1. Introduction
Throughout, we denote by
the largest prime divisor of an integer
. In [
1], for
, the sequence of integers
, where
and, for each
,
has been considered. For example, in the case when
and
, this sequence
is
Evidently, no two consecutive terms of the sequence
defined in (
1) can be a composite. Deleting all the composite terms and leaving only those elements of
that are primes, we will obtain a sequence of prime numbers
, where
if
is a prime number and
if
is a composite number, satisfying
for each
. Accordingly, removing the composite terms from (
2), we obtain the following sequence of primes
satisfying (
3) with the first term
and
:
The sequences (
1) and (
3) are both iterative sequences of integers
where
f is a map from the set
to itself. The most known sequence of this type is the Collatz sequence defined by
for
x odd and
for
x even; see [
2] and some recent papers [
3,
4,
5,
6] on the original Collatz problem and its variations. The results are very far from the conjecture asserting that the Collatz sequence starting from an arbitrary positive integer is ultimately periodic with the period
. Some other versions of iterative integer sequences have been considered in [
7] (where
) and subsequently in [
8,
9].
In [
1], it was shown that the sequence (
1) is periodic for any
and any initial choice of
. Now, we will give a different proof of this fact by deriving an explicit upper bound on the largest element of this sequence in terms of
and
k. Of course, this immediately implies the periodicity of
, because, by (
1), for each
, the element
is uniquely determined by its predecessor
.
To present our result, we will use the following notation. For a given
, by
we denote the smallest prime number that does not divide
k. For odd
k, it is clear that
Here are the first 15 values of
for
k even.
k | 2 | 4 | 6 | 8 | 10 | 12 | 14 | 16 | 18 | 20 | 22 | 24 | 26 | 28 | 30 |
| 3 | 3 | 5 | 3 | 3 | 5 | 3 | 3 | 5 | 3 | 3 | 5 | 3 | 3 | 7 |
By the definition of
, it follows that
for each
with equality if and only if
. For large
k, the upper bound for
is much better than that in (
4). Indeed, let
be the least prime number not dividing
. Then, all the primes smaller than
q must divide
k. Thus, their product
divides
k and hence
Using the asymptotical formula
as
, we deduce that for any
there is a constant
such that
In fact, by [
10] (Theorem 4), the lower bound
holds for
, so an explicit
in (
5) in terms of
can be determined.
With this notation, we can state our first result:
Theorem 1. All the elements of the sequence (1) are smaller than or equal to while all the elements of the sequence (3) are smaller than or equal toIn particular, the sequences (1) and (3) are both periodic. For
and
, the sequence (
1) is
, while the right-hand side of (
6) is
. For
and
, the sequence (
3) is
, whereas the right-hand side of (
7) is 3. So, formally, the inequalities (
6) and (
7) are the best possible. For
, these bounds can be improved, but we will not go into the details.
More generally, for a fixed real number
satisfying
and a finite set
with
we will consider a class of integer sequences
consisting of all sequences
satisfying
and, for each
,
Note that the smallest composite number in
is 4, so some integer
satisfying (
10) can always be chosen due to
and
.
In particular, if
is a singleton set (this notation is consistent with (
8)), then, by (
9) and (
10), for each
,
It is clear that
and
For each
, we will show the following:
Theorem 2. Assume that and that has the largest element k. Then, the elements of the sequence (as defined in (9) and (10)) are all smaller than Furthermore, in a particular case, when , all the elements of are smaller than or equal towhile all the prime elements of S do not exceedwhere is the first prime element of the sequence S. Note that parts (
12) and (
13) of Theorem 2 imply Theorem 1. Indeed, the sequence (
1) belongs to the class
, where
is singleton set and
. (The largest prime factor of a composite integer
does not exceed
.) Thus, the upper bound (
6) follows from (
12) with
, whereas (
7) follows from (
13). If
is the pair of positive integers with the smallest index
i and the smallest difference
satisfying
, then the sequence (
1) is ultimately periodic with period
. (The same is true for the sequence (
3) and the first pair of primes in it satisfying
.)
Of course, the sequences in , although bounded, are not necessarily all periodic. All the cases when they are all periodic are described by the next theorem:
Theorem 3. Assume that and . Then, the sequences in are all periodic if and only if one of the following holds:
- (i)
and ;
- (ii)
and .
In all other cases, the class contains infinitely many nonperiodic sequences.
In the next section, we will give three auxiliary lemmas. Then, in
Section 3 and
Section 4, we will prove Theorems 2 and 3, respectively. (As we already observed above, Theorem 2 implies Theorem 1.) In the last section, we will show that the class
always contains nonperiodic sequences in the case when
is infinite.
2. Auxiliary Lemmas
Lemma 1. For any integers , the arithmetic progressioncontains a composite number. Moreover, if , then the arithmetic progressioncontains a composite number. For example, for
, we have
. Selecting
, we see that the first 10 numbers in (
15)
are all primes, while the eleventh number
is a composite. This shows that for
, the list (
15) cannot be replaced by the shorter list
. See the Wikipediaarticle (
https://en.wikipedia.org/wiki/Primes_in_arithmetic_progression) (accessed on 2 January 2024) for some further nontrivial examples of primes that form long (in terms of
k) arithmetic progressions with the difference
k.
Proof. Consider the list of integers (
15) modulo
. If for some integers
satisfying
the numbers
and
were equal modulo
q, then
. Because
q is a prime and
, this forces
, which is not the case by the definition of
. Therefore, the integers (
15) are all distinct modulo
q, which means that exactly one of them, say
,
, is divisible by
q. This number is composite, unless
. Note that for
, we have
which is impossible using (
4). Hence, the equality
occurs only for
and
. This proves the second assertion, because then
. Of course, for
, this also proves the first assertion. On the other hand, if
, then the last number in the list (
14), namely,
, is composite. This completes the proof of the first assertion of the lemma. □
The next lemma is (1.12) from [
11].
Lemma 2. For any real numbers and , the interval contains at most prime numbers.
This result of Montgomery and Vaughan is related to the famous Hardy–Littlewood conjecture, which asserts that for the prime-counting function
the inequality
holds for any integers
, see [
12] (p. 54). This inequality has been proved only under some assumptions on
x and
y; roughly, when
x and
y are of similar size, see, e.g., [
13,
14,
15,
16]. More references can be found in [
17]. However, in our situation,
y can be small compared to
x, so the bound with an extra factor 2 as given in Lemma 2 seems to be the best available known result for our purposes. In fact, as a result of Hensley and Richards [
18], the conjecture of Hardy and Littlewood is incompatible with the so-called prime
k-tuples conjecture, which is widely believed to be true. In view of this, it is not clear at all if the constant 2 in Lemma 2 can be replaced by a constant arbitrarily close to 1.
To state our next lemma, we need the following definition. We say that a finite string of positive integers
is an
s-cycle in the class
if
,
for
and the purely periodic sequence
belongs to the class
. This means that the elements of the sequence
are all in
and satisfy (
9) and (
10). (Of course, it is sufficient to verify this for
, because
and the sequence
is periodic.)
For example, consider the case
and
. Note that if
, then, by (
10), as
we can select, for instance, 3 or 6. Hence,
and
are both 3-cycles in the class
. (Their first element is 3, and 3 is the only element in both strings
.)
Lemma 3. Assume that for some integer , the class has at least two distinct s-cycles. Then, contains infinitely many nonperiodic sequences.
Proof. Let C and be two distinct s-cycles in . Take any nonperiodic sequence with two letters of the alphabet . Then, replace in it with their corresponding strings of integers, say, and . We claim that the resulting sequence is nonperiodic.
Assume that S is periodic. Then, without the loss of generality, we may assume that some period in it starts with s and ends at a certain integer . The next element of S must be s again; so, in the period, we can replace the strings back to the letters C and . Because S is periodic, this means that a nonperiodic sequence on from a certain place is also represented by a periodic sequence on the same two letters. Consequently, at some stage, say from the gth element, we must have the cycles C and both starting from the same element . As , the cycles C and cannot be of the same length. Indeed, otherwise, the sequence of , starting from the element , is uniquely determined, and a nonperiodic sequence on these two letters cannot be represented by a periodic one.
Assume that C has more elements than , i.e., . Recall that the cycles C and both start from . But then, as after we have C or , the element of the sequence S must be s, which is not allowed by the definition of C (s is only the first element of C). The case can be treated with the same argument.
Therefore, the sequence S obtained as a nonperiodic combination of two s-cycles and then replacing them with their corresponding strings of numbers in is indeed nonperiodic.
Finally, observe that, by taking any composite integer
greater than
and adding it to the beginning of the above constructed nonperiodic sequence
we will obtain a new nonperiodic sequence
in
; see the property (
10). This completes the proof of the lemma, because there are infinitely many choices of such integers
. □
3. Proof of Theorem 2
Let
be a sequence from the class
. For the simplicity of exposition, we present this sequence in a binary alphabet
, where the letter
p stands for
if
is prime, and the letter
c stands for
if
is composite. For example, the sequence (
2) is
We clearly have
if the letter
p stands for
, and
if the letter
c stands for
.
Let be a subsequence of S obtained from S by deleting its composite elements, so simply enumerate the letters p. If the sequence were finite, then we would have for each sufficiently large i, say, for . But then, for each from , we deduce that is a decreasing sequence of integers. This is impossible, because for all . Consequently, the sequence is infinite. In the notation with p and c, this means that the sequence S contains infinitely many letters p.
Next, we consider a subsequence of obtained by removing from the primes from consecutive patterns of all primes except for the first one. In particular, we will have , while for each , , between and , first there are possibly a few prime elements of S and then there must be one of several composite elements of S.
Now, we will prove (
13). (Recall that
.) We claim that
for each
.
We will use the induction on
i. Of course, (
16) trivially holds for
because then
. Assume that (
16) is true for some
, where
. Suppose that between
and
there are
primes (letters
p) and then
composite elements of
S (letters
c). By Lemma 1, we have
Thus, the first composite element is smaller than or equal to
. The
lth composite element (the one that appears just before
, say,
) is therefore at most
. Hence,
Now, by our inductive assumption
, it remains to verify the inequality
However, the latter inequality is equivalent to
, which is true by the definition of
M in (
16). This completes the proof of (
16).
Next, note that each
is of the form
with some integers
and
. Furthermore, we must have
by Lemma 1. Hence, by (
16), each
,
, is smaller than or equal to
. This completes the proof of (
13).
In order to prove (
12), we first observe that, by Lemma 1 and the definition of
, each element of the sequence
S is smaller than or equal to
Hence, by the definition of
M in (
16), all the elements of
S do not exceed
Because
, (
18) does not exceed the right-hand side of (
12).
It remains to prove (
11) for the set
with the largest element
k. This time, we claim that
for each
.
It is clear that (
19) is true for
. Assume that (
19) is true for
with
. As above, suppose that between
and
first there are
prime elements and then
composite elements of
S. We will show that
By
, it is clear that (
20) holds for
, so assume that
. The inequality (
20) also holds for
K being a singleton set by (
4) and (
17) because
. Thus, we can assume that
. The
consecutive elements of
S
where
, are all prime, and the first composite element of
S following them is
If
, there are also other composite elements between this element and
, but they all appear in descending order. This means that
Also, the interval
contains at least
prime numbers, for example,
distinct primes that are listed in (
21). Here,
and
Therefore, using
and (
8), we obtain
Hence, by Lemma 2, it follows that
because the function
is increasing for
. Inequality (
23) implies
, which yields (
20).
Next, by (
20) and (
22), we obtain
Using the inductive assumption
, from (
24) we deduce that
, which is less than or equal to
by the definition of
in (
19). Hence,
. This concludes the proof of (
19) for each
.
Because the bound on
ℓ in (
20) is independent of
i, the largest prime element of
S does not exceed
Consider the subsequence
of
, where
and
u is the smallest integer with this property. By (
25), the largest element of this subsequence is less than
This proves (
11) in the case when
. Assume that
. Then, the largest element of
is either less than
(if it is among
) or is equal to
. Because
,
does not exceed the right-hand side of (
11). On the other hand, the element
is also strictly smaller than the right-hand side of (
11). Consequently, all the elements of
S are smaller than
. This finishes the proof of the theorem.
4. Proof of Theorem 3
Consider the case (i). Let
be a sequence in the class
, where
. If
, then, by (
9) and (
10),
S is a purely periodic sequence with period
or
or
. We will show that any
is ultimately periodic with one of those three periods. The proof is by induction on
. Assume that
and that we already established the periodicity of the sequence
S in case it has an element at most
. For
, at least one of the numbers
is composite and the next element of
S is smaller than
. The periodicity now follows due to our inductive assumption.
Now, consider the case (ii). Let
be a sequence in
, where
. We will show that each
S is ultimately periodic with one of the possible periods,
or
or
or
or
or
. If
, then, by (
9) and (
10),
S is purely periodic with period
or
. If
, then after several steps we reach
(possibly
), so the next element
is less than
. If
is 2 or 4, then the sequence becomes ultimately periodic with period
or
. Otherwise,
. If it is always 3, namely,
for every
, then the sequence is purely periodic with period
or
or
or
. If otherwise
for some
, then it is ultimately periodic with period
or
. For
, one of the integers
is composite, so the next element of
S is less than
. Hence, it is at most
, which concludes the proof by induction on
s.
Assume that and K are such that neither (i) nor (ii) is satisfied. We first consider the case when the set K contains an element k satisfying . In view of , it is sufficient to show that contains infinitely many nonperiodic sequences.
Suppose first that is even. Then, is composite. Thus, is a 2-cycle of . Moreover, if is a composite number, then is also a 2-cycle of . On the other hand, if is a prime number, then and is not a prime. In that case, is a 2-cycle of . Therefore, in both cases, for even , the class contains at least two distinct 2-cycles. Consequently, by Lemma 3, it contains infinitely many nonperiodic sequences.
Likewise, for odd, and , where is composite, are both 3-cycles of , so the result follows by Lemma 3. If is a prime, then and is a 3-cycle, because is composite and greater than or equal to 6. The result again follows by Lemma 3.
In the remaining cases , we will explicitly present the corresponding 2-cycles in . For , in there are two distinct 2-cycles, and . For , there are two distinct 2-cycles, and . For , there are two distinct 2-cycles, and . Finally, for , in there are two distinct 2-cycles, and . In all the above cases, the required result follows from Lemma 3.
Now, it remains to consider the case . Suppose first that . Then, the class contains two distinct 2-cycles, and , so the proof is concluded by Lemma 3. Because the cases (i) and (ii) are already considered, we are left with two possibilities , and , . If and , then in there are the following two distinct 2-cycles: and . Finally, if and , then the class also contains two distinct 3-cycles, for instance, and . In both cases, the proof is concluded by Lemma 3 as before.
5. Concluding Remarks
The main result of this paper’s Theorem 2 shows that the sequences of the class are all bounded. More precisely, the largest element of is bounded from above in terms of and the maximal element of K no matter how large the finite set K is.
What about the case when the set is infinite, which is possibly a very sparse set? We will show that then no result similar to Theorem 2 is possible, because the class always contains unbounded sequences for any infinite and any .
Indeed, let us start the construction of such
from any prime number
. Because
K is infinite, we can choose
so large that
Take the least positive integer
j for which the number
is composite. By Lemma 1, this
j does not exceed
. Then, by the rule (
9), because
are all primes, the numbers
can be chosen as the consecutive elements of
S. Because
is composite, by the rule (
10), as the next element
of
S we can choose any integer from the interval
This is indeed possible, because the right endpoint of the interval (
27) is
, while its left endpoint is
due to (
26).
Note that the interval (
27) is of the form
, with
. Therefore, by Bertrand’s postulate, it contains a prime number, say,
. Let us choose
. Because
belongs to the interval (
27), using (
26), we deduce
so
.
Now, arguing with
as before, namely, choosing
so large that
we will construct another prime element
of
S satisfying
. Continuing this process, we will obtain a sequence
containing an infinite subsequence of primes
The latter sequence of primes is unbounded, so
is unbounded too.