VASP + wannier90 always generates parameters automatically

[liang-wen_pi]

Hello everyone, ！
Recently when I was using VASP + wannier90 to calculate graphene, I always encountered a problem where VASP automatically set num_wann = NBANDS, instead of using my own settings. The specific situation is as follows:
part of INCAR:

LWANNIER90 = true
WANNIER90_WIN = "
NUM_WANN=2
Begin Projections
C : pz
End Projections

# Disentanglement
dis_win_min = -15.0 ! 能量窗口下限 (eV)
dis_win_max = 10.0 ! 能量窗口上限 (eV)
dis_froz_min = -4.0 ! 冷冻窗口下限
dis_froz_max = 0.0 ! 冷冻窗口上限 (包含了费米面附

guiding_centres = true
"

This is the content of the generated "wannier90.win" file.

Begin Projections
C : pz
End Projections

# Disentanglement
dis_win_min = -15.0 ! 能量窗口下限 (eV)
dis_win_max = 10.0 ! 能量窗口上限 (eV)
dis_froz_min = -4.0 ! 冷冻窗口下限
dis_froz_max = 0.0 ! 冷冻窗口上限 (包含了费米面附

guiding_centres = true

# This part was generated automatically by VASP
num_bands = 32
num_wann = 8
begin unit_cell_cart
2.4680118 0.0000000 0.0000000
-1.2340056 2.1373611 0.0000000
0.0000000 0.0000000 15.0000000
end unit_cell_cart
begin atoms_cart
C 1.2340060 0.7124537 7.5000000
C 0.0000002 1.4249074 7.5000000
end atoms_cart
mp_grid = 18 18 1

The erro : param_get_projections: too few projection functions defined
If someone can answer this question, I would be extremely grateful.

[christopher_sheldon1]

Dear Liang-Wen Pi,

Thank you for your question. This is because NUM_WANN is set inside WANNIER90_WIN. If you set it outside, e.g:

Code: Select all

NUM_WANN=2
WANNIER90_WIN = "
Begin Projections
C : pz
...

Then the Wannier90 calculation runs smoothly. You can also run Wannier90 within VASP using WANNIER90_WIN; this worked for me.

Does this solve your problem?

Best wishes,

Chris