def parse_bands_file(bands_lines):
'''
Parses the returned bands.1 and bands.2 file and returns a complete
bandsData object. bands.1 has the form: k value, energy
:param bands_lines: string of the read in bands file
'''
# TODO: not finished
# read bands out of file:
nrows = 0 # get number of rows (known form number of atom types
bands_values = [] # init an array of arrays nkpoint * ...
bands_labels = [] # label for each row.
# fill and correct fermi energy.
bands_values = []
# TODO: we need to get the cell from StructureData node
# and KpointsData node from inpxml
fleur_bands = BandsData()
# fleur_bands.set_cell(cell)
#fleur_bands.set_kpoints(kpoints, cartesian=True)
fleur_bands.set_bands(bands=bands_values, units='eV', labels=bands_labels)
for line in bands_lines:
pass
return fleur_bands
def bands_to_bandsdata(bands_info, kpoints, bands):
"""
Convert the result of parser_dot_bands into a BandsData object
:param bands_info: A dictionary of the informations of the bands file.
contains field such as eferemi, units, cell
:param kpoints: An array of the kpoints of the bands, rows are
(kindex, kx, ky, kz, weight)
:param bands: The actual bands array
:return: A BandsData object
:rtype: ``aiida.orm.bands.data.array.bands.BandsData``
"""
bands_node = BandsData()
# Extract the index of the kpoints
kpn_array = np.asarray(kpoints)
k_index = kpn_array[:, 0]
# We need to restore the order of the kpoints
k_sort = np.argsort(k_index)
# Sort the kpn_array
kpn_array = kpn_array[k_sort]
_weights = kpn_array[:, -1]
kpts = kpn_array[:, 1:-1]
bands_node.set_kpoints(kpts, weights=_weights)
# Sort the bands to match the order of the kpoints
bands_array = np.asarray(bands)[k_sort]
# We need to swap the axes from kpt,spin,engs to spin,kpt,engs
bands_array = bands_array.swapaxes(0, 1)
# Squeeze the first dimension e.g when there is a single spin
if bands_array.shape[0] == 1:
bands_array = bands_array[0]
bands_array = bands_array * units['Eh']
bands_info = dict(bands_info) # Create a copy
# Convert the units for the fermi energies
if isinstance(bands_info['efermi'], list):
bands_info['efermi'] = [x * units['Eh'] for x in bands_info['efermi']]
else:
bands_info['efermi'] = bands_info['efermi'] * units['Eh']
bands_node.set_bands(bands_array, units="eV")
# PBC is always true as this is PW DFT....
bands_node.set_cell(bands_info['cell'], pbc=(True, True, True))
# Store information from *.bands in the attributes
# This is needs as we need to know the number of electrons
# and the fermi energy
for key, value in bands_info.items():
bands_node.set_attribute(key, value)
return bands_node
def test_bandsexport_single_kp(self):
"""
Plot band for single k-point (issue #2462).
"""
kpnts = KpointsData()
kpnts.set_kpoints([[0., 0., 0.]])
bands = BandsData()
bands.set_kpointsdata(kpnts)
bands.set_bands([[1.0, 2.0]])
bands.store()
# matplotlib
options = [str(bands.id), '--format', 'mpl_singlefile']
res = self.cli_runner.invoke(cmd_bands.bands_export,
options,
catch_exceptions=False)
self.assertIn(
b'p.scatter', res.stdout_bytes,
'The string p.scatter was not found in the bands mpl export')
# gnuplot
with self.cli_runner.isolated_filesystem():
options = [str(bands.id), '--format', 'gnuplot', '-o', 'bands.gnu']
self.cli_runner.invoke(cmd_bands.bands_export,
options,
catch_exceptions=False)
with open('bands.gnu', 'r') as gnu_file:
res = gnu_file.read()
self.assertIn(
'vectors nohead', res,
'The string "vectors nohead" was not found in the gnuplot script'
)
def _has_empty_bands(self, bands_data: BandsData, thresh=0.005):
"""
Check for the occupation of the BandsData
There should be some empty bands if the calculation is a not a fixed occupation one.
Otherwise, the final energy and forces are not reliable.
"""
# Check if occupation is allowed to vary
param = self.node.inputs.parameters.get_dict()['PARAM']
# If it is a fixed occupation calculation we do not need to do anything about it....
fix_occ = (param.get('fix_occupancy', False)
or param.get('metals_method', 'dm').lower() == 'none'
or param.get('elec_method', 'dm').lower() == 'none')
if fix_occ:
return True
_, occ = bands_data.get_bands(also_occupations=True)
nspin, nkppts, _ = occ.shape
problems = []
for ispin in range(nspin):
for ikpts in range(nkppts):
if occ[ispin, ikpts, -1] >= thresh:
problems.append((ispin, ikpts, occ[ispin, ikpts, -1]))
if problems:
for ispin, ikpts, val in problems:
self.logger.warning(
"No empty bands for spin %d, kpoint %d - occ: %.5f", ispin,
ikpts, val)
return False
return True
def _parse_stdout(self, out_folder):
"""CP2K output parser"""
from aiida.orm import BandsData, Dict
# pylint: disable=protected-access
fname = self.node.process_class._DEFAULT_OUTPUT_FILE
if fname not in out_folder._repository.list_object_names():
raise OutputParsingError("Cp2k output file not retrieved")
abs_fn = os.path.join(
out_folder._repository._get_base_folder().abspath, fname)
with io.open(abs_fn, mode="r", encoding="utf-8") as fobj:
result_dict = parse_cp2k_output(fobj)
if "nwarnings" not in result_dict:
raise OutputParsingError("CP2K did not finish properly.")
if "kpoint_data" in result_dict:
bnds = BandsData()
bnds.set_kpoints(result_dict["kpoint_data"]["kpoints"])
bnds.labels = result_dict["kpoint_data"]["labels"]
bnds.set_bands(
result_dict["kpoint_data"]["bands"],
units=result_dict["kpoint_data"]["bands_unit"],
)
self.out("output_bands", bnds)
del result_dict["kpoint_data"]
self.out("output_parameters", Dict(dict=result_dict))
def create_aiida_bands_data(fleurinp, retrieved):
"""
Creates :py:class:`aiida.orm.BandsData` object containing the kpoints and eigenvalues
from the `banddos.hdf` file of the calculation
:param fleurinp: :py:class:`~aiida_fleur.data.fleurinp.FleurinpData` for the calculation
:param retrieved: :py:class:`aiida.orm.FolderData` for the bandstructure calculation
:returns: :py:class:`aiida.orm.BandsData` for the bandstructure calculation
:raises: ExitCode 300, banddos.hdf file is missing
:raises: ExitCode 310, banddos.hdf reading failed
:raises: ExitCode 320, reading kpointsdata from Fleurinp failed
"""
from masci_tools.io.parsers.hdf5 import HDF5Reader, HDF5TransformationError
from masci_tools.io.parsers.hdf5.recipes import FleurSimpleBands #no projections only eigenvalues for now
from aiida.engine import ExitCode
try:
kpoints = fleurinp.get_kpointsdata_ncf(only_used=True)
except ValueError as exc:
return ExitCode(
320,
message=f'Retrieving kpoints data from fleurinp failed with: {exc}'
)
if 'banddos.hdf' in retrieved.list_object_names():
try:
with retrieved.open('banddos.hdf', 'rb') as f:
with HDF5Reader(f) as reader:
data, attributes = reader.read(recipe=FleurSimpleBands)
except (HDF5TransformationError, ValueError) as exc:
return ExitCode(310,
message=f'banddos.hdf reading failed with: {exc}')
else:
return ExitCode(300,
message='banddos.hdf file not in the retrieved files')
bands = BandsData()
bands.set_kpointsdata(kpoints)
nkpts, nbands = attributes['nkpts'], attributes['nbands']
eigenvalues = data['eigenvalues_up'].reshape((nkpts, nbands))
if 'eigenvalues_down' in data:
eigenvalues_dn = data['eigenvalues_down'].reshape((nkpts, nbands))
eigenvalues = [eigenvalues, eigenvalues_dn]
bands.set_bands(eigenvalues, units='eV')
bands.label = 'output_banddos_wc_bands'
bands.description = (
'Contains BandsData for the bandstructure calculation')
return bands
def _parse_stdout(self):
"""Advanced CP2K output file parser."""
from aiida.orm import BandsData
from aiida_cp2k.utils import parse_cp2k_output_advanced
fname = self.node.process_class._DEFAULT_OUTPUT_FILE # pylint: disable=protected-access
if fname not in self.retrieved.list_object_names():
raise OutputParsingError("Cp2k output file not retrieved")
try:
output_string = self.retrieved.get_object_content(fname)
except IOError:
return self.exit_codes.ERROR_OUTPUT_STDOUT_READ
result_dict = parse_cp2k_output_advanced(output_string)
# nwarnings is the last thing to be printed in th eCP2K output file:
# if it is not there, CP2K didn't finish properly
if 'nwarnings' not in result_dict:
raise OutputParsingError("CP2K did not finish properly.")
if "aborted" in result_dict:
return self.exit_codes.ERROR_OUTPUT_CONTAINS_ABORT
# Compute the bandgap for Spin1 and Spin2 if eigen was parsed (works also with smearing!)
if 'eigen_spin1_au' in result_dict:
if result_dict['dft_type'] == "RKS":
result_dict['eigen_spin2_au'] = result_dict['eigen_spin1_au']
lumo_spin1_idx = result_dict['init_nel_spin1']
lumo_spin2_idx = result_dict['init_nel_spin2']
if (lumo_spin1_idx > len(result_dict['eigen_spin1_au'])-1) or \
(lumo_spin2_idx > len(result_dict['eigen_spin2_au'])-1):
#electrons jumped from spin1 to spin2 (or opposite): assume last eigen is lumo
lumo_spin1_idx = len(result_dict['eigen_spin1_au']) - 1
lumo_spin2_idx = len(result_dict['eigen_spin2_au']) - 1
homo_spin1 = result_dict['eigen_spin1_au'][lumo_spin1_idx - 1]
homo_spin2 = result_dict['eigen_spin2_au'][lumo_spin2_idx - 1]
lumo_spin1 = result_dict['eigen_spin1_au'][lumo_spin1_idx]
lumo_spin2 = result_dict['eigen_spin2_au'][lumo_spin2_idx]
result_dict['bandgap_spin1_au'] = lumo_spin1 - homo_spin1
result_dict['bandgap_spin2_au'] = lumo_spin2 - homo_spin2
if "kpoint_data" in result_dict:
bnds = BandsData()
bnds.set_kpoints(result_dict["kpoint_data"]["kpoints"])
bnds.labels = result_dict["kpoint_data"]["labels"]
bnds.set_bands(
result_dict["kpoint_data"]["bands"],
units=result_dict["kpoint_data"]["bands_unit"],
)
self.out("output_bands", bnds)
del result_dict["kpoint_data"]
self.out("output_parameters", Dict(dict=result_dict))
return None
def test_valid_node():
"""Test that the correct exceptions are thrown for incompatible nodes."""
from aiida.orm import ArrayData, BandsData
# Invalid node type
node = ArrayData().store()
with pytest.raises(ValueError):
get_highest_occupied_band(node)
# The `occupations` array is missing
node = BandsData()
node.set_array('not_occupations', numpy.array([]))
node.store()
with pytest.raises(ValueError):
get_highest_occupied_band(node)
# The `occupations` array has incorrect shape
node = BandsData()
node.set_array('occupations', numpy.array([1., 1.]))
node.store()
with pytest.raises(ValueError):
get_highest_occupied_band(node)
def _parse_stdout(self, out_folder):
"""Advanced CP2K output file parser"""
from aiida.orm import BandsData
from .parser_functions import parse_cp2k_output_advanced
# pylint: disable=protected-access
fname = self.node.process_class._DEFAULT_OUTPUT_FILE
if fname not in out_folder._repository.list_object_names():
raise OutputParsingError("Cp2k output file not retrieved")
abs_fn = os.path.join(
out_folder._repository._get_base_folder().abspath, fname)
with io.open(abs_fn, mode="r", encoding="utf-8") as fobj:
result_dict = parse_cp2k_output_advanced(fobj)
# nwarnings is the last thing to be printed in th eCP2K output file:
# if it is not there, CP2K didn't finish properly
if 'nwarnings' not in result_dict:
raise OutputParsingError("CP2K did not finish properly.")
# Compute the bandgap for Spin1 and Spin2 if eigen was parsed (works also with smearing!)
if 'eigen_spin1_au' in result_dict:
if result_dict['dft_type'] == "RKS":
result_dict['eigen_spin2_au'] = result_dict['eigen_spin1_au']
lumo_spin1_idx = result_dict['init_nel_spin1']
lumo_spin2_idx = result_dict['init_nel_spin2']
if (lumo_spin1_idx > len(result_dict['eigen_spin1_au'])-1) or \
(lumo_spin2_idx > len(result_dict['eigen_spin2_au'])-1):
#electrons jumped from spin1 to spin2 (or opposite): assume last eigen is lumo
lumo_spin1_idx = len(result_dict['eigen_spin1_au']) - 1
lumo_spin2_idx = len(result_dict['eigen_spin2_au']) - 1
homo_spin1 = result_dict['eigen_spin1_au'][lumo_spin1_idx - 1]
homo_spin2 = result_dict['eigen_spin2_au'][lumo_spin2_idx - 1]
lumo_spin1 = result_dict['eigen_spin1_au'][lumo_spin1_idx]
lumo_spin2 = result_dict['eigen_spin2_au'][lumo_spin2_idx]
result_dict['bandgap_spin1_au'] = lumo_spin1 - homo_spin1
result_dict['bandgap_spin2_au'] = lumo_spin2 - homo_spin2
if "kpoint_data" in result_dict:
bnds = BandsData()
bnds.set_kpoints(result_dict["kpoint_data"]["kpoints"])
bnds.labels = result_dict["kpoint_data"]["labels"]
bnds.set_bands(
result_dict["kpoint_data"]["bands"],
units=result_dict["kpoint_data"]["bands_unit"],
)
self.out("output_bands", bnds)
del result_dict["kpoint_data"]
self.out("output_parameters", Dict(dict=result_dict))
def example_bands_v2(fresh_aiida_env):
"""Example eigen values and occupations"""
from aiida.orm import BandsData
bdata = BandsData()
bdata.set_kpoints([[0, 0, 0], [0.25, 0.25, 0.25]])
occ = np.array([[1, 1, 1, 0, 0, 0], [1, 1, 1, 0, 0, 0]], dtype=np.float64)
eigen = np.array([[-1, -1, -1, 0, 0, 0], [-0.5, -0.5, -0.5, 0, 0, 0]])
bdata.set_bands(bands=eigen, occupations=occ)
return bdata
def bands_from_castepbin(seedname, fmanager):
"""
Acquire and prepare bands data from the castep_bin file instead
"""
with fmanager.open(seedname + '.castep_bin', 'rb') as handle:
binfile = CastepbinFile(fileobj=handle)
bands_node = BandsData()
kidx = binfile.kpoints_indices
sort_idx = np.argsort(kidx)
# Generated sorted arrays
kpoints = binfile.kpoints[sort_idx, :]
weights = binfile.kpoint_weights[sort_idx].astype(float)
eigenvalues = binfile.eigenvalues[:, sort_idx, :]
occupancies = binfile.occupancies[:, sort_idx, :]
efermi = binfile.fermi_energy
bands_node.set_kpoints(kpoints, weights=weights)
bands_node.set_bands(eigenvalues, occupations=occupancies, units="eV")
bands_node.set_cell(binfile.cell, pbc=(True, True, True))
bands_node.set_attribute('efermi', efermi)
return bands_node
def connect_structure_bands(strct): # pylint: disable=unused-argument
alat = 4.
cell = np.array([
[alat, 0., 0.],
[0., alat, 0.],
[0., 0., alat],
])
kpnts = KpointsData()
kpnts.set_cell(cell)
kpnts.set_kpoints([[0., 0., 0.], [0.1, 0.1, 0.1]])
bands = BandsData()
bands.set_kpointsdata(kpnts)
bands.set_bands([[1.0, 2.0], [3.0, 4.0]])
return bands
def test_spin_unpolarized():
"""Test the function for a non spin-polarized calculation meaning there will be a single spin channel."""
from aiida.orm import BandsData
occupations = numpy.array([
[2., 2., 2., 2., 0.],
[2., 2., 2., 2., 0.],
[2., 2., 2., 2., 0.],
[2., 2., 2., 2., 0.],
])
bands = BandsData()
bands.set_array('occupations', occupations)
bands.store()
h**o = get_highest_occupied_band(bands)
assert h**o == 4
def connect_structure_bands(structure):
alat = 4.
cell = np.array([
[alat, 0., 0.],
[0., alat, 0.],
[0., 0., alat],
])
k = KpointsData()
k.set_cell(cell)
k.set_kpoints([[0., 0., 0.], [0.1, 0.1, 0.1]])
b = BandsData()
b.set_kpointsdata(k)
b.set_bands([[1.0, 2.0], [3.0, 4.0]])
return b
def test_spin_polarized(self, fixture_database):
"""Test the function for a spin-polarized calculation meaning there will be two spin channels."""
from aiida.orm import BandsData
occupations = numpy.array([
[
[2., 2., 2., 2., 0.],
[2., 2., 2., 2., 0.],
],
[
[2., 2., 2., 2., 0.],
[2., 2., 2., 2., 0.],
]
])
bands = BandsData()
bands.set_array('occupations', occupations)
bands.store()
h**o = get_highest_occupied_band(bands)
assert h**o == 4
def _parse_gsr(self):
"""Abinit GSR parser."""
# Output GSR Abinit NetCDF file - Default name is aiidao_GSR.nc
fname = f'{self.node.get_attribute("prefix")}o_GSR.nc'
# Absolute path of the folder in which aiidao_GSR.nc is stored
path = self.node.get_remote_workdir()
if fname not in self.retrieved.list_object_names():
return self.exit_codes.ERROR_MISSING_OUTPUT_FILES
with abilab.abiopen(path + '/' + fname) as gsr:
gsr_data = {
'abinit_version':
gsr.abinit_version,
'cart_stress_tensor':
gsr.cart_stress_tensor.tolist(),
'cart_stress_tensor' + UNITS_SUFFIX:
DEFAULT_STRESS_UNITS,
'is_scf_run':
bool(gsr.is_scf_run),
# 'cart_forces': gsr.cart_forces.tolist(),
# 'cart_forces' + units_suffix: DEFAULT_FORCE_UNITS,
'forces':
gsr.cart_forces.tolist(), # backwards compatibility
'forces' + UNITS_SUFFIX:
DEFAULT_FORCE_UNITS,
'energy':
float(gsr.energy),
'energy' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_localpsp':
float(gsr.energy_terms.e_localpsp),
'e_localpsp' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_eigenvalues':
float(gsr.energy_terms.e_eigenvalues),
'e_eigenvalues' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_ewald':
float(gsr.energy_terms.e_ewald),
'e_ewald' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_hartree':
float(gsr.energy_terms.e_hartree),
'e_hartree' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_corepsp':
float(gsr.energy_terms.e_corepsp),
'e_corepsp' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_corepspdc':
float(gsr.energy_terms.e_corepspdc),
'e_corepspdc' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_kinetic':
float(gsr.energy_terms.e_kinetic),
'e_kinetic' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_nonlocalpsp':
float(gsr.energy_terms.e_nonlocalpsp),
'e_nonlocalpsp' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_entropy':
float(gsr.energy_terms.e_entropy),
'e_entropy' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'entropy':
float(gsr.energy_terms.entropy),
'entropy' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_xc':
float(gsr.energy_terms.e_xc),
'e_xc' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_xcdc':
float(gsr.energy_terms.e_xcdc),
'e_xcdc' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_paw':
float(gsr.energy_terms.e_paw),
'e_paw' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_pawdc':
float(gsr.energy_terms.e_pawdc),
'e_pawdc' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_elecfield':
float(gsr.energy_terms.e_elecfield),
'e_elecfield' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_magfield':
float(gsr.energy_terms.e_magfield),
'e_magfield' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_fermie':
float(gsr.energy_terms.e_fermie),
'e_fermie' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_sicdc':
float(gsr.energy_terms.e_sicdc),
'e_sicdc' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_exactX':
float(gsr.energy_terms.e_exactX),
'e_exactX' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'h0':
float(gsr.energy_terms.h0),
'h0' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_electronpositron':
float(gsr.energy_terms.e_electronpositron),
'e_electronpositron' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'edc_electronpositron':
float(gsr.energy_terms.edc_electronpositron),
'edc_electronpositron' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e0_electronpositron':
float(gsr.energy_terms.e0_electronpositron),
'e0_electronpositron' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'e_monopole':
float(gsr.energy_terms.e_monopole),
'e_monopole' + UNITS_SUFFIX:
DEFAULT_ENERGY_UNITS,
'pressure':
float(gsr.pressure),
'pressure' + UNITS_SUFFIX:
DEFAULT_STRESS_UNITS
}
try:
# will return an integer 0 if non-magnetic calculation is run; convert it to a float
total_magnetization = float(gsr.ebands.get_collinear_mag())
gsr_data['total_magnetization'] = total_magnetization
gsr_data['total_magnetization' +
UNITS_SUFFIX] = DEFAULT_MAGNETIZATION_UNITS
except ValueError as valerr:
# get_collinear_mag will raise ValueError if it doesn't know what to do
if 'Cannot calculate collinear magnetization' in valerr.args[
0]:
pass
else:
raise valerr
try:
bands_data = BandsData()
bands_data.set_kpoints(gsr.ebands.kpoints.get_cart_coords())
bands_data.set_bands(np.array(gsr.ebands.eigens),
units=str(gsr.ebands.eigens.unit))
self.out('output_bands', bands_data)
except: # pylint: disable=bare-except
pass
self.out('output_parameters', Dict(dict=gsr_data))
def create_structure_bands():
"""Create bands structure object."""
alat = 4. # angstrom
cell = [
[
alat,
0.,
0.,
],
[
0.,
alat,
0.,
],
[
0.,
0.,
alat,
],
]
strct = StructureData(cell=cell)
strct.append_atom(position=(0., 0., 0.), symbols='Fe')
strct.append_atom(position=(alat / 2., alat / 2., alat / 2.),
symbols='O')
strct.store()
@calcfunction
def connect_structure_bands(strct): # pylint: disable=unused-argument
alat = 4.
cell = np.array([
[alat, 0., 0.],
[0., alat, 0.],
[0., 0., alat],
])
kpnts = KpointsData()
kpnts.set_cell(cell)
kpnts.set_kpoints([[0., 0., 0.], [0.1, 0.1, 0.1]])
bands = BandsData()
bands.set_kpointsdata(kpnts)
bands.set_bands([[1.0, 2.0], [3.0, 4.0]])
return bands
bands = connect_structure_bands(strct)
bands_isolated = BandsData()
bands_isolated.store()
# Create 2 groups and add the data to one of them
g_ne = Group(label='non_empty_group')
g_ne.store()
g_ne.add_nodes(bands)
g_ne.add_nodes(bands_isolated)
g_e = Group(label='empty_group')
g_e.store()
return {
DummyVerdiDataListable.NODE_ID_STR: bands.id,
DummyVerdiDataListable.NON_EMPTY_GROUP_ID_STR: g_ne.id,
DummyVerdiDataListable.EMPTY_GROUP_ID_STR: g_e.id
}
def test_threshold():
"""Test the `threshold` parameter."""
from aiida.orm import BandsData
threshold = 0.002
bands = BandsData()
bands.set_array('occupations',
numpy.array([[2., 2., 2., 2., 0.001, 0.0015]]))
bands.store()
# All bands above the LUMO (occupation of 0.001) are below `2 * threshold`
h**o = get_highest_occupied_band(bands, threshold=threshold)
assert h**o == 4
bands = BandsData()
bands.set_array('occupations',
numpy.array([[2., 2., 2., 2., 0.001, 0.003]]))
bands.store()
# A band above the LUMO (occupation of 0.001) has an occupation above `2 * threshold`
with pytest.raises(ValueError):
get_highest_occupied_band(bands, threshold=threshold)
def band_parser_legacy(band_dat, band_kpt, special_points, structure): # pylint: disable=too-many-locals
"""
Parsers the bands output data, along with the special points retrieved
from the input kpoints to construct a BandsData object which is then
returned. Cannot handle discontinuities in the kpath, if two points are
assigned to same spot only one will be passed. Used for wannier90 < 3.0
:param band_dat: list of str with each str stores one line of aiida_band.dat file
:param band_kpt: list of str with each str stores one line of aiida_band.kpt file
:param special_points: special points to add labels to the bands a dictionary in
the form expected in the input as described in the wannier90 documentation
:return: BandsData object constructed from the input params,
and a list contains warnings.
"""
import numpy as np
from aiida.orm import BandsData
from aiida.orm import KpointsData
warnings = []
warnings.append((
"Note: no file named SEEDNAME_band.labelinfo.dat found. "
"You are probably using a version of Wannier90 before 3.0. "
"There, the labels associated with each k-points were not printed in output "
"and there were also cases in which points were not calculated "
"(see issue #195 on the Wannier90 GitHub page). "
"I will anyway try to do my best to assign labels, "
"but the assignment might be wrong "
"(especially if there are path discontinuities)."))
# imports the data
out_kpt = np.genfromtxt(band_kpt, skip_header=1, usecols=(0, 1, 2))
out_dat = np.genfromtxt(band_dat, usecols=1)
# reshaps the output bands
out_dat = out_dat.reshape(len(out_kpt), (len(out_dat) // len(out_kpt)),
order="F")
# finds expected points of discontinuity
kpath = special_points['path']
cont_break = [(i, (kpath[i - 1][1], kpath[i][0]))
for i in range(1, len(kpath))
if kpath[i - 1][1] != kpath[i][0]]
# finds the special points
special_points_dict = special_points['point_coords']
# We set atol to 1e-5 because in the kpt file the coords are printed with fixed precision
labels = [
(i, k) for k in special_points_dict for i in range(len(out_kpt))
if all(
np.isclose(special_points_dict[k], out_kpt[i], rtol=0, atol=1.e-5))
]
labels.sort()
# Checks and appends labels if discontinuity
appends = []
for x in cont_break:
# two cases the break is before or the break is after
# if the break is before
if labels[x[0]][1] != x[1][0]:
# checks to see if the discontinuity was already there
if labels[x[0] - 1] == x[1][0]:
continue
insert_point = x[0]
new_label = x[1][0]
kpoint = labels[x[0]][0] - 1
appends += [[insert_point, new_label, kpoint]]
# if the break is after
if labels[x[0]][1] != x[1][1]:
# checks to see if the discontinuity was already there
if labels[x[0] + 1] == x[1][1]:
continue
insert_point = x[0] + 1
new_label = x[1][1]
kpoint = labels[x[0]][0] + 1
appends += [[insert_point, new_label, kpoint]]
appends.sort()
for i, append in enumerate(appends):
labels.insert(append[0] + i, (append[2], six.text_type(append[1])))
bands = BandsData()
k = KpointsData()
k.set_cell_from_structure(structure)
k.set_kpoints(out_kpt, cartesian=False)
bands.set_kpointsdata(k)
bands.set_bands(out_dat, units='eV')
bands.labels = labels
return bands, warnings
def spin_dependent_subparser(out_info_dict):
"""This find the projection and bands arrays from the out_file and out_info_dict. Used to handle the different
possible spin-cases in a convenient manner.
:param out_info_dict: contains various technical internals useful in parsing
:return: ProjectionData, BandsData parsed from out_file
"""
out_file = out_info_dict['out_file']
spin_down = out_info_dict['spin_down']
od = out_info_dict #using a shorter name for convenience
# regular expressions needed for later parsing
WaveFraction1_re = re.compile(r'\=(.*?)\*') # state composition 1
WaveFractionremain_re = re.compile(r'\+(.*?)\*') # state comp 2
FunctionId_re = re.compile(r'\#(.*?)\]') # state identity
# primes arrays for the later parsing
num_wfc = len(od['wfc_lines'])
bands = np.zeros([od['k_states'], od['num_bands']])
projection_arrays = np.zeros([od['k_states'], od['num_bands'], num_wfc])
try:
for i in range(od['k_states']):
if spin_down:
i += od['k_states']
# grabs band energy
for j in range(i * od['num_bands'], (i + 1) * od['num_bands'], 1):
out_ind = od['e_lines'][j]
try:
# post ~6.3 output format "e ="
val = out_file[out_ind].split()[2]
float(val)
except ValueError:
# pre ~6.3 output format? "==== e("
val = out_file[out_ind].split()[4]
bands[i % od['k_states']][j % od['num_bands']] = val
#subloop grabs pdos
wave_fraction = []
wave_id = []
for k in range(od['e_lines'][j] + 1, od['psi_lines'][j], 1):
out_line = out_file[k]
wave_fraction += WaveFraction1_re.findall(out_line)
wave_fraction += WaveFractionremain_re.findall(out_line)
wave_id += FunctionId_re.findall(out_line)
if len(wave_id) != len(wave_fraction):
raise IndexError
for l in range(len(wave_id)):
wave_id[l] = int(wave_id[l])
wave_fraction[l] = float(wave_fraction[l])
#sets relevant values in pdos_array
projection_arrays[i % od['k_states']][j % od['num_bands']][
wave_id[l] - 1] = wave_fraction[l]
except IndexError:
raise QEOutputParsingError(
'the standard out file does not comply with the official documentation.'
)
bands_data = BandsData()
# Attempts to retrieve the kpoints from the parent calc
parent_calc = out_info_dict['parent_calc']
try:
parent_kpoints = parent_calc.get_incoming(
link_label_filter='kpoints').one().node
except ValueError:
raise QEOutputParsingError(
'The parent had no input kpoints! Cannot parse from this!')
try:
if len(od['k_vect']) != len(parent_kpoints.get_kpoints()):
raise AttributeError
bands_data.set_kpointsdata(parent_kpoints)
except AttributeError:
bands_data.set_kpoints(od['k_vect'].astype(float))
bands_data.set_bands(bands, units='eV')
orbitals = out_info_dict['orbitals']
if len(orbitals) != np.shape(projection_arrays[0, 0, :])[0]:
raise QEOutputParsingError('There was an internal parsing error, '
' the projection array shape does not agree'
' with the number of orbitals')
projection_data = ProjectionData()
projection_data.set_reference_bandsdata(bands_data)
projections = [projection_arrays[:, :, i] for i in range(len(orbitals))]
# Do the bands_check manually here
for projection in projections:
if np.shape(projection) != np.shape(bands):
raise AttributeError('Projections not the same shape as the bands')
#insert here some logic to assign pdos to the orbitals
pdos_arrays = spin_dependent_pdos_subparser(out_info_dict)
energy_arrays = [out_info_dict['energy']] * len(orbitals)
projection_data.set_projectiondata(orbitals,
list_of_projections=projections,
list_of_energy=energy_arrays,
list_of_pdos=pdos_arrays,
bands_check=False)
# pdos=pdos_arrays
return bands_data, projection_data
def parse(self, **kwargs): # noqa: MC0001 - is mccabe too complex funct -
"""
Receives in input a dictionary of retrieved nodes. Does all the logic here.
"""
from aiida.engine import ExitCode
parser_info = {}
parser_info['parser_info'] = 'AiiDA Siesta Parser V. {}'.format(
self._version)
try:
output_folder = self.retrieved
except exceptions.NotExistent:
raise OutputParsingError("Folder not retrieved")
output_path, messages_path, xml_path, json_path, bands_path, basis_enthalpy_path = \
self._fetch_output_files(output_folder)
if xml_path is None:
raise OutputParsingError("Xml file not retrieved")
xmldoc = get_parsed_xml_doc(xml_path)
result_dict = get_dict_from_xml_doc(xmldoc)
if output_path is None:
raise OutputParsingError("output file not retrieved")
output_dict = dict(
list(result_dict.items()) + list(parser_info.items()))
warnings_list = []
if json_path is not None:
from .json_time import get_timing_info
global_time, timing_decomp = get_timing_info(json_path)
if global_time is None:
warnings_list.append(["Cannot fully parse the time.json file"])
else:
output_dict["global_time"] = global_time
output_dict["timing_decomposition"] = timing_decomp
if basis_enthalpy_path is not None:
the_file = open(basis_enthalpy_path)
bas_enthalpy = float(the_file.read().split()[0])
the_file.close()
output_dict["basis_enthalpy"] = bas_enthalpy
output_dict["basis_enthalpy_units"] = "eV"
else:
warnings_list.append(["BASIS_ENTHALPY file not retrieved"])
have_errors_to_analyse = False
if messages_path is None:
# Perhaps using an old version of Siesta
warnings_list.append([
'WARNING: No MESSAGES file, could not check if calculation terminated correctly'
])
else:
have_errors_to_analyse = True
#succesful when "INFO: Job completed" is present in message files
succesful, from_message = self._get_warnings_from_file(
messages_path)
warnings_list.append(from_message)
output_dict["warnings"] = warnings_list
# An output_parametrs port is always return, even if only parser's info are present
output_data = Dict(dict=output_dict)
self.out('output_parameters', output_data)
#
# When using floating sites, Siesta associates 'atomic positions' to them, and
# the structure and forces in the XML file include these fake atoms.
# In order to return physical structures and forces, we need to remove them.
# Recall that the input structure is the physical one, and the floating sites
# are specified in the 'basis' input
#
physical_structure = self.node.inputs.structure
number_of_real_atoms = len(physical_structure.sites)
# If the structure has changed, save it
#
if output_dict['variable_geometry']:
in_struc = self.node.inputs.structure
# The next function never fails. If problems arise, the initial structure is
# returned. The input structure is also necessary because the CML file
# traditionally contains only the atomic symbols and not the site names.
# The returned structure does not have any floating atoms, they are removed
# in the `get_last_structure` call.
success, out_struc = get_last_structure(xmldoc, in_struc)
if not success:
self.logger.warning(
"Problem in parsing final structure, returning inp structure in output_structure"
)
self.out('output_structure', out_struc)
# Attempt to parse forces and stresses. In case of failure "None" is returned.
# Therefore the function never crashes
forces, stress = get_final_forces_and_stress(xmldoc)
if forces is not None and stress is not None:
from aiida.orm import ArrayData
arraydata = ArrayData()
arraydata.set_array('forces',
np.array(forces[0:number_of_real_atoms]))
arraydata.set_array('stress', np.array(stress))
self.out('forces_and_stress', arraydata)
#Attempt to parse the ion files. Files ".ion.xml" are not produced by siesta if ions file are used
#in input (`user-basis = T`). This explains the first "if" statement. The SiestaCal input is called
#`ions__El` (El is the element label) therefore we look for the str "ions" in any of the inputs name.
if not any(["ions" in inp for inp in self.node.inputs]): #pylint: disable=too-many-nested-blocks
from aiida_siesta.data.ion import IonData
ions = {}
#Ions from the structure
in_struc = self.node.inputs.structure
for kind in in_struc.get_kind_names():
ion_file_name = kind + ".ion.xml"
if ion_file_name in output_folder._repository.list_object_names(
):
ion_file_path = os.path.join(
output_folder._repository._get_base_folder().abspath,
ion_file_name)
ions[kind] = IonData(ion_file_path)
else:
self.logger.warning(f"no ion file retrieved for {kind}")
#Ions from floating_sites
if "basis" in self.node.inputs:
basis_dict = self.node.inputs.basis.get_dict()
if "floating_sites" in basis_dict:
floating_kinds = []
for orb in basis_dict["floating_sites"]:
if orb["name"] not in floating_kinds:
floating_kinds.append(orb["name"])
ion_file_name = orb["name"] + ".ion.xml"
if ion_file_name in output_folder._repository.list_object_names(
):
ion_file_path = os.path.join(
output_folder._repository._get_base_folder(
).abspath, ion_file_name)
ions[orb["name"]] = IonData(ion_file_path)
else:
self.logger.warning(
f"no ion file retrieved for {orb['name']}")
#Return the outputs
if ions:
self.out('ion_files', ions)
# Error analysis
if have_errors_to_analyse:
# No metter if "INFO: Job completed" is present (succesfull) or not, we check for known
# errors. They might apprear as WARNING (therefore with succesful True) or FATAL
# (succesful False)
for line in from_message:
if u'split options' in line:
min_split = get_min_split(output_path)
if min_split:
self.logger.error(
"Error in split_norm option. Minimum value is {}".
format(min_split))
return self.exit_codes.SPLIT_NORM
if u'sys::die' in line:
#This is the situation when siesta dies with no specified error
#to be reported in "MESSAGES", unfortunately some interesting cases
#are treated in this way, we explore the .out file for more insights.
if is_polarization_problem(output_path):
return self.exit_codes.BASIS_POLARIZ
if u'SCF_NOT_CONV' in line:
return self.exit_codes.SCF_NOT_CONV
if u'GEOM_NOT_CONV' in line:
return self.exit_codes.GEOM_NOT_CONV
#Because no known error has been found, attempt to parse bands if requested
if bands_path is None:
if "bandskpoints" in self.node.inputs:
return self.exit_codes.BANDS_FILE_NOT_PRODUCED
else:
#bands, coords = self._get_bands(bands_path)
try:
bands = self._get_bands(bands_path)
except (ValueError, IndexError):
return self.exit_codes.BANDS_PARSE_FAIL
from aiida.orm import BandsData
arraybands = BandsData()
#Reset the cell for KpointsData of bands, necessary
#for bandskpoints without cell and if structure changed
bkp = self.node.inputs.bandskpoints.clone()
if output_dict['variable_geometry']:
bkp.set_cell_from_structure(out_struc)
else:
bkp.set_cell_from_structure(self.node.inputs.structure)
arraybands.set_kpointsdata(bkp)
arraybands.set_bands(bands, units="eV")
self.out('bands', arraybands)
#bandsparameters = Dict(dict={"kp_coordinates": coords})
#self.out('bands_parameters', bandsparameters)
#At the very end, return a particular exit code if "INFO: Job completed"
#was not present in the MESSAGES file, but no known error is detected.
if have_errors_to_analyse:
if not succesful:
self.logger.error(
'The calculation finished without "INFO: Job completed", but no '
'error could be processed. Might be that the calculation was killed externally'
)
return self.exit_codes.UNEXPECTED_TERMINATION
return ExitCode(0)
def _get_output_nodes(self, output_path, messages_path, xml_path,
json_path, bands_path):
"""
Extracts output nodes from the standard output and standard error
files. (And XML and JSON files)
"""
from aiida.orm import TrajectoryData
import re
parser_version = '1.0.1'
parser_info = {}
parser_info['parser_info'] = 'AiiDA Siesta Parser V. {}'.format(
parser_version)
parser_info['parser_warnings'] = []
#No need for checks anymore
xmldoc = get_parsed_xml_doc(xml_path)
in_struc = self.node.inputs.structure
try:
in_settings = self.node.inputs.settings
except exceptions.NotExistent:
in_settings = None
result_dict = get_dict_from_xml_doc(xmldoc)
# Add timing information
#if json_path is None:
###TODO###
#Not sure how to implement what once was logger.info
#Also not sure I understood the purpose of this file
if json_path is not None:
from .json_time import get_timing_info
global_time, timing_decomp = get_timing_info(json_path)
if global_time is None:
raise OutputParsingError(
"Cannot fully parse the time.json file")
else:
result_dict["global_time"] = global_time
result_dict["timing_decomposition"] = timing_decomp
# Add warnings
if messages_path is None:
# Perhaps using an old version of Siesta
warnings_list = ['WARNING: No MESSAGES file...']
else:
successful, warnings_list = self.get_warnings_from_file(
messages_path)
###TODO### or maybe not
#How do we process this successfull booelan?
result_dict["warnings"] = warnings_list
# Add parser info dictionary
parsed_dict = dict(
list(result_dict.items()) + list(parser_info.items()))
output_data = Dict(dict=parsed_dict)
self.out('output_parameters', output_data)
# If the structure has changed, save it
if is_variable_geometry(xmldoc):
# Get the input structure to copy its site names,
# as the CML file traditionally contained only the
# atomic symbols.
#
struc = get_last_structure(xmldoc, in_struc)
# result_list.append((self.get_linkname_outstructure(),struc))
self.out(self.get_linkname_outstructure(), struc)
# Save forces and stress in an ArrayData object
forces, stress = get_final_forces_and_stress(xmldoc)
if forces is not None and stress is not None:
# from aiida.orm.nodes.array import ArrayData
from aiida.orm import ArrayData
arraydata = ArrayData()
arraydata.set_array('forces', np.array(forces))
arraydata.set_array('stress', np.array(stress))
# result_list.append((self.get_linkname_fs_and_stress(),arraydata))
self.out(self.get_linkname_fs_and_stress(), arraydata)
# Parse band-structure information if available
if bands_path is not None:
bands, coords = self.get_bands(bands_path)
from aiida.orm import BandsData
arraybands = BandsData()
f = self.node.inputs.bandskpoints #Temporary workaround due to a bug
#f._set_reciprocal_cell() #in KpointData (issue #2749)
arraybands.set_kpoints(f.get_kpoints(cartesian=True))
arraybands.labels = f.labels
arraybands.set_bands(bands, units="eV")
self.out(self.get_linkname_bands(), arraybands)
bandsparameters = Dict(dict={"kp_coordinates": coords})
self.out(self.get_linkname_bandsparameters(), bandsparameters)
return result_dict
def band_parser(band_dat, band_kpt, band_labelinfo, structure): # pylint: disable=too-many-locals
"""
Parsers the bands output data to construct a BandsData object which is then
returned. Used for wannier90 >= 3.0
:param band_dat: list of str with each str stores one line of aiida_band.dat file
:param band_kpt: list of str with each str stores one line of aiida_band.kpt file
:param band_labelinfo: list of str with each str stores one line in aiida_band.labelinfo.dat file
:return: BandsData object constructed from the input params
"""
import numpy as np
from aiida.orm import BandsData
from aiida.orm import KpointsData
warnings = []
# imports the data
out_kpt = np.genfromtxt(band_kpt, skip_header=1, usecols=(0, 1, 2))
out_dat = np.genfromtxt(band_dat, usecols=1)
# reshaps the output bands
out_dat = out_dat.reshape(len(out_kpt), (len(out_dat) // len(out_kpt)),
order="F")
labels_dict = {}
for line_idx, line in enumerate(band_labelinfo, start=1):
if not line.strip():
continue
try:
# label, idx, xval, kx, ky, kz = line.split()
label, idx, _, _, _, _ = line.split()
except ValueError:
warnings.append(
('Wrong number of items in line {} of the labelinfo file - '
'I will not assign that label')).format(line_idx)
continue
try:
idx = int(idx)
except ValueError:
warnings.append((
"Invalid value for the index in line {} of the labelinfo file, "
"it's not an integer - I will not assign that label"
)).format(line_idx)
continue
# I use a dictionary because there are cases in which there are
# two lines for the same point (e.g. when I do a zero-length path,
# from a point to the same point, just to have that value)
# Note the -1 because in fortran indices are 1-based, in Python are
# 0-based
labels_dict[idx - 1] = label
labels = [(key, labels_dict[key]) for key in sorted(labels_dict)]
bands = BandsData()
k = KpointsData()
k.set_cell_from_structure(structure)
k.set_kpoints(out_kpt, cartesian=False)
bands.set_kpointsdata(k)
bands.set_bands(out_dat, units='eV')
bands.labels = labels
return bands, warnings
def _aiida_bands_data(self, data, cell, kpoints_dict):
if not data:
return False
kpt_idx = sorted(data.keys()) # list of kpoint indices
try:
k_list = [kpoints_dict[i]
for i in kpt_idx] # list of k-point triplet
except KeyError:
# kpoint triplets are not present (true for .qp and so on, can not use BandsData)
# We use the internal Yambo Format [ [Eo_1, Eo_2,... ], ...[So_1,So_2,] ]
# QP_TABLE [[ib_1,ik_1,isp_1] ,[ib_n,ik_n,isp_n]]
# Each entry in DATA has corresponding legend in QP_TABLE that defines its details
# like ib= Band index, ik= kpoint index, isp= spin polarization index.
# Eo_1 => at ib_1, ik_1 isp_1.
pdata = ArrayData()
QP_TABLE = []
ORD = []
Eo = []
E_minus_Eo = []
So = []
Z = []
for ky in data.keys(): # kp == kpoint index as a string 1,2,..
for ind in range(len(data[ky]['Band'])):
try:
Eo.append(data[ky]['Eo'][ind])
except KeyError:
pass
try:
E_minus_Eo.append(data[ky]['E-Eo'][ind])
except KeyError:
pass
try:
So.append(data[ky]['Sc|Eo'][ind])
except KeyError:
pass
try:
Z.append(data[ky]['Z'][ind])
except KeyError:
pass
ik = int(ky)
ib = data[ky]['Band'][ind]
isp = 0
if 'Spin_Pol' in list(data[ky].keys()):
isp = data[ky]['Spin_Pol'][ind]
QP_TABLE.append([ik, ib, isp])
pdata.set_array('Eo', numpy.array(Eo))
pdata.set_array('E_minus_Eo', numpy.array(E_minus_Eo))
pdata.set_array('So', numpy.array(So))
pdata.set_array('Z', numpy.array(Z))
pdata.set_array('qp_table', numpy.array(QP_TABLE))
return pdata
quasiparticle_bands = BandsData()
quasiparticle_bands.set_cell(cell)
quasiparticle_bands.set_kpoints(k_list, cartesian=True)
# labels will come from any of the keys in the nested kp_point data,
# there is a uniform set of observables for each k-point, ie Band, Eo, ...
# ***FIXME BUG does not seem to handle spin polarizes at all when constructing bandsdata***
bands_labels = [
legend for legend in sorted(data[list(data.keys())[0]].keys())
]
append_list = [[] for i in bands_labels]
for kp in kpt_idx:
for i in range(len(bands_labels)):
append_list[i].append(data[kp][bands_labels[i]])
generalised_bands = [numpy.array(it) for it in append_list]
quasiparticle_bands.set_bands(bands=generalised_bands,
units='eV',
labels=bands_labels)
return quasiparticle_bands
