Hi,

By setting

LWANNIER90 = .TRUE.

LWRITE_MMN_AMN=.TRUE.

You will get the amn, mmn, eig and win files.

You can use them wherever you want afterwards.

Thanks Stephan. But I mean an alternative way to that one. The problem is that the version of VASP installed does not have wannier90, and we cannot upgrade it at this moment.

Greetings

Dear Diego Benalcázar Ruiz and Stepan S. Tsirkin

Which version of VASP support wannier90. Is there any special installation of VASP with wannier90 or just normal installation ?

Thank You

Shilendra

Hello Shilendra. To be honest I cannot answer you with complete certainty since I am not in charge of VASP installation

Diego, altertatively you may write a code to get all needed files from the WAVECAR file. In principle, all needed data is there. However the coding will take some work. 

Shilendra, I work with VASP 5.4, which supports wannier, not sure about earlier versions, you may try. However, VASP should be compiled with wannier90 as a library. I do not know details either, installation is done by our computing centre administrators.

Interesting. Thanks Stepan

i am just want to get fermi surface of my system and in my system, there is a partial filled "d" state so in principle i should use LDA+U method in INCAR and according to tutorial given in this link https://github.com/JMSkelton/VASP-Examples/tree/master/Wannier90_Si i have to use these tags in 2nd step. LHFCALC = .TRUE. HFSCREEN = 0.2 but two days have passed and still my calculations are running. i am using 32 processors and my system contains only 3 atoms.my INCAR is as follow ALGO = All EDIFF = 1.0e-8 ENCUT = 500 ISIF = 2 ISMEAR = 0 SIGMA=0.01 LORBIT=11 NSW=0 ISTART=0 ICHARGE=2 KPAR = 2 LCHARG = .FALSE. LREAL = .FALSE. NBANDS = 50 NEDOS = 1500 NPAR = 4 IBRION = 2 PREC = Accurate LDAU = .TRUE. LDAUTYPE = 2 LDAUL = 3 -1 LDAUU = 7 0 LDAUJ = 0.0 0.0 ISPIN = 2 MAGMOM = 1*10 2*0.6 LHFCALC = .TRUE. HFSCREEN = 0.2 can anyone please tell me that am i doing something wrong !or there is any quick or short method to get fermi surface?

Ali, do you really want LDA+U or HSE?

I guess for LDA you may check https://cms.mpi.univie.ac.at/vasp/vasp/On_site_Coulomb_interaction_L_S_DA_U.html

If you really want to do HSE calculations, they are much more time-consuming, than GGA/LDA. So try to reduce the number of k-points and other parameters, (prec=low, for example, ENCUT, EDIFF = 1.0e-6)

Hello again.

We finally recompiled vasp with wannier but there are a couple of questions I hope you can help me with.

I am using this INCAR:

ISMEAR = 0

SIGMA = 0.01

GGA = PE

LWANNIER90=.TRUE.

LWRITE_UNK=.TRUE.

LWRITE_MMN_AMN=.TRUE.

The .win, .mmn and .eig are created in the first run. However the .amn is not generated. Moreover, a .wout is

also generated though LWANNIER90_RUN is not set true. If I run it again, the .amn is still not created unless I change

the first lines of the previously created .win from

num_wann = 32 ! set to NBANDS by VASP

to:

num_wann = 8

num_bands=8

Begin Projections

Si:sp3

End Projections

I do not understand why VASP does not generate the .amn directly and if there is any way to correct this. I also do not get

why the .wout is created.

Thank you for your help.

VASP does not create projectuins, because it does not know, which projections you want. If you specify the projections in the win file, VASP will write the .amn from the first run. It will also create a .win file with structure parameters, k-points etc.

Thus, the minimal .win file that you need to provide should contain num_wann and projections, already from the first run.

I see.

Thank you very much Stephan

Hello,

I am also doing vasp2wannier90 calculation of Fe. But I am getting wannier90.win file with only one comment num_wann = 11 . How will i overcome this problem? What should i add extra in INCAR?

Hello,

I have compiled VASP2WANNIER90 as per the procedure successfully with VASP.5.3. and Wannier90-v.1.2. With Setting additional input tags in INCAR file of VASP as follows,

LWANNIER90=.TRUE.

LWRITE_MMN_AMN=.TRUE.

LWRITE_UNK=.TRUE.

I am getting .win, .up.mmn, .dn.mmn, up.eig, dn.eig, .nnkp & .wout files but not seedname.amn file. What necessary changes should i do in wannier90 input file to get it?

How should i get hamiltonian matrix (hr.dat file) from wannier90? I have tried with setting write_hr=T & hamiltonian_write_hr in input file.

Looking forward for any necessary suggestions in this regard. Please point out what parameter i am missing.

With Regards.