Cyclic Subgroups Z mod nZ

Let's put all of these into practice. Let G=D8=r,s|r4=s2=1,rs=sr1={1,r,r2,r3,s,sr,sr2,sr3} be a Dihedral Groups. We made some Subgroup S=r={1,r,r2,r3}.

Subgroup or Subset?

All of these definitions allow for subsets, and don't restrict to subgroups; however, for this class you can expect we'll usually only consider subgroups.

Let's make a conjugation table for gxg1:

For the sake of simplicity let G={x1,,x8} as described in the order above (so x3=r2 for example. For each we compute gxg1 and find the corresponding xi in the table).

g is below, x is right x1 x2 x3 x4
1 x1 x2 x3 x4
r x1 x2 x3 x4
r2 x1 x2 x3 x4
r3 x1 x2 x3 x4
s x1 x4 x3 x2
sr x1 x4 x3 x2
sr2 x1 x4 x3 x2
sr3 x1 x4 x3 x2

For example:

sr2grx(sr2)1=sr2rr2s1=srs1=r3x4

Notice that this table alone tells you what happens to S, and helps you identify Stabilizers and Centralizers. Namely:

CD8(r)={gD8|g1g1=1,grg1=r,,gr3g1=r3}

Which means look at the rows where all the column entries don't change. We see that only for the first four rows (the g's), so then:

CD8(r)={1,r,r2,r3}

which convieniently happened to be the subgroupo r itself. For the Normalizer:

ND8(r)={gD8|g1g1r,,gr3g1r}

is a weaker condition (one can show that NC here). Here everyone is in that set (no one has an s in our table) so then:

=G=D8

Knowing the Subgroups

It'd be nice, when computing the Centralizer, if we can make it out of known subgroups. We want a picture of the subgroups of a group. Let's highlight with some tiny examples.

Look at Z2Z={0,1}. There's really only two subgroups, the trivial one and {0} (since they need to have the identity 0). Notice that:

{0}=0

and

{0,1}=1

Notice here that the latter includes the former, ie: 10. But this may not always be the case! For Z3Z then:

1=2=G

So again these subgroups include the trivial subgroup 0.

What about Z4Z? Notice that 1=3=G. But:

{0,2}=2

So we have some sort of tree of inclusions:

What about Z5Z? Here 1,2,3,4 generate the whole subgroups (I'm treating the numbers as their modulus equivalent, so 11), so there's no intermediate subgroups.

What about Z6Z?

You may notice that any generator's subgroup also is equivalent to the inverse generator's subgroup (ex: 2=4 as they are inverses).

How can we generalize this? Here's some facts we can try to prove for ZnZ (although we can prove this for any Cyclic Group of the same order, then use Isomorphism to show the properties still hold):

  1. For any aZ then |a|=ngcd(a,n) (see Order (Groups) for more info).
  2. Every subgroup of ZnZ is cyclic, so they are of the form a (so you only need one generator).
  3. a=bgcd(a,n)=gcd(b,n)|a|=|b|.
  4. There's exactly one unique subgroup for each divisor d of n, and it's nd.
  5. abb|a assuming a,b|n and a>b for simplicity.

To highlight these, consider Z12Z. The divisors of 12 are: 1,2,3,4,6,12. Then:

This picture we draw is a lattice of subgroups of our group Z12Z.