Koh Khee Meng

NTING
Tay Eng Guan

World Scientific
COUNTING
This page is intentionally left blank
COUNTING

Koh Khee Meng

National University of Singapore, Singapore

Tay Eng Guan

National Institute of Education
Nanyang Technological University, Singapore

V f e World Scientific
WM New Jersey • London *Sini
• Singapore • Hong Kong
Published by
World Scientific Publishing Co. Pte. Ltd.
P O Box 128, Farrer Road, Singapore 912805
USA office: Suite IB, 1060 Main Street, River Edge, NJ 07661
UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

The authors and publisher would like to thank the following organizations for their permission to
reproduce the selected problems in this book:

The Cambridge International Examinations

The MAA American Mathematics Competitions

COUNTING
Copyright © 2002 by World Scientific Publishing Co. Pte. Ltd.
All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means,
electronic ormechanical, including photocopying, recording or any information storage and retrieval
system now known or to be invented, without written permission from the Publisher.

For photocopying of material in this volume, please pay a copying fee through the Copyright
Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to
photocopy is not required from the publisher.

ISBN 981-238-063-9
ISBN 981-238-064-7 (pbk)

Printed in Singapore by Mainland Press

 Preface

Combinatorics is a branch of mathematics dealing with discretely struc-

tured problems. Its scope of study includes selections and arrangements
of objects with prescribed conditions, configurations involving a set of
nodes interconnected by edges (called graphs), and designs of experi-
mental schemes according to specified rules. Combinatorial problems
and their applications can be found not only in various branches of
mathematics, but also in other disciplines such as engineering, computer
science, operational research, management sciences and the life sciences.
Since computers require discrete formulation of problems, combinato-
rial techniques have become essential and powerful tools for engineers
and applied scientists, in particular, in the area of designing and ana-
lyzing algorithms for various problems which range from designing the
itineraries for a shipping company to sequencing the human genome in
the life sciences.
The counting problem, which seeks to find out how many arrange-
ments there are in a particular situation, is one of the basic problems
in combinatorics. Counting has been used in the social sciences for cal-
culating the Shapley-Shubik power index which measures the power of
a player in a decision-making body (such as a shareholders' meeting,
a parliament or the UN Security Council). In Chemistry, Cayley used
graphs to count the number of isomers of saturated hydrocarbons; while
Harary and Read counted the number of certain organic compounds
built up from benzene rings by representing them as configurations of
VI Counting

hexagons joined together along a common edge. In Genetics, by count-

ing all possibilities for a DNA chain made up of the four bases, scientists
arrive at an astoundingly large number and so are able to understand
the tremendous possible variation in genetic makeup. Counting has
been used as well to study the primary and secondary structures of
RNA.
This booklet is intended as an introduction to basic counting tech-
niques for upper secondary and junior college students, and teachers.
We believe that it would also be of interest to those who appreciate
mathematics and to avid puzzle-solvers.
The variety of problems and applications in this booklet is not only
useful for building up an aptitude in counting but is a rich source for
honing basic skills and techniques in general problem-solving. Many of
the problems evade routine and, as a desired result, force the reader to
think hard in his attempts to solve them. In fact, the diligent reader will
often discover more than one way of solving a particular problem, which
is indeed an important awareness in problem-solving. This booklet thus
helps to provide students an early start to learning problem-solving
heuristics and thinking skills.
The first two chapters cover two basic principles, namely, the
Addition Principle and the Multiplication Principle. Both principles
are commonly used in counting, even by those who would never count
themselves as students of mathematics! However, these principles have
equally likely been misunderstood and misused. These chapters help to
avoid this by stating clearly the conditions under which the principles
can be applied. Chapter 3 introduces the concepts of combinations and
permutations by viewing them as subsets and arrangements of a set of
objects, while Chapter 4 provides various applications of the concepts
learnt.
Many apparently complex counting problems can be solved with just
"a change of perspective". Chapter 5 presents an important principle
along this line, i.e. the Bijection Principle; while Chapter 6 introduces
a very useful perspective to which many counting problems can be con-
verted to, i.e. the distribution of balls into boxes. The next three chap-
ters flesh out the Bijection Principle and the distribution perspective
with a number of applications and variations.
 Preface vn

In Chapter 3, we introduce a family of numbers which are denoted

by (™) or C™. The last three chapters put this family of numbers in
the context of the binomial expansion and Pascal's Triangle. A num-
ber of useful identities are proven and problems are posed where these
identities surprisingly appear.
Chapter 13 closes this booklet with a collection of interesting prob-
lems in which the approaches to solving them appear as applications of
one or more concepts learnt in all the earlier sections. Problems in this
and other sections marked with (C) are reproduced by permission of the
University of Cambridge Local Examinations Syndicate and those with
(AIME) are from the American Invitational Mathematics Examination.
We would like to express our gratitude to the above organizations for
allowing us to include these problems in the book.
This booklet is based on the first six from a series of articles on
counting that first appeared in Mathematical Medley, a publication
of the Singapore Mathematical Society. We would like to thank Tan
Ban Pin who greatly helped the first author with the original series.
Many thanks also to our colleagues, Dong Fengming, Lee Tuo Yeong and
Toh Tin Lam for reading through the draft and checking through the
problems — any mistakes that remain are ours alone.
For those who find this introductory work interesting and would like
to know more about the subject, a recommended list of publications for
further reading is provided at the end of this book.

Koh Khee Meng

Tay Eng Guan
This page is intentionally left blank
 Contents

Preface v

1. The Addition Principle 1

2. The Multiplication Principle 9

3. Subsets and Arrangements 17

4. Applications 25

5. The Bijection Principle 35

6. Distribution of Balls into Boxes 47

7. More Applications of (BP) 53

8. Distribution of Distinct Objects into Distinct Boxes 63

9. Other Variations of the Distribution Problem 67

10. The Binomial Expansion 73

11. Some Useful Identities 77

12. Pascal's Triangle 85

x Counting

13. Miscellaneous Problems 95

Books Recommended for Further Reading 105

Answers to Exercises 107

Index 111
 Chapter 1

The Addition Principle

In the process of solving a counting problem, there are two very simple
but basic principles that we always apply. They are called the Addition
Principle and the Multiplication Principle. In this chapter, we shall
introduce the former and illustrate how it is applied.
Let us begin with a simple problem. Consider a 4-element set A =
{a,b,c,d}. In how many ways can we form a 2-element subset of A1
This can be answered easily by simply listing all the 2-element subsets:
{a, b}, {a, c}, {a, d}, {b, c}, {b, d}, {c, d} .
Thus, the answer is 6.
Let us try a slightly more complicated problem.
Example 1.1 A group of students consists of 4 boys and 3 girls. How
many ways are there to select 2 students of the same sex from the groupl
Solution As the problem requires us to select students of the same
sex, we naturally divide our consideration into two distinct cases: both
of the two students are boys, or, both are girls. For the former case,
this is the same as selecting 2 elements from a 4-element set. Thus, as
shown in the preceding discussion, there are 6 ways. For the latter case,
assume the 3 girls are g\, g% and #3. Then there are 3 ways to form such
a pair; namely,

{91,92}, {§1,9z}, {52,53} •

Thus, the desired number of ways is (6 + 3), which is 9. •
1
2 Counting

In dealing with counting problems that are not so straightforward,

we quite often have to divide our consideration into cases which are
disjoint (like boys or girls in Example 1.1) and exhaustive (besides boys
or girls, no other cases remain). Then the total number of ways would
be the sum of the numbers of ways from each case. More precisely, we
have:

The Addition Principle

Suppose that there are n\ ways for the event E\ to occur
and n2 ways for the event E2 to occur. If all these ways are (1.1)
distinct, then the number of ways for E\ or E2 to occur is
ni + n2.

For a finite set A, the size of A or the cardinality of A, denoted by |^4|,

is the number of elements in A. For instance, if A — {u,v,w,x,y,z},
then \A\ = 6; if A is the set of all the letters in the English alphabet,
then \A\ — 26; if 4> denotes the empty (or null) set, then \4>\ = 0.
Using the language of sets, the Addition Principle simply states the
following.

If A and B are finite sets with AD B = <p,

(1.2)
then | A U B | = |A| + |B|.

Two sets A and B are disjoint if A n B — <fi. Clearly, the above

result can be extended in a natural way to any finite number of pairwise
disjoint finite sets as given below.

(AP) If A1,A2,..., An,n > 2, are finite sets which are

pairwise disjoint, i.e. Ai C\ Aj = (f> for all i, j with 1 < i <
j <n, then

|;4i U A2 U • • • U An\ = \AX\ + \A2\ + • • • + \An\, (1.3)

or, in a more concise form:

= i£14
=l
 The Addition Principle 3

Example 1.2 From town X to town Y, one can travel by air, land or
sea. There are 3 different ways by air, A different ways by land and 2
different ways by sea as shown in Figure 1.1.

air

Figure 1.1

How many ways are there from X to Y?

Let A\ be the set of ways by air, A2 the set of ways by land and A3
the set of ways by sea from XtoY. We are given that

|Ai| = 3, |A 2 | = 4 and |A 3 | = 2.
Note that AiDA2 = AinA3 = A2(~)A3 = <j> and AiLiA2UA3 is the set of
ways from X to Y. Thus, the required number of ways is IA1U.A2UVI3I,
which, by (AP), is equal to
|i4i| + |Ai| + |i4i| = 3 + 4 + 2 = 9. •

Example 1.3 Find the number of squares contained in the Ax A array

(where each cell is a square) of Figure 1.2.

Figure 1.2
4 Counting

Solution The squares in the array can be divided into the following
4 sets:
A\: the set of 1x 1 squares,
Aim. the set of 2x 2 squares,
A3: the set of 3x 3 squares, and
A±: the set of 4x 4 squares.
There are sixteen "1 x 1 squares". Thus \A\\ = 16. There are nine "2 x 2
squares". Thus IA2I = 9. Likewise, |A 3 | = 4 and |A 4 | = 1.
Clearly, the sets A\, A2, A3, A4 are pairwise disjoint and A\ U A% U
A3 U A4 is the set of the squares contained in the array of Figure 1.2.
Thus, by (AP), the desired number of squares is given by

Uti^ = E t i \M = 16 + 9 + 4 + 1 = 30. •
Example 1.4 Find the number of routes from X toY in the one-way
system shown in Figure 1.3.

Figure 1.3

Solution Of course, one can count the number of such routes by

simply listing all of them: I 4 4 - } fl -> 7, X -+ A -> E ->•
y , . . . , x -> c -> G -4 y.
 The Addition Principle 5

Let us, however, see how to apply (AP) to introduce a more general
method.
Call a route from X to Y an X—Y route. It is obvious that just
before reaching Y along any X—Y route, one has to reach D, E, F or G.
Thus, by (AP), the number of X—Y routes is the sum of the numbers
of X—D routes, X—E routes, X—F routes and X—G routes.
How many X—D routes are there? Just before reaching D along any
X—D route, one has to reach either A or B, and thus, by (AP), the
number of X—D routes is the sum of the numbers of X—A routes and
X—B routes. The same argument applies to others (X—E routes,...)
as well.

G(2)

It is clear that the number of X—A routes (X—B routes and X—C
routes) is 1. With these initial values, one can compute the numbers of
X—D routes, X—E routes, etc., using (AP) as explained above. These
are shown in brackets at the respective vertices in Figure 1.4. Thus, we
see that the total number of possible X—Y routes is 2 + 3 + 3 + 2,
i.e. 10. •
6 Counting

Exercise

1.1 We can use 6 pieces of to cover a 6 x 3 rectangle, for

example, as shown below:

In how many different ways can the 6 x 3 rectangle be so covered?

1.2 Do the same problem as in Example 1.3 for 1 x 1, 2 x 2, 3 x 3 and
5 x 5 square arrays. Observe the pattern of your results. Find, in
general, the number of squares contained in an n x n square array,
where n > 2.
1.3 How many squares are there in
(i) the following 4 x 3 array (where each cell is a square)?

(ii) an n x 3 array (where each cell is a square), with n > 5?

1.4 How many squares are there in the following array (where each
cell is a square)?

1.5 Find the number of triangles in the following figure.

1.6 Find the number of triangles in the following figure.

 The Addition Principle 7

1.7 How many squares are there in the following configuration (where
each cell is a square with diagonals)?

1.8 Following the arrows given in the diagram, how many different
routes are there from N to S?

1.9 Following the arrows given in the diagram, how many different
routes are there from N to 5?

/iX / X ^iX
This page is intentionally left blank
 Chapter 2

The Multiplication Principle

Mr. Tan is now in town X and ready to leave for town Z by car. But
before he can reach town Z, he has to pass through town Y. There are
4 roads (labeled 1, 2, 3, 4) linking X and Y, and 3 roads (labeled as a,
b, c) linking Y and Z as shown in Figure 2.1. How many ways are there
for him to drive from X to Zl

Figure 2.1

Mr. Tan may choose road " 1 " to leave X for Y, and then select "a"
from Y to Z. For simplicity, we denote such a sequence by (I, a). Thus,
by listing all possible sequences as shown below:

(l,a),(l,6),(l,c),
(2, a), (2,6), (2, c),
(3,a),(3,6),(3,c),
(4, a), (4, b), (4, c),
9
10 Counting

we get the answer ( 4 x 3 =)12.

Very often, to accomplish a task, one may have to split it into ordered
stages and then complete the stages step by step. In the above example,
to leave X and reach Z, Mr. Tan has to split his journey into 2 stages:
first from X to Y, and then Y to Z. There are 4 roads to choose from
in Step 1: To each of these 4 choices, there are 3 choices in Step 2. Note
that the number of choices in Step 2 is independent of the number of
choices in Step 1. Thus, the number of ways from X to Z is given by
4 x 3 (= 12). This illustrates the meaning of the following principle.

The Multiplication Principle

Suppose that an event E can be split into two events E\
and Ei in ordered stages. If there are n\ ways for E\ to (2.1)
occur and n<i ways for Ei to occur, then the number of
ways for the event E to occur is n\U2.

Example 2.1 How many ways are there to select 2 students of dif-
ferent sex from a group of 4 boys and 3 girls?

E2

E: forming a pair consisting of a boy and a girl;

E\\ selecting a boy;
E-i. selecting a girl.

Figure 2.2
 The Multiplication Principle 11

Solution The situation when the 2 students chosen are of the same
sex was discussed in Example 1.1. We now consider the case where the 2
students chosen are of different sex. To choose 2 such students, we may
first choose a boy and then select a girl. There are 4 ways for step 1
and 3 ways for step 2 (see Figure 2.2). Thus, by the Multiplication
Principle, the answer is 4 x 3 (= 12). D
The Addition Principle can be expressed using set language. The
Multiplication Principle can likewise be so expressed. For the former,
we make use of the union AU B of sets A and B. For the latter, we shall
introduce the cartesian product Ax B of sets A and B. Thus given two
sets A and B, let
AxB = {(x,y) :xeA, yeB};
namely, Ax B consists of all ordered pairs (x,y), where the first coordi-
nate, "x", is any member in the first set A, and the second coordinate,
"y", is any member in the second set B. For instance, if A = {1,2,3,4}
and B = {a, b, c}, then
A x B = {(1, a), (1,6), (1, c), (2, a), (2,6), (2, c), (3, a), (3,6),
(3, c), (4, a), (4, b), (4, c)}.
Assume that A and B are finite sets. How many members (i.e. or-
dered pairs) are there in the set A x B? In forming ordered pairs in
A x B, a member, say "x" in A is paired up with every member in
B. Thus there are \B\ ordered pairs having "x" as the first coordinate.
Since there are \A\ members in A, altogether we have \A\ \B\ ordered
pairs in Ax B. That is,

\AxB\ = \A\\B\ (2.2)

Principle (2.1) and result (2.2) are two different forms of the same
fact. Indeed, an event E which is split into two events in ordered stages
can be regarded as an ordered pair (a, b), where "a" stands for the first
event and "6" the second; and vice versa.
Likewise, Principle (2.1) can be extended in a very natural way as
follows:
12 Counting

(MP) Suppose that an event E can be split into k events

Ei, E2,..., Ejc in ordered stages. If there are n\ ways for
E\ to occur, n,2 ways for E2 to occur,..., and n^ ways for (2.3)
Ek to occur, then the number of ways for the event E to
occur is given by n\n2 • • • n\..

By extending the cartesian product A x B of two sets to A\ x

A2 x • • • x vlfc of k sets, we shall also derive an identity which extends
(2.2) and expresses (2.3) using set language.
Let Ai,A2, • •., Ak be k finite sets, and let
Ai x A2 x • • • x Ak
= {(xi, X2, •. •, Xk) : xi G Ai for each i — 1, 2, ,*}•

Then

(MP) |AX x ^. 2 x • • • x Afc| = 1^4x1 |A 2 | - - - |Afc| (2.4)

Example 2.2 There are four 2-digit binary sequences: 00,01,10,11.

There are eight 3-digit binary sequences: 000,001,010,011,100,101,
110, 111. How many 6-digit binary sequences can we form?
Solution The event of forming a 6-digit binary sequence can be split
into 6 ordered stages as shown in Figure 2.3.
0 0 0 0 0 0

_i S<^_j__^yl ^ 1 ^ >L^__l_^> <^_J___^J< \_

1st 2nd 3rd 4th 5th 6th

Figure 2.3

Thus, by (2.3), the desired number of sequences is 2 x 2 x 2 x 2 x 2 x 2 = 2 6 .

Using set language, the same problem can be treated as follows. We
have
A1=A2 = --- = A6 = {0,1},
 The Multiplication Principle 13

The members in A\ x Ai x • • • x AQ can be identified with 6-digit

binary sequences in the following way:
(1,1,0,1,0,1) o 110101,
(0,1,1,0,1,1) ++011011,
etc.

Thus, the number of 6-digit binary sequences is given by

\Ai x A2 x • • • x Ae\, which, by (2.4), is equal to
\Ai\ \A2\ • • • \Ae\ = 2 x 2 x 2 x 2 x 2 x 2 = 2 6 . •
Prom now on, (MP) shall refer to Principle (2.3) or the identity (2.4).
Example 2.3 Figure 2.4 shows 9 fixed points a,b,c,... , i which are
located on the sides ofAABC. If we select one such point from each side
and join the selected points to form a triangle, how many such triangles
can be formed?

Solution To form such a triangle, we first select a point on AB,

then a point on BC and finally a point on CA. There are 3 ways
in step 1 (one of a,b,c), 4 ways in step 2 (one of d,e,f,g) and 2 ways
in step 3 (either h or i). Thus by (MP), there are 3 x 4 x 2 (= 24) such
triangles. •
We have seen in both the preceding and the current chapters some
problems that can be solved by applying (AP) or (MP) individually.
Indeed, more often than not, problems that we encounter are more
14 Counting

complex and these require that we apply the principles together. The
following is an example.
Example 2.4 (Continuation of Example 2.3) Find the number of
triangles that can be formed using the 9 fixed points of Figure 2.4 as
vertices.
Solution This problem is clearly more complex than that of Exam-
ple 2.3 as there are other triangles whose three vertices are not neces-
sarily chosen from three different sides; but then, where else can they
be chosen from? The answer is: two from one side and one from the
remaining two sides. In view of this, we shall now classify the required
triangles into the following two types.
Type 1 — Triangles whose three vertices are chosen from three different
sides.
As shown in Example 2.3, there are 3 x 4 x 2 (= 24) such
triangles.
Type 2 — Triangles having two vertices from one side and one from the
other two sides.
We further split our consideration into three subcases.
(i) Two vertices from AB and one from BC or CA.
There are 3 ways to choose two from AB (namely,
{a,b},{a,c} or {b, c}) and 6 ways to choose one from
the other sides (namely, d,e,f,g,h,i). Thus, by (MP),
there are 3 x 6 (= 18) such triangles,
(ii) Two vertices from BC and one from CA or AB.
There are 6 ways to choose two from BC (why?) and
5 ways to choose one from the other sides (why?). Thus,
by (MP), there are 6 x 5 (= 30) such triangles,
(iii) Two vertices from CA and one from AB or BC.
There is only one way to choose two from CA and there
are 7 ways to choose one from the other sides. Thus, by
(MP), there are 1 x 7 (= 7) such triangles.
Summing up the above discussion, we conclude that by (AP), there are
18 + 30 + 7 (= 55) triangles of Type 2.
 The Multiplication Principle 15

As there are 24 triangles of Type 1 and 55 triangles of Type 2, the

required number of triangles is thus, by (AP), 24 + 55 (= 79). •

Exercise

2.1 Following the arrows given in the diagram, how many different
routes are there from W to E?

W
y
2.2 In the following figure, ABCD and FEC are two perpendicular
lines.

B D

(i) Find the number of right-angled triangles AXCY that can be

formed with X, Y taken from A, B, D, E, F.
(ii) Find the number of triangles that can be formed with any
three points A,B,C,D,E,F as vertices.
2.3 There are 2 distinct terms in the expansion of a(p + q):
a
(P + q) = ap + aq.

There are 4 distinct terms in the expansion of (a + b)(p + q):

(a + b)(p + q) = ap + aq + bp + bq .
 Counting

How many distinct terms are there in each of the expansions of

(i) (a + b + c + d)(p + q + r + s + t),
(ii) (xi + x2 H + xm)(2/i + 7/2 H 1- J/n), and
(iii) (xi + x2 + --- + xm)(yi + y2-\ + yn){zi + z2-\ h zt)?
In how many different ways can the following configuration be
covered by nine 2 x 1 rectangles?

A ternary sequence is a sequence formed by 0, 1 and 2. Let n be

a positive integer. Find the number of n-digit ternary sequences
(i) with no restrictions;
(ii) which contain no "0";
(iii) which contain at most one "0";
(iv) which contain at most one "0" and at most one " 1 " .
6 The
following diagram shows 12 distinct points: 01,02,03,61,
c i , . . . , C5 chosen from the sides of AABC.

(i) How many line segments are there joining any two points,
each point being from a different side of the triangle?
(ii) How many triangles can be formed from these points?
(iii) How many quadrilaterals can be formed from these points?
 Chapter 3

Subsets and Arrangements

There are 25 students in the class. How many ways are there to choose
5 of them to form a committee? If among the chosen five, one is to be
the chairperson, one the secretary and one the treasurer, in how many
ways can this be arranged? In this chapter, our attention will be focused
on the counting problems of the above types. We shall see how (MP)
is used to solve such problems, and how (MP), by incorporating (AP),
enables us to solve more complicated problems.
From now on, for each natural number n, we shall denote by N n the
set of natural numbers from 1 to n inclusive, i.e.,

Nn = { l , 2 , 3 , . . . , n } .

Consider the 4-element set N4 = {1,2,3,4}. How many subsets of N4

are there? This question can be answered readily by listing all the
subsets of N4. Table 3.1 lists all the subsets according to the number

Table 3.1

Number of Subsets of N4 Number of

elements subsets of N4

0 <t> 1
1 {1}, {2}, {3}, {4} 4
2 {1, 2}, {1, 3}, {1, 4}, {2, 3}, {2, 4}, {3, 4} 6
3 {1, 2, 3}, {1, 2, 4}, {1, 3, 4}, {2, 3, 4} 4
4 {1, 2, 3, 4} 1

17
18 Counting

of elements they possess: It is now easy to count the total number of

subsets of N4(= 16).
We note that the 5 numbers, namely, 1, 4, 6, 4, 1 (whose sum is
16) shown in the right hand column of Table 3.1 are the corresponding
numbers of r-element subsets of N4, where r = 0,1,2,3,4. These num-
bers are very interesting, useful and important in mathematics, and
mathematicians have introduced special symbols to denote them.
In general, given two integers n and r with 0 < r < n, we denote
by ("), the number of r-element subsets of N„. Thus, Table 3.1 tells
us that

(!)-• C)-<- (J)-- (!)-• O -

The symbol (") is read "n choose r". Some other symbols for this
quantity include C™ and nCr.
Now, what is the value of (2)? Since (2) counts, by definition, the
number of 2-element subsets of N5, we may list all these subsets as
shown below:

{1,2},{1,3},{1,4},{1,5},{2,3},{2,4},{2,5},{3,4},{3,5},{4,5} ,

and see that there are 10 in total. Thus, we have (2) = 10.
You may ask: How about ( 6 )? We are sure that we are too busy
to have time to compute ( 6 ) by listing all the 6-element subsets of
Nioo- Thus, a natural question arises: Is there a more efficient way to
compute (1060)? The answer is "Yes", and we are going to show you.
Let us first consider a different but related problem. How many
ways are there to arrange any three elements of N4 = {1,2,3,4} in a
row? With a little patience, we can list all the required arrangements
as shown in Table 3.2.

Table 3.2
123 132 213 231 312 321
124 142 214 241 412 421
134 143 314 341 413 431
234 243 324 342 423 432
 Subsets and Arrangements 19

Thus, there are 24 ways to do so. The answer is "correct" but the
method is "naive". Is there a cleverer way to get the answer?
Imagine that we wish to choose 3 numbers from N4 and put them
one by one into 3 spaces as shown.

1st 2nd 3rd

This event can be thought of as a sequence of events: We first select

a number from N4 and place it in the 1st space; we then select another
number and place it in the 2nd space; finally, we select another number
and place it in the 3rd space. There are 4 choices for the first step, 3
choices (why?) for the second and 2 choices (why?) for the last. Thus,
by (MP), there are 4 - 3 - 2 (= 24) ways to do so. The answer agrees
with what we obtained above. Isn't this method better?
This method is better not only in shortening our solution, but also
in giving us an idea on how to generalize the above result.
In the above example, we considered the number of ways of arranging
3 elements of N4 in a row. We now ask a general question: Given integers
r, n with 0 < r < n, how many ways are there to arrange any r elements
of N n in a row?
Consider the r spaces shown in Figure 3.1.

(T 1, 2, 3, ...,n 1

n n-\ n—2 n-(r-2) n-0-l)

choices choices choices choices choices
...
1st 2nd 3rd (r-l)th rth

Figure 3.1
20 Counting

We wish to choose r elements from {1, 2 , . . . ,n} to fill the r spaces,

where the ordering of elements matters. There are n choices for the
1st space. After fixing one in the 1st space, there are n — 1 choices
remaining for the 2nd space. After fixing one in the 2nd space, there
are n — 2 choices left for the 3rd space, and so on. After fixing one in
the (r — l)th space, there are n — (r — 1) choices left for the rth space.
Thus, by (MP), the number of ways to arrange any r elements from N n
in a row is given by

n(n - l)(n - 2 ) . . . (n - r + 1).

For convenience, let us call an arrangement of any r elements from N n
in a row, an r-permutation of N n , and denote by P™ the number of
r-permutations of N n . Thus, we have

P r " = n(n - l)(n - 2 ) . . . (n - r + 1) (3.1)

For instance, all the arrangements in Table 3.2 are 3-permutations of

N4, and, by (3.1), the number of 3-permutations of N4 is given by
P34 = 4 • 3 • 2 = 2 4 ,

which agrees with what we have counted in Table 3.2.

The expression (3.1) for P™ looks a bit long. We shall make it more
concise by introducing the following useful notation. Given a positive
integer n, define n! to be the product of the n consecutive integers
n, n — 1 , . . . , 3,2,1; that is,

n! = n ( n - l ) ( n - 2 ) . . . 3 - 2 - l (3.2)

Thus 4! = 4 • 3 • 2 • 1 = 24. The symbol "n!" is read "n factorial". By

convention, we define 0! = 1.
Using the "factorial" notation, we now have
P r n = n(n - 1 ) . . . (n - r + 1)
n{n — 1 ) . . . (n — r + l)(n — r)(n — r — 1 ) . . . 2 • 1 _ n\
(n — r)(n — r — 1 ) . . . 2 • 1 (n — r)\ '
 Subsets and Arrangements 21

That is,

(3.3)

When n = 4 and r = 3, we obtain

4! 4! 4• 3• 2• 1
P4 - = 4 • 3 • 2 = 24,
r
3 — (4 - 3)! 1! 1
which agrees with what we found before.
The expression (3.3) is valid when 0 < r < n. Consider two extreme
cases: when r = 0 and r = n respectively. When r = 0, by (3.3),
n!
pn __ = ^ = 1
0
(n - 0)! n\
(How can this be explained?) When r = n, an n-permutation of N n is
simply called a permutation of N„. Thus, by (3.3) and that 0! = 1, the
number of permutations of N n is given by

n!
" {n-n)\ 0!
i.e.,

P n n = «!- (3.4)

Thus, for example, P | counts the number of permutations of N5, and

we have, by (3.4), P* = 5 • 4 • 3 • 2 • 1 = 120.

Let us now return to the problem of evaluating the quantity (").

We know from (3.3) that the number P™ of r-permutations of N n is
given by (n"i'r\, • We shall now count this number (namely the number
of r-permutations of N n ) in a different way.
To get an r-permutation of N„, we may proceed in the following
manner: first select an r-element subset of Nn, and then arrange the
chosen r elements in a row. The number of ways for the first step is, by
22 Counting

definition, ("), while that for the second step is, by (3.4), r\. Thus, by
(MP), we have

r?=[nri-«
As
p n = rc!
r
(n - r)\ '
we have
n \ n!
•r! = (n — r)\ '
and thus

(3.5)

For instance,

5\ = _iL_ = 4L = 10, while ( ?6 ) = 6S! 9 4 ! = H92052400.

1) 2!(5-2)! 2!3! V /
Note that when r = 0 or n, we have

; - i - ; , . i .

Again, by convention, we define

= 1.

By applying (3.5), one can show that the following identity holds (see
Problem 3.1):

(3.6)

Thus, (™) = («>) = 45 and Q = ^f) = 1192052400.

 Subsets and Arrangements 23

We define P™ as the number of r-permutations and (") as the number

of r-element subsets of N n . Actually, in these definitions, N n can be
replaced by any n-element set since it is the number of the elements but
not the nature of the elements in the set that matters. That is, given
any n-element set S, P™ (respectively, (")) is defined as the number of
r-permutations (respectively, r-element subsets) of S. Any r-element
subset of S is also called an r-combination of S.
We have introduced the notions of r-permutations (or permutations)
and r-combinations (or combinations) of a set S. Always remember that
these two notions are closely related but different. While a "combina-
tion" of S is just a subset of S (and thus the ordering of elements is
immaterial), a "permutation" of S is an arrangement of certain elements
of S (and thus the ordering of elements is important).

Exercise

3.1 Show that for nonnegative integers r and n, with r < n,

(i) (?) = L-J;
(ii) rft) = n(?:}), where r > l ;
(hi) ( n - r ) e ) = n ( V 1 ) ;
H r£) = (n-r + l ) ^ ) , where r > 1.
3.2 Show that for 1 < r < n,
(i) P?+1 = P? + rP?_1;
(ii) P r n + 1 = r! + r{P^ + P^1 + ••• + P;_ x );
(iii) (n - r)P? = riP?'1;
(iv) P r n = ( n - r + l)P r n _ i ;
(v) P r " = nP^i-
3.3 Prove that the product of any n consecutive integers is divisible
by n!.
3.4 Find the sum
l - l ! + 2-2! + 3-3! + --- + n - n ! .
This page is intentionally left blank
 Chapter 4

Applications

Having introduced the concepts of r-permutations and r-combinations

of an n-element set, and having derived the formulae for P? and (™),
we shall now give some examples to illustrate how these can be applied.

Example 4.1 There are 6 boys and 5 men waiting for their turn in a
barber shop. Two particular boys are A and B, and one particular man
is Z. There is a row of 11 seats for the customers. Find the number of
ways of arranging them in each of the following cases:

(i) there are no restrictions;

(ii) A and B are adjacent;
(iii) Z is at the centre, A at his left and B at his right (need not be
adjacent);
(iv) boys and men alternate.

Solution (i) This is the number of permutations of the 11 persons.

The answer is 11!.
(ii) Treat {A, B} as a single entity. The number of ways to arrange the
remaining 9 persons together with this entity is (9 + 1)!. But A and B
can permute themselves in 2 ways. Thus the total desired number of
ways is, by (MP), 2-10!.

z
25
26 Counting

As shown in the diagram above, A has 5 choices and B has 5 choices as

well. The remaining 8 persons can be placed in 8! ways. By (MP), the
total desired number of ways is 5 • 5 • 8!.
(iv) The boys (indicated by b) and the men (indicated by m) must be
arranged as shown below.

b m b m b m b m b m b

The boys can be placed in 6! ways and the men can be placed in 5!
ways. By (MP), the desired number of ways is 6!5!. D
Example 4.2 In each of the following cases, find the number of inte-
gers between 3000 and 6000 in which no digit is repeated:
(i) there are no additional restrictions;
(ii) the integers are even.

Solution Let abed be a required integer.

(i) As shown in the diagram below, a has 3 choices (i.e. 3, 4, or 5), say
a = 3.

{3,4,5}

Since no digit is repeated, a way of forming "feed" corresponds to

a 3-permutation from the 9-element set {0,1,2,4,5,... , 9}. Thus the
required number of integers is 3 • P$.
(ii) Again, a = 3, 4 or 5. We divide the problem into two cases.
Case (1) a = 4 (even)

A b e d

In this case, d has 4 choices (i.e. 0, 2, 6, 8), say d = 2. Then

a way of forming "fee" is a 2-permutation from the 8-element set
{0,1,3,5,6,7,8,9}. Thus the required number of integers is 4 • P%.
Case (2) a = 3 or 5 (odd)
 Applications 27

In this case, d has 5 choices, and the number of ways to form "fee" is
Pf. The required number of integers is 2 • 5 • Pf.
By (AP), the desired number of integers is
4 • P28 + 2 • 5 • Pf = 14 • P 2 8 . D
Example 4.3 There are 10 pupils in a class.
(i) ffou; many ways are there to form a 5-member committee for the
class?
(ii) How many ways are there to form a 5-member committee in which
one is the Chairperson, one is the Vice-Chairperson, one is the
Secretary and one is the Treasurer?
(iii) How many ways are there to form a 5-member committee in
which one is the Chairperson, one is the Secretary and one is the
Treasurer?
Solution (i) This is the same as finding the number of 5-combinations
of a 10-element set. Thus the answer is ( 5 ) = 252.
(ii) This is the same as choosing 5 pupils from the class and then placing
them in the following spaces.

Chairperson V-Chairperson | Secretary Treasurer Member

Clearly, this is a "permutation" problem, and the answer is Pg° = ^ =

7620480.
(iii) This problem can be counted in the following procedure: we first
select one for Chairperson, then one for Secretary, then one for Trea-
surer, and finally two from the remainder for committee members as
shown below:

Chairperson Secretary Treasurer 2 Members

(choose 2 from the
(10 choices) (9 choices) (8 choices) remaining 7)

Figure 4.1
28 Counting

Thus, by (MP), the answer is given by 10 • 9 • 8 • Q = 1814400. •

Note There are different ways to solve (iii). You may want to try
your own ways.
Example 4.4 As shown in Example 2.2, the number of 6-digit
binary sequences is 2 6 . How many of them contain exactly two 0's
(e.g. 001111,101101,...)?
Solution Forming a 6-digit binary sequence with two 0's is the same
as choosing two spaces from the following 6 spaces into which the two
0's are put (the rest are then occupied by l's) as shown below:
(1) (2) (3) (4) (5) (6)

e.g. 0 0

1 0 1 0 1 1

Thus, the number of such binary sequences is (2) • d

Example 4.5 Figure 4.2 shows 9 distinct points on the circumference
of a circle.
(i) How many chords of the circle formed by these points are there?
(ii) / / no three chords are concurrent in the interior of the circle, how
many points of intersection of these chords within the circle are
there?

Figure 4.2
 Applications 29

Solution (i) Every chord joins two of the nine points, and every two
of the nine points determine a unique chord. Thus, the required number
of chords is (2).

(ii) Every point of intersection of two chords corresponds to four of

the nine points, and every four of the nine points determine a point
of intersection. Thus the required number of points of intersection
^ ffl- •
Example 4.6 At a National Wages Council conference, there are
19 participants from the government, the unions and the employers.
Among them, 9 are from the unions.
In how many ways can a 7'-member committee be formed from these
participants in each of the following cases:

(i) there are no restrictions?

(ii) there is no unionist in the committee?
(iii) the committee consists of unionists only?
(iv) there is exactly one unionist in the committee?
(v) there is at least one unionist in the committee?

Solution (i) This is the number of 7-element subsets of a 19-element

set. By definition, the desired number is ( 7 ) .
(ii) This is the number of ways to form a 7-member committee from the
10 non-unionists. Thus, the desired number is (17°).
(iii) Obviously, the desired number is ( 7 ).
(iv) We first select a member from the 9 unionists and then select the
remaining 6 from the 10 non-unionists. By (MP), the desired number
tafflff)=8ffl-
30 Counting

(v) There are 7 cases to consider, namely, having r unionists, where

r = 1,2,3,4,5,6,7. Thus, by (AP) and (MP), the desired number is
given by
0
TeKX^'+QG )^)-
Indeed, we can have a shorter way to solve this part by using the idea
of "complementation".
By (i), there are (^) 7-member committees we can form from
19 participants. Among them, there are (17°) such committees which
contain no unionist by (ii). Thus, the number of 7-member committees
which contain at least one unionist should be (179) — ( x 7 ). (The reader
may check that these two answers agree.) •
The second solution given in (v) for the above example is just an
instance of applying the following principle.

Principle of Complementation (CP)

Let A be a subset of a finite set B. (4.1)
Then \B\A\ = \B\-\A\, where B\A = {x : x € B but x £ A}

If you revisit Example 2.4, you may then observe that the problem
can also be solved by (CP). There are (3) ways to form a 3-vertex subset
from the given 9 vertices. Among them, the 3 on AB and any 3 on BC
do not form a triangle. Thus, the number of triangles one can form is,
by (CP),

which is 79.
We have seen from the above examples how, by applying (CP), we
are able to considerably shorten the work needed to solve a counting
problem. When a direct approach involves a large number of cases for
which a certain condition holds, the complementary view of the smaller
number of cases in which the condition does NOT hold allows a quicker
solution to the problem. What follows then is that we count the number
of ways afforded by the smaller number of "complementary" cases and
 Applications 31

finally obtain the required answer by subtracting this from the total
number of ways.

Exercise

4.1 (Continuation from Example 4.1)

(v) A and B are at the two ends;
(vi) Z is at the centre and adjacent to A and B;
(vii) A, B and Z form a single block (i.e. there is no other person
between any two of them);
(viii) all men form a single block;
(ix) all men form a single block and all boys form a single block;
(x) no two of A, B and Z are adjacent;
(xi) all boys form a single block and Z is adjacent to A;
(xii) Z is between A and B (need not be adjacent).
4.2 (Continuation from Example 4.2)
(iii) the integers are odd;
(iv) the integers are divisible by 5;
(v) the integers are greater than 3456.
4.3 Four people can be paired off in three ways as shown below:
(1) {{A,B},{C,D}},
(2) {{A,C},{B,D}},
(3) {{A,D},{B,C}}.
In how many ways can 10 people be paired off?
4.4 If n points on the circumference of a circle are joined by straight
lines in all possible ways and no three of these lines meet at a
single point inside the circle, find
(i) the number of triangles formed with all vertices lying inside
the circle;
(ii) the number of triangles formed with exactly two vertices
inside the circle;
(iii) the number of triangles formed with exactly one vertex in-
side the circle;
(iv) the total number of triangles formed.
32 Counting

4.5 Three girls and seven boys are to be lined up in a row. Find the
number of ways this can be done if
(i) there is no restriction;
(ii) the girls must form a single block;
(iii) no two girls are adjacent;
(iv) each boy is adjacent to at most one girl.
4.6 Eight students are in a sailing club. In how many ways can they
form a team consisting of 4 Laser pairs, where the order of the
pairs does not matter? (Note: A Laser is a sailing boat that
takes a crew of two.)
4.7 There are three boys and two girls.
(i) Find the number of ways to arrange them in a row.
(ii) Find the number of ways to arrange them in a row so that
the two girls are next to each other,
(iii) Find the number of ways to arrange them in a row so that
there is at least one boy between the two girls.
4.8 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) the committee must include exactly 2 teachers?
(ii) the committee must include at least 3 teachers?
(iii) a particular teacher and a particular student cannot be both
in the committee?
4.9 A palindrome is a number that remains the same when it is read
backward, for example, 2002 is a palindrome. Find the number
of n-digit palindromes.
4.10 A team of 6 people is to be chosen from a list of 10 candidates.
Find in how many ways this can be done
(i) if the order of the people in the team does not matter;
(ii) if the team consists of 6 people in a definite order;
(iii) if the team consists of a first pair, a second pair and a third
pair but order within each pair does not matter. /rr.
\SJ)
4.11 Find how many three figure numbers, lying between 100 and 999
inclusive, have two and only two consecutive figures identical.
(C)
 Applications 33

4.12 Find t h e number of ways in which 10 persons can be divided into

(i) two groups consisting of 7 and 3 persons;
(ii) three groups consisting of 4, 3 and 2 persons with 1 person
rejected. ,„>.
(M
4.13 (i) Find the number of integers from 100 to 500 that do not
contain the digit "0".
(ii) Find the number of integers from 100 to 500 that contain
exactly one "0" as a digit.
4.14 Calculate the number of ways of selecting 2 points from 6 distinct
points. Six distinct points are marked on each of two parallel
lines. Calculate the number of
(i) distinct quadrilaterals which may be formed using 4 of the
12 points as vertices;
(ii) distinct triangles which may be formed using 3 of the
12 points as vertices. ,„,.

4.15 (a) A tennis team of 4 men and 4 women is to be picked from

6 men and 7 women. Find the number of ways in which this
can be done,
(b) It was decided that 2 of the 7 women must either be se-
lected together or not selected at all. Find how many pos-
sible teams could be selected in these circumstances. The
selected team is arranged into 4 pairs, each consisting of a
man and a woman. Find the number of ways in which this
can be done. ,n-.
This page is intentionally left blank
 Chapter 5

The Bijection Principle

We have introduced three basic principles for counting, namely, the

(AP), the (MP) and the (CP). In this chapter, we shall introduce an-
other basic principle for counting which we call the Bijection Principle,
and discuss some of its applications.

A A A A .Qt
J?k y*Tk jyy J5L25L Figure 5.1

Suppose that there are 200 parking lots in a multi-storey carpark.

The carpark is full with each vehicle occupying a lot and each lot being
occupied by a vehicle (see Figure 5.1). Then we know that the number
of vehicles in the carpark is 200 without having to count the vehicles
one by one. The number of vehicles and the number of lots are the same
because there is a one to one correspondence between the set of vehicles
and the set of lots in the carpark. This is a simple illustration of the
Bijection Principle that we will soon state.
Let us first recall some concepts on mappings of sets. Suppose A and
B are two given sets. A mapping f from A to B, denoted by

f:A^B,

35
36 Counting

is a rule which assigns to each element a in A a unique element, denoted

by f(a), in B. Four examples of mappings are shown pictorially in
Figure 5.2.
Certain kinds of mappings are important. Let / : A —> B be a
mapping. We say that / is injective (or one to one) if f(x) 7^ f(y) in B
whenever x / y in A. Thus, in Figure 5.2, /2 and / 4 are injective, while

(i) Ai Bi

(ii) A2 B2

(iii) A3 B3

(iv) A4 B4

Figure 5.2
 The Bijection Principle 37

fx and /3 are not (why?). We say that / is surjective (or onto) if for any
b in B, there exists an a in A such that / ( a ) = b. Thus, in Figure 5.2,
/3 and /4 are surjective whereas / i and /2 are not (why?). We call / a
bijection from A to £? if / is both injective and surjective. (Sometimes,
a bijection from A to 5 is referred to as a one-to-one correspondence
between A and J5.) Thus, in Figure 5.2, f\ is the only bijection. These
observations on the four mappings are summarized in Table 5.1.

Table 5.1

Injection Surjection Bijection

h X X X
h / X X
h X / X
U / / /

Let A and B be two finite sets. Suppose there is a mapping / : A —>

B that is injective. Then, by definition, each element a in A corresponds
to its image f(a) in B, and distinct elements in A correspond to distinct
images in B. Thus, we have:

The Injection Principle (IP)

Let A and B be finite sets. If there exists a one-to-one
mapping f : A-t B, then (5.1)

\A\ < \B\.

Suppose further, that the one-to-one mapping / : A —t B is onto.

Then each element b in B has a unique preimage a in A such that
/ ( a ) = b. In this case, we clearly have:

The Bijection Principle ( B P )

Let A and B be finite sets. If there exists a bijection
/ : A -> B, then (5.2)

\A\ = \B\.
38 Counting

In this chapter, we shall focus on (BP). Through the discussions on

a number of problems, we shall show you how powerful this principle is.
First of all, let us revisit a problem we studied in Chapter 4. In Ex-
ample 4.5, we counted the number of chords and the number of points
of intersection of the chords joining some fixed points on the circumfer-
ence of a circle. Let us consider a similar problem. Figure 5.3 shows
five distinct points on the circumference of a circle.
How many chords can be formed by these points?
Let A be the set of such chords, and B, the set of 2-element subsets
of {1,2,3,4,5}. Given a chord a in A, define f(a) = {p, q}, where p, q
are the two points (on the circumference) which determine the chord a
(see Figure 5.4). Then / is a mapping from A to B. Clearly, if a and
j3 are two distinct chords in A, then f(a) ^ /(/?)• Thus, / is injective.
On the other hand, for any 2-element subset {p, q} in B (say, p = 2 and

Figure 5.3

•U,5}
5

> {2,4}

Figure 5.4
 The Bijection Principle 39

q = 5), there is a chord a in A (in this instance, a is the chord joining

points 2 and 5) such that f(a) = {p,q}. Thus, / is onto.
Hence, / : A ->• B is a bijection and, by (BP), we have \A\ = \B\.
As 5 is the set of all 2-element subsets of {1,2,3,4,5}, \B\ = g ) . We
thus conclude that \A\ = | 5 | = (jj).
Next we ask: How many points of intersection (of these (j) chords)
that lie in the interior of the circle are there if no three chords are
concurrent in the interior of the circle?
Let A be the set of such points of intersection and B, the set
of 4-element subsets of {1,2,3,4,5}. Figure 5.5 exhibits a bijection
between A and B (figure out the rule which defines the bijection!).
Thus, by (BP), |A| = \B\. Since \B\ = (%) by definition, we have

a- ->{1,2,3,5)
b- ->U,2,4,5}
c- ->{1,3,4,5}
d- ->{2,3,4,5}
e- ->{1,2,3,4}

Figure 5.5

Let us proceed to show some more applications of (BP).

Example 5.1 Figure 5.6 shows a 2 x 4 rectangular grid with two
specified corners P and Q. There are 12 horizontal segments and 10
vertical segments in the grid. A shortest P—Q route is a continuous
path from P to Q consisting of 4 horizontal segments and 2 vertical

— • —
^ T

.r
1

PO

Figure 5.6
40 Counting

segments. An example is shown in Figure 5.6. How many shortest

P—Q routes in the grid are there?
Solution Certainly, we can solve the problem directly by listing all
the possible shortest routes. This, however, would not be practical if
we wish to solve the same problem in, say, a 190 x 100 rectangular grid.
We look for a more efficient way.
There are two types of segments: horizontal and vertical. Let us
use a "0" to represent a horizontal segment, and a " 1 " to represent a
vertical segment. Thus, the shortest P—Q route shown in Figure 5.6
can accordingly be represented by the binary sequence with four "0"s
and two "l"s as shown below:

-> 010100
P •-

Likewise, we can have:

_J -> 000101

and so on.
Now, let A be the set of all shortest P—Q routes and B, the set of
all 6-digit binary sequences with two l's. Then we see that the above
way of representing a shortest P—Q route in A by a binary sequence
in B defines a mapping / : A -> B. Clearly, different shortest P—Q
routes in A correspond to different sequences in B under / . Thus, / is
one-to-one. Further, for any sequence b in B, say, b = 100010, one can
find a shortest P—Q route, a in A, in this case,

-%Q

r
 The Bijection Principle 41

so that f(a) = b. Thus / is onto, and so it is a bijection. Now, by (BP),

we conclude that \A\ = \B\. But how does this simplify our effort to
find the number of shortest P—Q routes?
Let us explain. What is the set Bl B is the set of all 6-digit binary
sequences with two l's. Can we count \B\1 Oh, yes! We have already
solved it in Example 4.4. The answer is | S | = ( 2 ). Accordingly, we have
\A\ = \B\ = ®. D

Example 5.2 The power set of a set S, denoted by V(S), is the set
of all subsets of S, inclusive of S and the empty set (j). Thus, for N n =
{1,2,... ,n}, 1 < n < 3, we have

V(N1) = {<P,{1}},

P(N 2 ) = {«/>,{!}, {2}, {1,2}},

V(N3) = {<j>, {1}, {2}, {3}, {1, 2}, {1,3}, {2,3}, {1, 2,3}} .

Note that \V{Ni)\ = 2, |7 , (N 2 )| = 4, |7>(N3)| = 8. Table 3.1 shows that

|P(N 4 )| = 16. What is the value of |P(N 5 )|?

Solution For convenience, let A = 73(N5); that is, A is the power set
of {1,2,3,4,5}. Represent these subsets by 5-digit binary sequences as
follows:

4> • ooooo
{1} • 10000
{2} • 01000

{5} >• 00001

{1,2} • 11000

{4,5} •00011

{1,3,5} • 10101

{1,2,3,4,5} > 11111

42 Counting

The rule is that the ith digit of the corresponding binary sequence is
"1" if "i" is in the subset; and "0" otherwise. Let B be the set of all
5-digit binary sequences. Clearly, the above rule establishes a bijection
between A and B. Thus, by (BP), |A| = |B|. Since | B | = 2 5 (see
Example 2.2), \A\ = 2 5 . D
Note that |P(Ni)| = 2 = 2!, |P(N 2 )| = 4 = 2 2 , |P(N 3 )| = 8 = 2 3 ,
|7>(N4)| = 16 = 2 4 , and now |P(N 5 )| = 2 5 . What is |P(N„)| for n > 1?
(See Exercise 5.3.)
Finally, let us introduce a counting problem related to the notion of
divisors of natural numbers. We shall denote by N, the set of natural
numbers; i.e.
N = {1,2,3,...}.
Assume that d,n € N. We say that d is a divisor of n if when n is
divided by d, the remainder is zero. Thus, 3 is a divisor of 12, 5 is a
divisor of 100, but 2 is not a divisor of 9.
Let n € N, n > 2. Clearly, n has at least two divisors, namely 1 and
n. How many divisors (inclusive of 1 and n) does n have? This is a
type of problem that can often be found in mathematical competitions.
We shall tackle this problem and see how (MP) and (BP) are used in
solving the problem.
To understand the solution, we first recall a special type of numbers
called prime numbers and state an important result relating natural
numbers and prime numbers.
A natural number p is said to be prime (or called a prime) if p > 2
and the only divisors of p are 1 and p. All prime numbers less than 100
are shown below:

2,3,5,7,11,13,17,19,23,29,31,37,41,
43,47,53,59,61,67,71,73,79,83,89,97.
The primes are often referred to as building blocks of numbers
because every natural number can always be expressed uniquely as a
product of some primes. For example,
108 = 2 2 x 3 3 , 1620 = 2 2 x 3 4 x 5 ,
1815 = 3 x 5 x l l 2 , 215306 = 2 x 7 2 x 13 3 .
 The Bijection Principle 43

This fact is so basic and important to the study of numbers that it is

called the Fundamental Theorem of Arithmetic (FTA).

(FTA) Every natural number n > 2 can be factorized as

n = p?W •Pkk

for some distinct primes pi,P2, • • • ,Pk a n d for some natural numbers
mi, m,2,..., mk. Such a factorization is unique if the order of primes
is disregarded.

FTA was first studied by the Greek mathematician, Euclid (c. 450-
380 BC) in the case where the number of primes is at most 4. It was
the German mathematician, Carl Friedrich Gauss (1777-1855), known
as the Prince of Mathematicians, who stated and proved the full result
in 1801.
Let us now return to the problem of counting the number of divisors
of n. How many divisors does the number 72 have? Since 72 is not a big
number, we can get the answer simply by listing all the divisors of 72:
1,2,3,4,6,8,9,12,18,24,36,72.
The way of counting the divisors of n by listing as shown above is
certainly impractical when n gets larger. We look for a more efficient
way.
Let us look at the example when n = 72 again and try to get some
information about 72 and its divisors by FTA.

!
2&s%

The images above are those of Euclid on a stamp of the Maldives and Gauss on a
German banknote.
44 Counting

Observe that 72 = 2 3 x 3 2 . Suppose x is a divisor of 72. Clearly, x

does not contain prime factors other than 2 and 3. That is, x must be
of the form

x = 2px3«

where, clearly, p E {0,1,2,3} and q € {0,1, 2}. On the other hand, any
such number 2P x 3 9 is a divisor of 72. Indeed,
1 = 2° x 3° 9 = 2° x 32 ,
2 = 21 x 3° 12 = 2 2 x 31,
3 - 2° x 31 18 = 21 x 32 ,
4 = 2 2 x 3° 24 = 2 3 x 3 1 ,
6 = 21 x 31 36 = 2 2 x 32 ,
8 = 2 3 x 3° 72 = 23 x 32 .

Let A be the set of divisors of 72 and B = {{p,q) : 0 < p < 3,0 <
q < 2} = {0,1,2,3} x {0,1,2}. Then the above list implies that the
mapping / defined by

/(1) == (0,0), /(9) == (0,2)

/(2) == (1,0), /(12)== (2,1)
/(3) == (o,i), /(18) == (1,2)
/(4) == (2,0), /(24) == (3,1)
/(6) == (1,1), /(36) == (2,2)
/(8) == (3,0), /(72)== (3,2)

is a bijection from A to B. Thus, by (BP) and (MP), \A\ = \B\ =

|{0,1,2,3} x {0,1,2}| = |{0,1,2,3}| x |{0,1,2}| = 4 x 3 = 12, which
agrees with the above listing.
The following example extends what we discussed above.
Example 5.3 Find the number of divisors of 12600.
Solution Observe that 12600 = 2 3 x 3 2 x 5 2 x 7 1 .
Thus a number z is a divisor of 12600 if and only if it is of the form

z = 2° x 3 6 x 5C x 7d
 The Bijection Principle 45

where a,b,c,d are integers such that 0 < a < 3 , 0 < 6 < 2 , 0 < c < 2
and 0 < d < 1.
Let A be the set of divisors z of 12600 and B = {(a,b,c,d) : a —
0,1,2,3; 6 = 0,1,2; c = 0,1,2; d = 0,1}. Clearly, the mapping / denned
by

f{z) = (a,b,c,d),

is a bijection from A to B. Then, by (BP) and (MP), \A\ = | £ | =

4 • 3 • 3 • 2 = 72. D
We have seen from the above examples how crucial applying (BP)
is as a step towards solving a counting problem. Given a finite set A,
the objective is to enumerate \A\, but of course, this is not easy. In
the course of applying (BP), we look for a more familiar finite set B
and try to establish a bijection between these two sets. Once this is
done, the harder problem of counting |J4| is transformed to an easier
problem (hopefully) of counting \B\. It does not matter how different
the members in A and those in B are in nature. As long as there exists
a bijection between them, we get |vl| = |B|.

Exercise

5.1 (a) Find the number of positive divisors of n if

(i) n = 31752;
(ii) n = 55125.
(b) In general, given an integer n > 2, how do you find the number
of positive divisors of re?
5.2 In each of the following cases, find the number of shortest P—Q
routes in the grid below:

Q
vc
«
\\ft uB
%t
<

p
46 Counting

(i) the routes must pass through A;

(ii) the routes must pass through AB;
(iii) the routes must pass through A and C;
(iv) the segment AB is deleted.
5.3 For each positive integer n, show that ^ ( N ^ l = 2" by establishing
a bijection between V(Nn) and the set of n-digit binary sequences.
5.4 Let n and r be integers with 1 < r < n. Prove that (") = ( n " r )
by establishing a bijection between the set of r-element subsets of
N n and the set of (n — r)-element subsets of N n .
5.5 The number 4 can be expressed as a sum of one or more positive
integers, taking order into account, in the following 8 ways:

4=4=1+3
=3+1=2+2
=1+1+2=1+2+1
= 2 + 1 + 1 = 1 + 1 + 1 + 1.
Show that every natural number n can be so expressed in 2 n _ 1
ways.
5.6 How many rectangles are there in the following 6 x 7 grid?

5.7 Find the number of parallelograms which are contained in the

configuration below and which have no sides parallel to BC. (Hint:
Adjoin a new row at the base of the triangle.)
 Chapter 6

Distribution of Balls into Boxes

Figure 6.1 shows three distinct boxes into which seven identical
(indistinguishable) balls are to be distributed. Three different ways
of distribution are shown in Figure 6.2. (Note that the two vertical bars
at the two ends are removed.)

(i) (2) (3)

Figure 6.1

(a)
OOOO OO O

(b)
OO O O OOO

(c)
oooooo
Figure 6.2

47
48 Counting

In how many different ways can this be done? This is an example of

the type of problem we shall discuss in this chapter. We shall see how
problems of this type can be solved by applying (BP).
In Figure 6.2, by treating each vertical bar as a "1" and each ball as
a "0", each way of distribution becomes a 9-digit binary sequence with
two l's. For instance,
(a) • 000010010,
(b) > 001001000,
(c) • 0 110 0 0 0 0 0.
Obviously, this correspondence establishes a bijection between the set
of ways of distributing the balls and the set of 9-digit binary sequences
with two l's. Thus, by (BP), the number of ways of distributing the
seven identical balls into three distinct boxes is ( 2 ).
In general, we have:

The number of ways of distributing r identical balls into

ni distinct
distil boxes is given by ( r ^" x ), which is equal to (6.1)
r+ 1
( r ),by(3.6).

In the distribution problem discussed above, some boxes may be

vacant at the end. Supposing no box is allowed to be vacant, how many
ways are there to distribute the seven identical balls into three distinct
boxes?
To meet the requirement that no box is vacant, we first put a ball in
each box and this is counted as one way because the balls are identical.
We are then left with 4 (= 7 — 3) balls, but we are now free to distribute
these 4 balls into any box. By the result (6.1), the number of ways this
can be done is (^t^1) = (2) • Thus, the number of ways to distribute 7
identical balls into 3 distinct boxes so that no box is empty is (,).
In general, suppose we wish to distribute r identical balls into n
distinct boxes, where r > n, in such a way that no box is vacant. This
can be done in the following steps: First, we put one ball in each box;
and then distribute the remaining r - n balls to the n boxes in any
arbitrary way. As the balls are identical, the number of ways for the
 Distribution of Balls into Boxes 49

first step to be done is 1. On the other hand, by the result (6.1), the
number of ways to do the second step is

(r — n) + n — 1
n-1

Thus, by (MP) and upon simplification, we arrive at the following result.

The number of ways to distribute r identical balls into n

distinct boxes, where r > n, so that no box is empty is (6.2)
given by (£~J), which is equal to £ ~ J ) .

Example 6.1 There are 11 men waiting for their turn in a barber
shop. Three particular men are A, B and C. There is a row of 11 seats
for the customers. Find the number of ways of arranging them so that
no two of A, B and C are adjacent.

Solution There are different ways to solve this problem. We shall see
in what follows that it can be treated as a distribution problem.
First of all, there are 3! ways to arrange A, B and C. Fix one of the
ways, say A—B—C. We then consider the remaining 8 persons. Let us
imagine tentatively that these 8 persons are identical, and they are to be
placed in 4 distinct boxes as shown in Figure 6.3 so that boxes (2) and
(3) are not vacant (since no two of A,B and C are adjacent). To meet
this requirement, we place one in box (2) and one in box (3). Then the
remaining six can be placed freely in the boxes in i6^^1) = (jj) ways
by (6.1). (Figure 6.4 shows a way of distribution.)

Oo
o
ooo
S~
o B o
(1) (2) (3) (4)

Figure 6.3
50 Counting

QO A OO B O C OOP
(1) (2) (3) (4)

Figure 6.4

But the eight persons are actually distinct. Thus, to each of these
(3) ways, there are 8! ways to arrange them.
Hence by (MP), the required number of ways is 3!(3)8!, which is
8! 9 • 8 • 7. •

Remark The answer, 8! 9 • 8 • 7, suggests that the problem can be

solved in the following way. We first arrange the 8 persons (excluding
A, B and C) in a row in 8! ways. Fix one of these ways, say

X\ X2 X-} X4 X$ ^6 X7 Xg
(1) (2) (3) (4) (5) (6) (7) (8) (9)

We now consider A. There are 9 ways to place A in one of the 9 boxes,

say box (4):

X\ X2 X3 A A4 X5 X$ X-] Xg
HT ~W (3) (4) (5) ~W ~W (8) (9)
Next, consider B. Since A and B cannot be adjacent, B can be placed
only in one of the remaining 8 boxes. Likewise, C can be placed only in
one of the remaining 7 boxes. The answer is thus 8! 9 • 8 • 7.

Exercise

6.1 There are four types of sandwiches. A boy wishes to place an

order of 3 sandwiches. How many such orders can he place?
6.2 Calculate the number of distinct 9-letter arrangements which can
be made with letters of the word SINGAPORE such that no two
vowels are adjacent.
6.3 There is a group of 10 students which includes three particular
students A, B and C. Find the number of ways of arranging the
10 students in a row so that B is always between A and C. {A
and B, or B and C need not be adjacent.)
 Distribution of Balls into Boxes 51

6.4 Six distinct symbols are transmitted through a communication

channel. A total of 18 blanks are to be inserted between the
symbols with at least 2 blanks between every pair of symbols.
In how many ways can the symbols and blanks be arranged?
This page is intentionally left blank
 Chapter 7

More Applications of (BP)

We shall give additional examples in this chapter to show more appli-

cations of (BP).
Consider the following linear equation:
xi + x2 + x3 = 7. (1)
If we put x\ = 4, X2 = 1 and x% = 2, we see that (1) holds. Since
4, 1, 2 are nonnegative integers, we say that (xi,X2,xz) = (4,1,2) is
a nonnegative integer solution to the linear equation (1). Note that
(x\,X2,xs) = (1,2,4) is also a nonnegative integer solution to (1), and
so are (4, 2, 1) and (1, 4, 2). Other nonnegative integer solutions to (1)
include
(0,0,7), (0,7,0), (1,6,0), (5,1,1),... .
Example 7.1 Find the number of nonnegative integer solutions to (1).
Solution Let us create 3 distinct "boxes" to represent xi,x2 and
£3, respectively. Then each nonnegative integer solution (xi,X2,X3) =
(a,b,c) to (1) corresponds, in a natural way, to a way of distributing 7
identical balls into boxes so that there are a, b and c balls in boxes (1),
(2) and (3) respectively (see Figure 7.1).
This correspondence clearly establishes a bijection between the set of
nonnegative integer solutions to (1) and the set of ways of distributing 7
identical balls in 3 distinct boxes. Thus, by (BP) and the result of (6.1),
the number of nonnegative integer solutions to (1) is ( ^_Y ) — (2)- ^

53
54 Counting

(4, 1, 2) -> 0000 | 0 | 00

(1) (2) (3)

(2, 5,0) -> 00 | 00000 1

(1) (2) (3)

Figure 7.1

By generalizing the above argument and applying the results (6.1)

and (6.2), we can actually establish the following general results.

Consider the linear equation

x\ + a>2 H \-xn = r (2)

where r is a nonnegative integer.
(i) The number of nonnegative integer solutions to (2) is (7.1)
given by (r+^1).
(ii) The number of positive integer solutions (x\,X2,...,
xn) to (2), with each Xj > 1, is given by (£Z„)> where
r > n and i = 1,2,... ,n.

Example 7.2 Recall that the number of 3-element subsets {a, b, c} of

the set N 1 0 = { 1 , 2 , 3 , . . . ,10} is (13°). Assume that a < b < c and
suppose further that

b-a>2 and c - b > 2 (3)

(i.e. no two numbers in {a,b,c} are consecutive). For instance, {1,3,8}
and {3,6,10} satisfy (3) but not {4,6,7} and {1,2,9}. How many such
3-element subsets o/Nio are there?
Solution Let us represent a 3-element subset {a, b, c} of Nio satisfying
(3) by a 10-digit binary sequence as follows:

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
{1,3,8} - + 1 0 1 0 0 0 0 1 0 0
{3,6,10} - + 0 0 1 0 0 1 0 0 0 1
 More Applications of (BP) 55

Note that the rule is similar to the one introduced in Example 5.2.
Clearly, this correspondence is a bijection between the set A of 3-element
subsets of Nio satisfying (3) and the set B of 10-digit binary sequences
with three l's in which no two l's are adjacent. Thus \A\ = \B\. But
how do we count \B\1 Using the method discussed in Example 6.1, we
obtain

w-C?-*4-1)-®-
Thus |A| = (5). •
Example 7.3 Two tennis teams A and B, consisting of 5 players
each, will have a friendly match playing only singles tennis with no ties
allowed. The players in each team are arranged in order:
A : ai,a2,a 3 ,a 4 ,a5 ,
B : h,b2,b3,b4,b5.
The match is run in the following way. First, a\ plays against b\. Sup-
pose ai wins (i.e. b\ is eliminated). Then a\ continues to play against
62; if ai is beaten by 62 (&-e. a\ is eliminated), then 62 continues to play
against a2, and so on. What is the number of possible ways in which all
the 5 players in team B are eliminated? (Two such ways are shown in
Figure 7.2.)
Solution Let Xi be the number of games won by player a^, i —
1,2,3,4,5. Thus, in Figure 7.2(i),
x\ = 2, £2 = 0 j £3 = 3 , £4 = X5 = 0
and in Figure 7.2(h),
xi = x2 — 0, X3 = 2 , £4 = 1, £5 = 2.
In order for the 5 players in team B to be eliminated, we must have
xi + x2 + x3 + £4 + x5 = 5 (4)
and the number of ways this can happen is, by (BP), the number
of nonnegative integer solutions to (4). Thus, the desired answer is
( 5 + 4 - 1 ) = (4),bytheresult7.1(i). •
56 Counting

a,\ a-i a-?, «4 a$

b\ b2 b3 b4 b5
(i)

a\ ai an, a± a$

\AAAA
b\ 62 ^3 i>4 b$
(ii)
"o ->• 6" means "a beats 6"

Figure 7.2

Example 7.4 Eight letters are to be selected from the five vowels
a,e,i,o,u with repetition allowed. In how many ways can this be done
if
(i) there are no other restrictions?
(ii) each vowel must be selected at least once?

Solution (i) Some examples of ways of the selection are given below:

(1) a,a,u,u,u,u,u,u;
(2) a, e,i,i,i,o, o, u;
(3) e,e,i,i,o,o,u,u.

As shown in Figure 7.3, these selections can be treated as ways of dis-

tributing 8 identical objects into 5 distinct boxes.

(i) < > • • 1 I 1 IS 11

(2) < > • • • • • • • •
(3) <—> l»«l»»l«»l»r
a e i o u
Figure 7.3
 More Applications of (BP) 57

Thus, by (BP) and the result (6.1), the number of ways of selection
is given by f ^ - 1 ) , i.e. ft2).
(ii) As shown in the second row of Figure 7.3, a way of selection
which includes each vowel can be treated as a way of distribution such
that no box is empty. Thus, by (BP) and the result (6.2), the number
of ways of selection is given by (g~g), i.e. Q . •
Example 7.5 Consider the following two 13-digit binary sequences:

1110101110000,
1000110011110.
For binary sequences, any block of two adjacent digits is of the form
00,01,10 or 11. In each of the above sequences, there are three 00, two
01, three 10 and four 11. Find the number ofl3-digit binary sequences
which have exactly three 00, two 01, three 10 and four 11.

Solution To have exactly three 10 and two 01 in a sequence, such

a sequence must begin with 1, end with 0, and have the changeovers
of 1 and 0 as shown below, where each of the boxes (1), (3) and (5)
(respectively (2), (4) and (6)) contains only l's (respectively 0's) and at
least one 1 (respectively 0).

[10] [01] [10] [01] [10]

I 1 I 0 I 1 1 0 I 1 I 0 I
(1) (2) (3) (4) (5) (6)

For instance, the two sequences given in the problem are of the form:

I 111 I 0 I 1 I 0 I 111 I 0000 |

(1) (2) (3) (4) (5) (6)

1 1 | 000 | 11 1 00 1 1111 | 0 |
(1) (2) (3) (4) (5) (6)

To have three 00 and four 11 in such a sequence, we must

(i) put in three more 0's in boxes (2), (4) or (6) (but in an arbitrary
way), and
(ii) put four more l's in boxes (1), (3) or (5) (also in an arbitrary way).
58 Counting

(Check that there are 13 digits altogether.) The number of ways to do

(i) is ( 3 + ^ 1 ) , i.e. g ) ; while that of (ii) is ( 4 + 4 _ 1 ), i.e. (5). Thus, by
(MP), the number of such sequences is (2) (2)) l-e- 150. •

Example 7.6 Consider the following three arrangements of 5 persons

A,B,C,D,E in a circle:

(i)

Two arrangements ofn objects in a circle are considered different if and

only if there is at least one object whose neighbour on the right is differ-
ent in the two arrangements. Thus arrangements (i) and (ii) above are
considered identical, while arrangement (iii) is considered different from
(i) and (ii). (Note that the right neighbour of A in arrangement (iii) is
C while that in both (i) and (ii) is B.) Find the number of arrangements
of the 5 persons in a circle.

Solution For each arrangement of the 5 persons in a circle, let us line

the 5 in a row as follows: We always start with A at the left end. Then
we place the right neighbour of A (in the circle) to the right of A in the
row. We continue, in turn, to place the right neighbour (in the circle)
of the last placed person to his right in the row until every person is
arranged in the row. (We can also visualize this as cutting the circle at A
and then unraveling it to form a line.) Then each circular arrangement
of the 5 persons corresponds to an arrangement of 5 persons in a row
with A at the left end. Now, since A is always fixed at the left end,
he can be neglected and the arrangement of 5 persons in a row can be
seen to correspond to an arrangement of only 4 persons (B,C,D,E) in
a row (see Figure 7.5).
 More Applications of (BP) 59

B,
-> ABCDE- -> BCDE

'D

->ACBDE- ->CBDE

Figure 7.5

This correspondence clearly establishes a bijection between the set

of arrangements of 5 persons in a circle and the set of arrangements of
4 persons in a row. Thus, by (BP) and the result of (3.1), the number
of arrangements of 5 persons in a circle is 4!. •

By generalizing the above argument, we can establish the following

result:

The number of ways of arranging n distinct objects in a

(7.2)
circle is given by (n — 1)!.

Exercise

7.1 Find the number of integer solutions to the equation:

in each of the following cases:

(i) Xi > 0 for each i = 1,2,..., 5;
(ii) x\ > 3, X2 > 5 and Xi > 0 for each i = 3,4,5;
(iii) 0 < #1 < 8 and X{ > 0 for each i = 2,3,4,5;
(iv) X\ + X2 = 10 and x; > 0 for each i = 1,2,..., 5;
60 Counting

(v) Xi is positive and odd (respectively, even) for each i =

1,2,...,5.
7.2 An illegal gambling den has 8 rooms, each named after a different
animal. The gambling lord needs to distribute 16 tables into the
rooms. Find the number of ways of distributing the tables into
the rooms in each of the following cases:
(i) Horse Room holds at most 3 tables.
(ii) Each of Monkey Room and Tiger Room holds at least 2 ta-
bles.
7.3 The number 6 can be expressed as a product of three factors in
9 ways as follows:
1-1-6, 1-6-1, 6-1-1, 1-2-3, 1-3-2, 2-1-3, 2-3-1, 3-1-2, 3-2-1.
In how many ways can each of the following numbers be so
expressed?
(i) 2592
(ii) 27000
7.4 Find the number of integer solutions to the equation:

£i + ^2 + ^3 + X4 — 30
in each of the following cases:
(i) Xi > 0 for each i = 1,2,3,4;
(ii) 2 < x\ < 7 and Xi > 0 for each i = 2,3,4;
(iii) x\ > —5,X2 > —1,0:3 > 1 and £4 > 2.
7.5 Find the number of quadruples (w, x, y, z) of nonnegative integers
which satisfy the inequality
w + x + y + z < 2002.
7.6 Find the number of nonnegative integer solutions to the equation:

5a:i + x2 + 23 + £4 = 14.
7.7 There are five ways to express 4 as a sum of two nonnegative
integers in which the order matters:

4=4+0=3+1=2+2=1+3=0+4.
 More Applications of (BP) 61

Given r, n € N, what is the number of ways to express r as a sum

of n nonnegative integers in which the order matters?
7.8 There are six ways to express 5 as a sum of three positive integers
in which the order matters:

5 = 3+1+1 = 2+2+1 = 2+1+2 = 1+3+1 = 1+2+2 = 1+1+3.

Given r, n € N with r >n, what is the number of ways to express
r as a sum of n positive integers in which the order matters?
7.9 Find the number of 4-element subsets {a, b, c, d} of the set N20 =
{ 1 , 2 , . . . , 20} satisfying the following condition

6-a>2,c-6>3 and d-c>4.

7.10 In a sequence of coin tosses, one can keep a record of the

number of instances when a tail is immediately followed by a
head, a head is immediately followed by a head, etc. We de-
note these by TH,HH, etc. For example, in the sequence
HHTTHHHHTHHTTTT of 15 coin tosses, we observe that
there are five HH, three HT, two TH and four TT subsequences.
How many different sequences of 15 coin tosses will contain ex-
actly two HH, three HT, four TH and five TT subsequences?
(AIME)
7.11 Show that the number of ways of distributing r identical objects
into n distinct boxes such that Box 1 can hold at most one object
is given by

/r + n - 3 \ fr + n-2\
( , - , ) + ( r J'
7.12 In a new dictatorship, it is decided to reorder the days of the
week using the same names of the days. All the possible ways of
doing so are to be presented to the dictator for her to decide on
one. How many ways are there in which Sunday is immediately
after Friday and immediately before Thursday?
7.13 Five couples occupy a round table at a wedding dinner. Find the
number of ways for them to be seated if:
(i) every man is seated between two women;
62 Counting

(ii) every man is seated between two women, one of whom is his
wife;
(iii) every man is seated with his wife;
(iv) the women are seated on consecutive seats.
7.14 The seats at a round table are numbered from 1 to 7. Find the
number of ways in which a committee consisting of four men and
three women can be seated at the table
(i) if there are no restrictions;
(ii) if all the men sit together. ,„-.

7.15 Four men, two women and a child sit at a round table. Find
the number of ways of arranging the seven people if the child is
seated
(i) between the two women;
(ii) between two men. ,_,*
 Chapter 8

Distribution of Distinct Objects into

Distinct Boxes

We have seen from the various examples given in Chapters 6 and 7

that the distribution problem, which deals with the counting of ways
of distributing objects into boxes, is a basic model for many counting
problems. In distribution problems, objects can be identical or distinct,
and boxes too can be identical or distinct. Thus, there are, in general,
four cases to be considered, namely
Table 8.1
Objects Boxes

(1) identical distinct

(2) distinct distinct
(3) distinct identical
(4) identical identical

We have considered Case (1) in Chapters 6 and 7. Cases (3) and (4) will
not be touched upon in this booklet. In this chapter, we shall consider
Case (2).
Suppose that 5 distinct balls are to be put into 7 distinct boxes.
Example 8.1 In how many ways can this be done if each box can
hold at most one ball?
Example 8.2 In how many ways can this be done if each box can
hold any number of balls?
63
64 Counting

Solution Before we proceed, we would like to point out that the

ordering of the distinct objects in each box is not taken into con-
sideration in the discussion in this chapter.
We first consider Example 8.1. As shown in Figure 8.1, let a,b,c,d
and e denote the 5 distinct balls. First, we put a (say) into one of the
boxes. There are 7 choices.

I
(1) (2) (3) (4) (5) (6) (7)

Figure 8.1

Next, we consider b (say). As each box can hold at most one ball,
and one of the boxes is occupied by a, there are now 6 choices for
b. Likewise, there are, respectively, 5, 4 and 3 choices for c, d and
e. Thus, by (MP), the number of ways of distribution is given by
7-6-5-4-3.
Note that the above answer can be expressed as P j which, as denned
in Chapter 3, is the number of ways of arranging any 5 objects from
7 distinct objects. The fact that the above answer is P j does not sur-
prise us as there is a 1-1 correspondence between the distributions of
5 distinct balls into 7 distinct boxes and the arrangements of 5 distinct
objects from 7 distinct objects as shown in Figure 8.2. (Find out the
rule of the correspondence.)

{a,b,c,d,e} {1,2,3,4,5,6,7}

b c a e d -> 41275
e d c b a -> 74321
(1) (2) (3) (4) (5) (6) (7)

Figure 8.2
 Distribution of Distinct Objects into Distinct Boxes 65

In general, we have:

The number of ways of distributing r distinct objects into

n distinct boxes such that each box can hold at most one
(8.1)
object (and thus r < n) is given by P™, which is equal to
n\/(n — r)\.

We now consider Example 8.2. There are 7 ways of putting a in

the boxes. As each box can hold any number of balls, there are also 7
choices for each of the remaining balls b, c, d and e. Thus, by (MP), the
answer is 7 5 .
In general, we have:

The number of ways of distributing r distinct objects into

n distinct boxes such that each box can hold any number (8.2)
of objects is given by nr.

Exercise

8.1 Find the number of ways for a teacher to distribute 6 different

books to 9 students if
(i) there is no restriction;
(ii) no student gets more than one book.
8.2 Let A be the set of ways of distributing 5 distinct objects into
7 distinct boxes with no restriction, and let B be the set of 5-digit
numbers using 1, 2, 3, 4, 5, 6, 7 as digits with repetition allowed.
Establish a bijection between A and B.
8.3 Five friends go to a Cineplex which contains 6 theatres each
screening a different movie and 2 other theatres screening the
current blockbuster. Find the number of ways the friends can
watch a movie in each of the following cases:
(i) two of the friends must be together;
(ii) the theatres do not matter, only the movies do.
 Counting

Find the number of ways of distributing 8 distinct objects into

3 distinct boxes if each box must hold at least 2 objects.
Suppose that m distinct objects are to be distributed into n dis-
tinct boxes so that each box contains at least one object. State
a restriction on m with respect to n. In how many ways can the
distribution be done if
(i) m — n?
(ii) m = n + 1?
(iii) m = n + 2?
 Chapter 9

Other Variations of the

Distribution Problem

Two cases of the distribution problem were discussed in the preceding

chapters. In this chapter, we shall study some of their variations.
When identical objects are placed in distinct boxes, whether the
objects in each box are ordered or not makes no difference. The situa-
tion is no longer the same if the objects are distinct as shown in
Figure 9.1.

| de \ ab c \ | [ ed \ cba | |
(1) (2) (3) (1) (2) (3)

Figure 9.1

In Chapter 8, we did not consider the ordering of objects in each

box. In our next example, we shall take it into account.

Example 9.1 Suppose that 5 distinct objects a, b, c, d, e are distributed

into 3 distinct boxes, and that the ordering of objects in each box matters.
In how many ways can this be done?

Solution First, consider a (say). Clearly, there are 3 choices of a box

for a to be put in (say, a is put in box (2)). Next, consider b. The object
b can be put in one of the 3 boxes. The situation is special if 6 is put
in box (2) because of the existence of a in that box. As the ordering of
objects in each box matters, if b is put in box (2), then there are two

67
68 Counting

choices for b, namely, left of a or right of a as indicated in Figure 9.2.

Thus, altogether, there are 4 choices for b.

Assume that b is put in box (3). Now, consider c. As shown in

Figure 9.3, c has 5 choices.

Figure 9.3

Continuing in this manner, we see that d and e have, respectively, 6 and

7 choices. Thus, the answer is given by 3 • 4 • 5 • 6 • 7. •
Let us try a different approach to solve the above problem. First, we
pretend that the objects a, b, c, d, and e are all identical. The number
of ways of distributing 5 identical objects into 3 distinct boxes is, by
result (6.1), ( 5 + 5 _ 1 ), i-e. Q . Next, take such a way of distribution, say,

00 000
(1) (2) (3)

Since the 5 objects are actually distinct and the ordering of objects mat-
ters, such a distribution for identical objects corresponds to 5! different
distributions of distinct objects. Thus, by (MP), the answer is given by
(2) • 5! which agrees with the answer 3 • 4 • 5 • 6 • 7.
In general, we have:
 Other Variations of the Distribution Problem 69

The number of ways of distributing r distinct objects into

n distinct boxes such that the ordering of objects in each
box matters is given by

(9.1)

which is equal to
n(n + l)(n + 2 ) . . . (n + r - 1).

In our previous discussion on the distribution problem, objects were

either all identical or all distinct. We now consider a case that is a
mixture of these two.
Example 9.2 Four identical objects "a", three identical objects "b"
and two identical objects "c" are to be distributed into 9 distinct boxes
so that each box contains one object. In how many ways can this be
done?
Solution Let's start with the four a's. Among the 9 boxes, we choose
4 of them, and put one a in each chosen box. Next, we consider the
three 6's. From among the 5 remaining boxes, we choose 3, and put one
b in each chosen box (see Figure 9.4). Finally, we put one c in each of
the 2 remaining boxes.

Figure 9.4

There are Q ways for step 1, (3) ways for step 2 and (%)(= 1) way for
step 3. Thus, by (MP), the answer is given by
70 Counting

Remark In the above solution, a is considered first, followed by b

and finally c. The answer is independent of this order. For instance,
if b is considered first, followed by c and then a, by applying a similar
9!
argument we arrive at (3) (2) (4), which is again 4!3!2! •
There is a 1-1 correspondence between the distributions considered
in Example 9.2 and the arrangements of 4 a's, 3 6's and 2 c's in a row
as shown in Figure 9.5.

1a c a b a a b b c | <- acabaabbc

1c c b b b a a a a I <- ccbbbaaaa

Figure 9.5

Thus, by the result of Example 9.2, the number of arrangements of 4

a's, 3 b's and 2 c's in a row is given by
9!
4!3!2!
In general,

Suppose there are n\ identical objects of type 1, n<i

identical objects of type 2 , . . . , and n& identical objects
of type k. Let n — n\ + ri2 +... + rik- Then the number
of arrangements of these n objects in a row is given by

n \ j n — n\ n •ni ra/fc-i (9.2)

n\) \ n2 «fc
which is equal to
n\
n\\n2\. ,nk\

Let us reconsider Example 9.1. We observe that there is a 1-1 cor-

respondence between the distributions considered in Example 9.1 and
the arrangements of a, b, c, d, e and two l's as shown in Figure 9.6.
By the above result, the number of arrangements of a,b,c,d,e and
two l's is given by | | , which agrees also with the earlier two answers.
 Other Variations of the Distribution Problem 71

be a dc ; > be\a\dc
cde ba : > cde 11 ba

(1) (2) (3)

Figure 9.6

Exercise
9.1 Calculate the number of different arrangements which can be made
using all the letters of the word BAN AN ARAM A.
9.2 Calculate the number of distinct 8-letter arrangements which can
be made with letters of the word INFINITE. How many of these
begin with III?
(C)
9.3 Find the number of arrangements of 4 identical squares, 5 identical
pentagons and 6 identical hexagons in a row if
(i) there is no restriction;
(ii) no two pentagons are adjacent;
(iii) any two squares are separated by at least two other polygons.
9.4 Let A = { 1 , 2 , . . . , m} and B = { 1 , 2 , . . . , n} where m, n > 1. Find
the number of
(i) mappings from A to B;
(ii) 1-1 mappings from A to B (here m < n);
(iii) mappings f : A —t B such that /(«') < f(j) in B whenever
i < j in A (here m < n);
(iv) mappings f : A-* B such that / ( l ) = 1.
9.5 Let A = { 1 , 2 , . . . , m} and B = { 1 , 2 , . . . , n}. Find the number of
onto mappings from A to B in each of the following cases:
(i) m = n;
(ii) m = n + 1;
(iii) m = n + 2.
(Compare this problem with Problem 8.5.)
9.6 Ten cars take part in an Automobile Association of Singapore
autoventure to Malaysia. At the causeway, 4 immigration counters
are open. In how many ways can the 10 cars line up in a 4-line
queue?
72 Counting

9.7 Solve Problem 8.5 with an additional condition that the ordering
of objects in each box counts.
9.8 Show that
n\ (n — ni\ (n — ni — • • • — rik-\\ n\
n
n\) \ 712 / V h ) n\\n^. • • • n\^.
where n = n\ + ni -\ 1- n^.
 Chapter 10

The Binomial Expansion

In Chapter 3, we introduced a family of numbers which were denoted

by (™) or C™. Given integers n and r with 0 < r < n, the number (") is
defined as the number of r-element subsets of the set N n = { 1 , 2 , . . . , n}.
That is, (") is the number of ways of selecting r distinct objects from a
set of n distinct objects. We also derived the following formula for ("):

(10.1)

By applying (10.1), or otherwise, we can easily derive some interesting

identities involving these numbers such as:

(10.2)
\r J \n — r J '

0 -C-lH'r 1
)- (10.3)

1 2 1
(10.4)
( - ( - l ) - ' -

73
74 Counting

(10.5)

In this chapter, we shall learn more about this family of numbers

and derive some other important identities involving them.
In algebra, we learn how to expand the algebraic expression (1 + x)n
for n = 0,1,2,3. Their expansions are shown below:

(1 + *)° = 1,
(l + x)1 = l + xt
(1 + x)2 = 1 + 2x + x2,
(1 + x)3 = 1 + 3x + 3x 2 + x 3 .

Notice that the coefficients in the above expansions are actually numbers
of the form ("). Indeed, we have:

- ( c ) ' •-(!)• ' - ( 0 -

- G ) . - ( ; ) • - ( ; ) • HD-
What can we say about the coefficients in the expansion of (1 + x) 4 ?
Will we obtain

^"••(;) + (0- + (J)- ,+ (J)' + (J)' ,T

Let us try to find out the coefficient of x 2 in the expansion of (1 + x) 4 .
We may write

(1 + x) 4 = (1 + x)(l + x)(l + x)(l + x ) .

(1) (2) (3) (4)
 The Binomial Expansion 75

Table 10.1

(1) (2) (3) (4)

X X
X X
X X
X X
X X

X X

Observe that in the expansion, each of the factors (1), (2), (3) and (4)
contributes either 1 o n , and they are multiplied together to form a
term. For instance, to obtain x2 in the expansion, two of (1), (2), (3)
and (4) contribute x and the remaining two contribute 1. How many
ways can this be done? Table 10.1 shows all the possible ways, and the
answer is 6.
Thus, there are 6 terms of a;2 and the coefficient of a;2 in the expansion
of (1 + x) 4 is therefore 6. Indeed, to select two x's from four factors
(1 + x), there are (*,) ways (while the remaining two have no choice
but to contribute "1"). Thus the coefficient of x 2 in the expansion of
(1 + x) 4 is (2) which is 6. Using a similar argument, one can readily see
that

In general, what can be said about the expansion of (1 + x)n, where

n is any natural number?
Let us write
(l + x) n = (l + x)(l + x ) . . . ( l + x ) . (*)
(1) (2) (n)
To expand (1 + x) n , we first select 1 or x from each of the n factors
(1 + x), and then multiply the n chosen l's and x's together. The
general term thus obtained is of the form x r , where 0 < r < n. What
is the coefficient of xr in the expansion of (1 + x)n if the like terms are
grouped? This coefficient is the number of ways to form the term xr in
the product (*). To form a term xr, we choose r factors (1 + x) from the
76 Counting

n factors (1 + x) in (*) and select x from each of the r chosen factors.

Each of the remaining n — r factors (1 + x) has no other option but
to contribute 1. Clearly, the above selection can be done in (™) ways.
Thus, the coefficient of xr in the expansion of (1 + x)n is given by (™).
We thus arrive at the following result:

The Binomial Theorem (BT)

For any natural number n,

(i+.,--(;)+ l +1: ; ) x2 + --- (10.6)

::)•
»• + ... + )xn.
< >

Exercise

10.1 By applying Identity (10.1), or otherwise, derive the following

identities:

(n) oo = {r)Cz)-
10.2 In the expansion of (1 + x) 1 0 0 , it is known that the coefficients of
xr and x3r, where 1 < r < 33, are equal. Find the value of r.
10.3 What is the largest value of k such that there is a binomial ex-
pansion (1 + x)n in which the coefficients of k consecutive terms
are in the ratio 1 : 2 : 3 : . . . : kl Identify the corresponding
expansion and the terms.
 Chapter 11

Some Useful Identities

We gave four simple identities involving binomial coefficients, namely

(10.2)-(10.5), in Chapter 10. In this chapter, we shall derive some more
identities involving binomial coefficients from (BT). These identities,
while interesting in their own right, are also useful in simplifying certain
algebraic expressions.
Consider the expansion of (1 + x)n in (BT). If we let x = 1, we then
obtain from (BT) the following

(n\ fn^ (n*) fn\

+ + + • • + = 2n .
W W
Example 11.1 In Example 5.2, we discussed a counting problem on
V(S), the set of all subsets of a finite set S. If S is an n-element
set (i.e. | 5 | = n), it can be shown (see Problem 5.3) by establishing
a bijection between V(S) and the set of n-digit binary sequences that
there are exactly 2 n subsets of S inclusive of the empty set (f> and the
set S itself (i.e. ^(S 1 )! = 2n). We can now give a more natural proof
for this fact. Assume that \S\ — n. By definition, the number of

0-element subsets of S is
CD
1-element subsets of S is
"
77
78 Counting

2-element subsets of S is n

Thus,
n-element subsets of S is
0
n n
\ns)\ = +1 l )+ + ••• +
(by (Bl)) •
Example 11.2 The number 4 can be expressed as a sum of one or
more positive integers, taking order into account, in the following 8
ways:
4=4=1+3
=3+1=2+2
=1+1+2=1+2+1
= 2 + 1 + 1 = 1 + 1 + 1 + 1.
Show that every natural number n can be so expressed in 2 n _ 1 ways.
Solution This is in fact Problem 5.5. Let us see how (Bl) can be
used to prove the result. But first of all, consider the special case above
when n = 4.
We write 4 = 1 + 1 + 1 + 1 and note that there are three "+"s in the
expression. Look at the following relation.
4 <—y 1 + 1 + 1 + 1 (no "+" is chosen)
4

> (one "+" is chosen)

 Some Useful Identities 79

l + l + 2<—• 1 i 1 ®1+1
1 1 2
1 + 2 + 1 ^—> 1 i + ie l (2 "+"s are chosen)
l 2 1

2 + 1 + 1 <—n + i e l © I
2 1 1
1+ 1 + 1 + 1^—>-^l^©l©l (3 "+"s are chosen)
1 1 1

This correspondence is actually a bijection between the set of all such

expressions of 4 and the set of all subsets of three "+"s. Thus, by (BP)
and (Bl), the required answer is

; • : + « • : > - * •
In general, write

n = l + l + --- + l + l
" v '
n
and note that there are n — 1 "+"s in the above expression. We now
extend the above technique by establishing a bijection between the set
of all such expressions of n and the set of all subsets of n — 1 "+"s.
Thus, by (BP) and (Bl), the number of all such expressions of n is

n - l \ (n — \ \ (n — \ \ „_i
+ + + 1=2 D
o i - U-i
Consider again the expansion of (1 + x)n in (BT). If we now let
x — — 1, we then have

where the terms on the LHS alternate in sign. Thus, if n is even, say
n = 2k, then

n \ (n ,
o) +
( 2
l +
- +
U) - (" + f
3 +
- +
Ufc-1
80 Cc writing

and if n is odd, say n = 2k + 1, then

(:) + (;) + - + U)-(T) + (:) + -- n

2k + 1
As

+ + + = 2n
[(:)•(;)••••] [Q G) -]
by (Bl), we have:

c; + it + e !+•
n
= -(2 ) = 2 " -1
.
e +-
(B2)

Li

Example 11.3 yl /mite set S is said to be "even" ("odd") if \S\ is

even (odd). Consider Ns = { 1 , 2 , . . . , 8}. How many even (odd) subsets
does Ns have?

Solution The number of even subsets of Ns is

and the number of odd subsets of Ns is

By (B2),

!+G'+
'8
0 + + ••• +
+
= 2 8 _ 1 .= 2 7 = 128. •
Consider the following binomial expansion once again:

(1 + x)n = n n n n z3 + n
+ l1)*+l2l*' + + n
x
 Some Useful Identities 81

If we treat the expressions on both sides as functions of x, and differen-

tiate them with respect to x, we obtain:

^i+.r- = (-)+8(;),+,(;)^+...+-(;)^-.
By letting x — 1 in the above identity, we have:

n
In * fn \
E*l -f? )+2| | + » | c n ^l + - • + n ( ' : )
k=i
\ k i 3

= n2 n "-1

Let us try to derive (B3) by a different way. Consider the following

problem. Suppose that there are n(n > 1) people in a group, and they
wish to form a committee consisting of people from the group, including
the selection of a leader for the committee. In how many ways can this
be done?
Let us illustrate the case when n = 3. Suppose that A, B, C are the
three people in the group, and that a committee consists of k members
from the group, where 1 < k < 3. For k = 1, there are 3 ways to do so
as shown below.

Committee members Leader

A A
B B
C C

For k — 2, there are 6 ways to do so as shown below.

Committee members Leader

A, B A
A, B B
A, C A
A, C C
B, C B
B, C C
82 Counting

For k = 3, there are 3 ways to do so as shown below.

Committee members Leader

A, B, C A
A, B, C B
A, B, C C

Thus, there are altogether 3 -f- 6 + 3 = 12 ways to do so.

In general, from a group of n people, there are (£) ways to form a
A;-member committee, and k ways to select a leader from the k mem-
bers in the committee. Thus, the number of ways to form a fc-member
committee including the selection of a leader is, by (MP), &(£). As k
could be 1,2,... ,n, by (AP), the number of ways to do so is given by

Ek
fc=i
Let us count the same problem via a different approach as follows.
First, we select a leader from the group, and then choose k — 1 members,
where k = 1,2,..., n, from the group to form a fc-member committee.
There are n choices for the first step and
n-1 n 1 n —l
4- + ••• +
0 n-1
ways for the second step. Thus, by (MP) and (Bl), the required number
is
n-1 n-1 n-1 = n2"_1
n 0 + + ••• +
n-1
Since both

and n2,n—1
fc=l
count the same number, identity (B3) follows.
In the above discussion, we establish identity (B3) by first intro-
ducing a "suitable" counting problem. We then count the problem in
two different ways so as to obtain two different expressions. These two
 Some Useful Identities 83

different expressions must be equal as they count the same quantity.

This way of deriving an identity is quite a common practice in com-
binatorics, and is known as "counting it twice".

Exercise

11.1 By applying Identity (10.5) or otherwise, show that

k=r
11.2 Show that
n-1

£(V)-
fc=0 V y
o2n-2

11.3 Show that

by integrating both sides of (1 + x)n = "^2^=0 (fc)3^ w ** n res

Pect
to x.
11.4 Show that

f>s(Y)=n(n+1)2"-*.
v J
fc=l

11.5 Solve Example 11.2 by using result (7.1)(ii).

This page is intentionally left blank
 Chapter 12

Pascal's Triangle

In Chapter 10, we established the Binomial Theorem (BT) which states

that for all nonnegative integers n,

( i+ *> B =E(?y
r=0 ^ '
Let us display the binomial coefficients row by row following the
increasing values of n as shown in Figure 12.1. We observe from
Figure 12.1 the following.
1. The binomial coefficient at a lattice point counts the number of
shortest routes from the top lattice point (representing (g)) to the
lattice point concerned. For example, there are (2) (= 6) shortest

Figure 12.1

85
86 Counting

routes from the lattice point representing (Q) to the lattice point (2)
(also see Example 5.1).
2. The number pattern is symmetric with respect to the vertical line
through the top lattice point, and this observation corresponds to
the identity (?) = ( ^ r ) (see (10.2)).
3. Any binomial coefficient represented by an interior lattice point is
equal to the sum of the two binomial coefficients represented by the
lattice points on its "shoulders" (see Figure 12.2). This observation
corresponds to the identity (?) = (?~J) + (n;1) (see (10.3)).

Figure 12.2

4. The sum of the binomial coefficients in the nth row is equal to 2 n

and this fact corresponds to the identity

The number pattern of Figure 12.1 was known to Omar Khayyam

and Jia Xian around 1100 AD. The pattern was also found in the book
written by the Chinese mathematician Yang Hui in 1261, in which Yang
Hui called it, the Jia Xian triangle. The number pattern in the form
of Figure 12.3 was found in another book written by another Chinese
mathematician Zhu Shijie in 1303.
However, the number pattern of Figure 12.1 is generally called
Pascal's Triangle in memory of the great French mathematician Blaise
Pascal (1623-1662) who also applied the "triangle" to the study of
probability, a subject dealing with "chance". For a history of this number
pattern, readers are referred to the book Pascal's Arithmetical Triangle
by A. W. F. Edwards (Oxford University Press (1987)).
 Pascal's TYiangle 87

M$ & * & %

At^k^m <wk%*k*&m>m

Figure 12.3

Blaise Pascal

Look at Pascal's triangle of Figure 12.4.

What is the sum of the six binomial coefficients enclosed in the
shaded rectangle? The answer is 56. Note that this answer appears
as another binomial coefficient located at the right side of 21 in the
next row. Is this situation just a coincidence? Let us take a closer look.
88 Counting

1 1
1 2 1
1 3 3 1
1 4 6 4 1
l 5 10 10 5 1
1 6 IS 20 15 6 1
1 7 21 35 35 21 7
1 8 28 56 70 56 28 8

Figure 12.4

Observe that

. + . + . + » + » + »-G) + G) + G) + G) + (;) + G)
+ +
-G)*G) G)+G)*G) G)
(- G) • G))
+
-G)*G)*G)*G) G)
+
-G) G)*G)*G)
+ +
-G) G) G)
-GMD
-G) (=56)
by applying the identity ( £ } ) + ( n ; x ) = (?).
The above result is really a special case of a general situation. As
a matter of fact, the above argument can also be used to establish the
following general result (also see Figure 12.5):
 Pascal's Triangle

Figure 12.5

Figure 12.6

(B4)

By the symmetry of Pascal's triangle, one obtains the following accom-

panying identity of (B4) (also see Figure 12.6):

(k\ fk + l> (k + m\ fk + m + l\ (B5)

+ +• • + == (
K A ;f { m J \ m )

To end this chapter, we show an application of identity (B4) in

the solution of the following problem which appeared in International
Mathematical Olympiad 1981.
90 Counting

Example 12.1 Let 1 < r < n and consider all r-element subsets of
the set { 1 , 2 , . . . ,n}. Each of these subsets has a smallest member. Let
F(n, r) denote the arithmetic mean of these smallest numbers. Prove
that
n+ 1
F(n,r) =
r+1

Solution As an illustration of this problem, we consider the case

when n = 6 and r = 4. There are (^) (= 15) 4-element subsets of
the set {1,2,3,4,5,6}. They and their "smallest members" are listed in
Table 12.1.
Table 12.1

4-element subsets of {1, 2 , . . . , 6} Smallest member

{1, 2, 3, 4}
{1, 2, 3, 5}
{1, 2, 3, 6}
{1, 2, 4, 5}
{1,2,4,6}
{1, 2, 5, 6}
{1, 3, 4, 5}
{1, 3, 4, 6}
{1, 3, 5, 6}
{1, 4, 5, 6}
{2, 3, 4, 5} 2
{2, 3, 4, 6} 2
{2, 3, 5, 6} 2
{2, 4, 5, 6} 2
{3, 4, 5, 6} 3

By definition,

F(6,4) = (10 • 1 + 4 • 2 + 1 • 3) -r-15

_ 7

and this is equal to j^± when n = 6 and r = 4.

Write N n = { 1 , 2 , . . . , n } . To evaluate F(n,r), it is clear that we
need to first find out
 Pascal's Triangle 91

1. which numbers in N n could be the smallest member of an r-element

subset of N n (in the above example, these are 1, 2, 3 but not 4, 5,
6), and
2. how many times such a smallest member occurs (in the above
example, 1 occurs ten times, 2 four times and 3 once);
and then sum these smallest numbers up, and finally divide by ("), the
number of r-element subsets of N n , to obtain the "average".
The last r elements (according to the magnitude) of the set N n are:
n — r + l , n — r + 2 , . . . , n — r + r ( = n) .

It follows that n — r +1 is the largest possible number to be the smallest

member of an r-element subset of N n . Hence, 1,2,3,... ,n — r + 1 are
all the possible candidates to be the smallest members of r-element
students of N n .
Let k € { 1 , 2 , 3 , . . . , n —r+1}. Our next task is to find out how many
times k occurs as the smallest member. To form an r-element subset
of N n containing k as the smallest member, we simply form an (r — 1)-
element subset from the (n—fc)-element set {ft+1, k+2,... ,n} and then
add k to it. The number of (r —l)-element subsets of {k+1, k+2,... ,n}
is given by ("I*). Thus, k occurs ("!*) times as the smallest member.
Let S denote the sum of all these smallest members. Then, as k =
1,2,..., n — r + 1, we have

s=>(::;) + <;-"0 +3 ("-) + ••••<-—>(- ( ;:r +1) )

- ( .-, + «(;:i) + - + .(;:;) + .(;:j) + i(;:})
) + - + (;: 1 ') + (;:;) + (::i)l
+ + ••• +
n — r + l rows of
+ ••• + summands

+ ) •
92 Counting

Now, by applying (B4) to each summand above except the last one
and noting that (£"}) = (£), E can be simplified to

n\ /n —l\ fr
r r r
) V J \
n—r+1

By applying (B4) once again, we have

n+ 1
r+ 1
Finally, by definition of F(n,r), it follows that

'n\ /n + l\ (n
F(n,r) = E + . . . , . . .

(n + 1)! r\(n-r)\
(r + l)!(n - r)! n!
_ n+ 1
_
r+ 1
as desired. D

Exercise

12.1 Find the coefficient of x5 in the expansion of

(l + x) 5 + (l + x) 6 + --- + (l + x ) 1 0 0 .

12.2 Consider the rows of Pascal's Triangle. Prove that if a row is

made into a single number by using each element as a digit of
the number (carrying over when an element itself has more than
one digit), the number is equal to l l n _ 1 . (For example, from the
first row 1 = 11°, from the second row 11 = l l 1 , from the third
row 121 = l l 2 , and from the 6th row 15(10)(10)51 = 15(11)051 =
161051 = ll 5 .)
12.3 On the rth day of an army recruitment exercise, r men register
themselves. Each day, the recruitment officer chooses exactly
 Pascal's Triangle 93

k of the men and line them up in a row to be marched to the

barracks. Show that the sum of the numbers of all the possible
rows in the first 2k days is equal to the number of possible rows
in the (2k + l)th day.
4 The greatest integer not exceeding a real number x is denoted by
[x\. Show that
(i) ( ? ) < ( J ) i f o < i < i < L f J ;
(ii) (?) > © if Lf J <i<j<n, with equality if and only if
anc
* = LfJjJ = Lf J + ! * n is odd.
5 Evaluaten! + ^ + ^ + ... + ^ + ... + g$.
This page is intentionally left blank
 Chapter 13

Miscellaneous Problems

1 One commercially available ten-button lock may be opened by

depressing — in any order — the correct five buttons. The sam-
ple shown below has {1,2,3,6,9} as its combination. Suppose
that these locks are redesigned 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?

l EZ~3 ^
2

5 CZD CZZI io

(AIME)
2 Calculate in how many ways each of the following choices can
be made.
(i) 4 books are to be chosen from a list of 10 titles to be taken
away for reading during a holiday.
(ii) 20 people have sent in winning entries for a newspaper
competition, and three are to be chosen and placed in order
of merit so as to receive the 1st, 2nd and 3rd prizes.
95
 Counting

(iii) A team of 6 people is to be chosen from a list of 10 possibles;

the team consists of a 1st pair, a 2nd pair and a 3rd pair,
but order within each pair does not matter.
(C)
3 A society is planning a ballot for the office of president. There
are 5 candidates for the office. In order to eliminate the order
of the candidates on the ballot as a possible influence on the
election, there is a rule that on the ballot slips, each candidate
must appear in each position the same number of times as any
other candidate. What is the smallest number of different ballot
slips necessary?
4 In the waiting area of a specialist clinic, patients sit on chairs
arranged 10 to a row with an aisle on either side. Ten patients
are sitting in the second row. How many ways are there for all
the patients in the second row to see the doctor if at least one
patient has to pass over one or more other patients in order to
reach an aisle?
5 In how many ways can 4 a's, 4 6's, 4 c's and 4 d's be arranged in
a 4 x 4 array so that exactly one letter occurs in each row and in
each column? (Such an arrangement is called a Latin square.)
6 A card is drawn from a full pack of 52 playing cards. If the card
is a King, Queen or Jack, two dice are thrown and the total T
is taken to be the sum of the scores on the dice. If any other
card is drawn, only one die is thrown and T is taken to be the
sum of the scores on the card (an Ace is considered as 1) and
the die. Find the number of ways for each of the following:
(i) T < 2;
(ii) T > 13;
(iii) T is odd.
7 In each of the following 5-digit numbers

25225,33333, 70007,11888,...

every digit appears more than once. Find the number of such
5-digit numbers.
 Miscellaneous Problems 97

8 The following list contains some permutations of Ng in which

each of the digits 2, 3, 4 appears in between 1 and 9:

814736259,569324178,793548216,....

Find the number of such permutations of Ng.

9 The following list contains some permutations of Ng in which
each of the digits 1, 2, 3 appears to the right of 9:

458971263,695438172,854796123,....

Find the number of such permutations of Ng.

10 Find the number of 0's at the end of 1 x 2 x 3 x • • • x 2002.
11 Find the number of 15-digit ternary sequences (formed by 0, 1
and 2) in each of the following cases:
(i)there is no restriction;
(ii)there are exactly three 0's;
(iii)there are exactly four 0's and five l's;
(iv) there are at most two 0's;
(v) there is at least one pair of consecutive digits that are the
same;
(vi) there are exactly one "00", three "11", three "22",
three "02", two "21" and two "10" (for instance,
002211102221102).
12 Find the number of (i) positive divisors (ii) even positive divisors
of 2160.
13 Find all natural numbers which are divisible by 30 and have
exactly 30 different divisors.
14 Consider the following grid:

—<>
R
i

I I —
98 Counting

Find in the grid

(i) the number of shortest P-R routes;
(ii) the number of shortest P-Q routes.
13.15 The following figure shows 10 distinct points on the circumfer-
ence of a circle.
(i) How many chords of the circle formed by these points are
there?
(ii) If no three chords are concurrent within the circle, how
many points of intersection of these chords within the circle
are there?

13.16 In a shooting match, eight clay targets are arranged in two hang-
ing columns of three each and one column of two, as pictured.

A marksman is to break all eight targets according to the fol-

lowing rules:
(1) The marksman first chooses a column for which a target is
to be broken.
 Miscellaneous Problems 99

(2) The marksman must 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)
17 Six scientists are working on a secret project. They wish to lock
up the documents in a cabinet such 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?
18 A team for a boxing competition consists of a heavyweight, a
middleweight and a lightweight. There are 5 teams in the com-
petition.
(i) If each person fights with each person of a similar weight
class, how many fights take place?
(ii) At the end of the competition, everyone shakes hands
exactly once with every other person, except his teammates
(they have to tend to each other's wounds later). How many
handshakes take place?
19 Find the number of paths in the array which spell out the word
COUNTING.
con

c c
c o 0 c
c 0 u N u o c
c o u N T N u o c0
c 0 u N T I T N u c
o u N T I N I T N u 0
u N T I N G N I T N u
o U N T I N I T N u o
c o U N T I T N U o c
c 0 U N T N U O c
c 0 U N U O c
c 0 U 0 c
c 0 c
c

20 Let A = {1,2,...,500}. Find

(i) the number of 2-element subsets of A;
 Counting

(ii) the number of 2-element subsets {a, b} of A such that a • b

is a multiple of 3;
(iii) the number of 2-element subsets {a, b} of A such that a + b
is a multiple of 3.
21 There are 7 ways to divide 4 distinct objects a, b, c, d into two
nonempty groups as shown below:

{a, b, c} U {d}, {a, b, d} U {c}, {a, c, d} U {b}, {b, c, d} U {a} ,

{a, b} U {c, d], {a, c} U {b, d}, {a, d} U {b, c} .

How many ways are there to divide n distinct objects, where

n > 2, into two nonempty groups?
22 Two integers p and q, with p > 2 and q > 2, are said to be
coprime if p and g have no common prime factor. Thus 8 and
9 are coprime while 4 and 6 are not.
(i) Find the number of ways to express 360 as a product of
two coprime numbers (the order of these two numbers is
unimportant),
(ii) In general, given an integer n > 2, how do you find the
number of ways to express n as a product of two coprime
numbers where the order is immaterial?
23 The lattice points of the following m x n(n < m) grid are named
as shown:

(0,n) (!,«) (2,n) (m-1, ri) (m, ri)

1 iI ii 1i 1

i ( 1 , 2) (2, 2) (/n-1 , 2 ) i
(0,2) 1 » 1 ii (m, 2)

( 1 , 1) (2, 1) (m-1 1)
(0, D' 1 (m, 1)

» « i < i

(0,0) (1,0) (2,0) (m-1,0) (jfi,0)

 Miscellaneous Problems 101

For k € {1, 2 , . . . , n), let p be the number of shortest (k, k — 1)-

(m, n) routes and q be the number of shortest (k — 1, k)-(m, n)
routes. Show that p(n + 1 — k) = q(m + 1 — k).
24 The face cards (Kings, Queens and Jacks) are removed from
a pack of playing cards. Six cards are drawn one at a time
from this pack of cards such that they are in increasing order of
magnitude. How many ways are there to do this?
25 There are 12 coins on a table. I pick up a number (non-zero) of
coins each time. Find the number of ways of picking up all the
12 coins in the following cases:
(i) I pick up all the 12 coins in an even number of picks,
(ii) I pick up an even number of coins each time.
26 Find the number of 4-tuples of integers
(i) (a, b, c, d) satisfying 1 < a < b < c < d < 30;
(ii) (p, q, r, s) satisfying l < p < g < r < s < 3 0 .
27 Consider the following two 15-digit ternary sequences (formed
by 0, 1 and 2):

000111220011222
012220000111122
Observe that each of the sequences contains exactly three 00,
three 11, three 22, two 01, two 12 and one 20. Find the number
of such ternary sequences.
28 There are n upright cups in a row. At each step, I turn over
n — 1 of them. Show that I can end up with all the cups upside
down if and only if n is even. Find the number of ways this can
be done in a minimum number of steps.
29 The following diagram shows 15 distinct points: toi, W2, u>3,
xi,..., X4, y\,..., y6, z\, Z2 chosen from the sides of rectangle
ABCD.
(i) How many line segments are there joining any two points
each on different sides?
(ii) How many triangles can be formed from these points?
(iii) How many quadrilaterals can be formed from these points?
102 Counting

M>i W2 W3 D

,Xi
Z24
,x2

Z\\ X4

B y* ys y* J3 yi y\

(iv) If no three line segments are concurrent in the interior of

the rectangle, find the number of points of intersection of
these line segments in the interior of rectangle ABCD.
13.30 A ternary sequence is a sequence formed by 0, 1 and 2. Let n be
a positive integer. Find the number of n-digit ternary sequences
(i) which contain at least one "0";
(ii) which contain one "0" and one " 1 " ;
(iii) which contain three 2's.
13.31 Each of the following six configurations consists of 4 vertices
w,x,y,z with some pairs of vertices joined by lines. We are
now given five colours 1, 2, 3, 4, 5 to colour the 4 vertices such
that

(1) each vertex is coloured by one colour and

(iii)

-0
 Miscellaneous Problems 103

(2) any two vertices which are joined by a line must be coloured
by different colours.
How many different ways are there to colour each configuration?
32 If repetitions are not allowed, find the number of different 5-digit
numbers which can be formed from 0 , 1 , 2 , . . . , 9 and are
(i) divisible by 25;
(ii) odd and divisible by 25;
(iii) even and divisible by 25;
(iv) greater than 75000;
(v) less than 75000;
(vi) in the interval [30000, 75000] and divisible by 5.
33 There are 12 boys and 8 girls, including a particular boy B and
two particular girls G\ and G2, in a class. A class debating team
of 4 speakers and a reserve is to be formed for the inter-class
games. Find the number of ways this can be done if the team
is to contain

(i) exactly one girl;

(ii) exactly two girls;
(iii) at least one girl;
(iv) at most two girls;
(v) Gi;
(vi) no B;
(vii) B and G\\
(viii) neither B nor G\\
(ix) exactly one from G\ and Gi\
(x) an odd number of girls.
34 A group of 6 people is to be chosen from 7 couples. Find the
number of ways this can be done if the group is to contain
(i) three couples;
(ii) no couples;
(iii) exactly one couple;
(iv) exactly two couples;
(v) at least one couple.
 Counting

Find the number of ways in which 6 people can be divided into

(i) 3 groups consisting of 3, 2, and 1 persons;
(ii) 3 groups with 2 persons in each group;
(iii) 4 groups consisting of 2, 2, 1 and 1 persons;
(iv) 3 groups with 2 persons in each group, and the groups are
put in 3 distinct rooms.
 Books Recommended for
Further Reading

1. K. P. Bogart, Introductory Combinatorics (3 r d ed.), S. I. Harcourt Brace

College Publishers, 1998.
2. R. A. Brualdi, Introductory Combinatorics (3 r d ed.), Prentice Hall, 1999.
3. C. C. Chen and K. M. Koh, Principles and Techniques in Combinatorics,
World Scientific, 1992.
4. D. I. A. Cohen, Basic Techniques of Combinatorial Theory, John Wiley &
Sons, 1978.
5. R. L. Graham, D. E. Knuth and 0 . Patashnik, Concrete Mathematics (2 n d
ed.), Addison-Wesley, 1994.
6. B. W. Jackson and T. Dmitri, Applied Combinatorics with Problem Solv-
ing, Addison-Wesley, 1990.
7. C. L. Liu, Introduction to Combinatorial Mathematics, McGraw-Hill, 1968.
8. F. Roberts and B. Tesman, Applied Combinatorics (2 n d ed.), Prentice Hall,
2002.
9. A. Tucker, Applied Combinatorics (4 t h ed.), John Wiley & Sons, 2002.

105
This page is intentionally left blank
 Answers to Exercises

1.1 6 1.2 1, 5, 14, 55, 2^r=lr

1.3 6n-4 1.4 29 1.5 27
1.6 60 1.7 31 1.8 29
1.9 14 2.1 90 2.2(i) 6
2.2(ii) 15 2.3(i) 20 2.3(ii) mn
2.3(iii) rant 2.4 30 2.5(i) 3 n

2.5(ii) 2n 2.5(iii) 2"" 1 (l + n) 2.5(iv) n2 + n +

2.6(i) 47 2.6(ii) 205 2.6(iii) 378
3.4 (n + 1)! - 1 4.1(v) 2-9! 4.1(vi) 2-8!
4.1(vii) 9!3! 4.1(viii) 7!5! 4.1 (ix) 2 • 5!6!
4.1(x) 8!9 • 8 • 7 4.1(xi) 5!4!2
4.1(xii) 2 - 4 - 5 - 6 - 7 - 8 - 9 - 1 0 - 11 4.2 (iii) 1008
4.2(iv) 392 4.2(v) 1823 4.3 9-7-5-3
4.4(i) (S) 4.4(H) 5(S) 4.4(iii) 4(2)
4.4(iv) ( e ) + 5 ( s ) + < : ) + (!) 4.5(1) 10! 4.5(ii) 8!3!
4.5 (iii) 7! • 8 • 7 • 6 4.5(iv) 7!3!70 4.6 7-5-3
4.7(i) 5! 4.7(ii) 4!2! 4.7(iii) 2!fi)3!
4.8(i) fflQ 4.8(ii) (?) - (Q + 0 ( S + (Dffl)
4.8 (iii) (?) - (s) 4.9 9-ioi-VJ 4.io(o m
4.10(ii) p i o 4.10(iii) ( ? ) g ) g ) 4.11 162

107
108 Counting

4.12(1) (?) 4.12(H) ft°)g)g) 4.13(i) 405

4.13(H) 90 4.14 g) 4.14(i) g)g)
4.14(11) 2©g) 4.15(a) (J) (J)
4.15(b) g ) g ) 0 + g)G)C), 225.4-3.2
5.1(a)(1) 60 5.1(a)(ii) 36
5.1 (b) By FTA, express n as n = p™1?™2 • • • p™k. Then number of
positive divisors is ni=i( m * + -0-
5.2(1) g)g) 5.2(11) " g)g) 5.2(111) ©(})©
5.2(iv) (»)-fflfflB.6 Qg) 5.7 ffl
6.1 g) 6.2 5! • 6 • 5 • 4 • 3 6.3 2• $
2 55 4 7
6-4 ft ) 7.1(1) (4) 7.1(11) ( 4 )'
7.1(111) ( « ) - ( « ) 7.1(iv) (?)(«) 7.1(v) (247),0
7.2(1) ( ? ) _ ( » ) 7.2(H) (?) 7.3(1) Q g)
7.3(11) g) 3 7.4(1) (333) 7.4(H) ft1) - (?)
7.4(111) (336) 7.5 (20f)
7.6 (?)+(?)+© 7.7 r;-1)
7-8 Gil) 7.9 ff)7.10 g)(J)
7.12 4! 7.13(1) 4!5! 7.13(H) 4!2
7.13(111) 4!25 7.13(iv) 5!5! 7.14(1) 7!
7.14(H) 3!4!7 7.15(1) 4!2! 7.15(H) 4-3-4!
8.1(1) 96 8.1(H) P% 8.3 (i) 84
8.3(H) 7* 8.4 fflg)g)3+g)g)g)3
n l
8.5(1) n! 8.5(H) ( + ) n!
8.5(111) rf)n!+I(nf)(2)^ 9-1 ^
9.2 S! 5!
3!2!' 2!
9.3(i) JL 9.3(H) (yjft1)
9.3 (Hi) ft?)® 9.4(1) nm 9.4(H) P£
9.4(iii) C) 9.4(iv) n™" 1
9.5(i) n!
9.5(H) ("?>! 9.5(iii)
9.6 4io
9.7 n!,n(n + 1)!, n(n + 2)! + Q)(n + 2)!
 Answers to Exercises 109

10.2 25 10.3 k = 3; (1 + x) 14 ; ft), ft), ft)

12-1 (T) 12-5 (Sgyr 13.1 2" - 2 - ft)
2
13.2(1) ft) 13.2(11) P ° 13.2(111) ffiQg)
13.3 5 13.4 10!-29 13.5 576
13.6(1) 16 13.6(11) 40 13.6(iii) 336
13.7 813 13.8 2!3!6-7-8-9
13.9 315-6-7-8-9 13.10 499
15 12
13.11(1) 3 13.11(11) ft)2 13.11(111) ft) ft)
15 14 13
13.11(iv) 2 + 15 • 2 + ft)2 13.11(v) 3 15 - 3 • 214
13.11(vi)120 13.12(i) 40 13.12(H) 32
13.13 2 • 3 • 5 ,2 • 3 • 5 ,2 • 3 • 5 4 ,2 2 • 3 4 • 5,24 • 32 • 5,24 • 3 • 52
2 4 4 2 2

13.14(1) © 13.14(11) Qffl + Offl + fflffl

13.15(1) ft) 13.15(11) ft) 13.16 i
13.17 15, 10 13.18(i) 3g) 13.18(H) ft) - 5(3)
13.19 4-27-4 13.20(i) (50,°)
13.20(11) ft6) + ft6) (334) 13.20(111) (166) + ft6) ft7)
13.21 2n-1-l 13.22(i) 3
13.22(H) By FTA, express n as n = pfx p™2 • • -p™*. Then number
of ways is 2k~1 - 1. 13.24 ft0) 46
13.25(i) 210 13.25(H) 25 13.26(1) ft)
13.26(H) ft) 13.27 64 13.28 (n - 1 ) !
13.29(1) ft) - ((3) + (4) + g) + ( 2 ))
13.29(11) ft) - (g) + Q + ffi)
13.29(111) ft) - (l2@ + llffl + 9g) + (J) + 0 )
13.29(iv)ft^)-(l2(3) + ll( 4 ) + 9© + ( 4 ) + («))
13.30(1) 3 n - 2 n 13.30(H) n(n - 1) 13.30(iii) (!}) 2 n - 3
13.31(1) 5 - 4 3 13.31(H) 5 • 43 13.31 (Hi)5 • 4 2 • 3
110 Counting

13.31(iv) 4 • 13 13.31(v) 5 • 4 • 32 13.31(vi) 5

13.32(i) 7-i6 + 7-7-6-2 13.32 (ii) 7-
13.32(iii) 7- 6 13.32(iv) 4 - 8 - 7 - 6 + 2 - 9 - 8 - 7 - I
13.32(v) 9- 8 • 7 • 6 - (4 • 8 7 - 6 + 2 - 9 - 8 - 7 - 6 )
13.32(vi) 8- 7 . 6 + 9 - 7 - 6
13.33(i) (?)© + (?) ft1)©
13.33(ii) ft^fflffl + Cflft1)®
13.33(iii) (DftVfiK)
13.33(iv) (\2)ft1)+ft2)® +ft2)ft1)® +ft2)® (I) +ft2)ft1)®
13.33(v) ft9)ft6)+ ft9) 13.8S(vi) (»)(?)
13.33(vii) 2ft8) + 8
ft ) 7
ft ) 13.33(viii) (»)ff)
13.33(ix)
2(ft8)ft5) + ft8))
13.33(x) ft2) (!) +ft2)ft1)(!) +ft2)® (?) +ft2)ft1)© +ft2)(!)
13.34(i) © 13.34(ii) (J) 2* 13.34(iii) (J)(J)2*
13.34(iv) Q®2»13.S4(v) (?) - Q*» 13.35(i) g) Q
13.35(H) mm
3!
13.35(iii) WE) 13.35(iv) ®g)
2!2!
 Index

Addition Principle, 1, 2, 11 Injection, 37

Multiplication, 1, 9-11
bijection, 35, 37, 39, 41, 42, 44-46, 48, of Complementation, 30
53, 55, 59, 65, 77, 79 Principle of Complementation (CP), 30
Bijection Principle (BP), 37
binomial coefficient, 85-87 sequence, 9, 12, 16, 19, 28, 40, 42, 48,
Binomial Theorem, 76, 85 61, 102
binary, 12, 28, 40, 42, 48
combination, 23, 95 ternary, 16, 102
r-, 23 surjection, 37

distribution problem, 48, 49, 63, 67, 69

Fundamental Theorem of
Arithmetic, 43

injection, 37
Injection Principle, 37

mapping, 35-38, 40, 44, 45

one-to-one, 37
Multiplication Principle, 1, 9-11

Pascal's Triangle, 85-87, 89, 92

permutation, 20, 21, 23, 27
r-, 20, 21
power set, 41
Principle
Addition, 1, 2, 11
Bijection, 35, 37

111
COUNTING
This book is a useful, attractive introduction to basic
counting techniques for upper secondary and junior
college students, as well as teachers. Younger students
and lay people who appreciate mathematics, not to
mention avid puzzle solvers, will also find the book
interesting. The various problems and applications
here are good for building up proficiency in counting.
They are also useful for honing basic skills and
techniques in general problem solving. Many of the
problems avoid routine and the diligent reader will
often discover more than one way of solving a
particular problem, which is indeed an importan
awareness in problem solving. The book thus help;
to give students an early start to learning proble
solving heuristics and thinking skills.

www. worldscientific. com

5018 he