relax SVN archive

Reverted r as it contains several unrelated changes.

The command used was:
svn merge -r 13931:13924 .

This commit contains several unrelated changes in /generic_fns/structure/main.py and /prompt/structure.py.

Redesign of the structure.load_spins user function back end.

This redesign is required due to problems encountered by Han in the xyz branch. See the thread
starting at https://mail.gna.org/public/relax-devel/2011-07/msg00021.html.

This commit consists of a number of changes:
- The combine_models argument has been removed as it is a relic, and isn't used at all.
- The spins are now created in the standard way via the generic_fns.mol_res_spin.create_spin()
function, eliminating all of the issues and complications of the old code.
- The atomic positions are now converted to numpy arrays.
- Only the created spins are included in the print out, and not replicated for each model.

Changing the value of the unit arg in the function 'vectors' in 'prompt/structure.py'

The argument of the keyword 'unit' in the function 'vectors' in 'prompt/structure.py' has been changed back from 'False' to 'True' since it causes many test failure.

Merged revisions 13926-13927,13929 via svnmerge from
svn+ssh://bugman@.../svn/relax/1.3

........
r13926 | bugman | 2011-07-27 16:01:22 +0200 (Wed, 27 Jul 2011) | 5 lines

The create_molecule, create_residue, and create_spin functions now returned the created container.

This should not affect any of the current code using these, as normally None would be returned.
........
r13927 | bugman | 2011-07-27 16:05:49 +0200 (Wed, 27 Jul 2011) | 6 lines

Invisible fix for the test_opendx_theta_phi_da() system test.

The spins are extracted from the structure, so they don't need to be read from the relaxation data
file as well. This is causing problems in the xyz branch, but is not an issue in the main line.
........
r13929 | bugman | 2011-07-27 16:15:41 +0200 (Wed, 27 Jul 2011) | 3 lines

Modified the create_spin() function to overwrite the first spin if empty.
........

Modified the create_spin() function to overwrite the first spin if empty.

Fix for the dauvergne_protocol auto-analysis - the analysis status is in status.auto_analysis!

Invisible fix for the test_opendx_theta_phi_da() system test.

The spins are extracted from the structure, so they don't need to be read from the relaxation data
file as well. This is causing problems in the xyz branch, but is not an issue in the main line.

The create_molecule, create_residue, and create_spin functions now returned the created container.

This should not affect any of the current code using these, as normally None would be returned.

Created transparent background images of the diffusion tensors for the model-free auto-analysis tab.

Deleting the function 'vector_xyz' in the generic_fns/structure/main.py

The code of the function 'vector_xyz' was removed since it will be not used.