relax SVN archive

Updated the cpmg_fit results for the Korzhnev et al., 2005 single data sets.

The starting point for optimisation is now the solution for using all data together. This allows a
much better solution to be found for each script.

Added cmpg_fit input and results files for all single CPMG data combinations from Korzhnev et al., 2005.

Added cmpg_fit input and results files for the 15N SQ CPMG data from Korzhnev et al., 2005.

Big redesign of the dispersion point returning and loop_*() functions.

These are the functions in the specific_analyses.relax_disp.disp_data module. The
return_cpmg_frqs() and return_spin_lock_nu1() functions now no longer take the spins and spin_ids
arguments. Instead they determine if a dispersion point exists for the given experiment and
spectrometer frequency using the intensity keys and data in the base of the data pipe.

The specific_analyses.relax_disp.disp_data.loop_*() functions now accept the return_indices argument
which if True will cause all of the relevant experiment type, spectrometer frequency, dispersion
point, and relaxation time indices to be returned.

The behaviour of the loop_point() method is now different. Instead of looping over all possible
dispersion points, it only loops over those points present for the given experiment and spectrometer
frequency. This change allows for many simplifications and latent bug fixes in the dispersion
analysis.

Fixes for the dispersion optimisation Disp_result_command.run() method.

The dispersion point loop is fixed over all dispersion points, but the 'missing' data structure has
a variable length.

Created the Relax_disp.test_korzhnev_2005_15N_sq_data system test.

This is used to check the optimisation of the 15N SQ CPMG data using the 'MMQ 2-site' model.

Removed some debugging printouts.

Added a relax script to optimise just the 15N SQ CPMG data from Korzhnev et al., 2005.

The corresponding log file has also been added to the repository.

Changed an import order.

Different relaxation time periods for each experiment is now taken into account in the dispersion code.

Previously only the first relaxation time period was being used. This was fine for single data type
models, but was preventing the MMQ-type models from working. Now the return_r2eff_data() function
of the specific_analyses.relax_disp.disp_data module assembles and returns the relax_times data
structure which has two dimensions - the experiment type and the magnetic field strength.