relax SVN archive

Corrected an error in the definition of eta.

This difference arises from a different definition of the constant c... Basically, this only affects
the obtained F_eta value by dividing it by a factor of sqrt(3) independent of the magnetic field.
Hence, consitency evaluation was fine before this correction, except that values were multiplied by
sqrt(3).

Also, the test-suite values for F_eta were changed so the test still passes.

These changes follow a thread started at:
https://mail.gna.org/public/relax-devel/2008-06/msg00013.html
(message ID: 48626A6B.1010104@...)

and continued at:
https://mail.gna.org/public/relax-devel/2008-07/msg00001.html
(message ID: 486D13EB.1000100@...).

Improved the diffusion tensor system tests.

Moved the PDB and sequence loading with the rest of the diffusion tensors initialization and
uncommented the sphere and spheroid parts of the structure.create_diff_tensor_pdb() function test.

Added a missing import.

Added the loading of a sequence for the corresponding PDB.

Corrected a path from relative to absolute.

Corrected a test for the loading of a PDB.

This is one step towards the structure.create_diff_tensor_pdb() function working.

Added a test within the diffusion tensor system tests.

This one tests the structure.create_diff_tensor_pdb() function.

Fix for the function arg unit test of eliminate().

Added a missing import.

Caught all tuple types in the args arg unit test of eliminate().