Incorporate internally consistent probe and source coordinates

Is your feature request related to a problem? Please describe. At present, the different classes rely on different parameters names declaring extracellular recording sites:

PointSourcePotential and similar uses ndarrays x,y,z in Cartesian coordinates,
OneSphereVolumeConductor uses ndarray r in spherical coordinates,
FourSphereVolumeConductor uses r_electrodes in Cartesian,
InfiniteHomogeneousVolCondMEG uses sensor_locations etc.
For class InfiniteVolumeConductor the user must specify relative measurement location r as input to the get_transformation_matrix,
NYheadmodel something else altogether

These internal inconsistencies needs to be resolved and conform to the same structure, e.g.,

class Whatever(LinearModel):
    '''Some description

    Parameters
    ------------
    cell: CellGeometry object
    r_probe: ndarray
        shape (n_contacts, 3) array of extracellular measurement locations in absolute Cartesian coordinates (µm)
    '''
    def __init__(self, cell, r_probe, **kwargs):
        super.__init__(cell=cell)
        self.r_probe = r_probe
        ....

In a similar vein the source locations in case of dipoles should be declared with the same parameter, e.g., r_source.

A good timeline to incorporate would be for future LFPy-2.3 (https://github.com/LFPy/LFPy/projects/2) which may break backwards compatibility anyway.

LFPy / LFPykit

Incorporate internally consistent probe and source coordinates #158