[Pw_forum] Error from coset

[Paolo Giannozzi]

On Monday 21 November 2005 15:26, Sergey Lisenkov wrote:

>      from coset : error #         0
>      The order of the group is not a multiple of that of the subgroup
>
> What does it mean?

it means that the symmetry group of your system is not a group. 
The algorithm that finds the point group symmetries of the crystal
uses a threshold to decide whether the symmetry is there or not.
Apparently you found a set of coordinates and of lattice vectors 
for which seven candidate symmetry operations pass the test, the 
eight (inversion) fails.

Possible options:
- disable fractionary translations (set fractional_translations = .false.
  in PW/sgam_at.f90): the six operations in addition to identity have
  nonzero fractionary translations
- change the threshold (accep in PW/eqvect.f90)
- use option 'nosym' to disable symmetry (read carefully what it does)
- change a little bit atomic positions so as to break symmetry for good
- best of all: find a better algorithm to determine the symmetry ...

By the way: what should be the symmetry of this system?

Paolo

-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy