Interesting. The previous symmetry-searching algorithm (v.4.3.2 and earlier) used to distinguish between hexagonal and all other lattices. The current one (v.5.0 and later) is basically the same but makes the same checks for all lattices. Unfortunately it is confused by your rather peculiar setting for the hexagonal lattice: one axis along the z axis (and this is ok), one along (110), one forming a 15 degrees angle with the x axis. The previous algorithm finds only 12 symmetries (instead of 24). The current one performs better, since it finds 16, but unfortunately they do not form a group, unlike the old 12. This will be fixed sooner or later, but for the time being the only solution is to rotate your system (nosym=.true. will not do the job: it disables symmetries after the lattice symmetries have been found., i.e. too late).
P. > Yes, the code stops with the error message... > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > Error in routine multable (2): > > Not a group > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > -- Paolo Giannozzi, IOM-Democritos and University of Udine, Italy
