refactored code to include a _initialise_structure function, a comput…

[AndySAnker]

…e_iq function, etc. This modulates the code and makes it shorter. It also makes it possible to calculate Fq and Gr directly from Sq data, which can also be done on experimental datasets.

[FrederikLizakJohansen]

I've done some benchmarking to see if this version performs as well as the old, and it seems to be!

• Running gr with low statistics (10 reps) on bigger particles reveals no difference in performance. Although at generation time, we do see a small increase in memory usage. This however is so small in comparison to the memory at calculation time, that I would argue that the utility of this refactored version outweighs the small increase in memory.
• Running gr with high statistics (1k reps) on smaller particles sizes (needed for consistency), reveal no noticeable difference. Although again we see the same, but insignificant, increase in memory at generation time.

[AndySAnker]

Great work Frederik, I propose that we merge.