[Pw_forum] wrong number of symmetries!

Paolo Giannozzi p.giannozzi at gmail.com
Sat Oct 15 11:13:51 CEST 2016


In order to find symmetries, the code uses a threshold to decide if a
rotated vector is the same as the original one. It may sometimes happen
that some "quasi-symmetries" are recognized as symmetries, some are not,
leading to a "symmetry group" that is not a group. This is likely what
happened in your case.  Paolo

Il 15/ott/2016 01:27 AM, "Li Mingkai" <mingkaili at gmail.com> ha scritto:

> I tried to do the scf calculation on a perturbated cell. I used
> "CELL_PARAMETERS alat" to specify vector. However, I got an error.
>
>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
> %%%%%%%%%%%%%%%%%%%
>      Error in routine set_sym_bl (3):
>      wrong number of symmetries! Use standard orientations for axis
>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
> %%%%%%%%%%%%%%%%%%%
>
> Since the cell is perturbated, there is no symmetry on it and space group
> should be P1. But it seemed that the pw.x tried to find some symmetry and
> got wrong number of symmetries. I tried nosym, nosym_vec, no_inv. Nothing
> works. I tried old version down to 5.0, which worked. After version 5.1,
> the error emergied. Of course, I tried to use celldm(1)~(6). It works well
> for all version. Since this input file is generated by a script
> automatically. I don't tend to modify the script a lot and try to figure
> out what happened before change everything.
>
> --
> Mingkai Li
> =======================================================
> School of Material Science and Engineering
> Hubei University
> Wuhan, Hubei Province
> P. R. China
> =======================================================
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20161015/8f98851d/attachment.html>


More information about the users mailing list