Module #21 - Relations
University of Florida
Dept. of Computer & Information Science & Engineering
COT 3100
Applications of Discrete Structures
Dr. Michael P. Frank
Slides for a Course Based on the Text
Discrete Mathematics & Its Applications
(5th Edition)
by Kenneth H. Rosen
10/3/2015
(c)2001-2003, Michael P. Frank
1
Module #21 - Relations
Module #21:
Relations
Rosen 5th ed., ch. 7
~35 slides (in progress), ~2 lectures
10/3/2015
(c)2001-2003, Michael P. Frank
2
Module #21 - Relations
Binary Relations
• Let A, B be any sets. A binary relation R from A to B,
written (with signature) R:A×B, or R:A,B, is (can be
identified with) a subset of the set A×B.
– E.g., let < : N↔N :≡ {(n,m) | n < m}
• The notation a R b or aRb means that (a,b)R.
– E.g., a < b means (a,b) <
• If aRb we may say “a is related to b (by relation R)”,
– or just “a relates to b (under relation R)”.
• A binary relation R corresponds to a predicate function
PR:A×B→{T,F} defined over the 2 sets A,B;
– e.g., predicate “eats” :≡ {(a,b)| organism a eats food b}
10/3/2015
(c)2001-2003, Michael P. Frank
3
Module #21 - Relations
Complementary Relations
• Let R:A,B be any binary relation.
• Then, R:A×B, the complement of R, is the
binary relation defined by
R :≡ {(a,b) | (a,b)R} = (A×B) − R
• Note this is just R if the universe of
discourse is U = A×B; thus the name
complement.
• Note the complement of R is R.
Example: < = {(a,b) | (a,b)<} = {(a,b) | ¬a<b} = ≥
10/3/2015
(c)2001-2003, Michael P. Frank
4
Module #21 - Relations
Inverse Relations
• Any binary relation R:A×B has an inverse
relation R−1:B×A, defined by
R−1 :≡ {(b,a) | (a,b)R}.
E.g., <−1 = {(b,a) | a<b} = {(b,a) | b>a} = >.
• E.g., if R:People→Foods is defined by
a R b  a eats b, then:
b R−1 a  b is eaten by a. (Passive voice.)
10/3/2015
(c)2001-2003, Michael P. Frank
5
Module #21 - Relations
Relations on a Set
• A (binary) relation from a set A to itself is
called a relation on the set A.
• E.g., the “<” relation from earlier was
defined as a relation on the set N of natural
numbers.
• The (binary) identity relation IA on a set A is
the set {(a,a)|aA}.
10/3/2015
(c)2001-2003, Michael P. Frank
6
Module #21 - Relations
Reflexivity
• A relation R on A is reflexive if aA, aRa.
– E.g., the relation ≥ :≡ {(a,b) | a≥b} is reflexive.
• A relation R is irreflexive iff its complementary
relation R is reflexive.
– Example: < is irreflexive, because ≥ is reflexive.
– Note “irreflexive” does NOT mean “not reflexive”!
• For example: the relation “likes” between people is not
reflexive, but it is not irreflexive either.
– Since not everyone likes themselves, but not everyone dislikes
themselves either!
10/3/2015
(c)2001-2003, Michael P. Frank
7
Module #21 - Relations
Symmetry & Antisymmetry
• A binary relation R on A is symmetric iff
R = R−1, that is, if (a,b)R ↔ (b,a)R.
– E.g., = (equality) is symmetric. < is not.
– “is married to” is symmetric, “likes” is not.
• A binary relation R is antisymmetric if
a≠b, (a,b)R → (b,a)R.
– Examples: < is antisymmetric, “likes” is not.
– Exercise: prove this definition of antisymmetric is
equivalent to the statement that RR−1  IA.
10/3/2015
(c)2001-2003, Michael P. Frank
8
Module #21 - Relations
Transitivity
• A relation R is transitive iff (for all a,b,c)
(a,b)R  (b,c)R → (a,c)R.
• A relation is intransitive iff it is not
transitive.
• Some examples:
– “is an ancestor of” is transitive.
– “likes” between people is intransitive.
– “is located within 1 mile of” is… ?
10/3/2015
(c)2001-2003, Michael P. Frank
9
Module #21 - Relations
Totality
• A relation R:A×B is total if for every aA,
there is at least one bB such that (a,b)R.
• If R is not total, then it is called strictly
partial.
• A partial relation is a relation that might be
strictly partial. (Or, it might be total.)
– In other words, all relations are considered
“partial.”
10/3/2015
(c)2001-2003, Michael P. Frank
10
Module #21 - Relations
Functionality
• A relation R:A×B is functional if, for any aA, there is at most 1
bB such that (a,b)R.
– “R is functional”  aA: ¬b1≠b2B: aRb1  aRb2.
– Iff R is functional, then it corresponds to a partial function R:A→B
• where R(a)=b  aRb; e.g.
• R is
– E.g., The relation aRb :≡ “a + b = 0” yields the function −(a) = b.
antifunctional if its inverse relation R−1 is functional.
– Note: A functional relation (partial function) that is also antifunctional is an
invertible partial function.
• R is a total function R:A→B if it is both functional and total, that is,
for any aA, there is exactly 1 b such that (a,b)R.
I.e., aA: ¬!b: aRb.
– If R is functional but not total, then it is a strictly partial function.
– Exercise: Show that iff R is functional and antifunctional, and both it and its
inverse are total, then it is a bijective function.
10/3/2015
(c)2001-2003, Michael P. Frank
11
Module #21 - Relations
Lambda Notation
• The lambda calculus is a formal mathematical language for
defining and operating on functions.
– It is the mathematical foundation of a number of functional (recursive
function-based) programming languages, such as LISP and ML.
• It is based on lambda notation, in which “λa: f(a)” is a way to
denote the function f without ever assigning it a specific symbol.
– E.g., (λx. 3x2+2x) is a name for the function f:R→R where f(x)=3x2+2x.
• Lambda notation and the “such that” operator “” can also be used
to compose an expression for the function that corresponds to any
given functional relation.
– Let R:A×B be any functional relation on A,B.
– Then the expression (λa: b  aRb) denotes the function f:A→B
where f(a) = b iff aRb.
• E.g., If I write: f :≡ (λa: b  a+b = 0),
this is one way of defining the function f(a)=−a.
10/3/2015
(c)2001-2003, Michael P. Frank
12
Module #21 - Relations
Composite Relations
• Let R:A×B, and S:B×C. Then the composite SR
of R and S is defined as:
SR = {(a,c) | b: aRb  bSc}
• Note that function composition fg is an example.
• Exer.: Prove that R:A×A is transitive iff RR = R.
• The nth power Rn of a relation R on a set A can be
defined recursively by:
R0 :≡ IA ;
Rn+1 :≡ RnR for all n≥0.
– Negative powers of R can also be defined if desired, by
R−n :≡ (R−1)n.
10/3/2015
(c)2001-2003, Michael P. Frank
13
Module #21 - Relations
§7.2: n-ary Relations
• An n-ary relation R on sets A1,…,An,
written (with signature) R:A1×…×An or
R:A1,…,An, is simply a subset
R  A1× … × An.
• The sets Ai are called the domains of R.
• The degree of R is n.
• R is functional in the domain Ai if it contains at
most one n-tuple (…, ai ,…) for any value ai
within domain Ai.
10/3/2015
(c)2001-2003, Michael P. Frank
14
Module #21 - Relations
Relational Databases
• A relational database is essentially just an
n-ary relation R.
• A domain Ai is a primary key for the
database if the relation R is functional in Ai.
• A composite key for the database is a set of
domains {Ai, Aj, …} such that R contains at
most 1 n-tuple (…,ai,…,aj,…) for each
composite value (ai, aj,…)Ai×Aj×…
10/3/2015
(c)2001-2003, Michael P. Frank
15
Module #21 - Relations
Selection Operators
• Let A be any n-ary domain A=A1×…×An,
and let C:A→{T,F} be any condition
(predicate) on elements (n-tuples) of A.
• Then, the selection operator sC is the
operator that maps any (n-ary) relation R on
A to the n-ary relation of all n-tuples from R
that satisfy C.
– I.e., RA, sC(R) = {aR | sC(a) = T}
10/3/2015
(c)2001-2003, Michael P. Frank
16
Module #21 - Relations
Selection Operator Example
• Suppose we have a domain
A = StudentName × Standing × SocSecNos
• Suppose we define a certain condition on A,
UpperLevel(name,standing,ssn) :≡
[(standing = junior)  (standing = senior)]
• Then, sUpperLevel is the selection operator that takes
any relation R on A (database of students) and
produces a relation consisting of just the upperlevel classes (juniors and seniors).
10/3/2015
(c)2001-2003, Michael P. Frank
17
Module #21 - Relations
Projection Operators
• Let A = A1×…×An be any n-ary domain, and
let {ik}=(i1,…,im) be a sequence of indices
all falling in the range 1 to n,
– That is, where 1 ≤ ik ≤ n for all 1 ≤ k ≤ m.
• Then the projection operator on n-tuples
is defined by:
P{ i k } : A  Ai1    Ai m
P{ i k } ( a1 ,..., a n )  ( a i1 ,..., a i m )
10/3/2015
(c)2001-2003, Michael P. Frank
18
Module #21 - Relations
Projection Example
• Suppose we have a ternary (3-ary) domain
Cars=Model×Year×Color. (note n=3).
• Consider the index sequence {ik}= 1,3. (m=2)
• Then the projection P{i }simply maps each tuple
k
(a1,a2,a3) = (model,year,color) to its image:
( a i1 , a i 2 )  ( a1 , a 3 )  ( model , color )
• This operator can be usefully applied to a whole
relation RCars (a database of cars) to obtain a
list of the model/color combinations available.
10/3/2015
(c)2001-2003, Michael P. Frank
19
Module #21 - Relations
Join Operator
• Puts two relations together to form a sort of
combined relation.
• If the tuple (A,B) appears in R1, and the
tuple (B,C) appears in R2, then the tuple
(A,B,C) appears in the join J(R1,R2).
– A, B, and C here can also be sequences of
elements (across multiple fields), not just single
elements.
10/3/2015
(c)2001-2003, Michael P. Frank
20
Module #21 - Relations
Join Example
• Suppose R1 is a teaching assignment table,
relating Professors to Courses.
• Suppose R2 is a room assignment table
relating Courses to Rooms,Times.
• Then J(R1,R2) is like your class schedule,
listing (professor,course,room,time).
10/3/2015
(c)2001-2003, Michael P. Frank
21
Module #21 - Relations
§7.3: Representing Relations
• Some ways to represent n-ary relations:
– With an explicit list or table of its tuples.
– With a function from the domain to {T,F}.
• Or with an algorithm for computing this function.
• Some special ways to represent binary
relations:
– With a zero-one matrix.
– With a directed graph.
10/3/2015
(c)2001-2003, Michael P. Frank
22
Module #21 - Relations
Using Zero-One Matrices
• To represent a binary relation R:A×B by an |A|×|B|
0-1 matrix MR = [mij], let mij = 1 iff (ai,bj)R.
• E.g., Suppose Joe likes Susan and Mary, Fred likes
Mary, and Mark likes Sally.
• Then the 0-1 matrix
Susan
Mary
Sally
representation
Joe
1
0 
 1
of the relation


Fred
0
1
0
Likes:Boys×Girls


relation is:
 0
Mark
0
1 

10/3/2015
(c)2001-2003, Michael P. Frank

23
Module #21 - Relations
Zero-One Reflexive, Symmetric
• Terms: Reflexive, non-reflexive, irreflexive,
symmetric, asymmetric, and antisymmetric.
– These relation characteristics are very easy to
recognize by inspection of the zero-one matrix.
1

1

 any thing

any-
thing
1




1
0

0


anything

any- 
thing
0




0
Reflexive:
Irreflexive:
all 1’s on diagonal all 0’s on diagonal
10/3/2015


1




1
0


0




Symmetric:
all identical
across
diagonal
(c)2001-2003, Michael
P. Frank


1




0
0
0
0


1




Antisymmetric:
all 1’s are across
from 0’s
24
Module #21 - Relations
Using Directed Graphs
• A directed graph or digraph G=(VG,EG) is a set VG of
vertices (nodes) with a set EGVG×VG of edges
(arcs,links). Visually represented using dots for nodes, and
arrows for edges. Notice that a relation R:A×B can be
represented as a graph GR=(VG=AB, EG=R).
Matrix representation MR:
Joe
Fred
Mark
10/3/2015
Susan
 1

0

 0
Mary
1
1
0
Sally
0 

0

1 
Graph
rep. GR:
(c)2001-2003, Michael P. Frank
Edge set EG
(blue arrows)
Joe
Fred
Mark
Susan
Mary
Sally
Node set VG
(black dots)
25
Module #21 - Relations
Digraph Reflexive, Symmetric
It is extremely easy to recognize the reflexive/irreflexive/
symmetric/antisymmetric properties by graph inspection.
 



 
Reflexive:
Every node
has a self-loop
 
Irreflexive:
No node
These are asymmetric & non-antisymmetric
10/3/2015
Symmetric:
bidirectional
 
Antisymmetric:
bidirectional
These are non-reflexive & non-irreflexive
(c)2001-2003, Michael P. Frank
26
Module #21 - Relations
§7.4: Closures of Relations
• For any property X, the “X closure” of a set A is defined as
the “smallest” superset of A that has the given property.
• The reflexive closure of a relation R on A is obtained by
adding (a,a) to R for each aA. I.e., it is R  IA
• The symmetric closure of R is obtained by adding (b,a) to
R for each (a,b) in R. I.e., it is R  R−1
• The transitive closure or connectivity relation of R is
obtained by repeatedly adding (a,c) to R for each
(a,b),(b,c) in R.
– I.e., it is
R 
*
R
n Z
10/3/2015
n

(c)2001-2003, Michael P. Frank
27
Module #21 - Relations
Paths in Digraphs/Binary Relations
• A path of length n from node a to b in the directed
graph G (or the binary relation R) is a sequence
(a,x1), (x1,x2), …, (xn−1,b) of n ordered pairs in EG
(or R).
– An empty sequence of edges is considered a path of
length 0 from a to a.
– If any path from a to b exists, then we say that a is
connected to b. (“You can get there from here.”)
• A path of length n≥1 from a to itself is called a
circuit or a cycle.
• Note that there exists a path of length n from a to
b in R if and only if (a,b)Rn.
10/3/2015
(c)2001-2003, Michael P. Frank
28
Module #21 - Relations
Simple Transitive Closure Alg.
A procedure to compute R* with 0-1 matrices.
procedure transClosure(MR:rank-n 0-1 mat.)
A := B := MR;
for i := 2 to n begin
A := A⊙MR; B := B  A {join}
{note A represents Ri, B represents
end
R }
return B {Alg. takes Θ(n4) time}
i
i
j 1
10/3/2015
(c)2001-2003, Michael P. Frank
29
Module #21 - Relations
A Faster Transitive Closure Alg.
procedure transClosure(MR:rank-n 0-1 mat.)
A := MR;
for i := 1 to log2 n begin
2 1
A := A⊙(A+In); {A represents  R i }
j 1
end
return A
{Alg. takes only Θ(n3 log n) time!}
i 1
10/3/2015
(c)2001-2003, Michael P. Frank
30
Module #21 - Relations
Roy-Warshall Algorithm
• Uses only Θ(n3) operations!
Procedure Warshall(MR : rank-n 0-1 matrix)
W := MR
for k := 1 to n
for i := 1 to n
for j := 1 to n
wij := wij  (wik  wkj)
return W {this represents R*}
wij = 1 means there is a path from i to j going only through nodes ≤k
10/3/2015
(c)2001-2003, Michael P. Frank
31
Module #21 - Relations
§7.5: Equivalence Relations
• An equivalence relation (e.r.) on a set A is
simply any binary relation on A that is
reflexive, symmetric, and transitive.
– E.g., = itself is an equivalence relation.
– For any function f:A→B, the relation “have the
same f value”, or =f :≡ {(a1,a2) | f(a1)=f(a2)}
is an equivalence relation,
• e.g., let m=“mother of” then =m = “have the same
mother” is an e.r.
10/3/2015
(c)2001-2003, Michael P. Frank
32
Module #21 - Relations
Equivalence Relation Examples
• “Strings a and b are the same length.”
• “Integers a and b have the same absolute
value.”
• “Real numbers a and b have the same
fractional part.” (i.e., a − b  Z)
• “Integers a and b have the same residue
modulo m.” (for a given m>1)
10/3/2015
(c)2001-2003, Michael P. Frank
33
Module #21 - Relations
Equivalence Classes
• Let R be any equiv. rel. on a set A.
• The equivalence class of a,
[a]R :≡ { b | aRb }
(optional subscript R)
– It is the set of all elements of A that are “equivalent” to
a according to the eq.rel. R.
– Each such b (including a itself) is called a
representative of [a]R.
• Since f(a)=[a]R is a function of a, any equivalence
relation R can be defined using
aRb :≡ “a and b have the same f value”, given f.
10/3/2015
(c)2001-2003, Michael P. Frank
34
Module #21 - Relations
Equivalence Class Examples
• “Strings a and b are the same length.”
– [a] = the set of all strings of the same length as a.
• “Integers a and b have the same absolute value.”
– [a] = the set {a, −a}
• “Real numbers a and b have the same fractional
part (i.e., a − b  Z).”
– [a] = the set {…, a−2, a−1, a, a+1, a+2, …}
• “Integers a and b have the same residue modulo
m.” (for a given m>1)
– [a] = the set {…, a−2m, a−m, a, a+m, a+2m, …}
10/3/2015
(c)2001-2003, Michael P. Frank
35
Module #21 - Relations
Partitions
• A partition of a set A is the set of all the
equivalence classes {A1, A2, … } for some
equivalence relation on A.
– The Ai’s are all disjoint and their union = A.
• They “partition” the set into pieces. Within
each piece, all members of that set are
equivalent to each other.
10/3/2015
(c)2001-2003, Michael P. Frank
36
Module #21 - Relations
§7.6: Partial Orderings
• A relation R on A is called a partial ordering or
partial order iff it is reflexive, antisymmetric, and
transitive.
– We often use a symbol looking something like ≼ (or
analogous shapes) for such relations.
– Examples: ≤, ≥ on real numbers, ,  on sets.
– Another example: the divides relation | on Z+.
• Note it is not necessarily the case that either a≼b or b≼a.
• A set A together with a partial order ≼ on A is
called a partially ordered set or poset and is
denoted (A, ≼).
10/3/2015
(c)2001-2003, Michael P. Frank
37
Module #21 - Relations
Posets as Noncyclical Digraphs
• There is a one-to-one correspondence
between posets and the reflexive+transitive
closures of noncyclical digraphs.
– Noncyclical: Containing no directed cycles.
• Example: consider the poset ({0,…,10},|)
– Its minimal
digraph:
10/3/2015
1
2
3
5
7
4
6
9
10
(c)2001-2003, Michael P. Frank
8
0
38
Module #21 - Relations
More to come…
• More slides on partial orderings to be added
in the future…
10/3/2015
(c)2001-2003, Michael P. Frank
39