relax SVN archive

Wrote the pcs.back_calc() user function back end.

Initialising and alignment tensor now adds the ID to the alignment ID list.

Fix for the alignment tensor rotation matrix - it is now always in the right handed universe.

Added a system test checking the conversion of alignment tensor to chi tensor.

Major patch for integrating the GUI into the noe calculation.

This patch is from Michael Bieri (https://gna.org/users/michaelbieri) and is from file #9490
(https://gna.org/support/download.php?file_id=9490) attached to task #6847
(https://gna.org/task/?6847).

Changes include:
- output filename can be specified
- folder, where results files will be placed, can be specified
- label of hetero nucleus and proton of peak lists and pdb file can be selected
- sequence is read either by sequence file or pdb file
- removed white spaces and progress output.

Merged revisions 11291-11296 via svnmerge from
svn+ssh://michaelbieri@.../svn/relax/1.3

........
r11291 | bugman | 2010-07-08 19:45:21 +1000 (Thu, 08 Jul 2010) | 3 lines

Docstring improvements - the geometric info should be in Angstroms.
........
r11292 | bugman | 2010-07-08 19:49:35 +1000 (Thu, 08 Jul 2010) | 9 lines

Better support for the optimisation and mapping of the Ln3+ position.

The position is now treated in Angstrom, and is properly scaled. A few additional API methods have
been added for the mapping of the optimisation space, and a few modified to support the Ln3+
position.
........
r11293 | bugman | 2010-07-08 19:56:42 +1000 (Thu, 08 Jul 2010) | 3 lines

Fix for the Ln3+ position optimisation with no populations, the target function was overwritten.
........
r11294 | bugman | 2010-07-08 21:03:13 +1000 (Thu, 08 Jul 2010) | 5 lines

Fix for the N-state model _assemble_scaling_matrix() method for the 2-domain model.

This is a recently introduced bug. For the 2-domain model, no alignment parameters are optimised.
........
r11295 | bugman | 2010-07-08 21:49:04 +1000 (Thu, 08 Jul 2010) | 3 lines

Changed the alignment tensor parameter scaling back to 1 as this was slowing down the optimisation.
........
r11296 | bugman | 2010-07-08 23:26:24 +1000 (Thu, 08 Jul 2010) | 3 lines

Created the n_state_model.elim_no_prob() to remove structures with zero probability.
........

Created the n_state_model.elim_no_prob() to remove structures with zero probability.

Changed the alignment tensor parameter scaling back to 1 as this was slowing down the optimisation.

Fix for the N-state model _assemble_scaling_matrix() method for the 2-domain model.

This is a recently introduced bug. For the 2-domain model, no alignment parameters are optimised.

Fix for the Ln3+ position optimisation with no populations, the target function was overwritten.