[Wannier] gamma_only= true - wannier90.mmn has not the right number of nearest neighbours

[Mohammad Sadegh Talezadehlari]

Dear All,


I am currently attempting to perform Wannierization for a relatively large supercell comprising 96 atoms. In my DFT calculations using VASP, my structure involves repeating MAPbI3 units (8 times). For the KPOINTS setting, I have employed the gamma point, and in the corresponding wannier input file, I have set "gamma_only = true". However, upon enabling the gamma_only option, I encountered the following error message:

"wannier90.mmn has not the right number of nearest neighbours".

To generate the initial wannier90 files, I incorporated the following commands into my INCAR file, which successfully produced the .amn, .mmn, .eig, and .win files:

LWANNIER90 = .TRUE.
NUM_WANN = 104
WANNIER90_WIN = "
gamma_only = true
Begin Projections
Pb:l=1
Pb:l=0
I:l=1
End Projections
"
My .mmn file started with:
File generated by VASP: unknown system
         252          64           6
    1    2    0    0    0
    0.909230482043    0.409856179950
   -0.000126326113   -0.000013000425
.
.
.

I attempted the LSCDM approach along with random projections; but, I encountered the same error.

Can you please help me with this issue?
Additionally, I would appreciate your insights on whether it is meaningful to carry out a DFT calculation with a single KPOINT (gamma point), while performing the subsequent Wannierization with kpoints set at 2 2 2 or 4 4 4, etc.


Thank you very much in advance for your support,

Kind regards,
Sadegh Talezadeh

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/wannier/attachments/20230802/304d6653/attachment.html>