def conformer_to_stat_values(conformer):
"""Beam transform to produce stats values for later aggregation.
Each output will be a tuple of primary_key, secondary_key and these will be
aggregated as counts.
Args:
conformer: dataset_pb2.Conformer
Yields:
primary_key, secondary_key
"""
# Yield the values for all the relevant error fields.
for field in [
'status', 'warn_t1', 'warn_t1_excess', 'warn_bse_b5_b6',
'warn_bse_cccsd_b5', 'warn_exc_lowest_excitation',
'warn_exc_smallest_oscillator', 'warn_exc_largest_oscillator',
'warn_vib_linearity', 'warn_vib_imaginary', 'warn_num_neg',
'error_nstat1', 'error_nstatc', 'error_nstatt', 'error_frequencies'
]:
yield 'errors.' + field, getattr(conformer.properties.errors, field)
yield 'fate', dataset_pb2.Conformer.FateCategory.Name(conformer.fate)
yield 'num_initial_geometries', len(conformer.initial_geometries)
yield 'num_duplicates', len(conformer.duplicate_of)
if not conformer.duplicated_by:
yield 'num_topologies', len(conformer.bond_topologies)
for field in smu_utils_lib.find_zero_values(conformer):
yield 'zero_field', field
def molecule_to_stat_values(molecule):
"""Beam transform to produce stats values for later aggregation.
Each output will be a tuple of primary_key, secondary_key and these will be
aggregated as counts.
Args:
molecule: dataset_pb2.Molecule
Yields:
primary_key, secondary_key
"""
# Yield the values for all the relevant error fields.
for field in [
'status', 'warn_t1', 'warn_t1_excess', 'warn_bse_b5_b6',
'warn_bse_cccsd_b5', 'warn_exc_lowest_excitation',
'warn_exc_smallest_oscillator', 'warn_exc_largest_oscillator',
'warn_vib_linearity', 'warn_vib_imaginary', 'warn_num_neg',
'error_nstat1', 'error_nstatc', 'error_nstatt', 'error_frequencies'
]:
yield 'errors.' + field, getattr(molecule.properties.errors, field)
yield 'fate', dataset_pb2.Properties.FateCategory.Name(
molecule.properties.errors.fate)
yield 'num_initial_geometries', len(
[g for g in molecule.initial_geometries if g.atom_positions])
yield 'num_duplicates', len(molecule.duplicate_of)
for field in smu_utils_lib.find_zero_values(molecule):
yield 'zero_field', field
if not molecule.duplicated_by and molecule.properties.errors.status < 512:
yield 'num_topologies', len(molecule.bond_topologies)
yield 'num_topologies_itc', len([
None for bt in molecule.bond_topologies
if bt.source & dataset_pb2.BondTopology.SOURCE_ITC
])
yield 'num_topologies_mlcr', len([
None for bt in molecule.bond_topologies
if bt.source & dataset_pb2.BondTopology.SOURCE_MLCR
])
yield 'num_topologies_csd', len([
None for bt in molecule.bond_topologies
if bt.source & dataset_pb2.BondTopology.SOURCE_CSD
])
for bt in molecule.bond_topologies:
yield 'bt_source', bt.source
def test_excitation(self): conformer = get_stage2_conformer() conformer.properties.excitation_energies_cc2.value[2] = 0.0 got = list(smu_utils_lib.find_zero_values(conformer)) self.assertEqual(got, ['excitation_energies_cc2'])
def test_scalar(self): conformer = get_stage2_conformer() conformer.properties.lumo_b3lyp_6_31ppgdp.value = 0.0 got = list(smu_utils_lib.find_zero_values(conformer)) self.assertEqual(got, ['lumo_b3lyp_6_31ppgdp'])
def test_no_zeroes(self): conformer = get_stage2_conformer() got = list(smu_utils_lib.find_zero_values(conformer)) self.assertEqual(got, [])
def test_atomic(self): conformer = get_stage2_conformer() conformer.properties.partial_charges_esp_fit_hf_6_31gd.values[3] = 0.0 got = list(smu_utils_lib.find_zero_values(conformer)) self.assertEqual(got, ['partial_charges_esp_fit_hf_6_31gd'])