Feature/magn

[Alvkuzin]

The magnetic field is added. I created a new file magn from which I import the needed stuff to other modules. The two main additions are:

1. In BaseVerticalStructure, I added two functions returning zero but to be overwritten by mixins: an additional pressure at the photosphere and an additional source of $dP/dz$. Then I created a class MagneticField which defines these functions (I collected all magn. fields args into a dictionary magn_args) and passed this MagneticField to BaseVerticalStructure or BaseMesaVerticalStructure as a mixin, creating a "magnetic" class for every VS non-mangetic one. I did it for all non-magnetic VS classes without irradiation; irrad. classes were not changed. Everywhere below when I say: 'it was tested for everything' or 'everything takes a new argument', I mean "everything that's without an irradiation".

2. I created a class ViscousTorque which takes the same dictionary + some other parameters. This class takes care of everything that can be found prior to the VS calculation: torques, inner radius, Teff..., including the functions that determine $\dot{M}$ from $T_\mathrm{eff}$ or $F_\mathrm{vis}$. In case of magn_args=None, it gives the standard expressions: $r_{in} = r_{ISCO} = 3r_g, F = F_{in} + \dot{M} h (1 -\sqrt{(r_{in}/r)} )$, while for the magnetic field, the equations are generalized. The inner radius is now found using the function magn.radius_inner , and there is a magnetic contribution $F_\mathrm{magn}$ to $F$ calculated by magn.magnetic_torque. Note that the mixin MagneticField also initializes ViscousTorques inside, as it needs to calculate $r_0$ from $F$ (e.g., it recalculates $F \rightarrow \dot{M} \rightarrow r_{in}$).

So introducing these two main things, I also rewrote all references to the inner radius and viscous torques everywhere in profiles, e.g., now in StructureChoice the class Torques = ViscousTorques(...) is created, and it takes care of converting Mdot <-> F <-> Teff, and calculates r_in. Now in StructureChoice, if magn_args is None, the old classes are initialized, while if magn_args is not None, the new magnetic classes for VS are chosen. Subsequently, other classes of profiles: VerticalProfile, S_Curve, and Radial_Structure are now taking magn_args as an argument.

The BaseVerticalStructure, and all subsequent functions (VS classes and classes defined in profile) can now take a boolean argument rad_pressure_include which determines whether to take the radiation pressure into account in calculations. Everywhere when it was clear that something like $...+4\sigma_b T^4/3c $ appears, I changed it to rad_pressure(T) that returns the same if rad_pressure_include==True and zero if rad_pressure_include==False.

The classes VerticalProfile, S_Curve, and Radial_Structure now can return namedtuples (if to_return=True) containing some results of calculations. These classes now all have an argument to_save which tells whether to save results in a file. They all now have their corresponding verbose_scurve/verbose_rad arguments which determine whether to print the information as the calculation proceeds.

All additions to the existing methods and functions were made as to follow the backward compatibility. If a user uses the new model as they were using it before, nothing should change for this user.

Using the provided 'tests', I made sure that the tests are successful for the non-magnetic case.

I changed ReadMe slightly; I updated a version number in pyproject.toml, and added a .gitignore.