Skip to content

03 Ordering Relations

Ordering Relations & Lattices

Partially-ordered sets(POSETS)

A relation R defined on a set S is said to be partially ordered if it is

  1. reflexive
  2. Anti-symmetric (uni-directional)
  3. Transitive

Then (S, R) is a POSET.

Eg: \((\mathbb{Z}, \le), (\mathbb{Z}, \ge), (\mathbb{Z}^+ , /), (P(S) , \subseteq)\)

Notes

  1. Any partial ordering relation is denoted by \(\preceq\)

  2. if \(a \prec b\) denotes that \(a \preceq b\) but \(a \ne b\)

  3. if R is a partial order relation on S, the R-1 is also a partial order relation on S, where \(R^{-1} = \{ (b, a) | (a,b) \in R \}\) (S, R-1) is called the dual of (S, R)

  4. Let\(a, b \in S\) where \((S, \preceq)\) is a POSET. a and b are said to be comparable, if either \(a \preceq b\) or \(b \preceq a\) otherwise a and b are not comparable

Eg: \((\mathbb{Z}^+, /)\) (2, 10) , (16, 8) is comparable; but (2, 7) isn't comparable as neither 2 nor 7 can divide each other

Totally ordered Set

\(\preceq\) is a totally-ordered relation if every 2 elements of S is comparable, and S is a totally-ordered set

Eg: \((Z, \le), (D_8,/)\) (divisors of 8)

Not totally ordered

eg: \((Z^+, /), (D_{12}, /)\) (divisors of 12)

Dn: set of all positive divisors of positive integer n it is always a POSET, but not necessarily a TOSET

Well Ordered Set

A relation that is

  1. totally ordered
  2. every subset of S has least element in the Hasse diagram doesn't have to be the least mathematically, such as the case of \((Z^-, \ge)\)

WOSET \(\implies\) TOSET not all TOSETs are WOSETs, but all finite TOSETs are

Eg: \((N, \le), (Z^+, \le), (Z^-, \ge)\) \((Z^-, \le)\) is TOSET but not WOSET, as there is no subset (\(- \infin\) is the least element)

POSET/HASSE Diagram

  1. draw from lower to upper direction
  2. no loops
  3. eliminate edges that are implied by transitiveness
  4. no arrows

Elements

Minimal elements

indegree = 0 (excluding self)

a is a minimal element in S if there is no \(b \in S\) such that \(b \preceq a\)

Maximal elements

outdegree = 0 (excluding self)

a is a maximal element in S if there is no \(b \in S\) such that \(a \preceq b\)

Least element

Element a is called least element of S, if \(a \preceq b, \forall b \in S\)

The lowermost element of Hasse diagram

It has to be unique - ie the only minimal element

Greatest Element

Element a is called greatest element of S if \(b \preceq a, \forall b\in S\)

The uppermost element of Hasse diagram

It has to be unique - ie the only maximal element

Summary

Minimal and maximal points are end points that are related to all elements of the question set B, while least and greatest point are unique end point related to all elements of the question set B

Bounds

Let A be a subset of S

Bound is a set of the above elements

Upper bound

If u is an element of S such that \(a \preceq u, \forall a \in A\), then u is called upper bound of A

should be related to both a and b

LUB/Supremum

Least upper bound

L==U==B/S==u==premum

Lower bound

If \(l\) is an element of S such that \(l \preceq a, \forall a \in A\), then \(l\) is called lower bound of A

should be related to both a and b

GLB/Infimum

Greatest lower bound

Lattice

If the HASSE diagram starts and ends with a single point, it's called as a lattice.

A lattice is a POSET \((S, \preceq)\) in which each pair of elements has

  1. LUB LUB of a and b: \(a \lor b\) (join of {a, b}) -> sup(a, b)
  2. GLB GUB of a and b: \(a \land b\) (meet of {a,b}) -> inf(a,b)

eg: \((D_6, /)\)

Examples

\[ \left(P(S), \subseteq\right)\\ A, B \in P(S): A \lor B = A \cup B, A \land B = A \cap B \]
\[ \left(P(S), \supseteq \right)\\ A, B \in P(S): A \lor B = A \cap B, A \land B = A \cup B \]
\[ \left(P(S), \le\right)\\ A, B \in P(S): A \lor B = \text{max}(A,B), A \land B = \text{min}(A,B) \]
\[ \left(D_n, / \right)\\ A, B \in D_n: A \lor B = \text{lcf}(A,B), A \land B = \text{hcf}(A,B) \]

Semi-Lattice

Join Semi-Lattice

Lattice with only LUB

multiple starting points

Eg: \(( \{2, 3, 60,180\}, /)\)

Meet Semi-Lattice

Lattice with only GLB

multiple ending points

Eg: \((I_{12}, /)\)

Properties of Lattices

Let \((L, \lor, \land)\) be an algebraic system defined by lattice \((L, \preceq)\)

  1. Idempotency
  2. \(a \land a = a\)
  3. \(a \lor a = a\)
  4. Commutative
  5. \(a \land b = b \and a\)
  6. \(a \lor b = b \lor a\)
  7. Associative
  8. \((a \land b) \land c = a \land (b \land c)\)
  9. \((a \lor b) \lor c = a \lor (b \lor c)\)
  10. Absorption (opposite operation)
  11. \(a \land (a \lor b) = a\)
  12. \(a \lor (a \land b) = a\)
  13. Distributive (not all lattices are distributable)
  14. \(a \land (b \lor c) = (a \land b) \lor (a \land c)\)
  15. \(a \lor (b \land c) = (a \lor b) \land (a \lor c)\)
  16. Consistency \(a \land b = a \text{ and } a \lor b = b\)
Last Updated: 2023-01-25 ; Contributors: AhmedThahir

Comments