2

The group $\langle (\mathbb Z/9)^\times,\times\rangle$ has elements given by: $\{1, 2, 4, 5, 7, 8\},$ and their orders are:

$o(1)= 1;$
$o(2)=|\{2,4,8, 7, 5, 1\}| = 6;$
$o(4)=|\{4,7,1\}|= 3;$
$o(5)=|\{5, 7, 8, 4, 2, 1\}| = 6;$
$o(7)= |\{7, 4, 1\}|= 3; $
$o(8)=|\{8, 1\}|= 2.$

The cyclic group $\left<\mathbb Z/6\mathbb Z, +\right>$ has the set of elements $=\{0,1,2,3,4,5\},$ with their order as:

$o(0)= 1;$
$ o(1)= |\{1,2,3,4,5,6\}|= 6;$
$ o(2)= |\{2,4,0\}|=3;$
$ o(3)=|\{3,0\}|= 2;$
$ o(4)=|\{4, 2, 0\}|= 3;$
$ o(5)=|\{5,4,3,2,1,0\}|=6.$

Hence, both have two generators, two elements of order $3,$ one element each of order $2,$ and $1.$

The number of Isomorphisms possible are: $2,$ given by :

1.$\varphi(2)=1; \varphi(5)=5; \varphi(1)= 0; \varphi(8)=3; $

2.$\xi(2)=5; \xi(5)=1; \xi(1)= 0; \xi(8)=3; $

Consider the case of $\varphi$:

As, given two cyclic groups of order six; so, using any generator's map to the co-domain, should be able to find the map for $\varphi(4), \varphi(7).$

But, am unable to do this for finding the map of $\varphi(7),$ as shown below:

$\varphi(4)= \varphi(2\cdot 2)= 1+1= 2.$

Had, chosen $\xi$, with $\xi(2)=5$ then $\xi(4)=\xi(2\cdot 2)= 5+5=10\equiv 4\pmod 6.$

But, how to find $\varphi(7),$ is unclear as the prime term $'7'$ inside the brackets should be expressed as product of smaller terms, and have to use $\varphi(2).$

jiten
  • 4,960
  • @Bumblebee But, request to provide the logic why that is applicable. Also, you seem to agree that $\varphi(7),$ cannot be found directly. – jiten Oct 27 '22 at 03:47
  • When you write : "$\xi(4)=\xi(2\cdot 2)= 5+5=10$", you make the usual abuse of notations in $\mathbb Z/6\mathbb Z$. Example : "$10$" means $10+6\mathbb Z$: it is a class of numbers and you know it. So you can't write "$10\equiv4 \pmod 6$", can you? – Stéphane Jaouen Oct 27 '22 at 06:28

1 Answers1

4

You noted (sort of), that $7\equiv 2^4\pmod9 $, so knowing $\varphi (2)$ makes it easy to get $\varphi (7)$. Namely it's $\varphi (7)=\varphi (2)^4=4\varphi (2)$. So you get the values under each isomorphism. That's the point of the statement "the (iso)morphism is determined by generators". It's actually true even when you have more than one generator. But the case of a cyclic group is particularly simple: one generator.

Similarly, since $4\equiv 2^2\pmod9 $, we have $\varphi (4)=\varphi (2)^2=2\varphi (2)$. You can use either generator to do this. There's $2$ because Euler's totient function evaluated at $6$ is $2$. In general for a cyclic group there are $\phi(n)$ generators. It's also the number of isomorphisms, and (of course, automorphisms) because generators go to generators.

  • 3
    $\varphi(2)^4$ is bad notation, since $\varphi(2)$ lives in an additive group. Should be $4\varphi(2)$. – Arturo Magidin Oct 27 '22 at 04:59
  • That's true. I almost pointed that out. There's, at least borderline, abuse of notation. You and I can follow, the OP and others are a different story. I looked at it, and it's not the only confusing point. I say two generators and at another point one, when of course I mean "singly generated". @ArturoMagidin – suckling pig Oct 27 '22 at 05:04
  • @codeofsilence Please tell my error, as am confused by your last comment. – jiten Oct 27 '22 at 05:20
  • @codeofsilence Also, request to provide a similar problem (if possible), on isomorphism; where the group's generating set has at least two generators. – jiten Oct 27 '22 at 05:23
  • There's no error that I have in mind, it's just that it's sure to be confusing at this point. – suckling pig Oct 27 '22 at 05:29
  • See this: https://math.stackexchange.com/a/4076175/1070376 Note that just the targets for the generators are specified. – suckling pig Oct 27 '22 at 05:32
  • Please tell what is confusing. Also, thanks for the link. But, seems couldn't find any example there, except in the answer at: https://math.stackexchange.com/a/4076175/513178, by @ahulpke. But even that is not clear to me. Seems, need some elaboration, as how that example (in the answer); or the OP (of the link) is about finding map of elements in a domain. – jiten Oct 27 '22 at 05:40
  • @codeofsilence sorry, as couldn't wait for more than few minutes. If you can, please join chat. – jiten Oct 27 '22 at 06:52