[Wannier] Spread in Wannier90

[Payal Wadhwa]

Dear Wannier experts,

I am new in wannier tools and facing a problem in choosing frozen windows.
YSb is a semi-metal and its Fermi energy is 5.2043, with 8 filled bands. To
incorporate the contribution from valence as well as conduction bands i
have chosen

num_wann =    18  ! set to NBANDS by VASP
num_bands = 36
exclude_bands = 19-36
num_iter=5000
dis_num_iter=30000
#dis_win_min = -18.0
#dis_win_max = 40.0
dis_froz_min = -6.0000
dis_froz_max =  6.000

begin projections
Y : s; px; py; pz; dxy; dyz; dxz; dx2-y2; dz2
Sb : s; px; py; pz; dxy; dyz; dxz; dx2-y2; dz2
end projections

With this, i am getting spread as
*  5000    -0.135E-05     0.0012524516       22.2875048331    6442.97  <--
CONV*
*        O_D=      0.0020980 O_OD=      5.4577301 O_TOT=     22.2875048 <--
SPRD*
* Delta: O_D= -0.4243946E-07 O_OD= -0.1308148E-05 O_TOT= -0.1350587E-05 <--
DLTA*

Then after adding dis_win_min = -6.0, dis_win_max = 35.0
i got
   5000    -0.924E-12     0.0000092488       46.4034625264    3777.49  <--
CONV
        O_D=      0.0000000 O_OD=     17.3872465 O_TOT=     46.4034625 <--
SPRD
 Delta: O_D=  0.1967317E-21 O_OD= -0.9308110E-12 O_TOT= -0.9237056E-12 <--
DLTA

Here my convergence limit reaches to 10^-12 but spread increases.
I have also tried with skipping the inner atomic states by decreasing the
num_wann.
Can somebody please suggest to me what window should I choose to have
minimum spread and converged MLWF.

I would be very thankful to you.

Thanks
With Regards,
Payal Wadhwa
Research Scholar
Department of Physics
Indian Institute of Technology,Ropar.

"Everybody is a genius. But if you judge a fish by its ability to climb a
tree, it will live its whole life believing that it is stupid"
--Albert Einstein
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/wannier/attachments/20190219/5f923c62/attachment.html>