How to prove it?
How to prove \(Z_n =<[d]>~~~iff~~~gcd(d,n)=1\)
@nerdguy2535
i think a double set inclusion should work here. Clearly \(\langle [d]\rangle \subseteq \mathbb Z_n\). We just need the other inclusion.
Since \(\langle [1]\rangle =\mathbb Z_n \) (i.e. \(\mathbb Z_n\) is cyclic), we just need to show that \([1]\in \langle [d]\rangle\). Then it follows that \(\mathbb Z_n\subseteq \langle [d]\rangle \).
To show that \([1]\in \langle [d]\rangle \), we need to show there exists an integer \(m\) such that: $$m[d]=[1].$$Can you think of a way to use the hypothesis that \(\gcd(d,n)=1\) to show the existence of such an m?
Why do we have to do that? for this way (-->), is it not that if \(\mathbb Z_n =\langle[d]\rangle \) means elements in d is in Z_n??
Ah, I didn't realize its an iff. I'm doing \(\Longleftarrow\) right now. Is \(\Longrightarrow \) the one you would rather tackle first?
either of them. I have to generalize the case in my attachment formally. We cannot "Notice" or "Observe" in our argument, right? :)
Right. The key to using the gcd condition is the following theorem: $$\gcd(d,n)=1\iff \text{There exist }x,y\in \mathbb{Z}\text{ such that }dx+ny=1.$$
Have you seen this theorem before?
yes, but I don't see the link between generator d with it
I think you should play with that theorem some more then. The gcd(d,n) is very much related to what \(\langle [d]\rangle \) generates, and that theorem is pretty much the bridge that connects them.
This is what I did, but I go nowhere. Suppose gcd (d,n ) = a, so that a|d and a|n and d = a a1 , n = a a2 that gives me d/a1= n/a2 --> d = \(\dfrac{a_1}{a_2}n\)
if [d] generate Z_n, then 1[d] = element in Z_n 2[d] = same as above 3[d]= ........ and we know that [d] = d (modn)
and replace d =a1/a2 n I have 1d = a1/a2 * n if I have \(\dfrac{a_2}{a_1} d= n\) then I can show this n does not belong to Z_n so that gcd (d,n) must be 1
But it is ambiguous. :(
When working with \(\mathbb Z_n \), you don't want to have fractions flying around. So I don't think working with statements like \(d=\frac{a_1}{a_2}n\) are very helpful =/
It should always be about integers, or equivalence classes of integers.
yes,
If \(\langle[d]\rangle=\mathbb Z_n\), then for all \([y]\in \mathbb Z _n\), there exists an integer \(x\) such that \(x[d]=[y].\) In particular, there exists an integer \(x_0 \) such that \(x_0 [d]=[1]\). Can you use this to find a linear combination of d and n that equals 1?
We have theorem like: if [d] generates Zn , order of [d], namely k, then gcd(k, n) =1 does it help us?
I think you might be misquoting that theorem >.< as written its not true. For example, take \([5]\in \mathbb Z_6\). [5] generates the group, so the order of [5] is 6, and the gcd(6,6)=6, not 1.
What the youtube theorem says is "If [d] generates Zn, then k[d] also generates Zn if and only of gcd(k,n)=1". That statement has nothing to do with the order of [d].
:( I spent 2 days on this problem, make a lot of short research, watch a lot of youtube tutoring but get nothing. :(
:(
All the information you need is in this thread now (for the most part). You just have to put it together.
How?
Its all in the theorem. $$\gcd(d,n)=1\iff \text{there exists}x,y\in \mathbb Z\text{ such that }dx+ny=1.$$
Because: $$dx+ny=1\iff [xd]=[1]\iff x[d]=[1]\iff [1]\in \langle [d]\rangle$$
then?
\([1]\in \langle [d]\rangle \iff \langle [1]\rangle \subseteq \langle [d]\rangle\iff \mathbb Z_n\subseteq \langle [d]\rangle \iff \mathbb Z_n =\langle [d]\rangle .\)
oh, I see, you prove if gcd (d, n) =1 then Z_n = <[d]> the leftover is backward, right?
right right.
I gave up. It's 1:19 am here. I miss my bed. hihihi.... Thanks for your time. Please, forgive me for my impoliteness. I am so tired.
no worries.
Join our real-time social learning platform and learn together with your friends!