TheoremComplete

Equivalence Relations and Partitions

Equivalence relations are among the most important structures in mathematics, providing a way to identify objects that share essential properties. They formalize the notion of "sameness" and lead naturally to quotient structures.

DefinitionEquivalence Relation

A binary relation RR on a set AA (i.e., RβŠ†AΓ—AR \subseteq A \times A) is an equivalence relation if it satisfies three properties:

  1. Reflexive: βˆ€a∈A(aRa)\forall a \in A(aRa)
  2. Symmetric: βˆ€a,b∈A(aRbβ†’bRa)\forall a, b \in A(aRb \rightarrow bRa)
  3. Transitive: βˆ€a,b,c∈A(aRb∧bRcβ†’aRc)\forall a, b, c \in A(aRb \land bRc \rightarrow aRc)

We often write a∼ba \sim b instead of aRbaRb for equivalence relations.

Equivalence relations capture the idea of objects being "the same" in some sense, even if they are not literally equal. The three axioms ensure that this notion of sameness behaves as expected.

ExampleStandard Equivalence Relations
  1. Equality: The relation == on any set is an equivalence relation (the finest equivalence)
  2. Congruence modulo n: On Z\mathbb{Z}, define a∼ba \sim b iff n∣(aβˆ’b)n | (a - b). This is an equivalence relation
  3. Similarity: On the set of matrices, A∼BA \sim B iff there exists an invertible PP with B=Pβˆ’1APB = P^{-1}AP
  4. Cardinality: On the class of all sets, A∼BA \sim B iff ∣A∣=∣B∣|A| = |B| (i.e., there exists a bijection)
  5. Parallel lines: In Euclidean geometry, lines are equivalent if they are parallel (including the case where a line is parallel to itself)
DefinitionEquivalence Class

Let ∼\sim be an equivalence relation on AA. For a∈Aa \in A, the equivalence class of aa is:

[a]={b∈A:a∼b}[a] = \{b \in A : a \sim b\}

The set of all equivalence classes is called the quotient set and denoted A/∼A/\sim or A/RA/R:

A/∼={[a]:a∈A}A/\sim = \{[a] : a \in A\}

Equivalence classes partition the set: every element belongs to exactly one equivalence class, and two equivalence classes are either identical or disjoint.

TheoremFundamental Theorem of Equivalence Relations

Let ∼\sim be an equivalence relation on AA. Then:

  1. Every element a∈Aa \in A belongs to some equivalence class (its own): a∈[a]a \in [a]
  2. If a∼ba \sim b, then [a]=[b][a] = [b]
  3. If a≁ba \not\sim b, then [a]∩[b]=βˆ…[a] \cap [b] = \emptyset
  4. A=⋃a∈A[a]A = \bigcup_{a \in A} [a], i.e., the equivalence classes cover AA

Therefore, the equivalence classes form a partition of AA.

Proof

(1) follows from reflexivity: a∼aa \sim a, so a∈[a]a \in [a].

(2) Suppose a∼ba \sim b. We show [a]=[b][a] = [b] by double inclusion.

  • Let c∈[a]c \in [a], so a∼ca \sim c. By symmetry, b∼ab \sim a. By transitivity, b∼cb \sim c, so c∈[b]c \in [b]. Thus [a]βŠ†[b][a] \subseteq [b].
  • Similarly, [b]βŠ†[a][b] \subseteq [a]. Therefore [a]=[b][a] = [b].

(3) Suppose [a]∩[b]β‰ βˆ…[a] \cap [b] \neq \emptyset, so there exists c∈[a]∩[b]c \in [a] \cap [b]. Then a∼ca \sim c and b∼cb \sim c. By symmetry, c∼bc \sim b. By transitivity, a∼ba \sim b. This contradicts a≁ba \not\sim b.

(4) is immediate from (1).

β– 
DefinitionPartition

A partition of a set AA is a collection P={Ai:i∈I}\mathcal{P} = \{A_i : i \in I\} of non-empty subsets of AA such that:

  1. ⋃i∈IAi=A\bigcup_{i \in I} A_i = A (the sets cover AA)
  2. If iβ‰ ji \neq j, then Ai∩Aj=βˆ…A_i \cap A_j = \emptyset (the sets are pairwise disjoint)

The sets AiA_i are called the blocks or parts of the partition.

TheoremCorrespondence Between Equivalence Relations and Partitions

There is a bijective correspondence between equivalence relations on AA and partitions of AA:

  • Given an equivalence relation ∼\sim, the equivalence classes {[a]:a∈A}\{[a] : a \in A\} form a partition
  • Given a partition P\mathcal{P}, define a∼ba \sim b iff aa and bb belong to the same block. This is an equivalence relation

These constructions are inverse to each other.

ExampleQuotient Structures
  1. Integers modulo n: The quotient Z/nZ\mathbb{Z}/n\mathbb{Z} consists of nn equivalence classes: [0],[1],…,[nβˆ’1][0], [1], \ldots, [n-1]. We can define addition: [a]+[b]=[a+b][a] + [b] = [a+b], giving the ring Zn\mathbb{Z}_n

  2. Rational numbers: We construct Q\mathbb{Q} from ZΓ—(Zβˆ–{0})\mathbb{Z} \times (\mathbb{Z} \setminus \{0\}) by the equivalence (a,b)∼(c,d)(a, b) \sim (c, d) iff ad=bcad = bc. The equivalence class of (a,b)(a, b) is denoted a/ba/b

  3. Projective space: In projective geometry, points in Pn\mathbb{P}^n are equivalence classes of (n+1)(n+1)-tuples under the relation (x0,…,xn)∼(Ξ»x0,…,Ξ»xn)(x_0, \ldots, x_n) \sim (\lambda x_0, \ldots, \lambda x_n) for Ξ»β‰ 0\lambda \neq 0

  4. Homeomorphism: In topology, we often study spaces "up to homeomorphism," treating homeomorphic spaces as equivalent

DefinitionCanonical Projection

Given an equivalence relation ∼\sim on AA, the canonical projection (or quotient map) is the function:

Ο€:Aβ†’A/∼,a↦[a]\pi: A \to A/\sim, \quad a \mapsto [a]

This function is always surjective. It "collapses" equivalent elements to a single representative.

TheoremUniversal Property of Quotients

Let ∼\sim be an equivalence relation on AA, and let f:Aβ†’Bf: A \to B be a function that is compatible with ∼\sim (meaning a∼aβ€²a \sim a' implies f(a)=f(aβ€²)f(a) = f(a')). Then there exists a unique function fΛ‰:A/βˆΌβ†’B\bar{f}: A/\sim \to B such that f=fΛ‰βˆ˜Ο€f = \bar{f} \circ \pi:

Aβ†’fB↓π↗fΛ‰A/∼\begin{array}{ccc} A & \xrightarrow{f} & B \\ \downarrow{\pi} & \nearrow{\bar{f}} & \\ A/\sim & & \end{array}

The function fˉ\bar{f} is defined by fˉ([a])=f(a)\bar{f}([a]) = f(a), and compatibility ensures this is well-defined.

Remark

This universal property is fundamental in category theory and abstract algebra. It says that compatible functions through AA can be "factored" through the quotient A/∼A/\sim. This construction is ubiquitous: group homomorphisms factor through normal subgroups, continuous maps factor through quotient topologies, etc.

ExampleFirst Isomorphism Theorem

In group theory, if Ο†:Gβ†’H\varphi: G \to H is a homomorphism, we can define an equivalence relation on GG by g1∼g2g_1 \sim g_2 iff Ο†(g1)=Ο†(g2)\varphi(g_1) = \varphi(g_2). The equivalence classes are the cosets of ker⁑(Ο†)\ker(\varphi), and the quotient G/ker⁑(Ο†)G/\ker(\varphi) is isomorphic to im(Ο†)\text{im}(\varphi) via the map Ο†Λ‰([g])=Ο†(g)\bar{\varphi}([g]) = \varphi(g).

This is the first isomorphism theorem, which holds analogously for rings, modules, vector spaces, and many other algebraic structures. The underlying set-theoretic machinery is the universal property of quotients.

Equivalence relations and quotients are among the most powerful tools in mathematics, allowing us to study structure by ignoring irrelevant distinctions. They appear everywhere from number theory to topology to category theory, providing a uniform language for discussing "identification" and "modding out."