topologyobjects.Angle does not know about the periodic boundaries nor is this behaviour documented

[matsikora]

Expected behaviour

topologyobjects.Angle.angle (and probably dihedral as well) should have a "box" option to allow for correct calculation of angles when contributing atoms cross periodic boundaries. Analogous to lib.distances.calc_angles().

Actual behaviour

angle function is not aware of PBC and this behaviour is not documented.

Code to reproduce the behaviour

Currently version of MDAnalysis:

0.17.0

[orbeckst]

@richardjgowers I assigned to you as you're working on PR #1939 already