Qhull algorithm used to create the VoronoiNN class does not take care of the periodic boundary conditions.

[Anny-tech]

trafficstars

Hi

I had been trying to use the VoronoiNN class to find nearest neighbors w.r.t a particular site within the structures I am studying. However, I found from literature, that the quickhull algorithm used to create VoronoiNN does not take care of periodic boundary conditions. In that case, how can I be sure whether the nearest neighbors analysis I am doing using the VoronoiNN class is good enough?

Does the 'allow_pathological' parameter help?

[CompRhys]

You can handle periodic systems by taking a supercell, applying qhull to the super cell and then folding the images back into the unit cell. This is the approach used in pymatgen.