Set Partition

Context: FIT1058_MOC · splits a set into disjoint covering parts · each element in exactly one part · counted by Bell numbers

Quick Revision

  • 🎯 Objective: carve into nonempty disjoint parts whose union is ➔ each element in exactly one part.
  • 📦 Core Components: nonempty ➔ pairwise disjoint ➔ covering ➔ coarsest/finest.
  • ⚡ Key Constraint: counted by Bell numbers ; correspond to equivalence relations.

📝 Core

1. The Partition (Three Conditions)

  • Definition ➔ a set of nonempty, pairwise disjoint subsets (parts/blocks) of whose union is .
  • One-line equivalent ➔ every element lies in exactly one part.
  • Extremes ➔ coarsest = single part ; finest = all singletons.

2. Why Each Condition

  • Nonempty ➔ no block.
  • Disjoint ➔ no element in two parts.
  • Covering ➔ union equals (nothing left out).

3. Counting (Bell Numbers)

  • ➔ number of partitions of an -set.
  • Sequence ().

Key identities:

⚖️ Core Decision Matrix

ConceptRestrictionCount
partition of -set3 conditionsBell
Power Setany subset collection
coarsest partitionone block1
finest partitionall singletons1

When It Flips: partitions correspond exactly to equivalence relations on (each part = an equivalence class). A partition is a restricted subcollection of , so but still grows super-exponentially.

📊 Exam Execution Trace

Manual Execution Trace

Partitions of by block count:

Step / StateBlocksPartition
0 (Init)
11
22
33

⚠️ Common Mistakes

  • 💡 Each element in exactly one part fails (1); fails (2); fails (3).

🧠 Active Recall