relax SVN archive

Created the add_model() structural API method.

This is implemented within the internal structural object.

Created the test data for the iso cone, free rotor frame order model.

Shifted the add_molecule() method to the correct position.

All private methods have been renamed by replacing '__*()' with '_*()'.

Added MC sims to the Frame_order.test_cam_rotor system test.

This is just a quick check with 3 simulations to make sure the code is functional.

Added Monte Carlo simulations to the second rotor frame order model testing script.

This is to check that the PyMOL representations are correct, as this cannot be done in a system
test.

Monte Carlo simulations are now implemented for the frame order analysis.

Eliminated one of the eigenframe parameters from the rotor frame order model.

In the rotor frame order model, the x and y axes of the eigenframe of the motion are not defined.
Therefore the 3 eigenframe Euler angles have been replaced by 2 spherical angles defining the
rotation axis. This removes a lot of optimisation issues with this model.

Fixes for the rotor frame order system test data.

The first structure of the distribution was missing!

The frame order torsion angle cone_sigma_max is now forced to be positive.

Optimisation sometimes comes up with the negative of the angle, as both the negative and positive
angles give the same result, so now the angle is always unpacked after optimisation as the absolute
value.