relax SVN archive

Removed a double full stop in the relax_disp.select_model docstring.

Fix for the model equivalence setup in the nesting() method of the dispersion auto-analysis.

This is the use of the analytic CR72 model parameters for the numeric models to avoid the grid
search.

Updated the Noe.test_noe_analysis system test for the changes of r20385.

These were the changes to the lib.software.grace.write_xy_header() function.

Added the 'LM63 3-site' dispersion model to the model list in the GUI.

It is not selected by default.

Added a warning not to use the 'LM63 3-site' model to the dispersion chapter of the user manual.

Added the relax results for the 'LM63 3-site' dispersion model test data.

Added the CPMGFit results for the 'LM63 3-site' dispersion model test data.

Another update of the 'LM63 3-site' dispersion model test data.

This data now makes CPMGFit happy!

Modified the 'LM63 3-site' dispersion model test data again.

This time the data has been changed to be that of two residues rather than two spins.

Made the 'LM63 3-site' dispersion model test data more realistic.

Previously all the rates were within a few decimal places of the R20 values. Now the dispersion is
much more significant.