Chapter 5 & 7

Counting
 Objectives

⚫ Some questions we can face:

– “How can we determine the complexity of an
algorithm?”
– “How many cases possible are there to arrange n
objects?”
– “How can we generate input data from given initial
data?”
– A password consists of six characters. How many
such password with some criteria?
– How many bit strings of length n that do not have two
consecutive 0s?
 Contents

⚫ The basics of Counting

⚫ Recurrence relations
⚫ Divide – and – Conquer Algorithms and
recurrence relations
 5.1- The Basics of Counting
⚫ Basic Counting Principles : Product rule, sum
rule, The Inclusion-Exclusion Principle
⚫ Tree Diagrams
 Basic Counting Principle

⚫ Ex: Distribute 3 v A
distinct balls to 2 bags v
v B
⚫ We must do 3 times,
A
each time we have 2 v
v A
ways to select a bags. A v
⚫ 2.2.2 = 23 ways v B
B

v A
v

There are rn ways to A

v B
v
distribute n distinct v A
objects to r bags B v

B v B
 Basic Counting Principle

⚫ 12 teams. How many ways to select 3 teams for the

first, second and third teams

I
II III

Select 1/ 12 After select 1 team for the first After select 1 team for the
→ 12 ways position, one of 11 teams can second position, one of 10
be select into the second team remaining teams can be select
➔ 11 ways into the third team
➔ 10ways

12.11.10= 1320 ways

 Basic Counting Principle: Product rule

⚫ Suppose that a procedure can be broken

down into a sequence of two tasks. If there is
n1 ways to do the first task, there is n2 ways to
do the second task, then there are n1n2 ways to
do the procedure.
⚫ Example 1: 2 persons, 12 rooms. How many ways are there
to assign different room to these two person?
12 ways can be applied to assign the first
11 ways can be applied to assign the second after the first
assigned.
➔ 12.11= 132 ways.
Examples 3, 4, 5,6,7,8,9: Refer to pages 336, 337
 Basic Counting Principle: Sum rule

⚫ If a task can be done either in one of n1 ways

or in one of n2 ways, where none of the set of
n1 ways is the same as any of the set of n2
ways, then there are n1 + n2 ways to do this
task.
⚫ Example 11: There are 37 faculties, 83 students. A
person either a faculty or student can be choose to
attend a committee. How many ways are there to
select such person?
37 + 83= 120 ways
Examples 12, 13: Refer to page 339
 Basic Counting Principles…

⚫ Example 14: Name of a variable in BASIC language, an

case insensitive language, is a string of 1 or 2 characters
in which the first must be a character, the second may be
an alphanumeric characters, and must be different from
5 two-character keywords . How many different variable
names are there?

26 characters (a..z), 10 digits (0..9)

Name with 1 character: 26
Name with 2 character: 26.36 -5 =931
➔ There are 957 different names
 Basic Counting Principles…
⚫ Example 16 – page 341:Counting Internet Address
⚫ Class A address: largest networks
⚫ Class B address: medium-sized networks
⚫ Class C address: smallest networks.
IPv4
 Basic Counting Principles:
The Inclusion-Exclusion Principle

⚫ Suppose that a task can be done in n1 or n2

ways, but that some ways in the set of n1 ways
are the same as some ways in the set of n2
ways.
⚫ ns : number of ways that are the same in two
sets of ways. |AB|=|A| + |B| - |AB|
⚫ Total ways n= n1+ n2 –nS
Example 17: How many 8-bit strings either start
with 1 or end with 00?
27 + 26 – 25 = 160 – Refer to page 342
 Basic Counting Principles:
The Inclusion-Exclusion Principle
Example 18:
350 applications, in which
A1: 220 people majored in IT
A2: 147 people majored in business
A3: 51 people majored both in IT and business
How many applicants majored neither in IT nor
business?

Number of applicants who majored either IT or business =

(|A1| + |A2| - |A3|) = 220+147-51=316
Result: 350 – 316 = 34
 Tree Diagrams
⚫ Counting problem
can be solved
using Tree
Diagram.
⚫ Example: How
many 4-bit strings
do not have two
consecutive 1s?
⚫ Example 20: A playoff between two teams consists of
at lost 5 games. The first team that wins 3 games wins the
playoff. In how many different ways can the playoff occur?
⚫ Example 22: T-Shirts come in 5 sizes: S, M, L, XL, XXL.
Each size comes in 4 colors, W, R, G, B excepts for XL, which
comes only in R, G, B, and XXL, which comes only in G and B.
⚫ How many different shirts a souvenir shop have to stock to have at
least one of each available size and color of the T_shirt?
 Review

1. How many different bit strings of length seven are

there?
2. Counting Functions How many functions are there
from a set with m elements to a set with n elements?
3. Counting One-to-One Functions How many one-to-
one functions are there from a set with m elements to
one with n elements?
4. How many bit strings of length eight either start with
a 1 bit or end with the two bits 00?
7.1. Recurrence Relations (Hệ thức đệ quy)
Fibonacci numbers of Leonardo Pisano
⚫ Example 4: : f1=f2=1; fn= fn-2 + fn-1
The Tower of Hanoi Problem- Ex5, page 452
How many steps this
problem is solved if
there are n disks on
the peg 1?
 The Tower of Hanoi Problem- Ex5, page 452

Let Hn is the number of moves needed to solve the Tower of Hanoi

problem. We will set up a recurrence relation for the sequence Hn.
-Begin with n disk on peg 1.
-We transfer the top n-1 disk from peg1 to peg3 → Hn-1 moves .
-We transfer the largest disk from peg1 to peg 2 → 1 move
-We transfer n-1 disk from peg3 to peg 2
➔Hn= 2Hn-1 + 1
H1=1 ➔Hn= 2n -1 ( See page 453)
n=64 → 264-1=18 446 744 073 709 551 615. With 1 move/sec → 500
billion years.
Find recurrence relation and give initial conditions for the
number of bit strings of length n that do not have two
consecutive 0s. How many such bit strings are there of
length five.
 Find recurrence relation and give initial conditions for the
number of bit strings of length n that do not have two
consecutive 0s. How many such bit strings are there of
length five.
⚫ an: number of bit string of length n that do not have two consecutive 0s.
⚫ One bit string can terminate with bit 1 or 0
⚫ Format of a bit string that do not have two consecutive 0s:

Any bit string of length n-1 with no two

End with 1 1
consecutive 0s
Any bit string of length n-2 with no
End with 0 1 0
two consecutive 0s

➔an = an-1 + an-2 for n ≥ 3 { Fibonacci sequence }

➔a1=2 { two string “0” and “1” }
➔a2= 3 { “10” , “01” , “11” }
With n=5
➔ a3 = a2+ a1 =5
➔ a4= a3 + a2 = 5+3=8
➔ a5=a4 + a3 = 8+5=13
 Counting valid codewords
valid codeword: String of decimal digits that contains
an even number of 0 digits.

⚫ an: number of valid n-digit codewords

⚫ Find a recurrence relation for an
⚫ a1=9 // “1”, “2”, …, “9” ➔ “0” is not used
⚫ Study an:
Valid codeword with n-1 digits 1..9 9an-1 ways
Invalid codeword with n-1
digits ( number of 0s is odd)
0 10n-1 - an-1 ways
All possible codewords
minus valid codewords

an= 9an-1 + 10n-1 - an-1 = 8an-1 + 10n-1

7.3. Divide-and-Conquer Algorithms and
recurrence Relations
7.3. Divide-and-Conquer Algorithms and
recurrence Relations
Divide-and-Conquer Recurrence Relations
 Recurrence Relations for Binary Search
procedure binary-search(x, i, j)
if i>j then location=0
m= (i+j)/2
if x= am then location =m
else if x< am then location= binary-search(x, i, m-1)
else location= binary-search(x, m+1, j)

f(n) = f(n/2) +2
 Recurrence Relations for Finding Maximum of a
sequence

procedure max(i,j: integer ,ai,a2+1,…,aj: integers)

if i=j then
begin f(n) = 2f(n/2) +1
max:= ai
end
else
begin
m= (i+j)/2
max1= max (i,m,ai,ai+1,…,am)
max2= max (m+1,j,am+1,am+2,…,aj)
if max1>max2 then max:= max1
else max:=max2
 Theorem 1

f (n) = C1nlogb a + C2
Using Theorem 1
Theorem 2: Master Theorem
Theorem 2: Master Theorem
 An Demonstration: Closest-Pair Problem

⚫ n points in the plane. How to determine the closest-pair of points?

⚫ (1) Determine the distance of every pair of points.
⚫ (2) Determine the pair of points that have minimum distance.
⚫ ➔ C(n,2)= n(n-1)/2= O(n2)
⚫ Michal Samos proposed an approach that is O(nlogn) only.
⚫ Michal Samos’s approach
(1) Sorting points in order of increasing x coordinates → O(nlog(n))
(2) Sorting points in order of increasing y coordinates → O(nlog(n))
 An Demonstration: Closest-Pair Problem
(3) Using recursive approach 16 points
to divide the problem into 2
subproblem with n/2 points
(left and right points based
on x coordinates). Let ℓ is
the line that partitions two
subproblems. If there is any
point on this dividing line, we
decide these points among
the two parts if necessary)
(4) Finding out closest-pair of
points in two side ( dL, dR)
(5) Let d=min(dL, dR)
8 points 8 points
 An Demonstration: Closest-Pair Problem

(6) Studying area [l-d,l+d].

This area may be
contains the result.
(7) Because we sorted
points by their y
coordinate. We
examine for points p in
the strip of width 2d
that has the line l as its
center with upward
direction.
 An Demonstration: Closest-Pair Problem

Total number of
points in the strip
does not exceed n
and there are at
most 8 points,
including p, can lie
in or on the 2dxd
rectangle.
➔A point will be computed with 7 others.
➔At most 7n distances need to be compare with d to find the minimum
distance between points.
➔ The increasing function f(n) satisfies the recurrence relation :
f(n) = 2f(n/2) + 7n
➔ By the Master Theorem, it follows that f(n) is O(nlogn)
 Summary

⚫ Product rule: Suppose that a procedure can be

broken down into a sequence of two tasks. If
there is n1 ways to do the first task, there is n2
ways to do the second task, then there are n1n2
ways to do the procedure.
⚫ Sum rule: If a task can be done either in one of
n1 ways or in one of n2 ways, where none of the
set of n1 ways is the same as any of the set of
n2 ways, then there are n1 + n2 ways to do this
task.
 Summary

⚫ Inclusion-Exclusion Principle: Suppose that a task can

be done in n1 or n2 ways, but that some ways in the set
of n1 ways are the same as some ways in the set of n2
ways.
 Summary

⚫ Recurrence relation on the sequence { an } is

an expression in which an is expressed in terms
of one or more of previous items for all n with n ≥
n0 . {an} is called a solution of recurrence
relation.
⚫ Initial conditions: Terms that precede the first
item where the recurrence takes effect.
⚫ Modeling with recurrence relation: Finding out
an recurrence relation for input of a problem.
 Summary

⚫ Divide: Dividing a problem into one or more

instances of the same problem of smaller size
⚫ Conquer: Using the solutions of the smaller
problems to find a solution of the original
problem, perhaps with some additional work
⚫ Recurrence relation: f(n) = af(n/b) + g(n)
 Summary – Theorem 1

⚫ Let f be an increasing function that satisfies the

recurrence relation f(n)= af(n/b) + c
whenever n is divisible by b, where a ≥ 1 , b is an
integer and greater than 1, and c is a positive real
number. Then

Furthermore, when n=bk, where k is a positive integer,

f (n) = C1nlogb a + C2
Where C1=f(1) + c/(a-1) and C2= -c/(a-1)
Proof: page 477
Summary – Theorem 2- Master Theorem

Let f be an increasing function that satisfies the recurrence

relation f(n) = af(n/b) + cnd
Whenever n= bk, where k is a positive integer, a ≥ 1, b is an integer
greater than 1, and c and d are real numbers with c positive and d
nonnegative. Then

 O(n d ) if a  b d

f (n) is O(n d log n) if a = b d
 O(n logb a ) if a  b d

⚫ THANKS