Small codebase updates

[MonkeyBreaker]

This PR addresses multiples small issues here and there in the code base that where on my to-do list:

• : Add templates to apparent pair related code to remove code logic duplication
• : Add verification of coefficient parameter. Checks that the parameter is lower than the maximal value supported and if the value is a prime number
• : Update naming of variables inside of binomial_coeff_table as requested by @ulupo
• : Update barcodes returned from the bindings to directly be of Numpy type, because we want to match float 32-bit precision used instead of Python default one float 64-bit precision. The return value is therefore composed of Numpy arrays.
• : Fixed an issue with recent version of GCC compiler with a header dependency that was removed from algorithm header.
• : Remove factor parameter who was unused
• : Remove num_edges return value who is unused and if needed, can be easily computed in Python.

[MonkeyBreaker]

@ulupo if you think of anything else that should be added to this PR.
Please let me know, I am not sure additional things can/should be added on this PR, it addresses small things I got in my to-do list.

[ulupo]

Great effort, thank you very much! A quick comment before a full review: I was under the impression that Ripser/ripser implemented its own is_prime function in C++. What is the reason for not incorporating it here?

[MonkeyBreaker]

I was under the impression that Ripser/ripser implemented its own is_prime function in C++. What is the reason for not incorporating it here

Well, the function is_prime was removed from the ripser source file.
In my opinion, a is_primefunction should be more in a utility file than inside of the ripser PH computation file.
The only reason I would think to have the C++ is_prime function back, is if this becomes a bottleneck computation.
@ulupo, what do you think ?

[ulupo]

Great work, thanks!