chengcheng-xiao / vasp2wan90_v2_fix

An updated version of the VASP2WANNIER90v2 interface

Hello there stranger!

My name is Chengcheng Xiao (肖承诚), a postdoctoral researcher @ Imperial College London.

• 🔭 I’m currently working on: 2D materials | various DFT methods.
• 🌱 I’m currently learning: QFT | Green's function method.
• 💬 Ask me about my reseasrch and my codes.
• 📃 Check out my blog.
• 📫 How to reach me:
  • Mail : [email protected]
  • Twitter : @iconxicon
  • WeChat : ICONXICON

vasp2wan90_v2_fix's Issues

.spn file?

The spn file contains:

• <\psi_m| S | \psi_n>, where S is the Pauli matrix, the pseudo part.
• <\psi_m| \sigma_{ij} q_{ij} |\beta_i> <\beta_j| |\psi_n>, the augmentation part.

The pseudo part can be easily obtained. For the augmentation part, refer to pw2wannier90.f90 (line 2469)
and VASP's PEAD_CALC_OVERLAP and CALC_OVERLAP_GN.

collinear spin calculations

Somehow, when doing collinear spin calculations, the code skips "number of bands" or "number of projections" error. If they occur, the whole calculation will just stop. Worth looking into.

write UNK.NC file?

This seems to be easier than the .spn file.

So, do this first.

WVFN(IND)=WCOLLECT(ISTRIP)%CR(NX+(NC-1)*WP%WDES%GRID%RL%NROW)

By the looks of it, we only need to halve the CR array and output it to two WVFN arrays and then print it out.

MPI issue?

Dear all,

I'd like to share my recent experience in case it may help others facing a similar issue in the future. I successfully resolved the problem by recompiling VASP with the patch available at https://github.com/Chengcheng-Xiao/VASP2WAN90_v2_fix. After recompiling, it's important to note that you should run VASP using only a SINGLE CPU core to generate mmn, amn, and spn. I encountered the same underestimation issue when attempting to generate mmn and amn using multiple CPUs. While I'm not an expert, I suspect this might be related to the use of openmpi or some other factor. Therefore, if you're struggling to obtain correct results, I recommend trying this patch and SINGLE CPU approach (note: reading converged WAVECAR and CHGCAR makes the SINGLE CPU approach affordable). Good luck!

Regards,
Hongyang

Originally posted by @ma455173220 in wannier-berri/wannier-berri#274 (comment)

Plotting MLWFs

Hi Xiao,
Hope you are good.
I want to plot and visualize the spread of maximally localized wannier functions (though I can see the latter in the output file).
If you don't mind sharing what workflow I should follow?
For example,
Step 1: SOC-SCF calculation
Step 2: Reading SOC-SCF with ICHARG=11, LWANNIER90=.TRUE. and generating wannier inputs
Step 3: ..???
I read that I need to set Wannier_plot=.true. but I am not sure at which step I should do this because this needs UNK files.
P.S> I am using vasp5.4.4 compiled with this patch.
Thanks in advance.

Does your patch works for VASP 6.12?

Dear Mr. Xiao

May I consult you that does your patch can be applied to VASP 6.12 as well?

My current VASP is version 6.12, and I am also running into projection problems when SOC is included. I would like to know your comments on this regarding using VASP version 6.12. Can you help? Tks!

with regards,
Peter

Direction of magnetization

Hi Chengcheng,

Thanks for the nice codes. I have a question regarding the direction of magnetization. In your example04 or the example18 from wannier90 tutorial, the magnetization is set to parallel to the minus z direction then the calculated AHC is positive. If we reverse the magnetization, the AHC will become negative. I want to study some other materials. Except for symmetry analysis, is there a better way to study the dependence of magnetization on AHC rather than reconstructing the wannierisation from scratch?

Thanks,
Wuzhang Fang

Regarding VASP2WANNIER Spn file

In the vasp2wannier calculation. I am using vasp6.3 with wannier90-developer code
I am trying to calculate the spin hall conductivity when I am running postw90.x it is showing an error in opening input file wannier90.spn.
But vasp6.3.0 is not generating wannier90.spn file
so I used vasp5.4 with wannier to generate the wannier90.spn file after that I ran vasp6.3 with wannier but it is showing
forrtl: severe (174): SIGSEGV, segmentation fault occurred
Please suggest me, how to resolve this problem.
Thank you in advance

Can this be used for VASP2WAN90_v3?

Hi,

The wannier90 has released v3 version, can this also be used for VASP2WAN90_v3?

Regards

spin component y in .spn output might be wrong

in the .patch file line 1446 we have:

•                CALL ECCP_NL_SPN(LMDIM,LMMAXC,CMPLX(0._q,1._q,q),CQIJ(1,1,NI,1+1+2*1),W2%CPROJ(NPRO_+1),W1%CPROJ(NPRO+1),CNLy)
  

Should that be:

•                CALL ECCP_NL_SPN(LMDIM,LMMAXC,CMPLX(0._q,1._q,q),CQIJ(1,1,NI,1+1+2*0),W2%CPROJ(NPRO_+1),W1%CPROJ(NPRO+1),CNLy)
  

Just to be consistent with the Pauli matricies.

Add spin quantisation axis support

The fix now reads in proj_s_qaxis. But all spinor projections are done without rotation of quantisation axis.

If Im not mistaken, the rotated spinor eigenvectors can be dotted directly onto the A_mn when the PAW pseudopotential is used (According to PW2WANNIER90.F90 for pwscf v6.4):

! general routine
! for quantisation axis (a,b,c) 
! 'up'    eigenvector is 1/sqrt(1+c) [c+1,a+ib]
! 'down'  eigenvector is 1/sqrt(1-c) [c-1,a+ib]
if(spin_eig(iw)==1) then
   fac(1)=(1.0_dp/sqrt(1+spin_qaxis(3,iw)))*(spin_qaxis(3,iw)+1)*cmplx(1.0d0,0.0d0,dp)
   fac(2)=(1.0_dp/sqrt(1+spin_qaxis(3,iw)))*cmplx(spin_qaxis(1,iw),spin_qaxis(2,iw),dp)
else
   fac(1)=(1.0_dp/sqrt(1-spin_qaxis(3,iw)))*(spin_qaxis(3,iw)-1)*cmplx(1.0d0,0.0d0,dp)
   fac(2)=(1.0_dp/sqrt(1-spin_qaxis(3,iw)))*cmplx(spin_qaxis(1,iw),spin_qaxis(2,iw),dp)
endif
ibnd1 = 0
DO ibnd = 1,nbnd
   IF (excluded_band(ibnd)) CYCLE
   amn=(0.0_dp,0.0_dp)
   DO ipol=1,npol
      istart = (ipol-1)*npwx + 1
      amn_tmp=(0.0_dp,0.0_dp)
      if (any_uspp) then
        amn_tmp = zdotc(npw,evc(istart,ibnd),1,sgf_spinor(istart,iw),1)
        CALL mp_sum(amn_tmp, intra_pool_comm)
        amn=amn+amn_tmp
      else
        amn_tmp = zdotc(npw,evc(istart,ibnd),1,sgf(1,iw),1)                        
        CALL mp_sum(amn_tmp, intra_pool_comm)
        amn=amn+fac(ipol)*amn_tmp
      endif
   enddo
   ibnd1=ibnd1+1
   IF (wan_mode=='standalone') THEN
      IF (ionode) WRITE(iun_amn,'(3i5,2f18.12)') ibnd1, iw, ik, amn
   ELSEIF (wan_mode=='library') THEN
         a_mat(ibnd1,iw+n_proj*(ipol-1),ik) = amn
      ELSE
         CALL errore('compute_amn',' value of wan_mode not recognised',1)
      ENDIF
   ENDDO
endif

The same could be done in VASP2WANNIER90 interface.

Add ability to choose spin channel? (collinear)

By default, we loop over all spin channle:

spin: DO ISP=1,WDES%ISPIN

This could be fixed by just adding a W90_spin tag to choose which channel we want.

Moreover, we should be able to use different projections for different spin channel ( using *.up.win and *.dn.win files ).

For this to work, we need to change seed_name.

Spinor projector not done right

With spinor calculation (LNONCOLLINEAR = .TRUE.), the spinor projection (i.e. C:pz(d) for Carbon's pz spin_2 channel orbital) is not done right.

Considering without spinor projection the projection yields correct result, I am thinking the ignored projections are, as a matter of fact, the spin_2 channel.

After checking with parameters.F90 in the WANNIER90 distro, this is exactly the case. A better fix would uses proj_s as the spin channel identifier, may even consider proj_s_qaxis for quantisation axis.

For now, I will only update the fix so that the spin channel in the projection scheme (using (d) or (u) for down and up channel) can work.

write spn files.

Thanks for your excellent code!
As you mentioned in the list of keywords, the spn file can be only written when VASP is exerted in serial. However, it is very time-consuming. Is there any solution that makes VASP is exerted in parallel (or only becomes serial when writing spn file, remaining parallel when doing other parts)?