Documentation issues

[ThomasCeulemans]

As the Magritte library is still quite new, the documentation is not yet on point everywhere. This issue will contain references to parts of the library which do not yet have a clear documentation. It will be updated when parts are noted where the documentation may be improved.

Current documentation issues:

• [ ] in Solver.tpp: solve_feautrier_order2: This is a giant computation with many constants, so it is not very clear how and what we are calculating. Either add a reference to a paper for this calculation, or add some more comments. Same issue with solve_shortchar_order_0 and update_Lambda.
• [x] In LineProducingSpecies: update_using_Ng_acceleration: same comment, too many constants without knowing what they are. Probably best to just add a reference to a paper on Ng-acceleration (although it should be well-known in the astrophysics community) or optionally: add some psuedocode with the calculation in a few lines.
• [ ] In a lot of .hpp files with a definition of a vector: also put some indication of what the indices represent (eg: species.hpp; Double2 abundance; ///< (current) abundance in every cell (point, species); also please add a bit more than a single letter to represent something (e.g. in radiation.hpp Tensor I; ///< intensity (r, p, f) /change to/ (ray, point, frequency)

Unused stuff that probably no longer needs to be there:

• [ ] In species.hpp/tpp: Is abundance_init still used in anything?
• [x] In multigrid branch: delete the Neighbors struct, superseded by Multiscale
• [x] In both branches: far too many commented-out functions (which are no longer used) around
• [ ] Scattering struct: seems not to do anything anymore

[ThomasCeulemans]

Also another thing that is not yet finished:

• [x] The readthedocs documentation is incomplete

[ThomasCeulemans]

And yet another issue:

• [x] Explicitely state that we use SI units for all input variables (currently minus the velocities, which is a fraction of the light speed)