This paper investigates the properties of an arithmetic function counting non-prime units modulo n, providing insights into its behavior and distribution within number theory.
Contribution
It introduces and analyzes the function f, a novel arithmetic function counting non-prime units modulo n, expanding understanding of unit groups.
Findings
01
Characterization of f for various n
02
Distribution patterns of non-prime units
03
Relationships between f and classical arithmetic functions
Abstract
In this work, we studied various properties of arithmetic function φ~, where \tilde{\varphi}(n)=|\{m\in \mathbb{N} | 1\le m\le n, (m,n)=1, \mbox{m is not a prime}\}|.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAnalytic Number Theory Research · Advanced Mathematical Theories · Advanced Mathematical Theories and Applications
Full text
Number of non-primes in the set of units modulo n
Abhijit A J, A. Satyanarayana Reddy
Abstract.
In this work, we studied various properties of arithmetic function φ~, where
\tilde{\varphi}(n)=|\{m\in{\mathbb{N}}|1\leq m\leq n,(m,n)=1,\mbox{m is not a prime}\}|.
Key Words: Greatest common divisor, Arithmetic function, Euler-totient function.
AMS(2010): 11A05,11A25.
1. Introduction
For a fixed positive integer n, let Un={k:1≤k≤n,gcd(k,n)=1}. If ∣S∣ denotes the cardinality of the set
S, then ∣Un∣=φ(n), the well known Euler-totient
function. Let E_{n}=\{m\in{\mathbb{N}}|1\leq m\leq n,(m,n)=1,\mbox{m is not a prime}\}.
That is E_{n}=\{m\in U_{n}|\mbox{m is not a prime}\}. It is known that φ(n)=n−1 if and only if n is prime, hence En={m∈Un∣φ(m)=m−1}.
Let φ~(n)=∣En∣. It is clear that φ~ is an arithmetic function. First twenty values of φ~ are given in the following tables.
nEnφ~(n)
1
{1}
1
2
{1}
1
3
{1}
1
4
{1}
1
5
{1,4}
2
6
{1}
1
7
{1,4,6}
3
8
{1}
1
9
{1,4,8}
3
10
{1,9}
2
nEnφ~(n)
11
{1,4,6,8,9,10}
6
12
{1}
1
13
{1,4,6,8,9,10,12}
7
14
{1,9}
2
15
{1,4,8,14}
4
16
{1,9,15}
3
17
{1,4,6,8,9,10,12,14,15,16}
10
18
{1}
1
19
{1,4,6,8,9,10,12,14,15,16,18}
11
20
{1,9}
2
The following results provides few immediate properties of φ~ function.
Recall that for n∈N, π(n),ω(n) denotes the number of primes less than or equal to n,
the number of distinct primes divisors of n respectively. We denote Nj=p1p2⋯pj, product of first j primes, for example
N3=30. It would be interesting to note that 2n<φ(n)<n−n and by prime number theorem, for a large enough n, π(n)≈log(n)n. But, ω(n) does not have any ordinary behaviour. In fact for any natural number k, one can find a subsequence {ni} of natural numbers such that ω(ni)=k for all i. To know more properties of π(n),ω(n),φ(n) and
Un refer any one of [1, 2, 4, 5].
Proposition 0**.**
If pk denotes the kth prime, then φ~(pk)=pk−k.
In general φ~(n)=φ(n)−π(n)+ω(n).
Proof.
First part follows from the observation that Epk=Upk∖{p1,p2,…,pk−1}.
Let n=q1α1q2α2⋯qkαk. We claim Un∪{q1,q2,…,qk}=En∪{p1,p2,…,pπ(n)}
Suppose x∈Un∪{q1,q2,…,qk}. Now if x is prime, then
then x∈{p1,p2,…,pπ(n)} and if x is not prime, then x∈En. Thus x∈En∪{p1,p2,…,pπ(n)}.
Suppose y∈En∪{p1,p2,…,pπ(n)}. If y∈En, then y∈Un if y∈{p1,p2,…,pπ(n)} then y∈Un or
y∈{q1,q2,…,qk} depending on y∤n or y∣n respectively.
By definition, En∩{p1,p2,…,pπ(n)}=Un∩{q1,q2,…,qk}=∅. Hence
φ~(n)+π(n)=∣En∣+∣{p1,p2,…,pπ(n)}∣=∣Un∣+∣{q1,q2,…,qk}∣=φ(n)+ω(n).
∎
Proposition 1**.**
(1)
If n0 is the square free part of n then φ~(n)≥φ~(n0).
2. (2)
*Let n,k∈N and n≥3. Then φ~(nk)<φ~(nk+1).
*
3. (3)
Let n,p,q∈N where p,q primes with p<q and (p,n)=1. Then φ~(np)≤φ~(nq).
4. (4)
If a∈N and ω(a)=i, then φ~(Ni)≤φ~(a).
Proof.
Proof of Part 1: Since n0 and n have same set of prime factors, En0⊆En.
Proof of Part 2: It is easy to see that Enk⊆Enk+1. Sufficient to show that
Enk+1∖Enk=∅.
Since n≥3
there exists ℓ∈{2,3,…} such that (n−1)ℓ−1≤nk<(n−1)ℓ. Consequently nk<(n−1)ℓ<(n−1)ℓ−1⋅n≤nk+1.
Thus we produced an element (n−1)ℓ in Enk+1∖Enk.
Proof of Part 3: Let x∈Enp and qm∣∣x for some m∈N∪{0}. Then x=qmy for some y∈N. Further
that (pmy,nq)=1 as (pm,n)=1,(y,q)=1. Now we define a map f:Enp→Enq as f(x)=pmy. It is easy to see that
f is one-one. Hence the result follows.
Proof of Part 4 Let q1,q2,…,qi be all distinct prime divisors of a and s=q1q2⋯qi.
Then from the proof of Part 1 we have φ~(s)≤φ~(a).
Hence the result follows from the facts that Ni≤s and ENi∖{q1,q2,…,qi}⊆Es∖{p1,p2,…,pi}.
∎
2. Main results
It is easy to see that φ~(n)=1 if and only if n∈{1,2,3,4,6,8,12,18,24,30}. That is φ~(ℓ)>1 whenever
ℓ≥31. In general we prove.
Theorem 2**.**
For every n∈N, there exists N∈N such that φ~(m)>n, for all m>N.
Lemma 3**.**
Let n∈N. There exists an Mn∈N such that φ~(Nk)>n for all k∈N, k≥Mn.
Proof.
Let pi denote the ith prime.
For i≥4, we define Pi={m∈N∣φ(m)=m−1,m∈UNi}, Qi={r∈Pi∣rpi+1<Ni}.
From Bertrand’s Postulate pi+1∈Qi. Suppose Qi={pi+1,pi+2,…,ph} then ∣Qi∣=h−i. Further
pi+2<2pi+1 and ph+2<2ph+1<4ph. Thus pi+2ph+2<8pi+1ph<8Ni<Ni+1 as pi+1>8.
Hence {pi+2,…,ph+2}⊆Qi+1 consequently ∣Qi∣<∣Qi+1∣. Now {rpi+1∣r∈Qi}⊂ENi.
Thus φ~(Ni)>∣Qi∣. Hence the result follows from the fact that ∣Qi∣ increases with i.
∎
Since ∣Q4∣=4, therefore for all n≥4, ∣Qn∣≥n and φ~(Nk)>k when k≥4. Hence Mn≤max{4,n}. Further, it is easy to see that φ~ is an increasing function on {N3,N4,N5,…}.
Lemma 4**.**
Let a,b∈N and A(a,b)={n∈N∣φ~(n)=a,ω(n)=b}. Then A(a,b) is finite.
Proof.
Let p be the (b+ℓ)th prime, where a<2ℓ(ℓ+1). Let n∈N with ω(n)=b and n>p2.
Since there are at least ℓ primes less than p,which are co-prime to n, we can choose two such primes q1,q2. Then q1q2<p2<n.
Thus q1q2∈En as a consequence φ~(n)>2ℓ(ℓ+1)>a. Hence the result follows.
∎
Let n∈N. There exists k∈N such that if ω(t)>k, then φ~(t)≥φ~(Nω(t))>n.
Now from Lemma 4 for each i∈N, i≤k there exists mi∈N such that for all t∈N with
ω(t)=i and t>mi, then φ~(t)>n.
Thus if N=max{m1,m2,…,mn} and ℓ>N, then φ~(ℓ)>n.
∎
3. Future work
The following question is natural.
Let k∈N be fixed. Find all n such that φ~(n)=k.
One can see the following result by calculating the appropriate bounds and checking accordingly.
Lemma 5**.**
Let s(k)={n∈N∣φ~(n)=k}. Then the set of values k∈{1,2,…,100} such that s(k) is empty set is {13,31,70}.
Conjecture: If M={φ~(n)∣n∈N}, then N∖M contains infinite number of elements.
Also recall the Carmichael conjecture [3]: if N(m)=∣{n∈N∣φ(n)=m}∣, then N(m)=1. This conjecture no longer true for φ~ as φ~(n)=16 only for n=144.
If s(k) is as defined in the Lemma 5, then then the values of k∈{1,…,100} such that s(k) contains a single element is {16,39,47,49,53,57,58,65,66,76,85,91,94}.
The following tables and observations may be useful in answering the above questions.
Given a positive integer k, the following tables provides the smallest value of n such that
φ~(n)=k. For example φ~(1)=φ~(2)=φ~(3)=φ~(4)=1, hence the smallest value n such that φ~(n)=2 is 5.
[TABLE]
[TABLE]
[TABLE]
[TABLE]
The following observations are useful for the question posed in the beginning of the section for the case k≤20.
(1)
φ~(n)=1⇔n∈{2,3,4,6,8,12,18,24,30}.
2. (2)
φ~(n)=2⇔n∈{5,10,14,20,42,60}.
3. (3)
φ~(n)=3⇔n∈{7,9,16,36,48,90}.
4. (4)
φ~(n)=4⇔n∈{15,22,54,84}.
5. (5)
φ~(n)=5⇔n∈{26,28,66,120}.
6. (6)
φ~(n)=6⇔n∈{11,21,32,40,72,78,210}.
7. (7)
φ~(n)=7⇔n∈{13,34,50}.
8. (8)
φ~(n)=8⇔n∈{38,44,70,150}.
9. (9)
φ~(n)=9⇔n∈{102,114,126}.
10. (10)
φ~(n)=10⇔n∈{17,27,46,56,96,108,180}.
11. (11)
φ~(n)=11⇔n∈{19,33,52,132}.
12. (12)
φ~(n)=12⇔n∈{25,45,80,168}.
13. (13)
φ~(n)=14⇔n∈{23,39,58,62,110,138}.
14. (14)
φ~(n)=15⇔n∈{35,64,68,156,240}.
15. (15)
φ~(n)=16⇔n=144.
16. (16)
φ~(n)=17⇔n∈{74,76,100,140}.
17. (17)
φ~(n)=18⇔n∈{198,270,330}.
18. (18)
φ~(n)=19⇔n∈{29,51,88,98,162,174,420}.
19. (19)
φ~(n)=20⇔n∈{31,63,82,130}.
Acknowledgment: We would like to thank referee for valuable comments.
Bibliography5
The reference list from the paper itself. Each links out to its DOI / PubMed record.
1[1] T. M. Apostol, Introduction to Analytic Number Theory , Springer-Verlag, 1976.
2[2] David M. Burton, Elementary Number Theory , Mc Graw-Hill Higher Education, 7th edition, 2010.
3[3] R.D. Carmichael Note on Euler’s- φ 𝜑 \varphi function , Bull. Amer.Math.Soc.28 (1922) 109-110.
4[4] G. H. Hardy and E. M. Wright, An Introduction to the Theory of Numbers , 5th edition, Oxford University Press, Oxford, 1979.
5[5] Gareth A.Jones and J.Mary Jones, Elementary Number Theory ,Springer undergraduate mathematics series,1998.