NowhereLog

Abstract Algebra 01 - The Introduction of Groups

December 25, 2020

This is the first post of the Introduction to Abstract Algebra series. In the first post, we introduce the basic concepts in abstract algebra, starting from mappings, equivalent relations, partitions, and extending to groups and homomorphisms. We will also encounter an important theorem: the Lagrange’s theorem.

Preliminaries

Definition 1.1: Cartesian Product

  • For two sets AA and BB, there Cartesian product is A×B={(a,b)aA,bB}A \times B = \{(a, b) \mid a \in A, b \in B\}.

Definition 1.2: Mapping

  • A mapping ff from the set AA to the set BB, denotes has f:ABf: A \to B, is an unambiguous association of an element in AA to an element in BB. Note that fA×Bf \subseteq A \times B.
  • AA is the domain of ff, BB is the codomain of ff, f(A)={f(a)aA}f(A) = \{f(a) \mid a \in A\} is the image/range of ff.
  • ff is onto/surjective if f(A)=Bf(A) = B.
  • ff is injective if a1,a2A,a1a2,f(a1)    f(a1)f(a2)\forall a_1, a_2 \in A, a_1 \neq a_2, f(a_1) \implies f(a_1) \neq f(a_2).
  • ff is bijective if ff is surjective and injective.

Definition 1.3: Equivalence Relation

  • Given a set AA, a relation RR is a subset of A×AA \times A, i.e., RA×AR \subseteq A \times A. For a1,a2Aa_1, a_2 \in A, if (a1,a2)R(a_1, a_2) \in R, put a1a2a_1 \sim a_2.
  • RR is is an equivalence relation if:

    1. (reflexivity) aa,aAa \sim a, \forall a \in A.
    2. (symmetry) ab    ba,a,bAa \sim b \implies b \sim a, \forall a, b \in A.
    3. (transitivity) ab,bc    ac,a,b,cAa \sim b, b \sim c \implies a \sim c, \forall a, b, c \in A.
  • For a set AA and an equivalence relation \sim on AA, [a]={xAxa}[a] = \{x \in A \mid x \sim a\} is an equivalence class. We define A/A/\sim as the set of equivalence classes.

Example 1.1: Set Z/nZ\mathbb Z/n\mathbb Z

An equivalence relation on Z\mathbb Z is R={(a,b)a,bZ,abmodn}R = \{(a, b) \mid a, b \in \mathbb Z, a \equiv b \mod n\}. This relation is called congruence modulo nn. The n1n-1 equivalence classes induced by congruence modulo nn, also referred to as congruence classes, form a set called Z/nZ\mathbb Z/ n \mathbb Z. For example, Z/2Z\mathbb Z/ 2 \mathbb Z contains two congruence classess: [0][0] and [1][1].

Definition 1.4: Paritition

  • Given a set AA, a partition is a set of non-empty subsets AiAA_i \subseteq A such that:

    1. iAi=A\cup_i A_i = A
    2. For iji \neq j, AiAj =A_i \cap A_j \ = \emptyset
  • Lemma: An equivalence relation A/A/\sim on AA is a partition of AA.

Groups

A group is a set and a binary operation. Intuitively, a group is simply some elements and a rule that describes how they interact.

Definition 1.5: Group

  • A group is a (G,)(G, \circ) is a set GG and a binary operation :G×GG\circ: G \times G \to G that satisfies the following axioms:

    1. (associativity) (ab)c=a(bc),a,b,cG(a \circ b) \circ c = a \circ (b \circ c), \forall a, b, c \in G.
    2. (identity) There exists an identity element eGe \in G such that ea=ae=a,aGe \circ a = a \circ e = a, \forall a \in G.
    3. (inverse) For each element aGa \in G, there exists an inverse a1Ga^{-1} \in G such that aa1=ea \circ a^{-1} = e. An equivalent statement of the three axioms is that a group is closed under multiplication and inverses. When the meaning is unambiguous, we often drop the binary operation and simply refers the group as GG.

Applying the three axioms, we can easily show several useful properties below.

Properties 1.6: Basic Properties of a Group

  • Given a group GG:

    1. The identity element is unique.x
    2. Every element has a unique inverse.
    3. If ab=ea \circ b = e, then ba=eb \circ a = e.
    4. (ab)1=b1a1(a \circ b)^{-1} = b^{-1} \circ a^{-1}.
    5. a11=ea^{-1^{-1}} = e.

Definition 1.7: Abelian Group A (G,)(G, \circ) is abelian if there for every two elements a,bGa, b \in G, ab=baa \circ b = b \circ a.

Definition 1.8: Order Given a group GG and a element gGg \in G, the order of gg, denoted as ord(g)ord(g) or g|g|, is the smallest kZ1k \in \mathbb Z_{\geq 1} such that gk=eg^k = e and gieg^i \neq e for i<ki < k, where gkg^k here means applying the binary operation on kk copies of gg. If such a kk does not exist, we say gg has infinite order.

Note that we also talk about the order of an group, which is the size of the underlying set.

Definition 1.9: Cyclic Group A group GG is cyclic if there exists a gGg \in G such that G={gkkZ}G = \{g^k \mid k \in \mathbb Z\}. We call gg the generator of GG. We say gg generates GG and write G=gG = \langle g \rangle. Obviously, ord(g)=Gord(g) = |G|.

Property 1.10: Properties of Cyclic Groups

Let GG be a finite cyclic group, G=xG = \langle x \rangle and G=n|G| = n. Then,

  1. For a positive integer kk, xk=n/gcd(n,k)|x^k| = n/gcd(n,k). This implies that the generators of GG has an order coprime with nn, so the number of generators of GG is ϕ(n)\phi(n), where again ϕ(n)\phi(n) is the Euler’s totient function.
  2. Every subgroup of GG is cyclic. Suppose HH is a subgroup of GG, and amHa^m \in H is the element of smallest mm in HH. Then for some anH,an=aqm+ra^n \in H, a^n = a^{qm+r} for some integers q,rZ,rmq, r\in \mathbb Z, r \leq m. Then ar=aqm+nHa^r = a^{-qm+n} \in H, but mm is the smallest order, so rmr \geq m. Thus, r=mr=m, and an=aqma^n = a^{qm}, which means H=amH = \langle a^m \rangle.
  3. For a positive integer kk, if knk \mid n, there exists a subgroup of GG of order kk which is xn/k\langle x^{n/k} \rangle.
  4. For a positive integer kk, xk=xgcd(n,k)\langle x^{k} \rangle = \langle x^{gcd(n,k)} \rangle.

We will define several important groups.

Example 1.2: Group Z/nZ\mathbb Z/n\mathbb Z

Recall from Example 1.1 that Z/nZ\mathbb Z/n \mathbb Z is a set of congruence classes modulo nn. We will introduce the group (Z/nZ,+)(\mathbb Z/n \mathbb Z, +). We define the binary operation ++ for two congruent classes as: [a]+[b]=[a+b][a]+[b] = [a+b], where the latter ”+” is our usual addition on numbers.

P.S. But to be more precise, the latter ”+” is not all separated form our new definitions. When n=0n=0, we get back a group that is isomorphic (for now understand it as equivalent) to Z\mathbb Z of integers.

You can check that this binary option is well-defined (unambiguous) and satisfies the above-mentioned axioms. For example, the identity element is clearly [0][0]. Also, Z/nZ\mathbb Z/n \mathbb Z is cyclic, and [1][1] is its generator.

Example 1.3: Group (Z/nZ)×(\mathbb Z/n\mathbb Z)^\times

First, we define the set (Z/nZ)×={[a]Z/nZgcd(a,n)=1}(\mathbb Z/n\mathbb Z)^\times = \{[a] \in \mathbb Z/n\mathbb Z \mid gcd(a,n)=1\}. In other words, (Z/nZ)×(\mathbb Z/n\mathbb Z)^\times contains the congruent classes that are coprime with nn. We define the binary operation ×\times as: [a]×[b]=[a×b][a] \times [b] = [a \times b], where the latter ”×\times” is our usual multiplication on numbers. It’s less trivial to show that ((Z/nZ)×,×)((\mathbb Z/n\mathbb Z)^\times, \times) is a group.

Property 1.11: The Euclidean Algorithm

Given a,bZ,x,yZa, b \in \mathbb Z, \exists x, y \in \mathbb Z, such that xa+yb=gcd(a,b)xa + yb = gcd(a,b).

ProofProof. This can be proven by the Euclidean algorithm.

Lemma 1.12: Group (Z/nZ)×(\mathbb Z/n\mathbb Z)^\times

ProofProof. It is straightforward to show that (Z/nZ)×(\mathbb Z/n\mathbb Z)^\times has the identity [1][1] and satisfies associativity. We will show every element in (Z/nZ)×(\mathbb Z/n\mathbb Z)^\times has an inverse by Property 1.8.

For a [k](Z/nZ)×,gcd(k,n)=1[k] \in (\mathbb Z/n\mathbb Z)^\times, gcd(k, n) = 1, so there exists x,yZx, y \in \mathbb Z such that xk+yn=gcd(k,n)=1xk + yn = gcd(k,n) = 1, or [x]×[k]=[1][x] \times [k] = [1]. So [x]=[k]1[x] = [k]^{-1}. Note that (Z/nZ)×=ϕ(n)|(\mathbb Z/n\mathbb Z)^\times| = \phi(n), where ϕ\phi is the Euler’s totient function, which counts numbers lesser than nn that are coprime with nn.

We will introduce other important groups like Sn,An,D2n,CnS_n, A_n, D_{2n}, C_n later.

Homomorphisms and Isomorphisms

Homomorphisms are mappings from groups to groups.

Definition 1.13: Homomorphism

A homomorphism from a group (G1,1)(G_1, \circ_1) to a group (G2,2)(G_2, \circ_2) is a map ϕ:G1G2\phi: G_1 \to G_2 such that ϕ(a1b)=ϕ(a)2ϕ(b)\phi(a \circ_1 b) = \phi(a) \circ_2 \phi(b).

ProofProof. By directly applying the definition of homomorphisms, we can show the following properties.

Property 1.14: Properties of Homomorphisms

Given a homomorphism ϕ:G1G2\phi: G_1 \to G_2,

  1. ϕ\phi must send e1G1e_1 \in G_1 to e2G2e_2 \in G_2, where e1,e2e_1, e_2 are identities.
  2. ϕ(a1)\phi(a^{-1}) = ϕ(a)1\phi(a)^{-1}

Definition 1.15: Isomorphisms

An isomorphism from a group (G1,1)(G_1, \circ_1) to a group (G2,2)(G_2, \circ_2) is a bijective homomorphism ϕ:G1G2\phi: G_1 \to G_2. We say G1G_1 and G2G_2 are isomorphic, denoted as G1G2G_1 \cong G_2.

Property 1.16: Properties of Isomorphisms

Given an isomorphism ϕ:G1G2\phi: G_1 \to G_2,

  1. |G1| = |G2|.
  2. G1 is cyclic    G2 is cyclicG_1 \text{ is cyclic} \iff G_2 \text{ is cyclic}.
  3. G1 is abelian    G2 is abelianG_1 \text{ is abelian} \iff G_2 \text{ is abelian}.
  4. If g1G1g_1 \in G_1 has order kk, ord(ϕ(g1))=kord(\phi(g_1)) = k.

Moreover,

  1. Any two cyclic groups of the same order are isomorphic. Just send the generator of one group to the generator of the other group.

The Lagrange’s Theorem

Definition 1.17: Subgroup Given a group (G,)(G, \circ), (H,)(H, \circ) is a subgroup of (G,)(G, \circ) if (H,)(H, \circ) is a well-defined group and HGH \subseteq G is a subset. {e}\{e\} and GG are trivial subgroups of GG, where {e}\{e\} is the identity. Again, we simply say HH is a subgroup of GG when there are no confusions on the binary operator.

Theorem 1.18: The Lagrange’s theorem

If GG is a finite group, and HGH \subset G is a subgroup. Then HG|H| \mid |G|.

Corollary: If GG is a finite group, and aGa \in G. Then ord(a)Gord(a) \mid |G|. This is by observing that aa generates a subgroup H=aH = \langle a \rangle which is of size ord(a)ord(a).

Corollary 1.19: The Fermat’s little theorem

If pp is a prime integer, aa is any integer, then apamodpa^p \equiv a \mod p.

ProofProof. We prove the Fermat’s little theorem by applying Lagrange’s Theorem to (Z/pZ)×(\mathbb Z/p\mathbb Z)^\times, where pp is a prime. We have that (Z/pZ)×=p1|(\mathbb Z/p\mathbb Z)^\times| = p-1, so by the Lagrange’s Theorem, for [a](Z/pZ)×,ord([a])p1[a] \in (\mathbb Z/p\mathbb Z)^\times, ord([a]) \mid p-1. Thus, we have [a]p1=[1]    [a]p=[a][a]^{p-1} = [1] \implies [a]^{p} = [a], or apamodpa^p \equiv a \mod p.

Corollary: In general, for nZ1,[a](Z/nZ)×,[a]ϕ(n)=[1]n \in \mathbb Z_{\geq 1}, [a] \in (\mathbb Z/n\mathbb Z)^\times, [a]^{\phi(n)} = [1].

Corollary 1.20: Prime Order Groups Are Cyclic

ProofProof. Given a group GG of order pp, where pp is a prime, for an aGa \in G, ord(a)pord(a) \mid p, so ord(a)=1 or pord(a) = 1 \text{ or } p. Since the identity is unique, we have p1p-1 generators.

Now we will prove the Lagrange’s theorem by defining cosets.

Definition 1.21: Coset

Given a group G,HGG, H \subset G is a subgroup. Define an equivalence relation on GG by g1,g2G,g1g2    g1H=g2H\forall g_1, g_2 \in G, g_1 \sim g_2 \iff g_1H = g_2H. Here for gG,gH={ghhH}g \in G, gH = \{gh \mid h \in H\}. We call gHgH a left coset of GG. Then G/G/\sim is the set of the equivalence classes, or the set of the left cosets. (Similarly, HgHg is a right coset).

ProofProof. By straightforwardly showing \sim is indeed an equivalence relation, we already have that G/G/\sim, or the set of cosets, is a partition of GG. But to make the idea more transparent, we will show this directly.

  1. We first note that every gGg \in G lies in some left coset. Namely, ggHg \in gH.
  2. We then note that g1H=g2H    g21g1H=H    g21g1H    g21g1=h for some hH    g1=g2hg_1H = g_2H \iff g_2^{-1}g_1 H = H \iff g_2^{-1}g_1 \in H \iff g_2^{-1}g_1 = h \text{ for some } h \in H \iff g_1 = g_2 h. The second     \iff can be shown by assuming that g21g1Gg_2^{-1}g_1 \notin G and noting that g21g1e=g21g1G    g21g1HHg_2^{-1}g_1 \circ e = g_2^{-1}g_1 \notin G \implies g_2^{-1}g_1H \neq H, a contradiction. This easy to prove property, gH=H    gHgH = H \iff g \in H, is a very useful property. Thus, if two left cosets g1Hg_1H intersects with g2Hg_2H at some element g3g_3, then g1H=g3Hg_1H = g_3H and g2H=g3Hg_2H = g_3H, so g1H=g2Hg_1H = g_2H and they are the same left coset.

ProofProof. Finally, we prove the Lagrange’s theorem by defining an isomorphism. Given a group GG, a subgroup HGH \subseteq G and any gGg \in G, define the mapping ϕ:HgH\phi: H \to gH as ϕ(h)=gh\phi(h) = gh. The mapping is clearly well-defined. It is injective since gh1=gh2    h1=h2gh_1 = gh_2 \iff h_1 = h_2. It is surjective since by definition, every element of gHgH is ghgh for some hHh \in H. Since HgHH \cong gH, H=gH|H| = |gH|. Thus we show that every left coset has the same cardinality as HH. Since G/G/\sim, the set of left cosets, is a partition of GG, we have that HG|H| \mid |G|.

Definition 1.22: Index

Given a group G,HGG, H \subset G is a subgroup, G/H|G|/|H|, the number of left cosets, is the index of HH in GG, denoted as [G:H][G:H].

References

[1] Siegel, Kyler. MATH GU4041 Introduction to Modern Algebra I. Department of Mathematics, Columbia University. 2019.

[2] Dummit, David and Foote, Richard. Abstract Algebra. Third Edition. John Wiley and Sons, Inc. 2004.


By NowhereMan who goes nowhere.


© 2024, NowhereLog by NowhereMan