MATH 415 Lecture 6

From Notes
Jump to navigation Jump to search

« previous | Thursday, September 12, 2013 | next »


Cayley's Theorem

Let G be a group. Choose element gG and consider ψg:GG such that ψg(h)=gh.

Lemma. ψg is a permutation of G.

Proof. ψg(h)=ψg(k) for some h,kG. This implies gh=gk and nhus h=k. Therefore ψg is injective.

Given

hG

,

ψg(g1h)=g(g1h)=(gg1)h=h

. Thus

ψg

is surjective.

quod erat demonstrandum


If f:AB is a function and HA, then the "image of H under f" is the set

f[H]={f(h)hH}B


Lemma 8.15

Lemma. Let G,G be two groups and ϕ:GG be a group homomorphism which is injective. Then

  1. ϕ[G] is a subgroup of G, and
  2. ϕ gives an isomorphism between G and ϕ[G].

Proof. First: why is ϕ[G] a subgroup? If x,yϕ[G], then x=ϕ(x) and y=ϕ(y) for some x,yG. If we take xy, then we have ϕ(x)ϕ(y)=ϕ(x,y) because it is a homomorphism. Therefore xyϕ[G].

If xϕ[G], then x=ϕ(x) for some xG. (x)1=(ϕ(x))1=ϕ(x1), so (x)1ϕ[G]. Therefore ϕ[G] is a subgroup of G.

Therefore

ϕ:Gϕ[G]

is an isomorphism.

quod erat demonstrandum


Theorem 8.16 (Cayley's Theorem)

Theorem. Every group is isomorphic to a group of permutations.

Proof. Let G be a group and consider SG (all permutations of G).

Define γ:GSG such that γ(g)=ψg, where ψg(h)=gh (from above). According to the first lemma, ψgSG, so γ is well-defined.

We claim that γ is injective.

If γ(g1)=γ(g2), then ψg1=ψg2. We find ψg1(e)=ψg2(e), where e is the identity element of G. Thus g1e=g2e and g1=g2. Thus γ is injective.

We claim that γ is a homomorphism.

γ(g1g2)=ψg1g2=g1g2h=ψg1ψg2 for all hG. Therefore γ is a homomorphism.

By Lemma 8.15,

Gγ[G]SG

.

quod erat demonstrandum


Section 9: Orbits, Cycles, and Alternating Groups

Orbits

Suppose σ is a permutation of a set A. σ naturally partitions A into equivalence classes. If a,bA, then

aBb=σn(a) for some n

We claim that is an equivalence relation:

Proof.:

  • It is reflexive since a=σ0(a).
  • It is symmetric since a=σn(b) (i.e. ba) for all b=σn(a) (i.e. ab)
  • It is transitive since b=σn(a) and c=σm(b) imply that c=σm(b)=σm(σn(a))=σn+m(a)=ac.
    quod erat demonstrandum


Let σ be a permutation of A. The equivalence classes of are called orbits of σ.

For example, let A={1,2,3,4,5,6,7,8}. If we take σ=e (identity permutation), then the orbits of σ are {1}, {2}, ..., {8}.

If we take σ=(1234567838674152), then the orbits of σ are {1,3,6},{2,8},{4,5,7}.


Cycles

Each orbit may be represented as its own permutation. For example, from above, the orbits can be written as:

μ1=(1234567832645178)μ2=(1234567818345672)μ3=(1234567812374658)

Observe that σ=μ1μ2μ3

A permutation σSn is called a cycle if it has at most one orbit containing more than one element. The length of a cycle is the number of elements in its largest orbit.

Let's find the orbits and lengths of each μ above:

  1. {1,3,6}, {2}, {4}, {5}, {7}, {8} (len = 3)
  2. {1}, {2,8}, {3}, {4}, {5}, {6}, {7} (len = 2)
  3. {1}, {2}, {3}, {4,7,5}, {6}, {8} (len = 3)

Convenient Notation for Cycles

Instead of writing out a permutation for each μ, we can use the following shortand:

μ1=(1,3,6)μ2=(2,8)μ3=(4,7,5)

By this we mean that (in μ1), 1 maps to 3, 3 maps to 6, and 6 maps back to 1. All other elements are understood to be identity-mapped.

Thus σ=(1,3,6)(2,8)(4,7,5)

Note that we wrote an arbitrary permutation as a product/composition of disjoint cycles, which leads to our next theorem.

Note: Composition of disjoint cycles is commutative. This is not true if they are not disjoint.

(1,4,5,6)(2,1,5)(2,1,5)(1,4,5,6)

  • On the left, 1 maps to 5 and then 5 maps to 6: 1 → 6
  • On the right, 1 maps to 4 and then 4 is id-mapped: 1 → 4


Theorem 9.8

Theorem. Every permutation σ of a finite set is a product of disjoint cycles.

Proof. Let B1,B2,,Br be the orbits of σ. Define a cycle μi as follows:

μi(x)={σ(x)xBixotherwise

Then clearly

σ=μ1μ2μr

.

quod erat demonstrandum


For example, write σ=(123456652431) as a product of disjoint cycles:

σ=(1,6)(2,5,3)


Transposition

A cycle of length 2 is called a transposition

Observation: (a1,a2,,an)=(a1,an)(a1,an1)(a1,a2)

Corollary. Any permutation of a finite set |A|2 is a product of transpositions.

for example, write the σ=(1,6)(2,5,3) above as a product of transpositions.

σ=(1,6)(2,3)(2,5)

Note that this representation is not unique because (1,6)(2,3)(2,5)(3,4)(4,3) is equivalent (we change 4 to 3 and then back to 4).


Theorem 9.15

Theorem. No permutation in Sn can be expressed both as a product of an even number of transpositions and as a product of an odd number of transpositions.

Proof (sketch).

detI=1

, but switching any two rows negates the determinant. We can think of this switch as a transposition. If we assume the negation of the theorem, then we can determine that

detσ(I)=1

and

detσ(I)=1

, which is a contradiction.

quod erat demonstrandum

This theorem can be applied to the shifting tiles puzzle. The following board is not solvable since all solutions have an even parity.

1 2 3 4 5
6 7 8 9 10
11 12 13 14 15
16 17 18 19 20
21 22 24 23  

Alternating Groups

Let An be the subset of Sn of all even permutations and Bn be the set of all odd permutations.

Lemma. |An|=|Bn|=|Sn|2=n!2

Proof. Let ϕ:AnBn by ϕ(σ)=(1,2)σ.

We need to show that ϕ is bijective.

Theorem 9.20

Theorem. If n2 then An is a subgroup of order n!2.

Proof. Even + even = even; the set of even numbers is closed under addition

quod erat demonstrandum

An is called the alternating group