Set TheoryTopic #13 of 40

Partial Orders

Posets, Hasse diagrams, lattices, total orders, well-ordering, and topological sorting.

Definition

A relation RR on a set AA is a partial order if it is:

  1. Reflexive: ∀a∈A:aRa\forall a \in A: aRa
  2. Antisymmetric: ∀a,b∈A:(aRb∧bRa)→a=b\forall a, b \in A: (aRb \land bRa) \rightarrow a = b
  3. Transitive: ∀a,b,c∈A:(aRb∧bRc)→aRc\forall a, b, c \in A: (aRb \land bRc) \rightarrow aRc

A set with a partial order is called a partially ordered set or poset, written (A,⪯)(A, \preceq).

Examples

Divisibility on Z+\mathbb{Z}^+

a∣ba \mid b (a divides b) is a partial order.

Subset Relation

A⊆BA \subseteq B on a power set P(S)\mathcal{P}(S).

Less Than or Equal (≤\leq)

≤\leq on R\mathbb{R} is a partial order (actually a total order).

Lexicographic Order

For strings: compare character by character.

Comparability

In a poset (A,⪯)(A, \preceq):

  • aa and bb are comparable if a⪯ba \preceq b or b⪯ab \preceq a
  • aa and bb are incomparable if neither a⪯ba \preceq b nor b⪯ab \preceq a

In (P({1,2,3}),⊆)(\mathcal{P}(\{1,2,3\}), \subseteq):

  • {1}\{1\} and {1,2}\{1, 2\} are comparable ({1}⊆{1,2}\{1\} \subseteq \{1,2\})
  • {1}\{1\} and {2}\{2\} are incomparable

Total Orders

A partial order is a total order (or linear order) if every pair is comparable: ∀a,b∈A:a⪯b∨b⪯a\forall a, b \in A: a \preceq b \lor b \preceq a

Examples:

  • ≤\leq on R\mathbb{R} — total order
  • ⊆\subseteq on P(S)\mathcal{P}(S) — NOT total (unless ∣S∣≤1|S| \leq 1)

Hasse Diagrams

A visual representation of a finite poset:

  1. Draw elements as points
  2. Draw line from aa (lower) to bb (higher) if a≺ba \prec b with no element between
  3. Omit reflexive loops and transitive edges

Example: Divisibility on {1,2,3,4,6,12}\{1, 2, 3, 4, 6, 12\}

        12
       /  \
      4    6
      |   /|
      2  3 |
       \ | /
         1

Special Elements

Maximal and Minimal

  • aa is maximal if ¬∃b∈A:a≺b\neg\exists b \in A: a \prec b (nothing above it)
  • aa is minimal if ¬∃b∈A:b≺a\neg\exists b \in A: b \prec a (nothing below it)

A poset can have multiple maximal/minimal elements.

Greatest and Least

  • aa is the greatest element (maximum) if ∀b∈A:b⪯a\forall b \in A: b \preceq a
  • aa is the least element (minimum) if ∀b∈A:a⪯b\forall b \in A: a \preceq b

A poset has at most one greatest/least element.

Upper and Lower Bounds

For subset S⊆AS \subseteq A:

  • uu is an upper bound of SS if ∀s∈S:s⪯u\forall s \in S: s \preceq u
  • ll is a lower bound of SS if ∀s∈S:l⪯s\forall s \in S: l \preceq s

Least Upper Bound (LUB) / Supremum

The least upper bound of SS, written sup⁡(S)\sup(S) or lub(S)\text{lub}(S), is the smallest upper bound.

Greatest Lower Bound (GLB) / Infimum

The greatest lower bound of SS, written inf⁡(S)\inf(S) or glb(S)\text{glb}(S), is the largest lower bound.

Lattices

A poset (L,⪯)(L, \preceq) is a lattice if every pair of elements has:

  • A least upper bound (join): a∨b=lub({a,b})a \lor b = \text{lub}(\{a, b\})
  • A greatest lower bound (meet): a∧b=glb({a,b})a \land b = \text{glb}(\{a, b\})

Examples

  • (P(S),⊆)(\mathcal{P}(S), \subseteq) with A∨B=A∪BA \lor B = A \cup B and A∧B=A∩BA \land B = A \cap B
  • (Z+,∣)(\mathbb{Z}^+, \mid) with a∨b=lcm(a,b)a \lor b = \text{lcm}(a,b) and a∧b=gcd⁡(a,b)a \land b = \gcd(a,b)

Well-Ordering

A poset (A,⪯)(A, \preceq) is well-ordered if:

  1. It is a total order
  2. Every non-empty subset has a least element

Examples

  • (N,≤)(\mathbb{N}, \leq) — well-ordered
  • (Z,≤)(\mathbb{Z}, \leq) — NOT well-ordered (no least element)
  • (Q+,≤)(\mathbb{Q}^+, \leq) — NOT well-ordered ({x∈Q∣x>0}\{x \in \mathbb{Q} \mid x > 0\} has no least element)

Well-Ordering Theorem

Every set can be well-ordered (equivalent to Axiom of Choice).

Topological Sorting

Given a partial order on a finite set, a topological sort is a total order consistent with the partial order.

Algorithm

  1. Find a minimal element
  2. Remove it and add to sorted list
  3. Repeat until empty

Example

For tasks with dependencies, topological sort gives a valid execution order.

Chains and Antichains

  • A chain is a totally ordered subset
  • An antichain is a set of pairwise incomparable elements

Dilworth's Theorem

In a finite poset, the minimum number of chains needed to cover all elements equals the maximum size of an antichain.