Quote:
Originally Posted by nesio
Sum means here the amount, quantity.

We have some language problems here. Just to clarify:
12 has divisors 1, 2, 3, 4, 6, 12. Is 6 the result you mean, or is it 28?
Quote:
Originally Posted by ThiloHarich
why should my model be very rough?

The rough model is the one he proposed, not yours:
Quote:
Originally Posted by nesio
You can check this rough (very rough) model of m on your source data of n:
maximum(S/(m^1/3)) and m < n^1/3,
where S  the sum of all divisors of m.
Here are S and m^1/3 have equal weights in maximum function.
What will be?

I interprete this as follows:
The quality of a multiplier m is about S/(m^1/3), where we consider m < n^1/3 only.
S might be
a) the sum of divisors of m, or
b) the number of disivors of m.
My first question in this post aims to clarify which of the two options is the correct one.
nesio, do I get that right so far?