The problem
Consider an equation of the following form
or, perhaps in a more readable notation
Lets refer to sums of this form as balanced sums. In other words, we are looking at sequences of natural numbers of length
Obviously,
The problem: given
For example, for
Straightforward solution
A straighforward constructive solution for this problem can be obtained by applying the formula for the sum of arithmetic progression to both sides of the equation:
From the last expression it is obvious that increasing
A second look
However... Looking at the results for different values of
q | n |
---|---|
1 | 1 |
2 | 1 |
3 | 3 |
4 | 1 |
5 | 3 |
6 | 3 |
7 | 3 |
8 | 1 |
... | ... |
q | n |
---|---|
30 | 9 |
31 | 3 |
32 | 1 |
33 | 9 |
34 | 3 |
35 | 9 |
36 | 5 |
37 | 3 |
... | ... |
q | n |
---|---|
60 | 9 |
61 | 3 |
63 | 15 |
64 | 1 |
65 | 9 |
66 | 9 |
67 | 3 |
68 | 3 |
... | ... |
All results are odd numbers. Powers of
then
which is nothing else than the number of odd divisors of
This raises an interesting question: how does this connection between the above balancing sums and odd divisors comes into existence?
Odd divisors
Let's apply some transformations to
It follows from
Also,
Finally,
So, the number of different odd values of the right-hand side of
And this is exactly what's expressed by
Some additional remarks
-
Since
is even, there is always at least one complementary even-odd pair of divisors that satisfies the above requirements: . Substituting into we get . I.e. every perfect square is a starting point for a balanced sum with terms on the left-hand side and terms on the right-hand sideConversely, all balanced sums of with "almost equal" number of terms on both sides (i.e.
) start from a pefect square.When
is a power of , a balanced sum of this kind is the only one available. -
Using a complementary even-odd pair of divisors
in in "reverse" order (i.e. substituting the larger divisor as ) results in negative value of . Of course, we'll still get a sum that satisfies the equality . For example, for : , for pair we can substitute and end up with . also provides us with a balanced sumbut it "doesn't count" in the original statement of the problem.
-
It is easy to show that if the smaller divisor produces value
, then the larger divisor will from the same pair will produce . Consequently, the the left-hand side sum corresponding to the negative value of will have the following formThe left-hand side includes a symmetrical sequence of
additional terms before , which all sum to zero. This means that a sequence of terms starting from will ultimately lead to the same sum as a sequence starting from , i.e. both divisors from a complementary pair lead to the same value of the sum and to the same right-hand side in (which is already ilustated by the above example) -
Obtaining
from is only possible when is a product of two consequitive natural numbers. For example, for : .Substituting
we obtain . Substituting we obtain . Again, the latter gives us a balanced sum starting from , but it is not eligible in the original statement of the problem. -
Finally, values of
representable as product of two consequitive natural numbers satisfyThe left-hand side represents triangular numbers. So, obtaining zero on the right-hand side of
is only possible when is a triangular number.