A
Archaea
Guest
Consider an integer y = 3a2 + 2a3, where a2 is any integer, positive or negative, which is not divisible by 2, and likewise, a3 is an integer not divisible by 3. It's pretty straight forward to show that y is not divisible by 2 or 3.
We can continue with this logic to find a formula for an integer which is not divisible by 2, 3 or 5. The formula is y = 3·5a2 + 2·5a3 + 2·3a5, where a2, a3 and a5 are not divisible by 2, 3 or 5, respectively.
More generally, we can devise a formula to find all prime numbers in an interval [pn, pn2], where pn is the nth prime number, if we know all the primes p1 to pn-1. The equation in all it's glory is:
Where ap is any number not divisible by p.
So it's pretty simple, and while it might make it slightly easier to compute primes, the task still requires all the same information as it did before. There are a few questions that remain however:
1) Will all primes have a solution to this equation? I think it's possible to show this by showing that if we replace all the ap's with bp's, where the bp's can be any integers, then any integer y has a solution to the equation.
2) Given an integer, is it possibly to use the equation to determine if it's prime? I think this is actually fairly straight forward. The problem is that it doesn't seem to me like it would be computationally faster than existing methods.
3) Can we use this equation to determine the number of primes in an interval [pn, pn2]? If the solutions can be ordered in a nice way, then I think the answer is yes, otherwise, I think the problem would resemble the Sieve of Eratosthenes in some way.
There are a set of 4 prime number problems known as Landau's problems. I think this formula might be useful for solving some of these. I've already had a go at Goldbach's conjecture and I'm satisfied that the conjecture's true, although I haven't proved it. I'll post what I've got later.
We can continue with this logic to find a formula for an integer which is not divisible by 2, 3 or 5. The formula is y = 3·5a2 + 2·5a3 + 2·3a5, where a2, a3 and a5 are not divisible by 2, 3 or 5, respectively.
More generally, we can devise a formula to find all prime numbers in an interval [pn, pn2], where pn is the nth prime number, if we know all the primes p1 to pn-1. The equation in all it's glory is:
Where ap is any number not divisible by p.
So it's pretty simple, and while it might make it slightly easier to compute primes, the task still requires all the same information as it did before. There are a few questions that remain however:
1) Will all primes have a solution to this equation? I think it's possible to show this by showing that if we replace all the ap's with bp's, where the bp's can be any integers, then any integer y has a solution to the equation.
2) Given an integer, is it possibly to use the equation to determine if it's prime? I think this is actually fairly straight forward. The problem is that it doesn't seem to me like it would be computationally faster than existing methods.
3) Can we use this equation to determine the number of primes in an interval [pn, pn2]? If the solutions can be ordered in a nice way, then I think the answer is yes, otherwise, I think the problem would resemble the Sieve of Eratosthenes in some way.
There are a set of 4 prime number problems known as Landau's problems. I think this formula might be useful for solving some of these. I've already had a go at Goldbach's conjecture and I'm satisfied that the conjecture's true, although I haven't proved it. I'll post what I've got later.