wanpy 0.16.1

A Wannier Tight-Binding Model Based Numerical Package

WanPy: A Wannier Tight-Binding Model Based Numerical Package

Requirements

• python>=3.11
• numpy
• scipy
• sympy
• pandas
• mpi4py
• h5py
• fortio
• hatch
• spglib>=2.3.0
• matplotlib (optional)
• phonopy (optional; required by findmsg)
• pymatgen (optional; required by wpyplotband)

Installation

Create a Conda environment and install wanpy:

conda create -n wanpy numpy scipy pandas sympy h5py mpi4py fortio hatch spglib matplotlib
conda activate wanpy 
pip install wanpy 

Note for compatibility with srun, it is recommended to use MPICH:

conda install -c conda-forge mpi4py==3.1.4 mpich

create a symbolic link for wanpyProjects at ~/opt/wanpyProjects:

wanpy-postinstall

(Optional) Configure Environment Variables If you are installing wanpy on a personal computer, you may want to set environment variables for convenience:

export WANPY_ROOT_DIR="your_work_dir"  # Set a default working directory for WanPy  
export PYGUI="True"  # Enable GUI features if applicable  

Get start

# show usage
wanpy -h

# collect Wannier TB data to a single .h5 file
wanpy htb

# Collect Wannier TB data to a single .h5 file
wanpy wannier

# Write _hr.dat and _r.dat (optional) from .h5 file
wanpy write_dat

# Twist the order of .amn (uudd or udud) 
wanpy twist_amn

# compare Wannier and VASP band structure. support HSE type calculation. 
wpyplotband

# check job status
statistic_cores ...

Collect Wannier TB data

WanPy need to collect necessary Wannier TB data into a single .h5 file as the input for further calculations. This can be done by:

Method 1 (recommended): from wanpy wannier, this will need POSCAR .nnkp, .wout, .chk, .eig, and WAVECAR if --spn. The nnkp file can be obtained by:

$ wannier90.x -pp wannier90.win 

To start collecting dat:

# start to collect Wannier TB data 
wanpy wannier [options]

# For detail of options, see
wanpy wannier -h

Method 2: from wanpy htb, this will need at least POSCAR, .wout, _hr.dat . The files .nnkp, _r.dat, _wsvec.dat, and _spin.dat are optional.

# start to collect Wannier TB data 
wanpy htb [options]

# For detail of options, see
wanpy htb -h

Symmetric Wannier tight-binding models (SWTB)

The SWTB can be obtained during collecting Wannier TB data. Set symmetry related parameters in symmetry.in, then start to get SWTB with an additional tag --symmetry

wanpy wannier --symmetry [options]

See Comput. Phys. Commun. 270, 108153 (2022) (arXiv:2012.08871) for detail.

Important notice:

• WanPy uses uudd order of Wannier orbitals for its internal calculations, which is the default setting when using v1.2 of wannier_setup. If a higher version vasp2wannier interface was used, one should utilize wanpy twist_amn to reorganize the .amn file into the uudd order and then redo the disentanglement.

• wannier_center_def must be specified to excute wanpy wannier. The available options are poscar and ws. By using wannier_center_def = poscar, the wannier centers will align with POSCAR. For wannier_center_def = ws, the wannier centers will be refined in range of [-0.5, 0.5). For example, if .amn file is generated by VASP 5.4.4, set wannier_center_def = poscar. If .amn file is generated by VASP 6.4.3, set wannier_center_def = ws.

Template input file for generating htb.h5:

This template file can also be produced by executing wanpy wannier --temp:

# Input file for building symmetric Wannier TB model

# Choose method, default is kspace
symmetric_method = rspace   # Options: {rspace, kspace}
rspace_use_ngridR = F       # use ngridR to resample the TB model, default is False

# Choose between manually setting symmops or automatically detecting symmops from magmoms
parse_symmetry = man        # Options: {man, auto}

# Choose how wannier centers are defined in calculating .amn
# Options: {poscar, ws}
#   poscar: the same as poscar. Use this if amn is generated by VASP 5.4.4 (see mlwf.F for details). 
#   ws: refined in range of [-0.5, 0.5). Use this if amn is generated by VASP 6.4.3 (see mlwf.F for details). 
wannier_center_def = poscar 

ngridR = 12 12 1            # in kspace method, use a slightly larger value than the original TB model 

# Parameters used when parse_symmetry = auto
symprec = 1e-5
&magmoms
0 0 1
0 0 -1
/

# Parameters used when parse_symmetry = man
&symmops
# TR  det  alpha  nx  ny  nz  taux  tauy  tauz
  0   1    0      0   0   1   0     0     0   # e
  0   1    180    0   0   1   0.5   0.5   0   # c2z
# Anti-unitary symmetry operations
  1   1    0      0   0   1   0.5   0.5   0   # T
  1   1    180    0   0   1   0     0     0   # Tc2z
/