GitLab

My test with a periodic system now works, but the interface to adding box
vectors should be improved.

Traceback (most recent call last):
  File "/path/to/python/tests/TestCharmmFiles.py", line 91, in test_HydrogenMass
    topology = self.psf_v.topology
  File "/path/to/lib/python2.7/site-packages/simtk/openmm/app/charmmpsffile.py", line 782, in topology
    topology.addAtom(atom.name, elem, residue)
UnboundLocalError: local variable 'elem' referenced before assignment

simulateCharmm.py to use the latest force fields. This is what CHARMM does for
their own tutorials, so this is fine.

to prevent users from actually using them.

ERROR: Test that altering the mass of hydrogens works correctly.
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/travis/build/SimTk/openmm/python/tests/TestCharmmFiles.py", line 91, in test_HydrogenMass
    topology = self.psf_v.topology
  File "/usr/local/lib/python2.7/dist-packages/simtk/openmm/app/charmmpsffile.py", line 777, in topology
    masselem = Element.getByMass(mass)
NameError: global name 'Element' is not defined

pickleability.

- Add a docstring to element.Element.getByMass describing what it does
- Add a new element virtualsite with atomic number 0 (and alias it to extrapoint
  and lonepair)

Include force groups when serializing forces

Cleanup to the recent serialization code changes

constructor _should_ be able to take all of files in the CHARMM parameter set
distribution and correctly determine file type from just the name and/or
filename extension.

Added missing serialization support for CustomNonbondedForce group interactions

Added defaults to make serialization of CustomNonbondedForce and NonbondedForce backward-compatible.

parameters to a ProteinStructure before calling createSystem.

Updated tests of serialization of CustomNonbondedForce, though test omits comparison of interaction group sets since I couldn't figure out how to do that.