• added more sample tests
• added tester function for better error feedback
• increased random tests from 5 to 100
• provided a wider range of random tests from 0 <= n <= 100 to ... ?
• made reference solution static
• added missing headers for random testing
• added srand() to seed actual randomization!
• streamlined testing format
• added larger tests to fixed tests
• made fixed / random tester static
• updated language verion from C11 to Clang 8/C18

Hello!

I don't know if solving it in this way has a name. Maybe constraint satisfaction?

I think it is one natural way of approaching optimization for the problem: constrain the next subset to be explored based on the current subset.

I'm interested in this O(f()^2 * logf()) approach that you mentioned. I'd like to learn this optimization. Could you point me to any online reference?

EDIT: I found your submission. Is this a popular approach? I can't find similar applications online. Would appreciate some references.

I like this solution because it does the minimum amount of programmer work.

This solution is worse than O(n^3) because of the nested for loops, making this an extremely costly solution in time. Storing each set of kPrimes and iterating over it would be a simple optimization. It can further be reduced by restricted the set a based on b and c (taking it to something less than O(f()^2 * logf()), where f() is based on # of primes below s). There are further improvements and maybe someone would have more insight into the time complexity.

Rejected as there are an insufficient amount of fixed assertions and no random assertions in the Submit tests.