Error in routine divide_class (1): > Wrong classes for C_3v about qe-doc HOT 3 CLOSED

You are probably right. There are two classes of hexagonal lattices. Probably there is an error in handling of the 4H type. I will try to take a look at it next week.
Regards,
P. Jochym

from qe-doc.

Dear Dr.Jochym

I was able to solve this issue by increasing the significant digits of the input data (pw.in), so that the Quantum espresso was able to resolve the conflict in choosing the symmetry.

Regards
Linu

from qe-doc.

Glad to hear it!
This is often a correct approach in case of close-to-perfect symmetries. The symmetry search problem is not solvable in general for practical cases. The solutions are approximations and best-likelihood estimations.

from qe-doc.