relax SVN archive

The CaM frame order system test base script now loads the molecule name from the PCS files.

This is needed to match the spin IDs of the loaded spins.

Changed the molecule name to 'C-dom' in all of the CaM frame order PCS test data.

Mass conversion of all of the CaM frame order RDC data used in the test suite.

This converts the files to the new interatomic data container design so that the file consists of
two spin IDs and the RDC and error.

Reverted r17258 as a special script will be used for mass conversion of all RDC data.

The command used was:
svn merge -r17258:r17257 .

....
r17258 | bugman | 2012-07-16 19:46:31 +0200 (Mon, 16 Jul 2012) | 3 lines
Changed paths:
M /branches/frame_order_testing/test_suite/shared_data/frame_order/cam/rotor/rdc_dy.txt
M /branches/frame_order_testing/test_suite/shared_data/frame_order/cam/rotor/rdc_er.txt
M /branches/frame_order_testing/test_suite/shared_data/frame_order/cam/rotor/rdc_tb.txt
M /branches/frame_order_testing/test_suite/shared_data/frame_order/cam/rotor/rdc_tm.txt

Converted the test suite CaM RDC data for the rotor frame order model to the interatomic design.
....

Fix for the structure.get_pos user function backend.

The structural API atom_loop() method has been rewritten in this branch to be much faster, and the
new version of this method does not work with models.

Fix for the frame order target function set ups for when no PCS data is present.

The numerical integration using the multi-processor framework should not and cannot be used without
PCS data.

Eliminated the analysis specific API _set_param_values() base method.

This is now in two parts, the _set_param_values_global() and _set_param_values_spin() methods. The
frame order analysis in the trunk had switched to _set_param_values_global(), and now this branch
does the same thing. Not sure why this change did not come across.

Fix for the frame_order.ref_domain call in the base script for the frame order CaM system tests.

Reimplementation of r15046 but within the auto-generated user function design.

The 'fix' flag of the frame_order.pivot user function was lost as the changes went into
prompt/frame_order.py.

.....
r15046 | bugman | 2011-12-07 18:04:11 +0100 (Wed, 07 Dec 2011) | 5 lines
Changed paths:
M /branches/frame_order_testing/prompt/frame_order.py
M /branches/frame_order_testing/specific_fns/frame_order.py

The frame_order.pivot user function now accepts the fix argument to allow the pivot to be fixed.

This is for optimisation.
.....

Reimplementation of r15044 but within the auto-generated user function design.

The merger of the redesigns from the trunk causes the user function front-end changes of r15044 in
prompt/align_tensor.py to be lost, as this file no longer exists. The align_tensor.init user
function arguments align_id and domain have been reintroduced.

.....
r15044 | bugman | 2011-12-07 16:32:01 +0100 (Wed, 07 Dec 2011) | 8 lines
Changed paths:
M /branches/frame_order_testing/generic_fns/align_tensor.py
M /branches/frame_order_testing/prompt/align_tensor.py

Converted the align_tensor.init user function to accept the align_id and domain arguments.

The domain argument complements the align_tensor.set_domain user function whereas the align_id arg
is new and is used to associate a tensor with an alignment. This is part of a redesign of how
alignment tensors are handled, as there can be multiple tensors per alignment if the molecule is
multi-domain.
.....