bandu 1.3.4

The BandU program constructs a rank ordered series of crystal orbitals using principal component analysis. These principal orbital components can then be projected on the Fermi surface and visualized

BandU

---

BandU OVERVIEW

A package that performs a principal component inspired analysis on the Bloch wavefunctions of periodic material to provide a real space visualization of the states that significantly contribute to the Fermi surface.

These real space functions can then be projected onto the Fermi surface to provide a clear visual for where a nesting vector may combine two points in reciprocal space.

This package is designed to be very straightforward in its use, offering Fermi surface and BandU function visualizations in as little as 5 lines of Python script. This package can also be used to just visual the Fermi surface, without BandU projections, if provided with the necessary k-point and eigenvalue data.

---

INSTALLATION INSTRUCTIONS

THROUGH GITHUB

1. Inside that directory type on the command line
   "git clone https://github.com/pcross0405/BandU.git"

2. Type "cd BandU"

3. Make sure you have python's build tool up to date with
   "python3 -m pip install --upgrade build"

4. Once up to date type
   "python3 -m build"

5. This should create a "dist" directory with a .whl file inside

6. On the command line type
   "pip install dist/*.whl"

THROUGH PIP

pip install bandu

---

DEPENDENCIES

REQUIRED FOR VISUALIZING FERMI SURFACE

• pyvista

• numpy

REQUIRED FOR CUSTOM COLORS

• matplotlib

WAVEFUNCTIONS THAT CAN BE READ DIRECTLY

Currently only reading directly from ABINIT 7 and 10 wavefunctions is supported. Reading eigenvalues from other DFT packages will come in future updates.

• ABINIT

---

REPORTING ISSUES

Please report any issues here

---

TUTORIAL

An example script that can run the different functions of the BandU program is given below.

from bandu.bandu import BandU
from bandu.abinit_reader import AbinitWFK
from bandu.isosurface_class import Isosurface
from bandu.plotter import Plotter
from bandu.colors import Colors

root_name = 'your file root name here' # root_name of WFK files and of XSF files
xsf_number = 1 # XSF file number to be read in
energy_level = 0.000 # Energy relative to the Fermi energy to be sampled
width = 0.0005 # Search half the the width above and below the specified energy level
wfk_path = f'path\to\WFK\file\{root_name}_o_WFK'
xsf_path = f'path\to\XSF\file\{root_name}_bandu_{xsf_number}'
bandu_name = f'{root_name}_bandu'

def main(
        principal_orbital_components:bool, 
        fermi_surface:bool, 
        fermi_surface_projection:bool,
        load_fermi_surface:bool
)->None:
    # this option will generate the principal orbital components 1 through 10
    # to generate more or less, adjust the range of the "nums" keyword in the ToXSF() function
    # the energy sampled can be set, relative to the Fermi energy, by changing the the "energy_level" global variable
    # states are included in the analysis if they are within +/- 1/2*width of the set energy_level 
    # to get fewer or more states, decrease or increase, respectively, the "width" global variable
    # by default, the prinicipal orbital components are generated from an irreducible wedge of the Brillouin Zone
    # to generate from the full BZ, change the "sym" attribute in the BandU class from "False" to "True"
    if principal_orbital_components:
        wfk_gen = AbinitWFK(wfk_path).ReadWFK(
            energy_level = energy_level,
            width=width
         )
        wfk = BandU(
            wfks = wfk_gen,
            energy_level = energy_level,
            width = width,
            sym = False
         )
        wfk.ToXSF(
            xsf_name = bandu_name,
            nums = [1,10]
         )
    # this option will only generate an energy isosurface and will not project principal component overlap onto the surface
    # the "energy_level" global variable is the energy at which the isosurface will be be generated, relative to the Fermi energy
    # so energy_level = 0.0 will generate the Fermi surface
    # the "width" global variable determines how many states are included in the generation of the isosurface
    # a small width (~10 meV or ~0.5 mHa) is best here as larger widths may introduce bands that do not cross the Fermi energy
    # the color of the surface can be changed to any string compatible with the matplotlib colors 
    # see named colors here: https://matplotlib.org/stable/gallery/color/named_colors.html
    # the Plot function has many other keywords to customize the visuals to the users liking, see the docstring for more
    elif fermi_surface: 
        contours = Isosurface(
            wfk_name = wfk_path,
            energy_level = energy_level,
            width = width
        )
        contours.Contour() # make contours
        plot = Plotter(
            isosurface = contours,
            save_file=f'{root_name}_bandu_{xsf_number}_fermi_surf.pkl'
         ) # create plotter object
        plot.Plot(
            color = 'silver',
        ) # plot contours
    # this option will generate an energy isosurface as well as project the overlap of a principal orbital component onto the surface
    # everything remains the same as the previous option, except now the principal orbtial component XSF file is needed 
    # also the color of the surface is done with the Colors module by default
    # other colors can be made with the Colors module, also any matplotlib colormap works
    elif fermi_surface_projection: 
        contours = Isosurface(
            wfk_name = wfk_path,
            energy_level = energy_level,
            width = width
        )
        contours.Contour() # make contours
        plot = Plotter(
            isosurface = contours,
            save_file=f'{root_name}_bandu_{xsf_number}_fermi_surf.pkl'
         ) # create plotter object
        overlap_vals = plot.SurfaceColor(
            wfk_path=wfk_path,
            xsf_path=xsf_path,
        ) # compute overlap between principal orbital component and states in Brillouin Zone
        plot.Plot(
            surface_vals = overlap_vals,
            colormap = Colors().blues,
        ) # plot contours
    # this option will load a previously generated and saved fermi surface file
    # update the "save_path" keyword to match the path and name of your save file
    elif load_fermi_surface:
        Plotter().Load(
            save_path='{root_name}_bandu_{xsf_number}_fermi_surf.pkl',
        )
# to run any of the options above, make sure to set that option to "True"
# also be sure that the other options (or at least all options that come before) are set to "False"
# the main function will only run which ever option is the first found to be "True" in top to bottom order
# in other words, the priority follows as most to least in the order:
# principal_orbital_components -> fermi_surface -> fermi_surface_projection -> load_fermi_surface
if __name__ == '__main__':
    main(
        principal_orbital_components=True, 
        fermi_surface=True, 
        fermi_surface_projection=True,
        load_fermi_surface=True
    )