relax SVN archive

Converted the Relax_disp.test_sprangers_cpmg_data_auto_analysis system test to not use the auto-analysis.

The test has been renamed to Relax_disp.test_sprangers_data_to_mq_ns_cpmg_2site. The optimisation
is now for the cluster and has been severely cut back. The 'MQ NS CPMG 2-site' model appears to be
rubbish anyway - it looks to be indeterminate with multiple solutions, and possibly infinite lines
of solutions. The test now passes, and quickly.

Artificially increased the errors in Sprangers ClpP data to match the publication.

The R2eff errors are simply multiplied by 5, as the errors from the paper cannot be replicated.

Fixes for the R2eff data files for Sprangers ClpP data.

Added a script and results files for the base 'R2eff' model for Remco Sprangers' ClpP data.

Simplified the Relax_disp.test_sprangers_cpmg_data_auto_analysis system test script.

The pA and kex parameters are now also pre-set to speed things up.

Changes to the Sprangers ClpP data analysis script.

Manually reverted the temporary change of r21094.

The command used was:
svn merge -r21094:r21093 .

.....
r21094 | bugman | 2013-10-14 15:40:53 +0200 (Mon, 14 Oct 2013) | 5 lines
Changed paths:
M /branches/relax_disp/test_suite/system_tests/relax_disp.py

Temporarily turned off the Relax_disp.test_sprangers_cpmg_data_auto_analysis system test.

This commit must be reverted!
.....

Modified the Relax_disp.test_sprangers_cpmg_data_auto_analysis system test so the models are programatically changed.

Replaced all usage of scipy.linalg.expm() with lib.linear_algebra.matrix_exponential.matrix_exponential().

This is for the functions of the lib.dispersion package used for the relaxation dispersion numeric
solution models. The change eliminates a bug in the scipy function which uses the Pade
approximation which fails horribly for the complex part of the matrix. The real part looks good,
but the complex part looks to have nasty truncation artefacts which is propagated and amplified
through the Bloch-McConnell equations.

Merged revisions 21111 via svnmerge from
svn+ssh://bugman@.../svn/relax/trunk

........
r21111 | bugman | 2013-10-15 13:20:25 +0200 (Tue, 15 Oct 2013) | 5 lines

Fix for the new lib.linear_algebra.matrix_exponential.matrix_exponential() function.

This function now returns a numpy array type rather than matrix type.
........