Proposition 17: Consider setsA, B,DandE. Then:
(A⇥B)\(D⇥E) = (A\D)⇥(B\E).
Proposition 18: De Morgan’s Laws a. AC[BC= (A\B)C
• or: (AC[BC)C=A\B b. AC\BC= (A[B)C
• or: (AC\BC)C=A[B
7.3 Proof Techniques for Statements about Sets
To prove setsA=B then show(A⇢B)^(B⇢A).
• i.e. Prove all elements inAare inBand vice versa.
To proveA*B:
• Then9y2As.t. y2Aandy /2B.
7.4 Cartesian Products
Given setsA, B, ..., Z, their cartesian product is the set of ordered tuples:
A⇥B⇥...⇥Z={(a, b, ..., z)|a2A, b2B, ..., z2Z}.
• Important: Order matters.
• A⇥B6=B⇥A.
• Order matters forA⇥A, ifx, y2Aandx6=y, then(x, y)6= (y, x).
ConsiderR2:
• A⇥P ={(a, p)2R2|a2A, p2P}
• B⇥Q={(b, q)2R2|b2B, q2Q}
12
Consider the cartesian product of a setS with itself repeatedktimes,Sk: Sk= S⇥S⇥...⇥S
| {z }
k times ={(s1, s2, ..., sk)|si2S, 8i2{1,2, ..., k}}
• The elements= (s1, s2, ..., sk)is a vector.
• Eachsi2S fori2{1, ..., k}is a co-ordinate of vectors.
• The diagonal of setSkis the following subset ofSk;{(x, x, ..., x)2Sk|x2 S}.
– E.g. If A = {1,2,3}, then the diagonal of A3 = A⇥A⇥A = {(a, a, a)2A3|a2A}={(1,1,1),(2,2,2),(3,3,3)}.
7.5 Euclidean Spaces
An example of a cartesian product is the Euclidean spaceRn=R⇥R⇥...⇥R for somen2N⇤.
• ConsiderRn, each elementx2Rnis a point or a vector of spaceRn.
• Every vectorx2Rnis written as: x= (x1, x2, ..., xn)2Rn. Vector operations:
• (Summation) Let vectorsx, y2Rn, thenx+y= (x1+y1, x2+y2, ..., xn+ yn).
• ( x) Let 2Randx2Rn, then x= ( x1, x2, ..., xn).
Canonical vectors: Considerx2Rn, the vectorxcan be decomposed as:
x=Pn
i=1xiei=x1e1+...+xnen
whereei2Rnbut are 0’s all but theithterm equals 1, the canonical vector.
• Canonical decomposition: The decompositionx=x1e1+...+xnen. Definition: x >> y () xi> yi8i2{1, ..., n}
Definition: x > y () (xi yi)^(9xk> yk)
• ‘at least one greater than’
Definition: x y () (xi yi)
• It is possible thatxi=yi8i={1,2, ..., n}.
13
7.6 Power Sets
Let A be a set. Then the power set, denoted 2A is defined as the set of all subsets ofA.
• If setAhask elements, then the power set2Ahas2k elements.
8 Partition of a Set
Definition: Partition of a set
A (finite) partition of a non-empty setS is a finite collection of non-empty subsetsB1, B2, ..., Bk⇢S, for somek2N⇤, such that:
Bi\Bj=;, ifi6=j and
S=Sk i=1Bi.
• B={B1, B2, ..., Bk}is a partition ofS.
• B is a set whose elements are subsets decomposing setS.
Definition: Refinement
Consider two partitionsBandCof a non-empty setS. If for every element Ci2C, there are elementsBj1, ..., Bjs2Bsuch that:
Ci=Ss r=1Bjr
thenBrefinesCor equivalentlyBis a refinement ofC.
• IfB refinesCthenCis a coarsening ofB.
• Any partitionB ofSrefines itself.
• IfBrefinesCandB refinesD, thenB is a common refinement ofCand D.
• IfBrefinesCand they are different partitions, thenB is a proper refine- ment ofC.
Example 29: SupposeS ={1,2,3,4,5}. Consider the following 3 partitions of S: R=P={{1,2},{3,4,5}}andQ={{1},{2},{3,4,5}}.
• Qis a proper refinement ofRandP.
• Qis a common refinement ofRandP.
• RandP are proper coarsening ofQ.
• P is a refinement ofR, but not a proper refinement.
• P is a coarsening ofR, but not a proper coarsening.
• P is a common coarsening ofRandQ.
14