section 2.3 counting - jerryandsusan.tripod.comjerryandsusan.tripod.com/sec23.pdf · ... eight...
TRANSCRIPT
Section 2.3 183 Combinatorics
183
Section 2.Section 2.Section 2.Section 2.3333 Combinatorics Combinatorics Combinatorics Combinatorics
Purpose of SectionPurpose of SectionPurpose of SectionPurpose of Section To introduce some basic tools of counting, such as the
multiplication principle, permutations and combinations.
Introduction Introduction Introduction Introduction
Someone once said every mathematician is part combinatorist since
combinatorial ideas pervade every area of mathematics, pure and applied.
Today, the interest in combinatorial analysis is fueled by important problems
in chemistry, biology, physics and computer science. Problems in chemistry
arise in studying the arrangement of atoms in molecules and crystals, in
physics in the study of statistical and quantum mechanics, in biology in
studying gene structure, and in computer science in a myriad of problems
ranging from computer networks to microcircuit design.
Combinatorics is a branch of mathematics that deals with concrete, easy
to understand problems and often requires no technical knowledge. However,
simply because a problem is easy to understand, that doesn’t imply it is easy
to solve. For example, given an ordinary 8 8× checkerboard, how many ways
can you cover it with dominoes (each domino covering 2 adjacent squares)?
Although the problem is easy to understand, it is not easy to solve. The total
number of different “coverings” was not solved until 1961 by M.E. Fischer,
who found the number to be ( )24
2 901 12,988,816⋅ = .
Although it is difficult to define such a multifaceted subject as
combinatoricscombinatoricscombinatoricscombinatorics, it is simply the subject of counting. Counting is so
commonplace, we generally don’t give it much thought, yet as the reader will
see, the techniques of counting are as ingenious as any in abstract
mathematics, and just as difficult. More than one good mathematician has
been embarrassed by a seemingly simply “counting” problem.
Multiplication PrincipleMultiplication PrincipleMultiplication PrincipleMultiplication Principle
One of the most basic principles of counting, if not the most basic, is the
multiplication principlemultiplication principlemultiplication principlemultiplication principle. Although the principle is simple, it has far reaching
consequences. For example, how many dots are there in the array of dots in
Figure 1?
Section 2.3 184 Combinatorics
184
Count the Dots
Figure 1
Your answer is 30 and we suspect you arrived at this by counting 10 dots in
the first row and then multiplied by 3. If this is true, then you used the
multiplication principle. As a small step up in complexity chain, try counting
the number of different paths from A to C in Figure 2. No doubt this problem
didn’t stump you either, getting 4 5 20⋅ = paths. Again, you used the
multiplication principle.
How Many Paths from A to C?
Figure 2
This leads us to the formal statement of the multiplication principle.
Multiplication RuleMultiplication RuleMultiplication RuleMultiplication Rule: If a procedure can be broken into successive stages, and if
there are 1s outcomes for the first stage, 2s outcomes for the second stage, …,
and n
s for the n th stage, then the entire procedure has 1 2...,n
s s s outcomes. .
Example Example Example Example 1 1 1 1 ((((CountingCountingCountingCounting SubsetsSubsetsSubsetsSubsets)))) Show that that a set { }1 2, ,...,n
A a a a=
containing n elements has 2n subsets.
Proof:Proof:Proof:Proof: A subset of A can be formed in n successive steps. On the first step,
one can pick or not pick 1a , on the second step one can pick or not pick 2a , …
and so on. On each step there are two options, to pick or not to pick. Using
the multiplication principle, the number of subsets that can be selected is
Section 2.3 185 Combinatorics
185
2 2 2 2n⋅ ⋅ ⋅⋅ =
Example 2: Example 2: Example 2: Example 2: ((((Counting FunctionsCounting FunctionsCounting FunctionsCounting Functions1111)))) How many functions are there from the set
{ }, ,A a b c= to the set of binary numbers { }0,1B = ?
Solution Solution Solution Solution For each of the 3 values in A the function can take on 2 values.
Hence, by the multiplication rule the number of functions is 32 2 2 2 8⋅ ⋅ = = . We
leave it to the reader to draw graphs of the eight functions. In general the
number of functions from a set with cardinality n to a set with cardinality m
is nm .
Although the multiplication principle is very simple, it has far reaching
results. One is the study of permutations.
PermutationsPermutationsPermutationsPermutations
A permupermupermupermutationtationtationtation is simply an arrangementarrangementarrangementarrangement of objects. For example the
permutations of the three letters abc are the six arrangements
, , , , ,abc acb bac bca cab cba
The number of permutations increases dramatically as the size of the set
increases. The number of permutations of the first 10 letters of the alphabet
abcdefghij is 3,628,800 . We certainly didn’t arrive at that number by listing each
arrangement. We simply argue that we can select the left-most member in our
arrangement 10 ways, and once this is done there are 9 choices for the second
member in the arrangement, the third member 8 ways, … and so down to the last
member. Then, using the multiplication principle, we find the number of
permutations or arrangements of 10 objects to be 10 factorial, or
10 10 9 8 7 6 5 4 3 2 1 3 628 800! , ,= ⋅ ⋅ ⋅ ⋅ ⋅ ⋅ ⋅ ⋅ ⋅ = .
Now, suppose four individuals , , ,a b c d are in a foot race and we wish to
determine the possible ways the runners can finish first and second. Each of the
four runners can finish first, and for each winner, there are 3 second-place
finishers. Hence by the multiplication principle there are 4 3 12⋅ = possible ways
the runners can finish first and second, which are.
1 We will talk more about functions in Chapter 4.
Section 2.3 186 Combinatorics
186
Permutations of Two Elements from a Set of Four Elements
Figure 3
This leads to the general definition of permutations of different sizes.
Big NumberBig NumberBig NumberBig Number:::: Factorials grow very fast. The number 50! is thirty vigintillion, four hundred and four novemdecillion, ninety three octodecillion, two hundred and one septendecillion, seven hundred and thirteen sexdecillion, three hundred and seventy eight quindecillion, forty three quatuordecillion, six hundred and twelve tredecillion, six hundred and eight dodecillion, one hundred and sixty six undecillion, sixty four decillion, seven hundred and sixty eight nonillion, eight hundred and forty four octillion, three hundred and seventy seven septillion, six hundred and forty one sextillion, five hundred and sixty eight quintillion, nine hundred and sixty quadrillion, five hundred and twelve trillion. That is
50 30414093201713378043612608166064768844377641568960512000000000000!=
Definition: Definition: Definition: Definition: A permutation permutation permutation permutation of r elements taken from a set of n elements is
simply an arrangement of r elements chosen from a set of n elements. The
number of such arrangements (or permutations) is denoted by ( ),P n r .
Using the multiplication principle, we find the number of permutations.
Theorem 1 Number of Theorem 1 Number of Theorem 1 Number of Theorem 1 Number of Permutations Permutations Permutations Permutations
The number of permutations of r elements taken from a set of size n is
Section 2.3 187 Combinatorics
187
( )( )
( ) ( ) ( )!
, 1 2 1!
nP n r n n n n r
n r= = ⋅ − ⋅ − − +
−�
Proof:Proof:Proof:Proof:
Choosing r elements from a set of size n , we have
▶ the first element can be selected n ways.
▶ the second element can be selected 1n − ways (since now
There are 1n − left).
▶ the third element can be selected 2n − ways (since
now there are 2n − left).
… … … … … … … … … … … …
▶ the thr element can be selected 1n r− + ways
Hence, by the principle of sequential counting (or the multiplication
rule), we have
( ) ( ) ( ) ( ) factors
, 1 2 1
r
P n r n n n n r= ⋅ − ⋅ − − +��������������
.
▶ When r n= we have the number of permutations of n objects to be
n factorial, or
( ) ( )( ) ( )( )1 2 2 1, !P n n n n n n= − − =�
Example Example Example Example 3333 Permutations Permutations Permutations Permutations of of of of a Set a Set a Set a Set with 3 Elements with 3 Elements with 3 Elements with 3 Elements
Find the permutations of 1, 2 and 3r = elements selected from { }, ,a b c .
SolutionSolutionSolutionSolution
The underlying set { }, ,a b c has 3n = elements. The permutations of 1,
2 and 3 elements from { }, ,a b c are listed in Table 1.
1r = 2r = 3r =
a ab abc
b ac acb
c ba bac
bc bca
Section 2.3 188 Combinatorics
188
ca cab
cb cba
Permutations ( ),P n r
Table 1
We don’t use set notation for writing permutations2 since order is important.
The permutation ab is not the same as ba .
Margin Note:Margin Note:Margin Note:Margin Note: Permutations: Order matters.
Margin Note:Margin Note:Margin Note:Margin Note: To evaluate ( ),P n r start at n and multiply r factors.
( )
( )
( )
( )
( )
4, 2 4 3 12
7,3 7 6 5 210
4,1 4 4
10,3 10 9 8 720
4, 4 4 3 2 1 24
P
P
P
P
P
= ⋅ =
= ⋅ ⋅ =
= =
= ⋅ ⋅ =
= ⋅ ⋅ ⋅ =
Example Example Example Example 4444
How many ways can one arrange the seven letters of the word SYSTEM?
Solution Solution Solution Solution
The two S’s in SYSTEM are indistinguishable3 so we must modify our
argument. If we momentarily call the two S’s as 1S and 2S , then we have
1 2S YS TEM , which has 6! permutations. But the two letters 1S and 2S can be
arranged 2! 2= ways, so we must divide 6! by 2!, getting the number of
distinguishable (permutations you can recognize with the naked eye)
permutations of SYSTEM to be
( )6!
6, 4 6 5 4 3 3602!
P = = ⋅ ⋅ ⋅ =
A few more distinguishable permutations are given in the following table.
2 Sometimes permutations are written with round parenthesis, such as ( )ab ab= . . 3 If we interchanged the two S’s we would get two different permutations, SYSTEM and SYSTEM, but they are
not distinguishable permutations since the S’s look alike.
Section 2.3 189 Combinatorics
189
Word Indistinguishable Permutations
too 3
32
!
!=
error 5
203
!
!=
toot 4
62 2
!
! !=
Mississippi 11
69 3004 4
!,
! !=
Distinguishable Permutations
Table 2
A second application of the multiplication principle is computing
combinations.
Combinations
Combinations are essentially permutations where order doesn’t matter.
For example, for a set with three members { }, ,c a t , a combination of two
members of this set is a subset of size two. There are three such subsets, or
combinations, which are
{ } { } { }, , , , ,a c a t c t
Note that we write these three combinations in set notation { },x y since the
order of the members in the combination does not matter. In other words, the
combination { },a c is the same as the combination { },c a . This is different from
permutations where the permutation ab is not the same as the permutation ba .
Definition: Definition: Definition: Definition: A combinationcombinationcombinationcombination is a subset of elements of a set.
Example Example Example Example 5555 Find the combinations or subsets of size 1, 2 and 3r = taken from
the set { }, ,a b c .
SolutionSolutionSolutionSolution
It is a simple task to enumerate these combinations, which are listed
Table 3. Note that the 8 subsets of { }, ,a b c with the exception of the empty
Section 2.3 190 Combinatorics
190
set ∅ are listed. If we wanted a subset of size 0 we would include the empty
set.
1r = 2r = 3r =
{ }a { },a b { }, ,a b c
{ }b { },a c
{ }c { },b c
Nonempty Subsets of { }, ,a b c
Table 3
TheTheTheTheorem 2:orem 2:orem 2:orem 2: Number of Combinations Number of Combinations Number of Combinations Number of Combinations The number of combinationscombinationscombinationscombinations (or
subsets) of size r which can be selected from a set of size n , which we
denote by ( ), or n
rC n r
, is
( )( )
!,
! !
n nC n r
r r n r
= =
− ....
Proof:Proof:Proof:Proof:
Recall that the number of permutations of size r elements taken from a
set of size n is ( )( )
!,
!
nP n r
n r=
−. However, these r elements can be
permuted in !r ways, so to count only one of these !r permutations, we
divide ( , )P n r by !r getting
( )( )
( ), !
,! ! !
P n r nC n r
r r n r= =
−.
The numbers ( ), or n
C n rr
are called binomial coefficientsbinomial coefficientsbinomial coefficientsbinomial coefficients because they are
the coefficients in the binomial expansion
Section 2.3 191 Combinatorics
191
( )
( )
( )
2 2 2 2 2
3 3 2 2 3 3 2 2 3
4 4 3 2 2 3 4 4 3 2 2 3 4
2 2 22
0 1 2
3 3 3 33 3
0 1 2 3
4 4 4 4 44 6 4
0 1 2 3 4
..
a b a ab b a ab b
a b a a b ab b a a b ab b
a b a a b a b ab b a a b a b ab b
+ = + + = + +
+ = + + + = + + +
+ = + + + + = + + + +
. ... ... ... ... ...
.
It helps in thinking about combinations to say n
r
as "n choose "r since it
denotes the number of ways one can choose r items from a set of n items.
For example 4
62
=
is read as “4 choose 2 is 6” which means there are 6
ways to choose 2 things from 4 things.
Margin Note: Margin Note: Margin Note: Margin Note: Combinations: Order does not matter
Example Example Example Example 6666 How many ways can 10 players choose up sides to play five-
on-five in a game of basketball?
SolutionSolutionSolutionSolution
As in many combinatorial problems, there is more than one way to do the
counting. Perhaps the simplest way here is to determine the number of ways
one player can choose his or her four teammates from the 9 other players. In
other words, the number of subsets of size four taken from a set of size 9, or
“nine choose four”, which is
9 9! 9 8 7 6
126 ways 4 4!5! 4 3 2 1
⋅ ⋅ ⋅= = =
⋅ ⋅ ⋅ .
Another approach is to find number of subsets of size 5 taken from a set of
size 10 and divide that number by 2, getting
101 1 10! 1 10 9 8 7 6
12652 2 5!5! 2 5 4 3 2 1
⋅ ⋅ ⋅ ⋅= ⋅ = ⋅ =
⋅ ⋅ ⋅ ⋅
Section 2.3 192 Combinatorics
192
Example 7 (Number of World Series)Example 7 (Number of World Series)Example 7 (Number of World Series)Example 7 (Number of World Series) How many possible ways can two
teams play a best-of-seven game World Series4? For example, two outcomes
are WWWW (sweep) and WLWWLW which means the winner wins games
1,34,6 in a 6-game series.
Solution:Solution:Solution:Solution: The series might run 4, 5, 6, or 7 games. However, since the winner
eventually wins 4 games, we simply ask how many ways can we place 4 W’s in
7 slots:
1 2 3 4 5 6 7
For example, a 4-game sweep would correspond to
WWWW WWWW WWWW WWWW 5 6 7
and a 6-game series in which the winner wins games 1,3,4, and 6 would
correspond to
WWWW 2 WWWW WWWW 5 WWWW 7
Hence, the total number of series that can be played is 7 choose 4, or
( )7 4 35,C = .
In general, the number of series that can be played in a best of 2 1n +
game series is ( )2 1 1,C n n+ + . We could analyze the series further by asking
the possible number of 4,5,6, and 7 game series. These would be
( )( )( )( )
Number of 4-game series 3 3 1
Number of 5-game series 4 3 4
Number of 6-game series 5 3 10
Number of 7-game series 6 3 20
,
,
,
,
C
C
C
C
= =
= =
= =
= =
The idea behind these numbers is that the winner must win 4 games and the
last game of the series is always won by the winner. Hence, we simply ask
the number of ways the teams can play the series before the last game. For
example, in a 6-game series, how many possible series games will be played
before the last game? The answer is the number of ways you can put 3 wins
in 5 games, or “5 choose 3” which is ( )5 3 10,C = .
4 We assume the reader knows the World Series is a best-of-7 game series.
Section 2.3 193 Combinatorics
193
It is interesting to note that of all the World Series between 1903 and 2009,
eleven series have never occurred, one being WLLLWWW when the winner wins
games 1,5,6, and 7. The most common outcome is the four-game sweep
WWWW, which has occurred 15 times
PPPPoker Handsoker Handsoker Handsoker Hands
In poker, five cards are dealt from a deck of 52 cards. The types of
poker hands (and examples) are
1111 Royal Flush:Royal Flush:Royal Flush:Royal Flush: A, K, Q, J, 10 of the same suit (10♠, J♠, Q♠,K♠,A♠ )
2 Straight Flush: Five cards of same suit in sequence (3♣,4♣,5♣,6♣,7♣ )
3 4 of a Kind: Four cards of the same rank (7♦,7♠,7♣,7♥)
4 Full House: Three of a kind plus a pair (7♠,7♣,7♥,9♦,9♥)
5 Flush: Five cards of the same suit (3♥,6♥,K♥,9♥,A♥ )
6 Straight: Five cards in sequence (4♥,5♦,6♠,7♥,8♠ )
7777 3 of a Kind:3 of a Kind:3 of a Kind:3 of a Kind: Three cards of the same rank (J♥, J♠, J♦)
8888 2 Pairs:2 Pairs:2 Pairs:2 Pairs: Two pairs of different rank (5♣, 5♥, 9♦, 9♠)
9 1 Pair: Two cards of the same rank (A♣, A♥)
Find a) the total number of poker hands, b) the number of 4-or-a-kinds, c) the
number of full houses, d) the number of 3-of-a-kind hands.
Solution:
a) Total Hands:Total Hands:Total Hands:Total Hands: Each hand represents a subset of size 5 from a set of size 52.
Hence the total number of hands a player can receive is “52 choose 5” or
52 52! 52 51 50 49 48
2,598,960 total hands5 5!47! 5 4 3 2
⋅ ⋅ ⋅ ⋅= = =
⋅ ⋅ ⋅
b) 4444----ofofofof----aaaa----kindkindkindkind:::: There are 13
131
=
ways to select the “quads” and
4848
1
=
choices for the remaining card. Using the multiplication principle, we get
Section 2.3 194 Combinatorics
194
48 13
48 13 6244 four of a kind hands1 1
= ⋅ =
c) Full House:Full House:Full House:Full House: There are 13
131
=
choices for the rank of the triple, and
1212
1
=
choices for the rank of the pair. There are also
44
3
=
ways to
choose the triple from a card of a given rank, and 4
62
=
ways to choose the
pair from four cards of the other rank. Hence, the multiplication principle gives
us
13 12 4 4
13 12 4 6 3,744 full house hands1 1 3 2
= ⋅ ⋅ ⋅ =
d) Three of a Kind:Three of a Kind:Three of a Kind:Three of a Kind: There are 13
131
=
choices for the rank of the triple, and
1266
2
=
choices for the rank of the remaining 2 cards. There are also
44
3
=
choices for the triple of the given rank, and
44
1
=
for each of the
remaining two cards. Using the multiplication principle, we have
313 12 4
54,912 3-of-a-kind hands1 2 1
=
Naming Naming Naming Naming Results Results Results Results in Mathematicsin Mathematicsin Mathematicsin Mathematics It is a sorry state of affairs that in mathematics,
the most precise of all disciplines, is so lax about giving credit to those who
made major discoveries. Pascal’s triangle was well known by many
mathematicians centuries before Pascal described it in a paper, so why is it
called “Pascal’s triangle?” It probably goes back to some person giving credit
to Mr. Pascal, and others picking up on that. After a while it becomes “Pascal’s
triangle.” Mathematics is rife with all sorts of equations and theorems
attributed to one person that were actually discovered by another.
Example Example Example Example 8888 (Going to the Movies) (Going to the Movies) (Going to the Movies) (Going to the Movies)
Three boys and two girls are going to a movie. How many ways can
they sit next to each other under the following conditions?
Section 2.3 195 Combinatorics
195
a) Neither boy sits next to each other.
b) The two girls sit next to each other.
SolutionSolutionSolutionSolution
a) The only way they can sit is boy-girl-boy-girl-boy. But the boys can
be permuted 3! 6= ways and the girls 2! 2= ways, so the total number of
arrangements is 3!2! 12= .
b) First think of the two girls as “one girl” so there are 4 persons; 3 boys
and 1 girl. But there are 4! 24= ways to permute the “girl” among the 3 boys.
But, for each of these arrangements, we can permute the two girls 2! 2=
ways, and so the total number of arrangements is 4!2! 48= .
Example Example Example Example 9999 (Marbles and Bins) (Marbles and Bins) (Marbles and Bins) (Marbles and Bins)
John is handing out 10 pieces of corn candy to 4 children in such a way
that every child gets at least one piece. How many different ways can John
distribute the candy?
SolutionSolutionSolutionSolution
Many combinatorial problems seem difficult until viewed from a certain
perspective. This is one of those problems. The idea here is to line up the 10
pieces of candy in a row like
And separate the candy with three bars like
which is our way of saying John gives the first child 2 pieces of candy, the
second child 3 pieces, the third child 4 pieces, and the last child 1 piece.
Another way to think of it is there are 9 gaps between the candy and we put
the bars in gaps 2,5, and 9 which is the same as picking the combination
Section 2.3 196 Combinatorics
196
{ }2 5 9, , from the set { }1 2 3 4 5 6 7 8 9, , , , , , , , . Hence, the number of ways to put 3
bars between 9 pieces of candy is the binomial coefficient
9 9
843 6 3
!
! !
= =
which is the number of ways John can distribute the 10 pieces of candy to four
children.
In general, the number of ways to distribute n pieces of candy to k children
where every child gets at least one piece is ( )1 1,C n k− − .
Example Example Example Example 10101010 (Finding Paths) (Finding Paths) (Finding Paths) (Finding Paths)
How many paths are there from Start to End in the road system in Figure
4, always moving to the right and down?
SolutionSolutionSolutionSolution
Counting Paths
Figure 4
SolutionSolutionSolutionSolution Since all paths pass through the “one-point” gap, the problem is
subdivided into two parts; finding the paths from Start to the gap, then finding
the paths from the gap to End, then multiplying the results together. From
Start to the gap, note that we travel a total of eight “blocks,” four blocks to
the right and four blocks down. Labeling each block as R or D , depending
whether the move to the right or down, all paths can be written
{ }, , , , , , ,x x x x x x x x , where four of the 'x s are R and four are D . Hence, the
total number of paths is the number of ways you can select 4 'D s (or 4 'R s )
from a set of size 8, which is “8 choose 4” or ( )8, 4 70C = . Similarly, the
Section 2.3 197 Combinatorics
197
number of paths from the “gap” to End is ( )7,3 35C = Hence, the total number
is
8 7
70 35 2, 450 paths.4 3
= ⋅ =
.
CCCCombinatorial Combinatorial Combinatorial Combinatorial Card Trickard Trickard Trickard Trick5555♠♠♠♠
Here is a little trick you and a friend can perform before an audience.
It’s very clever and involves two principles of combinatorics, the pigeonhole
principle6 and the fact that a set with three elements has 6 permutations. You
can be the magician and your friend the assistant. Beginning with a deck of 52
playing cards, choose a volunteer from the audience to select five cards at
random from the deck and gives them to the assistant. The assistant looks at
the cards, places one card face down on a table, and the other four face up.
You then perform your magic by studying the four upright cards, waving your
magic wand, and announcing the 5th card to the audience. How did you do it?
To show how the trick is done, someone the person in audience has
selected the cards shown in Figure 5.
Selected Cards
Figure 5
Now, since there are four suits in the deck and since five cards are selected,
the Pigeonhole Principle requires at least one suit will appear more than once.
After your assistant receives the cards, he (or she) focuses on the repeated
5 This trick has been attributed to mathematician William Fitch Cheney. For more variations on this
trick, google “Cheney’s Five Card Trick.” 6 The pigeonhole principlepigeonhole principlepigeonhole principlepigeonhole principle is a rather obvious principle from combinatorics which states if
m objects are attempted to be placed in n boxes and if m n> , then at least one box contains
more than one object. The principle is sometimes stated that if m pigeons try to nest in n
pigeonholes and if m n> , then at least one pigeonhole will contain more than one pigeon.
Section 2.3 198 Combinatorics
198
suit7 (hearts in our example) and computes the smaller clockwise distance
between the 7♥ and 9♥, as illustrated in Figure 6, which is 2. Note that the
minimum distance will always 1,2,3,4,5 or 6. The assistant then selects the
larger 9♥ as the “mystery card” and places it face down on a table. The
remaining four cards {7♥, 7♦, 9♣, 2♠} are placed upright on the table.
Counting clockwise no two cards are more than 6 places apart
Figure 6
But the cards are not placed randomly but in a specific order, and that is what
is crucial. Your assistant places the cards in the following order (from left to
right) as
order of the cards from left to right = (7♥ , 2♠ , 7♦ , 9♣)
So how does this ordering tell you that the mystery card is a 9 of hearts? The
first card on the left tells you that the mystery card is a heart. The remaining
three cards (2♠ , 7♦ , 9♣) give sufficient information to tell you that the
mystery card is two places past the 7♥ which is the 9♥. The idea behind the
code is to rank all 52 cards in the deck according to the ranking
spaces clubs < diamonds hearts< <
and within each suit
2 3A J Q K< < < < < <� .
7 If there is more than one repeated suit, pick one at random.
Section 2.3 199 Combinatorics
199
so the three cards can be ranked according to small (S), medium (M), and
large (L). In this way, your assistant can encode one of the numbers 1,2,3,4,5
or 6 by placing the cards in one of the 6 permutations
1 2
3 4
5 6
SML SLM
MSL MLS
LSM LMS
= =
= =
= =
In our example, your assistant has laid out the three cards as 2♠ 7♦ 9♣
whose ranking is SLM whose code is 2. So after viewing these cards you
perform some quick mental calculations, and decode the mystery cards and
move 2 places past the 7 and arrive at the 9♥, whereupon you wave your
magic wand and announce the card to the amazed audience8.
8 The audience will be amazed. The author has performed this trick. The audience will also be
suspicious, thinking your assistant passed the identity of the card to you in some surreptitious
manner.
Section 2.3 200 Combinatorics
200
SeSeSeSection 2.3, Combinatoricsction 2.3, Combinatoricsction 2.3, Combinatoricsction 2.3, Combinatorics
1. Compute the following.
a) (5,3)P Ans:Ans:Ans:Ans: 5 4 3 60⋅ ⋅ =
b) (4,1)P Ans:Ans:Ans:Ans: 4
c) (30, 2)P Ans:Ans:Ans:Ans: 30 29 870⋅ =
d) (4,1)C Ans:Ans:Ans:Ans: 4
41 3
!
! !=
e) (10,8)C Ans:Ans:Ans:Ans: 10
458 2
!
! !=
f) 7
2
Ans:Ans:Ans:Ans: 7
212 5
!
! !=
g) 9
2
Ans:Ans:Ans:Ans: 9
362 7
!
! !=
h) ( )6
a b+
Ans:Ans:Ans:Ans: 6 5 4 2 3 3 2 4 5 66 15 20 15 6a a b a b a b a b ab b+ + + + + +
Permutations and CombinaPermutations and CombinaPermutations and CombinaPermutations and Combinationstionstionstions
2. (Distinguishable Permutations)(Distinguishable Permutations)(Distinguishable Permutations)(Distinguishable Permutations) The number of distinguishable permuations distinguishable permuations distinguishable permuations distinguishable permuations
of the word TOOTS is 5!
2!2!. Since there are five letters in the word one
writes 5! in the numerator. However, we cannot distinguish the 2 Ts and the 2
Os in the word, hence we divide 2!2!. Find the number of distinguishable
permutations in the following words.
a) TO Ans:Ans:Ans:Ans: 2
21 1
!
! !=
b) TWO Ans:Ans:Ans:Ans: 3
31 1
!
! !=
c) TOO Ans:Ans:Ans:Ans: 3
32 1
!
! !=
Section 2.3 201 Combinatorics
201
d) BOOT Ans:Ans:Ans:Ans: 4
62 2
!
! !=
e) SNOOT Ans:Ans:Ans:Ans: 5
601 1 2 1
!
! ! ! !=
f) DALLAS Ans:Ans:Ans:Ans: 6
1201 2 2 1
!
! ! ! !=
g) TENNESSEE Ans:Ans:Ans:Ans: 9
37801 4 2 2
!
! ! ! !=
h) ILLINOIS Ans:Ans:Ans:Ans: 8
33603 2 1 1
!
! ! ! !=
3. (Goi(Goi(Goi(Going to the Moviesng to the Moviesng to the Moviesng to the Movies)))) Find the numbers of ways in which 4 boys and 4
girls can be seated in a row of 8 seats if they sit alternately and if there is a
boy named Jerry and a girl named Susan who cannot sit next to each other.
Ans: Ans: Ans: Ans: Calling B = boy, G = girl we first count the arrangements of alternating
boy and girl. We can arrange as
B,G,B,G,B,G,B,G 4 4 576 ways
G,B,G,B,G,B,G,B 4 4 576 ways
! !
! !
= × =
= × =
for a total of 576 + 576 = 1152 ways. We now must subtract from this
number the number of ways where Jerry and Susan are sitting next to each
other. The number of ways will be
Jerry, Susan, B,G,B,G,B,G 6 3 3 252 ways
Susan, Jerry, G,B,G,B,G,B 6 3 3 252 ways
! !
! !
= × × =
= × × =
For a total of 252 + 252 = 504 ways. To satisfy the condition that Jerry and
Susan do not sit next to each other we must subtract 504. Hence the total
number of desired arrangements is 1152 – 504 = 648 ways.
4. (Movies Again) (Movies Again) (Movies Again) (Movies Again) Four couples go to the movies and sit together in six seats.
How many ways can these people arrange themselves where each couple sits
next to each other?
Section 2.3 202 Combinatorics
202
Ans:Ans:Ans:Ans: Four couples can arrangement themselves 4! Number of ways, where
each couple can sit two ways. Hence, the total number of ways is
4 2 2 2 192 ways!× × × =
5. (More Movies) (More Movies) (More Movies) (More Movies) Mary and 4 friends go to a movie. How many ways can
they sit next to each other with Mary always between two friends?
Ans:Ans:Ans:Ans: The total number of ways they can sit together is
5 5 4 3 2 1 120!= × × × × = . The total number of ways they can sit together with
Mary on the left is 4!, which is the same as if Mary sat on the right. Hence,
the total number with Mary between two friends is
5 4 4 120 24 24 72 ways! ! !− − = − − =
6. (Baseball Season)(Baseball Season)(Baseball Season)(Baseball Season) A baseball league consists of 6 teams. How many
games will be played over the course of a year if each team plays every other
team exactly 5 times?
Ans:Ans:Ans:Ans: 6 6
5 5 752 2 4
!
! !
= =
7777. . . . (Disti(Disti(Disti(Distinct 3nct 3nct 3nct 3----digit numbers)digit numbers)digit numbers)digit numbers) How many 3-digit numbers are there in which
the number of even digits is 0,2,4,6, or 8 ?
AnsAnsAnsAns: Calling even digits by E and odd digits by O, we can have any 3-digit
numbers of the form EEO, EOE, OEE, and OOO. Now there are 5 even digits
0,2,4,6, 8 and there are 5 odd digits 1,3,5,7,9. However, the number 0 cannot
start the 3-digit number, else the number is a 2-digit number. Hence, using the
multiplication principle, we have
EEO: 4 5 5 100 3-digit numbers of this form
EOE: 4 5 5 100 3-digit numbers of this form
OEE: 5 5 5 125 3-digit numbers of this form
OOO: 5 5 5 125 3-digit numbers of this form
× × =
× × =
× × =
× × =
for a total of 450 possibilities.
8. (Interesting Problem)(Interesting Problem)(Interesting Problem)(Interesting Problem) How many 3-digit numbers 1 2 3
d d d are there whose
digits add up to 8? That is, 1 2 3
8d d d+ + − , Note: 063 is not a 3-digit number,
it is the 2-digit number 63.
Section 2.3 203 Combinatorics
203
Ans: Ans: Ans: Ans: Starting with a 3-digit number 1 2 3
d d d , we start with the hundreds digit
1d , which can have values 1 2 3 4 5 6 7 or 8, , , , , ,k = for a total of 8 possibilities (0
is not a possibility for the hundreds digit else1 2 3
d d d will only be a 2-digit
number. If we now pick the hundreds digit to be k , then the number of
possibilities for the 10-digit 2
d will be 8 1k− + (we need the +1 since now 0
is a possibility). The 1-digit 3
d is now determined and so if we add up
these possibilities of the hundred and ten digit we find
( )8 8 8
1 1 1
8 1 9
8 99 8
272 36
36 ways
k k k
k k= = =
− + = −
⋅= ⋅ −
= −
=
∑ ∑ ∑
9. (World Series Time(World Series Time(World Series Time(World Series Time) Two teams are playing in playoff series. If it is a
best of 3-game series there are three possible outcomes, WW, WLW, LWW if
we do not distinguish the teams. If the teams are distinguished, say team A
and team B, then there would be six outcomes in a 3-game series, AA, ABA,
BAA, BB, BAB, ABB.
a) How many possible games are there is a 5-game series if we do not
distinguish the teams?
b) How many possible games are there is a 2 1n + game series if we do not distinguish the teams?
Ans:Ans:Ans:Ans: a) If we denote the possible games as
1 2 3 4 5
the three W’s can be placed in any 3 slots. Hence the number of possible
games is
5
103
=
b) The W’s can be placed in any 1n + of the 2 1n + possible slots. Hence, the
number of possible games is 2 1
1
n
n
+ +
.
Section 2.3 204 Combinatorics
204
10. ((((BBBBell Numbersell Numbersell Numbersell Numbers)))) A partition of a set A is defined as asset of nonempty
subsets, pairwise disjoint, whose union is A. The number of such partitions
of a set of size n is called the Bell number n
B . For example when 3n = the
Bell number 3
5B = since there are 5 partitions of a set of size 3. For
example, if { }, ,A a b c= the 5 partitions are:
{ } { } { }{ }{ } { }{ }{ } { }{ }{ } { }{ }{ }{ }
, ,
, ,
, ,
, ,
, ,
a b c
a b c
b a c
c a b
a b c
Enumerate the partitions of the set { }, , ,a b c d to find the Bell number 4
B .
Ans:Ans:Ans:Ans: 4
15B =
11. (Two Committees)(Two Committees)(Two Committees)(Two Committees) How many ways can the Snail Darter Society, who has
25 members, elect an executive committee of 2 members and an
entertainment committee of 4 members if no member of the society can
serve on both committees?
Ans:Ans:Ans:Ans: 25 23 25 23
300 1771 531 3002 3 23 2 20 3
! !,
! ! ! !
= = × =
12. (Serving on More than One Committee) (Serving on More than One Committee) (Serving on More than One Committee) (Serving on More than One Committee) How many ways can the
Snail Darter Society, who has 25 members, elect an executive
committee of 2 members, an entertainment committee of 3 members,
and a welcoming committee of 2 members if members can serve on more
than one committee?
Ans:Ans:Ans:Ans: 25 25 25 25 23 25
2 3 2 23 2 20 3 23 2
! ! !
! ! ! ! ! !
=
13. (Counting Softball Teams) (Counting Softball Teams) (Counting Softball Teams) (Counting Softball Teams) A college softball team is taking 25 players on
a road trip. The traveling squad consists of 3 catchers, 6 pitchers, 8
Section 2.3 205 Combinatorics
205
infielders, and 8 outfielders. Assuming each player can only play her own
position, how many different teams can the coach put on the field?
Ans:Ans:Ans:Ans: 25 22 18 10
3 6 8 8
14. (Permutations as Groups)(Permutations as Groups)(Permutations as Groups)(Permutations as Groups) The 3! 6= permutations of set { }1, 2,3 can be
illustrated by the 2 3× arrays
1 2 3
1 2 3α
=
1 2 3
2 3 1β
=
1 2 3
3 1 2γ
=
1 2 3
1 3 2δ
=
1 2 3
2 1 3ε
=
1 2 3
3 2 1η
=
where the bottom row of each array shows how the top row is permuted.
Carrying out one permutation followed by another defines a multiplication
of the permutations. For instance ε δβ= means we do permutationβ first
then permutation δ second, which yields the permutation ε (check it
yourself). Compute the following.
a) αβ
Ans:Ans:Ans:Ans: 1 2 3 1 2 3 1 2 3
1 2 3 2 3 1 2 3 1αβ α
= = =
b) 2α
Ans:Ans:Ans:Ans: 21 2 3 1 2 3 1 2 3
1 2 3 1 2 3 2 3 1α α
= = =
c) 2β
Ans:Ans:Ans:Ans: 21 2 3 1 2 3 1 2 3
2 3 1 2 3 1 3 1 2β γ
= = =
d) δε
Ans:Ans:Ans:Ans: 1 2 3 1 2 3 1 2 3
1 3 2 2 1 3 3 1 2δε γ
= = =
Section 2.3 206 Combinatorics
206
After you get the hang of multiplying permutations, make a 6 6×
multiplication table of all products. This table describes what in group
theory is called the symmetric groupsymmetric groupsymmetric groupsymmetric group of order 3, denoted by 3S .
15. (The Josephus Problem)(The Josephus Problem)(The Josephus Problem)(The Josephus Problem) A thousand Roman slaves are put in a circle
numbered from 1 to 1000. All are to be shot except one lucky survivor.
Slave #1 graciously allows slave #2 to begin, so the order of shooting is
2,4,6,… . After a slave is shot, the body is removed from the circle and the
shooting continues. For example, if there were six slaves the order of the
shooting would be 2,4,6,3,1 and slave #59.
a) Find the position ( )p n of the surviving slave if there are n slaves for
1 2 10, ,...,n = .
Ans: Ans: Ans: Ans:
n 1111 2222 3333 4444 5555 6666 7777 8888 9999 10101010
( )p n 1111 1111 3333 1111 3333 5555 7777 1111 3333 5555
b) Verify that if n is odd, ( )p n satisfies the recurrence relation
( ) ( ) ( )2 1 2 1 1 1,p n p n p+ = + = .
Ans:Ans:Ans:Ans:
( ) ( )( ) ( )( ) ( )( ) ( )
3 2 1 1 2 1 1 3
5 2 2 1 2 1 1 3
7 2 3 1 2 3 1 7
9 2 4 1 2 1 1 3
p p
p p
p p
p p
= + = ⋅ + =
= + = ⋅ + =
= + = ⋅ + =
= + = ⋅ + =
c) Verify that if n is even, ( )p n satisfies the recurrence relation
( ) ( ) ( )2 2 1 2 1,p n p n p= − = .
Ans:Ans:Ans:Ans:
9 The problem is named after Flavius Josephus, a Jewish historian living in the 1st century.
According to legend, he and 40 fellow soldiers were trapped in a cave, surrounded by Romans.
They chose suicide over capture and decided that to form a circle and start killing themselves,
killing alternate persons in the circle.
Section 2.3 207 Combinatorics
207
( ) ( )( ) ( )( ) ( )( ) ( )( ) ( )
2 2 1 1 2 1 1 1
4 2 2 1 2 1 1 1
6 2 3 1 2 3 1 5
8 2 4 1 2 1 1 1
10 2 5 1 2 3 1 5
p p
p p
p p
p p
p p
= − = ⋅ − =
= − = ⋅ − =
= − = ⋅ − =
= − = ⋅ − =
= − = ⋅ − =
d) Which position would you choose if there were 100 slaves in the circle?
Ans:Ans:Ans:Ans: We can use the results from b), c), getting
( ) ( )( ) ( )( ) ( )( ) ( )( ) ( )( ) ( )
3 2 1 1 2 1 1 3
6 2 3 1 2 3 1 5
12 2 6 1 2 5 1 9
25 2 12 1 2 9 1 19
50 2 25 1 2 19 1 37
100 2 50 1 2 37 1 73
p p
p p
p p
p p
p p
p p
= + = ⋅ + =
= − = ⋅ − =
= − = ⋅ − =
= + = ⋅ + =
= − = ⋅ − =
= − = ⋅ − =
Hence, slave #73 is the lucky one.
16. (Catalan Numbers) (Catalan Numbers) (Catalan Numbers) (Catalan Numbers) The Catalan number 1 2 3, , , ,...n
C n = is the number of
different ways a convex polygon with 2n + sides can be subdivided into
triangles. Figure 6 shows the Catalan numbers 1 2 3 4
and , , ,C C C C . Can you
find the Catalan number 5
C ?
Section 2.3 208 Combinatorics
208
First 4 Catalan Numbers
Figure 6
Ans:Ans:Ans:Ans: 42. The general formula for the number of ways n
c to dissect a
convex polygon is 21
1n
nc
nn
= +
.
17. (Famous Apple Problem) (Famous Apple Problem) (Famous Apple Problem) (Famous Apple Problem) We wish to distribute 8 identical apples to 4
children. How many ways is this possible if each child gets at least one
apple?
Ans:Ans:Ans:Ans: Think of the 8 apples in a row like a a a a a a a a . We can distribute
the apples to the 4 children by placing 3 bars | in some of the 7 gaps
between the apples. For example, if we place the bars like
a a | a | a a a a | a
this means we give the first child 2 apples, the second child 1 apple, the
third apple 4 apples, and the 4th child 1 apple. So the total number of ways
to do this is the number of ways of selecting the 3 bars from 7 gaps, or
7 7
35 ways3 3 4
!
! !
= =
.
Section 2.3 209 Combinatorics
209
18181818. (Pigeon Hole Principle) . (Pigeon Hole Principle) . (Pigeon Hole Principle) . (Pigeon Hole Principle) Apply the pigeonhole to prove that at least two
people in New York City have the same number of hairs on their head. Hint:
You may want to make a few assumptions regarding the population of NYC
and the maximum number of hairs on the human head.
Ans:Ans:Ans:Ans: Since there are more people in NYC than any human has hairs on their
head, the pigeon hole states the given result.
19. (Derangements)(Derangements)(Derangements)(Derangements) A derangement is a permutation in which none of the
elements remain in their natural order. For example the only derangements
of ( )1,2,3 are ( ) ( )3,1, 2 and 2,3,1 . Hence we write !3 2= . Nicolas Bernoulli
proved that the number of derangements of a set of size n is
( )
1
1! !
!
kn
k
n nk=
−= ∑
How many derangements are there for the members ( )1,2,3, 4 ?
Enumerate them.
Ans:Ans:Ans:Ans: We begin by putting B,C,D in the first position. If B is in the first
position we put A,C,D in the second position, and so on. Continuing
in this manner we see that there are 9 derangements of 4 objects,
which are : BADC, BCDA, BDAC, CADB, CDAB, CDBA,DABC,
DCAB, DCBA
20202020. . . . (Counting Handshakes)(Counting Handshakes)(Counting Handshakes)(Counting Handshakes) There are 20 people in a room and each person
shakes hands with everyone else. How many handshakes are there?
Ans:Ans:Ans:Ans: The first person shakes hands with 19 other persons, the second
person shakes hands with 18 other persons (since we have already counted
the handshake with the first person). Continuing on, the third person shakes
hands with 17 persons, etc … so the total number of handshakes is
19 20
19 18 17 2 1 190 handshakes2
...×
+ + + + + = =
21. (Counting Functions)(Counting Functions)(Counting Functions)(Counting Functions) How many functions are there from { }, ,A a b c= to
{ }0,1, 2B = ? Write them down and draw the graphs for a few of them.
Section 2.3 210 Combinatorics
210
Ans:Ans:Ans:Ans: The value of the function at a can be either 0 or 1; likewise the value
of the function at b can be 0 or 1; likewise at c . Hence, the total number
of functions is 32 8= . The eight functions, calling them all by f , are
( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )
0 0 0
0 0 1
0 1 0
0 1 1
1 0 0
1 0 1
1 1 0
1 1 1
, ,
, ,
, ,
, ,
, ,
, ,
, ,
, ,
f a f b f c
f a f b f c
f a f b f c
f a f b f c
f a f b f c
f a f b f c
f a f b f c
f a f b f c
= = =
= = =
= = =
= = =
= = =
= = =
= = =
= = =
22. (Combinatorial Euclidean Algorithm)(Combinatorial Euclidean Algorithm)(Combinatorial Euclidean Algorithm)(Combinatorial Euclidean Algorithm) Finding the greatest integer that
divides two positive integers can be interpreted as a combinatorial problem
where one seeks all possibilities of divisors. However, the method of
finding the greatest common divisor, denoted ( )gcd ,n m , of two natural
numbers ,n m is called the Euclidean AlgorithmEuclidean AlgorithmEuclidean AlgorithmEuclidean Algorithm and avoids the combinatorial
headache of looking through all divisors. The algorithm makes the
fundamental observation that the quotient of divided by m n has the form:
, 0n r
q r mm m
= + ≤ <
where q is the quotient of /n m and r the remainder. Using the basic
property10 ( ) ( )gcd , gcd ,n m m r= , it is possible to find a sequence of
“decreasing” gcds, which eventually lead to an obvious gcd. For example,
the greatest common divisor of 255 and 68 is 17 as seen from the
“decreasing” sequence of gcds:
( ) ( ) ( ) ( )gcd 255,68 gcd 68,51 gcd 51,17 gcd 17,0 17= = = =
The numbers 51, 17, 0 in the gcds are the remainders of
10 This is true since any whole number that divides both and m n also divides and n r n mq= − ,
and vice versa. true.
Section 2.3 211 Combinatorics
211
255remainder of is 51
68
68remainder of is 17
51
51remainder of is 0
17
Use the Euclidean Algorithm to find the greatest common divisors of the
following pairs of numbers11.
a) ( )gcd 25,3
b) ( )gcd 54,36
c) ( )gcd 900, 22
d) ( )gcd 816,438
Ans:Ans:Ans:Ans:
a) ( )gcd 25,3 1=
b) ( )gcd 54,36 1=
c) ( )gcd 900, 22 6=
d) ( )gcd 816,438 6=
23. (Partitions) (Partitions) (Partitions) (Partitions) A partition of a natural number n is a sequence of natural
numbers 1 2a a≥ ≥� such that 1 2 kn a a a= + + +� for some k . There are two
partitions of 2, which are 2 2, 2 1 1= = + , which we denote by 2, 11. There are
three partitions of 3, which are 3 3, 3 2 1, 3 1 1 1= = + = + + , which we denote by 3,
21, 111. Find the partitions of 4, 5 and 6. How would you construct a total
order for the partitions of the natural numbers? You are relieved from finding
the partitions of 200n = since there are 3,972,999,029,388 of them.
Ans:Ans:Ans:Ans: The partitions of 4 are 1111, 211, 22, 31, and 4 . The partitions of
5 are 11111, 2111, 221, 311, 32, 41, 5 . There are 11 partitions of 6 ,
which are 111111, 21111, 2211, 3111, 222, 321, 411, 33, 42, 51, 6.
A total order " "≺ for a partition is 1 2 1 2a a b b+ + + +�≺ � if and only if
11 There are other methods for finding the greatest common divisor of two natural numbers. One
could factor each natural number into its prime components, find the common factors, then take the
product of the common factors. For example, 816 and 438 have factors of 2 and 3, and so 6 would
be their greatest common divisor.
Section 2.3 212 Combinatorics
212
1 2 1 2k ka a a b b b+ + + ≤ + + +� � for all 1k ≥ . With this ordering we have
4 : 1111 211 22 31 4
5 : 11111 2111 221 311 32 41 5
6 : 11111 21111 2211 222 3111 321 33 411 42 51 6
n
n
n
=
=
=
≺ ≺ ≺ ≺
≺ ≺ ≺ ≺ ≺ ≺
≺ ≺ ≺ ≺ ≺ ≺ ≺ ≺ ≺ ≺
24. . . . (Euler’s Theorem)(Euler’s Theorem)(Euler’s Theorem)(Euler’s Theorem) There are many identities related to integer partitions.
The first one was due to the Swiss mathematician Leonhard Euler who proved
that any number has as many partitions consisting of odd numbers as there are
partitions consisting of distinct numbers. For example, the number 5 has 3
partitions consisting only of odd numbers, 11111, 311, and 5; and 3 partitions
consisting of distinct numbers is 5, 41, 32. Show Euler’s partition identity
holds for the number 6.
AnsAnsAnsAns: Partitions consisting of odd numbers: 111111, 3111, 33, 51 and partitions
of distinct numbers: 321, 42, 51, and 6.
25. (Fun Problem)(Fun Problem)(Fun Problem)(Fun Problem) How many integers are there between 000 and 999
where the middle digit is the average of the other two digits? For example,
246 and 420 satisfy the condition.
AnsAnsAnsAns: Firet make the observation that the first two digits cannot be 0 and
that the only possible numbers where the last digit is 0 are the four
numbers 210, 420, 630, and 840. We now include the nine numbers 111,
222, 333, … , 999 where all the digits are the same. We now list the
remaining possibilities when the middle digit is 2,3,4,… 8. We have
middle digit 2: 123
middle digit 3: 135 234
middle digit 4: 147 246 345
middle digit 5: 159 258 357 456
middle digit 6: 369 468 567
middle digit 7: 579 678
middle digit 8: 789
,
, ,
, , ,
, ,
,
and
Section 2.3 213 Combinatorics
213
middle digit 2: 321
middle digit 3: 531, 432
middle digit 4: 741,642,543
middle digit 5: 951,852,753,654
middle digit 6: 963,864,765
middle digit 7: 975,876
middle digit 8: 987
Counting up these we have ( )2 16 9 4 45+ + = . Now, if you want to be clever,
realize that that the first and last digits must both be odd or both even.
Thus there are 5 5 25⋅ = possible numbers where the first and last digits are
odd, and 5 4 20⋅ = even-even pairs that do not use 0 in the first digit.
Hence we have a total of 25 + 20 = 45 numbers.
26. ((((Basic Basic Basic Basic Pizza CuttingPizza CuttingPizza CuttingPizza Cutting) ) ) ) You are given a circular pizza and ask to cut it into as
many pieces as possible, where a cut means any line that passes all the way
through the pizza, although not necessarily through the center. The picture
below shows a pizza with four typical slices dividing the pizza into 9 pieces of
pizza.12
Typical pizza with 4 cuts and 9 pieces
a) How many different pieces can you obtain with three slices?
b) How many different pieces can you obtain with four slices?
12 We could just as well said that the entire plane can be subdivided into 9 disjoint sections. In fact
normally the “pizza problem” is stated in terms of the entire plane and not simply a the inside of a circle.
Section 2.3 214 Combinatorics
214
c) Show how you can slice a pizza with 4 slices to obtain 5, 6, 7, 8, 9, 10,
and 11 pieces ?
Ans:Ans:Ans:Ans: a) With 3 slices one can obtain 4, 5, 6, or 7 pieces. If none of the
slices intersect then there will be 4 slices, if two slices intersect there will
be 5 slices, etc. Examples are
b) With 4 slices one can obtain 5-11 pieces, depending on the number of
intersections of the slices, assuming the intersections only involve two
slices. We leave this fun problem to the reader. (Note that if all the slices
are all made through the center of the pizza, then the number of pieces of
pizza will be twice the number of slices.)
c) We leave this fun problem for the reader.
Section 2.3 215 Combinatorics
215
27. (Pizza Cutter’s Formula)(Pizza Cutter’s Formula)(Pizza Cutter’s Formula)(Pizza Cutter’s Formula) A pizza cutter wants to cut a pizza in such a
way that it has the maximum number of pieces, not necessarily the same
size or shape. The only restriction on how the pizza is cut is that each cut
much pass all the way through the pizza, not necessary through the center.
Typical Pizza after 3 cuts with 7 pieces
a) If ( )P n is the maximum number of pieces from n cuts, then
( ) ( )1P n P n n= − + .
b) Show the pizza cutter’s formula is
( )2 1
2
n nP n
+ += .
Ans:Ans:Ans:Ans: a) The first line cuts the pizza into 2 pieces, the second piece into a
max of 4 pieces and the third line yields a maximum of 7 pieces, and so on.
We made the n th slice so it cuts the previous 1n − slices (we can do this by
making it very close to the previous slice), hence the n th slice is cut into
n segments, each of those segments cutting one piece of pizza into two
pieces, and so the n th slice yields n new pieces of pizza. Hence, if
( )1P n − is the maximum number of pieces after 1n − cuts, then to maximize
the number after n cuts we make the n -th cut intersect each of the 1n −
cuts, giving an extra n pieces. Hence ( ) ( )1P n P n n= − + .
Section 2.3 216 Combinatorics
216
b) We can write
( ) ( )( ) ( )( ) ( ) ( )
( )( )
( )
2
1
1 1
2 1 1
1 0 1 2
1 1 2
11
22
2
... ...
P n P n n
P n n n
P n n n n
P n
n
n n
n n
= − +
= − + − + = − + − + − +
= + + + + + = + + + +
+= +
+ +=
�
�
Note that for 3n ≥ this is larger than if the pizza cutter made all cuts
through the center of the pizza. If the cuts are made through the center of
the pizza, then the number of pieces is twice the number of cuts. Cutting a
pizza into pieces is the same as cutting the plane into pieces. More
problems along this line is to cut 3-dimensional space by planes, and
higher-dimensional spaces by hyper-planes. The reader can find more
about how to divide spaces by lines, circles, ellipses, etc by “googling”
phrases like “subdividing the plane into regions.”