1) Since the Lem. of a and d is the Lem. of b and a, « is commutative. @) It is proven in number theory that (a+) +e = a+(b 0), ie that the operation of Lem. is associative. (8) The number 1 is an identity clement since the Lem. of 1 and any number a is a, that is, 1+a =o for every aeN. (4) Sinco the Lem. of two numbers @ and is 1 if and only if @=1 and 6=1, the only number ‘which has an inverse is 1 and itis its own inverse, Consider the operation «:Qx QQ which is denoted and defined by (a,b) = aeb = atb—ab Here Q is the set of rational numbers. (1) Is « commutative? (2) Is « associative? (8) Find the identity element for a. (4) Do any of the elements in @ have an inverse and what is it? Solution: a ath = atb—abCHAP. 8] FURTHER THEORY OF FUNCTIONS, OPERATIONS 127 @ (aedjee = (@tb—ad)te = (atb—ab)+e-(atb—adjo (e+ b— ad) +o— (at b—adje = atb—ab+e~ae~be + abe at b+e~ab—ae~ bet abe as(bee) = aebte—bo) at (b+ e— be) — alb + 6— be) = atb+e~be—ab—ac+ abe Hence a is associative (8) An clement « follows: fan identity element for « if a*e =a for every clement ae@. Compute as ate =a atenae Accordingly, 0 is the identity clement. , emer = 0, ella) = 0, © (4) Im order for a to have an inverse x, we must have az = 0, since, by (3), 0 is the identity element. Compute as follows: ate = 0, ate-ar = 0, a= ar—z, a = s(a—1), 2 = alla—1) ‘Thus if a1, then a has an inverse and it is a/(a—1). 17, Prove Theorem 8.1: If ¢ and ¢’ are identity elements (for the same operation), then ene. Solution: By hypothesis, +e! and eve'=e. Thus e 18. Consider the operation of union of sets. (1) Find the identity element. (2) Which elements, if any, have inverses and what are they? Solution: (1) Note that AUQ = QUA =A. for any et A. Hence the null set @ is the identity element for the operation of union of sets. (2) Im order for a set A to have an inverse X, AUX =. Since AUX = Q implies A=9 and X=, the only set which has an inverse is tho empty set and it is its own inverse. 19, Let the operation «:Ax A> A, denoted by a(a,b) = aed be associative and have an identity element e. If b and b’ are inverses of the same element a, then b=’. (In other words, inverses are unique.) Proof: ‘Note that bs (aed) raee bee eee Since « is assoc beard) = beaey and therefore 20. Let J, be the set of all functions of A into A. Let « be the operation of composition of functions, (1) Is « commutative? (2) Is « associative? (3) Find the identity element for a. (4) Which elements, if any, have an inverse and what is it? Solution: () If A has more than one element, then is not commutative. For let aeA, beB and a6, and consider the constant functions f and g defined by f(z) =<, o(z) = 6. Then Geo) = foe) = 10) = « ofa) = oe) = ga) = b128 FURTHER THEORY OF FUNCTIONS, OPERATIONS (CHAP. 8 (2) By Theorem 41, « is an associative operation. (3) The identity function 1,:A +A is an identity clement since, as previously noted, (1,°/) = (Fo1,)) =f for any function fe J,. (4) The function f:A-+A has an inverse if and only if f is one-one and onto, and ite inverse is the inverse function f-! which was defined in Chapter 4, 21, Let the operation «:N x N+ N be defined by a(a,b) = aeb =a (1) Is a commutative? (2) Is « associative? (8) Is there an identity element? (4) Does any of the elements have an inverse and what is it? Solution: (1) Since a+b =a and bea =, a is not commutative, (2) Since (a+d)¥0=aee=a and av(bwe) = ab =a, « is associative: (3) Tf a has an identity element ¢ then, by definition of identity clement, e*a =a for every aeN. But by the definition of a, e+a —e. Hence there ia no identity element (4) It is meaningless to talk about an inverse when there does not exist an identity element. OPERATIONS AND SUBSETS 22. State whether or not each of the following subsets of N, the natural numbers, is closed under the operation of multiplication. (1) (1) (4) (1,8,5,7,...) = (| # is odd) @) 4,2) (5) (| x is prime} (3) (2,4,6,8,...) = (#| 2 is even) (6) {2,4,8,16,...) = (2 | #=2", neN} Soluti w (00) = 0, (40) = 0, (04) = 0, YA) = 4 Hence (0,1) is closed under multiplication, (@) Since (2)(2) = 4 ¢ (1,2), (1,2) is not closed under multiplication. (@) The product of even numbers is even; hence the set ja closed under multiplication. (4) The product of odd numbers is odd; hence this set is closed under multiplication, (8) Note that 2 and 3 are primes, but (2)(3)=6 is not a prime, Hence the set is not closed under ‘multiplication. (8) Since (29(2") = 2%, the set is closed under the operation of multiplication. 23, State whether or not each of the sets in the preceding problem is closed under the operation of addition. Solution: ‘The set of even numbers, (2,4,6,8,...}, is closed under addition since the sum of two even ‘numbers is even. But each of the other sets are not closed under addition since, for example, 141 =2¢ (0,1) 142 = 8 ¢ 4,2) B45 = 8 e (,3,5,7,..) 845 = 8 ¢ (| © is prime) 2+4= 6 ¥ (2,4,8,16,...}CHAP. 8) FURTHER THEORY OF FUNCTIONS, OPERATIONS 129 2%. Let (1) of be the family of finite sets of real numbers (2) B be the family of intervals (8) C be the family of supersets of the unit interval (0, 1). State whether or not each of the above families of real numbers is closed under the operation of (a) union, (b) intersection. Sotution: (2) Since the union and intersection of finite sets are finite, of is closed under both operations, (@) Since [1,2] U [8,4] is not an interval, is not closed under the operation of union. As shown in Chapter 3, the intersection of two intervals is an interval; hence 3 is closed under the operation of intersection. (@) Hf (0,1) CA and (0,1]CB, then [0,1] C (AUB) and [0,1] C(ANB). Accordingly, C is closed ‘under both operations. 25. Prove: Let «:A XA > A be an associative operation with an identity element ¢, and let B be the set of units in A, that is, the set of elements of A each of which has an inverse. ‘Then B is closed under the operation of «. Solution: Let ae B and beB. Then a has an inverse a~! and 6 has an inverse b-!. We must show that a,b) = atbeB that is, a # 6 has an inverse. Note that. (oe Oa ‘Thus @*b has an inverse b-!+ a", and therefore a *b eB. Supplementary Problems DIAGRAMS AND FUNCTIONS 28, In the diagram of functions, shown in Fig. 8-3 below, how many paths are there from A to D and what are they? I=! BA a—ts—*# ce Fig. 83 21. If the diagram shown in Fig, 8-4 above is commutative, which functions are equal? SET FUNCTIONS 28. Let W = (1,2,3,4,5) and let f:W> W be the following set of ordered pairs: £ = (4,9), 22, 5), 4,8) 6, D> Find: (1) 4,2,3), @) 9), (8) 2,8).180 FURTHER THEORY OF FUNCTIONS, OPERATIONS [cHar.s 1%, Let S = (a,b,c), T 44,2, 9) and . £ = (@,1), (6,8), I) Find the induced function f:2°-+ 2", REAL-VALUED FUNCTIONS 30. Let V = {a,b,c} and let f and g be the following real valued functions on V: F = (2, (b,-8) (Dh 9 = Ma, -2), (6,0), (0) Find: (1) 8, @) 9 +2 (8) f+, (4) 2-50, ©) fo, IM, DF, (8) [8f~ fa} 31. Find the domain of each of the following real-valued functions: () fe) = aKe+9) @) fe) = Va=e (2) He) = afe+8) wherez>0 (a) fle) = log te) 32, Let f be the real-valued function with domain [4,4] which is displayed on the following coordinate diagram: TF ay amas Sketch each of the following fonctions on a coordinate diagram: (1) f—3, (2) {fl CHARACTERISTIC FUNCTIONS 88, Let U = (a,b,¢,d,e}, and let A= (a,b,e}, B= {ed} and C= (ade). Find: ) xa (2) xe @) xe B4. Let_U = (a,b, 6d). Each of the following functions of U into (1,0) is a characteristic function of a subset of U. Find each subset, (1) a1), 6,0), (6.0, 41} 8) (a, 0), (0,0), (6,0), 2,0} 2) G0, (1, (0, GO} 4) (1), 1, (6, 0), (DY 85, If the characteristic fonction x, is a constant function, what ean one say about the set A? MISCELLANEOUS PROBLEMS ON FUNCTIONS 36, Consider the function J = (01,2), (8,5), 4, 6), 3) whose domain is {1,8,4,8}. In each of the following functions = and y appear as unknowns. For which values of 2 and y, if any, will each function be an extension of /? Qf = (2, 24, G2), 4,6), 6,8), @ W) @) fe = (2), @9, 29, 8), B,5) @) fs = 4,6), (@, 9), 8,5), (0), (1,2) fons can the restriction of a characteristic function x, to a set B, ie. x,|B, be a 31. Under what con constant function’CHAP. 8] FURTHER THEORY OF FUNCTIONS, OPERATIONS 181, 38, Consider the subsets Ai = {a,0,d), As = (e,d,e}, Ar = (0) of B for not each of the following functions of (As,As,As) into B (1) fe = (Ava), (AaB), (Ass OD @) fe = (Aya, (And), (As, (8) fe = (And), (Arse), (Ass (a,b,0d,e}. State whether '8 choice function. OPERATIONS. 39, Let a be the operation of intersection of sets, (1) Is « commutative? (2) Is « associative? (8) Find the identity element for a. (4) Which elements, if any, have inverses and what are they? 40, Let a be the operation of difference of sets. Note that A-B=AnB (1) Is @ commutative? (2) Is « associative (3) Show, by Venn diagrams, that the operation of union does not distribute over a, ie. in general, AUB-O # (AUB)—(AUQ) (4) Prove that the operation of intersection does distribute over «i.e. An(B-O) = (ANB)- (Ano) 41, Let e be the operation on sets defined (and denoted) by AMB = (AUB)~(ANB) ‘This operation i called the symmetric difference. (1) Is « com- mutative? (2) Ise associative? (8) Find the identity clement. (4) Find the inverse of an arbitrary set A. (6) Show by Venn diagrams that the operation of union does not distribute over «, ie. in general, AU) # (AUB)A(AUC) (6) Prove that the operation of intersection does distribute over a hey ‘AB is shaded An(Ba0) = (ANB) (ANC) 42, Let a be the operation on @X Q, the set of ordered pairs of rational numbers, defined (and denoted) by (a, 0) * (2) = (ax, ay+b) (a) Ise commutative? (2) Is « associative? (8) Find the identity element for a. (4) Which elements, if any, have inverses and what are they? 43, Let a be the operation on the real numbers defined (and denoted) by ash = atb+2ab (1) Is @ commutative? (2) Is @ associative? (3) Find the identity element for «. (4) Which ele- ments, if any, have inverses and what are they? OPERATIONS AND SUBSETS A, Lat B= (...,—4,-20,2,4, 0)» ie. the even integers. State whether or not Eis closed under the ‘operation of (1) addition, (2) subtraction, (2) multiplication, (4) division (except by zero). 45. Let F = (45-5, 8,1, 1,2,5, «..), ie. the odd integers. State whether or not F is closed under the operation of (1) addition, (2) subtraction, (3) multiplication, (4) division (except by 2¢r0). 46, ‘Let & be the family of all bounded sets of real numbers. State whether or not is closed under the ‘operation of (1) union, (2) intersection, (2) difference. 41. Let of be the family of all open-closed intervals (a, 6], together with the mull set, State whether or not of is closed under the operation of (1) union, (2) intersection, (8) difference. 48. Consider the family of of sets of real numbers which are supersets of (0}, ie. Acof if and only if Oecd. State whether or not of is closed under the operation of (1) union, (2) intersection, (8) difference,182 FURTHER THEORY OF FUNCTIONS, OPERATIONS [CHAP. 8 Answers to Supplementary Problems 2%, Three: ief, jog and johoy. = 80f, r= noe, ref = wonof = uct, w= vouct = vouseef = vorey, 8. (1) 82,5), 2 , @) (1) 2. Ma, b,€}) = (1,8), Hla, 0) = (1,3), feed) = 1}, (0) =, H{b,e}) = (1,3), fla) = @), fle) = 4) ou, wee=vor=vowrs, and wet=woaef = (a, 6), (6, ~9), (e, -8)) ©) fo 4), (6, 0), ( —1)) {(a, 0, (, 2), (3)) (8) Wl = (0,2), (6,3), (1) {+0 = (2,0), 9, (0) OF = (8), (b,-20, ( —1)) @ 2-50 = (@,1, (6-6), (—M) (8) 18/—Fa] = {(a, 10), (6,9), (29) B (1) @ | ee Re eB) @) (| 2>0) (8) fe | -B=2=2) A) fe | zeRe, 2 40} @ uw 38.) xx = CD, (0, (6 0), @,0), (6, D? 2) xe = {(a,0), (6,0), (¢, 1), (2, 2), (€, 0) xe = Ka, 2), (0), (6, 0), (@, 0, (e, 1) 4) fd, @ O, @ 9 & fda 85, Either A= or A=U. 36. (1) 2=5,y=8 @) 2=8 y=4 (8) 2=8, y can be any clement 37. Either B is a subset of A, or B is a subset of the complement of A. 38. (1) No, (2) Yes, (8) Yes 39. (1) Yes (2) Yes (8) U, the universal set, is the identity clement, (4) Only U has an inverse which is itself. 4. (1) No, (2) No ® AU(B=O) is shaded (AUB) ~ (AUC) is shadedCHAP. 8} a. FURTHER THEORY OF FUNCTIONS, OPERATIONS 133 o Statement Reason 1, (ANB) = (Ane) = (AnB)n(Ancy 1, Definition of difference 2 (AnBynia'uc) 2. De Morgan's Law 5. = [AnBnaUlAnBync] a wutive Law 4. But (ANB)na’ = (And)OB 4. Associative Law, Commutative Law 8. OnB 5, Complement Law 6. 9 6. Identity Law 1. 2. (ANB) ~ (Ane) = BullAnByncy 7. Substitution a (anBync 8. Identity Law a Anenc} 9. Associative Law 10, An(B-0) 10, Definition of difference (1) Yes, (2) Yes. (2) 9 is the identity element. (4) The inverse of any set is itself. 6) AU(BAC) is shaded (AUB) 4 (AUC) i shaded © Statement 1. (ANB) A(AN® = [ANB)UANG) — [ANB)NANG) 1. 2 But (AnB)U(An©) = An(Buo) 3. (ANB)N(ANC) = (An A)n(BNO) 8 4 = Anno, 4 5. (ANB) a (Ane) = [An(BUO] ~ [An(BnC)] 5 6. = An[(BUC) ~ BO) 6 1 = An(Bac) 1 (1) No. (2) Yes. (3) The ordered (1,0) ia the identity clement. hhas an inverse if a0, and its invers (fe, bie). Reason Definition 2. Distributive Law Associative Law, Commutative Law Tdempotent Law Substitution Intersection distributes over difference Definition (4) The ordered pair (a, ) (1) Yes, (2) Yes. (8) Zero is the identity clement. (4) If @1/2 then a has an inverse, and its inverse is —a/(1-+ 2a). (1) Yes (2) Yes (8) Yes (4) No (1) No (@) No (3) Yes (4) No () Yes @) Yes (8) Yes (1) No (2) Yes (8) Yes (1) Yee (2) Yes (8) NoPart IT — Cardinals, Ordinals, Transfinite Induction | Chapter 9 Cardinal Numbers EQUIVALENT SETS It is natural to ask whether or not any two sets have the same number of elements. For finite sets the answer can be found by counting the number of elements in each set. For infinite sets the answer depends upon how one defines two sets to have the same number of elements, that is, as we will say, to be equivalent. At one time all infinite sets were considered to be equivalent. The following definition, which has revolutionized the entire theory of sets, is attributed to the German mathematician Georg Cantor (1845-1918). Definition 9.1: | Set A is equivalent to set B, denoted by A~B if there exists a function fAoB which is both one-one and onto. The function f is then said to define a one-to-one correspondence between the sets A and B. Let R= (1,2,5,8) and T = (Mare, Brik, Paul, Betty). The following diagram defines a function of R into T which ie both one-one and onto, Hence Fis equiva Tent to 7 Example 12: Let M = {1,2,8) and N = (1,2). If we list all the functions of M into N, none cf them will be both one-one and onto. Hence Af is not equivalent to NV. If we examine the above two examples, it is not difficult to see that, in general, two finite sets are equivalent if and only if they contain the same number of elements. Hence, for finite sets, Definition 9.1 corresponds to the usual meaning of two sets containing the same number of elements. Example 13: Let G = (0,1) and H = [2,5], and let /:@- H be the function defined by fle) Note that / is both one-one and onto, Hence @ ~ H, Le. G is equivalent to H. Example 14: Let N= (1,2,3,...} and B= (2,4,6,...). ‘The function f:N + E, defined by f(z) = 22, 4s both one-one and onto. “Therefore N ~ E, In Example 1.4 we see that the infinite set N, the natural numbers, is equivalent to a proper subset of itself. This property is characteristic of infinite sets. In fact, we now state formally e+e 134CHAP. 9] CARDINAL NUMBERS 185 Defi m 9.2:] A set is infinite if it is equivalent to a proper subset of itself. Otherwise a set is finite. Example 15: Let A and B be any two sets. ‘Then A~Ax(} B~ Bx) since the functions fra+@2, acd 9:6 (2), beB tre both one-one and onto, Moreover, although A and B need not be disjoint, note that Ax} 0 BX = 0 since each ordered pair in A x (1) contains 1 a8 a second element, and each or- dered pair in B X {2} contains 2 as a second clement. We conclude this section with a theorem which we shall use later in the chapter. ‘Theorem 9.1: The relation in sets defined by A ~B, is an equivalence relation. Specifically, (1) A~A for any set A, (2) if A~B, then B~ A, (8) if A~B and B~C, then A~C. DENUMERABLE SETS The reader is familiar with the set of natural numbers, N = (1,2,8,. Definition 9.3:] If a set D is equivalent to N, the set of natural numbers, then D is called denumerable and is said to have cardinality a. Definition 9.4:] A set is called countable if it is finite or denumerable, and a set is called non-denumerable if it is infinite and if it is not equivalent to N, ie. if it is not countable. | Example 21: Any infinite sequence of distinet elements is denumerable, for a sequence is essentially @ function Jin) = a ‘whose domain is N. So if the a, are distinct, the function is one-one and onto. Hence each of the following sete is denumerable: (1, 1/2, 118) «ey Wty.) (1, 2 8, 420, (Atm, 2} (G1), 4,8), (2D, GP, -) Example 22: Consider the product set NX N as exhibited in Fi aoe GQ) 270.8) dae. Qn 22)” 23 ea”. Ze Ze a \* en 2 3” 3,0" ee 1 4,2) 4, 8) 4,49 77 Fig. ‘The set NX N can be written in an infinite soquence of distinct elements as follows: AG, 1), (2,1), (1,2), (Ly 8, (2), ...) (Note that the sequence is determined by “following the arrows” in Fig. 9-1.) Thus, for reasons stated in Example 2.1, NX N is denumerable. —136 Example 23: Let Mt CARDINAL NUMBERS [cHAP.9 (0,1, 2, ...) = NU{O}. Now each natural number ae N ean be written ‘uniquely in the form. a = 22641) where r,©M. Consider the function f:N > MX M defined by fla) = (r.9) where rand s are as above. Then / is one-one and onto, Hence MX M is denu- merable. Note that NN is a subset of XM. ‘The following theorems concern denumerable sets. ‘Theorem 9.2: ‘Theorem 9.3: Corollary 9.3: ‘Theorem 9.4: Corollary 94: Every infinite set contains a subset which is denumerable. A subset of a denumerable set is either finite or denumerable. A subset of a countable set is countable, Let As, As, As, ... be a denumerable family of pairwise disjoint sets, each of which is denumerable. Then the union of the sets Yrew Ay is denumerable. Let {A)},,, be a countable family of countable sets. Then U,,,A, is countable. ‘Next follows a very important, and not entirely obvious, example of a denumerable set. Example 24: Let Q* be the set of positive rational numbers and let Q~ be the set of negative rational numbers. ‘Then @ = Qu{ug* is the set of rational numbers. Let the function :@* > NX N be defined by Kola) = (a) where p/q is any member of Q* expressed at the ratio of two relatively prime positive integers. Note / is ono-one, and therefore Q* is equivalent to a subsct of 22, Q° is denumerable. Similarly Q- is, Hence the set of rational numbers, which is the union of Q°, {0 jenumerable, THE CONTINUUM Not every infinite set is denumerable. The next theorem gives a specific and extremely important example. Theorem 9.5: The unit interval (0,1] is non-denumerable. Defi Example 31: Let [a,)] be any closed interval ‘Two proofs of this theorem appear in the section on solved problems. Let a set A be equivalent to the interval [0,1]. Then A is said to have cardinality ¢ and to have the power of the continuum, nd let 4:10.) > [0,0] be the function defined by f@) = a+ b-ae Note that f is one-one and onto. ‘Thus (a, ] has cardinality ¢. Furthermore, we will prove that any open or half-open interval also has cardinality ¢CHAP. 9] CARDINAL NUMBERS 137 Example 82: The function :(~#/2,+/2) + R®, defined by f(a) = tanz, is one-one and onto; hence RE ~ (-e/2, 2/2) ‘Therefore the set of real numbers R# has the power of the continuum, ie, has cardinality CARDINAL NUMBERS Note once again, by Theorem 9.1, that the relation in sets defined by A~B is an equivalence relation. Hence by the Fundamental Theorem on Equivalence Relations, all sets are partitioned into disjoint classes of equivalent sets. Definition 9.6:] Let A be any set and let « denote the family of sets which are equivalent to A. Then « is called a cardinal number (or, simply, cardinal) and is denoted by «= #4) Definition 9.7:] The cardinal number of each of the sets D, (A), (1,2), (1, 2,8}, is denoted by 0,1,2,8,..., respectively, and is called a finite cardinal. lt Definition 9.8: |The cardinal numbers of N, the set of natural numbers, and the unit in- terval [0,1] are denoted by #W) = 0, #((0,1)) Remark 9.1: The symbol X, (read aleph-null) is also used to denote the cardinality of denumerable sets, i.e. #(N), since this is the symbol originally used by Cantor. CARDINAL ARITHMETIC In view of Definition 9.7, the cardinal numbers can be considered to be a superset of the finite cardinals ... that is, the natural numbers N and 0. The next definition essentially extends the ordinary operations of addition and multiplication for natural numbers to operations for all the cardinal numbers. Definition 9. Let « and f be cardinal numbers and let A and B be disjoint sets such that a= #(A), p= #8) Then ath ap ‘Theorem 96: Definition 9.9 is well-defined, that is, the definitions of a+ and af do not depend upon the particular sets A and B. In other words, if A~ A’, B~ BY, ANB=Q, A’nB = 0 #(AUB) #(AXB) then #(AUB) = #(4UB) #AXB) = #(A'xB)138 CARDINAL NUMBERS [omar. 9 The sets A and B are assumed to be disjoint in Definition 9.9. Since the sets A x (1} and B x (2) are always disjoint, regardless of A and B, the following definition can be substituted for Definition 9.9. Definition 99: | Let «= #(4) and p= #(B), then atp = #Ax() U Bx (2) a8 = #(A xB) Example 41: Note that 3 = 4((a, 2, }) and 4 = #((1,3,5,7)). ‘Then B44 = #((a,b,6,1,3,5,7)) = (4%) = Ula, b,6) X'C1,8,5,7)) = 12 In other words, the operations of addition and multiplication of finite cardinal numbers correspond to the ordinary operations of addition and multiplication of natural numbers. Example 42: Note that a = #((1, 8,5, ...}) = #((2,4,6,...)). Then ata=#(M) =a and aa = #VXN) =a ‘Theorem 9.7: The operations of addition and multiplication of cardinal numbers is associative and commutative; and addition distributes over multiplication, ie. for any cardinal numbers «, and y, Q) (+A) ty = at (B+y) (2) (e8)y = a(By) @) et+B = fta (a) of = fa (5) a(B+y) = of + ay Not every property of addition and multiplication of natural numbers holds for cardinal numbers in general. For example, for the natural numbers, the cancellation law is true, i.e., a+b =a+te implies b= ab = ac implies b=e Since, by Example 4.2, ata 1+a does not imply a=1 ea Ja does not imply a= 1 ‘ion law is not true for the operations of addition and multiplication of cardinal Remark 92: Exponents can also be introduced into the arithmetic of cardinal numbers as follows: Let «= #(A) and = #(B), and let BA denote the family of all functions from A (the exponent) into B. ‘Then pr = #(B4) In fact, the following properties of exponents, which are known to hold for the natural numbers, also hold for any cardinals o, 6 and »: () ar = aftr @) yr = @) (ay = wer Example 43: Let A = (0,5, 0) and B = (1,2). Then #(4)= 8, #(8) =2 and = 2B, But B* consists of exactly 8 functions: {@,0,0,D, (603, 6,29, 6,19, (6,29, (2), 6,2), 1), 0, (6,2) (627 {.2), (0,2, (8, {C2 (6,1), (2)), C2, (2), (60), (a.2), (6,2), -2))CHAP. 9} CARDINAL NUMBERS 139 ‘Therefore, as cardinals, was In other words, if m and m are finite cardinals then m* denotes the same number regardless of whether we consider m and m to be cardinals or natural numbers, INEQUALITIES AND CARDINAL NUMBERS An inequality relation is defined for the cardinal numbers as follows: [Definition 9.0:] Let @ = #(A) and g = #(B), Furthermore, let A be equivalent to a subset of B; that is, suppose there exists a function f:A > B which is one-one. ‘Then we write ASB which reads “A precedes B”, and e=

A is one-one; hence AS A. Also, ‘therefore, « = @ for any cardinal namber e Example 54: If {:A~ B is oneone and 9:B~ C is one-one, then the composition function °F: C is alao oneone, Therefore, ASB and BSC implies A (0,1) defined by the following diagram In other words, 2 if 2-0 fe) {ies if =m, men =e if 2 0,1/n, neN ‘The function f is one-one and onto. Consequently, [0,1] ~ (0,1). (#) The function f:[0,1] > (0,1) defined by fey = {une fs one-one and onto. (If is similar to the function in Part (a)). Hence (0,1) ~ (0,1). (©) Let (0,1) ~ (0,1] be the function defined by f(2) = 1s. Then f is one-one and onto and, therefore, [0,1) ~ (0,1}. By Part (0) and Theorem 9.4, [0,1] ~ (0,1). 2a Un, nen ze in, ne 3. Prove that each of the following intervals has the power of the continuum, ie. has eardi- nality c: (1) [a, 6], (2) (a,b), (8) (a,b), (4) (a,b). Herea f08) Sf.) .->@8 0.1 Cd) Each function is one-one and onto. Hence, by Problem 2 and ‘Theorem 9.1, each interval is equivalent to [0}, that is has the power of the condinuom.142, CARDINAL NUMBERS [cHAP.9 4. Prove Theorem 9.1: The relation A~B in sets is an equivalence relation. Specifically, (1) A~A for any set 4, (2) ifA~B then B~ A, (@) ifA~B and B~C then A~C. Solution: (1) The identity fonction 1,:A + A is one-one and onto; hence A~ A. @) If A~B, then there exists a function /:A + B which is one-one and onto, Hence f has an inverse function -*:B-> A’ which i also one-one and onto, ‘Therefore, A~B implies B~ A (@) If A~ B and B~C, then there exists functions. f:A > B and g:B > C which are one-one and onto, Then the product function g®/:A ~C ia also one-one and onto, Therefore, A~B and B~C implies A~C 5. Prove Theorem 9.2: Every infinite set A contains a subset D which is denumerable. Solution: Let f:2~ A be a choice function. Consider the following sequence: a = f(a) a = f(A ~ (a) a = f(A ~ (a, a) HA fee acy Since A is infinite, A — (ai, ..., der) is not empty for any meN. Furthermore, since / is a choice funetion, a * aq where i BX A defined by M(a,b)) = (a), (aed, beB) fs one-one and onto; hence AXB ~ BX A. 7. Prove: For any sets A,B and C, (Ax B)x0 ~ AXBXC ~Ax(BXC) tbe funtion f2(AX 3) x0 + AX B XC datned by Ne,00) = (6,0, (eA, deBoe) is oneone and onto; hence (AX) XC ~ AX BX C. Similarly, Ax (BXO)~ AXBXC. Thus, (xB) xO ~ AXBKC ~ AXWXC) 8. Prove: Let X be any set and let C(X) be the family of characteristic functions of X, that is, the family of functions f:X > (1,0). Then the family of subsets of X is equivalent to O(X), ie. 2° ~ C(X). Soluti Let A be any subset of X, ie. Ae2% Let 7:2" C(X) be defined by BA) = xy that is, f maps each subset A of X into x,, the characteristic function of A (relative to X). Then f is one-one and, as noted previously, onto, Hence 2¥ ~ C(X).onar. 9) CARDINAL NUMBERS 143 9. Prove Theorem 9.3: A subset of a denumerable set is either finite or denumerable. Solution: Let A= fay an.) 0 bbe any denumerable set and let B be a subset of A. If B If B49, then let on, bbe the first element in the sequence in (1) such that ar, eB: let ay, be the first element which follows fan, in the sequence in (1) such that ane B; ete. ‘Then’ B= (any any.) If the set of integers (n,m, -..) is bounded, then B is finite. Otherwise B is denumerable. 10. Prove Theorem 9.5: The unit interval A Solution: ‘Method 1. Assume the contrary; then [0,1] is not denumerable. A= lentes. i.e, the elements of A can be written in a sequence. ‘ach element in A can be written in the form of an infinite decimal as follows: a = O.enan 2 = Oananan -. an o and, for those numbers which can be written in the form of a decimal in two ‘ways, for example, 2 = 5000. 14999, (in one of them there is an infinite number of nines and in the other all except a finite aet of digits are zeros), write the infinite decimal in which an infinite number of nines appear. [Now construct the real number Y= Obi babe by which will belong to A, in the following way: choose b, s0 Dir ay and b.%0, choose bs 80 bs aux and bx0, ete. Note yz: since b,#au (and 6:40), yz since bran (and by0), ete, that ia, y% zm for neN; hence y¢A, which contradicts the fact that yc. ‘Thus the assumption that A is denumerable hhas led to a contradiction. Consequently, A is non-denumerable ‘Method 2. (In this second proof of Theorem 9.5, we use the following property of the real numbers: Let I: = [a1,bslp Ts = [aeyby), -.- be a sequence of closed intervals for which 1,2/,3"+-. ‘Then there exists a real number y with the property that y belongs to every interval.) Assume the contrary. Then, as above, A= lente) Now construct a sequence of closed intervals Iy,ls,... a8 follows. Consider the following three losed sub-intervals of [0,1], [o, 1/3}, (1/3, 2/8), (2/8, 1) ® ‘each having length }. Now 21 cannot belong to all three intervals. (If 2; is one of the endpoints, ‘then it could belong to two of the intervals.) Let J, = [a,,bi] be one of the intervals in (1) such that 21 ¢, Now consider the following three closed sub-intervale of Is = [a bi), (ay, a+ 1/9], far +1/9, a+ 2/9}, [ar +2/9, by] ® ‘each having length . Similarly, let J, be one of the intervals in (2) with the property that an does not belong tos. Continue in this manner. ‘Thus we have a sequence of closed interval Ihde ® such that 2af 1, for every neN.144 CARDINAL NUMBERS (CHAP. 9 By the above property of real numbers, there exists a real number y ¢ (0,1) such that y belongs to every interval in (3). But since ved = (yay...) y=2_ for some meN. Then by our construction y = suf Im, which contradiets the fact that y belongs to every interval in (8). Thus our assumption that is denumerable has led to a contradiction, Accordingly, Ais non-denumerable, CARDINAL NUMBERS AND CARDINAL ARITHMETIC 11, Let Ai,As,As and As be any sets. Define sets B;, B:, By and By such that #(As) + #(Aa) + #(4s) + H(A) = #(B,UBLUBLUB,) Solution: Let By = Aix (1), Be = Asx (2), By = Asx (8) and By = Agx (4). Then Be~ Ay 1,23,4; and BB; = if ij. Consequently, the above will be true, 12. Let (4,),,, be any family of sets. Define a family of sets (B,},,, such that Bi~ Ay iel and BOB,=Q it ixf Solution: Let B= A,X (i), (ie). Then the family {B:);e has the required properties, 13. Prove Theorem 9.7: For any cardinal numbers a, and y, () (a+ p)+y = at+(B+y) (2) (aB)y = a(By) @) atB=Pta (4) 6) Solution: Let A,B and C be pairwise disjoint sets such that a= #(A), # = #(B) and y = #(C). a (+A) +7 = #(AUB) + (0) = #(AUB)UC) e+ (B47) = #4 + #BUC) = #AU@UO) Since union of sets is associative, ie, (AUB)UC = AU(BUC), then Gta ty = «+ @+y @ (ay = #AX BHC) = #04 xB) xO) oR) = #ARBXC) = HA x(BXO) By Problem 7, (AX B)XC ~ AX(BXC). Therefore (op) = a(2y). (9) «+8 = #(AUB) = #(BUA) = B +a, since AUB = BUA, (4) Note of = #(4 xB) and fa = #(BXA). By Problem 6, AXB~ BX A; hence of (8) Note frst that BNC = 0 implies (A x B)N(A XC) = 9. Then (B+ y) = #(A)+#(BUC) = #4 x (BUG) ab + ay = HAXB) + FAXC) = (AX B)UIAXC) But A x (BUC) = (AX B)U(A XC). Therefore al +7) = of tay. 14. Prove: a Solution: Let Z = (..)-1,0,1,...} and A = (0,1). Furthermore, let f:Z%A > RE be defined by Mua) = ita ‘im other words, /((i} X (0,1)) is mapped onto [i, (+1). Then f is one-one and onto. Hence (x A) ~ RE Since #(Z) =a, #(A) =¢ and #(R2} then acCHAP. 9] CARDINAL NUMBERS 145 15. Prove: Let # be any infinite cardinal number. Then otp=B Solution: Lat A be an infinite set, B = {bi,bs,...) be denumerable and ANB =. Then the theorem is true if we show that AUB~ A Since A is infinite, A contains a denumerable subset D = (dude...) Let f: AUB A be defined by the following diagram: AUB A In other words, f: AUB A is defined by 2 it zeA-D ftey fins itso dn if = be ‘Then f is one-one and onto. Consequently, AUB ~ A and our theorem is true, INEQUALITIES AND CARDINAL NUMBERS 16. Prove Cantor’s Theorem: For any set A, A<2* and hence #(A) < #(24). Solution: ‘The function g:A + 24 which sends each element ae A into the set consisting of a alone, ic. which is defined by g(a) = (a}, is one-one; hence A <2. If we now show that A is not equivalent to 2, then the theorem will follow. Suppose the contrary, ive, let there exist a function f:A > 2° which is one-one and onto. Let ac be called a “bad? clement if a is not a member of the set which is its image, ie. if ay f(a), and let B be the set of “bad” elements. Specifically, B= (@| 2eA, 2¢fle) Note B is a subset of A, that is, Be. Hence, since f:A ~ 2* is onto, there exists an clement, beA with the property that f()=B. Is b “bad” or “good”? If beB then, by definition of B, by f(b) = B, which is impossible. Likewise, if 6B, then be J(0) = B which is also impossible. ‘Thus the original assumption, that A~ 24, has led to a contradiction. Hence the assumption is false ‘and, therefore, the theorem is true. 17. Prove the Schréder-Bernstein Theorem: Let XD¥2X; and let X~Xu; then X~Y. Solution: Since X~ X,, there exista a function f:X > X; which is one-one and onto, Furthermore, since XDY, the restriction of f to ¥, which we shall also denote by f, is algo one-one; hence Y is equivalent to a subset of X;, Le. ¥~¥i wher X2YoMoy146 CARDINAL NUMBERS [cHAP. 9 and f:¥ 7 Ys is one-one and onto, But now XsCY; for similar reasons, X:~ Xz where XDYOM ION and f:X, > Xs ia oneone and onto, Consequently, there exist equivalent sets X,Xi,Xa,... and ‘equivalent sets ¥,Y.,¥s,... such that XDYOM OMIM Let B= XaYaXnKakyn ya ‘Then (KY) UAV UYU UB (FX) U YQ U(X) Ue UB Note further that (=) ~ =r) ~ Hy ~ Spoeifically, the function FOV) > Ker Year) fs one-one and onto, Consider the function g:X + Y defined by the following diagram: In other words, fe) it ze XY, or zeX-¥ wo = {0 EKON SELB ‘Then g is one-one and onto, Therefore X ~ ¥. . Prove Theorem 9.12: ¢= 2". Solution: Let R# be the set of real numbers and 2* be the family of subsets of Q, the set of rational numbers. Furthermore, let the function f:R* > 2° be defined by Ka) = (@ | #eQ,2 % defined by fey —2i2+ 1/2 if zis odd 22 if is even hhas the required property. 22, Consider the following diagrams which eontain Nx N. ay 9a ao— 1 —S2) 9) a ht) 2)—@22 23 @4) oe oe en a2 aa 6) a D624 9— ai ey 8 A) @Y=—B2—BH GH) “@ ®28. Let A HAP. 9] CARDINAL NUMBERS 149 In view of these diagram: follows: XN ean be written as an infinite sequence of distinct elements as NXN = (2, GM, @2), 12, (48) 29), .-.) NXN = (42, 2) @1), 8) 22) GM, OM, «9 23. Hint. Show that Uj, Ay is equivalent to NX N. 25. Let S be the set of points in the plane with rational coordinates and let Q be the set of rational numbers. Note that S ~ QxQ, since each point 2eS corresponds to a unique ordered pair (aa) QQ and vice versa. But QxQ is denumerable since Q is denumerable. Therefore Sis denumerable. 26, Any interval T;, ieJ, contains at least one rational number gj. Furthermore, if 7,7, then a,x 4), since 7, and 7) are disjoint. ‘Therefore {(7)},,; is eqivalent to subsct (q),7 of the rational numbers. Hence (7,),-;_ is countable, 2. Hint. B®, which is non-denumerable, is the union of the algebraic and transcendental numbers, 0,1]. We show that Ax A has the power of the continuum. Let 2, ¢ [0,1]. ‘Then x and y ‘can be written uniquely in the form of an infinite decimal, 2 = Onn, y= Ommy ‘which contains an infinite number of non-zero digits, (e.g. for } write 0.4999... instead of 0.6000... Let f: AXA A be defined by Mew) = Omweney ‘Then fis one-one, Hence A A has cardinality at most ¢. But AXA has eardin for example, at least ¢ since, (0,2) | ze (0,1) which is a subset of Ax A, is equivalent to A. Consequently, AXA has cardinality c, ie, has the power of the continuum. Note that #(A) =<; hence ot = #(4 x A) 82, Let B= {0,1}. Thon #(B)=2 and B*=C(A), the set of characteristic functions of A. By Problem 8, 24~B'. Hence #(2") = #(B4). 38. Let a = H(A), B = #(B) and y = #(C) where B and C are disjoint. Then p+y = #(BUC). Note that. ae ABU and gBar = BAP KAS) Note also that AUC consists of all the functions with domain BUC and co-domain A. A® and AC hhave similar meaning. ‘The theorem is proven if we show that auc ~ ABx ac Let fe ABUC correspond to the ordered pair of functions WB, FO) the restriction of f to B and the restriction of f to C. Note that (f|B, f|C) belongs to AB x AC, ‘The function F;ABUC + AD AC’ defined by FO = UIBAIO is one-one and onto, Hence ABUC ~ AB x AC, ‘Therefore at*Y = a? ar. 34, Since ap, there exists a set B with a subset A such that «= #(A) and = #(B). Furthermore, let y= #10). (1) Let fea®, ie, lets: > A. Since ACB, f can alto be considered as a function from C to B, that is, feB°, Hence A® is @ subset of B® and, therefore, A°< B®. Since a¥ = #(A®) and f= #(B"), we can conclude that a” = 2, (2) Let fect, ie. let f:A > C. Moreover, let "be an extension of f to a function f':B > C. Note that if fg, then f's9' where g' is an extension of g. Hence the function F':C*-> C* defined by FY) =f" is one-one. Therefore C* 7 ~ defines a partinl order in W in the following way: 2 Sy if # = y or if one ean g0 from 2 to'y In the diagram, always moving in the indlented direction, Le. upward. Note that bo a,d a means a a means a bd for any two elements @ and b belonging to A. A set A together with a specific total order in A is called a totally ordered set. Example 21: ‘The partial order in any set A of real numbers (with the natural order) is a total order since any two numbers are comparable, Example 22: Let R be the partial order in V = (1,2, , 4,5, 6) defined by “e divides y”. ‘Then Bis not a total order in V since 8 and 5 are not comparable. Example 23: Let A and B be totally ordered sets. ‘Then the totally ordered as follows: (4,8) <(@,b) if a B which is one-one and onto and which has the property that, for any elements a,a’e A, a, or | (not comparable) between each pair of numbers: (@) 28, (b) 1824, (2) 9__8, (d) 515 (2) State whether or not each of the following subsets of N is totally ordered: (a) (24,2,6), (b) {8,15,5}, (c) (15,5,30), (4) (2,8,82,4}, (e) (1,2,8,...}, () (7) Solution: (1) (@) Since 2 divides 8, 2 precedes 8, ie. 2<8, (@) 18 does not divide 24, and 24 does not divide 18; hence 18 124 (6) Since 9 is divisible by 3, 9 > 8 (@) Since 6 divides 15, 5 <15. (2) (@) Since 2 divides 6 which divides 24, the sot is totally ordered. (@) Since 3 and 5 are not comparable, the set is not totally ordered. (©) The set ia totally ordered since 6 divides 15 which divides 30, (@) The set is totally ordered since 2<4~<8 < 32. (o) The set is not totally ordered since 2 and 3 are not comparable () Any set consisting of one element is totally ordered. 2 Let V {a,b,¢,d,e} be ordered by the following diagram: : v4 \ 4 / \, a (1) Insert the correct symbol, <, >, or t (not comparable) between each pair of elements: @a_e, (b)b¢, ()d—a, (d)c_d (2) Construet a diagram of the elements in V which defines the inverse order. Solution: (1) (@) Since there is a “path” from ¢ to ¢ to a, ¢ precedes a; hence a > e. (®) There is no path from to ¢, or vice versa; hence b Fe. (©) There is a path from d to b to a; hence d, between each of the following pairs of numbers: (@) 514, (b) 69, (c) 3__20, (a) 14__21 Solution: ‘The elements in N can be written as follows: i [18 [16 10 | a4 | 1 | 22 | 26 | a0 2 | 20 | 28 | 36 | a4 | 62 | 0 ‘Then a number in a higher row precedes a number in a lower row and, if two numbers are in the sme row, the number to the left precedes the number to the right. Consequently, (a) 5X14, 0) 6>9, () B<20, ( 14>21 4, Let N x N be ordered lexicographically. Insert the correct symbol, < or >, between each of the following pairs of elements of N x N: (a) (,78)__(7,1), (6) (4,6)__(4,2), (ce) (6,5). Solution: Note that according to the lexicoxraphical ordering, (a,8)<(#,b) if a (4,2), since 4=4 but 6>2 (©) (6,5) > (4,23), since 5 > 4, (@ (1,3)> (4,2), since 1=1 but 3>2 (4,28), (4) (1,8) (1,2) 5. Let A= (N,=), the natural numbers with the natural order, and let B = (N,=), the natural numbers with the inverse order. Furthermore, let A xB denote the lexico- graphical ordering of N x N according to the order of A and then B. Insert the correct symbol, < or >, between each of the following pairs of elements of N x N. © B9)—G,0, ) @)—@,8), (© B3)—@1), @ (4,9)_(7,15) Solution: ‘The following rule applies: (a, 5) <(o',b’ if fe for a= a! but b> (@) @,8)> (2,1), since 8>1; ie. 8>1 according to the order of A. @) (2,1) > 28), but 1<8; ie. 1>8 according to the order of B. (© (8,3) < (3,1), but 8>1; ie 8<1 according to the order of B. (@) (4,9) < (1,15), since 4< 1; ie, 4<7 according to the order of A. 6. Let of be the family of all subsets A of the natural numbers N where A has the follow- ing properties: A is finite and the greatest common divisor of the elements of A is 1. (1) State whether or not each of the following subsets of N belongs to of (a) (2,8,8) (c) (2,5) (e) (4,6, 8) (0) (2,8, 5,8) (d) (2,3,4,5,...) ) (2,8)CHAP. 10) PARTIALLY AND TOTALLY ORDERED SETS 159 (2) Now order of by set inclusion, ie. X<¥ if XCY, and let @ be the subfamily of of which consists of the sets in (1) which belong to of. Construct a diagram of 8. Solution: (2) ‘The greatest common divisor of {4,6,8) is 2, and {2,3,4.5, not belong to of. All the other sets do belong: to of and 2) A diagram of @ is as follows Fa 2,8, 8) (2,5) aly 1. et A= (ot) be ordered s follows: 7 ~ Let of be the family of all non-empty totally ordered subsets of A, and let of be partially ordered by set inclusion. Construct a diagram of of. Solution: ‘The totally ordered subsets of A are: {a}, (6), (c), {@,0), (e,e). Hence a diagram of of is as follows: fo.) {a,0) wy tay te) & Let B= (1,2,3,4,5) be ordered as follows: an i a Let B be the family of all totally ordered subsets of B which contain 2 or more elements, and let @ be partially ordered by set inclusion. Construct a diagram of B. sis fe eee ae tna Tap of 8's od ase ase ase 5, (1,4y ily a. ly Sassy an 18,5) MINIMAL, MAXIMAL, FIRST AND LAST ELEMENTS 9% Let A = (2,8,4,5,...) be ordered by “z divides ”. (1) Find all minimal elements. (2) Find all maximal elements. Solution: (1) If p is a prime number, then only p divides p (since 1¢); hence all the prime numbers ‘minimal elements. Furthermore, if aeA is not prime, then there is a number be A such that O divides a, ie. b ay Consider the subset E = (2,3,4) of D. (1) Find the set of upper bounds of B. (2) Find the set of lower bounds of B. (8) Does sup(E) exist? (4) Does inf (H) exist? wet Bohs ee Beh a Apa or ee Se ee 15. Consider Q, the set of rational numbers, and its subset A = {x | #eQ, «*<8). (1) Is A bounded above, i.e. does A have an upper bound? (2) Is A bounded below, i.e. does A have a lower bound? (8) Does sup(A) exist? (4) Does inf (A) exist? Solutio (1) 4 fs bounded above since, for example, 50 is an upper bound. (2) There are no lower bounds of A; hence A is not bounded below. (8) Sup (A) does not exist. Considering A as a subset of R#, the real numbers, then 3 would be the least upper bound of A; but as a subset of Q, sup (A) does not exist (4) Int (A) does not exist since the set of lower bounds is empty. 16. Let of be a family of sets which is partially ordered by set inclusion, and let ® be a subset of of. (1) Prove that if Beof is an upper bound of @, then (U,,,4) CB. (2) Is U,,.,A, an upper bound of 8? Solution: () Let 2 belong to Uses Av ‘Then there exists an A,, where je, such that ze Ay Since B is an ‘upper bound, AjcB: hence z belongs to B. Since 2 U;,1A; implies ze B, then (Uiey A) CB. (2) Even though (A)cer is a subfamily of of, it need not be true that the union Use; Ay is an clement in ef. ‘Therefore Uy. 4, is an upper bound of & if, and only if, U,ey Ay belongs to of. eet 17. Let N, the natural numbers, be ordered by “‘z divides y”, and let A = {a;,a1,...,dm) bea finite subset of N. (1) Does inf (A) exist? (2) Does sup(A) exist? Solution: (1) The greatest common divisor of the elements in A is inf (A) and it always exists, (2) The least common multiple of the elements in A is sup(A) and it always exists,162 PARTIALLY AND TOTALLY ORDERED SETS (CHAP. 10 SIMILAR SETS 18, Give an example of an ordered set X = (A,R) which is similar to Y = (A,R~), the set A with the inverse order. Solution: ‘The set of rational numbers Q, with the natural order, is similar to Q with the inverse order. In fact the function f:Q—@ defined by f(z) =z i a similarity mapping since, for any real numbers, 2 =y if'and only if -25—y ‘As a second example, consider the set W = (a, 5, ¢, 4, ¢, f) ordered as follows: \/ ! 7 \ ‘The following diagram, which is derived by inverting the original diagram and reversing the arrows, iv aX. Notice that the two diagrams are similar. The function F = (0), 6.0, (8, (0, (10, F,0) is a similarity mapping. 19, Let A be an ordered set and, for any element ae A, let S(a) be the set of elements which Precede , Le Se) = (| #eA,e of defined by f:2~ S(e) in a similarity mapping. If a, then 2 implien 2b; hence ab implies S(a}CS(). Also if S(@}CS(0), then ee Sle) also belongs to S(); hence S(e)S(0) implies ob. ‘Thus f preserves order. By definition, / is onto. We now show that it is one-one. If ax, then either a, or ! (not comparable) between each pair of numbers: 3% 28 (61, (3 _83 (2) ‘State whether or not each of the following subsets of N is totally ordered (@) (8,2,24) (©) 18,1,9) (0) (2,4, 6,8...) ® ©) @) 24,8, 24) ) (35, 3,9) — \7 “3 V (1) Insert the correct symbol, <>, oF (not comparable) between each pair of elements: @i_s (45 @5—1, (4_2 (2) Construct a diagram of the elements in W which defines the inverse order. (2) Find all the totally ordered subsets of W, each of which contains at least three elements. (4) Find all the totally ordered subsets of W with the inverse order, each of which contains at least three elements, 22, Let A= (N,*), the natural numbers with the natural order, let B = (N,™), the natural numbers swith the inverse order, and let A x B be ordered lexicographically. Insert the correct symbol, < or >, between each of the following pairs of elements of AB. @) 1,9 4,5), ©) &N— 18), © 4,304), (@ @,2) 05,18) \Z { ! ‘Let @ be the family of all non-empty totally ordered subsets of D ordered by set inclusion. Construct a diagram of 8.164 PARTIALLY AND TOTALLY ORDERED SETS [CHAP. 10 ‘MINIMAL, MAXIMAL, FIRST AND LAST ELEMENTS 2, Let B = (a,b, 0,d,¢,f} be ordered as follows: — aa (1) (@) Find all the minimal elements of B. _(c) Does B have a first element? (2) Find all the maximal elements of B. __(d) Does B have a last clement? (2) Let & be the family of all non-empty totally ordered subsets, and let & be ordered by set inclusion. (a) Find all maximal elements of B. (©) Does have a first element? (®) Find all minimal elements of . (@) Does B have Inst clement? 2 Let M 2, 3,4, ...) and let MX M be ordered as follows: (0,8) < (ed) if'a divides c and if b is less than or equal to d. (1) Find all the minimal elements, (2) Find all the maximal elements, 2%, Let M = (2,3,4,...} be ordered by “e divides y". Furthermore, let 1 be the family of all non-empty. ‘totally ordered subsets of M and let .4¢ be partially ordered by set inclusion. (1) Find all minimal elements of e¥. (2) Find all maximal elements of <1. 27, State whether each of the following statements is true or false and, if it false, give » counter-example. (1) If a partially ordered set A has only one maximal clement a, then a is also a last element. (2) Ifa finite partially ordered set A has only one maximal element a, then a is alto a last clement. (2) If a totally ordered set A has only one maximal element a, then a is also a last element. UPPER AND LOWER BOUNDS 28. Let W = (1,2,...,7,8) be ordered as follows: 1 N re ~ (1) Consider the subset A = (4,5, 7) of W. (q) Find the set of upper bounds of A. {(@) Does sup (A) exist? (6) Find the set of lower bounds of A. (@) Does inf (A) exist? (2) Consider the subset B = (2, 8,6) of W. (a) Find the set of upper bounds of B. (©) Does sup (B) exis (6) Find the set of lower bounds of B. (@) Does inf (2) exist? (8) Consider the subset C= (1, 2, 4,7) of W. (a) Find the set of upper bounds of C. (@) Does sup (C) exist? (8) Find the set of lower bounds of C. (@) Does inf (C) exist? 29, Consider @, the set of rational numbers with the natural order, and its subset A: A = (@| 2eQ,8<28< 15) (2) Is A bounded above? (2) Is A bounded below? (3) Does sup (A) exist? (4) Does inf (A) exist? SIMILAR SETS 30, Find the maximum number of pairwise non-similar partially ordered sets with three elements, and construct a diagram of each. 31. Prove Theorem 10.2: ‘The relation in seta defined by A~B is an equivalence relation.CHAP. 10] PARTIALLY AND TOTALLY ORDERED SETS 165 Answers to Supplementary Problems 2, (1) (a) 817, (®) 2>8, () 6<1, (d) 3>38 (2) (a) Yes, (b) Yes, (e) No, (a) Yes, (@) No, (/) No BL (1) (a) 1>6, (0) 415, (©) 511, @ 4<2 @ r 4 i N} | ON (8) {1,84 (1,8,0, (2,8,4, 2,8,61, (25,6) (@) The same sete aan @) 4.8)> 0,5, 6) GD G1, (6 4,80) <(6,0, 6,2) <5.15) BRB 24. (1) (a) d and f, (8) a, (€) No, (@) Yes, @ is a last clement. 2) (@) (a,5,d)}, {a,be0h, (eet (8) ‘The subsets consisting of one element: (a), {0}, {e), {a}, (eb, U0. (©) No, (a) No 25. (1) Any ordered pair (p,2), where p is a prime, is a minimal element. (2) ‘There is no maximal element. 26, (1) Each subset consisting of one element is @ minimal clement. (2) Each set of the form {p,,PiPz,PiP2Pss---) Where Py, Pz .+- 18 any sequence of primes, maximal element. 2. (1) False, Consider, for example, the set (a,1,2,3,...) ordered as follows: a a “ Mie Note that the subset (1,2,3,...} has the natural order. ‘Then a is a maximal clement, the only fone, and it is not a last element. (2) True, (8) True, In fact, a totally ordered set can have at most one maximal element and it will always be a last element. 28. (1) (a) {1,2,3), () {8}, (@) sup(A)=3, (a) inf A) @) (a) (2)," () (6,8), (6) sup(B) = 2, (d) intB (8) (a) Q." There are no upper bounds.” (6) {8}, (@) No, (a) inf (C 2, (1) Yes, (2) Yes, (8) No, (4) inf (A) = 2 80, ‘There are five non-similar ways to order three elements, ie. a set A = {a,b,c}: t > me « t we i x @ ® @) ® ®Chapter 11 Well-ordered Sets. Ordinal Numbers WELL-ORDERED SETS Not every ordered set, even if it is totally ordered, need have a first element. One of the fundamental properties of N, the set of natural numbers with the natural order, is that N and every subset of N does have a first element. We call an ordered set well-ordered if it has this property. Specifically, Definition Let A be an ordered set with the property that every subset of A contains a first element. Then A is called a well-ordered set. In particular, any well-ordered set A is totally ordered. For if a,be A, then the subset (a,b) of A contains a first element which, therefore, must precede the other; hence any two elements of A are comparable. ‘The following theorems follow directly from the above definition, ‘Theorem 11.1.1: Every subset of a well-ordered set is well-ordered. ‘Theorem 11.1.2: If A is well-ordered and B is similar to A, then B is well-ordered. Example 141: Consider the ordered subsets Ar = 0,8,8,-.) and Ar = (24,6...) of N which are also well-ordered, ‘Then the union (ordered from left to right) AVAL = (3,550.65 24,6)...3 is also well-ordered. ‘This example shows that it is possible for a set, such as N= A.UAs to be well-ordered in more than one way ‘The preceding example can be generalized as follows: Theorem 11.2: Let (A,),,, be a well-ordered family of pairwise disjoint well-ordered sets. Then the tnion of the sets U,,,A, is well-ordered. (The order of the union of a totally ordered family of totally ordered sets is defined in Example 2.4 of Chapter 10.) Example 12: Let V = (ai,as,...as} be any finite totally ordered set, ‘Then V can be written as : Vi = (ain diy 5 04) where the ai, are the original elements rearranged according to the order. Notice that V is wellordered. Furthermore, notice that any other totally ordered set of oe We (iy bay oo sy Big) is similar to V. In view of Example 1.2, we state ‘Theorem 11.3: All finite totally ordered sets with the same number of elements are well- ordered and are similar to each other. ‘TRANSFINITE INDUCTION ‘The reader is familiar with the Principle of Mathematical Induction Let S be a subset of N, the natural numbers, with the following properties: (Q) 1eS 2) neS implies n+108. ‘Then S is the set of natural numbers (ie. S=N). 16CHAP. 11] WELL-ORDERED SETS. ORDINAL NUMBERS 167 ‘The above principle is one of Peano’s Axioms for the natural numbers. The principle can be shown to be a consequence of the fact that N is well-ordered. In fact, there is a somewhat similar statement which is true for every well-ordered set. Principle of Transfinite Induetion Let S be a subset of a well-ordered set, A with the following properties: Q) aes (2) 9(a)CS implies ae. ‘Then S=A. Here, ay is the first element of A and s(2), called the initial segment of a, is defined to bbe the set of elements in A which strictly precede a. LIMIT ELEMENTS ‘An element b in an ordered set A is called an immediate successor of an element ae A, and a is called the immediate predecessor of b if a~ S(A) defined by f:x > a(z) is @ similarity mapping of A into S(4). SIMILARITY BETWEEN A WELL-ORDERED SET AND ITS SUBSET Consider N, the natural numbers, and the subset E = (2,4,6,...) of N. ‘The function {:N > E defined by f(z) = 2x is a similarity mapping of N into its subset E. Notice that, for every xeN, 2 sf) This property is true in general. . ‘Theorem 11.6: Let A be a well-ordered set, let B be a subset of A, and let the function f:A > B bea similarity mapping of A into B. Then, for every ae A, a < fla) The following important properties of well-ordered sets are consequences of the pre- ceding theorem. ‘Theorem 11.7: Let A and B be similar well-ordered sets. ‘Then there exists only one similarity mapping of A into B. ‘Theorem 11.8: A well-ordered set cannot be similar to one of its initial segments. COMPARISON OF WELL-ORDERED SETS ‘The next theorem gives an important relationship between any two well-ordered sets. ‘Theorem 11.9: Given any two well-ordered sets, either they are similar to each other or one of them is similar to an initial segment of the other. If a well-ordered set A is equivalent to an initial segment of a well-ordered set B, then A is said to be shorter than B or B is said to be longer than A. With these definitions, ‘Theorem 11.9 can be restated as follows: ‘Theorem 11.9: Let A and B be well-ordered sets. ‘Then A is shorter than B, A is similar to B or A is longer than B. ‘The preceding theorem can be strengthened as follows: ‘Theorem 11.10: Let of be any family of pairwise non-similar well-ordered sets. ‘Then there exists a set Acof such that A is shorter than every other set in cf. Example 41: Consider two finite well-ordered sets A= (yey) and B= {buy csp bad Then if n m then A would be longer than B.CHAP. 11] WELL-ORDERED SETS, ORDINAL NUMBERS 169 shorter than the well-ordered set (0,8,5) 0205 26466, 0} since it is similar to the initial segment (1,3, 5, ...}- Example 42: Note that N = (1,2,8,...) ORDINAL NUMBERS Notice once again, by Theorem 10.2, that the relation in ordered sets defined by A=B ie. A is similar to B, is an equivalence relation. Hence by the Fundamental Theorem on Equivalence Relations, all ordered sets, and in particular all well-ordered sets, are parti- tioned into disjoint classes of similar sets. Definition 11.3:] Let A be any well-ordered set and let A denote the family of well-ordered sets which are similar to A, Then A is called an ordinal number and it is denoted by = ord (A) Definition 11.4:] The ordinal number of each of the wellardred els D (1), (1,2), 1,28), is denoted by 0,1,2,8,... respectively, and is called a finite ordinal number. All other ordinals are called transjinite numbers. Definition 11.5:] The ordinal number of N, the natural numbers, is denoted by «= ord (N) Remark 112: Although the symbols 0,1,2,8,... are used to denote natural numbers, cardinal numbers and, now, ordinal numbers, the context in which the symbols appear determines their particular meaning. Furthermore, since by Theorem 11.8 any two finite well-ordered sets with the same number of jements are similar, 0,1,2,... are the only finite ordinal numbers. In view of the definition of the order type of a totally ordered set in the prec chapter, Definition 11.8 can be restated as follows: Definition 11.8’:] Let A be the order type of a totally ordered set A. If A is well-ordered then d is called an ordinal number. EQUALITIES AND ORDINAL NUMBERS An inequality relation is defined for the ordinal numbers as follow: Let A and » be two ordinal numbers and A and B be two well-ordered sets such that A = ord(A) and,» = ord(B) ‘Then A

if Ais longer than B 4p if Apora170 WELL-ORDERED SETS. ORDINAL NUMBERS (CHAP. 11 Consider two finite well-ordered sets 4 } and B ordered positionwise from left to right. Say m initial segment (by, ...,bs) of B. Hence ord(A) = ord (B). In other words, n= m as ordinal numbers if and only if bers. Thus the inequality relation for ordinal numbers inequality relation in the set of natural numbers. Example 52: Let A = ord ((1,3,5,...;2,4,6,...)). Since N, the natural numbers, js similar to the initial segment {1, 8, 8, .-), ba, ey bad m. Then A is similar to the fas natural num- an extension of the es ‘The next theorem is a direct consequence of Theorem 11.9 and the above definition. ‘Theorem 11.11: Any set of ordinal numbers is totally ordered by the relation A=). In view of Theorem 11.10, the preceding theorem can be strengthened as follows: ‘Theorem 11.12: Any set of ordinal numbers is well-ordered by the relation A=). Now let 4 be any ordinal number and s(A) denote the set of ordinal numbers less than A. By the preceding theorem, s(x) is a well-ordered set and, therefore, ord(s(i)) exists. Question: what is the relationship between and ord(s(A))?\ The answer is given in the next theorem. Theorem 11.13: Let (A) be the set of ordinals less than the ordinal a. Then A = ord (s(A)). Remark 11.3: Since the ordinal numbers are themselves well-ordered, every ordinal has an immediate successor. Some non-zero ordinals, for example w, do not have immediate predecessors; these are called limit ordinal numbers or, simply, limit numbers, ORDINAL ADDITION An operation of addition is defined for ordinal numbers as follows: Definition 11.7:] Let 4 and » be ordinal numbers such that A= ord(A) and » = ord(B), where A and B are disjoint. Then A+n = ord (fA; B)) Example 641: Note o = ord ({1,2,.-.)) and nm = ord ((a1,..2505)). Then me = ord lay sya 42.3) = 0 But etm = ord... aya) > 0 since NY is equivalent to s(a), the initial segment of a Thus we see, by Example 6.1, that the operation of addition of ordinal numbers is not commutative. However the following conditions do hold. Theorem 11.14: (1) Addition of ordinal numbers satisfies the associative law, i Qty ty = At (Ha) (2) The ordinal 0 is an additive identity element, OFA = atoaa Example 62: In is example we will denote the finite ordinals by 0%, 18, 2%, Consider, now, two finite well-ordered disjoint sets A= (dy ecya) and B= (by,CHAP. 11) WELLORDERED SETS. ORDINAL NUMBERS qm ‘Then nttmt = ordi(A) + ord(B) = ord((A; B) = (nt-mt ‘Thos the operation of addition for finite ordinal numbers corresponds to the opera- tion of addition for the aot of natural numbers Note once again that the set of ordinal numbers is itself a well-ordered set; hence every ordinal has an immediate successor. For the finite ordinals, ie. the natural numbers, it is easily seen that n+1 is the immediate successor to n. ‘The next theorem states that this property is true in general. ‘Theorem 11.15: Let A be any ordinal number. ‘Then A+1 is the immediate successor of 2. Addition of real numbers and, therefore, of natural numbers, is a binary operation ‘and can be extended by induction to any finite sum ay +a tor Hy of real numbers. ‘The sum of an infinite number of real numbers, such as 1+24+8+4+- L+e+dtbs: has no meaning (unless one introduces the concepts of limits). On the other hand, it is possible to define the sum of an infinite number of ordinal numbers as follows: Let {A),,, be any well-ordered set, finite or infinite, of ordinal numbers. In other words, I'is @ well-ordered set, and to each ie! there corresponds an ordinal number 2, Furthermore, let a Then the family of sets (A, x (i)},,, is a well-ordered family of pairwise disjoint well- ordered sets. By Theorem 11.2, User (AX) is a well-ordered set. We therefore state Definition 11.8:] Let (4),,, be a well-ordered set of ordinal numbers such that A, = ord (A). Then Died = ord(Ue (A,X )) Example 63: By the above definition, 141414 ++ then In fact, if each 2, is finite (and not 0) Mthtates = Sewh = 0 ORDINAL MULTIPLICATION An operation of multiplication is defined for ordinal numbers as follows: Definition 11.9:] Let 2 and , be ordinal numbers such that 4 = ord(A) and x = ord (B). ic Aw = ord ((A x B}) where (A x B} is ordered reverse lexicographically. Note that (A x B) is ordered reverse lexicographically means that (aa) < (0,0) if a’ 0 since N is similar to the initial segment (1,2), (2,4), . ‘Thus we see that the operation of multiplication of ordinal numbers is not commutative. However, the following conditions do hold. ‘Theorem 11.16: (1) The associative law for multiplication holds, i.e., Mun) = Only (2) The left distributive law of multiplication over addition holds, ic., Mau tn) = Met ay (8) The ordinal 1 is a multiplicative identity element, i.e. waasa STRUCTURE OF ORDINAL NUMBERS ‘We now write down many of the ordinal numbers according to their order. First come the finite ordinals 01,23 and then comes the first limit ordinal » and its successors owl, ot2, ... Note (see Example 7.1) that ord ((0,1,2,...; « e+, 42, ...)) = 02 Hence next comes the second limit number «2 and its successors 02, 2H, o2t2, o243, ‘The next limit number is «3. We proceed as follows: 08, BD oy dy By cee cece we at # is the limit number following the limit numbers an, where ne N. We continue: PHD iy FOB, ey cay a tet a Wl2 Here oo APH yet to tet, ‘Then 8, oy PB, ‘Then we have the powers of w: AHL ot Here w is the limit number after the limit numbers o*, where neN. We proceed: 8 oo (I oo (OO « After all these ordinals we have the ordinal «,. We can continue: wath Each of the ordinal numbers we have enumerated is still the ordinal number of a denumerable set. AUXILIARY CONSTRUCTION OF ORDINAL NUMBERS Note once again ‘Theorem 11.13: Let s(A) be the set of ordinal numbers which precede A. ‘Then A = ord(s(a))oHAP. 11] WELL-ORDERED SETS. ORDINAL NUMBERS 173 Some authors use this property of ordinal numbers to actually define the ordinal numbers. Roughly speaking, an ordinal number is defined to be the set of ordinal numbers which precede it. Specifically, Definition] 0 = 0 oF2 = (01,2... 041} 12 (0) : 2 = (0,1) 3 = {0,1,2) : 2 = (1. e0t...) 4 e241 © (0,1, soe thy 52.5 08) © = (01,2...) ot1 = (0,1,2,...,0) One main reason the ordinal numbers are developed as above is in order to avoid certain inherent contradictions which appear in the preceding development of the ordinal numbers (see Chapter 13). Solved Problems 1. Prove the Principle of Transfinite Induction: Let S be a subset of a well-ordered set A with the following properties: (1) dxeS, (2) s(a)CS implies aeS. Then S=A. Solution: Suppose SHA, ie. suppose A—S = 1 is not empty. Since A is well-ordered, T has a frst clement "a. Each clement zeal) precedes ts and, therefore, cannot belong to 7, ic. belongs to S; hence a(t) CS. By (2), eS, This contradicts the fact that. tee AS. Hence the original assumption that SA is not true; in other words, S=A. (Note that (1) is in fact a consequence of (2) since = sla) is a subset of S and, therefore, implies ave S:) 2 Prove Theorem 11.5: Let S(A) denote the family of all initial segments of elements in a well-ordered set A, and let S(A) be ordered by set inclusion. Then A is similar to ‘S(4) and, in particular, the function f:A > S(A) defined by f:2— s(z) is a similarity mapping of A into S(A). Solution: By definition f is onto, We show that / is one-one. Suppose zy. Then one of them, say 2, strictly recedes the other: hence zee(y). Bot, by definition of initial segment, xa(2). ‘Thus tle) way), and hence f is one-one. ‘We show that / preserves order, that Sy iand only if a(2) Catv) Let ey. If axel) then oz and hence «; then yea(2). But, by definition of initial segment, Veep); hence a(a)gely). In other words, # B bea similarity mapping of A into B. Then, for every ae A, a B and g:A~ B be similarity mappings. Suppose f 9. ‘Then there exists an element #eA such that f(z) #g(2). Consequently, either fz) T defined by fey =y it ae) ~ atu) is one-one and onto. Now let #',2eS, fle) =u fla’) =y' and 2’ <2, The theorem is proven if we ean show that yie Hence (14 1)o x lo +a. 19, Let (A,},,, be a well-ordered family of pairwise disjoint well-ordered sets, and let ord (1) = sand ord(A) =o for every iel. Find ord(U,,,4)- ord (Vier) = etetete = alti tit 20. Prove: «+e = «2. Solution: ‘Method 1. Note that ote = ltel = ott) = 2 Here, the left distributive law is used. ‘Method £. Consider the well-ordered sets A= (ayy, B= Oyby dy C= leney od D= trad Nets tt = ord(A) = ord(@) = on) and 8 = ord) hm ee ee «2 = ord({CxD}) ord ({(c1,1), (¢2, 7), (eis), (en 8), «+ 3) But the function f:4As B}-+ (CX) dasa by ie eee wo = {03 EF Is a similarity mapping of (A; B) into CXD. Hence ete = ord((A; BY) = ord {CxD) = 02 Supplementary Problems ‘2, Prove Theorem 11.1.2: If A is a well-ordered set and B lar to A, then B is well-ordered. 22, Prove Theorem 11.2: Let {Ajje be a well-ordered family of pairwise disjoint well-ordered sets. ‘Then the union of the sets U,,, A, is well-ordered. 23. Assume that NV, the set of natural numbers with the natural order, is well-ordered. Prove the Principle of Mathematical Induction: Let S be a subsot of N with the properties (1) eS and @) neS implies n+1eS; then S=N. 2. Prove that 0 is the identity clement for addition of ordinal numbers, that is, for any ordinal 2, OFA =at0=,178 WELL-ORDERED SETS, ORDINAL NUMBERS (CHAP. 11 25, Prove that 1 is the identity element for multiplication of ordinal numbers, that is, for any ordinal i, ema 2%, Prove: If each iy -ieN, is a finite ordinal, then ry + d+ 03+ -* = Bier 2, Prove: Let 2 be any infinite ordinal number. Then \ = «+7, where « is a limit number and » is a finite ordinal 28, State whether each of the following statements about ordinals is true or false; if it is true prove it, and if it is false give a counter-example: Q) edo, then e X, where (Aijie1 is a family of subsets of X, is called a choice function if f(A.) = ae Ay for every ie/. In other words, f “chooses” a Point ae Ay for each set A: Example L1: Let (AyAz,-.414q} be a finite family of sets. In Chapter 5, we defined the Cartesian product of the n seta, KAgX eX AQ = TP AL to be the set of m-tuples ay 4 dy) where aje A; for i= 1,...,n. But to each choice function f defined on {Ay An) there corresponds the unique n-tuple WAY, fd), AQ land vice versa. Accordingly, in the finite case, Definition 12.1 agrees with the previous definition of the Cartesian product. ‘The main reason for introducing Definition 12.1 is that it applies to any family of sets: finite, denumerable or even non-denumerable. The previous definition, which used the concept of n-tuples, applied only to a finite family of sets. Remark 12.1: Although a choice function is defined for a family of subsets, any family of sets (Ai}i+1 can be considered to be a family of subsets of their union User Ac AXIOM OF CHOICE The axiom of choice lies at the foundations of mathematics and, in particular, the theory of sets. This “innocent looking” axiom, which follows, has as a consequence some of the most powerful and important results in mathematics. Axiom of Choice: The Cartesian product of a non-empty family of non-empty sets is non-empty. In view of Definition 12.1, the Axiom of Choice can be stated as follows: Axiom of Choice: There exists a choice function for any non-empty family of non-empty sets. 179180 AXIOM OF CHOICE. ZORN'S LEMMA. WELL-ORDERING THEOREM (CHAP. 12 ‘The axiom of choice is equivalent to the following postulate: Zermelo’s Postulate: Let (Aijier be any non-empty family of disjoint non-empty sets. ‘Then there exists a subset B of Use: A, such that the intersection of B and each set A; consists of exactly one element. Observe that in Zermelo’s Postulate the sets are disjoint whereas in the Axiom of Choice may not be disjoint. WELL-ORDERING THEOREM, ZORN’S LEMMA ‘The following theorem is attributed to Zermelo who proved the theorem directly from the axiom of choice. Well-ord Zorn's Lemma, which follows, is one of the most important tools in mathematics; it establishes the existence of certain types of elements although no constructive process is given to find these elements, ig Theorem: Every set can be well-ordered. Zorn's Lemma: Let X be a non-empty partially ordered set in which every totally ordered subset has an upper bound in X. Then X contains at least one maximal element. We formally state and prove (Problem 4) the following basic result of set theory: Theorem 121; The following are equivalent: (i) Axiom of Choice, (li) Well-ordering ‘Theorem, (iii) Zorn’s Lemma. CARDINAL AND ORDINAL NUMBERS To each ordinal number A = ord(A) we can associate a unique cardinal number a= #(A). We call « the cardinal number of A and denote it by e=i This function from the ordinal numbers to the cardinal numbers is not one-one, that is, there are different ordinal numbers with the same cardinal number. For example, © = ord ({1,2,8, ...)) «2 = ord (a1, a, ...5 bi, ba, ...}) are both ordinal numbers of denumerable sets, i.e. sets with the same cardinal number a. In other words, : one ‘The well-ordering theorem implies that the above function from the ordinal numbers to the cardinal numbers is onto. For, suppose «= #(A) is any cardinal number. By the well-ordering theorem, A can be well-ordered; say A=ord(A). Then a=%. Hence a is the cardinal number of at least one ordinal number A. (Here, A is used both as the original set and then as the well-ordered set.) The following correspondence between the ordinal and cardinal numbers is easily established. ‘Theorem 122: Let «=X and f= 7 be cardinal numbers. Then (1) «

Вам также может понравиться

• Fear: Trump in the White House

  

  От Everand

  Fear: Trump in the White House

  Bob Woodward

  Рейтинг: 3.5 из 5 звезд

  3.5/5 (738)
• A Man Called Ove: A Novel

  

  От Everand

  A Man Called Ove: A Novel

  Fredrik Backman

  Рейтинг: 4.5 из 5 звезд

  4.5/5 (4609)
• A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story

  

  От Everand

  A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story

  Dave Eggers

  Рейтинг: 3.5 из 5 звезд

  3.5/5 (231)
• • Журналы
  • Подкасты
  • Нотное издание
• The Sympathizer: A Novel (Pulitzer Prize for Fiction)

  

  От Everand

  The Sympathizer: A Novel (Pulitzer Prize for Fiction)

  Viet Thanh Nguyen

  Рейтинг: 4.5 из 5 звезд

  4.5/5 (119)
• Never Split the Difference: Negotiating As If Your Life Depended On It

  

  От Everand

  Never Split the Difference: Negotiating As If Your Life Depended On It

  Chris Voss

  Рейтинг: 4.5 из 5 звезд

  4.5/5 (838)
• Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America

  

  От Everand

  Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America

  Gilbert King

  Рейтинг: 4.5 из 5 звезд

  4.5/5 (265)
• The Little Book of Hygge: Danish Secrets to Happy Living

  

  От Everand

  The Little Book of Hygge: Danish Secrets to Happy Living

  Meik Wiking

  Рейтинг: 3.5 из 5 звезд

  3.5/5 (399)
• Grit: The Power of Passion and Perseverance

  

  От Everand

  Grit: The Power of Passion and Perseverance

  Angela Duckworth

  Рейтинг: 4 из 5 звезд

  4/5 (587)
• The World Is Flat 3.0: A Brief History of the Twenty-first Century

  

  От Everand

  The World Is Flat 3.0: A Brief History of the Twenty-first Century

  Thomas L. Friedman

  Рейтинг: 3.5 из 5 звезд

  3.5/5 (2219)
• Yes Please

  

  От Everand

  Yes Please

  Amy Poehler

  Рейтинг: 4 из 5 звезд

  4/5 (1891)
• The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life

  

  От Everand

  The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life

  Mark Manson

  Рейтинг: 4 из 5 звезд

  4/5 (5794)
• Principles: Life and Work

  

  От Everand

  Principles: Life and Work

  Ray Dalio

  Рейтинг: 4 из 5 звезд

  4/5 (599)
• Team of Rivals: The Political Genius of Abraham Lincoln

  

  От Everand

  Team of Rivals: The Political Genius of Abraham Lincoln

  Doris Kearns Goodwin

  Рейтинг: 4.5 из 5 звезд

  4.5/5 (234)
• Rise of ISIS: A Threat We Can't Ignore

  

  От Everand

  Rise of ISIS: A Threat We Can't Ignore

  Jay Sekulow

  Рейтинг: 3.5 из 5 звезд

  3.5/5 (137)
• Shoe Dog: A Memoir by the Creator of Nike

  

  От Everand

  Shoe Dog: A Memoir by the Creator of Nike

  Phil Knight

  Рейтинг: 4.5 из 5 звезд

  4.5/5 (537)
• The Emperor of All Maladies: A Biography of Cancer

  

  От Everand

  The Emperor of All Maladies: A Biography of Cancer

  Siddhartha Mukherjee

  Рейтинг: 4.5 из 5 звезд

  4.5/5 (271)
• The Glass Castle: A Memoir

  

  От Everand

  The Glass Castle: A Memoir

  Jeannette Walls

  Рейтинг: 4.5 из 5 звезд

  4.5/5 (1711)
• The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are

  

  От Everand

  The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are

  Brene Brown

  Рейтинг: 4 из 5 звезд

  4/5 (1090)
• A Tree Grows in Brooklyn

  

  От Everand

  A Tree Grows in Brooklyn

  Betty Smith

  Рейтинг: 4.5 из 5 звезд

  4.5/5 (1929)
• Her Body and Other Parties: Stories

  

  От Everand

  Her Body and Other Parties: Stories

  Carmen Maria Machado

  Рейтинг: 4 из 5 звезд

  4/5 (821)
• The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  

  От Everand

  The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  Ben Horowitz

  Рейтинг: 4.5 из 5 звезд

  4.5/5 (344)
• John Adams

  

  От Everand

  John Adams

  David McCullough

  Рейтинг: 4.5 из 5 звезд

  4.5/5 (2409)
• The Woman in Cabin 10

  

  От Everand

  The Woman in Cabin 10

  Ruth Ware

  Рейтинг: 3.5 из 5 звезд

  3.5/5 (2322)
• Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race

  

  От Everand

  Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race

  Margot Lee Shetterly

  Рейтинг: 4 из 5 звезд

  4/5 (890)
• Sing, Unburied, Sing: A Novel

  

  От Everand

  Sing, Unburied, Sing: A Novel

  Jesmyn Ward

  Рейтинг: 4 из 5 звезд

  4/5 (1103)
• Wolf Hall: A Novel

  

  От Everand

  Wolf Hall: A Novel

  Hilary Mantel

  Рейтинг: 4 из 5 звезд

  4/5 (3811)
• Angela's Ashes: A Memoir

  

  От Everand

  Angela's Ashes: A Memoir

  Frank McCourt

  Рейтинг: 4.5 из 5 звезд

  4.5/5 (440)
• Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future

  

  От Everand

  Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future

  Ashlee Vance

  Рейтинг: 4.5 из 5 звезд

  4.5/5 (474)
• The Art of Racing in the Rain: A Novel

  

  От Everand

  The Art of Racing in the Rain: A Novel

  Garth Stein

  Рейтинг: 4 из 5 звезд

  4/5 (4200)
• The Unwinding: An Inner History of the New America

  

  От Everand

  The Unwinding: An Inner History of the New America

  George Packer

  Рейтинг: 4 из 5 звезд

  4/5 (45)
• The Yellow House: A Memoir (2019 National Book Award Winner)

  

  От Everand

  The Yellow House: A Memoir (2019 National Book Award Winner)

  Sarah M. Broom

  Рейтинг: 4 из 5 звезд

  4/5 (98)
• The Perks of Being a Wallflower

  

  От Everand

  The Perks of Being a Wallflower

  Stephen Chbosky

  Рейтинг: 4.5 из 5 звезд

  4.5/5 (2099)
• The Constant Gardener: A Novel

  

  От Everand

  The Constant Gardener: A Novel

  John le Carre

  Рейтинг: 3.5 из 5 звезд

  3.5/5 (104)
• The Outsider: A Novel

  

  От Everand

  The Outsider: A Novel

  Stephen King

  Рейтинг: 4 из 5 звезд

  4/5 (1839)
• The Light Between Oceans: A Novel

  

  От Everand

  The Light Between Oceans: A Novel

  M.L. Stedman

  Рейтинг: 4.5 из 5 звезд

  4.5/5 (789)
• Little Women

  

  От Everand

  Little Women

  Louisa May Alcott

  Рейтинг: 4 из 5 звезд

  4/5 (104)
• On Fire: The (Burning) Case for a Green New Deal

  

  От Everand

  On Fire: The (Burning) Case for a Green New Deal

  Naomi Klein

  Рейтинг: 4 из 5 звезд

  4/5 (73)
• Brooklyn: A Novel

  

  От Everand

  Brooklyn: A Novel

  Colm Toibin

  Рейтинг: 3.5 из 5 звезд

  3.5/5 (1937)
• Manhattan Beach: A Novel

  

  От Everand

  Manhattan Beach: A Novel

  Jennifer Egan

  Рейтинг: 3.5 из 5 звезд

  3.5/5 (792)
• Bad Feminist: Essays

  

  От Everand

  Bad Feminist: Essays

  Roxane Gay

  Рейтинг: 4 из 5 звезд

  4/5 (1015)
• Ambika Wauters Homeopathic Color Sound Remedies

  

  Документ6 страниц

  Ambika Wauters Homeopathic Color Sound Remedies

  antoscri

  Оценок пока нет
• Steve Jobs

  

  От Everand

  Steve Jobs

  Walter Isaacson

  Рейтинг: 4.5 из 5 звезд

  4.5/5 (806)
• Consciously Quantum: How You Make Everything Real: Newsletters

  

  Документ10 страниц

  Consciously Quantum: How You Make Everything Real: Newsletters

  antoscri

  Оценок пока нет
• Cccccreated With Sketch - Created With Sketch.: Object 1

  

  Документ11 страниц

  Cccccreated With Sketch - Created With Sketch.: Object 1

  antoscri

  Оценок пока нет
• Philosophical Writing Beyond Standard Essays

  

  Документ2 страницы

  Philosophical Writing Beyond Standard Essays

  antoscri

  Оценок пока нет
• Identical Magnetic

  

  Документ8 страниц

  Identical Magnetic

  antoscri

  Оценок пока нет
• Everything Is Grounded in Logic

  

  Документ3 страницы

  Everything Is Grounded in Logic

  antoscri

  Оценок пока нет
• Predictive Homeopathy II

  

  Документ4 страницы

  Predictive Homeopathy II

  antoscri

  Оценок пока нет
• Are Crows Capable of Conscious Thought

  

  Документ2 страницы

  Are Crows Capable of Conscious Thought

  antoscri

  Оценок пока нет
• Scientists Discover The Purpose of Sleep

  

  Документ2 страницы

  Scientists Discover The Purpose of Sleep

  antoscri

  Оценок пока нет
• There Never Was Time

  

  Документ3 страницы

  There Never Was Time

  antoscri

  Оценок пока нет
• FORMULAS

  

  Документ2 страницы

  FORMULAS

  antoscri

  Оценок пока нет
• 25 Great First Date Questions

  

  Документ1 страница

  25 Great First Date Questions

  h4leluy4h

  Оценок пока нет
• Predictive Homeopathy

  

  Документ5 страниц

  Predictive Homeopathy

  antoscri

  Оценок пока нет
• The Influence of Emotional Valence on False Memories in the Deese-Roediger-McDermott Paradigm

  

  Документ8 страниц

  The Influence of Emotional Valence on False Memories in the Deese-Roediger-McDermott Paradigm

  antoscri

  Оценок пока нет
• Nikola Tesla's Idea of Free Energy

  

  Документ6 страниц

  Nikola Tesla's Idea of Free Energy

  antoscri

  Оценок пока нет
• Unpredictability in Culture

  

  Документ3 страницы

  Unpredictability in Culture

  antoscri

  Оценок пока нет
• School of Protagoras.-Gorgias

  

  Документ8 страниц

  School of Protagoras.-Gorgias

  antoscri

  Оценок пока нет
• Computer Simulated Matrix

  

  Документ2 страницы

  Computer Simulated Matrix

  antoscri

  Оценок пока нет
• Insights From International Relations Theory

  

  Документ73 страницы

  Insights From International Relations Theory

  antoscri

  Оценок пока нет
• Review Article: Complementary and Alternative Medicine and Cardiovascular Disease: An Evidence-Based Review

  

  Документ8 страниц

  Review Article: Complementary and Alternative Medicine and Cardiovascular Disease: An Evidence-Based Review

  antoscri

  Оценок пока нет
• Fantasia Coral PDF

  

  Документ35 страниц

  Fantasia Coral PDF

  Araceli Pérez

  Оценок пока нет
• A Study of Meaning

  

  Документ6 страниц

  A Study of Meaning

  antoscri

  Оценок пока нет