Notation and Abbreviation N = {1,2,3,...}, (p.5) N* = {0,1,2,3,...}, (p.104) Ni = {1,2,3,...,k}, (p.62) Ni = {0,1,2,3,...,k}, (p91) Zz = {...,-2,-1,0,1,2,3,...}, (p.3) R = the set of real numbers, (p.187) (AP) : Addition Principle, (p.1) (BP) : Bijection Principle, (p.27) (CP) : Complementation Principle, (p.16) (IP) : Injection Principle, (p.27) (MP) : Multiplication Principle, (p.4) (PP) : Pigeonhole Principle, (p.120) (GPP) : Generalized Pigeonhole Principle, (p.133) (PIE) : Principle of Inclusion and Exclusion, (p.146) (GPIE) : Generalized Principle of Inclusion and Exclusion, (p.150) (RP) : Reflection Principle, (p.91) LHS : Left hand side, (p.192) RHS : Right hand side, (p.151) o : if and only if, (p.79) iff : if and only if, (p.95) viialb a fb Lz] fz] a=b (mod m) HCF LCM Isl a(r,n) S(r,n) P(rsriy..- Ta) Notation and Abbreviation a divides 6, (p.94) a does not divide 6, (p.94) the largest integer less than or equal to x (p.94) the smallest integer greater than or equal to z (p.124) a is congruent to b modulo m, i.e. m|(a — b) (p.94) highest common factor, (p.113) Lowest common multiple, (p.147) the number of elements in the finite set S, (p.2) Stirling number of the first kind the number of ways to arrange r distinct objects around n identical circles such that each circle has at least one object, (p.25) Stirling number of the second kind the number of ways of distributing r distinct objects into n identical boxes such that no box is empty, (p.47) t the rth Bell number = > S(r,n), (p.50) the number of r-element subsets of an n-element set ! Aer (p.17) the a of r-permutations of n distinct objects ce aoe? (#277), @37 r! rial ral? P34Dn D(n,r,k) P(X) v(n) R(p, 9) R(pi,.--, Pn) (osrnvtn) 11,125 009m, »(n) MO IMO APMO AIME Notation and Abbreviation ix the number of r-circular permutations of n distinct objects = (p18) the number of derangements of Nj, (p.160) the number of r-permutations of Nj, that have ex- actly k fixed points, (p.160) the power set of X, (p.28) the Euler y-function, (p.160) the smallest natural number “n” such that for any colouring of the edges of an n-clique by 2 colours: blue or red (one colour for each edge), there exists either a “blue p-clique” or a “red q-clique”, (p.132) the smallest natural number “n” such that for any colouring of the edges of an n-clique by k colours: colour 1, colour 2, ..., colour k, there exist a colour i(i=1,2,...,k) and a pj-clique in the resulting con- figuration such that all edges in the p;-clique are coloured by colour i, (p.136) ! aio , (p.96) the number of different partitions of n, (p.196) Mathematical Olympiad International Mathematical Olympiad Asian Pacific Mathematics Olympiad American Invitational Mathematics ExaminationContents Notation and Abbreviation . . Contents 1. Permutations and Combinations 1.1. Two Basic Counting Principles 1.2. Permutations 1.3. Circular Permutations 1.4. Combinations .... 1.5. The Injection and Bijection Principles ............. 7 1.6. Arrangements and Selections with Repetitions ........ 32 1.7. Distribution Problems ...............--.0000. 40 Exercise 1 2. Binomial Coefficients and Multinomial Coefficients .... 69 2.1. Introduction ©... 2... eee 2.2. The Binomial Theorem 2.3. Combinatorial Identities . 2.4. The Pascal’s Triangle .... 2.5. Chu Shih-Chieh’s Identity . . . 2.6. Shortest Routes in a Rectangular Grid” ausiee a . 85 2.7. Some Properties of Binomial Coefficients ........... 93 2.8. Multinomial Coefficients and the Multinomial Theorem . . 96 WGXORCIRG Desist itt sese ste ceeces eet tei ics 102 3. The Pigeonhole Principle and Ramsey Numbers .... . . 119 3.1. Introduction... 2.6... ee 119 3.2. The Pigeonhole Principle .................-54% 119 O37 |More Examples sien coe 122 3.4. Ramsey Type Problems and Ramsey Numbers ........ 129 3.5. Bounds for Ramsey Numbers ..............--0- 132 PExercasels cect sient testa tenrteare aera eee aene 137xii 4, Contents The Principle of Inclusion and Exclusion . 4.1. Introduction 4.2. The Principle ... . 4.3. A Generalization .. . - 145 4.4. Integer Solutions and Shor Routes ..........--- 153 4.5. Surjective Mappings and Stirling Numbers of the Second Raa Se re ieteeeee a teteaet oe estat esreeete et ceeteteaete a etterteeraeeaeteietee 158 4.6. Derangements and A Generalization .............. 160 4.7. The Sieve of Eratosthenes and Euler y-function ....... 163 4.8. The ‘Probléme des Ménages’ .............20005 169 Exercise 40... eee 173 Generating Functions ............-0--0 00000 e ee 185 5.1. Ordinary Generating Functions . 5.2. Some Modelling Problems .. . . 5.3. Partitions of Integers . . . ‘ 5.4. Exponential Generating Functions... Exercise 5 Recurrence Relations .. 1... 6... eee eee ee eee 225 clo: Introduchion tatsisrcsasrere setae termites sie reiericrne cece 225 6.2. Two Examples ....... 2.0.0 e eee eee eee 228 6.3. Linear Homogeneous Recurrence Relations .......... 234 6.4, General Linear Recurrence Relations ............. 241 6.5. Two Applications ... 2.2... 0.05.0 eee eee ee 244 6.6. A System of Linear Recurrence Relations ........... 251 6.7. The Method of Generating Functions . . 6.8. A Nonlinear Recurrence Relation and Catalan Numbers » . 259 6.9. Oscillating Permutations and an Exponential Generating PMAMCHION feet escictce te tee ieree te tecetietet eee atierietce te telietiercete teeta 262 Exercise 6 000... eee ee te tee eee 270 Bibliography, sees eee 287 BA as afr ease ae elepaeseoaee eerreaerier verreeniteerlarer ee teeenienraerearintes 289Chapter 1 Permutations and Combinations 1.1. Two Basic Counting Principles In our everyday lives, we often need to enumerate “events” such as, the arrangement of objects in a certain way, the partition of things under a certain condition, the distribution of items according to a certain specifi- cation, and so on. For instance, we may come across counting problems of the following types: “How many ways are there to arrange 5 boys and 3 girls in a row so that no two girls are adjacent?” “How many ways are there to divide a group of 10 people into three groups consisting of 4, 3 and 2 people respectively, with 1 person rejected?” These are two very simple examples of counting problems related to what we call “permutations” and “combinations”. Before we introduce in the next three sections what permutations and combinations are, we state in’ this section two principles that are fundamental in all kinds of counting problems. The Addition Principle (AP) Assume that there are my ways for the event Ej, to occur, n2 ways for the event 2 to occur, nx ways for the event Ej, to occur, where k > 1. If these ways for the different events to occur are pair- wise disjoint, then the number of ways for at least one of the events E,, E2,..., or Ey to occur is ny +ng+--- +n, = hin.2 Section 1.1. Two Basic Counting Principles Example 1.1.1. One can reach city Q from city P by sea, air and road. Suppose that there are 2 ways by sea, 3 ways by air and 2 ways by road (see Figure 1.1.1). Then by (AP), the total number of ways from P to Q by sea, air or road is2+34+2=7. y Figure 1.1.1. An equivalent form of (AP), using set-theoretic terminology, is given below. Let Ai,Ao,...,Ax be any k finite sets, where k > 1. If the given sets are pairwise disjoint, ie., A; A; = @ for i,j = 1,2,...,k, i # J, then k |A1 U Ag U-+-U-Aa] = > Ail. i=l Example 1.1.2. Find the number of ordered pairs (z,y) of integers such that 2? +y? <5.Chapter 1. Permutations and Combinations 3 Solution. We may divide the problem into 6 disjoint cases: 2? + y? = 0,1,...,5. Thus for i = 0,1,...,5, let Si = {(z,y) |z,yEZ, 2? +y? =i}. It can be checked that So = {(0,0)}, Si = {(1,0), (-1,0), (0,1), (0,-1)}, Se = {(1,1), (1,-1),(-1,1),(-1,-)}, 53 = 0, Sa = {(0,2), (0, -2), (2,0),(—2,0)}, and Ss = {(1,2), (1, -2), (2, 1), (2,-1); (-1, 2), (-1, -2), (2,1), (-2, -1)}. Thus by (AP), the desired number of ordered pairs is 5 Vo isil=1+44+440444+8= 21. #=0 Remarks. 1) In the above example, one can find out the answer “21” simply by listing all the required ordered pairs (x,y). The above method, however, provides us with a systematical way to obtain the answer. 2) One may also divide the above problem into disjoint cases: 2? = 0,1,...,5, find out the number of required ordered pairs in each case, and obtain the desired answer by applying (AP). The Multiplication Principle (MP) Assume that an event E can be decomposed into r ordered events Ei, F2,...,£,, and that there are n, ways for the event E, to occur, m2 ways for the event 2 to occur, n, ways forthe event E, to occur. Then the total number of ways for the event E to occur is given by: my X Ng X +X My =4 Section 1.1. Two Basic Counting Principles Example 1.1.3. To reach city D from city A, one has to pass through city B and then city C as shown in Figure 1.1.2. A }_, k? = 3n(n + 1)(2n + 1), we finally obtain Is] = i x 99 x 100 x 199 = 328350. » As mathematical statements, both (AP) and (MP) are really ‘trivial’. This could be a reason why they are very often neglected by students. Ac- tually, they are very fundamental in solving counting problems. As we shall witness in this book, a given counting problem, no matter how complicated it is, can always be ‘decomposed’ into some simpler ‘sub-problems’ that in turn can be counted by using (AP) and/or (MP). 1.2. Permutations At the beginning of Section 1.1, we mentioned the following problem: “How many ways are there to arrange 5 boys and 3 girls in a row so that no two girls are adjacent?” This is a typical example of a more general problem of arranging some distinct objects subject to certain additional conditions. Let A = {a}, a2, ...,¢n} be a given set of n distinct objects. For 0 (n|n € S) in part (ii). Observe that the 4 numbers in 5; can be paired off as {1,7} and {3,5} so that the sum of the two numbers in each pair is equal to 8 and the 12 numbers in Sz can be paired off as {13,75}, {15,73}, {17,71}, {35,53}, ... 80 that the sum of the two numbers in each pair is 88. Likewise, the 24 numbers in S3 and the 24 numbers in Sq can be paired off so that the sum of the two numbers in each pair is equal to 888 and 8888 respectively. Thus 4 12 24 24 a= 8x 5 + 88x > + 888 x > + 8888 x a = 117856. 1.3. Circular Permutations The permutations discussed in Section 1.2 involved arrangements of objects in a row. There are permutations which require arranging objects in a circular closed curve. These are called circular permutations. Consider the problem of arranging 3 distinct objects a,b, c in 3 positions around a circle. Suppose the 3 positions are numbered (1), (2) and (3) as shown in Figure 1.3.1. Then the three arrangements of a,b,c shown in the figure can be viewed as the permutations: abe, cab, bea respectively. a c. b () ()) Q) c bob aa ¢ (3) (2) (3) (2) (3) (2) abe cab bea Figure 1.3.1.Chapter 1. Permutations and Combinations 13 In this case, such “circular permutations” are identical with the usual permutations, and thus there is nothing new worth discussing. To get something interesting, let us now neglect the numbering of the positions (and thus only “relative positions” of objects are concerned). As shown in Figure 1.3.2, any of the 3 arrangements is a rotation of every other; i.e., the relative positions of the objects are invariant under rotation. In this case, we shall agree to say that the 3 arrangements of Figure 1.3.2 are identical. In general, two circular permutations of the same objects are identical if any one of them can be obtained by a rotation of the other. abe cab bea Figure 1.3.2. Let A be a set of n distinct objects. For 0 < r < n, an r-circular permutation of A is a circular permutation of any r distinct objects taken from A. Let Q? denote the number of r-circular permutations of A. We shall derive a formula for Q?. Example 1.3.1. Let A= {a,b,c,d}. There are altogether P$(= 24) 3-permutations of A and they are shown in Example 1.2.1. These 24 3- permutations are re-grouped into 8 subsets as shown below: abe _cab_bea ach bac cba abd dab bda | adb bad dba acd_dac _cda | ade cad dca bed _dbc_cdb | bde cbd deb It is noted that every 3-circular permutation of A gives rise to a unique14 Section 1.8. Circular Permutations such subset. For instance, a = — {acd,dac, eda} Conversely, every such subset corresponds to a unique 3-circular permuta- tion of A. For instance, {adb,bad,dba} = => Thus we see that Example 1.3.1 tells us that Q3 = $P}. What is the relation between Q® and P? in general? A circular permutation of r distinct objects 21, 29,...,2, shown below: a gives rise to a unique subset of r r-permutations: ©1XQ+++ Lp, LpX{LZQ-++ Ley, -.., Lares Lp TyChapter 1. Permutations and Combinations 15 obtained through a rotation of the circular permutation. Conversely, every such subset of r r-permutations of A corresponds to a unique r-circular permutation of A. Since all the r-permutations of A can be equally divided into such subsets, we have gam. (1.3.1) r In particular, pr Qh = <2 =(n-1)!. (1.3.2) Example 1.3.2. In how many ways can 5 boys and 3 girls be seated around a table if (i) there is no restriction? (ii) boy By and girl G; are not adjacent? (iii) no girls are adjacent? Solution (i) The number of ways is Q§ = 71. (ii) The 5 boys and 2 girls not including G1 can be seated in (71)! ways. Given such an arrangement as shown in Figure 1.3.3, G; has 5(= 7 — 2) choices for a seat not adjacent to B,. Thus the desired number of ways is 6! x 5 = 3600. Bs G: 7 2 By By Bs G3 Figure 1.3.3. ‘We may obtain another solution by using what we call the Principle of Complementation as given below:16 Section 1.3. Circular Permutations Principle of Complementation (CP) If A is a subset of a finite universal set U/, then W\ Al = Bal 1AL Now, the number of ways to arrange the 5 boys and 3 girls around a table so that boy B, and girl G, are adjacent (treating {B,,G,} as an entity) is (7-1)! x 2= 1440. ‘Thus the desired number of ways is by (CP), 7! — 1440 = 3600. (iii) We first seat the 5 boys around the table in (5 — 1)! = 4! ways. Given such an arrangement as shown in Figure 1.3.4, there are 5 ways to seat gitl G;. As no girls are adjacent, Gz and G3 have 4 and 3 choices respectively. Thus the desired number of ways is 4!x5x4x3=1440. g Ba By Figure 1.3.4. Example 1.3.3. Find the number of ways to seat n married couples around a table in each of the following cases: (i) Men and women alternate; (ii) Every woman is next to her husband.Chapter 1. Permutations ond) Combinations 17 Solution. (i) The n men can first be seated in (n — 1)! ways. The n women can then be seated in the n spaces between two men in n! ways. Thus the number of such arrangements is (n — 1)! x n!. (ii) Each couple is first treated as an entity. The number of ways to arrange the n entities around the table is (n — 1)!. Since the two people in each entity can be permuted in 2! ways, the desired number of ways is (n=1)!x 2". Remark. A famous and much more difficult problem related to the above problem is the following: How many ways are there to seat n married couples (n > 3) around a table such that men and women alternate and each woman is not adjacent to her husband? This problem, known as the problem of ménages, was first introduced by the French mathematician Francis Edward Anatole Lucas (1842 — 1891). A solution to this problem will be given in Chapter 4. 1.4. Combinations Let A be a set of n distinct objects. A combination of A is simply a subset of A. More precisely, for 0 < r < n, an r-combination of A is an r-element subset of A. Thus, for instance, if A = {a,b,c,d}, then the following consists of all the 3-combinations of A: {a,b,c}, {a,b,d}, {a,c,d}, {b, c,d}. There are 4 in number. Let C? or (") (which is read ‘n choose r’) denote the number of r-combinations of an n-element set A. Then the above example says that Cf = (3) = 4. We shall soon derive a formula for C?. What is the difference between a permutation and a combination of a set of objects? A permutation is an arrangement of certain objects and thus the ordering of objects is important, whereas a combination is just a set of objects and thus the ordering of objects is immaterial. As a matter of fact, every r-permutation of A can be obtained in the following way: Step 1. Form an r-combination B of A. Step 2. Arrange the r objects of B in a row.18 Section 1.4. Combinations ‘This provides us with a means to relate the numbers P” and C?. Indeed, we have by (MP): PP =OP xr! and thus = geese eee =r A@en! (144) In particular, Note that nt —r)\(n-(n—r))! Chars ()=(2) tas For convenience, we show in Table 1.4.1 the values of ("), where 0 G-Dimaonl A _ (n—Dir+(n = rin Combinatorial Proof. Let A = {1,2,...,n}. By definition, there are (") ways to form r-combinations S of A. We shall count the number of such S in a different way. Every r-combination S of A either contains “1” or not. If 1 € S, the number of ways to form S is (77). If 1 ¢ S, the number of ways to form Sis ("7"). Thus by (AP), we have (=a) C5) = + | r r-1 r Remark. In the second proof, we fix an enumeration problem and count it in two different ways, giving rise to an equality relating two different. expressions. This is a useful way to derive combinatorial identities. Example 1.4.2. By Example 1.1.4, there are 2” binary sequences of length 7. How many such sequences are there which contain 3 0’s and 4 1's? Solution. To form such a sequence of length 7: EEE ey (1) (2) (3) (4) (5) 6) (7) we first select 3 of the 7 spaces for ‘0’ and then leave the remaining spaces for ‘1’. There are (3) ways in the first step and (3) = 1 way in the next. Thus the number of such binary sequences is given by (3). =20 Section 1.4. Combinations Remarks. (1) In the above example, you may first select 4 of the 7 spaces for ‘1’ and obtain the answer (7), which is equal to (3) by identity (1.4.2). (2) In general, the number of binary sequences of length n with m 0’s and (n — m) 1’s, where 0 < m < n, is given by (*). Example 1.4.3. In how many ways can a committee of 5 be formed from a group of 11 people consisting of 4 teachers and 7 students if (i) there is no restriction in the selection? (ii) the committee must include exactly 2 teachers? (iii) the committee must include at least 3 teachers? (iv) a particular teacher and a particular student cannot be both in the committee? Solution. (i) The number of ways is (*2) = 11!/(5!6!) = 462. (ii) We first select 2 teachers from 4 and then (5 — 2) students from 7. The number of ways is 4) (7 () () =6 x 35 = 210. (iii) There are two cases: either 3 teachers or 4 teachers are in the committee. In the former case, the number of ways is ()@)=+-nm while in the latter, the number of ways is @@=* Thus by (AP), the desired number of ways is 84+ 7 = 91. (iv) Let T be the particular teacher and $ the particular student. We first find the number of ways to form a committee of 5 which includes both T and S. Evidently, such a committee of 5 can be formed by taking the union of {7,5} and a subset of 3 from the remaining 9 people. Thus the number of ways to form a committee of 5 including T and S is (§) = 84.Chapter 1. Permutations and Combinations 21 Hence the number of ways to form a committee of 5 which does not include both T and S is by (CP): (2) 2 () = 462—84=378, by (i). # Suppose that there are 8 players a,b,c,d,e, f,g,h taking part in the singles event of a tennis championship. In the first round of the competition, they are divided into 4 pairs so that the two players in each pair play against each other. There are several ways to do so. For instance, or () avsb, evsf, dvsh, evsg, (2) avsh, bvsg, cvsf, dvse. What is the number of ways that this arrangement can be made? To phrase this sort of questions mathematically, let A be a set of 2n distinct objects. A pairing of A is a partition of A into 2-element subsets; ie., a collection of pairwise disjoint 2-element subsets whose union is A. For instance, if A is the set of 8 players {a,b,c,d,e, f,9,h} as given above, then {{a,0}, {c, f}, {4,4}, fe, 9}} and {{a,h}, {5,9}, {c, f}, {d, e}} are different pairings of A. We note that the order of the subsets and the order of the 2 elements in each subset are immaterial. Example 1.4.4. Let A be a 2n-element set where n > 1. Find the number of different pairings of A. Solution. We shall give 3 different methods for solving the problem. Method 1. Pick an arbitrary element, say 2, of A. The number of ways to select z’s partner, say y, is 2n—1 (and {z,y} forms a 2-element subset). Pick an arbitrary element, say z, from the 2n—2 elements of A\{z, y}. The number of ways to select z’s partner is 2n — 3. Continuing in this manner and applying (MP), the desired number of ways is given by (2n — 1)(2n —8)---5-3-1.22 Section 1.4. Combinations Method 2. First, form a 2-element subset of A and put it in position (1) as shown below. There are (7s) ways to do so. eee eyed) Q @ (8) (n) Next, form a 2-element subset from the remainder of A and put it in the position (2). There are (°"7”) ways to do so. Continuing in this manner and applying (MP), we see that the number of ways of arranging the n 2-element subsets in a row is: Qn\ (2n—2\ (4) (2 2 2 2) \2} Since the order of the n subsets is immaterial, the desired number of ways is: GCs) | Method 3. We first arrange the 2n elements of A in a row by putting them in the 2n spaces as shown below: , bf , how , Q) ) (3) (4) (2n—1) — (2n) There are (2n)! ways to do so. Since the order of the elements in each 2-element subset and the order of the n subsets are immaterial, the desired number of ways is given by (2n)! _ _(2n)! 2! x 2! x2xn! nlx 2° ee es It can be checked that the above 3 answers are all the same. The above problem can be generalized in the following way. Let A be a set of kn distinct elements, where k,n € N. A k-grouping of A is a partition of A into k-element subsets; i.e., a collection of pairwise disjoint k-element subsets whose union is A. Thus if A = {a1,a2,...,a12}, then {{a1, 44,49, 412}, {a2, 45,48, 410}, {a3, 46, a7, a11}} is a 4-grouping of A. Clearly, a pairing of a 2n-element set A is a 2-grouping of A. What is the number of different k-groupings of a set with kn elements? (See Problem 1.43.)Chapter 1. Permutations and Combinations 23 Example 1.4.5. (IMO, 1989/3) Let n and k be positive integers and let S be a set of n points in the plane such that (i) no three points of S are collinear, and (ii) for any point P of S, there are at least k points of S equidistant from P. Prove that k << $+ -V2n. Proof. For convenience, we call a line segment in the plane an edge if it joins up any two points in S. Let £ be the number of edges in the plane. We shall consider the quantity £. First, since there are n distinct points in S and any two of them deter- mine an edge, we have, n t= 6) (1) Next, for each point P of S, by condition (ii), one can draw a circle with centre P whose circumference C(P) contains at least k points of S. Clearly, the points of S on C(P) determine at least (3) edges. As there are n points P in S, the total number of these edges, counted with repetition, is at least n(*). Now, let us look at those edges which are counted more than once. An edge is counted more than once when and only when it is a common chord of at least 2 circles. Since two circles can have at most one common chord and there are n such circles, the number of common chords, counted with repetition, is at most (5). Thus ee)-6) ° Combining (1) with (2), we have “()-@)<() 0) <6) BP k-2(n-1) <0. or which implies that24 Section 1.4. Combinations Hence ke 1+ V1+8(n-1) Ht 2 11 1 1, 8(r,r) =1 ifr >0, 8(r,1)=(r—1)! for r > 2, s(r,r—1)=(5) forr>2. The following result, which resembles (1.4.3) in Example 1.4.1, tells us how to compute s(r,n) for larger r and n from smaller r and n.26 Section 1.4. Combinations Example 1.4.7. Show that 8(r,n) = 8(r —1,n—1) +(r— 1)s(r — 1,n) (1.4.4) where r,n € N with n 1 and s(r,1) = (r—1)! for r > 1, and applying the identity (1.4.4), one can easily find out the values of s(r,n) for small r and n. For r,n withO B from A to B is injective (or one-one) if f(a1) # f(a2) in B whenever a; # a2 in A. f is surjective (or onto) if for any b € B, there exists a € A such that f(a) =b. f is bijective if f is both injective and surjective. Every injective (resp., surjective and bijective) mapping is also called an injection (resp., a surjection and a bijection). The Injection Principle (IP) Let A and B be two finite sets. If there is an injection from A to B, then |A| < |BI. The Bijection Principle (BP) Let A and B be two finite sets. If there is an bijection from A to B, then |A| = |B]. Just like (AP), (MP) and (CP), the two principles (IP) and (BP) as given above are also trivially true. However, as we will see below, they are also useful and powerful as tools for solving counting problems. Example 1.5.1. A student wishes to walk from the corner X to the corner Y through streets as given in the street map shown in Figure 1.5.1. How many shortest routes are there from X to Y available to the student? Y Figure 1.5.1.28 Section 1.5. The Injection and Bijection Principles Solution. Let A be the set of all shortest routes from X to Y. We shall find [A]. We first note that every route in A consists of 7 continuous segments (a segment is part of a street connecting two adjacent junctions) of which 4 are horizontal and 3 vertical. Thus if we use a ‘0’ to denote a horizontal segment and a ‘1’ to denote a vertical segment, then every route in A can be uniquely represented by a binary sequence of length 7 with 4 0’s and 3 1’s (for instance, the shortest route shown by bold line segments in Figure 1.5.1 is represented by 1001100). This way of representing a route clearly defines a mapping f from A to the set B of all binary sequences of length 7 with 4 0’s and 3 1’s. It is easy to see that f is both one-one and onto, and hence it is a bijection from A to B. Thus by (BP) and Example 1.4.2, we have l4|=|B1=@)- o Remark. The street map of Figure 1.5.1 is a 5 x 4 rectangular grid. In general, if it is an (m+ 1) x (n + 1) rectangular grid consisting of m+1 vertical streets and n + 1 horizontal streets, then the number of shortest routes form the southwest corner X to the northeast corner Y is equal to the number of binary sequences of length m+n with m 0’s and n 1’s, which by Remark (2) of Example 1.4.2, is given by ("a") = 2") Given a set X, the power set of X, denoted by P(X), is the set of all subsets of X, inclusive of X and the empty set 0. Thus, for instance, if X = {1,2,3}, then P(X) = {0, {1}, {2}, {3}, {1,2}, {1,3}, {2,3}, X}. We observe that |P(X)| = 8. In general, what can be said about |P(X)| if X consists of n distinct elements? Example 1.5.2. Show that if |X| = n, then [P(X)| = 2” for all neN.Chapter 1. Permutations and Combinations 29 Proof. We may assume that X = {1,2,...,n}. Now, let B= {aa9...d, |; =0 or 1, i=1,2,..,n} be the set of all binary sequences of length n. Define a mapping f : P(X) — B as follows: For each S € P(X) (ie., SCX), we put F(S) = bybe.. Bp where : ie i ifi eS, 0 ifi¢S. (For instance, if X = {1,2,3,4,5}, S1 = {4} and S2 = {2,3,5}, then F(S1) = 00010 and f(S2) = 01101.) It is easy to see that f is a bijection from P(X) to B. Thus by (BP), |P(X)| = |B]. Since |B| = 2” by Example 1.1.4, we have [P(X)| = 2", as required. Example 1.5.3. Let X = {1,2,...,n}, where n € N. Show that the number of r-combinations of X which contain no consecutive integers is given by (ort?) . , where 0 m for each i = 2,3,...,r. Thus S is 2-separated if and only if S contains no consecutive integers. Let X = {1,2,...,n}, where n € N. Using (BP), we can also find a formula for the number of r-element subsets of X which are m-separated. Readers who are interested may see Problem 1.91. In the above three examples, three sets A (namely, the set of shortest routes, the power set P(X) and the set of r-combinations containing no consecutive integers) and their counterparts B (namely, the set of binary sequences of length 7 with 4 0’s, the set of binary sequences of length n and the set of r-combinations, respectively) are considered. By establishing a bijection from A to B, we have by (BP), |A| = |B]. Note that in each case, the enumeration of |A| by itself is not straightforward, while that of |B| is fairly standard and so much easier. Thus with (BP), we can sometimes transform a hard problem into an easier one. This is always a crucial step in the process of solving a problem. The above three examples involve only the applications of (BP). In what follows, we shall give an interesting example which makes use of both (BP) and (IP).Chapter 1. Permutations and Combinations 31 Example 1.5.4. (IMO, 1989/6) A permutation 2122...22n of the set {1,2,...,2n}, where n € N, is said to have property P if |z; — zi41| =n for at least one i in {1,2,...,2n—1}. Show that, for each n, there are more permutations with property P than without. This problem, proposed by Poland, was placed last in the set of the six problems for the 1989 IMO, and was considered a more difficult problem among the six. The original proof given by the proposer makes use of recurrence relations which is rather long and looks hard. However, it was a pleasant surprise that a contestant from the China team was able to produce a shorter and more elegant proof of the result. Before we see this proof, let us first try to understand the problem better. Let S = {1,2,...,2n}. Clearly, a permutation of S does not have prop- erty P if and only if for each k = 1,2,...,n, the pair of numbers k and n+k are not adjacent in the permutation. For n = 2, the set A of permutations without P and the set B of permutations with P are given below: A ={1234, 1432, 2143, 2341, 3214, 3412, 4123, 4321}, B ={1243, 1324, 13.42, 1423, 2134, 2314, 2413, 2431, 3124, 3142, 3241, 3421, 4132, 4213, 4231, 4312}. Clearly, |B| = 16 > 8 = |AI. Proof. The case when n = 1 is trivial. Assume that n > 2. Let A (resp., B) be the set of permutations of S = {1,2,...,2n} without property P (resp., with P). To show that |B| > |A|, by (IP) and (BP), it suffices to establish a mapping f : A + B which is injective but not surjective. For convenience, any number in the pair {k,n + k} (k = 1,2,...,n) is called the partner of the other. If k and n+k are adjacent in a permutation, the pair {k,n +k} is called an adjacent pair of partners. Let a@ = 2122...22, be an element in A. Since a does not have property P, the partner of z, is z, where 3 0 for all i, can be represented by a row having r; 0’s within the interval under a;. If we treat each vertical stroke as an ‘1’, then every r-element multi-subset of M corresponds to a unique binary sequence of length r + n — 1 with r 0’s and (n — 1) 1’s. This correspondence is indeed a bijection between the family of all r-element multi-subsets of M and the family of all such binary sequences. Thus by (BP) and Remark (2) of Example 1.4.2, we obtain the following result. (Il) Let M = {00+ 41,00 -43,...,00+@,}. The number H? of r-element multi-subsets of M is given by Ht= Ga) Result (III) can be proved in various ways. We give another proof below. Another Proof of (III). For convenience, we represent a; by i, i= 1,2,...,n, and so M = {c0-1,00-2,...,00-n}. Let A be the family of all r-element multi-subsets of M and B be the family of all r-combinations of the set {1,2,...,r-+n—1}. Define a mapping f : A > B as follows: For each r-element multi-subset S = {b1, b2,...,6,} of M, where 1 < b; < bo < wa S bp Sr, let F(S) = {b1,b2 +1, b3 + 2,-1.5 bp + (nr — 1)}.40 Section 1.7. Distribution Problems It should be noted that members in f(S) are distinct and so f(S) € B. It is easy to see that f is injective. To show that f is surjective, let T = {c1,¢2,...,¢,} be an r-combination of {1,2,...,7 +n — 1} with c, < c2 < +++ < ¢. Consider S = {e1,¢2 — 1,¢3 — 2.4 — (r— 1)}. Observe that $ is an r-element multi-subset of M and by definition, f(S) = T. This shows that f is surjective. We thus conclude that f is a bijection from A to B. Hence by (BP), r+n—1 Hp = |4l= [BI = ( 7 y Remarks. (1) Let M’ = {p1-a1,p2-a9, ..., Pn-Gn} be amulti-set. From the above discussion, we see that the number of r-element multi-subsets {ri -a1,72 +42; ..51n * dn}, (0 p; for some i, then the above statement is invalid. This case will be studied in Chapter 5. (2) The reader might have noticed that there is some similarity be-~ tween the above proof and that given in Example 1.5.3. Indeed, the rule defining f here is the same as that defining f~1 there. 1.7. Distribution Problems We consider in this section the following problem: Count the number of ways of distributing r objects into n distinct boxes satisfying certain con- ditions. We split our consideration into two cases: (1) objects are distinct, (2) objects are identical (or indistinguishable). Case (1) Distributing r distinct objects into n distinct boxes. (i) If each box can hold at most one object, then the number of ways to distribute the objects is given by n(n— 1)(n—2)++-(n—r+ 1) = PB,Chapter 1. Permutations and Combinations 41 since object 1 can be put into any of the n boxes, object 2 into any of the n—1 boxes left, and so on. (ii) If each box can hold any number of objects, then the number of ways to distribute the objects is given by as each object can be put into any of the n boxes. (iii) Assume that each box can hold any number of objects and the orderings of objects in each box count. In this case, the 1st object, say a1, can be put in any of the n places (namely, the n boxes); and the 2nd object, say az, can be put in any of the n + 1 places (the n — 1 boxes not containing a; plus the left and right positions of a; in the box containing a1). Similarly, the 3rd object can be put in any of the n + 2 places due to the presents of a; and a2, and so on. Thus the number of ways that an arrangement can be made in this case is given by n(n +1)(n +2)---(n+(r—1)). left a box 1 box 2 box 3 ot box n There is another way to solve the problem. As shown below, 4a: a: @3a27_ +—+ agasla;lasa2 box 1 box 2 box 3 n=3,r=5 one can establish a bijection between the set of such distributions of r dis- tinct objects a1, a2, ...,@, into n distinct boxes and the set of arrangements of the multi-set {a1,a2,...,4,,(n — 1) - 1} (we treate each vertical stroke separating adjacent boxes as a ‘1’). Thus by (BP) and result (II) in Section 1.6, the desired number of ways is given by (n-1+r)! (n-1)! ’42 Section 1.7. Distribution Problems which agrees with the above result. Case (2) Distributing r identical objects into n distinct boxes. (i) Assume that each box can hold at most one object (and thus r n); i.e., no box is empty. In this case, we first put one object in each box to fulfill the requirement (this can be done in one way), and then distribute the remaining r—n objects in the boxes in an arbitrary way. By (MP) and the result in (ii), the desired number of ways is given by RE) By identity (1.4.2), this can also be written as ayChapter 1. Permutations and Combinations 43 Example 1.7.1. How many ways are there to arrange the letters of the word ‘VISITING?’ if no two I’s are adjacent? Solution. Method 1. The letters used are V,S, T, N, G and 3 I’s. We first arrange V, S, T, N, G in a row. There are 5! ways. Take one of these arrangements as shown below. EV see Essen ve gee EEN HEREC CteHeCaaay 1st 2nd 8rd 4th 5th 6th ‘There are 6 spaces separated by the 5 letters. The problem is now reduced to that of distributing the 3 identical I’s in the 6 places such that each place can hold at most one I (no 2 I’s are adjacent). By Case (2)(i), the number ", by (§). Thus by (MP), the desired number of ways 6) In the above method, we first consider “V, S, T, N, G” and then 3 I’s. In the next method, we reverse the order. of ways to do so is giv is Method 2. We first arrange the 3 I’s in a row in one way: Ree ee eeeeeee gee eee eee 1st 2nd 3rd 4th Then treat the 5 letters “V, S, T, N, G” as 5 identical “x”. Since no 2 I’s are adjacent, one ‘z’ must be put in the 2nd and 3rd places (this can be done in one way): 1st 2nd 3rd 4th Now, the remaining 3 z’s can be put in the 4 places arbitrarily in (**4-1) = ($) ways by Case (2)(ii). Finally, by restoring the original letters that each “z” represents (so 5! ways to arrange them) and by applying (MP), the desired number of ways is given by (es44 Section 1.7. Distribution Problems Example 1.7.2. (Example 1.2.3(ii) revisited) In how many ways can 7 boys and 3 girls be arranged in a row so that the 2 end-positions are occupied by boys and no girls are adjacent? In Example 1.2.3, this problem was solved by considering the arrange- ment of boys followed by that of girls. In the following solution, we reverse the order. Solution. The 3 girls can be arranged in 3! ways. Fix one of them: — G& — G — Gs — Ist 2nd 3rd 4th Then treat the 7 boys as 7 identical “x”. To meet the requirements, one ‘x” must be placed in each of the 4 places separated by the 3 girls as shown below: z Giz G, « G3 Now the remaining 3 2’s can be put in the 4 places arbitrarily in (*+{-*) = (2) ways by Case (2)(ii). Finally, by restoring the meaning of ‘x’ and applying (MP), we obtain the desired number of ways: 6 at. (§) = 71-6.5-4 ' Example 1.7.3. (Example 1.5.3 revisited) Let X = {1,2,...,n}, where n € N. Show that the number of r-combinations of X which contain no consecutive integers is given by ("~1*"), where O 0 and n>1. An integer solution to the equation (1.7.1) is an n-tuple (€1, €2,...,€n) of integers satisfying (1.7.1) when 2; is substituted by e;, for each i = 1,2,...,n. Thus, for instance, (—1,—4,7),(2,0,0),(0,1,1),(0,2,0) and (0,0,2) are some integer solutions to the equation 21 +29 +23 = 2.46 Section 1.7. Distribution Problems There are infinitely many integer solutions to (1.7.1). In this section, we shall confine ourselves to “nonnegative” integer solutions (i-e., e; > 0, for all i). Example 1.7.4. Show that the number of nonnegative integer solu- tions to equation (1.7.1) is given by r+n-1 is : Proof. Every nonnegative integer solution (€1,€2,.-.,én) to (1.7.1) corresponds to a way of distributing r identical objects to n distinct boxes as shown below: en ° box n Clearly, different solutions to (1.7.1) correspond to different ways of distri- bution. On the other hand, every such way of distribution corresponds to a nonnegative integer solution to (1.7.1). Thus by (BP) and the result in Case (2)(ii), the desired number is given by | We now review all the problems in this chapter which give rise to the important number ("*"~?).Chapter 1. Permutations and Combinations 47 =e} The number of ways of selecting r objects from n different types of objects with repetitions allowed = the number of r-element multi-subsets of the multi-set {00 -a1,00-a2,...,00-an} = the number of ways of distributing r identical objects into n distinct boxes = the number of nonnegative integer solutions to the equation tatzete +e, =r Some problems of distributing objects (identical or distinct) into distinct boxes have just been studied. In what follows, we shall study a problem of distributing distinct objects into identical boxes. Problems of distributing identical objects into identical boxes will be discussed in Chapter 5. Given nonnegative integers r and n, the Stirling number of the second kind, denoted by S(r,n), is defined as the number of ways of distributing r distinct objects into n identical boxes such that no box is empty. The following results are obvious. O) i) (i) (wv) ) (vi) $(0,0) =1, S(r,0) = S(0,n) =0 for all rn € N, S(r,n)>0ifr>n>1, S(rn) =0ifn>r>1, S(r,1) =1forr>1, S(r,r) =1forr>1. We also have (see Problem 1.84):48 Section 1.7. Distribution Problems (vii) S(r,2) = 27-2 -1, (viii) $(r,3) = 4087-1 +1) - 2°, (x) S(r,r-1) = (9), (x) S(r,r—2) = (§) + 3(4). The following result bears some analogy to those given in Example 1.4.1 and Example 1.4.7. Example 1.7.5. Show that S(r,n) = S(r —1,n— 1) +nS(r—1,n) (1.7.2) where r,n € N with r>n. Proof. Let a be a particular object of the r distinct objects. In any way of distributing the r objects into n identical boxes such that no box is empty, either (i) a; is the only object in a box or (ii) a; is mixed with others in a box. In case (i), the number of ways to do so is S(r — 1,n—1). In case (ii), the r — 1 objects (excluding a) are first put in the n boxes in S(r —1,n) ways; then a; can be put in any of the boxes in n ways (why?). Thus the number of ways this can be done in case (ii) is nS(r —1,n). The result now follows by (AP). § Using some initial values of S(r,n) and applying the identity (1.7.2), one can easily construct the following table: Let A = {1,2,...,r}. For n € N, an n-partition of A is a collection {S1, S2,...,Sn} of n nonempty subsets of A such that () SS; =O for #5 and (ii) & S= A. A partition of A is an n-partition of A for some n = 1, 2,.. A binary relation R on A is an equivalence relation on A if , @Ra for all a € A, (ii) R is symmetric; i.e., if a,b € A and aRb, then bRa, and (iii) R is transitive; i.e., if a,b,c € A, aRb and bRe, then aRc. (i) Ris reflexive; ie.Chapter 1. Permutations and Combinations 49 Table 1.7.1. The values of S(r,n),0 S(r,n) = the number of partitions of {1,2,...,7} nat the number of equivalence relations on {1,2,...,r}. The sum )77,-; S(r,n), usually denoted by B,, is called a Bell number after E.T. Bell (1883 — 1960). The first few Bell numbers are: By =1, By =2, Bs =5, By = 15, By = 52, Be = 203,. Exercise 1 1. Find the number of ways to choose a pair {a,b} of distinct numbers from the set {1,2,...,50} such that (i) |e-b| = 5; (ii) Ja— 0] < 5. 2. There are 12 students in a party. Five of them are girls. In how many ways can these 12 students be arranged in a row if (i) there are no restrictions? (ii) the 5 girls must be together (forming a block)? (iii) no 2 girls are adjacent? (iv) between two particular boys A and B, there are no boys but exactly 3 girls? 3. m boys and n girls are to be arranged in a row, where m,n EN. Find the number of ways this can be done in each of the following cases: (i) There are no restrictions; (ii) No boys are adjacent (m < n+ 1); (iii) The n girls form a single block; (iv) A particular boy and a particular girl must be adjacent. 4, How many 5-letter words can be formed using A, B, C, D, E, F, G, H, I,J, (i) if the letters in each word must be distinct? (ii) if, in addition, A, B,C, D, E,F can only occur as the first, third or fifth letters while the rest as the second or fourth letters?Chapter 1. Permutations and Combinations 51 5. Find the number of ways of arranging the 26 letters in the English alphabet in a row such that there are exactly 5 letters between z and y. 6. Find the number of odd integers between 3000 and 8000 in which no digit is repeated. 7. Evaluate 1-1142-2143-3!4---4n-nl, where nEN. See een eae et ee ae etn G+! * @+H! (tb? where n EN. 9. Prove that for each n € N, (n+1)(n+2)---(2n) is divisible by 2”. (Spanish Olympiad, 1985) 10. Find the number of common positive divisors of 104° and 20°°. 11. In each of the following, find the number of positive divisors of n (in- clusive of n) which are multiples of 3: (i) n= 210; (ii) n = 630; (iii) n = 151200. 12. Show that for any n € N, the number of positive divisors of n? is always odd. 13. Show that the number of positive divisors of “111...1” is even. 1992 14. Let n,r € N with r 2, and let T={(z,y,z)€S?|a2 6. Find the number of triangles formed by any 3 vertices of P that are pairwise nonadjacent in P.35. 36. 37. 38. 39. 4 Ss 42. 43. Chapter 1. Permutations and Combinations 55 . 6 boys and 5 girls are to be seated around a table. Find the number of ways that this can be done in each of the following cases: (i) There are no restrictions; (ii) No 2 girls are adjacent; (iii) All girls form a single block; (iv) A particular girl G is adjacent to two particular boys B; and Bo. Show that the number of r-circular permutations of n distinct objects, where 1 720; ® (") (") et (") (277), where n> m3 730. . Prove the identity (") = (,,",) by (BP). 41. Let X = {1,2,...,.n}, A={ACX [ng A}, and B= {AC X |n€ A}. Show that |A| = |B] by (BP). Let r,n € N. Show that the product (n+ 1)(n+2)---(n+7r) of r consecutive positive integers is divisible by r!. Let A be a set of kn elements, where k,n € N. A k-grouping of A is a partition of A into k-element subsets. Find the number of different k-groupings of A.56 44, 45. 46. Exercise 1 Twenty five of King Arthur’s knights are seated at their customary round table. Three of them are chosen - all choices of three being equally likely — and are sent off to slay a troublesome dragon. Let P be the probability that at least two of the three had been sitting next to each other. If P is written as a fraction in lowest terms, what is the sum of the numerator and denominator? (AIME, 1983/7) (Readers who wish to get more information about the AIME may write to Professor Walter E. Mientka, AMC Executive Director, Department of Mathematics & Statistics, University of Nebraska, Lincohn, NE 68588-0322, USA.) One commercially available ten-button lock may be opened by depress- ing — in any order — the correct five buttons. The sample shown below has {1,2,3,6,9} as its combination. Suppose that these locks are re- designed so that sets of as many as nine buttons or as few as one button could serve as combinations. How many additional combinations would this allow? (AIME, 1988/1) akeone In a shooting match, eight clay targets are arranged in two hanging columns of three each and one column of two, as pictured. A marksman is to break all eight targets according to the following rules: (1) The marksman first chooses a column from which a target is to be broken. (2) The marksman must then break the lowest remaining unbroken target in the chosen column. If these rules are followed, in how many different orders can the eight targets be broken? (AIME, 1990/8)47. 48. oo 49. S 50. 51. Chapter 1. Permutations and Combinations 57 Using the numbers 1, 2, 3, 4, 5, we can form 5!(= 120) 5-digit numbers in which the 5 digits are all distinct. If these numbers are listed in increasing order: 12345, 12354, 12435, 54321, Ast 2nd 3rd. 120th find (i) the position of the number 35421; (ii) the 100th number in the list. The P$(= 24) 3-permutations of the set {1,2,3,4} can be arranged in the following way, called the lexicographic ordering: 123, 124, 132, 134, 142, 143, 213, 214, 231, 234, 241, 243, 312, ..., 431, 432. Thus the 3-permutations “132” and “214” appear at the 3rd and 8th positions of the ordering respectively. There are P?(='3024) 4- permutations of the set {1,2,...,9}. What are the positions of the ~ 4-permutations “4567” and “5182” in the corresponding lexicographic ordering of the 4 permutations of {1,2, ..., 9}? The (§)(= 10) 3-element subsets of the set {1,2,3,4,5} can be arranged in the following way, called the lexicographic ordering: {1, 2,3}, {1,2,4}, {1,2,5}, {1,3,4}, {1,35}, {1,4,5}, {2,3,4}, {2,3,5}, {2,4,5}, {3,45}. Thus the subset {1,3,5} appears at the 5th position of the ordering. There are (1?) 4-clement subsets of the set {1,2,...,10}. What are the positions of the subsets {3,4,5,6} and {3,5,7,9} in the corresponding lexicographic ordering of the 4-element subsets of {1, 2, ..., 10}? Six scientists are working of a secret project. They wish to lock up the documents in a cabinet so that the cabinet can be opened when and only when three or more of the scientists are present. What is the smallest number of locks needed? What is the smallest number of keys each scientist must carry? A 10-storey building is to be painted with some 4 different colours such that each storey is painted with one colour. It is not necessary that all 4 colours must be used. How many ways are there to paint the building if (i) there are no other restrictions? (ii) any 2 adjacent stories must be painted with different colours?58 Exercise 1 52. Find the number of all multi-subsets of M = {r1-a1,r2-a2,+++,Tm Gn}. 53. Let r,b € N with r nk. Find the number of ways of distribut- ing r identical objects into n distinct boxes so that each box holds at least k objects. 61. Find the number of ways of arranging the 9 letters r,s,t,u,v,w,2,y,z in a row so that y always lies between x and z (z and y, or y and z need not be adjacent in the row). 62. Three girls A,B and C, and nine boys are to be lined up in a row. In how many ways can this be done if B must lie between A and C, and A,B must be separated by exactly 4 boys? 63. Five girls and eleven boys are to be lined up in a row such that from left to right, the girls are in the order: G1, G2, G3, Ga, Gs. In how many ways can this be done if G; and G2 must be separated by at least 3 boys, and there is at most one boy between G4 and Gs? 64. Given r,n € N with r > n, let L(r,n) denote the number of ways of distributing r distinct objects into n identical boxes so that no box is empty and the objects in each box are arranged in a row. Find L(r,n) in terms of r and n. 65. Find the number of integer solutions to the equation: 2, +22 +23 +24+25 + 25 = 60 in each of the following cases: (i) 2 > i-1 for each i = 1,2,...,6; (ii) 1 > 2, 22 >5,2< 23 <7, t4 > 1, 25 > 3 and a6 > 2. 66. Find the number of integer solutions to the equation: 21 +22+23 +24 = 30 in each of the following cases: (i) 2; > 0 for each i = 1,2,3, 4; (ii) 2 < 21 <7 and 2; > 0 for each i = 2,3,4; (iii) 21 > -5, 22 > -1, 23 > 1 and xq > 2.