[Wannier] kmesh_get: something wrong, found too many nearest neighbours

[Jonathan Yates]

On 9 Jan 2017, at 16:43, Jia Chen <jiachenchem at gmail.com> wrote:

> Dear All,
> 
> I am having the title error from wannier90. And I tried to set kmesh_tol=0.00001 or devel_flag=kmesh_degen, the problem still stays. I hope I can get some help here. Thank you.

Set 
 kmesh_tol=0.0000001
in the win file - and then it works ok. (this is an order of magnitude tighter than the default tolerance).


Why does that work?
  Your structure is just very slightly away from being symmetric - if we look at your unit cell in terms of cell lengths and cell angles this is more apparent:
        Real Lattice(A)                     
   6.5513444   0.0085398   0.0002025       
   0.0920464   6.5507774  -0.0080818   
   0.1010507   0.1085719   6.5496972      
       Lattice parameters(A)       Cell Angles
     a =    6.551350          alpha =   89.108776
     b =    6.551429          beta  =   89.113207
     c =    6.551376          gamma =   89.120294
This means that the shells of k-points are very closely spaced, and wannier90 needs a finer tolerance to distinguish them than usual. With the regular tolerance it was putting kpoints into the same shell that should belong to different shells.

Jonathan




-- 
Department of Materials, University of Oxford, Parks Road, Oxford, OX1 3PH, UK
tel: +44 (0)1865 612797                http://users.ox.ac.uk/~oums0549/