Proteuspy

Added

- Stronger type checking in general
- Additional error checking for slicing lists
- Several new Unittests

Changed

- Refactored both ``Disufulfide`` and ``DisulfideList`` and moved plotting and statistics into new classes ``DisulfideVisualization`` and ``DisulfideStats``.
- Simplification and unification of accessing Disulfides via class strings
- Generalized Disulfide class creation, can create up to 26-fold classes now.

Fixed

- DisulfideList.AverageConformation now uses a circular mean function to correctly handle the nature of -180 - 180 degree dihedral angle averaging.

Added

- Disulfide.spin() - Spin the disulfide about the Y axis.
- Test_DisplaySS.py - Unittest improvements.

Changed

- Completely rewrote ``DisulfideClass_Constructor`` methods that generate the binary and octant class strings for simplicity.
- The new ``DisulfideClass_Constructor.get_segment()`` is a more general version of ``DisulfideClass_Constructor.get_eighth_quadrant()``.

Added

- static method ``DisufideClass_Constructor.class_string_from_dihedral()`` - return the binary or octant class string for a single or array of dihedrals.
- properties ``Disulfide.binary_class_string`` and ``Disulfide.octant_class_string``
- ``DisulfideLoader.sslist_from_class()`` - now uses the index value for a specific disulfide that matches the classID from the TorsionDF dataframe. This allows for direct access to the DisulfideList containing the Disulfides via ``loader[index]``, which is significantly faster than looking up via ``loader[disulfide_name]``.

Changed

- vectorized binary and octant class string construction
- added binary and octant class strings to the master ``loader.TorsionDF`` DataFrame.
- Rewrote ``programs/DisulfideClass_Extractor.py`` to use the index-based address scheme described above. This resulted in about a 20x speedup in the program!

Added

- ``DisulfideList.plot_distances()``
- ``DisulfideList.plot_deviation_histograms()``
- Added the ability to access disulfide class strings directly through ``DisulfideLoader`` with indexing.

Changed

- ``Load_PDB_SS(verbose=True)`` now calls ``loader.describe()`` to print statistics for the database
- Optimized ``DisulfideList.create_deviation_dataframe()``.
- Moved various class plotting routines into ``DisulfideLoader`` class.
- Cleaned up ``DisulfideClasses.py``

Added

- ``DisulfideExtractor_mp.py`` moved into the package as a callable module.

``proteusPy.DisulfideExtractor`` from command line will launch the program

- Incorporated consensus structures into the ``DisulfideClass_Constructor`` object. This presumes these have been generated. The consensus structures are created through the program ``DisulfideClass_Analysis.py``.

Changed

- Corrected an error in ``DisulfideLoader`` that failed to initialize the torsion dataframe properly after filtering.
- Change to setup.py - 2q7q_seqsim.csv was not being included

Fixed

- Sg_distance was not being calculated with ``Disulfide.build_yourself()``
- phi and psi were not correctly populating in the torsion dataframe.
- There was a subtle error in the ``DisulfideLoader`` initialization that led to internal database inconsistencies after filtering. This has been corrected.

Added

- ``qt5viewer.py`` moved into the package as a callable module.

``proteusPy.qt5viewer`` from command line will launch the program

Changed

- logging cleanup in ``DisulfideLoader.py`` and ``DisulfideList.py``.
- ongoing documentation tweaks, cleanup