relax SVN archive

Removed some debugging printouts accidentally committed in the last commit (r21930).

Bug fix for the dispersion grid_search_setup() optimisation function.

This function was not updated for the recent addition of the spin-lock or hard pulse offset
dimension in the specific_analyses.relax_disp.disp_data module (and hence all structures used by the
dispersion target functions). The loop_exp_frq_point() function call has been replaced by a
loop_exp_frq_offset_point() function call to allow the 'R2eff' model parameters to be looped over.

For more details, see the thread http://thread.gmane.org/gmane.science.nmr.relax.scm/19685. This
solution was mentioned at http://thread.gmane.org/gmane.science.nmr.relax.scm/19685/focus=4859.

Epydoc docstring fix for the lib.dispersion.ns_mmq_3site.r2eff_ns_mmq_3site_sq_dq_zq() function.

This allows the API to be compiled correctly.

Capitalised 'Python' in the IO redirection messages.

Fix for the Relax_disp.test_bug_21076_multi_col_peak_list GUI test.

The peak intensity wizard is now closed at the end of the test so that subsequent tests can cleanly
operate. Without closing this wizard, launching it a second time in another test will always fail.

System test turned off until release is over.

Work in progress for Support Request #3044, (https://gna.org/support/index.php?3044) - Load spins from SPARKY list.

Added the ns_r1rho_3site module to the lib.dispersion package __all__ list.

This allows the unit tests to pass.

Fix for wrong use of relax_fit.relax_time instead of relax_disp.relax_time.

Work in progress for Support Request #3044, (https://gna.org/support/index.php?3044) - Load spins from SPARKY list.

Created the new Relax_disp.test_ns_r1rho_3site system test.

This was copied from the Relax_disp.test_ns_r1rho_3site_linear test and modified to use the new
'NS R1rho 3-site' model synthetic data.

Created synthetic R1rho dispersion data for the 'NS R1rho 3-site' model.

This is a simple modification of the data for the 'NS R1rho 3-site linear' model. The k_AC
parameter was simply changed from 0 to 1000. The cpmg_fit software was used to create the data.
Both cpmg_fit and relax results have been updated to the new model.