Bug fix in calc_from_real back calculations
Calculation of the pseudo motors back was wrong for all the geometries, except 2,6 and 8 (there was + instead of -). That made the table drift when using the pseudo motors.
Added some documentation, ran black and pylint.