use new orientation for magnetic models

[pkienzle]

theta and phi for magnetism direction do not operate like theta and phi. Consider changing to a latitude/longitude coordinate system. Make sure it is clear from the docs.

[pkienzle]

Setting the magnetic orientation relative to the particle would be more convenient than setting it relative to the lab frame.

If magnetic orientation is relative to the particle, then particle orientation distribution should lead to a magnetic orientation distribution. However, symmetric particles would then need psi and possibly theta and phi orientations to control the distribution.

Magnetic saturation leads to orientation in the lab frame. This can be done by tying magnetic orientation to negative particle orientation, but it will not work correctly with orientation distribution on the particles.