G0W0 flow with convergence study wrt nband

This script shows how to compute the G0W0 corrections in silicon. More specifically, we build a flow to analyze the convergence of the QP corrections wrt to the number of bands in the self-energy. More complicated convergence studies can be implemented on the basis of this example.

import os
import sys
import abipy.data as data
import abipy.abilab as abilab

from abipy import flowtk


def make_inputs(ngkpt, paral_kgb=1):
    # Crystalline silicon
    # Calculation of the GW correction to the direct band gap in Gamma
    # Dataset 1: ground state calculation
    # Dataset 2: NSCF calculation
    # Dataset 3: calculation of the screening
    # Dataset 4-5-6: Self-Energy matrix elements (GW corrections) with different values of nband

    multi = abilab.MultiDataset(structure=data.cif_file("si.cif"),
                                pseudos=data.pseudos("14si.pspnc"), ndtset=6)

    # This grid is the most economical, but does not contain the Gamma point.
    scf_kmesh = dict(
        ngkpt=ngkpt,
        shiftk=[0.5, 0.5, 0.5,
                0.5, 0.0, 0.0,
                0.0, 0.5, 0.0,
                0.0, 0.0, 0.5]
    )

    # This grid contains the Gamma point, which is the point at which
    # we will compute the (direct) band gap.
    gw_kmesh = dict(
        ngkpt=ngkpt,
        shiftk=[0.0, 0.0, 0.0,
                0.0, 0.5, 0.5,
                0.5, 0.0, 0.5,
                0.5, 0.5, 0.0]
    )

    # Global variables. gw_kmesh is used in all datasets except DATASET 1.
    ecut = 6

    multi.set_vars(
        ecut=ecut,
        timopt=-1,
        istwfk="*1",
        paral_kgb=paral_kgb,
        gwpara=2,
    )
    multi.set_kmesh(**gw_kmesh)

    # Dataset 1 (GS run)
    multi[0].set_kmesh(**scf_kmesh)
    multi[0].set_vars(
        tolvrs=1e-6,
        nband=4,
    )

    # Dataset 2 (NSCF run)
    # Here we select the second dataset directly with the syntax multi[1]
    multi[1].set_vars(
        iscf=-2,
        tolwfr=1e-12,
        nband=35,
        nbdbuf=5,
    )

    # Dataset3: Calculation of the screening.
    multi[2].set_vars(
        optdriver=3,
        nband=25,
        ecutwfn=ecut,
        symchi=1,
        inclvkb=0,
        ecuteps=4.0,
        ppmfrq="16.7 eV",
    )

    # Dataset4: Calculation of the Self-Energy matrix elements (GW corrections)
    kptgw = [
         -2.50000000E-01, -2.50000000E-01,  0.00000000E+00,
         -2.50000000E-01,  2.50000000E-01,  0.00000000E+00,
          5.00000000E-01,  5.00000000E-01,  0.00000000E+00,
         -2.50000000E-01,  5.00000000E-01,  2.50000000E-01,
          5.00000000E-01,  0.00000000E+00,  0.00000000E+00,
          0.00000000E+00,  0.00000000E+00,  0.00000000E+00,
      ]

    bdgw = [1, 8]

    # Convergence study wrt nband in sigma.
    for idx, nband in enumerate([10, 20, 30]):
        multi[3+idx].set_vars(
            optdriver=4,
            nband=nband,
            ecutwfn=ecut,
            ecuteps=4.0,
            ecutsigx=6.0,
            symsigma=1,
            #gw_qprange=0,
            #nkptgw=0,
        )
        multi[3+idx].set_kptgw(kptgw, bdgw)

    return multi.split_datasets()


def build_flow(options):
    # Working directory (default is the name of the script with '.py' removed and "run_" replaced by "flow_")
    if not options.workdir:
        options.workdir = os.path.basename(sys.argv[0]).replace(".py", "").replace("run_","flow_")

    # Change the value of ngkpt below to perform a GW calculation with a different k-mesh.
    scf, nscf, scr, sig1, sig2, sig3 = make_inputs(ngkpt=[2, 2, 2])

    return flowtk.g0w0_flow(options.workdir, scf, nscf, scr, [sig1, sig2, sig3], manager=options.manager)


# This block generates the thumbnails in the AbiPy gallery.
# You can safely REMOVE this part if you are using this script for production runs.
if os.getenv("READTHEDOCS", False):
    __name__ = None
    import tempfile
    options = flowtk.build_flow_main_parser().parse_args(["-w", tempfile.mkdtemp()])
    build_flow(options).graphviz_imshow()


@flowtk.flow_main
def main(options):
    return build_flow(options)


if __name__ == "__main__":
    sys.exit(main())
run si g0w0

Run the script with:

run_si_g0w0.py -s

The last three tasks (w0_t3, w0_t4, w0_t5) are the SigmaTask who have produced a netcdf file with the GW results with different number of bands. We can check this with the command:

abirun.py flow_si_g0w0/ listext SIGRES

Found 3 files with extension `SIGRES` produced by the flow
File                                        Size [Mb]    Node_ID  Node Class
----------------------------------------  -----------  ---------  ------------
flow_si_g0w0/w0/t3/outdata/out_SIGRES.nc         0.05     241325  SigmaTask
flow_si_g0w0/w0/t4/outdata/out_SIGRES.nc         0.08     241326  SigmaTask
flow_si_g0w0/w0/t5/outdata/out_SIGRES.nc         0.13     241327  SigmaTask

Let’s use the SIGRES robot to collect and analyze the results:

abirun.py flow_si_g0w0/ robot SIGRES

and then, inside the ipython terminal, type:

In [1]: df = robot.get_dataframe()
In [2]: df
Out[2]:
                                          nsppol     qpgap            ecutwfn  \
flow_si_g0w0/w0/t3/outdata/out_SIGRES.nc       1  3.627960  5.914381651684836
flow_si_g0w0/w0/t4/outdata/out_SIGRES.nc       1  3.531781  5.914381651684836
flow_si_g0w0/w0/t5/outdata/out_SIGRES.nc       1  3.512285  5.914381651684836

                                                     ecuteps  \
flow_si_g0w0/w0/t3/outdata/out_SIGRES.nc  3.6964885323070074
flow_si_g0w0/w0/t4/outdata/out_SIGRES.nc  3.6964885323070074
flow_si_g0w0/w0/t5/outdata/out_SIGRES.nc  3.6964885323070074

                                                   ecutsigx scr_nband  \
flow_si_g0w0/w0/t3/outdata/out_SIGRES.nc  5.914381651684846        25
flow_si_g0w0/w0/t4/outdata/out_SIGRES.nc  5.914381651684846        25
flow_si_g0w0/w0/t5/outdata/out_SIGRES.nc  5.914381651684846        25

                                         sigma_nband gwcalctyp scissor_ene  \
flow_si_g0w0/w0/t3/outdata/out_SIGRES.nc          10         0         0.0
flow_si_g0w0/w0/t4/outdata/out_SIGRES.nc          20         0         0.0
flow_si_g0w0/w0/t5/outdata/out_SIGRES.nc          30         0         0.0

                                          nkibz
flow_si_g0w0/w0/t3/outdata/out_SIGRES.nc      6
flow_si_g0w0/w0/t4/outdata/out_SIGRES.nc      6
flow_si_g0w0/w0/t5/outdata/out_SIGRES.nc      6

In [3]: %matplotlib
In [4]: df.plot("sigma_nband", "qpgap", marker="o")
QP results in Si plotted vs the KS energy e0.

Total running time of the script: (0 minutes 0.275 seconds)

Gallery generated by Sphinx-Gallery