CSCE 222

Discrete Structures for Computing

Relations
Dr. Hyunyoung Lee

Based on slides by Andreas Klappenecker

1
 Rabbits

Suppose we have three rabbits called

Albert, Bertram, and Chris that have
distinct heights.

Let us write (a,b) if a is taller than b.

Obviously, we cannot have both

(Albert, Bertram) and (Bertram,
Albert), so not all pairs of rabbit
names will occur.
Then the set of “taller than”
Suppose: Albert is taller than relation is:
Bertram, and Bertram is taller than
Chris. { (Albert, Bertram), (Bertram,
Chris), (Albert, Chris) }

2
 Rabbits

Let
A = { Albert, Bertram, Chris }
be the set of rabbits.

Then the “taller than” relation is a subset of the cartesian

product AxA, namely { (Albert, Bertram), (Bertram, Chris),
(Albert, Chris) } ⊆ A x A.

3
 Binary Relations

Let A and B be sets.

A binary relation from A to B is a subset of AxB.

A relation on a set A is a subset of AxA.

4
 Examples

Let us consider the following relations on the set of

integers:

A = { (a,b) in ZxZ | a <= b }

B = { (a,b) in ZxZ | a > b }

C = { (a,b) in ZxZ | a = b or a = -b }

5
 Notation

Let R be a relation from A to B. In other words, R

contains pairs (a,b) with a in A and b in B.
If (a,b) in R, then we say that a is related to b by R.

It is customary to use infix notation for relations.

Thus, we write a R b to express that a is related to b
by R. In other words, a R b if and only if (a,b) in R.

6
 Example

Let A be the set of city names of the USA. Let B be the set
of states. Define the relation C
C = { (a,b) in A x B | a is a city of b }
Then
(College Station, Texas)
(Austin, Texas)
(San Francisco, California)
all belong to the relation C.

7
 Remark

The concept of a relation generalizes the concept of a

function. A function f relates the argument x with its
function value f(x). The difference is that a relation can
relate an element x with more than one value.
For example, consider the relation

A = { (a,b) in ZxZ | a <= b }.

8
 Plan

We are going to study relations as mathematical

objects. This allows us to abstract from well-known
relations such as <=, =, “is taller than”, “likes the same
sport as”.
We identify some basic properties of relations. Then we
study relations generalizing the equality relation (so-
called equivalence relations), and relations generalizing
<= (so-called partial order relations).

9
Basic Properties of Relations

10
 Reflexivity

We call a relation R on a set A reflexive if and only if

(a,a) ∈ R holds for all a in A.
Example: The equality relation = on the set of integers
is reflexive, since a=a holds for all integers a.
The less than relation < on the set of integers is not
reflexive, since 1<1 does not hold.

11
 Test Yourself...

X1 Let | denote the divides relation on the set of

positive integers, so 2 | 4 means that there exists an
integer x such that 2x=4. Is the relation | reflexive?

X2 Let S be the set of students in this class. Consider

the relation R = “wears the same color shirt as.” Is
the relation R reflexive?

12
 Symmetry

We call a relation R on a set A symmetric if and only

if (a,b) ∈ R implies that (b,a) ∈ R holds.
Example: The equality relation = on the set of integers
is symmetric, since a=b implies that b=a.
The less than relation < on the set of integers is not
symmetric, since 1<2 but 2<1 does not hold.

13
 Test Yourself...

X1 Let | denote the divides relation on the set of

positive integers, so 2 | 4 means that there exists an
integer x such that 2x=4. Is the relation | symmetric?

X2 Let S be the set of students in this class. Consider

the relation R = “wears the same color shirt as”. Is
the relation R symmetric?

14
 Antisymmetry

We call a relation R on a set A antisymmetric if and

only if (a,b) ∈ R and (b,a) ∈ R imply that a=b.
Formally: ∀a∀b ((a,b) ∈ R ⋀ (b,a) ∈ R) a=b.
Example: The equality relation = on the set of integers
is antisymmetric, since a=b and b=a implies that a=b.
The less than relation < on the set of integers is
antisymmetric. Why?

15
 Test Yourself...

X1 Let | denote the divides relation on the set of

positive integers, so 2 | 4 means that there exists an
integer x such that 2x=4. Is the relation |
antisymmetric?
X2 Let S be the set of students in this class. Consider
the relation R = “wears the same color shirt as”. Is the
relation R antisymmetric?

16
 Warning

The meaning of antisymmetry is not opposite to the

meaning of symmetry! In fact, we have already seen
that the equality relation = on the set of integers is
both symmetric and antisymmetric.
You should very carefully study the meaning of these
terms.

17
 Transitive

We call a relation R on a set A transitive if and only if

(a,b) ∈ R and (b,c) ∈ R imply that (a,c) ∈ R

Example: The equality relation = on the set of integers

is transitive, since a=b and b=c implies that a=c.
The less than relation < on the set of integers is
transitive, since a<b and b<c imply that a<c.

18
 Test Yourself...

X1 Let | denote the divides relation on the set of

positive integers, so 2 | 4 means that there exists an
integer x such that 2x=4. Is the relation | transitive?

X2 Let S be the set of students in this class. Consider

the relation R = “wears the same color shirt as”. Is the
relation R transitive?

19
Equivalence Relations

20
 Equivalence Relation

A relation R on a set A is called an equivalence relation

if and only if R is reflexive, symmetric, and transitive.

- Reflexive: For all a in A, we have (a,a) ∈ R

- Symmetric: (a,b) in R (b,a) ∈ R

- Transitive: [ (a,b) ∈ R and (b,c) ∈ R ] (a,c) ∈ R

21
 Example: Equality

The equality relation = on the set of integers is an

equivalence relation.
Indeed,
the relation = is reflexive, since a=a holds for all
integers a.
the relation = is symmetric, since a=b implies that b=a.
the relation = is transitive, since a=b and b=c implies
that a=c.

22
 Example: Congruence mod m

Let m be a positive integer. For integers a and b, we write

a ≡ b (mod m)
if and only if m divides a-b.
For all a in Z, we have m | (a-a), since m 0 = 0 = a-a.
Thus, a ≡ a (mod m) holds for all integers a. Thus, the
relation is reflexive.
For a, b in Z, if a ≡ b (mod m), then this means that there
exists an integer k such that mk=a-b. Thus, m(-k) = b-a,
which implies b ≡ a (mod m). Thus, the relation is
symmetric.

23
 Example: Congruence mod m

If a ≡ b (mod m) and b ≡ c (mod m) holds, then this

means that there exist integers k and l such that
mk = a-b and ml = b-c
Hence, m(k+l) = a-b + b-c = a-c
This shows that a ≡ c (mod m) holds.
Therefore, the relation is transitive.

We can conclude that a ≡ b (mod m) is an equivalence

relation.

24
 Equivalence Classes

Let R be an equivalence relation on a set A. For an

element a in A, the set of elements
[a]R = { b in A | a R b }
is called the equivalence class of a.

25
 Example

Let us consider the equivalence relation a ≡ b (mod 4) on the

set of integers. Thus, two integers a and b are related
whenever their difference is a multiple of 4. Thus, the
equivalence classes are:
[0] = { ..., -8, -4, 0, 4, 8, ... }
[1] = { ..., -7,-3, 1, 5, 9, ... }
[2] = { ..., -6, -2, 2, 6,... }
[3] = { ..., -5, -1, 3, 7, ... }
Now note that [4] = [0], [5] = [1]. In fact, [0], [1], [2] and [3]
are all equivalence classes.

26
 Theorem

Let R be an equivalence relation on a set A. Then

the following statements are equivalent:
a) a R b
b) [a] = [b]
c) [a] ∩ [b] ≠ ∅

27
 = + by Induction Hypothesis
2 2
2n + 2 + n2 + n n2 + 3n + 2
= =
Proof
2 2
(n + 1)(n + 2)
=
2
Therefore, the claim follows by induction on n.
Let A(n) be the claimed equality.
Suppose that
Induction Step: We aRb
need toholds.
show thatWe 8n are going
1 : [A(n) to show
! A(n + 1)] that
[a] ✓ [b] holds.
As induction hypothesis,Let c 2that[a].
suppose A(n)This means that aRc
holds. Then,
holds. Since
n+1
X R is symmetric,
X n aRb implies that bRa.
fk = fn+1 + fk
By transitivity,
k=1
bRa and
k=1
aRc imply that bRc holds.
Hence, c 2 [b]. = fTherefore,
n+1 + fn+2 1 weby have
Inductionshown that [a] ✓
Hypothesis
[b]. The proof = fthat
n+3 1[b]by✓ [a] is similar. Hence, we
definition
have shown
Therefore, that
the claim statement
follows by inductiona) onimplies
n. statement b).
We will show now that b) implies c). Since a 2 [a],
we know that the equivalence class of a is not empty.
As [a] = [b] 6= ;, we have [a] \ [b] 6= ;.

28
 Proof (continued)

We will show now that c) implies a).

Suppose that [a] \ [b] 6= ;. Thus, there exists an
element c such that aRc and bRc. By symmetry, we
get cRb. It follows by transitivity that aRb holds. ⇤

29
Partial Order Relations

30
 Partial Orders

A relation R on a set A is called a partial order if and

only if it is reflexive, antisymmetric, and transitive.

A set A with a partial order is called a partially

ordered set (poset).

31
 Example 1

The “less than or equal to” relation ≤ on the set of

integers is a partial order relation.
Indeed, since a ≤ a holds for all integers a, the
relation ≤ is reflexive.
Since a ≤ b and b ≤ a implies that a = b, the relation is
antisymmetric.
Since a ≤ b and b ≤ c implies that a ≤ c, the relation is
transitive.

32
 Example 2

The divides relation | on the set of positive integers is a partial

order relation.
Indeed, since a|a for all positive integers a, the relation | is
reflexive.
If a|b and b|a, then there exist integers k and l such that a k
= b and b l = a. Therefore, a kl = a, so kl=1. This means that
either k=l=1 or k=l=-1. Since a and b are positive integers, we
cannot have a (-1) = b. Therefore, we must have k=l=1, which
means that a=b. Thus, | is an antisymmetric relation.
The relation | is transitive, since a|b and b|c means that there
exist integers k and l such that ak=b and bl=c, so a(kl)=c,
which implies that a | c.

33
 Test Yourself ...

X1 Is the less than relation < on the set of integers a

partial order relation?

X2 Let S be a set. Is the subset relation ⊆ on the set

P(S) a partial order relation?

34
 Comparable Elements

A partial order on a set S is often denoted by

symbols resembling the notation commonly used for
“less than or equal to”, namely ≤ or ⊑ or ≼
Let (S, ≼) be a partially ordered set. For two
elements a and b of S, we do not necessarily have
that one of the relations a ≼ b or b ≼ a holds. If one
of them holds, then we call a and b comparable
elements of S, otherwise a and b are incomparable.

35
 Total Orders

A partially ordered set (S, ≼) in which any two

elements are comparable is called a total order.

A totally ordered set is also called a chain.

For example, consider the set of positive integers N

with <=. Any two positive integers are comparable
with <=. It can form a chain such that 1 <= 2 <= 3
<= 4 <= …

36
 Lexicographic Ordering

Suppose that we have two partially ordered sets:

(A, ≼1 ) and (B, ≼2).
We can construct a partial order on AxB by defining
(a1, b1) ≼ (a2, b2)
if and only if (a1 = a2 and b1 ≼2 b2) or (a1 ≺1 a2) holds.
We call the relation ≼ the lexicographic order on the
cartesian product AxB.

37
 Example

Let Z be the set of integers, totally ordered with the

“less than or equal to” relation ≤.

In the lexicographic order ≼ on ZxZ, we have

(3,4) ≼ (4,2)
(3,7) ≼ (3,8)

38
 Hasse Diagram

Let (S, <=) be a finite partially ordered set.

Suppose that a and b are distinct elements of S such
that a <= b. We say that b covers a if and only if
there does not exist an element c in S such that a < c
< b.
The Hasse diagram of (S, <=) is a diagram in which an
element b of S is written above a and connected by a
line if and only if b covers a.

39
 Examples

Consider {2,4,5,10,12,20,25} with divisibility condition.

The Hasse diagram is given by
12 20
The cover relation for this
Hasse diagram is
4 10 25
{(2,4), (2,10), (4,12), (4,20),
(5,10), (5,25), (10,20)}.
2 5
The Hasse diagram is more economical than representing the
partial order relation by a directed graph (with an edge from
a to b whenever a <= b). Self-loops and transitively implied
relations are omitted.

40
Maximal and Minimal Elements

Let (S, <=) be a partially ordered set.

An element m in S is called maximal iff there does not
exist any element b in S such that m < b.

An element m in S is called minimal iff there does not

exist any element b in S such that b < m.

41
 Example

12 20
Determine the maximal elements of the set
{2,4,5,10,12,20,25}, 4 10 25

partially ordered by the divisibility relation. 2 5

The elements 12, 20, and 25 are the maximal elements.

Determine the minimal elements of the above partially
ordered set ( {2,4,5,10,12,20,25}, | ).
The elements 2 and 5 are the minimal elements.

42
 Least and Greatest Element

Let (S, <=) be a partially ordered set.

An element a in S is called the least element iff a <= b
holds for all b in S.
[A least element does not need to exist. If it does, then
it is uniquely determined.]
An element z in S is called the greatest element iff
b <= z holds for all b in S. 12 20

[A greatest element does not need to exist. 4 10 25

If it does, then it is uniquely determined.]
2 5

43
 Test Yourself...

X1 Determine the least and the greatest elements of

the set of positive integers partially ordered by
divisibility.

X2 Let S be a nonempty set. Partially order the power

set P(S) by inclusion. Determine the least and the
greatest elements of P(S).

44
Lattices

45
 Upper and Lower Bounds

Let (S, <=) be a partially ordered set.

Let A be a subset of S.
An element u of S is called an upper bound of A if and
only if a <= u holds for all a in A.
An element l of S is called a lower bound of A if and
only if l <= a holds for all a in A.

46
 Example

Let {2,4,5,10,12,20,25} with divisibility condition.

The Hasse diagram is given by
12 20

4 10 25

2 5
The subset A = {4,10} has 20 as an upper bound, and 2 as a
lower bound.
The subset A = {12} has 12 as an upper bound, and 2, 4 and 12
as lower bounds.
47
 Least Upper Bounds

Let (S, <=) be a partially ordered set, and A a subset of

S. An element u of S is called a least upper bound of A
if it is an upper bound that is less than any other
upper bound of A.

[Unlike upper bounds, the least upper bound is uniquely

determined if it exists]

48
 Greatest Lower Bounds

Let (S, <=) be a partially ordered set, and A a subset

of S. An element l of S is called a greatest lower
bound of A if it is a lower bound that is greater than
any other lower bound of A.

A greatest lower bound is uniquely determined if it

exists.

49
 Example

Consider the poset (S={2, 4, 6, 9, 12, 18, 27, 36, 48, 60, 72}, | ).
Draw the Hasse diagram: 72
60 48
36

12 18 27

4 6

2 9
What are the upper bounds of the subset A={2, 9} ?
What are the lower bounds of the subset B={60, 72} ?

50
 Lattices

A partially ordered set in which every pair has both a

least upper bound and a greatest lower bound is called
a lattice.

51
 Example

Consider the set (N,|) of positive integers that is

partially ordered with respect to the divisibility
relation.
Let a and b be two distinct positive integers. Then the
least upper bound of {a,b} is the least common multiple
of a and b. The greatest lower bound is the greatest
common divisor of {a,b}. Therefore, (N,|) is a lattice.

52