relax SVN archive

Updated the relaxation dispersion target function class to handle the recent changes.

Expanded the relax_disp.select_model user function documentation to cover R2eff and I0.

These parameters and how they are optimised are now better described.

Created the dispersion specific _exp_curve_loop() method for looping over each exponential curve.

This yields the index and key for each curve, simplifying the handling of this data.

The dispersion specific _assemble_param_vector() method now handles multiple R2eff and I0 values.

These spin structures are dictionaries holding multiple values.

Better management of the global relaxation dispersion data.

The user functions relax_disp.cpmg_frq, relax_disp.spin_lock_nu1, and relax_disp.relax_time now
maintain data structures in the current data pipe of the unique frequencies, fields, and times
(sorted) as well as the number of frequencies, fields, and times. This data is used by the minimise
user function back end to set up the target function, and will be required by many other parts of
the analysis.

The dispersion specific methods now handle one R2eff and I0 parameter per exponential curve.

Fixes for the dispersion specific analysis separating R2eff from R2.

There is one R2eff parameter per exponential curve, but only one R2 per model. The code now better
handles this.

Renamed cdp.curve_type to cdp.model to better explain the variable.

Better setup checking for the relaxation dispersion specific analysis minimise() method.

The relaxation dispersion analysis specific _param_num() method now takes the number of curves into account.