Issue with zero-radius atoms

[Marcello-Sega]

In case one or more atoms are buried within the excluded volume of another one the GITIM algorithm does not behave as one would expect. The typical case is hydrogen atoms in water models like SPCx, TIPxP. This happens because:

1. the resulting triangulation includes simplexes with acute angles
2. the kissing spheres algorithm might provide spurios soulutions.

Usually can be dealt with by excluding hydrogen atoms from the analysis group. As this can be not immediately obvious to new users, one possibility is to introduce the default behaviour of not including hydrogens or atoms with zero radius. As the atom types are not always reliable, probably the best option is to exclude atoms with zero radius.