Relations
Ex. The following are some binary relations on {1, 2, 3, 4}.
R1 = {(1, 1), (2, 2), (3, 3), (4, 4)}.
R2 = {(1, 1), (2, 2), (3, 3), (4, 4), (1, 2), (2, 1)}.
R3 = {(1, 2), (2, 1), (3, 4), (4, 3)}.
R4 = {(1, 2), (1, 3), (2, 3), (3, 4)}.
R5 = {(1, 1), (2, 2), (2, 3), (3, 4), (2, 4)}.
reflexive: R1, R2 irreflexive: R3, R4 symmetric: R1, R2, R3 asymmetric: R4
antisymmetric: R1, R4, R5
transitive: R1, R2, R5
Ex. Define R to be a binary relation on the set of integers, where a R b iff ab≥0.
R is reflexive and symmetric.
R is not transitive (for example, −5 R 0, 0 R 8, but −5 8).
Suppose that A={1, 2, …, n}.
1. There are 2n2−n reflexive binary relations on A.
Each reflexive binary relation on A must contain (1, 1), (2, 2), …, (n, n).
|A×A − {(i, i): 1≤i≤n}| = n2 −n.
2. There are 2(n2+n)/2 symmetric binary relations on A.
A×A = {(i, i): 1≤i≤n} + {(i, j): 1≤i≤n, 1≤j≤n, and i≠j}.
2n × 2(n2−n)/2 = 2(n2+n)/2.
3. There are 2(n2−n)/2 reflexive and symmetric binary relations on A.
4. There are 2n×3(n2−n)/2 antisymmetric binary relations on A.
Each (i, i) can be either included or excluded.
For each pair of (i, j) and (j, i), there are three choices:
(a) include (i, j) and exclude (j, i);
(b) exclude (i, j) and include (j, i);
(c) exclude (i, j) and (j, i).
⇒ 2n×3(n2−n)/2.
5. There is no general formula for counting the number of transitive binary relations on A.
A poset A is called a lattice, if every two elements of A have their least upper bound and greatest lower bound in A.
Ex. The poset A in the above example is not a lattice.
Ex. Let A be the power set of {1, 2, 3}.
Define R on A as follows: a R b iff a⊂b.
The least upper bound (greatest lower bound) of a and b is a∪b (a∩b).
⇒ A is a lattice.
Ex. Define R on the set Z of integers as follows:
a R b iff 4 divides a−b.
R is an equivalence relation, and its equivalence classes, denoted by [0], [1], [2] and [3], are as follows:
[0]={…, −8, −4, 0, 4, 8, …}={4k| k∈Z};
[1]={…, −7, −3, 1, 5, 9, …}={4k+1| k∈Z};
[2]={…, −6, −2, 2, 6, 10, …}={4k+2| k∈Z};
[3]={…, −5, −1, 3, 7, 11, …}={4k+3| k∈Z}.
Ex. Define R on the set Z of integers as follows:
a R b iff a2=b2.
R is an equivalence relation, and its equivalence classes are {0}, {−1, 1}, {−2, 2}, …, {−i, i}, … .
Ex. Suppose that R is an equivalence relation on {1, 2, …, 7}, and the induced partition is {{1, 2}, {3}, {4, 5, 7}, {6}}.
Then, R = {(1, 1), (2, 2), (1, 2), (2, 1)}∪{(3, 3)}∪
{(4, 4), (5, 5), (7, 7), (4, 5), (5, 4), (4, 7), (7, 4), (5, 7), (7, 5)}∪{(6, 6)}.
There is a one-to-one correspondence between the set of equivalence relations on {1, 2, …, n} and the set of partitions of {1, 2, …, n}.
Ex. What is the number of equivalence relations on A={1, 2, ..., 6}?
Let f(m) be the number of onto functions from A to BB= {b1, b2, …, bm}, which can be evaluated by the principle of inclusion and exclusion.
The answer is equal to
f(1)+f(2)/2!+f(3)/3!+f(4)/4!+f(5)/5!+f(6)/6!
= 1+31+90+65+15+1
= 203.
Boolean
Algebra
( , , , ) = + + + + +
f w x y z wx yz wxyz wx y z wx yz wxyz wx y z
( ) ( )
( ) ( )
( ( ) )
( , , , ) = + +
= + + + = +
= 1 + (1 )
f w x y z wxz y y wx y z z wxyz wx y z
wxz wx y wxyz wx y z wx z y z wx y yz
wx z y y z wx y z y
+ + +
+ +
+ +
(
+ +)
( ( ) ) ( ( ) )
= +
z wx z + y z + z wx y + z y+ y
( ) ( )
= +
= + + + wx z y wx y z wxz wx y wx y wxz
+ +
The elements of a finite Boolean algebra can be partially ordered.
Suppose that (K,
⋅
, +) is a Boolean algebra.For a, b∈K, define ap b iff a
⋅
b=a.Then, p is a partial ordering.
reflexisive: a
⋅
a=a ⇒ ap a.antisymmetric: ap b, bp a ⇒ a
⋅
b=a, b⋅
a=b⇒ a=b.
transitive: ap b, bp c ⇒ a
⋅
b=a, b⋅
c=b⇒ a=a
⋅
b=a⋅
(b⋅
c)=(a⋅
b)⋅
c= a
⋅
c⇒ ap c.
The Hasse diagrams for the Boolean algebra of page 12 is depicted below.
Notice that 0p a, ap 1 for every a∈K.
(0=1 and 1=30 for the example above)
A nonzero element a∈K is called an atom of K, if bp a implies b=0 or b=a, where b∈K.
The Boolean algebra of the example above has three atoms:
2, 3, 5.
Fact 1. If a is an atom of K, then a
⋅
b=0 or a⋅
b=a forevery b∈K.
Fact 2. If a1 ≠a2 are two atoms of K, then a1
⋅
a2 =0.Fact 3. Suppose that a1, a2, …, an are atoms of K, and b is a nonzero element in K. Without loss of generality, assume b
⋅
ai ≠0 for 1≤i≤k, and b⋅
ai =0 else.Then, b=a1 +a2 + … +ak.
Fact 4. If K has n atoms, then |K|=2n.
For the example above, we have 10=2+5, 30=2+3+5, and |K|=23=8.
The proofs of the four facts can be found in pages 738 and 739 of Grimaldi’s book.
Rings
Ex. Let R={a, b, c, d, e}, and define + and
⋅
as follows.(R, +,
⋅
) is a commutative ring with unity, but without proper divisors of zero.zero: a.
unity: b.
units: b, c, d, e.
Every nonzero element has a multiplicative inverse.
Ex. Let R={s, t, v, w, x, y}, and define + and
⋅
as follows.(R, +,
⋅
) is a commutative ring with unity.zero: s.
unity: t.
units: t, y.
(R, +,
⋅
) is not an integral domain, because v⋅
w=s.(R, +,
⋅
) is not a field, because v, w and x have no multiplicative inverses.Also notice that v
⋅
v=v⋅
y=x, i.e., the cancellation law of multiplication does not hold for this example.
Proof of Theorem 14.10 in Grimaldi’s book.
• z∈S.
Let a=b ⇒ b+(−b)=z ∈ S.
• For each b∈S, −b∈S.
Let a=z ⇒ z+(−b)= −b ∈ S.
• a+b∈S for all a, b∈S.
a, −b∈S ⇒ a+(−(−b))= a+b ∈ S.
When S is finite, we assume S={s1, s2, …, sn}.
For every a∈S, {a+s1, a+s2, …, a+sn}=S.
⇒ a=a+sk =sk +a for some 1≤k≤n
⇒ sk =z∈S
⇒ z=a+sl =sl +a for some 1≤l≤n
⇒ sl = −a∈S.
For 1≤a<n,
gcd(a, n)=1 ⇒ [a]−1 exists (i.e., [a] is a unit of Zn);
gcd(a, n)>1 ⇒ [a] is a proper zero divisor of Zn.
Ex. Find [25]−1 in Z72 (gcd(25, 72)=1).
(−23)×25+8×72=1.
(−23)×25 ≡ 1 (mod 72).
[25]−1=[−23]=[49].
Ex. Find x so that 25x ≡ 3 (mod 72).
(−23)×25 ≡ 1 (mod 72).
3×(−23)×25 ≡ 3 (mod 72).
(−69)×25 ≡ 3 (mod 72).
x∈[−69]=[3].
Ex. gcd(8, 18)=gcd(2×4, 2×9)=2.
⇒ [8]
⋅
[9]=[0]⇒ [8] is a proper zero divisor of Z .
Ex. How many units and how many proper zero divisors are there in Z72?
The number of units in Z72 is equal to the number of integers a such that 1≤a<72 and gcd(a, 72)=1.
The latter can be computed as φ(72)=φ(23×32)=72×(1− 1
2)×(1− 1
3)=24, where φ(n) is the Euler’s phi function (refer to
Example 8.8 in page 394 of Grimaldi’s book).
The number of proper zero divisors in Z72 is equal to 71−24=47.
Ex. How many units and how many proper zero divisors are there in Z117?
The number of units in Z117 is equal to φ(117)=φ(32×13)=117×(1− 1
3)×(1− 1
13)=72.
The number of proper zero divisors in Z117 is equal to 116−72=44.
The Chinese Remainder Theorem
m1, m2, …, mk : positive integers that are greater than 1 and are prime to one another , where k≥2.
0≤ai ≤mi for 1≤i≤k.
Mi =m1…mi−1mi+1…mk for 1≤i≤k.
Mixi ≡1(mod mi) for 1≤i≤k.
Then, x=a1M1x1 +a2M2x2 + … +akMkxk is a solution to x≡ai (mod mi) for 1≤i≤k.
Moreover, if y is another solution, then y≡x(mod m1m2…mk).
The proof can be found in page 702 of Grimaldi’s book.
Ex. x≡14(mod 31), x≡16(mod 32), and x≡18(mod 33).
(a1, a2, a3)=(14, 16, 18); (m1, m2, m3)=(31, 32, 33);
(M1, M2, M3)=(1056, 1023, 992).
M1x1 ≡1(mod m1) (i.e., gcd(x1, m1)=1)
⇒ [x1]=[M1]−1=[1056]−1=[2]−1=[16] in
m1
Z =Z31. Similarly, [x2]=[31] in
m2
Z =Z32 and [x3]=[17] in
m3
Z =Z33.
Therefore, x = (a1M1x1 +a2M2x2 +a3M3x3) mod 32736
= 32688 is a solution.
The general solution is
y≡x(mod 32736).
In the study of cryptology, we often need to compute be mod n, where b, e and n are large integers.
Ex. Compute 5143 mod 222.
Let E(i)=5i mod 222, where i≥0 integer.
⇒ E(2i)=(E(i))2 mod 222.
143=27+23+22+21+20.
⇒ E(143)=(E(27)×E(23)×E(22)×E(21)×E(20)) mod 222.
E(20)=5.
E(21)=(E(20))2 mod 222=25.
E(22)=(E(21))2 mod 222=181.
E(23)=(E(22))2 mod 222=127.
E(24)=(E(23))2 mod 222=145.
E(25)=(E(24))2 mod 222=157.
E(26)=(E(25))2 mod 222=7.
E(27)=(E(26))2 mod 222=49.
E(28)=(E(27))2 mod 222=181.
Therefore, E(143) = (49×127×181×25×5) mod 222
= ((49×127) mod 222)×((181×25×5) mod 222) mod 222
= 7×203 mod 222
= 89.
f
(R, +,
⋅
) −−−−−−−−−−−−> (S, ⊕, ~)a f(a)
b f(b)
a+b f(a+b)=f(a)⊕f(b) a
⋅
b f(a⋅
b)=f(a)~f(b)Performing + (or
⋅
) in R and then mapping is “equivalent”to mapping and then performing ⊕ (or ~) in S.
(a) For any s∈S, there exist r∈R with f(r)=s (because f is onto).
s=f(r)=f(ruR)=f(r)f(uR)=sf(uR).
Similarly, s=f(uR)s.
⇒ f(uR) is the unity of S.
(b) Suppose b=a−1.
⇒ ab=ba=uR
f(a)f(b)=f(ab)=f(uR)=uS (uS is the unity of S) Similarly, f(b)f(a)=uS.
⇒ f(a)f(b)=f(b)f(a)=uS
⇒ (f(a−1)=) f(b)=[f(a)]−1
Ex. (following the discussion of the Chinese remainder theorem)
(Z32736, +,
⋅
) is a ring, where 32736=31×32×33.(Z31 ×Z32 ×Z33, ⊕, ~) is a ring, where
(x1, x2, x3)⊕(y1, y2, y3)=(x1 +y1, x2 +y2, x3 +y3) and (x1, x2, x3)~(y1, y2, y3)=(x1
⋅
y1, x2⋅
y2, x3⋅
y3).Define f: (Z32736, +,
⋅
) → (Z31 ×Z32 ×Z33, ⊕, ~) as follow:f(x)=(x1, x2, x3), where x1 =x mod 31, x2 =x mod 32, and x3 =x mod 33.
f is an isomorphism from Z32736 to Z31 ×Z32 ×Z33. (Refer to Example 14.21 in page 700 of Grimaldi’s book.)
18152
⋅
18153 in Z32736 can be computed as follows.18152
⋅
18153 = f−1f(18152⋅
18153)= f−1(f(18152)~f(18153)) = f−1((17, 8, 2)~(18, 9, 3))
= f−1(17×18 mod 31, 8×9 mod 32, 2×3 mod 33)
= f−1(27, 8, 6)
(refer to the example of the Chinese
remainder theorem)
= 25416
In general, if n=n1 ×n2 × … ×nk, where ni >1 is an integer and gcd(ni, nj)=1, then (Zn, +,
⋅
) and( × ×
n1
Z Zn2 …× , ⊕, ~) are isomorphic.
nk
Z
As a result, computation on large integers in Zn can be achieved with (parallel) computation on smaller integers in × ×
n1
Z Zn2 …× .
nk
Z
Groups
Ex. G = {π0, π1, π2, r1, r2, r3}, where
π0 = ⎟
⎠
⎜ ⎞
⎝
⎛
3 2 1
3 2
1 π1 = ⎟
⎠
⎜ ⎞
⎝
⎛
2 1 3
3 2
1 π2 = ⎟
⎠
⎜ ⎞
⎝
⎛
1 3 2
3 2 1
r1 = ⎟
⎠
⎜ ⎞
⎝
⎛
3 1 2
3 2
1 r2 = ⎟
⎠
⎜ ⎞
⎝
⎛
2 3 1
3 2
1 r3 = ⎟
⎠
⎜ ⎞
⎝
⎛
1 2 3
3 2
1 .
(G,
⋅
) is a (nonabelian) group.π1
⋅
r1 = ⎟⎠
⎜ ⎞
⎝
⎛
2 1 3
3 2
1 ⎟
⎠
⎜ ⎞
⎝
⎛
3 1 2
3 2
1 = ⎟
⎠
⎜ ⎞
⎝
⎛
1 2 3
3 2
1 = r3.