relax SVN archive

Small corrections to r1rho settings script.

Shifted the 'NS CPMG 2-site expanded' model to the top of the CPMG numerical solutions in the manual.

This is because this is the default model which should be used in most cases.

Added test data where both the spin-lock time, the spin lock offset and the spin lock field is varied.

The data is published in:
A Folded Excited State of Ligand-Free Nuclear Coactivator Binding Domain (NCBD) Underlies Plasticity in Ligand Recognition.
Kjaergaard, M., Andersen, L., Nielsen, L.D. & Teilum, K. (2013).
Biochemistry 52, 1686–1693

with experimental conditions that
"off-resonance R1rho relaxation dispersion experiments on 15N were recorded at 18.8 T and 31 C."
"using the pulse sequence of Mulder et al. with spin-lock field strengths from 431 to 1649 Hz and offsets ranging from 0 to 10000 Hz."

Fixes for 2 Relax_disp GUI tests to match the model-free fixes of r21182.

The Fake_right_click.GetPosition() method now returns a wx.Point object.

Fix for a bug introduced at r21202 - the call to the calculate_r2eff() must also be made private.

Fix for the Mf.test_mf_auto_analysis system test on MS Windows.

The Fake_right_click.GetPosition() method now returns a valid position. This is the original
(10, 10) position.

The specific API calculate_r2eff() method for the dispersion analysis is now private.

This is not part of the API, so it must be made private for the test suite to pass.

Loosened the test_hansen_cpmg_data_to_ns_cpmg_2site_star_full system test for Mac OS X.

Even the calc user function does not help, the results are quite different between different systems.

Redesigned the Relax_disp.test_hansen_cpmg_data_to_ns_cpmg_2site_star_full system test.

The R2A and R2B rates cannot be distinguished for this data, therefore there was no unique solution.
This resulted in too much variability between 32 and 64-bit systems as well as different operating
systems. Instead a single calc user function call is used to determine the chi-squared value for a
fixed set of parameters.

Fixes for the calc user function for the dispersion analysis.

This now does something logical for the non-R2eff models. The chi-squared value is now being
calculated and stored. Previously this was only calculating the R2eff/R1rho values for fixed
relaxation time period data for the 'R2eff' model and failing for all others. Now the pre-existing
_back_calc_r2eff() method is used to back-calculate and store the chi-squared value.