: The Digit Sum Process.
: Curious Properties of Reiterated
: The Collatz Problem
We are going to study the DCS (digit cube sum) problem.
Definition 2.1.
We define the dcs function by
, where
is the list of the digits of an integer n.
If we apply dcs(n) function to a natural number n repeatedly, then we get the sequence . For the detail of the theory of the DCS problem see [7].
Example 2.1.
For Example, we start with n= 12. Then
dcs,
dcsdcs
dcsdcsdcsWe continue this operation and we get
Lemma 2.1.
We have

(2.1) 
for natural number such that
Proof.
Suppose that m has k digits, then
. Since each digit is less than or equal to 9,
dcs.
By Lemma 2.1 for we have
, and hence
dcs
If
and , then
dcs
Theorem 1
has been already proved in p.27 of
.
Theorem 2.
For any natural number the sequence
eventually will enter into a loop. In other words there exist numbers p and q such that
for any natural number k and any integer r such that
, where we call q the length of the loop. The sequence
enters into a loop when m = p.
Proof.
If we start with a natural number n, we have a sequence
by applying the function dcs to the number n repeatedly. When
, by Theorem 1
. If
, by Theorem 1 we have
. In this way this sequence decreases while the value of the sequence is bigger or equal to 2917. Therefore there exists a natural number t such that
. In this way we can prove that there are infinite number of m such that
, and hence there is a nonnegative integer u such that
for at least two values of m. Let v be the smallest integer such that
and let w be the second smallest integer such that
.
Then clearly
,...,
is a loop.
Example 2.2.
If we apply
dcs function to the number 115 for 20 times, then we get the sequence
{115, 127, 352, 160, 217, 352, 160, 217, 352, 160, 217, 352, 160, 217, 352, 160, 217, 352, 160, 217, 352}.
From this sequence we get Graph 2.1.
In Example 2.2 the
dcs function produces a sequence with a loop.
By Theorem 2 successive application of
dcs function will produce a sequence with a loop.
Now we are going to look for all the loops. By Theorem 1 we have only to look for loops of the sequences that start with a number
, and we can find loops by calculation of computers.
Example 2.3.
If we use the Mathematica function in Example 10.3, we can find all the loops. The output of the program is
{{1}, {55, 250, 133}, {133, 55, 250}, {136, 244}, {153}, {160, 217,
352}, {217, 352, 160}, {244, 136}, {250, 133, 55}, {352, 160,
217}, {370}, {371}, {407}, {919, 1459}, {1459, 919}}.
Loops of the length of 3 are {55, 250, 133},{160, 217,
352}, and loops of the length of 2 are {136, 244}, {919, 1459}.
Fixed points are {1}, {153}, {370}, {371}, {407}.
: The Digit Sum Process.
: Curious Properties of Reiterated
: The Collatz Problem