10

I'm interested in the ring $(\mathbb Z/30\mathbb Z,+,\times)$, the elements of which I'll write down in bold. For example $$\textbf{7}=7+30\mathbb Z=\{...-83,-53,-23,7,37,67...\}.$$ I'm interested in particular with $$U:=(\mathbb Z/30\mathbb Z)^\times=\{\textbf{1},\textbf{7},\textbf{11},\textbf{13},\textbf{17},\textbf{19},\textbf{23},\textbf{-1}\}$$ because every prime except $2,3$ and $5$ does belong to one of theses classes of numbers. There is a lot of things to say about $\mathbb Z/30\mathbb Z$, but those are the types of objects that interest me : $$\boxed{J\overset{def1}=\{p\in (\mathbb Z/30\mathbb Z)^\times|p+\textbf{6}\in (\mathbb Z/30\mathbb Z)^\times \}}$$For example, $\textbf{19}\notin J$ $$\boxed{G\overset{def2}=\{p\in (\mathbb Z/30\mathbb Z)^\times|6p+\textbf{1}\in (\mathbb Z/30\mathbb Z)^\times\}}$$About $J$ and $G$, there's a result that I find interesting about involution defined below$$\boxed{\varphi:U\to U, u\mapsto u^{-1}\text{ induces bijection }\varphi_{/J}:J\to G}$$since $u\in J\overset{def1}\iff u+\textbf{6}\in U \iff u^{-1}(u+\textbf{6})\in U\iff \textbf{1}+6u^{-1}\in U\overset{def2}\iff u^{-1}\in G$ enter image description here


Edit (to generalize, see question 3):

let $p_1=2,p_2=3,p_3,...$ the prime numbers. $$\forall n \in \mathbb N^*, p_n\#:=p_1...p_n\text{ (n-th primorial)}$$ $$U_n:=(\mathbb Z/p_n\#\mathbb Z)^\times$$ $$J_n:=\{p\in U_n|p+6\in U_n\}$$ $$G_n:=\{p\in U_n|6p+1\in U_n\}$$ $$\text{result :}|J_n|=2(p_n-2)...(p_2-2). \text{So, }\lim_{n \to \infty}|J_n|=+\infty$$


My questions:

1.- I was wondering if $J$ and $G$ were given names and if not, why?

2.-If you have other simple results like this in $(\mathbb Z/30\mathbb Z)^\times$, for example like $|J|=|G|=2(5-2)(3-2)=6$, where $|X|$ denotes the number of elements of a set $X$.

3.- These results easily extend to $\mathbb Z/210\mathbb Z(\text{for example }, |J_4|=|G_4|=2(7-2)(5-2)(3-2)), \mathbb Z/2310\mathbb Z, ..., \mathbb Z/p_n \# \mathbb Z$.

Does the switch to $\mathbb Z/210\mathbb Z$ bring any interesting new results?

  • 2
    I'd be interested to know why the downvote. What is not clear? What is not useful? $J, G $? – Stéphane Jaouen Mar 24 '24 at 03:16
  • 6
    @Kapil: Well, more than half the questions on this site would be too elementary if that were the standard. Also, I don’t see how the isomorphism you mention resolves this question. (Note that you can get proper angle brackets $\langle\rangle$ by using \langle and \rangle instead of $\lt$ and $\gt$ (which yield the wrong spacing because they’re interpreted as binary operators).) – joriki Mar 24 '24 at 05:20
  • 1
    @StéphaneJaouen Can you explain what you mean when you write $|J| = \ldots = 3$? Obviously that equality is false. Are you just trying to find a relationship between $|J|$ and the set ${2,3,5}$ that might extend to $(\mathbb{Z}/(p#)\mathbb{Z})^\times$? – Brian Moehring Mar 24 '24 at 05:26
  • @BrianMoehring : yes, the right "formula" would be $|J|=2(5-2)(3-2)=6$. – Stéphane Jaouen Mar 24 '24 at 08:10
  • @Bill Dubuque : I could generalize further by introducing for example $J_{n,m}:={p\in U_n|p+p_m#\in U_n}$ and further and further without bringing much to the table. But I'm worried that this will detract from the clarity of the initial post – Stéphane Jaouen Mar 24 '24 at 14:09
  • 3
    @Kapil, take a look here. In particular: "Mathematics Stack Exchange is for people studying mathematics at any level and professionals in related fields." The point is there are no too elementary questions on MSE if the asker put appropriate effort into the question they want help with. – Ennar Mar 24 '24 at 18:39
  • @Stéphane Jaouen: Your set $U$ has $\phi(30)$ elements. Is it casual or not? With other $U$, for example for your quoted $210$, we wil have $\phi(210)$ elements too? – Ataulfo Mar 26 '24 at 13:45
  • Obviously you know it but it is not the question. – Ataulfo Mar 26 '24 at 13:50
  • 1
    I speak about your choice of elements. Salut l'ami. – Ataulfo Mar 26 '24 at 13:53
  • I did not choose them : it's the group of unities of $Z/30$ – Stéphane Jaouen Mar 26 '24 at 13:55
  • 1
    I see okay. Names you ask are for $J$ and $G$, not for $U$. I am sorry. – Ataulfo Mar 26 '24 at 13:59
  • One little point, that may be underlying some of the phenomena, is that, for example, the subring of $\mathbb Z/15\mathbb Z$ of elements $m=1\mod 5$ and $m=0\mod 3$ is a subring, WITH_UNIT $6$, which is not a unit in the ambient ring. Such examples are easily found via Sun-Ze's theorem, etc. – paul garrett Mar 26 '24 at 21:27
  • 1
    You've edited this thread 27 times over the past few days. This is far too much. This site is not meant for blogging. Please try to coalesce your edits to minimize this in the future. – Bill Dubuque Mar 27 '24 at 19:11

2 Answers2

1

Since I have not received a reply, I do not think it is inappropriate to propose names to $J$ and $G$( in my answer, I will not use bold notation but the usual misuse of notation.)

To make the name I give them understood, it is necessary to generalize. Let $n,m\in \mathbb N^*$. $$U_n:=(\mathbb Z/p_n\#\mathbb Z)^\times$$ $$J_{n,m}:=\{p\in U_n|p+p_m\#\in U_n\}$$ $$G_{n,m}:=\{p\in U_n|p_m\#.p+1\in U_n\}$$

Examples :

$$p_1\#=2, p_2\#=p_1.p_2=2\times 3=6$$ $$J_{n,1}:=\{\color{red}p\in U_n|\color{red}{p+2}\in U_n\}, G_{n,1}:=\{\color{red}p\in U_n|\color{red}{2p+1}\in U_n\}$$ $$J=J_{3,2}, G=G_{3,2}$$

Defining these objects and studying them is obviously inspired by Sophie Germain's numbers and twins(Jumeaux in French). So I call them "m-order twins in $\mathbb Z/p_n\#\mathbb Z$" and "m-order Germains".

Hence my $J$ and $G$ notations of the original post. I hope that this will be a little clearer, as my questions are still open.


Coefficients $$\alpha_m=\frac{|J_{n,m}|}{|J_{n,1}|}$$do not depend on $n$.$$(\alpha_m)_{m\geq 2}=(2,\color{green}{\frac83},\frac{16}{5},...), \alpha_{m}=\alpha_{m-1}\frac{p_m-1}{p_m-2}$$


It is clear that the modular objects considered before have little to do with prime numbers. However, let's consider a long list of prime numbers and count $p$ such that $p+30$ is also prime and $p$ such that $p+2$ is also prime. If we calculate the ratio between these two results, we find a number close to $\color{green}{\frac83}$

1

As you already observed that $J$ and $G$ are in bijection, so let’s just focus on $J$. These results generalize to a wider setting, not necessarily in your formulation of $p_n^{\#},$ i.e. the $n$-th primorial. More precisely, your observations can be derived from the following results, which follow in turn from solving congruences and applying the Chinese Remainder Theorem.

Lemma 1. Let $n=p^m,m\geq 1, p~{\rm a ~prime ~and~}a\in {\mathbb Z}.$ Define $U_n$ and $J_n(a)$ by $$U_n=\left({\mathbb Z}/n{\mathbb Z}\right)^\times~{\rm and~}J_n(a)=\{u\in U_n~|~u+a\in U_n\}.$$ Then $$|J_n(a)|=\left\{\begin{array}{cc}|U_n|=\varphi(p^m)=(p-1)p^{m-1},&~{\rm if~}p \mid a\\ p^m-2p^{m-1}=(p-2)p^{m-1},&~{\rm if~}p\nmid a.\end{array}\right.$$ This motivates the following definition (not sure if this exists in the literature).

Definition 1. For a prime number $p,$ integer $m\geq 1,$ and $a\in {\mathbb Z},$ define $\varphi_a(p^m)$ by $\varphi_a(p^m)=\varphi(p^m)$ if $p\mid a$, and $\varphi_a(p^m)=p^m-2p^{m-1}$ if $p\nmid a.$

Lemma 2. Let $n=p_1^{m_1}\cdots p_r^{m_r},r\geq 2,m_1,\cdots,m_r\geq 1,~{\rm and~}p_1,\cdots,p_r$ be distinct primes. Let $$U_n=\left({\mathbb Z}/n{\mathbb Z}\right)^\times \simeq \prod_{k=1}^r\left({\mathbb Z}/p_k^{m_k}{\mathbb Z}\right)^\times$$ and $$J_n(a)=\{u\in U_n~|~u+a\in U_n\}.$$ Then $$|J_n(a)|=\prod_{k=1}^r|J_{p_k^{m_k}}(a)|=\prod_{k=1}^r\varphi_a(p_k^{m_k}),$$ where $\varphi_a$ is as in Definition 1.

As in the beginning paragraph, the proof is by the Chinese Remainder Theorem. Here one includes two examples.

Example 1. For $n=3^4\cdot 7\cdot 11^2,a=2\cdot 5\cdot 7,$ one has $$|J_n(a)|=(3^4-2\cdot 3^3)(7-1)(11^2-2\cdot 11)=16038.$$

Example 2. For $n=2^3\cdot 5\cdot 7^2\cdot 11,a=2^2\cdot 3\cdot 7,$ one has $$|J_n(a)|=(2^3-2^2)(5-2)(7^2-7)(11-2)=4536.$$

Note. In Question 2, you have $a=6=2\cdot 3, n=30=2\cdot 3\cdot 5,$ so by Lemma 2, $$|J|=(2-1)(3-1)(5-2)=6.$$ Similarly, in Question 3, $a=6=2\cdot 3,n=210=2\cdot 3\cdot 5\cdot 7,$ hence $$|J|=(2-1)(3-1)(5-2)(7-2)=30.$$

Pythagoras
  • 7,149
  • Thank you very much for your reply. I have several remarks:1) You don't give a proof of first Lemma .2) I find it surprising that you use my notations J and G since J comes from the French "Jumeaux" for Twins; "J and G are in bijection", as I already observed. I was wondering if J and G were given names and if not, why? – Stéphane Jaouen Mar 31 '24 at 07:25
  • I'm studying your lemma 1: it's easy to understand what's going on in J9(3) for example [using your notations]. Lemma 2 is clear by CRT but not lemma 1. – Stéphane Jaouen Mar 31 '24 at 07:28
  • I proved lemma 1 (without using CRT which has nothing to do with lemma 1). Where did you know these results from? – Stéphane Jaouen Mar 31 '24 at 10:09
  • The proof of Lemma 1 involves solving congruences mod $p$ and lifting it to classes mod $p^n$. The method is elementary but it may date back to Euler or at least Lipschitz. I did not invent J and G. I was using your notations: this seems cool, as for $a=2,$ $p+2$ is related to twin primes, and $2p+1$ is related to the Sophie Germain primes. – Pythagoras Mar 31 '24 at 15:09
  • 1
    Your observation was new to me, and I made a generalization to Lemma 1, which might have existed or not. If you would like to cite, just cite “Pythagoras”. – Pythagoras Mar 31 '24 at 15:13
  • Sophie Germain primes are cool(https://oeis.org/A005384); ...also $210p+1$(don't find OEIS associated, I know there is)... idem for $p_m#.p+1$ for all $m\in \mathbb N^*$. See my definition of $G_{n,m}$ in my answer below. Did you know the existence of bijection between $J_{n,m} $ and $G_{n,m}$? – Stéphane Jaouen Apr 13 '24 at 07:21
  • Do you require $p$ to be a prime or just a unit in $({\mathbb Z}/n{\mathbb Z})^\times$? I am not sure about the case for prime, but for $p$ to be a unit, your proof in your original post should go through. – Pythagoras Apr 14 '24 at 00:48
  • I require $p$ to be just a unit. Yes, my proof in OP goes through. – Stéphane Jaouen Apr 14 '24 at 05:22