Author: bugman
Date: Tue Jun 15 14:16:56 2010
New Revision: 11242
URL: http://svn.gna.org/viewcvs/relax?rev=11242&view=rev
Log:
Merged revisions
10937-10958,10961-10990,10999-11002,11004-11050,11053-11059,11070-11071,11085,11087-11099,11144-11175,11185-11189,11206-11210,11223-11224,11230-11233,11236,11240
via svnmerge from
svn+ssh://bugman@xxxxxxxxxxx/svn/relax/1.3
........
r10937 | bugman | 2010-02-25 12:36:30 +0100 (Thu, 25 Feb 2010) | 10 lines
Addition of the relax status singleton object.
See https://mail.gna.org/public/relax-devel/2010-01/msg00094.html and other
related posts in the
thread.
The idea is that various parts of relax deposit status information into
here that other parts of
relax can then use. For example the Monte Carlo code can place the current
simulation number in
there while the relax GUI controller can read these values.
........
r10938 | bugman | 2010-02-25 13:52:35 +0100 (Thu, 25 Feb 2010) | 6 lines
The Monte Carlo simulation number is now present in the status singleton.
The value of status.mc_number is set to the simulation number (starting at
0) and at the end is
reset to None.
........
r10939 | bugman | 2010-02-25 14:37:11 +0100 (Thu, 25 Feb 2010) | 11 lines
All of the relevant status information about the dAuvergne_protocol
auto-analysis is being stored.
This includes:
status.dAuvergne_protocol.diff_model = None # The global
diffusion model.
status.dAuvergne_protocol.round = None # The round of
optimisation, i.e. the global iteration.
status.dAuvergne_protocol.mf_models = None # The list of
model-free models for optimisation, i.e. the global iteration.
status.dAuvergne_protocol.local_mf_models = None # The list of
model-free local tm models for optimisation, i.e. the global iteration.
status.dAuvergne_protocol.current_model = None # The current
model-free model.
status.dAuvergne_protocol.convergence = False # The convergence of
the global model.
........
r10940 | bugman | 2010-02-25 14:59:08 +0100 (Thu, 25 Feb 2010) | 7 lines
Bug fix and updates to the generic_fns.angles.wrap_angles() function.
This was not wrapping correctly when the window was not between 0 and 2pi.
Epydoc tags have been
added to the docstring, the window size can now be specified and is checked
against the bounds as a
sanity check, and comments have been added all over.
........
r10941 | bugman | 2010-02-25 15:15:09 +0100 (Thu, 25 Feb 2010) | 5 lines
Fix for a bug introduced in the last revision (r10940).
This affects the spherical angles used to define the spheroidal diffusion
tensor major axis.
........
r10942 | bugman | 2010-02-25 15:16:53 +0100 (Thu, 25 Feb 2010) | 5 lines
Added a new function fold_spherical_angles() for removing symmetry from
spherical angles.
This will be useful for future use of these angles.
........
r10943 | bugman | 2010-02-25 15:19:38 +0100 (Thu, 25 Feb 2010) | 5 lines
The fold_spherical_angles() performs a second wrapping in case the angles
shift out of the window.
Window size checks have also been added.
........
r10944 | bugman | 2010-02-25 15:25:02 +0100 (Thu, 25 Feb 2010) | 5 lines
Added default lower and upper bounds to fold_spherical_angles().
More epydoc tags have been added.
........
r10945 | bugman | 2010-02-25 15:31:51 +0100 (Thu, 25 Feb 2010) | 3 lines
The spheroidal diffusion tensor major axis angles are now folded using
angles.fold_spherical_angles().
........
r10946 | bugman | 2010-02-25 15:55:32 +0100 (Thu, 25 Feb 2010) | 3 lines
Fix for the angles system test for the new folding of the spherical angles.
........
r10947 | bugman | 2010-02-25 16:37:43 +0100 (Thu, 25 Feb 2010) | 7 lines
Removed the iteration count check in the test_opt_rigid_rand_rot() system
test.
This number is different on a PowerPC chip with Mac OS X, even if the
parameter values are
identical. This was identified by Jack Howarth (howarth att bromo dot med
dot uc dot edu) in the
post https://mail.gna.org/public/relax-users/2010-02/msg00025.html.
........
r10948 | bugman | 2010-02-26 16:36:59 +0100 (Fri, 26 Feb 2010) | 9 lines
Partially removed the need for Fink specific modifications.
This removes the need for the patch given by Jack Howarth <howarth att
bromo dot med dot uc dot edu>
in the message
https://mail.gna.org/public/relax-users/2010-02/msg00031.html.
A similar python specific path was being used to find Python.h on all
architectures, so the include
path was copied and modified to be a bin path.
........
r10949 | bugman | 2010-03-09 11:06:06 +0100 (Tue, 09 Mar 2010) | 6 lines
Bug fix for the N-state model cone PDB creation _cone_pdb() method.
The internal structural API had changed. structure.add_struct() is now
called
structure.add_molecule().
........
r10950 | bugman | 2010-03-09 11:11:05 +0100 (Tue, 09 Mar 2010) | 5 lines
Fix for the N-state model user function n_state_model.cone_pdb().
This was not calling the correct method.
........
r10951 | bugman | 2010-03-09 11:15:26 +0100 (Tue, 09 Mar 2010) | 5 lines
Removed the pipe test from the pymol.cone_pdb() user function.
This check is not necessary as a data pipe is not needed for this.
........
r10952 | bugman | 2010-03-09 11:44:48 +0100 (Tue, 09 Mar 2010) | 5 lines
Added a new module for holding class objects defining different cone types.
This currently contains the Iso_cone class with the simple method phi_max()
to return the cone edge.
........
r10953 | bugman | 2010-03-09 11:50:48 +0100 (Tue, 09 Mar 2010) | 5 lines
The cone_edge() method now calls a function to obtain the maximum polar
angle.
The cone_edge() arg 'angle' has been changed to 'phi_max_fn'.
........
r10954 | bugman | 2010-03-09 11:51:38 +0100 (Tue, 09 Mar 2010) | 3 lines
The frame order and N-state model cone_pdb() methods now use the isotropic
cone object.
........
r10955 | bugman | 2010-03-09 13:14:32 +0100 (Tue, 09 Mar 2010) | 3 lines
Added a new cone type in the Pseudo_elliptic class.
........
r10956 | bugman | 2010-03-09 15:29:59 +0100 (Tue, 09 Mar 2010) | 3 lines
The limit_check() method has been added to the cone objects to determine if
a point is inside or not.
........
r10957 | bugman | 2010-03-09 19:11:18 +0100 (Tue, 09 Mar 2010) | 7 lines
Converted generate_vector_dist() to use a limit checking function and
simplified the code.
The angle arrays are now returned by uniform_vect_dist_spherical_angles()
so that
generate_vector_dist() can use any spherical point distribution. The
max_angle arg has been changed
to limit_check so that any type of vector distribution can be created.
........
r10958 | bugman | 2010-03-09 19:23:46 +0100 (Tue, 09 Mar 2010) | 5 lines
Shifted the calculation of the uniform sphere point positions into
angles_uniform().
This removes this replicated code from various functions, allowing for more
flexibility.
........
r10961 | bugman | 2010-03-10 10:07:35 +0100 (Wed, 10 Mar 2010) | 3 lines
Modified stitch_cone_to_edge() to use the new cone limits function.
........
r10962 | bugman | 2010-03-10 16:59:32 +0100 (Wed, 10 Mar 2010) | 9 lines
The diffusion tensor within the PDB frame can now be specified.
This avoids all Euler angle convention problems.
-This line, and those below, will be ignored--
M prompt/diffusion_tensor.py
M generic_fns/diffusion_tensor.py
........
r10963 | bugman | 2010-03-10 17:33:53 +0100 (Wed, 10 Mar 2010) | 5 lines
Shifted the model-free _back_calc() method into the analysis specific API
as back_calc_ri().
This is now being called by the relax_data.back_calc() user function.
........
r10964 | bugman | 2010-03-10 17:38:17 +0100 (Wed, 10 Mar 2010) | 3 lines
The back_calc_ri() analysis specific API method has been activated.
........
r10965 | bugman | 2010-03-10 17:39:33 +0100 (Wed, 10 Mar 2010) | 6 lines
Fixes for the back_calc() method.
The find_index() function was already defined (so renamed to
find_ri_index()), and the analysis
specific API back_calc_ri() method was incorrectly called.
........
r10966 | bugman | 2010-03-10 18:02:28 +0100 (Wed, 10 Mar 2010) | 3 lines
Comment fixes - relax handles spins, not residues now.
........
r10967 | bugman | 2010-03-10 18:27:06 +0100 (Wed, 10 Mar 2010) | 5 lines
Fix for the back_calc_ri() API method.
Spins are now skipped if no bond vector information is present (for the
spheroids and ellipsoids).
........
r10968 | bugman | 2010-03-10 18:28:30 +0100 (Wed, 10 Mar 2010) | 3 lines
Fix for the back_calc() function - if the back-calculation returns no
value, the spin is skipped.
........
r10969 | bugman | 2010-03-10 18:31:25 +0100 (Wed, 10 Mar 2010) | 5 lines
Updated the generate_ri.py sample script to work with the current relax
code base.
This code was old and broken. A copyright notice has been added.
........
r10970 | bugman | 2010-03-10 18:45:10 +0100 (Wed, 10 Mar 2010) | 3 lines
The back-calculation of Ri data via model-free and diffusion tensors has
been added to the system tests.
........
r10971 | bugman | 2010-03-10 20:18:42 +0100 (Wed, 10 Mar 2010) | 3 lines
Added the cones module to the __all__ package list.
........
r10972 | bugman | 2010-03-10 20:23:23 +0100 (Wed, 10 Mar 2010) | 5 lines
Added theta_max() methods to all the isotropic cone classes.
This is for determining theta from the phi polar angle.
........
r10973 | bugman | 2010-03-10 20:26:26 +0100 (Wed, 10 Mar 2010) | 3 lines
Modified the cone_edge() function to add atoms for the latitude lines as
well.
........
r10974 | bugman | 2010-03-11 11:31:06 +0100 (Thu, 11 Mar 2010) | 3 lines
Rewrote the stitch_cone_to_edge() function to handle the arbitrarily shaped
cones.
........
r10975 | bugman | 2010-03-11 11:46:06 +0100 (Thu, 11 Mar 2010) | 3 lines
cone_pdb() now takes the new EDG residue for the cone edge into account.
........
r10976 | bugman | 2010-03-11 13:36:22 +0100 (Thu, 11 Mar 2010) | 3 lines
Created the new fn create_cone_pdb() which can be used to create PDB file
representations of cones.
........
r10977 | bugman | 2010-03-11 13:40:24 +0100 (Thu, 11 Mar 2010) | 3 lines
Fix for create_cone_pdb() for when a rotation matrix is supplied.
........
r10978 | bugman | 2010-03-11 13:42:36 +0100 (Thu, 11 Mar 2010) | 5 lines
Bug fix for cone_edge() for when a rotation matrix is supplied.
The latitude points were not being rotated.
........
r10979 | bugman | 2010-03-11 13:44:26 +0100 (Thu, 11 Mar 2010) | 5 lines
Bug fix for create_cone_pdb() for when a rotation matrix is supplied.
The cone central axis was not being rotated!
........
r10980 | bugman | 2010-03-11 13:45:57 +0100 (Thu, 11 Mar 2010) | 3 lines
Renamed the pivot point to the apex in create_cone_pdb().
........
r10981 | bugman | 2010-03-11 15:54:13 +0100 (Thu, 11 Mar 2010) | 5 lines
Various point distributions can now be used for the cone PDB
representations.
This includes the original 'uniform' distribution and now the 'regular'
distribution.
........
r10982 | bugman | 2010-03-12 10:27:05 +0100 (Fri, 12 Mar 2010) | 3 lines
Fix for the cone_edge() function for a non-zero apex position.
........
r10983 | bugman | 2010-03-12 11:43:27 +0100 (Fri, 12 Mar 2010) | 6 lines
Bug fix for back_calc_ri()!
The bug was introduced at r10967. The check for missing data was not
performing correctly, causing
the Monte Carlo simulations to be stuck in an infinite loop!
........
r10984 | bugman | 2010-03-12 11:55:46 +0100 (Fri, 12 Mar 2010) | 5 lines
Bug fix for the generate_vector_dist() function.
The limits function was being called even if not supplied! This change
fixes that.
........
r10985 | bugman | 2010-03-12 15:02:29 +0100 (Fri, 12 Mar 2010) | 5 lines
Removed the pipe check from the pymol.command() user function.
This is not necessary.
........
r10986 | bugman | 2010-03-12 19:20:40 +0100 (Fri, 12 Mar 2010) | 3 lines
Added the import of pymol modules to the dependency checks.
........
r10987 | bugman | 2010-03-12 19:45:37 +0100 (Fri, 12 Mar 2010) | 7 lines
Started to redesign the generic_fns.pymol module to use the pymol python
package.
This involves much renaming of the PyMOL pipe object which is not the PyMOL
executable object. This
attempts to use the pymol python package, and if not available PyMOL will
be run from the command
line as a pipe.
........
r10988 | bugman | 2010-03-12 19:49:27 +0100 (Fri, 12 Mar 2010) | 3 lines
Renamed the prompt and generic_fns pymol modules to pymol_control to avoid
the clash with real PyMOL.
........
r10989 | bugman | 2010-03-12 19:56:55 +0100 (Fri, 12 Mar 2010) | 5 lines
Renamed the initialised PyMOL executable object from 'pymol' to 'pymol_ob'.
This is to avoid the name clash with the pymol python package.
........
r10990 | bugman | 2010-03-14 12:13:15 +0100 (Sun, 14 Mar 2010) | 3 lines
Added a series of new cone types to generic_fns.structure.cones.
........
r10999 | bugman | 2010-03-15 14:39:01 +0100 (Mon, 15 Mar 2010) | 6 lines
Redesigned the ImportErrorTestCase class so that the test failure clearly
explains what went wrong.
The traceback is now stored and later printed, the exception is not passed
in, rather strings with
the module name and traceback, and comments and docstrings have been added.
........
r11000 | bugman | 2010-03-15 14:48:49 +0100 (Mon, 15 Mar 2010) | 7 lines
Copied the relax installation path into __main__.install_path.
The import of the PyMOL clobbered sys.path!!! Therefore relax does not
touch this array any more.
Instead __main__.install_path is used to store where relax is installed.
This should avoid all
problems with sys.path.
........
r11001 | bugman | 2010-03-15 15:01:04 +0100 (Mon, 15 Mar 2010) | 3 lines
Fixes for the renaming of the pymol modules to pymol_control.
........
r11002 | bugman | 2010-03-15 15:01:28 +0100 (Mon, 15 Mar 2010) | 3 lines
More fixes for the install path change to __main__.install_path.
........
r11004 | bugman | 2010-03-16 00:27:51 +0100 (Tue, 16 Mar 2010) | 3 lines
Renamed the pymol modules to pymol_control in the __all__ lists and unit
tests.
........
r11005 | bugman | 2010-03-16 00:32:39 +0100 (Tue, 16 Mar 2010) | 5 lines
Improvements to the handling of the PyMOL Python packages.
The methods have been alphabetically ordered as well.
........
r11006 | bugman | 2010-03-16 01:34:20 +0100 (Tue, 16 Mar 2010) | 6 lines
Fix for the docstring fetching script.
The relax prompt interface is no longer run as the user functions are
directly located in the
_locals private dictionary.
........
r11007 | bugman | 2010-03-16 01:43:32 +0100 (Tue, 16 Mar 2010) | 5 lines
Converted the user function exec_info object to be private.
This was bad in 2 ways, it was visible to the user and it was killing the
fetch_docstring.py script.
........
r11008 | bugman | 2010-03-16 01:52:03 +0100 (Tue, 16 Mar 2010) | 5 lines
Fixes for some error messages in fetch_docstrings.py.
This should have been writing to stderr, not sys!
........
r11009 | bugman | 2010-03-16 01:55:36 +0100 (Tue, 16 Mar 2010) | 5 lines
Removed the epydoc tags from the on() and off() methods.
These are user functions, and the tags were killing fetch_docstrings.py.
........
r11010 | bugman | 2010-03-16 02:13:12 +0100 (Tue, 16 Mar 2010) | 8 lines
Temporary set up the svnmerge properties to merge the bieri_gui branch back
into the 1.3 line.
The command used was:
svnmerge.py init svn+ssh://bugman@xxxxxxxxxxx/svn/relax/branches/bieri_gui
This will be used solely to test if this branch can be merged.
........
r11011 | bugman | 2010-03-16 14:20:25 +0100 (Tue, 16 Mar 2010) | 6 lines
Reverted r11004 as this, for some unknown reason, undid the changes of
r11001.
The command used was:
svn merge -r11004:11003 .
........
r11012 | bugman | 2010-03-16 14:26:19 +0100 (Tue, 16 Mar 2010) | 6 lines
Bug fix for the minimisation user functions.
This was introduced accidentally in r11007. The reference self._exec_info
was accidentally set to
self.__exec_info.
........
r11013 | bugman | 2010-03-16 15:07:25 +0100 (Tue, 16 Mar 2010) | 6 lines
Reverted the temporary r11010 for testing the merging of the bieri_gui
branch.
The command used was:
svn merge -r11010:r11009 .
........
r11014 | bugman | 2010-03-16 18:05:47 +0100 (Tue, 16 Mar 2010) | 30 lines
Reverted r9348 as this breaks on Mac OS X newline characters!!!! Retarded
python 3.0!
The command used was:
[edau@localhost relax-1.3]$ svn merge -r9348:9347 .
--- Reverse-merging r9348 into '.':
U prompt/interpreter.py
.....
r9348 | bugman | 2009-08-21 10:50:08 +0200 (Fri, 21 Aug 2009) | 18 lines
Changed paths:
M /1.3/prompt/interpreter.py
The execfile statement has been replaced with a call to the
exec(compile()) function.
This was done with the python 2.x to 3.x script with the commands:
2to3 -w -f exec relax .
The change is in preparation for Python 3.0, and relax still works with
Python versions 2.4, 2.5,
and 2.6. The syntax has changed from:
execfile(script_file, local)
to:
exec(compile(open(script_file).read(), script_file, 'exec'), local)
Hmmm, not much of a language simplification!
.....
........
r11015 | bugman | 2010-03-16 18:18:35 +0100 (Tue, 16 Mar 2010) | 5 lines
Fix for the ellipsoid() function for the new param_types == 3 (Dxx, Dyy,
Dzz, Dxy, Dxz, Dyz) data.
The ordering Dx <= Dy <= Dz was not being used so the rhombicity was
incorrectly calculated.
........
r11016 | bugman | 2010-03-16 18:42:07 +0100 (Tue, 16 Mar 2010) | 5 lines
Another bug fix for the ellipsoid() function for the (Dxx, Dyy, Dzz, Dxy,
Dxz, Dyz) parameter set.
The Euler angles which were in radians were being converted from deg to
rad!!!
........
r11017 | bugman | 2010-03-17 01:31:41 +0100 (Wed, 17 Mar 2010) | 10 lines
Creation of the NOE auto-analysis module.
This is a modified copy of the sample script created by Michael Bieri.
The key references are:
Start: https://mail.gna.org/public/relax-devel/2010-03/msg00051.html,
The file:
https://mail.gna.org/public/relax-devel/2010-03/msg00052.html,
Details: https://mail.gna.org/public/relax-devel/2010-03/msg00054.html.
........
r11018 | bugman | 2010-03-17 15:01:49 +0100 (Wed, 17 Mar 2010) | 7 lines
Bug fix for the local_tm_models status value.
This was identified by Daniel Pérez <daniel dott perez att mol dott biol
dott ethz dott ch> in the
post at https://mail.gna.org/public/relax-users/2010-03/msg00004.html
(Message-id:
<3C30C4A9-75D1-4986-9CE5-0ECA364B723B@xxxxxxxxxxxxxxxx>).
........
r11019 | bugman | 2010-03-17 18:41:46 +0100 (Wed, 17 Mar 2010) | 3 lines
Added 4 system tests for the setting ellipsoidal diffusion tensor (the 4
different param sets).
........
r11020 | bugman | 2010-03-18 10:40:24 +0100 (Thu, 18 Mar 2010) | 3 lines
Added the noe module to the auto_analyses.__all__ list.
........
r11021 | bugman | 2010-03-18 15:09:00 +0100 (Thu, 18 Mar 2010) | 10 lines
Final bug fix for the ellipsoid() function for the (Dxx, Dyy, Dzz, Dxy,
Dxz, Dyz) parameter set.
The 2304 unique Euler angle values for a rank-2 tensor with no skew (12
axis orders, 2 frames of
ref., 2 directions (forwards + reverse), 4 rotational symmetries (pi
rotations about each
eigenvector), 2 handednesses (left + right), and 6 eigenvalue orders) are
now all properly handled.
The resultant Euler angle values are a single set of the 2304, matching
that already used in relax
[... 1184 lines stripped ...]
r11242 - in /branches/bmrb: ./ auto_analyses/ docs/ docs/latex/ generic_fns/ generic_fns/structure/ maths_fns/ prompt/ sampl...