Problem Solvers TOPOLOGY A Complete Solution Guide OM Uma ory Su ee Cnet Wu ) SM cele lame LT he let Code PU Merl ety Ce CS em eu Sa PE Sem a el Limes Ce Se ees ee TC enabling you to understand the subject fully @ Will save you hours of time in finding solutions to problems Research & Education AssociationThis page is intentionally left blankTHE TOPOLOGY PROBLEM SOLVER’ REGISTERED TRADEMARK, A Complete Solution Guide to Any Textbook Emil G. Milewski, Ph.D. Research and Education Association 61 Ethel Road West Ys Piscataway, New Jersey 08854THE TOPOLOGY PROBLEM SOLVER ® Copyright © 1994 by Research and Education Association. Allrights reserved. No part of this book may be reproduced. in any form without permission of the publisher. Printed in the United States of America Library of Congress Catalog Card Number 94-65507 International Standard Book Number 0-87891-925-2 PROBLEM SOLVER is a registered trademark of Research and Education Association, Piscataway, New JerseyWHAT THIS BOOK IS FOR Students have generally found topology a difficult subject to understand and learn. Despite the publication of hundreds of textbooks in this field, each one intended to provide an improvement over previous textbooks, students continue to remain perplexed as a result of the numerous conditions that must often be remembered and correlated in solving a problem. Various possible interpretations of terms used in topology have also contributed to much of the difficulties experienced by students. In a study of the problem, REA found the following basic reasons under- lying students’ difficulties with topology taught in schools: (a) No systematic rules of analysis have been developed which students may follow in a step-by-step manner to solve the usual problems encoun- tered. This results from the fact that the numerous different conditions and principles which may be involved in a problem, lead to many possible differ- ent methods of solution. To prescribe a set of rules to be followed for each of the possible variations, would involve an enormous number of rules and steps to be searched through by students, and this task would perhaps be more burdensome than solving the problem directly with some accompany- ing trial and error to find the correct solution route. (b) Textbooks currently available will usually explain a given principle in a few pages written by a professional who has an insight in the subject matter that is not shared by students. The explanations are often written in an abstract manner which leaves the students confused as to the application of the principle. The explanations given are not sufficiently detailed and exten- sive to make the student aware of the wide range of applications and differ- ent aspects of the principle being studied. The numerous possible variations of principles and their applications are usually not discussed, and it is left for the students to discover these for themselves while doing exercises. Accord- ingly, the average student is expected to rediscover that which has been long known and practiced, but not published or explained extensively. (C) The examples usually following the explanation of a topic are too few in number and too simple to enable the student to obtain a thorough grasp of the principles involved. The explanations do not provide sufficient basis to enable a student to solve problems that may be subsequently assigned for homework or given on examinations. The examples are presented in abbreviated form which leaves out much material between steps, and requires that students derive the omitted material iiithemselves. As a result, students find the examples difficult to understand— contrary to the purpose of the examples. Examples are, furthermore, often worded in a confusing manner. They do not state the problem and then present the solution. Instead, they pass through a general discussion, never revealing what is to be solved for. Examples, also, do not always include diagrams/graphs, wherever appro- priate, and students do not obtain the training to draw diagrams or graphs to simplify and organize their thinking. (d) Students can learn the subject only by doing the exercises themselves and reviewing them in class, to obtain experience in applying the principles with their different ramifications. In doing the exercises by themselves, students find that they are required to devote considerably more time to topology than to other subjects of com- parable credits, because they are uncertain with regard to the selection and application of the theorems and principles involved. It is also often necessary for students to discover those “tricks” not revealed in their texts (or review books), that make it possible to solve problems easily. Students must usually resort to methods of trial-and-error to discover these “tricks,” and as a result they find that they may sometimes spend several hours to solve a single problem. (e) When reviewing the exercises in classrooms, instructors usually re- quest students to take turns in writing solutions on the boards and explaining them to the class. Students often find it difficult to explain in a manner that holds the interest of the class, and enables the remaining students to follow the material written on the boards. The remaining students seated in the class are, furthermore, too occupied with copying the material from the boards, to listen to the oral explanations and concentrate on the methods of solution. This book is intended to aid students in topology overcoming the diffi- culties described, by supplying detailed illustrations of the solution methods which are usually not apparent to students. The solution methods are illus- trated by problems selected from those that are most often assigned for class work and given on examinations. The problems are arranged in order of complexity to enable students to learn and understand a particular topic by reviewing the problems in sequence. The problems are illustrated with de- tailed step-by-step explanations, to save the students the large amount of time that is often needed to fill in the gaps that are usually found between steps of illustrations in textbooks or review/outline books. ivThe staff of REA considers topology a subject that is best learned by allowing students to view the methods of analysis and solution techniques themselves. This approach to learning the subject matter is similar to that practiced in various scientific laboratories, particularly in the medical fields. In using this book, students may review and study the illustrated prob- lems at their own pace; they are not limited to the time allowed for explain- ing problems on the board in class. When students want to look up a particular type of problem and solution, they can readily locate it in the book by referring to the index which has been extensively prepared. It is also possible to locate a particular type of problem by glancing at just the material within the boxed portions. To facilitate rapid scanning of the problems, each problem has a heavy border around it. Fur- thermore, each problem is identified with a number immediately above the problem at the right-hand margin. To obtain maximum benefit from the book, students should familiarize themselves with the section, “How To Use This Book,” located in the front pages. Special thanks are due to Dr. Nathan Busch for his technical editing of the book. Max Foaiet, Ph.D. Program DirectorHOW TO USE THIS BOOK This book can be an invaluable aid to students in topology as a supple- ment to their textbooks. The book is subdivided into 19 chapters, each deal- ing with a separate topic. The subject matter is developed beginning with the fundamental concepts of sentence calculus, sets, mappings, and extending through metric and topological spaces, continuity, homeomorphisms, axi- oms, compactness, connectedness, and homotopy theory. TO LEARN AND UNDERSTAND A TOPIC THOROUGHLY 1. Refer to your class text and read the section pertaining to the topic. You should become acquainted with the principles discussed there. These principles, however, may not be clear to you at that time. 2. Then locate the topic you are looking for by referring to the “Table of Contents” in front of this book, “The Topology Problem Solver.” 3. Turn to the page where the topic begins and review the problems under each topic, in the order given. For each topic, the problems are arranged in order of complexity, from the simplest to the more difficult. Some problems may appear similar to others, but each problem has been selected to illustrate a different point or solution method. To learn and understand a topic thoroughly and retain its contents, it will be generally necessary for students to review the problems several times. Repeated review is essential in order to gain experience in recognizing the principles that should be applied, and in selecting the best solution technique. TO FIND A PARTICULAR PROBLEM To locate one or more problems related to a particular subject matter, refer to either the index, located at the back of the book or the indexes found at the beginning of each chapter. In using the indexes, be certain to note that the numbers given refer to problem numbers, not page numbers. This ar- rangement of the indexes is intended to facilitate finding a problem more rapidly, since two or more problems may appear on a page. If a particular type of problem cannot be found readily, it is recom- mended that the student refer to the “Table of Contents” in the front pages, viand then turn to the chapter which is applicable to the problem being sought. By scanning or glancing at the material that is boxed, it will generally be possible to find problems related to the one being sought, without consuming considerable time. After the problems have been located, the solutions can be reviewed and studied in detail. For this purpose of locating problems rapidly, students should acquaint themselves with the organization of the book as found in the “Table of Contents.” In preparing for an exam, locate the topics to be covered on the exam in the “Table of Contents,” and then review the problems under those topics several times. This should equip the student with what might be needed for the exam. viiThis page is intentionally left blankCONTENTS Chapter No. Page No. 1. INTRODUCTION TO SENTENCE CALCULUS 1 2 ALGEBRAOFSETS 27 oe 13 14 15 16 17 18 19 EQUIVALENCE RELATIONS, AXIOMATIC SET THEORY, QUANTIFIERS 60 MAPPINGS 90 POWER OF A SET, EQUIVALENT SETS 123 CARDINAL NUMBERS, CARDINAL ARITHMETIC 144 AXIOM OF CHOICE AND ITS EQUIVALENT FORMS 180 ORDINAL NUMBERS AND ORDINAL ARITHMETIC 219 FUNDAMENTAL CONCEPTS OF TOPOLOGY 250 METRIC SPACES 374 TOPOLOGICAL SPACES 413 CONTINUITY, HOMEOMORPHISMS, AND TOPOLOGICAL EQUIVALENCE 458 SEPARATION AXIOMS 499 CARTESIAN PRODUCTS 532 COUNTABILITY PROPERTIES 546 COMPACTNESS 568 CONNECTEDNESS 614 METRIZABLE SPACES 663 ELEMENTS OF HOMOTOPY THEORY 700 GLOSSARY 724 INDEX 727This page is intentionally left blankCHAPTER 1 INTRODUCTION TO SENTENCE CALCULUS Sentences, Sum, Product, Negation Aristotelian Logic DeMorgan’s Laws Implication Equivalent Statements Syllogism Law Contraposition Duns Scotus Law Clausius Law Examples Law of Absorption Symmetric Difference Digital Systems Logic Gates and Boolean Algebra AND, OR, INVERTER gates NOR, NAND gates TIL ICs Design of Circuits 1-1 1-2 1-3, 1-4 1-5, 1-6 1-5 1-6 1-6 1-6 1-6 1-7, 1-8, 1-9 1-10 1-11 1-12 1-13, 1-14 1-15 1-16 1-13, 1-14, 1-16@ PROBLEM 1-1 Let a, B, y, ... denote sentences, each of which has one of two logical values 0 or 1. We assign the value 0 to a false sentence and 1 to a true sentence. To express the fact that a is a true sentence we write a=1 or B=0 a if B is false. We consider here only sentences of mathematical nature, i.e., sen- tences which take values either 0 or 1. For example, the sentence: “two times two is four” is (no doubt) a true one. On the other hand, the statement: “Goethe is a great poet” is not a sentence even though most people would agree that he indeed was a great poet. Define the sum, the product and negation of sentences and establish the truth tables for each case. SOLUTION: If @ and £ are two sentences, then the sentence “a or B,” denoted by a v 6, is called the sum of o and £6. The sentence “a and B,” denoted by a a 8, is called the product of @ and £. Clearly, the sentence o v 6 is true if at least one of the components is a true proposition. It has 2 x 2 = 4 logical possibilities. Table 1 a 6 avp I 1 1 1 0 1 0 1 1 0 0 0 Sentence a a B is true if both factors are true sentences. Table 2 a 6 anBp 1 1 1 i 0 0 0 1 0 0 0 0The sum and product of sentences are commutative and associative, i.e. avBp=iva, anBpefpaa (2) av(Bvy)s(avB)vy, an(Pay)=(anB)ay. (3) The distributive law holds aan (Bvy)=(an 8) v (ay). (4) In the above formulas we used the equivalence sign =. The equivalence a = 6 holds, if and only if a and B have the same logical value. The operation of negation of a sentence a denoted by a’ (or by ] a or ~ @) is de- fined in such a way that if a is true, then a’ is false, and if a is false, then a’ is true. Table 3 From this we obtain the law of double negation av sa. (5) @ PROBLEM 1-2 Using the logical notation introduced in Problem 1-1, write down two fundamental theorems of Aristotelian logic. Prove thata v 1=1,a00 =0. SOLUTION: The first theorem is called the law of the excluded middle (principium tertii exclusi) and, in classical logic, is formulated in the following manner: From two contradictory sentences, one is true. Let @ represent a sentence, then «’ is its contradiction. The law can be writ- ten as ava'al. (1) The second theorem, called the law of contradiction, states: no sentence canbe true simultaneously with its negation. It can be written briefly as follows: ana’ s0. (2) Let us examine the truth table for the statement a v 1: a avi il 1 0 1 Hence, a v 1 is a statement which is always true. aviel. (3) Similarly, the truth table for a 0 is a aad 1 0 0 0 and and=0. @ @ PROBLEM 1-3 So far, we defined three operations on sentences: product, sum and negation. By applying DeMorgan’s laws we see that the number of these fundamental operations can be reduced to two. SOLUTION: DeMorgan’s laws play an important role in mathematical logic. They can be formulated as follows: (av B)’ sa’ ap’ (1) (an) sa’vp. (2) The first law (1) asserts that, if it is not true that one of the sentences a and B is true, then both of these sentences are false and conversely. Taking the negation of (1) we obtain: (av B)’ sav Be=(a'np). (3) 4Equation (3) can be treated as a definition of the sum of sentences. Here, the sum is defined in terms of two operations: product and nega- tion. Hence, we limit the number of fundamental operations to two, product and negation. Similarly, taking the negation of (2) we obtain anBa(a’v py. (4) Now, the product is defined in terms of sum and negation. Again the number of fundamental operations is reduced to two. Product is defined with the aid of sum and negation. It is easy to verify DeMorgan’s laws using the truth tables. Table 1 for (a v 8)’ =a’ a’ a 6 (av B)’ a’ ap’ 1 1 0 0 1 0 0 | [ 0 0 1 0 | 0 0 0 1 1 Table 2 for (a a B)’ =a’ v B* a 6 (anB) av 6" 1 1 0 0 1 0 1 ( 1 0 1 1 | 1 0 0 1 1 @ PROBLEM 1-4 Generalize DeMorgan’s laws for n components. SOLUTION: DeMorgan’s laws state that (a, va, = 0,/ 00, (1) (0,44) =a,°va,”. (2) 5We shall generalize (1) for the system of n sentences: ,, a, .. (a,va,v..va,,va)s[(a,va,..va_)va] = (3) The sum of the sentences is associative. Applying (1) to (3) we find a(a,va,v..va,_)'aa,s =[(a,va,v..va,,)va,_]aa,= (4) Applying (1) again we obtain (a,va,v...va,_) aa, s(a,vav...va, aa’, ra. (5) Observe that applying the above procedure to the sentence (a, v a, v ... v a.,)’, (2— 1) times we find (a, Va,Vv...Va) =O, AG, A... AG,” (6) Now, take the negation of the product, (a, A a, ... 4 a,)”. Since the product is associative, we have (a,nan..aa) s[(a,rar..aa,)raa]= (7) Applying (2) to (7) we obtain B(Q, AGA. AG) Va, = (8) This procedure can be repeated again: a[(a,raar...aa ra ]ova,” = B(G,A..AQ,,) VO,,Va,= (9) After (1 — 1) steps we obtain (Q,AQ,A..AQ) Ba,°VO,V...Va,% (10) @ PROBLEM 1-5 Another important operation on sentences is implication. It is defined by (a= 6) =(o’ vB) a) The sign “=>” is read “implies” or “if ..., then ...”. 1. Write the truth table for the implication a = B. 2. Explain the difference between “if ..., then ...” as defined in a strictly mathematical way by (1) and everyday “if ..., then ....”3. Show that if a => B and B > a, thena =f. (2) SOLUTION: Let us start with the truth table for a = B. Table 1 a B (a = B) = (a vB) corer oror FeO 1. The implication a = f is always true, except when a. is true (a = 1) and B is false (B = 0). 2. Implication has properties similar to deduction. There is, however, a small but significant difference. Let @ denote the statement “The sun is shin- ing” and let B denote the statement “I am swimming.” The statement “If the sun is shining, then I am swimming” makes sense. When is such a statement false? When the sun is shining, I am not swimming, i.e., a 0 6” = 1. But, that is equivalent to V=0s(anpy=a'vB (3) in agreement with Table 1. In everyday language a sentence of the form “If a, then B” means that f is true whenever a is true. In such a case, the state- ment “If two times two is five, then Paris is the capital of Germany” is re- garded as nonsense, since both components are false. The same statement, in terms of formal language, not only makes sense but is also true. To each of the four possibilities shown in Table 1 (for a = 6), the system of formal language assigns either 0 or 1, even though two of the cases (a = 0 and B = 0, a = 0 and f = 1) appear to be pure nonsense in ordi- nary language. 3. To prove (2), note that (a = B) and (8 > a) = (av B) a (PB va)s = (AAR) VG aav(BABR)vV(Paas =(a’ 0B’) v (an). (4)The sentence (2° 4 B’) v (a a 6) is true when both components @ and B have the same logical value; i.e., they are both true or both false. Hence, if o> Band B >a, thena =. (5) Symbols <= and = may be used interchangeably. The symbol <> is read as: “if, and only if,” and indicates iniplication in both directions. Sometimes, instead of if, and only if, we write iff. Table 2 lists the symbols used so far. Table 2 Symbol Meaning a, B, y, 6, «.- logical sentences “or ~or] negation (not) sum (or) product (and) implication (if ..., then ...) iff (if, and only if) $< @ PROBLEM 1-6 1. Prove the syllogism law (sometimes called the law of implication): Ifa = B and B => y, thena => y. (qd) 2. Prove the law of contraposition, which is the basis of indirect proof (proof by contradiction, reductio ad absurdum): (a= 8) = >a). (2) 3. Prove the law of Duns Scotus: Ifa =0, thena > 6p. (3) 4. Prove the law of Clausius: If a’ = B for each B, then a = 1. SOLUTION: 1. Let us write (1) in the form [(@’ v B) a Bo v y= (@’ vy). 6)Note that the implication a = f is always true, except in the case when @ = 1 and B = 0. Suppose (a vB)a (QB vy)=1 and a’ vy=0. (6) From a’ v y = 0, we conclude that a = 1 and y = 0. Substituting into (6) we obtain (OvB)a (PB v0)=1 or PaPsl (72) which is a contradiction. 2. => We shall prove that (4>8)>=@' >a) (8) which is equivalent to (a= 8) v@' >a) (9) and to (av By v@Gvay (10) and to (an B) vo vB). qd) The last statement is true for any combintation of o and B. = (8 =a) = (a >) (12) or (vay va’ vB) (13) or (Baa) vw vB) aL. (14) a 3. Suppose a = 0, then a’ = 1. Therefore (a> B)=(a’ vB)el. (15) 4. If for B = 0 and f = 1 we have (a >8)=(avB)=1 (16) then ae=l. (17)Throughout this book we will be using a box (m™) to indicate the end of a proof, theorem, or definition. Note that there is a shorter way to prove (2) (a= 8)=(' va’. (18) Indeed, (a’ v B) =(B" va). (19) @ PROBLEM 1-7 Prove that: - asav~p SOLUTION: 1. We shall apply the definition of implication (see Problem 1-5, (1)) (a=>B)sa'vB. (5) (a>avB)sa’ vavBPa(a’ va)vps =slvBpal. (6) Here, we applied the law of excluded middle (see Problem 1-2, (1)). 2. (a>a)sa’vaal. ( 3. In (2) we can replace a by a a B to obtain (3). 4. (aa B= B)= (a8) vB= sa’vBpvpal. (8) In (8) we applied DeMorgan’s law (See Problem 1-3, (2)). 10Note that each of the above problems can be solved “automatically using the truth tables. Table a Bp avB anpasavp asa anap>anp anp=>Bh 11 1 1 1 1 1 1 1 0 1 0 1 1 1 1 ol 1 0 1 1 1 1 0 0 0 0 1 1 1 1 Now it should be clear why this system is called a formal language. @ PROBLEM 1-8 1. Prove that: (a= B)=(@ay=>Bay). 2. Prove that: (4=B)= (anB)v (aa). SOLUTION: 1. From the definition of implication, we conclude that statement (1) is equivalent to (a= By v(any>Bay) @) which is equivalent to (a v BY v [aay vay) 4 From DeMorgan’s law, we find that (4) can be replaced by (arB) v@vy)v Gay). 6) Statement (5) is always true. Indeed, it is tue when a’ = 1 or y’ = 1. Suppose a’ =O and y’ = 0, then a = 1 and y = 1 and for any B, (1 4B’) v (Ba 11. Thus GAB)v@vy)vyBayel. (6) 2. Remember that = can be replaced by <> (See Problem 1-5). Then (a=Pp)=(a@P)e(a>P)raP>a)= 11=(a’ vB) AB’ va) = (a AB) v (a Aa) Vv (BAB) V (AaB) = = (a 4B’) v (aA B). fo) As always, the proofs can be carried out by means of the truth tables. @ PROBLEM 1-9 1, Prove that if a => B, then anapsa and avBp=B. 2. Prove that if a => £ and y => 6, then any>Bnd avy=Bvs. SOLUTION: 1. Since a = f, we have a’ vpal (5) and aszanlsan(a vp)e(ana)v(an fe Ban fp. (6) Similarly, since a’ v B = 1 we obtain (a’ vB)’ san P' 21°20 (7) and BeBvO=Bv(anB)eavB. (8) 2. Since a = f, then for any sentence y we have any=>Bay. (9) See Problem 1-8, (1). Similarly, since y = 6, then Bay=>Bad. (10) 12Combining (9) and (10) we obtain, by virture of Problem 1-6, any=Bad. (11) To prove (4), note that (a> p)>@ >a) (12) => 8)= 6 =). (13) Applying (11) to (12) and (13), we find Pa sa'ny. (14) Applying the law of contraposition (Problem 1-6), (2)) to (14) we obtain from (14) (wavy = (Basy (15) or (avy) (B v8). (16) @ PROBLEM 1-10 Prove the law of absorption: azav(anp)san(av Bp) and its more general version av(Bay)s(av p)a(avy). SOLUTION: We have an(avB)s(ana)v(anP)sav (anf). (3) Now we shall show that asav(an). (4) Indeed, if @ = 1, then av(anp)slv(iap)=lea. (5) If a =0, then av(anB)=0v (0a B)=08a. (6) That completes (1). To prove (2), note that 13(av B)a(@vy) sfaravy]vBavyls ) By applying (1) we obtain sav [Ba(avyjsav (an pv Bay= s[av(anB)] vBays (8) Again, by applying (1), we obtain from (8) (av B)aGvy)sav (Bay). (9) @ PROBLEM 1-11 The symmetric difference of the sentences @ and B is defined by (a+ B)=[(arB)v @' a B)]- Prove that: avBp2((a+B)+(an B)]. SOLUTION: The easiest way to prove (2) is to establish the truth table and to compare the corresponding values. The truth table for a + B is: Table 1 a 6 anp aap a+ 1 1 0 0 0 1 0 1 0 1 0 i 0 1 1 0 0 0 0 0 Utilizing Table 1, we obtain Table 2 a 6 avB (a+B)+(anB) 1 1 1 1 1 0 il 1 0 1 i) 1 0 0 0 0 14By comparing the results of the last two columns of Table 2, we find that (2) is true. We can also carry out the proof of (2) without the use of the truth table. (a+ B)+ (a B)=[(a+B) a (ar By] v [+BY AGAB)] = = {[(an B) v (a’ a B)] 4 (a v BY} v {an B)v ap) a (an B)} < = (a vB) a (an BI] v [le vB) A (a vB) Y VGA BY a(a’ Ay aan) = = (aA B) v (a ABV [av B) ACVB) AAA B= = (aA BY) v (a AB)V (4B) = B(AaB) v[Ba(ava)]= By (aaB)eavB. @) @ PROBLEM 1-12 The digital (logic) circuits operate in the binary mode where each input and output voltage is either 1 or 0. This enables us to use Boolean algebra, which differs from ordinary algebra in that Boolean variables and constants are only allowed to have two possible values, 1 or 0. The voltage values in the circuit are predetermined, in the sense that any voltage in the range of OV to 0.5V corresponds to “0” and any voltage in the range of 2V to 6V corresponds to “1.” The values between 0.5V and 2V should not occur in the circuit. Table 1 lists some commonly used terms. Table 1 Logic 1 Logic 0 True False On Off High Low Yes No Closed switch Open switch 1. Describe the OR gate. 1S2. Describe the AND gate. 3. Describe the NOT gate (inverter). SOLUTION: 1. Ina digital circuit, an OR gate has two or more inputs, and one output is equal to the sum of the inputs. a B avBp=x 1 1 1 a 0 1 1 0 0 0 FIGURE 1 Figure 1 shows the symbol of the OR gate with two inputs and its truth table. An OR gate with 4 inputs is shown in Figure 2. Jor xsavBvyvi orz2 we FIGURE 2 2. Atwo-input AND gate and its truth table is shown in Figure 3. a B anp [ AND aarp 1 1 1 B e—- 1 0 0 0 1 0 0 0 0 FIGURE 3 163. The NOT gate (inverter) has one input and one output. FIGURE 4 @ PROBLEM 1-13 1. Replace the following digital circuit with the simpler one: FIGURE 1 2. Illustrate DeMorgan’s law using logic gates. SOLUTION: 1. The circuit shown in Figure 1 has two inputs, one output, and it performs the function described by (Bo a ay a (av By. qd) By applying DeMorgan’s laws and some basic properties of sentence calcu- lus, we obtain (8 na)’ (a vB) =(Bvaya(anB)= =[erB) AB] [ar B) Aa] B(anPaB)v(ana ap) 17=(an0)v (On B)=0v 020. (2) No matter what the input, the output is always 0. To construct a circuit equivalent to the one shown in Figure 1, all that is needed is wire. See Figure 2. ———__ Output FIGURE 2 Luckily, not all of the computer logic circuits can be simplified to such an extent. 2. We will build the circuit for (aa By =a’ v Bp. G3) SD | Ep (a) (b) FIGURE 3 Circuits (a) and (b) shown in Figure 3 are equivalent. Now we will build the circuit for (av py sa’ np. (4) . C3 (a) (b) FIGURE 4 Circuits (a) and (b) of Figure 4 are equivalent. 18@ PROBLEM 1-14 Design the circuit that implements the expression 1. [a B)v (Boa yy. 2. [ar Bayv(a rp) ay. SOLUTION: 1. Step by step we shall move a’, which is realized by a > a a’ a Bis realized by a by" B B’ is realized by B > Br B’ a yis realized by qi 19(a’ a B) v (B’ A y) is realized by 8 (a a B) v (B’ ay) [(@’ 4 B) v (B’ a y)J’ is realized by FIGURE 1 2. aaB ay is implemented by a’ a B’ is implemented by a’ A BY 20[(aa Bay) v (a’ a B)] Ay’ is implemented by o oaB (erBanv rbd (a aBany @ ABA FIGURE 2 Note that since [arBayv(@ ap ay= BCAABAYAY) VAP Ay) SQ ABI AY @) the circuit in the last part of Figure 2 can be replaced by FIGURE 3 @ PROBLEM 1-15 1. Define NOR and NAND gates. 2. Show that any logic circuit can be built exclusively with the NAND gates. SOLUTION: 1. NOR gate consists of OR gate and an INVERTER. 21a (av By’ a 5) > 0 (a) : 0 « (av By B a) > (b) FIGURE 1—NOR Gate 8 oore avBp (av Bp) ore i) Both figures represent the same logic circuit. The symbol in Figure (b) is used for convenience. NAND gate and its symbol are shown in Figure 2. a (aap) a s {| >> il (a) : 0 e 0 5 | »—— (b) FIGURE 2—NAND Gate B : 0 1 0 aap (aarp) 1 0 0 1 0 1 0 1 2. All Boolean expressions consist of various combinations of the basic operations of OR, AND, and INVERT. Thus, any circiut can be implemented using OR, AND, and INVERT gates. It is possible, however, to build any logic circuit using only NAND gates. Using NAND gates, we can design any logic operations OR, AND, and INVERT. 22(a) INVERTER Indeed (an ay za’ (b) AND Gate Indeed [Can B) a(anB)] s(anB)v(anP)sanp. a avp a avp (c) OR Gate Indeed [Can a) a (BaP) ]) (apy save. FIGURE 3 NAND gates can be used to perform any Boolean function. 23@ PROBLEM 1-16 1. Design a logic circuit that implements the function (a B) v (y 48). qa) 2. Suppose you have the TTL IC’s shown in Figure 1 at your disposal. Each IC is a quad; that is, each chip contains four identical two-input gates. Using the minimum number of IC’s, design the circuit which performs function (1). FIGURE 1c — IC 7400 24SOLUTION: ie R (@rB)v (yrds > ar FIGURE 2 Figure 2 shows the circuit implementating operation (a A B) v (y 4 5). 2. From Figure 2 we conclude that in order to build the circuit (1), we have to use two gates from IC 7408 and one gate from IC 7432, as shown in Figure 3. ° 7408 }783) Y 8 7408 FIGURE 3 Each AND gate and OR gate can be replaced by a combination of NAND gates (See Problem 1-15). AND OR —_—_—_———. a B 1 & ee Double Inversion FIGURE 4 25By eliminating double inversion, we obtain (aa B) v (yA 8) FIGURE 5 Now to implement the operation, we only need three NAND gates. There- fore, instead of using two IC’s as in Figure 3, we can use one IC 7400, as shown in Figure 5 and Figure 6. a B IC 7400 OUTPUT = (a 4 B) v (y 4 8) FIGURE 6 26CHAPTER 2 ALGEBRA OF SETS Sets, Elements Equal Sets Inclusion Venn Diagrams Set Operations Power Set Some Formulas Space, Complement DeMorgan’s Theorem Constituents Independent Sets Symmetric Difference Rings Ideals, Filters Russell Paradox Axiomatic Set Theory Boolean Algebra 27 2-1 2-2 2-3 2-4 2-4, 2-5, 2-6, 2-7 2-8, 2-9 2-10, 2-11 2-12 2-13, 2-14 2-15, 2-16 2-17 2-18, 2-19 2-20 2-21, 2-22 2-23 2-24 2-25@ PROBLEM 2-1 Which of the following statements is true? 1. 7€A, where A={x:xENand5 («EA iffx EC) or (A=B)\(B=C)>(A=0). 29@ PROBLEM 2-3 1. Show that inclusion of sets is reflexive, anti-symmetric, and transi- tive; ie. a. ACA for all sets A. b. ACBand BCA imply A = B for all A, B. ce. IfACBandBCC, then A CC for all sets A, B, and C. 2. Prove [(A CB) (BCC) A(CCA)] = (A =B=0). SOLUTION: DEFINITION: INCLUSION The set A is included in B; i.e., A is a subset of B, which is written as ACB or BCA, when xGA>xEB (2) a 1. Obviously (x€A) => (x EA) for all A. (3) Hence ACA. (xxXEB)a(xEB>xEA)=> (xEA iff x © B). (4) Hence (ACB) a(BCA)=>A2=B. (5) (xEA >xEB)n(xEB>xEC)=> (EA >xEO). (6) Thus (ACB) A(BCO)=ACC. o) 2. To prove that A = B, we must prove that x € A iff x © B. Since A CB, it is 30enough to show that B CA. But B CC and C CA, thus B CA. And ACBandBCA=>A=B. (8) Similarly, we show that B = C. (CCA andA CB) =>CCB. (9) Thus (B CCand CCB) >C=B. @ PROBLEM 2-4 Using Venn diagrams, illustrate the union, intersection and difference of two sets A and B. Show that intersection is associative; i.e., that AN(BNC)=(ANB)NC for all sets A, B, and C. SOLUTION: The union of two sets A and B is the set whose elements are all the ele- ments of the set A and all the elements of the set B and which does not con- tain any other elements. The union of A and B is denoted by A U B; see Fig- ure 1. FIGURE 1. A U Bis shaded. The common part of two sets is called the intersection. The intersection of A and B contains those, and only those, elements which belong to A and to B. The intersection is denoted by AN B. 31FIGURE 2. A B is shaded. The difference of two sets A and B is the set consisting of those, and only those, elements which belong to A, but do not belong to B. The difference is denoted by A —B (sometimes A | B). FIGURE 3. A — Bis shaded. Now we will use Venn diagrams to prove (1). FIGURE 4. From Figure 4 we conclude that AN(BNO)=(ANB)NC. 32@ PROBLEM 2-5 By applying the sentence calculus, prove the following formula: AU(ANB)=A= AN(AUB). SOLUTION: Operations on sets are related to operations on sentences. To denote that x is an element of A we write xEA. Q) Similarly («€A) = (EA) @) where x € A means x is not an element of A. The following equivalences hold for all x: (e€A UB) = (x EA) v (xB) (4) (x EA NB) = (x EA) a (xB) (5) (x A -B) =(@ EA) A EBY (6) Using formulas (3) - (6) we can deduce theorems on the calculus of sets from analogous theorems of sentence calculus. Furthermore, let us note that ifx€A = x€B holds for all x, then A = B. (7) We shall now prove (1). [xEA U(ANB)] =(xEA) vV(KEANB) = = (x EA) v [(x EA) 4 (XE B)] = (EA). (8) Thus AU(ANB)=A. (9) Similarly [xEAN (A UB)] =(x EA) A (EA UB) = = (xGA) a[(x GA) v (xEB)] = (EA). (10) Hence AN(AUB)=A. re8)) 33Note that in (8) and (10) we applied the law of absorption of sentence calculus. av(anp)zazaa(av fp) (12) @ PROBLEM 2-6 Prove these basic formulas for unions and intersections: 1. AUA=A indempotency ANA=A . AUB=BUA commutativity ANB=BNA . AU(BUC) =(AUB) UC associativity AN(BNC) =(ANB) NC . AU(ANB) =A adjunction AN(AUB) =A . AU(BNC) =(AUB)N(AUGQ) AN(BUC) =(ANB)U(ANG | distributivity SOLUTION: 1. (e€A UA) = (x EA) v (x EA) = (x EA) (6) Hence AUA=A. (7) (xEA NA) = (EA) A (x EA) & (EA) (8) Hence ANA=A. 9) 34From (7) and (9) we see that, in contrast to arithmetic, neither multiples nor exponents arise in set theory. 2. («EA UB) =(eEA) v(x EB) = = (x €B) v(x EA) = (x BUA). (10) Thus AUB=BUA. (11) (x EA NB)=(@EA)A(KEB)= = (x €B) a («EA) = (x BNA). (12) Hence ANB=BNA. (13) 3. [x€AUBUO]=[(KEA) v(x E BUC) = =[(xEA) v(xEB) v(xEO]=[(KEAUB)V(xEQ]= (14) =[x€(AUB)UC] Thus AU(BUC)=(AUB)UC. (15) Replacing “v” by “a” in (14) we prove AN(BNC)=(ANB)NC. (16) 4. See Problem 2-5. 5. [xEA U(BNO)]=[(KEA) vV(KEBNO)] = ={(eEA) v[%EB) a(EO)} = = {[@ EA) v(XEB)] a [XK EA) Vv (XEO)]} = =(xEAUB)a(xEAUC) = =[xE(AUB)N(AUC). a7) Hence AU(BNC)= (AUB)N(AUO). (18) [xEAN BUC] =[(xEA) a (KEBUO)] = = {(xEA) a [(e EB) vV(KEO)]} = 35= {[(@ EA) AEB) v [(KEA) a (XEC)]} = =[(xEA NB) v(KEANC] = =[xE(ANB)UANC)]. (19) Hence AN(BUQ= ANB)U(ANOC. (20) @ PROBLEM 2-7 1. Find the elements of the set Az {eixt1=3,22 1}. 2. Prove that: IfA is a subset of the null set 9, then A = 9. 3. Prove that: (A-B)NB=96. SOLUTION: DEFINITION OF THE NULL SET The null set (or the empty set or the void set) is the set which contains no elements. The null set is denoted by . Often the null set is defined as = {erx% x}. (3) 1. Number x, such that x + 1 = 3 and x? = 1 does not exist, hence A=o. (4) 2. The null set @ is a subset of every set, thus @ C A. But, by hypothesis, AC 4, therefore A=6. (5) 3. [xE(A-B)NB]= 36=[(«€A-B) n(@EB)]= = [(x GA) a (EB) a (XE B)] =0. (6) Thus, the set (A —B) M B contains no elements (A-B)NB=6. (7) @ PROBLEM 2-8 1. Find the power set P(A) of the set A={I, 2, 3}. 2. Find the power set P(B) of the set B= {1, {2, 3}}. (2) 3. Set A consists of 10 elements. How many elements does the power set P(A) have? SOLUTION: DEFINITION OF POWER SET The power set P(A) of A is the class of all subsets of A. 1. The subsets of A are: Oelements: lelement: {1}, {2}, {3} (3) 2elements: {1, 2}, {1, 3}, {2, 3} 3elements: {1, 2,3} Hence POA) = {, {1}, {2}, {3}, {1,2}, {1,3}, {2,3}, {L 2, 3}} (4) P(A) contains 2° = 8 elements. 2. Note that B contains only two elements, 1 and {2, 3}. Thus, its power set P(B) is P(B)= {9, {1}, {2,3}, (1, {2, 3}}}- 6) 373. From the theory of permutations we have the equation for the number of different combinations that can be formed from n different elements, using m elements at a time. n (1) * ware © where ml=l2>... -(a-l)n (7) and Ol=1. Suppose set A contains n elements; then there are n () = 1 subsets with 0 elements n! () = Tam-i n subsets with 1 element subsets with 2 elements. ()-agig The total number of subsets of a set A consisting of n elements is OQ eQe(Q-awer — @ Set A consists of 10 elements, hence its power set P(A) consists of 2'° = 1024. @ PROBLEM 2-9 If A is any non-empty set, then P(A) (or 24) is the power set of A. Prove that: 1. 2AM 2822408 q) 22.028 eae 3. Show that 24 U 28 #24U8 38SOLUTION: 1. Let a represent an element of 2* 9 2° . Recall that the power set 2* of A is the class of all subsets of A. Suppose a consists of elements p,, p,, ... = {Py Py +} (4) Since a € 24 and a € 2°, each of p,, P,, -.. belongs to A M B. Thus, a is a subset of A M B and ae 2408 (5) Hence, 2 2eC 2408 (6) Now, suppose a € 24%: i.e., a is a subset of A M B, and each element of a= {p,, P,, -..} belongs to A M B. Hence, each of p,, p,, ... belongs to A and to * Therefore, a is a subset of A, a € 24; and a is a subset of B, a € 2°. ae 2028, (7) Hence, ABCA 28, (8) From (6) and (8) we conclude that ZA =2AK 2, (9) 2. Suppose a € 24 U 2°; that is, a € 2 or a € 2°. Assume a € 24, then a is a subset of A. If a is a subset of A, then a is a subset of A U B and ae Ue, (10) That proves AUC Wve (11) 3. Now we shall show that Cl (12) Let A = {a}, B = {b}, then 2% = {o, {a}} (13) 2? = {6, {b}} (14) and A U B= {a, b} 2402 = {6, {a}, {b}, {a, b}}. (14) 39Hence, 24 U 2? = {6, {a}, {b}} and 2028 42 (15) (16) @ PROBLEM 2-10 Prove the following formulas: - ANBCACAUB . A-BCA . IfACBandC CD, then AUCCBUD and ANCCBND . (ACB) =(AUB=B)=(ANB=A) . (AUB)N(AUC)=AU(BNO SOLUTION: 1. [xEANB] #=[(x GA) a (&EB)] > (KE A) Thus, ANBCA (x EA) [(@ EA) v (x EB)] = (x EA UB) Hence, ACAUB. 2. [x€A-B] =[(x GA) v (x EB)] > (x EA) Hence, A-BCA. 40 a) @) ) (10) (11) (12)3. We have to show that AUCCBUD (AC B)a(CCD)] => and (13) ANCCBND [xE€A UC] = [(x EA) v(xEO)] > => [(xEB) v (xED)] =(xE BUD) (14) Hence, AUCCBUD. (15) @EANC) B[(KEA) A(KEQ] => => [@EB)a(xED)] =(@EBND) (16) Hence, ANCCBND. a7) 4, Suppose A C B; then from (A C B) and (B C B) we conclude that (AUB)C(BUB)=B (18) by virtue of (13). But since B C (A U B), we have AUB=B. (19) Conversely from A U B = B, it follows that A C B. Hence, the relations ACBandA U B= Bare equivalent. (ACB) =(AUB=B). (20) Similarly, combining A C B with A C A, we obtain A C (A £ B). Since ANBCA, A=ANB. (21) Conversely from A = A M B we conclude that A C B. 5. (AUB)N(AUC)=(ANA)UANCUBNA)UBNO= =AU(ANC)U(ANB)U(BNC) (22) But A M B CA, hence, A U (A MB) =A andA MCCA, hence, A U(A NC) = A. Thus, (AUB)N(AUC)=AU(BNO. (23) 41@ PROBLEM 2-11 Prove the following formulas: 1. ANB=A-(A- B) 2. AU(B-A)=AUB 3. A-(ANB)=A-B 4. AN(B-C)=(ANB)-(ANG) 5. AU(B-C)#(AUB)-(AUC) SOLUTION: aL Hence, [x€A-(A-B)] = [(@EA) 4 (KEA-B)] ® = {(x EA) a [(x EA) a (xB) } = = {(x EA) a [x EA)’ v @EB)]} = = [(x GA) a (x EB)] =(@EANB). [x€A U(B-A)] =[@ EA) v(x EB-A)] = = {(x EA) v [(@ EB) A (xEA) TP = = {[(@@ EA) v (x EB] a [(x EA) v (XE A) T]} © =[(x EA) v (x EB)] =(@EAUB). AU(B-A)=AUB. [xEA -(ANB)] =[(xEA) a (®EANB)] = = {(xEA) A [(@ EA) A (x EB) } = = {(x EA) a [x EA)’ v (EB) ]} = 42 (6) (7) (8)= {[(@ EA) A (EA) ] v [( EA) A (EB) FE = [(x EA) a (x B)'] =(xEA-B). (9) Hence, A-(ANB)=A-B. (10) 4. [x€(ANB)-(ANO] =[(KEANB)A(EANC)] = = {(x EA) a (xEB) A [(K EA) A (KEC)]} ® = {(x A) a (xB) a [(x EA)’ V(EC)]} = = [(x EA) a (x EB) a (x EC) = = [(e EA) a (x EB-O)] =[x EAN (B-O)]. (11) 5. Suppose A, B, C are three disjoint sets as shown in the Venn diagram: © OO) FIGURE 1 Then AU(B-C)=AUB (12) and (AUB)-(AUC)=B (13) Hence, AU(B-C)#(AUB)-(AUC). (14) 43@ PROBLEM 2-12 In the theory of sets, we assume that all the sets under consideration are subsets of some fixed set, called the space (or universal set). For ex- ample, in geometry the space is the Euclidean space, and in analysis the space is the set of real or complex numbers. Throughout this book, all the sets considered will belong to the space which we shall denote by (1). Hence AC1 qd) for each of the sets considered. DEFINITION OF COMPLEMENT OF THE SET The complement of the set A with respect to the given space 1 is denoted by AC (or CA or ~A) and defined by AC=1-A See Figure 1. FIGURE 1. (x GAS) = (x GA) =X EA) Prove that: 16-9, C= AC =A AUAC=1, ANAC=OSOLUTION: From definition (2) we have 1=1-1=6 (1) o=1-o=1 CS) AC = (A = (1-A)° = =1-(1-A)=A (9) AUAC=AU(1-A)=1 (10) ANAC=AN(1-A)=6 qt @ PROBLEM 2-13 Prove DeMorgan’s Theorem: 1, (AUB)E = ACN BE 2A NB =ACUBC SOLUTION: L [x€(A UBY] = [xE1-(AUB)] = =[xEA UB] =[(x EA) v (WEB) = = [(x GAY A (x EB) ] 8 [(x EAD a (EB) = = [xE ACN BS. (3) Hence, (A UB) =ASN BS. (4) 2. Similarly [xE(ANB)] = [xE1-(ANB)]= 45=[xEA NB] =[(xEA) a (xEB)] = = [(x EA)’ v (x EB)'] = [(x EAD v EB] = = [xE ACU BY. (5) Hence, (ANB) =A UBS. (6) DeMorgan’s theorem can be generalized to any finite number of sets. (A,UA,U...UA)PSA CN ALN... AS (2) (A, NAN... NA)OEACSUALU...UAS. (8) @ PROBLEM 2-14 Prove that: 1. A-B=ANBC qa) 2. ACB if and only if BC AC (2) 3. AC-BC=B-A . (3) 4. Prove Equation (4) of Problem 2-11 using the notion of comple- ment. SOLUTION: E («EA NB = [(KEA) AEB] = =[(x EA) a (x E1-B)] = [(x EA) A (x EB) ] = =(xGA-B). (4) 2. Here, we shall apply the law of contraposition: if «= B, then B’ => a (5) (ACB) =[(KEA) > (x EB)] = = [BY => (eEA)] = [@ EB) => EAD] = 46= [(x © 1-B) > (xE 1-A)] = [(x EB) > (XE AD] # = (BCA). (6) Thus, ACB iff BOCAS. 1) 3. (x GAC - BY) = [(x EAS a (XE BY] = = [(x€1-A) s(xE1-B)'] = [(k EA) A (EB) J = = [(x © B) a (x EA)] = (« EB-A). (8) 4. We have to prove that (ANB)-(ANC)=AN(B-C) (9) By applying (1), we obtain (ANB)-(NC)=ANB)N(ANCL= =(ANB)N[1-(ANO]=(ANB)N (ASUC) = =(ANBNAQU(ANBNC)= =(ANACNB)U(ANBNC)= =ANBNCC=AN(B-C). (10) In (10), we used DeMorgan’s theorem. @ PROBLEM 2-15 Let Aj, Ag, ..., A, represent fixed subsets of the space 1. Let us denote Aj = 1-A,= AC Af =A, (1) A constituent of the space with respect to the sets Aj, Aa, ..., A, is any set of the form a a tn A/ NA, 9..04, 47where ij, ..., i, = 1 or 0. (2) 1. Show that constituents are disjoint. 2. Prove that the union of all constituents is equal to 1. SOLUTION: a a i, ‘ A &, 1. Let A, 9A, 9...NA, and A, OA, O...A, denote two con- stituents. Then at least one index i, exists, such that i,#k, @) Then the intersection of these two constituents is a 4 ‘a a ‘ ky 4 ‘, A) NA} ...04,,04;'04,0...04,"04/ 04) = A, NAS = 9. (4) 2. Now we shall show that the union of constituents is equal to 1. Let x € 1 represent an element of the space. We shall show that a con- stituent exists, such that x belongs to this constituent. Consider the sets A,,A,, +++) A, Two possibilities exist: xGA, or xGA,. IfxEA, =A", we set i, = 0. If x €A,, then x E1—A, =A,' and we set i, We repeat the above procedure for each setA,, A A, and obtain i, i, i,, The element x belongs to i xGA! % xGA, i, xGA, (5) Hence, i i i ct—7,V tay. Vea) ee (Al. (6) Since x is an arbitrary element of 1, we conclude that the union of con- stituents is equal to 1. 48@ PROBLEM 2-16 Let A, B, and C represent the subsets of the space. 1. Find the constituents of the space with respect to the sets A, B, and C. 2. Represent the set A-(BUC) (1) as the union of the constituents. SOLUTION: 1. We shall consider all possible sets of the form i i i A'NB* NC’ (2) where i,, i,, i, = Oor 1 and A°=A, B=B, C=C A'=AS, B'= BS C=C (3) The constituents are ANBNC ANBNC. ANBSNC ASNBNC ANBENC ASN BENC ASN BNC ASN BNC. (4) 2. Let us write A —(B U C) in the form A-(BUC)=(4-B)-C=(ANB)-C= =ANBENC.. (5) Note that AM BC M C‘ is itself a constituent. 49@ PROBLEM 2-17 The sets Aj, Ap, ..., A, are called independent if all the constituents are non-empty. Independent sets play an important role in the probability theory. Find the number of constituents of the independent sets A,, Ao, ..., A ne SOLUTION: The constituents are B ar fy A, NASN..NA, () where ij, iz, ..., i, take one of two values, 0 or 1. A? =A; and Aj! = 1-A; (2) The sets Aj, Ao, ..., A, are independent, therefore the constituents are non- empty sets. Setting 0 (3) izi=. in (1) we obtain A, NAQN...NA,#O- (4) Thus, none of the sets A;, Ao, ...,A, are empty sets. Similarly, setting i; = iz =... i, = 1, we find that AL NASON... NAS #0. (5) That is, none of the sets A,, Az, ...,A,, are whole spaces. Each of the superscripts i,, iz, ..., i, in i i i A} NA, n...94, takes up one of two values. Thus, the total number of combinations is 2x2x wx 2=2" (6) The number of constituents is 2". @ PROBLEM 2-18 The symmetric difference of the sets A and B is defined by A+B=(A-B)U(B- A). a) 50Prove the following formulas: 1, A+B=B+A 2. A+(B+OC)=(A+B)+C The symmetric difference is associative. 3. AUB=A+B+ANB SOLUTION: We shall use A-B=AN Bo to write (1) in the form A+B=(ANB9 U(ACNB). From (6) we obtain (A +B)o= (AN BIN (ASN BY = = (ANB) U(ACN BO. L A+B=(A-B)U(B-A) =(B-A) U(A-B)= =B+A. 2. A+(B+C)=[AN(B+CY]U[(B+OC) NAY = (2) 6) (6) ” (8) 2) ={A N[(CNB)U (CON BY} U {[(BNCY U (BEN OC] NA} = =(ANBNC) VANBENCY) UATNBNC) U(ASN BNC). Similarly, (A+B)+C=[(ANB)U (Ao NB) +C= = {[(A NB) U (ACN B)] NC} U L(A + BY NC] = (10) =(ANBEN CHU (ASN BNCOU {[(A 1B) U (ACN BO] NC} = =(ANBENCYU(ASTNBNCHU(ANBNOU (AEN BEN C). Comparing (9) and (10) we find 51(A+B)+C=A+(B+O)=A+B+C. (11) 3. A+B+ANB=(A+B)+(ANB)= =[(AN BYU (ACN B)] +(ANB)= = {[(A NB U (ACN B)] N(46U BF U U{{(A5UB)N AUBIN GNB)}= = (AN BY N (AE U BO] U [(AE NB) N (ACU BO] U U[ASNBYNANB) U[ANB)NANB)= =(AN BYU (ACN B)U (ANB) = =(A4- B)U(B-A) U(ANB)=AUB. (12) Similarly, we can show that A-B=A+ANB. (13) @ PROBLEM 2-19 Show that (A, UU... UA,) + (B, U... UB,) C(A, +B) U... U(A,*B,) (1) where A+B=(A-B)U(B-A). 2) SOLUTION: Let us denote A=A,U...UA, B=B,U...UB, (3) Suppose x€A+B=(A-B) U(B-A). (4) 52The sets A — B and B — A are disjoint sets (i.e., they have no common ele- ments). We can assume that x€A-B, then x GA and (x ¢ B) = (x © B)’. We have xEA,U...UA, and x@B,U...UB,. Therefore a set A, exists, such that x€A, and (EB) =(xEB,U...UB) = =(@EB)v@EB)v...v@eeEB)) =(@EB) A (xEB) A... A (xEB)’ =(xEB)A(KEB)A...A(eEB). Thus, xGA, andx€B, and x GA,+B,=(A,-B) U (B,-A). Hence, (A,...94)+6,N...B)C,+B)U...UG, +B). (6) 6) 0) (8) (9) (10) (11) (12) @ PROBLEM 2-20 DEFINITION OF A RING The Operations + and - form a commutative ring, if they satisfy the conditions: 1. xty=yex, x-ysyrx 2 xt(y+za(xty)tz x-(V-Za(e-y)-z 53 Q) (2)3. x-(y+z)=(x'y)+(x°2) (3) 4. An element 0 exists, such that x+0=x (4) 5. For every pair x, y, an element z exists, such that x=ytz (5) a Show that the sets form a ring with respect to the operations + and 1 but don’t form a ring with respect to the operations U and N. SOLUTION: , Let us replace in (1) - (5) operation + by ~, and operation - by M. We ob- tain A+B=B+A,ANB=BNA (6) which are obviously true. Then A+(B+C)=(A+B)+C, AN(BNO) =(ANB)NC. (7) The relationship AN(B+C)=(ANB)+(ANC) (8) is true. For an empty set ¢, we obtain A+o=A. (9) To prove condition 5, we have to show that (A=B+C)>(C=B~+A) (10) which is easy to prove because A=B+C=B+B+A=(B+B)+A=A., (11) Hence, operations + and M form a ring. Consider operations U and M. We have AUB=BUA, ANB=BNA (12) AU(BUCQ)=(AUB)UC, AN(BNC)=(AN=B)NC. (13) AN(BUC)=(ANB)U(AN ©) (14) AU {o} =A (5) 54It is not true that for every pair A, B, a set C exists, such that A=BUC. (16) For example, let A = {} and C # {¢}. Thus, U and M do not form a ring. @ PROBLEM 2-21 1. Show that the family of all subsets of a given set A is an ideal. 2. Show that the family of all sets B, such that A C B C 1, isa filter. SOLUTION: DEFINITION OF IDEAL A family is a collection of sets. A non-empty family / of subsets of 1 is called ideal if (AED a(BCA)>BET qa (AEN A(BEN=>(AUBED. (2) 1. Let P denote the family of all subsets of a given set A. If D € P, then DCA. If G CD, then G CA and G &€ P. Hence, condition (1) is fulfilled. If D € Pand GE P, then D CA and G CA. Hence, DUG CA and DU GE P. Therefore, the family of all subsets of a given set is an ideal. DEFINITION OF A FILTER A non-empty family F is called a filter if (AEF) A(ACB)>(BEF) @) (AEF) A(BEF) => (ANBEF) (4) 7 2. Let A represent a given set and R the family of all sets B, such that ACBC 1. Suppose DER and DCG, thnACDCGC1andGER. Condition (3) is fulfilled. If DE Rand G ER, then A CD andA CG. Therefore, A CDN G and DNGER. Hence, R is a filter. 55@ PROBLEM 2-22 Show that a family of sets is filter if and only if the family of the complements of these sets is an ideal. SOLUTION: Let S = {A, B, C, ...} represent a family of sets. We shall prove that ( = {A,B,C,...} ‘| (’ . WC) => a filter an ideal Let AC € P and BC C AS, then A C B. But A ES (which is a filter); hence, BES. Therefore B¢ € P. Let AC € P and BS EP, then A ES and B ES. Since S is a filter, ANB ES. But AC U BS = (ANB) EP. Similarly, we can show that (’ = {A‘,B,... "" (’ = {A,B,C,...} is an ideal a filter Let A © S and A CB. Then AC & P and BS C A“. Hence, BC € P and B eS. Let A ES and B ES. Then AC & P and BCE P. Since P is an ideal, AC U BE Pand ANB (ACUBYES. @ PROBLEM 2-23 In 1895, Georg Cantor created a theory of sets. At that time, it was accepted that a universal set (that is, “the set of all sets”) exists. In 1902, Bertrand Russell showed in his famous paradox that the admission of a set of all sets leads to a contradiction. Explain Russell’s paradox. SOLUTION: We shall show that the assumption that the universal set exists leads to contradictory statements. THEOREM 1 Suppose that the set of all sets, denoted by W, exists. Let 56A={xEWw:x Ex} Then or AgeA AeA Indeed, (x © A) = (x Ex); therefore (A € A) = (A EA). That leads to: THEOREM 2 A set of all sets does not exist. This fact was briefly described by Paul Halmos in this statement: “Noth- ing contains everything.” New paradoxes of set theory appeared shortly after Russell’s paradox. To overcome this difficulty, mathematicians created sev- eral variations of axiomatic set theory. None of these theories was com- pletely satisfactory, nevertheless they eliminated most of the antinomies. @ PROBLEM 2-24 Consider the axiomatic formulation of the algebra of sets. It consists of the concept of element, of set and of the relation of an element belong- ing to a set (x € A) and four fundamental axioms. Using this system, show that 1. the null set exists. 2. for any two sets A and B, only one set A U B exists. SOLUTION: We shall list the four fundamental axioms. |. UNIQUENESS AXIOM (Axiom of Extension). If the sets A and B have the same elements, then A and B are identical. 57I. UNION AXIOM. For two arbitrary sets A and B, a set exists which contains all the ele- ments of the set A, and all the elements of the set B and which does not con- tain any other element. Ul. DIFFERENCE AXIOM. For two arbitrary sets A and B, a set exists which contains those and only those elements of the set A which are not elements of the set B. IV. EXISTENCE AXIOM. At least one set exists. 1. Let us define the null set by the formula o=A-A. The existence of at least one set is guaranteed by Axiom IV. 2. For given sets A and B, one and only one set satisfying Axiom II exists. Hence, the operation U is unique. Since ANB=A-(A-B), it is not necessary to include an axiom on the existence of an intersection. The intersection can be defined in terms of the difference. @ PROBLEM 2-25 SOLUTION: Note that in most of the theorems of the algebra of sets the symbol € does not appear, although it occurs in their proofs. This suggests the possibil- ity of establishing the system of axioms, which will enable us to prove the theorems without referring to the relation €. BOOLEAN ALGEBRA The fundamental concepts are: 58The set > and the operations U, N, -. We assume the following axioms: 1. AUB=BUA 2. ANB=BNA 3. AU(BUC)= (AUB)UC 4. AN(BNO= (ANB)NC 5. AUO=A 6. ANM(AUB)=A 7. AU(ANB)=A 8 AN(BUC= (ANB)UANG 9 (A-B)UB=AUB 10. BN(A-B)=6 Ml. AN1=A 12. (ACB) =(AUB=B) From the above axioms we can deduce all the theorems of the algebra of sets. Axiom 12 defines the Inclusion Theory of sets as only one of the applications of Boolean algebra. Replacing in axioms the notion of a set by sentences, we obtain the algebra of sentences. This explains the close rela- tionship between the algebra of sentences and the algebra of sets. Sentence Sets v corresponds to U A corresponds to n * (negation) corresponds to ¢ (complement) By omitting axioms 8 - 12, we obtain the axiomatic formulation of the lattice theory, applied, for example, in electrical networks. 59CHAPTER 3 EQUIVALENCE RELATIONS, AXIOMATIC SET THEORY, QUANTIFIERS {909} Quantifiers V and 3 Ordered Pair Cartesian Product Properties of Cartesian Product Family of Sets Union and Intersection Properties of the Union and Intersection Relations Diagonal, Inverse Relations Composition of Relations Equivalence Relation Equivalence Class Canonical Mapping (Projection) Sentence Functions of Two Variables Axiomatic Formulation, Axiom of Choice 3-1 3-2, 3-3, 3-4 3.5 35 3-6, 3-7, 3-8 39 3-10 3-11, 3-12 3-13 3-14 3-15, 3-16 3-17, 3-18, 3-19, 3-20 3-21, 3-22 3-23 3-24, 3-25 3-26@ PROBLEM 3-1 1, The set of all values of the variable x, for which @(x) is a true sen- tence, is denoted by the symbol fx: @@)}. Q The following equivalence holds: For every a: [a € {x : g(x)}] = G(a). (2) Using (1) and (2), describe the set of the real numbers which are larger than zero and smaller than or equal to 5. 2. Prove the following formulas: {x : (x) v E(x)} = {xP} U {x : CO} fx: E(x) A POa)} = {x 2 CO} 1 {x: G)} {x P(x) 4 (EQ) = fe: OCD} - {e : C@)} and {x: [p(x)]"} = {x: @@)}E. SOLUTION: 1. The set can be described as follows: {x: (ER) a (x>0) a (xs5)}. (6) Z a & {x: g(x) v C(x)} = (a) v a) = = [a {x: o(x)}] v [2€ fx: O@) H = 2a E {x: G0} U (x: C0}. oO) We employ the strategy used in the solution of (3) to prove (4). aE {x: E(x) A GD} = Ea) A (a) = = (aE {x: Ca} [a E fx: G0) Hs aa {x:C@)}N fx: G0}. ) Proving (5), a E {x2 92) a [CIF = 9a) 4 [CC] = 61=[a& {x: @)}] 4 [a E {x:6@) Hs = aE {x: g(x} — (x: Ga}. @) Proving (6), a € {x: [9(%)]} = [9@] = a(a El)a [a¢ {x: G(x) }] = =a {x: g(x} (10) The symbol {x : g(x)} will be used very often, sometimes with some modifi- cations, throughout this book. @ PROBLEM 3-2 1, Using qualifiers, write the sentence: For every natural number, a larger natural number exists. 2. Explain the quantifiers by using the calculus of sentences. SOLUTION: 1. The quantifier “there exists” is denoted by J and the quantifier “for each” is denoted by V. Let A represent the domain of a general statement p(x), i.e., x EA. Then _— a asserts that for all x © A, the statement p(x) is true. Similarly, dxEA: p(x) (2) means that at least one x € A exists, such that p(x) is true. By denoting the set of natural numbers by N, we write Vadm:n [x : p@)). ©) @ PROBLEM 3-3 1. Using quantifiers write the statement: For each x there exists a y such that for all z the statement p(x, y, z) is true. 2. Write the negation of this statement. SOLUTION: We shall apply the following rule of quantifier negation: [Vx: p@)]’ = Ax: [p@)]” (dy (2x: p@) = Vx: POY 2) Equations (1) and (2) are generalized DeMorgan formulas. Note that by taking negation of (1), we can define a universal quantifier in terms of the existential quantifier and negation Vx: p(x) = (Ax: PO). @) Similarly, existential quantifiers can be defined in terms of the universal quantifier and negation Ax: p(x) = [Wx: [PQ] T- 4) 1. This statement can be written as Wx Ay Wz: p(x y, z). (5) 2. The negation of (5) is [Wx dy Wz: p(x y, 2) = = [Wx [y Vz: p(x y, 2)]] = = Ax [By Vz: psy, OY (6) 63Here we used (1). = ah [Ay [Vz: pte y, 2D] = ax Vy [Vz: p(x y, 2)’ = dx Vy az: [p(s y, 2)] = (7) @ PROBLEM 3-4 Using quantifiers, write down the definitions of: 1. Limit of a sequence of real numbers. 2. Continuous functions. 3. Uniformly continuous functions. SOLUTION: 1. Let (x,) denote a sequence of real numbers. This sequence is said to have a limit x. We write lim x, =x (1) ne when Ve>0 ImEN VWnem |x -x\0 F6>0 VxER! |x-x,|<8=>|f)-fx)| 0, 5 > 0 exists such that for every xER', if |x—x,| <6, implies | f(x) —f(x,) |<. 3. Function f: R' — R'is said to be uniformly continuous on R' if Ve>0 35>0 Vx,x ER! |[x-x'|[f%)-fx)| and B = 9, then elements a, b exist, such thata © A and b € B. Therefore, (a, b) EA x BandA x Bz. FIGURE 1. 3. We shall prove that (AxBCCxD) (ACC) a (BCD). (3) => Let (a, b) GA x B. Then (a, b) EC x D. Therefore, if aE A, thena EC. Similarly, if b € B, then bE D. <= Let (a, b) EA x B. Then a EA and a EC. Also bE Band b€ D. There- fore (a, b) € C x D and AxBCCxD. (4) @ PROBLEM 3-7 Show that Ax(BUC)=AxBUAxC Ax(BNC)=AKBNAxC Ax(B-C)=AxB-AxC 67